Depression is a heterogenous diagnostic construct; however, dynamic interactions between specific depressive symptoms across their qualitatively different profiles remain largely unknown. The study aimed to recognize the most prevalent profiles of depressive symptoms and assess their dynamics in young adults without a history of psychiatric treatment. Depressive symptoms were recorded using the Patient Health Questionnaire-9 (PHQ-9). The data were assessed for all theoretical and empirical combinations of depressive symptoms in participants with a positive screening for depression. The profiles identified in the majority of participants were analyzed using partial correlation and Bayesian networks. Data from 3583 individuals with a positive screening for depression were analyzed. Out of 382 theoretical profiles, 150 profiles (39.3%) were present in this dataset. The majority of participants (56.8%) showed 4 profiles of depressive symptoms including the profile with all depressive symptoms present, the profile without suicidal ideation, the profile without psychomotor impairment, and the profile without both psychomotor impairment and suicidal ideation. The profiles differed largely in terms of their dynamics and symptoms that are necessary to activate the whole network. The network characteristics within specific profiles did not differ significantly across the level of difficulties attributable to depressive symptoms. Our findings indicate that depression emerging in young adults shows a limited number of symptom profiles. However, dynamics of depressive symptoms differs largely between specific profiles regardless of functional impairment indicating the need to personalize therapeutic approaches. Future studies should further disentangle the heterogeneity of depressive symptoms, e.g., by dissecting the symptoms that are combined together by single PHQ-9 items (i.e., hypersomnia and insomnia; psychomotor agitation and retardation).
Keywords: Early intervention; Mood disorder; Personalized psychiatry; Psychopathology.
Copyright © 2024 The Author(s). Published by Elsevier Ltd.. All rights reserved.