Citation
Detection of submaximal effort in isotonic back strength testing

Material Information

Title:
Detection of submaximal effort in isotonic back strength testing determination of optimal resistance level
Alternate title:
Determination of optimal resistance level
Creator:
Sadler, Ian John, 1966-
Publication Date:
Language:
English
Physical Description:
vi, 42 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Clinical psychology ( jstor )
Exercise ( jstor )
Honesty ( jstor )
Human back ( jstor )
Low back pain ( jstor )
Philosophical psychology ( jstor )
Rehabilitative medicine ( jstor )
Research methods ( jstor )
Torque ( jstor )
Velocity ( jstor )
Back -- physiology ( mesh )
Department of Clinical and Health Psychology thesis Ph.D ( mesh )
Diagnostic Tests, Routine ( mesh )
Dissertations, Academic -- College of Health Professions -- Department of Clinical and Health Psychology -- UF ( mesh )
Exertion ( mesh )
Isotonic Contraction ( mesh )
Reproducibility of Results ( mesh )
Research ( mesh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1999.
Bibliography:
Bibliography: leaves 37-41.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Ian John Sadler.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Ian J. Sadler. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
029470304 ( ALEPH )
50910894 ( OCLC )

Downloads

This item has the following downloads:


Full Text









DETECTION OF SUBMAXIMAL EFFORT IN ISOTONIC BACK STRENGTH
TESTING: DETERMINATION OF OPTIMAL RESISTANCE LEVEL














By

IAN J. SADLER


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

UNIVERSITY OF FLORIDA


1999













ACKNOWLEDGMENTS

A number of individuals have made significant contributions to this dissertation in particular and to my graduate training in general. Michael Robinson stands out as the person who has contributed the most to my professional development and provided me with a solid foundation in both clinical health psychology and, especially, experimental methodology. This dissertation is a testimony to his commitment to the highest standards of empiricism and scholarly work. I would also like to acknowledge the contributions of my committee members, Duane Dede, Jon Kassel, Samuel Sears, and Mark Trimble. These individuals have seen me through the process of qualifying examinations, dissertation proposal, and dissertation defense. Their involvement and support has been appreciated. Although not on my dissertation committee, Cynthia Belar, Jim Rodrigue, and Michael Perri were also instrumental in my development as a clinician and as an academic professional at the University of Florida. Collectively, these individuals have provided me with a model of the scientist-practitioner in clinical psychology.

Several individuals have made contributions of time and energy to aspects of the present study, including Patrick O'Connor, Joseph Riley, and John Otis. Joseph Riley has been a supportive friend and colleague from whom I have learned a great deal both professionally and personally.








I have made close and lasting relationships throughout the years of graduate

school. These friends have seen me through the good times and the difficult times. To these individuals (Amy, Becca, Braden, Elena, Kathy, and Gregg), I say a heartfelt thank you.

Finally, I will forever be indebted to my parents, Bridget and Brian Sadler, as well as my sister, Sandy, for their innumerable sacrifices and unwavering love and support throughout my years of graduate education. I dedicate this work to my family.


















AC"wi. E W~TS

ADS hA


Staeu ofhe Prblen
S,nc ntyfff

The Choic orfio= Tesnn
Th prset Stud

2 METHD





Paitic ant a E L






4 DISUSSION


RE E N E













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

DETECTION OF SUBMAXIMAL EFFORT IN ISOTONIC BACK STRENGTH
TESTING: DETERMINATION OF OPTIMAL RESISTANCE LEVEL By

Ian J. Sadler

August 1999

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


The purpose of this study was to investigate the use of an isotonic back strength assessment task in the detection of submaximal effort and to test the hypothesis that a submaximal effort would yield greater variability as measured by the coefficient of variation. There was also an examination of the optimal resistance level at which effort level could best be discriminated. The hypothesis proposed that a resistance level of either 15%, 25%, or 35% would produce convincing results with respect to effort discriminability. It was also hypothesized that the calculation of the CV measure based on 6 repetitions would result in an improved submaximal effort detection rate versus a CV calculation based on 3 repetitions. Sixty- six participants were asked to undergo back extension assessment. Each participant completed 3 practice repetitions followed by 7 repetitions at a 50% effort level and 7 repetitions at a 100% effort level. There was a 10








minute rest period following the first set of 7 repetitions and order of conditions was counterbalanced. Results indicated that while participants were able to produce a submaximal effort, there was not a significant difference in variability between effort levels across the three resistance levels. There was no difference in the use of the CV based on

6 versus 3 repetitions with respect to effort discriminability. Implications of these results are discussed with recommendations for continued investigation of effort level in backs through use of isokinetic testing methodology.













CHAPTER 1
INTRODUCTION

Musculoskeletal injuries are the most common cause of frequent or permanent

impairment (Brena & Meacham, 1990). In a review of the literature, Frymoyer and CatsBaril (1991) reported that while low back disorders are extremely prevalent in all societies, and probably have not increased substantially over the last two decades, they note that what has increased is the rate of disability, the reasons for which are uncertain. According to these researchers what has changed is society's perception, most particularly relating to the disability that results. With an increasing rate of disability, the costs have increased to industry and government, and there has been an increased utilization of medical care, including surgical intervention. While Frymoyer and Cats-Baril (1987) reported a dramatic increase in disability of 14 times the rate of population growth from 1977 to 1981, further research has suggested that this disability rate relative to population growth continues to increase (Webster & Snook, 1994).

Although a precise estimate is impossible, it is plausible that the direct medical and indirect costs of these conditions are in the range of more then $50 billion per annum, and could be as high as $100 billion at the extreme (Frymoyer & Cats-Baril, 1991). Of these costs, 75% or more can be attributed to the 5% of people who become disabled temporarily or permanently from back pain. Further, it is estimated that the average worker's compensation claim results in costs of $4075 annually (Webster & Snook, 1994). Webster and Snook reported the mean cost per case of low back pain was $8321,








more than twice the amount for the average worker's compensation claim ($4075). Further the median cost per low back pain case was $396, an indication that low back pain costs are not normally distributed with a few cases accounting for most of the costs.

Clinical Assessment of Force

As health care providers in the current managed care environment attempt to care for patients with musculoskeletal disorders in a more cost effective manner, it is becoming increasingly important to identify which diagnostic and therapeutic procedures have the greatest impact on the cost of caring for patients with low back pain (Liu & Byrne, 1995). The measurement of applied force produced by voluntary muscular contraction across a range of motion is a common method used to assess the status of the musculoskeletal system. Strength testing is an important part of an assessment battery for evaluation of patients with chronic low back pain. A number of testing methods have been investigated including isotonic, isokinetic and isometric testing protocols. A brief description of these testing approaches is provided.

Isotonic, Isokinetic, and Isometric Testing

Various forms of musculoskeletal exercise are available for the researcher to use when assessing muscle strength. Today, three primary types of exercise are commonly employed- isometrics, isotonics, and isokinetics (Davies, Wilk, & Ellenbecker, 1997). While isotonic methodology will be used in the present study, a considerable amount of the research has involved the two other testing approaches. As a result, definitions of all three approaches follow:








Isometric Exercise: (Fixed Speed: Immovable, Fixed Resistance)

Isometric or static resistance exercises normally are performed against an

immovable object such as a wall, a barbell, or weight machine that is loaded beyond the maximal concentric strength of the individual (Sanders, 1997). This type of exercise or muscle contraction occurs when a muscle attempts to shorten but is unable to overcome the resistance (Albert, 1991).

Isotonic Exercise: (Variable Speed: Fixed Resistance)

This exercise is dynamic and is traditionally defined as a muscular contraction in which the muscle exerts a constant tension (velocity varies). Another way to think about isotonic exercise is that it is muscle movement with a constant load (Sanders, 1997). Isotonic exercise is divided into concentric and eccentric muscle loading. The individual muscle fibers actually shorten with a concentric contraction. In an eccentric action, muscle fibers actually go through a lengthening process (Davies et al., 1997). Isokinetic Exercise: (Fixed Speed: Accommodating Resistance)

Isokinetic exercise can be defined as exercise with a constant maximal speed and variable load (Sanders, 1997). In isokinetic exercise, the muscular contraction is performed at a constant maximal angular limb velocity, and the equipment varies the resistance. Isokinetics have been used extensively for muscle performance evaluation and objective documentation and for rehabilitation. The exercise is dynamic and most isokinetic machines provide both concentric and eccentric exercise options.

Measurement: Use of the Coefficient of Variation

Having discussed the different strength testing methodologies, it is now necessary to address the measurement approaches used in determination of effort level.








Hazard, Reid, Fenwick, and Reeves (1988) have examined the hypothesis that force/distance curve variability distinguishes submaximal from maximal efforts in isokinetic trunk and lifting strength tests. Thirty normal subjects were tested on the Cybex trunk extension/flexion (TEF) and Lift Task (LT) machines during maximal (100%) and submaximal efforts (50%). Considering each test separately, visual assessments of curve variability were indeterminate in 28% of TEF and 24% of LT tasks. Measurement models of curve variability were more clearly discriminating. Hazard et al. concluded that estimation of the degree of effort a given individual applies to a test of strength is essential to the clinical interpretation of that person's performance. While they assert that clinical judgment is required in evaluating effort during tests of isokinetic trunk and lifting strength, minimal attention to the need for more objective assessment measures was discussed. Further, this study points out a number of problems in the clinical application of the principle of rating degree of effort according to curve variability. First, if an individual produces curves of insufficient amplitude to allow assessment of variability, no effort rating can be made. Second, the visual technique used is subjective and inherently produces intermediate or debatable degrees of curve variability. Third, the correlation between curve variability and degree of effort is not perfect. It is possible for an inexperienced subject to produce consistent curves inadvertently while exerting a submaximal effort. Conversely, highly motivated subjects operating at maximal effort can produce variable curves.

Another method of measuring variability is the Coefficient of Variation (CV)

which has been used in numerous studies as a measure of subject effort (Carlsoo, 1986; Robinson, O'Connor, Riley, Kvaal, & Shirley, 1994; Simonsen, 1995; Lin, Robinson






5

Carlos, & O'Connor, 1996). It is a measure of relative variability, calculated as the ratio of standard deviation divided by the mean multiplied by 100 to yield a unitless percentage. It therefore expresses the standard deviation as a proportion of the mean, thereby accounting for differences in the magnitude of the mean. The coefficient of variation is most meaningful when comparing two distributions (Portney & Watkins, 1993).

The CV has been used as a measure of variability in strength measurements and it has been suggested that it might be useful for determining whether individuals are giving their maximal voluntary efforts. It is often assumed that a high CV indicates a lack of consistent effort; however, CV norms have not been established. Simonsen (1995) points out that if the CV can be used to validate an individual's effort, we must know how high a CV can be before we judge it as unacceptable. Agre and colleagues (1987) have reported that the CV from isometric muscle tests range from 5.1% to 8.3% in the upper arm and from 11.3% to 17.8% in the lower extremity. Bohannon and Smith (1988) examined isokinetic knee extension in healthy subjects. The CV for four trials for maximal and submaximal efforts were compared. Although on average, the submaximal efforts yielded higher CVs (mean 13.3%) than the maximal effort (mean 6.0%), no clear cutoff was identified above and below which performance could be dichotomized into the submaximal and maximal groups based on validity. Recent research (Lin et al., 1996; Robinson et al., 1994; Robinson, Sadler, O'Connor, & Riley, 1997) has indicated that use of a 10% CV cutoff can detect more submaximal efforts than any other cutoff. Harber and Soohoo (1984) suggest that there is no established validity in using the CV to discover malingering. However, the fact that the CV varies with different tests suggests that an absolute cutoff score may not be valid for all tests, even if one could be agreed on. A








relative cutoff that varies from test to test makes more sense. Bohannon (1987) mentions the need for norms to be established for a specific test before a cutoff score can be determined and judgments made regarding subjects' effort. Research continues to address the utility of using the CV in detection of submaximal efforts. Having provided a brief explanation of both the testing and measurement approaches in the assessment of musculoskeletal injury, the discussion now addresses how the research has responded to the question of how to best assess effort level.

Statement of the Problem

Given the review of the costs of musculoskeletal injury, the need for reliable

assessment of sincerity of effort in back strength testing is emphasized. Further, a factor contributing to lack of reliability in such testing is the potential for patients to be motivated to appear more disabled than they are. Patients with musculoskeletal injury may have incentive to produce less than maximal effort given concern for future injury, exacerbation of current pain, apprehension in a novel test environment as well as issues related to compensation (Mendelson, 1995). The degree to which "insincere" effort in testing contributes to the reported costs is difficult to determine.

In the testing of injured individuals, evaluation of sincerity of effort is an important factor in determination of functional capacity. The inability to accurately assess a patient's effort level can lead to procedures that are both unnecessary and ineffective (Robinson, O'Connor, Riley, Kvaal, & Shirley, 1994). Further, the failure to accurately determine a patient's sincerity of effort has implications regarding the lack of apparent response to treatment, as well as escalating health care and disability costs. In addition, failure to








accurately detect patient effort can create an atmosphere where suspicion dominates, which may then generalize to the genuinely disabled.

Sincerity of Effort

One approach to the assessment of sincerity of effort in both research and

practice has been based on the hypothesis that sincerity of effort may be determined by evaluating the consistency of individuals' responses to various strength testing tasks. According to this hypothesis, the performance of an individual who is sincerely attempting maximal exertion will be influenced only by physical limitations associated with his or her injury or impairment and the influence of these limitations will remain relatively consistent across trials. However, an individual whose performance is limited not only by physical impairment but by attempts to give less than full effort will not be able to maintain as consistent a level of performance, since other than physical factors are limiting his/her performance.

Previous research has proposed that when a patient is deliberately attempting to produce a submaximal effort, the obtained strength measurements will show increased variability and reduced reliability (Kishino et al., 1985). The assumption is that subjects cannot willfully reproduce a submaximal effort with the same consistency as a maximal effort. Although Kishino et al. (1985) speculated about this hypothesis, they offered no data to support it. Bohannon (1987) compared the variability of maximal and submaximal tests of isometric elbow flexion by calculating the coefficient of variation (CV). He reported that the CV of force measures was significantly greater during submaximal trials than during maximal trials, which supported the Kishino et al. hypothesis.








Further evidence supporting the "sincerity of effort" hypothesis is provided by other investigators who have also suggested that the variability of force production in injured patients undergoing strength assessment may be an indicator for submaximal effort (Gilbert & Knowlton, 1983; Chengalur, Smith, Nelson, & Sadoff, 1990; Niebuhr & Marion, 1990). Gilbert and Knowlton (1983) proposed that individuals may use different motor strategies in maximal and submaximal efforts. They propose that a maximal effort is a lower order motor task in which feedback from the muscle contraction is not crucial. On the other hand, a submaximal effort is a higher order motor task in which feedback is crucial in limiting the force of the contraction. To limit force, a submaximal task would require a greater degree of feedback from the muscles than would a maximal effort. An implication is that differing motor strategies would make feigned submaximal efforts detectable. While the sincerity of effort hypothesis shows promise, the validity of testing methods requires more empirical examination.
Static vs. Dynamic Testing

While the assessment of sincerity of effort is now frequently used in clinical

practice, employing both static (isometric) and dynamic (isokinetic and isotonic) testing approaches, the reliability and validity of current assessment methods have not been adequately demonstrated (Mooney & Anderson, 1994). The following review will initially consider the research using a static assessment approach. As the discussion proceeds, the rationale for the move to a dynamic testing protocol will be provided as will a review of the literature in this area.








Static Testing

Clinical practice has generally studied variability in effort level in isometric tasks. A number of studies have used the grip test strength assessment task in determination of effort level (Gilbert & Knowlton, 1983; Bohannon, 1987; Smith, Nelson, Sadoff, & Sadoff, 1989; Chengalur et al. 1990; Niebuhr & Marion, 1990; Robinson, Geisser, Hanson, & O'Connor, 1993). Robinson et al. (1993) used the CV to discriminate level of effort in grip strength testing. Twenty-nine asymptomatic subjects participated in two conditions of testing: 100% and 50% effort. The submaximal (50%) effort condition showed significantly more variability than the maximal effort condition in both sets of conditions. Intra-class correlation coefficients were very low for both maximal effort and submaximal efforts (.036 and .025) indicating very low stability for the coefficient of variation. Classification rates were also found to have unacceptably large errors with 69% of the submaximal efforts being classified as maximal with the traditional 15% cutoff, and 55% misclassification of submaximal efforts with an optimized 11% cutoff It was concluded that the currently practiced method of using a low number of repetitions to calculate the CV might result in very unstable measures. Further, grip strength testing, which is an isometric testing procedure, did not produce consistent results (suitable detection rates) that might validate its use in a clinical setting. Although its use continues in clinical practice the research does not adequately support such use.

While grip strength testing may provide helpful information, when we consider the costs and prevalence of musculoskeletal injury it is important to think about the relevance of sincerity of effort testing from a more ecologically valid perspective, that is, the assessment of the spinal region. Because of its significant economic impact, assessment of








lumbar spine disorders in workers and patients has become widespread. These tests are used in the prediction of back injury (Batti'e et al., 1989) and assessing the functional capacity of workers (Mayer et al., 1988). Although the methods of recording force production may differ, with research at this point having focused primarily on isometric and isokinetic testing, all methods have assumed that subjects giving less than the instructed full, maximal effort would be identified by their inconsistent recordings. However, despite the medicolegal implications of this assumption, there is insufficient data to support the hypothesis that subjects cannot willfully reproduce a submaximal effort with the same consistency as a maximal effort.

Robinson, MacMillan, O'Connor, Fuller and Cassisi (1991) conducted a study investigating the differences in test-retest reliability between maximal and submaximal efforts in an isometric lumbar extension task. They also tested the hypothesis that submaximal efforts would be less consistent than maximal efforts. Twenty subjects were asked to produce maximal voluntary contractions at seven different positions in a lumbar extension machine. Each subject was tested twice in a maximal effort condition and twice in a 50% effort condition. Results indicated high test-retest reliability at all angles in both conditions. There were no differences in test-retest reliability between effort conditions. Therapist ratings of consistency did not differ between conditions, and therapists could not discriminate between conditions on the basis of effort consistency. These authors concluded that claims that isometric strength testing could be used to determine sincerity of effort or "malingering" might be unfounded.








It is clear from these results that individuals are able to consistently show

"weakness" on an isometric strength test. This is likely to be the intent of an individual who is exaggerating or faking a low back injury. Although it was possible to show strength differences between maximal and submaximal effort conditions within a given subject, it was of little utility to attempt such comparisons across subjects. Previous research (Mayer, Smith, Keley, & Mooney, 1985; McIntyre, Glover, Seeds & Levene, 1990) has suggested that CLBP patients produce less torque from the lumbar musculature than normals do. However, in a population that may have incentives to show weakness, there are no data to suggest that those individuals giving a submaximal effort may be differentiated from those giving a maximal effort on the basis of strength measures alone.

In summary, given the reported findings there is insufficient evidence to suggest that reliable classification of effort level can be determined using isometric strength tasks. Further, despite the need for assessment of sincerity of effort in backs, a review of the literature indicates a relative paucity of research specifically addressing this area. As a result, the continued discussion of the transition from a static to a dynamic testing approach involves primarily a review of sincerity of effort in leg extension testing. The Move to Dynamic Testing

Our discussion so far has focused on the more static isometric assessment measure. However, while the use of isometric tasks appears to be of increasingly questionable value, there has been relatively little investigation of the potential of other strength assessment tasks. While this section will focus primarily on dynamic strength assessment, some of the research has included an isometric task in a continued effort to investigate and compare this methodology with dynamic tasks.








In the Hazard et al. (1988) isokinetic TEF and LT study, previously discussed in the context of the sincerity of effort hypothesis, the authors reported modest ability to correctly classify degree of effort with simple visual inspection of the force curves. However, using discriminant analysis techniques, the authors were able to improve the accuracy of classification to 80-83% for the trunk extension task and 75-82% for the lift task.

Although the results reported in their study were promising, Hazard et al. (1988) pointed out that some subjects were able to produce consistent submaximal curves and in contrast others produced inconsistent maximal effort curves, suggesting there is considerable between subject variability. In addition, these authors noted that clinical observation during the study was more accurate than the measures of curve variability. Finally, the absence of uniform criteria for strength variability assessment may contribute to the controversy in some of the studies mentioned above.

A recent study using the coefficient of variation measure compared the variability of torque production and velocity in both isometric and isotonic leg extension tasks in both maximal and submaximal effort conditions (Robinson et al., 1994). Fifteen asymptomatic subjects participated in a within subject counterbalanced design in which they were asked to perform maximally and submaximally in both isometric and isotonic leg extension tasks. Results indicated that both isometric and isotonic tasks showed greater variability (measured by coefficient of variation) in the submaximal effort condition. The ability to detect submaximal efforts without misclassification of any maximal efforts was much greater for the isotonic task (87% of efforts were correctly classified). While the isometric task had unacceptably poor classification rates (60% of efforts were correctly






13

classified), consistent with previous research, those of the isotonic task were high enough to show promise for clinical application. Further, in the isotonic task there were no false positives (maximal efforts called submaximal) and only 4 of 15 submaximal efforts were missed using a 10% CV cutoff This represents a significant improvement over the isometric testing method and previous research (Robinson et al., 1991), and provides support for the use of dynamic testing protocols.

Another recent study has investigated differences between maximal and

submaximal efforts in isometric and isokinetic knee extension using torque variability and slope measures (Lin et al., 1996). These authors note that the use of torque variability and slope measures to detect submaximal efforts has been studied in isometric tests, but not fully investigated in isokinetic tests. Thirty-two healthy students were studied to determine the extension torque of the knee during isometric and isokinetic exercise. The CV of average torque, CV of peak torque, and slope to peak torque were obtained from maximal and submaximal torque curves during isometric and two isokinetic tests (60o/s and 1800/s). Subjects were instructed to kick as hard as they could in the maximal effort condition and were told to give a 50% effort in the submaximal effort condition. The testing order of maximal vs. submaximal conditions was randomized across subjects to eliminate a possible order effect caused by fatigue.

Results of this study, which were consistent with previous findings (Robinson et al., 1993, 1994; Smith et al., 1989), indicated significant differences between effort levels (maximal and submaximal) for all variables in isometric and isokinetic tests. However, in the isometric condition, effort level classification based on the CV method was found to have unacceptably large errors, in that over 50% of submaximal efforts were misclassified






14

with an optimized (no false positives) CV cutoff With regard to the isokinetic condition, Lin et al. (1996) found that a 10% cutoff of CVpeak correctly classified 59% of submaximal efforts in fast isokinetic tests (1800/s) for an overall correct classification rate of 80% without any false positives. Using a 10% CVavg cutoff, 75% of submaximal efforts were detected for an overall correct classification rate of 88%. These results suggest that the isokinetic task may represent an approach that improves upon the discrimination of submaximal from maximal efforts. This appears to confirm the view of Robinson et al. (1994) in isometric and isotonic knee extension tests that the sensitivity to detect submaximal efforts was much greater for the dynamic condition. It seems that movement is the key difference between methods that better discriminate (isokinetic and isotonic), and methods that are less successful (isometric). As suggested, it may be that subjects are less able to reliably reproduce torque and velocity of submaximal efforts because of the dynamic nature of the task.

