What makes implementation intention interventions effective for promoting healthy eating behaviours? A meta-regression.

This study examines the efficacy of implementation intentions (II), a widely used self-regulatory strategy to help people achieve their goals. Although previous research has shown that the effect of II interventions is significantly higher in promoting healthy eating behaviours than in diminishing unhealthy eating behaviours, the factors that can moderate the effectiveness of these interventions remain unclear. In a meta-analysis of 70 interventions (N = 9689), we confirmed that II interventions for healthy eating behaviours yielded a medium significant effect size (d = 0.33) and a low significant effect size for unhealthy eating behaviors (d = 0.18). We show that the moderator variables of II interventions for healthy and unhealthy eating goals are very different. Regarding healthy eating, since moderator variables explain 53% of the variance in the heterogeneity of the effect sizes, the present study helps in gaining an understanding of the previous inconsistent results and offers suggestions for designing more efficient interventions. Effect size was negatively predicted by age, indicating that for younger people the effect size is higher, and II check, showing that if the instructor checks the plan it decreases its efficacy. Moreover, the effect of II interventions on students is significantly smaller than in non-student samples. In contrast, the effect size was positively predicted by initial training, off-line delivered interventions and, specific if-then and action plans versus complex plans. For unhealthy eating behaviours, our results show that there is less room to improve the intervention; there is only one moderator variable (plan formulation), and the heterogeneity found in the studies is lower for unhealthy eating behaviours (I2 = 46.70%) than for healthy eating behaviours (I2 = 73.25%), indicating that the intervention has low efficacy regardless of the design of the intervention.