Oxygen cost of walking and its relationship with body composition in multiple sclerosis

This cross-sectional study examined the relationship between the oxygen (O2) cost of walking and body composition metrics, while considering potential covariates such as disability status, step length, and cadence, in persons with multiple sclerosis (MS). The sample included 63 persons with MS across a wide distribution of body mass index (BMI). O2 cost of walking was assessed using portable, indirect calorimetry, and percent body fat (%Fat), fat-free mass (FFM), bone mineral content, bone mineral density (BMD), and weight/FFM were determined from dual-energy x-ray absorptiometry. Other outcome measures included step length, cadence, physical activity, and disability status. The O2 cost of walking had small-to-moderate associations with BMI (rs = –31, p = 0.015), %Fat (rs = –0.26, p = 0.041), and BMD (rs = –0.31, p = 0.013). O2 cost of walking was significantly associated with these outcomes even after controlling for age, sex, disability status, and gait outcomes. The O2 cost of walking was further significantly associated with shorter step length (rs = –0.40, p = 0.001), slower cadence (rs = –0.38, p = 0.002), and higher disability status (rs = 0.44, p < 0.001), but not physical activity. Body composition metrics were not associated with gait parameters, physical activity or disability status in our sample of persons with mild-to-moderate MS. The results indicated that higher O2 cost of walking was associated with lower fat and worse bone health after taking factors such as disability status into consideration. Researchers may focus on interventions that change body composition, or perhaps gait profiles, as possible approaches for changing O2 cost of walking and its consequences such as disability status in persons with MS.


INTRODUCTION
Multiple sclerosis (MS) is a chronic neurological disease characterized by demyelination and transection of axons as well as loss of neurons in the central nervous system with a prevalence of~1 million people in the United States and 2.8 million people worldwide [1,2]. One hallmark feature of MS is walking impairment [3], even in the early stages of the disease [4]. Walking impairment often results from changes in gait, including reduced step length and cadence [5]. Walking impairment might become particularly troublesome in MS when it occurs with poor walking efficiency, as indicated by an increase in the oxygen (O 2 ) cost of walking.
The O 2 cost of walking is defined as the amount of O 2 consumed per kilogram of body weight per unit distance walked, and reflects the contributions of disability-related gait abnormalities and manifestations and its interaction with external constraints [6]. Collectively, there is evidence that the O 2 cost of walking is higher in persons with MS than healthy controls and is linearly associated with worse disability status and perhaps other disease-related consequences [7][8][9]. Persons with MS who have higher O 2 cost of walking further have lower levels of daily physical activity and worse fatigue. Other correlates of higher O 2 cost of walking include reduced cadence, shorter step length, worse spasticity, and lower peak aerobic capacity [8,10,11].
Body composition may be another correlate of the O 2 cost of walking in persons with MS. To date, there is only one published study examining the relationship between weight status and the O 2 cost of walking in persons with MS. That study examined the O 2 cost of walking among persons with MS based on wellestablished body mass index (BMI) categories of normal-weight, overweight, and obese, and reported a trend whereby persons with lower BMI demonstrated higher O 2 cost of walking [12]. The primary limitation of that study was the inclusion of BMI, calculated from height and weight, as a marker of weight status rather than body composition. BMI has advantages of costeffectiveness and ease of administering, but the current categories of BMI significantly underestimate adiposity in persons with MS [13]; this can result in misclassifications of persons with MS in the obese category as non-obese [13]. Dual-energy x-ray absorptiometry (DXA) is considered the gold-standard method of measuring body composition outcomes and can differentiate fat mass, fat-free mass (FFM), and bone mineral content (BMC) of the body [14]. Such a methodology would provide a stronger and more comprehensive understanding of body composition as a correlate of the O 2 cost of walking in persons with MS, and suggest that a specific tissue(s) contributes to walking efficiency. The focus on components of body composition might identify the types of interventions for reducing the O 2 cost of walking. For example, interventions may target FFM differently than fat mass.
The present study examined the relationship between the O 2 cost of walking and body composition metrics, while considering potential covariates such as disability status, step length, cadence, and physical activity based on previous research [8,10]. We expected that persons with MS who demonstrate higher O 2 cost of walking would be significantly related to greater fat mass, lower FFM, and worse bone health based on previous research that examined DXA-based outcomes in persons with MS [15]. This study provides a further step in understanding the relationship among those variables for possibly developing health-promoting lifestyle interventions targeting body composition for improving walking efficiency and quality of life in persons with MS. If this study identifies a body composition outcome as an independent correlate of the O 2 cost of walking, such results would underscore the importance of developing interventions that target body composition profiles for the management of the O 2 cost of walking and its consequences in persons with MS.

