The nonlinear relationship between symptoms of internet addiction and depressive symptoms: an empirical study among Chinese university students

Object To examine the nonlinear dose-response association between IA and depressive symptoms, while also exploring potential variations in this association across demographic subgroups. Methods A cross-sectional study was administered to Chinese university students. Internet addiction (IA) was assessed using an 8-item standardized diagnostic instrument. Depressive symptoms were evaluated utilizing the Center for Epidemiologic Studies Depression Scale (CES-D). Restricted cubic spline (RCS) regression was employed to detect nonlinear relationships and identify inflection points. Multivariable logistic regression and Cochran-Mantel-Haenszel stratified analyses, controlling for potential confounders, were conducted to assess symptom-specific correlates and examine subgroup differences. Results A nonlinear association was observed between IA scores and depressive symptoms ( p < 0.005), characterized by inflection points at scores of 4 and 7. Higher odds of depressive symptoms were seen at scores exceeding 4, demonstrating a pronounced increase beyond 7, and lower odds below 4. Analysis of specific IA symptoms revealed: Preoccupation was significantly correlated with higher odds of depressive symptoms (OR = 2.60, p< 0.05); deception was associated with elevated odds of depressive symptoms (OR = 1.51, p < 0.05); whereas needing more time online exhibited an association with lower odds of depressive symptoms (OR = 0.63, p < 0.05). Each unit increase in IA score was associated with 31% higher overall odds of depressive symptoms (OR = 1.31, p < 0.05). This association persisted following covariate adjustment, demonstrating a robust and independent correlational association. The positive association remained consistent across nearly all demographic and socioeconomic subgroups examined. Boarding school attendance was linked to a stronger IA-depressive-symptom association ( p < 0.001). Furthermore, only children exhibited stronger correlational estimates between IA scores and depressive symptoms compared to non-only children (OR = 1.24, p = 0.030). Conclusion The study indicates a positive correlational link between IA Score and depressive symptoms. RCS regression shows a curved relationship, with turning points at scores 4 and 7. Higher IA Scores correlate with elevated odds of depressive symptoms, making it a potential screening indicator for depressive-symptom risk (correlational only) for college students. Elevated odds are also present with an IA Score below 4. Excessive online activity correlates with higher odds, while prolonged internet use for gratification shows an association with lower odds of depressive symptoms.