Chronic HIV infection is associated with immune activation, inflammation in various tissues, changes in hemostatic balance, and potentially with cardiovascular disease (CVD) risk.1 Inflammatory and hemostatic biomarkers seem to provide clinically meaningful information about how HIV infection affects the host.2,3 Further study is important for several reasons. First, treated and untreated HIV-infected individuals, even at high CD4+ T-cell counts, seem to have higher risk of mortality compared with HIV-uninfected individuals, and inflammation remains an important risk factor for death.4–6 Identification of specific inflammation and coagulation mediators involved in HIV disease would contribute understanding to these mechanisms of comorbidity. Second, biomarkers might supplement those used in clinical practice to predict patient prognosis including future risk of CVD events. Third, biomarkers that can be reproducibly shown to predict complications of HIV infection may suggest modalities of future targeted therapy. Fourth, although prior studies have described inflammation and coagulation biomarkers in HIV-infected women, few have included a group of HIV-uninfected controls who are well matched for other factors that may influence these markers.7–10 Finally, serum biomarkers and measures of subclinical atherosclerosis such as carotid artery intima–media thickness (CIMT) are potential endpoints that can be used to examine whether elevated CVD risk might persist even in effectively treated HIV-infected adults.
Among participants in the Women's Interagency HIV Study (WIHS), we examined the association of HIV infection and first initiation of highly active antiretroviral therapy (HAART), with biomarkers of inflammation and hemostasis including proinflammatory cytokines such as tumor necrosis factor (TNF) alfa and interleukin (IL) 6; soluble IL-2 receptor; IL-10, an anti-inflammatory TH2 cytokine produced by monocytes/macrophages and regulatory T cells (Tregs); monocyte chemoattractant protein 1 (MCP-1)/CCL2, a chemokine that has been associated with atherosclerosis in the HIV-infected population; soluble CD14, a pathogen recognition receptor expressed by monocytes that predicts CVD events among HIV-infected patients; D-dimer, which is produced during the degradation of the fibrin clot; and fibrinogen, an inflammatory marker that functions as a modulator of platelet and coagulation protein activity and is a fibrin precursor. Among women initiating HAART, we further examined the association of biomarker level both before and after initiating treatment, with subclinical atherosclerosis.
MATERIALS AND METHODS
Study Design and Variable Definition
The WIHS cohort enrolled 3766 HIV-infected and HIV-uninfected women who were recruited in 2 waves (1994–1995 and 2001–2002) at 6 US field centers. Every 6 months, WIHS participants are scheduled for study examinations, which involve collection of interview-administered questionnaire data, physical measurements, and biospecimens.11,12 Institutional review board approval and informed consent were obtained on all participants.
Using medication questionnaire data collected at each semiannual visit, we identified 769 HIV-infected women who first reported use of HAART without any prior reported use of antiretroviral therapy while enrolled in WIHS. Of these women, we identified 127 who had provided blood specimens at 6 consecutive semiannual WIHS study examinations, including 3 before and 3 after the use of HAART. Compared with the HIV-infected women in WIHS excluded from the study, the 127 women included had a lower median baseline age (33 years vs 35 years), but were similar in terms of race/ethnicity, smoking status, body mass index (BMI), and hepatitis C virus (HCV) antibody status. Our comparison population consisted of a group of HIV-uninfected women enrolled in WIHS who were individually matched to the HIV-infected women using propensity score matching that accounted for age, race/ethnicity, BMI, smoking status, HCV antibody status, and calendar time.13 To find the best matched HIV-uninfected woman for all 127 of the HIV-infected women who initiated HAART, 29 HIV-uninfected women were selected more than once; for these women, we chose different sequences of 6 visits that were matched with different HIV-infected HAART initiators. HAART was defined as reported use of combination therapy in accordance with US Department of Health and Human Services guidelines, with major classes defined by cornerstone therapies: protease inhibitors, nonnucleoside reverse transcriptase inhibitors, and nucleoside reverse transcriptase inhibitors.
