Hydrogen-rich water decreases serum LDL-cholesterol levels and improves HDL function in patients with potential metabolic syndrome.

We have found that hydrogen (dihydrogen; H2) has beneficial lipid-lowering effects in high-fat diet-fed Syrian golden hamsters. The objective of this study was to characterize the effects of H2-rich water (0.9-1.0 l/day) on the content, composition, and biological activities of serum lipoproteins on 20 patients with potential metabolic syndrome. Serum analysis showed that consumption of H2-rich water for 10 weeks resulted in decreased serum total-cholesterol (TC) and LDL-cholesterol (LDL-C) levels. Western blot analysis revealed a marked decrease of apolipoprotein (apo)B100 and apoE in serum. In addition, we found H2 significantly improved HDL functionality assessed in four independent ways, namely, i) protection against LDL oxidation, ii) inhibition of tumor necrosis factor (TNF)-α-induced monocyte adhesion to endothelial cells, iii) stimulation of cholesterol efflux from macrophage foam cells, and iv) protection of endothelial cells from TNF-α-induced apoptosis. Further, we found consumption of H2-rich water resulted in an increase in antioxidant enzyme superoxide dismutase and a decrease in thiobarbituric acid-reactive substances in whole serum and LDL. In conclusion, supplementation with H2-rich water seems to decrease serum LDL-C and apoB levels, improve dyslipidemia-injured HDL functions, and reduce oxidative stress, and it may have a beneficial role in prevention of potential metabolic syndrome.

consume 450-500 ml of H 2 -rich water two times/day for a total minimum consumption of 900 ml to a maximum consumption of 1,000 ml. The blood samples were collected at baseline (0 week) and after 10 weeks of drinking H 2 water.

Serum analysis
Serum lipids. Blood samples were obtained in the morning after an overnight fast. Serum glucose levels were measured by the glucose oxidase method. Serum TC, HDL-C, LDL-C, and triacylglycerols (TG) were measured by enzymatic methods on a chemical autoanalyzer (Hitachi Co, Tokyo, Japan). Lipoprotein profi les were obtained by fast-protein liquid chromatography (FPLC) using Superpose 6 10/300GL column ( 17 ). Briefl y, 100 l of fasting serum was applied to a Superose 6 column, and the samples were eluted in a mobile phase (0.15 M NaCl, 0.01% NaN 3 , and 2 mM EDTA, pH 7.5) at a rate of 0.3 ml/min in 60 fractions of 500 l. Lipid composition of fractions corresponding to VLDL, LDL, and HDL were quantifi ed by enzymatic assays using commercially available kits for TC (BioSino, Beijing, China).
Measurement of serum oxidative stress and oxidizability. Serum levels of malondialdehyde (MDA), a marker for oxidative stress, were determined by a spectrophotometric measurement of thiobarbituric acid-reactive substances (TBARS) according to the manufacturer's instructions (Nanjing Jiancheng Biochemistry, China). The activity of superoxide dismutase (SOD), which acts as antioxidant and protects cellular components from being oxidized by reactive oxygen species, was measured by a commercial kit (Nanjing Jiancheng Biochemistry, China) according to the manufacturer's instructions. The activity of paraoxonase-1 (PON1), an antioxidant enzyme associated with HDL, was measured by adding serum to 1 ml of Tris-HCl buffer (100 mM, pH 8.0) containing 1 mM CaCl 2 and 1 mM of phenylacetate (Sigma) as described previously ( 18 ). The rate of phenyl acetate hydrolysis was determined spectrophotometrically (Uvikon 930 spectrophotometer, Kontron) at 270 nm. PON1 activity was expressed in international units (U) per milliliter of serum. When measuring PON1 activity in lipoproteins, the activity was expressed in international units per gram of protein. The contents of the biologically active oxidized lipids in serum, including 12-hydroxy eicosatetraenoic acid (12-HETE), 13-hydroxy octadecadienoic acid (13-HODE), prostaglandin (PG) and 8-iso-prostaglandin F2 ␣ (8-iso-PGF2 ␣ ), were determined by ELISA (Bluegene, Shanghai, China) according to the manufacturer's instructions.

