icon-folder.gif   Conference Reports for NATAP  
48th Annual ICAAC / IDSA 46th Annual Meeting
October 25-28, 2008
Washington, DC
Back grey_arrow_rt.gif
Lower Risk of Diabetes With HIV in US Veterans Affairs Cohort
  48th ICAAC, October 25-28, 2008, Washington, DC
Mark Mascolini
Contrary to trends in other HIV cohorts, Veterans Aging Cohort Study (VACS) investigators found that HIV infection independently lowers the risk of diabetes about 15% [1]. But other diabetes risk factors--older age, minority race, higher weight, and HCV coinfection--raised the risk of diabetes more in people with HIV than in those without HIV. And people taking nucleosides or nonnucleosides had a higher diabetes risk than those who did not.
Two US cohort studies found higher risks of insulin resistance or diabetes in US populations with HIV, but in both studies the higher risk correlated with nucleoside reverse transcriptase inhibitors. A 1288-man Multicenter AIDS Cohort Study (MACS) analysis figured that every additional year of nucleoside therapy raised the risk of high insulin levels 8% and worsened the insulin resistance QUICKI score [2]. Comparing 1524 HIV-infected women with 564 uninfected women, Women's Interagency HIV Study (WIHS) investigators found nearly a doubled risk of diabetes in HIV-infected women exposed to nucleosides, though HIV itself did not affect diabetes risk [3]. Further work by the WIHS team placed particular blame on stavudine (d4T) for insulin resistance [4].
Veterans Affairs investigators compared 3327 HIV-infected people with 3240 uninfected people enrolled in the VACS in 6 cities: Atlanta, Baltimore, Houston, New York, Pittsburgh, and Washington, DC . The researchers recruited HIV-uninfected people from the general medical clinics at the same sites and matched them to cases by 5-year age block, race, and gender. They defined diabetes as (1) glucose at or above 200 mg/dL on two separate occasions, or (2) ICD-9 codes for diabetes (2outpatient or 1 inpatient) plus treatment with an oral hypoglycemic or insulin for at least 30 days, or (3) ICD-9 codes for diabetes (2outpatient or 1 inpatient) plus glucose at or above 120 mg/dL on two separate occasions, or (4) glucose at or above 200 mg/dL once plus treatment with an oral hypoglycemic or insulin for at least 30 days.
Compared with HIV-negative controls, people in the HIV group were younger, more likely to be black, more likely to be men, and more likely to have HCV infection. The HIV group had a significantly lower average body mass index and drank less alcohol than did controls, but HIV-infected people used recreational and illicit drugs more than the non-HIV group. Younger age and lower body mass index in the HIV group would make diabetes less likely, while being black and having HCV infection would raise the diabetes risk.
Multivariate analysis considering all these risk factors plus antiretroviral duration and CD4 count at study entry determined that HIV infection independently lowered the risk of diabetes 16% (odds ratio [OR] 0.84, 95% confidence interval [CI] 0.72-0.97). However, increasing age pushed up the risk of diabetes much more in the HIV group than in the non-HIV group. Compared with people under 35 years old, those in higher age brackets had the following increased risks of diabetes (95% CI in parentheses; CIs that do not cross 1.0 are statistically significant):
· 35-39: Overall 1.86 (0.98-3.54), HIV group 4.54 (1.01-20.34), non-HIV group 1.25 (0.58-2.68)
· 40-44: Overall 2.69 (1.50-4.83), HIV group 6.80 (1.61-28.71), non-HIV group 1.97 (1.01-3.84)
· 45-49: Overall 3.65 (2.08-6.42), HIV group 8.43 (2.03-35.06), non-HIV group 2.72 (1.44-5.12)
· 50-54: Overall 5.57 (3.18-9.77), HIV group 13.52 (3.26-56.07), non-HIV group 3.88 (2.07-7.27)
