Distinct stress, eating, and body image phenotypes among university students: a latent class analysis in Saudi Arabia

Disordered eating and body image disturbances are increasingly recognized as pressing public health concerns among young adults. While variable-centered approaches dominate the literature, they may obscure heterogeneity by assuming uniform associations between stress, body image, and eating behaviors. In Saudi Arabia, where rapid sociocultural change, Western beauty ideals, and rising obesity intersect with unique cultural values and academic pressures, university students may exhibit distinct psychosocial phenotypes. Therefore, the study applied latent class analysis (LCA) to identify distinct subgroups characterized by varying levels of stress, body image dissatisfaction, and disordered eating attitudes among students at the Saudi Electronic University.

A cross-sectional online survey was conducted from February to March, 2025 across 15 campuses of the Saudi Electronic University. Validated instruments included the Perceived Stress Scale (PSS-10), Eating Attitudes Test (EAT-26), and Body Shape Questionnaire (BSQ-8). Latent class models (2-5 classes) were estimated using Gaussian mixture modeling. Bayesian Information Criterion, model parsimony, and interpretability guided class selection. Demographic correlates (age, sex, BMI, marital status) were assessed using χ², ANOVA, and multinomial logistic regression.

Among 771 respondents (mean age 25·1 years, 69.65% women, mean BMI 24.94 kg/m²), a four-class solution best fit the data. Class 0 (22·8%) exhibited Body Image Concerned -High Eating Concerns; Class 1 (30·7%) high stress with severe body-image dissatisfaction; Class 2 (9·9%) high eating concerns but low body-image dissatisfaction; and Class 3 (36·6%) moderate stress and body-image vulnerable with modest eating attitudes. Higher BMI increased the odds of Class 0 but reduced the odds of Classes 1 and 3. Female sex predicted lower odds of Classes 1 and 3 but a tendency toward Class 2. Age and marital status were not consistent predictors.

Saudi university students exhibit heterogeneous stress-eating-body image profiles, including a large subclinical group and a behaviorally at-risk subgroup without overt body-image dissatisfaction. These findings support stratified prevention strategies: stress-management for stress-dominated groups, behavioral monitoring for Class 2, and early preventive interventions for Class 3. Culturally adapted, sex-sensitive approaches aligned with Saudi Vision 2030 are warranted to reduce the burden of disordered eating.