Gut Microbiota and the Polycystic Ovary Syndrome: Influence of Sex, Sex Hormones, and Obesity

Gut Microbiota and the Polycystic Ovary Syndrome: Influence of Sex, Sex Hormones, and Obesity Abstract Context Gut microbiota play a major role in health and disease by influencing physiology, metabolism, nutrition, and immune function. Objective To evaluate the composition of gut microbiota in women with polycystic ovary syndrome (PCOS), focusing on the influence of sex, sex hormones and obesity on the associations found. Design Cross-sectional study. Setting Academic hospital. Participants We recruited 15 women with PCOS, 16 nonhyperandrogenic control women, and 15 control men. Participants were classified as nonobese (<30 kg/m2) or obese (≥30 kg/m2) according to their body mass index. Interventions Standardization of diet for 3 consecutive days (at least 300 g of carbohydrates per day) followed by fecal sampling and a standard 75-g oral glucose tolerance test. Main Outcome Measures Analysis of bacterial abundance and composition of gut microbiota by massive sequencing of 16S ribosomal DNA amplicons in a MiSeq platform (Illumina). Results α Bacterial diversity was reduced in women compared with men, and β diversity was reduced particularly in obese patients with PCOS. Women with PCOS presented with specific abnormalities in gut microbiota consisting of an increased abundance of the Catenibacterium and Kandleria genera. When all participants as a whole were considered, indexes of bacterial diversity and the abundance of several bacterial genera correlated positively with serum androgen concentrations and negatively with estradiol levels. Conclusions The diversity and composition of the gut microbiota of young adults are influenced by the combined effects of sex, sex hormone concentrations, and obesity, presenting with specific abnormalities in women with PCOS. Polycystic ovary syndrome (PCOS) is now considered a complex multigenic disorder with strong environmental influences (1). Several metabolic disorders are associated with PCOS, including insulin resistance, diabetes, and obesity (2). The latter exerts a major effect on the metabolic associations and complications of the syndrome (3). Among other factors, obesity is clearly related to the infertility of PCOS and increases the risk for metabolic syndrome and the clustering of cardiovascular risk factors in these women (4, 5). During the last two decades, several studies revealed the major influence of gut microbiota on physiology, metabolism, nutrition, and immune function (6). Germ-free mice—which therefore lack gut microbiota—have reduced adiposity and improved tolerance to glucose and insulin when compared with colonized counterparts (7). Moreover, fecal transplantation of the obese gut microbiome into germ-free mice promotes increase in adiposity, resulting in obese phenotype (8). Moreover, several studies have reported the gut microbiota particularities related to not just obesity but also diabetes and the metabolic syndrome (9). The possible influence of host sex and/or sex steroids on the composition of gut microbiota remains unclear at present (10–13). The current study aims to evaluate the composition of gut microbiota in women with PCOS, focusing on the possible influence of sex, sex hormones, obesity, and their interactions, on any putative association. Materials and Methods Participants A total of 46 young adult volunteers (15 patients with PCOS, 16 nonhyperandrogenic women, and 15 healthy men) were recruited. Participants had similar mean body mass index (BMI) and age. They were classified as nonobese (<30 kg/m2) or obese (≥30 kg/m2) according to their BMI. The diagnosis of PCOS was based on the presence of clinical and/or biochemical hyperandrogenism, oligo-ovulation, and exclusion of secondary causes (14, 15). Before inclusion, the participants had no history of obesity-associated comorbidities, such as impaired glucose tolerance, hypertension, cardiovascular events, or sleep apnea. Smokers were excluded. None of the individuals had received treatment with oral contraceptives, antiandrogens, lipid-lowering or blood pressure–lowering drugs, or insulin sensitizers within the previous 3 months. None of the women in the control group had signs or symptoms of hyperandrogenism, history of menstrual dysfunction, or infertility. Men with androgen deficiency, as defined by a serum total testosterone concentration <8 nmol/L, were excluded. The Ethics Committee of Hospital Universitario Ramón y Cajal approved the study protocol, and all participants gave written informed consent. Study design All individuals underwent a comprehensive clinical, anthropometric, and physical evaluation. Participants were advised to follow a diet with 300 g of carbohydrate per day for 3 days before sampling to avoid false-positive results on the 75-g oral glucose load performed thereafter; the latter test was used not only for research purposes but also to check patients for disorders of glucose tolerance. In women, the study was performed during days 3 to 10 of the follicular phase of the menstrual cycle or in amenorrhea after exclusion of pregnancy. Occurrence of infectious disease, including vaginal candidiasis and periodontal disease, during the previous 3 months and recent use of antibiotics were recorded. Laboratory measurements Baseline blood samples were collected for hormonal and metabolic measurements. Plasma and serum aliquots were frozen at −80°C until use. Technical characteristics of the assays used for laboratory measurements have been described previously (16). The homeostasis model assessment of insulin resistance (HOMA-IR) was calculated from fasting glucose and insulin concentrations (17), and the composite insulin sensitivity index was estimated from the glucose and insulin concentrations measured during the oral glucose tolerance test (18). DNA extraction from fecal sample and next-generation sequencing Patients were instructed to collect fecal samples into sterile plastic bottle topped with a screw cap. The samples were immediately divided into aliquots and stored at −80°C after collection until analysis was performed. Samples were defrosted slowly at 4°C to avoid DNA degradation and further thawed at room temperature. An aliquot of 0.5 g of feces was solubilized in 5 mL of distilled water and centrifuged at 157g and 4°C for 10 minutes to eliminate undigested fiber. Total DNA from 1 mL of the supernatant was obtained with the QIAamp DNA Mini KitPellet commercial kit (Qiagen, Hilden, Germany). After DNA samples were checked for quality and concentration, they were sent to Fundacio per al Foment de la Investigacio i Biomedica (FISABIO, Valencia, Spain) for massive sequencing of 16S ribosomal DNA (rDNA) amplicons in a MiSeq platform (Illumina). Bioinformatics analysis of the readings was conducted at FISABIO, and only sequences longer than 200 nucleotides were considered. The bacterial diversity was estimated by using Shannon, Chao 1, Jaccard, and Sorensen indexes with their SDs. Taxonomic affiliations were assigned by using the RDP_classifier from the Ribosomal Database Project (RDP), and readings with RDP score value <0.8 were assigned to the upper taxonomic rank, leaving the last rank as unidentified. Statistical analysis Data are expressed as mean ± SD unless otherwise stated. We tested the normality of distribution by using the Kolmogorov-Smirnov test. Two-step transformations were applied as needed to ensure normal distribution of data (19). General linear models were used to determine the influence of group and obesity on hormonal, metabolic, and inflammatory variables and indexes of α and β diversity. The differences between the groups of participants in bacterial composition were evaluated by the Kruskal-Wallis test followed by Dunn nonparametric post hoc test (20, 21). The comparisons between nonobese and obese subgroups on bacterial composition and abundance were evaluated by Mann-Whitney U test. The interaction between groups and obesity were evaluated, in each subgroup of participants separately, by a Mann-Whitney U test. We applied a Bonferroni correction to the level of statistical significance to take into account multiple comparisons (P < 0.0125 was considered to indicate a statistically significant difference). Categorical variables were analyzed by a χ2 or Fisher exact test as appropriate. Relationships between described genera and clinical and biochemical variables were analyzed by Spearman correlation analysis. We used SPSS statistical software, version 18.0 (IBM, Chicago, IL) and PAST for statistical analyses of biodiversity. A P value <0.05 was considered to indicate statistical significance with the exception mentioned previously. Results Clinical, hormonal, and metabolic variables in patients with PCOS and controls The clinical, hormonal, and metabolic data are summarized in Table 1. Age and BMI did not differ between groups. The hirsutism score was higher in patients with PCOS than in control women (hirsutism score was not assessed in men). Besides their higher total testosterone, free testosterone and dehydroepiandrosterone-sulfate (DHEAS) levels, men also had higher waist-to-hip ratio and ratio of free testosterone to free estradiol than did control women, who, in contrast, had higher SHBG and free estradiol than did men. Women with PCOS had increased concentrations of free testosterone, ratio of free testosterone to free estradiol, and androstenedione compared with control women. In addition, women with PCOS presented lower levels of total and free testosterone and ratio of free testosterone to free estradiol, but higher SHBG and androstenedione levels, than did men. Table 1. Clinical, Hormonal, and Metabolic Characteristics of Participants Characteristic Control Women Women With PCOS Men Group P Value Obesity P Value Interaction P Value Nonobese (n = 8) Obese (n = 8) Nonobese (n = 7) Obese (n = 8) Nonobese (n = 8) Obese (n = 7) Age, y 27.3 ± 4.0 27.3 ± 6.5 23.0 ± 7.9 29.9 ± 5.1 23.3 ± 3.3 23.6 ± 3.2 0.116 0.129 0.138 BMI, kg/m2 23.4 ± 1.9 35.9 ± 4.2 24.4 ± 2.8 37.0 ± 4.8 22.9 ± 2.0 34.6 ± 3.4 0.309 <0.001 0.888 Waist-to-hip ratioa,b 0.74 ± 0.1 0.83 ± 0.12 0.74 ± 0.04 0.85 ± 0.1 0.84 ± 0.04 0.9 ± 0.1 0.005 <0.001 0.542 Hirsutism score 1.3 ± 1.3 1.8 ± 1.2 10.1 ± 5.1 9.3 ± 4.5 NA <0.001 0.887 0.589 Total testosterone, nmol/La,b 1.6 ± 0.3 2.0 ± 0.5 2.5 0.7 2.4 1.0 19.2 ± 3.4 16.6 ± 2.8 <0.001 0.993 0.134 FT, pmol/La,b 20 ± 5 31 ± 8 38 ± 10 45 ± 24 462 ± 106 464 ± 81 <0.001 0.012 0.022 Total estradiol, pmol/La,b 153 ± 66 276 ± 200 188 ± 230 149 ± 49 64 ± 13 95.8 ± 25.3 <0.001 0.024 0.598 FE, pmol/La 2.7 ± 1.2 5.3 ± 3 3.9 ± 4.9 3.5 ± 1.4 1.7 ± 0.5 2.7 ± 0.7 0.027 0.003 0.549 Ratio FT/FEa,b,c 8.3 ± 2.6 7.5 ± 4.6 19 ± 12.7 14 ± 5.4 286.2 ± 60.4 179.0 ± 39.3 <0.001 0.087 0.272 SHBG, nmol/La,b,c 60 ± 24 43 ± 14 44 ± 18 32 ± 13 28 ± 10 18 ± 5 <0.001 0.007 0.788 DHEAS, mmol/La 5.1 ± 0.9 5.3 ± 1.5 8.5 ± 3.9 5.7 ± 2.1 6.5 ± 1.8 8.5 ± 3.2 0.025 0.798 0.027 Androstenedione, nmol/Lb,c 9.1 ± 3.1 9.4 ± 2.8 16.1 ± 3.5 13.3 ± 5.9 7.3 ± 1.8 9.8 ± 4.2 <0.001 0.992 0.175 Total cholesterol, mmol/L 4.5 ± 0.9 4.7 ± 1.2 4.4 ± 1.2 4.4 ± 1.0 3.9 ± 0.6 4.61 ± 0.75 0.635 0.263 0.607 LDL-cholesterol, mmol/L 2.8 ± 0.7 2.8 ± 0.4 2.6 ± 1.0 2.6 ± 1. 2.4 ± 0.5 3.1 ± 0.6 0.811 0.263 0.291 HDL-cholesterol, mmol/La,b 1.3 ± 0.3 1.2 ± 0.2 1.4 ± 0.2 1.2 ± 0.2 1.2 ± 0.2 1.0 ± 0.1 0.011 0.020 0.619 Triglycerides, mmol/L 0.8 ± 0.2 0.9 ± 0.4 0.8 ± 0.2 1.2 ± 0.6 0.9 ± 0.3 1.2 ± 0.6 0.423 0.055 0.755 Fasting glucose, mmol/Lb,c 4.7 ± 0.3 5.6 ± 0.4 4.8 ± 0.4 4.9 ± 0.4 5.3 ± 0.5 5.3 ± 0.5 0.014 0.014 0.039 Fasting insulin, pmol/L 53 ± 28 92 ± 45 47 ± 35 83 ± 28 37 ± 14 76 ± 28 0.383 <0.001 0.969 HOMA-IR 1.6 ± 0.9 3.3 ± 1.6 1.5 ± 1.1 2.6 ± 0.9 1.3 ± 0.5 2.6 ± 1.0 0.348 <0.001 0.780 Insulin sensitivity index 7.0 ± 2.7 3.3 ± 1.2 8.2 ± 5.2 3.5 ± 1.4 7.6 ± 2.9 3.6 ± 1.7 0.899 <0.001 0.997 Antibiotics, n (%) 0 (0) 0 (0) 0 (0) 2 (25) 1 (12.5) 0 (0) 0.217 1.000 NA Periodontal or dental disease, n (%) 0 (0) 1 (12.5) 0 (0) 0 (0) 0 (0) 0 (0) 0.434 1.000 NA Urinary tract infection, n (%) 0 (0) 0 (0) 0 (0) 1 (12.5) 0 (0) 0 (0) 0.434 1.000 NA Vaginal candidiasis or infection, n (%) 0 (0) 0 (0) 0 (0) 1 (12.5) NA 0.396 1.000 NA Characteristic Control Women Women With PCOS Men Group P Value Obesity P Value Interaction P Value Nonobese (n = 8) Obese (n = 8) Nonobese (n = 7) Obese (n = 8) Nonobese (n = 8) Obese (n = 7) Age, y 27.3 ± 4.0 27.3 ± 6.5 23.0 ± 7.9 29.9 ± 5.1 23.3 ± 3.3 23.6 ± 3.2 0.116 0.129 0.138 BMI, kg/m2 23.4 ± 1.9 35.9 ± 4.2 24.4 ± 2.8 37.0 ± 4.8 22.9 ± 2.0 34.6 ± 3.4 0.309 <0.001 0.888 Waist-to-hip ratioa,b 0.74 ± 0.1 0.83 ± 0.12 0.74 ± 0.04 0.85 ± 0.1 0.84 ± 0.04 0.9 ± 0.1 0.005 <0.001 0.542 Hirsutism score 1.3 ± 1.3 1.8 ± 1.2 10.1 ± 5.1 9.3 ± 4.5 NA <0.001 0.887 0.589 Total testosterone, nmol/La,b 1.6 ± 0.3 2.0 ± 0.5 2.5 0.7 2.4 1.0 19.2 ± 3.4 16.6 ± 2.8 <0.001 0.993 0.134 FT, pmol/La,b 20 ± 5 31 ± 8 38 ± 10 45 ± 24 462 ± 106 464 ± 81 <0.001 0.012 0.022 Total estradiol, pmol/La,b 153 ± 66 276 ± 200 188 ± 230 149 ± 49 64 ± 13 95.8 ± 25.3 <0.001 0.024 0.598 FE, pmol/La 2.7 ± 1.2 5.3 ± 3 3.9 ± 4.9 3.5 ± 1.4 1.7 ± 0.5 2.7 ± 0.7 0.027 0.003 0.549 Ratio FT/FEa,b,c 8.3 ± 2.6 7.5 ± 4.6 19 ± 12.7 14 ± 5.4 286.2 ± 60.4 179.0 ± 39.3 <0.001 0.087 0.272 SHBG, nmol/La,b,c 60 ± 24 43 ± 14 44 ± 18 32 ± 13 28 ± 10 18 ± 5 <0.001 0.007 0.788 DHEAS, mmol/La 5.1 ± 0.9 5.3 ± 1.5 8.5 ± 3.9 5.7 ± 2.1 6.5 ± 1.8 8.5 ± 3.2 0.025 0.798 0.027 Androstenedione, nmol/Lb,c 9.1 ± 3.1 9.4 ± 2.8 16.1 ± 3.5 13.3 ± 5.9 7.3 ± 1.8 9.8 ± 4.2 <0.001 0.992 0.175 Total cholesterol, mmol/L 4.5 ± 0.9 4.7 ± 1.2 4.4 ± 1.2 4.4 ± 1.0 3.9 ± 0.6 4.61 ± 0.75 0.635 0.263 0.607 LDL-cholesterol, mmol/L 2.8 ± 0.7 2.8 ± 0.4 2.6 ± 1.0 2.6 ± 1. 2.4 ± 0.5 3.1 ± 0.6 0.811 0.263 0.291 HDL-cholesterol, mmol/La,b 1.3 ± 0.3 1.2 ± 0.2 1.4 ± 0.2 1.2 ± 0.2 1.2 ± 0.2 1.0 ± 0.1 0.011 0.020 0.619 Triglycerides, mmol/L 0.8 ± 0.2 0.9 ± 0.4 0.8 ± 0.2 1.2 ± 0.6 0.9 ± 0.3 1.2 ± 0.6 0.423 0.055 0.755 Fasting glucose, mmol/Lb,c 4.7 ± 0.3 5.6 ± 0.4 4.8 ± 0.4 4.9 ± 0.4 5.3 ± 0.5 5.3 ± 0.5 0.014 0.014 0.039 Fasting insulin, pmol/L 53 ± 28 92 ± 45 47 ± 35 83 ± 28 37 ± 14 76 ± 28 0.383 <0.001 0.969 HOMA-IR 1.6 ± 0.9 3.3 ± 1.6 1.5 ± 1.1 2.6 ± 0.9 1.3 ± 0.5 2.6 ± 1.0 0.348 <0.001 0.780 Insulin sensitivity index 7.0 ± 2.7 3.3 ± 1.2 8.2 ± 5.2 3.5 ± 1.4 7.6 ± 2.9 3.6 ± 1.7 0.899 <0.001 0.997 Antibiotics, n (%) 0 (0) 0 (0) 0 (0) 2 (25) 1 (12.5) 0 (0) 0.217 1.000 NA Periodontal or dental disease, n (%) 0 (0) 1 (12.5) 0 (0) 0 (0) 0 (0) 0 (0) 0.434 1.000 NA Urinary tract infection, n (%) 0 (0) 0 (0) 0 (0) 1 (12.5) 0 (0) 0 (0) 0.434 1.000 NA Vaginal candidiasis or infection, n (%) 0 (0) 0 (0) 0 (0) 1 (12.5) NA 0.396 1.000 NA Data are means ± SD or counts (percentage). For categorical variables, differences between groups were analyzed by χ2 tests. The effects of group and obesity on continuous variables were analyzed by a two-way general linear model after application of two-step transformation as needed to ensure a normal distribution of the variables. Abbreviations: FE, free estradiol; FT, free testosterone; NA, not applicable. a P < 0.05 for the differences between men and control women, independently of obesity. b P < 0.05 for the differences between men and women with PCOS, independently of obesity. c P < 0.05 for the differences between women with PCOS and control women, independently of obesity. View Large Table 1. Clinical, Hormonal, and Metabolic Characteristics of Participants Characteristic Control Women Women With PCOS Men Group P Value Obesity P Value Interaction P Value Nonobese (n = 8) Obese (n = 8) Nonobese (n = 7) Obese (n = 8) Nonobese (n = 8) Obese (n = 7) Age, y 27.3 ± 4.0 27.3 ± 6.5 23.0 ± 7.9 29.9 ± 5.1 23.3 ± 3.3 23.6 ± 3.2 0.116 0.129 0.138 BMI, kg/m2 23.4 ± 1.9 35.9 ± 4.2 24.4 ± 2.8 37.0 ± 4.8 22.9 ± 2.0 34.6 ± 3.4 0.309 <0.001 0.888 Waist-to-hip ratioa,b 0.74 ± 0.1 0.83 ± 0.12 0.74 ± 0.04 0.85 ± 0.1 0.84 ± 0.04 0.9 ± 0.1 0.005 <0.001 0.542 Hirsutism score 1.3 ± 1.3 1.8 ± 1.2 10.1 ± 5.1 9.3 ± 4.5 NA <0.001 0.887 0.589 Total testosterone, nmol/La,b 1.6 ± 0.3 2.0 ± 0.5 2.5 0.7 2.4 1.0 19.2 ± 3.4 16.6 ± 2.8 <0.001 0.993 0.134 FT, pmol/La,b 20 ± 5 31 ± 8 38 ± 10 45 ± 24 462 ± 106 464 ± 81 <0.001 0.012 0.022 Total estradiol, pmol/La,b 153 ± 66 276 ± 200 188 ± 230 149 ± 49 64 ± 13 95.8 ± 25.3 <0.001 0.024 0.598 FE, pmol/La 2.7 ± 1.2 5.3 ± 3 3.9 ± 4.9 3.5 ± 1.4 1.7 ± 0.5 2.7 ± 0.7 0.027 0.003 0.549 Ratio FT/FEa,b,c 8.3 ± 2.6 7.5 ± 4.6 19 ± 12.7 14 ± 5.4 286.2 ± 60.4 179.0 ± 39.3 <0.001 0.087 0.272 SHBG, nmol/La,b,c 60 ± 24 43 ± 14 44 ± 18 32 ± 13 28 ± 10 18 ± 5 <0.001 0.007 0.788 DHEAS, mmol/La 5.1 ± 0.9 5.3 ± 1.5 8.5 ± 3.9 5.7 ± 2.1 6.5 ± 1.8 8.5 ± 3.2 0.025 0.798 0.027 Androstenedione, nmol/Lb,c 9.1 ± 3.1 9.4 ± 2.8 16.1 ± 3.5 13.3 ± 5.9 7.3 ± 1.8 9.8 ± 4.2 <0.001 0.992 0.175 Total cholesterol, mmol/L 4.5 ± 0.9 4.7 ± 1.2 4.4 ± 1.2 4.4 ± 1.0 3.9 ± 0.6 4.61 ± 0.75 0.635 0.263 0.607 LDL-cholesterol, mmol/L 2.8 ± 0.7 2.8 ± 0.4 2.6 ± 1.0 2.6 ± 1. 