---
canonical_name: Dihydromyricetin
alternate_names: DHM, DMY, Ampelopsin, Ampeloptin
canonical_topic: Dihydromyricetin for Health & Longevity
short_topic_lc: dihydromyricetin
creation_date: 2026-1108-0149
creator_ai_fullname: Opus 5.5
ep_keywords: Flavonoids, Flavanonols, Dihydroflavonols, Polyphenols
---

# Dihydromyricetin for Health & Longevity

<section id="top" markdown="1"></section>  
Evidence Review created on 11/08/2026 using [AI4L](https://github.com/forever-healthy/AI4L) / Opus 5.5

**Also known as:** DHM, DMY, Ampelopsin, Ampeloptin

  
## Motivation

<!-- Author statement: this Motivation section was written only after all other sections of the document were completed, so that it reflects the full scope of the topic. -->

Dihydromyricetin (also called ampelopsin) is a natural plant compound found in high amounts in the leaves of Chinese vine tea and in the fruit of the Japanese raisin tree. It is sold as a dietary supplement, most often marketed as a remedy for the after-effects of drinking and increasingly promoted for metabolic health and healthy aging.

Vine tea has been brewed in southern China for centuries, and the raisin tree has a long history in East Asian folk medicine for alcohol intoxication. Scientific interest grew when laboratory researchers began studying its actions on the brain's main calming receptor and on cellular energy sensors linked to metabolism and aging. A small number of human trials have since examined its effects on metabolic markers such as liver tests and blood sugar.

This review examines what the human, animal and laboratory evidence shows about dihydromyricetin's benefits, risks, interactions and practical use, with a focus on adults who want to extend healthy lifespan, and where that evidence is strong, weak or missing.

**[Benefits](#expected-benefits) - [Risks](#potential-risks--side-effects) - [Protocol](#therapeutic-protocol) - [Conclusion](#conclusion)**

  
## Recommended Reading

<!-- Search statement: On 2026-11-07 the author ran web searches for "<expert> dihydromyricetin" for each priority expert and searched each priority platform on-site. foundmyfitness.com: on-site search blocked (d-browser 403 Cloudflare, d-fetch 403, d-proxy-1 Cloudflare, d-proxy-2 refused); the episode page "/episodes/alcohol" loaded via d-fetch and d-browser; the episode's chapter list includes "Dihydromyricetin (DHM) for reducing hangover severity — does it work?" (1:32:43), a brief discussion by name, with a matching show-notes paragraph on DHM; as by-name discussion from a priority expert the episode is listed. peterattiamd.com/?s=dihydromyricetin (d-fetch): "Nothing Found". hubermanlab.com/search-results?q=dihydromyricetin (d-browser): only unrelated semantic matches. chriskresser.com/?s=dihydromyricetin (d-fetch): no results. lifeextension.com search (d-browser): fuzzy product matches only; web searches found no Life Extension Magazine article. lifespan.io/?s=dihydromyricetin and ?s=ampelopsin (d-fetch): "No Articles Found". Non-priority expert commentary and narrative reviews were then searched on PubMed and the web. -->

These resources give a high-level overview of dihydromyricetin (DHM), from expert commentary on its hangover claims to narrative reviews of its biology.

- [The Truth About Alcohol: Risks, Benefits, and Everything In-Between](https://www.foundmyfitness.com/episodes/alcohol) - Rhonda Patrick

  A FoundMyFitness episode with a DHM chapter and show-notes section summarizing rodent evidence that DHM counters intoxication and withdrawal via GABA receptors (the brain's calming receptors), while noting no strong human evidence exists.

- [Can You Prevent Your Next Hangover By Using Cheers (Thrive+)?](https://www.mcgill.ca/oss/node/7153) - Jonathan Jarry

  A skeptical McGill Office for Science and Society analysis tracing how one rat study launched a DHM hangover brand, and why poor bioavailability (the share of a dose reaching the blood) undermines its claims.

- [The Versatile Effects of Dihydromyricetin in Health](https://pubmed.ncbi.nlm.nih.gov/28947908/) - Li et al., 2017

  A narrative review of DHM's antioxidant, anti-inflammatory and glucose- and lipid-regulating actions, noting that low bioavailability limits translation to people.

- [Preclinical Research of Dihydromyricetin for Brain Aging and Neurodegenerative Diseases](https://pubmed.ncbi.nlm.nih.gov/31780947/) - Martínez-Coria et al., 2019

  A narrative review of animal and cell studies of DHM in brain aging and in Alzheimer's, Parkinson's and Huntington's disease models, relevant to its longevity interest.

- [Dihydromyricetin: A review on identification and quantification methods, biological activities, chemical stability, metabolism and approaches to enhance its bioavailability](https://pubmed.ncbi.nlm.nih.gov/32288229/) - Liu et al., 2019

  Explains why DHM is unstable and poorly absorbed, how the body metabolizes it, and which delivery systems are being developed to improve uptake.

No directly relevant content on DHM was found from Peter Attia, Andrew Huberman, Chris Kresser, Life Extension Magazine or Lifespan.io; searches of their sites and of the web returned no article, episode or segment on the compound.

  
## Grokipedia

<!-- Search statement: On 2026-11-07 grokipedia.com/search?q=dihydromyricetin was loaded with d-browser (browser_navigate + page text); it returned 17 results, the first being the dedicated article "Ampelopsin", which covers dihydromyricetin under its synonym. The article page loaded with d-browser. d-fetch, d-proxy-1 and d-proxy-2 were not needed. -->

- [Ampelopsin](https://grokipedia.com/page/Ampelopsin)

  An AI-written encyclopedia entry covering DHM's chemistry, plant sources, pharmacokinetics (absorption and clearance in the body), preclinical research, clinical trials and safety; useful as a broad orientation, though secondary to the primary studies.

  
## Examine

<!-- Search statement: On 2026-11-07 examine.com/search/?q=dihydromyricetin and ?q=ampelopsin were loaded with d-browser (a Vercel security checkpoint cleared on reload); both returned "no search results". A search for "hovenia" with d-browser returned the article "Japanese Raisin Tree", which Examine describes as a source of dihydromyricetin (ampelopsin); it is Examine's only page covering the compound. d-fetch, d-proxy-1 and d-proxy-2 were not needed. -->

No Examine article on dihydromyricetin exists.

