The predictive power of autobiographical memory in shaping the mental health of young people : an individual participant data meta-analysis.

Abstract

Reduced autobiographical memory (AM) specificity, characterized by difficulty recalling specific past events, is a feature of multiple psychiatric disorders. While meta-analyses indicate that reduced AM specificity can predict future symptom severity, its role as a premorbid risk factor for mental illness onset in young people remains unclear. Our preregistered individual participant data meta-analysis (PROSPERO; CRD42022287786) synthesized longitudinal data from 14 community-based studies of children and adolescents (N = 9,165). Most studies reported symptom severity (92.9%), with one third also reporting diagnostic status (35.7%). Assessment timing ranged from 2 months to 8.3 years following the autobiographical memory task. Multivariate mixed-effect models found no support for AM specificity predicting future symptom severity in depression or posttraumatic stress disorder. Contrary to expectations, higher memory specificity significantly predicted higher anxiety symptoms, but the quality of evidence was low. Intriguingly, reduced AM specificity significantly predicted an increased risk of receiving a psychiatric diagnosis and an earlier onset of disorder (hazard ratio = 0.55, p = .018). This risk was most robust for depressive disorders (hazard ratio = 0.21, p < .001). Findings suggest reduced AM specificity could be a risk factor for the onset of functionally impairing psychiatric disorders, but it does not consistently predict elevated symptoms in community-based samples. Further theoretical development is needed.