Background

Gestational diabetes mellitus (GDM) is defined as “glucose intolerance of different degrees that occurs or is discovered for the first time during pregnancy” [1]. It has a global incidence of approximately 14.0% [2]. The prevalence of GDM has increased by more than 30% in many countries over the past 30 years, while the birth rate has declined [3]. GDM affects both pregnant women and their offspring. GDM is a risk factor for gestational hypertension, gestational preeclampsia, and diabetes [4, 5]. Compared with the offspring of healthy women, the offspring of women with GDM are at considerably higher risk of developing obesity and diabetes in childhood and adulthood [6]; thus, GDM is an important public health concern [7]. However, the specific mechanism and factors influencing GDM remain unclear [8]. Genetic and environmental factors play important roles in the etiology of GDM [9]. Recent studies have found that single nucleotide polymorphisms (SNPs) in TCF7L2, CAPN10, KCNQ1, and ADIPOQ are associated with the onset of gestational diabetes [10, 11]. Of these, the earliest and most intensively studied are genes encoding transcription factor 7-like 2 (TCF7L2) and calpain 10 (CAPN10). The relationship between TCF7L2 and CAPN10 and the susceptibility to GDM has been studied in populations in different geographical regions. However, most of these studies have been based on a limited number of loci, and the results have been inconsistent. In this study, we screened published literature for studies on the relationship between TCF7L2 and CAPN10 gene polymorphisms and the incidence of diabetes mellitus during pregnancy in different populations. We selected five loci reported in a large number of articles and included them in a meta-analysis to determine genetic susceptibility to GDM in different populations.

Methods

Literature search

We prioritized the selection of SNPs with well-documented correlations established in existing literature on GDM. These specific SNPs, identified for their substantiated association with the onset or progression of GDM in previous research, provide a solid foundation of relevance to this condition. Additionally, our study involved an exhaustive review of published articles to pinpoint epidemiological studies exploring the relationship between SNPs within TCF7L2 and CAPN10 and their connection to gestational diabetes. This rigorous approach aimed to ensure a comprehensive understanding of the genetic variations associated with GDM and specifically targeted these genes for deeper investigation within the context of this condition. We searched eight Chinese and English databases: Cochrane, Elton B. Stephens. Company (EBSCO), Embase, Scopus, Web of Science, China National Knowledge Infrastructure (CNKI), Wanfang, and China Science and Technology Journal Database (VIP) using the search terms TCF7L2, transcription factor 7 like 2 protein, T-cell-specific transcription factor 4, T cell specific transcription factor 4, T cell transcription factor 4, TCF7L2 transcription factor, transcription factor, TCF7L2, or t cell factor 4 as well as monogenic diabetes, (diabetes, pregnancy induced), pregnancy-induced diabetes, gestational diabetes, (diabetes mellitus, gestational), or gestational diabetes mellitus. We also searched the Chinese databases CNKI, Wanfang, and Weipu using “gestational diabetes” or “GDM” and “transcription factor 7 analog-2” or “TCF7L2” as search terms. All relevant articles published between the date the database was established and July 2022 were retrieved. The articles were sorted by the name of the first author, the publication date, and ethnicity of study participants, and genotype and allele distribution information of the cases and controls were extracted.

Inclusion criteria

The inclusion criteria were: (i) epidemiological studies on the correlation between TCF7L2 SNPs and GDM; (ii) well-designed cohort, case–control, case-cohort, and cross-sectional studies; (iii) studies containing sufficient genotype or allele frequency data; and (iv) all cases included in the studies are patients diagnosed with GDM based on the GDM diagnostic criteria established by the American International Association of the Diabetes and Pregnancy Study Groups and "Obstetrics and Gynecology (Ninth Edition)."

Exclusion criteria

The following publications were excluded: (i) abstracts, reviews, lectures, and dissertations; (ii) publications with incomplete or unavailable genotype or allele frequency data; (iii) studies reporting a combination of interventions; and (iv) studies reporting animal experiments.

