Metabolic complications of antiretroviral therapy (ART) in a South African black population..

Abstract

Aims To determine the prevalence and incidence of lipodystrophy (fat distribution [lipoatrophy and lipohypertrophy] and metabolic complications [insulin resistance-dysglycaemia and dyslipidemia]) in HIV-1 infected adult subjects of second generation Zulu descent at baseline and during 24 months of follow-up on antiretroviral therapy (ART). Methods The total study group included three groups: HIV infected ART naive patients eligible for ART (HIV-ART, n=150), age, gender and ethnically matched HIV infected not eligible for ART (HIV-no ART, n=88) and HIV negative (control, n=88) subjects. All participants had demographic, anthropometric, biochemical and radiological assessments at baseline; in addition, the HIV-ART group had follow-up assessments for 24 months on ART (tenofovir, lamivudine and nevirapine or efavirenz). Fat distribution was assessed using FRAM questionnaires, computerized tomography (CT) scans and dual energy absorptiometry X-ray (DXA). Disorders of glycaemia (diabetes mellitus (DM), impaired glucose tolerance and impaired fasting glucose) were defined using WHO criteria. Total, LDL, HDL cholesterol and triglycerides were measured for each group; CD4 cell count and HIV RNA for group 2 and 3, at baseline, 3, 6, 12, 18 and 24 months. Poisson approximations estimated incidence of disorders of glycaemia. Results At baseline, when compared with the control group, the mean BMI (kg/m2) was significantly lower in the HIV-ART and HIV-no ART subjects (26.4 vs. 28.6 vs. 29.1; p =0.01). Prevalence of lipoatrophy as measured by participant and physician examination questionnaires was similar in the three groups. Visceral and subcutaneous fat area by CT scan were similar between the groups but limb and trunk fat mass by DXA scan was significantly lower in the HIV-ART compared to control subjects. In the HIV-ART group, at the 24 month follow-up, there was a significant mean reduction in HIV RNA (p<0.0001) and increase in CD4 cell count (p<0.0001). The mean BMI increased to 29.4 kg/m2 and no lipoatrophy developed; DXA scan showed a 33.6% increase in trunk fat mass (mean difference 4.2 kg, p <0.0001) and 30.8% increase in total fat mass (mean difference 9.4 kg, p < 0.0001); visceral (p 0.005) and subcutaneous (p 0.0002) fat area also increased. At baseline, the prevalence of DM was 0% in HIV-ART and HIV-no ART and 4.9% in control subjects (p 0.005); the prevalence of “any dysglycaemia” was 3.7% in HIV-ART and HIV-no ART compared to 8.6% in control subjects. When compared with group 1, mean values in group 3 were lower for the following serum lipids: total cholesterol (p<0.0001), LDL (p=0.0007) and HDL (p<0.0001). There was no difference in mean total triglycerides in the three groups (p=0.3). During follow-up, in the HIV-ART group, using glucose-based WHO criteria, the incidence of diabetes mellitus was 2.3 per 100 person year follow-up (PYFU) and of “any dysglycaemia” 7.6 per 100 PYFU. The only independent predictor of DM was visceral: subcutaneous fat ratio measured by CT scan (HR 2.95 [95% CI 1.25-6.98], p 0.01). Significant predictors for development of “any dysglycaemia” included systolic blood pressure (HR 1.04 [95%CI 1.02-1.07], p=0.0006), serum albumin (HR 0.85 [95% CI 0.76-0.94], p=0.002), CD4 cell count (HR 0.988 [95%CI 0.978-0.997], p=0.01) and efavirenz (HR 6.27 [95%CI 1.65-23.80], p=0.01) Serum total (p<0.0001), LDL (p<0.0001) and HDL-cholesterol (p<0.0001) increased significantly during follow-up. Conclusion: In this cohort of South Africans with HIV-1 infection, at baseline (prior to ART) there was no significant fat redistribution or lipoatrophy and an absent to low prevalence of dysglycaemia. In the follow-up study, ART use was not associated with lipoatrophy although there was significant increase in BMI and in limb and trunk fat mass by DXA scan. ART was associated with increased incidence of dysglycaemia. These findings underscore the importance of clinical monitoring on ART. The association of efavirenz with dysglycaemia warrants further evaluation.