MATERIALS/SUBJECTS AND METHODS Participants
Persons with MS were recruited through word-of-mouth and flyers in the community, support groups, and local neurology clinics. We targeted a sample of 60 participants (20 persons of normal-weight [18.5-24.9 kg/m 2 ], overweight [25-29.9 kg/m 2 ], and obese [>30 kg/m 2 ] status) based on self-reported height and weight over the phone. We included BMI and BMI categories as a general strategy of recruitment for yielding variability in DXA body composition metrics; however, BMI was not the focal measure as a correlate of the O 2 cost of walking. Inclusion criteria were as follows: (1) diagnosis of MS; (2) age of 18+ years; (3) no relapse or sudden change in MS symptoms within the last 30 days; (4) ability to ambulate either with or without the use of an assistive device; and (5) no pregnancy. Participants who did not meet these criteria were excluded from the study. Of 127 persons who contacted the research team, 85 persons were assessed for eligibility. We lost contact with 9 persons after screening, and 12 persons were excluded due to the completion of recruitment for specific BMI categories. The remaining 64 persons with MS completed the inperson session; however, one person did not have usable metabolic data based on a recording error with the indirect calorimetry system. The final sample consisted of 63 persons with MS who were included in the data analysis. The sample size was selected based on a power calculation in G*Power 3.1. When designing the study, we estimated the sample size necessary to detect a correlation coefficient (i.e., effect size) as small as 0.26 based on the assumption of alpha of 0.05, beta of 0.8, and twosided test.

