Choline and TMAO: Eggs Still Don't Cause Heart Disease | Chris Kresser

Choline and TMAO: Eggs Still Don’t Cause Heart Disease

by

Last updated on

Could the choline in these eggs contribute to heart disease?
Do choline-rich eggs increase TMAO levels? Stockbyte/Stockbyte/Thinkstock

A 2013 study by published in the New England Journal of Medicine (NEJM) has proposed a new link between eggs and coronary heart disease (CHD) that doesn’t involve cholesterol. A team of researchers, led by Dr. Stanley Hazen, showed that dietary choline—a nutrient found largely in eggs, beef liver, and other animal foods—is metabolized by bacteria in our gut and then converted by the liver into TMAO.

They demonstrated this with a “choline challenge:” feeding volunteers two large hard-boiled eggs (with approximately 250 mg of choline each) along with 250 mg of supplemental choline that was tagged with a heavy isotope. The isotope acts like a chemical “label” that allowed the researchers to track what happened to the choline after it was ingested. Their data did indeed show an increase in both labeled TMAO and total TMAO (in urine and blood) in the volunteers after they consumed the eggs and supplemental choline.

In a second study, Dr. Hazen’s group showed that increased levels of TMAO in the blood are associated with cardiovascular disease (CVD). The researchers followed roughly 4,000 adults for three years. At the end of the study period, those with the highest levels of TMAO had a 2.5-fold increased risk of heart attack, stroke and death.

On the surface this sounds like very bad news for omnivores. But let’s take a closer look at the studies to see if it’s really time to swap your morning eggs for a tofu scramble.

Do Choline-Rich Foods Increase TMAO Levels?

Dr. Hazen’s team did show a temporary increase in total TMAO after eating eggs. However, as Dr. Chris Masterjohn pointed out to me in an email dialog, the researchers’ own data show that there’s no way that the “choline challenge” could have contributed to this increase in total TMAO. If it had, we would expect to see an initial increase in labeled TMAO followed by an increase in labeled TMAO. This would indicate that the labeled choline supplement (that participants ate with the eggs) had been metabolized by the gut bacteria and then converted into TMAO in the liver.

But that’s not what happened. I re-created Figure 1C and 1D from the study. Figure 1C (below) shows an increase in total serum TMAO at one hour after the choline challenge. But by hour four, total TMAO is back to baseline and by hour 8 it’s even below baseline (i.e. the participants had lower TMAO at 8 hours than they did before they ate the eggs/choline).

Figure 1C choline and eggs Chris Kresser

However, Figure 1D (below) shows that labeled TMAO did not increase at all until hour four, and it didn’t increase significantly until hour six! This shows that the eggs and supplemental choline the participants ate had nothing to do with the increase in total TMAO that occurred one hour after the challenge.

figure 1D choline and eggs Chris Kresser

What’s more, the researchers didn’t mention that other commonly eaten foods have a much more significant impact on TMAO than eggs. A 1999 study tested the effects of 46 different foods on the urinary excretion of TMAO in 6 human volunteers. (1) Eggs had no effect on TMAO excretion compared to a light control breakfast, yet 19 out of 21 types of seafood tested did. In fact, halibut generated over 53 times as much TMAO as eggs! This is not surprising, because although all species of seafood contain lower amounts of choline than eggs, they do contain trimethylamine and TMAO. Dr. Hazen’s team was aware of this study, because they referenced it briefly in the discussion section of the NEJM paper. They acknowledged that “TMAO has been identified in fish” and “the ingestion of fish raises urinary TMAO levels.” But remarkably, they did not explain how much greater fish’s impact on TMAO was when compared to eggs.

Finally, this paper did not prove that eating choline-rich foods (or any other foods) increases TMAO levels over time. In fact, the researchers themselves seem to suggest this is unlikely in the discussion section of the paper.

They said: “the high correlation between urine and plasma levels of TMAO argues for effective urinary clearance of TMAO.” In other words, even if eating food does increase total TMAO levels, most people are able to quickly and efficiently clear that TMAO from their blood by excreting it in the urine. This makes it doubtful that dietary factors alone explain chronic elevations in TMAO.

