Peak aerobic capacity from incremental shuttle walk test in chronic kidney disease
Texto Integral @ b-onSUMMARY
Background
Assessment of cardiorespiratory fitness is an important outcome in chronic kidney disease (CKD). We aimed to develop a predictive equation to estimate peak oxygen uptake (VO2peak) and power output (WPeak), as measured during a cardiopulmonary exercise test (CPET), from the distance walked (DW) during the incremental shuttle walk test (ISWT).
Methods
Thirty‐six non‐dialysing patients with CKD [17 male, age: 61 ± 12 years, eGFR: 25±7 ml/min/1.73 m2, body mass index (BMI): 31 ± 6 kg/m2] carried out laboratory‐based CPET on a cycle ergometer and ISWT on two separate occasions.
Results
Linear regression revealed that DW/BMI was a significant predictor of VO2Peak (r = 0.78) (VO2Peak (ml/min/kg) = [0.5688 × (DW/BMI) (m)] + 11.50). No difference (p = 0.66) between CPET VO2Peak (19.9 ± 5.5 ml/min/kg) and predicted VO2Peak (19.9 ± 4.3 ml/min/kg) was observed. DW multiplied by body mass (BM) was a significant predictor of WPeak (r = 0.80) [WPeak (W) = (0.0018 × (DW × BM)) + 50.47]. No difference (p = 0.97) between CPET WPeak (116.2 ± 38.9 W) and estimated WPeak (113.9 ± 30.1 W) was seen.
Conclusion
The present study demonstrates that VO2Peak and WPeak can be accurately estimated using the DW during an ISWT in CKD populations.
INTRODUCTION
Chronic kidney disease (CKD) is associated with reduced exercise capacity and increased cardiovascular disease risk (Go et al. 2004). A growing body of research has shown that cardiorespiratory fitness, which can be reflected by peak oxygen consumption (VO2Peak), is strongly related to a spectrum of health outcomes including cardiovascular disease and all‐cause mortality (Blair et al. 1996; Harber et al. 2017). In the general population, low cardiorespiratory fitness is a greater indicator of mortality than any other traditional risk factor (Myers et al. 2002; Laukkanen et al. 2004; Blair 2009; Shah et al. 2016). A study of young adults aged 18–30 over a period of 26 years reported a significant relationship between high cardiorespiratory fitness and reduced risk of all‐cause mortality (Shah et al. 2016). Furthermore, in men aged 42–60 years, with and without conventional risk factors for cardiovascular disease, VO2Peak represented a strong predictor of fatal and non‐fatal cardiac events (Laukkanen et al. 2004), and Myers et al. (2002) showed, in a sample of 6,213 men, each 1‐metabolic equivalent (MET) increase in exercise capacity, equating to 3.5 ml/min/kg, conferred a 12% improvement in survival.
Unsurprisingly, given its multi‐morbid nature, patients with CKD have markedly reduced cardiorespiratory fitness compared to age‐predicted values (Howden et al. 2015). Such poor levels of cardiorespiratory fitness are associated with greater cardiovascular burden (e.g. aortic stiffness and poor left ventricular function), and an overall elevated risk of further morbidity and mortality (Gulati et al. 2012; Howden et al. 2015; MacKinnon et al. 2018). Whilst no estimate exists in patients not requiring renal replacement therapy, in end stage kidney disease VO2Peak values > 17.5 ml/min/kg are a powerful predictor of survival (Sietsema et al. 2004). Consequently, measuring cardiorespiratory fitness in CKD is fundamental for assessment of current and prognostic health status and survival.
Cardiorespiratory fitness (measured as VO2Peak) represents the ability to utilise oxygen during increased metabolic demand (Bassett & Howley 2000). The ‘gold standard’ assessment of VO2Peak is through an incremental cardiopulmonary exercise test (CPET) (Palange et al. 2007) using breath‐by‐breath sampling methods. However, these are poorly tolerated in older patients with low fitness (Gill et al. 2000), and require specialized costly equipment. The incremental shuttle walk test (ISWT), a simple symptom‐limited field test, has been used as a surrogate marker for both VO2Peak and peak power output (WPeak) in clinical populations such as chronic obstructive pulmonary disease (COPD) (Singh et al. 1994; Arnardóttir et al. 2006). Both VO2Peak and WPeak can help set appropriate exercise intensities when prescribing exercise, and our group have previously shown the ISWT and CPET to have good reliability in CKD patients (Wilkinson et al. 2018).
