Working Paper · Nutritional Biochemistry
The liver recovers from drinking faster than almost any other organ. Liver stiffness drops measurably within a week of stopping. What decides the long-term outcome is how much you drink from here on, not which supplements you take. Coffee has the best human data of anything on the popular "protect your liver" lists, and most of the rest is mouse work or fatty-liver trials that don't involve alcohol.
Alcohol is broken down almost entirely in the liver, and the process does damage in three ways. It floods the cell with NADH, which shuts down fat burning and builds up fat. It makes acetaldehyde, which is toxic. And through CYP2E1 it generates oxidants. Alcohol also loosens the gut wall, so bacterial endotoxin reaches the liver and inflames it. Most people drinking more than about 40 g a day (roughly three US drinks) get a fatty liver. Only some go on to inflammation, scarring and cirrhosis.1 In animals, the fat eaten alongside the alcohol matters a lot: alcohol plus corn oil injures the liver, and alcohol plus saturated fat largely doesn't.3,4
Recovery starts quickly. In heavy drinkers admitted for withdrawal, liver stiffness fell within seven days of stopping, and it kept falling over two months in those who stayed off alcohol. In those who relapsed it rose again.6,7 Over years, staying abstinent is one of the strongest predictors of survival, along with fibrosis stage. Once cirrhosis is present, even moderate drinking worsens outcomes.8,9
Against that backdrop, this paper checks a 12-item "things that protect your liver" post by Ryan Ayala (@RyanJAyala, October 2026), founder of the lab-testing company Labify Health. The post cites nothing and links no products. Coffee and thiamine hold up. Milk thistle and vitamin E are the right ideas in the wrong form. Glycine, theanine, resveratrol, grapefruit and taurine rest on animal work, and in one 2026 mouse study high-dose taurine made alcohol injury worse. TUDCA has no alcohol data, and its parent drug failed in alcohol-related cirrhosis. The post leaves out the main factor, the drinking itself, along with choline, magnesium, the type of dietary fat, and the one supplement with a positive survival signal in alcoholic hepatitis (NAC).
Five parts
The liver heals on its own when alcohol stops coming in. Supplements are a small adjustment on top of that, and none has been shown to offset continued heavy drinking.
Part III is the item-by-item check and can be read on its own. Part V is the practical summary. Entries on this site for Milk Thistle, DHM, NAC, L-Theanine, Beet Juice, Thiamine, Glycine and Vitamin E go deeper on single items.
Strong Moderate Animal–in vitro Inferred Contested Speculative
The body has no way to store alcohol, so the liver clears it first, ahead of other fuels. Most goes through alcohol dehydrogenase (ADH) to acetaldehyde, then through aldehyde dehydrogenase (ALDH2) to acetate. Both steps turn NAD+ into NADH. Heavy or chronic drinking also switches on a second enzyme, CYP2E1, which oxidises alcohol while leaking reactive oxygen species.
The Nature Reviews primer summarises the damage as steatosis, oxidative stress, acetaldehyde toxicity and cytokine-driven inflammation.1 The gut-endotoxin route matters for this paper because several items on the list (glycine, taurine) act on it, and so does the type of dietary fat (I.3).
Strongtextbook biochemistry and decades of human and animal work.
The stages run from fatty liver to steatohepatitis (fat plus inflammation) to fibrosis to cirrhosis, with cancer and acute alcoholic hepatitis branching off. Most people drinking more than 40 g a day develop fatty liver, which is about three US standard drinks at 14 g each, but only a subset progress further. Genes, sex, body weight and diet explain much of the difference.1 Fatty liver is the reversible stage. Fibrosis can partly regress. Cirrhosis largely stays, but its outcome still depends on drinking (II.3).
The trouble is that people can't tell which stage they're in. In an Austrian biopsy series, half the patients with "early", symptom-free alcohol liver disease already had severe fibrosis or cirrhosis (F3–4). Their 10-year mortality was 45%, against 0% for F0–2.8
Strongfor the spectrum and the dose threshold; Moderatefor the hidden-fibrosis finding, which is one centre's biopsy series.