Data extraction

Three researchers independently screened original publications based on the inclusion criteria. Disagreements were resolved through discussion or by a fourth researcher. The following data were extracted from the publications that met the inclusion criteria: the name of the first author, the date of publication, definitions and characteristics of the case and control groups, and the distribution and frequency of alleles and genotypes in the included case and control groups.

Evaluating the quality of publications

The Newcastle–Ottawa Scale (NOS) [12] was used to evaluate the quality of the screened publications. The final 39 articles included case and control groups. Articles with quality scores of 0–4 were classified as low-quality studies, while those with scores of 5–8 were classified as high-quality studies.

Statistical analysis

Stata16.0 software was used to calculate odds ratios (OR) to represent the combined effect size index. Correlations between genotypes, alleles, and GDM were analyzed using different models. Heterogeneity between studies was tested before merging the results. Where there was large heterogeneity between studies, subgroup analysis was used to explore the source of heterogeneity. Heterogeneity between studies was quantitatively analyzed and the I2 value calculated. Where p < 0.1 or I2 > 40%, a random effect model was selected to calculate the combined OR value and the 95% confidence interval (Cl), otherwise, a fixed effect model was used. To evaluate the sensitivity of the analyses and to test the stability and reliability of the meta-analysis results, individual studies were excluded in turn and the combined OR and 95% Cl were recalculated. To avoid false positives due to multiple comparisons, pooled effect size values were corrected using the false discovery rate. Begg's and Egger's tests were used to determine publication bias [13]. The test level was set at α = 0.05.

Results

Description of the included studies

A total of 25 publications written in Chinese and 936 publications written in English were retrieved using the specified search terms. Duplicated publications were removed, leaving 16 articles written in Chinese and 799 articles written in English. Based on the exclusion and inclusion criteria, 644 articles were excluded after reading the abstracts. Three articles with NOS scores < 5 and three articles with incomplete data were excluded after reading the full text. A total of 39 articles reporting data on 8,795 cases and 16,290 controls were included in the final analysis. The process and outcome of database screening are shown in Fig. 1. Basic information on each included article is given in Table 1, while case data is provided in Table 2.

Fig. 1
figure 1

PRISMA flowchart

Full size image
Table 1 Characteristics of publications included in this study
Full size table
Table 2 TCF7L2 and CAPN10 allele distributions in gestational diabetes mellitus cases and controls
Full size table

Meta-analysis results

Association between rs7903146 and GDM

A total of 26 articles reporting data on 6,650 cases and 13,545 controls were included in this analysis. rs7903146 was the most common SNP in TCF7L2. Analysis using five gene models showed that polymorphisms in this locus were associated with the incidence of GDM in African, Asian, American, and Pacific populations.

Over-dominant gene model (CC + TT/CT) analysis showed very high heterogeneity (I2 = 81.8% P < 0.0001). Subgroup analysis conducted to reduce heterogeneity showed that rs7903146 was associated with the incidence of GDM in European (OR = 0.72, 95% CI: 0.65–0.86), American (OR = 0.61, 95% CI: 0.48–0.77), Asian (OR = 0.61, 95% CI: 0.48–0.77), African (0.73, 95% CI: 0.61–0.87), and Pacific (OR = 0.01, 95% CI: 0.00–0.02) populations (Fig. 2).

Fig. 2
figure 2

Forest plot of the relationship between GDM and rs 7,903,146 SNP under overdominance model. GDM, gestational diabetes mellitus; SNP, single nucleotide polymorphisms. Black diamonds denote the odds ratio

Full size image

Allelic model (T/C) subgroup analysis identified more mutations in this locus in the European (OR = 0.85, 95% CI: 0.78–0.9), African (OR = 2.68, 95% CI: 1.83–3.91), American (OR = 0.64, 95% CI: 0.53–0.78), Asian (OR = 0.60, 95% CI: 0.50–0.70), and Pacific (OR = 0.11, 95% CI: 0.07–0.18) populations (Fig. 3). The rs7903146 SNP variant, which is a risk factor for people from America, Asia, and the Pacific, may be a protective factor for Africa.

