Effects of kudzu extract on alcohol consumption among men and women seeking treatment for alcohol use disorder Authors: Diana R Keith, Jing Liu, Gordana Vitaliano, Scott E Lukas Published: 2026 ABSTRACT AbstractBackgroundWe previously demonstrated that standardized kudzu extract (NPI-031) reduced drinking in a four-week trial among non-treatment-seeking men diagnosed with alcohol abuse or dependence. The present study assessed the safety and efficacy of eight weeks of kudzu among treatment-seeking men and women with alcohol use disorder (AUD).MethodsIn this double-blind placebo-controlled study, nineteen individuals (12 M, 7 F; 27–65 years old) with AUD took either kudzu extract (750 mg isoflavones) or matched placebo daily, combined with weekly counseling for eight weeks. Twice weekly laboratory visits monitored alcohol consumption (via daily diaries and timeline follow-back), medication adherence (via urinary riboflavin), safety (via blood chemistry), and adverse events.ResultsThere were no significant differences in drinking at baseline, with 4.2 ± 0.7 heavy drinking days (HDDs) per week reported by the kudzu group (N = 9; 5 M, 4 F) and 4.7 ± 0.6 in the placebo group (N = 10; 7 M, 3 F). Medication adherence was excellent (>96%) with no adverse events or safety concerns noted. After 4 weeks of treatment, both groups decreased the number of HDDs per week by ~40%. Across the full 8-week treatment, the average reduction was 30% for kudzu extract and 44% for placebo.ConclusionsIn contrast to results from prior studies among non-treatment-seeking young adult men, kudzu did not perform better than placebo in reducing drinking among a treatment-seeking sample of men and women with AUD. Taken together, these data suggest that kudzu may help to reduce binge drinking but may not provide additional benefits to individuals who are enrolled in an effective AUD treatment program.Keywords: Kudzu extract, Alcohol use disorder, Treatment, Alternative medicine, Alkontrol-herbal Introduction Despite the availability of several medications for alcohol use disorder (AUD) (Kranzler and Soyka, 2018), most patients with AUD do not achieve therapeutic goals, often due to barriers to accessing prescriptions, limited efficacy, and/or side effects that affect medication adherence (Anton et al., 2014, Lohoff, 2022, Muller et al., 2014). Indeed, in 2024, it was estimated that only 2.5% of individuals with AUD received medication-assisted treatment in the past year (SAMHSA, 2025a, SAMHSA, 2025b, SAMHSA, 2025c). Further, while evidence suggests moderate efficacy of naltrexone and acamprosate, recent meta-analyses found there were no measurable effects in 40%-70% of patients (Palpacuer et al., 2018, Ray et al., 2019, Rosner et al., 2010, Rosner et al., 2010, Srisurapanont and Jarusuraisin, 2005). Taken together, it is clear that novel AUD medications are sorely needed. One promising herbal preparation that may address this gap is extract of the kudzu plant, Pueraria lobata. Kudzu is indigenous to Japan and China and has been used for thousands of years in Chinese herbal medicine to reduce alcohol intoxication (Keung and Vallee, 1998). A small but growing body of scientific evidence supports kudzu as a potential medication for AUD (Benlhabib et al., 2004, Bracken et al., 2011, Lukas et al., 2005, Lukas et al., 2013, Penetar et al., 2011, Penetar et al., 2015). Kudzu extract has several benefits relative to other pharmacotherapies, including low incidence of side effects, rapid onset of efficacy, lack of subjective effects, and availability without a prescription (Lukas et al., 2013). The low incidence of side effects even after chronic treatment may translate to enhanced medication adherence, which remains a key challenge and could potentially translate to greater adherence with other aspects of treatment. Further, kudzu extract does not require a prescription, thereby reducing stigma, and the rapid onset of effects may increase the likelihood of adherence. Three small human studies have shown promising efficacy of kudzu extract (Lukas et al., 2005, Lukas et al., 2013, Penetar et al., 2015). In one laboratory study of acute effects, participants took standardized kudzu extract 2.5 h prior to a free drinking session (1.5 h) where they could consume up to six beers (Penetar et al., 2015). Analyses revealed that kudzu increased the length of time between drinks, resulting in a significant reduction of one drink consumed during each binge