Skip to Content
MDPIMDPI
DiagnosticsDiagnostics
PDF
  • Article
  • Open Access

25 February 2026

Association of Genetic Polymorphisms with Gestational Diabetes in a Kazakh Population: A Case–Control Study

,
,
,
,
,
,
,
and
1
National Center of Endocrinology and Diabetes, JSC Research Institute of Cardiology and Internal Diseases, Almaty 050012, Kazakhstan
2
Scientific Department of the Center for Molecular Medicine, Center for Molecular Medicine, Almaty 050026, Kazakhstan
3
Department of Strategy and Science, JSC Scientific Center of Obstetrics, Gynecology and Perinatology, Almaty 050020, Kazakhstan
4
JSC Research Institute of Cardiology and Internal Diseases, Almaty 050000, Kazakhstan
Diagnostics2026, 16(5), 663;https://doi.org/10.3390/diagnostics16050663
This article belongs to the Section Pathology and Molecular Diagnostics
Version Notes
Order Reprints

Abstract

Background: Gestational diabetes mellitus (GDM) poses a growing public health challenge worldwide due to its increasing prevalence, associated pregnancy complications, and long-term metabolic risks for both mothers and offspring. Genetic factors are known to contribute to GDM susceptibility, yet little is known about their relevance in ethnic Kazakh population. The primary objective of this study was to evaluate associations between selected candidate SNPs involved in β-cell function and the risk of GDM in a Kazakh cohort. Secondary objectives included the assessment of potential gene–gene interactions. Methods: We conducted a case–control study among 365 pregnant Kazakh women. Of these, 217 were diagnosed with GDM, and 148 had normal glucose tolerance. Clinical and genealogical data were collected. Eight candidate SNPs that were previously associated with GDM or glucose metabolism were genotyped. Logistic regression was used to assess associations between SNPs and GDM risk. Gene–gene interactions were evaluated using multifactor dimensionality reduction (MDR). Results: In univariate analysis, MTNR1B rs10830963 demonstrated a statistically significant association under the additive model (OR 0.61, 95% CI 0.42–0.89), indicating a potential protective effect of the C allele. However, this association was not statistically significant after multivariable adjustment (adjusted OR 0.58, 95% CI 0.32–1.03) and correction for multiple testing. In the adjusted analysis, TCF7L2 rs7903146 showed a significant association under the dominant model (adjusted OR 2.29, 95% CI 1.01–5.46); however, this finding did not remain significant following FDR correction. MDR analysis showed that the best two-locus model included IGF2BP2 rs4402960 and CDKAL1 rs7754840 (CVC = 6/10; testing accuracy = 0.558; permutation p < 0.001). The most stable interaction was observed for the three-locus model comprising IGF2BP2 rs4402960, MTNR1B rs10830963, and PPARG rs1801282 (CVC = 9/10; testing accuracy = 0.576; permutation p < 0.001). Conclusions: The findings suggest that common variants in IGF2BP2, CDKAL1, MTNR1B, TCF7L2, PPARG, and GCK do not exert strong individual effects on GDM susceptibility in this cohort of ethnic Kazakh women. Instead, the results are more consistent with a modest polygenic architecture characterized by small effect sizes and possible weak gene–gene interactions. MDR analysis identified statistically significant interaction models; however, their limited predictive accuracy indicates that these findings should be interpreted as exploratory.

