Study design and Participants
The study was a randomized, double-blind, placebo-controlled clinical trial; that was conducted in Tabriz, Iran, from September 2018 through May 2019. The target population of the present study was obese women with PCOS. The PCOS was diagnosed by a gynecologist, based on the Rotterdam criteria, which includes suffering from two of the following symptoms: (i) amenorrhea or oligomenorrhea with eight or fewer menstruations in the previous 12 months, (ii) biochemical and/or clinical signs of hyperandrogenism, and (iii) polycystic ovaries on ultrasound examination (> 12 follicles, 2 to 9 mm in diameter and/or increased ovarian volume > 10 mL)[19]. The study population was recruited from the gynecology and infertility clinics of Alzahra Hospital in Tabriz, Iran, or Sheykholrayis Polyclinic in Tabriz, Iran. Medical records of patients in these clinical settings were carefully reviewed. Of those, forty-eight obese women (body mass index (BMI): 30–40 kg/m2) with PCOS and aged 20–45 years, were recruited. Exclusion criteria were as follows: any evidence of thyroid disease, adrenocortical dysfunction, or hyperprolactinemia (Prolactin > 30 mg/mL), being menopause, or pregnant, or lactating; smoking or being exposed to cigarette smoke; having co-morbidity with other gynecologic or endocrine disease, or hepatic, renal, or cardiovascular disease, diabetes and/or impaired glucose tolerance; taking any nutritional or herbal supplements during two months prior to the study, using ovulation induction agents or drugs affecting metabolic or insulin statuses such as statins, thiazolidinediones, corticosteroids, insulin, anti-obesity and anti-diabetic drugs (including Metformin). Comprehensive interviews on the general characteristics of the participants, containing age, family history of PCOS, disease and medication history or previously utilized treatments, and their lifestyle, were conducted at the beginning of the trial. If the patients had adopted a diet and/or a specific physical activity program, or any changes in medications, or experienced any detrimental events during the study, they were withdrawn from this clinical trial. This clinical trial was registered in the Iranian Registry of Clinical Trials (http://www.irct.ir; registration no. IRCT20140907019082N9).
Ethics approval and consent to participate
The eligible patients were given a detailed explanation of the study procedures. Written informed consent was obtained from those willing to participate in the trial. The study protocol was conducted according to the Declaration of Helsinki guideline and approved by the ethics committee of Research Vice-Chancellor of Tabriz University of Medical Sciences (Tabriz, Iran; Ethics code: IR.TBZMED.REC.1397.447).
Sample size
The sample size was calculated based on the previous study [20]. By considering the changes in FBS, with a confidence level of 95%, alpha = 0.05 and power of 90%, a total of 21 participants were calculated for each group, which was increased to 24 to cover a probable dropout rate of 15% (total sample size: 48 obese women with PCOS). For this calculation, Power Analysis and Sample Size Software (PASS; NCSS, LLC, US) version 15 was used.
Study protocol
The study participants were randomly allocated into one of the two experimental groups (in a 1:1 ratio) by a research assistant not otherwise involved in the study, using of the Random allocation software (RAS) and randomized block procedure of size two (age (< 33 vs. ≥33) and BMI (< 35 kg/m2 vs. ≥35 kg/m2)). The researchers and participants were blinded regarding the randomization, allocation, and subjects’ groups until the end of the study and the completion of final analyses. To ensure the blinding in the evaluation process, the person (completely irrelevant to the study) who prepared the supplement packages assigned a three-digit code to each of the treatments. The eligible subjects were assigned to receive 5 g/day of thylakoid-rich spinach extract or matching placebo as 5 g/day raw starch (one sachet before lunch) for 12 weeks. The sachets were completely identical in all other aspects (shape, weight, color). The primary outcomes were changes in anthropometric and metabolic status (FBS, insulin, insulin resistance, and lipid profile). The secondary outcomes were blood pressure, triglyceride glucose (TyG) index, lipid accumulation product (LAP), visceral adiposity index (VAI), and hs-CRP. At first, demographic and clinical questionnaires, a 3-days food record, and international physical activity questionnaire-short form (IPAQ-SF) were completed for all participants. Next, anthropometric assessments and blood pressure measurements were done. All mentioned assessments were performed at baseline and at the end of the study. For biochemical evaluations, blood samples were taken from each patient, after 10–12 h of overnight fasting, at baseline and end of the study. The participants were asked to keep their regular medication (i.e., OCP) and usual levels of physical activity throughout the study period. They were also advised to inform the researchers for any changes in their medical therapy program and also any adverse effects of the supplements.
