Validity of muscular fitness self-assessments in the ecofit smartphone application: A correlation study.

Background: Mobile app-based interventions have the potential for wide-reach and therefore may be a useful tool in up-scaling physical activity interventions. In larger-scale interventions, face-to-face assessments are less cost-effective, and researchers often rely on surveys or activity trackers to assess outcomes. However, there is limited evidence of valid muscular fitness assessments that can be self-administered within mHealth interventions. As such, this study will evaluate the concurrent validity of upper and lower body muscular fitness that have been independently assessed by participants via the ecofit app, and face-to-face assessments conducted by a trained researcher.

Methods: This study compared baseline data from two muscular fitness tests from the ecofit two-armed randomised controlled trial and self-assessed data collected via the ecofit smartphone app (i.e., validated 90-degree push-up and 60-second sit-to-stand test). To assess the concurrent validity, the self-assessed push-up and sit-to-stand tests (i.e., collected via the ecofit app) were correlated using Spearman's correlation coefficient against the researcher-assessed results (i.e., objective results collected during baseline assessment for the ecofit trial). Bland-Altman plots were also used to allow visualisation of the differences between the self- and research-assessed tests.

Results: Participants (N = 54) completed the push-up (24.1%) and sit-to-stand (100%) tests within 14-days of receiving the app. The results revealed a strong significant correlation for the push-up test (0.83, p<0.001) and a moderate significant correlation for the sit-to-stand test (0.63, p<0.001).

Conclusion: This study provides support for the concurrent validity of self-reported upper and lower body muscular fitness assessments (i.e., the push-up and sit-to-stand tests) in mHealth. While these tests may be a feasible option for large scale physical activity interventions, more research is needed to determine the generalisability of these results.