Digestion Supplement Guide: What The Evidence Shows

The complaints are ordinary enough to sound trivial and are not: a stomach that feels tight by the evening, a bowel habit that has slowed, wind that arrives with no obvious trigger, a general sense of heaviness after meals that used to pass unnoticed. Together they account for most of what fills this shelf.

Ordinary is not the same as unimportant. The 2023 AGA clinical practice update on belching, bloating and distention exists because the same sensation has several different drivers, and the update is blunt about how often the wrong one is assumed. The joint AGA and ACG guideline on chronic idiopathic constipation does the same job for the slower end of the problem.

Neither of those is difficult reading and MedlinePlus covers the same ground in plain language for nothing. The gap a supplement fits into is narrower than the shelf implies: symptoms that have already been looked at once, nothing on a prescription pad yet, and an interest in supporting a normal function while the situation settles.

One thing worth knowing before reading any further: almost nothing on this shelf has been tested as a finished product. The trials below tested ingredients, singly, at amounts printed in their methods sections, and a capsule containing the same plant is not the thing that was studied.

Psyllium husk is the ingredient this category should be built on. It is the seed coat of Plantago ovata, it forms a gel in water, and it has more controlled human evidence behind it than anything else on the shelf.

The primary-care trial that anchors the modern literature randomised patients in general practice to psyllium, bran or placebo and ran for twelve weeks at ten grams a day. The 2022 meta-analysis of fibre for chronic constipation pooled the randomised evidence and found the benefit sits above ten grams a day sustained for four weeks or more. The ACG guideline on irritable bowel syndrome and a network meta-analysis of soluble fibre reach compatible conclusions.

Two details decide whether a label is borrowing that evidence honestly. The first is the amount, which is measured in grams rather than milligrams. The second is the form: McRorie’s review of stool output sets out why gel-forming psyllium and coarse insoluble bran behave differently in the bowel, which is why they cannot be swapped on a weight-for-weight basis.

Fibre also does things outside the bowel that get borrowed for packaging. A lipid meta-analysis found LDL cholesterol moved at a median of about 10.2 grams a day, a weight meta-analysis reported a mean of 10.8 grams a day over nearly five months, and a satiety study gave 6.8 grams before two meals. Every one of those figures is in grams, and a capsule that carries a fraction of a gram in total is not buying into any of them.

Cascara sagrada bark, Chinese rhubarb root, buckthorn bark and the latex layer of the aloe leaf share a chemistry. All four carry hydroxyanthracene derivatives, and those compounds act on the colon to produce a bowel movement within hours. They are stimulant laxatives, and that is a description of how they work rather than a criticism of them.

The reason this matters on a supplement label is that it is rarely said out loud. A blend can be described as gentle, natural and non-habit-forming while containing four ingredients whose entire pharmacology is cathartic. The LactMed entry on buckthorn uses the phrase potent cathartic properties, and the entry on cascara sagrada covers the same ground.

The safety literature is genuinely contested and a reader deserves both halves of it. EFSA concluded in 2018 that hydroxyanthracene derivatives should be treated as genotoxic and carcinogenic absent data to the contrary, and could not set a safe daily intake. Its 2024 opinion reviewed further data on rhubarb, cascara and frangula by name and did not reverse that position.

Against that, Whorwell’s 2024 critical review argues that much of the harm literature is confounded, and Rao and Brenner’s systematic review of over-the-counter therapies graded senna, a stimulant of the same family, grade A for chronic constipation while grading psyllium B. A reader who takes only the first of those away has been misled as surely as one who takes only the second.

Two practical points survive the argument. Long use of anthraquinone herbs is associated with melanosis coli, a harmless pigmentation of the colon lining that reverses on stopping, and a worldwide meta-analysis examined whether it tracks with colorectal neoplasia. And the content of these compounds in finished products varies enormously: a 2025 survey of the Italian market measured them across botanical supplements and found the spread wide.

A probiotic on a supplement panel is the one ingredient where a milligram figure tells you almost nothing. What matters is the number of live organisms, written as colony forming units, and the strain they belong to.

The Lactobacillus acidophilus literature makes the point cleanly. A two-strain placebo-controlled trial in irritable bowel syndrome ran at one times ten to the tenth colony forming units a day for eight weeks, the earlier LAPIBSS trial used the same order of magnitude, and the DDS-1 trial ran a named strain pairing at the same daily count for six weeks.

Pooled, the picture is modest rather than dramatic. A 2026 systematic review with trial sequential analysis found the evidence base still short of the information size that would settle the question, which is a more useful summary than either enthusiasm or dismissal.

There is also a labelling problem worth carrying to the shelf. A shotgun metagenomics investigation of fifty widely sold American probiotics found two thirds free of labelling inaccuracies, which is the charitable way to describe what it found in the other third. A panel that names a species, gives no strain and prints no colony count has told you the least it could and still be truthful.

Two ingredients in this aisle are routinely credited with research that belongs to a different part of the same plant, and both substitutions are easy to miss because the species name on the panel is correct.

