Participants
In this double-blinded, placebo-controlled trial, 96 T2DM patients were recruited from the endocrinology and metabolism clinics of Golestan Hospital of Ahvaz Jundishapur University of Medical Science, Iran. Fifty participants were selected based on the inclusion criteria (Fig. 1).
The FBS ≥ 126 mg/dl and HbA1c ≥ 6.5% or 2-hglucose (2 hpp) ≥ 200 mg/dl were used as the criteria to define diabetes mellitus (DM) [30]. The inclusion criteria were: male or female, age of 30-60 years old, BMI 18.5-30 kg/m2, and having DM (for more than five years since the diagnosis). The exclusion criteria were having kidney and thyroid diseases, traveling for more than two weeks, pregnancy and lactation, smoking, having consumed alcohol over the past year, using medications such as immunosuppressive agents, insulin, anti-inflammatory drugs, and dietary supplements, and remarkable changes in using medications and treatment of diabetes or considerable changes in diet over the past six months. The compliance of the participants was assessed by counting the remaining tablets. Participants who consumed < 90% of the prescribed tablets were excluded.
The participants were allocated to intervention (n = 25) or placebo (n = 25) groups by a random block design. Accordingly, all the participants were assigned to six groups with four blocks in two steps plus two separate subjects using two codes of A and B according to the block design (AABB, BBAA, ABAB, BABA, ABBA, BABA). Randomization was performed by an independent third person in the health-care system who was not informed of the study. The researchers and patients were not informed of the allocation of participants to the groups (double-blinded). The participants in the intervention and control groups received 250 mg of either melatonin supplement (containing sodium starch glycolate, magnesium stearate, and 3 mg of net melatonin; purchased from the National Vitamin Company, Nature Made, USA) or placebo tablets (containing a few drops of peppermint oil, manufactured in the Faculty of Pharmacy, Ahvaz Jundishapur University of Medical Sciences, Iran). The melatonin and placebo tablets were prescribed for eight weeks as two tablets once a day one hour before bedtime. In terms of size, color, taste, and shape, the placebo and melatonin tablets were matched. The participants were asked to report any probable side effects of supplementation throughout the study. Patients were advised to follow any prescribed treatment and diabetic diet, and not to change their usual physical activity and diet pattern during the study.
Blood pressure (BP), mean arterial pressure (MAP), and pulse pressure (PP)
BP was recorded in a relaxed sitting position on the left arm after at least a 5min rest using a mercury sphygmomanometer. The measurement was repeated twice, and the mean of the two measurements was considered as the BP. MAP and PP were calculated according to the following equations [8, 9]:
MAP(mmHg)=[SBP+(2×DBP)]/3
PP(mmHg)=SBP(mmHg)-DBP(mmHg)
Traditional and new anthropometric indices and nutritional assessment
All anthropometric measurements were performed by a trained nutritionist. Traditional anthropometric indices included height, weight, BMI, WC, and HC. Height (without shoes in a standard situation with a precision of 0.1 cm) and weight (by Seca scales, with light clothes, and a precision of 0.1 kg) were measured. BMI was calculated as weight (Kg)/height (m2). WC as the smallest horizontal girth between the costal and iliac crests was measured using a non-stretchable measuring tape at minimal respiration. Hip circumference, as the greatest circumference at the level of greater trochanters (the widest portion of the hip) on both sides, was also measured.
All calculations for new anthropometric indices (ABSI, AVI, BAI, conicity, LAP, and WHR) were performed according to the following equations [31]:
A 3-day 24-h dietary recall was also obtained for dietary assessment at the beginning and end of the study. The Nutritionist IV software (Axxya Systems, Stafford, Texas, USA), modified for Iranian foods, was used for dietary data analysis.
Atherogenic index of plasma (AIP(
AIP was obtained by dividing the logarithmically transformed ratio of triglycerides (TGs) / high-density lipoprotein cholesterol
All the measurements and calculations were repeated after the intervention at the end of the study.
Blood sampling and biochemical measurements
Blood samples of 2 cc were collected from all the participants after fasting overnight for at least 12 hours, and were centrifuged at 3000 rpm for 5 min. The lipid profile parameters were immediately measured using Pars Azmoon kits (Pars Azmoon, Tehran, Iran). The serum concentrations of TG and HDL-C were determined through an enzymatic procedure using glycerol phosphate oxidase and phosphotungstic acid, respectively.
Statistical analysis
Statistical analyses were performed using SPSS software version 19.0 for Windows (PASW Statistics; SPSS Inc., Chicago, IL, USA). Analyses were conducted via an intention-to-treat (ITT) approach. The Kolmogorov-Smirnov test was used to check the normality of the data. An independent sample t-test and a paired sample t-test were also performed to compare the normal variables between and within groups, respectively, at baseline and after intervention. Post-intervention, the analysis of covariance (ANCOVA) was done to compare the results between the two groups after adjustment for confounding factors. We used two models, a crude model (without adjust) and with adjustment with age, sex, job, education, energy, physical activity, drugs, and disease duration (Model 1). A chi-square test was also used to compare the categorical variables. Data are expressed as mean ± standard deviation (SD) for normal distributions and median (Q1-Q3) for abnormal distributions. A Mann-Whitney U test was performed to compare changes between the two groups after the intervention. P-values < 0.05 were deemed significant.
Based on Ramezanpour et al.'s study [32], with 5% α (confidence interval of 95%) and 80% power of study, and considering SBP as the primary outcome, according to this formula:
the sample size was estimated to be 21 participants for each group; however, considering a 20% withdrawing rate, we selected 25 participants in each group.