If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Accelerometer measured physical activity and sedentary time in individuals with multiple sclerosis versus age matched controls: A systematic review and meta-analysis
Institute of Clinical Exercise & Health Sciences, School of Science and Sport, University of the West of Scotland, Stephenson Place, Hamilton International Technology Park, South Lanarkshire, Scotland G72 0HL, United Kingdom
Institute of Clinical Exercise & Health Sciences, School of Science and Sport, University of the West of Scotland, Stephenson Place, Hamilton International Technology Park, South Lanarkshire, Scotland G72 0HL, United Kingdom
Institute of Clinical Exercise & Health Sciences, School of Science and Sport, University of the West of Scotland, Stephenson Place, Hamilton International Technology Park, South Lanarkshire, Scotland G72 0HL, United Kingdom
Institute of Clinical Exercise & Health Sciences, School of Science and Sport, University of the West of Scotland, Stephenson Place, Hamilton International Technology Park, South Lanarkshire, Scotland G72 0HL, United Kingdom
People with Multiple Sclerosis (PwMS) find it more difficult to engage in physical activity (PA) than healthy controls. Accelerometers can be used to measure sedentary time and free-living physical activity, understanding the differences between PwMS and controls can help inform changes such as interventions to promote a more active lifestyle. This in turn will help prevent secondary conditions and reduce symptom progression.
Objective
To conduct a systematic review and meta-analysis on accelerometer measured sedentary behavior and physical activity between PwMS and healthy controls.
Methods
A systematic search of five databases (PubMed, Web of Science, Ovid, Science Direct and CINAHIL) from inception until 22nd November 2019. Inclusion criteria was (1) included a group of participants with a definite diagnosis of multiple sclerosis of any type; (2) have 3 or more days of PA monitoring using accelerometers during free living conditions; (3) include age matched healthy controls; (4) assess adults over the age of 18; (5) reported data had to have been reported in a manner suitable for quantitative pooling including: percent of time spent sedentary, minutes per day of sedentary, light, moderate, vigorous activity (moderate and vigorous totaled together), steps per day or counts per day.
Results
Initial search produced 9021 papers, after applying inclusion criteria 21 eligible papers were included in the study. One paper was a longitudinal study from which only baseline data was included. One paper was a reliability and validity study, with data for PwMS versus controls in the validity section. All other papers are cross sectional, with one being a pilot study and another a random control study. One paper used two devices in unison, only one set of data is included in the statistics. Outcome data was available for 1098 participants, 579 PwMS and 519 healthy controls. Significant differences were seen in all categories tested: (1) sedentary time (min/day), standard mean difference -0.286, P = 0.044, n = 4 studies; (2) relative sedentary time (%/day), standard mean difference -0.646, P = 0.000, n = 5 studies; (3) LPA (min/day), standard mean difference 0.337, P = 0.039, n = 5 studies; (4) relative LPA (%/day), standard mean difference 0.211, P = 0.152, n = studies; (5) MVPA (min/day), standard mean difference 0.801, P = 0.000, n = 8 studies; (6) relative MVPA (%/day), mean difference 0.914, P = 0.000, n = 5 studies; (7) step count, standard mean difference 0.894, P = 0.000, n = 8 studies; (8) activity count, standard mean difference 0.693, P = 0.000, n = 13 studies.
Conclusion
PwMS are more sedentary and engage in less LPA, MVPA, steps per day and accelerometer counts per day than healthy controls when measured using accelerometers during free-living conditions.
Multiple sclerosis (MS) is an autoimmune disease characterized by chronic inflammation, oligodendrocyte destruction and demyelination causing lesions throughout the central nervous system (CNS), from which the condition derives its name (‘many scars’) (
). However, the diffuse nature of damage to the CNS means that there is no consistent pattern as to which systems are affected, and consequently symptoms are frequently idiosyncratic, with wide variations in the type of impairments, the degree of impairment and, the rate of decline over time (
). Nevertheless, while the symptoms experienced by people with MS (PwMS) are heterogenous, common symptoms of the condition include muscle spasticity, tremor, impaired motor control, numbness, pain, fatigue, cognitive dysfunction, and depression.