  
## ConsumerLab

<!-- Search statement: On 2026-11-07 consumerlab.com/search/?q=dihydromyricetin was loaded with d-browser; it returned the CL Answer "Dihydromyricetin (DHM) – Does It Help and Is it Safe?" plus two clinical updates (arthritis, fatty liver) that are sections of that same article. The article page loaded with d-browser; its body is member-only. d-fetch, d-proxy-1 and d-proxy-2 were not needed. -->

- [Dihydromyricetin (DHM) – Does It Help and Is it Safe?](https://www.consumerlab.com/answers/dihydromyricetin-dhm-does-it-help-and-is-it-safe/dhm-supplement/)

  ConsumerLab's review of DHM for liver health, fatty liver disease, blood sugar, cholesterol, hangovers and arthritis, including safety concerns; the full analysis is available to members only.

  
## Systematic Reviews

<!-- Search statement: On 2026-11-07 PubMed was searched for (dihydromyricetin OR ampelopsin OR "Ampelopsis grossedentata" OR "vine tea") AND ("systematic review" OR "meta-analysis"); 10 records were returned, of which four are systematic reviews or meta-analyses that include DHM. A repeat search adding "Hovenia dulcis" found Roberts 2022, a systematic review of hangover trials that includes a Hovenia extract trial. Selection weighed relevance, recency and size; no systematic review of DHM's risks or side effects exists. -->

The following systematic reviews and meta-analyses evaluate DHM's effects in alcohol-related conditions and fatty liver disease.

- [Therapeutic Effects of Dihydromyricetin on Wholly Alcohol-Attributed Conditions: A Systematic Review](https://pubmed.ncbi.nlm.nih.gov/42514290/) - Skinner et al., 2026

  Reviews 22 studies narratively; finds robust preclinical liver protection but only two small human trials, both of Hovenia dulcis extracts rather than pure DHM.

- [The efficacy and tolerability of pharmacologically active interventions for alcohol-induced hangover symptomatology: a systematic review of the evidence from randomised placebo-controlled trials](https://pubmed.ncbi.nlm.nih.gov/34972259/) - Roberts et al., 2022

  Reviews 21 placebo-controlled hangover trials; a *Hovenia dulcis* extract trial, not purified DHM, cut symptom scores, but all evidence was very low quality.

- [Efficacy of dietary polyphenol supplement in patients with non-alcoholic fatty liver disease: a network meta-analysis](https://pubmed.ncbi.nlm.nih.gov/40416369/) - Wang et al., 2025

  Ranks 13 polyphenols across 54 randomized trials (chance-assigned studies) in fatty liver disease; DHM ranks well for lowering LDL (low-density lipoprotein, plaque-building) cholesterol.

- [Food-derived dihydromyricetin and metabolic dysfunction-associated steatotic liver disease: a preclinical systematic review and meta-analysis](https://pubmed.ncbi.nlm.nih.gov/42022551/) - Jin et al., 2026

  Pools 14 mouse studies; DHM lowered liver fat, liver enzymes and blood glucose in diet-induced fatty liver, with moderate-to-high heterogeneity (variation between study results).

- [Pharmacological effects of dihydromyricetin in preclinical models of NAFLD: a systematic review and meta-analysis stratified by dose and duration](https://pubmed.ncbi.nlm.nih.gov/42625734/) - Li et al., 2026

  Pools 13 animal studies (316 animals); lipid outcomes tracked dose and liver-enzyme outcomes tracked treatment duration, informing dose choice.

No systematic review or meta-analysis of DHM's risks, side effects or drug interactions exists; the principal risks are therefore unrepresented in this section.

  
## Mechanism of Action

DHM, a flavanonol (a flavonoid subclass), has several targets:

- **GABA-A receptor modulation:** In rats, DHM bound the benzodiazepine site (the target of sedative anti-anxiety drugs) of GABA-A receptors (the brain's main calming ion channels) and counteracted alcohol's actions; the site blocker flumazenil abolished this ([Shen 2012](https://pubmed.ncbi.nlm.nih.gov/22219299/)).
- **Energy and antioxidant signaling:** In mice fed alcohol, DHM activated AMPK (AMP-activated protein kinase, a cellular fuel sensor) ([Silva 2020](https://pubmed.ncbi.nlm.nih.gov/32267550/)). Across preclinical alcohol studies, a systematic review found it modulated Nrf2 (a master switch for antioxidant genes) ([Skinner 2026](https://pubmed.ncbi.nlm.nih.gov/42514290/)).
- **Epigenetic and senescence effects:** DHM inhibited DNMT1 (the enzyme maintaining gene-silencing DNA methylation tags) in skin cells, in work by cosmetics maker Beiersdorf ([Falckenhayn 2023](https://pubmed.ncbi.nlm.nih.gov/38500495/)), and dampened the inflammatory secretions of senescent (aged, non-dividing) cells via PRDX2 (a DNA-protecting antioxidant protein) ([Xu 2026](https://pubmed.ncbi.nlm.nih.gov/41792133/)).
- **Competing view on alcohol metabolism:** DHM raised levels of alcohol-clearing enzymes in mice ([Silva 2020](https://pubmed.ncbi.nlm.nih.gov/32267550/)), but did not change alcohol dehydrogenase (the first alcohol-clearing enzyme) or alcohol elimination in rats ([Skotnicová 2020](https://pubmed.ncbi.nlm.nih.gov/33656905/)).
- **Pharmacology:** In rats, oral bioavailability was 4.02% ([Liu 2017](https://pubmed.ncbi.nlm.nih.gov/27951743/)) and half-life 3.70 hours ([Liu 2017](https://pubmed.ncbi.nlm.nih.gov/27951743/)). Blocking P-glycoprotein (P-gp, an intestinal pump expelling compounds) raised that bioavailability from 3.84% to 6.84% ([Huang 2018](https://pubmed.ncbi.nlm.nih.gov/28795912/)). A review reports DHM is broken down by reduction, methylation and attachment of sugar or sulfate groups ([Liu 2019](https://pubmed.ncbi.nlm.nih.gov/32288229/)); its enzymes are uncharacterized. It reaches mouse brain ([Carry 2021](https://pubmed.ncbi.nlm.nih.gov/34299083/)). In laboratory tests it inhibits CYP3A4, CYP2E1 and CYP2D6 (drug-clearing liver enzymes) ([Liu 2017](https://pubmed.ncbi.nlm.nih.gov/28614988/)). It is non-selective. No human pharmacokinetic study is published.