2.4 ± 0.5 3.1 ± 0.6 0.811 0.263 0.291 HDL-cholesterol, mmol/La,b 1.3 ± 0.3 1.2 ± 0.2 1.4 ± 0.2 1.2 ± 0.2 1.2 ± 0.2 1.0 ± 0.1 0.011 0.020 0.619 Triglycerides, mmol/L 0.8 ± 0.2 0.9 ± 0.4 0.8 ± 0.2 1.2 ± 0.6 0.9 ± 0.3 1.2 ± 0.6 0.423 0.055 0.755 Fasting glucose, mmol/Lb,c 4.7 ± 0.3 5.6 ± 0.4 4.8 ± 0.4 4.9 ± 0.4 5.3 ± 0.5 5.3 ± 0.5 0.014 0.014 0.039 Fasting insulin, pmol/L 53 ± 28 92 ± 45 47 ± 35 83 ± 28 37 ± 14 76 ± 28 0.383 <0.001 0.969 HOMA-IR 1.6 ± 0.9 3.3 ± 1.6 1.5 ± 1.1 2.6 ± 0.9 1.3 ± 0.5 2.6 ± 1.0 0.348 <0.001 0.780 Insulin sensitivity index 7.0 ± 2.7 3.3 ± 1.2 8.2 ± 5.2 3.5 ± 1.4 7.6 ± 2.9 3.6 ± 1.7 0.899 <0.001 0.997 Antibiotics, n (%) 0 (0) 0 (0) 0 (0) 2 (25) 1 (12.5) 0 (0) 0.217 1.000 NA Periodontal or dental disease, n (%) 0 (0) 1 (12.5) 0 (0) 0 (0) 0 (0) 0 (0) 0.434 1.000 NA Urinary tract infection, n (%) 0 (0) 0 (0) 0 (0) 1 (12.5) 0 (0) 0 (0) 0.434 1.000 NA Vaginal candidiasis or infection, n (%) 0 (0) 0 (0) 0 (0) 1 (12.5) NA 0.396 1.000 NA Characteristic Control Women Women With PCOS Men Group P Value Obesity P Value Interaction P Value Nonobese (n = 8) Obese (n = 8) Nonobese (n = 7) Obese (n = 8) Nonobese (n = 8) Obese (n = 7) Age, y 27.3 ± 4.0 27.3 ± 6.5 23.0 ± 7.9 29.9 ± 5.1 23.3 ± 3.3 23.6 ± 3.2 0.116 0.129 0.138 BMI, kg/m2 23.4 ± 1.9 35.9 ± 4.2 24.4 ± 2.8 37.0 ± 4.8 22.9 ± 2.0 34.6 ± 3.4 0.309 <0.001 0.888 Waist-to-hip ratioa,b 0.74 ± 0.1 0.83 ± 0.12 0.74 ± 0.04 0.85 ± 0.1 0.84 ± 0.04 0.9 ± 0.1 0.005 <0.001 0.542 Hirsutism score 1.3 ± 1.3 1.8 ± 1.2 10.1 ± 5.1 9.3 ± 4.5 NA <0.001 0.887 0.589 Total testosterone, nmol/La,b 1.6 ± 0.3 2.0 ± 0.5 2.5 0.7 2.4 1.0 19.2 ± 3.4 16.6 ± 2.8 <0.001 0.993 0.134 FT, pmol/La,b 20 ± 5 31 ± 8 38 ± 10 45 ± 24 462 ± 106 464 ± 81 <0.001 0.012 0.022 Total estradiol, pmol/La,b 153 ± 66 276 ± 200 188 ± 230 149 ± 49 64 ± 13 95.8 ± 25.3 <0.001 0.024 0.598 FE, pmol/La 2.7 ± 1.2 5.3 ± 3 3.9 ± 4.9 3.5 ± 1.4 1.7 ± 0.5 2.7 ± 0.7 0.027 0.003 0.549 Ratio FT/FEa,b,c 8.3 ± 2.6 7.5 ± 4.6 19 ± 12.7 14 ± 5.4 286.2 ± 60.4 179.0 ± 39.3 <0.001 0.087 0.272 SHBG, nmol/La,b,c 60 ± 24 43 ± 14 44 ± 18 32 ± 13 28 ± 10 18 ± 5 <0.001 0.007 0.788 DHEAS, mmol/La 5.1 ± 0.9 5.3 ± 1.5 8.5 ± 3.9 5.7 ± 2.1 6.5 ± 1.8 8.5 ± 3.2 0.025 0.798 0.027 Androstenedione, nmol/Lb,c 9.1 ± 3.1 9.4 ± 2.8 16.1 ± 3.5 13.3 ± 5.9 7.3 ± 1.8 9.8 ± 4.2 <0.001 0.992 0.175 Total cholesterol, mmol/L 4.5 ± 0.9 4.7 ± 1.2 4.4 ± 1.2 4.4 ± 1.0 3.9 ± 0.6 4.61 ± 0.75 0.635 0.263 0.607 LDL-cholesterol, mmol/L 2.8 ± 0.7 2.8 ± 0.4 2.6 ± 1.0 2.6 ± 1. 2.4 ± 0.5 3.1 ± 0.6 0.811 0.263 0.291 HDL-cholesterol, mmol/La,b 1.3 ± 0.3 1.2 ± 0.2 1.4 ± 0.2 1.2 ± 0.2 1.2 ± 0.2 1.0 ± 0.1 0.011 0.020 0.619 Triglycerides, mmol/L 0.8 ± 0.2 0.9 ± 0.4 0.8 ± 0.2 1.2 ± 0.6 0.9 ± 0.3 1.2 ± 0.6 0.423 0.055 0.755 Fasting glucose, mmol/Lb,c 4.7 ± 0.3 5.6 ± 0.4 4.8 ± 0.4 4.9 ± 0.4 5.3 ± 0.5 5.3 ± 0.5 0.014 0.014 0.039 Fasting insulin, pmol/L 53 ± 28 92 ± 45 47 ± 35 83 ± 28 37 ± 14 76 ± 28 0.383 <0.001 0.969 HOMA-IR 1.6 ± 0.9 3.3 ± 1.6 1.5 ± 1.1 2.6 ± 0.9 1.3 ± 0.5 2.6 ± 1.0 0.348 <0.001 0.780 Insulin sensitivity index 7.0 ± 2.7 3.3 ± 1.2 8.2 ± 5.2 3.5 ± 1.4 7.6 ± 2.9 3.6 ± 1.7 0.899 <0.001 0.997 Antibiotics, n (%) 0 (0) 0 (0) 0 (0) 2 (25) 1 (12.5) 0 (0) 0.217 1.000 NA Periodontal or dental disease, n (%) 0 (0) 1 (12.5) 0 (0) 0 (0) 0 (0) 0 (0) 0.434 1.000 NA Urinary tract infection, n (%) 0 (0) 0 (0) 0 (0) 1 (12.5) 0 (0) 0 (0) 0.434 1.000 NA Vaginal candidiasis or infection, n (%) 0 (0) 0 (0) 0 (0) 1 (12.5) NA 0.396 1.000 NA Data are means ± SD or counts (percentage). For categorical variables, differences between groups were analyzed by χ2 tests. The effects of group and obesity on continuous variables were analyzed by a two-way general linear model after application of two-step transformation as needed to ensure a normal distribution of the variables. Abbreviations: FE, free estradiol; FT, free testosterone; NA, not applicable. a P < 0.05 for the differences between men and control women, independently of obesity. b P < 0.05 for the differences between men and women with PCOS, independently of obesity. c P < 0.05 for the differences between women with PCOS and control women, independently of obesity. View Large Regarding metabolic parameters, high-density lipoprotein cholesterol levels were higher in female controls and women with PCOS than in men. Fasting glucose levels were reduced in patients with PCOS compared with controls. Total cholesterol, low-density lipoprotein cholesterol, triglycerides, insulin sensitivity index, and fasting HOMA-IR scores did not differ among groups. Regardless of the study group, obesity was characterized by increased levels of free testosterone, fasting insulin, and glucose and HOMA-IR score and decreased SHBG, high-density lipoprotein cholesterol levels, and insulin sensitivity index. In addition, we observed interactions between group and obesity. Obese control women had higher levels of free testosterone and fasting glucose than did their nonobese counterparts. Obese women with PCOS had decreased levels of DHEAS compared with nonobese patients with PCOS, whereas obese men had increased levels of DHEAS compared with their nonobese counterparts. Infectious events and use of antibiotics were rare and did not differ in frequency among groups (Table 1). However, to avoid the possible effect of prior use of antibiotics on gut microbiota, the results from the three participants who had used these drugs during the previous 3 months were eliminated from further analyses of gut microbiota. Differences in bacterial diversity of gut microbiota according to study group, obesity, and circulating sex hormone concentrations Sequencing the V4 16S microbial rDNA generated a dataset consisting of 6,507,940 filtered high-quality 16S rDNA gene sequences with 1,659,832 reads for control women, 2,462,150 reads for women with PCOS, and 2,385,958 reads for men, with an average of 151,437 sequences per sample (minimum, 44,795; maximum, 281,779). We assessed gut microbiota α and β diversity by Chao 1, Shannon, Jaccard, and Sorensen estimators in the three groups. Regardless of obesity, the analysis of α diversity revealed statistically significantly larger values in men than in control women (Fig. 1). We also observed a trend toward higher Chao 1 values in men when compared with women with PCOS that was close to reaching statistical significance (P = 0.053) (Fig. 1). When women were considered as a whole, irrespective of the presence or absence of PCOS, sex also influenced gut microbiota by lowering the Chao1 index in women (Fig. 1), and a similar near-significant trend (P = 0.087) was observed in the Shannon index (Fig. 1). With all participants considered as a whole, the Chao 1 and Shannon indexes showed positive correlations with total testosterone (Chao 1: r = 0.345, P = 0.023; Shannon: r = 0.337, P = 0.027) concentrations and the ratio of free testosterone to free estradiol (Chao 1: r = 0.362, P = 0.017; Shannon: r = 0.408 P = 0.007) and showed a negative correlation with total estradiol (Chao 1: r = −0.313, P = 0.041; Shannon: r = −0.377, P = 0.013). Figure 1. View largeDownload slide Box-and-whisker plots of the interactions of study group and obesity on the bacterial α and β diversity of gut microbiota. White boxes are nonobese participants and gray boxes correspond to obese individuals. The box represents the interquartile range, which contains the median, and the whiskers indicate the 90th and 10th percentiles. *P < 0.05; †P < 0.01; ‡P < 0.001. Figure 1. View largeDownload slide Box-and-whisker plots of the interactions of study group and obesity on the bacterial α and β diversity of gut microbiota. White boxes are nonobese participants and gray boxes correspond to obese individuals. The box represents the interquartile range, which contains the median, and the whiskers indicate the 90th and 10th percentiles. *P < 0.05; †P < 0.01; ‡P < 0.001. Even though obesity did not show independent effects on bacterial diversity in our series of young adults, we observed an interaction of obesity with the effects of sex and of the study group on the indexes of bacterial β diversity. When control women and patients with PCOS were considered as a whole, the Sorensen and Jaccard indexes decreased in obese women compared with lean women, whereas these indexes increased in obese men compared with lean men (Fig. 1). The analysis considering control women and patients with PCOS as separate groups revealed that the aforementioned interaction was actually related to the findings in patients with PCOS because obese patients had decreased β diversity indexes compared with their lean counterparts, whereas such differences were not present in control women (Fig. 1). Differences in bacterial phyla of gut microbiota according to study group and obesity We then investigated bacterial abundance at the phylum level but found no statistically significant differences among the three groups of participants (data not shown). On the contrary, we found an increased relative abundance of the candidatus Saccharibacteria phylum in obese patients compared with nonobese participants regardless of group and sex (obese, 0.06; nonobese, 0.02; P = 0.018). Differences in bacterial genera of gut microbiota according to study group, obesity, and circulating sex hormone concentrations At the genus level, we observed differences in the relative abundance of three genera among the study groups. The relative abundance of Catenibacterium and Kandleria genera was higher in women with PCOS and in men compared with control women; in fact, Kandleria was absent in the latter. Moreover, the abundance of the Raoultella genus was higher in men than in women with PCOS (Fig. 2). Figure 2. View largeDownload slide Differences in bacterial relative abundance at the genus level among groups. Columns are means plus SEM. White columns are control women, gray columns are patients with PCOS, and black columns are men. Only bacterial genera showing statistically significance differences (P < 0.0125) in at least one of the comparisons are shown. Missing columns indicate that the genus was not detected in this particular category. *P < 0.0125; †P < 0.010. Figure 2. View largeDownload slide Differences in bacterial relative abundance at the genus level among groups. Columns are means plus SEM. White columns are control women, gray columns are patients with PCOS, and black columns are men. Only bacterial genera showing statistically significance differences (P < 0.0125) in at least one of the comparisons are shown. Missing columns indicate that the genus was not detected in this particular category. *P < 0.0125; †P < 0.010. In addition, we also analyzed the presence or absence of bacterial genera in the study groups and observed that men more frequently had Butyricimonas (86% vs 44%; OR, 7.0; 95% CI, 1.2 to 42.8), Megasphaera (71% vs 25%; OR, 7.5; 95% CI, 1.5 to 37.9), and Paraprevotella (86% vs 31%; OR, 13.2; 95% CI, 2.1 to 85.5) genera compared with control women and that women with PCOS more frequently had Catenibacterium (77% vs 19%; OR, 14.4; 95% CI, 2.4 to 87.4) and Oribacterium (69% vs 19%; OR, 9.7; 95% CI, 1.7 to 54.5) genera compared with control women. When all participants were considered as a whole, total and free testosterone and the free testosterone to free estradiol ratio showed positive correlations with the relative abundance of Raoultella genus (Table 2) and serum androstenedione concentrations showed positive correlations with the abundance of the Kandleria genus (Table 2). Table 2. Correlations Between the Relative Abundance of Bacterial Genera, Markers of Weight Excess and Adiposity, Circulating Sex Hormones, and Biochemical Markers of Intermediate Metabolism in All Participants as a Whole Genera Markers of Weight Excess and Adiposity BMI (kg/m2) Fat Mass (kg) Fat Mass (% of Total Body Weight) Waist-to-Hip Ratio r P Value r P Value r P Value r P Value Alloprevotella −0.542a <0.001 −0.369b 0.016 −0.312b 0.045 −0.250 0.105 Anaerofustis 0.409a 0.006 0.420a 0.006 0.344b 0.026 0.304b 0.048 Atopobium 0.378b 0.013 0.342b 0.027 0.267 0.087 0.293 0.057 Neisseria −0.375b 0.013 −0.274 0.080 −0.322b 0.037 −0.411a 0.006 Scardovia 0.449a 0.003 0.442a 0.003 0.245 0.118 0.344b 0.024 Genera Markers of Weight Excess and Adiposity BMI (kg/m2) Fat Mass (kg) Fat Mass (% of Total Body Weight) Waist-to-Hip Ratio r P Value r P Value r P Value r P Value Alloprevotella −0.542a <0.001 −0.369b 0.016 −0.312b 0.045 −0.250 0.105 Anaerofustis 0.409a 0.006 0.420a 0.006 0.344b 0.026 0.304b 0.048 Atopobium 0.378b 0.013 0.342b 0.027 0.267 0.087 0.293 0.057 Neisseria −0.375b 0.013 −0.274 0.080 −0.322b 0.037 −0.411a 0.006 Scardovia 0.449a 0.003 0.442a 0.003 0.245 0.118 0.344b 0.024 Circulating Sex Hormone Concentrations Total T (nmol/L) Free T (pmol/L) E2 (pmol/L) Free E2 (pmol/L) Free T/Free E2
(nmol/pmol) Androstenedione (nmol/L) r P Value r P Value r P Value r P Value r P Value r P Value Catenibacterium −0.169 0.278 −0.100 0.524 −0.069 0.659 −0.123 0.430 −0.037 0.813 0.247 0.110 Kandleria −0.023 0.883 0.059 0.705 0.047 0.765 0.069 0.659 0.021 0.895 0.301b 0.049 Raoultella 0.317b 0.038 0.316b 0.039 −0.219 0.159 −0.138 0.379 0.308b 0.045 −0.131 0.404 Circulating Sex Hormone Concentrations Total T (nmol/L) Free T (pmol/L) E2 (pmol/L) Free E2 (pmol/L) Free T/Free E2
(nmol/pmol) Androstenedione (nmol/L) r P Value r P Value r P Value r P Value r P Value r P Value Catenibacterium −0.169 0.278 −0.100 0.524 −0.069 0.659 −0.123 0.430 −0.037 0.813 0.247 0.110 Kandleria −0.023 0.883 0.059 0.705 0.047 0.765 0.069 0.659 0.021 0.895 0.301b 0.049 Raoultella 0.317b 0.038 0.316b 0.039 −0.219 0.159 −0.138 0.379 0.308b 0.045 −0.131 0.404 Markers of Intermediate Metabolism Glucose (mmol/L) Insulin (pmol/L) ISI Total Cholesterol (mmol/L) HDL (mmol/L) LDL (mmol/L) Triglycerides (mmol/L) r P Value r P Value r P Value r P Value r P Value r P Value r P Value Atopobium 0.256 0.097 0.368b 0.015 −0.421a 0.005 0.237 0.126 −0.142 0.364 0.251 0.105 0.263 0.088 Catenibacterium −0.406a 0.007 −0.230 0.137 0.269 0.081 −0.237 0.125 0.285 0.064 −0.325b 0.034 −0.143 0.361 Raoultella 0.352b 0.021 −0.067 0.668 0.006 0.971 −0.091 0.563 −0.215 0.167 −0.103 0.511 0.050 0.748 Scardovia 0.053 0.737 0.388b 0.010 −0.425a 0.005 0.271 0.079 0.043 0.783 0.101 0.518 0.318b 0.038 Markers of Intermediate Metabolism Glucose (mmol/L) Insulin (pmol/L) ISI Total Cholesterol (mmol/L) HDL (mmol/L) LDL (mmol/L) Triglycerides (mmol/L) r P Value r P Value r P Value r P Value r P Value r P Value r P Value Atopobium 0.256 0.097 0.368b 0.015 −0.421a 0.005 0.237 0.126 −0.142 0.364 0.251 0.105 0.263 0.088 Catenibacterium −0.406a 0.007 −0.230 0.137 0.269 0.081 −0.237 0.125 0.285 0.064 −0.325b 0.034 −0.143 0.361 Raoultella 0.352b 0.021 −0.067 0.668 0.006 0.971 −0.091 0.563 −0.215 0.167 −0.103 0.511 0.050 0.748 Scardovia 0.053 0.737 0.388b 0.010 −0.425a 0.005 0.271 0.079 0.043 0.783 0.101 0.518 0.318b 0.038 Abbreviations: E2, estradiol; HDL, high-density lipoprotein cholesterol; ISI, insulin sensitivity index; LDL, low-density lipoprotein cholesterol; T, testosterone. a P < 0.01. b P < 0.05. View Large Table 2. Correlations Between the Relative Abundance of Bacterial Genera, Markers of Weight Excess and Adiposity, Circulating Sex Hormones, and Biochemical Markers of Intermediate Metabolism in All Participants as a Whole Genera Markers of Weight Excess and Adiposity BMI (kg/m2) Fat Mass (kg) Fat Mass (% of Total Body Weight) Waist-to-Hip Ratio r P Value r P Value r P Value r P Value Alloprevotella −0.542a <0.001 −0.369b 0.016 −0.312b 0.045 −0.250 0.105 Anaerofustis 0.409a 0.006 0.420a 0.006 0.344b 0.026 0.304b 0.048 Atopobium 0.378b 0.013 0.342b 0.027 0.267 0.087 0.293 0.057 Neisseria −0.375b 0.013 −0.274 0.080 −0.322b 0.037 −0.411a 0.006 Scardovia 0.449a 0.003 0.442a 0.003 0.245 0.118 0.344b 0.024 Genera Markers of Weight Excess and Adiposity BMI (kg/m2) Fat Mass (kg) Fat Mass (% of Total Body Weight) Waist-to-Hip Ratio r P Value r P Value r P Value r P Value Alloprevotella −0.542a <0.001 −0.369b 0.016 −0.312b 0.045 −0.250 0.105 Anaerofustis 0.409a 0.006 0.420a 0.006 0.344b 0.026 0.304b 0.048 Atopobium 0.378b 0.013 0.342b 0.027 0.267 0.087 0.293 0.057 Neisseria −0.375b 0.013 −0.274 0.080 −0.322b 0.037 −0.411a 0.006 Scardovia 0.449a 0.003 0.442a 0.003 0.245 0.118 0.344b 0.024 Circulating Sex Hormone Concentrations Total T (nmol/L) Free T (pmol/L) E2 (pmol/L) Free E2 (pmol/L) Free T/Free E2
