Physiological responsivity to venipuncture and speech giving in insulin-dependent diabetic adolsecents at two levels of ...


Material Information

Physiological responsivity to venipuncture and speech giving in insulin-dependent diabetic adolsecents at two levels of diabetes control and their nondiabetic peers
Physical Description:
vii, 86 leaves : ill. ; 29 cm.
Gilbert, Brenda, 1947-
Publication Date:


Subjects / Keywords:
Diabetes Mellitus, Type I -- Adolescent   ( mesh )
Stress, Psychological -- Adolescent   ( mesh )
Adaptation, Physiological   ( mesh )
Clinical and Health Psychology thesis Ph.D   ( mesh )
Dissertations, Academic -- Clinical and Health Psychology -- UF   ( mesh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph.D.)--University of Florida, 1985.
Bibliography: leaves 81-85.
Statement of Responsibility:
by Brenda Gilbert.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000535807
oclc - 16878052
notis - ACV8816
sobekcm - AA00004874_00001
System ID:

Full Text







This is dedicated to my husband, David Gilbert,

and our two beautiful daughters,

Aline Marie, aged 11, and Elizabeth Ann, aged 3.


Special thanks are given to all my committee members

which include Hugh Davis, Ph.D., Randy Carter, Ph.D., James

Johnson, Ph.D., and Barbara Melamed, Ph.D., who have given

steady support and excellent consultation. Barbara

Melamed, Ph.D., and Peter Lang, Ph.D., provided me with

much needed direction in the design of this research and

the selection and analyses of the physiological data.

Janet Silverstein, M.D., Michael Kappy, M.D., and their

coworkers were indispensable in helping me understand

diabetes and select and analyze the metabolic data

collected. The study could not have been accomplished

without the great assistance of my fellow students and

coworkers who helped collect the data. These include Gary

Geffken, Ph.D., Marika Spevack, M.S., Carol Lewis, M.A.,

and Barbara Walker.

Extra special thanks are given to Suzanne B. Johnson,

Ph.D., my "wonderful" chairperson. Her excellent direction

and support were essential to the completion of this

dissertation. In addition, extra special thanks are given

to my husband, David G. Gilbert, Ph.D., who paid my bills

while I worked on my degree and provided steady,

unflinching support.




ACKNOWLEDGMENTS................. ........................ iii

ABSTRACT ... ..... ................ ....... ............ vi


1 INTRODUCTION ...... ........................... I

Major Hypothees.................................7
Exploratory Investigations......................9

2 METHODOLOGY...................... .. .... ..... ....15

Participants. .... ..................... ... ...... 15
Stress Manipulation Task.....................17
Speeches .............................. ......... 17
Major Dependent Measures ...................... 18
Additional Dependent Measures..................24
Procedure.................. ....... .... ..... 26

3 RESULTS ............................ .......... .... 29

Description of Sample...........................29
Duration of Diabetes and HA1 Values.............29
Time of Day and Location of Data Collection....30
Reliability of Observation Measurement..........30
Subject-Parent LEC Correlations and T-Tests....30
Inter-relationship Between Measures............ 31
Control Variables................................32
Analyses of Major Hypotheese ...................34

4 DISCUSSION AND SUMMARY......................... 46

Reported and Observed Anxiety of Tasks..........47
Physiological Response to Tasks................47
Metabolic Reactivity..............................52
Life Stress and Diabetes Control..............54
Personality Findings .......................... 57
Future Research................................58
Implications. .... ................ ......... ....60


A SPEECH TOPICS. .............................. 62

B HEART FUNCTIONING...... ............ .............63


D VENIPUNCTURE QUESTIONNAIRE............ ........ 67


MODIFIED FORM........................ ........... 69

G RANK ORDER OF TASK FORM......................74




BIBLIOGRAPHY... .............. ...... ......... .. ........ .81

BIOGRAPHICAL SKETCH... ............... ................... 86

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




August, 1985

Chairperson: Suzanne B. Johnson, Ph.D.
Major Department: Clinical Psychology

Fifteen adolescents with insulin-dependent diabetes in

good diabetes control were yoked with 15 insulin-dependent

adolescents in poor control matched for age, sex, duration

of diabetes, and race. The same number of nondiabetic

adolescents matched for age and sex were included.

Participants were involved in three stressful tasks

(venipuncture and two speeches). Each task was preceded by

a rest period and followed by a recovery period. Both

speeches were preceded by a plan period. Observational,

physiological (heart rate, skin conductance, blood and

urine measures), and self-report data were collected. Life

stress and personality information were collected.

Diabetic adolescents in poor control had higher heart rates

across all conditions but no differences in skin

conductance were found. Diabetic adolescents had less

desirable blood and urine outcomes compared to the

nondiabetic youth with the adolescents with poor diabetes

control having the least desirable outcomes. No

differences in life stress between groups were found.

Adolescents with well-controlled diabetes were less

neurotic than the nondiabetic adolescents.



A substantial number of people with insulin-dependent

diabetes have difficulty adequately controlling this

disease. They feel sick, miss school or work and have

serious problems carrying out normal activities. One very

serious, even life threatening, consequence of poor control

is ketoacidosis (Cahil, Etzwiler & Freinkel, 1976).

Repeated episodes of ketoacidosis are associated with

retinopathy and kidney failure. The increased incidence of

poor control and ketoacidosis in 12-18 year olds is well-

documented (Fallstrom, 1974; Koski & Kumento, 1975). The

relationship between the psychosocial and physiological

changes of adolescence and poor control in youngsters with

insulin-dependent diabetes is not entirely clear. The

question of what contributes to the onset and maintenance

of poor control and associated ketoacidotic symptoms in

some adolescents while others remain healthy is important

in our efforts to successfully manage this chronic illness.

Stress has been frequently implicated in cases of poor

diabetic control. Patients with insulin-dependent diabetes

may be particularly susceptible to stressful events because

they lack insulin, a hormone that counters other "stress"


hormones. More specifically, in both diabetic and

nondiabetic persons, stress results in increased levels of

catecholamines (epinephrine and norepinephrine). When

catecholamines are released into the blood a complex series

of events occurs. First, gluconeogenesis is stimulated

which increases blood glucose. Second, catecholamines act

directly on fat cells to increase lipolysis (fat breakdown

and mobilization of free fatty acids (FFA)).

Catecholamines also result in increased glucagon, which, in

turn, stimulates gluconeogenesis and ketogenesis in the

liver (Tarnow & Silverman, 1981-82).

Once the stress is over, there is typically an

increase in insulin production which "counters" the stress

hormones and permits the body to return to a normal

metabolic state. However, the youngster with diabetes does

not produce his own insulin and may not be able to counter

the effects of the stress hormones. Although exogenous

insulin replacement is helpful, the youngster is still left

with a system insensitive to rapidly changing stress

related blood glucose or ketone levels. When this system

is unable to effectively counteract the stress hormones,

ketoacidosis may result. Ketones (B-hydroxybutyric acid

and acetone) are produced in the liver from fatty acids

(fatty acids -> acetyl-Co A -> acetoacetyl-Co A -> B-

hydroxybutyric acid and acetone). Ketones provide a source

of energy, but in excessive amounts produce a low plasma

pH, acidosis. This results in rapid deep breathing,

hypotension, and ultimately coma. Ketoacidosis is also

associated with hyperglycemia, osmotic diuresis, with

electrolyte and fluid loss, vomiting and dehydration.

Sodium is markedly depleted in circulation along with a

lowered total body potassium (Ganong, 1971).

The role of epinephrine and norepinephrine in

stimulating free fatty acid production is supported by the

work of Baker, Barcai, Kay, and Hague (1969) and Pinter,

Peterfy, Cleghorn, and Pattee (1967). Pinter et al. (1967)

demonstrated that FFA levels may be increased by both

exposure to stress (i.e., an anxiety-provoking suggestion

to hypnotized subjects and spontaneous speeches produced by

subjects) and exogenous epinephrine administration. Using

a single case design, Baker et al. (1969) found increased

FFAs during a stressful interview compared to a nonstress

period in an adolescent with insulin-dependent diabetes.

The above two studies by Pinter et al. and Baker et al.

report that the administration of propranolol (a B-

adrenergic blocking agent) in the stressful situation

blocked the bulk of the increase in FFAs. However, Baker

et al. noted that the therapeutic effects of adrenergic

blocking agents with poorly controlled insulin-dependent

diabetic adolescents appear to be short lived. Although

this treatment was helpful for a temporary period of time,

eventually the problems with diabetic control returned

(Minuchin, Rosman & Baker, 1978).

Only a few studies have compared the effects of stress

in diabetic and nondiabetic subjects. Hinkle and Wolf

(1952) assessed nondiabetic and diabetic adult and

adolescent responses to a stressful interview and a

nonstressful control period. Both groups showed similar

responses in blood ketone production and urine output.

However, diabetics with elevated ketone levels in the

nonstressful control period exhibited a particularly

exaggerated increase in ketones when stressed.

Vandenbergh, Sussman, and Titus (1966) performed a similar

study comparing diabetic and nondiabetic adults' reactions

to an unpredictable shock. Both groups responded with

increases in FFA levels and urine volume, although the

increases were not significantly different between stress

and nonstress periods. However, this lack of difference

may have been a result of their small sample size (i.e.,

n = 6 in each group).

The work of Hinkle and Wolf (1952) and Vandenburgh et

al. (1966) suggest that increased free fatty acid

production is a likely result of stress. However, a number

of questions remain. First, it is unclear whether the

responses of insulin-dependent patients are different from

subjects having adult-onset diabetes. Most of the subjects

studied were adults, not adolescents, and the effects of


diabetes type were not analyzed. Second, although there

was some suggestion that patients in poor control may be

more stress sensitive, this was not explicitly studied.

Consequently, it is unclear whether adolescents with

insulin-dependent diabetes differ from normals in their

metabolic response to stressful stimuli and whether

adolescents in good versus poor diabetic control differ as

well. Finally, no objective or subjective measure of

stress was collected in either study. Since an external

stressor may have different effects on different subjects,

it is difficult to assess how many subjects felt stressed

and to what extent.

Within the diabetes literature, there is repeated

mention of youngsters who have "brittle" diabetes. These

patients' (who are often adolescents) diabetes is very

difficult to manage, and they have numerous episodes of

ketosis. Minuchin et al. (1978) have suggested that there

is a subgroup of youngsters who have "psychosomatic"

insulin-dependent diabetes. These youngsters show

excessive reactivity to stress which results in a "brittle"

condition. Support for this position is found in a study

by Minuchin et al. (1978) in which psychosomatic insulin-

dependent diabetic adolescents were compared to two groups

of good control insulin-dependent diabetic adolescents (one

composed of normal adolescents and the other of adolescents

referred for psychiatric treatment of behavioral

problems). Each group observed their parents discussing

unresolved family problems and later joined their parents

in this discussion. A major finding in this research

endeavor was that in the psychosomatic group the

adolescents with diabetes produced higher levels of FFA

which took longer to return to baseline when compared to

youngsters in the other groups. The psychosomatic

adolescents also produced higher FFA levels than their

parents which remained elevated after their parents' FFA

levels had returned to the baseline level (Minuchin et al.,

1978). No similar difference was found for the normal or

behavior problem diabetic groups.

The findings of Minuchin et al. (1978) support the

notion that a subgroup of insulin-dependent diabetic youth

have exaggerated response patterns to stress. Their

findings suggest that there are no major metabolic response

differences between well-controlled insulin-dependent

diabetic youngsters and their parents. However, only a

small number of patients were studied and the criteria for

placement in study groups (psychosomatic, normal, behavior

problem) was not clearly specified. Consequently, it is

unclear how many youngsters in poor diabetic control have

the "psychosomatic" or heightened stress reactivity that

Minuchin et al. postulate.