· 55-59: Overall 6.94 (3.95-12.19), HIV group 13.75 (3.30-57.23), non-HIV group 5.31 (2.82-9.97)
· 60-64: Overall 8.85 (4.89-16.02), HIV group 18.38 (4.27-79.13), non-HIV group 6.81 (3.47-13.39)
· 65-69: Overall 12.15 (6.58-22.41), HIV group 22.53 (5.08-99.93), non-HIV group 9.59 (4.79-19.22)
· 70 or older: Overall 10.68 (5.78-19.71), HIV group 17.04 (3.70-78.62), non-HIV group 8.03 (4.04-15.97)
Men had a 79% higher risk of diabetes than women (OR 1.79, 95% CI 1.23-2.59). Although that steeper gender-based risk appeared to be higher in the HIV group than in the uninfected group, the higher risk in HIV-infected people failed to reach statistical significance (OR 2.51, 95% CI 0.96-6.52), while the gender-based risk did reach statistical significance in the non-HIV group (OR 1.65, 95% CI 1.09-2.49). Compared with white race, being black independently raised the risk of diabetes 65% in the HIV group and 24% in the non-HIV group. Hispanic ethnicity independently inflated the risk 55% in people with HIV and 42% in people without HIV.
Compared with a body mass index under 20 kg/m(2), people in the 20-to-24.9 range had a 40% higher risk of diabetes
(68% higher with HIV and 20% higher without HIV), people in the 25-to-29.9 range had a 1.99 times higher risk (2.30 higher with HIV and 1.70 higher without HIV), and people at 30 kg/m(2) or more had a 4 times higher risk (5.35 higher with HIV and 3.25 higher without HIV). HCV coinfection independently raised the risk of diabetes in people with HIV (OR 1.36, 95% CI 1.06-1.73) but not in those without HIV (OR 1.28, 95% CI 0.96-1.66).
Treatment with protease inhibitors did not up the risk of diabetes in this cohort, but, as in MACS and WIHS, nucleoside therapy did. Every year of nucleoside treatment independently heightened the diabetes risk 6% (95% CI 1.02-1.10). Every added year of nonnucleoside therapy hoisted the diabetes risk even more--9% (95% CI 1.02-1.17).
More alcohol and illicit drug use correlated with a lower risk of diabetes in this cohort, regardless of HIV infection. Only when participants reported more than 60 drinks monthly did they run a higher risk of diabetes. Further study discerned no link between liver dysfunction (judged by alanine or aspartate aminotransferase level) and diabetes risk. Nor did the investigators uncover an association between drinking or illegal drugs and lower body mass index, which would protect against diabetes. Cohort members who drank alcohol tended to be younger than those who did not, but the VA team suggested that other undiscovered mechanisms contribute to the lower diabetes risk they found in alcohol and drug users.
Despite the independently lower risk of diabetes with HIV in the overall analysis, the consistently higher diabetes risk with HIV in all of the subanalyses supports close glucose and insulin monitoring in people with HIV.

1. Butt AA, McGinnis K, Rodriguez-Barradas M, et al. HIV infection and risk of diabetes mellitus. 48th Annual International Conference on Antimicrobial Agents and Chemotherapy (ICAAC). October 25-28, 2008. Washington, DC. Abstract H-2306.
2. Brown TT, Li X, Cole SR, et al. Cumulative exposure to nucleoside analogue reverse transcriptase inhibitors is associated with insulin resistance markers in the Multicenter AIDS Cohort Study. AIDS. 2005,19:1375-1383.
3. Tien PC, Schneider MF, Cole SR, et al. Antiretroviral therapy exposure and incidence of diabetes mellitus in the Women's Interagency HIV Study. AIDS. 2007,21:1739-1745.
4. Tien PC, Schneider MF, Cole SR, et al. Antiretroviral therapy exposure and insulin resistance in the Women's Interagency HIV Study. J Acquir Immune Defic Syndr. 2008 Oct 16. Epub ahead of print.