Citation
Sleep apnea activity and its concomitants in a subclinical population

Material Information

Title:
Sleep apnea activity and its concomitants in a subclinical population
Creator:
Berry, David Thomas Reed, 1958-
Publication Date:
Language:
English
Physical Description:
viii, 132 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Apnea ( jstor )
Chemical desaturation ( jstor )
Hypoxia ( jstor )
Index numbers ( jstor )
Memory ( jstor )
Oxygen ( jstor )
Rapid eye movement sleep ( jstor )
Sleep ( jstor )
Sleep apnea syndromes ( jstor )
Snoring ( jstor )
Apnea ( mesh )
Clinical and Health Psychology thesis Ph.D ( mesh )
Dissertations, Academic -- Clinical and Health Psychology -- UF ( mesh )
Respiration ( mesh )
Sleep Apnea Syndromes ( mesh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1985.
Bibliography:
Bibliography: leaves 123-131.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by David Thomas Reed Berry.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
000522493 ( ALEPH )
16877989 ( OCLC )
ACU4129 ( NOTIS )

Downloads

This item has the following downloads:


Full Text


















SLEEP APNEA ACTIVITY AND ITS CONCOMITANTS
IN A SUBCLINICAL POPULATION




By

DAVID THOMAS REED BERRY





















A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY


UNIVERSITY OF FLCOIDA

1985
















ACKEOWLEDGCI I:TS

The author would like to thank his chairman, Dr. Wilse

B. V.bt, for his moral and material support, as well as

guidance through the completion of the project. Also due

thanks are the other members of the supervisory committee,

Dr. A. J. Block, Dr. Russell M. Eauer, Dr. F. D. McGlynn,

ann Dr. Rudy Vuchinich, for their support, advice, and

contributions. Several people provided technical assistance

and are due a vote of thanks: Dan Switzer, Alex Kasterton,

Karla Smith, Alton Howard, Albert Briggs, and Edna Pearson.

The author wishes to acknowledge the moral and

financial support of his parents, Mr. and Mrs. John G.

Berry, without which the project would never have been

undertaken. Also of considerable help were the

encouragement and advice of C. i. Carswell, D. Haymaker, anc

S. Erody, all of whom deserve thanks. Cynthia Zimmerman

provided expert typing assistance for the manuscript.















TALE OF CONTENTS


Page

ACKNOWLEDG ENTS ....... .............. ..................... ii

LIST OF TABLES ........................................... v

ABSTRACT ...............................................vii

CHAPTER

C!E IITRODUCTION ..................................... 1

Definitions and Terminology......................2
History and Characteristics of Sleep Apnea
Syndrome ......................................
Incidence of Sleep Apnea Activity.................5
Temporal Characteristics of Apneas...............13
Variables Associated With Sleep Apnea
Syndrome and Activity.........................17
Deficits Associated With Sleep Apnea
Syndrome.....................................25
Cardiopulmonary Deficits.....................26
Arousal Deficits (Hypersomnolence)..........28
Oxygen-Deseturation and Cognitive Deficits..30
Measurement of Deficits Found in Sleep Apnea
Syndrome.......................................34
Measurement of Cardiopulmonary and Health
Deficits .................................. 34
Measurement of Arousal Deficits
(Hypersomnolence)........................35
Measurement of Hypoxia and Its Sequelae.....38
Statement of the Problem ........................46

TWO METHOD...................................... ... 48

Subjects ........................................48
Apparatus.......................................49
Measures........................................50
Procedure .. ................... ............... 54
Statistical Procedures..........................55

THREE RESULTS................. .... ................... 56

Demographics and Incidence of Nocturnal
Respiratory Disorder..........................56


iii









Nocturnal Respiratory and Health Variables......63
Nocturnal Respiratory and Sleep/Wake Variables..65
EEG Data.................................... 65
Daytime Sleepiness Data.....................69
Subjective Sleep Assessment.................70
Nocturnal Respiratory and Neuropsychological
Variables..................................... 74

FOUR DISCUSSION.....................................86

Demographics and Incidence of Ilocturnal
Respiratory Disorder..........................86
nocturnal Respiratory and Health Variables......90
l!octurnal Respiratory and Sleep/Wake Variables..92
EEG Data ....................................92
Daytime Sleepiness, Sleep Questionnaire,
and Sleep Log Data........................94
Nocturnal Respiratory and Neuropsychological
Variables.....................................96

FIVE SUMIIARY AND CONCLUSIONS........................103

APPENDIX

A SCORING PROTOCOL FOR RESPIRATORY VARIABLES..... 106

B RAk DATA TABLES................................109

REFERErCES...............................................123

EIOGRAPHICAL SKETCH.....................................132















LIST OF TABLES

Table Page

2-1 Interrater agreement for respiratory events
of 46 snoring males..............................51

3-1 Demographic variables from 46 snoring males.....56

3-2 Incidence of low and high levels of apnea/
hypopnea by age in 46 snoring males.............58

3-3 Significant (p<.05) Pearson correlations
between nocturnal respirEtory variables
and demographic variables in 46 snoring
males............................................ .

3-4 Means and standard deviations for selected
demographic and nocturnal respiratory
variables in 46 snoring males grouped by
level oi apnea/hypopnea.........................61

3-5 Means and standard deviations for health
related variables in 46 snoring males grouped
by level of apnea/hypopnea......................65

3-6 Means and standard deviations of selected
sleep variables from 43 snoring males............66

5-7 Meens end standard deviations of selected
sleep variables from 43 snoring males
groupec by level of apnea/hypopnea..............68

3-8 Leans and standard deviations for daytime
sleepiness variables in 46 snoring males
grouped by level of apnea/hypopnea..............69

3-9 Significant (p<.05) Pearson correlations
between nocturnal respiratory and sleep
questionnaire variables in 46 snoring males.....71

3-10 Means and standard deviations of sleep
questionnaire variables from 46 snoring
males groupec by level of apnea/hypopnea........71









3-11 Significant (p<.05) Pearson correlations
between nocturnal respiratory and sleep log
variables in 46 snoring males.....................75

3-12 Means and standard deviations of sleep log
variables in 46 snoring males grouped by
level of apnea/hypopnea......................... 73

3-13 Multivariate regression of demographic and
nocturnal respiratory variables on cognitive
scores in 46 snoring males......................76

3-14 Significant (p<.05) partial correlations,
controlling for age, between nocturnal
respiratory and cognitive variables in
46 snoring males................................ 79

3-15 Means and standard deviations of cognitive
scores in 46 snoring males grouped by
level of apnea/hypopnea.........................80

3-16 Multivariate regression of sleep and
nocturnal respiratory scores on cognitive
variables in 43 snoring males...................85

B-1 Demographic and medical variables from 46
snoring males...................................109

B-2 Daytime sleepiness variables from 46 snoring
males........................................... 111

B-3 Sleep questionnaire variables from 46 snoring
males...........................................113

8-4 Seven day sleep log means from 46 snoring
males ..........................................115

B-5 Neuropsychological variables from 46 snoring
males........................................... 117

B-6 Respiratory variables from 46 snoring males.....119

B-7 Sleep variables from 46 snoring males...........121















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

SLEEP APNEA ACTIVITY AND ITS CONCOMITANTS
IN A SUBCLINICAL POPULATION

By

DAVID THOMAS REED BERRY

August 1985

Chairman: Wilse B. Webb
Major Department: Clinical Psychology

Sleep Apnea Syndrome (SAS) is a nocturnal respiratory

disorder with serious consequences. Patients with SAS

experience apneas (pauses in breathing of 10 or more

seconds) and hypopneas (declines in the amplitude of

breathing accompanied by oxygen desaturation) while

asleep. Recently, it has become apparent that apneas and

hypopneas may occur in otherwise apparently normal

subjects. A sample selected to display a wide range of

sleep disordered breathing (heavy snoring males in good

health and over 30 years of age) received detailed

measurement of their nocturnal respiration, sleep,

cognitive skills, and health in an attempt to clarify the

impact of apneas and hypopneas in a subclinical population.

Forty-six heavy snoring males in good health with a

mean age of 50 years and a mean weight of 190 pounds

comprised the sample. During an experimental evening they

vii









received testing and filled out various questionnaires

before sleeping overnight with their sleep and breathing

continuously recorded. Sixty-two per cent of these subjects

experienced at least one event, while 13% had more than 5

events per hour of sleep. Most events occurred in light

slow wave or REM sleep and were linked to obesity. Subjects

with high levels of apnea/hypopnea (more than 5 per hour)

experienced significantly exacerbated oxygen desaturation

relative to the remaining subjects. Oxygen desaturation was

linked to higher blood pressure readings. A typical "first

night" effect on the sleep of the subjects in the lab was

noted, resulting in a lighter sleep than usual and obscuring

possible relationships between sleep and respiratory events,

which were not observed. Deteriorating overnight

respiratory indices were associated with increased

sleepiness and napping. High levels of apnea/hypopnea were

associated with declining scores on tests measuring non-

verbal intelligence, verbal and non-verbal memory,

expressive verbal fluency, and cognitive flexibility. It is

speculated that hypoxia induced changes in cerebral

cholinergic synthetic pathways underlie these changes.

It is concluded that a significant subgroup of heavy

snoring males who experience multiple apnea/hypopneas are at

increased risk of significant oxygen desaturation, daytime

sleepiness, and cognitive changes. This group may fall on a

continuum with sleep apnea syndrome patients.


viii















CHAPTER ONE
INTRODUCTION

Sleep apneas are respiratory pauses during sleep.

According to the Association of Sleep Disorders Centers

Nosology of Sleep Disorders (ASDC, 1979), a sleep apnea

syndrome (SAS) is a potentially lethal condition charac-

terized by multiple apneas, excessive daytime sleepiness,

alterations of consciousness, and cardiopulmonary

complications. The apnea episodes are thought to be the

causal factor behind the other deficits associated with

SAS (Guilleminault et al., 1978).

Polysomnographic recordings in various populations

have revealed the occurrence of sleep apneas in the

general population, and particularly in aging popula-

tions. Estimates of the prevalence of at least a few

sleep apneas in older subjects have ranged as high as 75%

(Carskadon and Dement, 1981b). Carskadon et al. (1980)

have suggested that sleep apneas in older subjects may be

implicated in insomnia, cardiac ailments, and even

senility.

However, certain reports have documented the presence

of multiple sleep apnea episodes in subjects without

apparent pathology (Crr et al., 1979). This evidence

obscures the role of sleep apneas in the other pathologies









noted in SAS patients. The link between apneas and patho-

logy may be more complex than was originally thought,

given its apparently benign occurrence in otherwise

asymptomatic subjects.

The present study will examine the correlates of

sleep apnea episodes in a population selected to display a

wide range of such events. The relationship between sleep

apnea indices and sleep characteristics, age, weight and a

variety of deficit measurements will be analyzed.

A review of the current data available on various

aspects of sleep apneas and sleep apnea syndromes will be

presented as background for the study. The diagnosis,

history, incidence, and characteristics of sleep apneas

and sleep apnea syndrome will be reviewed, followed by a

discussion of variables associated with sleep apneas.

Lastly, the deficits commonly found in sleep apnea syn-

drome will be detailed, along with measurements thought to

be sensitive to these deficits.

Definitions and Terminology

Sleep apneas have been defined as cessations of air-

flow at the mouth and nose which last for 10 or more

seconds (Guilleminault et al., 1976). In a sleep apnea

syndrome (SAS), multiple sleep apneas are accompanied by

snoring, excessive daytime sleepiness, cardiopulmonary

complications, and altered states of consciousness. The

ease of quantification of sleep apneas has led to their

use as an important tool in diagnosing SAS. Guilleminault









et al. (1976) first suggested that the occurrence of 30 or

more apneas in 7 or more hours of sleep was diagnostic of

SAS, while Guilleminault et al. (1978) proposed the use of

an apnea index (AI; number of apneas/number of hours

sleep) in excess of 5 as a cut-off score to diagnose the

syndrome. Later, Carskadon et al. (1980) combined apneas

and hypopneas (a reduction in airflow of 50% or more at

the mouth and nose) to form an apnea + hypopnea index

(AHI) with 5 utilized as a diagnostic level of sleep apnea

syndrome. As can be appreciated from the preceding dis-

cussion, the diagnostic criteria for SAS have undergone a

rapid evolution, with remarkably little validation work

published.

History and Characteristics of Sleep Apnea Syndrome

Sleep Apnea Syndrome falls within the larger category

of Sleep Disordered Breathing (Block, 1980). Respiratory

disorders within this category involve changes in

breathing patterns during sleep. Syndromes described

within this category include Pickwickian Syndrome; Sleep

Apnea Syndrome; Chronic Obstructive Lung Disease; and

syndromes of Primary and Secondary Alveolar

Hypoventilation (Block, 1980; Guilleminault et al.,

1976).

The first disorder described in this group was the

Pickwickian Syndrome. These patients were initially

identified by a chronic daytime hypersomnolence, obesity,

Cheyne-Stokes breathing, hypercapnia and hypoxemia. This









syndrome typically results in complications such as right

sided congestive heart failure, pulmonary hypertension,

and peripheral edema (Block, 1980). An exacerbation of

disordered breathing (apnea) as well as hypercapnia and

hypoxemia (often severe) occurs with sleep onset. Treat-

ment consists of oxygen therapy, respiratory stimulants,

and maintenance of wakefulness.

The discovery of a sleep induced exacerbation of the

symptoms accompanying Pickwickian Syndrome spurred a

systematic investigation of sleep respiratory events

(Gastaut et al., 1965). The new emphasis on noctural

respiration led to a series of investigations which docu-

mented sleep induced respiratory dysrhythmias in non-obese

subjects (Lugarasi et al., 1966). A new group of sleep

disordered breathers designated as Sleep Apnea Syndrome

emerged, with several features which distinguished them

from Pickwickians. Among these differences were weight

(i.e., SAS patients were not necessarily obese), daytime

respiratory control (i.e., SAS patients were typically not

hypercapnic during waking as were Pickwickians), sex

(i.e., while AO; of Pickwickians were female, very few

premenapausal females have been identified with SAS), and

age (i.e, SAS patients are typically younger than

Pickwickians)(Block, 1980).

Two large samples of SAS patients have been reported

in the literature, allowing characterization of typical

SAS patients. Guilleminault et al. (1976) reported on a









series of 50 patients diagnosed with SAS over a period of

six years. Age ranged from 28-62 with a mean age of

45.5. The sample consisted of 48 males and 2 females.

Thirty-nine were referred on the basis of excessive day-

time sleepiness, while 7 were referred for loud snoring

and abnormal movements reported by their spouses. Sixty

percent of the patients were more than 15% above ideal

weight. Other symptoms included hypnagogic halluci-

nations, automatic behaviors, intellectual deterioration,

personality changes, impotence, morning headaches, and

hypertension. Weitzman, Kahn and Pollak (1980) report

data on a group of 10 SAS patients seen in their clinic.

These subjects displayed the following characteristics:

age between 38 and 47, obesity, male sex, nicotine depen-

dence, hypertension, complaints of excessive daytime

sleepiness, and the presence of serious cardiac arrhyth-

mias. (This group was selected from a larger sample of 38

patients.) Thus, pathological symptomatology associated

with a diagnosed SAS include excessive daytime somnolence,

cardiopulmonary complications and cognitive/intellectual

changes; a "typical" SAS patient is a middle aged male

who snores and is obese.

Incidence of Sleep Apnea Activity

The incidence of sleep apnea activity and sleep apnea

syndrome have been frequently confuse. While sleep

apneas have been an important diagnostic tool for identi-

fying sleep apnea syndrome, they have not been









definitively demonstrated to be the necessary and suffi-

cient cause of the syndrome. Thus, for the purposes of

the following review, the focus will be on occurrence and

level of sleep apnea activity in both sleep apnea syndrome

subjects and other populations. The full range of sleep

apnea activity will be reflected by noting the presence of

sleep apnea activity, and an indication of the quantita-

tive level of the activity will be presented in the apnea

index (AI) or the apnea + hypopnea index (AHI). Recall

that an AI of 5 has been frequently utilized as a cut-off

score for identifying SAS patients.

A review of the literature suggests that at least

three sampling strategies have been used for studying the

incidence of sleep epnea activity. Subjects have been

selected on the basis of one of the following three

criteria: 1) absence of sleep or respiratory complaints,

2) presence of sleep or respiratory complaints, or 3)

sampling without regard to these variables.

Because diagnosis of SAS is often dependent on number

of sleep apneas (or level of SAA), most researchers report

number of apneas observed, although some report only on

number of subjects who exceed one of the criteria for

SAS. A review of studies with samples of 20 or greater

will be presented with careful note of the sampling

strategy utilized, as well as the dependent measure of

apnea.









One sampling strategy which has been pursued is to

exclude subjects who complain of sleep or respiratory

difficulties. Screening for sleep complaints varies

widely. Guilleminault et al. (1978) studied a sample of

20 normalss" (presumably noncomplaining subjects). These

subjects ranged in age from 40-60 years. Males showed a

mean number of apneas of 7 (range 1-25), while females

showed a mean number of apneas of 2 (range 0-5). Thus,

none of these noncomplaining subjects presented a clinical

syndrome of sleep apnea, although sleep apneas did occur

apparently without pathology. Block et al. (1979) studied

49 subjects (M = 30, F = 19). Males in the sample had a

mean age of 38, while females had a mean age of 29. Any

potential subjects who complained of breathing difficul-

ties or sleep disturbances were excluded from the study.

Twelve males (40%) had episodes of sleep apnea while only

3 females (15%) suffered as well. For those subjects,

males had a mean of 4.2 episodes, while females had a mean

of 3. Within this sample, the number of apneas was not

significantly correlated with increasing age, although

occurrence of oxygen desaturation (a concomittant of

disordered breathing) was positively correlated with

age. Additionally, although the mean number of sleep

apneas was not significantly different between sexes (M =

4.2, F = 3), the more sensitive measure of apneas with

desaturation was (M = 3, F = 0). Webb (1974) studied 2C

males with a mean age of 44. Subjects were excluded from









the study if they reported any "serious" sleep com-

plaint. Two of the subjects reported consuming hypnotics

for sleep induction, indicating that the criterion used in

this study was less stringent than that of Block et al.

(1979) report. Nine of the subjects (45%) suffered at

least one episode of sleep apnea with a mean number of

episodes of 2.5. All of the subjects with episodes of

sleep apnea were older than the mean age of the sample,

leading Webb to conclude that apnea is related to age.

Bixler et al. (1982) constructed an age stratified patient

representative sample of 100 subjects (M = 41, F = 59).

In perhaps the most rigorous screening for sleep com-

plaints in this group of studies, Bixler et al. eliminated

any subject with a sleep complaint or chronic medication

usage. For the entire sample, 6 males (14.6%) and 5

females (10.25) had at least one episode of sleep apnea.

For subjects under 30, 3.3%; for those between 30 and 50,

12.8e; and those 50-74, 19.4; suffered episodes of sleep

apnea. For the subsample over 60, 25% had sleep apnea

activity with 8.5 episodes for males and 10.5 episodes for

females. In this sample, there was a nonsignificant trend

toward more apnea activity in males and a significant

positive correlation between age and sleep apnea.

block et al. (1980) report data on 20 post menapausal

women with a mean age of 59. Eight women (40%) had epi-

sodes of sleep apnea, with a mean number of episodes of









5. There was a significant correlation between apnea and

age.

For noncomplaining subjects then, males suffer a

higher incidence of apnea activity in virtually every

study which compares sex effects. For males, incidences

range widely from a low of 14.6% (Bixler et al., 1980) to

a high of 45% (Webb, 1974); while females range from a

low of 10.2% (Bixler et al., 1982) to a high of 40% (Block

et al., 1980). These differences are probably due at

least in part to age differences in male samples, as well

as age differences in females (only as they reflect mena-

pausal status). Reported number of apneas range from 2.5

(Webb, 1974) to 8.5 (Bixler et al., 1982) in males and

from 2 (Guilleminault et al., 1978) to 10.5 (Bixler et

al., 1982) in females. Many studies report a positive

correlation between number of apneas and increasing age.

A second sampling technique which has been utilized

has been studying subjects with sleep and breathing com-

plaints. Kales et al. (1982) studied 200 subjects with

primary complaints of insomnia. There were 82 males and

118 females in the sample with a mean age of 42.3. Sleep

apnea activity was noted in 10.5% of this group (M =

13.4%, F = 8.5%) with a mean number of 11.2 episodes for

males and 6.6 for females. Subjects with sleep apnea

activity were significantly older (46.8) than those with-

out (40.6). Ancoli-Israel et al. (1981) recruited 24

subjects whose answers to questionnaires provoked









suspicion of insomnia or nighttime breathing and muscular

events. Eleven males (7 = 72.5) and 13 females (R = 68.5)

were studied. All of these subjects, save one male and

one female, suffered at least one episode of sleep

apnea. For those subjects with at least one episode of

sleep apnea, males had a mean of 53 episodes of apnea,

while females had a mean of 32 episodes. In contrast to

the Kales et al. study, 6 males had a high level of sleep

apnea (AI>5), while 3 females met this criterion. A rela-

tionship between age and apnea was not reported in this

sample, possibly because of restricted range. These two

studies of sleep dist'r'bed subjects provide conflicting

data. That is, the insomniac alone sample did not demon-

strate an apnea incidence which was substantially

different from a noncomplaining sample (10.5%), while the

sample recruited for sleep induction, respiratory, and

muscular differences showed very discrepant apnea inci-

dences (nearly 100%). One potential explanation involves

age, as the Ancoli-Israel sample is substantially older

than Kales', although the aging samples have not reported

this level of apnea before. An alternative explanation is

that sampling insomniacs does not generate a high number

of apneas, whereas aging sleep and respiratory disturbed

subjects show a very high incidence of apnea. It would

seem then that the relevant "complaint" would involve

those reported by Ancoli-Israel et al. which are in

addition to those noted by Kales et al. These are (as








nearly as can be judged) questions surrounding

respiration.

The final category of incidence studies involves

those which do not consider sleep complaints in their

sampling. As in the other studies, these vary in their

generalizability. Carskadon and Dement (1981b) recruited

40 aging subjects, including 18 males (R = 72.7) and 22

females (R = 74). Their exclusion criteria are somewhat

problematic, involving excluding subjects who "spon-

taneously complain of sleep problems." Thus, it is

unclear what this is a representative sample of, specifi-

cally whether the distribution of those not spontaneously

complaining of sleep disturbances may be different from

those who actually do not suffer from sleep distur-

bances. These considerations aside, Carskadon and Dement

report that 16 males (88%) and 16 females (72%) demon-

strated at least one sleep apnea episode, while 8 males

(44%) and 7 females (31.8%) have high levels of sleep

apnea (AI>5). Males with apnea showed a mean number of

51, while females showed a mean number of 34. Sex

differences appear to have fallen out at this age, while

the restricted range of this extreme age sample does not

lend itself to a correlational analysis of age versus

apnea. In another study of limited generalizability,

Kreiss et al. (1982) randomly selected inpatients on a

Veterans Administration medical ward. These patients, all

presumably male, numbered 26. Further data reported were









sketchy, but 7 (27%) met the sleep apnea syndrome

criterion of >30 episodes of sleep apnea. The fact that

these subjects were patients on a medical ward obviously

restricts generalizability of these findings. Finally, in

the only truly random sampling reported, Ancoli-Israel et

al. (1982) present initial data on randomly selected sub-

jects. Forty subjects (sex not reported) with a mean age

of 71.5 were studied. Thirteen (32.5%) met the sleep

apnea syndrome criterion of 30 or more apneas. The sleep

apnea subgroup was older, 74.2, than the full sample.

The data collected from relatively random sampling

are primarily based on aging populations (>60). The data

from these studies suggest that between 32.5% (Ancoli-

Israel et al., 1981) and 37% (Carskadon and Dement, 1981b)

of aging subjects suffer high levels of sleep apnea

(AI>5). An even larger percentage, perhaps greater than

75%, suffer at least some episodes of sleep apnea. A

major deficiency in these randomly sampled studies is a

lack of data on younger subjects.

Thus data collected under different sampling proce-

dures yield variable incidence rates of sleep apnea and

AI>S. Subjects without sleep or respiratory complaints

have a very low incidence (perhaps 0) of AI>5, between 14%

and 455 of males surveyed had at least some apnea, while

females ranged between 10% and 40% experiencing some

apnea. Sampling subjects with sleep or respiratory com-

plaints suggested that subjects suffering respiratory








difficulties had very high incidences of apnea (almost

100% for both males and females) with increase of inci-

dence of AI>5 (55% for males, 23% for females). Sampling

without regard to sleep or respiratory difficulties is

largely confined to aging populations. Incidences of

apnea in these aging populations would seem to be rather

high (88% males, 72% females) while incidence of AI>5 seem

to be in the range of 32%-37%.

It is clear that different populations (as defined by

sampling strategies) suffer different incidences of sleep

apneas and AI>5. The clinician dealing with sleep dis-

orders would be wise to consider the acturial base rates

of apnea and AI>5 provided by these data. It is apparent

that sex, age, and respiratory status are significantly

related to the incidence rates of SAA and SAS. These

factors must be considered in any study of apnea and asso-

ciated deficits.

Temporal Characteristics of Apneas

The temporal characteristics of sleep apneas repre-

sent a potentially important variable. Above it was shown

that a wide variability in number of sleep apneas occurs,

and the arbitrary nature of the division between SAA and

SAS subjects was noted. Therefore, it is of interest to

determine whether the duration of sleep apneas provides a

more reliable distinction between the two groups. Another

variable of interest involves the placement of apneas

within sleep, as Guilleminault et al. (1976) noted that









normals might experience apneas during REM without

pathological consequence. Therefore, data available on

these two issues will be reviewed, including only studies

with a sample size great than 10 (with one exception).

Reports on normal subjects below 30 years of age are

somewhat rare, although some do exist. Bixler et al.

(1982) included a group aged 18-29 years (M = 13, F =

17). One subject in this group had sleep apneas (12) with

a mean duration of 15 seconds. Bixler et al. noted that

most apneas occurred in Stages 1 and REN sleep. Part of

the sample of Block et al. (1979), the females, had a mean

age of 29. Three of these subjects showed apneas with a

mean duration of 14 seconds. Again, a disproportionate

number of apneas occurred in Stages 1 and REM. Thus in

normal subjects under 30, apneas are somewhat rare, with a

mean duration of approximately 15 seconds when they do

occur. Most apneas occur in light and REM sleep.

More data are available on normal subjects between 30

and 60. Guilleminault et al. (1978) do not report on

duration of apneas in their sample aged 40-60, but they

note that apneas occurred only in Stages 1 and REM.

Bixler et al. (1982) reported that, in their 50-44 year

old group, 5 subjects experienced sleep apnea with a mean

duration of 13.1 seconds. A disproportionate number ol

apneas occurred in Stages 1 and REM. Block et al. (1979)

studied 30 males with a mean age of 38, with 12

experiencing sleep apneas. These apneas had a mean









duration of 20 seconds and occurred most frequently

(30/60) in Stages 1 and REM. Eight of 20 aging females,

in the report by Block et al. (1980), had sleep apneas

with a mean duration of 19 seconds. Sixty-two percent of

apnea episodes occurred in Stages 1, 2 or REM. Thus in

middle aged normal subjects, incidence of sleep apnea

appears to rise, with a mean duration of between 13 and 20

seconds, perhaps slightly longer than in younger groups.

Again, apneas occurred mostly in light (Stages 1 and 2)

and REM sleep.

Duration data on aging normals are limited.

Carskadon and Dement's (1981b) study of aging normals (M =

72, F = 74) does not report duration of apnea events.

Similar, Ancoli-Israel et al. (1982) do not provide this

information on their aged normals (R = 71.5). Bixler et

al. (1982) noted a high incidence of apneas in their sub-

jects over 60, but failed to report a mean duration for

this group. Given the sketchy details reported for this

group, it is difficult to draw any conclusions about

changes in temporal duration or placement of apneas in

aging normals.

To summarize the temporal data in normals, apneas

appear somewhat rarely in normals below 30, with a mean

duration of 15 seconds. Incidence of apneas increase

between 30 and 60, with a wider variation in mean dura-

tions reported (13-20 seconds). Apneas occur most

frequently in light and REM sleep in normals of all ages.









Temporal data on subjects diagnosed with sleep apnea

syndrome (>30 or AI>5) primarily include patients between

the ages of 30 and 60. The reason for the lack of older

SAS subjects is unclear, particularly in light of data

reported above, demonstrating very high percentages of SAS

in certain groups of aging subjects.

Garay et al. (1981) studied 13 sleep apnea syndrome

subjects. They divided these 12 males and 1 female into

two groups, based on their daytime resting CO2 levels.

Subjects who were eucapnic during the day had a mean apnea

duration of 16 seconds, while those who were hypercapnic

had a mean duration of 17 seconds. No data on distribu-

tion relative to sleep stages were reported.

Guilleminault et al. (1978) studied 50 predominantly male

(R age 48) sleep apnea syndrome patients. Apneas in this

population averaged 22 seconds in duration. Apneas were

longest in REM, and few occurred in Stages 3 or 4 sleep.

Fujita et al. (1981) studied 12 male SAS patients with a

mean age of 43. Apnea episodes averaged 23 seconds in

duration. Weitzman et al. (1980) studied 10 SAS males

whose age ranged from 38-57. The mean duration of apneas

in this group was 30 seconds. All apneas occurred in

light or REM sleep. Orr et al. (1979) studied 4 hyper-

somnolent males diagnosed with SAS. A control group of 4

nonhypersomnolent subjects was matched on number of ob-

structions per minute. Mean age was 57 for asymptomatic

and 42 for symptomatic subjects. Asymptomatic subjects








had apneas with a mean duration of 25.9 seconds, while

symptomatic subjects had a mean duration of 20.8 seconds.

Several characteristics emerge from the group of

reports on SAS subjects. There is wide variability in

duration of apneas, with reported means ranging from 16-30

seconds. The data available on placement of apneas in

sleep suggest that in SAS subjects most apneas occur in

light and REM sleep.

Comparison of data on normal and SAS subjects reveals

substantial overlap between mean duration of apneas in SAS

subjects (13-20 seconds) and SAS subjects (16-30

seconds). Particularly interesting was the Orr et al.

(1979) study which demonstrated a lower mean duration of

apneas for hypersomnolent subjects than for nonhypersomno-

lent subjects. Data on temporal placement of apneas show

most events occurring in light or REM sleep for both

normal and SAS subjects. Taken together, these data

suggest that substantial overlap exists on temporal

variables between normal and SAS subjects indicating a

limited utility for diagnostic purposes.

Variables Associated With Sleep Apnea
Syndrome and Activity

It was noted above that several variables have been

suggested as associated with increased incidences of sleep

apneas and SAS. These variables represent potential

"markers" of subpopulations which might experience higher

incidences of sleep apnea activity. These primary









associated variables include sex, age, snoring and

obesity. The data available on these variables and their

relationships to sleep apnea activity will be reviewed,

followed by a brief discussion of their relevance to

experimental methodology.

Regarding sex variables, as previously noted,

Guilleminault et al. (1978) found 96% of their SAS

patients to be males. In the Block et al. (1979) study of

nonsleep disturbed subjects, 40% of the males in the

sample had at least one episode of apnea, while only 15"

of the females so suffered. Bixler et al. (1982) noted a

trend toward more sleep apnea activity in males. Block et

al. (1980) studied 20 post menapausal females and found

apnea activity in 40% of them, leading them to suggest

that some factor associated with menapause (e.g., pro-

gesterone) might somehow "protect" premenapausal females

from apneas, and that this factor along with its benign

influence was lost with menapause. This explanation pre-

dicted the low level of apnea activity in younger females,

as well as the heightened apnea activity of aging

females. While the data of Carsakdon and Dement (1981b)

show aged females to be close to aged males in percentage

exhibiting sleep apnea activity, a report by Smallwood et

al. (1983) found no evidence of sleep apnea activity in

their 6 elderly (post menapausal) females, while elderly

males showed significant levels of apnea activity. The

limitations of the sample size in the Smallwood et al.








report might explain this discrepancy as representing a

sampling bias. Taken together, this evidence strongly

supports the notion that in young and middle aged popula-

tions, sleep apnea is a male phenomenon. Additionally,

evidence points toward a closing of the gap between

females and age matched males as the former group passes

menapause.

A second factor which has been implicated in sleep

apnea activity is age. Webb (1974) found that his sub-

jects exhibiting sleep apnea activity were significantly

older than his full sample. This was also the case for

the subjects of Kales et al. (1982) and Ancoli-Israel et

al. (1982). Smallwood et al. (1983) found that sleep

apnea was an age dependent phenomenon with male subjects

over 50 exhibiting significantly more apneas than those

under 30. Other evidence includes that of Bixler et al.

(1982) who found a positive correlation between age and

apnea in their full sample, as did Block et al. (1980) in

their post menapausal females. It seems clear from these

data then that increasing age is a factor predisposing for

sleep apnea activity.

Snoring has been closely associated with sleep apnea

(Lugaresi et al., 1982). This may be due to a causal

relationship or a single common pathway. Lugaresi et al.

propose the latter explanation. Specifically, they

suggest that both snoring and SAS are due to a sleep

induced stenosis of the upper airway. With this premise









in mind, they conducted a survey of 1000 subjects on the

incidence of snoring. Chronic snorers included 31% of the

males and 19% of the females. In an elaboration of this

study, Lugaresi et al. (1980) questioned 5713 indivi-

duals. In this sample, 24% of the males and 13% of the

females were chronic snorers. Snoring increased with age,

and by 60 years of age 60% of the males and 40% of the

females snored. Of note was the observation that hyper-

tension occurred more frequently in snorers than non-

snorers, suggesting that the former was more at risk for

cardiovascular complications.

Virtually every study of SAS patients notes the pre-

sence of snoring in all these subjects (Sullivan and Issa,

1980; Block, 1980; Coverdale et al., 1980). This is

true for both predominantly obstructive, and predominately

central SAS subjects (Guilleminault et al., 1978). How-

ever, the presence of heavy snoring in asymptomatic

(nonhypersomnolent) apneas is also well documented (Orr et

al., 1979; Fisher et al., 1978; Berry and Block, 1983),

as well as in otherwise normal subjects. These data may

be taken as evidence for snoring as a necessary, but not

sufficient, condition for the occurrence of sleep apnea.

The role of body weight in SAS is less clear.

Recall that Pickwickian patients are invariably obese, as

well as hypercapnic. Guilleminault et al. (1978) stress

that SAS may occur in nonobese as well as obese sub-

jects. However, Block (1980) states that he has found SAS









subjects to be uniformly obese. A review of the

literature is in order to clarify this point. First

normal, then obese, then SAS groups will be reviewed.

Bixler et al. (1982) found that their subgroup of

subjects with sleep apnea activity weighed significantly

more than the remainder of the sample. Kales et al.

(1982) found that within their sample of 200 insomniacs

and 100 normals, a significant positive correlation

emerged between a weight:height ratio and apnea acti-

vity. Block et al. (1979) found a positive correlation

between apneas, oxygen desaturation, and weight within

their normal males, while females did not demonstrate this

relationship. Using an older female sample, Block et al.

(1980) found a positive correlation between weight and

oxygen desaturation, but not apnea. Thus in normal sub-

jects, there appears to be a positive relationship between

weight and apnea activity, although this may be restricted

to males.

A research approach which bears on this question

involves studying obese subjects alone. Two studies on

successive referrals for gastric bypass surgery were

noted. Sicklesteel et al. (1981) found that of 19 suc-

cessive referrals, all 14 males exhibited sleep apneas,

while only 5 of the females showed evidence of oxygen

desaturation alone. Harmon et al. (1981) studied gastric

bypass patients and found 5 of 6 males to suffer apnea, as








opposed to none of the females. This evidence provides

further evidence for a role of increasing weight in apnea.

The data on symptomatic SAS subjects are more com-

plex. Several studies (Sullivan and Issa, 1980; beitzman

et al., 1980; Garay et al., 1981; Zwillich et al., 1982)

report that all their SAS subjects were obese, although

definitions of obesity either vary widely or remain un-

specified. In contrast, Guilleminault et al. (1978)

reported that only 60% of their SAS sample was greater

than 15% above ideal weight. Coverdale et al. (1980)

indicate that only two of their subjects proved to be

obese, according to their definition (>125% of ideal

weight). These reports suggest that most, although not

all, SAS subjects are obese.

Several investigators have sought to compare weight

in symptomatic (somnolent) and asymptomatic subjects with

enough sleep apnea episodes to be diagnosed as SAS. Orr

et al. (1979) compared 4 symptomatic SAS subjects with

matched asymptomatic subjects. In this sample, symptoma-

tic SAS subjects were heavier than asymptomatic sub-

jects. Examination of the data of Berry and Block (1963)

indicates that their symptomatic subjects were heavier

than their asymptomatic subjects (118 kg vs. 91 kg).

Standing in contrast to these data is a study reported by

Fisher et al. (1978), which included 19 subjects referred

for SAS evaluation. Based on number of apneas per night,

Fisher et al. divided their subjects into 3 groups. Group








I had >100 apneas per night, Group II had >30 but <100

apneas per night, while Group III had less than 20 apneas

per night. Thus Groups I and II were diagnosable as SAS

(>30), while Group III was not. Obesity was present in

Groups I and II, but in only 3 subjects in Group III.

However, these differences were not statistically signifi-

cant.

Taken together, this evidence suggests a positive

relationship between weight and apnea in the general popu-

lation. However, SAS patients are not invariably obese.

Further, weight loss fails to result in a decrease in

apneas in many SAS patients. Thus the relationship

between obesity and full blown SAS is less than clear.

The preceding pages have investigated the currently

understood role of several variables (sex, age, snoring

and weight) in sleep apnea activity. Presently available

data suggest that in young and middle aged subjects, apnea

is a predominantly male phenomenon, although post mena-

pausal females may close the gap. Increasing age is re-

lated to sleep apnea activity. Snoring, which occurs in a

significant percentage of the general population, is

thought to be a necessary but not sufficient condition for

SAS. Weight was shown to have a positive relationship to

sleep apnea activity in the general population, although

its presence or role in SAS is equivocable.

These conclusions have important implications for

experimental designs directed at understanding sleep apnea








activity and its correlates in a subclinical population.

It is clear from the earlier review of apnea incidence

that sampling from the general population would result in

a relatively low population of individuals with sleep

apnea, an inefficient approach from the perspective of the

limited resources available for research. Thus selection

variables must be considered with one eye on efficiency

and the other on protecting the generalizability of

results. Selection criteria should also aim toward eluci-

dating the relative roles and possible interactions of

these variables in sleep apnea. Therefore, selection

variables should have as straight-forward a relationship

as possible with sleep apnea, seeking to answer more ques-

tions than they raise.

With these caveats in mind, the variables available

as selection criterion will be reviewed and conclusions

drawn about their relative merits. The first variable

considered, sex, has been relatively well investigated.

It is clear that sleep apnea activity is primarily a male

circumstance. Therefore, the selection of female subjects

is contraindicated from the standpoint of efficiency. Age

has been demonstrated to exhibit a positive relationship

with sleep apnea, indicating selection of older sub-

jects. Snoring is another variable with a close relation-

ship to sleep apnea. The apparently common mechanism of

snoring and sleep apnea (stenosis of the upper airway)

indicates that possibly very few subjects without snoring








suffer sleep apneas. A significant proportion of snorers

might be expected to suffer apneas, while few nonsnorers

might have apnea. The last variable, weight, while having

a positive correlation with apnea activity in the general

population, is of uncertain pertinence to SAS. Addi-

tionally, changes in weight may occur without changes in

apnea activity. Therefore, this variable has a less than

explicit connection to apnea activity and violates the

second principle adopted to consider these variables.

This review of available data suggests that sampling

aging, snoring, males would generate a high proportion of

sleep apnea activity. External validity for this idea is

drawn from its similarity to the "typical" SAS patient

described earlier. Extant data suggest that it may also

prove useful to utilize weight measures as a covariate, in

an attempt to control its apparent influence on apnea

activity.

Deficits Associated With Sleep Apnea Syndrome

Sleep Apnea Syndrome is thought to result in several

deficits. These include cardiopulmonary complications,

hypersomnolence, hypoxia and intellectual changes. Some

researchers have suggested that these deficits may be

present, in an attenuated form, in subclinical apnea

(Ancoli-Israel et al., 1981; Carskadon and Dement,

1981b). A discussion of the evidence for the presence of

these deficits will be presented, followed by proposals

for appropriate measures of these variables.









Cardiopulmonary Deficits

Cardiopulmonary complications have been reported as

concomittents of SAS (Guilleminault et al., 1978).

Coccagna et al. (1972) reported high incidences of hyper-

tension and congestive heart failure in their SAS sub-

jects. Schroeder et al. (1978) studied 22 SAS patients

with a mean age of 47. Six subjects had waking systemic

hypertension, and several had waking cardiac abnorma-

lities. During sleep studies, 20 patients developed

significant rises in systemic arterial pressure which

cycled with episodes of apnea, while 21 patients developed

pulmonary arterial hypertension. Tracheostomy reversed

all these abnormalities, strongly implicating the apneas

in their etiology. Tilkian et al. (1978) studied 25 male

SAS subjects with a mean age of 44. Twenty-four of the 25

showed sleeping sinus arrhythmias, while 9 developed more

serious symptoms such as asystole. Seventeen of these

patients received tracheostomy, which abolished all of

these abnormalities. Experimental occlusion of the

tracheostomy during sleep in 6 of these patients resulted

in a return of the cardiopulmonary symptoms. Zwillich et

al. (1982) studied six consecutive male SAS patients. In

all patients, bradycardia accompanied any apnea with

significant desaturation. Additionally, a significant

correlation was noted between degree of desaturation and

severity of bradycardia. Bradycardia was abolished or

attentuated during administration of 02 enriched air.









Zwillich et al. proposed that bradycardia during apnea

results from increased vagal tone mediated by carotid body

chemoreceptors. Fujita et al. (1981) reported on 12 male

SAS patients. All reportedly suffered cardiac arrhythmias

during sleep. The body of data suggests that SAS patients

frequently suffer from cardiac abnormalities such as

hypertension, arrhythmias, and bradycardia.

A second body of data exists which indicates that SAS

patients do not invariably have cardiopulmonary compli-

cations. Coverdale et al. (1980) studied 14 patients, of

whom 8 were diagnosed with SAS. Only 2 of these patients

had evidence of corpulmongle, while 2 more had systemic

hypertension. None of these patients demonstrated severe

brady or tachycardia during sleep. In Orr et al. (1979),

8 patients, all of whom were technically diagnosable as

having SAS, all symptomatic (somnolent) patients had evi-

dence of right sided heart failure as well as frequent

arrhythmias during apneas. Asymptomatic subjects had none

of these symptoms. Kreiss et al. (1982) studied 26

Veterans Administration ward patients and found 7 to have

a SAS. The SAS patients had significantly more signs of

congestive heart failure, but not hypertension or

angina. Ancoli-Israel et al. (1981) sampled 24 elderly

subjects with complaints of sleeping respiratory ois-

orders. Nine of these subjects were diagnosable as SAS,

but this group did not have significantly more heart








disease or hypertension. Thus a group of SAS patients

without evidence of heart disease exists.

Taken together, these data indicate that cardiac

complications frequently, but not inevitably, accompany a

sleep apnea syndrome. That a large number of apneas may

be present without a significant increase in cardio-

pulmonary complications is demonstrated by the Orr et al.

and Ancoli-Israel et al. studies. As the relationship

between these complications and a full blown SAS is

unclear, the possibility of cardiopulmonary sequelae from

a subclinical level of apneas is even more uncertain.

However, an assessment of cardiopulmonary status is

indicated by the frequent association of cardiopulmonary

complications with sleep apnea syndrome.

Arousal Deficits (Hypersomnolence)

Excessive Daytime Sleepiness (EDS), or hypersomno-

lence, has been described as the single most common result

of SAS (Dement et al., 1978). Indeed, patients who meet

the criterion of number of apnea episodes and exhibit EDL

are described as symptomatic, while those with number of

apneas alone are labeled asymptomatic (Orr et al.,

1979). As an aside, this distinction illustrates again a

central issue of SAS. If patients exhibit the number of

apneas qualifying them for SAS, but no other pathology,

are they suffering from a pathological process? Dement et

al. (1978) believe that they are. They suggest that most,

if not all, patients with a clinically elevated number of









apneas actually suffer LDS, but that it is masked by two

factors. Dement et al. noted that SAS patients may deny

somnolence while literally falling asleep before the

clinician. They speculate that this results from a re-

sponse bias against acknowledging illness or a change in

subjective frame of reference about just what alertness

is. Thus EDS may be denied by patients for reasons which

may be outside the clinician's control or knowledge.

Alternatively, Block et al. (1979) propose that asympto-

matic SAS patients are suffering a subclinical level of

fallout from their frequent apneas. As years pass, in-

creasing weight or cumulative effects of desaturation

eventually lead to a full blown hypersomnolence--Sleep

Apnea Syndrome. A final possibility is that nonhyper-

somnolent subjects with frequent apneas exhibit a com-

pletely benign process totally unrelated to that found in

symptomatic SAS patients. It would seem that delineation

of the level of EDS in subjects with subclinical levels of

apnea may help clarify this issue. This notion will be

elaborated below.

The somnolence deficit which is noted in full blown

SAS patients is commonly reported. Dement et al. (1978)

describe impairments in continuous performance of

virtually any activity. Sleep Apnea Syndrome patients

reportedly fall asleep at outdoor stadiums, in front of

classes, and while treating patients. Excessive Daytime

Somnolence is described in every SAS patient reported in








experimental protocols by Weitzman et al. (1980), Fujita

et al. (1981), Zwillich et al. (1982), Garay et al. (1981)

and Sullivan and Issa (1980). At the same time, subjects

wno are asymptomatic for EDS but with the necessary number

of apneas to be diagnosed as SAS are described by Orr et

al. (1979) and Smirne et al. (1980) as well as others.

However, it should be noted that the latter studies relied

on global reports of somnolence, rather than quantified

data. Recalling the criticism of Dement et al. (1978) of

this technique as vulnerable to subjective bias, these

asymptomatic patients may have unreported or undetected

somnolence.

Excessive Daytime Somnolence appears to be frequently

concomittant with Sleep Apnea Syndrome. Thus its possible

occurrence in subclinical populations seems worthy of

investigation.

Oxygen Desaturation and Cognitive Deficits

One deficit which appears in SAS patients is oxygen

desaturation which accompanies apneas. When apnea begins,

arterial oxygen saturation begins to fall, and may con-

tinue falling throughout the event. An early technique

utilized in studying oxygen saturation levels involved

periodic blood samples drawn from arterial sources.

Birchfield et al. (1958) studied 11 normal males with a

mean age of 23. Daytime oxygen saturation in these sub-

jects averaged 96.5%. These values fell during sleep to

95.3%. Later, Orr et al. (1979) studied 8 males who had








been matched on number of apnea events. Asymptomatic

(nonhypersomnolent) subjects fell from a baseline 02

saturation of 80 mm (Hg) to a mean maximum desaturation of

54 mm (Hg) during sleep. Symptomatic subjects fell from a

baseline of 54 mm (Hg) to a mean maximum desaturation of

35 mm (Hg) while asleep. Orr et al. suggested that the

more severe desaturation which accompanied apneas in

symptomatic subjects underlaid their somnolence and other

complications. However, these conclusions are vulnerable

to certain methodological criticisms, as Block (1980)

points out. Specifically, periodic sampling may well fail

to detect the multiple brief desaturations associated with

SAS, rendering this technique potentially insensitive to

important events.

In an attempt to provide more representative data on

desaturation, Block et al. (1979) utilized an ear oximeter

in studying saturation in normal subjects. An ear oxi-

meter provided a continuous, accurate readout of moment to

moment arterial oxygen levels. Thirty males (mean age 38)

and 19 females (mean age 27) were studied. Of the 30

males, seventeen suffered at least one episode of desatu-

ration, falling from a mean saturation of 95% to a maximum

desaturation of 84% during sleep. In contrast, no epi-

sodes of desaturation were noted in the sample of premena-

pausal females (baseline 02 saturation: 96%). The oxygen

desaturations of the males were always found in asso-

ciation with breathing abnormalities or snoring. A stay








of oxygen saturation in 20 post menapausal females was

reported by Block et al. (1980). Eleven of these women

exhibited desaturation with several subjects desaturating

to less than 85%. Dolly and Block (1982) studied a group

of 17 males and 3 females with a mean age of 49 years.

These normal subjects dropped from a mean baseline satura-

tion of 96.3% to a mean maximum desaturation of 88.3%.

This evidence indicates that desaturations of 10-127 are

common in normal males and post menapausal females while

premenapausal females rarely, if ever, desaturate.

Oxygen saturation levels in SAS patients are not

commonly reported. Using an ear oximeter, Garay et al.

(1981) studied 11 SAS patients who were divided into 6

daytime eucapnics and 7 daytime hypercapnics. Eucapnics

fell from a baseline saturation of 95% to a mean maximum

desaturation of 75.5%, while hypercapnics fell from a

baseline saturation of 89% to a maximum desaturation of

62.8% (calculated from tabled data). These data were

collected during daytime "nap" studies, however, and hence

is somewhat questionable. Berry and Block (1983) recorded

9 heavy snoring males who suffered many apneac events.

Four subjects were somnolent, while 5 were not. Symptoma-

tic subjects fell from a baseline of 95% to a mean maximum

desaturation of 44%, while asymptomatic subjects desetu-

rated from a baseline of 94% to a mean maximum desatura-

tion of 85%.








Data on desaturation are interesting from several

perspectives. First of all, Sleep Apnea Syndrome subjects

seem to desaturate more heavily than normals, with Berry

and Block's symptomatic group desaturating up to 50% and

the subjects of Garay et al. desaturating about 22%,

compared with normal desaturation levels of 10-12%.

Secondly, amount of desaturation distinguished symptomatic

(hypersomnolent) from asymptomatic subjects in the data of

both Orr et al. and Berry and Block. Additionally, two

points relevant to methodology emerge. Measures of oxygen

desaturation seem crucial to understanding sleep apnea and

its effects, and ear oximetry appears to provide a proven,

valid measure of this variable.

Weitzman (1979) has suggested that the pathological

process in SAS is the hypoxia of desaturation in apneac

episodes. While cognitive and intellectual changes in SAS

have been only anecdotally reported (Guilleminault et al.,

1978), hypoxia does seem a plausible potential cause of

these changes. Thus oxygen desaturation appears to be a

reliable consequence of sleep apnea syndrome, while cogni-

tive/intellectual changes are putative sequelae.

In summary, a review of deficits found in sleep apnea

syndrome subjects suggests that cardiopulmonary complica-

tions, excessive daytime sleepiness, oxygen desaturation

and cognitive/intellectual changes are frequently asso-

ciated with this syndrome. An assessment of these

variables in a subclinical population seems indicated by








these data. Below a discussion of measurement issues and

proposal assessment of these variables appears.

Measurement of Deficits Found in Sleep Apnea Syndrome

Above it was noted that cardiopulmonary complica-

tions, excessive daytime sleepiness, nocturnal oxygen

desaturation and cognitive/intellectual changes are common

deficits of sleep apnea syndrome. A discussion of

measurement issues and proposed measurement devices for

these variables will be presented below.

Measurement of Cardiopulmonary and Health Deficits

Cardiopulmonary deficits are thought to be frequent

sequelae of sleep apnea syndrome. Thus an assessment of

blood pressure as well as self reports of hypertension and

heart trouble seems indicated in subclinical apnea sub-

jects. Additionally, a general survey of health status,

such as that provided by the Cornell Medical Incex (CiI;

Broadman et al., 1949) would also screen for other health

deficits. The CMI provides an overall score indicating

number of symptoms endorsed, as well as subscales on

several symptom categories. Separate examination of sub-

scales on respiratory, cardiopulmonary, and neurological

subscales seems appropriate. Assessment of these

variables provides a broad health screening, as well as a

detailed analysis of symptoms found in sleep apnea syn-

drome.








Measurement of Arousal Deficits (Hypersomnolence)

Excessive daytime sleepiness is a common sequel of

sleep apnea syndrome. Measurements of the sleep/wake

cycle of subclinical apnea patients thus seems indi-

cated. A sleep questionnaire assessing trait sleep habits

seems in order, as well as 1 week sleep logs following the

experimental night. The sleep questionnaire and sleep

logs have been used in ongoing research in W.B. Webb's

studies of aging and sleep and seem an adequate assessment

of sleep patterns of subclinical apnea subjects. Addi-

tionally, electroencephalographic recordings of sleep

during the experimental night seem in order using a stan-

dard recording and scoring system (Agnaw and Webb,

1972). The assessment of daytime sleepiness is relatively

new, only recently being the subject of investigation.

William Dement and associates began to develop methods of

quantifying somnolence in the early 1970s. The first

attempt produced a subjective rating scale, the Stanford

Sleepiness Scale (SSS; Hoddes et al., 1973). The SSS was

a 7 point Likert scale with endpoints anchored on 1) feel

active and vital; and 7) almost in reverie, sleep onset

soon; with 5 points with statements reflecting various

degress of somnolence in between. Patients rated their

introspective level of sleepiness on this scale.

Unfortunately, the SSS proved to be vulnerable to exactly

the same difficulties as a global rating scale used on SAS








patients, i.e., response bias and altered frame of

reference (Dement et al., 1978).

It was clear then that an objective measure of day-

time somnolence was needed. With this goal in mind,

Carskadon and Dement (1977) introduced the Multiple Sleep

Latency Test (MSLT). This technique involved measuring

the objective sleep latency during multiple nap attempts

throughout the day. In a typical paradigm, a subject

would be wired to an EEG and attempt to fall asleep at 2

hour intervals through the day. The latency to sleep in

each nap was measured by determining the latency to the

first epoch of Stage 1 sleep. In order to prevent signi-

ficant amounts of sleep accumulating during testing, sub-

jects were awakened after an epoch of sleep was

observed. If no sleep was noted, subjects were discon-

nected after 20 minutes and allowed to resume other acti-

vities. Carskadon and Dement believe that this measures

underlying physiological sleep tendency. This underlying

tendency is thought to be modulated by alerting stimuli in

the environment (Carskadon and Dement, 1982a). The vali-

dity of the test was established on normal subjects, who

exhibited a biphasic curve of sleep tendency throughout a

24 hour period. Extension of sleep over typical lengths

in normals (who are thought to be moderately chronically

sleep deprived) saw a significant rise in average sleep

latencies on the MSLT (Carskadon and Dement, 1979b), while

chronic sleep restriction (5 hrs sleep per night) caused








significant declines in average sleep latencies (Carskadon

and Dement, 1981a). Finally, total sleep deprivation led

to a drastic decline in MSLT values, with averages

plummeting below one minute (Carskadon and Dement,

1979a). Mean MSLT values showed recovery to baseline

values after a full night's sleep for partially (5 hrs)

deprived subjects, and recovery after two nights of ad lib

sleep for totally sleep deprived subjects (Carskadon and

Dement, 1979; 1981a). This body of evidence suggests

that the MSLT is a reliable and valid index of sleepiness

in normal subjects.

With its internal validity established, the MSLT was

applied to somnolent populations. Dement et al. (1978) as

well as Richardson et al. (1978) demonstrated that SAS and

narcoleptic patients have significantly shorter latencies

to sleep on the MSLT than normal subjects do. Hartse et

al. (1979) found SAS subjects to have an average latency

to Stage 1 of 2.5 minutes, while narcoleptics had a mean

of 3.2 minutes. Both these latencies were significantly

lower than those of a miscellaneous control group whose

latencies were about 11.2 minutes. Hartse et al. (1980)

found latencies to Stage 1 sleep as follows for various

patient groups: Normals 12.2 m, Insomniacs 16.2 m, SAS

2.5 m, and Narcoleptics 3.2 m. The SAS and narcoleptic

patients had significantly shorter latencies to Stage 1

sleep than insomniacs or normals. Roth et al. (1980)

compared 10 SAS patients with 10 age matched controls.








The normals averaged 20.4 m to Stage 1 sleep while the SAS

patients averaged 1.9 m. Zorrick et al. (1982) found mean

MSLT latencies to Stage 1 sleep of 3.1 m for SAS subjects

and 2.9 m for narcoleptics, significantly shorter than for

patients with psychiatric disorders.

The chief difficulty remaining in judging the MSLT

"reliable and valid" is external validation demonstrating

that exceeding a certain mean latency is correlated with

other deficits of performance. In some ways, low laten-

cies on the MSLT may suffer the same limitations as sleep

deprivation protocols; that is, demonstrating serious

deficits is surprisingly difficult. At the same time, few

would deny the subjective unpleasantness of sleep depriva-

tion, which may be applicable to low MSLT scores. Thus

the MSLT appears indicated in assessing excessive daytime

sleepiness while the Stanford Sleepiness Scale (SSS) pro-

vides an index of the subjective experience of the sub-

ject. Application of both these measures to a subclinical

apnea population would seem to adequately assess daytime

sleepiness in these subjects.

Measurement of Hypoxia and Its Sequelae

Noctural desaturation is frequently noted in SAS

patients. Therefore, it seems in order to examine the

known sequelae of hypoxia of other etiologies in an

attempt to determine variables or behaviors likely to be

sensitive to the hypoxia found in SAS and SAA. Research

on hypoxia varies systematically along two dimensions,









human vs. animal subjects and acute vs. chronic hypoxia.

First, the data available on animals will be reviewed,

followed by that extant on human subjects. Finally,

behavioral variables sensitive to changes in these

variables will be summarized.

The effects of chronic hypoxia on animals have not

been exhaustively studied (Davis, 1975). One project

which did examine this variable involved exposure of rats

to up to 36 hours of 10% oxygen. Davis (1975) observed

the behavior of these animals and concluded that they

appeared normal. Although hyperventilation was noted, no

permanent effects of the hypoxia were found. Hanbauer et

al. (1981) examined the effects of 10% 02 on rats exposed

for up to 4 weeks. They found that relatively short term

(2 days) hypoxia resulted in an increase in dopamine (DA)

content in the carotid body. An increase in norepine-

phrine (NE) content was noted after 1 week of exposure.

As the carotid bodies are implicated in the control of

respiration, the authors concluded that different

mechanisms are operative in the adaptation to short and

long term hypoxia. Unfortunately, little other data have

been reported on the chronic effects of hypoxia on more

complex functioning. This is particularly problematic as

it are these functions which are thought to be affected in

humans exposed to hypoxia.

Certain aspects of acute hypoxia in rats have been

carefully examined. Gibson et al. (1981) reviewed the









evidence which has been collected on neurotransmitter

alterations in hypoxia. In brief, mild to moderate

hypoxia, cerebral levels of ATP are normal, implying that

energy supplies of neurons are not disturbed. In con-

trast, the turnover of certain neurotransmitters is

altered. Gibson et al. note that oxygen is an important

substrate for the synthesis of several neurotrans-

mitters. When rats are exposed to acute hypoxia, syn-

thesis of dopamine and serotonin is decreased, while

absolute levels of tryosine, tryptophan, catecholamines,

and serotonin remain constant. Continuing hypoxia leads

to an adaptation resulting in normal turnover of these

neurotransmitters. Gibson and Blass (1976) examined the

effect of acute hypoxia on central acetylcholine (ACh).

As in dopamine (DA) and serotonin (5-HT), acute hypoxia

results in a decrease in synthesis in ACh, but not in the

absolute level of ACh. Thus there is a substantial body

of evidence indicating changes in central neurotransmitter

levels in hypoxia. Generalizing from the animal data to

human subjects, Gibson and Duffy (1981) suggest that these

changes underlie the behavioral deficits of hypoxia in

humans (which will be reviewed below).

Other variables have been examined in acute hypoxia

in animals. Annau (1972) found that rats exposed to brief

hypoxia exhibited decreased appetite, thirst, and self

stimulation rates. Annau observed that brief hypoxia

depressed all behavioral variables. Gellhorn (1951) found








that brief hypoxia of 4-7.5% 02 abolished cortical re-

sponses to auditory stimulation. Sara (1974) trained rats

in active one way avoidance tasks. Immediately after

training, subjects were exposed to 3.5-4.0% oxygen. Rats

tested at 1 and 3 hours post hypoxia avoided correctly,

while those tested at 6 and 24 hours did not, leading Sara

to suggest that hypoxia resulted in a memory retrieval

deficit. Freides and Allweise (1978) also trained rats in

one way avoidance and exposed them to 2.0% hypoxia imme-

diately following training. These animals avoided at 12

and 200 minutes post hypoxia, but not at 90 minutes.

Friedes and Allweise suggested that a sensitive period for

hypoxia effects on training existed for 15 minutes after

learning. Tauber and Allweise (1975) followed the same

training/hypoxia paradigm and found avoidance impaired at

2.5 but not 4.5 hours. They suggest that hypoxia inter-

feres with a medium term memory. While some of these data

are conflicting, it appears that acute hypoxia depresses

certain behavioral variables and interferes with aspects

of memory processes.

While caution is in order when generalizing from

animals to humans, several seemingly relevant points

emerge from the review. Behaviorally, simple activities

may be spared, although some depression of appetitive

activities may occur, while more complex behavior depen-

dent on memory processes may be altered. Of particular

interest are the demonstrated alterations in









neurotransmitter turnover seen in hypoxia. It has been

suggested that these subtle alterations may underlie the

changes in complex behavior seen in hypoxic humans.

Data on human performance after hypoxia have neces-

sarily been reliant on fortuitous or weak manipulations.

Although this may limit the usefulness of these data, it

is at the very least highly suggestive of variables sensi-

tive to hypoxia. Van Liere and Stickney (1963) note that

acute hypoxia leads to confusion, headache, drowsiness,

weakness and incoordination. After effects include head-

ache, nausea and emotional liability. Richardson et al.

(1959) found that acute, fatal hypoxia from surgery

results in damage to cortical neurons, particularly in

layers III and IV. Susceptible subcortical neurons in-

clude those of the corpus striatum and cerebellum. Plum

et al. (1962) describe a syndrome of delayed postanoxic

encephalopathy. After the acute coma resolves in 4-5

days, a brief period of normal functioning returns.

Increasing irritability and confusion then emerge, with

loss of coordination and memory, and diminished attention

span. Typically, the lesion spares grey matter while

affecting white matter. Devereaux and Partnow (1975)

describe a patient who recovered from a delayed acute

encephalopathy. Although IQ was relatively spared, severe

dysarthria emerged. Relevant points which emerge from

studies of acute hypoxia in humans suggest that sublethal









levels of hypoxia affect coordination, memory, attention

span, and may result in headache, drowsiness, and nausea.

The effects of milder, chronic hypoxia are summarized

by Gibson et al. (1981). At the oxygen pressure equiva-

lent of 5,000 feet, impaired dark adaptation is noted. At

10,000 feet decreased concentration, hyperventilation, and

short term memory deficits are noted. By 15,000 feet,

euphoria and loss of both coordination and critical judge-

ment are observed. Christenson et al. (1977) found signi-

ficantly less visual signals detected at 17% 02 than at

21,%, suggesting reduced vigilance. McFarland (1937)

studied men in ascents to high altitudes and concluded

that subjects with slower rates of ascent performed better

on unspecified psychological tests. West (1984) reports

data collected on an expedition to Mt. Everest. Results

of a neuropsychological battery indicated deficits in

finger tapping, verbal fluency, verbal learning, short

term memory, and expressive language. Thus the general

affects of chronic mild hypoxia are thought to be deficits

in dark adaptation, short term memory, critical judgement,

motor coordination, verbal fluency, and vigilance.

One group of chronically hypoxia patients, those with

chronic obstructive lung disease (COLD), has been rela-

tively well studied with sensitive neuropsychological

instruments and is deserving of a more painstaking

review. Krop et al. (1973) studied COLD patients before

and after they received oxygen supplements. The WAIS,









Wechsler Memory Scale (WKS), Bender-Gestalt, Bender

Gestalt Interference Procedure, Facial Recognition and

Finger Tapping tests were administered to COLD and control

subjects. Subjects receiving supplementation improved

significantly more than controls on WAIS Full Scale IQ,

Performance IQ, Wechsler Memory Quotient, both Bender

procedures and Finger Tapping. Grant et al. (1982)

studied another group of COLD patients. They administered

the Halstead-Reitan Battery, Aphasia Screening Test,

Trailmaking Test, Russell modification of the Wechsler

Memory Scale, Tactile memory test, and the WAIS. In COLD

subjects, deficits were found relative to controls on all

variables except Reitan Rhythm, Aphasia Screening and WNS

Logical stories. Grant et al. concluded that COLD

patients exhibited deficits on a global impairment rating,

attention, abstracting ability, complex perceptual motor

tasks, simple sensory and motor tasks, and memory tasks.

Low but significant negative correlations were reported

between several measures of oxygen saturation and both

Halstead Impairment Index and Global Impairment Rating.

Grant et al. speculated that floor effects resulted in low

variability in PA02 levels among COLD patients and kept

these correlations low. They further suggested that the

neuropsychological deficits observed in these patients

resulted from the oxygen want found in these subjects.

Thus data available on humans indicate that dark

vision, memory, motor coordination, critical judgement,








verbal fluency, and signal detection are affected by acute

hypoxia. Animal data suggest that sensitivity of memory

processes to hypoxia, while data from chronic hypoxia

humans indicate that attention, vigilance, abstracting

ability, complex perceptual motor skills, as well as motor

and sensory abilities are also impaired by hypoxia. While

none of these findings are unequivocally directly

applicable to the brief, acute, desaturation occurring

during apnea, they are, at the very least, indicative of

abilities sensitive to hypoxia in humans.

A neuropsychological battery sensitive to these

abilities includes Wechsler Adult Intelligence Scale

(WAIS; Wechsler, 1955); Wechsler Memory Scale (WMS;

Wechsler, 1945) including the Russell delayed memory

aspects (Russell, 1975); Rey-Osterreith Complex Figure

(Osterreith, 1944); Hooper Visual Organization Test

(Hooper, 1958); Verbal Fluency Test (Halstead, 1947);

Wisconsin Card Sort (Berg, 1948); and the Finger Tapping

Test (Halstead, 1947). Administration of this test

battery provides a broad screening of neuropsychological

functioning, with a particular emphasis on those abilities

known to be sensitive to hypoxia. Application of this

battery to subclinical apnea subjects provides a sensitive

indicator of possible hypoxic brain dysfunction.

In summary, alterations of sleep/wake cycles appear

to be present in sleep apnea syndrome patients. Thus

assessment of these variables in a subclinical population









is indicated. An adequate description of the sleep/wake

cycle would seem to include sleep questionnaire, sleep

log, overnight EEG, multiple sleep latency test, as well

as several Stanford Sleepiness Scale ratings.

Statement of the Problem

Sleep Apnea Syndrome is a clinical entity with

seemingly serious consequences. Although the nocturnal

apneas found in this illness are thought to be the cause

of the associated deficits, the identification of

seemingly asymptomatic subjects with a number of apneas

comparable to those found in SAS throws doubt onto this

formulation. Despite this evidence, the recent discovery

tnat apneas occur in significant numbers in the general

population has led to speculation that this subclinical

level of apnea may be the cause of various pathological

processes. At present, there is little evidence to

support or refute this notion.

The present report evaluates this issue through a

detailed study of a group of subjects thought to be at

risk for sleep apneas. Aging, heavy snoring males were

recruited, and a comprehensive analysis of their respira-

tory, health, neuropsychological, and sleep/wake cycle

status was carried out. Thus potential relationships

between respiratory disturbances and variables known to be

disrupted in sleep apnea syndrome patients could be

evaluated. This approach also allowed exploration of the





47



possible risk factors associated with subclinical sleep

apneas.















CHAPTER TWO
METHOD

Subjects

Heavy snoring males were recruited via newspaper

advertisements, notices, and phone calls to a list of

aging subjects reported by Webb (1982). Participants were

required to be male, self reported heavy snorers, over 30

years of age, and in general good health (i.e., self

described healthy and normal and not under the active care

of a physician for illness). Subjects with a history of

head trauma or alcoholism were excluded. Potential

subjects who met these criteria were offered a chance to

participate in a "snoring study" in which they would

complete testing during an evening, sleep overnight in a

lab with several physiological parameters recorded, and

complete one week's worth of "sleep logs," for a payment

of fifty dollars. Volunteers were scheduled on a first

come, first served basis. A total of 60 subjects were

studied, with 9 cropped because of technical problems with

their overnight recordings and 5 excluded because of

health problems. Thus 46 subjects represented the final

sample. These 46 subjects had a mean age of 49.9 (sd

13.1) and represented a wide cross section of the

population, ranging from the unemployed to university








professors. All subjects completed and signed an informed

consent agreement.

Apparatus

During the evening testing, all subjects received a

blood pressure check with a standard hospital cuff. The

measurement was made from a seated position. Overnight

recordings were made in a quiet, darkened room with the

subject sleeping on a standard hospital bed. All noc-

turnal physiological recordings were routed through a

Grass Polygraph (model 7D). Electrodes for electro-

encephalographic recordings were mounted in three pairs

using sites F2/F8, PZ/T6, RE/LE from the international

10/20 system of placement. These electrodes were affixed

to the scalp with collodion soaked gauze pads dried with

an air hose. Respiratory measures included measurement of

chest and abdomen wall movement derived from impedence

measures monitored from surface electrodes (3M Ag/AgC1

#2246) placed on the lower chest and just above the

navel. The signal from these electrodes was passed

through an impedence convertor and onto the polygraph.

Oral and nasal airflow was recorded by thumistors (TCT-1R

Transducer-Grass) clipped to the respective orifices.

These signals were also routed through the polygraph.

Electrocardiograms were monitored from modified bipolar

chest leads (MCL2). Finally, blood oxygen saturation was

continuously monitored from a probe attached to the ear

lobe, and analyzed by a Biox ear oximeter (Ilodel IIA BTA









Co). All physiological variables were recorded on the

polygraph's chart paper which was run at a speed of 10

mm/s.

Measures

Two classes of independent variables are identified,

with the first being thumistor variables (those derived

primarily from the nose and mouth thumistors). These

included number of apneas and hypopneas, mean duration of

apneas and hypopneas, mean low oxygen saturation in apneas

and hypopneas, mean oxygen saturation change in apneas and

hypopneas, and total seconds spent in apneas and hypo-

pneas. Thumistor variables were separately scored by two

trained raters, whose agreement appears in Table 2-1. An

overall agreement of 85% was achieved. Disagreements were

resolved by consensus scoring. The second class of

independent variables included oxygen saturation variables

(derived from a minute by minute rating of highest and

lowest oxygen saturation from the overnight record).