Fig. 3
figure 3

Forest plot of the relationship between GDM and rs 7,903,146 SNP under allelic model. GDM, gestational diabetes mellitus; SNP, single nucleotide polymorphisms. Black diamonds denote the odds ratio

Full size image

Subgroup analysis based on the co-dominant gene model (TT/CC) showed that the polymorphism was associated with GDM pathogenesis in the European (OR = 0.66, 95% CI: 0.57–0.77), African (OR = 8.90, 95% CI: 3.75–21.12), American (OR = 0.56, 95% CI: 0.34–0.97), Asian (OR = 0.35, 95% CI: 0.22–0.56), and Pacific (OR = 0.12, 95% CI: 0.03–0.45) populations (Fig. 4). The co-dominant gene model (CT/CC) subgroup analysis showed that this polymorphism was associated with the incidence of GDM in the Pacific or Asian populations (OR = 22.45, 95% CI: 5.94–84.51), as shown in Fig. 5.

Fig. 4
figure 4

Forest plot of the relationship between GDM and rs 7,903,146 SNP under codominant model. GDM, gestational diabetes mellitus; SNP, single nucleotide polymorphisms. Black diamonds denote the odds ratio

Full size image
Fig. 5
figure 5

Forest plot of the relationship between GDM and rs 7,903,146 SNP under codominant model 2. GDM, gestational diabetes mellitus; SNP, single nucleotide polymorphisms. Black diamonds denote the odds ratio

Full size image

The dominant gene model (CT + TT/CC) subgroup analysis suggested that the SNP was associated with GDM pathogenesis in the European (OR = 0.81, 95% CI: 0.71–0.93 =) and African (OR = 3.00, 95% CI: 1.59–5.64) populations (Fig. 6).

Fig. 6
figure 6

Forest plot of the relationship between GDM and rs 7,903,146 SNP under dominant model. GDM, gestational diabetes mellitus; SNP, single nucleotide polymorphisms. Black diamonds denote the odds ratio

Full size image

Recessive gene model (TT/CC + CT) analysis showed that the polymorphism was associated with the onset of GDM in the European (OR = 0.69, 95% CI: 0.60–0.78), African (OR = 5.67, 95% CI: 2.74–11.74), American (OR = 0.57, 95% CI: 0.45–0.72), Asian (OR = 0.62, 95% CI: 0.51–0.75), and Pacific (OR = 0.01, 95% CI: 0.00–0.02) populations (Fig. 7).

Fig. 7
figure 7

Forest plot of the relationship between GDM and rs 7,903,146 SNP under recessive model. GDM, gestational diabetes mellitus; SNP, single nucleotide polymorphisms. Black diamonds denote the odds ratio

Full size image

Association between rs12255372 and GDM

There were 30 articles on rs12255372 (15 written in Chinese and 15 written in English) reporting data on 3,312 cases and 3,535 controls. The SNP was associated with the incidence of GDM (OR = 0.91, 95% CI: 0.52–1.30) in the African population based on analysis using all gene models apart from the over-dominant gene model. In the Asian population, this polymorphism was associated with the incidence of GDM (OR = 0.82, 95% CI: -1.50–0.13) only in the over-dominant gene model analysis (Figs. 8, 9, 10, 11, 12, 13).

Fig. 8
figure 8

Forest plot of the relationship between GDM and rs 12,255,372 SNP under over-dominant model. GDM, gestational diabetes mellitus; SNP, single nucleotide polymorphisms. Black diamonds denote the odds ratio

Full size image
Fig. 9
figure 9

Forest plot of the relationship between GDM and rs 12,255,372 SNP under allelic model. GDM, gestational diabetes mellitus; SNP, single nucleotide polymorphisms. Black diamonds denote the odds ratio

Full size image
Fig. 10
figure 10

Forest plot of the relationship between GDM and rs 12,255,372 SNP under codominant model 2. GDM, gestational diabetes mellitus; SNP, single nucleotide polymorphisms. Black diamonds denote the odds ratio

Full size image
Fig. 11
figure 11

Forest plot of the relationship between GDM and rs 12,255,372 SNP under codominant model 1. GDM, gestational diabetes mellitus; SNP, single nucleotide polymorphisms. Black diamonds denote the odds ratio

