Effects of drinking hydrogen-rich water in men at risk of peripheral arterial disease: a randomized placebo-controlled trial

Aims Smoking, hypertension, hyperlipidemia, and diabetes are considered to increase the incidence of peripheral arterial disease (PAD). They can activate endogenous free radicals, cause inammation and oxidative stress, and lead to endothelial cell dysfunction. Hydrogen (H 2 ) has been proven to decrease oxidative stress, improve cell function, and reduce chronic inammation. The purpose of this research was to validate the role of H 2 in individuals who are at risk of PAD. Methods Sixty subjects were randomly assigned to placebo (PBO) group or H 2 -rich water (HRW) group and drank either bottled pure water or H 2 -rich water (245 mL/time, 3 times/d) for ten weeks. Results The pulse wave velocity was ameliorated in the HRW group with no signicant change in the ankle-brachial index. The serum total cholesterol of the HRW group was signicantly reduced compared to the placebo group. In addition, compared to baseline, the levels of lipoprotein(a) was decreased, the malondialdehyde content was reduced, the superoxide dismutase activity was increased, and the expression of intercellular cell adhesion molecule-1 was decreased signicantly in the HRW group. The oxidized phospholipid of 1-palmitoyl-2-azelaoyl-sn-glycero-3-phosphatidylcholine level in the HRW group were signicantly reduced compared to the placebo group. Finally, H 2 signicantly improved the antioxidant, antiinammatory, and antiapoptotic abilities of high-density lipoprotein (HDL). Conclusions Drinking HRW can improve vascular sclerosis indicators, improve dyslipidemia, reduce vascular oxidative stress and inammation, and improve HDL function. H 2 may be used to prevent and relieve PAD caused by major risk factors such as smoking, hypertension, hyperlipidemia, and diabetes.


Introduction
Atherosclerotic diseases, including peripheral arterial disease (PAD), coronary artery disease, and cerebral artery disease are the leading causes of death worldwide. PAD is an abnormal narrowing of the arteries, and mainly includes disease of the aortoiliac, femoropopliteal, and infrapopliteal arterial segments. The symptomatic manifestations of PAD include leg pain, intermittent claudication, pain at rest, gangrene when the limb is severely ischemic, and even amputation 1 . Currently, there are more than 202 million patients with PAD worldwide, and it is predicted that up to 45 million patients with PAD will die from coronary or cerebrovascular disease over a 10-year period 2 . Smoking, hypertension, hypercholesteremia, and diabetes are the four major risk factors for PAD 3 . Smoking is a particularly strong risk factor for PAD with an obvious dose-response relationship, and heavy smokers are four times more likely to develop PAD than nonsmokers 4 . Diabetes, hypertension, and hypercholesterolemia are also highly associated with PAD, with an approximately two-to three-fold increased risk. The incidence of PAD is also increased with age 5,6 .
Planned exercise and lifestyle improvements are practical ways to reduce the risk and delay the progress of PAD. Currently, PAD is mainly treated with antiplatelet therapy, anticoagulant therapy, statins, antihypertensive therapy, and medications to improve circulatory ow 7 . These therapies are mainly aimed at the cause of PAD, and applied to patients with symptomatic PAD; however, such drugs are usually accompanied by side effects. At present, 17β-estradiol 8 , ginsenoside Rb3 9 , recombinant human Relaxin-2 10 and active substances from red wine 11 have shown promise in improving vascular damage caused by smoking. However, their curative effects have certain limitations. Patients with typical PAD symptoms account for only 20%, and approximately 50% of patients are asymptomatic 12 . People generally lack awareness of the diagnosis and treatment of PAD, and are unlikely to choose long-term treatment to slow the progression of PAD. Therefore, for asymptomatic patients with PAD, it is of great signi cance to choose a safe treatment with no side effects in order to prevent and alleviate the progression of the disease.
Hydrogen (H 2 ) is a bioactive gas that has bene cial effects in diseases such as metabolic syndrome 13,14 , type 2 diabetes 15 , chronic liver in ammation 16 ,and focal brain and ischemia/reperfusion injury 17 . the mechanisms of which are thought to be related to its antioxidative, antiin ammatory, and antiapoptotic properties. In addition, it has been previously demonstrated that H 2 has a protective effect on endothelial cells. Indeed, a previous report showed that H 2 -saturated water could promote the recovery of blood perfusion in a mouse PAD model by increasing angiogenesis and decreasing the level of oxidative stress 18 . However, the effect of H 2 in individuals at risk of PAD has not yet been demonstrated. In this study, we performed a randomized placebo-controlled trial to characterize the effect of H 2 -rich water drinking on PAD in men at risk of PAD, and its effects on oxidative stress and in ammatory factors. This is the rst randomized controlled trial of H 2 on risk of PAD.