The aim of the present study was to develop prediction equations to estimate VO2Peak and WPeak based on the distance walked (DW) during the ISWT. This can aid healthcare researchers and professionals in the evaluation of patient health status, disease prognosis and prescription of appropriate exercise intensities for rehabilitation interventions without having to subject patients to expensive and occasionally poorly tolerated laboratory testing.
METHODS
STUDY DESIGN
This is a secondary analysis of baseline data collected during an exercise trial (#ISRCTN36489137), which took place between December 2013 and October 2016. A full description of the methods for this trial can be found in Watson et al. (2018). The study was approved by the National Research Ethics Committee, East Midlands‐Northampton (#13/EM/0344) and all participants gave informed consent.
PARTICIPANTS
Patients were recruited from nephrology outpatient clinics and included if they had CKD stages 3b–5, and were not yet requiring dialysis. Exclusion criteria were: (a) aged < 18 years, (b) body mass index (BMI) > 40 kg/m2, (c) physical impairment preventing undertaking of the study assessments, (d) myocardial infarction within <6 months, (e) any unstable chronic condition (e.g. diabetes) and (f) inability to give informed consent.
CARDIOPULMONARY EXERCISE TESTING
VO2Peak was determined using a symptom‐limited, graded, maximal effort CPET performed on a cycle ergometer (Lode Excalibur, Groningen, The Netherlands). Participants underwent a full test familiarisation. Following a three‐minute warm‐up at a resistance of 50 W, the patients were instructed to cycle >60 revolutions per minute (RPM). Work rate was increased using a ramp protocol of incremental workloads of 1 W/4 seconds (15 W/min) (Cortex Metasoft CPX Software, Leipzig, Germany). The test was stopped if the RPM dropped <60 or if the participant reached volitional exhaustion. VO2Peak and WPeak (W) were calculated over a rolling 20‐second average value.
INCREMENTAL SHUTTLE WALK TEST
The ISWT involved participants walking up and down a 10‐m course (Singh et al. 1994; Wilkinson et al. 2018). Walking speed was controlled by an external auditory tone. The test was maximal and progressive, with an initial speed of 0.5m/s, increasing by 0.17 m/s every minute, for a maximum of 12 minutes. Standardised instructions were played before each test. The test was terminated when the participant felt they were unable to continue or if the participant was unable to sustain the speed for two continuous shuttles. The DW throughout the test was calculated. All participants completed a prior familiarisation test.
STATISTICAL ANALYSIS
As a CPET provides relative oxygen uptake (e.g. ml/min/kg), we normalised ISWT results by systematically individually multiplying and then dividing the relative age, BMI and body mass (BM) of the participants. These variables were selected as they are most likely to be available to healthcare professionals or researchers. Individual r values were calculated for each variable by assessing the relationship between each variable and the VO2Peak and WPeak (see Table 1). The variable with the highest r value was used to develop the prediction equations for VO2Peak and WPeak.
| VO2Peak (r) | WPeak (r) | |
|---|---|---|
| DW × Age | 0.52 | 0.55 |
| DW × BMI | 0.57 | 0.72 |
| DW × BM | 0.64 | 0.80aa Variable used to create the prediction equation for estimating WPeak. |
| DW/Age | 0.52 | 0.61 |
| DW/BMI | 0.78bb Variable used to create the prediction equation for estimating VO2Peak. |
0.59 |
| DW/BM | 0.73 | 0.48 |
| DW | 0.74 | 0.71 |
- BM: body mass; BMI: body mass index; DW: distance walked during an incremental shuttle walk test.
- a Variable used to create the prediction equation for estimating WPeak.
