Trajectories of psychotic-like experiences in youth and associations with lifestyle factors.

BACKGROUND

Persistent and/or distressing psychotic-like experiences (PLEs) during adolescence are associated with poorer subsequent psychiatric outcomes. Modifiable lifestyle factors (such as sleep quality or regular exercise) may improve mental health outcomes; however, it is unknown how lifestyle factors are linked to trajectories of PLEs.

METHODS

Using data from the Adolescent Brain Cognitive Development Study (N = 10,075, age 9-10 years at baseline), we characterized trajectories of PLEs using latent growth mixture models assessed using the Prodromal Questionnaire-Brief Child Version. We examined trajectories of Total and Distress scores. We used multinomial logistic regressions to examine associations between baseline lifestyle behaviors (including self-reported screen time, physical activity and caffeine intake, and parent-reported sleep disturbances and recreational activities) and PLE trajectories.

RESULTS

We identified four trajectories of distress-related PLEs: No Distress (27%), Rapid Decreasing (17%), Gradual Decreasing (36%), and Persistent Elevated Distress (21%). Compared with the No Distress trajectory, individuals in the Persistent Elevated Distress trajectory spent more time using screens (adjusted Odds Ratio [OR] 2.27, 95% confidence interval [CI] 2.03-2.53), had higher caffeine intake (OR 1.62, 95% CI 1.28-2.04), greater sleep disturbance (OR 1.58, 95% CI 1.45-1.73), participated in fewer recreational activities (OR 0.75, 95% CI 0.68-0.83) and less frequent physical activity (OR 0.81, 95% CI 0.74-0.89). Greater screen time and sleep disturbances further distinguished the most severe group from all other trajectories. Findings were similar when examining total scores. Results remained statistically significant when we included established risk factors of psychosis in each model.

CONCLUSIONS

Lifestyle factors associate with trajectories of PLE-related distress, providing novel tools for intervention and risk prediction.