Journal Logo

CLINCIAL SCIENCES: Clinical Investigations

Assessing the efficacy of exercise training in patients with chronic disease


Author Information
Medicine & Science in Sports & Exercise: August 2002 - Volume 34 - Issue 8 - p 1234-1241
  • Free


The majority of previously reported exercise training studies in patients with end stage renal disease (ESRD) have established therapeutic efficacy of exercise training by focusing on indices of peak exercise tolerance (V̇O2peak and peak power output) (5,10,21,22,23). It is somewhat surprising that no studies have reported the effects of exercise training on submaximal constant load exercise performance (V̇O2 kinetics) and relatively few (23,28) have reported the effects of exercise training on exercise capacity at ventilatory threshold (VT). Evaluation of the adaptive response of submaximal indices of exercise tolerance in addition to peak responses may be especially useful in severely deconditioned individuals.

Evaluation of the therapeutic efficacy of exercise training interventions in clinical populations has generally been based on a traditional hypothesis testing approach. However, several investigators (1,3,16) have recently questioned the use of statistical significance tests as the sole criterion by which to assess the value of research findings. It has been proposed that a more meaningful approach to assess the impact of exercise training interventions should incorporate information about changes observed in relation to the coefficient of variation (CV%) and standard error of measurement (SEM) for the parameters of interest (16). The value of this additional or alternative information is evidenced by the fact that it is clearly possible to achieve a statistically significant change in exercise tolerance for a group of patients that would not necessarily be reflected in favorable adaptations for all patients. Exploration of the relationship between the statistical significance of an intervention and the meaningfulness of the findings is a particularly relevant issue for clinical exercise rehabilitation, because individual patient assessment is of paramount importance in clinical settings. This is a particularly relevant issue for exercise training interventions involving patients with ESRD, as this patient population is particularly heterogeneous in terms of exercise tolerance, age, nutritional state, and coexisting disease, all factors that may influence the adaptive response to exercise training.

Therefore, the aim of this study was to evaluate the effects of exercise training on the nature and time course of adaptation of peak and submaximal exercise capacity of patients with ESRD, using a variety of analytical approaches ranging from conventional hypothesis testing through assessment of group mean changes to assessment of individual subject response characteristics. A secondary aim was to compare the group of patients with a group of age, gender, and physical activity-matched healthy individuals in order to determine the degree of “normalization” of ESRD patients after the end of the intervention period.


Thirty-four patients with ESRD initially volunteered to participate in the exercise program. After completion of all assessments at baseline, patients entered the exercise rehabilitation program.

Exclusion criteria to participate in the study were evidence of recent myocardial infarction (within 6 wk), uncontrolled arrhythmias, uncontrolled hypertension, unstable angina, severe uncontrolled diabetes, symptomatic left ventricular dysfunction, or neurological disorder with functional deficit, demonstrating an interdialytic weight ≥ 2.5 kg, pre dialysis potassium ≥ 5.5 mmol·L−1, and urea clearance kt/v ≤ 1 mL·min−1·L−1. Medication and dialysis prescription were kept optimal throughout the study period. All the patients, except one, was receiving EPO to maintain hemoglobin (Hb) constant above 10 g·dL−1.

The study was approved by the Local Scientific Merit and Ethical Committees. All the subjects were informed about the procedures of testing and training, and any risks associated with it, and provided a written informed consent to participate.

The healthy control group (HC) consisted of 14 men and 4 women, age, body mass, and physical activity-matched with the patient group. None of the healthy control subjects was smoking at the time of the study or was aware of any medical conditions that could influence exercise testing results. None of the subjects participated in any regular exercise activities other than physical activities of daily living.

Nutritional Assessment

Nutritional status of the patients was assessed using the seven-point subjective global assessment scale (SGA). This type of nutritional assessment is based on evaluation of i) weight change, ii) dietary intake, iii) gastrointestinal symptoms, iv) nutritionally related functional impairment, and v) physical examination for evidence of loss of subcutaneous fat, muscle wasting, and edema. This assessment classifies subjects by assigning an alphabetic category ranking of either A, A−, B+, B, B−, C, or C− according to the SGA evaluation. These categories translate into numerical categories ranging from the highest number of 7 corresponding to letter A, to the lowest of 1 corresponding to the letter C−, for statistical purposes. The validity and reliability of this method of nutritional assessment has previously been reported (11,28). Patients were categorized as adequately nourished (SGA > B+) or as malnourished (SGA ≤ B+).