1. Introduction

Gestational diabetes mellitus (GDM), defined as glucose intolerance first identified during pregnancy, represents an escalating global public health concern [1]. According to the International Diabetes Federation, in 2024, approximately one in six live births worldwide was affected by GDM [2], with substantial regional variation driven by differences in ethnicity, maternal age, body mass index (BMI), and diagnostic criteria. GDM is associated with a wide range of adverse maternal and neonatal outcomes, including preeclampsia, fetal macrosomia, neonatal hypoglycemia, preterm birth, cesarean delivery, and increased risk of neonatal intensive care unit (NICU) admission [3,4,5]. Beyond pregnancy, women with GDM face a markedly increased risk of type 2 diabetes mellitus (T2DM) [6,7], while offspring exposed to maternal hyperglycemia are predisposed to obesity, impaired glucose tolerance, and cardiovascular disease later in life [8].
The economic burden of GDM is substantial due to increased healthcare utilization, intensive prenatal monitoring, and long-term disease consequences for both mother and child. Evidence from large-scale economic evaluations highlights the financial implications of both early detection and delayed management. The TOBOGM trial, a multicenter randomized controlled study conducted between 2017 and 2022 across 17 hospitals in Australia, Austria, India, and Sweden, demonstrated that early diagnosis and treatment of GDM not only improved clinical outcomes but also showed a tendency toward cost savings compared with standard care, with an estimated mean cost difference of −$1373 (95% CI: −$3749 to $642) [9]. Similarly, population-level data from China indicate a substantial societal impact of GDM, with the annual economic burden estimated at approximately ¥19.36 billion (US $5.59 billion) in 2015 [10]. In addition to direct and indirect costs, GDM was associated with considerable health losses, corresponding to an estimated 260,000 quality-adjusted life years (QALYs) lost annually [10]. Together, these findings underscore the significant economic and public health consequences of GDM and highlight the importance of effective prevention, early detection, and targeted intervention strategies.
Preventive strategies for GDM have predominantly targeted modifiable risk factors, including lifestyle and metabolic interventions implemented before or during pregnancy. Identifying risk factors that can be modified is essential for lowering the likelihood of developing GDM and minimizing related health complications [11]. Evidence from systematic reviews indicates that interventions based on dietary modification, increased physical activity, and combined diet–exercise programs, as well as pharmacological and nutritional approaches such as metformin and myoinositol supplementation, are associated with a reduced incidence of GDM compared with standard care [12,13]. However, despite these preventive efforts, the broader burden of dysglycemia will continue to rise. Prediction models show that by 2050, in women, the global incidence, disability-adjusted life years (DALYs), and mortality associated with type 2 diabetes (T2D) will all increase over time—particularly among younger age groups—despite declining mortality trends in higher Socio-demographic Index (SDI) countries [14].
Genetic susceptibility contributes to interindividual variability in pancreatic β-cell compensation during pregnancy. Genome-wide association studies (GWASs) indicate that many loci associated with GDM overlap with those implicated in type 2 diabetes, particularly genes regulating insulin secretion rather than peripheral insulin sensitivity. This suggests that genetically determined β-cell vulnerability plays a central role in GDM pathogenesis. Single-nucleotide polymorphisms (SNPs) in genes such as IGF2BP2, CDKAL1, MTNR1B, TCF7L2, and PPARG have been consistently associated with GDM and T2DM in diverse populations [15,16]. For example, TCF7L2 rs7903146 has been linked to increased GDM risk in European cohorts [17], and MTNR1B rs10830963 has been associated with fasting glucose and GDM in Chinese populations [18]. Similarly, CDKAL1 rs7754840 and PPARG rs1801282 variants have been implicated in GDM susceptibility in multiple ethnic groups [19,20]. However, no significant association between IGF2BP2 rs4402960 and GDM was observed in a recent meta-analysis [21].
Research on the genetic epidemiology of GDM in Central Asia, particularly in Kazakhstan, remains limited, despite the rising burden of both DM and GDM in the country. Kazakhstan represents a historically admixed population shaped by Turkic, Mongolian, Slavic, and Central Asian ancestries [22]. Such genetic heterogeneity may influence allele frequencies, linkage disequilibrium patterns, and gene–environment interactions relevant to metabolic disease. Although numerous single-nucleotide polymorphisms (SNPs) have been associated with GDM and type 2 diabetes (T2D) in global studies, there is limited evidence regarding their relevance in the Kazakh population. Recent studies have begun to address this gap. For example, a 2024 case–control study by Sikhayeva et al. examined ADIPOQ gene polymorphisms and found significant associations with T2D and obesity among Kazakh individuals, suggesting a genetic basis for metabolic risk in this population [23]. Additional investigations into PPARG, TCF7L2, and ANRIL gene variants have revealed associations with prediabetes and coronary artery disease among Kazakh patients with diabetes, underscoring the importance of population-specific genetic research [24].
Notably, a 2022 study by Svyatova et al. examined allele frequencies of eight SNPs across seven genes related to glucose metabolism and insulin signaling in healthy Kazakh individuals [25]. The study reported high frequencies of risk alleles in MTNR1B (rs10830963), TCF7L2 (rs7903146), and GCK (rs4607517), suggesting potential genetic susceptibility to GDM in this population. A subsequent genome-wide study analyzed TCF7L2 (rs7903146) and PPARG (rs1801282) variants in 1800 ethnically Kazakh individuals and found that the protective G allele of PPARG occurred at a frequency of 13.8%, while the risk-associated T allele of TCF7L2 was present in 15.2% of the sample, indicating relevance for prediabetes risk [26]. Both studies have described allele frequencies of metabolic risk variants in healthy Kazakh cohorts; however, detailed analyses examining genotype–phenotype associations in GDM are lacking.
We hypothesized that variants in genes regulating pancreatic β-cell function would be associated with altered risk of GDM in a Kazakh cohort and that population substructure might influence genotype distribution. The primary objective of this study was to evaluate associations between selected candidate SNPs involved in β-cell function and the risk of GDM in a Kazakh cohort. Secondary objectives included the assessment of potential gene–gene interactions.

2. Materials and Methods

2.1. Study Design and Study Population

This study employed a case–control design to investigate clinical, genealogical, and genetic factors associated with GDM. Participants were recruited from obstetric clinics across four regions in Kazakhstan—Aktobe, Almaty, Almaty Region, and Semey—from April 2022 to June 2023 (15 months). A total of 365 pregnant women were enrolled between 24 and 28 weeks of gestation. Among them, 217 women were diagnosed with GDM based on the World Health Organization (WHO) 2013 criteria using a 75 g oral glucose tolerance test (OGTT), and 148 women with normal glucose tolerance served as non-GDM controls. Written informed consent was obtained from all participants prior to enrollment. Pregnant women were recruited at 24–28 weeks’ gestation prior to the classification of GDM status. Diagnosis was established according to the WHO 2013 criteria based on a 75 g oral glucose tolerance test (OGTT). Participants were consecutively enrolled, and no matching was performed. No a priori SNP-specific power calculation was performed; therefore, the study should be considered exploratory.
Inclusion criteria for both groups included age between 18 and 45 years, Kazakh ethnicity, singleton pregnancy, and availability of complete clinical and biochemical data. Exclusion criteria included pre-existing diabetes mellitus, chronic kidney or liver disease, autoimmune disorders, or other significant metabolic conditions. All participants provided written informed consent prior to enrollment. The study was approved by the Institutional Ethics Committee of the Asfendiyarov Kazakh National Medical University (Protocol No. 5 (111) dated 28 April 2021) and conducted in accordance with the Declaration of Helsinki.

2.2. Genealogical Data

In the Kazakh context, such self-reported genealogical identity is considered highly reliable because genealogical knowledge (shezhire) is deeply embedded in cultural and social life. From childhood, Kazakhs are taught to recite their zheti ata (“seven ancestors”), which encompasses paternal lineage (ata), tribe (ru), and socio-territorial subdivision (zhuz) [27]. This tradition serves both as a safeguard against close-kin marriage and as a marker of social cohesion and ancestral geography.
Genealogical records in the form of shezhire exist in both public and private formats. Publicly accessible shezhire sources typically provide extensive historical lineage information beginning at the zhuz level and extending through ru and their descendants, whereas privately maintained shezhire records often contain more detailed and up-to-date information on recent generations [28].
Historian Asfendiarov argued that the formation of the Kazakh zhuzes reflected distinct economic, cultural, and historical conditions shaped by the natural division of Kazakhstan into the Semirechie, Western, and Central regions [29]. Kishi zhuz (Junior) is concentrated in Western Kazakhstan. Orta zhuz (Middle) is dominant in the Central regions. Uly zhuz (Senior) is historically centered in southern Kazakhstan (Semirechie) [30].
In addition to the three main zhuz, the genealogical information includes Tore and Kozha, two historically distinct lineages [31]. Tore genealogies are traditionally considered descendants of Genghis Khan, while those of Kozha trace their ancestry to Islamic religious scholars. These groups do not belong to the formal zhuz classification and were therefore recorded and analyzed separately. Importantly, zhuz affiliation in this study is interpreted as a sociocultural indicator of potential population substructure rather than a direct measure of genetic ancestry. Because genome-wide ancestry-informative markers were not assessed, residual population stratification cannot be excluded. Zhuz affiliation was not included as a covariate in primary logistic regression models due to limited sample size and multicollinearity with the region.

