TMG (Trimethylglycine / Betaine Anhydrous)

TMG and betaine anhydrous are the same molecule sold under two names. Trimethylglycine is the chemistry-class label, betaine anhydrous is the supplement-industry label, and both refer to glycine with three methyl groups attached. The compound has two genuinely separate use cases with very different evidence quality, and the honest review has to keep them apart.

The first use case is homocysteine lowering, and this is where TMG has its strongest data. Olthof's 2003 dose-response RCT (n=12 healthy adults, six weeks at 1.5g, 3g, and 6g/day) showed fasting homocysteine drops of about 12, 15, and 20 percent respectively versus placebo. McRae's 2013 meta-analysis pooled five RCTs and found betaine lowered plasma homocysteine by 1.23 micromol/L (95% CI -1.61 to -0.85, p=0.01), with the threshold dose around 4g/day for 6 weeks or longer. Mechanism is clear: betaine donates a methyl group to remethylate homocysteine to methionine via betaine-homocysteine methyltransferase, an alternate pathway to the folate-B12 route. Whether dropping homocysteine translates to lower cardiovascular events in healthy people is a separate, more contested question, but the biomarker move is real and replicated. In clinical homocystinuria (the inherited CBS deficiency that produces 10-20x normal homocysteine), betaine anhydrous is the active ingredient in the FDA-approved drug Cystadane, dosed at 6-9g/day with documented mortality and event-rate improvement.

The second use case is power output and muscle endurance in resistance-trained men, and this is more modest. Hoffman 2009 (n=24 active college males, 1.25g twice daily for 15 days) saw improved squat endurance and number of quality reps, with no upper-body effect. Lee 2010 (n=12 trained men, 1.25g twice daily for 14 days) reported improved bench-throw power and isometric bench press force in a crossover design. Cholewa 2013 (n=23 strength-trained men, 2.5g/day for 6 weeks) found body composition improvements (lean mass up, fat mass down) and a trend toward better bench press work capacity. The effects are small, the trials are small, and they cluster around the same research group with industry funding from Danisco/DuPont. Cholewa's 2018 follow-up in collegiate females was flatly negative on body composition and performance. The honest read: real signal in trained men at 2.5g/day, weaker or absent signal in women, and effect sizes too small to be transformative.

The third use case, methylation support for NMN/NR users, is biochemically logical but not trial-tested. NMN and NR raise NAD+, which is consumed by sirtuins and PARPs in reactions that produce nicotinamide, which is methylated by NNMT and excreted as methyl-nicotinamide. This pathway uses methyl groups from S-adenosylmethionine, and the SAM pool draws on methyl donors including betaine. The Huberman-popularized stack of NMN plus TMG rests on this logic. The biochemistry checks out, but there is no head-to-head RCT showing that TMG co-supplementation produces better cognitive, metabolic, or longevity outcomes in NMN/NR users versus NMN/NR alone. People taking it for this reason are betting on a plausible mechanism, not on trial evidence.

Cognitive and depression claims are even thinner. The methyl-donor logic extends to SAM, which is involved in neurotransmitter synthesis, and SAMe (the methylated form of methionine) has actual antidepressant data. But TMG is two enzymatic steps upstream of SAMe and there are no good cognitive or depression RCTs in healthy adults. Treat these claims as theoretical.

Practical bottom line: 1.5-3g/day if your goal is homocysteine lowering and methylation support. 2.5g/day taken before training if you are a resistance-trained man hoping for a small power-output edge. The same 1.5-3g/day if you are stacking NMN/NR and want to backstop your methyl pool on biochemical principle. Do not expect TMG to do anything dramatic. The molecule is cheap, well-tolerated, and earns a respectable but bounded place on the evidence ladder.