A Critical Review of Exercise Training in Hemodialysis Patients: Personalized Activity Prescriptions Are Needed

INTRODUCTION

For several decades, well-intentioned researchers and clinicians from across the globe have been conducting clinical trials or implementing exercise programs in hemodialysis (HD) patients with the primary goal of improving their health. Data from these trials have been encouraging, as evidenced by the extensive list of benefits highlighted in numerous systematic reviews and meta-analyses of exercise, including improvements in physical function, markers of cardiovascular disease (CVD) risk, and quality of life (QOL) (^1–3). Based on this literature, many have argued that exercise should be included as a part of the standard of care for all patients with chronic kidney disease (CKD), including HD patients (^4,5). Several guidelines and position statements regarding exercise in HD also have been published by national and international organizations (^6–8). The guidelines are similar to recommendations for the general population in terms of frequency, intensity, and types of physical activity, including combinations of endurance, resistance, flexibility, and balance exercises on most days of the week (⁹).

Despite these efforts, exercise programs have not become a component of the standard of care for HD patients and continue to be the exception, rather than the rule, in HD clinics (¹⁰). More importantly, physical activity levels in HD patients continue to be discouragingly low, and poor physical function remains a hallmark of this disease (¹¹). This suggests that the typical approaches used to implement exercise programs in HD patients have largely failed, and novel strategies should be considered.

Numerous studies have been conducted to identify barriers to exercise in HD patients (^12,13). One major barrier is the high burden of comorbidities that significantly affects patient's health and QOL. Muscle wasting, bone disorders, and cardiovascular complications are especially common. In addition, two thirds of HD patients die within 5 yr of starting long-term dialysis treatment (¹⁴). Although the etiologies of these comorbidities are complex, physical inactivity is likely both a cause and consequence.

Another often-cited barrier includes lack of resources and expertise at HD clinics to implement or sustain exercise programs (^12,13). Most of the published interventions have used “intradialytic cycling” as the primary mode of exercise, which involves patients cycling on ergometers placed in front of their dialysis chairs or beds. HD patients typically dialyze 3 d a week for several hours and normally have little to do during treatments, so intradialytic exercise is viewed as relatively convenient and time efficient. However, intradialytic exercise constrains the type, volume, and intensity of exercise that can be performed, so a number of studies have also assessed the efficacy of out-of-center exercise programs. Although there are advantages and drawbacks to each approach, recent studies suggest the overall benefits of intradialytic and out-of-center (“interdialytic”) exercise are similar. This has held true for both endurance and resistance training interventions (²).

Data from many published surveys and other qualitative studies indicate that both patients and clinicians believe exercise is beneficial for HD patients (^12,13,15,16). Despite this, these same studies indicate that implementing exercise programs is not a priority for the medical staff at dialysis clinics. This suggests one of two things: 1) the medical staff does not care enough about the health of their patients to prioritize exercise programs, which is clearly absurd; or 2) the medical staff does not feel it is capable of implementing a program that is robust enough to realize the potential benefits that exercise has to offer. The latter is obviously the more plausible explanation. Indeed, qualitative data from these studies suggest that clinicians do not believe most patients want to exercise, they feel they lack the expertise to prescribe it, and other work-related obligations make exercise programs impractical to implement (^12,13,15,16). Taken together, a reasonable conclusion from this data is that clinicians are not convinced that exercise programs, as typically structured, provide enough benefits to commit the resources necessary to implement a program. As a result, the support needed to implement and sustain robust exercise programs has not been sufficient. This ambivalence may be due in part to the fact that data on the benefits of exercise from randomized clinical trials (RCTs) and other clinical trials have not been as robust as generally suggested (^17,18). Although outcomes from exercise-related clinical trials may often reach the threshold of statistical significance, the data have not been impressive enough to inspire the nephrology community to make exercise a part of the standard care in HD clinics.

The primary purpose of this review is to 1) briefly summarize the findings from recent major RCTs and other selected studies that have evaluated the efficacy of exercise training on markers of physical function and CVD risk in HD patients, 2) discuss reasons for some of the equivocal findings from these studies, and 3) suggest novel research approaches for increasing physical activity levels in this population. Our primary hypothesis is that exercise interventions often fail to produce clinically significant improvements in the health and QOL of HD patients, primarily because the volume and intensity of the exercise prescribed are insufficient.

Although not intended to be an exhaustive review of the literature, we have focused on results from several recent studies we believe to be the most significant to date in terms of sample size, intervention length, outcomes measured, and other methodological considerations. Our review seeks to highlight some of the major concerns in the literature, with the goal of using the lessons we have learned to help move the field forward with better ideas and approaches. We believe that if we continue to only highlight or exaggerate our successes, while ignoring the reality of our failures, we may never realize the true potential of exercise as a medicine for this critically ill patient population.

Endurance Exercise Training: Effects on Physical Function

Most exercise interventions in HD patients have investigated the effects of endurance training on physical function and related parameters, including aerobic capacity and muscle strength. Unfortunately, most of these studies have had significant shortcomings, such as small sample sizes, short intervention periods, or inadequate control groups, so it is often difficult to confidently interpret the data. By contrast, several larger RCTs examining the effects of endurance exercise on physical function have been published in recent years that are highly informative, including interventions by Manfredini et al. (¹⁹) and Koh et al. (²⁰). Although the results from the trial by Manfredini et al. generally showed significant benefits of exercise on physical function, the trial by Koh et al. did not. Reasons for the apparently discrepant results from these and other endurance training interventions will be discussed.

