Secondary Logo

Journal Logo

Basic and Translational Science

Microbiota-Dependent Marker TMAO Is Elevated in Silent Ischemia but Is Not Associated With First-Time Myocardial Infarction in HIV Infection

Haissman, Judith M. MD*; Knudsen, Andreas MD, PhD†,‡; Hoel, Hedda MD§; Kjær, Andreas MD, PhD; Kristoffersen, Ulrik S. MD, PhD; Berge, Rolf K. PhD‖,¶; Katzenstein, Terese L. MD, PhD*; Svardal, Asbjørn PhD; Ueland, Thor PhD#,**,††; Aukrust, Pål MD, PhD§,#,**,††; Lebech, Anne-Mette MD, PhD, DMSc; Nielsen, Susanne D. MD, DMSc*; Trøseid, Marius MD, PhD§,#,**,††,‡‡

Author Information
JAIDS Journal of Acquired Immune Deficiency Syndromes: February 1, 2016 - Volume 71 - Issue 2 - p 130-136
doi: 10.1097/QAI.0000000000000843



Since the introduction of antiretroviral therapy (ART), AIDS and HIV-related mortality has declined in HIV-infected persons, although coronary heart disease (CHD) has emerged as a leading cause of morbidity and mortality.1,2 A higher prevalence of traditional CHD risk factors, such as smoking, dyslipidemia, diabetes, family history of CHD, and hypertension, have been observed in HIV-infected persons compared with the general population.3,4 However, increasing evidence indicates that HIV infection is associated with increased risk of CHD beyond that explained by the higher burden of traditional risk factors among HIV-infected persons.5–7 Several studies have reported HIV-related inflammation, immune activation, and microbial translocation being associated with increased risk of CHD, still the pathogenesis of CHD in HIV infection is not fully understood.4,8,9

HIV infection causes substantial loss of intestinal CD4+ T cells and deterioration of gut-associated lymphoid tissue resulting in increased intestinal permeability.10,11 Subsequently, leakage of bacterial products such as lipopolysaccharide into the bloodstream occurs, promoting low-grade systemic inflammation through interaction with Toll-like receptors on monocytes/macrophages and dendritic cells.12–14 HIV-associated gut-associated lymphoid tissue depletion also causes changes in the composition of the gut microbiota.15 Recent studies have demonstrated that altered composition of microbiota in HIV infection is associated with increased systemic inflammation,16–18 suggesting a potential association between altered microbiota and CHD in HIV infection.

In the general population, a link between the microbiota-dependent metabolite trimethylamine-N-oxide (TMAO) and cardiovascular risk has recently been established.19–22 TMAO has been reported to promote atherosclerosis through foam cell formation and interference with reverse cholesterol transport from the atherosclerotic plaque,20,23,24 and elevated TMAO has been shown to be predictive of cardiovascular events.19,20 TMAO is formed in the liver from trimethylamine (TMA), a product exclusively generated by the gut microbiota from dietary phosphatidylcholine, choline, and carnitine, which in turn are abundant in various food sources including eggs, dairy products, and red meat.20,23 TMAO may also be generated from betaine which can be produced by oxidation of choline or endogenously, independent of choline.25

Only 1 published study has investigated TMAO in HIV infection. A link between TMA and coronary plaque burden in HIV-infected individuals was found, but this was not the case for TMAO.26 There are no published data regarding the use of TMAO as a marker of clinical CHD in HIV infection. Given the potential role of gut microbiota in HIV pathogenesis and the atherosclerotic process, we hypothesized that TMAO is a marker of both subclinical and clinical CHD in HIV-infected persons.