These variables included mean highest saturation, mean

lowest saturation, and number of desaturations of 4 ana

10% or more. In selecting these scoring criteria, several

considerations were relevant. Initially, a system with

reasonable interrater reliability was necessary. While

published criteria for scoring apneas were not problem-

atic, reliable scoring of hypopneas was not achieved until

a desaturation criterion of 10% was introduced. In

selecting measures of oxygen saturation, a mean high and










Table 2-1. Interrater agreement for respiratory events of
46 snoring males.


Consensus Initial Disagreements
Scoring Agreements #1 #2

(178) -
91 90 4 6
78 76 1 1
61 52 3 8
57 56 9 2
32 29 3 3
28 27 0 2
25 25 0 0
19 16 4 0
17 17 0 0
15 16 1 0
13 13 0 0
7 7 0 0
S44 0 0
4 4 0 0
3 2 0 1
3 3 0 0
3 0 7 2
3 3 0 0
3 3 0 0
1 1 0 0
1 1 0 0
1 1 0 0
1 1 0 1
1 0 0 1
1 1 0 0
1 1 0 0
1 1 0 0


474(652) 450


Note


Scored by
Consensus
Consensus
Consensus
Consensus
Consensus
Consensus


consensus
added 1
added 2
added 9
added 1
added 3
added 1


Consensus added 3

Consensus added 1




Consensus added 1

Consensus added 3






Consensus added 1


37 56


Note: 28 of
hypopnea.


the 46 subjects experienced at least 1 apnea or









low score seemed obvious. In order to assess the quan-

titative "impact" of numerous desaturations, a sum of 4%

and 10% desaturation was tabulated. These selected

parameters for thumistor and saturation variables allowed

reliable scoring and a description of several aspects of

nocturnal respiration.

Dependent measures fell into three broad classes:

measures of sleep/wake status, measures of health status,

and measures of neuropsychological status. Measures of

the sleep/wake cycle included one week's sleep logs

(derived variables included number of reports of daytime

sleepiness, hours asleep, estimated sleep latency, number

of wakenings after sleep onset, and minutes of waking

after sleep onset), a sleep questionnaire (derived

variables including hours of sleep usually obtained,

number of naps, hours napping, quality of sleep,

restfulness on awakening, depth of sleep, and amount of

daytime sleepiness), the mean of the several Stanford

Sleepiness Scale ratings, the two multiple sleep latency

tests (derived variables included mean sleep latency and

total minutes asleep in both naps) and the overnight LEG

variables derived from the scoring system of Agnew and

Webb (1972). Electroencephalographic variables included

time in bed; pure sleep time; sleep efficiency index;

number stage 0 periods; time 1 stage 0; sleep latency;

time % stages 1, 2, 3, 4 and REM; latency first hLM









period; mean REM period; mean REM cycle length; $ slow

wave sleep; and number of stage changes.

Measures of health status included height, weight,

diastolic and systolic blood pressure, as well as the

Cornell Medical Index. From the CMI, reports of hyper-

tension and heart trouble were noted, as well as the total

number of symptoms endorsed. Additionally, the scores for

the scales tapping respiratory, cardiac, and neurological

symptoms were tabulated separately.

Measures of neuropsychological status included the

WAIS Verbal and Performance IQ's, the WMS Memory Quotient,

delayed recall of the logical stories and the visual re-

production subtests of the WMS, delayed recall of the Rey

complex figure, number of words generated in the verbal

fluency test, number of correct sorts of the Wisconsin

Card Sort, mean number of taps for the right and left

handed finger tapping tests, and lastly the number of

correct trials on the Hooper test.

These neuropsychological tests were chosen with two

aims. First, an attempt was made to sample from each of

several loosely bounded areas of cognitive skill. These

included intelligence (both verbal and non-verbal),

memory, both immediate and delayed (verbal and non-

verbal), visuo-perceptual/organizational skills, language,

and frontal self-regulatory skills. A second emphasis was

on a relatively detailed analysis of areas previously

shown to be sensitive to hypoxia. These sensitive areas









included memory, visuo-organizational skills, verbal

fluency, and motor coordination. The chosen neuropsycho-

logical battery reflected these two aims. Thus the WAIS

tapped intellectual functioning; the Wechsler Memory

Scale with delayed recall of logical stories and visual

reproductions as well as digit span and Rey Figure heavily

sampled verbal and non-verbal memory; the Hooper tapped

visuo-organizational skills; Verbal fluency sampled lan-

guage skills; the Wisconsin card sort tapped frontal

self-regulatory skill; and Finger tapping sampled motor

functions. The chosen battery provided a broad screening

as well as a detailed assessment of neuropsychological

functions thought to be impaired by hypoxia.

Procedure

Subjects followed the schedule appearing below for

the experimental night:

1e00 Sign informed consent, blood pressure,

height/weight, finger tapping, verbal fluency,

subjective sleepiness (SSS-made every 1/2 hour

till bedtime)

1830 Wiring for EEG

1900 Begin MSLT I (20 m)

1930 WAIS

2000 WMS, Hooper

2030 Graphesthesia, Luria motor programs

2100 Begin MSLT II (20 m)

2130 Rey figure, Wisconsin card sort









2200 Sleep Questionnaire, Cornell Medical Index,

remaining testing

2230 Wiring for thumistor, EKG, respiratory effort,

ear oximeter

2300 Lights out, begin recording

A night technician remained with the subjects throughout

the night. Subjects usually awakened between 5 and 6 am,

and later completed the one week's sleep logs.

Statistical Procedures

Several statistical procedures were utilized in

evaluating the present data. One of the primary

approaches was correlational, considered appropriate

because of the exploratory nature of the study. Bearing

in mind the possible alpha inflation of large corre-

lational matrices, other approaches were also utilized to

add confidence. Thus the sample was stratified by level

of apnea/hypopnea, and oneway ANOVA procedures were used

to evaluate between group differences. Additionally, non-

parametric procedures were utilized where appropriate.

Positive correlational results were considered against

possible replication with other measures in the study, and

in some cases multivariate procedures were addea for

further confidence. It is conceded that correlational

approaches may carry the burden of possible spurious

results, but it is believed that the various additional

procedures outline above lend confidence to the findings.















CHAPTER THREE
RESULTS

Demographics and Incidence of
Nocturnal Respiratory Disorder

Subjects were selected to fulfill several sampling

requirements, including male sex, presence of heavy

snoring, and self report of general good health. A total

of 60 subjects were studied. Five subjects were later

excluded when questioning revealed the presence of chronic

alcoholism (2) and serious head trauma (3). Additionally

9 overnight records were rejected because of missing or

unusable data. The characteristics of the final 46

subjects are detailed in Table 3-1. These snoring males

had a mean age of approximately 50 years, a mean weight of

approximately 190 pounds, and a mean educational level of

15 years of schooling.




Table 3-1. Demographic variables from 46 snoring males.



K SD Range

Age 49.9 13.1 30-75
height (Ibs) 189.3 34.9 125-250
Weight:Height Ratio (lbs:inches) 2.7 0.5 1.9-3.6
Eaucation (years) 14.7 2.5 9-20









The respiratory records were annotated by two hour

intervals following bedtime. From a total of 652 events

scored, 150 (23%) occurred in the first two hours

following bedtime, 185 (23%) occurred in the second two

hours, 232 (36%) occurred in the third two hours, while 83

(12%) occurred in the final two hours. It should be noted

that wakeup times varied considerably, making interpre-

tation of the attenuated number of events in the final two

hours somewhat problematic.

From a total of 578 events in which reliable sleep

staging was achieved, 459 (78%) occurred in light slow-

wave sleep (stages 1-2), while 118 (20%) began in REi

sleep. Only 6 events (01%) occurred in slow wave sleep.

These precentages must be considered in light of the

relative distribution of sleep stages in these subjects:

stages 1-2 (64%), stage REM (12%), stages 3-4 (10%), stage

0 (14%). This distribution suggests that the sleep time

spent in stages, 1, 2, or REM must have provided the

overwhelming preponderance of events in these stages.

Sleep characteristics will be discussed in greater detail

below.

As noted below, 28 of the 46 subjects experienced at

least one episode of apnea or hypopnea. Table 3-2

presents the data on occurrence of at least one

apnea/hypopnea, as well as the occurrence of a high level

of apnea/hypopnea. Almost 2/3 of the sample had at least

one apnea or hypopnea, while 13% had high levels of









Table 3-2. Incidence of low and high levels of
apnea/hypopnea by age in 46 snoring males.



Low Apnea/ High Apnea/
Age Group N Hypopnea Hypopnea

30-39 14 .71 0
40-49 7 .57 .29
50-59 11 .72 .27
60+ 14 .43 .07
Overall 46 .62 .13



apnea/hypopnea. Note that subjects with high levels of

apnea/hypopnea are included in the overall total of

subjects with at least one event for the following

analyses. The classification of "high" levels of

apnea/hypopnea utilizes the criterion of an apnea +

hypopnea index greater than 5, a frequently used clinical

cut-off score. The table also presents the frequency of

respiratory distress broken down by age. No simple age

related trend is apparent, and a X2 analysis of the two

factors of age and respiratory distress was not signi-

ficant for those with at least one event (X2=3.2; p=NS),

or for those with high levels of apnea/hypopnea (X2=5.9;

p=KS).

A further investigation of possible age by level of

apnea activity involved division of the sample into those

above ano below 60 years of age crossed with those above

and below 5 events/hour. The X2 value for this comparison

was also nonsignificant (X2=0.6; p=NS). Subjects were









also divided into those above ano below the mean weight:

height ratio (2.7) and crossed with those with and without

at least one event or high levels of apnea/hypopnea. Both

these non-parametric tests were also non-significant

(X2=.01; p=lS; X2=1.5; p=HS).

lablt 5-3 utilizes Pearson correlations to explore

possible relationships between demographic variables (age,

weight, weight:height ratio) and nocturnal respiratory

indices (apnea index, apnea + hypopnea index, seconds in

events, mean hiFh saturation, mean low saturation, number

of desaturations >_4, and number of desaturations >10Q).

Again, age displays little relationship with respiratory

indices, but substantial correlations between weight and

various respiratory indices are apparent.




Table 3-3. Significant (p<.05) Pearson correlations between
nocturnal respiratory variables ana demographic
variables in 46 snoring males.



Respiratory Weight:Height
Variable Ape Weieht Ratio

Number of Apneas
Lumber of Apneas+Hypopneas .265
Apnea Index
Apnea+hypopnea Index .260 .506*
Mean High Saturation -.323* -.363*
Mean Low Saturation -.397* -.460*
Mean Saturation Change .319* .373*
Number of Desaturations >4 .267 .368* .450*
lumber of Desaturations >10


*p<.01









In summary, non-parametric analyses failed to reveal

significant relationships between age or weight and

overnight respiration. However, Pearson correlations

suggested that increasing weight is associated with

deterioration of several noctural respiratory indices.

Because increasing apnea/hypopnea levels have been

suggested to carry increasing risks for various deficits,

the sample was stratified by level of apnea/hypopnea. The

first group included those subjects without events (No

events), those with at least one event but less than 5

events per hour (Low apnea/hypopnea), and those with 5 or

more events per hour (High apnea/hypopnea). It should be

noted that this division of subjects is slightly different

from the earlier non-parametric analyses, because in this

grouping low apnea/hypopnea and high apnea/hypopnea are

mutually exclusive, while in the earlier analyses the

group with at least one apnea/hypopnea included those with

more than 5 events per hour.

Table 3-4 utilizes oneway ANOVA procedures to compare

these groups on demographic and nocturnal respiratory

variables. There were no significant between group

differences on any of the demographic factors: age,

weight, weight:height ratio, and education [F(2,43)=2.1,

0.6,0.8,0.7; p=lIS]. Logically, both apnea and hypopnea

indices were significantly different between groups

[F(2,43)=9.6, p<.CO04; 20.5, p<.0000]. Followup

comparisons (Scheffe) indicated that the high






61
















0 (U CU a N n m

v00 i- 0 -- .
O <0 Z NX
CO Ov


C.>
CO 0 CL ) o OaN




0. ..





C O*
o >> < Qr co O











10.
200.


O "
o 0 m IN- CU .~



aC O O ..
OU 0 JO m C

CCO C0




C UC** O**


SH 0)
SE -- 3 m o T- NOC~ro O^^






C D
3J O N
Co ( I C J O-(1J'30 0.U






0 0 0L00 0 c\ 0 r
-> Q CO N 00


C 0 0

CU La a
Dc t co It
Ei > a fu cu O -H l



CU 0 Cd


v 0 L 0 I0. 0 0 0 HU
0 c >Q. H -(




9, Vm "0 u 0 cu 0.'

00. CU C w C0 OU Z o o3
Ili>10 C: CUQu )m L. (U CUE-P
,u Q) a co w w m 0 M 4J-

0. a X 4 C 00 00c
*u-H 00D0. UP)
a)o 0 z Qa o o 7H










w Q -1.0'*-H-aOO Q -41 4-j
00 c0 > a 0 OO -H-H
*H 'a CL 0 O CCQ
*4 -L +*C to (0
c a'o W0W c U C ru)C C c I

C 0E-H-H=E(ECCQ C, WXCU E E
< 00 ) 400 0 C: oo o t o oo C, r









apnea/hypopnea group was significantly different from the

other two groups on these two indices. While mean

duration of apneas and hypopneas was not significantly

different between the high and low apnea groups

[F(1,25)=3.5, 0.1; p=NS], total seconds in respiratory

events was [F(2,43)=50.4, p<.0000], with followup

comparisons showing that the high apnea/hypopnea group was

significantly different from tne other two groups.

Turning to the oxygen saturation data, mean high

saturation failed to reach significance [F(2,43)=0.3,

p=NS]. However, mean low saturation, number of

desaturations >4% and >10% were all significant

[F(2,43)=11.6, p<.0001; 7.0, p<.C02; 21.4, p<.0000], and

followup comparisons showed that the group with high

levels of apnea/hypopnea was significantly different from

the other two groups.

Thus, division of the sample into levels of

apnea/hypopnea resulted in groups which were not

significantly different on demographic variables including

age, weight, weight:height ratio, and education. However,

the group with high levels of apnea/hypopnea had

significantly more nocturnal events, time in events, and

oxygen desaturation.

In summary, the present sample of 46 snoring males

appears representative of those predisposed to experience

sleep disordered breathing (middle-aged, snoring males).

Sixty-two per cent of these subjects had at least one









apnea or hypopnea, while 135 had high levels of

apnea/hypopnea. Most events occurred in light slow-wave

or REM sleep. No effect of age on level of sleep

disordered breathing was observed, although increasing

weight was correlated with nocturnal respiratory events.

When the sample was stratified by level of apnea/hypopnea

activity, subjects with high levels of events had

significantly more events and desaturations than the

remaining subjects, although no significant differences

were observed on age, weight or education.

Nocturnal Respiratory and Health Variables

Health variables including blood pressure readings

and several scores derived from the Cornell Medical Index

(overall number of symptoms endorsed, as well as number of

symptoms endorsed from the respiratory, cardiopulmonary

and neurological subscales) were evaluated initially

through a matrix of Pearson correlations which included

the respiratory variables (apnea index, apnea + hypopnea

index, seconds in events, as well as mean high and low

saturation and number of desaturations >4% and >10%).

From a total of 42 correlations, three were significant at

p<.05 or less: Systolic blood pressure and mean low

saturation (r=-.261); Systolic blood pressure and number

of desaturations >4% (r=.373); Overall number of symptoms

endorsed on the CMI and mean high saturation (r=-.254).

As a further exploration of possible relationships between

blood pressure and apnea/hypopnea activity, subjects were









asked whether they had ever been diagnosed with hyper-

tension. Thirty-three per cent of those with no events

had diagnosed hypertension, as did 33% of those with low

levels of apnea/hypopnea, while only 15% of those with

high levels of apnea/hypopnea had diagnosed hyperten-

sion. These data do not support an increasing level of

hypertension in subjects with high levels of

apnea/hypopnea.

The subjects were grouped into the earlier delineated

levels of apnea/hypopnea. Table 3-5 presents the means

and standard deviations of health variables from the three

groups. No dramatic trends are apparent in this table,

and oneway ANOVA procedures confirmed this observation

with no significant differences for: Diastolic BP,

F(2,43)=.01, p=NS; Systolic BP, F(2,43)=.1, p=NS; CMI

overall, F(2,42)=.2, p=liS: CMI respiratory, F(2,42)=.5,

p=NS; CMI cardiopulmonary, F(2,42)=.5, p=NS; CMI

neurological, F(2,42)=2.2, p=NS.

In summary, while correlational procedures suggested

relationships between several oxygen saturation measures

and blood pressure as well as overall number of symptoms

endorsed on the CMI, oneway ANOVAs failed to reveal

significant between group differences when subjects were

stratified by level of nocturnal distress.









Table 3-5. Means and standard deviations for health related
variables in 46 snoring males grouped by level
of apnea/hypopnea.



No Apnea/ Low Apnea/ High Apnea/
Hypopnea Hypopnea Hypopnea

Systolic BPa 129.1(14.2) 128.7(12.6) 127.0(7.5)
Diagtolic BP 83.2(13.4) 83.9(11.0) 81.6(7.6)
CMIU Overall Score 14.7 (7.4) 16.1(10.5) 13.8(7.4)
CMI Respiratory Score 1.4 (1.2) 1.1 (1.8) 2.0(1.4)
CMI Cardiac Score 1.6 (1.3) 1.2 (1.2) 1.0(1.2)
CMI Neurological Score 1.2 (0.9) 1.6 (1.5) 0.8(0.8)

SBP = blood pressure.
b CMI = Cornell Medical Index.


Nocturnal Respiratory and Sleep/Wake Variables

EEG Data

Electroencephographic variables were recorded

overnight for all subjects. Records were scored by a

trained technician who utilized the system of Arnew and

Webb (1972) to determine sleep stages. In three records,

EEG tracings were judged inadequate for differential sleep

staging. However, a judgement of sleep vs. wake was made

for these subjects in order to form a basis for the

respiratory indices. These three EEG records were

excluded from the following analyses, leaving 43 records

as a base for analysis.

Fifteen variables (time in bed, pure sleep time,

sleep efficiency index, number of stage 0 periods, time E

stage 0, sleep latency, time % stages 1, 2, 3, 4, and RLM,

latency 1st REM period, mean REM period length, mean REM









cycle length, 0 slow wave sleep), reflecting multiple

aspects of the sleep of these subjects, were derived from

the sleep stage scoring. Table 3-6 presents the means ana

standard deviations for these subjects. Although subjects

spent over six hours in bed, the table indicates that not

quite 5 hours were spent asleep. The remaining variables

suggest that an increase in sleep latency, and time awake,

is responsible for the diminished sleep time, while

lighter sleep is also increased.



Table 3-i. e!eans and standard deviations of selected sleep
variables from 43 snoring males.



Standard
Mean Deviation

Time in Bed 377.3 45.1
Sleep Latency 29.2 32.9
Pure Sleep Time 288.7 75.8
Sleep Eificiency Index .76 .16
Number Stage 0 Periods 8.6 5.6
Time ? Stage 0 14.3 12.7
Time % Stage 1 3.3 2.8
Time 4 Stage 2 60.8 10.3
Time % Stage 5 3.4 2.4
Time % Stage 4 7.9 6.4
Time % Stage REN 11.5 7.2
Latency 1st HEI Period 124.7 92.2
Mean REIi Period Length 14.9 9.4
RL! Cycle Length 74.9 57.7
% Slow Wave Sleep 12.9 8.3


Note: see text for explanation of N.


Partial correlations, controlling for age, were

calculated between these sleep and the respiratory

variables. Out of a total of 105 correlations, only one









was significant at p<.05 or less: Number of desaturations

>4% vs REM latency (r=-.369).

Table 3-7 presents the sleep variable means for the

subjects grouped by level of apnea/hypopnea. The group

with high apnea/hypopnea appears to spend more time awake,

less time asleep, and achieve less slow wave sleep than

the others. However, none of these differences proved

significant. In fact, only one out of 15 oneway ANOVAs

achieved significance: Time in bed, F(2,40)=0.4, p=NS;

Sleep latency, F(2,40)=1.7, p-iS; Pure sleep time,

F(2,40)=0.3, p=IS; Sleep efficiency index, F(2,40)=0.2,

p=nlS; Number of stage C periods, F(2,40)=0.5, p=NS; Time

5 stage 0, F(2,40)=1.3, p=NS; Time % stage 1,

F(2,40)=1.3, p=NS; Time % stage 2, F(2,40)=0.04, p=NS;

Time % stage 3, F(2,40)=3.64, p<.04; Time % stage 4,

F(2,40)=1.32, p=NS; Time % stage REM, F(2,40)=1.5,

p=NS: Latency 1st REM, F(2,40)=0.8, p=NS; Mean REM

period length, F(2,40)=0.3, p=NS: Mean REM cycle length,

F(2,40)=0.9, p=NS; slow wave sleep, F(2,40)=2.4,

p=NS. Scheffe followup procedures for time % stage 3

revealed no significant between group comparisons.

To summarize, EEG sleep measures indicated lighter

sleep then normal in this sample, with increases in

awakenings and light slow wave sleep observed. Neither

between group comparisons nor correlational procedures

revealed significant interactions between sleep and

nocturnal respiratory variables.






68














0 O ^^-LnOOrnri^-

C7 1m1;N2rNtavNC



o 0c f













to . .
*r41 cu^





H nUN O O =3 CNc CO -cz
0. . .

0(U n N n C- N K C- %D Z C -' C

ID (U. c 0 X
co 0 1-J N m nL-
C O3 Q 0 * ., .


- 4q
u0 0








0. C
C. -
.cu Q

4-1 a)
au ra 'Z
o 'oQo H-t-- c )i o









cu w-


C.;F=r0 N 0t
10 N1 a
'0 00 O4- 0






000 ti Oo
)o a c0 4 -

0W > *.
a0 C 004 P 0+

0a- d 0 J2 0



0W"ICCO 0nCiQ)0 Z (U


) a C0 0C *0
i 00000000000C
ac C) ma0aaa )w











*P(U -102 0 0E) 0
1; .c, E EE F ao



n E aUCLWZ) E- E-4 E L, oZ C: H
*o~~~~~~z t: u a i im iiau u









Daytime Sleepiness Data

Included among the measures of daytime sleepiness

were the mean of the several Stanford Sleepiness Scale

ratings, and two variables calculated from the two evening

naps: mean sleep latency and total minutes asleep during

the naps. These variables were correlated with the

respiratory variables, resulting in a Pearson correlation

matrix. From these 21 correlations, two emerged signi-

ficant at p<.05 or less, both involving the mean Stanford

Sleepiness Scale ratings. Mean SSS ratings were corre-

lated with apnea index (r=.297), and apnea + hypopnea

index (r=.283).

Table 3-8 presents the means and standard deviations

for the subjects grouped by level of apnea/hypopnea. One-

way ANOVA procedures were non-significant for all these

variables [SSS: F(2,38)=2.4, p=NS; Mean sleep latency:

F(2,35)=0.4, p=NS; Total minutes asleep: F(2,35)=0.3,

p=NS].



Table 3-8. Means and standard deviations for daytime
sleepiness variables in 46 snoring males grouped
by level of apnea/hypopnea.


ho Apnea/ Low Apnea/ High Apnea/
Hypopnea Hypopnea Hypopnea
Stanford Sleepiness
Scale 3.2 (.7) 3.6 (.8) 4.0 (.7)
Mean Nap Latency 14.5 (5.7) 15.5 (5.3) 13.4(3.4)
Total Miinutes of
Sleep in Naps 11.3(11.9) 9.6(10.0) 13.5(6.6)









In summary, correlational procedures suggested

relationships between nocturnal respiratory variables and

subjective sleepiness, while no reliable between group

difference were noted on the nap or SSS variables.

Subjective Sleep Assessment

Two assessments of subjective sleep parameters were

completed, including the sleep questionnaire and sleep

log. Several variables were selected from the sleep

questionnaire and included: usual hours of sleep,

judgement of whether usual sleep was adequate, number of

naps per week, hours napping per week, number of nightly

wakenings, minutes of nightly wakenings, usual sleep

latency, depth of sleep, and daytime sleepiness. These

variables were correlated with the respiratory variables,

for a Pearson matrix of 63 correlations. Table 3-9

presents the significant correlations which emerged from

this matrix. Several respiratory variables were

associated with napping variables, while mean low

saturation and number of desaturations >4% were associated

with reported minutes of nightly wakenings.

Table 3-10 presents the means and standard deviations

from the subjects grouped by level of apnea/hypopnea.

Oneway ANOVA procedures revealed no significant between

group differences on the selected sleep questionnaire

variables [Usual hours sleep, F(2,43)=G.6, p=NS; Adequate

sleep, F(2,43)=0.3, p=NS; number of naps, F(2,43)=2.9,

p=NS; Hours napping, F(2,43)=0.8, p=NS; Number of











Table 3-9. Significant (p<.05) Pearson correlations between
nocturnal respiratory and sleep questionnaire
variables in 46 snoring males.



Minutes
Number hours Nightly
of Naps Napping Awakening

Apnea Index
Apnea+Hypopnea Index .249
Seconds in Events .331
Pean High Saturation
Mean Low Saturation .307
Lumber of Desaturations >4% .283 -.291
number of Deseturations 10% .403 .376










Table 3-10. Means and standard deviations of sleep
questionnaire variables derived from 46 snoring
males grouped by level of apnea/hypopnea.



No Apnea/ Low Apnea/ High Apnea/
Hypopnea hypopnea Hypopnea
eean h. Sleep 7.4 (.9) 7.0(1.6) 7.5(1.0)
Adecuate Amount? 1.9 (.4) 1.8 (.4) 2.0(0)
Number of laps/Week 1.8(1.5) 2.3(2.6) 4.3(2.2)
hours Napping/Week 2.6(1.8) 3.5(2.7) 4.0(2.7)
Dumber of Nightly
lakenings 1.1 (.9) 1.5(1.3) 1.6 (.4)
Minutes Nightly
Awakenings 1.9(1.1) 1.5(1.0) 1.5 (.8)
Sleep Latency 2.0 (.9) 2.0 (.6) 2.0 (.6)
Depth of Sleep 2.4 (.8) 2.1 (.9) 3.0(C)
Daytime Sleepiness 1.6 (.6) 1.9 (.4) 1.8 (.4)









awakenings, F(2,43)=0.9, p=LS: Ilinutes of awakening,

F(2,43)=0.8, p=NS; Sleep latency, F(2,43)=0.01, p=NS;

Depth of sleep, F(2,43)=2.6, p=NS; Daytime sleepiness,

F(2,43)=1.8, p=NS].

Means were calculated from the 7 day sleep logs, and

the following variables were selected for analysis:

reports of daytime sleepiness, number of naps, minutes

napping, total bedtime, sleep latency, number of

awakenings, minutes of awakenings. These variables were

correlated with the respiratory indices to form a Pearson

correlation matrix totalling 49 correlations. Table 3-11

presents the 6 significant (p<.05) correlations which

emerged. Again, number of desaturations were associated

with reports of napping, while mean high saturation was

associated with total bedtime, sleep latency, and minutes

awake, and mean low saturation was associated with minutes

awake as well.

Table 3-12 presents the means and standard deviations

for the subjects divided into levels of apnea/hypopnea.

Oneway ANCVA procedures failed to reveal significant

between group differences [Daytime sleepiness,

F(2,35)=0.6, p=NS; Number of naps, F(2,35)=0.3, p=NS;

Minutes of napping, F(2,35)=1.3, p=NS; Total bedtime,

F(2,35)=0.7, p=NS; Sleep latency, F(2.35)=0.2, p=NS;

Number of awakenings, F(2,35)=0.2, p=NS; Minutes of

wakenings, F(2,35)=.0001, p=NS].












Table 3-11.


Significant (p<.05) Pearson correlations between
nocturnal respiratory and sleep log variables in
46 snoring males.


Number Minutes
of Total Sleep Nightly
Naps Bedtime Latency Wakenings

Apnea Index
Apnea+Hypopnea Index
Seconds in Events
Mean High Saturation .237 .250 .359
Mean Low Saturation .265
lumber of
Desaturations >4% .240
Number of
Desaturations >10% .240









Table 3-12. Means and standard deviations of sleep log
variables derived from 46 snoring males grouped
by level of apnea/hypopnea.



No Apnea/ Low Apnea/ High Apnea/
hypopnea hypopnea Hypopnea
Daytime Sleepiness 1.9 (.9) 1.9(1.1) 2.4 (.3)
Number of Naps 0.2 (.2) .3 (.4) .3 (.3)
Minutes Napping 0.0(0) .2 (.5) 0 (0)
Total Bed Time 6.5(3.1) 5.8(3.4) 7.5 (.7)
Sleep Latency 10.5(8.1) 11.1(9.8) 8.4(4.7)
Number of Wakenings 1.3(0.8) 1.3 (.9) 1.3 (.2)
Minutes making 2.0(1.4) 1.7(1.1) 1.7 (.3)









In summary, no between group differences were noted

for either sleep questionnaire or log variables when

subjects were grouped by level of apnea/hypopnea.

Correlational procedures suggested relationships between

several respiratory variables and reports of napping and

wakening after sleep onset on the sleep questionnaire.

These findings were cross validated in the correlational

analysis of sleep log data, in which reports of napping

and wakening after sleep onset were related to respiratory

variables, as were total bedtime and sleep latency.

Nocturnal Respiratory and Neuropsychological Variables

Scores derived from the cognitive/neuropsychological

battery included WAIS Performance IQ, WAIS Verbal IQ,

Wechsler Memory Quotient, delayed recall of logical

stories, delayed recall of visual reproductions, delayed

recall of Rey complex figure, digit span, Hooper score,

Wisconsin card sort score, finger tapping right and left,

and Verbal fluency.

As a preliminary analysis, these scores were compared

with selected respiratory and demographic variables in a

multivariate regression procedure. In this type of

analysis, a set of predictor scores is used to

simultaneously predict a set of criterion variables while

controlling for intercorrelations amongst the measures.

An overall test of significance, such as the Hottelings

T2, establishes the probability of obtaining the observed

relationships from chance alone. Upon obtaining a








significant overall p value, univariate regressions are

used to individually predict criterion variables with the

multiple predictors. Contribution of individual

predictors may also be examined in these univariate

regressions.

The results from the multivariate regression of the

above noted cognitive scores and selected nocturnal

respiratory variables (Apnea index, Apnea + hypopnea

index, lean high saturation, Number of desaturations 14%,

Age, Weight:Height ratio) are presented in Table 3-13. An

overall Hotellings T2 score was significant (p<.003),

allowing examination of the univariate regressions.

Criterion variables which were significantly predicted by

the respiratory and demographic variables included WAIS

PIQ, VMS NQ, delayed recall of Rey figure and visual

reproductions, Hooper, and Wisconsin card sort. These

relationships reflected negative relationship between

overnight respiratory indices and neuropsychological

scores. Respiratory indices formed significant components

of the predictor combinations for all these variables save

Hooper score, suggesting relationships between nocturnal

respiratory indices and non-verbal intelligence, verbal

and visual memory, and ability to shift sets while problem

solving.

Because age is correlated with nocturnal respiratory

parameters in past reports, and also with neuropsycho-

logical scores in other samples, the various overnight









Table 3-13. Multivariate regression of demographic and
nocturnal respiratory variables on cognitive
scores in 46 snoring males.


Predictor Variables
Age
Weight:Height Ratio
Apnea Index
Apnea+Hypopnea Index
Mean High Saturation
Number of Desaturations >4%


Criterion Variables
WAIS Verbal IQ
WAIS Performance IQ
Wechsler Memory Scale Memory
Quotient
Delayed Recall Logical Stories
Delayed Recall Visual
Reproduction
Delayed Recall Hey Figure
Digit Span
Hooper Test
Finger Tapping Left Hand
Finger Tapping Right Hand
Verbal Fluency
Wisconsin Card Sort


Multivariate test of significance

Hotellings T2 = 4.841 (p<.003)


Univariate Regression with
Variable
WAIS Verbal IQ
WAIS Perlormance IQ
Wechsler Memory Scale
Memory Quotient
Delayed Recall
Logical Stories
Delayed Recall
Visual Reproductions
Delayed Recall Rey Figure
Digit Span
Hooper Test
Finger Tapping Right Hand
Finger Tapping Left Hand
Verbal Fluency
Wisconsin Card Sort


(6,3e) d.f.
Multiple R
.3646
.5808

.5345

.4713

.6450
.6461
.2432
.6786
.3907
.4270
.4441
.5143


Breakdown of significantly predicted criterion regressions


Criterion
WAIS Performance
10


Predictor
Age
Weight:Height Ratio
Apnea Index
Apnea+Hypopnea Index
Mean High Saturation
Number of
Desaturations >4?'


Eeta
.1240
.0170
-.2636
-.1553
.0920


Sig.
.382
.914
.059
.344
.533


-.3978 .024


F.
.971
3.22

2.53

1.80

4.51
4.54
.39
5.40
1.14
1.41
1.55
2.27


Sig.
.458
.012

.037

.123

.002
.001
.875
.000
.358
.235
.187
.050









Table 3-13-continued.


Criterion
Wechsler Memory
Scale Eiemory
Quotient





Delayed Recall
Visual
Reproductions





Delayed Recall
Rey FiFure






Hooper Test







iisconsin Card
Sort


Predictor
Age
Weight:Height Ratio
Apnea Index
Apnea+Hypopnea Index
Mean High Saturation
Number of
Desaturations >44%

Age
Weight:Height Ratio
Apnea Index
Apnea+Hypopnea Index
Mean High Saturation
Number of
Desaturations >45,

Age
Weight:Height Ratio
Apnea Index
Apnea+Hypopnea Index
Mean High Saturation
Number of
Desaturations >45%

Age
l\eight:Height Ratio
Apnea Index
Apnea+Hypopnea Index
Mean High Saturation
Number of
Desaturations >,4%

Age
Weight:Height Ratio
Apnea Index
Apnea+Hypopnea Index
Nean High Saturation
Number of
Desaturations >45


Leta
.0272
.0996
-.3315
-.3421
.2384

-.50,7

-.3254
-.1214
-.4037
-.2026
.1368

-.2888

-.3550
.1795
-.2206
-.0572
.3191

-.3228

-.5658
.1202
-.1428
.1478
.1072

-.2042

-.3409
.0258
-.3205
.1029
.2383

-.0801


Lig.
.852
.546
.024
.049
.124

.007

.018
.416
.003
.191
.325

.078

.010
.231
.091
.709
.025

.049

.000
.403
.251
.318
.421

.190

.027
.887
.031
.550
.130

.657









scores were entered into a partial correlation matrix,

controlling for age, with the cognitive scores. Table

3-14 presents the resulting correlations significant at

p<.05 or less. The table incicates that every cognitive

test but digit span and Hooper correlated with at least

one overnight respiratory index. Of additional impact is

the observation that in every case of a significant

correlation, deteriorating respiratory indices are

associated with worsening neuropsychological scores.

Respiratory indices with particularly heavy loadings on

neuropsychological scores included those quantifying

apneas, (Apnea index, Apnea + hypopnea index) as well as

those more directly representing deepening hypoxia (Mean

low saturation and Number of desaturations >4%).

Subjects were grouped by level of nocturnal

respiratory disorder (no apneas/hypopneas, subclinical

level of apnea/hypopnea, and clinically significant level

of apnea/hypopnea). Means and standard deviations for the

neuropsychological scores from these groups are presented

in Table 3-15. The table suggests a general deterioration

of cognitive scores as level of nocturnal respiratory

distress increases. Oneway AIIOVA procedures were applied

to the neuropsychological scores from each test. Signifi-

cant between group differences emerged for WAIS PIQ

[F(2,43)=6.9, p<.002] delayed recall of logical stores

[F[2,43)=4.0, p<.03] delayed recall of visual reproduc-

tions [F(2,45)=3.4, p<.04], delayed recall of Hey complex














-OO 00 0










co
0 0
0 r- ?N1



*
I f I I I I I




E C E
C6 CO CUD N .f

*-" U-C- 0 0 L 0N 3


0 C

OCU
S4-) *
C4E o c
0 N 0 0 N o ,

0 I0 I .


0 <;
O C.
1 0 > ZC
4-<

CL











41 r r- o
O -P ( C X *










r-O co 0 9 z
-P C
V C











0U o 0

Q0 0 c O w
*Uc- t C







.H0 CO to
4OCO C40 F- C C0

e 33 a c
C "O 2i C >, a CUUflC
c] CU a 4-1 0 C a) w i


rC o) -r ai 4 W ., M aU
a:ECUE W0 0CU C: CrC

>CO 0 0. O C H- C





E cC 0 0 0 00 C. O *C -C .> *
ar) H 0 H .1 a H > a











Table 3-15. teans and standard deviations of cognitive
variables in 46 snoring males groupec by level
of apnea/hypopnea.


Variable

WAIS Full IQ
WAIS Verbal IQ
WAIS Performance IQ
Weschler Memory
Scale memory
Quotient
Delayed Recall
Logical Stories
Delayed Recall
Visual
productions
Digit Span
Delayed Recall
Rey Figure
Hooper Test
.isconsin Card Sort
Finger Tapping
Right Hand
Finger Tapping
Left Hard
Verbal Fluency


No Apnea/ Low Apnea/ High Apnea/
Hypopnea Hypopnea Hypopnea


124.9(11.2)
123.7(12)
122.8(11.5)


118.3(13.9)
118.5(14.5)
114.5(14.4)


111.5 (9.6)
114.3(10.4)
100.1(12.5)a


124.7(14.3) 117.1(19.0) 107.8(18.7)

9.9 (2.4) 8.6 (3.2) 5.8 (4.0)a


10.5 (5.7) 10.4 (3.4) (.1 (4.7)
6.8 (0.8) 6.8 (1.2) t.6 (.8)


25.6 (5.4)
25.2 (2.6)
4.1 (1.2)


23.2 (6.0)
25.5 (4.8)
3.4 (1.5)


15.1 (8.4)a,b
22.8 (4.1)
2.8(1.6)


62.6 (8.5) 57.5(10.4) 56.1 (8.3)

59.1 (9.0) 54.0 (9.7) 54.8(15.1)
14.6 (4.5) 13.1 (4.2) 9.3 (2.6)e


a Significantly different from no sleep apnea/hypopnea.
b Significantly different from low sleep apnea/hypopnea.








figure [F[2,45)=6.5, p<.OO3], and Verbal fluency,

[F[2,43)=3.6, p<.04]. Followup comparisons (Scheffe)

showed the group with high apnea/hypopnea to be

significantly impaired relative to those with no sleep

apnea/hypopnea on WAIS PIQ, delayed recall of logical

stories and Rey complex figure as well as verbal

fluency. Additionally, the high apnea/hypopnea group was

impaired relative to those with subclinical apnea/hypopnea

on aelayea recall of the Rey complex figure. These

findings are bolstered by the lack of significant between

group differences on age, weight, or education (see Table

3-15), all of which are potential alternative explanations

of the between group differences.

As a check, an analysis was run to evaluate the

relative contribution of nocturnal respiratory variables

and sleep variables reflecting fragmentation in predicting

cognitive/neuropsychological scores. A multivariate

regression of these scores was carried out. Demographic

and respiratory scores included age, apnea index, apnea +

hypopnea index, and number of desaturations >4%. A subset

of the earlier noted sleep variables was chosen to reflect

measures thought to be most responsive to sleep

fragmentation. These included sleep efficiency, number of

stage O periods, time % of stage 0, time % of stage 2, and

number of stage changes. These respiratory and sleep

parameters were used to predict the 5 cognitive scores

which were earlier shown to be predicted by nocturnal








respiratory scores: WAIS PIQ, Wechsler Memory Quotient,

delayed recall of visual reproductions and Hey figure, and

Wisconsin card sort. Table 3-16 presents the data from

this analysis. The Hotellings T2 was significant,

(p<.001) with delayed recall of visual reproductions and

Rey figure significantly predicted by the predictor

scores. Inspection of the individual regressions for

these two measures discloses that age, apnea index, and

number of desaturations >4% all formed significant

components of the prediction of delayed recall of visual

reproductions, while sleep efficiency was the sole

significant predictor of the Key figure. These data seem

supportive of an independent contribution of both

nocturnal respiratory variables and sleep variables at a

limited level, to the prediction of cognitive/neuro-

psychological scores.

In summary, various statistical procedures indicate

relationships between nocturnal respiratory measures and

non-verbal intelligence, verbal and non-verbal memory,

expressive verbal fluency, and ability to shift cognitive

set while problem solving. Division of subjects by level

of nocturnal respiratory distress indicated that subjects

with high amounts of apnea/hypopnea were impaired relative

to controls on measures of non-verbal intelligence, verbal

and non-verbal memory, and expressive verbal fluency, as

well as being impaired relative to subjects with

subclinical levels of apnea/hypopnea on measures of











Multivariate regression of sleep and nocturnal
respiratory on cognitive variables in 43
snoring males.


Predictor Variables

Age
Apnea Index
Apnea+Hypopnea Index
Number of Desaturations >4%
Sleep Efficiency
Number Stage 0 Periods
Time % Stage 0
Time C Stage 2
Lumber of Stage Changes


Criterion Variables

WAIS Performance IQ
Wechsler Memory Scale Memory
Quotient
Delayed Recall Visual
Reproduction
Delayed Recall Rey Figure
Verbal Fluency
Wisconsin Card Sort


Sultivariate test of significance


Hotellings T2 = 4.063

Univariate Regression
Variable
WAIS Performance IQ
Vechsler Memory Scale
Memory Quotient
Delayed Recall Visual
Reproductions
Delayed Recall Rey
Figure
Verbal Fluency
Wisconsin Card Sort


p<.001

with (9,33) d.f.
Multiple R
.610

.566

.655

.668
.578
.453


Breakdown of significantly predicted criterion regressions
Criterion Predictor Beta Sig.
Delayed Recall Age -.3596 .020
Visual Apnea Index -.3218 .021
Reproduction Apnea+Hypopnea Index .2529 .218
Number of
Desaturations >4% -.3778 .041
Sleep Efficiency .2303 .408
Number of Stage 0 Periods .3260 .209
Time % Stage 0 -.1136 .731
Time % Stage 2 .0005 .998
Number of Stage Changes -.2235 .391


Table 3-16.


F.
2.17

1.73

2.91

2.96
.85
.95


Si .
.051

.121

.012

.011
.096
.499









Table -1I-continuec.


Criterion
Delayed recall
Key Figure


Predictor
Age
Apnea Index
Apnea+Hypopnea Index
Lumber of
Desaturations >4%
Sleep Efficiency
lumber of Stage 0 Periods
Time S Stage 0
Time % Stage 2
Number of Stage Changes


Beta
-.2623
-.1519
-.3286

-.0795
.6832
.2487
.4559
-.0465
-. 054


.074
.259
.085

.b57
.016

.172
.802
.057








non-verbal memory. Addition of variables reflecting sleep

fragmentation to the analysis indicated that both

nocturnal respiratory and sleep variables make significant

and independent contributions to prediction of

cognitive/neuropsychological scores.














CHAPTER FOUR
DISCUSSION

Demographics and Incidence of
Nocturnal Respiratory Disorder

The present sample of 46 snoring males was selected

as a group likely to display a wide range of sleep

disordered breathing. These subjects, with a mean age of

50 years and a mean weight of 190 pounds, reflect a mostly

middle-aged to older male population which is moderately

obese. The demographic characteristics of this sample are

similar to those of the "typical" sleep apnea syndrome

patient (Guilleminault et al., 1978), although the

requirement of general good health insured that subjects

with a classic SAS were excluded. This sample may be

considered representative of subjects predisposed to sleep

disordered breathing, although at no more than a

subclinical level. Therefore it seems valid to consider

various risk factors as they relate to nocturnal

respiration.

In the present sample, 62% experienced at least one

apnea or hypopnea, while 13% suffered higher levels of

apnea/hypopnea. Only two other reports have described

samples selected for snoring or respiratory complaints.

Ancoli-Israel et al. (1981) examined 24 subjects with

complaints of respiratory difficulties, daytime








sleepiness, and muscular events (nocturnal myoclonus).

Further, these subjects had a mean age of almost 70 years,

indicating significant sampling and age differences from

the present study. Ninety-three per cent of these

subjects had at least one event, while 25% had high

numbers of apnea/hypopnea. These rates are much higher

than those found in the present sample. This is probably

attributable to the sampling differences noted above.

However, a recent report by Miles and Simmons (1984)

described apneas in a series of 190 patients referred for

complaints of heavy snoring. In this sample, 73% had at

least one event, while 9% had severe levels of apnea.

These data are quite comparable to the prevalences

observed in the present sample, 62% for at least one event

and 13i for high levels of apnea/hypopnea, and lend some

confidence to the present finding.

Thus the overall frequency of sleep disordered

breathing in middle-aged subjects with heavy snoring

appears to be between 62 and 75% for at least one event,

and between 9 and 135 for more severe levels of apnea.

The relatively close concordance of the two reports on

heavy snoring subjects is reassuring.

Factors which were found to be related to the

occurrence of sleep apneas in the previous review included

male sex, complaints of snoring, increasing age, and

obesity. The first two factors could not be examined in

the present sample because of their use as selection









criteria. A wide range of age was achieved in this

sample, but neither non-parametric nor correlational

procedures revealed significant relationships between

increasing age and sleep disordered breathing. Although

several previous reports have noted relationships between

increasing age and sleep disordered breathing, typically

this relationship is moderate for apnea (Block et al.,

1979; Carskadon et al., 1980; Bixler et al., 1982).

Additionally, no other correlational data are available in

subjects selected for sleep and breathing complaints. It

may be speculated that the effect of snoring masks the

effect of age in this group. Alternatively, some quirk of

sampling may have generated this effect. Correlational

procedures, however, revealed substantial relationships

between weight and sleep disordered breathing. This

finding is consistent with several past reports. To

summarize, the present data support relationships between

increasing weight and nocturnal events, while

relationships between age and sleep disordered breathing

are not found in this sample of heavy snoring males,

perhaps because the effect of age is masked by the effect

of snoring.

The temporal characteristics of the apneic events

assessed by two hour interval suggested an increase in

events between 3 and 5 am, although differences in sleep

onset time and wakeup time render any interpretation of








these data somewhat tenuous. This finding needs

replication before any conclusions may be drawn.

Ninety-eight per cent of apneic events occurred in

light slow wave and REM sleep, an observation which is not

surprising given that subjects spent nearly 80% of their

sleep in these stages. The preponderance of events in

these stages of sleep is consistent with the bulk of past

reports in this area (e.g, Bixler et al., 1982; Block et

al., 1979; Krieger et al., 1985).

When subjects were divided into groups by level of

apnea/hypopnea, significant differences emerged on oxygen

desaturation variables. Specifically, comparisons

indicated that the group with a high level of

apnea/hypopnea events experienced deeper desaturations

than the remaining two groups. This difference occurred

despite a lack of differences on age and weight among the

three groups, and appears to implicate the respiratory

events in the genesis of the desaturations. This is

consistent with an earlier report by McGinty et al. (1982)

who demonstrated that subjects with significant levels of

sleep disordered breathing were impaired on oxygen

desaturation variables, relative to subjects without

respiratory events. Subjects with heavy snoring and

multiple nocturnal apneic respiratory events appear to be

at risk for nocturnal oxygen desaturation.

In summary, the present sample of 46 snoring males

appears representative of those predisposed to experience









sleep disordered breathing (middle-aged, snoring males).

A prevalence of 62k was observed for occurrence of at

least one event, while 13% had high numbers of

apnea/hypopnea, data which are consistent with an earlier

report drawn from heavy snoring males. Most events

occurred in light slow wave or REM sleep. No effect of

age on level of sleep disordered breathing was observed,

although increasing weight was linked to nocturnal

respiratory events. When the sample was divided by level

of apnea/hypopnea, subjects with high numbers of

apnea/hypopnea suffered significantly exacerbated oxygen

desaturation, although no significant differences were

noted for weight or age among the groups.

ijocturnal Respiratory and Health Variables

The finding of only very limited relationships

between nocturnal respiratory events and health variables

must be tempered by an acknowledgement of possible bias

against such findings in this sample. Specifically, the

sample was selected to be normal and healthy, a criterion

which undoubtably restricted the range of possible health

pathology, and hence probably attenuated the correlational

findings. Additionally, the self-report health checklist,

the Cornell Medical Index, cannot be regarded as a

powerful measure of health pathology. In a very real

sense then, the odds were against finding reliable

relationships between nocturnal respiratory and health

variables.








In light of these caveats, it is of interest that two

measures of desaturation, mean low saturation, and number

of desaturations >04 were associated with increasing blood

pressure readings. While a report by Guilleminault et al.

(1978) noted an increasing prevalence of hypertension in

subjects diagnosed with sleep apnea syndrome, a

relationship between desaturation and hypertension has not

been previously reported. This finding was observed in

spite of the fact that subjects prescribed hypertensives

were not medication free for the pressure reading.

Additionally, a check of self reports of diagnosed

hypertension showed six subjects in the no event group as

having diagnosed hypertension, as opposed to six in the

low apnea/hypopnea group, and one in the high

apnea/hypopnea group, data which did not link

apnea/hypopnea, per se, with hypertension in these

subjects. The isolated finding of a relationship between

mean high saturation and overall number of symptoms

endorsed on the CMI is more puzzling and does not follow

any logical analysis.

In summary, the present study's assessment of health

variables must be regarded as somewhat limited. In spite

of this, an association between two measures of desatura-

tion and a measure of blood pressure was noted. It is

speculated that blood pressure may be more responsive to

nocturnal oxygen saturation tnan to the apneic events

alone.








Nocturnal Respiratory and Sleep/Wake Variables

EEG Data

The overnight sleep in the laboratory for these

snoring subjects was noteworthy for a general shift

towards lighter sleep than usual. A longer sleep latency,

more awakenings, more light sleep, and less slow wave and

REM sleep than normal are apparent. These findings

parallel the report by Block et al. (1979) who found

shifts to lighter sleep and more awakenings than in a

sample of normal subjects. These changes were said to

result from a first night in a strange bed and the

discomfort of the multiple clips and sensors used to

monitor respiration. Recording montages were identical

for the present subjects, and these two factors

undoubtably influenced the sleep of the present subjects

as well.

An intriguing observation was a lack of covariation

between respiratory and EEG sleep variables derived from

the recording night. This finding is counterintuitive, as

one would expect arousals from the respiratory events to

influence the sleep characteristics of subjects with

events. However, correlational procedures failed to

support this notion. It may be that a ceiling effect was

exerted on sleep measures sensitive to disturbance, one

that resulted from the first night effect compounded by

the discomfort of the recording procedure. The already

elevated disturbance sensitive sleep variables might not




Full Text
128
Hooper, H. (1958), The Hooper Visual Organization Test;
Hanue 1. Los Angeles; Western Psychological Services.
Kales, A., Bixler, E., Soldatos, C., Vela-Bueno, A.,
Caldwell, A., & Cadieux, R. (1982). Insomnia: The
role of sleep apnea and nocturnal myoclonus.
Psychosomatics, 25 187-195*
Kreiss, P., Kripke, D., & Ancoli-Israel, S. (1982).
Prevalence of sleep apnea among representative
inpatients on a veterans administration medical ward.
Abstract published for 1982 APSS meeting, San Antonio,
Texas.
Krieger, J., Turlot, J., Margin, P., & Kurtz, D. (1585).
Breathing during sleep in normal young and elderly
subjects: Hypopneas, apneas, and correlated factors.
Sleep, 6^(2), 106-120.
Krop, H., Block, A., & Cohen, E. (1975). neuropsycho
logical efiects of continuous oxygen therapy in chronic
obstructive pulmonary disease. Chest, 64, 517-522.
Krop, II., Block, A., Cohen, E., Croucher, H., & Schuster,
J. (1977). Neuropsychological effects of continuous
oxygen therapy in the aged. Chest, 72, 757-745*
Lugaresi, E., Cirignotta, F., Coccagna, G., & Baruzzi, A.
(1982). Snoring and the obstructive sleep apnea
syndrome. EEG, Suppl No. 55, 421-450.
Lugaresi, E., Cirignotta, F., Coccagna, G., & Piaa, C.
(1980). Some epidemiological data on snoring and
cardiocirculatory disturbances. Sleep, _3(3/4), 221-224.
Lugaresi, E., Coccagna, G., Berti, G., & Mantovani, M.
(1968). La "maledizione di Ondine" il disturbo del
respiro del sonno neil ipoventalizione alveolari
primaria. List Neurologie, 20, 27-37.
McFarland, R. (1937). Effects of partial acclimitization
on psychological tests at high altitudes. Journal of
Comparative Psychology, 23, 227-230.
McGinty, D., Littner, M., Beahm, E., Ruiz-Primo, E., Young,
L., & Sowers, J. (1982). Sleep related breathing
disorders in older men: A search for underlying
mechanisms. Neurobiology of Aging, 337-350.
leehl, P. (1977). Psychodiagnosis. New York: MAI. Norton
and Co.


62
apnea/hypopnea group was significantly different; from the
other two groups on these two indices. While mean
duration of apneas and hypopneas was not significantly
different between the high and low apnea groups
[F(1,25)=3*3, 0.1; p=NS], total seconds in respiratory
events was [F(2,43)=50.4, pC.0000], with followup
comparisons showing that the high apnea/hypopnea group was
significantly different from tne other two groups.
Turning to the oxygen saturation data, mean high
saturation failed to reach significance LF(2,43)=0.3,
p*NS]. However, mean low saturation, number of
desaturations and >10% were all significant
[F(2,43)=11.6, p<.0001; 7-0, p<.C02; 21.4, pC.0000], and
followup comparisons showed that the group with high
levels of apnea/hypopnea was significantly different from
the other two groups.
Thus, division of the sample into levels of
apnea/hypopnea resulted in groups which were not
significantly different on demographic variables including
age, weight, weight:height ratio, and education. However,
the group with high levels of apnea/hypopnea had
significantly more nocturnal events, time in events, and
oxygen desaturation.
In summary, the present sample of 46 snoring males
appears representative of those predisposed to experience
sleep disordered breathing (middle-aged, snoring males).
Sixty-two per cent of these subjects had at least one


76
Table 3-13. Multivariate regression of demographic and
nocturnal respiratory variables on cognitive
scores in 46 snoring males*
Criterion Variables
VAIS Verbal IQ
VAIS Performance IQ
Wechsler Memory Scale Memory
Quotient
Delayed Recall Logical Stories
Delayed Recall Visual
Reproduction
Delayed Recall Rey Figure
Digit Span
Hooper Test
Finger Tapping Left Hand
Finger Tapping Right Hand
Verbal Fluency
Wisconsin Card Sort
Multivariate test of significance
Hotellings T^ = 4.841 (p<*003)
Univariate Regression with
(6,3e) d.i.
Variable
Multiple R
F.
Sig.
VAIS Verbal IQ
.3646
.971
.458
WAIS Performance IQ
.5808
3.22
.012
Wechsler Memory Scale
Memory Quotient
5345
2.53
.037
Delayed Recall
Logical Stories
.4713
1.80
.123
Delayed Recall
Visual Reproductions
.6450
4.51
.002
Delayed Recall Rey Figure
. 6461
4.54
.001
Digit Span
.2432
.39
.875
Hooper Test
.6786
5.40
.000
Finger Tapping Right Hand
.3907
1.14
.358
Finger Tapping Left Hand
.4270
1.41
.235
Verbal Fluency
.4441
1.55
.187
Wisconsin Card Sort
.5143
2.27
.050
Breakdown of significantly
predicted criterion regressions
Criterion Predictor
£eta
Sig.
WAIS Performance Age
.1240
.382
10 Weight
¡Height Ratio
.0170
.914
Apnea
Index
-.2636
.059
Apnea+Hypopnea Index
-.1553
.34 4
Mean High Saturation
.0920
.533
Number
of
Desaturations >A%
-.3970
.024
Predictor Variables
AgG
Weight¡Height Ratio
Apnea Index
Apnea+Hypopnea Index
Mean High Saturation
Number of Desaturations >A%


40
evidence which has been collected on neurotransmitter
alterations in hypoxia. In brief, mild to moderate
hypoxia, cerebral levels of ATP are normal, implying that
energy supplies of neurons are not disturbed. In con
trast, the turnover of certain neurotransmitters is
altered. Gibson et al. note that oxygen is an important
substrate for the synthesis of several neurotrans
mitters. V/hen rats are exposed to acute hypoxia, syn
thesis of dopamine and serotonin is decreased, while
absolute levels of tryosine, tryptophan, catecholamines,
and serotonin remain constant. Continuing hypoxia leads
to an adaptation resulting in normal turnover of these
neurotransmitters. Gibson and Blass (1976) examined the
affect of acute hypoxia on central acetylcholine (ACh).
As in dopamine (DA) and serotonin (5-HT), acute hypoxia
results in a decrease in synthesis in ACh, but not in the
absolute level of ACh. Thus there is a substantial body
of evidence indicating changes in central neurotransmitter
levels in hypoxia. Generalizing from the animal data to
human subjects, Gibson and Duffy (1981) suggest that these
changes underlie the behavioral deficits of hypoxia in
humans (which will be reviewed below).
Other variables have been examined in acute hypoxia
in animals. Annau (1972) found that rats exposed to brief
hypoxia exhibited decreased appetite, thirst, and self
stimulation rates. Annau observed that brief hypoxia
depressed all behavioral variables. Gellhcrn (1951) found


104
fluency, and cognitive flexibility were observed. It was
speculated that hypoxic alterations in the cholinergic
synthetic pathways underlayed these changes. Apparently,
a significant subgroup of heavy snoring males fall on the
same continuum as subjects with sleep apnea syndrome,
occupying an intermediate position between normals and
those with SAS. From these and other data, it is
concluded that
1) between 62 and 73% of snoring males experience at
least one nocturnal respiratory event, while
between 9 and 13% appear to experience 5 or more
events per hour;
2) most events occur in light slow wave or REM
sleep;
3) increasing obesity is linked to nocturnal events
in snoring males;
4) nocturnal events are linked to increasing blood
pressure in snoring males;
5) nocturnal events are linked to increasing sleepi
ness in snoring males;
6) nocturnal events are linked to deteriorating
neuropsychological scores in snoring males;
7) a subset of heavy snoring men, those with more
than 5 events per hour, suffer increased oxygen
desaturation and decreased cognitive abilities in
the areas of non-verbal intelligence, verbal and


41
that brief hypoxia of 4-7*5% O2 abolished cortical re
sponses to auditory stimulation. Sara (1974) trained rats
in active one way avoidance tasks. Immediately after
training, subjects were exposed to 3.5-4.0# oxygen. Rats
tested at 1 and 3 hours post hypoxia avoided correctly,
while those tested at 6 and 24 hours did not, leading Sara
to suggest that hypoxia resulted in a memory retrieval
deficit. Freides and Allweise (1978) also trained rats in
one way avoidance and exposed them to 2,0% hypoxia imme
diately following training. These animals avoided at 12
and 200 minutes post hypoxia, but not at 90 minutes.
Friedes and Allweise suggested that a sensitive period for
hypoxia effects on training existed for 15 minutes after
learning. Tauber and Allweise (1975) followed the same
training/hypoxia paradigm and found avoidance impaired at
2.5 but not 4.5 hours. They suggest that hypoxia inter
feres with a medium term memory. While some of these data
are conflicting, it appears that acute hypoxia depresses
certain behavioral variables and interferes with aspects
of memory processes.
While caution is in order when generalizing from
animals to humans, several seemingly relevant points
emerge from the review. Behaviorally, simple activities
may be spared, although some depression of appetitive
activities may occur, while more complex behavior depen
dent on memory processes may be altered. Of particular
interest are the demonstrated alterations in


112
labie B-2-continued.
Mean Stanford
Mean Sleep
Minutes of
Subject
Sleepiness
Latency
Sleep
a
Hating
j- Naps
- Naps
42
1.5
06.0
30
43
2.5
-
-
44
3.8
05.5
35
43
3.4
-
-
46
2.1
-
-


2
noted in SAS patients. The link between apneas and patho
logy may be more complex than was originally thought,
given its apparently benign occurrence in otherwise
asymptomatic subjects.
The present study will examine the correlates of
sleep apnea episodes in a population selected to display a
wide range of such events. The relationship between sleep
apnea indices and sleep characteristics, age, weight and a
variety of deficit measurements will be analyzed.
A review of the current data available on various
aspects of sleep apneas and sleep apnea syndromes will be
presented as background for the study. The diagnosis,
history, incidence, and characteristics of sleep apneas
and sleep apnea syndrome will be reviewed, followed by a
discussion of variables associated with sleep apneas.
Lastly, the deficits commonly found in sleep apnea syn
drome will be detailed, along with measurements thought to
be sensitive to these deficits.
Definitions and Terminology
Sleep apneas have been defined as cessations of air
flow at the mouth and nose which last for 10 or more
seconds (Guilleminault et al., 1976). In a sleep apnea
syndrome (SAS), multiple sleep apneas are accompanied by
snoring, excessive daytime sleepiness, cardiopulmonary
complications, and altered states of consciousness. The
ease of quantification of sleep apneas has led to their
use as an important tool in diagnosing SAS. Guilleminault


35
Measurement of Arousal Deficits (Hypersomnolence)
Excessive daytime sleepiness is a common sequela of
sleep apnea syndrome. Measurements of the sleep/wake
cycle of subclinical apnea patients thus seems indi
cated. A sleep questionnaire assessing trait sleep habits
seems in order, as well as 1 week sleep logs following the
experimental night. The sleep questionnaire and sleep
logs have been used in ongoing research in W.E. Webb's
studies of aging and sleep and seem an adequate assessment
of sleep patterns of subclinical apnea subjects. Addi
tionally, electroencephalographic recordings of sleep
during the experimental night seem in order using a stan
dard recording and scoring system (Agnaw and Webb,
1972). The assessment of daytime sleepiness is relatively
new, only recently being the subject of investigation.
William Dement and associates began to develop methods of
quantifying somnolence in the early 1970s. The first
attempt produced a subjective rating scale, the Stanford
Sleepiness Scale (SSS; Hoddes et al., 1973)* The SSS was
a 7 point Likert scale with endpoints anchored on 1) feel
active and vital; and 7) almost in reverie, sleep onset
soon; with 5 points with statements reflecting various
degress of somnolence in between. Patients rated their
introspective level of sleepiness on this scale.
Unfortunately, the SSS proved to be vulnerable to exactly
the same difficulties as a global rating scale used on SAS


Table b-3
¡bleep questionnaire variables from 46 snoring males.
Subject
a
Mean
Hours
Sleep
Enough
?
Humber
of
Maps
Hours
Happing
Humber
of
Wakings
Minutes
Awake
Sleep
Latency
Depth
of
Sleep
Daytime
Sleepiness
1
7
2
4
4
2
0
1
3
2
2
o
2
6
5
1
0
1
*
>
2
3
7
1
"Z
J
3
0
2
2
3
2
4
8
2
0
0
1
2
2
3
2
5
6
2
0
0
0
0
3
3
2
6
7
2
1
2
0
0
2
3
2
7
8
2
0
0
1
2
3
1
2
8
1
3
2
3
3
2
1
3
2
9
8
2
7
4
3
2
2
2
2
10
9
2
2
2
2
2
3
2
2
11
7
2
1
1
1
1
2
2
2
12
'7
i
2
0
0
4
1
2
1
1
13
1
2
2
3
2
1
2
3
1
14
9
2
1
2
0
0
2
1
2
15
9
2
0
0
0
2
2
3
1
16
9
2
2
2
2
1
3
2
2
17
7
1
2
3
1
1
2
3
2
18
5
1
2
2
1
3
3
1
2
19
8
2
6
3
2
4
4
1
2
20
1
3
8
10
3
3
2
3
2
21
7
2
0
0

1
1
1
1
22
6
1
3
3
3
3
2
3
2
23
8
2
6
0
0
3
2
2
1
24
8
2
0
0
1
2
2
2
2


LIST OF TABLES
Table Page
2-1 Interrater agreement for respiratory events
of 46 snoring males 51
3-1 Demographic variables from 46 snoring males 56
3-2 Incidence of low and high levels of apnea/
hypopnea by age in 46 snoring males 58
3-3 Significant (p<.05) Pearson correlations
between nocturnal respiratory variables
and demographic variables in 46 snoring
males 59
3-4 Means and standard deviations for selected
demographic and nocturnal respiratory
variables in 46 snoring males grouped by
level ol apnea/hypopnee 61
3-5 Means and standard deviations for health
related variables in 46 snoring males grouped
by level of apnea/hypopnea 65
3-6 Means and standard deviations of selected
sleep variables from 43 snoring males 66
3-7 Means and standard deviations of selected
sleep variables from 43 snoring males
grouped by level of apnea/hypopnea 68
3-8 Leans and standard deviations for daytime
sleepiness variables in 46 snoring males
grouped by level of apnea/hypopnea 69
3-9 Significant (p<.05) Pearson correlations
between nocturnal respiratory and sleep
questionnaire variables in 46 snoring males 71
3-10 Means and standard deviations of sleep
questionnaire variables from 46 snoring
males groupec by level of apnea/hypopnea 71
v


Nocturnal Respiratory and Health Variables 63
Nocturnal Respiratory and Sleep/V.'ake Variables..65
ELG Data 65
Daytime Sleepiness Data 69
Subjective Sleep Assessment 70
Nocturnal Respiratory and Neuropsychological
Variables 74
FOUR DISCUSSION 86
Demographics and Incidence of Nocturnal
Respiratory Disorder 86
Nocturnal Respiratory and Health Variables 90
Nocturnal Respiratory and Sleep/Wake Variables,.92
EEG Data 92
Daytime Sleepiness, Sleep Questionnaire,
and Sleep Log Data 94
Nocturnal Respiratory and Neuropsychological
Variables 96
FIVE SUMMARY AND CONCLUSIONS 103
APPENDIX
A SCORING PROTOCOL FOR RESPIRATORY VARIABLES 106
B RAW DATA TABLES 109
REFERENCES 123
BIOGRAPHICAL SKETCH 132


CHAPTER FIVE
SUMMARY AND CONCLUSIONS
In this study, a detailed assessment of the nocturnal
respiration, health status, cognitive/neuropsychological
skills, and sleep/wake cycle of 46 snoring male subjects
was carried out. Sixty-two per cent of these subjects
experienced at least one event (apnea or hypopnea) while
13% exhibited more than 5 events per hour. Most events
occurred in light slow wave or REM sleep. Obesity was
linked to increasing nocturnal events. Subjects with more
than 5 events per hour suffered significantly exacerbated
oxygen desaturation relative to the remaining subjects.
Overnight oxygen desaturation was linked to increasing
blood pressure readings. As a whole, these subjects
experienced a longer sleep latency, increased wakening
after sleep onset, and more light sleep in the lab, the
classic "first night" effect. It was thought that this
first night effect obscured relationships between noc
turnal respiratory and sleep parameters, which were not
observed. However, overnight respiratory disturbances
were linked to increased sleepiness and napping. Addi
tionally, relationships between nocturnal events and
deteriorating scores on tests tapping non-verbal intelli
gence, verbal and non-verbal memory, expressive verbal
103


71
Table 3-9* Significant (p<.05) Pearson correlations between
nocturnal respiratory and sleep questionnaire
variables in 46 snoring males.
Minutes
Number
hours
Nightly
of Naps
Napping
Awakening
Apnea Index
Apnea+Hypopnea Index
.249
Seconds in Events
.331
Mean High Saturation
Mean Low Saturation
.307
Lumber of Desaturations >4%
.283
-.291
Lumber of Desaturations _>10%
.403
.378
Table 3-10. Means and standard deviations of sleep
questionnaire variables derived from 46 snoring
males grouped by level of apnea/hypopnea.
No Apnea/
Hypopnea
Low Apnea/
Hypopnea
High Apnea
Hypopnea
Mean h. Sleep
7.4 (.9)
7.0(1.6)
7.5(1.0)
Adequate Amount?
1.9 (.4)
1.6 (.4)
2.0(0)
Number of Naps/Y/eek
1.6(1.5)
2.3(2.6)
4.3(2.2)
Hours Napping/Y;eek
Number of Nightly
2.6(1.8)
3.5(2.7)
4.0(2.7)
Wakenings
Minutes Nightly
1.1 (.9)
1.5(1.3)
1.6 (.4)
V/akenings
1.9(1.1)
1.5(1.0)
1.5 (.8)
Sleep Latency
2.0 (.9)
2.C (.6)
2.0 (.6)
Depth of Sleep
2.4 (.8)
2.1 (.9)
3.0(C)
Daytime Sleepiness
1.6 (.6)
1.9 (.4)
1.8 (.4)


9
5. There was a significant correlation between apnea and
age.
For noncomplaining subjects then, males suffer a
higher incidence of apnea activity in virtually every
study which compares sex effects. For males, incidences
range widely from a low of 14.6% (Bixler et al., I960) to
a high of 45% (Webb, 1974); while females range from a
low of 10.2% (Bixler et al., 1982) to a high of 40% (Block
et al., 1980). These differences are probably due at
least in part to age differences in male samples, as well
as age differences in females (only as they reflect rnena-
pausal status). Reported number of apneas range from 2.5
(Webb, 1974) to 8.5 (Bixler et al., 1982) in males and
from 2 (Guilleminault et al., 1978) to 10.5 (Bixler et
al., 1982) in females. Many studies report a positive
correlation between number of apneas and increasing age.
A second sampling technique which has been utilized
has been studying subjects with sleep and breathing com
plaints. Kales et al. (1982) studied 200 subjects with
primary complaints of insomnia. There were 82 males and
118 females in the sample with a mean age of 42.5. Sleep
apnea activity was noted in 10.5% of this group (M =
13*4%, F = 8.5%) with a mean number of 11.2 episodes for
males and 6.6 for females. Subjects with sleep apnea
activity were significantly older (46.8) than those with
out (40.6). Ancoli-Israel et al. (1981) recruited 24
subjects whose answers to questionnaires provoked


129
Miles, L., & Simmons, F. (1964). Evaluation of 190
patients with loud and disruptive snoring. Sleep
Research, 13, 154.
Miller, W. (1962). Cardiac arrhythmias and conduction
disturbances in the sleep apnea syndrome. American
Journal of Medicine, 75* 517-521 .
Milner, E. (1970). Memory and the medial temporal regions
of the brain. In K. Pribram & D. Broadbent (eds),
Biology of Memory. New York: Academic Press.
Mitler, M., Gujaverty, K., Sampson, M., & Bowman, C.
(1982). Multiple nap approaches to evaluating the
sleepy patient. Sleep, , 119-127.
Orr, Vi., Martin, R., Imes, N., Rogers, R., & Stahl, M.
(1979). Hypersomnolent and nonhypersomnolent patients
with upper airway obstruction during sleep. Chest,
75(4), 418-422.
Csterrieth, P. (1944). Le test de copie d'une figure
complexe. Archives de Psychologie, 50. 206-556.
Plum, F., Posner, J., & Hain, R. (1962). Delayed
neurological deterioration after anoxia. Archives of
Internal Medicine, 110, 18-25*
Reynolds, C., Coble, P., Kupfer, D., & Holzer, B. (1982).
Application of tne multiple sleep latency tesx in
disorders of excessive somnolence.
Electroencephalography and Clinical Neurophysiology, 55.
445-452.
Richarcson, G., Carskadon, M., Flagg, Vi., van den Hoed, J.,
Dement, W., & Mitler, H. (1978). Excessive daytime
sleepiness in man: Multiple sleep latency measurements
in narcoleptic and control subjects.
Electroencephalograpny ano Clinical Neurophysiology, 45.
021-627.
Richardson, J., Chambers, R., & Heywood, P. (1959)*
Encephalopathies of anoxia and hypoglycemia. Archives
of Neurology, J_, 178-190.
Roth, T., Hartse, K., Zorick, F., & Conway, V. (1980).
Multiple naps in the evaluation of daytime sleepiness in
patients with upper airway sleep apnea. Sleep, (3/4)f
425-459.
Russell, E. (1975). A multiple scoring method for
assessment of complex memory functions. Journal of
Consulting and Clinical Psychology, 45. 800-809.


