Abstract
Objective
Some intestinal microbiota-generated metabolites of phosphatidylcholine are recognized to be pro-atherogenic. As the HIV population is vulnerable to cardiovascular disease (CVD) and can develop intestinal dysbiosis associated with systemic inflammation, we investigated the novel relationship between microbiota-derived metabolites of phosphatidylcholine and coronary atherosclerosis in HIV.
Design/Methods
155 HIV-infected and 67 non-HIV-infected subjects without known history of CVD were previously recruited to assess coronary plaque by CT angiography. In the current study, we evaluate whether serum choline, trimethylamine (TMA), or trimethylamine N-oxide(TMAO) levels are associated with plaque features.
Results
Young, asymptomatic HIV-infected subjects (age 47±7 years) demonstrated significantly higher prevalence of plaque (53 vs. 35%, P=0.01) and number of total plaque segments (1.8±2.5 vs. 1.2±2.2, P=0.03) when compared to well-matched non-infected subjects with similar co-morbidities. TMA was significantly associated with calcium score (r=0.22, P=0.006), number of total (r=0.20, P=0.02) and calcified (r=0.18, P=0.03) plaque segments, and calcium plaque volume (r=0.19, P=0.02) and mass (r=0.22, P=0.009) in the HIV cohort only. In multivariate modeling among HIV-infected subjects, TMA remained significantly associated with calcium score (P=0.008), number of total (P=0.005) and calcified (P=0.02) plaque segments, and calcium plaque volume (P=0.01) and mass (P=0.007), independent of Framingham Risk Score. In contrast, there was no association of TMAO to coronary plaque features in either cohort.
Conclusion
A link between TMA and atherosclerosis has not previously been established. The current study suggests that TMA may be a non-traditional risk factor related to the number of plaque segments and severity of calcified plaque burden in HIV.
Keywords: HIV, Trimethylamine, Trimethylamine-N-Oxide, Coronary Plaque, Gut Microbiota, Inflammation
Introduction
HIV-infected patients demonstrate increased risk of cardiovascular disease (CVD) beyond that predicted by traditional risk factors, such as age, gender, and tobacco use, when compared to control subjects [1–3]. Prior studies have suggested that unique, non-traditional mechanisms involving inflammation may potentiate the risk for atherosclerotic plaque in HIV [4–6], but the exact mechanism is yet to be determined. Emerging evidence suggests that alterations in intestinal microbial composition may contribute to CVD [7–9]. Indeed, HIV itself has been associated with changes in the gut microbiome [10] and disruption in intestinal wall integrity [11], which may in part contribute to a proatherogenic disease state in this population. In addition, recent studies demonstrate that altered gut microbiota is associated with increased inflammatory and immune activation markers among HIV-infected patients with good virologic control on ART (antiretrovial therapy) [12–14], thus further suggesting a relationship between gut microbiota and CVD in these patients.
Recent interest has focused on the metabolism of phosphatidylcholine in relationship to the gut as a potentially important pathway contributing to CVD. Dietary phosphatidylcholine is a major source of choline, which is further metabolized by intestinal flora to trimethylamine (TMA). In the liver, TMA is oxidized to trimethylamine-N-oxide (TMAO). Alternatively, dietary L-carnitine [15] and betaine [16] can also be metabolized to TMAO, but contribute less to TMAO production. Wang et al. proposed that metabolites of dietary phosphatidylcholine, while essential to certain biologic roles, also carry atherogenic potential [17]. In this regard, circulating levels of TMAO are associated with CVD among non HIV-infected patients [15, 17, 18]. However, no prior studies have investigated whether enhanced metabolism of phosphatidylcholine contributes to increased CVD in the HIV population.
The aim of this study was to compare serum phosphatidylcholine metabolites among HIV-infected subjects and healthy controls in relation to subclinical atherosclerotic disease. While prior studies have determined the association of TMAO to cardiovascular events in the non-HIV-infected population, this study investigated for the first time the relationship of choline-related metabolites to coronary plaque burden in HIV, assessed by cardiac CT angiography. This study therefore investigates a new potential mechanistic factor linking gut metabolism to CVD in HIV-infected patients. We hypothesized that HIV-infected subjects would demonstrate unfavorable coronary plaque characteristics in association with increased serum levels of TMAO.
Methods
Study Participants
155 HIV-infected subjects and 67 non-HIV-infected controls were previously recruited. HIV-infected subjects were recruited from Boston area HIV clinics, community health care centers, and newspaper advertisements. Healthy controls were recruited from the same communities to ensure similar demographic characteristics. Data on coronary plaque characteristics have been previously reported in this cohort [1, 19, 20], but data on choline-related metabolites have not previously been assessed.
Inclusion and exclusion factors were identical for both groups with the exception of HIV serostatus and are detailed elsewhere [1, 20]. In brief, subjects aged 18–60 years were recruited based on no known cardiac disease, or symptoms of heart disease (including current or prior angina, arrhythmias, valvular heart disease, pericarditis, congestive heart failure, or any prior treatment for coronary artery disease (CAD) or any heart disease). Subjects with known renal disease, creatinine levels >1.5 mg/dL or creatinine clearance <70 mL/min were excluded to minimize risk of contrast nephropathy. HIV-infected subjects were receiving stable ART for >3 months.
All participants provided informed consent to participate. This study was approved by the institutional review board of Massachusetts General Hospital.
Study Procedures and Assessment of Cardiovascular Risk Factors
Detailed interviews were performed to obtain sociodemographic information, medical and family history, cardiovascular risk factors, and tobacco, recreational intravenous drug and medication use. For HIV-infected subjects, duration of diagnosis and detailed history of past and present ART use were also obtained.
Cardiac CT Angiography
Cardiac CT angiography (CTA) imaging was performed using a 64-slice CT scanner (Sensation 64; Siemens Medical Solutions) according to a standardized protocol [1]. Assessment of coronary plaque type, calcified or non-calcified, and other plaque characteristics are detailed elsewhere [1]. A consensus reading of coronary atherosclerotic plaque segments was established between a cardiologist and radiologist with significant experience in cardiac CT interpretation who were blinded to subjects’ medical history, including HIV serostatus. Agatston calcium score was quantified using non-contrast CT imaging [21].