2.3. Genealogical Data Collection

Participants provided genealogical information via structured questionnaires at enrollment. They self-reported the zhuz (socio-territorial subdivision) affiliation of their maternal and paternal grandparents. In this study, the inclusion of participants from four geographically distinct regions, Almaty, Almaty Region (Semirechie), Semey (East and Central), and Aktobe (Western), was strategically designed to reflect the diversity of genealogical backgrounds and ensure a representative sampling of all major zhuz affiliations. Almaty city, as a metropolitan hub, is demographically diverse and includes representatives from all zhuz.

2.4. Clinical Measurements

Clinical assessments were conducted at the time of enrollment. Anthropometric measurements included height, weight, and body mass index (BMI), which was calculated as weight in kilograms divided by height in meters squared (kg/m2).

2.5. DNA Extraction and Genotyping

The eight candidate SNPs included in this study were selected based on prior genome-wide association signals in GDM or T2D, biological plausibility in β-cell or insulin signaling pathways, and reported minor allele frequencies sufficient to permit analysis in Central Asian populations. Preference was given to variants with replicated associations across multiple ethnic cohorts. A total of eight SNPs were selected for genotyping based on their previously reported associations with insulin signaling and glucose homeostasis. These included two variants in Insulin-Like Growth Factor 2 mRNA-Binding Protein 2 (IGF2BP2; rs1470579, TaqMan assay ID: C___2165184_10; rs4402960, C___2165199_10), and one SNP each in Cyclin-Dependent Kinase 5 Regulatory Subunit-Associated Protein 1-Like 1 (CDKAL1; rs7754840, C__29246232_10), Insulin Receptor Substrate 1 (IRS1; rs1801278, C___2384392_20), Melatonin Receptor 1B (MTNR1B; rs10830963, C___3256858_10), Transcription Factor 7-Like 2 (TCF7L2; rs7903146, C__29347861_10), Peroxisome Proliferator-Activated Receptor Gamma (PPARG; rs1801282, C___1129864_10), and Glucokinase (GCK; rs4607517, C__27917691_10).
Peripheral blood samples were collected from the cubital vein into EDTA-containing vacutainer tubes. Genomic DNA was extracted using magnetic M-PVA bead separation on the Prepito® automated nucleic acid extraction platform (PerkinElmer, Wallac, Finland), with the Prepito® DNA CytoPure reagent kit.
DNA purity was assessed by measuring absorbance ratios at 260/280 nm using a NanoDropLite UV spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). Samples with absorbance ratios around 1.8 were considered pure. The DNA was suspended in nuclease-free water and stored at –40 °C until further analysis. Genotyping was performed within ten days of extraction using a real-time polymerase chain reaction (PCR) on the StepOnePlus™ system (Applied Biosystems, Foster City, CA, USA), with TaqMan® allele discrimination assays (Thermo Fisher Scientific, Waltham, MA, USA). All assays were conducted according to the manufacturer’s protocol using standard TaqMan reagents and TE buffer (pH 8.0, 10 mM Tris-HCl, 1 mM EDTA; Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA). The quality of DNA used for genotyping was confirmed by purity ratios ranging from 1.7 to 2.0. High-quality DNA samples demonstrated clear clustering on allele discrimination plots. Any outliers that fell outside defined clusters were excluded from analysis, and DNA extraction and PCR were repeated. As part of quality control, 5% of samples were randomly re-genotyped, confirming 100% concordance and ensuring genotyping reliability. Genotyping call rates exceeded 95% for all SNPs. Minor allele frequencies (MAFs) were calculated in controls. Hardy–Weinberg Equilibrium (HWE) was assessed in controls using the exact test. Variables deviating from HWE were excluded from further analysis.

2.6. Statistical Analysis Plan

All statistical analyses were conducted using R version 4.5.1 (13 June 2025) [32] within the RStudio environment (version 2025.9.0.387) [33]. Statistical significance was set at a two-sided p-value of less than 0.05 for all comparisons. Descriptive analyses were first performed to summarize the baseline characteristics of the study population. Continuous variables were assessed for normality using the Shapiro–Wilk test. Non-normally distributed variables were reported as medians with interquartile ranges (IQRs). Categorical variables were described using frequencies and percentages.
To compare baseline characteristics between participants with GDM and non-GDM controls, appropriate statistical tests were selected based on data type and distribution. For continuous variables, comparisons were made using the non-parametric Mann–Whitney U test. Categorical variables were analyzed using Pearson’s chi-square test or Fisher’s exact test, as appropriate.
To examine the association between selected single-nucleotide polymorphisms (SNPs) and the risk of GDM, both univariate and multivariate logistic regression analyses were conducted. Crude odds ratios (ORs) and adjusted odds ratios (AORs), along with 95% confidence intervals (CIs), were estimated. Multivariate models were adjusted for potential confounders, including maternal body mass index (BMI), education, region, employment status, and parity. To account for multiple testing across SNPs, false discovery rate (FDR) correction was applied using the Benjamini–Hochberg procedure.
To assess potential epistatic (gene–gene) interactions among the studied SNPs, multifactor dimensionality reduction (MDR) analysis was performed in R [34]. Ten-fold cross-validation was used to validate model stability, and the best models were selected based on cross-validation consistency (CVC), testing accuracy, and permutation-based p-values. Post hoc statistical power was estimated based on the observed sample size (217 cases and 148 controls), case–control ratio, and typical minor allele frequencies. Assuming a two-sided α of 0.05 under an additive genetic model, the study had approximately 80% power to detect odds ratios ≥ 1.6, whereas power was substantially lower for detecting modest effect sizes (OR ≤ 1.3). All data analyzed in the present study are provided in Supplementary Table S1.