Instead, there are several other factors that are more likely to explain such an increase, including:

  • Impaired urinary clearance of TMAO due to impaired kidney function: This is at least partially supported by data in the NEJM paper. Those with the highest levels of TMAO had an average glomerular filtration rate (GFR) of 69 mL/min. According to National Kidney Foundation guidelines, a GFR between 60–89 ml/min is indicative of a reduced capacity to filter blood through the kidneys. (2)
  • Differences in the gut microbiota that predispose toward increase TMAO production: Previous work by Dr. Hazen’s group has shown that people with higher levels of Prevotella bacteria in their gut produce higher levels of TMAO. (3) (Interestingly enough, other research has shown that consumption of whole grains—not animal products—is associated with higher levels of Prevotella bacteria.) (4)
  • Enhanced conversion of trimethylamine to TMAO in the liver: An enzyme called Fmo3 carries out this conversion, and its activity is affected by genetic factors, iron or salt overload, and a number of common pharmaceutical drugs used to treat arthritis, GERD, and infections. (5)
  • Diabetes and metabolic syndrome: Fmo3 activity is upregulated in cases of insulin resistance and insulin deficiency. (6)

If food really did make a significant contribution to TMAO levels, and high TMAO levels cause heart disease, then we’d expect to see much higher rates of CHD among people who eat more fish—since fish has a much greater effect on TMAO than eggs. Yet this is the opposite of what studies indicate: Eating more fish (especially cold-water, fatty fish) has consistently been shown in both observational and randomized controlled trials to reduce the risk of death from heart disease. (78)

Do Choline-Rich Food Cause Heart Disease?

At the end of their paper, Dr. Hazen’s group cautions against “excessive consumption of dietary phosphatidylcholine and choline” and recommends a high-fiber, vegetarian diet as a means of protecting your heart health.

Yet as I’ve argued above, they failed to present convincing evidence that eating eggs significantly increases TMAO over time—especially when compared to other foods like fish. Moreover, if eating choline-rich foods did increase the risk of heart disease (via TMAO or any other mechanism), we’d expect to see higher rates of CHD in those that eat more eggs. Yet numerous studies have failed to find any such association. For example, a meta-analysis of prospective studies involving a total of 474,000 participants followed from 8 to 22 years published in the British Medical Journal found no association between higher egg consumption (up to one per day) and CHD or stroke. (9) An analysis of data from the National Health and Nutrition Examination Study found an inverse association between egg consumption and stroke, and a cohort study from Japan found that consumption of animal products including eggs was associated with reduced risk of death from stroke. (1011) The lack of association—or inverse association—between egg consumption and CHD is even more impressive when you consider that those who eat more eggs are also more likely to smoke and be physically inactive. (12)

Some studies suggest that eggs may even prevent heart disease. Egg consumption leads to the formation of larger, less dense LDL and HDL particles, which may be protective against atherosclerosis. (13) Eating eggs frequently may even lead to lower cholesterol; one study found that those eating four or more eggs per week had lower total serum cholesterol than those eating one or fewer per week.  (14)

This same study found that egg consumers had diets higher in nutrients that have been shown to reduce the risk of cardiovascular disease compared to non consumers, including vitamins E, B12, and folate.

Finally, as I pointed out above, some research suggests that consuming large amounts of whole grain increase Prevotella bacteria in the gut, which were associated with the highest levels of TMAO in Dr. Hazen’s previous study on TMAO. If this is the case, consuming large amounts of fiber from whole grains may actually increase the risk of heart disease.

The hypothesis that increased serum TMAO is associated with heart disease is interesting and should be investigated further. But the data presented by Dr. Hazen’s group doesn’t support the conclusion that dietary choline is a major cause of increased TMAO, nor does it support their advice to avoid choline-rich foods like eggs, liver, beef, and pork.

Now, I’d like to hear from you. Do you eat eggs (or another source of dietary choline) regularly? What are your thoughts on this research? Comment below and let me know.  

  1. ” . . . some research suggests that consuming large amounts of whole grain increase Prevotella bacteria in the gut, which were associated with the highest levels of TMAO in Dr. Hazen’s previous study on TMAO. If this is the case, consuming large amounts of fiber from whole grains may actually increase the risk of heart disease.”

    Whole grains. This is the culprit. I’ve been off whole grains about a month now. I feel so much better!

    I’m glad I read this article because I had also cut down on eggs and meats, and then I’m finding that TMAO and Lecithin are in just about everything.

    • Prevotella has recently been shown to also be high in the Hadza hunter gatherers so it is not all about grains.

  2. That’s right. I’m guessing he means unlabeled for the second one. So in those charts don’t make sense. The second one shows that it takes 4 to 8 hours for the choline supplement to show up as TMAO in the blood but then why does the total not go up in the first chart? And what is the spike attributed to? The y axis is different for the charts, what are the units?

  3. This is a very interesting study. I think researchers still need to figure out how TMAO is linked to heart disease. Some researchers are recommending to completely remove egg and meat from our diet.