(nmol/pmol) Androstenedione (nmol/L) r P Value r P Value r P Value r P Value r P Value r P Value Catenibacterium −0.169 0.278 −0.100 0.524 −0.069 0.659 −0.123 0.430 −0.037 0.813 0.247 0.110 Kandleria −0.023 0.883 0.059 0.705 0.047 0.765 0.069 0.659 0.021 0.895 0.301b 0.049 Raoultella 0.317b 0.038 0.316b 0.039 −0.219 0.159 −0.138 0.379 0.308b 0.045 −0.131 0.404 Circulating Sex Hormone Concentrations Total T (nmol/L) Free T (pmol/L) E2 (pmol/L) Free E2 (pmol/L) Free T/Free E2
(nmol/pmol) Androstenedione (nmol/L) r P Value r P Value r P Value r P Value r P Value r P Value Catenibacterium −0.169 0.278 −0.100 0.524 −0.069 0.659 −0.123 0.430 −0.037 0.813 0.247 0.110 Kandleria −0.023 0.883 0.059 0.705 0.047 0.765 0.069 0.659 0.021 0.895 0.301b 0.049 Raoultella 0.317b 0.038 0.316b 0.039 −0.219 0.159 −0.138 0.379 0.308b 0.045 −0.131 0.404 Markers of Intermediate Metabolism Glucose (mmol/L) Insulin (pmol/L) ISI Total Cholesterol (mmol/L) HDL (mmol/L) LDL (mmol/L) Triglycerides (mmol/L) r P Value r P Value r P Value r P Value r P Value r P Value r P Value Atopobium 0.256 0.097 0.368b 0.015 −0.421a 0.005 0.237 0.126 −0.142 0.364 0.251 0.105 0.263 0.088 Catenibacterium −0.406a 0.007 −0.230 0.137 0.269 0.081 −0.237 0.125 0.285 0.064 −0.325b 0.034 −0.143 0.361 Raoultella 0.352b 0.021 −0.067 0.668 0.006 0.971 −0.091 0.563 −0.215 0.167 −0.103 0.511 0.050 0.748 Scardovia 0.053 0.737 0.388b 0.010 −0.425a 0.005 0.271 0.079 0.043 0.783 0.101 0.518 0.318b 0.038 Markers of Intermediate Metabolism Glucose (mmol/L) Insulin (pmol/L) ISI Total Cholesterol (mmol/L) HDL (mmol/L) LDL (mmol/L) Triglycerides (mmol/L) r P Value r P Value r P Value r P Value r P Value r P Value r P Value Atopobium 0.256 0.097 0.368b 0.015 −0.421a 0.005 0.237 0.126 −0.142 0.364 0.251 0.105 0.263 0.088 Catenibacterium −0.406a 0.007 −0.230 0.137 0.269 0.081 −0.237 0.125 0.285 0.064 −0.325b 0.034 −0.143 0.361 Raoultella 0.352b 0.021 −0.067 0.668 0.006 0.971 −0.091 0.563 −0.215 0.167 −0.103 0.511 0.050 0.748 Scardovia 0.053 0.737 0.388b 0.010 −0.425a 0.005 0.271 0.079 0.043 0.783 0.101 0.518 0.318b 0.038 Abbreviations: E2, estradiol; HDL, high-density lipoprotein cholesterol; ISI, insulin sensitivity index; LDL, low-density lipoprotein cholesterol; T, testosterone. a P < 0.01. b P < 0.05. View Large We then investigated which genera of the gut microbiota changed in abundance with obesity regardless of the study group being analyzed. The Anaerofustis, Atopobium, and Scardovia genera were more abundant, and the Alloprevotella and Neisseria genera were less abundant, in obese than in nonobese persons (Fig. 3). Moreover, obese compared with nonobese participants more frequently had Anaerofustis (47.6% vs 9.1%; P = 0.006), Klebsiella (61.9% vs 31.8%; P = 0.047), Mogibacterium (19% vs 0%; P = 0.048), and Scardovia (42.9% vs 9.1%; P = 0.013), whereas the presence of Alloprevotella (59.1% vs 14.3%; P = 0.003), Gordonibacter (86.4% vs 52.4%; P = 0.017), Neisseria (81.8% vs 42.9%; P = 0.009), Propionibacterium (50% vs 19%; P = 0.034), and Pseudomonas (54.5% vs 19%; P = 0.017) was more frequent in nonobese than in obese participants. Figure 3. View largeDownload slide Differences in bacterial relative abundance at the genus level among nonobese and obese participants. Columns are means plus SEM. White columns are nonobese participants and gray columns are obese participants. Only bacterial genera showing statistically significant differences (P < 0.0125) in the comparisons are shown. *P < 0.0125; †P < 0.010. Figure 3. View largeDownload slide Differences in bacterial relative abundance at the genus level among nonobese and obese participants. Columns are means plus SEM. White columns are nonobese participants and gray columns are obese participants. Only bacterial genera showing statistically significant differences (P < 0.0125) in the comparisons are shown. *P < 0.0125; †P < 0.010. The interactions between study group and the effects of obesity revealed statistically significant differences in the abundance of two bacterial genera, Atopobium and Neisseria. Obesity exerted a different effect in control women and in women PCOS and men; it was associated with an increase in the abundance of Atopobium in control women but not in women with PCOS and men (Fig. 4). Moreover, obesity was associated with a decrease in the abundance of Neisseria in control women, an association that was not found in women with PCOS or in men (Fig. 4). Figure 4. View largeDownload slide Interactions between study group and obesity in bacterial relative abundance at the genus level. Columns are means plus SEM. White columns are nonobese participants and gray columns are obese participants. Only bacterial genera showing statistically significant differences (P < 0.0125) in the comparisons are shown. †P < 0.010. Figure 4. View largeDownload slide Interactions between study group and obesity in bacterial relative abundance at the genus level. Columns are means plus SEM. White columns are nonobese participants and gray columns are obese participants. Only bacterial genera showing statistically significant differences (P < 0.0125) in the comparisons are shown. †P < 0.010. Given the impact of obesity on the findings described above, we analyzed the correlations between the relative abundance of bacterial genera with several metabolic parameters. When all participants were considered as a whole, irrespective of group, sex, and obesity, BMI showed positive correlations with the abundance of Anaerofustis, Atopobium, and Scardovia and negative correlations with that of Alloprevotella and Neisseria (Table 2). Fat mass, expressed as total kilograms or percentage of total body weight, showed positive correlations with the abundances of Anaerofustis, Atopobium, and Scardovia and negative correlations with those of Neisseria and Alloprevotella (Table 2). Fasting glucose levels showed positive correlations with abundance of Raoultella and negative correlations with abundance of Catenibacterium, whereas insulin showed positive correlations, and the insulin sensitivity index showed negative correlations, with the abundance of Atopobium and Scardovia (Table 2). Discussion Our study showed that the diversity and composition of the gut microbiota were influenced by sex and obesity and that sex hormones may be related to these differences. Indexes of bacterial α diversity in the gut microbiota were reduced in women with and without androgen excess when compared with men. Such an effect of sex was not found in a recent study; that study showed no differences in α bacterial diversity between men and women (13). However, the fact that the women included in that study were postmenopausal (13) may have obscured the sexual dimorphism in the diversity of gut microbiota and possible influences of sex hormones suggested in the analysis of our young adults. Moreover, because α diversity showed positive correlations with testosterone levels and ratio of free testosterone to free estradiol and negative correlations with estradiol concentrations, our study further suggests that androgens and estrogens might be involved in the sexual dimorphism reported here. However, being derived from a cross-sectional study, our data cannot clarify the causality of these associations, and the possibility exists that gut microbes participate in the regulation sex hormones and, conversely, that sex hormones modify microbial diversity (12). Contrary to the results in α diversity, the changes in β bacterial diversity of gut microbiota found in our young adults was heavily influenced by obesity, exemplifying the complex interactions among sex, sex hormones, and obesity on gut bacterial diversity. This effect was particularly important in the subset of women with PCOS. Even though the current literature on the effect of obesity on bacterial diversity yielded conflicting results (22–25), our study addressed the role of sex and sex hormones on such an effect; this issue is of paramount importance because gonadal dysfunction, consisting of androgen excess in women and androgen deficiency in men, is prevalent in obese persons (26, 27). Moreover, the negative effect of obesity on the bacterial diversity of gut microbiota might be explained, at least in part, by the ecological theory, which predicts that species-rich communities, with different species specializing to each potentially limiting resource, are less susceptible to invasion because they use limiting resources more efficiently (22). Excess nutrient loading may result in decreased ecosystem diversity as a small number of species overgrow and outcompete everything else (22). In relative agreement with our finding of decreased β diversity in obese patients with PCOS, Lindheim et al. (28) and Torres et al. (29) reported reductions in indexes of α diversity in women with PCOS compared with control women. The fact that we did not find such a reduction in indexes of α diversity in our patients might be related to differences in the criteria used to diagnose PCOS—all patients had hyperandrogenism in our series compared with only 60% in previous studies (28, 29)—and in the characteristics of the populations analyzed because the aforementioned studies did not include obese participants (28, 29). At the phyla level, our study showed that the abundance of candidatus Saccharibacteria was higher in obese patients than in nonobese patients. This phylum plays a role in the degradation of various organic compounds as well as glucose, butyrate, oleic acid, and amino acids (30). However, studies in mice reported a reduced abundance of this phylum in response to a high-fat diet (31, 32). Such an effect was unlikely to influence our present results because all the participants maintained similar diets in the days preceding sampling. Sex also influenced the composition of gut microbiota at the genera level, with women showing reduced abundance of Raoultella genus when compared with men. Moreover, the Kandleria genus was not present in control women, and the Megasphaera, Paraprevotella, and Butyricimonas genera were less frequent in control women than in men. Most of these genera have not been related to sex before in human studies, with the notable exception of Butyricimonas, which was also decreased in women in earlier studies (13). The influence of sex on the composition of gut microbiota is in conceptual agreement with animal data that also demonstrate the remarkable interaction of these effects with host genetics and environmental factors, such as diet (12, 33). Accordingly, in our study Paraprevotella showed the strongest positive correlation with testosterone and negative associations with estradiol concentrations among the genera that correlated with sex hormones and sex. In addition, androgenized rats are characterized by an increase in the abundance of the Paraprevotella genus (34). But similar to what has been suggested for bacterial diversity, evidence suggests that gut microbiota may catalyze modifications of steroid hormones (35) and might play some role in the regulation of sex hormone concentrations in the body (36). The analysis of bacterial composition at the genus level revealed an abnormality characteristic of women with PCOS. These patients presented increased abundance of the Catenibacterium and Kandleria genera compared with both female and male controls. Moreover, the abundance of Kandleria correlated positively with circulating androstenedione concentrations when all participants were considered as a whole, suggesting an association with androgen levels. Catenibacterium are gram-positive anaerobic bacteria that use glucose to produce acetic, lactic, butyric, and isobutyric acids (37). Several studies reported an increase in the abundance of Catenibacterium in infectious and autoimmune disorders (38, 39), suggesting a link with the proinflammatory response. Moreover, the abundance of this genus in human microbiota increased with western-style and carbohydrate-rich diets (40, 41), whereas in ob/ob mice, feeding with prebiotics decreased the abundance of Catenibacterium in parallel with amelioration of low-grade inflammation and fat-mass accumulation (42). Kandleria are gram-positive anaerobic bacteria recently reclassified on the basis of phylogenetic information from the genus Lactobacillus and placement in family Erysipelotrichaceae (43). Kandleria metabolizes glucose producing lactic acid (43) and have not been related to metabolism or sexual hormones before in human studies. To the best of our knowledge, no data have been published linking differences in the Raoultella genus with PCOS and/or obesity. Obesity increased the abundance of Anaerofustis, Atopobium, and Scardovia and decreased the abundance of Alloprevotella and Neisseria. Of these, Atopobium increased specifically in obese control women compared with their nonobese counterparts. Even though the abundance of Atopobium has been reported to be decreased in fecal samples of patients with diabetes who presented with higher BMI than the controls (44) and in obese children (45), the abundance of this genus increased in male mice with high-fat diet–induced obesity, in agreement with our present results (46). A similar model of obesity in the rat resulted in a decrease in the abundance of Alloprevotella (47) similar to the increase found in our obese participants. Only a few human and animal studies studied gut microbiota genera in PCOS (28, 29, 48–50). In mice, letrozole-induced PCOS resulted into a reduction in overall species and phylogenetic richness, where the abundance of the Alistipes genus was decreased and that of the Allobaculum and Blautia genera were increased (49). A similar model in rats resulted into reduced Lactobacillus, Ruminococcus, and Clostridium abundances and increased Prevotella levels (50). Moreover, correction of this dysbiosis improved the menstrual pattern in these animals (50). In humans, available data suggest an association between the abundance of certain gut microbiota genera and PCOS. In Europeans, PCOS is associated with a reduced relative abundance of the phylum Tenericutes, two unclassified genera from the order ML615J-28 (belonging to the phylum Tenericutes), and the family S24-7 (belonging to the phylum Bacteroidetes) (28). However, as many as 40% of the women diagnosed with PCOS lacked androgen excess (28) in a phenotype that usually associates much milder metabolic disorders compared with the classic PCOS phenotype that motivated the diagnosis in our series, explaining the different outcomes of both