Variability of isometric and isotonic leg exercise: Utility for detection of submaximal effort

Discriminating between maximal and feigned isokinetic knee musculature performance using waveform similarity measures

BACKGROUND

Muscle strength test outcomes may aid in determination of impairment or disability rating following injury. In such settings, verification of participant effort during testing is imperative. This investigation explored the utilization of within-set moment waveform similarity measures, namely cross correlation and percent root mean square difference scores, to develop decision rules for discriminating between maximal and feigned efforts during isokinetic testing of the knee joint musculature.

METHODS

A mixed-gender sample of 46 participants performed non-reciprocal sets of maximal or feigned knee extension and flexion concentric and eccentric efforts at testing velocities of 30°s(-1) and 120°s(-1). Logistic regression and Monte Carlo simulations were used to derive decision rules for differentiating between the two effort types.

FINDINGS

Employing cutoff scores corresponding to 100% specificity; sensitivities of the knee extensor's velocity-specific decision rules were 92.4% and 84.8%, respectively. The velocity-specific knee flexor's test sensitivities were 56.5% and 46.7%.

INTERPRETATION

Utilizing the proposed decision rules, substantiating maximal effort performance of the knee extensors may be possible using this specific testing protocol. However, the proposed methods are limited in their ability to verify performance of maximal knee flexor efforts.

Differentiating between types and levels of isokinetic knee musculature efforts

This investigation assessed whether a measure of moment curve shape similarity, and a measure quantifying curve magnitude differences, enables differentiation between types (sincere vs. feigned) and levels (maximal vs. submaximal) of effort exerted during isokinetic testing of the knee. Healthy participants (n=37) performed four sets of six concentric knee extension-flexion repetitions on two occasions. The sets consisted of: (1) maximal effort; (2) self-perceived 75% of maximal effort; (3) self-perceived 50% of maximal effort; and (4) a set attempting to feign injury. Average cross-correlation and percent root mean square difference values were computed between moment curves in each direction. Logistic regression was used to derive decision rules for differentiating between maximal and submaximal effort levels; and between sincere and feigned effort types. Using a cutoff criteria corresponding to 100% specificity, maximal effort production could be ascertained with 96% sensitivity within the sample. Feigned efforts, however, could be ascertained with only 31% sensitivity due to overlap with sincere submaximal effort. Using the proposed models, clinicians may be able to ascertain whether maximal efforts were produced during isokinetic knee musculature testing. Additionally, evidence regarding participant's intentions with regard to influencing test results may be gauged, although to a lesser extent.

Critical issues in the use of muscle testing for the determination of sincerity of effort

Over the past 20 years, there have been numerous attempts to identify methodologies that are capable of the determination of sincerity of effort during muscle testing. The ensuing paper reviewed this literature and drew several conclusions. Injured patients and healthy volunteers do produce less force and more variable force while performing submaximal contractions than maximal contractions. However, submaximal efforts during strength testing can be reproduced and the use of force variability is not adequate to distinguish sincerity of effort. Visual examination of the shape of force output curves is also not adequate for distinguishing sincerity of effort. Furthermore, much of the research using strength ratios, difference scores, and an assortment of different parameters derived during strength testing has not established reliable and clinically useful methods of differentiating effort levels. Methods examining motion variability, radial/ulnar force output ratios, difference scores of eccentric-concentric ratios, and electromyography offer some promise, but numerous critical issues need to addressed. The use of the coefficient of variation, for example, is statistically untenable given the number of trials appropriate for clinical samples. Several studies have inadequate sample size to number of variable ratios. Many studies have questionable or at least unknown generalizability to patient samples and actual functional capacity. It is critical that other explanatory variables such as fear of injury, pain, medications, work satisfaction, and other motivational factors be considered. It is our opinion that there is not sufficient empirical evidence to support the clinical application of muscle testing to determine sincerity of effort.

Chronic pain disability exaggeration/malingering and submaximal effort research

OBJECTIVE

This is the first review of chronic pain (CP) malingering/disease simulation research. The purpose of this review was to determine the prevalence of malingering within CP patients (CPPs), whether evidence exists that malingering can be detected within CPPs, and to suggest some avenues of research for this topic.

