Walk Away to Lower Hospitalization?

#NephJC chat

Tuesday December 20th at 9 pm Eastern

Wednesday Dec 21 at 8 pm GMT, 12 noon Pacific

A bear, however hard he tries, grows tubby without exercise.
— A.A. Milne, Winnie-the-Pooh

Exercise in Patients on Dialysis: A Multicenter, Randomized Clinical Trial.

Manfredini F, Mallamaci F, D'Arrigo G, Baggetta R, Bolignano D, Torino C, Lamberti N, Bertoli S, Ciurlino D, Rocca-Rey L, Barillà A, Battaglia Y, Rapanà RM, Zuccalà A, Bonanno G, Fatuzzo P, Rapisarda F, Rastelli S, Fabrizi F, Messa P, De Paola L, Lombardi L, Cupisti A, Fuiano G, Lucisano G, Summaria C, Felisatti M, Pozzato E, Malagoni AM, Castellino P, Aucella F, ElHafeez SA, Provenzano PF, Tripepi G, Catizone L, Zoccali C

J Am Soc Nephrol. 2016 Dec 1. pii: ASN.2016030378.

PMID: 27909047 DOI: 10.1681/ASN.2016030378   (free text at that link courtesy JASN)


Supplementary data


There is no doubt about it, dialysis patients have it tough. Hemodialysis patients spend the best part of three days a week in a hospital/satellite dialysis unit receiving life-sustaining treatment and the symptom burden related to dialysis is high, with many patients experiencing high levels of fatigue, disordered sleep, pain and low mood. It is unsurprising, therefore, that physical activity levels among dialysis patients are low, with <50% dialysis patients reporting exercising once a week, and higher mortality rates in such patients. Current guidelines suggest we should be giving lifestyle advise to dialysis patients about increasing levels of activity and engaging in regular exercise – albeit the evidence to support these recommendations is graded as ‘weak’. In the general population, lifestyle changes that result in increased physical exercise lower mortality, and there are a number of studies that have investigated the benefits of exercise in dialysis. For hemodialysis patients these studies can be largely divided up in the following ways:

1.     Intradiaytic – Taking place whilst the patient is on dialysis, for instance cycling or resistance training. This has the advantage of having superior adherence rates to interdialytic programmes, but with inferior cardio-pulmonary adaptations.

2.     Interdialytic – Taking place between dialysis sessions, where greatest gains in VO2 Max have been shown, but with, historically, higher dropout rates.

3.     Resistance – Muscle strengthening exercise, using weights/resistance bands, often with flexibility and core stability exercises. These have often focused on adaptations in particular muscle groups.

4.     Aerobic – Cardio-pulmonary exercise to improve aerobic threshold and maximal oxygen usage

 When reviewing an exercise study it is also important to note whether the exercise intervention was structured (that is planned, with performance testing and plans for progression), or unstructured and whether it was delivered by a trained individual (e.g. exercise physiologist or physical therapist), or whether there were simply instructions and a plan for the patient to follow themselves. All these factors need to be considered because there have been enormous problems implementing the results from previous studies into clinical practice. There are undoubtedly patient, physician, healthcare worker and institution related barriers to these that need consideration. It is worth bearing these things in mind when you read the article by Manfredini et al and consider how it can be that something as simple as asking your patient to walk can seem like such a challenge.


This was a randomized controlled trial in dialysis patients comparing the effects of a home-based walking program (exercise group) vs standard care on functional capacity and quality of life (they describe these as the two primary outcomes…..we will come back to this…..). To ensure the groups were matched for important medical co-morbidity, randomization was stratified by NYHA class. Patients in both groups underwent baseline and 6 month study assessments including the 6-minute walking test (total distance covered in 6 minutes), the sit-to-stand-5 (how long it takes to stand and sit from a chair 5 times) and quality of life questionnaires including the kidney disease quality of life, short form, version 1.3 (KDQOL-SF). Outcome measures were carried out on non-dialysis days. Exclusion criteria were a little vague, but included severe angina, stage 4 NYHA or amputation and they excluded patients able to walk >550 metres on the 6-minute walk test as they were deemed as having a high degree of fitness.

A walking intervention sounds simple enough, but was a little more prescriptive than just ‘going for a walk’ and was tailored, in part, to individual patients (full details were included in supplementary data). Essentially patients were asked to walk for 10 minutes twice a day (broken up with rest for those with poorer baseline performance), three times a week on their non-dialysis days. The speed and cadence of their walking was determined by their baseline performance at the 6-minute walk test and they wore a metronome to ensure they walked at the correct pace and cadence (they included an example YouTube clip). After 15 weeks the intensity of the walking was increased according to a pre-determined protocol. Patients recorded their adherence, symptoms and comments in a diary and the battery drain on the metronome was assessed as a surrogate for intervention fidelity (sneaky!).