session. To determine the safety and efficacy of long-term use, a follow-up study quantified the effects of 4-weeks of kudzu in heavy drinking males who reported an average of five heavy-drinking-days (HDDs) per week but were not treatment-seeking (Lukas et al., 2013). Kudzu modestly reduced drinking by about one HDD per week, with no side effects or adverse events even among high rates of adherence (i.e., >90%). Interestingly, none of the participants reported a change in craving despite consuming fewer drinks. Taken together, these data suggest that kudzu extract reduces alcohol consumption, although it has primarily been studied in a young adult male cohort. These prior studies have focused on non-treatment-seeking moderate or binge drinkers who are typically younger and have not been drinking for many years. Thus, one goal of the current study was to focus on a treatment-seeking population and examine how kudzu extract may fit in as an adjunct to existing treatments for AUD. A second goal was to examine the safety and efficacy of eight weeks of treatment with kudzu extract among both men and women with AUD. Methods Participants A total of 19 adult female and male volunteers (7 female, 12 male) were recruited through online advertisements, in local and college newspapers, and flyers posted in the Boston area. Persons who responded to these advertisements were given a brief telephone screen and invited to the laboratory for further evaluations. After providing written informed consent, a psychiatric evaluation (Structured Clinical Interview for DSM-5 (First et al., 2015)) was performed and a physical exam, including electrocardiogram (ECG), vital signs, full hematology, blood chemistry, and urinalysis tests, was conducted. Volunteers 18–65 years old were enrolled if they drank a minimum of 10 drinks/week (an intermediate metric based on guidelines for men and women), met DSM-5 criteria for AUD and reported seeking treatment, a stable living situation, and if women of childbearing age agreed to use contraception while participating. Individuals were excluded if the screening physician determined detoxification was warranted or if they reported pregnancy, smoking > 10 cigarettes/day, heavy caffeine use (> 500 mg per day), unmanaged primary psychiatric disorders, past-year substance use disorder (excluding mild cannabis use disorder), BMI outside of 18–32 kg/m2, or medical contraindications (e.g., hepatitis or severe liver injury, head trauma, currently taking antipsychotic, antiseizure, or benzodiazepine medication, chronic neurological or infectious disease). Heavy smokers were excluded to minimize the impact of smoking on results due to recent trends towards lower smoking rates. The study was approved by the McLean Hospital Institutional Review Board and participants were paid for their participation. Medication Participant randomization, preparation of the kudzu extract (Alkontrol-Herbal™; NPI-031; Natural Pharmacia International, Inc., Burlington, MA) and placebo pills (650 mg gelatin/capsule, Nature’s Bounty), and quality control assessments were conducted as described in Lukas et al., (2013). Briefly, both the active and placebo groups took three pills, three times a day (e.g., morning, afternoon, and evening), with both groups receiving a riboflavin biomarker (25 mg) to monitor adherence and participants in active treatment receiving a total daily dose of 750 mg of isoflavones. Participants were asked to download an application to their phone to send them reminders to take their medication and refrain from using B-2 or B-complex vitamin supplements, which might interact with the riboflavin. Measures Medication adherence An ultraviolet (UV) light spectrophotometer was used to detect riboflavin in centrifuged urine samples (Del Boca et al., 1996) that were collected thrice weekly. Because riboflavin is cleared from the body quickly, fluorescing urine indicated compliance in the past 18–24 h. In addition, participants reported pill taking by checking time-of-day boxes on daily diaries. Urine samples were also assayed for puerarin (one active ingredient in kudzu) using a modified liquid chromatographic procedure developed for plasma (Ma et al., 2005). Urine puerarin concentrations were qualitatively analyzed by HPLC on a weekly basis starting on the second week of baseline and continuing through the follow-up visits. Alcohol consumption Participants were asked to come to the lab three times/week to complete