Western blots
For serum apolipoprotein measurement, 0.2 l of serum was denatured at 90°C for 10 min and then subjected to western blot analysis using anti-apoB, anti-apoE, and anti-apoAI antibodies (Abcam). The proteins were visualized and quantifi ed using a chemiluminescence method (Pierce) and Quantity One (Bio-Rad) software program.

Measurement of antioxidant properties of HDL
The serum of every 5-6 patients was pooled, and fasted serum lipoproteins were fractionated by ultracentrifugation at 40,000 rpm in a Beckman Optima LE-80K into VLDL (density less than 1.006 g/ml), LDL (density = 1.006-1.063 g/ml) and HDL3 (density = 1.125-1.21 g/ml) as described previously ( 19 ). Fractions were dialyzed in PBS at 4°C and LDL (100 g protein/ml) was incubated with freshly prepared CuSO 4 (10 mol/l) in the presence and dyslipidemia (7)(8)(9). Metabolic disease remains a serious concern in the world, and people with metabolic syndrome are at increased risk of developing cardiovascular disease, stroke, and type 2 diabetes ( 9,10 ). It is known that the functions and concentrations of serum HDL have strong inverse correlations with risk of metabolic syndrome and atherosclerotic cardiovascular disease (11)(12)(13). HDL is known to undergo dramatic modifi cation in structure and composition under the concerted actions of infl ammation and oxidative stress ( 14,15 ). As a result, HDL particles progressively lose normal biological activities and acquire altered properties. It is well known that H 2 is an electron donor and therefore has a high reducing ability, and the benefi cial effects of H 2 on different disease models are mostly dependent on its antioxidative, anti-infl ammatory, and antiapoptotic properties ( 16 ). Therefore, it is possible that dyslipidemia-injured HDL function might be improved by H 2 treatment in animals or patients with hyperlipidemia or metabolic syndrome. We have previously demonstrated that administration of H 2 -saturated saline markedly improves the functional properties of the HDL particle in mice and golden hamsters ( 5,6 ). However, whether the same effects of H 2 could be observed in humans has not yet been determined. Here, we aimed to characterize the effects of H 2 on functional properties of the HDL particle on 20 patients with potential metabolic syndrome.

Subjects
The study protocol was approved by the Ethics Committee of Taishan Medical University. We recruited 20 subjects у 43 years of age, males (n = 12) and females (n = 8), from existing patient databases or by advertisement. In this study, subjects were required to have one or more of the following conditions: prehypertension (diastolic blood pressure of 80-89 mmHg and systolic blood pressure of 139 mmHg or lower); prediabetes (fasting serum glucose from 5.2 to 6.9 mmol/l); TC > 5.18 mmol/l and/or LDL-C > 2.59 mmol/l; body mass index (BMI) between 25.0 and 34.9 kg/m 2 ; or waist circumference у 100 cm for males and у 88 cm for females. All participants provided written informed consent to participate before enrollment in the study.

Study design and preparation of H 2 water
The patients consumed 0.9-1.0 l/day of H 2 -rich pure water for 10 weeks. A plastic shelled product (Premium FDR, Friendear, Tokyo, Japan) consisting of metallic magnesium (99.9% pure) and natural stones in 500 ml of polypropylene water bottles (NongFu Spring, HangZhou, China) was used to produce H 2 . The product was capable of generating H 2 when placed in drinking water by the following chemical reaction: Mg + 2H 2 O → Mg (OH) 2 + H 2 . The H 2 water stick was placed into the sealed water bottle for 12 h before consumption by the patients. The sealed cap of the bottle protected the H 2 from escaping, thus preserving the H 2 ions and molecules in the water. The water was drunk by patients within 15 min after opening the sealed cap, and the H 2 concentration was maintained between 0.2 and 0.25 mM and pH between 7.8 and 8.2 measured by a H 2 sensor (Unisense, Denmark) for continuous 15 min after opening the sealed cap. Subjects were instructed to reuse the magnesium sticks by transferring the sticks to a new bottle of water after use. In summary, subjects were expected to of 0.1 M sodium hydroxide, and the radioactivity per aliquot was measured. The percentage cholesterol effl ux was calculated by dividing the media-derived radioactivity by the sum of the radioactivity in the media and the macrophages.

