Genetic association between LHCGR variants and polycystic ovary syndrome: a meta-analysis

Abstract

Background

The luteinizing hormone/choriogonadotropin receptor (LHCGR) is mapped at the 2p16.3 region, which was identified as a PCOS-susceptible locus in the first genome-wide association study (GWAS). Since then, several variants of LHCGR have been documented as susceptible loci for the development of PCOS, with striking differences due to diverse racial backgrounds. The present meta-analysis was conducted to unravel the association between LHCGR variants and PCOS.

Methods

Databases such as PubMed, PCOSkb, and Google Scholar were extensively searched to gather the relevant articles. To determine the heterogeneity, I² statistic was used followed by fixed effect and random effect models depending on the degree of heterogeneity present. Different genetic models were used to assess the association between LHCGR variants and the risk of PCOS.

Results

Out of the six studied variants, the rs2293275 and rs12470652 did not show an association with PCOS among any genetic models. The rs13405728 showed an association with all models in the overall analysis and after stratification it was associated but only in Asians. Only the recessive and additive models were found to be significantly associated with PCOS for the rs4539842 variant in the overall analysis. The rs4953616 variant displays an association under all models with the Asian and Indian populations during sub-group analysis. For the rs7371084 variant, dominant and allele models exhibited an association with PCOS during overall and sub-group analysis; however, in later only the Asians were significantly associated.

Conclusions

The rs4953616 variant exhibited the risk of PCOS in Indians however served as protective against the risk in Asians. The variants rs13405728, rs4539842, and rs7371084 were found to be acting as a protective factor against the development of PCOS.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10815-025-03434-4.

Introduction

The luteinizing hormone/choriogonadotropin receptor (LHCGR) belongs to a rhodopsin-like family of G protein-coupled receptors (GPCRs) and is primarily expressed in ovarian theca cells and granulosa cells [1]. LHCGR serves as a reception site for luteinizing hormone (LH) and, upon binding, results in steroidogenesis, folliculogenesis, and ovulation [2]. The enhanced sensitivity to LH and overexpression of LHCGR in theca and granulosa cells have been documented in women with polycystic ovary syndrome (PCOS) [1, 3, 4], which results in excess production of androgens which in turn disrupts the production of sex hormone binding globulin (SHBG) leading to more proportions of free testosterone in the bloodstream [5]. Hyperandrogenism along with ovulatory dysfunction, and multiple cysts form the basis of all three diagnostic criteria of PCOS, i.e. NIH, Rotterdam 2003, and AE-PCOS criteria. PCOS has been described as not only reproductive but metabolic and endocrine disorder affecting 6–26% of women of fertile age worldwide [6]. Being a multifactorial condition, the pathogenesis of PCOS involves the interactions among several genes and proteins under the influence of environmental factors [7].

The LHCGR is encoded by a single-copy gene mapped at the chromosome 2p16.3 region and consists of 11 exons and ten introns [8]. The LHCGR was identified as a candidate gene in the pathogenesis of PCOS in the first genome-wide association study (GWAS) published on the Han Chinese population [9]. After that, several genetic association studies were carried out to analyze the association between potential genetic variants of LHCGR and the risk of PCOS across different nations to narrow down the causal variants which resulted in conflicting outcomes due to discrepancies in sample sizes, diagnostic criteria, and ethnicity [10–17]. In order to eliminate the disparities in the association studies and arrive at a single comprehensive statement, meta-analysis has been considered the best approach. Meta-analysis integrates the data from several studies on the same subject and aids in reaching a single outcome with enhanced statistical robustness [18]. We have undertaken a thorough and updated meta-analysis of the studies published until now to establish a correlation between the susceptible variants of LHCGR and the risk of PCOS.

