Background: Psychotropic-associated weight gain contributes to non-adherence and cardiometabolic morbidity, but antipsychotics, antidepressants, and mood stabilizers are rarely compared in a common longitudinal framework. Methods: We harmonized 4945 participants from five NIMH effectiveness trials: CATIE and ACLAIMS (schizophrenia; antipsychotics), CO-MED (major depression; three antidepressant regimens—escitalopram monotherapy, bupropion–escitalopram, and venlafaxine–mirtazapine), and STEP-BD and LiTMUS (bipolar disorder; mood stabilizers). Co-primary outcomes were cumulative clinically significant weight gain (≥7%; restricted to the 4436 participants with ≥1 post-baseline weight) and the early weight-change slope (6 months); ≥7% gain within a common 6-month window was an additional, observation-time-matched comparison. Models adjusted for baseline weight, age, and sex where applicable. Results: Cumulative ≥7% gain occurred in 41% of antipsychotic, 29% of mood-stabilizer, and 16% of antidepressant-treated participants with a post-baseline weight (antipsychotic vs antidepressant OR 3.67, 95% CI 2.74–4.91); the cumulative proportions were similar across the two antipsychotic and the two mood-stabilizer trials. This cumulative difference partly reflects the longer follow-up and more frequent weight measurement in the antipsychotic trials: early weight-gain slopes did not differ significantly between antipsychotics and antidepressants, and within a common 6-month window no statistically significant difference in clinically significant gain was detected between the antipsychotic and antidepressant cohorts (OR 1.29, 95% CI 0.91–1.81, p = 0.148). Lower baseline weight predicted cumulative gain (p < 0.001) but was null for absolute (kg) weight gain (OR 1.002 per kg, p = 0.161), indicating it partly reflects the mathematics of a percentage-based threshold; younger age and male sex predicted faster early gain. In exploratory secondary analyses the antipsychotic trials showed serum-glucose increases, with a waist increase in CATIE. Conclusion: In these trial-defined class/diagnosis-regimen cohorts, the apparent antipsychotic excess was largely cumulative and associated with longer observation duration and more frequent measurement; within a common 6-month window no statistically significant difference in early weight gain was detected between the antipsychotic and antidepressant cohorts. Because drug class, diagnosis, and trial are aligned in this design, these differences should not be interpreted as an isolated drug-class effect. Younger patients—and, for proportional gain, leaner patients—warrant early metabolic monitoring across diagnoses.
Transdiagnostic medication-associated weight gain across five effectiveness trials: a harmonized longitudinal cohort analysis of antipsychotics, antidepressants, and mood stabilizers
Serretti, Alessandro
2026-01-01
Abstract
Background: Psychotropic-associated weight gain contributes to non-adherence and cardiometabolic morbidity, but antipsychotics, antidepressants, and mood stabilizers are rarely compared in a common longitudinal framework. Methods: We harmonized 4945 participants from five NIMH effectiveness trials: CATIE and ACLAIMS (schizophrenia; antipsychotics), CO-MED (major depression; three antidepressant regimens—escitalopram monotherapy, bupropion–escitalopram, and venlafaxine–mirtazapine), and STEP-BD and LiTMUS (bipolar disorder; mood stabilizers). Co-primary outcomes were cumulative clinically significant weight gain (≥7%; restricted to the 4436 participants with ≥1 post-baseline weight) and the early weight-change slope (6 months); ≥7% gain within a common 6-month window was an additional, observation-time-matched comparison. Models adjusted for baseline weight, age, and sex where applicable. Results: Cumulative ≥7% gain occurred in 41% of antipsychotic, 29% of mood-stabilizer, and 16% of antidepressant-treated participants with a post-baseline weight (antipsychotic vs antidepressant OR 3.67, 95% CI 2.74–4.91); the cumulative proportions were similar across the two antipsychotic and the two mood-stabilizer trials. This cumulative difference partly reflects the longer follow-up and more frequent weight measurement in the antipsychotic trials: early weight-gain slopes did not differ significantly between antipsychotics and antidepressants, and within a common 6-month window no statistically significant difference in clinically significant gain was detected between the antipsychotic and antidepressant cohorts (OR 1.29, 95% CI 0.91–1.81, p = 0.148). Lower baseline weight predicted cumulative gain (p < 0.001) but was null for absolute (kg) weight gain (OR 1.002 per kg, p = 0.161), indicating it partly reflects the mathematics of a percentage-based threshold; younger age and male sex predicted faster early gain. In exploratory secondary analyses the antipsychotic trials showed serum-glucose increases, with a waist increase in CATIE. Conclusion: In these trial-defined class/diagnosis-regimen cohorts, the apparent antipsychotic excess was largely cumulative and associated with longer observation duration and more frequent measurement; within a common 6-month window no statistically significant difference in early weight gain was detected between the antipsychotic and antidepressant cohorts. Because drug class, diagnosis, and trial are aligned in this design, these differences should not be interpreted as an isolated drug-class effect. Younger patients—and, for proportional gain, leaner patients—warrant early metabolic monitoring across diagnoses.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