3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay
HUVECs were grown to confl uence in 96-well plates and pretreated with or without HDL (100 g protein/ml) for 18 h and stimulated with TNF-␣ (15 ng/ml) for 6 h. Cells treated with medium only served as a negative control group. After removing the supernatant of each well and washing twice with PBS, 10 l of MTT solution (5 mg/ml in PBS) and 100 l of medium were then introduced. After incubation for another 4 h, the resultant formazan crystals were dissolved in dimethyl sulfoxide (150 l), and the absorbance intensity was measured by a microplate reader (Tecan, Sweden) at 490 nm with a reference wavelength of 620 nm. All experiments were performed in quadruplicate, and the relative cell viability was expressed as a percentage relative to the untreated control cells.

LDL oxidation assay
LDL was isolated by ultracentrifugation from the serum. The MDA content of LDL was determined by the measurement of TBARS and expressed as nanomoles per milligram of protein .
For determining LDL-mediated infl ammation, 100 g protein/ml of LDL was added to the cultured RAW264.7 macrophages and bone marrow-derived macrophages. After incubation for 24 h, quantitation of the secreted proinfl ammatory cytokines, including IL-6 and TNF-␣ , were performed from aliquots of conditioned medium by ELISA (Bluegene, Shanghai, China) according to the manufacturer's instructions. Bone marrow-derived macrophages in mice were obtained as described previously ( 24 ).

Statistical analysis
Statistical analysis was performed by one-way ANOVA (ANOVA) test with the GraphPad Prism v.4.0. Results are expressed as means ± SD. P < 0.05 was considered signifi cant.

Subject characteristics
The baseline demographics of subjects are presented in Table 1 . Subjects enrolled in the study included those or absence of the isolated HDL (200 g protein/ml). After incubation at 37°C for 2 h, the extent of LDL oxidation was assessed by measuring TBARS formation ( 20 ) via a spectrophotometric method according to the manufacturer's instructions (Nanjing Jiancheng Biochemistry, China).

Endothelial cell monocyte adhesion assay
Monocyte adhesion assays were performed under static conditions as previously described ( 21 ) with minor modifi cation. Human umbilical vein endothelial cells (HUVEC) were grown to confl uence in 24-well plates and pretreated with or without HDL (100 g protein/ml) for 18 h and stimulated with TNF-␣ (10 ng/ml) or LDL (100 g protein/ml) for 6 h. THP-1 cells were labeled with a fl uorescent dye, 2 ′ ,7 ′ -bis(2-carboxyethyl)-5(6)-carboxyfl uorescein acetoxymethyl ester (BCECF-AM), by incubation with 10 µmol/l BCECF-AM at 37°C for 1 h in RPMI-1640 medium and were subsequently washed with EBM-2. Confl uent HUVECs in 24-well plates were washed three times, and then labeled THP-1 cells (2 × 10 5 cells per 200 µl) were added to each well of HUVEC. THP-1 cells were allowed to adhere to HUVECs by incubation at 37°C for 60 min, and unbound THP-1 cells were removed by washing (three times, 5 min). THP-1 cells bound to HUVECs were counted under fl uorescent microscope. The number of adherent leukocytes was determined by counting four fi elds per 100× high-power fi eld well using fl uorescent microscopy (Nikon, Japan) and photographed. Four randomly chosen high-power fi elds were counted per well. Experiments were performed in duplicate or triplicate and were repeated at least three times. The person counting the adherent monocytes was unaware of the treatment.

