Lumbar strengthening in chronic low back pain patients


Material Information

Lumbar strengthening in chronic low back pain patients psychological and physiological benefits
Physical Description:
viii, 109 leaves : ; 29 cm.
Risch, Sherry V., 1953-
Publication Date:


Subjects / Keywords:
Back Pain -- therapy   ( mesh )
Exercise Therapy   ( mesh )
Lumbosacral Region   ( mesh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1990.
Includes bibliographical references (leaves 89-97).
Statement of Responsibility:
by Sherry V. Risch.
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:
oclc - 22534541
ddc - 617.564062
System ID:

This item is only available as the following downloads:

Full Text







This dissertation is dedicated to my husband, E. David

Risch, and my two children, Valerie and Kristopher, who have

encouraged and supported my efforts over the years. Their

faith and understanding will always be cherished.


I wish to express my appreciation and thanks to my

chairperson, Nancy Norvell, for working with me on this

study and affording me the opportunity to complete it under

her supervision. Most importantly, I appreciate her

continued support and her contributions of advice and

expertise in this study's formulation and completion. In

addition, a special thanks are extended to my committee

members, Michael Robinson, Anthony Greene, James Johnson,

and Michael Pollock, for their support and supervision in

this dissertation preparation. I also extend my thanks and

gratitude to the staff of E. David Risch and to the staff of

the Center for Exercise Science for their assistance in the

completion of this treatment study. Mostly, a special

thanks are extended to my husband and children for their

support and understanding while I worked on this

dissertation. Without their love and reassurance, I could

not have persevered in completing this work.




ACKNOWLEDGEMENTS.................................... iii

LIST OF TABLES..................................... vi

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

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

Chronic versus Acute Pain................. 2
Treatment of Chronic Pain................. 9
Self-Efficacy and Decreased Avoidance of
Physical Activities..................... 20
Physical Reconditioning as a Way of
Improving Muscular Strength and
Psychological Symptoms.................. 28
General Effects of Exercise on
Psychological Functioning............ 30
Exercise in Chronic Low Back Pain
Patients............................. 33
Exercise of the Low Back.................. 35
Hypotheses............................... 40

2 METHOD................................... .. 43

Subjects................................... 43
Equipment and Measures.................... 44
MedX Assessment and Training ........... 44
Self-Report Questionnaires............. 45
Procedures ............................... 51
Experimental Group..................... 52
Wait-List Control Group................ 52
Statistical Analysis................... 53

3 RESULTS.................................... 56

Physiological Results..................... 58
Psychological Results..................... 59

4 DISCUSSION ...... ......................... 70

Physiological Findings.................... 72
Psychological Findings.................... 74
Implications............... ....... ....... 80
Limitations of the Study.................. 84
Conclusions and Future Directions......... 87

REFERENCES......................................... 89

APPENDICES..................................... .... 98

A PAIN QUESTIONNAIRE....................... 98


BIOGRAPHICAL SKETCH................................ 109





STANDARD DEVIATIONS................... 64

VALUES............ ................... 68


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



Sherry V. Risch

May 1990

Chairman: Nancy K. Norvell, Ph.D.
Major Department: Clinical and Health Psychology

This study examined the effects of dynamic exercise on

the isometric strength of the lumbar extensors in a chronic

low back pain population. It was hypothesized that exercise

on the MedXTM machine would result in increased lumbar

strength which would be associated with decreases in pain

and physical and psychosocial impairments.

Fifty-five chronic low back pain subjects participated

in the study. Thirty-one patients were randomly assigned to

a treatment group and 23 patients were assigned to a wait-

list control group. Prior to participating in the study,

all subjects were tested on the MedXTM machine for isometric

strength and were administered psychological questionnaires

addressing physical and psychosocial dysfunction, pain,

stress, depression, and anxiety.

Treatment consisted of variable resistance dynamic

exercises two times each week for six weeks, followed by one

session each week for six weeks. The control group was

requested to make no changes in their current life-style or


methods of treating their pain. All subjects were re-tested

on the psychological questionnaires and re-tested for

isometric strength on the MedXT at the end of 12 weeks.

Multivariate analysis of covariance of change scores

revealed that the treatment group significantly improved in

strength as compared to the control group. Increased

strength was associated with a significant decrease in pain

and in physical and psychosocial dysfunction. In contrast,

the control group increased in reported pain and perceived

physical and psychosocial dysfunction. There were no

significant differences between groups on measures of

depression, anxiety, and stress.

Pre-treatment measures indicated that both groups

reported high levels of self-efficacy (internal locus of

control) prior to entering the treatment protocol, but post-

treatment analysis indicated that the control group changed

to decreased self-efficacy expectations (an external locus

of control). The treatment group maintained a high self-

efficacy rating post-treatment.

The findings of this study support the hypothesis that

exercise on the MedXTM machine is a beneficial treatment for

chronic low back pain patients. There is evidence that

increasing the strength of the low back muscles decreases

pain and improves physical and psychosocial functioning.

Self-efficacy expectations were significantly related to

treatment gains.



The importance of rehabilitating patients with low back

pain is evident from reports which estimate that 80% of the

general population and 85% of the industrial population will

at some time experience low back pain (Feuerstein, Papciak,

& Hoon, 1987; Mayer, Gatchel, Kishino et al., 1986;

Rosomoff, 1985; Schuchmann, 1988). More specifically, this

disabling condition has partially disabled five million

people in the United States and accounts for approximately

93 million lost work days resulting in a sixteen billion

dollar economic loss each year (Block, 1982; Snook &

Webster, 1987).

Low back pain is commonly associated with injuries of

the nerves, bone, joint or ligaments, with resulting

myofacial syndromes and degenerative diseases of the spine

(Feuerstein et al., 1987). Of note, a high percent of the

patients with low back pain appear to suffer from soft

tissue injury in the lumbar spine area, with 78% of disabled

low back pain patients having no demonstrable

pathophysiological basis for their pain (Bono & Zasa, 1988;

Loeser, 1980; Mayer et al., 1986). Thus, treatment of the

low back pain patient focuses not only on spinal

abnormalities but also on the surrounding soft tissue areas.

Chronic versus Acute Pain

The experience of pain is described as acute or

chronic. Acute pain is often defined as a neural response

to a perceptual experience caused by a noxious stimulus.

This experience then leads to suffering and negative

affective states. These pain-eliciting states are

interpreted by the individual as signals of a problem that

requires intervention (Bono & Zasa, 1988). Hence, the

experience of acute pain is believed to be correlated with a

physiological process and the associated pain subsides when

the symptoms are resolved.

Chronic pain is different from acute pain because the

symptoms are present and persistent for an extended period

of time. Chronic pain has been arbitrarily differentiated

from acute pain by establishing a time frame of

approximately six months. This time criteria was derived

from research by Sternbach (1974) which found personality

differences between patients who suffered pain for less than

six months compared to those who suffered for more than six

months. Patients who suffered longer durations of pain were

noted to have higher elevations on the clinical scales of

the Minnesota Multiphasic Personality Inventory (MMPI),

suggesting increasing psychological distress with increasing

durations of pain.

Chronic pain patients have been described as being

poorly psychologically adjusted. As the chronicity of pain

exceeds six months, these individuals evidence behavioral

changes (i.e., restricted activities), emotional distress

(i.e., anxiety or depression), heavy medication usage, and

increased reliance on the health care system. These factors

as well as many others are often explained in the literature

as a result of increased psychological distress created by

the patients' long-term disabilities in relation to their

psychosocial and economical environments (Bono & Zasa, 1988;

Urban, 1982). For example, Ackerman and Stevens (1989)

studied 110 patients presenting for treatment of acute and

chronic pain. These authors found that both groups

experienced increased levels of state anxiety, but the

chronic pain group was significantly differentiated from the

acute pain group in their experience of a negative affective

state (e.g., depression) which was again believed to be

associated with the chronicity of pain.

Bono and Zasa (1988) further differentiated the chronic

pain experience according to the individual's adjustment and

reaction to his/her pain. These authors reported that not

all chronic pain patients fail to adapt to their

difficulties and some continue to function well in society.

They identified patients with "chronic pain" as those who

are without deleterious psychosocial/environmental

disruptions, and those patients who are dysfunctional, as

exhibiting a "chronic pain syndrome."

The chronic pain syndrome has been associated with

western medicine and industrialized civilizations. Waddell

(1987) commented that prior to modern medicine and

industrialization, the prevalence and incidence of back pain

was the same as it is now, but those individuals did not

stop daily activities or become permanently disabled. He

postulated that seeking health care for chronic pain is a

function of the individual's perception and interpretation

of physical symptoms and a function of the availability and

expectations of medical treatment. Hence, the chronic pain

syndrome has been associated with psychological distress,

depression, the experience of failed treatment

interventions, and a sick role adaptation; not as a response

to physiological activity.

There are many theories that attempt to describe and

define the chronic pain experience. The original theories

of pain were based on a somatosensory model. This model

postulated that feelings of pain were directly proportional

to peripheral damage (Blackwell, Galbraith, & Dahl, 1984;

Tursky, Jamner, & Friedman, 1982). However, current

theories of pain disagree with this model. More recent

theories propose that the subjective feeling of pain is not

directly related to tissue damage, but instead, results from

an interaction of physical, socio-environmental, and

psychological experiences (Bandura, O'Leary, Taylor,

Gauthier, & Gossard, 1987; Craig, 1983; Turk, Meichenbaum, &

Genest, 1983; Weisenberg, 1977).

An often-cited theory of pain is Melzack and Wall's

(1965) Gate Control Theory. This theory postulates a

neurophysiological basis for the role of psychological

factors in the pain experience. It describes the

interaction between sensory-discriminative, affective-

motivational, and cognitive-evaluative systems (Turk & Flor,

1984; Turskey et al., 1982; Melzack & Dennis, 1978). The

model incorporates the importance of psychological factors

in the central nervous system's modulation of pain. The

perception of pain is said to be influenced by the

individual's attitudes and attentions which act through a

central control mechanism. The control mechanism operates

via descending neural fibers which modulate the excitation

of afferent pain receptors in the dorsal horn of the spinal

column. Opening and closing of this control mechanism

increases or decreases sensory sensitivity. In summary,

suffering with pain is correlated with psychological states

that exacerbate the neurophysiological mechanisms (Clark &

Yang, 1983; Melzack & Dennis, 1978; Turk & Flor, 1984;

Tursky et al., 1982).

In support of this model of pain, experimental studies

have shown that reported pain severity and reactions to pain

are related to cultural variables (Sternbach & Tursky,

1965), to the influence of social modeling (Craig, Best, &

Ward, 1975), to the mood states of the patient (Clark, 1974;

Turk & Kerns, 1984), to previous learning and medical

management (Fordyce, 1988), and to a physiological

hyperactivity (Flor & Turk, 1989).

For example, Sternbach and Tursky (1965) interviewed 60

subjects from four different ethnic backgrounds and obtained

information regarding the subject's past reactions and

perceptions to pain. Subjects then participated in a

laboratory experiment which tested pain thresholds,

estimated stimulus intensities, and autonomic reactivity.

Findings of this study indicated that different subcultural

groups hold different attitudes (e.g., "pain is to be

suffered in silence" versus "fear the worst") about pain and

that based on these attitudes, they experience pain


Another experimental study by Craig et al. (1975)

exposed 30 undergraduate students to various levels of

shock. This study found that subjects differed in their

reported experience of pain as a function of being exposed

to a pain-tolerant model versus a pain-intolerant model.

Subjects exposed to a pain-intolerant model reported more

pain. Hence, this study lends support not only for the

cognitive-mediation in the experience of pain, but also in

the influential role of social-environmental mediators.


Several studies have found that psychological variables

are frequently associated with reports of pain, especially

depression and anxiety. Not only do clinically depressed

patients frequently report pain symptoms, but pain patients

frequently report depressive symptoms (Weisenberg, 1977).

For example, a study comparing 33 chronic low back pain

patients to 35 healthy controls found that depression and

anxiety were commonly reported in the low back pain group,

not in the control group. It was further noted that the

severity of depression and anxiety were positively

correlated with increased pain reports (Feuerstein, Sult, &

Houle, 1985). The role of anxiety in pain is further noted

in Weisenberg's (1977) statement that patients who trust

their physicians reduce their presenting anxiety and this

leads to decreased pain reports and increased responsiveness

to placebo treatments. For example, laboratory studies

utilizing psychological procedures such as attention

diversion, cognitive restructuring, and hypnotic inductions

have been found to significantly reduce anxiety and

subsequently the experience of pain (Bandura et al., 1987;

Spanos, Perlini, & Robertson, 1989), Additionally,

psychological measures of depression and anxiety decrease

significantly with decreased pain after treatment (Gatchel,

Mayer, Capra, Diamond, & Barnett, 1986).