3. Results

Original Study Results

Table 1 summarizes the baseline characteristics of participants with GDM and non-GDM controls. The median age of GDM patients was 32.0 years (IQR: 27.0–35.0), compared to 30.0 years (IQR: 27.0–35.0) among controls, with no significant difference observed (p = 0.59). GDM patients had significantly higher BMI values than controls, with 29.49 [24.52, 34.08] and 25.34 [21.77, 31.2], respectively (p = 0.0187). Significant differences were found between groups in the distribution of genealogical zhuz affiliations for both maternal and paternal grandparents (all p-values < 0.001). A higher proportion of GDM patients belonged to the Kishi zhuz, whereas controls were more frequently associated with the Orta zhuz. Marital status did not differ significantly between groups, with the majority of both GDM patients (94%) and controls (97%) being married (p = 0.242). Educational attainment, however, showed a statistically significant difference (p = 0.038), with a higher proportion of controls having a bachelor’s degree or higher (66%) compared to GDM patients (54%). Regional distribution varied markedly (p < 0.001). Participants with GDM were more likely to reside in Aktobe (44%) or the Almaty Region (27%), whereas controls were predominantly from Semey (63%). Employment status also differed significantly (p < 0.001), with a higher percentage of controls currently not employed (69%) compared to GDM participants (35%). Conversely, a greater proportion of GDM patients reported being employed (40% vs. 18%). Family history of diabetes did not differ significantly between groups (p = 0.062), nor did gravidity (median gravidity = 2.0 for both groups; p = 0.47). However, parity showed a significant difference (p = 0.009), with a higher proportion of controls being multiparous (89% vs. 78%). Smoking status was comparable across groups (p = 0.090), with the majority of both groups being non-smokers.
Table 1. Baseline characteristics of study participants.
Table 2 presents the results of the HWE analysis conducted in the control group using the exact test. Genotype distributions for all investigated SNPs were consistent with HWE expectations (exact p ≥ 0.05), indicating no evidence of significant deviation from equilibrium and supporting the reliability of genotyping for these variants. The only exception was IRS1 rs1801278, which significantly deviated from HWE (p < 0.0001) and was therefore excluded from further association analyses.
Table 2. Hardy–Weinberg Equilibrium in control group (exact test).
Table 3 presents the associations between selected SNPs and the risk of GDM, based on univariate and multivariate logistic regression analyses adjusted for BMI, education, region, employment status, and parity. In univariate analysis, MTNR1B rs10830963 demonstrated a statistically significant association under the additive model (OR 0.61, 95% CI 0.42–0.89), indicating a potential protective effect of the C allele. However, this association was attenuated and no longer statistically significant after multivariable adjustment (adjusted OR 0.58, 95% CI 0.32–1.03) and correction for multiple testing. In the adjusted analysis, TCF7L2 rs7903146 showed a significant association under the dominant model (adjusted OR 2.29, 95% CI 1.01–5.46); however, this finding did not remain significant following FDR correction. No statistically significant associations were observed for the remaining SNPs under additive, dominant, or recessive genetic models after adjustment for covariates and multiple testing.
Table 3. Univariate and multivariate logistic regression of SNPs with gestational diabetes.
Table 4 presents the top-performing gene–gene interaction models identified using 10-fold cross-validation. The best two-locus model included IGF2BP2 rs4402960 and CDKAL1 rs7754840 (CVC = 6/10; testing accuracy = 0.558; permutation p < 0.001). The most stable interaction was observed for the three-locus model comprising IGF2BP2 rs4402960, MTNR1B rs10830963, and PPARG rs1801282 (CVC = 9/10; testing accuracy = 0.576; permutation p < 0.001). Although statistically significant, the modest testing accuracy suggests limited predictive capacity.
Table 4. Best MDR gene–gene interaction models (10-fold CV).

