Do adults with cardiovascular disease risk show meaningful reactivity to physical activity measurement? Coordinated analysis across six studies

Objectives
To estimate the extent of physical activity (PA) measurement reactivity among adults ages 40–60 with risk factors for cardiovascular disease (CVD), to inform best practices for addressing reactivity in PA research and intervention.
Design
Coordinated secondary analysis across six datasets from studies that used 6–7 days of observation following the introduction of PA measurement devices. Moderators of interest were demographic and study design characteristics.
Methods
We included data from participants ages 40–60 with ≥1 CVD factors who provided device-assessed PA behaviour across 6–7 days (N = 1825). We used multilevel modelling to examine participants' PA behaviour (i.e., activity units, steps per day) across days, with decreases in activity indicating reactivity. The threshold for statistical significance was set at p < .05 and standardized effect sizes of interest were semipartial correlation coefficients (srs) ≥.25; we also report conversions to Cohen's d and corresponding equivalence tests.
Results
No patterns met both criteria for significance for either main or moderation effects, including tests of study design features. Results from one small study showed a decrease in steps per day across days of observation (p = .15, sr = .26, d = .23, 90% CI: −.03, .50), though men showed an increase in steps per day (whereas women decreased).
Conclusions
Adults in midlife with CVD risk factors show little evidence of PA measurement reactivity. We recommend continuing to examine PA patterns in individual studies, though widespread use of burdensome procedures to prevent reactivity is not warranted in this at-risk population.