Benefits
Traditional Korean root with its own coumarin profile
Korean angelica root (cham-dang-gwi) has a long history in Korean folk medicine, including use as a sedative. Its main compounds, decursin and decursinol angelate, are absent from Chinese dong quai (Angelica sinensis) and Japanese Angelica acutiloba, so studies of those plants do not carry over to this one.
Blood triglycerides in a 12-week placebo-controlled trial
In 100 Korean adults with fasting triglycerides of 130 to 200 mg/dL, 1 g a day of an Angelica gigas extract product (200 mg extract) for 12 weeks lowered triglycerides, VLDL cholesterol and the triglyceride to HDL ratio versus placebo. Triglycerides fell about 15 mg/dL while rising in the placebo group. These are blood markers from one government-funded trial.
Absorption and conversion of decursin compounds
In 20 healthy US adults, a single dose of an ethanol root extract supplying 119 mg decursin and 77 mg decursinol angelate was largely converted to decursinol, which peaked about 3 hours later. A Korean study of 4.6 g root extract powder in 10 men found faster peaks, shorter half-lives and much lower decursinol levels. These are blood level studies, not effects.
Urinary symptom scores in men taking a two-herb blend
In a 12-week placebo-controlled trial in 100 Korean men with urinary symptoms linked to an enlarged prostate, 600 mg a day of SHPro® (Angelica gigas and Astragalus Extract Blend), two parts Angelica gigas to one part astragalus, improved total symptom scores versus placebo at 12 weeks but not at the interim visit. The result cannot be split between the two herbs.
Pain and sleep behavior in rodent studies
In mice, oral decursinol reduced responses in heat, chemical and inflammatory pain tests in a dose-dependent way. In rodents, oral decursinol angelate reduced movement and lengthened pentobarbital-induced sleep, and in rat brain cells it boosted GABA-A signaling. These are animal results; no human pain or sleep trial of Angelica gigas alone was found.
Memory tasks in mice given an Angelica gigas extract
INM-176, the ethanol extract used in the US supplement Cogni.Q, reduced scopolamine- and amyloid beta-induced memory loss in mice and inhibited acetylcholinesterase in brain tissue. A Korean human trial was published only in Korean, and a 280-person trial of INM-176 completed in 2011 has no published results.
Mechanism of action
Conversion to decursinol
In US single-dose studies of an ethanol extract, decursin and decursinol angelate were absorbed and largely converted to decursinol, which reached far higher blood levels than the parent compounds; a Korean study of root extract powder found low decursinol levels. Lab enzyme work showed CYP2C19, then CYP3A4, metabolize both compounds, and an esterase (CES2) also breaks down decursin.
GABA-A signaling (animal and cell work)
In rat brain cells, decursinol angelate increased chloride flow and raised levels of GABA-A receptor subunits and the GABA-making enzyme GAD65/67, which fits its sedative-like effects in rodents.
Acetylcholinesterase inhibition (lab and mouse work)
The INM-176 extract inhibited acetylcholinesterase, the enzyme that breaks down the memory-related messenger acetylcholine, in mouse hippocampal tissue in the test tube and after dosing, and limited nerve cell damage in mice. This has not been measured in people.
Platelet activity (lab)
Decursin and decursinol angelate inhibited platelet aggregation triggered by arachidonic acid or a thromboxane mimetic in lab tests, which is relevant to the caution with blood thinners.
Clinical trials
Randomized, double-blind, placebo-controlled 12-week trial (NCT03079648) of capsules supplying 1 g a day (200 mg Angelica gigas extract), developed by Wellness Up Research Inc.; primary outcome was change in triglycerides. Funded by the Korean Rural Development Administration; authors declared no conflicts (Jung et al. 2020, Nutrients)
100 Korean adults with fasting triglycerides of 130 to 200 mg/dL; 93 completed (48 extract, 45 placebo).
Triglycerides fell by 14.8 mg/dL with the extract and rose by 19.9 mg/dL with placebo, a significant between-group difference (p < 0.05); VLDL cholesterol and the triglyceride to HDL ratio also improved versus placebo, while total, LDL and HDL cholesterol did not differ; free fatty acids rose slightly with the extract and fell with placebo, also a significant between-group difference. There were 17 mild or moderate adverse events with the extract and 10 with placebo (no significant difference), one serious event was judged unrelated, and lab safety measures did not change significantly. These are blood markers, not functional outcomes.
