Common cytokine polymorphisms and predisposition to polycystic ovary syndrome: a meta-analysis

Abstract

The results of studies on the relationship between cytokine polymorphisms and polycystic ovary syndrome (PCOS) have been controversial. This meta-analysis was thus designed to more precisely assess the relationship between TNF-α/IL-1/IL-6/IL-10 polymorphisms and PCOS by pooling the results of published studies. A search of PubMed, Embase, Web of Science, and CNKI databases turned up 23 studies that were pooled and analyzed in this meta-analysis. The overall results showed that the distributions of TNF-α –238 G/A, TNF-α –857 C/T, and IL-1B –51 C/T polymorphisms among patients and controls differed significantly. Additionally, the distributions of TNF-α –308 G/A and IL-1B –51 C/T polymorphisms among patients and controls from Asian populations differed significantly, whereas the distributions of IL-6 –174 G/C and IL-1A –889 C/T polymorphisms among patients and controls from Caucasian populations also differed significantly. In conclusion, our meta-analysis demonstrated that TNF-α –238 G/A, TNF-α –857 C/T, and IL-1B –51 C/T polymorphisms might influence susceptibility to PCOS in the overall pooled population. Moreover, TNF-α –308 G/A and IL-1B –51 C/T polymorphisms might influence susceptibility to PCOS in Asians, whereas IL-6 –174 G/C and IL-1A –889 C/T polymorphisms might influence susceptibility to PCOS in Caucasians.

POLYCYSTIC OVARY SYNDROME (PCOS), characterized by polycystic ovaries, chronic anovulation, and hyperandrogenism, is the most common reproductive endocrine disorder in women of childbearing age [1, 2]. Although its definite etiologies and pathogenesis mechanisms have yet to be clarified, accumulating evidence suggests that genetic architecture greatly influences its development. Firstly, the incidence of PCOS in different populations differs significantly [3, 4], indicating that genetic background is probably one of the key factors underlying this phenomenon. Secondly, previous association studies have also detected numerous predisposing gene loci of PCOS in different populations [5, 6]. However, the etiologies and pathogenesis mechanisms of PCOS are extremely, and the precise genetic factors that contribute to the development of PCOS still need to be intensively explored.

Cytokines play vital roles in modulating immune responses and are involved in the pathogenesis of various inflammatory disorders [7, 8]. Previous studies have demonstrated that PCOS share similar properties with many chronic inflammatory disorders, and classical inflammatory mediators such as C-reactive protein (CRP), tumor necrosis factor-α (TNF-α) and interleukin-6 (IL-6) have also been found to be elevated in patients with PCOS [9-11]. Moreover, pro-inflammatory cytokines and its associated over-activated immune responses have also been shown to be associated with higher androgen levels, and have been shown to be able to impact ovarian function and jeopardize the processes of ovulation in PCOS [12, 13]. Therefore, if a polymorphism can impact gene expression or the protein structure of cytokines, it is highly likely that it might alter inflammation status and influence predisposition to PCOS.

In the last two decades, investigators around the world have extensively explored the relationship between cytokine polymorphisms and PCOS, particularly for polymorphisms of TNF-α, interleukin-1 (IL-1), interleukin-6 (IL-6), and interleukin-10 (IL-10), yet the relationship between TNF-α/IL-1/IL-6/IL-10 polymorphisms and PCOS still remains controversial and ambiguous. Thus, we designed this meta-analysis to obtain a more statistically reliable conclusion regarding the relationship between TNF-α/IL-1/IL-6/IL-10 polymorphisms and PCOS by pooling the results of already published studies.

Materials and Methods

The PRISMA guideline was followed by the authors when conducting this meta-analysis [14].

Literature search and inclusion criteria

Literature search of PubMed, Web of Science, Embase, and CNKI was conducted using the following terms: (Tumor necrosis factor-α or TNF-α or interleukin-10 or IL-10 or interleukin 10 or IL 10 or interleukin-6 or IL-6 or interleukin 6 or IL 6 or interleukin-1 or IL-1 or interleukin 1 or IL 1) and (polymorphism or variant or variation or mutation or SNP or genome-wide association study or genetic association study or genotype or allele) and (polycystic ovary syndrome or PCOS). The references of the articles obtained were also checked for additional related studies.

