The Right Tool for the Task: Body-Weight Supported Treadmill or Total Body Recumbent Stepper for Mobility-Adapted Cardiopulmonary Exercise Testing in Multiple Sclerosis Patients with Disability

INTRODUCTION

Multiple sclerosis (MS) is the most prevalent neurological disorder among young and middle-aged people [1]. Adopting an inactive lifestyle [2] results in inactivity-related health problems among persons with MS (PwMS), such as osteoporosis, cardiovascular disease, fatigue, and depression [3]. Aerobic capacity is a crucial health [4] and performance marker [5] [6], and PwMS exhibit lower values of maximal aerobic power when compared to age and sex-matched healthy adults [7]. Importantly, lower risk of cardiovascular disease [6], better walking performance [8], faster cognitive processing speed [9], and enhanced neuroprotection and brain health [10], have all been associated with greater aerobic capacity in MS [6]. Testing aerobic fitness accurately, as a metric to calculate personalized exercise intensity and as a study outcome measure, is recommended [6].

The gold standard assessment of aerobic fitness measures the maximum rate of oxygen consumption (V̇O_2max), in relation to the rate of carbon dioxide (V̇CO₂) produced during a progressive maximal cardiopulmonary exercise test (CPET). CPET is feasible and safe in PwMS [11-13]. While several different exercise testing modalities (e.g., treadmill, bicycle ergometer, recumbent stepper) have been used to assess aerobic capacity, there is no evidence-based consensus regarding the best instrument that should preferentially be used in people having mobility disability [13].

Weakness and fatigue of the limbs can limit the ability to reach maximum aerobic capacity during a CPET [14, 15]. Additionally, leg weakness and fear of falling could impede maximal effort, even when using treadmills with overhead harness support [16]. Although yet to be determined, modalities that permit the user to exercise in a seated position, such as total body recumbent steppers (TBRS), may permit CPET metrics that represent true cardiopulmonary capacity, less influenced by leg impairments [17]. Obtaining stable and representative CPET values would be especially useful in longitudinal studies, as there may be an emergence of MS-related relapses or accumulation of disability within an individual.

Investigators have tested several adapted and non-adapted methods for CPET among PwMS, including bicycle ergometers, arm ergometers, and TBRS [11, 18]. For instance, Pilutti and colleagues compared arm ergometers to TBRS, and reported higher peak aerobic capacity obtained with TBRS, especially among people with low levels of disability and higher fitness levels [19]. Although useful in determining optimal CPET methodology, the authors did not provide comprehensive reporting of criteria necessary for V̇O_2max [19, 20], so whether maximum CPET was achieved cannot be assured. The researchers recommended the use of a TBRS over the arm ergometer for CPET in PwMS. Similarly, evidence from our laboratory, testing two common adapted CPET methods (TBRS and Body Weight Supported Treadmill [BWST]) suggested less fatigue of the soleus muscle when the workload was distributed between four limbs (using a TBRS) rather than two (using a BWST) [21].

We aimed to compare CPET metrics using two adapted modalities (BWST and TBRS) among PwMS and matched controls. Metabolic parameters (V̇O_2max, % normative values for V̇O_2max [% V̇O_2max], heart rate maximum [HR_max], and age-predicted HR_max) were obtained along with the ability of participants to meet criteria for achieving a maximum CPET[22]. We also aimed to determine, in a preliminary way, whether the level of disability influenced the capacity to achieve the indicators of V̇O_2max. We hypothesized that: a) Controls would achieve similar, if not higher, V̇O_2max on the BWST and TBRS; b) PwMS would achieve higher values and be more likely to meet V̇O_2max criteria on TBRS, especially for those having greater mobility disability.

