TMAO is thought to be atherogenic (plaque causing) because it is associated with cardiovascular disease (i.e., people who get heart attacks tend to have more TMAO in their blood). This study itself established a correlation (n=2595 cohort) between cardiovascular disease and greater blood concentrations of TMAO.
TMAO's molecular structure is common to both carnitine and choline, and thus it is thought they can bioconvert in the body. TMAO is thought to contribute to atherosclerosis by promoting cholesterol influx into macrophages, which then promotes foam cell formation (and then the foam cells themselves become plaque); this was confirmed in ApoE-/- mice (genetically modified mice used for cholesterol research purposes) in this study, and appeared to influence macrophages derived from wild-type mice as well. This mechanism is plausible.
The study above noted that following consumption of l-carnitine (180mg via sirloin and 250mg via radiolabelled capsules), TMAO was confirmed in both the blood and urine. The role of intestinal microflora was confirmed when participants were put on antibiotics (to suppress the microflora), which abolished diet-induced TMAO increases. Furthermore, this increase was detected in omnivores but not long term (more than 5 years) vegetarians and vegans, and seemed to be related to a higher gut population of Prevotella bacteria.
Finally, the researchers conducted a study in APoE-/- mice. The mice fed carnitine had roughly double the plaque buildup vs mice that were not fed carnitine. This was again abolished with antibiotic treatment (to destroy the gut:TMAO link). It should be noted that APoE-/- mice are genetically modified to exacerbate any bad cholesterol effects (for research purposes).
Note: The previous version of this article made note of '1.3% carnitine in the diet' which was a misreading from the paper. Carnitine was supplemented at 1.3% of the drinking water. Quantification of carnitine ingested is difficult, but assuming an intake of 5mL of water daily (somewhat in the ballpark of mouse studies) this would translate to 65mg carnitine in mice (and assuming 25g in weight, which is in the ballpark for a 10 week old female mouse, 2,600mg/kg) and a preliminary human estimate of 208mg/kg. This oral dose is not practically achieved via supplementation or diet despite what the first draft stated; we apologize for any confusion
Overall, what can we conclude from this study?
There is a correlation between higher TMAO in serum and cardiovascular disease risk in humans.
There appears to be microbial fermentation of dietary carnitine into TMAO in non-vegetarian humans.
TMAO in serum from dietary carnitine causes atherosclerosis in APoE-/- mice.
What cannot be concluded from this study?
TMAO in humans causes heart disease. ApoE-/- mice are used since it is easy to research pro-atherogenic cholesterol metabolism, but these are mice with a specific genetic fault. This research would have to be replicated in regular mice (and then in humans). Additionally, the human correlational research does suggest there is a link but cannot state that it is a causative link, and past research on choline does suggest that TMAO is fairly rapidly excreted via the kidneys (suggesting some effects that are differential dependent on species).
Carnitine causes heart disease, as the link between TMAO and heart disease is not fully established in humans.
TMAO is bad. To nip this in the bud, most molecules have good and bad associated with them (good example: cortisol) and it would be improper to conclude that TMAO must be avoided at all cost. To reiterate as well, the correlation between TMAO and cardiovascular disease could very well be a biomarker of some other lifestyle habits.
What can plausibly be concluded, but requires future research to confirm?
Carnitine could contribute to heart disease via pro-inflammatory mechanisms, and this pro-atherogenic effect can likely be negated by other nutraceuticals (including statin drugs such as red yeast extract or other hypocholesterolemic compounds like Berberine).
Probiotic supplementation could modulate heart disease risk via a TMAO link, but this assumes that the TMAO link is actually relevant in humans.
The study found that in genetically modified mice, a high (but not impossible) dosage of l-carnitine did double plaque buildup. This may or may not be related to TMAO, we cannot say. This may or may not happen in humans, we cannot say. Overall? It's just preliminary research that should only interest other researchers, not the layperson.
At this time, restricting your carnitine consumption is not a prudent response for most people.
Want to learn more? See if red meat causes cancer.