Eligible studies had to meet all three inclusion criteria: 1. formally published study evaluating relationship between TNF-α/IL-1/IL-6/IL-10 polymorphisms and PCOS; 2. provided genotypic distribution of TNF-α/IL-1/IL-6/IL-10 polymorphisms in patients with PCOS and controls; 3. the entire manuscript was available in English or Chinese. Articles were excluded when any one of the following three conditions was fulfilled: 1. studies were not concerning TNF-α/IL-1/IL-6/IL-10 polymorphisms and PCOS; 2. studies were reviews or expert comments; 3. case series only involved participants with PCOS. When duplicate reports were uncovered during literature search, only the most recent one was included for pooled meta-analysis.

Data extraction and quality assessment

The following five items were extracted from the studies included: 1. surname of first author; 2. year of online publication; 3. country and ethnicity of involved participants; 4. number of patients and controls; 5. genotypic distributions of TNF-α/IL-1/IL-6/IL-10 polymorphisms in patients and control subjects. The p values of Hardy-Weinberg equilibrium (HWE) were also calculated based on genotypic distributions of TNF-α/IL-1/IL-6/IL-10 polymorphisms.

Newcastle-Ottawa scale (NOS) was used to assess the quality of the studies included [15]. Its score range is from zero to nine, and the methodology quality of an article is considered to be good if it can get a score of greater than seven.

Data extraction and quality assessment of the studies included were independently performed by two authors. We wrote to the corresponding authors of eligible studies for additional data if we failed to extract the necessary information from the studies included.

Statistical analyses

Review Manager software was used to pool mete-analysis results. Z test was used to evaluate the relationship between TNF-α/IL-1/IL-6/IL-10 polymorphisms and predisposition to PCOS with the statistical significant threshold at 0.05. I² statistics was used to estimate heterogeneity. DerSimonian-Laird method was used to pool the results if I² is was larger than 50%. Otherwise, Mantel-Haenszel method was used to pool the results. The authors also conducted subgroup analyses by ethnicity. Stabilities of pooled results were examined by omitting one study each time and pooling the results of the other studies. Publication biases were examined by using funnel plots.

Results

Characteristics of included studies

Of the 173 articles were retrieved by the literature search, 44 were assessed for eligibility after omitting unrelated and repeated reports. Twelve reviews and nine studies with incomplete data were further excluded for a total of 23 studies finally pooled in our meta-analysis (Fig. 1). Extracted data of eligible studies are summarized in Table 1. Rotterdam criterion was used by most of the eligible studies for diagnosing PCOS, whereas Mao et al., Milner et al., Korhonen et al., Walch et al. and Kolbus et al. used NIH criterion, and Thathapudi et al. used 2006 AE-PCOS criterion.