METHODS

RESULTS

Metabolic Parameters

V̇O_2max

As expected, the Control Group’s V̇O_2max was significantly higher than PwMS overall, as a main effect was observed for the group (MS vs. Control) with a large effect size (F _{(1, 15)} = 10.0, p = 0.006, µ² = 0.4; table 3 and figure 1). Although the PwMS achieved similar V̇O_2max using BWST and TBRS, the Control group’s V̇O_2max was higher on the BWST than when using the TBRS (Interaction effect, F _(1,15) = 19.3, p < 0.001, µ² = 0.5) (figure 1). PwMS had lower V̇O_2max than Controls on the BWST (24.2±7.8 vs. 40.2±7.6 mL·min^-1·kg^-1) but not on the TBRS (26.5±8.7 vs. 34.3±7.1 mL·min^-1·kg^-1). A preliminary analysis suggested differences between the Controls and MS disability subgroups were more apparent when using BWST, likely because the Controls obtained higher V̇O_2max on BWST. (table 3; figure 2). After V̇O_2max values were transformed into % normative values for age and sex (table 3, figures 3 & 4), similar results were observed. Controls achieved higher values than PwMS (F_(1,15) = 19.2, µ² = 0.56, p < 0.001). Controls obtained 83% normative values using BWST, significantly higher than TBRS (62%; F _(1,15) = 13.9, p = 0.002, µ² = 0.48). PwMS performed similarly on the BWST (18%) versus the TBRS (28%), both significantly below the BWST and TBRS obtained from Controls (figure 3). When examining the MS disability subgroups, Controls were higher than both EDSS ≤ 2 (p = 0.001) and EDSS > 2 groups (p < 0.001) on BWST but not using TBRS (figure 4).

• Download figure
• Open in new tab

Figure 1

V̇O_2max in Multiple Sclerosis and Control (mean ± SD) when both groups performed CPET on BWST and TBRS. The Control group achieved a significantly higher V̇O_2max on the BWST (#; p < 0.001); Multiple Sclerosis participants had a significantly lower V̇O_2max than the Control on BWST (*; p < 0.001), but not on the TBRS.

• Download figure
• Open in new tab

Figure 2

V̇O_2max differences (mean ± SD) between the two Multiple Sclerosis-severity participants (EDSS ≤ 2, EDSS ˃ 2) and Control. Participants with higher Multiple Sclerosis severity (EDSS > 2) were significantly lower than the Control, only on BWST (#, p < 0.001, 95% C.I. = 11.1, 31.7). Multiple Sclerosis participants with lower severity (EDSS ≤ 2) were also lower than the Control only on BWST (*, p = 0.04, 95% C.I. = 0.27, 20.9).

• Download figure
• Open in new tab

Figure 3

%V̇O_2max (% normative values) in Multiple Sclerosis and Control; the Control group was significantly higher on BWST than TBRS (#, p = 0.002). MS was lower than the Control on both BWST (**, p < 0.001) and TBRS (*, p = 0.02) (F _{(1, 15)} = 13.9, p = 0.002, µ² = 0.48).

• Download figure
• Open in new tab

Figure 4

V̇O_2max (% normative values) differences between the groups of Multiple Sclerosis participants (EDSS ≤ 2; EDSS >2) and the Control. The Multiple Sclerosis participants with higher Multiple Sclerosis severity (EDSS >2) had significantly lower values than the Control, only on BWST (**, p < 0.001, 95% C.I. = 49.2, 109.5). Also, Multiple Sclerosis participants with lower Multiple Sclerosis severity (EDSS ≤ 2) had significantly lower values than the Control only on BWST (*, p = 0.001, 95% C.I. = 20.4, 80.7).

View this table:

• View inline
• View popup

Table 3 Summary of Metabolic Parameters and Criteria for Achieving V̇O_2max on TBRS and BWST

HR_maxand age-predicted HR_max

In terms of HR_max values, there were no differences between MS and Controls (Group, F _(1,15) = 3.46, p = 0.08, µ² = 0.18), between modalities (F _(1,15) = 0.57, p = 0.4, µ² = 0.03), or interaction effect (F _(1,15) = 1.64, p = 0.2, µ² = 0.09). The mean values of HR_max were only different between Control and MS subgroup EDSS > 2 (p = 0.002, 95% C.I. = 16.6, 73.6) on BWST but not on TBRS (F _(2,14) = 3.0, p = 0.08, µ² = 0.3) (table 2). In terms of age-predicted HR_max,, both PwMS and Controls reached 90% or greater (table 3), ([F _{(1, 15)} = 2.23, p = 0.1, µ² = 0.13)], with no differences between the modalities [(TBRS vs, BWST), (F _{(1, 15)} = 0.49, p = 0.4, µ² = 0.03)], or interaction effect [(F _{(1, 15)} = 1.75, p = 0.2, µ² = 0.1]). Also, there were no differences between the PwMS subgroups: EDSS ≤ 2, EDSS > 2, and Control on either of BWST (F _{(2, 14)} = 3.26, p = 0.06, µ² = 0.31), or TBRS (F _{(2, 14)} = 0.63, p = 0.5, µ² = 0.08).

