A 3‐minute recumbent stepper test in chronic stroke

Persons with stroke often have difficulty achieving target heart rate (HR) during graded exercise testing (GXT), which is known to limit test sensitivity for detecting clinically relevant cardiac conditions. A novel Recumbent Stepper 3‐minute (RS 3Min) “all out” test may increase sensitivity of stress testing after stroke.


INTRODUCTION
Aerobic exercise has been shown to improve mobility, cardiovascular health, aerobic capacity and quality of life after stroke and is therefore recommended in current stroke rehabilitation recommendations and guidelines. 1,2owever, clinical implementation of aerobic exercise in stroke rehabilitation can be challenging. 3,4For example, some guidelines recommend a symptom-limited graded exercise test (GXT) with electrocardiography (ECG) prior to exercise prescription in persons with stroke, primarily to screen for silent myocardial ischemia. 1,2However, the sensitivity of such testing is dependent on the peak heart rate (HR) achieved, 5 and persons with stroke often have difficulty reaching the typical target HR (85% of agepredicted maximal HR [AP-HRmax]) during a GXT due to neurologic impairments. 6aseline GXTs are also used to prescribe an individualized exercise intensity target, usually at some percentage of peak HR or peak HR reserve (HRR). 7,8his could also be problematic after stroke because peak HR on a GXT may be limited by motor function before an anaerobic end point is reached. 9,10When peak aerobic capacity is uncertain, one option for exercise prescription is to directly measure thresholds between different intensity domains. 11For example, the first ventilatory threshold, measured with gas exchange analysis during a GXT, represents the beginning of the transition into increased anaerobic metabolism and occurs at the upper end of the moderate intensity domain (at about 50%-60% HRR). 11,12Critical power, the highest physiologically sustainable workload (ie, with steady state oxygen consumption and blood lactate), is a different intensity threshold that occurs at the upper end of the vigorous intensity domain (at about 70%-80% HRR) and can be measured without gas exchange analysis. 11Ventilatory threshold can be reliably measured after stroke, 9 but no previous studies in this population have assessed critical power.
We propose that a novel Recumbent Stepper 3-minute (RS 3Min) "all out" test may enable persons with stroke to achieve higher peak HR than a standard GXT for cardiovascular screening.Second, the higher peak HR combined with the critical power measurements may assist in individualized exercise intensity prescription.This 3-minute test has been previously studied among healthy adults using a cycle ergometer. 13It requires a participant to perform an all-out effort, stepping as fast as possible against a low constant resistance for 3 minutes.However, the feasibility of an RS 3Min test had not been previously tested for persons with stroke.In this population, an RS has been shown to enable higher aerobic intensities than a traditional cycle ergometer during a GXT. 14Therefore, the aims of this study were to evaluate a 3-minute exercise test on an RS post stroke to assess its feasibility for (1) cardiovascular screening and (2) measuring critical power.

Overview
This study used a within-participant, nonrandomized, repeated measures design.After informed consent and screening, 15 participants with chronic stroke each had two study visits separated by 1 week.During the first visit, participants performed a symptom-limited treadmill GXT with gas exchange analysis.During the second visit, participants performed a symptom-limited modified total body RS (mTBRS) GXT followed by an RS 3Min test after 15 minutes of rest.Each session was performed under the direction of a licensed physical therapist with support from a research assistant for data collection.

Setting and participants
The study was approved by the University of Cincinnati Institutional Review Board and was performed in a rehabilitation research laboratory and cardiovascular stress laboratory.Participants were recruited from the community and signed informed consent prior to participation.Inclusion criteria were age 21-80 years; unilateral paresis from stroke experienced >6 months prior to enrollment (to help ensure stable deficits); residual gait impairment; able to walk with assistive devices as needed and no physical assistance; stable cardiovascular condition; discharged from formal rehabilitation; and able to communicate with investigators and correctly answer consent comprehension questions.Exclusion criteria were evidence of significant arrhythmia or myocardial ischemia on treadmill exercise test; hospitalization for cardiopulmonary disease within 3 months; pacemaker or implanted defibrillator; lower extremity claudication; severe lower limb spasticity, that is, Ashworth >2; weight bearing pain >4/10; and pregnancy.

