Perceptual differences between chronic low-back pain patients and healthy volunteers using magnitude matching and clinic...

MISSING IMAGE

Material Information

Title:
Perceptual differences between chronic low-back pain patients and healthy volunteers using magnitude matching and clinically relevant stimuli
Physical Description:
v, 48 leaves : ill. ; 29 cm.
Language:
English
Creator:
Fuller, Adam Kittinger, 1963-
Publication Date:

Subjects

Subjects / Keywords:
Low Back Pain -- psychology   ( mesh )
Pain Measurement   ( mesh )
Perception   ( mesh )
Pain Threshold   ( mesh )
Exercise Test   ( mesh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1994.
Bibliography:
Includes bibliographical references (leaves 45-47).
Statement of Responsibility:
by Adam Kittinger Fuller.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 50408762
ocm50408762
System ID:
AA00011168:00001

Full Text











PERCEPTUAL DIFFERENCES BETWEEN CHRONIC LOW-BACK PAIN
PATIENTS AND HEALTHY VOLUNTEERS USING MAGNITUDE MATCHING
AND CLINICALLY RELEVANT STIMULI
















By

ADAM KITTINGER FULLER


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

UNIVERSITY OF FLORIDA


1994













ACKNOWLEDGMENTS

Several people have been integral in the completion of

this project. Michael Robinson, the chairman of my

dissertation committee and my research mentor, has been a

major contributor to this project and the driving force of

my research throughout my graduate career. I have valued

his ideas, his availability, his critical comments, and his

enthusiasm for research. I also want to thank Fred Shirley,

the physical therapist who assisted with this project.

Without him, his humor, and especially his willingness to

use his time to run the MedX exercise equipment for every

experimental and pilot subject, the entire project never

would have been completed. Finally, I want to thank my

wife, Pamela, for her support and listening ear. Even in

the throes of her own dissertation, she was always there.

In addition, I want to acknowledge my dissertation

committee members, Cynthia Belar, Russell Bauer, Michael

Geisser, and Linda Shaw, for their critical comments and

time spent on this project.














TABLE OF CONTENTS

page

ACKNOWLEDGMENTS.... .................................. ii

ABSTRACT. ....... ....................................... iv

INTRODUCTION..... ............. ......... ....... 1

REVIEW OF LITERATURE.................................... 3

Perceptual Differences Found in
Chronic-pain Patients........................ 3
Signal Detection Theory ............................ 8
Magnitude Matching ................................ 11
Magnitude Matching in Terms of
Signal Detection Theory....................... 15
Signal Detection and Magnitude Matching
in the Present Study........................... 18

METHOD.................................................. 20

Subjects........................................... 20
Apparatus ......................................... 22
Procedure.......................................... 23
Analysis............................................. 26

RESULTS................................ ............ 29

DISCUSSION.............................................. 35

APPENDIX (BACKGROUND QUESTIONNAIRE)..................... 44

REFERENCES.................................. .................. 45

BIOGRAPHICAL SKETCH..................................... 48











iii













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

PERCEPTUAL DIFFERENCES BETWEEN CHRONIC LOW-BACK PAIN
PATIENTS AND HEALTHY VOLUNTEERS USING MAGNITUDE MATCHING
AND CLINICALLY RELEVANT STIMULI

By

Adam Kittinger Fuller

August 1994

Chairperson: Michael E. Robinson, Ph.D.
Major Department: Clinical and Health Psychology



It has been demonstrated by numerous investigators that

patients with chronic pain who are exposed to experimental

pain stimuli are less able to discriminate pain stimuli than

normal controls. Conflictual studies have also shown that

patients with chronic-pain may either have a bias toward

under-reporting the painfulness of experimental stimuli (the

"adaptation" hypothesis) or toward over-reporting the

painfulness of stimuli (the "hypervigilance" hypothesis).

It has also been suggested that chronic-pain patients show

differences in the perception of non-painful stimuli. The

present study tested these findings using clinically

relevant stimuli requiring pain- and non-pain related

judgements. Magnitude matching was used to compare 16

chronic low-back pain patients with 16 pain-free controls in

iv









their perception of pain/discomfort during low-back exercise

on a lumbar exercise apparatus as well as their perception

of heaviness of weights lifted during a separate set of

exercise.

In contrast to previous studies that have used less

relevant stimuli, the chronic low-back pain subjects in this

study were able to discriminate relevant pain-related

stimuli better than normals, and they were equal to normal

controls in their ability to discriminate the relevant

pain stimuli of heaviness. In addition, consistent with the

adaptation model, subjects with chronic-pain tended to

under-report the painfulness of lifting weights with their

backs, showing no difference in pain ratings during low-back

exercise compared to pain-free controls. The chronic low-

back pain subjects also consistently underestimated the

heaviness of weights lifted with their lower-backs, which

may represent a generalization of an under-reporting style

to relevant but non-pain related stimuli. A cognitive

distortion hypothesis is proposed to explain the under-

reporting, or under-estimating style of the chronic-pain

subjects, and a nociceptive noise hypothesis is proposed to

explain the discrimination data from this and previous

studies.













INTRODUCTION

Traditionally, pain perception has been investigated by

subjecting healthy volunteers to a variety of experimental

pain stimuli. The goal of such research, whether directly

stated or not, is to further understanding of nociception

and ultimately to apply this knowledge to the treatment and

management of chronic and acute pain in human patients.

Although the use of healthy volunteers is common, it has

been demonstrated by numerous investigators that patients

with chronic pain who are exposed to experimental pain

stimuli exhibit altered pain perception compared to the

typical healthy subjects (Malow, Grimm, & Olson, 1980;

Naliboff, Cohen, Schandler, & Heinrich, 1981; Yang, Richlin,

Brand, Wagner, & Clark, 1985). It has also been shown that

chronic-pain patients show differences in the perception of

non-painful stimuli (Seltzer and Seltzer, 1986; Flor,

Schugens, & Birbaumer, 1992).

Evidence of perceptual changes across modalities

suggests that there may be an underlying central processing

deficit that exists in people who have experienced prolonged

pain. A question remains whether these perceptual

differences exist only in the context of experimental

stimuli, such as radiant-heat applied to the forearm or a

lucite-edge pressed on to a finger, or if such differences

1









occur in situations with more relevant stimuli. A central

processing deficit hypothesis would predict altered

perception of clinically relevant stimuli in chronic-pain

patients. Whether or not this hypothesis is true, the

evidence for altered perceptions in chronic-pain patients

indicates that the traditional method of using pain-free

subjects and experimental pain stimuli may not provide

adequate external validity necessary for drawing conclusions

about the pain experience of the chronic-pain populations in

question.

The present study combined a signal detection theory

conceptualization with magnitude matching methodology. The

intent was to assess differences between chronic low-back

pain patients and healthy control subjects in two

dimensions: the first being their perception of clinically

relevant, potentially painful stimuli (standard exercise

with their lower back), and second, a non-pain related, but

still relevant, perceptual dimension (judgement of heaviness

of a weight lifted with the lower back).