Criteria for Achieving a Maximal Cardiorespiratory Fitness Test

Overall, all Controls achieved the criteria for a V̇O_2max test using both the TBRS and BWST (table 4). Notably, all Control participants reached their predicted HR_max using BWST, and 86% did so using the TBRS. All Controls reached volitional exhaustion using both modalities. Lastly, 57% reached a VO₂ plateau using the BWST, whereas 14% reached a VO₂ plateau using the TBRS. Five of the 10 PwMS were unable to continue on the BWST (due to leg weakness, pain or fatigue) which precluded the ability to determine whether VO₂ plateau had been reached (table 4). 90% of PwMS achieved the criteria for a V̇O_2max test (i.e., two or more of the individual criterion) using the TBRS, whereas 60% achieved the criteria using the BWST. The RER criterion was reached by 60% of PwMS using the TBRS and 20% using the BWST. In the MS subgroup with greater disability, more CPET criteria were reached using TBRS. These 5 participants reached 13/20 criteria on TBRS and only 5/20 on BWST.

View this table:

• View inline
• View popup
• Download powerpoint

Table 4 Criteria to satisfy the achievement of maximal aerobic capacity-

DISCUSSION

We undertook this study to optimize methods for achieving maximal CPET among PwMS with and without mobility disability, especially in longitudinal studies in which abilities, particularly walking ability, can change over time. We compared standard parameters between TBRS and BWST, including V̇O_2max, HR_max, age- and sex-predicted %V̇O_2max, RER, RPE reached, age-predicted HR_max, and time to test completion. We also examined the criteria for reaching a maximal VO₂ between modalities, including VO₂ plateau (primary criterion), RER of at least 1.1, achieving at least 10% of age-predicted HR_max, and a minimum of 7/10 RPE scale.

We report three key findings. First, contrary to our hypotheses, Controls achieved higher V̇O_2max and %V̇O_2max using BWST compared to TBRS. We also expected that PwMS would achieve higher values on TBRS but in fact, PwMS performed similarly on both devices (figures 1 and 3). This likely contributed to significantly higher V̇O_2max in Controls compared to PwMS only on BWST and not TBRS. Secondly, while performing BWST, CPET was compromised among PwMS since half could not continue walking due to leg symptoms, which prevented the determination of VO₂ plateau (table 4). As for PwMS, 9/10 achieved the criteria for V̇O_2max (i.e., two or more of the individual criterion) using the TBRS, whereas 6/10 achieved criteria using the BWST. All Controls obtained criteria for a maximum CPET on both devices (table 4). Finally, a preliminary analysis of an MS subgroup having higher levels of mobility disability (n=5) suggested that they achieved most of the criteria for a maximum CPET using TBRS (13/20 criteria achieved (five participants * four criteria) compared to CPET using BWST (5/20 criteria achieved; table 4). Inability to meet criteria for a maximum CPET using BWST in participants having higher mobility disability likely contributed to significant differences between Controls and the PwMS having EDSS > 2 (figures 2 and 4).

Cycling ergometers and treadmills serve as gold standards for CPETs in healthy subjects [33], however, the recumbent seated stepper has been suggested as an ideal device for PwMS [34]. CPET is complicated among persons with mobility disability because lower values [6] could be explained by either cardiopulmonary deconditioning or limb disability [35]. This duality is important because the CPET is meant to measure cardiopulmonary capacity rather than leg endurance. Not all adapted CPET devices perform equally, even in apparently healthy subjects. For instance, lower V̇O_2max on TBRS compared to BWST, evident in our Controls, was also observed among healthy highly active participants [30]. Previous researchers suggested that the support provided by the TBRS seat may localize fatigue to the upper and lower limbs accounting for a 5-20% reduction in V̇O_2max values on TBRS [36]. Similarly, our results point to achievement of higher V̇O_2max and predicted V̇O_2max using BWST compared to TBRS in apparently healthy controls. Controls had little difficulty meeting criteria for CPET, the same could not be said for PwMS. They were more likely to reach criteria for a CPET on the TBRS which likely provided a more representative V̇O_2max value. We propose, although the PwMS performed similarly on both devices (figures 1 and 3), the TBRS more reliably represented true V̇O_2max.

Few studies disclose whether persons with neurological disability meet predefined criteria to satisfy reaching V̇O_2max [37, 38]. In one study, Mackay-Lyons and Makrides [20] reported CPET criteria performed 1-month post-stroke with 76% of patients achieving one or more of the V̇O_2max criteria on BWST [20]. Although 62% of the stroke patients achieved RER criteria on BWST [20], only 20% did in our patient sample. Similarly, there is a lack of information related to achieving CPET criteria in PwMS. Pilutti and group [19] compared TBRS and arm ergometry (EDSS=3.0), reporting that participants achieved similar V̇O_2max on both devices (25.2±6.8 vs. 22.5±10.1 mL*min^-1kg^-1) [19], whether participants achieved the necessary criteria for V̇O_2max was not reported.

