The assessment of metabolic syndrome in UK patients with HIV using two different definitions: CREATE 2 study.

OBJECTIVE

To determine the prevalence and clinical associations of the metabolic syndrome (M-IRS) in an HIV cohort.

METHODS AND

DESIGN

Data was collected prospectively on demographics, anthropometry, HIV disease, drug regimens and cardiometabolic risk factors using a two-centre cross-sectional cohort study design. M-IRS was diagnosed by National Cholesterol Education Program (NCEP) and International Diabetes Federation (IDF) criteria.

RESULTS

The prevalence of M-IRS in 678 subjects was 14% by NCEP and 10% by IDF. One feature of the M-IRS was present in 68%, while 37% had two or more features. Increased waist circumference was found in 32% by NCEP or by IDF criteria, hypertriglyceridaemia in 32%, reduced HDL-C in 27%, 18% had raised systolic blood pressure and 13% had dysglycaemia. Protease inhibitor (PI) usage was similar in both M-IRS categories (43 vs. 38%; p = 0.38) but increased use of efavirenz was seen in M-IRS (47 vs. 36%; p = 0.07) and nevirapine in the non-M-IRS groups (10 vs. 20%; p = 0.05). Multiple drug therapies were associated with raised triglyceride levels while nevirapine therapy was associated with raised HDL-C and abacavir with dysglycaemia.

CONCLUSIONS

The prevalence of M-IRS in this HIV cohort was similar to the general population and independent of current or previous highly active antiretroviral therapy (HAART) or its duration. Given the relationship between individual drugs and features of M-IRS its significance must be interpreted in the light of probable accrual bias in prescribing. Prospective studies are required to ascertain the cardiometabolic risk factors to include in a prognostically useful HIV disease-specific definition of M-IRS.