Purpose: Reduction in exercise capacity in patients with chronic heart failure (CHF) has been partially attributed to decreased muscle strength. Training studies reported important variations in strength increases during rehabilitation, ranging between 5% and 70% and depending on the measurement technique: isokinetic dynamometry or the one-repetition maximum (1RM) methods. Therefore, the question arises if both techniques assess the changes in muscle strength in a comparable way.
Methods: Thirty patients with CHF, New York Heart Association class II-III, with mean baseline V˙O2peak of 14.8 ± 3.0 mL·kg−1·min−1 and mean baseline left ventricular ejection fraction of 23.5% ± 5.5%, were assessed for knee extensor and knee flexor strength before and after 40 training sessions by isokinetic dynamometry and 1RM method. These two strength measurement techniques were compared using the Bland and Altman method for agreement.
Results: Knee extensor muscle strength increased by 7.4% and knee flexor strength increased by 18.7% if measured by isokinetic testing. With the 1RM method, knee extensor and flexor strength increased by 36% and 100%, respectively. Both techniques were not in agreement for muscle strength assessment; improvements were more important with the 1RM method. Furthermore, statistical analysis showed that the 1RM technique was overestimating strength increases in comparison with isokinetic evaluation, especially for higher strength levels.
Conclusions: In the follow-up of exercise training programs in patients with CHF, isokinetic measurements should be preferentially used to limit bias between measurements at different times.
1Cardiology Department, Centre Hospitalier de Luxembourg, LUXEMBOURG; and 2Clinical Epidemiology and Public Health Unit, Centre de Recherche Publique-Santé, LUXEMBOURG
Address for correspondence: Patrick Feiereisen, Cardiology Department, Centre Hospitalier de Luxembourg, 4, rue Barblé, L-1210, Luxembourg; E-mail: email@example.com.
Submitted for publication October 2009.
Accepted for publication April 2010.
Skeletal muscle atrophy, change in muscle fiber distribution with an increased percentage of type II fibers, and loss of muscle strength are common features of patients experiencing chronic heart failure (CHF). They are major components of exercise intolerance, frequently observed in these patients. The loss of skeletal muscle strength has been investigated by various methods, mostly isometric (3,7,16,18,20,24,25) or isokinetic dynamometry (19,24,25).
To follow up improvements in skeletal muscle strength after exercise training in patients with CHF, the one-repetition maximum (1RM) method is commonly applied in addition to isometric or isokinetic dynamometry. So far, there is no consensus concerning the most appropriate measurement method and protocol to use. This is in contrast with cardiopulmonary exercise testing, where standardized ramp protocols are used to determine V˙O2peak and adequate guidelines have been established (34).
In healthy subjects, Abernethy and Jürimäe (2) showed, among others, that improvements in strength are partially explained by neurological adaptations and that the results of dynamic training are dependent on the evaluation method that is used. They found that after 12 wk of training using free weights and weight machines, strength increased most significantly when the 1RM method was used for evaluation, followed by isokinetic dynamometry, and the least significant increase was measured by the isometric evaluation.
So far, the fact that strength increases might vary depending on the evaluation method has not been taken into consideration in patients with CHF. In a recently published study (13), where we compared the effects of strength to endurance and combined strength and endurance training, we measured the change in knee extensor and knee flexor muscle strength before and after training using an isokinetic dynamometer. In the strength training (ST) and combined training (CT) subgroups, 1RM was also determined (to calculate adequate training loads) at the beginning and at the end of the study for knee extension and knee flexion. In the previous study (13), we only used isokinetic dynamometry to analyze changes in muscle strength. The goals of the present study were to compare both strength measurement methods and to determine whether isokinetic dynamometry and 1RM measurement document the changes in muscle performance in a comparable way.