Fig. 1

Flowchart of study selection for this meta-analysis

Table 1 The characteristics of included studies in this meta-analysis

First author, year	Country	Ethnicity	Diagnostic criteria	Sample size	Genotype distribution		p-value for HWE	NOS score
					Cases	Controls
TNF-α –238 G/A rs361525					GG/GA/AA
Bhatnager 2019	India	Mixed	2003 Rotterdam criterion	200/200	148/50/2	117/83/0	<0.001	8
Kordestani 2018	Iran	Mixed	2003 Rotterdam criterion	111/105	104/6/1	96/6/3	<0.001	8
Wen 2013	China	Asian	2003 Rotterdam criterion	144/72	137/7/0	68/4/0	0.808	8
Xie 2016	China	Asian	2003 Rotterdam criterion	102/96	99/3/0	93/3/0	0.876	8
TNF-α –308 G/A rs1800629					GG/GA/AA
Bhatnager 2019	India	Mixed	2003 Rotterdam criterion	200/200	121/79/0	135/63/2	0.067	8
Deepika 2013	India	Mixed	2003 Rotterdam criterion	283/306	270/10/3	293/10/3	<0.001	8
Mao 2000	China	Asian	1990 NIH criterion	118/54	88/29/1	37/13/4	0.089	7
Milner 1999	Australia	Caucasian	1990 NIH criterion	84/108	59/23/2	63/42/3	0.194	7
Peng 2010	China	Asian	2003 Rotterdam criterion	130/175	118/11/1	146/27/2	0.557	8
Vural 2010	Turkey	Caucasian	2003 Rotterdam criterion	97/95	78/16/3	77/15/3	0.055	8
Wen 2013	China	Asian	2003 Rotterdam criterion	103/59	89/14/0	52/7/0	0.628	8
TNF-α –857 C/T rs1799724					CC/CT/TT
Korhonen 2002	Finland	Caucasian	1990 NIH criterion	87/115	74/11/2	96/16/3	0.036	8
Bhatnager 2019	India	Mixed	2003 Rotterdam criterion	200/200	120/80/0	136/64/0	0.007	8
Thathapudi 2014	India	Mixed	2006 AE-PCOS criterion	204/204	178/26/0	186/18/0	0.510	8
Xie 2016	China	Asian	2003 Rotterdam criterion	102/96	32/62/8	49/43/4	0.147	8
TNF-α –1,031 T/C rs1799964					TT/TC/CC
Bhatnager 2019	India	Mixed	2003 Rotterdam criterion	200/200	152/48/0	129/69/2	0.027	8
Deepika 2013	India	Mixed	2003 Rotterdam criterion	283/306	107/170/6	162/139/5	<0.001	8
Yun 2011	China	Asian	2003 Rotterdam criterion	217/144	144/71/2	122/22/0	0.321	7
IL-6 –174 G/C rs1800795					GG/GC/CC
Erdogan 2008	Turkey	Caucasian	2003 Rotterdam criterion	153/234	93/51/9	63/148/23	<0.001	7
Eser 2017	Turkey	Caucasian	2003 Rotterdam criterion	76/27	32/39/5	11/13/3	0.772	7
Tumu 2013	India	Mixed	2003 Rotterdam criterion	104/156	69/31/4	77/73/6	0.024	7
Vural 2010	Turkey	Caucasian	2003 Rotterdam criterion	97/95	59/34/4	46/42/7	0.536	8
Walch 2004	Australia	Caucasian	1990 NIH criterion	62/84	24/30/8	43/25/16	0.002	8
IL-1A –889 C/T rs1800587					CC/CT/TT
Eser 2017	Turkey	Caucasian	2003 Rotterdam criterion	76/27	33/32/11	16/8/3	0.235	7
Kolbus 2007	Australia	Caucasian	1990 NIH criterion	105/102	42/56/7	55/41/6	0.649	7
Wang 2009	China	Asian	2003 Rotterdam criterion	205/177	171/32/2	137/40/0	0.090	8
IL-1B –51 C/T rs16944					CC/CT/TT
Kolbus 2007	Australia	Caucasian	1990 NIH criterion	105/102	43/47/15	40/48/14	0.947	7
Mu 2010	China	Asian	2003 Rotterdam criterion	200/177	60/76/64	64/87/26	0.684	8
Xia 2013	China	Asian	2003 Rotterdam criterion	59/56	13/21/25	18/31/7	0.257	7
Yang 2010	China	Asian	2003 Rotterdam criterion	118/86	30/56/32	34/26/26	<0.001	8
IL-1B +3,953 C/T rs1143634					CC/CT/TT
Kolbus 2007	Australia	Caucasian	1990 NIH criterion	105/102	48/46/11	60/37/5	0.817	7
Yang 2010	China	Asian	2003 Rotterdam criterion	118/86	114/4/0	74/12/0	0.487	8
IL-10 –1,082 A/G rs1800896					AA/AG/GG
Karadeniz 2008	Turkey	Caucasian	2003 Rotterdam criterion	91/75	44/37/10	31/34/10	0.888	7
Talaat 2016	Egypt	Mixed	2003 Rotterdam criterion	61/80	28/23/10	28/41/11	0.511	8
Vural 2010	Turkey	Caucasian	2003 Rotterdam criterion	97/95	25/57/15	25/54/16	0.152	7
IL-10 –819 C/T rs1800871					CC/CT/TT
Karadeniz 2008	Turkey	Caucasian	2003 Rotterdam criterion	91/75	36/45/10	36/31/8	0.733	7
Talaat 2016	Egypt	Mixed	2003 Rotterdam criterion	61/80	32/18/11	43/33/4	0.462	8

Abbreviations: HWE, Hardy-Weinberg equilibrium; NOS, Newcastle-Ottawa scale; NA, Not available.

