Diabetes Rate 3.8% Higher in People Living With HIV in National U.S. Comparison
Diabetes prevalence stood at 10.3% among HIV-positive adults in a nationally representative U.S. sample. After adjustment for sex, age and other factors, diabetes prevalence was 3.8% higher with HIV than in the general U.S. population. Among people with HIV, women, persons 20 to 44 years old and nonobese individuals all had significantly higher diabetes prevalence than comparable groups in the general population.
Diabetes mellitus affects 29 million people across the United States, more than one-quarter of them undiagnosed. Metabolic diseases including diabetes affect growing proportions of aging HIV populations surviving longer on antiretroviral therapy. But there has been no nationally representative estimate of diabetes prevalence in U.S. residents with HIV, and diabetes risk with versus without HIV remains controversial. To address these issues, Centers for Disease Control and Prevention (CDC) researchers conducted this nationally representative comparison of people with versus without HIV.
To estimate prevalence of diagnosed diabetes in people with HIV, the CDC team used 2009-2010 data from the nationally representative Medical Monitoring Project (MMP), which includes adults who receive HIV care in the United States. General population data came from the National Health and Nutrition Examination Survey (NHANES) for 2009-2010. The researchers used logistic regression to determine and compare weighted prevalences of diabetes mellitus in the two populations. They also used logistic regression adjusted for age, race, obesity and other variables to identify factors associated with diabetes in the HIV population.
The study samples included 8,610 nonpregnant adults with HIV and 5,604 nonpregnant adults in NHANES. Compared with the general population, the HIV group included a higher proportion of men (73.6% versus 49.3%), a lower proportion of whites (34.6% versus 67.3%), a higher proportion of blacks (41.3% versus 11.7%), a higher proportion of 45- to 60-year-olds (50.3% versus 29.8%), a lower proportion 60 or older (9.6% versus 21.6%), a higher proportion living at or below the poverty level (43.5% versus 8.5%), a higher proportion positive for hepatitis C (HCV) (20.6% versus 1.7%) and a lower proportion with obesity (25.5% versus 36.0%).
Unadjusted prevalence of diagnosed diabetes came to 10.3% with HIV and 8.3% in the general population. After adjustment for age, sex, race, obesity, poverty, education and HCV, diabetes prevalence was 3.8% higher among people in care for HIV. The largest adjusted differences between the HIV group and the general population involved people positive for HCV (6.3%), those with at least high-school education (5.1%), women (5.0%), non-Hispanic whites (4.9%), people at or below the poverty line (4.6%), obese people (4.4%), people 20 to 44 years old (4.1%) and nonobese people (3.5%).
In an analysis restricted to people in care for HIV, adjusted weighted diabetes prevalence was highest in people 60 or older (19.6%), obese people (18.9%), people without AIDS and with a nadir CD4 count above 500 cells/mm3 (13.5%), people diagnosed with HIV for 10 or more years (12.3%), people 45 to 60 years old (11.9%) and people positive for HCV (11.6%). Factors independently associated with total diabetes in people with HIV were older age, obesity, increasing time since HIV diagnosis and geometric mean CD4 count. Prescription of antiretroviral therapy in the past year did not predict prevalent diabetes.
The CDC investigators note that their analysis offers the first nationally representative estimate of diabetes prevalence in adults in care for HIV infection in the United States. Compared with the general population, these people with HIV were more likely to have diabetes at younger ages and without obesity. The researchers observe that the association between diabetes and longer time since HIV diagnosis may reflect older age, cumulative antiretroviral exposure and exposure to HIV-related chronic inflammation. The CDC investigators recommend that clinicians follow current diabetes screening guidelines for people with HIV. They add that further research must address whether these guidelines should be modified to include HIV as a risk factor.