To explore this hypothesis, we measured TMAO and its precursors choline and betaine in 2 previously described cohorts: One cross-sectional cohort of asymptomatic HIV-infected persons and uninfected asymptomatic controls assessed with myocardial perfusion scintigraphy, intima-media thickness (IMT), and coronary artery calcium scoring (CACS),27 and 1 cohort of HIV-infected individuals with first-time myocardial infarction (MI) compared with HIV-infected individuals without MI.28


Cross-sectional Cohort

The cross-sectional cohort comprised 105 asymptomatic HIV-infected persons on continuous ART and with no previous history of CHD. All participants were recruited at the outpatient clinic at the Department of Infectious Diseases, University of Copenhagen, Hvidovre University Hospital. The control population comprised 105 asymptomatic HIV-uninfected persons matched for age, sex, and smoking status and with no previous history of CHD. The cohort has previously been described in detail.27 Myocardial perfusion scintigraphy, CACS, carotid IMT, the 10-year risk of cardiovascular events (Framingham risk score), and biochemical analyses including CD4+ count, HIV-RNA, cholesterol, triglycerides, and glucose were performed and have previously been described.27 Clinical characteristics of the cross-sectional cohort are given in Table 1. All study subjects had blood samples performed after inclusion in the trial. Although a fasting state was generally encouraged, this was not systematically recorded. Plasma was stored at −80°C until analysis.

Baseline Characteristics of the Cross-Sectional Cohort

Nested Case–control Study

The nested case–control study comprised 55 HIV-infected persons with confirmed first-time MI and no known previous CHD or diabetes mellitus. A total of 182 HIV-infected controls matched for age, duration of ART, sex, and smoking status were included. Smoking was recorded with a dichotomous status of either smoking (ever) or nonsmoking (never). Cases with MI from January 1998 to December 2008 were identified through the Danish National Hospital Registry and the Danish HIV Cohort Study, as previously described.28 Due to the matching criteria, cases and controls were recruited from the same period. All participants were followed at the Departments of Infectious Diseases at University of Copenhagen, Rigshospitalet or Hvidovre University Hospital, Denmark. As part of the routine visits by HIV-infected patients, plasma samples are systematically collected and stored at −80°C. Although a fasting state was generally encouraged, this was not systematically recorded. For each participant 4 samples were selected; (1) before initiation of ART {a median of 2133 days before MI [interquartile range (IQR) 1039–3045]}, (2) 3 months after initiation of ART [a median of 2020 days before MI (IQR 916–2990)], (3) 1 year before MI [a median of 334 days before MI (IQR 292–367)], and (4) the last sample available before MI [median of 51 days before MI (IQR 27–82), range: 0–239 days] and similar time point for controls; (1) before initiation of ART [a median of 2223 days before case patient's MI (IQR 1057–3231)], (2) 3 months after initiation of ART [a median of 2092 days before case patient's MI (IQR 882–2980)], (3) 1 year before MI [a median of 368 days before case patient's MI (IQR 334–408)], and (4) the last sample available before case patient's MI [median: 0 (IQR 0–0 days)]. The cohort has previously been described in detail, including demographic data, CD4+ count, HIV-RNA, lipid status, ART exposure, and coinfection with hepatitis B virus (HBV) and hepatitis C virus (HCV).28 Clinical characteristics of the cross-sectional cohort are given in Table 2.

Baseline Characteristics of the Nested Case–control Study

Ethics and Informed Consent

Written informed consent, including consent to store plasma and perform further analysis on blood samples, was obtained from all participants. The studies were performed in accordance with the Helsinki Declaration and approved by the Scientific Ethical Committee of the Capital Region of Denmark (H-C-2008-060 and H-D-2008-108).

Measurement of TMAO, Choline, and Betaine

Stable isotope dilution liquid chromatography with tandem mass spectrometric was used for quantification of TMAO, choline, and betaine as previously described.29,30 Valid measurements were obtained for all HIV-infected persons and uninfected controls from the cross-sectional cohort. For the nested case–control study, valid measurements were obtained for 51 cases and 164 controls at time point 1; 49 cases and 160 controls at time point 2; 50 cases and 171 controls at time point 3; 52 cases and 174 controls at time point 4, giving a total number of measurements of 202 from cases and of 676 from controls.

Statistical Analyses

Student t test or Mann Whitney U test were used to compare groups, and continuous data are given as mean (95% confidence interval), except TMAO, betaine, choline, triglycerides, and HIV-RNA, which are given as median (IQR). Wilcoxon paired test was used to evaluate differences in TMAO between time points. Trend analyses through tertiles of TMAO were performed with χ2 test for categorical data and with analysis of variance for continuous data, and skewed data (triglycerides and HIV-RNA) were log-transformed as appropriate. In the nested case–control study, each of the 4 time points were analyzed separately. In addition, area under the curve (AUC) for all 4 time points (n = 202 cases and 676 controls) was analyzed. Analyses were performed using IBM SPSS 21 (SPSS Inc., Quarry Bay, Hong Kong), and a P-value less than 0.05 was considered statistically significant.


