HIV and carotid atherosclerosis: A mediational model

Abstract

Cardiovascular disease (CVD) is the leading cause of death in the US and is a significant contributor to morbidity and mortality for people living with HIV (PLWH). This study examined the association between HIV infection, cocaine usage, and inflammatory markers, and their combined association with carotid atherosclerosis among young and middle-aged adults with HIV. Participants (N=494) were enrolled based on HIV status and cocaine use. Blood pressure, body mass index (BMI), and cocaine use were assessed. Cytokines and growth factors, IL-1a, IL-6, TNFα and VEGF, and immune activation markers, sCD14 and sCD163 were measured. Participant age was 36.2 years (SD = 9.5); 50% were male, 49% female and 1% transgender; 39% were HIV-positive, 50% were current or past smokers, and 39% endorsed cocaine use. A path analysis showed an indirect effect of HIV serostatus on the presence of carotid atherosclerotic plaques (Indirect Effect = 0.048, SE=0.024, p=.043), when controlling age, BMI, smoking, and cocaine use. This effect was mediated by inflammatory markers and changes in blood pressure. Findings point to putative underlying mechanisms leading to atherosclerosis among PLWH.

Introduction

Cardiovascular disease (CVD) is the leading cause of death in the US and a significant contributor to morbidity and mortality for people living with HIV (PLWH) (Beires et al., 2018). CVD is commonly associated with coronary artery narrowing due to atherosclerosis (Frostegård, 2013). Individuals with CVD have increased cytokines, immune activation markers, and vascular growth factors (e.g., interlukin-6 or IL-6, tumor necrosis factor or TNF-a, soluble CD14 and CD163, or sCD14 and sCD163, vascular endothelial growth factor or VEGF, and C-reactive protein or CRP, among others) (Jones et al., 2017). Studies suggest that chronic psychosocial stressors may lead to inflammation and may result in CVD (Miller and Blackwell, 2006). However, the association between inflammation and CVD may be bidirectional (Levi, van der Poll, & Büller, 2004).

Among PLWH, markers of immune activation contribute to CVD and other comorbidities (Beires, et al., 2018; Kelly et al., 2016). The mechanism by which HIV increases the risk of CVD in PLWH has not been elucidated due to the difficulty of separating traditional risk factors and effects of chronic infection. Substance use is prevalent among PLWH, and cocaine can particularly contribute to cardiovascular morbidity. Chronic cocaine use increases cardiovascular events by causing myocardial vasoconstriction and ischemia, facilitating plaque formation (Sandfort et al., 2017). Cocaine use is associated with increased cytokines, CRP, and growth factors, which may also exacerbate CVD risk among PLWH (Zhang et al., 2016).

Previous studies have examined the relationship between HIV, cocaine, inflammatory markers and CVD independently, but the mechanism by which HIV and cocaine use may increase CVD risk has not been evaluated. Because etiologies of inflammation and CVD are complex, the goal of this study was to evaluate two contributing factors (cocaine and HIV) to both inflammation and CVD risk. This study examined the effects of HIV, cocaine usage, and inflammatory markers, and their association with carotid atherosclerosis among young and middle-aged adults living with HIV. The goal was to use results to inform development of interventions to decrease CVD risk among PLWH.

Methods

Participants

Study approval was obtained from the University of Miami Miller School of Medicine Institutional Review Board. Written informed consent was obtained prior to participation. Recruitment was conducted from December 2014-June 2018. Participants were aged 18 to 50. To evaluate the effect of cocaine and HIV, the sample included HIV infected and uninfected control cocaine and non-cocaine users.

Measures

Demographics and Lifestyle.

Demographic and lifestyle information included gender, age, race, income, and cigarette smoking (past/current; yes/no).

Physical Assessment.

Body mass index (BMI) was calculated. Three measurements of systolic and diastolic blood pressure were taken at rest and averaged. Mean arterial pressure (MAP) was calculated (Townsend et al., 2015).

Cocaine Use

The Structured Clinical Interview for Diagnostic and Statistical Manual of Mental Disorders, 4^th Edition, non-patient version (SCID-IV-NP), a well-validated, semi-structured interview, was administered to assess cocaine use (Segal, Hersen, & Van Hasselt, 1994). Cocaine usage was limited to the past three months.