4. Discussion

This study did not identify any SNP that remained significantly associated with GDM after correction for multiple testing. Although MTNR1B rs10830963 demonstrated a statistically significant association under the additive model in univariate analysis, this association did not remain significant after adjustment. Similarly, TCF7L2 rs7903146 showed a nominally significant association under the dominant model in adjusted analysis; however, this finding also did not persist after correction for multiple testing. No statistically robust independent association was observed for other investigated variants.
These findings suggest that common variants in IGF2BP2, CDKAL1, MTNR1B, TCF7L2, PPARG, and GCK do not exert strong independent effects on GDM susceptibility in this cohort of Kazakh women. The suggestive signal observed for MTNR1B rs10830963 is biologically plausible, given the established role of MTNR1B in glucose homeostasis and β-cell function [35]. Variants in MTNR1B have been associated with fasting glucose levels and type 2 diabetes risk in multiple populations [36,37]. Likewise, TCF7L2 rs7903146 is one of the most consistently replicated loci for type 2 diabetes [37,38]. The observed effect sizes were modest and confidence intervals were wide, consistent with limited statistical power to detect small genetic effects typical of complex metabolic disorders.
Beyond SNP analyses, MDR analysis suggested potential gene–gene interaction models involving IGF2BP2, CDKAL1, MTNR1B, PPARG, and GCK. While several interaction models demonstrated statistical significance in permutation testing, their overall testing accuracy was modest (≤0.58), indicating limited predictive capacity. Thus, these interaction findings should be regarded as exploratory rather than confirmatory.
Taken together, the updated results do not support a strong independent genetic determinant of GDM among the investigated candidate variants in this population. Instead, the data are more consistent with a polygenic architecture characterized by modest individual effects and possible weak epistatic interactions. Given the moderate sample size and the typical small effect sizes of common metabolic variants (OR ≈ 1.1–1.3), the study was likely underpowered to detect subtle associations after correction for multiple testing.
Ethnic background and population structure may further influence genetic associations. The Kazakh population has a complex genealogical structure shaped by historical migration patterns and zhuz affiliation. While genealogical differences between cases and controls were observed, multivariable regression models were not adjusted for zhuz affiliation due to substantial multicollinearity between zhuz and region. Our findings suggest that any potential relationship between ancestry-related structure and GDM risk requires further investigation in larger, well-powered cohorts with more balanced regional and genealogical representation.
Our results extend existing evidence. Previous case–control research in China reported no significant association between IGF2BP2 polymorphisms and GDM, a finding that was consistent with subsequent meta-analytic results showing no association across multiple genetic models [21]. In line with this evidence, a study from Poland involving women with gestational diabetes mellitus and those with normal glucose tolerance found no association between IGF2BP2 rs4402960 and rs11705701 variants and the risk of GDM [39]. A study conducted in Russia also found no association between the IGF2BP2 rs4402960 variant and the risk of GDM [40]. Genetic background, including ethnicity, plays an important role in those associations.
A prior study reported marked ethnic differences in genetic susceptibility to gestational diabetes mellitus. Among White women, GDM was associated with rs7901695 variants in TCF7L2, rs10830963 variants in MTNR1B, and rs780094 variants in GCKR, which are also linked to T2DM and fasting glucose regulation in non-pregnant populations [41]. In contrast, among African American women, an elevated risk was observed in individuals homozygous for the TSPAN8 rs7961581 C allele, whereas carriers of the JAZF1 rs864745 T allele demonstrated a reduced risk [41]. The protective association observed for the TT genotype compared with the GG and GT genotypes of the IGF2BP2 rs4402960 polymorphism in our study contrasts with the findings reported in previous studies. These discrepancies may be explained by ethnic-specific linkage disequilibrium patterns, gene–environment interactions, or differences in diagnostic criteria. The lack of significant associations for all investigated SNPs highlights the need for population-specific research and may reflect underlying genetic heterogeneity.
In Kazakhstan, projections indicate a decline in type 2 diabetes prevalence—from 3758 cases in 2020 to 857 by 2030—suggesting possible improvements in public health outcomes [32]. However, GDM continues to impose a significant and enduring economic and social burden, affecting both maternal and offspring health across generations. In this context, leveraging all available tools for early detection, risk stratification, and prevention of GDM is paramount. Although our study did not identify robust independent genetic associations after correction for multiple testing, the observed suggestive signals and exploratory interaction patterns highlight the complexity of GDM susceptibility in this population. Rather than supporting immediate clinical integration of specific genetic markers, our findings underscore the importance of cautious interpretation and the need for larger, well-powered studies before genetic information can be incorporated into screening strategies. By recognizing the genetic diversity and population substructure specific to Kazakhstan, particularly among ethnic Kazakh women, future research may better clarify whether modest polygenic effects or gene–environment interactions contribute meaningfully to GDM risk. The consideration of zhuz-related ancestral clustering as both a potential source of genetic insight and a source of confounding remains essential to advancing genomic research in populations with complex genealogical structures. At present, however, the clinical utility of the investigated SNPs for risk stratification appears limited. Future research should build on these findings by incorporating larger cohorts, longitudinal designs, and multi-omics approaches to further unravel the complex interplay between genetics, ancestry, and metabolic health during pregnancy.
Despite its strengths, this study has several limitations. The sample size was determined by accrual rather than by an a priori power calculation, resulting in modest statistical power and limited ability to detect small effect sizes typical of common genetic variants, thereby increasing the risk of type II error. Regional and genealogical imbalances between cases and controls may have introduced residual confounding, and regression models were not adjusted for zhuz affiliation due to substantial multicollinearity with region, limiting the ability to fully disentangle ancestry-related effects. Although nominal associations were observed in selected models, none remained significant after correction for multiple testing, and the predefined candidate SNP approach may have overlooked population-specific loci. Additionally, MDR interaction findings demonstrated modest predictive accuracy and should be considered exploratory. The absence of statistically significant associations after correction for multiple testing may reflect limited power rather than the absence of a biological effect. Finally, the observational design precludes causal inference, and replication in larger, ancestry-aware cohorts with genome-wide approaches is warranted.

5. Conclusions

The findings suggest that common variants in IGF2BP2, CDKAL1, MTNR1B, TCF7L2, PPARG, and GCK do not exert strong individual effects on GDM susceptibility in this cohort of ethnic Kazakh women. Instead, the results are more consistent with a modest polygenic architecture characterized by small effect sizes and possible weak gene–gene interactions. MDR analysis identified statistically significant interaction models; however, their limited predictive accuracy indicates that these findings should be interpreted as exploratory.
Taken together, the present study does not support the immediate clinical utility of the investigated SNPs for GDM risk prediction or screening in the Kazakh population. Nevertheless, the observed suggestive patterns and the unique genealogical structure of this population underscore the importance of larger, ancestry-aware, and methodologically robust studies. Future research incorporating genome-wide approaches, formal adjustment for population structure, and functional validation will be essential to clarify the genetic contribution to GDM and to determine whether polygenic or ancestry-specific effects meaningfully influence disease risk.

Supplementary Materials

The following supporting information can be downloaded at https://www.mdpi.com/article/10.3390/diagnostics16050663/s1: Table S1: Clinical and genetic data of study participants analyzed in the study.