studies. In addition, patients with PCOS were younger than controls in this study (28) and therefore, such results may have been influenced by age. In the Chinese population nonobese and obese control women and women with PCOS showed differences in relative abundance in 28 operational taxonomic units (48). However, most differences in abundance in these studies (28, 48) pertained to unclassified bacterial genera, making difficult a direct comparison with our present results. Among the advantages of our present study, we highlight the nontargeted approach to the analysis of the gut microbiome, which may have permitted the finding of associations of gut microbiota with sex, sex hormones, and PCOS with genera that might have been missed by targeted approaches. But we also acknowledge that our experiments are not free from limitations, especially with regard to the cross-sectional design used, the limited sample size analyzed, and the high interindividual variability observed. We tried to compensate for these limitations by comparing only rigorously defined and homogeneous groups of participants of similar age and BMI. Similarly, host specificities or diet-related differences may explain variations observed in the composition of gut microbiota, yet we tried to minimize these confounding factors by excluding from analysis the fecal samples from the three participants who had used antibiotics and by standardizing the diet of all these persons for 3 days before sampling. However, we cannot definitely exclude a putative impact of differences in the diet occurring before these 3 days on the microbiota of these individuals. The data presented were obtained for participants living in a specific geographical region, which does not necessarily reflect the environment in other locations. Similarly, we included only patients with hyperandrogenic phenotypes of PCOS and, accordingly, our results should not be extrapolated to nonhyperandrogenic patients with PCOS. Finally, a more detailed structural and functional analysis of intestinal microbiota is required to better define lifestyle- or diet-related effects. Discrimination at the species or even strain level might be of interest for some bacterial groups in order to detect subtle variations in dominant groups. In summary, our findings indicate that the diversity and composition of gut microbiota are influenced by the combined effects of sex, sex hormone concentrations, and obesity. Intervention studies designed to determine the causality of these associations may permit targeting gut microbiota as putative diagnostic and therapeutic targets for PCOS, obesity, and associated metabolic comorbidities. Abbreviations: Abbreviations: BMI body mass index DHEAS dehydroepiandrosterone-sulfate HOMA-IR homeostasis model assessment of insulin resistance PCOS polycystic ovary syndrome rDNA ribosomal DNA Acknowledgments Financial Support: Supported by grants PI11/00357, PI15/01686 SEXMETAB, and PIE16/00050 SEXCOMPLEX from Instituto de Salud Carlos III, Spanish Ministry of Economy and Competitiveness (H.F.E.M.). CIBERDEM and IRYCIS are also initiatives of Instituto de Salud Carlos III. Supported in part by Fondo Europeo de Desarrollo Regional (FEDER, European Union). Disclosure Summary: The authors have nothing to disclose. References 1. Escobar-Morreale HF , Luque-Ramírez M , San Millán JL . The molecular-genetic basis of functional hyperandrogenism and the polycystic ovary syndrome . Endocr Rev . 2005 ; 26 ( 2 ): 251 – 282 . 2. Escobar-Morreale HF , San Millán JL . Abdominal adiposity and the polycystic ovary syndrome . Trends Endocrinol Metab . 2007 ; 18 ( 7 ): 266 – 272 . 3. Escobar-Morreale HF , Samino S , Insenser M , Vinaixa M , Luque-Ramírez M , Lasunción MA , Correig X . Metabolic heterogeneity in polycystic ovary syndrome is determined by obesity: plasma metabolomic approach using GC-MS . Clin Chem . 2012 ; 58 ( 6 ): 999 – 1009 . 4. Luque-Ramírez M , Escobar-Morreale HF . Polycystic ovary syndrome as a paradigm for prehypertension, prediabetes, and preobesity . Curr Hypertens Rep . 2014 ; 16 ( 12 ): 500 . 5. Wild RA , Carmina E , Diamanti-Kandarakis E , Dokras A , Escobar-Morreale HF , Futterweit W , Lobo R , Norman RJ , Talbott E , Dumesic DA . Assessment of cardiovascular risk and prevention of cardiovascular disease in women with the polycystic ovary syndrome: a consensus statement by the Androgen Excess and Polycystic Ovary Syndrome (AE-PCOS) Society . J Clin Endocrinol Metab . 2010 ; 95 ( 5 ): 2038 – 2049 . 6. Clemente JC , Ursell LK , Parfrey LW , Knight R . The impact of the gut microbiota on human health: an integrative view . Cell . 2012 ; 148 ( 6 ): 1258 – 1270 . 7. Bäckhed F , Ding H , Wang T , Hooper LV , Koh GY , Nagy A , Semenkovich CF , Gordon JI . The gut microbiota as an environmental factor that regulates fat storage . Proc Natl Acad Sci USA . 2004 ; 101 ( 44 ): 15718 – 15723 . 8. Turnbaugh PJ , Ley RE , Mahowald MA , Magrini V , Mardis ER , Gordon JI . An obesity-associated gut microbiome with increased capacity for energy harvest . Nature . 2006 ; 444 ( 7122 ): 1027 – 1031 . 9. Sonnenburg JL , Bäckhed F . Diet-microbiota interactions as moderators of human metabolism . Nature . 2016 ; 535 ( 7610 ): 56 – 64 . 10. Kovacs A , Ben-Jacob N , Tayem H , Halperin E , Iraqi FA , Gophna U . Genotype is a stronger determinant than sex of the mouse gut microbiota . Microb Ecol . 2011 ; 61 ( 2 ): 423 – 428 . 11. Lay C , Rigottier-Gois L , Holmstrøm K , Rajilic M , Vaughan EE , de Vos WM , Collins MD , Thiel R , Namsolleck P , Blaut M , Doré J . Colonic microbiota signatures across five northern European countries . Appl Environ Microbiol . 2005 ; 71 ( 7 ): 4153 – 4155 . 12. Yurkovetskiy L , Burrows M , Khan AA , Graham L , Volchkov P , Becker L , Antonopoulos D , Umesaki Y , Chervonsky AV . Gender bias in autoimmunity is influenced by microbiota . Immunity . 2013 ; 39 ( 2 ): 400 – 412 . 13. Haro C , Rangel-Zúñiga OA , Alcalá-Díaz JF , Gómez-Delgado F , Pérez-Martínez P , Delgado-Lista J , Quintana-Navarro GM , Landa BB , Navas-Cortés JA , Tena-Sempere M , Clemente JC , López-Miranda J , Pérez-Jiménez F , Camargo A . Intestinal microbiota is influenced by gender and body mass index . PLoS One . 2016 ; 11 ( 5 ): e0154090 . 14. Zawadzki JK , Dunaif A . Diagnostic criteria for polycystic ovary syndrome: Towards a rational approach. In: Dunaif A , Givens JR , Haseltine FP , and Merriam GR , eds. Polycystic ovary syndrome . Boston : Blackwell Scientific Publications ; 1992 : 377 – 384 . 15. Azziz R , Carmina E , Dewailly D , Diamanti-Kandarakis E , Escobar-Morreale HF , Futterweit W , Janssen OE , Legro RS , Norman RJ , Taylor AE , Witchel SF ; Task Force on the Phenotype of the Polycystic Ovary Syndrome of The Androgen Excess and PCOS Society . The Androgen Excess and PCOS Society criteria for the polycystic ovary syndrome: the complete task force report . Fertil Steril . 2009 ; 91 ( 2 ): 456 – 488 . 16. Escobar-Morreale HF , Sanchón R , San Millán JL . A prospective study of the prevalence of nonclassical congenital adrenal hyperplasia among women presenting with hyperandrogenic symptoms and signs . J Clin Endocrinol Metab . 2008 ; 93 ( 2 ): 527 – 533 . 17. Matthews DR , Hosker JP , Rudenski AS , Naylor BA , Treacher DF , Turner RC . Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man . Diabetologia . 1985 ; 28 ( 7 ): 412 – 419 . 18. Matsuda M , DeFronzo RA . Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp . Diabetes Care . 1999 ; 22 ( 9 ): 1462 – 1470 . 19. Templeton GF . A two-step approach for transforming continuous variables to normal: implications and recommendations for IS research . Comm Assoc Info Syst . 2011 : 28 :41–51 . 20. White JR , Nagarajan N , Pop M . Statistical methods for detecting differentially abundant features in clinical metagenomic samples . PLOS Comput Biol . 2009 ; 5 ( 4 ): e1000352 . 21. Jonsson V , Österlund T , Nerman O , Kristiansson E . Statistical evaluation of methods for identification of differentially abundant genes in comparative metagenomics . BMC Genomics . 2016 ; 17 ( 1 ): 78 . 22. Lozupone CA , Stombaugh JI , Gordon JI , Jansson JK , Knight R . Diversity, stability and resilience of the human gut microbiota . Nature . 2012 ; 489 ( 7415 ): 220 – 230 . 23. Le Chatelier E , Nielsen T , Qin J , Prifti E , Hildebrand F , Falony G , Almeida M , Arumugam M , Batto JM , Kennedy S , Leonard P , Li J , Burgdorf K , Grarup N , Jørgensen T , Brandslund I , Nielsen HB , Juncker AS , Bertalan M , Levenez F , Pons N , Rasmussen S , Sunagawa S , Tap J , Tims S , Zoetendal EG , Brunak S , Clément K , Doré J , Kleerebezem M , Kristiansen K , Renault P , Sicheritz-Ponten T , de Vos WM , Zucker JD , Raes J , Hansen T , Bork P , Wang J , Ehrlich SD , Pedersen O ; MetaHIT consortium . Richness of human gut microbiome correlates with metabolic markers . Nature . 2013 ; 500 ( 7464 ): 541 – 546 . 24. Walters WA , Xu Z , Knight R . Meta-analyses of human gut microbes associated with obesity and IBD . FEBS Lett . 2014 ; 588 ( 22 ): 4223 – 4233 . 25. Kasai C , Sugimoto K , Moritani I , Tanaka J , Oya Y , Inoue H , Tameda M , Shiraki K , Ito M , Takei Y , Takase K . Comparison of the gut microbiota composition between obese and non-obese individuals in a Japanese population, as analyzed by terminal restriction fragment length polymorphism and next-generation sequencing . BMC Gastroenterol . 2015 ; 15 ( 1 ): 100 . 26. Escobar-Morreale HF , Alvarez-Blasco F , Botella-Carretero JI , Luque-Ramírez M . The striking similarities in the metabolic associations of female androgen excess and male androgen deficiency . Hum Reprod . 2014 ; 29 ( 10 ): 2083 – 2091 . 27. Escobar-Morreale HF , Santacruz E , Luque-Ramírez M , Botella Carretero JI . Prevalence of ‘obesity-associated gonadal dysfunction’ in severely obese men and women and its resolution after bariatric surgery: a systematic review and meta-analysis . Hum Reprod Update . 2017 ; 23 ( 4 ): 390 – 408 . 28. Lindheim L , Bashir M , Münzker J , Trummer C , Zachhuber V , Leber B , Horvath A , Pieber TR , Gorkiewicz G , Stadlbauer V , Obermayer-Pietsch B . Alterations in gut microbiome composition and barrier function are associated with reproductive and metabolic defects in women with polycystic ovary syndrome (PCOS): A pilot study . PLoS One . 2017 ; 12 ( 1 ): e0168390 . 29. Torres PJ , Siakowska M , Banaszewska B , Pawelczyk L , Duleba AJ , Kelley ST , Thackray VG . Gut microbial diversity in women with polycystic ovary syndrome correlates with hyperandrogenism . J Clin Endocrinol Metab . 2018 ; 103 ( 4 ): 1502 – 1511 . 30. Kindaichi T , Yamaoka S , Uehara R , Ozaki N , Ohashi A , Albertsen M , Nielsen PH , Nielsen JL . Phylogenetic diversity and ecophysiology of Candidate phylum Saccharibacteria in activated sludge . FEMS Microbiol Ecol . 2016 ; 92 ( 6 ): fiw078 . 31. Lu Y , Fan C , Li P , Lu Y , Chang X , Qi K . Short chain fatty acids prevent high-fat-diet-induced obesity in mice by regulating G protein-coupled receptors and gut microbiota . Sci Rep . 2016 ; 6 ( 1 ): 37589 . 32. Moya-Pérez A , Neef A , Sanz Y . Bifidobacterium pseudocatenulatum CECT 7765 reduces obesity-associated inflammation by restoring the lymphocyte-macrophage balance and gut microbiota structure in high-fat diet-fed mice . PLoS One . 2015 ; 10 ( 7 ): e0126976 . 33. Org E , Mehrabian M , Parks BW , Shipkova P , Liu X , Drake TA , Lusis AJ . Sex differences and hormonal effects on gut microbiota composition in mice . Gut Microbes . 2016 ; 7 ( 4 ): 313 – 322 . 34. Moreno-Indias I , Cardona F , Tinahones FJ , Queipo-Ortuño MI . Impact of the gut microbiota on the development of obesity and type 2 diabetes mellitus . Front Microbiol . 2014 ; 5 : 190 . 35. Lombardi P , Goldin B , Boutin E , Gorbach SL . Metabolism of androgens and estrogens by human fecal microorganisms . J Steroid Biochem . 1978 ; 9 ( 8 ): 795 – 801 . 36. Donova MV , Egorova OV . Microbial steroid transformations: current state and prospects . Appl Microbiol Biotechnol . 2012 ; 94 ( 6 ): 1423 – 1447 . 37. Kageyama A , Benno Y . Catenibacterium mitsuokai gen. nov., sp. nov., a gram-positive anaerobic bacterium isolated from human faeces . Int J Syst Evol Microbiol . 2000 ; 50 ( Pt 4 ): 1595 – 1599 . 38. Wagner BD , Frank DN , Zipris D . Response to comment on Alkanani et al. Alterations in intestinal microbiota correlate with susceptibility to type 1 diabetes. Diabetes 2015;64:3510-3520 . Diabetes . 2015 ; 64 ( 10 ): e41 . 39. Lozupone CA , Li M , Campbell TB , Flores SC , Linderman D , Gebert MJ , Knight R , Fontenot AP , Palmer BE . Alterations in the gut microbiota associated with HIV-1 infection . Cell Host Microbe . 2013 ; 14 ( 3 ): 329 – 339 . 40. Turnbaugh PJ , Ridaura VK , Faith JJ , Rey FE , Knight R , Gordon JI . The effect of diet on the human gut microbiome: a metagenomic analysis in humanized gnotobiotic mice . Sci Transl Med . 2009 ; 1 ( 6 ): 6ra14 . 41. Wu GD , Chen J , Hoffmann C , Bittinger K , Chen YY , Keilbaugh SA , Bewtra M , Knights D , Walters WA , Knight R , Sinha R , Gilroy E , Gupta K , Baldassano R , Nessel L , Li H , Bushman FD , Lewis JD . Linking long-term dietary patterns with gut microbial enterotypes . Science . 2011 ; 334 ( 6052 ): 105 – 108 . 42. Everard A , Lazarevic V , Derrien M , Girard M , Muccioli GG , Neyrinck AM , Possemiers S , Van Holle A , François P , de Vos WM , Delzenne NM , Schrenzel J , Cani PD . Responses of gut microbiota and glucose and lipid metabolism to prebiotics in genetic obese and diet-induced leptin-resistant mice [published correction appears in Diabetes. 2011;60(12):3307] . Diabetes . 2011 ; 60 ( 11 ): 2775 – 2786 . 43. Salvetti E , Felis GE , Dellaglio F , Castioni A , Torriani S , Lawson PA . Reclassification of Lactobacillus catenaformis (Eggerth 1935) Moore and Holdeman 1970 and Lactobacillus vitulinus Sharpe et al. 1973 as Eggerthia catenaformis gen. nov., comb. nov. and Kandleria vitulina gen. nov., comb. nov., respectively . Int J Syst Evol Microbiol . 2011 ; 61 ( Pt 10 ): 2520 – 2524 . 44. Sato J , Kanazawa A , Ikeda F , Yoshihara T , Goto H , Abe H , Komiya K , Kawaguchi M , Shimizu T , Ogihara T , Tamura Y , Sakurai Y , Yamamoto R , Mita T , Fujitani Y , Fukuda H , Nomoto K , Takahashi T , Asahara T , Hirose T , Nagata S , Yamashiro Y , Watada H . Gut dysbiosis and detection of “live gut bacteria” in blood of Japanese patients with type 2 diabetes . Diabetes Care . 2014 ; 37 ( 8 ): 2343 – 2350 . 45. Nagata S , Chiba Y , Wang C , Yamashiro Y . The effects of the Lactobacillus casei strain on obesity in children: a pilot study . Benef Microbes . 2017 ; 8 ( 4 ): 535 – 543 . 46. Lee JC , Lee HY , Kim TK , Kim MS , Park YM , Kim J , Park K , Kweon MN , Kim SH , Bae JW , Hur KY , Lee MS . Obesogenic diet-induced gut barrier dysfunction and pathobiont expansion aggravate experimental colitis . PLoS One . 2017 ; 12 ( 11 ): e0187515 . 47. Lin H , An Y , Hao F , Wang Y , Tang H . Correlations of fecal metabonomic and microbiomic changes induced by high-fat diet in the pre-obesity state . Sci Rep . 2016 ; 6 ( 1 ): 21618 . 48. Liu R , Zhang C , Shi Y , Zhang F , Li L , Wang X , Ling Y , Fu H , Dong W , Shen J , Reeves A , Greenberg AS , Zhao L , Peng Y , Ding X . Dysbiosis of gut microbiota associated with clinical parameters in polycystic ovary syndrome . Front Microbiol . 2017 ; 8 : 324 . 49. Kelley ST , Skarra DV , Rivera AJ , Thackray VG . The gut microbiome is altered in a letrozole-induced mouse model of polycystic ovary syndrome . PLoS One . 2016 ; 11 ( 1 ): e0146509 . 50. Guo Y , Qi Y , Yang X , Zhao L , Wen S , Liu Y , Tang L . Association between polycystic ovary syndrome and gut microbiota . PLoS One . 2016 ; 11 ( 4 ): e0153196 . Copyright © 2018 Endocrine Society http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Clinical Endocrinology and Metabolism Oxford University Press

Gut Microbiota and the Polycystic Ovary Syndrome: Influence of Sex, Sex Hormones, and Obesity

Loading next page...