Randomized, double-blind, placebo-controlled 12-week trial of 600 mg a day (2 tablets) of a 2:1 mix of ethanol Angelica gigas extract and hot-water astragalus extract; primary outcome was the change in total prostate symptom score (IPSS). Funded by the Korean Rural Development Administration; products supplied by Suheung Co., and three authors worked at the Suheung Technology Research Institute; no registry number was given in the paper (Lee et al. 2025, Investig Clin Urol)
100 Korean men aged 40 to 75 with symptom scores of 8 to 19 related to an enlarged prostate; 84 analyzed per protocol (39 blend, 45 placebo).
The primary symptom score did not differ from placebo at the interim visit (p=0.0944) but did at 12 weeks (down 3.41 vs 1.36 points, p=0.0219); the incomplete emptying item and erectile function scores also improved versus placebo. Ten mild adverse events (3 with the blend, 7 with placebo) were judged unrelated; liver enzymes (AST, ALT) and LDL cholesterol rose more with the blend than with placebo but stayed within normal ranges. Because two herbs were given together, the result cannot be credited to Angelica gigas alone.
Open-label single oral dose pharmacokinetic study (NCT02114957) of the Angelica gigas ethanol extract supplement Cogni.Q, 4 capsules supplying 119 mg decursin and 77 mg decursinol angelate; NIH-supported, no efficacy outcomes (Zhang et al. 2015, PLoS One)
20 healthy adults (10 men, 10 women) in Texas.
Both compounds were absorbed and extensively converted to decursinol, which peaked at about 3.3 hours at concentrations far higher than the parent compounds. Half-lives were about 17 to 19 hours for decursin and decursinol angelate and 7.4 hours for decursinol; men absorbed the parent compounds faster.
Single oral dose pharmacokinetic study (KCT0001118) within an analytical method paper; 4.6 g of root extract powder, described as the usual dose approved by Korea's Ministry of Food and Drug Safety. Funded by the Korean Ministry of Health and Welfare; no efficacy outcomes (Kim et al. 2018, Molecules)
10 healthy fasting Korean men aged 23 to 25.
Decursin, decursinol angelate, nodakenin and decursinol all reached peak blood levels within about 0.3 to 0.7 hours, with half-lives of about 2.6 to 6.3 hours, shorter than in the US Cogni.Q study; the authors attributed the gap to different doses and formulations. Decursinol peaked far lower than decursin (0.92 vs 11.87 ng/mL).
Open-label phase 1 sequential single-dose escalation of Cogni.Q (INM-176 extract) at 800, 1,200, 1,600 and 2,000 mg, with no placebo; primary outcomes were ECG and blood safety tests. Sponsored by Penn State Milton S. Hershey Medical Center with the National Cancer Institute; results posted on the registry, no journal paper found (Milton S. Hershey Medical Center 2026, ClinicalTrials.gov NCT05375539)
12 men aged 40 or older with a history of prostate cancer, not on hormone therapy; 10 reached the 1,600 mg dose and 5 the 2,000 mg dose.
No deaths or serious adverse events. Other reported effects included digestive complaints, dizziness, restlessness, euphoria and low blood potassium; one man had a nonspecific ECG T-wave change at 2,000 mg, and two stopped after the first dose because of adverse events. One participant taking warfarin had his INR rise from 2 to 8 after the first 800 mg dose, and the registry states that people on warfarin should not take Angelica gigas supplements. Peak decursinol rose with dose.
Animal toxicology study following OECD test guidelines of an Angelica gigas extract, not the Decursinol-50 extract; gavage doses of 125, 250, 500, 1000 and 2000 mg/kg a day for 13 weeks, plus Ames, chromosome aberration and micronucleus tests. (Yun et al. 2015, Regul Toxicol Pharmacol)
Male and female rats.
No significant adverse effects on food intake, body weight, deaths, blood tests, organ weights or tissue exams at any dose; the no observed adverse effect level was above 2000 mg/kg a day, the highest dose tested. The genotoxicity tests were negative. No human data.