HDL-induced cholesterol effl ux assay
Cholesterol effl ux experiments were performed as described by Smith et al. ( 22 ). Acetyl LDL (AcLDL) was prepared according to the methods of Basu et al. ( 23 ). RAW264.7 macrophages at 50% confl uence were cholesterol loaded and labeled in 1 ml of RGGB (RPMI 1640 supplemented with 50 mM glucose, 2 mM glutamine, and 0.1% BSA) containing [ 3 H]cholesterol (1 Ci/ml) and AcLDL (100 g protein/ml) for 30 min. Then macrophages were washed twice with 0.1% BSA-PBS and equilibrated with RGGB for 24 h. On the following day, the medium was then replaced with RGGB containing 200 g protein/ml of HDL. After 12 h of incubation, the culture was centrifuged to remove cell debris, and 100 l of the medium was removed for determination of radioactivity. At the end of the chase period, the macrophages were dissolved in 0.4 ml signifi cantly ( Fig. 3A ), and the activity of SOD, which acts as an antioxidant and protects cellular components from being oxidized by reactive oxygen species, was increased after H 2 water consumption ( Fig. 3B ). However, the activity of PON1, an antioxidant enzyme associated with HDL, was not altered in either serum or HDL3 fractions by H 2 water ( Fig. 3C, D ). In addition, we detected serum concentrations of several biologically active oxidized lipids, including 12-HETE, 13-HODE, PG, and 8-iso-PGF2 ␣ . The data in Fig. 3E, F shows that 13-HODE and 8-iso-PGF2 ␣ levels were improved following the H 2 treatment; however, 12-HETE and PG levels were not altered by H 2 . There was no significant effect of intake of H 2 water on serum levels of infl ammatory biomarkers, including TNF-␣ and IL-6 ( Fig. 3G, H ).

H 2 treatment decreases apoB and apoE protein levels in serum
Serum LDL and HDL are particles composed of a variety of lipids and protein components; thus, it is necessary to clarify which content of the lipoprotein could be affected who had TC > 5.18 mmol/l (n = 17), LDL-C >2.59 mmol/l (n = 18), BMI 25-34.9 (n = 17), and/or smokers (n = 10). All subjects showed mean normal clinical levels of baseline biometric parameters, clinical chemistry, and hematology. All smokers were occasional smokers.

Effect of H 2 on serum lipoprotein profi les and serum levels of antioxidative and infl ammatory biomarkers
The serum lipid levels of each individual are presented in Table 2 , and the distribution of lipid levels is shown in Fig. 1 . We can see that serum TC and LDL-C levels were signifi cantly decreased after 10 weeks of H 2 treatment in all of the 20 patients with potential metabolic syndrome. Among 20 patients, 18 had decreased and 1 (patient 14) had increased TC and LDL-C levels, and 1 (patient 8) was not altered after H 2 consumption.
In addition, among the 20 patients, 10 were smokers and 10 were nonsmokers. As shown in Table 2 , H 2 treatment decreased serum TC and LDL-C levels not only in smokers but also in nonsmokers. And it seems that the lipid-lowering effects of H 2 on smokers were better than those on nonsmokers, although there is no signifi cant difference. Serum levels of HDL-C and glucose were not altered by consumption of H 2 ( Table 2 and Fig. 1C, D ). Interestingly, serum TG levels were decreased by H 2 treatment in 10 smokers, although there is no signifi cant difference ( Table 2 ), which caused a slight decrease in TG levels in all of the 20 patients ( Table 2 and Fig. 1E ).
Moreover, serum lipoprotein profi le by FPLC further confi rmed the decrease in serum LDL-C levels in H 2 -treated subjects and revealed that serum VLDL-C and HDL-C remained unchanged after intake of H 2 water ( Fig. 2 ).
Changes in biomarkers of oxidative and infl ammatory status after 10 weeks of consumption of H 2 water are shown in Fig. 3 . Serum levels of MDA, one of the most frequently used indicators of lipid peroxidation, were decreased   . 5A ), suggesting that H 2 improves the oxidation of HDL particle.
Second, the biological effect of H 2 on the antioxidative functionality of HDL was tested, namely, the protection of LDL particles from oxidation. As shown in Fig.  5B , H 2 treatment signifi cantly inhibited the formation of TBARS.
Third, the effect of H 2 on the anti-infl ammatory properties of HDL was tested, including protection of cytokine-induced monocyte adhesion to endothelial cells and stimulation of endothelial nitric oxide production, which has been suggested as an important endothelial-atheroprotective effect of HDL. As shown in Fig. 5C , after incubation of HUVECs for 6 h with TNF-␣ , adhesion of monocytes to HUVECs was signifi cantly increased, and preincubation of HUVECs with HDL3 isolated from patients after 10 weeks of drinking H 2 water (HDL-H 2 -A) markedly reduced TNF-␣ -induced adhesion of monocyte to HUVECs compared with those of HDL3 isolated at baseline (0 week) (HDL-H 2 -B). These data suggest that the anti-infl ammatory function of HDL was improved by H 2 water. We also determined the effects of HDL-H 2 -A compared with HDL-H 2 -B on endothelial nitric oxide production. Unfortunately, the nitric oxide production by H 2 treatment. Consistent with the difference observed for LDL-C, apoB100, the major protein on LDL, was signifi cantly decreased by consumption of H 2 water in patients with potential metabolic syndrome ( Fig. 4 ). The apoE protein, which is mainly present in VLDL and LDL particles, was also lowered with the treatment of H 2 ( Fig. 4 ). ApoAI, the major protein on HDL, was not altered after intake of H 2 water ( Fig. 4 ), which is consistent with the changes observed for HDL-C. These data suggested that H 2 could decrease the expression of the major protein constituents of LDL and VLDL.

