Regression analysis of psychophysiological stress, cumulative trauma disorder symptoms, and cognitive performance decrement


Material Information

Regression analysis of psychophysiological stress, cumulative trauma disorder symptoms, and cognitive performance decrement
Physical Description:
vi, 74 leaves : ill. ; 29 cm.
Lane, John C
Publication Date:


Subjects / Keywords:
Exercise and Sport Sciences thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Exercise and Sport Sciences -- UF   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 2004.
Includes bibliographical references.
Additional Physical Form:
Also available online.
Statement of Responsibility:
by John C. Lane.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 022438601
oclc - 847494697
System ID:

Full Text








I thank the members of my supervisory committee, James H. Cauraugh, Ph.D.

(chair), Ira S. Fischler, Ph.D., Christopher M. Janelle, Ph.D., and Mark D. Tillman, Ph.D.

Their patience during this effort is appreciated.

I also wish to express my gratitude to Deborah J. Boeff, RN, BSN, COHN-S, and

to Trina M. Girimont, MSN, ARNP, COHN-S, successive Directors of Occupational

Health Services, Human Resources Division, Shands HealthCare. Their dedication to

employee health and safety made this undertaking possible.


ACKNOWLEDGEMENTS.................................................................. ii

A B STRA C T .................................................................................... v



Stress and H ealth..................................................................... 10
Stress and Performance.............................................................. 14
Definitional Shortcomings.......................................................... 18
Correlating the Impact of Stress................................................... 19
Theoretical Bases for the Proposed Research........................... 21
Physiological Bases for the Proposed Research........................ 26

2 METHOD............................................................................. 35

Research Design...................................................................... 35
Measurement Instruments........................................................... 35
Cumulative Trauma Disorder Symptoms................................ 35
Cognitive Performance Decrement....................................... 36
Cognitive Anxiety.......................................................... 36
Physiological Arousal...................................................... 37
Participants............................................................................ 38
Minority Inclusion.......................................................... 38
Recruitm ent................................................................... 39
Procedures........................................................................... 41
Statistical A nalysis................................................................... 43
Response Variables......................................................... 43
Predictor Variables.......................................................... 43
Regression Models.......................................................... 44

3 RESULTS............................................................................. 45

Cumulative Trauma Disorder Symptoms......................................... 49
Cognitive Performance Decrement................................................ 51

4 DISCUSSION........................................................................ 53


A DISCOMFORT SURVEY.......................................................... 62





F SAMPLE AROUSAL DATA REPORT.......................................... 68

REFERENCE LIST........................................................................... 69

BIOGRAPHICAL SKETCH................................................................ 74

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



John C. Lane

May, 2004

Chair: James H. Cauraugh
Major Department: Exercise and Sport Sciences

The objective of this research was to develop a means by which safety and health

practitioners could quantify the contribution ofpsychophysiological stress to the etiology

of cumulative trauma disorders (CTDs) among employees. The specific aim was to

develop two regression models that quantify the stress/injury and stress/performance

relationships theorized by Andersen and Williams. The CTD regression model and the

experimental method that produced the model could then be used to justify stress

management therapies in the treatment of CTDs.

The two dependent variables for the regression models were (1) a self-reported

scale rating of each participant's level of musculoskeletal discomfort, and (2) each

participant's score on the Stroop Color-Word Interference Test. For both models, the

independent variables were a combination of cognitive trait anxiety and state anxiety

physiological arousal measurements. Specifically, the cognitive anxiety variables were

the participants' scores on the Brief Stress Inventory (BSI). The physiological arousal

variables were the participants' Autonomic Lability Scores for (1) frontalis EMG, (2)

digital temperature, and (3) digital electrodermal responses.

Independent variable physiological arousal measurements were taken during a

two-min baseline and during the first 10 min of three randomly assigned treatment

conditions (1) during surveys for chest and upper extremity, and back and lower

extremity musculoskeletal discomfort (the scores for the first dependent variable), (2)

during the Stroop Color-Word Interference Test (with the average response times used as

the second dependent variable), and (3) while the participants completed the BSI (the

scores for the remaining independent variables). As an observational survey, the

experimental design exploited the reciprocal interaction between stress and stressor, and

circumvented the theoretical and definitional complications of discriminating between the


Multiple regression analyses of the data produced by the three conditions

provided two predictive equations for CTDs and cognitive performance. The BSI

cognitive anxiety subscales for frustration and self-esteem were significant and plausible

predictors for CTD symptoms. The experimental method regarding CTD symptoms

could be repeated to serve as a noninvasive intervention tool to justify stress management

therapies in the treatment of CTDs.


Imagine a middle-aged employee who has worked as a transcriptionist for most of

her adult life. Her work is not easy; day after day she types for at least eight hours.

About six months ago a painful tingling in her hands woke her several times a night.

Five months ago, when the pain had not subsided and she began to experience the pain

while typing, she reluctantly informed her supervisor. Her supervisor advised her to

report to the company's Occupational Health Services Department for evaluation. She

was diagnosed as having carpal tunnel syndrome.

During this same period, other life events were prompting change. For one, a

granddaughter was born. This was a happy occasion but stressful nonetheless. To make

matters worse, in spite of the pain it caused to her wrists, the grandmother could not resist

picking up and holding her new granddaughter. Within a month, the progress made in

rehab had disappeared and the pain became almost constant. Nights of interrupted sleep

continued. She now felt tired most of the time.

With only three years left to retirement, but financially bound to working until

then, staying at the keyboard became a matter of labored endurance. In addition, she

realized that her productivity was decreasing as a consequence of her mental state. The

worries and the pain were distracting. As a last resort, to escape the months of pain and

the threat to her livelihood, she agreed to carpal tunnel release surgery.

Now imagine that you are an ergonomist whose job is to prevent such work-

related injuries. You view this progression to surgery as failure, and are left to ponder

alternatives to avoid surgery. To answer this question, the ergonomist investigates the

etiologies of cumulative trauma disorders. Was it age, the repetitive strain of typing, an

inherited weakness, poor ergonomic work practices, the psychosocial stress associated

with the injury and other life events, sleep loss and fatigue, or a combination of all of

these? The logical response is to prevent and eliminate as many of the plausible risk

factors for injury and performance decrement as possible.

Most work-related musculoskeletal injury prevention and treatment practices are

well established: job design and ergonomic work practices that fit the job to the physical

and psychological capabilities of the employee, icing, therapeutic exercise, splinting, rest,

and antiinflammatories. A glaring exception is how work-related injury prevention and

treatment practices virtually ignore addressing the causal role ofpsychophysiological

stress. One reason this circumstance prevails is that the contribution of stress as a causal

factor, though abundantly theorized, and retrospectively correlated, has never been

quantified in a manner that readily accommodates intervention. Baker and Karasek

(2000) contend that there is no simple, yet comprehensive, instrument for measuring

stressors and stress reactions. Consequently, health maintenance organizations and third

party administrators reimburse for surgery, but are not inclined to pay for stress

management training as a conservative treatment for select work-related musculoskeletal


This experiment was designed to develop two regression models that quantify the

relationship ofpsychophysiological stress to work-related musculoskeletal disorder


symptoms, and to cognitive performance decrement. Health and safety professionals

could use these regression models and the experimental methodology to justify stress

management interventions as treatments for certain work-related musculoskeletal


In 1995, the Occupational Safety and Health Administration defined cumulative

trauma disorder (CTD) as an injury or an illness of the muscles, tendons, ligaments,

peripheral nerves, joints, cartilage, bones and/or supporting blood vessels in either the

upper or lower extremities, or back; these injuries are associated with workplace risk

factors and are not the result of instantaneous events. Clinical diagnoses of CTD include

low back pain, sciatica, tendonitis, epicondylitis, rotator cuff tendonitis, synovitis,

DeQuervain's disease, carpal tunnel syndrome, and Raynaud's phenomenon.

According to the National Institute for Occupational Safety and Health (1994),

employees experienced more disabling symptoms from such musculoskeletal problems

than from any other category of disease. In the year 2000 for example, over 577,800

musculoskeletal disorders were reported, accounting for more than one out of three of the

injuries and illnesses involving lost work days (U. S. Department of Labor, 2002).

Typical employment stressors of a somatic nature which may contribute to these

injuries include circadian desynchronization, sleep deprivation, excessive overtime,

repetition, force, posture, and shift work (Putz-Anderson, 1988). Current responses to

CTD include engineering controls, work practice changes, and administrative measures.

However, Moon's (1996) epidemiological research suggested that additional

psychosocial factors might create job stress that contributes to the physical symptoms of

CTD. Moon noted that CTDs are found even in offices that have invested in

ergonomically sound equipment. Accordingly, Carayon, Smith, and Haims (1999)

recommend examining both physical ergonomic and psychosocial work factors

simultaneously. Psychosocial stressors include, but are not limited to, life event stresses,

excessive demands placed on workers (e.g., time pressures, several jobs, and

distractions), too-low demands (e.g., repetitive or monotonous tasks), failures in

performance, dissatisfaction with one's job, social isolation, interpersonal conflicts,

changes in habits, loss of relatives, conflict in decisions between several alternatives, and

uncertainty about future events (Hackfort & Schwenkmezger, 1993). In addition, Smith

and Carayon-Sainfort (1989) suggested that the psychological climate regarding

socialization, career and job security, the level of employee training, the availability of

assistance, and supervisory relations should be considered.

Such stressors impact individual employees as, among other ways, psycho-

physiological stress. Stress-induced reactivity of the psychophysiological system creates

several problems including (1) difficulty maintaining concentration on a task, (2)

increased energy mobilization above what is actually needed to perform the task, (3)

activation of energetical mechanisms unnecessary for task execution, and (4)

prolongation of activation which persists even after the task has been completed

(Caldwell et al., 1994). Thus, there are psychobiological mechanisms that make a

connection between job stress and CTDs/cognitive performance plausible and likely

(Blair, 1996; Sauter & Swanson, 1996; Smith & Carayon, 1996). These problems and

their consequences obviously have a negative impact on employee productivity.

Consequently, there is a tremendous need to develop an intervention tool to

reduce stress-related risks to health and performance. Epidemiological statistics indicate

that stress costs the U.S. at least $200 billion a year in worker absenteeism, lowered

productivity, increased compensation claims, higher health insurance rates, and higher

direct medical expenses (Hales & Hales, 1996). Sheffield Hallam University (2001)

serves as one example of how an institution attempts to reduce stress-related risks to

health and performance for its employees. The university's On-line Health and Safety

Manual states that it is the university's policy to ensure, so far as is reasonably

practicable, that no member of its staff is subjected at work to a level of stress which is

detrimental to health. Managers are directed to assess the risks arising from stress via

checklists and questionnaires.

Standardized job stress questionnaires are available, but most of these are

designed for epidemiological studies rather than for assessment of an individual worker

(Baker & Karasek, 2000). One such questionnaire, the Schedule of Recent Experience

(SRE) Life Changes Questionnaire, was developed by Rahe (1974). Forty two life

changes contained in the SRE were scaled as Life Change Units (LCU). Rahe's

retrospective studies found that people who recorded 150 LCU or less in the preceding

year reported good health for the succeeding year. If the person's yearly LCU sum

ranged between 150 and 300, an illness was reported during the following year for

approximately half of these people. For those registering more than 300 LCU per year,

an illness was recorded during the following year in 70 percent of the cases. A wide

variety of illnesses were reported; i.e., the LCU sum did not predict the type of ensuing

health change. Rahe also found a positive relationship between the LCU sum and the

severity of the subsequent illness.

The same somatic and psychosocial stressors cited earlier as contributing to injury

also contribute to decrements in cognitive performance. However, a distinction should be

made between mental load and stress. Mental load should not be equated with stress.

With mental load, extra energy is mobilized via mental effort that focuses attention and

improves performance efficiency (Caldwell et al., 1994). In contrast, under stressing

conditions, the increased activation is distracting and reduces efficiency. Mental load

mobilizes energy as required by the demands of the task for the period during which the

task is executed, while in a stressful state this activation persists and inhibits recovery.

Finally, mental effort is oriented towards the execution of the task, while stress is

oriented towards self-protection. The National Research Council (1995) points out that

the emphasis on quality introduced by Total Quality Management has highlighted gaps in

the human factors' scientific knowledge base. One important kind of quality flaw stems

from operator error. However, there is a paucity of studies on the causes and sources of

operator error, and even less is known about designing jobs and processes to reduce the

frequency and seriousness of errors. Wickens (1992), however, noted that stress and

human errors are tightly linked. Becoming aware of our errors induces stress. On the

other hand, high levels of stress cause more errors.

Psychophysiological stress is a state of disharmony or threatened homeostasis

(Baker & Karasek, 2000; Caldwell et al., 1994). Epidemiological studies have identified

high positive correlations between stress and the onset of illness (Rahe, 1974) and

accident involvement (Haakonson, 1980). Accordingly, numerous theories offer

descriptive models of the interaction of stressors and the psychology, physiology, and

behavior of the individual (Hancock & Warm, 1989). Regarding behavior, it is theorized

that people may, under psychological stress, develop specific physical symptoms to

legitimize their general psychological discomfort (Mechanic, 1961). However, an

objective means for correlating the level at which stress negatively impacts an

individual's health or cognitive performance has not evolved from theoretical work.

According to Haakonson (1980), this circumstance is partially due to the fact that it is

impossible to produce an exhaustive list of all of the stressors that may influence an

individual at any given time or in any given operation.

A brief theoretical overview of the stress/injury and stress/performance

relationships endorses Haakonson's (1980) view. However, the field of sport/exercise

psychology offers an interactional theoretical model for stress and injury (Figure 1-1).