In a follow-up study, focusing on a dynamic strength assessment task, Robinson et al. (1997) investigated the stability of the coefficient of variation over time in asymptomatic normals using an isokinetic leg extension strength assessment task. Previous research (Robinson, Geisser, Hanson & O'Connor, 1993) had suggested that the CV for isometric strength testing might be unstable because of the usual method of its calculation based on a small number of trials (typically 3). Although significant classification of asymptomatic normals to effort condition was obtained in two recent studies (Lin et al., 1996; Robinson et al., 1994), instability of the CV would make replication of the results more difficult and would jeopardize the clinical utility of the method. This more recent study calculated the CV with a larger number of repetitions and








used a test-retest protocol. It was hypothesized that calculation of the CV based on 5 repetitions would be more stable over time than that based on 3 repetitions. Further, it was hypothesized that submaximal efforts would be more variable than maximal efforts.

Twenty asymptomatic subjects participated in isokinetic leg extension strength

testing. The coefficient of variation of peak torque (CVpeak) and coefficient of variation of average torque (CVavg) were obtained from maximal and submaximal torque curves during fast isokinetic (180*/s) leg extension strength tests. Subjects participated in two conditions of testing: 100% effort and 50% effort. Each subject completed 6 knee extensions at 100% effort and 6 knee extensions at 50% effort on Day I of testing. This was repeated at least 48 hours later.

The results of this study replicated previous research (Lin et al., 1996), providing evidence that submaximal efforts as measured using isokinetic leg extension testing were significantly more variable than maximal efforts. However, effort level classification, although improved compared to the use of isometric assessment techniques, resulted in misclassification of submaximal efforts. Using the clinically accepted cutoff of 15%, a large number of submaximal efforts were misclassified. However, adjusting the cutoffto an optimized 10%, resulted in a significant reduction in misclassification rate with no false positives. CVpeak (5 trials) resulted in a detection rate of 75% for an overall classification rate of 88% on Day 1 of testing. Lin et al. (1996) also found that a 10% cutoff detected more submaximal efforts than any other cutoff greater than it, with no false positives. However, they used the CVave measure, obtaining a detection rate of 75% of submaximal effort for an overall classification rate of 88% with no false positives. Using the CVpeak (4 trials) measure, they detected 59%/ of submaximal efforts for an overall classification








rate of 80%. The more recent study represents an improvement in CVpeak classification rates compared to the study by Lin et al. This CVpeak measure was used in the present study.

In addition, based on these recent results, it appears that a lower CV cutoff(10%) than the standard 15% cutoff might be warranted in the detection of submaximal efforts. Determination of classification stability also indicated that a 10% CV cutoff appears to be the optimal detection rate at which no false positives were obtained.

Concerns regarding the stability of using a low number of repetitions (usually three) to calculate the CV were also addressed in the study by Robinson et al. (1997). The calculation of this measure on such a small number of trials raises the possibility that the measure may be very unstable. Results indicated that use of 5 repetitions versus 3 repetitions in the calculation of CVpeak resulted in increased stability across trials. Results indicated that the CVave measure (calculated on 3 trials only) was unstable. The trend in these findings suggest that increasing the number of repetitions used in CV calculation may prove beneficial.

The previous research discussed suggests that use of a dynamic methodology

appears to provide better discrimination of effort level. The following discussion provides a helpful analysis and summary of the dynamics proposition. A Final Comment on Dynamics

Research has suggested that the specific relationship between the force produced and force variability has been proposed to be linear (Schmidt, Zelaznik, Hawkins, Frank,& Zuinn, 1979), an inverted-U (Sherwood & Schmidt, 1980), or a curvilinear function (Sherwood, Schmidt, & Walter, 1988), with variability generally decreasing as function of








the force applied. These functions have also been found in studies using dynamic tasks (Sherwood et al., 1988; Carlton & Newell, 1988). These tasks are thought to be inherently more complex because of increased changes involving a number of variables including muscle length, muscle velocity, and limb displacement. In their review, Newell, Carlton and Hancock (1984) found that many functions of variability in motor control appear, depending on the task-related relationships of movement distance, time to peak force, magnitude of the force applied, and rate of force production. Newell and Carlton (1988) noted that separating the effects of these variables has been difficult as a result of the change in one variable leading to changes in one of the other variables. These effects have been suggested to be accounted for peripherally by factors such as proprioceptive feedback, the nature of the muscle contraction mechanism, information transmission ability of the peripheral anatomical unit as well as centrally by noise in the neural control mechanism (Newell et al., 1984). Given these factors, it is likely that the increased complexity of the dynamic tasks results in a greater variability from repetition to repetition, making these tasks more susceptible to manipulation. In addition, attempting submaximal efforts leads to an even greater complexity of the task, with a resulting increase in the likelihood of error and variability from trial to trial. This complexity is likely to make dynamic tasks the better choice for detecting submaximal efforts.

The Choice of Isotonic Testing

Having now provided a rationale for a dynamic testing approach, the discussion now turns to consideration of the use of an isotonic testing procedure. A significant proportion of the research in the area concerning sincerity of effort has involved use of






18

isometric and isokinetic methodology. The evaluation of sincerity of effort using isotonic testing methods has not been addressed to any significant extent.

Even within the more extensively researched isometric and isokinetic

methodology, research has focused primarily on assessment of strength with less attention to detection of sincerity of effort. However, it is important to acknowledge that that Dvir and colleagues have employed isokinetic testing methodology and reported promising results with respect to detection of sub-optimal performance (Dvir, 1997a, 1997b; David, Dvir, Mackintosh, & Brien, 1996; Dvir & David, 1996).

In a review of the literature related to trunk muscle performance, Beimborn and Morrisey (1988) discussed aspects such as differences in trunk extension force versus trunk flexion force. They noted that there was great variation in methods, procedures, and equipment, as well as type and speed of contraction used in these studies. In addition, the subjects and their physical condition, weight, sex, age, and height differed in each study. Due to the extreme variability in methodology and subject samples, it was very difficult to assemble all of the data and have an accurate set of normative values. Despite these difficulties, an attempt was made to present a compilation of all the work and to categorize the many studies according to sample type, equipment used, motions tested, and the mode of contraction evaluated.

Their review focused on studies investigating torque producing capacity in the trunk musculature. The studies were classified into three categories: 1) isometric, 2) isotonic and 3) isokinetic analysis. In this review they reported twenty-five studies employing isometric testing, and seventeen studies employing isokinetic testing methods. Providing further evidence of the lack of research involving use of isotonic testing






19

methods, they reported only five older studies using this methodology (Flint, 1955; Kluck, 1967; Mayer & Greenberg, 1942; Nachemson, & Lindh, 1969). Considering the paucity of research investigating isotonic methodology in assessment of trunk musculature torque producing capacity, it is not surprising that there are even fewer studies investigating the specific area of sincerity of effort. In fact, an extensive search of the literature considering isotonic methodology indicates that the majority of this research has used this paradigm in evaluation of rehabilitation/exercise treatment programs for medical patients with back problems (Greenleaf, Bernauer, Ertl, Trowbridge, & Wade, 1989; Holmes et al., 1996) and heart disease (Hammond & Froelicher, 1985; Cantor et al., 1987).

However, Robinson, Cassisi, O'Connor & MacMillan (1992), in a study

investigating lumbar iEMG during isotonic exercise of the lumbar spine, have suggested that while not a direct comparison of the clinical efficacy of isotonic versus isokinetic testing, their study establishes that clinical isotonic testing can be used in a fashion similar to clinical isokinetic testing. Further, Robinson et al. (1994) suggest that using the isotonic method allows for the investigation of differences in velocity variability between maximal and submaximal efforts. It may be that subjects are less able to reliably reproduce submaximal velocity efforts than they are able to reproduce submaximal strength or torque efforts because of the dynamic nature of the task (Robinson et al., 1994). In addition, the use of dynamic tasks may better represent the normal use of the joint system, since isometric muscle contractions do not represent a high frequency behavior.








Given the lack of research involving use of isotonic methodology in determining sincerity of effort, as well as a similar paucity of research assessing this concept in the spine musculature, a recent pilot study was completed in our lab at the University of Florida.

Sadler, Robinson, Otis, O'Connor and Riley (unpublished data) assessed lumbar strength variability in 20 asymptomatic controls, using the Dynatrac Back Tester (BTE Dynatrac, Hanover, Maryland). Participants were asked to complete 6 maximal and 6 submaximal isotonic lumbar extension exercises with a 10 minute break between trials. Order of condition was counterbalanced. A resistance level of 60% of participants' body weight was set for each subject.

While a significant difference was found between maximal and submaximal efforts with regard to velocity, there were no significant differences found in the CV between effort conditions. Accurate classification to effort condition was not possible. These results clearly did not support recent findings using isokinetic leg extension testing (Robinson et al., 1997). It was speculated that perhaps the anatomy of the spine and its musculature, might somehow result in less variability being observed in back strength testing.

As a result of the poor discriminability found in this recent study, a follow-up pilot study was conducted. It was observed at the time of testing in the previous study, that a number of subjects were experiencing some discomfort with the procedure. As a result, it was speculated that a lowered resistance level might provide a better opportunity to discriminate maximal from submaximal efforts. Eight subjects participated in the pilot study with the same methodology; however, resistance level was set at 25% of body






21

weight. Results indicated that first, participants were more accurate in their attempt at a 50% effort compared to the previous study using a 60% resistance level. In addition, as in the previous study using 20 subjects, there was a significant difference in the velocities produced between maximal and submaximal effort conditions. Finally, results indicated that there was a clear trend toward significant differences (p<.06) between CV of maximal and submaximal effort conditions (d=.80).

The Present Study

The above literature review suggests that the dynamic isotonic testing method appears to show the most promise. Therefore this study considered detection of submaximal effort in an isotonic back strength test. Given the results of the more recent studies using different resistance levels, an attempt was made to determine an optimal resistance level that might lead to improved detection of submaximal effort. In addition, as a result of trends in recent findings, suggesting somewhat increased stability, a larger number of repetitions (6) was used in the determination of the CV. Previous results had raised doubt about the stability of using a low number of repetitions (usually three) to calculate the CV. The calculation of this measure on such a small number of trials raised the possibility that the measure might be very unstable.

First, it was hypothesized that submaximal efforts would be more variable than maximal efforts across resistance levels. It was also hypothesized that a resistance level lower than the 60% level previously addressed, would result in increased discriminability between effort level. Finally, it was hypothesized that use of the CVpeak 6 trial measure would provide better discriminability between effort level than the CVpeak 3 trial measure






22

given trends in recent findings suggesting moderate increases in stability with increased number of repetitions used in CV calculation.













CHAPTER 2
METHOD

Participants

Sixty-six asymptomatic normals were recruited from the psychology department and physical therapy department and asked to participate in isotonic back extension strength testing. These participants were divided into three groups of twenty-two subjects. Results of a power analysis indicated that this would be the appropriate number of subjects per group, given the pilot study results with an effect size of .8. An even number of males and females were recruited. Exclusion criteria consisted of any history of neuromuscular disease or any injury to the lower extremity that has required medical or surgical intervention. Informed consent was obtained from the participants in accordance with the Institutional Review Board.

Apparatus

Back extension assessment was measured by the Baltimore Therapeutic Equipment Co. Dynatrac Back Tester (BTE DynatracTM, Hanover, MD). Research investigating the reliability and validity of the BTE Dynatrac' Back Tester has suggested that the Dynatrac can be considered both reliable and objective (Richards, Bailey, & Castasqno, 1996). The Dynatrac computer provides test reports in which both concentric (as in lifting) and eccentric (as in lowering) strength is exercised and analyzed. While in the back tester, the patient remains in one stabilized position for flexion/extension and rotation. Any size patient can be accommodated.






24

This apparatus allows for isometric assessment as well as isotonic testing at a preset resistance. The dynamometer can be locked for isometric testing or provide a specific resistance to be maintained throughout movement in the case of isotonic strength testing. With Dynatrac isotonic loading, the muscular force the patient must exert during a test is prescribed and set by the experimenter. This is in contrast to isokinetic loading, where the patient is told to exert a maximal force throughout his/her range of motion. The Dynatrac isotonic load can be set by the experimenter to safely suit the patient's capability, from virtually zero to high performance torque levels. Isotonic strength assessment was conducted in this study.

Procedure

Participants were asked to repeat a back extension exercise in the Dynatrac Back Tester. Participants were divided into three groups of twenty-two subjects. Three resistance levels were measured: 15% of body weight, 25% of body weight and 35% of body weight. These values were chosen based on results from the pilot study which showed a moderate effect size for weights in this range. Demographic information including, height, weight and gender was collected.

A graduate student previously trained by a physical therapist and blind to order of condition conducted the testing. On both trials, the graduate student conducted testing consistent with maximal effort instructions (Ready! Go! As hard and as fast as you can!); however, participants were told by the experimenter assistant to give a 100% effort in the maximal effort condition and were told to give a 50% effort in the submaximal condition. The assistant informed participants of order of condition.






25

First, the participants were secured to the Dynatrac back testing device. Patients were placed in a comfortable, semi-kneeling posture with the pelvis at a 35 degree forward tilt. This placed the spine in a neutral position, maintaining the lumbar lordosis and the normal angle of the sacrum. The semi-kneeling posture allowed full unrestricted flexion and extension. Simple adjustments accommodated any size participant, tall or short. Padded fixation points were adjusted for variations in limb length, pelvis height, pelvis width, and shoulder height, while isolating the back muscles for testing. With this patient positioning system, test subjects were easily positioned in less than a minute with solid stabilization.

Once secured in the back tester, participants were told that 3 practice repetitions would be performed to ensure understanding of the protocol. This was followed by a one minute rest period. Participants were then asked to perform 7 repetitions at a 50% effort level (submaximal) and 7 repetitions at a 100% effort level. Order of conditions was counterbalanced and separated by a 10 minute rest period. Finally, upon completion of the trials subjects were required to indicate the order of effort level that they were assigned.













CHAPTER 3
RESULTS

Demographics

There were 66 participants in the study (see Table 1 for breakdown). The mean

age of participants was 26.2 years (SD=6.4); mean height was 171.9 cm (SD=9.9cm); and mean weight was 67.4 kg. (SD=15.7kg). A manipulation check which required subjects to state the order of effort conditions indicated that all participants followed the assigned protocol.


Table 1
Number of Participants by Group

Group 1 (15%) Group 2 (25%) Group 3 (35%) Total Males 9 9 10 28 Females 13 13 12 38 N=66


Coefficient of Variation

The CVpeak velocity for 6 repetitions was calculated using repetitions 2-7. The CVpeak velocity for 3 repetitions was calculated using repetitions 2-4. Trial I was not used in the calculations as this was not considered to be an accurate record of effort level given the possibility that acclimation to the protocol was necessary.








Submaximal vs. Maximal Effort Level

Mean velocity scores were calculated across repetitions for each group and in each condition using both 3 and 6 repetitions. These scores were subjected to paired t-tests to assess differences between conditions. Results indicated that the submaximal effort yielded significantly lower velocity than the maximal effort across groups as well as number of repetitions used in calculation (see Table 2). Table 2
Mean Velocity Differences between Test Conditions Group Maximal Submaximal t(21) 15%-6 reps 221. ld/s 127.2d/s 10.2 .001 15%-3reps 206.3d/s 126.9d/s 18.8 .001 25%-6reps 225.8d/s 140.6d/s 12.4 .001 25%-3reps 211.2d/s 136.5d/s 11.5 .001 35%-6reps 219.4d/s 124.3d/s 14.5 .001 35%-3reps 202.9d/s 121.4d/s 12.5 .001 Total(66)-6rep 222.1d/s 130.7d/s 20.9 .001 Total(66)-3rep 206.8d/s 126.9d/s 18.8 .001



In addition, the mean CVpeak was compared between 3 and 6 repetitions for each effort level. Results indicated that the CV based on 6 repetitions was significantly higher for both maximal and submaximal efforts when collapsed across groups (Maximal: t(65)=2.94; p<.005: Submaximal: t(65)=3.22; p<.005) ( see Table 3).

Analysis of Variance

The mean CVpeak between effort levels and across resistance levels are presented in Figure 1(based on 6 reps). Similar findings were evident for the calculation of the CV based on 3 repetitions. Table 4 specifically provides the means and standard deviations for









Table 3
Mean CV Differences between 3 vs 6 repetitions by Group

Maximal Effort Grp 1 (15%) Grp 2 (25%) Maximal Effort (3) 7.7 4.4 9.7 5.5 Maximal Effort (6) 10.7 3.8 10.5 4.3 Submaximal Effort Grp 1 (15%) Grp 2 (25%) Submaximal Effort (3) 8.7 . 4.5 8.43.9 Submaximal Effort (6) 10.2 . 4.8 10 4.4

*p<.005


Grp 3 (35%) 11.1 5.1 11.6 3.8 Grp 3 (35%)

9.2 6 10.4 3.9


All Groups

9.5 5.1* 10.8 3 .9

All Groups

8.8 5 10.4 4.2*


the CVpeak (6 reps) directly comparing maximal and submaximal effort levels. Paired t-tests indicated that there were no significant differences for either the CVpeak based on

6 or 3 repetitions with respect to the ability to discriminate effort levels within each resistance level (see Table 4 for 6rep. t-test results).


Table 4
CVpeak: Means and Standard Deviations across Groups

Grp 1(15%)' Grp 2(25%)' Grp 3(35%)3 All Groups4 Maximal Effort 10.73.8 10.54.3 11.63.8 10.83.9 Submaximal Effort 10.24.8 104.4 10.4 3.9 10.44.2 I t(21)= -.38, p<.707; 2t(21)=.57, p<.574; 't(21)=1.11; p<.279; 4t(21)=.66,p <.513


A two-way mixed repeated measures analysis of variance was used to examine main effects across resistance and effort level as well as an effort (maximal and submaximal) by condition (15%, 25%, and 35% resistance levels) interaction for the CV velocity measure (see Table 5). There were no significant effects found for within (effort level) (F(2,63) = .43, p < .516; F(2,63)= .64, p < .531) and between (resistance level) group(F(2,63) = .32, p < .725) factors. Sex differences in the CV were also examined in an








analysis of variance. There were no significant effects found for within (effort level) (Fo(I64) = .58, p< .449; F ,") = .64, p< .457) and between (sex) group (F(,4) = 1.92; p<.171) (see Table 6).


Table 5
Analvsis of Variance: CV


Differences/Group by Effort L


Source df SS MS F p Between Subject Effects
Within & Residual 63 .12 .00 Group 2 .00 .00 .32 .725 Within Subjects Effects
Within & Residual 63 .09 .00 Effort 1 .00 .00 .43 .516 Group by Effort 2 .00 .00 .64 .531


Table 6
Analysis of Variance: CV Differences/Sex By Effort Level

Source df SS MS F p Between Subject Effects
Within & Residual 64 .12 .00 Sex 1 .00 .00 1.92 .171 Within Subjects Effects
Within & Residual 64 .09 .00 Effort 1 .00 .00 .58 .449 Sex by Effort 1 .00 .00 .64 .457


evel


.d


An. vsi of.V.r.a.c.....











12 +4.8 +3.8 +4.4 +4.3


S50%
CV 0 100%




0
Group Group2 Group3
15% 25% 35%


Figure 1. CV Peak Across Groups













CHAPTER 4
DISCUSSION

As with the previous studies in the sincerity of effort program of research that have been conducted in our lab, results of the present study indicated that there was a significant difference between aggregate maximal and submaximal effort data. Participants produced a submaximal effort at a significantly lower velocity than a maximal effort. On average, the submaximal effort was approximately 57- 60% of the maximal effort for calculations based on both 3 and 6 repetitions. These findings are consistent with previous research in our lab. (Robinson et al. 1994; Lin et al., 1996; Robinson et al., 1997)

Results did not support the hypothesis that the submaximal effort condition would be more variable than the maximal effort condition or that a resistance level lower than the 60% level previously addressed would result in increased discriminability between effort level. Across the three resistance levels used in this study, we could not discriminate between effort levels based on the proposed coefficient of variation measure. Previous research had suggested that a high CV indicated a lack of consistent effort (Carlsoo, 1986; Robinson et al., 1994; Lin et al., 1996; Robinson et al., 1997). As a result, we expected to observe a significantly higher CV in our submaximal effort condition than in our maximal effort condition. We did not find a significant difference between effort levels. Further, our results indicated that the maximal effort condition across resistance levels produced a slightly higher CV. This was in contrast to the results found by Robinson et al. (1994), a study which prompted the use of isotonic strength






32

testing in the present study. In the study by Robinson et al., results indicated that isotonic strength testing showed promise in the detection of submaximal effort in leg extension testing. This study found that in addition to significant differences between aggregate maximal and submaximal effort data, much better sensitivity was achieved with an isotonic task. In fact, 87% of efforts were correctly classified with no false positives. Robinson et al. had proposed that using the isotonic method allowed for the investigation of differences in velocity variability between maximal and submaximal efforts. Again, the proposition was that perhaps participants were less able to reproduce submaximal velocity efforts than they were able to reproduce submaximal strength or torque efforts because of the dynamic nature of the task.

While the present study did not replicate the results found in the Robinson et al. isotonic leg extension study, previous research employing isokinetic testing methodology in leg extension testing which had found viable submaximal effort detection rates (Lin et al, 1996; Robinson et al., 1997) was also not supported. It is possible that the musculoskeletal differences between the lower extremity (a less stable structure) and the spinal region (a more stable structure) could have accounted for the differences in the ability to reproduce consistent/inconsistent effort. In addition, the nature in which participants were secured in the Dynatrac Back extension device compared to the stabilization of the leg in the previous studies is different from both a subjective and objective perspective. For example, in the back study, participants were more likely to comment on the extreme nature of the stabilization procedure. The extent to which this stabilization procedure played a role in reducing the likelihood of more variability in performance, although difficult to determine cannot be underestimated.








Further, although this study which employed a dynamic assessment protocol did not replicate the results found in the leg extension studies, the findings were consistent with previous research employing isometric testing methodology in a lumbar extension task (Robinson et al., 1991; 1992). In these isometric lumbar extension studies, results indicated that participants were able to reliably reproduce a submaximal effort making it difficult to discriminate between effort level. These earlier studies had set the precedent for pursuing a dynamic testing approach in assessment of sincerity of effort of the lumbar musculature.

In the present study there was no specific hypothesis attempting to determine

which of the three resistance levels would produce the best submaximal effort detection rate. Results of the present study did not indicate any significant differences between these resistance levels with respect to discriminability. In the pilot study, a 25% resistance level had resulted in a moderate effect size in which there was an 8%-12% difference between maximal and submaximal effort coefficient of variation respectively. We expected that we might find similar results in the current study or, perhaps a somewhat higher or lower resistance level might prove more efficacious in detection of submaximal efforts.

In addition, while the CV value based on 6 repetitions was higher than the CV based on 3 repetitions for both maximal and submaximal effort levels, there was no difference between the two measures with respect to discriminability of effort level. Further, the higher CV calculated for 6 repetitions may simply represent a fatigue effect because the latter repetitions (5, 6 and 7) were added in order to calculate the CV based on 6 repetitions. In a previous study by Robinson et al. (1997) a trend suggesting increased stability of the CV based on a greater number of repetitions was observed. In








the present study there was no indication that the CV calculation based on 3 versus 6 repetitions represented any significant difference with respect to discriminability of the measure. However, given the findings in a number of recent studies (Robinson et al., 1994; Lin et al., 1996; Robinson et al., 1997) we believe that the CV continues to represent a promising construct in determining effort level. Continued investigation of the measure with respect to both its ability to discriminate effort level and its stability is warranted.