Self-Reported Physical Activity Assessment

Levels of physical activity were assessed using the 7-Day Physical Activity Recall Interview Questionnaire (7-d RQ) (4). This method of assessment was employed to match subjects by activity levels. The 7-d RQ allows the calculation of the total number of calories per body mass expended in 1 wk based on questions regarding the hours spent in 1 wk, for sleep, moderate, hard, and very hard activities. In this study, we report the 7-d RQ data in kcal·kg −1·d−1.

Incremental Exercise Testing

All tests were performed on a nondialysis day for the hemodialysis (HD) patients. The continuous ambulatory peritoneal dialysis (CAPD) patients performed all the tests with 2 L of fluid in the abdomen.

Each subject performed an incremental exercise test on a friction-braked cycle ergometer (Monark 814, Varberg, Sweden) for the determination of V̇O2peak and VT, according to the methods described by Koufaki et al. (19). Respiratory gas response data were measured throughout the V̇O2peak test by using a fast response breath by breath gas analyzer (Quark b2, COSMED, Rome, Italy). Heart rate (HR) and rhythm were monitored continuously using a 3-lead ECG (Tango SunTech Medical Instruments, Moreton-in-Marsh, Gloucestershire, UK). Blood pressure (BP) was also measured and displayed on an oscilloscope (Tango SunTech Medical Instruments).

Reproducibility data, expressed as percent coefficients of variation (CV%) and SEM, for patients’ exercise responses at peak exercise and at VT has been previously established in our laboratory. The determination of these parameters was performed using the same protocol of exercise testing on 12 patients who agreed to repeat testing on the same day and approximately the same time 1 wk apart (19).

Oxygen Uptake Kinetics

V̇O2 kinetics response was determined on a different day from the incremental test. To determine the V̇O2 kinetics on response (time constant), all subjects performed one 6-min constant load exercise bout on a cycle ergometer at an exercise intensity corresponding to 90% of previously determined VT according to the methods described by Koufaki et al. (20). Oxygen uptake response was measured using a fast response breath-by-breath gas analyzer (Quark b2, COSMED).

Coefficient of variation and SEM for the time constant of V̇O2 kinetics were established following the same protocol on 12 patients who agreed to perform one more transition at 90%-VT after 1 wk at approximately the same time of the day from the first testing. Determination of these parameters was performed according to the methods extensively described by Koufaki et al. (19).

Training Protocol

Patients completed 6 months of aerobic training on a cycle ergometer. CAPD patients exercised three times per week in the Renal Rehabilitation Gym, under the supervision of an exercise physiologist. HD patients exercised during the first 2 h of dialysis also under the supervision of an exercise physiologist with a cycle ergometer (Monark 814) attached to the dialysis chair. Patients were carefully monitored in terms of BP and HR before and after each exercise session and within an exercise session. Patients who had to pause their training for greater than 2 wk had to start from the beginning again.

Exercise intensity in W, was initially set to correspond to 90% of VT. Oxygen uptake, work rate, ratings of perceived exertion (RPE; Borg RPE 6–20 scale), HR, and BP were recorded at the VT point. The 90% of V̇O2 at VT was calculated and then the corresponding power output in W was determined from the associated exercise test responses. Each exercise session was divided into a warm-up, conditioning, and cool-down section. Exercise training started gently with all the patients having to perform three separate bouts on a cycle ergometer, each of 6- to 8-min duration.

Exercise duration of cycling was gradually increased by adding 1 min to each conditioning bout, also depending on patients’ response to training. The aim was by the end of 12 wk of training that all the patients would be able to do two separate bouts of continuous cycling of 20 min each on the cycle ergometer. As soon as a patient was able to do two bouts of 20 min, then the duration of the first bout was increased by 1 or 2 min while the duration of the second one was decreased by 1 or 2 min so as to achieve eventually 40 min of continuous cycling.

