Feasibility and effects of intra-dialytic low-frequency electrical muscle stimulation and cycle training: A pilot randomized controlled trial

Background and objectives Exercise capacity is reduced in chronic kidney failure (CKF). Intra-dialytic cycling is beneficial, but comorbidity and fatigue can prevent this type of training. Low–frequency electrical muscle stimulation (LF-EMS) of the quadriceps and hamstrings elicits a cardiovascular training stimulus and may be a suitable alternative. The main objectives of this trial were to assess the feasibility and efficacy of intra-dialytic LF-EMS vs. cycling Design, setting, participants, and measurements Assessor blind, parallel group, randomized controlled pilot study with sixty-four stable patients on maintenance hemodialysis. Participants were randomized to 10 weeks of 1) intra-dialytic cycling, 2) intra-dialytic LF-EMS, or 3) non-exercise control. Exercise was performed for up to one hour three times per week. Cycling workload was set at 40–60% oxygen uptake (VO2) reserve, and LF-EMS at maximum tolerable intensity. The control group did not complete any intra-dialytic exercise. Feasibility of intra-dialytic LF-EMS and cycling was the primary outcome, assessed by monitoring recruitment, retention and tolerability. At baseline and 10 weeks, secondary outcomes including cardio-respiratory reserve, muscle strength, and cardio-arterial structure and function were assessed. Results Fifty-one (of 64 randomized) participants completed the study (LF-EMS = 17 [77%], cycling = 16 [80%], control = 18 [82%]). Intra-dialytic LF-EMS and cycling were feasible and well tolerated (9% and 5% intolerance respectively, P = 0.9). At 10-weeks, cardio-respiratory reserve (VO2 peak) (Difference vs. control: LF-EMS +2.0 [95% CI, 0.3 to 3.7] ml.kg-1.min-1, P = 0.02, and cycling +3.0 [95% CI, 1.2 to 4.7] ml.kg-1.min-1, P = 0.001) and leg strength (Difference vs. control: LF-EMS, +94 [95% CI, 35.6 to 152.3] N, P = 0.002 and cycling, +65.1 [95% CI, 6.4 to 123.8] N, P = 0.002) were improved. Arterial structure and function were unaffected. Conclusions Ten weeks of intra-dialytic LF-EMS or cycling improved cardio-respiratory reserve and muscular strength. For patients who are unable or unwilling to cycle during dialysis, LF-EMS is a feasible alternative.

Version 1 -07. 11.2013 have CHF 30 . In addition to fibrotic LVH and CHF, abnormal electrolyte concentrations, arrhythmia and ischemic coronary artery disease lead to an increase in sudden cardiac death 33 , although myocardial infarction accounts for relatively few events 34 . Further pathological mechanisms of increased CVD risk in CKD are extensive, including dyslipidaemia, systemic inflammation, decreased bioavailability of nitric oxide, increased RAS and sympathetic activity, impaired endothelial function, vitamin D deficiency, hyperparathyroidism and altered mineral metabolism 9,28,35 . Collectively, the complex interaction between the kidneys and the heart is known as the cardio-renal syndrome 36 with the destructive effects of CHF being the predominant feature.

