Beef, Microbes in the Gut, and Heart Disease

An Expert Interview With Stanley L. Hazen , MD, PhD

Linda Brookes, MSc; Stanley L. Hazen, MD, PhD


June 19, 2013

In This Article

The Role of TMAO in Cardiovascular Residual Risk

Medscape: Could this represent the residual risk for cardiovascular disease identified in patients with low LDL cholesterol levels on statin therapy?[8] Recent studies that focused on reducing this risk by increasing low levels of HDL cholesterol do not appear to have been successful.[9,10] Could lowering TMAO be the way to reduce that residual risk?

Dr. Hazen: Perhaps. In the New England Journal of Medicine paper[2] we reported that, statistically, inclusion of TMAO as a covariate resulted in a net 8.6% reclassification for improvement in risk estimation over traditional risk factors (net reclassification improvement, a significant P < .001). I hesitate to call it "residual risk"; it is independent of LDL- and HDL cholesterol levels. Once we can find people at risk, the next step is whether we can inhibit or treat that risk. Obviously these studies need to be done in the future, both attacking this pathway and seeing whether it represents a new therapeutic angle for lowering cardiovascular risk. It does not even have to be drugs, because it seems that nutrition and diet approaches can make a difference. However, we will also need to determine whether, by identifying someone who is at risk, in the same way as with high-sensitivity C-reactive protein, we can identify a population that has a higher risk and then get them to more aggressive goals -- for example, by intensifying statin therapy. Would these people benefit from more aggressive preventive efforts?

Medscape: Would you risk creating a "worried well" population of patients with no symptoms but raised TMAO?

Dr. Hazen: In our cohort of over 1000 primary prevention patients, we found that TMAO predicted future risk quite well. It is almost a CAD risk equivalent, a twofold increase, which is equivalent to the risk of having diabetes in our cohort. And those were people who just had a coronary angiogram and were told that there is less than 50% stenosis in all of their major vessels and they did not have anything to worry about. So, more studies need to be done to see who should have their TMAO measured. The present study was not a small number of people, and the data are pretty strong.

Medscape: How can TMAO be measured nowadays?

Dr. Hazen: TMAO is straightforward to measure. We currently use mass spectrometry, but this is time-consuming and not something that one can order from the hospital lab. We are currently involved in the development of a test using nuclear magnetic resonance (NMR). It turned out that the hydrogens on TMAO are unique and have a different NMR signature. The Cleveland Clinic has licensed the technology for measuring TMAO to LipoScience (Raleigh, North Carolina).[11,12] LipoScience uses NMR-based technology to measure lipoprotein particle subfractions and does about 2 million lipoprotein assays per year. The same technique that measures LDL and HDL particles can be applied to measure TMAO. The plan is for our clinical assay to be available for measuring TMAO through this platform by the end of summer 2013. Their machine has recently been cleared by the US Food and Drug Administration for availability in hospital labs.[13] It will not be available so quickly for TMAO measurement, but at least TMAO testing will be available for research studies before the end of the year.

In the future, just like we do a blood test for cholesterol and triglycerides, we may do a blood test for TMAO for advice on dietary patterns as well as for other bacterial products that we think are biologically linked to glucose metabolism. These are going to be the subjects of future papers.

We should be thinking of our intestinal microbial community as the largest endocrine organ in our body. The nutrients you ingest go through the filter of the intestinal microbes, and depending on the nutrient input and microbial composition, you have a different capacity to make different biologically active compounds. Compounds that diffuse in the blood and act at a different site meet all the definitions of a hormone, so this is why the microbial community can be considered like an endocrine organ, in a way.

Medscape: How would one lower high levels of TMAO? Would it be by diet, or would it mean treating individualized gut microbiota?

Dr. Hazen: Dietary efforts, including reduction in meat consumption, appear to be associated with reduced TMAO levels. As for treating an "individual gut microbe"? There are trillions of microbes, but if you look at the top 95% of them, in terms of broad classifications of family and genus they are pretty much the same in everyone. Proportions may be shifted in someone who eats a more vegetarian diet with more roughage compared with someone who has more carnivorous eating patterns, but the 95% are similar except for some proportional changes. In that 5% is where there is huge variability. If you are talking about a microbe making a product that is biologically active, it need not be abundant; it can be a really small proportion, because it is making something catalytically. It does not have to be one of the major microbe types; it just has to be present in sufficient numbers to produce the compound at a blood level that will allow us to see if it causes whatever phenotype we are looking at.

One of the really intriguing findings of our L-carnitine paper was how substantially different the metabolism was of carnitine in vegetarians and vegans compared with omnivores. Chronic exposure to a diet that included L-carnitine shifted the microbiota composition. It was a subtle shift, but the proportions of the microbes that tracked with TMAO levels in vegan or vegetarian status still accounted for a very small proportion of the microbes in the entire intestines. The vast majority of the microbes are similar among omnivores, vegetarians, and vegans.

Medscape: Does that mean that if you eat any meat at all on a regular basis, whether a small amount or large portions, your gut microbiota are similar?

Dr. Hazen: We do not know. To be defined as a vegetarian or a vegan in our studies, participants had to claim to have no meat product during the past year. How long you have to change your diet before you see a shift, we do not know. But at least now we can measure blood levels of TMAO, and it is going to be much easier than trying to look at the entire microbiome in the stool.

Medscape: In your studies you used antibiotics to suppress the gut microbiota, but presumably treatment with antibiotics to suppress production of TMAO would not be a good idea.