APPENDIX A
SCORING PROTOCOL FOR RESPIRATORY VARIAELES *
Data from the nose and mouth thermisters, as well as
the oximeter and strain gauges* were used in scoring
respiratory events.
An apnea was scored when a cessation of airflow at the
mouth and nose lasting for 10 or more seconds was
observed. Up to 3 mm of "noise" were tolerated in an apnea,
as long as no rhythmic pattern which might represent shallow
breathing was observed. An event began at the end of the
last exhalation preceding the apnea. Duration, lowest
saturation, and per cent saturation change were noted.
A hypopnea was scored when a reduction in amplitude of
breathing at the mouth and nose of 50# or mere was
accompanied by a 10# or larger desaturation. The event
began at the end of the last full exhalation. Duration,
lowest saturation, and per cent saturation change were
noted.
Each record was separately scored by two raters. Later
each record was jointly rescored and disagreements were
resolved by consensus.
A separate scoring of oxygen saturation was made by
rating each minute of the record for highest and lowest
saturation. Additionally, a separate tabulation of
106


59
also divided into those above and below the mean weight:
height ratio (2.7) and crossed with those with ana without
t least one event or high levels o apnea/bypopnea. Both
these non-parametrie tests were also non-significant
(X2=.01; p=I\S; X2=1.5; p-NS).
labia 3-3 utilizes Pearson correlations to explore
possible relationships between demographic variables (age,
weight, weight:height ratio) and nocturnal respiratory
indices (apnea index, apnea + hypopnea index, seconds in
events, mean high saturation, mean low saturation, number
of desaturations _>4^, and number of desaturations >10%).
Again, age displays little relationship with respiratory
indices, but substantial correlations between weight and
various respiratory indices are apparent.
Table 3-3. Significant (p<.05) Pearson correlations between
nocturnal respiratory variables and demographic
variables in 46 snoring males.
Respiratory
Variable
Age
Weight
WeightrHeight
Ratio
Number of Apneas
Number of Apneas+Hypopneas
Apnea Index
.265
Apnea+hypopnea Index
.260
.306*
Mean High Saturation
-.323*
-.363*
Mean Low Saturation
-.397*
-.460*
Mean Saturation Change
.319*
.373*
Number of Desaturations
Humber of Desaturations
>4 .287
2.10
.368*
.430*
*p<.01


90
sleep disordered breathing (middle-aged, snoring males).
A prevalence of 62& was observed for occurrence of at
least one event, while 135 had high numbers of
apnea/hypopnea, data which are consistent with an earlier
report drawn from heavy snoring males. Most events
occurred in light slow wave or REM sleep. No effect of
age on level of sleep disordered breathing was observed,
although increasing weight was linked to nocturnal
respiratory events. When the sample was divided by level
of apnea/hypopnea, subjects with high numbers of
apnea/hypopnea suffered significantly exacerbated oxygen
desaturation, although no significant differences were
noted for weight or age among the groups.
Nocturnal Respiratory and Health Variables
The finding of only very limited relationships
between nocturnal respiratory events and health variables
must be tempered by an acknowledgement of possible bias
against such findings in this sample. Specifically, the
sample was selected to be normal and healthy, a criterion
which undoubtably restricted the range of possible health
pathology, and hence probably attenuated the correlational
findings. Additionally, the self-report health checklist,
the Cornell Medical Index, cannot be regarded as a
powerful measure of health pathology. In a very real
sense then, the odds were against finding reliable
relationships between nocturnal respiratory and health
variables.


51
Table 2-1. Interrater agreement for respiratory events of
46 snoring males.
Consensus
Initial
Disagreements
Scoring
Agreements
#1
n
Note
(178)
_
.
Scored by
consensus
91
90
4
6
Consensus
added 1
78
76
1
1
Consensus
added 2
61
52
3
8
Consensus
added 9
57
56
9
2
Consensus
added 1
32
29
3
3
Consensus
added 3
28
27
0
2
Consensus
added 1
25
25
0
0
19
16
4
0
Consensus
added 3
17
17
0
0
15
16
1
0
Consensus
added 1
13
13
0
0
7
7
0
0
4
t
0
0
4
4
0
0
3
2
0
1
Consensus
added 1
3
3
0
0
3
0
7
2
Consensus
added 3
3
3
0
0
3
3
0
0
1
1
0
0
1
1
0
0
1
1
0
0
1
1
0
1
1
0
0
1
Consensus
added 1
1
1
0
0
1
1
0
0
1
1
0
0
474(652)
450
37
36
Note: 26 of the 46 subjects experienced at least 1 apnea or
hypopnea.


CHAPTER ONE
INTRODUCTION
Sleep apneas are respiratory pauses during sleep.
According to the Association of Sleep Disorders Centers
Nosology of Sleep Disorders (ASDC, 1979) a sleep apnea
syndrome (SAS) is a potentially lethal condition charac
terized by multiple apneas, excessive daytime sleepiness,
alterations of consciousness, and cardiopulmonary
complications. The apnea episodes are thought to be the
causal factor behind the other deficits associate^ with
SAS (Guilleminault et al., 1978).
Polysomnographic recordings in various populations
have revealed the occurrence of sleep apneas in the
general population, and particularly in aging popula
tions. Estimates of the prevalence of at least a few
sleep apneas in older subjects have ranged as high as 75£
(Carskadon and Dement, 1981b). Carskadon et al. (1SSC)
have suggested that sleep apneas in older subjects may be
implicated in insomnia, cardiac ailments, and even
senility.
However, certain reports have documented the presence
of multiple sleep apnea episodes in subjects without
apparent pathology (Orr et al., 1979)* This evidence
obscures the role of sleep apneas in the other pathologies
1


84
Pable ;;-16-con tinuec.
Criterion
Predictor
Beta
Sig
Delayed Recall
Age
-.2623
.074
Key Figure
Apnea Index
-.1519
259
Apnea+Hypopnea Inoex
Number of
-.3286
.085
Desaturations >4%
-0795
b57
Sleep Efficiency
.6832
.018
Number oi Stage 0 Periods
2487
354
Time % Stage 0
455S
.172
Time % Stage 2
-.0405
.602
Number oi Stage Changes
-.5054
.057


43
levels of hypoxia affect coordination, memory, attention
span, and may result in headache, drowsiness, and nausea.
The effects of milder, chronic hypoxia are summarized
by Gibson et al. (1981) At the oxygen pressure equiva
lent of 5*000 feet, impaired dark adaptation is noted. At
10,000 feet decreased concentration, hyperventilation, and
short term memory deficits are noted. By 15*000 feet,
euphoria and loss of both coordination and critical judge
ment are observed. Christenson et al. (1977) found signi
ficantly less visual signals detected at 17% O2 than at
21%, suggesting reduced vigilance. McFarland (1937)
studied men in ascents to high altitudes and concluded
that subjects with slower rates of ascent performed better
on unspecified psychological tests. West (1984) reports
data collected on an expedition to Mt. Everest. Results
of a neuropsychological battery indicated deficits in
finger tapping, verbal fluency, verbal learning, short
term memory, and expressive language. Thus the general
affects of chronic mild hypoxia are thought to be deficits
in dark adaptation, short term memory, critical judgement,
motor coordination, verbal fluency, and vigilance.
One group of chronically hypoxia patients, those with
chronic obstructive lung aisease (COLD), has been rela
tively well studied with sensitive neuropsychological
instruments and is aeserving of a more painstaking
review. Krop et al. (1973) studied COLD patients before
and after they received oxygen supplements. The WAIS,


15
duration of 20 seconds and occurred most frequently
(30/60) in Stages 1 and REM. Eight of 20 aging females,
in the report by Block et al. (1980), had sleep apneas
with a mean duration of 19 seconds. Sixty-two percent of
apnea episodes occurred in Stages 1, 2 or REM. Thus in
middle aged normal subjects, incidence of sleep apnea
appears to rise, with a mean duration of between 13 and 20
seconds, perhaps slightly longer than in younger groups.
Again, apneas occurred mostly in light (Stages 1 and 2)
and REM sleep.
Duration data on aging normals are limited.
Carskadon and Dement's (1981b) study of aging normals (M =
72, F = 74) does not report duration of apnea events.
Similary, Ancoli-Israel et al. (1982) do not provide this
information on their aged normals (x = 71.5). Bixler et
al. (1982) noted a high incidence of apneas in their sub
jects over 60, but failed to report a mean duration for
this group. Given the sketchy details reported for this
group, it is difficult to draw any conclusions about
changes in temporal duration or placement of apneas in
aging normals.
To summarize the temporal data in normals, apneas
appear somewhat rarely in normals below 30, with a mean
duration of 15 seconds. Incidence of apneas increase
between 30 and 60, with a wider variation in mean dura
tions reported (13-20 seconds). Apneas occur most
frequently in light and REM sleep in normals of all ages.


130
Sara, S. (1974). Delayed development of amnestic behavior
after hypoxia. Physiology and behavior, 13* 693-696.
Schmidt, H. (1982). Pupilometric assessment of disoraers
of arousal. Sleep, 5., 157-164.
Schroeder, J., Motta, J., & Guilleminault, C. (1978).
Hemodynamic studies in sleep apnea. In C. Guilleminault
& W. Dement (eds.), Sleep Apnea Syndromes. New York:
Alan R. Liss Inc.
Sicklesteel, J., Zorick, F., Wittig, R., Conway, W., & Roth,
T. (1981). Daytime somnolence and insufficient
sleep: A case series. Sleep Research, 10, 233-
Smallwood, R., Vitiello, M., Giblin, E., & Prinz, P.
(1983). Sleep apnea: Relationship to age, sex, anc
Alzheimers Dementia. Sleep, 60), 16-22.
Smirne, S., Francheschi, M., Comi, G., & Leder, F.
(1980). Sleep apneas in Alzheimers disease. Sleep
Research, 9., 224.
Sullivan, C., & Issa, F. (1980). Pathophysiological
mechanisms in obstructive sleep apnea syndrome. Sleep,
M3/4), 235-246.
Tauber, B., & Allweise, C. (1975). Effects of acute
hypoxia on memory. Israeli Journal of Medical Science,
11, 71.
Tilkian, A., Motta, J., & Guilleminault, C. (1978).
Cardiac arrhythmias in sleep apnea. In C. Guilleminault
& W. Dement (eds), Sleep Aonea Syndromes. New York:
Alan R. Liss Inc.
Van Liere, E., & Stickney, J. (1963). Hypoxia. Chicago:
University of Chicago Press.
Webb, P. (1974). Breathing during sleep. Journal of
Applied Physiology, 17(6), 899-903.
Webb, W.B. (1975). SleepThe Gentle Tyrant. Englewood
Cliffs, N.J.: Prentice-Hall Inc.
Webb, W.B. (1982). Measurement and characteristics of
sleep in older subjects. Neurobiology of Aging, 3_, 311-
319.
Wechsler, D. (1945). A standardized memory scale for
clinical use. Journal of Psychology, 19, 87-95*


33
Data on desaturation are interesting from several
perspectives. First of all, Sleep Apnea Syndrome subjects
seem to desaturate more heavily than normals, with Berry
and Block's symptomatic group desaturating up to 50% and
the subjects of Garay et al. desaturating about 22%,
compared with normal desaturation levels of 10-12%.
Secondly, amount of desaturation distinguished symptomatic
(hypersomnolent) from asymptomatic subjects in the data of
both Orr et al. and Berry and Block. Additionally, two
points relevant to methodology emerge. Measures of oxygen
desaturation seem crucial to understanding sleep apnea and
its effects, and ear oximetry appears to provide a proven,
valid measure of this variable.
Weitzman (1979) has suggested that the pathological
process in SAS is the hypoxia of desaturation in apneac
episodes. While cognitive and intellectual changes in SAS
have been only anecdotally reported (Guilleminault et al.,
1976), hypoxia does seem a plausible potential cause of
these changes. Thus oxygen desaturation appears to be a
reliable consequence of sleep apnea syndrome, while cogni
tive/intellectual changes are putative sequelae.
In summary, a review of deficits found in sleep apnea
syndrome subjects suggests that cardiopulmonary complica
tions, excessive daytime sleepiness, oxygen cesaturation
and cognitive/intellectual changes are frequently asso
ciated with this syndrome. An assessment of these
variables in a subclinical population seems indicated by


Table B-5
Ueuropsycholgica1 variables
V/AiS lw
Sub j.
II
Verba 1
Ferior-
mance
Full
WI'ISa
HQ
UI L>KVKC
Digit
Span
1
118
089
11b
157
11
11
7
2
07 5
07 b
074
Ob 7
02
00
4
3
1 2.'j
11b
121
12b
11
13
8
4
135
094
127
120
11
11
8
5
120
10b
115
135
10
11
8
b
122
110
118
124
12
14
7
7
108
10b
108
105
Ob
09
b
8
101
117
108
10b
Ob
11
7
9
120
140
130
129
Ob
08
8
10
128
122
12?
137
15
14
7
11
134
154
13b
124
11
12
8
12
1U2
112
10b
101
Ob
14
7
13
102
090
100
095
05
04
b
14
140
151
140
143
09
15
8
15
104
103
104
107
09
11
5
1b
115
125
118
089
05
10
5
17
113
111
113
094
05
14
b
18
122
110
118
112
11
11
7
19
095
121
107
100
Ob
00
b
20
151
159
1 5b
145
11
13
7
21
120
113
118
118
08
10
7
22
124
125
125
13o
10
10
8
23
119
118
120
142
10
14
8
24
127
119
125
120
10
15
7
25
122
120
122
120
11
11
7
2u
120
100
110
105
11
1u
r
(
46 snoring males
Hooper
L'l Finger Tapping
Hight Lelt Verbal
WOSb Hand Hand Fluency
17
24
4
44
40
03
Cb
12
1
25
2'j
04
25
30
5
52
47
08
19
22
1
54
52
1U
5b
25
5
55
55
15
24
28
4
59
58
13
1b
2
5
53
58
12
14
23
5
5b
53
07
24
28
5
55
53
20
28
28
4
b1
55
13
25
28
4
bO
53
1b
29
27
5
04
51
10
1b
24
3
5b
32
09
20
28
5
55
32
15
30
29
1
59
b1
13
24
29
5
53
30
12
24
29
5
57
59
12
27
28
2
7b
bo
17
22
22
2
b1
b2
03
29
2/
4
72
r *
/ c
19
25
25
2
bb
o4
09
28
25
1
5b
32
13
31
27
5
57
35
20
2b
27
5
b4
bO
18
1b
24
4
54
35
12
2b
25
s
bO
49
14


37
significant declines in average sleep latencies (Carskadon
and Dement, 1981a). Finally, total sleep deprivation led
to a drastic decline in MSLT values, with averages
plummeting below one minute (Carskadon and Dement,
1979a). Mean MSLT values showed recovery to baseline
values after a full night's sleep for partially (5 hrs)
deprived subjects, and recovery after two nights of ad lib
sleep for totally sleep deprived subjects (Carskadon and
Dement, 1979; 1981a). This body of evidence suggests
that the MSLT is a reliable and valid index of sleepiness
in normal subjects.
With its internal validity established, the MSLT was
applied to somnolent populations. Dement et al. (1978) as
well as Richardson et al. (1978) demonstrated that SAS and
narcoleptic patients have significantly shorter latencies
to sleep on the MSLT than normal subjects do. Hartse et
al. (1979) found SAS subjects to have an average latency
to Stage 1 of 2.5 minutes, while narcoleptics had a mean
of 3.2 minutes. Both these latencies were significantly
lower than those of a miscellaneous control group whose
latencies were about 11.2 minutes. Hartse et al. (1980)
found latencies to Stage 1 sleep as follows for various
patient groups: Normals 12.2 m, Insomniacs 16.2 m, SAS
2.5 m, and Narcoleptics 3.2 m. The SAS and narcoleptic
patients had significantly shorter latencies to Stage 1
sleep than insomniacs or normals. Roth et al. (1980)
compared 10 SAS patients with 10 age matched controls.


26
Cardiopulmonary Deficits
Cardiopulmonary complications have been reported as
concomittants of SAS (Guilleminault et al., 1978).
Coccagna et al. (1972) reported high incidences of hyper
tension and congestive heart failure in their SAS sub
jects. Schroeder et al. (1978) studied 22 SAS patients
with a mean age of 47. Six subjects had waking systemic
hypertension, and several had waking cardiac abnorma
lities. During sleep studies, 20 patients developed
significant rises in systemic arterial pressure which
cycled with episodes of apnea, while 21 patients developed
pulmonary arterial hypertension. Tracheostomy reversed
all these abnormalities, strongly implicating the apneas
in their etiology. Tilkian et al. (1978) studied 25 male
SAS subjects with a mean age of 44* Twenty-four of the 25
showed sleeping sinus arrhythmias, while 9 developed more
serious symptoms such as asystole. Seventeen of these
patients received tracheostomy, which abolished all of
these abnormalities. Experimental occlusion of the
tracheostomy during sleep in 6 of these patients resulted
in a return of the cardiopulmonary symptoms. Zwillich et
al. (1982) studied six consecutive male SAS patients. In
all patients, bradycardia accompanied any apnea with
significant desaturation. Additionally, a significant
correlation was noted between degree of desaturation and
severity of bradycardia. Bradycardia was abolished or
attentuated during administration of O2 enriched air.


53
period; mean REM period; mean REM cycle length; % slow
wave sleep; and number of stage changes.
Measures of health status included height, weight,
diastolic and systolic blood pressure, as well as the
Cornell Medical Index. From the CMI, reports of hyper
tension and heart trouble were noted, as well as the total
number of symptoms endorsed. Additionally, the scores for
the scales tapping respiratory, cardiac, and neurological
symptoms were tabulated separately.
Measures of neuropsychological status included the
WAIS Verbal and Performance IQs, the WMS Memory Quotient,
delayed recall of the logical stories and the visual re
production subtests of the WMS, delayed recall of the Rey
complex figure, number of words generated in the verbal
fluency test, number of correct sorts of the Wisconsin
Card Sort, mean number of taps for the right and left
handed finger tapping tests, and lastly the number of
correct trials on the Hooper test.
These neuropsychological tests were chosen with two
aims. First, an attempt was made to sample from each of
several loosely bounded areas of cognitive skill. These
included intelligence (both verbal and non-verbal),
memory, both immediate and delayed (verbal and non
verbal), visuo-perceptual/organizational skills, language,
and frontal self-regulatory skills. A second emphasis was
on a relatively detailed analysis of areas previously
shown to be sensitive to hypoxia. These sensitive areas


Table 3-14.
Significant (p<.05) partial correlations, controlling lor age, between
nocturnal respiratory and cognitive variables in 46 snoring males.
Apnea
Apnea/
ilypopnea
Index
Index
WAIS Verbal IQ
WAIS Performance IQ
-.311
-.414*
Wechsler Memory Scale
Memory Quotient
-.290
Delayed Recall Logical
Stories
-.343
Delayed Recall Visual
Reproductions
-.416
Digit Span
Delayed Recall Rey Figure
-.284*
-.305
Hooper Test
Wisconsin Card Sort
-338
Finger Tapping Right Hand
-.259
Finger Tapping Left Hand
-.258
Verbal Fluency
-.364*
Mean
Humber
of
Fvent
Saturations
L'esatura
tions
Time
High
Low
>A%
>_10>>
.274
-.321
.421*
.521*
-.461*
.302
.251
-.311
-.260
.260
-.276
.437*
.349*
.489*
-.337
.369*
.252
.332
. 265
.274
.329
.365*
-.329
vO
*p<.01.


85
non-verbal memory. Addition of variables reflecting sleep
fragmentation to the analysis indicated that both
nocturnal respiratory and sleep variables make significant
and independent contributions to prediction of
cognitive/neuropsychological scores.


APPENDIX E
RAW DATA TABLES


18
associated variables include sex, age, snoring and
obesity. The data available on these variables and their
relationships to sleep apnea activity will be reviewed,
followed by a brief discussion of their relevance to
experimental methodology.
Regarding sex variables, as previously noted,
Guillerainault et al. (1978) found 96% of their SAS
patients to be males. In the Block et al. (1979) study of
nonsleep disturbed subjects, 40% of the males in the
sample had at least one episode of apnea, while only 15%
of the females so suffered. Bixler et al. (1982) noted a
trend toward more sleep apnea activity in males. Block et
al. (1980) studied 20 post menapausal females and found
apnea activity in 40% of them, leading them to suggest
that some factor associated with menapause (e.g., pro
gesterone) might somehow "protect" premenapausal females
from apneas, and that this factor along with its benign
influence was lost with menapause. This explanation pre
dicted the low level of apnea activity in younger females,
as well as the heightened apnea activity of aging
females. While the data of Carsakdon and Dement (1981b)
show aged females to be close to aged males in percentage
exhibiting sleep apnea activity, a report by Smallwood et
al. (1983) found no evidence of sleep apnea activity in
their 6 elderly (post menapausal) females, while elderly
males showed significant levels of apnea activity. The
limitations of the sample size in the Smallwood et al.


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EOGSOKC06_8RP790 INGEST_TIME 2012-09-24T13:31:42Z PACKAGE AA00011824_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


Table 3-4. Means and standard deviations for selected demographic'ana nocturnal
respiratory variables in 46 snoring
apnea/hypopnea.
males
grouped
by level
of
No Apnea/
Low
Apnea/
High Apnea/
Nypopnea
Hypopnea
Hypcpnea
Age
52.3(12.7)
46.1
(13-5)
56.5
(9.3)
Weight (lbs)
185.6(34.8)
188.6
(37.6)
203-3
(24.3)
Weight:Height Ratio
2.7 (0.4)
2.7
(0.5)
2.9
(0.4)
Education (years)
15.2 (2.9)
14.6
(2.0)
13.8
(2.6) .
Number of Apneas
0
5.0
(7.7)
24.5
(34.1)a,fc
Number of Apneas+Hypopneas
0
7.3
(9.2)
81.7
(52.6)
Apnea Index
0
0.9
(1.1)
4.5
(5.y)*.b
Apnea+Hypopnea Index
0
1.4
(1.5)
19.5
(19 4)0 f3
Seconds in Apneas/Hypopneas
0
129.4(167.1)
2325 *5(1469.5)aD
Meen Apnea Duration
-
13.4
(4.0)
17.8
(1.3)
Mean Hypcpnea Duration
-
29.6
(21.2)
33.2
(14.5)
Mean High Saturation
95.5 (1.7)
95.3
(1.3)
94.9
(1.2) .
Mean Low Saturation
94.3 (2.1)
93.1
(1.6)
89.0
(4.6)0>l)
Number of Desaturations _>4/
29.5(54.1)
56.2
(36.6)
143.2
(1O5.9)0,b
Number of Desaturations >J\0%
0.2 (0.7)
3.5
(6.3)
51.0
(48.7)ab
Significantly different from no sleep apnea/hypopnea.
Significantly different from low sleep apnea/hypopnea.


99
with greater physical deterioration (Feinberg et al.,
19^7), subjects selected for health do not display this
effect. A third embarrassment for the sleep fragmentation
hypothesis lies in the study of Orr et al. (1979) who
described one group of symptomatic sleep apnea syndrome
subjects and a group of asymptomatic snoring subjects that
were matched on number of nocturnal events. Presumably,
arousals resulting from events were matched in the two
groups, yet the symptomatic subjects had serious impair
ments relative to the asymptomatic groups.
Despite these difficulties, the present study found
that one of the five cognitive scores significantly pre-
cicted by the nocturnal respiratory indices was also
correlated with sleep efficiency on the experimental
night. However, given the earlier observed problems with
the experimental nights* sleep, the inconsistent past
evidence, and the limited (1) relationship observed, this
finding remains inconclusive. A more ambitious study,
examining sleep and respiration across several nights, is
necessary to clarify this issue.
An alternative, although admittedly speculative,
explanation of the cognitive changes lies in the cerebral
hypoxia accompanying respiratory events. Brain dysfunc
tion is known to occur in mild to moderate hypoxia. Y/hile
the brain possess various homeostatic mechanisms to com
pensate for chronic hypoxia (hence the adaptation of long
term residents of mountainous regions) it may be that the


100
multiple brief, acute desaturations experienced by the
subject with apnea/hypopnea fail to activate these protec
tive mechanisms effectively (Gibson et al., 1981).
Recently, a body of evidence has accumulated which indi
cates that turnover of certain neurotransmitters is drama
tically altered by acute hypoxia. The cholinergic synthe
tic pathway, which results in the formation of the neuro-
transmitter acetylcholine (ACh) appears to be particularly
sensitive to the effects of acute hypoxia (Gibson et al.,
1981)* That ACh may be the neurotransmitter most affected
by hypoxia of nocturnal events is provocative, in light of
the "Cholinergic hypothesis" (Bartus et al., 1985) of the
memory and cognitive impairment often found in aging sub
jects. This formulation notes that cholinergic blockade
in young subjects inhibits memory, while various aspects
of the cholinergic synthetic pathv/ay are altered in aged
subjects with memory and cognitive deficits. Alterations
in the cnolinergic synthetic pathway are thought to under
lie the cognitive changes in both populations. It may be
speculated that the cognitive dysfunction found in the
present subjects with multiple apnea/hypopnea is also a
product of a disruption of cholinergic synthesis.
To summarize the logic of the "cerebral hypoxia hypo
thesis," it is first noted that subjects with nocturnal
respiratory events experience multiple, acute hypoxic
episodes. This acute hypoxia appears to evade cerebral
protective mechanisms which normally respond to chronic


19
report might explain this discrepancy as representing a
sampling bias. Taken together, this evidence strongly
supports the notion that in young and middle aged popula
tions, sleep apnea is a male phenomenon. Additionally,
evidence points toward a closing of the gap between
females and age matched males as the former group passes
menapause.
A second factor which has been implicated in sleep
apnea activity is age. Webb (1974) found that his sub
jects exhibiting sleep apnea activity were significantly
older than his full sample. This was also the case for
the subjects of Kales et al. (1982) and Ancoli-Israel et
al. (1982). Smallwood et al. (1983) found that sleep
apnea was an age dependent phenomenon with male subjects
over 50 exhibiting significantly more apneas than those
under 50. Other evidence includes that of Bixler et al.
(1982) who found a positive correlation between age and
apnea in their full sample, as did Block et al. (1980) in
their post menapausal females. It seems clear from these
data then that increasing age is a factor predisposing for
sleep apnea activity.
Snoring has been closely associated with sleep apnea
(Lugaresi et al., 1982). This may be due to a causal
relationship or a single common pathway. Lugaresi et al.
propose the latter explanation. Specifically, they
suggest that both snoring and SAS are due to a sleep
induced stenosis of the upper airway. With this premise


63
apnea or hypopnea, while 13% had high levels of
apnea/hypopnea. Most events occurred in light slow-wave
or REM sleep. No effect of age on level of sleep
disordered breathing was observed, although increasing
weight was correlated with nocturnal respiratory events.
When the sample was stratified by level of apnea/hypopnea
activity, subjects with high levels of events had
significantly more events and desaturations than the
remaining subjects, although no significant differences
were observed on age, weight or education.
Nocturnal Respiratory and Health Variables
health variables including blood pressure readings
and several scores derived from the Cornell Medical Index
(overall number of symptoms endorsed, as well as number of
symptoms endorsed from the respiratory, cardiopulmonary
and neurological subscaies) were evaluated initially
through a matrix of Pearson correlations which included
the respiratory variables (apnea index, apnea + hypopnea
index, seconds in events, as well as mean high and low
saturation and number of desaturations >A% and >10%)*
From a total of 42 correlations, three were significant at
p<.05 or less: Systolic blood pressure and mean low
saturation (r=-.21); Systolic blood pressure and number
of desaturations >A% (r=.373); Overall number of symptoms
endorsed on the CMI and mean high saturation (r=-.254).
As a further exploration of possible relationships between
blood pressure and apnea/hypopnea activity, subjects were


52
low score seemed obvious. In order to assess the quan
titative "impact" of numerous desaturations, a sum of 4%
and 10% desaturation was tabulated. These selected
parameters for thumistor and saturation variables allowed
reliable scoring and a description of several aspects of
nocturnal respiration.
Dependent measures fell into three broad classes:
measures of sleep/wake status, measures of health status,
and measures of neuropsychological status. Measures of
the sleep/wake cycle included one week's sleep logs
(derived variables included number of reports of daytime
sleepiness, hours asleep, estimated sleep latency, number
of wakenings after sleep onset, and minutes of waking
after sleep onset), a sleep questionnaire (derived
variables including hours of sleep usually obtained,
number of naps, hours napping, quality of sleep,
restfulness on awakening, depth of sleep, and amount of
daytime sleepiness), the mean of the several Stanford
Sleepiness Scale ratings, the two multiple sleep latency
tests (derived variables included mean sleep latency and
total minutes asleep in both naps) and the overnight EEC
variables derived from the scoring system of Agnew and
Webb (1972). Electroencephalographic variables included
time in bed; pure sleep time; sleep efficiency index;
number stage 0 periods; time % stage 0; sleep latency;
time % stages 1, 2, 3, 4 and REM; latency first REM


89
these data somewhat tenuous- This finding needs
replication before any conclusions may be drawn.
Ninety-eight per cent of apneic events occurred in
light slow wave and REK sleep, an observation which is not
surprising given that subjects spent nearly 80£ of their
sleep in these stages. The preponderance of events in
these stages of sleep is consistent with the bulk of past
reports in this area (e.g, Bixler et al., 1982; Block et
al., 1979; Krieger et al., 1983).
When subjects were divided into groups by level of
apnea/hypopnea, significant differences emerged on oxygen
desaturation variables. Specifically, comparisons
indicated that the group with a high level of
apnea/hypopnea events experienced deeper desaturations
than the remaining two groups. This difference occurred
despite a lack of differences on age and weight among the
three groups, and appears to implicate the respiratory
events in the genesis of the desaturations. This is
consistent with an earlier report by McGinty et al. (1982)
who demonstrated that subjects with significant levels of
sleep disordered breathing were impaired on oxygen
desaturation variables, relative to subjects without
respiratory events. Subjects with heavy snoring and
multiple nocturnal apneic respiratory events appear to be
at risk for nocturnal oxygen desaturation.
In summary, the present sample of 46 snoring males
appears representative of those predisposed to experience


97
included Krop et al. (1973) and Krop et al. (1977) both
of which notea xhr.it continuous nocturnal oxygenation
treatment significantly improved neuropsychological scores
in normal aged and chronic obstructive lung disease
patients, respectively. The present results are also
consistent with a report by Grant et al. (1982)
demonstrating considerable neuropsychological impairment
in chronic obstructive lung disease patients, a group
knov/n to experience hypoxia which is significantly
exacerbated during sleep (Block, 1981). Additionally,
West (1984) found impairments in finger tapping speed,
shcrt-term memory, and expressive verbal fluency in a
group of climbers ascending F:t. Everest. Taken together,
the present results and past reports present a consistent
picture of relationships between both daytime and
nighttime hypoxia and neuropsychological impairments.
Although the selection requirement of general good
health ruled out the inclusion of subjects with a full
blown sleep apnea syndrome, six of the present subjects
demonstrated levels of respiratory disturbance sufficient
to meet current diagnostic criteria for SAS-(A+H)I>5
These subjects showed significant impairments, relative to
those with no sleep apnea/hypopnea, on measures tapping
non-verbal intelligcnce, delayed verbal and non-verbal
memory, and verbal fluency. These data are suggestive
that heavy snoring males with multiple apneas/hypopneas
may be at risk for cognitive sequelae. If the impairments


UNIVERSITY OF FLORIDA
illlllllllH
3 1262 08554 7908


75
significant overall p value, univariate regressions are
used to individually predict criterion variables with the
multiple predictors. Contribution of individual
predictors may also be examined in these univariate
regressions.
The results from the multivariate regression of the
above noted cognitive scores and selected nocturnal
respiratory variables (Apnea index, Apnea + hypopnea
index, Mean high saturation, Number of desaturations >4%,
Age, Veight:Height ratio) are presented in Table 5-13* An
overall Hotellings T2 score was significant (p<.003),
allowing examination of the univariate regressions.
Criterion variables which were significantly preaicted by
the respiratory and demographic variables included WAIS
PIQ, VMS MQ, delayed recall of Hey figure and visual
reproductions, Hooper, and Wisconsin card sort. These
relationships reflected negative relationship between
overnight respiratory indices and neuropsychological
scores. Respiratory indices formed significant components
of the predictor combinations for all these variables save
Hooper score, suggesting relationships between nocturnal
respiratory indices and non-verbal intelligence, verbal
and visual memory, and ability to shift sets while problem
solving.
Because age is correlated with nocturnal respiratory
parameters in past reports, and also with neuropsycho
logical scores in other samples, the various overnight


3
et al. (1976) first suggested that the occurrence of 30 or
more apneas in 7 or more hours of sleep was diagnostic of
SAS, while Guilleminault et al. (1978) proposed the use of
an apnea index (AI; number of apneas/number of hours
sleep) in excess of 5 as a cut-off score to diagnose the
syndrome. Later, Carskadon et al. (1980) combined apneas
and hypopneas (a reduction in airflow of 50% or more at
the mouth and nose) to form an apnea + hypopnea index
(AHI) with 5 utilized as a diagnostic level of sleep apnea
syndrome. As can be appreciated from the preceding dis
cussion, the diagnostic criteria for SAS have undergone a
rapid evolution, with remarkably little validation work
published.
History and Characteristics of Sleep Apnea Syndrome
Sleep Apnea Syndrome falls within the larger category
of Sleep Disordered Breathing (Block, 1980). Respiratory
disorders within this category involve changes in
breathing patterns during sleep. Syndromes described
within this category include Pickwickian Syndrome; Sleep
Apnea Syndrome; Chronic Obstructive Lung Disease; and
syndromes of Primary and Secondary Alveolar
Hypoventilation (Block, 1980; Guilleminault et al.,
1976).
The first disorder described in this group was the
Pickwickian Syndrome. These patients were initially
identified by a chronic daytime hypersomnolence, obesity,
Cheyne-Stokes breathing, hypercapnia and hypoxemia. This


17
had apneas with a mean duration of 25.9 seconds, while
symptomatic subjects had a mean duration of 20.8 seconds.
Several characteristics emerge from the group of
reports on SAS subjects. There is wide variability in
duration of apneas, with reported means ranging from 16-30
seconds. The data available on placement of apneas in
sleep suggest that in SAS subjects most apneas occur in
light and REM sleep.
Comparison of data on normal and SAS subjects reveals
substantial overlap between mean duration of apneas in SAS
subjects (13-20 seconds) and SAS subjects (16-30
seconds). Particularly interesting was the Orr et al.
(1979) study which demonstrated a lower mean duration of
apneas for hypersomnolent subjects than for nonhypersomno-
lent subjects. Data on temporal placement of apneas show
most events occurring in light or REM sleep for both
normal and SAS subjects. Taken together, these data
suggest that substantial overlap exists on temporal
t
variables between normal and SAS subjects indicating a
limited utility for diagnostic purposes.
Variables Associated With Sleep Apnea
Syndrome anc Activity
It was noted above that several variables have been
suggested as associated with increased incidences of sleep
apneas and SAS. These variables represent potential
"markers" of subpopulations which might experience higher
incidences of sleep apnea activity. These primary


92
Nocturnal Respiratory and Sleep/Wake Variables
EEG Data
The overnight sleep in the laboratory for these
snoring subjects was noteworthy for a general shift
towards lighter sleep than usual. A longer sleep latency,
more awakenings, more light sleep, and less slow wave and
REM sleep than normal are apparent. These findings
parallel the report by Block et al. (1979) who found
shifts to lighter sleep and more awakenings than in a
sample of normal subjects. These changes were said to
result from a first night in a strange bed and the
discomfort of the multiple clips and sensors used to
monitor respiration. Recording montages were identical
for the present subjects, and these two factors
undoubtably influenced the sleep of the present subjects
as well.
An intriguing observation was a lack of covariation
between respiratory and EEG sleep variables derived from
the recording night. This finding is counterintuitive, as
one would expect arousals from the respiratory events to
influence the sleep characteristics of subjects with
events. However, correlational procedures failed to
support this notion. It may be that a ceiling effect was
exerted on sleep measures sensitive to disturbance, one
that resulted from the first night effect compounded by
tne discomfort of the recording procedure. The already
elevated disturbance sensitive sleep variables might not


6
definitively demonstrated to be the necessary and suffi
cient cause of the syndrome. Thus, for the purposes of
the following review, the focus will be on occurrence and
level of sleep apnea activity in both sleep apnea syndrome
subjects and other populations. The full range of sleep
apnea activity will be reflected by noting the presence of
sleep apnea activity, and an indication of the quantita
tive level of the activity will be presented in the apnea
index (AI) or the apnea + hypopnea index (AHI). Recall
that an AI of 5 has been frequently utilized as a cut-off
score for identifying SAS patients.
A review of the literature suggests that at least
three sampling strategies have been used for studying the
incidence of sleep apnea activity. Subjects have been
selected on the basis of one of the following three
criteria: 1) absence of sleep or respiratory complaints,
2) presence of sleep or respiratory complaints, or 3)
sampling without regard to these variables.
because diagnosis of SAS is often dependent on number
of sleep apneas (or level of SaA), most researchers report
number of apneas observed, although some report only on
number of subjects who exceed one of the criteria for
SAS. A review of studies with samples of 20 or greater
will be presented with careful rote of the sampling
strategy utilized, as well as the dependent measure of
apnea.


44
Wechsler Memory Scale (WMS), Bender-Gestalt, Bender
Gestalt Interference Procedure, Facial Recognition and
Finger Tapping tests were administered to COLD and control
subjects. Subjects receiving supplementation improved
significantly more than controls on WAIS Full Scale IQ,
Performance IQ, Wechsler Memory Quotient, both Bender
procedures and Finger Tapping. Grant et al. (1982)
studied another group of COLD patients. They administered
the Halstead-Reitan Battery, Aphasia Screening Test,
Trailmaking Test, Russell modification of the Wechsler
Memory Scale, Tactile Memory test, and the WAIS. In COLD
subjects, deficits were found relative to controls on all
variables except Reitan Rhythm, Aphasia Screening and WMS
Logical stories. Grant et al. concluded that COLD
patients exhibited deficits on a global impairment rating,
attention, abstracting ability, complex perceptual motor
tasks, simple sensory and motor tasks, and memory tasks.
Low but significant negative correlations were reported
between several measures of oxygen saturation and both
Halstead Impairment Index ana Global Impairment Rating.
Grant et al. speculated that floor effects resulted in low
variability in PA2 levels among COLD patients and kept
these correlations low. They further suggested that the
neuropsychological deficits observed in these patients
resulted from the oxygen want found in these subjects.
Thus data available on humans indicate that dark
vision, memory, motor coordination, critical judgement,


74
In summary, no between group differences were noted
for either sleep questionnaire or log variables when
subjects were grouped by level of apnea/hypopnea.
Correlational procedures suggested relationships between
several respiratory variables and reports of napping and
wakening after sleep onset on the sleep questionnaire.
These findings were cross validated in the correlational
analysis of sleep log data, in which reports of napping
and wakening after sleep onset were related to respiratory
variables, as were total bedtime and sleep latency.
Nocturnal Respiratory and Neuropsychological Variables
Scores derived from the cognitive/neuropsychological
battery included WAIS Performance IQ, WAIS Verbal IQ,
YJechsler Memory Quotient, delayed recall of logical
stories, delayed recall of visual reproductions, delayed
recall of Rey complex figure, digit span, Hooper score,
Wisconsin card sort score, finger tapping right and left,
and Verbal fluency.
As a preliminary analysis, these scores were compared
with selected respiratory and demographic variables in a
multivariate regression procedure. In this type of
analysis, a set of predictor scores is used to
simultaneously predict a set of criterion variables while
controlling for intercorrelations amongst the measures.
An overall test of significance, such as the Hottelings
T2, establishes the probability of obtaining the observed
relationships from chance alone. Upon obtaining a


102
at risk for cognitive sequelae. These data may be taken
as limitec support for the anecdotal reports of cognitive
changes in SAS patients. Two explanations of possible
underlying mechanisms of the cognitive changes may be
considered. The first, the "sleep fragmentation hypo
thesis," suggests that the cognitive deficits result from
the multiple arousals experienced by the subject. This
notion is suspect as a plausible explanation because sleep
deprivation and sleep measures tapping sleep fragmentation
fail to intercorrelate consistently with cognitive scores
in other populations. Additionally, the extremely limited
relationship observed in the present study is less than
conclusive. It is proposed that cerebral hypoxia induces
alterations in synthesis of ACh, a neurotransmitter impli
cated in cognitive abilities in other populations, and
thus causes the cognitive changes. The cerebral hypoxia
theory remains speculative, although several testable
hypotheses may be generated from it. Further research
into factors affecting the sleeping brain may prove
heuristic for neuropsychologists.


16
Temporal data on subjects diagnosed with sleep apnea
syndrome (>50 or AI>5) primarily include patients between
the ages of 30 and 60. The reason for the lack of older
SAS subjects is unclear, particularly in light of data
reported above, demonstrating very high percentages of SAS
in certain groups of aging subjects.
Garay et al. (1981) studied 13 sleep apnea syndrome
subjects. They divided these 12 males and 1 female into
two groups, based on their daytime resting CO2 levels.
Subjects who were eucapnic during the day had a mean apnea
duration of 16 seconds, while those who were hypercapnic
had a mean duration of 17 seconds. No data on distribu
tion relative to sleep stages were reported.
Guilleminault et al. (1978) studied 50 predominantly male
(x age 48) sleep apnea syndrome patients. Apneas in this
population averaged 22 seconds in duration. Apneas were
longest in REM, and few occurred in Stages 3 or 4 sleep.
Fujita et al. (1981) studied 12 male SAS patients with a
mean age of 43* Apnea episodes averaged 23 seconds in
duration. Weitzman et al. (198C) studied 10 SAS males
whose age ranged from 38-57* The mean duration of apneas
in this group was 30 seconds. All apneas occurred in
light or REM sleep. Orr et al. (1979) studied 4 hyper-
somnolent males diagnosed with SAS. A control group of 4
nonhypersomnolent subjects was matched on number of ob
structions per minute. Mean age was 57 for asymptomatic
and 42 for symptomatic subjects. Asymptomatic subjects


10
suspicion o insomnia or nighttime breathing and muscular
events. Eleven males (x = 72.5) and 13 females (x = 68.5)
were studied. All of these subjects, save one male and
one female, suffered at least one episode of sleep
apnea. For those subjects with at least one episode of
sleep apnea, males had a mean of 53 episodes of apnea,
while females had a mean of 32 episodes. In contrast to
the Kales et al. study, 6 males had a high level of sleep
apnea (AI>5), while 3 females met this criterion. A rela
tionship between age and apnea was not reported in this
sample, possibly because of restricted range. These two
studies of sleep disturbed subjects provide conflicting
data. That is, the insomniac alone sample did not demon
strate an apnea incidence which was substantially
different from a noncomplaining sample (10.5%) while the
sample recruited for sleep induction, respiratory, and
muscular differences showed very discrepant apnea inci
dences (nearly 100£). One potential explanation involves
age, as the Ancoli-Israel sample is substantially older
than Kales', although the aging samples have not reported
this level of apnea before. An alternative explanation is
that sampling insomniacs does not generate a high number
of apneas, whereas aging sleep and respiratory disturbed
subjects show a very high incidence of apnea. It would
seem then that the relevant "complaint" would involve
those reported by Ancoli-Israel et al. which are in
addition to those noted by Kales et al. These are (as


30
experimental protocols by Weitzman et al. (1980), Fujita
et al. (1981), Zwillich et al. (1982), Garay et al. (1981)
and Sullivan and Issa (1980). At the same time, subjects
wno are asymptomatic for EDS but with the necessary number
of apneas to be diagnosed as SAS are described by Orr et
al. (1979) and Smirne et al. (1980) as well as others.
However, it should be noted that the latter studies relied
on global reports of somnolence, rather than quantified
data. Recalling the criticism of Dement et al. (1978) of
this technique as vulnerable to subjective bias, these
asymptomatic patients may have unreported or undetected
somnolence.
Excessive Daytime Somnolence appears to be frequently
concomittant with Sleep Apnea Syndrome. Thus its possible
occurrence in subclinical populations seems worthy of
investigation.
Oxygen Desaturation and Cognitive Deficits
One deficit which appears in SAS patients is oxygen
desaturation which accompanies apneas. V.'hen apnea begins,
arterial oxygen saturation begins to fall, and may con
tinue falling throughout the event. An early technique
utilized in studying oxygen saturation levels involved
periodic blood samples drawn from arterial sources.
Eirchfield et al. (1958) studied 11 normal males with a
mean age of 23* Daytime oxygen saturation in these sub
jects averaged 96.5$. These values fell during sleep to
95.3#. Later, Orr et al. (1979) studied 8 males who had


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
SLEEP APNEA ACTIVITY AND ITS CONCOMITANTS
IN A SUBCLINICAL POPULATION
By
DAVID THOMAS REED BERRY
August 1985
Chairman: Wilse B. Webb
Major Department: Clinical Psychology
Sleep Apnea Syndrome (SAS) is a nocturnal respiratory
disorder with serious consequences. Patients with SAS
experience apneas (pauses in breathing of 10 or more
seconds) and hypopneas (declines in the amplitude of
breathing accompanied by oxygen desaturation) while
asleep. Recently, it has become apparent that apneas and
hypopneas may occur in otherwise apparently normal
subjects. A sample selected to display a wide range of
sleep disordered breathing (heavy snoring males in good
health and over 30 years of age) received detailed
measurement of their nocturnal respiration, sleep,
cognitive skills, and health in an attempt to clarify the
impact of apneas and hypopneas in a subclinical population.
Forty-six heavy snoring males in good health with a
mean age of 50 years and a mean weight of 190 pounds
comprised the sample. During an experimental evening they
vii


11
nearly as can be judged) questions surrounding
respiration.
The final category of incidence studies involves
those which do not consider sleep complaints in their
sampling. As in the other studies, these vary in their
generalizabil.ity. Carskadon and Dement (1981b) recruited
40 aging subjects, including 18 males (x = 72.7) and 22
females (x = 74). Their exclusion criteria are somewhat
problematic, involving excluding subjects who "spon
taneously complain of sleep problems." Thus, it is
unclear what this is a representative sample of, specifi
cally whether the distribution of those not spontaneously
complaining of sleep disturbances may be different from
those who actually do not suffer from sleep distur
bances. These considerations aside, Carskadon and Dement
report that 16 males (88%) and 16 females (72.%) demon
strated at least one sleep apnea episode, while 8 males
(44$) and 7 females (31.8%) have high levels of sleep
apnea (AI>5). Males with apnea showed a mean number of
51, while females showed a mean number of 34. Sex
differences appear to have fallen out at this age, while
the restricted range of this extreme age sample does not
lend itself to a correlational analysis of age versus
apnea. In another study of limited generalizability,
Kreiss et al. (1982) randomly selected inpatients on a
Veterans Administration medical ward. These patients, all
presumably male, numbered 26. Further data reported were


received testing and filled out various questionnaires
before sleeping overnight with their sleep and breathing
continuously recorded. Sixty-two per cent of these subjects
experienced at least one event, while 13% had more than 5
events per hour of sleep. Most events occurred in light
slow wave or REM sleep and were linked to obesity. Subjects
with high levels of apnea/hypopnea (more than 5 per hour)
experienced significantly exacerbated oxygen desaturation
relative to the remaining subjects. Oxygen desaturation was
linked to higher blood pressure readings. A typical "first
night*' effect on the sleep of the subjects in the lab was
noted, resulting in a lighter sleep than usual and obscuring
possible relationships between sleep and respiratory events,
which were not observed. Deteriorating overnight
respiratory indices were associated with increased
sleepiness and napping. High levels of apnea/hypopnea were
associated with declining scores on tests measuring non
verbal intelligence, verbal and non-verbal memory,
expressive verbal fluency, and cognitive flexibility. It is
speculated that hypoxia induced changes in cerebral
cholinergic synthetic pathways underlie these changes.
It is concluded that a significant subgroup of heavy
snoring males who experience multiple apnea/hypopneas are at
increased risk of significant oxygen desaturation, daytime
sleepiness, and cognitive changes. This group may fall on a
continuum with sleep apnea syndrome patients.
viii


4b
is indicated. An adequate description of the sleep/wske
cycle would seem to include sleep questionnaire, sleep
log, overnight EEG, multiple sleep latency test, as well
as several Stanford Sleepiness Scale ratings.
Statement of the Problem
Sleep Apnea Syndrome is a clinical entity with
seemingly serious consequences. Although the nocturnal
apneas found in this illness are thought to be the cause
of the associated deficits, the identification of
seemingly asymptomatic subjects with a number of apneas
comparable to those found in SAS throws doubt onto this
formulation. Despite this evidence, the recent discovery
that apneas occur in significant numbers in the general
population has led to speculation that this subclinical
level of apnea may be the cause of various pathological
processes. At present, there is little evidence to
support or refute this notion.
The present report evaluates this issue through a
detailed study of a group of subjects thought to be at
risk for sleep apneas. Aging, heavy snoring males were
recruited, and a comprehensive analysis of their respira
tory, health, neuropsychological, and sleep/wake cycle
status was carried out. Thus potential relationships
between respiratory disturbances and variables known to be
disrupted in sleep apnea syndrome patients could be
evaluated. This approach also allowed exploration of the


29
apneas actually suffer EDS, but that it is masked by two
factors. Dement et al. noted that SAS patients may deny
somnolence while literally falling asleep before the
clinician. They speculate that this results from a re
sponse bias against acknowledging illness or a change in
subjective frame of reference about just what alertness
is. Thus EDS may be denied by patients for reasons which
may be outside the clinician's control or knowledge.
Alternatively, Block et al. (1979) propose that asympto
matic SAS patients are suffering a subclinical level of
fallout from their frequent apneas. As years pass, in
creasing weight or cumulative effects of desaturation
eventually lead to a full blown hypersomnolenceSleep
Apnea Syndrome. A final possibility is that nonhyper-
somnolent subjects with frequent apneas exhibit a com
pletely benign process totally unrelated to that found in
symptomatic SAS patients. It would seem that delineation
of the level of EDS in subjects with subclinical levels of
apnea may help clarify this issue. This notion will be
elaborated below.
The somnolence deficit which is noted in full blown
SAS patients is commonly reported. Dement et al. (1978)
describe impairments in continuous performance of
virtually any activity. Sleep Apnea Syndrome patients
reportedly fall asleep at outdoor stadiums, in front of
classes, and while treating patients. Excessive Daytime
Somnolence is described in every SAS patient reported in


I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
Wilse B. V/ebb, Chairman
Graduate Research Professor of
Psychology
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
Assistant Professor of Clinical
Psychology
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
T. 37Block
Processor of Medicine
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
F. Dudley McGlyriri
Professor of Clinical
Psychology


49
professors. All subjects completed and signed an informed
consent agreement.
Apparatus
During the evening testing, all subjects received a
blood pressure check with a standard hospital cuff. The
measurement was made from a seated position. Overnight
recordings were made in a quiet, darkened room with the
subject sleeping on a standard hospital bed. All noc
turnal physiological recordings were routed through a
Grass Polygraph (model 7D). Electrodes for electro-
encephalographic recordings were mounted in three pairs
using sites F2/F8, PZ/T6, RE/LE from the international
10/20 system of placement. These electrodes were affixed
to the scalp with collodion soaked gauze pads dried with
an air hose. Respiratory measures included measurement of
chest and abdomen wall movement derived from impedence
measures monitored from surface electrodes (3F Ag/AgCl
#2246) placed on the lower chest and just above the
navel. The signal from these electrodes was passed
through an impedence convertor and onto the polygraph.
Oral and nasal airflow was recorded by thumistors (TCT-1R
Transducer-Grass) clipped to the respective orifices.
These signals were also routed through the polygraph.
Electrocardiograms were monitored from modified bipolar
chest leads (MCL2). Finally, blood oxygen saturation was
continuously monitored from a probe attached to the ear
lobe, and analyzed by a Biox ear oximeter (flodel IIA BTA


61
figure [F[2,43)=6.5> p<.003], and Verbal fluency,
[F[2,43)=3-6 p<.04]. Followup comparisons (Scheffe)
showed the group with high apnea/hypopnea to be
significantly impaired relative to those with no sleep
apnea/hypopnea on WAIS PIQ, delayed recall of logical
stories and Rey complex figure as well as verbal
fluency- Additionally, the high apnea/hypopnea group was
impaired relative to those with subclinical apnea/hypopnea
on aelayea recall of the Hey complex figure. These
findings are bolstered by the lack of significant between
group differences on age, weight, or education (see Table
3-15) all of which are potential alternative explanations
of the between group differences.
As a check, an analysis was run to evaluate the
relative contribution of nocturnal respiratory variables
and sleep variables reflecting fragmentation in predicting
cognitive/neuropsychological scores. A multivariate
regression of these scores was carried out. Demographic
and respiratory scores included age, apnea index, apnea +
hypopnea index, and number of desaturations >4%. A subset
of the earlier noted sleep variables was chosen to reflect
measures thought to be most responsive to sleep
fragmentation. These included sleep efficiency, number of
stage 0 periods, time % of stage 0, time % of stage 2, ano
number of stage changes. These respiratory and sleep
parameters were used to predict the 5 cognitive scores
which were earlier shown to be predicted by nocturnal


7e
scores were entered into a partial correlation matrix,
controlling for age, with the cognitive scores. Table
5-14 presents the resulting correlations significant at
p<.05 or less. The table incicates that every cognitive
test but digit span and Hooper correlated with at least
one overnight respiratory index. Of additional impact is
the observation that in every case of a significant
correlation, deteriorating respiratory indices are
associated with worsening neuropsychological scores.
Respiratory indices with particularly heavy loadings on
neuropsychological scores included those quantifying
apneas, (Apnea index, Apnea + hypopnea index) as well as
those more directly representing deepening hypoxia (Mean
low saturation and Humber of desaturations ^>4^).
Subjects were grouped by level of nocturnal
respiratory disorder (no apneas/hypopneas, subclinical
level of apnea/hypopnea, and clinically significant level
of apnea/hypopnea). Means and standard deviations for the
neuropsychological scores from these groups are presented
in Table 5-15. The table suggests a general deterioration
of cognitive scores as level of nocturnal respiratory
distress increases. Oneway AHOVA procedures were applied
to the neuropsychological scores from each test. Signifi
cant between group differences emerged for V/AIS PIQ
LF(2,45)=6.9 p<.002] delayed recall of logical stores
[F[2,45)=4.0, p<.05] delayed recall of visual reproduc
tions [F(2,45)=5.4, p<.04], delayed recall of Rey complex


sleepiness, and muscular events (nocturnal myoclonus).
Further, these subjects had a mean age of almost 70 years,
indicating significant sampling and age differences from
the present study. Ninety-three per cent of these
subjects had at least one event, while 25% had high
numbers of apnea/hypopnea. These rates are much higher
than those found in the present sample. This is probably
attributable to the sampling differences noted above.
However, a recent report by Miles and Simmons (1984)
described apneas in a series of 1S0 patients referred for
complaints of heavy snoring. In this sample, 73% had at
least one event, while 9% had severe levels of apnea.
These data are quite comparable to the prevalences
observed in the present sample, 62% for at least one event
and 13% for high levels of apnea/hypopnea, and lend some
confidence to the present finding.
Thus the overall frequency of sleep disordered
breathing in middle-aged subjects with heavy snoring
appears to be between 62 and 73% for at least one event,
and between 9 and 13% for more severe levels of apnea.
The relatively close concordance of the two reports on
heavy snoring subjects is reassuring.
Factors which were found to be related to the
occurrence of sleep apneas in the previous review included
male sex, complaints of snoring, increasing age, and
obesity. The first two factors could not be examined in
the present sample because of their use as selection


28
disease or hypertension. Thus a group of SAS patients
without evidence of heart disease exists.
Taken together, these data indicate that cardiac
complications frequently, but not inevitably, accompany a
sleep apnea syndrome. That a large number of apneas may
be present without a significant increase in cardio
pulmonary complications is demonstrated by the Orr et al.
and Ancoli-Israel et al. studies. As the relationship
between these complications and a full blown SAS is
unclear, the possibility of cardiopulmonary sequelae from
a subclinical level of apneas is even more uncertain.
However, an assessment of cardiopulmonary status is
indicated by the frequent association of cardiopulmonary
complications with sleep apnea syndrome.
Arousal Deficits (Hypersomnolence)
Excessive Daytime Sleepiness (EDS), or hypersomno
lence, has been described as the single most common result
of SAS (Dement et al., 1978). Indeed, patients who meet
the criterion of number of apnea episodes and exhibit EDS
are described as symptomatic, while those with number of
apneas alone are labeled asymptomatic (Orr et al.,
1979). As an aside, this distinction illustrates again a
central issue of SAS. If patients exhibit the number of
apneas qualifying them for SAS, but no other pathology,
are they suffering from a pathological process? Dement et
al. (1978) believe that they are. They suggest that most,
if not all, patients with a clinically elevated number of


116
A-CCOC^>aM^CO I CTiCT\^L^COCOt>-
OOOOOOOt- ioooooooo
v~ Al Al AJ Al Al AJ I Al fA A fA AJ Al
CNJ t v~t~ CNJ IrrrrrtMrtM
vO m ltn "3- co ^ v- i r-i>-c\j>-m^£jcom
r~ rf'i rr (\| r- t
OOr-OO--r-v IOOt-OOOt-O
Os Al fA C\] f A AJ Al AJ lAJAJs-rAAJIAAJfA
O
I t-N- i
I OO l
i c\j i
I C\i AJ I
I lAvO I
I OO I
I Al AJ I
vCNCU^O^-tMi'Av N CO CA O Al f A ^ LA M)
AJ Al Al Aj f A f A ¡A fA NA fA (A fA fA rA-v^-^^--^'^'-^-


126
Feinberg, I., Koresko, R.L., & Heller, N. (1967). EEG
sleep patterns as a function of normal and pathological
aging in man. Journal of Psychiatric Research, 107-
144.
Fisher, J., de la Pena, A,, Mayfield, D., & Flickinger, R,
(1978). Psychobiologic determinants of subgroups of
sleep apnea. Sleep Research, _7, 221.
Francheschi, M., Zamponi, P.f Crippa, D., & Smirna, S.
(1982). EDS: A 1 year study in an unselected inpatient
population. Sleep, (3), 239-247.
Freides, E., & Allweise, C. (1978). Transient hypoxic
amnesia. Behavioral Biology, 22, 178-189.
Fujita, S., Zorick, F., Koshoreck, G., Wittig, R., Conway,
W. & Roth, T. (1981). Treatment of upper airway sleep
apnea with UPP: Recent experience. Sleep Research, 10,
197.
Garay, S., Rapaport, D., Sorkin, E., Epstein, H., Feinberg,
I.f & Goldring, R. (1981). Regulation of ventilation
in the obstructive sleep apnea syndrome. American
Review of Respiratory Disease, 124, 451-457.
Gastaut, H., Tassinari, C., & Duron, B. (1965). Etude
polygraphique des manifestations episoiques diurnes et
nocturnes du syndome de Pickwick. Revue Neurolgie, 112,
573-579.
Gellhorn, E. (1951). Sensitivity of the auditory
projection area to anoxia. American Journal of
Physiology, 164. 748-751.
Gibson, G., & Blass, J. (1976). Impaired synthesis of
acetylcholine accompanying mild hypoxia and
hypoglycemia. Journal of Neurochemistry, 27. 37-46.
Gibson, G., & Duffy, J. (1981). Impaired synthesis of
acetylcholine by mild hypoxia or nitrous oxide. JournaI
of Neurochemistry, 36(1), 28-53.
Gibson, G., Pulsinelli, W., Blass, J., & Duffy, T.
(1981). Brain dysfunction in mild to moderate
hypoxia. American Journal of Medicine, 70, 1247-1254.
Grant, I., Heaton, R., McSweeny, J., Adams, K., & Timms,
R. (1982). Neuropsychological findings in hypoxemic
chronic obstructive pulmonary disease. Arcnives ci
Internal Medicine, 142, 1470-1479.


e
the study if they reported any "serious" sleep com
plaint. Two of the subjects reported consuming hypnotics
for sleep induction, indicating that the criterion used in
this study was less stringent than that of Block et si.
(1979) report. Nine of the subjects (45%) suffered at
least one episode o sleep apnea with a mean number of
episodes of 2.5. All of the subjects with episodes of
sleep apnea were older than the mean age of the sample,
leading Webb to conclude that apnea is related to age.
Bixler et al. (1982) constructed an age stratiiied patient
representative sample of 100 subjects (M =41, F = 59).
In perhaps the most rigorous screening for sleep com
plaints in this group of studies, Bixler et al. eliminated
any subject with a sleep complaint or chronic medication
usage. For the entire sample, 6 males (14.6%) and 5
females (10.2%) had at least one episode of sleep apnea.
For subjects under 30, 3*3%; for those between 30 and 50,
12.8%; and those 50-74, 19.4% suffered episodes of sleep
apnea. For the subsample over 60, 25% had sleep apnea
activity with 8.5 episodes for males and 10.5 episodes for
females. In this sample, there was a nonsignificant trend
toward more apnea activity in males and a significant
positive correlation between age and sleep apnea.
Block et al. (1980) report data on 20 post menapausal
women with a mean age of 59. Eight women (40%) had epi
sodes of sleep apnea, with a mean number of episodes of


CHAPTER FOUR
DISCUSSION
Demographics and Incidence o
Nocturnal Respiratory Disorder
The present sample of 46 snoring males was selected
as a group likely to display a wide range of sleep
disordered breathing. These subjects, with a mean age of
50 years and a mean weight of 190 pounds, reflect a mostly
middle-aged to older male population which is moderately
obese. The demographic characteristics of this sample are
similar to those of the "typical" sleep apnea syndrome
patient (Guilleminault et al., 1978), although the
requirement of general good health insured that subjects
with a classic SAS were excluded. This sample may be
considered representative of subjects predisposed to sleep
disordered breathing, although at no more than a
subclinical level. Therefore it seems valid to consider
various risk factors as they relate to nocturnal
respiration.
In the present sample, 62% experienced at least one
apnea or hypopnea, while 13% suffered higher levels of
apnea/hypopnea. Only two other reports have described
samples selected for snoring or respiratory complaints.
Ancoli-Israel et al. (1981) examined 24 subjects with
complaints of respiratory difficulties, daytime
86


12
sketchy, but 7 (27%) met the sleep apnea syndrome
criterion of >30 episodes of sleep apnea. The fact that
these subjects were patients on a medical ward obviously
restricts generalizability of these findings. Finally, in
the only truly random sampling reported, Ancoli-Israel et
al. (1982) present initial data on randomly selected sub
jects. Forty subjects (sex not reported) with a mean age
of 71.5 were studied. Thirteen (32.5$) met the sleep
apnea syndrome criterion of 30 or more apneas. The sleep
apnea subgroup was older, 74.2, than the full sample.
The data collected from relatively random sampling
are primarily based on aging populations (>60). The data
from these studies suggest that between 32.5$ (Ancoli-
Israel et al., 1981) and 37$ (Carskadon and Dement, 1981b)
of aging subjects suffer high levels of sleep apnea
(AI>5). An even larger percentage, perhaps greater than
75$ suffer at least some episodes of sleep apnea. A
major deficiency in these randomly sampled sxudies is a
lack of data on younger subjects.
Thus data collected under different sampling proce
dures yield variable incidence rates of sleep apnea and
AI>S. Subjects without sleep or respiratory complaints
have a very low incidence (perhaps 0) of AI>5, between 14$
and 45$ of males surveyed had at least some apnea, while
females ranged between 10$ and 40$ experiencing some
apnea. Sampling subjects with sleep or respiratory com
plaints suggested that subjects suffering respiratory


Tabic B-4
Seven day sleep log means from 46 snoring males
ubject
f?
Daytime
Sleepiness
Dumber
of
Daps
Sleep
Latency
Number
of
Wakenings
Minutes
Waking
Minutes
Napping
Total
Bedtime
1
2
0
3
1
2
0
7
2
2
1
5
0
1
0
9
3
2
0
15
4
2
0
6
4
2
0
7
2
3
0
8
5
3
0
12
3
2
0
6
6
3
0
16
2
1
0
6
7






8
3
1
9
2
2
0
7
9
2
1
21
2
4
1
7
10
3
0
19
2
3
0
9
11
3
0
6
2
2
0
8
12
-
-
-
-
-
-
_
13
3
0
9
1
1
0
7
14
2
0
17
1
1
0
9
15

-
-
-
-
-
-
16
3
0
12
2
3
0
9
17
-
-

-

-
-
18
2
0
35
2
3
0
n
/
19
2
1
30
2
5
0
7
20
2
0
7
2
2
0
7
21
3
0
14
1
2
0
9
22
2
0
14
2
4
0
7
23
3
0
9
0
1
0
8
24

-
-
*
-
-
-
25
1
1
5
1
o
2
5


82
respiratory scores: WAIS PIQ, Wechsler Memory Quotient,
delayed recall of visual reproductions and Hey figure, and
Wisconsin card sort. Table 3-16 presents the data from
this analysis. The Hotellings T2 was significant,
(p<.001) v;ith delayed recall of visual reproductions and
Rey figure significantly predicted by the predictor
scores. Inspection of the individual regressions for
these two measures discloses that age, apnea index, and
number of desaturations >A% all formed significant
components of the prediction of delayed recall of visual
reproductions, while sleep efficiency was the sole
significant predictor of the Hey figure. These data seem
supportive of an independent contribution of both
nocturnal respiratory variables and sleep variables at a
limited level, to the prediction of cognitive/neuro
psychological scores.
In summary, various statistical procedures indicate
relationships between nocturnal respiratory measures and
non-verbal intelligence, verbal and non-verbal memory,
expressive verbal fluency, and ability to shift cognitive
set while problem solving. Division of subjects by level
of nocturnal respiratory distress indicated that subjects
with high amounts of apnea/hypopnea were impaired relative
to controls on measures of non-verbal intelligence, verbal
and non-verbal memory, and expressive verbal fluency, as
well as being impaired relative to subjects with
subclinical levels of apnea/hypopnea on measures of


I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy,
Assistant Professor of Clinical
Psychology
This dissertation was submitted to the Graduate Faculty of
the College of Health Related Professions and to the
Graduate School and was accepted as partial fulfillment of
the requirements for the degree of Doctor of Philosophy.
August 1985
(?
Dean, College of Health Related
Professions
Dean, Graduate School


Table E-7
Sleep variables from 46 snoring
Number
ub j.
tt
TOT
bleep
La tency
bleep
Liliciency
bleep
Tine
Time '/
Sta^e 0
Ota£e 0
Periods
1_
1
}Ob
0
0.0
103
30.4
3
1.7
2
373
3
O.b
210
12.9
0
5.0
5
A
401
0
0.0
304
4.2
10
5.9
j
374
b
0.7
20 3
22. b
12.0
3.0
0
487
2
o.o
450
3.6
0
5.9
7
431
1d
o.o
371
10.1
14
5.3
0
30b
14
0.7
203
23.1
9
3.2
4bri
54
O.b
3b5
7.5
2
1.9
1U
33b
94
0.3
110
37.0
3
1.9
11
40/
30
o.b
341
0-3
7
2.1
12
402
154
0.5
199
9.9
6
1.4
13
300
50
0.7
2b7
23.5
7
3-7
14
33b
135
0.6
203
0.1
9
0.5
1b
302
34
O.b
335
3.7
7
0.3
1b
33b
30
0.6
219
26.0
10
2.0
17
374
25
O.b
30b
1.3
2
0.9
1d
343
11
0.9
320
3.6
7
2.7
1V
320
24
0.2
73
33.0
2
1.9
20
44b
67
0.7
321
13.7
7
4 .0
21
323
17
0.7
23b
15.1
9
1.1
22
348
9
O.b
200
17.4
7
3.9
2 >
2b 4
t>
0.9
2bb
3.3
4
1 .b
24
37 b
19
0.0
290
19.2
12
2.0
2b
32o
0
o.o
203
7.5
12
2.3
2b
35b
23
0.6
221
33.b
33
2.1
27
3b4
2b
O.b
232
20.5
1b
6.9
2d
327
19
0.0
2b5
4.3
4
2.9
20
344
b
0.0
27 3
19-2
11
2.7
50
340
20
o.o
295
7.0
6
3.1
31
-
-
-
-

-

ales
% Slow Number
lime
Jo OLcl/IG
Wave
UEft
Mean
MEM
HEM
2
3.
4
bleep
Chan/'es
La tency
HEM
6ycii
0
60.0
o
0
0
14
0
0
0
0
71.0
9.9
0.4
11.9
51
0
0
0
17.9
40.3
0.7
16.5
26.3
63
02
1b
90.7
10.9
46.3
4.9
11.5
21.2
32
75
20
233.0
16.2
55.5
2.9
15.0
19.5
47
72
13
bb. wj
10.2
69-9
2.1
1.7
4.3
64
163
1 3
61 .0
12.6
63.0
1.0
0.5
3.1
38
190
23.5
bb.o
25.4
47.3
6.1
11.0
19.3
36
3b
2b. 5
1 19.7
1.5
26.0
1.9
9.9
2b. 1
1b
221
4
0
21 .0
65.4
2.7
1.6
4.7
42
bb
27
145
0
07.0
0
0
0
16
0
0
0
2.9
60.5
2.6
6.9
12.3
30
247
5
49
17.0
65.5
4.5
4.5
9.8
32
02
19
167.0
17.0
59.2
4.3
15.5
20.6
23
212
29.3
10b. 0
2.6
b2.5
6.0
0.0
8.2
44
292
0
0
17.1
bb. 3
4.5
9.7
14.4
21
97
2b.5
120
10.1
53.b
2.1
19.9
22.8
30
71
15
75.7
0
41.7
2.5
0
5.4
11
0
0
0
1b.7
57.3
2.9
7.5
12.1
40
153
31
122
14.0
32.5
7.5
9.7
20.1
43
131
19.5
114
7.9
55.9
2.3
14.3
20.4
32
07
15.3
1/8
9.0
07.2
6.9
13.7
21.3
34
114
15.5
93
0.9
61.3
2.2
5.0
9.6
39
291
52
0
21.1
60.4
2.6
7.1
10.5
44
43
10.0
47.0
0
64.3
0
0
0
71
0
0
0
9.b
63
0
0
0
50
262
2b
0
13
6b. 4
4.7
0.7
13.9
3o
01
9
64
13.9
62.1
6
0
0
32
153
1b
3b. 3
15
63.4
4.7
6
3.1
30
65
1b
111
(V)


65
Table 3-5. Means and standard deviations for health related
variables in 46 snoring males grouped by level
of apnea/hypopnea.
No Apnea/
Low Apnea/
High Apnea/
Hypopnea
Hypopnea
Hypopnea
Systolic BPa
129-1(14.2)
128.7(12.6)
127.0(7.5)
Dia¡
stolic BP
85.2(13.4)
83.9(11.0)
81.6(7.6)
CMI*
3 Overall Score
14.7 (7.4)
16.1(10.5)
13.8(7.4)
CMI
Respiratory Score
1.4 (1.2)
1.1 (1.8)
2.0(1.4)
CMI
Cardiac Score
1.6 (1.3)
1.2 (1.2)
1.0(1.2)
CMI
Neurological Score
1.2 (0.9)
1.6 (1.5)
0.8(0.6)
"f BP = blood pressure.
b CMI = Cornell Medical Index.
Nocturnal Respiratory and Sleep/V/ake Variables
EEG Data
Electroencephographic variables were recorded
overnight for all subjects. Records were scored by a
trained technician who utilized the system of Agnev and
Webb (1972) to determine sleep stages. In three records,
EEG tracings were judged inadequate for differential sleep
staging. However, a judgement of sleep vs. wake was made
for these subjects in order to form a basis for the
respiratory indices. These three EEG records were
excluded from the following analyses, leaving 43 records
as a base for analysis.
Fifteen variables (time in bed, pure sleep time,
sleep efficiency index, number of stage 0 periods, time %
stage 0, sleep latency, time % stages 1, 2, 3, 4, and REM,
latency 1st REM period, mean REM period length, mean REM


131
Wechsler, D. (1955)* Wechsler Adult Intelligence Scale:
Manual. New York: Psychological Corporation.
Weitzman, E. (1979). The syndrome o hypersomnia and sleep
induced apnea. Chest, 75, 414-415*
Weitzman, E., Kahn, E., & Poliak, C. (1980). Quantitative
analysis o sleep and sleep states before and after
tracheostomy in patients with the hypersomnia sleep
apnea syndrome. Sleep, (3/4), 407-423*
Weitzman, E., Poliak, C., Borowieki, B., Burak, B.,
Shprintzen, R., & Rakoff, S. (1978). The hypersomnia
sleep apnea syndrome: Site and mechanism of upper
airway obstruction. In C. Guilleminault & W. Dement
(eds), Sleep Apnea Syndromes. New York: Alan R. Liss
Inc.
West, J. (1984). Human pnysiology at extreme altitudes.
Science, 223, 784-786.
Zorrick, T., Roehrs, T., Koshorek, G.,
Hartse, K., Wittig, R., & Roth, T.
sleepiness in various disorders of
174.
Sicklesteei, J.,
(1982). Patterns of
EDS. Sleep, , 165-
Zwillich, C., Devlin, T., White, D., Douglass, N., Weil, J.,
& Martin, R. (1982). Bradycardia during sleep apnea.
Journal of Clinical Investigations, 69, 1286.