Carotid Artery Ultrasound
During 2004–2005, we obtained B-mode ultrasound carotid artery measurements of the intima–media thickness of the far wall of the right common carotid artery (CIMT). Standardized carotid artery ultrasound images were centrally measured by automated computerized edge detection software (Patents, 2005, 2006, 2011).14–16
HIV infection was determined via serologic testing using enzyme-linked immunosorbent assay and confirmed using Western blot assays. Plasma HIV RNA levels were quantified using nucleic acid sequence–based amplification commercial assays with a lower limit of quantification of 80 copies per milliliter (bioMérieux, Boxtel, the Netherlands), and total peripheral CD4+ T-cell counts were measured with standard flow cytometric methods. HCV antibody testing was performed using enzyme immunoassays (version 2.0 or 3.0; Abbott, Abbott Park, IL). HCV RNA levels were measured by a polymerase chain reaction (Roche Diagnostics, Indianapolis, IN).17 Enzyme-linked immunosorbent assay methods were used to measure soluble IL-2 receptor (DR2A00; R&D Systems, Minneapolis, MN), IL-6 (Q6000B; R&D Systems), and soluble CD14 (DC140; R&D Systems). Fibrinogen activity was measured using a clot-based assay (00674) and D-dimer using immunoturbidimetric methods (00515) (Stago Diagnostics, Parsippany, NJ). MCP-1, TNF-α, and IL-10 were measured on a bead-based immunoassay multiplex platform (MPXHCYTO-60K-04 Cytokine Panel; Millipore Corporation, Billerica, MA). We previously described an association between HAART initiation and increased C-reactive protein levels in the WIHS cohort and did not repeat this measurement in the present study.10
We compared characteristics of the HIV-infected and HIV-uninfected women using χ2 and Mann–Whitney tests, for categorical and continuous variables, respectively. We examined summary statistics of each biomarker and correlations among inflammatory and hemostasis biomarkers. We also examined the association of biomarker levels with current and nadir CD4+ T-cell count, HIV RNA, and clinical characteristics including age, HCV antibody status, smoking, and BMI. We computed median biomarker values and Wilcoxon signed rank tests to compare HIV-infected and HIV-uninfected women and to examine changes in biomarkers associated with first use of HAART among the HIV-infected group. Linear mixed-effects models were also used to assess statistical significance of biomarker level comparisons, summarized across all pre-HAART and post-HAART visits. Models incorporated random intercept and slope terms for each individual to account for the within-individual correlation among the 6 contributed measurements and heterogeneous rates of change across individuals. Results obtained using the linear mixed-effects models were similar to those from the Wilcoxon signed rank test–based comparisons of the raw data, so are not presented.
After comparing biomarker levels in HIV-infected women before treatment and after HAART initiation relative to matched HIV-uninfected women, we repeated analyses while limiting visits among HIV-infected women to those where they were treated and aviremic (HIV RNA below 80 copies/mL). HCV was a common coinfection that may contribute to inflammation or influence production of liver-derived coagulation and inflammation markers; therefore, we used stratified analysis and tests of statistical interaction to examine whether HCV coinfection (HCV RNA+ at study entry) was an effect modifier of the associations of HIV infection status and use of HAART with inflammation and coagulation biomarkers. Finally, we used multivariable linear regression to assess the difference in CIMT in micrometers associated with a 10% increase in biomarker level, in models that were adjusted for age, race/ethnicity, smoking, BMI, and time between biomarker measurements and carotid artery ultrasound visit. Analyses were performed using SAS version 9.2 (Cary, NC).
HIV-infected and HIV-uninfected women were both of median age 37 years and were similar on other matching factors (Table 1). In both groups, well over half of women were overweight or obese, approximately half were current smokers, and nearly one-third had detectable HCV antibodies.
All HIV-infected women initiated HAART while under study observation. The most common antiretroviral medications used were, in the protease inhibitor class: nelfinavir (used at 16% of treated study visits), indinavir (16%), ritonavir (14%), and atazanavir (11%); in the nonnucleoside reverse transcriptase inhibitor class: efavirenz (20%) and nevirapine (16%); in the nucleoside reverse transcriptase inhibitor class: lamivudine (68%), zidovudine (44%), tenofovir (25%), emtricitabine (22%), stavudine (21%), and abacavir (11%); and fixed-dose combinations: Combivir (lamivudine and zidovudine, 23%; Glaxo SmithKline, Research Triangle Park, NC) and Truvada (tenofovir and emtricitabine, 17%; Gilead Sciences, Foster City, CA).
