Kava and Anxiety: The Trial Evidence and the Liver Risk

What kava is

Kava is a drink made from the root of Piper methysticum, a Pacific Island plant used ceremonially for centuries. Its active compounds, kavalactones, produce a calm, relaxed state without the cognitive fog of alcohol.

The liver caution, up front. Kava has been linked to rare but serious liver injury, including cases of acute liver failure that needed a transplant. The US National Institutes of Health’s LiverTox database places it in its highest likelihood category, “A: well known cause of clinically apparent liver injury” (NIH LiverTox). Everything below about benefit has to be read alongside that. This article is educational: it does not recommend products, brands or how to take kava.

How kavalactones work, in plain words. A small group of kavalactones accounts for most of the activity. Laboratory studies suggest they enhance signalling at GABA-A receptors, the brain’s main braking system, and damp the ion channels that let neurons fire rapidly, though the mechanism is not settled. In the 6-week Australian trial, variants in a GABA-transporter gene tracked with response, which fits a GABA-centred mechanism (Sarris 2013); the 16-week trial did not replicate it (Sarris 2020). That same GABA action is why kava adds to the effects of alcohol, sedatives and sleep medicines. Why an overactive threat system responds to GABA signalling is covered in anxiety and the brain.

The evidence for anxiety

Kava has arguably the most placebo-controlled trial support of any herb for anxiety symptoms. A Cochrane systematic review of randomized trials found kava significantly better than placebo for anxiety symptoms, with modest effect sizes in mostly short trials (Pittler and Ernst 2003; NIH LiverTox overview); a later review of 12 studies focused on generalized anxiety disorder pooled only three placebo-controlled trials (130 people) and found effects in kava’s favour that did not reach statistical significance (Ooi 2018). Most trials used 60 to 250 mg of kavalactones per day.

The individual reviews and trials, with their limits:

• Cochrane review. 11 double-blind trials, 645 people, 1 to 24 weeks. Pooling the six that used the Hamilton Anxiety scale, kava beat placebo by 5.0 points (95% CI 1.1 to 8.8), and adverse events were “mild, transient and infrequent” (Pittler and Ernst 2003). All predate the first water-extract trial (2009) and used the commercial ethanol- or acetone-based extracts later implicated in liver cases (Sarris 2009; Teschke 2010).
• The 12-study review. Focused on generalized anxiety disorder, it found support in two placebo-controlled trials and one comparator trial, one negative trial, and a pooled estimate from three placebo-controlled trials that favoured kava (effect sizes 0.59 to 0.99) without reaching statistical significance; its authors called the evidence promising but insufficient (Ooi 2018). A separate 2018 analysis of responder rates across five trials (330 people) found kava 1.5 times as likely as placebo to produce a response, with adverse events no different from placebo (Smith and Leiras 2018).
• KADSS, 3 weeks. 60 adults with at least a month of elevated generalized anxiety took a water-extracted root preparation at 250 mg kavalactones a day in a crossover design. Hamilton scores fell 9.9 points on kava versus 0.8 on placebo in the first phase, a very large effect, with no liver signal over three weeks (Sarris 2009).
• GAD, 6 weeks. 58 analysed adults with diagnosed generalized anxiety disorder took an aqueous extract at 120 to 240 mg kavalactones a day. The effect was moderate (Cohen’s d 0.62); 26 percent remitted versus 6 percent on placebo; liver tests did not differ; headaches were more common (Sarris 2013).
• The largest and longest trial, 16 weeks. 171 unmedicated adults with generalized anxiety disorder took an aqueous root extract standardised to 120 mg kavalactones twice daily, or placebo. Kava was not better than placebo (the difference favoured placebo by 1.37 points, p = 0.25), remission was 17.4 percent versus 23.8 percent, memory complaints and tremor were more common on kava, and liver function test abnormalities were significantly more frequent in the kava group, although no participant met criteria for herb-induced liver injury (Sarris 2020).