  
## Historical Context & Evolution

- **Original use:** Vine tea (*Ampelopsis grossedentata*, now *Nekemias grossedentata*) has been drunk in China for more than 1,200 years against heatstroke, sore throat and fatigue, and DHM is its main flavonoid, according to a review ([Zeng 2023](https://pubmed.ncbi.nlm.nih.gov/37343650/)). *Hovenia dulcis* has long served as an East Asian folk remedy for alcohol intoxication ([Kim 2017](https://pubmed.ncbi.nlm.nih.gov/28750942/)).
- **Hangover era:** A 2012 rat study reported that injected DHM counteracted alcohol intoxication and withdrawal and reduced drinking ([Shen 2012](https://pubmed.ncbi.nlm.nih.gov/22219299/)). A student read it and founded the hangover brand now called Cheers ([McGill OSS article](https://www.mcgill.ca/oss/node/7153)). A cohort study of its product was funded by the maker, Cheers Health ([Song 2026](https://pubmed.ncbi.nlm.nih.gov/42588049/)). Korean hangover trials of *Hovenia* extract drinks were run partly by employees of Kwangdong Pharmaceutical ([Lee 2024](https://pubmed.ncbi.nlm.nih.gov/39766963/); [Paik 2024](https://pubmed.ncbi.nlm.nih.gov/39767027/)).
- **Metabolic turn:** University researchers in Chongqing ran placebo-controlled trials in fatty liver disease ([Chen 2015](https://pubmed.ncbi.nlm.nih.gov/26032587/)) and type 2 diabetes ([Ran 2019](https://pubmed.ncbi.nlm.nih.gov/30089792/)). A Greek trial tested a DHM tablet supplied by its maker, Uni-Pharma ([Michailidou 2026](https://pubmed.ncbi.nlm.nih.gov/41612982/)).
- **Longevity interest:** Reports of DNA-methylation effects ([Falckenhayn 2023](https://pubmed.ncbi.nlm.nih.gov/38500495/)) and senescence effects ([Xu 2026](https://pubmed.ncbi.nlm.nih.gov/41792133/)) and a skin-serum study by the cosmetics maker Beiersdorf ([Qi 2026](https://pubmed.ncbi.nlm.nih.gov/42034839/)) broadened the focus.
- **What changed:** A rat study questioning effects on alcohol metabolism ([Skotnicová 2020](https://pubmed.ncbi.nlm.nih.gov/33656905/)) and a null hangover-score trial ([Paik 2024](https://pubmed.ncbi.nlm.nih.gov/39767027/)) tempered the alcohol claims, while metabolic and aging findings opened new questions; the human evidence on each remains small.

  
## Expected Benefits

<!-- Search statement: On 2026-11-07 the author searched PubMed (DHM/ampelopsin/vine tea/Hovenia dulcis with randomized, placebo, clinical trial, patients, volunteers, hangover, sleep, lifespan, senescence, arthritis), Europe PMC, ClinicalTrials.gov, the web, LiverTox and the Examine and ConsumerLab pages before writing this section. The search covered studies of any design finding no effect or the opposite effect: Ran 2019 (no lipid change), Paik 2024 (no hangover-score difference), Michailidou 2026 and LiverTox (no change in liver fat), and Skotnicová 2020 (no effect on alcohol metabolism) were found and are cited. -->

### High 🟩 🟩 🟩

No benefit reaches High: every human outcome comes from single small trials or from trials by one research group, and the hangover data test whole-plant extracts rather than purified DHM.

### Medium 🟩 🟩

#### Lower Liver Enzymes in Fatty Liver Disease ⚠️ Conflicted

In 60 adults with nonalcoholic fatty liver disease (NAFLD, liver fat not caused by alcohol), DHM 600 mg/day for 3 months lowered ALT, AST and GGT (alanine aminotransferase, aspartate aminotransferase and gamma-glutamyl transferase, liver-injury enzymes) versus placebo ([Chen 2015](https://pubmed.ncbi.nlm.nih.gov/26032587/)). A 12-month trial of DHM plus vitamins C and E and choline lowered liver stiffness ([Michailidou 2026](https://pubmed.ncbi.nlm.nih.gov/41612982/)). Liver fat did not fall in that trial ([Michailidou 2026](https://pubmed.ncbi.nlm.nih.gov/41612982/)), and LiverTox reports it fell no more than with placebo in the 3-month trial ([LiverTox record](https://pubmed.ncbi.nlm.nih.gov/37643278/)). Net reading: enzymes improve; liver fat does not.

**Magnitude:** Combined ALT and GGT normalization at 12 months in 32.1% versus 3.7% of patients by intention-to-treat (counting everyone randomized) with the DHM combination tablet versus placebo ([Michailidou 2026](https://pubmed.ncbi.nlm.nih.gov/41612982/)).

#### Improved Glycemic Control and Insulin Sensitivity

In adults with type 2 diabetes, 10 g of vine tea daily delivering 970 mg DHM for one month lowered fasting glucose and glycated albumin (a 2–3-week blood sugar marker) versus the same tea with DHM removed ([Ran 2019](https://pubmed.ncbi.nlm.nih.gov/30089792/)). In fatty liver disease, purified DHM lowered fasting glucose and HOMA-IR (an insulin-resistance index from fasting glucose and insulin) versus placebo ([Chen 2015](https://pubmed.ncbi.nlm.nih.gov/26032587/)). Both trials come from one research group (Mi and colleagues, Chongqing).

**Magnitude:** Fasting glucose and glycated albumin fell significantly more than with DHM-free tea (p < 0.05, a probability below 5% that chance alone explains the difference) ([Ran 2019](https://pubmed.ncbi.nlm.nih.gov/30089792/)), and fasting glucose and HOMA-IR fell significantly more than with placebo ([Chen 2015](https://pubmed.ncbi.nlm.nih.gov/26032587/)); the available trial reports give no outcome figure.

### Low 🟩

#### Kidney Function Markers in Type 2 Diabetes

In a 1-month trial, vine tea delivering DHM lowered cystatin C (a blood marker of kidney filtration) and retinol-binding protein 4 (an insulin-resistance marker) versus DHM-free tea ([Ran 2019](https://pubmed.ncbi.nlm.nih.gov/30089792/)). The evidence is indirect: whole vine tea, not purified DHM, was tested. No kidney outcome has been tested.

**Magnitude:** Cystatin C and retinol-binding protein 4 fell significantly more than with the DHM-free tea (p < 0.05) ([Ran 2019](https://pubmed.ncbi.nlm.nih.gov/30089792/)); the available trial report gives no outcome figure.

