Polycystic ovary syndrome in women is associated with longer anogenital distance,a new potential biomaker for PCOS

Abstract

Background

Fetal androgen exposure plays a pivotal role in Polycystic ovary syndrome (PCOS) development and may result in elevated Anogenital distance (AGD). This meta-analysis aimed to investigate the clinical link between PCOS and AGD.

Methods

A literature search was performed across various databases to identify studies evaluating AGD in adults with PCOS and without, regardless of language, up to December 2024. The quality of the studies was evaluated using the Newcastle-Ottawa Scale (NOS) scoring system. Random-effects models were utilized to determine mean differences (MDs) and 95% confidence intervals (CIs) in cases of high heterogeneity. This meta-analysis encompassed 4 studies involving a total of 837 participants.

Results

The pooled analysis found a noteworthy increase of the AGD-ac and AGD-af in PCOS patient compared with the control groups, with an overall MD of AGD-ac = 5.23, 95% CI (2.60, 7.85), P-value < 0.0001, I²= 57%, and with an overall MD of AGD-af = 2.19, 95% CI (0.04, 4.35), P-value = 0.05, I²= 89%.

Conclusion

The meta-analysis results indicated that women diagnosed with PCOS exhibit elongated AGD. This potential association between AGD and PCOS could serve as a novel clinical marker for the diagnosis of PCOS. Fetal androgen exposure may play a role in the pathogenesis of PCOS.

Keywords

Anogenital distance polycystic ovary syndrome androgens prenatal exposures

1 Introduction

Polycystic ovary syndrome (PCOS) is one of the most common endocrine and metabolic disease in women, with 4%-12% incidences worldwide. It has complex etiology and high clinical heterogeneity, like hyperandrogenism, anovulation, infertility and increased risk of metabolic diseases besides psychosocial dysfunction.^1–3 Rotterdam criteria is the most common used one which still exist some controversy.⁴ Since the classical features of PCOS, including hirsutism, acne and androgenetic alopecia, polycystic ovarian morphology and biochemical hyperandrogenism, lack some reliable criteria to evaluate and varies significantly with race and obesity.^5,6 Therefore, researchers still try to find more reliable biomarkers to diagnose or reflect PCOS.The etiology of PCOS is unclear, and emerging data suggest that PCOS may originate from intrauterine life and prenatal exposure to androgen hormones is considered an important factor of PCOS.^7,8 However, accurate diagnosis is critical, particularly in distinguishing PCOS from conditions like congenital adrenal hyperplasia (CAH) and non-classical CAH, which can significantly overlap in symptoms. Future studies should ensure the exclusion of CAH to validate the link between AGD and PCOS more effectively. During the early embryonic period, tissues and organs of the body undergo a sensitive developmental phase due to rapid cell differentiation. When the body experiences stimuli and injuries during this sensitive phase, such as abnormal hyperandrogenic stimulation in utero, lifelong effects can occur. Animal experiments demonstrated that the excess androgen exposure during prenatal period may cause PCOS-like Characteristics for their offspring, like prolonged or irregular estrous cycles, fewer corpus luteum and preovulatory follicles and more preantral follicles and antral follicles than the controls.^9–11

Anogenital distance (AGD) has been suggested to represent a phenotypic signature reflecting in utero androgen action.¹² In rodents, the anogenital distance (AGD) is a sensitive biomarker of androgen exposure during a critical embryonic window of testis development. In humans, the measurement method of AGD is: women were asked to lay down in the lithotomy position with their thighs at 45° to the examination table. A digital caliper was used to measure AGD-ac and AGD-af. The AGD-ac was measured from the anterior clitoral surface to the anus, and AGD-af was measured from the posterior fourchette to the anus (Fig. 1). Several epidemiological studies have shown alterations in AGD associated with prenatal exposure to several chemicals with potential endocrine disrupting activity.^12,13 Studies have shown that AGD is positively correlated with male androgen levels, and shorter AGD may be related to testicular dysfunction.¹⁴ In women, longer adult AGD is associated with higher testosterone levels and larger ovarian follicle numbers, whereas shorter AGD is associated with endometriosis, a disease that may be induced by prenatal exposure to high estrogen and low testosterone in the intrauterine environment,¹⁵ and AGD is positively correlated with serum testosterone levels.¹⁶ The offspring of mice with PCOS induced by intrauterine androgen treatment during pregnancy show increased AGD in adulthood.⁹ Four studies found that adult PCOS patients have longer AGD than control PCOS patients.^17–20 PCOS is one of the most prevalent endocrine and metabolic disorders affecting women of childbearing age. The challenge is to accurately diagnose PCOS. Artificial intelligence (AI) technology has recently become a popular approach in the healthcare industry to diagnose and manage multiple complex diseases, such as COPD.^21–23 Artificial intelligence uses algorithms to analyze imaging images, anthropometry and biochemical test result data for deep learning, integrating different data sources to build diagnostic models and provide doctors with more comprehensive diagnostic support. We believe that AGD, as a anthropometric indicator, has the potential to be applied to artificial intelligence, combined with in-depth analysis of patients’ multidimensional data, to assist doctors in early and accurate diagnosis of PCOS patients, because early detection and effective management of PCOS are crucial to improve patients’ quality of life and reduce the occurrence of related health complications.²⁴

Figure 1.