Personality --- History of Stressors 4- Coping Resources

Hardiness Life Events General Coping
Locus of Control Daily Hassles Behaviors
Sense of Coherence Previous Injuries Social Support System
Competitive Trait Stress Management
Anxiety and Mental Skills
Achievement Motivation Medication, Self or
| Prescribed

Stress Response
Cognitive Appraisal of Physiological/ Attentional
Demands Aspects
Potentially Resources _______________ Increased General
Stressful Athletic t Consequences Muscle Tension -Injury
Situations Narrowing of Visual

Cognitive Restructuring 0 Relaxation Skills
Thought Stoppage Autogenics/Meditation
Confidence Training 0 Imagery/Mental Rehearsal
Fostering Realistic Expectations 0 Distraction Desensitization
Fostering Team Cohesiveness 0 Medication Modification

Figure 1-1. A model of stress and athletic injury
Note. From "Stress and Athletic Injury," by M. B. Andersen and J. M. Williams, 1988,
Journal of Sport and Exercise Psychology (Vol. 10, No. 3), p. 297. Copyright 1988 by
Human Kinetics Publishers, Inc. Reprinted with permission.


Similarly, human factors science offers a model ofjob stress and injury (Figure 1-

2). Most germane to the rationale and methodology proposed in the current experiments

are the interaction and feedback of the various components of these models. The

experimental designs of this study will mimic this stressor/stress reciprocal interaction.

V___ VT I
* the physical ergonomic risk factors Psychobiological Cardiovascular diseases
* the psychosocial risk factors Psychological Gastrointestinal disorders
SBehavioral Mental health disorders
Work-related musculoskeletal

V Perceptions
SHealth status
o etc...

Figure 1-2. Model of job stress
Note. From "Work Organization, Job Stress, and Work-Related Musculoskeletal
Disorders," by P. Carayon, M. J. Smith, and M. C. Haims, 1999, Human Factors (Vol.
41, No. 4), p. 646. Copyright 1999 by Human Factors and Ergonomics Society.
Reprinted with permission.

Given that theoretical efforts have failed to objectively evaluate stress i priori

complicates the practical application of what is known about stress. For example,

ergonomists are typically responsible for preventing, detecting, and eliminating the

workplace risk factors known to cause CTDs among employees, and to enhance

employee job performance through the reduction of stress. Yet, while stress is known to

mediate the etiology of CTDs as well as impair cognitive performance, practitioners (e.g.,

ergonomists and safety professionals) still have no reliable, scientifically based tool at

their disposal to objectively determine when to intervene to interrupt this causal


Objectively quantifying the level at which stress negatively impacts health and

cognitive performance is not a straightforward endeavor. The goal of quantifying the

stress level is not to minimize it, because up to a point, stress enhances health and

performance (Selye, 1976). However, the point at which stress becomes detrimental to

health and performance is obfuscated by individual differences that preclude making

generalizations about how much stress is too much. Objectively quantifying that point, in

spite of complications such as individual differences, is a necessary first step toward

developing and using effective interventions.

Return to the case history of the middle-aged woman discussed earlier.

Hypothesizing that there is a quantifiable relationship between injury/cognitive

performance and psychophysiological stress, mathematical models could be developed.

The employee could then be evaluated, and her results compared to the predictions of the

models. If her results were similar to others who had demonstrated a relationship

between psychphysiological stress and injury, proactive intervention would be justified.

Referral to the Employee Assistance Program for stress management training may well

have precluded the need for surgery.

This paper provides a comprehensive review of the stress/anxiety/arousal

literature pertaining to health and cognitive performance in an effort to both summarize

what is known and to point out theoretical and definitional shortcomings. Second, a

rationale based on psychophysiological theory and the Andersen and Williams (1988)

sport/exercise psychology stress/injury model is offered. The premise of this rationale is

that regression equations that included measures of both cognitive anxiety and

psychophysiological arousal as independent variables should have predictive value. The

underlying hypothesis is that there is a relationship between at least one of these

measures and the dependent variables of CTD symptomology and/or cognitive

performance decrement.

The absence of a unifying theory about what stress is and how it influences health

and cognitive performance poses a major problem when examining occupational stress.

Lacking a unifying theory, the study of stress entails exposure to a vast amount of

multidisciplinary literature with differing emphases on the physiological, cognitive, or

behavioral facets of the stress reaction. A selection of the literature follows that supports

a stress/health/performance interaction, as well as introduces the potential confounds that

need to be considered by any attempt to quantify the level at which stress might

negatively impact health and cognitive performance.

Stress and Health

The National Institute for Occupational Safety and Health (1999) defines job

stress as harmful physical and emotional responses that occur when the requirements of

the job do not match the capabilities, resources, or needs of the worker. Occupational

stress evolved from the concept that particular work activities produce worker

maladaptations that lead to ill health. This concept is grounded in many theories that

attempt to describe the interaction of environmental conditions, stress, and disease

(Aruin, 2002; Smith, 1987). For example, as early as 1914, Cannon proposed that

emotional states could influence hormone production and related physiological reactions

such as heart rate. He further stated that the relationship of emotional states, hormone

production, and physiological reactions is reciprocal. Thus, increased hormone response

due to emotional distress could lead to greater emotional distress. Today, researchers

accept that reciprocal relationships exist but are left without a generalizable means to

interpret the hormone response or the emotional distress.

Further, Cannon's (1914) theory illustrated the complication of discriminating

between the stress and the stressor. As noted by Martens, Burton, Vealey, Bump, and

Smith (1990), stress has historically been one of the most ambiguous psychological

constructs in the behavioral sciences. Stress has been defined as a stimulus, as an

intervening variable, and as a response variable by different researchers. As a stimulus

variable, stress is a precipitator; as an intervening variable, a mediator; and as a response

variable, a behavior. The reciprocity of the hormone/distress relationship further clouds

the distinction between stress and stressor. However, in Selye's (1976) framework, stress

is the behavior while the stressor is the precipitator.

Going beyond Cannon's predictions, Selye (1976) advanced the endocrinology of

stress. He pioneered the medical model that described how noxious stimuli produced a

specific syndrome of disease. After years of extensive research, Selye proposed the

General Adaptation Syndrome of stress. The evidence in favor of the General Adaptation

Syndrome indicated that the primary adverse health effects of stress are caused through

overextension of energy mobilization and disease fighting systems of the body. Possibly

Selye's most usable contributions involved demonstrating that the General Adaptation

Syndrome was a nonspecific response, and that the stress syndrome developed in three

stages (1) alarm reaction, (2) stage of resistance, and (3) stage of exhaustion. During the

alarm reaction, general resistance to the particular stressor by which the General

Adaptation Syndrome had been elicited falls below normal. Then, as adaptation is

acquired in the stage of resistance, the capacity to resist rises above normal, but

eventually, in the stage of exhaustion, resistance again drops below normal. Selye's

contributions, nonetheless, again complicated the endeavor to objectively assess the

impact of stress because he demonstrated that both eustress (good stress) and distress

(bad stress) provoked the same precursors of the syndrome.

Thus, while both Cannon (1914) and Selye (1976) established basic relationships

between environmental stressors and bodily reactions, neither researcher thoroughly

addressed the perceptual and psychological processes (e.g., cognitive appraisal,

hardiness, and motivation) that mediate the stress response. These shortcomings left

individual differences unexplained (Smith, 1987). McGrath (1976) addressed the need

for considering individual differences concerning the psychological conditions of stress.

He argued that stress occurred when an environmental situation is perceived as presenting

a demand that threatened to exceed an individual's capabilities and inherent resources.

Accordingly, identical conditions can affect individuals differently depending on the

individual's varying resources. The intensity with which people experience (or

anticipate) the losses that would ensue should they fail to meet a challenge is another

complication. Thus, ignorance or special sensitivity to situational conditions can

preclude or exacerbate the perception of psychological threat (National Research

Council, 1995).

Consistent with Cannon and Selye, Lazarus (1977) studied the hormonal

responses to environmental stimuli. He addressed some of the perceptual and

psychological aspects of the stress response and proposed that cognitive appraisal and

resultant behavior define the hormone responses to environmental stimuli. Lazarus

surmised that personality is influenced by psychosocial factors such as beliefs, motives,

fears, and desires which interact with environmental conditions to shape responses. The

quality or intensity of the emotion and its resultant behavior depend on the cognitive

appraisal of the present or anticipated significance of the interaction with the environment

in terms of the person's well-being (Smith, 1987).

A different approach at incorporating the perceptual and psychological

contributions to the stress response was proposed by French (1963) and Harrison (1978).

They argued in favor of a systems approach toward understanding the relationship

between stress and health. The systems approach focuses on the degree of fit between a

person and his/her environment. Such an approach involves an objective

person/environment and a subjective person/environment. The objective person refers to

the true physical, mental, and emotional attributes of the person, whereas the subjective

person is the individual's perception of self and his/her abilities. Similarly, the objective

environment refers to the physical and social circumstances as they actually are, while the

subjective environment is represented through the individual's perception. Greater

discrepancies between the objective and the subjective are hypothesized to generate more


Siegrist (1996) offered an occupational stress model that emphasized an

imbalance between the effort required for a job and the rewards provided by the job as a

potential stressor. In this model, job stress arises both from the immediate conditions of

work and from the broader context of the role of work as a link to important emotional

and motivational needs in a person's life.

Overall, the conclusions of McGrath (1976), Lazarus (1977), French (1963),

Harrison (1978) and Siegrist (1996) contributed to the understanding of the interaction

between stress and health by addressing the perceptual and psychological aspects of the

stress response. A general conclusion derived from these theories is that if working

conditions are perceived as unpleasant for a prolonged time, the resultant stress reactions

influence the development of diseases (possibly including cumulative trauma disorders

(Baker & Karasek, 2000)). That is, the worker plays a significant role in exacerbating or

mitigating the potential for disease and/or accidents (Smith, 1987). Further, the above

theories illustrated the number of factors potentially demanding consideration in any

attempt to predict a level at which stress would negatively impact health. However,

toward developing a regression model to correlate the relationship between

psychophysiological stress and health, at this point, theory has offered only a starting


YHealth = 3Po+31(Psychophysiological Stress)+c

Stress and Performance

In addition to the stress/health issues, a second ergonomic concern is the

relationship between stress and cognitive performance. The emphasis of the performance

and stress/anxiety/arousal literature differs from that of the health models reviewed

above. For example, Welford (1973) emphasized the relationship between demand and

capacity. He argued that people function best under conditions where a moderate

demand is made upon them. Performance is less than optimum if the demand is either

too high or too low. Stress varies with the environmental and social conditions affecting

demand, as well as with the native endowment, training, and bodily conditions affecting

capacity. Demand and capacity are conceived in terms of both the physical (e.g.,

strength) and information processing (e.g., cognition). Moreover, for stress to occur, the

person concerned must regard the consequences of failure to meet the demand as

important. Consequently, Welford linked the stress response to motivation.

Two dominant theories of motivation (e.g., Need Achievement Theory,

Attribution Theory) agree on the basic principle that action arises in an effort to improve

conditions that are less than optimum. Stress arises whenever there is a departure from

optimal conditions which the organism is unable, or not easily able, to correct. However,

defining optimal conditions is a difficult task. Welford (1973) stated that at least three

principles contribute to optimal conditions (1) people tend to avoid extreme stimulation

and to seek moderate levels, (2) people prefer stimulation that has a moderate level of

predictability, and (3) people prefer a moderate degree of conflict, either of cognitive data

or of potential action. Signs or symbols of departure from optimal conditions may also

trigger an action. In this way, anticipatory or imaginative action is made possible.

Furthermore, several motives can operate simultaneously on different time scales to

determine any one behavior. Thus, people may cope successfully with their immediate

tasks, and yet suffer stress because they feel unable to attain their long-term goals.

Welford's (1973) thesis is based on the premise that people strive to secure an

optimal level of arousal. The broad relationship of stress on arousal and performance

was described by the Inverted-U Hypothesis. Arousal was assumed to be a unitary

construct embodying both the psychological and the physiological response systems

(Hardy & Parfitt, 1991). The Inverted-U Hypothesis posits that the organism is

insensitive and inert if the level of arousal is too low, or tense and disorganized if the

arousal level is too high (Landers, 1980; Welford, 1973). This model is attractive;

however it does not advance practical understanding because of the ambiguous

operational definitions of the optimal conditions and explanations about why optimal

conditions differ from one task to another.

Hardy and Parfitt (1991) further noted that there were difficulties with the basic

constructs of the Inverted-U Hypothesis. Indeed, the Inverted-U Hypothesis has been

described as a relationship between stress and performance, arousal and performance, and

anxiety and performance. Whichever of these constructs is adopted, anecdotal evidence

suggests that the Inverted-U should not have the symmetric shape usually presented.

Hardy and Parfitt argued that when performers go "over the top," performance appears to

drop dramatically rather than gradually; and once this has happened, it is very difficult to

achieve even a mediocre level of performance. This consequence suggests that small

reductions in anxiety do not really make any difference to performance once this peak

level has been reached. However, the Inverted-U Hypothesis implies that performance

should return to its optimal level with such an intervention. Such construct difficulties

are not surprising when one recalls that the Inverted-U Hypothesis originated from a

study of habit strength formation in mice under varying punishment frequencies.

Returning to the arousal/performance relationship, based on current views,

arousal is a multidimensional construct that refers to an energizing function of the mind

and body. Arousal includes a general physiological response, but behavioral responses

and cognitive processes are also indicators of arousal (Zaichkowsky & Takenaka, 1993).