CD4+ T-Cell Count and HIV RNA
Our analyses used data from 6 sequential semiannual visits that spanned a mean period of 2.5 years (range 2.1–2.8 years). Among HIV-infected women, mean CD4+ T-cell count was 414 cells per cubic millimeter at the first of the 6 visits, which was 12–18 months before first use of HAART (Fig. 1). Among these women, mean CD4+ T-cell count was 332 cells per cubic millimeter at the third visit, which was the last visit before initiation of HAART. After HAART initiation, mean CD4+ T-cell count among HIV-infected women reached a high of 479 cells per cubic millimeter at the sixth visit, which was 12–18 months after HAART initiation. Of the 127 HAART initiators, 31 failed to achieve a CD4+ T-cell count above 350 cells per cubic millimeter at any of the 3 semiannual visits after HAART initiation.
Mean log10 HIV RNA was 4.4 at the last visit before HAART initiation (Fig. 1). Viral suppression to <80 copies per milliliter was achieved by 43%, 58%, and 51% of women at the first, second, and third semiannual visit after initiation of HAART, respectively.
Correlation of Biomarkers With Clinical Variables
In cross-sectional analyses, higher HIV RNA was associated with higher levels of TNF-α, soluble IL-2 receptor, IL-10, MCP-1, and D-dimer, both pre and post HAART (see Table, Supplemental Digital Content 1, http://links.lww.com/QAI/A328). Cross sectionally, both lower current and lower nadir CD4+ T-cell counts were associated with higher levels of soluble CD14, TNF-α, IL-6, and IL-10 both before and after initiating HAART; only nadir CD4+ T-cell count was associated with pre- and post-HAART soluble IL-2 receptor levels (see Table, Supplemental Digital Content 2, http://links.lww.com/QAI/A328). Using the longitudinal data, we found that the degree of response to HAART, as defined by changes in CD4+ T-cell count and HIV RNA after treatment initiation, was significantly associated with the magnitude of changes in soluble CD14, TNF-α, soluble IL-2 receptor, IL-10, and MCP-1 levels (see Table, Supplemental Digital Content 3, http://links.lww.com/QAI/A328). Age was correlated with increased soluble CD14 (only before HAART initiation), IL-6, and fibrinogen. Number of years smoked was correlated with increased IL-6. BMI was correlated with increased post-HAART fibrinogen levels. Presence of HCV RNA was positively correlated with soluble CD14 (only before HAART initiation), IL-6, IL-10, TNF-α (only after HAART initiation), and soluble IL-2 receptor (only after HAART initiation) (see Table, Supplemental Digital Content 4, http://links.lww.com/QAI/A328).
Inflammation- and Hemostasis-Related Biomarkers: Association With Untreated HIV Infection
As compared with HIV-uninfected women, HIV-infected women in the period before HAART initiation had significantly higher levels of TNF-α (P < 0.0001), IL-10 (P < 0.0001), soluble IL-2 receptor (P < 0.0001), soluble CD14 (P < 0.0001), MCP-1 (P < 0.001), and D-dimer (P < 0.01) (Fig. 2). IL-6 and fibrinogen levels were not elevated among untreated HIV-infected compared with HIV-uninfected women.
Inflammation- and Hemostasis-Related Biomarkers: Effect of HAART Initiation
After HAART initiation, levels of TNF-α and soluble IL-2 receptor decreased among HIV-infected women, but over the period after HAART had been initiated, levels of these biomarkers still remained elevated among HIV-infected women compared with HIV-uninfected controls (Fig. 2). In contrast, after initiation of HAART, levels of MCP-1 and D-dimer decreased and were no longer elevated among HAART-treated HIV-infected women relative to matched HIV-uninfected controls. IL-10 levels decreased gradually and were no longer elevated at the third semiannual HAART-treated visit. Results for soluble CD14 varied significantly by HCV status, as described below. As compared with HIV-uninfected women, in treated HIV-infected women, IL-6 levels were similar or slightly reduced and fibrinogen levels were also similar.
Analyses of Women Achieving Viral Suppression on HAART
When we limited the analysis to 91 HAART-treated HIV-infected women who achieved HIV RNA <80 copies per milliliter, TNF-α levels remained elevated relative to HIV-uninfected controls. Levels of TNF-α were 4.2 pg/mL among HIV-infected women at HAART-treated visits where viral suppression was achieved, compared with 3.4 pg/mL among HIV-uninfected women (P = 0.0002). In this group of HAART-treated aviremic women, lower CD4+ T-cell count was significantly correlated with higher levels of TNF-α (r = −0.31, P = 0.002).