The patients had their peak and submaximal exercise capacity reassessed at the end of the 3-month period, and exercise intensity in W during training was readjusted to the value corresponding to the new 90% of VT. RPE responses were used as a signal to adjust exercise intensity due to inherent variability in BP and HR in the patients and also because it provided a more sensitive picture of how patients felt during exercise. Initial conditioning RPE responses were set to corresponding to RPE at the 90% of VT power output. When patients consistently responded to a 1-unit decrease in RPE, the target power output in W was increased by approximately 5%.

Statistical Analysis: Analytical Approach

Results are presented as mean ± SD. Baseline comparison between the HC and patient groups for anthropometric characteristics and physical activity levels was performed using independent t-tests. Comparison of physical and clinical characteristics of the patient group was performed using repeated measures analysis of variance. Bonferroni post hoc analysis was performed if there were significant time main effects. Exercise training responses were analyzed using the following types of analyses: i) conventional statistical hypothesis testing using repeated measures analysis of variance with Bonferroni post hoc comparisons for time main effects, ii) group mean CV% to evaluate group mean percentage changes, and iii) SEM analysis to evaluate individual patient responses. All statistical analyses were performed using SPSS 9.0 for Windows (Chicago, IL). A P-value of ≤ 0.05 was considered statistically significant.


Exercise Training Responses

Of the 34 patients initially recruited, 18 patients (10 CAPD, 4 women) completed the 6 months of training (47% drop-out rate). Reasons for dropping out from the exercise program included: injuries unrelated to exercise program (N = 1), loss of interest (N = 3), noncompliance with dialysis treatment (N = 2), noncompliance with exercise prescription (N = 1), transportation problems to the renal gym (N = 2), development of medical conditions that required surgery or long term treatment (N = 2), frailty (N = 1), and death (N = 4).

By the end of 3 months of training, all patients could complete two bouts of conditioning of 20 min each within a single session at a mean workload of 62.9 ± 20.2 W corresponding to 114.5% of the baseline VT. By the end of the 6-month period, every patient, apart from two, was able to do 40 min of continuous cycling at a group mean workload of 66.5 ± 21.8 W corresponding to 121% of the baseline VT. Most of the patients tolerated and responded to the exercise training program well.

Physical and Clinical Characteristics

The physical and clinical characteristics of the patients are presented in Table 1. The group mean 7-d RQ score was 35.5 ± 6.6 kcal·kg−1·d−1. The anthropometric and physical activity characteristics of the HC group at baseline were as follows: age, 57 ± 16.5 yr; body mass index (BMI), 25.7 ± 3 (kg·m−2)−1; body mass, 74.7 ± 8.5 kg; and 7-d RQ, 38.8 ± 5.1 kcal·kg−1·d−1. No statistically significant differences (P > 0.05) existed at baseline between the patient and the HC group in any of the anthropometric characteristics and physical activity levels. Seven patients at baseline were taking β-blockers for blood pressure control. Three were taking angiotensin-converting enzyme inhibitors and/or calcium channel blockers, whereas the remainder were not taking any blood pressure medication at the time of the study.

Physical and clinical characteristics of the patients.

After exercise training, no changes were observed (P > 0.05) for body mass, BMI, Hb, albumin, total carbon dioxide, or parathyroid hormone. A significant improvement of 14.2% was observed in nutritional status, as assessed by SGA (5.6 ± 1.3 vs 6.4 ± 1, vs 6.5 ± 0.8). Post hoc analysis revealed that this improvement occurred after 3 months of training with no further improvement thereafter.

Peak Exercise Responses

Repeated measures analysis of variance with Bonferroni post hoc analysis revealed significant (P < 0.05) improvements for group mean V̇O2peak, power output, and time to exhaustion after 3 months of exercise rehabilitation. No further significant changes were observed thereafter (Table 2). Overall group mean percentage increases observed at the end of the training period were 19.6 ± 18.5% for V̇O2peak, 22.2 ± 20.9% for peak power output, and 32 ± 22.5% for time to exhaustion. Improvements in the mean values of the above variables exceeded the predetermined CV% of 4.7%, 10.3%, and 13%, respectively (19).

Exercise responses at peak exercise.