More recent theories have mimicked the Inverted-U Hypothesis regarding the

impact of stress on arousal and performance. Examples include Hanin's (1989) Zone of

Optimal Functioning, and Martens' (1987) Optimal Energy Zone. Hanin added insight

regarding interindividual differences (i.e., some individuals will perform best when

highly aroused, others when arousal is low, and still others when it is moderate). Martens

divided the Inverted-U into zones, thereby depicting the lack of precision that is

associated with the measurement of arousal. However, Zaichkowsky and Takenaka

(1993) pointed out that both the Zone of Optimal Functioning and the Optimal Energy

Zone were merely extensions of the Inverted-U Hypothesis, with different labels on the x

and y axes.

Dissatisfied with the one-dimensional nature of the Inverted-U Hypothesis and its

clones, Hardy (1990) proposed a catastrophe model of anxiety and performance. An

attractive feature of this model is the simulation of the interactive effects of anxiety and

physiological arousal on performance. His model proposed that cognitive anxiety acts as

a splitting factor that determines whether the effects of physiological arousal are small

and smooth, large and catastrophic, or somewhere in between these two extremes. In

contrast to Martens' multidimensional anxiety theory (Martens et al., 1990), Hardy's

catastrophe model took exception. Specifically, the multidimensional anxiety theory

argues that at least two different components can be distinguished in the anxiety

response: a cognitive component associated with fear about the consequences of failure,

and a somatic component reflecting perceptions of the physiological response to

psychological stress. Hardy questioned multidimensional anxiety theory because it

attempts to explain a three-dimensional relationship among cognitive anxiety, somatic

anxiety, and performance, in terms of a series of two-dimensional relationships. Martens

only makes predictions about the separate relationships between cognitive anxiety and

performance, and somatic anxiety and performance when the theory implies an

interaction model of at least three dimensions. Appropriately, Hardy's cusp catastrophe

model is three-dimensional.

Hardy (1990) further disagreed with Martens et al. (1990) regarding the two-

predictor variables chosen. Rather than Martens' cognitive and somatic components of

anxiety, Hardy investigated the interaction of cognitive anxiety and arousal on

performance. Hardy predicted that only under conditions of high cognitive anxiety

would increased physiological arousal be detrimental to performance. However, in spite

of Hardy's theoretical refinements and supporting empirical data, his experimental

methodology limits the theory's applicability to the stated purpose of this paper. Hardy

manipulated arousal (heartbeat) with exercise, a methodology that would compromise the

ecological setting of the workplace.

The above stress/performance theories were consistent with the stress/health

theories discussed earlier in that they offered primarily qualitative descriptions of the

impact of stress/anxiety/arousal. As with the stress/health literature, theory offers only a

conceptual regression model:

YPerformance = Po+Pi(Psychophysiological Stress)+e

Definitional Shortcomings

Beyond the absence of a unifying theory about what stress is and how it affects

health and cognitive performance, the stress/anxiety/arousal literature offers definitional

complications that an applied intervention tool must address. Because anxiety results in

increased central and autonomic nervous system activity, anxiety has unfortunately been

confused with arousal. Although anxiety results in increased physiological activity,

anxiety refers to states that provide feelings of discomfort and worry (Zaichkowsky &

Takenaka, 1993). The two constructs (anxiety and arousal) are not the same.

Additionally, cognitive anxiety (i.e., worry) is differentiated from somatic anxiety (i.e.,

positive or negative perceptions regarding physiological arousal) (Rotella & Lemrner,

1993). Further, Hardy, Jones, and Gould (1996) distinguished between somatic anxiety

and physiological arousal. This distinction was made on the premise that physiological

arousal may influence performance either directly or indirectly, while somatic anxiety

only affects performance indirectly.

Another potential confound is the distinction between state and trait anxiety.

Trait anxiety is defined as an acquired disposition causing an individual to subjectively

perceive a wide range of objectively not very dangerous circumstances as threatening.

State anxiety reactions can be described as subjective, consciously perceived feelings of

inadequacy and tension accompanied by an increased arousal in the autonomic nervous

system (Hackfort & Schwenkmezger, 1993). In spite of these complications, based on

the opinions of the authors cited in this section, it may be appropriate to combine and

refine the two aforementioned regression models as follows:

YHealth and Performance = P3o+13p (Cognitive Trait Anxiety)+p2(State Physiological Arousal)+s

Correlating the Impact of Stress

Given the theoretical and definitional considerations cited earlier, an assessment

instrument to predict workplace CTD injuries should quantify the outcome of the

interaction of environmental input, the stress response, and health, while circumventing

the reciprocity of stress/stressor problem encountered by Cannon (1914). In addition, the

intervention tool should accommodate Selye's (1976) finding that distress and eustress

both present the same precursors of the stress response. Further, the tool must not be

confounded by the measurement similarities of Selye's alarm and exhaustion stages of

the General Adaptation Syndrome. Similarly, the tool must accommodate the temporally

changing perceptual and psychological processes that contribute to the intra- and inter-

individual differences theorized by McGrath (1976), Lazarus (1977) and Siegrist (1996),

or by the systems approaches of French (1963) and Harrison (1978).

An assessment instrument to predict cognitive performance decrement would

have to quantify the outcome of the interaction of the changing demand, capacity and

motivation cited by Welford (1973), whether the demand was physical, or psychosocial,

and from the past, present, or future. It would have to provide predictive powers whether

the stress/performance interaction was one-dimensional as envisioned by the Inverted-U

Hypothesis and its clones, or multidimensional as theorized by Hardy (1990) and Martens

et al. (1990).

Obviously, the difficulties that an applied intervention method must address are

complex. Haakonson (1980) stated that an all-encompassing stress/health/performance

model must be developed to predict the impact on health and performance of

environmental, sociological, psychological and physiological stressors.

The resolution of this conceptual proposal involves three main points. First, in

terms of the stress theories reviewed, deviant outputs of the nervous system would take

the form of some combination of cognitive anxiety, somatic anxiety, and arousal.

Cognitive and somatic anxiety can be measured via surveys. Arousal can be measured by

psychophysiological instrumentation. Second, the reciprocal interaction of stressor and

stress noted earlier could be used to an advantage in the experimental design.

Specifically, rather than manipulate the stressor per se, the participant's attention could

simply be directed toward matters of potential cognitive anxiety and somatic anxiety,

while measuring arousal liability under these conditions. Third, correlation of the anxiety

and arousal measurements with CTD symptoms and cognitive performance decrement

(CPD), in the form of two regression models, would provide the proactive intervention

tools sought as the purpose of this proposal.

Given that the envisioned predictive assessment instruments would be used in the

applied setting of the workplace, the manipulation of arousal as used in the research

supporting the catastrophe model (Hardy et al., 1996) is undesirable. For the same

reason, minimizing measurement modalities and instrumentation complexity is desirable.

This seems to be an attainable goal in that Caldwell et al. (1994) observed the

measurement of end-organ physiological responses such as heart rate, blood pressure,

muscle activity, and skin conductance can determine overall sympathetic activation.

Such simplified measurements are attractive from a practical perspective and are in sharp

contrast to other possibilities (Carayon et al., 1999). Selecting the independent variables

for the two following conceptual regression models is therefore the next step in

developing predictive intervention tools:

YCTD = 3Po+Pi(Cognitive Trait Anxiety)+p32(State Physiological Arousal)+-

YCPD = Po+P3i (Cognitive Trait Anxiety)+32(State Physiological Arousal)+e

Theoretical Bases for the Proposed Research

Andersen and Williams (1988) proposed an interactional theoretical model of

stress and athletic injury that could be adapted as the all-encompassing stress/health/

performance model called for by Haakonson (1980). This interactional theoretical model

provides details about how various factors influence stress, health, and performance.

Included are the cognitive, physiological, attentional, behavioral, intrapersonal, social,

and stress history variables that may influence injury occurrence. This model assumes

that two of the basic mechanisms behind the stress/injury relationship are increases in

general muscle tension and deficits in attention during stress. Andersen and Williams

hypothesized that individuals dealing with excessive stress, who have personality traits

that tend to exacerbate the stress response, and few coping resources, will exhibit greater

muscle tension and attentional changes, and thus be at greater risk of injury. Andersen

and Williams cite the need for research to address whether certain individuals, under

stress, exhibit greater increases in generalized muscle tension than others do.

Muscle tension caused by unwanted simultaneous contraction of agonist and

antagonist muscle groups (often called bracing, splinting, or guarding) is a common

response to stressors. With increased levels of norepinephrine, the tension in muscles has

the potential to be greater, as does the extent of recruitment of the muscle fibers in

performing an activity (Westgaard, 1996). Additionally, however, McNulty, Gevirtz,

Hubbard, and Berkoff (1994) suggest another mechanism by which emotional factors

influence muscle tension. Under a stressful condition (mental arithmetic), participants

demonstrated increased trigger point needle electromyographic activity in the myofascia

of the trapezius. McNulty et al. attribute this activity to nonvascular sympathetic nervous

system innervation to the intrafusal fibers of the muscle spindles. The end result of this

combination of muscle tensing mechanisms is a greater risk for incurring injuries such as

sprains, strains, and other musculoskeletal injuries.

Concerning attentional changes, Andersen and Williams (1988) argued that under

stress the individual is distracted and attends to stimuli not relevant to the task at hand.

They concluded that their model provides a broad theoretical foundation for future

investigations into the prediction and prevention of injury and the many variables to be

considered in the stress/injury relationship. Because their theory reflects multiple

interactions, the Andersen and Williams model is an ideal basis for the rationale of the

current research.

Andersen and Williams' (1988) theoretical model is multicomponent in that the

components span three major areas: personality factors, history of stressors, and coping

resources. While there has been no research to test the complete model, experiments

directed at evaluating the model's three major areas have been generally supportive. In a

review of the studies that examined the relationship of stress history to sport injury risk,

Williams and Roepke (1993) indicated that 18 of the 20 studies they reviewed found a

positive relationship between high life stress and injury. Regarding personality factors,

the six personality variables proposed in the stress/injury model were merely suggestions

for research rather than an exhaustive list of potential factors (Williams & Andersen,

1998). Mixed results occurred when researchers examined three of the personality

variables, i.e., locus of control, trait anxiety, and sensation seeking. Also, the stress-

injury model includes coping resources consisting of social support, stress management

coping skills, and coping behaviors such as good sleeping patterns and nutritional habits.

According to Williams and Andersen, the preponderance of evidence clearly supports the

conclusion that coping resources directly affect injury outcome, moderate the life stress-

injury relationship, or do both. There has been no subsequent research, based on the

stress/injury model, addressing three of the six proposed personality variables: hardiness,

sense of coherence, and achievement motivation.

While the Andersen and Williams (1988) model provides a basis for the rationale

of this proposal for investigating stress, psychophysiological theory advocates the

investigation of psychophysiological responses to stress beyond only muscle tension. For

example, Caldwell et al. (1994) proposed that the investigation of physiological variables

within a psychophysiological framework is best accomplished by detecting deviant

physiological signals that indicate a less than optimal state, or tendencies toward a

disrupted state indicating stress. Consistent with Andersen and Williams, they stated that

such information might be used to optimize performance, to prevent reduced well being

or health, or to prevent accidents. However, Caldwell et al. emphasized that it is

important to collect multiple measures due to the large variance among individuals in

terms of how they respond to the environment and/or internal events. For example, some

people respond with reactive changes in the cardiovascular system, while others respond

primarily with changes in the central nervous system. Similarly, Landers (1980) argued

that stress/anxiety/arousal is a multidimensional phenomenon, and that multimethod

procedures should be used to examine it. Finally, the individual differences in

physiological reactivity alluded to by Caldwell et al. include the ramifications of the Law

of Initial Values, which will be discussed next. Lacey (1956) revealed an intriguing way

to circumvent this obstacle via the Autonomic Lability Score.

Based on extensive experimental investigations, Lacey (1956) concluded that

autonomic response was dependent on the prestress level of autonomic activity; the

autonomic response is negatively correlated with the initial value. Accordingly, any


induced excitation of an autonomically innervated structure instantly initiates a series of

changes that serve to nullify the disturbance. The recorded autonomic response is a

function both of the induced magnitude of autonomic activation and of the promptness

and vigor of secondarily induced autonomic changes that serve to restrain the effects of

the initial disturbance. As the prestimulus level of functioning increases, there is a

disproportionately greater homeostatic restraint, both in increased magnitude and

decreased latency and, as the magnitude of induced activation increases, there is a

disproportionately greater increment of counterreaction. Thus, stress measurements

reflect a phenomenon of physiological homeostasis and must be dealt with as such. This

Law of Initial Values is a concern because investigations in psychophysiology require

estimates of the relative standings of individuals within a group with respect to the

amount of reaction shown by a given effector system. Similarly, when reactivity in

different effector systems are to be compared, there is a need for a metric of common

units. For these reasons, Lacey proposed using the regression of the stress level on the

initial level. He called the resultant score the intersubject Autonomic Lability Score. The

equation for calculating resultant score follows:

50 + 10 [yi x,iry]/ (1 ry)1/2

* 50 + 10 are constants to eliminate negative numbers and to permit results to be

expressed in terms of frequency distributions with a means of 50 and standard

deviations of 10.

* yz,i is the measure of response expressed as standard deviations from a mean.

* Xz,i is the prestimulus level expressed as standard deviations from a mean.

* rxy is the product-moment correlation.

(1- r2xy)"2 is the error of estimate about the line of regression when x andy are

expressed in units of standard deviation.

