Stroke volume and cardiac output during 6 minute-walk tests are strong predictors of maximal oxygen uptake in people after stroke.

BACKGROUND AND OBJECTIVES: The 6-minute walk test (6MWT) is a field test commonly used to predict peak oxygen consumption (VO2peak) in people after stroke. Inclusion of cardiodynamic variables measured by impedance cardiography (ICG) during a 6MWT has been shown to improve prediction of VO2peak in healthy adults but these data have not been considered in people after stroke. This study investigates whether the prediction of VO2peak can be improved by the inclusion of cardiovascular indices derived by impedance cardiography (ICG) during the 6MWT in people after stroke.
METHODS: This was a cross-sectional study. Patients diagnosed with stroke underwent in random order, a maximal cardiopulmonary exercise test (CPET) and 6MWT in separate dates. Heart rate (HR), stroke volume (SV) and cardiac output (CO) were measured by ICG during all tests. Oxygen consumption was recorded by a metabolic cart during the CPET. Recorded data were subjected to multiple regression analyses to generate VO2peak prediction equations.
RESULTS: Fifty-nine patients, mean age 50.0±11.7 years were included in the analysis. The mean distance covered in the 6MWT (6MWD) was 294±13 m, VO2peak was 19.2±3.2 ml/min/kg. Mean peak HR, SV and CO recorded during 6MWT were 109±6 bpm, 86.3±8.8 ml, 9.4±1.2 L/min and during CPET were 135±14 bpm, 86.6±9 ml, 11.7±2 L/min respectively. The prediction equation with inclusion of cardiodynamic variables: 16.855 + (-0.060 x age) + (0.196 x BMI) + (0.01 x 6MWD) + (-0.416 x SV6MWT) + (3.587 x CO 6MWT) has a higher squared multiple correlation (R2) and a lower standard error of estimate (SEE) and SEE% compared to the equation using 6MWD as the only predictor.
CONCLUSION: Inclusion of SV and CO measured during the 6MWT in stroke patients further improved the VO2peak prediction power compared to using 6MWD as a lone predictor.

Introduction

Neurological deficits, a sedentary lifestyle and impaired diastolic function are all factors which impair exercise capacity of people after stroke [1,2]. Peak oxygen consumption (VO2peak) measured after stroke is usually significantly lower than in sex-matched sedentary healthy individuals [3-5]. Suboptimal exercise capacity not only leads to low quality of life but also results in increased risks of recurrent stroke and accident [1,6].

Therapeutic exercise training is known for its potential in increasing aerobic capacity, an essential element for functional recovery in people after stroke [7,8]. Appropriate and effective prescription of a rehabilitation exercise program is guided by accurate assessment and prediction of the participant’s aerobic capacity. Aerobic capacity, either predicted or from direct measurement, is an essential outcome for progress monitoring and program evaluation. Progressive cardiopulmonary exercise testing (CPET) is the gold standard for assessment of aerobic capacity [9]. However, in people after stroke, neuro-motor impairment, poor balance, and spasticity often limit them from reaching their maximum aerobic capacity [10].

The 6-minute walk test (6MWT) is a common field test used to monitor and evaluate submaximal aerobic capacity in people with chronic cardiopulmonary disease [11,12], and after stroke [13,14]. The outcome used in a 6MWT is the distance covered during this submaximal test (6MWD) and reflects the VO2peak (aerobic capacity) of the participant. However, the ‘distance’ covered during the test, particularly in the stroke population, can be influenced by non-hemodynamic factors. Heart rate (HR), stroke volume (SV) and cardiac output (CO) are the primary contributing factors to exercise capacity [15,16]. These cardiodynamic parameters can now be measured non-invasively and conveniently during a 6MWT. Information from these ‘direct’ cardio-dynamic factors could more reliably apprise the aerobic capacity of the participant, compared to using only the 6MWD as the outcome. Further, the predictive power of 6MWD for VO2peak in the stroke population has been questioned [10]. Inclusion of these parameters during a 6MWT for prediction VO2peak has recently been shown to be more accurate in predicting VO2peak than using 6MWD alone in young adults [17]. Whether such a relationship can be extended to people after stroke has not been determined.

This study aimed to explore, in a cohort of patients after stroke, (1) the relationship between cardiodynamic parameters (HR/SV/CO) and VO2peak recorded during a CPET and a 6MWT, and (2) whether inclusion of cardiodynamic parameters recorded during a 6MWT in the linear regression equation for prediction of VO2peak can further improve the accuracy and stability of the linear regression model, compared to using 6MWD alone as the predictive parameter.

Materials and methods

This was a cross-sectional study approved by the Institutional Review Board of Shenzhen Second People’s Hospital (Ethics approval number: KS20191119004). The study protocol (Register number: ChiCTR1900028393) is available at the Chinese Clinical Trial Register Center website: www.chictr.org.cn. The protocol of this study is also available at: dx.doi.org/10.17504/protocols.io.6qpvr6pp2vmk/v1.

Participants

People diagnosed with stroke and receiving treatment at the Second People’s Hospital, Shenzhen, China, between December 2019 and December 2021, were invited to participate in the study through in-hospital poster advertising. Interested participants were screened for inclusion eligibility (see criteria below). The nature of the study was explained and written informed consent was obtained from all participants prior to data collection.

The inclusion criteria were: (1) age ≥ 18 years, (2) clinically diagnosed with ischemic and/or hemorrhagic stroke, (3) period since stroke diagnosis ranging from 3 to 12 months after diagnosis, (4) able to independently ambulate with or without an assistive device for ≥ 100 meters, (5) medically stable and with no significant limitation due to pain, and (6) able to clearly comprehend the exercise testing instructions.