Screening examinations
Participants completed a physical examination, over ground 10-meter walk test, 7,15 resting ECG and a baseline GXT with ECG monitoring to confirm eligibility and obtain baseline characteristics.

Exercise testing protocols
During all testing sessions, participant monitoring included 12-lead ECG activity, manually measured blood pressure (BP) and continuous observation for signs or symptoms of cardiorespiratory insufficiency, orthopedic injury, or new/worsening neurologic impairment, using accepted stop criteria. 9,16Participants wore a safety harness during treadmill testing to protect from falls.For participants on β-blocker medication, the medication was not withdrawn for testing.All tests within participant were scheduled at the same time of day and standardized dietary instructions were given.

Treadmill graded exercise test (treadmill GXT)
The symptom-limited treadmill GXT followed the stroke-specific protocol of Macko et al. 7 Speed remained constant throughout the test at a predetermined level from a treadmill screening test. 7Grade started at 0% and was increased 2%-4% every 2 minutes until volitional fatigue, severe gait instability, or a cardiovascular safety limit.

Recumbent stepper graded exercise test (RS GXT)
The symptom-limited mTBRS GXT was performed on a T5XR RS (Nustep Inc., Ann Arbor, MI) and followed the stroke-specific protocol of Billinger et al. 14 Paretic upper limb use was prohibited to prevent injury and impact on test results.Participants stepped at a constant cadence of 80 steps/min.Resistance was increased 15 watts every 2 minutes until volitional fatigue, inability to maintain 80 steps/min, or a cardiovascular safety limit.

Recumbent stepper 3-min test (RS 3Min)
The RS 3Min test was performed on the RS with the same limbs as the GXT.Each participant rested a minimum of 15 minutes, had a recovery HR near resting values and at minimum below 100 bpm, and subjectively determined readiness prior to the RS 3Min test.Participants were instructed to step as fast as possible for 3 minutes against a constant resistance.Resistance was predetermined by a NuStep steep ramp test conducted on the initial screening visit where resistance was increased 15 watts every 15 seconds until the participant could no longer maintain 80 steps/min.Resistance for the RS 3Min test was then set at 25% of the maximum resistance successfully completed during the steep ramp test.During the RS 3Min test, if power output was still systematically changing in the last 30 seconds, the test was extended until the power output remained constant for 30 seconds.

Data collection
HR was measured continuously during each test with an ECG (Nasiff CardioCard ECG).Rate of perceived exertion (RPE) was measured at the conclusion of each test using the Borg 6-20 scale. 17BP during the treadmill GXT (GaitKeeper 2200T) and mTBRS GXT (NuStep T5XR) was measured every 2 minutes and immediately post exercise.BP for the RS 3Min test was measured only immediately post exercise and was taken on the nonparetic arm.Oxygen consumption (VO 2 ) and respiratory exchange ratio (RER) were measured continuously using the TrueOne 2400 (Parvo Medics, Sandy, UT) and a facemask interface.Power output, in watts, was measured continuously for 12 participants.All of these variables except BP and RPE were averaged in 20-second intervals for analysis.

Statistical analysis
Potential utility of the RS 3Min test for cardiovascular screening Peak HR, BP, and RPE were compared between the RS 3Min test, the mTBRS GXT, and the treadmill GXT to determine the relative potential utility of the RS 3Min test for cardiovascular screening.The percentage of participants who achieved target HR (85% AP-HRmax) was calculated for each test and the percentage achieving target HR was compared between tests.For continuous variables, analysis was done using the general linear model with an unstructured covariance matrix to account for the within-participant correlations between each of the tests.For the percentage achieving target HR, analysis was done using generalized estimating equations with a binomial distribution, a logit link function and an unstructured covariance matrix.When the omnibus test found a significant difference among the protocols (p < .05),pairwise testing was conducted between the RS 3Min test and each of the GXTs, using Dunnett adjustment for multiple comparisons.

