Secondary Logo

Journal Logo

Epidemiology and Prevention

HIV Infection, Cardiovascular Disease Risk Factor Profile, and Risk for Acute Myocardial Infarction

Paisible, Anne-Lise MD*; Chang, Chung-Chou H. PhD; So-Armah, Kaku A. PhD; Butt, Adeel A. MD§; Leaf, David A. MD; Budoff, Matthew MD; Rimland, David MD#; Bedimo, Roger MD**; Goetz, Matthew B. MD; Rodriguez-Barradas, Maria C. MD††; Crane, Heidi M. MD‡‡; Gibert, Cynthia L. MD§§; Brown, Sheldon T. MD‖‖; Tindle, Hilary A. MD, MPH*; Warner, Alberta L. MD¶¶; Alcorn, Charles MA##; Skanderson, Melissa MSW***; Justice, Amy C. MD, PhD†††; Freiberg, Matthew S. MD, MSc‡‡‡

Author Information
JAIDS Journal of Acquired Immune Deficiency Syndromes: February 1, 2015 - Volume 68 - Issue 2 - p 209-216
doi: 10.1097/QAI.0000000000000419



With the advent of antiretroviral medications, persons with HIV are living long enough to face significant morbidity and mortality from chronic illness such as cardiovascular disease (CVD).1–5 Traditional CVD risk factors (eg, diabetes, hypertension, dyslipidemia, smoking), HIV-related risk factors (eg, renal disease), and other risk factors [eg, antiretroviral therapy (ART), substance abuse] contribute to increased risk of CVD in HIV-infected patients.6,7 Although traditional CVD risk factors are often assessed individually, there is strong evidence that they occur in clusters,8,9 which can be categorized as CVD risk factor profiles.10 Comparisons among infected and uninfected people with similar traditional CVD risk factor profiles are needed to more accurately estimate the independent effect of HIV on acute myocardial infarction (AMI) risk. One way to assess the independent effects of HIV versus comorbidity on CVD risk is to compare people with low traditional CVD risk factor burden or even optimal cardiac health, a phenomenon whose prevalence is low among uninfected people but unknown among HIV-infected people.11,12 Our objectives were to compare the association of HIV status and incident AMI within specific cardiac health profiles and to assess the prevalence of the optimal cardiac health profile by HIV status.


Subject Selection

The Veterans Aging Cohort Study Virtual Cohort (VACS VC) is a prospective longitudinal cohort of HIV-infected and age, gender, race/ethnicity, and clinical site matched uninfected participants who were identified from United States Department of Veterans Affairs (VA) administrative data in the fiscal years 1998–2003 using a modified existing algorithm.13

This cohort has been described in detail elsewhere.2,13 Briefly, this cohort consists of data from the immunology case registry, the VA HIV registry, the pharmacy benefits management database, the VA Decision Support System, the National Patient Care Database, and Health Factor data, which are data collected from physician clinical reminders within the VA electronic medical record system.

For this analysis, we considered all VACS VC participants alive and enrolled in VACS VC on or after 2003. The baseline was a participant's first clinical encounter on or after April 1, 2003. All participants were followed from their baseline date to an AMI event, death, or the last follow-up date. Participants were followed until December 31, 2009.

AMI event data were obtained from Medicare and the Ischemic Heart Disease Quality Enhancement Research Initiative, an initiative designed to improve the quality of care and health outcomes of Veterans with IHD.14 Subjects with prevalent CVD based on ICD-9 codes for AMI, unstable angina, cardiovascular revascularization, stroke or transient ischemic attack, peripheral vascular disease, or heart failure (N = 17,229)15,16 were excluded from all analyses. Given the J-shaped mortality curve associated with blood pressure (BP),17 those with systolic/diastolic BP less than 90/60 mm Hg were also excluded to avoid misclassifying people with hypotension as having optimal cardiac health when their low BP may be more reflective of poor overall health. After these exclusions, 81,322 Veterans (33% HIV+) were eligible for this study.