Repetitive solution of the equation is necessary to obtain response measures for each


In addition, Lacey's (1956) finding shows that individuals tend to reproduce

response patterns (i.e., stress levels of autonomic arousal show substantial

reproducibility). He called this tendency the stereotypy of response pattern. Lacey

concluded that the stress levels and Autonomic Lability Scores were positively correlated

and that the Autonomic Lability Score represented the stress level, corrected for baseline

level. Thus, the evaluation of stress level is important from the viewpoint of

predisposition to the behavioral consequences of autonomic responses.

Physiological Bases for the Proposed Research

Given that Lacey's (1956) Autonomic Lability Score is based on measurements of

autonomic activity, this section addresses the physiological bases for the conceptual

proposal. Guyton (1976) provided an overview of physiological arousal. The autonomic

nervous system (ANS) normally operates at the subconscious level and controls many

bodily functions, including the action of the heart, arterial pressure, sweating, body

temperature, and secretion by different glands. The nervous system, in general, regulates

rapid muscular and secretary activity while the hormonal system regulates the slowly

reacting metabolic functions. Autonomic impulses are transmitted to the body through

two major subdivisions, the sympathetic and parasympathetic systems. The sympathetic

system is strongly activated in many emotional states. This activation is often called the

fight or flight response. One purpose of the sympathetic system is to provide extra

activation of the body in states of stress. This activation is therefore also called the

sympathetic stress reaction.

Hormonal stress response theories focus on two mechanisms: the adrenal

medullary response involving epinephrine and norepinephrine, and the adrenal cortical

response involving cortisol. The adrenal medullary response is activated by time

pressures, increased job demands, and a range of workplace social situations. Long-term

activation can lead to difficulties with relaxation and a state of chronic overarousal. The

adrenal cortical response is a response of defeat and withdrawal. Adrenal cortical output

occurs in a situation where the person faces stress over which he or she has little control,

i.e., a decreased decision latitude (Baker & Karasek, 2000; Berggren, Hane, & Ekberg,


According to Guyton (1976), sympathetic discharge increases the capability for

muscle activity by increasing the rate of cellular metabolism, increasing blood glucose

concentration, constricting the adrenergic alpha vasculature while dilating the cholinergic

and adrenergic beta vasculature, and increasing glycolysis (glycogenolysis) in muscle.

Muscle contraction results from the more or less synchronous contraction of the many

muscle fibers that comprise a muscle. Each muscle fiber is actuated by an electrical

signal carried by the motor unit. This electrical activity can be sensed via surface

electromyography. The frontalis muscle of the forehead is typically the site chosen to

monitor for generalized muscle tension (Schwartz, 1987).

The sweat glands secrete large quantities of sweat when the sympathetic nerves

are stimulated. Since sweat contains salts that make it electrically conductive, sweaty

skin is more conductive to electricity than dry skin. Therefore, skin conductance

corresponds well to sweat gland activity. Increases in skin conductance are a sensitive

index to behavioral changes related to boredom, fatigue, and monotony (Baker &

Karasek, 2000). This electrodermal response is measured by a device that applies a very

small electrical pressure to the skin and measures the amount of electrical current that the

skin will allow to pass ( Schwartz, 1987).

Except for the capillaries, sphincters, and most of the metaarterioles, all of the

blood vessels of the body are supplied with sympathetic nerve fibers. Sympathetic tone

normally keeps almost all the blood vessels of the body constricted to approximately half

their maximum diameter. When the sympathetic vasoconstrictor nerve fibers are

maximally stimulated, they decrease blood flow through the muscles to about one fourth

normal. Sympathetic stimulation also causes mild vasoconstriction of the large vessels of

the cerebral vasculature. The majority of blood vessels, especially those of the skin of

the limbs are constricted by sympathetic stimulation (Guyton, 1976). An indirect way of

monitoring peripheral vasoconstriction is to measure digital temperature. Since

constricted vessels pass less warm blood than dilated vessels, the surrounding

temperature tends to warm and cool as vascular diameter increases and decreases.

When applied in a therapeutic setting, biofeedback training makes use of the

phenomenon cited by Caldwell et al. (1994) regarding determination of overall

sympathetic activation via observation of end-organ physiological responses. In fact, the

target behavior of biofeedback training, in operant conditioning terms, is to decrease a

behavioral excess, (i.e., to decrease the neuromuscular and/or the sympathetic nervous

system activity of the monitored end organs). However, biofeedback therapy is reactive.

The biofeedback coach uses psychophysiological measurements to gauge the client's

success at decreasing chronic behavioral excesses that have caused or contributed to a

clinical diagnosis. This study proposes that the same psychophysiological measurements

of neuromuscular activity and of sympathetic arousal be used proactively. Accordingly,

the proposed regression models are expanded:

YCTD = po30+ 1(Anxiety)+p32(EMG ALS)+p33(TEM ALS)+34(EDR ALS)+s

YCPD = Po30+P31(Anxiety)+p2(EMG ALS)+p3(TEM ALS)+04(EDR ALS)+e

The preceding paragraphs illustrate that the technology exists to measure the

sympathetic stress reaction noninvasively with instrumentation suitable for the workplace

applied settings. There are several physiological bases by which the sympathetic stress

reaction could contribute to the etiologies of cumulative trauma disorders or performance

decrement. A review of these models is offered next.

Cumulative trauma disorder symptoms. Static work occurs when a muscle

remains in a contracted state for an extended period. Static work can be caused by

sustained awkward posture or by high strength demands. When a muscle contracts, its

blood vessels are compressed. Vascular resistance increases with the level of muscle

tension and the blood supply to the working muscle decreases. If the muscle cannot relax

periodically, the demand for metabolic nutrients exceeds the supply and metabolic wastes

accumulate. Blood lactate levels have been measured to document residual fatigue in

muscles 24 hours after sustained handgrip exertions at only 10% maximum strength

(Keyserling, 2000). Baker and Karasek (2000) cite the large number of studies that have

demonstrated an association between job stress and musculoskeletal problems, especially

upper extremity disorders among office workers. Typically, the mechanisms blamed for

these associations are that stress-related tension causes increased static loading of the

muscles, or that workers under stress alter their behaviors to increase musculoskeletal

strain. The normal fight or flight response would not produce a CTD. However, with

chronic bracing, or a chronic abnormal autonomic vasoconstriction reaction, ischemic

conditions could become the norm and deprive the tissues of sufficient recovery.

One might not intuitively conclude that autonomic vasoconstriction in skeletal

muscle would be part of the fight or flight response. However, the sympathetic stress

reaction is only part of the fight or flight response. The purpose of the sympathetic

vasoconstriction in skeletal muscle is twofold. First, sympathetic vasoconstriction is a

reflex compensation for any blood loss that might ensue, thus ensuring maintenance of

arterial pressure and the prevention of shock. Second, sympathetic vasoconstriction acts

to reduce circulation to all skeletal muscle in preparation for making blood available to

the skeletal muscles involved in the fight or flight. Thus, sympathetic vasoconstriction

accommodates the next part of the response the vessels in the active muscles dilate as a

result of vasodilator effects in the muscles themselves. For instance, when a person

performs whole-body exercise such as running, there is extreme vasodilation occurring in

large masses of muscle. On the other hand, when a person performs exercise under very

tense conditions but uses only a few muscles, the sympathetic response still occurs

throughout the body but vasodilation occurs in only a few muscles (Guyton, 1976).

Consequently, jobs performed with a static posture (i.e., computer operators) could have

widespread ischemia throughout their skeletal muscles if they are under stress, and the

condition would be worsened if they also chronically brace.

Alternatively, cumulative trauma disorders could be viewed in the light of a Selye

model. In his classic book, The Stress of Life (1976), Selye proposed a Local Adaptation

Syndrome, along with his General Adaptation Syndrome. Selye argued that perhaps one

of the most important indicators of stress was its effect on inflammation. Normally,

stress applied to a limited part of the body causes inflammation. However, the ability of

the body part to respond locally with inflammation is impaired when the whole body is

under stress. In response to an alarm reaction, inflammation will be produced in the

involved body part during the stage of resistance. However, if the stress is chronic and

the stage of exhaustion ensues, the body's protective mechanisms are broken down. Thus

the General Adaptation Syndrome and the Local Adaptation Syndrome are

interdependent. General stress can influence local stress reactions. In this instance the

ischemia is relative. While circulation to the damaged tissue may be normal, the

increased circulation consequent to inflammation is needed to provide sufficient oxygen

and nutrients. Without the inflammatory response, the outcome is the same as in the

previous model; the tissue is denied the chance to recuperate.

In light of the interdependence of the General and Local Adaptation Syndromes,

what is the cause of cumulative trauma disorder: repetitive strain, muscle bracing,

sympathetic vasoconstriction, or fatigue? The answer is probably all of the above. Such

a compounded etiology of injury illustrates the stress/stressor reciprocity problem that

must be circumvented by the attempt to quantify the impact of stress in the development

of cumulative trauma disorders. However, such a compounded etiology of injury is not

peculiar to CTDs. Selye (1976) addressed this issue and coined the term "pluricausal

disease." His central notion was the idea that a microbe is in or around everyone all the

time, yet causes no disease until exposed to stress. When questioned about whether the

cause of the illness was the microbe or the stress, Selye believed that both are, and

equally so. Intuitively, this approach makes a great deal of sense. Certainly there are

jobs that expose the employee to intensities of repetitive strain of sufficient duration to

solely cause cumulative trauma disorder. At the other end of the spectrum are the jobs

where the intensity and duration of repetitive strain would injure only an employee with

adaptive and recuperative capabilities compromised by stress. Hadler (1990, 1992)

believes that stress is the primary cause of the symptomology associated with many upper

extremity CTDs. While ergonomists are trained to prevent the injuries posed by the more

traditional risk factors such as repetition, force, and posture, they are not equipped to

judge when the level of stress is sufficient to be a causal factor of injury. There might

also be a vicious circle of stress leading to CTDs and vice versa. For instance, someone

who is stressed might grip a tool too tightly. In turn, the resulting symptoms could

increase anxiety (Carayon et al., 1999). Generally, maladaptive coping behaviors have

been related to poor overall health, less energy, and greater general fatigue. This could

make people more susceptible to injury and lead to a diminished capacity to work.

According to Fitzgerald (1992), both conditions increase the potential for CTDs.

Cognitive performance decrement. While not presented in the terms of Hardy's

(1990) catastrophe model, the experiment needed to evaluate this conceptual proposal

nonetheless could be couched in his terms. As stated earlier, Hardy predicted that under

conditions of high cognitive anxiety, physiological arousal would be detrimental to

performance. The experiment needed to support the proposed concept will also have to

measure cognitive anxiety, physiological arousal, and performance. Hardy's model

focuses on what happens to performance under conditions of high cognitive anxiety and

physiological arousal, without dwelling on how these changes come about. Accordingly,

he does not address any proposed physiological chain of events. Psychological

explanations, such as interference between the task at hand and preoccupation with

personal concerns, are offered instead. Such arguments would be equally applicable to

similar outcomes for the research proposed by this paper.

A more physiologically oriented model of the stress reaction's impact on

cognitive performance was briefly mentioned earlier. Guyton's (1976) account of the

stress reaction included that sympathetic stimulation causes mild vasoconstriction of the

large vessels of the cerebral vasculature, although he does not define mild. However, if

mild were interpreted as a reduction in circulation that could be functionally equated to a

drop in the blood oxygen saturation to a resultant hypoxia (e.g., 87%), this would

certainly have an effect on performance. Signs of such a mild hypoxic state are

negligible under resting conditions, but visual and cognitive performance are nonetheless

sufficiently degraded to require the use of supplemental oxygen to regain sea level

performance parameters.

The theoretical basis for the proposed experiments is that CTDs and performance

decrement are significantly affected by reduced circulation to the involved tissues,

mediated by the vasoconstriction and increased muscle tension of a chronic stress

reaction. Testing this theory in an ergonomics framework is a novel approach.

In summary, despite the diversity of stress/anxiety/arousal theory and the

definitional complications presented, there is sufficient information about stress to enable

quantification and prediction of its impact on health and performance. The rationale of

this proposal is based on two primary premises. First, the study of stress is currently

mired in semantics. This paper proposes a possible solution to advance the study of

stress is to circumvent the theoretical and definitional issues by focusing on the cognitive

and physiological end products. In light of the Andersen and Williams (1988) model of

stress and injury, no matter the environmental inputs or the perceptual and psychological

mediations, the individual's cognitive and autonomic responses are end products, and the

technology currently exists to measure these outputs. The cognitive output can be

measured with a survey score. The ideal survey instrument would have been designed

for the general population and produce an overall stress score.

The second premise is that investigating autonomic responses to stressors, outside

the Lacey (1956) context (as an Autonomic Lability Score representing threatened

homeostasis), merely measures arousal rather than psychophysiological stress.

Conversely, this manuscript holds that when the autonomic responses are measured with

biofeedback instrumentation, and interpreted as Autonomic Lability Scores, one is

measuring stress. Based on the above theoretical and physiological bases, expected

findings include linear relationships of one or more measures of cognitive anxiety and/or

sympathetic stress reactions with CTD symptoms and cognitive performance. As the

independent variables of regression models, cognitive and physiological scores together

would allow for the quantification (via the coefficients of determination) of the impact of

stress in the etiologies of cumulative trauma disorders and cognitive performance



Research Design

The two studies were conducted simultaneously. The first study investigated the

relationship between CTD symptoms and psychophysiological stress measurements. The

second study examined the relationship between cognitive performance decrement and

psychophysiological stress measurements. Both experiments were observational studies.

