Link between diabetes and antiretrovirals may particularly affect resource-poor settings

An understanding of the links between antiretroviral drugs and diabetes is growing, with previous data particularly implicating protease inhibitors (PIs). Now, a Swiss study, published in the July 1^st edition of Clinical Infectious Diseases has found that other HIV drugs, often used in resource-poor settings, may also increase the risk in populations that may already have a high risk of diabetes.

Bruno Ledergerber (University Hospital, Zurich, Switzerland) and colleagues examined data from the Swiss HIV Cohort Study to clarify the associations with treatments and also coinfection with hepatitis B or C (HBV and HCV, respectively). All participants in the Swiss Cohort were included if they did not already have diabetes and if they had at least two visits and one year follow-up since March 1, 2000. Diabetes was diagnosed by criteria of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus and confirmed by blood glucose sampling.

By July 2006, 123 of 6,513 participants experienced diabetes mellitus during 27,798 person-years of follow-up (PYFU), which gives an incidence of 4.4 cases per 1,000 PYFU. In multivariable models, male gender, older age, African or Asian ethnicity, an AIDS diagnosis, and central obesity were strong predictors of diabetes. Obesity should become a major target for diabetes prevention, as it is a modifiable risk factor, say the authors. Of note, no markers of HBV or HCV infection were associated with an increased risk of diabetes, although the team acknowledges that the study size may have been too low to show a small increased risk.

Importantly, current treatment with an antiretroviral regime containing nucleoside reverse-transcriptase inhibitors (NRTIs) and PIs, or NRTIs plus PIs and non-nucleoside reverse-transcriptase inhibitors (NNRTIs) were strongly associated with diabetes, with similar incidence rate ratios to other predictors. Treatment with NRTIs plus NNRTIs was not linked with diabetes nor did cumulative exposure to antiretrovirals increase the risk. In a separate analysis, current treatments associated with the greatest risk of diabetes were indinavir, lamivudine-stavudine, didanosine-stavudine, and didanosine-tenofovir.

The incidence of diabetes in this study is fairly comparable with that found in general population studies in Europe. However, much higher estimates have been found in HIV-positive populations in the USA. For example, the Multicenter AIDS Cohort Study (MACS) found 47 cases of diabetes per 1,000 PYFU in people on combination antiretroviral treatment, versus 17 cases for HIV-positive people not on treatment and 14 cases in the HIV-negative population. The authors propose three factors to explain the difference: the diagnosis of diabetes, and the older age and much higher body-mass index found in the MACS study.

The findings in regard to antiretroviral drugs confirm the studies done so far, and add weight to the few studies that implicate NRTIs in abnormalities of glucose metabolism. Mechanisms may differ, with protease inhibitors affecting glucose transport and nucleoside reverse-transcriptase inhibitors affecting mitochondrial function. “Because of their association with other metabolic disorders, regimens containing stavudine and didanosine are avoided as long as possible in developed countries, but they belong to first-line regimens in resource-limited areas,” note the authors. Together with the elevated diabetes risk associated with Asian and African ethnicity, “this may have an important impact on the long-term tolerability of anti-HIV treatment in the regions that are most affected,” they conclude.