Subject's baseline characteristics
The selection process is shown in Fig. 1. 59 subjects participate in the study, and one subject withdrew at week 10 due to his work outside. The speci c baseline clinical indicators are shown in Table 1. The study subjects were divided into the placebo (PBO) group or the H 2 -rich water (HRW) group at random. The study subjects each drank three bottles of placebo water or H 2 -rich water (245 mL/bottle) per day for 10 weeks. Blood samples were collected at the beginning and after the 10-week trial. Breath hydrogen measurement following administration of H 2 -rich water As shown in Fig. 2, after drinking H 2 -rich water (650-700 µmol/L, 245 mL), the exhaled H 2 concentration rose rapidly over the course of 5 min and reached the peak in the 10th min, with a value of approximately 7 ppm. Subsequently, the concentration gradually decreased and returned to baseline in the 60th min. There was no signi cant difference between the results of males and females (data not shown).
Effect of H 2 on PWV and ABI Table 2 depicts the changes in ABI and PWV after 10 weeks of H 2 -rich water intervention. The low value of ABI (the minimum value of the measured ABI of the left and right lower limbs) and the ABI of the left and right lower limbs were not signi cantly different between the HRW group and the PBO group). The low value of PWV (the minimum value of the measured PWV of the left and right lower limbs) and the PWV of the left lower limb decreased signi cantly after drinking H 2 -rich water (P < 0.05). The PWV of the right lower limb also showed a decreasing trend after drinking H 2 -rich water (P > 0.05). shown as mean ± SD. ‡ Statistical analysis was performed by nonparametric tests for nonparametric data, and Student's t test for normally distributed data. *P < 0.05.

Effect of H 2 on lipid pro les
After 10 weeks of intervention, the TC level in the HRW group was signi cantly lower than that in the PBO group (P < 0.05, Table 3). The serum Lp(a) levels were signi cantly decreased after 10 weeks of H2 treatment in the HRW group (P < 0.05, Table 3). In addition, the levels of TG, VLDL, and LDL-C showed a slight decreasing trend in the HRW group after 10 weeks of intervention compared with the HRW group before intervention or the PBO group (Table 3). No obvious changes in Apo A , Apo B, and HDL-C levels were observed (Table 3). P § represents the comparison between the hydrogen-rich water group and the placebo group at 10week.*P < 0.05. † Non-normal data is represented by median and interquartile range and the remaining results are shown as mean ± SD. ‡ Statistical analysis was performed by nonparametric tests for nonparametric data, and Student's t test for normally distributed data.

Effect of H 2 on oxidative and in ammatory biomarkers
The serum levels of MDA decreased and the activity of SOD increased in the HRW group after 10 weeks of intervention (P < 0.05, Table 3). The levels of oxidized phospholipids also decreased, especially PAzPC (P < 0.05, Table 3). The HRW group demonstrated signi cant attenuation of the in ammatory biomarker of ICAM-1 (P < 0.05, Table 3). The concentration of MMP-1 and CCL1 was not signi cantly different following intervention (Table 3). H 2 improves the oxidation and the functional properties of HDL H 2 signi cantly reduced thiobarbituric acid-reactive substances generated by LDL oxidation (Fig. 3A), indicating that H 2 can improve HDL antioxidant function. We detected cell viability by CCK-8 to verify the antiapoptotic function of HDL after H 2 intervention. We found that H 2 inhibited endothelial cell apoptosis induced by ox-LDL (Fig. 3B). Ox-LDL can induce HUVECs to increase the adhesion of monocytes to endothelial cells. Following incubation with HDL, we found that the adhesion of monocytes in the HRW group was signi cantly reduced after H 2 intervention (Fig. 3C), showing that H 2 can enhance the antiin ammatory effect of HDL.