The Exercise Introduction to Enhance Performance in Dialysis (EXCITE) trial conducted by Manfredini et al. (¹⁹) examined the effects of a 6-month, low-intensity, home-based walking program on physical function. In many regards, the EXCITE trial was one of the most impressive exercise-related RCTs conducted to date in HD patients. The protocol included a relatively long-term exercise intervention (6 months) with a very large sample size (total N = 296 patients) and also included a nonexercising control group. By comparison, most exercise interventions in HD patients include much smaller sample sizes (e.g., 10 to 20 patients per group) and a shorter intervention period (e.g., 3 to 4 months), and often lack a nonexercising control group. The primary outcome in the EXCITE trial was significantly improved performance on the 6-min walk test (~12%) and other functional measures in the exercise group, compared with controls. Importantly, patients with the highest adherence to the exercise protocol had the largest improvements in their performance, indicating a dose-response effect from the exercise. These findings are both encouraging and generally consistent with other studies in the literature that have shown modest improvements in physical function measures after 3 to 6 months of exercise training (^1–3).

Another exercise-related RCT that shares some similar features with the EXCITE trial was published a few years earlier by Koh et al. (²⁰). In this study, HD patients were randomized to one of three groups for 6 months: 1) home-based walking program (unsupervised), 2) intradialytic cycling (supervised), or 3) nonexercising control group (total n = 70). Both activity groups had a goal of exercising at a moderate intensity for ~30 to 45 min⋅d⁻¹, 3 d⋅wk⁻¹. The primary outcome being assessed was a change in performance on the 6-min walk test. However, in contrast to the results from the EXCITE trial, changes in the 6-min walk test did not differ significantly between the exercise and control groups. Moreover, self-reported physical function actually regressed in the intradialytic cycling group compared with the controls.

At first glance, these trials seem to have highly discrepant results. However, a deeper analysis indicates many similarities, particularly in terms of the volume of exercise, adherence rates, and even the primary outcomes. For example, the average increase in performance on the 6-min walk test in the exercise group in the EXCITE trial was 39 m (12%), compared with an increase of just 2 m in the control group (P < 0.0001 for comparison between groups). In the study by Koh et al., the magnitude of improvement in the 6-min walk test was 63 m (14%), 49 m (11%), and 21 m (5%) in the intradialytic cycling, at-home walking program, and control group, respectively. However, because the Koh et al. trial had a much smaller sample size (70 vs 296), the changes in walking performance were not statistically significant. In a post hoc power analysis, the authors determined that 40 patients per group would have been needed to detect a statistically significant improvement.

Although the study by Koh et al. was underpowered, it is important to emphasize that the magnitude of improvements in physical function in both studies was modest, and it is unclear if the changes that were seen are clinically meaningful. Some studies suggest improvements in the 6-min walk test of just 14 to 31 m are clinically significant (²¹), whereas other data suggest changes of at least 54 to 80 m are the minimum needed to be clinically relevant, in part due to the inherent variability in performance on this test (²²).

Another similar characteristic between these two trials was the low volume and intensity of the exercise prescription, which may have accounted for the relatively modest improvements in function. In Koh et al., it was determined that the energy expended during a typical intradialytic cycling bout at the end of the 6-month intradialytic cycling intervention averaged just 35 kcal per session. Although similar data are not available for the EXCITE trial, the volume of exercise prescribed was indeed low. The exercise protocol required patients to walk for a maximum of 20 min, 3 d⋅wk⁻¹ (60 min⋅wk⁻¹ total), which is about 60% lower than the 150 min⋅wk⁻¹ of moderate intensity aerobic exercise that is suggested in most exercise guidelines (⁹). Despite this modest exercise prescription in the EXCITE trial, there was a significant dropout (31%) of participants in the exercise group (compared with 15% in the control group), and nearly 50% of those completing the intervention had low adherence to the exercise protocol (defined as completing <60% of exercise sessions). The study also excluded patients with both low mobility and high fitness, so just 30% of patients at the participating clinics met the study inclusion criteria, and 36% of those refused to participate.

Although by many metrics the EXCITE trial was one of the best exercise interventions conducted to date in HD patients, just 14% of the patients in the participating clinics were enrolled and able to adhere to a 6-month walking program that is 60% below the recommended volume of aerobic exercise suggested in standard physical activity guidelines. In addition, those who did comply with the exercise prescription only had modest improvements in physical function.

Unfortunately, the low volume and intensity of the exercise prescriptions, poor compliance and/or adherence, and high dropout rates are the norm, rather than the exception, in the exercise literature in HD patients (^19,20,23,24). The Table highlights examples of aerobic exercise training studies in which the volume and intensity of exercise were quantified in some manner. Whether expressed as total exercise time per week, average watts, or estimated energy expenditure per session, it is obvious that the exercise prescriptions are extremely modest. It is important to note that most interventions published in the literature to date, including studies from our own laboratories, often fail to accurately assess or report exercise work rates. Based on our experience, this is due in part to the difficulty in calibrating the ergometers commonly used for intradialytic cycling. Regardless, the exercise prescriptions described in the Table seem to be similar in volume and intensity to most of the published studies in the literature. Developing strategies for increasing the volume and intensity of exercise in HD patients has to be an important priority if we hope to significantly improve the efficacy of our interventions.

In summary, data from several meta-analyses and systematic reviews that pool data from multiple small and larger studies indicate statistically significant improvements in physical function after 3 to 6 months of endurance exercise training in HD patients (^1–3). Although these analyses are informative, it also is important to highlight the deficiencies in the data and to critique some of the approaches that have been used to collect it. Indeed, details from robust RCTs such as those conducted by Manfredini et al. (¹⁹) and Koh et al. (²⁰) highlight some of the major challenges associated with implementing exercise programs in HD patients and suggest areas for improvements that should be considered in future study designs.