The measurements of anogenital distance. (The AGD-ac was measured from the anterior clitoral surface to the anus, and AGD-af was measured from the posterior fourchette to the anus).

Therefore, the aim of this study was to conduct a meta-analysis to find the association between AGD measurements as a biomarker of prenatal androgen environment and the presence of PCOS.

2 Materials and methods

2.1 Databases

This systematic review was registered prospectively on the International Platform of Registered Systematic Review and Meta-analysis Protocols with the registration number INPLASY202430065. The review follows the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement for reporting.

Studies comparing AGD in women with vs. without PCOS were identified by searching PubMed, MEDLINE, EMBASE, Cochrane Library, CBM, CNKI, WANFANG, and VIP. Databases were searched last in Feb. 2024 for publications, with no time limit specified for trial inclusion and without language restriction since data inception. The search terms search in PubMed were included as follows: (polycystic ovary syndrome [Mesh] OR (Ovary Syndrome, Polycystic[Title/Abstract]) OR (Syndrome, Polycystic Ovary[Title/Abstract]) OR (Syndrome, Stein-Leventhal[Title/Abstract]) OR (Sclerocystic Ovarian Degeneration[Title/Abstract]) OR (Ovarian Degeneration, Sclerocystic[Title/Abstract]) OR (Sclerocystic Ovary Syndrome[Title/Abstract]) OR (Polycystic Ovarian Syndrome[Title/Abstract]) OR (Ovarian Syndrome, Polycystic[Title/Abstract]) OR (Polycystic Ovary Syndrome[Title/Abstract]) OR (Polycystic Ovary Syndrome[Title/Abstract]) OR (Ovary, Sclerocystic[Title/Abstract]) OR (Sclerocystic Ovary[Title/Abstract]) AND ((anogenital distance) OR (ano-genital distance[Title/Abstract]) OR (Anogenital distance[Title/Abstract]) OR (anogenital distanc[Title/Abstract]) OR (ano-genital distanc[Title/Abstract])), which was then modified for each database searched. Only fully published, peer-reviewed papers were included, whereas grey literature was not eligible. No language restrictions were placed on the search.

2.2 Study selection

Cohort, case-control, and cross-sectional observational studies that reported on anogenital distance (AGD) in women with and without polycystic ovary syndrome (PCOS) were included in the review, irrespective of the setting in which the studies were conducted. The trials included reproductive-aged women diagnosed with PCOS based on criteria such as those from the National Institute of Health (1990) and the Rotterdam criteria.

In this meta-analysis, it is imperative to ensure accurate diagnosis of PCOS, distinctly differentiating it from conditions such as congenital adrenal hyperplasia (CAH) and non-classical CAH. Given that the carrier rate of CAH is notably high in certain populations, misdiagnosis could obscure the true prevalence of PCOS and its correlation with anogenital distance (AGD). Therefore, we strictly included studies where clinicians confirmed the exclusion of CAH/non-classical CAH, which is essential for validating the associations drawn between AGD and PCOS. This rigorous approach strengthens the reliability of our findings, suggesting that AGD may serve as a novel biomarker for PCOS, potentially reflecting underlying fetal androgen exposure. The review included studies that used diagnostic criteria such as the ESHRE/ASRM (2003) or the AE-PCOS Criteria (2006) for the diagnosis of polycystic ovary syndrome (PCOS). The studies were required to provide means of anogenital distance (AGD) and their standard deviations (SDs), or have sufficient data available to calculate these parameters. Studies were excluded if they (1) were letters, case reports, editorials, animal experiments, or conference abstracts;(2) Types of participants: males, adolescent females, post-menopausal women, women without PCOS. Selection of eligible studies was independently performed by two authors (Shumin Chen and Yue Wu). In case of doubt, papers were discussed, and consensus was reached. As consensus was reached in all cases, no additional co-author was involved.

2.3 Data extraction

Two authors independently extracted the characteristics of included studies. Discrepancies were noted and resolved by discussing with the third author (Weiwen Fan).

