DMG (Dimethylglycine).

DMG is what betaine turns into after it hands off its methyl group, so the two are consecutive steps in the same pathway rather than two separate ingredients. Taking both raises the downstream metabolite while also supplying its precursor. A 2025 genome-wide analysis mapped genetic variation in circulating choline, betaine and DMG and in the ratios between them, which is where individual differences in this pathway show up. Those were circulating concentrations, which are markers rather than outcomes.

Choline sits two steps upstream of DMG through the choline to betaine to dimethylglycine sequence. Supplemental choline therefore raises endogenous DMG production without any DMG being taken. The 2025 study found genetic loci influencing where individuals sit along this sequence. Circulating concentrations were the measurement, so read them as pathway markers.

The enzyme that demethylates DMG to sarcosine carries FAD as a covalently bound cofactor, and FAD comes from dietary riboflavin. Poor riboflavin status slows that step and DMG accumulates upstream. This is a settled cofactor relationship, not an inferred one. It explains why DMG clearance is not uniform across people.

Demethylation of DMG to sarcosine, and sarcosine to glycine, hands methyl groups to tetrahydrofolate, which links the betaine pathway back into folate metabolism. The two routes for remethylating homocysteine, folate-dependent and betaine-dependent, share this junction. Folate status therefore affects how the DMG-derived methyl groups get used. This is textbook one-carbon chemistry.

Homocysteine has two remethylation routes: the B12- and folate-dependent methionine synthase, and the betaine-dependent BHMT that generates DMG. When B12 status is poor, flux shifts toward the betaine route and DMG output rises. The two pathways compensate for each other, which is why looking at one in isolation is misleading. This is pathway architecture, not a measured outcome.

DMG is N,N-dimethylglycine, and stripping both methyl groups yields glycine. Supplemental DMG therefore contributes to the glycine pool, though as a minor contribution compared with dietary glycine or glycine itself. The value of the intermediate is the methyl groups it carries, not the glycine at the end. Anyone taking DMG for glycine is taking the long way round.

Homocysteine can be remethylated, which is where DMG comes in, or committed to transsulfuration through the B6-dependent cystathionine beta-synthase. B6 status determines the split between the two fates. Supporting only the remethylation side leaves the disposal route unaddressed. This is standard one-carbon and transsulfuration biochemistry.

Sarcosine sits one demethylation step below DMG in the same catabolic sequence. Taking both means supplying an intermediate and its own product. Sarcosine is itself further demethylated to glycine by a folate-dependent enzyme. The relationship is sequential and well characterised.

The combination was tested in quail chicks using a response surface model, which explores dose combinations rather than testing a single fixed pairing. Whatever was observed there was in birds, so it grounds a formulation hypothesis and nothing about people. The three-way design also makes it hard to attribute any effect to the DMG component alone. Read it as animal work.

Tryptophan was one of three factors in the response surface design, so its contribution is entangled with the other two. The animals were quail chicks, which limits what can be carried across. Tryptophan is the precursor to serotonin and then melatonin, so including both in the same design creates its own overlap. Nothing here supports a claim in people.