H 2 improves the oxidation and the functional properties of the HDL particle
HDL is known to undergo dramatic modifi cation in structure and composition under the concerted actions of inflammation and oxidative stress ( 14,15 ). Recent evidence indicates that H 2 acts as a therapeutic medical gas in a variety of disease models by exerting antioxidant and anti-infl ammatory effects ( 3,25,26 ). Therefore, it is possible that administration of H 2 might improve the functional quality of HDL particle. First, H 2 water supplementation tended to decrease the MDA content in HDL3 H 2 water signifi cantly decreased LDL-induced monocyte adhesion to endothelial cells. These data reveal that H 2 reduces LDL-mediated infl ammatory properties in culture.

DISCUSSION
In a previous study, we found that H 2 has benefi cial lipidlowering effects in high-fat diet-fed Syrian golden hamsters ( 6 ). However, it remains unknown whether H 2 has effects on lipid and lipoprotein metabolism in humans. The key fi nding of our present study is that the novel antioxidant chemical element H 2 seems to alleviate lipid metabolism disorder, including hyperlipidemia and defective HDL function, in patients with potential metabolic syndrome. Our results were not consistent with that of human studies in type 2 diabetes mellitus, as they showed signifi cant decreases in modifi ed LDL-C levels and no effect on total and LDL-C ( 3 ). The inconsistency might be attributable to the difference in pathological conditions, the dosage of administration, or the intervention period. Despite this possibility, the discrepancy might be explained by the fact that modifi ed LDL, like oxidized LDL, largely existed in the hyperlipidemia model, which was used in our study.
Subanalysis was conducted on 20 subjects, and we found the serum TC levels of 17 subjects with hyperlipidemia (TC > 5.18 mmol/l) was decreased by intake of H 2 water for 10 weeks, and the serum TC levels of the remaining 3 subjects without hyperlipidemia (TC < 5.18 mmol/l) was not signifi cantly altered by H 2 treatment, although the TC level of 1 subject was slightly increased after treatment ( Table 2 , patient 14). The data gave us a clue that H 2 might exert a lipid-lowering effect on the patients with hyperlipidemia but have little effect on the population without hyperlipidemia.
It is well known that high cholesterol level is one of the important risk factors for atherosclerosis. Therefore, the elucidation of the mechanism by which H 2 decreases serum cholesterol will provide solid evidence for the application of H 2 in cardiovascular disease therapy. Serum LDL and VLDL are particles composed of a variety of lipids and protein components; thus, it is necessary to clarify which content of the lipoprotein could be affected by H 2 treatment. First, LDL is a contributing factor to the development of atherosclerosis. ApoB100 is the major protein present in LDL particles, and like LDL-C, the serum apoB level has been positively correlated with risk for atherosclerotic disease. In the present study, we found H 2 treatment not only decreased serum LDL-C levels, but also remarkably lowered apoB100 and apoE protein levels in serum. The data gave us a clue that H 2 might regulate the metabolism of LDL-C by decreasing the synthesis of apoB or increasing the catabolism of LDL-C . Second, like LDL-C, VLDL-C is considered a type of ''bad'' cholesterol because elevated levels are associated with an increased risk of coronary artery disease ( 27 ). We found that apoB and apoE, the major proteins in VLDL, were lowered by H 2 , although cholesterol analysis by FPLC revealed no changes of was not altered signifi cantly after intake of H 2 water (data not shown), which might be attributable to the sensitivity of Griess method.
Fourth, the ability of the isolated HDL particles to elicit effl ux from cholesterol-loaded macrophages was tested. As shown in Fig. 5D , HDL particles isolated from the serum after H 2 treatment exhibited dramatically higher effl ux properties compared with the HDL particles isolated from the serum before H 2 treatment, indicating that the cholesterol effl ux ability mediated by HDL particles was increased by H 2 .
Finally, the biological effect of H 2 on the antiapoptotic functionality of HDL was determined. As shown in Fig. 5E , H 2 treatment signifi cantly inhibited TNF-␣ induced endothelial cell apoptosis. These data indicate that dyslipidemia-injured HDL functions, including the ability to protect against LDL oxidation, the ability to inhibit cytokineinduced monocyte adhesion to endothelial cells, the ability to stimulate cholesterol effl ux from macrophage foam cells, and the ability to protect endothelial cell apoptosis, were markedly improved by administration of H 2 water in patients with potential metabolic syndrome.