95
disturbed nocturnal breathing. In the first case,
increasing respiratory disturbance causes sleep
deprivation, manifested in increased subjective sleepiness
(SSS ratings) and compensatory sleeping (increased
napping). The other side of this coin is a sleep pressure
v/hicn causes disturbed subjects to fall asleep quickly
(lowered sleep latency), and subjectively fail to report
nightly wakenings, while subjects with intact respiration
and hence less sleep pressure experience a longer sleep
latency and awareness of nightly awakenings. Admittedly
this is speculative, and two immediate problems with this
formulation are evident. First, the lack of correlation
between LEG sleep and respiration of the experimental
night is problematic, as well as the lack of relationships
between respiration and the evening "naps." However, it
has already been speculated that an exaggerated first
night effect obscured correlations between sleep and
respiration, and it is new speculated that the two naps,
at 7 and S pm, missed the afternoon trough of sleepiness
which might have better reflected sleepiness.
The interpretation of increasea sleepiness from
nocturnal respiratory disturbance is useful in explaining
the pattern of findings observed in this study, although
some problems with this formulation were noted.
To summarize the main aspects of the discussion of
sleep/vake variables, it is noted that these snoring
subjects achieved a night of sleep comparable to those


21
subjects to be uniformly obese* A review of the
literature is in order to clarify this point. First
normal, then obese, then SAS groups will be reviewed.
Bixler et al. (1982) found that their subgroup of
subjects with sleep apnea activity weighed significantly
more than the remainder of the sample. Kales et al.
(1982) found that within their sample of 200 insomniacs
and 100 normals, a significant positive correlation
emerged between a weight:height ratio and apnea acti
vity. Block et al. (1979) found a positive correlation
between apneas, oxygen desaturation, and weight within
their normal males, while females did not demonstrate this
relationship. Using an older female sample, Block et al.
(1980) found a positive correlation between weight and
oxygen desaturation, but not apnea. Thus in normal sub
jects, there appears to be a positive relationship between
weight and apnea activity, although this may be restricted
to males.
A research approach which bears on this question
involves studying obese subjects alone. Two studies on
successive referrals for gastric bypass surgery were
noted* Sicklesteel et al. (1981) found that of 19 suc
cessive referrals, all 14 males exhibited sleep apneas,
while only 5 of the females showed evidence of oxygen
desaturation alone. Harmon et al. (1981) studied gastric
bypass patients and found 5 of 6 males to suffer apnea, as


25
suffer sleep apneas. A significant proportion of snorers
might be expected to suffer apneas, while few nonsnorers
might have apnea. The last variable, weight, while having
a positive correlation with apnea activity in the general
population, is of uncertain pertinence to SAS. Addi
tionally, changes in weight may occur without changes in
apnea activity. Therefore, this variable has a less than
explicit connection to apnea activity and violates the
second principle adopted to consider these variables.
This review of available data suggests that sampling
aging, snoring, males would generate a high proportion of
sleep apnea activity. External validity for this idea is
drawn from its similarity to the "typical'' SAS patient
described earlier. Extant data suggest that it may also
prove useful to utilize weight measures as a covariate, in
an attempt to control its apparent influence on apnea
activity.
Deficits Associated With Sleep Apnea Syndrome
Sleep Apnea Syndrome is thought to result in several
deficits. These include cardiopulmonary complications,
hypersomnolence, hypoxia and intellectual changes. Some
researchers have suggested that these deficits may be
present, in an attenuated form, in subclinical apnea
(Ancoli-Israel et al., 1981; Carskadon and Dement,
1981b). A discussion of the evidence for the presence of
these deficits will be presented, followed by proposals
for appropriate measures of these variables.


22
opposed to none of the females. This evidence provides
further evidence for a role of increasing weight in apnea.
The data on symptomatic SAS subjects are more com
plex* Several studies (Sullivan and Issa, 1980; Iveitzman
et al., 198C; Garay et al., 1981; Zwillich et al., 1982)
report that all their SAS subjects were obese, although
definitions of obesity either vary widely or remain un
specified. In contrast, Guilleminault et al. (1978)
reported that only 60# of their SAS sample was greater
than 15/S above ideal weight. Coverdale et al. (1980)
indicate that only two of their subjects proved to be
obese, according to their definition (>125# of ideal
weight). These reports suggest that most, although not
all, SAS subjects are obese.
Several investigators have sought to compare weight
in symptomatic (somnolent) and asymptomatic subjects with
enough sleep apnea episodes to be diagnosed as SAS. Orr
et al. (1979) compared 4 symptomatic SAS subjects with
matched asymptomatic subjects. In this sample, symptoma
tic SAS subjects were heavier than asymptomatic sub
jects. Examination of the data of Berry and Block (1963)
indicates that their symptomatic subjects were heavier
than their asymptomatic subjects (118 kg vs. 91 kg).
Standing in contrast to these data is a study reported by
Fisher et al. (1978), which included 19 subjects referred
for SAS evaluation. Based on number of apneas per night,
Fisher et al. divided their subjects into 3 groups. Group


91
In light of these caveats, it is of interest that two
measures of desaturation, mean low saturation, and number
of desaturations >4% were associated with increasing blood
pressure readings. While a report by Guilleminault et al.
(1978) noted an increasing prevalence of hypertension in
subjects diagnosed with sleep apnea syndrome, a
relationship between desaturation and hypertension has not
been previously reported. This finding was observed in
spite of the fact That subjects prescribed hypertensives
were not medication free for the pressure reading.
Additionally, a check of self reports of diagnosed
hypertension showed six subjects in the no event group as
having diagnosed hypertension, as opposed to six in the
low apnea/hypopnea group, and one in the high
apnea/hypopnea group, data which did not link
apnea/hypopnea, per se, with hypertension in these
subjects. The isolated finding of a relationship between
mean high saturation and overall number of symptoms
endorsed on the CMI is more puzzling and does not follow
any logical analysis.
In summary, the present study's assessment of health
variables must be regarded as somewhat limited. In spite
of this, an association between two measures of desatura
tion and a measure of blood presure was noted. It is
speculated that blood presure may be more responsive to
nocturnal oxygen saturation than to the apneic events
alone.


66
cycle length, % slow wave sleep), reflecting multiple
aspects of the sleep of these subjects, were derived from
the sleep stage scoring. Table 3-6 presents the means ana
standard deviations for these subjects. Although subjects
spent over six hours in bed, the table indicates that not
quite 5 hours were spent asleep. The remaining variables
suggest that an increase in sleep latency, and time awake,
is responsible for the diminished sleep time, while
lighter sleep is also increased.
Table 3-6. Means and standard deviations of selected sleep
variables from 43 snoring males.
Time in Bed
Sleep Latency
Pure Sleep Time
Sleep Efficiency Index
Number Stage 0 Periods
Time % Stage 0
Time % Stage 1
Time % Stage 2
Time % Stage 3
Time % Stage 4
Time £ Stage REM
Latency 1st REM Period
Mean REM Period Length
REM Cycle Length
% Slow Wave Sleep
Standard
Mean Deviation
377.3
45.1
29.2
32.9
288.7
75.8
.76
.16
8.6
5.6
14.3
12.7
3.3
2.8
60.8
10.3
3*4
2.4
7.9
6.4
11.5
7.2
124.7
92.2
14.9
9.4
74.9
57.7
12.9
8.3
Note: see text for explanation of N.
Partial correlations, controlling for age, were
calculated between these sleep and the respiratory
variables. Out of a total of 105 correlations, only one


114
C\JC\]C\JC\C\jT-C\U\irO|CSjT-C\J C\]C\JtC\]C\JC\Jt-C\IOOJ
T- fO, k\ io NA fi"\ n rr\ 1^1 CM ro 00 r^i O to
C\| C\¡ rr\ CM C\l C\J OJ C\J CM rO A C\J i^r(M(MtOr rr Or
r- C\j CM CM O OJ CM rr\ jo* x- M ¡*^ s- CM CM lA t- CM CM O r-
^0CMC\JCM0x-C\JCMT-T-C\^-C\J^-T-r-0 O CM CM ^ r- h*, O r- ro> K\ CM CM ^rf'.OrOCO
O v-
COT-C\Jirur\t-tACOC\Jf^C\JC\J rCM^OCVO^v-Or
r-C\jC\JC\¡C\JC\JC\JC\jC\JC\JC\¡C\jC\|C\jC\iC\jCvJ\]C\Jv-OC\J
O C > 00 N- CQ O'. C ce C"- N- ¡>
CO NO CO
m^NCj OiOr-CM W'N ^ iT\O N CC C>0 r* C\J
CM CM CM CM CM rO NN n fA ro IN n ^ ^ ' N '


39
human vs. animal subjects and acute vs. chronic hypoxia.
First, the data available on animals will be reviewed,
followed by that extant on human subjects. Finally,
behavioral variables sensitive to changes in these
variables will be summarized.
The effects of chronic hypoxia on animals have not
been exhaustively studied (Davis, 1975). One project
which did examine this variable involved exposure of rats
to up to 36 hours of *10% oxygen. Davis (1975) observed
the behavior of these animals and concluded that they
appeared normal. Although hyperventilation was noted, no
permanent effects of the hypoxia were found. Hanbauer et
al. (1981) examined the effects of 10? O2 on rats exposed
for up to 4 weeks. They found that relatively short term
(2 days) hypoxia resulted in an increase in dopamine (DA)
content in the carotid body. An increase in norepine
phrine (WE) content was noted after 1 week of exposure.
As the carotid bodies are implicated in the control of
respiration, the authors concluded that different
mechanisms are operative in the adaptation to short and
long term hypoxia. Unfortunately, little other data have
been reported on the chronic effects of hypoxia on more
complex functioning. This is particularly problematic as
it are these functions which are thought to be affected in
humans exposed to hypoxia.
Certain aspects of acute hypoxia in rats have been
carefully examined. Gibson et al. (1981) reviewed the


105
non-verbal memory, expressive verbal fluency, anc
cognitive flexibility;
8) the subset of heevy snoring males with more than
5 events per hour appear to occupy an inter
mediate position between normals and those with
sleep apnea syndrome.


88
criteria. A wide range of age was achieved in this
sample, but neither non-parametric nor correlational
procedures revealed significant relationships between
increasing age and sleep disordered breathing. Although
several previous reports have noted relationships between
increasing age and sleep disordered breathing, typically
this relationship is moderate for apnea (Block et al.,
1979; Carskadon et al., 1980; Bixler et al., 1982).
Additionally, no other correlational data are available in
subjects selected for sleep and breathing complaints. It
may be speculated that the effect of snoring masks the
effect of age in this group. Alternatively, some quirk of
sampling may have generated this effect. Correlational
procedures, however, revealed substantial relationships
between weight and sleep disordered breathing. This
finding is consistent with several past reports. To
summarize, the present data support relationships between
increasing weight and nocturnal events, while
relationships between age and sleep disordered breathing
are not found in this sample of heavy snoring males,
perhaps because the effect of age is masked by the effect
of snoring.
The temporal characteristics of the apneic events
assessed by two hour interval suggested an increase in
events between '5 and 5 am, although differences in sleep
onset time and wakeup time render any interpretation of


2 V
10U
090
099
066
01
00
28
129
133
133
135
09
12
29
122
116
120
114
09
09
30
137
133
138
120
10
09
31
121
102
114
097
03
02
32
123
114
120
118
08
11
35
123
112
119
109
13
O'j
34
133
110
127
129
09
10
33
113
120
11(1
120
09
12
36
122
102
114
121
10
08
37
133
127
134
133
15
14
36
138
116
1 31
118
07
00
39
134
133
137
133
08
12
4
100
109
105
107
03
09
41
130
123
129
133
08
12
42
134
135
130
143
15
12
43
111
111
112
100
09
03
44
123
133
131
13b
09
13
43
136
142
131
143
13
14
46
120
114
119
137
13
11
WMSMQ = Wechsler Memory Scale Memory yuotient
Delayed Recall Logical Stories
Delayed Recall Visual Reproductions
^ Delayed Recoil Rey figure
Wisconsin Card Sort
6
OO
16
2
57
60
09
8
20
26
3
70
bO
19
8
25
26
4
57
5b
09
b
24
26
5
60
68
1b
7
1y
2U
0
53
44
11
u
20
27
4
70
77
13
5
3
24
1
49
42
Ot
7
30
26
4
80
72
11
7
21
27
4
71
70
1
7
26
25
5
54
50
11
7
34
25
5
66
6/
17
0
20
12
4
55
47
26
b
30
25
5
73
74
17
7
19
23
2
58
49
10
7
30
28
5
69
61
14
5
2b
28
5
53
50
18
7
13
17
3
50
49
17
7
28
26
4
74
65
10
7
2/
25
5
55
48
22
9
25
29
2
6b
60
13
3 U


93
have enough tree variance to respond to the respiratory
events of the night, Alternatively, the LEG scoring
system, which averages across 1 minute periods, might miss
the short arousals ana fragmentation resulting from
events. Whatever the reason, the lack of covariation of
sleep and respiratory measures is puzzling in light of the
report of Guilleminault et al. (1978) who described
arousals as a frequent concommitant of nocturnal
respiratory events, although Bixler et al. (1982) failed
to replicate this in a group of normal subjects.
Between group differences on the sleep variables were
also absent when subjects grouped by level of respiratory
distress were compared on sleep characteristics. Again it
may be speculated that the effect of an exacerbated first
night might obscure possible differences.
In summary, these snoring subjects appeared to
achieve a night of sleep comparable to tnose observed in
similar studies. Longer sleep latency, more awakenings,
and lighter sleep were observed. Significantly, these
changes characterize a first night effect (Agnew, Webb,
and Williams, 1966) of lighter sleep in a strange bed, and
may have been increased by the discomfort of the recording
procedures. Less than a chance number of correlations
were observed between nocturnal respiratory and sleep
parameters. This surprising result is attributed to an
exacerbated first night effect which probably functioned
to obscure possible respiratory disturbed sleep parameter


58
Table 3-2. Incidence of low and high levels of
apnea/hypopnea by age in 46 snoring males.
Age Group
U
Low Apnea/
Hypopnea
High Apnea/
Hypopnea
50-39
14
.71
0
40-49
7
.57
.29
50-59
11
.72
.27
60+
14
.43
.07
Overall
46
.62
.15
apnea/hypopnea. Note that subjects with high levels of
apnea/hypopnea are included in the overall total of
subjects with at least one event for the following
analyses. The classification of "high" levels of
apnea/hypopnea utilizes the criterion of an apnea +
hypopnea index greater than 5 a frequently used clinical
cut-off score. The table also presents the frequency of
respiratory distress broken down by age. No simple age
related trend is apparent, and a X2 analysis of the two
factors of age and respiratory distress vas not signi
ficant for those with at least one event (X2=3.2; p=NS),
or for those with high levels of apnea/hypopnea (X2=5*9;
p=NS).
A further investigation of possible age by level of
apnea activity involved division of the sample into those
above ana below 60 years of age crossed with those above
and below 5 events/hour. The X2 value for this comparison
was also nonsignificant (X2=0.6; p=NS).
Subjects were


TAELE OF CONTENTS
Page
ACKNOWLEDGMENTS ii
LIST OF TABLES v
ABSTRACT vii
CHAPTER
ChE INTRODUCTION 1
Definitions and Terminology 2
History and Cheracteristics of Sleep Apnea
Syndrome 3
Incidence of Sleep Apnea Activity 5
Temporal Characteristics of Apneas 13
Variables Associated With Sleep Apnea
Syndrome and Activity 17
Deficits Associated With Sleep Apnea
Syndrome 23
Cardiopulmonary Deficits 26
Arousal Deficits (Hypersomnolence) 28
Oxygen-Desaturation and Cognitive Deficits..30
Measurement of Deficits Found in Sleep Apnea
Syndrome 34
Measurement of Cardiopulmonary and Health
Deficits 34
Measurement of Arousal Deficits
(Hypersomnolence) 35
Measurement of Hypoxia and Its Sequelae 38
Statement of the Problem 46
TWO METHOD 48
Subjects 48
Apparatus 49
Measures 50
Procedure 54
Statistical Procedures.. 55
THREE RESULTS 56
Demographics and Incidence of Nocturnal
Respiratory Disorder 56
iii


32
of oxygen saturation in 20 post menapausal females was
reported by Block et al. (1980). Eleven of these women
exhibited desaturation with several subjects desaturating
to less than 85#. Dolly and Block (1982) studied a group
of 17 males and 3 females with a mean age of 49 years.
These normal subjects dropped from a mean baseline satura
tion of 96*3% to a mean maximum desaturation of 88.3$.
This evidence indicates that desaxurations of 10-12?' are
common in normal males and post menapausal females while
premenapausa1 females rarely, if ever, desaturate.
Oxygen saturation levels in SAS patients are not
commonly reported. Using an ear oximeter, Garay et al.
(1981) studied 11 SAS patients who were divided into 6
daytime eucapnics and 7 daytime hypercapnics. Eucapnics
fell from a baseline saturation of 95% to a mean maximum
desaturation of 73*5/, while hypercapnics fell from a
baseline saturation of Q9% to a maximum desaturation of
62.8# (calculated from tabled data). These date were
collected during daytime "nap" studies, however, and hence
is somewhat questionable. Berry and Elock (1983) recorded
9 heavy snoring males who suffered many apneac events.
Four subjects were somnolent, while 5 were not. Symptoma
tic subjects fell from a baseline of 95# to a mean maximum
desaturation of 44#, while asymptomatic subjects desatu-
rated from a baseline of 94# to a mean maximum desatura
tion of 83#.


CHAPTER TWO
METHOD
Subjects
Heavy snoring males were recruited via newspaper
advertisements, notices, and phone calls to a list o
aging subjects reported by Webb (1982), Participants were
required to be male, self reported heavy snorers, over 30
years o age, and in general good health (i.e., self
described healthy and normal and not under the active care
of a physician for illness). Subjects with a history of
head trauma or alcoholism were excluded. Potential
subjects who met these criteria were offered a chance to
participate in a "snoring study" in which they would
complete testing during an evening, sleep overnight in a
lab with several physiological parameters recorded, and
complete one week's worth of "sleep logs," for a payment
of fifty dollars. Volunteers were scheduled on a first
come, first served basis. A total of 60 subjects were
studied, with 9 dropped because of technical problems with
their overnight recordings and 5 excluded because of
health problems. Thus 46 subjects represented the final
sample. These 46 subjects had a mean age of 49*9 (sd
13-1) and represented a wide cross section of the
population, ranging from the unemployed to university
48


27
Zwillich et al. proposed that bradycardia during apnea
results from increased vagal tone mediated by carotid body
chemoreceptors. Fujita et al. (1981) reported on 12 male
SAS patients. All reportedly suffered cardiac arrhythmias
during sleep. The body of data suggests that SAS patients
frequently suffer from cardiac abnormalities such as
hypertension, arrhythmias, and bradycardia.
A second body of data exists which indicates that SAS
patients do not invariably have cardiopulmonary compli
cations. Coverdale et al. (1980) studied 14 patients, of
whom 8 were diagnosed with SAS. Only 2 of these patients
had evidence of corpulmonle, while 2 more had systemic
hypertension. None of these patients demonstrated severe
brady or tachycardia during sleep. In Orr et al. (1979),
8 patients, all of whom were technically diagnosable as
having SAS, all symptomatic (somnolent) patients had evi
dence of right sided heart failure as well as frequent
arrhythmias curing apneas. Asymptomatic subjects had none
of these symptoms. Kreiss et al. (1982) studied 26
Veterans Administration ward patients and iound 7 to have
a SAS. The SAS patients had significantly more signs of
congestive heart failure, but not hypertension or
angina. Ancoli-Israel et al. (1981) sampled 24 elderly
subjects with complaints of sleeping respiratory ais-
orders. Nine of these subjects were diagnosable as SAS,
but this group did not have significantly more heart


5
series of 50 patients diagnosed with SAS over a period of
six years. Age ranged from 28-62 with a mean age of
45.5* The sample consisted of 48 males and 2 females.
Thirty-nine were referred on the basis of excessive day
time sleepiness, while 7 were referred for loud snoring
and abnormal movements reported by their spouses. Sixty
percent of the patients were more than 15$ above ideal
weight. Other symptoms included hypnagogic halluci
nations, automatic behaviors, intellectual deterioration,
personality changes, impotence, morning headaches, and
hypertension. Veitzman, Kahn and Poliak (1980) report
data on a group of 10 SAS patients seen in their clinic.
These subjects displayed the following characteristics:
age between 38 and 47, obesity, male sex, nicotine depen
dence, hypertension, complaints of excessive daytime
sleepiness, and the presence of serious cardiac arrhyth
mias. (This group was selected from a larger sample of 38
patients.) Thus, pathological symptomatology associated
with a diagnosed SAS include excessive daytime somnolence,
cardiopulmonary complications and cognitive/intellectual
changes; a "typical" SAS patient is a middle aged male
who snores and is obese.
Incidence of Sleep Apnea Activity
The incidence of sleep apnea activity and sleep apnea
syndrome have been frequently confused. While sleep
apneas have been an important diagnostic tool for identi
fying sleep apnea syndrome, they have not been


13
difficulties had very high incidences of apnea (almost
100J2 for both males and females) with increase of inci
dence of AI>5 (55% for males, 23% for females). Sampling
without regard to sleep or respiratory difficulties is
largely confined to aging populations. Incidences of
apnea in these aging populations would seem to be rather
high (88% males, 72% females) while incidence of AI>5 seem
to be in the range of 32%-37%.
It is clear that different populations (as defined by
sampling strategies) suffer different incidences of sleep
apneas and AI>5. The clinician dealing with sleep dis
orders would be wise to consider the acturial base rates
of apnea and AI>5 provided by these data. It is apparent
that sex, age, and respiratory status are significantly
related to the incidence rates of SAA and SAS. These
factors must be considered in any study of apnea and asso
ciated deficits.
Temporal Characteristics of Apneas
The temporal characteristics of sleep apneas repre
sent a potentially important variable. Above it was shown
that a wide variability in number of sleep apneas occurs,
and the arbitrary nature of the division between SAA and
SAS subjects was noted. Therefore, it is of interest to
determine whether the duration of sleep apneas provides a
more reliable distinction between the two groups. Another
variable of interest involves the placement of apneas
within sleep, as Guilleminault et al. (1976) noted that


120
rO'rO^riO^OOOOOr OtOOO^OOr
N(\j r- v C\|
^oot-o o O'sTinv md ovcoA r- r--
CM ^-vD mo CO CO CM CT\ v t-cOf-T- 't lA
CM CM C\J
a 7- v- ^^r\jN^tri^vDNirvntn^inoir\'j)c\j
c^ cts xi o^ en o c'i ao cr. en o\ en ^ en en en en en en en en en
O A >- O A A O ^ X) CM CQ CO IAt- LHC0 NfOtO m
N'lTiLTiM^in' iT.VOsDNNO NC0 T
cncncncncncnencncncncncncn CTv cncncncncncncncn
OOCMs-^rOAOCMOaoOvO OfflOOOOOt^
a) Ti^r OCCO^ A KO
\S\ r~ \S\"~ V-
rA y- rA
CMOO^iTiO^O^OOOO 00 OOOOOOOOCNJ
OCMf ^O.^^MOOOO O O-OOOOOO
OOOOCn 0^-0^0000 C0 oooooooo
OOOOfMOrOCMOOOOOOOOOOOOO
A vC C0 Cn Q C\J A ^ AAD X- CQ O O v- C\J A -vi- LTVvG
CM CM CM CM CM ¡A -A A A rA A IA A '


£* ^ ^ ^ ^ V^J ^ \^> ,\J (V
C ui ^ ^ re
^ CT CT O' CT C? Ui er V CT A** en en efi en *o. c
e* O VC C C C \T Wj -* C ko ^vCU*^^
^ _i [\j u ^ i rvj 4* ^ i i ^ .__!
unroo^^cr rocrcroiv^j^crc^ecrro
CT -< CT cw cr -< cr -%: c cr -c ^ c -< -e c cr
ro-*-ii\j->i\;r\j-> -i-iiNjivro-^roivro-*
-i^cr.^c:^ ^uia^vc c ^c r
vj* rv er en Cz ru vx c C C re \r u. r\j [\j e* r\j e>j e*j ro ro f\: rv re fv ro i\j -> r\j re
* en o- vji C ^ crouiO'C^CvC^JC: crctvc^-
o Co f\: c o c n Cat c !T c tt u v£
f\J
^rccT'V)^iV^Cf\jf\-'C e?
CCC^O^UiCCa^CC
IV
c
ro rv v v
c->c^ococcoccccccccc
cr c v' c1 cr < v s c: *>; vc c: a cr c: -X t
crovicco^uicccc^cGNccc
ccccccccccccccccccc
cen-Ceri->ercc^
CCCCCCCCGCCCCCCCCCC
C^WCCCCC^fUCN'^fy-V;. V. e>- X
cccccoccccccccccccc
rv^f\j-k^^CJV-^CCUi re e> C -1
cccccocccoccccccccc
v ror po-^c
ou


ACKNOWLEDGMENTS
The author would like to thank his chairman, Dr. V/ilse
E. Webb, for his moral and material support, as well as
guidance through the completion of the project. Also due
thanks are the other members of the supervisory committee,
Dr. A. J. Elock, Dr. Russell M. Eauer, Dr. F. D. McGlynn,
and Dr. Rudy Vuchinich, for their support, advice, and
contributions. Several people provided technical assistance
and are due a vote of thanks: Dan Switzer, Alex Masterton,
Karla Smith, Alton Howard, Albert Briggs, and Edna Pearson.
The author wishes to acknowledge the moral anti
financial support of his parents, Mr. and Mrs. John G.
Berry, without which the project would never have been
undertaken. Also of considerable help were the
encouragement and advice of C. M. Carswell, D. Haymaker, and
S. Brody, all of whom deserve thanks. Cynthia Zimmerman
provided expert typing assistance for the manuscript.
ii


73
Table 3-11* Significant (p<.05) Pearson correlations between
nocturnal respiratory and sleep log variables in
46 snoring males.
Number
Minutes
of Total
Sleep
Nightly
Naps Bedtime
Latency
Wakenings
Apnea Index
Apnea+Hypopnea Index
Seconds in Events
Mean High Saturation
.237
.250
.359
Mean Low Saturation
Kumber of
.265
Desaturations >_4%
.240
Number of
Desaturations y]Q%
.240
Table 3-12. Means and
variables
by level
standard deviations of sleep log
derived from 46 snoring males grouped
of apnea/hypopnea.
No Apnea/
Low Apnea/
High Apnea/
Hypopnea
Hypopnea
Hypopnea
Daytime Sleepiness
1.9 (.9)
1.9(1.1)
2.4 (.3)
Number of Naps
0.2 (.2)
.3 (.4)
.3 (.3)
Minutes Mapping
0.0(0)
.2 (.5)
0 (0)
Total Bed Time
6.5(3.1)
5.8(3.4)
7.5 (.7)
Sleep Latency
10.5(8.1)
11.1(9.8)
8.4(4.7)
Number of Wakenings
1.3(0.8)
1.3 (.9)
1.3 (.2)
Minutes V, a king
2.0(1.4)
1.7(1.1)
1.7 (.3)


1C1
hypoxia. Acute hypoxia has recently been shown to signi
ficantly alter the synthesis of ACh, a neurotransrr.itter
implicated in memory and cognitive changes in young sub
jects experiencing cholinergic blockade, and older sub
jects experiencing age related disruption of cholinergic
pathways. The hypoxic alteration of ACh turnover is
proposed as the proximal cause of cognitive/memory
impairments found in subjects with nocturnal respiratory
events.
Although the cerebral hypoxia hypothesis is highly
speculative, it does generate predictions which might be
used to tost its efiicacy. For instance, subjects with
differing levels of hypoxia might be expected to exhibit
diiiering levels of ACh turnover, as well as cognitive
differences. Administration of nocturnal oxygen should
improve both cognitive and cholinergic variables. Addi
tionally, manipulation of cholinergic pathways might
improve cognition in subjects with nocturnal respiratory
events. However, until a demonstration of this sort is
carried out, this line of reasoning remains highly specu
lative.
In summary, the present findings indicate clear rela
tionships between nocturnal respiratory and cogni-
tive/neuropsychological variables. This result is con
sistent with past reports of neuropsychological
impairments in subjects experiencing hypoxia. Apparently,
heavy snorers with multiple apnea/hypopnea episodes may be


80
Table 3-15* Keans and standard deviations of cognitive
variables in 46 snoring males groupec by level
of apnea/hypopnea.
No Apnea/
Low Apnea/
High Apnea/
Variable
Hypopnea
Hypopnea
Hypopnea
VAIS Full IQ
124.9(11.2)
118.3(13.9)
111.5 (9.6)
WAIS Verbal IQ
123.7(12)
118.5(14.5)
114.3(10.4)
WAIS Performance IQ
122.8(11.5)
114.5(14.4)
100.1(12.5)a
Y/eschler Memory
Scale Memory
Quotient
124.7(14.5)
117.1(19.0)
107.8(18.7)
Delayed Recall
Logical Stories
9-9 (2.4)
6.6 (3.2)
5.8 (4.0)a
Delayed Recall
Visua1
Reproductions
10.5 (3.7)
10.4 (3.4)
f.1 (4.7)
Digit Span
6.6 (0.6)
6.8 (1.2)
6.6 (.6)
Delayed Recall
Rey Figure
25.6 (5.4)
23.2 (6.0)
15.1 (fc.4)s,b
Hooper Test
25.2 (2.6)
25.5 (4.6)
22.8 (4.1)
Wisconsin Card Sort
4.1 (1.2)
3.4 (1.5)
2.6(1.6)
Finger Tapping
Right Hand
62.6 (6.5)
57.5(10.4)
56.1 (8.3)
Finger Tapping
Left Hand
59.1 (9.0)
54.C (9.7)
54.8(13.1)
Verbal Fluency
14.6 (4.5)
13.1 (4.2)
So (2.6)
a
b
Significantly different from no sleep apnea/hypopnea.
Significantly different from low sleep apnea/hypopnea.


found in these subjects are taken as minimally comparable
with subjects experiencing the classic sleep apnea
syndrome, some support for the anecdotal reports of in
tellectual deterioration in SAS (Cuilleminault et el.,
1978) is derived. However, both these conclusions stand
in need of replication belore they may be confidently
asserted.
The mechanism underlying the link between nocturnal
respiratory and cognitive variables is a fascinating
enigma. To cate, two explanations have been proposed as
proximal causes of the changes observed in sleep apnea
syndrome. The first, or "sleep fragmentation" hypothesis
(Carskadon et al., 1980) notes the large number of
arousals which may accompany nocturnal respiratory events
and suggests that the resultant sleep fragmentation may
lead to daytime impairments, particularly in the
elderly. While intuitively appealing, there are several
difficulties with this explanation. First, the demonstra
tion of reliable cognitive deficits after sleep depriva
tion hrs proved surprisingly difficult (Webb, 1975).
Changes in cognitive test scores such as those used in the
present study have not been reliably observed following
sleep deprivation. A second line of evidence inconsistent
with the sleep fragmentation hypothesis is the observed
lack of correlation between sleep structure and cognitive
variables in normal subjects (Berry anc Webb, In press).
Although this covariation has been observed in subjects



PAGE 1

6/((3 $31($ $&7,9,7< $1' ,76 &21&20,7$176 ,1 $ 68%&/,1,&$/ 3238/$7,21 %\ '$9,' 7+20$6 5((' %(55< $ ',66(57$7,21 35(6(17(' 72 7+( *5$'8$7( 6&+22/ 2) 7+( 81,9(56,7< 2) )/25,'$ ,1 3$57,$/ )8/),//0(17 2) 7+( 5(48,5(0(176 )25 7+( '(*5(( 2) '2&725 2) 3+,/2623+< 81,9(56,7< 2) )/25,'$

PAGE 2

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

PAGE 3

7$(/( 2) &217(176 3DJH $&.12:/('*0(176 LL /,67 2) 7$%/(6 Y $%675$&7 YLL &+$37(5 &K( ,1752'8&7,21 'HILQLWLRQV DQG 7HUPLQRORJ\ +LVWRU\ DQG &KHUDFWHULVWLFV RI 6OHHS $SQHD 6\QGURPH ,QFLGHQFH RI 6OHHS $SQHD $FWLYLW\ 7HPSRUDO &KDUDFWHULVWLFV RI $SQHDV 9DULDEOHV $VVRFLDWHG :LWK 6OHHS $SQHD 6\QGURPH DQG $FWLYLW\ 'HILFLWV $VVRFLDWHG :LWK 6OHHS $SQHD 6\QGURPH &DUGLRSXOPRQDU\ 'HILFLWV $URXVDO 'HILFLWV +\SHUVRPQROHQFHf 2[\JHQ'HVDWXUDWLRQ DQG &RJQLWLYH 'HILFLWV 0HDVXUHPHQW RI 'HILFLWV )RXQG LQ 6OHHS $SQHD 6\QGURPH 0HDVXUHPHQW RI &DUGLRSXOPRQDU\ DQG +HDOWK 'HILFLWV 0HDVXUHPHQW RI $URXVDO 'HILFLWV +\SHUVRPQROHQFHf 0HDVXUHPHQW RI +\SR[LD DQG ,WV 6HTXHODH 6WDWHPHQW RI WKH 3UREOHP 7:2 0(7+2' 6XEMHFWV $SSDUDWXV 0HDVXUHV 3URFHGXUH 6WDWLVWLFDO 3URFHGXUHV 7+5(( 5(68/76 'HPRJUDSKLFV DQG ,QFLGHQFH RI 1RFWXUQDO 5HVSLUDWRU\ 'LVRUGHU LLL

PAGE 4

1RFWXUQDO 5HVSLUDWRU\ DQG +HDOWK 9DULDEOHV 1RFWXUQDO 5HVSLUDWRU\ DQG 6OHHS9nDNH 9DULDEOHV (/* 'DWD 'D\WLPH 6OHHSLQHVV 'DWD 6XEMHFWLYH 6OHHS $VVHVVPHQW 1RFWXUQDO 5HVSLUDWRU\ DQG 1HXURSV\FKRORJLFDO 9DULDEOHV )285 ',6&866,21 'HPRJUDSKLFV DQG ,QFLGHQFH RI 1RFWXUQDO 5HVSLUDWRU\ 'LVRUGHU 1RFWXUQDO 5HVSLUDWRU\ DQG +HDOWK 9DULDEOHV 1RFWXUQDO 5HVSLUDWRU\ DQG 6OHHS:DNH 9DULDEOHV ((* 'DWD 'D\WLPH 6OHHSLQHVV 6OHHS 4XHVWLRQQDLUH DQG 6OHHS /RJ 'DWD 1RFWXUQDO 5HVSLUDWRU\ DQG 1HXURSV\FKRORJLFDO 9DULDEOHV ),9( 6800$5< $1' &21&/86,216 $33(1',; $ 6&25,1* 35272&2/ )25 5(63,5$725< 9$5,$%/(6 % 5$: '$7$ 7$%/(6 5()(5(1&(6 %,2*5$3+,&$/ 6.(7&+

PAGE 5

/,67 2) 7$%/(6 7DEOH 3DJH ,QWHUUDWHU DJUHHPHQW IRU UHVSLUDWRU\ HYHQWV RI VQRULQJ PDOHV 'HPRJUDSKLF YDULDEOHV IURP VQRULQJ PDOHV ,QFLGHQFH RI ORZ DQG KLJK OHYHOV RI DSQHD K\SRSQHD E\ DJH LQ VQRULQJ PDOHV 6LJQLILFDQW Sf 3HDUVRQ FRUUHODWLRQV EHWZHHQ QRFWXUQDO UHVSLUDWRU\ YDULDEOHV DQG GHPRJUDSKLF YDULDEOHV LQ VQRULQJ PDOHV 0HDQV DQG VWDQGDUG GHYLDWLRQV IRU VHOHFWHG GHPRJUDSKLF DQG QRFWXUQDO UHVSLUDWRU\ YDULDEOHV LQ VQRULQJ PDOHV JURXSHG E\ OHYHO RO DSQHDK\SRSQHH 0HDQV DQG VWDQGDUG GHYLDWLRQV IRU KHDOWK UHODWHG YDULDEOHV LQ VQRULQJ PDOHV JURXSHG E\ OHYHO RI DSQHDK\SRSQHD 0HDQV DQG VWDQGDUG GHYLDWLRQV RI VHOHFWHG VOHHS YDULDEOHV IURP VQRULQJ PDOHV 0HDQV DQG VWDQGDUG GHYLDWLRQV RI VHOHFWHG VOHHS YDULDEOHV IURP VQRULQJ PDOHV JURXSHG E\ OHYHO RI DSQHDK\SRSQHD /HDQV DQG VWDQGDUG GHYLDWLRQV IRU GD\WLPH VOHHSLQHVV YDULDEOHV LQ VQRULQJ PDOHV JURXSHG E\ OHYHO RI DSQHDK\SRSQHD 6LJQLILFDQW Sf 3HDUVRQ FRUUHODWLRQV EHWZHHQ QRFWXUQDO UHVSLUDWRU\ DQG VOHHS TXHVWLRQQDLUH YDULDEOHV LQ VQRULQJ PDOHV 0HDQV DQG VWDQGDUG GHYLDWLRQV RI VOHHS TXHVWLRQQDLUH YDULDEOHV IURP VQRULQJ PDOHV JURXSHF E\ OHYHO RI DSQHDK\SRSQHD Y

PAGE 6

6LJQLILFDQW Sf 3HDUVRQ FRUUHODWLRQV EHWZHHQ QRFWXUQDO UHVSLUDWRU\ DQG VOHHS ORJ YDULDEOHV LQ VQRULQJ PDOHV 0HDQV DQG VWDQGDUG GHYLDWLRQV RI VOHHS ORJ YDULDEOHV LQ VQRULQJ PDOHV JURXSHG E\ OHYHO RI DSQHDK\SRSQHD 0XOWLYDULDWH UHJUHVVLRQ RI GHPRJUDSKLF DQG QRFWXUQDO UHVSLUDWRU\ YDULDEOHV RQ FRJQLWLYH VFRUHV LQ VQRULQJ PDOHV 6LJQLILFDQW Srf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r % 5HVSLUDWRU\ YDULDEOHV IURP VQRULQJ PDOHV % 6OHHS YDULDEOHV IURP VQRULQJ PDOHV YL

PAGE 7

$EVWUDFW RI 'LVVHUWDWLRQ 3UHVHQWHG WR WKH *UDGXDWH 6FKRRO RI WKH 8QLYHUVLW\ RI )ORULGD LQ 3DUWLDO )XOILOOPHQW RI WKH 5HTXLUHPHQWV IRU WKH 'HJUHH RI 'RFWRU RI 3KLORVRSK\ 6/((3 $31($ $&7,9,7< $1' ,76 &21&20,7$176 ,1 $ 68%&/,1,&$/ 3238/$7,21 %\ '$9,' 7+20$6 5((' %(55< $XJXVW &KDLUPDQ :LOVH % :HEE 0DMRU 'HSDUWPHQW &OLQLFDO 3V\FKRORJ\ 6OHHS $SQHD 6\QGURPH 6$6f LV D QRFWXUQDO UHVSLUDWRU\ GLVRUGHU ZLWK VHULRXV FRQVHTXHQFHV 3DWLHQWV ZLWK 6$6 H[SHULHQFH DSQHDV SDXVHV LQ EUHDWKLQJ RI RU PRUH VHFRQGVf DQG K\SRSQHDV GHFOLQHV LQ WKH DPSOLWXGH RI EUHDWKLQJ DFFRPSDQLHG E\ R[\JHQ GHVDWXUDWLRQf ZKLOH DVOHHS 5HFHQWO\ LW KDV EHFRPH DSSDUHQW WKDW DSQHDV DQG K\SRSQHDV PD\ RFFXU LQ RWKHUZLVH DSSDUHQWO\ QRUPDO VXEMHFWV $ VDPSOH VHOHFWHG WR GLVSOD\ D ZLGH UDQJH RI VOHHS GLVRUGHUHG EUHDWKLQJ KHDY\ VQRULQJ PDOHV LQ JRRG KHDOWK DQG RYHU \HDUV RI DJHf UHFHLYHG GHWDLOHG PHDVXUHPHQW RI WKHLU QRFWXUQDO UHVSLUDWLRQ VOHHS FRJQLWLYH VNLOOV DQG KHDOWK LQ DQ DWWHPSW WR FODULI\ WKH LPSDFW RI DSQHDV DQG K\SRSQHDV LQ D VXEFOLQLFDO SRSXODWLRQ )RUW\VL[ KHDY\ VQRULQJ PDOHV LQ JRRG KHDOWK ZLWK D PHDQ DJH RI \HDUV DQG D PHDQ ZHLJKW RI SRXQGV FRPSULVHG WKH VDPSOH 'XULQJ DQ H[SHULPHQWDO HYHQLQJ WKH\ YLL

PAGE 8

UHFHLYHG WHVWLQJ DQG ILOOHG RXW YDULRXV TXHVWLRQQDLUHV EHIRUH VOHHSLQJ RYHUQLJKW ZLWK WKHLU VOHHS DQG EUHDWKLQJ FRQWLQXRXVO\ UHFRUGHG 6L[W\WZR SHU FHQW RI WKHVH VXEMHFWV H[SHULHQFHG DW OHDVW RQH HYHQW ZKLOH b KDG PRUH WKDQ HYHQWV SHU KRXU RI VOHHS 0RVW HYHQWV RFFXUUHG LQ OLJKW VORZ ZDYH RU 5(0 VOHHS DQG ZHUH OLQNHG WR REHVLW\ 6XEMHFWV ZLWK KLJK OHYHOV RI DSQHDK\SRSQHD PRUH WKDQ SHU KRXUf H[SHULHQFHG VLJQLILFDQWO\ H[DFHUEDWHG R[\JHQ GHVDWXUDWLRQ UHODWLYH WR WKH UHPDLQLQJ VXEMHFWV 2[\JHQ GHVDWXUDWLRQ ZDV OLQNHG WR KLJKHU EORRG SUHVVXUH UHDGLQJV $ W\SLFDO ILUVW QLJKWrn HIIHFW RQ WKH VOHHS RI WKH VXEMHFWV LQ WKH ODE ZDV QRWHG UHVXOWLQJ LQ D OLJKWHU VOHHS WKDQ XVXDO DQG REVFXULQJ SRVVLEOH UHODWLRQVKLSV EHWZHHQ VOHHS DQG UHVSLUDWRU\ HYHQWV ZKLFK ZHUH QRW REVHUYHG 'HWHULRUDWLQJ RYHUQLJKW UHVSLUDWRU\ LQGLFHV ZHUH DVVRFLDWHG ZLWK LQFUHDVHG VOHHSLQHVV DQG QDSSLQJ +LJK OHYHOV RI DSQHDK\SRSQHD ZHUH DVVRFLDWHG ZLWK GHFOLQLQJ VFRUHV RQ WHVWV PHDVXULQJ QRQn YHUEDO LQWHOOLJHQFH YHUEDO DQG QRQYHUEDO PHPRU\ H[SUHVVLYH YHUEDO IOXHQF\ DQG FRJQLWLYH IOH[LELOLW\ ,W LV VSHFXODWHG WKDW K\SR[LD LQGXFHG FKDQJHV LQ FHUHEUDO FKROLQHUJLF V\QWKHWLF SDWKZD\V XQGHUOLH WKHVH FKDQJHV ,W LV FRQFOXGHG WKDW D VLJQLILFDQW VXEJURXS RI KHDY\ VQRULQJ PDOHV ZKR H[SHULHQFH PXOWLSOH DSQHDK\SRSQHDV DUH DW LQFUHDVHG ULVN RI VLJQLILFDQW R[\JHQ GHVDWXUDWLRQ GD\WLPH VOHHSLQHVV DQG FRJQLWLYH FKDQJHV 7KLV JURXS PD\ IDOO RQ D FRQWLQXXP ZLWK VOHHS DSQHD V\QGURPH SDWLHQWV YLLL

PAGE 9

&+$37(5 21( ,1752'8&7,21 6OHHS DSQHDV DUH UHVSLUDWRU\ SDXVHV GXULQJ VOHHS $FFRUGLQJ WR WKH $VVRFLDWLRQ RI 6OHHS 'LVRUGHUV &HQWHUV 1RVRORJ\ RI 6OHHS 'LVRUGHUV $6'& f} D VOHHS DSQHD V\QGURPH 6$6f LV D SRWHQWLDOO\ OHWKDO FRQGLWLRQ FKDUDFn WHUL]HG E\ PXOWLSOH DSQHDV H[FHVVLYH GD\WLPH VOHHSLQHVV DOWHUDWLRQV RI FRQVFLRXVQHVV DQG FDUGLRSXOPRQDU\ FRPSOLFDWLRQV 7KH DSQHD HSLVRGHV DUH WKRXJKW WR EH WKH FDXVDO IDFWRU EHKLQG WKH RWKHU GHILFLWV DVVRFLDWHA ZLWK 6$6 *XLOOHPLQDXOW HW DO f 3RO\VRPQRJUDSKLF UHFRUGLQJV LQ YDULRXV SRSXODWLRQV KDYH UHYHDOHG WKH RFFXUUHQFH RI VOHHS DSQHDV LQ WKH JHQHUDO SRSXODWLRQ DQG SDUWLFXODUO\ LQ DJLQJ SRSXODn WLRQV (VWLPDWHV RI WKH SUHYDOHQFH RI DW OHDVW D IHZ VOHHS DSQHDV LQ ROGHU VXEMHFWV KDYH UDQJHG DV KLJK DV e &DUVNDGRQ DQG 'HPHQW Ef &DUVNDGRQ HW DO 66&f KDYH VXJJHVWHG WKDW VOHHS DSQHDV LQ ROGHU VXEMHFWV PD\ EH LPSOLFDWHG LQ LQVRPQLD FDUGLDF DLOPHQWV DQG HYHQ VHQLOLW\ +RZHYHU FHUWDLQ UHSRUWV KDYH GRFXPHQWHG WKH SUHVHQFH RI PXOWLSOH VOHHS DSQHD HSLVRGHV LQ VXEMHFWV ZLWKRXW DSSDUHQW SDWKRORJ\ 2UU HW DO fr 7KLV HYLGHQFH REVFXUHV WKH UROH RI VOHHS DSQHDV LQ WKH RWKHU SDWKRORJLHV

PAGE 10

QRWHG LQ 6$6 SDWLHQWV 7KH OLQN EHWZHHQ DSQHDV DQG SDWKRn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n GURPH ZLOO EH GHWDLOHG DORQJ ZLWK PHDVXUHPHQWV WKRXJKW WR EH VHQVLWLYH WR WKHVH GHILFLWV 'HILQLWLRQV DQG 7HUPLQRORJ\ 6OHHS DSQHDV KDYH EHHQ GHILQHG DV FHVVDWLRQV RI DLUn IORZ DW WKH PRXWK DQG QRVH ZKLFK ODVW IRU RU PRUH VHFRQGV *XLOOHPLQDXOW HW DO f ,Q D VOHHS DSQHD V\QGURPH 6$6f PXOWLSOH VOHHS DSQHDV DUH DFFRPSDQLHG E\ VQRULQJ H[FHVVLYH GD\WLPH VOHHSLQHVV FDUGLRSXOPRQDU\ FRPSOLFDWLRQV DQG DOWHUHG VWDWHV RI FRQVFLRXVQHVV 7KH HDVH RI TXDQWLILFDWLRQ RI VOHHS DSQHDV KDV OHG WR WKHLU XVH DV DQ LPSRUWDQW WRRO LQ GLDJQRVLQJ 6$6 *XLOOHPLQDXOW

PAGE 11

HW DO f ILUVW VXJJHVWHG WKDW WKH RFFXUUHQFH RI RU PRUH DSQHDV LQ RU PRUH KRXUV RI VOHHS ZDV GLDJQRVWLF RI 6$6 ZKLOH *XLOOHPLQDXOW HW DO f SURSRVHG WKH XVH RI DQ DSQHD LQGH[ $, QXPEHU RI DSQHDVQXPEHU RI KRXUV VOHHSf LQ H[FHVV RI DV D FXWRII VFRUH WR GLDJQRVH WKH V\QGURPH /DWHU &DUVNDGRQ HW DO f FRPELQHG DSQHDV DQG K\SRSQHDV D UHGXFWLRQ LQ DLUIORZ RI b RU PRUH DW WKH PRXWK DQG QRVHf WR IRUP DQ DSQHD K\SRSQHD LQGH[ $+,f ZLWK XWLOL]HG DV D GLDJQRVWLF OHYHO RI VOHHS DSQHD V\QGURPH $V FDQ EH DSSUHFLDWHG IURP WKH SUHFHGLQJ GLVn FXVVLRQ WKH GLDJQRVWLF FULWHULD IRU 6$6 KDYH XQGHUJRQH D UDSLG HYROXWLRQ ZLWK UHPDUNDEO\ OLWWOH YDOLGDWLRQ ZRUN SXEOLVKHG +LVWRU\ DQG &KDUDFWHULVWLFV RI 6OHHS $SQHD 6\QGURPH 6OHHS $SQHD 6\QGURPH IDOOV ZLWKLQ WKH ODUJHU FDWHJRU\ RI 6OHHS 'LVRUGHUHG %UHDWKLQJ %ORFN f 5HVSLUDWRU\ GLVRUGHUV ZLWKLQ WKLV FDWHJRU\ LQYROYH FKDQJHV LQ EUHDWKLQJ SDWWHUQV GXULQJ VOHHS 6\QGURPHV GHVFULEHG ZLWKLQ WKLV FDWHJRU\ LQFOXGH 3LFNZLFNLDQ 6\QGURPH 6OHHS $SQHD 6\QGURPH &KURQLF 2EVWUXFWLYH /XQJ 'LVHDVH DQG V\QGURPHV RI 3ULPDU\ DQG 6HFRQGDU\ $OYHRODU +\SRYHQWLODWLRQ %ORFN *XLOOHPLQDXOW HW DO f 7KH ILUVW GLVRUGHU GHVFULEHG LQ WKLV JURXS ZDV WKH 3LFNZLFNLDQ 6\QGURPH 7KHVH SDWLHQWV ZHUH LQLWLDOO\ LGHQWLILHG E\ D FKURQLF GD\WLPH K\SHUVRPQROHQFH REHVLW\ &KH\QH6WRNHV EUHDWKLQJ K\SHUFDSQLD DQG K\SR[HPLD 7KLV

PAGE 12

V\QGURPH W\SLFDOO\ UHVXOWV LQ FRPSOLFDWLRQV VXFK DV ULJKW VLGHG FRQJHVWLYH KHDUW IDLOXUH SXOPRQDU\ K\SHUWHQVLRQ DQG SHULSKHUDO HGHPD %ORFN f $Q H[DFHUEDWLRQ RI GLVRUGHUHG EUHDWKLQJ DSQHDf DV ZHOO DV K\SHUFDSQLD DQG K\SR[HPLD RIWHQ VHYHUHf RFFXUV ZLWK VOHHS RQVHW 7UHDWn PHQW FRQVLVWV RI R[\JHQ WKHUDS\ UHVSLUDWRU\ VWLPXODQWV DQG PDLQWHQDQFH RI ZDNHIXOQHVV 7KH GLVFRYHU\ RI D VOHHS LQGXFHG H[DFHUEDWLRQ RI WKH V\PSWRPV DFFRPSDQ\LQJ 3LFNZLFNLDQ 6\QGURPH VSXUUHG D V\VWHPDWLF LQYHVWLJDWLRQ RI VOHHS UHVSLUDWRU\ HYHQWV *DVWDXW HW DO f 7KH QHZ HPSKDVLV RQ QRFWXUDO UHVSLUDWLRQ OHG WR D VHULHV RI LQYHVWLJDWLRQV ZKLFK GRFXn PHQWHG VOHHS LQGXFHG UHVSLUDWRU\ G\VUK\WKPLDV LQ QRQREHVH VXEMHFWV /XJDUDVL HW DO f $ QHZ JURXS RI VOHHS GLVRUGHUHG EUHDWKHUV GHVLJQDWHG DV 6OHHS $SQHD 6\QGURPH HPHUJHG ZLWK VHYHUDO IHDWXUHV ZKLFK GLVWLQJXLVKHG WKHP IURP 3LFNZLFNLDQV $PRQJ WKHVH GLIIHUHQFHV ZHUH ZHLJKW LH 6$6 SDWLHQWV ZHUH QRW QHFHVVDULO\ REHVHf GD\WLPH UHVSLUDWRU\ FRQWURO LH 6$6 SDWLHQWV ZHUH W\SLFDOO\ QRW K\SHUFDSQLF GXULQJ ZDNLQJ DV ZHUH 3LFNZLFNLDQVf VH[ LH ZKLOH $b RI 3LFNZLFNLDQV ZHUH IHPDOH YHU\ IHZ SUHPHQDSDXVDO IHPDOHV KDYH EHHQ LGHQWLILHG ZLWK 6$6f DQG DJH LH 6$6 SDWLHQWV DUH W\SLFDOO\ \RXQJHU WKDQ 3LFNZLFNLDQVf%ORFN f 7ZR ODUJH VDPSOHV RI 6$6 SDWLHQWV KDYH EHHQ UHSRUWHG LQ WKH OLWHUDWXUH DOORZLQJ FKDUDFWHUL]DWLRQ RI W\SLFDO 6$6 SDWLHQWV *XLOOHPLQDXOW HW DO f UHSRUWHG RQ D

PAGE 13

VHULHV RI SDWLHQWV GLDJQRVHG ZLWK 6$6 RYHU D SHULRG RI VL[ \HDUV $JH UDQJHG IURP ZLWK D PHDQ DJH RI r 7KH VDPSOH FRQVLVWHG RI PDOHV DQG IHPDOHV 7KLUW\QLQH ZHUH UHIHUUHG RQ WKH EDVLV RI H[FHVVLYH GD\n WLPH VOHHSLQHVV ZKLOH ZHUH UHIHUUHG IRU ORXG VQRULQJ DQG DEQRUPDO PRYHPHQWV UHSRUWHG E\ WKHLU VSRXVHV 6L[W\ SHUFHQW RI WKH SDWLHQWV ZHUH PRUH WKDQ DERYH LGHDO ZHLJKW 2WKHU V\PSWRPV LQFOXGHG K\SQDJRJLF KDOOXFLn QDWLRQV DXWRPDWLF EHKDYLRUV LQWHOOHFWXDO GHWHULRUDWLRQ SHUVRQDOLW\ FKDQJHV LPSRWHQFH PRUQLQJ KHDGDFKHV DQG K\SHUWHQVLRQ 9HLW]PDQ .DKQ DQG 3ROLDN f UHSRUW GDWD RQ D JURXS RI 6$6 SDWLHQWV VHHQ LQ WKHLU FOLQLF 7KHVH VXEMHFWV GLVSOD\HG WKH IROORZLQJ FKDUDFWHULVWLFV DJH EHWZHHQ DQG REHVLW\ PDOH VH[ QLFRWLQH GHSHQn GHQFH K\SHUWHQVLRQ FRPSODLQWV RI H[FHVVLYH GD\WLPH VOHHSLQHVV DQG WKH SUHVHQFH RI VHULRXV FDUGLDF DUUK\WKn PLDV 7KLV JURXS ZDV VHOHFWHG IURP D ODUJHU VDPSOH RI SDWLHQWVf 7KXV SDWKRORJLFDO V\PSWRPDWRORJ\ DVVRFLDWHG ZLWK D GLDJQRVHG 6$6 LQFOXGH H[FHVVLYH GD\WLPH VRPQROHQFH FDUGLRSXOPRQDU\ FRPSOLFDWLRQV DQG FRJQLWLYHLQWHOOHFWXDO FKDQJHV D W\SLFDO 6$6 SDWLHQW LV D PLGGOH DJHG PDOH ZKR VQRUHV DQG LV REHVH ,QFLGHQFH RI 6OHHS $SQHD $FWLYLW\ 7KH LQFLGHQFH RI VOHHS DSQHD DFWLYLW\ DQG VOHHS DSQHD V\QGURPH KDYH EHHQ IUHTXHQWO\ FRQIXVHG :KLOH VOHHS DSQHDV KDYH EHHQ DQ LPSRUWDQW GLDJQRVWLF WRRO IRU LGHQWLn I\LQJ VOHHS DSQHD V\QGURPH WKH\ KDYH QRW EHHQ

PAGE 14

GHILQLWLYHO\ GHPRQVWUDWHG WR EH WKH QHFHVVDU\ DQG VXIILn FLHQW FDXVH RI WKH V\QGURPH 7KXV IRU WKH SXUSRVHV RI WKH IROORZLQJ UHYLHZ WKH IRFXV ZLOO EH RQ RFFXUUHQFH DQG OHYHO RI VOHHS DSQHD DFWLYLW\ LQ ERWK VOHHS DSQHD V\QGURPH VXEMHFWV DQG RWKHU SRSXODWLRQV 7KH IXOO UDQJH RI VOHHS DSQHD DFWLYLW\ ZLOO EH UHIOHFWHG E\ QRWLQJ WKH SUHVHQFH RI VOHHS DSQHD DFWLYLW\ DQG DQ LQGLFDWLRQ RI WKH TXDQWLWDn WLYH OHYHO RI WKH DFWLYLW\ ZLOO EH SUHVHQWHG LQ WKH DSQHD LQGH[ $,f RU WKH DSQHD K\SRSQHD LQGH[ $+,f 5HFDOO WKDW DQ $, RI KDV EHHQ IUHTXHQWO\ XWLOL]HG DV D FXWRII VFRUH IRU LGHQWLI\LQJ 6$6 SDWLHQWV $ UHYLHZ RI WKH OLWHUDWXUH VXJJHVWV WKDW DW OHDVW WKUHH VDPSOLQJ VWUDWHJLHV KDYH EHHQ XVHG IRU VWXG\LQJ WKH LQFLGHQFH RI VOHHS DSQHD DFWLYLW\ 6XEMHFWV KDYH EHHQ VHOHFWHG RQ WKH EDVLV RI RQH RI WKH IROORZLQJ WKUHH FULWHULD f DEVHQFH RI VOHHS RU UHVSLUDWRU\ FRPSODLQWV f SUHVHQFH RI VOHHS RU UHVSLUDWRU\ FRPSODLQWV RU f VDPSOLQJ ZLWKRXW UHJDUG WR WKHVH YDULDEOHV EHFDXVH GLDJQRVLV RI 6$6 LV RIWHQ GHSHQGHQW RQ QXPEHU RI VOHHS DSQHDV RU OHYHO RI 6D$f PRVW UHVHDUFKHUV UHSRUW QXPEHU RI DSQHDV REVHUYHG DOWKRXJK VRPH UHSRUW RQO\ RQ QXPEHU RI VXEMHFWV ZKR H[FHHG RQH RI WKH FULWHULD IRU 6$6 $ UHYLHZ RI VWXGLHV ZLWK VDPSOHV RI RU JUHDWHU ZLOO EH SUHVHQWHG ZLWK FDUHIXO URWH RI WKH VDPSOLQJ VWUDWHJ\ XWLOL]HG DV ZHOO DV WKH GHSHQGHQW PHDVXUH RI DSQHD

PAGE 15

2QH VDPSOLQJ VWUDWHJ\ ZKLFK KDV EHHQ SXUVXHG LV WR H[FOXGH VXEMHFWV ZKR FRPSODLQ RI VOHHS RU UHVSLUDWRU\ GLIILFXOWLHV 6FUHHQLQJ IRU VOHHS FRPSODLQWV YDULHV ZLGHO\ *XLOOHPLQDXOW HW DO f VWXGLHG D VDPSOH RI QRUPDOV SUHVXPDEO\ QRQFRPSODLQLQJ VXEMHFWVf 7KHVH VXEMHFWV UDQJHG LQ DJH IURP \HDUV 0DOHV VKRZHG D PHDQ QXPEHU RI DSQHDV RI UDQJH f ZKLOH IHPDOHV VKRZHG D PHDQ QXPEHU RI DSQHDV RI UDQJH f 7KXV QRQH RI WKHVH QRQFRPSODLQLQJ VXEMHFWV SUHVHQWHG D FOLQLFDO V\QGURPH RI VOHHS DSQHD DOWKRXJK VOHHS DSQHDV GLG RFFXU DSSDUHQWO\ ZLWKRXW SDWKRORJ\ %ORFN HW DO f VWXGLHG VXEMHFWV 0 ) f 0DOHV LQ WKH VDPSOH KDG D PHDQ DJH RI ZKLOH IHPDOHV KDG D PHDQ DJH RI $Q\ SRWHQWLDO VXEMHFWV ZKR FRPSODLQHG RI EUHDWKLQJ GLIILFXOn WLHV RU VOHHS GLVWXUEDQFHV ZHUH H[FOXGHG IURP WKH VWXG\ 7ZHOYH PDOHV f KDG HSLVRGHV RI VOHHS DSQHD ZKLOH RQO\ IHPDOHV f VXIIHUHG DV ZHOO )RU WKRVH VXEMHFWV PDOHV KDG D PHDQ RI HSLVRGHV ZKLOH IHPDOHV KDG D PHDQ RI :LWKLQ WKLV VDPSOH WKH QXPEHU RI DSQHDV ZDV QRW VLJQLILFDQWO\ FRUUHODWHG ZLWK LQFUHDVLQJ DJH DOWKRXJK RFFXUUHQFH RI R[\JHQ GHVDWXUDWLRQ D FRQFRPLWWDQW RI GLVRUGHUHG EUHDWKLQJf ZDV SRVLWLYHO\ FRUUHODWHG ZLWK DJH $GGLWLRQDOO\ DOWKRXJK WKH PHDQ QXPEHU RI VOHHS DSQHDV ZDV QRW VLJQLILFDQWO\ GLIIHUHQW EHWZHHQ VH[HV 0 ) f WKH PRUH VHQVLWLYH PHDVXUH RI DSQHDV ZLWK GHVDWXUDWLRQ ZDV 0 ) f :HEE f VWXGLHG & PDOHV ZLWK D PHDQ DJH RI 6XEMHFWV ZHUH H[FOXGHG IURP

PAGE 16

H WKH VWXG\ LI WKH\ UHSRUWHG DQ\ VHULRXV VOHHS FRPn SODLQW 7ZR RI WKH VXEMHFWV UHSRUWHG FRQVXPLQJ K\SQRWLFV IRU VOHHS LQGXFWLRQ LQGLFDWLQJ WKDW WKH FULWHULRQ XVHG LQ WKLV VWXG\ ZDV OHVV VWULQJHQW WKDQ WKDW RI %ORFN HW VL f UHSRUW 1LQH RI WKH VXEMHFWV bf VXIIHUHG DW OHDVW RQH HSLVRGH R VOHHS DSQHD ZLWK D PHDQ QXPEHU RI HSLVRGHV RI $OO RI WKH VXEMHFWV ZLWK HSLVRGHV RI VOHHS DSQHD ZHUH ROGHU WKDQ WKH PHDQ DJH RI WKH VDPSOH OHDGLQJ :HEE WR FRQFOXGH WKDW DSQHD LV UHODWHG WR DJH %L[OHU HW DO f FRQVWUXFWHG DQ DJH VWUDWLLLHG SDWLHQW UHSUHVHQWDWLYH VDPSOH RI VXEMHFWV 0 ) f ,Q SHUKDSV WKH PRVW ULJRURXV VFUHHQLQJ IRU VOHHS FRPn SODLQWV LQ WKLV JURXS RI VWXGLHV %L[OHU HW DO HOLPLQDWHG DQ\ VXEMHFW ZLWK D VOHHS FRPSODLQW RU FKURQLF PHGLFDWLRQ XVDJH )RU WKH HQWLUH VDPSOH PDOHV bf DQG IHPDOHV bf KDG DW OHDVW RQH HSLVRGH RI VOHHS DSQHD )RU VXEMHFWV XQGHU rb IRU WKRVH EHWZHHQ DQG b DQG WKRVH b VXIIHUHG HSLVRGHV RI VOHHS DSQHD )RU WKH VXEVDPSOH RYHU b KDG VOHHS DSQHD DFWLYLW\ ZLWK HSLVRGHV IRU PDOHV DQG HSLVRGHV IRU IHPDOHV ,Q WKLV VDPSOH WKHUH ZDV D QRQVLJQLILFDQW WUHQG WRZDUG PRUH DSQHD DFWLYLW\ LQ PDOHV DQG D VLJQLILFDQW SRVLWLYH FRUUHODWLRQ EHWZHHQ DJH DQG VOHHS DSQHD %ORFN HW DO f UHSRUW GDWD RQ SRVW PHQDSDXVDO ZRPHQ ZLWK D PHDQ DJH RI (LJKW ZRPHQ bf KDG HSLn VRGHV RI VOHHS DSQHD ZLWK D PHDQ QXPEHU RI HSLVRGHV RI

PAGE 17

7KHUH ZDV D VLJQLILFDQW FRUUHODWLRQ EHWZHHQ DSQHD DQG DJH )RU QRQFRPSODLQLQJ VXEMHFWV WKHQ PDOHV VXIIHU D KLJKHU LQFLGHQFH RI DSQHD DFWLYLW\ LQ YLUWXDOO\ HYHU\ VWXG\ ZKLFK FRPSDUHV VH[ HIIHFWV )RU PDOHV LQFLGHQFHV UDQJH ZLGHO\ IURP D ORZ RI b %L[OHU HW DO ,f WR D KLJK RI b :HEE f ZKLOH IHPDOHV UDQJH IURP D ORZ RI b %L[OHU HW DO f WR D KLJK RI b %ORFN HW DO f 7KHVH GLIIHUHQFHV DUH SUREDEO\ GXH DW OHDVW LQ SDUW WR DJH GLIIHUHQFHV LQ PDOH VDPSOHV DV ZHOO DV DJH GLIIHUHQFHV LQ IHPDOHV RQO\ DV WKH\ UHIOHFW UQHQD SDXVDO VWDWXVf 5HSRUWHG QXPEHU RI DSQHDV UDQJH IURP :HEE f WR %L[OHU HW DO f LQ PDOHV DQG IURP *XLOOHPLQDXOW HW DO f WR %L[OHU HW DO f LQ IHPDOHV 0DQ\ VWXGLHV UHSRUW D SRVLWLYH FRUUHODWLRQ EHWZHHQ QXPEHU RI DSQHDV DQG LQFUHDVLQJ DJH $ VHFRQG VDPSOLQJ WHFKQLTXH ZKLFK KDV EHHQ XWLOL]HG KDV EHHQ VWXG\LQJ VXEMHFWV ZLWK VOHHS DQG EUHDWKLQJ FRPn SODLQWV .DOHV HW DO f VWXGLHG VXEMHFWV ZLWK SULPDU\ FRPSODLQWV RI LQVRPQLD 7KHUH ZHUH PDOHV DQG IHPDOHV LQ WKH VDPSOH ZLWK D PHDQ DJH RI 6OHHS DSQHD DFWLYLW\ ZDV QRWHG LQ b RI WKLV JURXS 0 rb ) bf ZLWK D PHDQ QXPEHU RI HSLVRGHV IRU PDOHV DQG IRU IHPDOHV 6XEMHFWV ZLWK VOHHS DSQHD DFWLYLW\ ZHUH VLJQLILFDQWO\ ROGHU f WKDQ WKRVH ZLWKn RXW f $QFROL,VUDHO HW DO f UHFUXLWHG VXEMHFWV ZKRVH DQVZHUV WR TXHVWLRQQDLUHV SURYRNHG