Carotid Atherosclerosis.

Ultrasonography of the carotid artery was performed using a high-resolution B-mode carotid ultrasound machine. Four hundred images were obtained at 1-centimeter segment of the distal common carotid artery near and far wall, 1-centimeter of bifurcation and 1-centimeter proximal internal carotid artery near and far wall. These images were analyzed using an automated edge tracking system, M’Ath (Intelligence in Medical Technologies, Inc., Paris, France), which scans the vicinity for wall boundaries using an intensity gradient algorithm to identify plaque.

Cytokines and Growth factors.

IL-1a, IL-6, TNFα and VEGF plasma levels were computed using a tailored MILLIPLEX Human Cytokine magnetic bead panel (EMD Millipore, Billerica, MA). Mean Fluorescence Intensity (MFI) data were assessed via MILLIPLEX Analyst Software V.3.5 (EMD Millipore). Mean cytokine concentrations were resolved using standard curve and represented in pg/ml.

Immune markers.

Plasma levels of sCD14 and sCD163 were determined by ELISA using the Human Quantikine Immunoassay (R&D Systems). To measure sCD14, samples were diluted 400-fold and results were expressed in pg/ml. For sCD163, samples were diluted 30-fold and results were expressed in pg/ml.

Data Analysis

Univariate statistics were used to describe the sample. Continuous variables were first standardized, and inflammatory markers were logarithmically transformed. A path analysis outlined in Figure 1 was carried out in Mplus version 8 to assess whether HIV and cocaine was related to number of plaques through inflammatory markers and MAP after controlling for age, BMI, and smoking. Indirect effects were tested from both HIV and cocaine to plaques. A bias-corrected bootstrap (n=5000) was utilized in calculating 95% confidence intervals for the indirect effects (MacKinnon, Lockwood, & Williams, 2004). Since the outcome of the path analysis was atherosclerotic plaques, a Poisson distribution and a maximum likelihood estimator was used. An α threshold of .05 assessed for statistical significance.

Figure 1.

Mediational model.

Results

Sample demographic characteristics, risk factors, and inflammatory markers

Participants (N=494) had a mean age of 36.15 years (SD = 9.48); 50% were male, 49% female, and 1% transgender; 39% were HIV-seropositive, 50% endorsed smoking tobacco, and 39% endorsed cocaine use. Two-thirds of the sample (62%) self-identified as African American. Table 1 provides more detail of key study variables, sociodemographic characteristics, risk factors, MAP, and inflammatory markers.

Table 1.

Sample Statistics

Variable	Mean (Standard Deviation)
Age	36.15 (9.48)
Body Mass Index	29.65 (8.32)
Mean Arterial Pressure	90.28 (10.92)
CRP	1955.84 (788.61)
TNF-α	18.15 (22.14)
IL-6	3.98 (11.19)
VEGF	79.44 (96.9)
SCD163	611.75 (371.73)
SCD14	1785.7 (520.77)
Characteristics of Participants with HIV (n = 191)	Mean (Median; SD) / n (%)
Year of Diagnosis	2002.5 (2002; 9.63)
Infection Vector
Sex	154 (80.6%)
Drug Use	5 (2.6%)
Transfusion	1 (0.5%)
Don’t Know	8 (4.2%)
Other	20 (10.5%)
Perinatal Transmission	20 (100%)
Opioid Use	2 (1%)
Injection Cocaine Use	2 (1%)
CD34	0.45 (0.21; 1.13)
CD133	1 (0.41; 3.52)
Viral Load	6.07 (1.45; 7.04)

Note. CRP = C-reactive protein, TNF-α = Tumor Necrosis Factor Alpha, IL-6 = Interlukin-6, VEGF = Vascular Endothelial Growth Factor, sCD163 = Soluble CD163, sCD14 = Soluble CD14. 3 observations missing for infection vector and year since diagnosis, 30 for CD34 and CD133, 189 for opioids and injectable cocaine, 106 for viral load.