The pattern: kava’s support comes mostly from short trials - the two positive Australian trials ran three and six weeks - in mixed or milder anxiety; the one large, 16-week trial in a diagnosed disorder was negative. Its authors’ own reading was that kava may act as a short-term aid for situational anxiety rather than a long-term option for generalized anxiety disorder.

The liver question, in context

Kava’s reputation took a hit after early-2000s reports of liver injury led several countries to restrict it. The fuller picture is more nuanced: most serious cases involved solvent (acetone or ethanol) extracts, use of leaves and stems rather than the root, very high doses, or people with pre-existing liver disease or heavy alcohol use. A 2007 World Health Organization review concluded liver reactions are rare and tied to those factors, and Germany later lifted its ban (NIH LiverTox). Still, the risk is not zero.

The numbers behind “rare”: between 50 and 100 cases of clinically apparent liver injury have been published, with some reviews counting more than 100. Injury has ranged from enzyme elevations to fulminant liver failure. Onset was typically 2 to 24 weeks after starting, and most cases resolved within 1 to 3 months of stopping (NIH LiverTox). Germany’s 2002 ban was overturned by a court in 2014. Case reviews also note that many patients were taking other drugs or supplements (Teschke 2010).

Two cautions against over-reassurance. A detailed pathogenesis review found cases with traditional water-based extracts as well as solvent extracts, and attributed most injury to daily overdose, prolonged use, poor-quality raw material and co-medication rather than to the solvent alone (Teschke 2010). And in the 16-week Australian trial, which used a water-extracted root, liver-test abnormalities were still more frequent on kava than placebo. “Water-based” lowers the risk; it does not remove it.

What lowers the risk, according to researchers and regulators

Researchers who reviewed the liver cases proposed a six-point standard: a “noble” cultivar at least five years old, peeled and dried root and rhizome only, water extraction, a proposed ceiling of 250 mg of kavalactones a day for medicinal products, rigorous trials, and enforced quality control in producing countries (Teschke, Sarris and Lebot 2011). In consumer terms that is the basis of the familiar advice:

• Prefer water-based extracts of the root only; check the label for kavalactone content and which plant part was used.
• Do not combine with alcohol, acetaminophen, or other liver-stressing drugs.
• Limit use to about 3 months without a medical review, and avoid it if you have liver disease, are pregnant, or take sedatives or other medications. The trials that found benefit ran three and six weeks, well inside that limit.

Warning signs of liver injury that call for stopping and seeing a doctor promptly: yellowing of the skin or eyes, dark urine, pale stools, unusual fatigue, nausea or pain under the right ribs.

Interactions and who should talk to a clinician

• Alcohol, benzodiazepines, sleep medicines, opioids and sedating antihistamines add to kava’s sedation and impair driving and coordination.
• Liver-stressing medicines, including acetaminophen (paracetamol), interact through shared burden on the liver. Laboratory studies suggest kavalactones inhibit several drug-clearing liver enzymes; human studies point to a modest effect on one enzyme, so the interaction picture is incomplete (NIH LiverTox).
• Heavy, long-term use has been associated with a scaly skin condition (kava dermopathy) (NIH LiverTox).

Talk to a clinician first if you have liver disease or abnormal liver tests, drink alcohol regularly, take any prescription medicine, are pregnant or breastfeeding, or have a diagnosed anxiety or depressive disorder, where the trial evidence does not support kava as a substitute for treatment.

Where the evidence is weak

Most positive trials were short, used different extracts, and several came from a single research group. The largest and longest trial was negative. There is no long-term safety data, only small comparisons with standard anxiety medicines, and no way to predict in advance who will develop liver injury. Other approaches studied for occasional stress are covered in ashwagandha and rhodiola and magnesium.

Bottom line

Kava is one of the few herbs with genuine trial support for anxiety. Used as the traditional water-extracted root, short-term, and without alcohol or liver-taxing drugs, its risk appears low - but the liver caveat is real, so loop in a clinician, especially if you take other medications. The best trials say short-term, situational relief is plausible and long-term use for a diagnosed disorder is not supported.