Endurance Exercise Training: Effects on Markers of CVD Risk

In 1986, Goldberg et al. (²⁵) conducted one of the first trials to examine the effects of exercise on CVD risk in HD patients. This was a 12-month intervention in which patients cycled at 80% of their V̇O_2max for 10–20 min during a 45-min exercise session. The primary outcomes included improvements in V̇O_2max and plasma lipids, and a reduction in hypertensive medications. Since this early study, many other reports have been published that generally confirm a beneficial effect of exercise training on traditional CVD risk factors in HD patients (^1–3). However, the apparent reverse epidemiology between some CVD risk factors (e.g., serum cholesterol) and CVD-related outcomes and mortality (²⁶) makes it difficult to interpret these findings.

By contrast, markers of cardiac function, endothelial function, and arterial stiffness have been consistently associated with CVD mortality in patients with CKD (²⁷). In healthy populations, exercise improves cardiovascular and arterial function (²⁸). It has been hypothesized that exercise may also have similar benefits in those with CKD, but the data supporting this are, surprisingly, weak. In nondialysis patients with CKD, recent data suggest no effect of endurance training on blood pressure, endothelial function, or arterial stiffness (^29,30). By contrast, Kirkman et al. (³¹) showed that 12 wk of endurance exercise significantly improved microvascular function in patients with CKD, but central arterial hemodynamics and arterial stiffness once again did not improve.

The data on exercise and arterial function in HD patients also are mixed. Although at least three small pilot studies noted modest improvements in markers of arterial stiffness after several months of endurance exercise training (^32–34), a larger exercise RCT did not (²⁰). Details of these studies examining the impact of exercise on arterial stiffness in HD patients will be described, as well as potential reasons for the discrepant results.

One study consistently cited as evidence that endurance exercise reduces arterial stiffness in HD patients was a crossover study conducted by Toussaint et al. (³³). In this study, 19 HD patients were randomized to two groups that either performed 30–45 min of intradialytic cycling for 3 months or remained sedentary during treatment before crossover. The primary outcome was pulse wave velocity (PWV), a “gold standard” measure of arterial stiffness. The primary conclusion in the article states that “there was an overall improvement in PWV,” suggesting that exercise reduces arterial stiffness. Although this conclusion is generally supported by their data, some details are needed to provide a proper context. First, in the primary analysis, it should be noted that there was a trend for a reduction (improvement) in PWV that was not statistically significant. However, when data from both groups were combined, the average PWV after the exercise period was significantly lower than the average PWV after the control period. Another factor that should be considered was the low exercise volume and intensity performed. According to data provided in the article, patients cycled an average of just 0.9 km in about 39 min. This is an average velocity of ~1.4 km⋅h⁻¹ or ~0.9 miles⋅h⁻¹. Although this is obviously a very modest volume and intensity of exercise, it is generally consistent with what has been prescribed in the other studies described previously (Table). It is reasonable to question whether this is a sufficient amount of exercise to stimulate arterial wall remodeling, especially given the significant vascular calcification that is a hallmark of renal failure.

A similar study cited as evidence that exercise improves arterial stiffness in HD patients was conducted by Mustata et al. (³²). This study demonstrated that 3 months of aerobic exercise, consisting of a 1-h exercise class twice per week, reduced arterial stiffness. Moreover, the changes in stiffness regressed toward baseline levels after a 1-month detraining period. Although these findings are again encouraging, there were several limitations to this study. First, this was a small pilot study (N = 11) that did not include a nonexercising control group. Second, arterial stiffness was estimated using pulse wave analysis to determine the “augmentation index” (AIx). Although the AIx provides valuable information about wave reflection that may suggest structural or functional changes in the arterial wall, it is not considered as precise as PWV or other methods for quantifying arterial stiffness.

To our knowledge, the previously mentioned study from Koh et al. (²⁰) is the only RCT to date that has investigated the effects of endurance exercise training on arterial stiffness in HD patients. In this study, neither 6 months of intradialytic cycling or an at-home walking program improved measures of arterial stiffness (PWV or AIx). The authors suggest the negative results may be due in part to changes in medication use or perhaps to the relatively modest exercise prescription used (~35 kcal of work performed per session; Table).

Taken together, these studies provide limited evidence for a benefit of exercise on arterial stiffness in HD patients. This may be due to a combination of factors, including chronic volume overload, impaired glucose metabolism, systemic inflammation, and alterations in mineral metabolism, that contribute to significant atherosclerosis and vascular calcification in advanced CKD (³⁵). The vascular remodeling mediated by these factors may limit the ability of vessels to respond to an exercise intervention, especially given the modest volume of physical activity that is typically prescribed.

Benefits of Resistance Exercise Training in HD

HD patients have a high prevalence of protein energy wasting (PEW) (³⁶). Reductions in muscle mass limit the patient's ability to perform activities of daily living, exacerbate the development of common comorbidities such as CVD, and increase mortality rates. Resistance training is known to be a more potent anabolic stimuli for skeletal muscle than endurance exercise, and so may be especially important in HD patients. Unfortunately, there are fewer studies examining the benefits of resistance training compared with endurance exercise in this population. This may be due in part to difficulty of performing resistance exercises during dialysis, as well as the perception that resistance training outside of the clinic is often impractical because many patients have limited access to proper equipment.