questionnaires and provide urine, breath, and blood samples to monitor drug use, alcohol drinking, and liver function, respectively. Alcohol drinking was measured via three methods: 1) thrice weekly breathalyzer samples (AlcoSensor, Intoximeter, St Louis, MO), 2) thrice weekly timeline follow back (TFLB), and 3) daily diaries (take-home questionnaire). TLFB (Sobell and Sobell, 1992) consisted of using a calendar to help participants recall drinking behavior over the past several days. Daily diaries were completed by participants every morning upon awakening and were collected thrice weekly. Depression and anxiety Participants were assessed with the Beck Depression Inventory - II (BDI-2; (Beck et al., 1996)), and the Beck Anxiety Inventory (BAI; (Beck and Steer, 1993)) once per week during the baseline, treatment, and follow-up phases. Experimental procedure Baseline phase During a 2-week Baseline Phase, participants wore an ActiWatch device (ActiWatch. Score, MiniMitter Co., OR) on their non-dominant wrist which tracked sleep/wake cycles (data not shown), completed daily diaries, and attended 3 visits/week where they provided urine and breath samples and filled out questionnaires. Treatment phase In the 8-week Treatment Phase, participants continued the above procedures but also took kudzu or placebo capsules as directed. In addition, medical management sessions were scheduled once a week. Medical Management is a manualized treatment deriving from a number of empirically tested manualized therapies (Pettinati et al., 2000) designed to approximate a primary care approach to alcohol dependence (Volpicelli et al., 1997). The initial session lasted 40–60 min and involved reviewing the AUD diagnosis and negative consequences from drinking, a recommendation to abstain, medication information, strategies to enhance medication adherence, and referral to support groups such as Alcoholics Anonymous. In subsequent weekly visits (20–30 min), there was an assessment of recent drinking patterns, overall daily functioning, medication adherence, and side effects. Follow-up phase The 4-week Follow-up Phase was structured like previous lab visits, except participants came in once/week. Safety, side effects, and adverse event reporting The safety of kudzu extract was monitored via several methods including 1) twice weekly assessments of vital signs (resting pulse, blood pressure), and collecting adverse events and side effects data from 2) daily diaries and 3) thrice weekly in-lab assessments. Laboratory assessments utilized an industry-standard adverse event reporting form and a standard side effect checklist with 32 items addressing a wide range of somatic complaints. Participants also provided a blood sample at screening and during treatment for laboratory testing including: hematology, blood chemistry, renal function, and liver function. In addition, participants were required to have an ECG performed at the screening visit and at the end of treatment and follow up phases. Data analysis Data from the TLFB was the primary source for analyses while the daily diary served as a backup to verify daily totals and fill in for sporadic missing data. Data due to dropouts were handled as missing data and no values were entered. The primary efficacy endpoint was the percent change in heavy drinking days (HDD) during the treatment phase (weeks 3–10) and follow-up phase (weeks 11–14) relative to the baseline phase (weeks 1–2), with anxiety and depression as secondary outcomes. Following NIAAA guidelines, HDDs were defined as days where 5 or more standard drinks are consumed for men, or 4 or more for women. The primary data analysis was an intent-to-treat analysis, which included all randomized subjects. A general linear model ANOVA for data with repeated measurements were used to analyze percent change in HDD. Treatment (kudzu or placebo) was analyzed as a fixed effect and time (weeks since randomization) as a repeated-measurements effect. In all analyses, biological sex, age, and baseline HDD per week were modeled as covariates. All tests were two-sided with significance level of 0.05 and were performed using SPSS software (version 19.0, IBM SPSS, Inc, Chicago, IL). Results Participant flow, completion, and demographics Of the 148 individuals who responded to advertisements, 28 were found eligible after telephone screening. Following an in-person screening, nine individuals were either not eligible because of medical reasons, they