  4. Chris and fellow knowledgable posters,
    I’m avoiding eggs for the next 25 days and counting as part of an autoimmune protocol test. I’m also 4.5 months pregnant and know I need the nutrients provided by eggs. Can you offer some advice on how best to substitute/replace egg nutrients such as choline, cholesterol and fat-soluble vitamins in the right proportions? Thanks!

    • The egg white is the potential allergen and there may be no reason to avoid the yolk (which has the nutrients you want) if you can separate it cleanly. I’d look online for tips about ding that. Failing that, chicken livers are probably an OK substitute, chop them up finely and fry them in butter.

  5. Egg got the heart disease reputation from the ’70s and since then the reactions have been varied.I’m of the view that it is not bad at all,if consumed in moderation.Those arguing that the study focused on lecithin and not the egg and saying it’s sold in nutritional components don’t get the point.Why then should we eat while we can just get supplements of the fat list of thiamin,choline,niacin,betaine and the never-ending list of vits,minerals and what-not-all present in a single egg?

    I’ll go with the egg on any given day.And the almost half a million participants followed for 16 years in the meta-analysis of studies that found no link between eggs and CHD PLUS stroke is a number I can take heart from.

    • The point about lecithin is this.

      Some people buy lecithin as a supplement NOT to supplant. Therefore the question remains:

      Is consuming EXTRA lecithin (i.e. as a supplement) unwise, advantageous or ‘the jury is still out’

  6. “we would expect to see an initial increase in labeled TMAO followed by an increase in labeled TMAO. ”
    This does not make sense, can you explain? Is there a typo?
    Also the units for TMAO are not depicted and there is a very large difference in the peak TMAO levels obtained in the two graphs. The first graph indicates a peak = 8 at hour one, the second graph shows a max = 300 at hour eight. The level of 8 (at one hour) pales in comparison to the level of 300 at hour 8. I would expect that GI transit, intestinal bacteria processing, absorption and hepatic metabolism to take several hours, which is when the TMAO levels reached the highest.

  7. Yo! Chris
    Where does this put L-Carnitine supplementation in the TMAO theory.

    Also (OT) sorry..a slight dilemma.
    If you have a Barrett’s and are prescribed a PPI and wish to get the gut acid friendly with Betaine HCL
    supplementation…can this be accomplished or is it contra-indicated?

    Thanks & regards

  8. Dear Dr. Chris,

    Im glad to hear I can continue to eat eggs. I just got hooked on mixing them with fermented sauerkraut. Any recommendation on Udo’s Oil? 2-1 Omega-6/omega-3 ratio, although not ideal (not 1-1), I tend to use it for fueling on long runs and it seems to help.

    RIch

  9. This part doesn’t make sense to me. Can you rephrase it?

    “we would expect to see an initial increase in labeled TMAO followed by an increase in labeled TMAO”

  10. If I wanted to supplement phosphatidylcholine I’d buy lecithin.
    What’s that made from? Soya beans.
    What’s it used for? Emulsifying chocolate and other processed foods.
    Another reason why soy is bad for you – TMAO.
    LMFTMAO.

  11. Dear Dr. Chris,

    I’m starting a blog/website for nutrition supplements, how can I use the wordage in your articles? Is it legal?

    Scott

    • You can quote passages from the article with proper attribution (and preferably a link back to the article), but reposting an entire article without permission is a violation of copyright—and just not very nice.

  12. what makes me laugh in all this is the fact choline is a required by our bodies in order to live. choline is found in cell membranes i.e phosphatidylcholine also in important in liver and brain health. In fact phosphatidylcholine plays a huge role in what is allowed going in and out of the cell membrane

    • This is true, but the fact a nutrient is required for health doesn’t mean that too much of it can’t cause problems. This is really the rule rather than the exception, with everything from vitamins to minerals to water.

      The question is whether choline-rich foods like eggs and liver contribute to an increase in total serum TMAO. This study does not support that hypothesis.

      • Their contention that TMAO causes atherosclerosis depends at one stage on it elevating LDL and depressing HDL. All trans fats do this, including ruminant trans fats. Yet observational studies consistently fail to find a link between consumption of ruminant trans fats and CVD.
        Ergo, inhibition of reverse cholesterol transport is not sufficient of itself to produce the effects attributed to it in these papers. Either that or ruminant fats contain some magic ingredient that counters the harmful offect of the trans fats. I wonder what that is – palmitic acid? Stearic acid? Cholesterol?

  13. Chris, if you think your line of reasoning is sufficiently solid could you, maybe together with Chris Masterjohn, write a comment about this article to the New England Journal of Medicine? That would force Stanley Hazen to reply. I am curious what they would say.

