Classic Risk Factors Drive Heart Attack Rates More Than HIV Variables

Avoiding traditional cardiovascular risk factors -- high total cholesterol, hypertension and smoking -- would have the biggest impact on cutting myocardial infarction (MI) rates in North Americans with HIV, according to a 29,515-person analysis. Eliminating any of the three traditional risk factors would cut MI prevalence by about 40%, while avoiding a CD4 count below 200 cells/mm3 would avert only 10% of MIs, and avoiding a detectable viral load or clinical AIDS would have even smaller impacts.

Research ties HIV infection to a 50% higher MI risk after adjusting for classic risk factors, comorbidities and substance use. But the relative contribution of classic MI risk factors versus HIV-related factors remains difficult to sort out because HIV populations typically carry high burdens of traditional factors such as smoking and dyslipidemia.

To address this issue, NA-ACCORD researchers figured population-attributable fractions for classic and HIV risk factors, defined as "the proportion of MIs that could be avoided in HIV-infected adults if all were unexposed to the modifiable risk factor of interest." These fractions can be figured by combining data on risk factor prevalence and adjusted hazard ratios for each risk factor.

Related: No Evidence of Accelerated Brain Aging in HIV-Positive People on Effective ART

The study population came from seven cohorts contributing to NA-ACCORD, an observational cohort established in 2006 and designed to provide a representative sample of people in HIV care in the United States and Canada. The study period for the MI analysis extended from January 2000 through December 2013, and the main outcome was incident validated type 1 MI.

The NA-ACCORD investigators assessed the population-attributable fraction of nine risk factors: cigarette smoking, elevated total cholesterol, treated hypertension, diabetes, stage 4 chronic kidney disease (CKD), CD4 count below 200 cells/mm3, HIV RNA at or above 400 copies/mL, clinical AIDS diagnosis and hepatitis C (HCV) coinfection. In a sensitivity analysis including 57% of participants, the researchers considered the impact of body mass index (BMI) of 30 kg/m2 or higher (obesity).

Through a median follow-up of 3.5 years (interquartile range 1.5 to 6.9), 29,515 cohort members had 347 MIs. Substantial differences between proportions of people who had an MI and those who did not were age under 40 years (18% versus 47%), age 50 to 59 (31% versus 14%), age 60 or older (10% versus 3%), male gender (86% versus 80%), white race (56% versus 46%), Hispanic ethnicity (6% versus 11%) and antiretroviral naive status (26% versus 39%).

Hazard ratios adjusted for the risk factors being scrutinized -- and for age, sex and race --determined that smoking (ever versus never), treated hypertension, diabetes, stage 4 CKD and CD4 count below 200 cells/mm3 were each independently associated with higher MI risk. When the analysis also adjusted for BMI indicating obesity, those variables remained independently associated with higher MI risk and two others joined the list: elevated total cholesterol and HCV coinfection.

Adjusted population-attributable fraction was highest for elevated total cholesterol at 43%, meaning 43% of MIs could be avoided if no one in the study population ever had high total cholesterol and if the analysis held other risk factors constant. Population-attributable fractions for hypertension (41%) and smoking (38%) were almost as high. Results were similar in the analysis that factored in BMI indicating obesity. In the main analysis, population-attributable fractions were much lower for CD4 count below 200 cells/mm3 (10%), HCV coinfection (8%), viral load at or above 400 copies/mL (6%), stage 4 CKD (3%), a clinical AIDS diagnosis (2%) and diabetes (2%).

The NA-ACCORD investigators propose that intervention to prevent MIs "must expand beyond successful [antiretroviral treatment] to address traditional risk factors." They suggest that efforts focus on preventing high total cholesterol, hypertension, smoking and HCV coinfection.