A social cognitive theory and theory of planned behavior–based serial multiple mediation model of social media–related factors and eating behaviors among sports science students

Background/objectives Social media has become a pervasive digital environment closely associated with individuals’ dietary preferences, eating attitudes, and consumption behaviors. Although prior research has documented associations between social media use and eating behaviors, the underlying cognitive, attitudinal, and psychological mechanisms remain insufficiently clarified, particularly among university students. Accordingly, this study examines the serial mediating roles of digital healthy nutrition literacy, attitudes toward functional foods, food craving acceptance, and emotional eating in the relationship between social media use and eating behaviors. Methods This cross-sectional study was conducted with 1,576 sports sciences students from 38 public universities in Türkiye. Sports science students were selected because of their high engagement with health-related content, intensive social media exposure, and potential future role in promoting healthy lifestyle behaviors. Data were collected using validated instruments. The proposed serial multiple mediation model was tested using PROCESS Macro (Model 6) with 5,000 bootstrap resamples. Results Social media–related factors were significantly associated with e-healthy diet literacy ( β = 0.31, p < 0.001), attitudes toward functional foods ( β = 0.31, p < 0.001), food craving acceptance ( β = −0.12, p < 0.001), and emotional eating ( β = 0.12, p < 0.001). In turn, e-healthy diet literacy ( β = 0.27, p < 0.001), attitudes toward functional foods ( β = 0.21, p < 0.001), food craving acceptance ( β = 0.29, p < 0.001), and emotional eating ( β = 1.32, p < 0.001) were all significantly associated with eating attitudes. The model explained 50% of the variance in eating attitudes (R 2 = 0.50). The full serial pathway was statistically significant, although the indirect effect size was small (B = 0.01, 95% CI [0.01, 0.01]). The total indirect effect was statistically significant (B = 0.51, SE = 0.04, 95% CI [0.44, 0.58]), while the full serial mediation pathway also reached significance (B = 0.01, 95% CI [0.01, 0.01]). Conclusion The findings indicate that the relationship between social media use and eating behaviors involves a sequence of cognitive, attitudinal, and emotional processes rather than a single direct pathway. Digital healthy nutrition literacy and acceptance-based regulation processes appear to play key roles within this framework. Given their intensive social media use and potential role-model status, sports sciences students represent an important target group for such approaches.