Material and methodology

Literature review and selection criteria

A comprehensive literature survey was performed on PubMed, PCOSkb, SCOPUS, Embase, and Google Scholar databases to select the published work on the LHCGR variants (rs2293275, rs12470652, rs13405728, rs4539842, rs4953616, and rs7371084) and PCOS. The six variants chosen for this meta-analysis are the most extensively and frequently studied for their potential risk of developing PCOS. These variants have been linked with PCOS development and might be associated with hormonal dysregulation, menstrual irregularities, increased risk of type 2 diabetes, dyslipidemia, cardiovascular risk, and psychological issues [13–17]. The keywords used for the literature search are as follows: PCOS, LHCGR, LHCGR polymorphisms, genetic association, and LHCGR genetic variants. The studies (1) with a case–control design, (2) following Hardy–Weinberg equilibrium (HWE), (3) that worked on human blood samples (4), with English language (5), where genotypic data for cases and controls were available, and (6) with association between LHCGR variants and PCOS analyzed were included for the present meta-analysis. However, the studies (1) not obeying HWE, (2) that is not a case–control design, (3) that worked on animal models, (4) with language used other than English, and (5) where genotypic data were not available were excluded.

Data extraction and quality evaluation

The three authors individually gathered data from the chosen studies and worked with the fourth author to address any inconsistencies. The information obtained comprises the first author’s name, country of origin, year of publication, total number of cases and controls, diagnostic criteria used, methods for genotyping, and genotypic frequencies. The adopted checklist (S1 table) and flow diagram were based on the PRISMA 2020 (Preferred Reporting Items for Systematic Reviews and Meta-Analysis).

Statistics

The Review Manager 5.4.1 software was used to carry out the meta-analysis. To evaluate the deviations from HWE, a goodness of fit Chi-square test was applied. The strength of the association between LHCGR variants and the risk of PCOS was measured using pooled odds ratios (ORs) with 95% confidence intervals (CI). A p < 0.05 is considered statistically significant. The I² statistic was employed to determine the heterogeneity by applying a fixed effect model (FEM) when the I² value was below 50% (heterogeneity is absent). In contrast, a random effect model (REM) was applied when the I² value exceeded 50% (heterogeneity is present). The genetic association analysis was conducted by undertaking four genetic models which are as follows: (1) dominant model (GG vs. GA + AA), (2) recessive model (GG + GA vs. AA), (3) additive model (GG vs. AA), and (4) allele model (G vs. A). The Begg-Mazumdar test was performed to address the publication bias.

Results

Characteristics of the selected studies

The present meta-analysis was performed on six LHCGR variants and the PCOS risk. We identified 263 potential studies after searching the literature, and 232 studies were excluded as these did not meet the inclusion criteria. Out of 232 publications, 127 irrelevant articles were eliminated after examining the titles and abstracts. After that, a thorough assessment was carried out on each publication resulting in the removal of 26 papers classified as review articles. Additionally, 46 and 11 more studies were eliminated as these were not case–control studies and were duplicates, respectively. Furthermore, 14 studies that did not have sufficient data for meta-analysis and eight studies disobeying HWE were excluded. In conclusion, 31 studies were selected for the present meta-analysis. Search proceedings and results are represented as a flowchart in Fig. 1. The genotype frequency, HWE p value, and characteristic features of each study are mentioned in Table 1.

Fig. 1.

Flowchart for the selection process of included studies for meta-analysis

Table 1.