Thirty subjects, who had been previously included in the ST or the CT groups of a randomized controlled study about the effects of different training methods in CHF (13), were evaluated in this analysis. Inclusion criteria were New York Heart Association (NYHA) class II-III, left ventricular ejection fraction less than 35%, and/or V˙O2peak less than 20 mL·kg−1·min−1. Etiology of CHF was either CAD or idiopathic dilated cardiomyopathy. Exclusion criteria were NYHA class IV, malignant ventricular arrhythmias, renal dysfunction, orthopedic limitations, chronic obstructive pulmonary disease (COPD), and stroke. The study was approved by the local ethics committee, and written informed consent was obtained from all subjects.
Isokinetic strength testing for knee extensors and knee flexors was realized in an upright sitting position on a BIODEX Pro 3 dynamometer (Shirley, NY) after a 10-min warm-up at 30 W on a bicycle. Seat position was adjusted to guarantee that the knee axis was aligned to the axis of the dynamometer, using the lateral femoral epicondyle as an anatomical landmark. The resistance pad at the end of the lever arm was strapped around the distal part of the tibia. The distal border of the pad was placed 5 cm over the epicondyle of the lateral malleolus, which served as an anatomical landmark. Patients were stabilized in the chair by using straps around the thigh, the pelvis, and the chest. Identical testing position between baseline and posttraining tests was guaranteed by storing exact setup data. Range of motion was adjusted to exactly 90°, from knee extension to 90° of knee flexion. Gravity correction was obtained at 45° of knee flexion, and all measured variables were automatically adjusted for gravity. After these adjustments, patients were instructed about testing procedure. They first executed four to five submaximal repetitions of reciprocal knee extension and flexion, aiming to familiarize the patients with the isokinetic contraction mode. Afterward, four maximal repetitions at an angular velocity of 60°·s−1 were realized. Torque production was followed online at the computer screen, and the maximal repetition had to be reached within the first two repetitions and the variance between the four repetitions had to be less than 10% to guarantee maximal effort. The right leg was tested first, followed by the left leg. The dominant leg was used for data interpretation. Peak torque (PT) was defined as the highest value achieved during the four repetitions. All the tests were performed by the same technician. Verbal encouragements were provided during the tests. The tests were scheduled at the same time of the day (between 10:00 a.m. and noon), and they were performed within 2-5 d after the last training session.
The 1RM method can be defined as the maximum amount of weight lifted one time using proper form during a standard weightlifting exercise (23). For the execution of the 1RM evaluation, an initial weight, believed to be close to the persons' maximum lifting capacity, is chosen. Then, weight is progressively added (in steps of 2.5 kg in our case) until the person reaches their maximum lifting capacity, the 1RM. The rest interval during successive attempts was 1 min to provide enough recovery time before lifting the next heavier load (23). In this study, the 1RM for knee extension was measured in a sitting position. At the start, the knees were bent at 90° of flexion, and the subject was encouraged to move the legs until total extension. For knee flexion, the starting position consisted of knee extension, and the subjects were encouraged to bend their knee until 90° of flexion. The 1RM tests were performed using both legs simultaneously. For both measures, patients were positioned to guarantee that the knee axis and the rotation axis of the lever arm coincided. Anatomical landmarks used for pad placement during the 1RM evaluation were identical with those described in the previous paragraphs. In contrast to isokinetic testing, there were no straps available around the chest, thigh, or leg during the 1RM testing. As for isokinetic evaluation, the same technician performed all tests. The time of the day for testing was identical for the 1RM testing and the isokinetic testing.
1RM testing and isokinetic testing were separated by an interval of 2-3 d; isokinetic testing was always performed first.
Individual technique analyses of repeated measurements.
Each strength evaluation technique was independently analyzed for the effects of training group and time. The objective was to estimate the effect with each technique by using a random effect model accounting for the group and time effect (26). Least square means (±SEM) are presented to indicate the average difference between groups and times. Significant effects were declared at P < 0.05.
Strength evaluation technique bias.
To estimate the mean difference and the SD of differences about the mean, variance for repeated differences between the two methods on the same subject and that for the differences between the averages of the two strength evaluation techniques across subjects were estimated.
