Abstract
CONTEXT: Metabolic disorders are a growing global concern, especially in developed countries, due to their increasing prevalence. Serum lipid profiles, including triglycerides (TG), total cholesterol (TC), high-density lipoprotein (HDL), and low-density lipoprotein (LDL), are commonly used clinical biomarkers for monitoring the progression of these metabolic abnormalities. In recent decades, hydrogen-rich water (HRW) has gained attention as a safe and effective treatment, with regulatory effects on lipid peroxidation and inflammatory responses in clinical trials. OBJECTIVES: This systematic review and meta-analysis aim to evaluate the effectiveness of HRW therapy on blood lipid profiles in randomized controlled trials (RCTs) for metabolic disorders. METHODS: Following the PRISMA guidelines, a search for RCT studies was conducted in the PubMed, Web of Science, Embase, and Google Scholar databases up to January 2024. Eight studies that met all eligibility criteria, including RCTs involving metabolic dysfunctions and evaluations of lipid profiles, were included for further analysis. Data extraction was followed by quality evaluation using the Jadad scoring system. Meta-analysis was conducted using STATA software. RESULTS: The eight selected RCTs included a total of 357 patients with various metabolic disorders. All studies showed either no or low risk of bias. The overall levels of TG [95% CI: -0.27 (-0.47, -0.07)], TC [95% CI: -0.07 (-0.32, -0.18)], and LDL [95% CI: -0.06 (-0.28, 0.15)] demonstrated slight decreases across the studies. However, there was some heterogeneity in HDL levels [95% CI: -0.11 (-0.37, 0.14)] among the studies (I² = 37.32%). Meta-regression analysis further indicated a positive association between the outcomes and the duration of the intervention as a moderating factor. CONCLUSIONS: Hydrogen-rich water demonstrated modest lipid-lowering effects in patients with metabolic disorders. However, due to the observed heterogeneity in HDL variations, further long-term trials involving larger populations are needed to clarify these inconsistencies.