Several of these symptoms make it difficult for individuals with MS to participate in physical activity (PA). Fatigue, will frequently leave PwMS with little energy to engage in PA (
). Moreover, psychological factors such as depression, and impaired cognitive function may make managing daily tasks difficult, and prevent engagement with more active lifestyles (
). Overall, these factors contribute to an increase in time spent being sedentary and reduced time engaged in moderate or vigorous intensity PA (MVPA) resulting in an elevated risk of secondary comorbidities including cardiovascular disease (CVD), stroke and type 2 diabetes (
Early work assessing levels of PA in PwMS used questionnaire and self-reported measures of PA, while more recently, objective assessment using accelerometers has become the dominant technique. Indeed, accelerometry has been validated as a measure of walking performance (
, combined self-reported and objective assessments of PA commenting on effect size when comparing between PwMS and controls, preventing effective pooling of accelerometer only outcomes as the data was not available. Additionally, there are concerns over the accuracy of self-reported measures. Participants have been found to over report their activity levels, especially those with lower levels of fitness (
). Additionally factors such as social desirability (the tendency to keep to cultural ‘norms’) and social approval (the need to obtain a ‘good’ test score (
). Additionally, the increase in studies using objective measures since then, means that an updated review using only objective measures is warranted. More recently a meta-analysis by
assessed objective measures of PA in people with MS and while comprehensive, they chose to compare their data to NAHNES activity data rather than to non-MS control groups. Since the NAHNES data is specific to the US it is not clear if its use as a control variable is appropriate for studies in other locations. Moreover, there have been some criticisms of the validity and reliability of some NAHNES data sets, primarily under reporting of data sets including body mass index (BMI) and total energy expenditure (TEE) among other variables (
assessed remote activity monitoring in a variety of neurological conditions using a variety of activity monitoring devices, including accelerometers, step-counters, and making conclusions about objectively measured PA in PwMS difficult. Moreover, because of these different methods of data collection they were unable to undertake statistical pooling of PA outcomes.
Consequently, there are no current reviews that have compared physical activity levels, sedentary time or step and activity counts of people with MS to healthy controls within the same study, which have used accelerometry. Therefore, the aim of this review is to systematically review the literature regarding objective assessment of sedentary time, MVPA, LPA, step and activity counts in PwMS compared to healthy, matched controls, and to provide quantitative data pooling to determine if differences exist in time spent in different PA domains between PwMS and healthy controls.
2. Methods
This systematic review and meta-analysis followed the reporting guidelines outlined in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement by
a systematic search processes, evaluation, analysis, and reporting was conducted.
2.1 Search strategy
An electronic database search was conducted to procure English language papers comparing accelerometer data from people with multiple sclerosis and controls. Five databases (PubMed, Web of Science, Ovid, Science Direct and CINAHIL) were searched from inception until 22nd November 2019. The search included the keywords: (“multiple sclerosis AND Actigraph OR accelerometer”), (“multiple sclerosis AND physical activity OR sedentary behaviour”), (“multiple sclerosis AND MVPA”), (“multiple sclerosis AND light physical activity”), (multiple sclerosis AND step count”), (“multiple sclerosis AND sedentary time”). Light physical activity was used instead of the abbreviation LPA as this incurred search results relating to lysophosphatidic acid. A second search to find journals relating to sedentary time was also conducted after realizing that sedentary behavior is a term relating to posture, although it has been used frequently in the past to denote sedentary time which is studied in this review. A specific search for Actigraph accelerometers was applied as they are the most frequently utilized monitors on the market for objective physical activity measurements in MS populations (
). A manual search of previously published relevant meta-analysis and systematic reviews was also conducted, as was a review of the reference lists of studies included in this review.
2.2 Inclusion criteria
To be included studies had to: (1) included a group of participants with a definite diagnosis of multiple sclerosis of any type; (2) have 3 or more days of monitoring free living conditions with an accelerometer; (3) include age matched healthy controls; (4) assess adults over the age of 18; (5) reported data had to have been reported in a manner suitable for quantitative pooling including: percent of time spent sedentary, minutes per day of sedentary, light, moderate, vigorous activity (moderate and vigorous totaled together), steps per day or counts per day.