could not be contacted, or they declined to participate. A total of 19 individuals were randomized to receive kudzu extract (n = 9; 5 M, 4 F) or placebo (n = 10; 7 M, 3 F). A research pharmacist (not affiliated with the study) stratified the participants on the basis of treatment group (A or B) and sex assigned at birth. Two participants in each group lost to follow-up before the end of the 8-week treatment phase. In the placebo group, one female had missing data starting in Week 3 and one male provided no data for the final week of treatment. In the kudzu group, one male had missing data starting in Week 4 and one male had missing data starting in Week 6. The demographic profiles of the participants in the two groups are provided in Table 1. There were no significant differences among any of these variables. Table 1Demographic profiles of participants at baseline1.VariableKudzu ExtractPlacebop-valueN (# male)9 (5 male)10 (7 male)0.515Age (years)42.9 ± 14.542.2 ± 12.90.319Body Mass Index (BMI)24.7 ± 3.527.1 ± 2.90.329Race (Ethnicity)7 Caucasian;1 African American;1 Uknown(1 Hispanic)8 Caucasian;1 African American;1 Asian0.568AUD severity(mild/mod/severe)3 / 5 / 12 / 4 / 40.356Current alcohol(# drinks/week)25.7 ± 8.622.1 ± 8.60.386Age of first drink15.9 ± 1.815.1 ± 3.60.555Years drinking regularly23.3 ± 14.722.2 ± 13.00.865No. of Heavy Drinking Days(HDD per week)4.2 ± 2.84.7 ± 1.50.547Tobacco smokers00N/ACannabis use frequency(daily/monthly/yearly or less)1 / 2 / 62 / 2 / 60.8691Values are means ± SD Demographic profiles of participants at baseline1. 1Values are means ± SD Medication adherence A total of 186 urine samples were collected from the kudzu-treated participants, with 96% (n = 178) positive for riboflavin. Of note, half of the urines that tested negative for riboflavin (n = 4) were collected in the first week of treatment/medication. Among the placebo-treated group, 100% of the urine samples were positive for riboflavin (n = 154/154). As kudzu extract is widely available as an over-the-counter (OTC) supplement, the urine of all participants was tested for puerarin up to three times per week (minimum of 8 and a maximum of 24 tests) to confirm medication adherence, such that participants in the active group were expected to test positive and those in the placebo group test negative. Of note, puerarin data from two participants in the kudzu group and three participants in the placebo group were lost/missing. A total of 132 urine samples from the kudzu group were tested, with 88.6% positive for puerarin (n = 117/132). On average, each participant in the active group had two negative urine samples out of 19 samples collected across 8 weeks (mean: 2.4 urine samples/week). For the placebo group, 147 urine samples were tested, with 15.6% positive for puerarin (n = 23/147). Participants in the placebo group provided a mean of 2.7 urine samples/week. A review of the data revealed this rate was driven by a single participant that tested positive for puerarin in 22 out of 22 samples. Of note, statistical analyses were conducted with and without this participant, with no difference in results. Safety, side effects, and adverse events Blood chemistry results for hematology and renal function fell within the normal range for all participants during all phases of the study (data not shown). In contrast, there were significant differences at baseline between the two groups in some measures of liver functioning (i.e., AST, ALT, and GGT), with values within the normal range for participants in the placebo group and elevated in the kudzu group (Table 2). However, all values across the three phases of the present study were within the expected range for individuals presenting with AUD (Magzoub et al., 2024). Table 2Hepatic Functioning.GroupKudzu ExtractPlaceboBlood Test(Normal Range)ScreenWeek 2Week 6Week 8ScreenWeek 2Week 6Week 8Protein, total(6.1–8.1 g/dL)7.37.17.17.27.17.17.07.1Albumin(3.6–5.1 g/dL)4.74.64.64.64.64.54.44.4Globulin(1.9–3.7 g/dL)2.62.52.52.62.52.62.62.7Albumin/Globulin Ratio(1.0–2.5)1.91.91.81.81.91.71.71.6Bilirubin, total(0.2–1.2 mg/dL)0.60.60.70.60.60.70.90.8Bilirubin, direct(≤0.2 mg/dL)0.10.10.20.20.10.20.20.2Bilirubin, indirect(0.2–1.2 mg/dL)0.50.50.60.40.50.50.70.7Alkaline Phosphatase(33–115 U/L)78.170.867.872.257.569.054.259.8AST2(10–30 U/L)54.0*45.0*46.3*60.6*21.821.424.226.0ALT3(6–29 U/L)64.8*56.3*60.8*80.4*20.518.619.817.8GGT4(3 – 40 U/L)82.8*85.6*95.6*112.8*30.945.3*38.032.6*Indicates values are above the normal range. 