The exclusion criteria were: (1) current use of beta-blocker medications; (2) neurological condition other than stroke and/or an orthopedic condition causing motor deficit (e.g. fracture, degenerative joint change, clinical instability of the hip or knee joint), (3) psychiatric impairment, such as severe depression or panic disorder, (4) pregnancy, (5) uncontrolled hypertension, arrhythmia, or an unstable cardiovascular status, (6) recent pulmonary embolism, subacute systemic illness or infection, and (7) brain injury affecting the respiratory and circulatory centers, e.g. brainstem injury.

Procedure

Demographic data for each eligible participant (including age, sex, height, weight, body mass index (BMI), percent body fat and lean body mass) were recorded. The stroke type (cerebral infarction, intracerebral hemorrhage), medical history of comorbidity including hypertension, diabetes mellitus, cardiovascular disease, lipidemia, kidney disease, pulmonary disease, National Institutes of Health Stroke Scale (NIHSS), Modified Rivermead Mobility Index (MRMI), Berg Balance Scale (BBS), Barthel index (BI) were retrieved from the patient’s medical record.

NHISS—a 15-item impairment scale used to measure stroke severity by means of a score ranging from 0 to 42 [18]. The higher the NHISS score, the more severe the stroke. MRMI is an 8-item scale with score ranging from 0 to 40 [19]. Higher MRMI scores imply a greater level of independence during transfers, balance and walking.

BBS is a scale which includes 14 items that measure balance with a total score between 0 and 56. A higher BBS score indicates better balance [20].

BI is a 10-item instrument assessing the basic activities of daily living for an individual. The minimum score of 0 reflects totally dependent, and a maximum score of 100 indicates totally independent living [21].

Participants were requested to attend the hospital cardiopulmonary laboratory twice, 72 hours apart. They were randomized to either a CPET or two 6MWTs at their first visit. The two 6MWTs were conducted consecutively with a separation of 30 minutes. Thus, if a patient had CPET during their first visit, the 6MWTs would be conducted 72 hours later, and vice versa. Each visit was scheduled at least 2 hours after a light meal. Participants were also requested to avoid caffeine-containing products, nicotine, and alcohol for at least 12 hours before attending the laboratory. Breath-by-breath oxygen consumption was recorded during the CPET and cardiodynamic parameters (HR/SV/CO) were recorded with impedance cardiography (ICG) during both the CPET and 6MWTs (see below).

Measurement of cardiac parameters

The HR, SV and CO during the tests were measured at 1-second intervals using a PhysioFlow®PF07 EnduroTM (PhysioFlow Enduro, Paris, France). The PhysioFlow®PF07 is a portable, non-invasive device that adopts real-time wireless monitoring of morphology-based impedance cardiography signals via a blue tooth USB adapter. Electrode placement was conducted as described by Tonelli and colleagues [22]. Auto-calibration of the device was performed as instructed by the manufacturer before data collection. Variables were measured by the EnduroTM at 1-second intervals, prior to, during and after, each 6MWT and CPET. HR was derived from the electrocardiograph (ECG). Variations in the impedance signal during cardiac ejection generate a specific waveform from which the SV was calculated [23]. CO was computed by multiplication of the SV and HR. The reliability of ICG in people after stroke has been previously reported [24].

6-minute walk test (6MWT)

The 6MWT was performed in a 30-meter hospital hallway. Two 6MWT trials were performed to accommodate any possible learning effect and to ensure maximal effort, each test was conducted according to the American Thoracic Society (ATS) guidelines [9]. A rest period of at least 30 minutes was allowed between tests. Each participant was asked to rest in a sitting position for 10 minutes before and after each 6MWT. Apart from ICG recording as described above, oxygen saturation (SpO2) was recorded by a pulse oximeter (Heal Force, POD-3, China) immediately before, and at minute intervals during, and at the end of the 6MWT. Blood pressure (BP) was measured by an electronic blood pressure monitor (OMRON, U30, China) immediately before and after the 6MWT, and at 2-minute intervals during the 10-minute rest period post 6MWT. Perceived fatigue sensation (modified Borg 0–10 Scale [25]) was recorded immediately at the end of 6MWT. Both the 6-minute walk distance (6MWD) as well as the cardiodynamic data recorded during the ‘better’ performed 6MWT were used for analysis.

Cardiopulmonary exercise test (CPET)

Participants were requested to perform a symptom limited, graded CPET on a cycle ergometer (Ergoline GmbH, ergoselect 200, Germany). Throughout the test, 12-lead electrocardiography was continuously recorded. Participants were required to wear a mask and breathe through a calibrated volume sensor attached to a metabolic cart (MasterScreenTM CPX, CareFusion, Germany). Breath-by-breath oxygen consumption, carbon dioxide consumption and respiratory exchange ratio (RER) were measured. The gas analysis system was fully calibrated immediately before every test in accordance with the manufacturer’s instructions. HR, SV, and CO data were recorded by ICG at 1-second intervals during the CPET. Blood pressure was measured at 2-minute intervals, and SpO2 was measured at 1-minute intervals during the CPET.

CPET protocol

The CPET protocol commenced with a rest period of 3 minutes sitting on the cycle ergometer to establish a steady state, then a 3-minute warm-up stage with pedaling without resistance. Participants were then required to pedal at increasing intensity of 4–8 W increments each minute, to ensure that the total exercise time remained in a range of 8–12 minutes. Participants were instructed to maintain a cycling speed of 55–65 revolutions per min. Strong verbal encouragement was given throughout the test. The test was terminated once the participant was unable to maintain the required pedaling rate despite encouragement, or should any signs of risk to health, as prescribed in the guidelines of the American College of Sports Medicine (ACSM) [26] become manifest. Respiratory exchange ratio (RER) > 1.15 and Borg scores at the level “very hard” were used to signify a “maximal” exercise test performance [27]. Participants were asked to rate their sensation of fatigue level immediately after the test using the modified Borg 0–10 scale administered during the 6MWT.