PAGE 18

VXVSLFLRQ R LQVRPQLD RU QLJKWWLPH EUHDWKLQJ DQG PXVFXODU HYHQWV (OHYHQ PDOHV [ f DQG IHPDOHV [ f ZHUH VWXGLHG $OO RI WKHVH VXEMHFWV VDYH RQH PDOH DQG RQH IHPDOH VXIIHUHG DW OHDVW RQH HSLVRGH RI VOHHS DSQHD )RU WKRVH VXEMHFWV ZLWK DW OHDVW RQH HSLVRGH RI VOHHS DSQHD PDOHV KDG D PHDQ RI HSLVRGHV RI DSQHD ZKLOH IHPDOHV KDG D PHDQ RI HSLVRGHV ,Q FRQWUDVW WR WKH .DOHV HW DO VWXG\ PDOHV KDG D KLJK OHYHO RI VOHHS DSQHD $,!f ZKLOH IHPDOHV PHW WKLV FULWHULRQ $ UHODn WLRQVKLS EHWZHHQ DJH DQG DSQHD ZDV QRW UHSRUWHG LQ WKLV VDPSOH SRVVLEO\ EHFDXVH RI UHVWULFWHG UDQJH 7KHVH WZR VWXGLHV RI VOHHS GLVWXUEHG VXEMHFWV SURYLGH FRQIOLFWLQJ GDWD 7KDW LV WKH LQVRPQLDF DORQH VDPSOH GLG QRW GHPRQn VWUDWH DQ DSQHD LQFLGHQFH ZKLFK ZDV VXEVWDQWLDOO\ GLIIHUHQW IURP D QRQFRPSODLQLQJ VDPSOH bf} ZKLOH WKH VDPSOH UHFUXLWHG IRU VOHHS LQGXFWLRQ UHVSLUDWRU\ DQG PXVFXODU GLIIHUHQFHV VKRZHG YHU\ GLVFUHSDQW DSQHD LQFLn GHQFHV QHDUO\ ef 2QH SRWHQWLDO H[SODQDWLRQ LQYROYHV DJH DV WKH $QFROL,VUDHO VDPSOH LV VXEVWDQWLDOO\ ROGHU WKDQ .DOHVn DOWKRXJK WKH DJLQJ VDPSOHV KDYH QRW UHSRUWHG WKLV OHYHO RI DSQHD EHIRUH $Q DOWHUQDWLYH H[SODQDWLRQ LV WKDW VDPSOLQJ LQVRPQLDFV GRHV QRW JHQHUDWH D KLJK QXPEHU RI DSQHDV ZKHUHDV DJLQJ VOHHS DQG UHVSLUDWRU\ GLVWXUEHG VXEMHFWV VKRZ D YHU\ KLJK LQFLGHQFH RI DSQHD ,W ZRXOG VHHP WKHQ WKDW WKH UHOHYDQW FRPSODLQW ZRXOG LQYROYH WKRVH UHSRUWHG E\ $QFROL,VUDHO HW DO ZKLFK DUH LQ DGGLWLRQ WR WKRVH QRWHG E\ .DOHV HW DO 7KHVH DUH DV

PAGE 19

QHDUO\ DV FDQ EH MXGJHGf TXHVWLRQV VXUURXQGLQJ UHVSLUDWLRQ 7KH ILQDO FDWHJRU\ RI LQFLGHQFH VWXGLHV LQYROYHV WKRVH ZKLFK GR QRW FRQVLGHU VOHHS FRPSODLQWV LQ WKHLU VDPSOLQJ $V LQ WKH RWKHU VWXGLHV WKHVH YDU\ LQ WKHLU JHQHUDOL]DELOLW\ &DUVNDGRQ DQG 'HPHQW Ef UHFUXLWHG DJLQJ VXEMHFWV LQFOXGLQJ PDOHV [ f DQG IHPDOHV [ f 7KHLU H[FOXVLRQ FULWHULD DUH VRPHZKDW SUREOHPDWLF LQYROYLQJ H[FOXGLQJ VXEMHFWV ZKR VSRQn WDQHRXVO\ FRPSODLQ RI VOHHS SUREOHPV 7KXV LW LV XQFOHDU ZKDW WKLV LV D UHSUHVHQWDWLYH VDPSOH RI VSHFLILn FDOO\ ZKHWKHU WKH GLVWULEXWLRQ RI WKRVH QRW VSRQWDQHRXVO\ FRPSODLQLQJ RI VOHHS GLVWXUEDQFHV PD\ EH GLIIHUHQW IURP WKRVH ZKR DFWXDOO\ GR QRW VXIIHU IURP VOHHS GLVWXUn EDQFHV 7KHVH FRQVLGHUDWLRQV DVLGH &DUVNDGRQ DQG 'HPHQW UHSRUW WKDW PDOHV bf DQG IHPDOHV bf GHPRQn VWUDWHG DW OHDVW RQH VOHHS DSQHD HSLVRGH ZKLOH PDOHV f DQG IHPDOHV bf KDYH KLJK OHYHOV RI VOHHS DSQHD $,!f 0DOHV ZLWK DSQHD VKRZHG D PHDQ QXPEHU RI ZKLOH IHPDOHV VKRZHG D PHDQ QXPEHU RI 6H[ GLIIHUHQFHV DSSHDU WR KDYH IDOOHQ RXW DW WKLV DJH ZKLOH WKH UHVWULFWHG UDQJH RI WKLV H[WUHPH DJH VDPSOH GRHV QRW OHQG LWVHOI WR D FRUUHODWLRQDO DQDO\VLV RI DJH YHUVXV DSQHD ,Q DQRWKHU VWXG\ RI OLPLWHG JHQHUDOL]DELOLW\ .UHLVV HW DO f UDQGRPO\ VHOHFWHG LQSDWLHQWV RQ D 9HWHUDQV $GPLQLVWUDWLRQ PHGLFDO ZDUG 7KHVH SDWLHQWV DOO SUHVXPDEO\ PDOH QXPEHUHG )XUWKHU GDWD UHSRUWHG ZHUH

PAGE 20

VNHWFK\ EXW bf PHW WKH VOHHS DSQHD V\QGURPH FULWHULRQ RI HSLVRGHV RI VOHHS DSQHD 7KH IDFW WKDW WKHVH VXEMHFWV ZHUH SDWLHQWV RQ D PHGLFDO ZDUG REYLRXVO\ UHVWULFWV JHQHUDOL]DELOLW\ RI WKHVH ILQGLQJV )LQDOO\ LQ WKH RQO\ WUXO\ UDQGRP VDPSOLQJ UHSRUWHG $QFROL,VUDHO HW DO f SUHVHQW LQLWLDO GDWD RQ UDQGRPO\ VHOHFWHG VXEn MHFWV )RUW\ VXEMHFWV VH[ QRW UHSRUWHGf ZLWK D PHDQ DJH RI ZHUH VWXGLHG 7KLUWHHQ f PHW WKH VOHHS DSQHD V\QGURPH FULWHULRQ RI RU PRUH DSQHDV 7KH VOHHS DSQHD VXEJURXS ZDV ROGHU WKDQ WKH IXOO VDPSOH 7KH GDWD FROOHFWHG IURP UHODWLYHO\ UDQGRP VDPSOLQJ DUH SULPDULO\ EDVHG RQ DJLQJ SRSXODWLRQV !f 7KH GDWD IURP WKHVH VWXGLHV VXJJHVW WKDW EHWZHHQ $QFROL ,VUDHO HW DO f DQG &DUVNDGRQ DQG 'HPHQW Ef RI DJLQJ VXEMHFWV VXIIHU KLJK OHYHOV RI VOHHS DSQHD $,!f $Q HYHQ ODUJHU SHUFHQWDJH SHUKDSV JUHDWHU WKDQ } VXIIHU DW OHDVW VRPH HSLVRGHV RI VOHHS DSQHD $ PDMRU GHILFLHQF\ LQ WKHVH UDQGRPO\ VDPSOHG V[XGLHV LV D ODFN RI GDWD RQ \RXQJHU VXEMHFWV 7KXV GDWD FROOHFWHG XQGHU GLIIHUHQW VDPSOLQJ SURFHn GXUHV \LHOG YDULDEOH LQFLGHQFH UDWHV RI VOHHS DSQHD DQG $,!6 6XEMHFWV ZLWKRXW VOHHS RU UHVSLUDWRU\ FRPSODLQWV KDYH D YHU\ ORZ LQFLGHQFH SHUKDSV f RI $,! EHWZHHQ DQG RI PDOHV VXUYH\HG KDG DW OHDVW VRPH DSQHD ZKLOH IHPDOHV UDQJHG EHWZHHQ DQG H[SHULHQFLQJ VRPH DSQHD 6DPSOLQJ VXEMHFWV ZLWK VOHHS RU UHVSLUDWRU\ FRPn SODLQWV VXJJHVWHG WKDW VXEMHFWV VXIIHULQJ UHVSLUDWRU\

PAGE 21

GLIILFXOWLHV KDG YHU\ KLJK LQFLGHQFHV RI DSQHD DOPRVW IRU ERWK PDOHV DQG IHPDOHVf ZLWK LQFUHDVH RI LQFLn GHQFH RI $,! b IRU PDOHV b IRU IHPDOHVf 6DPSOLQJ ZLWKRXW UHJDUG WR VOHHS RU UHVSLUDWRU\ GLIILFXOWLHV LV ODUJHO\ FRQILQHG WR DJLQJ SRSXODWLRQV ,QFLGHQFHV RI DSQHD LQ WKHVH DJLQJ SRSXODWLRQV ZRXOG VHHP WR EH UDWKHU KLJK b PDOHV b IHPDOHVf ZKLOH LQFLGHQFH RI $,! VHHP WR EH LQ WKH UDQJH RI bb ,W LV FOHDU WKDW GLIIHUHQW SRSXODWLRQV DV GHILQHG E\ VDPSOLQJ VWUDWHJLHVf VXIIHU GLIIHUHQW LQFLGHQFHV RI VOHHS DSQHDV DQG $,! 7KH FOLQLFLDQ GHDOLQJ ZLWK VOHHS GLVn RUGHUV ZRXOG EH ZLVH WR FRQVLGHU WKH DFWXULDO EDVH UDWHV RI DSQHD DQG $,! SURYLGHG E\ WKHVH GDWD ,W LV DSSDUHQW WKDW VH[ DJH DQG UHVSLUDWRU\ VWDWXV DUH VLJQLILFDQWO\ UHODWHG WR WKH LQFLGHQFH UDWHV RI 6$$ DQG 6$6 7KHVH IDFWRUV PXVW EH FRQVLGHUHG LQ DQ\ VWXG\ RI DSQHD DQG DVVRn FLDWHG GHILFLWV 7HPSRUDO &KDUDFWHULVWLFV RI $SQHDV 7KH WHPSRUDO FKDUDFWHULVWLFV RI VOHHS DSQHDV UHSUHn VHQW D SRWHQWLDOO\ LPSRUWDQW YDULDEOH $ERYH LW ZDV VKRZQ WKDW D ZLGH YDULDELOLW\ LQ QXPEHU RI VOHHS DSQHDV RFFXUV DQG WKH DUELWUDU\ QDWXUH RI WKH GLYLVLRQ EHWZHHQ 6$$ DQG 6$6 VXEMHFWV ZDV QRWHG 7KHUHIRUH LW LV RI LQWHUHVW WR GHWHUPLQH ZKHWKHU WKH GXUDWLRQ RI VOHHS DSQHDV SURYLGHV D PRUH UHOLDEOH GLVWLQFWLRQ EHWZHHQ WKH WZR JURXSV $QRWKHU YDULDEOH RI LQWHUHVW LQYROYHV WKH SODFHPHQW RI DSQHDV ZLWKLQ VOHHS DV *XLOOHPLQDXOW HW DO f QRWHG WKDW

PAGE 22

QRUPDOV PLJKW H[SHULHQFH DSQHDV GXULQJ 5(0 ZLWKRXW SDWKRORJLFDO FRQVHTXHQFH 7KHUHIRUH GDWD DYDLODEOH RQ WKHVH WZR LVVXHV ZLOO EH UHYLHZHG LQFOXGLQJ RQO\ VWXGLHV ZLWK D VDPSOH VL]H JUHDW WKDQ ZLWK RQH H[FHSWLRQf 5HSRUWV RQ QRUPDO VXEMHFWV EHORZ \HDUV RI DJH DUH VRPHZKDW UDUH DOWKRXJK VRPH GR H[LVW %L[OHU HW DO f LQFOXGHG D JURXS DJHG \HDUV 0 ) f 2QH VXEMHFW LQ WKLV JURXS KDG VOHHS DSQHDV f ZLWK D PHDQ GXUDWLRQ RI VHFRQGV %L[OHU HW DO QRWHG WKDW PRVW DSQHDV RFFXUUHG LQ 6WDJHV DQG 5(0 VOHHS 3DUW RI WKH VDPSOH RI %ORFN HW DO f WKH IHPDOHV KDG D PHDQ DJH RI 7KUHH RI WKHVH VXEMHFWV VKRZHG DSQHDV ZLWK D PHDQ GXUDWLRQ RI VHFRQGV $JDLQ D GLVSURSRUWLRQDWH QXPEHU RI DSQHDV RFFXUUHG LQ 6WDJHV DQG 5(0 7KXV LQ QRUPDO VXEMHFWV XQGHU DSQHDV DUH VRPHZKDW UDUH ZLWK D PHDQ GXUDWLRQ RI DSSUR[LPDWHO\ VHFRQGV ZKHQ WKH\ GR RFFXU 0RVW DSQHDV RFFXU LQ OLJKW DQG 5(0 VOHHS 0RUH GDWD DUH DYDLODEOH RQ QRUPDO VXEMHFWV EHWZHHQ DQG *XLOOHPLQDXOW HW DO f GR QRW UHSRUW RQ GXUDWLRQ RI DSQHDV LQ WKHLU VDPSOH DJHG EXW WKH\ QRWH WKDW DSQHDV RFFXUUHG RQO\ LQ 6WDJHV DQG 5(0 %L[OHU HW DO f UHSRUWHG WKDW LQ WKHLU \HDU ROG JURXS VXEMHFWV H[SHULHQFHG VOHHS DSQHD ZLWK D PHDQ GXUDWLRQ RI VHFRQGV $ GLVSURSRUWLRQDWH QXPEHU RI DSQHDV RFFXUUHG LQ 6WDJHV DQG 5(0 %ORFN HW DO f VWXGLHG PDOHV ZLWK D PHDQ DJH RI ZLWK H[SHULHQFLQJ VOHHS DSQHDV 7KHVH DSQHDV KDG D PHDQ

PAGE 23

GXUDWLRQ RI VHFRQGV DQG RFFXUUHG PRVW IUHTXHQWO\ f LQ 6WDJHV DQG 5(0 (LJKW RI DJLQJ IHPDOHV LQ WKH UHSRUW E\ %ORFN HW DO f KDG VOHHS DSQHDV ZLWK D PHDQ GXUDWLRQ RI VHFRQGV 6L[W\WZR SHUFHQW RI DSQHD HSLVRGHV RFFXUUHG LQ 6WDJHV RU 5(0 7KXV LQ PLGGOH DJHG QRUPDO VXEMHFWV LQFLGHQFH RI VOHHS DSQHD DSSHDUV WR ULVH ZLWK D PHDQ GXUDWLRQ RI EHWZHHQ DQG VHFRQGV SHUKDSV VOLJKWO\ ORQJHU WKDQ LQ \RXQJHU JURXSV $JDLQ DSQHDV RFFXUUHG PRVWO\ LQ OLJKW 6WDJHV DQG f DQG 5(0 VOHHS 'XUDWLRQ GDWD RQ DJLQJ QRUPDOV DUH OLPLWHG &DUVNDGRQ DQG 'HPHQWnV Ef VWXG\ RI DJLQJ QRUPDOV 0 ) f GRHV QRW UHSRUW GXUDWLRQ RI DSQHD HYHQWV 6LPLODU\ $QFROL,VUDHO HW DO f GR QRW SURYLGH WKLV LQIRUPDWLRQ RQ WKHLU DJHG QRUPDOV [ f %L[OHU HW DO f QRWHG D KLJK LQFLGHQFH RI DSQHDV LQ WKHLU VXEn MHFWV RYHU EXW IDLOHG WR UHSRUW D PHDQ GXUDWLRQ IRU WKLV JURXS *LYHQ WKH VNHWFK\ GHWDLOV UHSRUWHG IRU WKLV JURXS LW LV GLIILFXOW WR GUDZ DQ\ FRQFOXVLRQV DERXW FKDQJHV LQ WHPSRUDO GXUDWLRQ RU SODFHPHQW RI DSQHDV LQ DJLQJ QRUPDOV 7R VXPPDUL]H WKH WHPSRUDO GDWD LQ QRUPDOV DSQHDV DSSHDU VRPHZKDW UDUHO\ LQ QRUPDOV EHORZ ZLWK D PHDQ GXUDWLRQ RI VHFRQGV ,QFLGHQFH RI DSQHDV LQFUHDVH EHWZHHQ DQG ZLWK D ZLGHU YDULDWLRQ LQ PHDQ GXUDn WLRQV UHSRUWHG VHFRQGVf $SQHDV RFFXU PRVW IUHTXHQWO\ LQ OLJKW DQG 5(0 VOHHS LQ QRUPDOV RI DOO DJHV

PAGE 24

7HPSRUDO GDWD RQ VXEMHFWV GLDJQRVHG ZLWK VOHHS DSQHD V\QGURPH RU $,!f SULPDULO\ LQFOXGH SDWLHQWV EHWZHHQ WKH DJHV RI DQG 7KH UHDVRQ IRU WKH ODFN RI ROGHU 6$6 VXEMHFWV LV XQFOHDU SDUWLFXODUO\ LQ OLJKW RI GDWD UHSRUWHG DERYH GHPRQVWUDWLQJ YHU\ KLJK SHUFHQWDJHV RI 6$6 LQ FHUWDLQ JURXSV RI DJLQJ VXEMHFWV *DUD\ HW DO f VWXGLHG VOHHS DSQHD V\QGURPH VXEMHFWV 7KH\ GLYLGHG WKHVH PDOHV DQG IHPDOH LQWR WZR JURXSV EDVHG RQ WKHLU GD\WLPH UHVWLQJ &2 OHYHOV 6XEMHFWV ZKR ZHUH HXFDSQLF GXULQJ WKH GD\ KDG D PHDQ DSQHD GXUDWLRQ RI VHFRQGV ZKLOH WKRVH ZKR ZHUH K\SHUFDSQLF KDG D PHDQ GXUDWLRQ RI VHFRQGV 1R GDWD RQ GLVWULEXn WLRQ UHODWLYH WR VOHHS VWDJHV ZHUH UHSRUWHG *XLOOHPLQDXOW HW DO f VWXGLHG SUHGRPLQDQWO\ PDOH [ DJH f VOHHS DSQHD V\QGURPH SDWLHQWV $SQHDV LQ WKLV SRSXODWLRQ DYHUDJHG VHFRQGV LQ GXUDWLRQ $SQHDV ZHUH ORQJHVW LQ 5(0 DQG IHZ RFFXUUHG LQ 6WDJHV RU VOHHS )XMLWD HW DO f VWXGLHG PDOH 6$6 SDWLHQWV ZLWK D PHDQ DJH RI r $SQHD HSLVRGHV DYHUDJHG VHFRQGV LQ GXUDWLRQ :HLW]PDQ HW DO &f VWXGLHG 6$6 PDOHV ZKRVH DJH UDQJHG IURP r 7KH PHDQ GXUDWLRQ RI DSQHDV LQ WKLV JURXS ZDV VHFRQGV $OO DSQHDV RFFXUUHG LQ OLJKW RU 5(0 VOHHS 2UU HW DO f VWXGLHG K\SHU VRPQROHQW PDOHV GLDJQRVHG ZLWK 6$6 $ FRQWURO JURXS RI QRQK\SHUVRPQROHQW VXEMHFWV ZDV PDWFKHG RQ QXPEHU RI REn VWUXFWLRQV SHU PLQXWH 0HDQ DJH ZDV IRU DV\PSWRPDWLF DQG IRU V\PSWRPDWLF VXEMHFWV $V\PSWRPDWLF VXEMHFWV

PAGE 25

KDG DSQHDV ZLWK D PHDQ GXUDWLRQ RI VHFRQGV ZKLOH V\PSWRPDWLF VXEMHFWV KDG D PHDQ GXUDWLRQ RI VHFRQGV 6HYHUDO FKDUDFWHULVWLFV HPHUJH IURP WKH JURXS RI UHSRUWV RQ 6$6 VXEMHFWV 7KHUH LV ZLGH YDULDELOLW\ LQ GXUDWLRQ RI DSQHDV ZLWK UHSRUWHG PHDQV UDQJLQJ IURP VHFRQGV 7KH GDWD DYDLODEOH RQ SODFHPHQW RI DSQHDV LQ VOHHS VXJJHVW WKDW LQ 6$6 VXEMHFWV PRVW DSQHDV RFFXU LQ OLJKW DQG 5(0 VOHHS &RPSDULVRQ RI GDWD RQ QRUPDO DQG 6$6 VXEMHFWV UHYHDOV VXEVWDQWLDO RYHUODS EHWZHHQ PHDQ GXUDWLRQ RI DSQHDV LQ 6$6 VXEMHFWV VHFRQGVf DQG 6$6 VXEMHFWV VHFRQGVf 3DUWLFXODUO\ LQWHUHVWLQJ ZDV WKH 2UU HW DO f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

PAGE 26

DVVRFLDWHG YDULDEOHV LQFOXGH VH[ DJH VQRULQJ DQG REHVLW\ 7KH GDWD DYDLODEOH RQ WKHVH YDULDEOHV DQG WKHLU UHODWLRQVKLSV WR VOHHS DSQHD DFWLYLW\ ZLOO EH UHYLHZHG IROORZHG E\ D EULHI GLVFXVVLRQ RI WKHLU UHOHYDQFH WR H[SHULPHQWDO PHWKRGRORJ\ 5HJDUGLQJ VH[ YDULDEOHV DV SUHYLRXVO\ QRWHG *XLOOHUDLQDXOW HW DO f IRXQG b RI WKHLU 6$6 SDWLHQWV WR EH PDOHV ,Q WKH %ORFN HW DO f VWXG\ RI QRQVOHHS GLVWXUEHG VXEMHFWV b RI WKH PDOHV LQ WKH VDPSOH KDG DW OHDVW RQH HSLVRGH RI DSQHD ZKLOH RQO\ b RI WKH IHPDOHV VR VXIIHUHG %L[OHU HW DO f QRWHG D WUHQG WRZDUG PRUH VOHHS DSQHD DFWLYLW\ LQ PDOHV %ORFN HW DO f VWXGLHG SRVW PHQDSDXVDO IHPDOHV DQG IRXQG DSQHD DFWLYLW\ LQ b RI WKHP OHDGLQJ WKHP WR VXJJHVW WKDW VRPH IDFWRU DVVRFLDWHG ZLWK PHQDSDXVH HJ SURn JHVWHURQHf PLJKW VRPHKRZ SURWHFW SUHPHQDSDXVDO IHPDOHV IURP DSQHDV DQG WKDW WKLV IDFWRU DORQJ ZLWK LWV EHQLJQ LQIOXHQFH ZDV ORVW ZLWK PHQDSDXVH 7KLV H[SODQDWLRQ SUHn GLFWHG WKH ORZ OHYHO RI DSQHD DFWLYLW\ LQ \RXQJHU IHPDOHV DV ZHOO DV WKH KHLJKWHQHG DSQHD DFWLYLW\ RI DJLQJ IHPDOHV :KLOH WKH GDWD RI &DUVDNGRQ DQG 'HPHQW Ef VKRZ DJHG IHPDOHV WR EH FORVH WR DJHG PDOHV LQ SHUFHQWDJH H[KLELWLQJ VOHHS DSQHD DFWLYLW\ D UHSRUW E\ 6PDOOZRRG HW DO f IRXQG QR HYLGHQFH RI VOHHS DSQHD DFWLYLW\ LQ WKHLU HOGHUO\ SRVW PHQDSDXVDOf IHPDOHV ZKLOH HOGHUO\ PDOHV VKRZHG VLJQLILFDQW OHYHOV RI DSQHD DFWLYLW\ 7KH OLPLWDWLRQV RI WKH VDPSOH VL]H LQ WKH 6PDOOZRRG HW DO

PAGE 27

UHSRUW PLJKW H[SODLQ WKLV GLVFUHSDQF\ DV UHSUHVHQWLQJ D VDPSOLQJ ELDV 7DNHQ WRJHWKHU WKLV HYLGHQFH VWURQJO\ VXSSRUWV WKH QRWLRQ WKDW LQ \RXQJ DQG PLGGOH DJHG SRSXODn WLRQV VOHHS DSQHD LV D PDOH SKHQRPHQRQ $GGLWLRQDOO\ HYLGHQFH SRLQWV WRZDUG D FORVLQJ RI WKH JDS EHWZHHQ IHPDOHV DQG DJH PDWFKHG PDOHV DV WKH IRUPHU JURXS SDVVHV PHQDSDXVH $ VHFRQG IDFWRU ZKLFK KDV EHHQ LPSOLFDWHG LQ VOHHS DSQHD DFWLYLW\ LV DJH :HEE f IRXQG WKDW KLV VXEn MHFWV H[KLELWLQJ VOHHS DSQHD DFWLYLW\ ZHUH VLJQLILFDQWO\ ROGHU WKDQ KLV IXOO VDPSOH 7KLV ZDV DOVR WKH FDVH IRU WKH VXEMHFWV RI .DOHV HW DO f DQG $QFROL,VUDHO HW DO f 6PDOOZRRG HW DO f IRXQG WKDW VOHHS DSQHD ZDV DQ DJH GHSHQGHQW SKHQRPHQRQ ZLWK PDOH VXEMHFWV RYHU H[KLELWLQJ VLJQLILFDQWO\ PRUH DSQHDV WKDQ WKRVH XQGHU 2WKHU HYLGHQFH LQFOXGHV WKDW RI %L[OHU HW DO f ZKR IRXQG D SRVLWLYH FRUUHODWLRQ EHWZHHQ DJH DQG DSQHD LQ WKHLU IXOO VDPSOH DV GLG %ORFN HW DO f LQ WKHLU SRVW PHQDSDXVDO IHPDOHV ,W VHHPV FOHDU IURP WKHVH GDWD WKHQ WKDW LQFUHDVLQJ DJH LV D IDFWRU SUHGLVSRVLQJ IRU VOHHS DSQHD DFWLYLW\ 6QRULQJ KDV EHHQ FORVHO\ DVVRFLDWHG ZLWK VOHHS DSQHD /XJDUHVL HW DO f 7KLV PD\ EH GXH WR D FDXVDO UHODWLRQVKLS RU D VLQJOH FRPPRQ SDWKZD\ /XJDUHVL HW DO SURSRVH WKH ODWWHU H[SODQDWLRQ 6SHFLILFDOO\ WKH\ VXJJHVW WKDW ERWK VQRULQJ DQG 6$6 DUH GXH WR D VOHHS LQGXFHG VWHQRVLV RI WKH XSSHU DLUZD\ :LWK WKLV SUHPLVH

PAGE 28

LQ PLQG WKH\ FRQGXFWHG D VXUYH\ RI VXEMHFWV RQ WKH LQFLGHQFH RI VQRULQJ &KURQLF VQRUHUV LQFOXGHG RI WKH PDOHV DQG b RI WKH IHPDOHV ,Q DQ HODERUDWLRQ RI WKLV VWXG\ /XJDUHVL HW DO f TXHVWLRQHG LQGLYLn GXDOV ,Q WKLV VDPSOH b RI WKH PDOHV DQG b RI WKH IHPDOHV ZHUH FKURQLF VQRUHUV 6QRULQJ LQFUHDVHG ZLWK DJH DQG E\ \HDUV RI DJH b RI WKH PDOHV DQG b RI WKH IHPDOHV VQRUHG 2I QRWH ZDV WKH REVHUYDWLRQ WKDW K\SHUn WHQVLRQ RFFXUUHG PRUH IUHTXHQWO\ LQ VQRUHUV WKDQ QRQ VQRUHUV VXJJHVWLQJ WKDW WKH IRUPHU ZDV PRUH DW ULVN IRU FDUGLRYDVFXODU FRPSOLFDWLRQV 9LUWXDOO\ HYHU\ VWXG\ RI 6$6 SDWLHQWV QRWHV WKH SUHn VHQFH RI VQRULQJ LQ DOO WKHVH VXEMHFWV 6XOOLYDQ DQG ,VVD %ORFN &RYHUGDOH HW DO f 7KLV LV WUXH IRU ERWK SUHGRPLQDQWO\ REVWUXFWLYH DQG SUHGRPLQDWHO\ FHQWUDO 6$6 VXEMHFWV *XLOOHPLQDXOW HW DO f +RZn HYHU WKH SUHVHQFH RI KHDY\ VQRULQJ LQ DV\PSWRPDWLF QRQK\SHUVRPQROHQWf DSQHDV LV DOVR ZHOO GRFXPHQWHG 2UU HW DO )LVKHU HW DO %HUU\ DQG %ORFN f DV ZHOO DV LQ RWKHUZLVH QRUPDO VXEMHFWV 7KHVH GDWD PD\ EH WDNHQ DV HYLGHQFH IRU VQRULQJ DV D QHFHVVDU\ EXW QRW VXIILFLHQW FRQGLWLRQ IRU WKH RFFXUUHQFH RI VOHHS DSQHD 7KH UROH RI ERG\ ZHLJKW LQ 6$6 LV OHVV FOHDU 5HFDOO WKDW 3LFNZLFNLDQ SDWLHQWV DUH LQYDULDEO\ REHVH DV ZHOO DV K\SHUFDSQLF *XLOOHPLQDXOW HW DO f VWUHVV WKDW 6$6 PD\ RFFXU LQ QRQREHVH DV ZHOO DV REHVH VXEn MHFWV +RZHYHU %ORFN f VWDWHV WKDW KH KDV IRXQG 6$6

PAGE 29

VXEMHFWV WR EH XQLIRUPO\ REHVHr $ UHYLHZ RI WKH OLWHUDWXUH LV LQ RUGHU WR FODULI\ WKLV SRLQW )LUVW QRUPDO WKHQ REHVH WKHQ 6$6 JURXSV ZLOO EH UHYLHZHG %L[OHU HW DO f IRXQG WKDW WKHLU VXEJURXS RI VXEMHFWV ZLWK VOHHS DSQHD DFWLYLW\ ZHLJKHG VLJQLILFDQWO\ PRUH WKDQ WKH UHPDLQGHU RI WKH VDPSOH .DOHV HW DO f IRXQG WKDW ZLWKLQ WKHLU VDPSOH RI LQVRPQLDFV DQG QRUPDOV D VLJQLILFDQW SRVLWLYH FRUUHODWLRQ HPHUJHG EHWZHHQ D ZHLJKWKHLJKW UDWLR DQG DSQHD DFWLn YLW\ %ORFN HW DO f IRXQG D SRVLWLYH FRUUHODWLRQ EHWZHHQ DSQHDV R[\JHQ GHVDWXUDWLRQ DQG ZHLJKW ZLWKLQ WKHLU QRUPDO PDOHV ZKLOH IHPDOHV GLG QRW GHPRQVWUDWH WKLV UHODWLRQVKLS 8VLQJ DQ ROGHU IHPDOH VDPSOH %ORFN HW DO f IRXQG D SRVLWLYH FRUUHODWLRQ EHWZHHQ ZHLJKW DQG R[\JHQ GHVDWXUDWLRQ EXW QRW DSQHD 7KXV LQ QRUPDO VXEn MHFWV WKHUH DSSHDUV WR EH D SRVLWLYH UHODWLRQVKLS EHWZHHQ ZHLJKW DQG DSQHD DFWLYLW\ DOWKRXJK WKLV PD\ EH UHVWULFWHG WR PDOHV $ UHVHDUFK DSSURDFK ZKLFK EHDUV RQ WKLV TXHVWLRQ LQYROYHV VWXG\LQJ REHVH VXEMHFWV DORQH 7ZR VWXGLHV RQ VXFFHVVLYH UHIHUUDOV IRU JDVWULF E\SDVV VXUJHU\ ZHUH QRWHGr 6LFNOHVWHHO HW DO f IRXQG WKDW RI VXFn FHVVLYH UHIHUUDOV DOO PDOHV H[KLELWHG VOHHS DSQHDV ZKLOH RQO\ RI WKH IHPDOHV VKRZHG HYLGHQFH RI R[\JHQ GHVDWXUDWLRQ DORQH +DUPRQ HW DO f VWXGLHG JDVWULF E\SDVV SDWLHQWV DQG IRXQG RI PDOHV WR VXIIHU DSQHD DV

PAGE 30

RSSRVHG WR QRQH RI WKH IHPDOHV 7KLV HYLGHQFH SURYLGHV IXUWKHU HYLGHQFH IRU D UROH RI LQFUHDVLQJ ZHLJKW LQ DSQHD 7KH GDWD RQ V\PSWRPDWLF 6$6 VXEMHFWV DUH PRUH FRPn SOH[r 6HYHUDO VWXGLHV 6XOOLYDQ DQG ,VVD ,YHLW]PDQ HW DO & *DUD\ HW DO =ZLOOLFK HW DO f UHSRUW WKDW DOO WKHLU 6$6 VXEMHFWV ZHUH REHVH DOWKRXJK GHILQLWLRQV RI REHVLW\ HLWKHU YDU\ ZLGHO\ RU UHPDLQ XQn VSHFLILHG ,Q FRQWUDVW *XLOOHPLQDXOW HW DO f UHSRUWHG WKDW RQO\ RI WKHLU 6$6 VDPSOH ZDV JUHDWHU WKDQ 6 DERYH LGHDO ZHLJKW &RYHUGDOH HW DO f LQGLFDWH WKDW RQO\ WZR RI WKHLU VXEMHFWV SURYHG WR EH REHVH DFFRUGLQJ WR WKHLU GHILQLWLRQ RI LGHDO ZHLJKWf 7KHVH UHSRUWV VXJJHVW WKDW PRVW DOWKRXJK QRW DOO 6$6 VXEMHFWV DUH REHVH 6HYHUDO LQYHVWLJDWRUV KDYH VRXJKW WR FRPSDUH ZHLJKW LQ V\PSWRPDWLF VRPQROHQWf DQG DV\PSWRPDWLF VXEMHFWV ZLWK HQRXJK VOHHS DSQHD HSLVRGHV WR EH GLDJQRVHG DV 6$6 2UU HW DO f FRPSDUHG V\PSWRPDWLF 6$6 VXEMHFWV ZLWK PDWFKHG DV\PSWRPDWLF VXEMHFWV ,Q WKLV VDPSOH V\PSWRPDn WLF 6$6 VXEMHFWV ZHUH KHDYLHU WKDQ DV\PSWRPDWLF VXEn MHFWV ([DPLQDWLRQ RI WKH GDWD RI %HUU\ DQG %ORFN f LQGLFDWHV WKDW WKHLU V\PSWRPDWLF VXEMHFWV ZHUH KHDYLHU WKDQ WKHLU DV\PSWRPDWLF VXEMHFWV NJ YV NJf 6WDQGLQJ LQ FRQWUDVW WR WKHVH GDWD LV D VWXG\ UHSRUWHG E\ )LVKHU HW DO f ZKLFK LQFOXGHG VXEMHFWV UHIHUUHG IRU 6$6 HYDOXDWLRQ %DVHG RQ QXPEHU RI DSQHDV SHU QLJKW )LVKHU HW DO GLYLGHG WKHLU VXEMHFWV LQWR JURXSV *URXS

PAGE 31

, KDG DSQHDV SHU QLJKW *URXS ,, KDG EXW DSQHDV SHU QLJKW ZKLOH *URXS ,,, KDG OHVV WKDQ DSQHDV SHU QLJKW 7KXV *URXSV DQG ,, ZHUH GLDJQRVDEOH DV 6$6 !f ZKLOH *URXS ,,, ZDV QRW 2EHVLW\ ZDV SUHVHQW LQ *URXSV DQG ,, EXW LQ RQO\ VXEMHFWV LQ *URXS ,,, +RZHYHU WKHVH GLIIHUHQFHV ZHUH QRW VWDWLVWLFDOO\ VLJQLILn FDQW 7DNHQ WRJHWKHU WKLV HYLGHQFH VXJJHVWV D SRVLWLYH UHODWLRQVKLS EHWZHHQ ZHLJKW DQG DSQHD LQ WKH JHQHUDO SRSXn ODWLRQ +RZHYHU 6$6 SDWLHQWV DUH QRW LQYDULDEO\ REHVH )XUWKHU ZHLJKW ORVV IDLOV WR UHVXOW LQ D GHFUHDVH LQ DSQHDV LQ PDQ\ 6$6 SDWLHQWV 7KXV WKH UHODWLRQVKLS EHWZHHQ REHVLW\ DQG IXOO EORZQ 6$6 LV OHVV WKDQ FOHDU 7KH SUHFHGLQJ SDJHV KDYH LQYHVWLJDWHG WKH FXUUHQWO\ XQGHUVWRRG UROH RI VHYHUDO YDULDEOHV VH[ DJH VQRULQJ DQG ZHLJKWf LQ VOHHS DSQHD DFWLYLW\ 3UHVHQWO\ DYDLODEOH GDWD VXJJHVW WKDW LQ \RXQJ DQG PLGGOH DJHG VXEMHFWV DSQHD LV D SUHGRPLQDQWO\ PDOH SKHQRPHQRQ DOWKRXJK SRVW PHQD SDXVDO IHPDOHV PD\ FORVH WKH JDS ,QFUHDVLQJ DJH LV UHn ODWHG WR VOHHS DSQHD DFWLYLW\ 6QRULQJ ZKLFK RFFXUV LQ D VLJQLILFDQW SHUFHQWDJH RI WKH JHQHUDO SRSXODWLRQ LV WKRXJKW WR EH D QHFHVVDU\ EXW QRW VXIILFLHQW FRQGLWLRQ IRU 6$6 :HLJKW ZDV VKRZQ WR KDYH D SRVLWLYH UHODWLRQVKLS WR VOHHS DSQHD DFWLYLW\ LQ WKH JHQHUDO SRSXODWLRQ DOWKRXJK LWV SUHVHQFH RU UROH LQ 6$6 LV HTXLYRFDEOH 7KHVH FRQFOXVLRQV KDYH LPSRUWDQW LPSOLFDWLRQV IRU H[SHULPHQWDO GHVLJQV GLUHFWHG DW XQGHUVWDQGLQJ VOHHS DSQHD

PAGE 32

DFWLYLW\ DQG LWV FRUUHODWHV LQ D VXEFOLQLFDO SRSXODWLRQ ,W LV FOHDU IURP WKH HDUOLHU UHYLHZ RI DSQHD LQFLGHQFH WKDW VDPSOLQJ IURP WKH JHQHUDO SRSXODWLRQ ZRXOG UHVXOW LQ D UHODWLYHO\ ORZ SRSXODWLRQ RI LQGLYLGXDOV ZLWK VOHHS DSQHD DQ LQHIILFHQW DSSURDFK IURP WKH SHUVSHFWLYH RI WKH OLPLWHG UHVRXUFHV DYDLODEOH IRU UHVHDUFK 7KXV VHOHFWLRQ YDULDEOHV PXVW EH FRQVLGHUHG ZLWK RQH H\H RQ HIILFLHQF\ DQG WKH RWKHU RQ SURWHFWLQJ WKH JHQHUDOL]DELOLW\ RI UHVXOWV 6HOHFWLRQ FULWHULD VKRXOG DOVR DLP WRZDUG HOXFLn GDWLQJ WKH UHODWLYH UROHV DQG SRVVLEOH LQWHUDFWLRQV RI WKHVH YDULDEOHV LQ VOHHS DSQHD 7KHUHIRUH VHOHFWLRQ YDULDEOHV VKRXOG KDYH DV VWUDLJKWIRUZDUG D UHODWLRQVKLS DV SRVVLEOH ZLWK VOHHS DSQHD VHHNLQJ WR DQVZHU PRUH TXHVn WLRQV WKDQ WKH\ UDLVH :LWK WKHVH FDYHDWV LQ PLQG WKH YDULDEOHV DYDLODEOH DV VHOHFWLRQ FULWHULRQ ZLOO EH UHYLHZHG DQG FRQFOXVLRQV GUDZQ DERXW WKHLU UHODWLYH PHULWV 7KH ILUVW YDULDEOH FRQVLGHUHG VH[ KDV EHHQ UHODWLYHO\ ZHOO LQYHVWLJDWHG ,W LV FOHDU WKDW VOHHS DSQHD DFWLYLW\ LV SULPDULO\ D PDOH FLUFXPVWDQFH 7KHUHIRUH WKH VHOHFWLRQ RI IHPDOH VXEMHFWV LV FRQWUDLQGLFDWHG IURP WKH VWDQGSRLQW RI HIILFLHQF\ $JH KDV EHHQ GHPRQVWUDWHG WR H[KLELW D SRVLWLYH UHODWLRQVKLS ZLWK VOHHS DSQHD LQGLFDWLQJ VHOHFWLRQ RI ROGHU VXEn MHFWV 6QRULQJ LV DQRWKHU YDULDEOH ZLWK D FORVH UHODWLRQn VKLS WR VOHHS DSQHD 7KH DSSDUHQWO\ FRPPRQ PHFKDQLVP RI VQRULQJ DQG VOHHS DSQHD VWHQRVLV RI WKH XSSHU DLUZD\f LQGLFDWHV WKDW SRVVLEO\ YHU\ IHZ VXEMHFWV ZLWQRXW VQRULQJ

PAGE 33

VXIIHU VOHHS DSQHDV $ VLJQLILFDQW SURSRUWLRQ RI VQRUHUV PLJKW EH H[SHFWHG WR VXIIHU DSQHDV ZKLOH IHZ QRQVQRUHUV PLJKW KDYH DSQHD 7KH ODVW YDULDEOH ZHLJKW ZKLOH KDYLQJ D SRVLWLYH FRUUHODWLRQ ZLWK DSQHD DFWLYLW\ LQ WKH JHQHUDO SRSXODWLRQ LV RI XQFHUWDLQ SHUWLQHQFH WR 6$6 $GGLn WLRQDOO\ FKDQJHV LQ ZHLJKW PD\ RFFXU ZLWKRXW FKDQJHV LQ DSQHD DFWLYLW\ 7KHUHIRUH WKLV YDULDEOH KDV D OHVV WKDQ H[SOLFLW FRQQHFWLRQ WR DSQHD DFWLYLW\ DQG YLRODWHV WKH VHFRQG SULQFLSOH DGRSWHG WR FRQVLGHU WKHVH YDULDEOHV 7KLV UHYLHZ RI DYDLODEOH GDWD VXJJHVWV WKDW VDPSOLQJ DJLQJ VQRULQJ PDOHV ZRXOG JHQHUDWH D KLJK SURSRUWLRQ RI VOHHS DSQHD DFWLYLW\ ([WHUQDO YDOLGLW\ IRU WKLV LGHD LV GUDZQ IURP LWV VLPLODULW\ WR WKH W\SLFDOnn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f $ GLVFXVVLRQ RI WKH HYLGHQFH IRU WKH SUHVHQFH RI WKHVH GHILFLWV ZLOO EH SUHVHQWHG IROORZHG E\ SURSRVDOV IRU DSSURSULDWH PHDVXUHV RI WKHVH YDULDEOHV

PAGE 34

&DUGLRSXOPRQDU\ 'HILFLWV &DUGLRSXOPRQDU\ FRPSOLFDWLRQV KDYH EHHQ UHSRUWHG DV FRQFRPLWWDQWV RI 6$6 *XLOOHPLQDXOW HW DO f &RFFDJQD HW DO f UHSRUWHG KLJK LQFLGHQFHV RI K\SHUn WHQVLRQ DQG FRQJHVWLYH KHDUW IDLOXUH LQ WKHLU 6$6 VXEn MHFWV 6FKURHGHU HW DO f VWXGLHG 6$6 SDWLHQWV ZLWK D PHDQ DJH RI 6L[ VXEMHFWV KDG ZDNLQJ V\VWHPLF K\SHUWHQVLRQ DQG VHYHUDO KDG ZDNLQJ FDUGLDF DEQRUPDn OLWLHV 'XULQJ VOHHS VWXGLHV SDWLHQWV GHYHORSHG VLJQLILFDQW ULVHV LQ V\VWHPLF DUWHULDO SUHVVXUH ZKLFK F\FOHG ZLWK HSLVRGHV RI DSQHD ZKLOH SDWLHQWV GHYHORSHG SXOPRQDU\ DUWHULDO K\SHUWHQVLRQ 7UDFKHRVWRP\ UHYHUVHG DOO WKHVH DEQRUPDOLWLHV VWURQJO\ LPSOLFDWLQJ WKH DSQHDV LQ WKHLU HWLRORJ\ 7LONLDQ HW DO f VWXGLHG PDOH 6$6 VXEMHFWV ZLWK D PHDQ DJH RI r 7ZHQW\IRXU RI WKH VKRZHG VOHHSLQJ VLQXV DUUK\WKPLDV ZKLOH GHYHORSHG PRUH VHULRXV V\PSWRPV VXFK DV DV\VWROH 6HYHQWHHQ RI WKHVH SDWLHQWV UHFHLYHG WUDFKHRVWRP\ ZKLFK DEROLVKHG DOO RI WKHVH DEQRUPDOLWLHV ([SHULPHQWDO RFFOXVLRQ RI WKH WUDFKHRVWRP\ GXULQJ VOHHS LQ RI WKHVH SDWLHQWV UHVXOWHG LQ D UHWXUQ RI WKH FDUGLRSXOPRQDU\ V\PSWRPV =ZLOOLFK HW DO f VWXGLHG VL[ FRQVHFXWLYH PDOH 6$6 SDWLHQWV ,Q DOO SDWLHQWV EUDG\FDUGLD DFFRPSDQLHG DQ\ DSQHD ZLWK VLJQLILFDQW GHVDWXUDWLRQ $GGLWLRQDOO\ D VLJQLILFDQW FRUUHODWLRQ ZDV QRWHG EHWZHHQ GHJUHH RI GHVDWXUDWLRQ DQG VHYHULW\ RI EUDG\FDUGLD %UDG\FDUGLD ZDV DEROLVKHG RU DWWHQWXDWHG GXULQJ DGPLQLVWUDWLRQ RI 2 HQULFKHG DLU

PAGE 35

=ZLOOLFK HW DO SURSRVHG WKDW EUDG\FDUGLD GXULQJ DSQHD UHVXOWV IURP LQFUHDVHG YDJDO WRQH PHGLDWHG E\ FDURWLG ERG\ FKHPRUHFHSWRUV )XMLWD HW DO f UHSRUWHG RQ PDOH 6$6 SDWLHQWV $OO UHSRUWHGO\ VXIIHUHG FDUGLDF DUUK\WKPLDV GXULQJ VOHHS 7KH ERG\ RI GDWD VXJJHVWV WKDW 6$6 SDWLHQWV IUHTXHQWO\ VXIIHU IURP FDUGLDF DEQRUPDOLWLHV VXFK DV K\SHUWHQVLRQ DUUK\WKPLDV DQG EUDG\FDUGLD $ VHFRQG ERG\ RI GDWD H[LVWV ZKLFK LQGLFDWHV WKDW 6$6 SDWLHQWV GR QRW LQYDULDEO\ KDYH FDUGLRSXOPRQDU\ FRPSOLn FDWLRQV &RYHUGDOH HW DO f VWXGLHG SDWLHQWV RI ZKRP ZHUH GLDJQRVHG ZLWK 6$6 2QO\ RI WKHVH SDWLHQWV KDG HYLGHQFH RI FRUSXOPRQ£OH ZKLOH PRUH KDG V\VWHPLF K\SHUWHQVLRQ 1RQH RI WKHVH SDWLHQWV GHPRQVWUDWHG VHYHUH EUDG\ RU WDFK\FDUGLD GXULQJ VOHHS ,Q 2UU HW DO f SDWLHQWV DOO RI ZKRP ZHUH WHFKQLFDOO\ GLDJQRVDEOH DV KDYLQJ 6$6 DOO V\PSWRPDWLF VRPQROHQWf SDWLHQWV KDG HYLn GHQFH RI ULJKW VLGHG KHDUW IDLOXUH DV ZHOO DV IUHTXHQW DUUK\WKPLDV FXULQJ DSQHDV $V\PSWRPDWLF VXEMHFWV KDG QRQH RI WKHVH V\PSWRPV .UHLVV HW DO f VWXGLHG 9HWHUDQV $GPLQLVWUDWLRQ ZDUG SDWLHQWV DQG LRXQG WR KDYH D 6$6 7KH 6$6 SDWLHQWV KDG VLJQLILFDQWO\ PRUH VLJQV RI FRQJHVWLYH KHDUW IDLOXUH EXW QRW K\SHUWHQVLRQ RU DQJLQD $QFROL,VUDHO HW DO f VDPSOHG HOGHUO\ VXEMHFWV ZLWK FRPSODLQWV RI VOHHSLQJ UHVSLUDWRU\ DLV RUGHUV 1LQH RI WKHVH VXEMHFWV ZHUH GLDJQRVDEOH DV 6$6 EXW WKLV JURXS GLG QRW KDYH VLJQLILFDQWO\ PRUH KHDUW

PAGE 36

GLVHDVH RU K\SHUWHQVLRQ 7KXV D JURXS RI 6$6 SDWLHQWV ZLWKRXW HYLGHQFH RI KHDUW GLVHDVH H[LVWV 7DNHQ WRJHWKHU WKHVH GDWD LQGLFDWH WKDW FDUGLDF FRPSOLFDWLRQV IUHTXHQWO\ EXW QRW LQHYLWDEO\ DFFRPSDQ\ D VOHHS DSQHD V\QGURPH 7KDW D ODUJH QXPEHU RI DSQHDV PD\ EH SUHVHQW ZLWKRXW D VLJQLILFDQW LQFUHDVH LQ FDUGLRn SXOPRQDU\ FRPSOLFDWLRQV LV GHPRQVWUDWHG E\ WKH 2UU HW DO DQG $QFROL,VUDHO HW DO VWXGLHV $V WKH UHODWLRQVKLS EHWZHHQ WKHVH FRPSOLFDWLRQV DQG D IXOO EORZQ 6$6 LV XQFOHDU WKH SRVVLELOLW\ RI FDUGLRSXOPRQDU\ VHTXHODH IURP D VXEFOLQLFDO OHYHO RI DSQHDV LV HYHQ PRUH XQFHUWDLQ +RZHYHU DQ DVVHVVPHQW RI FDUGLRSXOPRQDU\ VWDWXV LV LQGLFDWHG E\ WKH IUHTXHQW DVVRFLDWLRQ RI FDUGLRSXOPRQDU\ FRPSOLFDWLRQV ZLWK VOHHS DSQHD V\QGURPH $URXVDO 'HILFLWV +\SHUVRPQROHQFHf ([FHVVLYH 'D\WLPH 6OHHSLQHVV ('6f RU K\SHUVRPQRn OHQFH KDV EHHQ GHVFULEHG DV WKH VLQJOH PRVW FRPPRQ UHVXOW RI 6$6 'HPHQW HW DO f ,QGHHG SDWLHQWV ZKR PHHW WKH FULWHULRQ RI QXPEHU RI DSQHD HSLVRGHV DQG H[KLELW ('6 DUH GHVFULEHG DV V\PSWRPDWLF ZKLOH WKRVH ZLWK QXPEHU RI DSQHDV DORQH DUH ODEHOHG DV\PSWRPDWLF 2UU HW DO f $V DQ DVLGH WKLV GLVWLQFWLRQ LOOXVWUDWHV DJDLQ D FHQWUDO LVVXH RI 6$6 ,I SDWLHQWV H[KLELW WKH QXPEHU RI DSQHDV TXDOLI\LQJ WKHP IRU 6$6 EXW QR RWKHU SDWKRORJ\ DUH WKH\ VXIIHULQJ IURP D SDWKRORJLFDO SURFHVV" 'HPHQW HW DO f EHOLHYH WKDW WKH\ DUH 7KH\ VXJJHVW WKDW PRVW LI QRW DOO SDWLHQWV ZLWK D FOLQLFDOO\ HOHYDWHG QXPEHU RI

PAGE 37

DSQHDV DFWXDOO\ VXIIHU ('6 EXW WKDW LW LV PDVNHG E\ WZR IDFWRUV 'HPHQW HW DO QRWHG WKDW 6$6 SDWLHQWV PD\ GHQ\ VRPQROHQFH ZKLOH OLWHUDOO\ IDOOLQJ DVOHHS EHIRUH WKH FOLQLFLDQ 7KH\ VSHFXODWH WKDW WKLV UHVXOWV IURP D UHn VSRQVH ELDV DJDLQVW DFNQRZOHGJLQJ LOOQHVV RU D FKDQJH LQ VXEMHFWLYH IUDPH RI UHIHUHQFH DERXW MXVW ZKDW DOHUWQHVV LV 7KXV ('6 PD\ EH GHQLHG E\ SDWLHQWV IRU UHDVRQV ZKLFK PD\ EH RXWVLGH WKH FOLQLFLDQnV FRQWURO RU NQRZOHGJH $OWHUQDWLYHO\ %ORFN HW DO f SURSRVH WKDW DV\PSWRn PDWLF 6$6 SDWLHQWV DUH VXIIHULQJ D VXEFOLQLFDO OHYHO RI IDOORXW IURP WKHLU IUHTXHQW DSQHDV $V \HDUV SDVV LQn FUHDVLQJ ZHLJKW RU FXPXODWLYH HIIHFWV RI GHVDWXUDWLRQ HYHQWXDOO\ OHDG WR D IXOO EORZQ K\SHUVRPQROHQFHf§6OHHS $SQHD 6\QGURPH $ ILQDO SRVVLELOLW\ LV WKDW QRQK\SHU VRPQROHQW VXEMHFWV ZLWK IUHTXHQW DSQHDV H[KLELW D FRPn SOHWHO\ EHQLJQ SURFHVV WRWDOO\ XQUHODWHG WR WKDW IRXQG LQ V\PSWRPDWLF 6$6 SDWLHQWV ,W ZRXOG VHHP WKDW GHOLQHDWLRQ RI WKH OHYHO RI ('6 LQ VXEMHFWV ZLWK VXEFOLQLFDO OHYHOV RI DSQHD PD\ KHOS FODULI\ WKLV LVVXH 7KLV QRWLRQ ZLOO EH HODERUDWHG EHORZ 7KH VRPQROHQFH GHILFLW ZKLFK LV QRWHG LQ IXOO EORZQ 6$6 SDWLHQWV LV FRPPRQO\ UHSRUWHG 'HPHQW HW DO f GHVFULEH LPSDLUPHQWV LQ FRQWLQXRXV SHUIRUPDQFH RI YLUWXDOO\ DQ\ DFWLYLW\ 6OHHS $SQHD 6\QGURPH SDWLHQWV UHSRUWHGO\ IDOO DVOHHS DW RXWGRRU VWDGLXPV LQ IURQW RI FODVVHV DQG ZKLOH WUHDWLQJ SDWLHQWV ([FHVVLYH 'D\WLPH 6RPQROHQFH LV GHVFULEHG LQ HYHU\ 6$6 SDWLHQW UHSRUWHG LQ

PAGE 38

H[SHULPHQWDO SURWRFROV E\ :HLW]PDQ HW DO f )XMLWD HW DO f =ZLOOLFK HW DO f *DUD\ HW DO f DQG 6XOOLYDQ DQG ,VVD f $W WKH VDPH WLPH VXEMHFWV ZQR DUH DV\PSWRPDWLF IRU ('6 EXW ZLWK WKH QHFHVVDU\ QXPEHU RI DSQHDV WR EH GLDJQRVHG DV 6$6 DUH GHVFULEHG E\ 2UU HW DO f DQG 6PLUQH HW DO f DV ZHOO DV RWKHUV +RZHYHU LW VKRXOG EH QRWHG WKDW WKH ODWWHU VWXGLHV UHOLHG RQ JOREDO UHSRUWV RI VRPQROHQFH UDWKHU WKDQ TXDQWLILHG GDWD 5HFDOOLQJ WKH FULWLFLVP RI 'HPHQW HW DO f RI WKLV WHFKQLTXH DV YXOQHUDEOH WR VXEMHFWLYH ELDV WKHVH DV\PSWRPDWLF SDWLHQWV PD\ KDYH XQUHSRUWHG RU XQGHWHFWHG VRPQROHQFH ([FHVVLYH 'D\WLPH 6RPQROHQFH DSSHDUV WR EH IUHTXHQWO\ FRQFRPLWWDQW ZLWK 6OHHS $SQHD 6\QGURPH 7KXV LWV SRVVLEOH RFFXUUHQFH LQ VXEFOLQLFDO SRSXODWLRQV VHHPV ZRUWK\ RI LQYHVWLJDWLRQ 2[\JHQ 'HVDWXUDWLRQ DQG &RJQLWLYH 'HILFLWV 2QH GHILFLW ZKLFK DSSHDUV LQ 6$6 SDWLHQWV LV R[\JHQ GHVDWXUDWLRQ ZKLFK DFFRPSDQLHV DSQHDV 9nKHQ DSQHD EHJLQV DUWHULDO R[\JHQ VDWXUDWLRQ EHJLQV WR IDOO DQG PD\ FRQn WLQXH IDOOLQJ WKURXJKRXW WKH HYHQW $Q HDUO\ WHFKQLTXH XWLOL]HG LQ VWXG\LQJ R[\JHQ VDWXUDWLRQ OHYHOV LQYROYHG SHULRGLF EORRG VDPSOHV GUDZQ IURP DUWHULDO VRXUFHV (LUFKILHOG HW DO f VWXGLHG QRUPDO PDOHV ZLWK D PHDQ DJH RI r 'D\WLPH R[\JHQ VDWXUDWLRQ LQ WKHVH VXEn MHFWV DYHUDJHG 7KHVH YDOXHV IHOO GXULQJ VOHHS WR /DWHU 2UU HW DO f VWXGLHG PDOHV ZKR KDG

PAGE 39

EHHQ PDWFKHG RQ QXPEHU RI DSQHD HYHQWV $V\PSWRPDWLF QRQK\SHUVRPQROHQWf VXEMHFWV IHOO IURP D EDVHOLQH VDWXUDWLRQ RI PP +Jf WR D PHDQ PD[LPXP GHVDWXUDWLRQ RI PP +Jf GXULQJ VOHHS 6\PSWRPDWLF VXEMHFWV IHOO IURP D EDVHOLQH RI PP +Jf WR D PHDQ PD[LPXP GHVDWXUDWLRQ RI PP +Jf ZKLOH DVOHHS 2UU HW DO VXJJHVWHG WKDW WKH PRUH VHYHUH GHVDWXUDWLRQ ZKLFK DFFRPSDQLHG DSQHDV LQ V\PSWRPDWLF VXEMHFWV XQGHUODLG WKHLU VRPQROHQFH DQG RWKHU FRPSOLFDWLRQV +RZHYHU WKHVH FRQFOXVLRQV DUH YXOQHUDEOH WR FHUWDLQ PHWKRGRORJLFDO FULWLFLVPV DV (ORFN f SRLQWV RXW 6SHFLILFDOO\ SHULRGLF VDPSOLQJ PD\ ZHOO IDLO WR GHWHFW WKH PXOWLSOH EULHI GHVDWXUDWLRQV DVVRFLDWHG ZLWK 6$6 UHQGHULQJ WKLV WHFKQLTXH SRWHQWLDOO\ LQVHQVLWLYH WR LPSRUWDQW HYHQWV ,Q DQ DWWHPSW WR SURYLGH PRUH UHSUHVHQWDWLYH GDWD RQ GHVDWXUDWLRQ %ORFN HW DO f XWLOL]HG DQ HDU R[LPHWHU LQ VWXG\LQJ VDWXUDWLRQ LQ QRUPDO VXEMHFWV $Q HDU R[Ln PHWHU SURYLGHG D FRQWLQXRXV DFFXUDWH UHDGRXW RI PRPHQW WR PRPHQW DUWHULDO R[\JHQ OHYHOV 7KLUW\ PDOHV PHDQ DJH f DQG IHPDOHV PHDQ DJH f ZHUH VWXGLHG 2I WKH PDOHV VHYHQWHHQ VXIIHUHG DW OHDVW RQH HSLVRGH RI GHVDWXn UDWLRQ IDOOLQJ IURP D PHDQ VDWXUDWLRQ RI b WR D PD[LPXP GHVDWXUDWLRQ RI b GXULQJ VOHHS ,Q FRQWUDVW QR HSLn VRGHV RI GHVDWXUDWLRQ ZHUH QRWHG LQ WKH VDPSOH RI SUHPHQD SDXVDO IHPDOHV EDVHOLQH VDWXUDWLRQ bf 7KH R[\JHQ GHVDWXUDWLRQV RI WKH PDOHV ZHUH DOZD\V IRXQG LQ DVVRn FLDWLRQ ZLWK EUHDWKLQJ DEQRUPDOLWLHV RU VQRULQJ $ VWXG\

PAGE 40

RI R[\JHQ VDWXUDWLRQ LQ SRVW PHQDSDXVDO IHPDOHV ZDV UHSRUWHG E\ %ORFN HW DO f (OHYHQ RI WKHVH ZRPHQ H[KLELWHG GHVDWXUDWLRQ ZLWK VHYHUDO VXEMHFWV GHVDWXUDWLQJ WR OHVV WKDQ 'ROO\ DQG %ORFN f VWXGLHG D JURXS RI PDOHV DQG IHPDOHV ZLWK D PHDQ DJH RI \HDUV 7KHVH QRUPDO VXEMHFWV GURSSHG IURP D PHDQ EDVHOLQH VDWXUDn WLRQ RI rb WR D PHDQ PD[LPXP GHVDWXUDWLRQ RI 7KLV HYLGHQFH LQGLFDWHV WKDW GHVD[XUDWLRQV RI "n DUH FRPPRQ LQ QRUPDO PDOHV DQG SRVW PHQDSDXVDO IHPDOHV ZKLOH SUHPHQDSDXVD IHPDOHV UDUHO\ LI HYHU GHVDWXUDWH 2[\JHQ VDWXUDWLRQ OHYHOV LQ 6$6 SDWLHQWV DUH QRW FRPPRQO\ UHSRUWHG 8VLQJ DQ HDU R[LPHWHU *DUD\ HW DO f VWXGLHG 6$6 SDWLHQWV ZKR ZHUH GLYLGHG LQWR GD\WLPH HXFDSQLFV DQG GD\WLPH K\SHUFDSQLFV (XFDSQLFV IHOO IURP D EDVHOLQH VDWXUDWLRQ RI b WR D PHDQ PD[LPXP GHVDWXUDWLRQ RI rr ZKLOH K\SHUFDSQLFV IHOO IURP D EDVHOLQH VDWXUDWLRQ RI 4b WR D PD[LPXP GHVDWXUDWLRQ RI FDOFXODWHG IURP WDEOHG GDWDf 7KHVH GDWH ZHUH FROOHFWHG GXULQJ GD\WLPH QDS VWXGLHV KRZHYHU DQG KHQFH LV VRPHZKDW TXHVWLRQDEOH %HUU\ DQG (ORFN f UHFRUGHG KHDY\ VQRULQJ PDOHV ZKR VXIIHUHG PDQ\ DSQHDF HYHQWV )RXU VXEMHFWV ZHUH VRPQROHQW ZKLOH ZHUH QRW 6\PSWRPDn WLF VXEMHFWV IHOO IURP D EDVHOLQH RI WR D PHDQ PD[LPXP GHVDWXUDWLRQ RI ZKLOH DV\PSWRPDWLF VXEMHFWV GHVDWX UDWHG IURP D EDVHOLQH RI WR D PHDQ PD[LPXP GHVDWXUDn WLRQ RI

PAGE 41

'DWD RQ GHVDWXUDWLRQ DUH LQWHUHVWLQJ IURP VHYHUDO SHUVSHFWLYHV )LUVW RI DOO 6OHHS $SQHD 6\QGURPH VXEMHFWV VHHP WR GHVDWXUDWH PRUH KHDYLO\ WKDQ QRUPDOV ZLWK %HUU\ DQG %ORFNnV V\PSWRPDWLF JURXS GHVDWXUDWLQJ XS WR b DQG WKH VXEMHFWV RI *DUD\ HW DO GHVDWXUDWLQJ DERXW b FRPSDUHG ZLWK QRUPDO GHVDWXUDWLRQ OHYHOV RI b 6HFRQGO\ DPRXQW RI GHVDWXUDWLRQ GLVWLQJXLVKHG V\PSWRPDWLF K\SHUVRPQROHQWf IURP DV\PSWRPDWLF VXEMHFWV LQ WKH GDWD RI ERWK 2UU HW DO DQG %HUU\ DQG %ORFN $GGLWLRQDOO\ WZR SRLQWV UHOHYDQW WR PHWKRGRORJ\ HPHUJH 0HDVXUHV RI R[\JHQ GHVDWXUDWLRQ VHHP FUXFLDO WR XQGHUVWDQGLQJ VOHHS DSQHD DQG LWV HIIHFWV DQG HDU R[LPHWU\ DSSHDUV WR SURYLGH D SURYHQ YDOLG PHDVXUH RI WKLV YDULDEOH :HLW]PDQ f KDV VXJJHVWHG WKDW WKH SDWKRORJLFDO SURFHVV LQ 6$6 LV WKH K\SR[LD RI GHVDWXUDWLRQ LQ DSQHDF HSLVRGHV :KLOH FRJQLWLYH DQG LQWHOOHFWXDO FKDQJHV LQ 6$6 KDYH EHHQ RQO\ DQHFGRWDOO\ UHSRUWHG *XLOOHPLQDXOW HW DO f K\SR[LD GRHV VHHP D SODXVLEOH SRWHQWLDO FDXVH RI WKHVH FKDQJHV 7KXV R[\JHQ GHVDWXUDWLRQ DSSHDUV WR EH D UHOLDEOH FRQVHTXHQFH RI VOHHS DSQHD V\QGURPH ZKLOH FRJQLn WLYHLQWHOOHFWXDO FKDQJHV DUH SXWDWLYH VHTXHODH ,Q VXPPDU\ D UHYLHZ RI GHILFLWV IRXQG LQ VOHHS DSQHD V\QGURPH VXEMHFWV VXJJHVWV WKDW FDUGLRSXOPRQDU\ FRPSOLFDn WLRQV H[FHVVLYH GD\WLPH VOHHSLQHVV R[\JHQ FHVDWXUDWLRQ DQG FRJQLWLYHLQWHOOHFWXDO FKDQJHV DUH IUHTXHQWO\ DVVRn FLDWHG ZLWK WKLV V\QGURPH $Q DVVHVVPHQW RI WKHVH YDULDEOHV LQ D VXEFOLQLFDO SRSXODWLRQ VHHPV LQGLFDWHG E\

PAGE 42

WKHVH GDWD %HORZ D GLVFXVVLRQ RI PHDVXUHPHQW LVVXHV DQG SURSRVDO DVVHVVPHQW RI WKHVH YDULDEOHV DSSHDUV 0HDVXUHPHQW RI 'HILFLWV )RXQG LQ 6OHHS $SQHD 6\QGURPH $ERYH LW ZDV QRWHG WKDW FDUGLRSXOPRQDU\ FRPSOLFDn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n MHFWV $GGLWLRQDOO\ D JHQHUDO VXUYH\ RI KHDOWK VWDWXV VXFK DV WKDW SURYLGHG E\ WKH &RUQHOO 0HGLFDO ,QFH[ &0, %URDGPDQ HW DO f ZRXOG DOVR VFUHHQ IRU RWKHU KHDOWK GHILFLWV 7KH &0, SURYLGHV DQ RYHUDOO VFRUH LQDLFDWLQJ QXPEHU RI V\PSWRPV HQGRUVHG DV ZHOO DV VXEVFDOHV RQ VHYHUDO V\PSWRP FDWHJRULHV 6HSDUDWH H[DPLQDWLRQ RI VXEn VFDOHV RQ UHVSLUDWRU\ FDUGLRSXOPRQDU\ DQG QHXURORJLFDO VXEVFDOHV VHHPV DSSURSULDWH $VVHVVPHQW RI WKHVH YDULDEOHV SURYLGHV D EURDG KHDOWK VFUHHQLQJ DV ZHOO DV D GHWDLOHG DQDO\VLV RI V\PSWRPV IRXQG LQ VOHHS DSQHD V\Qn GURPH