Discussion
We performed a randomized placebo-controlled trial and found that H 2 can relieve vascular sclerosis, regulate lipid metabolism disorder, improve antioxidant capacity, and reduce in ammation in people who are at risk of PAD. Since ischemia/reperfusion injury that occurs during PAD is accompanied by an increase in reactive oxygen species (ROS) formation, antioxidant therapy may be a viable countermeasure 19 . H 2 , as a novel antioxidant, has been shown to have roles in various diseases, and a previous report has demonstrated the therapeutic effect of H 2 in PAD in mice 18 . Therefore, it is important to further verify the effect of H 2 in patients with PAD.
We measured the H 2 concentration in the exhaled gas of patients in order to study the in vivo kinetics of H 2 after drinking H 2 -rich water. Our results demonstrated that it took 10 min for H 2 to reach the peak, and it gradually returned to baseline in 60 min. These results were consistent with previous reports that showed that the breath H 2 concentration reaches the maximum 10-15 minutes after intake, and then decreases gradually, returning to baseline levels 45-150 min after drinking H 2 -rich water [20][21][22] . The peak concentration in the current study was approximately 7 ppm after drinking 245 mL of H 2 -rich water; this is lower than previous reports that found peak breath hydrogen levels of 30-60 ppm after drinking 200-300 mL of H 2 -rich water [20][21][22] .
PWV and ABI are becoming increasingly important in the assessment of arterial stiffness. ABI refers to the pressure ratio of the left and right upper arms to the left and right ankles, and has a normal range of 1.0-1.4. PWV is the most widely used measure of arterial stiffness, and an increase in the propagation speed of the PWV in the artery is associated with an increase in arterial stiffness. The uctuating conduction velocity between two heartbeats is used to judge the elasticity of arterial wall 23 . The normal PWV is less than 1400 cm/s, and our study showed that H 2 had a signi cant impact on PWV following 10-week intervention.
The lipid-lowering effect of H 2 -rich water has been veri ed in patients with potential metabolic syndrome 24 . In our study, the TC level was lower in the HRW group than the PBO group, and the 10-week H 2 intervention reduced the Lp(a) level in individuals at risk of PAD. Furthermore, our previous study also showed the reducing effect of H 2 -rich water on TC in patients with potential metabolic syndrome or hypercholesterolemia 13,25 . In addition, studies have shown that the serum TC levels were signi cantly decreased in ethanol-induced fatty liver and nonalcoholic fatty liver mice after H 2 -rich water treatment 26,27 . It is well known that a high cholesterol level is one of the most important risk factors for PAD.
Therefore, elucidation of the mechanism by which H 2 decreases serum cholesterol will provide solid evidence for the application of H 2 in PAD therapy.
Lp(a) is a cholesterol-rich, LDL-like particle, with speci c apolipoprotein (a) and apolipoprotein B-100. The Lp(a) levels of plasma independently predict atherosclerotic cardiovascular disease (CVD) and PAD, although its mechanism of action in atherosclerosis remains unclear 28 . Therapy to lower the plasma levels of Lp(a) has gained much attention in recent years, and our results reveal that H 2 -rich water can reduce the level of plasma Lp(a). Lp(a) is regarded as a preferential carrier of oxidized phospholipids (OxPLs) in human plasma 29 , and we demonstrated that H 2 -rich water can reduce the levels of oxidized phospholipids, especially PAzPC. Moreover, a previous study showed that H 2 can suppress the autoxidation of linoleic acid and PAPC in a pure chemical system 30 . The mechanism by which H 2 decreased the formation of OxPLs in vivo is worthy of further study.
