PGX

Also sold as: PolyGlycopleX, Glucomannan Complex

A proprietary cross-linked complex of three viscous polysaccharides. Its claim is rheology, not binding: it thickens gut contents so glucose arrives more slowly — which makes it the one product in this category whose central claim an instrument can check.

How PGX actually works

Three viscous polysaccharides cross-linked in a proprietary process: konjac glucomannan, sodium alginate and xanthan gum. The claim is rheological rather than chemical — nothing is bound. High viscosity in the lumen slows gastric emptying, thickens the unstirred water layer that glucose has to diffuse across to reach SGLT1, and slows the mixing that brings amylase to starch. All three change the rate at which glucose appears in blood; none changes the total that eventually arrives. Downstream, konjac glucomannan is not absorbed and is heavily fermented, so short-chain fatty acids are a second candidate mechanism nobody has separated from the first.

Where to get PGX

The evidence for PGX

Graded by what exists behind each claim.

✅ Clinically validated

• Granular PGX at 2.5, 5.0 and 7.5 g with breakfast cut the incremental blood glucose area by up to 50%, linearly with dose.
• Against wheat dextrin — a soluble fiber with essentially no viscosity — 5 g of PGX roughly doubled post-meal fullness and significantly lowered the glucose area.
• In a 52-week randomized trial, PGX was 2.6 kg below control at 12 months; a psyllium comparator in the same trial was 2.4 kg below control at 6 months and no longer significant at 12.
• Capsules taken with the evening meal lowered the next morning's glycemia by up to 28%, while the same capsules taken 15–45 minutes before a meal did nothing — a capsule needs time to hydrate before it is viscous.

📊 Correlative data

• Konjac glucomannan is not absorbed intact and is significantly fermented by gut bacteria, so short-chain fatty acids are a second candidate mechanism that no PGX trial has separated from viscosity.

🧪 Theoretical / extrapolated benefits

• Whether the cross-linked complex beats plain konjac glucomannan — sold by the kilogram — at equal grams has never been tested head to head.

How to read these tiers: they say how much human evidence exists, not how well something works — and ✗ flags harm, never a disappointing trial. How the evidence tiers work →

What PGX actually does

This one does not claim to bind anything, and that is the most important sentence on the page. PolyGlycopleX is a processed complex of three polysaccharides -- konjac glucomannan, sodium alginate and xanthan gum -- and its claim is rheological. It is supposed to make the contents of the stomach and small intestine thicker, and thickness is a physical property with a unit, which makes this the one product in its category whose central claim can be checked with an instrument rather than argued about.

The three components are chemically distinct and each is viscous for its own reason. Konjac glucomannan is a beta-1,4 linked chain of glucose and mannose in roughly a 1 to 1.6 ratio, decorated with acetyl groups, and it reaches molecular weights in the high hundreds of thousands to millions of daltons -- which is why it is among the most viscous soluble fibers in food use. Sodium alginate is a block copolymer of beta-D-mannuronate and alpha-L-guluronate from brown algae -- the same family of additives the European food code numbers E 400 to E 404 EFSA ANS Panel 2017 -- and its guluronate blocks chelate calcium into egg-box junction zones, which is the standard polymer chemistry behind an alginate gelling in acid and behind an alginate raft antacid. Xanthan gum is a bacterial heteropolysaccharide with a cellulose backbone and trisaccharide side chains, and it is strongly shear-thinning: very viscous at rest, much less so while it is being stirred. The proprietary step is a processing reaction intended to make the three behave as one network whose viscosity exceeds the sum of the parts.

Scale is worth fixing before any of the trial numbers, because the regulatory file gives it. Konjac is permitted as a food additive at a maximum of 10 g per kg of food, and the calculated refined exposure across all population groups came out below 0.1 mg per kg of body weight per day -- roughly 7 mg a day for a 70 kg adult EFSA ANS Panel 2017. The year-long supplement trials dose 15 g a day of the blend Pal 2016. Those two numbers are three orders of magnitude apart, and the distance is worth holding on to, because the safety record of konjac as a thickener in a sauce is quoted constantly as though it were the safety record of konjac as a 15 g daily supplement. It is not the same exposure and it is not the same question.