Metabolic, Biochemical, Inflammatory and Body Composition Parameters
Creatinine and glucose and lipid parameters were determined using standard techniques. Plasma high-sensitivity interleukin 6 (hsIL-6) (R&D Systems) and high sensitivity C-reactive protein (hsCRP) (Labcorp) were quantified by enzyme-linked immunosorbent assay (ELISA). The endpoint limulus amebocyte lysate assay (Associates of Cape Cod) was used to evaluate LPS levels [20]. All subjects fasted prior to blood draws. A cross-sectional CT scan at the level of the L4 pedicle was performed to assess abdominal visceral and subcutaneous adipose tissue (VAT and SAT, respectively) area [22].
Dietary Assessment of Choline and Betaine Levels
Dietary intake was collected through a 4-day food record and analyzed using Nutrition Data System for Research software (Nutrition Coordinating Center (NCC), University of Minnesota, Minneapolis, MN). Dietary choline and betaine were assessed.
Measurements of Choline Related Serum Metabolites
Serum samples were obtained after a 12 hour fast and aliquoted into 2 mL Sarstedt microtubes and stored at −80°C. A mass spectrometry based method was specifically designed to assay choline, betaine, L-carnitine, TMA and TMAO in serum samples. Serum aliquots were spiked with internal standards, mixed with acetonitrile and injected on a UPLC® chromatographic system (Waters Corp, Saint-Quentin en Yvelines, France) coupled to a Q-Exactive™ hybrid quadrupole-Orbitrap mass spectrometer (Thermo Fisher Scientific, Illkirch, France). The system was fitted with an Acquity BEH HILIC column and a corresponding guard column (Waters Corp, Saint-Quentin en Yvelines, France). Data were acquired in full scan positive ion mode with a resolution of 70,000 FWHM at a scan range of m/z 50–400. Target analyte signals were extracted with an accuracy of 5 ppm. Additional studies were performed to assess correlations between plasma and serum and metabolite assessment in different storage conditions (see Supplemental Methods).
Statistical Analysis
Normality of distribution was determined using the Shapiro-Wilk test. Data are presented as mean ± standard deviation or median (interquartile range [IQR]), depending on normality of the distribution; all categorical variables are reported as proportions. Comparisons were made between HIV-infected and non-HIV-infected subjects using the Student’s t test for normally distributed continuous variables, the Wilcoxon rank sum test for non-normally distributed data, and the χ2 test for categorical variables. Non-normally distributed variables were log-transformed to establish a normal distribution. Linear regression was performed using Pearson’s correlation coefficient within the HIV and non-HIV-infected cohorts, separately. Multivariate regression analysis was performed to assess the effects of individual choline-related serum metabolites on coronary plaque features among HIV-infected subjects controlling for Framingham Risk Score (FRS), a composite score of traditional risk factors known to affect plaque. The focus of these analyses was TMA, as this metabolite was most strongly associated with coronary plaque features in univariate analyses. Sensitivity analyses including LPS in the models were also performed, as LPS was the systemic inflammatory marker most significantly associated with TMA. In addition, sensitivity analyses controlling for HDL, hepatitis C Virus (HCV) status, and PI, NNRTI, atanazavir (ATV) and efavirenz (EFV) use were performed. Statistical significance was defined as P<0.05. All statistical analysis was performed using SAS JMP (version 9.0).
Results
Demographics and Traditional CVD Risk Factors
HIV-infected and non-infected subjects were well matched with respect to age, gender, and race. Similar proportions of HIV subjects and non-HIV-infected controls were identified to have dyslipidemia, hypertension, and diabetes mellitus. The HIV cohort presented with a long duration of HIV for 14 ± 6 years, and 99% of this population had exposure to ART for 8 ± 5 years. Current use of PI, NRTI, and NNRTI treatments in the HIV-infected subjects was 55%, 95%, and 37%, respectively. The majority of the HIV subjects had evidence of good immunological control with CD4+ cell count 552 ± 290 cell/μL. Undetectable viral load was demonstrated among 86% of the cohort. There was a higher prevalence of HCV in the HIV subjects. With respect to metabolic parameters, HIV-infected subjects demonstrated significantly increased fasting triglycerides compared to the non-infected subjects. There was a trend towards lower SAT and in the HIV cohort. The prevalence of smoking was similar between the groups. Other lipid parameters, including total cholesterol, LDL, and HDL, were not significantly different between populations. Furthermore, FRS did not differ between the HIV and control populations. The HIV cohort demonstrated elevated liver transaminases when compared to healthy subjects. In addition, LPS was increased in the HIV-infected subjects compared to the non-HIV infected subjects (Table 1).
Table 1.
