Body mass index and health-related lifestyle behaviors among university students in Saudi Arabia: a single center cross-sectional study
Abstract
Overweight and obesity are important public-health concerns among young adults. University students experience changes in dietary, physical-activity, sleep, and other health behaviors, but evidence from Saudi university populations remains heterogeneous. To explore associations between BMI and selected dietary, physical-activity, sleep, and sociodemographic characteristics among university students. This cross-sectional study analyzed electronic questionnaire responses collected from January to April 2025 on a sample of 196 students. Weight and height were self-reported and BMI was recalculated as weight (kg)/height 2 (m 2 ). BMI was categorized as underweight (<18.5), normal (18.5 to <24.9), overweight (25 to <30), or obesity (≥30 kg/m 2 ). Associations were examined using contingency-table analyses and limited exploratory multivariable linear and logistic regression models. Among 196 analyzed students, 17 (8.7%) were underweight, 106 (54.1%) had normal BMI, 38 (19.4%) were overweight, and 35 (17.9%) had obesity; thus, 73 (37.2%) had overweight or obesity. Associations of BMI category with lifestyle were not statistically significant. Electronic-device use before bed showed a sparse-cell association ( χ 2 =23.438, df = 9, Monte Carlo p = 0.010) that should be considered exploratory. In adjusted linear regression, male sex ( β = 5.26 kg/m 2 , 95% CI 2.36–8.16, p < 0.001), and students aged >23 years had higher BMI ( β = 2.86, 95% CI 0.44–5.29, p = 0.021). In adjusted logistic regression, male sex (OR = 2.71, 95% CI 1.19–6.19, p = 0.018) and age >23 years (OR = 3.15, 95% CI 1.17–8.43, p = 0.023) were associated with overweight/obesity. Overweight and obesity were common. Gender and age showed associations with BMI status, whereas evidence for independent associations of the selected lifestyle measures was limited and imprecise. The cross-sectional design and self-reported measurements preclude causal inference