- b Variable used to create the prediction equation for estimating VO2Peak.
Following selection of the most appropriate variable, linear regression was carried out: (a) between the DW (m) divided by BMI (kg/m2) and the VO2Peak (ml/min/kg) determined using CPET; and (b) between the DW (m) multiplied by BM (kg) and the WPeak (W) determined during the CPET. This elicited adjusted prediction equations for the VO2Peak and WPeak adjusting for BMI and BM respectively. To elicit unadjusted prediction equations for VO2Peak and WPeak in conditions where BMI and BM measurements are absent, linear regression was also carried out: (a) between DW and the VO2Peak (ml/min/kg) determined using CPET; and (b) between the DW (m) and the WPeak (W) determined during the CPET.
Non‐normally distributed data underwent log transformation. Paired sample t tests were used to test for significant differences between values assessed using CPET and the estimated VO2Peak and WPeak using the prediction equation. Paired sample t tests were also used to test for significant differences between the estimated results using adjusted and unadjusted prediction equations (VO2Peak: adjusted for BMI; WPeak: adjusted for BM). Mean differences are shown along with 95% confidence intervals (95% CI). Bland and Altman analysis plots assessed the agreement between values obtained during CPET and those estimated using the prediction equations. Significance was set at <0.050. Data analysis was performed using IBM SPSS Statistics for Windows, Version 24.0 (IBM Corp., Armonk, NY).
RESULTS
Out of the 54 patients consenting to the main trial, 36 patients completed the CPET and the ISWT. Baseline characteristics for these patients are found in Table 2.
| n = 36 | |
|---|---|
| Age (years) | 61 ± 12 |
| eGFR (ml/min/1.73 m2) | 25 ± 7 |
| Number of males | 17 |
| Ethnicity | |
| White | 25 |
| Indian/South Asian | 9 |
| Black Caribbean | 2 |
| Body mass index (kg/m2) | 31 ± 6 |
| Systolic/diastolic blood pressure (mmHg) | 137 ± 13/80 ± 9 |
| VO2Peak (ml/min/kg) | 19.9 ± 5.5 |
| ISWT distance walked (m) | 430 ± 190 |
| Haemoglobin (g/dl) | 119 ± 15 |
| Comorbid conditions | |
| Diabetes, n (%) | 9 (25) |
| Hypertension, n (%) | 15 (42) |
- eGFR: estimated glomerular filtration rate; ISWT: incremental shuttle walk test.
ADJUSTED PREDICTION EQUATIONS
PREDICTION OF VO2Peak USING ADJUSTED EQUATIONS
(A) the correlation between VO2Peak measured during a cardiopulmonary exercise test (CPET) and the distance walked during an incremental shuttle walking test (ISWT) divided by body mass index (BMI); (B) a Bland and Altman scatterplot comparing VO2Peak values obtained during CPET and estimated using the prediction equation; (C) the correlation between WPeak (W) measured during CPET and the distance walked during the ISWT multiplied by body mass; (D) a Bland and Altman scatterplot comparing WPeak values obtained during CPET and estimated using the prediction equation.
PREDICTION OF WPeak USING ADJUSTED EQUATIONS
(A) the correlation between VO2Peak measured during a cardiopulmonary exercise test (CPET) and the distance walked during an incremental shuttle walking test (ISWT); (B) a Bland and Altman scatterplot comparing VO2Peak values obtained during CPET and estimated using the prediction equation; (C) the correlation between WPeak measured during CPET and the distance walked during an ISWT; (D) a Bland and Altman scatterplot comparing WPeak values obtained during CPET and estimated using the prediction equation.
UNADJUSTED PREDICTION EQUATIONS
PREDICTION OF VO2Peak USING UNADJUSTED EQUATIONS
PREDICTION OF WPeak USING UNADJUSTED EQUATIONS
DISCUSSION
CPET is considered the ‘gold standard’ for measuring aerobic capacity, an important risk factor determining morbidity and mortality in CKD (Gulati et al. 2012; Howden et al. 2015). This was demonstrated by Gulati et al. (2012) who reported that VO2Peak values <17.5 ml/min/kg (~5 MET's) combined with estimated glomerular filtration rate values of <45 ml/min/1.73 m2 were associated with higher mortality rates compared to those with a better aerobic capacity and eGFR. In the present study, we report an average of the VO2Peak of 19.9 ml/min/kg determined using CPET; these values are below that of comparative age‐matched healthy sedentary individuals (Herdy & Uhlendorf 2011).
