Prevalence of Problematic Smartphone Use and Its Association with Socio-economic, Physical, and Psychological Conditions: A Cross- Sectional Study among University Students in Bangladesh

Abstract Background Problematic smartphone use (PSU) denotes excessive, poorly controlled smartphone engagement linked to impaired functioning; it is best treated as a screening construct rather than a formal diagnosis. University students are a high-risk group. In Bangladesh, prevalence estimates diverge widely (~ 28–61%), and few studies jointly model socio-economic, lifestyle, and psychological correlates with transparent measurement reporting and full model diagnostics. Methods Cross-sectional survey of 450 students across four faculties at Noakhali Science and Technology University, Bangladesh, reported per STROBE. PSU was measured with the SAS-SV (cut-offs ≥ 31 males, ≥ 33 females). Three SAS-SV items were data-entry duplicates; classification used seven unique items rescaled to the 10-item metric. Socio-demographics, sleep, physical activity, sedentary time, and a four-item PHQ-4-style distress screener were collected. Chi-square tests and a parsimonious multivariable logistic model with crude/adjusted ORs and diagnostics (VIF, Hosmer–Lemeshow, AUC) were used. Results Corrected PSU prevalence was 54.0% (243/450; 95% CI 49.4–58.6%), versus 63.8% with duplicated items. Males (73.0%) exceeded females (33.6%; χ²=68.36, p < 0.001). In the final model (LR χ²=269.20, df = 14, p < 0.001; Nagelkerke R²=0.602; Hosmer–Lemeshow p = 0.70; AUC = 0.90; all VIF ≤ 2.51), female gender predicted lower odds (AOR = 0.13, 95% CI 0.07–0.24), while senior year (AOR = 2.18, 1.19–4.00), sleep outside 7–8 h/night (AORs 5.83–38.76), greater sedentary time (AOR = 1.41 per category, 1.10–1.81), and higher distress (AOR = 1.50 per point, 1.30–1.71) predicted higher odds; physical activity was protective (AOR = 0.76 per level, 0.58–0.99). Conclusions PSU was common and linked to male gender, senior standing, non-optimal sleep, sedentariness, low activity, and distress. Cross-sectional design and low scale reliabilities (SAS-SV 7-item α = 0.44; distress α = 0.34) limit causal inference; associations may guide campus sleep, activity, and mental-health programmes. Transparent item-correction and sample-specific alphas are a methodological contribution.