While the results of the present study do not support the findings by Robinson et al using isotonic methodology or the previous research involving use of an isokinetic methodology we are left to speculate as to why we were unable to discriminate between effort level. An explanation can perhaps be found in Schmidt's general assertion that as we move more rapidly, we become more inaccurate in terms of the goal we are trying to achieve. Schmidt (1988) has proposed that the basic laws in motor behavior may be seen as analogous to the fundamental principles of physics. The simple laws relating the mass, velocity and acceleration of objects when forces are applied to them for example, have served as the cornerstone of the physical sciences, and hence they deserve a special status. Schmidt notes that in the same way, the field of motor behavior has analogous principles that are somehow fundamental to all the rest. These describe such things as the relationship between the speed a limb moves and its resulting accuracy. While there are a well defined set of simple principles that can be stated for the various branches of the physical sciences, we should not expect something similar for the behavioral sciences, or for motor control in particular (Schmidt, 1988). Schmidt states that for a number of reasons, in motor control we find far fewer statements possessing sufficient generality to






35

be termed a law. One reason is that motor-control principles have been far more difficult to discover, based as they are on data from biological systems that are more variable and complex than the physical systems.

The speed-accuracy relationship was first discussed by Woodsworth (1899) who found that accuracy decreased as the movement speed increased. Fitts (1954) provided a more formal mathematical relationship. Given that the results of the present study did not discriminate between effort level and in fact the increased variability however minor was observed in the maximal (higher speed) effort condition, this speed-accuracy relationship provides some rationale for the findings. While the isotonic leg extension study by Robinson et al. did provide promising submaximal effort detection rates, perhaps it was a function of lower extremity testing versus back extension testing.

Since this theory is based on the use of speed in motor control, perhaps it would prove beneficial if we applied the current methodology using an isokinetic testing paradigm in which the dependent variable would be the torque or force produced by the subject rather than the velocity. Future research employing an isokinetic protocol in investigation of sincerity of effort in back extension assessment is recommended. Dvir (1997) has reported promising findings in a study investigating differentiation of submaximal from maximal effort in an isokinetic trunk extension task. However, these researchers have discriminated between effort levels based on eccentric/concentric ratio differences, asserting that the highest differentiating power among the experimental conditions was attributed to the intervelocity difference between the concentric and eccentric contractions. Further with respect to measures used in discriminating effort level, Dvir and David (1996) note that consistency as defined by the absolute variability of








the test findings cannot serve as a reliable indicator for optimality of effort. In contrast, other research has provided promising results in detecting submaximal effort in both isokinetic and isotonic strength tasks using the CV as the sole discriminating variable. (Lin et al., 1996, Robinson et al., 1994; Robinson et al., 1997)

In summary, this study indicated that an isotonic assessment protocol like the isometric assessment strategies previously used were not effective in discriminating between maximal and submaximal effort levels. This may be due to both stabilization and musculoskeletal between the leg and back structures. While the CV measure did not indicate differences invariability between effort levels, continued investigation of this measurement technique is warranted.

Finally, the investigation of sincerity of effort in strength assessment testing remains an important issue with respect to issues of treatment, rehabilitation, and compensation. Continued research in this area is necessary in order that the reliability and validity of measurement techniques in this area of research can be improved.













REFERENCES


Agre, J., Magness, J., Hull, S., Wright, K., Baxter, T., & Patterson, R. (1987). Strength testing with a portable dynamometer: reliability of upper and lower extremities. Archives of Physical Medicine and Rehabilitation, 68, 454-458.

Albert, M. (1991). Eccentric muscle training in sports and orthopaedics. New York: Churchill Livingstone.

Batti'e, M., Bigos, S.J., Fisher, L.D., Hanson, T.H., Jones, M.E., & Ley, M.D.,
(1989). Isometric lifting strength as a predictor of individual back pain reports. Spine, 14, 851-856.

Beimborn, D.S., & Morrissey, M.C. (1988). A review of the literature related to trunk muscle performance. Spine, 13, 655-660.

Bohannon, R.W. (1987). Differentiation of maximal from submaximal static
elbowflexor efforts by measurement variability. American Journal of Physical Medicine and Rehabilitation, 66, 213-218.

Bohannon, R., & Smith, M. (1988). Differentiation of maximal and submaximal knee extension efforts by isokinetic testing. Clinical Biomechanics, 1, 16-24.

Brena, S. & Meacham, A. (1990). Evaluation of function and disability. In J. Bonica, J. Loeser, & C.R. Chapman (Eds.), The management of pain (pp.641-648). Philadelphia: Lea and Fiberger.

Cantor, A., Gold, B., Gueron, M., Cristal, N., Praqjrod, G., & Shapiro, Y. (1987). Isotonic (dynamic) and isometric effort in the assessment and evaluation of diastolic hypertension: correlation and clinical use. Cardiology. 74, 141-146.

Carlsoo, S. (1986). With what degree can voluntary static force be repeated. Scandinavian Journal of Rehabilitation Medicine, 18, 1-13.

Carlton, L.G., & Newell, K.M. (1988). Force variability and movement accuracy in space-time. Journal of Experimental Psychology. 14, 24-36.

Cassisi, J.E., Robinson, M.E., O'Connor, P.D., & MacMillan, M.M. (1993).
Trunk strength and lumbar paraspinal muscle activity during isometric exercise in chronic low-back pain patients and controls. Spine, 18, 245-251.








Chengalur, S.N., Smith, G.A., Nelson, R.C., & Sadoff, A.M. (1990). Assessing sincerity of effort in maximal grip strength tests. American Journal of Physical Medicine & Rehabilitation, 69, 148-153.

David, G., Dvir, Z., Mackintosh, S., & Brien, C. (1996). Validity study of a novel test protocol for the identification of submaximal muscular effort. Isokinetics and Exercise, 6, 139-144.

Davies, G.J., Wilk, K., & Ellenbecker, T.S. (1997). Assessment of strength. In T.R. Malone, T. McPoil, & A.J. Nitz (Eds.), Orthopedic and sports physical therapy (3' ed., pp. 217-233). St Louis: Mosby-Year Book, Inc.

Dvir, Z. (1997a). Differentiation of submaximal from maximal trunk extension effort: An Isokinetic study using a new testing protocol. Spine, 22 2672-2676.

Dvir, Z. (1997b). An isokinetic study of submaximal effort in elbow flexion. Perceptual and Motor Skills, 84, 1431-1438.

Dvir, Z., & David, G. (1996). Suboptimal muscular performance: Measuring isokinetic strength of knee extensors with new testing protocol. Archives of Physical Medicine and Rehabilitation, 77 578-581.

Fitts, P.M. (1954). The information capacity of the human motor system in
controlling the amplitude of movement. Journal of Experimental Psychology, 47, 381391.

Flint, M. (1955). Effect of increasing back and abdominal muscle strength on low back pain. Research Quarterly, 29, 160-171.

Frymoyer, J.W., & Cats-Baril, W.L. (1987). Predictors of low back pain disability. Clinics of Orthopedics, 221, 89-98.

Frymoyer, J.W., & Cats-Baril, W.L. (1991). An overview of the incidences and costs of low back pain. Orthopedic Clinics of North America, 22 263-271.

Gilbert, J.C., & Knowlton, R.G. (1983). Simple method to determine sincerity of effort during a maximal isometric test of grip strength. American Journal of Physical Medicine and Rehabilitation, 62, 135-144.

Greenleaf, J.E., Bernauer, E.M., Ertl, A.C., Trowbridge, T.S., & Wade, C.E. (1989). Work capacity during 30 days of bedrest with isotonic and isokinetic exercise training. Journal of Applied Physiology, 67, 1820-1826.

Hammond, H.K., & Froelicher, V.F. (1985). Normal and abnormal heart rate responses to exercise. Progress in Cardiovascular Disease, 27, 271-196.








Harber, P., & Soohoo, K. (1984). Static ergonomic strength testing in evaluating occupational back pain. Journal of Occupational Medicine, 26, 887-884.

Hazard R., Reid, S., Fenwick, J., & Reeves, V. (1988). Isokinetic trunk and lifting strength measurements: Variability as an indicator of effort. Spine, 13, 54-57.

Holmes, B., Leggett, S., Mooney, V., Nichols, J., Negri, S., & Hoeyberghs, A.
(1996). Comparison of female geriatric lumbar extension strength: Asymptomatic versus chronic low back pain patients and their response to active rehabilitation. Journal of Spinal Disorders, 9, 17-22.

Kishino, N.D., Mayer, T.G., Gatchel, R.J., McCrate Parrish, M., Anderson, C., Gustin, L., & Mooney, V. (1985). Quantification of lumbar function. Part 4: Isometric and isokinetic lifting simulation in normal subjects and low-back dysfunction patients. Spine, 10, 921-927.

Kluck, D.J. (1967). A study of the strength ratio of the back extensors to the
trunk flexors. Masters Thesis. Department of Physical Therapy, University of Iowa, Iowa City.

Lin, P.C., Robinson, M.E., Carlos, J., & O'Connor, P. (1996) Detection of submaximal efforts in isometric and isokinetic knee extension tests. Journal of Orthopaedic and Sports Physical Therapy, 24, 19-24.

Liu, A.C., & Byrne, E., (1995). Cost of care for ambulatory patients with low back pain. The Journal of Family Practice, 40, 449-455.

Mayer, L. Greenberg, B.B. (1942). Measurements of the strengths of trunk muscles. American Journal of Bone and Joint Surgery, 24, 842-856.

Mayer, T.G., Barnes, D., Kishino, N.D., Nichols, G., Gatchel, R.J., Mayer, H., & Mooney, V. (1988). Progressive isoinertial evaluation in a standardized protocol and normative data base. Spine. 13, 993-997.

Mayer, T.G., Smith, S.S., Keley, P.T., & Mooney, V. (1985). Quantification of lumbar function: Part 2. Saggital plane trunk strength in chronic low back pain patients. Spine, 10. 765-772.

McIntyre, D.R., Glover, L.H., Seeds, R.H., & Levene, J.A. (1990). The
characteristics of preferred low back motion. Journal of Spinal Disorders, 3, 1454 155.

Mendelson, G. (1995). Compensation neurosis revisited: Outcome studies of the effects of litigation. Journal of Psychosomatic Research, 39, 695-706.








Mooney, V.M., & Andersson, G.B.J. (1994). Controversies: Trunk strength testing in patient evaluation and treatment. Spine, 19, 2483-2485.

Nachemson, A.L., & Lindh, M. (1969). Measurement of abdominal and back muscle strength with and without low back pain. Scandinavian Journal of Rehabilitation Medicine, 1, 60-65.

Newell, K.M., & Carlton, L.G. (1988). Force variability in isometric responses. Journal of Experimental Psychology: Human Perception and Performance, 1 37-44.

Newell, K.M., Carlton, L.G., & Hancock, P.A. (1984). Kinetic analysis of response variability. Psychology Bulletin, 1, 133-151.

Niebuhr, B., & Marion, R. (1990). Voluntary control of submaximal grip strength. American Journal of Physical Medicine & Rehabilitation, 69, 96-101.

Portney, L.G., & Watkins, M.P. (1993). Foundations of clinical research and applications to practice. Norwalk, CT: Appleton & Lange.

Richards, J.G., Quigley, E.J., & Castaqno, P.W. (1996). Validity and reliability of the BTE Dynatrac. Medical Science Sports Exercise, 28 913-920.

Robinson, M.E., Cassisi, J.E., O'Connor, P.D., & MacMillan, M. (1992).
Lumbar iEMG during isotonic exercise: Chronic low back pain patients versus controls. Journal of Spinal Disorders, 5 8-15.

Robinson, M.E., Geisser, M.E., Hanson, C.S., & O'Connor, P. (1993). Detecting submaximal efforts in grip strength testing with the coefficient of variation. Journal of Occupational Rehabilitation, 3, 45-50.

Robinson, M.E., MacMillan, M., O'Connor, P., Fuller, A., & Cassisi, J. (1991). Reproducibility of maximal versus submaximal efforts in an isometric lumbar extension task. Journal of Spinal Disorders, 4, 444-448.

Robinson, M.E., O'Connor, P., MacMillan, M., Fuller, A., & Cassisi, J. (1992). Effect of instructions to simulate a back injury on torque reproducibility in an isometric lumbar extension task. Journal of Occupational Rehabilitation, 2, 1-9.

Robinson, M.E., O'Connor, P., Riley, J.L., Kvaal, S., & Shirley, F.R. (1994). Variability of isometric and isotonic leg exercise: Utility for detection of submaximal efforts. Journal of Occupational Rehabilitation, 4, 163-169.

Robinson, M.E., Sadler, I.J., O'Connor, P., & Riley, J.L. (1997). Detection of submaximal effort and assessment of stability of the coefficient of variation. Journal of Occupational Rehabilitation, 7, 207-215.








Sanders, B. (1997). Exercise and Rehabilitation concepts. In T.R. Malone, T.
McPoil, & A.J. Nitz (Eds.), Orthopedic and sports physical therapy (3 ed., pp. 236-247). St Louis: Mosby-Year Book, Inc.

Schmidt, D.E., Zelaznik, H.N., Hawkins, B., Frank, J.S., & Zuinn, J.T. (1979). Motor output variability: A theory for the accuracy of rapid motor acts. Psychology Review, 86, 415-451.

Schmidt, R.A. (1988). Motor control and learning: A behavioral emphasis. Urbana-Champaign, IL: Human Kinetics Publishers.

Sherwood, D.E., & Schmidt, R.A. (1980). The relationship between force and force variability in minimal and near maximal static and dynamic contractions. Journal of Motor Behavior, 12, 75-89.

Sherwood, D.E., Schmidt, R.A., & Walter, C.B. (1988). The force/forcevariability relationship under controlled temporal conditions. Journal of Motor Behavior, 20 106-116

Simonsen, J.C. (1995). Coefficient of variation as a measure of subject effort. Archives of Physical Medicine and Rehabilitation, 76, 516-520.

Smith, G.A., Nelson, R.C., Sadoff, S.J., & Sadoff, A.M. (1989). Assessing
sincerity of effort in maximal grip strength tasks. American Journal of Physical Medicine and Rehabilitation, 68 73-80.

Stokes, I.A.F., Rush, S., Moffroid, M., Johnson, G.B., & Haugh, L.D. (1987). Trunk extensor EMG-torque relationship. Spine. 12, 770-776.

Webster, B.S., & Snook, S.H. (1994). The cost of 1989 worker's compensation low back pain claims. Spine. 19. 1011-1016.

Woodsworth, R.S. (1899). The accuracy of voluntary movement. Psychology Review, 3 (Suppl. 2).













BIOGRAPHICAL SKETCH

Ian John Sadler was born in Georgetown, Guyana, S.A., on March 16, 1966. He emigrated to Toronto, Ontario, in May of 1976. Mr. Sadler graduated from Silverthorn Collegiate Institute in Toronto in June of 1980. He attended McMaster University in Hamilton, Ontario, where he completed his Bachelor of Science degree in honors biology and psychology in 1990. Following graduation, he departed for Tokyo, Japan, where he spent one year as an English teacher before returning to Toronto in the fall of 1991. Upon his return, he spent one year conducting research at the University of Toronto before entering graduate school. In the fall of 1992 he entered Connecticut College in New London, Connecticut, where he completed his Master of Arts degree in psychology. He entered the graduate program in clinical and health psychology at the University of Florida in 1994. He will be completing his predoctoral internship training year at the University of California, San Diego, starting July, 1, 1998.








I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.


'Michael E Robinson, Chair Associate Professor of Clinical and Health Psychology


I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.


Duarne E. Dede
Assistant Professor of Clinical and Health Psychology


I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.


JoW assel
Assistant Professor of Clinical and Health Psychology


I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.


gamuel F. Sears
Assistant Professor of Clinical and Health Psychology


I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosopy


Mark Trimble P
Assistant Professor f Physical Therapy








This dissertation was submitted to the Graduate Faculty of the College of Health Professions and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy.


August 1999


Dean, College of Health Professions


Dean, Graduate School




Full Text
REFERENCES
Agre, J., Magness, J., Hull, S., Wright, K., Baxter, T., & Patterson, R. (1987).
Strength testing with a portable dynamometer: reliability of upper and lower extremities.
Archives of Physical Medicine and Rehabilitation. 68, 454-458.
Albert, M. (1991). Eccentric muscle training in snorts and orthopaedics. New
York: Churchill Livingstone.
Battie, M., Bigos, S.J Fisher, L.D., Hanson, T.H., Jones, ME., & Ley, M.D.,
(1989). Isometric lifting strength as a predictor of individual back pain reports. Spine, 14,
851-856.
Beimbom, D.S., & Morrissey, M.C. (1988). A review of the literature related to
trunk muscle performance. Spine, 13. 655-660.
Bohannon, R.W. (1987). Differentiation of maximal from submaximal static
elbowflexor efforts by measurement variability. American Journal of Physical Medicine
and Rehabilitation. 66, 213-218.
Bohannon, R., & Smith, M. (1988). Differentiation of maximal and submaximal
knee extension efforts by isokinetic testing. Clinical Biomechanics. 1, 16-24.
Brea, S. & Meacham, A. (1990). Evaluation of function and disability. In J.
Bonica, J. Loeser, & C .R. Chapman (Eds ), The management of pain (pp.641-648).
Philadelphia: Lea and Fiberger
Cantor, A., Gold, B Gueron, M., Cristal, N., Praqjrod, G., & Shapiro, Y. (1987).
Isotonic (dynamic) and isometric effort in the assessment and evaluation of diastolic
hypertension: correlation and clinical use. Cardiology, 74. 141-146.
Carlsoo, S. (1986). With what degree can voluntary static force be repeated.
Scandinavian Journal of Rehabilitation Medicine, 18. 1-13.
Carlton, L.G., & Newell, KM. (1988) Force variability and movement accuracy
in space-time. Journal of Experimental Psychology, 14. 24-36.
Cassisi, J.E., Robinson, M E., OConnor, P.D., & MacMillan, MM. (1993).
Trunk strength and lumbar paraspinal muscle activity during isometric exercise in chronic
low-back pain patients and controls. Spine. 18. 245-251.
37


15
used a test-retest protocol It was hypothesized that calculation of the CV based on 5
repetitions would be more stable over time than that based on 3 repetitions Further, it was
hypothesized that submaximal efforts would be more variable than maximal efforts.
Twenty asymptomatic subjects participated in isokinetic leg extension strength
testing. The coefficient of variation of peak torque (CVpeak) and coefficient of variation
of average torque (CVavg) were obtained from maximal and submaximal torque curves
during fast isokinetic (1807s) leg extension strength tests. Subjects participated in two
conditions of testing: 100% effort and 50% effort. Each subject completed 6 knee
extensions at 100% effort and 6 knee extensions at 50% effort on Day 1 of testing. This
was repeated at least 48 hours later.
The results of this study replicated previous research (Lin et al., 1996), providing
evidence that submaximal efforts as measured using isokinetic leg extension testing were
significantly more variable than maximal efforts. However, effort level classification,
although improved compared to the use of isometric assessment techniques, resulted in
misclassification of submaximal efforts. Using the clinically accepted cutoff of 15%, a
large number of submaximal efforts were misclassified. However, adjusting the cutoff to
an optimized 10%, resulted in a significant reduction in misclassification rate with no false
positives. CVpeak (5 trials) resulted in a detection rate of 75% for an overall classification
rate of 88% on Day 1 of testing. Lin et al. (1996) also found that a 10% cutoff detected
more submaximal efforts than any other cutoff greater than it, with no false positives.
However, they used the CVave measure, obtaining a detection rate of 75% of submaximal
effort for an overall classification rate of 88% with no false positives Using the CVpeak
(4 trials) measure, they detected 59% of submaximal efforts for an overall classification


5
Carlos, & OConnor, 1996). It is a measure of relative variability, calculated as the ratio of
standard deviation divided by the mean multiplied by 100 to yield a unitless percentage. It
therefore expresses the standard deviation as a proportion of the mean, thereby accounting
for differences in the magnitude of the mean. The coefficient of variation is most
meaningful when comparing two distributions (Portney & Watkins, 1993).
The CV has been used as a measure of variability in strength measurements and it
has been suggested that it might be useful for determining whether individuals are giving
their maximal voluntary efforts. It is often assumed that a high CV indicates a lack of
consistent effort; however, CV norms have not been established. Simonsen (1995) points
out that if the CV can be used to validate an individuals effort, we must know how high a
CV can be before we judge it as unacceptable. Agre and colleagues (1987) have reported
that the CV from isometric muscle tests range from 5.1% to 8.3% in the upper arm and
from 11.3% to 17.8% in the lower extremity. Bohannon and Smith (1988) examined
isokinetic knee extension in healthy subjects. The CV for four trials for maximal and
submaximal efforts were compared. Although on average, the submaximal efforts yielded
higher CVs (mean 13.3%) than the maximal effort (mean 6.0%), no clear cutoff was
identified above and below which performance could be dichotomized into the submaximal
and maximal groups based on validity Recent research (Lin et al., 1996; Robinson et al.,
1994; Robinson, Sadler, OConnor, & Riley, 1997) has indicated that use of a 10% CV
cutoff can detect more submaximal efforts than any other cutoff. Harber and Soohoo
(1984) suggest that there is no established validity in using the CV to discover
malingering. However, the fact that the CV varies with different tests suggests that an
absolute cutoff score may not be valid for all tests, even if one could be agreed on. A


11
It is clear from these results that individuals are able to consistently show
weakness on an isometric strength test This is likely to be the intent of an individual
who is exaggerating or faking a low back injury. Although it was possible to show
strength differences between maximal and submaximal effort conditions within a given
subject, it was of little utility to attempt such comparisons across subjects. Previous
research (Mayer, Smith, Keley, & Mooney, 1985; McIntyre, Glover, Seeds & Levene,
1990) has suggested that CLBP patients produce less torque from the lumbar musculature
than normals do. However, in a population that may have incentives to show weakness,
there are no data to suggest that those individuals giving a submaximal effort may be
differentiated from those giving a maximal effort on the basis of strength measures alone.
In summary, given the reported findings there is insufficient evidence to suggest
that reliable classification of effort level can be determined using isometric strength tasks.
Further, despite the need for assessment of sincerity of effort in backs, a review of the
literature indicates a relative paucity of research specifically addressing this area. As a
result, the continued discussion of the transition from a static to a dynamic testing
approach involves primarily a review of sincerity of effort in leg extension testing.
The Move to Dynamic Testing
Our discussion so far has focused on the more static isometric assessment measure.
However, while the use of isometric tasks appears to be of increasingly questionable
value, there has been relatively little investigation of the potential of other strength
assessment tasks. While this section will focus primarily on dynamic strength assessment,
some of the research has included an isometric task in a continued effort to investigate and
compare this methodology with dynamic tasks.


3
Isometric Exercise. (Fixed Speed; Immovable. Fixed Resistance)
Isometric or static resistance exercises normally are performed against an
immovable object such as a wall, a barbell, or weight machine that is loaded beyond the
maximal concentric strength of the individual (Sanders, 1997). This type of exercise or
muscle contraction occurs when a muscle attempts to shorten but is unable to overcome
the resistance (Albert, 1991).
Isotonic Exercise: (Variable Speed; Fixed Resistance)
This exercise is dynamic and is traditionally defined as a muscular contraction in
which the muscle exerts a constant tension (velocity varies). Another way to think about
isotonic exercise is that it is muscle movement with a constant load (Sanders, 1997).
Isotonic exercise is divided into concentric and eccentric muscle loading. The individual
muscle fibers actually shorten with a concentric contraction. In an eccentric action,
muscle fibers actually go through a lengthening process (Davies et al., 1997).
Isokinetic Exercise: (Fixed Speed; Accommodating Resistance)
Isokinetic exercise can be defined as exercise with a constant maximal speed and
variable load (Sanders, 1997). In isokinetic exercise, the muscular contraction is
performed at a constant maximal angular limb velocity, and the equipment varies the
resistance. Isokinetics have been used extensively for muscle performance evaluation and
objective documentation and for rehabilitation. The exercise is dynamic and most
isokinetic machines provide both concentric and eccentric exercise options.
Measurement: Use of the Coefficient of Variation
Having discussed the different strength testing methodologies, it is now necessary
to address the measurement approaches used in determination of effort level.


