Chitosan

Also sold as: Chitosan Fat Binder

A cationic shell-derived polysaccharide sold as a fat binder. It really does bind fatty acids and bile salts — in a beaker. In a person, the measured amount is about a gram of fat a day.

How Chitosan actually works

A polysaccharide made by boiling the acetyl groups off crab or shrimp chitin, which exposes a primary amine with a pKa near 6.4. That amine is protonated in gastric acid and neutral in the duodenum, so the polymer is a cationic ion-exchange resin in the stomach and an inert precipitate below it. While charged it binds anionic fatty acids and bile salts and flocculates emulsified fat droplets, shrinking the interface pancreatic lipase works on. Both the charge density (degree of deacetylation) and the chain length (molecular weight) set how much it binds, and neither appears on a label.

Where to get Chitosan

The evidence for Chitosan

Graded by what exists behind each claim.

✅ Clinically validated

• A 12-day controlled-feeding balance study at 4.5 g/day against 133 g of dietary fat raised fecal fat from 6.1 to 7.2 g/day — an extra 1.1 g, about 9.9 kcal.
• A second balance study found 1.8 g/day extra fecal fat in men at 2.5 g/day and no measurable effect at all in women.
• A 24-week randomized trial in 250 adults at 3 g/day found a 0.4 kg difference from placebo, which the investigators called not clinically significant; fecal fat and fat-soluble vitamins did not differ between groups.
• A Cochrane review of 15 trials (1,219 participants) found −1.7 kg overall, no clear difference in fecal fat excretion, and substantially smaller effects once analysis was restricted to the larger, longer, allocation-concealed trials.

📊 Correlative data

• In simulated gastrointestinal digestion, 670 kDa chitosan at 90% deacetylation cut free fatty acid release by about 77%, with 30% bile salt binding and 57% lipolysis inhibition. That in-vitro number is what the product is sold on.

🧪 Theoretical / extrapolated benefits

• The gap between the beaker and the person is a pH window: the amine that carries the charge is protonated in the stomach and neutral in the duodenum, which is where fat is actually digested.

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 Chitosan actually does

The mechanism is one chemical fact: a sugar with a positive charge on it. Chitin is poly(N-acetyl-D-glucosamine), the beta-1,4 linked polysaccharide that makes up a crab shell. Boiling it in concentrated sodium hydroxide strips acetyl groups off the C2 nitrogen and leaves a free primary amine behind. Strip enough of them and the polymer becomes soluble in dilute acid, and at that point it is called chitosan. The proportion stripped is the degree of deacetylation, and it is the single number the whole product depends on, because the amine is what does the work and the acetamide it replaced does nothing.

That amine has a pKa of roughly 6.3 to 6.5, which puts the mechanism inside a specific compartment. In gastric juice at pH 1.5 to 3.5 the amine is protonated and the chain is a polycation carrying hundreds of positive charges. In the duodenum at pH 6 to 7 it is largely neutral, loses solubility and precipitates. So chitosan is an ion-exchange resin that is switched on in the stomach and switched off in the intestine, and lipolysis happens overwhelmingly in the intestine. Everything difficult about this product is downstream of that mismatch.

What a polycation can grab, it grabs by charge. Free fatty acids liberated by lipase carry a carboxylate with a pKa near 4.8. Bile salts -- glycocholate, taurocholate -- carry a carboxylate or a sulfonate and are anionic across the whole gut. Both are electrostatic targets for a protonated amine, and there is a third effect that is purely physical: a cationic polymer bridges the anionic surfaces of emulsified oil droplets and flocculates them, which collapses the interfacial area that pancreatic lipase has to work on. Fat is not digested in bulk, it is digested at a surface, so shrinking the surface slows the enzyme without touching it.