 
/lp/ou_press/gut-microbiota-and-the-polycystic-ovary-syndrome-pcos-influence-of-sex-0DUsT0pdwt
Publisher
Endocrine Society
Copyright
Copyright © 2018 Endocrine Society
ISSN
0021-972X
eISSN
1945-7197
D.O.I.
10.1210/jc.2017-02799
Publisher site
See Article on Publisher Site

Abstract

Abstract Context Gut microbiota play a major role in health and disease by influencing physiology, metabolism, nutrition, and immune function. Objective To evaluate the composition of gut microbiota in women with polycystic ovary syndrome (PCOS), focusing on the influence of sex, sex hormones and obesity on the associations found. Design Cross-sectional study. Setting Academic hospital. Participants We recruited 15 women with PCOS, 16 nonhyperandrogenic control women, and 15 control men. Participants were classified as nonobese (<30 kg/m2) or obese (≥30 kg/m2) according to their body mass index. Interventions Standardization of diet for 3 consecutive days (at least 300 g of carbohydrates per day) followed by fecal sampling and a standard 75-g oral glucose tolerance test. Main Outcome Measures Analysis of bacterial abundance and composition of gut microbiota by massive sequencing of 16S ribosomal DNA amplicons in a MiSeq platform (Illumina). Results α Bacterial diversity was reduced in women compared with men, and β diversity was reduced particularly in obese patients with PCOS. Women with PCOS presented with specific abnormalities in gut microbiota consisting of an increased abundance of the Catenibacterium and Kandleria genera. When all participants as a whole were considered, indexes of bacterial diversity and the abundance of several bacterial genera correlated positively with serum androgen concentrations and negatively with estradiol levels. Conclusions The diversity and composition of the gut microbiota of young adults are influenced by the combined effects of sex, sex hormone concentrations, and obesity, presenting with specific abnormalities in women with PCOS. Polycystic ovary syndrome (PCOS) is now considered a complex multigenic disorder with strong environmental influences (1). Several metabolic disorders are associated with PCOS, including insulin resistance, diabetes, and obesity (2). The latter exerts a major effect on the metabolic associations and complications of the syndrome (3). Among other factors, obesity is clearly related to the infertility of PCOS and increases the risk for metabolic syndrome and the clustering of cardiovascular risk factors in these women (4, 5). During the last two decades, several studies revealed the major influence of gut microbiota on physiology, metabolism, nutrition, and immune function (6). Germ-free mice—which therefore lack gut microbiota—have reduced adiposity and improved tolerance to glucose and insulin when compared with colonized counterparts (7). Moreover, fecal transplantation of the obese gut microbiome into germ-free mice promotes increase in adiposity, resulting in obese phenotype (8). Moreover, several studies have reported the gut microbiota particularities related to not just obesity but also diabetes and the metabolic syndrome (9). The possible influence of host sex and/or sex steroids on the composition of gut microbiota remains unclear at present (10–13). The current study aims to evaluate the composition of gut microbiota in women with PCOS, focusing on the possible influence of sex, sex hormones, obesity, and their interactions, on any putative association. Materials and Methods Participants A total of 46 young adult volunteers (15 patients with PCOS, 16 nonhyperandrogenic women, and 15 healthy men) were recruited. Participants had similar mean body mass index (BMI) and age. They were classified as nonobese (<30 kg/m2) or obese (≥30 kg/m2) according to their BMI. The diagnosis of PCOS was based on the presence of clinical and/or biochemical hyperandrogenism, oligo-ovulation, and exclusion of secondary causes (14, 15). Before inclusion, the participants had no history of obesity-associated comorbidities, such as impaired glucose tolerance, hypertension, cardiovascular events, or sleep apnea. Smokers were excluded. None of the individuals had received treatment with oral contraceptives, antiandrogens, lipid-lowering or blood pressure–lowering drugs, or insulin sensitizers within the previous 3 months. None of the women in the control group had signs or symptoms of hyperandrogenism, history of menstrual dysfunction, or infertility. Men with androgen deficiency, as defined by a serum total testosterone concentration <8 nmol/L, were excluded. The Ethics Committee of Hospital Universitario Ramón y Cajal approved the study protocol, and all participants gave written informed consent. Study design All individuals underwent a comprehensive clinical, anthropometric, and physical evaluation. Participants were advised to follow a diet with 300 g of carbohydrate per day for 3 days before sampling to avoid false-positive results on the 75-g oral glucose load performed thereafter; the latter test was used not only for research purposes but also to check patients for disorders of glucose tolerance. In women, the study was performed during days 3 to 10 of the follicular phase of the menstrual cycle or in amenorrhea after exclusion of pregnancy. Occurrence of infectious disease, including vaginal candidiasis and periodontal disease, during the previous 3 months and recent use of antibiotics were recorded. Laboratory measurements Baseline blood samples were collected for hormonal and metabolic measurements. Plasma and serum aliquots were frozen at −80°C until use. Technical characteristics of the assays used for laboratory measurements have been described previously (16). The homeostasis model assessment of insulin resistance (HOMA-IR) was calculated from fasting glucose and insulin concentrations (17), and the composite insulin sensitivity index was estimated from the glucose and insulin concentrations measured during the oral glucose tolerance test (18). DNA extraction from fecal sample and next-generation sequencing Patients were instructed to collect fecal samples into sterile plastic bottle topped with a screw cap. The samples were immediately divided into aliquots and stored at −80°C after collection until analysis was performed. Samples were defrosted slowly at 4°C to avoid DNA degradation and further thawed at room temperature. An aliquot of 0.5 g of feces was solubilized in 5 mL of distilled water and centrifuged at 157g and 4°C for 10 minutes to eliminate undigested fiber. Total DNA from 1 mL of the supernatant was obtained with the QIAamp DNA Mini KitPellet commercial kit (Qiagen, Hilden, Germany). After DNA samples were checked for quality and concentration, they were sent to Fundacio per al Foment de la Investigacio i Biomedica (FISABIO, Valencia, Spain) for massive sequencing of 16S ribosomal DNA (rDNA) amplicons in a MiSeq platform (Illumina). Bioinformatics analysis of the readings was conducted at FISABIO, and only sequences longer than 200 nucleotides were considered. The bacterial diversity was estimated by using Shannon, Chao 1, Jaccard, and Sorensen indexes with their SDs. Taxonomic affiliations were assigned by using the RDP_classifier from the Ribosomal Database Project (RDP), and readings with RDP score value <0.8 were assigned to the upper taxonomic rank, leaving the last rank as unidentified. Statistical analysis Data are expressed as mean ± SD unless otherwise stated. We tested the normality of distribution by using the Kolmogorov-Smirnov test. Two-step transformations were applied as needed to ensure normal distribution of data (19). General linear models were used to determine the influence of group and obesity on hormonal, metabolic, and inflammatory variables and indexes of α and β diversity. The differences between the groups of participants in bacterial composition were evaluated by the Kruskal-Wallis test followed by Dunn nonparametric post hoc test (20, 21). The comparisons between nonobese and obese subgroups on bacterial composition and abundance were evaluated by Mann-Whitney U test. The interaction between groups and obesity were evaluated, in each subgroup of participants separately, by a Mann-Whitney U test. We applied a Bonferroni correction to the level of statistical significance to take into account multiple comparisons (P < 0.0125 was considered to indicate a statistically significant difference). Categorical variables were analyzed by a χ2 or Fisher exact test as appropriate. Relationships between described genera and clinical and biochemical variables were analyzed by Spearman correlation analysis. We used SPSS statistical software, version 18.0 (IBM, Chicago, IL) and PAST for statistical analyses of biodiversity. A P value <0.05 was considered to indicate statistical significance with the exception mentioned previously. Results Clinical, hormonal, and metabolic variables in patients with PCOS and controls The clinical, hormonal, and metabolic data are summarized in Table 1. Age and BMI did not differ between groups. The hirsutism score was higher in patients with PCOS than in control women (hirsutism score was not assessed in men). Besides their higher total testosterone, free testosterone and dehydroepiandrosterone-sulfate (DHEAS) levels, men also had higher waist-to-hip ratio and ratio of free testosterone to free estradiol than did control women, who, in contrast, had higher SHBG and free estradiol than did men. Women with PCOS had increased concentrations of free testosterone, ratio of free testosterone to free estradiol, and androstenedione compared with control women. In addition, women with PCOS presented lower levels of total and free testosterone and ratio of free testosterone to free estradiol, but higher SHBG and androstenedione levels, than did men. Table 1. Clinical, Hormonal, and Metabolic Characteristics of Participants Characteristic Control Women Women With PCOS Men Group P Value Obesity P Value Interaction P Value Nonobese (n = 8) Obese (n = 8) Nonobese (n = 7) Obese (n = 8) Nonobese (n = 8) Obese (n = 7) Age, y 27.3 ± 4.0 27.3 ± 6.5 23.0 ± 7.9 29.9 ± 5.1 23.3 ± 3.3 23.6 ± 3.2 0.116 0.129 0.138 BMI, kg/m2 23.4 ± 1.9 35.9 ± 4.2 24.4 ± 2.8 37.0 ± 4.8 22.9 ± 2.0 34.6 ± 3.4 0.309 <0.001 0.888 Waist-to-hip ratioa,b 0.74 ± 0.1 0.83 ± 0.12 0.74 ± 0.04 0.85 ± 0.1 0.84 ± 0.04 0.9 ± 0.1 0.005 <0.001 0.542 Hirsutism score 1.3 ± 1.3 1.8 ± 1.2 10.1 ± 5.1 9.3 ± 4.5 NA <0.001 0.887 0.589 Total testosterone, nmol/La,b 1.6 ± 0.3 2.0 ± 0.5 2.5 0.7 2.4 1.0 19.2 ± 3.4 16.6 ± 2.8 <0.001 0.993 0.134 FT, pmol/La,b 20 ± 5 31 ± 8 38 ± 10 45 ± 24 462 ± 106 464 ± 81 <0.001 0.012 0.022 Total estradiol, pmol/La,b 153 ± 66 276 ± 200 188 ± 230 149 ± 49 64 ± 13 95.8 ± 25.3 <0.001 0.024 0.598 FE, pmol/La 2.7 ± 1.2 5.3 ± 3 3.9 ± 4.9 3.5 ± 1.4 1.7 ± 0.5 2.7 ± 0.7 0.027 0.003 0.549 Ratio FT/FEa,b,c 8.3 ± 2.6 7.5 ± 4.6 19 ± 12.7 14 ± 5.4 286.2 ± 60.4 179.0 ± 39.3 <0.001 0.087 0.272 SHBG, nmol/La,b,c 60 ± 24 43 ± 14 44 ± 18 32 ± 13 28 ± 10 18 ± 5 <0.001 0.007 0.788 DHEAS, mmol/La 5.1 ± 0.9 5.3 ± 1.5 8.5 ± 3.9 5.7 ± 2.1 6.5 ± 1.8 8.5 ± 3.2 0.025 0.798 0.027 Androstenedione, nmol/Lb,c 9.1 ± 3.1 9.4 ± 2.8 16.1 ± 3.5 13.3 ± 5.9 7.3 ± 1.8 9.8 ± 4.2 <0.001 0.992 0.175 Total cholesterol, mmol/L 4.5 ± 0.9 4.7 ± 1.2 4.4 ± 1.2 4.4 ± 1.0 3.9 ± 0.6 4.61 ± 0.75 0.635 0.263 0.607 LDL-cholesterol, mmol/L 2.8 ± 0.7 2.8 ± 0.4 2.6 ± 1.0 2.6 ± 1. 2.4 ± 0.5 3.1 ± 0.6 0.811 0.263 0.291 HDL-cholesterol, mmol/La,b 1.3 ± 0.3 1.2 ± 0.2 1.4 ± 0.2 1.2 ± 0.2 1.2 ± 0.2 1.0 ± 0.1 0.011 0.020 0.619 Triglycerides, mmol/L 0.8 ± 0.2 0.9 ± 0.4 0.8 ± 0.2 1.2 ± 0.6 0.9 ± 0.3 1.2 ± 0.6 0.423 0.055 0.755 Fasting glucose, mmol/Lb,c 4.7 ± 0.3 5.6 ± 0.4 4.8 ± 0.4 4.9 ± 0.4 5.3 ± 0.5 5.3 ± 0.5 0.014 0.014 0.039 Fasting insulin, pmol/L 53 ± 28 92 ± 45 47 ± 35 83 ± 28 37 ± 14 76 ± 28 0.383 <0.001 0.969 HOMA-IR 1.6 ± 0.9 3.3 ± 1.6 1.5 ± 1.1 2.6 ± 0.9 1.3 ± 0.5 2.6 ± 1.0 0.348 <0.001 0.780 Insulin sensitivity index 7.0 ± 2.7 3.3 ± 1.2 8.2 ± 5.2 3.5 ± 1.4 7.6 ± 2.9 3.6 ± 1.7 0.899 <0.001 0.997 Antibiotics, n (%) 0 (0) 0 (0) 0 (0) 2 (25) 1 (12.5) 0 (0) 0.217 1.000 NA Periodontal or dental disease, n (%) 0 (0) 1 (12.5) 0 (0) 0 (0) 0 (0) 0 (0) 0.434 1.000 NA Urinary tract infection, n (%) 0 (0) 0 (0) 0 (0) 1 (12.5) 0 (0) 0 (0) 0.434 1.000 NA Vaginal candidiasis or infection, n (%) 0 (0) 0 (0) 0 (0) 1 (12.5) NA 0.396 1.000 NA Characteristic Control Women Women With PCOS Men Group P Value Obesity P Value Interaction P Value Nonobese (n = 8) Obese (n = 8) Nonobese (n = 7) Obese (n = 8) Nonobese (n = 8) Obese (n = 7) Age, y 27.3 ± 4.0 27.3 ± 6.5 23.0 ± 7.9 29.9 ± 5.1 23.3 ± 3.3 23.6 ± 3.2 0.116 0.129 0.138 BMI, kg/m2 23.4 ± 1.9 35.9 ± 4.2 24.4 ± 2.8 37.0 ± 4.8 22.9 ± 2.0 34.6 ± 3.4 0.309 <0.001 0.888 Waist-to-hip ratioa,b 0.74 ± 0.1 0.83 ± 0.12 0.74 ± 0.04 0.85 ± 0.1 0.84 ± 0.04 0.9 ± 0.1 0.005 <0.001 0.542 Hirsutism score 1.3 ± 1.3 1.8 ± 1.2 10.1 ± 5.1 9.3 ± 4.5 NA <0.001 0.887 0.589 Total testosterone, nmol/La,b 1.6 ± 0.3 2.0 ± 0.5 2.5 0.7 2.4 1.0 19.2 ± 3.4 16.6 ± 2.8 <0.001 0.993 0.134 FT, pmol/La,b 20 ± 5 31 ± 8 38 ± 10 45 ± 24 462 ± 106 464 ± 81 <0.001 0.012 0.022 Total estradiol, pmol/La,b 153 ± 66 276 ± 200 188 ± 230 149 ± 49 64 ± 13 95.8 ± 25.3 <0.001 0.024 0.598 FE, pmol/La 2.7 ± 1.2 5.3 ± 3 3.9 ± 4.9 3.5 ± 1.4 1.7 ± 0.5 2.7 ± 0.7 0.027 0.003 0.549 Ratio FT/FEa,b,c 8.3 ± 2.6 7.5 ± 4.6 19 ± 12.7 14 ± 5.4 286.2 ± 60.4 179.0 ± 39.3 <0.001 0.087 0.272 SHBG, nmol/La,b,c 60 ± 24 43 ± 14 44 ± 18 32 ± 13 28 ± 10 18 ± 5 <0.001 0.007 0.788 DHEAS, mmol/La 5.1 ± 0.9 5.3 ± 1.5 8.5 ± 3.9 5.7 ± 2.1 6.5 ± 1.8 8.5 ± 3.2 0.025 0.798 0.027 Androstenedione, nmol/Lb,c 9.1 ± 3.1 9.4 ± 2.8 16.1 ± 3.5 13.3 ± 5.9 7.3 ± 1.8 9.8 ± 4.2 <0.001 0.992 0.175 Total cholesterol, mmol/L 4.5 ± 0.9 4.7 ± 1.2 4.4 ± 1.2 4.4 ± 1.0 3.9 ± 0.6 4.61 ± 0.75 0.635 0.263 0.607 LDL-cholesterol, mmol/L 2.8 ± 0.7 2.8 ± 0.4 2.6 ± 1.0 2.6 ± 1. 2.4 ± 0.5 3.1 ± 0.6 0.811 0.263 0.291 HDL-cholesterol, mmol/La,b 1.3 ± 0.3 1.2 ± 0.2 1.4 ± 0.2 1.2 ± 0.2 1.2 ± 0.2 1.0 ± 0.1 0.011 0.020 0.619 Triglycerides, mmol/L 0.8 ± 0.2 0.9 ± 0.4 0.8 ± 0.2 1.2 ± 0.6 0.9 ± 0.3 1.2 ± 0.6 0.423 0.055 0.755 Fasting glucose, mmol/Lb,c 4.7 ± 0.3 5.6 ± 0.4 4.8 ± 0.4 4.9 ± 0.4 5.3 ± 0.5 5.3 ± 0.5 0.014 0.014 0.039 Fasting insulin, pmol/L 53 ± 28 92 ± 45 47 ± 35 83 ± 28 37 ± 14 76 ± 28 0.383 <0.001 0.969 HOMA-IR 1.6 ± 0.9 3.3 ± 1.6 1.5 ± 1.1 2.6 ± 0.9 1.3 ± 0.5 2.6 ± 1.0 0.348 <0.001 0.780 Insulin sensitivity index 7.0 ± 2.7 3.3 ± 1.2 8.2 ± 5.2 3.5 ± 1.4 7.6 ± 2.9 3.6 ± 1.7 0.899 <0.001 0.997 Antibiotics, n (%) 0 (0) 0 (0) 0 (0) 2 (25) 1 (12.5) 0 (0) 0.217 1.000 NA Periodontal or dental disease, n (%) 0 (0) 1 (12.5) 0 (0) 0 (0) 0 (0) 0 (0) 0.434 1.000 NA Urinary tract infection, n (%) 0 (0) 0 (0) 0 (0) 1 (12.5) 0 (0) 0 (0) 0.434 1.000 NA Vaginal candidiasis or infection, n (%) 0 (0) 0 (0) 0 (0) 1 (12.5) NA 0.396 1.000 NA Data are means ± SD or counts (percentage). For categorical variables, differences between groups were analyzed by χ2 tests. The effects of group and obesity on continuous variables were analyzed by a two-way general linear model after application of two-step transformation as needed to ensure a normal distribution of the variables. Abbreviations: FE, free estradiol; FT, free testosterone; NA, not applicable. a P < 0.05 for the differences between men and control women, independently of obesity. b P < 0.05 for the differences between men and women with PCOS, independently of obesity. c P < 0.05 for the differences between women with PCOS and control women, independently of obesity. View Large Table 1. Clinical, Hormonal, and Metabolic Characteristics of Participants Characteristic Control Women Women With PCOS Men Group P Value Obesity P Value Interaction P Value Nonobese (n = 8) Obese (n = 8) Nonobese (n = 7) Obese (n = 8) Nonobese (n = 8) Obese (n = 7) Age, y 27.3 ± 4.0 27.3 ± 6.5 23.0 ± 7.9 29.9 ± 5.1 23.3 ± 3.3 23.6 ± 3.2 0.116 0.129 0.138 BMI, kg/m2 23.4 ± 1.9 35.9 ± 4.2 24.4 ± 2.8 37.0 ± 4.8 22.9 ± 2.0 34.6 ± 3.4 0.309 <0.001 0.888 Waist-to-hip ratioa,b 0.74 ± 0.1 0.83 ± 0.12 0.74 ± 0.04 0.85 ± 0.1 0.84 ± 0.04 0.9 ± 0.1 0.005 <0.001 0.542 Hirsutism score 1.3 ± 1.3 1.8 ± 1.2 10.1 ± 5.1 9.3 ± 4.5 NA <0.001 0.887 0.589 Total testosterone, nmol/La,b 1.6 ± 0.3 2.0 ± 0.5 2.5 0.7 2.4 1.0 19.2 ± 3.4 16.6 ± 2.8 <0.001 0.993 0.134 FT, pmol/La,b 20 ± 5 31 ± 8 38 ± 10 45 ± 24 462 ± 106 464 ± 81 <0.001 0.012 0.022 Total estradiol, pmol/La,b 153 ± 66 276 ± 200 188 ± 230 149 ± 49 64 ± 13 95.8 ± 25.3 <0.001 0.024 0.598 FE, pmol/La 2.7 ± 1.2 5.3 ± 3 3.9 ± 4.9 3.5 ± 1.4 1.7 ± 0.5 2.7 ± 0.7 0.027 0.003 0.549 Ratio FT/FEa,b,c 8.3 ± 2.6 7.5 ± 4.6 19 ± 12.7 14 ± 5.4 286.2 ± 60.4 179.0 ± 39.3 <0.001 0.087 0.272 SHBG, nmol/La,b,c 60 ± 24 43 ± 14 44 ± 18 32 ± 13 28 ± 10 18 ± 5 <0.001 0.007 0.788 DHEAS, mmol/La 5.1 ± 0.9 5.3 ± 1.5 8.5 ± 3.9 5.7 ± 2.1 6.5 ± 1.8 8.5 ± 3.2 0.025 0.798 0.027 Androstenedione, nmol/Lb,c 9.1 ± 3.1 9.4 ± 2.8 16.1 ± 3.5 13.3 ± 5.9 7.3 ± 1.8 9.8 ± 4.2 <0.001 0.992 0.175 Total cholesterol, mmol/L 4.5 ± 0.9 4.7 ± 1.2 4.4 ± 1.2 4.4 ± 1.0 3.9 ± 0.6 4.61 ± 0.75 0.635 0.263 0.607 LDL-cholesterol, mmol/L 2.8 ± 0.7 2.8 ± 0.4 2.6 ± 1.0 2.6 ± 1. 