Importantly it was the dialysis staff who managed the intervention delivery and whilst the ‘rehabilitation team’ were available to discuss, advise and help as needed, they did not deliver the intervention or manage the progression of training.


The CONSORT diagram shows that 41.5% of the available population were enrolled and randomized in the study (n=296). 47 out of the 151 patients randomized to the exercise intervention dropped out; 5 due to transplantation and 2 due to death. This was more than twice the number who dropped out of the control arm.

Figure 1: CONSORT diagram.

The majority of patients were hemodialysis patients, but there were a small and equal number of CAPD patients in the exercise and control arms. The patient groups were well matched, with a non-significant trend to higher BP’s in the exercise arm.

The 104 patients who completed the study in the exercise arm, managed an average of 83% of the prescribed exercise (mean 119 out of 144 sessions). 49 of the 104 patients were classed as having low adherence to the exercise regimen.

There were no differences in changes in clinical or dialysis-related parameters between the two groups. There were significant improvements in both the 6 minute walk test and the sit-to-stand 5 in the exercise group compared to the control group on both intention to treat analysis, and when they analyzed only patients who completed the trial.

Figure 2: Effect of study intervention on 6 minute walk test and sit-to-stand-5 compared to control

The authors also described a dose response to improvement in physical performance tests from exercise, based on the level of adherence to the exercise regimen:

Figure 3: Dose response to levels of exercise undertaken.

There were no major changes in quality of life scores for patients, with only two components of the KDQOL-SF changing significantly: Both cognition and social interaction scores were significantly worse in the control group at the end of the study than the start, with no changes in the exercise group between baseline and study completion.

They undertook Kaplan-Meier analysis for hospitalizations and intention-to-treat analysis revealed no significant differences between rates of hospitalization between the groups. When they included patients that completed the study, this became significant:

Figure 4

Figure 4


It is great to be discussing an exercise study, as an intervention like this is low cost and potentially of great benefit to patients well-being. If previous chats are anything to go by, people will raise points I didn’t even consider, so here are just a few things to consider that are worth a thought or two for the upcoming discussions:

  • I will be interested to see what people make of the design of this study. Surely it is a contradiction in terms to have more than one primary outcome measure? The power calculation for sample size in this case was based on the 6-minute walk test. They calculated that to detect a 40 metre difference in walking distance between exercise and control groups with 80% power they needed 180 patients. That is fine. What I don’t think is fine is to then say you have two other ‘primary outcome measures’ and give no justification of the sample required to confidently detect a significant change. I expect significantly more patients would be required to power this study with a questionnaire as the primary outcome measure.
  • Commensurate with previous studies, the dropout rates from the intervention arm are very high. 30% dropout is a lot and if it were a drug we would be questioning those compliance rates. Moreover, 50% of those who completed the study in the exercise arm were found were considered to have a low adherence to the protocol. It is surprising that with such a small amount of exercise actually undertaken that the results are still positive. This is not really mentioned by the authors in the discussion and there is little in the discussion to acknowledge the additional work that needs to be undertaken to understand the motivators and barriers to exercise for patients and staff, without which these programs will never make it into clinical practice.
  • The exercise intervention is relatively light (the AHA recommends 150 minutes moderate or 75 minutes vigorous activity per week) and I would encourage people to think about whether they think this is indeed an ‘exercise’ intervention or whether this is actually an intervention of increased physical activity. This might sound like semantics, but there are differences in the definitions and the physiological responses you can expect to both of these.
  • Whilst the majority of patients were hemodialysis patients, there were an important number of CAPD patients – no mention of when/how they undertook their exercise is mentioned, though the authors do state there was no significant difference between response to the intervention between HD and PD patients.
  • Whilst it is nice to have a positive trial in nephrology, positive trials are only useful if we can adopt the practices. I will be interested to see:
  1. What people think of the clinical relevance of the outcome measures described
  2. How easy it would be to implement this protocol in a run-of-the-mill dialysis unit, and who would implement it?
  3. I commend the authors for their continued work in this important area. One final thought from me: The findings here support what everybody will ‘know’ – if you are very sedentary, doing more activity improves your functional capacity. Knowing a lifestyle change is good for you is one thing, the question we really need to ask is how can we get patients more physically active? Identifying a way to research this question should be a priority for clinicians and research who believe in exercise and physical activity as medicine.

Summary by Matt Graham-Brown, @DrMattGB

Disclosure – I am working on the randomized controlled trial CYCLE-HD (@CYCLEHD15)


#NephJC chat

Tuesday December 20th at 9 pm Eastern

Wednesday Dec 21 at 8 pm GMT, 12 noon Pacific