Latent Profile Analysis of Parent-Reported Mood Symptoms in Youth: A Person-Centered Approach to Classification

Objective To evaluate whether latent profile analysis (LPA) of dimensional mood symptoms can identify clinically meaningful subgroups of youth, using a measure that captures both positive (manic) and negative (depressive) valence, and to assess convergence with psychiatric diagnoses and interview-based ratings of mood symptom severity. Method We conducted a secondary analysis of data from 1369 treatment-seeking youths (ages 5–18) recruited from an urban community mental health center and an academic medical center. Consensus DSM diagnoses were derived from KSADS interviews, family history, and treatment records. LPA was applied to 20 parent-rated symptom facets from the Parent General Behavior Inventory (PGBI), modeling two continuous mood dimensions. Profiles were compared against mood diagnoses and symptom severity based on the Children's Depression Rating Scale-Revised (CDRS-R) and the Young Mania Rating Scale (YMRS), both scored via direct parent-youth interviews by trained clinicians (with PGBI scores masked). Results A 5-profile solution was supported by fit indices (BIC, entropy) and interpretability (minimum cluster size: n =123). Profile membership was strongly associated with consensus diagnoses (χ 2 [16]=579.39, p <.000005) and current mood state ratings ( χ 2 [20]=583.99, p <.000005). Profiles varied significantly in rates of depressive, manic, and mixed states, and in average CDRS-R and YMRS scores, as well as global functioning and family history of bipolar disorder. Conclusion Latent profiles based on dimensional ratings of positive and negative mood symptoms aligned well with structured interview-derived diagnoses and symptom states. These profiles may offer a data-driven method to enhance mood disorder classification and support early identification efforts in youth psychiatry.