2.4 ± 0.5 3.1 ± 0.6 0.811 0.263 0.291 HDL-cholesterol, mmol/La,b 1.3 ± 0.3 1.2 ± 0.2 1.4 ± 0.2 1.2 ± 0.2 1.2 ± 0.2 1.0 ± 0.1 0.011 0.020 0.619 Triglycerides, mmol/L 0.8 ± 0.2 0.9 ± 0.4 0.8 ± 0.2 1.2 ± 0.6 0.9 ± 0.3 1.2 ± 0.6 0.423 0.055 0.755 Fasting glucose, mmol/Lb,c 4.7 ± 0.3 5.6 ± 0.4 4.8 ± 0.4 4.9 ± 0.4 5.3 ± 0.5 5.3 ± 0.5 0.014 0.014 0.039 Fasting insulin, pmol/L 53 ± 28 92 ± 45 47 ± 35 83 ± 28 37 ± 14 76 ± 28 0.383 <0.001 0.969 HOMA-IR 1.6 ± 0.9 3.3 ± 1.6 1.5 ± 1.1 2.6 ± 0.9 1.3 ± 0.5 2.6 ± 1.0 0.348 <0.001 0.780 Insulin sensitivity index 7.0 ± 2.7 3.3 ± 1.2 8.2 ± 5.2 3.5 ± 1.4 7.6 ± 2.9 3.6 ± 1.7 0.899 <0.001 0.997 Antibiotics, n (%) 0 (0) 0 (0) 0 (0) 2 (25) 1 (12.5) 0 (0) 0.217 1.000 NA Periodontal or dental disease, n (%) 0 (0) 1 (12.5) 0 (0) 0 (0) 0 (0) 0 (0) 0.434 1.000 NA Urinary tract infection, n (%) 0 (0) 0 (0) 0 (0) 1 (12.5) 0 (0) 0 (0) 0.434 1.000 NA Vaginal candidiasis or infection, n (%) 0 (0) 0 (0) 0 (0) 1 (12.5) NA 0.396 1.000 NA Characteristic Control Women Women With PCOS Men Group P Value Obesity P Value Interaction P Value Nonobese (n = 8) Obese (n = 8) Nonobese (n = 7) Obese (n = 8) Nonobese (n = 8) Obese (n = 7) Age, y 27.3 ± 4.0 27.3 ± 6.5 23.0 ± 7.9 29.9 ± 5.1 23.3 ± 3.3 23.6 ± 3.2 0.116 0.129 0.138 BMI, kg/m2 23.4 ± 1.9 35.9 ± 4.2 24.4 ± 2.8 37.0 ± 4.8 22.9 ± 2.0 34.6 ± 3.4 0.309 <0.001 0.888 Waist-to-hip ratioa,b 0.74 ± 0.1 0.83 ± 0.12 0.74 ± 0.04 0.85 ± 0.1 0.84 ± 0.04 0.9 ± 0.1 0.005 <0.001 0.542 Hirsutism score 1.3 ± 1.3 1.8 ± 1.2 10.1 ± 5.1 9.3 ± 4.5 NA <0.001 0.887 0.589 Total testosterone, nmol/La,b 1.6 ± 0.3 2.0 ± 0.5 2.5 0.7 2.4 1.0 19.2 ± 3.4 16.6 ± 2.8 <0.001 0.993 0.134 FT, pmol/La,b 20 ± 5 31 ± 8 38 ± 10 45 ± 24 462 ± 106 464 ± 81 <0.001 0.012 0.022 Total estradiol, pmol/La,b 153 ± 66 276 ± 200 188 ± 230 149 ± 49 64 ± 13 95.8 ± 25.3 <0.001 0.024 0.598 FE, pmol/La 2.7 ± 1.2 5.3 ± 3 3.9 ± 4.9 3.5 ± 1.4 1.7 ± 0.5 2.7 ± 0.7 0.027 0.003 0.549 Ratio FT/FEa,b,c 8.3 ± 2.6 7.5 ± 4.6 19 ± 12.7 14 ± 5.4 286.2 ± 60.4 179.0 ± 39.3 <0.001 0.087 0.272 SHBG, nmol/La,b,c 60 ± 24 43 ± 14 44 ± 18 32 ± 13 28 ± 10 18 ± 5 <0.001 0.007 0.788 DHEAS, mmol/La 5.1 ± 0.9 5.3 ± 1.5 8.5 ± 3.9 5.7 ± 2.1 6.5 ± 1.8 8.5 ± 3.2 0.025 0.798 0.027 Androstenedione, nmol/Lb,c 9.1 ± 3.1 9.4 ± 2.8 16.1 ± 3.5 13.3 ± 5.9 7.3 ± 1.8 9.8 ± 4.2 <0.001 0.992 0.175 Total cholesterol, mmol/L 4.5 ± 0.9 4.7 ± 1.2 4.4 ± 1.2 4.4 ± 1.0 3.9 ± 0.6 4.61 ± 0.75 0.635 0.263 0.607 LDL-cholesterol, mmol/L 2.8 ± 0.7 2.8 ± 0.4 2.6 ± 1.0 2.6 ± 1. 2.4 ± 0.5 3.1 ± 0.6 0.811 0.263 0.291 HDL-cholesterol, mmol/La,b 1.3 ± 0.3 1.2 ± 0.2 1.4 ± 0.2 1.2 ± 0.2 1.2 ± 0.2 1.0 ± 0.1 0.011 0.020 0.619 Triglycerides, mmol/L 0.8 ± 0.2 0.9 ± 0.4 0.8 ± 0.2 1.2 ± 0.6 0.9 ± 0.3 1.2 ± 0.6 0.423 0.055 0.755 Fasting glucose, mmol/Lb,c 4.7 ± 0.3 5.6 ± 0.4 4.8 ± 0.4 4.9 ± 0.4 5.3 ± 0.5 5.3 ± 0.5 0.014 0.014 0.039 Fasting insulin, pmol/L 53 ± 28 92 ± 45 47 ± 35 83 ± 28 37 ± 14 76 ± 28 0.383 <0.001 0.969 HOMA-IR 1.6 ± 0.9 3.3 ± 1.6 1.5 ± 1.1 2.6 ± 0.9 1.3 ± 0.5 2.6 ± 1.0 0.348 <0.001 0.780 Insulin sensitivity index 7.0 ± 2.7 3.3 ± 1.2 8.2 ± 5.2 3.5 ± 1.4 7.6 ± 2.9 3.6 ± 1.7 0.899 <0.001 0.997 Antibiotics, n (%) 0 (0) 0 (0) 0 (0) 2 (25) 1 (12.5) 0 (0) 0.217 1.000 NA Periodontal or dental disease, n (%) 0 (0) 1 (12.5) 0 (0) 0 (0) 0 (0) 0 (0) 0.434 1.000 NA Urinary tract infection, n (%) 0 (0) 0 (0) 0 (0) 1 (12.5) 0 (0) 0 (0) 0.434 1.000 NA Vaginal candidiasis or infection, n (%) 0 (0) 0 (0) 0 (0) 1 (12.5) NA 0.396 1.000 NA Data are means ± SD or counts (percentage). For categorical variables, differences between groups were analyzed by χ2 tests. The effects of group and obesity on continuous variables were analyzed by a two-way general linear model after application of two-step transformation as needed to ensure a normal distribution of the variables. Abbreviations: FE, free estradiol; FT, free testosterone; NA, not applicable. a P < 0.05 for the differences between men and control women, independently of obesity. b P < 0.05 for the differences between men and women with PCOS, independently of obesity. c P < 0.05 for the differences between women with PCOS and control women, independently of obesity. View Large Regarding metabolic parameters, high-density lipoprotein cholesterol levels were higher in female controls and women with PCOS than in men. Fasting glucose levels were reduced in patients with PCOS compared with controls. Total cholesterol, low-density lipoprotein cholesterol, triglycerides, insulin sensitivity index, and fasting HOMA-IR scores did not differ among groups. Regardless of the study group, obesity was characterized by increased levels of free testosterone, fasting insulin, and glucose and HOMA-IR score and decreased SHBG, high-density lipoprotein cholesterol levels, and insulin sensitivity index. In addition, we observed interactions between group and obesity. Obese control women had higher levels of free testosterone and fasting glucose than did their nonobese counterparts. Obese women with PCOS had decreased levels of DHEAS compared with nonobese patients with PCOS, whereas obese men had increased levels of DHEAS compared with their nonobese counterparts. Infectious events and use of antibiotics were rare and did not differ in frequency among groups (Table 1). However, to avoid the possible effect of prior use of antibiotics on gut microbiota, the results from the three participants who had used these drugs during the previous 3 months were eliminated from further analyses of gut microbiota. Differences in bacterial diversity of gut microbiota according to study group, obesity, and circulating sex hormone concentrations Sequencing the V4 16S microbial rDNA generated a dataset consisting of 6,507,940 filtered high-quality 16S rDNA gene sequences with 1,659,832 reads for control women, 2,462,150 reads for women with PCOS, and 2,385,958 reads for men, with an average of 151,437 sequences per sample (minimum, 44,795; maximum, 281,779). We assessed gut microbiota α and β diversity by Chao 1, Shannon, Jaccard, and Sorensen estimators in the three groups. Regardless of obesity, the analysis of α diversity revealed statistically significantly larger values in men than in control women (Fig. 1). We also observed a trend toward higher Chao 1 values in men when compared with women with PCOS that was close to reaching statistical significance (P = 0.053) (Fig. 1). When women were considered as a whole, irrespective of the presence or absence of PCOS, sex also influenced gut microbiota by lowering the Chao1 index in women (Fig. 1), and a similar near-significant trend (P = 0.087) was observed in the Shannon index (Fig. 1). With all participants considered as a whole, the Chao 1 and Shannon indexes showed positive correlations with total testosterone (Chao 1: r = 0.345, P = 0.023; Shannon: r = 0.337, P = 0.027) concentrations and the ratio of free testosterone to free estradiol (Chao 1: r = 0.362, P = 0.017; Shannon: r = 0.408 P = 0.007) and showed a negative correlation with total estradiol (Chao 1: r = −0.313, P = 0.041; Shannon: r = −0.377, P = 0.013). Figure 1. View largeDownload slide Box-and-whisker plots of the interactions of study group and obesity on the bacterial α and β diversity of gut microbiota. White boxes are nonobese participants and gray boxes correspond to obese individuals. The box represents the interquartile range, which contains the median, and the whiskers indicate the 90th and 10th percentiles. *P < 0.05; †P < 0.01; ‡P < 0.001. Figure 1. View largeDownload slide Box-and-whisker plots of the interactions of study group and obesity on the bacterial α and β diversity of gut microbiota. White boxes are nonobese participants and gray boxes correspond to obese individuals. The box represents the interquartile range, which contains the median, and the whiskers indicate the 90th and 10th percentiles. *P < 0.05; †P < 0.01; ‡P < 0.001. Even though obesity did not show independent effects on bacterial diversity in our series of young adults, we observed an interaction of obesity with the effects of sex and of the study group on the indexes of bacterial β diversity. When control women and patients with PCOS were considered as a whole, the Sorensen and Jaccard indexes decreased in obese women compared with lean women, whereas these indexes increased in obese men compared with lean men (Fig. 1). The analysis considering control women and patients with PCOS as separate groups revealed that the aforementioned interaction was actually related to the findings in patients with PCOS because obese patients had decreased β diversity indexes compared with their lean counterparts, whereas such differences were not present in control women (Fig. 1). Differences in bacterial phyla of gut microbiota according to study group and obesity We then investigated bacterial abundance at the phylum level but found no statistically significant differences among the three groups of participants (data not shown). On the contrary, we found an increased relative abundance of the candidatus Saccharibacteria phylum in obese patients compared with nonobese participants regardless of group and sex (obese, 0.06; nonobese, 0.02; P = 0.018). Differences in bacterial genera of gut microbiota according to study group, obesity, and circulating sex hormone concentrations At the genus level, we observed differences in the relative abundance of three genera among the study groups. The relative abundance of Catenibacterium and Kandleria genera was higher in women with PCOS and in men compared with control women; in fact, Kandleria was absent in the latter. Moreover, the abundance of the Raoultella genus was higher in men than in women with PCOS (Fig. 2). Figure 2. View largeDownload slide Differences in bacterial relative abundance at the genus level among groups. Columns are means plus SEM. White columns are control women, gray columns are patients with PCOS, and black columns are men. Only bacterial genera showing statistically significance differences (P < 0.0125) in at least one of the comparisons are shown. Missing columns indicate that the genus was not detected in this particular category. *P < 0.0125; †P < 0.010. Figure 2. View largeDownload slide Differences in bacterial relative abundance at the genus level among groups. Columns are means plus SEM. White columns are control women, gray columns are patients with PCOS, and black columns are men. Only bacterial genera showing statistically significance differences (P < 0.0125) in at least one of the comparisons are shown. Missing columns indicate that the genus was not detected in this particular category. *P < 0.0125; †P < 0.010. In addition, we also analyzed the presence or absence of bacterial genera in the study groups and observed that men more frequently had Butyricimonas (86% vs 44%; OR, 7.0; 95% CI, 1.2 to 42.8), Megasphaera (71% vs 25%; OR, 7.5; 95% CI, 1.5 to 37.9), and Paraprevotella (86% vs 31%; OR, 13.2; 95% CI, 2.1 to 85.5) genera compared with control women and that women with PCOS more frequently had Catenibacterium (77% vs 19%; OR, 14.4; 95% CI, 2.4 to 87.4) and Oribacterium (69% vs 19%; OR, 9.7; 95% CI, 1.7 to 54.5) genera compared with control women. When all participants were considered as a whole, total and free testosterone and the free testosterone to free estradiol ratio showed positive correlations with the relative abundance of Raoultella genus (Table 2) and serum androstenedione concentrations showed positive correlations with the abundance of the Kandleria genus (Table 2). Table 2. Correlations Between the Relative Abundance of Bacterial Genera, Markers of Weight Excess and Adiposity, Circulating Sex Hormones, and Biochemical Markers of Intermediate Metabolism in All Participants as a Whole Genera Markers of Weight Excess and Adiposity BMI (kg/m2) Fat Mass (kg) Fat Mass (% of Total Body Weight) Waist-to-Hip Ratio r P Value r P Value r P Value r P Value Alloprevotella −0.542a <0.001 −0.369b 0.016 −0.312b 0.045 −0.250 0.105 Anaerofustis 0.409a 0.006 0.420a 0.006 0.344b 0.026 0.304b 0.048 Atopobium 0.378b 0.013 0.342b 0.027 0.267 0.087 0.293 0.057 Neisseria −0.375b 0.013 −0.274 0.080 −0.322b 0.037 −0.411a 0.006 Scardovia 0.449a 0.003 0.442a 0.003 0.245 0.118 0.344b 0.024 Genera Markers of Weight Excess and Adiposity BMI (kg/m2) Fat Mass (kg) Fat Mass (% of Total Body Weight) Waist-to-Hip Ratio r P Value r P Value r P Value r P Value Alloprevotella −0.542a <0.001 −0.369b 0.016 −0.312b 0.045 −0.250 0.105 Anaerofustis 0.409a 0.006 0.420a 0.006 0.344b 0.026 0.304b 0.048 Atopobium 0.378b 0.013 0.342b 0.027 0.267 0.087 0.293 0.057 Neisseria −0.375b 0.013 −0.274 0.080 −0.322b 0.037 −0.411a 0.006 Scardovia 0.449a 0.003 0.442a 0.003 0.245 0.118 0.344b 0.024 Circulating Sex Hormone Concentrations Total T (nmol/L) Free T (pmol/L) E2 (pmol/L) Free E2 (pmol/L) Free T/Free E2
(nmol/pmol) Androstenedione (nmol/L) r P Value r P Value r P Value r P Value r P Value r P Value Catenibacterium −0.169 0.278 −0.100 0.524 −0.069 0.659 −0.123 0.430 −0.037 0.813 0.247 0.110 Kandleria −0.023 0.883 0.059 0.705 0.047 0.765 0.069 0.659 0.021 0.895 0.301b 0.049 Raoultella 0.317b 0.038 0.316b 0.039 −0.219 0.159 −0.138 0.379 0.308b 0.045 −0.131 0.404 Circulating Sex Hormone Concentrations Total T (nmol/L) Free T (pmol/L) E2 (pmol/L) Free E2 (pmol/L) Free T/Free E2
(nmol/pmol) Androstenedione (nmol/L) r P Value r P Value r P Value r P Value r P Value r P Value Catenibacterium −0.169 0.278 −0.100 0.524 −0.069 0.659 −0.123 0.430 −0.037 0.813 0.247 0.110 Kandleria −0.023 0.883 0.059 0.705 0.047 0.765 0.069 0.659 0.021 0.895 0.301b 0.049 Raoultella 0.317b 0.038 0.316b 0.039 −0.219 0.159 −0.138 0.379 0.308b 0.045 −0.131 0.404 Markers of Intermediate Metabolism Glucose (mmol/L) Insulin (pmol/L) ISI Total Cholesterol (mmol/L) HDL (mmol/L) LDL (mmol/L) Triglycerides (mmol/L) r P Value r P Value r P Value r P Value r P Value r P Value r P Value Atopobium 0.256 0.097 0.368b 0.015 −0.421a 0.005 0.237 0.126 −0.142 0.364 0.251 0.105 0.263 0.088 Catenibacterium −0.406a 0.007 −0.230 0.137 0.269 0.081 −0.237 0.125 0.285 0.064 −0.325b 0.034 −0.143 0.361 Raoultella 0.352b 0.021 −0.067 0.668 0.006 0.971 −0.091 0.563 −0.215 0.167 −0.103 0.511 0.050 0.748 Scardovia 0.053 0.737 0.388b 0.010 −0.425a 0.005 0.271 0.079 0.043 0.783 0.101 0.518 0.318b 0.038 Markers of Intermediate Metabolism Glucose (mmol/L) Insulin (pmol/L) ISI Total Cholesterol (mmol/L) HDL (mmol/L) LDL (mmol/L) Triglycerides (mmol/L) r P Value r P Value r P Value r P Value r P Value r P Value r P Value Atopobium 0.256 0.097 0.368b 0.015 −0.421a 0.005 0.237 0.126 −0.142 0.364 0.251 0.105 0.263 0.088 Catenibacterium −0.406a 0.007 −0.230 0.137 0.269 0.081 −0.237 0.125 0.285 0.064 −0.325b 0.034 −0.143 0.361 Raoultella 0.352b 0.021 −0.067 0.668 0.006 0.971 −0.091 0.563 −0.215 0.167 −0.103 0.511 0.050 0.748 Scardovia 0.053 0.737 0.388b 0.010 −0.425a 0.005 0.271 0.079 0.043 0.783 0.101 0.518 0.318b 0.038 Abbreviations: E2, estradiol; HDL, high-density lipoprotein cholesterol; ISI, insulin sensitivity index; LDL, low-density lipoprotein cholesterol; T, testosterone. a P < 0.01. b P < 0.05. View Large Table 2. Correlations Between the Relative Abundance of Bacterial Genera, Markers of Weight Excess and Adiposity, Circulating Sex Hormones, and Biochemical Markers of Intermediate Metabolism in All Participants as a Whole Genera Markers of Weight Excess and Adiposity BMI (kg/m2) Fat Mass (kg) Fat Mass (% of Total Body Weight) Waist-to-Hip Ratio r P Value r P Value r P Value r P Value Alloprevotella −0.542a <0.001 −0.369b 0.016 −0.312b 0.045 −0.250 0.105 Anaerofustis 0.409a 0.006 0.420a 0.006 0.344b 0.026 0.304b 0.048 Atopobium 0.378b 0.013 0.342b 0.027 0.267 0.087 0.293 0.057 Neisseria −0.375b 0.013 −0.274 0.080 −0.322b 0.037 −0.411a 0.006 Scardovia 0.449a 0.003 0.442a 0.003 0.245 0.118 0.344b 0.024 Genera Markers of Weight Excess and Adiposity BMI (kg/m2) Fat Mass (kg) Fat Mass (% of Total Body Weight) Waist-to-Hip Ratio r P Value r P Value r P Value r P Value Alloprevotella −0.542a <0.001 −0.369b 0.016 −0.312b 0.045 −0.250 0.105 Anaerofustis 0.409a 0.006 0.420a 0.006 0.344b 0.026 0.304b 0.048 Atopobium 0.378b 0.013 0.342b 0.027 0.267 0.087 0.293 0.057 Neisseria −0.375b 0.013 −0.274 0.080 −0.322b 0.037 −0.411a 0.006 Scardovia 0.449a 0.003 0.442a 0.003 0.245 0.118 0.344b 0.024 Circulating Sex Hormone Concentrations Total T (nmol/L) Free T (pmol/L) E2 (pmol/L) Free E2 (pmol/L) Free T/Free E2
