What are the most effective techniques in changing obese individuals' physical activity self-efficacy and behaviour: a systematic review and meta-analysis.

Increasing self-efficacy is generally considered to be an important mediator of the effects of physical activity interventions. A previous review identified which behaviour change techniques (BCTs) were associated with increases in self-efficacy and physical activity for healthy non-obese adults. The aim of the current review was to identify which BCTs increase the self-efficacy and physical activity behaviour of obese adults. A systematic search identified 61 comparisons with obese adults reporting changes in self-efficacy towards engaging in physical activity following interventions. Of those comparisons, 42 also reported changes in physical activity behaviour. All intervention descriptions were coded using Michie et al's (2011) 40 item CALO-RE taxonomy of BCTs. Meta-analysis was conducted with moderator analyses to examine the association between whether or not each BCT was included in interventions, and size of changes in both self-efficacy and physical activity behaviour. Overall, a small effect of the interventions was found on self-efficacy (d = 0.23, 95% confidence interval (CI): 0.16-0.29, p < 0.001) and a medium sized effect on physical activity behaviour (d = 0.50, 95% CI 0.38-0.63, p < 0.001). Four BCTs were significantly associated with positive changes in self-efficacy; 'action planning', 'time management', 'prompt self-monitoring of behavioural outcome' and 'plan social support/social change'. These latter two BCTs were also associated with positive changes in physical activity. An additional 19 BCTs were associated with positive changes in physical activity. The largest effects for physical activity were found where interventions contained 'teach to use prompts/cues', 'prompt practice' or 'prompt rewards contingent on effort or progress towards behaviour'. Overall, a non-significant relationship was found between change in self-efficacy and change in physical activity (Spearman's Rho = -0.18 p = 0.72). In summary, the majority of techniques increased physical activity behaviour, without having discernible effects on self-efficacy. Only two BCTs were associated with positive changes in both physical activity self-efficacy and behaviour. This is in contrast to the earlier review which found a strong relationship between changes in physical activity self-efficacy and behaviour. Mechanisms other than self-efficacy may be more important for increasing the physical activity of obese individuals compared with non-obese individuals.