Background: Mood and psychotic disorders, including major depressive disorder (MDD), bipolar disorder (BD), and schizophrenia (SCZ), show overlapping symptoms that challenge diagnostic boundaries and may inform personalised treatments. We examined the contribution of genetic liability to transdiagnostic symptom dimensions in treatment outcomes across disorders using polygenic scores (PGSs). Methods: We analysed two MDD cohorts (total N = 2548), one BD cohort (N = 755), and one SCZ cohort (N = 449). Outcomes included treatment resistance, symptomatic remission, and change in functioning defined using cohort-specific data. PGSs for anhedonia, anxiety, sociability, resilience, cognitive and sleep-related traits were computed using SBayesRC. Regressions adjusted for potential confounders were run in each cohort and results meta-analysed with random-effects models; heterogeneity was assessed with leave-one-out and influence diagnostics, and meta-regressions. Multiple testing was controlled using a Bonferroni correction adjusted for the effective number of independent tests (αadj = 0.0034). Results: In meta-analyses, no result survived multiple testing correction. The strongest signal was for the Trail Making Test Part B PGS, indexing worse executive function/processing speed, and non-remission (OR = 1.13, p = 0.012; I2 = 13 %); modelling diagnosis (mood disorders vs SCZ) reduced heterogeneity to I2 = 0 %, and leave-one-out excluding SCZ reached statistical significance (OR = 1.17, p = 0.001). Verbal-numerical reasoning PGS, indexing predisposition to higher fluid intelligence, was nominally associated with improved functioning (β = −0.06, p = 0.016), confirmed in leave-one-out excluding BD (β = −0.07, p = 0.0095). Conclusion: PGSs for cognitive traits showed trait- and diagnosis-specific associations with treatment outcomes. Cross-diagnostic analyses may identify shared genetic influences, but variability in symptom expression across disorders may introduce heterogeneity and reduce the detectability of such effects.
Polygenic predisposition to transdiagnostic symptom dimensions and treatment outcomes across psychiatric disorders
Serretti A.;
2025-01-01
Abstract
Background: Mood and psychotic disorders, including major depressive disorder (MDD), bipolar disorder (BD), and schizophrenia (SCZ), show overlapping symptoms that challenge diagnostic boundaries and may inform personalised treatments. We examined the contribution of genetic liability to transdiagnostic symptom dimensions in treatment outcomes across disorders using polygenic scores (PGSs). Methods: We analysed two MDD cohorts (total N = 2548), one BD cohort (N = 755), and one SCZ cohort (N = 449). Outcomes included treatment resistance, symptomatic remission, and change in functioning defined using cohort-specific data. PGSs for anhedonia, anxiety, sociability, resilience, cognitive and sleep-related traits were computed using SBayesRC. Regressions adjusted for potential confounders were run in each cohort and results meta-analysed with random-effects models; heterogeneity was assessed with leave-one-out and influence diagnostics, and meta-regressions. Multiple testing was controlled using a Bonferroni correction adjusted for the effective number of independent tests (αadj = 0.0034). Results: In meta-analyses, no result survived multiple testing correction. The strongest signal was for the Trail Making Test Part B PGS, indexing worse executive function/processing speed, and non-remission (OR = 1.13, p = 0.012; I2 = 13 %); modelling diagnosis (mood disorders vs SCZ) reduced heterogeneity to I2 = 0 %, and leave-one-out excluding SCZ reached statistical significance (OR = 1.17, p = 0.001). Verbal-numerical reasoning PGS, indexing predisposition to higher fluid intelligence, was nominally associated with improved functioning (β = −0.06, p = 0.016), confirmed in leave-one-out excluding BD (β = −0.07, p = 0.0095). Conclusion: PGSs for cognitive traits showed trait- and diagnosis-specific associations with treatment outcomes. Cross-diagnostic analyses may identify shared genetic influences, but variability in symptom expression across disorders may introduce heterogeneity and reduce the detectability of such effects.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