REVIEW OF LITERATURE

Perceptual Differences Found in Chronic-pain Patients

One of the first comparisons of pain perception between

chronic-pain patients and normal subjects was done by Malow

et al. (1980). Using a signal detection analysis and

threshold measures, these authors found that a group of

myofascial pain dysfunction patients had lower pain

thresholds and were more likely to report the experimental

stimuli to be painful as compared to pain free normal

subjects. The authors suggest that the experience of

chronic pain sensitizes the entire nervous system to

stimulation resulting in an over-responsivity in the

chronic-pain patients "as if the slightest touch makes a

person in chronic pain jump" (Malow et al., 1980, p. 307).

Malow and Olson (1981) provided supporting data by

demonstrating that patients who were successfully treated

for myofascial pain showed increased pain thresholds, and a

decreased likelihood of reporting stimuli as painful

compared to patients whose pain was not alleviated by

treatment. Both of the above mentioned studies provide

evidence for a "hypervigilance" model similar to that

described by Chapman (1978) which hypothesizes that chronic-

pain patients are overly focused on bodily sensations,

3









particularly painful sensations, resulting in a tendency to

over-report pain.

Other studies have not lent support to the

hypervigilance model. Naliboff et al. (1981) and Cohen,

Naliboff, Schandler, and Heinrich (1983) found that chronic

low-back pain patients had higher thresholds for radiant-

heat pain than normal controls, which contradicts the

findings of the Malow et al. (1980), and Malow and Olson

(1981) studies. Naliboff et al. (1981) concluded that

rather than a hypervigilance model, their results best fit

an adaptation model originally described by Helson (1964)

and extended to pain perception by Rollman (1979). This

model predicts that chronic-pain patients would exhibit

higher pain thresholds and would be less likely to report

experimental pain stimuli as painful because of their

extensive previous pain experience. In other words, in

comparison to their extensive experience with clinical pain,

the chronic-pain patients no longer label typically painful

sensations as painful, rather, they save that label for the

atypical, higher intensity sensations from their chronic

pain site, with which they have become familiar.

Other studies have supported this adaptation model of

altered pain perception. For example, using signal

detection and threshold measures, Yang et al. (1985) found

that chronic low-back pain patients were less likely to

report radiant-heat stimuli as painful and showed higher

thresholds for reporting radiant-heat pain as compared to









healthy volunteers. In addition, Clark and Clark (1980)

reported that Nepalese baggage porters showed higher

thresholds to noxious experimental electrical stimuli and

were less likely to report the stimuli as painful compared

to western subjects who were used to living more

comfortably. Clark and Clark (1980) concluded that harsh

living conditions increased the stoicism of the Nepalese and

were at least in part responsible for differences in pain

report. In this case, harsh living conditions, rather than

chronic pain, served as the referent previous experience,

but the interpretation remains consistent with the

adaptation model.

Cohen et al. (1983) attempted to explain the

discrepancies in the data supporting the hypervigilance

model versus the adaptation model. In particular, these

authors suggested that the results of the Malow et al.

(1980), and Malow and Olson (1981) studies may reflect

different psychological components of myofascial pain

disorder as compared to chronic low-back pain. In addition,

the studies supporting the hypervigilance model used pain

stimuli that measured the length of time the subjects could

tolerate pressure pain as opposed to ratings of stimulus

intensity as seen in the studies that used radiant heat or

electrical stimuli.

It is important to note that despite these differences

in populations tested, stimuli used, and seemingly

contradictory results, there has been one finding that has









been consistent across studies. The results of the Malow et

al. (1980), Malow and Olson (1981), Naliboff et al. (1981),

Cohen et al. (1983), and Yang et al. (1985) studies all

demonstrated that chronic-pain patients, whether

experiencing myofascial or chronic low-back pain, exhibit a

significant deficit in their ability to discriminate varying

levels of experimental pain stimuli as compared to pain-free

control subjects. Neither the hypervigilance model nor the

adaptation model clearly explain these findings of

discrimination deficits in chronic-pain patients. However,

this deficit is a clear example of a perceptual difference

between chronic-pain patients and pain-free subjects.

There are some studies that have examined whether

perceptual differences in chronic-pain patients, such as a

discrimination deficit, extend to other sensory modalities,

beyond the perception of traditional nociceptive stimuli.

However, the results have been mixed. Seltzer and Seltzer

(1986) used threshold methods to determine the average

distance between two non-painful points of stimulation

applied to the forearm that was necessary for the subjects

to report perceiving two points instead of one. Their

results showed that the chronic-pain patients required a

significantly greater distance between the points before

they reported perception of both points. In a similar test

of two-point threshold, Peters and Schmidt (1991) found no

difference between groups of chronic low-back pain patients

and controls. These authors claim that the results of









Seltzer and Seltzer (1986) were probably due to

methodological flaws.

Despite these conflicting results, other evidence of

non-pain related perceptual differences has been found.

Flor, Schugens, and Birbaumer (1992) found that chronic low-

back pain patients and temporomandibular pain disorder

showed significant deficits in their ability to discriminate

varying amounts of muscle tension, both at the site of their

pain and at peripheral sites, compared to pain-free

controls. These authors suggest that the poor

discrimination seen in these patients represents a basic

deficit, not the result of local changes at the site of the

pain.

Cohen et al. (1981) provide additional evidence for

perceptual differences across sensory modalities. Their

study found that chronic-pain patients had higher thresholds

for reporting discomfort from increasingly louder tones.

The subjects did not exhibit problems discriminating between

the tones in this case, but the results, like the Seltzer

and Seltzer (1986) and Flor et al. (1992) studies, suggest

that chronic-pain patients may have perceptual differences

across sensory modalities, in addition to differences in

perception of painful stimuli.

Even with the evidence of global discrimination

deficits in chronic-pain patients, it is possible that these

perceptual differences exist when experimental stimuli are

applied, but not in the presence of true, clinically









relevant stimuli. For example, Yang et al. (1985) suggest

that chronic low-back pain may increase the level of

endogenous opioids resulting in enough of an analgesic

effect to reduce discriminability on the milder cutaneous

pain from experimental stimuli, but not enough to overcome

the severe clinical pain. In support, Fuller and Robinson

(1993) showed that strenuous exercise, which also raises

central and peripheral levels of endogenous opioids, caused

a similar effect of reducing discriminability of milder

experimental cutaneous stimuli but not higher intensities of

the same stimuli. Based on these data and the suggestion by

Yang and his colleagues, it is possible that chronic pain

may cause mild perceptual disturbance, but the effect is not

enough to influence clinically relevant perceptual

processes.

Signal Detection Theory

Some information about signal detection theory is

useful in understanding its terminology and applications to

the present study. The essence of signal detection theory

is that it provides a mathematical method of separating

perceptual experiences into two distinct measures, one

labelled discriminability and the other response-bias.

Discriminability represents how well a subject can

distinguish varying intensities of a stimulus.

Discriminability requires a certain degree of sensory or

physiological aptitude, especially when stimuli intensities

are close or if a stimulus is just on the edge of threshold.