What high viscosity does in a gut is three separate things, and none of them is blocking. First, it slows gastric emptying, so chyme is delivered to the duodenum over a longer period. Second, it thickens the unstirred water layer -- the near-stationary film between the bulk lumen and the brush border -- which is a diffusion barrier, and glucose has to cross it before it can reach SGLT1. Third, it slows convective mixing, so amylase and starch meet more slowly. Every one of those changes the RATE at which glucose appears in blood. None of them changes the total amount that eventually gets there. Over 24 hours the carbohydrate is absorbed; what changes is the shape of the curve, and the shape is where the satiety and the glycemic claims live.

The dose-response is exactly what a viscosity mechanism predicts. Granular PGX at 0, 2.5, 5.0 and 7.5 g with breakfast reduced the incremental area under the blood glucose curve by up to 50 percent, and did it in a linear dose-response fashion, p < 0.001 Brand-Miller 2010. A linear relationship between grams of polymer and reduction in glucose area is the signature of a bulk physical property, not of a receptor: receptors saturate, viscosity does not until the material stops hydrating.

The end of the fiber is fermentation, and it is not silent. Konjac gum and konjac glucomannan are unlikely to be absorbed intact and are significantly fermented by the intestinal microbiota EFSA ANS Panel 2017; guar gum, the closest comparator with its own regulatory file, is described in the same terms -- practically undigested, not absorbed intact, significantly fermented by enteric bacteria in humans EFSA ANS Panel 2017. So the material ends as short-chain fatty acids and gas in the colon, which supplies a second candidate mechanism for anything measured the following morning, and the two mechanisms have never been separated in a trial of this product.

Cell, rodent, human — and where it stops

The acute human work is careful, it is small, and it contains one result that the marketing has never repeated. Three trials of 10 subjects each, mean age 24.4 +/- 2.6 years Brand-Miller 2010. Study 1 established the linear dose-response above. Study 2 asked when to take it: 5 g of granules or 4.5 g in capsules, given at 60, 45, 30 and 15 minutes before, at, and 15 minutes after a bread meal. The granular form reduced glycemia by up to 28 percent across the window from 45 minutes before to 15 minutes after, p < 0.001. The capsules did not. Study 3 then gave capsules at 3, 4.5 and 6 g with the evening meal and found glycemia at the following morning's breakfast reduced by up to 28 percent, p < 0.001.

Study 2 and study 3 together say something precise about capsules, and it is a usage instruction rather than a verdict. A capsule has to disintegrate and the powder has to hydrate before anything is viscous. Given 15 to 45 minutes it does not get there; given overnight it does. So the same material in the same dose either works or does not depending on how much time it is given to become a gel, which is what a mechanism built on viscosity should look like and is a stronger argument for the mechanism than any of the outcome data.

The satiety work is where the comparator choice needs reading. Fourteen subjects trained as a satiety panel (n = 14), double-blind randomized crossover, a standard meal plus 5 g of PGX granules, or 4.5 g of PGX softgels, or 5 g of wheat dextrin Solah 2016. Fullness area under the curve was 477 +/- 121 for granules and 454 +/- 242 for softgels against 215 +/- 261 cm.min for wheat dextrin, p < 0.001, and the glucose area was significantly lower after PGX than after wheat dextrin, p < 0.001. Wheat dextrin is a soluble fiber with essentially no viscosity, so this is a clean test of the mechanism and not a test of the product against its market. The paper's own mechanistic sentence is worth quoting for what it concedes: the high viscosity reported for PGX is a likely mechanism behind the effects observed. Reported, and likely. The viscosity was not measured in the study that attributed the result to it.

The appetite work under real caloric restriction is consistent and modest. Forty-five women randomized and 35 completing (n = 45 randomized, n = 35 analyzed), a 1,000 kcal per day diet for 3 days with 5 g of PGX or placebo at breakfast, lunch and dinner, crossover with a 3-week washout Kacinik 2011. PGX lowered the area under the curve for hunger on day 3 (p = 0.048), for prospective consumption on day 3 (p = 0.017) and across the 3-day average (p = 0.026), with the effect concentrated in the afternoon and evening. Three days is three days; this is a demonstration that the sensation changes, not that a body composition does.