The SEM for V̇O2peak was calculated to be 0.07 L·min−1 (19). Individual assessment of patients’ V̇O2peak in response to exercise training revealed that in two (11%) patients V̇O2peak did not exceed the SEM either at 3 or 6 months of training. Eleven patients (61%) improved by more than 1 SEM after 3 months of training, and 16 (88.8%) improved by more than 1 SEM after 6 months of training (Fig. 1). Patients’ group mean data for V̇O2peak in reference to the HC group before during and after the exercise intervention is illustrated in Figure 2.

Individual patient responses to exercise training for V̇O2peak. Solid and broken lines indicate the cut-offs for patients exceeding 1 or 2 SEM after 3 or 6 months of exercise rehabilitation, respectively.
Group mean percentage deficits of ESRD patients compared with the group of HC at baseline and after 3 and 6 months of exercise training.

Exercise Responses at VT

Group mean V̇O2, time to the onset of the VT, and power output significantly improved (P < 0.05) after 6 months of training by 24.5 ± 25.4%, 47 ± 61%, and 23.7 ± 33.3%, respectively (Table 3). The percentage improvement was higher than the CV% of 6.6%, 11.7%, and 9.7%, respectively (19). The SEM for V̇O2-VT was calculated to be 0.06 L·min−1 (19). Individual patient assessment revealed that three (16%) patients did not alter their V̇O2 at VT by more or less than 1 SEM either after 3 or 6 months of exercise. Nine patients (50%) improved by more than 1 SEM after 3 months of training, and 14 patients (77.7%) improved after 6 months of training.

Exercise responses at VT and V̇O2 kinetics.

The HC group at baseline achieved a V̇O2 at VT of 1.19 L·min−1, corresponding to a 61.6 ± 10.7% of the V̇O2peak value. The respective numbers for the patient group at baseline were 0.86 L·min−1 and 61.9%. The patient group mean V̇O2 response at VT after training in reference to the HC group is presented in Figure 2.

Oxygen Uptake Kinetics

Table 3 demonstrates that the oxygen uptake kinetic response at the relativized exercise intensity of 90% of VT was significantly faster (P < 0.05) after 3 months of training and remained unchanged after 6 months of training. The percentage improvement of 24.2 ± 16.1% and 23.8 ± 20.2% at the end of the 3- and 6-month training periods, respectively, exceeded the CV% of 19.8%. In eight patients (∼45%), V̇O2 kinetics response at the relativized exercise intensity remained within the SEM (12.3 s) after 3 and 6 months of exercise. Figure 3 illustrates that in 10 patients (∼55%) the improvement in the time constant exceeded the SEM after 3 months on the program, with no further increases in the number of patients exceeding the SEM after 6 months of exercise training.

Individual patient responses to exercise training for V̇O2 kinetics. The solid and broken lines indicate the cut-offs for patients exceeding 1 or 2 SEM after 3 or 6 months of exercise rehabilitation, respectively.

The HC group had a group mean time constant of 46.9 ± 13.1 s at the relativized exercise intensity of 90% of VT. The patient group at baseline had a slower kinetic response compared with the HC group by ∼ 44%. After the end of the training period, the patient group was slower by ∼ 10% compared with the HC group at the exercise intensity at 90% of VT.


This study attempted to characterize the adaptive response of peak and submaximal exercise tolerance of patients with ESRD in response to a 6-month program of exercise rehabilitation. It also highlighted the substantial degree of individual variability of adaptive response in this patient group.

Patients from the present study had a V̇O2peak within the range of values reported in the literature (14–22 mL·kg−1) (5,10,18,21,22,28). Patients and HC subjects were characterized as sedentary, based on their self-reported weekly energy expenditure on physical activities (17). The HC group examined had a lower V̇O2peak compared with other healthy individuals reported elsewhere (5,10,18). This is acceptable, however, considering the older age (∼ 10 yr older) of the healthy subjects assessed for the purposes of the present study.

At baseline, the percentage deficit in V̇O2peak of the ESRD patients compared with HC group was consistent with the value of about 30–40% (see Fig. 2.) identified in most previous reports of exercise tolerance (5,10,18). After 6 months of exercise training, the patient group mean V̇O2peak was still about 20% lower than the HC group’s, indicating the persistence of oxygen transport and utilization limitations even in “trained” ESRD patients.