The plateau in VO₂ stands out as a primary criterion in the measurement of aerobic capacity [39-41], and is the best evidence that a true V̇O_2max is achieved [42]. However, VO₂ plateau (≤150 ml/min^-1) has been criticized because of the lack of theoretical and statistical basis, as well as its insufficient specificity to the testing protocols [43]. A significant variability in the percentage of subjects who showed a plateau in VO₂ has also been reported [44]. In this study, achieving VO₂ plateau was the most challenging criterion to meet using BWST or TBRS for both PwMS and Controls. Still, 30% of our PwMS achieved the VO₂ plateau on BWST; more than the 17% reported previously in stroke [20] and similar to 34% of stroke patients tested on a semi-recumbent cycle ergometer [45]. Notably, 40% of our PwMS achieved this criteria using TBRS. Some authors propose concerns about falling may prevent PwMS from pushing themselves hard enough on the BWST [16]. With greater muscle fatigue on BWST [21], TBRS may be a better option for PwMS when conducting longitudinal studies. With challenges like the participants’ physical fitness status and the protocols used influencing the incidence of VO₂ plateaus [40, 46], further investigations need to standardize the VO₂ plateau achievement. However, based on our results, using TBRS likely represents a true VO₂ in PwMS (and potentially other groups experiencing mobility disability) based on VO₂ plateau results.

Time to complete a CPET should be between 8 and 12 minutes [22]. Most PwMS and all but one Control finished the test within 13 and 17 minutes, respectively, using the TBRS. The time ranged from 13 to 35 minutes using BWST across both groups, suggesting that TBRS was more efficient at obtaining V̇O_2max. Overall, older individuals or patients may require a longer time to achieve V̇O_2max than healthy, trained, or active subjects [47]. Grover et al. reported that PwMS had lower peak plantarflexor torque following 30-minute BWST exercise compared to the same intensity and duration using TBRS [21]. Participants in Grover’s study also took a longer time to achieve their peak torques on BWST [21] which may relate to leg fatigue using BWST. Leg fatigue may limit the patient’s perception of cardiopulmonary exertion on BWST [12, 16] lengthening the time taken to achieve V̇O_2max on BWST compared to the recommended 8-12 minutes by Mezzani [48].

STUDY LIMITATIONS

We provided a thorough reporting of CPET metrics in Controls and PwMS with different degrees of mobility disability. Although we provide comprehensive CPET data, the sample is small, so analyses are exploratory. Although the time to complete a CPET has been suggested to be 8-12 minutes, PwMS exercised about 9 minutes longer on BWST compared to TBRS. Therefore, fatigue may be considered as a limiting factor before cardiorespiratory exhaustion.

CONCLUSIONS

PwMS achieved similar V̇O_2max and %V̇O_2max on the BWST and TBRS, but Controls achieved higher values using BWST. Determination of VO₂ plateau -the main criteria- was more challenging on BWST for both groups, predictable in PwMS, probably due to MS symptoms in their legs limiting their ability to continue walking. Also, a higher percentage of PwMS achieved overall criteria for V̇O_2max using TBRS. All Controls achieved overall criteria on both modalities. Lastly, PwMS subgroup having higher levels of mobility disability, EDSS > 2 achieved most of the criteria for a maximum CPET using TBRS. During the preliminary investigation of PwMS subgroups with lower and higher mobility disability, CPET using BWST exaggerated already lower CPET metrics in persons with mobility disability. Based on our results, using TBRS more likely represents a true VO₂ in PwMS (and potentially other groups experiencing mobility disability).

HIGHLIGHTS

• Testing aerobic fitness accurately is crucial in persons with multiple sclerosis.

• There is no evidence-based consensus regarding the best testing modality that should preferentially be used in patients with mobility disability.

• Comparing two modalities (body weight-supported treadmill vs. total body recumbent stepper) showed that patients achieve more of the criteria required to satisfy an actual maximal aerobic capacity on the stepper.

• Considering the leg symptoms in MS patients, testing aerobic capacity using a stepper is suggested.

DECLARATION OF INTEREST

The authors declared no potential conflicts of interest related to the research, authorship, and publication.

FUNDING

This research was supported by MS Canada, Grant#2631 (MP); Canada Research Chairs Program Grant #950-232532 (MP) and Canada Foundation for Innovation Grant #211716 (MP).