Feasibility of the RS 3Min test for measuring critical power
For an RS 3Min test to provide a valid assessment of critical power, the participant has to complete the test, power output must decline then plateau (end power significantly closer to minimum power than to peak power), and VO 2 should increase systematically then plateau at VO 2 max (peak VO 2 during 3Min test not significantly lower than the peak VO 2 reached during the RS GXT).Therefore, validity of the RS 3Min test was analyzed by: 1. Describing the trajectory of mean power output during the test.2. Testing the relative closeness of the end-test power (critical power measurement) to the peak power versus the minimum power.This analysis compared the peak-end difference to the end-minimum difference using a general linear model with an unstructured covariance matrix to account for the correlation between the two difference measures obtained during the same test.3. Describing the trajectory of mean VO 2 during the test.
4. Measuring the percentage of participants with a VO 2 plateau (change <2.1 mL/kg/min) after reaching their peak VO 2 value during the test. 18.Comparing VO 2 -peak from the RS 3Min test to the mTBRS GXT and the treadmill GXT.This analysis was done using the general linear model with an unstructured covariance matrix and Dunnett correction for multiple comparisons, as described previously.6. Comparing power output, cadence, resistance, and RER from the RS 3Min test to the RS GXT.For most of these variables, analysis was done using the general linear model with an unstructured covariance matrix.However, this was not possible for cadence because there was no cadence variability for the mTBRS GXT, which used a prescribed cadence of 80 steps/min.Thus, cadence from the RS 3Min test was analyzed with a one-sample t-test against the null value of 80 from the mTBRS GXT.
All analyses were performed with SAS v9.4 and missing data were handled with the method of maximum likelihood.

RESULTS
Fifteen participants performed the RS 3Min test in addition to the two GXTs.Participant characteristics are presented in Table 1.One participant's HR data were excluded from Table 2 because he was in atrial fibrillation for all testing sessions.All other data from this participant were included.

Potential utility of the RS 3Min test for cardiovascular screening
Average target HR (85% AP-HRmax) was 144 ± 5 beats/ min and average HRpeak values during the RS 3Min test, treadmill GXT, and mTBRS GXT were 140 ± 19, 137 ± 21, and 133 ± 16 beats/min, respectively.The number of participants who achieved target HR (85% AP-HRmax) for the RS 3Min test, treadmill GXT, and mTBRS GXT were 9, 5, and 4, respectively (out of 14).Peak HR was significantly higher for the RS 3Min test compared to the mTBRS GXT (adjusted p = .03),but there was no significant difference in peak HR between the RS 3Min test and the treadmill GXT (adjusted p = .73;Tables 2 and 3).BP and RPE were not significantly different among the three tests (p = .050to .244).

Feasibility of the RS 3Min test for measuring critical power
Fourteen of 15 (93%) participants completed the full RS 3Min test on the first attempt.One participant requested to stop the test early on their first attempt but was able to complete it on the second attempt after a 15-minute rest.
On average, participants reached peak power output early in the RS 3Min test, then power output began declining at 40 seconds and plateaued at 140 seconds (Figure 1A).End-test power (estimated critical power; 110 ± 41 W) was significantly closer to minimum power (103 ± 39 W) than to peak power (210 ± 139 W; p = .016).
On average, VO 2 during the RS 3Min test increased to peak VO 2 then plateaued from 100 seconds to the end of the test (Figure 1B).This VO 2 plateau (change <2.1mL/kg/min after reaching peak VO 2 ) was exhibited by 12 of 15 (80%) participants.VO 2 -peak from the RS 3Min test was not significantly different from either of the GXTs (Tables 2 and 3).

Safety
No serious adverse events occurred during any of the exercise testing sessions.ECG was monitored throughout all sessions for signs of significant arrhythmias or evidence of myocardial ischemia.Fourteen of the 15 participants were free of arrhythmias or ischemia with one participant displaying atrial fibrillation while at rest and during exercise.One participant reported moderate knee pain when questioned after both the mTBRS GXT and RS 3Min test.One participant reported mild transient lightheadedness when being questioned after the mTBRS GXT and RS 3Min test.No other abnormal signs or symptoms were found.Diastolic BP for one participant exceeded current poststroke exercise guidelines, 1 reaching 120 mmHg at the end of the RS 3Min test.