Response-bias represents a more motivational or cognitively

based decision criteria regarding how willing or likely a

subject is to report having perceived a stimulus.

In order to illustrate the perceptual dimensions of

discriminability and response-bias, Lloyd and Appel (1976)

describe a radar operator who is required to both

discriminate a blip on his radar screen amidst background

static and then decide whether or not to report this

ambiguous blip as enemy aircraft. In one case, the radar

operator is a British man during World War II. This man

would have lenient decision criteria (response-bias)

regarding what he would report as enemy aircraft. He would

tend to report any possible blip as the enemy because the

consequences of missing a true signal would result in the

enemy getting through undetected. His response would be

biased toward reporting the signal. In the second case, the

radar operator lives in peace time. This person, seeing the

same blip, would have much more strict criteria and would

therefore be much less likely to report the blip as enemy

aircraft.

Normally, our perceptual experience is a blend of these

discriminative and psychological or cognitive/motivational

dimensions. Signal detection theory enables researchers to

separate the predominantly physiological discriminative

aspect from the cognitive or psychological components of

perceptual tasks, thereby assessing the differential

contributions of each to the report of any given perception.









Numerous researchers such as Fuller and Robinson

(1993), Janal, Colt, Clark, and Glusman (1984), Yang, Clark,

Ngai, Berkowitz, and Spector (1979), Clark and Mehl (1971),

and the above mentioned Clark and Clark (1980), Malow et al.

(1980), Malow and Olson (1981), Naliboff et al. (1981),

Cohen et al. (1983), and Yang et al. (1985), have applied

signal detection theory to measuring pain perception. The

fundamental concepts of the traditional theory remain the

same within this application. The report of pain is based

on the combination of the ability to discriminate varying

levels of stimuli and the willingness or motivation to

report a given stimulus as being painful.

In a typical experiment involving signal detection and

pain perception, the experimenter uses a device that is

capable of generating a range of discreet and reliably

reproduced stimuli, such as selected temperatures of radiant

heat, which are applied to some part of the subject's body.

The experimenter chooses in advance several intensity levels

of the stimuli that would encompass a range containing some

painful and some typically non-painful stimulus intensities.

The experimenter then presents numerous trials of each

intensity level to the subject; sometimes up to 100 or 200

stimulus presentations per subject. The subject is asked to

rate each stimulus on a categorical scale that typically

ranges from no endorsement of pain to a very strong

endorsement of pain. These ratings are then used to

mathematically calculate values of discriminability (how









well the subject could tell apart the different stimulus

levels) and response-bias (how likely the subject was to

report stimuli as being painful).

Magnitude Matching

Although the signal detection concepts of

discriminability and response-bias will be used in this

study, the traditional method of calculating these measures

will not be used. Instead, these concepts will be

incorporated into a magnitude matching methodology.

Magnitude matching is a procedure defined by Stevens

and Marks (1980), who describe it as a useful method for

comparing groups of normal subjects with groups who may have

pathological sensory modalities. The fundamental concept of

magnitude matching is that each subject is exposed to a

predetermined range of discrete stimuli from two different

sensory modalities. The task of the subject is to provide a

rating of each stimulus on a common scale of sensory

magnitude. For example, in the Stevens and Marks (1980)

study, the subjects were exposed to alternating trials of

standardized light and sound stimuli. The subjects were

instructed to rate "several brightnesses of light and

several loudnesses of a sound by assigning to each

event a number that best seems to match its intensity" (p.

382). The subjects were allowed to use any range of numbers

that they deemed appropriate to the intensity of the stimuli

with the only restriction being that they assume a rating of

zero would represent no sensation experienced.









Once subjects provide their ratings, mathematical

procedures are then used to "normalize" one modality in

terms of the other and generate cross-modality matching

functions. These functions can be plotted on a graph of

normalized magnitude estimations versus units of one of the

modalities, such as decibels in the case of auditory

stimuli. Such a graph represents a linear, or near linear,

plot. Therefore, after collapsing the ratings of all the

subjects within a group, the entire group can be represented

on this graph by a specific line identified by its own slope

and intercept. Group differences can be determined by

calculating differences in slopes and intercepts of the

linear functions generated for each group. Differences can

also be determined by comparing group median or mean ratings

of each of the stimulus levels (Stevens, Plantinga, and

Cain, 1982; Duncan, Feine, Bushnell, and Boyer, 1988) or by

comparing the grand mean for each group of all the subjects'

ratings for all stimulus levels (Feine, Bushnell, Miron, and

Duncan, 1991).

The magnitude matching method has been used to compare

group differences within several sensory modalities

including olfaction (Stevens et al., 1982; Cometto-Muniz and

Noriega, 1985) and taste (Calvino, 1986). With special

import to the present study, magnitude matching has also

been used in assessing pain perception (Duncan et al., 1988;

Feine et al., 1991) and perceived exertion (Marks, Borg, and

Ljunggren, 1983).









As an example of the use of magnitude matching in

assessing pain perception, Feine et al. (1991) compared pain

perception between male and female subjects using this

method. In this study, in order to determine gender

differences in pain perception, male and female subjects

were asked to rate the intensity of painful heat stimuli and

the brightness of visual stimuli on the same scale of

magnitude with zero representing no sensation of pain or no

perception of the visual stimulus. In this case, the visual

stimuli were considered the "control" stimuli, and the heat

was considered the experimental stimuli. The control

modality is assumed to have the same psychophysical function

across all subjects (Marks et al., 1983). Therefore the

subjects' ratings of brightness were used to mathematically

normalize the ratings of the experimental pain stimuli in

order to produce the linear functions characteristic of each

group as described above.

Numerous other control stimuli have been used such as

ratings of sweetness (Cometto-Muniz and Noriega, 1985), and

notably for the present study, auditory tone signals

(Stevens and Marks, 1980; Stevens et al., 1982; Marks et al.

1983). In the Marks et al. (1983) perceived exertion study,

the authors presented sounds of predetermined decibel levels

through headphones. These control stimuli were presented

alternately with 4-minute bouts of exercise at standard work

levels on a stationary bicycle. In a manner consistent with

the magnitude matching method, the subjects were asked to









rate the sounds and their exertion on the same scale of

perceived magnitude.

Duncan et al. (1988) state that the use of magnitude

matching in pain research "should be most appropriate when

comparisons of pain perception or analgesia must be made

between different groups of subjects" (p. 389). The authors

also state that the unrestricted range of ratings used by

the subjects reduces the tendency toward regression bias and

ceiling and floor effects. In addition, the matching of two

distinct sensory dimensions mathematically cancels out the

individual differences in response style (such as having one

subject range from one to ten and another range from zero to

100). This cancelling effect is due to the tendency of

subjects who use small or large ranges on one dimension to

do the same on the other dimension (Stevens and Marks, 1980;

Stevens et al., 1982).

A further advantage to using magnitude matching is that

it has been found to be sensitive enough to detect

differences in pain perception between groups with as few as

five subjects in each group and with far fewer stimulus

trials than required for signal detection calculations,

making this methodology particularly suitable for studying

pain perception (Duncan et al., 1988). In addition, Duncan

et al. (1988) concluded that subjects are quite able to

equate the magnitude of such disparate sensations as the

pain from thermal stimuli and brightness of light.