Characteristics of Control and HIV-Infected Subjects
| Control Subjects (n=67) | HIV-Infected Subjects (n=155) | P-value | |
|---|---|---|---|
| Demographics | |||
| Age, y | 46 ± 7 | 47 ± 7 | 0.13 |
| Gender, % | |||
| Male | 58 | 61 | 0.67 |
| Race, % | |||
| White | 49 | 53 | 0.37 |
| Framingham Risk Score | 9 (5–11) | 10 (7–12) | 0.09 |
| Family history of premature CHD by NCEP, % | 24 | 21 | 0.71 |
| Dyslipidemia by NCEP, % | 24 | 34 | 0.13 |
| Hypertension, % | 15 | 24 | 0.15 |
| Diabetes mellitus, % | 7 | 10 | 0.50 |
| Current smoker, % | 42 | 44 | 0.74 |
| HCV, % | 9 | 27 | 0.002 |
| History of IVDU, % | 13 | 23 | 0.12 |
| HIV Disease Related Parameters | |||
| Duration since HIV diagnosis, y | N/A | 14 ± 6 | N/A |
| Ever on antiretroviral therapy, % | N/A | 99 | N/A |
| Duration of antiretroviral therapy, y | N/A | 8 ± 5 | N/A |
| Current PI treatment, % | N/A | 55 | N/A |
| Current NRTI treatment, % | N/A | 95 | N/A |
| Current NNRTI treatment, % | N/A | 37 | N/A |
| CD4+ T-lymphocytes, cells/μL | N/A | 552 ± 290 | N/A |
| Log HIV RNA viral load, copies/mL | N/A | 1.8 ± 0.5 | N/A |
| Undetectable HIV RNA < 50 copies/mL, % | N/A | 86 | N/A |
| Anthropometric Parameters | |||
| Body mass index, kg/m2 | 27 ± 5.0 | 27 ± 5.0 | 0.39 |
| Waist Circumference (iliac crest), cm | 93.9 (84.4–105.1) | 96.2 (85.1–108.4) | 0.60 |
| VAT area, cm2 | 100 (52–172) | 109 (61–210) | 0.28 |
| SAT area, cm2 | 238 (146–343) | 199 (126–288) | 0.07 |
| Metabolic Indices | |||
| Systolic blood pressure, mm Hg | 117 ± 15 | 119 ± 14 | 0.22 |
| Diastolic blood pressure, mm Hg | 76 ± 9 | 76 ± 9 | 0.70 |
| Fasting glucose, mg/dL | 86 (80–93) | 88 (80–95) | 0.47 |
| Hemoglobin A1c, % | 5.5 (5.4–5.8) | 5.5 (5.2–5.8) | 0.08 |
| Total cholesterol, mg/dL | 176 (156–200) | 178 (155–206) | 0.71 |
| HDL cholesterol, mg/dL | 49 (42–66) | 50 (41–62) | 0.71 |
| LDL cholesterol, mg/dL | 105 ± 30 | 102 ± 35 | 0.60 |
| Triglycerides, mg/dL | 83 (62–119) | 97 (77–172) | 0.001 |
| Creatinine, mg/dL | 0.94 (0.77–1.13) | 0.93 (0.81–1.10) | 0.81 |
| ALT, U/dL | 24 ± 15 | 35 ± 28 | 0.0001 |
| AST, U/dL | 28 ± 15 | 39 ± 30 | 0.0008 |
| Inflammatory Indices | |||
| CRP, mg/dL | 0.13 (0.05–0.36) | 0.14 (0.05–0.39) | 0.76 |
| hsIL-6, pg/mL | 0.96 (0.61–1.83) | 1.06 (0.74–1.86) | 0.38 |
| LPS, ng/mL | 0.07 (0.02–0.09) | 0.09 (0.04–0.12) | 0.003 |
Data reported as mean ± standard deviation, percentage, or median (interquartile range). Abbreviations: CHD, coronary heart disease; NCEP, National Cholesterol Education Program; HCV, hepatitis C virus; IVDU, intravenous drug use; PI, protease inhibitor; N/A, not applicable; NRTI, nucleoside/nucleotide reverse transcriptase inhibitors; NNRTI, non-nucleoside reverse transcriptase inhibitors; VAT, visceral adipose tissue; SAT, subcutaneous adipose tissue; HDL, high-density lipoprotein; LDL, low-density lipoprotein; AST, aspartate aminotransferase; ALT, alanine aminotransferase; CRP, C-reactive protein; hsIL-6, high sensitivity interleukin-6; LPS, lipopolysaccharide
Dietary Intake and Serum Levels of Choline Related Metabolites
Dietary intake of choline and betaine and measured serum levels of choline, betaine, L-carnitine, TMA, and TMAO did not differ between HIV-infected and non-infected subjects (Table 2). Dietary choline was related to serum TMA in both HIV-infected (r=0.19, P=0.05) and control (r=0.33, P=0.05) groups. There were no other significant associations between dietary intake and serum measures of choline-related metabolites. Serum TMA did not differ by current PI use status vs. not (0.4[0.3, 1.4] vs. 0.4[0.3, 1.5], P=0.30), but was higher in those on NNRTI-containing regimens vs. not (0.4[0.3, 1.6] vs. 0.3[0.3, 0.4], P=0.001). Specific correlations to ART demonstrated that serum TMA levels were significantly higher in HIV-infected subjects on efavirenz-containing regimens vs. non-efavirenz containing regimens (0.4[0.3, 1.6] vs. 0.3[0.3, 0.4], P=0.03). TMA did not differ by atanazavir-containing regimens vs. non-atanazavir containing regimens (0.3[0.3, 0.4] vs. 0.3[0.3, 1.5], P=0.38).
Table 2.
Dietary and Serum Choline-Related Metabolites among Controls and HIV-Infected Subjects
| Control Subjects (n=67) | HIV-Infected Subjects (n=155) | P-value | |
|---|---|---|---|
| Dietary Intake | |||
| Betaine, mg | 155.9 (90.3–208.9) | 128.9 (94.2–182.3) | 0.43 |
| Choline, mg | 349.2 (251.5–464.8) | 330.7 (218.9–429.4) | 0.60 |
| Serum Metabolites | |||
| Betaine, μmol/L | 6.1 (4.7–7.7) | 5.8 (5.1–7.0) | 0.32 |
| L-Carnitine, μmol/L | 5.9 (4.9–6.7) | 5.8 (4.5–6.9) | 0.58 |
| Choline, μmol/L | 2.0 (1.8–2.2) | 1.9 (1.7–2.3) | 0.88 |
| TMA, μmol/L | 0.3 (0.3–0.4) | 0.3 (0.3–1.5) | 0.50 |
| TMAO, μmol/L | 0.7 (0.5–1.0) | 0.7 (0.5–1.1) | 0.79 |
Data reported as median (interquartile range). Abbreviations: TMA, trimethylamine; TMAO, trimethylamine-N-oxide.
Food record data were available for 36 control subjects and 100 HIV-infected subjects.
Coronary Plaque Characteristics
HIV-infected subjects demonstrated a significantly higher prevalence of plaque (53 vs. 35%, P=0.01) and number of total plaque segments (1.8± 2.5 vs. 1.2± 2.2, P=0.03) when compared to control subjects. Both cohorts had a similar number of calcified plaque segments in addition to other parameters related to calcified plaque, including calcium score and calcium plaque volume and mass, but the HIV cohort was observed to have more non-calcified segments compared to healthy controls (Table 3).
Table 3.