(nmol/pmol) Androstenedione (nmol/L) r P Value r P Value r P Value r P Value r P Value r P Value Catenibacterium −0.169 0.278 −0.100 0.524 −0.069 0.659 −0.123 0.430 −0.037 0.813 0.247 0.110 Kandleria −0.023 0.883 0.059 0.705 0.047 0.765 0.069 0.659 0.021 0.895 0.301b 0.049 Raoultella 0.317b 0.038 0.316b 0.039 −0.219 0.159 −0.138 0.379 0.308b 0.045 −0.131 0.404 Circulating Sex Hormone Concentrations Total T (nmol/L) Free T (pmol/L) E2 (pmol/L) Free E2 (pmol/L) Free T/Free E2
(nmol/pmol) Androstenedione (nmol/L) r P Value r P Value r P Value r P Value r P Value r P Value Catenibacterium −0.169 0.278 −0.100 0.524 −0.069 0.659 −0.123 0.430 −0.037 0.813 0.247 0.110 Kandleria −0.023 0.883 0.059 0.705 0.047 0.765 0.069 0.659 0.021 0.895 0.301b 0.049 Raoultella 0.317b 0.038 0.316b 0.039 −0.219 0.159 −0.138 0.379 0.308b 0.045 −0.131 0.404 Markers of Intermediate Metabolism Glucose (mmol/L) Insulin (pmol/L) ISI Total Cholesterol (mmol/L) HDL (mmol/L) LDL (mmol/L) Triglycerides (mmol/L) r P Value r P Value r P Value r P Value r P Value r P Value r P Value Atopobium 0.256 0.097 0.368b 0.015 −0.421a 0.005 0.237 0.126 −0.142 0.364 0.251 0.105 0.263 0.088 Catenibacterium −0.406a 0.007 −0.230 0.137 0.269 0.081 −0.237 0.125 0.285 0.064 −0.325b 0.034 −0.143 0.361 Raoultella 0.352b 0.021 −0.067 0.668 0.006 0.971 −0.091 0.563 −0.215 0.167 −0.103 0.511 0.050 0.748 Scardovia 0.053 0.737 0.388b 0.010 −0.425a 0.005 0.271 0.079 0.043 0.783 0.101 0.518 0.318b 0.038 Markers of Intermediate Metabolism Glucose (mmol/L) Insulin (pmol/L) ISI Total Cholesterol (mmol/L) HDL (mmol/L) LDL (mmol/L) Triglycerides (mmol/L) r P Value r P Value r P Value r P Value r P Value r P Value r P Value Atopobium 0.256 0.097 0.368b 0.015 −0.421a 0.005 0.237 0.126 −0.142 0.364 0.251 0.105 0.263 0.088 Catenibacterium −0.406a 0.007 −0.230 0.137 0.269 0.081 −0.237 0.125 0.285 0.064 −0.325b 0.034 −0.143 0.361 Raoultella 0.352b 0.021 −0.067 0.668 0.006 0.971 −0.091 0.563 −0.215 0.167 −0.103 0.511 0.050 0.748 Scardovia 0.053 0.737 0.388b 0.010 −0.425a 0.005 0.271 0.079 0.043 0.783 0.101 0.518 0.318b 0.038 Abbreviations: E2, estradiol; HDL, high-density lipoprotein cholesterol; ISI, insulin sensitivity index; LDL, low-density lipoprotein cholesterol; T, testosterone. a P < 0.01. b P < 0.05. View Large We then investigated which genera of the gut microbiota changed in abundance with obesity regardless of the study group being analyzed. The Anaerofustis, Atopobium, and Scardovia genera were more abundant, and the Alloprevotella and Neisseria genera were less abundant, in obese than in nonobese persons (Fig. 3). Moreover, obese compared with nonobese participants more frequently had Anaerofustis (47.6% vs 9.1%; P = 0.006), Klebsiella (61.9% vs 31.8%; P = 0.047), Mogibacterium (19% vs 0%; P = 0.048), and Scardovia (42.9% vs 9.1%; P = 0.013), whereas the presence of Alloprevotella (59.1% vs 14.3%; P = 0.003), Gordonibacter (86.4% vs 52.4%; P = 0.017), Neisseria (81.8% vs 42.9%; P = 0.009), Propionibacterium (50% vs 19%; P = 0.034), and Pseudomonas (54.5% vs 19%; P = 0.017) was more frequent in nonobese than in obese participants. Figure 3. View largeDownload slide Differences in bacterial relative abundance at the genus level among nonobese and obese participants. Columns are means plus SEM. White columns are nonobese participants and gray columns are obese participants. Only bacterial genera showing statistically significant differences (P < 0.0125) in the comparisons are shown. *P < 0.0125; †P < 0.010. Figure 3. View largeDownload slide Differences in bacterial relative abundance at the genus level among nonobese and obese participants. Columns are means plus SEM. White columns are nonobese participants and gray columns are obese participants. Only bacterial genera showing statistically significant differences (P < 0.0125) in the comparisons are shown. *P < 0.0125; †P < 0.010. The interactions between study group and the effects of obesity revealed statistically significant differences in the abundance of two bacterial genera, Atopobium and Neisseria. Obesity exerted a different effect in control women and in women PCOS and men; it was associated with an increase in the abundance of Atopobium in control women but not in women with PCOS and men (Fig. 4). Moreover, obesity was associated with a decrease in the abundance of Neisseria in control women, an association that was not found in women with PCOS or in men (Fig. 4). Figure 4. View largeDownload slide Interactions between study group and obesity in bacterial relative abundance at the genus level. Columns are means plus SEM. White columns are nonobese participants and gray columns are obese participants. Only bacterial genera showing statistically significant differences (P < 0.0125) in the comparisons are shown. †P < 0.010. Figure 4. View largeDownload slide Interactions between study group and obesity in bacterial relative abundance at the genus level. Columns are means plus SEM. White columns are nonobese participants and gray columns are obese participants. Only bacterial genera showing statistically significant differences (P < 0.0125) in the comparisons are shown. †P < 0.010. Given the impact of obesity on the findings described above, we analyzed the correlations between the relative abundance of bacterial genera with several metabolic parameters. When all participants were considered as a whole, irrespective of group, sex, and obesity, BMI showed positive correlations with the abundance of Anaerofustis, Atopobium, and Scardovia and negative correlations with that of Alloprevotella and Neisseria (Table 2). Fat mass, expressed as total kilograms or percentage of total body weight, showed positive correlations with the abundances of Anaerofustis, Atopobium, and Scardovia and negative correlations with those of Neisseria and Alloprevotella (Table 2). Fasting glucose levels showed positive correlations with abundance of Raoultella and negative correlations with abundance of Catenibacterium, whereas insulin showed positive correlations, and the insulin sensitivity index showed negative correlations, with the abundance of Atopobium and Scardovia (Table 2). Discussion Our study showed that the diversity and composition of the gut microbiota were influenced by sex and obesity and that sex hormones may be related to these differences. Indexes of bacterial α diversity in the gut microbiota were reduced in women with and without androgen excess when compared with men. Such an effect of sex was not found in a recent study; that study showed no differences in α bacterial diversity between men and women (13). However, the fact that the women included in that study were postmenopausal (13) may have obscured the sexual dimorphism in the diversity of gut microbiota and possible influences of sex hormones suggested in the analysis of our young adults. Moreover, because α diversity showed positive correlations with testosterone levels and ratio of free testosterone to free estradiol and negative correlations with estradiol concentrations, our study further suggests that androgens and estrogens might be involved in the sexual dimorphism reported here. However, being derived from a cross-sectional study, our data cannot clarify the causality of these associations, and the possibility exists that gut microbes participate in the regulation sex hormones and, conversely, that sex hormones modify microbial diversity (12). Contrary to the results in α diversity, the changes in β bacterial diversity of gut microbiota found in our young adults was heavily influenced by obesity, exemplifying the complex interactions among sex, sex hormones, and obesity on gut bacterial diversity. This effect was particularly important in the subset of women with PCOS. Even though the current literature on the effect of obesity on bacterial diversity yielded conflicting results (22–25), our study addressed the role of sex and sex hormones on such an effect; this issue is of paramount importance because gonadal dysfunction, consisting of androgen excess in women and androgen deficiency in men, is prevalent in obese persons (26, 27). Moreover, the negative effect of obesity on the bacterial diversity of gut microbiota might be explained, at least in part, by the ecological theory, which predicts that species-rich communities, with different species specializing to each potentially limiting resource, are less susceptible to invasion because they use limiting resources more efficiently (22). Excess nutrient loading may result in decreased ecosystem diversity as a small number of species overgrow and outcompete everything else (22). In relative agreement with our finding of decreased β diversity in obese patients with PCOS, Lindheim et al. (28) and Torres et al. (29) reported reductions in indexes of α diversity in women with PCOS compared with control women. The fact that we did not find such a reduction in indexes of α diversity in our patients might be related to differences in the criteria used to diagnose PCOS—all patients had hyperandrogenism in our series compared with only 60% in previous studies (28, 29)—and in the characteristics of the populations analyzed because the aforementioned studies did not include obese participants (28, 29). At the phyla level, our study showed that the abundance of candidatus Saccharibacteria was higher in obese patients than in nonobese patients. This phylum plays a role in the degradation of various organic compounds as well as glucose, butyrate, oleic acid, and amino acids (30). However, studies in mice reported a reduced abundance of this phylum in response to a high-fat diet (31, 32). Such an effect was unlikely to influence our present results because all the participants maintained similar diets in the days preceding sampling. Sex also influenced the composition of gut microbiota at the genera level, with women showing reduced abundance of Raoultella genus when compared with men. Moreover, the Kandleria genus was not present in control women, and the Megasphaera, Paraprevotella, and Butyricimonas genera were less frequent in control women than in men. Most of these genera have not been related to sex before in human studies, with the notable exception of Butyricimonas, which was also decreased in women in earlier studies (13). The influence of sex on the composition of gut microbiota is in conceptual agreement with animal data that also demonstrate the remarkable interaction of these effects with host genetics and environmental factors, such as diet (12, 33). Accordingly, in our study Paraprevotella showed the strongest positive correlation with testosterone and negative associations with estradiol concentrations among the genera that correlated with sex hormones and sex. In addition, androgenized rats are characterized by an increase in the abundance of the Paraprevotella genus (34). But similar to what has been suggested for bacterial diversity, evidence suggests that gut microbiota may catalyze modifications of steroid hormones (35) and might play some role in the regulation of sex hormone concentrations in the body (36). The analysis of bacterial composition at the genus level revealed an abnormality characteristic of women with PCOS. These patients presented increased abundance of the Catenibacterium and Kandleria genera compared with both female and male controls. Moreover, the abundance of Kandleria correlated positively with circulating androstenedione concentrations when all participants were considered as a whole, suggesting an association with androgen levels. Catenibacterium are gram-positive anaerobic bacteria that use glucose to produce acetic, lactic, butyric, and isobutyric acids (37). Several studies reported an increase in the abundance of Catenibacterium in infectious and autoimmune disorders (38, 39), suggesting a link with the proinflammatory response. Moreover, the abundance of this genus in human microbiota increased with western-style and carbohydrate-rich diets (40, 41), whereas in ob/ob mice, feeding with prebiotics decreased the abundance of Catenibacterium in parallel with amelioration of low-grade inflammation and fat-mass accumulation (42). Kandleria are gram-positive anaerobic bacteria recently reclassified on the basis of phylogenetic information from the genus Lactobacillus and placement in family Erysipelotrichaceae (43). Kandleria metabolizes glucose producing lactic acid (43) and have not been related to metabolism or sexual hormones before in human studies. To the best of our knowledge, no data have been published linking differences in the Raoultella genus with PCOS and/or obesity. Obesity increased the abundance of Anaerofustis, Atopobium, and Scardovia and decreased the abundance of Alloprevotella and Neisseria. Of these, Atopobium increased specifically in obese control women compared with their nonobese counterparts. Even though the abundance of Atopobium has been reported to be decreased in fecal samples of patients with diabetes who presented with higher BMI than the controls (44) and in obese children (45), the abundance of this genus increased in male mice with high-fat diet–induced obesity, in agreement with our present results (46). A similar model of obesity in the rat resulted in a decrease in the abundance of Alloprevotella (47) similar to the increase found in our obese participants. Only a few human and animal studies studied gut microbiota genera in PCOS (28, 29, 48–50). In mice, letrozole-induced PCOS resulted into a reduction in overall species and phylogenetic richness, where the abundance of the Alistipes genus was decreased and that of the Allobaculum and Blautia genera were increased (49). A similar model in rats resulted into reduced Lactobacillus, Ruminococcus, and Clostridium abundances and increased Prevotella levels (50). Moreover, correction of this dysbiosis improved the menstrual pattern in these animals (50). In humans, available data suggest an association between the abundance of certain gut microbiota genera and PCOS. In Europeans, PCOS is associated with a reduced relative abundance of the phylum Tenericutes, two unclassified genera from the order ML615J-28 (belonging to the phylum Tenericutes), and the family S24-7 (belonging to the phylum Bacteroidetes) (28). However, as many as 40% of the women diagnosed with PCOS lacked androgen excess (28) in a phenotype that usually associates much milder metabolic disorders compared with the classic PCOS phenotype that