Data analysis

All data were analyzed using IBM SPSS Statistics for Windows, Version 25.0 (Armonk, NY: IBM Corp). Demographic data and clinical characteristics for all participants were summarized using descriptive statistics. Variables of interval-ratio data meeting a normality assumption were compared using the independent t test for gender difference. The Chi-square test was used for the analysis of gender specific categorical data. HR, SV, CO were averaged every 10 seconds. Differences in HR, SV, CO recorded at rest, at the peak of CPET and at the end of the 6MWT were analyzed using repeated-measures ANOVA. The associations between measured VO2peak and recorded variables (including 6MWD, HR, SV, CO) at the end of 6MWT and at the end of the CPET were analyzed by Pearson’s correlation coefficients. The p-value of <0.05 was considered statistically significant.

Similar statistical analyses, as described in our previous work, were adopted [17]. Three sets of multiple linear regression analyses were primarily performed to explore the optimal predictor variables of VO2peak. The first regression equation was built by 6MWD as the single variable predicting VO2peak. The second regression equation was generated using BMI, age, gender, time after stroke, 6MWD, plus HR, SV and CO recorded by ICG at the end of the better performed 6MWT as predictors for VO2peak. The third regression equation utilised the same variables as in the 2nd equation but with peak HR, SV and CO measured during CPET. The stepwise backward regression method was adopted to determine the significant predictor variables to be retained in the regression equations. The appropriateness and precision of the regression parameters were evaluated with the squared multiple correlation (R2), the standard error of estimate (SEE) as well as the ‘SEE/meanVO2peak’ ratio, which was expressed as a percentage (SEE%). The predicted residual sum of squares (PRESS) statistic [28] was computed to estimate the degree of R2 shrinkage (R2) when the VO2peak regression equation was used for cross-validation across similar but independent samples. PRESS-derived R2, SEEp and SEEp% for each regression model were compared.

Two further analyses were conducted to examine the unique contribution of changes in HR or SV to any changes in CO, during 6MWT and CPET; forced entry regression method with HR change and SV change as predictor variables were used.

Sample size estimation

The PASS 15.0.5 (Kaysville, Utah: NCSS) was used to calculate the sample size required for the multiple regression analyses. A sample size of at least 54 participants were required, with 8 predictor variables (age, gender, BMI, time after stroke, 6MWD, HR, SV, CO) modeled for estimated effect size of 0.30, α level of 0.05 and power of 0.80.

Results

A total of 67 patients met the inclusion criteria and participated in the exercise tests. Eight participants either had RER below1.15 or were unable to tolerate the CPET and thus were excluded from the analysis. A flow chart illustrating the number of patients entered for final analysis is displayed in Fig 1.

Fig 1

A flow chart illustrating the number of patients entered for final analysis.

The mean age of the 59 participants (52 males) included was 50.0±11.7 years. Demographic data and clinical characteristics of these participants are displayed in Table 1.

Table 1

Demographic data and clinical characteristics of the 59 participants.

Data in n (%) or mean±SD.

	Total (n = 59)	Male (n = 52)	Female (n = 7)	P value
Age (years)	50.0±11.7	49.6±11.6	53.1±12.5	0.451
Height (cm)	167 ± 6	168.2±4.9	159.0±3.9	<0.001
Weight (kg)	65.5±8.6	66.5±8.2	58.5±9.0	0.020
BMI(kg/m2)	23.4±2.7	23.4±2.5	23.3±4.1	0.915
Lean body mass (kg)	49.8±5.2	50.8±4.6	42.9±4.6	<0.001
Diagnosis        Cerebral hemorrhage        Cerebral infarction	21 (36%)38 (64%)	17(33%)35(67%)	4 (57%)3 (43%)	0.024
Lesion side        Left        Right	26(44%)33(56%)	22(42%)30(58%)	4(42%)3(58%)	0.393
Time after stroke (months)        <3        3–6        7–12	12 (23%)29 (49%)18 (28%)	9(17%)25(48%)18(35%)	3(43%)4(57%)0(0%)	0.030
Comorbidities        Hypertension        Diabetes mellitus        Hyperlipidemia        Heart disease        Kidney disease        Pulmonary disease	24 (41%)28 (48%)9 (15%)1 (2%)0 (0%)0 (0%)	22(43%)24(46%)6(12%)1(2%)0 (0%)0 (0%)	2(29%)4(57%)3(43%)0(0%)0 (0%)0 (0%)	0.024
Ambulation        Independent    With one walking stick	56 (97%)3 (3%)	50(96%)2(4%)	6(86%)1(14%)	0.229
NHISS score	2.5±1.8	2.5±1.8	2.6±1.7	0.923
MRMI score	37.6±3.2	37.7±3.2	36.9±3.4	0.533
BI score	89.4±11.9	89.4±12.1	89.3±11.0	0.977
Berg score	49.8±9.3	49.9±9.7	48.6±5.8	0.767
Borg score at the end of the test with greater 6MWD	5±2	5±2	5±1	0.725
Borg score at the end of the CPET	8±1	8±1	8±1	0.775
Peak work rate achieved during CPET (W)	68.8±23.1	71.5±22.7	48.9±16.1	0.013
VO2peak achieved during CPET (ml/min/kg)	19.2±3.2	19.5±3.1	16.8±2.5	0.038
6MWD (m)	294±131	298±131	265±135	0.539

BMI = Body Mass Index; NHISS = National Institutes of Health Stroke Scale; MRMI = Modified Rivermead Mobility Index; BI = Barthel Index; 6MWD = the farther distance covered in the two 6 minute walk tests; VO2peak = peak oxygen consumption, CPET = cardiopulmonary exercise test.

The 6MWD, peak HR, SV, CO during 6MWT and CPET and VO2peak recorded at CPET are displayed in Table 2. Peak HR and CO during CPET were significantly higher than that achieved at the end of the 6MWT, with a mean difference in HR and CO being 26 (95% CI: 22.3 to 29.3) and 2.3 (95% CI: 1.8 to 2.8), respectively. However, there was no difference in maximum SV between the two tests, the mean difference of maximum SV was 0.3 (95% CI: -3.2 to 2.5).