Coronary Plaque Characteristics by CT Angiography of Control and HIV-Infected Subjects
| Control Subjects (n=67) | HIV-Infected Subjects (n=155) | P-value | |
|---|---|---|---|
| Coronary Plaque Characteristics | |||
| Log Calcium Score | 1.5 ± 2.7 | 1.6 ± 2.5 | 0.81 |
| Presence of Plaques, % | 35 | 53 | 0.01 |
| Total Plaque Segments, # | 1.2 ± 2.2 | 1.8 ± 2.5 | 0.03 |
| Non-Calcified Segments, # | 0.5 ± 1.2 | 1.0 ± 1.6 | 0.003 |
| Calcified Segments, # | 0.3 ± 0.8 | 0.2 ± 0.6 | 0.41 |
| Total Calcium Volume of Plaque, mm3 | 19.9 ± 49.2 | 21.3 ± 59.7 | 0.35 |
| Total Calcium Mass of Plaque, mg | 4.2 ± 10.5 | 4.9 ± 14.9 | 0.35 |
Data reported as mean ± standard deviation (SD) or percentage. For segments and plaque characteristics, results are reported as mean ± SD to provide more information because of the low number of lesions. P value comparison by Wilcoxon Rank Sums where appropriate.
Relationship of Serum Metabolites and Coronary Plaque Features among HIV-infected and non-HIV-infected Subjects
Among the HIV-infected subjects, serum TMA was significantly and positively associated with calcium score (r=0.22, P=0.006), number of total plaque segments (r=0.20, P=0.02), number of calcified plaque segments (r=0.18, P=0.03), calcium plaque volume (r=0.19, P=0.02) and calcium plaque mass (r=0.22, P=0.009) (Table 4a). Serum betaine was not consistently related to plaque parameters, and no relationships were observed between serum choline, L-carnitine, or TMAO and plaque characteristics (Table 4a). In addition, dietary intake of choline and betaine did not relate to plaque parameters (data not shown). Serum TMA levels inversely correlated with HDL (r= −0.17, P=0.03) among HIV-infected subjects.
Table 4.
| Table 4a. Univariate Associations with Choline Related Metabolites among HIV-Infected Subjects | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Parameter | Log Serum Betaine | Log Serum L-Carnitine | Log Serum Choline | Log Serum TMA | Log Serum TMAO | |||||
| r | P-value | r | P-value | r | P-value | r | P-value | r | P-value | |
| Coronary Plaque Characteristics | ||||||||||
| Log Calcium Score | −0.02 | 0.83 | −0.006 | 0.94 | 0.10 | 0.22 | 0.22 | 0.006 | 0.06 | 0.47 |
| Total Plaque Segments, # | −0.001 | 0.99 | −0.07 | 0.43 | 0.03 | 0.71 | 0.20 | 0.02 | 0.007 | 0.93 |
| Noncalcified Plaque Segments, # | 0.03 | 0.76 | −0.13 | 0.13 | 0.02 | 0.82 | 0.11 | 0.18 | 0.007 | 0.93 |
| Calcified Plaque Segments, # | −0.19 | 0.03 | 0.10 | 0.23 | −0.03 | 0.70 | 0.18 | 0.03 | 0.02 | 0.85 |
| Total Calcium Volume of Plaque, mm3 | −0.07 | 0.42 | 0.07 | 0.39 | 0.09 | 0.26 | 0.19 | 0.02 | 0.04 | 0.67 |
| Total Calcium Mass of Plaque, mg | −0.07 | 0.42 | 0.08 | 0.38 | 0.14 | 0.10 | 0.22 | 0.009 | 0.04 | 0.66 |
| Inflammatory Indices | ||||||||||
| Log CRP, mg/dL | 0.10 | 0.22 | 0.12 | 0.15 | −0.14 | 0.08 | 0.11 | 0.18 | 0.06 | 0.47 |
| Log hsIL-6, pg/mL | 0.05 | 0.58 | −0.14 | 0.11 | 0.006 | 0.94 | −0.10 | 0.25 | 0.15 | 0.08 |
| Log LPS, ng/mL | 0.05 | 0.59 | 0.02 | 0.81 | 0.03 | 0.75 | 0.19 | 0.03 | 0.14 | 0.12 |
| Table 4b. Univariate Associations with Choline Related Metabolites among Control Subjects | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Parameter | Log Serum Betaine | Log Serum L-Carnitine | Log Serum Choline | Log Serum TMA | Log Serum TMAO | |||||
| r | P-value | r | P-value | r | P-value | r | P-value | r | P-value | |
| Coronary Plaque Characteristics | ||||||||||
| Log Calcium Score | 0.02 | 0.86 | 0.08 | 0.52 | −0.08 | 0.54 | 0.09 | 0.47 | 0.16 | 0.21 |
| Total Plaque Segments, # | 0.02 | 0.87 | 0.14 | 0.27 | −0.05 | 0.72 | 0.14 | 0.29 | 0.02 | 0.90 |
| Noncalcified Plaque Segments, # | −0.03 | 0.80 | 0.18 | 0.17 | −0.07 | 0.61 | 0.19 | 0.15 | −0.14 | 0.27 |
| Calcified Plaque Segments, # | 0.06 | 0.65 | 0.07 | 0.58 | 0.06 | 0.63 | −0.17 | 0.17 | 0.18 | 0.15 |
| Total Calcium Volume of Plaque, mm3 | 0.22 | 0.08 | 0.24 | 0.06 | −0.06 | 0.64 | 0.0003 | 1.00 | 0.12 | 0.34 |
| Total Calcium Mass of Plaque, mg | 0.21 | 0.09 | 0.24 | 0.06 | −0.07 | 0.61 | −0.02 | 0.89 | 0.13 | 0.29 |
| Inflammatory Indices | ||||||||||
| Log CRP, mg/dL | −0.0002 | 1.00 | 0.14 | 0.25 | −0.10 | 0.44 | 0.16 | 0.20 | −0.03 | 0.80 |
| Log hsIL-6, pg/mL | −0.21 | 0.13 | −0.10 | 0.48 | −0.05 | 0.71 | 0.04 | 0.79 | 0.09 | 0.54 |
| Log LPS, ng/mL | 0.15 | 0.28 | 0.15 | 0.28 | 0.07 | 0.64 | −0.01 | 0.92 | −0.08 | 0.56 |
r represents Pearson’s correlation coefficient.
Abbreviations: TMA, trimethylamine; TMAO, trimethylamine-N-oxide; CRP, C-reactive protein; hsIL-6, high sensitivity interleukin-6; LPS, lipopolysaccharide.
Among the non-HIV-infected control subjects, no significant relationships with serum TMA or other metabolites were seen (Table 4b).