To summarize the main points made thus far, stress

exposure leads to metabolic changes associated with insulin

efficiency in both nondiabetic and diabetic persons. These

metabolic changes are related to sympathetic increases of

epinephrine and norepinephrine production although other

body hormones and transmitter substances are involved. The

normal person has finely-tuned metabolic processes which

act to keep their metabolic responses within normal

limits. However, due to the insulin-dependent diabetic

individual's inefficient insulin response capability, more

extreme metabolic responses may become more likely.

Evidence to date suggests that at least some subgroups of

diabetics may be more metabolically reactive than other

groups, although the relationship between metabolic

responsivity and control level has not been clearly

specified. There is less evidence to suggest clear-cut

response differences between diabetics as a group and

nondiabetics, although this hypothesis has received little

research attention.

Major Hypotheses

The primary purpose of the present investigation was

to study the effects of stress on youngsters with insulin-

dependent diabetes. The stress responses of youngsters

with well-controlled diabetes were compared to youngsters

in poor control and both diabetic groups were compared to

nondiabetic adolescents. Self-report, behavioral,

psychophysiological, and metabolic effects of stress were

assessed. Few differences on any of the measures were

expected between the well-controlled diabetic youngsters

and their nondiabetic counterparts. In contrast, those

youngsters in poor control were expected to show greater

psychophysiological and metabolic reactivity to stress than

either of the two other groups.

The study's hypotheses were as follows:

1. Compared to well controlled diabetics or

nondiabetics, insulin-dependent diabetic

adolescents in poor control who are stressed in a

laboratory setting will exhibit (a) heightened

metabolic reactivity, (b) heightened

psychophysiological reactivity, and (c) slower

rates of psychophysiological recovery subsequent to

the stressful experience.

2. Well controlled diabetics will show increased

metabolic effects to stress compared to nondiabetic


3. Few differences between groups are expected on the

self and behavioral measures of stress and

anxiety. However, should differences exist they

should be between the poorly controlled diabetic

youngsters and the other two groups. If youngsters

in poor diabetic control are more stress-reactive,

they may acknowledge greater stress and appear more

anxious to an observer.

The present investigation differs from past attempts

to study the effects of stress on persons with diabetes in

a number of important respects. First, only insulin-

dependent adolescents were studied. Second, distinctions

between those in good versus poor control were made.

Third, in both groups of youngsters with diabetes the

youngster and parent confirmed that the prescribed insulin

dose was given the night before and morning of the

experiment. Fourth, self-report, behavioral, and

psychophysiological effects of the stress were measured.

In past research efforts, no attempt has been made to

quantify the stress experienced by the subject either

through subjective ratings or by more objective behavioral

or psychophysiological measurement. And finally, a sample

size of fifteen subjects per group was obtained.

Exploratory Investigations

In addition to the major purposes and hypotheses

outlined previously, this study explored two other

variables potentially related to diabetes control. These

are life stress and the personality dimensions of

extraversion and neuroticism.

Life Stress

Evidence has accumulated suggesting that diabetic

adolescents with high scores on a life stress/change scale

or who have lost a parent show increased ketoacidosis and

related symptoms (Chase & Jackson, 1981; Koski & Kumento,

1975). Bradley (1979) found that the number of stressful

life events was associated with diabetes control in

adults. Furthermore, the insulin treated group had higher

levels of diabetes disturbance (glycosuria, prescription

changes, and clinic visits) compared to the tablet treated

group, although there was little difference between

reported levels of life stress in the two groups.

These findings support the hypothesis that life stress

is associated with diabetes control particularly in

insulin-dependent diabetics.


Hans Eysenck (1967) has postulated two basic

dimensions of personality (introversion-extraversion and

stability-neuroticism) based on biological inheritance and

its interaction with environmental learning (Eysenck,

1967). His neuroticism dimension measures emotional

responsivity and reactivity and he suggests that this

dimension may be involved in psychosomatic disorders

(Eysenck, 1967). Persons high on neuroticism have low

tolerance for stress, whether physical or psychological,

tend to avoid stressful stimuli, and are "overly emotional,

reacting too strongly to all sorts of stimuli, and find it

difficult to get back on an even keel after each

emotionally arousing experience" (Eysenck & Eysenck, 1975,

p. 5). Furthermore, Eysenck (1967) suggests that whereas

introversion-extraversion is associated with cortical

functioning/arousal, neuroticism is associated with

autonomic arousal which is mediated by the limbic-

hypothalamic brain area (Eysenck, 1967).

Eysenck's personality theory and related research

suggest that introversion also may be associated with

psychosomatic proneness. The unstable introvert, due to

his introverted propensities, is more aware of subtle

stimuli earlier and the stimuli are more "amplified" in the

sense that the introvert is more cortically aroused by

it. One may think of the introvert as "geared to inspect"

stimuli. This earlier greater awareness of a threatening

stimulus should increase autonomic responsivity.

Furthermore, Gray (1975) has modified Eysenck's model of

extraversion-introversion by suggesting that introverts are

more sensitive to particular aversive stimulation, i.e.,

punishment and frustrative-nonreward and thus more

conditionable to situations involving punishment or

frustration-nonreward. Both Eysenck and Gray agree that

trait anxiety as measured by the Taylor Manifest Anxiety

Scale and the Spielberger Trait Anxiety Scale are

correlated with Eysenck's neuroticism and introversion

scales. In fact, Gray (1975) suggests that a 45 degree

rotation of Eysenck's factors would provide more relevant

dimensions although not independent factors. The new

factors could be labeled trait anxiety and impulsivity.

Regardless of the theoretical differences between these

men, both agree introversion should predispose an

individual toward increased emotional responding; Eysenck

via increased cortical arousal leading to earlier awareness

of more subtle threatening cues, and Gray via a

biologically increased sensitivity to aversive and

frustrative-nonreward cues. Gray's issue is not with the

validity of Eysenck's scales, but with the choice of factor

rotation and the nature of the underlying biological


In summary, high introversion and neuroticism may

predispose insulin-dependent diabetics toward higher

autonomic arousal and slower return to baseline. This

autonomic over-reactivity may predispose them toward more

diabetic control problems. It is interesting to note that

the two personality descriptions (from parental ratings)

that Simonds (1977) found to differentiate well controlled

from poorly controlled diabetic youth were anxious and

depressed. These are the same personality trait terms that

Eysenck (1967) uses to describe an introverted neurotic or

dysthymic personality. Due to the strong association

between sympathetic reactivity and neuroticism hypothesized

by Eysenck, it is reasonable to expect that if this

relationship exists it would show up in a diabetic

population where sympathetic influences may be magnified by

an easily disrupted metabolic system.

Although Eysenck's personality theory has not been

applied to problems of diabetic control in insulin-

dependent youth, there is extensive literature assessing

psychophysiological reactivity in introverts and neurotics

compared to other control groups. For example, Stelmack

(1981) concluded that differences in electrodermal activity

between introverts and extroverts has support with

electrodermal activity generally greater for introverts and

electrodermal habituation faster for extroverts. In a

sample of college women, Harvey and Hirshmann (1980) found

significant heart rate differences for introverted-neurotic

and extraverted-stable groups who viewed slides of violent

death. The introverted-neurotic groups exhibited heart

rate increase whereas their counterparts showed heart rate


Some additional evidence exists suggesting a

relationship between neuroticism and reported illness.

Denny and Frisch (1981) found neuroticism to be a predictor

of self-reported illness in two samples of college

students. Akerstedt and Theorell (1976) found increased

physical complaints from neurotic vs. stable railway

workers (utilizing Eysenck's personality scale) who were

switched from day to night shifts. (Eysenck's scale was

administered prior to the shift change.) Less direct

evidence for a relationship between neuroticism and illness

comes from a study by Mehrabian and Ross (1977). These

authors utilized a stimulus screening questionnaire (high

stimulus screening theoretically was related to low

arousability) devised by Mehrabian and Ross (1977), which

correlated negatively (r = -.54) with Eysenck's measure of

neuroticism. The high arousability group reported more

psychosomatic health complaints and nonrecurring illness.

In summary, the following exploratory hypotheses were


1. Adolescents in poor diabetic control will have

higher self-reported levels of life stress than

diabetic adolescents in good control or normal

nondiabetic youth.

2. Adolescents in poor diabetic control will have

higher scores on Eysenck's Neuroticism and

Introversion Scales than diabetic youngsters in

good control or nondiabetic normals.

3. Subjects scoring high on Eysenck's Neuroticism and

Introversion Scales will show heightened

psychophysiological reactivity to the laboratory

stresses and slower habituation than subjects

scoring low on these scales.



Participants consisted of adolescents with insulin-

dependent diabetes and nondiabetic adolescents aged

11-18. Diabetic participants were obtained from lists of

patients treated through the North Florida Regional

Diabetes Program located at the J. Hillis Miller Health

Center, Gainesville, Florida; the University of South

Florida Diabetes Program located in Tampa, Florida; and

lists of campers attending a summer camp for diabetic

youngsters run by these two programs. The nondiabetic

youth were recruited from aged 12-18 students at the P.K.

Yonge School associated with the University of Florida and

through staff at the J. Hillis Miller Health Center.

Subjects were contacted based on their hemoglobin Al

(HA1) values, a physiological measure used to assess

diabetes control over several weeks time (Tarnow and

Silverman, 1981-82). The HA1 is a measure of the amount of

glucose adhering to hemoglobin in the blood and reflects

amount of blood glucose over a period of time. In this

study, adolescents in good diabetic control had HA1 values

of 12 or less and those in poor control had values 15 or

over. These cutoff scores are based on Harkavy's (1981)

study in which similar scores discriminated good and poor

control as defined by diabetologists' ratings. The HA1

value obtained at the time of the study served as the final

criterion and in three cases a participant whose HA1 was

slightly above 12 was kept as a good control subject if

his/her match had a value above 16.

All adolescents having diabetes at least one year and

meeting the HA1 criteria for good control were asked to

participate if a potential poor control match could be

identified. Participants were matched on sex, age, race,

and duration of diabetes. Matched subjects had to be of

the same sex, within 2 years of age, and of the same ethnic

group (except in one case where a white male was

substituted for a black male). Poor control subjects could

not have had diabetes more than 1 year longer than their


Each potential subject and his/her parent indicated

that the recommended insulin dose was regularly

administered and specifically acknowledged that the

required dose was administered at the regular time the

night before and the morning of the experimental session.

If the subject or parent indicated that this was not the

case, the adolescent was not used as a subject. This

occurred with two potential poor control subjects.

The nondiabetic group was obtained by advertising at

the P.K. Yonge School and through staff at the J. Hillis

Miller Health Center. They were of the same sex, race, and

within 2 years in age of both their diabetic matches. All

study participants were paid money; $15 the first year data

were collected and $25 the second year.

Stress Manipulation Task

Each youngster participated in three potentially

stressful tasks. First a heparin lock was inserted in the

participant's arm or hand for the initial blood

withdrawal. The needle insertion was preceded by a 3

minute rest period and followed by a 3 minute recovery

period. Secondly, each participant was asked to give two 3

minute speeches. Each speech was preceded by a 3 minute

rest period and planning period. Both speeches were

followed by a 3 minute recovery period. Participants

remained seated for both blood withdrawal and speech

giving. All three events were videotaped. Each rest and

recovery period was preceded with the request to close

their eyes and rest and relax as completely as possible "as

if they were going to sleep."