#### Lower LDL Cholesterol ⚠️ Conflicted

DHM lowered LDL cholesterol versus placebo over 3 months ([Chen 2015](https://pubmed.ncbi.nlm.nih.gov/26032587/)), and a 12-month DHM combination tablet lowered it from baseline only in the treated group ([Michailidou 2026](https://pubmed.ncbi.nlm.nih.gov/41612982/)). A larger 1-month vine-tea trial found no lipid change ([Ran 2019](https://pubmed.ncbi.nlm.nih.gov/30089792/)). Net reading: longer trials lowered LDL; the shorter trial did not.

**Magnitude:** LDL cholesterol and apolipoprotein B (a plaque-particle protein) fell significantly more than with placebo over 3 months ([Chen 2015](https://pubmed.ncbi.nlm.nih.gov/26032587/)), while lipids did not change versus DHM-free tea over 1 month ([Ran 2019](https://pubmed.ncbi.nlm.nih.gov/30089792/)); the available trial reports give no outcome figure.

#### Fewer Hangover Symptoms ⚠️ Conflicted

Controlled data test *Hovenia dulcis* (Japanese raisin tree) extracts, not purified DHM. Extract eased hangover symptoms in 26 men ([Kim 2017](https://pubmed.ncbi.nlm.nih.gov/28750942/)) and gut symptoms in 25 adults ([Lee 2024](https://pubmed.ncbi.nlm.nih.gov/39766963/)), but not total scores in 30 adults ([Paik 2024](https://pubmed.ncbi.nlm.nih.gov/39767027/)). Net reading: inconsistent and indirect.

**Magnitude:** Acute Hangover Scale totals of 3.0–3.8 points (0–7 scale) with no difference between extract and placebo drinks ([Paik 2024](https://pubmed.ncbi.nlm.nih.gov/39767027/)).

#### Smoother Skin and Lower Skin Methylation Age ⭕️ Not Central to Health & Longevity

Applied to the face twice daily for 8 weeks by 60 adults, a DHM serum reduced wrinkle visibility and roughness and lowered an epidermal DNA-methylation age estimate ([Qi 2026](https://pubmed.ncbi.nlm.nih.gov/42034839/)). The study was uncontrolled and used skin application, not oral intake. It bears on skin appearance.

**Magnitude:** Epidermal DNA-methylation age fell by a mean of about 2.1 years after 8 weeks, no control group ([Qi 2026](https://pubmed.ncbi.nlm.nih.gov/42034839/)).

### Speculative 🟨

#### Senescent Cell Modulation

DHM calmed the inflammatory output of senescent cells and eased tissue aging in prematurely aged mice ([Xu 2026](https://pubmed.ncbi.nlm.nih.gov/41792133/)). The basis is cell and mouse work only.

#### Longer Lifespan in Model Organisms

DHM extended lifespan in fruit flies ([Fan 2020](https://pubmed.ncbi.nlm.nih.gov/33291074/)) and raised mean lifespan by 9.1–15.0% in roundworms ([Jo 2026](https://pubmed.ncbi.nlm.nih.gov/42939825/)). No mammalian lifespan study exists.

#### Brain Aging and Neurodegeneration

In Alzheimer's model mice, DHM cleared senescent brain immune cells from amyloid plaques (protein clumps) and eased symptoms ([Xu 2026](https://pubmed.ncbi.nlm.nih.gov/41792133/)); a review reports similar signals in other neurodegeneration models ([Martínez-Coria 2019](https://pubmed.ncbi.nlm.nih.gov/31780947/)). No human data exist.

#### Less Alcohol Intoxication and Drinking ⚠️ Conflicted

Injected DHM reduced intoxication, withdrawal and voluntary drinking in rats ([Shen 2012](https://pubmed.ncbi.nlm.nih.gov/22219299/)). Oral DHM did not change alcohol metabolism in rats ([Skotnicová 2020](https://pubmed.ncbi.nlm.nih.gov/33656905/)). Net reading: animal-only, route-dependent and unconfirmed.

#### Exercise-Like Irisin Release

DHM raised circulating irisin, a muscle hormone released during exercise, in rats and humans ([Zhou 2015](https://pubmed.ncbi.nlm.nih.gov/26054747/)). Irisin is an unvalidated biomarker; no human performance or metabolic outcome was measured.

  
## Benefit-Modifying Factors

- **Genetic polymorphisms:** The positive *Hovenia* hangover trial enrolled only men carrying one inactive copy of ALDH2 (the enzyme that clears acetaldehyde, alcohol's toxic by-product), and variants of CYP2E1, which also breaks down alcohol, appeared to modify the response ([Kim 2017](https://pubmed.ncbi.nlm.nih.gov/28750942/)).
- **Baseline biomarkers:** Liver-enzyme and glucose benefits were seen in people with fatty liver disease ([Chen 2015](https://pubmed.ncbi.nlm.nih.gov/26032587/)) or diabetes ([Ran 2019](https://pubmed.ncbi.nlm.nih.gov/30089792/)); no trial tested people with normal values, who have less room to improve.
- **Sex:** No human trial reported results by sex. In mice, total oral exposure was 2.5 µM × h in males versus 0.7 µM × h in females ([Carry 2021](https://pubmed.ncbi.nlm.nih.gov/34299083/)).
- **Pre-existing conditions and weight loss:** In the 12-month fatty liver trial, patients whose ALT and GGT normalized lost a median 5 kg versus 1 kg in non-responders ([Michailidou 2026](https://pubmed.ncbi.nlm.nih.gov/41612982/)), but only DHM use independently predicted normalization ([Michailidou 2026](https://pubmed.ncbi.nlm.nih.gov/41612982/)).
- **Age:** No oral DHM trial targeted older adults. With the topical serum, epigenetic age reduction did not differ across 5-year age groups ([Qi 2026](https://pubmed.ncbi.nlm.nih.gov/42034839/)).

  
## Potential Risks & Side Effects

<!-- Search statement: On 2026-11-07 the author searched for DHM's side-effect profile in drug reference sources before writing this section. NIH LiverTox has a dihydromyricetin record (loaded with d-browser); drugs.com search returned 403 to d-fetch; no WebMD, Mayo Clinic or DailyMed monograph for dihydromyricetin was found by web search; DHM has no prescribing information because it is not an approved drug. The EFSA opinion on a Hovenia dulcis extract and PubMed (toxicity, safety evaluation, drug interaction, CYP, P-gp, adverse events) were also searched. Studies finding no excess risk were sought and found: no drug-related adverse events in either arm of Michailidou 2026, no liver-injury cases in LiverTox, and stable drinking in Song 2026. -->

### High 🟥 🟥 🟥

No risk reaches High: no trial from any research group has documented an adverse event attributable to DHM, and the safety record rests on a few short placebo-controlled trials.