Animal studies of alcohol and dietary fat point the same way as the Linoleic Acid paper. Amin Nanji's rat work found that liver pathology from alcohol only appeared when the alcohol was fed with polyunsaturated fat. The proposed mechanisms were more oxidative stress, more endotoxin and more inflammatory cytokines.3 A 2012 Louisville mouse study made the gut link explicit. With alcohol plus corn oil (linoleic acid), the mice had liver injury, fatty liver, a leakier ileum and more endotoxin in the blood. Alcohol plus saturated fat (MCT and beef tallow) did none of this. Corn oil alone, without alcohol, already weakened the gut's tight junctions.4
Animal–in vitroconsistent across labs and decades, but untested in people. Inferredthat a drinker who cooks in butter or tallow instead of seed oil takes less liver damage.
The same lab models that make milk thistle, glycine and resveratrol look good also show that the background diet largely decides whether alcohol injures the liver at all. If the animal pattern holds in people, swapping seed oil for butter, tallow, or red palm oil would matter more than anything on the X list. Red palm oil is the one fat the list itself names.
A healthy liver clears a night's alcohol within hours. The hangover is mostly acetaldehyde, dehydration (alcohol blocks the hormone that holds water), poor sleep and inflammation, not liver damage. The only systematic review of hangover remedies found 21 placebo-controlled trials in 386 people. No two trials tested the same thing, and every efficacy result was rated very low quality. Clove extract, tolfenamic acid, pyritinol, Hovenia (the DHM plant), L-cysteine, red ginseng and Korean pear juice each had one positive study.10 None of the 12 items in the X post has a hangover trial. See DHM and Molecular Hydrogen for the two remedies covered on this site.
Strongthat one night isn't a liver event for a healthy liver; Speculativefor any specific hangover remedy.
Two French studies used FibroScan (an ultrasound measure of liver stiffness) on heavy drinkers admitted for withdrawal.
Liver fat is the fastest part to clear, and GGT and ALT come down over the following weeks. That makes a 4–8 week break a meaningful test: if the numbers normalise, the liver was mostly inflamed and fatty rather than scarred.
Moderatetwo prospective cohorts, consistent direction, short follow-up.
In the Austrian biopsy cohort, staying abstinent predicted survival at both early and advanced stages.8 In the French CIRRAL cohort (650 people with compensated alcohol-related cirrhosis), drinking during follow-up worsened survival in a dose-dependent way. The authors concluded that no amount of alcohol is safe once cirrhosis has developed. A third of people who had stopped had started again within five years.9 The AASLD and Nature Reviews summaries agree that abstinence is the best treatment at every stage of alcohol liver disease.1,2
Strongobservational, but large, consistent and dose-dependent, and no trial is ethically possible.
The post in full: "12 things that protect your liver: Thiamine (B1). Milk thistle (tea, not pills). Glycine. Taurine (a few weeks at a time). TUDCA (a few weeks at a time). Coffee, only after eating, and only if your gut is fine. Grapefruit (skip it if you're on meds). Cranberry extract. Resveratrol. Beetroot juice. L-theanine. Vitamin E from natural sources, like red palm oil." No sources were given, so each item is checked against the best evidence I could find, giving priority to alcohol-specific human data.
| Claim | What the literature shows | Holds? |
|---|---|---|
| Ryan Ayala (@RyanJAyala) · October 2026 | ||
| Thiamine (B1) | Chronic drinking causes thiamine deficiency three ways: less in the diet, less absorbed from the gut, and less used in the cells. The damage is to the brain (Wernicke–Korsakoff).11 UK guidance is to give B vitamins by injection at the outset to heavy drinkers with a poor diet.12 Oral thiamine reaches high blood levels at high doses because it's absorbed by passive diffusion as well as by transporters.13 In a 24-week trial of 600 mg benfotiamine in 120 actively drinking alcohol-dependent people, women cut their drinking more than on placebo (N=21 women).14 See Thiamine. | Holds. The most important item on the list, though it protects the brain more than the liver. |
| Milk thistle, "tea, not pills" | The trials used standardised extract capsules. Ferenci 1989 (140 mg three times a day, 170 cirrhotics): 4-year survival 58% vs 39%, driven by the alcoholic subgroup.15 Parés 1998 (150 mg three times a day, 200 alcoholic cirrhotics): no effect.16 Cochrane, 18 trials: no effect on overall mortality. Liver-related deaths were halved across all trials, but not in the high-quality ones.17 On tea: boiling water extracts the polar flavonolignans (taxifolin, silychristin) well, but ethanol extracts the most silymarin overall.18 At 100 °C, yields took about 200 minutes to peak.19 A 5-minute steep gets a small fraction, and little of the silybin the trials dosed. See Milk Thistle. | Half holds. The herb has a contested signal; "tea, not pills" gets the form backwards. |