Meta-analyses of TNF-α polymorphisms and PCOS

Twelve studies were eligible for estimation of relationship between TNF-α polymorphisms and PCOS. TNF-α –238 G/A (dominant comparison: OR = 1.78, p = 0.002; over-dominant comparison: OR = 0.55, p = 0.002; allele comparison: OR = 1.60, p = 0.005) and –857 C/T (dominant comparison: OR = 0.67, p = 0.004; over-dominant comparison: OR = 1.46, p = 0.007; allele comparison: OR = 0.72, p = 0.007) polymorphisms were found to be significantly associated with PCOS in overall pooled population, and –308 G/A polymorphism was also found to be significantly associated with PCOS in Asians (allele comparison: OR = 1.50, p = 0.05). The pooled meta-analyses did not reveal any significant associations for TNF-α –1,031 T/C polymorphism and PCOS (see Table 2).

Table 2 Pooled meta-analyses results of the current study

Polymorphisms	Population	Sample size	Dominant comparison		Recessive comparison		Over-dominant comparison		Allele comparison
			p value	OR (95%CI)	p value	OR (95%CI)	p value	OR (95%CI)	p value	OR (95%CI)
TNF-α –238 G/A	Overall	557/473	0.002	1.78 (1.24–2.56)	0.97	0.97 (0.22–4.27)	0.002	0.55 (0.38–0.80)	0.005	1.60 (1.15–2.21)
	Asian	246/168	0.83	1.12 (0.41–3.03)	NA	NA	0.83	0.89 (0.33–2.43)	0.83	1.11 (0.42–2.99)
TNF-α –308 G/A	Overall	1,015/997	0.55	1.07 (0.85–1.36)	0.16	0.58 (0.27–1.23)	0.89	0.98 (0.77–1.26)	0.35	1.11 (0.90–1.37)
	Asian	351/288	0.11	1.42 (0.92–2.21)	0.08	0.24 (0.05–1.17)	0.31	0.79 (0.51–1.25)	0.05	1.50 (1.01–2.24)
	Caucasian	181/203	0.22	1.34 (0.84–2.11)	0.89	0.92 (0.27–3.08)	0.22	0.74 (0.46–1.19)	0.26	1.26 (0.84–1.88)
TNF-α –857 C/T	Overall	593/615	0.004	0.67 (0.51–0.88)	0.41	1.53 (0.56–4.14)	0.007	1.46 (1.11–1.93)	0.007	0.72 (0.57–0.91)
TNF-α –1,031 T/C	Overall	700/650	0.42	0.70 (0.29–1.68)	0.84	1.11 (0.42–2.91)	0.41	1.42 (0.62–3.25)	0.47	0.77 (0.38–1.57)
IL-6 –174 G/C	Overall	492/596	0.16	1.61 (0.83–3.15)	0.05	0.63 (0.39–1.01)	0.36	0.72 (0.36–1.45)	0.40	1.25 (0.75–2.08)
	Caucasian	388/440	0.38	1.50 (0.61–3.69)	0.04	0.58 (0.35–0.97)	0.66	0.81 (0.32–2.05)	0.68	1.15 (0.59–2.25)
IL-1A –889 C/T	Overall	386/306	0.52	0.79 (0.39–1.61)	0.43	1.39 (0.61–3.17)	0.66	1.18 (0.57–2.44)	0.52	0.86 (0.54–1.36)
	Caucasian	181/129	0.01	0.56 (0.35–0.89)	0.64	1.23 (0.52–2.91)	0.03	1.71 (1.06–2.75)	0.04	0.68 (0.48–0.98)
IL-1B –51 C/T	Overall	482/421	0.04	0.74 (0.56–0.98)	0.12	1.83 (0.85–3.93)	0.64	0.87 (0.48–1.58)	0.03	0.70 (0.52–0.96)
	Asian	377/319	0.01	0.66 (0.48–0.90)	0.11	2.20 (0.84–5.77)	0.71	0.85 (0.36–2.01)	0.001	0.63 (0.47–0.83)
IL-1B +3,953 C/T	Overall	223/188	0.67	1.54 (0.20–11.65)	0.14	2.27 (0.76–6.78)	0.56	0.59 (0.10–3.57)	0.66	1.53 (0.23–10.24)
IL-10 –1,082 A/G	Overall	249/250	0.23	1.26 (0.87–1.83)	0.86	0.95 (0.58–1.58)	0.30	0.83 (0.58–1.18)	0.35	1.13 (0.87–1.46)
	Caucasian	188/170	0.55	1.14 (0.73–1.79)	0.62	0.86 (0.48–1.56)	0.83	0.96 (0.63–1.45)	0.51	1.11 (0.82–1.50)
IL-10 –819 C/T	Overall	151/155	0.37	0.81 (0.52–1.28)	0.32	1.99 (0.51–7.82)	0.86	0.93 (0.41–2.12)	0.13	0.77 (0.55–1.08)