There are other theories and models attempting to

explain the experience of chronic pain. For example,

researchers have studied pain with psychophysical

definitions of pain threshold levels, with intensity

ratings, with magnitude estimations (utilizing cross

modality matching scales for intensity and quality of pain),

and with observable behaviors for discrimination and

detection of pain (Chapman, Casey, Dubner, Foley, Gracely &

Reading, 1985). Additionally, chronic pain has been

described as a psychobiological disorder; as a symptom in

which pain fulfills the needs of a dysfunctional family

system; as a classical conditioning paradigm or respondent

model of pain in which a pain-tension cycle is created; as

an operant model where the pain is believed to be under the

control of contingencies of reinforcement; and finally, as a

diathesis stress model where physical, psychological and

social factors are believed to have preconditioned the

patient for hyper-reactivity with a responsive stereotypic

style diathesiss) resulting from a genetic predisposition

(Chaturvedi, Varma & Malhotra, 1984; Turk & Kerns 1984).

Theories of pain sensation and sensitivity are

beneficial in explaining the nociception of experimentally

induced pain, but they fall short in describing the clinical

patient presenting with the chronic pain syndrome.

Laboratory studies have successfully documented the role of

endogenous opioids and cognitive mediational processes in

the control of pain (Bandura et al., 1987; Yang, Richlin,

Brand, Wagner, & Clark, 1985), but as mentioned earlier,

there is a lack of understanding between chronic pain and

the chronic pain syndrome (Bigos & Battie, 1987). Hence,

theories of pain that incorporate psychological, physical,

and environmental influences are more proficient at

explaining the chronic pain experience.

In summary, there are many models and explanations of

acute and chronic pain experience, most of which include

psychological and social phenomena as a prominent factor.

Therefore, a biopsychosocial conceptualization of chronic

pain (which incorporates the role of physiological,

social/environmental, and psychological responses) is viewed

as the predominant working model (Waddell, 1987).

Treatment of Chronic Pain

Interventions and treatment of the chronic pain patient

vary depending on the practitioner's theoretical formulation

of the chronic pain syndrome. Traditionally, medical

treatments have consisted of conservative management which

included bed rest, traction and medication, or surgical

interventions. Unfortunately, these treatment modalities do

not appear to resolve the pain for a significant majority of

chronic pain patients. For example, short-term pain relief

often follows surgery, but for some patients, surgery does

not provide long-term pain relief. Gottlieb et al. (1977)

cited unpublished data by Shealy and Beckner (1975) stating

that 30% of the patients undergoing a traditional

neurosurgical procedure failed to experience post-operative


pain relief and at the end of five years, 90% of the surgery

patients failed to experience satisfactory pain relief. In

a study of conservative medical treatment, Finneson (1973)

reported findings that only 50% of the population studied

reported significant pain relief with conservative treatment

after three years. Thus, it was concluded that traditional

medical treatments did not exceed a 50% success rate and,

therefore, treatment successes could not be distinguished

from spontaneous remission rates.

Fordyce (1988) reported a prospective study of acute

low back pain patients in which he addressed differential

treatment recommendations as they effected long-term

outcome. Patients presenting to a hospital emergency room

with acute low back pain were randomly assigned to one of

two groups. The first group (Group A) received open ended

instructions such as to take medication as their pain

dictated, to exercise when their pain subsided or as

tolerated, and to remain in bed for an unspecified time. In

contrast, the second group (Group B) was instructed to take

medication at a fixed time interval and for a specified

duration, to remain in bed for only a specific period of

time, and to exercise according to a pre-set regimen. At

six weeks follow-up, there were no differences between the

groups, but at 9 to 12 months follow-up there were

significant differences found. Group A (open-ended

instructions) was found to report more physical impairments,


disrupted vocational status, increased levels of health care

utilization, and higher levels of pain. The conclusions

drawn from this study were that subjects with vague open-

ended guidelines for treatment experienced increased pain

from muscular disuse and then potentially interpreted their

pain as a failed healing process. These findings are of

particular import in that they document the role of

expectancy in the efficacious treatment of acute pain and

its relationship to the later development of chronic pain.

This is consistent with Waddell's (1987) contention that

physical illness, illness behavior and psychological

distress combine to produce disability. Bigos and Battie

(1987) reached a similar conclusion stating that there is an

important difference between treating back pain and

preventing chronic back pain disability. Hence, the

interaction between physical illness, pain behaviors and

psychological states may determine the outcome of treatment

of acute pain, while continued disability, work loss, and

failure to return to work may be more related to social

factors than physical disease (Waddell, 1987).

In order to better address and treat the chronic pain

phenomenon, many treatment interventions focus on

psychological variables associated with chronic pain. In

addition to the subjective experience of pain, chronic pain

patients frequently report psychological distress, including

symptoms of anxiety and depression. Thus, as Gottlieb et


al. (1977) state, the patient's distress may be as important

as any structural damage in treating the chronic pain

patient. Treatments with a physical-psychological focus

often use operant conditioning principles that attempt to

extinguish pain behaviors by reinforcing "well behavior" and

ignoring "sick behavior." Within this treatment protocol,

there is a strong emphasis on increasing the patient's

physical activities (well behavior) and helping the patient

develop coping strategies which will lead to reductions in

sick behaviors (Block, 1982; Keefe & Gil, 1986; Turk, Wack,

& Kerns, 1985). This treatment approach endorses the notion

that chronic pain behaviors potentially come under the

control of environmental contingencies. For example, if

pain is experienced following the lifting of a heavy object,

avoiding lifting that object will decrease the pain

(respondent conditioning). With the continued avoidance of

lifting the heavy object, the behavioral changes become

operant in nature and may generalize to lifting all objects

or even to the entire work place. A reinforcement

contingency then develops in the absence of the original

pain-eliciting event. Hence, the individual generalizes

pain behaviors to other situations. A secondary

reinforcement contingency may also develop from potential

positive reinforcement received by others deriving from

their changed pain behavior patterns operantt conditioning)

(Bono & Zasa, 1988). The operant treatment approach,


therefore, attempts to extinguish pain behaviors by changing

the contingencies of reinforcement to well behaviors.

Specific goals of these physical-psychological

treatments are to reduce pain behaviors such as medication

use, inactivity, and verbal and nonverbal pain

communications (such as guarding and bracing). Medication

reduction is often achieved by administering "pain

cocktails" or by reinforcing patients in their own attempts

at reducing medication utilization. Low activity levels are

usually assessed by self-report measures of "up-time" (time

that the patient is not in a reclined position). Some

studies report more objective measures of assessment such as

video taping while the patient engages in various activities

(e.g., sitting, standing, and walking). Treatment

interventions for activity levels most often include

reinforcement for up-time. Patients are encouraged to

gradually increase their time out of bed, walking, and doing

various activities. Patients' perceived ability to perform

increased activity levels are also encouraged and reinforced

through their participation in general physical therapy

modalities. Treatment for verbal and nonverbal pain

behavior is usually accomplished by ignoring inappropriate

behaviors (body postures or verbal complaints) and

reinforcing appropriate behaviors. Hence, the physical

component to these treatment programs relies heavily on the

psychological concept of operant conditioning.

Psychological interventions usually include cognitive-

behavioral techniques. For example, Fordyce (1976) proposes

that pain behaviors are maintained by contingent events and

thus are amenable to behavioral interventions such as

undesirable consequences for certain pain behaviors.

Similarly, within this behavioral framework, increases in

activity levels are encouraged and rewarded. Since social

factors have been found to have a significant impact on the

chronic pain experience, many psychological interventions

include family therapy to address the role of positive and

negative social reinforcement from significant others.

Patients are often instructed in relaxation and biofeedback

techniques, and various coping strategies are taught (such

as pacing daily activities). In terms of more cognitive

interventions, faulty cognitions are challenged. This

teaches the patient to identify and relabel negative and

defeating self-statements, which leads to more adaptive

functioning. Psychological interventions are also

accomplished with individual psychotherapy or group therapy

sessions. These psychological interventions focus on faulty

cognitions, affective reactions, and ineffective coping

strategies, while at the same time they encouraged the

patient to increase his/her physical activities. Hence, the

psychological component of these programs is heavily laden

with a physical component--increased activity (Block, 1982).


Many treatment programs emphasize a multimodal approach

to the treatment of chronic pain. These treatment protocols

include psychological, physical, and vocational

interventions. Such multidisciplinary interventions appear

to produce significant improvement in chronic pain patients.

In a review of multimodal treatment studies which included

follow-up, Block (1982) found that studies including a 10-

month follow-up reported an average of 58% of the subjects

had reduced their intake of analgesic medications. Other

improvements include 75% either employed or in employment

training programs, 70% with significantly increased activity

levels, and 74% received no additional treatment. Block

(1982) also reports that in studies with 3-year follow-up,

an average of 60% of the subjects had reduced their

analgesic medications, and 100% maintained increased

activity levels.

Although Block's (1982) review suggests that these

treatment programs are successful, the studies reviewed

provide only limited information in that they were quasi-

experimental. Additionally, patients participating in

multidisciplinary pain treatment centers are a select sample

and only represent a small percentage of individuals

suffering with chronic low back pain. Participation in

these multidisciplinary programs usually requires

authorization and payment by insurance companies (mostly

workmen's compensation) and is often contingent on the

patient receiving a final disability determination at the

end of treatment. Differences in treatment responsivity and

sample characteristics in different treatment settings were

documented in a study by Deyo, Bass, Walsh, Schoenfeld and

Ramamurthy (1988). These investigators recruited subjects

for a clinical treatment of low back pain by advertising for

an outpatient treatment program (clinical group), and they

recruited another sample of subjects with low back pain in a

multidisciplinary pain clinic. Although the two groups of

subjects did not differ on duration or intensity of pain,

there were significant differences found between the groups

pre- and post-treatment. The clinical outpatient group was

more likely to be working, using no medication, having

higher incomes and education, and not receiving workmen's

compensation pre-treatment. Additionally, post-treatment

findings indicated that the clinical group significantly

benefited more from treatment compared to the

multidisciplinary treatment group. Hence, the authors

suggested that patients participating in multidisciplinary

treatment programs tend to represent a small, specific

sample who experience not only chronic pain, but have lower

incomes and education and are more reliant on insurance and

work related compensations which impacts on treatment

outcome (Deyo et al., 1988).

Another problem in this area of research is that few

experimental studies have been conducted which examine the

specific physical or psychological effects of a

comprehensive treatment program. Philips (1987) reported a

cognitive-behavioral treatment program that utilized random

assignment to a wait-list control or treatment group. He

found that 83% of the subjects in the treatment group

improved relative to the control group. Treatment consisted

of teaching the patients techniques to manage and control

their pain and techniques to increase their exercise levels

and physical fitness. Hence, improvement was described as

increased physical fitness levels and decreased

psychological distress (Philips, 1987).

In a study by Heinrich, Cohen, Naliboff, Collins and

Bonebakker (1985), chronic pain patients were assigned to a

physical therapy treatment or to a behavior therapy program.

Both treatments addressed increasing activity levels. They

found that both groups improved after treatment, and

interestingly, there were no significant differences between

the two groups at 6- and 12-month follow-ups.

Studies addressing physical reconditioning and

cognitive behavioral treatments in multimodal treatment

programs report significant improvements after treatment and

at long-term follow-up. Improvements typically are defined

as return to work or in training programs, decreased

psychological dysfunction, and increased physical

functioning. Subjects participating in these studies are

considered to be refractory to previous medical treatments

and are not candidates for surgical interventions.

Treatment success is high in these studies, but it should be

noted that insurance companies pay for these programs and

they expect their clients to return to work or receive final

disability at completion. Hence, it is not surprising that

subjective measures of experienced pain are not considered

predictive of a successful outcome, but instead, return to

work is the measure of successful outcome. Additionally,

there is a selection bias inherent in these studies due to

the fact that some insurance companies refuse to pay and

some subjects refuse to participate due to the knowledge of

a potential risk to their economic situations (Gatchel et

al., 1986; Mayer, Gatchel, Mayer, Kishino, Keeley, & Mooney,

1987; Mayer et al., 1986; Mayer, Smith, Keeley, & Mooney,

1985; Mayer, Gatchel, Kishino, Keeley, Capra, Mayer,

Barnett, & Mooney, 1985).

Given that chronic pain patients suffer differentially

with physiological symptoms, psychological distress, and

environmental disruption, it is important to understand the

components of multimodal treatment programs that are

beneficial. Since treatment success is often defined as

increased functional activity (Gottlieb et al., 1977), it is

reasonable to assume that the effective ingredient might be

the physical reconditioning component. On the other hand,

improved psychological well-being may lead to improved

functioning and increased activity (Weisenberg, 1987).

Therefore, in order to better understand the effects of

these treatment programs, there is a need for studies to be

conducted within an experimental design (Tan, 1982).

Recent studies contend that the important element in

the treatment of chronic low back pain patients is increased

functional strength. Quantitative strength changes and

physical improvement have been postulated to be important

for surgical decisions, designing rehabilitation techniques,

and for disability determinations (Mayer et al., 1985).