1Values are means; 2Aspartate Aminotransferase (AST); 3Alanine aminotransferase (ALT); 4Gamma-glutamyl transferase (GGT). Hepatic Functioning. *Indicates values are above the normal range. 1Values are means; 2Aspartate Aminotransferase (AST); 3Alanine aminotransferase (ALT); 4Gamma-glutamyl transferase (GGT). There were no adverse events related to study medication. The overall incidence of side effects in the kudzu extract group was 2.1%, with a range from 0% to 7% across participants. For the placebo group, the incidence of side effects was 1.7%, with a range of 0–6% across participants. Side effects data are displayed in Table 3. The most commonly reported side effect among both groups was ”Other” (for details, see Table 4, Table 5). Table 3Side Effects.Kudzu Extract1Placebo2SymptomPercentCountPercentCountDry Mouth0.6511.092Drooling or too much saliva0.0000.000Lost weight0.0000.000Gained weight0.6510.541Constipation8.44130.000Diarrhea0.6518.7016Nausea, vomitting1.3021.092Stomach pain, cramping0.6511.633Increased thirst or hunger0.6511.092Trouble urinating0.0000.000Urinating frequently3.2550.000Blurred vision0.0000.000Problems with sexual function0.0000.000Sweating, increased body temperature3.9060.000Changes in menstrual periods0.6510.000Increased heart rate or blood pressure1.9530.000Headache1.3022.174Lightheaded, Dizziness3.2551.092Sleepiness, sleeping a lot0.0001.633Trouble getting to or staying asleep7.79123.807Feeling restless or jittery0.6510.541Muscle stiffness1.9530.541Slowness, trouble getting moving0.0000.000Shaking or muscle trembling2.6041.092Memory problems, forgetfulness1.3020.000Low energy, easily tired2.6041.092Finding it hard to concentrate0.6510.541No motivation, loss of interest in usual things0.6511.092Anxiety4.5571.633Depression0.0001.092Agitation0.6510.541Other16.23259.78181Percent and raw number of times each symptom was endorsed across 154 visits by 9 participants in the kudzu extract group; 2Percent and raw number of times each symptom was endorsed across 184 visits by 10 participants in the placebo group Side Effects. 1Percent and raw number of times each symptom was endorsed across 154 visits by 9 participants in the kudzu extract group; 2Percent and raw number of times each symptom was endorsed across 184 visits by 10 participants in the placebo group Table 4Side Effects, Ranked by Frequency.PLACEBOSymptomPercent(Ranked) 1Other9.782Diarrhea8.703Trouble getting to or staying asleep3.804Headache2.175Stomach pain, cramping1.63KUDZU EXTRACTSymptomPercent(Ranked) 1Other16.232Constipation8.443Trouble getting to or staying asleep7.794Anxiety4.555Sweating, increased body temperature3.90 Side Effects, Ranked by Frequency. Table 5Summary and Details on “Other” Side Effects1.PLACEBONumber of times reported2Number of participants3Percent of “Other”4Total reports of“Other” symptoms186100%Sinus/Cold/Flusymptoms8444%Face flushing5128%Tinnitus or earache2211%Difficultly/ fallingwhile walking2211%Back pain115%KUDZU EXTRACTNumber of times reported2Number of participants3Percent of “Other”4Total reports of“Other” symptoms255100%Pain/Sorenessfrom car accident515160%Sinus/Cold/Flusymptoms4216%Weird taste3312%Stronger effects ofalcohol3112%Numbness in hand114%1Summary and details on the specific symptoms reported by participants under the “Other” category (free response) of Side Effects. 2The total number of times each symptom was reported under the “Other” category. 3 The number of participants who reported this symptom under the “Other” category. 4The number of times a specific symptom was reported under the “Other” category divided by the total number of times “Other” symptoms were reported in each group. 5A single participant reported they were involved in a car accident that occurred in Week 1 of active treatment. Summary and Details on “Other” Side Effects1. 1Summary and details on the specific symptoms reported by participants under the “Other” category (free response) of Side Effects. 2The total number of times each symptom was reported under the “Other” category. 3 The number of participants who reported this symptom under the “Other” category. 4The number of times a specific symptom was reported under the “Other” category divided by the total number of times “Other” symptoms were reported in each group. 