Hydrogen-rich water reduces liver fat accumulation and improves liver enzyme profiles in patients with non-alcoholic fatty liver disease: a randomized controlled pilot trial
Korovljev D, Stajer V, Ostojic J, LeBaron TW, Ostojic SM. · Clin Res Hepatol Gastroenterol. 2019;43(6):688-693.
Study at a glance
A twelve-person pilot reported less liver fat on MRI after four weeks of hydrogen water. Weight, blood sugar and blood fats did not improve, and liver-enzyme benefit was uncertain. The source also contains an unresolved percentage discrepancy.
Human
Mild-to-moderate NAFLD outpatients in Novi Sad, August–December 2018; seven women/five men, age 56.2 ± 10 years, BMI 37.7 ± 5.3 kg/m². The authors’ no-other-chronic-disease description coexists with Table 1 glucose 8.0 ± 3.9 mmol/L; diagnoses are not reclassified here.
1 L/day, 500 mL before breakfast and dinner, for 28 days per condition. Rejuvenation magnesium/magnesium-malate tablets in lukewarm water, measured by gas chromatography approximately five minutes after dissolution and immediately before consumption; 80 mg magnesium per 500 mL. · 3 mM dissolved H₂ versus below 0.1 mM control. Magnesium/water generation; no gas inhalation.
28 days per condition, fourteen-day washout.
Regional MRI fat signals were lower with hydrogen (P < 0.05). Reported liver-fat concentration was 284.0 ± 118.1 to 256.5 ± 108.3 mM. The paper also reports 2.9% change (95% CI 0.5–5.5), inconsistent with those raw concentrations. AST decline 10.0%, 95% CI −23.2 to 3.4, includes zero; an AST benefit is not established. Other liver enzymes were nonsignificant and within reference ranges; glucose/lipids were unaffected with data not shown. Weight, BMI and body composition were nonsignificant. Insulin 39.5% versus 20.5%, P = 0.15; leptin 10.5% versus 31.7%, P = 0.11; ghrelin 96.1% versus 34.7%, P = 0.21; all nonsignificant. Compliance 90.0% versus 93.3%, P = 0.38.
Twelve participants, short periods, no histology or long-term clinical outcomes, no severe-disease generalization, outlier-removal and outcome-reporting limitations, no clearly reported carryover/period analysis. The printed chemical reaction is not reproduced as a validated reaction.
What kind of evidence is this?
Human
Double-blind randomized placebo-controlled crossover pilot; all twelve completed both periods. Endpoint exclusions under the outlier rule were not enumerated.
Liver health
H₂-rich water
Information not yet classified
Some editorial classification fields are still pending. The source-reported outcomes and result are shown below; Hydrogenology does not infer a positive or negative signal from prose automatically.
Methods
Mild-to-moderate NAFLD outpatients in Novi Sad, August–December 2018; seven women/five men, age 56.2 ± 10 years, BMI 37.7 ± 5.3 kg/m². The authors’ no-other-chronic-disease description coexists with Table 1 glucose 8.0 ± 3.9 mmol/L; diagnoses are not reclassified here.
Twelve participants in crossover, not twelve new people per condition. All completed; endpoint-specific outlier exclusions not fully accounted for.
28 days per condition, fourteen-day washout.
1 L/day, 500 mL before breakfast and dinner, for 28 days per condition. Rejuvenation magnesium/magnesium-malate tablets in lukewarm water, measured by gas chromatography approximately five minutes after dissolution and immediately before consumption; 80 mg magnesium per 500 mL.
Hydrogen-rich water
3 mM dissolved H₂ versus below 0.1 mM control. Magnesium/water generation; no gas inhalation.
Tap water with non-H₂-producing magnesium tablets, same 80 mg magnesium per dose; appearance/taste matched with CO₂ bubbles. HRW Natural Health Products supplied both.
Outcomes and reported result
Primary: MRI liver-fat deposition; fasting biochemistry, body composition and adverse-event questionnaire.
Regional MRI fat signals were lower with hydrogen (P < 0.05). Reported liver-fat concentration was 284.0 ± 118.1 to 256.5 ± 108.3 mM. The paper also reports 2.9% change (95% CI 0.5–5.5), inconsistent with those raw concentrations. AST decline 10.0%, 95% CI −23.2 to 3.4, includes zero; an AST benefit is not established. Other liver enzymes were nonsignificant and within reference ranges; glucose/lipids were unaffected with data not shown. Weight, BMI and body composition were nonsignificant. Insulin 39.5% versus 20.5%, P = 0.15; leptin 10.5% versus 31.7%, P = 0.11; ghrelin 96.1% versus 34.7%, P = 0.21; all nonsignificant. Compliance 90.0% versus 93.3%, P = 0.38.
Main-text methods, results, limitations and declarations were reviewed. Source versions, missing information and conflicting findings are stated in this record.
No side effects reported in the short trial; this does not establish long-term safety.
A reported association, difference or mechanism is not automatically a clinical benefit.
Limitations and applicability
Twelve participants, short periods, no histology or long-term clinical outcomes, no severe-disease generalization, outlier-removal and outcome-reporting limitations, no clearly reported carryover/period analysis. The printed chemical reaction is not reproduced as a validated reaction.
Limited to the population/model and protocol described above. These findings do not provide a treatment, concentration or device recommendation.
Partial support: Serbian Ministry 175037; Provincial Secretariat 114-451-710; Faculty of Sport and Physical Education; HRW Natural Health Products Inc.
Authors declare no competing interests; the commercial tablet supplier also funded the study. T. W. LeBaron’s Molecular Hydrogen Institute affiliation is disclosed.
No single-study GRADE certainty rating is assigned. Read how records and evidence assessments are prepared.
Sources and record status
PMID: 30982748 · DOI: 10.1016/j.clinre.2019.03.008
Complete publisher-typeset ARTICLE IN PRESS, +Model CLINRE-1263, six pages, hosted by Aquacentrum (hydrogen-water equipment retailer). This is not the final paginated issue version.
AI-assisted editorial review of the available main full text, 29 October 2026. No independent clinical or statistical assessment has been performed.
Six article-in-press pages: Patients; Experimental protocol; statistics; Table 1; Figures 1–3; Results; Discussion; Disclosure and Acknowledgments.
AI-assisted editorial review of the available main full text, 29 October 2026. Complete publisher-typeset ARTICLE IN PRESS, +Model CLINRE-1263, six pages, hosted by Aquacentrum (hydrogen-water equipment retailer). This is not the final paginated issue version.
NCT03625362; ethics 0417-HRW/2018. This registration distinguishes the trial from other NAFLD publications.
29 October 2026
Help people check the facts before they buy.
Hydrogenology keeps 692 source-checked records accessible without advertising or product promotion.
Help us correct the recordReport a discrepancy
Send the exact difference between this page and the linked source. The study reference is attached automatically.