Then the 52-week trial, which is the one that answers the question this product actually has to answer. One hundred and fifty-nine people (n = 159) randomized to rice flour control, PGX, or a proprietary psyllium product, 5 g taken with 500 mL of water 5 to 10 minutes before meals, with no other change to diet or exercise Pal 2016. Against control, PGX was 1.6 kg lower at 3 months, 2.6 kg at 6 months and 2.6 kg at 12 months. Against control, psyllium was 1.1 kg lower at 3 months and 2.4 kg at 6 months, and the 12-month comparison was not significant. Body fat against control: PGX 1.8 kg lower at 6 months and 1.9 kg at 12; psyllium 1.9 kg lower at 6 months and 1.4 kg at 12. As a fraction of starting weight the 12-month figures were a 2.8 percent difference for PGX and a 1.5 percent difference for psyllium against control.

Read those two columns side by side, because they are the honest answer to whether the proprietary blend beats a plain viscous fiber. At 6 months the two are 0.2 kg apart on weight, and on body fat the psyllium arm is numerically ahead. The case for the blend rests entirely on the 12-month time point, where psyllium's advantage has faded and PGX's has not -- and that case is built from two separate comparisons against a shared control rather than from a direct head-to-head test between the two fibers. It is a real observation and it is a weaker form of evidence than a difference measured between the two arms directly. The comparator was also a proprietary psyllium supplied within the trial, not the ispaghula husk on a supermarket shelf.

PGX — which form, and does it matter

Physical form is not a preference here, it is the variable that decided one of the three published acute studies. Granules worked in the pre-meal window and capsules did not, in the same experiment on the same subjects Brand-Miller 2010. Anything sold in a hard capsule and taken 15 minutes before eating is being asked to hydrate on a timetable the material has been measured failing to meet.

But capsule is not one thing, and the distinction is worth having. Softgels containing 4.5 g gave a satiety area under the curve of 454 +/- 242 against 477 +/- 121 for 5 g of granules -- essentially the same Solah 2016. A softgel carries a pre-dispersed fill rather than a dry powder, so it has less hydrating to do -- and note that it matched the granules on satiety while carrying 10 percent less material, 4.5 g against 5 g. And capsules taken with the evening meal reduced next-morning glycemia perfectly well Brand-Miller 2010, because overnight is long enough. Form interacts with timing; neither is a property of the product alone.

Water is part of the dose and should be written on the label as such. The 52-week trial specified 5 g with a total of 500 mL of water, 5 to 10 minutes before meals Pal 2016. A gel-forming polymer given too little water does not become a viscous solution; it becomes a bolus, which is the mechanism behind the obstruction risk in the safety section below and not a separate issue.

Doses that have actually been studied. Single doses of 2.5 to 7.5 g of granules Brand-Miller 2010; 5 g three times a day Kacinik 2011; 5 g before each meal, 15 g a day, for 52 weeks Pal 2016 Pal 2022. That is a coherent and well-bounded dosing literature, which is more than most of the shelf can say.

Now the price question, stated plainly. All three components are commodity food additives with their own regulatory dossiers -- konjac as E 425 EFSA ANS Panel 2017, alginates as E 400 to E 404 EFSA ANS Panel 2017 -- and plain konjac glucomannan is sold by the kilogram at a small fraction of the cost of the branded complex. The premium buys the cross-linking process. The evidence for that premium is the 12-month divergence in one trial against a proprietary psyllium Pal 2016, and there is no published head-to-head against plain glucomannan at equal grams at all.

And the specification the panel does not carry is the one the mechanism is made of. No PGX label states an apparent viscosity at a defined concentration and shear rate, a molecular weight for the glucomannan fraction, or the ratio of the three polymers. For a product whose entire claim is a rheological property, that is the equivalent of a protein powder declining to state its protein content -- and it is measurable on a benchtop rheometer for the price of an afternoon.

What would have to be true, and how you would know it was not

1. The breakfast curve, on granules, at 45 minutes. Same white-bread breakfast at the same hour on alternate days, six pairs, with a continuous glucose monitor. Prediction: a visibly flattened and lowered peak on the PGX mornings, on the order of a 30 to 50 percent reduction in incremental area at 2.5 to 7.5 g Brand-Miller 2010. This is the cheapest real experiment in this whole file and anyone with a sensor can run it in a fortnight.

2. The second-meal effect, which distinguishes the two candidate mechanisms. Capsules at 3 to 6 g with dinner reduced the following morning's glycemia by up to 28 percent Brand-Miller 2010. Prediction: fasting glucose on waking is lower after a PGX dinner than after a control dinner. That matters because pure viscosity should be gone by morning -- the stomach is empty -- while colonic fermentation of the fiber and the short-chain fatty acids it produces should not be EFSA ANS Panel 2017. A persistent overnight effect is evidence for fermentation; an effect confined to the meal itself is evidence for viscosity.