2.4 Quality assessment

The quality of selected studies was assessed using the Newcastle-Ottawa Scale (NOS) scoring system.²⁵ According to the quality score assessment, the total score ranged from 0 to 9. Studies with a score of 7 or above were considered high-quality, and studies with a score of 4 or below were considered low-quality. Studies with a score between 4 and 7 were considered medium quality. Evaluation of evidence quality (high-quality, medium-quality, or low-quality) was determined by 2 review authors (Shumin Chen and Yue Wu) independently, with differences resolved by the discussion (Quilin Cui and Weiwen Fan).

2.5 Data synthesis

Outcomes across each trial were presented as continuous data. Mean differences (MDs) with 95% confidence intervals (CIs) were adopted to calculate the overall estimates and were presented in forest plots. The heterogeneity variance using the Q-statistic and inconsistency index (I²) was calculated. I² values of 0%, 25%, 50% and 75% representing no, low, moderate and high heterogeneity, respectively. According to the Cochrane review guidelines, if severe heterogeneity was present at I²> 50%, the random effect models were chosen, otherwise the fixed effect models were used and the random-effects method for meta-analysis was utilized to combine data, given the potential heterogeneity related to study populations, number of included study and assessment methods. Subgroup analyses was conducted based on geographical region, measure method and history of vaginal delivery. Moreover, sensitivity analyses were conducted where appropriate to determine the robustness of the results. Two-sided P < 0.05 was considered of statistically significance. All analyses were carried out in Review Manager software (Version 5.3; Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014) and STATA software (Version 13.0; Stata Corporation, College Station, TX).

3 Results

3.1 Study selection

Initially, a total of 136 citations were identified through electronic database searches as mentioned previously. After removing duplicate entries, 57 unique studies were left for further evaluation. From these, 30 studies were excluded based on screening of their titles and abstracts. The remaining 27 articles underwent full-text screening, leading to the exclusion of 21 articles due to reasons such as being animal research (1), letters (4), reviews (3), or not aligning with the theme of the review (13). After the exclusions, 6 studies were found to meet the inclusion criteria; however, 2 of these studies were extensions of one of the included studies, resulting in a final inclusion of 5 unique studies for the review. Finally, 4 case-control studies^17–19,26 fit all inclusion and exclusion criteria and were used for data extraction in our meta-study. The selection process was documented with a flowchart. (Fig. 2)

Figure 2.

PRISMA flow diagram.

3.2 Characteristics of included studies

The details of the included clinical studies were listed in Table 1. The included clinical studies are all case-control study. A total of 390 of PCOS and 447of healthy controls were reported. Three studies^18,19,26 were conducted in Europe, and 1 study¹⁷ was based in China. Age between the PCOS group and control group were comparable in Wu's and Coskun Simsir's studies as reported. BMI between the PCOS group and control group were comparable in Wu's and Henrike E. Peters's studies as reported. PCOS in all studies was diagnosed based on Rotterdam criteria. Woman in three included studies were with no history of vaginal delivery and in Marıa T. Prieto-S anchez's study, PCOS patients have similarly vaginal delivery rate, compared with control participants (18.3% versus 20.1%, P = 0.69).¹⁹

Table 1.
Characteristics of included studies.

Sample size

First author year Area Diagnostic criterion Study design Age of PCOS (year) Age of CONT (year) BMI of PCOS BMI of CONT VDR (P/C) PCOS CONT Main outcomes

Mar ıa T. Prieto-S anchez 2017 Spain Rotterdam criteria Case-control study 28(19–35.7) 32(21–39) 24.1 (18.6–36.3) 22.3 (18.1–32.7) 18.3/20.1 126 159 T, FSH, LH, AGD-ac, AGD-af

Henrike E. Peters 2019 Netherlands Rotterdam criteria Case-control study 32(29–35) 35(33–37) 26.4 (21.8–29.2) 23.2 (21–27.2) 0/0 43 43 AGD-ac, AGD-af, clinical hyperandrogenism

Yingchen Wu 2017 China Rotterdam criteria Case-control study 25.5(3.4) 26.5(3.5) 22.1(2.9) 21.5(2.7) 0/0 156 180 T, FSH, LH, AGD-ac, AGD-af

Coskun Simsir 2019 Turkey Rotterdam criteria Case-control study 25.7(3.6) 26.6(4.3) 27.9(5.1) 25.8(3.2) 0/0 65 65 T, FAI, FSH, LH, AGD-ac, AGD-af