Cross-sectional Study

The cross-sectional cohort included 105 symptomatic HIV-infected persons with a mean age of 47 years (range: 18–70 years), consisted of 89% men, and 47% current or former smokers. Participants had all been on ART for a minimum of 1 year, 90% had undetectable viral load (HIV-RNA < 40 copies/mL), and the mean CD4+ T cell count was 636 cells per microliter (Table 1). The 105 uninfected controls were matched for age, sex, and smoking status (Table 1).

TMAO and Subclinical Atherosclerosis

No difference was found in plasma TMAO or choline between HIV-infected persons and uninfected controls, whereas betaine was lower in HIV-infected persons (Table 3).

Comparison of HIV-Infected Persons and Controls and HIV-Infected Persons With and Without MPD

Among the HIV-infected persons, 18% had myocardial perfusion defects (MPD) vs. 0% among the uninfected controls.27 When comparing HIV-infected persons with and without MPD, TMAO was elevated in those with perfusion defects [3.9 μM (3.1–6.3) vs. 3.1 μM (2.2–4.0), P = 0.025, Table 3]. Furthermore, in light of previous studies suggesting a potential threshold above which TMAO becomes predictive of cardiovascular events,19,29 we divided the HIV-infected population in groups according to TMAO levels above or below 5 μM, revealing that individuals with TMAO levels higher than 5 μM had a significantly higher proportion of MPD compared with the group of HIV-infected individuals with TMAO levels below 5 μM; 8 of 21 (38.1%) vs. 11 of 84 (13.1%), P = 0.008. In contrast, no difference was found in choline and betaine when comparing HIV-infected persons with and without MPD (Table 3).

Tertiles of TMAO were not associated with CACS score, IMT, Framingham risk score, or any traditional or HIV-related risk factors (Table 1). Furthermore, there was a significant trend to increasing ART duration and use of protease inhibitors (PI) by increasing TMAO tertiles (Table 1). However, comparing subjects on PIs vs. nonnucleoside reverse transciptase inhibitors (NNRTIs), there was no significant difference in levels of TMAO (P = 0.2).

Nested Case–control Study

HIV-infected persons included in the case–control study had a median age of 49 years, consisted of 92% men, and were 95% current or former smokers. A total of 55 HIV-infected persons had confirmed first-time MI (cases), including 32 cases with ST-elevation MI, 16 cases with non-ST-elevation MI, and 7 cases with nonspecified MI. A total of 182 HIV-infected controls without MI and matched for age, duration of ART, sex, and smoking were also included.28 All HIV-infected cases and controls were followed from before initiation of ART to last sample before case's MI (sample 4, collected at median: 51, range: 0–239 days before MI). After introduction of ART, all patients had been exposed to nucleoside reverse transciptase inhibitors, an equal proportion of cases and controls received PI, whereas a significantly higher proportion of cases had received NNRTIs.28 More than 80% of the patients, both among cases and controls, had suppressed viral load (<400 copies/mL) and CD4+ T cell count above 400 cells per microliter when plasma sample 4 was collected.28

TMAO in Relation to MI

MI cases and controls were compared at all 4 time points. TMAO, choline, and betaine did not differ between MI cases and controls at any time point (Fig. 1). Analyzing all 4 time points together (AUC), there were still no differences in concentration of TMAO, choline, or betaine between cases and controls (see Table S1, Supplemental Digital Content, Looking at number of MI cases through tertiles of TMAO, there was no trend toward increasing number of MI with higher concentration of TMAO at time point 1 (χ2P = 0.2), time point 2 (P = 0.6), time point 3 (P = 0.3), or time point 4 (P = 0.9). Furthermore, at baseline before initiation of ART, tertiles of TMAO were not associated with traditional cardiovascular risk factors, CD4+ T cell count, HIV-RNA, HBV status, or HCV status (Table 2). In contrast, an association with age (positive) and renal function (negative) was found (Table 2). Moreover, using TMAO levels above 5 μM as a cutoff in logistic regression analyses, univariate odds ratios decreased after introduction of ART, from 1.6 at time point 1 and 2 to 0.9 at time point 3 and to 0.8 at time point 4.

