Multivariate associations between psychiatric drug intake and grey matter volume changes in individuals at early stages of psychosis and depression

• Published in: European psychiatry 2024-04, Vol.67 (S1), p.S272-S273
• Main Authors: Weyer, C., Popovic, D., Ruef, A., Hahn, L., Sarişik, E., Fanning, J., Kambeitz, J., Salokangas, R. K., Hietala, J., Bertolino, A., Borgwardt, S., Brambilla, P., Upthegrove, R., Wood, S. J., Lencer, R., Meisenzahl, E., Falkai, P., Koutsouleris, N.
• Format: Article
• Language: English
• Subjects: Antidepressants; Antipsychotics; Mental depression; Personality; Psychosis; Psychotropic drugs
• ISSN: 0924-9338; 1778-3585
• DOI: 10.1192/j.eurpsy.2024.573

IntroductionPsychiatric drugs, including antipsychotics and antidepressants, are widely prescribed, even in young and adolescent populations at early or subthreshold disease stages. However, their impact on brain structure remains elusive. Elucidating the relationship between psychotropic medication and structural brain changes could enhance the understanding of the potential benefits and risks associated with such treatment.ObjectivesInvestigation of the associations between psychiatric drug intake and longitudinal grey matter volume (GMV) changes in a transdiagnostic sample of young individuals at early stages of psychosis or depression using an unbiased data-driven approach.MethodsThe study sample comprised 247 participants (mean [SD] age = 25.06 [6.13] years, 50.61% male), consisting of young, minimally medicated individuals at clinical high-risk states for psychosis, individuals with recent-onset depression or psychosis, and healthy control individuals. Structural magnetic resonance imaging was used to obtain whole-brain voxel-wise GMV for all participants at two timepoints (mean [SD] time between scans = 11.15 [4.93] months). The multivariate sparse partial least squares (SPLS) algorithm (Monteiro et al. JNMEDT 2016; 271:182-194) was embedded in a nested cross-validation framework to identify parsimonious associations between the cumulative intake of psychiatric drugs, including commonly prescribed antipsychotics and antidepressants, and change in GMV between both timepoints, while additionally factoring in age, sex, and diagnosis. Furthermore, we correlated the retrieved SPLS results to personality domains (NEO-FFI) and childhood trauma (CTQ).ResultsSPLS analysis revealed significant associations between the antipsychotic classes of benzamides, butyrophenones and thioxanthenes and longitudinal GMV decreases in cortical regions including the insula, posterior superior temporal sulcus as well as cingulate, postcentral, precentral, orbital and frontal gyri (Figure 1A-C). These brain regions corresponded most closely to the dorsal and ventral attention, somatomotor, salience and default network (Figure 1D). Furthermore, the medication signature was negatively associated with the personality domains extraversion, agreeableness and conscientiousness and positively associated with the CTQ domains emotional and physical neglect.Image:ConclusionsPsychiatric drug intake over a period of one year was linked to distinct GMV reductions in key cortical hubs. Thes