Abbreviations: OR, Odds ratio; CI, Confidence interval; NA, Not available.

Meta-analyses of IL-6 polymorphisms and PCOS

Five studies were eligible for estimation of relationship between IL-6 polymorphisms and PCOS. IL-6 –174 G/C polymorphism was found to be significantly associated with PCOS in Caucasians (recessive comparison: OR = 0.58, p = 0.04). Nevertheless, we did not observe any other significant associations for IL-6 –174 G/C polymorphism in pooled meta-analyses (see Table 2).

Meta-analyses of IL-1 polymorphisms and PCOS

Six studies were eligible for estimation of relationship between IL-1 polymorphisms and PCOS. IL-1B –51 C/T polymorphism was found to be significantly associated with PCOS in overall pooled population (dominant comparison: OR = 0.74, p = 0.04; allele comparison: OR = 0.70, p = 0.03) and Asians (dominant comparison: OR = 0.66, p = 0.01; allele comparison: OR = 0.63, p = 0.001). Moreover, IL-1A –889 C/T polymorphism was found to be significantly associated with PCOS in Caucasians (dominant comparison: OR = 0.56, p = 0.01; over-dominant comparison: OR = 1.71, p = 0.03; allele comparison: OR = 0.68, p = 0.04). The pooled meta-analyses did not reveal any significant associations for IL-1B +3,953 C/T polymorphism and PCOS (see Table 2).

Meta-analyses of IL-10 polymorphisms and PCOS

Three studies were eligible for estimation of relationship between IL-10 polymorphisms and PCOS. The pooled meta-analyses did not reveal any significant associations for IL-10 polymorphisms and PCOS (see Table 2).

Sensitivity analyses

Stabilities of pooled meta-analyses results were examined by omitting one study each time and pooling the results of the other studies. The trends of associations remained unchanged in sensitivity analyses, indicating that our pooled meta-analyses results were statistically stable.

Publication biases

Publication biases were examined by funnel plots. Funnel plots were overall symmetrical, suggesting that our pooled meta-analyses results were not likely to be severely influenced by publication biases.

Discussion

The meta-analysis results demonstrated that TNF-α –238 G/A, TNF-α –857 C/T, and IL-1B –51 C/T polymorphisms might influence susceptibility to PCOS in overall pooled population. Moreover, we found that TNF-α –308 G/A and IL-1B –51 C/T polymorphisms might influence susceptibility to PCOS in Asians, whereas IL-6 –174 G/C and IL-1A –889 C/T polymorphisms might influence susceptibility to PCOS in Caucasians. The trends of associations remained unchanged in sensitivity analyses, suggesting that our pooled meta-analysis results were quite stable statistically.

