Older anti-HIV drugs associated with increased risk of diabetes

New cases of diabetes in patients taking HIV therapy in France peaked between 1999 and 2000 but have since fallen sharply, investigators report in the online edition of AIDS. Treatment with older anti-HIV drugs such as indinavir (Crixivan), d4T (stavudine, Zerit) and ddI (didanosine, Videx) was associated with the development of diabetes. However, there was no evidence that this was the case for newer anti-HIV drugs.

“The importance of exposure of first-generation antiretroviral drugs in the development of new-onset diabetes is consistent with the drop in incidence of new-onset diabetes in recent years,” write the investigators. “[This] correlates with the use of new antiretroviral drugs.”

Effective antiretroviral therapy has significantly extended the life expectancy of many patients with HIV. However, non-AIDS-related diseases are an increasingly important cause of illness and death in HIV-positive patients. One such disease is type-2 diabetes.

Therefore investigators from the French ANRS-CO8 APROCO-COPILOTE cohort designed a study to establish the incidence of new cases of diabetes and its risk factors in a group of 1046 patients who started HIV therapy between 1997 and 1999.

The patients were followed until 2009. A total of 7,846 person-years of follow-up were available for anaylsis (median 9.6 years per patient). The most widely used antiretroviral drugs were indinavir (54%), d4T (75%) and ddI (52%).

New-onset diabetes was diagnosed if a patient’s fasting glucose increased to above 7.0 mmol/l or if their two-hour oral glucose tolerance test was above 11.1 mmol/l. Individuals who started anti-diabetic therapy were also diagnosed as having new-onset diabetes. 

Overall, there were 111 cases of new-onset diabetes. This provided an incidence of 14.1 cases per 1000 person-years. Incidence was similar in men and women (14.6 vs 12.6 per 1000 person-years).

Incidence peaked between 1999 and 2000 at 23.2 cases per 1000 person-years. It then fell steadily to 5 cases per 1000 person years in the period 2007 to 2009. The investigators note that such an incidence is similar to that seen in other HIV observational cohorts and the general French population.

Risk factors for the development of diabetes included older age (aged between 40 and 50 vs age below 40, HR = 2.13, 95% CI, 1.36-3.34; aged over 50 vs under 40, HR = 3.63, 95% CI, 2.22-5.92).

Higher body mass index (BMI) was also predictive of new-onset diabetes (BMI 25-30 kg/m² vs below 25 mg/m², HR = 1.91, 95% CI, 1.22-2.99; above 30 kg/m² vs below 25 kg/m², HR = 2.85, 95% CI, 1.35-6.04). Hip-to-waist ratio was also a significant predictor of the condition (above 97 men/92 women, HR = 3.87, 95% CI, 2.31-6.49).

Analysis also indicated a significant relationship between lipoatrophy and diabetes (HR = 2.14, 95% CI, 1.33-3.44).

Therapy with several older anti-HIV drugs (whose use is no longer recommended) was also predictive of the condition.

Treatment with indinavir lasting up to one year increased the risk of diabetes (HR = 2.53, 95% CI, 1.34-4.79). Therapy with d4T lasting up to two years was also associated with new-onset diabetes (HR = 2.65, 95% CI, 1.23-5.72) as was up to three years of treatment with ddI (HR = 3.16, 95% CI, 1.34-7.42).

“We found that diabetes occurred more frequently in HIV-infected patients exposed to indinavir, stavudine and didanosine,” comment the investigators.

However, once treatment with these drugs was stopped, the risk of developing new-onset diabetes was reduced to background levels.

Reassuringly, there was no evidence that newer anti-HIV drugs, such as lopinavir/ritonavir (Kaletra) or atazanavir (Reyataz), was associated with diabetes.

The investigators conclude that routine HIV care “should include measure of adiposity markers (waist circumference and BMI) and fasting glycemia at least yearly to identify at-risk patients. It is of utmost importance that diabetes is detected and adequately treated in order to favourably influence long-term cardiovascular health.”