Abstract: Background: Individuals experiencing a first episode of psychosis (FEP) frequently exhibit rapid and clinically significant weight gain, especially during the first year of antipsychotic treatment. Such weight gain considerably increases the risk for metabolic and cardiovascular diseases, contributing to higher morbidity and reduced life expectancy. Previous evidence suggests that incorporating polygenic scores (PGS) for body mass index improves weight gain prediction in patients with a FEP beyond clinical parameters alone. However, the predictive contribution of other PGS remains unexplored. We evaluated whether cardiometabolic and cardiovascular PGS enhance the prediction of BMI at 3 months (BMI3M), as well as at 1 (BMI1Y) and 3 (BMI3Y) years, and whether BMI changes are also predicted at these time points (?BMI3, ?BMI1Y, and ?BMI3Y) in individuals with FEP.
Methods: We calculated the PGS for BMI (PGSBMI) and the additional cardiometabolic (type II diabetes: PGST2D, metabolic syndrome: PGSMS, hypercholesterolemia: PGSHC) and cardiovascular traits (blood pressure: PGSBP, non-ischemic cardiomyopathy: PGSNIC) using GWAS data in a sample of 381 FEPs (PAFIP cohort). Nested clinical and genetic models were built, incorporating clinical-demographic information (population stratification principal components, age, sex, diagnosis, antipsychotic type and dose, tobacco and cannabis use) and different PRSs as predictors. Cross-sectional BMI measures (BMI3M, BMI1Y, and BMI3Y) and longitudinal BMI changes (BMI3M, BMI1Y, and BMI3Y) were used as dependent variables. We assessed the incremental predictive value of the models regarding BMI outcomes.
Results: The analyses indicate that individual PGSs models significantly predict BMI3M, BMI1Y and BMI3Y, as well as the BMI changes at 3 months and 1 and 3 years, explaining between 5% and 17% of the variance. Nonetheless, individual PGS did not enhance the prediction achieved by the clinical and PGSBMI models. On the other hand, the integration of different cardiometabolic PRSs demonstrated a significant value for BMI trajectories, with an incremental predictive capacity over the clinical and PRSBMI models alone.
Discussion: Our results align with previous evidence indicating that different genetic factors significantly influence weight gain trajectories during the early stages of antipsychotic treatment. The increased predictive ability observed when combining multiple PGS reflects the contribution of partially distinct genetic factors and other biological mechanisms that are not fully captured by BMI polygenicity alone. Therefore, integrating various PGS may better represent the complex genetic architecture of weight gain and may enhance early identification of patients at elevated risk, potentially informing personalised intervention strategies to mitigate long-term health comorbidities associated with psychosis. Funding: MICIU/AEI co-funded by ESF+ grant-JDC2023-050677-I to MG-R; IISCIII co-funded by ERDF/ESF project-PI21/00612.
Other conference communications or articles related to authors from the University of Cantabria