Each subject was asked to give two speeches. She was

told that she would have 3 minutes to plan the speech and a

clock was pointed out to time the planning. No pencils or

paper was available for notetaking in the plan period. The

topics were "the last big argument I had or my most recent

big disappointment" and "a recent fun or pleasant time I

had or something very nice that happened to me." See

Appendix A for the instructions that accompanied each

topic. Each subject was told that the speech would be

videotaped and a small audience would listen. At least one

male and female were present during speech giving. Matched

subjects were yoked to the same speech order and the order

of the two speeches was alternated between sets of matched

subjects. If a subject "froze" in his speech, one of the

audience would say a phrase designed to keep the talk

going. Examples included "tell us some more about that,"

"keep going," "what else." Generally the audience was

supportive and friendly and listened to the subject with an

interested affect.

Major Dependent Measures

Heart Rate

One sympathetic response to emotional stress is

increased heart rate. Increased heart rate has been used

as an indicant of emotional arousal and remains sensitive

across several consecutive stressors punctuated by brief

rest periods (Harvey & Hirschmann, 1980; Shipman, Heath &

Oken, 1979).

Whether or not heart rate accelerates or decelerates

is greatly dependent on the nature of the task or

stimulus. Siddle and Turpin (1980) point out that heart

rate decelerates in response to simple stimuli and

accelerates to intense or threatening stimuli, during

periods of word association tasks, and during mental

arithmetic tasks. Heart rate increases in both the

anticipatory and performance phases of public speaking

(Borkovec & O'Brien, 1977; Knight & Borden, 1979; Levenson,

Jaffee & McFall, 1978). Please refer to Appendix B for a

brief discussion of the heart and primary theories

regarding its regulation.

Heart rate data were collected on a Lafayette four-

channel datagraph. Paper speed was 10 mm/sec. Heart rate

was obtained by counting the systolic spikes associated

with the cardiac contraction of the recorded pulses of the

photoplethysmographic transducer. The photoplethysmo-

graphic transducer was attached to the thumb of the left

hand unless this arm held the heparin lock, in which case

it was attached to the thumb of the right hand.

Heart rate in beats per minute (bpm) was tabulated for

each 1-minute segment of each period. Each period except

blood withdrawal (rest, plan, speech, recovery) lasted 3

minutes. Blood withdrawal lasted 2 minutes. The mean

heart rate of each period served as the respective period

score (rest, blood withdrawal, recovery, plan, and speech).

Skin Conductance

Measurement of the conductance of an electrical

current through skin tissue is often used as a

physiological indicant of arousal (Martin & Venables,

1980). Due to the high density of eccrine sweat glands

(which are innervated by the sympathetic nervous system) on

palmar and plantar skin surfaces, these sites are typically

used to obtain electrodermal information. Between group

differences in skin conductance activity on stress-inducing

tasks have been found (Knight & Borden, 1979; Levenson,

Jaffee & McFall, 1978). Please see Appendix C for a brief

note on skin conductance.

Skin conductance data were collected on a Lafayette

four-channel datagraph with paper speed of 10 mm/sec. Skin

conductance level and responses were recorded via bipolar

leads from the middle phalanges of the first and second

fingers of the left hand (unless this arm held the heparin

lock in which case the right hand was used) using Beckman

silver/silver chloride miniature electrodes with K-Y

Lubricating Jelly (Johnson & Johnson) as an electrode


Skin conductance was measured in micromhos. Skin

conductance levels were measured at each 20 second point

for each 1-minute segment during the 5-minute rest period,

the 3-minute plan period, the 3-minute task period (or 2-

minute blood withdrawal), and the recovery period. For

each period the mean of the skin conductance levels was

calculated and served as the score for each period. The

number of spontaneous conductance fluctuations equaling or

exceeding 0.1 micromhos was counted for each 5-minute


Blood Measures (FFA and Glucose)

Free fatty acids (FFA) and blood glucose increase in

response to stressful stimuli and are related to metabolic

disruption in diabetes (Tarnow & Silverman, 1981-82).

These blood measures were analyzed from blood samples drawn

at the beginning and end of the experimental session.

Urine Measures (Ketones, Glucose, and Volume)

Ketones increase in response to stress (Tarnow &

Silverman, 1981-82) and urine volume and urine sugar have

been shown to increase in response to stress exposure

(Hinkle & Wolf, 1952; Vandenbergh et al., 1966). These

urine measures were analyzed from urine samples taken at

the beginning and end of the experimental session-task


Venipuncture Questionnaire (VQ)

The VQ is a two-item Likert scale developed by the

researcher to allow the participant to rate venipuncture

(see Appendix D). The participant rated the task on two 5-

point scales asking how bothered by and painful the

venipuncture procedure was for them.

State Trait Inventory for Children (STAIC)

The state portion of the STAIC was administered to all

adolescents after each of the three tasks. The STAIC is

designed to measure both transitory anxiety specific to

stressful events (state anxiety) and stable anxiety with

consistency and permanence across time and events (trait

anxiety). Only the 20-item state anxiety portion of the

instrument was used in this study. The state portion of

the STAIC has good split-half reliability, r = .89 (Finch,

Montgomery & Deardorff, 1974), and has shown changes as a

function of stress (Bedell & Roitzch, 1976; Finch, Kendall,

Montgomery & Morris, 1975). The STAIC has been used

predominantly with children aged 8 and over. The STAIC was

orally administered and the subject responded to the task

portion of the procedure (blood withdrawal, speech).

Venipuncture Observation Scale (VOS)

The VOS was developed by the researcher to assess

observed anxiety in the venipuncture situation.

Appropriate items were selected from the Self-Injection

Behavior Profile Rating Scale (previously developed by the

researcher and S. Johnson, 1982). Other items were

developed from observed signs of nervousness noted by

medical staff involved in venipuncture. Ratings were

obtained from videotaped venipuncture session and

interrater reliabilities were obtained. Presence or

absence of each item was assessed for each 20 second

interval. Please see Appendices E and F for the VOS and

scoring procedure.

Rank Order of Task Form (ROTF)

The ROTF is a form developed by the researcher to

allow the participant to rank order the tasks from most

stressful to least stressful (Appendix G).

Personal Report of Confidence as Speaker
Short Form (PRCS)

The PRCS is a 50-item questionnaire revised by Paul

(1966) from an earlier and longer version in order to

improve the form's psychometric properties and make its

completion easier and quicker for the informant. The

instrument is designed to assess self-reported public

speaking anxiety.

Time Behavioral Checklist for Performance
Anxiety (TBCL) Modified Form

The TBCL was developed by Paul (1966) to assess

performance anxiety exhibited during a public speaking

exercise. The instrument lists 20 observable

manifestations of anxiety and is scored for their presence

or absence during consecutive 20-second observation

periods. The TBCL assesses behaviors reflecting

interference with performance (e.g., stammering) and

observable effects of arousal on behavior (e.g., heavy

breathing). The TBCL was modified by deleting items that

were inappropriate for videotaped speeches by seated

persons (e.g., pacing). Several items were added from

other facial rating protocols by Ekman and Friesen (1975)

including miserable smile and facial emblem negative. Paul

(1966) reports the average interobserver reliability after

training exceeded r = .95. Other investigators using the

TBCL have reported interrater reliabilities after training

between .71 to .96 (Ciminero, Calhoun & Adams, 1977).

Ratings were obtained from videotaped performances and

interrater reliabilities computed. Raters were trained

prior to scoring. See Appendices F and H for TBCL-M form

and scoring procedures.

Additional Dependent Measures

Junior Eysenck Personality Questionnaire (JEPQ)

The JEPQ is a personality inventory designed to

measure extraversion, neuroticism, psychoticism, and

conventionality for children aged 7-15 (Eysenck & Eysenck,

1978). Only the neuroticism, extraversion, and

conventionality scales were used. Six month test-retest

reliabilities for each scale by age and sex are as

follows: extraversion range--.38-.82, with all remaining

coefficients in the .60's and .70's; neuroticism range--.66

to .77, the conventionality scale range--.59 to .83; with

all remaining coefficients in the .60's and .70's. One

month test-retest reliabilities were considerably higher

with a range across scales by age and sex of .59 to .89,

with most coefficients in the .70's and .80's.

The JEPQ is an extension for children of the EPQ with

normative and reliability data available, but lacking

extensive validational studies. A manual to help interpret

scores is available.

Eysenck Personality Questionnaire (EPQ)

The EPQ is a personality inventory designed to measure

the same personality factors as the JEPQ in adults aged 16

and older. The extraversion, neuroticism, and

conventionality scales of the EPQ were administered to the

16-18 year-old subjects. One month test-retest

reliabilities by group and scale are primarily in the

.80's, with a range of .72 to .92; overall reliability

coefficients were as follows: extraversion was .90 and

.86; neuroticism was .89 and .80; and conventionality was

.86 and .86, for men and women, respectively. The

extraversion and neuroticism scale were produced through

factor analytic procedures and are orthogonal factors.

Eysenck and Eysenck (1975) report that others have

reproduced this factor pattern and they report validity

data using twin and other experimental studies (Eysenck &

Eysenck, 1975). A manual to help interpret scores is


Life Events Checklist (LEC)

The LEC is a 46-item (plus four blank spaces for

individual responses) inventory of significant life events

for adolescents (Johnson & McCutcheon, 1980). The

respondent is requested to check those events (s)he

experienced during the preceding year, rate the event as

good or bad, and rate the degree of impact the event on

his/her life on a 4-point scale from no effect = 0 to great

effect = 3. This instrument developed by Gad and Johnson

(1980) is relatively new and was developed to overcome

specific deficiencies in the existing life change/stress

inventories. Specific items were selected from existing

life change/stress inventories and nominations by an

adolescent sample. It has been administered to adolescents

aged 12 to 17 and found to correlate with indices of

physical and emotional health.


The families were asked to participate either by

telephone or in a letter. A letter was used only when

there was no telephone in the home. If both parents and

adolescent agreed to participate, an appointment time was

scheduled. Parents were asked to observe the insulin

injection of their child the night before and morning of

the experiment. In all cases, an informed consent was

obtained from both the adolescent and parent. The parent

was given an LEC form to complete while the youngster was

taken to the study area.

First the participant voided urine into a container.

Then she was taken to the room where the experiment was to

transpire and the equipment was explained. Words like

electrodes were avoided and attempts were made to use

nonthreatening and understandable words to explain the

different pieces of equipment (videorecorder and camera,

physiograph, leads). Although visible, the tray with

venipuncture equipment was placed somewhat behind the

youngster. The adolescent was seated in a comfortable

chair and the heart rate and skin conductance electrodes

were attached to the hand. Then the participant was asked

to close his/her eyes and rest for 3 minutes. Afterward

the videorecorder was turned on and the heparin lock

inserted. When the venipuncture procedure was completed,

the recorder was turned off and the adolescent asked to

rest with closed eyes for 3 minutes. After the recovery

period, the following questionnaires were administered (by

the researcher reading the items outloud): STAIC (applied

to blood withdrawal), VQ, JEPQ or EPQ, LEC, and PRCS.

After completion of the questionnaires which took about 30

minutes, the 3-minute baseline rest for the first speech

took place, followed by presenting the topic to the subject

and giving him/her 3 minutes to plan the speech. Then the

audience came in the room, the videorecorder was turned on,

and the adolescent gave the speech. This was followed by

the 3-minute recovery period. The STAIC (for the speech)

was administered a second time. Then the same procedure

was followed for the second speech. After the recovery

period for the second speech, the STAIC (for the second


speech) was readministered and the final blood withdrawal

done and heparin lock discontinued. The ROTF was

administered, the electrodes were removed from the

youngster's hand, and the Peabody Picture Vocabulary Test

was given. The participant was debriefed and paid.