2.3 Study selection
All papers were transferred to a reference manager (Zotero: V 5.0.60, Fairfax, VA, USA). Articles were screened for duplicates. The remaining papers were screened using title, then abstracts. Subsequently, remaining papers were then analyzed by reading the full text identifying relevant studies. Abstract only papers, conference papers and posters were excluded. If there were no results reported of an original study i.e. reviews, secondary analysis or study protocols, they were eliminated. Papers were further excluded if they did not provide accelerometer data, only correlations and other statistical measures.
2.4 Data extraction
Data was extracted and entered in a spreadsheet (Microsoft® Excel 2016, Microsoft Corporation, Redmond, WA, USA). The following fields were collected: MS type, years diagnosed, gender, age, disease severity (using Expanded Disability Status Scale (EDSS) or Patient Determined Disease Steps (PDDS)), intervention duration, objectives, findings, other outcomes, biometric outcomes, and the presence of other cardio metabolic diseases. Accelerometer fields included: outcomes, make/model, cut points used, calibration, position worn, valid days for collection, duration, wear time criteria, and wear time. Accelerometer outcomes were further examined, data procured from the different studies included: percent of time spent sedentary, minutes per day sedentary, light physical activity, moderate PA, vigorous PA, MVPA, steps per day and counts per day.
Studies that provided percent of wear time for the different categories of physical activity were equated to minutes per day by using the percentage from the average daily wear time (min) provided for each group. All other data was converted to give a value per day, weekly data was divided by seven, hourly data was multiplied by hours per day the device was worn. The quality of each study was evaluated using a 20 question appraisal of cross sectional studies form (AXIS) provided by the British Medical Journal (BMJ) (
Meta-analyses were executed using Comprehensive Meta-Analysis (Biostat, V 2.2.064, Englewood, NJ, USA). Pooled data using a random-effects model were used to investigate differences between healthy controls and PwMS. Due to the likely differences in device, wear time, and calibration protocols, studies were assessed using standardized mean differences (SMD) rather than differences in means. Mean, standard deviation and sample size for PwMS and healthy controls for each variable of interest were used to determine overall effect size using a random effects model.
3. Results
3.1 Search results
The search criteria and review process are outlined in Fig. 1. The initial search using the five databases produced 9021 papers. One paper was extracted after reviewing similar systematic review and meta-analysis. After removal of duplicates 5314 papers remained. Initial filtering for inclusion and exclusion criteria via title and abstract resulted in the removal of 5280 papers. Full papers of the remaining 34 articles were assessed with a further 13 removed (two reviews, two secondary analysis, two included children, one did not age match controls, three provided less than three days of data, three papers displayed or collected results in a different format; one only shared five-minute counts, one measured walking speed, and one provided no written results only a graph). Subsequently, 21 eligible papers were included in the study.
Fig. 1The PRISMA flow diagram with numbers of included and excluded articles at each step of the review process.
One paper was a longitudinal study with accelerometer data collected in baseline and week eight. Only the baseline data was extracted for use in this study due to the statistical powers of analyzing the number of participants, also, to ensure the behavioral change of the study did not reflect on the study as all other papers used baseline data (
Recovery of peripheral muscle function from fatiguing exercise and daily physical activity level in patients with multiple sclerosis: a case-control study.
used two devices in unison (Actigraph 7164 and GT3X), only data collected using the 7164 are included in the statistics due to the GT3X being calculated as one axis making comparison to the three axis mode in all other studies using the GT3X model collected in this review balanced.
3.2 Demographic information
Twenty-one papers included in the analysis had a total of 1098 participants, including 519 controls (73.5% female) and 579 people with a definite diagnosis of multiple sclerosis (76.7% female). The mean age of controls was 46.6 ± 10.79 years versus 47.9 ± 9.48 years for PwMS. 15 papers included figures on years since diagnosis, with a resulting mean was 10.9 ± 6.8 years (
Recovery of peripheral muscle function from fatiguing exercise and daily physical activity level in patients with multiple sclerosis: a case-control study.