4
Hazard, Reid, Fenwick, and Reeves (1988) have examined the hypothesis that
force/distance curve variability distinguishes submaximal from maximal efforts in
isokinetic trunk and lifting strength tests. Thirty normal subjects were tested on the Cybex
trunk extension/flexion (TEF) and Lift Task (LT) machines during maximal (100%) and
submaximal efforts (50%). Considering each test separately, visual assessments of curve
variability were indeterminate in 28% of TEF and 24% of LT tasks Measurement models
of curve variability were more clearly discriminating. Hazard et al. concluded that
estimation of the degree of effort a given individual applies to a test of strength is essential
to the clinical interpretation of that persons performance. While they assert that clinical
judgment is required in evaluating effort during tests of isokinetic trunk and lifting
strength, minimal attention to the need for more objective assessment measures was
discussed. Further, this study points out a number of problems in the clinical application
of the principle of rating degree of effort according to curve variability. First, if an
individual produces curves of insufficient amplitude to allow assessment of variability, no
effort rating can be made. Second, the visual technique used is subjective and inherently
produces intermediate or debatable degrees of curve variability. Third, the correlation
between curve variability and degree of effort is not perfect. It is possible for an
inexperienced subject to produce consistent curves inadvertently while exerting a
submaximal effort. Conversely, highly motivated subjects operating at maximal effort can
produce variable curves.
Another method of measuring variability is the Coefficient of Variation (CV)
which has been used in numerous studies as a measure of subject effort (Carlsoo, 1986;
Robinson, OConnor, Riley, Kvaal, & Shirley, 1994; Simonsen, 1995; Lin, Robinson


14
with an optimized (no false positives) CV cutoff With regard to the isokinetic condition,
Lin et al. (1996) found that a 10% cutoff of CVpeak correctly classified 59% of
submaximal efforts in fast isokinetic tests (180/s) for an overall correct classification rate
of 80% without any false positives. Using a 10% CVavg cutoff, 75% of submaximal
efforts were detected for an overall correct classification rate of 88%. These results
suggest that the isokinetic task may represent an approach that improves upon the
discrimination of submaximal from maximal efforts. This appears to confirm the view of
Robinson et al. (1994) in isometric and isotonic knee extension tests that the sensitivity to
detect submaximal efforts was much greater for the dynamic condition. It seems that
movement is the key difference between methods that better discriminate (isokinetic and
isotonic), and methods that are less successful (isometric). As suggested, it may be that
subjects are less able to reliably reproduce torque and velocity of submaximal efforts
because of the dynamic nature of the task.
In a follow-up study, focusing on a dynamic strength assessment task, Robinson et
al. (1997) investigated the stability of the coefficient of variation over time in
asymptomatic normals using an isokinetic leg extension strength assessment task
Previous research (Robinson, Geisser, Hanson & OConnor, 1993) had suggested that the
CV for isometric strength testing might be unstable because of the usual method of its
calculation based on a small number of trials (typically 3). Although significant
classification of asymptomatic normals to effort condition was obtained in two recent
studies (Lin et al., 1996; Robinson et al., 1994), instability of the CV would make
replication of the results more difficult and would jeopardize the clinical utility of the
method. This more recent study calculated the CV with a larger number of repetitions and


DETECTION OF SUBMAXIMAL EFFORT N ISOTONIC BACK STRENGTH
TESTING: DETERMINATION OF OPTIMAL RESISTANCE LEVEL
By
IAN J SADLER
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE
UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1999

ACKNOWLEDGMENTS
A number of individuals have made significant contributions to this dissertation in
particular and to my graduate training in general. Michael Robinson stands out as the
person who has contributed the most to my professional development and provided me
with a solid foundation in both clinical health psychology and, especially, experimental
methodology. This dissertation is a testimony to his commitment to the highest standards
of empiricism and scholarly work. I would also like to acknowledge the contributions of
my committee members, Duane Dede, Jon Kassel, Samuel Sears, and Mark Trimble.
These individuals have seen me through the process of qualifying examinations,
dissertation proposal, and dissertation defense. Their involvement and support has been
appreciated. Although not on my dissertation committee, Cynthia Belar, Jim Rodrigue,
and Michael Perri were also instrumental in my development as a clinician and as an
academic professional at the University of Florida. Collectively, these individuals have
provided me with a model of the scientist-practitioner in clinical psychology.
Several individuals have made contributions of time and energy to aspects of the
present study, including Patrick OConnor, Joseph Riley, and John Otis. Joseph Riley has
been a supportive friend and colleague from whom I have learned a great deal both
professionally and personally.
11

I have made close and lasting relationships throughout the years of graduate
school. These friends have seen me through the good times and the difficult times. To
these individuals (Amy, Becca, Braden, Elena, Kathy, and Gregg), I say a heartfelt thank
you.
Finally, I will forever be indebted to my parents, Bridget and Brian Sadler, as well
as my sister, Sandy, for their innumerable sacrifices and unwavering love and support
throughout my years of graduate education. I dedicate this work to my family.
in

> CONTENTS
ACKNOWLEDGMENTS
ABSTRACT
CHAPTERS
1 INTRODUCTION.
Clinical Assessment of Force
The Present Study
2 METHOD
Participants
BIOGRAPHICAL SKETCH

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
DETECTION OF SUBMAXIMAL EFFORT IN ISOTONIC BACK STRENGTH
TESTING: DETERMINATION OF OPTIMAL RESISTANCE LEVEL
By
Ian J. Sadler
August 1999
Chairperson: Michael E. Robinson
Major Department: Clinical and Health Psychology
The purpose of this study was to investigate the use of an isotonic back strength
assessment task in the detection of submaximal effort and to test the hypothesis that a
submaximal effort would yield greater variability as measured by the coefficient of
variation. There was also an examination of the optimal resistance level at which effort
level could best be discriminated. The hypothesis proposed that a resistance level of either
15%, 25%, or 35% would produce convincing results with respect to effort
discriminability. It was also hypothesized that the calculation of the CV measure based
on 6 repetitions would result in an improved submaximal effort detection rate versus a CV
calculation based on 3 repetitions. Sixty- six participants were asked to undergo back
extension assessment. Each participant completed 3 practice repetitions followed by 7
repetitions at a 50% effort level and 7 repetitions at a 100% effort level. There was a 10
v

minute rest period following the first set of 7 repetitions and order of conditions was
counterbalanced. Results indicated that while participants were able to produce a
submaximal effort, there was not a significant difference in variability between effort levels
across the three resistance levels. There was no difference in the use of the CV based on
6 versus 3 repetitions with respect to effort discriminability. Implications of these results
are discussed with recommendations for continued investigation of effort level in backs
through use of isokinetic testing methodology.
vi

CHAPTER 1
INTRODUCTION
Musculoskeletal injuries are the most common cause of frequent or permanent
impairment (Brea & Meacham, 1990). In a review of the literature, Frymoyer and Cats-
Baril (1991) reported that while low back disorders are extremely prevalent in all
societies, and probably have not increased substantially over the last two decades, they
note that what has increased is the rate of disability, the reasons for which are uncertain.
According to these researchers what has changed is societys perception, most particularly
relating to the disability that results. With an increasing rate of disability, the costs have
increased to industry and government, and there has been an increased utilization of
medical care, including surgical intervention. While Frymoyer and Cats-Baril (1987)
reported a dramatic increase in disability of 14 times the rate of population growth from
1977 to 1981, further research has suggested that this disability rate relative to population
growth continues to increase (Webster & Snook, 1994).
Although a precise estimate is impossible, it is plausible that the direct medical and
indirect costs of these conditions are in the range of more then $50 billion per annum, and
could be as high as $100 billion at the extreme (Frymoyer & Cats-Baril, 1991). Of these
costs, 75% or more can be attributed to the 5% of people who become disabled
temporarily or permanently from back pain. Further, it is estimated that the average
worker's compensation claim results in costs of $4075 annually (Webster & Snook, 1994).
Webster and Snook reported the mean cost per case of low back pain was $8321,
1

2
more than twice the amount for the average workers compensation claim ($4075).
Further the median cost per low back pain case was $396, an indication that low back pain
costs are not normally distributed with a few cases accounting for most of the costs.
Clinical Assessment of Force
As health care providers in the current managed care environment attempt to care
for patients with musculoskeletal disorders in a more cost effective manner, it is becoming
increasingly important to identify which diagnostic and therapeutic procedures have the
greatest impact on the cost of caring for patients with low back pain (Liu & Byrne, 1995).
The measurement of applied force produced by voluntary muscular contraction across a
range of motion is a common method used to assess the status of the musculoskeletal
system. Strength testing is an important part of an assessment battery for evaluation of
patients with chronic low back pain. A number of testing methods have been investigated
including isotonic, isokinetic and isometric testing protocols. A brief description of these
testing approaches is provided.
Isotonic. Isokinetic, and Isometric Testing
Various forms of musculoskeletal exercise are available for the researcher to use
when assessing muscle strength. Today, three primary types of exercise are commonly
employed: isometrics, isotonics, and isokinetics (Davies, Wilk, & Ellenbecker, 1997).
While isotonic methodology will be used in the present study, a considerable amount of
the research has involved the two other testing approaches. As a result, definitions of all
three approaches follow:

3
Isometric Exercise. (Fixed Speed; Immovable. Fixed Resistance)
Isometric or static resistance exercises normally are performed against an
immovable object such as a wall, a barbell, or weight machine that is loaded beyond the
maximal concentric strength of the individual (Sanders, 1997). This type of exercise or
muscle contraction occurs when a muscle attempts to shorten but is unable to overcome
the resistance (Albert, 1991).
Isotonic Exercise: (Variable Speed; Fixed Resistance)
This exercise is dynamic and is traditionally defined as a muscular contraction in
which the muscle exerts a constant tension (velocity varies). Another way to think about
isotonic exercise is that it is muscle movement with a constant load (Sanders, 1997).
Isotonic exercise is divided into concentric and eccentric muscle loading. The individual
muscle fibers actually shorten with a concentric contraction. In an eccentric action,
muscle fibers actually go through a lengthening process (Davies et al., 1997).
Isokinetic Exercise: (Fixed Speed; Accommodating Resistance)
Isokinetic exercise can be defined as exercise with a constant maximal speed and
variable load (Sanders, 1997). In isokinetic exercise, the muscular contraction is
performed at a constant maximal angular limb velocity, and the equipment varies the
resistance. Isokinetics have been used extensively for muscle performance evaluation and
objective documentation and for rehabilitation. The exercise is dynamic and most
isokinetic machines provide both concentric and eccentric exercise options.
Measurement: Use of the Coefficient of Variation
Having discussed the different strength testing methodologies, it is now necessary
to address the measurement approaches used in determination of effort level.

4
Hazard, Reid, Fenwick, and Reeves (1988) have examined the hypothesis that
force/distance curve variability distinguishes submaximal from maximal efforts in
isokinetic trunk and lifting strength tests. Thirty normal subjects were tested on the Cybex
trunk extension/flexion (TEF) and Lift Task (LT) machines during maximal (100%) and
submaximal efforts (50%). Considering each test separately, visual assessments of curve
variability were indeterminate in 28% of TEF and 24% of LT tasks Measurement models
of curve variability were more clearly discriminating. Hazard et al. concluded that
estimation of the degree of effort a given individual applies to a test of strength is essential
to the clinical interpretation of that persons performance. While they assert that clinical
judgment is required in evaluating effort during tests of isokinetic trunk and lifting
strength, minimal attention to the need for more objective assessment measures was
discussed. Further, this study points out a number of problems in the clinical application
of the principle of rating degree of effort according to curve variability. First, if an
individual produces curves of insufficient amplitude to allow assessment of variability, no
effort rating can be made. Second, the visual technique used is subjective and inherently
produces intermediate or debatable degrees of curve variability. Third, the correlation
between curve variability and degree of effort is not perfect. It is possible for an
inexperienced subject to produce consistent curves inadvertently while exerting a
submaximal effort. Conversely, highly motivated subjects operating at maximal effort can
produce variable curves.
Another method of measuring variability is the Coefficient of Variation (CV)
which has been used in numerous studies as a measure of subject effort (Carlsoo, 1986;
Robinson, OConnor, Riley, Kvaal, & Shirley, 1994; Simonsen, 1995; Lin, Robinson

5
Carlos, & OConnor, 1996). It is a measure of relative variability, calculated as the ratio of
standard deviation divided by the mean multiplied by 100 to yield a unitless percentage. It
therefore expresses the standard deviation as a proportion of the mean, thereby accounting
for differences in the magnitude of the mean. The coefficient of variation is most
meaningful when comparing two distributions (Portney & Watkins, 1993).
The CV has been used as a measure of variability in strength measurements and it
has been suggested that it might be useful for determining whether individuals are giving
their maximal voluntary efforts. It is often assumed that a high CV indicates a lack of
consistent effort; however, CV norms have not been established. Simonsen (1995) points
out that if the CV can be used to validate an individuals effort, we must know how high a
CV can be before we judge it as unacceptable. Agre and colleagues (1987) have reported
that the CV from isometric muscle tests range from 5.1% to 8.3% in the upper arm and
from 11.3% to 17.8% in the lower extremity. Bohannon and Smith (1988) examined
isokinetic knee extension in healthy subjects. The CV for four trials for maximal and
submaximal efforts were compared. Although on average, the submaximal efforts yielded
higher CVs (mean 13.3%) than the maximal effort (mean 6.0%), no clear cutoff was
identified above and below which performance could be dichotomized into the submaximal
and maximal groups based on validity Recent research (Lin et al., 1996; Robinson et al.,
1994; Robinson, Sadler, OConnor, & Riley, 1997) has indicated that use of a 10% CV
cutoff can detect more submaximal efforts than any other cutoff. Harber and Soohoo
(1984) suggest that there is no established validity in using the CV to discover
malingering. However, the fact that the CV varies with different tests suggests that an
absolute cutoff score may not be valid for all tests, even if one could be agreed on. A

6
relative cutoff that varies from test to test makes more sense. Bohannon (1987) mentions
the need for norms to be established for a specific test before a cutoff score can be
determined and judgments made regarding subjects effort. Research continues to address
the utility of using the CV in detection of submaximal efforts. Having provided a brief
explanation of both the testing and measurement approaches in the assessment of
musculoskeletal injury, the discussion now addresses how the research has responded to
the question of how to best assess effort level.
Statement of the Problem
Given the review of the costs of musculoskeletal injury, the need for reliable
assessment of sincerity of effort in back strength testing is emphasized. Further, a factor
contributing to lack of reliability in such testing is the potential for patients to be
motivated to appear more disabled than they are. Patients with musculoskeletal injury
may have incentive to produce less than maximal effort given concern for future injury,
exacerbation of current pain, apprehension in a novel test environment as well as issues
related to compensation (Mendelson, 1995). The degree to which "insincere" effort in
testing contributes to the reported costs is difficult to determine.
In the testing of injured individuals, evaluation of sincerity of effort is an important
factor in determination of functional capacity. The inability to accurately assess a patient's
effort level can lead to procedures that are both unnecessary and ineffective (Robinson,
OConnor, Riley, Kvaal, & Shirley, 1994). Further, the failure to accurately determine a
patients sincerity of effort has implications regarding the lack of apparent response to
treatment, as well as escalating health care and disability costs. In addition, failure to

7
accurately detect patient effort can create an atmosphere where suspicion dominates,
which may then generalize to the genuinely disabled
Sincerity of Effort
One approach to the assessment of sincerity of effort in both research and
practice has been based on the hypothesis that sincerity of effort may be determined by
evaluating the consistency of individuals' responses to various strength testing tasks.
According to this hypothesis, the performance of an individual who is sincerely attempting
maximal exertion will be influenced only by physical limitations associated with his or her
injury or impairment and the influence of these limitations will remain relatively consistent
across trials. However, an individual whose performance is limited not only by physical
impairment but by attempts to give less than full effort will not be able to maintain as
consistent a level of performance, since other than physical factors are limiting his/her
performance.
Previous research has proposed that when a patient is deliberately attempting to
produce a submaximal effort, the obtained strength measurements will show increased
variability and reduced reliability (Kishino et al., 1985). The assumption is that subjects
cannot willfully reproduce a submaximal effort with the same consistency as a maximal
effort. Although Kishino et al. (1985) speculated about this hypothesis, they offered no
data to support it. Bohannon (1987) compared the variability of maximal and submaximal
tests of isometric elbow flexion by calculating the coefficient of variation (CV). He
reported that the CV of force measures was significantly greater during submaximal trials
than during maximal trials, which supported the Kishino et al. hypothesis.

8
Further evidence supporting the sincerity of effort hypothesis is provided by
other investigators who have also suggested that the variability of force production in
injured patients undergoing strength assessment may be an indicator for submaximal effort
(Gilbert & Knowlton, 1983; Chengalur, Smith, Nelson, & Sadoff, 1990; Niebuhr &
Marion, 1990). Gilbert and Knowlton (1983) proposed that individuals may use different
motor strategies in maximal and submaximal efforts. They propose that a maximal effort
is a lower order motor task in which feedback from the muscle contraction is not crucial.
On the other hand, a submaximal effort is a higher order motor task in which feedback is
crucial in limiting the force of the contraction. To limit force, a submaximal task would
require a greater degree of feedback from the muscles than would a maximal effort. An
implication is that differing motor strategies would make feigned submaximal efforts
detectable. While the sincerity of effort hypothesis shows promise, the validity of testing
methods requires more empirical examination.
Static vs. Dynamic Testing
While the assessment of sincerity of effort is now frequently used in clinical
practice, employing both static (isometric) and dynamic (isokinetic and isotonic) testing
approaches, the reliability and validity of current assessment methods have not been
adequately demonstrated (Mooney & Anderson, 1994). The following review will
initially consider the research using a static assessment approach. As the discussion
proceeds, the rationale for the move to a dynamic testing protocol will be provided as will
a review of the literature in this area.

9
Static Testing
Clinical practice has generally studied variability in effort level in isometric tasks. A
number of studies have used the grip test strength assessment task in determination of
effort level (Gilbert & Knowlton, 1983, Bohannon, 1987; Smith, Nelson, Sadoff, & Sadoff,
1989; Chengalur et al. 1990; Niebuhr & Marion, 1990; Robinson, Geisser, Hanson, &
OConnor, 1993). Robinson et al. (1993) used the CV to discriminate level of effort in grip
strength testing. Twenty-nine asymptomatic subjects participated in two conditions of
testing: 100% and 50% effort. The submaximal (50%) effort condition showed
significantly more variability than the maximal effort condition in both sets of conditions.
Intra-class correlation coefficients were very low for both maximal effort and submaximal
efforts (.036 and .025) indicating very low stability for the coefficient of variation.
Classification rates were also found to have unacceptably large errors with 69% of the
submaximal efforts being classified as maximal with the traditional 15% cutoff, and 55%
misclassification of submaximal efforts with an optimized 11% cutoff It was concluded
that the currently practiced method of using a low number of repetitions to calculate the
CV might result in very unstable measures. Further, grip strength testing, which is an
isometric testing procedure, did not produce consistent results (suitable detection rates)
that might validate its use in a clinical setting. Although its use continues in clinical
practice the research does not adequately support such use.
While grip strength testing may provide helpful information, when we consider the
costs and prevalence of musculoskeletal injury it is important to think about the relevance
of sincerity of effort testing from a more ecologically valid perspective, that is, the
assessment of the spinal region. Because of its significant economic impact, assessment of

10
lumbar spine disorders in workers and patients has become widespread. These tests are
used in the prediction of back injury (Battie et al., 1989) and assessing the functional
capacity of workers (Mayer et al., 1988). Although the methods of recording force
production may differ, with research at this point having focused primarily on isometric
and isokinetic testing, all methods have assumed that subjects giving less than the
instructed full, maximal effort would be identified by their inconsistent recordings.
However, despite the medicolegal implications of this assumption, there is insufficient data
to support the hypothesis that subjects cannot willfully reproduce a submaximal effort with
the same consistency as a maximal effort
Robinson, MacMillan, OConnor, Fuller and Cassisi (1991) conducted a study
investigating the differences in test-retest reliability between maximal and submaximal
efforts in an isometric lumbar extension task. They also tested the hypothesis that
submaximal efforts would be less consistent than maximal efforts. Twenty subjects were
asked to produce maximal voluntary contractions at seven different positions in a lumbar
extension machine. Each subject was tested twice in a maximal effort condition and twice
in a 50% effort condition. Results indicated high test-retest reliability at all angles in both
conditions. There were no differences in test-retest reliability between effort conditions.
Therapist ratings of consistency did not differ between conditions, and therapists could not
discriminate between conditions on the basis of effort consistency These authors
concluded that claims that isometric strength testing could be used to determine sincerity
of effort or malingering might be unfounded

11
It is clear from these results that individuals are able to consistently show
weakness on an isometric strength test This is likely to be the intent of an individual
who is exaggerating or faking a low back injury. Although it was possible to show
strength differences between maximal and submaximal effort conditions within a given
subject, it was of little utility to attempt such comparisons across subjects. Previous
research (Mayer, Smith, Keley, & Mooney, 1985; McIntyre, Glover, Seeds & Levene,
1990) has suggested that CLBP patients produce less torque from the lumbar musculature
than normals do. However, in a population that may have incentives to show weakness,
there are no data to suggest that those individuals giving a submaximal effort may be
differentiated from those giving a maximal effort on the basis of strength measures alone.
In summary, given the reported findings there is insufficient evidence to suggest
that reliable classification of effort level can be determined using isometric strength tasks.
Further, despite the need for assessment of sincerity of effort in backs, a review of the
literature indicates a relative paucity of research specifically addressing this area. As a
result, the continued discussion of the transition from a static to a dynamic testing
approach involves primarily a review of sincerity of effort in leg extension testing.
The Move to Dynamic Testing
Our discussion so far has focused on the more static isometric assessment measure.
However, while the use of isometric tasks appears to be of increasingly questionable
value, there has been relatively little investigation of the potential of other strength
assessment tasks. While this section will focus primarily on dynamic strength assessment,
some of the research has included an isometric task in a continued effort to investigate and
compare this methodology with dynamic tasks.