PAD is an atherosclerotic disease of arterial vessels, in which the cycles of ischemia and reperfusion induced by PAD leads to increased mitochondrial ROS 19 . Smoking is one of the main risk factors for PAD, and Smokers with hypertension, hyperlipidemia, and diabetes will increase the incidence of PAD. Cigarette products are known to activate the production of oxidative free radicals in the body 31 . In our study, H 2 , as a novel antioxidant, could signi cantly reduce MDA and signi cantly increase the activity of SOD. MDA is the product of lipid peroxidation and a marker of oxidative stress, while SOD is an antioxidant enzyme that can remove ROS from the body 32 . Studies have shown a reduction in serum MDA and the increase in SOD activity in patients with potential metabolic syndrome 33 or type 2 diabetes 20 after H 2 treatment.
Previous reports have also shown that H 2 can inhibit formation of MDA in ethanol/acetaminopheninduced fatty liver 34 or liver ischemia reperfusion injury in mice 35 , and reduce MDA and increase SOD activity to relieve oxidative stress induced by chronic intermittent hypoxia in rats 36 . Thus, H 2 may reduce the formation of lipid peroxides in people at risk PAD by improving antioxidant enzyme activity.
ICAM-1 plays key roles in immune-mediated and in ammatory processes; it can be induced by interleukin-1 and tumor necrosis factor, and expressed by the vascular endothelium, macrophages, and lymphocytes. ICAM-1 is also involved in local plaque formation, and has been shown to be an independent predictor of the development and progression of PAD 37 . Previous studies have shown that H 2 can inhibit the expression of ICAM-1 and reduce the in ammatory response in different animal models, including a pressure ulcer mouse model 38 , sepsis mouse model 39 , ,and noise-induced hearing loss guinea pig model 40 . In our study, the reduced expression of serum ICAM-1 may have led to the alleviation of PAD.
HDL is known to have vasoprotective actions and antiatherogenic effects 41 . The underlying mechanism is mainly related to its function in promoting cholesterol e ux from macrophage foam cells and stimulating endothelial cell NO production to improve endothelial cell function 41,42 . HDL also plays roles in the reduction of in ammation and oxidative stress 43 , and is bene cial to endothelial cells by protection of cytokine-induced monocyte adhesion 44 . It has been shown that H 2 can improve HDL function by enhancing the cholesterol e ux ability mediated by HDL, preventing LDL oxidation, and reducing ox-LDL-induced endothelial cell apoptosis and monocyte adhesion to endothelial cells 25,33 . In this study, we isolated HDL from different groups before and after intervention and tested the functions in vivo. Our results showed that H 2 can improve the antioxidant, antiin ammatory, and antiadhesion effects of HDL, which may improve oxidation and in ammation, and reduce vascular damage in individuals at risk of PAD.
This study shows the effects of H 2 on relieving vascular sclerosis by antioxidation and antiin ammatory mechanisms, as well as improving HDL function. It is likely that H 2 exerts antioxidant and antiin ammatory effects by improving SOD activity and HDL function; this, in turn, decreases the levels of MDA, OxPLs, ICAM-1, CCL-1, and even Lp(a), thereby alleviating artistic stiffness as measured by PWV. At the same time, H 2 exerts a lipid-lowering effect by reducing plasma TC, which also functions to lower the risk of PAD .
This study has several limitations. First, the number of participants is small, and an expanded sample is needed to verify the results. Second, the dose-effect of H 2 -rich water was not studied, and a further study is needed to determine the best dose. Third, the intervention only lasted for 10 weeks, and the effect of long-term intervention needs further veri cation. Finally, besides the antioxidative and antiin ammatory properties, the molecular mechanisms of H 2 need further study, especially its regulatory effects on lipid metabolism disorder and its effect on Lp(a).
In conclusion, our data show that H 2− rich water can improve vascular sclerosis indicators and lipid disorders, and reduce oxidative stress and in ammatory factor in ltration. H 2 may be used in adjunctive therapy for alleviating PAD.