5A single participant reported they were involved in a car accident that occurred in Week 1 of active treatment. Alcohol consumption- baseline During the two-week baseline phase, the average number of HDDs per week was 4.70 ± 0.58 (SEM) for the placebo group and 4.16 ± 0.65 (SEM) for the kudzu-treated group. These values were not statistically different (p > 0.5). Each participant was also asked to provide a breath sample at a total of 6 laboratory visits (i.e., 3/week for 2 weeks). The kudzu group (N = 9) provided 48 of a possible 54 samples (i.e., 89% attendance) with 4 breath samples positive for alcohol (i.e., 8%; mean BAC=0.06; range=0.035–0.092) and the placebo group (N = 10) provided 58 of a possible 60 samples (i.e., 97% attendance) with 7 positive samples (i.e., 12%; mean BAC=0.06; range=0.011–0.059). Alcohol consumption- treatment Fig. 1 displays the percent change in HDD relative to the baseline phase for both kudzu- and placebo-treated groups. Percent Change in Heavy Drinking Days (HDDS) Relative to Baseline by Treatment Group. Mean percent change in HDDs across the eight treatment weeks and Week 1 and 4 of the follow-up phase among the placebo (open square) and kudzu extract (filled squares) treatment groups. Error bars represent ±1 standard error of the mean (SEM). There were no significant main effects of treatment group or week, or significant interactions between the two, although the groups began to diverge starting in Week 6 of treatment. There were no significant main effects of treatment group (F(1, 117)= 2.469, p = 0.119, η2 partial = 0.085) or week (F(7, 117)= 0.235, p = 0.976, η2 partial = 0.034), and no significant interaction between treatment and week (F(7117)= 0.658, p = 0.707, η2 partial = 0.082). Age (F(1117)= 0.087, p = 0.768, η2 partial = 0.007) and sex (F(1117)= 3.877, p = 0.051, η2 partial = 0.026) were also not significant, although there was a trend for differences according to sex. After 4 weeks of treatment, both groups decreased the number of HDDs per week by approximately 40% (i.e., from 4.5 HDDs/week at baseline to 2.5–3.0 HDDs/week during treatment). As seen in Fig. 1, the two groups started diverging around Week 6, with the placebo group maintaining approximately 40–50% reduction in HDDs/week while the kudzu-treated group exhibited a smaller, approximately 20% reduction. Averaged across the full 8-week treatment, the reduction in HDDs was 30% for kudzu and 44% for placebo (i.e., reduction of 1.26 HDDs for kudzu and 2.07 HDDs for the placebo). Participants were asked to provide a breath sample at a total of 24 laboratory visits (i.e., 3/week for 8 weeks). The kudzu group (N = 9) provided 153 of a possible 216 samples (i.e., 71% attendance) with 14 breath samples positive for alcohol (i.e., 9%; mean BAC=0.066; range=0.010–0.171) and the placebo group (N = 10) provided 189 of a possible 240 samples (i.e., 79% attendance) with 7 positive samples (i.e., 4%; mean BAC=0.06; range=0.007–0.175). Alcohol consumption- follow-up In the first week of follow-up, the placebo group maintained an approximately 50% reduction in HDDs while the kudzu-treated group exhibited a reduction of approximately 20% (Fig. 1). By the fourth week of follow-up, both groups exhibited approximately 30% reduction in HDDs. Participants were also asked to provide a breath sample at a total of 4 laboratory visits (i.e., 1/week for 4 weeks). The kudzu group (N = 9) provided 22 of a possible 36 samples (i.e., 61% attendance) with 3 breath samples positive for alcohol (i.e., 14%; mean BAC=0.025; range=0.016–0.037) and the placebo group (N = 10) provided 30 of a possible 40 samples (i.e., 75% attendance) with 2 positive samples (i.e., 7%; mean BAC=0.021; range=0.010–0.031). Other effects- depression and anxiety BDI and BAI scores remained unchanged during treatment for both the kudzu extract and placebo groups. The BDI for the kudzu extract-treated participants was 10.13 ± 4.57 during baseline and 7.00 ± 3.65 during treatment. For placebo-treated participants, BDI measures were 7.89 ± 6.41 at baseline and 7.10 ± 6.46 during treatment. A similar pattern was observed for BAI. Kudzu extract-treated participants reported an average BAI of 7.23 ± 6.71 at baseline and 4.05 ± 1.86 during treatment. In contrast, the placebo group had a BAI of 5.74 ± 4.51 at baseline and 4.01 ± 3.50 during treatment. Discussion As kudzu extract showed promising results in a 4-week clinical trial among non-treatment-seeking young adult men (Lukas et al., 2013), the primary goal of the current study was to examine how it may fit in as an adjunct to 8-weeks of medication management in a