Benchmarking the effectiveness of interventions to promote physical activity: A metasynthesis.

Abstract

The effectiveness of physical activity interventions is typically evaluated using null hypothesis significance testing or conventional interpretations of effect size (i.e., "small," "medium," or "large"). As these criteria have recently attracted criticism, we conducted a quantitative integration of meta-analyses (a metasynthesis) in order to provide precise, numerical estimates of intervention effectiveness. The research aimed to specify the percentile distribution of effect sizes in meta-analyses of physical activity trials, and their corresponding values for different types of activity (i.e., steps per day, minutes of moderate/vigorous physical activity [MVPA], meeting World Health Organization (WHO) physical activity guidelines). Computerized searches identified 104 meta-analyses incorporating findings from 2,762 trials that met the inclusion criteria for the metasynthesis. The median effect size across all meta-analyses was d+ = .21, equivalent to an increase of 1,320 steps per day, 15.6 additional minutes of daily MVPA, and a 4.3% increase in the proportion of participants meeting WHO guidelines. Separate percentile distributions were computed for different samples (e.g., children, older adults, cancer survivors), measures of physical activity (i.e., objective vs. self-report), settings (e.g., schools, workplace), and follow-up periods (3+, 6+, and 12+ months). Conventional interpretations of effect size gravely misrepresent the effectiveness of interventions to promote physical activity. The percentile values for effect sizes, steps per days, minutes of MVPA, and percentage meeting guidelines reported here can be used to benchmark the effectiveness of future trials and should enable more informed judgments about trade-offs between effectiveness and considerations such as reach, burden, and cost. (PsycInfo Database Record (c) 2021 APA, all rights reserved).