Furthermore, treatment programs that offer passive treatment

modalities (i.e., spinal manipulation, electrical

stimulation, and biofeedback) potentially increase a

patient's dependence on the health care system and

potentiate their manifestation of the sick role. Given the

complexity of these multidisciplinary treatment programs,

evaluation of successful treatment is difficult in that it

incorporates medical, legal, psychological, and

socioeconomic problems. One of the most current assertions

is that the major deficit in chronic low back pain is a

physical deficit from muscular disuse caused by prolonged

and excessive protection of the spine (Mayer et al., 1987).

Hence, there is a strong need for future research to address

the relative importance of the various therapeutic elements

(i.e., physical and psychological therapies) offered in the

multidisciplinary treatment programs (Hazard, Fenwick,

Kalish, Redmond, Reeves, Reid, & Frymoyer, 1989; Mayer et

al., 1985). In summary, the impact of improved physical

conditioning on functional abilities as well as

psychological status has not been sufficiently examined in

chronic low back pain patients.

Self-Efficacy and Decreased Avoidance of Physical Activities

Dolce (1987) postulated that the chronic pain patient's

desire to avoid pain results in more severe limitations in

activity than do actual physical limitations. This

avoidance of physical activity is similar to the behaviors

seen in phobic patients. A chronic pain patient fears

certain activities because an earlier activity was

experienced as aversive which created pain, tension and

anxiety. Hence, the patient continues to avoid activities

perceived as similar to the initial anxiety/pain-inducing

situation. The avoidance behavior is then reinforced by the

fact that the patient avoids the experience of increased

pain (Philips, 1987).

Self-efficacy theory as proposed by Bandura (1982) is

useful in understanding the cognitive mediators involved in

the chronic pain patient's avoidance of physical activity.

Self-efficacy is defined as an individual's belief in

his/her ability to perform a particular behavior. This

belief, in turn, can influence a certain outcome. For

example, Bandura (1982) suggests that thoughts about the

self mediate between knowledge of how to perform a behavior,

motivation to perform a behavior, and the actual performance

of a behavior. If the patient believes that he/she is

incapable of completing a certain behavior/activity, the

activity will be avoided. Self-efficacy beliefs will

determine not only the performance of an activity, but also

the amount of effort an individual will expend and persist

in an activity. In summary, the individual's belief that

he/she can master a certain behavior will predict actual

performance of a given behavior. If self-efficacy

expectations are negative, performance will be impaired in

spite of functional abilities (Dolce, 1987).

Bandura has shown that the phobic patient's perception

(self-efficacy expectancies) of his/her ability to cope with

and master in-vivo exposure to feared stimuli are predictive

of the patient's actual behavioral performance. He found

that with gradual exposure to a feared stimulus, the phobic

patient's prior self-efficacy expectations predicted his/her

performance of previously avoided behaviors, and that as

he/she performed these avoided behaviors, self-efficacy

expectancies increased.

Self-efficacy expectations have not only been shown to

predict behavioral change in phobia patients, but they have

also been predictive of behavioral change in other patient

populations. For example, Ewart, Taylor, Resse, and DeBusk

(1983) studied 40 males referred for treadmill exercise

following a myocardial infarction. Self-efficacy was

measured by a self-report questionnaire that asked the

patients to rate their self-perceived ability to perform

various activities (e.g., walking, running, and climbing

stairs). Self-efficacy was found to increase after

successful completion of treadmill exercise and was

predictive of the subject's participation in subsequent

activities such as walking, running, and climbing stairs.

Hence, enhanced self-efficacy expectations were

significantly correlated with the subject's actual intensity

and duration of subsequent physical activities. Along

similar lines, Manning and Wright (1983) found that self-

efficacy expectancies regarding an individual's perceived

ability to run a 10 Km race were more predictive of race

performance than the subject's actual running history.

Hence, the chronic pain patient's avoidance of physical

activity may not be in response to the experience of painful

sensations, but instead may be an avoidant coping mechanism

which is employed to reduce fear and anxiety associated with


Only a few studies have looked at the role of self-

efficacy expectations and the perception of pain. The role

of self-efficacy theory is important in that patients avoid

activities because pain signals a problem. Hence, the

patient develops avoidance behaviors independent of pain.

The impact of the patient's self-efficacy then is

determinant of their participation in treatment

interventions. For example, Kleinke & Spangle (1988) found

that patients exhibiting high levels of pain behavior and

reporting high levels of pain preferred to have an inactive

role in treatment (e.g., passive therapies such as ice or

heat packs) whereas patients with less pain behavior and

lower self-reported pain chose to take an active role in

their rehabilitation in the form of active physical


Manning and Wright (1983) evaluated self-efficacy

expectancies (e.g., perceived ability to undergo childbirth

without the use of medication) in 52 pregnant women. Self-

efficacy expectations were found to predict persistence in

self-controlled pain during labor in that higher self-

efficacy ratings were correlated with decreased medication

usage. Of note, this finding was independent of the length

of time the subject experienced labor pain. These authors

concluded that self-efficacy expectations were a better

predictor of outcome (i.e., decreased pain complaints) than

were previously exhibited behaviors (e.g., behaviors during

child birth classes).

Holroyd, Penzien, Hursey et al. (1984) studied the role

of self-efficacy expectations in the control of tension

headaches. These authors recruited 43 recurrent tension

headache subjects from a college population for an

electromyography (EMG) biofeedback treatment study. This

study utilized a 2 x 2 factorial design in which subjects

were randomly assigned to one of four conditions. Treatment

conditions consisted of increased versus decreased EMG

activity feedback, and high versus moderate successful

control feedback. Findings of this study were that changes

in post-treatment headache activity were induced by

performance feedback and were not related to actual EMG

activity. In summary, whether the subjects were told that

they successfully increased or decreased EMG activity was

not significant in their subsequent reduction of headache

pain. Instead, the high success feedback significantly

decreased headache pain post-treatment regardless of the

direction of EMG activity. Hence, these authors proposed

that high success feedback increased the subject's self-

efficacy expectations which subsequently reduced their

headache pain.

Self-efficacy expectancies were evaluated in a study of

chronic pain patients involved in a nine week outpatient

treatment program (Philips, 1987). This treatment program

was done in a small group format and focused on teaching the

patients new coping strategies and on increasing their

activity levels. Self-efficacy ratings were taken for 40

consecutive subjects who presented for treatment of chronic

pain. Twenty five subjects were assigned to treatment

groups and 15 subjects were assigned to a wait-list control

group. Subjects completed questionnaires on their number of

avoidance behaviors and on their perception of the severity

of their pain problem. There were no differences between


the groups pre-treatment in avoidance behaviors or negative

perceptions, but at post-treatment there was a significant

difference. The treatment group reported significantly

higher levels of self-control over their pain (increased

self-efficacy) and perceived their pain as less of a problem

when compared to the wait-list control group. Hence, in

this study, learning coping strategies in therapy increased

their self-efficacy expectancies, enhanced their ability to

function, and decreased their negative perceptions of their

pain problem (Philips, 1987).

Dolce and colleagues (1986) reported several studies

addressing the role of setting quotas of performance,

feelings of mastery at different levels, and subsequent

self-efficacy ratings for future performance. They reported

a laboratory study in which 64 college undergraduates

participated in an experimental pain procedure (cold pressor

test). These subjects were asked to rate their self-

efficacy expectations on their ability to tolerate

increasing levels of pain. These investigators found that

setting systematic quotas for pain tolerance increased the

subject's ability to tolerate pain. In addition, there was

a significant correlation between self-efficacy expectancies

and actual pain tolerance. Of note, a group that took a

placebo medication supposedly to improve pain tolerance

actually decreased in pain tolerance. These authors

concluded that setting quotas was more beneficial than

rewards and reinforcement and that self-efficacy

expectations were significantly related to actual abilities

to tolerate pain (Dolce, Doleys, Raczynski, Lossie, Poole, &

Smith, 1986).

Another set of laboratory studies utilizing the cold

pressor test for pain stimuli reached similar conclusions.

Self-efficacy expectations in these studies were addressed

in conjunction with the role of cognitive mediation and

endogenous opioid involvement following pain stimulation.

Evidence of endogenous opioid mechanisms was found in

response to pain stimuli in addition to nonopioid cognitive

mechanisms. In these studies, a placebo medication group

increased their pain tolerance due to endogenous opioid

involvement, but did not increase their self-efficacy

ratings. In contrast, the cognitive strategy group

increased both their self-efficacy and pain tolerance.

Conclusions for the chronic pain patient based on these

laboratory findings were that strong self-efficacy may lead

to increased activities despite pain which potentially

exacerbates the pain and results in a cognitive loss of

control over the painful experience. Stress associated with

a sense of failing control with increasing pain then

activates endogenous opioid mechanisms for additional pain

control. Hence, it was noted that cognitive appraisals did

not effect the experience of pain, but instead increased the

subject's ability to endure pain which resulted in increased

emotional responses and subsequent opioid activation. It

was further concluded that self-efficacy is a causal

determinant of performance and is the moderating factor on

the subject's pain tolerance (Bandura et al., 1987; Litt,

1988; Taylor, 1989).

This line of research has also examined the role of

setting exercise quotas and subsequent self-efficacy ratings

in the clinical treatment of chronic pain patients. Two

articles were reviewed in which subjects were referred to a

pain-management clinic and assessment measures were taken

with respect to their actual ability to accomplish set

exercise quotas and their self-efficacy expectations. Self-

efficacy ratings were found to be predictive of actual

participation in physical exercise and in maintaining

physical activities post-treatment (Dolce, Crocker,

Moletteire & Doleys, 1986; Dolce, Crocker, & Doleys, 1986).

These studies suggest that chronic pain patients need

to be desensitized to their fear of activities, need to

develop a sense of mastery over their pain by actual

activity performance, need to develop the belief that the

demands of an aversive situation (exercise) are not greater

than their coping skills, and need to have an internal

attribution for treatment success. In addition, these

investigators found that the patient's past learning history

of successes and failures were often a primary determinant

of their self-efficacy ratings. In summary, self-efficacy

expectations are a useful predictor of pain coping behaviors

and of post-treatment maintenance of therapeutic gains

(Dolce et al., 1986a; Dolce et al., 1986b).

In conclusion, perceived control over a situation is

important when addressing the chronic pain patient's

motivation and ability to engage in exercise treatment

programs. Exercise is emphasized in the treatment of

chronic pain and is described as "well behavior". Efficacy

expectations are important considerations and effect how

much effort an individual will exert and how long they will

persist in the face of aversive experiences such as exercise

regimens or quotas that may initially exacerbate their

fears, anxieties, and pain.

Physical Reconditioning as a Way of Improving
Muscular Strength and Psychological Symptoms

Sedentary life styles and the lack of physical fitness

have been proposed as causal agents in maintaining low back

pain (Rosomoff, 1985). Not only do chronic low back pain

patients suffer with the nociception of their original pain,

but their pain experience is exacerbated by muscle atrophy

resulting from subsequent inactivity. Similarly,

psychological distress may not only be a consequence of low

back pain, but also has a role in the exacerbation of

existing pain levels (Flor & Turk, 1989: Gottlieb et al.,

1977). According to the Gate Control Theory of pain, stress

can lead to increased autonomic arousal and increased muscle

activity. This in turn leads to peripheral stimulation

(pain) and emotional reactions (such as stress and anxiety).

Hence, psychological distress may precipitate low back pain

by creating muscle tension in the weaker spinal muscles. If

these weaker, but tense muscles are subjected to a work-

load, the individual may then experience pain. Therefore, a

pain-stress cycle can begin with pain, or stress itself can

be a precipitating factor (Keefe & Gil, 1986; Linton, 1987).

Although the experimental findings on the relationship

between pain and stress with respect to chronic pain is

inconclusive, there is some evidence that individuals

manifest specific physiological responses to perceived

stress (Flor & Turk, 1989).

The pain-stress model suggests that psychological

distress can exacerbate pain in weaker muscle groups, and

that weak and painful muscles can exacerbate psychological

distress. Bortz (1984) suggested that "use" is a biological

principle inherent in all species and that inactivity

results in the muscles stiffening and decreasing in fiber

diameter which results in muscular atrophy. Hence, chronic

pain may be viewed as a "disuse syndrome". If muscular

atrophy and psychological distress contribute to the

experience of low back pain, treatment of the chronic low

back patient should address both areas. The studies

reviewed above share a focus on improving the chronic pain

patient's psychological and functional status by increasing

their activity levels. Hence, physical reconditioning with

exercise is a viable treatment modality with a chronic low

back pain population.

General Effects of Exercise on Psychological Functioning

In terms of psychological benefits, previous studies

suggest that physical exercise can decrease psychological

stress, anxiety, and depression in certain populations

(Dishman, 1985; Folkins & Sime, 1981; Keller & Seraganian,

1984; Levine, 1971; Reiter, 1981; Weisenberg, 1987). In

addition, physical fitness is proposed to enhance an

individual's ability to effectively manage emotional stress

and to improve adaptive interactions with the environment.