Aloe is the clearer case. The inner leaf gel and the latex layer just under the rind are different materials with different chemistry. The ulcerative colitis trial used oral aloe vera gel at 100 millilitres twice daily for four weeks. The irritable bowel meta-analysis and the AVH200 pilot study also worked with gel preparations. The latex is where the anthraquinones sit, which is why a panel reading aloe leaf is making a laxative statement and a panel reading aloe gel is not. NCCIH’s aloe page keeps the two apart.

Oat is the quieter case. The cholesterol meta-analysis everyone cites is about oat beta-glucan at three grams a day or more, and beta-glucan comes from the grain, the bran and the oatmeal. Oat straw is the stem left after threshing, and a 2024 chemical study of it describes a lipophilic profile that has nothing to do with the soluble fibre in porridge. The trials on oat herb extract went in a different direction again, towards cognition rather than digestion.

The reading rule is simple enough to use in a shop. Find the plant part on the panel, then check which part the study you were told about used. Where they differ, the research does not transfer, however correct the Latin binomial is.

Some ingredients on this shelf arrive with a long history and almost no human trial data, and a guide that pretends otherwise is doing the same thing as the packaging.

Bitters are the honest version of that category. Gentian root has been used to stimulate appetite and digestion for centuries, a chemical profile of four Siberian gentians measured their iridoid and flavonoid content and a gastric stimulatory effect in laboratory work, and a review arguing for a new paradigm in how bitters are studied sets out how thin the human evidence remains. A label may reasonably call gentian a traditional bitter. It may not claim a digestive outcome for it.

Clay binders are the harder case, because the marketing language around them is confident and the literature is not. Bentonite is an aluminium phyllosilicate with a large surface area and a real capacity to adsorb, and searching for a controlled human trial of it for bloating or constipation returns nothing usable. What does exist is a case report of severe hypokalaemia after oral and rectal bentonite in a child, at quantities far above anything a capsule holds, which is a caution rather than a dose.

Alfalfa belongs in the same column. The LactMed entry is the most usable source on it and it is entirely about cautions: vitamin K content that matters alongside warfarin, and L-canavanine, which a 2025 scoping review of immunostimulatory herbs discusses in connection with autoimmune flares.

These three words carry a great deal of packaging in this aisle and almost no mechanism. They are not regulated terms, they describe no measurable process, and the nearest thing to a controlled test of the idea is worth knowing about.

A trial of a purported detoxification supplement gave healthy adult women a 1,350 milligram proprietary blend built on cascara sagrada bark and senna leaf, and measured body composition, waist circumference, blood markers and gastrointestinal symptoms. None of them moved. That is one trial of one blend, and it is more evidence than the vocabulary has otherwise generated.

The scale of the vocabulary is itself measurable. The STRIPED dietary supplement label explorer was built specifically to identify products sold with weight-loss, muscle-building and cleanse or detox claims, because the category is large enough to need a tool.

What a buyer can do with this is straightforward. Treat detox as a marketing word, look past it to the panel, and ask what each named ingredient does and at what amount. NCCIH’s guidance on using supplements wisely is a reasonable starting point for that habit.

Every product on this shelf is a dietary supplement, and United States law lets a supplement say it supports a normal body function while forbidding it from saying it treats, prevents or cures anything. That is why the front of every box reads the same and why the sentences sound deliberately unambitious.

The constraint is a feature for a reader. A box that promises to cure constipation is either breaking the rule or is a medicine, and either way something has gone wrong. A box that says it supports healthy digestion has told you it is a supplement and left the evidence question open, which is the honest position.

What that ceiling does not excuse is a panel nobody can check. Named ingredients with printed amounts, plant parts stated, colony counts given where a live culture is claimed: those are the disclosures that let a reader do the comparison themselves. An NIH perspective on proprietary blends sets out what is lost when a label prints one figure for everything inside it.

That is the whole point of the companion page to this one. The scorecard turns the paragraphs above into seven questions you can answer standing in front of a shelf, and then runs them on the product this website stocks.

• Whether a given finished formula works. Almost nothing on this shelf has been trialled as a finished product, and a blend of ingredients with individual evidence is not an ingredient with evidence.
• Whether the amount in the capsule is the amount on the panel. That is a laboratory question, and a printed figure is a declaration made at manufacture.
• Whether a fast result is the result you wanted. A stimulant laxative produces a bowel movement; that is not the same as an underlying problem having improved.
• Whether an ingredient suits you personally. Existing conditions and prescriptions are a conversation with whoever prescribes for you, and goldenseal, alfalfa and the anthranoids each have a specific one attached.
• Whether the price is fair. Cost per day is separate arithmetic, and a cheap capsule of the wrong plant part is not a bargain at any price.
• Whether your symptoms need something else entirely. That is the question worth answering before any of the other five.

None of the above is an argument that a digestion supplement is pointless. It is an argument about the order things happen in and about reading the panel rather than the front of the box. The constipation guideline and the MedlinePlus page between them cover what else is on the table, and both are free.