Şanlıdağ, Ahmet; Aydın, Tuğba; Uzel, Melih Can; Navruz, Çiğdem Yıldırım

Objective: This study aimed to examine the association between problematic smartphone use, sleep quality and psychosocial factors among dental students, and to identify key behavioural and psychological predictors of high-risk smartphone use within this academically demanding population. Materials and Methods: A cross-sectional survey was conducted during the 2024–2025 academic year at Faculty of Dentistry, Atatürk University including 485 students from all academic years (mean age = 22.31 ± 1.77; 63.1% female). Participants were classified into low- and high-risk groups based on gender-specific Smartphone Addiction Scale–Short Version (SAS-SV) cut-offs (≥ 31 for males; ≥ 33 for females). Sleep quality was assessed using the Pittsburgh Sleep Quality Index (PSQI). For the separate analysis of sleep disturbance, the raw sum of PSQI items 5b–5j was used, with a theoretical range of 0–27, rather than the conventional 0–3 PSQI sleep disturbance component score. Psychosocial status was evaluated using the Depression Anxiety Stress Scales-21 (DASS-21). Group comparisons were performed using appropriate parametric or non-parametric tests, and predictors of high-risk smartphone use were identified through binary logistic regression analysis. Results: Overall, 54.5% of participants were classified as high-risk users (mean SAS-SV = 33.26 ± 10.57). Demographic variables such as age, sex and class year were not associated with risk classification (all p > 0.90). Sleep onset latency was significantly shorter in the high-risk group (19.90 ± 17.53 min) compared to the low-risk group (24.50 ± 18.36 min; p = 0.005), while total sleep duration showed no difference (p = 0.531). Among PSQI subcomponents, only sleep disturbance was significantly associated with high-risk smartphone use (OR = 1.05, 95% CI: 1.01–1.09, p = 0.029). Categorical analyses further showed that poor subjective sleep quality, sleep disturbance, daytime dysfunction and elevated stress, anxiety and depression were significantly more frequent among high-risk users, whereas sleep medication use did not differ between groups. In the psychosocial model, stress scores predicted high-risk smartphone use (OR = 1.09, 95% CI: 1.02–1.17, p = 0.015), whereas anxiety and depression scores were non-significant (p ≥ 0.62). Conclusion: While sleep disturbance and stress predicted high-risk PSU in domain-specific models, sleep disturbance lost significance in the fully adjusted combined model, where daytime dysfunction emerged as the primary sleep-related predictor and stress remained only marginally significant (p = 0.049). These findings should therefore be interpreted with caution.