Despite this, several recent systematic reviews of resistance training in HD patients have been published, and these generally indicate increases in muscle hypertrophy, strength, and objective measures of physical function after progressive resistance training (^1,2,37). Moreover, a recent study by Chen et al. (³⁸) showed that even low-intensity strength training can yield modest improvements in physical function, strength, and body composition. However, similar to the data with endurance training, the benefits from many of the resistance training studies are either modest or inconsistent. For example, several resistance training interventions in HD patients with the largest sample sizes and longest intervention periods have not yielded significant increases in muscle hypertrophy (^39–41). A recent meta-analysis concluded that although muscle strength seems to improve with resistance training, muscle mass does not (³⁷). Finally, several recent studies found improved muscle size or composition in response to resistance training, but failed to improve objective measures of physical function (^39,42,43). Reasons for some of these inconsistent findings are discussed.

One highly referenced RCT examining the effects of resistance training on muscle mass in HD patients is the Nandrolone and Exercise Trial (NEXT) (⁴²). This 12-week intervention examined the individual and combined effects of thrice weekly nandrolone decanoate treatment and intradialytic resistance training on body composition, strength, and physical function. The training protocol consisted of five lower-body resistance exercises using ankle weights 3 d⋅wk⁻¹. The intensity of the program was vigorous, starting with two sets of each exercise at 60% of each patient's three repetition max (3-RM) strength, progressing to three sets per session, with additional weight added as tolerated. An important finding from this study was that both the nandrolone and resistance training resulted in significant increases in the cross-sectional area (CSA) of the main muscle group being exercised — the quadriceps (measured by magnetic resonance imaging (MRI)), and the combined treatments had an additive effect. In addition, the resistance training, but not the nandrolone treatment, increased the leg muscle strength. Surprisingly, the resistance training did not improve several functional measures such as gait speed, stair climbing, or ability to rise from a chair. Furthermore, there was a paradoxical increase in whole-body fat mass (assessed by dual-energy x-ray absorptiometry) in the resistance training group, whereas lean mass did not change. In summary, this progressive resistance training program had some important benefits (increased quadriceps CSA and strength) but did not improve whole-body lean mass or several measures of physical function.

Another highly cited resistance training RCT in HD patients is the Progressive Exercise for Anabolism in Kidney Disease (PEAK) study (³⁹). The PEAK intervention was similar in many regards to the training protocol used in NEXT. It included 12 wk of progressive resistance training involving 10 different exercises (two sets of eight repetitions for each) that were performed 3 d⋅wk⁻¹ during dialysis. Ankle weights were used for leg exercises, whereas free weights (dumbbells) were used for upper-body exercises. Vascular access arm exercises were performed in the clinic lobby before starting dialysis. The primary outcome in PEAK was the change in thigh muscle CSA measured by computed tomography scan. Unfortunately, muscle CSA did not increase significantly in the resistance trained group (+1.3%, P = NS) compared with the control group (−0.7%, P = NS). The authors suggested that this may have been due to the fact that the CSA measurements captured muscles that were not targeted (hamstring), as well as some that were targeted (quadriceps), by the exercise intervention. Despite no improvement in muscle CSA, there was a significant improvement in muscle “quality,” as measured by a reduction in muscle lipid infiltration. Leg muscle strength was also improved, but physical function, as measured by the 6-min walk test, did not.

In a follow-up to the PEAK study (⁴⁴), the resistance training was extended to 24 wk, with the 12-wk usual care control group crossing over to an identical 12-wk resistance training program. In the 24-wk training group, there was a continued (nonsignificant) trend for an increase in thigh CSA from wk 12 to 24 (+1.8% vs baseline, P = NS), and muscle strength and exercise capacity also continued to improve. By contrast, thigh CSA continued to decline in the control/crossover group from wk 12 to 24 (−1.4% vs baseline, P = NS) despite undergoing resistance training during this period. The authors' primary conclusion from this study was that resistance training did not improve muscle CSA or quality (lipid content) at 24 wk. Despite this, the statistical trends for an increase in thigh muscle CSA in the 24-wk exercise group and the significant improvements in strength at 12 and 24 wk are somewhat encouraging. This appears to be yet another example of a reasonably robust resistance training intervention in HD patients that yielded mixed results that are difficult to interpret.

Kopple et al. (²³) also conducted an RCT to examine potential mechanisms by which exercise training improves exercise capacity in HD patients. In this study, patients were randomized into one of the following four groups: 1) endurance training alone, 2) resistance training alone, 3) endurance and resistance training, or 4) no training, with all exercise conducted immediately before dialysis (3 d⋅wk⁻¹). The endurance training consisted of cycling at a moderate intensity for up to 40 min per session. The resistance training was two sets of three leg exercises (extension, curl, and press) on machines at a resistance up to 80% of their 5-RM strength. The combined endurance and resistance exercise group performed approximately one half of each exercise protocol. Unfortunately, there were no changes in body composition (whole-body or regional lean and fat mass) in any of the exercise-trained groups. Despite this, the training did result in several beneficial changes in mRNA levels that should promote muscle anabolism. Most of these beneficial transcriptional changes were related to increases in the anabolic protein IGF-1 and reductions in the protein synthesis inhibitor myostatin. In summary, similar to the NEXT (⁴²) and PEAK (^39,44) studies, there were both positive and negative findings that make it difficult to evaluate the efficacy of exercise in HD patients based on these data.

