How can progress in research on health behavior change be accelerated? Experimental medicine (EM) offers an approach that can help investigators specify the research questions that need to be addressed and the evidence needed to test those questions. Whereas current research draws predominantly on multiple overlapping theories resting largely on correlational evidence, the EM approach emphasizes experimental tests of targets or mechanisms of change and programmatic research on which targets change health behaviors and which techniques change those targets. There is evidence that engaging particular targets promotes behavior change; however, systematic studies are needed to identify and validate targets and to discover when and how targets are best engaged. The EM approach promises progress in answering the key question that will enable the science of health behavior change to improve public health: What strategies are effective in promoting behavior change, for whom, and under what circumstances?

We conducted a meta-analysis of randomized controlled trials (RCTs) to promote health behavior change based on self-determination theory (SDT). The review aimed to (a) quantify the impact of SDT interventions on health behaviors, (b) test mediation by theoretically specified variables (autonomous motivation and perceived competence), and (c) identify moderators of intervention effectiveness. Computerized searches and additional strategies identified 56 articles that yielded 65 independent tests of SDT interventions. Random effects meta-analysis and metaregressions were conducted via STATA; meta-analytic structural equation modeling (MASEM) was used to test mediation. The sample-weighted average effect size for SDT interventions was d+ = .23, and there were significant effects for physical activity, sedentary behavior, diet, alcohol consumption, and smoking cessation (.16 ≥ d+ ≥ .29). Effect sizes exhibited both publication bias and small sample bias but remained significantly different from zero, albeit of smaller magnitude, after correction for bias (d+ ≥ .15). MASEM indicated that autonomous motivation and perceived competence mediated intervention effects on behavior. Metaregression analyses indicated that features of the sample, intervention, or methodology generally did not moderate effect sizes. The present review indicates that SDT interventions have a significant but small effect on health behavior change and suggests several directions for future research. (PsycInfo Database Record (c) 2020 APA, all rights reserved).

Although risk perception is a key concept in many health behavior theories, little research has explicitly tested when risk perception predicts motivation to take protective action against a health threat (protection motivation). The present study tackled this question by (a) adopting a multidimensional model of risk perception that comprises deliberative, affective, and experiential components (the TRIRISK model), and (b) taking a person-by-situation approach. We leveraged a highly intensive within-subjects paradigm to test features of the health threat (i.e., perceived severity) and individual differences (e.g., emotion reappraisal) as moderators of the relationship between the three types of risk perception and protection motivation in a within-subjects design. Multi-level modeling of 2968 observations (32 health threats across 94 participants) showed interactions among the TRIRISK components and moderation both by person-level and situational factors. For instance, affective risk perception better predicted protection motivation when deliberative risk perception was high, when the threat was less severe, and among participants who engage less in emotional reappraisal. These findings support the TRIRISK model and offer new insights into when risk perceptions predict protection motivation.

Despite abundant observational and experimental tests, it is not yet clear whether enhancing autonomous motivation or perceived competence leads to health behavior change. We identified interventions that aimed to change these constructs and quantified the magnitude of changes in behavior observed when interventions generated increases in autonomous motivation, perceived competence, or both. Computerized searches and additional strategies identified 67 articles that yielded 135 effect sizes relevant to our research questions. Random effects meta-analyses were conducted via STATA. Interventions increased autonomous motivation and perceived competence in 31% and 38% of tests, respectively. Increasing autonomous motivation led to a medium change in health behaviors (d+ = .47, 95% CI [.44, .83]) and increasing perceived competence generated a small-to-medium change (d+ = .34, 95% CI [.22, .47]). Interventions that failed to generate significant improvements in autonomous motivation and perceived competence had much smaller effects on behavior change (d+ = .13 and .10, respectively). There was little evidence of synergistic effects. Changing both autonomous motivation and perceived competence (d+ = .42) did not lead to a larger effect on behavior compared with changing autonomous motivation on its own (d+ = .61), but had a larger effect compared with changing perceived competence on its own (d+ = .21). The present review suggests that autonomous motivation and perceived competence are valid targets for interventions to promote health behavior change but also indicates that research is warranted to ensure that interventions more effectively engage these targets. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

The current meta-analysis estimated the magnitude of the impact of asking intention and self-prediction questions on rates of subsequent behavior, and examined mediators and moderators of this question–behavior effect (QBE). Random-effects meta-analysis on 116 published tests of the effect indicated that intention/prediction questions have a small positive effect on behavior ( d₊ = 0.24). Little support was observed for attitude accessibility, cognitive dissonance, behavioral simulation, or processing fluency explanations of the QBE. Multivariate analyses indicated significant effects of social desirability of behavior/behavior domain (larger effects for more desirable and less risky behaviors), difficulty of behavior (larger effects for easy-to-perform behaviors), and sample type (larger effects among student samples). Although this review controls for co-occurrence of moderators in multivariate analyses, future primary research should systematically vary moderators in fully factorial designs. Further primary research is also needed to unravel the mechanisms underlying different variants of the QBE.