Furthermore, based on comparisons with a control group, they









found no indication that interposing the rating of an

unrelated stimulus modality interfered with the ability of

the subjects to rate the experimental pain stimuli.

Magnitude Matching in Terms of Signal Detection Theory

As stated above, magnitude matching allows researchers

to determine group differences in perception of a given

sensory modality by comparing the slopes and intercepts of

the linear cross-modality matching functions generated for

each group. It is possible to translate this comparison

into signal detection terminology by viewing slope measures

as representing the overall discriminability of that group,

and intercept measures representing response criterion. For

example, recall the cross-modality plots described above.

Assume the Y-axis is a scale of the range of perceived

magnitude estimates and the X-axis is a scale of the actual

physical stimulus intensities presented, as seen in Figure

1. If one group has a significantly steeper slope (group

A), this would indicate that the subjects in that group

tended to rate the different stimuli as farther apart on the

range of perceived magnitude. Conceptually, this is

identical to the signal detection discriminability measure

which is generally defined as the degree to which a subject

can tell apart different levels of a given stimulus. In

other words, referring to Figure 1, the subjects in group A

perceive bigger differences between stimuli, or discriminate

better, than the subjects in group B.












3





O2





tQ)
-4

o

1 2 3 4
Stimulus Level

Fig. 1. Sample Plot of Discrimination
Difference

To illustrate group differences in response-bias,

Figure 2 shows a line representing one of the groups as

having a significantly higher Y-intercept (group A). This

would indicate that subjects in group A had a general

tendency to rate the experimental stimuli as more intense,

or of a higher magnitude, as compared to group B.

Conceptually, this interpretation is identical to the signal

detection interpretation of response-bias, which would state

that subjects in group A (Figure 2) were more likely to rate

a given stimulus as being of greater magnitude than group B

subjects.

The Feine et al. study (1991), which compared pain

perception between males and females, provides an















4J




4-)

.-* Group B

Oil



0
Z
0 I I t t

1 2 3 4
Stimulus Level

Fig. 2. Sample Plot of Response-bias
Difference

illustrative example of discriminability and response-bias

concepts in a magnitude matching paradigm. In this study,

the authors state that the female group "produced a steeper

stimulus-response function than did males (which

suggests) that females showed a more pronounced

differentiation of the various temperature intensities" (p.

259). In addition, they showed that overall mean normalized

ratings at each temperature level were higher for the female

subjects than the male subjects. Translating the more

"pronounced differentiation" and higher overall pain ratings

into signal detection terminology it could be said that the

women in this study exhibited higher discriminability and









lower response criteria (i.e. greater likelihood to report

pain) than men.

Signal Detection and Magnitude Matching in the Present Study

Whereas the Feine et al. (1991) study, and the other

examples of pain research using magnitude matching, assessed

differences within one modality (i.e. perception of thermal

stimuli), the present study used the same method to assess

differences across two dimensions, one potentially noxious

and one non-noxious modality. The noxious modality was

represented by pain or discomfort elicited by varying weight

levels during individual low-back exercise repetitions. The

non-noxious modality was the judgement of how heavy the

different levels of weight were during a separate set of

low-back exercise repetitions. In this way, it was possible

to determine both how well chronic-pain subjects are able to

discriminate different levels of heaviness, and how well

they are able to discriminate varying levels of pain or

discomfort when lifting different amounts of weight with

their backs. Response-bias, in this experiment, is defined

as a measure of how likely a subject is to label a given

exercise repetition as painful or a particular weight as

heavy.

If the discrimination data from previous research

generalizes to clinically relevant situations, it would be

expected that in this study the chronic-pain patients would

be less able to discriminate the pain and heaviness stimuli.

However, if previous data does not apply to situations that









are more similar to the clinical experience of the patients,

it would be expected in this study that the chronic-pain

patients would be able to discriminate the stimuli as well

as pain-free controls, or perhaps better than controls

considering that the pain patients may be receiving

additional sensory information from their chronic-pain site.

The design of this study also made it possible to

determine whether chronic low-back pain patients use an

adaptation-model response style or a hypervigilant style

when the experimental stimuli more closely approximate the

pain patients typical clinical experience. The

hypervigilance model would predict that the patients would

be cautious and sensitized to sensations in their low-backs

and would provide higher ratings of pain and heaviness

during low-back exercise. The adaptation model would

predict a more stoical response style in which the patients

would not rate the pain during exercise higher than the

ratings of pain-free controls, and they would be more

conservative in their ratings of heaviness.













METHOD

Subjects

There were a total of 32 subjects used in this study,

divided into two groups. The first group was composed of 12

men and 4 women with chronic low-back pain (CLBP). On

average these subjects had been experiencing low-back pain

on a daily or near daily basis for 8.5 years with a range

from 1.5 to 25 years. The CLBP subjects' ages ranged from

22 to 66 years old, with a mean age of 39.3. None of the

subjects were taking narcotic medication at the time of

their participation in the study. Six of the CLBP subjects

were taking non-steroidal anti-inflammatory medications,

three were regularly taking aspirin, four were taking muscle

relaxants, two were taking anti-depressants, one was taking

an anti-anxiety medication. Four of the CLBP subjects were

not taking any medications.

The CLBP group was recruited from a hospital

orthopedics clinic, and a hospital-based rehabilitation

program which specifically serves people with chronic low-

back pain typically derived from previous injury. An

orthopedic physician determined that all the CLBP subjects

had structurally stable spines and were capable of safely

participating in isotonic low-back exercise.

20









The second group was composed of healthy, pain-free

control subjects with no history of back pain in the last

two years. These subjects were selected to match the CLBP

subjects on gender and age. Therefore, the control group

also had 12 men and 4 women with a mean age of 39.2, ranging

from 22 to 65 years of age. Six of the control subjects

were employees of the health center or university where the

study took place, the other ten subjects were recruited from

the local community. See Table 1 for summary descriptives

of both groups.

Of the CLBP subjects, nine had no experience with the

MedX apparatus within the last six months. Due to the

necessary requirements of their rehabilitation, three of the

CLBP subjects had used the MedX for the first time in the

week prior to their participation in the study, and four of

the CLBP subjects had used the MedX approximately 6-8 times

over the two months prior to their participation. Test-

retest pilot data from a group of normal subjects showed no

change in discriminability or response-bias in the pain or

heaviness conditions after one previous experience using the

MedX apparatus.

All subjects were fully informed of the procedures.

They also were informed that they have the option to

withdraw from the experiment at any time without penalty.

All subjects signed written informed consent in accordance

with Institutional Review Board requirements. Each subject

was paid $15 for participating.










Table 1. Summary


N =

Mean


of subject descriptives

CLBP

16 (12M, 4F) N

s.d. Range Mei


Control

16 (12M, 4F)

s.d. Range


Age 39.3 11.4 22-66 yrs 39.2 11.8 22-65 yri

Education 12.8 2.01 10-17 yrs 15.9 1.7 12-17 yr!