TMAO, choline, and betaine are not predictive of MI concentrations of TMAO (A), choline (B), and betaine (C) in cases and controls in the nested case–control study at 4 consecutive time points (T); (1) before initiation of ART, (2) 3 months after initiation of ART, (3) 1 year before the case's MI or similar time point for controls, and (4) the last sample available before the case's MI (sample collected median: 51, range: 0–239 days before MI) or similar time point for controls. Data are given as median and IQR. Time points were compared using Wilcoxon paired test. No significant differences were observed between cases and controls at any time point. TMAO = trimethylamine-N-oxide.


Assessment of the impact of ART exposure, including PI and NNRTI on TMAO, was performed in the nested case–control study cohort. A significant increase in TMAO was observed from pre-ART time point 1 [3.42 (IQR 2.09–5.14) μM] to first time point after ART introduction, time point 2 [3.61 (2.53–5.42) μM, P = 0.032]. TMAO continued to increase significantly during ART from time point 2 to time point 3 [4.43 (2.74–6.41) μM, P = 0.007], reaching a steady level with no further increase between time point 3 and time point 4 [4.25 (2.72–6.92) μM, P = 0.657] (Fig. 2). Use of PI was associated with higher TMAO concentration at time point 2 [2.91 (2.06–4.51) vs. 3.82 (2.76–6.11) μM, P = 0.025] and time point 4 [3.69 (2.33–5.76) vs. 4.28 (3.02–6.86) μM, P = 0.015]. The same pattern was seen in cases and controls. The use of NNRTI was not associated with increase of TMAO at any time point.

Increase in TMAO after ART introduction. TMAO concentrations in the total cohort of the nested case–control study at 4 consecutive time points (T); (1) before initiation of ART, (2) 3 months after initiation of ART, (3) 1 year before the case's MI or similar time point for controls, and (4) the last sample available before the case's MI (sample collected median: 51, range: 0–239 days before MI) or similar time point for controls. Data are given as median and IQR. Time points were compared using Wilcoxon paired test. *P < 0.05, **P < 0.001 given for comparison of versus T1. †P < 0.05, ††P < 0.01 given for comparison vs. T2. No significant differences were found comparing T3 vs. T4. TMAO, trimethylamine-N-oxide.


In this study, the prognostic value of the microbiota-dependent metabolite TMAO in relation to both silent ischemia and MI was evaluated using 2 different cohorts. Major findings were (1) TMAO was elevated in HIV-infected persons with silent ischemia; (2) TMAO was not associated with first-time MI in HIV-infected persons; and (3) ART and in particular use of PI induced an increase in TMAO. These findings are in contrast to several publications from HIV-uninfected populations, where TMAO has been found to give prognostic information in relation to cardiovascular events.19,20,23,25,29,31–33

Elevated TMAO has been reported to predict atherosclerosis-related cardiovascular events such as MI and stroke in patients undergoing elective coronary angiography,25 and we and others have recently showed that TMAO is predictive of clinical outcome in patients with heart failure.29,32,33 To date, very few studies have evaluated the association between TMAO and atherosclerosis in HIV infection. A recent study reported a link between TMA, but not TMAO, and coronary plaque burden in asymptomatic HIV-infected individuals.26 In contrast, a published abstract reported an association between TMAO and IMT, an association which was attenuated after multivariate adjustment.34

In our cross-sectional cohort, we found elevated TMAO in HIV-infected persons with MPD but no association with IMT or CACS score. Furthermore, TMAO was not associated with first-time MI. The same negative findings were seen both when using baseline TMAO before initiation of ART, TMAO during follow-up, and TMAO tertiles or AUC for all time points. Although relatively few patients were studied, this could question the usefulness of TMAO as a biomarker to predict cardiovascular events in HIV infection.