Independent Variable

Participants were categorized into mutually exclusive CVD risk profiles. Components of the risk profiles were diabetes, current smoking, total cholesterol, BP, hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA) reductase inhibitor use, and antihypertensive medication use (Table 1). Diabetes was identified using outpatient and clinical laboratory data collected closest to the baseline date. Specifically, diabetes was diagnosed using glucose measurements, use of insulin or oral hypoglycemic agents, and/or ≥1 inpatient and/or 2 outpatient ICD-9 codes.18 Smoking was measured from the VA Health Factors data.19 Cholesterol measurements were obtained from the VA Decision Support System. Systolic and diastolic BP was averaged across the 3 routine outpatient clinical BP measurements performed closest to the baseline date. HMG-CoA reductase inhibitor and antihypertensive medication use were based on pharmacy data.

Mutually Exclusive Risk Factor Categories

Cardiac health risk profiles were based on prior work10 and categorized as optimal, nonoptimal, elevated risk factors, and major risk factors (Table 1). Optimal cardiac health was defined as having no history of diabetes, not currently smoking, total cholesterol <180 mg/dL, and BP of 90–120/60–80 mm Hg without antihypertensive medication. Nonoptimal cardiac health was defined as having no history of diabetes, not currently smoking, total cholesterol of 180–199 mg/dL, and untreated BP of 120–139/80–89 mm Hg. Elevated risk factor profile was defined as no history of diabetes, not currently smoking, total cholesterol of 200–239 mg/dL, and untreated BP of 140–159/90–99 mm Hg. Major risk factors were defined as having 1, 2, or 3 or more of the following: diagnosis of diabetes, current smoking, use of HMG-CoA reductase inhibitors or untreated total cholesterol ≥240 mg/dL, or BP ≥160/100 mm Hg. Participants were placed in the highest risk category ascertainable. For example, someone with a BP of 120/80 mm Hg, total cholesterol of 190 mg/dL who smoked, and had no other major risk factors was considered to have 1 major CVD risk factor.

Participants with missing CVD risk factor data were categorized based only on nonmissing data and were placed in the highest risk profile ascertainable. For example, a smoker with diabetes, no other major risk factors, and missing cholesterol would be categorized as having 2 major risk factors; however, the missing cholesterol could be in the major risk factor range.

Dependent Variable

The protocol for incident AMI determination has previously been described.2 Briefly, we determined AMI incidence using adjudicated VA data, and Medicare and death certificate data. Documentation of AMI in the discharge summary, along with a review of the VA physician notes and medical chart (including elevation of serum markers of myocardial damage and EKG findings), was required to confirm diagnosis of AMI. For participants with non-VA AMI events who were not transferred to the VA, we used ICD-9 code, 410, which had strong agreement with adjudicated AMI outcomes in the Cardiovascular Health Study.15 Using Cardiovascular Health Study criteria, fatal AMI was designated as definite or possible fatal AMI as previously described.2 Definite fatal AMI was defined as a death within 4 weeks of a clinically confirmed AMI, and possible fatal AMI was determined by death certificate documenting AMI as the underlying cause (ICD-10 code: I21.0–I21.9). The following were used to identify deaths: VA vital status file, the Social Security Administration death master file, the Beneficiary Identification and Records Locator Subsystem, and the Veterans Health Administration medical Statistical Analysis Systems inpatient data sets. Causes of death were obtained from the National Death Index.


Covariates included sociodemographic data (age, sex, and race/ethnicity). Body mass index was measured from Health Factors data; renal disease and anemia were measured using outpatient and clinical laboratory data collected closest to the baseline date. Renal disease was defined as an estimated glomerular filtration rate (eGFR) of less than 60 mL·min−1·1.73 m−2 per National Kidney Foundation Kidney Disease Outcomes Quality Initiative thresholds for chronic kidney disease.20 Hepatitis C virus (HCV) infection was defined as a positive HCV antibody test or ≥1 inpatient and/or ≥2 outpatient ICD-9 codes for this diagnosis.21 History of cocaine and alcohol abuse or dependence was defined using ICD-9 codes.22