We found the relationship (r = 0.78) between DW during an ISWT and the VO2Peak obtained during CPET support the findings of others on COPD [r = 0.72 (20), r = 0.88 (Singh et al. 1994; Onorati et al. 2003)]. Similar relationships (r = 0.74) have also been reported in idiopathic pulmonary fibrosis (Moloney et al. 2003), and heart failure (r = 0.83) (Green et al. 2001). It has been shown that the VO2Peak determined using a treadmill‐based CPET is higher than the VO2Peak values measured using a cycle‐based CPET (Christensen et al. 2004; Myers et al. 2009), thus a regression equation derived using a non‐cycle walking‐based test could be expected to overestimate VO2Peak determined using an incremental cycle CPET. However, in the present study, average VO2Peak values determined during CPET (19.9 ± 5.5 ml/min/kg) were almost identical to estimated VO2Peak values determined using both formulae adjusted (19.9 ± 4.3 ml/min/kg) and unadjusted (19.9 ± 4.0 ml/min/kg) for BMI.
We also used the DW during the ISWT to develop a prediction equation of WPeak. The relationship (r = 0.80) we observed between the WPeak obtained during CPET and the one estimated using the prediction equation adjusted for BM supports research by Arnardóttir et al. (2006), who also showed a strong relationship (r = 0.88) between DW during the ISWT after controlling for BM and the WPeak obtained during CPET in 93 COPD patients. Given the high prevalence of cardiovascular disease in CKD, many patients are prescribed β‐blockers to manage hypertension. The confounding effects of this medication makes it difficult to prescribe exercise intensities according to heart rate target zones in this patient population. Therefore, the prediction equation in our study estimating WPeak through DW during the ISWT may be more useful than the use of the peak heart rate for the prescription of exercise in this patient population. Furthermore, patients whom use a cycle ergometer in the gym or at home, and without knowledge of their heart rate, could set appropriate exercise intensities using WPeak.
No significant differences were seen between data obtained using adjusted and unadjusted equations for VO2Peak and WPeak (p = 0.93; p = 0.14, respectively). This suggests that in the absence of BM or BMI measurements, the unadjusted equations still provide a valid means of estimating VO2Peak and WPeak. Being able to measure VO2Peak and WPeak using the ISWT does not make the ISWT a perfect substitute for the laboratory‐based incremental cycle test or other forms of laboratory‐based CPET. In the present study, slight variations were seen between estimated and measured peak values on an individual basis (as seen in the Bland and Altman plots with measures falling outside the 95% CI) however, when grouped, the values showed no significant difference, making the ISWT together with our prediction equations a valid means of obtaining this data in the absence of expensive laboratory tests or when these tests are poorly tolerated by the patients.
Arnardóttir et al. (2006) discussed the possibility that the prediction equations they provided for COPD could be inversed, making it possible to calculate walking speed from WPeak determined during CPET. This is also possible with the equations presented in the present study. Given that the equations work both ways, it is possible to estimate the DW during an ISWT and therefore, walking speed from WPeak determined during CPET. This may be useful in clinics that have laboratory‐based cycle testing as routine practice but want to prescribe walking‐based exercise instead of gym‐ or home‐based exercise.