### Medium 🟥 🟥

No risk reaches Medium: no single trial or controlled observational study has shown a harm; the concerns below rest on case registries, uncontrolled cohorts, or laboratory and animal work.

### Low 🟥

#### Liver Injury ⚠️ Conflicted

Many herbal supplements can injure the liver, so this is the key safety question for daily use. LiverTox lists no DHM-related case and rates it an unlikely cause ([LiverTox record](https://pubmed.ncbi.nlm.nih.gov/37643278/)), and liver enzymes fell rather than rose in a placebo-controlled trial ([Chen 2015](https://pubmed.ncbi.nlm.nih.gov/26032587/)). Net reading: no signal of liver harm.

**Magnitude:** No drug-related adverse event was reported in the DHM combination group or in the placebo group over 12 months ([Michailidou 2026](https://pubmed.ncbi.nlm.nih.gov/41612982/)).

#### Heavier Drinking From Milder Hangovers

DHM is sold mainly as a hangover aid, so users may drink more if next-day effects feel milder, while alcohol's organ damage continues. In a maker-funded cohort of moderate drinkers, scores on AUDIT-C (a 3-question alcohol-consumption screen) stayed stable ([Song 2026](https://pubmed.ncbi.nlm.nih.gov/42588049/)).

**Magnitude:** AUDIT-C scores were stable from baseline to exit in 68 participants, no control group ([Song 2026](https://pubmed.ncbi.nlm.nih.gov/42588049/)); the study reports no score figure.

### Speculative 🟨

#### Altered Drug Levels ⚠️ Conflicted

DHM inhibited CYP3A4, CYP2E1 and CYP2D6 in human liver tissue in the laboratory ([Liu 2017](https://pubmed.ncbi.nlm.nih.gov/28614988/)), but another laboratory study found little CYP3A4 inhibition ([Dombi 2024](https://pubmed.ncbi.nlm.nih.gov/39344282/)). Net reading: unresolved.

#### Additive Sedation

DHM acts at the brain's benzodiazepine binding site in rats ([Shen 2012](https://pubmed.ncbi.nlm.nih.gov/22219299/)), so stacking it with sedatives could deepen sedation. No human sedation reports exist; the basis is mechanistic.

#### Off-Target Epigenetic Effects

DHM inhibits DNMT1 in cultured skin cells ([Falckenhayn 2023](https://pubmed.ncbi.nlm.nih.gov/38500495/)); whether daily oral intake alters DNA methylation in other tissues is untested. The basis is mechanistic only.

  
## Risk-Modifying Factors

- **Genetic polymorphisms:** People with reduced-function CYP3A4 or CYP2D6 variants, or on drugs cleared by them, may be more exposed to any laboratory-predicted inhibition (inferred; no human data). ALDH2-deficient drinkers still accumulated acetaldehyde with *Hovenia* extract, which did not lower it ([Kim 2017](https://pubmed.ncbi.nlm.nih.gov/28750942/)).
- **Baseline biomarkers:** Already raised liver enzymes make any later rise hard to attribute; a baseline panel separates pre-existing disease from a new effect. Low-normal fasting glucose on diabetes drugs leaves less margin before hypoglycemia (low blood sugar).
- **Sex:** No sex differences in human side effects have been reported. Lower oral exposure in female mice ([Carry 2021](https://pubmed.ncbi.nlm.nih.gov/34299083/)) suggests sex may alter exposure; pregnancy is unstudied.
- **Pre-existing conditions:** Advanced liver disease, transplant immunosuppression, insulin-treated diabetes and alcohol use disorder raise the stakes of interactions or risk compensation; trials enrolled people with early fatty liver disease ([Michailidou 2026](https://pubmed.ncbi.nlm.nih.gov/41612982/)) or diabetes ([Ran 2019](https://pubmed.ncbi.nlm.nih.gov/30089792/)).
- **Age:** Older adults more often take several drugs cleared by CYP3A4 and are more prone to falls from sedation; no DHM safety trial has enrolled them specifically.

  
## Key Interactions & Contraindications

<!-- Search statement: On 2026-11-07 the author searched PubMed for dihydromyricetin with "drug interaction", "pharmacokinetics", cytochrome P450, CYP3A4, CYP2E1, liver microsomes and P-glycoprotein before writing this section. No human interaction or pharmacokinetic study was found; the evidence is in vitro (Liu 2017, Dombi 2024) and in rats (Huang 2018, Deng 2020). ClinicalTrials.gov lists one first-in-human pharmacokinetic study (NCT05623501) with no results. -->

**Documented interactions:**

- **Alcohol (tested with *Hovenia dulcis* extract, not purified DHM):** No interaction found on blood alcohol or acetaldehyde versus placebo ([Kim 2017](https://pubmed.ncbi.nlm.nih.gov/28750942/)); another trial found small reductions at 0.5 and 6 hours only ([Paik 2024](https://pubmed.ncbi.nlm.nih.gov/39767027/)). Intoxication is not meaningfully shortened, so impairment follows intake as usual.