Details of included studies in the current meta-analysis

Study	Country	Year	Diagnostic criteria	Genotyping method	Sample size (case/control)	Genotype						HWE
						Case			Control
rs2293275						GG	GA	AA	GG	GA	AA
Valkenburg et al	Netherlands	2009	Rotterdam criteria	Taqman assay	503/2916	184	240	79	978	1426	512	0.84
Capalbo et al	Italy	2012	Rotterdam criteria	PCR–RFLP	159/180	63	76	20	103	65	12	0.68
Bassiouny et al	Egypt	2014	Rotterdam criteria	PCR–RFLP	100/60	47	27	26	42	18	0	0.17
Robeva et al	Bulgaria	2018	Rotterdam criteria	Sequencing	60/40	13	20	27	5	14	21	0.29
Lidaka et al	Latvia	2021	ESHRE	PCR–RFLP	63/67	26	26	11	22	31	14	0.61
Alarcón-Granados et al	Colombia	2022	Rotterdam criteria	MassARRAY System	49/49	20	23	6	23	24	2	0.16
Atoum et al	Jordan	2022	Rotterdam criteria	PCR–RFLP	55/55	26	20	9	23	22	10	0.25
rs12470652						TT	TC	CC	TT	TC	CC
Capalbo et al	Italy	2012	Rotterdam criteria	PCR–RFLP	180/153	174	6	0	144	9	0	0.7
Robeva et al	Bulgaria	2018	Rotterdam criteria	Sequencing	59/40	55	4	0	37	3	0	0.8
rs13405728						AA	AG	GG	AA	AG	GG
Chen et al. (a)	China	2011	Rotterdam criteria	Affymetrix SNP 6.0 chips	744/895	491	227	26	472	356	67	0.99
Chen et al. (b)	China	2011	Rotterdam criteria	Affymetrix SNP 6.0 chips	2840/5012	1854	881	105	2918	1812	282	0.97
Chen et al. (c)	China	2011	Rotterdam criteria	Affymetrix SNP 6.0 chips	498/780	334	148	16	417	306	57	0.96
Lerchbaum et al	Caucasian	2011	Rotterdam criteria	Real-time PCR	513/317	445	66	2	282	32	3	0.06
Goodarzi et al	Europe	2012	NIH criteria	TaqMan assay	939/957	859	78	2	854	99	4	0.53
Erikson et al	Denmark	2012	Rotterdam criteria	Real-time PCR	169/248	156	13	0	220	27	1	0.86
Cui et al	China	2013	Rotterdam criteria	SNP array	1548/4525	1016	483	49	2630	1643	252	0.82
Almawi et al	Bahrain	2015	Rotterdam criteria	Real-time PCR	203/211	168	35	0	170	40	1	0.4
Ha et al	China	2015	Rotterdam criteria	Sequencing	151/99	99	41	11	49	38	12	0.28
Vishnubotla et al	India	2020	Rotterdam criteria	Sequencing	300/300	243	54	3	197	92	11	0.94
Wan et al	China	2021	Rotterdam criteria	Sequencing	400/480	263	123	14	270	178	32	0.71
Alarcón-Granados et al	Colombia	2022	Rotterdam criteria	MassARRAY system	49/49	40	8	1	36	12	1	1
Dadachanji et al	India	2024	Rotterdam criteria	Sequencing	504/234	352	139	13	161	66	7	0.94
rs4539842						non/non	non/ins	ins/ins	non/non	non/ins	ins/ins
Valkenburg et al	Caucasian	2009	Rotterdam criteria	TaqMan Assay	511/2344	283	202	26	1220	934	190	0.54
Caplbo et al	Italy	2012	Rotterdam criteria	PCR–RFLP	156/174	73	74	9	90	76	8	0.1
El-Shal et al	Egypt	2015	Rotterdam criteria	PCR–RFLP	210/200	109	92	9	99	90	11	0.1
rs4953616						CC	CT	TT	CC	CT	TT
Almawi et al	Bahrain	2015	Rotterdam criteria	Real-time PCR	203/211	42	91	70	27	81	103	0.08
Alarcón-Granados et al	Colombia	2022	Rotterdam criteria	MassARRAY system	49/49	4	23	22	3	21	25	0.6
Singh et al	India	2023	Rotterdam criteria	ARMS	443/380	92	217	134	100	198	82	0.38
rs7371084						TT	TC	CC	TT	TC	CC
Almawi et al	Bahrain	2015	Rotterdam criteria	Real-time PCR	203/211	184	17	2	172	37	2	0.99
Alarcón-Granados et al	Colombia	2022	Rotterdam criteria	MassARRAY system	49/49	37	11	1	28	21	0	0.56
Singh et al	India	2023	Rotterdam criteria	PCR–RFLP	443/380	397	42	4	329	49	2	0.9

Meta-analysis

Four different genetic models, such as dominant, recessive, additive, and allele, were used to evaluate the association of included genetic variants with the risk of PCOS, and the results are illustrated in Table 2 and Forest plots are mentioned in supporting information, S2 (Fig. 2–11).

Table 2.