We obtained data on HIV-1 RNA, CD4+ T-lymphocyte counts (CD4+ cell counts), and current use of ART. CD4+ cell counts and HIV-1 RNA measurements were obtained as part of clinical care within 180 days of our baseline date. ART was categorized by regimens defined as protease inhibitors plus nucleoside reverse-transcriptase inhibitors (NRTI), nonnucleoside reverse-transcriptase inhibitors plus NRTI, other, and no ART use (ie, referent group). We included all ART medications that were on VA formulary during the study period. A prior study using a nested sample demonstrated that 96% of HIV+ Veterans on ART obtain their medications from the VA.13

Statistical Analysis

We compared baseline characteristics by CVD risk factor profile using the χ2 and Kruskal–Wallis tests. We used similar tests to compare baseline characteristics by HIV status and CVD risk factor profile. We calculated average AMI rates across the study period and performed Cox proportional hazards regression to estimate the independent effect of CVD risk factor profile and HIV status on AMI risk. The referent group for the Cox analyses consisted of those with no major CVD risk factors (ie, those with an optimal, 1+ nonoptimal, and 1+ elevated CVD risk factor profile). These analyses were adjusted for age, race/ethnicity, HCV infection, body mass index, eGFR, history of cocaine abuse/dependence, and alcohol abuse/dependence. Models restricted to HIV-infected people were additionally adjusted for CD4+ cell count, HIV-1 RNA, and ART regimen at baseline.


Over a median follow-up of 5.9 [mean (SD): 4.9 (2.0)] years, 858 AMI events occurred (42% were among HIV-infected Veterans). Less than 2% of the cohort had optimal cardiac health (58% in optimal group were HIV infected). Twelve percent of the cohort had no major CVD risk factors, 46% had 1 major CVD risk factor, 20% had 2 major CVD risk factors, and 7% had 3 major CVD risk factors. HIV-infected Veterans had a higher prevalence of having a single major CVD risk factor and uninfected Veterans had a higher prevalence of multiple major CVD risk factors (Table 2).

TABLE 2-a:
Baseline Characteristics by CVD Risk Factor Profile Stratified by HIV Status
TABLE 2-b:
Baseline Characteristics by CVD Risk Factor Profile Stratified by HIV Status

In this cohort, compared to those with optimal CVD risk profiles, those with 1 or more major CVD risk factors were older, more likely to be black, obese (Table 2), and have low-density lipoprotein cholesterol ≥160 mg/dL (0.4% vs. 12.6%), triglycerides ≥150 mg/dL (25.0% and 43.5%), renal disease (eGFR <60; 3.9% vs. 5.9%), and a history of cocaine (4.9% vs. 10.5%) or alcohol abuse (7.1% vs. 16.4%), respectively. Among HIV-infected Veterans, immune depletion (CD4+ cell count <200 cells/mm3) and unsuppressed viremia (HIV-1 RNA ≥500 copies/mL) were more common among those in the optimal cardiac health group compared with other groups (Table 2). Veterans with only 1 major CVD risk factor risk were likely to have smoking as their 1 major risk factor. Those with 2 major risk factors were often diabetic smokers, whereas those with 3 major risk factors were typically diabetic smokers taking HMG-CoA reductase inhibitors (Table 2).

An optimal CVD risk profile was associated with low AMI rates {6.0/10,000 person-years (py) [95% confidence interval (CI): 1.9 to 18.8]; age/race/ethnicity adjusted}. Veterans with 1, 2, 3, or more major CVD risk factors had significantly higher AMI rates [18.5/10,000 py (95% CI: 15.7 to 21.8), 34.5/10,000 py (29.2 to 40.9), 42.5/10,000 py (95% CI: 34.4 to 52.6), respectively] compared with those with optimal CVD risk factors. Compared with uninfected people with the same CVD risk factor profile, HIV-infected Veterans had higher AMI rates (age/race/ethnicity adjusted), particularly among those with at least 1 major CVD risk factor present (Fig. 1). The CVD risk factor categorization was based on prior work and only considered current smoking (and not past smoking) as a major CVD risk factor. A sensitivity analysis excluding past smokers showed very similar absolute AMI rates overall and by HIV status (see Figure S1, Supplemental Digital Content,

Age-/race-/ethnicity-adjusted rates of acute myocardial infarction (AMI) by cardiovascular disease risk factor profile (CVDRF) stratified by HIV status.