Another resistance training study with similar, puzzling findings was recently published by Kirkman et al. (⁴³). In this pilot RCT, 23 HD patients were randomized to a control group or a group that underwent a progressive resistance training protocol for 12 wk. The intervention consisted of three sets of leg presses performed at high intensity during dialysis, using specially designed equipment placed at the end of the patients' chair. The primary findings in the study included an increase in thigh muscle volume, as assessed by MRI. Muscle strength also improved, but once again, physical function, as measured by the 6-min walk test, the 30-s sit-to-stand test, and the 8-ft timed up and go test, did not improve.

In summary, these data provide a lot of contradictory information that is hard to decipher. Although it appears fairly clear that resistance training consistently improves strength in the primary muscles that are trained, its impact on muscle hypertrophy has been mixed at best, and at least several of the resistance training RCTs with the largest sample sizes and highest-intensity training have not demonstrated improvements in various measures of physical function.

Combined Exercise and Nutrition-related Interventions

Data from the studies described previously suggest that HD patients may have an anabolic resistance to exercise. In the elderly, combining exercise with protein supplementation produces a potentiated anabolic response (⁴⁵). Likewise, early studies in HD patients showed that a single bout of either endurance or resistance exercise potentiates the anabolic response to nutritional supplementation (^46,47). However, data from longer-term interventions (3–12 months) in HD patients that have combined exercise training and nutrition supplementation have been less encouraging, as none have demonstrated additive benefits on strength, physical function, or body composition (^48–52). Details of these studies are reviewed, and some of the potential reasons for these modest results are described.

One of the first chronic interventions in HD patients to examine the combined effects of exercise and nutrition was published by Dong et al. (⁴⁰). In this study, patients were randomized to one of two groups: 1) oral nutritional supplementation (ONS) or 2) ONS and resistance exercise training. The ONS was a mixed macronutrient renal supplement (Nepro, Abbott Labs). Patients in the resistance training group performed 3 sets of 12 repetitions on a leg press machine at an intensity of 70% of their 1-RM strength, immediately before their dialysis session, 3 d⋅wk⁻¹. The primary finding from this study was that there were no significant differences between groups in terms of the change in body weight, total lean mass, leg lean mass, or leg strength at 6 months. When data from the two groups were combined, there was a significant increase in 1-RM leg strength; however, lean mass was still not significantly increased over baseline levels. Because the study did not include a control group that did not receive a nutritional supplement, it was not possible to discern if the nutritional supplement alone or the nutrition and resistance training preserved strength or body mass compared with a group receiving usual care.

Martin-Alemany et al. (⁵¹) recently published data from a pilot with a very similar protocol to the study described previously from Dong et al. In this study, HD patients were randomized to one of two groups for 12 wk: 1) ONS (Nepro) or 2) ONS and resistance training. The resistance training consisted of four intradialytic exercises at a light intensity, including three leg exercises and one upper-body exercise, all performed using relatively light resistance. After 12 wk, there were significant improvements in several markers of nutritional status, as well as an increase in handgrip strength. However, the resistance training did not potentiate the benefits of the ONS alone.

A recent study by Molsted et al. (⁴⁸) also had similar features. This was a 16-wk intervention in which HD patients were randomized into two groups: 1) strength training with protein supplementation (250 kcal) or 2) strength training with an isocaloric nonprotein supplement. The progressive resistance training protocol included three different leg exercises (curl, press, and extension) performed on a nondialysis day, 3 d⋅wk⁻¹. The primary finding from this study was that resistance training improved muscle strength, power, and physical function. However, as with most of the other resistance training RCTs, muscle hypertrophy was not seen. Moreover, the protein-supplemented group did not improve more than the nonprotein supplement group.

Although the intervention protocols in these three trials differed in some important ways, a surprising conclusion that can be made is that there is little evidence that combining resistance exercise and nutritional supplementation in HD provides additive benefits on muscle strength, function, or hypertrophy. This could be due to a number of factors, including alterations in muscle protein metabolism in HD patients that may inhibit the beneficial effects of either exercise or dietary protein supplementation (^53,54). However other factors that may contribute to these modest results also should be considered, particularly related to the nature of the exercise interventions. Similar to many of the endurance training studies discussed, the volume of exercise conducted in many resistance training studies in HD patients has been low. In the study by Dong et al., the training protocol included just three sets of 12 repetitions of leg presses, 3 d⋅wk⁻¹ (nine total sets of leg presses per week). Although the intensity of the training was relatively high, the volume was well below what is recommended in most physical activity guidelines (⁹). In the study by Martin-Alemany et al., the volume of exercise was a bit higher (three sets of four different exercises, twice per week), but the intensity (resistance) appeared to be rather low. By contrast, the volume and intensity of the resistance training protocol in the study by Molsted et al. appeared to be much higher (three exercises, thrice per week, at high intensity), yet muscle hypertrophy was still not achieved in the group receiving the combined resistance training and nutritional supplement. Taken together, these data suggest that combining resistance training and nutritional support may still be insufficient as a means for achieving robust improvements in muscle mass and function.

At least two prospective randomized trials have examined the combined effects of endurance exercise training and concomitant nutritional support (^50,52). In a pilot study by Hristea et al. (⁵⁰) (N = 21), HD patients with clinically defined PEW were randomly assigned to one of two groups: 1) nutritional support or 2) nutritional support and intradialytic cycling. Cycling was performed 3 d⋅wk⁻¹ for 30 min at a moderate intensity. The primary finding from the study was that combining intradialytic exercise and nutritional support was unable to reverse PEW, although some measures of physical function (6-min walk test) were improved.