Arterial remodelling in chronic kidney disease
The impact of arteriosclerosis is significant in CKD. Thickening of the medial arterial layer, due to increased collagen content, hyperplasia and hypertrophy of the vascular smooth muscle cells, combines with concentric calcification to increase arterial stiffness 37 . A strong association between arterial stiffness and mortality has been reported in ESRD 38 . Loss of arterial distensibility leads to increased risk of CHF, arrhythmia and stroke from microvascular damage caused by the higher systolic pressures required to maintain arterial hemodynamics 39 . Simultaneously, an early return of the reflected forward pressure wave to the central circulation causes a decreased diastolic pressure, thereby compromising myocardial perfusion via the coronary arteries 40 . Ultimately, the higher LV afterload resulting from aortic stiffness leads to increased LV mass 41 , LV stiffness 42 and LV filling pressures 43 , increasing myocardial O2 consumption and energy expenditure. Cardiovascular reserve is reduced and hemodynamic stability compromised, exposing the myocardium to further LV structural maladaptation and functional impairment 44 .
Endothelial dysfunction also occurs early in the arteriosclerotic process and is integral to the progression of arterial stiffness 45 . As a measure of endothelial function, flow-mediated vasodilation of the brachial artery is inversely related to arterial stiffness in hypertensive patients 46 . Impaired endothelial function has been demonstrated in the large and small arteries of CKD patients 46,47 , relating strongly to kidney and CVD outcomes 48,49 . Chronic kidney disease related low grade inflammation, oxidative stress and hormonal factors appear to reduce the bioavailability of nitric oxide (NO) 45 , a key protagonist of endothelial derived vasodilation. Arterial remodelling as a whole, involving increased stiffness and impaired flow-mediated vasodilation, is a key cardiovascular feature of CKD. In combination with cardiac maladaptation and skeletal muscle abnormalities this contributes to the significantly impaired exercise capacity observed in the CKD population.

Skeletal muscle dysfunction in chronic kidney disease
A further piece of the pathophysiological jigsaw dictating morbidity and mortality in CKD is skeletal muscle wasting 50 . The key driver of this process is an inflammatory cytokine and inactivity mediated imbalance in protein homeostasis, resulting in a catabolic destruction of structural and functional proteins with debilitating muscle wasting 51 . As kidney function declines and haemodialysis is initiated, poor nutritional status is mediated by anorexia, increased resting energy expenditure and a disproportionate increase in protein catabolism as a result of activation of the inflammatory cascade and proteolytic pathways 52 . Hypercatabolic protein energy wasting is a complex biomolecular process and the understanding of the phenomenon continues to evolve. Medical and pharmacological therapies to combat this state remain ineffective and it is clear that exercise training may be the most promising therapy 51 .

Exercise training in chronic kidney disease
When conducted between dialysis sessions (inter-dialytic), dynamic CV exercise and strength training can improve catabolic destruction of proteins, muscular fitness, resting systolic and diastolic blood pressure Version 1 -07.11.2013 and QOL 12 . The most effective programmes appear to be supervised and longer than 4 months in duration, adopting a mixed CV and strength training approach at intensities >60% VO2 reserve 12 . However, despite the high incidence of CHF in CKD, only three studies have examined the effect of exercise training on the heart 53-55 , all of which reported no change in standard structural measures. Compliance and adherence to an inter-dialytic exercise regimen can be challenging for chronically fatigued patients 12 . Limited recent work has confirmed that exercise performed during dialysis (intradialytic) is safe and indicates favourable outcomes in functional capacity, muscular strength, physical activity participation, arterial compliance and biochemical markers (serum phosphate and potassium) [56][57][58][59][60] . Intradialytic exercise may therefore offer an effective alternative as appointment attendance is essential, supervision is readily available and time on dialysis is used effectively. To date, benefits gained from this approach have not been maximised as trials have been characterised by poorly designed and controlled exercise interventions with insufficient prior assessment of exercise capacity and inaccurate regulation of exercise intensity.

Low frequency electrical muscle stimulation (LF-EMS)
With chronic and progressive moypathic changes in cardiac and skeletal muscle, it is common for dynamic exercise to be unachievable in patients with ESRD 7 . In these patients, alternative exercise stimuli should be considered. Low frequency electrical muscle stimulation (LF-EMS) of the leg muscles is well tolerated and has been shown to produce similar benefits to aerobic training, increasing exercise capacity, reducing dyspnoea and improving QOL in chronic heart failure 61-63 , chronic obstructive pulmonary disease 64 and other conditions. Neuromuscular electrical stimulation of this origin elicits rhythmical, subtetanic, isometric contractions of the quadriceps and hamstrings, mimicking the pattern of muscle activation associated with shivering 65 . Energy expenditure is increased leading to improvements in exercise capacity [66][67][68] . Intra-dialytic LF-EMS may offer a suitable alternative to dynamic intra-dialytic exercise. Indeed, comparable benefit has been demonstrated in the only previous study examining LF-EMS during dialysis 69 . Muscular strength, six minute walk distance, QOL and urea clearance were shown to be equally improved in patients completing 20 weeks of intra-dialytic dynamic exercise or LF-EMS. Importantly, LF-EMS was well tolerated and there were no adverse events throughout the duration of the study.