										Sample size
Mar ıa T. Prieto-S anchez	2017	Spain	Rotterdam criteria	Case-control study	28(19–35.7)	32(21–39)	24.1 (18.6–36.3)	22.3 (18.1–32.7)	18.3/20.1	126	159	T, FSH, LH, AGD-ac, AGD-af
Henrike E. Peters	2019	Netherlands	Rotterdam criteria	Case-control study	32(29–35)	35(33–37)	26.4 (21.8–29.2)	23.2 (21–27.2)	0/0	43	43	AGD-ac, AGD-af, clinical hyperandrogenism
Yingchen Wu	2017	China	Rotterdam criteria	Case-control study	25.5(3.4)	26.5(3.5)	22.1(2.9)	21.5(2.7)	0/0	156	180	T, FSH, LH, AGD-ac, AGD-af
Coskun Simsir	2019	Turkey	Rotterdam criteria	Case-control study	25.7(3.6)	26.6(4.3)	27.9(5.1)	25.8(3.2)	0/0	65	65	T, FAI, FSH, LH, AGD-ac, AGD-af

CONT: CONTROL, BMI: body mass index, Kg/m², T: testosterone, FAI: free androgen index, LH: luteinizing hormone, FSH: follicle-stimulating hormone, VDR(P/C): Vaginal delivery Rate(PCOS/CONTROL)

3.3 Quality assessment

Table 2 presents the quality assessment of the individual studies included in the review. Using the Newcastle-Ottawa Scale (NOS) criteria, the scores ranged from 7 to 9 stars, with 4 studies rated as high quality. These 4 studies demonstrated good comparability between cases and controls in terms of their design and analysis. Due to the limited number of included studies, a publication bias analysis was not conducted as part of this review.

Table 2.
Methodological quality assessment of included studies.

selection exposure

First author year Adequate definition of the cases Representativeness of the cases Selection 2 of controls Definition of controls Comparability of the design or analysis^a Ascertainment of exposure Same method of ascertainment for cases and controls Non-response rate Total score

Mar ıa T. 2017 * * * * * * * * 8

Henrike E. 2019 * * * * * * * * 8

Yingchen Wu 2017 * * * * ** * * * 9

Coskun Simsir 2019 * * * * ** * * * 9

		selection		exposure
Mar ıa T.	2017	*	*	*	*	*	*	*	*	8
Henrike E.	2019	*	*	*	*	*	*	*	*	8
Yingchen Wu	2017	*	*	*	*	**	*	*	*	9
Coskun Simsir	2019	*	*	*	*	**	*	*	*	9

One star (*) gives one point. ^a: A maximum of 2 stars can be allotted in this category.

3.4 Primary outcome

The meta-analysis conducted using the Review Manager software based on the random-effects model of the four studies revealed a significant increase in AGD-ac measurements in patients with PCOS compared to control groups. The overall Mean Difference for AGD-ac was 5.23, with a 95% confidence interval (CI) of 2.60 to 7.85 and a P-value less than 0.0001. The level of heterogeneity was noted to be high at 57%. Additionally, there was a significant increase in AGD-af measurements between the PCOS and control groups, with a Mean Difference of 2.19, a 95% CI of 0.04 to 4.35, and a P-value of 0.05. The I2 value indicated high heterogeneity among these studies (Fig. 3).

Figure 3.

Forest plot about AGD-ac and AGD-af. (A: Forest plot about AGD-ac,B: Forest plot about AGD-af).

3.5 Subgroup analysis

The heterogeneity among the 4 studies on AGD-ac and AGD-af were substantial in the overall meta-analysis (I²= 57%; P = 0.07, and I²= 89%; P < 0.00001). In order to figure out the possible sources of heterogeneity, the subgroup analyses were conducted based on geographical region and history of vaginal delivery by Review Manager software, and the subgroup analysis result was showed in Table 3 and Table 4. When divided by different area, except for 1 study from China, pooled analyses of another 3 studies in Europe demonstrated consistently longer MD of AGD-ac (4.00(95% CI (1.85,6.15) and longer MD of AGD-af (1.34(95% CI (0.37, 2.31) with a low heterogeneity (I²= 0%; P = 0.77 and I²= 0%; P = 0.57), meaning that the area was a significant factor causing high heterogeneity among all the 4 studies. In Marıa T. Prieto-S anchez's study, participants partly have history of vaginal delivery, and people had history of vaginal delivery were excluded in the other 3 three studies. The pooled analysis of AGD-ac and AGD-af of those three studies still present high heterogeneity with I²= 59%; P = 0.09 and I²= 89%; P = 0.0008.

Table 3.
Subgroup analysis of AGD-ac.