Analyses of Women With HCV Coinfection
HCV coinfection modified effects of HIV and HAART on levels of soluble CD14 (Pinteraction = 0.02). In the subgroup of HCV-infected women, HAART initiation reduced soluble CD14 levels in the HIV-infected group and also equalized soluble CD14 levels when comparing the HAART-treated HIV/HCV-coinfected women with the HIV-uninfected HCV-infected group (see Table and Figure, Supplemental Digital Contents 5 and 6, http://links.lww.com/QAI/A328). In contrast, soluble CD14 remained persistently elevated with HIV infection and was unaffected by HAART in the group of HCV-uninfected women. For all other biomarkers, the effects of HIV infection or HAART did not differ across HCV-coinfected and non–HCV-coinfected subgroups (data not shown).
Inflammation- and Hemostasis-Related Biomarkers: Association With Subclinical Atherosclerosis
Among the 127 HIV-infected women who initiated HAART, 81 had a carotid artery ultrasound measurement performed subsequent to the biomarker measurements. When measured before HAART initiation, biomarker levels were not associated with CIMT. By contrast, when the same biomarkers were measured after initiation of HAART, greater CIMT was associated with higher levels of soluble IL-2 receptor [per 10% higher level of biomarker, difference (Δ) in CIMT = 6.0 μm, 95% confidence interval (CI): 1.0 to 11.0 μm, P = 0.02], IL-6 (Δ CIMT = 3.1 μm, 95% CI: −0.1 to 6.3 μm, P = 0.05), D-dimer (Δ CIMT = 3.5 μm, 95% CI: 0.4 to 6.6 μm, P = 0.03), and (of borderline significance) MCP-1 (Δ CIMT = 4.2 μm, 95% CI: −0.5 to 9.0, P = 0.08) (Table 2). Other inflammation- and hemostasis-related markers measured after HAART initiation were not associated with CIMT. These analyses were adjusted for confounders including age, race/ethnicity, smoking, and BMI; further adjustment for CD4+ T-cell count, HIV RNA, and antiretroviral drug class did not change the results appreciably.
We characterized biomarkers of inflammation and hemostasis before and after first use of HAART (Table 3 summarizes key findings). As compared with HIV-uninfected controls, HIV-infected women studied in the 18 months before HAART initiation had increased circulating levels of the macrophage pathogen recognition receptor CD14, several inflammation-related cytokines (TNF-α, IL-10), the chemokine MCP-1/CCL2, soluble IL-2 receptors, and the fibrin clot degradation marker D-dimer. Initiation of HAART tended to normalize levels of most inflammation and hemostasis biomarkers, reducing them among women using effective antiretroviral treatment to levels observed in HIV-uninfected controls. On the other hand, elevated levels of TNF-α persisted even in HIV-infected women who were treated with HAART and had viral suppression to below detectable limits (HIV RNA < 80 copies/mL). Finally, HIV-infected women with higher levels of IL-6, soluble IL-2 receptors, and D-dimer while on HAART had significantly higher subclinical atherosclerosis as measured by CIMT. Thus, although HAART may cause adverse metabolic disturbances, this may be balanced in effectively treated patients by improvements in other CVD-related risk markers.
High IL-6 and D-dimer levels are well-known risk factors for CVD in individuals free of HIV infection18 and for mortality in HIV-infected patients.2 We extend these prior findings by linking these biomarkers with subclinical atherosclerosis (CIMT) in treated HIV-infected women. In addition, we report an association of higher soluble IL-2 receptor levels with treated HIV infection and CIMT. IL-2 is a cytokine mainly secreted by T cells; the biologic relevance of circulating soluble IL-2 receptors is uncertain. However, elevated soluble IL-2 receptors is a potential marker for T-cell activation,19 and our present findings bolster prior evidence linking T-cell activation in HIV-seropositive individuals with preclinical vascular disease.20,21 The association between MCP-1 levels, another biomarker that was increased in HIV-infected women, and CIMT was of borderline statistical significance but is consistent with prior evidence, suggesting a role for MCP-1 in elevated CVD risk among HIV-infected adults.22,23 This observation may be explained by the involvement of MCP-1, which is expressed by smooth muscle cells infected by HIV,24 in transmigration of monocytes from the circulation into the subendothelium. Although inflammation and coagulation markers were associated with increased CIMT among HAART-treated HIV-infected women, the same measures were not associated with CIMT when measured before the first use of HAART. Possibly, in HAART-treated patients, the presence of continued inflammation and activated coagulation despite effective HIV therapy reflects a host factor, an HIV factor, or cofactor (eg, viral coinfection) that persistently contributes to CVD over the duration of treated HIV infection or reflects the inflammation associated with atherosclerosis itself.