12
In the Hazard et al. (1988) isokinetic TEF and LT study, previously discussed in
the context of the sincerity of effort hypothesis, the authors reported modest ability to
correctly classify degree of effort with simple visual inspection of the force curves.
However, using discriminant analysis techniques, the authors were able to improve the
accuracy of classification to 80-83% for the trunk extension task and 75-82% for the lift
task.
Although the results reported in their study were promising, Hazard et al. (1988)
pointed out that some subjects were able to produce consistent submaximal curves and in
contrast others produced inconsistent maximal effort curves, suggesting there is
considerable between subject variability. In addition, these authors noted that clinical
observation during the study was more accurate than the measures of curve variability.
Finally, the absence of uniform criteria for strength variability assessment may contribute
to the controversy in some of the studies mentioned above.
A recent study using the coefficient of variation measure compared the variability
of torque production and velocity in both isometric and isotonic leg extension tasks in
both maximal and submaximal effort conditions (Robinson et al., 1994). Fifteen
asymptomatic subjects participated in a within subject counterbalanced design in which
they were asked to perform maximally and submaximally in both isometric and isotonic leg
extension tasks. Results indicated that both isometric and isotonic tasks showed greater
variability (measured by coefficient of variation) in the submaximal effort condition. The
ability to detect submaximal efforts without misclassification of any maximal efforts was
much greater for the isotonic task (87% of efforts were correctly classified). While the
isometric task had unacceptably poor classification rates (60% of efforts were correctly

13
classified), consistent with previous research, those of the isotonic task were high enough
to show promise for clinical application Further, in the isotonic task there were no false
positives (maximal efforts called submaximal) and only 4 of 15 submaximal efforts were
missed using a 10% CV cutoff. This represents a significant improvement over the
isometric testing method and previous research (Robinson et al., 1991), and provides
support for the use of dynamic testing protocols.
Another recent study has investigated differences between maximal and
submaximal efforts in isometric and isokinetic knee extension using torque variability and
slope measures (Lin et al., 1996). These authors note that the use of torque variability and
slope measures to detect submaximal efforts has been studied in isometric tests, but not
fully investigated in isokinetic tests. Thirty-two healthy students were studied to determine
the extension torque of the knee during isometric and isokinetic exercise. The CV of
average torque, CV of peak torque, and slope to peak torque were obtained from maximal
and submaximal torque curves during isometric and two isokinetic tests (60/s and 1807s).
Subjects were instructed to kick as hard as they could in the maximal effort condition and
were told to give a 50% effort in the submaximal effort condition. The testing order of
maximal vs. submaximal conditions was randomized across subjects to eliminate a possible
order effect caused by fatigue.
Results of this study, which were consistent with previous findings (Robinson et
al., 1993, 1994; Smith et al., 1989), indicated significant differences between effort levels
(maximal and submaximal) for all variables in isometric and isokinetic tests. However, in
the isometric condition, effort level classification based on the CV method was found to
have unacceptably large errors, in that over 50% of submaximal efforts were misclassified

14
with an optimized (no false positives) CV cutoff With regard to the isokinetic condition,
Lin et al. (1996) found that a 10% cutoff of CVpeak correctly classified 59% of
submaximal efforts in fast isokinetic tests (180/s) for an overall correct classification rate
of 80% without any false positives. Using a 10% CVavg cutoff, 75% of submaximal
efforts were detected for an overall correct classification rate of 88%. These results
suggest that the isokinetic task may represent an approach that improves upon the
discrimination of submaximal from maximal efforts. This appears to confirm the view of
Robinson et al. (1994) in isometric and isotonic knee extension tests that the sensitivity to
detect submaximal efforts was much greater for the dynamic condition. It seems that
movement is the key difference between methods that better discriminate (isokinetic and
isotonic), and methods that are less successful (isometric). As suggested, it may be that
subjects are less able to reliably reproduce torque and velocity of submaximal efforts
because of the dynamic nature of the task.
In a follow-up study, focusing on a dynamic strength assessment task, Robinson et
al. (1997) investigated the stability of the coefficient of variation over time in
asymptomatic normals using an isokinetic leg extension strength assessment task
Previous research (Robinson, Geisser, Hanson & OConnor, 1993) had suggested that the
CV for isometric strength testing might be unstable because of the usual method of its
calculation based on a small number of trials (typically 3). Although significant
classification of asymptomatic normals to effort condition was obtained in two recent
studies (Lin et al., 1996; Robinson et al., 1994), instability of the CV would make
replication of the results more difficult and would jeopardize the clinical utility of the
method. This more recent study calculated the CV with a larger number of repetitions and

15
used a test-retest protocol It was hypothesized that calculation of the CV based on 5
repetitions would be more stable over time than that based on 3 repetitions Further, it was
hypothesized that submaximal efforts would be more variable than maximal efforts.
Twenty asymptomatic subjects participated in isokinetic leg extension strength
testing. The coefficient of variation of peak torque (CVpeak) and coefficient of variation
of average torque (CVavg) were obtained from maximal and submaximal torque curves
during fast isokinetic (1807s) leg extension strength tests. Subjects participated in two
conditions of testing: 100% effort and 50% effort. Each subject completed 6 knee
extensions at 100% effort and 6 knee extensions at 50% effort on Day 1 of testing. This
was repeated at least 48 hours later.
The results of this study replicated previous research (Lin et al., 1996), providing
evidence that submaximal efforts as measured using isokinetic leg extension testing were
significantly more variable than maximal efforts. However, effort level classification,
although improved compared to the use of isometric assessment techniques, resulted in
misclassification of submaximal efforts. Using the clinically accepted cutoff of 15%, a
large number of submaximal efforts were misclassified. However, adjusting the cutoff to
an optimized 10%, resulted in a significant reduction in misclassification rate with no false
positives. CVpeak (5 trials) resulted in a detection rate of 75% for an overall classification
rate of 88% on Day 1 of testing. Lin et al. (1996) also found that a 10% cutoff detected
more submaximal efforts than any other cutoff greater than it, with no false positives.
However, they used the CVave measure, obtaining a detection rate of 75% of submaximal
effort for an overall classification rate of 88% with no false positives Using the CVpeak
(4 trials) measure, they detected 59% of submaximal efforts for an overall classification

16
rate of 80%. The more recent study represents an improvement in CVpeak classification
rates compared to the study by Lin et al This CVpeak measure was used in the present
study.
In addition, based on these recent results, it appears that a lower CV cutoff (10%)
than the standard 15% cutoff might be warranted in the detection of submaximal efforts.
Determination of classification stability also indicated that a 10% CV cutoff appears to be
the optimal detection rate at which no false positives were obtained.
Concerns regarding the stability of using a low number of repetitions (usually
three) to calculate the CV were also addressed in the study by Robinson et al. (1997).
The calculation of this measure on such a small number of trials raises the possibility that
the measure may be very unstable. Results indicated that use of 5 repetitions versus 3
repetitions in the calculation of CVpeak resulted in increased stability across trials. Results
indicated that the CVave measure (calculated on 3 trials only) was unstable. The trend in
these findings suggest that increasing the number of repetitions used in CV calculation
may prove beneficial.
The previous research discussed suggests that use of a dynamic methodology
appears to provide better discrimination of effort level. The following discussion provides
a helpful analysis and summary of the dynamics proposition
A Final Comment on Dynamics
Research has suggested that the specific relationship between the force produced
and force variability has been proposed to be linear (Schmidt, Zelaznik, Hawkins, Frank,&
Zuinn, 1979), an inverted-U (Sherwood & Schmidt, 1980), or a curvilinear function
(Sherwood, Schmidt, & Walter, 1988), with variability generally decreasing as function of

17
the force applied. These functions have also been found in studies using dynamic tasks
(Sherwood et al., 1988; Carlton & Newell, 1988) These tasks are thought to be
inherently more complex because of increased changes involving a number of variables
including muscle length, muscle velocity, and limb displacement In their review, Newell,
Carlton and Hancock (1984) found that many functions of variability in motor control
appear, depending on the task-related relationships of movement distance, time to peak
force, magnitude of the force applied, and rate of force production. Newell and Carlton
(1988) noted that separating the effects of these variables has been difficult as a result of
the change in one variable leading to changes in one of the other variables. These effects
have been suggested to be accounted for peripherally by factors such as proprioceptive
feedback, the nature of the muscle contraction mechanism, information transmission ability
of the peripheral anatomical unit as well as centrally by noise in the neural control
mechanism (Newell et al., 1984). Given these factors, it is likely that the increased
complexity of the dynamic tasks results in a greater variability from repetition to
repetition, making these tasks more susceptible to manipulation. In addition, attempting
submaximal efforts leads to an even greater complexity of the task, with a resulting
increase in the likelihood of error and variability from trial to trial. This complexity is
likely to make dynamic tasks the better choice for detecting submaximal efforts.
The Choice of Isotonic Testing
Having now provided a rationale for a dynamic testing approach, the discussion
now turns to consideration of the use of an isotonic testing procedure. A significant
proportion of the research in the area concerning sincerity of effort has involved use of

18
isometric and isokinetic methodology. The evaluation of sincerity of effort using isotonic
testing methods has not been addressed to any significant extent.
Even within the more extensively researched isometric and isokinetic
methodology, research has focused primarily on assessment of strength with less attention
to detection of sincerity of effort. However, it is important to acknowledge that that Dvir
and colleagues have employed isokinetic testing methodology and reported promising
results with respect to detection of sub-optimal performance (Dvir, 1997a, 1997b; David,
Dvir, Mackintosh, fe Brien, 1996; Dvir <& David, 1996).
In a review of the literature related to trunk muscle performance, Beimbom and
Morrisey (1988) discussed aspects such as differences in trunk extension force versus
trunk flexion force. They noted that there was great variation in methods, procedures, and
equipment, as well as type and speed of contraction used in these studies. In addition, the
subjects and their physical condition, weight, sex, age, and height differed in each study.
Due to the extreme variability in methodology and subject samples, it was very difficult to
assemble all of the data and have an accurate set of normative values. Despite these
difficulties, an attempt was made to present a compilation of all the work and to
categorize the many studies according to sample type, equipment used, motions tested,
and the mode of contraction evaluated.
Their review focused on studies investigating torque producing capacity in the
trunk musculature. The studies were classified into three categories: 1) isometric,
2) isotonic and 3) isokinetic analysis. In this review they reported twenty-five studies
employing isometric testing, and seventeen studies employing isokinetic testing methods.
Providing further evidence of the lack of research involving use of isotonic testing

19
methods, they reported only five older studies using this methodology (Flint, 1955; Kluck,
1967; Mayer & Greenberg, 1942; Nachemson, & Lindh, 1969). Considering the paucity
of research investigating isotonic methodology in assessment of trunk musculature torque
producing capacity, it is not surprising that there are even fewer studies investigating the
specific area of sincerity of effort. In fact, an extensive search of the literature considering
isotonic methodology indicates that the majority of this research has used this paradigm in
evaluation of rehabilitation/exercise treatment programs for medical patients with back
problems (Greenleaf, Bemauer, Ertl, Trowbridge, & Wade, 1989; Holmes et al., 1996)
and heart disease (Hammond & Froelicher, 1985; Cantor et al., 1987).
However, Robinson, Cassisi, OConnor & MacMillan (1992), in a study
investigating lumbar iEMG during isotonic exercise of the lumbar spine, have suggested
that while not a direct comparison of the clinical efficacy of isotonic versus isokinetic
testing, their study establishes that clinical isotonic testing can be used in a fashion similar
to clinical isokinetic testing. Further, Robinson et al. (1994) suggest that using the
isotonic method allows for the investigation of differences in velocity variability between
maximal and submaximal efforts. It may be that subjects are less able to reliably
reproduce submaximal velocity efforts than they are able to reproduce submaximal
strength or torque efforts because of the dynamic nature of the task (Robinson et al.,
1994). In addition, the use of dynamic tasks may better represent the normal use of the
joint system, since isometric muscle contractions do not represent a high frequency
behavior.

20
Given the lack of research involving use of isotonic methodology in determining
sincerity of effort, as well as a similar paucity of research assessing this concept in the
spine musculature, a recent pilot study was completed in our lab at the University of
Florida.
Sadler, Robinson, Otis, OConnor and Riley (unpublished data) assessed lumbar
strength variability in 20 asymptomatic controls, using the Dynatrac Back Tester (BTE
Dynatrac, Hanover, Maryland). Participants were asked to complete 6 maximal and 6
submaximal isotonic lumbar extension exercises with a 10 minute break between trials.
Order of condition was counterbalanced. A resistance level of 60% of participants body
weight was set for each subject.
While a significant difference was found between maximal and submaximal efforts
with regard to velocity, there were no significant differences found in the CV between
effort conditions. Accurate classification to effort condition was not possible. These
results clearly did not support recent findings using isokinetic leg extension testing
(Robinson et ai., 1997). It was speculated that perhaps the anatomy of the spine and its
musculature, might somehow result in less variability being observed in back strength
testing.
As a result of the poor discriminability found in this recent study, a follow-up pilot
study was conducted. It was observed at the time of testing in the previous study, that a
number of subjects were experiencing some discomfort with the procedure. As a result, it
was speculated that a lowered resistance level might provide a better opportunity to
discriminate maximal from submaximal efforts. Eight subjects participated in the pilot
study with the same methodology; however, resistance level was set at 25% of body

21
weight. Results indicated that first, participants were more accurate in their attempt at a
50% effort compared to the previous study using a 60% resistance level. In addition, as in
the previous study using 20 subjects, there was a significant difference in the velocities
produced between maximal and submaximal effort conditions. Finally, results indicated
that there was a clear trend toward significant differences (p<06) between CV of
maximal and submaximal effort conditions (d=.80).
The Present Study
The above literature review suggests that the dynamic isotonic testing method
appears to show the most promise Therefore this study considered detection of
submaximal effort in an isotonic back strength test. Given the results of the more recent
studies using different resistance levels, an attempt was made to determine an optimal
resistance level that might lead to improved detection of submaximal effort. In addition, as
a result of trends in recent findings, suggesting somewhat increased stability, a larger
number of repetitions (6) was used in the determination of the CV. Previous results had
raised doubt about the stability of using a low number of repetitions (usually three) to
calculate the CV. The calculation of this measure on such a small number of trials raised
the possibility that the measure might be very unstable
First, it was hypothesized that submaximal efforts would be more variable than
maximal efforts across resistance levels. It was also hypothesized that a resistance level
lower than the 60% level previously addressed, would result in increased discriminability
between effort level. Finally, it was hypothesized that use of the CVpeak 6 trial measure
would provide better discriminability between effort level than the CVpeak 3 trial measure

22
given trends in recent findings suggesting moderate increases in stability with increased
number of repetitions used in CV calculation

CHAPTER 2
METHOD
Participants
Sixty-six asymptomatic normals were recruited from the psychology department
and physical therapy department and asked to participate in isotonic back extension
strength testing. These participants were divided into three groups of twenty-two subjects.
Results of a power analysis indicated that this would be the appropriate number of
subjects per group, given the pilot study results with an effect size of .8. An even number
of males and females were recruited. Exclusion criteria consisted of any history of
neuromuscular disease or any injury to the lower extremity that has required medical or
surgical intervention. Informed consent was obtained from the participants in accordance
with the Institutional Review Board.
Apparatus
Back extension assessment was measured by the Baltimore Therapeutic Equipment
Co. Dynatrac Back Tester (BTE Dynatrac, Hanover, MD). Research investigating the
reliability and validity of the BTE Dynatrac Back Tester has suggested that the
Dynatrac can be considered both reliable and objective (Richards, Bailey, & Castasqno,
1996). The Dynatrac computer provides test reports in which both concentric (as in
lifting) and eccentric (as in lowering) strength is exercised and analyzed. While in the back
tester, the patient remains in one stabilized position for flexion/extension and rotation.
Any size patient can be accommodated.
23

24
This apparatus allows for isometric assessment as well as isotonic testing at a pre
set resistance. The dynamometer can be locked for isometric testing or provide a specific
resistance to be maintained throughout movement in the case of isotonic strength testing.
With Dynatrac isotonic loading, the muscular force the patient must exert during a test is
prescribed and set by the experimenter. This is in contrast to isokinetic loading, where the
patient is told to exert a maximal force throughout his/her range of motion. The Dynatrac
isotonic load can be set by the experimenter to safely suit the patients capability, from
virtually zero to high performance torque levels. Isotonic strength assessment was
conducted in this study.
Procedure
Participants were asked to repeat a back extension exercise in the Dynatrac Back
Tester. Participants were divided into three groups of twenty-two subjects. Three
resistance levels were measured: 15% of body weight, 25% of body weight and 35% of
body weight. These values were chosen based on results from the pilot study which
showed a moderate effect size for weights in this range. Demographic information
including, height, weight and gender was collected.
A graduate student previously trained by a physical therapist and blind to order of
condition conducted the testing. On both trials, the graduate student conducted testing
consistent with maximal effort instructions (Ready! Go! As hard and as fast as you can!);
however, participants were told by the experimenter assistant to give a 100% effort in the
maximal effort condition and were told to give a 50% effort in the submaximal condition.
The assistant informed participants of order of condition.

25
First, the participants were secured to the Dynatrac back testing device. Patients
were placed in a comfortable, semi-kneeling posture with the pelvis at a 35 degree forward
tilt. This placed the spine in a neutral position, maintaining the lumbar lordosis and the
normal angle of the sacrum The semi-kneeling posture allowed full unrestricted flexion
and extension. Simple adjustments accommodated any size participant, tall or short.
Padded fixation points were adjusted for variations in limb length, pelvis height, pelvis
width, and shoulder height, while isolating the back muscles for testing. With this patient
positioning system, test subjects were easily positioned in less than a minute with solid
stabilization.
Once secured in the back tester, participants were told that 3 practice repetitions
would be performed to ensure understanding of the protocol. This was followed by a one
minute rest period Participants were then asked to perform 7 repetitions at a 50% effort
level (submaximal) and 7 repetitions at a 100% effort level. Order of conditions was
counterbalanced and separated by a 10 minute rest period. Finally, upon completion of
the trials subjects were required to indicate the order of effort level that they were
assigned.

CHAPTER 3
RESULTS
Demographics
There were 66 participants in the study (see Table 1 for breakdown). The mean
age of participants was 26.2 years (SD=6.4); mean height was 171.9 cm (SD=9.9cm); and
mean weight was 67.4 kg. (SD=15.7kg). A manipulation check which required subjects
to state the order of effort conditions indicated that all participants followed the assigned
protocol.
Table 1
Number of Participants bv Group
Group 1 (15%)
Group 2 (25%)
Group 3 (35%)
Total
Males 9
9
10
28
Females 13
13
12
38
N=66
Coefficient of Variation
The CVpeak velocity for 6 repetitions was calculated using repetitions 2-7. The
CVpeak velocity for 3 repetitions was calculated using repetitions 2-4. Trial 1 was not
used in the calculations as this was not considered to be an accurate record of effort level
given the possibility that acclimation to the protocol was necessary.
26

27
Submaximal vs. Maximal Effort Level
Mean velocity scores were calculated across repetitions for each group and in each
condition using both 3 and 6 repetitions. These scores were subjected to paired t-tests to
assess differences between conditions. Results indicated that the submaximal effort
yielded significantly lower velocity than the maximal effort across groups as well as
number of repetitions used in calculation (see Table 2).
Table 2
Mean Velocity Differences between Test Conditions
Group
Maximal
Submaximal
t(21)
P
15%-6 reps
221. ld/s
127.2d/s
10.2
.001
15%-3reps
206.3d/s
126.9d/s
18.8
.001
25%-6reps
225.8d/s
140.6d/s
12.4
.001
25%-3reps
211.2d/s
136.5d/s
11.5
.001
35%-6reps
219.4d/s
124 3d/s
14.5
.001
35%-3reps
202.9d/s
121 4d/s
12.5
.001
Total(66)-6rep
222. ld/s
130.7d/s
20.9
.001
Total(66)-3rep
206.8d/s
126.9d/s
18.8
.001
In addition, the mean CVpeak was compared between 3 and 6 repetitions for each
effort level. Results indicated that the CV based on 6 repetitions was significantly higher
for both maximal and submaximal efforts when collapsed across groups (Maximal:
t(65)=2.94; p< 005: Submaximal: t(65)=3.22; p< 005) ( see Table 3).
Analysis of Variance
The mean CVpeak between effort levels and across resistance levels are presented
in Figure 1 (based on 6 reps). Similar findings were evident for the calculation of the CV
based on 3 repetitions. Table 4 specifically provides the means and standard deviations for

28
Table 3
Maximal Effort
Grp 1 (15%)
Grp 2 (25%)
Grp 3 (35%)
All Groups
Maximal Effort (3)
Maximal Effort (6)
7.7 4.4
10.7 3.8
9.7 5.5
10.5 4.3
11.1 5.1
11.6 3.8
9.5 5.1*
10.8 3 .9*
Submaximal Effort
Grp 1 (15%)
Grp 2 (25%)
Grp 3 (35%)
All Groups
Submaximal Effort (3)
Submaximal Effort (6)
8.7 4.5
10.2 4.8
8.43.9
10 4.4
9.2 6
10.4 3.9
8.8 5*
10.4 4.2*
Kp<005
the CVpeak (6 reps) directly comparing maximal and submaximal effort levels. Paired
t-tests indicated that there were no significant differences for either the CVpeak based on
6 or 3 repetitions with respect to the ability to discriminate effort levels within each
resistance level (see Table 4 for 6rep t-test results).
Table 4
CVpeak: Means and Standard Deviations across Groups
Grp 1(15%)* Grp 2(25%)2 Grp 3(3 5%)3 All Groups4
Maximal Effort 10.73.8 10.54.3 11.63.8 10.83.9
Submaximal Effort 10.24.8 104.4 10.43.9 10.44.2
11(21)= -.38, p<707; 2t(21)=.57, p< 574; 3t(21)=l.ll; p<279; 4t(21)=.66,p <513
A two-way mixed repeated measures analysis of variance was used to examine
main effects across resistance and effort level as well as an effort (maximal and
submaximal) by condition (15%, 25%, and 35% resistance levels) interaction for the CV
velocity measure (see Table 5). There were no significant effects found for within (effort
level) (F(2>63)= .43, p < .516; F(2>63) = .64, p < .531) and between (resistance level)
group(F(263) = .32, p < .725) factors. Sex differences in the CV were also examined in an

29
analysis of variance. There were no significant effects found for within (effort level)
(F(i,64) = -58, p< .449; F(164) .64, p< .457) and between (sex) group (F(164) = 1.92;
p< .171) (see Table 6).
Table 5
Analysis of Variance: CV Differences/Group by Effort Level
Source df SS MS F p
Between Subject Effects
Within & Residual
63
.12
.00
Group
2
.00
.00
.32
.725
Within Subjects Effects
Within & Residual
63
.09
.00
Effort
1
.00
.00
.43
.516
Group by Effort
2
.00
.00
.64
.531
Table 6
Analysis of Variance: CV Differences/Sex By Effort Level
Source
df
SS
MS
F
P
Between Subject Effects
Within & Residual
64
.12
.00
Sex
1
.00
.00
1.92
.171
Within Subjects Effects
Within & Residual
64
.09
.00
Effort
1
.00
.00
.58
.449
Sex by Effort
1
.00
.00
.64
.457

30
+3.8
Groupl Group2 Group3
15% 25% 35%
50%
100%
Figure 1. CV Peak Across Groups

CHAPTER 4
DISCUSSION
As with the previous studies in the sincerity of effort program of research that have
been conducted in our lab, results of the present study indicated that there was a
significant difference between aggregate maximal and submaximal effort data. Participants
produced a submaximal effort at a significantly lower velocity than a maximal effort. On
average, the submaximal effort was approximately 57- 60% of the maximal effort for
calculations based on both 3 and 6 repetitions These findings are consistent with previous
research in our lab. (Robinson et al. 1994; Lin et al., 1996; Robinson et al., 1997)
Results did not support the hypothesis that the submaximal effort condition would
be more variable than the maximal effort condition or that a resistance level lower than
the 60% level previously addressed would result in increased discriminability between
effort level. Across the three resistance levels used in this study, we could not
discriminate between effort levels based on the proposed coefficient of variation measure
Previous research had suggested that a high CV indicated a lack of consistent effort
(Carlsoo, 1986; Robinson et al., 1994; Lin et al., 1996; Robinson et al., 1997). As a
result, we expected to observe a significantly higher CV in our submaximal effort
condition than in our maximal effort condition We did not find a significant difference
between effort levels. Further, our results indicated that the maximal effort condition
across resistance levels produced a slightly higher CV. This was in contrast to the results
found by Robinson et al. (1994), a study which prompted the use of isotonic strength
31

32
testing in the present study. In the study by Robinson et al., results indicated that isotonic
strength testing showed promise in the detection of submaximal effort in leg extension
testing. This study found that in addition to significant differences between aggregate
maximal and submaximal effort data, much better sensitivity was achieved with an isotonic
task. In fact, 87% of efforts were correctly classified with no false positives. Robinson et
al. had proposed that using the isotonic method allowed for the investigation of
differences in velocity variability between maximal and submaximal efforts. Again, the
proposition was that perhaps participants were less able to reproduce submaximal velocity
efforts than they were able to reproduce submaximal strength or torque efforts because of
the dynamic nature of the task.
While the present study did not replicate the results found in the Robinson et al.
isotonic leg extension study, previous research employing isokinetic testing methodology
in leg extension testing which had found viable submaximal effort detection rates (Lin et
al, 1996; Robinson et al., 1997) was also not supported. It is possible that the
musculoskeletal differences between the lower extremity (a less stable structure) and the
spinal region (a more stable structure) could have accounted for the differences in the
ability to reproduce consistent/inconsistent effort. In addition, the nature in which
participants were secured in the Dynatrac Back extension device compared to the
stabilization of the leg in the previous studies is different from both a subjective and
objective perspective. For example, in the back study, participants were more likely to
comment on the extreme nature of the stabilization procedure The extent to which this
stabilization procedure played a role in reducing the likelihood of more variability in
performance, although difficult to determine cannot be underestimated.

