Are we making progress? Testing whether interventions to promote key health behaviors improved over the past 30 years

Conceptual and methodological developments could contribute to increasing efficacy of health behavior change interventions over time. We undertook a meta-synthesis to determine whether and to what extent intervention effects have improved during the past 30+ years.

We retrieved data from 10 meta-analyses of interventions for physical activity, diet, weight loss, alcohol consumption, and smoking cessation (total k = 1,701). Random effects meta-regression analysis was used to test associations between the year of publication of the primary study and the effect size observed for the intervention.

Year of publication was not related to effect size in physical activity, weight loss, or smoking cessation interventions (ps > 0.05). Associations for dietary interventions (B = -0.039, p = .001) and for digital and brief interventions to reduce alcohol consumption (B = 6.35 and 2.32, respectively, p < .001) indicated that efficacy was lower in more recent compared to earlier trials. Controlling for potential confounds (e.g., risk of bias, intervention dosage, sample characteristics) did not alter these results.

Findings from the present meta-synthesis indicate no improvement over time in the efficacy of interventions to promote health behavior change across 5 domains. Rapid changes in products and contexts, the paucity of behavior change trials, and evaluation practices that focus on the efficacy of individual trials rather than cumulative efficacy across trials offer potential explanations of the findings. Innovations are needed to improve intervention efficacy in future trials.