motivated the diagnosis in our series, explaining the different outcomes of both studies. In addition, patients with PCOS were younger than controls in this study (28) and therefore, such results may have been influenced by age. In the Chinese population nonobese and obese control women and women with PCOS showed differences in relative abundance in 28 operational taxonomic units (48). However, most differences in abundance in these studies (28, 48) pertained to unclassified bacterial genera, making difficult a direct comparison with our present results. Among the advantages of our present study, we highlight the nontargeted approach to the analysis of the gut microbiome, which may have permitted the finding of associations of gut microbiota with sex, sex hormones, and PCOS with genera that might have been missed by targeted approaches. But we also acknowledge that our experiments are not free from limitations, especially with regard to the cross-sectional design used, the limited sample size analyzed, and the high interindividual variability observed. We tried to compensate for these limitations by comparing only rigorously defined and homogeneous groups of participants of similar age and BMI. Similarly, host specificities or diet-related differences may explain variations observed in the composition of gut microbiota, yet we tried to minimize these confounding factors by excluding from analysis the fecal samples from the three participants who had used antibiotics and by standardizing the diet of all these persons for 3 days before sampling. However, we cannot definitely exclude a putative impact of differences in the diet occurring before these 3 days on the microbiota of these individuals. The data presented were obtained for participants living in a specific geographical region, which does not necessarily reflect the environment in other locations. Similarly, we included only patients with hyperandrogenic phenotypes of PCOS and, accordingly, our results should not be extrapolated to nonhyperandrogenic patients with PCOS. Finally, a more detailed structural and functional analysis of intestinal microbiota is required to better define lifestyle- or diet-related effects. Discrimination at the species or even strain level might be of interest for some bacterial groups in order to detect subtle variations in dominant groups. In summary, our findings indicate that the diversity and composition of gut microbiota are influenced by the combined effects of sex, sex hormone concentrations, and obesity. Intervention studies designed to determine the causality of these associations may permit targeting gut microbiota as putative diagnostic and therapeutic targets for PCOS, obesity, and associated metabolic comorbidities. Abbreviations: Abbreviations: BMI body mass index DHEAS dehydroepiandrosterone-sulfate HOMA-IR homeostasis model assessment of insulin resistance PCOS polycystic ovary syndrome rDNA ribosomal DNA Acknowledgments Financial Support: Supported by grants PI11/00357, PI15/01686 SEXMETAB, and PIE16/00050 SEXCOMPLEX from Instituto de Salud Carlos III, Spanish Ministry of Economy and Competitiveness (H.F.E.M.). CIBERDEM and IRYCIS are also initiatives of Instituto de Salud Carlos III. Supported in part by Fondo Europeo de Desarrollo Regional (FEDER, European Union). Disclosure Summary: The authors have nothing to disclose. References 1. Escobar-Morreale HF , Luque-Ramírez M , San Millán JL . The molecular-genetic basis of functional hyperandrogenism and the polycystic ovary syndrome . Endocr Rev . 2005 ; 26 ( 2 ): 251 – 282 . 2. Escobar-Morreale HF , San Millán JL . Abdominal adiposity and the polycystic ovary syndrome . Trends Endocrinol Metab . 2007 ; 18 ( 7 ): 266 – 272 . 3. Escobar-Morreale HF , Samino S , Insenser M , Vinaixa M , Luque-Ramírez M , Lasunción MA , Correig X . Metabolic heterogeneity in polycystic ovary syndrome is determined by obesity: plasma metabolomic approach using GC-MS . Clin Chem . 2012 ; 58 ( 6 ): 999 – 1009 . 4. Luque-Ramírez M , Escobar-Morreale HF . Polycystic ovary syndrome as a paradigm for prehypertension, prediabetes, and preobesity . Curr Hypertens Rep . 2014 ; 16 ( 12 ): 500 . 5. Wild RA , Carmina E , Diamanti-Kandarakis E , Dokras A , Escobar-Morreale HF , Futterweit W , Lobo R , Norman RJ , Talbott E , Dumesic DA . Assessment of cardiovascular risk and prevention of cardiovascular disease in women with the polycystic ovary syndrome: a consensus statement by the Androgen Excess and Polycystic Ovary Syndrome (AE-PCOS) Society . J Clin Endocrinol Metab . 2010 ; 95 ( 5 ): 2038 – 2049 . 6. Clemente JC , Ursell LK , Parfrey LW , Knight R . The impact of the gut microbiota on human health: an integrative view . Cell . 2012 ; 148 ( 6 ): 1258 – 1270 . 7. Bäckhed F , Ding H , Wang T , Hooper LV , Koh GY , Nagy A , Semenkovich CF , Gordon JI . The gut microbiota as an environmental factor that regulates fat storage . Proc Natl Acad Sci USA . 2004 ; 101 ( 44 ): 15718 – 15723 . 8. Turnbaugh PJ , Ley RE , Mahowald MA , Magrini V , Mardis ER , Gordon JI . An obesity-associated gut microbiome with increased capacity for energy harvest . Nature . 2006 ; 444 ( 7122 ): 1027 – 1031 . 9. Sonnenburg JL , Bäckhed F . Diet-microbiota interactions as moderators of human metabolism . Nature . 2016 ; 535 ( 7610 ): 56 – 64 . 10. Kovacs A , Ben-Jacob N , Tayem H , Halperin E , Iraqi FA , Gophna U . Genotype is a stronger determinant than sex of the mouse gut microbiota . Microb Ecol . 2011 ; 61 ( 2 ): 423 – 428 . 11. Lay C , Rigottier-Gois L , Holmstrøm K , Rajilic M , Vaughan EE , de Vos WM , Collins MD , Thiel R , Namsolleck P , Blaut M , Doré J . Colonic microbiota signatures across five northern European countries . Appl Environ Microbiol . 2005 ; 71 ( 7 ): 4153 – 4155 . 12. Yurkovetskiy L , Burrows M , Khan AA , Graham L , Volchkov P , Becker L , Antonopoulos D , Umesaki Y , Chervonsky AV . Gender bias in autoimmunity is influenced by microbiota . Immunity . 2013 ; 39 ( 2 ): 400 – 412 . 13. Haro C , Rangel-Zúñiga OA , Alcalá-Díaz JF , Gómez-Delgado F , Pérez-Martínez P , Delgado-Lista J , Quintana-Navarro GM , Landa BB , Navas-Cortés JA , Tena-Sempere M , Clemente JC , López-Miranda J , Pérez-Jiménez F , Camargo A . Intestinal microbiota is influenced by gender and body mass index . PLoS One . 2016 ; 11 ( 5 ): e0154090 . 14. Zawadzki JK , Dunaif A . Diagnostic criteria for polycystic ovary syndrome: Towards a rational approach. In: Dunaif A , Givens JR , Haseltine FP , and Merriam GR , eds. Polycystic ovary syndrome . Boston : Blackwell Scientific Publications ; 1992 : 377 – 384 . 15. Azziz R , Carmina E , Dewailly D , Diamanti-Kandarakis E , Escobar-Morreale HF , Futterweit W , Janssen OE , Legro RS , Norman RJ , Taylor AE , Witchel SF ; Task Force on the Phenotype of the Polycystic Ovary Syndrome of The Androgen Excess and PCOS Society . The Androgen Excess and PCOS Society criteria for the polycystic ovary syndrome: the complete task force report . Fertil Steril . 2009 ; 91 ( 2 ): 456 – 488 . 16. Escobar-Morreale HF , Sanchón R , San Millán JL . A prospective study of the prevalence of nonclassical congenital adrenal hyperplasia among women presenting with hyperandrogenic symptoms and signs . J Clin Endocrinol Metab . 2008 ; 93 ( 2 ): 527 – 533 . 17. Matthews DR , Hosker JP , Rudenski AS , Naylor BA , Treacher DF , Turner RC . Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man . Diabetologia . 1985 ; 28 ( 7 ): 412 – 419 . 18. Matsuda M , DeFronzo RA . Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp . Diabetes Care . 1999 ; 22 ( 9 ): 1462 – 1470 . 19. Templeton GF . A two-step approach for transforming continuous variables to normal: implications and recommendations for IS research . Comm Assoc Info Syst . 2011 : 28 :41–51 . 20. White JR , Nagarajan N , Pop M . Statistical methods for detecting differentially abundant features in clinical metagenomic samples . PLOS Comput Biol . 2009 ; 5 ( 4 ): e1000352 . 21. Jonsson V , Österlund T , Nerman O , Kristiansson E . Statistical evaluation of methods for identification of differentially abundant genes in comparative metagenomics . BMC Genomics . 2016 ; 17 ( 1 ): 78 . 22. Lozupone CA , Stombaugh JI , Gordon JI , Jansson JK , Knight R . Diversity, stability and resilience of the human gut microbiota . Nature . 2012 ; 489 ( 7415 ): 220 – 230 . 23. Le Chatelier E , Nielsen T , Qin J , Prifti E , Hildebrand F , Falony G , Almeida M , Arumugam M , Batto JM , Kennedy S , Leonard P , Li J , Burgdorf K , Grarup N , Jørgensen T , Brandslund I , Nielsen HB , Juncker AS , Bertalan M , Levenez F , Pons N , Rasmussen S , Sunagawa S , Tap J , Tims S , Zoetendal EG , Brunak S , Clément K , Doré J , Kleerebezem M , Kristiansen K , Renault P , Sicheritz-Ponten T , de Vos WM , Zucker JD , Raes J , Hansen T , Bork P , Wang J , Ehrlich SD , Pedersen O ; MetaHIT consortium . Richness of human gut microbiome correlates with metabolic markers . Nature . 2013 ; 500 ( 7464 ): 541 – 546 . 24. Walters WA , Xu Z , Knight R . Meta-analyses of human gut microbes associated with obesity and IBD . FEBS Lett . 2014 ; 588 ( 22 ): 4223 – 4233 . 25. Kasai C , Sugimoto K , Moritani I , Tanaka J , Oya Y , Inoue H , Tameda M , Shiraki K , Ito M , Takei Y , Takase K . Comparison of the gut microbiota composition between obese and non-obese individuals in a Japanese population, as analyzed by terminal restriction fragment length polymorphism and next-generation sequencing . BMC Gastroenterol . 2015 ; 15 ( 1 ): 100 . 26. Escobar-Morreale HF , Alvarez-Blasco F , Botella-Carretero JI , Luque-Ramírez M . The striking similarities in the metabolic associations of female androgen excess and male androgen deficiency . Hum Reprod . 2014 ; 29 ( 10 ): 2083 – 2091 . 27. Escobar-Morreale HF , Santacruz E , Luque-Ramírez M , Botella Carretero JI . Prevalence of ‘obesity-associated gonadal dysfunction’ in severely obese men and women and its resolution after bariatric surgery: a systematic review and meta-analysis . Hum Reprod Update . 2017 ; 23 ( 4 ): 390 – 408 . 28. Lindheim L , Bashir M , Münzker J , Trummer C , Zachhuber V , Leber B , Horvath A , Pieber TR , Gorkiewicz G , Stadlbauer V , Obermayer-Pietsch B . Alterations in gut microbiome composition and barrier function are associated with reproductive and metabolic defects in women with polycystic ovary syndrome (PCOS): A pilot study . PLoS One . 2017 ; 12 ( 1 ): e0168390 . 29. Torres PJ , Siakowska M , Banaszewska B , Pawelczyk L , Duleba AJ , Kelley ST , Thackray VG . Gut microbial diversity in women with polycystic ovary syndrome correlates with hyperandrogenism . J Clin Endocrinol Metab . 2018 ; 103 ( 4 ): 1502 – 1511 . 30. Kindaichi T , Yamaoka S , Uehara R , Ozaki N , Ohashi A , Albertsen M , Nielsen PH , Nielsen JL . Phylogenetic diversity and ecophysiology of Candidate phylum Saccharibacteria in activated sludge . FEMS Microbiol Ecol . 2016 ; 92 ( 6 ): fiw078 . 31. Lu Y , Fan C , Li P , Lu Y , Chang X , Qi K . Short chain fatty acids prevent high-fat-diet-induced obesity in mice by regulating G protein-coupled receptors and gut microbiota . Sci Rep . 2016 ; 6 ( 1 ): 37589 . 32. Moya-Pérez A , Neef A , Sanz Y . Bifidobacterium pseudocatenulatum CECT 7765 reduces obesity-associated inflammation by restoring the lymphocyte-macrophage balance and gut microbiota structure in high-fat diet-fed mice . PLoS One . 2015 ; 10 ( 7 ): e0126976 . 33. Org E , Mehrabian M , Parks BW , Shipkova P , Liu X , Drake TA , Lusis AJ . Sex differences and hormonal effects on gut microbiota composition in mice . Gut Microbes . 2016 ; 7 ( 4 ): 313 – 322 . 34. Moreno-Indias I , Cardona F , Tinahones FJ , Queipo-Ortuño MI . Impact of the gut microbiota on the development of obesity and type 2 diabetes mellitus . Front Microbiol . 2014 ; 5 : 190 . 35. Lombardi P , Goldin B , Boutin E , Gorbach SL . Metabolism of androgens and estrogens by human fecal microorganisms . J Steroid Biochem . 1978 ; 9 ( 8 ): 795 – 801 . 36. Donova MV , Egorova OV . Microbial steroid transformations: current state and prospects . Appl Microbiol Biotechnol . 2012 ; 94 ( 6 ): 1423 – 1447 . 37. Kageyama A , Benno Y . Catenibacterium mitsuokai gen. nov., sp. nov., a gram-positive anaerobic bacterium isolated from human faeces . Int J Syst Evol Microbiol . 2000 ; 50 ( Pt 4 ): 1595 – 1599 . 38. Wagner BD , Frank DN , Zipris D . Response to comment on Alkanani et al. Alterations in intestinal microbiota correlate with susceptibility to type 1 diabetes. Diabetes 2015;64:3510-3520 . Diabetes . 2015 ; 64 ( 10 ): e41 . 39. Lozupone CA , Li M , Campbell TB , Flores SC , Linderman D , Gebert MJ , Knight R , Fontenot AP , Palmer BE . Alterations in the gut microbiota associated with HIV-1 infection . Cell Host Microbe . 2013 ; 14 ( 3 ): 329 – 339 . 40. Turnbaugh PJ , Ridaura VK , Faith JJ , Rey FE , Knight R , Gordon JI . The effect of diet on the human gut microbiome: a metagenomic analysis in humanized gnotobiotic mice . Sci Transl Med . 2009 ; 1 ( 6 ): 6ra14 . 41. Wu GD , Chen J , Hoffmann C , Bittinger K , Chen YY , Keilbaugh SA , Bewtra M , Knights D , Walters WA , Knight R , Sinha R , Gilroy E , Gupta K , Baldassano R , Nessel L , Li H , Bushman FD , Lewis JD . Linking long-term dietary patterns with gut microbial enterotypes . Science . 2011 ; 334 ( 6052 ): 105 – 108 . 42. Everard A , Lazarevic V , Derrien M , Girard M , Muccioli GG , Neyrinck AM , Possemiers S , Van Holle A , François P , de Vos WM , Delzenne NM , Schrenzel J , Cani PD . Responses of gut microbiota and glucose and lipid metabolism to prebiotics in genetic obese and diet-induced leptin-resistant mice [published correction appears in Diabetes. 2011;60(12):3307] . Diabetes . 2011 ; 60 ( 11 ): 2775 – 2786 . 43. Salvetti E , Felis GE , Dellaglio F , Castioni A , Torriani S , Lawson PA . Reclassification of Lactobacillus catenaformis (Eggerth 1935) Moore and Holdeman 1970 and Lactobacillus vitulinus Sharpe et al. 1973 as Eggerthia catenaformis gen. nov., comb. nov. and Kandleria vitulina gen. nov., comb. nov., respectively . Int J Syst Evol Microbiol . 2011 ; 61 ( Pt 10 ): 2520 – 2524 . 44. Sato J , Kanazawa A , Ikeda F , Yoshihara T , Goto H , Abe H , Komiya K , Kawaguchi M , Shimizu T , Ogihara T , Tamura Y , Sakurai Y , Yamamoto R , Mita T , Fujitani Y , Fukuda H , Nomoto K , Takahashi T , Asahara T , Hirose T , Nagata S , Yamashiro Y , Watada H . Gut dysbiosis and detection of “live gut bacteria” in blood of Japanese patients with type 2 diabetes . Diabetes Care . 2014 ; 37 ( 8 ): 2343 – 2350 . 45. Nagata S , Chiba Y , Wang C , Yamashiro Y . The effects of the Lactobacillus casei strain on obesity in children: a pilot study . Benef Microbes . 2017 ; 8 ( 4 ): 535 – 543 . 46. Lee JC , Lee HY , Kim TK , Kim MS , Park YM , Kim J , Park K , Kweon MN , Kim SH , Bae JW , Hur KY , Lee MS . Obesogenic diet-induced gut barrier dysfunction and pathobiont expansion aggravate experimental colitis . PLoS One . 2017 ; 12 ( 11 ): e0187515 . 47. Lin H , An Y , Hao F , Wang Y , Tang H . Correlations of fecal metabonomic and microbiomic changes induced by high-fat diet in the pre-obesity state . Sci Rep . 2016 ; 6 ( 1 ): 21618 . 48. Liu R , Zhang C , Shi Y , Zhang F , Li L , Wang X , Ling Y , Fu H , Dong W , Shen J , Reeves A , Greenberg AS , Zhao L , Peng Y , Ding X . Dysbiosis of gut microbiota associated with clinical parameters in polycystic ovary syndrome . Front Microbiol . 2017 ; 8 : 324 . 49. Kelley ST , Skarra DV , Rivera AJ , Thackray VG . The gut microbiome is altered in a letrozole-induced mouse model of polycystic ovary syndrome . PLoS One . 2016 ; 11 ( 1 ): e0146509 . 50. Guo Y , Qi Y , Yang X , Zhao L , Wen S , Liu Y , Tang L . Association between polycystic ovary syndrome and gut microbiota . PLoS One . 2016 ; 11 ( 4 ): e0153196 . Copyright © 2018 Endocrine Society

Journal

Journal of Clinical Endocrinology and MetabolismOxford University Press

Published: Apr 20, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off