33
Further, although this study which employed a dynamic assessment protocol did
not replicate the results found in the leg extension studies, the findings were consistent
with previous research employing isometric testing methodology in a lumbar extension
task (Robinson et al., 1991; 1992). In these isometric lumbar extension studies, results
indicated that participants were able to reliably reproduce a submaximal effort making it
difficult to discriminate between effort level. These earlier studies had set the precedent
for pursuing a dynamic testing approach in assessment of sincerity of effort of the lumbar
musculature.
In the present study there was no specific hypothesis attempting to determine
which of the three resistance levels would produce the best submaximal effort detection
rate. Results of the present study did not indicate any significant differences between these
resistance levels with respect to discriminability. In the pilot study, a 25% resistance level
had resulted in a moderate effect size in which there was an 8%-12% difference between
maximal and submaximal effort coefficient of variation respectively. We expected that we
might find similar results in the current study or, perhaps a somewhat higher or lower
resistance level might prove more efficacious in detection of submaximal efforts.
In addition, while the CV value based on 6 repetitions was higher than the CV
based on 3 repetitions for both maximal and submaximal effort levels, there was no
difference between the two measures with respect to discriminability of effort level.
Further, the higher CV calculated for 6 repetitions may simply represent a fatigue effect
because the latter repetitions (5, 6 and 7) were added in order to calculate the CV based
on 6 repetitions. In a previous study by Robinson et al. (1997) a trend suggesting
increased stability of the CV based on a greater number of repetitions was observed. In

34
the present study there was no indication that the CV calculation based on 3 versus 6
repetitions represented any significant difference with respect to discriminability of the
measure. However, given the findings in a number of recent studies (Robinson et al.,
1994; Lin et al., 1996; Robinson et al., 1997) we believe that the CV continues to
represent a promising construct in determining effort level. Continued investigation of the
measure with respect to both its ability to discriminate effort level and its stability is
warranted
While the results of the present study do not support the findings by Robinson et
al using isotonic methodology or the previous research involving use of an isokinetic
methodology we are left to speculate as to why we were unable to discriminate between
effort level. An explanation can perhaps be found in Schmidts general assertion that as
we move more rapidly, we become more inaccurate in terms of the goal we are trying to
achieve. Schmidt (1988) has proposed that the basic laws in motor behavior may be seen
as analogous to the fundamental principles of physics. The simple laws relating the mass,
velocity and acceleration of objects when forces are applied to them for example, have
served as the cornerstone of the physical sciences, and hence they deserve a special status.
Schmidt notes that in the same way, the field of motor behavior has analogous principles
that are somehow fundamental to all the rest. These describe such things as the
relationship between the speed a limb moves and its resulting accuracy. While there are a
well defined set of simple principles that can be stated for the various branches of the
physical sciences, we should not expect something similar for the behavioral sciences, or
for motor control in particular (Schmidt, 1988). Schmidt states that for a number of
reasons, in motor control we find far fewer statements possessing sufficient generality to

35
be termed a law. One reason is that motor-control principles have been far more difficult
to discover, based as they are on data from biological systems that are more variable and
complex than the physical systems.
The speed-accuracy relationship was first discussed by Woodsworth (1899) who
found that accuracy decreased as the movement speed increased. Fitts (1954) provided a
more formal mathematical relationship. Given that the results of the present study did not
discriminate between effort level and in fact the increased variability however minor was
observed in the maximal (higher speed) effort condition, this speed-accuracy relationship
provides some rationale for the findings. While the isotonic leg extension study by
Robinson et al. did provide promising submaximal effort detection rates, perhaps it was a
function of lower extremity testing versus back extension testing.
Since this theory is based on the use of speed in motor control, perhaps it would
prove beneficial if we applied the current methodology using an isokinetic testing
paradigm in which the dependent variable would be the torque or force produced by the
subject rather than the velocity. Future research employing an isokinetic protocol in
investigation of sincerity of effort in back extension assessment is recommended. Dvir
(1997) has reported promising findings in a study investigating differentiation of
submaximal from maximal effort in an isokinetic trunk extension task. However, these
researchers have discriminated between effort levels based on eccentric/concentric ratio
differences, asserting that the highest differentiating power among the experimental
conditions was attributed to the intervelocity difference between the concentric and
eccentric contractions. Further with respect to measures used in discriminating effort
level, Dvir and David (1996) note that consistency as defined by the absolute variability of

36
the test findings cannot serve as a reliable indicator for optimality of effort. In contrast,
other research has provided promising results in detecting submaximal effort in both
isokinetic and isotonic strength tasks using the CV as the sole discriminating variable (Lin
et al., 1996, Robinson et al., 1994, Robinson et al., 1997)
In summary, this study indicated that an isotonic assessment protocol like the
isometric assessment strategies previously used were not effective in discriminating
between maximal and submaximal effort levels. This may be due to both stabilization and
musculoskeletal between the leg and back structures. While the CV measure did not
indicate differences invariability between effort levels, continued investigation of this
measurement technique is warranted
Finally, the investigation of sincerity of effort in strength assessment testing
remains an important issue with respect to issues of treatment, rehabilitation, and
compensation. Continued research in this area is necessary in order that the reliability and
validity of measurement techniques in this area of research can be improved.

REFERENCES
Agre, J., Magness, J., Hull, S., Wright, K., Baxter, T., & Patterson, R. (1987).
Strength testing with a portable dynamometer: reliability of upper and lower extremities.
Archives of Physical Medicine and Rehabilitation. 68, 454-458.
Albert, M. (1991). Eccentric muscle training in snorts and orthopaedics. New
York: Churchill Livingstone.
Battie, M., Bigos, S.J Fisher, L.D., Hanson, T.H., Jones, ME., & Ley, M.D.,
(1989). Isometric lifting strength as a predictor of individual back pain reports. Spine, 14,
851-856.
Beimbom, D.S., & Morrissey, M.C. (1988). A review of the literature related to
trunk muscle performance. Spine, 13. 655-660.
Bohannon, R.W. (1987). Differentiation of maximal from submaximal static
elbowflexor efforts by measurement variability. American Journal of Physical Medicine
and Rehabilitation. 66, 213-218.
Bohannon, R., & Smith, M. (1988). Differentiation of maximal and submaximal
knee extension efforts by isokinetic testing. Clinical Biomechanics. 1, 16-24.
Brea, S. & Meacham, A. (1990). Evaluation of function and disability. In J.
Bonica, J. Loeser, & C .R. Chapman (Eds ), The management of pain (pp.641-648).
Philadelphia: Lea and Fiberger
Cantor, A., Gold, B Gueron, M., Cristal, N., Praqjrod, G., & Shapiro, Y. (1987).
Isotonic (dynamic) and isometric effort in the assessment and evaluation of diastolic
hypertension: correlation and clinical use. Cardiology, 74. 141-146.
Carlsoo, S. (1986). With what degree can voluntary static force be repeated.
Scandinavian Journal of Rehabilitation Medicine, 18. 1-13.
Carlton, L.G., & Newell, KM. (1988) Force variability and movement accuracy
in space-time. Journal of Experimental Psychology, 14. 24-36.
Cassisi, J.E., Robinson, M E., OConnor, P.D., & MacMillan, MM. (1993).
Trunk strength and lumbar paraspinal muscle activity during isometric exercise in chronic
low-back pain patients and controls. Spine. 18. 245-251.
37

38
Chengalur, S.N., Smith, G.A., Nelson, R.C., & Sadoff, A M. (1990). Assessing
sincerity of effort in maximal grip strength tests. American Journal of Physical Medicine
& Rehabilitation. 69. 148-153.
David, G., Dvir, Z., Mackintosh, S., & Brien, C. (1996) Validity study of a novel
test protocol for the identification of submaximal muscular effort. Isokinetics and
Exercise. 6. 139-144.
Davies, G.J., Wilk, K., & Ellenbecker, T.S. (1997). Assessment of strength. In
T.R. Malone, T. McPoil, & A.J Nitz (Eds ), Orthopedic and sports physical therapy (3rd
ed., pp. 217-233). St Louis: Mosby-Year Book, Inc.
Dvir, Z. (1997a). Differentiation of submaximal from maximal trunk extension
effort: An Isokinetic study using a new testing protocol. Spine. 22. 2672-2676.
Dvir, Z. (1997b). An isokinetic study of submaximal effort in elbow flexion.
Perceptual and Motor Skills. 84. 1431-1438.
Dvir, Z & David, G. (1996). Suboptimal muscular performance: Measuring
isokinetic strength of knee extensors with new testing protocol. Archives of Physical
Medicine and Rehabilitation. 77. 578-581.
Fitts, P.M. (1954). The information capacity of the human motor system in
controlling the amplitude of movement. Journal of Experimental Psychology. 47. 381-
391.
Flint, M. (1955). Effect of increasing back and abdominal muscle strength on low
back pain. Research Quarterly. 29. 160-171.
Frymoyer, J.W., & Cats-Baril, W.L. (1987). Predictors of low back pain
disability. Clinics of Orthopedics. 221. 89-98.
Frymoyer, J.W., & Cats-Baril, W.L. (1991). An overview of the incidences and
costs of low back pain. Orthopedic Clinics of North America. 22. 263-271
Gilbert, J.C., & Knowlton, R.G. (1983). Simple method to determine sincerity of
effort during a maximal isometric test of grip strength. American Journal of Physical
Medicine and Rehabilitation. 62, 135-144.
Greenleaf, J.E., Bemauer, E.M., Ertl, A.C., Trowbridge, T.S., & Wade, C.E.
(1989). Work capacity during 30 days of bedrest with isotonic and isokinetic exercise
training. Journal of Applied Physiology, 67. 1820-1826.
Hammond, H.K., & Froelicher, V.F. (1985). Normal and abnormal heart rate
responses to exercise. Progress in Cardiovascular Disease. 27. 271-196.

39
Harber, P., & Soohoo, K. (1984). Static ergonomic strength testing in evaluating
occupational back pain. Journal of Occupational Medicine. 26. 887-884
Hazard R., Reid, S., Fenwick, J., & Reeves, V (1988). Isokinetic trunk and lifting
strength measurements: Variability as an indicator of effort. Spine. 13. 54-57.
Holmes, B Leggett, S., Mooney, V., Nichols, J., Negri, S., & Hoeyberghs, A.
(1996). Comparison of female geriatric lumbar extension strength: Asymptomatic versus
chronic low back pain patients and their response to active rehabilitation. Journal of
Spinal Disorders, 9, 17-22.
Kishino, N.D., Mayer, T.G., Gatchel, R.J., McCrate Parrish, M., Anderson, C ,
Gustin, L., & Mooney, V. (1985). Quantification of lumbar function. Part 4: Isometric
and isokinetic lifting simulation in normal subjects and low-back dysfunction patients.
Spine. 10. 921-927.
Kluck, D.J. (1967). A study of the strength ratio of the back extensors to the
trunk flexors. Masters Thesis. Department of Physical Therapy, University of Iowa, Iowa
City.
Lin, P C., Robinson, M E., Carlos, J., & O'Connor, P. (1996) Detection of
submaximal efforts in isometric and isokinetic knee extension tests. Journal of
Orthopaedic and Sports Physical Therapy. 24, 19-24.
Liu, A.C., & Byrne, E., (1995). Cost of care for ambulatory patients with low
back pain. The Journal of Family Practice. 40. 449-455.
Mayer, L. Greenberg, B.B. (1942). Measurements of the strengths of trunk
muscles. American Journal of Bone and Joint Surgery. 24. 842-856.
Mayer, T.G., Barnes, D., Kishino, N.D., Nichols, G., Gatchel, R.J., Mayer, H., &
Mooney, V. (1988). Progressive isoinertial evaluation in a standardized protocol and
normative data base. Spine. 13. 993-997.
Mayer, T.G., Smith, S.S., Keley, P.T., & Mooney, V. (1985). Quantification of
lumbar function: Part 2. Saggital plane trunk strength in chronic low back pain patients.
Spine. 10. 765-772.
McIntyre, D.R., Glover, L.H., Seeds, R.H., & Levene, J.A. (1990). The
characteristics of preferred low back motion. Journal of Spinal Disorders. 3. 145-155.
Mendelson, G. (1995). Compensation neurosis revisited: Outcome studies of the
effects of litigation. Journal of Psychosomatic Research. 39. 695-706.

40
Mooney, V.M., & Andersson, G.B.J. (1994). Controversies: Trunk strength
testing in patient evaluation and treatment. Spine. 19. 2483-2485.
Nachemson, A.L., & Lindh, M. (1969). Measurement of abdominal and back
muscle strength with and without low back pain. Scandinavian Journal of Rehabilitation
Medicine. 1. 60-65.
Newell, K.M., & Carlton, L.G. (1988). Force variability in isometric responses.
Journal of Experimental Psychology: Human Perception and Performance. 1. 37-44.
Newell, K.M., Carlton, L.G., & Hancock, P.A. (1984). Kinetic analysis of
response variability. Psychology Bulletin. 1. 133-151.
Niebuhr, B., & Marion, R. (1990). Voluntary control of submaximal grip
strength. American Journal of Physical Medicine & Rehabilitation. 69, 96-101.
Portney, L G., & Watkins, M.P (1993). Foundations of clinical research and
applications to practice. Norwalk, CT: Appleton & Lange.
Richards, J.G., Quigley, E.J., & Castaqno, P.W. (1996). Validity and reliability of
the BTE Dynatrac. Medical Science Sports Exercise. 28. 913-920.
Robinson, M E., Cassisi, J.E., OConnor, P.D., & MacMillan, M. (1992).
Lumbar iEMG during isotonic exercise: Chronic low back pain patients versus controls.
Journal of Spinal Disorders. 5. 8-15.
Robinson, ME., Geisser, M E., Hanson, C.S., & O'Connor, P. (1993). Detecting
submaximal efforts in grip strength testing with the coefficient of variation. Journal of
Occupational Rehabilitation. 3. 45-50.
Robinson, M.E., MacMillan, M., O'Connor, P, Fuller, A., & Cassisi, J. (1991).
Reproducibility of maximal versus submaximal efforts in an isometric lumbar extension
task. Journal of Spinal Disorders. 4. 444-448.
Robinson, M E., O'Connor, P., MacMillan, M., Fuller, A., & Cassisi, J. (1992).
Effect of instructions to simulate a back injury on torque reproducibility in an isometric
lumbar extension task. Journal of Occupational Rehabilitation. 2. 1-9.
Robinson, M.E., O'Connor, P, Riley, J.L., Kvaal, S., & Shirley, F.R. (1994).
Variability of isometric and isotonic leg exercise: Utility for detection of submaximal
efforts. Journal of Occupational Rehabilitation. 4, 163-169.
Robinson, M E., Sadler, I.J., OConnor, P., & Riley, J.L. (1997). Detection of
submaximal effort and assessment of stability of the coefficient of variation. Journal of
Occupational Rehabilitation. 7. 207-215.

41
Sanders, B. (1997). Exercise and Rehabilitation concepts. In T.R. Malone, T.
McPoil, & A.J. Nitz (Eds.), Orthopedic and sports physical therapy (3rd ed., pp 236-247).
St Louis: Mosby-Year Book, Inc.
Schmidt, D.E., Zelaznik, H.N., Hawkins, B Frank, J.S., & Zuinn, J.T. (1979).
Motor output variability: A theory for the accuracy of rapid motor acts. Psychology
Review. 86. 415-451.
Schmidt, R.A. (1988). Motor control and learning: A behavioral emphasis.
Urbana-Champaign, IL: Human Kinetics Publishers.
Sherwood, D.E., & Schmidt, R.A. (1980). The relationship between force and
force variability in minimal and near maximal static and dynamic contractions. Journal of
Motor Behavior. 12. 75-89
Sherwood, D.E., Schmidt, R.A., & Walter, C.B (1988). The force/force-
variability relationship under controlled temporal conditions. Journal of Motor Behavior.
20, 106-116
Simonsen, J.C. (1995). Coefficient of variation as a measure of subject effort.
Archives of Physical Medicine and Rehabilitation. 76. 516-520.
Smith, G.A., Nelson, R.C., Sadoff, S.J., & Sadoff, A M. (1989). Assessing
sincerity of effort in maximal grip strength tasks. American Journal of Physical Medicine
and Rehabilitation. 68. 73-80.
Stokes, I.A.F., Rush, S., Moffioid, M., Johnson, G.B., & Haugh, L.D. (1987).
Trunk extensor EMG-torque relationship. Spine. 12. 770-776.
Webster, B.S., & Snook, S.H. (1994). The cost of 1989 worker's compensation
low back pain claims. Spine. 19. 1011-1016.
Woodsworth, R.S. (1899). The accuracy of voluntary movement. Psychology
Review. 3 (Suppl. 2).

BIOGRAPHICAL SKETCH
Ian John Sadler was bom in Georgetown, Guyana, S.A., on March 16, 1966. He
emigrated to Toronto, Ontario, in May of 1976. Mr. Sadler graduated from Silverthom
Collegiate Institute in Toronto in June of 1980. He attended McMaster University in
Hamilton, Ontario, where he completed his Bachelor of Science degree in honors biology
and psychology in 1990. Following graduation, he departed for Tokyo, Japan, where he
spent one year as an English teacher before returning to Toronto in the fall of 1991. Upon
his return, he spent one year conducting research at the University of Toronto before
entering graduate school. In the fall of 1992 he entered Connecticut College in New
London, Connecticut, where he completed his Master of Arts degree in psychology. He
entered the graduate program in clinical and health psychology at the University of Florida
in 1994. He will be completing his predoctoral internship training year at the University of
California, San Diego, starting July, 1, 1998.
42

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
lichael E. Robinson, Chair
Associate Professor of Clinical and Health
Psychology
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
£ < \
Duane E. Dede
Assistant Professor of Clinical and Health
Psychology
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Jo
Assistant Professor of Clinical and Health
Psychology
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Samuel F. Sears
Assistant Professor of Clinical and Health
Psychology
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.

This dissertation was submitted to the Graduate Faculty of the College of Health
Professions and to the Graduate School and was accepted as partial fulfillment of the
requirements for the degree of Doctor of Philosophy.
August 1999
Dean, College of Health Professions
Dean, Graduate School

UNIVERSITY lllllllllllUl!
l! 08554 7833



CHAPTER 3
RESULTS
Demographics
There were 66 participants in the study (see Table 1 for breakdown). The mean
age of participants was 26.2 years (SD=6.4); mean height was 171.9 cm (SD=9.9cm); and
mean weight was 67.4 kg. (SD=15.7kg). A manipulation check which required subjects
to state the order of effort conditions indicated that all participants followed the assigned
protocol.
Table 1
Number of Participants bv Group
Group 1 (15%)
Group 2 (25%)
Group 3 (35%)
Total
Males 9
9
10
28
Females 13
13
12
38
N=66
Coefficient of Variation
The CVpeak velocity for 6 repetitions was calculated using repetitions 2-7. The
CVpeak velocity for 3 repetitions was calculated using repetitions 2-4. Trial 1 was not
used in the calculations as this was not considered to be an accurate record of effort level
given the possibility that acclimation to the protocol was necessary.
26


21
weight. Results indicated that first, participants were more accurate in their attempt at a
50% effort compared to the previous study using a 60% resistance level. In addition, as in
the previous study using 20 subjects, there was a significant difference in the velocities
produced between maximal and submaximal effort conditions. Finally, results indicated
that there was a clear trend toward significant differences (p<06) between CV of
maximal and submaximal effort conditions (d=.80).
The Present Study
The above literature review suggests that the dynamic isotonic testing method
appears to show the most promise Therefore this study considered detection of
submaximal effort in an isotonic back strength test. Given the results of the more recent
studies using different resistance levels, an attempt was made to determine an optimal
resistance level that might lead to improved detection of submaximal effort. In addition, as
a result of trends in recent findings, suggesting somewhat increased stability, a larger
number of repetitions (6) was used in the determination of the CV. Previous results had
raised doubt about the stability of using a low number of repetitions (usually three) to
calculate the CV. The calculation of this measure on such a small number of trials raised
the possibility that the measure might be very unstable
First, it was hypothesized that submaximal efforts would be more variable than
maximal efforts across resistance levels. It was also hypothesized that a resistance level
lower than the 60% level previously addressed, would result in increased discriminability
between effort level. Finally, it was hypothesized that use of the CVpeak 6 trial measure
would provide better discriminability between effort level than the CVpeak 3 trial measure


16
rate of 80%. The more recent study represents an improvement in CVpeak classification
rates compared to the study by Lin et al This CVpeak measure was used in the present
study.
In addition, based on these recent results, it appears that a lower CV cutoff (10%)
than the standard 15% cutoff might be warranted in the detection of submaximal efforts.
Determination of classification stability also indicated that a 10% CV cutoff appears to be
the optimal detection rate at which no false positives were obtained.
Concerns regarding the stability of using a low number of repetitions (usually
three) to calculate the CV were also addressed in the study by Robinson et al. (1997).
The calculation of this measure on such a small number of trials raises the possibility that
the measure may be very unstable. Results indicated that use of 5 repetitions versus 3
repetitions in the calculation of CVpeak resulted in increased stability across trials. Results
indicated that the CVave measure (calculated on 3 trials only) was unstable. The trend in
these findings suggest that increasing the number of repetitions used in CV calculation
may prove beneficial.
The previous research discussed suggests that use of a dynamic methodology
appears to provide better discrimination of effort level. The following discussion provides
a helpful analysis and summary of the dynamics proposition
A Final Comment on Dynamics
Research has suggested that the specific relationship between the force produced
and force variability has been proposed to be linear (Schmidt, Zelaznik, Hawkins, Frank,&
Zuinn, 1979), an inverted-U (Sherwood & Schmidt, 1980), or a curvilinear function
(Sherwood, Schmidt, & Walter, 1988), with variability generally decreasing as function of


36
the test findings cannot serve as a reliable indicator for optimality of effort. In contrast,
other research has provided promising results in detecting submaximal effort in both
isokinetic and isotonic strength tasks using the CV as the sole discriminating variable (Lin
et al., 1996, Robinson et al., 1994, Robinson et al., 1997)
In summary, this study indicated that an isotonic assessment protocol like the
isometric assessment strategies previously used were not effective in discriminating
between maximal and submaximal effort levels. This may be due to both stabilization and
musculoskeletal between the leg and back structures. While the CV measure did not
indicate differences invariability between effort levels, continued investigation of this
measurement technique is warranted
Finally, the investigation of sincerity of effort in strength assessment testing
remains an important issue with respect to issues of treatment, rehabilitation, and
compensation. Continued research in this area is necessary in order that the reliability and
validity of measurement techniques in this area of research can be improved.