Duration of
Pain prob. 8.4 7.9 1.5-25 yrs 0.0 0.0 0

Avg. pain* 4.8 1.8 2.2-7.9 0.0 0.0 0

Present pain* 4.6 2.2 0.8-8.3 0.0 0.0 0

*Average and present pain ratings were measured in
centimeters based on the subjects mark on a 10-cm visual
analogue scale line ranging from "No pain" to "Worst
possible pain."


Apparatus

A MedX low-back exercise apparatus was used to present

trials of four stimuli ranging from 35 pounds to 65 pounds

in ten pound increments. The MedX is a lumbar extension

device that seats subjects in a chair-like position. The

padded back of the chair moves through a range of motion

from 0 degrees of extension to 72 degrees of forward

flexion. One repetition of exercise entails starting at 72

degrees of forward flexion and pushing backwards against the

back pad to 0 degrees extension, which lifts the selected

weight, then lowering the weight by moving back to the

starting position. Femur and lap restraints stabilize the

pelvis and limit the subject's mobility so that only the

lumbar region of the back is the focus of the exercise


s
S


an









movement, precluding use of other major muscle groups that

would otherwise be incorporated into the lift.

Auditory tones, generated from a Hewlett Packard

Function Generator Model 203A, served as the control

stimuli. Tones of 1000-Hz were presented to the subjects

through headphones. Four amplitude levels of the tone were

used between the range of 72 to 90 decibels in 6 decibel

increments (adapted from Stevens and Marks, 1980). Each 6

decibel increment represents a doubling of the intensity of

the tone from one level to the next higher level. The

duration of each tone was approximately 1-sec. The

apparatus was calibrated with a Quest 215 sound-level meter

with an octave band filter and a 2-cc coupler.

A background information questionnaire was used to

collect demographic data. This questionnaire also included

two visual analogue scales. The subjects were asked to use

the scales to rate their present pain intensity and their

average pain intensity. On either end of the 10-cm visual

analogue lines were the verbal anchors "No Pain" and "Worst

possible pain" (see Appendix).

Procedure

After the subjects signed the consent form and

completed the questionnaire, the experimenter read the

following instructions (adapted from Stevens and Marks,

1980; Stevens et al., 1982; and Cometto-Muniz and Noriega,

1985):








"For this experiment, you will be asked to do
two sets of 12 repetitions in the MedX machine
with a ten minute rest in between sets. For the
first set I am going to give you a variety of
weights and ask you to rate the discomfort or
painfulness of each repetition. After each
repetition you will also be asked to rate the
loudness of a tone through headphones. I would
like you to judge the pain/discomfort of the
exercise and loudness of the tones on the same
number scale. Imagine that zero on the scale
represents no discomfort or pain and no sound at
all. The top end of the scale is open, but think
of it as approaching the loudest sound you can
imagine and the worst possible pain you can
imagine. During the exercise I will show you a
card reminding you of these zero points.
Basically, I want you to use bigger numbers for
louder sounds and more painful weights, and
smaller numbers for quiet sounds and weights that
are less uncomfortable, just make sure you think
of the pain and loudness on the same scale. You
can use fractions and decimals if you want.
A number of different tones and weights will
be used. There is no connection between the tone
and weight pairs. In other words, an
uncomfortable weight will not always be followed
by a really loud sound etc., they will all be
mixed up, so you will need to just rate each
repetition and each sound as you experience it but
keeping in mind that you are rating them on the
same scale.
For the second set of 12 repetitions, after a
ten minute rest, you will be asked to do the same
procedure I just described but instead of rating
pain/discomfort I will ask you to rate the
heaviness of the exercise weight. You will still
alternate between weights and tones but you will
be rating heaviness and loudness on the same
scale. In this case the zero point will be 'no
noticeable weight' and again 'no sound at all.'
You should think of the open top end of this scale
as approaching the heaviest weight you can imagine
trying to lift and again the loudest possible
sound.
I will talk you through some practice trials
which will also give you an idea of the range of
weights and tones we will be using."

As noted in these instructions, this study was composed

of two sets of exercise. The first set (Condition I)


represents the magnitude matching of tones with









discomfort/pain ratings. The second set (Condition II)

represents matching of tones with the judgement of the

heaviness of different weights. The conditions were

counter-balanced for order within each group to account for

possible order effects. The instructions to the subjects

were appropriately modified depending on which condition the

subject participated in first.

In both conditions, the stimuli were presented in three

blocks, each block consisting of the four weight stimuli

alternated with the four tones. The order of the levels of

stimuli were randomized within each block, but no stimulus

was repeated until all the other levels were presented.

This procedure resulted in three presentations of each of

the four stimulus levels to be rated by the subjects in each

condition. Therefore, in total for both conditions, each

subject performed 24 repetitions on the MedX alternated with

24 presentations of tones. The range of weights used and

the number of exercise repetitions per condition were chosen

to be within the safe limits of the exercise capacity of the

typical chronic low-back pain patients seen at the hospital

recruitment sites.

Each condition was preceded by two practice repetitions

of tone and weight pairs selected from the range of the

experimental stimuli. The subjects received paraphrased

instructions and additional practice trials if necessary.

Throughout the practice and experimental trials, a card was

placed before the subjects that displayed a written reminder









of the zero points of the scales that they were using in the

condition in which they were currently participating. The

card for Condition I stated "Zero=no sound at all, and

Zero=no feelings of discomfort or pain." For Condition II

the card read "Zero=no sound at all, and Zero=No noticeable

weight."

Analysis

Calculation of the cross-modality matching functions

derived from the magnitude matching procedure was based on

the methods used by Stevens and Marks (1980), Duncan et al.

(1988), and Feine et al. (1991). The procedures were the

same for each condition resulting in separate analyses for

the pain related ratings in Condition I and the heaviness

ratings in Condition II.

In the first step of the computations, the mean rating

of each experimental stimulus level (each of the four weight

levels) was calculated for each subject. For example, from

Condition I, the subject's three ratings of the painfulness

of the 35 pound weight were summed and divided by three, and

likewise this was done with each of the other weight levels.

These values were "normalized" by dividing each of these

mean values for each subject by that subject's grand-mean

rating of all the control stimuli (tones) from that

condition. Therefore, in each condition, each subject

generated four normalized values, one at each weight level.

In total, this resulted in 32 data points (16 per group) at

each of the four weight levels in each condition.









Regression equations were calculated for each group in

each condition with the normalized ratings as the dependent

variable, and weight level (35, 45, 55, and 65 pounds) and

group (experimental or control) as independent variables.

Potential differences in the slopes of the regression lines

were assessed by determining if an additional interaction

variable (group x weight level) accounted for a significant

proportion of the variance beyond what was accounted for by

the individual variables. By this method, if the additional

variance of the interaction variable does not reach

significant levels, it is assumed that a common single slope

can represent both groups equally. Pedhazur (1982) states

that a test of slope differences such as this has a risk of

a type II error (accepting the null hypothesis that there is

no slope difference when in fact there is). Therefore,

Pedhazur (1982) recommends using an alpha level of .10 or

.25 in order to minimize the type II error.