The percentage increase in V̇O2peak after 6 months of training in the present study (∼20%) was lower than the improvement reported recently (∼ 48% increase) for exercise studies of the same duration (10,21). Variations in exercise mode, frequency, and intensity of training may have accounted for this discrepancy because the exercise program employed by the latter investigators was reported to be more intense. The differences in the percentage improvements observed across studies could also reflect the large heterogeneity of the renal dialysis population in terms of age, comorbid conditions, nutritional status, and/or motivation.

This heterogeneity in the response of individual patient V̇O2peak to exercise training is evident in the present study. The group mean percentage changes for V̇O2peak after training (13.9 ± 13.2% at 3 months and 19.6 ± 18.5% at 6 months) exceeded the group mean CV% of 4.7%, suggesting that a meaningful change of substantial magnitude was likely. However, the large SDs observed in the group percentage changes underscore the heterogeneity of individual responses, indicating that some patients might have improved by a substantial amount but some others might have not improved at all. Evaluation of patients’ response on an individual basis using the SEM showed that 61% of the patients improved V̇O2peak by more than 1 SEM after 3 months on the exercise program. Yet the percentage of patients that improved after 6 months of exercise training reached 89%, even though conventional statistical hypothesis testing did not show any significant improvements from 3 to 6 months. This type of evaluation therefore added more sensitivity to the information regarding the time course of adaptation for the improvement in V̇O2peak in patients with ESRD and helped to identify the minimum period of exercise training needed to maximize the number of subjects benefited from the intervention. This outcome would have to coincide with the initial target set. For example, some investigators might want to deem an intervention successful only if more than 70% of the patients respond positively. Conventional hypothesis testing alone is not sufficient to determine whether this target has been achieved.

There were, however, two patients that did not show any improvement in V̇O2peak after training. One of these patients did perform three exercise sessions a week but did not completely comply with the prescribed exercise training intensity. The second patient trained hard and fully complied with the exercise program and prescription, yet failed to show any improvements. The latter patient was an insulin-dependent diabetic patient for years and at the same time of the study had extremely high levels of parathyroid hormone. Increased levels of parathyroid hormone are known to impair skeletal muscle function and energy production both in vitro and in vivo (7). It is possible, therefore, that these conditions compromised the adaptive response to exercise training in this patient. More interestingly, several investigators have recently drawn attention to the possibility of an existing genetic variation of physical fitness and adaptability to exercise training (15,26). The extent to which genetics can influence physical fitness has not yet been investigated in patients with ESRD. However, it is likely that these genetically predetermined “limitations” also apply in the ESRD population and thus prevent dramatic increases in V̇O2 or normalization of the same parameter.

In contrast with the peak exercise responses, a longer period of exercise training was required to induce statistically significant improvements in selected parameters measured at VT. A significant increase of ∼ 25% in the V̇O2 at VT group mean response was observed in the present study at the end of the training program. The group mean CV% for V̇O2 at VT was found to be 6.6% (19), suggesting that at least ∼18% of the observed improvement was likely due to exercise training. Changes in VT responses of ESRD patients after training interventions are rare to find in the literature. Zabetakis et al. (28) reported a 12% increase in V̇O2 after 10 wk of aerobic exercise in middle-aged, no comorbid, anaemic HD patients. Another training study by Ota et al. (23) conducted in elderly anaemic HD patients showed a group mean increase of 6%. The meaningfulness of such “improvement” is questionable given the CV% of 6.6 for this parameter reported by Koufaki et al. (19). Moreover, a major limitation present in both studies was the very small number of patients (8 and 5, respectively), a fact that limits the generalizability of these findings. The present study showed that it takes about 3 months of moderate-intensity exercise training for 50% of the patients to demonstrate a meaningful adaptation that exceeded at least 1 SEM. Nearly 80% of the patients exceeded the SEM after 6 months of training, again indicating that to optimize the number of the patients that positively respond to this type of intervention, longer periods of exercise training may be necessary.

Three patients (16%) were classified as nonresponders to exercise training by using the SEM criterion. Two of these patients were the same patients who did not increase V̇O2peak, probably because of the same reasons mentioned earlier. No reasonable explanation exists for the third one, despite the fact that he too was “a good exerciser.” Nevertheless, even though the group of patients and the patients as individuals demonstrated meaningful improvements in the V̇O2 response at the VT, there was still a ∼ 12% mean group deficit in comparison with the HC group after the end of the training period (see Fig. 2). This persistence of impaired submaximal exercise tolerance may be indicative of factors mentioned earlier or relating to coexisting medical conditions of various degrees of severity. It could also reflect an insufficiency of exercise training volume, or its components to induce adaptations that would result in normalization of exercise tolerance at VT.