Compared with those without major CVD risk factors, both HIV-infected and uninfected Veterans showed a stepwise increase in AMI risk with increasing number of major CVD risk factors (Table 3). Compared with uninfected people with no major CVD risk factors, HIV-infected people with no major CVD risk factors had a 2-fold increased risk of AMI (HR: 2.1; 95% CI: 1.1 to 4.0; Table 4). This association was slightly attenuated after covariate adjustment (HR: 2.0; 95% CI: 1.0 to 3.9, P = 0.044; Table 4).

AMI Risk by Cardiovascular Disease Risk Factor (CVDRF) Profile Stratified by HIV Status (Separate Referent Groups for HIV− and HIV+)
Rates and Risk of AMI by Cardiovascular Disease Risk Factor (CVDRF) and HIV Status (Common Referent Group)

Sensitivity analyses limiting the sample to those without missing cholesterol, smoking, or BP data still showed increased AMI risk among HIV-infected compared with uninfected people with similar CVD risk factors (see Table S1, Supplemental Digital Content,


Among Veterans without major CVD risk factors, HIV-infected Veterans had a 2-fold increased risk of AMI compared with uninfected Veterans. The prevalence of optimal cardiac health was low in this population of Veterans, regardless of their HIV status. The presence of any major CVD risk factors was associated with a 2- to 7-fold increased risk of AMI regardless of HIV status.

Our results support prior observations in the general population showing lowest CVD risk among those with optimal cardiac health and increased risk among those with major CVD risk factors present.10,11,23 Prior studies have described increased risk for AMI and other CVDs among HIV-infected compared with uninfected people.2,24–26 These analyses typically adjusted for CVD risk factors individually. Risk factor clustering has been of increasing importance in CVD research in the general population.27–29 This study supports these findings and extends them by specifically comparing HIV-infected to uninfected people with similar levels of global cardiovascular risk. Our findings suggest that the rates of AMI with increasing burden of CVD risk factors are significantly higher among HIV infected with at least 1 major CVD risk factor compared with uninfected people with at least 1 major CVD risk factor. For example, HIV-infected Veterans with 3 or more major CVD risk factors had absolute AMI rates that were 30 events per 10,000 py higher than those for uninfected Veterans with the same CVD risk factor profile compared with 20 and 7 events per 10,000 py for those with 2 or 1 major CVD risk factors, respectively (Fig. 1).

Although optimal health was associated with lower AMI risk overall, among HIV-infected Veterans, it was not associated with an optimal HIV biomarker profile. As compared with HIV-infected Veterans with a higher burden of CVD risk factors, those with an optimal profile were more likely to have HIV-1 RNA ≥500 copies per milliliter or CD4+ count <200 cells per cubic millimeter. Although the reason for this finding is not clear, HIV seroconversion without initiation of or with poor adherence to ART is associated with decreases in low-density lipoprotein and total cholesterol and weight loss.30–32 Those with poor HIV control may have had more extreme decreases in these lipids and weight loss making them appear healthier from a traditional CVD risk factor perspective. However, their risk is likely higher than that of uninfected veterans because of independent effect of an unsuppressed HIV viremia on AMI risk.2

Our findings have important clinical implications for reducing AMI risk in the HIV population. First, optimal cardiac health is rare, however, associated with a very low rate of AMI. These results suggest that interventions focusing on primary prevention of CVD risk factors in this population are needed. Second, most HIV-infected Veterans have CVD risk factors, and increasing risk factor burden substantially increases AMI risk. These results suggest that future studies comparing various strategies for the implementation of CVD risk factor management in the HIV population are also needed. For example, comparing whether managing all CVD risk factors equally and simultaneously is more effective in reducing CVD risk among HIV-infected people than a personalized and prioritized approach is an important area of research. The latter approach has been suggested as a means of improving outcomes in a health care environment where clinicians rarely have time to fully evaluate and implement all recommended clinical guidelines.33 Furthermore, in this health care environment, polypharmacy among those with multimorbidity is common and associated with decreased medication adherence, serious adverse drug events, organ system injury, hospitalization, and mortality.34

This study has limitations that warrant discussion. Missing data on CVD risk factors may have led to some misclassification in assigning CVD risk factor profiles. However, it is unlikely that these Veterans with missing data had optimal cardiac health because the rates and risk of AMI in the missing risk factor group were more consistent with those for Veterans who had 1 major CVD risk factor. Furthermore, sensitivity analyses excluding participants with missing cholesterol, smoking, and BP data did not change our conclusions. Our analyses do not consider changes in AMI risk factor management, development of new AMI risk factors over time, duration of risk factor prevalence, or treatment heterogeneity within risk factor categories. As the number of women in the VACS VC is small, our findings may not be generalizable to women.