Outcome measures
Demographic and clinical characteristics. Participants completed a demographic and clinical characteristics questionnaire regarding age, sex, race, MS type, and disease duration. Neurological status was assessed using the Expanded Disability Status Scale (EDSS) [16], a clinically-administered examination conducted by a research team member who is a Neurostatus-certified assessor. Height and weight were measured using a calibrated stadiometer and scale. Height and weight were used to calculate BMI with the following formula: BMI = weight (kg)/height (m 2 ).
Indirect calorimetry for O 2 cost of walking. Oxygen consumption (VO 2 ) was measured during a 5-min seated rest and 6-min overground walking bout. VO 2 was measured using a portable, indirect calorimetry system (K5, COSMED, Rome, Italy) [17]. We calibrated the spirometer and gas analyzers for the calorimetry system, as per manufacturer's recommendations. We measured VO 2 (ml kg −1 min −1 ) for an estimate of resting energy expenditure, expressed as the average of four, 30-s values over the last two minutes of an initial 5-min period of seated rest (i.e., resting-state VO 2 ). Participants then performed the walking bout around a course with the instructions of walking at normal comfortable pace for 6 minutes [18]; this protocol was based on previous examinations of the O 2 cost of walking [9,11,19]. The course was set up in an open space within the laboratory as an oval that was 31 feet long and 13 feet wide. While wearing the portable calorimetry system for VO 2 measurement, participants refrained from talking during both the 5-min rest and the 6-min period of walking. We measured distance walked using a calibrated measuring wheel (Stanley MW50, New Briton, CT, USA). We measured steady-state VO 2 as the average VO 2 during the last three minutes of the walking bout (i.e., steady-state VO 2 ) [7]. O 2 cost of walking was expressed as ml kg −1 m −1 by dividing steady-state VO 2 in ml kg −1 min −1 by actual walking speed in m min −1 (O 2 cost of walking = (steady-state VO 2 )/speed); we operationalized steadystate VO 2 as the last 3 min of the waking bout based on previous research of O 2 kinetics during a 6-min walking bout in MS [6]; see our data supporting steady-state VO 2 in Fig. 1. We recorded steps taken during the walk using a hand-tally counter for measuring cadence (steps/min), and determining step length based on distance traveled divided by steps taken (distance/step).
Body composition. Body composition values of interest for our analyses were derived from the whole-body DXA scan as per the manufacturer's guidelines. We used a GE Lunar Prodigy Primo DXA scanner (GE Healthcare, Wisconsin, USA) to assess soft tissue and bone composition for the whole body. Outcomes of interest were percent body fat (%Fat), FFM, BMC, bone mineral density (BMD), and weight/FFM of the leg.
Physical activity. The ActiGraph model GT3X + accelerometer was used to measure physical activity. We initialized the accelerometer to collect raw data (g force) at a sample rate of 100 Hz. The device was secured in a pouch on an elastic belt. Participants wore the device on the waist above the nondominant hip during waking hours, except during water-based activities, for a 7-day period. Participants were instructed to record the date and wear time, and we inspected the log during data processing to verify the days of wear time. The data were downloaded using ActiLife software, reintegrated into activity counts in 60-s epochs with low frequency extension, and then scored for time spent in moderate-to-vigorous physical activity (MVPA) based on established cut-points in persons with MS [20] and steps. We considered a day as valid if there was a minimum of 10 h or 600 min of wear time without continuous zeros exceeding 30 minutes, and participants with 1 or more valid were included in the analyses. Outcomes of interest were minutes/day of MVPA and steps/day.

Procedure
We obtained ethical approval to undertake this study from the University Institutional Review Board. We recruited participants and collected data from September 2021 through February 2022, and data were collected in a university-based laboratory. All participants provided signed informed consent. Participants provided demographic and clinical information and then underwent the EDSS as well as a whole-body DXA scan. Upon completion of the scan, participants performed a 6-min walk while wearing the portable, indirect calorimetry system to assess the O 2 cost of walking. The order of assessments per testing session was intentionally standardized across all participants. Participants were sent home with an accelerometer, belt, and daily log as well as a pre-paid, pre-addressed envelope for return service through the United States Postal Service. Participants were remunerated for coming into the laboratory and returning the accelerometer.

Data analysis
All data were analyzed using SPSS version 28.0. Descriptive statistics were used to summarize demographic and clinical characteristics, O 2 cost of walking, body composition outcomes, gait parameters, and physical activity of the sample. Values were presented as mean (standard deviation, SD), unless otherwise specified. Spearman's rank-order bivariate correlations (r s ) estimated the relationships among O 2 cost of walking, body composition outcomes, gait parameters, physical activity, and disability status, as this type of correlation coefficient is unbiased by potential outliers and/or non-normal data. We conducted further partial Spearman's correlations controlling for age, sex, gait parameters, and disability status. The magnitude of correlation coefficients of 0.1, 0.3, and 0.5 were expressed as small, moderate, and large, respectively [21].