The within-subject bias, e.g., the mean difference that we called the strength evaluation technique bias, was estimated from the mean of the single differences between pairs of measurements (4-6).
The mean of these two techniques as well as the difference between the two techniques was calculated. The difference between the two techniques was regressed on the mean of the two techniques to examine the effect of the magnitude of the response. A Bland-Altman plot was done for each movement measurement at baseline and after 3 months of training. The model used for the analysis of the difference between the two analytical techniques allowed for group, time, group × time interaction, and other term effects. These terms were the linear regressions of the differences between the two techniques on the means of the two techniques and the interactions between group and time (26). Scatter plots of differences with regression lines were used to illustrate the agreement of analytical techniques over the range of values observed for the sample results. Residual variances were reported to assess the precision of each analytical technique for the combinations of group and time.
The variability associated with the analytical techniques, group, and time were evaluated and compared using the Akaike Information Corrected Criterion as a "goodness-of-fit" indicator. Estimates of the residual variances for each combination of group and time and for each analytical technique are reported.
The baseline characteristics of the study population are presented in Table 1. The data of baseline and posttraining values for PT and 1RM for knee extensors and knee flexors are presented in Table 2.
Individual technique analyses of repeated measurements.
In the analysis of the PT in extension, both the ST and the CT groups did not differ from each other within each strength evaluation technique. There was, however, a difference in time because the 3 months of evaluation had significantly higher values than that of the baseline evaluation within the 1RM technique only. This was not the case within the group × time interaction (Table 3).
In the analysis of the PT in flexion, both the ST and the CT groups did not differ from each other within each technique. There was, however, a difference in time because the 3 months of evaluation had significantly higher values than that of the baseline evaluation within both techniques. This was not true within the group × time interaction, but a borderline significant P value could be observed in the 1RM flexion movement, where the 3 months of evaluation from both ST and CT groups had higher values than the baseline evaluation from both groups (Table 3).
Strength evaluation technique bias.
The estimated bias, the mean difference, was estimated from the mean of the individual differences (6). The averages were 142.84 for movement in extension and 76.37 for movement in flexion. Using the square root of total variance, the 95% limits of agreement were estimated: the PT in knee extension as measured by the 1RM in patients with CHF would be between 64.96 and 220.72 U of the isokinetic technique. For knee flexion, the 1RM results would be between 29.48 and 123.25 U of the isokinetic technique. However, we had to assume that the repeated differences for a single subject were independent in these analyses. This might be a rather strong assumption. There was an expected change between baseline and 3 months of evaluation because the subjects were undergoing an exercise therapy. Therefore, the measures at baseline and after 3 months in a single subject were not independent because the same parameters were recorded using the two different techniques. Furthermore, the previous analyses did not account for the group therapy, which might influence the results of the measures.
Whatever the movement and the time of measurement, when the mean of the two techniques increases, the difference between techniques increases as well. This demonstrates that, with the increasing mean, the agreement between techniques decreases. At baseline, most of the sample means were between 40 and 150 and most of the differences were in the maximum acceptable difference (MAD) range (mean ± 2 SD) (Figs. 1 and 2). Therefore, the 1RM technique showed the highest differences at higher strength levels. After 3 months of training, whereas sample means varied between 25 and 85, a similar result was observed (Figs. 3 and 4). Furthermore, two observations are outside the MAD range. The 1RM technique obviously does not agree with either isokinetic extension measures or flexion measures at baseline and at 3 months of evaluation: a significant slope bias (regression P < 0.0001, analysis not shown) is observed in all plots.
The last three lines of Table 4 relate to the differences between slopes, whereas the first three lines show the mean differences of the average of techniques between groups and time. There was no evidence of differences in the means between group and time or between knee extension and flexion. There were no significant slopes except the already observed significant slope bias (individual regression models by group and time). Likewise, there is no evidence that the mean differences are a function of group and time.