treatment-seeking population of men and women with AUD. In contrast to our prior results, kudzu extract did not perform better than placebo in reducing drinking among this treatment-seeking sample of men and women with AUD. We previously demonstrated in double-blind, placebo-controlled experiments that kudzu extract decreased alcohol intake in heavy-drinking males who were not seeking treatment, both in a simulated binge-drinking paradigm in the laboratory (Penetar et al., 2015) and in a four-week outpatient setting (Lukas et al., 2013). Importantly, results from these prior studies showed modest reductions in drinking, amounting to a reduction of approximately one drink per binge drinking session in the laboratory and one HDD per week during the four-week outpatient trial. These data are consistent with the present results indicating a reduction of 1.26 HDDs/week among the kudzu group. However, the strong placebo response (i.e., reduction of 2.07 HDDs/week) in the present study overshadowed any potential effect of kudzu. This inconsistency in results may reflect the higher motivation of treatment-seeking individuals, or alternatively, the efficacy of the primary care intervention, which the addition of kudzu did not improve. Baseline differences in biomarkers of hepatic functioning also suggest that individuals in the kudzu group may have drank more heavily than those in the placebo group at the onset of the trial, which could also help explain the inconsistent results. Taken together, these data may suggest that kudzu extract is effective in reducing binge drinking, particularly among young adult men, but may not provide any additional benefits in standard primary care approaches for AUD. However, further studies may be warranted as the present study included a small sample size. Of note, several other factors were different between current and prior studies, which may partially account for differences in results. Relative to the prior outpatient trial, participants in the present trial included both men and women, were significantly older (mean age of 45 vs. 24 years old), and had been drinking regularly for a much longer period (mean of 23 vs. 7 years). It is important to note that sex differences in drinking outcomes were very close to reaching significance (p = 0.051), which may have a played a role in the inconsistent results. In particular, dropouts over the course of the trial changed the proportion of women in each group over time, which may be related to the unexpected divergence between the groups observed in Treatment Week 6 (BL placebo 3 F, 7 M; kudzu 4 F, 5 M; Week 6 placebo: 2 F, 7 M; kudzu: 4 F, 3 M). However, given the small sample sizes, it is difficult to speculate about the role of sex differences at this early stage of research. Women were excluded in several of our prior studies utilizing longer-term chronic dosing (Lukas et al., 2013, Penetar et al., 2015) due to an abundance of caution regarding preclinical data suggesting weak estrogenic activity of diadzin, one of the active isoflavones in kudzu (Amer et al., 2010). However, one of our first pilot studies showing promising results of kudzu extract included three women who served as their own controls in a crossover study design (n = 14) where they received 7 days of kudzu extract or placebo prior to participating in a simulated binge drinking paradigm (Lukas et al., 2005). Of note, participants in this prior study did not meet criteria for alcohol dependence and were not motivated to reduce their drinking. While neither the prior nor the present study was powered to detect sex differences, taken together, these data suggest it may be important for future studies to more fully examine the role of sex in the response to kudzu among both individuals who drink heavily and those with AUD. A second goal of the present study was to examine the safety of eight weeks of kudzu treatment. We previously demonstrated that four-weeks of kudzu treatment produced minimal side effects and no changes in liver or renal function among a population of healthy, young adult, heavy-drinking males. The present data replicate and extend these results by demonstrating the safety profile of extended use of kudzu extract among an AUD population that has been drinking heavily for many years. While participants randomized to the kudzu-treated group exhibited elevated values of AST, ALT, and GGT (indicating worse hepatic functioning) at baseline relative to the placebo