Fig. 4: Results of PLS-SEM analysis for the associations among bullying victimization, help-seeking intention, Glx, and subclinical psychotic experiences.

Path analyses were performed to determine the relationships among variables including bullying victimization, help-seeking intention, age-sex-SES-IQ-adjusted Glx levels, and subclinical psychotic experiences using PLS-SEM. Results of a the time lagged model and of b the latent change score model are illustrated here. Squares represent observed variables and circles represent latent variables. Solid arrows represent direct effects and dashed arrows represent moderation effects. Path coefficients (p values in parentheses) are shown on corresponding arrows. Indirect effects were also assessed. In the time lagged model (a), Glx levels at Time 1 had a significant total indirect effect on subclinical psychotic experience at Time 2, and bullying victimization had a trend-level significant total indirect impact on subclinical psychotic experience at Time 2. In the latent change score model (b), bullying victimization and bullying victimization × help-seeking intention interaction had significant total indirect effects on differences in Glx levels. BV bullying victimization, HSI help-seeking intention, GLX combined glutamate-glutamine, SPE subclinical psychotic experience, T1 Time 1, T2 Time 2 BASE baseline, DIFF difference.