Beyond IL-6, D-dimer, soluble IL-2 receptors, and MCP-1, several other inflammation-related biomarkers have been previously associated with CVD risk, even though we did not find that they predicted CIMT in our study. TNF-α, a proinflammatory cytokine expressed by cells of both the innate immune system (macrophages) and adaptive immune system (T cells), remained elevated in HIV-infected women who were using HAART and had undetectable or very low levels of circulating HIV RNA. TNF-α drives inflammation and apoptosis in HIV infection and probably contributes to propagation of HIV replication.25 TNF-α levels predict risk of CVD events in the non–HIV-infected population26 and might therefore be hypothesized as a CVD risk factor in untreated and treated HIV-infected populations, despite the lack of association with CIMT in our study. Although several proinflammatory biomarkers were elevated with HIV infection, at the same time, HIV infection was also associated with elevated levels of IL-10, which is an anti-inflammatory cytokine secreted by TH2 cells, Tregs, and monocytes/macrophages. IL-10 has potential antiatherogenic properties,27 and high IL-10 production is associated with reduced risk of stroke in non–HIV-infected populations.28 Although no association between IL-10 and CIMT was observed in the present cohort, it remains possible that IL-10 or other inflammatory mediators invoked by HIV infection and reduced by HAART may possibly reduce CVD in certain HIV-infected individuals.
We found that circulating soluble CD14, a pathogen recognition receptor expressed by monocytes, was elevated in patients who had HIV infection, HCV infection, or both. Because CD14 is produced by hepatocytes,29 it is therefore unclear whether high levels of soluble CD14 may reflect liver function, receptor shedding from activated monocytes, and/or other aspects of HIV-related and non–HIV-related disease processes. Soluble CD14 levels predict increased risk of CVD events and immunologic disease progression in HIV-infected populations.30,31 Accordingly, it will be important to understand why soluble CD14 levels remain elevated in HIV monoinfected and HIV/HCV-coinfected individuals even with antiretroviral therapy. Although no overall association between soluble CD14 and CIMT was observed, we lacked statistical power to analyze this association in HCV-infected and HCV-uninfected subgroups, which is an important limitation.
Prior evidence describing inflammation and hemostatic perturbation in HIV-infected adults is only partially consistent with the present data. Among participants in the Strategies for Management of Antiretroviral Therapy (SMART) trial, elevated levels of IL-6 and D-dimer were observed in treated, well-controlled, HIV-infected patients compared with controls from external population-based cohorts.8 In contrast, among HIV-infected women in the present study, IL-6 levels were not elevated in either the untreated or the treated phase of HIV infection and D-dimer levels were similar between HAART-treated HIV-infected women and HIV-uninfected women. However, we did find an association in post-HAART levels of both biomarkers with greater CIMT, supporting the conclusions from the SMART study that these are important markers of vascular risk. It is important to note that our WIHS cohort was limited to women, although several studies have highlighted possible biomarker differences by sex.8,10,32
Given the nature of the WIHS cohort, caution should be used when generalizing the results and making comparisons with previously published reports. The female WIHS cohort participants are predominantly African American and Latina and have low socioeconomic status and a high burden of obesity, poor oral health,33 and smoking. These cofactors may interact with HIV infection to promote inflammation and aberrations in coagulation markers. Levels of several biomarkers were high in the HIV-infected and HIV-uninfected WIHS participants compared with previously described comparison populations.8,26,34 Many WIHS participants have impaired liver function due to hepatitis infection, medications, or obesity. Our results agree with prior evidence that liver function has an important role in influencing levels of many circulating inflammation and hemostasis-related biomarkers.8,32 Future studies are needed to address the additional contributions of factors such as liver fibrosis, renal dysfunction, diabetes, and hypercholesterolemia, which were not measured in the present study. In addition, WIHS is a cohort study where antiretroviral medication treatment strategies were chosen by physicians rather than determined by a study protocol. It should be noted that this study period includes use of several older antiretroviral drugs, which could both be less effective and more metabolically toxic than current drugs,35,36 and might therefore have contributed to the nonnormalization of biomarker levels among HAART-treated HIV-infected women. Finally, CIMT is a well-established measure of atherosclerosis but may not fully capture the effect of HIV on the risk of cardiovascular events.