group, these values tended to improve when alcohol drinking was reduced. The one exception was gamma-glutamyl transferase (GGT), a biomarker for alcohol consumption and liver function, which slowly increased over the course of the 12-week study among the kudzu-treated group, from 82 to 112 U/L. However, recent studies among individuals who drink heavily have observed mean GGT concentrations of 125 ± 109 U/L (Magzoub et al., 2024), suggesting these values represent a commonly observed, low to moderate GGT elevation. All other vital signs, measures of blood chemistry, and liver and renal functioning tests were within normal ranges. It is important to note that the present study had several limitations including 1) a 3 pills, 3 times/day daily dosing interval, which might not be optimal; 2) a small sample size; and 3) the lack of power to detect sex differences. While we observed high rates of medication adherence, presumably due in part to dosing reminders and the lack of side effects, it is possible that the requirement of morning, afternoon, and evening doses may be difficult for individuals to remember on an ongoing basis. However, as an OTC supplement, kudzu extract is highly accessible and does not require a visit to the doctor or a prescription, which could reduce barriers to accessing treatment. Further, the increased privacy and decreased stigma associated with OTC medications may increase its appeal to those who drink heavily in social situations but are on the threshold of meeting criteria for AUD. Indeed, individuals who drink heavily but are on the edge of dependence represent the highest percentage of those with problematic drinking in the US (Esser et al., 2014), and thus kudzu extract may be an important option for those who wish to reduce their drinking but are not yet ready to seek formal treatment. In summary, while significant reductions in drinking were observed in prior trials, the present results suggest that kudzu may not reduce drinking in highly motivated individuals seeking treatment for AUD. Taken together, these studies suggest that kudzu extract fills an important niche to help reduce binge drinking among individuals that drink heavily. CRediT authorship contribution statement Lukas, Scott E: Writing – review & editing, Supervision, Project administration, Methodology, Investigation, Funding acquisition, Conceptualization. Vitiliano, Gordana: Writing – review & editing, Supervision, Project administration, Methodology, Investigation. Liu, Jing: Writing – review & editing, Validation, Methodology, Formal analysis. Keith, Diana R: Writing – review & editing, Writing – original draft, Formal analysis, Data curation. Ethics declaration Written informed consent to take part in the study and to publish the article has been obtained from all participants or their legal representatives. The privacy rights of participants have been observed. This study was performed in compliance with relevant laws, regulatory frameworks and guidelines where the research took place. This study was approved by the McLean IRB. (Approval No. 2017P000109; 2023P001365) The results of this clinical trial and any associated work have been posted in a registry. This clinical trial was registered with number NCT03099590 (Clinicaltrials.gov). Funding sources This research was supported by the Norman E. Zinberg Fellowship in Addiction Psychiatry Research and a Livingston Award from the Department of Psychiatry/Harvard Medical School (DRK) and joint NCCIH and NIAAA Grant R44AA022577 (SEL). Prior to the award of the STTR Grant and the fellowships, McLean Hospital Corporation negotiated a licensing agreement for kudzu extract with Natural Pharmacia International (NPI), Inc., Burlington, MA. None of the authors are affiliated with NPI, Inc. Declaration of Competing Interest All authors declare that they have no conflicts of interest. Prior to receiving funding to conduct this study, McLean Hospital Corporation negotiated a licensing agreement for kudzu extract with Natural Pharmacia International (NPI), Inc., Burlington, MA. None of the authors are affiliated with NPI, Inc. Acknowledgements We thank and acknowledge the late David W.Y. Lee, PhD, for his important contributions and collaboration in the development and grant funding of this study. Dr. Lee oversaw the preparation of the kudzu extract via working with NPI, Inc., but he was not involved in conduct, data collection or analysis of the clinical study. References