In conclusion, less than 2 percent of HIV-infected and uninfected Veterans have an optimal cardiac profile, whereas almost 75% have at least 1 or more major CVD risk factors. Compared with HIV− veterans, AMI rates among HIV+ veterans with the same CVD risk factor profile were higher and increased faster with each additional major CVD risk factor. Preventing or reducing AMI risk factor burden may result in a substantial reduction in AMI risk among HIV-infected people. Future studies, therefore, should focus on new strategies and/or compare current implementation strategies designed to prevent and manage existing CVD risk factors in this high-risk population.


The authors acknowledge the veterans who participate in the Veterans Aging Cohort Study and the study coordinators and staff at each of our sites and at the West Haven Coordinating Center. Without the commitment and care of these individuals, this research would not be possible. The authors would also like to acknowledge the substantial in-kind support we receive from the Veterans Affairs Healthcare System.


1. Oramasionwu CU, Morse GD, Lawson KA, et al.. Hospitalizations for cardiovascular disease in African Americans and Whites with HIV/AIDS. Popul Health Manag. 2013;16:201–207.
2. Freiberg MS, Chang C-CH, Kuller LH, et al.. HIV infection and the risk of acute myocardial infarction. JAMA Intern Med. 2013;173:614–622.
3. Esser S, Gelbrich G, Brockmeyer N, et al.. Prevalence of cardiovascular diseases in HIV-infected outpatients: results from a prospective, multicenter cohort study. Clin Res Cardiol. 2013;102:203–313.
4. Durand M, Sheehy O, Baril J-G, et al.. Association between HIV infection, antiretroviral therapy, and risk of acute myocardial infarction: a cohort and nested case-control study using Québec's public health insurance database. J Acquir Immune Defic Syndr. 2011;57:245–253.
5. Triant VA, Lee H, Hadigan C, et al.. Increased acute myocardial infarction rates and cardiovascular risk factors among patients with human immunodeficiency virus disease. J Clin Endocrinol Metab. 2007;92:2506–2512.
6. D'Agostino RB. Cardiovascular risk estimation in 2012: lessons learned and applicability to the HIV population. J Infect Dis. 2012;205(Suppl):S362–S367.
7. Schillaci G, Maggi P, Madeddu G, et al.. Symmetric ambulatory arterial stiffness index and 24-h pulse pressure in HIV infection: results of a nationwide cross-sectional study. J Hypertens. 2013;31:560–567.
8. Genest J, Cohn JS. Clustering of cardiovascular risk factors: targeting high-risk individuals. Am J Cardiol. 1995;76:8A–20A.
9. Bøg-Hansen E, Lindblad U, Bengtsson K, et al.. Risk factor clustering in patients with hypertension and non-insulin-dependent diabetes mellitus. The Skaraborg Hypertension Project. J Intern Med. 1998;243:223–232.
10. Lloyd-Jones DM, Leip EP, Larson MG, et al.. Prediction of lifetime risk for cardiovascular disease by risk factor burden at 50 years of age. Circulation. 2006;113:791–798.
11. Berry JD, Dyer A, Cai X, et al.. Lifetime risks of cardiovascular disease. N Engl J Med. 2012;366:321–329.
12. Jousilahti P, Tuomilehto J, Korhonen HJ, et al.. Trends in cardiovascular disease risk factor clustering in eastern Finland: results of 15-year follow-up of the North Karelia Project. Prev Med. 1994;23:6–14.
13. Fultz SL, Skanderson M, Mole LA, et al.. Development and verification of a “virtual” cohort using the National VA Health Information System. Med Care. 2006;44:S25–S30.
14. Every NR, Fihn SD, Sales AE, et al.. Quality Enhancement Research Initiative in ischemic heart disease: a quality initiative from the Department of Veterans Affairs. QUERI IHD Executive Committee. Med Care. 2000;38:I49–I59.