All three of those have been measured, in the same experiment, with numbers. Chitosan variants spanning 3.2 to 670 kDa in molecular weight and 71.2 to 92.5 percent deacetylation were run through a simulated gastrointestinal digestion against corn oil Zhang 2025. Every one of them suppressed digestion to some degree. The best -- 670 kDa at 90 percent deacetylation -- cut the free fatty acid release rate by about 77 percent, and the paper separates the contributions: severe emulsion flocculation limiting enzyme access, 30 percent bile salt binding, and 57 percent inhibition of lipolysis. That is a coherent, quantified, three-part mechanism, and it is why the product exists.

Now hold that 77 percent next to what a person excretes, because that is the whole page. A 12-day controlled feeding study gave 15 men 4.5 g of chitosan a day as 10 capsules, 30 minutes before each of five meals, on an unrestricted diet averaging 133 g of fat a day, and collected every stool for 11 days with charcoal markers to split the collection into a control period and a supplement period Gades 2003. Fecal fat went from 6.1 g/day to 7.2 g/day. The chitosan effect was 1.1 g of fat per day. Per capsule that is 0.11 g of fat trapped by 0.45 g of chitosan, and in energy terms it is 9.9 kcal a day.

The gap between 77 percent and 1.1 g is not a contradiction, it is the mechanism failing at the pH boundary. A beaker holds the polymer, the oil and the enzyme together at whatever pH the experimenter chose, for as long as the experiment runs. A person moves the mixture out of the acid compartment where chitosan is charged into the alkaline one where it is not, in the presence of bile that re-emulsifies what was flocculated, at a transit rate nobody controls. The binding capacity is real and the residence time under charged conditions is too short for it to matter.

Cell, rodent, human — and where it stops

This is one of the few supplements where the human end of the chain is the well-populated end, and it is the reason the answer is unusually firm. Fecal fat balance is a hard endpoint: you feed a known amount of fat, you collect everything that comes out, you extract and weigh the fat in it. There is no surrogate and no self-report.

Study one: 15 men, 4.5 g/day, 133 g fat/day. Fecal fat rose 1.1 +/- 1.8 g/day, p = 0.02 Gades 2003. Statistically real, and the authors did the arithmetic themselves: at 9.9 kcal a day this cannot move energy balance. A pound of body fat is roughly 3,500 kcal, so 10 capsules a day for about a year buys one pound, assuming nothing else in the body compensates.

Study two is the one that should be quoted more, because it splits by sex and the split is informative. Twelve men and twelve women, 2.5 g/day, same design Gades 2005. Men ate 137 +/- 31 g of fat a day and excreted an extra 1.8 +/- 2.4 g/day, p = 0.02. Women ate 89 +/- 16 g of fat a day and excreted an extra 0.0 +/- 1.4 g/day, p = 0.99. Nothing. The authors put it in plain English: for men it would take more than 7 months to lose one pound of body fat, and for women no fat was trapped. Note what actually differed -- each two-capsule dose met 28 +/- 11 g of fat in the men and 18 +/- 7 g in the women. The likeliest reading is not that female physiology resists chitosan but that the binding is proportional to the fat present per dose, and below some threshold the effect disappears into the measurement noise.

Study three took it to 250 people for 24 weeks and measured the outcome people actually buy it for. Three grams a day against placebo, everyone given the same dietary and lifestyle advice Mhurchu 2004. The chitosan arm lost 0.4 kg, a 0.4 percent loss; the placebo arm gained 0.2 kg. The difference reached p = 0.03 and the authors' own conclusion is that it was not a clinically significant loss of body weight. Total and LDL cholesterol and glucose moved by similarly small amounts. Fecal fat and fat-soluble vitamins were among the secondary outcomes and showed no significant difference between groups -- which matters in both directions, because it is simultaneously a failure of the mechanism and a reassurance about the commonest warning attached to the product.

Pooled, the picture is a textbook demonstration of what trial quality does to an effect size. A Cochrane review of 15 trials and 1,219 participants found a weighted mean difference of -1.7 kg (95% CI -2.1 to -1.3), a total cholesterol fall of 0.2 mmol/L, and lower systolic and diastolic pressure Jull 2008. It also found no clear difference in fecal fat excretion, which is the mechanism the weight loss is supposed to come from. Restricting to trials that met allocation concealment criteria, or were larger, or ran longer, produced substantially smaller decreases in both weight and cholesterol. The review's conclusion is that the effect on body weight is minimal and unlikely to be of clinical significance.