The first experiment consisted of a 2-min baseline arousal measurement period,

and two treatment conditions of varying durations (1) a cognitive anxiety inventory, and

(2) a symptom survey. The second experiment consisted of the same 2-min baseline

arousal measurement period of the first experiment, plus a 10-min cognitive performance

test. To avoid any possible order effect, the three treatment conditions were presented in

random order. Physiological arousal measurements were collected for the first 10 min of

each of the three treatment conditions.

Measurement Instruments

Cumulative Trauma Disorder Symptoms

The Chest and Upper Extremity, and Back and Lower Extremity Symptoms

Surveys (Appendices B and C) provided discrete, self-reported scores of musculoskeletal

discomfort. These surveys were developed by the Occupational Safety and Health

Administration (1995), and modified to include a symptom rating scale (question 13 on

both forms) validated by Price, Bush, Long, and Harkins (1994). Survey questions apply

to both work-related and nonwork-related symptoms. Because this study is concerned

with CTDs, work-related symptoms reported as having been of "gradual" onset were of

primary concern. However, because both work and nonwork-related symptoms, of both

sudden and gradual onset, could affect both anxiety and arousal, scores for all

musculoskeletal symptoms were sought. This approach is consistent with the Andersen

and Williams' (1988) model wherein they theorize that bracing is relevant to the

reporting of acute athletic injuries. The reciprocal interaction of stress and stressor was

exploited by asking the participants to report their discomfort "now," and simultaneously

recording psychophysiological arousal measurements while the participants were exposed

to the potentially state anxiety producing act of completing the symptom survey forms.

Cognitive Performance Decrement

A modified Stroop Color-Word Interference Test (MacLeod, 1991; Muhonen,

2003) was administered as the cognitive performance task. Performance of the task

produced discrete response times, which were interpreted as the measure for cognitive

performance. Siska (2002) recommends the Stroop Color-Word Interference Test for

studying the influence ofpsychosocial stress. Completing the task potentially induced

state anxiety; thus measurements of psychophysiological arousal were simultaneously

recorded. Elliott, Bankart, and Light (as cited in Kahneman, 1973) found that palmar

conductance rose while participants took the Stroop Color-Word Interference Test.

Cognitive Anxiety

The Brief Stress Inventory (Press, 1988) was given as a measure of cognitive

anxiety. Consistent with the Andersen and Williams (1988) model, the Brief Stress

Inventory (BSI) examines various potentially stress inducing aspects of the individual's

life, gathers information on specific symptoms of stress the person is experiencing, and

explores the coping strategies used. The Brief Stress Inventory asks about 100 questions

(depending on how some of the questions are answered), covering 16 areas or subscales,

and takes approximately 20 min to complete. See Appendix D regarding the 16 subscales

of the BSI. The average internal reliability of the scales, as measured by coefficient

alpha, varies from a high of 0.88 to a low of 0.56. The BSI compiles each participant's

answers into an individualized report.

Regarding the distinction between trait and state anxiety, the BSI questions

emphasize trait characteristics. Consequently, the proposed regression models are

modified as:

YCTD = Po+P1(BSI Scores)+p2(EMG ALS)+p33(TEM ALS)+34(EDR ALS)+-

YCPD = po+P31(BSI Scores)+p2(EMG ALS)+33(TEM ALS)+p4(EDR ALS)+s

Again taking advantage of stress/stressor reciprocal interaction, the BSI doubled

as a stressor by potentially inducing state anxiety for participants while completing the

survey, as indicated by simultaneously recorded psychophysiological arousal data.

Physiological Arousal

The apparatus for measuring psychophysiological arousal was the J&J 1-330

Personal Computer Physiological Monitoring System with USE language version 1.2.

Modality instrumentation included the Electromyograph Module M-501, and two

Temperature/ Electrodermograph T-601 Modules. The J&J Standard Volumes software

was reprogrammed to create a setup tailored to the modalities and treatment conditions of

this study. Software timing was programmed so that data were collected every 2 sec and

averaged every 30 sec.

As discussed earlier, this arousal data more accurately depicts stress when

expressed as Lacey's (1956) interindividual Autonomic Lability Scores. Due to the

nature of the Autonomic Lability Score, it is unnecessary to control for circadian arousal



Forty female Shands HealthCare employees, whose jobs include computer keying

for at least two hours per day, were recruited as participants. Participation was restricted

to one gender to decrease the variance of the physiological data. In addition, females

were selected because they have a higher incidence of CTDs.

The mean age of the participants was 42.8, with a maximum of 60 years and a

minimum of 21 years. All of the participants were generally healthy enough to be

employed, and none were currently under any work restrictions for medical reasons.

Minority Inclusion

In Alachua County, approximately 73% of the population are Caucasians, 19%

are African Americans, and 8 % are of other ethnic origins (U.S. Census Bureau). The

recruitment plan mirrored this ethnic representation via the sample depicted in Table 2-1.

Table 2-1. Participant ethnic representation

White Black or American Asian Native Hawaiian Hispanic or Total
African Indian and and Other Latino (of
American Alaska Native Pacific Islander any race)
29 8 0 2 0 1 40


Prospective participants were asked to complete the preliminary Discomfort

Survey (Appendix A), and to read and sign the Informed Consent Form approved by the

UF Health Center Institutional Review Board. In general, participants were included in

the study if they indicated a willingness to complete the Brief Stress Inventory (Press,

1988), and agreed to have psychophysiological measurements taken (e.g., frontalis

electromyography, digital temperature, etc.), as evidenced by their signing the Consent

Form. However, toward the end of sampling, participants were selected only if their

highest preliminary Discomfort Survey score contributed to achieving a balance among

low, medium, and high symptom scores. Volunteers were recruited until the proposed n

of 40, with the distribution of highest preliminary Discomfort Survey scores shown in

Figure 2-1, was attained.

10 --...------.

0 1 2 3 4 5 6 7 8 9 10
No Discomfort Worst
Figure 2-1. Participant's preliminary Discomfort Survey scores

Employees would have been excluded from the study if they had ever received

biofeedback training. This subpopulation would have possibly confounded the

physiological measurements used as independent variables. However no such

participants were among the volunteers.

Prospective participants were informed that their involvement in the study would

require 50-60 min. They were told that they were free to withdraw their consent and to

discontinue their participation at any time without penalty.

Regarding the physiological sensors, participants were told that three surface

electrodes would be applied to their forehead to record muscle tension, that one surface

thermistor would be attached to a finger of their dominant hand to monitor temperature,

and that two surface electrodes would be attached to two fingers of their dominant hand

to record sweat activity. They were further told that their foreheads would be wiped with

rubbing alcohol prior to sensor placement. Further, it was explained that these sensors

are identical with those normally used during biofeedback training and, as such, present

no more than minimal risk. Participants were informed that this physiological data would

be collected for approximately 30 min.

Participants were informed that there would be four experimental conditions of

varying durations (1) a baseline period, (2) the Brief Stress Inventory (Press, 1988), (3)

the Stroop Color-Word Interference Test, and (4) a symptoms survey. Participants were

informed that all data would be kept confidential and that only the principal investigator

would know their identity. Finally, they were informed that questions or concerns about

their rights as a research participant could be directed to the UF Health Science Center

Institutional Review Board.

Data obtained from individually identifiable participants were used for research

purposes only. The sources of data included the participant's responses on (1) the

Discomfort Survey (Appendix A), (2) the Brief Stress Inventory (Press, 1988), (3) the

Chest and Upper Extremity Symptoms Survey (Appendix B), (4) the Back and Lower

Extremity Symptoms Survey (Appendix C), (5) the participant's EMG, temperature, and

electrodermal physiological data collected during the baseline period and the three

experimental conditions, and (6) the participant's score on the Stroop Color-Word

Interference Test.


The study was conducted in room 1002, the Occupational Health Services

Department of Shands at UF. After collecting the Informed Consent forms, participants

were seated and the thermistor for digital temperature was attached at the base of the nail

bed of the middle finger of the dominant hand. The finger temperature was allowed to

stabilize as the remaining connections necessary for recording the physiological variables

were made. The two sensors for recording the digital electrodermal response (EDR) were

attached to the palmar side of the middle segments of the ring and index fingers of the

dominant hand. The three electromyograph (EMG) electrodes were centered on the

frontalis muscle about one inch above the eyebrows. The EMG site was cleaned with

alcohol and the EMG and EDR sensors treated with electrode cream prior to attachment.

Room temperature was maintained at 75F.

As soon as the digital temperature stabilized and all sensor attachments were

verified to be providing accurate readings, participants were asked to relax for a 2-min

baseline arousal measurement recording. After the recording of baseline arousal data,

participants were then asked to begin the first of the three randomly assigned conditions.

Each participant proceeded at her own pace, with no time limit established, and arousal

data were recorded for the first 10 min of each condition.

During the BSI condition, the participants were asked to complete the

computerized stress survey, but that they did not have to answer any question that they

did not wish to answer. It was explained that by completing the survey, they might gain

insight into what stresses them, or what their usual coping strategies are. They were

apprised of the fact that they would receive a printout of their stress profile and a copy of

the accompanying text "Stress? Find Your Balance."

For the cognitive performance task, the object of the Stroop Color-Word

Interference Test was explained. Participants were encouraged to respond as quickly and

as accurately as they could to the presentation of stimulus colored words. Congruent and

incongruent colored words were presented for 200 trials. Arousal recording was resumed

and the participants performed the task for 10 min, after which the recording was halted.

The participants' average congruent and incongruent response times were recorded after

each set of 20 trials.

As the third randomly presented condition, the Chest and Upper Extremity, and

Back and Lower Extremity Symptoms Surveys (Appendices B and C) were explained to

the participants. They were instructed to score both work and nonwork-related

musculoskeletal symptoms, on a scale of 0 (No Discomfort) to 10 (Worst Imaginable

Discomfort). Participants were asked to complete these forms at their own pace while

arousal data were recorded for 10 min.

While the psychophysiological sensors were being removed, the BSI scores were

saved on a disk, and an individualized report printed. Next the arousal data were saved

on a disk and printed. Before leaving the testing area, each participant was given her

individualized BSI report, a copy of the accompanying textbook, offered a copy of the

Informed Consent, and thanked for volunteering. (A sample Graphic Summary from a

BSI Individualized Report, and a sample arousal data report are submitted as

informational items in Appendices E and F.)

Statistical Analysis

Statistical Analysis was performed with Number Cruncher Statistical System

(NCSS) 2001 software. Hintze (2001) recommends 10 participants per variable. As

depicted in Table 2-2, a sample size of 40 achieves 70% power to detect an R2 of 0.20

attributed to three independent variables using an F-Test with a significance level (alpha)

of 0.05.

Table 2-2. Multiple regression power analysis
mInd. Variables
Power N Alpha Beta Cnt R2
0.70969 40 0.05000 0.29031 3 0.20000

Response Variables

The two dependent variables are (1) a self-reported intensity of the participant's

musculoskeletal discomfort, and (2) the participant's average response time on the Stroop

Color-Word Interference Test. The research hypothesis is that there will be a linear

relationship between these response variables and at least one of the predictor variables.

Predictor Variables

The quantitative factors are (1) stress as estimated by the participant's scores

(continuous) on the Brief Stress Inventory, and, (2) the Autonomic Lability Scores (ALS)

(continuous) during the three experimental conditions (a) for frontalis muscle tension

(EMG) (microvolts), (b) for digital temperature (TEM F) and, (c) for the digital

electrodermal response (EDR) (micromhos). There are no qualitative or extraneous

factors included in the equations.

Regression Models

The proposed regression models are therefore:

YCTD = Po0+P31(BSI Scores)+p32(EMG ALS bsi,ss)+133(TEM ALS bsi,ss)+P34(EDR ALS bsiss)+e

YCPD = Po+PI31(BSI Scores)+p2(EMG ALSst)+P3(TEM ALSst)+P4(EDR ALSst)+c

where the arousal predictors are subscripted bsi for the Brief Stress Inventory condition,

ss for the symptom survey condition, and st for the Stroop Color-Word Interference Test



The symptom survey and Stroop Color-Word Interference Test experimental

conditions produced the dependent variable scores shown in Table 3-1.

Table 3-1. Participants' dependent variable scores

CTD Total Average CTD Total Average
n Score Discomfort Stroop n Score Discomfort Stroop
Score RT (sec) _________Score RT (sec)
1 7 10 1.3055 21 27 27 1.1290
2 0 0 1.2570 22 10 10 1.4050
3 1 1 1.2745 23 28 28 1.2145
4 5 5 1.1155 24 18 18 1.0970
5 0 2 1.3155 25 21 27 1.2250
6 32 32 1.2565 26 8 12 1.4650
7 5 6 1.0980 27 5 5 1.1415
8 3 7 1.1480 28 0 0 1.1710
9 0 0 1.3410 29 13 13 1.5125
10 12 12 1.4230 30 2 2 1.5285
11 2 2 1.3115 31 1 2 1.4245
12 14 14 1.7145 32 19 19 0.9285
13 2 2 1.1625 33 33 33 1.1630
14 16 16 1.2615 34 33 33 1.3440
15 0 0 1.2210 35 0 2 1.1985
16 18 18 1.0735 36 3 3 1.0985
17 25 28 1.0880 37 13 13 1.0395
18 0 0 1.0250 38 2 4 1.1125
19 5 8 1.3720 39 0 0 1.2870
20 3 6 1.4500 40 0 0 1.4420

Histograms of the participants' CTD scores and average Stroop Color-Word

Interference Test response times from Table 3-1 are depicted in Figures 3-1 and 3-2.

