Table 1.

Summary of major microbiome studies involving PCOS

Study Author/year	Country	diagnostic criteria	Sample groups	Sequencing	diversity change	Microbial composition change
Kelley 2016 [50]	USA	NA	Letrozole-induced PCOS mouse model/placebo	16S rDNA sequencing V4	↓α diversity β diversity: PCoA of the UniFrac distances: a significant shift	↑Lachnospiraceae, Erysipelotrichaceae, Ruminococcacea ↑Allobaculum, Blautia, and Ruminococcus ↓S24–7 and 1 in the family Rikenellaceae, genus Alistipes
Guo Y 2016 [7]	China	NA	Letrozole-treated rats (n = 24)/control (n = 8)	16S rDNA sequencing V3	NA	↓Lactobacillus, Ruminococcus, and Clostridium ↑Prevotella
Lindheim 2017 [18]	Austria	Rotterdam Criteria	PCOS (n = 24)/control (n = 19)	16S rDNA sequencing V1-V2	↓α diversity β diversity: Unweighted UniFrac analysis: significant clustering	↓Phylum Tenericutes, the order ML615J-28 (belonging to the phylum Tenericutes) ↓Family S24–7 (belonging to the phylum Bacteroidetes)
Liu 2017 [18]	China	Rotterdam Criteria	PCOS (non obese PCOS,n = 12; obese PCOS,n = 21)/controls (n = 9 non obese control;n = 6 obese control)	16S rDNA sequencing V3–4	↓α diversity β diversity: CO group was more similar with that of PN and PO groups, rather than CN group	↑LPS-producing bacteria ↓Spore forming species ↓Akkermensia (of the phylum Verrucomicrobia) and Alistipes, Corprococcus, Ruminococcus
Torres 2018 [20]	Poland	Rotterdam Criteria	Women with PCOM (n = 42)/control healthy women (n = 48)/women with PCOS (n = 73)	16S rDNA sequencing V4	↓α diversity in women with PCOS compared with healthy women β diversity was correlated with hyperandrogenism	↑Porphyromonas spp., Bacteroides coprophilus, Blautia spp., Faecalibacterium prausnitzii in women with PCOS compared with healthy women
Insenser 2018 [21]	Spain	Rotterdam Criteria	PCOS (n = 15)/16 non hyperandrogenic control women (n = 16) and control men (n = 15)	16S rDNA sequencing V4	↓α diversity in women compared with men ↓β diversity in obese patients with PCOS compared with their lean counterparts	↑Catenibacterium, Kandleria genera
Zeng 2018 [22]	China	Rotterdam Criteria	IR-PCOS (n = 9, PCOS with insulin resistance), NIR-PCOS (n = 8, PCOS alone), and healthy controls (n = 8)	16S rDNA sequencing V3–4	↓α diversity in IR-PCOS compared to healthy controls β diversity(weighted UniFrac and PCoA analysis):samples from the HC group can be clearly discriminated from the IR-PCOS group	↑Prevotellaceae in IR-PCOS and NIR-PCOS compared to healthy controls ↑Bacteroides in IR-PCOS and NIR-PCOS compared to healthy controls
Sheman 2018 [84]	USA	NA	Model of prenatal androgen exposure rats/maternal hyperandrogenemia by single-injection of testosterone cypionate or sesame oil vehicle	16S rDNA sequencing	↑α diversity in PNA animals β diversity: the PNA animal samples were further apart.	↑Nocardiaceae and Clostridiaceae ↓Akkermansia, Bacteroides, Lactobacillus, and Clostridium
Zhang 2019 [52]	China	NA	PCOS (n = 38)/control (n = 26)	16S rDNA sequencing V3-V4, deep shotgun metagenomic sequencing	β diversity: weighted and unweighted UniFrac analysis: significant clustering	↓Faecalibacterium, Bifidobacterium, and Blautia ↑Parabacteroides, Bacteroides, Lactobacillus, Oscillibacter, Escherichia/ Shigella, and Clostridium
Chen 2019 [98]	China	Chinese Criteria 2012	non⁃acne PCOS patients(n = 10), MSA⁃PCOS,n = 10)/ healthy controls (n = 10)	16S rDNA sequencing V3-V4	↓α diversity β diversity: weighted UniFrac analysis:samples from the HC group can be clearly discriminated from PCOS	↑Bacteroides ↓Prevotella, Faecalibacteruim
Zhou 2020 [99]	China	Rotterdam Criteria	obese PCOS (OG,n = 30), non-obese PCOS (NG,N = 30) /healthy women (NC,n = 30) healthy but obese women (OC,n = 11)	16S rDNA sequencing V3-V4	α diversity: Simpson index of OTU: NG was significantly higher than that in NC β diversity: No significant difference	↑OG:Lactococcus, Paraprevotella, Alloprevotella ↑NG:Coprococcus_2, Lactobacillus, Prevotella_7
Qi 2019 [53]		NA	PCOS (n = 50)/control (n=43)	Whole-genome shotgun sequencing	α diversity:no significant difference	↑Bacteroides vulgatus