Table 2

Cardiodynamic parameters measured by ICG at end of 6MWT and CPET.

Data in mean±SD.

	6MWT							CPET
		Heart rate		Stroke volume		Cardiac output		VO2	Heart rate		Stroke volume		Cardiac output
	6MWD(m)	RHR-6MWT(bpm)	PHR-6MWT(bpm)	RSV-6MWT(ml)	PSV-6MWT(ml)	RCO-6MWT(L/min)	PCO-6MWT(L/min)	Peak VO2(ml/min/kg)	RHR-CPET(bpm)	PHR- CPET(bpm)	RSV-CPET(ml)	PSV- CPET(ml)	RCO- CPET(l/min)	PCO- CPET(l/min)
Malen = 52	298±131	81±11	110±5	66.7±10.8	86.6±8.8	5.3±0.8	9.5±1.2	19.5±3.1	82±11	136±14	66.6±11.6	87.3±8.9	5.4±0.8	11.9±2
Femalen = 7	265±135	80±8	104±5	60.7±7.4	84.4±9.6	4.8±0.6	8.8±1.4	16.8±2.5	85±10	125±13	63±10	82.1±9.1	5.4±1.1	10.3±1.8
Totaln = 59	294±131	81±11	109±6	66±10.6	86.3±8.8	5.3±0.8	9.4±1.2	19.2±3.2	82±11	135±14	66.2±11.4	86.6±9	5.4±0.8	11.7±2

ICG = impedance cardiography; 6MWT = 6 minute walk test; 6MWD = the farther distance covered in the two 6MWTs; CPET = cardiopulmonary exercise test; RHR–6MWT = Resting heart rate–6MWT; PHR–6MWT = Peak heart rate–6MWT; RSV–6MWT = Resting stroke volume–6MWT; PSV–6MWT = Peak stroke volume–6MWT; RCO–6MWT = Resting cardiac output–6MWT; PCO–6MWT = Peak cardiac output–6MWT; VO2 = oxygen consumption; RHR–CPET = Resting heart rate–cardiopulmonary exercise test; PHR–CPET = Peak heart rate–cardiopulmonary exercise test; RSV–CPET = Resting stroke volume–cardiopulmonary exercise test; PSV–CPET = Peak stroke volume–cardiopulmonary exercise test; RCO–CPET = Resting cardiac output–cardiopulmonary exercise test; PCO–CPET = Peak cardiac output–cardiopulmonary exercise test.

The VO2peak achieved during CPET correlated well with 6MWD (r = 0.66), peak HR (r = 0.82) and peak CO (r = 0.74) during CPET, as well as with HR (r = 0.73) and CO (r = 0.6) recorded at the end of the 6MWT. However, correlations between measured VO2peak and SV during either CPET (r = 0.39) and 6MWT (r = 0.4) were poor (Table 3).

Table 3

Correlation between VO2peak and 6MWD, maximal HR, SV and CO during both 6MWT and CPET.

Variable	Correlation with VO2peakPearson r	P value
HRPeak (bpm)	0.82	<0.001
SVPeak (ml)	0.39	<0.05
COPeak (L/min)	0.74	<0.001
HRend (bpm)	0.73	<0.001
SVend (ml)	0.40	<0.05
COend (L/min)	0.60	<0.001
6MWD	0.66	<0.001

6MWT = 6–minute walk test; 6MWD = the farther distance covered in the two 6MWTs; CPET = Cardiopulmonary exercise testing; VO2peak = measured peak oxygen consumption at CPET; HRPeak = peak heart rate at CPET; SVPeak = peak stroke volume at CPET; COPeak = peak cardiac output at CPET; HRend = heart rate at the end of 6MWT; SVend = stroke volume at the end of 6MWT; COend = cardiac output at the end of 6MWT.

Regression analyses for prediction of VO2peak

Three multiple linear regression models were generated for prediction of VO2peak (Table 4). Model 1, with 6MWD as the only predictor variable, produced a R2 of 0.44 and a SEE% of 12.5%. Model 2 used SV and CO recorded during the maximal CPET as predictor variables and this equation produced a R of 0.69 and SEE% of 9.2%. Model 3: 16.855+(-0.060 x age(years)) + (0.196 x BMI (kg/m2) + (0.01 x 6MWD(m)) + (-0.416 x SV6MWT (ml)) + (3.587 x CO6MWT (L/min)) was associated with the highest R (0.7) and lowest SEE (1.75 mL/kg/min) and SEE% (9.1); this model is comparable to Model 2 for prediction of VO2peak measured during CPET from data measured during the 6MWT. The PRESS derived regression model illustrated a Rp2 of 0.66, SEEp of 1.83mL/kg/min and SEE p % of 9.5%. This model also produced the highest correlation coefficient between measured and predicted VO2peak (r = 0.847, p<0.001) (Table 5).

Table 4

Multiple regression analyses for prediction of VO2peak.