3. The prediction that argues against the price. PGX 5 g against plain konjac glucomannan 5 g, matched for water volume and timing, same test meal, incremental glucose area as the endpoint. Prediction: the difference is small enough that a trial would need a large sample to resolve it, because both are doing the same physics Pal 2016. If PGX halves the peak and plain glucomannan does not touch it, the argument on this page is wrong and the cross-linking is buying something real.

4. HbA1c and fasting insulin at 12 weeks, with a direction for each. Prediction: a small fall in HbA1c and very little movement in fasting insulin, because a change in the rate of glucose appearance is not a change in insulin sensitivity. If fasting insulin falls substantially while HbA1c does not, the mechanism described on this page is not the mechanism operating, and the page should be rewritten rather than the result explained away.

5. Micronutrients at 12 months, because 3 months was measured and was not enough. Serum micronutrient concentrations showed no significant between-group difference after 3 months of 15 g a day of PGX or psyllium Pal 2022. Prediction: ferritin and vitamin D are still unchanged at 12 months on the same dose. If ferritin falls, then a viscous fiber is doing something to mineral absorption that a 3-month window was too short to reveal, and the reassurance in that trial has been over-read by everyone quoting it, this page included.

What nobody has tested yet

Nobody has measured the viscosity and the glycemic response in the same study. The trial that attributes its results to viscosity describes it as the viscosity reported for PGX Solah 2016. For a product whose mechanism is a physical measurement, the single experiment that would join mechanism to outcome -- rheometry on the actual test dose alongside the actual glucose curve -- has not been published. The reading is a single number: apparent viscosity in millipascal-seconds at a stated shear rate and a stated concentration in simulated gastric fluid, at 37 degrees Celsius. One afternoon on a benchtop instrument would put the mechanism and the outcome in the same paper for the first time.

Nobody has published a gastric emptying study on PGX. Slowed gastric emptying is the first step of the stated mechanism and it is routinely measurable, by scintigraphy, by a stable isotope breath test or by the paracetamol absorption method. None of the three has been done for this product, so the first link in its own causal chain is assumed rather than shown.

There is no head-to-head against plain konjac glucomannan. The published comparators are wheat dextrin, chosen precisely for having no viscosity Solah 2016, and a proprietary psyllium Pal 2016. The comparison a buyer needs -- branded complex against the cheapest single ingredient inside it, at equal grams -- has never been run, and it is the one that decides whether the product is worth its price.

Nobody has measured short-chain fatty acids on PGX. Konjac glucomannan is significantly fermented EFSA ANS Panel 2017, which means a fermentation mechanism is available for every result the viscosity mechanism is credited with, and fecal or plasma acetate, propionate and butyrate have not been reported in any PGX trial. Two mechanisms, one set of results, no measurement that separates them.

And the composition of the commercial product is not in the public record. The ratio of glucomannan to alginate to xanthan, the molecular weight of the glucomannan used, and the chemistry of the cross-linking step are proprietary. That is a commercial right and it is also a limit on what any reader can check, so it belongs here rather than being written around.

PGX — its own safety story, not its category's

The risk that belongs to this product is mechanical, and it is the one genuine hazard in this whole file. A polymer that forms a gel is supposed to do it in the stomach. Swallowed with too little water, or lying down, or immediately before sleep, it can hydrate in the esophagus instead, and a gel in the esophagus is an obstruction. The dosing protocol from the 52-week trial is the mitigation and should be treated as part of the dose rather than as advice: 5 g with a total of 500 mL of water, taken 5 to 10 minutes before a meal, sitting upright Pal 2016. Anyone with dysphagia, an esophageal stricture or ring, achalasia, gastroparesis, a previous bezoar, or any narrowing anywhere in the gut should not start a high-viscosity fiber without a clinician's say-so.