Description of Sample

Forty-five adolescents participated including 18

females, 27 males, 8 black and 37 nonblack subjects. Each

good control diabetic subject (GCDS) was yoked to a same

sex, same race subject with the exception of one black GCDS

whose poor control diabetic subject (PCDS) match was


An ANOVA was performed for age and no significant

difference' between groups was found (F(2,42) = .616, p =

.54). Mean ages by group were GCDS--14.75 years, PCDS--

13.92 years, and nondiabetic subjects (NDS)--14.33 years.

The range of age was 11 to 18.8 years.

Duration of Diabetes and HA1 Values

A t-test was computed to determine differences between

GCDS and PCDS in duration of diabetes. No significant

differences were found (t(28) = .82, p = .421). The mean

number of years for duration of diabetes for the GCDS was

5.62 and for the PCDS 4.47.

A t-test showed significant difference for HA1 values

between the GCDS and PCDS (T = -9.77, P < .001). The mean

value for the GCDS was equal to 11.26 with a range of 8.5

to 13. The mean value for the PCDS was equal to 16.23 with

a range of 15 to 18.5.

Time of Day and Location of Data Collection

A 3 X 3 chi-square statistic was computed and no

evidence emerged suggesting differences between groups in

time of day subjects were run (chi-square = 3.6, p = .50).

An equivalent number of GCDS and PCDS were run in

Gainesville and in clinics outside Gainesville. Eight GCDS

and eight PCDS were run in Gainesville and the remaining

seven in each group outside Gainesville. All NDS were run

in Gainesville.

Reliability of Observation Measurement

Pearson product moment correlations were computed for

interscorer reliability of observed anxiety (VOC, TBCL).

The following reliability coefficients were obtained: for

the VOC, r = .79, p < .001; for the first speech (TBCL), r

= .75, P < .001; and for the second speech (TBCL), r = .71,

p < .001. No significant differences were found between

scorers with t values and probabilities as follows: VOC (t

= .94, P = .36); first speech TBCL (t = .66, p = .52); and

second speech TBCL (t = 1.6, p = .13).

Subject-Parent LEC Correlations and T-Tests

Pearson product moment correlations were computed

between the subject and parent forms of the LEC. Pearson

product moment correlations were obtained for various

possible scoring methods and the following validity

coefficients were obtained: total events rated good--r =

.40, p = .05; total events rated bad--r = .54, P = .004;

total events rated good and weighted for impact on life--r

= .48, p = .01; total events rated bad and weighted for

impact on life--r = .60, p = .001, sum of total number of

events--r = .35, p = .08; and sum of total number of events

weighted for impact on life--r = .50, p = .009. Paired t-

tests were computed between the subjects' and parents' LEC

scores. There was a significant difference between all

scores from the various possible methods; youngsters,

compared to the parents, reporting more events and having

higher weighted score.

Inter-relationships Between Measures

Pearson product moment correlations were completed

between the physiological, self-report, and observation

data. Appendix I is a compilation of a selection of these

Pearson coefficients. Physiological variables had

coefficients ranging from .04 to .72 (post-blood sugar and

urine volume) with the majority between .20 and .40 (HRs

with FFA and blood sugars). Self-report measures had

coefficients between .02 and .59 (the first STAIC with

rated venipuncture fear) with most higher than .25 (STAIC,

JEPQ, Speech Fear Questionnaire). Validity coefficients

between self-report, physiological, and observation data

ranged from no relationship to .52 (observed video anxiety

for venipuncture with rated venipuncture pain).

Control Variables

Speech Fear (PRCS)

No significant differences were found between the

three groups on reported speech fear (PRCS) using a oneway

ANOVA (F(2,42) = .51, p < .60). The following mean scores

by experimental group were found: GCDS = 11.47, PCDS =

10.67, NDS = 13.20; where the higher the score, the more

fear indicated.

Venipuncture Fear (VQ)

An ANOVA was computed between all groups for rated

venipuncture fear (VQ) and no statistically significant

differences emerged (F(2,42) = 1.58, p < .22). The

following mean ratings by experimental group were found:

GCDS = 4.33, PCDS = 3.93, NDS = 3.60; where the lower the

score, the more fear indicated.

Venipuncture Pain (VQ)

An ANOVA was computed between all groups for

venipuncture pain and no statistically significant

difference occurred (F(2,42) = .16, p = .85). The mean

ratings by experimental group were as follows: GCDS =

4.13, PCDS = 4.00, NDS = 3.93; where the lower the score,

the more pain indicated.

Venipuncture Observation Scale (VOS)

An ANOVA was computed for observed venipuncture

anxiety for all groups and no statistically significant

differences were found (F(2,28) = 1.60, p = .22).

Observed Speech Anxiety (TBCL)

An ANOVA was computed between all groups for observed

speech anxiety for both speeches. No statistically

significant differences were found for either speech [first

speech--(F(2,34) = .327, p = .72), second speech--(F(2,33)

= .03, p = .97)].

Reported Anxiety (STAIC) for Venipuncture and Speeches

An ANOVA was computed between all groups for the STAIC

for venipuncture and no statistically significant results

emerged (F(2,42) = .44, P = .65). ANOVAs were performed on

the STAIC administered for each of the speeches. No

significant differences were found for the second speech

(F(2,42) = .28, p = .76) but a tendency toward significance

occurred on the first speech (F(2,42) = 2.96, p = .06). A

Duncan's Range Test at the .05 level of significance showed

the PCDS reported more anxiety than the GCDS. Means for

the three groups are as follows: GCDS = 33.73, PCDS =

40.13, NDC = 35.67.

Rank Order of Task Form (ROTF)

Two 3 X 3 chi-square statistics were computed for the

rank order of tasks by subjects. The first chi-square

analysis found no difference between groups in rank

ordering venipuncture, first speech, and second speech

(chi-square = 5.4, P < .50). The second analysis found no

difference between groups in rank ordering venipuncture,

the speech on a pleasant topic, or the speech on an

unpleasant topic (chi-square = 6.0, p < .20).

Speech Order and Heart Rate

A 5 X 2 X 4 repeated measures ANOVA was computed for

experimental group, speech order and period (rest, plan,

speech, recovery) for both speeches. No main or

interaction effects were found for speech order in either

speech (F(1,37) = 1.17, p = .29) and (F(1,39) = .59, P =

.45), respectively.

Speech Order and Skin Conductance

A 3 X. 2 X 4 repeated measures ANOVA was performed for

experimental group, speech order, and period (rest, plan,

speech, recovery) for both speeches. No significant

differences were found for speech order in either speech

(F(1.38) = 1.76, p = .19) and (F(1,37) = 2.87, p = .10),


Analyses of Major Hypotheses

Heart Rate for Venipuncture

A 5 X 2 X 3 repeated measures ANOVA was computed for

experimental group, sex, and venipuncture period (rest,

blood withdrawal, recovery). Significant main effects were

found for experimental group (F(2,39) = 6.25, p = .004),

sex (F(1,59) = 9.63, p = .004), and period (F(2,78)

= 14.26, p < .001). A significant period X sex interaction

occurred (F(2,78) = 3.40, p = .04). Subsequent ANOVAs for

experimental group accompanied by Duncan's Multiple Range

Tests (at the .05 level of significance) found the heart

rate for PCDS was significantly higher than the two

remaining groups at each period of the venipuncture

procedure [rest (F(2,42) = 4.17, p =.02); blood withdrawal

(F(2,42) = 6.58, p = .003); recovery (F(2,42) = 4.37, p =

.02)]. Figure 1 illustrates the magnitude of the HR


Overall females had a higher heart rate in beats per

minute (BPM) with a mean equal to 90.19 BPM compared to

80.35 BPM for males.

Utilizing paired t-tests it was found that HR during

blood withdrawal was significantly higher than HR during

the rest or recovery period (t(44) = -5.14, p < .001) and

(t(44) = 3.76, p = .001), respectively. ANOVAs were

computed for sex by period with subsequent Duncan's

Multiple Range Tests performed. Whereas males had

significantly lower HRs in the rest and blood withdrawal

periods of the venipuncture procedure (F(1,43) = 10.11, p =

.003 and F(1,43) = 8.77, p = .005, respectively), no such

sex difference was found in the recovery period (F(1,43) =

1.43, p = .24).

(Wdg) 3aIV IUV3H


w In



















)ir -i
0) (

*Q 0

*^ fl






~ pO

0 0 0 0 0
0o oo

Heart Rate for First Speech

A 3 X 2 X 4 repeated measures ANOVA was computed

between experimental group, sex, and period. Significant

main effects were found for experimental group (F(2,39) =

5.65, p = .007), sex (F(1,39) = 9.51, p = .004), and period

(F(3,117) = 32.35, p = .000). In each case the PCDS had

significantly higher HRs than the NDS and, with the

exception of the speech period higher than the GCDS. See

Figure 2 for the comparison of HR by experimental group and


Overall females had higher HRs with a mean HR of 89.54

BPM compared to an 80.13 mean for males. Paired t-tests

were computed between all possible combinations of periods

in the first speech and the rest and recovery periods

significantly differed (were less) from the plan (t(44) =

-3.17, p = .003 and t(44) = 40.57, p = .000, respectively),

and speech periods (t(44) = -7.32, p = .000 and t(44) =

8.39, p = .000). Heart rate during the plan period

likewise was less than the speech period (t(43) = 39.17, P

< .001).

Heart Rate for Second Speech

A 3 X 2 X 4 repeated measures ANOVA was computed for

experimental group, sex, and period. Significant main

effects were found for sex (F(1,37) = 4.19, p = .05) and

period (F(3,111) = 31.65, p < .001). A tendency for

experimental group to be significant was found

S I /









(wag) aiva IHV3H








.4 1-
0 1

0 C

E 4




ca 00
W a
*P S
(B *-

(F(2,37) = 2.54, p < .09) with PCDS having higher HRs that

GCDS and NDS. Female subjects had higher heart rates

(86.83 BPM compared to the 80.6 BPM of their male


Paired t-tests showed that the rest and recovery

periods had lower HRs than the plan (t(44) = -3.53, p =

.001 and t(42) = 1.87, p = .07) and speech periods (t(44) =

-7.84, p < .001 and t(42) = 8.89, p < .001). The planning

period had a lower HR than the speech period (t(44) =

-8.00, p < .001).

Skin Conductance for Venipuncture

A 3 X 2 X 3 repeated measures ANOVA was computed for

experimental group, sex, and period. A significant main

effect was found for period (F(2,78) = 44.76, p < .001).

Paired t-tests showed each period significantly different

from the others with the highest level of skin conductance

in the blood withdrawal period and lowest level in the rest

period. The T values were as follows: rest-blood

withdrawal (t(44) = -7.99, p < .001); rest-recovery (t(44)

= -6.28, p < .001); and blood withdrawal-recovery (t(44) =

3.10, p = .003).
Skin Conductance for First Speech

A 3 X 2 X 4 repeated measures ANOVA for experimental

group, sex, and period was computed which found a

significant effect for period (F(3,114) = 17.35, p <

.001). Paired t-tests were computed and skin conductance

levels were less in the rest and recovery periods when

compared with the plan (t(43) = -2.77, p < .008) and (t(45)

= 2.4, p < .02) and speech (t(43) = -5.74, p < .001) and

(t(43) = 5.78, p < .001) periods. The speech period had
higher levels of skin conductance when compared with the

plan period (t(43) = -3.16, p < .003).