The beneficial effects of exercise training in CHF are largely due to the improvement of endothelial (14) and peripheral muscle functions (13). Various techniques, including imaging techniques such as computed tomographic scan (3,16,31), magnetic resonance imaging (22,25), or dual-energy x-ray absorptiometry (8), EMG (31), and muscle strength assessment (3,7,16,19,20,24,25) have been used to document peripheral muscle morphology and function. Unlike cardiopulmonary exercise testing, where guidelines recommend a ramp protocol for exercise testing that is unanimously used (34), different methods have been used for strength evaluation: isokinetic, isometric, or 1RM measurement. So far, there are no recommendations for peripheral muscle strength assessment in patients with CHF in daily clinical practice, and the different methods of muscle assessment have not been thoroughly investigated in patients with CHF. In the only study where the reliability of strength measurements was analyzed (29), isokinetic and isometric strength measurements were used. These authors found intraclass correlation coefficients of 0.96-0.99 for knee extensors and 0.94-0.96 for knee flexors and concluded that isokinetic testing was a reliable method for assessing thigh muscle strength in patients with CHF. Furthermore, these authors observed cardiovascular response and recorded the degree of dyspnea during testing. They concluded that this form of testing could be considered as safe in the absence of adverse events or inadequate hemodynamic responses. The intraclass correlation coefficient in patients with CHF is comparable to that of the athletic population (r < 0.9) (1). It is to be noted that the reliability of the 1RM method is also high in the experienced athletic population (r = 0.92-0.98) (1).
Most training studies in patients with CHF report gains in V˙O2peak consistently ranging between 12% and 31% (27), which is similar to those observed in our patients (CT group = +14% and ST group = +17% ). In contrast, the range of muscle strength improvements with training is much larger, approximately from 5% to 70% (9-11,15,17,20,21,28,32,33). This is because different evaluation methods have been used. The results in the literature are in accordance with our results (Table 2), showing more important strength gains when assessed by the 1RM technique (17,20,21,28) in comparison to the isokinetic technique (9-11,15,32). Some studies, using the 1RM method, even combine different muscle groups to, finally, find an overall improvement of muscle strength (21). Therefore, conclusions concerning the evolution of muscle strength have to be related to the methods used for strength assessment.
In our study, the statistical analysis demonstrated that the isokinetic and the 1RM techniques were not in agreement concerning the muscle strength assessment. Patients underwent resistive ST or combined endurance-resistive ST (13). Both training modalities induced an improvement of muscular strength. However, strength improvement was greater when assessed with the 1RM technique.
We found that, although most of the values were between the lower and upper limits of MAD, the difference between the two techniques increased when the mean of the two techniques increased. The 1RM technique was biased in comparison with the isokinetic evaluation, overestimating strength gains with the increasing strength levels. This bias was explained neither by the training method nor by the time of measurement. Therefore, the isokinetic technique seems a more accurate method of measurement than the 1RM technique. The use of the isokinetic method, especially in longitudinal measurements, is strongly advised to limit bias between measurements at a different time, particularly when performing trials of exercise therapy, which may improve muscle performance and, therefore, values as measured by the technique.
Increases in muscle mass reported after exercise training are small compared with the increases in muscle strength. In a study of Senden et al. (33), muscle mass did not increase with exercise training; however, in the control group, muscle mass decreased. Other studies observed increases between 3% and 15% in muscle mass (13,20,28,30). Although none of these studies could demonstrate a positive correlation between the change of muscle strength and the change of muscle volume, it seems that the range of changes in muscle strength measured by isokinetic dynamometry, compared with 1RM, is much closer to the range of changes in muscle volume. Isokinetic assessment seems to be less influenced by neuromuscular adaptations to strength exercise than 1RM measurements. This is easily understandable because the 1RM testing and the ST are executed on the same device, favoring a specificity of adaptations not transferable to isokinetic testing (2). Therefore, isokinetic dynamometry may be documenting muscle mass changes more effectively than the 1RM. Together with our statistical findings, this clinical consideration gives further support to the use of isokinetic dynamometry over the 1RM method.