Author Contributions

Conceptualization, L.D., G.S., G.B., B.S. and I.K.; methodology, L.D., B.S., A.T. and I.K.; software, A.S. and M.Z.; validation, G.S., G.B. and R.T.; formal analysis, I.K. investigation, G.S., G.B., B.S., A.T., A.S. and M.Z.; resources, L.D. and R.T.; data curation, I.K.; writing—original draft preparation, L.D. and I.K.; writing—review and editing, L.D., G.S., G.B., R.T., B.S., A.T., A.S., M.Z. and I.K.; visualization, B.S., A.T., A.S. and M.Z.; supervision, L.D., G.S., G.B. and R.T.; project administration, L.D., B.S. and I.K.; funding acquisition, L.D. and R.T. All authors have read and agreed to the published version of the manuscript.

Funding

This study was conducted as part of the scientific and technical program “Development of innovative and highly efficient technologies to reduce the risk of premature mortality from diseases of the circulatory system, chronic respiratory diseases and diabetes” (IRN: BR11065383), Ministry of Healthcare of the Republic of Kazakhstan.

Institutional Review Board Statement

The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Ethics Committee of the Asfendiyarov Kazakh National Medical University (Protocol No. 5 (111) dated 28 April 2021).

Data Availability Statement

The original contributions presented in this study are included in this article. Further inquiries can be directed to the corresponding authors (B.S., and I.K.).

Acknowledgments

The authors gratefully acknowledge Azhar Abilbay for assistance with reference management and image preparation during the manuscript drafting process.

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript:
ALTAlanine Aminotransferase
AORAdjusted Odds Ratio
ASTAspartate Aminotransferase
BMIBody Mass Index
CIConfidence Interval
CVCCross-Validation Consistency
DALYDisability-Adjusted Life Year
DMDiabetes Mellitus
GDMGestational Diabetes Mellitus
HDLHigh-Density Lipoprotein
IQRInterquartile Range
LDLLow-Density Lipoprotein
MDRMultifactor Dimensionality Reduction
OGTTOral Glucose Tolerance Test
OROdds Ratio
PCRPolymerase Chain Reaction
QALYQuality-Adjusted Life Year
RBCsRed Blood Cells
SDStandard Deviation
SDISocio-demographic Index
SNPSingle-Nucleotide Polymorphism
T2DMType 2 Diabetes Mellitus
TCTotal Cholesterol
TE bufferTris-EDTA Buffer (pH 8.0, used in molecular biology)
TGTriglycerides
TTGTissue Transglutaminase Antibodies
WHOWorld Health Organization

