Estimating Changes in Weight and Metabolic Parameters Before and After Treatment With Cariprazine: A Retrospective Study of Electronic Health Records

Purpose: Weight gain and associated negative cardiometabolic effects can occur as a result of mental illness or treatment with second-generation antipsychotics (SGAs), leading to increased rates of morbidity and mortality. In this analysis, we evaluated the effect of the SGA cariprazine on weight and metabolic parameters in a real-world, retrospective, observational dataset. Methods: Electronic health records from the Optum Humedica database (October 1, 2014-December 31, 2020) were analyzed during the 12-month period before starting cariprazine (baseline) and for up to 12 months fol-lowing cariprazine initiation; approved and off-label indications were included. Body weight trajectories were estimated in the overall patient cohort and at 3-, 6-, and 12-month timepoints (primary objective). Changes in hemoglobin A1c (HbA1c), low-density lipoprotein (LDL), high-density lipoprotein (HDL), and triglycerides were also evaluated (secondary objectives). Percentages of patients with clinically relevant shifts in body weight, total cholesterol, and fasting triglycerides were also determined. Discontinuation rates for metabolic regulating medications were calculated. Average predicted values were estimated by linear mixed-effects regression models. Findings: A total of 2,301 patients were included; average duration of follow-up was 133.7 days. Average predicted weight change for patients during the cariprazine overall follow-up period was + 2.4 kg, with predicted weight changes of + 0.8 kg (n = 811), + 1.1 kg (n = 350), and + 1.4 kg (n = 107) at months 3, 6, and 12, respectively. Over-all, the majority of patients did not experience clinically significant ( >= 7%) weight gain (82.8%) or loss (90.5%) after starting cariprazine. Average predicted HbA1c levels (n = 189) increased during baseline (0.15%/year) and decreased during cariprazine treatment (- 0.2%/year). Average predicted triglyceride levels (n = 257) increased during baseline (15.0 mg/dL/year) and decreased during cariprazine treatment (- 0.7 mg/dL/year). Predicted LDL (n = 247) and HDL (n = 255) values decreased during baseline (- 7.3 and - 1.1 mg/dL/year, respectively); during cariprazine treatment, LDL increased by 5.6 mg/dL/year and HDL decreased by - 0.6 mg/dL/year. During follow-up, most patients did not shift from normal/borderline to high total cholesterol ( < 240 to >= 240 mg/dL; 522 [90.2%]) or fasting triglyceride ( < 200 to >= 200 mg/dL; 143 [88.8%] patients) levels; shifts from high to normal/borderline levels occurred in 44 (61.1%) patients for total cholesterol and 38 (57.6%) patients for fasting triglycerides. After starting cariprazine, the discontinuation rate per 100 patient-years was 60.4 for antihyper-glycemic medication and 87.4 for hyperlipidemia medication. Implications: These real-world results support short-term clinical trial findings describing a neutral weight and metabolic profile associated with cariprazine treatment and they expand the dataset to include long-term follow-up.