The time to the onset of VT and power output also increased after 6 months by 47% and 23.7%, respectively. The significant prolongation of the time to onset of the VT at higher exercise intensity may have significant implications in everyday life, especially for individuals with chronic disease and disability. Such an adaptation actually translates into a significant delay of the onset of fatigue at higher absolute metabolic demand and thus possibly the prevention of early cessation of a physical task. Lactate threshold and its sometime “proxy,” ventilatory threshold, have been associated with changes in muscle mitochondrial content as assessed through mitochondrial enzyme activity (9). Increases in muscle mitochondrial oxidative capacity as a consequence of training are related to better muscular regulation of substrate utilization with enhanced fat oxidation, diminished carbohydrate breakdown, and thus less lactate production (12).

By using conventional statistical hypothesis testing, it was demonstrated that V̇O2 kinetics at an exercise intensity corresponding to 90% of the VT of a group of ESRD patients were significantly speeded up after 3 months of aerobic exercise training by ∼24 ± 16.1% with no further significant changes observed onward. The group CV% was found to be 19.8%, and therefore one can postulate that at least 4% of the improvement was due to exercise training intervention.

The effects of exercise training on the response of V̇O2 kinetics in clinical populations are rare to find in the literature (6,24). Otsuka et al. (24), investigating the effects of an 8 wk exercise program on V̇O2 kinetics response of patients with obstructive pulmonary disease, reported that the V̇O2 kinetic response became faster by 16% (from 63.2 to 53.2 s). V̇O2 kinetics in that study were determined at a workload corresponding to 90% of their patients’ lactate threshold. Results from the study by Otsuka et al. (24) may be questionable because no data on reproducibility for V̇O2 kinetics was reported for their patients, who were severely deconditioned (V̇O2peak was 13.6 mL·kg−1 after training). Given the observations on measurement error of V̇O2 kinetics, it is possible that the magnitude of improvement they reported (16%) could be within the limits of measurement error.

In the present study, analysis of the individual adaptive responses of oxygen uptake kinetics was also performed using the SEM criterion (12.3 s). Individual breakdown of the responses revealed that only about half of the patients (55%) actually became faster by more than 12.3 s after 3 months of training with no increases in the number of patients exceeding the SEM after 6 months of training. In an attempt to identify potential factors explaining the lack of training response in nearly half of the patients, a retrospective analysis was performed. The group of patients was divided into subgroups of responders (exceeded the SEM) and nonresponders (not exceeded the SEM). Independent t-tests analysis revealed that the only significantly different parameter at baseline between the two groups was the nutritional state as assessed by SGA with the group of responders exhibiting a better nutritional status than the nonresponders (mean group SGA: 6.5 ± 0.5 vs 4.4 ± 1.1, respectively;P < 0.05). Nutritional state and its effects on muscle protein catabolism that results in loss of muscle mass and worsening of quality of the remaining muscle is well established in renal failure (7,8,14). It would seem plausible therefore to speculate that the lower SGA may reflect poorer quality and lesser amount of functional muscle mass. In turn, this quantitative difference may be indicative of a qualitative disturbance of intrinsic muscle characteristics that may have inhibited the adaptation of V̇O2 kinetics.

Nevertheless, it would also seem reasonable to speculate based on the V̇O2 kinetic information for the 10 patients who did become faster after the exercise intervention that factors mostly related to peripheral adaptations to exercise training, such as increased mitochondrial activity, increased capillarity, or better local blood flow control, are responsible for the faster adjustment to the onset of a submaximal exercise task (2,13,24,25). Determinants of V̇O2 kinetics, such as cardiac output and HR, were likely to be eliminated in the present study because the first 15 s were excluded from the kinetic analysis and the exercise transition to the 90% of VT was initiated from a 2-min period of loadless cycling.

Compared with the HC group, the patient group demonstrated a slower group mean kinetic response at baseline of ∼ 45%. This percentage deficit was reduced to ∼10% by the end of the 6-month training period. It should be noted, however, that this dramatic improvement and near normal response of the patient group is driven by only half of the patients, as shown by the individual response data, and thus its meaningfulness remains in question.