Jeong et al. (⁵²) also recently published the largest RCT to date examining the combined effects of nutritional supplementation and aerobic exercise training in HD patients. In this study, 138 HD patients (age 58 ± 12 yr) were assigned for 12 months to 1) control (CON), 2) intradialytic protein (PRO), or 3) PRO and exercise (PRO and EX). PRO and PRO and EX consumed an oral protein supplement (30 g of whey) 3 d⋅wk⁻¹ during dialysis. PRO and EX also cycled for 30–45 min during treatment at a moderate intensity (rating of perceived exertion (RPE) 12–14). The primary outcome was change in physical function at 12 months, with secondary outcomes including changes in cardiovascular structure and function, muscle strength, and lean mass. In total, 101 patients completed the intervention. Despite the larger sample size, at 12 months, there were no significant differences between groups in the primary outcome related to physical function (the shuttle walk test). Although there were trends for improvements in some secondary measures of physical function (e.g., gait speed), muscle strength, or cardiovascular structure and function in PRO and PRO and EX at either 6 or 12 months, these generally did not reach statistical significance.

Although there is a clear theoretical rationale for combining exercise- and nutritional-related interventions in HD patients, the data from the trials conducted to date provide little evidence that doing so yields significant benefits. This suggests that a greater exercise stimulus or enhanced nutritional support may be needed to demonstrate the potentiated additive benefits of these lifestyle interventions that are often seen in healthier populations.

Why Is the Data on Exercise in HD Patients Not Stronger?

Many factors are likely to contribute to the modest benefits of exercise in HD patients. A primary factor not often discussed is that the exercise prescriptions used in many studies to date may have been below the threshold at which significant benefits may be realized. The majority of exercise trials in HD patients have used intradialytic training, which is often preferred because of the reduced patient burden and the ability to track compliance. Unfortunately, intradialytic exercise limits the intensity and types of exercise protocols to those that can be performed while confined to a chair (e.g., cycling and resistance training with bands or ankle weights). Of all the studies in the literature that have assessed the benefits of endurance exercise training (including those using intradialytic or out-of-center exercise), none that we are aware of has prescribed the 150 min of moderate activity per week that is the minimum recommended by most guidelines (⁵⁵). Indeed, the “typical” protocol in most endurance training studies has been ~30 min of cycling or walking at a low or moderate intensity 2 to 3 d⋅wk⁻¹. At best, that is about 40% below recommendations for aerobic activity for healthy populations (Table).

The resistance training literature in HD patients is similar. In healthy populations, full-body strength training protocols using combinations of free weights and/or resistance training machines above 60% of 1-RM strength is a preferred approach for improving muscle strength and mass (⁵⁶). Although this type of training protocol should be encouraged for all capable patients, it is impractical to expect it will be adopted by HD patients on a wide scale. Moreover, resistance training in HD has most often used lower-intensity training using resistance bands or ankle weights to complete lower-body exercises, whereas upper-body exercises are often restricted to the nonvascular access arm, if any upper-body exercise is done at all. It is clear that creative approaches will be needed to overcome these significant barriers to resistance training in HD patients.

Another factor to consider in the exercise literature in HD is that the length of the interventions may not have been sufficient to capture all of the potential benefits. A general hypothesis in most studies has been that exercise training would improve specific health outcomes. However, considering the multiple comorbid conditions and severe decline in physical function that many HD patients experience, a more appropriate goal may be to simply maintain patient's health and functioning, relative to a usual care control group. The length of the exercise intervention in most studies has been between 3 and 6 months, which may not have been sufficient to fully capture expected declines in the control group, if indeed a control group was even included. In many studies, trends for improvements in strength, function, and body composition were noted, and it is possible that statistically significant differences may have been seen had the interventions been extended for 9–12 months or longer. Although this would be ideal, dropout rates and low compliance with interventions are common in this population due primarily to high morbidity and mortality rates. There also is some evidence of a leveling off of benefits of exercise in some studies after approximately 3–6 months (^52,57). We speculate this may be due in part to a reduced enthusiasm for a mundane exercise prescription, such as thrice weekly intradialytic cycling, over a prolonged period. This suggests that conducting longer-term interventions in this population will be extremely challenging.

Another concerning explanation for this equivocal data is that the multiple comorbidities in HD patients may inhibit some of the beneficial effects of exercise. In the few resistance training studies that have prescribed an adequate volume of moderate to high-intensity progressive resistance training, muscle strength seems to improve, but muscle hypertrophy does not (^39,48). This suggests that HD patients may have an anabolic resistance to the effects of resistance training. Previous research has shown that HD patients have an anabolic resistance to protein supplementation, possibly due to an accelerated basal rate of muscle protein synthesis that creates a stimulatory ceiling effect and/or reduced amino acid availability (⁵⁴). These same factors could limit the anabolic response to resistance exercise. The chronic inflammation that is a hallmark of CKD promotes malnutrition and muscle wasting through a variety of mechanisms, including increased skeletal muscle protein catabolism, cytokine-mediated hypermetabolism, suppression of nutrient intake, and disruption of the growth hormone and IGF-1 axis, which may prevent skeletal muscle anabolism (⁵⁸). Furthermore, Chiang et al. (⁵⁹) recently demonstrated that low testosterone levels in HD patients are related to poor physical function, frailty, and muscle wasting, and low testosterone can also inhibit the anabolic response to resistance training (⁶⁰). Indeed, many HD patients are likely too frail to participate in activity of a sufficient volume or intensity to realize robust benefits from an exercise program. This suggests that bigger gains could be made by focusing exercise interventions in the nondialysis patients with CKD whose health has not deteriorated to such a significant extent. There may be similar reasons for the equivocal effects of exercise on cardiovascular function in HD patients (^20,32–34), as the vascular remodeling and calcification could limit the capacity for their vessels to respond to exercise. Chronic volume overload also promotes cardiovascular dysfunction and postdialysis fatigue (⁶¹), and so could inhibit both the capacity to exercise and the cardiovascular adaptations from it. In the absence of addressing these comorbid conditions, the efficacy of simplistic exercise interventions may be limited.