Result Heterogeneity

Subgroups Number of studies MD (95% CI) P I² (%) P value

Area

Europe 3 4.00(1.85,6.15) 0.0003 0 0.77

China 1 7.80(5.82,9.78) 0.0001

History of vaginal delivery

Yes 1 4.50(1.95,7.05) 0.0005

No 3 5.22(1.31,9.13) 0.009 59 0.09

		Result	Heterogeneity
Area
Europe	3	4.00(1.85,6.15)	0.0003	0	0.77
China	1	7.80(5.82,9.78)	0.0001
History of vaginal delivery
Yes	1	4.50(1.95,7.05)	0.0005
No	3	5.22(1.31,9.13)	0.009	59	0.09

MD: mean difference; 95% CI: 95% confidence interval; P: P values of MD; I²: the value of I-squared statistics;

Table 4.

Subgroup analysis of AGD-af.

		Result		Heterogeneity
Subgroups	Number of studies	MD (95% CI)	P	I² (%)	P value
Area
Europe	3	1.34(0.37, 2.31)	0.007	0	0.57
China	1	4.60(3.77, 5.43)	0.00001	89	0.00001
History of vaginal delivery
Yes	1	1.30(0.04, 2.56)	0.04
No	3	2.50(−0.12, 5.12)	0.06	86	0.0008

MD: mean difference; 95% CI: 95% confidence interval; P: P values of MD; I²: the value of I-squared statistics;

3.6 Sensitivity analyses

We finished sensitivity analyses by STATA software (Version 13.0; Stata Corporation, College Station, TX). As for AGD-ac, the significance of the effect size remained robust when leave-one-out models were used for the sensitivity test. No study has an extreme influence on the pooled MD of AGD-ac between PCOS and control women. When sequentially excluding one study at a time, MD of AGD-af don’t have significant difference after the study of Coskun Simsir¹⁸ and Marıa T. Prieto-Sanchez²⁶ was omitted, respectively. (Fig. 4)

Figure 4.

Sensitivity analysis of AGD-ac and AGD-af (A: Sensitivity analysis of AGD-ac, B: Sensitivity analysis of AGD-af).

4 Discussion

It's the first study that reveal the associations between AGD and PCOS by using meta-analysis to obtain a powerful conclusion. Our study presented a significant increase of the AGD-ac and AGD-af between PCOS and controls, with an overall MD of AGD-ac = 5.23, 95% CI (2.60, 7.85) and MD of AGD-af = 2.19, 95% CI (0.04, 4.35). As we know, the most commonly diagnostic criteria of PCOS are the Rotterdam criteria and that criteria still have some limitation. The count of ovarian follicle numbers and measurement of testosterone levels hirsutism and acne involved in this diagnostic criterion still lack exact cut-off values.^27–29 There is an urgent need for more safe and effective markers to figure out PCOS patients.

María T. Prieto-Sánche has been shown that the measurement of AGD could improve the diagnostic accuracy when combine with AMH level, suggesting that AGD can be a reliable marker.¹⁹ Wu. also found a positive association with the T levels and AGD-af in PCOS patients, and Prieto-Sánchez reported that the presence of hirsutism in PCOS patients, was positively associated with AGD-ac, indicating that longer AGD and hyperandrogenic may share a common fetal hyperandrogenic environmental origin in PCOS, which can interfere with the function of the hypothalamus-pituitary gonadal axis and contribute to reproductive dysfunction in adults.¹⁷ In young women, longer AGD has been related to irregularities in their mother's menstrual cycle before pregnancy. Some studies showed association between shorter AGD and the presence and severity of endometriosis, which has been hypothesized to be etiologically associated with exposure to estrogen in utero and a state of low testosterone.^30,31 AGD was initially used to define sexual differentiation of animals and is a sensitive androgenic biomarker of androgenic environment during the development of the reproductive system.³² Therefore, AGD has the potential clinical value for evaluating reproductive health and could be a good, economical and non-invasive techniques measure for PCOS.

In our study, we use the leave-one-out models for the sensitivity test of AGD-ac, and no study has an extreme influence on the pooled MD of AGD-ac of women with PCOS compared with control women. It means that the stability of this study is good. The results of the sensitivity analysis of AGD-af are not as satisfactory as those of AGD-ac, since after the study of Coskun Simsir¹⁸ and Marıa T. Prieto-Sanchez²⁶ was left out, respectively, MD of AGD-af don’t have significant difference. (Figure 4)