Summary
In a low functioning, inactive patient group, at increased risk of arterial and ventricular remodelling and skeletal muscle and other protein wasting, the development of safe and effective exercise strategies is essential to promote functional independence, improve QOL and potentially reduce mortality. An understanding of the CV and skeletal muscle adaptation to these interventions is required to quantify their impact and to inform effective exercise intervention design in CKD patients on dialysis. Through the use of validated techniques (echocardiography, cardiopulmonary exercise testing, leg dynamometry, physical activity monitoring, arterial applanation tonometry, flow mediated vascular dilatation, biochemistry, QOL questionnaire), comprehensive evaluation of the effects of intra-dialytic exercise training will likely provide insight into the most appropriate future strategies for the management of patients with ESRD, maximising the impact of these therapies in the context of the limited availability of effective alternative treatments. Version 1 -07.11.2013

Hypotheses:
1.Intra-dialytic exercise training is well tolerated and improves well-being in stable, chronic dialysis patients. 2.Intra-dialytic exercise training significantly improves cardiovascular and skeletal muscle function with reversal of the catabolic process. 3.The effect of intra-dialytic LF-EMS exercise on cardiovascular and skeletal muscle function and its tolerability is comparable to dynamic exercise training effects.
Aims are: 1.To assess the tolerability of intra-dialytic dynamic exercise and LF-EMS interventions compared to conventional treatment without exercise intervention in patients on regular dialysis for end stage renal failure. 2.To assess the improvement of quality of life with intra-dialytic dynamic exercise and LF-EMS interventions compared to conventional treatment without exercise intervention in patients on regular dialysis for end stage renal failure. 3.To evaluate the effects of intra-dialytic dynamic exercise training compared to conventional treatment without exercise intervention on measures of functional cardiovascular and skeletal muscle function. 4.To evaluate the effects of LF-EMS intra-dialytic exercise training compared to conventional treatment without exercise intervention on measures of functional cardiovascular and skeletal muscle function. 5.To qualitatively compare the effects of LF-EMS with dynamic intra-dialytic exercise training on measures of functional cardiovascular and skeletal muscle function.

Participants:
Patients on chronic and stable haemodialysis, under the renal care of Nephrologists at the University Hospital Coventry and Warwickshire NHS Trust will be invited to take part and will be randomised after informed consent is received. 1.Clinically significant valvular insufficiency 2.Clinically significant dysrythmia 3.Uncontrolled blood pressure: systolic > 160, diastolic >95 during the months before enrolment 4.Excessive fluid accumulation between dialysis sessions (>3 liters), more than twice pulmonary edema over 3 months before enrolment deemed to be due to excess fluid intake 5.Haemoglobin unstable and below 9.0 6.Ischemic cardiac event or intervention in the last 3 months 7.Morbidly obese, mid-thigh circumference of more than 60cm (EMS straps limit) 8.Clinically significant, still active inflammatory or malignant process 9.Pacemaker or cardiac device (contraindicated for bioelectrical impedance) 10.Patient highly physically active on their own accord. 11.Planned kidney transplant during study period.

Study design
This study is an open labeled, randomised controlled trial. Patients will be pre-screened using the clinical information available on their dialysis documentation. Eligible patients will be approached, patient information provided and then informed consent obtained. After randomisation patients will be allocated to one of the three study groups outlined in figure 1 below, with 12 weeks of dynamic exercise, LF-EMS or usual care.

Week 3
Week 12  Study flow: After patients have been approached and consented, the duration of their participation will be 20-24 weeks, undergoing 20 weeks study specific procedures.

