Prediction of early weight gain during psychotropic treatment using a combinatorial model with clinical and genetic markers
BACKGROUND: Psychotropic drugs can induce significant (>5%) weight gain (WG) already after 1 month of treatment, which is a good predictor for major WG at 3 and 12 months. The large interindividual variability of drug-induced WG can be explained in part by genetic and clinical factors.
AIM: The aim of this study was to determine whether extensive analysis of genes, in addition to clinical factors, can improve prediction of patients at risk for more than 5% WG at 1 month of treatment.
METHODS: Data were obtained from a 1-year naturalistic longitudinal study, with weight monitoring during weight-inducing psychotropic treatment. A total of 248 Caucasian psychiatric patients, with at least baseline and 1-month weight measures, and with compliance ascertained were included. Results were tested for replication in a second cohort including 32 patients.
RESULTS: Age and baseline BMI were associated significantly with strong WG. The area under the curve (AUC) of the final model including genetic (18 genes) and clinical variables was significantly greater than that of the model including clinical variables only (AUCfinal: 0.92, AUCclinical: 0.75, P<0.0001). Predicted accuracy increased by 17% with genetic markers (Accuracyfinal: 87%), indicating that six patients must be genotyped to avoid one misclassified patient. The validity of the final model was confirmed in a replication cohort. Patients predicted before treatment as having more than 5% WG after 1 month of treatment had 4.4% more WG over 1 year than patients predicted to have up to 5% WG (P≤0.0001).
CONCLUSION: These results may help to implement genetic testing before starting psychotropic drug treatment to identify patients at risk of important WG.
Vandenberghe, F., Saigí-Morgui, N., Delacrétaz, A., Quteineh, L., Crettol, S., Ansermot, N., ... Eap, C. B. (2016). Prediction of early weight gain during psychotropic treatment using a combinatorial model with clinical and genetic markers. Pharmacogenetics and Genomics, 26(12), 547-557. DOI: 10.1097/FPC.0000000000000249