Prevalence of carotid atherosclerosis

One quarter (n=122) of participants had carotid atherosclerosis, defined by a plaque in the carotid artery; the number of plaques ranged from 1 to 6. Of these, most had 1 plaque (n=58), n=31 had 2 plaques, n=21 had 3 plaques, n=10 had 4 plaques, and n=2 had 6 plaques.

Effect of HIV on atherosclerosis mediated by inflammatory markers and MAP

As illustrated in Figure 1 (AIC=−3123.101 and BIC=−2891.961), HIV was associated with increased inflammatory markers: CRP (β=0.036, SE=0.005, p=.004), VEGF (β=0.034, SE=0.009, p<.001), TNF-α (β=0.085, SE=0.008, p<.001), sCD163 (β=0.054, SE=0.009, p<.001), and sCD14 (β=0.062, SE=0.009, p<.001). HIV was also associated with cocaine use (β=0.158, SE=0.045, p<.001). BMI, a covariate, was associated with increases in both CRP (β=0.036, SE=0.005, p<.001) and sCD163 (β=0.012, SE=0.004, p=.001). Age, also a covariate, was associated with increases in both TNF-α (β=0.006, SE=0.003, p=.034) and IL-6 (β=0.008, SE=0.003, p<.023). CRP (β=1.239, SE=0.419, p=.003) and VEGF (β=1.395, SE=0.510, p=.006) predicted greater MAP per standard deviation increase in the inflammatory markers, and MAP was significantly related to plaque number (β=0.330, SE=0.079, p<.001).

When testing the association of HIV serostatus and cocaine use with number of plaques through inflammatory markers, the total indirect effect of HIV through inflammatory markers was significant (Total Indirect Effect = 0.048, SE=0.024, p=.043). Examining the specific pathways, HIV predicted MAP and number of plaques through both CRP (Indirect Effect = 0.011, SE=0.006, p=.058) and VEGF (Indirect Effect = 0.015, SE=0.008, p=.057).

Discussion

This study evaluated the association between HIV, cocaine use, and carotid atherosclerosis, as mediated by inflammation, immune activation and blood pressure, among young and middle-aged adults utilizing a mechanistic modeling approach. Results indicated that HIV was associated with atherosclerosis through increased MAP, mediated by increased inflammatory markers.

Earlier studies have described high rates of CVD in aging PLWH and have evaluated traditional risk factors and inflammation (Beires, et al., 2018; Kelly, et al., 2016). However, previous studies have not addressed young and middle age PLWH and the co-occurring impact of cocaine use. The current study is unique as it included a younger population of PLWH, a direct marker of CVD risk (carotid plaques), and a path modeling approach.

Despite multiple risk factors for CVD in this younger population, the observed blood pressure was within normal limits. However, findings suggest that even modest increases in blood pressure were associated with plaques. The pathway for plaque formation may be mediated by increases in inflammatory markers and blood pressure, and highlights the importance of controlling blood pressure (Whelton et al., 2018). Our results support the role of reducing systemic inflammation among PLWH to reduce CVD, which is currently being evaluated in ongoing clinical trials (Hoffmann et al., 2019).

The occurrence of myocardial ischemia among cocaine users has been previously attributed to vasospastic events, with atherosclerosis occurring after chronic use (Sandfort, et al., 2017). This study did not find an association between cocaine and systemic inflammation, elevation in the blood pressure, or plaques. Cocaine was likely not associated as cocaine does not promote plaque formation, or as cocaine use had been too brief to cause sustained damage to the arterial wall.

Study limitations included the cross-sectional design, the lack of CVD risk evaluation by established criteria, individuals with HIV infection were not on ART at the assessment, and as history of prior ART was not collected, this may have contributed to inflammation and atherosclerosis. Lastly, cocaine use was neither associated with inflammatory markers nor cardiovascular outcomes. However, the lack of association may have been due to a short period of cocaine exposure. Future studies should evaluate longer periods of exposure.

In summary, this study found HIV to be associated with carotid atherosclerotic plaques through increases in inflammatory markers and blood pressure. Interventions to monitor blood pressure to reduce CVD risk should be considered, as well as targeting traditional factors, such as smoking, among PLWH.