Study design and participants
This study was a randomized, double-blind, placebo-controlled clinical trial, conducted from September 2018 through May 2019, in Tabriz, Iran. This clinical trial was registered in the Iranian Registry of Clinical Trials (http://www.irct.ir; registration no. IRCT20140907019082N9). The target population of the present study was obese women with PCOS who were recruited from Sheykholrayis Polyclinic and the gynecology and infertility clinics of Alzahra Hospital in Tabriz, Iran. All patients were taking oral contraceptive pills (OCPs) as a routine medical therapy under the supervision of their physician. Forty-eight obese women (body mass index (BMI): 30–40 kg/m2) aged 20–45 years were enrolled. They were diagnosed with PCOS based on Rotterdam criteria [20]. Two of the following three features have to be present for the PCOS diagnosis: (i) oligomenorrhea with eight or fewer menstruations in the previous 12 months or amenorrhea, (ii) clinical and/or biochemical 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). The diagnosis of PCOS also required that there was no evidence of thyroid disease, adrenocortical dysfunction, or hyperprolactinemia (prolactin > 30 mg/mL). The main exclusion criteria for participation in this study were: menopause, or pregnancy, or lactation; smoking or being exposed to cigarette smoke (passive smoking); 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 status such as statins, thiazolidinediones, corticosteroids, insulin, anti-obesity and anti-diabetic drugs (i.e., Metformin, sulfonylureas,..). 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activity program, or any changes in medications, or experienced any detrimental events during the study, they were withdrawn from the clinical trial.
Ethics approval and consent to participate
The study 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); After being given a full explanation of the study procedures, written informed consent was obtained from all patients.
Sample size
The sample size for the study was calculated based on the results mean (standard deviation; SD) for FBG as reported by Stenblom et al. [21], with a confidence interval (CI) of 95%, and power of 90% in two-sided tests using power analysis and sample size software (PASS; NCSS, LLC, US) version 15, the sample size was 21 per group, which was increased to 24, considering a probable 15% dropout rate.
Study protocol
The study participants were randomly allocated into one of the two experimental groups (1:1), by the use of the Random allocation software (RAS) and randomized block procedure of size 2 (age (< 33 vs. ≥33) and BMI (< 35 kg/m2 vs. ≥35 kg/m2)). Randomization and allocation were concealed to the researchers and participants until the final analyses were completed. The eligible subjects were assigned to receive 5 g/day of thylakoid-rich spinach extract or matching placebo as 5 g/day of raw corn starch (one sachet, 30 min before lunch) for 12 weeks. The sachets were completely identical in all other aspects. The primary outcomes were changes in anthropometric measurements and metabolic status. At first, demographic and clinical questionnaires were completed for all participants. They were asked to complete a 3-day food record questionnaire including, two weekdays and one weekend day in the week just prior to the study and during weeks 6 and 12. The International Physical Activity Questionnaire - short form (IPAQ-SF) was also used to assess the physical activity level, at baseline, week 6, and week 12. Next, anthropometric measurements, bioelectric impedance analysis (BIA), and blood pressure measurements were performed. After 12-h overnight fasting, blood was taken from each patient for biochemical evaluations before and after the intervention. 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 of any changes in their medical therapy program and any adverse effects of the supplements.
Intervention protocol
Participants were 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 cornstarch as placebo + low-calorie diet (n = 24). The choice of 12-week intervention duration and a dose of 5 g/ day thylakoid was based on previously reported beneficial effects of thylakoid supplementation on obesity status and related metabolic profiles in participants who received thylakoid supplements for 12 weeks [16, 18]. All the participants received a calorie-restricted diet designed by a dietitian. For planning the calorie-restricted 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, a calorie deficit of 500 kcal per day was made for each person. Macronutrient distribution was organized as 30, 15, and 55% of energy from fat, protein, and carbohydrates, respectively. This distribution of macronutrients, as moderate and conventional ratios instead of higher or lower ratios (such as low fat or low carbohydrate or high protein), was planned based on existing clinical nutrition guidelines and aimed to increase adherence to the prescribed diet [23]. The participants were asked to pursue healthy eating recommendations, including substituting solid fats and animal fats with non-hydrogenated vegetable oils as well as limiting added sugar, desserts, sugar-sweetened beverages, saturated and trans fats, and fast foods as well as changing cooking methods to healthier ways. The application of food exchange lists was thoroughly explained to the participants, and they were advised of the alternative options with equal calories when a particular item from the corresponding food group is not affordable or available. Adherence to the recommended diet was evaluated using dietary intake records for 3 days (2 weekdays and 1 weekend) at the beginning, middle, and end of the study. Daily intakes of macro- and micro-nutrients were calculated by analyzing food data using nutritionist IV software (First Databank, San Bruno, CA).
