Emotional and Uncontrolled Eating as Mediators Between Perceived Stress and Adiposity Indicators in University Students: A Cross‐Sectional Study
Sahar Afeef, Sundus Malaikah, Nehal AlSharif, Noor Algoregry, Reoff Alobeidi, Eram AlbajriABSTRACT
Background
Perceived stress has been associated with maladaptive eating behaviors, yet evidence from young adults in non‑Western populations remains limited. This study examined whether eating behavior traits mediate the relationship between stress and adiposity indicators among university students.
Methods
In a cross‐sectional study, undergraduate students (aged 18–26 years) completed Arabic versions of the Perceived Stress Scale (PSS‐10) and the Three‐Factor Eating Questionnaire‐Revised 18 (TFEQ‐R18) to assess perceived stress and three eating behavior traits (i.e., cognitive restraint, uncontrolled eating, and emotional eating). Anthropometric data, including body mass index (BMI), waist circumference (WC), and handgrip strength, were measured. Statistical analyses included group comparisons, Spearman correlations, and sex‐ and age‐adjusted parallel mediation to assess indirect effects of perceived stress on adiposity indicators via eating behaviors.
Results
Of 297 students (66.7% female), males had higher BMI, WC, and handgrip strength ( p < 0.001), while females reported greater perceived stress and cognitive restraint ( p ≤ 0.032). No direct association was found between stress and BMI ( b = –0.014, 95% CI: –0.107–0.079) or WC ( b = –0.037, 95% CI: –0.248–0.174). However, perceived stress was indirectly associated with higher BMI via emotional eating ( b = 0.031, 95% CI: 0.004–0.066; β = 0.034) and higher WC via uncontrolled eating ( b = 0.048, 95% CI: 0.002–0.118; β = 0.018), although these effects were very small. Cognitive restraint was not an indirect pathway for either outcome.
Conclusion
Perceived stress showed very small indirect associations with BMI and WC via emotional and uncontrolled eating, respectively. These findings suggest potential behavioral correlates, but as cross‑sectional associations, they do not provide evidence of causal mediation or clinical significance and require longitudinal study.