In patients with CHF, there is not only a decrease in muscle mass but also a change in muscle fiber distribution with a higher amount of type II muscle fibers and a decrease in type I fibers (12). There may be a concern that this shift in fiber-type distribution might interfere with strength testing results. Indeed, it has been shown that isokinetic testing at high angular velocities could be associated with muscle fiber composition (35). This, however, is not the case for isokinetic testing at lower angular velocities like those that we used for our patients (35) or with 1RM testing. Therefore, we do not believe that an altered muscle fiber distribution in patients with CHF can explain the higher increase in muscle strength as assessed by the 1RM technique compared with the isokinetic technique.
To assess changes in muscle strength with exercise training, isokinetic strength testing seems to be less biased than the 1RM method and should therefore be recommended. 1RM, however, can be used to calculate the proper training loads for ST in patients with CHF.
The authors declare that they have no relationship to disclose regarding the present study.
The results of the present study do not constitute endorsement by the American College of Sports Medicine.
1. Abernethy P, Wilson G, Logan P. Strength and power assessment: issues, controversies and challenges. Sports Med
2. Abernethy PJ, Jürimäe J. Cross-sectional and longitudinal uses of isoinertial, isometric, and isokinetic dynamometry. Med Sci Sports Exerc
3. Anker SD, Swan JW, Volterrani M, et al. The influence of muscle mass, strength and fatigability and blood flow on exercise capacity in cachectic and non-cachectic patients with chronic heart failure. Eur Heart J
4. Altman DG, Bland JM. Measurement in medicine: the analysis of method comparison studies. Statistician
5. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet
6. Bland JM, Altman DG. Agreement between methods of measurement with multiple observations per individual. J Biopharm Stat
7. Buller NP, Jones D, Poole-Wilson PA. Direct measurements of skeletal muscle fatigue in patients with chronic heart failure. Br Heart J
8. Cicoira M, Zanolla L, Franceschini L, et al. Skeletal muscle mass independently predicts peak oxygen consumption and ventilatory response during exercise in non cachectic patients with chronic heart failure. J Am Coll Cardiol
9. Cider A, Tygesson H, Hedberg M, Seligman L, Wennerblom B, Sunnerhagen KS. Peripheral muscle training with clinical signs of heart failure. Scand J Rehab Med
10. Delagardelle C, Feiereisen P, Krecké R, Essamri B, Beissel J. Objective effects of a 6 months' endurance and strength training program in outpatients with congestive heart failure. Med Sci Sports Exerc
11. Delagardelle C, Feiereisen P, Autier P, Shita R, Krecké R, Beissel J. Strength/endurance training versus endurance training in congestive heart failure. Med Sci Sports Exerc
12. Drexler H, Riede U, Munzel T, Konig H, Funke E, Just H. Alterations of skeletal muscle in chronic heart failure. Circulation
13. Feiereisen P, Delagardelle C, Vaillant M, Lasar Y, Beissel J. Is strength training the more efficient training modality in chronic heart failure? Med Sci Sports Exerc
14. Hambrecht R, Fiehn E, Weigl C, et al. Regular physical exercise corrects endothelial function and improves exercise capacity in patients with chronic heart failure. Circulation
15. Hare DL, Ryan TM, Selig SE, Pellizer AM, Wrigley TV, Krum H. Resistance exercise training increases muscle strength, endurance and blood flow in patients with chronic heart failure. Am J Cardiol
16. Harrington D, Anker SD, Chua TP, et al. Skeletal muscle function and its relationship to exercise intolerance in chronic heart failure. J Am Coll Cardiol
17. Levinger I, Bronks R, Cody DV, Linton I, Davie A. Resistance training for chronic heart failure patients on beta blocker medication. Int J Cardiol
18. Lipkin D, Jones D, Round J, Poole-Wilson P. Abnormalities of skeletal muscle in patients with chronic heart failure. Int J Cardiol
19. Magnusson G, Isberg B, Karlberg G, Sylven C. Skeletal muscle strength and endurance in chronic congestive heart failure secondary to idiopathic dilated cardiomyopathy. Am J Cardiol
20. Magnusson G, Gordon A, Kaijser L, et al. High intensity knee extensor training, in patients with chronic heart failure. Major skeletal muscle improvement. Eur Heart J
21. Maiorana A, O'Driscoll G, Cheetham C, et al. Combined aerobic and resistance training improves functional capacity and strength in CHF. J Appl Physiol
22. Mancini DM, Walter G, Reichek N, et al. Contribution of skeletal muscle atrophy to exercise intolerance and altered muscle metabolism in heart failure. Circulation
23. McArdle WD, Katch FI, Katch VL. Exercise Physiology. Energy, Nutrition and Human Performance
. 5th ed. Philadelphia (PA): Lippincott, Williams & Wilkins; 2001. p. 502.