References

  1. Mora-Ortiz, M.; Rivas-García, L. Gestational Diabetes Mellitus: Unveiling Maternal Health Dynamics from Pregnancy Through Postpartum Perspectives. Open Res. Eur. 2024, 4, 164. [Google Scholar] [CrossRef]
  2. International Diabetes Federation Global. Prevalence of Gestational Diabetes (GDM). Available online: https://diabetesatlas.org/data-by-indicator/hyperglycaemia-in-pregnancy-hip-20-49-y/prevalence-of-gestational-diabetes-mellitus-gdm/ (accessed on 23 December 2025).
  3. Greco, E.; Calanducci, M.; Nicolaides, K.H.; Barry, E.V.H.; Huda, M.S.B.; Iliodromiti, S. Gestational Diabetes Mellitus and Adverse Maternal and Perinatal Outcomes in Twin and Singleton Pregnancies: A Systematic Review and Meta-Analysis. Am. J. Obstet. Gynecol. 2024, 230, 213–225. [Google Scholar] [CrossRef] [PubMed]
  4. Karkia, R.; Giacchino, T.; Shah, S.; Gough, A.; Ramadan, G.; Akolekar, R. Gestational Diabetes Mellitus: Association with Maternal and Neonatal Complications. Medicina 2023, 59, 2096. [Google Scholar] [CrossRef]
  5. Su, M.; Wang, Y.; Yuan, Q.; Tang, D.; Lu, Y.; Wu, X.; Xiong, W.; Li, Y.; Liu, T.; Zeng, S.; et al. The Impact of Gestational Diabetes Mellitus on Maternal-Fetal Pregnancy Outcomes and Fetal Growth: A Multicenter Longitudinal Cohort Study. Front. Pediatr. 2025, 13, 1592550. [Google Scholar] [CrossRef] [PubMed]
  6. Nielsen, H.; Windolf-Nielsen, A.; Scheuer, S.H.; Damm, P.; Andersen, A.M.N.; Andersen, G.S.; Nielsen, K.K. Type 2 Diabetes Risk After Gestational Diabetes According to Country/Region of Origin: A Nationwide Register-Based Study. J. Clin. Endocrinol. Metab. 2024, 109, e2196–e2204. [Google Scholar] [CrossRef] [PubMed]
  7. Bellamy, L.; Casas, J.P.; Hingorani, A.D.; Williams, D. Type 2 Diabetes Mellitus after Gestational Diabetes: A Systematic Review and Meta-Analysis. Lancet 2009, 373, 1773–1779. [Google Scholar] [CrossRef]
  8. Al Bekai, E.; Beaini, C.E.; Kalout, K.; Safieddine, O.; Semaan, S.; Sahyoun, F.; Ghadieh, H.E.; Azar, S.; Kanaan, A.; Harb, F. The Hidden Impact of Gestational Diabetes: Unveiling Offspring Complications and Long-Term Effects. Life 2025, 15, 440. [Google Scholar] [CrossRef]
  9. Haque, M.M.; Tannous, W.K.; Herman, W.H.; Immanuel, J.; Hague, W.M.; Teede, H.; Enticott, J.; Cheung, N.W.; Hibbert, E.; Nolan, C.J.; et al. Cost-Effectiveness of Diagnosis and Treatment of Early Gestational Diabetes Mellitus: Economic Evaluation of the TOBOGM Study, an International Multicenter Randomized Controlled Trial. eClinicalMedicine 2024, 71, 102610. [Google Scholar] [CrossRef]
  10. Xu, T.; Dainelli, L.; Yu, K.; Ma, L.; Silva Zolezzi, I.; Detzel, P.; Fang, H. The Short-Term Health and Economic Burden of Gestational Diabetes Mellitus in China: A Modelling Study. BMJ Open 2017, 7, e018893. [Google Scholar] [CrossRef]
  11. Canday, M. Identifying Gestational Diabetes Mellitus and Assessing Risk Factors in Affected Women: A Comprehensive Study. Eur. Rev. Med. Pharmacol. Sci. 2024, 28, 734–746. [Google Scholar] [CrossRef]
  12. Duan, B.; Zhou, Z.; Liu, M.; Liu, Z.; Zhang, Q.; Liu, L.; Ma, C.; Gou, B.; Liu, W. Development and Acceptability of a Gestational Diabetes Mellitus Prevention System (Better pregnancy) Based on a User-Centered Approach: A Clinical Feasibility Study. Digit. Health 2024, 10, 1–14. [Google Scholar] [CrossRef]
  13. Takele, W.W.; Vesco, K.K.; Josefson, J.; Redman, L.M.; Hannah, W.; Bonham, M.P.; Chen, M.; Chivers, S.C.; Fawcett, A.J.; Grieger, J.A.; et al. Effective Interventions in Preventing Gestational Diabetes Mellitus: A Systematic Review and Meta-Analysis. Commun. Med. 2024, 4, 75. [Google Scholar] [CrossRef] [PubMed]
  14. Shi, F.; Zhao, Q.; Yang, Y.; Liu, L.; Zhang, X.; Kim, H.J.; Wang, Y.; Pu, J. Global Burden of Diabetes in Women from 1990 to 2021, with Projections to 2050: Population-Based Study. BMC Med. 2025, 23, 538–548. [Google Scholar] [CrossRef]
  15. Prasad, R.B.; Kristensen, K.; Katsarou, A.; Shaat, N. Association of Single Nucleotide Polymorphisms with Insulin Secretion, Insulin Sensitivity, and Diabetes in Women with a History of Gestational Diabetes Mellitus. BMC Med. Genom. 2021, 14, 274. [Google Scholar] [CrossRef] [PubMed]
  16. Ding, M.; Chavarro, J.; Olsen, S.; Lin, Y.; Ley, S.H.; Bao, W.; Rawal, S.; Grunnet, L.G.; Thuesen, A.C.B.; Mills, J.L.; et al. Genetic Variants of Gestational Diabetes Mellitus: A Study of 112 SNPs among 8722 Women in Two Independent Populations. Diabetologia 2018, 61, 1758–1768. [Google Scholar] [CrossRef] [PubMed]
  17. Cruciat, G.; Florian, A.R.; Chaikh-Sulaiman, M.S.; Staicu, A.; Caracostea, G.V.; Procopciuc, L.M.; Stamatian, F.; Muresan, D. TCF7L2 Polymorphism Rs7903146 (C/T) and Gestational Diabetes Influence on Obstetric Outcome: A Romanian Case-Control Study. Int. J. Mol. Sci. 2024, 25, 4039. [Google Scholar] [CrossRef]
  18. Zeng, Q.; Liu, X.; Liu, J.; Wu, Y.; Zhang, Y.; He, Z.; Wei, Y.; Zeng, G.; Zou, D.; Guo, R. MTNR1B Variants Increase Gestational Diabetes Mellitus Risk in Young Chinese Pregnant Women. Sci. Rep. 2025, 15, 19643. [Google Scholar] [CrossRef]
  19. Gyan, W.R.; Zhang, H.; Shao, T.; Yang, T.; Wei, Y.; Li, M.; Che, X.; Zeng, Q.; Guo, R. Association of CDKAL1 Gene Polymorphisms Variations with Gestational Diabetes Mellitus Risk in Women: A Case-Control Study and Meta-Analysis. BMC Endocr. Disord. 2025, 25, 125. [Google Scholar] [CrossRef]
  20. Dmitrenko, O.; Karpova, N.; Nurbekov, M. Association of Polymorphisms Rs1801282 of the PPARG Gene, Rs8192678 of the PPARGC1A Gene and Rs7895833 of the SIRT1 Gene with the Risk of Preeclampsia in Pregnant Women with Gestational Diabetes in the Russian Population. Genet. Genom. Sci. 2022, 7, 1–8. [Google Scholar] [CrossRef]