Response-bias differences were determined in one of two

ways. The first method tests for intercept differences by

determining the significance level of the regression

coefficient of the group variable in the original regression

equation for that condition. This method is preferred when

the regression lines for the two groups are determined to be

parallel. The alternate method of testing response-bias

differences was to use t tests to determine group

differences in normalized ratings at each of the four weight

levels.









Additional analyses were completed to test the validity

of the research design. Oneway analyses of variance were

calculated to determine if there were any differences in the

subjects' ratings of the tone stimuli within and between

groups in each condition. Oneway ANOVAs were also used to

determine if there were any differences within conditions or

between groups in the linearity of the regression functions

generated by each subject in each condition. Finally, the

regression equations were used to determine if the slopes of

the lines for each group in each condition were

significantly different from zero.













RESULTS

The mean of the CLBP group's present pain intensity

visual analogue ratings was 4.6-cm on a 10-cm scale (with 10

representing the "worst possible pain"). The CLBP subjects'

ratings ranged from .8 to 8.3-cm. The control group had a

mean rating of 0 with no range. The CLBP group rated their

typical pain intensity as 4.8 with a range of 2.2 to 7.9-cm.

The control group rated their typical pain intensity as 0,

again with no range. These results, in addition to the

demographic data previously described, confirm that

appropriate subjects were selected for both groups. See

Table 1 for a summary of subject descriptives.

The analyses of the tone ratings showed that there were

no differences in the subjects' ratings of the tone stimuli

within and between groups in Condition I (see Table 2).

This indicates that the tone was an adequate control

stimulus which was perceived equally by subjects in both

groups. Therefore, with the tone ratings constant, any

differences in normalized stimulus ratings can be considered

to be due to differences in the subjects' ratings of pain or

heaviness.

Analysis of the Pearson correlation coefficients of the

individual best-fit lines for each subject in each condition
29












Table 2.


Condition


Oneway ANOVA comparison of tone ratings between
groups

I (Pain)


Condition


CLBP

Control

II (Heaviness)


CLBP

Control


Table 3.


Condition


Oneway ANOVA comparison of Pearson correlation


Oneway ANOVA comparison of Pearson correlation
coefficients of regression lines between groups

I (Pain)


Condition


CLBP

Control

TI (Heaviness)


CLBP


Control


Mean

6.01

5.52



Mean

5.65

5.74


F

.04





F

.002


df

1,30





df

1,30


E

.84





p

.96


Mean

.96

.93



Mean

.96

.97


F

3.44





F

.19


df

1,30





df

1,30


p

.07







.66


condition I (Pain)---


condition


m J


condition II (Heavinessl---










showed- that all the subjects. generated. highly linear

functions in both conditions with- no differences between.

groups (see Table 3).

Analysis- of the slopes of the regression equations:

generated by each group showed that in Condition I (pain

ratings) the CLBP group had a significantly steeper slope

than the control group (t=l.8, df=3,124-, p<.10), as seen in

figure 3. In addition, the slopes for both groups were

significantly different from zero (control t=3.48, df=1,62,

p<.001; CLBP t=7.31, df=1,62, p<.0001). This suggests that

when using pain information, the chronic-pain subjects

discriminate the experience of varying stimulus intensities

in their lower-back's better than normal subjects, presumably





"2 CLBP
Z 1.30


ig. 3 C I Control




N




1 2. 3 4
Weight Level
Fig. 3. Condition I: Pain Ratings



















4J Control /










0
a a






1 2 3 4


Fig. 4. Condition II: Heav.ness Ratings



because they are receiving additional sensory information

that the control subjects are not.

In Condition II (heaviness ratings) there was no

difference in slopes between groups (t=.25, df=3,124,

p>.10), as seen in figure 4. The common slope for both

groups was significantly different from zero (t=6.88,

df=2,125, p<.0001). Equal slopes in Condition II suggests

that when the discrimination task focuses away from pain and

onto a still-relevant perceptual dimension in which the two

groups have more in common (e.g. judgement of heaviness),









both groups are able to equally discriminate varying levels

of stimulus intensity.

Analysis of intercept differences showed that in

Condition II (heaviness), the CLBP group had a significantly

lower intercept than the control group (t=-3.45, df=2,125,

p<.001). Figure 4 shows that as a whole, the chronic-pain

subjects tended to systematically rate the weight stimuli as

less heavy compared to the control subjects' ratings.

Since there was a significant difference in slopes in

Condition I, it was not appropriate to calculate response-

bias differences using the intercept comparison method.

However, to determine if there were differences in pain

ratings at any stimulus level, individual t test comparisons

were calculated between the normalized pain ratings of each

group at each of the four weight levels. These analyses

showed no difference in pain ratings between groups at any

of the weight levels [stimulus 1 (35 lbs.) t=-.79, p>.40;

stimulus 2 (45 lbs.) t=-.55, p>.50; stimulus 3 (55 lbs.) t=-

.06, p>.90; stimulus 4 (65 lbs.) t=1.59, p>.12; df=30].

Overall, these data show that the chronic-pain subjects

were better at discriminating pain sensations during

exercise (i.e., pain patients had steeper slopes in

Condition I), but they did not tend to rate the exercise as

being any more painful than the normals reported their own

experience to be, at least within the range of chosen weight

levels (i.e. no differences found in t tests of pain ratings

between groups at any weight level). In addition, the






34


chronic-pain subjects were equal to normals in their ability

discriminate varying levels of heaviness (the slopes of the

groups' heaviness ratings were parallel in Condition II),

but as a group the chronic-pain subjects rated the weights

to be less heavy compared to the control subjects' ratings

(the CLBP group had a significantly lower intercept).













DISCUSSION

The response-bias data from this study best address the

question of the appropriateness of the hypervigilance model

versus the adaptation model when chronic-pain patients are

exposed to clinically relevant stimuli. The hypervigilance

model would predict that CLBP subjects would show elevated

ratings of painfulness during low-back exercise, which was

not the case in this study. In addition, if chronic-pain

patients were hypervigilant with respect to non-pain related

(but still clinically relevant) stimulus modalities, it

would be expected that the subjects in this study would rate

the stimulus weights as heavier compared to the control

subjects' ratings. This prediction also was not supported

by the present data. The chronic-pain subjects in this

study did not rate the stimuli as more painful, and their

ratings of heaviness were actually lower than those of the

controls, in direct contrast to hypervigilance model

predictions.

The adaptation model may fit better with the present

data, consistent with other studies that have used CLBP

subjects. The adaptation model would predict that the

chronic-pain subjects in some way would be more stoic than

the control subjects. In the present study, there was no

35









difference in the ratings of pain during exercise between

the groups despite the fact that the CLBP subjects were

exercising the very part of their bodies that gives them a

great deal of pain. Since previous studies have shown CLBP

subjects to provide lower pain ratings than controls when

using traditional experimental pain stimuli, the present

findings of no difference in pain ratings may only represent

indirect evidence of the stoic, under-reporting style

predicted by the adaptation model. However, as noted above,

this under-reporting effect seemed to generalize to a

relevant but non-pain related modality, namely the

perception of heaviness of weights lifted with the painful

body part as seen in Condition II. This phenomenon, not

previously reported in other studies, may be additional

evidence for an adaptation style of responding in chronic-

pain patients.