Relationship of Serum TMA and Systemic Inflammatory Markers among HIV-infected and non-HIV-infected Subjects
Among the HIV-infected subjects, serum TMA was positively correlated with LPS (r=0.19, P=0.03), while there was no significant association with hsIL-6 and hsCRP (Table 4a). Serum TMA did not differ based on HCV status in the HIV cohort (TMA (0.3[0.3, 1.5] vs. 0.3[0.3, 0.4], P=0.38, no HCV vs. presence of HCV).
Among the non-HIV-infected control subjects, neither TMA nor other choline related metabolites were associated to systemic inflammatory markers (Table 4b).
Multivariate Regression Modeling Among HIV-infected Subjects to Assess Effects of Serum TMA on Coronary Plaque Features
As the relationship to serum TMA and coronary plaque features was significant in the HIV cohort, we performed multivariate modeling assessing the relationship of TMA to coronary plaque features simultaneously controlling for the effects of other common risk factors for calcified coronary plaque, using the composite Framingham Risk Score. In separate models for specific coronary plaque features, serum TMA remained significantly associated with calcium score (β=25.7192, P=0.008), number of total plaque segments (β=0.9643, P=0.005), number of calcified plaque segments (β=0.1895, P=0.02), calcium plaque volume (β=21.0937, P=0.01), and calcium plaque mass (β= 5.7989, P=0.007), independent of FRS (Table 5a). In a model including HDL, rather than the aggregate total Framingham Risk Score, TMA remained significantly associated to these coronary plaque parameters independent of HDL (Supplemental Table 1). Similarly, TMA remained related to these plaque indices in multivariate regression models controlling for PI, NNRTI, ATV and EFV use and HCV status (Supplemental Table 1). Moreover, TMA remained a significant and independent predictor of calcium score, number of total plaque segments, and calcium plaque mass and tended to be independently associated with number of calcified plaque segments and calcium plaque volume after further controlling for both FRS and LPS (Table 5b).
Table 5.
Models to assess effects of Serum TMA on Coronary Plaque Parameters in HIV-Infected Subjects
| Table 5a. | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Calcium Score (n=146) (R2=0.08; P=0.002) |
Total Plaque Segments (#)(n=139) (R2=0.16; P=<0.0001) |
Calcified Plaque Segments (#)(n=139) (R2=0.08; P=0.004) |
Calcium Volume of Plaque (n=138) (R2=0.08; P=0.003) |
Calcium Mass of Plaque (n=138) (R2=0.08; P=0.004) |
||||||
| β Estimate | P Value | β Estimate | P Value | β Estimate | P Value | β Estimate | P Value | β Estimate | P Value | |
| Variables | ||||||||||
| Serum TMA | 25.7192 | 0.008 | 0.9643 | 0.005 | 0.1895 | 0.02 | 21.0937 | 0.01 | 5.7989 | 0.007 |
| FRS | 2.8206 | 0.01 | 0.1718 | <0.0001 | 0.0266 | 0.009 | 2.6245 | 0.01 | 0.5439 | 0.04 |
| Table 5b | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Calcium Score (n=123) (R2=0.08; P=0.02) |
Total Plaque Segments (#)(n=118) (R2=0.15; P=0.0003) |
Calcified Plaque Segments (#)(n=119) (R2=0.08; P=0.02) |
Calcium Volume of Plaque (n=117) (R2=0.09; P=0.02) |
Calcium Mass of Plaque (n=117) (R2=0.08; P=0.03) |
||||||
| β Estimate | P Value | β Estimate | P Value | β Estimate | P Value | β Estimate | P Value | β Estimate | P Value | |
| Variables | ||||||||||
| Serum TMA | 22.9497 | 0.04 | 0.8232 | 0.03 | 0.1720 | 0.07 | 18.2025 | 0.06 | 5.2065 | 0.03 |
| FRS | 3.2799 | 0.02 | 0.1605 | 0.0008 | 0.0293 | 0.02 | 3.1836 | 0.01 | 0.6456 | 0.04 |
| LPS | 36.0131 | 0.76 | 4.9143 | 0.21 | 0.6241 | 0.53 | 8.0591 | 0.94 | 6.7358 | 0.80 |
R2 represents the coefficient of determination and the proportion of variance explained by the model. P value represents significance by the whole model ANOVA test.
Abbreviations: FRS, Framingham Risk Score; TMA, trimethylamine; LPS, lipopolysaccharide.
Discussion
This is the first study to evaluate the association of phosphatidylcholine metabolites to detailed plaque phenotype and burden of calcified plaque both in healthy controls and the HIV population. Our data demonstrate that serum TMA, but not serum TMAO, is associated with the presence of coronary plaque, and more specifically calcified plaque in HIV-infected patients. Furthermore, other parameters of calcified plaque, including calcium score and calcified plaque mass and volume were similarly related to serum TMA. The significant association of TMA to these detailed calcified plaque indices was demonstrated to be independent of traditional CVD risk factors, as represented by the Framingham Risk Score and HDL. Therefore, our data suggest that TMA may be a non-traditional CVD risk factor with relevance to HIV-infected patients. Prior studies have related CVD event rates and degree of coronary artery stenosis to circulating TMAO in non-HIV-infected subjects, but have not yet evaluated the relationship of plaque phenotype to this metabolite or to TMA and other metabolites in HIV.
Adequate amounts of dietary choline, commonly found in eggs, beef, and animal liver, are essential to biological roles in cholinergic neurotransmission, hepatic VLDL secretion, and synthesis of integral phospholipids [23]. Moreover, through its intermediate oxidation to betaine, choline is an effective methyl donor in the conversion of homocysteine to methionine. Due to its ability to decrease plasma homocysteine, dietary supplementation of choline may be cardioprotective [24]. However, contrary evidence suggests that choline may, in fact, serve as a substrate for atherosclerotic disease through metabolism by intestinal microbiota to TMA and subsequent production of TMAO [17, 18]. It is possible that these contrasting data in the choline literature may be explained by a U-shaped curve, such that either deficiency or excess of choline could be detrimental to biologic function.