Recovery of peripheral muscle function from fatiguing exercise and daily physical activity level in patients with multiple sclerosis: a case-control study.
), which was converted to provide BMI figures. The combined BMI average of the 18 papers was 25.49 for controls and 26.09 for PwMS. Thirteen studies reported on EDSS with five providing an average (
Recovery of peripheral muscle function from fatiguing exercise and daily physical activity level in patients with multiple sclerosis: a case-control study.
). All participants were ambulatory with some requiring an assistive device i.e. a cane or walking frame. Demographic information shown in Tables 1 and 2.
Recovery of peripheral muscle function from fatiguing exercise and daily physical activity level in patients with multiple sclerosis: a case-control study.
Recovery of peripheral muscle function from fatiguing exercise and daily physical activity level in patients with multiple sclerosis: a case-control study.
). One study used both the 7164 and the GT3X device simultaneously, however for the purposes of the current analysis the results used in this analysis were of the 7164 (
Recovery of peripheral muscle function from fatiguing exercise and daily physical activity level in patients with multiple sclerosis: a case-control study.
). The Stepwatch, Step Activity Monitor (SAM) which is a pedometer which uses accelerometry, therefore is included in this review was used in two studies (
Recovery of peripheral muscle function from fatiguing exercise and daily physical activity level in patients with multiple sclerosis: a case-control study.
Recovery of peripheral muscle function from fatiguing exercise and daily physical activity level in patients with multiple sclerosis: a case-control study.
Recovery of peripheral muscle function from fatiguing exercise and daily physical activity level in patients with multiple sclerosis: a case-control study.
). Twelve papers included information on minimum valid days of accelerometer data need for inclusion. Four papers accepted a minimum of three days, which is the lowest end of the scale (
Recovery of peripheral muscle function from fatiguing exercise and daily physical activity level in patients with multiple sclerosis: a case-control study.
). Several studies did adopt, state, or reach, the criteria of a valid day as ≥10 h wear time without periods exceeding 60 min of continuous zeroes per day, with at least 3 valid days of wear time as the inclusion criteria in their subsequent analyses (
aired on the more conservative side with an inclusion criteria of ≥8 h per day for all 7 days of monitoring. Furthermore, from the results and reasons for exclusion it appears that other studies did achieve the validated monitoring period although they did not expressly state it in their requirements (
Recovery of peripheral muscle function from fatiguing exercise and daily physical activity level in patients with multiple sclerosis: a case-control study.
). In order to accommodate as many papers as possible into this study and due to a lack of information there is no specified wear time in terms of hours that validate a day or minimum number of days data required, only a requirement for the device to be sent out to participants for a minimum of three days Information shown in Table 3.
MS specific cut points were used but were determined from Multiple Sclerosis Walking Scale – 12, not activity counts. VA- vertical axis, 2D- motion detected in 2 axes, VM – vector magnitude, motion detected in all 3 axes.
Recovery of peripheral muscle function from fatiguing exercise and daily physical activity level in patients with multiple sclerosis: a case-control study.
MS specific cut points were used but were determined from Multiple Sclerosis Walking Scale – 12, not activity counts. VA- vertical axis, 2D- motion detected in 2 axes, VM – vector magnitude, motion detected in all 3 axes.
Recovery of peripheral muscle function from fatiguing exercise and daily physical activity level in patients with multiple sclerosis: a case-control study.
). The average time spent sedentary was 532.13 ± 89.67 min for people with MS and 506.37 ± 81.55 min for controls. This equates to PwMS being sedentary for 25.042 min per day more than their sedentary counterparts. To account for differences in wear time, comparisons were assessed as standardized mean difference (SMD). This equated to an SMD of -0.286 (p = 0.044; Fig. 2).
Fig. 2Forest plot of the comparison of sedentary intensity physical activity in minutes per day between people with multiple sclerosis and healthy participants. Sample size PwMS and Control; Standard Difference in means; Lower limit; Upper limit; p-Value; Standard difference in means and CI: 95% Confidence interval.
00966-X&id=gr1.jpg){kind=link}
BMI calculated from height and weight figures so no standard deviation available.
00966-X&id=gr2.jpg){kind=link}