A few points need to be considered when interpreting our findings. First, previous experimental studies have demonstrated that all investigated polymorphisms are correlated with altered gene expression or protein structure of corresponding cytokines [16, 17]. Thus, it is likely that these variations might influence normal functioning of TNF-α/IL-1/IL-6/IL-10, lead to immune dysfunction, and influence predisposition to PCOS. In this meta-analysis, we did not observe any positive findings for IL-10 polymorphisms. Nevertheless, since only three studies were pooled analyzed, it is possible that our pooled meta-analysis is still not statistically sufficient to detect the real associations between IL-10 polymorphisms and PCOS, and future studies involving larger populations are still needed to get a more statistically robust finding. Second, the etiologies and pathogenesis mechanisms of PCOS are extremely sophisticated, although many literatures support the finding that inflammation and immune-regulatory genes are involved in the pathogenesis of PCOS, we also need to point out that several reports failed to find any significant association between inflammatory gene pathways and PCOS, such as Bhatt et al. [18]. Considering that PCOS is a multi-factorial disease, further association studies also need to be conducted to consider the potential influence of gene-gene interactions to more precisely measure the effects of certain gene polymorphisms on predisposition to PCOS [19]. Third, we also intended to analyze gene polymorphisms of other cytokines such as IL-4, IL-8, and IL-12 at the beginning. However, since we could not find at least two eligible studies for gene polymorphisms in these cytokines, we had to focus only on TNF-α, IL-1, IL-6 and IL-10 in this meta-analysis. Fourth, we did not set any restrictions on the investigated polymorphisms, nevertheless we noticed that previous investigators even failed to analyze some very common polymorphisms of IL-6 and IL-10 such as IL-6 –572 G/C, IL-6 –634 G/C, and IL-10 –592 C/A. In the future, we hope to analyze the associations between these IL-6 or IL-10 polymorphisms and PCOS. Fifth, since different ethnicities were involved in our overall pooled analysis, the distributions of cytokine polymorphisms varied from study to study, especially when those studies were of different ethnic groups. This suggests that the effects of cytokine polymorphisms on predisposition to PCOS may vary from population to population, and that the positive findings in one ethnic subgroup should not be generalized to broader populations.

Like all meta-analyses, a few limitations of our pooled meta-analyses should also be acknowledged. Firstly, our pooled meta-analysis results were derived from pooling unadjusted findings of eligible studies since the authors did not have the raw data from them [20]. Secondly, environmental factors might also influence the relationship between TNF-α/IL-1/IL-6/IL-10 polymorphisms and PCOS. However, most investigators only focused on genetic associations in their works, and so the genetic-environmental interactions were not explored in this meta-analysis [21]. Thirdly, we did not consider grey literature. Therefore, despite the fact that funnel plots were overall symmetrical, publication biases still might affect the robustness of our pooled results [22].

In conclusion, this meta-analysis demonstrated that TNF-α –238 G/A, TNF-α –857 C/T, and IL-1B –51 C/T polymorphisms might influence susceptibility to PCOS in overall pooled population. Moreover, we found that TNF-α –308 G/A and IL-1B –51 C/T polymorphisms might influence susceptibility to PCOS in Asians, whereas IL-6 –174 G/C and IL-1A –889 C/T polymorphisms might influence susceptibility to PCOS in Caucasians. These results also suggested that TNF-α, IL-1, and IL-6 might be involved in the development of PCOS and that they might serve as potential therapeutic targets for PCOS.

Authors’ Contributions

Yunli Zhang and Jiaoyan He designed this meta-analysis. Yunli Zhang and Lina Che searched the literature databases. Mingyan Zhang analyzed the data. Yunli Zhang and Jiaoyan He wrote the manuscript. All authors approved the final manuscript as submitted.

Acknowledgments

The authors would like to thank Dr. Hongfang Qi from Changyi People’s Hospital for her help in deciding the final key words for the literature search.

Funding

None.

Conflict of Interest

None.

Ethical Statement

Ethical approval and informed consent are not applicable to meta-analyses.

Data Availability Statement

Data sharing is not applicable to meta-analyses.