Pooled and sub-group analysis

Genetic model	I2	OR (CI), p value
		Overall analysis	Caucasian studies	Asian studies	Indian studies
rs2293275
Dominant	74%	1.14 (0.75–1.73), 0.54	1.05 (0.89–1.23), 0.56	0.80 (0.38– 1.70), 0.57	-
Recessive	59%	1.19 (0.72–1.96), 0.49	1.07 (0.87–1.32), 0.53	0.88 (0.33–2.37), 0.80	-
Additive	70%	1.26 (0.65–2.43), 0.50	1.07 (0.84–1.35), 0.59	0.80 (0.28–2.30), 0.67	-
Allele	84%	1.20 (0.82–1.75), 0.34	1.04 (0.93–1.17), 0.45	0.85 (0.49–1.48), 0.58	-
rs12470652
Dominant	0%	0.64 (0.27–1.53), 0.32	-	-	-
Recessive	-	-	-	-	-
Additive	-	-	-	-	-
Allele	0%	0.65 (0.28–1.53), 0.32	-	-	-
rs13405728
Dominant	55%	0.69 (0.62–0.78), 0.00001	0.85 (0.68–1.07), 0.17	0.69 (0.65–0.73), 0.00001	0.95 (0.68–1.33), 0.78
Recessive	0%	0.56 (0.48–0.65), 0.00001	0.52 (0.18–1.49), 0.22	0.56 (0.48–0.65), 0.00001	0.86 (0.34–2.18), 0.75
Additive	0%	0.50 (0.43–0.58), 0.00001	0.51 (0.18–1.47), 0.21	0.49 (0.42–0.58), 0.00001	0.85 (0.33–2.17), 0.73
Allele	54%	0.71 (0.64–0.78), 0.00001	0.84 (0.68–1.04), 0.11	0.71 (0.68–0.75), 0.00001	0.95 (0.71–1.27), 0.73
rs4539842
Dominant	0%	0.92 (0.78–1.08), 0.31	-	-	-
Recessive	0%	0.69 (0.48–0.97), 0.04	-	-	-
Additive	17%	0.68 (0.47–0.97), 0.03	-	-	-
Allele	27%	0.90 (0.79–1.02), 0.10	-	-	-
rs4953616
Dominant	75%	0.88 (0.43–1.80), 0.72	0.73 (0.16–3.46), 0.70	0.56 (0.33–0.95), 0.03	1.36 (0.99–1.88), 0.06
Recessive	88%	0.89 (0.42–1.92), 0.77	0.78 (0.35–1.73), 0.54	0.55 (0.37–0.82), 0.003	1.58 (1.15–2.17), 0.005
Additive	88%	0.84 (0.27–2.56), 0.76	0.66 (0.13–3.28), 0.61	0.44 (0.25–0.77), 0.004	1.78 (1.20–2.64), 0.004
Allele	90%	0.89 (0.50–1.58), 0.69	0.82 (0.44–1.52), 0.53	0.62 (0.47–0.82), 0.001	1.33 (1.09–1.61), 0.004
rs7371084
Dominant	18%	0.60 (0.43–0.82), 0.001	0.43 (0.18–1.02), 0.06	0.46 (0.25–0.82), 0.009	0.75 (0.49–1.14) 0.18
Recessive	0%	1.57 (0.49–5.11), 0.45	3.06 (0.12–77.02), 0.50	1.04 (0.15–7.45), 0.97	1.72 (0.31–9.45), 0.53
Additive	0%	1.42 (0.44–4.62), 0.56	2.28 (0.09–58.07), 0.62	0.93 (0.13–6.71), 0.95	1.66 (0.30–9.11), 0.56
Allele	0%	0.66 (0.49–0.88), 0.005	0.56 (0.26–1.20), 0.13	0.51 (0.29–0.87), 0.01	0.80 (0.54–1.19), 0.27

rs2293275 (G > A)

In this variant, none of the genetic models display any association with PCOS in the overall analysis (Table 2; S1_Fig 2), and even after stratification, no relation was observed for the rs2293275 variant (Table 2; S1_Fig 3). Significant heterogeneity was observed under all genetic models in overall and sub-group analysis.

rs12470652 (T > C)