RESULTS
The mean (SD) age of the overall sample was 46.7(11.1) years, and the sample consisted mostly of females (73%) and persons with relapsing-remitting MS (92%). The mean disease duration was 12.2(7.1) years with a range of 1-35 years. EDSS scores of the sample ranged between 0 and 6.5 with a median score of 3.0, indicating mild walking disability. The final sample consisted of 56 persons did not use assistive devices, 4 persons who used canes, and 3 persons who used walkers or rollators. Mean height and weight were 166.4(11.0) cm and 80.3(18.6) kg, respectively, yielding a mean BMI of 28.7(6.3) kg m −2 ; this is comparable with the value of 27.5 kg m −2 reported in previous research that examined weight status in persons with MS [12].
The O 2 cost of walking, body composition metrics, and gait outcomes of the sample are presented in Table 1. The value of the O 2 cost of walking in our sample was comparable with values reported by one study that examined O 2 cost of walking across normal-weight, overweight and obese BMI groups in persons with MS [12]. Regarding body composition outcomes, values of %Fat, FFM, BMC, and BMD were consistent with previous research examining DXA-based body composition metrics in persons with MS [15,22].
Step length and cadence were comparable with gaitrelated research in persons with MS [12]. Our sample demonstrated similar amounts of MVPA and steps/day compared to previous research [23,24]. Overall, the characteristics of the sample were typical of MS.
Spearman's rank-order correlations among O 2 cost of walking, body composition metrics, gait outcomes, and disability status in the sample of 63 persons with MS are presented in Table 2. Partial correlations among O 2 cost of walking and body composition metrics, controlling for age, sex, disability status, step length, and cadence, of our sample are presented in Table 3. The O 2 cost of walking was moderately associated with BMI (r s = -0.41, p = 0.002), %Fat (r s = -0.37, p = 0.006), BMC (r s = -0.28, p = 0.041), and BMD (r s = -0.37, p = 0.006) even after controlling for covariates.