The modest benefits of exercise in some studies may also just be a manifestation of the difficulties of conducting large, long-term exercise training studies in this population. The numerous metabolic abnormalities and high prevalence of comorbidities in HD patients may significantly reduce compliance with exercise interventions, both in terms of how often they are willing to exercise and the intensity at which they are willing to work, as well as the outcomes used to evaluate the efficacy of interventions. For example, it is unlikely that patients with chronic volume overload or anemia will be able to either initiate or sustain a robust exercise program. Furthermore, changes in medication use, differences in dialysis dose and frequency, and a multitude of other treatment-related factors that are difficult to control may significantly impact the results of an exercise intervention. These confounding factors introduce variability in the data and make it unclear if standard outcome measures used to evaluate the efficacy of exercise interventions in most exercise trials are valid in HD patients. These factors make it extremely difficult to interpret results from these trials, but do not necessarily mean that a particular intervention is ineffective in the subset of patients that are able to comply with the prescription.

Research Needs

Many in the research community continue to argue that the evidence for the benefits of exercise is clear, and it is time for exercise programs to be included as a component of the standard of care for HD patients (^4,5). Despite this, few HD clinics support exercise programs or services (¹⁰), and physical activity levels in HD patients remain low (¹¹). This suggests that either the nephrology community is not convinced of the benefits of exercise or it is unclear how to implement effective and sustainable programs. An important first step for addressing this disconnect is for the research community to acknowledge the many deficiencies in the exercise literature contributing to this skepticism. By doing this, steps can be taken to improve exercise interventions, provide better guidance to clinics, and better serve our patients.

One of the most glaring weaknesses in the exercise literature in HD is the remarkably low intensity and volume of exercise that has been prescribed in most studies. As described previously, most endurance and resistance training interventions in HD, including studies from our own laboratory (^52,62), typically prescribe exercise of an intensity and volume that is significantly less than what is recommended by standard physical activity guidelines (Table). RCTs incorporating training protocols with greater volumes and intensities that we would expect to improve cardiovascular performance, elicit anabolic responses in skeletal muscle, and promote significant metabolic adaptations that inhibit the progression of comorbid disease are sorely needed. Of course, this is an incredible challenge given the high prevalence of frailty, low physical function, and other comorbidities in many HD patients. Furthermore, many of the studies to date have focused on the healthiest patients who are deemed more capable of moderate exercise and have therefore excluded a large portion of the dialysis population. The most functionally limited of dialysis patients may greatly benefit from exercise, so an emphasis should be placed on including less-healthy subsets of patients when possible. On the other hand, starting an exercise program when a patient reaches the need for dialysis might be too late, and it is reasonable to assume that bigger gains could be made by focusing exercise interventions in the nondialysis patients with CKD whose health has not deteriorated to such a significant extent.

Compounding this concern is the fact that many of the comorbidities in HD patients may attenuate the beneficial effects of exercise on factors such as skeletal muscle hypertrophy and cardiovascular health. It should not be surprising that we see limited benefits in studies where we underprescribe exercise to a population in which blunted results are expected. Because of these difficulties, we may need to accept that the best we can hope for is that exercise attenuates declines in the health of our patients, as opposed to providing positive adaptations. To test this, we will need to conduct RCTs with longer intervention periods. Most of the exercise literature include interventions with relatively short intervention periods (<6 months), but RCTs of 12 months or more may be needed to see declines in health-related parameters in nonexercising control groups. If these declines are attenuated in the exercise training groups, this would provide needed incentives to promote exercise programs to maintain patient's health.

Finally, outcome studies exploring the effects of exercise training on morbidity and mortality are sorely needed. This is of particular interest because self-reported physical activity has been strongly associated with hospitalizations and mortality (¹¹). Outcome studies should be a top priority among researchers because of the implications that positive findings would have on promoting exercise as an important component of patient's standard care. Collecting data on hospitalizations and other important treatment-related factors (e.g., dialysis compliance, changes in medication usage) will also be important for evaluating the cost-effectiveness of these programs and providing additional incentives for promoting exercise in this population.

Moving Toward a New Approach for Exercise Interventions in HD Patients

The data reviewed here suggest a grim reality: implementing exercise programs that produce robust benefits in HD patients will be a daunting challenge. However, there also are plenty of reasons for optimism. Subsets of patients in exercise trials have clearly benefitted from incorporating more physical activity into their lives, even if much of the evidence for this is anecdotal. So, it is apparent that the equivocal findings seen in many studies are likely a manifestation of the difficulties of conducting large, long-term exercise training studies in HD patients, rather than evidence that exercise is ineffective. Yet, an important question remains: “How can we make exercise interventions more effective for a larger percentage of patients?”

To answer this, it will first be critically important to include HD patients, their family, and other caregivers at all stages in the design and implementation of exercise interventions. Previous research has clearly shown that the priorities of patients often differ significantly from the priorities of the medical and research communities (⁶³). As such, significant efforts should be made to consider the unique perspective of patients, as exemplified by moving commentaries on this topic (^64,65).

