RESUMEN
Objective: A systematic review of health interventions using avatars (N = 18) was conducted to provide comprehensive knowledge of the effectiveness of using avatars to promote healthy behaviors, specifically in relationship to healthy eating and exercising. Materials and Methods: Two researchers identified field or laboratory studies that had quantified study results, which were published in peer-reviewed journals in English from January 2000 to March 2019. Databases (PsychInfo, PubMed, and Web of Science), forward reference, and manual searches were used to identify the studies. Search terms included avatar, the Proteus effect, exercise, and diet, among others. Two field interventions and 16 laboratory studies were identified. Information on sample characteristics, technologies used, study design and conditions pertaining to avatars, outcome measures, results, and conclusion were extracted. Results: Six different avatar characteristics used to elicit health outcomes were identified, which are the similarity with the user, avatar body size, self-domain (e.g., ideal-self), customizability, body transformation, and avatar's behaviors. Only a few studies had a no-avatar control group; thus, it was not possible to conclude whether employing avatars in health interventions increases the effectiveness of the interventions in comparison to not using an avatar. Conclusion: The results indicate that using an avatar that is physically active, fit, and similar-looking (to the user) is effective in eliciting healthy behaviors.