H 2 reduces the oxidation of LDL and LDL-mediated infl ammation
The oxidation of LDL plays an important role in atherogenesis and may infl uence lipid metabolism. In this study, we determined the effects of H 2 water on the oxidation of LDL and LDL-mediated infl ammation. As shown in Fig. 6A , the MDA content of the isolated LDL was reduced by H 2 treatment, suggesting that H 2 reduces the oxidation of LDL. To determine LDL-mediated infl ammation, 100 g protein/ml of the isolated LDL was added to the cultured RAW264.7 macrophages and bone marrow-derived macrophages for 24 h. As shown in Fig. 6B, C , the secretion of TNF-␣ and IL-6 by macrophages was reduced by H 2 treatment. Furthermore, we tested the effects of H 2 on LDL-induced monocyte adhesion to endothelial cells. As shown in Fig. 6D , consumption of  of lipoprotein particles and improves the serum lipoprotein profi le. Previous studies have demonstrated that administration of H 2 -rich saline improves insulin sensitivity ( 3 ), which, in our view, could partly contribute to the improved lipid metabolism in our study. Furthermore, liver cells sense the reduced levels of hepatic intracellular cholesterol and seek to compensate by synthesizing LDL receptors to draw cholesterol out of the circulation ( 28 ). Future studies on hepatic HMG-CoA reductase and LDL receptors may elucidate the mechanism by which H 2 acts on lipoprotein regulation. VLDL-C after H 2 treatment. The inconsistencies in protein and cholesterol levels might be explained by the suspicion that the functional target of H 2 might be apolipoprotein expression, not cholesterol. Taken together, the data indicate that molecular H 2 dissolved in water might have a benefi cial regulating effect on lipid abnormality in patients with metabolic syndrome, especially in patients with hyperlipidemia. This effect is partially related to its regulation of lipid and protein contents of LDL and VLDL. Further experiments are needed to identify the mechanisms by which H 2 regulates the lipid and protein contents Serum of every 5-6 patients was pooled, and LDL was isolated by ultracentrifugation from the serum. (A) MDA content of the isolated LDL. (B) One hundred micrograms of protein per milliliter of isolated LDL was added to cultured RAW264.7 macrophages. Supernatants were harvested at 24 h for measurement of TNF-␣ and IL-6 by ELISA assay. (C) One hundred micrograms of protein per milliliter of isolated LDL was added to the cultured bone marrow-derived macrophages, and cytokines in medium were measured after 24 h incubation. (D) HUVECs were stimulated with LDL (100 g protein/ml) for 6 h and THP-1 cells, labeled with a fl uorescent dye BCECF-AM, were added to HUVECs. After adhering, THP-1 cells bound to HUVECs were counted under fl uorescent microscope as described in Methods. Results are expressed as means ± SD (n = 3-4 pooled serum samples ). * P < 0.05. reveals that H 2 seems to reduce the oxidation of LDL and LDL-mediated infl ammation, and they further support our prediction that H 2 might improve lipid metabolism in patients with hyperlipidemia by inhibiting LDL-mediated infl ammation. Previous studies have demonstrated that administration of H 2 reduces atherogenesis in apoE Ϫ / Ϫ mice ( 5 ); in our view, this could partly contribute to the observed protective effects of H 2 on the oxidation of LDL and LDL-mediated infl ammation in our study.