Model	Predictor variables (Xn)			Coefficients	β	R2	SEE	SEE%	Rp2	SEEp	SEEp%	Regression equation
	Variables put into model	Removed from model because ofnon-significance	Retained
Model 16MWD as sole predictor	6MWD	-	constant6MWD	14.4420.016	0.663	0.44	2.4	12.5	0.39	2.46	12.8	VO2peak = 14.442 + (0.016 x 6MWD)
Model 2with data from CPET	BMI, Age, Gender,Time after stroke, HRpeak,SVpeak,COpeak	AgeGenderBMIHRpeak	constantSVpeakCOpeak	15.869-0.2422.076	-0.6851.303	0.69	1.77	9.2	0.67	1.82	9.5	VO2peak = 15.869 + (-0.242 x SVpeak) +(2.076 x COpeak)
Model 3with data from 6MWT	BMI, Age, Gender,Time after stroke, 6MWD,HRend, SVend, COend	GenderTime after strokeHRend	constantBMIAge6MWDSVendCOend	16.8550.196-0.0600.010-0.4163.587	0.167-0.2230.398-1.1581.402	0.70	1.75	9.1	0.66	1.83	9.5	VO2peak = 16.855+(-0.060 x age) + (0.196 x BMI)+ (0.01 x 6MWD)+ (-0.416 x SVend) +(3.587 x COend)
Model 3adata from 6MWT excluding age, BMI	HRend, SVend, COend 6MWD	HRend	constant6MWDSVendCOend	15.2530.009-0.3643.465	0.383-1.0151.355	0.64	1.91	9.9	0.61	1.98	10.3	VO2peak = 15.253 + (0.009 x 6MWD) + (-0.364 x SVend) + (3.465 x COend)

VO2peak = Peak oxygen consumption during the CPET; 6MWD = the farther distance covered in the two 6minute walk tests; HRend = Heart rate at the end of 6MWT; SVend = Stroke volume at the end of 6MWT; COend = Cardiac output at the end of 6MWT; HRpeak = Peak heart rate during the CPET; SVpeak = Peak stroke volume during the CPET; COpeak = Peak cardiac output during the CPET.

Table 5

Correlation between measured and predicted VO2peak generated by different equation models.

VO2peak prediction Model		Correlation with measured VO2peakr values	P value
Model 1	VO2peak= 14.442 + (0.016 x 6MWD)	0.663	<0.001
Model 2	VO2peak= 15.869 + (-0.242x SVpeak) +(2.076xCOpeak)	0.838	<0.001
Model 3	VO2peak= 16.855+(-0.060 x age) + (0.196 x BMI) + (0.01 x 6MWD) + (-0.416 x SVend) +(3.587 x COend)	0.847	<0.001

r = Pearson correlation coefficient.

VO2peak = Peak oxygen consumption during the CPET; 6MWD = the farther distance covered in the two 6minute walk tests; SVend = Stroke volume at the end of 6MWT; COend = Cardiac output at the end of 6MWT; SVpeak = Peak stroke volume during the CPET; COpeak = Peak cardiac output during the CPET.

Regression analyses to determine unit contribution of HR or SV changes to CO

Multiple linear regression analyses further revealed that the changes in HR or SV contributed equally to the change of CO during the 6MWT. During the CPET however, the change in CO was mainly a response to HR change rather than changes in SV (Table 6).

Table 6

Contribution of unique changes in HR and SV to changes in CO (Multiple linear regression analyses using change in CO as an outcome variable).

	At the end of 6MWT			At the end of CPET
	male	Female	all	male	female	All
HR changestandardized beta coefficient	0.73	0.64	0.73	0.79	0.71	0.78
SV changeStandardized beta coefficient	0.69	0.44	0.67	0.55	0.34	0.52
HR changesemipartial correlation	0.73	0.49	0.73	0.78	0.46	0.78
SV changesemipartial correlation	0.69	0.34	0.67	0.55	0.21	0.52
HR changeunique contribution to CO change (%)	54	24	53	61	21	60
SV changeunique contribution to CO change (%)	48	11	45	30	5	27
6MWD (m)	298±131	265±135	294±131	-	-	-
Peak VO2(ml/min/kg)	-	-	-	19.5±3.1	16.8±2.5	19.2±3.2

6MWT = 6–minute walk test; CPET = Cardiopulmonary exercise testing; 6MWD = the farther distance covered in the two 6MWTs; HR = heart rate; SV = stroke volume; CO = cardiac output.

Discussion

This study explored the hemodynamic responses recorded during a CPET and a 6MWT in the same cohort of people after stroke. Our study revealed that measured VO2peak during the CPET correlated well with HR, SV and CO data recorded during both CPET and the 6MWT (Table 4). The 6MWT is a convenient field test commonly used as an alternative method of assessment of aerobic fitness, especially in people with chronic impairment of cardiorespiratory and neurological function [11,29]; and as the 6MWT is a sub-maximal exercise test, it is not surprising that the peak heart rate and maximal cardiac output achieved at the end of the 6MWT were lower than peak HR and CO recorded during the CPET (Table 2). Consequently, the level of exertion as expressed by the Borg’s score recorded at the end of the 6MWT was only 5 compared to 8 at the end of the CPET, suggesting that the 6MWT appeared less strenuous compared to the exercise demands during the CPET.

Poor cardiovascular fitness in people after stroke affects quality of life [3]. Exercises that aim to increase aerobic fitness are recommended for inclusion in rehabilitation programs for stroke survivors [6,8]. An accurate assessment of VO2peak is necessary to prescribe an optimum exercise program which permits accurate monitoring and evaluation, of aerobic capacity. The primary aim of this current study was to identify an equation that most accurately predicts VO2peak in people after stroke. Cardiodynamic data recorded from the 6MWT and CPET were used to generate an equation to predict the VO2peak achieved during the CPET, the gold standard for evaluation of maximal aerobic capacity. In accord with our previous work with young healthy adults [17], this study reveals that the inclusion of age, BMI and SV and CO recorded by ICG (a non-invasive and simple to apply procedure during a 6MWT) generated more accurate prediction equation for VO2peak achieved by the CPET in our cohort of patients after stroke. Using 6MWD as the sole predictor of VO2peak, the correlation coefficient between predicted and measured VO2peak was only 0.66, but this correlation coefficient increased to 0.85 with the inclusion of age, BMI, and SV and CO recorded during 6MWT (Table 5). Not only was this model associated with a higher R and lower SEE and SEE% compared to the equation which relied on 6MWD as the predictor of VO2peak, the R2 and SEE values associated with this equation suggest that the predictive power of this equation is as accurate as using an individual’s peak SV and CO during the gold standard CPET as predictors of VO2peak (Model 2, Table 4). Additional analysis showed that with removal of the predictors age and BMI (Model 3a, Table 4), the R2 value of the equation remains comparable to that of Model 2. This suggests that SV and CO measured by ICG during a 6MWT are strong predictors of the measured VO2peak in our cohort of participants with stroke. The R and SEE values of VO2peak prediction equation in people with aneurysmal subarachnoid hemorrhage was previously reported by Harmsen and colleagues to be 0.56 and 4.12 respectively [30]. The mean age of our patient cohort and that reported by Harmsen’s group were similar, but the mean 6MWD achieved by our cohort (294 ±131m) was comparatively much lower than the Harmsen cohort (498±98m). The 6MWD has been commonly used to predict VO2peak but because functional capacity varies according to stroke severity, 6MWD has not been considered as a suitable indicator of aerobic capacity in people after stroke [31]. This current study shows that inclusion of SV and CO data recorded during a 6MWT provided a more precise prediction of maximal aerobic capacity in our stroke cohort.