And here is the comparison that makes the number legible, because a drug exists that does what chitosan claims to do. Orlistat is a covalent lipase inhibitor, and its dose-response was fitted across 11 phase I studies and 171 subjects on doses from 30 to 1,200 mg/day, with fecal fat excretion as the endpoint Zhi 1994. The maximum mean fraction of ingested fat recovered in stool was about 32 percent, against 5 percent on placebo, with half the maximum effect at 98 mg/day and a plateau above roughly 400 mg/day. On the 133 g/day fat intake of the chitosan balance study, 32 percent is about 43 g of fat a day. Chitosan delivered 1.1 g. Both numbers come from the same kind of study, in the same units, and they differ by a factor of roughly 40. That is what a real absorption blocker looks like, and it is the benchmark this product is competing against whether or not the label says so.

Chitosan — which form, and does it matter

Two numbers decide whether a given chitosan binds anything, and no retail panel prints either. They are the degree of deacetylation -- how many amines are available to carry charge -- and the molecular weight, which sets how many binding sites travel together on one chain and how viscous the solution is. A bottle that says 500 mg chitosan has told you the mass of a polymer whose two functional specifications are unstated.

The in-vitro work says the two together are the answer. Across variants from 3.2 to 670 kDa and 71.2 to 92.5 percent deacetylation, the strongest suppression of lipid digestion came from the high end of both: 670 kDa at 90 percent Zhang 2025. Higher charge density and a longer chain, together.

The manufacturing work says you cannot simply order both. Crayfish shell chitosan prepared at a low and a high degree of deacetylation ran 76.98 to 94.98 percent by potentiometric titration and 73.79 to 92.06 percent by elemental analysis, and as the deacetylation step ran longer and the degree rose, molecular weight fell from 424.03 to 334.66 kDa, apparent viscosity fell from 16.82 to 9.63 cP, water binding capacity fell from 481.29 to 428.04 percent and fat binding capacity fell from 419.30 to 355.75 percent Kadak 2023. The hot alkali that creates the amines also cuts the backbone. So the process that improves one specification degrades the other, and a manufacturer is choosing a point on that trade-off that the buyer never sees.

Even the acid used in demineralization changes the product. Shrimp chitosans extracted with different organic and mineral acids differ in crystallinity, in the order hydrochloric > citric > sulfuric > lactic > acetic El-Araby 2022. Crystallinity governs how readily the polymer hydrates and swells, which governs how much of it is available to bind anything at all in the 30 minutes before a meal. Two capsules identical on the panel can come off two different process lines.

Source species is the one specification a label sometimes does carry, and it is the one that matters for a different reason. Most commercial chitosan is crustacean -- shrimp, crab, crayfish. It is also made from squid pen and, importantly, from fungal mycelium, which contains no crustacean protein at all. If the shellfish question is live for a particular reader, fungal-source chitosan removes it rather than managing it.

Salt form is a real distinction and is usually invisible. Chitosan free base is insoluble at neutral pH; chitosan hydrochloride, glutamate and ascorbate are pre-salted and dissolve faster. A faster dissolving salt reaches its charged state earlier in the gastric window and is mechanistically the better bet, but no human balance study has compared salts, so this is an argument from chemistry and not a measured result.

Doses actually studied: 2.5 g/day Gades 2005, 3 g/day Mhurchu 2004 and 4.5 g/day Gades 2003. The highest of those produced 1.1 g of fecal fat. There is no published dose above 4.5 g/day with a balance endpoint, which means nobody can say what 9 or 12 g/day would do -- and that, rather than the form question, is the open experiment.