Dr. Hazen: Right. In fact, not only would it not be a good idea, it would be futile. What we already saw, at least in mice, is that when we initially did our experiments with only one antibiotic, we completely eliminated TMAO levels at the beginning, but when we harvested the aortas just over half a year later, we found that blood TMAO levels were completely back to normal because the intestinal microbes had developed tolerance. So, even if, let us say, a fraction of 1% has a resistance to the antibiotic, in the beginning it looks as though the TMAO is 99% inhibited. But at 20 weeks it is back up to normal, because bacteria have numbers on their side. They have a doubling time in hours, and the very low-abundance resistant forms have a selective advantage because all the susceptible bacteria are being eliminated. None of the clinical trials of chronic antibiotic use in atherosclerosis have been successful to date.[14,15,16] My suspicion is that if the antibiotic had inhibited TMAO production in the beginning, I see no reason why it would do so after a continuous year of use of the antibiotic, as any resistant microbial form would eventually take over and repopulate the intestines.

Medscape: So, that is not something that physicians will be considering in the future, but do you think they can be doing anything else at this stage?

Dr. Hazen: Two things. One is that the test for TMAO will be available on a limited basis -- not widely available in every lab -- before the end of the year. Second, these studies basically reinforce a lot of existing knowledge in terms of diet. These kinds of studies do not address how much you eat and how much is linked to heart risk. That is done by epidemiology studies like the Health Professionals Follow-Up Study or the Nurses' Health Study, in which over 100,000 people have been followed for over 20 years with over 1 million patient-years of follow-up and thousands of mortality events. It is that kind of data, which involves very detailed food questionnaires given out every few years over the course of several decades, that allow us to conclude that eating 1 portion of red meat per day accounts for about a 13% increase in risk for total, cardiovascular, and cancer mortality over an average follow-up period of 22 years.[17]

From our studies we perhaps now understand at a mechanistic level why red meat is more associated with cardiovascular risk than would be predicted simply by its cholesterol or saturated fat content. If you look at the mortality data from large studies, and then at the cholesterol content of red meat and the saturated fat content, neither is high enough to account for the enhanced mortality risk,[18] and that is why people have argued that there may be another contributor, such as grilling meat, which produces high benzo[a]pyrene concentrations, or that people who eat steak also increase their sodium intake.

Medscape: There is also a carcinogenic effect.

Dr. Hazen: There is, and whether or not this pathway is linked is something that needs to be determined.

Medscape: Can you say how much people should cut back on foods containing L-carnitine or phosphatidylcholine?

Dr. Hazen: They should try cutting back on foods that are particularly rich in them, which goes almost hand in hand with cutting back on foods rich in cholesterol and saturated fat. That is because animal cells do not just have free-floating fat and cholesterol; they are bound in membranes inside the cells. The major building block of membranes is lecithin or phosphatidylcholine, so almost every type of food that is high in fat and cholesterol is also high in phosphatidylcholine, the main dietary source of choline. Free choline also exists, but usually it is in the form of phosphatidylcholine, unless you are following a vegetarian or vegan diet, and then you are probably getting more free choline. However, I want to be clear that I am hesitant to say that we are studying nutrition. We are just studying the fundamental biochemical pathways linked to nutrition, and it just happens that we have discovered a link between some of these compounds via the gut flora that make the metabolite. Essentially we first found the link between TMAO and atherosclerosis in subjects. We then reverse-engineered the process and determined that it comes from what we eat. It just so happens that those foods are high in saturated fat and cholesterol. So, our approach is not to say or not to make recommendations on different food choices; instead it is to determine what the pathway is. Personally, I like a good steak, and in the future I want to be able to have a tablet so that I can continue to eat a small portion occasionally. I do not think that a wholly vegan diet is the way to go, because that has other associated risks, such as being high in carbohydrate and low in vitamins B12 and D...

Medscape: Where do you see this research heading now?

Dr. Hazen: Our data are opening up 2 exciting clinical possibilities. One is the need to identify new pathways for cardiovascular risk, because even lowering the LDL cholesterol below 70 mg/dL is not sufficient. There is still significant residual risk that is not being treated with our current lifestyle changes and our current therapeutic approaches.

The other clinical possibility follows from the major genetic studies that have looked at attributable cardiovascular risk but have not even been able to attribute 10% of the risk. So, if it is not genes, it is environment, and our biggest environmental exposure is what we eat, which is a foreign body being ingested. Two different people can experience the same food differently because they have different gut flora. One person may generate a little more of a compound like TMAO than the other. That concept is a new way of thinking about complex diseases like atherosclerosis and other cardiometabolic diseases. It can also apply to obesity and insulin resistance. Data links intestinal flora involvement in those phenotypes in both mice and humans.[19,20,21]

There have been very exciting data from a gut flora transplant in persons with metabolic syndrome who received either their own fecal samples or those from a lean donor. Persistent changes in insulin sensitivity occurred just by transplanting the intestinal flora from one individual to another.[22] That is a whole new way of thinking, and it might lead to opportunities for new therapeutics.

I now think of statins, HMG-CoA reductase inhibitors, lowering LDL cholesterol, as our Homo sapiens enzyme inhibitors. I predict in our future that we will also have in our medicine cabinets drugs that target bacterial enzymes -- not antibiotics, but compounds that just inhibit the microbe enzyme activity so that they continue to live but without being able to generate the metabolite that we are trying to suppress. I think that is going to be a new and exciting kind of therapeutic for heart disease in the future.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.