DESIGN

A computer and manual literature search produced 328 references related to malingering, disease simulation, dissimulation, symptom magnification syndrome, and submaximal effort. Of these, 68 related to one of these topics and to pain. The references were reviewed in detail, sorted into 12 topic areas, and placed into tabular form. These 12 topic areas addressed the following: existence of malingering within the CP setting; dissimulation, identification simulated (faked) facial expressions of pain; identification of malingering by questionnaire; identification of malingered sensory impairment; identification of malingered loss of hand grip strength; identification of submaximal effort by isometric strength testing; identification of submaximal or malingered effort by isokinetic strength testing; identification of submaximal or malingered effort by the method of coefficient of variation; self-deception; symptom magnification syndrome; and miscellaneous malingering identification studies. Each report, in each topic area, was rated for scientific quality according to guidelines developed by the Agency for Health Care, Policy and Research (AHCPR) for rating the level of evidence presented in the reviewed study. The AHCPR guidelines were then used to rate the strength and consistency of the research evidence in each topic area based on the type of evidence the reports represented. All review conclusions were based on the results of these ratings.

SETTING

Any medical setting reporting on either malingering or disease simulation, or dissimulation, or submaximal effort and pain.

PATIENTS

Normal volunteers, CPPs, or any group asked to produce a submaximal or malingered effort or a malingered test profile.

RESULTS

The reviewed studies indicated that malingering and dissimulation do occur within the CP setting. Malingering may be present in 1.25-10.4% of CPPs. However, because of poor study quality, these prevalence percentages are not reliable. The study evidence also indicated that malingering cannot be reliably identified by facial expression testing, questionnaire, sensory testing, or clinical examination. There was no acceptable scientific information on symptom magnification syndrome. Hand grip testing using the Jamar dynamometer and other types of isometric strength testing did not reliably discriminate between a submaximal/malingering effort and a maximal/best effort. However, isokinetic strength testing appeared to have potential for discriminating between maximal and submaximal effort and between best and malingered efforts. Repetitive testing with the coefficient of variation was not a reliable method for discriminating a real/best effort from a malingered effort.

CONCLUSIONS

Current data on the prevalence of malingering within CPPs is not consistent, and no conclusions can be drawn from these data. As yet, there is no reliable method for detecting malingering within CPPs, although isokinetic testing shows promise. Claims by professionals that such a determination can be made should be viewed with caution.

Detection of a "faked" strength task effort in volunteers using a computerized exercise testing system

The objective of this study was to develop an experimental method to separate a "faked" strength effort from a "best" effort in volunteers. Thirty-four pain-free volunteers (18 males, 16 females) performed a shoulder press and pull-down on an isokinetic computerized exercise testing system (CETS), giving a best effort followed by a faked effort. Two months later, a randomly selected subgroup (6 males) repeated the experiment to test the predictive validity of the derived variables. In the statistical analysis, best efforts were first compared with fake efforts by paired ttest for 80 CETS variables for males and females separately. Variables showing a strong difference between the best and faked effort were then selected for further analysis. In the second step of the analysis, the method of multiple correlations (r2 method) was used to reduce the number of redundant CETS variables to five in both the male and female groups. In the third step, a stepwise discriminant analysis was used to select predictor variables for the male and female groups. For the variables selected by the discriminant analysis for both males and females, sensitivities and specificities were calculated. Finally, the developed discriminant formula was used in the predictive validity part of the study to determine the sensitivities and specificities of the developed method. The discriminant analysis selected the following CETS variables for male and female groups, respectively: duty cycle down, work weight/down, peak value up (males); and average power up, 40% repetition down, duty cycle up (females). For males, using their three variables, the discriminant function classified 77.14% of the efforts correctly with 88.9% sensitivity and 64.7% specificity. For females, using their three variables, the discriminant function classified 90.63% of the efforts correctly with 100% sensitivity and 81.3% specificity. In the predictive validity group, the discriminant function classified 75% of the efforts correctly with 83.3% sensitivity and 66.7% specificity. This pilot study indicates that the method developed here may be useful in the experimental study for the discrimination between faked and best efforts on this isokinetic CETS machine. Future studies using this method will need to involve a larger number of volunteers.

Coefficient of variation in maximal and feigned static and dynamic grip efforts

The purpose of this study was to investigate the validity of the coefficient of variation as an identifier of feigned grip effort. Seventeen healthy female aged 20 to 25 yr participated in the study. Maximal and feigned efforts were measured isometrically and isokinetically (concentric and eccentric) using the Jamar and KinCom dynamometers, respectively. Findings indicated that, in all situations, the coefficient of variation derived from the maximal effort was significantly (P < 0.0001) lower than that derived from the feigned effort. However, the extent of overlapping between the two was sufficiently large to render the test sensitivities very low. Consequently, regardless of the measurement method, the coefficient of variation is not a valid tool for identifying feigned grip effort in healthy subjects.