Full size image
Fig. 12
figure 12

Forest plot of the relationship between GDM and rs 12,255,372 SNP under dominant model. GDM, gestational diabetes mellitus; SNP, single nucleotide polymorphisms. Black diamonds denote the odds ratio

Full size image
Fig. 13
figure 13

Forest plot of the relationship between GDM and rs 12,255,372 SNP under recessive model. GDM, gestational diabetes mellitus; SNP, single nucleotide polymorphisms. Black diamonds denote the odds ratio

Full size image

Association between rs7901695 and GDM

There were eight articles on rs7901695 written in Chinese and English, reporting data on 1,515 cases and 3,157 controls. All five gene models showed that this SNP was associated with the onset of GDM (OR = 0.69, 95% CI: 0.61–0.79). All subgroup-based gene model analyses of this SNP showed associated incidence of GDM in the European population. Further, this SNP was associated with the incidence of GDM in the American population based on the allelic model (OR = 0.56, 95% CI: 0.38–0.82), the co-dominant genes model (R = 0.32, 95% CI: 0.14–0.74), and the recessive gene model (OR = 0.45, 95% CI: 0.25–0.81) (Figs. 14, 15, 16, 17, 18).

Fig. 14
figure 14

Forest plot of the relationship between GDM and rs 7,901,695 SNP under allelic model. GDM, gestational diabetes mellitus; SNP, single nucleotide polymorphisms. Black diamonds denote the odds ratio

Full size image
Fig. 15
figure 15

Forest plot of the relationship between GDM and rs 7,901,695 SNP under implicit model. GDM, gestational diabetes mellitus; SNP, single nucleotide polymorphisms. Black diamonds denote the odds ratio

Full size image
Fig. 16
figure 16

Forest plot of the relationship between GDM and rs 7,901,695 SNP under overdominant model. GDM, gestational diabetes mellitus; SNP, single nucleotide polymorphisms. Black diamonds denote the odds ratio

Full size image
Fig. 17
figure 17

Forest plot of the relationship between GDM and rs 7,901,695 SNP under dominant model. GDM, gestational diabetes mellitus; SNP, single nucleotide polymorphisms. Black diamonds denote the odds ratio

Full size image
Fig. 18
figure 18

Forest plot of the relationship between GDM and rs 7,901,695 SNP under codominant model. GDM, gestational diabetes mellitus; SNP, single nucleotide polymorphisms. Black diamonds denote the odds ratio

Full size image

Association between rs290487 and GDM

There were five articles on rs290487 written in Chinese and English, reporting data on 1,361 cases and 1,361 controls. The allele model (OR = 0.73, 95% CI: 0.66–0.82) and recessive gene model (OR = 0.80, 95% CI: 0.69–0.93) demonstrated that this SNP was associated with the incidence of GDM (Figs. 19 and 20).

Fig. 19
figure 19

Forest plot of the relationship between GDM and rs 290,487 SNP under allelic model. GDM, gestational diabetes mellitus; SNP, single nucleotide polymorphisms. Black diamonds denote the odds ratio

Full size image
Fig. 20
figure 20

Forest plot of the relationship between GDM and rs 290,487 SNP under recessive model. GDM, gestational diabetes mellitus; SNP, single nucleotide polymorphisms. Black diamonds denote the odds ratio

Full size image

Association between rs2975760 and GDM

Rs2975760 is the most studied SNP in CAPN10. A total of five studies on rs2975760, reporting data on 887 cases and 1,913 controls, were included in this analysis. Recessive gene model analysis demonstrated that this mutation was associated with the onset of GDM (OR = 1.70, 95% CI: 1.16–2.50) (Fig. 21).

Fig. 21
figure 21

Forest plot of the relationship between GDM and rs 2,975,760 SNP under recessive model. GDM, gestational diabetes mellitus; SNP, single nucleotide polymorphisms. Black diamonds denote the odds ratio

Full size image

Discussion

In recent years, despite the large number of studies exploring the correlation between SNPs and GDM, there is a notable scarcity of high-quality meta-analyses within this domain.