PAGE 43

0HDVXUHPHQW RI $URXVDO 'HILFLWV +\SHUVRPQROHQFHf ([FHVVLYH GD\WLPH VOHHSLQHVV LV D FRPPRQ VHTXHOD RI VOHHS DSQHD V\QGURPH 0HDVXUHPHQWV RI WKH VOHHSZDNH F\FOH RI VXEFOLQLFDO DSQHD SDWLHQWV WKXV VHHPV LQGLn FDWHG $ VOHHS TXHVWLRQQDLUH DVVHVVLQJ WUDLW VOHHS KDELWV VHHPV LQ RUGHU DV ZHOO DV ZHHN VOHHS ORJV IROORZLQJ WKH H[SHULPHQWDO QLJKW 7KH VOHHS TXHVWLRQQDLUH DQG VOHHS ORJV KDYH EHHQ XVHG LQ RQJRLQJ UHVHDUFK LQ :( :HEEnV VWXGLHV RI DJLQJ DQG VOHHS DQG VHHP DQ DGHTXDWH DVVHVVPHQW RI VOHHS SDWWHUQV RI VXEFOLQLFDO DSQHD VXEMHFWV $GGLn WLRQDOO\ HOHFWURHQFHSKDORJUDSKLF UHFRUGLQJV RI VOHHS GXULQJ WKH H[SHULPHQWDO QLJKW VHHP LQ RUGHU XVLQJ D VWDQn GDUG UHFRUGLQJ DQG VFRULQJ V\VWHP $JQDZ DQG :HEE f 7KH DVVHVVPHQW RI GD\WLPH VOHHSLQHVV LV UHODWLYHO\ QHZ RQO\ UHFHQWO\ EHLQJ WKH VXEMHFW RI LQYHVWLJDWLRQ :LOOLDP 'HPHQW DQG DVVRFLDWHV EHJDQ WR GHYHORS PHWKRGV RI TXDQWLI\LQJ VRPQROHQFH LQ WKH HDUO\ V 7KH ILUVW DWWHPSW SURGXFHG D VXEMHFWLYH UDWLQJ VFDOH WKH 6WDQIRUG 6OHHSLQHVV 6FDOH 666 +RGGHV HW DO fr 7KH 666 ZDV D SRLQW /LNHUW VFDOH ZLWK HQGSRLQWV DQFKRUHG RQ f IHHO DFWLYH DQG YLWDO DQG f DOPRVW LQ UHYHULH VOHHS RQVHW VRRQ ZLWK SRLQWV ZLWK VWDWHPHQWV UHIOHFWLQJ YDULRXV GHJUHVV RI VRPQROHQFH LQ EHWZHHQ 3DWLHQWV UDWHG WKHLU LQWURVSHFWLYH OHYHO RI VOHHSLQHVV RQ WKLV VFDOH 8QIRUWXQDWHO\ WKH 666 SURYHG WR EH YXOQHUDEOH WR H[DFWO\ WKH VDPH GLIILFXOWLHV DV D JOREDO UDWLQJ VFDOH XVHG RQ 6$6

PAGE 44

SDWLHQWV LH UHVSRQVH ELDV DQG DOWHUHG IUDPH RI UHIHUHQFH 'HPHQW HW DO f ,W ZDV FOHDU WKHQ WKDW DQ REMHFWLYH PHDVXUH RI GD\n WLPH VRPQROHQFH ZDV QHHGHG :LWK WKLV JRDO LQ PLQG &DUVNDGRQ DQG 'HPHQW f LQWURGXFHG WKH 0XOWLSOH 6OHHS /DWHQF\ 7HVW 06/7f 7KLV WHFKQLTXH LQYROYHG PHDVXULQJ WKH REMHFWLYH VOHHS ODWHQF\ GXULQJ PXOWLSOH QDS DWWHPSWV WKURXJKRXW WKH GD\ ,Q D W\SLFDO SDUDGLJP D VXEMHFW ZRXOG EH ZLUHG WR DQ ((* DQG DWWHPSW WR IDOO DVOHHS DW KRXU LQWHUYDOV WKURXJK WKH GD\ 7KH ODWHQF\ WR VOHHS LQ HDFK QDS ZDV PHDVXUHG E\ GHWHUPLQLQJ WKH ODWHQF\ WR WKH ILUVW HSRFK RI 6WDJH VOHHS ,Q RUGHU WR SUHYHQW VLJQLn ILFDQW DPRXQWV RI VOHHS DFFXPXODWLQJ GXULQJ WHVWLQJ VXEn MHFWV ZHUH DZDNHQHG DIWHU DQ HSRFK RI VOHHS ZDV REVHUYHG ,I QR VOHHS ZDV QRWHG VXEMHFWV ZHUH GLVFRQn QHFWHG DIWHU PLQXWHV DQG DOORZHG WR UHVXPH HWKHU DFWLn YLWLHV &DUVNDGRQ DQG 'HPHQW EHOLHYH WKDW WKLV PHDVXUHV XQGHUO\LQJ SK\VLRORJLFDO VOHHS WHQGHQF\ 7KLV XQGHUO\LQJ WHQGHQF\ LV WKRXJKW WR EH PRGXODWHG E\ DOHUWLQJ VWLPXOL LQ WKH HQYLURQPHQW &DUVNDGRQ DQG 'HPHQW Df 7KH YDOLn GLW\ RI WKH WHVW ZDV HVWDEOLVKHG RQ QRUPDO VXEMHFWV ZKR H[KLELWHG D ELSKDVLF FXUYH RI VOHHS WHQGHQF\ WKURXJKRXW D KRXU SHULRG ([WHQVLRQ RI VOHHS RYHU W\SLFDO OHQJWKV LQ QRUPDOV ZKR DUH WKRXJKW WR EH PRGHUDWHO\ FKURQLFDOO\ VOHHS GHSULYHGf VDZ D VLJQLILFDQW ULVH LQ DYHUDJH VOHHS ODWHQFLHV RQ WKH 06/7 &DUVNDGRQ DQG 'HPHQW Ef ZKLOH FKURQLF VOHHS UHVWULFWLRQ KUV VOHHS SHU QLJKWf FDXVHG

PAGE 45

VLJQLILFDQW GHFOLQHV LQ DYHUDJH VOHHS ODWHQFLHV &DUVNDGRQ DQG 'HPHQW Df )LQDOO\ WRWDO VOHHS GHSULYDWLRQ OHG WR D GUDVWLF GHFOLQH LQ 06/7 YDOXHV ZLWK DYHUDJHV SOXPPHWLQJ EHORZ RQH PLQXWH &DUVNDGRQ DQG 'HPHQW Df 0HDQ 06/7 YDOXHV VKRZHG UHFRYHU\ WR EDVHOLQH YDOXHV DIWHU D IXOO QLJKWnV VOHHS IRU SDUWLDOO\ KUVf GHSULYHG VXEMHFWV DQG UHFRYHU\ DIWHU WZR QLJKWV RI DG OLE VOHHS IRU WRWDOO\ VOHHS GHSULYHG VXEMHFWV &DUVNDGRQ DQG 'HPHQW Df 7KLV ERG\ RI HYLGHQFH VXJJHVWV WKDW WKH 06/7 LV D UHOLDEOH DQG YDOLG LQGH[ RI VOHHSLQHVV LQ QRUPDO VXEMHFWV :LWK LWV LQWHUQDO YDOLGLW\ HVWDEOLVKHG WKH 06/7 ZDV DSSOLHG WR VRPQROHQW SRSXODWLRQV 'HPHQW HW DO f DV ZHOO DV 5LFKDUGVRQ HW DO f GHPRQVWUDWHG WKDW 6$6 DQG QDUFROHSWLF SDWLHQWV KDYH VLJQLILFDQWO\ VKRUWHU ODWHQFLHV WR VOHHS RQ WKH 06/7 WKDQ QRUPDO VXEMHFWV GR +DUWVH HW DO f IRXQG 6$6 VXEMHFWV WR KDYH DQ DYHUDJH ODWHQF\ WR 6WDJH RI PLQXWHV ZKLOH QDUFROHSWLFV KDG D PHDQ RI PLQXWHV %RWK WKHVH ODWHQFLHV ZHUH VLJQLILFDQWO\ ORZHU WKDQ WKRVH RI D PLVFHOODQHRXV FRQWURO JURXS ZKRVH ODWHQFLHV ZHUH DERXW PLQXWHV +DUWVH HW DO f IRXQG ODWHQFLHV WR 6WDJH VOHHS DV IROORZV IRU YDULRXV SDWLHQW JURXSV 1RUPDOV P ,QVRPQLDFV P 6$6 P DQG 1DUFROHSWLFV P 7KH 6$6 DQG QDUFROHSWLF SDWLHQWV KDG VLJQLILFDQWO\ VKRUWHU ODWHQFLHV WR 6WDJH VOHHS WKDQ LQVRPQLDFV RU QRUPDOV 5RWK HW DO f FRPSDUHG 6$6 SDWLHQWV ZLWK DJH PDWFKHG FRQWUROV

PAGE 46

7KH QRUPDOV DYHUDJHG P WR 6WDJH VOHHS ZKLOH WKH 6$6 SDWLHQWV DYHUDJHG P =RUULFN HW DO f IRXQG PHDQ 06/7 ODWHQFLHV WR 6WDJH VOHHS RI P IRU 6$6 VXEMHFWV DQG P IRU QDUFROHSWLFV VLJQLILFDQWO\ VKRUWHU WKDQ IRU SDWLHQWV ZLWK SV\FKLDWULF GLVRUGHUV 7KH FKLHI GLIILFXOW\ UHPDLQLQJ LQ -XGJLQJ WKH 06/7 UHOLDEOH DQG YDOLGf LV H[WHUQDO YDOLGDWLRQ GHPRQVWUDWLQJ WKDW H[FHHGLQJ D FHUWDLQ PHDQ ODWHQF\ LV FRUUHODWHG ZLWK RWKHU GHILFLWV RI SHUIRUPDQFH ,Q VRPH ZD\V ORZ ODWHQn FLHV RQ WKH 06/7 PD\ VXIIHU WKH VDPH OLPLWDWLRQV DV VOHHS GHSULYDWLRQ SURWRFROV WKDW LV GHPRQVWUDWLQJ VHULRXV GHILFLWV LV VXUSULVLQJO\ GLIILFXOW $W WKH VDPH WLPH IHZ ZRXOG GHQ\ WKH VXEMHFWLYH XQSOHDVDQWQHVV RI VOHHS GHSULYDn WLRQ ZKLFK PD\ EH DSSOLFDEOH WR ORZ 06/7 VFRUHV 7KXV WKH 06/7 DSSHDUV LQGLFDWHG LQ DVVHVVLQJ H[FHVVLYH GD\WLPH VOHHSLQHVV ZKLOH WKH 6WDQIRUG 6OHHSLQHVV 6FDOH V66f SURn YLGHV DQ LQGH[ RI WKH VXEMHFWLYH H[SHULHQFH RI WKH VXEn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

PAGE 47

KXPDQ YV DQLPDO VXEMHFWV DQG DFXWH YV FKURQLF K\SR[LD )LUVW WKH GDWD DYDLODEOH RQ DQLPDOV ZLOO EH UHYLHZHG IROORZHG E\ WKDW H[WDQW RQ KXPDQ VXEMHFWV )LQDOO\ EHKDYLRUDO YDULDEOHV VHQVLWLYH WR FKDQJHV LQ WKHVH YDULDEOHV ZLOO EH VXPPDUL]HG 7KH HIIHFWV RI FKURQLF K\SR[LD RQ DQLPDOV KDYH QRW EHHQ H[KDXVWLYHO\ VWXGLHG 'DYLV f 2QH SURMHFW ZKLFK GLG H[DPLQH WKLV YDULDEOH LQYROYHG H[SRVXUH RI UDWV WR XS WR KRXUV RI rb R[\JHQ 'DYLV f REVHUYHG WKH EHKDYLRU RI WKHVH DQLPDOV DQG FRQFOXGHG WKDW WKH\ DSSHDUHG QRUPDO $OWKRXJK K\SHUYHQWLODWLRQ ZDV QRWHG QR SHUPDQHQW HIIHFWV RI WKH K\SR[LD ZHUH IRXQG +DQEDXHU HW DO f H[DPLQHG WKH HIIHFWV RI " 2 RQ UDWV H[SRVHG IRU XS WR ZHHNV 7KH\ IRXQG WKDW UHODWLYHO\ VKRUW WHUP GD\Vf K\SR[LD UHVXOWHG LQ DQ LQFUHDVH LQ GRSDPLQH '$f FRQWHQW LQ WKH FDURWLG ERG\ $Q LQFUHDVH LQ QRUHSLQHn SKULQH :(f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f UHYLHZHG WKH

PAGE 48

HYLGHQFH ZKLFK KDV EHHQ FROOHFWHG RQ QHXURWUDQVPLWWHU DOWHUDWLRQV LQ K\SR[LD ,Q EULHI PLOG WR PRGHUDWH K\SR[LD FHUHEUDO OHYHOV RI $73 DUH QRUPDO LPSO\LQJ WKDW HQHUJ\ VXSSOLHV RI QHXURQV DUH QRW GLVWXUEHG ,Q FRQn WUDVW WKH WXUQRYHU RI FHUWDLQ QHXURWUDQVPLWWHUV LV DOWHUHG *LEVRQ HW DO QRWH WKDW R[\JHQ LV DQ LPSRUWDQW VXEVWUDWH IRU WKH V\QWKHVLV RI VHYHUDO QHXURWUDQVn PLWWHUV 9KHQ UDWV DUH H[SRVHG WR DFXWH K\SR[LD V\Qn WKHVLV RI GRSDPLQH DQG VHURWRQLQ LV GHFUHDVHG ZKLOH DEVROXWH OHYHOV RI WU\RVLQH WU\SWRSKDQ FDWHFKRODPLQHV DQG VHURWRQLQ UHPDLQ FRQVWDQW &RQWLQXLQJ K\SR[LD OHDGV WR DQ DGDSWDWLRQ UHVXOWLQJ LQ QRUPDO WXUQRYHU RI WKHVH QHXURWUDQVPLWWHUV *LEVRQ DQG %ODVV f H[DPLQHG WKH DIIHFW RI DFXWH K\SR[LD RQ FHQWUDO DFHW\OFKROLQH $&Kf $V LQ GRSDPLQH '$f DQG VHURWRQLQ +7f DFXWH K\SR[LD UHVXOWV LQ D GHFUHDVH LQ V\QWKHVLV LQ $&K EXW QRW LQ WKH DEVROXWH OHYHO RI $&K 7KXV WKHUH LV D VXEVWDQWLDO ERG\ RI HYLGHQFH LQGLFDWLQJ FKDQJHV LQ FHQWUDO QHXURWUDQVPLWWHU OHYHOV LQ K\SR[LD *HQHUDOL]LQJ IURP WKH DQLPDO GDWD WR KXPDQ VXEMHFWV *LEVRQ DQG 'XII\ f VXJJHVW WKDW WKHVH FKDQJHV XQGHUOLH WKH EHKDYLRUDO GHILFLWV RI K\SR[LD LQ KXPDQV ZKLFK ZLOO EH UHYLHZHG EHORZf 2WKHU YDULDEOHV KDYH EHHQ H[DPLQHG LQ DFXWH K\SR[LD LQ DQLPDOV $QQDX f IRXQG WKDW UDWV H[SRVHG WR EULHI K\SR[LD H[KLELWHG GHFUHDVHG DSSHWLWH WKLUVW DQG VHOI VWLPXODWLRQ UDWHV $QQDX REVHUYHG WKDW EULHI K\SR[LD GHSUHVVHG DOO EHKDYLRUDO YDULDEOHV *HOOKFUQ f IRXQG

PAGE 49

WKDW EULHI K\SR[LD RI rb 2 DEROLVKHG FRUWLFDO UHn VSRQVHV WR DXGLWRU\ VWLPXODWLRQ 6DUD f WUDLQHG UDWV LQ DFWLYH RQH ZD\ DYRLGDQFH WDVNV ,PPHGLDWHO\ DIWHU WUDLQLQJ VXEMHFWV ZHUH H[SRVHG WR R[\JHQ 5DWV WHVWHG DW DQG KRXUV SRVW K\SR[LD DYRLGHG FRUUHFWO\ ZKLOH WKRVH WHVWHG DW DQG KRXUV GLG QRW OHDGLQJ 6DUD WR VXJJHVW WKDW K\SR[LD UHVXOWHG LQ D PHPRU\ UHWULHYDO GHILFLW )UHLGHV DQG $OOZHLVH f DOVR WUDLQHG UDWV LQ RQH ZD\ DYRLGDQFH DQG H[SRVHG WKHP WR b K\SR[LD LPPHn GLDWHO\ IROORZLQJ WUDLQLQJ 7KHVH DQLPDOV DYRLGHG DW DQG PLQXWHV SRVW K\SR[LD EXW QRW DW PLQXWHV )ULHGHV DQG $OOZHLVH VXJJHVWHG WKDW D VHQVLWLYH SHULRG IRU K\SR[LD HIIHFWV RQ WUDLQLQJ H[LVWHG IRU PLQXWHV DIWHU OHDUQLQJ 7DXEHU DQG $OOZHLVH f IROORZHG WKH VDPH WUDLQLQJK\SR[LD SDUDGLJP DQG IRXQG DYRLGDQFH LPSDLUHG DW EXW QRW KRXUV 7KH\ VXJJHVW WKDW K\SR[LD LQWHUn IHUHV ZLWK D PHGLXP WHUP PHPRU\ :KLOH VRPH RI WKHVH GDWD DUH FRQIOLFWLQJ LW DSSHDUV WKDW DFXWH K\SR[LD GHSUHVVHV FHUWDLQ EHKDYLRUDO YDULDEOHV DQG LQWHUIHUHV ZLWK DVSHFWV RI PHPRU\ SURFHVVHV :KLOH FDXWLRQ LV LQ RUGHU ZKHQ JHQHUDOL]LQJ IURP DQLPDOV WR KXPDQV VHYHUDO VHHPLQJO\ UHOHYDQW SRLQWV HPHUJH IURP WKH UHYLHZ %HKDYLRUDOO\ VLPSOH DFWLYLWLHV PD\ EH VSDUHG DOWKRXJK VRPH GHSUHVVLRQ RI DSSHWLWLYH DFWLYLWLHV PD\ RFFXU ZKLOH PRUH FRPSOH[ EHKDYLRU GHSHQn GHQW RQ PHPRU\ SURFHVVHV PD\ EH DOWHUHG 2I SDUWLFXODU LQWHUHVW DUH WKH GHPRQVWUDWHG DOWHUDWLRQV LQ

PAGE 50

QHXURWUDQVPLWWHU WXUQRYHU VHHQ LQ K\SR[LD ,W KDV EHHQ VXJJHVWHG WKDW WKHVH VXEWOH DOWHUDWLRQV PD\ XQGHUOLH WKH FKDQJHV LQ FRPSOH[ EHKDYLRU VHHQ LQ K\SR[LF KXPDQV 'DWD RQ KXPDQ SHUIRUPDQFH DIWHU K\SR[LD KDYH QHFHVn VDULO\ EHHQ UHOLDQW RQ IRUWXLWRXV RU ZHDN PDQLSXODWLRQV $OWKRXJK WKLV PD\ OLPLW WKH XVHIXOQHVV RI WKHVH GDWD LW LV DW WKH YHU\ OHDVW KLJKO\ VXJJHVWLYH RI YDULDEOHV VHQVLn WLYH WR K\SR[LD 9DQ /LHUH DQG 6WLFNQH\ f QRWH WKDW DFXWH K\SR[LD OHDGV WR FRQIXVLRQ KHDGDFKH GURZVLQHVV ZHDNQHVV DQG LQFRRUGLQDWLRQ $IWHU HIIHFWV LQFOXGH KHDGn DFKH QDXVHD DQG HPRWLRQDO ODELOLW\ 5LFKDUGVRQ H[ DO f IRXQG WKDW DFXWH IDWDO K\SR[LD IURP VXUJHU\ UHVXOWV LQ GDPDJH WR FRUWLFDO QHXURQV SDUWLFXODUO\ LQ OD\HUV ,,, DQG ,9 6XVFHSWLEOH VXEFRUWLFDO QHXURQV LQn FOXGH WKRVH RI WKH FRUSXV VWULDWXP DQG FHUHEHOOXP 3OXP HW DO f GHVFULEH D V\QGURPH RI GHOD\HG SRVWDQR[LF HQFHSKDORSDWK\ $IWHU WKH DFXWH FRPD UHVROYHV LQ GD\V D EULHI SHULRG RI QRUPDO IXQFWLRQLQJ UHWXUQV ,QFUHDVLQJ LUULWDELOLW\ DQG FRQIXVLRQ WKHQ HPHUJH ZLWK ORVV RI FRRUGLQDWLRQ DQG PHPRU\ DQG GLPLQLVKHG DWWHQWLRQ VSDQ 7\SLFDOO\ WKH OHVLRQ VSDUHV JUH\ PDWWHU ZKLOH DIIHFWLQJ ZKLWH PDWWHU 'HYHUHDX[ DQG 3DUWQRZ f GHVFULEH D SDWLHQW ZKR UHFRYHUHG IURP D GHOD\HG DFXWH HQFHSKDORSDWK\ $OWKRXJK ,4 ZDV UHODWLYHO\ VSDUHG VHYHUH G\VDUWKULD HPHUJHG 5HOHYDQW SRLQWV ZKLFK HPHUJH IURP VWXGLHV RI DFXWH K\SR[LD LQ KXPDQV VXJJHVW WKDW VXEOHWKDO

PAGE 51

OHYHOV RI K\SR[LD DIIHFW FRRUGLQDWLRQ PHPRU\ DWWHQWLRQ VSDQ DQG PD\ UHVXOW LQ KHDGDFKH GURZVLQHVV DQG QDXVHD 7KH HIIHFWV RI PLOGHU FKURQLF K\SR[LD DUH VXPPDUL]HG E\ *LEVRQ HW DO f} $W WKH R[\JHQ SUHVVXUH HTXLYDn OHQW RI r IHHW LPSDLUHG GDUN DGDSWDWLRQ LV QRWHG $W IHHW GHFUHDVHG FRQFHQWUDWLRQ K\SHUYHQWLODWLRQ DQG VKRUW WHUP PHPRU\ GHILFLWV DUH QRWHG %\ r IHHW HXSKRULD DQG ORVV RI ERWK FRRUGLQDWLRQ DQG FULWLFDO MXGJHn PHQW DUH REVHUYHG &KULVWHQVRQ HW DO f IRXQG VLJQLn ILFDQWO\ OHVV YLVXDO VLJQDOV GHWHFWHG DW b 2 WKDQ DW b VXJJHVWLQJ UHGXFHG YLJLODQFH 0F)DUODQG f VWXGLHG PHQ LQ DVFHQWV WR KLJK DOWLWXGHV DQG FRQFOXGHG WKDW VXEMHFWV ZLWK VORZHU UDWHV RI DVFHQW SHUIRUPHG EHWWHU RQ XQVSHFLILHG SV\FKRORJLFDO WHVWV :HVW f UHSRUWV GDWD FROOHFWHG RQ DQ H[SHGLWLRQ WR 0W (YHUHVW 5HVXOWV RI D QHXURSV\FKRORJLFDO EDWWHU\ LQGLFDWHG GHILFLWV LQ ILQJHU WDSSLQJ YHUEDO IOXHQF\ YHUEDO OHDUQLQJ VKRUW WHUP PHPRU\ DQG H[SUHVVLYH ODQJXDJH 7KXV WKH JHQHUDO DIIHFWV RI FKURQLF PLOG K\SR[LD DUH WKRXJKW WR EH GHILFLWV LQ GDUN DGDSWDWLRQ VKRUW WHUP PHPRU\ FULWLFDO MXGJHPHQW PRWRU FRRUGLQDWLRQ YHUEDO IOXHQF\ DQG YLJLODQFH 2QH JURXS RI FKURQLFDOO\ K\SR[LD SDWLHQWV WKRVH ZLWK FKURQLF REVWUXFWLYH OXQJ DLVHDVH &2/'f KDV EHHQ UHODn WLYHO\ ZHOO VWXGLHG ZLWK VHQVLWLYH QHXURSV\FKRORJLFDO LQVWUXPHQWV DQG LV DHVHUYLQJ RI D PRUH SDLQVWDNLQJ UHYLHZ .URS HW DO f VWXGLHG &2/' SDWLHQWV EHIRUH DQG DIWHU WKH\ UHFHLYHG R[\JHQ VXSSOHPHQWV 7KH :$,6

PAGE 52

:HFKVOHU 0HPRU\ 6FDOH :06f %HQGHU*HVWDOW %HQGHU *HVWDOW ,QWHUIHUHQFH 3URFHGXUH )DFLDO 5HFRJQLWLRQ DQG )LQJHU 7DSSLQJ WHVWV ZHUH DGPLQLVWHUHG WR &2/' DQG FRQWURO VXEMHFWV 6XEMHFWV UHFHLYLQJ VXSSOHPHQWDWLRQ LPSURYHG VLJQLILFDQWO\ PRUH WKDQ FRQWUROV RQ :$,6 )XOO 6FDOH ,4 3HUIRUPDQFH ,4 :HFKVOHU 0HPRU\ 4XRWLHQW ERWK %HQGHU SURFHGXUHV DQG )LQJHU 7DSSLQJ *UDQW HW DO f VWXGLHG DQRWKHU JURXS RI &2/' SDWLHQWV 7KH\ DGPLQLVWHUHG WKH +DOVWHDG5HLWDQ %DWWHU\ $SKDVLD 6FUHHQLQJ 7HVW 7UDLOPDNLQJ 7HVW 5XVVHOO PRGLILFDWLRQ RI WKH :HFKVOHU 0HPRU\ 6FDOH 7DFWLOH 0HPRU\ WHVW DQG WKH :$,6 ,Q &2/' VXEMHFWV GHILFLWV ZHUH IRXQG UHODWLYH WR FRQWUROV RQ DOO YDULDEOHV H[FHSW 5HLWDQ 5K\WKP $SKDVLD 6FUHHQLQJ DQG :06 /RJLFDO VWRULHV *UDQW HW DO FRQFOXGHG WKDW &2/' SDWLHQWV H[KLELWHG GHILFLWV RQ D JOREDO LPSDLUPHQW UDWLQJ DWWHQWLRQ DEVWUDFWLQJ DELOLW\ FRPSOH[ SHUFHSWXDO PRWRU WDVNV VLPSOH VHQVRU\ DQG PRWRU WDVNV DQG PHPRU\ WDVNV /RZ EXW VLJQLILFDQW QHJDWLYH FRUUHODWLRQV ZHUH UHSRUWHG EHWZHHQ VHYHUDO PHDVXUHV RI R[\JHQ VDWXUDWLRQ DQG ERWK +DOVWHDG ,PSDLUPHQW ,QGH[ DQD *OREDO ,PSDLUPHQW 5DWLQJ *UDQW HW DO VSHFXODWHG WKDW IORRU HIIHFWV UHVXOWHG LQ ORZ YDULDELOLW\ LQ 3$ OHYHOV DPRQJ &2/' SDWLHQWV DQG NHSW WKHVH FRUUHODWLRQV ORZ 7KH\ IXUWKHU VXJJHVWHG WKDW WKH QHXURSV\FKRORJLFDO GHILFLWV REVHUYHG LQ WKHVH SDWLHQWV UHVXOWHG IURP WKH R[\JHQ ZDQW IRXQG LQ WKHVH VXEMHFWV 7KXV GDWD DYDLODEOH RQ KXPDQV LQGLFDWH WKDW GDUN YLVLRQ PHPRU\ PRWRU FRRUGLQDWLRQ FULWLFDO MXGJHPHQW

PAGE 53

YHUEDO IOXHQF\ DQG VLJQDO GHWHFWLRQ DUH DIIHFWHG E\ DFXWH K\SR[LD $QLPDO GDWD VXJJHVW WKDW VHQVLWLYLW\ RI PHPRU\ SURFHVVHV WR K\SR[LD ZKLOH GDWD IURP FKURQLF K\SR[LD KXPDQV LQGLFDWH WKDW DWWHQWLRQ YLJLODQFH DEVWUDFWLQJ DELOLW\ FRPSOH[ SHUFHSWXDOn PRWRU VNLOOV DV ZHOO DV PRWRU DQG VHQVRU\ DELOLWLHV DUH DOVR LPSDLUHG E\ K\SR[LD :KLOH QRQH RI WKHVH ILQGLQJV DUH XQHTXLYRFDOO\ GLUHFWO\ DSSOLFDEOH WR WKH EULHI DFXWH GHVDWXUDWLFQ RFFXUULQJ GXULQJ DSQHD WKH\ DUH DW WKH YHU\ OHDVW LQGLFDWLYH RI DELOLWLHV VHQVLWLYH WR K\SR[LD LQ KXPDQV $ QHXURSV\FKRORJLFDO EDWWHU\ VHQVLWLYH WR WKHVH DELOLWLHV LQFOXGHV :HFKVOHU $GXOW ,QWHOOLJHQFH 6FDOH :$,6 :HFKVOHU f :HFKVOHU 0HPRU\ 6FDOH :06 :HFKVOHU f LQFOXGLQJ WKH 5XVVHOO GHOD\HG PHPRU\ DVSHFWV 5XVVHOO f 5H\2VWHUUHLWK &RPSOH[ )LJXUH 2VWHUUHLWK f +RRSHU 9LVXDO 2UJDQL]DWLRQ 7HVW +RRSHU f 9HUEDO )OXHQF\ 7HVW +DOVWHDG f :LVFRQVLQ &DUG 6RUW %HUJ f DQG WKH )LQJHU 7DSSLQJ 7HVW +DOVWHDG f $GPLQLVWUDWLRQ RI WKLV WHVW EDWWHU\ SURYLGHV D EURDG VFUHHQLQJ RI QHXURSV\FKRORJLFDO IXQFWLRQLQJ ZLWK D SDUWLFXODU HPSKDVLV RQ WKRVH DELOLWLHV NQRZQ WR EH VHQVLWLYH WR K\SR[LD $SSOLFDWLRQ RI WKLV EDWWHU\ WR VXEFOLQLFDO DSQHD VXEMHFWV SURYLGHV D VHQVLWLYH LQGLFDWRU RI SRVVLEOH K\SR[LF EUDLQ G\VIXQFWLRQ ,Q VXPPDU\ DOWHUDWLRQV RI VOHHSZDNH F\FOHV DSSHDU WR EH SUHVHQW LQ VOHHS DSQHD V\QGURPH SDWLHQWV 7KXV DVVHVVPHQW RI WKHVH YDULDEOHV LQ D VXEFOLQLFDO SRSXODWLRQ

PAGE 54

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n WRU\ KHDOWK QHXURSV\FKRORJLFDO DQG VOHHSZDNH F\FOH VWDWXV ZDV FDUULHG RXW 7KXV SRWHQWLDO UHODWLRQVKLSV EHWZHHQ UHVSLUDWRU\ GLVWXUEDQFHV DQG YDULDEOHV NQRZQ WR EH GLVUXSWHG LQ VOHHS DSQHD V\QGURPH SDWLHQWV FRXOG EH HYDOXDWHG 7KLV DSSURDFK DOVR DOORZHG H[SORUDWLRQ RI WKH

PAGE 55

SRVVLEOH ULVN IDFWRUV DVVRFLDWHG ZLWK VXEFOLQLFDO VOHHS DSQHDVA

PAGE 56

&+$37(5 7:2 0(7+2' 6XEMHFWV +HDY\ VQRULQJ PDOHV ZHUH UHFUXLWHG YLD QHZVSDSHU DGYHUWLVHPHQWV QRWLFHV DQG SKRQH FDOOV WR D OLVW R DJLQJ VXEMHFWV UHSRUWHG E\ :HEE f 3DUWLFLSDQWV ZHUH UHTXLUHG WR EH PDOH VHOI UHSRUWHG KHDY\ VQRUHUV RYHU \HDUV R DJH DQG LQ JHQHUDO JRRG KHDOWK LH VHOI GHVFULEHG KHDOWK\ DQG QRUPDO DQG QRW XQGHU WKH DFWLYH FDUH RI D SK\VLFLDQ IRU LOOQHVVf 6XEMHFWV ZLWK D KLVWRU\ RI KHDG WUDXPD RU DOFRKROLVP ZHUH H[FOXGHG 3RWHQWLDO VXEMHFWV ZKR PHW WKHVH FULWHULD ZHUH RIIHUHG D FKDQFH WR SDUWLFLSDWH LQ D VQRULQJ VWXG\ LQ ZKLFK WKH\ ZRXOG FRPSOHWH WHVWLQJ GXULQJ DQ HYHQLQJ VOHHS RYHUQLJKW LQ D ODE ZLWK VHYHUDO SK\VLRORJLFDO SDUDPHWHUV UHFRUGHG DQG FRPSOHWH RQH ZHHNnV ZRUWK RI VOHHS ORJV IRU D SD\PHQW RI ILIW\ GROODUV 9ROXQWHHUV ZHUH VFKHGXOHG RQ D ILUVW FRPH ILUVW VHUYHG EDVLV $ WRWDO RI VXEMHFWV ZHUH VWXGLHG ZLWK GURSSHG EHFDXVH RI WHFKQLFDO SUREOHPV ZLWK WKHLU RYHUQLJKW UHFRUGLQJV DQG H[FOXGHG EHFDXVH RI KHDOWK SUREOHPV 7KXV VXEMHFWV UHSUHVHQWHG WKH ILQDO VDPSOH 7KHVH VXEMHFWV KDG D PHDQ DJH RI r VG f DQG UHSUHVHQWHG D ZLGH FURVV VHFWLRQ RI WKH SRSXODWLRQ UDQJLQJ IURP WKH XQHPSOR\HG WR XQLYHUVLW\

PAGE 57

SURIHVVRUV $OO VXEMHFWV FRPSOHWHG DQG VLJQHG DQ LQIRUPHG FRQVHQW DJUHHPHQW $SSDUDWXV 'XULQJ WKH HYHQLQJ WHVWLQJ DOO VXEMHFWV UHFHLYHG D EORRG SUHVVXUH FKHFN ZLWK D VWDQGDUG KRVSLWDO FXII 7KH PHDVXUHPHQW ZDV PDGH IURP D VHDWHG SRVLWLRQ 2YHUQLJKW UHFRUGLQJV ZHUH PDGH LQ D TXLHW GDUNHQHG URRP ZLWK WKH VXEMHFW VOHHSLQJ RQ D VWDQGDUG KRVSLWDO EHG $OO QRFn WXUQDO SK\VLRORJLFDO UHFRUGLQJV ZHUH URXWHG WKURXJK D *UDVV 3RO\JUDSK PRGHO 'f (OHFWURGHV IRU HOHFWUR HQFHSKDORJUDSKLF UHFRUGLQJV ZHUH PRXQWHG LQ WKUHH SDLUV XVLQJ VLWHV )) 3=7 5(/( IURP WKH LQWHUQDWLRQDO V\VWHP RI SODFHPHQW 7KHVH HOHFWURGHV ZHUH DIIL[HG WR WKH VFDOS ZLWK FROORGLRQ VRDNHG JDX]H SDGV GULHG ZLWK DQ DLU KRVH 5HVSLUDWRU\ PHDVXUHV LQFOXGHG PHDVXUHPHQW RI FKHVW DQG DEGRPHQ ZDOO PRYHPHQW GHULYHG IURP LPSHGHQFH PHDVXUHV PRQLWRUHG IURP VXUIDFH HOHFWURGHV ) $J$J&O f SODFHG RQ WKH ORZHU FKHVW DQG MXVW DERYH WKH QDYHO 7KH VLJQDO IURP WKHVH HOHFWURGHV ZDV SDVVHG WKURXJK DQ LPSHGHQFH FRQYHUWRU DQG RQWR WKH SRO\JUDSK 2UDO DQG QDVDO DLUIORZ ZDV UHFRUGHG E\ WKXPLVWRUV 7&75 7UDQVGXFHU*UDVVf FOLSSHG WR WKH UHVSHFWLYH RULILFHV 7KHVH VLJQDOV ZHUH DOVR URXWHG WKURXJK WKH SRO\JUDSK (OHFWURFDUGLRJUDPV ZHUH PRQLWRUHG IURP PRGLILHG ELSRODU FKHVW OHDGV 0&/f )LQDOO\ EORRG R[\JHQ VDWXUDWLRQ ZDV FRQWLQXRXVO\ PRQLWRUHG IURP D SUREH DWWDFKHG WR WKH HDU OREH DQG DQDO\]HG E\ D %LR[ HDU R[LPHWHU IORGHO ,,$ %7$

PAGE 58

&Rf $OO SK\VLRORJLFDO YDULDEOHV ZHUH UHFRUGHG RQ WKH SRO\JUDSKnV FKDUW SDSHU ZKLFK ZDV UXQ DW D VSHHG RI & UDPV 0HDVXUHV 7ZR FODVVHV RI LQGHSHQGHQW YDULDEOHV DUH LGHQWLILHG ZLWK WKH ILUVW EHLQJ WKXPLVWRU YDULDEOHV WKRVH GHULYHG SULPDULO\ IURP WKH QRVH DQG PRXWK WKXPLVWRUVf 7KHVH LQFOXGHG QXPEHU RI DSQHDV DQG K\SRSQHDV PHDQ GXUDWLRQ RI DSQHDV DQG K\SRSQHDV PHDQ ORZ R[\JHQ VDWXUDWLRQ LQ DSQHDV DQG K\SRSQHDV PHDQ R[\JHQ VDWXUDWLRQ FKDQJH LQ DSQHDV DQG K\SRSQHDV DQG WRWDO VHFRQGV VSHQW LQ DSQHDV DQG K\SRn SQHDV 7KXPLVWRU YDULDEOHV ZHUH VHSDUDWHO\ VFRUHG E\ WZR WUDLQHG UDWHUV ZKRVH DJUHHPHQW DSSHDUV LQ 7DEOH $Q RYHUDOO DJUHHPHQW RI ZDV DFKLHYHG 'LVDJUHHPHQWV ZHUH UHVROYHG E\ FRQVHQVXV VFRULQJ 7KH VHFRQG FODVV RI LQGHSHQGHQW YDULDEOHV LQFOXGHG R[\JHQ VDWXUDWLRQ YDULDEOHV GHULYHG IURP D PLQXWH E\ PLQXWH UDWLQJ RI KLJKHVW DQG ORZHVW R[\JHQ VDWXUDWLRQ IURP WKH RYHUQLJKW UHFRUGf 7KHVH YDULDEOHV LQFOXGHG PHDQ KLJKHVW VDWXUDWLRQ PHDQ ORZHVW VDWXUDWLRQ DQG QXPEHU RI GHVDWXUDWLRQV RI DQD RU PRUH ,Q VHOHFWLQJ WKHVH VFRULQJ FULWHULD VHYHUDO FRQVLGHUDWLRQV ZHUH UHOHYDQW ,QLWLDOO\ D V\VWHP ZLWK UHDVRQDEOH LQWHUUDWHU UHOLDELOLW\ ZDV QHFHVVDU\ :KLOH SXEOLVKHG FULWHULD IRU VFRULQJ DSQHDV ZHUH QRW SUREOHPn DWLF UHOLDEOH VFRULQJ RI K\SRSQHDV ZDV QRW DFKLHYHG XQWLO D GHVDWXUDWLRQ FULWHULRQ RI ZDV LQWURGXFHG ,Q VHOHFWLQJ PHDVXUHV RI R[\JHQ VDWXUDWLRQ D PHDQ KLJK DQG

PAGE 59

7DEOH ,QWHUUDWHU DJUHHPHQW IRU UHVSLUDWRU\ HYHQWV RI VQRULQJ PDOHV &RQVHQVXV ,QLWLDO 'LVDJUHHPHQWV 6FRULQJ $JUHHPHQWV Q 1RWH f B 6FRUHG E\ FRQVHQVXV &RQVHQVXV DGGHG &RQVHQVXV DGGHG &RQVHQVXV DGGHG &RQVHQVXV DGGHG &RQVHQVXV DGGHG &RQVHQVXV DGGHG &RQVHQVXV DGGHG &RQVHQVXV DGGHG W &RQVHQVXV DGGHG &RQVHQVXV DGGHG &RQVHQVXV DGGHG f 1RWH RI WKH VXEMHFWV H[SHULHQFHG DW OHDVW DSQHD RU K\SRSQHD

PAGE 60

ORZ VFRUH VHHPHG REYLRXV ,Q RUGHU WR DVVHVV WKH TXDQn WLWDWLYH LPSDFW RI QXPHURXV GHVDWXUDWLRQV D VXP RI b DQG b GHVDWXUDWLRQ ZDV WDEXODWHG 7KHVH VHOHFWHG SDUDPHWHUV IRU WKXPLVWRU DQG VDWXUDWLRQ YDULDEOHV DOORZHG UHOLDEOH VFRULQJ DQG D GHVFULSWLRQ RI VHYHUDO DVSHFWV RI QRFWXUQDO UHVSLUDWLRQ 'HSHQGHQW PHDVXUHV IHOO LQWR WKUHH EURDG FODVVHV PHDVXUHV RI VOHHSZDNH VWDWXV PHDVXUHV RI KHDOWK VWDWXV DQG PHDVXUHV RI QHXURSV\FKRORJLFDO VWDWXV 0HDVXUHV RI WKH VOHHSZDNH F\FOH LQFOXGHG RQH ZHHNnV VOHHS ORJV GHULYHG YDULDEOHV LQFOXGHG QXPEHU RI UHSRUWV RI GD\WLPH VOHHSLQHVV KRXUV DVOHHS HVWLPDWHG VOHHS ODWHQF\ QXPEHU RI ZDNHQLQJV DIWHU VOHHS RQVHW DQG PLQXWHV RI ZDNLQJ DIWHU VOHHS RQVHWf D VOHHS TXHVWLRQQDLUH GHULYHG YDULDEOHV LQFOXGLQJ KRXUV RI VOHHS XVXDOO\ REWDLQHG QXPEHU RI QDSV KRXUV QDSSLQJ TXDOLW\ RI VOHHS UHVWIXOQHVV RQ DZDNHQLQJ GHSWK RI VOHHS DQG DPRXQW RI GD\WLPH VOHHSLQHVVf WKH PHDQ RI WKH VHYHUDO 6WDQIRUG 6OHHSLQHVV 6FDOH UDWLQJV WKH WZR PXOWLSOH VOHHS ODWHQF\ WHVWV GHULYHG YDULDEOHV LQFOXGHG PHDQ VOHHS ODWHQF\ DQG WRWDO PLQXWHV DVOHHS LQ ERWK QDSVf DQG WKH RYHUQLJKW ((& YDULDEOHV GHULYHG IURP WKH VFRULQJ V\VWHP RI $JQHZ DQG :HEE f (OHFWURHQFHSKDORJUDSKLF YDULDEOHV LQFOXGHG WLPH LQ EHG SXUH VOHHS WLPH VOHHS HIILFLHQF\ LQGH[ QXPEHU VWDJH SHULRGV WLPH b VWDJH VOHHS ODWHQF\ WLPH b VWDJHV DQG 5(0 ODWHQF\ ILUVW 5(0

PAGE 61

SHULRG PHDQ 5(0 SHULRG PHDQ 5(0 F\FOH OHQJWK b VORZ ZDYH VOHHS DQG QXPEHU RI VWDJH FKDQJHV 0HDVXUHV RI KHDOWK VWDWXV LQFOXGHG KHLJKW ZHLJKW GLDVWROLF DQG V\VWROLF EORRG SUHVVXUH DV ZHOO DV WKH &RUQHOO 0HGLFDO ,QGH[ )URP WKH &0, UHSRUWV RI K\SHUn WHQVLRQ DQG KHDUW WURXEOH ZHUH QRWHG DV ZHOO DV WKH WRWDO QXPEHU RI V\PSWRPV HQGRUVHG $GGLWLRQDOO\ WKH VFRUHV IRU WKH VFDOHV WDSSLQJ UHVSLUDWRU\ FDUGLDF DQG QHXURORJLFDO V\PSWRPV ZHUH WDEXODWHG VHSDUDWHO\ 0HDVXUHV RI QHXURSV\FKRORJLFDO VWDWXV LQFOXGHG WKH :$,6 9HUEDO DQG 3HUIRUPDQFH ,4fV WKH :06 0HPRU\ 4XRWLHQW GHOD\HG UHFDOO RI WKH ORJLFDO VWRULHV DQG WKH YLVXDO UHn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f PHPRU\ ERWK LPPHGLDWH DQG GHOD\HG YHUEDO DQG QRQn YHUEDOf YLVXRSHUFHSWXDORUJDQL]DWLRQDO VNLOOV ODQJXDJH DQG IURQWDO VHOIUHJXODWRU\ VNLOOV $ VHFRQG HPSKDVLV ZDV RQ D UHODWLYHO\ GHWDLOHG DQDO\VLV RI DUHDV SUHYLRXVO\ VKRZQ WR EH VHQVLWLYH WR K\SR[LD 7KHVH VHQVLWLYH DUHDV

PAGE 62

LQFOXGHG PHPRU\ YLVXRRUJDQL]DWLRQDO VNLOOV YHUEDO IOXHQF\ DQG PRWRU FRRUGLQDWLRQ 7KH FKRVHQ QHXURSV\FKRn ORJLFDO EDWWHU\ UHIOHFWHG WKHVH WZR DLPV 7KXV WKH :$,6 WDSSHG LQWHOOHFWXDO IXQFWLRQLQJ WKH :HFKVOHU 0HPRU\ 6FDOH ZLWK GHOD\HG UHFDOO RI ORJLFDO VWRULHV DQG YLVXDO UHSURGXFWLRQV DV ZHOO DV GLJLW VSDQ DQG 5H\ )LJXUH KHDYLO\ VDPSOHG YHUEDO DQG QRQYHUEDO PHPRU\ WKH +RRSHU WDSSHG YLVXRRUJDQL]DWLRQDO VNLOOV 9HUEDO IOXHQF\ VDPSOHG ODQn JXDJH VNLOOV WKH :LVFRQVLQ FDUG VRUW WDSSHG IURQWDO VHOOUHJXODWRU\ VNLOO DQG )LQJHU WDSSLQJ VDPSOHG PRWRU IXQFWLRQV 7KH FKRVHQ EDWWHU\ SURYLGHG D EURDG VFUHHQLQJ DV ZHOO DV D GHWDLOHG DVVHVVPHQW RI QHXURSV\FKRORJLFDO IXQFWLRQV WKRXJKW WR EH LPSDLUHG E\ K\SR[LD 3URFHGXUH 6XEMHFWV IROORZHG WKH VFKHGXOH DSSHDULQJ EHORZ IRU WKH H[SHULPHQWDO QLJKW 6LJQ LQIRUPHG FRQVHQW EORRG SUHVVXUH KHLJKWZHLJKW ILQJHU WDSSLQJ YHUEDO IOXHQF\ VXEMHFWLYH VOHHSLQHVV 666PDGH HYHU\ 9 KRXU WLOO EHGWLPHf :LULQJ IRU ((* %HJLQ +6/7 Pf :$,6 906 +RRSHU *UDSKHVWKHVLD /XULD PRWRU SURJUDPV %HJLQ 06/7 ,, Pf 5H\ ILJXUH :LVFRQVLQ FDUG VRUW

PAGE 63

6OHHS 4XHVWLRQQDLUH &RUQHOO 0HGLFDO ,QGH[ UHPDLQLQJ WHVWLQJ :LULQJ IRU WKXPLVWRU (.* UHVSLUDWRU\ HIIRUW HDU R[LPHWHU /LJKWV RXW EHJLQ UHFRUGLQJ $ QLJKW WHFKQLFLDQ UHPDLQHG ZLWK WKH VXEMHFWV WKURXJKRXW WKH QLJKW 6XEMHFWV XVXDOO\ DZDNHQHG EHWZHHQ DQG DP DQG ODWHU FRPSOHWHG WKH RQH ZHHNnV VOHHS ORJV 6WDWLVWLFDO 3URFHGXUHV 6HYHUDO VWDWLVWLFDO SURFHGXUHV ZHUH XWLOL]HG LQ HYDOXDWLQJ WKH SUHVHQW GDWD 2QH RI WKH SULPDU\ DSSURDFKHV ZDV FRUUHODWLRQDO FRQVLGHUHG DSSURSULDWH EHFDXVH RI WKH H[SORUDWRU\ QDWXUH RI WKH VWXG\ %HDULQJ LQ PLQG WKH SRVVLEOH DOSKD LQIODWLRQ RI ODUJH FRUUHn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

PAGE 64

&+$37(5 7+5(( 5(68/76 'HPRUUDSKLFV DQG ,QFLGHQFH RLn QRFWXUQDO 5HVSLUDWRU\ 'LVRUGHU 6XEMHFWV ZHUH VHOHFWHG WR IXOILOO VHYHUDO VDPSOLQJ UHTXLUHPHQWV LQFOXGLQJ PDOH VH[ SUHVHQFH RI KHDY\ VQRULQJ DQG VHOI UHSRUW RI JHQHUDO JRRG KHDOWK $ WRWDO RI VXEMHFWV ZHUH VWXGLHG )LYH VXEMHFWV ZHUH ODWHU H[FOXGHG ZKHQ TXHVWLRQLQJ UHYHDOHG WKH SUHVHQFH RI FKURQLF DOFRKROLVP f DQG VHULRXV KHDG WUDXPD f $GGLWLRQDOO\ RYHUQLJKW UHFRUGV ZHUH UHMHFWHG EHFDXVH RI PLVVLQJ RU XQXVDEOH GDWD 7KH FKDUDFWHULVWLFV RI WKH ILQDO E VXEMHFWV DUH GHWDLOHG LQ 7DEOH 7KHVH VQRULQJ PDOHV KDG D PHDQ DJH RI DSSUR[LPDWHO\ \HDUV D PHDQ ZHLJKW RI DSSUR[LPDWHO\ SRXQGV DQG D PHDQ HGXFDWLRQDO OHYHO RI \HDUV RI VFKRROLQJ 7DEOH 'HPRJUDSKLF YDULDEOHV IURP VQRULQJ PDOHV 0 6' 5DQJH $JH :HLJKW OEVf :HLJKW+HLJKW 5DWLR OEVLQFKHVf (GXFDWLRQ \HDUVf

PAGE 65

7KH UHVSLUDWRU\ UHFRUGV ZHUH DQQRWDWHG E\ WZR KRXU LQWHUYDOV IROORZLQJ EHGWLPH )URP D WRWDO RI HYHQWV VFRUHG bf RFFXUUHG LQ WKH ILUVW WZR KRXUV IROORZLQJ EHGWLPH bf RFFXUUHG LQ WKH VHFRQG WZR KRXUV bf RFFXUUHG LQ WKH WKLUG WZR KRXUV ZKLOH bf RFFXUUHG LQ WKH ILQDO WZR KRXUV ,W VKRXOG EH QRWHG WKDW YDNHXS WLPHV YDULHG FRQVLGHUDEO\ PDNLQJ LQWHUSUHn WDWLRQ RI WKH DWWHQXDWHG QXPEHU RI HYHQWV LQ WKH ILQDO WZR KRXUV VRPHZKDW SUREOHPPDWLF )URP D WRWDO RI HYHQWV LQ ZKLFK UHOLDEOH VOHHS VWDJLQJ ZDV DFKLHYHG bf RFFXUUHG LQ OLJKW VORZ ZDYH VOHHS VWDJHV f ZKLOH bf EHJDQ LQ 5(0 VOHHS 2QO\ HYHQWV bf RFFXUUHG LQ VORZ ZDYH VOHHS 7KHVH SUHFHQWDJHV PXVW EH FRQVLGHUHG LQ OLJKW RI WKH UHODWLYH GLVWULEXWLRQ RI VOHHS VWDJHV LQ WKHVH VXEMHFWV VWDJHV bf VWDJH 5(0 bf VWDJHV bf VWDJH bf 7KLV GLVWULEXWLRQ VXJJHVWV WKDW WKH VOHHS WLPH VSHQW LQ VWDJHV RU 5(0 PXVW KDYH SURYLGHG WKH RYHUZKHOPLQJ SUHSRQGHUDQFH RI HYHQWV LQ WKHVH VWDJHV 6OHHS FKDUDFWHULVWLFV ZLOO EH GLVFXVVHG LQ JUHDWHU GHWDLO EHORZ $V QRWHG EHORZ RI WKH VXEMHFWV H[SHULHQFHG DW OHDVW RQH HSLVRGH RI DSQHD RU K\SRSQHD 7DEOH SUHVHQWV WKH GDWD RQ RFFXUUHQFH RI DW OHDVW RQH DSQHDK\SRSQHD DV ZHOO DV WKH RFFXUUHQFH RI D KLJK OHYHO RI DSQHDK\SRSQHD $OPRVW RI WKH VDPSOH KDG D[ OHDVW RQH DSQHD RU K\SRSQHD ZKLOH b KDG KLJK OHYHOV RI

PAGE 66

7DEOH ,QFLGHQFH RI ORZ DQG KLJK OHYHOV RI DSQHDK\SRSQHD E\ DJH LQ VQRULQJ PDOHV $JH *URXS 8 /RZ $SQHD +\SRSQHD +LJK $SQHD +\SRSQHD 2YHUDOO DSQHDK\SRSQHD 1RWH WKDW VXEMHFWV ZLWK KLJK OHYHOV RI DSQHDK\SRSQHD DUH LQFOXGHG LQ WKH RYHUDOO WRWDO RI VXEMHFWV ZLWK DW OHDVW RQH HYHQW IRU WKH IROORZLQJ DQDO\VHV 7KH FODVVLILFDWLRQ RI KLJK OHYHOV RI DSQHDK\SRSQHD XWLOL]HV WKH FULWHULRQ RI DQ DSQHD K\SRSQHD LQGH[ JUHDWHU WKDQ } D IUHTXHQWO\ XVHG FOLQLFDO FXWRII VFRUH 7KH WDEOH DOVR SUHVHQWV WKH IUHTXHQF\ RI UHVSLUDWRU\ GLVWUHVV EURNHQ GRZQ E\ DJH 1R VLPSOH DJH UHODWHG WUHQG LV DSSDUHQW DQG D ; DQDO\VLV RI WKH WZR IDFWRUV RI DJH DQG UHVSLUDWRU\ GLVWUHVV YDV QRW VLJQLn ILFDQW IRU WKRVH ZLWK DW OHDVW RQH HYHQW ; S 16f RU IRU WKRVH ZLWK KLJK OHYHOV RI DSQHDK\SRSQHD ; r S 16f $ IXUWKHU LQYHVWLJDWLRQ RI SRVVLEOH DJH E\ OHYHO RI DSQHD DFWLYLW\ LQYROYHG GLYLVLRQ RI WKH VDPSOH LQWR WKRVH DERYH DQD EHORZ \HDUV RI DJH FURVVHG ZLWK WKRVH DERYH DQG EHORZ HYHQWVKRXU 7KH ; YDOXH IRU WKLV FRPSDULVRQ ZDV DOVR QRQVLJQLILFDQW ; S 16f 6XEMHFWV ZHUH

PAGE 67

DOVR GLYLGHG LQWR WKRVH DERYH DQG EHORZ WKH PHDQ ZHLJKW KHLJKW UDWLR f DQG FURVVHG ZLWK WKRVH ZLWK DQD ZLWKRXW W OHDVW RQH HYHQW RU KLJK OHYHOV R DSQHDE\SRSQHD %RWK WKHVH QRQSDUDPHWULH WHVWV ZHUH DOVR QRQVLJQLILFDQW ; S ,?6 ; S16f ODELD XWLOL]HV 3HDUVRQ FRUUHODWLRQV WR H[SORUH SRVVLEOH UHODWLRQVKLSV EHWZHHQ GHPRJUDSKLF YDULDEOHV DJH ZHLJKW ZHLJKWKHLJKW UDWLRf DQG QRFWXUQDO UHVSLUDWRU\ LQGLFHV DSQHD LQGH[ DSQHD K\SRSQHD LQGH[ VHFRQGV LQ HYHQWV PHDQ KLJK VDWXUDWLRQ PHDQ ORZ VDWXUDWLRQ QXPEHU RI GHVDWXUDWLRQV B!A DQG QXPEHU RI GHVDWXUDWLRQV !bf $JDLQ DJH GLVSOD\V OLWWOH UHODWLRQVKLS ZLWK UHVSLUDWRU\ LQGLFHV EXW VXEVWDQWLDO FRUUHODWLRQV EHWZHHQ ZHLJKW DQG YDULRXV UHVSLUDWRU\ LQGLFHV DUH DSSDUHQW 7DEOH 6LJQLILFDQW Sf 3HDUVRQ FRUUHODWLRQV EHWZHHQ QRFWXUQDO UHVSLUDWRU\ YDULDEOHV DQG GHPRJUDSKLF YDULDEOHV LQ VQRULQJ PDOHV 5HVSLUDWRU\ 9DULDEOH $JH :HLJKW :HLJKWU+HLJKW 5DWLR 1XPEHU RI $SQHDV 1XPEHU RI $SQHDV+\SRSQHDV $SQHD ,QGH[ $SQHDK\SRSQHD ,QGH[ r 0HDQ +LJK 6DWXUDWLRQ r r 0HDQ /RZ 6DWXUDWLRQ r r 0HDQ 6DWXUDWLRQ &KDQJH r r 1XPEHU RI 'HVDWXUDWLRQV +XPEHU RI 'HVDWXUDWLRQV r r rS

PAGE 68

,Q VXPPDU\ QRQSDUDPHWULF DQDO\VHV IDLOHG WR UHYHDO VLJQLILFDQW UHODWLRQVKLSV EHWZHHQ DJH RU ZHLJKW DQG RYHUQLJKW UHVSLUDWLRQ +RZHYHU 3HDUVRQ FRUUHODWLRQV VXJJHVWHD WKDW LQFUHDVLQJ ZHLJKW LV DVVRFLDWHG ZLWK GHWHULRUDWLRQ RI VHYHUDO QRFWXUDO UHVSLUDWRU\ LQGLFHV %HFDXVH LQFUHDVLQJ DSQHDK\SRSQHD OHYHOV KDYH EHHQ VXJJHVWHG WR FDUU\ LQFUHDVLQJ ULVNV IRU YDULRXV GHILFLWV WKH VDPSOH ZDV VWUDWLILHG E\ OHYHO RI DSQHDK\SRSQHD 7KH ILUVW JURXS LQFOXGHG WKRVH VXEMHFWV ZLWKRXW HYHQWV +R HYHQWVf WKRVH ZLWK DW OHDVW RQH HYHQW EXW OHVV WKDQ HYHQWV SHU KRXU /RZ DSQHDK\SRSQHDf DQG WKRVH ZLWK RU PRUH HYHQWV SHU KRXU +LJK DSQHDK\SRSQHDf ,W VKRXOG EH QRWHG WKDW WKLV GLYLVLRQ RI VXEMHFWV LV VOLJKWO\ GLIIHUHQW IURP WKH HDUOLHU QRQSDUDPHWULF DQDO\VHV EHFDXVH LQ WKLV JURXSLQJ ORZ DSQHDK\SRSQHD DQG KLJK DSQHDK\SRSQHD DUH PXWXDOO\ H[FOXVLYH ZKLOH LQ WKH HDUOLHU DQDO\VHV WKH JURXS ZLWK DW OHDVW RQH DSQHDK\SRSQHD LQFOXGHG WKRVH ZLWK PRUH WKDQ HYHQWV SHU KRXU 7DEOH XWLOL]HV RQHZD\ $129$ SURFHGXUHV WR FRPSDUH WKHVH JURXSV RQ GHPRJUDSKLF DQG QRFWXUQDO UHVSLUDWRU\ YDULDEOHV 7KHUH ZHUH QR VLJQLILFDQW EHWZHHQ JURXS GLIIHUHQFHV RQ DQ\ RI WKH GHPRJUDSKLF IDFWRUV DJH ZHLJKW ZHLJKWLKHLJKW UDWLR DQG HGXFDWLRQ >)f S 86@ /RJLFDOO\ ERWK DSQHD DQG K\SRSQHD LQGLFHV ZHUH VLJQLILFDQWO\ GLIIHUHQW EHWZHHQ JURXSV >)f S& S&@ )ROORZXS FRPSDULVRQV 6FKHIIHf LQGLFDWHG WKDW WKH KLJK

PAGE 69

7DEOH 0HDQV DQG VWDQGDUG GHYLDWLRQV IRU VHOHFWHG GHPRJUDSKLFnDQD QRFWXUQDO UHVSLUDWRU\ YDULDEOHV LQ VQRULQJ DSQHDK\SRSQHD PDOHV JURXSHG E\ OHYHO RI 1R $SQHD /RZ $SQHD +LJK $SQHD 1\SRSQHD +\SRSQHD +\SFSQHD $JH f f f :HLJKW OEVf f f f :HLJKW+HLJKW 5DWLR f f f (GXFDWLRQ \HDUVf f f f 1XPEHU RI $SQHDV f fDIF 1XPEHU RI $SQHDV+\SRSQHDV f f $SQHD ,QGH[ f \frE $SQHD+\SRSQHD ,QGH[ f f f f I 6HFRQGV LQ $SQHDV+\SRSQHDV f rfDf' 0HHQ $SQHD 'XUDWLRQ f f 0HDQ +\SFSQHD 'XUDWLRQ f f 0HDQ +LJK 6DWXUDWLRQ f f f 0HDQ /RZ 6DWXUDWLRQ f f f!Of 1XPEHU RI 'HVDWXUDWLRQV B! f f 2fE 1XPEHU RI 'HVDWXUDWLRQV !-?b f f fDfE p 6LJQLILFDQWO\ GLIIHUHQW IURP QR VOHHS DSQHDK\SRSQHD 6LJQLILFDQWO\ GLIIHUHQW IURP ORZ VOHHS DSQHDK\SRSQHD

PAGE 70

DSQHDK\SRSQHD JURXS ZDV VLJQLILFDQWO\ GLIIHUHQW IURP WKH RWKHU WZR JURXSV RQ WKHVH WZR LQGLFHV :KLOH PHDQ GXUDWLRQ RI DSQHDV DQG K\SRSQHDV ZDV QRW VLJQLILFDQWO\ GLIIHUHQW EHWZHHQ WKH KLJK DQG ORZ DSQHD JURXSV >)f r S 16@ WRWDO VHFRQGV LQ UHVSLUDWRU\ HYHQWV ZDV >)f S&@ ZLWK IROORZXS FRPSDULVRQV VKRZLQJ WKDW WKH KLJK DSQHDK\SRSQHD JURXS ZDV VLJQLILFDQWO\ GLIIHUHQW IURP WQH RWKHU WZR JURXSV 7XUQLQJ WR WKH R[\JHQ VDWXUDWLRQ GDWD PHDQ KLJK VDWXUDWLRQ IDLOHG WR UHDFK VLJQLILFDQFH /)f Sr16@ +RZHYHU PHDQ ORZ VDWXUDWLRQ QXPEHU RI GHVDWXUDWLRQV DQG !b ZHUH DOO VLJQLILFDQW >)f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f 6L[W\WZR SHU FHQW RI WKHVH VXEMHFWV KDG DW OHDVW RQH

PAGE 71

DSQHD RU K\SRSQHD ZKLOH b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f ZHUH HYDOXDWHG LQLWLDOO\ WKURXJK D PDWUL[ RI 3HDUVRQ FRUUHODWLRQV ZKLFK LQFOXGHG WKH UHVSLUDWRU\ YDULDEOHV DSQHD LQGH[ DSQHD K\SRSQHD LQGH[ VHFRQGV LQ HYHQWV DV ZHOO DV PHDQ KLJK DQG ORZ VDWXUDWLRQ DQG QXPEHU RI GHVDWXUDWLRQV !$b DQG !bfr )URP D WRWDO RI FRUUHODWLRQV WKUHH ZHUH VLJQLILFDQW DW S RU OHVV 6\VWROLF EORRG SUHVVXUH DQG PHDQ ORZ VDWXUDWLRQ U f 6\VWROLF EORRG SUHVVXUH DQG QXPEHU RI GHVDWXUDWLRQV !$b U f 2YHUDOO QXPEHU RI V\PSWRPV HQGRUVHG RQ WKH &0, DQG PHDQ KLJK VDWXUDWLRQ U f $V D IXUWKHU H[SORUDWLRQ RI SRVVLEOH UHODWLRQVKLSV EHWZHHQ EORRG SUHVVXUH DQG DSQHDK\SRSQHD DFWLYLW\ VXEMHFWV ZHUH

PAGE 72

DVNHG ZQHWKHU WKH\ KDG HYHU EHHQ GLDJQRVHG ZLWK K\SHUn WHQVLRQ 7KLUW\WKUHH SHU FHQW RI WKRVH ZLWK QR HYHQWV KDG GLDJQRVHG K\SHUWHQVLRQ DV GLG b RI WKRVH ZLWK ORZ OHYHOV RI DSQHDK\SRSQHD ZKLOH RQO\ RI WKRVH ZLWK KLJK OHYHOV RI DSQHDK\SRSQHD KDG GLDJQRVHG K\SHUWHQn VLRQ 7KHVH GDWD GR QRW VXSSRUW DQ LQFUHDVLQJ OHYHO RI K\SHUWHQVLRQ LQ VXEMHFWV ZLWK KLJK OHYHOV RI DSQHDK\SRSQHD 7KH VXEMHFWV ZHUH JURXSHG LQWR WKH HDUOLHU GHOLQHDWHG OHYHOV RI DSQHDK\SRSQHD 7DEOH SUHVHQWV WKH PHDQV DQG VWDQGDUG GHYLDWLRQV RI KHDOWK YDULDEOHV IURP WKH WKUHH JURXSV 1R GUDPDWLF WUHQGV DUH DSSDUHQW LQ WKLV WDEOH DQG RQHZD\ $129$ SURFHGXUHV FRQILUPHG WKLV REVHUYDWLRQ ZLWK QR VLJQLILFDQW GLIIHUHQFHV IRU 'LDVWROLF %3 )fr S16 6\VWROLF %3 )Ifm S 16 &0, RYHUDOO )f S 16 &0, UHVSLUDWRU\ )f r S ,?6 &0, FDUGLRSXOPRQDU\ )f S 16 &0, QHXURORJLFDO )f S 16 ,Q VXPPDU\ ZKLOH FRUUHODWLRQDO SURFHGXUHV VXJJHVWHG UHODWLRQVKLSV EHWZHHQ VHYHUDO R[\JHQ VDWXUDWLRQ PHDVXUHV DQG EORRG SUHVVXUH DV ZHOO DV RYHUDOO QXPEHU RI V\PSWRPV HQGRUVHG RQ WKH &0, RQHZD\ $129$V IDLOHG WR UHYHDO VLJQLILFDQW EHWZHHQ JURXS GLIIHUHQFHV ZKHQ VXEMHFWV ZHUH VWUDWLILHG E\ OHYHO RI QRFWXUQDO GLVWUHVV

PAGE 73

7DEOH 0HDQV DQG VWDQGDUG GHYLDWLRQV IRU KHDOWK UHODWHG YDULDEOHV LQ VQRULQJ PDOHV JURXSHG E\ OHYHO RI DSQHDK\SRSQHD 1R $SQHD /RZ $SQHD +LJK $SQHD +\SRSQHD +\SRSQHD +\SRSQHD 6\VWROLF %3D f f f 'LDc VWROLF %3 f f f &0,r 2YHUDOO 6FRUH f f f &0, 5HVSLUDWRU\ 6FRUH f f f &0, &DUGLDF 6FRUH f f f &0, 1HXURORJLFDO 6FRUH f f f I %3 EORRG SUHVVXUH E &0, &RUQHOO 0HGLFDO ,QGH[ 1RFWXUQDO 5HVSLUDWRU\ DQG 6OHHS9DNH 9DULDEOHV ((* 'DWD (OHFWURHQFHSKRJUDSKLF YDULDEOHV ZHUH UHFRUGHG RYHUQLJKW IRU DOO VXEMHFWV 5HFRUGV ZHUH VFRUHG E\ D WUDLQHG WHFKQLFLDQ ZKR XWLOL]HG WKH V\VWHP RI $JQHY DQG :HEE f WR GHWHUPLQH VOHHS VWDJHV ,Q WKUHH UHFRUGV ((* WUDFLQJV ZHUH MXGJHG LQDGHTXDWH IRU GLIIHUHQWLDO VOHHS VWDJLQJ +RZHYHU D MXGJHPHQW RI VOHHS YV ZDNH ZDV PDGH IRU WKHVH VXEMHFWV LQ RUGHU WR IRUP D EDVLV IRU WKH UHVSLUDWRU\ LQGLFHV 7KHVH WKUHH ((* UHFRUGV ZHUH H[FOXGHG IURP WKH IROORZLQJ DQDO\VHV OHDYLQJ UHFRUGV DV D EDVH IRU DQDO\VLV )LIWHHQ YDULDEOHV WLPH LQ EHG SXUH VOHHS WLPH VOHHS HIILFLHQF\ LQGH[ QXPEHU RI VWDJH SHULRGV WLPH b VWDJH VOHHS ODWHQF\ WLPH b VWDJHV DQG 5(0 ODWHQF\ VW 5(0 SHULRG PHDQ 5(0 SHULRG OHQJWK PHDQ 5(0

PAGE 74

F\FOH OHQJWK b VORZ ZDYH VOHHSf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b 6WDJH 7LPH b 6WDJH 7LPH b 6WDJH 7LPH b 6WDJH 7LPH b 6WDJH 7LPH e 6WDJH 5(0 /DWHQF\ VW 5(0 3HULRG 0HDQ 5(0 3HULRG /HQJWK 5(0 &\FOH /HQJWK b 6ORZ :DYH 6OHHS 6WDQGDUG 0HDQ 'HYLDWLRQ r 1RWH VHH WH[W IRU H[SODQDWLRQ RI 1 3DUWLDO FRUUHODWLRQV FRQWUROOLQJ IRU DJH ZHUH FDOFXODWHG EHWZHHQ WKHVH VOHHS DQG WKH UHVSLUDWRU\ YDULDEOHV 2XW RI D WRWDO RI FRUUHODWLRQV RQO\ RQH