Methods
Subjects and study design The study was a 10-week, randomized, placebo-controlled trial. The study protocol was authorized by the Ethics Committee of Shandong First Medical University (NO.2019121, Date 08/10/2019). This study was registered in the Chinese Clinical Trial Registry (www.chictr.org.cn, Registration number Chi CTR 2000035232, Date: 04/08/2020). All participants who were eligible and agreed to participate in the randomized assignment were required to sign a written informed consent before participating in the study. This study followed CONSORT guidelines. All methods were performed in accordance with the relevant guidelines and regulations for research involving humans. Sixty-three subjects over 40 years old were enrolled from Zhoudian community (Tai'an, China). The enrollment conditions were as follows: Current smoker or quit smoking within the past 10 years, ankle-brachial index (ABI) < 1.0 or smoking index (smoking intensity × duration of smoking) ≥ 200, with or without risk factors for diabetes, high blood pressure, and hyperlipidemia, and able to complete the questionnaire independently or with the help of the researcher. Antioxidant properties of HDL HDL was isolated from the pooled serum by ultracentrifugation (n = 3 samples for each group, each comprising the serum of 4-5 subjects) as described 45 . LDL (100 µg/mL) from healthy people and HDL (200 µg/mL) isolated from each group were incubated with freshly prepared CuSO 4 (10 µmol/L) at 37 °C for 2 h. The extent of LDL oxidation was assessed by measuring the level of MDA via a spectrophotometric method according to the manufacturer's instructions (Nanjing Jiancheng Biochemistry, China).
Endothelial cell -monocyte adhesion assay The monocyte adhesion assay was slightly modi ed as described previously 46 . Human umbilical vein endothelial cells (HUVECs) were cultured at 37 °C in a humidi ed 95% air-5% CO 2 atmosphere, grown to 70-80% con uence in 96-well plates, and stimulated with ox-LDL (100 µg/mL) in the presence or absence of HDL (100 µg/mL) for 24 h. THP-1 monocytes at a density of 2 × 10 5 were labeled with 10 µmol/L 2',7'-bis(2-carboxyethyl)-5(6)-carboxy uorescein, acetoxymethyl ester (BCECF-AM) at 37 °C for 1 h in RPMI-1640 medium and rinsed with serum-free RPMI-1640 medium. HUVECs in 96-well plates were washed three times and incubated with 100 µL THP-1 cells for 1 h. Then, each well was rinsed three times with PBS to remove unbound THP-1 cells. THP-1 cells bound to HUVECs were visualized with a uorescent microscope (Nikon, Japan) at 4 elds per × 100 high-power-eld well. Experiments were performed at least three times and the selection of high-power elds to count separate wells was performed at random. Cell viability determined by CCK-8 assay HUVECs were seeded in 96-well plates and pretreated with or without HDL (100 µg /mL) for 6 h and stimulated with ox-LDL (100 µg/mL) for 18 h. The viability of the HUVECs was measured by CCK-8 assay (Med Chem Express, USA), and the absorbance was measured at 450 nm using a microplate spectrophotometer system (Tecan, Sweden). The percentage viability was calculated using the following formula: HUVECs viability % = (OD sample -OD blank )/ (OD control -OD blank ) × 100% 33 .

Statistical analysis
Descriptive statistics were used to compare the baseline characteristics of the subjects in the two groups (means ± SD). Statistical analysis was performed by Student's t test for normally distributed data and by nonparametric tests for nonparametric data. The SPSS program (version 22.0) was used for all statistical analyses, and all data were plotted with GraphPad Prism 8. P-values < 0.05 were considered signi cant.

Declarations
Con ict of Interest: The authors have no con icts of interest to report.   Exhaled H2 concentration after drinking hydrogen-rich water.