Intervention protocol
Participants randomly recruited to a 12-week intervention arm consisting of 5 g/day of thylakoid-rich spinach extract powder + low-calorie diet (n = 24) or to a control arm of 5 g/day powdered raw starch as placebo + low-calorie diet (n = 24). The choice of 12-weeks intervention duration and a dose of 5 g/day thylakoid, was based on observed beneficial effects of thylakoid supplementation on obesity status and related metabolic profiles in consumer subjects [21]. All of the participants received a calorie-restricted diet planned by an expert dietitian. For planning this diet, total energy expenditure was calculated based on resting energy expenditure (REE), which was calculated based on the Mifflin equation [22], physical activity level, and thermic effect of food (10% of total energy expenditure). After calculating the daily required energy for each participant, by a 500 kcal deficit from it, individualized diets were designed. The assigned diet contained 30% fat, 55% carbohydrate, and 15% protein. The participants were requested to follow healthy eating recommendations, including changing cooking methods to healthier ways and limiting fast foods, saturated fats, high-fat foods, sugar, sweets, and sugar-sweetened beverages. Food Exchange utilization was thoroughly explained to the participants, and replacing the foods they did not have access to, by the foods of equal calorie from the corresponding food groups was instructed. Adherence to the recommended diet was evaluated using a 3-day food diary (2 weekdays and 1 weekend day) at baseline and end of the study. Daily intakes of macro- and micro-nutrients were analyzed by Nutritionist IV software (First Databank Inc., Hearst Corp., San Bruno, CA, USA).
Preparation of spinach thylakoids and placebo
Fresh baby spinach leaves (Spinacia oleracea) were used to prepare of thylakoid membranes according to the previously registered protocols [16, 23, 24]. The required spinach was collected from Tabriz, East Azerbaijan Province, Iran in spring, 2018; some plant samples were delivered to the Herbarium Center of the Faculty of Pharmacy, Tabriz University of Medical Sciences. The scientific name of the collected specimen is Spinacia oleracea L. belonging to the Oleracea family with the herbarium number TBZ-fph-1898. The thylakoid supplement used in this investigation was prepared based on the method described by Emerk et al. [24], at an experimental scale in the Synthesis Laboratory of Drug Applied Research Center, Tabriz University of Medical Sciences. Fresh spinach leaves after removing the stems and veins, were washed and drenched. 1000 g spinach leaves were homogenized with 1250 ml water in a blender and filtered through four layers of Monodur polyester mesh (20 µm). This obtained filtrate was diluted 10 times with distilled water, and its pH adjusted to 4.7 with Hydrochloric acid (HCl). PH 4.7 is the isoelectric point of the thylakoids, and maximum precipitation occurs at this pH. The thylakoids flocculated, and a green precipitate with a clear, a bit yellowish supernatant was obtained after 4 h standing in the cold (-4oC). The supernatant was removed, and the green precipitate was collected from the filtrate thylakoids at pH 4.7 and washed in water by repeated centrifugation; the precipitation was repeated at the same pH. The washed thylakoids were collected, and after adjusting to the desired pH (pH 7.0), the final sediments freeze-dried to obtain a green thylakoid powder. Large scale production, of this freeze-dried thylakoid powder, was conducted by the Iran Darook Pharmaceutical Co., Tehran, Iran. Placebo consisted of corn starch, which was colored in edible green color, and like thylakoid powder, flavored with kiwifruit essence. Therefore, green powders with kiwifruit flavor were made of thylakoid or placebo, which appearances (shape, size, and color) and also, flavor were similar in placebo and thylakoid. Next, they were packed in completely identical sachets; and each sachet contained 5 g of thylakoid or 5 g of starch. The contents of the sachets were dissolved in a glass of water and drunk before lunch. Both participants and researchers were unaware of group assignment. Packages were coded; and distributed monthly by a third person, who was irrelevant to the study. For reminding the consumption of the supplements, a chart was designed for the participants to complete and return in each visit. Participants received a brief daily cell phone reminder and weekly a phone call to take the supplement and to minimize withdrawal and ensure their adherence to the study protocol. The participants were requested to return the remaining sachets at each visit; counting these sachets enabled us to evaluate compliance. Consuming ≥ 80% of the supplements was considered compliance.