Skin Conductance for Second Speech

A 5 X 2 X 4 repeated measures ANOVA was computed. A

significant main effect was found for period. Subsequent

paired t-tests were computed and found speech skin

conductance differed significantly (was higher) from all

other speech periods with t values as follows: from rest

(t(44) = -5.88, p < .001); from plan (t(43) = -2.0, p <

.05); and from recovery (t(42) = 4.95, P < .001). Skin

conductance in the planning period was higher than in the

rest period (t(43) = -4.29, p < .001).

Skin Conductance Fluctations-Venipuncture

A 5 X 2 X 3 repeated measures ANOVA was computed

between experimental group, sex and period. A significant

main effect was found for period (F(2,76) = 26.55, p <

.001). Paired t-tests were utilized to compare each period
with the other periods. The blood withdrawal period had a

higher number of skin conductance fluctuations than the

rest or recovery period (t(44) = -7.02, p < .001 and t(44)

= -1.85, p < .07, respectively).

Skin Conductance Fluctuations for First Speech

A 3 X 2 X 4 repeated measures ANOVA was computed for

experimental group, sex, and period. A significant main

effect for period (F(3,114) = 76.89, p < .001) and an

interaction between period and experimental group (F(6,114)

= 2.38, p < .04) were found. The rest and recovery periods

had fewer SC fluctuations than the plan (t(43) = -8.4, P <

.001 and t(43) = 7.29, p < .001, respectively) and speech

periods (t(43) = -10.71, p < .001 and t(43) = 9.91, P <

.001, respectively). The plan period had fewer

fluctuations than the speech period (t(43) = -4.72, p <

.001). Experimental group was significant only in the

speech period (F(2,41) = 3.08, p < .06) with a subsequent

Duncan's Multiple Range Test showing that the GCDS had more

fluctuations than the PCDS with the following means: GCDS

= 14.53, PCDS = 9.57, and NDS = 10.87.

Skin Conductance Fluctuations for Second Speech

A 3 X 2 X 4 repeated measures ANOVA was computed for

experimental group, sex, and period. A significant main

effect was found for period (F(3,111) = 41.64, p < .001).

Each period was compared with the remaining three periods

utilizing paired t-tests. The rest and recovery periods

had the lowest number of fluctuations compared to the plan

(t(43) = -6.59, p < .001 and t(42) = 6.29, p < .001,
respectively) and speech periods (t(44) = -9.11, p < .001

and t(42) = 8.41, p < .001, respectively). The speech

period had the highest number of SC fluctuations (t(43) =

-4.44, p < .001).
Life Events Checklist (LEC)

A oneway ANOVA was computed using the LEC completed by

the adolescent for each of several methods of scoring the

LEC and none reached statistical significance. The F

values and probabilities for each scoring method are as

follows: total events rated good (F(2,41) = .43, p = .65);

total events rated bad (F(2,41) = .40, p = .68); total

events rated good and weighted by impact on life (F(2,41) =

.08, p = .92); total events rated bad and weighted by

impact on life (F(2,41) = .70, p = .50); sum of total good

and bad events (F(2,41) = .31, p = .74); and sum of total

good and bad life events weighted for impact on life

(F(2,41) = .17, p = .85).

Eysenck Personality Questionnaire (EPQ, JEPQ)

A oneway ANOVA was computed for experimental group for

each dimension of the EPQ extraversionn, neuroticism,

psychoticism, conventionality). All EPQ and JEPQ scores

were converted to t scores with mean = 50, SD = 10 (based

on age and sex norms in Eysenck & Eysenck, 1975, 1978)
before analyses. No significant differences were found for

extraversion (F(2,42) = 1.00, p = .38) or psychoticism

(F(2,42) = .38, p = .69). A significant F value was

obtained for neuroticism (F(2,42) = 5.9, p < .005) and a

subsequent Duncan's Multiple Range Test (at the .05 level

of significance) was performed and found the good control

diabetic group differed significantly (less neurotic) from

the remaining two groups. The mean t score for the GCDS

was 45.13 compared to 52.87 and 55.75 for the PCDS and NDS,

respectively. An ANOVA was performed for conventionality

which showed a tendency (F(2,42) = 2.48, p = .10) for GCDS

to positively endorse conventional items more frequently

than the NDS utilizing Duncan's Multiple Range Test. The

mean t score for the GCDS was 58.47 compared to 55.20 and

50.80 by the PCDS and NDS, respectively.

A series of Pearson product moment correlations were

computed between extraversion and neuroticism and HR and

skin conductance for all the individual periods in the

venipuncture procedure and the speeches. No significant

correlations for HR or skin conductance were found. As a

result, no further analyses of extraversion or neuroticism

and psychophysiological arousal were done.

Free Fatty Acid (FFA)

A 3 X 2 X 2 repeated measures analysis was computed

for experimental group, sex and time of measurement (pre-

post blood withdrawal). A tendency toward significance for

experimental group was found (F(2,34) = 2.8, p = .07). A

Duncan's Multiple Range Test at the .05 level of

significance found that the PCDS differed from the NDS at

the first blood withdrawal. Mean values were as follows:

GCDS = .44; PCDS = .59; NDS = .31. Although not

statistically significant, the same pattern of mean values

was found for the post FFA sample, i.e., GCDS = .49, PCDS =

.55, and NDS = .36.

Blood Sugars

A 3 X 2 X 2 repeated measures ANOVA was computed for

experimental group, sex and trial (pre- and post-blood

withdrawal). A significant main effect for experimental

group (F(2,28) = 14.12, p < .001) and a trial by sex

interaction (F(1,28) = 9.82, p < .004) were found. A

oneway ANOVA and subsequent Duncan Multiple Range Test were

computed for both pre- and post-blood withdrawals. The

overall model was significant for both blood withdrawals

(F(2,35) = 15.38, p < .001 and F(2,32) = 15.89, p < .001,

respectively). At the .05 level of significance the Duncan

Multiple Range Test found all three experimental groups

differed significantly from each other at the initial

withdrawal but only the NDS differed from both diabetic

groups at the second withdrawal. Mean blood sugar values

for the first and second blood withdrawals, respectively,

were as follows: GCDS = 210.73, 215.17; PCDS = 294.5,

275.84; and NDS = 59.8, 65.64. Although at the first blood

withdrawal females had significantly higher levels of blood

sugar, no significant difference was found at the second

blood withdrawal.

Urine Volume

A oneway ANOVA was computed for experimental group

with urine volume. Overall significance for the model was

found at (F(2,42) = 5.67, p = .007). A Duncan's Multiple

Range Test showed that the NDS differed from the two

diabetic groups with a mean volume of 69.67 cc compared to

183.8 for the GCDS and 256.67 for the PCDS.

Pre- and Post-Urine Sugar

A Wilcoxin rank sum test (equivalent to the Mann-

Whitney U test) was computed for GCDS and PCDS urine sugar

levels for pre- and post-experimental session. No

significant differences were found between groups at the

beginning of the experimental session (z = .41, p = .66)

but the PCDS had larger values at the end of the session (z

= 1.85, p = .04). All NDS had 0 percent urine sugar.

Pre- and Post-Urine Ketones

A Wilcoxin Rank Sum Test was computed for urine

ketones before and after the experimental session. No

significant differences were found pre-session (z = .39, p

= .35) although a tendency for PCDS to have higher levels

of urine ketones was found at post-experiment (z = 1.29, p

= .10). No traces of urine ketones were found for any NDS.

Pre- and Post-Plasma Ketones

No evidence of plasma ketones was found in any

experimental group at pre- or post-blood withdrawals.


The major contribution of this study was to clarify

where and how insulin-dependent diabetic and noninsulin-

dependent diabetic youth differ in their physiological and

metabolic responsivity to a psychological stress.

Furthermore, differences between the responsiveness to the

stress by the level of diabetes control was addressed.

Basically the findings suggested that the insulin dependent

diabetic adolescent responds similarly to psychological

stress as his nondiabetic peer but generally with higher

and less desirable levels of response. However, little

evidence of excessive physiological responsivity or slower

return to baselines accrued. This study generally

replicated the directions of the findings of Hinkle and

Wolf (1952) and Vandenbergh et al. (1966). However, the

outcomes of this study failed to support the findings of

Minuchin et al. (1978). Specifically, no group responded

with a comparatively extreme increase in FFAs and slower

return to baseline FFA levels. In fact, in this study the

PCDS actually had a slightly lower FFA level after the

stressful task compared to pre-experimentally. This does

not support the notion that ketoacidosis and other serious

illness in insulin dependent diabetes are the result of an

extreme physiological response to stess.

What this study has not done is to definitively tell

us the reasons for the physiological and metabolic

differences found between groups. Possible explanations

and their merits are more fully discussed in the following

sections. Likewise, the exploratory hypotheses regarding

life stress and personality traits are discussed.

Reported and Observed Anxiety of Tasks

Overall there were few differences between the GCDS,

PCDS, and NDS in self-reports of anxiety, task rank

ordering, or observed anxiety on any of the tasks

(venipuncture or speeches). The one exception, in the

predicted 'direction, was that the PCDS tended to report a

higher level of anxiety during the first speech but this

tendency was lost during the second speech.

Physiological Response to Tasks

Heart rate, but not skin conductance, significantly

differentiated the PCDS from the GCDS and NDS. Heart rate

in PCDS was consistently about 10 BPM higher than the GCDS

or NDS which were highly similar. However, no differences

between groups were found for skin conductance. In fact, a

tendency of higher skin conductance fluctuations in the

GCDS was found compared to the PCDS and NDS. Furthermore,

no evidence of heightened reactivity and slower return to

baseline emerged. Instead, a consistent pattern of higher

HR level for poor control adolescents across all conditions

whether rest, blood withdrawal, speech, or recovery was


Heart Rate

The finding of higher heart rate in the PCDS is

consistent with the hypothesis that this group had higher

levels of catecholamines. This hypothesis is further

supported by the finding that PCDS had higher FFA, blood

sugars, urine sugars, urine volume, and urine ketones. In

effect, stress triggers the introduction of catecholamines

and other stress hormones which set off the stress response

of increased HR, FFA and blood sugar. This is the normal

physiological process stimulated by stress. In the PCDS

these stress responses were higher and less desirable than

in the good control group or nondiabetic group. In the

normal stress response insulin counters the effects of the

stress hormones and operates to reduce FFA and blood

sugars. By reducing the influence of the catecholamines

and stress hormones insulin contributes to the recovery of

heart rate to pre-stress levels. The skin conductance

findings and failure to find differences between groups on

self-reported and observation measures of anxiety do not

support this explanation.

Other Explanations of Higher Heart Rate in PCDS

The PCDS may have more morphologic damage to

circulatory organs (veins, arteries, heart) which reduces

the overall efficiency and intactness of the circulatory

system. One well-known risk of insulin-dependent diabetes

is heart disease. The retinal damage that is a serious

complication of IDDM is contributed to by vascular

hemorrhages, aneurysms, and neovascularizations. To make

up for the inefficiency resulting from these morphologic

abnormalities/damage the heart rate may be increased to

provide the blood flow required for normal body function.

Another potential explanation is that our PCDS had

higher rates of autonomic neuropathy. Naliboff (1985)

reports that estimates have been made that 40% of persons

with diabetes have at least mild symptoms of autonomic

neuropathy. He found a higher resting heart rate in a

group of adult subjects with both insulin-dependent and

noninsulin-dependent diabetes. Upon further examination of

these individuals some evidence of autonomic neuropathy was

found in almost all diabetics.

