Metabolism Supplement Guide: What The Evidence Shows

Start with the two reviews that looked at the whole field rather than at one plant. A systematic review of dietary supplements for this purpose and a meta-analysis of isolated compounds both reached the same shape of conclusion: individual studies are small, many are short, and where a pooled effect survives it is modest and rarely the size the packaging implies.

That is not an argument that nothing works. It is an argument about calibration. A reader who expects a bottle to do what a change in diet does will be disappointed by every product in this category, including the ones with the better evidence behind them.

The second thing those reviews establish is that the category runs on ingredient names rather than on finished products. Almost no supplement sold in this space has been trialled as sold. What exists is evidence about the names on the label, at doses the label usually does not match, and the honest reading of any bottle starts by keeping those two things apart.

Apple cider vinegar is the one ingredient a buyer will meet on almost every bottle in this category, and it is also the one with the most human data behind it. A 2025 meta-analysis pooled the controlled trials and reported effects on body composition and on several blood measures, with the usual caveats about study size and duration.

The dose matters more here than the conclusion. Trials generally ran at around 30 mL a day, taken over periods up to twelve weeks. That is two tablespoons of liquid vinegar, daily, for three months. A dropper bottle holding 2 fl oz / 60 mL in total could not deliver that amount for a week, let alone twelve, which tells you immediately that a vinegar row on a dropper label is not the row the trials tested.

Johnston’s vinegar dosing work and an earlier mealtime study are worth reading for the same reason. Both are about timing relative to a meal, both used measured liquid volumes, and neither describes a few drops under the tongue.

Both names appear regularly on metabolism labels and both have real human trials. Neither set of trials measured what the packaging usually implies.

The Cochrane bitter melon review and a later meta-analysis both assessed glycaemic control in type 2 diabetes. The insulin-secretion trial ran at 2000 mg a day for three months, and a 2024 analysis added lipid outcomes. Blood glucose and lipids are the outcomes, and they are clinical outcomes in people who already have a diagnosis.

Wild yam is narrower still. The crossover trial looked at menopausal symptoms, a 12-month study of Dioscorea alata ran at 24 mg a day, and a review of herbal menopause preparations places it in that literature rather than in this one.

The lesson is transferable. An ingredient having trials is not the same as an ingredient having trials for the thing you are buying it for, and the abstract of the paper usually settles which in one sentence.

Raspberry ketone is the clearest example in this category of a name that travelled a long way on very little. Its reputation rests on a rodent study and on cell work: adipocyte lipolysis experiments and a follow-up in the same cell line.

There is one human study, and it is a trial of a multi-ingredient product taken alongside exercise and a diet programme. Nothing in it can be attributed to the ketone, because nothing in its design separates the ketone from the other ingredients or from the exercise.

A reader can apply this test to any name on any label in five minutes. Search the ingredient with the word randomised. If everything that comes back is in cells or in rodents, the mechanism may be real and the human effect is simply unknown. That is a defensible thing for a label to carry. It is not a defensible thing for a label to imply it has tested.

Gotu kola, horse chestnut, butcher’s broom and grape seed turn up together often enough to be recognisable as a group. They are the standard phlebotonic quartet, and their evidence is among the better evidence in botanical medicine. It is about leg swelling in chronic venous insufficiency.

The Cochrane phlebotonics review covers the class. The horse chestnut review is the strongest single result, at 600 mg a day of seed extract standardised to 100 mg of aescin, supported by transcapillary filtration work and a placebo-controlled oedema trial.

The butcher’s broom trial and the pooled analysis of the same preparation share a complication worth knowing: the product tested contained 150 mg of Ruscus extract with 150 mg of hesperidin methyl chalcone and 100 mg of vitamin C. It is a three-part capsule, so the result belongs to the combination rather than to the plant on its own.

Grape seed is the same story in a different outcome. A meta-analysis of 16 randomised trials and a dose-response analysis both report cardiovascular measures, chiefly blood pressure and endothelial function.

None of that is an argument against these four plants. It is an argument about what their presence on a label entitles a seller to say. A quartet with good evidence for leg swelling, printed on a bottle sold for something else, has borrowed a reputation rather than earned one, and the buyer is the person who should be told.

Chinese hawthorn appears on metabolism labels and its research is cardiac. The Cochrane hawthorn review and the WS 1442 dose-ranging trial assessed chronic heart failure; the trial found only the highest arm, 1800 mg a day, separated from placebo.

The part to carry away is the interaction rather than the efficacy. A controlled study examined hawthorn alongside digoxin, and separate work showed hawthorn can disturb the immunoassay used to measure digoxin. That second finding is the unusual one: it can affect the test as well as the patient, which makes it worth mentioning to a prescriber even when the dose is small.

A clinical review of hawthorn sets the rest of the safety picture out at length. On a label that prints no amount, an interaction caution is the one piece of information that still applies regardless of how much is in the bottle.

Everything above turns on a comparison between a dose on a label and a dose in a trial. A label that prints no amount removes the reader’s half of that comparison entirely, and no amount of research on the ingredient names puts it back.

The Dietary Supplement Ingredient Database work exists because a printed amount and a measured amount are two different numbers, and an analysis of labels in this category found the two parting company often enough to be a documented problem. A product that prints an amount can at least be checked by that method. A product that prints none cannot be checked by any method at all.

Work on chemical analysis in supplement liver-injury cases makes the same point from the other end: when something goes wrong, the investigation starts from what the label said was in the bottle.

So the first question to ask of any label in this category is not which ingredients are on it. It is whether an amount is printed against each one. Everything else follows from the answer, and the scorecard starts there for exactly that reason.