People taking the nucleoside reverse transcriptase inhibitors (NRTIs) Zerit (stavudine), Retrovir (zidovudine) and Videx (didanosine) were found to have a higher risk of developing diabetes mellitus, say the authors of a large international study published in the February 11 issue of Diabetes Care. The data also suggest that the non-nucleoside reverse transcriptase inhibitor (NNRTI) Viramune (nevirapine) and, surprisingly, the protease inhibitor (PI) Norvir (ritonavir) are protective against new onset diabetes.

The risk of a heart attack is more than doubled among HIV-positive patients with diabetes mellitus, according to previously reported results from the Data Collection on Adverse Events of Anti-HIV Drugs (D:A:D). While traditional risk factors for diabetes—such as obesity, genetics and sedentary lifestyles—are believed to play major roles in onset of the disease in HIV-positive people, it has also been suggested that the use of antiretroviral medications can contribute to the impairment of glucose metabolism in the body. Early studies suggested that PIs were a likely culprit. 

To more completely assess the prevalence of diabetes in HIV-positive people, as well as its potential HIV-related risk factors, Stephane De Wit, MD, PhD, from the Centre Hospitalier Universitaire Saint-Pierre, in Brussels, and his colleagues revisited data from the D:A:D. Their analysis included 33,389 people living with HIV, of whom 774 developed diabetes during the course of the study.

Dr. De Wit’s team found that the longer a person spent on combination antiretroviral therapy, the higher their risk for developing diabetes became. This was true even when the team controlled for known diabetes risk factors. The drug most strongly associated with an increased risk for diabetes was the NRTI Zerit, although Retrovir and Videx were also associated with an increased risk.

Use of the NNRTI Viramune was found to lower the risk of developing diabetes. Most surprising, the PI Norvir—previously linked to an increased risk of diabetes—was actually shown to have a slight protective effect. In fact, the D:A:D data did not show a significant relationship between the use of any PIs and new onset diabetes. 

People with lipodystrophy, a syndrome that is defined by gains and losses of body fat and unhealthy changes to cholesterol and triglycerides (lipids), were also at increased risk of developing diabetes. However, when De Wit’s team controlled for lipodystrophy, Zerit, Retrovir and Videx were still associated with an increased risk for diabetes. This was statistically significant, meaning that the association was too large to have occurred by chance.

The authors hypothesize that Zerit and Retrovir may be contributing to diabetes by damaging mitochondria, the source of energy inside cells, and thereby reducing the body’s ability to use insulin to control blood sugar, a condition known as insulin resistance.

As for the suggestion that Norvir is not associated with new onset diabetes, a possible explanation is that much lower doses of the PI are now used solely to boost other antiretrovirals in the bloodstream, whereas early studies connecting the PI to diabetes involved patients using full therapeutic doses—600 mg twice a day—of the drug.