| Glycine | Thurman's lab, rats on continuous alcohol: glycine sped up alcohol breakdown in the stomach, so less reached the liver, and it cut steatosis and necrosis.20 Given after 6 weeks of alcohol was stopped, a glycine diet brought ALT and AST down about 30% faster, and the pathology score was 0.5 vs 2.6. The proposed mechanism is glycine calming Kupffer cells through chloride channels.21 No human trial. See Glycine & Collagen. | Plausible, animal only. The recovery study is the closest match to what the post claims. |
| Taurine, "a few weeks at a time" | No human trial in alcohol liver disease found. A 2026 Tulane/Louisville mouse study: high-dose taurine (3 g/kg/day) made alcohol injury worse. ALT and AST were higher, there was more fat and inflammation, hydrogen-sulfide-producing gut bacteria increased, the gut barrier was impaired and more bacteria reached the liver. A low dose (0.2 g/kg/day) improved it.22 By standard body-surface-area scaling, those doses equal roughly 17 g and 1 g a day for a 70 kg adult. A reply under the post ("you do not have to cycle taurine") is just as unsupported as the post: nobody has tested cycling. | Half holds. Dose is what the data speak to, and cycling is untested. |
| TUDCA, "a few weeks at a time" | No trial in alcohol liver disease. TUDCA performed the same as its parent drug UDCA in primary biliary cholangitis (199 patients; ALP response 76% vs 81%).23 UDCA itself, in 226 people with alcohol cirrhosis and jaundice: 6-month survival was 69% on UDCA vs 82% on placebo. That difference wasn't significant after adjusting for bilirubin, and UDCA certainly didn't help.24 High-dose UDCA in sclerosing cholangitis improved liver tests yet more than doubled the rate of serious endpoints.25 Better enzyme numbers didn't mean a better liver. | Does not hold for alcohol. It's a drug for bile-flow problems. |
| Coffee, "only after eating, and only if your gut is fine" | Meta-analysis of 9 studies (432,000 people): each extra 2 cups a day was linked to a 44% lower risk of cirrhosis, and a 38% lower risk of alcoholic cirrhosis.26 UK Biobank (495,000 people): coffee drinkers had 21% less chronic liver disease and 49% fewer deaths from it. Decaf, instant and ground coffee looked alike, so caffeine isn't the whole story.27 All observational. | Holds. The best human data on the list. The after-food and gut caveats are his own, not from studies. |
| Grapefruit, "skip it if you're on meds" | The drug warning is right: grapefruit switches off CYP3A4 in the gut wall for about 24 hours and raises blood levels of dozens of drugs. The interaction was found by accident, when grapefruit juice was used to mask the taste of ethanol in a felodipine–alcohol study.28 I found no human trial of grapefruit for the liver. The benefit claim rests on naringenin in rodents and zebrafish. Alcohol is cleared by ADH and CYP2E1, not CYP3A4, so grapefruit doesn't speed it up. | Warning holds; benefit unsupported. |
| Cranberry extract | Two small Iranian trials in non-alcoholic fatty liver. 144 mg a day for 6 months (110 enrolled): ALT unchanged, insulin resistance and triglycerides lower, more patients' steatosis improved.29 288 mg a day plus a weight-loss diet (41 patients): ALT fell more, steatosis grade no different from placebo.30 No alcohol data. | Weak. Fatty liver, not alcohol. |
| Resveratrol | In alcohol-fed mice it restored SIRT1–AMPK signalling and prevented fatty liver.31 In people with fatty liver, 3 g a day for 8 weeks did nothing for liver fat or insulin resistance, and raised ALT and AST.32 An umbrella review of the meta-analyses: the evidence does not support it.33 | Does not hold in people. |
| Beetroot juice | Karolinska mice: dietary nitrate prevented fatty liver through AMPK, and did nothing in germ-free mice, so it depends on bacteria to convert it. Two authors hold patents on therapeutic nitrate.34 An Iraqi trial of 180 people with non-alcoholic fatty liver found beet juice (with or without a Mediterranean diet) lowered ALT and steatosis versus control. But the participants also lost about 4.5 kg, and the abstract says the juice "had 250 mg of beetroot", which can't be right.35 See Beet Juice. | Plausible. Good mechanism, weak trial, no alcohol data. |
| L-theanine | Two mouse studies of acute alcohol: theanine lowered blood alcohol and acetaldehyde, raised ADH and ALDH, and damped CYP2E1 and lipid peroxidation.36,37 The 2005 study injected it. No human data. See L-Theanine. | Animal only. |
| Vitamin E "from natural sources, like red palm oil" | The best fatty-liver trial on the list: PIVENS, 800 IU natural vitamin E for 96 weeks in 247 non-diabetic adults, improved NASH in 43% vs 19% on placebo, though fibrosis didn't improve.38 In alcoholic hepatitis, 1,000 IU for 3 months didn't improve liver tests.39 Palm vitamin E is about 70% tocotrienols.41 Delta-tocotrienol at 300 mg twice daily matched alpha-tocopherol in fatty liver, in a trial with no placebo arm.40 Food delivers milligrams, while PIVENS dosed about 536 mg. See Vitamin E. | Half holds. Right nutrient, but the trial dose is far above food and the trial disease isn't alcohol. Red palm oil is a good cooking fat for another reason (I.3). |
Coffee (large cohorts, includes alcoholic cirrhosis). Thiamine (real deficiency in drinkers; protects the brain).