In summary, our findings confirm and extend current concepts about immune, inflammatory, and hemostatic responses that are invoked by chronic HIV infection. These perturbations are observed even in the context of effective antiretroviral treatment. Additionally, higher post-HAART levels of certain inflammatory and hemostatic biomarkers were associated with increased subclinical CVD. Further investigation of inflammation and hemostatic mediators might identify pathways that can be targeted to ameliorate the long-term complications of HIV infection.
The authors would like to thank Dr. Alexandra Levine for her thoughtful insight and significant contributions to this article. Data in this article were collected by the WIHS Collaborative Study Group with centers (Principal Investigators) at New York City/Bronx Consortium (K.A.), Brooklyn, NY (Howard Minkoff), Washington, DC, Metropolitan Consortium (M.Y.); The Connie Wofsy Study Consortium of Northern California (Ruth Greenblatt), Los Angeles County/Southern California Consortium (Alexandra Levine); Chicago Consortium (Mardge Cohen); Data Coordinating Center (S.J.G.).
1. Triant VA, Grinspoon SK. Immune dysregulation and vascular risk in HIV-infected patients: implications for clinical care. J Infect Dis. 2011;203:439–441.
2. Kuller LH, Tracy R, Belloso W, et al.. Inflammatory and coagulation biomarkers and mortality in patients with HIV infection. PLoS Med. 2008;5:e203.
3. Rodger AJ, Fox Z, Lundgren JD, et al.. Activation and coagulation biomarkers are independent predictors of the development of opportunistic disease in patients with HIV infection. J Infect Dis. 2009;200:973–983.
4. Lodwick RK, Sabin CA, Porter K, et al.. Death rates in HIV-positive antiretroviral-naive patients with CD4 count greater than 350 cells per microL in Europe and North America: a pooled cohort observational study. Lancet. 2010;376:340–345.
5. Cockerham L, Scherzer R, Zolopa A, et al.. Association of HIV infection, demographic and cardiovascular risk factors with all-cause mortality in the recent HAART era. J Acquir Immune Defic Syndr. 2010;53:102–106.
6. Tien PC, Choi AI, Zolopa AR, et al.. Inflammation and mortality in HIV-infected adults: analysis of the FRAM study cohort. J Acquir Immune Defic Syndr. 2010;55:316–322.
7. Levine AM, Vigen C, Gravink J, et al.. Progressive prothrombotic state in women with advancing HIV disease. J Acquir Immune Defic Syndr. 2006;42:572–577.
8. Neuhaus J, Jacobs DR Jr, Baker JV, et al.. Markers of inflammation, coagulation, and renal function are elevated in adults with HIV infection. J Infect Dis. 2010;201:1788–1795.
9. Baker JV, Neuhaus J, Duprez D, et al.. Changes in inflammatory and coagulation biomarkers: a randomized comparison of immediate versus deferred antiretroviral therapy in patients with HIV infection. J Acquir Immune Defic Syndr. 2011;56:36–43.
10. Palella FJ Jr, Gange SJ, Benning L, et al.. Inflammatory biomarkers and abacavir use in the Women's Interagency HIV Study and the Multicenter AIDS Cohort Study. AIDS. 2010;24:1657–1665.
11. Barkan SE, Melnick SL, Preston-Martin S, et al.. The Women's Interagency HIV Study. WIHS Collaborative Study Group. Epidemiology. 1998;9:117–125.
12. Bacon MC, von Wyl V, Alden C, et al.. The Women's Interagency HIV Study: an observational cohort brings clinical sciences to the bench. Clin Diagn Lab Immunol. 2005;12:1013–1019.
13. Rosenbaum PR. Model-based direct adjustment. J Am Stat Assoc. 1987;82:387–394.
14. Selzer RH, Hodis HN, Kwong-Fu H, et al.. Evaluation of computerized edge tracking for quantifying intima-media thickness of the common carotid artery from B-mode ultrasound images. Atherosclerosis. 1994;111:1–11.