Consenting (week -3 to-1):
Eligible patients will be approached by the research nurse or the investigator and information provided. At the second visit, while on dialysis, the patients will be consented and all the potential procedures explained again. This will include a video explaining the two modalities of exercise, dynamic and LF-EMS.

Baseline visit (week -1):
This will coincide with a dialysis day and include obtaining clinical and QOL data, medication list, and a 'Sensewear' physical activity monitor armband will be fitted. 'Senswear' will allow validated measurement of free living physical activity over a 7 day period 70,71 .

Randomisation visit (week 0):
This visit will occur on a non-dialysis day, during a short inter-dialytic interval and at least 12 hours after the last dialysis session. The visit will include an echocardiogram, bioelectrical impedance, CPEX with lung function test, applanation tonometry, flow mediated vascular dilatation, measurement of muscle strength and blood sampling for routine tests and for future testing of novel serum markers. Outcomes assessors will be blinded to the group allocation of the patient.
Participants will be randomised to one of 3 groups: 1) the intra-dialytic LF-EMS group, 2) the dynamic intra-dialytic exercise group or 3) the usual care without any dynamic exercise or LF-EMS group.

Treatmentfamiliarisation period (week 0-3):
Exercise under close instruction and supervision will be commenced in patients randomised to either of the exercise groups. Over the initial 3 weeks, LF-EMS and dynamic exercise duration will be gradually increased to 30 mins per dialysis session. Previous work has indicated that 3 weeks of regular LF-EMS exposure is required to achieve an intensity that will elicit a cardiovascular training stimulus 72 (i.e. 40-60% VO2 reserve 73 ). LF-EMS intensity will be gradually increased, as tolerated, with the aim of achieving the appropriate level by week 3. Dynamic exercise work load will be gradually increased to achieve a heart rate and workload equivalent to that achieved at 40-60% VO2 reserve during baseline CPEX. As the use of heart rate during dialysis may not always provide an accurate reflection of workload 73 , rating of perceived exertion (RPE) will also be used to monitor exercise intensity in both groups 72 .

Graded LF-EMS test (week 3):
to ensure that LF-EMS tolerance has improved to a level that elicits a workload of 40-60% VO2 reserve, a graded LF-EMS test will be performed in this group following the 3 week familiarisation period. This will be performed during a dialysis session. Stimulation intensity will be incrementally increased every 5 minutes until the tolerable level is achieved 72 and oxygen consumption measured with the respiratory gas analysis system.

Treatment visit (weeks 3-12):
Patients in the exercise groups will progress from 30-45 mins of exercise at an intensity of 40-60% VO2 reserve as regulated by heart rate and RPE. Where this isn't possible, maximal tolerated exercise intensity will be continued.

Treatment completion visit (week 13):
This visit will occur on a non-dialysis day, during a short interdialytic interval and at least 12 hours after last dialysis session. The visit will include recording of any clinical events, QOL assessment, free living with 'Senswear' for the following 7 days, echocardiogram, bioelectrical impedance, CPEX with lung function test, applanation tonometry, flow mediated vascular dilatation, measurement of muscle strength and blood sampling for routine tests and for future testing of novel serum markers. Outcomes assessors will be blinded to the group allocation of the patient. Version 1 -07.11.2013

Wash out and study completion visit (week 20):
This part of the study will only be performed, if further funding can be obtained. This visit will occur on a non-dialysis day, during a short inter-dialytic interval and at least 12 hours after last dialysis session. The visit will include recording of any clinical events, QOL assessment and free living physical activity with 'Senswear' for the following 7 days, echocardiogram, bioelectrical impedance, CPEX with lung function test, applanation tonometry, flow mediated vascular dilatation, measurement of muscle strength and blood sampling for routine tests and for future testing of novel serum markers. During the 8 week 'washout' period, patients will undertake usual care which will not involve dynamic exercise training or LF-EMS.
Randomisation: Stratified permuted block randomisation with a variable block size (sizes 3 and 6 used) will be employed. Stratification variables are age (young, middle-aged, old) and gender (male, female), and a separate permuted block randomisation will be performed within each of the six strata. The randomisation list will be created and held by a Warwick Medical School statistician who the study team are going to contact to obtain the group allocation of a recruited patient. The randomisation list will not be made available to the study team.