PAGE 75

ZDV VLJQLILFDQW DW S* RU OHVV 1XPEHU RI GHVDWXUDWLRQV !b YV 5(0 ODWHQF\ U f 7DEOH SUHVHQWV WKH VOHHS YDULDEOH PHDQV IRU WKH VXEMHFWV JURXSHG E\ OHYHO RI DSQHDK\SRSQHD 7KH JURXS ZLWK KLJK DSQHDK\SRSQHD DSSHDUV WR VSHQG PRUH WLPH DZDNH OHVV WLPH DVOHHS DQG DFKLHYH OHVV VORZ ZDYH VOHHS WKDQ WKH RWKHUV +RZHYHU QRQH RI WKHVH GLIIHUHQFHV SURYHG VLJQLILFDQW ,Q IDFW RQO\ RQH RXW RI RQHZD\ $129$V DFKLHYHG VLJQLILFDQFH 7LPH LQ EHG )f S 16 6OHHS ODWHQF\ )f S16 3XUH VOHHS WLPH )f & S 16 6OHHS HIILFLHQF\ LQGH[ )f S 16 1XPEHU RI VWDJH SHULRGV )f } S 16 7LPH b VWDJH )f } S 16 7LPH b VWDJH )f S 16 7LPH b VWDJH )f S 16 7LPH b VWDJH } )f S& 7LPH b VWDJH } )f S 16 7LPH b VWDJH 5(0 )f S 16 /DWHQF\ VW 5(0 )f S 16 0HDQ 5(0 SHULRG OHQJWK )f S 16 0HDQ 5(0 F\FOH OHQJWK )f S 16 b VORZ ZDYH VOHHS )f S 16 6FKHIIH IROORZXS SURFHGXUHV IRU WLPH b VWDJH UHYHDOHG QR VLJQLILFDQW EHWZHHQ JURXS FRPSDULVRQV 7R VXPPDUL]H ((* VOHHS PHDVXUHV LQGLFDWHF OLJKWHU VOHHS WKHQ QRUPDO LQ WKLV VDPSOH ZLWK LQFUHDVHV LQ DZDNHQLQJV DQG OLJKW VORZ ZDYH VOHHS REVHUYHG 1HLWKHU EHWZHHQ JURXS FRPSDULVRQV QRU FRUUHODWLRQDO SURFHGXUHV UHYHDOHG VLJQLILFDQW LQWHUDFWLRQV EHWZHHQ VOHHS DQG QRFWXUQDO UHVSLUDWRU\ YDULDEOHV

PAGE 76

7DEOH 0HDQV DQG VWDQGDUG GHYLDWLRQV VHOHFWHG VOHHS YDULDEOHV IURP VQRULQJ PDOHV JURXSHG E\ OHYHO RLn DSQHDK\SRSQHD 0R $SQHD /RZ $SQHD +LJK $SQHD +\SRSQHD +\SRSQHD +\SRSQHD 7LPH LQ %HG f f f 6OHHS /DWHQF\ f f f 3XUH 6OHHS 7LPH f f f 6OHHS (IILFLHQF\ ,QGH[ f f f 1XPEHU 6WDJH 3HULRGV f f f 7LPH b 6WDJH f f f 7LPH b 6WDJH f f f 7LPH b 6WDJH f f f 7LPH b 6WDJH f f f 7LPH b 6WDJH f f f 7LPH b 6WDJH 5(0 &f f f /DWHQF\ VW 5(0 3HULRG f f f 0HDQ 5(0 3HULRG /HQJWK f f f 5(0 &\FOH /HQJWK f f f b 6ORZ :DYH VOHHS f f f 1RWH VHH WH[W IRU H[SODQDWLRQ RI 1

PAGE 77

'D\WLPH 6OHHSLQHVV 'DWD ,QFOXGHG DPRQJ WKH PHDVXUHV RI GD\WLPH VOHHSLQHVV ZHUH WKH PHDQ RI WKH VHYHUDO 6WDQIRUG 6OHHSLQHVV 6FDOH UDWLQJV DQG WZR YDULDEOHV FDOFXODWHG IURP WKH WZR HYHQLQJ QDSV PHDQ VOHHS ODWHQF\ DQG WRWDO PLQXWHV DVOHHS GXULQJ WKH QDSV 7KHVH YDULDEOHV ZHUH FRUUHODWHG ZLWK WKH UHVSLUDWRU\ YDULDEOHV UHVXOWLQJ LQ D 3HDUVRQ FRUUHODWLRQ PDWUL[ )URP WKHVH FRUUHODWLRQV WZR HPHUJHG VLJQLn ILFDQW DW S RU OHVV ERWK LQYROYLQJ WKH PHDQ 6WDQIRUG 6OHHSLQHVV 6FDOH UDWLQJV 0HDQ 666 UDWLQJV ZHUH FRUUHn ODWHG ZL[K DSQHD LQGH[ U fr DQG DSQHD K\SRSQHD LQGH[ U f 7DEOH SUHVHQWV WKH PHDQV DQG VWDQGDUG GHYLDWLRQV IRU WKH VXEMHFWV JURXSHG E\ OHYHO RI DSQHDK\SRSQHD 2QHn ZD\ $129$ SURFHGXUHV ZHUH QRQVLJQLILFDQW IRU DOO WKHVH YDULDEOHV >666 )f S 1 0HDQ VOHHS ODWHQF\ )f S 16 7RWDO PLQXWHV DVOHHS )f S 16@ 7DEOH 0HDQV DQG VWDQGDUG GHYLDWLRQV IRU GD\WLPH VOHHSLQHVV YDULDEOHV LQ VQRULQJ PDOHV JURXSHG E\ OHYHO RI DSQHDK\SRSQHD 1R $SQHD /RZ $SQHD +LJK $SQHD +\SRSQHD +\SRSQHD +\SRSQHD 6WDQIRUG 6OHHSLQHVV 6FDOH f Hf f 0HDQ 1DS /DWHQF\ 7RWDO 0LQXWHV RI f f f 6OHHS LQ 1DSV f f f

PAGE 78

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b ZHUH DVVRFLDWHG ZLWK UHSRUWHG PLQXWHV RI QLJKWO\ ZDNHQLQJV 7DEOH SUHVHQWV WKH PHDQV DQG VWDQGDUG GHYLDWLRQV IURP WKH VXEMHFWV JURXSHG E\ OHYHO RI DSQHDK\SRSQHD 2QHZD\ $129$ SURFHGXUHV UHYHDOHG QR VLJQLILFDQW EHWZHHQ JURXS GLIIHUHQFHV RQ WKH VHOHFWHG VOHHS TXHVWLRQQDLUH YDULDEOHV >8VXDO KRXUV VOHHS )f S16 $GHTXDWH VOHHS )f S 16 1XPEHU RI QDSV )f} S 16 +RXUV QDSSLQJ )f S 16 1XPEHU RI

PAGE 79

7DEOH r 6LJQLILFDQW Sf 3HDUVRQ FRUUHODWLRQV EHWZHHQ QRFWXUQDO UHVSLUDWRU\ DQG VOHHS TXHVWLRQQDLUH YDULDEOHV LQ VQRULQJ PDOHV 0LQXWHV 1XPEHU KRXUV 1LJKWO\ RI 1DSV 1DSSLQJ $ZDNHQLQJ $SQHD ,QGH[ $SQHD+\SRSQHD ,QGH[ 6HFRQGV LQ (YHQWV 0HDQ +LJK 6DWXUDWLRQ 0HDQ /RZ 6DWXUDWLRQ /XPEHU RI 'HVDWXUDWLRQV !b /XPEHU RI 'HVDWXUDWLRQV B!b 7DEOH 0HDQV DQG VWDQGDUG GHYLDWLRQV RI VOHHS TXHVWLRQQDLUH YDULDEOHV GHULYHG IURP VQRULQJ PDOHV JURXSHG E\ OHYHO RI DSQHDK\SRSQHD 1R $SQHD +\SRSQHD /RZ $SQHD +\SRSQHD +LJK $SQHD +\SRSQHD 0HDQ K 6OHHS f f f $GHTXDWH $PRXQW" f f f 1XPEHU RI 1DSV
PAGE 80

DZDNHQLQJV )f } S16 0LQXWHV RI DZDNHQLQJ )f H S 16 6OHHS ODWHQF\ )f S 16 'HSWK RI VOHHS )nf Sr16 'D\WLPH VOHHSLQHVV )f S 16@ 0HDQV ZHUH FDOFXODWHG IURP WKH GD\ VOHHS ORJV DQG WKH IROORZLQJ YDULDEOHV ZHUH VHOHFWHG IRU DQDO\VLV UHSRUWV RI GD\WLPH VOHHSLQHVV QXPEHU RI QDSV PLQXWHV QDSSLQJ WRWDO EHGWLPH VOHHS ODWHQF\ QXPEHU RI DZDNHQLQJV PLQXWHV RI DZDNHQLQJV 7KHVH YDULDEOHV ZHUH FRUUHODWHG ZLWK WKH UHVSLUDWRU\ LQGLFHV WR IRUP D 3HDUVRQ FRUUHODWLRQ PDWUL[ WRWDOOLQJ FRUUHODWLRQV 7DEOH SUHVHQWV WKH VLJQLILFDQW Sf FRUUHODWLRQV ZKLFK HPHUJHG $JDLQ QXPEHU RI GHVDWXUDWLRQV ZHUH DVVRFLDWHG ZLWK UHSRUWV RI QDSSLQJ ZKLOH PHDQ KLJK VDWXUDWLRQ ZDV DVVRFLDWHG ZLWK WRWDO EHGWLPH VOHHS ODWHQF\ DQG PLQXWHV DZDNH DQG PHDQ ORZ VDWXUDWLRQ ZDV DVVRFLDWHG ZLWK PLQXWHV DZDNH DV ZHOO 7DEOH SUHVHQWV WKH PHDQV DQG VWDQGDUG GHYLDWLRQV IRU WKH VXEMHFWV GLYLGHG LQWR OHYHOV RI DSQHDK\SRSQHD 2QHZD\ $1&9$ SURFHGXUHV IDLOHG WR UHYHDO VLJQLILFDQW EHWZHHQ JURXS GLIIHUHQFHV >'D\WLPH VOHHSLQHVV )f S 16 1XPEHU RI QDSV )f S 16 0LQXWHV RI QDSSLQJ )fr! S 16 7RWDO EHGWLPH )f S 16 6OHHS ODWHQF\ )fm S 16 1XPEHU RI DZDNHQLQJV )f S 16 0LQXWHV RI ZDNHQLQJV )f S 16@

PAGE 81

7DEOH r 6LJQLILFDQW Sf 3HDUVRQ FRUUHODWLRQV EHWZHHQ QRFWXUQDO UHVSLUDWRU\ DQG VOHHS ORJ YDULDEOHV LQ VQRULQJ PDOHV 1XPEHU 0LQXWHV RI 7RWDO 6OHHS 1LJKWO\ 1DSV %HGWLPH /DWHQF\ :DNHQLQJV $SQHD ,QGH[ $SQHD+\SRSQHD ,QGH[ 6HFRQGV LQ (YHQWV 0HDQ +LJK 6DWXUDWLRQ 0HDQ /RZ 6DWXUDWLRQ .XPEHU RI 'HVDWXUDWLRQV !Bb 1XPEHU RI 'HVDWXUDWLRQV \@4b 7DEOH 0HDQV DQG YDULDEOHV E\ OHYHO VWDQGDUG GHYLDWLRQV RI VOHHS ORJ GHULYHG IURP VQRULQJ PDOHV JURXSHG RI DSQHDK\SRSQHD 1R $SQHD /RZ $SQHD +LJK $SQHD +\SRSQHD +\SRSQHD +\SRSQHD 'D\WLPH 6OHHSLQHVV f f f 1XPEHU RI 1DSV f f f 0LQXWHV 0DSSLQJ f f f 7RWDO %HG 7LPH f f f 6OHHS /DWHQF\ f f f 1XPEHU RI :DNHQLQJV f f f 0LQXWHV 9 D NLQJ f f f

PAGE 82

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

PAGE 83

VLJQLILFDQW RYHUDOO S YDOXH XQLYDULDWH UHJUHVVLRQV DUH XVHG WR LQGLYLGXDOO\ SUHGLFW FULWHULRQ YDULDEOHV ZLWK WKH PXOWLSOH SUHGLFWRUV &RQWULEXWLRQ RI LQGLYLGXDO SUHGLFWRUV PD\ DOVR EH H[DPLQHG LQ WKHVH XQLYDULDWH UHJUHVVLRQV 7KH UHVXOWV IURP WKH PXOWLYDULDWH UHJUHVVLRQ RI WKH DERYH QRWHG FRJQLWLYH VFRUHV DQG VHOHFWHG QRFWXUQDO UHVSLUDWRU\ YDULDEOHV $SQHD LQGH[ $SQHD K\SRSQHD LQGH[ 0HDQ KLJK VDWXUDWLRQ 1XPEHU RI GHVDWXUDWLRQV !b $JH 9HLJKW+HLJKW UDWLRf DUH SUHVHQWHG LQ 7DEOH r $Q RYHUDOO +RWHOOLQJV 7 VFRUH ZDV VLJQLILFDQW Sf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n ORJLFDO VFRUHV LQ RWKHU VDPSOHV WKH YDULRXV RYHUQLJKW

PAGE 84

7DEOH 0XOWLYDULDWH UHJUHVVLRQ RI GHPRJUDSKLF DQG QRFWXUQDO UHVSLUDWRU\ YDULDEOHV RQ FRJQLWLYH VFRUHV LQ VQRULQJ PDOHVr &ULWHULRQ 9DULDEOHV 9$,6 9HUEDO ,4 9$,6 3HUIRUPDQFH ,4 :HFKVOHU 0HPRU\ 6FDOH 0HPRU\ 4XRWLHQW 'HOD\HG 5HFDOO /RJLFDO 6WRULHV 'HOD\HG 5HFDOO 9LVXDO 5HSURGXFWLRQ 'HOD\HG 5HFDOO 5H\ )LJXUH 'LJLW 6SDQ +RRSHU 7HVW )LQJHU 7DSSLQJ /HIW +DQG )LQJHU 7DSSLQJ 5LJKW +DQG 9HUEDO )OXHQF\ :LVFRQVLQ &DUG 6RUW 0XOWLYDULDWH WHVW RI VLJQLILFDQFH +RWHOOLQJV 7A Srf 8QLYDULDWH 5HJUHVVLRQ ZLWK Hf GL 9DULDEOH 0XOWLSOH 5 ) 6LJ 9$,6 9HUEDO ,4 :$,6 3HUIRUPDQFH ,4 :HFKVOHU 0HPRU\ 6FDOH 0HPRU\ 4XRWLHQW f 'HOD\HG 5HFDOO /RJLFDO 6WRULHV 'HOD\HG 5HFDOO 9LVXDO 5HSURGXFWLRQV 'HOD\HG 5HFDOO 5H\ )LJXUH 'LJLW 6SDQ +RRSHU 7HVW )LQJHU 7DSSLQJ 5LJKW +DQG )LQJHU 7DSSLQJ /HIW +DQG 9HUEDO )OXHQF\ :LVFRQVLQ &DUG 6RUW %UHDNGRZQ RI VLJQLILFDQWO\ SUHGLFWHG FULWHULRQ UHJUHVVLRQV &ULWHULRQ 3UHGLFWRU eHWD 6LJ :$,6 3HUIRUPDQFH $JH :HLJKW c+HLJKW 5DWLR $SQHD ,QGH[ $SQHD+\SRSQHD ,QGH[ 0HDQ +LJK 6DWXUDWLRQ 1XPEHU RI 'HVDWXUDWLRQV !$b 3UHGLFWRU 9DULDEOHV $J* :HLJKWc+HLJKW 5DWLR $SQHD ,QGH[ $SQHD+\SRSQHD ,QGH[ 0HDQ +LJK 6DWXUDWLRQ 1XPEHU RI 'HVDWXUDWLRQV !$b

PAGE 85

7DEOH FRQWLQXHG &ULWHULRQ 3UHGLFWRU %HWD 6LJ :HFKVOHU 0HPRU\ $JH 6FDOH 0HPRU\ :HLJKWc+HLJKW 5DWLR 4XRWLHQW $SQHD ,QGH[ $SQHD+\SRSQHD ,QGH[ 0HDQ +LJK 6DWXUDWLRQ 1XPEHU RI 'HVDWXUDWLRQV !b 'HOD\HG 5HFDOO $JH 9LVXDO 9HLJKW+HLJKW 5DWLR 5HSURGXFWLRQV $SQHD ,QGH[ $SQHD+\SRSQHD ,QGH[ 0HDQ +LJK 6DWXUDWLRQ 1XPEHU RI 'HVDWXUDWLRQV !b 'HOD\HG 5HFDOO $JH 5H\ )LJXUH 9HLJKW c+HLJKW 5DWLR $SQHD ,QGH[ $SQHD+\SRSQHD ,QGH[ 0HDQ +LJK 6DWXUDWLRQ 1XPEHU RI 'HVDWXUDWLRQV !Bb +RRSHU 7HVW $JH 9HLJKW+HLJKW 5DWLR $SQHD ,QGH[ $SQHD+\SRSQHD ,QGH[ 0HDQ +LJK 6DWXUDWLRQ 1XPEHU RI 'HVDWXUDWLRQV !b ?YL VFRQ VLQ &DUG $JH 6RUW 9HLJKW c+HLJKW 5DWLR $SQHD ,QGH[ $SQHD+\SRSQHD ,QGH[ 0HDQ +LJK 6DWXUDWLRQ 1XPEHU RI 'HVDWXUDWLRQV B!Ar

PAGE 86

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f DV ZHOO DV WKRVH PRUH GLUHFWO\ UHSUHVHQWLQJ GHHSHQLQJ K\SR[LD 0HDQ ORZ VDWXUDWLRQ DQG +XPEHU RI GHVDWXUDWLRQV A!Af 6XEMHFWV ZHUH JURXSHG E\ OHYHO RI QRFWXUQDO UHVSLUDWRU\ GLVRUGHU QR DSQHDVK\SRSQHDV VXEFOLQLFDO OHYHO RI DSQHDK\SRSQHD DQG FOLQLFDOO\ VLJQLILFDQW OHYHO RI DSQHDK\SRSQHDf 0HDQV DQG VWDQGDUG GHYLDWLRQV IRU WKH QHXURSV\FKRORJLFDO VFRUHV IURP WKHVH JURXSV DUH SUHVHQWHG LQ 7DEOH 7KH WDEOH VXJJHVWV D JHQHUDO GHWHULRUDWLRQ RI FRJQLWLYH VFRUHV DV OHYHO RI QRFWXUQDO UHVSLUDWRU\ GLVWUHVV LQFUHDVHV 2QHZD\ $+29$ SURFHGXUHV ZHUH DSSOLHG WR WKH QHXURSV\FKRORJLFDO VFRUHV IURP HDFK WHVW 6LJQLILn FDQW EHWZHHQ JURXS GLIIHUHQFHV HPHUJHG IRU 9$,6 3,4 /)f } S@ GHOD\HG UHFDOO RI ORJLFDO VWRUHV >)>f S@ GHOD\HG UHFDOO RI YLVXDO UHSURGXFn WLRQV >)f S@ GHOD\HG UHFDOO RI 5H\ FRPSOH[

PAGE 87

7DEOH 6LJQLILFDQW Sf SDUWLDO FRUUHODWLRQV FRQWUROOLQJ ORU DJH EHWZHHQ QRFWXUQDO UHVSLUDWRU\ DQG FRJQLWLYH YDULDEOHV LQ VQRULQJ PDOHV $SQHD $SQHD LO\SRSQHD ,QGH[ ,QGH[ :$,6 9HUEDO ,4 :$,6 3HUIRUPDQFH ,4 r :HFKVOHU 0HPRU\ 6FDOH 0HPRU\ 4XRWLHQW 'HOD\HG 5HFDOO /RJLFDO 6WRULHV 'HOD\HG 5HFDOO 9LVXDO 5HSURGXFWLRQV 'LJLW 6SDQ 'HOD\HG 5HFDOO 5H\ )LJXUH r +RRSHU 7HVW :LVFRQVLQ &DUG 6RUW f )LQJHU 7DSSLQJ 5LJKW +DQG )LQJHU 7DSSLQJ /HIW +DQG 9HUEDO )OXHQF\ r 0HDQ +XPEHU RI )YHQW 6DWXUDWLRQV /nHVDWXUD WLRQV 7LPH +LJK /RZ !$b !B!! r r r r r r r r Y2 rS

PAGE 88

7DEOH r .HDQV DQG VWDQGDUG GHYLDWLRQV RI FRJQLWLYH YDULDEOHV LQ VQRULQJ PDOHV JURXSHF E\ OHYHO RI DSQHDK\SRSQHD 1R $SQHD /RZ $SQHD +LJK $SQHD 9DULDEOH +\SRSQHD +\SRSQHD +\SRSQHD 9$,6 )XOO ,4 f f f :$,6 9HUEDO ,4 f f f :$,6 3HUIRUPDQFH ,4 f f fD
PAGE 89

ILJXUH >)>f S@ DQG 9HUEDO IOXHQF\ >)>f } S@ )ROORZXS FRPSDULVRQV 6FKHIIHf VKRZHG WKH JURXS ZLWK KLJK DSQHDK\SRSQHD WR EH VLJQLILFDQWO\ LPSDLUHG UHODWLYH WR WKRVH ZLWK QR VOHHS DSQHDK\SRSQHD RQ :$,6 3,4 GHOD\HG UHFDOO RI ORJLFDO VWRULHV DQG 5H\ FRPSOH[ ILJXUH DV ZHOO DV YHUEDO IOXHQF\ $GGLWLRQDOO\ WKH KLJK DSQHDK\SRSQHD JURXS ZDV LPSDLUHG UHODWLYH WR WKRVH ZLWK VXEFOLQLFDO DSQHDK\SRSQHD RQ DHOD\HD UHFDOO RI WKH +H\ FRPSOH[ ILJXUH 7KHVH ILQGLQJV DUH EROVWHUHG E\ WKH ODFN RI VLJQLILFDQW EHWZHHQ JURXS GLIIHUHQFHV RQ DJH ZHLJKW RU HGXFDWLRQ VHH 7DEOH f} DOO RI ZKLFK DUH SRWHQWLDO DOWHUQDWLYH H[SODQDWLRQV RI WKH EHWZHHQ JURXS GLIIHUHQFHV $V D FKHFN DQ DQDO\VLV ZDV UXQ WR HYDOXDWH WKH UHODWLYH FRQWULEXWLRQ RI QRFWXUQDO UHVSLUDWRU\ YDULDEOHV DQG VOHHS YDULDEOHV UHIOHFWLQJ IUDJPHQWDWLRQ LQ SUHGLFWLQJ FRJQLWLYHQHXURSV\FKRORJLFDO VFRUHV $ PXOWLYDULDWH UHJUHVVLRQ RI WKHVH VFRUHV ZDV FDUULHG RXW 'HPRJUDSKLF DQG UHVSLUDWRU\ VFRUHV LQFOXGHG DJH DSQHD LQGH[ DSQHD K\SRSQHD LQGH[ DQG QXPEHU RI GHVDWXUDWLRQV !b $ VXEVHW RI WKH HDUOLHU QRWHG VOHHS YDULDEOHV ZDV FKRVHQ WR UHIOHFW PHDVXUHV WKRXJKW WR EH PRVW UHVSRQVLYH WR VOHHS IUDJPHQWDWLRQ 7KHVH LQFOXGHG VOHHS HIILFLHQF\ QXPEHU RI VWDJH SHULRGV WLPH b RI VWDJH WLPH b RI VWDJH DQR QXPEHU RI VWDJH FKDQJHV 7KHVH UHVSLUDWRU\ DQG VOHHS SDUDPHWHUV ZHUH XVHG WR SUHGLFW WKH FRJQLWLYH VFRUHV ZKLFK ZHUH HDUOLHU VKRZQ WR EH SUHGLFWHG E\ QRFWXUQDO

PAGE 90

UHVSLUDWRU\ VFRUHV :$,6 3,4 :HFKVOHU 0HPRU\ 4XRWLHQW GHOD\HG UHFDOO RI YLVXDO UHSURGXFWLRQV DQG +H\ ILJXUH DQG :LVFRQVLQ FDUG VRUW 7DEOH SUHVHQWV WKH GDWD IURP WKLV DQDO\VLV 7KH +RWHOOLQJV 7 ZDV VLJQLILFDQW Sf YLWK GHOD\HG UHFDOO RI YLVXDO UHSURGXFWLRQV DQG 5H\ ILJXUH VLJQLILFDQWO\ SUHGLFWHG E\ WKH SUHGLFWRU VFRUHV ,QVSHFWLRQ RI WKH LQGLYLGXDO UHJUHVVLRQV IRU WKHVH WZR PHDVXUHV GLVFORVHV WKDW DJH DSQHD LQGH[ DQG QXPEHU RI GHVDWXUDWLRQV !$b DOO IRUPHG VLJQLILFDQW FRPSRQHQWV RI WKH SUHGLFWLRQ RI GHOD\HG UHFDOO RI YLVXDO UHSURGXFWLRQV ZKLOH VOHHS HIILFLHQF\ ZDV WKH VROH VLJQLILFDQW SUHGLFWRU RI WKH +H\ ILJXUH 7KHVH GDWD VHHP VXSSRUWLYH RI DQ LQGHSHQGHQW FRQWULEXWLRQ RI ERWK QRFWXUQDO UHVSLUDWRU\ YDULDEOHV DQG VOHHS YDULDEOHV DW D OLPLWHG OHYHO WR WKH SUHGLFWLRQ RI FRJQLWLYHQHXURn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

PAGE 91

7DEOH r 0XOWLYDULDWH UHJUHVVLRQ RI VOHHS DQG QRFWXUQDO UHVSLUDWRU\ RQ FRJQLWLYH YDULDEOHV LQ VQRULQJ PDOHV 3UHGLFWRU 9DULDEOHV &ULWHULRQ 9DULDEOHV $JH :$,6 3HUIRUPDQFH ,4 $SQHD ,QGH[ 9HFKVOHU 0HPRU\ 6FDOH 0HPRU\ $SQHD+\SRSQHD ,QGH[ 4XRWLHQW 1XPEHU RI 'HVDWXUDWLRQV !b 'HOD\HG 5HFDOO 9LVXDO 6OHHS (IILFLHQF\ 5HSURGXFWLRQ 1XPEHU 6WDJH 3HULRGV 'HOD\HG 5HFDOO 5H\ )LJXUH 7LPH b 6WDJH 9HUEDO )OXHQF\ 7LPH b 6WDJH 1XPEHU RI 6WDJH &KDQJHV :LVFRQVLQ &DUG 6RUW 0XOWLYDULDWH WHVW RI VLJQLILFDQFH +RWHOOLQJV S 8QLYDULDWH 5HJUHVVLRQ ZLWK f GI 9DULDEOH 0XOWLSOH 5 ) 9$,6 3HUIRUPDQFH ,4 9HFKVOHU 0HPRU\ 6FDOH 0HPRU\ 4XRWLHQW 'HOD\HG 5HFDOO 9LVXDO 5HSURGXFWLRQV 'HOD\HG 5HFDOO 5H\ )LJXUH 9HUEDO )OXHQF\ f :LVFRQVLQ &DUG 6RUW (UHDNGRZQ RI VLJQLILFDQWO\ SUHGLFWHG FULWHULRQ UHJUHVVLRQV &ULWHULRQ 3UHGLFWRU (HWD 6LJ 'HOD\HG 5HFDOO $JH 9LVXDO $SQHD ,QGH[ 5HSURGXFWLRQ $SQHD+\SRSQHD ,QGH[ 1XPEHU RI 'HVDWXUDWLRQV !b 6OHHS (IILFLHQF\ 1XPEHU RI 6WDJH 3HULRGV 7LPH b 6WDJH 7LPH b 6WDJH f 1XPEHU RI 6WDJH &KDQJHV f

PAGE 92

3DEOH FRQ WLQXHF &ULWHULRQ 3UHGLFWRU %HWD 6LJ} 'HOD\HG 5HFDOO $JH .H\ )LJXUH $SQHD ,QGH[ f $SQHD+\SRSQHD ,QRH[ 1XPEHU RI 'HVDWXUDWLRQV !b f f E 6OHHS (IILFLHQF\ 1XPEHU RL 6WDJH 3HULRGV f f 7LPH b 6WDJH f 6 7LPH b 6WDJH 1XPEHU RL 6WDJH &KDQJHV

PAGE 93

QRQYHUEDO PHPRU\ $GGLWLRQ RI YDULDEOHV UHIOHFWLQJ VOHHS IUDJPHQWDWLRQ WR WKH DQDO\VLV LQGLFDWHG WKDW ERWK QRFWXUQDO UHVSLUDWRU\ DQG VOHHS YDULDEOHV PDNH VLJQLILFDQW DQG LQGHSHQGHQW FRQWULEXWLRQV WR SUHGLFWLRQ RI FRJQLWLYHQHXURSV\FKRORJLFDO VFRUHV

PAGE 94

&+$37(5 )285 ',6&866,21 'HPRJUDSKLFV DQG ,QFLGHQFH R 1RFWXUQDO 5HVSLUDWRU\ 'LVRUGHU 7KH SUHVHQW VDPSOH RI VQRULQJ PDOHV ZDV VHOHFWHG DV D JURXS OLNHO\ WR GLVSOD\ D ZLGH UDQJH RI VOHHS GLVRUGHUHG EUHDWKLQJ 7KHVH VXEMHFWV ZLWK D PHDQ DJH RI \HDUV DQG D PHDQ ZHLJKW RI SRXQGV UHIOHFW D PRVWO\ PLGGOHDJHG WR ROGHU PDOH SRSXODWLRQ ZKLFK LV PRGHUDWHO\ REHVH 7KH GHPRJUDSKLF FKDUDFWHULVWLFV RI WKLV VDPSOH DUH VLPLODU WR WKRVH RI WKH W\SLFDO VOHHS DSQHD V\QGURPH SDWLHQW *XLOOHPLQDXOW HW DO f DOWKRXJK WKH UHTXLUHPHQW RI JHQHUDO JRRG KHDOWK LQVXUHG WKDW VXEMHFWV ZLWK D FODVVLF 6$6 ZHUH H[FOXGHG 7KLV VDPSOH PD\ EH FRQVLGHUHG UHSUHVHQWDWLYH RI VXEMHFWV SUHGLVSRVHG WR VOHHS GLVRUGHUHG EUHDWKLQJ DOWKRXJK DW QR PRUH WKDQ D VXEFOLQLFDO OHYHO 7KHUHIRUH LW VHHPV YDOLG WR FRQVLGHU YDULRXV ULVN IDFWRUV DV WKH\ UHODWH WR QRFWXUQDO UHVSLUDWLRQ ,Q WKH SUHVHQW VDPSOH b H[SHULHQFHG DW OHDVW RQH DSQHD RU K\SRSQHD ZKLOH b VXIIHUHG KLJKHU OHYHOV RI DSQHDK\SRSQHD 2QO\ WZR RWKHU UHSRUWV KDYH GHVFULEHG VDPSOHV VHOHFWHG IRU VQRULQJ RU UHVSLUDWRU\ FRPSODLQWV $QFROL,VUDHO HW DO f H[DPLQHG VXEMHFWV ZLWK FRPSODLQWV RI UHVSLUDWRU\ GLIILFXOWLHV GD\WLPH

PAGE 95

VOHHSLQHVV DQG PXVFXODU HYHQWV QRFWXUQDO P\RFORQXVf )XUWKHU WKHVH VXEMHFWV KDG D PHDQ DJH RI DOPRVW \HDUV LQGLFDWLQJ VLJQLILFDQW VDPSOLQJ DQG DJH GLIIHUHQFHV IURP WKH SUHVHQW VWXG\ 1LQHW\WKUHH SHU FHQW RI WKHVH VXEMHFWV KDG DW OHDVW RQH HYHQW ZKLOH b KDG KLJK QXPEHUV RI DSQHDK\SRSQHD 7KHVH UDWHV DUH PXFK KLJKHU WKDQ WKRVH IRXQG LQ WKH SUHVHQW VDPSOH 7KLV LV SUREDEO\ DWWULEXWDEOH WR WKH VDPSOLQJ GLIIHUHQFHV QRWHG DERYH +RZHYHU D UHFHQW UHSRUW E\ 0LOHV DQG 6LPPRQV f GHVFULEHG DSQHDV LQ D VHULHV RI 6 SDWLHQWV UHIHUUHG IRU FRPSODLQWV RI KHDY\ VQRULQJ ,Q WKLV VDPSOH b KDG DW OHDVW RQH HYHQW ZKLOH b KDG VHYHUH OHYHOV RI DSQHD 7KHVH GDWD DUH TXLWH FRPSDUDEOH WR WKH SUHYDOHQFHV REVHUYHG LQ WKH SUHVHQW VDPSOH b IRU DW OHDVW RQH HYHQW DQG b IRU KLJK OHYHOV RI DSQHDK\SRSQHD DQG OHQG VRPH FRQILGHQFH WR WKH SUHVHQW ILQGLQJ 7KXV WKH RYHUDOO IUHTXHQF\ RI VOHHS GLVRUGHUHG EUHDWKLQJ LQ PLGGOHDJHG VXEMHFWV ZLWK KHDY\ VQRULQJ DSSHDUV WR EH EHWZHHQ DQG b IRU DW OHDVW RQH HYHQW DQG EHWZHHQ DQG b IRU PRUH VHYHUH OHYHOV RI DSQHD 7KH UHODWLYHO\ FORVH FRQFRUGDQFH RI WKH WZR UHSRUWV RQ KHDY\ VQRULQJ VXEMHFWV LV UHDVVXULQJ )DFWRUV ZKLFK ZHUH IRXQG WR EH UHODWHG WR WKH RFFXUUHQFH RI VOHHS DSQHDV LQ WKH SUHYLRXV UHYLHZ LQFOXGHG PDOH VH[ FRPSODLQWV RI VQRULQJ LQFUHDVLQJ DJH DQG REHVLW\ 7KH ILUVW WZR IDFWRUV FRXOG QRW EH H[DPLQHG LQ WKH SUHVHQW VDPSOH EHFDXVH RI WKHLU XVH DV VHOHFWLRQ

PAGE 96

FULWHULD $ ZLGH UDQJH RI DJH ZDV DFKLHYHG LQ WKLV VDPSOH EXW QHLWKHU QRQSDUDPHWULF QRU FRUUHODWLRQDO SURFHGXUHV UHYHDOHG VLJQLILFDQW UHODWLRQVKLSV EHWZHHQ LQFUHDVLQJ DJH DQG VOHHS GLVRUGHUHG EUHDWKLQJ $OWKRXJK VHYHUDO SUHYLRXV UHSRUWV KDYH QRWHG UHODWLRQVKLSV EHWZHHQ LQFUHDVLQJ DJH DQG VOHHS GLVRUGHUHG EUHDWKLQJ W\SLFDOO\ WKLV UHODWLRQVKLS LV PRGHUDWH IRU DSQHD %ORFN HW DO &DUVNDGRQ HW DO %L[OHU HW DO f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n DQG DP DOWKRXJK GLIIHUHQFHV LQ VOHHS RQVHW WLPH DQG ZDNHXS WLPH UHQGHU DQ\ LQWHUSUHWDWLRQ RI

PAGE 97

WKHVH GDWD VRPHZKDW WHQXRXV 7KLV ILQGLQJ QHHGV UHSOLFDWLRQ EHIRUH DQ\ FRQFOXVLRQV PD\ EH GUDZQ 1LQHW\HLJKW SHU FHQW RI DSQHLF HYHQWV RFFXUUHG LQ OLJKW VORZ ZDYH DQG 5(. VOHHS DQ REVHUYDWLRQ ZKLFK LV QRW VXUSULVLQJ JLYHQ WKDW VXEMHFWV VSHQW QHDUO\ e RI WKHLU VOHHS LQ WKHVH VWDJHV 7KH SUHSRQGHUDQFH RI HYHQWV LQ WKHVH VWDJHV RI VOHHS LV FRQVLVWHQW ZLWK WKH EXON RI SDVW UHSRUWV LQ WKLV DUHD HJ %L[OHU HW DO %ORFN HW DO .ULHJHU HW DO f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f ZKR GHPRQVWUDWHG WKDW VXEMHFWV ZLWK VLJQLILFDQW OHYHOV RI VOHHS GLVRUGHUHG EUHDWKLQJ ZHUH LPSDLUHG RQ R[\JHQ GHVDWXUDWLRQ YDULDEOHV UHODWLYH WR VXEMHFWV ZLWKRXW UHVSLUDWRU\ HYHQWV 6XEMHFWV ZLWK KHDY\ VQRULQJ DQG PXOWLSOH QRFWXUQDO DSQHLF UHVSLUDWRU\ HYHQWV DSSHDU WR EH DW ULVN IRU QRFWXUQDO R[\JHQ GHVDWXUDWLRQ ,Q VXPPDU\ WKH SUHVHQW VDPSOH RI VQRULQJ PDOHV DSSHDUV UHSUHVHQWDWLYH RI WKRVH SUHGLVSRVHG WR H[SHULHQFH

PAGE 98

VOHHS GLVRUGHUHG EUHDWKLQJ PLGGOHDJHG VQRULQJ PDOHVf $ SUHYDOHQFH RI t ZDV REVHUYHG IRU RFFXUUHQFH RI DW OHDVW RQH HYHQW ZKLOH }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

PAGE 99

,Q OLJKW RI WKHVH FDYHDWV LW LV RI LQWHUHVW WKDW WZR PHDVXUHV RI GHVDWXUDWLRQ PHDQ ORZ VDWXUDWLRQ DQG QXPEHU RI GHVDWXUDWLRQV !b ZHUH DVVRFLDWHG ZLWK LQFUHDVLQJ EORRG SUHVVXUH UHDGLQJV :KLOH D UHSRUW E\ *XLOOHPLQDXOW HW DO f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nV DVVHVVPHQW RI KHDOWK YDULDEOHV PXVW EH UHJDUGHG DV VRPHZKDW OLPLWHG ,Q VSLWH RI WKLV DQ DVVRFLDWLRQ EHWZHHQ WZR PHDVXUHV RI GHVDWXUDn WLRQ DQG D PHDVXUH RI EORRG SUHVXUH ZDV QRWHG ,W LV VSHFXODWHG WKDW EORRG SUHVXUH PD\ EH PRUH UHVSRQVLYH WR QRFWXUQDO R[\JHQ VDWXUDWLRQ WKDQ WR WKH DSQHLF HYHQWV DORQH

PAGE 100

1RFWXUQDO 5HVSLUDWRU\ DQG 6OHHS:DNH 9DULDEOHV ((* 'DWD 7KH RYHUQLJKW VOHHS LQ WKH ODERUDWRU\ IRU WKHVH VQRULQJ VXEMHFWV ZDV QRWHZRUWK\ IRU D JHQHUDO VKLIW WRZDUGV OLJKWHU VOHHS WKDQ XVXDO $ ORQJHU VOHHS ODWHQF\ PRUH DZDNHQLQJV PRUH OLJKW VOHHS DQG OHVV VORZ ZDYH DQG 5(0 VOHHS WKDQ QRUPDO DUH DSSDUHQW 7KHVH ILQGLQJV SDUDOOHO WKH UHSRUW E\ %ORFN HW DO f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

PAGE 101

KDYH HQRXJK WUHH YDULDQFH WR UHVSRQG WR WKH UHVSLUDWRU\ HYHQWV RI WKH QLJKW $OWHUQDWLYHO\ WKH /(* VFRULQJ V\VWHP ZKLFK DYHUDJHV DFURVV PLQXWH SHULRGV PLJKW PLVV WKH VKRUW DURXVDOV DQD IUDJPHQWDWLRQ UHVXOWLQJ IURP HYHQWV :KDWHYHU WKH UHDVRQ WKH ODFN RI FRYDULDWLRQ RI VOHHS DQG UHVSLUDWRU\ PHDVXUHV LV SX]]OLQJ LQ OLJKW RI WKH UHSRUW RI *XLOOHPLQDXOW HW DO f ZKR GHVFULEHG DURXVDOV DV D IUHTXHQW FRQFRPPLWDQW RI QRFWXUQDO UHVSLUDWRU\ HYHQWV DOWKRXJK %L[OHU HW DO f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f RI OLJKWHU VOHHS LQ D VWUDQJH EHG DQG PD\ KDYH EHHQ LQFUHDVHG E\ WKH GLVFRPIRUW RI WKH UHFRUGLQJ SURFHGXUHV /HVV WKDQ D FKDQFH QXPEHU RI FRUUHODWLRQV ZHUH REVHUYHG EHWZHHQ QRFWXUQDO UHVSLUDWRU\ DQG VOHHS SDUDPHWHUV 7KLV VXUSULVLQJ UHVXOW LV DWWULEXWHG WR DQ H[DFHUEDWHG ILUVW QLJKW HIIHFW ZKLFK SUREDEO\ IXQFWLRQHG WR REVFXUH SRVVLEOH UHVSLUDWRU\ GLVWXUEHG VOHHS SDUDPHWHU

PAGE 102

FRUUHODWLRQV LQ WKLV VDPSOH $GGLWLRQDOO\ DQ ((& VFRULQJ V\VWHP ZKLFK DYHUDJHV DFURVV RQH PLQXWH SHULRGV PLJKW REVFXUH EULHI VOHHS IUDJPHQWDWLRQ IURP QRFWXUQDO HYHQWV $OWKRXJK WKH ODFN RI FRUUHODWLRQ LV SX]]OLQJ %L[OHU HW DO f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f $ VHFRQG DVSHFW RI WKLV UHODWLRQVKLS LQYROYHG UHODWLRQVKLSV EHWZHHQ GHFUHDVHG VDWXUDWLRQ DQG UHSRUWV RI GHFUHDVHG VOHHS ODWHQF\ DQG QLJKWO\ ZDNHQLQJV RQ WKH VOHHS ORJV DQG TXHVWLRQQDLUHV 7KHVH VHHPLQJO\ GLVSDUDWH ILQGLQJV PD\ EH LQWHUSUHWHG DV WKH VHTXHODH RI VOHHS RLVUXSWLRQ UHVXOWLQJ IURP

PAGE 103

GLVWXUEHG QRFWXUQDO EUHDWKLQJ ,Q WKH ILUVW FDVH LQFUHDVLQJ UHVSLUDWRU\ GLVWXUEDQFH FDXVHV VOHHS GHSULYDWLRQ PDQLIHVWHG LQ LQFUHDVHG VXEMHFWLYH VOHHSLQHVV 666 UDWLQJVf DQG FRPSHQVDWRU\ VOHHSLQJ LQFUHDVHG QDSSLQJf 7KH RWKHU VLGH RI WKLV FRLQ LV D VOHHS SUHVVXUH YKLFQ FDXVHV GLVWXUEHG VXEMHFWV WR IDOO DVOHHS TXLFNO\ ORZHUHG VOHHS ODWHQF\f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

PAGE 104

UHSRUWHG LQ VLPLODU VWXGLHV $ ILUVW QLJKW HIIHFW SURn EDEO\ FDXVHG LQFUHDVHG VOHHS ODWHQF\ ZDNHQLQJV DQG OLJKWHU VOHHS D UHVXOW ZKLFK PD\ KDYH IXQFWLRQHG WR REVFXUH SRVVLEOH UHODWLRQVKLSV EHWZHHQ VOHHS DQG UHVSLUDn WRU\ YDULDEOHV ZKLFK ZHUH QRW REVHUYHG $ SDWWHUQ RI FRUUHODWLRQV EHWZHHQ LQFUHDVHG UHVSLUDWRU\ GLVWXUEDQFH DQG VOHHSLQHVV ZDV QRWHG SHUKDSV EHFDXVH WKHVH YDULDEOHV UHIOHFWHG VXEMHFWLYH UHSRUWV ZKLFK ZHUH ODUJHO\ LQGHSHQn GHQW RI WKH H[SHULPHQWDO QLJKWVf VOHHS 666 UDWLQJV PDGH EHIRUH VOHHS VOHHS TXHVWLRQQDLUHV UHIOHFWLQJ WUDLW VOHHS YDULDEOHV DQG VOHHS ORJV DYHUDJHG RYHU D GD\ SHULRG IROORZLQJ WKH H[SHULPHQWDO QLJKWVn VOHHSf 0XOWLSOH QLJKW UHFRUGLQJV DSSHDU QHFHVVDU\ WR VRUW RXW D SRVVLEOH V\QHUn JLVWLF LQWHUDFWLRQ EHWZHHQ WKH ILUVW QLJKW HIIHFW DQG WKH GLVFRPIRUW RI WKH UHFRUGLQJ SURFHGXUHV ,FFWXUQUO KHVSUDWRU\ FQF Kn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

PAGE 105

LQFOXGHG .URS HW DO f DQG .URS HW DO f} ERWK RI ZKLFK QRWHD [KULW FRQWLQXRXV QRFWXUQDO R[\JHQDWLRQ WUHDWPHQW VLJQLILFDQWO\ LPSURYHG QHXURSV\FKRORJLFDO VFRUHV LQ QRUPDO DJHG DQG FKURQLF REVWUXFWLYH OXQJ GLVHDVH SDWLHQWV UHVSHFWLYHO\ 7KH SUHVHQW UHVXOWV DUH DOVR FRQVLVWHQW ZLWK D UHSRUW E\ *UDQW HW DO f GHPRQVWUDWLQJ FRQVLGHUDEOH QHXURSV\FKRORJLFDO LPSDLUPHQW LQ FKURQLF REVWUXFWLYH OXQJ GLVHDVH SDWLHQWV D JURXS NQRYQ WR H[SHULHQFH K\SR[LD ZKLFK LV VLJQLILFDQWO\ H[DFHUEDWHG GXULQJ VOHHS %ORFN f $GGLWLRQDOO\ :HVW f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f,!f 7KHVH VXEMHFWV VKRZHG VLJQLILFDQW LPSDLUPHQWV UHODWLYH WR WKRVH ZLWK QR VOHHS DSQHDK\SRSQHD RQ PHDVXUHV WDSSLQJ QRQYHUEDO LQWHOOLJFQFH GHOD\HG YHUEDO DQG QRQYHUEDO PHPRU\ DQG YHUEDO IOXHQF\ 7KHVH GDWD DUH VXJJHVWLYH WKDW KHDY\ VQRULQJ PDOHV ZLWK PXOWLSOH DSQHDVK\SRSQHDV PD\ EH DW ULVN IRU FRJQLWLYH VHTXHODH ,I WKH LPSDLUPHQWV

PAGE 106

IRXQG LQ WKHVH VXEMHFWV DUH WDNHQ DV PLQLPDOO\ FRPSDUDEOH ZLWK VXEMHFWV H[SHULHQFLQJ WKH FODVVLF VOHHS DSQHD V\QGURPH VRPH VXSSRUW IRU WKH DQHFGRWDO UHSRUWV RI LQn WHOOHFWXDO GHWHULRUDWLRQ LQ 6$6 &XLOOHPLQDXOW HW HO f LV GHULYHG +RZHYHU ERWK WKHVH FRQFOXVLRQV VWDQG LQ QHHG RI UHSOLFDWLRQ EHORUH WKH\ PD\ EH FRQILGHQWO\ DVVHUWHG 7KH PHFKDQLVP XQGHUO\LQJ WKH OLQN EHWZHHQ QRFWXUQDO UHVSLUDWRU\ DQG FRJQLWLYH YDULDEOHV LV D IDVFLQDWLQJ HQLJPD 7R FDWH WZR H[SODQDWLRQV KDYH EHHQ SURSRVHG DV SUR[LPDO FDXVHV RI WKH FKDQJHV REVHUYHG LQ VOHHS DSQHD V\QGURPH 7KH ILUVW RU VOHHS IUDJPHQWDWLRQ K\SRWKHVLV &DUVNDGRQ HW DO f QRWHV WKH ODUJH QXPEHU RI DURXVDOV ZKLFK PD\ DFFRPSDQ\ QRFWXUQDO UHVSLUDWRU\ HYHQWV DQG VXJJHVWV WKDW WKH UHVXOWDQW VOHHS IUDJPHQWDWLRQ PD\ OHDG WR GD\WLPH LPSDLUPHQWV SDUWLFXODUO\ LQ WKH HOGHUO\ :KLOH LQWXLWLYHO\ DSSHDOLQJ WKHUH DUH VHYHUDO GLIILFXOWLHV ZLWK WKLV H[SODQDWLRQ )LUVW WKH GHPRQVWUDn WLRQ RI UHOLDEOH FRJQLWLYH GHILFLWV DIWHU VOHHS GHSULYDn WLRQ KUV SURYHG VXUSULVLQJO\ GLIILFXOW :HEE f &KDQJHV LQ FRJQLWLYH WHVW VFRUHV VXFK DV WKRVH XVHG LQ WKH SUHVHQW VWXG\ KDYH QRW EHHQ UHOLDEO\ REVHUYHG IROORZLQJ VOHHS GHSULYDWLRQ $ VHFRQG OLQH RI HYLGHQFH LQFRQVLVWHQW ZLWK WKH VOHHS IUDJPHQWDWLRQ K\SRWKHVLV LV WKH REVHUYHG ODFN RI FRUUHODWLRQ EHWZHHQ VOHHS VWUXFWXUH DQG FRJQLWLYH YDULDEOHV LQ QRUPDO VXEMHFWV %HUU\ DQF :HEE ,Q SUHVVf $OWKRXJK WKLV FRYDULDWLRQ KDV EHHQ REVHUYHG LQ VXEMHFWV

PAGE 107

ZLWK JUHDWHU SK\VLFDO GHWHULRUDWLRQ )HLQEHUJ HW DO Af VXEMHFWV VHOHFWHG IRU KHDOWK GR QRW GLVSOD\ WKLV HIIHFW $ WKLUG HPEDUUDVVPHQW IRU WKH VOHHS IUDJPHQWDWLRQ K\SRWKHVLV OLHV LQ WKH VWXG\ RI 2UU HW DO f ZKR GHVFULEHG RQH JURXS RI V\PSWRPDWLF VOHHS DSQHD V\QGURPH VXEMHFWV DQG D JURXS RI DV\PSWRPDWLF VQRULQJ VXEMHFWV WKDW ZHUH PDWFKHG RQ QXPEHU RI QRFWXUQDO HYHQWV 3UHVXPDEO\ DURXVDOV UHVXOWLQJ IURP HYHQWV ZHUH PDWFKHG LQ WKH WZR JURXSV \HW WKH V\PSWRPDWLF VXEMHFWV KDG VHULRXV LPSDLUn PHQWV UHODWLYH WR WKH DV\PSWRPDWLF JURXSV 'HVSLWH WKHVH GLIILFXOWLHV WKH SUHVHQW VWXG\ IRXQG WKDW RQH RI WKH ILYH FRJQLWLYH VFRUHV VLJQLILFDQWO\ SUH FLFWHG E\ WKH QRFWXUQDO UHVSLUDWRU\ LQGLFHV ZDV DOVR FRUUHODWHG ZLWK VOHHS HIILFLHQF\ RQ WKH H[SHULPHQWDO QLJKW +RZHYHU JLYHQ WKH HDUOLHU REVHUYHG SUREOHPV ZLWK WKH H[SHULPHQWDO QLJKWVr VOHHS WKH LQFRQVLVWHQW SDVW HYLGHQFH DQG WKH OLPLWHG f UHODWLRQVKLS REVHUYHG WKLV ILQGLQJ UHPDLQV LQFRQFOXVLYH $ PRUH DPELWLRXV VWXG\ H[DPLQLQJ VOHHS DQG UHVSLUDWLRQ DFURVV VHYHUDO QLJKWV LV QHFHVVDU\ WR FODULI\ WKLV LVVXH $Q DOWHUQDWLYH DOWKRXJK DGPLWWHGO\ VSHFXODWLYH H[SODQDWLRQ RI WKH FRJQLWLYH FKDQJHV OLHV LQ WKH FHUHEUDO K\SR[LD DFFRPSDQ\LQJ UHVSLUDWRU\ HYHQWV %UDLQ G\VIXQFn WLRQ LV NQRZQ WR RFFXU LQ PLOG WR PRGHUDWH K\SR[LD
PAGE 108

PXOWLSOH EULHI DFXWH GHVDWXUDWLRQV H[SHULHQFHG E\ WKH VXEMHFW ZLWK DSQHDK\SRSQHD IDLO WR DFWLYDWH WKHVH SURWHFn WLYH PHFKDQLVPV HIIHFWLYHO\ *LEVRQ HW DO f 5HFHQWO\ D ERG\ RI HYLGHQFH KDV DFFXPXODWHG ZKLFK LQGLn FDWHV WKDW WXUQRYHU RI FHUWDLQ QHXURWUDQVPLWWHUV LV GUDPDn WLFDOO\ DOWHUHG E\ DFXWH K\SR[LD 7KH FKROLQHUJLF V\QWKHn WLF SDWKZD\ ZKLFK UHVXOWV LQ WKH IRUPDWLRQ RI WKH QHXUR WUDQVPLWWHU DFHW\OFKROLQH $&Kf DSSHDUV WR EH SDUWLFXODUO\ VHQVLWLYH WR WKH HIIHFWV RI DFXWH K\SR[LD *LEVRQ HW DO fr 7KDW $&K PD\ EH WKH QHXURWUDQVPLWWHU PRVW DIIHFWHG E\ K\SR[LD RI QRFWXUQDO HYHQWV LV SURYRFDWLYH LQ OLJKW RI WKH &KROLQHUJLF K\SRWKHVLV %DUWXV HW DO f RI WKH PHPRU\ DQG FRJQLWLYH LPSDLUPHQW RIWHQ IRXQG LQ DJLQJ VXEn MHFWV 7KLV IRUPXODWLRQ QRWHV WKDW FKROLQHUJLF EORFNDGH LQ \RXQJ VXEMHFWV LQKLELWV PHPRU\ ZKLOH YDULRXV DVSHFWV RI WKH FKROLQHUJLF V\QWKHWLF SDWKYD\ DUH DOWHUHG LQ DJHG VXEMHFWV ZLWK PHPRU\ DQG FRJQLWLYH GHILFLWV $OWHUDWLRQV LQ WKH FQROLQHUJLF V\QWKHWLF SDWKZD\ DUH WKRXJKW WR XQGHUn OLH WKH FRJQLWLYH FKDQJHV LQ ERWK SRSXODWLRQV ,W PD\ EH VSHFXODWHG WKDW WKH FRJQLWLYH G\VIXQFWLRQ IRXQG LQ WKH SUHVHQW VXEMHFWV ZLWK PXOWLSOH DSQHDK\SRSQHD LV DOVR D SURGXFW RI D GLVUXSWLRQ RI FKROLQHUJLF V\QWKHVLV 7R VXPPDUL]H WKH ORJLF RI WKH FHUHEUDO K\SR[LD K\SRn WKHVLV LW LV ILUVW QRWHG WKDW VXEMHFWV ZLWK QRFWXUQDO UHVSLUDWRU\ HYHQWV H[SHULHQFH PXOWLSOH DFXWH K\SR[LF HSLVRGHV 7KLV DFXWH K\SR[LD DSSHDUV WR HYDGH FHUHEUDO SURWHFWLYH PHFKDQLVPV ZKLFK QRUPDOO\ UHVSRQG WR FKURQLF

PAGE 109

& K\SR[LD $FXWH K\SR[LD KDV UHFHQWO\ EHHQ VKRZQ WR VLJQLn ILFDQWO\ DOWHU WKH V\QWKHVLV RI $&K D QHXURWUDQVUULWWHU LPSOLFDWHG LQ PHPRU\ DQG FRJQLWLYH FKDQJHV LQ \RXQJ VXEn MHFWV H[SHULHQFLQJ FKROLQHUJLF EORFNDGH DQG ROGHU VXEn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n WLRQDOO\ PDQLSXODWLRQ RI FKROLQHUJLF SDWKZD\V PLJKW LPSURYH FRJQLWLRQ LQ VXEMHFWV ZLWK QRFWXUQDO UHVSLUDWRU\ HYHQWV +RZHYHU XQWLO D GHPRQVWUDWLRQ RI WKLV VRUW LV FDUULHG RXW WKLV OLQH RI UHDVRQLQJ UHPDLQV KLJKO\ VSHFXn ODWLYH ,Q VXPPDU\ WKH SUHVHQW ILQGLQJV LQGLFDWH FOHDU UHODn WLRQVKLSV EHWZHHQ QRFWXUQDO UHVSLUDWRU\ DQG FRJQL WLYHQHXURSV\FKRORJLFDO YDULDEOHV 7KLV UHVXOW LV FRQn VLVWHQW ZLWK SDVW UHSRUWV RI QHXURSV\FKRORJLFDO LPSDLUPHQWV LQ VXEMHFWV H[SHULHQFLQJ K\SR[LD $SSDUHQWO\ KHDY\ VQRUHUV ZLWK PXOWLSOH DSQHDK\SRSQHD HSLVRGHV PD\ EH

PAGE 110

DW ULVN IRU FRJQLWLYH VHTXHODH 7KHVH GDWD PD\ EH WDNHQ DV OLPLWHF VXSSRUW IRU WKH DQHFGRWDO UHSRUWV RI FRJQLWLYH FKDQJHV LQ 6$6 SDWLHQWV 7ZR H[SODQDWLRQV RI SRVVLEOH XQGHUO\LQJ PHFKDQLVPV RI WKH FRJQLWLYH FKDQJHV PD\ EH FRQVLGHUHG 7KH ILUVW WKH VOHHS IUDJPHQWDWLRQ K\SRn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n FDWHG LQ FRJQLWLYH DELOLWLHV LQ RWKHU SRSXODWLRQV DQG WKXV FDXVHV WKH FRJQLWLYH FKDQJHV 7KH FHUHEUDO K\SR[LD WKHRU\ UHPDLQV VSHFXODWLYH DOWKRXJK VHYHUDO WHVWDEOH K\SRWKHVHV PD\ EH JHQHUDWHG IURP LW )XUWKHU UHVHDUFK LQWR IDFWRUV DIIHFWLQJ WKH VOHHSLQJ EUDLQ PD\ SURYH KHXULVWLF IRU QHXURSV\FKRORJLVWV

PAGE 111

&+$37(5 ),9( 6800$5< $1' &21&/86,216 ,Q WKLV VWXG\ D GHWDLOHG DVVHVVPHQW RI WKH QRFWXUQDO UHVSLUDWLRQ KHDOWK VWDWXV FRJQLWLYHQHXURSV\FKRORJLFDO VNLOOV DQG VOHHSZDNH F\FOH RI VQRULQJ PDOH VXEMHFWV ZDV FDUULHG RXW 6L[W\WZR SHU FHQW RI WKHVH VXEMHFWV H[SHULHQFHG DW OHDVW RQH HYHQW DSQHD RU K\SRSQHDf ZKLOH b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n WXUQDO UHVSLUDWRU\ DQG VOHHS SDUDPHWHUV ZKLFK ZHUH QRW REVHUYHG +RZHYHU RYHUQLJKW UHVSLUDWRU\ GLVWXUEDQFHV ZHUH OLQNHG WR LQFUHDVHG VOHHSLQHVV DQG QDSSLQJ $GGLn WLRQDOO\ UHODWLRQVKLSV EHWZHHQ QRFWXUQDO HYHQWV DQG GHWHULRUDWLQJ VFRUHV RQ WHVWV WDSSLQJ QRQYHUEDO LQWHOOLn JHQFH YHUEDO DQG QRQYHUEDO PHPRU\ H[SUHVVLYH YHUEDO

PAGE 112

IOXHQF\ DQG FRJQLWLYH IOH[LELOLW\ ZHUH REVHUYHG ,W ZDV VSHFXODWHG WKDW K\SR[LF DOWHUDWLRQV LQ WKH FKROLQHUJLF V\QWKHWLF SDWKZD\V XQGHUOD\HG WKHVH FKDQJHV $SSDUHQWO\ D VLJQLILFDQW VXEJURXS RI KHDY\ VQRULQJ PDOHV IDOO RQ WKH VDPH FRQWLQXXP DV VXEMHFWV ZLWK VOHHS DSQHD V\QGURPH RFFXS\LQJ DQ LQWHUPHGLDWH SRVLWLRQ EHWZHHQ QRUPDOV DQG WKRVH ZLWK 6$6 )URP WKHVH DQG RWKHU GDWD LW LV FRQFOXGHG WKDW f EHWZHHQ DQG b RI VQRULQJ PDOHV H[SHULHQFH DW OHDVW RQH QRFWXUQDO UHVSLUDWRU\ HYHQW ZKLOH EHWZHHQ DQG b DSSHDU WR H[SHULHQFH RU PRUH HYHQWV SHU KRXU f PRVW HYHQWV RFFXU LQ OLJKW VORZ ZDYH RU 5(0 VOHHS f LQFUHDVLQJ REHVLW\ LV OLQNHG WR QRFWXUQDO HYHQWV LQ VQRULQJ PDOHV f QRFWXUQDO HYHQWV DUH OLQNHG WR LQFUHDVLQJ EORRG SUHVVXUH LQ VQRULQJ PDOHV f QRFWXUQDO HYHQWV DUH OLQNHG WR LQFUHDVLQJ VOHHSLn QHVV LQ VQRULQJ PDOHV f QRFWXUQDO HYHQWV DUH OLQNHG WR GHWHULRUDWLQJ QHXURSV\FKRORJLFDO VFRUHV LQ VQRULQJ PDOHV f D VXEVHW RI KHDY\ VQRULQJ PHQ WKRVH ZLWK PRUH WKDQ HYHQWV SHU KRXU VXIIHU LQFUHDVHG R[\JHQ GHVDWXUDWLRQ DQG GHFUHDVHG FRJQLWLYH DELOLWLHV LQ WKH DUHDV RI QRQYHUEDO LQWHOOLJHQFH YHUEDO DQG

PAGE 113

QRQYHUEDO PHPRU\ H[SUHVVLYH YHUEDO IOXHQF\ DQF FRJQLWLYH IOH[LELOLW\ f WKH VXEVHW RI KHHY\ VQRULQJ PDOHV ZLWK PRUH WKDQ HYHQWV SHU KRXU DSSHDU WR RFFXS\ DQ LQWHUn PHGLDWH SRVLWLRQ EHWZHHQ QRUPDOV DQG WKRVH ZLWK VOHHS DSQHD V\QGURPH

PAGE 114

$33(1',; $ 6&25,1* 35272&2/ )25 5(63,5$725< 9$5,$(/(6 r 'DWD IURP WKH QRVH DQG PRXWK WKHUPLVWHUV DV ZHOO DV WKH R[LPHWHU DQG VWUDLQ JDXJHVr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

PAGE 115

DHVDWXUDWLFQV H[FHHGLQJ DQG UHODWLYH HDVH RI WKLV UDWLQJ VDWXUDWLRQ UHFRUGV b ZDV NHSW EHFDXVH RI WKH RQO\ RQH UDWHU VFRUHG

PAGE 116

$33(1',; ( 5$: '$7$ 7$%/(6

PAGE 117

7DEOH % 'HPRJUDSKLF DQD PHG LFDO YDL U L D E H V IURP VQRULQJ PDOHV :HLJKW &RUQHOL +H GLFDO LQGH[ 6XE M EGXFD +HLJKW +HLJKW %ORRG +UHVVXUH 2YHUD L .HV &DUGLR KHXUR D $JH WLRQ LQf :HLJKW +X WLR ELDVWROLF 6\V WRLF VFRUH SLUD WRU\ SXOPRQX U\ ORJLFD L 8 ?M EE 22 8 RR E n } \ 2 E 822 &}X RR R E E E E E 8 2O 2E2 &2 8 RR E R X 8 & X f

PAGE 118

er A A A A 9AA ?A! ?9 & XL A A UH A &7 &7 2n &7 &" 8L HU 9 &7 $rr HQ HQ HIL HQ rR F Hr 2 9& & & & ?7 :M r & NR AY&8rAA A BL >?M X A r L UYM r A L r L A f§BB XQURRAAFU URFUFURLYAMAFUFAHFUUR &7 &7 FZ FU FU ‘ b F FU F A F H F FU U F R ?Q r YR URYFUHXLHUAHU FW[rAFFA URrLL?M!L?U?M! f§ LLL1MLYURAURLYURr LAFUAFA AXLDAYF F AF U YMr UY HU HQ &] UX Y[ F & & UH ?U UHQ F & R X U?M >?M Hr U?M H!M HrM UR UR I? UY UH IY UR L?M U?M UH f§r HQ R YML & A FURXL2n&A&Y&A-& FUFWYFA R &R I? F R F Q &DW F 7 F WW X Ye I?AUFF7n9fAL9A&I?MI?n& f§ H" &&&A2A8L&&DA&& ,9 F UR UY Y mY F!FARFRFFRFFFFFFFFF FU F Yn F FU Y V F r! YF F D FU F ; W FURYLFFRAXLFFFFAF*1FFF FFFFFFFFFFFFFFFFFFF FHQUAUR!L?HQFYY!&HUL!HUFFA &&&&&&&&*&&&&&&&&&& &A:&&&&&AI8&1nAI\9 9 H! ; FFFFFRFFFFFFFFFFFFF UYAI?MNAA&-9A&&8L f§ UH H! & FFFFFRFFFRFFFFFFFFF Y URU SRAF RX

PAGE 119

7DEOH % 'D\WLPH VOHHSLQHVV PDOHV YDULDEOHV IURP VQRULQJ 6XEMHFW 0HDQ 6WDQIRUG 6OHHSLQHVV 0HDQ 6OHHS /DWHQF\ 0LQXWHV RI 6OHHS LI 5DWLQJ 1DSV 1DSV B R & f & &

PAGE 120

OfDELH %FRQWLQXHG 0HDQ 6WDQIRUG 0HDQ 6OHHS 0LQXWHV RI 6XEMHFW 6OHHSLQHVV /DWHQF\ 6OHHS D +DWLQJ M 1DSV 1DSV

PAGE 121

7DEOH E cEOHHS TXHVWLRQQDLUH YDULDEOHV IURP VQRULQJ PDOHV 6XEMHFW D 0HDQ +RXUV 6OHHS (QRXJK +XPEHU RI 0DSV +RXUV +DSSLQJ +XPEHU RI :DNLQJV 0LQXWHV $ZDNH 6OHHS /DWHQF\ 'HSWK RI 6OHHS 'D\WLPH 6OHHSLQHVV R r = n L

PAGE 122

&?-&?@&?-&?&?M7&?8?LU2_&6M7&?&?@&?-Wf&?@&?-&?-W&?,227 I2 N? LR} 1$ IL? Q UU? A &0 UR UAL 2 WR &?_ &?c UU? &0 &?O &?2&?&0 U2 }$ &?LAU00W2U UU 2U U &?M &0 &0 2 2&0 UU? MRr [ 0 crA V &0 &0 O$ W &0 &0 2 mU fA‘&0&?-&0[&?-&077&?A&?-A7U0&?MAU 2 &0 &0 A U Kr 2 U UR! .? &0 &0 AUIn2U2&2 2 Y &27&?-LUXU?WW$&2&?-IA&?-&?U&0A2&92AY2U U&?M&?-&?c&?-&?-&?-&?M&?-&?-&?c&?M&?_&?M&?L&?M&Y-?@&?-Y2&?2 &f§ !f§ 1 &4 2n & FH & 1 c! &2 12 &2 PA1&M 2fL2U&0 :n1 A L7?2 1 && &! Ur &?&0 &0 &0 &0 &0 U2 11 Q I$ UR ,1 Q A A n 1 n

PAGE 123

7DELF % 6HYHQ GD\ VOHHS ORJ PHDQV IURP VQRULQJ PDOHV XEMHFW I" 'D\WLPH 6OHHSLQHVV 'XPEHU RI 'DSV 6OHHS /DWHQF\ 1XPEHU RI :DNHQLQJV 0LQXWHV :DNLQJ 0LQXWHV 1DSSLQJ 7RWDO %HGWLPH f§ f§ f§ f§ f§ f§ B f§ f§ f§ Q f§ r R

PAGE 124

$&&2&A!}D0A&2 &7L&7?kA/A&2&2W! 2222222W LRRRRRRRR Ya $O $O $$O $O $, $O I$ $ I$ $$O &1Wf§ YaWa &1,UUUUUW0UW0 Y2 P OWQ FR A Y L UL!F?M!PAeMFRP Ua UInL UU ?_ U Wf§ 22U22UY ,22W222W2 2V $O I$ &?@ I $ $$O $O$-$-VU$$-,$$-I$ 2 W1 L 22 O L F?M L &?L $, O$Y2 , 22 , $O $, Y&1&8A2AW0Ln$Y 1 &2 &$ 2 $O I $ A /$ 0f $$O $O $M I $ I $ c$ I$ 1$ I$ $ I$ I$ U$YfAAAAnAnA

PAGE 125

7DEOH % 8HXURSV\FKROJLFD YDULDEOHV 9$L6 OZ 6XE M ,, 9HUED )HULRU PDQFH )XOO :,n,6D +4 8,/6E /!.9.& 'LJLW 6SDQ E E 2E nM E E E E E 2E E E 2E 2E E 8 E 2E E E E 2E E E R X X U VQRULQJ PDOHV +RRSHU /nO.WrG OHVW )LQJHU 7DSSLQJ +LJKW /HOW 9HUEDO :26E +DQG +DQG )OXHQF\ &E nM 8 E E E E E2 E E E E E ER E E } n U r F EE R E E E E2 E E V E2

PAGE 126

9 8 2nM E :0604 :HFKVOHU 0HPRU\ 6FDOH 0HPRU\ \XRWLHQW 'HOD\HG 5HFDOO /RJLFDO 6WRULHV 'HOD\HG 5HFDOO 9LVXDO 5HSURGXFWLRQV A 'HOD\HG 5HFRLO 5H\ ILJXUH :LVFRQVLQ &DUG 6RUW 22 E2 E E E \ 8 X  2W  E E E 8

PAGE 127

7DEOH Of 5HVSLUDWRU\ YDULDEOHV IURP 6XEMHFW $SQHD LMB ,QGH[ r r $SQHD +\SRSQHD ,QGH[ 6HFRQGV LQ /YHQWV r r 2& ULQe PDOHV 5XPEHU RI 0HDQ 6DWXUDWLRQ 'HVDWXUD WLRQV 5LIK /RZ !b r r r ?'