AGD is an easily accessible measures that represents a phenotypic trait with high intra-examiner reproducibility and not large errors over multiple measurements. Many studies have revealed that the AGD is fixed in early gestation, since AGD is influenced by androgen exposure during a specific window.³³ It has been shown that AGD is stable across the woman's menstrual cycle and is stable in early childhood for both males and females.^34,35 But a cross-sectional study with adult women suggests that AGD is positively related to age and inversely related to the use of hormonal contraception, showing some plasticity in adulthood.^36,37 In our study, all the participants were women of childbearing age, and due to insufficient data, we could not complete the analysis of the correlation between age and AGD. Besides, the perineal length may be changed by nurture factors, like Perineal surgery and history of vaginal delivery. In this meta-analysis, participants of only Marıa T. Prieto-S anchez's study partly have history of vaginal delivery, but the vaginal delivery rate of PCOS patients is similarly with control participate (18.3% versus 20.1%, P = 0.69).¹⁹ In a way, we thought this would reduce the interference of vaginal birth history with the results.

This meta-analysis, involved 4 observational studies, showed some heterogeneity among these studies. Based on geographical region, expect for 1 study from China, the subgroup analyses of another 3 studies in Europe show similarly longer MD both of AGD-ac and AGD-af (4.00(95% CI (1.85,6.15) and1.34(95% CI (0.37, 2.31)) with a low heterogeneity (I²= 0%; P = 0.77 and I²= 0%; P = 0.57). What's more, we noticed that the MD of AGD between PCOS and the control in Chinese was the greatest one, indicating that ethnic origin may influence perineal length and Chinese PCOS patient may suffer more serious fetal hyperandrogenic environment. We also did the subgroup analyses based on vaginal birth history, except for Marıa T. Prieto-Sanchez's study, the pooled analyses about AGD-ac and AGD-af of the other three studies that participates without history of vaginal delivery still showed high heterogeneity (I²= 59%; P = 0.09 and I²= 89%; P = 0.0008), suggesting that vaginal birth history is not the source of heterogeneity of these studies. (Table 3 and Table 4).

Meta-analysis is indeed a valuable method for synthesizing and reviewing quantitative research to address clinical questions effectively. While our study represents the first meta-analysis to examine the association between PCOS and longer AGD, it is important to acknowledge certain limitations. The inclusion of a limited number of studies in the analysis raises the possibility of potential publication bias. Therefore, it is crucial to validate the conclusions drawn from this meta-analysis with further research studies in the future. Continued investigation and expansion of the evidence base will aid in establishing more robust and reliable conclusions regarding the relationship between PCOS and AGD.

In conclusion, our analysis is the first meta-analysis that synthesized evidence on PCOS associating with longer AGD, providing evidence for fetal androgen exposure as the possible pathogenesis behind PCOS and suggesting that measurements of AGD could hope be a useful biomarker which may optimize the PCOS diagnostic standardization. Although this meta-analysis suggests a potential association between AGD and PCOS that could lead to AGD being considered a novel clinical marker, we emphasize the necessity for additional studies. Such research should focus on confirming the findings while ruling out confounding factors, including CAH, to strengthen the evidence supporting AGD as a reliable biomarker for PCOS.

Footnotes

Acknowledgments

The authors have no acknowledgments.

Ethics declaration

This meta-analysis and all the included studies meet all the ethical standards described in the declaration of Helsinki. No ethical committee approval was required for this study.

Funding

This work was supported by the Shenzhen Science and Technology Program (grant numbers RCBS20221008093243058).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Correction (September 2025):

In this article affiliation 1 has been updated to include “the Seventh Affiliated Hospital”.

References

Rotterdam ESHRE/ASRM-Sponsored PCOS consensus workshop group . Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome (PCOS). Hum Reprod 2004; 19: 41–47.

Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group . Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome. Fertil Steril 2004; 81: 19–25.

Lizneva

Kirubakaran

Mykhalchenko

, et al. Phenotypes and body mass in women with polycystic ovary syndrome identified in referral versus unselected populations: systematic review and meta-analysis. Fertil Steril 2016; 106: 1510–1520.e2.

Azziz

. PCOS: a diagnostic challenge. Reprod Biomed Online 2004; 8: 644–648.

Azziz

Carmina

Dewailly

, et al. Positions statement: criteria for defining polycystic ovary syndrome as a predominantly hyperandrogenic syndrome: an androgen excess society guideline. J Clin Endocrinol Metab 2006; 91: 4237–4245.

Rosner

Auchus

Azziz

, et al. Position statement: utility, limitations, and pitfalls in measuring testosterone: an endocrine society position statement. J Clin Endocrinol Metab 2007; 92: 405–413.

Cattrall

Vollenhoven

Weston

. Anatomical evidence for in utero androgen exposure in women with polycystic ovary syndrome. Fertil Steril 2005; 84: 1689–1692.

Abbott

Barnett

Bruns

, et al.