Improved cardiovascular functioning with exercise and

reduced resting muscle action potentials following exercise

are suggested to be associated with decreased tension and

psychological distress. Exercise is also purported to

provide an individual with a sense of mastery and control

which is then associated with feelings of psychological

well-being (Folkins & Sime, 1981).

However, there is inconsistency in the literature

regarding the specific psychological benefits of exercise,

and as Hughes (1984) suggests, many studies are quasi-

experimental and flawed with experimental biases. His

review of 12 studies which employed acceptable experimental

methodology, suggests inconclusive findings regarding the

effects of exercise on depression and anxiety. In keeping

with Hughes' (1984) review of a lack of consistent positive

findings for the psychological benefits of exercise, Hughes,

Casal, and Leon (1986) studied sedentary men assigned to an

exercise or a control condition. These authors found no

psychological benefits (such as reduced anger, depression,

or mood disturbance) derived from the exercise condition

when compared to controls. Another study confirmed these

findings with adult women. These women participated in an

exercise program and were compared to a no-exercise control

group. Again, no psychological benefits were found

(Coleman, Price & Washington, 1985).

In contrast, aerobic and anaerobic exercise have been

reported as beneficial in elevating mood states, improving

self-concept, and decreasing anxiety (Folkins & Sime, 1981;

Reiter, 1981; Sime, 1984). For example, Doyne Ossip-Klein,

Bowman et al. (1987) studied 40 depressed females and found

that both aerobic and anaerobic exercise significantly

decreased depression compared to a wait-list control group.

There were no differences found between the two types of

exercise, and treatment gains were maintained for both

groups at a one year follow-up.

These inconsistent findings suggest that the initial

level of psychological distress (anxiety, depression, or

self-esteem) is an important factor when exercise produces

psychological benefits. For example, low self-esteem

appears to improve following exercise, but the studies

looking at self-esteem only find changes if initial self-

esteem is very low (Hughes, 1984). Similarly, studies

looking at depression and anxiety find that mild to moderate

distress (as opposed to severe distress) can be modified by

exercise (Sinyor, Schwartz, Peronnet et al., 1983).

Dishman, Sallis, and Orenstein (1985) in their review

of the exercise literature found that as the severity of

psychological distress increased, there was a related

increase in withdrawal from exercise programs. These

authors conclude that continued participation in exercise

programs of all types was related to the individual's past

exercise history, perceived health, education, self-

motivation, and positive support from a spouse or

significant other. These conclusions are tentative at best

in that the literature they reviewed included diverse

populations and settings, various research traditions and

disciplines, and a variety of differing interpretations.

Although professionals and laymen alike, promote the

concept of "healthy body, healthy mind" (Sachs, 1982), most

of the research addressing this area is by anecdotal report,

or quasi-experimental designs. There are only a few studies

experimentally done which have investigated the

psychological benefits of exercise, and these have resulted

in conflicting results. These inconsistencies are

attributable to the lack of experimental designs, and to the

fact that many of the subjects had low levels of

psychological distress prior to treatment. In contrast, the

chronic low back pain patient is a unique population which

appears to experience significant psychological turmoil

(Gottlieb et al., 1977). Hence, the role of exercise may

have a more pronounced effect in this population. This

suggestion is consistent with the finding that treatment

programs that focus on increasing activity levels (well-

behavior) and decreasing sick behaviors (inactivity) have

been shown to be an effective treatment for chronic low back

pain patients (Block, 1982).

In summary, if exercise and physical reconditioning

does enhance an individual's ability to cope with stress,

and decreases anxiety and depressive symptomatology,

increasing the chronic low back pain patient's physical

activity through a structured exercise program should result

in his/her experiencing an improved psychological well-being

and decreased pain.

Exercise in Chronic Low Back Pain Patients

Studies addressing exercise in chronic low back pain

have found that exercise is beneficial in reducing pain and

increasing physical functioning (Bigos & Battie, 1987).

Exercise is reported as a major focus of treatment in 100%

of pain treatment clinics and in 86.8% of all patients with

reports of pain (Tollison, Kriegel, & Satterthwait, 1989),

but the beneficial role of exercise is not understood in the

treatment of chronic low back pain. For example, Jackson

and Brown (1983) state that the reasons for recommending

exercise include reducing pain, increasing strength,

decreasing mechanical stress on the spine, improving overall

physical fitness and preventing future injury, stabilizing

hypermobile spinal segments, improving posture, improving

mobility, and finally, recommending something "when all else


Although the role of exercise in chronic low back pain

is poorly understood, patients do improve following exercise

programs. For example, Manniche, Hesselsoe, Bentzen,

Christensen, and Lundberg (1988) reported significant

improvements in chronic back pain patients following two to

three months of exercise. Additionally, they found that

intensive exercise (multiple strengthening exercises for 1

1/2 hours for 30 sessions) was more effective than moderate

exercise (1/5 the intensity and time). In a large study of

the benefits of exercise conducted through the YMCA on

11,809 people, Kraus and Nagler (1983) reported on a

subsample of 546 post-surgical back pain patients that had

significant decreases in pain and increases in strength

following participation in a routine exercise program for

six weeks. Reilly, Lovejoy, Williams, and Roth (1989)

studied 40 males and females with chronic low back pain (10

each group) who were randomly assigned to a six month

supervised or home exercise program. They found that the

supervised exercise subjects were not only healthier, but

they had decreased their pain reports, had decreased their

medical visits, and had fewer relapses. Additionally, the

significant factors found to relate to decreased pain were

the number of exercise sessions completed, decreased body

fat, and increased aerobic fitness.

In summary, the studies reviewed on the relationship

between exercise and psychological functioning are

inconclusive. Despite these discrepant findings, there does

appear to be some psychological benefits associated with

exercise. The role of exercise in treating chronic low back

pain also appears to be a beneficial treatment modality, but

the role of exercise in the chronic pain syndrome needs

further investigation. Given the suggestion that trunk

extension strength is the most severely affected area in low

back pain patients, exercise and rehabilitation of this area

is a viable treatment modality open for investigation

(Smidt, Herring, Amundsen, Rogers, Russell, & Lehmann,


Exercise of the Low Back

Recent investigators who have suggested that physical

reconditioning can be beneficial for the low back pain

patient emphasize that exercise regimens should recondition

the specific atrophied muscles (Feuerstein et al., 1987;

Mayer et al., 1985). Chronic disuse of specific spinal

muscles can exacerbate existing low back pain (Rosomoff,

1985). In addition to preventing exacerbation of the pain

experience, strengthening the muscles of the low back and

lower extremities may also lead to improvement in other

areas of functioning. For example, Fredrickson, Trief,

VanBeveren, Yuan and Baum (1987) followed 80 chronic pain

patients referred to a multimodal six week outpatient

treatment program. They found that increased functional

strength after treatment was the best predictor of

successful outcome which was defined as decreased pain

experienced, increased activity levels, and/or return to


Increasing the strength and elasticity of the lumbar

extension muscles should decrease chronic low back pain.

This is consistent with Keefe and Gil's (1986) pain-spasm-

pain model of the chronic pain patient. These authors state

that an initial painful event leads to a reflexive muscle

spasm, vasoconstriction, and the release of pain producing

substances. With time, this minimizes the pain and

additional spasm through reduced movement. This limited

movement, in turn, leads to muscle shortening and the

inactive muscles atrophy. The shortened, atrophied muscles

then predisposes the patient to more spasm and increased

pain. Therefore, increasing the physical fitness of the

lumbar muscles should decrease pain in patients with chronic

low back pain.

Jones, Pollock, Graves et al. (1988) propose three

primary reasons for myofacial low back pain: specific

muscular responses, type of muscle fiber, and chronic disuse

atrophy. They describe two muscle fiber types in the low

back--fast and slow twitch fibers. The fiber types

determine not only the patient's strength, but their ability

to endure work-loads for varying lengths of time. Fast

twitch muscle fibers are suggested to fatigue much more

readily than slow twitch fibers which may predispose the

individual to reduced muscular strength and subsequently

lead to injury and pain. These investigators report that

based on endurance measures taken across a large healthy

sample, 30% of the general population has predominantly fast

twitch lower back muscles, 10% has predominantly slow

twitch, and 60% has a fairly even mixture of the two types

of fibers. Since a majority of the population has a

combination of fiber types, these authors suggest that

understanding of an individual's predominant fiber type

should be determined by computing a ratio of their lower

back muscles' strength to their endurance under a work load.

In summary, results of unpublished studies testing a

diverse population of subjects on the MedXTM lumbar

extension machine suggest that 30% of the general population

has a high percentage of fast twitch fibers and 60% has a

mixture of fast and slow twitch fibers. Based on these

studies, it is proposed that the lumbar extensor muscles are

proportionately weaker in the extended position and stronger

in the flexed. These authors further conclude that

individuals with predominantly high twitch muscle fibers


will have a low work endurance ability and are therefore at

high risk for injury to the low back (Jones et al., 1988).

Additionally, these investigators suggest that current

modes of exercising these muscle groups are ineffective in

that the hip extensors (legs and buttocks muscles)

predominate and interfere with the strengthening of the

lumbar extensor muscles. Exercise on the MedXT lumbar

extension machine is proposed to be a superior exercise

modality in that the pelvis and large thigh and gluteal

muscles are restrained and stabilized which isolates the

lumbar extensors for maximum exercise benefits. Studies

utilizing the MedXT lumbar extension machine have found

that normal subjects trained on the MedXTM lumbar extension

machine have significant increases in low back strength,

despite their previous exercise experience (Pollock,

Leggett, Graves et al., 1989). Hence, Jones et al. (1988)

propose that these muscle groups must be isolated and

exercised in order to increase their strength and thereby

reduce the risk of low back injuries. They conclude further

that routine exercise on the MedXT lumbar extension machine

is the only available method which isolates and strengthens

the lumbar extensor muscles.

There is currently only one study addressing the

benefits of MedXT training for individuals with low back

pain. This study examined a population of 12 mild low back

pain subjects who exercised on the MedXT lumbar extension

machine for 10 weeks. Training produced significant

increases in lumbar extension strength, as well as decreases

in self-reported pain, and increases in functional status as

measured by the Oswestry Low Back Pain Disability

Questionnaire (MacMillan et al., 1988, unpublished data).

While this study suggests that lumbar extensor

strengthening may be beneficial in treating patient's with

chronic, mild, low back pain, there are several

methodological weaknesses. It is unclear how the term

"mild" low back pain was operationalized. It is also

unclear if mild represents limited pathophysiological

findings, limited psychological distress, or both. In

addition, a low back pain control group was not included.

Previous research suggests that the MedXT lumbar

extension machine effectively strengthens lumbar extensor

muscles and that this strengthening is superior to other

methods of lumbar exercise for healthy individuals (Graves

et al., 1990; Pollock et al., 1989). In addition, pilot

data (MacMillan et al., 1988) suggests that exercise on the

MedXT lumbar extension machine may be a beneficial

treatment modality for patients with mild, low back pain.

Since low back pain patients are a heterogeneous population

comprised of differing levels of physical, psychological and

social involvement with their pain, more research is needed

to investigate the benefits of the MedXT lumbar extension

machine for these patients.


The literature reviewed suggests that physical

reconditioning is important in treating patients with

chronic low back pain, particularly in order to recondition

specific atrophied muscles. The chronic back pain patient

avoids activity in an attempt to decrease pain which in

turn, may result in increased pain and psychological

distress. Feuerstein et al. (1987) suggested that

strengthening the muscles of the low back and lower

extremities is an important correlate of improvement after

treatment in the chronic low back pain patient, but this

area of investigation has been neglected. While many

treatment programs for the chronic pain patient include

physical and psychological components, the specific effects

of physical reconditioning on functional and psychological

outcomes are poorly understood.

Jones et al. (1988) propose that the MedXT lumbar

extension machine has the ability to isolate and strengthen

the weaker extensor muscles of the low back in healthy

subjects and subjects with mild low back pain. The meaning

of the previous research with mild low back pain patients is

unclear in that there is no differentiation of mild with

respect to mild physical or mild psychological involvement.

In summary, there has been no research to date investigating

this machine as a treatment modality for a diverse chronic

low back pain population.