15. Hodis HN, Mack WJ, Lobo RA, et al.. Estrogen in the prevention of atherosclerosis. A randomized, double-blind, placebo-controlled trial. Ann Intern Med. 2001;135:939–953.
16. Selzer RH, Mack WJ, Lee PL, et al.. Improved common carotid elasticity and intima-media thickness measurements from computer analysis of sequential ultrasound frames. Atherosclerosis. 2001;154:185–193.
17. Operskalski EA, Mack WJ, Strickler HD, et al.. Factors associated with hepatitis C viremia in a large cohort of HIV-infected and -uninfected women. J Clin Virol. 2008;41:255–263.
18. Tracy RP. Thrombin, inflammation, and cardiovascular disease: an epidemiologic perspective. Chest. 2003;124:49S–57S.
19. Pett SL, Kelleher AD, Emery S. Role of interleukin-2 in patients with HIV infection. Drugs. 2010;70:1115–1130.
20. Kaplan RC, Sinclair E, Landay AL, et al.. T cell activation and senescence predict subclinical carotid artery disease in HIV-infected women. J Infect Dis. 2011;203:452–463.
21. Kaplan RC, Sinclair E, Landay AL, et al.. T cell activation predicts carotid artery stiffness among HIV-infected women. Atherosclerosis. 2011;217:207–213.
22. Floris-Moore M, Fayad ZA, Berman JW, et al.. Association of HIV viral load with monocyte chemoattractant protein-1 and atherosclerosis burden measured by magnetic resonance imaging. AIDS. 2009;23:941–949.
23. Alonso-Villaverde C, Coll B, Parra S, et al.. Atherosclerosis in patients infected with HIV is influenced by a mutant monocyte chemoattractant protein-1 allele. Circulation. 2004;110:2204–2209.
24. Eugenin EA, Morgello S, Klotman ME, et al.. Human immunodeficiency virus (HIV) infects human arterial smooth muscle cells in vivo and in vitro: implications for the pathogenesis of HIV-mediated vascular disease. Am J Pathol. 2008;172:1100–1111.
25. Herbein G, Khan KA. Is HIV infection a TNF receptor signalling-driven disease? Trends Immunol. 2008;29:61–67.
26. Cesari M, Penninx BW, Newman AB, et al.. Inflammatory markers and onset of cardiovascular events: results from the Health ABC study. Circulation. 2003;108:2317–2322.
27. Kelly JA, Griffin ME, Fava RA, et al.. Inhibition of arterial lesion progression in CD16-deficient mice: evidence for altered immunity and the role of IL-10. Cardiovasc Res. 2010;85:224–231.
28. van Exel E, Gussekloo J, de Craen AJ, et al.. Inflammation and stroke: the Leiden 85-Plus Study. Stroke. 2002;33:1135–1138.
29. Meuleman P, Steyaert S, Libbrecht L, et al.. Human hepatocytes secrete soluble CD14, a process not directly influenced by HBV and HCV infection. Clin Chim Acta. 2006;366:156–162.
30. Brenchley JM, Price DA, Schacker TW, et al.. Microbial translocation is a cause of systemic immune activation in chronic HIV infection. Nat Med. 2006;12:1365–1371.
31. Sandler NG, Wand H, Roque A, et al.. Plasma levels of soluble CD14 independently predict mortality in HIV infection. J Infect Dis. 2011;203:780–790.
32. Reingold J, Wanke C, Kotler D, et al.. Association of HIV infection and HIV/HCV coinfection with C-reactive protein levels: the fat redistribution and metabolic change in HIV infection (FRAM) study. J Acquir Immune Defic Syndr. 2008;48:142–148.
33. Mulligan R, Phelan JA, Brunelle J, et al.. Baseline characteristics of participants in the oral health component of the Women's Interagency HIV Study. Community Dent Oral Epidemiol. 2004;32:86–98.
34. Yan AT, Yan RT, Cushman M, et al.. Relationship of interleukin-6 with regional and global left-ventricular function in asymptomatic individuals without clinical cardiovascular disease: insights from the Multi-Ethnic Study of Atherosclerosis. Eur Heart J. 2010;31:875–882.
35. 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.
36. 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;49:369–376.