Sex-specific diagnostic trajectories and time to transition from non-SMI to severe mental illness in Chinese adolescent inpatients

TL;DR

The study finds that Chinese adolescent inpatients with non-severe mental illness have distinct sex-specific diagnostic paths and a higher risk of transitioning to severe mental illness 4-8 years after initial diagnosis.

Contribution

This is the first longitudinal study in a non-Western population to identify sex-specific diagnostic trajectories and a 4-8 year high-risk window for transitioning from non-SMI to SMI in adolescents.

Findings

SMI diagnoses (schizophrenia spectrum and bipolar disorders) showed high stability over time.

Males with externalizing or obsessive-compulsive disorders and females with internalizing or stress-related disorders were more likely to transition to SMI.

Each 1-year increase in baseline age was associated with a 38% higher risk of transitioning to SMI.

Abstract

Adolescence is a high-risk period for mental disorders, and early clinical presentations often show uncertainty and pluripotentiality. Longitudinal evidence on transitions from non-specific diagnoses to severe mental illness (SMI—defined as schizophrenia spectrum and bipolar disorders) in non-Western populations remains limited. Using real-world data, we aimed to characterize diagnostic stability, transitions from non-SMI to SMI, and sex- and age-related predictors in Chinese adolescent inpatients.

This retrospective longitudinal cohort study utilized electronic medical records from a large tertiary psychiatric hospital in Eastern China (2010–2026). We included 884 first-time inpatients aged 12–17 years with ≥3 years of follow-up and at least two complete inpatient records. ICD-10 diagnoses were grouped into SMI and non-SMI categories. Sankey diagrams and transition matrices were used to describe diagnostic trajectories from baseline to the last admission. Among patients with non-SMI at baseline, Kaplan–Meier analyses examined the time to transition to SMI, and multivariable logistic regression tested the independent effects of sex and baseline age on SMI conversion.

Over a median follow-up of 4.60 years (IQR 3.63–6.53), SMI diagnoses showed high stability: 81%(243/300)of schizophrenia spectrum disorders and 74%(104/141) of bipolar disorders remained unchanged. Overall, 39.2%(346/884)of patients experienced at least one diagnostic change, primarily within non-SMI categories; depressive disorders were the most frequent antecedent of bipolar disorder (13%(22/171) converted). Sex-stratified analyses suggested that certain externalising and obsessive–compulsive presentations in males, and internalising and stress-related presentations in females, were more frequently followed by SMI; however, several subgroup estimates were based on small numbers and should be considered exploratory. Kaplan–Meier curves indicated that the risk of transition from non-SMI to SMI clustered between 4 and 8 years after the first admission. Each 1-year increase in baseline age was associated with a 38% higher risk of SMI conversion (OR = 1.38, 95% CI 1.19–1.60, P < 0.001), and, after adjusting for age, males had approximately twice the risk of SMI conversion compared with females (OR = 1.90, 95% CI 1.17–3.07, P = 0.009).

Adolescent psychiatric diagnoses show substantial longitudinal evolution, with relatively stable SMI once established but appreciable medium- to long-term progression from non-SMI to SMI. The identified sex-specific pathways and the 4-to-8-year high-risk window support longitudinal, developmentally informed monitoring—particularly for older male adolescents with severe or atypical non-SMI presentations.

Figures