DISCUSSION
This study examined the relationship between the O 2 cost of walking and body composition outcomes in persons with MS. The bivariate correlation analysis indicated that higher O 2 cost of walking was significantly associated with lower BMI, percent fat, and BMD, even after controlling for age, sex, disability status, and gait outcomes. Such results might support the design of interventions that target body composition, or gait profiles, for managing the O 2 cost of walking and its possible consequences such as disability status in MS.
There is evidence that the O 2 cost of walking is elevated in persons with MS, and we examined body composition outcomes as potential factors influencing O 2 cost of walking. To date, this is the first study to examine the relationship between body composition outcomes using DXA and the O 2 cost of walking in persons with MS. One study examined O 2 cost of walking across BMI categories in persons with MS and reported a moderate effect size for the difference in the O 2 cost of walking between persons of the normal and obese categories; persons in the obese category had lower O 2 cost of walking than persons in the normal weight category; however, there was no significant difference between the two groups [12]. We extended previous research by using DXA for body composition outcomes, rather than just BMI, for the differentiation between fat mass, FFM, and BMC and recruited persons with a wide range of body composition profiles. The results of our study indicated a similar, stronger pattern, in that higher BMI and adiposity were significantly related to lower O 2 cost of walking. This suggests that DXA-based outcomes do not provide much value beyond BMI, a clinical variable and a proxy of adiposity, in persons with MS. One possible explanation for this negative relationship may be that persons with more adiposity use adaptative or energy conservation mechanisms that minimize energy expenditure while walking at preferred speeds, a strategy that non-obese persons adopt in daily life [25]. This relationship between adiposity and O 2 cost of walking does not necessarily infer that we should not focus on reducing adiposity, but rather future interventions may target body composition for improvements in other consequences associated with excess adiposity (e.g., hormones, intestinal microbiota) [26,27]. Regarding the O 2 cost of walking, other research efforts may possibly examine physiological mechanisms that contribute to the O 2 cost of walking in persons with MS, as higher O 2 cost of walking may further cause early onset of fatigue, hindering the performance of tasks necessary for daily living, and has been linearly associated with disability in persons with MS [9]. Our results indicated that lower adiposity was associated with higher O 2 cost of walking even after controlling for age, sex, disability status, and gait outcomes; this suggests that these factors may not have a significant influence over the relationship between body fat and O 2 cost of walking in persons with MS. The results further indicated stronger associations between lower BMC and BMD and higher O 2 cost of walking after controlling for age, sex, disability status, and gait outcomes, suggesting that this relationship between bone health and walking efficiency is even more important when accounting for the overlapping variance. The results are comparable with previous research reporting that persons who with worse bone health had impaired mobility disability [28,29]. Bone mineral data were not related to EDSS scores, a measure of overall disability, but with O 2 cost of walking, a physiological measure of walking disability, in our sample of persons with MS. If we had included a greater distribution of EDSS scores, perhaps we may have had different results, especially if more persons with more severe disability status take higher doses of steroids and immunosuppressant medications, which may contribute to greater bone loss. Moreover, we examined weight/ FFM of the legs as a proxy of lower-extremity strength/power outcomes, and the association between O 2 cost of walking and weight/FFM became nearly significant (r s = -0.26, p = 0.060).
Walking involves the ability of lower extremities (i.e., muscle strength, power) to move a person's body weight, and there is evidence that lower-extremity muscle strength and power, especially during fast concentric contractions, is markedly reduced in persons with MS compared with controls [30]. This pattern suggests that future interventions may possibly target increasing leg muscle strength/power as a function for improving O 2 cost of walking in persons with MS.
Step O 2 cost of walking was associated with EDSS scores, cadence, and step length in our sample of persons with MS. This is consistent with previous research that has identified disability status and those two gait parameters as significant correlates of the O 2 cost of walking in persons with MS [8][9][10]. Those with worse disability have higher O 2 cost of walking, and those with slower cadence and shorter step length further have lower O 2 cost of walking. We expected that persons with higher disability status would experience greater abnormalities in gait and, therefore, a higher O 2 cost of walking. Of note, the relationship between the O 2 cost of walking and average steps taken per day was nearly significant (r s = -0.24, p = 0.061). Based on the results of previous and current research, the findings support the design of interventions that target gait in persons with MS, especially for persons with more severe disability, as an approach for reducing O 2 cost of walking and managing its potential consequences [8][9][10].
Of note, there were no associations between body composition metrics and gait parameters in our sample of persons with MS who had a wide range of body composition profiles (i.e., normalweight, overweight, and obese categories based on BMI). Other researchers have reported no relationship between BMI and gait parameters [12], and our research replicated and extended those results, in that BMI and body composition outcomes measured using DXA were not associated with either cadence or step length. Moreover, no body composition metrics were related to devicemeasured physical activity, contrary to general knowledge of physical activity, or exercise, as an approach to manage body composition. Our results further indicated no significant association between body composition outcomes and disability status in our sample of persons with mild-to-moderate MS, and this is inconsistent with previous research that reported significantly higher body fat in moderate MS [15]. Despite the variability in body composition metrics, the range of EDSS scores in this study might not have been adequate for examining the association between body composition and the O 2 cost of walking. The majority of persons were independently ambulatory, but perhaps a sample that included more persons who used a cane, walker or rollator for ambulation might have yielded different results.
There are limitations to consider when interpreting the results. The study involved a cross-sectional design, and this precluded inferences regarding causality. A limitation of the study is the lack of reliability data on protocol for this 6-min bout of over-ground walking. Another limitation is that our protocol did not measure O 2 cost of walking across various speeds (i.e., slower, faster). Persons with different body composition profiles may adjust their energy efficiency accordingly when walking at different speeds. Another limitation is the lack of control of additional factors that may influence metabolic processes or walking such as diet, medication, sleep quality, smoking, muscle mass or mitochondria capacity and function. Since resting VO 2 is not the same as resting metabolic rate, a potential research direction may be to examine the relationship between the O 2 cost of walking and true resting metabolic rate. Future research may consider examining these factors as possible correlates or consequences of the O 2 cost of walking, which may inform the development of targeted interventions to manage O 2 cost of walking.
Overall, the study indicated that higher O 2 cost of walking was associated with lower percent fat and worse bone health after taking factors such as disability status into consideration. Elevated O 2 cost of walking was further associated with slower cadence, reduced step length, and worse disability status. We proposed that researchers may focus on interventions that change body composition, or perhaps gait profiles, as possible approaches for changing O 2 cost of walking and its consequences such as disability status in persons with MS.

DATA AVAILABILITY
The data may be made available upon request from the corresponding author. The data are not publicly available in accordance with funding requirements and participant privacy.