The gastrointestinal effects are dose-related and documented in the regulatory file rather than the marketing. After 3,000 mg a day of konjac glucomannan for 12 weeks in adults, several individuals experienced abdominal discomfort including diarrhea or constipation EFSA ANS Panel 2017. The same opinion records a no-observed-effect level of 1,250 mg per kg of body weight per day in a 90-day rat study, no genotoxicity concern, and a conclusion that no numerical acceptable daily intake was needed for its use as a food additive -- provided total intake from all sources stays below 3 g a day. Read that last figure precisely: it is an additive intake limit, it is not a verdict on supplement use, and PGX trials dose 15 g a day of the blend. A reader is entitled to know that the food-additive ceiling for one component sits below the studied supplement dose, and that these are two different regulatory questions with two different evidence bases.

The drug-timing rule follows from the mechanism and is not optional. A substance whose stated action is to slow gastric emptying and thicken the diffusion layer will slow the absorption of anything else in the lumen. Separate oral medication by at least 2 hours, and treat that as firm rather than cautious for a narrow-therapeutic-index drug -- levothyroxine, lithium, an antiepileptic, a direct oral anticoagulant.

The glucose interaction is real precisely because the product works. A 50 percent reduction in the incremental glucose area Brand-Miller 2010 on top of a fixed mealtime insulin dose or a sulfonylurea is two glucose-lowering actions on one meal. That is a reason for the prescriber to know, not a reason to avoid it.

On micronutrients the evidence is better than the reputation, and it is short. Three months of 15 g a day of PGX or psyllium produced no significant between-group difference in serum micronutrient concentrations, and the authors' conclusion was that fiber supplementation is unlikely to compromise nutritional status in the short term, with longer or higher-dose research recommended Pal 2022. Three months is not 12, so this is reassurance with a stated expiry date rather than a clean bill.

And the class comparison worth keeping in view. Guar gum, the nearest chemical relative with a full regulatory file, is likewise practically undigested, not absorbed intact, significantly fermented, well tolerated orally in adults, and carries an acceptable daily intake of 'not specified' -- a category allocated by the Joint FAO/WHO Expert Committee on Food Additives in 1970, 1974 and 1975 and endorsed by the Scientific Committee for Food in 1977, and reaffirmed on re-evaluation with no adverse effects at the highest dose tested in subchronic and carcinogenicity studies EFSA ANS Panel 2017. Viscous fibers as a class are not toxic. What they are is mechanically consequential in a narrow tube with too little water, which is the whole of the risk and the reason the water is not optional. Nothing here is medical advice, this page does not diagnose or treat obesity, diabetes or any other condition, and none of these statements has been evaluated by the Food and Drug Administration.

Sources read for this page

• Brand-Miller JC. Effects of PGX, a novel functional fibre, on acute and delayed postprandial glycaemia. Eur J Clin Nutr 2010 · PMID 20924393
• Solah VA. Consumption of the Soluble Dietary Fibre Complex PolyGlycopleX Reduces Glycaemia and Increases Satiety of a Standard Meal Postprandially. Nutrients 2016 · PMID 27164135
• Pal S. Effect on body weight and composition in overweight/obese Australian adults over 12 months consumption of two different types of fibre supplementation in a randomized trial. Nutr Metab (Lond) 2016 · PMID 27891167
• Kacinik V. Effect of PGX, a novel functional fibre supplement, on subjective ratings of appetite in overweight and obese women consuming a 3-day structured, low-calorie diet. Nutr Diabetes 2011 · PMID 23154443
• Pal S. Micronutrient status of individuals with overweight and obesity following 3 months' supplementation with PolyGlycopleX (PGX) or psyllium: a randomized controlled trial. BMC Nutr 2022 · PMID 35505399
• Mortensen A. Re-evaluation of konjac gum (E 425 i) and konjac glucomannan (E 425 ii) as food additives. EFSA J 2017 · PMID 32625526
• Younes M. Re-evaluation of alginic acid and its sodium, potassium, ammonium and calcium salts (E 400-E 404) as food additives. EFSA J 2017 · PMID 32625343
• Mortensen A. Re-evaluation of guar gum (E 412) as a food additive. EFSA J 2017 · PMID 32625396

PGX — safety & side effects

Bloodwork to run alongside PGX

Baseline first, then again at 8–12 weeks.

The Metabolic Health & Prediabetes panel covers these in one order — 8 markers, $81.45 with the discount applied.

Check results you already have → · All 103 markers A–Z

PGX — frequently asked questions

What PGX is used for

PGX appears under 1 goal in the goal router.

Related Metabolic & Weight supplements

← Browse the full Supplement Vault

↑ Back to on this page