Androgen excess fetal programming of female reproduction: a developmental aetiology for polycystic ovary syndrome?

Hum Reprod Update. 2005; 11: 357–374.

Risal

Pei

, et al. Prenatal androgen exposure and transgenerational susceptibility to polycystic ovary syndrome. Nat Med 2019; 25: 1894–1904.

10.

Abbott

Kraynak

Dumesic

, et al.

In utero androgen excess: a developmental commonality preceding polycystic ovary syndrome?

Front Horm Res 2019; 53: 1–17.

11.

, et al. Endocrine traits of polycystic ovary syndrome in prenatally androgenized female sprague-dawley rats. Endocr J 2010; 57: 201–209.

12.

Thankamony

Pasterski

Ong

, et al. Anogenital distance as a marker of androgen exposure in humans. Andrology 2016; 4: 616–625.

13.

Peretz

Vrooman

Ricke

, et al. Bisphenol A and reproductive health: update of experimental and human evidence, 2007–2013. Environ Health Perspect 2014; 122: 775–786.

14.

Swan

Sathyanarayana

Barrett

, et al. First trimester phthalate exposure and anogenital distance in newborns. Hum Reprod 2015; 30: 963–972.

15.

Dural

Kurbanova

Yasa

, et al. Idiopathic primary ovarian insufficiency is associated with anogenital distance, a marker for prenatal environment. Eur J Obstet Gynecol Reprod Biol 2021; 258: 304–308.

16.

Mira-Escolano

Mendiola

Mínguez-Alarcón

, et al. Longer anogenital distance is associated with higher testosterone levels in women: a cross-sectional study. BJOG 2014; 121: 1359–1364.

17.

Zhong

Chen

, et al. Polycystic ovary syndrome is associated with anogenital distance, a marker of prenatal androgen exposure. Hum Reprod 2017; 32: 937–943.

18.

Simsir

Pekcan

Aksoy

, et al. The ratio of anterior anogenital distance to posterior anogenital distance: a novel-biomarker for polycystic ovary syndrome. J Chin Med Assoc 2019; 82: 782–786.

19.

Sanchez-Ferrer

Mendiola

Hernandez-Penalver

, et al. Presence of polycystic ovary syndrome is associated with longer anogenital distance in adult Mediterranean women. Hum Reprod 2017; 32: 2315–2323.

20.

Peters

Laeven

CHC

Trimbos

CJMA

, et al. Anthropometric biomarkers for abnormal prenatal reproductive hormone exposure in women with Mayer-Rokitanksy-Ku Euro ster-Hauser syndrome, polycystic ovary syndrome, and endometriosis. Fertil Steril 2020; 114: 1297–1305.

21.

Hafizović

Čaušević

Deumić

, et al. The use of artificial intelligence in diagnostic medical imaging: systematic literature review. 2021 IEEE 21st International Conference on Bioinformatics and Bioengineering (BIBE), Kragujevac, Serbia, 2021, pp. 1–6.

22.

Hrvat

Spahić

Pokvić

, et al. Artificial neural networks for prediction of medical device performance based on conformity assessment data: infusion and perfusor pumps case study. 2020 9th Mediterranean Conference on Embedded Computing (MECO), Budva, Montenegro, 2020, pp. 1–4.

23.

Badnjevic

Cifrek

Koruga

. Classification of Chronic Obstructive Pulmonary Disease (COPD) using integrated software suite. In: Roa Romero

(ed.) XIII Mediterranean conference on medical and biological engineering and computing 2013. IFMBE Proceedings, vol 41. Cham: Springer, 2014, pp.911–914.

24.

Veljović

Špirtović-Halilović

Muratović

, et al. artificial neural network and docking study in design and synthesis of xanthenes as antimicrobial agents. In: Badnjevic

(ed.) CMBEBIH 2017. IFMBE Proceedings, vol 62. Singapore: Springer, 2017, pp.617–626.

25.

Stang

. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol 2010; 25: 603–605.

26.

Hernandez-Penalver

Sanchez-Ferrer

Mendiola

, et al. Assessment of anogenital distance as a diagnostic tool in polycystic ovary syndrome. Reprod Biomed Online 2018; 37: 741–749.

27.

Ehrmann

. Polycystic ovary syndrome. N Engl J Med 2005; 352: 1223–1236.

28.

Rutkowska

Diamanti-Kandarakis

. Polycystic ovary syndrome and environmental toxins. Fertil Steril 2016; 106: 948–958.

29.

Franks

McCarthy

Hardy

. Development of polycystic ovary syndrome: involvement of genetic and environmental factors. Int J Androl 2006; 29: 278–285.

30.