Developments in molecular technologies have led to increased efforts to identify the genes associated with susceptibility to GDM as well as to develop molecular-based strategies for preventing and treating the disease [51]. The underlying mechanism of action of TCF7L2 and CAPN10 in the pathogenesis of GDM remains unclear [52], although TCF7L2 and CAPN10 may directly affect the function of pancreatic islet β-cells [53], resulting in decreased secretion of insulin and glucagon-like peptides, and subsequently leading to increased production of endogenous glucose [54]. Some studies have suggested that TCF7L2, an important gene associated with the pathogenesis of type 2 diabetes mellitus (T2DM), may interfere with GLP1, reduce the expression of GLP1R and glucose-dependent insulinotropic polypeptide/gastrointestinal inhibitory peptide (GIP) receptor (GIP-R), and inhibit β cell function by stimulating insulin secretion (in conjunction with GIP). TCF7L2 may also lead to abnormal insulin conversion and activation of the Wnt signaling pathway, leading to T2DM [55, 56].

GDM has a similar pathogenesis to T2DM [57], and some studies have found that the two may also share genetic characteristics [58]. TCF7L2 polymorphisms are also associated with susceptibility to GDM [59]. Studies in Sweden, Poland, and other European countries have suggested that the TCF7L2 SNPs rs7903146 ​​and rs12255372 are associated with GDM pathogenesis [57]. Similar findings have been reported in Asian countries such as South Korea and China [43]. However, reports on the association between rs7901695 and rs290487 and GDM are still inconclusive, similar to the results of the subgroup analyses conducted in this study. The TCF7L2 ​​SNP rs7903146 is a risk factor for GDM in Asian, European, American, African, and Pacific people. Unlike in previous studies, the subgroup analysis in our study showed that rs12255372, rs7901695, and rs290487 have different associations with GDM in different populations. The rs7903146 SNP variant, identified as a risk factor for individuals America, Asia, and the Pacific, may exhibit a protective role in the African population. This indicates that morbidity was associated with the onset of GDM, and a specific variant may serve as a protective or risk factor, depending on the population. For a more in-depth analysis of the reasons, it is crucial to consider the concept of genetic variability across populations. Populations demonstrate genetic diversity due to historical, geographical, and demographic factors. Therefore, a genetic variant that appears to confer protection in one population may not necessarily have the same effect in another. The interplay between genetic makeup and environmental factors can result in varying phenotypic expressions, making it essential to account for population-specific nuances in genetic studies [58]. Furthermore, the multifactorial nature of many health conditions necessitates a nuanced understanding of the role of specific variants. A variant might act as a protective factor in the presence of certain environmental conditions or in combination with other genetic factors. Conversely, the same variant may manifest as a risk factor under different circumstances. This emphasizes the need for comprehensive analyses that consider gene–gene and gene-environment interactions to unravel the intricate relationships between genetics and health [59]. Additionally, the observed variability in the effects of genetic variants might be attributed to gene flow and evolutionary processes. Migration and historical events can lead to the dissemination of certain genetic variants in specific populations, influencing their prevalence and impact on health outcomes [43].

Compared with TCF7L2, CAPN10 has only few studies analyzing its association with GDM. rs2975760 and rs5030952 have been associated with GDM pathogenesis [46, 48, 49]. However, only rs2975760 was associated with the onset of GDM in this study.

This study had some limitations. First, we only retrieved articles written in Chinese and English. Hence, we may have missed articles written in other languages. Additionally, the meta-analysis might be susceptible to publication bias, where studies with positive results are more likely to be published than those with null findings. This can lead to an overestimation of the true effect size. Addressing these limitations would strengthen the robustness and applicability of the meta-analysis, providing a more comprehensive and nuanced understanding of the relationship between TCF7L2 and CAPN10 gene polymorphisms and GDM across different geographical regions.

Conclusions

We searched eight Chinese and English databases: Cochrane, Elton B. Stephens. Company (EBSCO), Embase, Scopus, Web of Science, China National Knowledge Infrastructure (CNKI), Wanfang, and China Science and Technology Journal Database and retrieved all relevant articles published between the inception of the database and July 2022. As such, we found that rs7903146, rs12255372, rs7901695, rs290487, and rs2975760 were associated with the incidence of GDM in different populations.