STUDY LIMITATIONS
The patient population in the present study was an opportunistic sample obtained from a randomised clinical trial carried out by our group, and therefore the findings are limited by the characteristics of the cohort included. For example, the prediction equations may not be applicable to CKD patients outside the eGFR, BMI or age range studied. A larger and more diverse population would be required to tailor prediction equations for possible confounding factors such as BMI, gender and ethnicity. Additionally, the present method of estimating VO2Peak is only limited to people that can effectively carry out the walking test. Balance issues or visual impairments may influence the correct application of the walking test, thus affecting predicted VO2Peak values. Another possible limitation is our use of a cycle ergometer for the CPET, whereas a treadmill test may have provided a more relevant comparison to the ISWT. Indeed, previous studies have shown higher rates of O2 consumption and CO2 production in peak tests using a treadmill than a cycle protocol (Muscat et al. 2015), although other exercise parameters did not differ between the two protocols. In this study, we chose to use the ergometer for patients safety reasons and because it is easier to standardise than a treadmill test and therefore likely provides more reproducible results, Furthermore, many patient participants are unfamiliar with exercise equipment and feel more confident to exert themselves seated on an ergometer than walking or running on a treadmill.
IMPLICATIONS FOR CLINICAL PRACTICE
Cardiorespiratory fitness is a strong predictor of outcome and an important measure of health status in the general population and in chronic disease states, including CKD. CPET is the gold‐standard method for measuring this but requires specialized equipment and a trained operator and is therefore costly and often inaccessible for general use by renal clinicians, nurses and other allied healthcare staff. In the present study, we have identified that the ISWT is a good predictor of cardiorespiratory fitness in non‐dialysis CKD. We have developed adjusted and unadjusted formulae that use the DW during an ISWT and measures of BMI and BM to predict both VO2Peak and WPeak. This is of clinical importance as the ISWT is a considerably cost‐effective and simpler method for estimating aerobic capacity. These formulae provide a simple means to estimate both VO2Peak and WPeak in non‐dialysis CKD patients, which can be easily used as a field test by the renal multidisciplinary team.
CONCLUSIONS
The present study demonstrates that peak aerobic capacity and peak exercise capacity can be estimated using an ISWT with similar accuracy as CPET in non‐dialysis CKD populations staged 3b–5. Early measures of cardiorespiratory and exercise capacity represent a quantifiable and prognostic biomarker of survival in this patient population. Therefore, the ability to evaluate and monitor improvements in peak aerobic capacity easily, coupled with the ability to set appropriate exercise intensities using WPeak without the use of expensive equipment can provide a means to evaluate and improve the risk of adverse events.
ACKNOWLEDGEMENTS
We would like to thank Amy Clarke, Barbara Vogt, Darren Churchward, Patrick Highton and Charlotte Grantham, researchers at the Leicester Kidney Lifestyle Team, for collection of some of the assessment outcome data.
FUNDING
This research was co‐funded by the National Institute for Health Research (NIHR) Leicester Biomedical Research Centre (BRC) and the Stoneygate Trust. The views expressed in this publication are those of the author(s) and not necessarily those of the NHS, NIHR Leicester BRC or the Department of Health. At the time of writing this manuscript, E.L.W. was supported by a Kidney Research UK Post‐Doctoral Fellowship (PDF2–2015).
CONFLICT OF INTEREST
The authors declare that there are no conflicts of interest.
AUTHOR CONTRIBUTIONS
SX and TJW: responsible for generation and collection of data, performed data analysis and interpretation and drafted and approved the final version of the submitted manuscript. ACS, ELW and JLV: involved in conception and study design and revised and approved the final version of the manuscript. DWG responsible for generation/collection of data and revised and approved the final version of the manuscript.
Biography
-
Professor Alice Smith has worked in kidney research since 1989. Over the last 15 years she has built up the Leicester Kidney Lifestyle Team (LKLT), a 20?strong multidisciplinary research team who aim to optimise the health and well‐being of kidney patients through improved lifestyle management. Her bench‐to‐bedside translational research programme encompasses basic laboratory investigations of muscle metabolism and cardiovascular factors in CKD and the impact of lifestyle factors; clinical trials of the feasibility, efficacy and effectiveness of lifestyle interventions; evaluation of outcomes measures; qualitative research of patient and stakeholder perspectives; and the design, development, testing and implementation of behavioural interventions to encourage healthy lifestyle. All the research is underpinned by extensive and genuine patient and public involvement and engagement.