**Theoretical interactions (inferred from mechanism, not studied in humans):**

- **CYP3A4-cleared prescription drugs (tacrolimus, cyclosporine, simvastatin, midazolam):** Inferred from laboratory CYP3A4 inhibition ([Liu 2017](https://pubmed.ncbi.nlm.nih.gov/28614988/)); higher drug levels could raise toxicity, such as kidney injury with tacrolimus. Drug-level monitoring after starting mitigates this for narrow-margin drugs.
- **CYP2D6-cleared drugs (metoprolol, a beta-blocker that slows heart rate; codeine; dextromethorphan):** Inferred from laboratory CYP2D6 inhibition ([Liu 2017](https://pubmed.ncbi.nlm.nih.gov/28614988/)); possible stronger metoprolol effects or altered codeine activation. Heart-rate and pain-control monitoring after starting mitigates this.
- **Acetaminophen and other CYP2E1 substrates (chlorzoxazone):** Inferred from laboratory CYP2E1 inhibition ([Liu 2017](https://pubmed.ncbi.nlm.nih.gov/28614988/)); acetaminophen liver toxicity could shift, in a direction unknown in people. Acetaminophen use within label doses limits this risk.
- **P-gp substrates (digoxin, dabigatran):** Inferred from rat studies in which DHM raised triptolide exposure ([Deng 2020](https://pubmed.ncbi.nlm.nih.gov/31092113/)) and P-gp blockade raised DHM absorption ([Huang 2018](https://pubmed.ncbi.nlm.nih.gov/28795912/)); possible digoxin toxicity (nausea, heart-rhythm disturbances) or bleeding with dabigatran. Digoxin-level monitoring mitigates this.
- **Benzodiazepines (sedative anti-anxiety drugs) and Z-drugs (non-benzodiazepine sleep drugs) (diazepam, alprazolam, zolpidem):** Inferred from benzodiazepine-site binding in rats ([Shen 2012](https://pubmed.ncbi.nlm.nih.gov/22219299/)); sedation could be deepened or blunted. Not driving until the combined effect is known mitigates this.
- **Glucose-lowering drugs (metformin; sulfonylureas, which make the pancreas release insulin, such as glipizide; insulin):** Additive glucose lowering inferred from trial glucose reductions ([Ran 2019](https://pubmed.ncbi.nlm.nih.gov/30089792/)); risk of hypoglycemia. More frequent glucose checks in the first weeks mitigate this.
- **Statins (cholesterol-lowering drugs such as atorvastatin) and red yeast rice:** Additive LDL lowering inferred from trial data ([Chen 2015](https://pubmed.ncbi.nlm.nih.gov/26032587/)), and shared CYP3A4 clearance for simvastatin and red yeast rice; possible higher statin exposure and muscle pain. Unexplained muscle aches are the warning sign.
- **Glucose-lowering supplements (berberine, chromium, cinnamon extract):** Additive blood sugar lowering; risk of hypoglycemia when stacked with diabetes drugs. Adding one at a time with fasting-glucose tracking mitigates this.
- **Sedating supplements and over-the-counter sleep aids (kava, valerian, diphenhydramine):** Additive calming effects inferred from GABA-A activity ([Carry 2021](https://pubmed.ncbi.nlm.nih.gov/34299083/)); possible daytime drowsiness. Not combining them before driving mitigates this.

**Populations who should avoid Dihydromyricetin:**

- **Pregnant or breastfeeding women:** Inferred from mechanism: DHM inhibits DNMT1, which maintains DNA methylation, in cultured cells ([Falckenhayn 2023](https://pubmed.ncbi.nlm.nih.gov/38500495/)), an effect untested in fetal development; no human data exist for DHM in pregnancy or lactation.
- **Chronic liver disease, including cirrhosis:** Inferred from mechanism: cirrhosis already slows drug clearance, and DHM inhibits CYP3A4, CYP2E1 and CYP2D6 in the laboratory ([Liu 2017](https://pubmed.ncbi.nlm.nih.gov/28614988/)); the longest DHM trial enrolled early disease only, median liver stiffness F0–F1 (no-to-mild scarring) ([Michailidou 2026](https://pubmed.ncbi.nlm.nih.gov/41612982/)).
- **Children and adolescents:** Inferred from the same untested DNMT1 inhibition in developing tissues ([Falckenhayn 2023](https://pubmed.ncbi.nlm.nih.gov/38500495/)); every DHM trial enrolled adults only, and no pediatric data exist.
- **Transplant recipients and others on narrow-margin CYP3A4 drugs:** Inferred from laboratory CYP3A4 inhibition; no human data.
- **People with alcohol use disorder using DHM instead of treatment:** Inferred from the risk-compensation concern; no human data show DHM treats alcohol use disorder.

  
## Risk Mitigation Strategies

Doses and timings below follow common practice unless cited.

- **Trial-dose ceiling:** 600 mg daily was studied for 3 months ([Chen 2015](https://pubmed.ncbi.nlm.nih.gov/26032587/)) and 12 months ([Michailidou 2026](https://pubmed.ncbi.nlm.nih.gov/41612982/); abstract states 300 mg/day, methods 600 mg/day); staying at trial doses limits exposure to untested dose-dependent liver or drug-interaction effects.
- **Baseline and 3-month liver panel:** ALT, AST and GGT measured before starting and after 3 months, with any rise from a normal baseline prompting medical review, guard against unrecognized liver injury.
- **Drug-level checks for narrow-margin medicines:** Prescriber rechecks of tacrolimus, cyclosporine or digoxin levels 1–2 weeks after starting catch CYP3A4- or P-gp-related rises before toxicity.
- **Glucose tracking with diabetes drugs:** Fasting glucose checks several times weekly for 2–4 weeks, with readings below 70 mg/dL, the hypoglycemia threshold ([ADA Standards of Care 2026](https://pubmed.ncbi.nlm.nih.gov/41358894/); members earn no revenue from it), prompting medical review, guard against additive hypoglycemia.
- **Pre-set drinking limits:** A drink count fixed before the occasion and kept unchanged with DHM prevents risk compensation and the organ damage a milder hangover hides.
- **No sedative combinations before driving:** Not combining DHM with benzodiazepines, Z-drugs, kava or sedating antihistamines before driving, until the combined effect is known, reduces additive-sedation risk.
- **Single-ingredient, dose-disclosed products:** Products stating DHM milligrams per serving prevent accidental overdosing or underdosing; 59% of DHM-containing hangover products did not report the dose ([Verster 2021](https://pubmed.ncbi.nlm.nih.gov/34225031/)).

  
## Therapeutic Protocol

Timing, cycling and titration steps below reflect common practice unless cited.