PAGE 128

U2nU2AUL2A22222U 2W222A22U 1?M U Y &?_ ARRWR R 2nV7LQY PG RYFR$ U U &0 AY' PR &2 &2 &0 &7? Yf§ WF2I7 nW O$ &0 &0 &?D Y AAU?M1AWULAY'1LUYQWQALQRLU?nMfF?M FA FWV [L RA HQ R FnL DR FU HQ R? HQ A HQ HQ HQ HQ HQ HQ HQ HQ HQ 2 $ 2 $ $ 2 A ;f &0 &4 &2 ,$W /+& 1I2W2 P 1nO7L/7L0ALQn L7f“92V'112 1& 7 FQFQFQFQFQFQHQFQFQFQFQFQFQ &7Y FQFQFQFQFQFQFQFQ 22&0VAU2$2&02DR2Y2 2IIO22222WA Df 7LAU 2&&2A $ .2 ?6? Ua ?6?a 9 U$ \ U$ &022AL7L2A2A2222 22222222&12&0I A2AA02222 2 2222222 2222&Q AA & RRRRRRRR 2222I02U2&02222222222222 $ Y& & &Q 4 &?n $ A $$' ; &4 2 2 Y &?$ YL /79Y* &0 &0 &0 &0 &0 c$ f$ $ $ U$ f$ ,$ $ n

PAGE 129

7DEOH ( 6OHHS YDULDEOHV IURP VQRULQJ 1XPEHU XE M WW 727 EOHHS /D WHQF\ EOHHS /LOLFLHQF\ EOHHS 7LQH 7LPH n} 6WDAH 2WDeH 3HULRGV B `2E 2E $ fM E E RR G RR E EUL 2E E 8 E RE E E E 2E E E 2E E G 9 E E 2E E W! EE E E E R RR E E E E E 2E E G E E RR f§ f§ DOHV b 6ORZ 1XPEHU OLPH -R 2/FO,* :DYH 8(IW 0HDQ 0(0 +(0 EOHHS &KDQnHV /D WHQF\ +(0 f\FLL R E EE ZM EER E E E E EE E E EE E E E E E E R E E E 9f

PAGE 130

' r 2A  \ E 9 E  R\ E E \ R\ E E E f } R E R\ E E E R R8 \ E rE T E ‘ 8 R E f R E E f r U?M UR

PAGE 131

5()(5(1&(6 $JQHY +: t :HEE :( f 0HDVXUHPHQW RI VOHHS RQVHW E\ ((* FULWHULD $PHULFDQ -RXUQDO RI ((* 7HFKQRORJ\ $JQHZ +9 :HEE :% t :LOOLDPV 5/ f 7KH ILUVW QLJKW HIIHFW $Q ((* VWXG\ RI VOHHS 3V\FKRSK\VLRORJ\ B $QFROL,VUDHO 6 .ULSNH 0DVRQ 9 t .DSODQ : f $ SUHYDOHQFH VWXG\ RI VOHHS GLVRUGHUV LQ VHQLRUV 6L[PRQWKV UHVXOWV $EVWUDFW SXEOLVKHG IRU $36( PHHWLQJ 6DQ $QWRQLR $QFROL,VUDHO 6 .ULSNH 0DVRQ : t 0HVVLQ 6 f 6OHHS DSQHD DQG QRFWXUQDO P\RFORQXV LQ D VHQLRU SRSXODWLRQ 6OHHS Bf $QQDX = f 7KH FRPSDUDWLYH HIIHFWV RI K\SR[LF DQG FDUERQ PRQR[LGH K\SR[LD RQ EHKDYLRU ,Q % :HLVV t 8 /DLWHV HGVf %HKDYLRUDO 7R[LFRORJ\ 1HZ
PAGE 132

%HUU\ 5 t %ORFN $f} 3RVLWLYH QDVDO SUHVVXUH $ FXUH IRU VQRULQJ $PHULFDQ 5HYLHZ RI 5HVSLUDWRU\ 'LVHDVH } %LUFKILHOG 5 6HLNHU + t +H\PDQ $ f $OWHUDWLRQV LQ EORRG JDVVHV LQ QDWXUDO VOHHS DQG QDUFROHSV\ 1HXURORJ\ B& %L[OHU ( .DOHV $ 6ROGDWRV & 9HOD%XHQR $ -DFRE\ t 6FDURQH 6 f 6OHHS DSQHDF DFWLYLW\ LQ D QRUPDO SRSXODWLRQ 5HVHDUFK &RPPXQLFDWLRQV LQ &KHPLFDO 3DWKRORJ\ DQG 3KDUPDFRORJ\ f %ORFN $f 5HVSLUDWRU\ GLVRUGHUV GXULQJ VOHHS SDUW f +HDUW DQG /XQJ f %ORFN $Df 5HVSLUDWRU\ GLVRUGHUV GXULQJ VOHHS SDUW f +HDUW DQG /XQJ f %ORFN $Ef 3RO\VRPQRJUDSK\ 6RPH GLIILFXOW TXHVWLRQV $QQDOV RI ,QWHUQDO 0HGLFLQH f %ORFN $ %R\VHQ 3 :\QQH t +XQW / f 6OHHS DSQHD K\SRSQHD DQG R[\JHQ GHVDWXUDWLRQ LQ QRUPDO VXEMHFWV 1HZ (QJODQG -RXUQDO RI 0HGLFLQH f %ORFN $ :\QQH t (R\VHQ 3 f 6OHHS GLVRUGHUHG EUHDWKLQJ DQG QRFWXUQDO R[\JHQ GHVDWXUDWLRQ LQ SRVWPHQDSDXVDO ZRPHQ $PHULFDQ -RXUQDO RI 0HGLFLQH } (URDGPDQ (UGPDQQ $ /RUJH :ROI + t %URDDEHQW 7 f 7KH &RUQHOO 0HGLFDO ,QGH[ -RXUQDO RI WKH $PHULFDQ +HFWLFDO $VVRFLDWLRQ %URXJKWHQ 5 f 3HUIRUPDQFH DQG HYRNHG SRWHQWLDO PHDVXUHV RI YDULRXV VWDWHV RI DD\WLPH VOHHSLQHVV 6OHHSW f &DUVNDGRQ 0 t 'HPHQW : f 6OHHS WHQGHQF\ $Q REMHFWLYH PHDVXUH RI VOHHS ORVV 6OHHS 5HVHDUFK &DUVNDGRQ 0 t 'HPHQW : Df 6OHHS WHQGHQF\ GXULQJ H[WHQVLRQ RI QRFWXUQDO VOHHS 6OHHS 5HVHDUFK B &DUVNDGRQ 0 t 'HPHQW : Ef (IIHFWV RI WRWDO VOHHS ORVV RQ VOHHS WHQGHQF\ 3HUFHSWXDO DQG 0RWRU 6NLOOV

PAGE 133

&DUVNDGRQ 0 t 'HPHQW 9L Df &XPXODWLYH HIIHFWV RI VOHHS UHVWULFWLRQ RQ GD\WLPH VOHHSLQHVV 3V\FKRSK\VLRORJ\ &DUVNDGRQ t 'HPHQW : Ef 5HVSLUDWLRQ GXULQJ VOHHS LQ WKH DJHG KXPDQ -RXUQDO RI *HURQWRORJ\ f &DUVNDGRQ 0I t 'HPHQW 9 Df 7KH PXOWLSOH VOHHS ODWHQF\ WHVW :KDW GRHV LW PHDVXUH" 6OHHS B &DUVNDGRQ 0 t 'HPHQW : Ef 1RFWXUQDO GHWHUPLQDQWV RI GD\WLPH VOHHSLQHVV 6OHHS B &DUVNDGRQ 0 YDQ GHQ +RHG t 'HPHQW 9 f ,QVRPQLD DQG VOHHS GLVWXUEDQFHV LQ WKH HOGHUO\ -RXUQDO RI *HULDWULF 3V\FKLDWU\ &KULVWHQVRQ & *OLQHU +RUYDWK 6 t 9DJQHU f (IIHFWV R WKUHH NLQGV RI K\SR[LD RQ YLJLODQFH SHUIRUPDQFH $YLDWLRQ 6SDFH DQG (QYLURQPHQWDO 0HGLFLQH f &RFFDJQD & +DQWRYDQL + EHUJPDQL 7 3DUFKL & E /XJDUHVL ( f 7UDFKHRVWRP\ LQ K\SHUVRPQLD ZLWK SHULRGLF EUHDWKLQJ %XOOHWLQ RI 3K\VLRORJLFDO 3DWKRORJ\ DQG 5HVSLUDWLRQ B &RYHUFWDOH 6 5HDG 9RROFRFN $ t 6FKRHIIHO 5 f 7KH LPSRUWDQFH RI VXVSHFWLQJ VOHHS DSQHD DV D FRPPRQ FDXVH RI H[FHVVLYH GD\WLPH VOHHSLQHVV $XVWUDOLDQ DQG 1HZ =HDODQG -RXUQDO RI 0HGLFLQH t 'DYLV f $GDSWDWLRQ RI EUDLQ PRQRDPLQH V\QWKHVLV WR KYSR[LD LQ WKH UDW -RXUQDO RI $SSOLHG 3K\VLRORJ\ f 'HPHQW 9 t &DUVNDGRQ 0 f &XUUHQW SHUVSHFWLYH LQ GD\WLPH VOHHSLQHVV 7KH LVVXHV 6OHHS 'HPHQW 9 &DUVNDGRQ 0 t 5LFKDUGVRQ f ([FHVVLYH GD\WLPH VOHHSLQHVV LQ WKH VOHHS DSQHD V\QGURPH ,Q & &XLOOHPLQDXOW t 9 'HPHQW HRVf 6OHHS $SQHD 6\QGURPHV 1HZ
PAGE 134

)HLQEHUJ .RUHVNR 5/ t +HOOHU 1 f ((f* VOHHS SDWWHUQV DV D IXQFWLRQ RI QRUPDO DQG SDWKRORJLFDO DJLQJ LQ PDQ -RXUQDO RI 3V\FKLDWULF 5HVHDUFK )LVKHU GH OD 3HQD $ 0D\ILHOG t )OLFNLQJHU 5 f 3V\FKRELRORJLF GHWHUPLQDQWV RI VXEJURXSV RI VOHHS DSQHD 6OHHS 5HVHDUFK B )UDQFKHVFKL 0 =DPSRQL 3I &ULSSD t 6PLUQD 6 f ('6 $ \HDU VWXG\ LQ DQ XQVHOHFWHG LQSDWLHQW SRSXODWLRQ 6OHHS f )UHLGHV ( t $OOZHLVH & f 7UDQVLHQW K\SR[LF DPQHVLD %HKDYLRUDO %LRORJ\ )XMLWD 6 =RULFN ) .RVKRUHFN :LWWLJ 5 &RQZD\ : t 5RWK 7 f 7UHDWPHQW RI XSSHU DLUZD\ VOHHS DSQHD ZLWK 833 5HFHQW H[SHULHQFH 6OHHS 5HVHDUFK *DUD\ 6 5DSDSRUW 6RUNLQ ( (SVWHLQ + )HLQEHUJ ,I t *ROGULQJ 5 f 5HJXODWLRQ RI YHQWLODWLRQ LQ WKH REVWUXFWLYH VOHHS DSQHD V\QGURPH $PHULFDQ 5HYLHZ RI 5HVSLUDWRU\ 'LVHDVH *DVWDXW + 7DVVLQDUL & t 'XURQ % f (WXGH SRO\JUDSKLTXH GHV PDQLIHVWDWLRQV HSLVRLTXHV GLXUQHV HW QRFWXUQHV GX V\QGRPH GH 3LFNZLFN 5HYXH 1HXUROJLH *HOOKRUQ ( f 6HQVLWLYLW\ RI WKH DXGLWRU\ SURMHFWLRQ DUHD WR DQR[LD $PHULFDQ -RXUQDO RI 3K\VLRORJ\ *LEVRQ t %ODVV f ,PSDLUHG V\QWKHVLV RI DFHW\OFKROLQH DFFRPSDQ\LQJ PLOG K\SR[LD DQG K\SRJO\FHPLD -RXUQDO RI 1HXURFKHPLVWU\ *LEVRQ t 'XII\ f ,PSDLUHG V\QWKHVLV RI DFHW\OFKROLQH E\ PLOG K\SR[LD RU QLWURXV R[LGH -RXUQD, RI 1HXURFKHPLVWU\ f *LEVRQ 3XOVLQHOOL : %ODVV t 'XII\ 7 f %UDLQ G\VIXQFWLRQ LQ PLOG WR PRGHUDWH K\SR[LD $PHULFDQ -RXUQDO RI 0HGLFLQH *UDQW +HDWRQ 5 0F6ZHHQ\ $GDPV t 7LPPV 5 f 1HXURSV\FKRORJLFDO ILQGLQJV LQ K\SR[HPLF FKURQLF REVWUXFWLYH SXOPRQDU\ GLVHDVH $UFQLYHV FL ,QWHUQDO 0HGLFLQH

PAGE 135

*XLOOHPLQDXOW & YDQ GHQ +RHG t 0LWOHU 0 f &OLQLFDO RYHUYLHZ RI WKH VOHHS DSQHD V\QGURPHV ,Q & *XLOOHPLQDXOW t 9 'HPHQW HGVf 6OHHS $SQHD 6\QGURPHV 1HZ
PAGE 136

+RRSHU + f 7KH +RRSHU 9LVXDO 2UJDQL]DWLRQ 7HVW +DQXH /RV $QJHOHV :HVWHUQ 3V\FKRORJLFDO 6HUYLFHV .DOHV $ %L[OHU ( 6ROGDWRV & 9HOD%XHQR $ &DOGZHOO $ t &DGLHX[ 5 f ,QVRPQLD 7KH UROH RI VOHHS DSQHD DQG QRFWXUQDO P\RFORQXV 3V\FKRVRPDWLFV } r .UHLVV 3 .ULSNH t $QFROL,VUDHO 6 f 3UHYDOHQFH RI VOHHS DSQHD DPRQJ UHSUHVHQWDWLYH LQSDWLHQWV RQ D YHWHUDQV DGPLQLVWUDWLRQ PHGLFDO ZDUG $EVWUDFW SXEOLVKHG IRU $366 PHHWLQJ 6DQ $QWRQLR 7H[DV .ULHJHU 7XUORW 0DUJLQ 3 t .XUW] f %UHDWKLQJ GXULQJ VOHHS LQ QRUPDO \RXQJ DQG HOGHUO\ VXEMHFWV +\SRSQHDV DSQHDV DQG FRUUHODWHG IDFWRUV 6OHHS Af .URS + %ORFN $ t &RKHQ ( f QHXURSV\FKRn ORJLFDO HILHFWV RI FRQWLQXRXV R[\JHQ WKHUDS\ LQ FKURQLF REVWUXFWLYH SXOPRQDU\ GLVHDVH &KHVW .URS ,, %ORFN $ &RKHQ ( &URXFKHU + t 6FKXVWHU f 1HXURSV\FKRORJLFDO HIIHFWV RI FRQWLQXRXV R[\JHQ WKHUDS\ LQ WKH DJHG &KHVW r /XJDUHVL ( &LULJQRWWD ) &RFFDJQD t %DUX]]L $ f 6QRULQJ DQG WKH REVWUXFWLYH VOHHS DSQHD V\QGURPH ((* 6XSSO 1R /XJDUHVL ( &LULJQRWWD ) &RFFDJQD t 3LDD & f 6RPH HSLGHPLRORJLFDO GDWD RQ VQRULQJ DQG FDUGLRFLUFXODWRU\ GLVWXUEDQFHV 6OHHS Bf /XJDUHVL ( &RFFDJQD %HUWL t 0DQWRYDQL 0 f /D PDOHGL]LRQH GL 2QGLQH LO GLVWXUER GHO UHVSLUR GHO VRQQR QHLO LSRYHQWDOL]LRQH DOYHRODUL SULPDULD /LVW 1HXURORJLH 0F)DUODQG 5 f (IIHFWV RI SDUWLDO DFFOLPLWL]DWLRQ RQ SV\FKRORJLFDO WHVWV DW KLJK DOWLWXGHV -RXUQDO RI &RPSDUDWLYH 3V\FKRORJ\ 0F*LQW\ /LWWQHU 0 %HDKP ( 5XL]3ULPR (
PAGE 137

0LOHV / t 6LPPRQV ) f (YDOXDWLRQ RI SDWLHQWV ZLWK ORXG DQG GLVUXSWLYH VQRULQJ 6OHHS 5HVHDUFK 0LOOHU : f &DUGLDF DUUK\WKPLDV DQG FRQGXFWLRQ GLVWXUEDQFHV LQ WKH VOHHS DSQHD V\QGURPH $PHULFDQ -RXUQDO RI 0HGLFLQH r 0LOQHU ( f 0HPRU\ DQG WKH PHGLDO WHPSRUDO UHJLRQV RI WKH EUDLQ ,Q 3ULEUDP t %URDGEHQW HGVf %LRORJ\ RI 0HPRU\ 1HZ
PAGE 138

6DUD 6 f 'HOD\HG GHYHORSPHQW RI DPQHVWLF EHKDYLRU DIWHU K\SR[LD 3K\VLRORJ\ DQG EHKDYLRU r 6FKPLGW + f 3XSLORPHWULF DVVHVVPHQW RI GLVRUDHUV RI DURXVDO 6OHHS 6FKURHGHU 0RWWD t *XLOOHPLQDXOW & f +HPRG\QDPLF VWXGLHV LQ VOHHS DSQHD ,Q & *XLOOHPLQDXOW t : 'HPHQW HGVf 6OHHS $SQHD 6\QGURPHV 1HZ
PAGE 139

:HFKVOHU fr :HFKVOHU $GXOW ,QWHOOLJHQFH 6FDOH 0DQXDO 1HZ
PAGE 140

%,2*5$3+,&$/ 6.(7&+ 'DYLG 7 5 %HUU\ ZDV ERUQ RQ $SULO LQ &DOLIRUQLD +H DWWHQGHG KLJK VFKRRO LQ 0RQWJRPHU\ $ODEDPD UHFHLYHG D EDFKHORUnV GHJUHH LQ SV\FKRORJ\ IURP $XEXUQ 8QLYHUVLW\ DQG D PDVWHUnV GHJUHH LQ H[SHULPHQWDO SV\FKRORJ\ IURP WKH 8QLYHUVLW\ RI $ODEDPD LQ %LUPLQJKDP $IWHU FRPSOHWLQJ DQ LQWHUQVKLS LQ FOLQLFDO SV\FKRORJ\ LQ &KLFDJR KH SODQV WR VHHN D IDFXOW\ SRVLWLRQ LQ SV\FKRORJ\

PAGE 141

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

PAGE 142

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

PAGE 143

81,9(56,7< 2) )/25,'$ LOOOOOOOOO+


70
In summary, correlational procedures suggested
relationships between nocturnal respiratory variables and
subjective sleepiness, while no reliable between group
difference were noted on the nap or SSS variables.
Subjective Sleep Assessment
Two assessments of subjective sleep parameters were
completed, including the sleep questionnaire and sleep
log. Several variables were selected from the sleep
questionnaire and included: usual hours of sleep,
judgement of whether usual sleep was adequate, number of
naps per week, hours napping per week, number of nightly
wakenings, minutes of nightly wakenings, usual sleep
latency, depth of sleep, and daytime sleepiness. These
variables were correlated with the respiratory variables,
for a Pearson matrix of 63 correlations. Table 3-9
presents the signficant correlations which emerged from
this matrix. Several respiratory variables were
associated with napping variables, while mean low
saturation and number of desaturations >4% were associated
with reported minutes of nightly wakenings.
Table 3-10 presents the means and standard deviations
from the subjects grouped by level of apnea/hypopnea.
Oneway ANOVA procedures revealed no significant between
group differences on the selected sleep questionnaire
variables [Usual hours sleep, F(2,43)=G.6, p-NS; Adequate
sleep, F(2,43)=0.3> p=NS; Number of naps, F(2,43)-2.9
p=NS; Hours napping, F(2,43)=0.8, p=NS; Number of


45
verbal fluency, and signal detection are affected by acute
hypoxia. Animal data suggest that sensitivity of memory
processes to hypoxia, while data from chronic hypoxia
humans indicate that attention, vigilance, abstracting
ability, complex perceptual' motor skills, as well as motor
and sensory abilities are also impaired by hypoxia. While
none of these findings are unequivocally directly
applicable to the brief, acute, desaturaticn occurring
during apnea, they are, at the very least, indicative of
abilities sensitive to hypoxia in humans.
A neuropsychological battery sensitive to these
abilities includes Wechsler Adult Intelligence Scale
(WAIS; Wechsler, 1955); Wechsler Memory Scale (WMS;
Wechsler, 1945) including the Russell delayed memory
aspects (Russell, 1975); Rey-Osterreith Complex Figure
(Osterreith, 1944); Hooper Visual Organization Test
(Hooper, 1958); Verbal Fluency Test (Halstead, 1947);
Wisconsin Card Sort (Berg, 1948); and the Finger Tapping
Test (Halstead, 1947). Administration of this test
battery provides a broad screening of neuropsychological
functioning, with a particular emphasis on those abilities
known to be sensitive to hypoxia. Application of this
battery to subclinical apnea subjects provides a sensitive
indicator of possible hypoxic brain dysfunction.
In summary, alterations of sleep/wake cycles appear
to be present in sleep apnea syndrome patients. Thus
assessment of these variables in a subclinical population


Table 3-16* Multivariate regression of sleep and nocturnal
respiratory on cognitive variables in 43
snoring males.
Predictor Variables
Criterion Variables
Age
WAIS Performance IQ
Apnea Index
V/echsler Memory Scale
Memory
Apnea+Hypopnea Index
Quotient
Number of Desaturations >4%
Delayed Recall Visual
Sleep Efficiency
Reproduction
Number Stage 0 Periods
Delayed Recall Rey Figure
Time % Stage 0
Verbal Fluency
Time % Stage 2
Number of Stage Changes
Wisconsin Card Sort
Multivariate test of significance
Hotellings = 4.063 p<.001
Univariate Regression with
(9.33) d.f.
Variable
Multiple R F.
VAIS Performance IQ
V.echsler Memory Scale
.610 2.17
.051
Memory Quotient
Delayed Recall Visual
.566 1.73
.121
Reproductions
Delayed Recall Rey
.655 2.91
.012
Figure
.668 2.96
.011
Verbal Fluency
578 .85
.096
Wisconsin Card Sort
.453 .95
.499
Ereakdown of significantly
predicted criterion regressions
Criterion
Predictor
Eeta
Sig.
Delayed Recall
Age
-.3596
.020
Visual
Apnea Index
-.3218
.021
Reproduction
Apnea+Hypopnea Index
Number of
.2329
.218
Desaturations >4%
-.3778
.041
Sleep Efficiency
.2303
.408
Number of Stage 0 Periods
.3260
.209
Time % Stage 0
-.1136
.731
Time % Stage 2
.0005
998
Number of Stage Changes
-.2235
391


24
activity and its correlates in a subclinical population.
It is clear from the earlier review of apnea incidence
that sampling from the general population would result in
a relatively low population of individuals with sleep
apnea, an inefficent approach from the perspective of the
limited resources available for research. Thus selection
variables must be considered with one eye on efficiency
and the other on protecting the generalizability of
results. Selection criteria should also aim toward eluci
dating the relative roles and possible interactions of
these variables in sleep apnea. Therefore, selection
variables should have as straight-forward a relationship
as possible with sleep apnea, seeking to answer more ques
tions than they raise.
With these caveats in mind, the variables available
as selection criterion will be reviewed and conclusions
drawn about their relative merits. The first variable
considered, sex, has been relatively well investigated.
It is clear that sleep apnea activity is primarily a male
circumstance. Therefore, the selection of female subjects
is contraindicated from the standpoint of efficiency. Age
has been demonstrated to exhibit a positive relationship
with sleep apnea, indicating selection of older sub
jects. Snoring is another variable with a close relation
ship to sleep apnea. The apparently common mechanism of
snoring and sleep apnea (stenosis of the upper airway)
indicates that possibly very few subjects witnout snoring


124
Berry, R., & Block, A.J. (1983) Positive nasal
pressure: A cure for snoring. American Review of
Respiratory Disease, 127 84.
Birchfield, R., Seiker, H., & Heyman, A. (1958).
Alterations in blood gasses in natural sleep and
narcolepsy. Neurology, _C, 107-112.
Bixler, E., Kales, A., Soldatos, C., Vela-Bueno, A., Jacoby,
J., & Scarone, S. (1982). Sleep apneac activity in a
normal population. Research Communications in Chemical
Pathology and Pharmacology, 56(1), 141-152.
Block, A.J. (1980). Respiratory disorders during sleep
(part 1). Heart and Lung. 9.(6), 1101-1124.
Block, A.J. (1981a). Respiratory disorders during sleep
(part 2). Heart and Lung, 10(1), 90-96.
Block, A.J. (1981b). Polysomnography: Some difficult
questions. Annals of Internal Medicine, 95(5), 644-656.
Block, A., Boysen, P., Wynne, J., & Hunt, L. (1979). Sleep
apnea, hypopnea, and oxygen desaturation in normal
subjects. New England Journal of Medicine. 300(10),
513-517.
Block, A., Wynne, J., & Eoysen, P. (1980). Sleep
disordered breathing and nocturnal oxygen desaturation
in post-menapausal women. American Journal of Medicine,
69 75-79.
Eroadman, K., Erdmann, A., Lorge, I-, Wolf, H., & Broaabent,
T. (1949). The Cornell Medical Index. Journal of the
American Hectical Association, 140, 530-534.
Broughten, R. (1982), Performance and evoked potential
measures of various states of aaytime sleepiness.
Sleept (5), 135-146.
Carskadon, M., & Dement, W. (1977). Sleep tendency: An
objective measure of sleep loss. Sleep Research, 6,
200.
Carskadon, M., & Dement, W. (1979a). Sleep tendency during
extension of nocturnal sleep. Sleep Research, 8_, 147.
Carskadon, M., & Dement, W. (1979b). Effects of total
sleep loss on sleep tendency. Perceptual and Motor
Skills, 48, 495-506.


47
possible risk factors associated with subclinical sleep
apneas


Carskadon, M., & Dement, Vi. (1981a). Cumulative effects of
sleep restriction on daytime sleepiness.
Psychophysiology, 16, 107-113.
Carskadon, K., & Dement, W. (1981b). Respiration during
sleep in the aged human. Journal of Gerontology, 36(4),
420-423.
Carskadon, M.f & Dement, V. (1982a). The multiple sleep
latency test: What does it measure? Sleep, _5, 367-372.
Carskadon, M., & Dement, W. (1982b). Nocturnal
determinants of daytime sleepiness. Sleep, _5, 573-581 .
Carskadon, M., van den Hoed, J., & Dement, V. (1980).
Insomnia and sleep disturbances in the elderly. Journal
of Geriatric Psychiatry, 13, 135-151.
Christenson, C., Gliner, J., Horvath, S., & Vagner, J.
(1977). Effects o three kinds of hypoxia on vigilance
performance. Aviation Space and Environmental Medicine,
18(6), 491-497.
Coccagna, C., Hantovani, H. bergmani, T., Parchi, C., b
Lugaresi, E. (1972). Tracheostomy in hypersomnia with
periodic breathing. Bulletin of Physiological Pathology
and Respiration, _8, 1217-1227.
Coverctale, S., Read, D. Voolcock, A., & Schoeffel, R.
(1980). The importance of suspecting sleep apnea as a
common cause of excessive daytime sleepiness.
Australian and New Zealand Journal of Medicine, 10, 284-
20&:
Davis, J. (1975). Adaptation of brain monoamine synthesis
to hvpoxia in the rat. Journal of Applied Physiology,
39(2), 215-220.
Dement, V., & Carskadon, M. (1982). Current perspective in
daytime sleepiness: The issues. Sleep, 5., 556-566.
Dement, V/., Carskadon, M. & Richardson, G. (1978).
Excessive daytime sleepiness in the sleep apnea
syndrome. In C. Cuilleminault & V. Dement (eos), Sleep
Apnea Syndromes. New York: Alan R. Liss Inc.
Devereaux, M., & Partnow, M. (1975). Delayed hypoxic
encephalopathy without cognitive dysfunction. Archives
of Neurology, 32, 704-705.
Dolly, F., & Block, A. (1982). Effect of flurazepam on
sleep disordered breathing in asymptommatic subjects.
American Journal of Medicine, 73. 239-243.


55
2200 Sleep Questionnaire, Cornell Medical Index,
remaining testing
2230 Wiring for thumistor, EKG, respiratory effort,
ear oximeter
2300 Lights out, begin recording
A night technician remained with the subjects throughout
the night. Subjects usually awakened between 5 and 6 am,
and later completed the one week's sleep logs.
Statistical Procedures
Several statistical procedures were utilized in
evaluating the present data. One of the primary
approaches was correlational, considered appropriate
because of the exploratory nature of the study. Bearing
in mind the possible alpha inflation of large corre
lational matrices, other approaches were also utilized to
add confidence. Thus the sample was stratified by level
of apnea/hypopnea, and oneway ANOVA procedures were used
to evaluate between group differences. Additionally, non-
parametric procedures were utilized where appropriate.
Positive correlational results were considered against
possible replication with other measures in the study, and
in some cases multivariate procedures were added for
further confidence. It is conceded that correlational
approaches may carry the burden of possible spurious
results, but it is believed that the various additional
procedures outlined above lend confidence to the findings.


42
neurotransmitter turnover seen in hypoxia. It has been
suggested that these subtle alterations may underlie the
changes in complex behavior seen in hypoxic humans.
Data on human performance after hypoxia have neces
sarily been reliant on fortuitous or weak manipulations.
Although this may limit the usefulness of these data, it
is at the very least highly suggestive of variables sensi
tive to hypoxia. Van Liere and Stickney (1963) note that
acute hypoxia leads to confusion, headache, drowsiness,
weakness and incoordination. After effects include head
ache, nausea and emotional lability. Richardson ex al.
(1959) found that acute, fatal hypoxia from surgery
results in damage to cortical neurons, particularly in
layers III and IV. Susceptible subcortical neurons in
clude those of the corpus striatum and cerebellum. Plum
et al. (1962) describe a syndrome of delayed postanoxic
encephalopathy. After the acute coma resolves in 4-5
days, a brief period of normal functioning returns.
Increasing irritability and confusion then emerge, with
loss of coordination and memory, and diminished attention
span. Typically, the lesion spares grey matter while
affecting white matter. Devereaux and Partnow (1975)
describe a patient who recovered from a delayed acute
encephalopathy. Although IQ was relatively spared, severe
dysarthria emerged. Relevant points which emerge from
studies of acute hypoxia in humans suggest that sublethal


SLEEP APNEA ACTIVITY AND ITS CONCOMITANTS
IN A SUBCLINICAL POPULATION
By
DAVID THOMAS REED BERRY
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1985


D
*=93
O.^
3/2
3.1
5
2.0
33
4 9
2
0.9
5/y
11.1
8
1 b
3 4
378
33
0. V
2b 4
0
10
13.8
35
555
51
o.y
240
21.8
b
1.5
5<>
321
2 b
0.9
284
2.7
4
5.8
3/
424
28
0.8
55y
8.9
9
2.5
58
325
21
o.y
221
9.4
7
1 b
39
41b
5
0.9
505
b. 5
7
2.2
40
41
42
372
1
0.8
519
0.9
10
3.2
35o
1
0.9
354
3.7
11
11.5
43
442
28
0.8
552
8.8
6
3.6
44
425
5
0.9
575
9.b
1.
J
2.9
45
4 54
22
o.y
528
20.4
25
5.b
4b
5/0
54
0.8
305
2.8
b
0.8
I
1o.9
oU. 9
5.2
11.2
18.y
40
70
18. 3
100.7
0.5
85.3
0.7
3.1
4.2
27
102
1
0
10.0
99.7
1.2
5.3
5.7
37
67
1 1
106
1.5
09.1
1 .b
5*b
9.2
23
q298
2
5
10.b
56.1
5.1
28.4
50.3
29
147
15.5
1U9
9.9
75.1
1 .0
5.5
6.9
35
1o9
13
133
14.8
85.9
4.9
5.5
11.5
30
127
18
95
10.7
67.2
1.9
13.6
1 b o
37
177
14.7
98
25.5
59.5
2.4
7.9
10.9
40
91
26.7
68
12.4
81.7
->.2
7.5
11.1
12.4
81.7
10.8
<2
15.5
47.9
7.5
19.2
26.9
5b
113
20
123.5
10.b
84 6
2.4
6.9
12.5
40
95
17.3
63.7
5.4
80.2
5.4
7.0
13.1
70
393
7
17
16.7
03.1
5.9
11.2
17.2
29
123
28
1 3*5
r\j
ro


34
these data. Below a discussion of measurement issues and
proposal assessment of these variables appears.
Measurement of Deficits Found in Sleep Apnea Syndrome
Above it was noted that cardiopulmonary complica
tions, excessive daytime sleepiness, nocturnal oxygen
desaturation and cognitive/intellectual changes are common
deficits of sleep apnea syndrome. A discussion of
measurement issues and proposed measurement devices for
these variables will be presented below.
Measurement of Cardiopulmonary and Health Deficits
Cardiopulmonary deficits are thought to be frequent
sequelae of sleep apnea syndrome. Thus an assessment of
blood pressure as well as self reports of hypertension and
heart trouble seems indicated in subclinical apnea sub
jects. Additionally, a general survey of health status,
such as that provided by the Cornell Medical Incex (CMI;
Broadman et al., 1949) would also screen for other health
deficits. The CMI provides an overall score inaicating
number of symptoms endorsed, as well as subscales on
several symptom categories. Separate examination of sub
scales on respiratory, cardiopulmonary, and neurological
subscales seems appropriate. Assessment of these
variables provides a broad health screening, as well as a
detailed analysis of symptoms found in sleep apnea syn
drome.


CHAPTER THREE
RESULTS
Demorraphics and Incidence oi'
nocturnal Respiratory Disorder
Subjects were selected to fulfill several sampling
requirements, including male sex, presence of heavy
snoring, and self report of general good health. A total
of 60 subjects were studied. Five subjects were later
excluded when questioning revealed the presence of chronic
alcoholism (2) and serious head trauma (3). Additionally
9 overnight records were rejected because of missing or
unusable data. The characteristics of the final 4b
subjects are detailed in Table 3-1- These snoring males
had a mean age of approximately 30 years, a mean weight of
approximately 190 pounds, and a mean educational level of
15 years of schooling.
Table 3-1. Demographic variables
from 46
snoring
males.
M
SD
Range
Age
49.9
13.1
30-75
Weight (lbs)
189.3
34.9
125-250
Weight:Height Ratio (lbs:inches)
2.7
0.5
1.9-3.6
Education (years)
14.7
2.5
9-20
56


20
in mind, they conducted a survey of 1000 subjects on the
incidence of snoring. Chronic snorers included 31# of the
males and 19% of the females. In an elaboration of this
study, Lugaresi et al. (1980) questioned 5713 indivi
duals. In this sample, 24% of the males and 13% of the
females were chronic snorers. Snoring increased with age,
and by 60 years of age 60% of the males and 40% of the
females snored. Of note was the observation that hyper
tension occurred more frequently in snorers than non-
snorers, suggesting that the former was more at risk for
cardiovascular complications.
Virtually every study of SAS patients notes the pre
sence of snoring in all these subjects (Sullivan and Issa,
1980; Block, 1980; Coverdale et al., 1980). This is
true for both predominantly obstructive, and predominately
central SAS subjects (Guilleminault et al., 1978). How
ever, the presence of heavy snoring in asymptomatic
(nonhypersomnolent) apneas is also well documented (Orr et
al., 1979; Fisher et al., 1978; Berry and Block, 1983),
as well as in otherwise normal subjects. These data may
be taken as evidence for snoring as a necessary, but not
sufficient, condition for the occurrence of sleep apnea.
The role of body weight in SAS is less clear.
Recall that Pickwickian patients are invariably obese, as
well as hypercapnic. Guilleminault et al. (1978) stress
that SAS may occur in nonobese as well as obese sub
jects. However, Block (1980) states that he has found SAS


67
was significant at p<.G5 or less: Number of desaturations
>4% vs REM latency (r=-.369).
Table 3-7 presents the sleep variable means for the
subjects grouped by level of apnea/hypopnea. The group
with high apnea/hypopnea appears to spend more time awake,
less time asleep, and achieve less slow wave sleep than
the others. However, none of these differences proved
significant. In fact, only one out of 15 oneway ANOVAs
achieved significance: Time in bed, F(2,40)=0.4, p=NS;
Sleep latency, F(2,40)=1.7> p-NS; Pure sleep time,
F(2,40)=C.3, p=NS; Sleep efficiency index, F(2,40)=0.2,
p=NS; Number of stage 0 periods, F(2,40)=0.5 p=NS; Time
% stage 0, F(2,40)=1.3 p=NS; Time % stage 1,
F(2,40)=1.3, p=NS; Time % stage 2, F(2,40)=0.04, p=NS;
Time % stage 3 F(2,40)=3.64, p<.C4; Time % stage 4
F(2,40)=1.32, p=NS; Time % stage REM, F(2,40)=1.5,
p=NS: Latency 1st REM, F(2,40)=0.6, p=NS; Mean REM
period length, F(2,40)=0.3> p=NS: Mean REM cycle length,
F(2,40)=0.9, p=NS; % slow wave sleep, F(2,40)=2.4,
p=NS. Scheffe followup procedures for time % stage 3
revealed no significant between group comparisons.
To summarize, EEG sleep measures indicatec lighter
sleep then normal in this sample, with increases in
awakenings and light slow wave sleep observed. Neither
between group comparisons nor correlational procedures
revealed significant interactions between sleep and
nocturnal respiratory variables.


64
asked wnether they had ever been diagnosed with hyper
tension. Thirty-three per cent of those with no events
had diagnosed hypertension, as did 33% of those with low
levels of apnea/hypopnea, while only 15# of those with
high levels of apnea/hypopnea had diagnosed hyperten
sion. These data do not support an increasing level of
hypertension in subjects with high levels of
apnea/hypopnea.
The subjects were grouped into the earlier delineated
levels of apnea/hypopnea. Table 3-5 presents the means
and standard deviations of health variables from the three
groups. No dramatic trends are apparent in this table,
and oneway ANOVA procedures confirmed this observation
with no significant differences for: Diastolic BP,
F(2,43)-*01, p-NS; Systolic BP, F(2f43).1, p=NS; CMI
overall, F(2,42)=.2, p=NS: CMI respiratory, F(2,42)=.5*
p=I\!S; CMI cardiopulmonary, F(2,42)=.5, p=NS; CMI
neurological, F(2,42)=2.2, p=NS.
In summary, while correlational procedures suggested
relationships between several oxygen saturation measures
and blood pressure as well as overall number of symptoms
endorsed on the CMI, oneway ANOVAs failed to reveal
significant between group differences when subjects were
stratified by level of nocturnal distress.


111
Table B-2.
Daytime sleepiness
males.
variables from
46 snoring
Subject
Mean Stanford
Sleepiness
Mean Sleep
Latency
Minutes of
Sleep
if
Rating
- Naps
- Naps
1
_
09.0
22
2
-
16.0
05
5
-
11.5
18
4
4.5
-
-
5
3.5
06.5
26
6
5.5
-
-
7
4.7
20.0
01
8
2.7
13.5
13
o
3.8
09.5
17
1C
4.3
20.0
00
11
3.4
20.0
00
12
3.3
16.0
09
13)
-
-
-
14
4.5
20.0
00
15
4.1
14.0
14
16
3.6
20.0
00
17
2.8
08.5
25
18
2.5
11 .5
18
19
-
15.C
11
20
3.6
20.0
00
21
3.7
20.0
00
22
2.4
14.5
05
23
4-5
14.5
12
24
4.4
20.0
00
25
3.3
04.0
34
26
3.3
-
-
27
4.8
19.0
03
28
3.5
20.0
00
29
4.1
15.0
10
30
2.6
17.5
07
31
4.3
14.0
13
32
3.1
11.0
18
33
5.1
11.5
18
34
3-1
-
-
35
3.8
20.0
00
36
3.4
20.0
00
37
3.4
13.5
14
38
2.2
20.0
00
39
2.9
06.5
25
40
3.9
19.0
08
41
4.0
C8.0
27


96
reported in similar studies. A first night effect pro
bably caused increased sleep latency, wakenings, and
lighter sleep, a result which may have functioned to
obscure possible relationships between sleep and respira
tory variables, which were not observed. A pattern of
correlations between increased respiratory disturbance and
sleepiness was noted, perhaps because these variables
reflected subjective reports which were largely indepen
dent of the experimental nights sleep (SSS ratings made
before sleep, sleep questionnaires reflecting trait sleep
variables, and sleep logs averaged over a 7 day period
following the experimental nights' sleep). Multiple night
recordings appear necessary to sort out a possible syner
gistic interaction between the first night effect and the
discomfort of the recording procedures.
I.ccturnrl hespratory cnc h'c-crophysi oloriccl Variables
The major finding of the respiratory/cognitive
analysis was the demonstration of relationships between
nocturnal respiratory parameters and cognitive/neuro-
psychological scores tapping non-verbal intelligence,
verbal and non-verbal memory, expressive verbal fluency,
and cognitive flexibility. Most of these relationships
remained when age, weight, and educational differences
were ruled out as plausible alternatives.
These findings extend the boundaries set by previous
work which indicated a relationship between hypoxia and
neuropsychologica1 scores. These previous reports


BIOGRAPHICAL SKETCH
David T. R. Berry was born on April 3, 1938, in
California. He attended high school in Montgomery, Alabama,
received a bachelor's degree in psychology from Auburn
University, and a master's degree in experimental psychology
from the University of Alabama in Birmingham. After
completing an internship in clinical psychology in Chicago,
he plans to seek a faculty position in psychology.
132


38
The normals averaged 20.4 m to Stage 1 sleep while the SAS
patients averaged 1.9 m. Zorrick et al. (1982) found mean
MSLT latencies to Stage 1 sleep of 3.1 m for SAS subjects
and 2.9 m for narcoleptics, significantly shorter than for
patients with psychiatric disorders.
The chief difficulty remaining in Judging the MSLT
"reliable and valid is external validation demonstrating
that exceeding a certain mean latency is correlated with
other deficits of performance. In some ways, low laten
cies on the MSLT may suffer the same limitations as sleep
deprivation protocols; that is, demonstrating serious
deficits is surprisingly difficult. At the same time, few
would deny the subjective unpleasantness of sleep depriva
tion, which may be applicable to low MSLT scores. Thus
the MSLT appears indicated in assessing excessive daytime
sleepiness while the Stanford Sleepiness Scale ( sSS) pro
vides an index of the subjective experience of the sub
ject. Application of both these measures to a subclinical
apnea population would seem to adequately assess daytime
sleepiness in these subjects.
Measurement of Hypoxia and Its Sequelae
Moctural desaturation is frequently noted in SAS
patients. Therefore, it seems in order to examine the
known sequelae of hypoxia of other etiologies in an
attempt to determine variables or behaviors likely to be
sensitive to the hypoxia found in SAS and SAA. Research
on hypoxia varies systematically along two dimensions,


4
syndrome typically results in complications such as right
sided congestive heart failure, pulmonary hypertension,
and peripheral edema (Block, 1980). An exacerbation of
disordered breathing (apnea) as well as hypercapnia and
hypoxemia (often severe) occurs with sleep onset. Treat
ment consists of oxygen therapy, respiratory stimulants,
and maintenance of wakefulness.
The discovery of a sleep induced exacerbation of the
symptoms accompanying Pickwickian Syndrome spurred a
systematic investigation of sleep respiratory events
(Gastaut et al., 1965). The new emphasis on noctural
respiration led to a series of investigations which docu
mented sleep induced respiratory dysrhythmias in non-obese
subjects (Lugarasi et al., 1966). A new group of sleep
disordered breathers designated as Sleep Apnea Syndrome
emerged, with several features which distinguished them
from Pickwickians. Among these differences were weight
(i.e., SAS patients were not necessarily obese), daytime
respiratory control (i.e., SAS patients were typically not
hypercapnic during waking as were Pickwickians), sex
(i.e., while A0% of Pickwickians were female, very few
premenapausal females have been identified with SAS), and
age (i.e, SAS patients are typically younger than
Pickwickians)(Block, 1980).
Two large samples of SAS patients have been reported
in the literature, allowing characterization of typical
SAS patients. Guilleminault et al. (1976) reported on a


54
included memory, visuo-organizational skills, verbal
fluency, and motor coordination. The chosen neuropsycho
logical battery reflected these two aims. Thus the WAIS
tapped intellectual functioning; the Wechsler Memory
Scale with delayed recall of logical stories and visual
reproductions as well as digit span and Rey Figure heavily
sampled verbal and non-verbal memory; the Hooper tapped
visuo-organizational skills; Verbal fluency sampled lan
guage skills; the Wisconsin card sort tapped frontal
sell-regulatory skill; and Finger tapping sampled motor
functions. The chosen battery provided a broad screening
as well as a detailed assessment of neuropsychological
functions thought to be impaired by hypoxia.
Procedure
Subjects followed the schedule appearing below for
the experimental night:
1800 Sign informed consent, blood pressure,
height/weight, finger tapping, verbal fluency,
subjective sleepiness (SSS-made every V2 hour
till bedtime)
1830 Wiring for EEG
1900 Begin HSLT I (20 m)
1930 WAIS
2000 VMS, Hooper
2030 Graphesthesia, Luria motor programs
2100 Begin MSLT II (20 m)
2130 Rey figure, Wisconsin card sort


14
normals might experience apneas during REM without
pathological consequence. Therefore, data available on
these two issues will be reviewed, including only studies
with a sample size great than 10 (with one exception).
Reports on normal subjects below 30 years of age are
somewhat rare, although some do exist. Bixler et al.
(1982) included a group aged 18-29 years (M = 13, F =
17). One subject in this group had sleep apneas (12) with
a mean duration of 15 seconds. Bixler et al. noted that
most apneas occurred in Stages 1 and REM sleep. Part of
the sample of Block et al. (1979), the females, had a mean
age of 29. Three of these subjects showed apneas with a
mean duration of 14 seconds. Again, a disproportionate
number of apneas occurred in Stages 1 and REM. Thus in
normal subjects under 30, apneas are somewhat rare, with a
mean duration of approximately 15 seconds when they do
occur. Most apneas occur in light and REM sleep.
More data are available on normal subjects between 30
and 60. Guilleminault et al. (1978) do not report on
duration of apneas in their sample aged 40-60, but they
note that apneas occurred only in Stages 1 and REM.
Bixler et al. (1962) reported that, in their 30-44 year
old group, 5 subjects experienced sleep apnea with a mean
duration of 13.1 seconds. A disproportionate number of
apneas occurred in Stages 1 and REM. Block et al. (1979)
studied 30 males with a mean age of 38, with 12
experiencing sleep apneas. These apneas had a mean


72
awakenings, F(2,43)=0.9 p-NS: Minutes of awakening,
F(2,43)=0,e, p=NS; Sleep latency, F(2,43)=0.01, p=NS;
Depth of sleep, F(2',43)=2.6, p*NS; Daytime sleepiness,
F(2,43)=1.8, p=NS].
Means were calculated from the 7 day sleep logs, and
the following variables were selected for analysis:
reports of daytime sleepiness, number of naps, minutes
napping, total bedtime, sleep latency, number of
awakenings, minutes of awakenings. These variables were
correlated with the respiratory indices to form a Pearson
correlation matrix totalling 49 correlations. Table 3-11
presents the 6 significant (p<.05) correlations which
emerged. Again, number of desaturations were associated
with reports of napping, while mean high saturation was
associated with total bedtime, sleep latency, and minutes
awake, and mean low saturation was associated with minutes
awake as well.
Table 3-12 presents the means and standard deviations
for the subjects divided into levels of apnea/hypopnea.
Oneway ANCVA procedures failed to reveal significant
between group differences [Daytime sleepiness,
F(2,35)-0.6, p=NS; Number of naps, F(2,35)=0.3, p=NS;
Minutes of napping, F(2,35)-1*3> p=NS; Total bedtime,
F(2,35)=0.7, p=NS; Sleep latency, F(2.35)0.2, p=NS;
Number of awakenings, F(2,35)=0.2, p=NS; Minutes of
wakenings, F(2,35)=.0001, p=NS].


127
Guilleminault, C., van den Hoed, J., & Mitler, M. (1978).
Clinical overview of the sleep apnea syndromes. In C.
Guilleminault & V. Dement (eds), Sleep Apnea
Syndromes. New York: Alan R. Liss Inc.
Guilleminault, C., Simmons, F., Motta, J., Cumminsky, J.,
Rosekind, M., Shroeder, J., & Dement, W. (1981).
Obstructive sleep apnea syndrome and tracheostomy.
Archives of Internal Medicine, 141, 985-988.
Guilleminault, C., Tilkian, A., & Dement, W. (1976). The
Sleep Apnea Syndromes. Annual Review oi Medicine,27,
465-470.
Halstead, W. (1947). Drain and Intelligence. Chicago:
University of Chicago Press.
Hanbauer, I., Karoum, F., Hellstrom, S., & Lahiri, S.
(1981). Effects of hypoxia lasting up to one monrh on
the catecholamine content in rat carotid body.
Neuroscience, 6., 81-86.
Harmon, E., Wynne, J., Block, A., & Mailoy-Fisher, L.
(1981). Sleep-disordered breathing and oxygen
desaturation in obese patients. Chest, 79, 256-260.
Hartse, K., Roth, T., & Zorrick, F. (1982). Daytime
sleepiness and daytime wakefulness: The effect of
instruction. Sleep, , 107-118.
Hartse, K., Zorick, F., Roth, T*, Kaffeman, M., & Hoyles,
T. (1980). Daytime sleep tendency in normal,
insomniac, and somnolent populations. Sleep Research,
9., 204.
Hartse, K., Zorick, F., Sicklesteel, J., Piccione, P., &
Roth, T. (1979). Nap recordings in the diagnosis of
daytime somnolence. Sleep Research, _6, 190.
Hebb, D. (1961). Distinctive features of learning in the
higher animal. In D. Delafresneye (ed), Brain-
Mechanisms and Learning. London: Oxford University
Press.
Hoddes, E., Dement, W., & Zarcone, V. (1972). The
development and use of the Stanford Sleepiness Scale.
Psychophysiology, 15C-160.
Hoddes, E., Zarcone, V., Smythe, H., Phillips, R., & Dement,
Vf. (1973). Quantification of sleepiness: A new
approach. Psychophysiology, 10, 431-436.


31
been matched on number of apnea events. Asymptomatic
(nonhypersomnolent) subjects fell from a baseline 02
saturation of 80 mm (Hg) to a mean maximum desaturation of
54 mm (Hg) during sleep. Symptomatic subjects fell from a
baseline of 54 mm (Hg) to a mean maximum desaturation of
35 mm (Hg) while asleep. Orr et al. suggested that the
more severe desaturation which accompanied apneas in
symptomatic subjects underlaid their somnolence and other
complications. However, these conclusions are vulnerable
to certain methodological criticisms, as Elock (1980)
points out. Specifically, periodic sampling may well fail
to detect the multiple brief desaturations associated with
SAS, rendering this technique potentially insensitive to
important events.
In an attempt to provide more representative data on
desaturation, Block et al. (1979) utilized an ear oximeter
in studying saturation in normal subjects. An ear oxi
meter provided a continuous, accurate readout of moment to
moment arterial oxygen levels. Thirty males (mean age 38)
and 19 females (mean age 27) were studied. Of the 30
males, seventeen suffered at least one episode of desatu
ration, falling from a mean saturation of 95% to a maximum
desaturation of 84% during sleep. In contrast, no epi
sodes of desaturation were noted in the sample of premena-
pausal females (baseline 02 saturation: 96%). The oxygen
desaturations of the males were always found in asso
ciation with breathing abnormalities or snoring. A study


107
aesaturaticns exceeding 4 and
relative ease of this rating,
saturation records.
10% was kept. because of the
only one rater scored


3-11 Significant (p<.05) Pearson correlations
between nocturnal respiratory and sleep log
variables in 46 snoring males 73
3-12 Means and standard deviations of sleep log
variables in 46 snoring males grouped by
level of apnea/hypopnea 73
3-13 Multivariate regression of demographic and
nocturnal respiratory variables on cognitive
scores in 46 snoring males 76
3-14 Significant (p<*05) partial correlations,
controlling for age, between nocturnal
respiratory and cognitive variables in
46 snoring males 79
3-15 Means and standard deviations of cognitive
scores in 46 snoring males grouped by
level of apnea/hypopnea 80
3-16 Multivariate regression of sleep and
nocturnal respiratory scores on cognitive
variables in 43 snoring males 83
B-1 Demographic and medical variables from 46
snoring males 109
B-2 Daytime sleepiness variables from 46 snoring
males 111
B-3 Sleep questionnaire variables from 46 snoring
males 115
B-4 Seven day sleep log means from 46 snoring
males 115
B-5 Neuropsychological variables from 46 snoring
males 117
B-6 Respiratory variables from 46 snoring males...,119
B-7 Sleep variables from 46 snoring males 121
vi


7
One sampling strategy which has been pursued is to
exclude subjects who complain of sleep or respiratory
difficulties. Screening for sleep complaints varies
widely. Guilleminault et al. (1978) studied a sample of
20 "normals" (presumably noncomplaining subjects). These
subjects ranged in age from 40-60 years. Males showed a
mean number of apneas of 7 (range 1-25), while females
showed a mean number of apneas of 2 (range 0-5). Thus,
none of these noncomplaining subjects presented a clinical
syndrome of sleep apnea, although sleep apneas did occur
apparently without pathology. Block et al. (1979) studied
49 subjects (M = 30, F = 19). Males in the sample had a
mean age of 38, while females had a mean age of 29. Any
potential subjects who complained of breathing difficul
ties or sleep disturbances were excluded from the study.
Twelve males (40$) had episodes of sleep apnea while only
3 females (15$) suffered as well. For those subjects,
males had a mean of 4.2 episodes, while females had a mean
of 3. Within this sample, the number of apneas was not
significantly correlated with increasing age, although
occurrence of oxygen desaturation (a concomittant of
disordered breathing) was positively correlated with
age. Additionally, although the mean number of sleep
apneas was not significantly different between sexes (M =
4.2, F = 3), the more sensitive measure of apneas with
desaturation was (M =3, F = 0). Webb (1974) studied 2C
males with a mean age of 44. Subjects were excluded from


77
Table 2,-15-continued.
Criterion
Predictor
Beta
Sig.
Wechsler Memory
Age
.0272
.852
Scale Memory
Weight¡Height Ratio
.0996
.546
Quotient
Apnea Index
-.3315
.024
Apnea+Hypopnea Index
-.3421
.049
Mean High Saturation
Number of
.2384
.124
Desaturations >4%
-.5037
.007
Delayed Recall
Age
-.3254
.018
Visual
V/eight:Height Ratio
-.1214
.416
Reproductions
Apnea Index
-.4037
.005
Apnea+Hypopnea Index
-.2026
.191
Mean High Saturation
Number of
.1368
.325
Desaturations >4%
-.2888
.078
Delayed Recall
Age
-.3550
.010
Rey Figure
V/eight ¡Height Ratio
.1795
.251
Apnea Index
-.2206
.091
Apnea+Hypopnea Index
-.0572
.709
Mean High Saturation
Number of
.3191
.025
Desaturations >_4%
-.3228
.049
Hooper Test
Age
-.5658
.000
V.eight:Height Ratio
.1202
.405
Apnea Index
-.1428
.251
Apnea+Hypopnea Index
.1478
.318
Mean High Saturation
Number of
.1072
.421
Desaturations >4%
-.2042
.190
\vi scon sin Card
Age
-.3409
.027
Sort
V/eight ¡Height Ratio
.0258
.887
Apnea Index
-.3205
.031
Apnea+Hypopnea Index
.1029
.550
Mean High Saturation
Number of
.2383
.130
Desaturations _>4^
-.0801
.657


Table B-
1. Demographic ana med
ical vai
r i a b 1 e s
from 46
snoring
males.
Weight:
Corneli He
dical index
Sub j.
bduca-
Height
Height
Blood
Hressure
Overa 1 i
Kes-
Cardio
heuro-
a
Age
tion
(in)
Weight
Hu tio
biastolic
Sys to!ic
score
pira t.ory
pulmonu ry
logica 1
i
58
12
00
250
5.67
130
090
2
00
12
U5
225
5.57
140
100
022
01
\j
02
5
52
10
07
148
2.10
110
065
020
01
00
01
4
bb
15
70
106
2.4
140
100
007
00
01
OO
5
52
12
74
245
3.31
120
085
018
03
00
U1
0
50
10
07
155
2.51
110
060
015
00
02
02
7
51
12
09
155
1.95
130
085
00/
oo
00
01
b '
4
09
07
155
2.31
120
090
005
01
02
00
y
00
20
O0
157
2.57
130
100
014
00
02
05
10
50
10
08
255
3.45
155
090
005
00
00
02
11
49
18
70
250
5.28
120
100
030
01
02
01
12
5b
15
74
250
5.10
125
090
UOO
00
01
01
15
48
15
70
187
2.67
120
070
007
00
01
01
14
58
15
75
190
2.55
120
070
005
01
00
-Cu
15
54
14
07
125
1.99
114
076
000
00
oo
01
1o
50
10
71
100
2.25
150
080
014
02
04
01
17
32
1b
72
107
2.31
120
085
005
01
00
02
1b
5b
1b
b9
180
2.00
120
090
0;50
02
01
01
19
57
12
72
190
2.71
120
090
015
01
02
01
2U
50
16
71
250
3.52
155
100
042
06
02
Ol
21
47
14
71
198
2.79
125
090
007
00
02
02
22
66
14
05
155
2.04
120
ObO
018
02
00
CO
25
55
16
72
145
2.01
120
074
013
01
01
01
24
51
10
72
105
2.29
115
07U
010
04
oo
02
25
51
15
00
100
2.35
122
004
025
00
01
04
2b
04
10
00
215
1o
149
100
u10
U1
C1
02
27
75
12
00
185
2.09
14u
090
020
0.-)
00
01


36
patients, i.e., response bias and altered frame of
reference (Dement et al., 1978).
It was clear then that an objective measure of day
time somnolence was needed. With this goal in mind,
Carskadon and Dement (1977) introduced the Multiple Sleep
Latency Test (MSLT). This technique involved measuring
the objective sleep latency during multiple nap attempts
throughout the day. In a typical paradigm, a subject
would be wired to an EEG and attempt to fall asleep at 2
hour intervals through the day. The latency to sleep in
each nap was measured by determining the latency to the
first epoch of Stage 1 sleep. In order to prevent signi
ficant amounts of sleep accumulating during testing, sub
jects were awakened after an epoch of sleep was
observed. If no sleep was noted, subjects were discon
nected after 20 minutes and allowed to resume ether acti
vities. Carskadon and Dement believe that this measures
underlying physiological sleep tendency. This underlying
tendency is thought to be modulated by alerting stimuli in
the environment (Carskadon and Dement, 1982a). The vali
dity of the test was established on normal subjects, who
exhibited a biphasic curve of sleep tendency throughout a
24 hour period. Extension of sleep over typical lengths
in normals (who are thought to be moderately chronically
sleep deprived) saw a significant rise in average sleep
latencies on the MSLT (Carskadon and Dement, 1979b), while
chronic sleep restriction (5 hrs sleep per night) caused


Table 3-7. Means and standard deviations 01 selected sleep variables from 43
snoring males grouped by level oi' apnea/hypopnea.
Mo Apnea/
Low Apnea/
High Apnea/
Hypopnea
Hypopnea
Hypopnea
Time in Bed
376.1 (45.3)
382.2
(47.4)
361.2 (38.0)
Sleep Latency
20.2 (12.7)
38.3
(43.5)
20.4 (18.7)
Pure Sleep Time
292.7 (73.5)
290.9
(81.5)
265.4 (69.5)
Sleep Efficiency Index
.77 (.17)
.70
( .16)
.73 (.14)
Number Stage 0 Periods
9.5 (7.6)
7.7
(3.2)
9.2 (5-5)
Time % Stage 0
15.8 (12.9)
11 .6
(12.4)
20.9 02.6)
Time % Stage 1
4.1 (3.7)
2.6
(1.8)
3.5 (2.0)
Time % Stage 2
61.3 (8.0)
60.3
(12.9)
61.3 (1.3)
Time % Stage 3
2.8 (1.9)
4.3
(2.5)
1.5 (2.3)
Time % Stage 4
8.5 (7.2)
8.5
(5-6)
3.6 (5-1)
Time % Stage REM
9.7 (6.C)
13.3
(7.8)
9.1 (7.6)
Latency 1st REM Period
136.5(106)
107.5
(71.5)
156.4(125.4)
Mean REM Period Length
13.9 (0.8)
16.1
(9.8)
13.5 (11.1)
REM Cycle Length
80.0 (56.1)
78.7
(61.7)
41.6 (42.7)
% Slow Wave sleep
12.8 (8.0)
14.6
(7.4)
5.9 (0.1)
Note: see text for explanation of N.


23
I had >100 apneas per night, Group II had >30 but <100
apneas per night, while Group III had less than 20 apneas
per night. Thus Groups I and II were diagnosable as SAS
(>30), while Group III was not. Obesity was present in
Groups I and II, but in only 3 subjects in Group III.
However, these differences were not statistically signifi
cant.
Taken together, this evidence suggests a positive
relationship between weight and apnea in the general popu
lation. However, SAS patients are not invariably obese.
Further, weight loss fails to result in a decrease in
apneas in many SAS patients. Thus the relationship
between obesity and full blown SAS is less than clear.
The preceding pages have investigated the currently
understood role of several variables (sex, age, snoring
and weight) in sleep apnea activity. Presently available
data suggest that in young and middle aged subjects, apnea
is a predominantly male phenomenon, although post mena-
pausal females may close the gap. Increasing age is re
lated to sleep apnea activity. Snoring, which occurs in a
significant percentage of the general population, is
thought to be a necessary but not sufficient condition for
SAS. Weight was shown to have a positive relationship to
sleep apnea activity in the general population, although
its presence or role in SAS is equivocable.
These conclusions have important implications for
experimental designs directed at understanding sleep apnea


REFERENCES
Agnev, H.W., & Webb, W.E. (1972). Measurement of sleep
onset by EEG criteria. American Journal of EEG
Technology, 12, 127-134.
Agnew, H.V., Webb, W.B., & Williams, R.L. (1966). The
first night effect: An EEG study of sleep.
Psychophysiology, _2, 263-260.
Ancoli-Israel, S., Kripke, D., Mason, V., & Kaplan, W.
(1982). A prevalence study of sleep disorders in
seniors: Six-months results. Abstract published for
1982 APSE meeting, San Antonio.
Ancoli-Israel, S., Kripke, D., Mason, W., & Messin, S.
(1981). Sleep apnea and nocturnal myoclonus in a senior
population. Sleep, _4(4), 349-338.
Annau, Z. (1972). The comparative effects of hypoxic and
carbon monoxide hypoxia on behavior. In B. Weiss & U.
Laites (eds), Behavioral Toxicology. New York: Plenum
Press.
Appel, D., Schmidt-Nowara, V., Poliak, C., & Wietzman, E.
(1982). Effects of tracheostomy closure on sleep and
breathing in sleep apnea patients with long term
tracheostomy. Abstract published for APSS meeting 1982,
San Antonio, p. 131.
Association of Sleep Disorder Centers (1979). Diagnostic
classification of sleep and arousal disorders. Sleep,
2_, 1-137.
Bartus, R., Dean, R., & Bear, B. (1983). The cholinergic
hypothesis of impaired memory in the aged. In D. Harman
& K. Gray (eds), Aging: Nutrition anc Aging. New
York: Raupa Press.
Berg, I. (1948). A simple objective technique for
measuring flexibility in thinking. Journal of General
Psychology, 39 15-22.
Berry, D., l. Webb, W.B. (In Press). Sleep and cognitive
variables in normal aging subjects. Journal of
Gerontology
123


Table l-6
Respiratory variables from 46
Subject Apnea
ij_ Index
1
0.0
2
1.7
3
0.5
4
2.3
5
0.4
6
3*4
7
0.2
8
0.0
9
3*6
10
0.5
11
0.0
12
0.3
13
13.0
14
0.8
15
0.2
16
0.8
17
0.2
10
0.2
19
0.0
20
0.7
21
0.0
22
0.0
23
0.0
24
0.0
Apnea+
Hypopnea
Index
Seconds
in
Lvents
38.4
4058
4.2
350
0.3
36
2.5
144
0.9
44
3*7
500
0.2
10
0.0
0
4.8
522
0.5
10
0.0
0
0.9
109
13*0
1498
0.8
36
0.2
23
0.8
48
0.2
21
0.2
10
0.0
0
0.7
40
0.2
20
0.0
0
0.0
0
O.C
0
rin£ males
Rumber of
Mean
Saturation
Desatura
tions
Rifh
Low
>4%
>105-
93*4
80.9
220
73
93*4
88.7
214
21
94.9
93.0
26
0
96.9
94.0
74
2
95*6
92.0
123
2
95.8
91.7
90
7
95.6
91.7
5
0
93.6
94.9
0
0
98.4
95-4
145
9
95.0
94.2
2
0
91.6
90.0
22
0
95.1
92.1
148
3
94.3
90.0
3
0
95.1
93.7
10
0
95.6
94.5
4
0
95.1
94.1
2
0
95.4
94.1
7
0
94.3
93.2
0
0
96.5
95.1
18
0
95.1
93.3
49
0
94.7
92.3
60
2
94.9
94.3
1
0
95.4
94.9
0
0
92.8
92.0
1
0
\D


69
Daytime Sleepiness Data
Included among the measures of daytime sleepiness
were the mean of the several Stanford Sleepiness Scale
ratings, and two variables calculated from the two evening
naps: mean sleep latency and total minutes asleep during
the naps. These variables were correlated with the
respiratory variables, resulting in a Pearson correlation
matrix. From these 21 correlations, two emerged signi
ficant at p<.05 or less, both involving the mean Stanford
Sleepiness Scale ratings. Mean SSS ratings were corre
lated wixh apnea index (r=.297)* and apnea + hypopnea
index (r=.283).
Table 3-6 presents the means and standard deviations
for the subjects grouped by level of apnea/hypopnea. One
way ANOVA procedures were non-significant for all these
variables [SSS: F(2,38)=2.4, p=N5; Mean sleep latency:
F(2,35)-0.4, p=NS; Total minutes asleep: F(2,35)=0.3,
p=NS].
Table 3-8. Means and standard deviations for daytime
sleepiness variables in 46 snoring males grouped
by level of apnea/hypopnea.
No Apnea/
Low Apnea/
High Apnea/
Hypopnea
Hypopnea
Hypopnea
Stanford Sleepiness
Scale
3.2 (.7)
3.6 (.e)
4.0 (.7)
Mean Nap Latency
Total Minutes of
14.5 (5.7)
15.5 (5.3)
13.4(3.4)
Sleep in Naps
11.3(11.9)
9.6(10.6)
13.5(6.6)


57
The respiratory records were annotated by two hour
intervals following bedtime. From a total of 652 events
scored, 150 (23%) occurred in the first two hours
following bedtime, 185 (23%) occurred in the second two
hours, 232 (36%) occurred in the third two hours, while 83
(12%) occurred in the final two hours. It should be noted
that vakeup times varied considerably, making interpre
tation of the attenuated number of events in the final two
hours somewhat problemmatic.
From a total of 578 events in which reliable sleep
staging was achieved, 459 (78%) occurred in light slow-
wave sleep (stages 1-2), while 118 (20%) began in REM
sleep. Only 6 events (01%) occurred in slow wave sleep.
These precentages must be considered in light of the
relative distribution of sleep stages in these subjects:
stages 1-2 (64%), stage REM (12%), stages 3-4 (10%), stage
0 (14%). This distribution suggests that the sleep time
spent in stages, 1, 2, or REM must have provided the
overwhelming preponderance of events in these stages.
Sleep characteristics will be discussed in greater detail
below.
As noted below, 28 of the 46 subjects experienced at
least one episode of apnea or hypopnea. Table 3-2
presents the data on occurrence of at least one
apnea/hypopnea, as well as the occurrence of a high level
of apnea/hypopnea. Almost 2/3 of the sample had ax least
one apnea or hypopnea, while 13% had high levels of


94
correlations in this sample. Additionally, an EEC scoring
system which averages across one minute periods might
obscure brief sleep fragmentation from nocturnal events.
Although the lack of correlation is puzzling, Bixler et
al. (1982) reported similar findings. In their normal
sample, few apnea induced arousals were noted. Multiple
night recordings of all these parameters appear necessary
to clarify the possible interaction o respiration and
sleep fragmentation in a first night effect.
Daytime Sleepiness, Sleep Questionnaire, and
Sleep Lop Data
The three assessments of subjective aspects of sleep
and sleepiness will be discussed together, as the pattern
of findings lend themselves to this organization. The
core finding here is a relationship between nocturnal
respiratory variables and sleepiness variables manifesting
itself in two forms. Daytime sleepiness, reflected in the
mean SSS ratings and self report of napping on both the
sleep questionnaire and logs, was positively related to
various indices of nocturnal events (apnea index, apnea +
hypopnea index, seconds in events, and number of
desaturations). A second aspect of this relationship
involved relationships between decreased saturation and
reports of decreased sleep latency and nightly wakenings
on the sleep logs and questionnaires.
These seemingly disparate findings may be interpreted
as the sequelae of sleep oisruption resulting from


60
In summary, non-parametric analyses failed to reveal
significant relationships between age or weight and
overnight respiration. However, Pearson correlations
suggestea that increasing weight is associated with
deterioration of several noctural respiratory indices.
Because increasing apnea/hypopnea levels have been
suggested to carry increasing risks for various deficits,
the sample was stratified by level of apnea/hypopnea. The
first group included those subjects without events (Ho
events), those with at least one event but less than 5
events per hour (Low apnea/hypopnea), and those with 5 or
more events per hour (High apnea/hypopnea). It should be
noted that this division of subjects is slightly different
from the earlier non-parametric analyses, because in this
grouping low apnea/hypopnea and high apnea/hypopnea are
mutually exclusive, while in the earlier analyses the
group with at least one apnea/hypopnea included those with
more than 5 events per hour.
Table 3-4 utilizes oneway ANOVA procedures to compare
these groups on demographic and nocturnal respiratory
variables. There were no significant between group
differences on any of the demographic factors: age,
weight, weightiheight ratio, and education [F(2,43)-2.1,
0.6,0.8,0.7; p=US]. Logically, both apnea and hypopnea
indices were significantly different between groups
[F(2,43)=9-6, p<-C004; 20.5, pC.0000]. Followup
comparisons (Scheffe) indicated that the high


50
Co). All physiological variables were recorded on the
polygraph's chart paper which was run at a speed of 1C
ram/s.
Measures
Two classes of independent variables are identified,
with the first being thumistor variables (those derived
primarily from the nose and mouth thumistors). These
included number of apneas and hypopneas, mean duration of
apneas and hypopneas, mean low oxygen saturation in apneas
and hypopneas, mean oxygen saturation change in apneas and
hypopneas, and total seconds spent in apneas and hypo
pneas. Thumistor variables were separately scored by two
trained raters, whose agreement appears in Table 2-1. An
overall agreement of 85# was achieved. Disagreements were
resolved by consensus scoring. The second class of
independent variables included oxygen saturation variables
(derived from a minute by minute rating of highest and
lowest oxygen saturation from the overnight record).
These variables included mean highest saturation, mean
lowest saturation, and number of desaturations of 4 ana
10# or more. In selecting these scoring criteria, several
considerations were relevant. Initially, a system with
reasonable interrater reliability was necessary. While
published criteria for scoring apneas were not problem
atic, reliable scoring of hypopneas was not achieved until
a desaturation criterion of 10# was introduced. In
selecting measures of oxygen saturation, a mean high and