15. Ives DG, Fitzpatrick AL, Bild DE, et al.. Surveillance and ascertainment of cardiovascular events. The Cardiovascular Health Study. Ann Epidemiol. 1995;5:278–285.
16. Petersen LA, Wright S, Normand SL, et al.. Positive predictive value of the diagnosis of acute myocardial infarction in an administrative database. J Gen Intern Med. 1999;14:555–558.
17. Boutitie F, Gueyffier F, Pocock S, et al.. J-shaped relationship between blood pressure and mortality in hypertensive patients: new insights from a meta-analysis of individual-patient data. Ann Intern Med. 2002;136:438–448.
18. Butt AA, Fultz SL, Kwoh CK, et al.. Risk of diabetes in HIV infected veterans pre- and post-HAART and the role of HCV coinfection. Hepatology. 2004;40:115–119.
19. McGinnis KA, Brandt CA, Skanderson M, et al.. Validating smoking data from the Veteran's Affairs Health Factors dataset, an electronic data source. Nicotine Tob Res. 2011;13:1233–1239.
20. K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis. 2002;39:S1–S266.
21. Goulet JL, Fultz SL, McGinnis KA, et al.. Relative prevalence of comorbidities and treatment contraindications in HIV-mono-infected and HIV/HCV-co-infected veterans. AIDS. 2005;19(suppl 3):S99–S105.
22. Kraemer KL, McGinnis KA, Skanderson M, et al.. Alcohol problems and health care services use in human immunodeficiency virus (HIV)-infected and HIV-uninfected veterans. Med Care. 2006;44:S44–S51.
23. Wilkins JT, Ning H, Berry J, et al.. Lifetime risk and years lived free of total cardiovascular disease. JAMA. 2012;308:1795–1801.
24. Savès M, Chêne G, Ducimetière P, et al.. Risk factors for coronary heart disease in patients treated for human immunodeficiency virus infection compared with the general population. Clin Infect Dis. 2003;37:292–298.
25. Kaplan RC, Kingsley LA, Sharrett AR, et al.. Ten-year predicted coronary heart disease risk in HIV-infected men and women. Clin Infect Dis. 2007;45:1074–1081.
26. Currier JS, Taylor A, Boyd F, et al.. Coronary heart disease in HIV-infected individuals. J Acquir Immune Defic Syndr. 2003;33:506–512.
27. Mancia G. Total cardiovascular risk: a new treatment concept. J Hypertens Suppl. 2006;24:S17–S24.
28. Kadota A, Hozawa A, Okamura T, et al.. Relationship between metabolic risk factor clustering and cardiovascular mortality stratified by high blood glucose and obesity: NIPPON DATA90, 1990-2000. Diabetes Care. 2007;30:1533–1538.
29. Weycker D, Nichols GA, O'Keeffe-Rosetti M, et al.. Risk-factor clustering and cardiovascular disease risk in hypertensive patients. Am J Hypertens. 2007;20:599–607.
30. Riddler SA, Smit E, Cole SR, et al.. Impact of HIV infection and HAART on serum lipids in men. JAMA. 2003;289:2978–2982.
31. Brown TT, Chu H, Wang Z, et al.. Longitudinal increases in waist circumference are associated with HIV-serostatus, independent of antiretroviral therapy. AIDS. 2007;21:1731–1738.
32. Brown TT, Xu X, John M, et al.. Fat distribution and longitudinal anthropometric changes in HIV-infected men with and without clinical evidence of lipodystrophy and HIV-uninfected controls: a substudy of the Multicenter AIDS Cohort Study. AIDS Res Ther. 2009;6:8.
33. Taksler GB, Keshner M, Fagerlin A, et al.. Personalized estimates of benefit from preventive care guidelines: a proof of concept. Ann Intern Med. 2013;159:161–168.
34. Edelman EJ, Gordon KS, Glover J, et al.. The next therapeutic challenge in HIV: polypharmacy. Drugs Aging. 2013;30:613–628.

HIV; optimal cardiovascular health; myocardial infarction

Supplemental Digital Content

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.