The DML extract used was manufactured by a good manufacturing practices-certified company (Natural F&P Corp., Cheonwon, Korea). Fresh leaves from DML were obtained from Hambakjae Biopharm Co., Ltd. (Jeju, Korea). The plant was authenticated by two practitioners of traditional Asian medicine. Leaves from the plant samples (15 kg) were chopped, blended and soaked in 2 liters of 80% ethanol, and refluxed three times at 20°C for 2 h. The insoluble materials were removed by centrifugation at 10,000 × g for 30 min, and the resulting supernatant was concentrated and freeze-dried to obtain a powder. For the test product preparation, DML, stearin magnesium and silicon dioxide were mixed. For the placebo, microcrystalline cellulose 102 was mixed with lactose powder, and dextrin was added to obtain a tablet form. Food colors (red, yellow and blue) were also added to produce the same color as the test product (Table I). The raw materials used in the test product and placebo were placed in a mixer and blended sufficiently to ensure homogeneity; subsequently, these were made into tablet at 300 mg/tablet.

The test product and placebo were assessed for stability; moisture content (%); coliform group; lead (mg/kg), arsenic (mg/kg) and sodium (mg/100 g) levels; and disintegration by outsourcing to ‘Korea Health Supplements Association’ (Seongnam, South Korea). All test results were acceptable as per the requirements set by the Food Sanitation Act. No preservative was added to the test product.

In accordance with the Functional Foods for Health Act (http://www.mfds.go.kr/files/upload/eng/FOOD_SANITATION_ACT.pdf), all issues associated with human trials were approved by the Institutional Review Board for Human Trials of Hallym University (Hwaseong, Korea) (IRB approval no. 2014-074); the study was conducted in agreement with the Declaration of Helsinki and performed in accordance with International Conference on Harmonization guidelines. In addition, the study was retrospectively registered at the WHO Primary Registries in Republic of Korea (Clinical Research Information Service) (registration no. KCT0001472, May 8, 2015).

Subjects were male and female volunteers aged 23–65 years (n=61). The subjects were informed of the details of the trial, and only those who fully understood the purpose of trial participation and gave consent for participation were selected as subjects. All the subjects were considered suitable for the trial as they had normal liver functions, as assessed by measuring glutamate-oxaloacetate transaminase (AST) and glutamate-pyruvate transaminase (ALT) levels in the blood. Exclusion criteria consisted of intestinal absorption abnormalities and liver functions. Subjects with ALT or AST >60 U/T were also excluded. One subject was excluded due to high levels of ALT and AST. The others (n=60) (Fig. 1) were randomly assigned to treatment groups, minimizing any intentional elements of group assignment, in a double-blind procedure to ensure that the subjects did not know which group they were assigned to (test group or comparative control group) (Fig. 1). The sample size of a minimum 9 per treatment group would provide ≥95% power to detect a difference of Cd values in the serum based on the previous animal study (11).

For heavy metal, lipid peroxidation and antioxidant analysis, the assay was performed prior and subsequent to administration of the test product and the placebo (Fig. 2). All the subjects were instructed to ingest the tablets after breakfast over a 60-day period. The placebo group was comparable in all characteristics to the DML group. Food and water intake was not limited.

By vein puncture with a syringe, blood samples were collected in tubes containing ethylenediaminetetraacetic acid. The tubes were immediately centrifuged at 4°C and 500 × g for 10 min to obtain serum and plasma for measuring Cd, Hg and MDA levels and SOD1 activity. These parameters were measured using routine automated methods. Adverse experiences were rated by the investigators for intensity and their association to the DML extract. Efficacy and safety measurements were conducted at a Hallym University Dongtan Sacred Heart Hospital (Dongtan, South Korea) laboratory and at a Seoul National University College of Veterinary Medicine (Seoul, South Korea) by technicians or graduate students blinded to the placebo and treatment groups.

The measured Hg and Cd levels were calculated into excretion ratio (delta% value), as shown in Fig. 2. The MDA levels and SOD1 activities 60 days after placebo or DML consumption are shown. All the data are expressed as mean ± standard error of the mean. To determine the effects of the DML tablet, differences between the means of the two groups were statistically analyzed using two-way analysis of variance with repeated measures and Bonferroni's post hoc test using GraphPad Prism 5.01 (GraphPad Software, La Jolla, CA, USA) software.

The baseline characteristics and biochemical values of the subjects did not significantly differ between the placebo and DML groups (data not shown). In addition, no significant differences between male and female participants were identified.

Serum Cd levels were increased in the male and female subjects in the placebo group and decreased in the DML group; however, no significant difference between the two groups was identified (P=0.3008) (Fig. 3). Serum Hg levels decreased in all the groups, but serum levels were more prominently decreased in the DML-male and DML-female groups compared to the placebo-male and placebo-female groups, respectively (P=0.1336 and P=0.1937, respectively) (Fig. 3).

Serum MDA levels were increased in the DML-male group compared to the placebo-male group (P=0.3125), while these were decreased in the DML-female group compared to the placebo-female group (P=0.0842) (Fig. 4); however, no significant differences were identified.

Serum SOD1 activity was increased in DML-treated groups, but the SOD1 activity was significantly increased in the DML-female group compared to the placebo-female group (P=0.02031). However, no significant difference in SOD1 activity was observed between the placebo-male and DML-male group (P=0.2031) (Fig. 4).