Autonomic neuropathy tends to be manifested earlier in

the parasympathetic nervous system as opposed to the

sympathetic nervous system. This fact may help explain the

desynchrony between heart rate and skin conductance which

is primarily innervated by the sympathetic nervous

system. Since heart rate is heavily influenced by

parasympathetic processes as well as sympathetic processes,

autonomic neuropathy may show heart rate effects before

skin conductance effects. That is, a relatively greater

deterioration of parasympathetic inhibition of heart rate

in the PCDS would lead to an increased heart rate. If this

explanation is supported, it suggests that children with

insulin-dependent diabetes should be monitored for

autonomic neuropathy symptoms earlier than currently is


Skin Conductance

The failure to find differences between experimental

groups in skin conductance levels is interesting. This in

conjunction with the failure to find differences between

groups in self-report or observed anxiety supports the

hypothesis that the groups were not differentially

stressed. As a result support for an alternate explanation

to increased catecholamine levels to account for the heart

rate finding is suggested.

Other Explanations for Skin Conductance Findings

Skin temperature is positively associated with skin

conductance response (Haroian, Lykken & Huser, 1984;

Venables & Christie, 1980). If vasoconstriction or poor

circulation was more pronounced in the PCDS, finger

temperature would have been reduced in the PCDS. Such

reductions of skin temperature in the PCDS may have acted

to mask any heightened sympathetic input to the sweat

glands that are responsible for skin conductance

activity. Since we did not measure vasoconstriction,

finger temperature, or epinephrine, we cannot rule out this


Also, heart rate and skin conductance do not always

jointly distinguish between groups although they may

function similarly in both groups. For instance in our

study, both skin conductance and heart rate levels

increased in the plan, speech, and blood withdrawal periods

and decreased in the rest and recovery periods. However,

only heart rate distinguished between experimental

groups. This type of finding in the literature is not

unusual. Defining the mechanisms that underlie the

discrepancies in physiological response systems is very

difficult. For example, until recently the measurement of

epinephrine was both difficult and lacked reliability.

Currently, the measurement of epinephrine is improved but

remains difficult and very expensive.

Skin Conductance Fluctuations

The tendency of the GCDS to have higher numbers of

skin conductance fluctuations suggests that this group had

a higher propensity to notice and respond to nuances in the

environment whether the nuance consisted of physiograph

noises, movements inside or outside the room, etc. Skin

conductance level is an indication of more than anxiety.

Like heart rate, skin conductance responds to new or novel

stimuli. Instead of decreasing in the orienting situation

as heart rate does, skin conductance increases (Lacey,

1967). The GCDS low heart rate level, lower FFAs, lower

urine volume, lower blood and urine sugars, and similar

self-report, ratings, and behavioral measures of anxiety

provide consistent data to rule out that this group was

more anxious or aroused by the tasks. This supports the

hypothesis that the GCDS are more alert to environmental

changes. If so, this finding leads to the question of

whether or not such a propensity might make the GCDS more

alert to both internal and external changes in the

environment which aid them in better decision making in the

care of their diabetes. Since persons with

insulin-dependent diabetes must make daily decisions

directly influencing the management of their disorder such

as when a snack is needed, when their insulin dose requires

adjustment, or when exercise is needed, this extra

awareness may benefit them.

Metabolic Reactivity

Clear metabolic differences emerged between at least

one of the diabetic and the nondiabetic groups on all

dependent measures. The nondiabetic group had a lower

level of FFA, lower urine volume, and lower blood and urine

sugars. In addition, although not always statistically

significant, in every case the PCDS had the highest values

for FFA, urine volume, urine ketones and blood and urine

sugars with the GCDS having the middle values.

Only minimal support for the hypothesis that PCDS

would show heightened reactivity emerged. Instead, a

picture of sustained higher levels of HR, FFA, and blood

sugars was found. In fact, in the case of FFAs and blood

sugars, the post-experimental values were reduced enough

for statistical significance to be lost when it was found

pre-experimentally. One possible exception to this

generalization is that although no significant differences

were found pre-experimentally in urine sugars and urine

ketones between the GCDS and PCDS, differences were found

post-experimentally with the PCDS having higher levels.

Both Hinkle and Wolf (1952) and Vandenbergh et al.

(1966) found that their diabetic participants tended to

have a blood glucose decrease following stress. The drop

in the diabetic group tended to be higher while in the

nondiabetic group the level tended to stay the same or

increase slightly (as occurred in our GCDS and NDS). Both

researchers point out that increased urine sugars did not

help account for the blood sugar drop in the poorly

controlled group. In our case, urine sugars and ketones

were higher post-experimentally in the PCDS and may help

explain the reduction in blood sugar and FFAs. That is,

sugar and ketones were filtered from the blood into the

urine. Hinkle and Wolf and Vandenbergh et al. had small

and more hetereogeneous samples which may have prevented a

similar finding. It should be noted that our findings did

not parallel the findings of Minuchin et al. (1978).

Whereas the PCDS FFA level post-experimentally was lower

than pre-experimentally and the NDS and GCDS slightly

higher, the psychosomatic diabetic group in the Minuchin et

al. study showed higher FFA levels post-experimentally and

the other diabetic and nondiabetic groups lower FFA levels.

Another explanation of the decreased final blood sugar

and FFA level in our sample is that the PCDS had a higher

metabolic need for energy to sustain normal body

functions. The increased HR across tasks provides support

for this notion. That is, the body must burn more energy

(FFA, blood sugar) to deal with a higher heart rate.

Life Stress and Diabetes Control

The hypothesis that number of reported life stress

events would be related to level of diabetes control was

not confirmed. No relationship between diabetes control,

determined in this study by Hemoglobin Al value, and amount

of positive, negative, or combined positive and negative

life events was found. Likewise no differences emerged

between the diabetic and nondiabetic groups. The failure

to find differences between groups on the LEC suggests that

it is not the amount of reported life stress per se that

influences level of diabetes control. This conclusion is

supported further by the lack of reported differences

between groups in stressfulness of blood withdrawal and

speech giving even though physiological responses differed

between groups.

Brand, Johnson and Johnson (1983) report a finding

similar to that of this study. They found no relationship

between Hemoglobin Al and the LEC in diabetic youth aged 10

to 17.8 years attending a summer camp. The only

correlation that they found which approached significance

was between negative life change and urine ketones.

In the two studies that found a relationship between

diabetes health variables, in particular Hemoglobin Al, and

reported life stress, several differences emerge.

Bradley's (1979) study differed from ours in several

ways. First, she included British adult subjects with

insulin-dependent diabetes and adult onset diabetes.

Secondly, she used another life events change instrument

than the LEC. Chase and Jackson (1981) also used another

life stress questionnaire and measured life stress over the

preceding 3 months instead of year as in the current

study. As in this study, they looked at adolescents with

insulin-dependent diabetes. They found a high correlation

between amount of reported life stress and Hemoglobin Al

values (r = .41). In contrast, the current study found a

nonsignificant Pearson correlation between negative life

events and Hemoglobin Al of .16 and total life events of

.02. The participants in this study were matched for

duration of diabetes, sex, age, and race while those in

Chase and Jackson's were not. It was also confirmed from

both parents and adolescent that the prescribed insulin

dosage was administered at the same time generally and

specifically the night before and morning of the

experiment. As a result the sample was different.

Perhaps an instrument more sensitive to everyday

stresses would show a stronger relationship. Kanner,

Coyne, Schaefer and Lazarus (1981) suggest that a life

stress scale designed to measure smaller, more frequent

daily stressors might yield better results in predicting

psychological and health outcomes than the major life event

scales such as the LEC. Such scales are now available.

Although more refined techniques to assess stress

might yield higher predictive power, other factors need

assessment to account for significant amounts of the

variance. For instance, in our study we found for the most

part that the experimental groups reported and behaved in

ways suggesting that they viewed the tasks with similar

levels of stress/anxiety. Yet the PCDS were in much poorer

control of their disorder and physiologically dealt less

efficiently and effectively with the stresses. It is

reasonable to expect that the physiological

predispositions/states, behavioral tendencies, as well as

cognitive traits of the individual, must be considered to

make the best predictions of the effect of stress on


Personality Findings

On the JEPQ/EPQ the GCDS were less neurotic than both

other groups and more conventional than the NDS. This

finding is similar to Simonds (1977) who found (using

parental reports) that the poorly controlled group was more

likely to be anxious and depressed, the same terms Eysenck

uses to describe an introverted neurotic personality.

Likewise Simonds reported that his good control group

reported fewer conflicts than the nondiabetic group and had

an unusually low incidence of parental divorce. Our

findings in addition to Simonds suggest that adolescents

with insulin-dependent diabetes in good control may be

unusually well-adjusted and conventional or socially rule

oriented. Also in accord with Simonds, they suggest that

in general poorly controlled diabetic adolescents have

average or normal psychological adjustment.

No evidence emerged supporting differential

physiological or metabolic responsiveness due to

neuroticism. This suggests that neuroticism is not related

to poor control through a clearly identifiable

physiological or metabolic mechanism. (See Appendix J for

the correlation coefficients between neuroticism and

physiological and metabolic variables.) It appears more

reasonable to postulate that the more conventional,

socially appropriate self-report response tendencies of the

GCDS may be related to behavioral tendencies to follow

prescribed instructions and advice regarding care. This

may reach into the general living style of the

respondent. She may be more inclined to eat properly,

exercise as directed, and avoid less healthful life styles.

Future Research

Several lines of research are suggested by the present

findings. The finding of desynchrony of heart rate and

skin conductance levels raises several questions. One is

whether the differences between PCDS and GCDS were due to

different levels of catecholamine response. Measurement of

plasma catecholamine would help answer this question. If

the PCDS had higher levels of catecholamine, that group

would be more physiologically stressed. If no differences

in catecholamines were found, our skin conductance findings

would be expected. We would not have to consider other

explanations of why no skin conductance differences

emerged, e.g., increased circulatory damage and/or

autonomic neuropathy.

Another research question relates to the presence of

increased circulatory system damage or autonomic damage.

Several avenues of research are available to explore these

hypotheses including medical examination and tests to

assess circulatory and autonomic neuropathy damage.

However, whether or not increased damage is identified, we

still may not know the degree to which the damage

contributes to our findings of increased heart rate.

Another approach to this question is to see if the

increased heart rate and metabolic responses can be

reversed. That is, can a poor control diabetic group be

treated by some method and as a result respond

physiologically and metabolically like the good control

group. To the extent that circulatory or autonomic

neuropathy is resistant to intervention, reversal of heart

rate findings would not be expected.

Another line of investigation relates to the question

of whether or not the PCDS is more stress sensitive but,

due to their initial high level of being stressed,

differential rates of response were lost. More stress

sensitive persons might become more physiologically aroused

about participating in the experiment and their heightened

initial arousal may not be generalizable to other

situations. Measurement of heart rate and some metabolic

measures (perhaps FFA and urine volume) in ordinary and

nonintrusive ways would be very informative. Heart

monitoring devices could be attached and worn over time so

that the participant "forgets" about the device. Blood

sample would be more difficult to accomplish. However, as

metabolic measures such as glycosolated hemoglobin become

available, indicants of the metabolic state over time will

help address this question.

Several other questions were raised by the findings.

One was whether or not the GCDS were more sensitive to

novel or changing stimuli in the environment as was

suggested by their increased skin conductance

fluctuations. Likewise, are they more aware of internal

states and changes in the body. As suggested by Simonds

(1977) and our personality findings, are the GCDS more

psychologically well-adjusted compared to both their

diabetic and nondiabetic peers? Furthermore, are they more

rule-oriented and likely to follow medical advice and live

more health-oriented life styles?