In the overall analysis, only dominant and allele models were examined because of the absence of a homozygous mutant genotype (CC), and the variant showed no significant association with the development of PCOS (Table 2; S1_Fig 4). The sub-group analysis could not be done because of the limited number of studies meeting the selection criteria, and these were from the same ethnicity; hence, no heterogeneity was observed for rs12470652.

rs13405728 (A > G)

A significant association was found in all genetic models under pooled analysis (dominant: OR = 0.69, CI = 0.62 – 0.78, p = 0.00001; recessive: OR = 0.56, CI = 0.48–0.65, p = 0.00001; additive: OR = 0.50, CI = 0.43–0.58, p = 0.00001; allele: OR = 0.71, CI = 0.64–0.78, p = 0.00001) (Table 2; S1_Fig 5). Following stratification, the rs13405728 variant displayed an association only in the Asian population among all genetic models (dominant: OR = 0.69, CI = 0.65–0.73, p = 0.00001; recessive: OR = 0.56, CI = 0.48–0.65, p = 0.00001; additive: OR = 0.49, CI = 0.42–0.58, p = 0.00001; allele: OR = 0.71, CI = 0.68–0.75, p = 0.00001) (Table 2; S1_Fig 6). The heterogeneity was evident in the dominant and allele models but was absent in recessive and additive models under pooled analysis. However, when performing sub-group analysis, heterogeneity was only seen in the dominant and allele models within the Asian population. Our findings suggested the protective role of the rs13405728 variant against the risk of PCOS development in Asians.

rs4539842 (non > ins)

In the overall analysis, the recessive (OR = 0.69, CI = 0.48–0.97, p = 0.04) and additive (OR = 0.68, CI = 0.47–0.97, p = 0.03) models showed significant association with PCOS; however, no association was noted for the dominant and allele models (Table 2; S1_Fig 7). The stratification of data for rs4539842 was not possible due to the similar ancestry of the studied population. Furthermore, we did not observe heterogeneity in any of the four genetic models. The results of the meta-analysis for rs4539842 demonstrated a protective effect against the risk of PCOS.

rs4953616 (C > T)

During the overall analysis, no significant association of the rs4953616 variant was found with PCOS under any genetic model; however, we did observe the heterogeneity among all genetic models (Table 2; S1_Fig 8). Following sub-group analysis, all four genetic models appeared significantly associated with the Asian (dominant: OR = 0.56, CI = 0.33–0.95, p = 0.03; recessive: OR = 0.55, CI = 0.37–0.82; p = 0.003; additive: OR = 0.44, CI = 0.25–0.77, p = 0.004; allele: OR = 0.62, CI = 0.47–0.82, p = 0.001) and Indian (dominant: OR = 1.36, CI = 0.99–1.88, p = 0.06; recessive: OR = 1.58, CI = 1.15–2.17, p = 0.005; additive: OR = 1.78, CI = 1.20–2.64, p = 0.004; allele: OR = 1.33, CI = 1.09–1.61, p = 0.004) populations (Table 2; S1_Fig 9). The heterogeneity could not be calculated in sub-group analysis as only one study was available in each group.

rs7371084

The dominant (OR = 0.60, CI = 0.43–0.82, p = 0.001) and allele (OR = 0.66, CI = 0.49–0.88, p = 0.005) genetic models were significantly associated with PCOS in the overall analysis, and heterogeneity was absent among the four genetic models (Table 2; S1_Fig 10). After stratification, only the dominant (OR = 0.46, CI = 0.25–0.82, p = 0.009) and allele (OR = 0.51, CI = 0.29–0.87, p = 0.01) models displayed significant association with PCOS in the Asian population (Table 2; S1_Fig 11). In this variant, heterogeneity analysis was also not calculated because there was only one study in each group.

Publication bias

Begg’s funnel plots were utilized to assess the publication bias, which rendered no evidence of biases in any of the genetic models for studied genetic variants, except for the rs2293275 variant. A leave-one-out sensitivity analysis was performed to check the stability of the meta-analysis. After systematically excluding each study, statistically comparable results were still obtained, showing statistically valid findings from our meta-analysis.