It is interesting to note that while the CO during CPET was significantly higher than that recorded during a 6MWT, the maximal SV recorded during both tests were similar (Tables 2), inferring that the extra CO demands during CPET were met by way of an increase in HR, rather than SV. This inference is further supported by analysis of the standardized beta coefficient associated with changes in HR and SV. As illustrated in Table 6, the unique contribution of HR and SV changes to the change in CO during the 6MWT was 53% and 45% respectively. However, the contribution of HR change to CO was twice the change in SV seen during the CPET, affirming that the increase in HR was the main contributor to the increase in CO during the CPET. This finding accords with our previous report [17]. There are multiple explanations for dampened SV responses to increased exercise demands in the stroke population. It has been shown in athletes with paraplegia, that an increase in CO was accompanied by a significant increase in HR [32], whereas able-bodied individuals attained CO with a lower HR and higher SV; these authors postulated that this anomaly was due to a comparative reduction in venous return during exercise in people with paraplegia, and consequently a reduced SV. Further, it has been postulated that poor functional capacity is associated with suboptimal left ventricular diastolic function in people after stroke [2], it thereby follows that impaired left ventricular diastolic function could dampen the SV response to increased exercise intensity in this population. Similarly, in people with heart failure, maintenance of CO during CPET was reportedly associated with a compensatory increase in HR [33]; an increase in heart rate leads to a reduced diastolic filling time thereby reducing stroke volume [34].

Implications of the study

Although the 6-minute walk test (6MWT) is a common field test used to monitor and evaluate submaximal aerobic capacity, the predictive power of 6MWD on VO2peak in people after stroke has been considered unreliable. The 6MWD is an outcome often used to reflect the VO2peak (aerobic capacity) of the participant, however, in the stroke population, the ‘distance’ covered during the test can be influenced by non-haemodynamic factors. Heart rate (HR), stroke volume (SV) and cardiac output (CO) are primary contributing factors to exercise capacity and can now be measured noninvasively and conveniently by ICG during a 6MWT. Information from these ‘direct’ cardiodynamic factors should more reliably inform the clinician of the aerobic capacity of the participant. Further, the predictive power of 6MWD for VO2peak in the stroke population has been questioned [10]. Our findings show that the inclusion of SV and CO during 6MWT combined with 6MWD improves the predictive power of aerobic capacity in people after stroke compared to the predictive power of 6MWD alone. Thus the accuracy of a prescription for an effective rehabilitation exercise program can be enhanced. However, whether using this new predictive equation results in a more effective outcome following the prescribed program warrants further investigation.

Limitations of the study

The majority (90%) of patients admitted to our center during the data collection period were males. Unfortunately, it is not clear why this occurred. The participants were those with moderate motor deficits post-stroke and a mean 6MWD of less than 300m. It may not be appropriate to apply inferences drawn from our data analysis to female patients, however the pattern of response to exercise stress between male and female patients appeared to be similar. Our data suggest that further studies with a larger sample size to explore any effect of age, gender, and a wide range of motor capacity on ICG measured cardiodynamic parameters in response to exercise training, are warranted.

Conclusion

This study demonstrated that inclusion of SV and CO measured during the 6MWT in stroke patients further improved the VO2peak prediction power compared to using 6MWD as a lone predictor. Further studies exploring the regression model with wider demographic profiles are warranted.

STROBE statement—checklist of items that should be included in reports of observational studies.

(DOCX)

Click here for additional data file.

12 Jul 2022

26 Jul 2022

We are grateful for all reviewers’ valuable and insightful comments, and we have revised our manuscript following their suggestions.

Reviewer #1:

Dear Editors and Authors

Thank you for the opportunity to review the paper "Stroke volume and cardiac output during 6 minute-walk tests are strong predictors of maximal oxygen uptake in people with stroke" that concerns interesting research area - how to simply and objectively assess exercise capacity in patients with stroke. The authors use measurements of cardiovascular hemodynamics while 6MWT to predict VO2 peak in CPET. The concept is good because CPET is quite a difficult and time consuming procedure. The results of the study revealed that VO2 peak can be estimated from 6MWT+ICG with clinically acceptable accuracy. The paper is well written.

Authors response: We are grateful for Reviewer 1’s supportive comments.

I have only minor comments to be considered:

1. Abstract verses 16-18 - please put units after numbers, not in brackets .ie. "6MWD was 294+/-13 m"

Authors response: Units are now inserted as suggested. Please see lines 16-19 in the Abstract.

2. Please rephrase the second sentence of Introduction (verses 29-31) to be more general and simply, i.e. "Peak oxygen consumption (VO2 peak) measured after stroke is usually significantly lower than in sex-matched...."

Authors response: The sentence is modified as suggested. Please see lines 30-32.

3. Please provide information how the data was transferred from Physioflow device to the Physioflow software - was it real-time wireless transfer or did you used memory card with post-exam cable transfer to PC?