This study has several practical implications. First,

the high heart rate in the PCDS may be indicative of

serious medical problems in this group that to date have

not been expected to be manifested at so young an age.

Generally, adolescents of this age group are not closely

monitored for symptoms of autonomic neuropathy or

circulatory disease. These findings suggest that regular

monitoring should be taking place for these youngsters.

A second implication is that the PCDS do not build up

excessive FFAs or blood sugars in response to a stress but

return to their pre-stress baseline as their GCDS and NDS

counterparts. This does not provide support for a

"psychosomatic" hypothesis, for example as suggested by

Minuchin et al. (1978), in which predisposed

insulin-dependent diabetic youngsters become sick when

exposed to a stress (for Minuchin et al., the family stress

of a "psychosomatic family").

A final implication is that the GCDS are more

psychologically healthy and may be behaviorally oriented

toward better caretaking of their disease. This requires

much more investigation and can only be hypothesized about

at this time. It fits nicely with the current finding of

other researchers (Simonds, 1977).


Speech Topic 1

Hello. Your speech will be on the topic of "a recent

fun or pleasant time I had or something very nice that

happened to me." You are free to pick any event or angle

you wish. Some ideas include a grade you especially like;

a good time with your friend, parent, brother, sister,

etc.; special equipment that you got like a bike or record

player; an event you got to go to, etc. You can talk about

when it occurred, how it came about, what it was like for

you, how you still feel, what happened later, etc.

Speech Topic 2

Hello. Your speech will be on the topic of "the last

big argument I had or my most recent big disappointment."

You are free to pick any event or angle you wish. Some

ideas include a grade you did not like; an argument with

your friend, parent, brother, sister, etc.; equipment that

broke like a bike or record player; an event you had to

miss, etc. You can talk about when it occurred, how it

came about, what it was like for you, how you still feel,

what happened later, etc.


The heart in humans and higher mammals is a complexly

innervated organ whose activity frequently is used as an

indicator of a psychological process. Heart rate (HR)

acceleration and deceleration in response to various

stimuli have long been noted by psychologists. Heart rate

decelerates in response to simple stimuli and accelerates

to intense or threatening stimuli, during periods of word

association, and during mental arithmetic. Sokolov

referred to a HR increase in response to a stimulus as a

defense response whereas HR deceleration was called an

orienting response. Lacey (Siddle & Turpin, 1980)

hypothesized that this 'directional fractionation' could be

explained by the nature of the stimuli. Stimuli requiring

environmental intake and consequent sensory integration

would lead to HR deceleration. Lacey further suggested

that the HR deceleration was due to an indirect effect of

HR on cortical activity. Another explanation that has been

offered is that lowered body activity in general

facilitates sensory intake by reducing distraction. The

second part of the 'directional fractionation' is that HR

accelerates in response to stimuli requiring spurring

environmental rejection. Obrist (1976) introduces a new

principle in HR activity. He uses the term cardiac-somatic

coupling to describe the principle that HR changes in

accordance with somatic need. In other words, as somatic

activity increases HR increases and as somatic activity

decreases HR decreases. However, Obrist points out that

cardiac-somatic coupling breaks down in those situations

related to active avoidance of aversive stimuli. These

situations result in substantial HR increase that is

unrelated to somatic activity.

The heart is neurally innervated by two interactive

inputs from the sympathetic and parasympathetic branches of

the autonomic nervous system (ANS). The sympathetic input

consists of adrenergic fibers originating in the spinal

cord via the stellate and caudal cervicle ganglia. The

neural transmitter substances are epinephrine and

norepinephrine. Excitation of these fibers increases HR

and blood pressure and is associated with myocardiac

contractile force. Parasympathetic fibers (cholinergic)

emanate from the vagus nerve. Excitation of these fibers

reduces HR and contractile force and, in general, is

antagonistic (opposite in effect) to sympathetic

excitation. In general, the higher the sympathetic input,

the higher the parasympathetic input. Heart activity

influences baroreceptors in the vagal nerve which respond

according to the level of heart activity by either


inhibiting HR when HR is high and reducing parasympathetic

inhibition when HR is low.


The eccrine sweat glands are of particular importance

to the psychophysiologist and are sympathetically

innervated. These glands appear to play a role in thermal

regulation only for very hot temperatures. Eccrine sweat

glands are widespread over the body but are particularly

dense on the palmar and plantar surfaces. Martin and

Venables (1980, p. 10) point out that it is realistic to

think of the principle effector mechanism in SC measurement

as sweat glands arranged as resistors in parallel. The

eccrine sweating produces SC changes which are related to

orienting or signal responses. At the skin surface sweat

is both discharged and reabsorbed.


Please rate how

bothered you are in general by having blood



not bothered
at all

Please rate how painful the venipuncture procedure was for



not painful
at all


Subject Name
Rater Name


Rating Date

1 2 3 4 5 6 SUM
Time: *

Verbalized Pain *
Verbalized Anxiety *
Verbal Delay *
Looks Away *
Facial Grimaces *
Moisten Lips *
Swallow *
Heavy Breathing *
Smile Miserable *
Tearing/Crying *
Behavioral Delay *
Facial Emblem Negative *
*Facial Emblem Neutral *
*Smile False *
*Smile Spontaneous *
*Laughs *
*Talks Other *
*Blink Number *




*Anxiety General 0 1 2 3 4 5 6 7 8 9
*Activity General 0 1 2 3 4 5 6 7 8 9
*Positive General 0 1 2 5 4 5 6 7 8 9

* not scored or analyzed


General Procedure

Both the Venipuncture Observation Checklist (VOC) and

the Timed Behavior Checklist-Modified Form (TBCL-M) have

similar formats for scoring. First of all the videotapes

are readied for display on the Betamax videorecorder set.

The viewers) arrange themselves in comfortable seats

placed in a position maximizing their view of the tapes.

If two or more viewers are present, each is situated so

that no one can observe another scoring. This is done to

prevent inflation of the reliability measures by influences

other than the videotapes. Each scorer should have a

clipboard, pen or pencil, and scoring sheet. Before

viewing the videotape, scoring sheets should be completed

for subject's name or initials, ID number, scorer's name,

date, and whether the scorer is a reliability checker. On

the TBCL-M the speech number should be recorded (either 1

or 2) depending on whether the first or second speech is

being scored. The time settings delineating each 20 second

period should be filled in at the top of the form. The

initial time setting is obtained from the videorecorder

clock and subsequent times figured by adding 20 seconds to

the preceding time period.

Scoring Forms

Both forms have similar layouts. Each has a list of

specific behaviors going down the left hand column. To the

right of the list of behaviors, time period columns appear

with a separate box for each specific behavior. Each time

period accounts for 20 seconds for behavior.


Scoring consists of checking behaviors that appear

during each time segment. Behaviors are scored for their

presence or absence. If a behavior occurs during a time

segment, the corresponding box is checked. If it does not

occur, a zero is placed in the corresponding box. For both

forms (VOC, TBCL-M) the videotape is viewed for 20

seconds. The videotape is stopped by pressing the stop

button. The scorer then marks the appropriate box for the

behavior(s) that occurred during that 20 second period.

Each segment will require viewing several times to maximize

adequate scoring.

Venipuncture Observation Checklist

Two minutes of the venipuncture procedure will be

scored. This accounts for six time periods. The 60

seconds immediately before and after the needle insertion

will be scored. If needle insertion occurs before 60

seconds has lapsed continue to score after needle insertion

until six time periods have been scored.

Definitions and Descriptions of Behavioral Categories

Verbalized pain: says hurts, painful, ouch, ohhh, or other

verbal indication of pain/discomfort.

Verbalized anxiety: says scary, afraid, anxious, doesn't

like, or asked if it will hurt; exclude painful.

Verbal delay: makes excuses to delay venipuncture; example

includes asking phlebotomist to "wait a second."

Looks away at time of injection: this is scored only at

the time of needle insertion.

Facial grimaces: includes noncommunicative facial

movements such as tics and other uncoordinated muscle

movements; includes involuntary flinches.

Moistens or bites lips: licks or bites lips.

Swallows: swallows (note closed mouth and throat


Heavy and/or uneven breathing: involves obvious and clear

heavy and/or uneven breathing; include heavy and uneven

breathing associated with crying.

Smile miserable: involves a smile combined with clear

negative affect; usually a smile with a contracted

upper lip and may involve other facial expressions

indicative of negative affect; it is differentiated

from facial emblem negative by the smile which is

absent in the facial emblem negative.

Tearing/crying: tearing or crying; noticable welling or

tears in the eyes is scored.

Behavioral delay: involves a behavioral gesture that

delays venipuncture; examples include withdrawal of

arm, failure to extend arm when appropriate, or

covering site of injection.

Facial emblem negative: involves the coordinated tensing

and movement of facial muscles to provide a facial

expression that communicates negative affect to the

viewer; exclude if a smile is present; example includes

a snarled upper lip and nose or gritted teeth with

forehead frown.

The Time Behavior Checklist-Modified Form

The TBCL-M is scored in the same manner as the VOC

except that there are nine 20 second periods to be

scored. Many of the categories of behavior are the same

and the same definitions and descriptions apply in both

scales. Both include the following behavioral

categories: facial grimaces, moistens or bites lips,

swallows, smile miserable, heavy or uneven breathing, and

facial emblem negative. The distinct categories for the

TBCL-M are as follows:

No eye contact: fewer than three contacts with total

duration of all contacts less than two seconds.

Face deadpan: face looks bland, emotionless, flat for

entire 20 second period; associated with minimal eye

and head movement.

Vocal quivering: voice noticably quivers, breaks, or has

obvious pitch changes; includes noticable speech flow

or rhythm changes.

Speech blocks: includes evidence of speech blocks where

speaker cannot continue; examples include asking

researcher how much time left to speak, 3 seconds of

silence, having to repeat speech, saying have run out

of things to say.

Stammer/stutter: includes stammering or stuttering in

which at least two unnecessary sounds occur

consecutively; examples include "a a" or "f-f-friend";

excludes insertion of unnecessary words such as "you

know," unless phrase is repeated twice consecutively;

score if 5 or more breaks in flow occur in the 20

second period.


Please rank the three tasks according to how

stressful/anxious each one was for you. Place a 1 beside

the most stressful task for you and a 3 beside the least

stressful task. Place a 2 beside the task that fell

between these tasks in stressfulness for you.