Crespi

Evans

. Prenatal origins of endometriosis pathology and pain: reviewing the evidence of a role for low testosterone. J Pain Res 2023; 16: 307–316.

31.

Crespi

. Variation among human populations in endometriosis and PCOS A test of the inverse comorbidity model. Evol Med Public Health 2021; 9: 295–310.

32.

Buggio

Barbara

Dridi

, et al. Anogenital distance and gynaecological diseases: a narrative review. Italian J Gynaecol Obst 2020; 32: 200–207.

33.

Jain

Goyal

Chowdhary

, et al. Anogenital distance is determined during early gestation in humans. Hum Reprod 2018; 33: 1619–1627.

34.

Barrett

Parlett

Swan

. Stability of proposed biomarkers of prenatal androgen exposure over the menstrual cycle. J Dev Orig Health Dis 2015; 6: 149–157.

35.

Thankamony

Ong

Dunger

, et al. Anogenital distance from birth to 2 years: a population study. Environ Health Perspect 2009; 117: 1786–1790.

36.

Wainstock

Shoham-Vardi

Sheiner

, et al. Fertility and anogenital distance in women. Reprod Toxicol 2017; 73: 345–349.

37.

Verma

Maan

Gautam

, et al. Unveiling the role of Artificial Intelligence (AI) in Polycystic Ovary Syndrome (PCOS) diagnosis: a comprehensive review. Reprod Sci 2024; 31(10): 2901–2915.

										Sample size
First author	year	Area	Diagnostic criterion	Study design	Age of PCOS (year)	Age of CONT (year)	BMI of PCOS	BMI of CONT	VDR (P/C)	PCOS	CONT	Main outcomes
Mar ıa T. Prieto-S anchez	2017	Spain	Rotterdam criteria	Case-control study	28(19–35.7)	32(21–39)	24.1 (18.6–36.3)	22.3 (18.1–32.7)	18.3/20.1	126	159	T, FSH, LH, AGD-ac, AGD-af
Henrike E. Peters	2019	Netherlands	Rotterdam criteria	Case-control study	32(29–35)	35(33–37)	26.4 (21.8–29.2)	23.2 (21–27.2)	0/0	43	43	AGD-ac, AGD-af, clinical hyperandrogenism
Yingchen Wu	2017	China	Rotterdam criteria	Case-control study	25.5(3.4)	26.5(3.5)	22.1(2.9)	21.5(2.7)	0/0	156	180	T, FSH, LH, AGD-ac, AGD-af
Coskun Simsir	2019	Turkey	Rotterdam criteria	Case-control study	25.7(3.6)	26.6(4.3)	27.9(5.1)	25.8(3.2)	0/0	65	65	T, FAI, FSH, LH, AGD-ac, AGD-af

		selection					exposure
First author	year	Adequate definition of the cases	Representativeness of the cases	Selection 2 of controls	Definition of controls	Comparability of the design or analysis^a	Ascertainment of exposure	Same method of ascertainment for cases and controls	Non-response rate	Total score
Mar ıa T.	2017	*	*	*	*	*	*	*	*	8
Henrike E.	2019	*	*	*	*	*	*	*	*	8
Yingchen Wu	2017	*	*	*	*	**	*	*	*	9
Coskun Simsir	2019	*	*	*	*	**	*	*	*	9

Polycystic ovary syndrome in women is associated with longer anogenital distance,a new potential biomaker for PCOS

Abstract

Background

Methods

Results

Conclusion

Keywords

1 Introduction

2.1 Databases

2.2 Study selection

2.3 Data extraction

2.4 Quality assessment

2.5 Data synthesis

3 Results

3.1 Study selection

Table 3. Subgroup analysis of AGD-ac. Result Heterogeneity Subgroups Number of studies MD (95% CI) P I2 (%) P value Area Europe 3 4.00(1.85,6.15) 0.0003 0 0.77 China 1 7.80(5.82,9.78) 0.0001 History of vaginal delivery Yes 1 4.50(1.95,7.05) 0.0005 No 3 5.22(1.31,9.13) 0.009 59 0.09

Footnotes

Acknowledgments

Ethics declaration

Funding

Declaration of conflicting interests

Availability of data and materials

Correction (September 2025):

References

Table 3.
Subgroup analysis of AGD-ac.

Result Heterogeneity

Subgroups Number of studies MD (95% CI) P I² (%) P value

Area

Europe 3 4.00(1.85,6.15) 0.0003 0 0.77

China 1 7.80(5.82,9.78) 0.0001

History of vaginal delivery

Yes 1 4.50(1.95,7.05) 0.0005

No 3 5.22(1.31,9.13) 0.009 59 0.09