TMA is generated by the gut microbiota from dietary phosphatidylcholine and is converted in the liver to TMAO by hepatic flavin-containing monooxygenase 3 (FMO3). Thus, it has been speculated that impaired liver function or interaction with ART metabolism in the liver could impact concentration of TMAO in HIV-infected persons.26 Of note, TMAO increased after initiation of ART, which is in contrast to most cardiovascular biomarkers that tend to decrease after ART initiation.35 Highest TMAO was found in patients receiving PI. Hence, use of ART and PI, in particular, may have influenced TMAO and have diluted a potential association with first-time MI in the nested case–control study. Correspondingly, in the cross-sectional cohort, there was a trend to increasing ART duration and use of PI by increasing TMAO tertiles. Our findings are in line with recent data showing increased TMAO in patients receiving ART compared with untreated patients.34 In contrast, no difference in TMAO was found according to HBV status or HCV status in the nested case–control study or by alanine aminotransferase in the cross-sectional cohort, in line with a previous report,26 suggesting that PI induced effects on FMO3 rather than impaired liver function influenced TMAO levels. Finally, the association between TMAO and renal function and age in the nested case–control study is in line with our previous report from heart failure patients.29

The different ability of TMAO to predict cardiovascular events in HIV infection as compared with other patient populations could be influenced by several confounding factors, in addition to potentially PI-induced effects on FMO3. In particular, the exceptionally high-frequency of smokers (ever) in the nested case–control study may reduce the potential impact of other risk factors, including TMAO. In addition, the relatively low levels of TMAO in our study could be relevant, as in the original article on HIV seronegative subjects19 and also in a previous article by our group,29 there seems to be a potential threshold, above which TMAO becomes predictive of cardiovascular events. Hence, using increased TMAO levels above 5 μM as a cutoff for logistic regression analyses in the case control study, we found that univariate odds ratios for first-time MI decreased after introduction of ART, and it is possible that TMAO is a better biomarker of cardiovascular events in untreated populations. Another point is the relatively young age of our cohorts compared with the initial published study of HIV seronegative subjects,19 as we have previously found TMAO levels to correlate with increasing age.29 Hence, although TMAO was not associated with first-time MI in this study, other conclusions could be reached in other cohorts. In light of a rapidly aging HIV-population, we suggest that separate studies should be undertaken in elderly cohorts of HIV-infected subjects.

The diverging findings of TMAO being elevated in silent ischemia without being predictive of MI are not intuitively understood. Of note, TMAO has been proposed to interfere with lipid transportation from the circulation and further to the vessel wall,20,23,24 a mechanism that could have contributed to the association of TMAO with MPD in the cross-sectional cohort. However, despite an association with stable silent atherosclerosis, TMAO may not necessarily be associated with unstable, rupture-prone plaques, which occur more frequently in HIV infection.36 Moreover, as MPD are indicative of increased risk of MI, TMAO may still play a pathogenic role in development of atherosclerosis in HIV infection, which could translate into increased risk of clinical events over time.

To the best of our knowledge, no published studies have evaluated the specific HIV-associated microbiota and its relation to TMA and TMAO before and after initiation of ART. It is assumed that TMA-producing bacteria are typically localized in the colon and mostly anaerobic species.37 As it has been reported that several taxae from the phylum proteobacteria, and the genus Prevotella, which are enriched in HIV-infected subjects,15 have the capacity to produce TMA,37 we hypothesized that TMAO levels would be elevated in HIV-infected persons compared with uninfected controls. Although the lack of difference in TMAO levels between HIV-infected and uninfected could have been a power issue, our findings fit with the few studies published.26,34 Interestingly, one of the precursors of TMAO, betaine, was lower in HIV-infected persons. Betaine has been reported to be negatively associated with the metabolic syndrome,38 and as HIV-infected persons had higher blood pressure, triglycerides, and glucose in this study, negative association with the metabolic syndrome could have been a contributing factor for the lower betaine levels and lack of elevated TMAO levels in HIV-infected individuals.

Limitations of our study include the low number of cases in the nested case–control study, and we cannot exclude the possibility of statistical type II errors. However, although subtle differences may have been overlooked, the absolute levels of TMAO were of similar magnitude in cases and controls, even when comparing AUC based on all 4 time points, and there was no trend toward increasing number of MI cases through tertiles of TMAO. Other limitations are the potential confounders discussed previously, in particular, the relatively low age and the high-frequency of smokers. Furthermore, as the plasma samples were collected as part of routine clinical care, dietary intake of carnitine or phosphatidylcholine could have influenced TMAO levels. Moreover, detailed data about dietary habits and gut microbiota analyses would have provided valuable information.