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

1. The balance test, and it is the one that settles the product. Four days of controlled feeding at a known fat intake with full stool collection, off and on the label dose. Prediction: the change in fecal fat is under 2 g/day, because that is what 4.5 g/day produced against 133 g of dietary fat Gades 2003. If a commercial chitosan puts an extra 20 g of fat a day in the stool, the argument on this page is wrong and the product is a different one from the products that have been tested.

2. Body weight at 24 weeks, measured properly. Prediction: roughly half a kilogram of difference against an honest control, which is the 24-week randomized result Mhurchu 2004. Weigh at the same time on the same scale and use a 7-day rolling average, because day-to-day water swings are several times larger than the entire effect being looked for. An effect this small cannot be seen at all by weighing once a fortnight.

3. The lipid panel, at 12 weeks, and the specific shape of the answer. If bile salt binding is operating -- 30 percent binding in the simulated gut Zhang 2025 -- then it is a weak version of what a bile acid sequestrant does, and the signature of a sequestrant is that LDL falls while HDL does not. Prediction: a total cholesterol fall on the order of 0.2 mmol/L, concentrated in LDL Jull 2008. If nothing at all moves in the lipid panel over 12 weeks, then the bile-binding half of the mechanism is not operating at supplement doses either.

4. The prediction that contradicts the usual warning. Vitamin D as 25-hydroxyvitamin D, and vitamin A, at baseline and 24 weeks. Prediction: no fall. The 250-person trial measured fat-soluble vitamins as a secondary outcome at 3 g/day for 24 weeks and found no significant difference between groups Mhurchu 2004, which is exactly what a compound that removes 1.1 g of fat a day should do to fat-soluble vitamin status: nothing. If vitamin D does fall, the honest conclusion is not that the warning was right all along but that the product being taken is binding more than the trial product did, and then the first prediction above should have been positive too.

5. The experiment that would rescue the category if anything could. A dose-response with fecal fat as the endpoint, the way orlistat got one: four arms at 1.5, 4.5, 9 and 15 g/day, fixed fat intake, full collection Zhi 1994. Prediction: fecal fat rises with dose but does not approach the roughly 32 percent of ingested fat that a lipase inhibitor achieves, and the gastrointestinal tolerability limit arrives before any interesting dose does. If instead the curve keeps climbing, then chitosan was simply never dosed high enough and 40 years of small trials tested the wrong amount.

What nobody has tested yet

Nobody has run a fecal fat balance study on a chitosan whose degree of deacetylation and molecular weight were declared. Both human balance studies used commercial products and neither reports either specification Gades 2003 Gades 2005, while the in-vitro work says those two numbers are what decide the answer Zhang 2025. The two literatures cannot be joined, and joining them costs one study.

Nobody has looked inside the duodenum. The mechanistic claim on this page is that the chitosan-fatty acid complex formed at gastric pH does not survive the pH rise and the bile salt load downstream. That is testable by intubation -- aspirate duodenal contents after a labeled fat meal with and without chitosan, and measure bound versus free fatty acid. It has not been done, so the pH argument, which is the best explanation of the discrepancy, remains an inference from chemistry.

No controlled oral challenge of chitosan has been published in crustacean-allergic adults. The clinically important shellfish allergens are proteins; chitosan is a polysaccharide, so the real question is how much residual shell protein a given process leaves behind. No manufacturer publishes a residual protein figure and no trial has enrolled allergic participants. This is an unanswered question and not a reassurance.

Nobody has tested whether chitosan binds drugs. A polycation loose in the stomach is a plausible binder of anionic drug molecules, which is a mechanism shared with bile acid sequestrants, and sequestrants carry formal separation instructions for exactly that reason. No pharmacokinetic interaction study of chitosan with any oral drug has been published, which means the caution below is derived from chemistry rather than from data.

And nobody has compared fungal chitosan with crustacean chitosan on any human endpoint. They differ in acetylation pattern and in molecular weight distribution as well as in allergen risk Kadak 2023. A head-to-head balance study would answer the safety question and the efficacy question in one experiment.