In addition to response-bias issues, the experimental

literature has also hypothesized that chronic-pain subjects

have a perceptual discrimination deficit. As stated

previously, studies that have used traditional experimental

pain stimuli have shown chronic-pain patients to be less

able to discriminate varying levels of stimulus intensity.

The present study found no discrimination deficit when

clinically relevant stimuli were used. In this study, the

chronic-pain subjects actually were better able to

discriminate sensations in their affected site when they

were told to use discomfort or pain information in making









their discrimination. In view of this discrimination data

combined with the t test pain-rating comparisons, it appears

that the pain patients may effectively use pain information

to discriminate sensations in their affected site, but they

do this without rating the stimuli as being generally more

painful.

The findings of this study also stand in contrast to

the suggestion that chronic-pain patients may have a global

discrimination deficit that extends to non-pain perceptions.

When the discrimination task required a non-pain related

judgement of clinically relevant stimuli (heaviness of

weights lifted with their lower-backs), the chronic-pain

subjects were equally accurate at discriminating the

heaviness of the weights, even though they systematically

underestimated the weight.

It is important to note that the difference in

discriminability in Condition I was not due to a limited

range of ratings on the part of the control subjects. One

of the risks of using weight stimuli that are within the

typical range of weights for beginning rehabilitation

patients is that the stimuli may not adequately test the

range of discomfort or pain in the control subjects.

However, the range of normalized pain ratings for the

control group in this study was 0 to 2.67, s.d.=.76. The

chronic pain group ranged from 0 to 3.93, s.d.=.75. In

addition, the high Pearson r correlation for the controls

and the non-zero slope of the control group regression line









suggest that the control group provided an adequate range of

ratings.

Overall, the present data suggest that chronic-pain

patients have a distortion of clinically relevant

perceptions, but it is not the same distortion that is seen

with the traditional less relevant experimental pain

stimuli. Instead of a global discrimination deficit and a

bias toward under-reporting pain, the chronic-pain patients

seem to have a global bias toward under-reporting relevant

stimuli (as seen by their underestimation of heaviness and

what may be a stoic pain reporting style) and a superior

ability to discriminate sensations from their affected site.

One issue to be addressed concerns how the present

results can be explained in the context of the previous

experimental literature. As mentioned previously, Yang et

al. (1985) have suggested that the body responds to chronic-

pain by increasing the level of endogenous opioids, which

produces an analgesic effect on milder cutaneous pain but

not on the more severe clinical pain. This would suggest

that the decreases in discrimination found in chronic-pain

patients in laboratory settings has more to do with the

artificial nature of the experimental pain stimuli, whereas

in their daily lives the patients actually are very

sensitive to relevant pain stimuli. This conclusion is

supported by the present study which showed that chronic-

pain subjects discriminated pain-relevant stimuli better

than normals.









Another explanation may be that instead of increasing

endogenous opioids, the chronic-pain adds "noise" to the

pain-perception system, to use signal detection theory

terminology, making it more difficult to discriminate other

nociceptive stimulation that is less relevant or less

intense. Therefore, the additional stimulation, or noise in

the nociceptive system, from chronic-pain may have caused

discrimination deficits in the studies that used traditional

experimental stimuli, but not in this study which used

stimuli more directly relevant to their clinical pain

problem.

The source of the perceptual underestimation, or under-

reporting seen on the part of the CLBP subjects in this

study, remains unclear. Since response-bias is generally

viewed as a more cognitive or psychologically based measure,

perhaps there are cognitive or psychologically related

factors associated with this response style seen in chronic-

pain patients. For example, it may be that the true under-

estimation seen in Condition II is not of heaviness in

general, but rather the heaviness the pain patient thinks he

or she is capable of lifting. This would result in an over-

limiting of activities due to the patient's underestimation

of the work they are actually capable of doing. In the

context of the present study, if the CLBP subjects

underestimate their capacity and yet find that they are

capable of lifting all the stimulus weights, they would

assume that the stimulus weights must be lower, within their









capacity, resulting in overall lower ratings in the

heaviness condition.

Possibly chronic-pain patients have other cognitive-

sets regarding how much a given activity will hurt. They

may think that if they attempt to lift a certain weight the

pain would be too great and they would be unable to

successfully complete the lift. In this study, when the

subjects found that they were able to lift the weights, they

may have been biased toward lower ratings of pain to

maintain their cognitive-set.

Another possible influence on response-bias may be the

cognitive correlates and overall suppressed responsiveness

often seen with depression, which is common among chronic-

pain patients. Future research may need to focus on

determining if psychological factors correlate with this

under-reporting style. If so, this would have ramifications

for rehabilitation by emphasizing the necessity to address

psychological issues related to chronic pain.

Some ramifications for rehabilitation from the present

data are already evident. If CLBP patients tend to

underestimate how heavy an object is during a lift, then

they are at risk of lifting a weight that may be too heavy

and potentially harmful to their injury site. In addition,

a cognitive under-estimation of their capabilities before a

lift or activity, may lead them to over-limit themselves and

not make attempts at activities that are actually within

their capacity.









In rehabilitation, it is common for CLBP patients to be

told they must not exceed a restricted range of weights.

Rehabilitation professionals may find value in training

chronic low-back pain patients to recognize specific weights

within their range of safety and true capacity rather than

providing a verbal range of weights and relying on the pain-

patients' judgements of appropriate heaviness.

The discrimination difference found in the pain

condition may have less direct clinical application. It

makes intuitive sense that the additional pain information

the chronic-pain subjects receive would help them to

discriminate better than subjects who are receiving less

information on which to base their discrimination.

However, this additional information available to the

chronic-pain patients may not add any benefit to their

performance of functional activities. Post-hoc analyses of

the data from this study showed that the slopes and

intercepts of the pain patients' regression equations for

pain ratings and heaviness ratings were not significantly

different. This means that there was no additive effect in

which pain information improved the pain-subjects'

judgements of heaviness. In fact, as shown above, being a

pain-patient was related to a poor ability to judge the

heaviness of the weight stimuli.

To summarize, the data from this study suggest that

people with chronic low-back pain exhibit perceptual

differences when exposed to clinically relevant stimuli of









both pain-related and non-pain related dimensions. In

contrast to previous studies that have used less relevant

stimuli, the CLBP patients in this study were able to

discriminate relevant pain-related stimuli better than

normals, and they exhibited no discrimination deficit in

relevant non-pain perceptions of heaviness. In addition,

the CLBP subjects showed no difference in pain ratings

during low-back exercise compared to pain-free controls,

which is suggestive of the stoic, under-reporting response

style of the adaptation model. This under-reporting

tendency also generalized to relevant non-pain related

ratings, as seen in the consistent underestimation of

heaviness by the chronic-pain subjects when lifting weights

with their lower backs.