0 8

0 2 4 6 8 1012141618202224262830

Figure 3-1. Histogram ofparticipants' CTD scores

10- \
0 4
0.80 1.00 1.20 1.40 1.60 1.80
Figure 3-2. Histogram ofparticipants' average Stroop Test response times

The raw data provided by the Brief Stress Inventory experimental condition were

sent to Preventive Measures, Inc. for batch processing. To avoid multicollinearity

problems, these data were processed without the linear dependencies normally used to

derive the scales reflected on the Graphic Summary of the BSI's Individualized Report.

This processing resulted in the return of 13 subscale scores, per participant, for use as

trait anxiety predictor variables. As illustrated in Table 3-2, 49% of the participants'

scores indicated moderate to severe cognitive anxiety regarding the subscales of the Brief

Stress Inventory.

Table 3-2. Number and percent of employees at each stress level for each BSI subscale

Minimal Stress Moderate Stress Severe Stress
Work or Primary Activity 27 67.5% 12 30% 1 2.5%
Marriage/Primary Relationship 24 60% 7 17.5% 3 7.5%
Friends and Social Life 29 72.5% 11 27.5% 0 0%
Physical Health 18 45% 15 37.5% 7 17.5%
Family Relationships 18 45% 17 42.5% 5 12.5%
Self-esteem 26 65% 14 35% 0 0%
Physical Appearance 8 20% 14 35% 18 45%
Time 5 12.5% 13 32.5% 22 55%
Joys (or lack thereof) 18 45% 21 52.5% 1 2.5%
Physical Symptoms 20 50% 17 42.5% 3 7.5%
Feeling Discontent 25 62.5% 14 35% 1 2.5%
Feeling Dissatisfied 19 47.5% 18 45% 3 7.5%
Feeling in Charge of Life 27 67.5% 11 27.5% 2 5%
Scores X Subscales =514 264/514= 51% 184/514 = 36% 66/514=13%

The Marriage/Primary Relationship subscale was not used because six

participants responded that they were not married or in a relationship. Such rows with

missing values are ignored by the all possible regressions procedure of NCSS 2001.

As per Lacey's (1956) recommendations, the EMG, temperature, and EDR

measurements of maximum sympathetic activity were standardized and calculated as

Autonomic Lability Scores for use as state anxiety predictor variables. For the baseline

period, the minimum 30-sec average for digital temperature, and the maximum 30-sec

averages for EMG and EDR for each participant were centered by subtracting their mean

and scaled by dividing their standard deviation to become the Xz,i terms of the

aforementioned formula. For each of the three experimental conditions, the minimum

temperature 30-sec average and the maximum 30-sec averages for EMG and EDR for

each participant were similarly centered and scaled to become the Yz,i terms of the

formula. The correlations of these Xz,i and Yz,i terms for the three experimental

conditions and the three modalities produced nine rxy terms for use in the formula. With

these terms, computation via the ALS formula yielded 40 EMG scores, 40 temperature

scores, and 40 EDR scores for each of the three conditions.

The three treatment conditions all acted as stressors inasmuch as the mean arousal

measurement of central and/or autonomic nervous system activity was greater than the

corresponding mean baseline measurement, during at least one of the conditions, for

every participant. Table 3-3 best illustrates activation of the autonomic nervous system

as reflected by the electrodermal response.

Table 3-3. Percent of participants whose mean arousal increased and decreased during
each treatment condition relative to their baseline mean arousal

Modality Brief Stress Inventory Stroop Test Symptom Survey
EMG Increased .45 .45 .57
Decreased .55 .55 .43
Temperature Increased .57 .55 .55
Decreased .43 .45 .45
EDR Increased .93 .93 .98
______ Decreased .07 .07 .02

However, more specifically to Lacey's (1956) interindividual Autonomic Lability

Scores, Table 3-4 illustrates that the preponderance of treatment condition Yi terms

reflect nervous system activation relative to the corresponding baseline Xi terms.


Table 3-4. Percent ofparticipants whose treatment condition Yi arousal terms surpassed
their corresponding Xi baseline arousal terms

Modality _____Brief Stress Inventory Stroop Test Symptom Survey
EMG Max> .95 .93 1.00
______Base Max____________________
Temperature Min < .75 .60 .73
~_____Base Min_____
EDR Max> .98 .98 .98
~_____Base Max_______________

According to Lacy's theory, however, the data of both Tables 3-3 and 3-4 represent only

arousal. The derived Autonomic Lability Scores are the measures of stress.

Cumulative Trauma Disorder Symptoms

Data preparation was accomplished by running the dependent variables through

the NCSS 2001 data screening procedure. Consequently, three participants with outlying

CTD symptom scores were deleted from the spreadsheet, leaving an n of 37 for this


Variable selection was accomplished by running NCSS 2001's all possible

regressions procedure on the 12 BSI subscale scores, and the averages of the EMG,

temperature, and EDR Autonomic Lability Scores collected during the BSI and symptom

survey conditions. Adhering to the recommendation of 10 observations per variable, the

best two-variable model was subjected to the multiple regression procedure. Hierarchical

forward with switching subset selection, with up to one-way model. terms, was employed

for this multiple regression procedure.

The resultant estimated model, analysis of variance (Table 3-5), and normal

probability plot (Figure 3-3) follow:

CTD Symptom Score = 4.9172 6.0980 SELF ESTEEM + 8.6822 FRUSTRATION

Table 3-5. Analysis of variance of CTD symptom score regression model
Source DF R2 F-Ratio Level
Intercept 1
Model 2 0.2906 6.964 0.0029
Self-esteem 1 0.1003 4.806 0.0353
Frustration 1 0.2899 13.894 0.0007
Error 34 0.7094
Total 36 1.0000

Normal Probability Plot of Residuals of CTD_Score

o 11.3

WQ -0< ^ y6\, ^

-15.0 . 11 / . . . . . . . .
-3.0 -1.5 0.0 1.5 3.0
Expected Normals

Figure 3-3. Normal probability plot of residuals of CTD symptom scores

The above CTD Symptom Score regression equation produces lower R2 values

for the total pain scores (R2 = 0.27), or with the highest pain scores (R2 = 0.15) as the

dependent variables. Similarly, when hierarchical forward with switching subset

selection included using two-way model terms, the best two-variable model remained the


Figure 3-4 illustrates the linear relationship between the participant's CTD score

and their Frustration subscale score.



S20.0- o
0 -0 0

o 10.0

0 0 0
0 0 00 00 0
0.0 e
1.0 2.2 3.3 4.5

Figure 3-4. Linear relationship of CTD score and Frustration subscale score

Cognitive Performance Decrement

Using the data for the same 37 participants, variable selection was accomplished

by running NCSS 2001 's all possible regressions procedure on the 12 BSI subscale

scores, and the EMG, temperature, and EDR Autonomic Lability Scores collected during

the Stroop Color-Word Interference Test condition. Adhering to the recommendation of

10 observations per variable, the best two-variable model was subjected to the multiple

regression procedure. Hierarchical forward with switching subset selection, with up to

one-way model terms, was employed for this multiple regression procedure.

The resultant estimated model, analysis of variance (Table 3-6), and normal

probability plot (Figure 3-3) follow:

Stroop average RT = 1.2099 0.0068 ALSiEDRst + 0.0076 ALSiTEMst

Table 3-6. Analysis of variance of Stroop average response time recession model
Source DF R2 F-Ratio Level
Intercept 1
Model 2 0.2924 7.026 0.0028
ALSiEDRst 1 0.1586 7.619 0.0092
ALSiTEMst 1 0.2061 9.906 0.0034
Error 34 0.7076
Total 36 1.0000

Normal Probability Plot of Residuals StroopAVGRT


2 0.1

'( -0
(r_ _0.3,

Figure 3-3. Normal probability plot of residuals of Stroop average response times

Figure 3-4 illustrates the linear relationship between the participants' average

Stroop Color-Word Interference Test RT and their ALSiTEMst.

ALSiTEMst vs Stroop AVG_RT


0 0
o o o- o o

0 00 0
0 o
0 o 0
0 0

'0. 46. 63. 80.

Figure 3-4. Linear relationship of Stroop average RT and the ALSiTEMst





The stated purposes of this study were accomplished. First, two regression

models that quantify the psychophysiological stress/CTD and psychophysiological

stress/performance relationships were developed. Second, the experimental method

regarding CTD symptoms could be repeated with individual employees to serve as a

noninvasive intervention tool to justify stress management therapies in the treatment of


The two regression models produced have their strengths. CTDs, with their

typical gradual onset, were best predicted by trait cognitive anxiety characteristics

subscales. Similarly, real-time cognitive performance was best predicted by state

psychophysiological measurements. Relative to expectations, however, the above results

are mixed. The linear relationship among CTD symptoms and two cognitive anxiety

subscale scores is significant and plausible. However, sympathetic stress reaction

measures were not among the regressors. Second, the linear relationship found among

average response times on the Stroop Color-Word Interference Test and two sympathetic

stress reaction measures is significant. However, the signs on both regressors are

opposite those expected. Perhaps this circumstance was caused by an underdefined

regression model (Montgomery & Peck, 1992), but could reflect that this is a nonsense

relationship. Indeed, the correlation of CTD scores to average response times is -.20. On

the other hand, it is possible that, as a group, the participants were not sufficiently


psychophysiologically stressed by the experimental condition. The group means during

the Stroop Color-Word Interference Test condition were YiEMGst = 2.85 microvolts;

YiTEMst = 84.45 F; YiEDRst = 12.33 micromhos. If the group was not sufficiently

psychophysiologically stressed, even those participants with an EMG ALS bsi,ss > 50,

TEM ALS bsi,ss < 50, and EDR ALS bsi,ss> 50 would be aroused versus

psychophysiologically stressed. This experiment would then be serving as a vigilance,

versus performance decrement study, and the signs obtained for the regressors would be


That the experimental design reflects the Andersen and Williams (1988) model,

by accommodating the reciprocal interaction of the components of the model, is viewed

as a major strength of this experiment. Thoughts generated by completing the cognitive

anxiety survey affected arousal, in turn potentially inducing anxiety that could be

reflected in the Brief Stress Inventory (1988) subscale scores. Thoughts generated by

completing the symptoms survey affected arousal, in turn potentially producing anxiety

that could impact the symptoms survey scores. Anxiety caused by performing the Stroop

Color-Word Interference Test affected arousal, in turn potentially causing more anxiety

that could result in longer response times. Participant knowledge that arousal

measurements were being collected during all three conditions would conceivably

increase any somatic anxiety experienced during these reciprocal interactions.

Moreover, inclusion of the Brief Stress Inventory (1988) in the experimental

design is considered a strength of this study. The BSI accommodates the interaction of

stressors and coping skills as per the Andersen and Williams (1988) theoretical model, as

well as recognizes minimal stress (subscale scores less than 2) as a strength (Selye, 1976;


Welford, 1973). With its 16 subscales, the BSI covers stressors similar to those cited by

Hackfort and Schwenkmezger (1993) and Smith and Carayon-Sainfort (1989).

That two cognitive anxiety subscales are the predictors in the CTD regression

model could be considered a limitation of the study. To fully support the stated rationale,

the preferred predictors would have been the Overall Stress subscale of the Brief Stress

Inventory (1988), and the TEM ALS bsi, to indicate causal ischemia. For one overall

cognitive anxiety score to capture the interactions of the Andersen and Williams (1988)

model would have been elegant. Similarly, a TEM ALS bsi. predictor would have

provided biological validity to the regression formula by demonstrating that ischemia was

the probable bridge between cognitive anxiety and cumulative trauma disorder. Such a

regression equation could possibly serve as an operational definition of psycho-

physiological stress, and be used in other health/injury/performance studies. It is also

possible, as conjectured earlier regarding the Stroop Color-Word Interference Test

regression, that the TEM ALS bi, ss did not emerge as a predictor of CTDs because as a

group the participants were not sufficiently psychophysiologically stressed by the

experimental conditions. The group means during the BSI and symptom survey

conditions were, = 3.85 microvolts; YiTEMbsi,ss = 84.22 F; YiEDRbsiss =

13.08 micromhos. In support of this possibility, regression analysis with the same n = 37

on the model

YCTD = 3Po+P31(BSI Scores)+P2(EMG ALS bsi,ss)+P3(TEM ALS bsi,ss)+P4(EDR ALS bsi,ss)+8

with the Frustration and Self-Esteem BSI scores, and with filters EMG ALS bsi,ss > 50,

TEM ALS bsi,ss < 50, and EDR ALS bsi,ss > 50, provides an n = 29 and produces an R2 =

0.44 andp = .0127.


Evaluating the R2 values obtained by the two regression equations reflects that the

models produced by this study were purposefully underdefined, with predictors related

only to psychophysiological stress. To expand the CTD regression model with predictors

regarding the durations, frequencies, forces, and postures associated with the participants'

job tasks was beyond the goals of this effort. Similarly, no attempt was made to include

measurements of attention or distraction, or to discriminate among reaction, movement,

and response times in the second study. The rationale for the second study was merely

that if a given level of stress correlated with both CTDs and cognitive performance

decrement, the finding would provide added incentive for managers to reduce workplace


As stated earlier, this study intended to test Andersen and Williams' (1988)

assumption that certain individuals, under stress, exhibit greater increases in generalized

muscle tension. As previously depicted in Table 3-3, this study supports their

assumption. Relative to their baseline values, 45% of the participants increased their

mean frontalis EMG during the Brief Stress Inventory, 45% increased their mean

frontalis EMG during the Stroop Color-Word Interference Test condition, and 57%

increased their mean frontalis EMG during the symptom survey. However, participants

also exhibited greater autonomic reactivity in the electrodermal response, or in decreasing

their digital temperature as well. These interindividual response differences give

credence to Caldwell et al.'s (1994) and Landers' (1980) recommendations to collect

multiple psychophysiological measures when conducting stress research. More to the

point regarding Andersen and Williams' assumption, the correlation of the CTD scores

and the EMG Autonomic Lability Scores during the conditions that produced the

regression model was only .14. Perhaps rather than generalizing muscle tension to the

frontalis muscle, those susceptible to CTDs may compartmentalize muscle tension in a

specific way. For example, in light of the McNulty et al. (1994) findings, employees who

suffer upper extremity disorders are possibly more prone to express a sympathetic stress

response via the intrafusal fibers of the spindles of their trapezius muscles.

