Changes in non-high-density lipoprotein cholesterol levels and triglyceride/high-density lipoprotein cholesterol ratios among patients randomized to aripiprazole versus olanzapine.Schizophr Res. 2008 Dec; 106(2-3):300-7.SR
Non-high-density lipoprotein cholesterol (non-HDL-C) and the triglyceride to high-density lipoprotein cholesterol ratio (TG:HDL-C) are predictors of cardiovascular risk. This post-hoc analysis assessed changes in these parameters during treatment with the atypical antipsychotics olanzapine or aripiprazole using pooled data from three randomized, long-term clinical studies in patients with schizophrenia.
Data were pooled from one open-label and two double-blind (26- or 52-week) studies in patients randomized to olanzapine (5-20 mg/day) or aripiprazole (15-30 mg/day). Change from baseline in non-HDL-C levels between groups was analyzed in the Observed Case (OC) dataset at each time point and Last Observation Carried Forward (LOCF) dataset at endpoint using analysis of covariance, with treatment as main effect and baseline non-HDL-C as covariate. Differences between groups in median changes from baseline in TG:HDL-C were assessed with Kruskal-Wallis tests.
This analysis included 546 patients (olanzapine, n=274; aripiprazole, n=272). Mean changes from baseline in non-HDL-C levels were significantly different (p<0.0001) with olanzapine versus aripiprazole at Weeks 26 (+13.0 versus -7.5 mg/dL) and 52 (+12.2 versus -8.1 mg/dL). Baseline TG:HDL-C was high in the olanzapine (3.73) and aripiprazole (3.79) groups. Differences in median changes from baseline in TG:HDL-C were significant with olanzapine versus aripiprazole at Weeks 26 (+0.22 versus -0.54; p<0.0001) and 52 (+0.24 versus -0.62; p=0.004).
Long-term aripiprazole treatment is associated with improvements in lipid profiles of schizophrenia patients versus no improvement or worsening during olanzapine treatment. Consideration of cardiovascular risk is needed when prescribing antipsychotics, as is close monitoring for metabolic changes during treatment.