  21. Liu, J.; Song, G.; Zhao, G.; Meng, T. Lack of Association between IGF2BP2 Rs4402960 Polymorphism and Gestational Diabetes Mellitus: A Case–Control Study, Meta-Analysis and Trial Sequential Analysis. Biosci. Rep. 2020, 40, 20200990. [Google Scholar] [CrossRef]
  22. Lei, C.; Liu, J.; Zhang, R.; Pan, Y.; Lu, Y.; Gao, Y.; Ma, X.; Yang, Y.; Guan, Y.; Mamatyusupu, D.; et al. Ancestral Origins and Admixture History of Kazakhs. Mol. Biol. Evol. 2024, 41, msae144. [Google Scholar] [CrossRef] [PubMed]
  23. Sikhayeva, N.; Bolatov, A.; Zholdybayeva, E.; Akhmetollayev, I.; Iskakova, A. Association of ADIPOQ Gene Polymorphisms with Type 2 Diabetes and Obesity Risk in the Kazakh Population: A Case–Control and Population-Based Study. Genes 2024, 15, 669. [Google Scholar] [CrossRef]
  24. Aitkaliyev, A.; Bekenova, N.; Vochshenkova, T.; Kassiyeva, B.; Benberin, V. Polymorphism Rs564398 of the ANRIL Gene as a Coronary-Artery-Disease-Associated SNP in Diabetic Patients of the Kazakh Population. Diagnostics 2024, 14, 2412. [Google Scholar] [CrossRef] [PubMed]
  25. Svyatova, G.; Berezina, G.; Danyarova, L.; Kuanyshbekova, R.; Urazbayeva, G. Genetic Predisposition to Gestational Diabetes Mellitus in the Kazakh Population. Diabetes Metab. Syndr. Clin. Res. Rev. 2022, 16, 102675. [Google Scholar] [CrossRef]
  26. Svyatova, G.; Berezina, G.; Murtazaliyeva, A.; Dyussupov, A.; Belyayeva, T.; Faizova, R.; Dyussupova, A. Genetic Predisposition to Prediabetes in the Kazakh Population. Curr. Issues Mol. Biol. 2024, 46, 10913–10922. [Google Scholar] [CrossRef] [PubMed]
  27. Sairambay, Y. Young People’s Perspectives on How ‘Zhuz’ and ‘Ru’ Clans Affect Them: Evidence from Three Cities in Post-Soviet Qazaqstan. Stud. Transit. States Soc. 2019, 11, 39–54. [Google Scholar]
  28. Tumalas.kz. Alash Genealogical Data. Available online: https://tumalas.kz/mobi/#/ (accessed on 30 December 2025).
  29. Shakeyeva, B.; Shaikemelev, M.; Aitymbetov, N. Tribalism in Kazakhstan: Tradition Reborn or Social Instrument. Cent. Asia Cauc. 2020, 21, 33–41. [Google Scholar] [CrossRef]
  30. Khussainova, E.; Kisselev, I.; Iksan, O.; Bekmanov, B.; Skvortsova, L.; Garshin, A.; Kuzovleva, E.; Zhaniyazov, Z.; Zhunussova, G.; Musralina, L.; et al. Genetic Relationship Among the Kazakh People Based on Y-STR Markers Reveals Evidence of Genetic Variation Among Tribes and Zhuz. Front. Genet. 2022, 12, 801295. [Google Scholar] [CrossRef]
  31. Turusbekov, Y.; Sabitov, Z.; Daulet, B.; Sadykov, M.; Khalidullin, O. The Kazakhstan DNA Project Hits First Hundred Y-Profiles for Ethnic Kazakhs. Russ. J. Genet. Geneal. 2012, 2, 69–84. [Google Scholar]
  32. R Core Team. A Language and Environment for Statistical Computing—R; R Core Team: Vienna, Austria, 2025. [Google Scholar]
  33. Posit Team RStudio: Integrated Development Environment for R. Available online: http://www.posit.co/ (accessed on 22 January 2024).
  34. Winham, S.J.; Motsinger-Reif, A.A. An R Package Implementation of Multifactor Dimensionality Reduction. BioData Min. 2011, 4, 24. [Google Scholar] [CrossRef]
  35. Sparsø, T.; Bonnefond, A.; Andersson, E.; Bouatia-Naji, N.; Holmkvist, J.; Wegner, L.; Grarup, N.; Gjesing, A.P.; Banasik, K.; Cavalcanti-Proença, C.; et al. G-Allele of Intronic Rs10830963 in MTNR1B Confers Increased Risk of Impaired Fasting Glycemia and Type 2 Diabetes through an Impaired Glucose-Stimulated Insulin Release: Studies Involving 19,605 Europeans. Diabetes 2009, 58, 1450–1456. [Google Scholar] [CrossRef] [PubMed]
  36. Huang, Y.; Dou, X.; He, M.; Su, Y.; Lin, H.; Yang, Y. The G-Allele of Rs10830963 in MTNR1B Exerts Stage-Specific Effects Across the Trajectory of Type 2 Diabetes: A Multi-State Analysis. Int. J. Mol. Sci. 2025, 26, 7855. [Google Scholar] [CrossRef]
  37. Lyssenko, V.; Nagorny, C.L.F.; Erdos, M.R.; Wierup, N.; Jonsson, A.; Spégel, P.; Bugliani, M.; Saxena, R.; Fex, M.; Pulizzi, N.; et al. Common Variant in MTNR1B Associated with Increased Risk of Type 2 Diabetes and Impaired Early Insulin Secretion. Nat. Genet. 2009, 41, 82–88. [Google Scholar] [CrossRef]
  38. Grant, S.F.A.; Thorleifsson, G.; Reynisdottir, I.; Benediktsson, R.; Manolescu, A.; Sainz, J.; Helgason, A.; Stefansson, H.; Emilsson, V.; Helgadottir, A.; et al. Variant of Transcription Factor 7-like 2 (TCF7L2) Gene Confers Risk of Type 2 Diabetes. Nat. Genet. 2006, 38, 320–323. [Google Scholar] [CrossRef]
  39. Tarnowski, M.; Bujak, J.; Kopytko, P.; Majcher, S.; Ustianowski, P.; Dziedziejko, V.; Safranow, K.; Pawlik, A. Effect of FTO and IGF2BP2 Gene Polymorphisms on Duration of Pregnancy and Apgar Scores in Women with Gestational Diabetes. J. Obstet. Gynaecol. 2019, 39, 151–156. [Google Scholar] [CrossRef] [PubMed]
  40. Popova, P.V.; Klyushina, A.A.; Vasilyeva, L.B.; Tkachuk, A.S.; Vasukova, E.A.; Anopova, A.D.; Pustozerov, E.A.; Gorelova, I.V.; Kravchuk, E.N.; Li, O.; et al. Association of Common Genetic Risk Variants With Gestational Diabetes Mellitus and Their Role in GDM Prediction. Front. Endocrinol. 2021, 12, 628582. [Google Scholar] [CrossRef] [PubMed]
  41. Stuebe, A.M.; Wise, A.; Nguyen, T.; Herring, A.; North, K.E.; Siega-Riz, A.M. Maternal Genotype and Gestational Diabetes. Am. J. Perinatol. 2014, 31, 69–76. [Google Scholar] [CrossRef]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Article Metrics

Citations

Article Access Statistics

Journal Statistics
Multiple requests from the same IP address are counted as one view.
A Data-Efficient Machine Learning Approach for Breast Ultrasound Lesion Classification Integrating Image-Derived Features and Sonographic Descriptors
Previous
Prognostic Value of Blood Urea Nitrogen for Acute Kidney Injury and Mortality in Vasculitis: A Large Cohort Study Using Multivariate Joint Model and Machine Learning
Next