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

The honest headline is that the risk here is mostly opportunity cost, and that is worth saying before the specifics. Adverse events were not clearly different from placebo across 15 randomized trials and 1,219 participants Jull 2008. This is a poorly absorbed polysaccharide; it is not toxic, and the thing it is most likely to cost is the money and the attention that a person could have spent on something with a larger effect.

The shellfish question is specific and it is genuinely unresolved. Most chitosan on the market is made from crustacean shell. The allergens that cause crustacean anaphylaxis are proteins, and chitosan is a purified polysaccharide, so the risk is a function of residual protein left by a particular process -- a process that is already known to vary with the acid used at the demineralization step El-Araby 2022 and with how long the alkali step is run Kadak 2023. Nobody publishes that residual figure and nobody has run an oral challenge in allergic people. Anyone with a crustacean allergy severe enough to carry adrenaline should treat a shell-derived chitosan as unknown, and fungal-source chitosan avoids the question entirely rather than reducing it.

The fat-soluble vitamin warning is repeated everywhere and the one trial that measured it did not find it. Serum fat-soluble vitamins were a secondary outcome in a 250-person, 24-week randomized trial at 3 g/day and did not differ from placebo Mhurchu 2004. That is one dose, one duration and one product, so it is not a guarantee for a person taking twice as much for twice as long -- but it is the evidence that exists, and it is consistent with a compound removing about a gram of fat a day. Saying so is more useful than repeating a warning the data do not support.

The interaction caution is mechanistic, and it should be labeled as such. No interaction study exists. What is known is that a protonated polyamine in the stomach is the same chemistry that makes bile acid sequestrants bind warfarin, levothyroxine, digoxin and fat-soluble vitamins, and sequestrants therefore carry dosing-separation instructions. Treating chitosan the same way -- oral medication at least 2 hours away from a dose -- costs nothing and is the cautious reading of a chemistry that has never been tested. Anyone on warfarin should raise it with the prescriber rather than assume either way.

Two groups where this is not a self-service decision. Pregnancy, where there are no controlled exposure data at supplemental doses and the population is the least able to afford an unknown; and anyone whose weight or lipids are being managed clinically, because a small effect on total and LDL cholesterol has been reported Jull 2008 and a prescriber titrating a statin should know what else is in the mix. Nothing here is medical advice, this page does not diagnose or treat obesity or any other condition, and none of these statements has been evaluated by the Food and Drug Administration.

Sources read for this page

• Gades MD. Chitosan supplementation and fecal fat excretion in men. Obes Res 2003 · PMID 12740459
• Gades MD. Chitosan supplementation and fat absorption in men and women. J Am Diet Assoc 2005 · PMID 15635349
• Mhurchu CN. The effect of the dietary supplement, Chitosan, on body weight: a randomised controlled trial in 250 overweight and obese adults. Int J Obes Relat Metab Disord 2004 · PMID 15311218
• Jull AB. Chitosan for overweight or obesity. Cochrane Database Syst Rev 2008 · PMID 18646097
• Zhang J. Impact of chitosan on lipid digestion under simulated gastro-intestinal conditions. Food Chem X 2025 · PMID 41030815
• Kadak AE. Preparation and Characterization of Crayfish (Astacus leptodactylus) Chitosan with Different Deacetylation Degrees. Iran J Biotechnol 2023 · PMID 37228624
• El-Araby A. Physicochemical Properties and Functional Characteristics of Ecologically Extracted Shrimp Chitosans with Different Organic Acids during Demineralization Step. Molecules 2022 · PMID 36500378
• Zhi J. Retrospective population-based analysis of the dose-response (fecal fat excretion) relationship of orlistat in normal and obese volunteers. Clin Pharmacol Ther 1994 · PMID 8033498

Chitosan — safety & side effects

Bloodwork to run alongside Chitosan

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

Chitosan — frequently asked questions

What Chitosan is used for

Chitosan appears under 1 goal in the goal router.

Related Metabolic & Weight supplements

← Browse the full Supplement Vault

↑ Back to on this page