JMIR Publications

ABSTRACT

Background:

Although previous cross-sectional studies have found a positive association between internet gaming disorder (IGD) and suicidal behaviors, longitudinal evidence is limited, and the direction of this relationship remains unclear.

Objective:

This large-scale, two-wave study aimed to investigate the bidirectional relationship between IGD and suicidal behaviors among Chinese adolescents.

Methods:

A total of 57,985 adolescents were followed over one year, from November 2021 (baseline) to November 2022 (follow-up). IGD was assessed using the Internet Gaming Disorder Scale-Short Form (IGDS9-SF), while suicidal behaviors (including ideation and attempts) were measured at both time points. Covariates included demographic information and shared mental health risk factors (depression, anxiety, sleep problems, conduct problems, and hyperactivity/inattention). Univariate and multivariate logistic regression analyses were performed to examine: (1) the cross-sectional association between IGD and suicidal behavior at baseline (n = 57, 985); (2) the longitudinal association between baseline IGD and incident suicidal behavior among adolescents without suicidal behavior at baseline (n = 46, 370); and (3) the longitudinal association between baseline suicidal behavior and incident IGD among adolescents without IGD at baseline (n = 47,681). Sensitivity and subgroup analyses were also performed.

Results:

(1) At baseline, the prevalence of suicidal behaviors and IGD was 20.0% and 17.8%, respectively. After controlling for demographics and mental health issues, IGD was independently associated with suicidal behaviors (adjusted odds ratio [AOR] = 1.63, 95% CI = 1.53–1.73, p < 0.001). (2) Among 46,370 adolescents without suicidality at baseline, 9.5% developed suicidal behaviors during the follow-up. Baseline IGD was a significant predictor of subsequent suicidal behavior (AOR = 1.46, 95% CI = 1.34–1.59, p < 0.001). (3) Among 47,681 adolescents without IGD at baseline, 10.4% developed IGD during follow-up. Baseline suicidal behavior was significantly associated with the onset of IGD (AOR = 1.38, 95% CI = 1.26–1.50, p < 0.001). (4) Subgroup analyses showed consistent associations between IGD and suicidality across various demographic groups. Sensitivity analyses confirmed the robustness of the IGD-suicidal behavior and suicidality-IGD associations when treating suicidal ideation and attempts as separate outcomes.

Conclusions:

This study provides strong evidence for a bidirectional relationship between IGD and suicidal behaviors among adolescents, independent of shared mental health risk factors. These findings underscore the importance of screening for suicidal behaviors in adolescents with IGD and vice versa. Targeted interventions addressing both conditions simultaneously may offer more effective support for at-risk adolescents. Clinical Trial: Not applicable.