Relationships between the use of second-generation antipsychotics and changes in total cholesterol levels in children and adolescents perinatally infected with HIV

Abstract

PURPOSE: Perinatally HIV-infected children, who are increasingly aging into adolescence and early adulthood, have significant rates of psychiatric co-morbidities, some of which are treated with second-generation antipsychotics (SGAs). SGAs have been associated with elevated total cholesterol (TC) in youth, but no studies have examined this association in perinatally HIV-infected youth. This study examined changes in TC levels of youth with perinatally acquired HIV infection and co-morbid psychiatric conditions treated with SGAs. PATIENTS AND METHODS: Long-term changes in TC levels were examined using data from the US multisite prospective Pediatric AIDS Clinical Trials Group 219C cohort study. The change in TC levels from baseline to 12 months after initiating SGA use was compared between 52 SGA-exposed and 148 matched SGA-unexposed perinatally HIV-infected youth, using generalized estimating equation models adjusting for other covariates. The prevalence and time to incident hypercholesterolemia were also compared between these 2 groups. RESULTS: After adjustment for confounders, 52 youth with prescriptions for SGAs had a larger increase in TC levels than 148 matched youth without antipsychotic prescriptions (mean difference = 9 mg/dL, z = 1.96, df = 1, P = 0.0496). Among youth with TC below 220 mg/dL at baseline, 27% of SGA-exposed youth developed hypercholesterolemia (defined as two consecutive TC measurements ≥220 mg/dL), compared with 13% of SGA-unexposed patients (Fisher's exact test, P = 0.046). CONCLUSIONS: Caution should be used in prescribing SGAs to perinatally HIV-infected youth with psychiatric co-morbidities due to increased risk of hypercholesterolemia. Patients should be monitored, and alternative evidence-based treatments considered when available.

Figures

Figure 1

Mean total cholesterol over time, by SGA exposure, with 95% confidence intervals. The mean TC at a given time point included measurements that occurred within 3 months of that time point. If a participant had more than 1 measurement in this window, then the measurement closest to the time point was used. Abbreviations: SGA, second-generation antipsychotic; TC, total cholesterol.

Figure 2

Estimated probability of remaining free of hypercholesterolemia, by SGA-exposure. Abbreviation: SGA, second-generation antipsychotic.