Sample size calculation:
Since this is a pilot study, no formal sample size calculation was conducted. The sample size was chosen based on Kieser and Wassmer (1996), who recommended recruiting 20 to 25 patients per group for a pilot study (with a normally distributed outcome) where the aim is to minimise the overall sample size of pilot and large scale RCT. Being conservative, a drop-out rate of 20% is assumed and thus a sample size of 30 patients per group has been chosen.

Low frequency electrical stimulation (LF-EMS) (treatment group):
Bilateral neoprene straps will be applied to the quadriceps and hamstrings 3 times per week during the second hour of dialysis. Rhythmical contractions will occur at a frequency of 3-4 Hz. Participants will be introduced to the EMS protocol very gradually to ensure tolerance and adherence. On the first occasion the straps will be activated for 5 minutes at a very light current amplitude (milliamps, mA) (less than visible contraction), so that participants can become accustomed to the unusual sensation. During the following 8 familiarisation sessions (3 weeks), duration will be incrementally increased to 30 mins and intensity increased as tolerated. By the end of the third week the intention is that the patient will be accustomed to using the straps for 30 mins at an intensity equivalent to 40-60% VO2 reserve 72 . This will be assessed with an intradialytic graded LF-EMS CPEX (as above). During LF-EMS sessions, exercise intensity will be regulated using a combination of heart rate and rating of perceived exertion (RPE). Intensity will be further increased over the 9 week period of the intervention to the maximum tolerated level. This is necessary to ensure that, whilst cardiovascular reserve and skeletal muscle function improve, intensity remains at a level that will elicit a cardiovascular exercise response. Duration will also be increased to 45 mins by the second week of the intervention and will remain at this level throughout. Each session will be preceded by a 10 minute 'ramp-up' period during which intensity will be gradually increased to the training level 72 . Each session will be concluded with a 5 minute cool-down at a gradually reducing intensity. The LF-EMS units will record the intensity and duration of the stimulation which can be downloaded for analysis.

Dynamic exercise (exercise control group):
Dynamic exercise will be performed 3 times per week during the second hour of dialysis using a recumbent exercise bike, or alternatively a leg ergometer whilst seated in a dialysis chair. On the first occasion, patients will exercise for 10-15 mins at an intensity with which they feel comfortable. During the following 8 familiarisation sessions (3 weeks) exercise duration will be Version 1 -07.11.2013 gradually increased to 30 mins. Intensity will be simultaneously increased until patients are capable of achieving 40-60% VO2 reserve, as determined from CPEX. Intensity will be regulated and monitored using a combination of heart rate and rating of perceived exertion (RPE). Intensity will be periodically increased, as tolerated, over the 9 week period of the intervention. This is necessary to ensure that, whilst cardiovascular reserve increases and skeletal muscle function improves, intensity remains at a level that will elicit a cardiovascular exercise response. Duration will also be increased to 45 mins by the second week of the intervention and will remain at this level throughout. Each session will be preceded by a 5 minute warm-up period during which intensity will be gradually increased to the training level. Each session will be concluded with a 5 min cool down at a gradually reducing intensity. The intensity and duration of exercise performed at each session will be recorded.

LF-EMS Dynamic exercise
Usual care (no exercise control group): Participants randomised into this group will be supported, but will not receive any intra-dialytic exercise. They will also not be encouraged to participate in interdialytic exercise on their own. The amount of exercise will be determined with a questionnaire and with measurement of free living physical activity over a 7 day period ('Senswear') before randomisation, at the end of the intervention period (week 12) and at end of study (week 20).

Study outcome
Study aim 1: Tolerability of intra-dialytic dynamic exercise and LF-EMS interventions compared to conventional treatment without exercise intervention.