- **Metabolic and liver regimen:** 300 mg twice daily (600 mg/day) for 3 months in fatty liver disease ([Chen 2015](https://pubmed.ncbi.nlm.nih.gov/26032587/)).
- **Combination-tablet regimen:** Two 150 mg DHM tablets (with vitamins C and E and choline) twice daily, 600 mg/day for 12 months ([Michailidou 2026](https://pubmed.ncbi.nlm.nih.gov/41612982/)); the paper's abstract states 300 mg/day, but its methods give 600 mg/day.
- **Vine tea regimen:** 10 g of vine tea daily, delivering 970 mg DHM, for one month in type 2 diabetes ([Ran 2019](https://pubmed.ncbi.nlm.nih.gov/30089792/)).
- **Typical supplement range:** 300–1,000 mg daily ([LiverTox record](https://pubmed.ncbi.nlm.nih.gov/37643278/)).
- **Hangover-oriented approach:** Commercial blends such as Cheers Restore supply 1,200 mg DHM per serving around drinking ([Song 2026](https://pubmed.ncbi.nlm.nih.gov/42588049/)); this use was popularized by the Cheers brand ([McGill OSS article](https://www.mcgill.ca/oss/node/7153)) after rat research by Jing Liang's group ([Shen 2012](https://pubmed.ncbi.nlm.nih.gov/22219299/)).
- **Time of day:** The DHM-versus-metformin trial dosed before breakfast and dinner ([NCT03606694](https://clinicaltrials.gov/study/NCT03606694)), and both fatty liver trials dosed twice daily ([Chen 2015](https://pubmed.ncbi.nlm.nih.gov/26032587/); [Michailidou 2026](https://pubmed.ncbi.nlm.nih.gov/41612982/)); hangover products are taken around the drinking occasion.
- **Half-life:** No human half-life is published; in rats the oral half-life was 3.70 hours ([Liu 2017](https://pubmed.ncbi.nlm.nih.gov/27951743/)), consistent with short action.
- **Single versus split dose:** The short animal half-life and twice-daily trial regimens favor splitting the daily dose into two.
- **Genetic polymorphisms:** No pharmacogenetic dosing data exist; in a trial limited to ALDH2-heterozygous men, CYP2E1 variants may have modified the *Hovenia* hangover response ([Kim 2017](https://pubmed.ncbi.nlm.nih.gov/28750942/)), so hangover benefit may vary by genotype.
- **Sex:** No human sex-specific dosing data exist; lower oral exposure in female mice ([Carry 2021](https://pubmed.ncbi.nlm.nih.gov/34299083/)) has not been tested in people.
- **Age:** No dose data exist for adults over 65; starting at 300 mg daily and rising to the trial dose of 600 mg ([Chen 2015](https://pubmed.ncbi.nlm.nih.gov/26032587/)) after 2 weeks is a cautious approach for older adults taking several medicines.
- **Baseline biomarkers:** Benefits appeared in people with raised liver enzymes or glucose; those with normal baseline values have no trial evidence of further improvement.
- **Pre-existing conditions:** Studied populations were fatty liver disease and type 2 diabetes; people with other conditions are outside the tested evidence.

  
## Discontinuation & Cycling

- **Duration:** Trials lasted from 1 month ([Ran 2019](https://pubmed.ncbi.nlm.nih.gov/30089792/)) to 12 months ([Michailidou 2026](https://pubmed.ncbi.nlm.nih.gov/41612982/)); no data support lifelong use, so a defined trial period with reassessment fits the evidence.
- **Withdrawal effects:** None reported; no trial described rebound symptoms after stopping.
- **Tapering:** Not needed; DHM can be stopped at once, though drowsiness may shift in people also taking sedatives.
- **Cycling:** No evidence that cycling maintains efficacy; whether benefits persist after stopping is unknown.
- **Reassessment point:** Liver enzymes and glucose rechecked after 3 months show whether use is helping; no change suggests stopping.

  
## Sourcing and Quality

- **Source plant:** Purified DHM is mostly extracted from vine tea; *Hovenia dulcis* hot-water extracts are about 90% carbohydrate and contain only small amounts of DHM, according to the European Food Safety Authority (EFSA) ([EFSA opinion](https://pubmed.ncbi.nlm.nih.gov/32884582/)), so "raisin tree" products deliver far less.
- **Purity and testing:** Quality markers are a certificate of analysis stating DHM purity by HPLC (high-performance liquid chromatography, a standard purity test), heavy-metal testing and manufacture under good manufacturing practice.
- **Dose disclosure:** Many hangover blends hide doses in proprietary mixes; DHM doses were unreported in 59% of DHM-containing products ([Verster 2021](https://pubmed.ncbi.nlm.nih.gov/34225031/)).
- **Stability:** A review reports that DHM degrades with heat, alkaline pH and metal ions ([Liu 2019](https://pubmed.ncbi.nlm.nih.gov/32288229/)); cool, dry storage and not dissolving powder in hot or alkaline drinks preserve it.
- **Enhanced formulations:** A review describes delivery systems designed to raise DHM's stability and bioavailability ([Hou 2025](https://pubmed.ncbi.nlm.nih.gov/40306054/)), but no human outcome trial has tested them.
- **Products on the market:** Hepatrat (Uni-Pharma) was tested in a 12-month placebo-controlled trial ([Michailidou 2026](https://pubmed.ncbi.nlm.nih.gov/41612982/)); single-ingredient capsules are sold by brands such as Double Wood and Nootropics Depot, and Cheers sells multi-ingredient blends.

  
## Practical Considerations

- **Time to effect:** Blood glucose markers shifted within 1 month ([Ran 2019](https://pubmed.ncbi.nlm.nih.gov/30089792/)) and liver enzymes within 3 months ([Chen 2015](https://pubmed.ncbi.nlm.nih.gov/26032587/)); hangover effects, if any, occur within hours.
- **Common pitfalls:** Treating DHM as permission to drink more, buying blends with undisclosed doses, expecting liver-fat loss that trials did not show, and relying on DHM alone in place of weight loss and diet.
- **Regulatory status:** In the US, DHM is sold as a dietary supplement and is not approved for any disease ([LiverTox record](https://pubmed.ncbi.nlm.nih.gov/37643278/)). EFSA concluded that the safety of a *Hovenia dulcis* extract as a novel food was not established ([EFSA opinion](https://pubmed.ncbi.nlm.nih.gov/32884582/)).
- **Food status in China:** Vine tea leaves are listed as a new food resource under Chinese food safety standards, according to a 2023 review ([Zeng 2023](https://pubmed.ncbi.nlm.nih.gov/37343650/)).
- **Cost and access:** DHM is inexpensive and widely sold online, so cost is not a barrier.

  
## Interaction with Foundational Habits

- **Sleep:** Possibly potentiating, unproven. DHM acts on GABA-A receptors in animals ([Shen 2012](https://pubmed.ncbi.nlm.nih.gov/22219299/)), and a 288-person sleep trial of 200 mg daily is unpublished ([NCT05280561](https://clinicaltrials.gov/study/NCT05280561)). It does not undo alcohol's disruption of sleep, so limiting late-evening drinking matters more.
- **Nutrition:** Indirect, potentiating. In the 12-month fatty liver trial, patients whose liver enzymes normalized lost more weight, though DHM use was the only independent predictor ([Michailidou 2026](https://pubmed.ncbi.nlm.nih.gov/41612982/)), so a calorie-controlled, low-sugar diet may add to any DHM effect. No nutrient depletion is known; vine tea is a dietary source.
- **Exercise:** Possibly potentiating, unproven. DHM raised irisin, an exercise-released muscle hormone, in rats and humans ([Zhou 2015](https://pubmed.ncbi.nlm.nih.gov/26054747/)); no study shows blunted training adaptation or improved performance, and no timing rule around workouts exists.
- **Stress management:** No established interaction. DHM eased anxiety-like behavior in rats via GABA-A receptors ([Shen 2012](https://pubmed.ncbi.nlm.nih.gov/22219299/)), but no human cortisol or stress-outcome data have been published.