In our study, both HIV-infected and non-HIV infected cohorts had similar dietary intake of choline and betaine, as well as measured serum levels of all substrates betaine, choline, L-carnitine, TMA, and TMAO. Choline can be synthesized endogenously or obtained through the diet. The dietary intake for choline in both cohorts was well below the proposed daily recommendation for adequate intake, measured to be 425mg for women and 550mg for men [25]. Food records may under report the total amount of nutrients. Thus, serum metabolites under fasting conditions were obtained for more accurate comparisons.
Potential mechanisms for the presumed role of TMAO in atherosclerosis include inhibition of reverse cholesterol transport and altered cholesterol metabolism [15]. Moreover, TMAO may enhance formation of foam cells from macrophages [17]. The gut microbiome appears to be an essential component in the production of TMAO. After administration of antibiotics, TMAO production is attenuated, but production rebounds upon recolonization of the gut flora [17]. Indeed, Tang et al. showed that in human studies, TMAO is correlated with an increased risk of major cardiovascular events, independent of traditional cardiovascular risk factors [26]. In contrast, circulating levels of choline and betaine have been shown to be less predictive of major adverse cardiac events when compared to TMAO [16]. Furthermore, the relationship of dietary choline and betaine to CVD has been equivocal in prior studies [23, 27, 28]. Some prior studies that have cited a link between TMAO and CVD have incorporated a choline challenge. Tang et al. asked participants to digest hard boiled eggs and radiolabelled phosphatidylcholine [26]. In contrast, our subjects were observed after routine diet and under fasting conditions. Prior studies have not assessed serum measures of TMA and TMAO, but serum and plasma have given similar results for these analytes in our lab.
Limited details are available as to the function of TMA and its direct impact on atherosclerosis, apart from its central role as the serum intestinal microbiota-generated precursor to TMAO, and prior studies have not investigated the relationship between TMA and atherosclerotic disease. In this study, we investigated TMAO and additionally its precursor TMA in relationship to detailed indices of plaque morphology. There is site-specific enzymatic conversion of TMA to TMAO in the liver predominantly by flavin-containing monooxygenase 3 (FMO3). Our data demonstrate a novel relationship between calcified plaque features and TMA, but not TMAO, in this first study of HIV-infected patients. While FMO3 expression may reportedly differ based on gender and dietary bile acid [18], other mechanisms may account for these findings.
First, our HIV cohort demonstrated significantly higher transaminases, suggestive of underlying hepatic inflammation. Since FMO3 is unique to the liver, chronic inflammation may promote downregulation of this enzyme and in turn affect the conversion of TMA to TMAO. Trimethylaminuria, a clinical model of disease in which there is a FMO3 loss of function mutation and concomitant elevated TMA levels, exists in varying degrees of severity, but has not yet been associated with increased CVD. Bennett et al. created a murine model of FMO3 deficiency, which resulted in a 90% reduction of FMO3 mRNA in the liver, a 47% reduction in plasma TMAO, and more than doubling of plasma TMA levels when compared to control mice [18]. In two murine models of inflammation, induced by introduction of C. rodentium and LPS, mRNA levels of FMO3 were significantly decreased [29]. In addition, more than half of HIV-infected subjects in this study received a PI-based regimen, which could inhibit the activity of liver drug metabolizing enzymes. In addition, atanazavir can affect bile metabolism. In our HIV cohort, we did not see an association with serum TMA and PI use, but we did see an association of serum TMA to NNRTI use. We performed additional analyses to investigate the contribution of efavirenz, a NNRTI known to affect liver specific enzymes. There was a significant correlation between efavirenz use and TMA; however, the relationship of TMA to coronary plaque characteristics persisted regardless of efavirenz use. The strong independent relationship between TMA and plaque features in the current human study, suggest that TMA may itself have atherogenic potential. Lastly, TMA may also be a marker for other processes, including microbial translocation, which may contribute to CVD.
Specific gut microbiota compositions have been associated with either circulating levels of TMA or TMAO, respectively [15]. During HIV infection, there is a relative reduction in a CD4+ subset of Th17 cells that regulate the intestinal flora [30]. Moreover, taxa have been demonstrated to discriminate between HIV-infected and non-infected subjects [10]. Dysbiosis is present even in those HIV-infected subjects well controlled on suppressive ART. Relative enrichment of Enterobacteriaceae and depletion of Bacteriodes have been profiled in the gut microbiome in HIV-infected subjects when compared to non-HIV-infected subjects [12, 14]. In addition, specific microbiota have been identified that relate to increases in markers of microbial translocation and systemic inflammation in HIV-infected subjects on ART-suppressive regimens [12–14]. Alterations in gut-associated T-cell regulation in HIV may impair the mucosal barrier and enhance bacterial translocation in the gut with subsequent elevation in systemic bacterial endotoxins, including LPS. Indeed, increased levels of LPS have been associated with atherosclerotic disease in HIV-infected patients [31, 32]. Prior studies have demonstrated that alterations in gut flora may be linked to cardiometabolic disease, inflammation and atherosclerosis. In our HIV cohort, LPS levels were increased and positively associated with serum TMA, but the relationship of TMA to plaque features remained independent of LPS. Assessment of specific gut flora was beyond the scope of the current data, but it will be interesting to compare gut flora with serum TMA in future studies of choline metabolism in HIV and non-HIV-infected subjects.
In this study, HIV-infected and non-infected subjects had similar number of calcified plaque segments, but HIV-infected subjects had a significantly higher number of non-calcified plaque segments. Our group has previously shown that HIV-infected patients demonstrate a higher prevalence of non-calcified plaque—a “vulnerable” plaque feature more prone to rupture and increased atherosclerotic events [1, 20]. However, in this study, we demonstrate that serum TMA was most consistently related to calcified plaque characteristics. As calcified and non-calcified plaque phenotypes have divergent pathophysiologies, factors other than choline-related metabolites, including gender, inflammation and chronic immune activation, may play a more considerable role in non-calcified plaque formation in the HIV population [20]. In addition, longstanding effects of choline metabolism on nascent non-calcified lesions may over time contribute to calcified plaque formation.
This study has limitations. The study design was cross-sectional, so causality cannot be inferred. In addition, the role of TMA in CVD, independent of TMAO, needs further clarification. A choline challenge may also be warranted to more clearly accentuate pathophysiologic differences in phosphatidylcholine metabolism in HIV-infected subjects compared to controls. Studies in the HIV population should evaluate expression of FMO3 and composition of the intestinal microbiome to investigate these relationships to TMA production. In addition, the effect of ART use and liver inflammation on FMO3 expression should be investigated in HIV.