Authors response: Data from the Physioflow device was transmitted directly to the Physioflow software via real-time wireless transfer. This information is now included in the Methodology section, under Measurement of Cardiac parameters, lines 120-124.

4. Please comment on possible low quality of ICG while CPET and 6MWT - were any problems with it. If yes - please comment in Limitations.

Authors response: We did not encounter any issues associated with the quality of ICG data recorded. Signals appeared stable during both CPET and 6MWT.

5. verse 106 - provide full name of 6MWT in subheading.

Authors responses: The full name of 6MWT has been added in the subheading, line 132.

6. Table 1 - "Duration of stroke" - better "Time from stroke"

Authors response: “Duration of stroke” is now replaced with “Time after stroke” in the text and in Tables 1 and 4.

7. Verse 237-240 - the difference between model 2 and 3 are very slight, do not write that model 2 was better, it was comparable (as you correctly write in discussion).

Authors response: This sentence is now reworded as suggested, please see lines 272-273.

Reviewer #2:

Overall, a thorough and well-executed work that is clear and well described.

However, I am missing a small important addition in the introduction in relation to why this study is relevant, which the authors argue for in the discussion.

I would also find it relevant to elaborate a bit more about how these findings is relevant and can contribute in the clinical practice of stroke rehabilitation.

If the authors will consider the following comments, I find the manuscript contributing to gaining knowledge in the field of rehabilitation stroke survivors.

Authors response: We are thankful for Reviewer 2’s insightful comments, we have included a more detailed explanation of the rationale for our study in the Introduction, as suggested.

Comments for the manuscript:

1. In the introduction (line 27-52) I lack to understand why it is important to assess ICG to 6MWT in stroke patients? Why is it not enough just to have the 6MWT as an expression for the submaximal VO2test?

Authors response: The primary aim of our study was indeed an attempt to answer this important question. The outcome used in a 6MWT is the distance covered during this submaximal test and the distance is used to reflect the VO2peak (aerobic capacity) of the participant. However, the ‘distance’ covered during the test in the stroke population can be influenced by non-haemodynamic factors. Heart rate, stroke volume and cardiac output are the primary contributing factors to exercise capacity. As these can now be measured noninvasively and conveniently during a 6MWT, we believe these ‘direct’ factors measured during the stress of an exercise test could more reliability reflect the aerobic capacity of the participant, than merely the distance. Further, the predictive power of 6MWD for VO2peak in the stroke population has been questioned. This forms the basis of the rationale of this study. We have now included this explanation in the introduction. Please refer to lines 45-54.

2. How and why is this important in the clinic? It is presented in the discussion line 306-307, but this should be presented in the introduction as well.

Authors response: The importance of using an accurate prediction of aerobic capacity to guide exercise prescription is now included in the Introduction, please see lines 35-39.

Methods:

3. Line 71-72: Did you use a score for the ambulation ability such as FAC?

Authors response: Thank you for this suggestion. In hindsight, it would have been an informative addition to reflect the mobility status of our cohort. In this study, apart from the 6MWD, we have only included the Modified Rivermead Mobility Index (MRMI) to reflect the mobility status of our participants.

4. Line 85-87 “National Institutes of Health Stroke Scale (NIHSS), Modified Rivermead Mobility Index (MRMI), Berg Balance Scale (BBS), Barthel index (BI)”

The tests should be short described according to score range indication no/severe dependence/impairment and a reference.

Authors response: A short description of these instruments is now included. Please see lines 100-108.

5. Line 88-90: “Participants were then requested to attend the hospital cardiopulmonary laboratory 89 twice (72 hours apart) to perform, in random order, a progressive cycle ergometer test 90 or two 6MWTs (at least 30 minutes apart)”. —I do not understand. Did the participants first do the one kind of test and then the day after the other kind of test? I am not sure by the description in this section.

Authors response: We apologise for the confusion. Description of this procedure is now clarified. Please see lines 109-113.

6. Line 193-194” The VO2peak achieved during CPET correlated well with 6MWD, peak HR and peak CO during CPET, as well as with HR and CO recorded at the end of the 6MWT”—You are stating something but not presenting the numbers. Please present the numbers and move the statement to the discussion section.

Authors response: The correlation coefficients were now included in the text (lines 237-240) as well as in Table 3.

7. Line 221: Table 2 6MWD is not clarified in the table note

Authors response: 6MWD is now defined in the table note. Please refer to the updated Table 2.

8. Line 224: Table 3 Why is 6MWD represented as it is stated that the best of the two tests is represented. What is it compared with? The same goes with COpeak (L/min).

Authors response: Two 6MWTs were performed by each participant, 30 min apart. We selected the better 6MWT, i.e. the test that resulted in the farthest 6MWD as the representation of the 6MWT. This is now clarified in the table note of Tables 2 and 3 as well as in the text. Please refer to lines 144-146.

9. Regression analyses for prediction of VO2peak line 233-245 I must say it I rather confusing but I am not the right person to verify if this method is the correct.

Authors response: Presentation of results of the multiple linear regression analyses was confirmed and checked by our statistician and is consistent with previous work published in PLoS One. We are happy to follow any suggestions which could make the interpretation easier.

Discussion:

10. Results line 185: 59 participants (52 males) —How come so many males? Do you have any comment on that?

Authors response: Indeed, this is a limitation of our study. For reasons unknown to us, the majority of patients admitted to our center who met the inclusion criteria were males during the data collection period. We were not able to explain this but have addressed this under the Limitations of the study. Please see line 405-410.

11. Line 188-191: “Peak HR and CO during CPET were significantly higher than that achieved at the end of the 6MWT, with a mean difference in HR and CO being 26 (95% CI: 22.3 to 29.3) and 2.3 (95% CI: 1.8 191 to 2.8), respectively.”

—That is some mean differences, why?