  
## Monitoring Protocol & Defining Success

Before starting, a baseline panel records liver enzymes, fasting glucose, HbA1c (3-month average blood sugar) and lipids, so later changes can be attributed and pre-existing disease identified. For people taking narrow-margin drugs such as tacrolimus or digoxin, the baseline also includes a drug level.

Ongoing monitoring follows this cadence: drug levels and glucose checks at 1–2 weeks for those on interacting medicines, then the full panel at 3 months, then every 6–12 months while use continues. Success means a measurable fall in ALT, GGT or fasting glucose from the person's own baseline by 3 months; no change by then suggests stopping. A rise in liver enzymes, recurrent low glucose readings, or new daytime drowsiness is a reason to stop and seek medical review.

| Biomarker | Optimal Functional Range | Why Measure It? | Context/Notes |
|-----------|--------------------------|-----------------|---------------|
| ALT | 7–55 U/L (standard reference range) | Expected to fall; also a liver safety check | Laboratory ranges vary; no fasting needed; avoid heavy exercise 48 hours before testing |
| GGT | 8–61 U/L (standard reference range) | Expected to fall | Alcohol raises GGT; record recent intake; pair with ALT and AST |
| Fasting glucose | 70–99 mg/dL (standard reference range) | Expected to fall; safety check with diabetes drugs | 8-hour fast; morning draw |
| HbA1c | 4.0–5.6% (standard reference range) | Expected to fall | Reflects about 3 months; fell from baseline only in the treated group of a 12-month DHM combination-tablet trial, not compared between groups ([Michailidou 2026](https://pubmed.ncbi.nlm.nih.gov/41612982/)) |
| LDL cholesterol | No established target; track change from own baseline | Expected to fall | Pair with apolipoprotein B; one trial lowered it, another did not |
| Cystatin C | 0.5–1.0 mg/L (standard reference range) | Expected to fall in type 2 diabetes | Less affected by muscle mass than creatinine |
| Drug trough level (tacrolimus, cyclosporine, digoxin) | No established target; compare with the level before starting DHM | Safety check | Trough means the lowest level, drawn just before the next dose; check 1–2 weeks after starting |

Qualitative markers:

- Hangover severity and next-day energy after a matched number of drinks
- Number of drinks per week, compared with the pre-DHM baseline
- Sleep quality and time to fall asleep
- Daytime drowsiness, especially when combined with other calming agents
- Digestive comfort

  
## Emerging Research

- **First-in-human dose and absorption study ([NCT05623501](https://clinicaltrials.gov/study/NCT05623501)):** Phase 1, 12 healthy volunteers, 300 or 900 mg with or without L-lysine; primary endpoints blood levels and adverse events. Estimated completion December 2025 has passed; status unknown, no results posted. Adequate blood levels would support systemic claims; negligible levels would weaken them.
- **DHM versus metformin in new type 2 diabetes ([NCT03606694](https://clinicaltrials.gov/study/NCT03606694)):** Phase 2, 24 participants, 300 mg twice daily for 12 weeks; primary endpoints glucose, HbA1c and insulin sensitivity. Status unknown since 2023, no results posted. A positive result would strengthen the glycemic item; a null result would weaken it.
- **Stress-related sleep ([NCT05280561](https://clinicaltrials.gov/study/NCT05280561)):** 288 adults, 200 mg daily for 20 days versus placebo; primary endpoints sleep duration and stress. Completed November 2022, no results posted. Positive results would add a sleep benefit; null results would leave sleep claims unsupported.
- **Multi-ingredient hangover powder ([NCT06575972](https://clinicaltrials.gov/study/NCT06575972)):** 60 adults, DHM-containing JoyRise powder versus placebo; primary endpoint hangover questionnaire. Completed August 2024, no results posted. A positive result would add industry-run support for blends; a null result would add to the conflicting hangover data.
- **Senotherapeutic (targeting aged, non-dividing cells) potential:** Cell and mouse work suggests DHM modulates senescent cells ([Xu 2026](https://pubmed.ncbi.nlm.nih.gov/41792133/)); human trials of senescence markers could move this from Speculative toward Low, while a failure to reach active tissue levels would weaken it.
- **Bioavailability formulations:** A review describes new bio-based delivery systems that aim to improve DHM's stability and bioavailability ([Hou 2025](https://pubmed.ncbi.nlm.nih.gov/40306054/)); human trials showing higher blood levels would strengthen systemic claims, while failure would confirm absorption as the limiting step.
- **Evidence weakening the alcohol case:** Null findings on hangover scores ([Paik 2024](https://pubmed.ncbi.nlm.nih.gov/39767027/)) and on alcohol metabolism ([Skotnicová 2020](https://pubmed.ncbi.nlm.nih.gov/33656905/)) set the bar for a definitive purified-DHM hangover trial.

  
## Conclusion

Dihydromyricetin is a natural plant compound from Chinese vine tea and the Japanese raisin tree, sold mainly as a hangover aid and increasingly promoted for liver, metabolic and healthy-aging benefits. For a health-focused adult, the most credible signals are modest improvements in liver blood tests and blood sugar markers in people who already have fatty liver disease or type 2 diabetes. These come from a handful of small, short trials, mostly from one research group, and the longest trial tested a combination tablet. The effect on cholesterol is inconsistent, and no trial has shown less liver fat.

The hangover evidence relies on raisin-tree extracts rather than the purified compound, and the results disagree. The aging-related findings, on aged cells, lifespan in worms and flies, and brain aging, come only from laboratory and animal work. The compound is poorly absorbed, and its behavior in the human body has not been measured.

Safety looks reassuring so far: trials reported no related side effects, and no liver injury has been linked to it. The open questions are possible changes to how the liver clears certain medicines, extra drowsiness with sedatives, and a temptation to drink more when hangovers feel milder. Much of the hangover and skin research was paid for or run by the companies selling the products, and one liver trial used tablets supplied by their maker.

Overall, the evidence is early and thin, with a favorable safety record but limited proof of benefit.

**[Top](#top) - [Benefits](#expected-benefits) - [Risks](#potential-risks--side-effects) - [Protocol](#therapeutic-protocol)**