The current study examined the specific benefits of

exercise performed on the MedXTM lumbar extension machine

with a diverse chronic low back pain population. It was

hypothesized that subjects completing MedXM training would

show significant improvement in low back muscle strength as

compared to the control group. It was further hypothesized

that this strengthening program would lead to significant

improvement in post-treatment physical and psychosocial

dysfunction as assessed by subscales of the Sickness Impact

Profile. Physical reconditioning on the MedXT lumbar

extension machine was also hypothesized to result in

decreased pain reports and psychological distress

(specifically anxiety and depression). These hypotheses

were based on previous research which suggested that

exercise can decrease anxiety and depression. Reductions in

self-reported pain were believed to be important given that

it is verbal complaints that result in patient referrals for

continued treatment (Kleinke & Spangler, 1988). Finally, it

was hypothesized that adherence to the training program on

the MedXT lumbar extension machine and maintenance of

therapeutic gains would be predicted by the subject's self-

efficacy expectations, and social support (measured by the

Exercise Locus of Control Scale and the Social Support

subscale of the West Haven-Yale Multidimensional Pain

Inventory, respectively). This hypothesis is consistent

with the finding that self-efficacy expectations influence


an individual's performance in a given activity, and with

the finding that social reinforcement effects an

individual's pain coping abilities and adherence to exercise

programs. Given the hypothesis that patients avoid exercise

based on their negative past experiences, self-efficacy

beliefs were believed to contribute to the individual's

motivation to continue in a rigorous exercise program and

hence predict attrition (or treatment drop-outs).



Fifty-five subjects (34 males and 21 females) were

referred for rehabilitation on the MedXTM lumbar extension

machine by an orthopaedic surgeon specializing in disorders

of the spine. The average age of the subjects was 45 years

(ranging from 22 to 70 years). The sample was predominantly

caucasian (91%) and married (76%). Subjects experienced low

back pain for an average of eight years (ranging from 1 to

26 years), had two surgeries or less, were ambulatory, and

were not dependent (daily use) on narcotic analgesics.

Fifty-four percent of the subjects were receiving workmen's

compensation or disability payments as their primary source

of income and the average time out of work due to pain was

37 months (ranging from 0 to 168 months). Thirty-five

percent of the subjects were employed full-time and 46% were

unemployed due to back pain. The onset of pain was

described as sudden and was related to an automobile or work

accident for 83% of the sample.

Equipment and Measures

MedXT Assessment and Training

MedXT testing consisted of an isometric test of lumbar

extension strength at seven different testing points within

each subjects range of motion up to a maximum arch of 72

degrees. Subjects were seated in the MedXTM lumbar

extension machine and their knees were positioned so that

their femurs were parallel to the seat. Subjects were

secured in place by specially designed femur and thigh

restraints used to stabilize the pelvis. To begin the test,

subjects were locked into between 48 and 72 degrees of

flexion and instructed to extend their back against the

upper back pad. Once maximal tension had been achieved, the

subjects were instructed to maintain a maximal contraction

for one to two seconds before relaxing. A 10 second rest

interval was provided between each isometric contraction

while the next angle of measurement was set. During the

contractions, the subjects were provided visual feedback of

their generated force and were verbally encouraged to give a

maximum effort. A maximum isometric contraction was

measured at each of the seven angles and a computerized

force curve was obtained. The subject then exercised

through flexion and extension with a work load of 1/2 their

peak isometric strength until fatigued. At this point, a

second force curve was generated. The shape of the two

force curves were noted for similarity and the difference

between them was a measure of fatigue. Hence, the first

MedXT assessment provided a baseline isometric strength


For the next two consecutive clinic visits, subjects

were instructed in the proper use of the MedXTM lumbar

extension machine under the supervision of one of three

registered physical therapists. MedXTM instruction and

training consisted of isotonic exercise at a work load of

1/2 the subject's peak isometric strength (torque in N.m).

When the subject exceeded 12 repetitions, the resistance was

increased 5 foot pounds. When subjects returned for their

fourth clinic visit, they were again administered the

isometric strength test which generated a second force

curve. In order to control for familiarity and variable

learning effects, this testing session was used as the pre-

treatment strength measure. Subsequent training sessions

were performed twice per week for four weeks followed by

once per week for an additional six weeks. Strength curves

were then re-generated for assessment of changes in muscle

strength post-exercise.

Self-Report Questionnaires

All subjects completed a demographic questionnaire

which consisted of questions to assess race, sex, age,

marital status, occupation, work status, and litigation

status. Questions pertaining to the subject's pain history

included whether there was a precipitating event to their

pain problem, the duration of their pain problem, previous

treatment interventions (i.e., prior hospitalizations,

surgeries, physical therapy modalities, and other

medical/psychological interventions), and current

medications. At the end of the study, subjects were

administered a second questionnaire to assess their

utilization of medical interventions during the study (See

Appendix A and B.).

Measures of self-efficacy

Exercise Objectives Locus of Control Scale. This

measure is a standardized instrument which was used to

assess the subject's perception of control over and ability

to master the MedXT exercise regimen. The Exercise

Objectives Locus of Control Scale (EOLCS) was developed by

McCready and Long in 1985. The test has 18-items (e.g., My

own actions will determine whether or not I achieve my

exercise objectives) which are responded to on a scale of 1

to 5 (l=strongly agree, and 5=strongly disagree). Three

scores are obtained: an internal control factor, a powerful

other factor, and a chance control factor. These three

factors reflect the individual's perceived ability to master

an exercise objective and provided information as to their

perceived competence in completing the task. This test has

good psychometric properties and was normed on males and

females ranging from 17 to 55 years in age. Cronbach's

alphas were 0.86 for internal control, 0.79 powerful other,

and 0.57 for chance. Test-retest (three to four months)

were high for the powerful other and chance subscales (0.72

and 0.60, respectively), but were somewhat lower for the

internal scale (0.32). This test was originally developed

as a measure of predicting attrition, but it subsequently

was not fund to be solely predictive of attrition. It was

found to be predictive of adherence to an exercise program

based on the individual's self-efficacy expectations and if

the individual's participation was perceived as a stress and

tension reducing mechanisms, in contrast to a perceived

social event.

Activity Recall Questionnaire

This measure was originally developed by Blair,

Haskell, Ping Ho et al. in 1985 to assess changes in an

individual's habitual exercise patterns. It provides a

general measure of the subjects' routine activities (e.g.,

climbing stairs, walking, and household chores) in addition

to their participation over a one week period in

recreational activities. It has been used in a large

epidemiology study which examined exercise habits in over

2,000 subjects (Blair, Haskell, Ping Ho et al., 1985). This

activity measure was found to significantly correlate with

dietary energy intake in a study of 495 males and 545

females. It has also been shown to be positively and

significantly correlated with physiological changes in

maximum oxygen uptake and body fat after exercise (Sallis,

Haskell, Wood et al., 1985). Since it was suggested that

past exercise experience may predict outcome, this measure

was used to predict compliance and completion of the

treatment program.

West Haven-Yale Multidimensional Pain Inventory

Subjects completed the West Haven-Yale Multidimensional

Pain Inventory (WHYMPI; Kerns, Turk, & Rudy, 1985). This is

a 52-item self-report questionnaire that is consistent with

a cognitive-behavioral conceptualization of the chronic pain

syndrome. Studies using this brief measure report that it

is psychometrically sound and when used in treatment outcome

studies it has been found to be a reliable and valid

assessment instrument for chronic pain patients (internal

consistency ranging from 0.70 to 0.90, and stability

coefficients ranging from 0.62 to 0.91).

There are three parts of this inventory that examine the

impact of pain on the subject's lives by addressing their

perception of positive and negative social support, their

activity levels, and their current levels of experienced

pain and depression. The positive/negative social support

subscales were used to predict attrition and therapeutic

gains. A subject's perception of their pain intensity was

also measured with this questionnaire by having them rate

their current pain and their pain for the last week.

Mental Health Inventory

This measure was included to assess changes in overall

psychological well-being, and changes in anxiety and

depression. Factor analysis of the Mental Health

Inventory's (MHI) 38 questions indicates that there are two

higher order factors and five lower order factors (Veit &

Ware, 1983). The factor of psychological well-being is

defined by two lower order factors of General Positive

Affect and Emotional Ties. A second factor of psychological

distress is defined by three lower order factors of Anxiety,

Depression, and Loss of Behavioral/Emotional Control.

Additionally, a Life Satisfaction score is obtained. The

MHI is a good psychometric instrument with an internal

consistency reliability coefficient of 0.94 and a one week

test-retest reliability of 0.80 (Veit & Ware, 1983). The

two higher order factors were used in this study to assess

psychological well-being, and depression and anxiety.

Perceived Stress Scale

The Perceived Stress Scale (PSS, Cohen, Kamarck, &

Mermelstein, 1983) is a 14-item questionnaire used to

measure a general area of perceived stress. There are three

general subscales generated with this measure; physical

disability, psychosocial distress, and work/recreation

dysfunction. This measure was incorporated due to the

findings that chronic pain patients experience psychological

distress and to the belief that a particular event is not

stressful in itself, but it is the perception of the

individual that an event is stressful that results in a

concomitant stress response. This measure has moderate

test-retest reliability (0.85 after two days, and 0.55 after

six weeks) and validity coefficients are reported from 0.52

to 0.76. In addition, this measure is reported to be a good

predictor of health related outcomes as assessed with PSS

scores and subsequent seeking of medical services (Cohen,

Kamarck, & Mermelstein, 1983).

Sickness Impact Profile

The Sickness Impact Profile (SIP, Bergner, Bobbitt,

Pollard, Martin & Gilson, 1976) is a 136-item test which was

selected to assess changes in functional abilities. The

measure provides scores for the impact of an illness in

three general dimensions; physical disability, psychosocial

dysfunction, and other dysfunction (work and recreation).

For purposes of this study, the physical and psychosocial

dysfunction dimensions were utilized. Original validation

was performed on a diverse medical population and it was

found to be significantly correlated with the patient's

self-report of physical disability, physician ratings of

their disability, and with daily activity diaries (0.29 to

0.52, Pearson correlation coefficients).

More recently, this measure was used to assess

functional status changes in chronic low back pain patients

following treatment interventions (Follick, Smith, & Ahern,

1985). For this population, validity of this measure was

found in significant correlations between up/down time, and

emotional distress as measured by the MMPI (0.29, Pearson

correlation coefficient, and 0.52-0.64 canonical variate



Subjects were consecutively referred by an orthopaedic

surgeon for rehabilitation on the MedXTM lumbar extension

machine. All subjects received a complete medical

evaluation by the orthopaedic spinal specialist in order to

rule out the presence of organic pathology which would

require surgical intervention. Patient referrals were

informed that the exercise machine was involved in a

research protocol and were requested to participate. After

completing an informed consent, subjects completed the

demographic and self-report questionnaires. The

experimental protocol was approved by the institutional

review board of the University of Florida.

Subjects then returned for a baseline test followed by

two orientation and practice exercise sessions on the MedXTM

lumbar extension machine. These orientations provided the

subjects information on proper use of the equipment and

enabled the investigators to obtain reliable lumbar

extension force curves. Subjects then returned for a fourth

session and were tested for lumbar extension strength as was

previously described.

After the fourth session, subjects were randomly

assigned to a treatment condition which consisted of 14

exercise sessions or to a ten week wait-list control group.

Random assignment of the first 55 eligible patients resulted

in 32 subjects being assigned to the treatment group and 23

to a wait-list control group.

Before the fourth MedxTM session, subjects were re-

administered the EOLCS and the pain rating scale. This

enabled the investigators to evaluate the effects of

exposure to the training machine.

Experimental Group

Subjects began exercising on the MedXTM lumbar

extension machine twice per week for six weeks followed by

one day per week for an additional six weeks. MedXT

training consisted of variable resistant isotonic exercises

at a work load of 1/2 the subject's peak isometric strength.

At the end of 12 weeks, subjects were re-tested for lumbar

extension strength and the self-report questionnaires were

completed (treatment history, EOLCS, WHYMPI, MHI, PSS, SIP,

and activity questionnaire).

Wait-List Control Group

This group of subjects received the same complete

medical evaluation, MedXT orientation, and MedXT lumbar

strength testing as the experimental group. They also

completed all questionnaires prior to exposure to the MedXT

lumbar extension machine. They then repeated the EOLCS and

pain rating prior to their second lumbar strength test.

After lumbar strength testing, this group was asked to wait

10 weeks before beginning treatment. They were requested

not to change their current exercise and activities or their

methods of treating their pain. At the end of 10 weeks,

these subjects again completed all self-report

questionnaires and were retested on the MedXT lumbar

extension machine. They then followed the same treatment

protocol as the experimental group.

Statistical Analysis

The design of this study enabled the investigator to

measure changes on measures of psychological status,

functional abilities, and physical strength as a function of

treatment. Subjects were randomly assigned to a wait-list

control or treatment group. Pre- and post-treatment

differences on demographic variables and medical treatment

histories were addressed with multivariate analysis of

variance and Chi-square statistics.

Analysis of physical strength

Due to the chronicity and physical limitations of this

population, most subjects were unable to complete a full

range of motion within a standard arch of 72 degrees. In

order to standardize the data for group comparisons,

regression equations (Torque=angle(x) + intercept) for pre-

and post-treatment MedXT testing sessions were computed for

each subject based on their actual performance within their


varying range of motions. This enabled the investigator to

obtain the slope and intercept of the regression line

resulting from the generated force curve for each subject.