In summary, although conventional statistical analyses indicated that exercise training favorably altered peak and submaximal exercise capacity of patients with ESRD, it is apparent that considerable interindividual variability exists in the response to training. Consideration of the SEM data underscores the heterogeneity of adaptive response in this patient group. After 3 and 6 months of exercise training, this type of evaluation highlighted potentially important variations in the percentage of patients with substantially improved V̇O2peak and V̇O2 at VT, for instance, compared with those with substantially improved V̇O2 kinetics at 90%-VT. That “substantial” amount is defined, in our study, by the SEM, which allows one to identify, illustrate, and categorize the magnitude of the individual improvement, something that is not possible to do by only using CV% or group mean % change values. For example, by applying the SEM, you can identify how many people changed more or less than 1, 2, or even 3 SEMs. In this way, one can enhance interpretation of change data by relating it to numbers of individual patients improving (or otherwise) as well as quantifying meaningfully the extent (magnitude) of improvement seen. Also, as discussed earlier, the number of the nonresponders can also be described. But to do that, one needs an objective criterion beyond or below which a subject can be described as a responder or not. In renal failure, for instance, the kt/v needs to be at least higher than 1 in order for clinicians to assure that adequate hemodialysis is occurring. In exercise rehabilitation science, one also needs to define some limits above or below which the exercise physiologist is confident that adequate rehabilitation has been achieved and not rely just on a significance of a P-value. Moreover, by applying the SEM analysis, additional sensitivity was also added to information concerning the observed time course of adaptation for the improvement in various parameters measured over the course of the entire intervention period. Therefore, it may be important for future studies to establish and utilize sample-specific measurement error (reproducibility) data in the evaluation of exercise training effectiveness for patients with end stage renal disease.

Pelagia Koufaki was a recipient of a Jansen-Cilag Ltd research scholarship.

Address for correspondence: Pelagia Koufaki, Centre for Biophysical and Clinical Research into Human Movement, Department of Exercise and Sport Science, Manchester Metropolitan University, Alsager Campus, Hassall Road ST7 2HL, Alsager Stoke on Trent, United Kingdom; E-mail: [email protected]


