Anxiety, emotional abuse, and physical activity as correlates of AI-related affective dependence among adolescents: mediation, moderation, and network analyses.

Emotional abuse was positively associated with anxiety and AI dependence and negatively associated with physical activity among adolescents.

• Sample: 1,589 Chinese adolescents who completed self-report questionnaires in November 2025.
• Measures covered emotional abuse (EA), anxiety, physical activity (PA), and AI-related affective dependence.
• Analyses included descriptive statistics and Pearson correlations.

Anxiety statistically accounted for a small but significant proportion of the association between emotional abuse and AI dependence.

• Indirect effect = 0.034, 95% CI (0.013, 0.055).
• The indirect effect represented 16.43% of the total association.
• Mediation is described as statistical; the cross-sectional data do not support temporal or causal mechanisms.

Physical activity moderated the association between emotional abuse and anxiety, with a weaker association at higher physical activity levels.

• Tested using moderated mediation models.
• The positive EA-anxiety association was weaker among adolescents reporting higher PA.

In the covariate-adjusted partial-correlation network, the EA-anxiety edge was the strongest conditional association and EA had the highest node strength.

• Network analysis was adjusted for covariates.
• The authors state these findings should be interpreted descriptively rather than causally.

Common method bias was evaluated with two approaches.

• Methods used were Harman's single-factor test and a CFA-based unmeasured latent method construct (ULMC) approach.
• All data were self-reported.

The study's cross-sectional design limits inference to statistical associations, not temporal or causal mechanisms.

• The framing draws on attachment and emotion-regulation perspectives.
• Authors conclude anxiety and PA are meaningful correlates in the EA-AI dependence association.