The pre-treatment slopes and intercepts were used as a

covariate in an analysis of variance for between group

differences post-treatment. Change scores between pre- and

post-treatment were used as the dependent variable.

Additionally, the estimated torque values based on the

regression analysis at standardized angles (angles of 0, 12,

24, 36, 48, 60, and 72 degrees) within the subject's range

of motion were compared post-treatment utilizing the pre-

treatment standard values as the covariate in an analysis of


Analysis of psychological measures

Psychological measures for between group comparisons

were analyzed using pre-treatment scores as the covariate in

a multivariate analysis of covariance for between group

differences. Change scores were used as the dependent

variable. Secondly, locus of control (internal and

external), activities, positive social support, and the pain

rating scale were entered into a step-wise multiple

regression analysis for prediction of outcome as assessed by

their relationship with the post-treatment intercept.

Finally, Pearson Correlation Coefficients were computed

between the variables of post-treatment strength (as

measured by the changes in intercept scores post-treatment),


self-reported pain, and psychological variables of distress,

dysfunction, and social support to address the magnitude of

the relationship between psychological measures and changes

in strength.


There were no significant differences between groups on

demographic variables or pain histories pre-treatment.

There were also no differences between groups in their

medical examination (i.e., grimacing, ambulation, or

cooperativeness). Subjects were most frequently diagnosed

with combinations of low back pain with sciatica (56%), low

back pain without sciatica (43%) myofacial syndrome (50%),

spinal stenosis (28%), lumbar spondylosis (46%) and lumbar

instability (43%).

There were no significant differences between groups in

age, duration of their pain problem, medication usage, or

previous treatment histories. There was a significant

difference found between groups in the time since last

worked with the control group reporting more time since they

last worked compared to the treatment group (F(1,42)=4.10, R

< 0.05). When addressing medical treatments and activities

during the study, there were no differences found at the end

of the protocol between the control and treatment groups.

(See Table I.)


Treatment Control
Group Group

N=31 N=23


Unemployed due to
Back Pain


Low back pain with
Low back pain
without sciatica
Myofacial Syndrome
Spinal Stenosis
Lumbar Spondylosis
Lumbar Instability

Average age

Average duration
of Pain in months

Time since
last worked
in months

Daily Hours in


44 (22-70 yrs.)

84 (12-312 mos.)

22 (0-132 mos.)

13 (2-24 hrs.)


47 (25-70 yrs.)

89 (12-288 mos.)

56 (1-168 mos.)*

15 (2-24 hrs.)

* E < 0.05

Physiological Results

It was hypothesized that exercise on the MedXTM lumbar

extension machine would strengthen the low back extensor

muscles. T-Test results indicated that there were no pre-

treatment differences between groups in measures of

isometric strength as defined by the pre-treatment intercept

scores and slopes of the regression line. At post-treatment

utilizing an analysis of covariance with pre-treatment

scores as a covariate and the change scores as the dependent

variable, there were no differences between groups in the

slope of the isometric strength curve, but there were

significant differences between the intercepts,

(F(2,52)=6.50, R < 0.01). Results indicated that the mean

intercept significantly increased in the treatment group

while it remained the same for the control group. Given the

drastic decline in the number of subjects at the 72 degree

angle, the slope of the regression line has been graphically

portrayed twice. Figure I represents the regression line

with a full range of motion with only 24% of the subjects

completing the 72 degree angle, and Figure II represents the

regression line up to 60 degrees which encompasses 65% of

the sample. (See Figures I and II.). These findings

suggest that while the linear increase in strength was

consistent across the subject's range of motion between the

two groups, the treatment group generated consistently more

force post-treatment. When addressing each standard angle

utilizing the pre-treatment torque as a covariate, the

results were consistent with the increases in intercept and

indicated that the treatment group significantly increased

their strength at all angles within the subject's range of

motion. (See Table II.)

Psychological Results

The second hypothesis suggested that increased strength

in the low back muscles would be associated with increased

functional abilities. The physical and psychosocial

subscales of the Sickness Impact Profile were utilized to

address this hypothesis. Results of the pre-treatment

analysis indicated that the control group reported

significantly more physical and psychosocial dysfunction

when compared to the treatment group. Controlling for pre-

treatment differences by utilizing the pre-treatment scores

as a covariate in an analysis of covariance, there were

significant differences between the groups in pre/post-

treatment change scores on the physical dysfunction scale.

The treatment group reduced their scores in reported

physical dysfunction after exercise and the control group

increased in their reported physical limitations

(F(2,52)=4.77, R < 0.03). A similar trend was found on the

psychosocial subscale. Controlling for pre-treatment

differences by utilizing the pre-treatment scores as a

covariate in an analysis of covariance, there were

significant differences between the groups in pre/post-

12 24 36 48 60 72














0 12 24 36 48 60 72









73.2 104.1
(45.1) (63.5)

1.4 1.4
(0.8) (0.7)

73.4 73.6 F(2,52)=6.50
(54.6) (58.8) R<0.01

1.9 1.7 ns
(1.1) (0.9)




















































treatment change scores on the psychosocial dysfunction

scale. The treatment group decreased their scores and the

control group increased in their reported psychosocial

dysfunction (F(2,52)=5.05, E < 0.03). There were no

pre/post-treatment differences in recreational activities or

routine exercise regimens on the activities questionnaire.

(See Table III.)

The third hypothesis proposed that exercise and

increased low back strength would result in beneficial

reductions in psychological distress and pain. Analysis of

pre-treatment scores on perceived stress and measures of

anxiety and depression (defined as psychological distress on

the Mental Health Inventory) indicated that both groups

experienced severely high levels of perceived stress and

psychological distress pre-treatment and the control group

expressed significantly higher levels than the treatment

group. Analysis of covariance which used pre-treatment

scores as the covariate and change scores as the dependent

measure revealed no treatment differences between groups for

stress, anxiety and depression, or psychological well-being.

Although all subjects reported high levels of stress and

distress over the course of the study, there were no pre-

treatment difference between groups on the level of self-

reported pain as measured by the pain subscale of the

WHYMPI. In contrast, utilizing pre-treatment pain scores as

the covariate in an analysis of covariance for changes in


Treatment Group Control Group
Pre/posttreatment Pre/posttreatment
(SD) (SD) p Value

Sickness Impact



Mental Health



West Haven-Yale
Pain Inventory:

Pain Subscale

Positive Support

Negative Support

9.1 7.7 15.2 19.3
(9.3) (9.4) (10.4) (15.6)

12.5 10.3 20.8 24.8
(14.3) (12.8) (18.0) (23.7)



58.8 59.0 71.7 70.3
(18.8) (20.9) (28.9) (32.5)

51.3 52.2 45.1 46.8
(13.9) (14.5) (18.1) (19.0)

3.4 2.9
(1.6) (1.7)

3.6 3.4
(1.3) (1.5)

1.2 1.2
(1.0) (1.1)

3.7 4.1
(1.6) (1.5)

2.6 3.0
(1.7) (1.5)

2.1 1.7
(1.5) (1.4)




Exercise Locus
of Control:

23.3 23.9
(5.2) (4.4)

12.0 11.7
(3.7) (3.7)

21.8 19.9
(5.2) (6.7)

11.6 13.1
(3.8) (3.8)





reported pain, there was a significant difference found

between groups with the treatment group reporting a

significant reduction in pain and the control group

reporting an increase in pain (F(2,51)=6.83, R < 0.002).

It was originally hypothesized that social support,

self-efficacy, and pain reports would predict attrition or

adherence to the exercise program, and therapeutic gains.

Since attrition was limited (i.e., one drop-out), factors

associated with attrition could not be ascertained.

Analysis of the Social Support subscales of the WHYMPI

indicated pre-treatment differences in perceived support

with the treatment group reporting higher levels of positive

support and the control group higher levels of negative

support. In contrast, there were no post-treatment

differences in perceived negative or positive support which

may be reflective of a simple regression towards the mean.

In assessing the individual's perceived ability to master

the exercise program, there were no pre-treatment

differences between groups on internal or external locus of

control, but there were significant post-treatment

differences found. An analysis of covariance with pre/post-

treatment change scores indicated that the treatment group

maintained a high internal locus of control whereas the

control group decreased on internal locus of control

(F(2,50)=6.07, R < 0.02) and changed to a perceived external

locus of control (E(2,50)=4.59, R < 0.04). (See Table III.)

In order to assess pre-treatment predictors of

therapeutic gain, the five variables measuring positive

support, pain, past week activity levels, and internal or

external locus of control were entered into a stepwise

regression model with post-treatment intercepts as the

dependent variable. The regression analysis indicated that

28% of the variance was attributable to pre-treatment

measures of the past weeks activity level, pain, and

external locus of control (F(3,52)=6.22, R < 0.001). The

relationship between pre-treatment pain and post-treatment

intercepts accounted for 19% of the total variance in the

model and was the only significant variable influencing

strength outcome [F(4,48)=4.81, p < 0.002). (See Table IV.)

In order to assess the relationship between physical

strength changes, self-reported pain, and psychological

distress, a Pearson's correlational analysis was performed

looking at pre/post-treatment intercept change scores and

post-treatment psychological measures. Results indicated

that increased strength was positively correlated with an

internal locus of control (r=0.34, R<0.01), and negatively

correlated with higher levels of dysfunction on the physical

and psychosocial impact scales of SIP (r=-0.56, R<0.0001 and

[=-0.45, E<0.0006, respectively), with higher pain reports

(r=-0.39, E<0.004), and with increased psychological

distress on the MHI (r=-0.30, R<0.03). On the other hand,

high reports of pain post-treatment were not only negatively

related to increased strength, but with higher levels of

internal locus of control (r=-0.53, E<0.0001) and with

increased psychological well-being as measured by the MHI

(r=-0.56, R<0.0001). Pre/post-treatment changes in self-

reported pain were not significantly correlated with any of

the above variables. (See Table V.)



Step Variable Model Partial F p Beta
R2 R2 Statistic Value Weight

1 Pain 0.19 0.19 12.12 0.001 -12.3

2 Activity 0.24 0.05 3.05 0.08 12.7

3 External LOC 0.28 0.04 2.52 0.12 3.5

4 Internal LOC 0.28 0.01 0.69 0.41 1.3


VARIABLE 1 2 3 4 5 6 7 8 9 10 11 12

PAIN -.34
7. DISTRESS -.30
8. WELL-
9. STRESS -.25
LOC .34
LOC .04


.68 .67

.65 .64 .72

.21 .20
.65 .84




-.40**-.64** -.56** -.40** -.10
.54 .68 .67 .44 .18

-.63 -.67 -.53 -.41 -.26

.18 .23

.14 .38

.13 .01

.88 -.78

-.57 .39 -.54

.10 .02 .21 .21 -.09

.23 .26 -.02 .40 -.40

.19 -.38

.24 -.13 -.03

.13 .06 .07 -.02 .09 .05 -.27 .19 -.40

* p < 0.05
** E < 0.01


This study examined the effects of exercise on the

MedXT lumbar extension machine in a chronic low back pain

syndrome population. It was hypothesized that after a 12

week exercise program subjects would improve in lumbar

extensor strength and that improvements in strength would be

associated with decreased physical and psychological


The findings of this study confirmed that exercise on

the MedxTM lumbar extension machine substantially increased

low back strength in a sample of chronic low back pain

patients. These findings are of particular importance given

the chronicity and disability reported by this sample.

Additionally, increased strength was related to improved

physical and psychosocial functioning as measured by the

Sickness Impact Profile. On this measure, the treatment

group decreased in their report of dysfunction while the

control group increased on measures of physical and

psychosocial impairment. Interestingly, despite reporting

improved physical and psychosocial functioning, there were

no differences or changes in daily activity levels.

Exercise on the MedXTM lumbar extension machine was also

related to significant changes in self-reported pain.


Again, the treatment group decreased their pain reports

while the control group reported even higher levels of pain

at the end of the study. Changes in pain and physical

dysfunction were hypothesized to result in decreases in

psychological distress (depression, anxiety, and stress),

but the findings of this study did not support this

hypothesis. Subjects in this study reported significantly

high levels of depression, anxiety, and stress prior to

entering the protocol and at the end of treatment, both

groups continued to report significantly high levels of

psychological distress.

It was hypothesized that pre-treatment measures of

self-efficacy, pain, and perceived social support would be

related to adherence to the treatment program and to

therapeutic gains. All but one subject completed the

treatment program; therefore, attrition could not be

investigated. With respect to predicting treatment outcome

from pre-treatment measures of pain, activity levels, self-

efficacy, and social support; pre-treatment pain was most

predictive of post-treatment strength changes.

Additionally, there was a strong relationship between

changes in strength and perceived self-efficacy (as measured

by the Internal Locus of Control subscale of the Exercise

Objectives Locus of Control Scale). Contrary to the

original hypotheses, social support was not related to

treatment outcome measures.

