Association of First- vs. Second-Generation Antipsychotics with Lipid Abnormalities in Individuals with Severe Mental...

BACKGROUND AND OBJECTIVE: Individuals with severe mental illness experience increased morbidity and mortality as a result of metabolic problems that may partly be related to the adverse effects of antipsychotics. Compared with first-generation antipsychotics, second-generation antipsychotics collectively are considered to have stronger associations with lipid abnormalities, but evidence for this specific claim has not been systematically reviewed. The objective of this review was to evaluate the risk of dyslipidaemia with second-generation versus first-generation antipsychotics amongst individuals with severe mental illness.

Major electronic databases were searched until November 2018. Studies were eligible if they were cross-sectional, cohort, case-control or interventional, where any individual second-generation antipsychotic was directly compared with first-generation antipsychotics in individuals with severe mental illness, and where lipid metabolism was a primary or secondary outcome. The evidence was reviewed and appraised according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.

In total, 18 studies were eligible. The reported associations between second-generation antipsychotics vs. first-generation antipsychotics with dyslipidaemia were inconsistent, with high variability between studies and only a full qualitative synthesis was feasible. We had sufficient data, however, to undertake limited meta-analyses for clozapine, olanzapine and risperidone, all showing mildly elevated associations with dyslipidaemia "caseness" (clozapine, odds ratio 1.26, 95% confidence interval 1.16-1.38; olanzapine, odds ratio 1.29, 95% confidence interval 0.89-1.87; risperidone, odds ratio 1.05, 95% confidence interval 0.80-1.37) compared with first-generation antipsychotics, but heterogeneity was high (all I² > 50%, p < 0.05). Clozapine was also associated with increased triglycerides (standardised mean difference = 0.51, 95% confidence interval 0.21-0.81, I² = 5.74%), but not with cholesterol. Compared with haloperidol, neither olanzapine nor risperidone was associated with statistically significant increases in cholesterol or triglycerides.

There was considerable variation in study design and methodologies. Determining the comparative risk of second-generation vs. first-generation antipsychotics as a group of antipsychotics for lipid dysregulation may be of limited clinical utility, as drugs from either group have the potential to cause such adversity to varying degrees. It is therefore more valuable to consider the metabolic risks of specific antipsychotics rather than focusing on collective metabolic effects belonging to either antipsychotic group.