Study aim 3:
Effects of intra-dialytic dynamic exercise training compared to conventional treatment without exercise intervention on measures of functional cardiovascular and skeletal muscle function.

Study aim 4:
Effects of LF-EMS intra-dialytic exercise training compared to conventional treatment without exercise intervention on measures of functional cardiovascular and skeletal muscle function.

Study aim 5:
Qualitatively compare the effects of LF-EMS with dynamic intra-dialytic exercise training on measures of functional cardiovascular and skeletal muscle function. Version 1 -07.11.2013

Outcome measures
The primary outcome measure is the tolerability, safety and acceptance of LF-EMS, compliance/ adherence to LF-EMS and improvement of QOL measures with LF-EMS in comparison to dynamic exercise and a conventional treatment approach.
The secondary outcome measures are cardiovascular and skeletal muscle improvement with LF-EMS in comparison to dynamic exercise and a conventional treatment approach.

Qualitative
Acceptance and tolerability: This will be assessed by recruitment rate into the study (target = 20% of the available population recruited (i.e. 90 out of 450) and recruitment targets met in the specified timescale of 18 months), willingness of participants to be randomised to LF-EMS, dynamic exercise or control; tolerance to the two interventions and compliance/adherence to the protocol (no more than 20% study dropout; reasons will be recorded).

Quality of life (QoL):
Health related quality of life will be assessed using the Kidney Disease Quality of Life (KDQOL-36 TM ) questionnaire which has been extensively validated in CKD populations 74,75 . In addition, the Kansas City Cardiomyopathy Questionnaire will be administered which has been validated for use in stable and decompensated heart failure populations 76 .

Nutritional status:
To provide a global assessment of nutritional status, the Malnutrition-Inflammation Score 77 will be calculated following assessment by a renal dietitian, within 20 minutes of completion of a dialysis session. This tool is a validated measure of nutritional status in ESRD and is predictive of mortality in haemodialysis patients 78,79 .

Safety of intervention:
During and following each session of LF-EMS or dynamic exercise, clinical parameters will be recorded (HR, BP, O2 saturation) and the patient will be monitored for visible signs of clinical instability. Any adverse events will be recorded. The same data will be recorded in the usual care group to allow comparison.

Quantitative Cardiovascular
Free living physical activity: The 'Senswear' physical activity monitor will allow validated measurement of free living physical activity over a 7 day period 70,71 . This device, a small armband worn on the upper arm, measures motion with a tri-axial accelerometer, galvanic skin response (electrical conductivity of the skin, which changes in response to sweat and emotional stimuli), skin temperature and heat flux (amount of heat dissipating from the body). The following parameters will be recorded: total energy expenditure (kcal/min), active energy expenditure (kcal/min), METS (metabolic equivalent of activity), total number of steps, physical activity levels and duration, sleep duration and efficiency, lying down time, On/Off body time. Currently, there is no literature examining physical activity participation, assessed with this technique, in the CKD population.
CPEX: CPEX is an increasingly established tool in the medical management of CHF patients [80][81][82] . The clinical significance of this technology is demonstrated in its ability to quantify physical functional Version 1 -07.11.2013 capabilities, establish the severity of CHF and stratify risk of early death among these patients. CPEXderived measures of maximal aerobic capacity (VO2max) reflect the degree of ventricular function (pumping capacity), vascular function (O2 delivery) and skeletal muscle metabolic capacity (O2 utilization). This test will be carried out on the first non-dialysis day, at least 12 hours after the last dialysis session in order to avoid the effects of interdialytic weight gains and hemodialysis-induced myocardial stunning 83,84 . Tests will be performed in accordance with American Thoracic Society guidelines 85 . A calibrated, electronically braked, upright exercise bicycle will be used in combination with an exercise respiratory gas analysis system. Gas analyser and volume calibrations will be undertaken daily and prior to each test respectively, as dictated by manufacturer's guidelines. Breath by breath respiratory gas exchange measurements of oxygen uptake (VO2), carbon dioxide production (VCO2) and minute ventilation (VE) will be recorded and used to derive ventilatory threshold (VT) and ventilatory efficiency (VE/VCO2 slope) 86 . The 'V slope' method and/or the dual criteria 87, 88 will be utilised to identify VT and the VE/VCO2 slope will be measured from the start of unloaded pedalling to maximal exercise and calculated using linear regression. Commonly, a true VO2 max (as determined by a plateau in VO2) is not achieved in cardiovascular disease populations and thus the measure of peak VO2 (VO2 peak) is preferred 89,90 . Peak VO2 will be determined in the last 30 seconds of exercise as the highest measured average of the mid five of every seven breaths. In addition to gas exchange variables, a twelve lead ECG will be continuously monitored and blood pressure and rating of perceived exertion (RPE) recorded every two minutes. A standard ramp protocol will be employed with increments of 10, 15, 20 or 25 W/min, calculated to ensure an optimal test duration of 9-12 minutes 90 . Three minutes of rest, followed by three minutes of unloaded pedalling, will precede the test and subjects will be strongly encouraged to maintain a cadence of 70 rpm until symptom limited volitional fatigue. In accordance with published criteria, a respiratory exchange ratio (RER) of >1.15 will be considered indicative of maximal effort 89 .

