The algorithm of dissatisfaction: social media and body image crisis among university students

INTRODUCTION: To map and critically synthesize evidence (2020-2025) on the relationship between social media use/exposure and body image among university students, we conducted a scoping review guided by Arksey and O'Malley's methodological framework and reported according to PRISMA-ScR. A descriptive numerical summary and qualitative thematic synthesis were produced; meta-analysis was not conducted due to heterogeneity in study designs and outcome measures. METHODOLOGY: Seven databases (PubMed/MEDLINE, PsycINFO, Embase, CINAHL, Scopus, Web of Science, and Google Scholar) were searched using controlled vocabulary and free-text terms related to body image constructs, social media platforms/content, and higher education populations. Eligible peer-reviewed studies (quantitative, qualitative, mixed-methods, and systematic/scoping reviews) included samples with ≥ 70% university students and assessed a body image outcome in relation to social media exposure. RESULTS: A total of 81 studies (N = 19,947; ages predominantly 18-25; ~70% women) met inclusion criteria: 55.6% were cross-sectional, 18.5% experimental, 14.8% mixed-methods, and 11.1% longitudinal or EMA-based. DISCUSSION: Six non-exclusive thematic groups were identified: [1] body dissatisfaction/shame/low physical self-esteem (63.0%); [2] eating disorders and weight control behaviours (32.1%); [3] cosmetic procedure seeking and probable body dysmorphic disorder (BDD) (11.1%); [4] protective/positive content (body positivity/neutrality, self-compassion), including brief "social media breaks" (16.0%); [5] psychological mediators/moderators (including self-objectification) (33.3%); and (6) methodological contributions (e.g., EMA designs, longitudinal and experimental approaches, and social media-specific measurement advances) (3.7%). Across all designs, the emotional the emotional valence, degree of idealization, and appearance-focused or stigmatizing nature of contentconsistently predicted outcomes more strongly than time-based exposure metrics. Brief, theory-based, digitally delivered interventions showed modest short-term improvements, although questions remain about their scalability and long-term efficacy. CONCLUSION: Despite methodological advances, the evidence base is still constrained by cross-sectional self-report data, varied measurement tools, and underrepresentation of male and gender-diverse populations. Universities are encouraged to incorporate media literacy, comparison reappraisal, and self-compassion into student wellbeing programs, while future research should prioritize harmonized effect size reporting, mechanism-focused trials, and ethically governed digital trace methodologies.