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 [17, 24, 25]. 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. [25], 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 soaked in cold water and washed. Spinach leaves (1000 g) 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 (− 4 °C). 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. Corn starch, a white, tasteless, safe, non-toxic, non-irritant, and non-allergenic odorless powder, is the most frequently used substance in the food and pharmaceutical industries, as an inert (inactive) substance without any therapeutic effect. Due to its versatility and flexibility in application, and ease of modification, we modified it to green powder with Kiwi odor, exactly like that extracted thylakoid from spinach. Therefore, green powder of thylakoid or placebo was made, with kiwifruit flavor and identical appearances (shape, size, and color). Next, they were packed in identical sachets with each sachet containing 5 g of thylakoid or 5 g of cornstarch powder. The contents of the sachets were dissolved in a glass of water and consumed 30 min before lunch. Packages were coded and distributed monthly by a third person who was not involved in any other aspects of the study. To remind participants to consume their supplements, a supplement consumption chart was provided. This chart was required to be returned at each visit to ensure compliance. Participants received a daily text message reminder and weekly a phone call reminds them to take the supplement. This helped to minimize study withdrawal and ensure adherence to the study protocol. The participants were asked to return the remaining sachets at each visit; counting these sachets allowed us to evaluate compliance. Consuming ≥80% of the supplements was considered compliant.
Anthropometric and body composition measurements
The height, weight, waist and hip circumferences, and body composition of the participants were measured at baseline, mid, and end of the study. BMI was calculated as weight (kg) divided by height in meters squared (m2). Body composition was assessed using Tanita MC-780 S MA (Amsterdam, the Netherlands). Tanita MC-780 MA is a body composition analyzer based on bioelectrical impedance analysis, which uses BIA technology, as quick, non-invasive, and one of the most thorough and reliable ways, to measure body composition [26].
Blood sampling and biochemical measurements
At pre- and post-intervention, a 10 ml venous blood sample was drawn from each participant. Blood sampling was done between 8: 00 a.m. and 9:00 a.m. after a 12-h overnight fasting and resting in bed 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 min, and serum samples were separated from whole blood and were frozen immediately at − 80 °C until assay time. Measurements of serum non-esterified fatty acids (NEFA), FBG, and insulin were done on the sampling day. Serum FBG and NEFA were measured through enzymatic methods using the colorimetric technique, by commercial kits (Pars-Azmoon Co., Tehran, Iran; and Biovision Research Product, USA, respectively) by an auto-analyzer (Hitachi-917, Tokyo, Japan). Serum insulin level was measured by chemiluminescence (IMMULITE 2000, SIEMENS); and the homeostatic model of assessment for insulin resistance (HOMA-IR) and β cell function (HOMA-B) and quantitative insulin sensitivity check index (QUICKI) were calculated on the basis of suggested formulas [27]. Hormonal profiles including LH, FSH, dehydroepiandrosterone sulfate (DHEAS), testosterone, and SHBG, 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 [28].
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 quantitative data, and frequency (percent) for qualitative data. To compare the two groups at the baseline, independent sample t-tests, Mann–Whitney, and 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 (changes in baseline values of evaluated parameters and potential covariates including age, baseline BMI, and the changes in weight, calorie intake, and physical activity level during the study). Post hoc paired comparisons were made by using a Sidak test. Results with P values of < 0.05 were considered statistically significant.