Reliability and convergent validity of self-reported physical activity questionnaires for people with mental disorders: a systematic review and meta-analysis

Purpose: To examine the reliability and convergent validity of self-reported questionnaires (SRQs) to measure physical activity (PA). Methods: Systematic review with meta-analysis. The validity and reliability of SRQs to assess PA in people with mental disorders (January 20, 2020). Random-effects meta-analyses were performed pooling (1) test–retest correlations or (2) the convergent validity between the SRQs and objective measures (eg, accelerometry). Associations were provided in r values with the 95% confidence interval. Methodological quality was assessed. Results: A total of 9 unique studies (N = 1344; 40.5% females) were included. The authors found a moderate correlation test–retest reliability for PA SRQs in the assessment of vigorous PA (r = .69 [.38 to .85]; P = .001), moderate to vigorous PA (r = .63 [.25 to .84]; P = .003), moderate PA (r = .63 [.39 to .79]; P = .001), and good correlation total PA (r = .75 [.37 to .92]; P = .001). The SRQs have moderate correlations with objective measures for moderate to vigorous PA (r = .25 [.18 to .32]; P = .0001) and total PA (r = .47 [.28 to .62]; P = .005), a poor correlation for moderate PA (r = .18 [.03 to .36]; P = .047), and no correlation with vigorous PA (r = .06 [−.10 to .22]; P = .440). Conclusion: Current evidence indicates that SRQs are reliable over time to assess moderate, vigorous, and total PA levels and valid when assessing moderate PA.