32
testing in the present study. In the study by Robinson et al., results indicated that isotonic
strength testing showed promise in the detection of submaximal effort in leg extension
testing. This study found that in addition to significant differences between aggregate
maximal and submaximal effort data, much better sensitivity was achieved with an isotonic
task. In fact, 87% of efforts were correctly classified with no false positives. Robinson et
al. had proposed that using the isotonic method allowed for the investigation of
differences in velocity variability between maximal and submaximal efforts. Again, the
proposition was that perhaps participants were less able to reproduce submaximal velocity
efforts than they were able to reproduce submaximal strength or torque efforts because of
the dynamic nature of the task.
While the present study did not replicate the results found in the Robinson et al.
isotonic leg extension study, previous research employing isokinetic testing methodology
in leg extension testing which had found viable submaximal effort detection rates (Lin et
al, 1996; Robinson et al., 1997) was also not supported. It is possible that the
musculoskeletal differences between the lower extremity (a less stable structure) and the
spinal region (a more stable structure) could have accounted for the differences in the
ability to reproduce consistent/inconsistent effort. In addition, the nature in which
participants were secured in the Dynatrac Back extension device compared to the
stabilization of the leg in the previous studies is different from both a subjective and
objective perspective. For example, in the back study, participants were more likely to
comment on the extreme nature of the stabilization procedure The extent to which this
stabilization procedure played a role in reducing the likelihood of more variability in
performance, although difficult to determine cannot be underestimated.


35
be termed a law. One reason is that motor-control principles have been far more difficult
to discover, based as they are on data from biological systems that are more variable and
complex than the physical systems.
The speed-accuracy relationship was first discussed by Woodsworth (1899) who
found that accuracy decreased as the movement speed increased. Fitts (1954) provided a
more formal mathematical relationship. Given that the results of the present study did not
discriminate between effort level and in fact the increased variability however minor was
observed in the maximal (higher speed) effort condition, this speed-accuracy relationship
provides some rationale for the findings. While the isotonic leg extension study by
Robinson et al. did provide promising submaximal effort detection rates, perhaps it was a
function of lower extremity testing versus back extension testing.
Since this theory is based on the use of speed in motor control, perhaps it would
prove beneficial if we applied the current methodology using an isokinetic testing
paradigm in which the dependent variable would be the torque or force produced by the
subject rather than the velocity. Future research employing an isokinetic protocol in
investigation of sincerity of effort in back extension assessment is recommended. Dvir
(1997) has reported promising findings in a study investigating differentiation of
submaximal from maximal effort in an isokinetic trunk extension task. However, these
researchers have discriminated between effort levels based on eccentric/concentric ratio
differences, asserting that the highest differentiating power among the experimental
conditions was attributed to the intervelocity difference between the concentric and
eccentric contractions. Further with respect to measures used in discriminating effort
level, Dvir and David (1996) note that consistency as defined by the absolute variability of


22
given trends in recent findings suggesting moderate increases in stability with increased
number of repetitions used in CV calculation


41
Sanders, B. (1997). Exercise and Rehabilitation concepts. In T.R. Malone, T.
McPoil, & A.J. Nitz (Eds.), Orthopedic and sports physical therapy (3rd ed., pp 236-247).
St Louis: Mosby-Year Book, Inc.
Schmidt, D.E., Zelaznik, H.N., Hawkins, B Frank, J.S., & Zuinn, J.T. (1979).
Motor output variability: A theory for the accuracy of rapid motor acts. Psychology
Review. 86. 415-451.
Schmidt, R.A. (1988). Motor control and learning: A behavioral emphasis.
Urbana-Champaign, IL: Human Kinetics Publishers.
Sherwood, D.E., & Schmidt, R.A. (1980). The relationship between force and
force variability in minimal and near maximal static and dynamic contractions. Journal of
Motor Behavior. 12. 75-89
Sherwood, D.E., Schmidt, R.A., & Walter, C.B (1988). The force/force-
variability relationship under controlled temporal conditions. Journal of Motor Behavior.
20, 106-116
Simonsen, J.C. (1995). Coefficient of variation as a measure of subject effort.
Archives of Physical Medicine and Rehabilitation. 76. 516-520.
Smith, G.A., Nelson, R.C., Sadoff, S.J., & Sadoff, A M. (1989). Assessing
sincerity of effort in maximal grip strength tasks. American Journal of Physical Medicine
and Rehabilitation. 68. 73-80.
Stokes, I.A.F., Rush, S., Moffioid, M., Johnson, G.B., & Haugh, L.D. (1987).
Trunk extensor EMG-torque relationship. Spine. 12. 770-776.
Webster, B.S., & Snook, S.H. (1994). The cost of 1989 worker's compensation
low back pain claims. Spine. 19. 1011-1016.
Woodsworth, R.S. (1899). The accuracy of voluntary movement. Psychology
Review. 3 (Suppl. 2).


9
Static Testing
Clinical practice has generally studied variability in effort level in isometric tasks. A
number of studies have used the grip test strength assessment task in determination of
effort level (Gilbert & Knowlton, 1983, Bohannon, 1987; Smith, Nelson, Sadoff, & Sadoff,
1989; Chengalur et al. 1990; Niebuhr & Marion, 1990; Robinson, Geisser, Hanson, &
OConnor, 1993). Robinson et al. (1993) used the CV to discriminate level of effort in grip
strength testing. Twenty-nine asymptomatic subjects participated in two conditions of
testing: 100% and 50% effort. The submaximal (50%) effort condition showed
significantly more variability than the maximal effort condition in both sets of conditions.
Intra-class correlation coefficients were very low for both maximal effort and submaximal
efforts (.036 and .025) indicating very low stability for the coefficient of variation.
Classification rates were also found to have unacceptably large errors with 69% of the
submaximal efforts being classified as maximal with the traditional 15% cutoff, and 55%
misclassification of submaximal efforts with an optimized 11% cutoff It was concluded
that the currently practiced method of using a low number of repetitions to calculate the
CV might result in very unstable measures. Further, grip strength testing, which is an
isometric testing procedure, did not produce consistent results (suitable detection rates)
that might validate its use in a clinical setting. Although its use continues in clinical
practice the research does not adequately support such use.
While grip strength testing may provide helpful information, when we consider the
costs and prevalence of musculoskeletal injury it is important to think about the relevance
of sincerity of effort testing from a more ecologically valid perspective, that is, the
assessment of the spinal region. Because of its significant economic impact, assessment of


20
Given the lack of research involving use of isotonic methodology in determining
sincerity of effort, as well as a similar paucity of research assessing this concept in the
spine musculature, a recent pilot study was completed in our lab at the University of
Florida.
Sadler, Robinson, Otis, OConnor and Riley (unpublished data) assessed lumbar
strength variability in 20 asymptomatic controls, using the Dynatrac Back Tester (BTE
Dynatrac, Hanover, Maryland). Participants were asked to complete 6 maximal and 6
submaximal isotonic lumbar extension exercises with a 10 minute break between trials.
Order of condition was counterbalanced. A resistance level of 60% of participants body
weight was set for each subject.
While a significant difference was found between maximal and submaximal efforts
with regard to velocity, there were no significant differences found in the CV between
effort conditions. Accurate classification to effort condition was not possible. These
results clearly did not support recent findings using isokinetic leg extension testing
(Robinson et ai., 1997). It was speculated that perhaps the anatomy of the spine and its
musculature, might somehow result in less variability being observed in back strength
testing.
As a result of the poor discriminability found in this recent study, a follow-up pilot
study was conducted. It was observed at the time of testing in the previous study, that a
number of subjects were experiencing some discomfort with the procedure. As a result, it
was speculated that a lowered resistance level might provide a better opportunity to
discriminate maximal from submaximal efforts. Eight subjects participated in the pilot
study with the same methodology; however, resistance level was set at 25% of body


38
Chengalur, S.N., Smith, G.A., Nelson, R.C., & Sadoff, A M. (1990). Assessing
sincerity of effort in maximal grip strength tests. American Journal of Physical Medicine
& Rehabilitation. 69. 148-153.
David, G., Dvir, Z., Mackintosh, S., & Brien, C. (1996) Validity study of a novel
test protocol for the identification of submaximal muscular effort. Isokinetics and
Exercise. 6. 139-144.
Davies, G.J., Wilk, K., & Ellenbecker, T.S. (1997). Assessment of strength. In
T.R. Malone, T. McPoil, & A.J Nitz (Eds ), Orthopedic and sports physical therapy (3rd
ed., pp. 217-233). St Louis: Mosby-Year Book, Inc.
Dvir, Z. (1997a). Differentiation of submaximal from maximal trunk extension
effort: An Isokinetic study using a new testing protocol. Spine. 22. 2672-2676.
Dvir, Z. (1997b). An isokinetic study of submaximal effort in elbow flexion.
Perceptual and Motor Skills. 84. 1431-1438.
Dvir, Z & David, G. (1996). Suboptimal muscular performance: Measuring
isokinetic strength of knee extensors with new testing protocol. Archives of Physical
Medicine and Rehabilitation. 77. 578-581.
Fitts, P.M. (1954). The information capacity of the human motor system in
controlling the amplitude of movement. Journal of Experimental Psychology. 47. 381-
391.
Flint, M. (1955). Effect of increasing back and abdominal muscle strength on low
back pain. Research Quarterly. 29. 160-171.
Frymoyer, J.W., & Cats-Baril, W.L. (1987). Predictors of low back pain
disability. Clinics of Orthopedics. 221. 89-98.
Frymoyer, J.W., & Cats-Baril, W.L. (1991). An overview of the incidences and
costs of low back pain. Orthopedic Clinics of North America. 22. 263-271
Gilbert, J.C., & Knowlton, R.G. (1983). Simple method to determine sincerity of
effort during a maximal isometric test of grip strength. American Journal of Physical
Medicine and Rehabilitation. 62, 135-144.
Greenleaf, J.E., Bemauer, E.M., Ertl, A.C., Trowbridge, T.S., & Wade, C.E.
(1989). Work capacity during 30 days of bedrest with isotonic and isokinetic exercise
training. Journal of Applied Physiology, 67. 1820-1826.
Hammond, H.K., & Froelicher, V.F. (1985). Normal and abnormal heart rate
responses to exercise. Progress in Cardiovascular Disease. 27. 271-196.


minute rest period following the first set of 7 repetitions and order of conditions was
counterbalanced. Results indicated that while participants were able to produce a
submaximal effort, there was not a significant difference in variability between effort levels
across the three resistance levels. There was no difference in the use of the CV based on
6 versus 3 repetitions with respect to effort discriminability. Implications of these results
are discussed with recommendations for continued investigation of effort level in backs
through use of isokinetic testing methodology.
vi


29
analysis of variance. There were no significant effects found for within (effort level)
(F(i,64) = -58, p< .449; F(164) .64, p< .457) and between (sex) group (F(164) = 1.92;
p< .171) (see Table 6).
Table 5
Analysis of Variance: CV Differences/Group by Effort Level
Source df SS MS F p
Between Subject Effects
Within & Residual
63
.12
.00
Group
2
.00
.00
.32
.725
Within Subjects Effects
Within & Residual
63
.09
.00
Effort
1
.00
.00
.43
.516
Group by Effort
2
.00
.00
.64
.531
Table 6
Analysis of Variance: CV Differences/Sex By Effort Level
Source
df
SS
MS
F
P
Between Subject Effects
Within & Residual
64
.12
.00
Sex
1
.00
.00
1.92
.171
Within Subjects Effects
Within & Residual
64
.09
.00
Effort
1
.00
.00
.58
.449
Sex by Effort
1
.00
.00
.64
.457


2
more than twice the amount for the average workers compensation claim ($4075).
Further the median cost per low back pain case was $396, an indication that low back pain
costs are not normally distributed with a few cases accounting for most of the costs.
Clinical Assessment of Force
As health care providers in the current managed care environment attempt to care
for patients with musculoskeletal disorders in a more cost effective manner, it is becoming
increasingly important to identify which diagnostic and therapeutic procedures have the
greatest impact on the cost of caring for patients with low back pain (Liu & Byrne, 1995).
The measurement of applied force produced by voluntary muscular contraction across a
range of motion is a common method used to assess the status of the musculoskeletal
system. Strength testing is an important part of an assessment battery for evaluation of
patients with chronic low back pain. A number of testing methods have been investigated
including isotonic, isokinetic and isometric testing protocols. A brief description of these
testing approaches is provided.
Isotonic. Isokinetic, and Isometric Testing
Various forms of musculoskeletal exercise are available for the researcher to use
when assessing muscle strength. Today, three primary types of exercise are commonly
employed: isometrics, isotonics, and isokinetics (Davies, Wilk, & Ellenbecker, 1997).
While isotonic methodology will be used in the present study, a considerable amount of
the research has involved the two other testing approaches. As a result, definitions of all
three approaches follow:


25
First, the participants were secured to the Dynatrac back testing device. Patients
were placed in a comfortable, semi-kneeling posture with the pelvis at a 35 degree forward
tilt. This placed the spine in a neutral position, maintaining the lumbar lordosis and the
normal angle of the sacrum The semi-kneeling posture allowed full unrestricted flexion
and extension. Simple adjustments accommodated any size participant, tall or short.
Padded fixation points were adjusted for variations in limb length, pelvis height, pelvis
width, and shoulder height, while isolating the back muscles for testing. With this patient
positioning system, test subjects were easily positioned in less than a minute with solid
stabilization.
Once secured in the back tester, participants were told that 3 practice repetitions
would be performed to ensure understanding of the protocol. This was followed by a one
minute rest period Participants were then asked to perform 7 repetitions at a 50% effort
level (submaximal) and 7 repetitions at a 100% effort level. Order of conditions was
counterbalanced and separated by a 10 minute rest period. Finally, upon completion of
the trials subjects were required to indicate the order of effort level that they were
assigned.


This dissertation was submitted to the Graduate Faculty of the College of Health
Professions and to the Graduate School and was accepted as partial fulfillment of the
requirements for the degree of Doctor of Philosophy.
August 1999
Dean, College of Health Professions
Dean, Graduate School


BIOGRAPHICAL SKETCH
Ian John Sadler was bom in Georgetown, Guyana, S.A., on March 16, 1966. He
emigrated to Toronto, Ontario, in May of 1976. Mr. Sadler graduated from Silverthom
Collegiate Institute in Toronto in June of 1980. He attended McMaster University in
Hamilton, Ontario, where he completed his Bachelor of Science degree in honors biology
and psychology in 1990. Following graduation, he departed for Tokyo, Japan, where he
spent one year as an English teacher before returning to Toronto in the fall of 1991. Upon
his return, he spent one year conducting research at the University of Toronto before
entering graduate school. In the fall of 1992 he entered Connecticut College in New
London, Connecticut, where he completed his Master of Arts degree in psychology. He
entered the graduate program in clinical and health psychology at the University of Florida
in 1994. He will be completing his predoctoral internship training year at the University of
California, San Diego, starting July, 1, 1998.
42


30
+3.8
Groupl Group2 Group3
15% 25% 35%
50%
100%
Figure 1. CV Peak Across Groups


13
classified), consistent with previous research, those of the isotonic task were high enough
to show promise for clinical application Further, in the isotonic task there were no false
positives (maximal efforts called submaximal) and only 4 of 15 submaximal efforts were
missed using a 10% CV cutoff. This represents a significant improvement over the
isometric testing method and previous research (Robinson et al., 1991), and provides
support for the use of dynamic testing protocols.
Another recent study has investigated differences between maximal and
submaximal efforts in isometric and isokinetic knee extension using torque variability and
slope measures (Lin et al., 1996). These authors note that the use of torque variability and
slope measures to detect submaximal efforts has been studied in isometric tests, but not
fully investigated in isokinetic tests. Thirty-two healthy students were studied to determine
the extension torque of the knee during isometric and isokinetic exercise. The CV of
average torque, CV of peak torque, and slope to peak torque were obtained from maximal
and submaximal torque curves during isometric and two isokinetic tests (60/s and 1807s).
Subjects were instructed to kick as hard as they could in the maximal effort condition and
were told to give a 50% effort in the submaximal effort condition. The testing order of
maximal vs. submaximal conditions was randomized across subjects to eliminate a possible
order effect caused by fatigue.
Results of this study, which were consistent with previous findings (Robinson et
al., 1993, 1994; Smith et al., 1989), indicated significant differences between effort levels
(maximal and submaximal) for all variables in isometric and isokinetic tests. However, in
the isometric condition, effort level classification based on the CV method was found to
have unacceptably large errors, in that over 50% of submaximal efforts were misclassified


DETECTION OF SUBMAXIMAL EFFORT N ISOTONIC BACK STRENGTH
TESTING: DETERMINATION OF OPTIMAL RESISTANCE LEVEL
By
IAN J SADLER
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE
UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1999


CHAPTER 1
INTRODUCTION
Musculoskeletal injuries are the most common cause of frequent or permanent
impairment (Brea & Meacham, 1990). In a review of the literature, Frymoyer and Cats-
Baril (1991) reported that while low back disorders are extremely prevalent in all
societies, and probably have not increased substantially over the last two decades, they
note that what has increased is the rate of disability, the reasons for which are uncertain.
According to these researchers what has changed is societys perception, most particularly
relating to the disability that results. With an increasing rate of disability, the costs have
increased to industry and government, and there has been an increased utilization of
medical care, including surgical intervention. While Frymoyer and Cats-Baril (1987)
reported a dramatic increase in disability of 14 times the rate of population growth from
1977 to 1981, further research has suggested that this disability rate relative to population
growth continues to increase (Webster & Snook, 1994).
Although a precise estimate is impossible, it is plausible that the direct medical and
indirect costs of these conditions are in the range of more then $50 billion per annum, and
could be as high as $100 billion at the extreme (Frymoyer & Cats-Baril, 1991). Of these
costs, 75% or more can be attributed to the 5% of people who become disabled
temporarily or permanently from back pain. Further, it is estimated that the average
worker's compensation claim results in costs of $4075 annually (Webster & Snook, 1994).
Webster and Snook reported the mean cost per case of low back pain was $8321,
1


> CONTENTS
ACKNOWLEDGMENTS
ABSTRACT
CHAPTERS
1 INTRODUCTION.
Clinical Assessment of Force
The Present Study
2 METHOD
Participants
BIOGRAPHICAL SKETCH


27
Submaximal vs. Maximal Effort Level
Mean velocity scores were calculated across repetitions for each group and in each
condition using both 3 and 6 repetitions. These scores were subjected to paired t-tests to
assess differences between conditions. Results indicated that the submaximal effort
yielded significantly lower velocity than the maximal effort across groups as well as
number of repetitions used in calculation (see Table 2).
Table 2
Mean Velocity Differences between Test Conditions
Group
Maximal
Submaximal
t(21)
P
15%-6 reps
221. ld/s
127.2d/s
10.2
.001
15%-3reps
206.3d/s
126.9d/s
18.8
.001
25%-6reps
225.8d/s
140.6d/s
12.4
.001
25%-3reps
211.2d/s
136.5d/s
11.5
.001
35%-6reps
219.4d/s
124 3d/s
14.5
.001
35%-3reps
202.9d/s
121 4d/s
12.5
.001
Total(66)-6rep
222. ld/s
130.7d/s
20.9
.001
Total(66)-3rep
206.8d/s
126.9d/s
18.8
.001
In addition, the mean CVpeak was compared between 3 and 6 repetitions for each
effort level. Results indicated that the CV based on 6 repetitions was significantly higher
for both maximal and submaximal efforts when collapsed across groups (Maximal:
t(65)=2.94; p< 005: Submaximal: t(65)=3.22; p< 005) ( see Table 3).
Analysis of Variance
The mean CVpeak between effort levels and across resistance levels are presented
in Figure 1 (based on 6 reps). Similar findings were evident for the calculation of the CV
based on 3 repetitions. Table 4 specifically provides the means and standard deviations for


19
methods, they reported only five older studies using this methodology (Flint, 1955; Kluck,
1967; Mayer & Greenberg, 1942; Nachemson, & Lindh, 1969). Considering the paucity
of research investigating isotonic methodology in assessment of trunk musculature torque
producing capacity, it is not surprising that there are even fewer studies investigating the
specific area of sincerity of effort. In fact, an extensive search of the literature considering
isotonic methodology indicates that the majority of this research has used this paradigm in
evaluation of rehabilitation/exercise treatment programs for medical patients with back
problems (Greenleaf, Bemauer, Ertl, Trowbridge, & Wade, 1989; Holmes et al., 1996)
and heart disease (Hammond & Froelicher, 1985; Cantor et al., 1987).
However, Robinson, Cassisi, OConnor & MacMillan (1992), in a study
investigating lumbar iEMG during isotonic exercise of the lumbar spine, have suggested
that while not a direct comparison of the clinical efficacy of isotonic versus isokinetic
testing, their study establishes that clinical isotonic testing can be used in a fashion similar
to clinical isokinetic testing. Further, Robinson et al. (1994) suggest that using the
isotonic method allows for the investigation of differences in velocity variability between
maximal and submaximal efforts. It may be that subjects are less able to reliably
reproduce submaximal velocity efforts than they are able to reproduce submaximal
strength or torque efforts because of the dynamic nature of the task (Robinson et al.,
1994). In addition, the use of dynamic tasks may better represent the normal use of the
joint system, since isometric muscle contractions do not represent a high frequency
behavior.


