Statistical methods to model and evaluate physical activity programs, using step counts: A systematic review

Abstract

BackgroundPhysical activity reduces the risk of noncommunicable diseases and is therefore an essential component of a healthy lifestyle. Regular engagement in physical activity can produce immediate and long term health benefits. However, physical activity levels are not as high as might be expected. For example, according to the global World Health Organization (WHO) 2017 statistics, more than 80% of the world’s adolescents are insufficiently physically active. In response to this problem, physical activity programs have become popular, with step counts commonly used to measure program performance. Analysing step count data and the statistical modeling of this data is therefore important for evaluating individual and program performance. This study reviews the statistical methods that are used to model and evaluate physical activity programs, using step counts.MethodsAdhering to PRISMA guidelines, this review systematically searched for relevant journal articles which were published between January 2000 and August 2017 in any of three databases (PubMed, PsycINFO and Web of Science). Only the journal articles which used a statistical model in analysing step counts for a healthy sample of participants, enrolled in an intervention involving physical exercise or a physical activity program, were included in this study. In these programs the activities considered were natural elements of everyday life rather than special activity interventions.ResultsThis systematic review was able to identify 78 unique articles describing statistical models for analysing step counts obtained through physical activity programs. General linear models and generalized linear models were the most popular methods used followed by multilevel models, while structural equation modeling was only used for measuring the personal and psychological factors related to step counts. Surprisingly no use was made of time series analysis for analysing step count data. The review also suggested several strategies for the personalisation of physical activity programs.ConclusionsOverall, it appears that the physical activity levels of people involved in such programs vary across individuals depending on psychosocial, demographic, weather and climatic factors. Statistical models can provide a better understanding of the impact of these factors, allowing for the provision of more personalised physical activity programs, which are expected to produce better immediate and long-term outcomes for participants. It is hoped that this review will identify the statistical methods which are most suitable for this purpose.

Leisure but Not Occupational Physical Activity and Sedentary Behavior Associated With Better Health

“…These variables may be related to physical activity level and the health outcome, and therefore, could confound the findings if not controlled for. [34][35][36] General linear models are consistent with the methodology of similar studies 10,37 and is recommended for group analysis. 38 A chi-squared test was used to test independence of LTPA and OPA, and LTSB and OSB.…”

Gamified Text Messaging Contingent on Device-Measured Steps: Randomized Feasibility Study of a Physical Activity Intervention for Cancer Survivors

“…These models included terms for group-by-time interactions with random coefficients for participants nested within pairs. We used a linear growth model with random intercepts and slopes to assess between-group differences in changes in Fitbit-measured step counts during the intervention [39]. As the intervention duration varied among pairs, we modeled each participant's average daily step count as a function of quartiles of the intervention period.…”

Sedentary behaviour, physical activity and psychobiological stress reactivity: A systematic review

“…First, an adapted Downs and Black checklist (Downs and Black, 1998) was used, but only items relevant to this systematic review were retained, as per others (e.g., Plourde et al, 2017). Total scores could range from 0-15, and tertiles were formed to indicate high (0-5), moderate (6-10) and low (>10) risk of bias (e.g., Silva, Jayawardana and Meyer, 2018). Second, the Risk Of Bias In Non-randomized Studies of Interventions (ROBINS-I; Sterne et al, 2016) tool was employed, which comprised items devised specifically for this review (e.g., Slavish and Szabo, 2019).…”

“…Seven studies exclusively tested healthy participants (See Tables 1 and 2), whereas others assessed healthy individuals with parental history of hypertension (Buckworth, Dishman and Cureton, 1994), temporarily abstinent smokers (Taylor and Katomeri, 2006), individuals with moderately elevated psychological distress (Poole et al, 2011) and patients with obstructive sleep apnoea (Ferreira-Silva et al, 2018). Body mass index (BMI) ranged from 22.1 kg/m 2 (Gerber et al, 2017) to 29.0 kg/m 2 (Ferreira- Silva et al, 2018) and two studies tested participants with an overweight BMI (Hong et al, 2004;Ferreira-Silva et al, 2018). The remaining participant characteristics are summarised in Tables 1 and 2.…”