  14. Thank u for this. I was starting to look at my morning egg yolk as a yellow version of Darth Vader.

  15. The study was less about eggs and more about lecithin, which is in eggs. Why did your analysis not deal with lecithin, which, incidentally sold as a supplement? Do we avoid lecithin OR take it as a supplement I. E. welcome it?

  16. Chris,
    I have read information stating that when you cook an egg yolk you oxidize the good fat in the yolk, which in turn increases the bad cholesterol. Any truth to this or is ok to eat poached, hard boiled and a variety of eggs?

    • That “info” on cooking eggs is sometimes applied to any method that breaks the yolk or mixes the white and yolk, as if oxidation occurs instantaneously. If smashing the yolk is so bad, how come all the wild mammals who smash eggs and readily eat them are not suffering from “bad cholesterol”? Unless you get the eggs from your home flock, or from someone you know, I would think any problems from cooking are outweighed by the potential for bacterial contamination if you consume them raw.

    • Christine,
      If you’d like a good primer on cholesterol, search “the straight dope on cholesterol”; it’ll be on Dr. Peter Attia’s blog, The Eating Academy…I do believe that he and Chris link to each other. He particularly debunks the myth that dietary cholesterol has anything to do with LDL et al; actually, according to the data he references, most of the “cholesterol” that we consume is in a form that our body doesn’t even bother to absorb, as the liver already produces a much more readily re-absorbable form of cholesterol.

  17. In reading Dr. Hazen’s article it almost seems like he had a pre-conceived idea and set about to prove it regardless of what else he found. How much money did he get for this research?

    • It would be interesting to find out who funded Dr. Hazen’s study. The funding source might shed some light on the apparent bias.

    • Indeed it does Elizabeth. He has managed to get his name in the news, which seems to be a major goal for researchers going for government grants and the like. As anything else in life, one must ask who benefits from his headline grabbing ‘research’?

      Thank you Chris for writing this article. R

    • Hazen and the others are listed as the inventor(s) of a 2009 patent application for what seems to be a standard test protocol for plasma TMAO (inter alia) .

      Patent at http://www.google.com/patents/US20120157397

      So presumably if their conclusions are accepted we’ll all be trotting off for another test of questionable value to go with the current lipid panel – which I guess will work out pretty well for them in terms of royalties.

      I guess it could be co-incidence that they developed the test four years ago and now their research (which has been heavily marketed in the public domain) concludes that we all need a TMAO test. Or possibly not.

  18. I agree with Mr. Shield that this borders on fraud, though it is more likely the be the result of the garbage that passes for “scientific” reporting for non-scientists; too many of these writers are simply not equipped for the task of understanding the study and reporting accurately. However, I am not going to rule out the existence of scientists who have an agenda, which this one gave away with his recommendation that people scared by his study eat a high fiber (meaning lots of whole grains) vegetarian diet. That is a pretty remarkable leap in logic from a fairly limited study, and especially one that has yet to be replicated by other scientists and discussed more fully within the field. So, same old, same old from the militant, everyone-must-become-vegetarian-or-vegan types.

  19. This all borders on fraud to me. I mean, they’re making headlines about beef and eggs, yet as you and others noted, some veggies are just as “bad” w/TMAO and seafood much worse…. If I ever have a heart attack it will because of reading nonsense rather than eating the eggs.

    • The fraud to which you refer unfortunately is a common occurrence in scientific research, especially in the medical field. The poor design and data cherry-picking in so many of these studies likely explain why there has been so much conflicting advice on nutrition by so-called experts in the last few decades. For insights on the problems associated with so much of this research, check out Ben Goldacre’s books, “Bad Science” and “Bad Pharma”. Dr. Goldacre is a British doc whose exposes will make your blood pressure rise (unlike red meat and eggs). What’s especially shocking is that no one seems to do anything about the fraud and the perversion of the scientific method.

    • This quote:

      “Finally, as I pointed out above, some research suggests that consuming large amounts of whole grain increase Prevotella bacteria in the gut, which were associated with the highest levels of TMAO”

      This could indicate that Dr Hazen’s group have used a group of test persons that are vegan or vegetarian that due to their diet have their guts allready full of Prevotella bacterias.

    • Maybe the impact of plant choline in the gut is different than that when consuming red meat or even eggs.

      There is still too much unknown about all this to be making decisions confidently.

    • Speaking of fraud, half as much effort on the Diet-Heart Study would result in lack of proof that saturated fats do not impact CHD.

[if lte IE 8]
[if lte IE 8]