Forty-six studies (4,765 participants) were included in the review.
There was a statistically significant positive effect of interventions on physical activity behaviour (ES 0.54, 95% CI 0.39 to 0.69; 46 studies). There was significant heterogeneity (I2=83%) and evidence of publication bias for this analysis. When the two studies contributing to publication bias were removed from the analysis, the result remained statistically significant and heterogeneity remained significant (I2=76%).
Moderator analyses indicated that the number of steps (nine comparisons) and "other PA indexes" (six comparisons) showed larger effect sizes than energy expenditure (14 comparisons), cardiovascular health (seven comparisons), length of physical activity (15 comparisons) and physical test performance (11 comparisons).
Analyses indicated that shorter intervention periods (less than six months) had greater effects than longer periods. Neither the number nor frequency of sessions (each 37 comparisons) had any significant effect on the results. Analysis also showed that there were no significant changes in the level of physical activity of participants from the end of the intervention to the follow-up assessment (nine studies).