Figure 4-1 illustrates the rationale behind Lacey's (1956) Autonomic Lability

Score. The Law of Initial Values is reflected by the higher score (82) being assigned to


1.5 82 ALS
1 -, -71 ALS
0.5 --__--_ _50 ALS
0 --------- --27ALS
-0.5 -

Figure 4-1. Electrodermal Autonomic Lability Scores

the participant with the greater baseline value, even though the sympathetic response (the

slope) is almost identical to that with a score of 71. A homeostatic, or autonomically

neutral response relative to the other participants, is assigned a score of 50. A

parasympathetic dominant response is assigned an ALS of 27, even though the baseline

value was high relative to the group. This experience with Lacey's Autonomic Lability

Score affords the following observations. There is no argument with Lacey's contention

that there are pitfalls awaiting the investigator who simply quantifies autonomic

responses as an algebraic or percentage change. Further, there is no objection to the Law

of Initial Values. However, due to the Autonomic Lability Score's reliance on the

maximum sympathetic prestimulus measurement (xz,i) and maximum sympathetic

response measurement (Yz,i), the ALS might not capture all of the relevant facets of the

sympathetic stress reaction. Figure 4-2 illustrates dissimilar electrodermal responses

during the baseline and Stroop Color-Word Interference Test conditions for two

participants, both of whom have Autonomic Lability Scores of 50. The ALS might be a

12.00 Y Variables
10.83 A Base EDR

9.67 AEDR Stroop
8.50 -

7.33 -
6.17 -
5.00 -______
0 1 3 5 6 8 10
M minutes

20.00 Y Variables
18.33 OBase EDR

16.67- &EDR Stroop
15.00 -

13.33 -
11.67 .

10.00^ -~
10.00 0. 1 3 5 6 8 1,0
M minutes

Figure 4-2. Dissimilar electrodermal responses with identical Autonomic Lability Scores

fuller measure of the autonomic response if the areas under the baseline and response

curves were used as the xz,,i and yz,i terms in Lacey's formula. The Law of Initial Values

would still be observed. Similarly, Lacey's admonition to use only measures that are

based on physiological theory, versus mathematical transformations, would be honored.

Magnitude, frequency, and duration dimensions for stressors and stress reactions are

fundamental to Selye's (1976) theory, and these are the dimensions offered by the areas

under the curves. With richer data, the psychophysiological stress portion of the CTD

regression formula might be strengthened in predictive value. As stated earlier, such a

regression equation would fill the need cited by Baker and Karasek (2000) for a simple

yet comprehensive instrument for measuring stressors and stress reactions.

The combined ischemic effects of centrally and autonomically mediated muscle

bracing and sympathetic vasoconstriction that occur with stress, do result in ischemic

tissue. The sympathetic stress reaction therefore probably plays a role in the

development of cumulative trauma disorders. Figure 4-3 attempts to unify the concepts
o^sasI __ Adaptatiop, e.g., Learning _A Rest/Treatment
Perception ,+ Restore Capabilities
Homeostasis"! Stressors Capabilities Eustress L_
Hoe is S+ Hazard Controls
I_ Adaptatioh, e.g., Conditioning Reduce Stressors

Feedback Maladaptation, e.g., Anxiety -+-----Performance
Loops iDecrease/Error
> Capabilities Distress
Positive :- | Fatigue
Feedback Maladaptation, e.g., Bracing 4TjuInjury/Illness--

Alarm Resistance Exhaustion
Arousal Psychophysiological Stress

Figure 4-3. Stress Feedback Loop Model of the General Adaptation Syndrome in
Ergonomics/Human Factors

and terms of Selye (1976), Lacey (1956), McGrath (1976), Welford (1973), and Caldwell

et al. (1994) within an ergonomic context. Within this model's framework, the

sympathetic stress reaction's role in the development of CTDs is effected via the

diminution of the employee's internal resources to meet the demands of the external

stressors. This circumstance mirrors Selye's description of a pluricausal illness. Given

that these internal resources are temporally changing perceptual and psychological

processes (McGrath, 1976; Lazarus, 1977; Siegrist, 1996), implies that psycho-

physiological studies are amenable only to observational, quasi-experimental research.

Mediators and behaviors feed backward and forward to also act as stressors. As Lacey

noted, responses of the autonomic nervous system are ubiquitous. One cannot stimulate

the organism, however innocuously, without producing some evidence of disturbance of

autonomic equilibrium. At some point, this disturbance crosses a threshold separating the

psychological and physiological energizing responses of arousal, and the maladaptive

psychophysiological stress responses such as excessive anxiety, muscular bracing, and

ischemia. In Selye's terms, these maladaptive responses are overextensions of the energy

mobilization and disease fighting systems of the body.

In future attempts to uncover this causal role, the experimental design to produce

a CTD regression equation that includes an arousal predictor should address the

following (1) Ensure a strong correlation between the participants' recruiting and

experimental CTD scores by using the same survey for both purposes, or alternatively,

using a survey only during the experiment. (2) Recruit a sample of participants whose

CTD scores' distribution is representative of the workforce of interest. (3) Use only

upper extremity CTD scores. (4) Exclude participants who are experiencing the hot

flashes associated with menopause (there is a sudden and dramatic increase in the digital

temperature). (5) Include photoplethysmography (PPG), a more direct measure of

circulating blood, among the physiological monitoring modalities. (6) Measure EMG at


the trapezius and forearm versus the frontalis muscle. (7) Explore using the areas under

the curves as the baseline and response terms when calculating the Autonomic Lability


In summary, as noted by Chengalur, Rodgers, and Bernard (2004), with a leaner

workforce and increased job demands of recent years, psychosocial factors in the job

have become increasingly important. There is increasing interest in the quantification of

psychosocial factors to better understand why some people seem to be more susceptible

to occupational injuries and illnesses, such as musculoskeletal disorders, than are others

on the same job. The findings and methodology of this study have contributed toward

this understanding. Further, the CTD regression equation derived in this study justifies

the use of stress management therapies for the treatment of CTDs. The author shares the

conclusion of McNulty et al. (1994), regarding their search for a connection between

musculoskeletal trigger points and stress, that helping individuals modify their

sympathetic response to stress, with therapeutic cognitive and behavioral techniques,

should be a major goal of treatment.


Participant ID

Think about how you feel RIGHT NOW.

1. Shade in all the areas of discomfort on the figure.
2. Rate the discomfort for each left and right side body area named in the box
below. Use the scale below, and write the score in each box.

0- -1-

-2- -3- -4- -5- -6- -7- -8- -9- -10

Rating Score

Discomfort Area Right Left







Lower leg





Think about how you feel RIGHT NOW.

3. Shade in all the areas of discomfort on the figure.
4. Rate the discomfort for each left and right side body area named in the box below.
Use the scale below, and write the score in each box.

0- -1- -2-

-3- -4- -5- -6- -7- -8-

-9- -10

Rating Score

Discomfort Area Right Left








Lower leg





Please complete a column for each area
that bothers you

Front of Front of
Neck Shoulder





El Left El Left 1 Left M Left Left E.] Left
1. Which side bothers you? []Right [] Right [] Right [] Right D] Right D Right
[E Both El Both E] Both El Both Both E] Both
2. In what year did you first notice the
problem?_______ _____ __________ _____
Lessthan lHr E] El El El El El
1 Hr to l Day El El El El El El
3.Howlongwas IDayto I Week D El El El El
each episode? 1 Week to 1 Month D EE El] El El
I to 6 months l E E El El ElD
More than 6 Months El El El E El El
Constant El El El El E
4. How often have Daily El E E El El El
you had separate Once a Week El El El El El El
episodes in the last Once a Month D E El El E El
year? Every 2 to 3 Months E] El El E El
_____ More than 6 Months El El El El El El

5. What do you think caused the problem? _______ ____ ____ _______

6. Was the onset of symptoms sudden or Sudden Sudden Sudden Sudden Sudden Sudden
gradual? (Circle one) Gradual Gradual Gradual Gradual Gradual Gradual
7. Is this problem interfering with your El Yes El Yes E Yes E Yes El Yes El Yes
ability to do your job? [I No [] No El No E No [ No El No
8. Have you told your supervisor about E Yes El Yes E Yes E Yes El Yes [! Yes
these symptoms? El No [E No [El No E No [E No [E No
9. Do these symptoms clear up over the El Yes E Yes El Yes E Yes El Yes El Yes
weekend? El No [ No El No [E No ElNo El No
10. Do you have any health problems or E Yes El Yes El Yes El Yes El Yes [E Yes
injuries that might account for them? [] No El No [E No [E No [E No [E No
11. Has a Health Care Worker diagnosed El Yes E Yes E Yes El Yes E Yes E Yes
your symptoms? [E No E No E No E No [ No [I No
12. Has a Health Care Worker prescribed El Yes El Yes El Yes E Yes D Yes [E Yes
work limitations? El No E No E No E No ElNo [ No
13. On the 0-10 scale below, how would
you rate this problem right now? E El El E E El
No Worst Imaginable
Discomfort Discomfort
0--------....--- I 2 ---------- 3 ------- 4 ------- 5 --------- 6 ------- 7 --------- 8 --. 9 ---- 10
14. How many days of work did you lose
in the last year due to this problem? days days days days days days
15. Days of restricted duty in the last year
due to this problem? days days days days days days


Please complete a column for each area
that bothers you

Back of

Back of




l Left E Left E Left D Left U Left D Left
1. Which side bothers you? I"Right D Right D Right n Right [ Right [D Right
S__________ _[Both E Both n Both D Both []Both EL Both
2. In what year did you first notice the
problem?______________ _________________
Lessthanl Hr D U U U U U
I Hrtol Day E] U U U U
3. Howlongwas 1 Day to I Week n U U U U U
each episode? 1 Week to 1 Month U U U U U U
1 to 6 months U U U U U U
__________More than 6 Months U U U U U
Constant U U U U U
4. How often have Daily U U U U U ]
you had separate Once a Week U U U U U U
episodes in the last Once a Month U U U U U U
year? Every 2 to 3 Months [ U U U U U
______More than 6 Months E] U U U U UD

5. What do you think caused the problem?_______________ _____

6. Was the onset of symptoms sudden or Sudden Sudden Sudden Sudden Sudden Sudden
gradual? (Circle one) Gradual Gradual Gradual Gradual Gradual Gradual
7. Is this problem interfering with your U Yes U Yes U Yes U Yes U Yes U Yes
ability to do your job? [U No U No U No [ No [ No [ No
8. Have you told your supervisor about U Yes U Yes D Yes U Yes U Yes U Yes
these symptoms? No No D No U No No [ No
9. Do these symptoms clear up over the U Yes U Yes U Yes U Yes 0 Yes D Yes
weekend? U No [ No [] No [ No No [ No
10. Do you have any health problems or D Yes U Yes D Yes U Yes U Yes U Yes
injuries that might account for them? [] No U No U No U No U No U No
11. Has a Health Care Worker diagnosed [ Yes D Yes U Yes U Yes U Yes U Yes
your symptoms? Ul No [ No U No U No [ No [ No
12. Has a Health Care Worker prescribed U Yes U Yes U Yes U Yes U Yes 0 Yes
work limitations? [U No No [ No ] No U No No

13. On the 0-10 scale below, how would
you rate this problem right now? fl [] F] ] E]
No Worst Imaginable
Discomfort Discomfort
0 ------------ I --------- - 2 ------------ 3 -------4 ------------ 5 -------6---- 6 -------7---- 7 -------8---- 8 --------9--- 9 ------------ 10

14. How many days of work did you lose
in the last year due to this problem? days days days days days days
15. Days of restricted duty in the last year
due to this problem? days days days days days days


Self-Report on Overall Stress

Work or Primary Activity

Marriage or Primary Relationship

Friends and Social Life

Family Relationships

Physical Health

Eating Habits


Physical Appearance


Joys (or their lack)

Ability to Make Changes

Physical Symptoms

Feeling Discontent

Feeling Unsuccessful/Dissatisfied

Feeling in Charge of your Life


Work or Primary Activity ************

Marriage/Primary Relationship ** *

Friends and Social Life ************

Physical Health ************ ************* *

Family Relationships ********

Self-Esteem ************ *****

Physical Appearance ************

Time ************ ************* **

Joys (or lack thereof) ************

Physical Symptoms ************

Feeling Discontent ************

Feeling Dissatisfied ************ ****

Feeling in Charge of Your Life ************

Copyright 1986-89, Preventive Measures, Inc., Lawrence, Kansas


Current timing: Seconds/Average=30.000 Averages/Trial =1

STATISTICS at period level
PREBASE-------- Al Cl Dl
Mean: 0.60 74.75 11.37
Stan. Deviation: 0.09 0.37 0.19
Maximum: 0.75 75.02 11.70
Minimum: 0.53 74.13 11.22
Data Points: 4 4 4
BSI------------- Al Cl D1
Mean: 0.60 72.00 12.89
Stan. Deviation: 0.04 0.93 0.61
Maximum: 0.69 73.54 13.96
Minimum: 0.56 70.58 11.90
Data Points: 20 20 20
STROOP ---------Al Cl D1
Mean: 0.56 68.11 13.20
Stan. Deviation: 0.07 0.23 0.56
Maximum: 0.76 68.66 14.664
Minimum: 0.47 67.82 12.50
Data Points: 20 20 20
SYMPTOM-------- Al Cl Dl
Mean: 0.45 68.20 13.08
Stan. Deviation: 0.11 0.10 0.65
Maximum: 0.73 68.41 14.41
Minimum: 0.36 68.00 12.22
Data Points: 10 10 10

AVERAGES at trial level
PREBASE -------- Al Cl Dl
Tr#l 1 0.53 75.02 11.27
Tr#2 0.57 75.02 11.22
Tr#3 0.55 74.82 11.29
Tr#4 0.75 74.13 11.70


Andersen, M. B., & Williams, J. M. (1988). A model of stress and athletic injury:
Prediction and prevention. Journal of Sport & Exercise Psychology, 10, 294-306.