Discussion

The present meta-analysis analyzed the association of six LHCGR variants with PCOS, of which three variants fall in the exonic region. The rs2293275 and rs12470652 variants are located in exon 10 of LHCGR, resulting in a change of amino acids [10]. A total of seven studies were included for rs2293275 involving 989 cases and 3367 controls, and it was found that none of the genetic models show an association with PCOS and after stratification still no association appeared in any of the models. In 2018, Zou and colleagues conducted a meta-analysis and documented no association of rs2293275 with PCOS in Asians; however, they did observe a 4.115 times increase in the risk of PCOS in Caucasian women carrying AA genotype [1]. For the rs12470652 variant, only two studies were available to evaluate the association of this variant with PCOS which belonged to the Caucasian population and comprised 239 cases and 193 controls. Only dominant and allele models were utilized because of the lack of mutant genotype and neither model revealed any association with the risk of PCOS and data could not be stratified as both studies were carried out on the Caucasians. The rs13405728 is the most extensively studied variant of the LHCGR gene as it was identified as the most significant locus in the first GWAS performed on the Han Chinese population [9]. The current meta-analysis included 13 studies composed of 8858 women with PCOS and 14,107 women without PCOS. All the studied genetic models were significantly associated with PCOS in an overall analysis, and after stratification, the models remain significant but only in the Asian population. We also observed the rs13405728 variant to be associated with a reduced risk of PCOS in women from Asia. Our results are in agreement with a meta-analysis by Zou et al. [1]. However, another meta-analysis by Wan and colleagues documented the rs13405728 variant as a susceptible locus for PCOS in two genetic models [19]. If a genetic variant has been documented as a risk factor in one population and appears to have a protective role in another, that signifies the complex and heterogeneous nature of a disorder. The underlying mechanism for these conflicting findings could be attributed to gene-environment interactions, population differences, study design, etc. PCOS is a multifactorial disorder involving the complex interplay between several genes and proteins under the influence of environmental factors. Women from diverse geographical regions encounter distinct environmental conditions, lifestyles, diets, and socioeconomic challenges which can impact the expression of genes to varying degrees resulting in differential expression of genetic variants. In addition, different populations exhibit varying frequencies of genotypes (such as wild-type, heterozygous, and mutant genotype), and different environments and other factors can cause alterations in the effect of genetic variants. Another possible factor that can result in contrasting outcomes is the study design like disparity in sample sizes, prevalence of a particular disorder in a population, diagnostic criteria used, and the difference in the number of studies included in the meta-analysis. Besides, genetic heterogeneity could also be considered a factor responsible for inconsistent findings. Within a single population, individuals may possess diverse genetic backgrounds and ethnicities which could lead to differential expression of a genetic variant as the same gene would interact differently with other genes among individuals with unique allele frequencies. These factors might be responsible for the rs13405728 variant serving as a risk factor in one study while showing a protective effect in another and could also describe why the rs4953616 variant was found to be associated with enhanced risk of PCOS in Indian women yet appeared as a protective variant among Asians. The third exonic variant rs4539842 analyzed in the present meta-analysis is an insertion of two amino acids at the 18th position in exon 1. Three studies comprised 877 PCOS cases, and 2718 controls were included for analysis. This variant showed an association only with recessive and additive models as a protection against the risk of PCOS in Caucasians. Similar findings were also obtained by Zou et al. in their meta-analysis [1]. The rs4953616 and rs7371084 are the least studied variants of LHCGR; nevertheless, these variants have been documented as susceptible loci for the development of PCOS [13]. To the best of our knowledge, only three studies are available in the literature that have evaluated the association of rs4953616 and rs7371084 variants with PCOS and the current meta-analysis undertook these studies involving 695 women with PCOS and 640 healthy controls [13, 15, 17]. For rs4953616, no association was observed in the overall analysis. However, after sub-group analysis, this variant was identified as a protective locus in Asians whereas it was found to significantly elevate the risk of PCOS in Indian women. The rs7371084 variant was also revealed to be a protective variant in Asians in dominant and allele models; however, it did not show any association in women from India. After a genetic variant is found to be significantly associated with a particular disease regardless of whether the variant is increasing or decreasing the risk of that disease in a specific population, then it could prove beneficial in clinical practice. Clinicians can identify women at a higher risk before the onset of disease by genotyping analysis allowing for early intervention and possibly better outcomes. Early implementation of medical care or preventive actions is crucial in handling complex diseases like PCOS. Besides, the knowledge of PCOS-susceptible variants can also be used for large-scale screening programs and the designing of personalized medical interventions for an individual.