A significant problem to date is that almost all HD exercise trials have relied on a “one-size-fits-all” prescription like intradialytic cycling, which may have limited benefits (¹⁷). Sticking a bike in front of patients and asking them to pedal for 20–30 min during their thrice weekly dialysis sessions is not an ideal exercise prescription for many reasons, and we must change this approach. Mandated prescriptions such as this are boring for most patients, and although they may provide modest benefits in the short term, they are often unsustainable. Most of the resistance training and walking programs conducted in HD patients have the same problems. There are a few wonderful examples where programs with simple, mandated exercise prescriptions have apparently provided robust and sustained benefits (⁶⁶), but these are clearly exceptions to the rule (¹⁰).

There are better options. One alternative to these mundane exercise prescriptions is to provide patients with more autonomy to select types of activities in which they choose to participate. An example of this approach, called the “Life Readiness Program,” was piloted by Tawney et al. (⁶⁷). The activity prescription in this trial focused on accumulating 30 min⋅d⁻¹ of lifestyle-type physical activity (e.g., housework, gardening, using walking as transportation). A liberalized activity prescription such as this may result in greater participation and more sustained changes in patient's lifestyles. Unfortunately, higher-intensity exercises were not a significant part of the Life Readiness Program, and it will be important to include these types of activities to maximize benefits in patients willing and able to do them. It is clear that an exercise specialist will be needed to design an individually tailored, progressive exercise training program for each patient based on his or her preferred activities. This approach would help ensure that patients recognize the importance of different types and intensities of activities that are needed to maximize the benefits they receive from exercising. One concern with this approach is that it will be hard to quantify benefits in a clinical trial. This is because individualized training programs will target different outcomes, as per patient preferences, thus diluting the overall impact on any specific outcome. For example, 1) patients targeting weight loss to become eligible for a transplant may prefer and benefit the most from an aerobic training program; 2) those with signs of muscle wasting may prefer a resistance training–focused program designed to increase muscle mass; and 3) frail patients may desire a combined aerobic and resistance training program targeting improvements in physical function, strength, bone health, or cardiovascular function. This is a concern from a research perspective because it will be difficult to test the effect of the intervention on specific outcomes (e.g., physical function, strength, or muscle mass) since each individual will be doing unique activities that may have different benefits. However, it seems likely that an individualized approach has a better chance of improving patient's QOL, or even “hard” outcomes such as cardiovascular events or mortality, so these should be targets in clinical trials aimed at assessing the efficacy of this approach.

There also is a need to transition away from using exercise as a stand-alone intervention. Instead, exercise should be included as a component of a multifactorial lifestyle modification strategy. The complex comorbidities associated with renal failure mandate comprehensive solutions. There are many examples of such programs in the exercise literature. For example, the Diabetes Prevention Program included an intensive lifestyle intervention that involved weight loss through a low-fat diet and engaging in moderate-intensity physical activity (e.g., brisk walking) for at least 150 min per week. The intervention included a 16-lesson curriculum covering diet, exercise, and behavior modification to help the participants achieve these goals (⁶⁸). This comprehensive strategy reduced the incidence of diabetes in patients with insulin resistance and is now lauded as an example of the importance of comprehensive lifestyle modification to prevent disease progression. The Italian Diabetes and Exercise Study (IDES) (⁶⁹) is another example from the diabetes literature that included a comprehensive behavioral intervention including individual counseling sessions with both diabetes and exercise specialists as part of a comprehensive behavioral intervention to reduce sedentary time and increase physical activity. Unfortunately, we are not aware of good examples of multifactorial lifestyle intervention strategies such as these in the HD literature.

In designing similar programs for HD patients, it will be important to consider the significant behavioral, personal, and environmental barriers that contribute to their poor compliance with exercise and other healthy behaviors (^12,13). In the Figure, we have presented a framework for a comprehensive lifestyle intervention strategy that may help address some of these challenges. In brief, this strategy suggests moving away from mandating simplistic exercise prescriptions that fail to address the wants or needs of such a diverse patient population. Instead, we should provide patients guidance on how to incorporate physical activity into their lives by any means possible, including low-intensity “lifestyle” activities, intradialytic exercises, and progression to higher-intensity activities and recreational sports for patients who are willing and able to participate. Ideally, an exercise specialist (e.g., kinesiologist, physiotherapist) would serve as a “lifestyle coach” that works directly with patients during their dialysis sessions to design a personalized physical activity prescription. The exercise specialist would also facilitate support from the dialysis clinic staff and the patient's family or other caregivers, possibly by counseling them on how to incorporate more activity into their own lives and/or how to help motivate or exercise with the patients. Additional nutritional support may be needed to address problems such as malnutrition and chronic volume overload that may be inhibiting a patient's ability to exercise. Implementing such a program will require a culture change in which physical activity and wellness are prioritized in HD clinics. Although this is a significant challenge, there is a model of this commitment to behavior change from HD clinics in Tassin, France and Izmir, Turkey. In these regions, blood pressure has been normalized in 90%–95% of HD patients using a comprehensive volume control strategy that includes an intensive, committed effort to reducing dietary sodium intake (⁷⁰). This is a behavioral modification with barriers every bit as difficult to address as physical inactivity.

CONCLUSIONS

In summary, although many consider the benefits of exercise in HD patients to be clear, a detailed examination of evidence from important clinical trials indicates that the data may not be as robust as often stated. A significant concern in the literature is that a one-size-fits-all approach to exercise prescription may be inadequate to address the myriad of health concerns in HD patients. The days of sticking bikes in front of HD patients and calling it an exercise prescription needs to end. Instead, a new strategy is needed to move our field toward using more personalized approaches that include physical activity as a component of a comprehensive behavior change intervention strategy. In the absence of this shift, we may continue to see only modest benefits from exercise-related interventions and waste opportunities to optimize the health and QOL of our patients.