In conclusion, this study tested the external validity

of previous research that has investigated perceptual

differences in chronic-pain patients, which has typically

used traditional experimental pain stimuli. The present

study used stimuli that more closely approximated CLBP

patients typical experience while still maintaining the

experimental control seen in previous research. In

addition, the present methodology represents an initial

conceptual bridge between two psychophysical methods, namely

signal detection theory and magnitude matching. With this

bridge, direct comparisons are possible between findings

from previous SDT pain research and data from the present

and future studies using magnitude matching. A trend toward






43


using magnitude matching in measuring pain perception in

clinical populations has advantages, especially if

clinically relevant stimuli are used, since magnitude

matching requires fewer pain trials and can be used with

smaller groups of subjects. Future attempts at replicating

the present study using similar methods, relevant stimuli,

and a variety of subject populations will be important to

ensure the reliability of the present findings.












APPENDIX


Background Information Sheet


Name:

Sex (circle): M F

Age:


1. Education (circle highest completed):

1 2 3 4 5 6 7 8 9 10 11 12 college (1 2 3 4) post-college


2. Length of current back pain problem (specify months or
years):


3. Are you currently taking narcotic medication? Y N

4. Please list current medications:




5. Please put a mark on the line below in a place that you
think best represents your present pain intensity:


No Pain


Worst
Possible Pain


6. Please mark your average or typical pain intensity:


No Pain


Worst
Possible Pain













REFERENCES


Calvino, A. (1986). Perception of sweetness: The effects of
concentration and temperature. Physiology and Behavior,
36(6), 1021-1028.

Chapman, C. (1978). Pain: the perception of noxious events.
In R. Sternbach (ed.), The psychology of pain. New York:
Raven Press.

Clark, W., & Clark, S. (1980). Pain responses in Nepalese
porters. Science, 209, 410-412.

Clark, W., & Mehl, L. (1971). Thermal pain: A sensory
decision theory analysis of the effect of age and sex on
d', various response criteria, and 50% pain threshold.
Journal of Abnormal Psychology, 78, 202-212.

Cohen, M., Naliboff, B., Schandler, S., & Heinrich, R.
(1983). Signal Detection and threshold measures to loud
tones and radiant heat in chronic low back pain patients
and cohort controls. Pain, 16, 245-252.

Cometto-Moniz, J., & Noriega, G. (1985). Gender differences
in the perception of pungency. Physiology and Behavior,
34, 385-389.

Duncan, G., Feine, J., Bushnell, M., & Boyer, M. (1988). Use
of magnitude matching for measuring group differences in
pain perception. In R. Dubner, G. Gebhart, & M. Bond
(Eds.), Proceedings of the Vth World Congress on Pain
(pp. 383-390). Amsterdam: Elsevier Science Publishers.

Feine, J., Bushnell, M., Miron, D., & Duncan, G. (1991). Sex
differences in the perception of noxious heat stimuli.
Pain, 44, 255-262.

Flor, H., Schugens, M., & Birbaumer, N. (1992).
Discrimination of muscle tension in chronic pain patients
and healthy controls. Biofeedback and Self Regulation,
17, 165-177.

Fuller, A., & Robinson, M. (1993). A test of exercise
analgesia using signal detection theory and a within-
subjects design. Perceptual and Motor Skills, 76, 1299-
1310.








Helson, H. (1964). Adaptation-level theory: An experimental
and systematic approach to behavior. New York: Harper &
Row.

Janal, M., Colt, E., Clark, C., & Glusman, M. (1984). Pain
sensitivity, mood and plasma endocrine levels in man
following long distance running: Effects of naloxone.
Pain, 19, 13-25.

Lloyd, M., & Appel, J. (1976). Signal detection theory and
the psychophysics of pain: An introduction and review.
Psychosomatic Medicine, 38, 79-94.

Malow, R., Grimm, L., & Olson, R. (1980). Differences in
pain perception between myofascial pain dysfunction
patients and normal subjects: A signal detection
analysis. Journal of Psychosomatic Research, 24, 303-309.

Malow, R., & Olson, R. (1981). Changes in pain perception
after treatment for chronic pain. Pain, 11, 65-72.

Marks, L., Borg, G., & Ljunggren, G. (1983). Individual
differences in perceived exertion assessed by two new
methods. Perception and Psychophysics, 34(3), 280-288.

Naliboff, B., Cohen, M., Schandler, S., & Heinrich, R.
(1981). Signal detection and threshold measures for
chronic back pain patients, chronic illness patients, and
cohort controls to radiant heat stimuli. Journal of
Abnormal Psycholoyv, 90(3), 271-274.

Peters, M., & Schmidt, A. (1991). A comparison of two-point
discrimination threshold of tactual, non-painful stimuli
between chronic low back pain patients and controls.
Pain, 44, 57-60.

Rollman, G. (1979). Signal detection theory pain measures:
Empirical validation studies and adaptation-level
effects. Pain, 6, 9-21.

Seltzer, S., & Seltzer, J. (1986). Tactual sensitivity of
chronic pain patients to non-painful stimuli. Pain, 27,
291-295.

Stevens, J., & Marks, L. (1980). Cross-modality matching
functions generated by magnitude estimation. Perception
and Psychophysics, 22(5), 379-389.

Stevens, J., Plantinga, A., & Cain, W. (1982). Reduction of
odor and nasal pungency associated with aging.
Neurobiology of Aging, 3, 125-132.







47

Yang, J., Clark, W., Nagai, S., Berkowitz, B., & Spector, S.
(1979). Analgesic action and pharmokinetics of morphine
and diazepam in man: An evaluation by sensory decision
theory. Anesthesiology, 51, 495-502.

Yang, J., Richlin, D., Brand, L., Wagner, J., & Clark, W.
(1985). Thermal sensory decision theory indices and pain
threshold in chronic pain patients and healthy
volunteers. Psychosomatic Medicine, 47(5), 461-468.













BIOGRAPHICAL SKETCH

Adam Kittinger Fuller was born on July 5, 1963, in

Bartlesville, Oklahoma. He was raised in Brockport, New

York, where he had particular interests in soccer and

travel. In 1984, he graduated from St. Lawrence University

in Canton, New York, with a Bachelor of Science degree in

psychology. After four years of working in the field of

head-injury rehabilitation, he was admitted to the graduate

program in clinical psychology at the University of Florida,

where he received a degree of Master of Science in 1990.








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.





lichael Robinson, Chair
Assistant Professor of Clinical
and Health 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.





Cyn a Belar
Professor of Clinical and Health
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.





ussell Bau
associate Professor of Clinical
and Health 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.





Michael Geisser
Assistant Professor of Clinical
and Health 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.





Linda Shaw
Assistant Professor of
Rehabilitation Counseling


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 1994 C
Dean, College of Heaglh Related
Professions


Dean, Graduate School







































UNIVERSITY OF FLORIDA
II I II I I 1 111 I lI I 11111
3 1262 08557 0470