Authors response: We postulate that the differences were a result of the fact that the 6MWT is only a submaximal exercise test while the CPET is a maximal exercise test. This observation is explained in the discussion section, please refer to lines 325-331.

12. Your findings: SV and CO measured during the 6MWT in stroke patients further improved the VO2peak prediction power compared to using 6MWD as a lone predictor.

–By how much is the prediction improved in including impedance cardiography (ICG) during a 6MWT?? And is the difference clinical relevant?

Authors response: In Model 1, the 6MWD was the only predictor variable and the results showed a small R2 of 0.44 and a larger SEE% of 12.5%. However, when the SV and CO were added with 6MWD, the R2 increased to 0.64 and the SEE% decreased to 9.9% (as shown in Table 5). A higher R2 value suggests a more reliable and stable predictive power, and a lower SEE% suggests a lower risk of error in estimation. Thus, mathematically this improvement is considered significant. However, the clinical relevance of this new predictive equation needs further investigation. We have included a paragraph on “Implications of the study” to address this important point. Please see lines 387-403.

13. I would find it relevant to elaborate a bit more about how these findings is relevant and can contribute in the clinical practice of stroke rehabilitation.

Authors response: We hope the paragraph “Implications of the study” mentioned above suitably addresses this point.

Reviewer #3:

Dear Authors

I appreciate the authors and PLOS ONE for the opportunity to evaluate this paper. The authors led an interesting study that the prediction of VO2peak can be improved by the inclusion of cardiovascular indices derived by impedance cardiography (ICG) during the 6MWT in people with stroke. They found that the prediction equation with inclusion of cardio dynamic variables: 16.855 + (-0.060 x age) + (0.196 x BMI) +20 (0.01 x 6MWD) + (-0.416 x SV6MWT) + (3.587 x CO 6MWT) has a higher squared multiple correlation (R2) and a lower standard error of estimate (SEE) and SEE% compared to the equation using 6MWD as the only predictor. These findings will be of interest to clinicians, as well as researchers in the field.

Authors response: We are thankful for the Reviewer’s positive comments.

I have following concerns：

Introduction

1. What is the clinical implication of this study? Is it significant that it has been conducted in young adults but not in stroke? If so, I would think a comparison with trends in young adults would be in order.

Authors response: We are thankful for the reviewer’s insightful comment. The rationale for this study was based on reports that the predictive power of 6MWD for aerobic capacity in people with stroke was questionable. While we have previously shown that inclusion of cardio-dynamic data recorded from ICG could improve the predictive power of aerobic capacity in young adults, we believe the response to exercise stress between people with and without stroke may be different, and the regression model could vary. We have included a more detailed explanation for the rationale of our study in the Introduction, as well as inclusion of a paragraph under the subheading “implications of the study” in the discussion (lines 387-403).

Methods

2. Are patients with arrhythmias excluded? There are descriptions of uncontrolled arrhythmias, but what about controlled Af? Additionally, were there any patients with brainstem lesions? There may be some impact because the respiratory and circulatory centers are located.

Authors response: We did not include patients with brainstem injury, although it was not stated in the exclusion criteria. We have now included this in our list of exclusion criteria (please see line 90-91). The patients with uncontrolled arrhythmias were excluded (this was listed at point 5 in the exclusion criteria). If a stroke patient with arrhythmia was stable, the patient would be included in our study.

Results

3. A patient flow chart of inclusion/exclusion criteria may be helpful to understand the patient’s characteristics.

Authors response: A flow chart of participant enrollment and reasons for exclusion is now included as Figure 1.

4. My greatest concern is that most of the subjects are male. It is mentioned in the limitation, but we need to be very careful about generalizing. Authors should consider presenting male-only results.

Authors response: Indeed, this is a limitation of our study. For reasons unknown to us, the majority of patients admitted to our center were males during the data collection period. We were not able to explain why this is so but have addressed this under the Limitations of the study. Please see line 405-410.

5. I think the p-value in Table 1 should be a specific number, not p>0.05.

Authors response: The specific p values are now included in Table 1.

6. Kidney disease and pulmonary disease were recorded, but there are no data in Table1.

Authors response: The data for kidney disease and pulmonary disease are now included in Table 1.

7. Table 6 “At the end of end”. Is it “At the end of 6MWT”?

Authors response: Thank you for pointing out this typo error. The error in Table 6 is now rectified.

Discussion

8. It may not be appropriate to apply inferences drawn from our data analysis to female patients, however the pattern of response to exercise stress between male and female patients appeared to be similar. The tendency of the pattern of response may be similar, but the degree to which they are different is quite different.

Authors response: Indeed, we are aware of the limitation of gender bias. However, the pattern of response to exercise stress between male and female patients appeared to be similar. Therefore, we included the results of all the male and female participants in the report.

Submitted filename: Response to reviewers.docx

Click here for additional data file.

16 Aug 2022

Stroke volume and cardiac output during 6 minute-walk tests are strong predictors of maximal oxygen uptake in people after stroke

PONE-D-22-12074R1

Dear Dr. Jones,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Shane Patman, PhD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: All comments have been addressed

Reviewer #3: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Yes

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

Reviewer #3: I Don't Know

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: Yes

Reviewer #3: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: Thank you for including me in this process and for having the confidence in letting me review your interesting work. Comments and suggestions for improvement have been met and made and I recommend the article to be published.

Reviewer #3: Dear Authors

The authors did an excellent job addressing my previous comments and their response seems reasonable. The manuscript and associated figures are now more informative. I have no further comments.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: Yes: Henriette Busk

Reviewer #3: No

**********

Submitted filename: Letter to the editor August 2022.docx

Click here for additional data file.

19 Aug 2022

PONE-D-22-12074R1

Stroke volume and cardiac output during 6 minute-walk tests are strong predictors of maximal oxygen uptake in people after stroke

Dear Dr. Jones:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Assoc Prof Shane Patman

Academic Editor

PLOS ONE