In conclusion, TMAO was elevated in HIV-infected persons with MPD but was not associated with first-time MI. However, TMAO increased after initiation of ART and was associated with the use of PI. Hence, our data question TMAO as a useful biomarker of cardiovascular risk in HIV infection, at least in ART-treated individuals.


All study participants are thanked sincerely for their participation in the studies.


1. Triant VA. Epidemiology of coronary heart disease in patients with human immunodeficiency virus. Rev Cardiovasc Med. 2014;15(suppl 1):S1–S8.
2. Causes of death in HIV-1-infected patients treated with antiretroviral therapy, 1996-2006: collaborative analysis of 13 HIV cohort studies. Clin Infect Dis. 2010;50:1387–1396.
3. Triant VA, Lee H, Hadigan C, et al.. Increased acute myocardial infarction rates and cardiovascular risk factors among patients with human immunodeficiency virus disease. J Clin Endocrinol Metab. 2007;92:2506–2512.
4. So-Armah K, Freiberg MS. Cardiovascular disease risk in an aging HIV population: not just a question of biology. Curr Opin HIV AIDS. 2014;9:346–354.
5. Silverberg MJ, Leyden WA, Xu L, et al.. Immunodeficiency and risk of myocardial infarction among HIV-positive individuals with access to care. J Acquir Immune Defic Syndr. 2014;65:160–166.
6. Freiberg MS, Chang CC, Kuller LH, et al.. HIV infection and the risk of acute myocardial infarction. JAMA Intern Med. 2013;173:614–622.
7. Islam FM, Wu J, Jansson J, et al.. Relative risk of cardiovascular disease among people living with HIV: a systematic review and meta-analysis. HIV Med. 2012;13:453–468.
8. Hunt PW. HIV and inflammation: mechanisms and consequences. Curr HIV/AIDS Rep. 2012;9:139–147.
9. Zanni MV, Grinspoon SK. HIV-specific immune dysregulation and atherosclerosis. Curr HIV/AIDS Rep. 2012;9:200–205.
10. Stockmann M, Schmitz H, Fromm M, et al.. Mechanisms of epithelial barrier impairment in HIV infection. Ann N Y Acad Sci. 2000;915:293–303.
11. 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.
12. Troseid M, Manner IW, Pedersen KK, et al.. Microbial translocation and cardiometabolic risk factors in HIV infection. AIDS Res Hum Retroviruses. 2014;30:514–522.
13. Latz E, Visintin A, Lien E, et al.. Lipopolysaccharide rapidly traffics to and from the golgi apparatus with the toll-like receptor 4-MD-2-CD14 complex in a process that is distinct from the initiation of signal transduction. J Biol Chem. 2002;277:47834–47843.
14. da SB, Singer W, Fong IW, et al.. In vivo cytokine and neuroendocrine responses to endotoxin in human immunodeficiency virus-infected subjects. J Infect Dis. 1999;180:106–115.
15. Lozupone CA, Rhodes ME, Neff CP, et al.. HIV-induced alteration in gut microbiota: driving factors, consequences, and effects of antiretroviral therapy. Gut Microbes. 2014;5:562–570.
16. Dinh DM, Volpe GE, Duffalo C, et al.. Intestinal microbiota, microbial translocation, and systemic inflammation in chronic HIV infection. J Infect Dis. 2015;211:19–27.
17. Mutlu EA, Keshavarzian A, Losurdo J, et al.. A compositional look at the human gastrointestinal microbiome and immune activation parameters in HIV infected subjects. PLoS Pathog. 2014;10:e1003829.
18. Vujkovic-Cvijin I, Dunham RM, Iwai S, et al.. Dysbiosis of the gut microbiota is associated with HIV disease progression and tryptophan catabolism. Sci Transl Med. 2013;5:193ra91.
19. Tang WH, Wang Z, Levison BS, et al.. Intestinal microbial metabolism of phosphatidylcholine and cardiovascular risk. N Engl J Med. 2013;368:1575–1584.
20. Wang Z, Klipfell E, Bennett BJ, et al.. Gut flora metabolism of phosphatidylcholine promotes cardiovascular disease. Nature. 2011;472:57–63.