Physiological Findings

Subjects were tested for isometric strength of the

lumbar extensor muscles on the MedXTM lumbar extension

machine at the beginning and end of this 12 week treatment

protocol. Due to the chronicity and physical limitations of

this sample, many of the subjects were unable to complete a

full 72 degrees range of motion. In fact, 75% of the

population could not flex at the 72 degree angle and 35%

could not flex at 60 degrees. Despite these limitations in

range of motion, statistical analysis resulted in finding

that the treatment group as compared to the control group

significantly increased in their generated torque values

following only 14 exercise sessions over a 10 week period.

The statistics used to assess strength changes pre- and

post-treatment standardized the data for group comparisons.

Given that subjects could not complete a full range of

motion, isometric testing points varied a few degrees across

subjects within their differing abilities. Hence,

regression equations were computed for each subject pre- and

post-treatment and the resultant slopes and intercepts of

the regression lines were utilized for measuring change.

Results of the statistical analysis indicated that there

were no differences between groups in the slopes of the

regression line pre- or post-treatment. However, the

control group failed to generate a straight line and

exhibited a marked deviation at the 72 degree angle. Given

the small sample size at this angle (18% of the control

group) and the variability in the torque values generated,

it could be assumed that this is a measurement error and not

indicative of a non-linear relationship. Additionally, when

the slopes of the strength curve are compared ranging to

only 60 degrees of flexion (72% of the total sample), the

discrepancy in the linear relationship is no longer evident.

Based on the findings that the slopes of the isometric

strength curves were similar pre- and post-treatment,

comparisons were performed between groups on changes in the

intercept. There were no pre-treatment differences between

the intercepts of the control group and the treatment group.

Results suggest that the control group was stronger pre-

treatment at the 60 degree angle. Again, given the smaller

sample size and the increased variation in torque values at

this angle, these findings may also be an artifact of

measurement. On the other hand, given that the control

group's post-wait testing remained higher than the treatment

group's pre-treatment testing, the control group may have

been stronger at this measurement point. Nevertheless, the

decline in torque for the control group as compared to the

increase in torque for the treatment group support the

hypothesis that exercise on the MedXT lumbar extension

machine increases lumbar extension strength. Analyses of

changes in intercept indicated that the treatment group

significantly changed their intercept scores as compared to

the control group. Since an increase in intercept is

indicative of an increase in. torque generated, it was

concluded that isotonic exercise on the MedXT lumbar

extension machine resulted in increased muscle strength


An additional analysis comparing change scores at each

standard angle supported the results of the changes in the

intercepts. Standard angles were calculated for each

subject based on their actual range of motion when their

testing positions were found to vary a few degrees.

Comparisons were then performed utilizing pre-treatment

standard angles as the covariate in an analysis of

covariance with changes in torque as the dependent variable.

This analysis supported the previous analysis and indicated

that training on the MedXT lumbar extension machine

increased lumbar strength by 11 to 17 percent whereas the

control group remained the same.

Psychological Findings

One of the most pronounced findings in this study was

that psychological factors found to be associated with

treatment outcome consistently improved in the treatment

group and worsened in the control group. Specifically,

treatment on the MedXT lumbar extension machine resulted in

significant improvements in reported physical and

psychosocial dysfunction as measured by the Sickness Impact

Profile. Over the course of the 12 week protocol, the

treatment group reported less physical and psychosocial

disruption while the control group reported more

impairments. This measure of impairment covers a broad

array of physical, emotional, and social situations in which

illness behaviors have a negative impact. As is evident in

the correlational analysis, the two subscales are highly

correlated and therefore, it is not surprising that finding

a change in one subscale resulted in a change in the other.

Interestingly, there were no concomitant changes found in

weekly activity levels, depression and anxiety, or stress.

These findings may be indicative of the sensitivity of the

measuring instrument, or that perceived increased physical

abilities does not readily result in increased activities.

Additionally, global measures of psychosocial functioning

may change more rapidly than more specific measures of

depression, anxiety, and stress.

The subjects participating in this study experienced

high pre-treatment levels of depression, anxiety, and stress

as measured by the Mental Health Inventory and Perceived

Stress Survey. Pre-treatment scores on the Sickness Impact

Profile were also significantly elevated. Of concern is

that the control group generally reported higher levels of

psychological distress pre-treatment as compared to the

treatment group. In addition, the control group was found

to have been unemployed due to pain for a longer duration of

time than the treatment group. It might be speculated that

the control group reported higher levels of psychological

distress in response to their less stable economic and

social situation as a result of longer durations of

unemployment. This is only speculative, but given that

physical findings, activities, pain behaviors, and pain

report were not significantly different between the groups

pre-treatment, it appears to be a reasonable hypothesis.

Analysis of differences in measures of depression,

anxiety, and stress indicated no differences between groups

at the end of the treatment study. These psychological

measures of distress were significantly elevated pre-

treatment for both groups and remained high over the 12 week

study. Studies addressing the effects of exercise on

depression and anxiety have found that mild to moderately

elevated distress appears to improve with exercise, not

severe elevations of distress (Sinyor et al., 1983). Hence,

the level of psychological distress in this study's

population was sufficiently high that more intensive

interventions may be needed to specifically address these

patients' depression, anxiety, and stress. Additionally, as

was noted in the pre-treatment analysis, the subjects

participating in this study experienced pain for extended

durations of time, were out of work for long time periods,

and half of the sample were dependent on disability or

workmen's compensation for their primary source of income.

The impact on psychological well-being of increased

chronicity of pain and the associated socio-environmental

influences was beyond the scope of this study, but these

factors surely play a significant role in the maintenance of

psychological distress. Hence, distress potentially

associated with socio-environmental factors would not be

expected to respond to an intervention of exercise, but

would require interventions aimed at ameliorating the

distressing situations. In summary, exercise on the MedXTM

lumbar extension machine improved low back strength, but did

not result in significant decreases in depression, anxiety,

and stress in this patient population.

After addressing the effects of exercise on self-

reported pain, this study found that the treatment group

decreased their pain reports while the control group

increased their report of pain. There were no pre-treatment

differences in reported pain. Hence, these findings lend

support to the hypothesis that patients with low back pain

may avoid activities that previously created pain which

results in muscular disuse and atrophy. Atrophied muscles

potentially increase pain independent of the original pain

stimuli (Feuerstein et al., 1987; Rosomoff, 1985). In

summary, exercise of the low back muscles resulted in lower

levels of reported pain in the treatment group while the

control group's report of pain continued to escalate.

An important element of this study addressed the role

of self-efficacy in treatment adherence and outcome. There


was only one drop out in this study; therefore adherence was

not examined. Self-efficacy was measured with the Exercise

Locus of Control Scale. This measure assessed the subjects

perceived ability to master the exercise program in terms of

internal and external locus of control. Interestingly,

there were no pre-treatment differences between groups in

locus of control with both groups reporting a higher score

on internal locus of control. At post-treatment, the

control group's mean scores changed from a higher internal

locus of control to a higher external locus of control. In

contrast, the treatment group maintained their orientation

for an internal locus of control. These findings are

consistent with self-efficacy theory and indicate that

mastery of a given behavior will influence self-efficacy

expectations which will subsequently influence behavioral


In this study, internal locus of control (increased

self-efficacy) was the only psychological measure positively

correlated with changes in strength. Pre-treatment scores

of internal locus of control (self-efficacy) were not found

to be predictive of therapeutic gain as originally

hypothesized, but increased strength (treatment success) was

highly associated with maintenance of high self-efficacy

ratings. The change of focus from an internal to an

external locus of control in the control group suggests that

as pain persists and passive treatment modalities (i.e.,


hot/cold packs and massage) fail to ameliorate the symptoms,

the subject's self-efficacy expectations decline and in

turn, they seek external resources for meeting their needs.

On the other hand, subjects that experienced success with

exercise and subsequent reductions in pain were more apt to

internalize their treatment goals and maintain strong self-

efficacy expectations.

It was originally hypothesized that social support

would predict attrition and therapeutic gains. Attrition

was not addressed in this study due to a high adherence

rate. Neither negative nor positive social support as

measured on the WHYMPI were predictive of changes in

strength at the end of treatment. Although the majority of

this sample were married, a supportive relationship was not

related to therapeutic changes in strength or pain. The

exercise literature suggests that social support is an

important factor in maintaining exercise objectives over

time. In contrast in the chronic pain patient, the role of

pain may be the predominant factor associated with

therapeutic endeavors which supersedes the role of social


In conclusion, changes in strength were positively

associated with decreased pain reports and increased

physical and psychosocial functioning. There was no

relationship between changes in strength and psychological

distress. Additionally, the predictive ability of pre-

treatment measures of activities, locus of control, pain,

and social support was limited to 28 percent of the variance

being explained by reported pain, the previous week's

activity level, and external and internal locus of control.

Of these measures, pain was the single most influential

variable and accounted for 19 percent of the variance.

Social support did not contribute to the predictive ability

of therapeutic gain. There was only one positive

correlation with strength changes and that was with internal

locus of control (self-efficacy). Measures of psychological

distress, pain, and physical and psychosocial impairments

were significantly correlated with decreased changes in



The findings of this study support the hypothesis that

strengthening the lumbar extensor muscles in a chronic pain

population is a beneficial treatment modality. The

literature reviewed suggested that increased physical

fitness and strength was an important treatment modality,

but the specific role of exercise in a chronic low back pain

syndrome population has not been clearly documented.

Although the subjects participating in the current study

were treated in an out-patient program, they do resemble

subjects found in multidimensional pain treatment programs

in terms of their chronicity of pain problem, psychological

distress, and socio-environmental disruption. Hence, the

findings of this study have potential implications for the

severely disabled low back pain population as well as the

less disabled back pain patient as is typically found in

out-patient treatment programs. Of particular interest is

the fact that those who suffer with significantly high

levels of psychological distress and physical disabilities

exhibit beneficial changes with lumbar extension exercise.

This would suggest that a less severely disabled group might

demonstrate greater therapeutic gains with low back exercise

because of lower levels of pre-treatment pain and

psychological distress. However, the improvements found in

this chronic pain syndrome population may have been greater

given the high levels of dysfunction.

In addition to increased strength with exercise on the

MedXT lumbar extension machine, patients reported less pain

post-treatment. Decreased pain reports are important in

that it is the subjective experience of pain that results in

patients seeking health care services and it is the report

of pain that often determines treatment interventions

(Kleinke & Spangler, 1988). Hence, exercise and increased

strength of low back muscles reduces reported pain and has

the potential for a reduction in the over utilization of the

health care system associated with this population.

Although exercise on the MedXT lumbar extension machine did

not result in reported increased activity levels, there was

a decline in perceived physical and psychosocial


limitations. This finding suggests that life-style changes

in daily activities may not be related to perceived changes

in limitations. Hence, exercise alone does not impact on

activities of daily living and instead, socio-environmental

factors may be more determinant and should be addressed


Exercise alone did not influence measures of

psychological distress. This finding is consistent with the

literature reviewed in that these subjects reported severe

levels of psychological distress. It may be that as the

chronicity of pain and disability increase, that

psychological distress is more closely associated with

socio-environmental influences (e.g., workmen's

compensations, unemployment, availability of work, financial

stability, and adaptation to a sick role) than to physical

limitations and pain. The implications from these findings

are that patients exhibiting a chronic low back pain

syndrome improve in physical strength, and they experience a

reduction in pain following physical exercise. The

perceived or actual limitations for returning to gainful

function in society remains impaired and potentially

exacerbates psychological distress. Hence, the failure to

find improvement in psychological distress in this study

suggests that multidisciplinary treatment programs that

emphasize physical reconditioning need to address vocational

and socio-environmental factors for the successful

rehabilitation of the chronic low back pain syndrome


Treatment success was predicted by decreased reports of

pain, increased activity levels, and higher self-efficacy

expectations, and not by social support. Self-efficacy

expectations are important in the rehabilitation process and

are related to an internal attribution following treatment

gains. Findings of this study suggest that instructions in

the benefits of exercise and increased strength potentially

increased self-efficacy expectations and that following an

inability to experience a change with treatment

interventions, the patient's self-efficacy expectations

changed to an external reliance for help. These findings

suggest that patients who are encouraged to take an active

role in their rehabilitation adapt an internal attribution

for treatment success which is associated with actual

therapeutic gains. In contrast, patients who wait for

helpful treatment interventions may become more dependent on

others for help which potentially increases their

psychological distress and pain. Hence, the role of

internal locus of control (self-efficacy expectations) in

managing the chronic low back pain syndrome has important

implications for future improvements in performance and


In this study, treatment success was defined by

increased strength and decreased pain, and not by return to

work. This study attempted to focus on the specific effects

of exercise in the rehabilitation of chronic pain patients.

