TMG Supplement: Homocysteine Fell 20%, LDL Rose 0.36 mmol/L

TMG has had a remarkable second life. For decades it was betaine, a compound isolated from sugar beets and prescribed at gram doses for a rare inherited disorder of methionine metabolism. Now it sells as trimethylglycine, a “methylation support” and longevity supplement, on the strength of one genuinely well-established effect and a great deal of extrapolation from it.

The established effect is that it lowers homocysteine. That is true, it is dose-dependent, and it has been replicated. What the label does not say is that the same laboratory that produced the best homocysteine data went on to publish what happened to blood lipids in the same people, and the answer was that LDL cholesterol went up, measurably, within two weeks.

Both findings come from the same trials. You cannot have one without the other.

The homocysteine result, with the numbers

The cleanest dose-finding trial was run at Wageningen and published in the Journal of Nutrition in 2003. Four groups of 19 healthy men and women took placebo or 1.5, 3 or 6 g of betaine a day for six weeks. The group also ran methionine-loading tests, which measure how well the body clears a surge of homocysteine.

After six weeks, fasting homocysteine was 12%, 15% and 20% lower than placebo at the three doses. The post-methionine rise was 23%, 30% and 40% smaller. The authors noted that ordinary dietary intake of betaine is about 0.5 to 2 g a day, so the lower doses were within what a diet rich in beets, spinach and whole grains could supply.

A 2013 meta-analysis in the Journal of Chiropractic Medicine pooled five placebo-controlled trials that had used at least 4 g a day for 6 to 24 weeks and found a combined reduction of 1.23 micromol/L. For scale, the obese adults in the Finnish trial started at 8.76 micromol/L, so that is a real shift, not a rounding error.

This is the part of the TMG story that survives scrutiny. It is the only part.

The lipid result, from the same group

Two years after the dose-finding trial, the same authors published a paper in PLoS Medicine that almost nobody selling TMG quotes. Its background section explains why they wrote it: earlier studies in kidney patients and in obese people on weight-loss diets had suggested betaine raises cholesterol, and nobody had checked in healthy people.

So they went back to their own trials, three placebo-controlled betaine studies with 151 healthy participants between them, and pooled the lipid measurements. The result:

Betaine supplementation (6 g/d) for 6 wk increased blood LDL cholesterol concentrations by 0.36 mmol/l (95% confidence interval: 0.25-0.46), and triacylglycerol concentrations by 0.14 mmol/l (0.04-0.23) relative to placebo.

In the units on an American lab report, 0.36 mmol/L of LDL is about 14 mg/dL. The total-to-HDL ratio rose by 0.23. HDL did not change. Lower doses also pushed LDL up, though not to statistical significance, and the effect was already evident after 2 wk of intervention.

The earlier Finnish trial they were following up had found the same direction. Schwab and colleagues gave 42 obese adults on a reducing diet 6 g of betaine or placebo for 12 weeks; homocysteine fell, body composition did not change any more than diet alone, and total and LDL cholesterol were higher in the betaine group.

The Wageningen authors did not soften their conclusion:

The adverse effects on blood lipids may undo the potential benefits for cardiovascular health of betaine supplementation through homocysteine lowering.

And they named the alternative. Folic acid at 0.8 mg a day had lowered homocysteine in their trials with no effect on lipids, and therefore remains the preferred treatment for lowering of blood homocysteine concentrations.

One note of disclosure that the paper itself makes: the Wageningen Centre for Food Sciences that funded the work was an alliance of industrial partners including Unilever, DSM and the sugar cooperative Cosun. That is a reason to read the methods, which are sound, not a reason to dismiss a finding that runs against the interests of a sugar-beet processor.

The step nobody has taken: does any of it prevent anything?

Lowering homocysteine was supposed to prevent heart attacks. That hypothesis has been tested, hard, though not with betaine. The 2017 Cochrane review of homocysteine-lowering interventions collected 15 randomized trials of B6, folate and B12 in 71,422 people, followed for one to 7.3 years, and found:

• no difference in myocardial infarction compared with placebo, high-quality evidence;
• no difference in death from any cause, 11.7% versus 12.3%, relative risk 1.01, high-quality evidence;
• a reduction in stroke, 4.3% versus 5.1%, relative risk 0.90 (95% CI 0.82 to 0.99).

So with the agents that have been tested at scale, the marker moves and the heart attacks do not. A modest stroke benefit is the one signal left standing, and it was produced by B vitamins, not by betaine.

No betaine trial has ever measured a cardiovascular event. The chain of reasoning behind “TMG for heart health” therefore runs: betaine lowers homocysteine (true), homocysteine lowering prevents heart attacks (tested and not supported), and betaine’s own effect on LDL does not matter (tested and contradicted). Our article on B12 dosing covers the more direct question of what the B vitamins do and do not achieve.

The gym claims, at their actual sample size

The other market for TMG is strength and physique, and the trial that anchors it is small. Cholewa and colleagues randomized 23 experienced lifters to 2.5 g of betaine or placebo alongside six weeks of periodized training: 11 people in the betaine group, 12 on placebo.

The betaine group improved on arm cross-sectional area and on the authors’ skinfold-based body-composition estimates, and increased bench press training volume at two of the three micro-cycles. But the strength outcomes, bench press and back squat one-repetition maximum, and vertical jump, were not different between groups, and the effect on power was a trend at p = 0.07. Set against the 12-week Finnish trial at more than twice the dose, which found no body-composition effect at all, the physique evidence is best described as unresolved.

Note that 2.5 g is also below the dose at which the LDL effect became significant, which is not the same as saying it is below the dose at which it exists. The pooled analysis found LDL rising at lower doses too, just not to statistical significance in a sample of 151.

What the capsule dose actually is

Every number on this page was produced at 1.5 to 6 g a day. Typical supplement capsules contain 500 mg to 1 g. A 500 mg capsule is a third of a normal day’s dietary betaine and one-twelfth of the top trial dose. If the trials are the reason you are buying, a capsule reproduces none of them; a powder measured with a scoop is the only form that matches any published protocol.

The molecule matters too. TMG is betaine anhydrous. Betaine hydrochloride, sold for stomach acid, is a different product with different behavior, and the HCl on the label means it is not what any of these trials used. And if the glycine part of the name is what caught your eye, our glycine and sleep article covers a different molecule with a different, and equally small, evidence base.

What follows from all this

If you take TMG for its homocysteine or methylation story, the evidence points to three things, and none of them is a bigger bottle. Take a dose you can measure in grams, because the capsules do not reach any trial. Take a lipid panel before you start and another at six weeks, because the trials say that is where the cost appears, and it appeared by week two. And know that the authors who did the best work on this compound recommended a different one, folic acid, for the job most people are buying TMG to do.