Aruin, A. S. (2002). The biomechanical foundations of a safe labor environment:
Bernstein's vision in 1930. Motor Control, 6, 1-18.

Baker, D. B., & Karasek, R. A. (2000). Stress. In B. S. Levy & D. H. Wegman (Eds.),
Occupational health: Recognizing and preventing work-related disease and injury
(pp. 419-445). Philadelphia: Lippincott Williams & Wilkins.

Berggren, T., Hane, M., & Ekberg, K. (1988). Stress at work: An interactive model for
work environment analysis. In C. Zenz (Ed.), Occupational medicine: Principles
and practical applications (pp. 1075-1086). Chicago: Year Book Medical.

Blair, S. J. (1996). Pathophysiology of cumulative trauma disorders: Some possible
humoral and nervous system mechanisms. In S. D. Moon, & S. L. Sauter (Eds.),
Beyond biomechanics: Psychosocial aspects ofmusculoskeletal disorders in office
work (pp. 91-97). London: Taylor & Francis.

Caldwell, J. A., Wilson, G. F., Cetinguc, M., Gaillard, A. W. K., Gunder, A., Lagarde,
D., et al. (1994). Psychophysiological assessment methods (AGARD-AR-324).
Loughton, Essex: North Atlantic Treaty Organization Advisory Group for
Aerospace Research and Development.

Cannon, W. B. (1914). The interrelations of emotions as suggested by recent
physiological researchers. American Journal of Psychology, 25, 256-282.

Carayon, P., Smith, M. J., & Haims, M. C. (1999). Work organization, job stress, and
work-related musculoskeletal disorders. Human Factors, 41(4), 644-663.

Chengalur, S. N, Rodgers, S. H., & Bernard, T. E. (2004). Kodak's ergonomic design
for people at work (2nd ed.). Hoboken, NJ: John Wiley & Sons.

Fitzgerald, T. E. (1992). Psychosocial aspects of work-related musculoskeletal
disability. In J. C. Quick, L. R. Murphy, & J. J. J. Hurrell (Eds.), Stress and well-
being at work (pp. 117-133). Washington, DC: American Psychological

French, J. R. P. (1963). The social environment and mental health. Journal of Social
Issues, 19, 39-56.

Guyton, A, C. (1976). A textbook of medical physiology (5th ed.). Philadelphia: W. B.

Haakonson, N. H. (1980). Investigation of life change as a contributing factor in aircraft
accidents: A prospectus. Aviation, Space, and Environmental Medicine 51(9),

Hackfort, D., & Schwenkmezger, P. (1993). Anxiety. In R. N. Singer, M. Murphey, &
L. K. Tennant (Eds.), Handbook of research on sport psychology (pp. 328-364).
New York: Macmillan.

Hadler, N. M. (1990). Cumulative trauma disorders An iatrogenic concept. Journal of
Occupational Medicine, 32(1), 38-41.

Hadler, N. M. (1992). Arm pain in the workplace A small area analysis. Journal of
Occupational Medicine, 113-119.

Hales, D., & Hales, R. E. (1996, March 17). You can beat stress on the job. Parade, 24-

Hancock, P. A., & Warm, J. S. (1989). A dynamic model of stress and sustained
attention. Human Factors, 31(5), 519-537.

Hanin, Y. L. (1989). Interpersonal and intergroup anxiety in sports. In D. Hackfort, &
C. Spielberger (Eds.), Anxiety in sports: An international perspective (pp. 19-28).
New York: Hemisphere.

Hardy, L. (1990). A catastrophe model of performance in sport. In J. G. Jones, & L.
Hardy (Eds.), Stress and performance in sport (pp. 81-106). New York: Wiley &

Hardy, L., Jones, G., & Gould, D. (1996). Understanding psychological preparation for
sport: Theory and practice of elite performers. New York: Wiley & Sons.

Hardy, L., & Parfitt, G. (1991). A catastrophe model of anxiety and performance.
British Journal of Psychology, 82, 163-178.

Harrison, R. V. (1978). Person-environment fit and job stress. In C. Cooper, & R Payne
(Eds.), Stress at work (pp. 175-209). New York: Wiley.

Hintze, J. L. (2001). Power Analysis and Sample Size (Version 2000) [Computer
software]. Kaysville, UT: Number Cruncher Statistical System.

Kahneman, D. (1973). Attention and effort. Englewood Cliffs, NJ: Prentice-Hall.

Keyserling, W. M. (2000). Occupational ergonomics promoting safety and health
through work design. In B. S. Levy & D. H. Wegman (Eds.), Occupational
Health: Recognizing and preventing work-related disease and injury (pp. 419-
445). Philadelphia: Lippincott Williams & Wilkins.

Lacey, J. I. (1956). The evaluation of autonomic responses: Toward a general solution.
Annals of the New York Academy of Sciences, 67, 123-164.

Landers, D. M. (1980). The arousal-performance relationship revisited. In M. J. Safrit
(Ed.), Research Quarterly for Exercise and Sport, 51(1), 77-90.

Lazarus, R. S. (1977). Cognitive and coping processes in emotion. In A. Monat, & R.
Lazarus (Eds.), Stress and coping. (pp.145-158). New York: Columbia
University Press.

MacLeod, C. M. (1991). Half a century of research on the Stroop effect: An integrative
review. Psychological Bulletin, 109(2), 163-203.

Martens, R. (1987). Coaches guide to sport psychology. Champaign, IL: Human

Martens, R., Burton, D., Vealy, R. S., Bump, L. A., & Smith, D. E. (1990). Competitive
state anxiety inventory-2. In R. Martens, R. Vealy, & D. Burton (Eds.),
Competitive anxiety (pp. 119-126). Champaign, IL: Human Kinetics.

McGrath, J. E. (1976). Stress and behavior in organizations. In M. Dunnette (Ed.),
Handbook of industrial and organizational psychology (pp. 1353-1395). Chicago:
Rand McNally.

McNulty, W. H., Gevirtz, R. N., Hubbard, D. R., & Berkoff, G. M. (1994). Needle
electromyographic evaluation of trigger point response to a psychological
stressor. Psychophysiology, 31, 313-316.

Mechanic, D. (1961). Stress, illness behavior, and the sick role. American Sociological
Review, 26, 51-58.

Montgomery, D. C., Peck, E. A. (1992). Introduction to linear regression analysis (2nd
ed.). New York: John Wiley & Sons.

Moon, S. D. (1996). A psychosocial view of cumulative trauma disorders. In S. D.
Moon, & S. L. Sauter (Eds.), Beyond biomechanics: Psychosocial aspects of
musculoskeletal disorders in office work (pp.109-143). Bristol, PA: Taylor &

Muhonen, G. (2003). Confusing colors. Retrieved November 16, 2003, from

National Institute for Occupational Safety and Health. (1994). Cumulative trauma
disorders: A manual for musculoskeletal diseases of the upper limbs. Putz-
Anderson, V. (Ed.). Bristol, PA: Taylor & Francis.

National Institute for Occupational Safety and Health. (1999). Stress at work (NIOSH
Publication No. 99-101). Washington, DC: U.S. Government Printing Office.

National Research Council. (1995). Emerging needs and opportunities for human
factors research. Washington, DC: National Academy Press.

Occupational Safety and Health Administration. (1995). "Ergonomic protection
standard" (Proposed standard). Washington, DC: Author.

Press, A. N. (1988). Brief Stress Inventory [Computer software]. Lawrence, KS:
Preventive Measures.

Price, D. D., Bush, F. M., Long, S., & Harkins, W. (1994). A comparison of pain
measurement characteristics of mechanical visual analogue and simple numerical
rating scales. Pain 56, 217-226.

Putz-Anderson, V. (1988). (Ed.). Cumulative trauma disorders A manualfor
musculoskeletal disease of the upper limbs. London: Taylor & Francis.

Rahe, R. H. (1974). Life change and subsequent illness reports. In E. K. E. Gunderson,
& R. H. Rahe (Eds.), Life stress and illness (pp. 58-104). Springfield, IL: Charles
C. Thomas.

Rotella, R. J., & Lerner, J. D. (1993). Responding to competitive pressure. In R. N.
Singer, M. Murphey, & L. K. Tennant (Eds.), Handbook of research on sport
psychology (pp. 528-541). New York: Macmillan.

Sauter, S. L., & Swanson, N. G. (1996). An ecological model of musculoskeletal
disorders in office work. In S. D. Moon, & S. L. Sauter (Eds.), Beyond
biomechanics: Psychosocial aspects of musculoskeletal disorders in office work
(pp. 3-21). London: Taylor & Francis.

Schwartz, M. S. (1987). Biofeedback: A practitioner's guide. New York: Guilford.

Selye, H. (1976). The stress of life (Rev. ed.). New York: McGraw-Hill.

Sheffield Hallam University. (2001, January 6). Stress in the workplace. In On-Line
Health and Safety Manual. Retrieved November 16, 2003, from

Siegrist, J. (1996). Adverse health effects of high-effort/low reward conditions. Journal
of Occupational Health Psychology, 1, 27-41.

Siska, E. (2002). The Stroop colour-word test in psychology and biomedicine. Acta
Universitatis Palackianae Olomucensis Gymnica, 32(1), 45-51.

Smith, M. J. (1987). Occupational Stress. In G. Salvendy (Ed.), Handbook of human
factors (pp. 844-860). New York: Wiley & Sons.

Smith, M. J., & Carayon, P. (1996). Work organization, stress, and cumulative trauma
disorders. In S. D. Moon, & S. L. Sauter (Eds.), Beyond biomechanics:
Psychosocial aspects of musculoskeletal disorders in office work (pp. 23-42).
London: Taylor & Francis.

Smith, M. J., & Carayon-Sainfort, P. (1989). A balance theory ofjob design for stress
reduction. International Journal of Industrial Ergonomics, 4, 67-79.

United States Census Bureau. (n.d.). State and County Quick Facts. Retrieved
November 16, 2003, from

United States Department of Labor. (2002, April 10). News (Bureau of Labor Statistics,
USDL 02-196). Washington DC: Office of Publications & Special Services.

Welford, A. T. (1973). Stress and performance. Ergonomics, 16(5), 567-580.

Westgaard, R. H. (1996). Effects of psychological demand and stress on neuromuscular
function. In S. D. Moon, & S. L. Sauter (Eds.), Beyond biomechanics:
Psychosocial aspects of musculoskeletal disorders in office work (pp. 75-89).
London: Taylor & Francis.

Wickens, C. D. (1992). Engineering psychology and human performance (2nd ed.).
New York: Harper Collins.

Williams, J. M., & Andersen, M. B. (1998). Psychosocial antecedents of sport injury:
Review and critique of the stress and injury model. Journal of Applied Sport
Psychology, 10, 5-25.

Williams, J. M., & Roepke, N. (1993). Psychology of injury and injury rehabilitation.
In R. N. Singer, M. Murphey, & L. K. Tennant (Eds.), Handbook of research on
sport psychology (pp. 815-839).New York: Macmillan.

Zaichkowsky, L., & Takenaka, K.(1993). Optimizing arousal level. In R. N. Singer, M.
Murphey, & L. K. Tennant (Eds.), Handbook of research on sport psychology
(pp. 511-527).New York: Macmillan.


John C. Lane holds a Bachelor of Science in biology and a Master of Science in

occupational health and safety. He is currently a doctoral candidate in the College of

Health and Human Performance at the University of Florida. He is employed as the

ergonomist for Shands HealthCare. He served in the United States Navy for twenty

years, as a hospital corpsman, as an aviation physiologist, and as an aeromedical safety

officer. His interests in human factors and ergonomics include technology transfer into

health care, hypo/hyberbaric environments, visual problems in traffic safety, and failure

mode effect analysis.

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.

,fessor of Exercise and Sport

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy. j

S. ischler-
Professor of Psychology

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully ad quate, in scope and quality,
as a dissertation for the degree of Doctor ofPhilosophy./ j '

Christopher PJanelle
Associate Prfessor of Exercise and
Sport Sciences

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.

Mark D. Tillman
Assistant Professor of Exercise and
Sport Sciences

This dissertation was submitted to the Graduate Faculty of the College of Health
and Human Performance and to the Graduate School and was accepted as partial
fulfillment of the requirements for the degree of Doctor of Philosophy.

May, 2004

DO,, College of Health and Human

Dean, Graduate School



IIUlllll llttii ll3 1262 08557 2096
3 1262 08557 2096