34
the present study there was no indication that the CV calculation based on 3 versus 6
repetitions represented any significant difference with respect to discriminability of the
measure. However, given the findings in a number of recent studies (Robinson et al.,
1994; Lin et al., 1996; Robinson et al., 1997) we believe that the CV continues to
represent a promising construct in determining effort level. Continued investigation of the
measure with respect to both its ability to discriminate effort level and its stability is
warranted
While the results of the present study do not support the findings by Robinson et
al using isotonic methodology or the previous research involving use of an isokinetic
methodology we are left to speculate as to why we were unable to discriminate between
effort level. An explanation can perhaps be found in Schmidts general assertion that as
we move more rapidly, we become more inaccurate in terms of the goal we are trying to
achieve. Schmidt (1988) has proposed that the basic laws in motor behavior may be seen
as analogous to the fundamental principles of physics. The simple laws relating the mass,
velocity and acceleration of objects when forces are applied to them for example, have
served as the cornerstone of the physical sciences, and hence they deserve a special status.
Schmidt notes that in the same way, the field of motor behavior has analogous principles
that are somehow fundamental to all the rest. These describe such things as the
relationship between the speed a limb moves and its resulting accuracy. While there are a
well defined set of simple principles that can be stated for the various branches of the
physical sciences, we should not expect something similar for the behavioral sciences, or
for motor control in particular (Schmidt, 1988). Schmidt states that for a number of
reasons, in motor control we find far fewer statements possessing sufficient generality to


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
DETECTION OF SUBMAXIMAL EFFORT IN ISOTONIC BACK STRENGTH
TESTING: DETERMINATION OF OPTIMAL RESISTANCE LEVEL
By
Ian J. Sadler
August 1999
Chairperson: Michael E. Robinson
Major Department: Clinical and Health Psychology
The purpose of this study was to investigate the use of an isotonic back strength
assessment task in the detection of submaximal effort and to test the hypothesis that a
submaximal effort would yield greater variability as measured by the coefficient of
variation. There was also an examination of the optimal resistance level at which effort
level could best be discriminated. The hypothesis proposed that a resistance level of either
15%, 25%, or 35% would produce convincing results with respect to effort
discriminability. It was also hypothesized that the calculation of the CV measure based
on 6 repetitions would result in an improved submaximal effort detection rate versus a CV
calculation based on 3 repetitions. Sixty- six participants were asked to undergo back
extension assessment. Each participant completed 3 practice repetitions followed by 7
repetitions at a 50% effort level and 7 repetitions at a 100% effort level. There was a 10
v


17
the force applied. These functions have also been found in studies using dynamic tasks
(Sherwood et al., 1988; Carlton & Newell, 1988) These tasks are thought to be
inherently more complex because of increased changes involving a number of variables
including muscle length, muscle velocity, and limb displacement In their review, Newell,
Carlton and Hancock (1984) found that many functions of variability in motor control
appear, depending on the task-related relationships of movement distance, time to peak
force, magnitude of the force applied, and rate of force production. Newell and Carlton
(1988) noted that separating the effects of these variables has been difficult as a result of
the change in one variable leading to changes in one of the other variables. These effects
have been suggested to be accounted for peripherally by factors such as proprioceptive
feedback, the nature of the muscle contraction mechanism, information transmission ability
of the peripheral anatomical unit as well as centrally by noise in the neural control
mechanism (Newell et al., 1984). Given these factors, it is likely that the increased
complexity of the dynamic tasks results in a greater variability from repetition to
repetition, making these tasks more susceptible to manipulation. In addition, attempting
submaximal efforts leads to an even greater complexity of the task, with a resulting
increase in the likelihood of error and variability from trial to trial. This complexity is
likely to make dynamic tasks the better choice for detecting submaximal efforts.
The Choice of Isotonic Testing
Having now provided a rationale for a dynamic testing approach, the discussion
now turns to consideration of the use of an isotonic testing procedure. A significant
proportion of the research in the area concerning sincerity of effort has involved use of


8
Further evidence supporting the sincerity of effort hypothesis is provided by
other investigators who have also suggested that the variability of force production in
injured patients undergoing strength assessment may be an indicator for submaximal effort
(Gilbert & Knowlton, 1983; Chengalur, Smith, Nelson, & Sadoff, 1990; Niebuhr &
Marion, 1990). Gilbert and Knowlton (1983) proposed that individuals may use different
motor strategies in maximal and submaximal efforts. They propose that a maximal effort
is a lower order motor task in which feedback from the muscle contraction is not crucial.
On the other hand, a submaximal effort is a higher order motor task in which feedback is
crucial in limiting the force of the contraction. To limit force, a submaximal task would
require a greater degree of feedback from the muscles than would a maximal effort. An
implication is that differing motor strategies would make feigned submaximal efforts
detectable. While the sincerity of effort hypothesis shows promise, the validity of testing
methods requires more empirical examination.
Static vs. Dynamic Testing
While the assessment of sincerity of effort is now frequently used in clinical
practice, employing both static (isometric) and dynamic (isokinetic and isotonic) testing
approaches, the reliability and validity of current assessment methods have not been
adequately demonstrated (Mooney & Anderson, 1994). The following review will
initially consider the research using a static assessment approach. As the discussion
proceeds, the rationale for the move to a dynamic testing protocol will be provided as will
a review of the literature in this area.


6
relative cutoff that varies from test to test makes more sense. Bohannon (1987) mentions
the need for norms to be established for a specific test before a cutoff score can be
determined and judgments made regarding subjects effort. Research continues to address
the utility of using the CV in detection of submaximal efforts. Having provided a brief
explanation of both the testing and measurement approaches in the assessment of
musculoskeletal injury, the discussion now addresses how the research has responded to
the question of how to best assess effort level.
Statement of the Problem
Given the review of the costs of musculoskeletal injury, the need for reliable
assessment of sincerity of effort in back strength testing is emphasized. Further, a factor
contributing to lack of reliability in such testing is the potential for patients to be
motivated to appear more disabled than they are. Patients with musculoskeletal injury
may have incentive to produce less than maximal effort given concern for future injury,
exacerbation of current pain, apprehension in a novel test environment as well as issues
related to compensation (Mendelson, 1995). The degree to which "insincere" effort in
testing contributes to the reported costs is difficult to determine.
In the testing of injured individuals, evaluation of sincerity of effort is an important
factor in determination of functional capacity. The inability to accurately assess a patient's
effort level can lead to procedures that are both unnecessary and ineffective (Robinson,
OConnor, Riley, Kvaal, & Shirley, 1994). Further, the failure to accurately determine a
patients sincerity of effort has implications regarding the lack of apparent response to
treatment, as well as escalating health care and disability costs. In addition, failure to


40
Mooney, V.M., & Andersson, G.B.J. (1994). Controversies: Trunk strength
testing in patient evaluation and treatment. Spine. 19. 2483-2485.
Nachemson, A.L., & Lindh, M. (1969). Measurement of abdominal and back
muscle strength with and without low back pain. Scandinavian Journal of Rehabilitation
Medicine. 1. 60-65.
Newell, K.M., & Carlton, L.G. (1988). Force variability in isometric responses.
Journal of Experimental Psychology: Human Perception and Performance. 1. 37-44.
Newell, K.M., Carlton, L.G., & Hancock, P.A. (1984). Kinetic analysis of
response variability. Psychology Bulletin. 1. 133-151.
Niebuhr, B., & Marion, R. (1990). Voluntary control of submaximal grip
strength. American Journal of Physical Medicine & Rehabilitation. 69, 96-101.
Portney, L G., & Watkins, M.P (1993). Foundations of clinical research and
applications to practice. Norwalk, CT: Appleton & Lange.
Richards, J.G., Quigley, E.J., & Castaqno, P.W. (1996). Validity and reliability of
the BTE Dynatrac. Medical Science Sports Exercise. 28. 913-920.
Robinson, M E., Cassisi, J.E., OConnor, P.D., & MacMillan, M. (1992).
Lumbar iEMG during isotonic exercise: Chronic low back pain patients versus controls.
Journal of Spinal Disorders. 5. 8-15.
Robinson, ME., Geisser, M E., Hanson, C.S., & O'Connor, P. (1993). Detecting
submaximal efforts in grip strength testing with the coefficient of variation. Journal of
Occupational Rehabilitation. 3. 45-50.
Robinson, M.E., MacMillan, M., O'Connor, P, Fuller, A., & Cassisi, J. (1991).
Reproducibility of maximal versus submaximal efforts in an isometric lumbar extension
task. Journal of Spinal Disorders. 4. 444-448.
Robinson, M E., O'Connor, P., MacMillan, M., Fuller, A., & Cassisi, J. (1992).
Effect of instructions to simulate a back injury on torque reproducibility in an isometric
lumbar extension task. Journal of Occupational Rehabilitation. 2. 1-9.
Robinson, M.E., O'Connor, P, Riley, J.L., Kvaal, S., & Shirley, F.R. (1994).
Variability of isometric and isotonic leg exercise: Utility for detection of submaximal
efforts. Journal of Occupational Rehabilitation. 4, 163-169.
Robinson, M E., Sadler, I.J., OConnor, P., & Riley, J.L. (1997). Detection of
submaximal effort and assessment of stability of the coefficient of variation. Journal of
Occupational Rehabilitation. 7. 207-215.


ACKNOWLEDGMENTS
A number of individuals have made significant contributions to this dissertation in
particular and to my graduate training in general. Michael Robinson stands out as the
person who has contributed the most to my professional development and provided me
with a solid foundation in both clinical health psychology and, especially, experimental
methodology. This dissertation is a testimony to his commitment to the highest standards
of empiricism and scholarly work. I would also like to acknowledge the contributions of
my committee members, Duane Dede, Jon Kassel, Samuel Sears, and Mark Trimble.
These individuals have seen me through the process of qualifying examinations,
dissertation proposal, and dissertation defense. Their involvement and support has been
appreciated. Although not on my dissertation committee, Cynthia Belar, Jim Rodrigue,
and Michael Perri were also instrumental in my development as a clinician and as an
academic professional at the University of Florida. Collectively, these individuals have
provided me with a model of the scientist-practitioner in clinical psychology.
Several individuals have made contributions of time and energy to aspects of the
present study, including Patrick OConnor, Joseph Riley, and John Otis. Joseph Riley has
been a supportive friend and colleague from whom I have learned a great deal both
professionally and personally.
11


I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
lichael E. Robinson, Chair
Associate Professor of Clinical and Health
Psychology
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
£ < \
Duane E. Dede
Assistant Professor of Clinical and Health
Psychology
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Jo
Assistant Professor of Clinical and Health
Psychology
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Samuel F. Sears
Assistant Professor of Clinical and Health
Psychology
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.


33
Further, although this study which employed a dynamic assessment protocol did
not replicate the results found in the leg extension studies, the findings were consistent
with previous research employing isometric testing methodology in a lumbar extension
task (Robinson et al., 1991; 1992). In these isometric lumbar extension studies, results
indicated that participants were able to reliably reproduce a submaximal effort making it
difficult to discriminate between effort level. These earlier studies had set the precedent
for pursuing a dynamic testing approach in assessment of sincerity of effort of the lumbar
musculature.
In the present study there was no specific hypothesis attempting to determine
which of the three resistance levels would produce the best submaximal effort detection
rate. Results of the present study did not indicate any significant differences between these
resistance levels with respect to discriminability. In the pilot study, a 25% resistance level
had resulted in a moderate effect size in which there was an 8%-12% difference between
maximal and submaximal effort coefficient of variation respectively. We expected that we
might find similar results in the current study or, perhaps a somewhat higher or lower
resistance level might prove more efficacious in detection of submaximal efforts.
In addition, while the CV value based on 6 repetitions was higher than the CV
based on 3 repetitions for both maximal and submaximal effort levels, there was no
difference between the two measures with respect to discriminability of effort level.
Further, the higher CV calculated for 6 repetitions may simply represent a fatigue effect
because the latter repetitions (5, 6 and 7) were added in order to calculate the CV based
on 6 repetitions. In a previous study by Robinson et al. (1997) a trend suggesting
increased stability of the CV based on a greater number of repetitions was observed. In


10
lumbar spine disorders in workers and patients has become widespread. These tests are
used in the prediction of back injury (Battie et al., 1989) and assessing the functional
capacity of workers (Mayer et al., 1988). Although the methods of recording force
production may differ, with research at this point having focused primarily on isometric
and isokinetic testing, all methods have assumed that subjects giving less than the
instructed full, maximal effort would be identified by their inconsistent recordings.
However, despite the medicolegal implications of this assumption, there is insufficient data
to support the hypothesis that subjects cannot willfully reproduce a submaximal effort with
the same consistency as a maximal effort
Robinson, MacMillan, OConnor, Fuller and Cassisi (1991) conducted a study
investigating the differences in test-retest reliability between maximal and submaximal
efforts in an isometric lumbar extension task. They also tested the hypothesis that
submaximal efforts would be less consistent than maximal efforts. Twenty subjects were
asked to produce maximal voluntary contractions at seven different positions in a lumbar
extension machine. Each subject was tested twice in a maximal effort condition and twice
in a 50% effort condition. Results indicated high test-retest reliability at all angles in both
conditions. There were no differences in test-retest reliability between effort conditions.
Therapist ratings of consistency did not differ between conditions, and therapists could not
discriminate between conditions on the basis of effort consistency These authors
concluded that claims that isometric strength testing could be used to determine sincerity
of effort or malingering might be unfounded


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EBOUHTXYM_X9V9GF INGEST_TIME 2015-04-22T21:09:28Z PACKAGE AA00029774_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


39
Harber, P., & Soohoo, K. (1984). Static ergonomic strength testing in evaluating
occupational back pain. Journal of Occupational Medicine. 26. 887-884
Hazard R., Reid, S., Fenwick, J., & Reeves, V (1988). Isokinetic trunk and lifting
strength measurements: Variability as an indicator of effort. Spine. 13. 54-57.
Holmes, B Leggett, S., Mooney, V., Nichols, J., Negri, S., & Hoeyberghs, A.
(1996). Comparison of female geriatric lumbar extension strength: Asymptomatic versus
chronic low back pain patients and their response to active rehabilitation. Journal of
Spinal Disorders, 9, 17-22.
Kishino, N.D., Mayer, T.G., Gatchel, R.J., McCrate Parrish, M., Anderson, C ,
Gustin, L., & Mooney, V. (1985). Quantification of lumbar function. Part 4: Isometric
and isokinetic lifting simulation in normal subjects and low-back dysfunction patients.
Spine. 10. 921-927.
Kluck, D.J. (1967). A study of the strength ratio of the back extensors to the
trunk flexors. Masters Thesis. Department of Physical Therapy, University of Iowa, Iowa
City.
Lin, P C., Robinson, M E., Carlos, J., & O'Connor, P. (1996) Detection of
submaximal efforts in isometric and isokinetic knee extension tests. Journal of
Orthopaedic and Sports Physical Therapy. 24, 19-24.
Liu, A.C., & Byrne, E., (1995). Cost of care for ambulatory patients with low
back pain. The Journal of Family Practice. 40. 449-455.
Mayer, L. Greenberg, B.B. (1942). Measurements of the strengths of trunk
muscles. American Journal of Bone and Joint Surgery. 24. 842-856.
Mayer, T.G., Barnes, D., Kishino, N.D., Nichols, G., Gatchel, R.J., Mayer, H., &
Mooney, V. (1988). Progressive isoinertial evaluation in a standardized protocol and
normative data base. Spine. 13. 993-997.
Mayer, T.G., Smith, S.S., Keley, P.T., & Mooney, V. (1985). Quantification of
lumbar function: Part 2. Saggital plane trunk strength in chronic low back pain patients.
Spine. 10. 765-772.
McIntyre, D.R., Glover, L.H., Seeds, R.H., & Levene, J.A. (1990). The
characteristics of preferred low back motion. Journal of Spinal Disorders. 3. 145-155.
Mendelson, G. (1995). Compensation neurosis revisited: Outcome studies of the
effects of litigation. Journal of Psychosomatic Research. 39. 695-706.


UNIVERSITY lllllllllllUl!
l! 08554 7833


18
isometric and isokinetic methodology. The evaluation of sincerity of effort using isotonic
testing methods has not been addressed to any significant extent.
Even within the more extensively researched isometric and isokinetic
methodology, research has focused primarily on assessment of strength with less attention
to detection of sincerity of effort. However, it is important to acknowledge that that Dvir
and colleagues have employed isokinetic testing methodology and reported promising
results with respect to detection of sub-optimal performance (Dvir, 1997a, 1997b; David,
Dvir, Mackintosh, fe Brien, 1996; Dvir <& David, 1996).
In a review of the literature related to trunk muscle performance, Beimbom and
Morrisey (1988) discussed aspects such as differences in trunk extension force versus
trunk flexion force. They noted that there was great variation in methods, procedures, and
equipment, as well as type and speed of contraction used in these studies. In addition, the
subjects and their physical condition, weight, sex, age, and height differed in each study.
Due to the extreme variability in methodology and subject samples, it was very difficult to
assemble all of the data and have an accurate set of normative values. Despite these
difficulties, an attempt was made to present a compilation of all the work and to
categorize the many studies according to sample type, equipment used, motions tested,
and the mode of contraction evaluated.
Their review focused on studies investigating torque producing capacity in the
trunk musculature. The studies were classified into three categories: 1) isometric,
2) isotonic and 3) isokinetic analysis. In this review they reported twenty-five studies
employing isometric testing, and seventeen studies employing isokinetic testing methods.
Providing further evidence of the lack of research involving use of isotonic testing


CHAPTER 2
METHOD
Participants
Sixty-six asymptomatic normals were recruited from the psychology department
and physical therapy department and asked to participate in isotonic back extension
strength testing. These participants were divided into three groups of twenty-two subjects.
Results of a power analysis indicated that this would be the appropriate number of
subjects per group, given the pilot study results with an effect size of .8. An even number
of males and females were recruited. Exclusion criteria consisted of any history of
neuromuscular disease or any injury to the lower extremity that has required medical or
surgical intervention. Informed consent was obtained from the participants in accordance
with the Institutional Review Board.
Apparatus
Back extension assessment was measured by the Baltimore Therapeutic Equipment
Co. Dynatrac Back Tester (BTE Dynatrac, Hanover, MD). Research investigating the
reliability and validity of the BTE Dynatrac Back Tester has suggested that the
Dynatrac can be considered both reliable and objective (Richards, Bailey, & Castasqno,
1996). The Dynatrac computer provides test reports in which both concentric (as in
lifting) and eccentric (as in lowering) strength is exercised and analyzed. While in the back
tester, the patient remains in one stabilized position for flexion/extension and rotation.
Any size patient can be accommodated.
23


7
accurately detect patient effort can create an atmosphere where suspicion dominates,
which may then generalize to the genuinely disabled
Sincerity of Effort
One approach to the assessment of sincerity of effort in both research and
practice has been based on the hypothesis that sincerity of effort may be determined by
evaluating the consistency of individuals' responses to various strength testing tasks.
According to this hypothesis, the performance of an individual who is sincerely attempting
maximal exertion will be influenced only by physical limitations associated with his or her
injury or impairment and the influence of these limitations will remain relatively consistent
across trials. However, an individual whose performance is limited not only by physical
impairment but by attempts to give less than full effort will not be able to maintain as
consistent a level of performance, since other than physical factors are limiting his/her
performance.
Previous research has proposed that when a patient is deliberately attempting to
produce a submaximal effort, the obtained strength measurements will show increased
variability and reduced reliability (Kishino et al., 1985). The assumption is that subjects
cannot willfully reproduce a submaximal effort with the same consistency as a maximal
effort. Although Kishino et al. (1985) speculated about this hypothesis, they offered no
data to support it. Bohannon (1987) compared the variability of maximal and submaximal
tests of isometric elbow flexion by calculating the coefficient of variation (CV). He
reported that the CV of force measures was significantly greater during submaximal trials
than during maximal trials, which supported the Kishino et al. hypothesis.


I have made close and lasting relationships throughout the years of graduate
school. These friends have seen me through the good times and the difficult times. To
these individuals (Amy, Becca, Braden, Elena, Kathy, and Gregg), I say a heartfelt thank
you.
Finally, I will forever be indebted to my parents, Bridget and Brian Sadler, as well
as my sister, Sandy, for their innumerable sacrifices and unwavering love and support
throughout my years of graduate education. I dedicate this work to my family.
in


CHAPTER 4
DISCUSSION
As with the previous studies in the sincerity of effort program of research that have
been conducted in our lab, results of the present study indicated that there was a
significant difference between aggregate maximal and submaximal effort data. Participants
produced a submaximal effort at a significantly lower velocity than a maximal effort. On
average, the submaximal effort was approximately 57- 60% of the maximal effort for
calculations based on both 3 and 6 repetitions These findings are consistent with previous
research in our lab. (Robinson et al. 1994; Lin et al., 1996; Robinson et al., 1997)
Results did not support the hypothesis that the submaximal effort condition would
be more variable than the maximal effort condition or that a resistance level lower than
the 60% level previously addressed would result in increased discriminability between
effort level. Across the three resistance levels used in this study, we could not
discriminate between effort levels based on the proposed coefficient of variation measure
Previous research had suggested that a high CV indicated a lack of consistent effort
(Carlsoo, 1986; Robinson et al., 1994; Lin et al., 1996; Robinson et al., 1997). As a
result, we expected to observe a significantly higher CV in our submaximal effort
condition than in our maximal effort condition We did not find a significant difference
between effort levels. Further, our results indicated that the maximal effort condition
across resistance levels produced a slightly higher CV. This was in contrast to the results
found by Robinson et al. (1994), a study which prompted the use of isotonic strength
31


28
Table 3
Maximal Effort
Grp 1 (15%)
Grp 2 (25%)
Grp 3 (35%)
All Groups
Maximal Effort (3)
Maximal Effort (6)
7.7 4.4
10.7 3.8
9.7 5.5
10.5 4.3
11.1 5.1
11.6 3.8
9.5 5.1*
10.8 3 .9*
Submaximal Effort
Grp 1 (15%)
Grp 2 (25%)
Grp 3 (35%)
All Groups
Submaximal Effort (3)
Submaximal Effort (6)
8.7 4.5
10.2 4.8
8.43.9
10 4.4
9.2 6
10.4 3.9
8.8 5*
10.4 4.2*
Kp<005
the CVpeak (6 reps) directly comparing maximal and submaximal effort levels. Paired
t-tests indicated that there were no significant differences for either the CVpeak based on
6 or 3 repetitions with respect to the ability to discriminate effort levels within each
resistance level (see Table 4 for 6rep t-test results).
Table 4
CVpeak: Means and Standard Deviations across Groups
Grp 1(15%)* Grp 2(25%)2 Grp 3(3 5%)3 All Groups4
Maximal Effort 10.73.8 10.54.3 11.63.8 10.83.9
Submaximal Effort 10.24.8 104.4 10.43.9 10.44.2
11(21)= -.38, p<707; 2t(21)=.57, p< 574; 3t(21)=l.ll; p<279; 4t(21)=.66,p <513
A two-way mixed repeated measures analysis of variance was used to examine
main effects across resistance and effort level as well as an effort (maximal and
submaximal) by condition (15%, 25%, and 35% resistance levels) interaction for the CV
velocity measure (see Table 5). There were no significant effects found for within (effort
level) (F(2>63)= .43, p < .516; F(2>63) = .64, p < .531) and between (resistance level)
group(F(263) = .32, p < .725) factors. Sex differences in the CV were also examined in an


24
This apparatus allows for isometric assessment as well as isotonic testing at a pre
set resistance. The dynamometer can be locked for isometric testing or provide a specific
resistance to be maintained throughout movement in the case of isotonic strength testing.
With Dynatrac isotonic loading, the muscular force the patient must exert during a test is
prescribed and set by the experimenter. This is in contrast to isokinetic loading, where the
patient is told to exert a maximal force throughout his/her range of motion. The Dynatrac
isotonic load can be set by the experimenter to safely suit the patients capability, from
virtually zero to high performance torque levels. Isotonic strength assessment was
conducted in this study.
Procedure
Participants were asked to repeat a back extension exercise in the Dynatrac Back
Tester. Participants were divided into three groups of twenty-two subjects. Three
resistance levels were measured: 15% of body weight, 25% of body weight and 35% of
body weight. These values were chosen based on results from the pilot study which
showed a moderate effect size for weights in this range. Demographic information
including, height, weight and gender was collected.
A graduate student previously trained by a physical therapist and blind to order of
condition conducted the testing. On both trials, the graduate student conducted testing
consistent with maximal effort instructions (Ready! Go! As hard and as fast as you can!);
however, participants were told by the experimenter assistant to give a 100% effort in the
maximal effort condition and were told to give a 50% effort in the submaximal condition.
The assistant informed participants of order of condition.


12
In the Hazard et al. (1988) isokinetic TEF and LT study, previously discussed in
the context of the sincerity of effort hypothesis, the authors reported modest ability to
correctly classify degree of effort with simple visual inspection of the force curves.
However, using discriminant analysis techniques, the authors were able to improve the
accuracy of classification to 80-83% for the trunk extension task and 75-82% for the lift
task.
Although the results reported in their study were promising, Hazard et al. (1988)
pointed out that some subjects were able to produce consistent submaximal curves and in
contrast others produced inconsistent maximal effort curves, suggesting there is
considerable between subject variability. In addition, these authors noted that clinical
observation during the study was more accurate than the measures of curve variability.
Finally, the absence of uniform criteria for strength variability assessment may contribute
to the controversy in some of the studies mentioned above.
A recent study using the coefficient of variation measure compared the variability
of torque production and velocity in both isometric and isotonic leg extension tasks in
both maximal and submaximal effort conditions (Robinson et al., 1994). Fifteen
asymptomatic subjects participated in a within subject counterbalanced design in which
they were asked to perform maximally and submaximally in both isometric and isotonic leg
extension tasks. Results indicated that both isometric and isotonic tasks showed greater
variability (measured by coefficient of variation) in the submaximal effort condition. The
ability to detect submaximal efforts without misclassification of any maximal efforts was
much greater for the isotonic task (87% of efforts were correctly classified). While the
isometric task had unacceptably poor classification rates (60% of efforts were correctly