In our meta-analysis, we observed LHCGR variants to be associated with PCOS but in a population-specific manner, and this would aid in a better understanding of the genetic predisposition of a particular population toward a specific disease. The data yielded from population-specific studies can provide insights that can be used for the advancement of personalized treatment plans signifying the importance of studies involving a single population. However, there are some circumstances where the results from targeted studies do not apply to other populations restricting the wider impact of findings. For instance, GWAS by Chen et al. in the Han Chinese population concluded the highly significant association of the LHCGR rs13405728 variant with PCOS development; however, these findings were not found to be significant in the women of European ancestry. Although population-specific research may yield more accurate and reliable data for certain groups, it may not represent the whole spectrum of genetic diversity that exists among distinct populations across the globe. So, in contrast to studies focused on a single population, studies involving multiple populations help in identifying the genetic variations that may be universal or particular to certain ethnic groups. Considering the significance of both types of studies, a balanced approach would be most effective where multi-population studies offer broad knowledge about genetic associations, while population-specific research highlights how these associations may differ by geographic region or ethnicity.

Limitations

To date, only three studies have been carried out to evaluate the influence of rs4953616 and rs7371084 variants on PCOS development with a combined sample size of 695 cases and 640 controls. In contrast, sufficient studies are available for other variants. Nonetheless, the present meta-analysis has documented the association of both variants with PCOS in a population-specific manner. The disparity in the number of studies for different variants and the availability of varying sample sizes among populations could impact the final results. Since the dataset is primarily composed of Asian populations and smaller sample sizes for other ethnic groups, this may affect the results. So larger sample sizes encompassing diverse ethnic groups are required to arrive at a robust conclusion. The rs4953616 variant exhibited the greatest degree of heterogeneity in comparison to other variants, potentially because of the limited number of studies available representing only three ethnicities. Consequently, it is not feasible to generalize the findings of this variant as in meta-analysis; when there is a rise in the level of heterogeneity, it becomes challenging to substantiate the conclusion.

Conclusion

The present meta-analysis investigated the association of six variants of the LHCGR gene with PCOS and concluded that four of them (rs13405728, rs4539842, rs4953616, and rs7371084) are influencing the development of PCOS. The rs4953616 was linked with an increased risk of PCOS in Indian women but served as a protective factor in Asians. The other three variants (rs13405728, rs4539842, and rs7371084) were found to be protective against the risk of PCOS. The data can be utilized to facilitate the advancement of therapeutic interventions and screening purposes to take preventive measures before onset of disease. Our meta-analysis contained a higher number of publications than previous ones, and to the best of our knowledge, this is the first meta-analysis to assess the association of rs4953616 and rs7371084 variants with PCOS. Further investigations are needed to be done with larger sample sizes encompassing diverse genetic backgrounds to obtain deeper insights into how LHCGR variants impact the development of PCOS and arrive at a more applicable conclusion with stronger statistical significance.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contribution

Conceptualization: Sukhjashanpreet Singh, Mandeep Kaur.

Data curation: Sukhjashanpreet Singh, Mandeep Kaur.

Formal analysis: Sukhjashanpreet Singh, Mandeep Kaur.

Investigation: Sukhjashanpreet Singh, Mandeep Kaur.

Methodology: Sukhjashanpreet Singh, Mandeep Kaur, Pallvi Thapar.

Supervision: Anupam Kaur.

Writing—original draft: Sukhjashanpreet Singh.

Writing—review and editing: Sukhjashanpreet Singh, Mandeep Kaur, Pallvi Thapar, Anupam Kaur.

Data availability

N.A.

Declarations

Ethics approval

This meta-analysis was conducted on published case–control studies, so ethical approval was not required.

Competing interests

The authors declare no competing interests.