24. Minotti JR, Christoph I, Oka R, Weiner MW, Wells L, Massie BM. Impaired skeletal muscle function in patients with congestive heart failure. J Clin Invest
25. Minotti JR, Pillay P, Oka R, Wells L, Christoph I, Massie BM. Skeletal muscle size: relationship to muscle function in heart failure. J Appl Physiol
26. Peterson AB, Douglass LW. Adopting new techniques: statistical procedures to determine the agreement between two laboratory methods. In: Proceedings of 18th Annual Conference Proceedings of the Northeast SAS Users Group Inc; Sept 11-14, 2005
: Portland (ME); 2005. p. AN11.
27. Pina Il, Apstein CS, Balady GJ, et al. Exercise and heart failure: a statement from the American Heart Association Committee on Exercise, Rehabilitation and Prevention. Circulation
28. Pu CT, Johnson MT, Forman DE, et al. Randomized trial of progressive resistance training to counteract the myopathy of chronic heart failure. J Appl Physiol
29. Quittan M, Wiesinger GF, Crevenna R, et al. Isokinetic strength testing in patients with chronic heart failure-a reliability study. Int J Sports Med
30. Quittan M, Wiesinger GF, Sturm B, et al. Improvement of thigh muscles by neuromuscular electrical stimulation in patients with refractory heart failure. Am J Phys Med Rehabil
31. Schulze PC, Linke A, Schoene N, et al. Functional and morphological skeletal muscle abnormalities correlate with reduced electromyographic activity in chronic heart failure. Eur J Cardiovasc Prev Rehabil
32. Selig SE, Carey MF, Menzies DG, et al. Moderate-intensity resistance training in patients with chronic heart failure improves strength, endurance, heart rate variability and forearm blood flow. J Card Fail
33. Senden PJ, Sabelis LW, Zonderland ML, Hulzebos EH, Bol E, Mosterd WL. The effect of physical training on workload, upper leg muscle function and muscle areas in patients with chronic heart failure. Int J Cardiol
34. Task Force of the Italian Working Group on Cardiac Rehabilitation Prevention; Working Group on Cardiac Rehabilitation and Exercise Physiology of the European Society of Cardiology, Piepoli MF, et al. Statement on cardiopulmonary exercise testing in chronic heart failure due to left ventricular dysfunction: recommendations for performance and interpretation. Part I: definition of cardiopulmonary exercise testing parameters for appropriate use in chronic heart failure. Eur J Cardiovasc Prev Rehabil
35. Wrigley T. Correlation between isokinetic dynamometry and muscle fiber type. In: Brown LE, editor. Isokinetics in Human Performance
. Champaign (IL): Human Kinetics; 2000. p. 50.
Keywords:©2010The American College of Sports Medicine
EXERCISE TRAINING; PERIPHERAL MUSCLE STRENGTH ASSESSMENT; EVALUATION TECHNIQUES; BLAND AND ALTMAN METHOD FOR AGREEMENT