Milk thistle: one positive and one null trial in alcoholic cirrhosis; Cochrane unconvinced. Standardised extract, not tea.
Vitamin E (strong for NASH at 800 IU, null in alcoholic hepatitis). Cranberry, beet juice (small trials).
Glycine, L-theanine, grapefruit. Taurine, where the effect depends on dose.
Resveratrol (null, ALT rose). TUDCA (no alcohol data; UDCA null-to-harmful in alcohol cirrhosis).
Not one of the 12 items mentions drinking less. Every human study in Part II says that's what decides the outcome. A reply under the post (@primal_edge) did better without any supplements: enough protein, eggs for choline, collagen for glycine, meat and seafood for taurine, avoid excess alcohol, keep insulin sensitivity, sleep, and "give your liver fewer problems instead of endlessly searching for another detox." To be fair to the post, it doesn't sell anything, and its two most expensive-sounding items (TUDCA, resveratrol) are the ones with the weakest human data. That pattern is common: supplement lists put mechanism and novelty ahead of the outcome data.
One reply asked "NAC?" It's the omission with the best data. In a French trial of 174 people with severe alcoholic hepatitis, adding 5 days of IV NAC to prednisolone cut 1-month mortality from 24% to 8%, and deaths from kidney failure caused by the liver (hepatorenal syndrome) and infections fell too. But 6-month mortality, the primary endpoint, missed significance (27% vs 38%, p=0.07).43 That's a hospital drug regimen. For an outpatient drinker, NAC's case is the glutathione logic covered in the NAC and Glutathione entries, with no oral trial in alcohol-related liver disease.
Moderatefor IV NAC in severe alcoholic hepatitis; Inferredfor oral NAC in ordinary drinkers.
Choline is needed to export fat from the liver as VLDL. In a UNC depletion study, 77% of men and 80% of postmenopausal women on a choline-free diet developed fatty liver or muscle damage within weeks. Six men did so even at the official Adequate Intake.42 Eggs (about 150 mg each) and beef liver are the dense sources. A drinker eating chicken breast and protein shakes may be short.
Strongthat choline deficiency causes fatty liver in people; Inferredthat it adds to alcohol's effect.
A 2022 Glasgow review of 46 studies found observational evidence that magnesium deficiency speeds alcohol liver injury, and weak or negative evidence for folate, vitamin D, zinc and selenium. It supports supplementing only when someone is actually deficient.5 Heavy drinkers lose magnesium in their urine, and serum levels miss tissue deficiency. See Magnesium.
Moderateobservational.
Covered in I.3. It's left out of every supplement list because nobody sells it.
Get medical help to stop. Withdrawal after daily heavy drinking can cause seizures and delirium tremens, which can kill. Detox is done with medication, and thiamine is given before any glucose, because glucose given to someone low on thiamine can bring on Wernicke's encephalopathy.12
Supplements: Milk Thistle · DHM · NAC · L-Theanine · Beet Juice · Molecular Hydrogen · Green Tea
Nutrients: Thiamine · Magnesium · Glycine & Collagen · Glutathione · Cysteine · Vitamin E · Vitamin A · Sodium. No entry yet for taurine, TUDCA, choline, coffee, cranberry or resveratrol.
The trial this field needs is boring: heavy drinkers cutting down, randomised to a seed-oil-free, choline- and protein-adequate diet or their usual diet, with FibroScan and enzymes at 8 and 24 weeks. That would test the dietary background the animal studies point to, and nobody sells it.