21. Tang WH, Hazen SL. The contributory role of gut microbiota in cardiovascular disease. J Clin Invest. 2014;124:4204–4211.
22. Ettinger G, MacDonald K, Reid G, et al.. The influence of the human microbiome and probiotics on cardiovascular health. Gut Microbes. 2014;5:719–728.
23. Koeth RA, Wang Z, Levison BS, et al.. Intestinal microbiota metabolism of L-carnitine, a nutrient in red meat, promotes atherosclerosis. Nat Med. 2013;19:576–585.
24. Bennett BJ, de Aguiar Vallim TQ, Wang Z, et al.. Trimethylamine-N-oxide, a metabolite associated with atherosclerosis, exhibits complex genetic and dietary regulation. Cell Metab. 2013;17:49–60.
25. Wang Z, Tang WH, Buffa JA, et al.. Prognostic value of choline and betaine depends on intestinal microbiota-generated metabolite trimethylamine-N-oxide. Eur Heart J. 2014;35:904–910.
26. Srinivasa S, Fitch KV, Lo J, et al.. Plaque burden in HIV-infected patients is associated with serum intestinal microbiota-generated trimethylamine. AIDS. 2015;29:443–452.
27. Kristoffersen US, Lebech AM, Wiinberg N, et al.. Silent ischemic heart disease and pericardial fat volume in HIV-infected patients: a case-control myocardial perfusion scintigraphy study. PLoS One. 2013;8:e72066.
28. Knudsen A, Moller HJ, Katzenstein TL, et al.. Soluble CD163 does not predict first-time myocardial infarction in patients infected with human immunodeficiency virus: a nested case-control study. BMC Infect Dis. 2013;13:230.
29. Troseid M, Ueland T, Hov JR, et al.. Microbiota-dependent metabolite trimethylamine-N-oxide is associated with disease severity and survival of patients with chronic heart failure. J Intern Med. 2015;277:717–726.
30. Bjorndal B, Burri L, Wergedahl H, et al.. Dietary supplementation of herring roe and milt enhances hepatic fatty acid catabolism in female mice transgenic for hTNFalpha. Eur J Nutr. 2012;51:741–753.
31. Backhed F. Meat-metabolizing bacteria in atherosclerosis. Nat Med. 2013;19:533–534.
32. Tang WH, Wang Z, Fan Y, et al.. Prognostic value of elevated levels of intestinal microbe-generated metabolite trimethylamine-N-oxide in patients with heart failure: refining the gut hypothesis. J Am Coll Cardiol. 2014;64:1908–1914.
33. Tang WH, Wang Z, Shrestha K, et al.. Intestinal microbiota-dependent phosphatidylcholine metabolites, diastolic dysfunction, and adverse clinical outcomes in chronic systolic heart failure. J Card Fail. 2015;21:91–96.
34. Sinha A, Ma Y, Carroll C. TMAO and HIV-Associated Atherosclerosis [Abstract]. In: Sinha A, Ma Y, Carroll C. Proceedings of the Conference on Retroviruses and Opportunistic Infections, Seattle, WA, February 1, 2015; Abstract No 755.
35. Kristoffersen US, Kofoed K, Kronborg G, et al.. Reduction in circulating markers of endothelial dysfunction in HIV-infected patients during antiretroviral therapy. HIV Med. 2009;10:79–87.
36. Zanni MV, Abbara S, Lo J, et al.. Increased coronary atherosclerotic plaque vulnerability by coronary computed tomography angiography in HIV-infected men. AIDS. 2013;27:1263–1272.
37. Koeth RA, Levison BS, Culley MK, et al.. Gamma-Butyrobetaine is a proatherogenic intermediate in gut microbial metabolism of L-carnitine to TMAO. Cell Metab. 2014;20:799–812.
38. Konstantinova SV, Tell GS, Vollset SE, et al.. Divergent associations of plasma choline and betaine with components of metabolic syndrome in middle age and elderly men and women. J Nutr. 2008;138:914–920.

microbiota; TMAO; coronary heart disease; cardiovascular; myocardial infarction; atherosclerosis; ischemia; HIV; ART; protease inhibitor

Supplemental Digital Content

Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.