Novel data from the current study suggest a potential contribution of choline metabolism to CVD in HIV-infected patients, assessed in relationship to plaque burden using a sensitive CT angiographic technique. This relationship may derive from altered gut flora or microbial translocation unique to HIV-infected patients, and further studies are required to answer these important questions, particularly as to the role of the gut microbiome and related metabolism in the HIV population.
Supplementary Material
Acknowledgments
Funding: This work was supported by Bristol Myers Squibb, Inc; the National Institutes of Health K24 DK064545 and R01 HL095123-04 to S.K.G., K23 HL092792 to J.L., NS37654, M01 RR01066, 1 UL1 RR025758-01 and UL1 TR001102, Harvard Clinical and Translational Science Center, from the National Center for Research Resources, and P30 DK040561 from the Nutrition Obesity Research Center at Harvard; the Institute of Cardiometabolism and Nutrition (ICAN, ANR-10-IAHU-05) to D.G. and J.C.; the European Commission (METACARDIS HEALTH-F4-2012-305312) to D.G; and an equipment grant from Ile de France Region (CORDDIM 120167) to D.G. Funding sources had no role in the design of the study, data analysis or the writing of the manuscript.
The investigators would like to thank Justine Dubus for technical assistance and the patients and the nursing staff on the MGH CRC for their dedicated patient care as well as the volunteers who participated in this study.
Footnotes
Clinical Trial Registration: NCT00455793
Principal contributions of the authors are study conception (JL, SKG), study design (JL, SKG), subject recruitment and implementation of the protocol (KVF, JL), data acquisition and analysis and database management (SS, KVF, JL, HK, RK, KW, SA, DG, JC, FB), statistical analysis and interpretation (SS, KVF, SKG), drafting of the manuscript (SS, KVF, RK, HK, DG), critical revision of the manuscript (SS, KVF, JL, HK, SA, DG, JC, FB, SKG), and supervision of the study (SKG).
Declaration of Interest: SS, KVF, JL, HK, RK, KW, SA and DG have nothing to declare. JC has received research funding from Janssen, ViiV and Merck and honoraria for conferences from Janssen, ViiV, Merck, Gilead, Bristol-Myers Squibb and Chugai unrelated to this manuscript. FB has received honoraria (advisory board and lectures) from Amgen, AstraZeneca and Sanofi unrelated to this manuscript. SKG has received research funding from Bristol-Myers Squibb, Immunex, Gilead, and EMD Serono and served as a consultant for Navidea Inc., Theratechnologies, Novo Nordisk and BMS/AstraZeneca unrelated to this manuscript.
References
- 1.Lo J, Abbara S, Shturman L, Soni A, Wei J, Rocha-Filho JA, et al. Increased prevalence of subclinical coronary atherosclerosis detected by coronary computed tomography angiography in HIV-infected men. AIDS. 2010;24:243–253. doi: 10.1097/QAD.0b013e328333ea9e. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Triant VA, Lee H, Hadigan C, Grinspoon SK. Increased acute myocardial infarction rates and cardiovascular risk factors among patients with human immunodeficiency virus disease. J Clin Endocrinol Metab. 2007;92:2506–2512. doi: 10.1210/jc.2006-2190. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Lang S, Mary-Krause M, Cotte L, Gilquin J, Partisani M, Simon A, et al. Increased risk of myocardial infarction in HIV-infected patients in France, relative to the general population. AIDS. 2010;24:1228–1230. doi: 10.1097/QAD.0b013e328339192f. [DOI] [PubMed] [Google Scholar]
- 4.Subramanian S, Tawakol A, Burdo TH, Abbara S, Wei J, Vijayakumar J, et al. Arterial inflammation in patients with HIV. JAMA. 2012;308:379–386. doi: 10.1001/jama.2012.6698. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Boccara F. Cardiovascular complications and atherosclerotic manifestations in the HIV-infected population: type, incidence and associated risk factors. Aids. 2008;22 (Suppl 3):S19–26. doi: 10.1097/01.aids.0000327512.76126.6e. [DOI] [PubMed] [Google Scholar]
- 6.Desvarieux M, Boccara F, Meynard JL, Bastard JP, Mallat Z, Charbit B, et al. Infection duration and inflammatory imbalance are associated with atherosclerotic risk in HIV-infected never-smokers independent of antiretroviral therapy. AIDS. 2013;27:2603–2614. doi: 10.1097/QAD.0b013e3283634819. [DOI] [PubMed] [Google Scholar]
- 7.Tremaroli V, Backhed F. Functional interactions between the gut microbiota and host metabolism. Nature. 2012;489:242–249. doi: 10.1038/nature11552. [DOI] [PubMed] [Google Scholar]
- 8.Turnbaugh PJ, Ley RE, Mahowald MA, Magrini V, Mardis ER, Gordon JI. An obesity-associated gut microbiome with increased capacity for energy harvest. Nature. 2006;444:1027–1031. doi: 10.1038/nature05414. [DOI] [PubMed] [Google Scholar]
- 9.Karlsson FH, Fak F, Nookaew I, Tremaroli V, Fagerberg B, Petranovic D, et al. Symptomatic atherosclerosis is associated with an altered gut metagenome. Nat Commun. 2012;3:1245. doi: 10.1038/ncomms2266. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Lozupone CA, Li M, Campbell TB, Flores SC, Linderman D, Gebert MJ, et al. Alterations in the gut microbiota associated with HIV-1 infection. Cell Host Microbe. 2013;14:329–339. doi: 10.1016/j.chom.2013.08.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Nazli A, Chan O, Dobson-Belaire WN, Ouellet M, Tremblay MJ, Gray-Owen SD, et al. Exposure to HIV-1 directly impairs mucosal epithelial barrier integrity allowing microbial translocation. PLoS Pathog. 2010;6:e1000852. doi: 10.1371/journal.ppat.1000852. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Dinh DM, Volpe GE, Duffalo C, Bhalchandra S, Tai AK, Kane AV, et al. Intestinal Microbiota, Microbial Translocation, and Systemic Inflammation in Chronic HIV Infection. J Infect Dis. 2014 doi: 10.1093/infdis/jiu409. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Vujkovic-Cvijin I, Dunham RM, Iwai S, Maher MC, Albright RG, Broadhurst MJ, et al. Dysbiosis of the gut microbiota is associated with HIV disease progression and tryptophan catabolism. Sci Transl Med. 2013;5:193ra191. doi: 10.1126/scitranslmed.3006438. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Mutlu EA, Keshavarzian A, Losurdo J, Swanson G, Siewe B, Forsyth C, et al. A compositional look at the human gastrointestinal microbiome and immune activation parameters in HIV infected subjects. PLoS Pathog. 2014;10:e1003829. doi: 10.1371/journal.ppat.1003829. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Koeth RA, Wang Z, Levison BS, Buffa JA, Org E, Sheehy BT, et al. Intestinal microbiota metabolism of L-carnitine, a nutrient in red meat, promotes atherosclerosis. Nat Med. 2013;19:576–585. doi: 10.1038/nm.3145. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Wang Z, Tang WH, Buffa JA, Fu X, Britt EB, Koeth RA, et al. Prognostic value of choline and betaine depends on intestinal microbiota-generated metabolite trimethylamine-N-oxide. Eur Heart J. 2014;35:904–910. doi: 10.1093/eurheartj/ehu002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Wang Z, Klipfell E, Bennett BJ, Koeth R, Levison BS, Dugar B, et al. Gut flora metabolism of phosphatidylcholine promotes cardiovascular disease. Nature. 2011;472:57–63. doi: 10.1038/nature09922. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Bennett BJ, de Aguiar Vallim TQ, Wang Z, Shih DM, Meng Y, Gregory J, et al. Trimethylamine-N-oxide, a metabolite associated with atherosclerosis, exhibits complex genetic and dietary regulation. Cell Metab. 2013;17:49–60. doi: 10.1016/j.cmet.2012.12.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Burdo TH, Lo J, Abbara S, Wei J, DeLelys ME, Preffer F, et al. Soluble CD163, a novel marker of activated macrophages, is elevated and associated with noncalcified coronary plaque in HIV-infected patients. J Infect Dis. 2011;204:1227–1236. doi: 10.1093/infdis/jir520. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Fitch KV, Srinivasa S, Abbara S, Burdo TH, Williams KC, Eneh P, et al. Noncalcified Coronary Atherosclerotic Plaque and Immune Activation in HIV-Infected Women. J Infect Dis. 2013;208:1737–1746. doi: 10.1093/infdis/jit508. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Agatston AS, Janowitz WR, Hildner FJ, Zusmer NR, Viamonte M, Jr, Detrano R. Quantification of coronary artery calcium using ultrafast computed tomography. J Am Coll Cardiol. 1990;15:827–832. doi: 10.1016/0735-1097(90)90282-t. [DOI] [PubMed] [Google Scholar]
- 22.Borkan GA, Gerzof SG, Robbins AH, Hults DE, Silbert CK, Silbert JE. Assessment of abdominal fat content by computed tomography. Am J Clin Nutr. 1982;36:172–177. doi: 10.1093/ajcn/36.1.172. [DOI] [PubMed] [Google Scholar]
- 23.Rajaie S, Esmaillzadeh A. Dietary choline and betaine intakes and risk of cardiovascular diseases: review of epidemiological evidence. ARYA Atheroscler. 2011;7:78–86. [PMC free article] [PubMed] [Google Scholar]
- 24.Olthof MR, Brink EJ, Katan MB, Verhoef P. Choline supplemented as phosphatidylcholine decreases fasting and postmethionine-loading plasma homocysteine concentrations in healthy men. Am J Clin Nutr. 2005;82:111–117. doi: 10.1093/ajcn.82.1.111. [DOI] [PubMed] [Google Scholar]
- 25.Yates AA, Schlicker SA, Suitor CW. Dietary Reference Intakes: the new basis for recommendations for calcium and related nutrients, B vitamins, and choline. J Am Diet Assoc. 1998;98:699–706. doi: 10.1016/S0002-8223(98)00160-6. [DOI] [PubMed] [Google Scholar]
- 26.Tang WH, Wang Z, Levison BS, Koeth RA, Britt EB, Fu X, et al. Intestinal microbial metabolism of phosphatidylcholine and cardiovascular risk. N Engl J Med. 2013;368:1575–1584. doi: 10.1056/NEJMoa1109400. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Bidulescu A, Chambless LE, Siega-Riz AM, Zeisel SH, Heiss G. Usual choline and betaine dietary intake and incident coronary heart disease: the Atherosclerosis Risk in Communities (ARIC) study. BMC Cardiovasc Disord. 2007;7:20. doi: 10.1186/1471-2261-7-20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Dalmeijer GW, Olthof MR, Verhoef P, Bots ML, van der Schouw YT. Prospective study on dietary intakes of folate, betaine, and choline and cardiovascular disease risk in women. Eur J Clin Nutr. 2008;62:386–394. doi: 10.1038/sj.ejcn.1602725. [DOI] [PubMed] [Google Scholar]
- 29.Zhang J, Chaluvadi MR, Reddy R, Motika MS, Richardson TA, Cashman JR, et al. Hepatic flavin-containing monooxygenase gene regulation in different mouse inflammation models. Drug Metab Dispos. 2009;37:462–468. doi: 10.1124/dmd.108.025338. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Brenchley JM, Price DA, Schacker TW, Asher TE, Silvestri G, Rao S, et al. Microbial translocation is a cause of systemic immune activation in chronic HIV infection. Nat Med. 2006;12:1365–1371. doi: 10.1038/nm1511. [DOI] [PubMed] [Google Scholar]
- 31.Pedersen KK, Pedersen M, Troseid M, Gaardbo JC, Lund TT, Thomsen C, et al. Microbial translocation in HIV infection is associated with dyslipidemia, insulin resistance, and risk of myocardial infarction. J Acquir Immune Defic Syndr. 2013;64:425–433. doi: 10.1097/QAI.0b013e31829f919d. [DOI] [PubMed] [Google Scholar]
- 32.Kelesidis T, Kendall MA, Yang OO, Hodis HN, Currier JS. Biomarkers of microbial translocation and macrophage activation: association with progression of subclinical atherosclerosis in HIV-1 infection. J Infect Dis. 2012;206:1558–1567. doi: 10.1093/infdis/jis545. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.