References

• 1   Ehrmann  DA (2015) Polycystic ovary syndrome. N Engl J Med 352: 1223–1236.
• 2   Pasquali  R,  Gambineri  A (2006) Polycystic ovary syndrome: a multifaceted disease from adolescence to adult age. Ann NY Acad Sci 1092: 158–174.
• 3   Kim  JJ,  Choi  YM (2019) Phenotype and genotype of polycystic ovary syndrome in Asia: ethnic differences. J Obstet Gynaecol Res 45: 2330–2337.
• 4   Ding  T,  Hardiman  PJ,  Petersen  I,  Wang  FF,  Qu  F, et al. (2017) The prevalence of polycystic ovary syndrome in reproductive-aged women of different ethnicity: a systematic review and meta-analysis. Oncotarget 8: 96351–96358.
• 5   Urbanek  M (2007) The genetics of the polycystic ovary syndrome. Nat Clin Pract Endocrinol Metab 3: 103–111.
• 6   Luque-Ramirez  M,  San Millan  JL,  Escobar-Morreale  HF (2006) Genomic variants in polycystic ovary syndrome. Clin Chim Acta 366: 14–26.
• 7   Dinarello  CA (1997) Role of pro- and anti-inflammatory cytokines during inflammation: experimental and clinical findings. J Biol Regul Homeost Agents 11: 91–103.
• 8   Lin  WW,  Karin  M (2007) A cytokine-mediated link between innate immunity, inflammation, and cancer. J Clin Invest 117: 1175–1183.
• 9   Ojeda-Ojeda  M,  Murri  M,  Insenser  M,  Escobar-Morreale  HF (2013) Mediators of low-grade chronic inflammation in polycystic ovary syndrome (PCOS). Curr Pharm Des 19: 5775–5791.
• 10   Deligeoroglou  E,  Vrachnis  N,  Athanasopoulos  N,  Iliodromiti  Z,  Sifakis  S, et al. (2012) Mediators of chronic inflammation in polycystic ovarian syndrome. Gynecol Endocrinol 28: 974–978.
• 11   Bhatnager  R,  Jalthuria  J,  Sehrawat  R,  Nanda  S,  Dang  AS (2019) Evaluating the association of TNF α promoter haplotype with its serum levels and the risk of PCOS: a case control study. Cytokine 114: 86–91.
• 12   González  F (2012) Inflammation in polycystic ovary syndrome: underpinning of insulin resistance and ovarian dysfunction. Steroids 77: 300–305.
• 13   Boots  CE,  Jungheim  ES (2015) Inflammation and human ovarian follicular dynamics. Semin Reprod Med 33: 270–275.
• 14   Moher  D,  Liberati  A,  Tetzlaff  J,  Altman  DG, PRISMA group (2009) Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med 151: 264–269.
• 15   Stang  A (2010) Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol 25: 603–605.
• 16   Smith  AJ,  Humphries  SE (2009) Cytokine and cytokine receptor gene polymorphisms and their functionality. Cytokine Growth Factor Rev 20: 43–59.
• 17   Kazemi  S,  Saidijam  M,  Hashemi  SH,  Karami  M,  Vaisi-Raygani  A, et al. (2016) Analysis of IL-10 and IL-6 gene polymorphisms and their serum levels in patients with brucellosis: a case control study. Immunol Invest 45: 107–115.
• 18   Bhatt  S,  Mutharasan  P,  Garcia  OA,  Jafari  N,  Legro  RS, et al. (2014) The inflammatory gene pathway is not a major contributor to polycystic ovary snydrome. J Clin Endocrinol Metab 99: E567–E571.
• 19   Nishi  A,  Milner DA  Jr,  Giovannucci  EL,  Nishihara  R,  Tan  AS, et al. (2016) Integration of molecular pathology, epidemiology and social science for global precision medicine. Expert Rev Mol Diagn 16: 11–23.
• 20   Park  JH,  Li  L,  Choi  JW,  Baek  KH (2016) The Association of –429T>C and –374T>A Polymorphisms in the RAGE Gene with Polycystic Ovary Syndrome. Int J Med Sci 13: 451–456.
• 21   Xue  H,  Zhao  H,  Liu  X,  Zhao  YR,  Chen  ZJ, et al. (2016) Association of single-nucleotide polymorphisms rs2197076 and rs2241883 of FABP1 gene with polycystic ovary syndrome. J Assist Reprod Genet 33: 75–83.
• 22   Liu  Z,  Wang  Z,  Hao  C,  Tian  Y,  Fu  J (2018) Effects of ADIPOQ polymorphisms on PCOS risk: a meta-analysis. Reprod Biol Endocrinol 16: 120.