Measurement variability and sincerity of effort: clinical utility of isokinetic strength coefficient of variation scores

Although the use of measures of strength variability as a means of judging sincerity of effort is becoming common practice, the accuracy of doing so has been questioned. Coefficient of variation (CV) cut-off points, indicating the upper limit of variability for repeated maximal efforts, are routinely used to identify workers providing submaximal efforts during various strength tests. However, the stability of the CV itself has not been considered when comparing an individual's observed CV score to these cut-off points. The purpose of the present study was to examine the day-to-day variability of the CV calculated from maximal isokinetic knee extension efforts, and to describe how this measurement error affects the accuracy of the CV as a distinguishing criterion between maximal and submaximal efforts. Thirty-one healthy males (mean age 25 +/- 4.5 years) completed three maximal and three submaximal isokinetic knee extension efforts on two separate occasions. Although submaximal CVs were significantly greater than maximal CVs (15.6 versus 3.7%; p < 0.01), there was considerable overlap between submaximal and maximal CV frequency distributions. More importantly, an individual observed CV could vary +/- 3.1% as a result of day-to-day variation or measurement error. This range in possible CV scores should be considered when comparing an individual's score to proposed cut-off points. Since individual CVs vary considerably from day-to-day, and since precise cut-off values distinguishing between maximal and submaximal conditions cannot be identified, CV scores must be interpreted cautiously, and the potential errors in relying extensively on this approach to identifying insincere efforts should be recognised.

Bias effects in three common self-report pain assessment measures

OBJECTIVE

Past research has shown response biases to influence the accuracy of results from self-report measures. In pain assessment, where a percentage of patients have financial and other reasons to minimize or exaggerate psychological disturbance, it becomes especially important to identify the influence of response bias in self-report of adjustment. This study investigated the susceptibility of three commonly used self-report pain assessment measures to response bias.

DESIGN

This study used a within-subjects (asymptomatic subjects) design with two experimental conditions and nonequivalent control group (chronic pain patients).

SUBJECTS

Experimental group: 40 students enrolled in an occupational therapy program at a major southeastern United States university.

CONTROL GROUP

200 subjects referred to a multidisciplinary pain clinic at a major teaching hospital.

MEASURES

Coping Strategies Questionnaire, Multidimensional Pain Inventory, and Pain Beliefs and Perceptions Inventory.

RESULTS

With few exceptions, asymptomatic subjects scored significantly differently on these measures while portraying themselves as either coping well or coping poorly. In addition, when using the "coping poorly" response set, asymptomatic subjects reproduced scores similar to those of symptomatic chronic pain patients.

CONCLUSION

The susceptibility to manipulation appeared constant across the three measures, a finding that highlighted the difficulties clinicians and researchers encounter in accurate interpretation of results from these measures in the absence of validity indicators. This study also emphasizes the ease with which subjects with sufficient motivation can present themselves in an untruthful and manipulative manner and can generate scores that are, on their own, difficult to distinguish from those of a group of typical chronic pain patients.

Detection of submaximal effort in isometric and isokinetic knee extension tests

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. The purpose of this study was to investigate differences between maximal and submaximal efforts in isometric and isokinetic knee extension using torque variability and slope measures obtained from 32 volunteers (age = 25.2 +/- 4.7 years). The coefficient of variation of average torque, coefficient of variation of peak torque, and slope to peak torque were obtained from maximal and submaximal torque curves during isometric and two isokinetic tests (60 degrees/sec and 180 degrees/sec). Significant differences between effort levels (maximal and submaximal) were shown for all variables in isometric and isokinetic tests. An optimized cut-off value to determine submaximal efforts without false positives was determined for each variable. The coefficient of variation of average torque detected 75% of submaximal efforts at 180 degrees/sec and slope to peak torque detected 63% of submaximal efforts at 60 degrees/sec. For both speeds, combining the coefficient of variation of average torque with slope improved the submaximal detection rate to 84%. No variable provided a satisfactory detection rate for the isometric test. Therefore, submaximal isokinetic knee extensions are detectable with torque variability and slope measures, but submaximal isometric exercise is not detectable.