Echocardiography with Tissue Doppler (TDI) and Strain Imaging:
Echocardiography will allow the assessment of cardiac structure and function using traditional and novel analysis techniques. Standard Mmode, 2-dimensional echocardiography, and Doppler blood flow measurements will be performed 91 .
Calculations will include LV diastolic and systolic volumes, mass, ejection fraction (LVEF) according to Simpson's method and isovolumic relaxation time. Mitral inflow measurements will include peak early (E), peak late (A) flow velocities and the E/A ratio. The TDI of the mitral annulus will be obtained from the apical 4-chamber view. A 1.5 sample will be placed sequentially at the lateral and septal annular sites. Analysis will be performed for the early (E') diastolic peak velocity. The ratio of early transmitral flow velocity to annular mitral velocity of the lateral LV wall (E/E') will be taken as an estimate of LV filling pressure. Myocardial deformation (strain and strain rate) will be measured in three planes (longitudinal, circumferential and radial) in the apical 4-chamber view, and the parasternal short axis views at the base (mitral valve) and apex using 2D speckle tracking 92 .
Bioelectrical impedance: Fluid volume status and body composition will be assessed with direct segmental multi-frequency bioelectrical impedance analysis (Biospace Inbody 720). This is currently the recommended method for assessment of fluid status in haemodialysis patients 93 . The primary aim of this assessment is to ensure that fluid volume is comparable at the baseline testing visit and the study completion visit, thus minimising the impact of fluid volume on the assessment of cardiac and vascular function. Patients will stand on the bioelectrical impedance machine whilst eight electrodes are attached to the body.

Arterial applanation tonometry:
Resting brachial blood pressure will be measured following 10 minutes of supine rest. Pulse wave analysis will be performed on the radial artery using a high fidelity micromanometer in accordance with a validated protocol 94 . Aortic pulse wave-form, augmentation Version 1 -07.11.2013 pressure, augmentation index (AIx) and time to reflection (Tr) will be derived using a validated radial-toaortic transfer function (SphygmoCor). The systolic part of the aortic wave-form is characterised by the incident LV ejection pressure peak followed by the second pressure peak caused by the reflected wave. The augmented pressure is the difference between these two peak pressures and the ratio of the augmented pressure to pulse pressure (=systolic minus diastolic blood pressures) defines the AIx. As AIx is influenced by heart rate, an index adjusted to a heart rate of 75 beats/minute (AIx75) will be recorded 95 . Tr is a measure of time of the return of the reflected waves to the central circulation. Using a similar technique, aortic (carotid-femoral) pulse wave velocity (PWV) will be determined by sequential recording of ECG-gated carotid and femoral waveforms.
Functional vascular assessment: Endothelial function will be assessed using flow-mediated dilation (FMD) of the non-fistula arm brachial artery 96 . Baseline artery blood flow and diameter will be measured using ultrasound. A blood pressure cuff will be inflated to >200 mmHg for five minutes. After rapid cuff release, brachial artery ultrasound will be repeated to demonstrate hyperaemia. Arterial diameter will be measured continuously for three minutes following cuff deflation, with maximal diameter being recorded. FMD is expressed as the percentage change in artery diameter after cuff deflation, compared to the preinflation measurement.