DISCUSSION
The primary finding of this study was that the RS 3Min exercise test conducted after a mTBRS GXT elicited significantly higher peak HR than a standard RS GXT.Thus, the inclusion of the RS 3Min exercise test after a standard protocol GXT could increase test sensitivity for cardiovascular screening.At the same time, the critical power measurement that can be attained from this test without a metabolic cart appears to represent a participant's highest physiologically sustainable exercise intensity.Similar to healthy adults and athletic populations, knowing the critical power (the upper intensity limit of physiologically sustainable exercise) provides valuable information to guide intensity prescription for both continuous and interval exercise. 19,20tential utility of the RS 3Min test for cardiovascular screening The ability to reach maximal levels of HR has been shown to increase GXT test sensitivity with regard to cardiovascular screening in nonstroke populations. 16In persons with stroke, neurological impairments add an additional challenge as a significant portion of this population has difficulty reaching the typical target HR intensity of 85% of AP-HRmax. 6The preliminary results from this study indicate that the addition of the RS 3Min test can be helpful in getting patients up to and to stay above the desired 85% threshold.Nine of our 14 participants were able to exceed the 85% threshold whereas only four and five were able to do so on the RS and treadmill, respectively.In addition, the odds of achieving target HR (85% AP-HRmax) were significantly enhanced when conducting the RS 3Min test relative to the mTBRS GXT (1.8 vs 0.4; adjusted p = .02).
Although American College of Sports Medicine guidelines 8 recognize the variability associated with prediction HRs and accept values +/-10 beats/min as peak effort, the addition of an RS 3Min test in this study enhanced the chances of reaching higher peak HR compared to the RS or treadmill alone.By increasing the potential for patients to achieve the 85% threshold we in turn may have identified a way to enhance the overall test sensitivity for cardiovascular screening in the stroke population.
Our HR results indicate the 3-minute test could be just as effective as currently accepted treadmill GXT stress testing protocols while eliminating many issues associated with the treadmill GXT.Treadmill exercise testing typically elicits higher aerobic intensities than seated modalities among healthy adults, 11 but exercise testing on an RS may be more feasible than treadmill testing for some persons with stroke and in some settings, for example, a person with motor impairments that prevent them from walking or a setting where there is not the necessary safety equipment to perform treadmill walking post stroke.Therefore, it is promising that aerobic intensities for the RS 3Min test were similar to the treadmill GXT in this study.

Feasibility of the RS 3Min test for measuring critical power
To our knowledge this is the first study to address the measurement of critical power in persons of stroke.
Based on test completion rates, trends in power output, and VO 2 responses, our results support the feasibility of the RS 3Min test in measuring critical power in persons after stroke.Although one participant did require a second attempt to complete the RS 3Min test, we believe this was an adjustment to the novelty of vigorous exercise and that they needed psychological acclimation for the level of exertion required.
In accordance with our hypothesis, the RS 3Min test demonstrated an early onset peak power output followed by a rapid decline in power production that concluded with a plateauing of power output.The decline in power output was associated with both the rise in HR along with the decline in stepper cadence.These results fit the expected trend reflecting an early reliance on anaerobic energy followed by a taxing of the anaerobic energy supply with an eventual transition into sustainable aerobic energy sourcing.Although the peak power output of our participants varied widely (210 ± 139 watts), the plateau in power output observed at the end of the RS 3Min test represents the level of aerobically sustainable power output, or critical power. 14,21he critical power designation is further supported by the finding that the end-test power was significantly closer to the minimum power recorded than to the peak power.Without the observed plateau, the designation of a sustainable critical power output would not be valid.Specific knowledge of the critical power has direct influence on the establishment of exercise programming for cardiovascular benefit.
Our VO 2 results further support the validity of the critical power measurements obtained from the RS 3Min test.Twelve of our 15 participants demonstrated a plateauing of their VO 2 response after reaching an early peak.The three individuals who did not meet VO 2 leveling criteria were still exercising at > 81% of their peak capacity from the mTBRS GXT at completion of the 3-minute test.In addition, the VO 2 peak elicited by the RS 3Min test was not significantly different from either of the mTBRS GXT peak or the treadmill GXT peak and therefore less likely to have been a submaximal VO 2 .
Peak blood pressure, RPE, VO 2 , and RER (Table 2) Our measurement of peak systolic BP, RPE, and VO 2 values found no statistical differences across the three tests.The lack of a significant difference in these traditional measures of cardiovascular effort, suggests support for the need to identify a true peak HR and critical velocity.With HR and RPE the most readily available tools for a clinician monitoring exercise prescription intensity, knowing the true attainable peak HR and the critical velocity values could provide enhanced guidelines for establishing an exercise prescription that can maximize the stroke person's cardiovascular adaptations while eliminating reliance upon measurements that may not be as accurate.
The higher RER values seen in the mTBRS GXT (adjusted p = .006)may be the result of having a test that extends further into anaerobic energy production allowing for the additional accumulation of anaerobic debt.The design of the RS 3Min test is opposite in nature in that it starts anaerobically and begins a transition into sustainable aerobic production.The timeline of the test does not allow for an accumulation of anaerobic debt that is seen in the traditional GXT.