Indeed, our data give us a clue that H 2 might have the potential to be used as a novel lipid-regulating agent with the advantage of no toxicity compared with other commonly used lipid-regulating drugs that have side-effects on the liver and kidney. Further understanding of the mechanisms underlying the signaling pathways involved in the ability of H 2 to infl uence lipid and cellular metabolism is required to fully exploit intake of H 2 gas as a therapeutic strategy.
There are several limitations to the present study. First, the number of subjects recruited is small and it is hard to draw a solid conclusion, particularly having smokers among the subjects. It should, however, be emphasized that we analyzed the data in smokers and nonsmokers separately, and the response to H 2 in smokers seems to be stronger than that in nonsmokers, but the difference did HDL is known to protect against the development of atherosclerosis and is widely documented as a "negative risk factor" for coronary heart diseases ( 29 ). The antiatherogenic activity of HDL is principally attributable to a variety of antioxidative, anti-infl ammatory, and antiapoptotic properties and the reverse transport of cholesterol ( 30 ). In the present study, we that found H 2 treatment seems to improve the functionality of HDL3 without altering HDL-C serum levels. It is known that the benefi cial effects of H 2 on different disease models are mostly dependent on its antioxidative, anti-infl ammatory, and antiapoptotic properties ( 16 ). Therefore, it is possible that the protective effect of H 2 on HDL function in our study is attributable, at least in part, to the antioxidative and anti-infl ammatory properties of H 2 . A number of therapeutic strategies are being developed to target HDL-C in an attempt to inhibit the progression or induce regression of atherosclerosis and reduce cardiovascular events. Therefore, the protective effect of H 2 on HDL function in the our study might provide a further evidence for H 2 application in vascular disease therapy. We did not observe any unwanted side effects of H 2 , including headache, diarrhea, and vomiting (data not shown), suggesting fewer toxic and adverse effects of H 2 .
The oxidation of LDL plays an important role in atherogenesis and may infl uence lipid metabolism. Our data not reach statistical signifi cance. It would be useful to enlarge the sample size and examine the reaction to H 2 in smokers due to the antioxidative property of H 2 . Another limitation of the study was that it was not a double-blind, randomized, controlled trial comparing H 2 -rich water to placebo and comparing experimental endpoints from H 2treated metabolic syndrome subjects with those measured in untreated healthy individuals. It is possible that during the 10 weeks of the study, the subjects altered their lifestyles, which could have affected the parameters studied. Therefore, it is diffi cult for us to draw a solid conclusion, and it limits the conclusions to stating that H 2 treatment appear to regulate lipid metabolism in patients.
In summary, our data show that in vivo administration of H 2 water seems to decrease serum TC and LDL-C levels and improve HDL functions in patients with potential metabolic syndrome, suggesting that H 2 may be used as a newer pharmacological agent to treat or control lipid metabolism disorder.