Assessment of anthropometric variables and blood pressure
The body weight, height, waist, and hip circumferences of the participants were measured at baseline and at the end of the study. Body weight (to the nearest 0.1 kg) and height (to the nearest 0.1 cm) were measured using a Seca digital scale (Hamburg, Germany), in an overnight fasting state with minimal clothing and no shoes on. BMI was calculated as weight (kg) divided by height squared (m2). The hip circumference (HC) was measured at the level of the maximum extension of the buttocks, and waist circumference (WC) was measured at the midpoint between the lower ribs and the iliac crest. Waist to height ratio (WHtR) was calculated as WC (cm) divided by height (cm). Waist to hip ratio (WHR) was calculated as WC (cm) divided by HC (cm). Blood pressure was measured in a sitting position three times after resting for 10 min, by a digital sphygmomanometer (Omron M3, Kyoto, Japan), and the mean of three measurements was reported as the final systolic and diastolic blood pressure.
Blood sampling and measurements of biochemical parameters
The venous blood sample (10 ml) was drawn from each participant at baseline and end of the study after overnight fasting, between 8:00 a.m. and 9:00 a.m. during the early follicular phase (d 2–5) of a spontaneous or P-induced menstrual cycle. Blood samples immediately centrifuged at 3500 rpm for 10 minutes to the separation of serum samples from whole blood. Serum samples were used to measure lipid profile, FBS, insulin, testosterone, sex hormone-binding globulin (SHBG), and hs-CRP. The biochemical parameters, including a lipid profile, glycemic parameters, and hs-CRP were immediately measured after sample collection, and the remaining serum was frozen immediately at − 80 °C, until the end of the study. Serum FBS, triglycerides (TG), total cholesterol (TC), and high-density lipoprotein cholesterol (HDL-C) concentrations were measured through enzymatic methods by an auto-analyzer (Hitachi-917, Tokyo, Japan) using the colorimetric technique, by commercial kits (Pars-Azmoon Co., Tehran, Iran). The Friedewald equation was used to calculate low-density lipoprotein cholesterol (LDL-C) [25]. Serum insulin level was measured by chemiluminescence (IMMULITE 2000, SIEMENS); and the homeostatic model of assessment for insulin resistance (HOMA-IR) was calculated, on the basis of the suggested formula [26]. Serum hs-CRP concentration was determined using an immunoturbidimetric assay (Pars Azmoon Co., Tehran, Iran). Testosterone and SHBG concentrations were determined using ELISA kits (Bioassay Technology Laboratory, Shanghai Korean Biotech, Shanghai City, China) according to the manufacturer’s instructions with inter- and intra-assay coefficient variances (CVs) lower than 7%. The free androgen index (FAI) was calculated based on suggested formulas [27]. TyG index, LAP, and VAI were calculated based on suggested formulas as follows [28]:
TyG = Ln [TG (mg/dl) × FBS (mg/dl)/2],
LAP = [WC (cm) -58] × [TG (mmol/L)] ,
VAI = [WC(cm)/36.58 + (1.89 × BMI)] × (TG(mmol/L)/0.81) × (1.52/HDL-C(mmol/L)).
Statistical analysis
All statistical analyses were performed using SPSS version 23 (SPSS Inc., Chicago, IL, USA). The Kolmogorov-Smirnov test was performed to determine the normality of data distribution. Distribution of data was expressed as mean (SD) for normally distributed and median (percentiles 25 and 75) for not normally distributed quantitative data and frequency (percent) for qualitative data. To compare the two groups at the baseline, independent sample t-tests or Mann–Whitney or chi-squared tests were used. Assessments of differences within the group were made by paired-samples t-tests or nonparametric Wilcoxon signed-rank test, and sign test. A comparison of the two groups at the end of the study was completed by the analysis of covariance (ANCOVA) after adjusting for the baseline parameters and covariates. Post hoc paired comparisons were made by using a Sidak test. Results with P values of < 0.05 were considered statistically significant.