1. Atkinson, G., and A. M. Nevill. Statistical methods for assessing measurement error (reliability) in variables relevant to sports medicine. Sports Med. 26: 217–238, 1998.
2. Babcock, M. A., D. H. Paterson, and D. A. Cunningham. Effects of aerobic endurance training on gas exchange kinetics of older men. Med. Sci. Sports Exerc. 26: 447–452, 1994.
3. Batterham, A. M., and K. P. George. Reliability in evidence-based clinical practice: a primer for allied health professionals. Phys. Ther. Sport. 1: 54–62, 2000.
4. Blair, S. N., W. L. Haskell, P. Ho, et al. Assessment of habitual physical activity by a seven day recall in a community survey and controlled experiments. Am. J. Epidemiol. 122: 794–804, 1985.
5. Beasley, C. R. W., D. A. Smith, T. J. Neale. Exercise capacity in chronic renal failure in patients managed by continuous ambulatory peritoneal dialysis. Aust. N. Z. J. Med. 16: 5–9, 1986.
6. Branderburg, S. L., J. E. B. Reusch, T. A. Bauer, B. W. Jeffers, W. R. Hiatt, and J. G. Regensteiner. Effects of exercise training on oxygen uptake kinetic responses in women with type 2 diabetes. Diabetes Care 22: 1640–1646, 1999.
7. Brautbar, N. Skeletal myopathy in uremia: abnormal energy metabolism. Kidney Int. 16: S81–S86, 1983.
8. Castaneda, C., L. Grossi, and J. Dwyer. Potential benefits of resistance exercise training on nutritional status in renal failure. J. Ren. Nutr. 8: 2–10, 1998.
9. Basset, D. R., and E. T. Howley. Limiting factors for maximum oxygen uptake and determinants of endurance performance. Med. Sci. Sports Exerc. 32: 70–84, 2000.
10. Deligiannis, A., E. Kouidi, E. Tassoulas, P. Gigis, A. Tourkantonis, and A. Coats. Cardiac effects of exercise rehabilitation in hemodialysis patients. Int. J. Cardiol. 70: 253–266, 1999.
11. Enia, G., C. Sisuso, G. Alati, and C. Zoccali. Subjective global assessment of nutrition in dialysis patients. Nephrol. Dial. Transplant. 8: 1094–1098, 1993.
12. Gollnick, P. D., and D. R. Hodgson. Enzymatic adaptation and its significance for metabolic response to exercise. Int. Series Sports Sci. 16: 191–200, 1986.
13. Grassi, B., L. B. Gladden, M. Samaja, C. M. Stary, and M. C. Hogan. Faster adjustment of O2 delivery does not affect VO2 on-kinetics in isolated in situ canine muscle. J. Appl. Physiol. 85: 1394–1403, 1998.
14. Guarnieri, G., G. Toigo, R. Situlin, et al. Muscle biopsy studies in chronically uremic patients: evidence for malnutrition. Kidney Int. 24: 187–193, 1983.
15. Hagberg, J. M., R. E. Ferrell, S. D. Mccole, K. R. Wilund, and G. E. Moore. VO2max is associated with ACE genotype in postmenopausal women. J. Appl. Physiol. 85: 1842–1846, 1998.
16. Hopkins, W. Measures of reliability in sports medicine and science. Sports Med. 30: 1–15, 2000.
17. Johansen, K. L., G. M. Chertow, A. V. Ng, et al. Physical activity levels in patients on hemodialysis and healthy sedentary controls. Kidney Int. 57: 2564–2570, 2000.
18. Kettner, A., A. Goldberg, J. Hagberg, J. Delmez, and H. Harter. Cardiovascular and metabolic responses to submaximal exercise in hemodialysis patients. Kidney Int. 26: 66–71, 1984.
19. Koufaki, P., T. H. Mercer, and P. F. Naish. Reproducibility of exercise tolerance in patients with end stage renal disease. Arch. Phys. Med. Rehabil. 82: 1421–1424, 2001.
20. Koufaki, P., T. H. Mercer, and P. F. Naish. Dialysis mode does not affect exercise intolerance of patients with end stage renal disease. Clin. Exerc. Physiol. 3: 154–160, 2001.
21. Kouidi, E., M. Albanis, K. Natsis, et al. The effects of exercise training on muscle atrophy in haemodialysis patients. Nephrol. Dial. Transplant. 13: 685–699, 1998.
22. Lo, C. Y., L. Li, W. K. Lo, et al. Benefits of exercise training in patients on continuous ambulatory peritoneal dialysis. Am. J. Kidney Dis. 32: 1011–1018, 1998.
23. Ota, S., K. Takashashi, H. Suzuki, S. Nishimura, H. Makino, Z. Ota, K. Taniai, T. Kaneshige. Exercise rehabilitation for elderly patients on chronic hemodialysis. Geriatr. Nephrol. Urol. 5: 157–165, 1996.
24. Otsuka, T., N. Kurihara, T. Fujii, S. Fujimoto, and J. Yoshikawa. Effect of exercise training and detraining on gas exchange kinetics in patients with chronic pulmonary disease. Clin. Physiol. 17: 287–297, 1997.
25. Phillips, S. M., H. J. Green, M. J. Macdonald, and R. L. Hughson. Positive effect of endurance training on VO2 kinetics at the onset of submaximal exercise. J. Appl. Physiol. 79: 1914–1920, 1995.
26. Rankinen, T., L. Perusse, J. Gagnon, et al. Angiotensin-converting enzyme ID polymorphism and fitness phenotype in the HERITAGE Family Study. J. Appl. Physiol. 88: 1029–1035, 2000.
Visser, R., F. W. Dekker, E. W. Boeschoten, P. Stevens, and R. T. Krediet. Reliability of the 7-point subjective global assessment scale in assessing nutritional status of dialysis patients. Adv. Perit. Dial. 15: 222–225, 1999.
28. Zabetakis, P. M., G. W. Gleim, F. L. Pasternack, A. Saraniti, J. A. Nicholas, and M. F. Michelis. Long duration submaximal exercise conditioning in hemodialysis patients. Clin. Nephrol. 18: 17–22, 1982.


©2002The American College of Sports Medicine