Quantitative skeletal muscle
Muscle strength: Leg strength will be measured using a validated hand held dynamometry technique 97 . Subjects will sit in an elevated chair with foam padding and stabilisation straps for the waist and thigh. The subject will attempt to straighten their leg and perform 3 maximal contractions (with 30 seconds rest between each) against a dynamometer which will measure the force of contraction.
Serum and plasma will be stored for further analysis: 1.Cardiac stress: NT-pro-BNP; NT-pro-BNP is representative of LV wall stress and thus indicative of LV hemodynamic compromise 98 . Raised NT-pro-BNP is incrementally associated with poor prognosis and adverse outcomes in CHF 99 . The independent increase in B-type natriuretic peptide (BNP) in response to increasing LV mass index has been observed in CKD patients 100 . With exercise training in CHF, meta-analyses have reported reductions in NT-pro-BNP (37% ), indicating improved LV hemodynamics 101 . This has not been studied in CKD. Troponin T (cTnT): raised cTnT demonstrates a strong association with pathologic LV hypertrophy in CKD 102 and is independently associated with cardiovascular events and mortality 103 . 2.Inflammation: hsCRP, Calprotectin, Cathelicidin; inflammation is one of the main markers of poor outcome in CKD. CRP is associated with all-cause and CV mortality in ESRD 104 . Novel Version 1 -07. 11.2013 inflammatory biomarkers such as Calprotectin and Cathelicidin represent potential future indicators of an improved inflammatory environment with exercise training in CKD. 3.Protein catabolism: Gelsolin, an actin-binding protein that is a key regulator of actin filament assembly and disassembly, and Myostatin, a negative regulator of skeletal muscle mass, are novel markers of hypercatabolism in CKD. Limited data suggests an exercise induced reduction in myostatin in CHF and obesity 105,106 , whereas the effects of exercise on Gelsolin have not been studied. 4.Iron metabolism: Iron, Transferrin and saturation, Hepcidin, an acute phase reactant protein and a key regulator of iron homeostasis is increased with inflammation in CKD, limiting the availability of iron, thus contributing to anaemia 107 . The potential for improved iron metabolism through an exercise induced reduction in hepcidin has not been explored in the CKD population.

Statistical analyses
The two primary objectives of this study are to generate estimates, in particular variance estimates, for a sample size calculation of a subsequent fully powered RCT and to determine the likelihood of a successful conduct of such an RCT. The latter will be assessed by estimating recruitment and drop-out rate indication feasibility of a large-scale trial. Since this study is a pilot the statistical analysis will be primarily descriptive. Point estimates and corresponding 95% confidence intervals will be calculated for all outcome measures in the three arms and the aforementioned rates. Furthermore the distribution of outcome measures will be assessed to identify appropriate statistical analysis methods for a future RCT. Between group comparisons via statistical tests (unpaired two-sample t-test for continuous outcomes, Chi-squared tests for categorical (binary) outcomes) will be conducted, however, it will be clearly stated that these yield only preliminary results since the study is not powered for such investigations. The results of these tests will only be used to identify overwhelming evidence of inferiority of the two experimental treatments relative to usual care so that plans of a full trial can be abandoned or an arm dropped. In accordance with Kieser and Wassmer (1996), the upper 90% confidence limit will be chosen as the estimate of the variance of the treatment effect for a future trial's sample size calculation. Intention-to-treat and per protocol analyses will both be conducted to determine potential ranges of results. Version 1 -07.11.2013 Version 1 -07.11.2013