Speaking on a recent pleasant time

Speaking on a recent argument or disappointment

The venipuncture procedure


Subject Nan
Rater Name



Rating Date Speech #

1 2 3 4 5 6 7 8 9 SUM
Time: *

No Eye Contact *
Facial Grimaces *
Face Deadpan ?
Moisten Lips *
Swallows *
Smile Miserable *
Vocal Quivering *
Speech Blocks *
Stammer/Stutter *
Heavy Breathing *
Facial Emblem Neg. *
*Facial Emblem Neut. *
*Smile False *
*Smile Spontaneous *
*Laughs *
*Blink Number *

*Anxiety General 0 1 2
*Activity General 0 1 2
*Positive General 0 1 2

* not scored or analyzed

Moderate Extreme
3 4 5 6 7 8 9
3 4 5 6 7 8 9
3 4 5 6 7 8 9



DS Isaa

ll q)aadS


VAA Isod

awnloA auTJn isod

ae8ns pooqg isod








0C\J (NJ

N oN

0 t-I K --
% -

0 0' E0- n'\
0 Or"OO"O--
Tr- I I I I




'- I

O --0 O

- I I

* *
0 ir, N



0 O 0 D

N +c>
-0C 0 a

E-4 w w c
l a0. (
6-1 CE M




,-4 -

0-1 -1

S0 (0

.-I0 C0
m 0

.t* *
CO 0 t- I- N

.r>~. \No C. 0, O
00 0 0 0'-

T- 0 -

0 0 w- O t- -0

-M 0 r- CM

-1 0 M -- 0-" 1 1



0 C
0 0


4-- 4a> M 4 0
0W 0 a)0 00
O1 MOOCi)0)
0 0 0 H (U

DS Isau

UH II qoaads

Hil Isaa

VAA isod

awnlOA UTlan qsod

'IfnS poolg Isod



uorsia~c ixs



0 0



N') 0 CO
0 -" L- 0

o n n

0 -" 0 0
* *


0 C
c I
l -.q 0
(0 0 -r

> 4-3

00 0
() X O0
M i M Z U
< Y u ra (
E-c rC

t- o
0 O)

'- m

o 0
'H *H

0 >

*.H 0

SU) 0
I ci) fl
u)n ci
H 0
E-1 S:




C! o C

N t( N %

0 00 0
r* >

00 H -

00 0 K

U) E 0
000m10 0

, aCO 0w c w 0)

o 0 0 d^a 0
aac0 3r

3S isal

HH IUemRJpqIT poolg

Nil Isaa

VJad sod

awnloA auiln isod

io2ns pooIqg sod





I si wl 'bA

I wall 'bA


* e


o C


* r-



'o co


* *




CI -Ct 0

** *

*C *H0

o0 0 H

r 4( 00

4'- *i) X C

C. 0 z-
cf-l 4- + 0


Ij 0 %t- M
K0 C0\


O %" o o -

oNOt N cooo



0 0 0 -- -0

oC- c-\ 0" t-i- o d tO
0 0 00
111 III

... *.*

K0 m 0 4O N

(0 (V 0 (m0

o oa oa .,-.

3 0 0 -0 -4

a, a. a. W C W3
00 c0 00














-4 0

0) -Li


bOo, U)
C ()X
hO- 6
**-< *T -
en 2: x
* *


Variables Extraversion Neuroticism

Post Blood Sugar .09 (S=.31) -.17 (S=.17)

Post Urine Volume .08 (S=.30) -.13 (S=.20)

Post FFA -.19 (S=.12) .12 (S=.25)

Task Venipuncture
Hr -.12 (S=.22) .04 (S=.40)
SC -.07 (S=.31) -.17 (S=.13)
HR Change -.26 (S=.04) .16 (S=.15)
SC Change .07 (S=.33) .17 (S=.13)

Task Speech I
HR .04 (S=.39) -.09 (S=.28)
SC .06 (S=.35) -.15 (S=.6)
HR Change .26 (S=.04) .04 (S=.39)
SC Change .20 (S=.10) 1.0 (S=.25)



Akerstedt, T., & Theorell, T. (1976). Exposure to night
work serum gastrin reaction, psychosomatic complaints
personality variables. Journal of Psychosomatic
Research, 20, 479-484.

Baker, L., Barcai, A., Kay, R., & Hague, N. (1969). Beta
adrenergic blockade and juvenile diabetes. Journal of
Pediatrics, 75, 19-29.

Bedell, J., & Roitzch, J. (1976). The effects of stress on
state and trait anxiety in emotionally disturbed,
normal, and delinquent children. Journal of Abnormal
Child Psychology, 4, 175-177.

Borkovec, T., & O'Brien, G. (1977). Relation of autonomic
perception and its manipulation to the maintenance and
reduction of fear. Journal of Abnormal Psychology, 86,

Bradley, C. (1979). Life events and the control of
diabetes mellitus. Journal of Psychosomatic Research,
23, 159-162.

Brand, A., Johnson, J., & Johnson, S. (1983). Life stress
and diabetic control in children and adolescents with
insulin-dependent diabetes. Unpublished manuscript.

Cahill, G., Etzwiler, D., & Freinkel, N. (1976). Control
and diabetes. New England Journal of Medicine, 294,

Chase, H., & Jackson, G. (1981). Stress and sugar control
in children with insulin-dependent diabetes mellitus.
Brief Clinical and Laboratory Observations, 6, 1011-

Ciminero, R., Calhoun, K., & Adams, H. (1977). Handbook of
behavioral assessment. New York: John Wiley and Sons.

Denny, D., & Frisch, M. (1981). The role of neuroticism in
relation to life stress and illness. Journal of
Psychosomatic Research, 25, 303-307.

Ekman, P., & Friesen, W. (1975). Unmasking the face: A
guide to recognizing emotions from facial clues.
Princeton, New Jersey: Prentice Hall.

Ekman, P., & Friesen, W. (1982). Felt, false, and
miserable smiles. Journal of Nonverbal Behavior, 6(4),

Eysenck, H. (1967). The biological basis of personality.
Springfield, IL: Charles C. Thomas.

Eysenck, H., & Eysenck, S. (1975). Manual: Eysenck
Personality Questionnaire. San Diego: Educational and
Industrial Testing Service.

Eysenck, H., & Eysenck, S. (1978). Manual of the Eysenck
Personality Questionnaire: Junior and adult. San
Diego: Education and Industrial Testing Services.

Fallstrom, K. (1974). On the personality structure in
diabetic school children. Acta Paediatrica, 254, 5-71.

Finch, A., Kendall, P., Montgomery, L., & Morris, T.
(1975). Effects of two types failure on anxiety.
Journal of Abnormal Psychology, 84, 583-585.

Finch, A., Montgomery, L., & Deardorff, P. (1974).
Reliability of state-trait anxiety with emotionally
disturbed children. Journal of Abnormal Child
Psychology, 2, 67-69.

Gad, M., & Johnson, J. (1980). Correlates of adolescent
life stress as related to race, SES, and levels of
perceived social support. Journal of Clinical Child
Psychology, 9, 13-16.

Ganong, W. (1971). Review of medical physiology (6th
ed.). Los Angeles: Lange Medical Publications.

Gilbert, B., & Johnson, S. (1982). Effects of a peer-
modeling film in anxiety-reduction and skill
acquisition in children learning self-injection of
insulin. Behavior Therapy, 13, 186-193.

Gray, J. (1975). Elements of a two-process of learning.
London: Academic Press.

Harkavy, J. (1981). A study of the relationship of
knowledge, behavior, and control in juvenile
diabetes. Unpublished master's thesis, University of
Florida, Gainesville, Florida.

Haroian, K., Lykken, D., & Huser, R. (1984). SPR
Abstracts: The effect of hand temperature on
electrodermal measurement. Psychophysiology, 21(5),

Harvey, F., & Hirschmann, R. (1980). The influence of
extraversion and neuroticism on heart rate responses to
aversive visual stimuli. Personality and Individual
Differences, 1, 97-100.

Hinkle, L., & Wolf, S. (1952). Importance of life stress
in course and management of diabetes mellitus. Journal
of American Medical Association, 148, 513-525.

Johnson, J., & McCutcheon, S. (1980). Assessing life
stress in older children and adolescents: Preliminary
findings with the life events checklist. In I. Sarason
& C. Spielberger (Eds.), Stress and anxiety (Vol. 7,
pp. 111-126). Washington, DC: Hemisphere.

Johnson, J., & Sarason, I. (1982). Life stress research:
Where we have been--where we are going. Unpublished

Johnson, S. (1980). Psychosocial factors in juvenile
diabetes: A review. Journal of Behavioral Medicine,
3, 95-116.

Kanner, A., Coyne, J., Schaefer, C., & Lazarus, R.
(1981). Comparison of two modes of stress measurement
daily hassles and uplifts versus major life events.
Journal of Behavioral Medicine, 4, 1-39.

Knight, M., & Borden, R. (1979). Autonomic and affective
reactions of high and low socially-anxious individuals
awaiting public performance. Psychophysiology, 16,

Koski, M., & Kumento, A. (1975). Adolescent development
and behavior: A psychosomatic followup study of
childhood diabetes. In Z. Laron (Ed.), Diabetes in
juveniles: Medical and rehabilitation aspects. Modern
problems in paediatrics (Vol. 12, pp. 348-355).
Basel: Karger.

Lacey, J. (1967). Somatic response patterning and
stress: Some revisions of activation theory. In M. H.
Appley & R. Trumbull (Eds.), Psychological stress:
Issues in research (pp. 14-36). New York: Appleton-

Levenson, R., Jaffee, L., & McFall, R. (1978, April).
Heightened responsibility to stress in nonassertive,
low self-confident subjects. Presented at Society of
Psychophysiological Research, Vancouver, BC.

Martin, I., & Venables, P. (1980). Techniques in
psychophysiology. New York: John Wiley and Sons.

Mehrabian, A., & Ross, M. (1977). Quality of life change
and individual differences in stimulus screening in
relation to incidence of illness. Psychological
Reports, 41, 267-278.

Minuchin, S., Rosman, B., & Baker, L. (1978).
Psychosomatic families. Cambridge, MA: Harvard
University Press.

Naliboff, B. (1985). Biobehavioral studies of stress and
stress management in diabetes. Unpublished manuscript.

Obrist, P. (1976). The cardiovascular behavioral
interaction--as it appears today. Psychophysiology, 3,

Paul, G. (1966). Insight vs. desensitization in
psychotherapy. Palo Alto, CA: Stanford University

Pinter, E., Peterfy, G., Cleghorn, J., & Pattee, C.
(1967). The influence of emotional stress on fat
mobilization: The role of endogenous catecholamines
and B adrenergic receptors. American Journal of
Medical Science, 254, 634-651.

Shipman, W., Heath, H., & Oken, D. (1979). Response
specificity among muscular and autonomic variables.
Archives of General Psychiatry, 23, 369-374.

Siddle, D., & Turpin, G. (1980). Measurement,
quantification, and analysis of cardiac activity. In
I. Martin & P. Venables (Eds.), Techniques in
psychotherapy (pp. 139-246). New York: John Wiley and

Simonds, J. (1977). Psychiatric status of diabetic youth
matched with a control group. Diabetes, 26, 921-925.

Stelmack, R. (1981). The psychophysiology of extraversion
and neuroticism. In H. Eysenck (Ed.), A model for
personality. New York: Springer-Verlag.


Tarnow, J., & Silverman, S. (1981-82). The
psychophysiologic aspects of stress in juvenile
diabetes. International Journal of Psychiatry in
Medicine, 11, 25-44.

Vandenbergh, R., Sussman, K., & Titus, C. (1966). Effects
of hypnotically induced acute emotional stress on
carbohydrate and lipid metabolism in patients with
diabetes mellitus. Stress and Metabolism in Diabetes,
28, 382-389.

Venables, P., & Christie, M. (1980). Electrodermal
activity. In I. Martin & P. Venables (Eds.),
Techniques of psychophysiology (pp. 5-62). New York:
John Wiley and Sons.


Brenda Gilbert was born in Knoxville, Tennessee, in

January of 1947. She obtained her M.S.W. degree from

Florida State University in 1972 and worked 5 years as a

social worker. She returned to the University of Florida

and earned an M.A. and Ph.D. in clinical psychology. Her

research interests are in the areas of medical psychology

and her focus has been on coping with chronic illness in

children and adolescents. She is married and the mother of

two beautiful girls. Currently, she is the coordinator of

an adolescent inpatient program.

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.

Suzanne B/. Johnson, Chairman
Associate Professor of Clinical

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.

Hugh Davy
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.

Barbara Melamed
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.

James Johnson
Associate Professor of Clinical

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.

Randy Cart'r
Associate Professor of

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 _0 __ _A......_
Dean, College of Health Related

Dean, Graduate School

11111111111 1111111111111111111111i 1 11 lllll 111111111
3 1262 08554 3790