Studies addressing treatment outcome with return to work are

heavily weighted with socio-environmental influences and it

is therefore difficult to evaluate the beneficial treatment

components. In summary, exercise of the low back in a

chronic low back pain syndrome population is a necessary

treatment component, but it does not in and of itself,

ameliorate psychological distress or socio-environmental

influences. This implies that physical rehabilitation,

psychological rehabilitation, and return to work are

distinct areas of the treatment process and need to be

evaluated independently.

Limitations of the Study

This study examined the relationship between

psychological distress and exercise of the low back muscles

after twelve weeks of treatment in a chronic low back pain

syndrome population. Although subjects were randomized at

the beginning of treatment, the control group was found to

exhibit higher levels of psychological distress and

dysfunction pre-treatment. The finding of increased

psychological distress and pain in the control group as

compared to the treatment group may have been influenced by

these pre-treatment differences.

The subjects participating in this study exhibited a

chronic pain syndrome and this specific patient population

represents only a small portion of all individuals suffering

with chronic low back pain. Thus, these findings are

relevant only to patients with severe low back pain and

psychosocial dysfunction. Hence, a more heterogeneous

population with mild to moderate chronic low back pain who

were better psychologically and socially adjusted may have

exhibited different results.

This study only addressed one component of treatment

(exercise of the low back) in a chronic pain population and

compared changes in physical and psychological functioning

post-treatment with a wait-list control group. Other

important comparisons for investigation would have

potentially provided even more information regarding the

effectiveness of exercise in the treatment of chronic low

back pain. For example, an additional group engaged in a

commonly prescribed general exercise program for chronic low

back pain compared with MedXTM training might have been more


The subjects participating in this study were

significantly impaired in their ability to exercise through

a complete 72 degree range of motion. Performances on the

MedXT lumbar extension machine were highly variable and

group comparisons required statistical analysis for

standardizing the measuring points. Ideally, each subject

would have tested at the standard angles through the full

range of motion (0 to 72 degrees) and a standard

multivariate analysis of variance could have been utilized.

The examination of differing exercise regimens on the

MedXT lumbar extension machine would have also been

informative. Although previous research in healthy

individuals found that one exercise session per week was as

effective as two and three sessions per week, there is no

research addressing the effects of exercise frequency in a

chronic pain syndrome population (Graves et al., 1990). In

this study, all treatment subjects exercised twice a week

for six weeks followed by once a week for six weeks. Given

the chronicity of this samples disability, it can only be

speculated that exercise twice a week for the entire 12 week

treatment program may have resulted in even more dramatic

strength changes which may have potentially resulted in

beneficial changes in activities and psychological distress.

The current study demonstrated that the control group

increased in measures of pain and psychological distress

which suggests that the role of expectancy for getting

better may have influenced the outcome of the treatment and

control groups. The increased distress and pain evident in

the control group may have been a function of having to wait

for treatment. On the other hand, the increased pain and


psychological distress in the wait-list group may have been

a result of the continued lack of beneficial treatment


Conclusions and Future Directions

This study found that after 14 exercise sessions on the

MedXT lumbar extension machine, a chronic low back pain

syndrome population significantly increased in measured

strength and decreased in reported pain. There were

significant changes found in physical and psychosocial

dysfunction post-treatment. Treatment gains were associated

with enhanced self-efficacy which has the potential for

positive impacts on future performance and disability.

Future research is needed to address the different

patient populations with chronic low back pain that would

benefit from exercise on the MedXTM lumbar extension

machine. Perhaps, patients with less physical and

psychosocial impairments would have experienced even greater

physical and psychological gains.

Given the chronicity of pain and deficits in physical

strength of this patient population, a treatment program

longer than 12 weeks may be necessary. A longer treatment

program may produce different results in terms of increases

in strength as well as increases in range of motion.

Additional research is needed to address the long-term

benefits of exercise on the MedXTM lumbar extension machine

in a chronic low back pain syndrome population. Future


research is currently under way to investigate the long-term

effects of exercise on the MedXTM lumbar extension machine

on psychological distress, return to work, and changes in

daily activities.

In conclusion, this study demonstrated that exercise on

the MedXT lumbar extension machine significantly

strengthened the low back extensor muscles in a population

of chronic low back pain syndrome patients. Associated

improvements were also found in the experience of pain and

physical and psychosocial dysfunction. Given that strength

changes with specific exercises of the extensor muscles have

been shown to increase strength in both healthy and low back

pain subjects, future research is needed to address the

potential benefits of this exercise modality on prevention

of future back injuries and the chronic low back pain



Ackerman, M.D. and Stevens, M.J. (1989). Acute and chronic
pain: pain dimensions and psychological status. Journal
of Clinical Psychology, 76, 223-228.

Bandura, A. (1982). Self-efficacy mechanism in human
agency. American Psychologist, 37, 122-147.

Bandura, A., O'Leary, A., Taylor, C.B., Gauthier, J., and
Gossard, D. (1987). Perceived self-efficacy and pain
control: opiate and nonopioid mechanisms. Journal of
Personality and Social Psychology, 53, 563-571.

Bergner, M., Bobbitt, R.A., Pollard, W.E., Martin, D.P., and
Gilson, B.S. (1981). The sickness impact profile:
Validation of a health status measure. Medical Care, 14,

Bigos, S.J. and Battie, M.C. (1987). Acute care to prevent
back disability. Ten years of progress. Clinical
Orthopaedics and Related Research, 221, 121-130.

Blackwell, B., Galbraith, J.R., and Dahl, D.S. (1984).
Chronic pain management. Hospital and Community
Psychiatry, 35, 999-1008.

Blair, S.N., Haskell, W.L., Ping Ho, Paffenbarger, R.S.,
Jr., Vranizan, K.M., Farquhar, J.W., and Wood, P.D.
(1985). Assessment of habitual physical activity by a
seven-day recall in a community survey and controlled
experiments. American Journal of Epidemiology, 122, 794-

Block, A.R. (1982). Multidisciplinary treatment of chronic
low back pain: A review. Rehabilitation Psychology, 27,

Bono, S.F. and Zasa, M.L. (1988). Chronic low back pain and
therapy: A critical review and overview. The Behavior
Therapist, 11, 189-198.

Bortz II., W.M. (1984). The disuse syndrome. The Western
Journal of Medicine, 141, 691-694.

Chapman, C.R., Casey, K.L., Dubner, R., Foley, K.M.,
Gracely, R.H., and Reading, A.E. (1985). Pain
measurement: An overview. Pain, 22, 1-31.

Chaturvedi, S.K., Varma, V.K., and Malhotra, A. (1984).
Non-organic chronic intractable pain: A comparative
study. Pain, 19, 87-94.

Clark, W. C. (1974). Pain sensitivity and the report of
pain. Anesthesiolorv, 40, 272-287.

Clark, W.C. and Yang, J.C. (1983). Applications of sensory
decision theory to problems in laboratory and clinical
pain. In R. Melzack (Ed.), Pain Measurement and
Assessment, (pp. 15-25). New York: Raven Press.

Cohen, S., Kamarck, T., and Mermelstein, R. (1983). A
global measure of perceived stress. Journal of Health
and Social Behavior, 24, 385-396.

Coleman, M., Price, S., and Washington, M. (1985). Physical
exercise, social background, and the well-being of older
adult women. Perceptual and Motor Skills, 60, 737-738.

Craig, K.D. (1983). A social learning perspective on pain
experience. In M. Rosenbaum, C.M. Franks, & Y. Jaffe
(Eds.), Perspectives on Behavior Therapy in the Eighties.
(pp. 311-327). New York: Springer Publishing.

Craig, K.D., Best, H., and Ward, L.M. (1975). Social
modeling influences on psychophysical judgments of
electrical stimulation. Journal of Abnormal Psychology,
84, 366-373.

Deyo, R.A., Bass, J.E., Walsh, N.E., Schoenfeld, L.S., and
Ramamurthy, R. (1988). Prognostic variability among
chronic pain patients: Implications for study design,
interpretation, and reporting. Archives of Physical
Medicine and Rehabilitation, 69, 174-178.

Dishman, R.K. (1985). Medical psychology in exercise and
sport. Medical Clinics of North America, 69, 123-143.

Dishman, Sallis, and Orenstein (1985). The determinants of
physical activity and exercise. Public Health Reports,
100, 158-171.

Dolce, J.J. (1987). Self-efficacy and disability beliefs in
behavioral treatment of pain. Behavior Research and
Therapy, 25, 289-299.

Dolce, J.J., Crocker, M.F., and Doleys, D.M. (1986).
Prediction of outcome among chronic pain patients.
Behavior Research and Therapy, 24, 313-319.

Dolce, J.J., Crocker, M.F., Moletteire, C., and Doleys, D.M.
(1986). Exercise quotas, anticipatory concern and self-
efficacy expectancies in chronic pain: A preliminary
report. Pain, 24, 365-372.

Dolce, J.J., Doleys, D.M., Raczynski, J.M., Lossie, J.,
Poole, L., and Smith, M. (1986). The role of self-
efficacy expectancies in the prediction of pain
tolerance. Pain, 27, 261-272.

Dolce, J.J. and Raczynski, J.M. (1985). Neuromuscular
activity and electromyography in painful backs:
Psychological and biomechanical models in assessment and
treatment. Psychological Bulletin, 97, 502-520.

Doyne, E.J., Ossip-Klein, D.J., Bowman, E.D., Osborn, K.M.,
McDougall-Wilson, I.B., and Neimeyer, R.A. (1987).
Running versus weight lifting in the treatment of
depression. Journal of Consulting and Clinical
Psychology, 55, 748-754.

Ewart, C.K., Taylor, C.B., Reese, L.B., and DeBusk, R.F.
(1983). Effects of early postmyocardial infarction
exercise testing on self-perception and subsequent
physical activity. The American Journal of Cardiology,
51, 1076-1080.

Feuerstein, M., Papciak, A.S., and Hoon, P.E. (1987).
Biobehavioral mechanisms of chronic low back pain.
Clinical Psychology Review, 7, 243-273.

Feuerstein, M., Sult, S., and Houle, M. (1985).
Environmental stressors and chronic low back pain: Life
events, family and work environment. Pain, 22, 295-307.

Finneson, B. (1973). Low Back Pain. Philadelphia:

Flor, H. and Turk, D.C. (1989). Psychophysiology of chronic
pain. Do chronic pain patients exhibit symptom-specific
psychophysiological responses? Psychological Bulletin,
105, 215-259.

Folkins, C.H., Lynch, S., and Gardner, M.M. (1972).
Psychological fitness as a function of physical fitness.
Archives of Physical Medicine and Rehabilitation, 53,

Folkins, C.H. and Sime, W.E. (1981). Physical fitness
training and mental health. American Psychologist, 36,

Follick, M.J., Smith, T.W., and Ahern, D.K. (1985). The
sickness impact profile: A global measure of disability
in chronic low back pain. Pain, 21, 67-76.

Fordyce, W.E. (1988). Pain and suffering. American
Psychologist, 43, 276-283.

Fredrickson, B.E., Trief, P.M., VanBeveren, P., Yuan, H.A.,
and Baum, G. (1988). Rehabilitation of the patient with
chronic back pain: A search for outcome predictors.
Spine, 13, 351-353.

Gatchel, R.J., Mayer, T.G., Capra, P., Barnett, J., and
Diamond, P. (1986). Million Behavioral Health Inventory:
Its utility in predicting physical function in patients
with low back pain. Archives of Physical Medicine
Rehabilitation, 67, 878-882.

Gatchel, R.J., Mayer, T.G., Capra, P., Diamond, P., and
Barnett, J. (1986). Quantification of lumbar function
part 6: the use of psychological measures in guiding
physical functional restoration. Spine, 11, 36-42.

Gottlieb, H., Strite, L.C., Koller, R., Madorsky, A.,
Hockersmith, V., Kleeman, M., and Wagner, J. (1977).
Comprehensive rehabilitation of patients having chronic
low back pain. Archives of Physical Medical
Rehabilitation, 58, 101-108.

Graves, J.E., Pollock, M.L., Foster, D., Leggett, S.H.,
Carpenter, D.M., Vuoso, R., and Jones, A. (1990). Effect
of training frequency and specificity on isometric lumbar
extension strength. Spine, In Press.

Hazard, R.G., Fenwick, J.W., Kalisch, S.M., Redmond, J.,
Reeves, V., Reid, S., and Frymoyer, J.W. (1989).
Functional restoration with behavioral support: A one
year prospective study of patients with chronic low back
pain. Spine, 14, 157-161.

Heinrich, R., Cohen, M., Naliboff, B., Collins, G., and
Bonebakker, A. (1985). Comparing physical and behavior
therapy for chronic low back pain on physical
abilities, psychological distress, and patient's
perceptions. Journal of Behavioral Medicine, 8, 61-78.

Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EEV6FGFZT_8TSO2Y INGEST_TIME 2013-01-23T14:51:13Z PACKAGE AA00011141_00001