Limitations
The primary limitation to this study is that the three exercise tests were performed in the same order (treadmill GXT, mTBRS GXT, RS 3Min test) potentially influencing the results.The participants in this study were a subset of participants that did not allow us to alter testing protocol.A second limitation may be that the mTBRS GXT and RS 3Min were performed on the same day, separated only by a 15-minute rest.It is also important to note that this study did not test for ischemia during the RS 3Min although the aim was to increase HR beyond the GXT.All participants were screened for ischemia during the baseline treadmill stress test.Instead, this study examined the peak HR and BP achieved during the 3-minute test, relative to standard GXTs, to infer its potential utility for cardiovascular screening.This utility will require further evaluation.Measurement of reliability of the RS 3Min test was not conducted and needs future verification.Finally, the lack of ischemia testing during the RS 3Min test coupled with the small sample size limits our declaration of safety with regard to the RS 3Min testing.

CONCLUSION
A novel RS 3Min test appears to be a promising method for cardiovascular screening post stroke, while providing an apparently valid measure of critical power.Adding the RS 3Min test after a standard mTBRS GXT resulted in significantly higher peak HR than the mTBRS GXT alone, which may increase test sensitivity for cardiovascular disease.The critical power measurement obtained from this test may also be a valuable assessment of a participant's highest physiologically sustainable workload and may provide a biologically based, individualized target for effective exercise prescription.Further studies are needed to extend these results to the broader target population in this promising area of research.

F
I G U R E 1 (A) Individual participant and mean peak power outputs related to duration.(B) Individual participant and mean peak VO 2 measurements relative to peak VO 2 from the RS GXT.(C) Individual participant and HR measurements from the RS GXT.(D) Mean peak power and % peak VO 2 without individual participants.GXT, graded exercise testing; HR, heart rate; RS, recumbent stepper.
This research was supported by an award from the University of Cincinnati Neuroscience Institute.Institutional support was provided by the National Center for Research Resources and the National Center for Advancing Translational Sciences, National Institutes of Health (NIH), through Grant 8 UL1 TR000077-04.The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.DISCLOSURE No conflicts of interest, financial or otherwise, are declared by the authors.ORCID Daniel L. Carl https://orcid.org/0000-0002-7807-7417REFERENCES 1. Billinger SA, Arena R, Bernhardt J, et al.Physical activity and exercise recommendations for stroke survivors: A statement for healthcare professionals from the American Heart Results from statistical models presented as mean ± SD or N (%).P values are from omnibus test for a significant difference among protocols.Abbreviations: AP-HRmax, age-predicted maximal heart rate; BP, blood pressure; GXT, graded exercise testing; HR, heart rate; mTBRS, modified total body recumbent stepper; RER, respiratory exchange ratio; RPE, rate of perceived exertion; RS, recumbent stepper.
Note:a One participant was in atrial fibrillation for multiple sessions making HR data unusable.bSignificantlydifferentfrom RS 3Min test after Dunnett adjustment for multiple comparisons.Note: Pairwise comparisons of tests shown in Table2.Estimates are mean differences or odds ratios from statistical models.P values are Dunnett adjusted for multiple comparisons.Abbreviations: AP-HRmax, age-predicted maximal heart rate; BP, blood pressure; GXT, graded exercise testing; HR, heart rate; mTBRS, modified total body recumbent stepper; RER, respiratory exchange ratio; RPE, rate of perceived exertion; RS, recumbent stepper.a One participant was in atrial fibrillation for multiple sessions making HR data unusable.