1. Introduction
Gestational diabetes mellitus (GDM) refers to glucose intolerance recognized first during pregnancy and has become a growing public health concern worldwide. Recent epidemiological evidence estimates its global prevalence at around 14%, although this figure varies considerably between regions depending on screening practices, diagnostic criteria, and the underlying risks of the population.[1] The increasing incidence of GDM parallels the worldwide rise in obesity, type 2 diabetes, and advanced maternal age.[2]
Beyond its short-term obstetric and neonatal complications – such as preeclampsia, preterm delivery, and neonatal metabolic disturbances – GDM also confers long-term metabolic consequences for both mothers and their offspring, including heightened susceptibility to metabolic syndrome and type 2 diabetes later in life.[3,4] The economic implications are substantial; in the United States alone, GDM-related healthcare costs exceed $1.6 billion annually, mainly due to the management of associated complications.[1]
Despite progress in treatment, including diet therapy, exercise, and pharmacologic interventions with insulin or metformin, optimal glucose control remains challenging to achieve in many cases.[2,5] Although these strategies can mitigate immediate adverse pregnancy outcomes, they may not adequately address the psychological burden of GDM. A growing body of research has shown that affected women frequently experience higher levels of stress, anxiety, and depressive symptoms during and after pregnancy than those without GDM.[6,7] However, current clinical approaches emphasize glycemic regulation and physical health, while often neglecting the mental health dimension of this condition.
Increasing observational evidence links GDM with an elevated risk of perinatal mental disorders, including postpartum depression, anxiety, and even suicidality.[6,8] However, such findings are difficult to interpret due to the confounding and reverse causation inherent in observational designs.[9] Shared determinants, such as obesity, older maternal age, or socioeconomic stress, can predispose individuals to both GDM and psychiatric symptoms, obscuring the underlying direction of association. Consequently, although the relationship between GDM and perinatal mental health problems is widely reported, strong causal evidence remains limited.
To bridge this gap, approaches capable of distinguishing causal effects from correlation are required. Mendelian randomization (MR) provides such an opportunity by using genetic variants as instrumental variables to proxy modifiable exposures.[10,11] Because the allele distribution is random at conception, MR reduces confounding and minimizes reverse causation, producing more reliable causal estimates.[10,12] The growing number of genome-wide association studies (GWAS) that identify GDM-related loci now allows MR applications to explore whether genetically predicted GDM contributes to perinatal mental health outcomes.[13] Consequently, this study used a two-sample magnetic resonance imaging design to assess whether GDM has a causal effect on perinatal psychiatric disorders. Using independent genetic instruments from large GWAS datasets, our objective was to overcome the key limitations of traditional observational studies.
The overall design and analytic workflow are shown in Figure 1. Our objective was to determine whether genetically predicted GDM increases susceptibility to perinatal depression, anxiety, and suicidality. This study intends to provide genetic evidence supporting potential causal links and to emphasize the need for integrated clinical care – including systematic mental health screening and targeted interventions – for women with GDM.
2. Materials and methods
We implemented a two-sample MR design using the summary-level data (Table 1). Exposure data for GDM (finn-b-GEST DIABETES) and outcome data for peripartum mental disorders (finn-b-KRA_PSY_ANYMENTAL_SUICID_PREG_NERV) were obtained from the FinnGen database. Single-nucleotide polymorphisms (SNPs) significantly associated with GDM (P < 5 × 10−6) were selected as genetic instruments.
To ensure independence, SNPs were clumped using a 10,000 kb window and linkage disequilibrium threshold of r2 < 0.001. The causal effects were estimated using inverse-variance weighted (IVW) as the primary method, along with the MR-Egger, weighted median, simple mode, and weighted mode approaches (Table 2). Heterogeneity was tested using the Cochran’s Q statistic, horizontal pleiotropy was assessed using the MR-Egger intercept, and directionality was evaluated using the Steiger test (Table 3).
All analyses were performed using R version 4.3.2 (R Foundation for Statistical Computing, Vienna, Austria), with TwoSampleMR and MRPRESSO packages.
3. Results
Nineteen independent SNPs were included as instrumental variables (Table 1). The IVW analysis revealed a significant positive association between genetically predicted GDM and peripartum mental disorders (β = 0.048, SE = 0.014, P = 8.34 × 10−4, odds ratio [OR] = 1.049, 95% confidence interval [CI] = 1.020–1.079). The weighted median approach yielded consistent results (β = 0.056, SE = 0.021, P = .007, OR = 1.057, 95% CI = 1.015–1.101).
The MR-Egger regression produced similar directional estimates (β = 0.086, P = .018, OR = 1.090, 95% CI = 1.022–1.163). No evidence of heterogeneity (Q = 17.19, P = .51) or horizontal pleiotropy (Egger’s intercept = −0.0086, P = .217) was observed (Table 3). The Steiger directionality test confirmed the correct causal direction from GDM to mental disorders.
The forest and scatter plots displaying these causal estimates are shown in Figures 2 and 3, respectively. The funnel plots and leave-one-out sensitivity analyses further demonstrated the robustness of the results (Figs. 4 and 5).
The causal estimates at the SNP level with 95% CI are shown in Figure 6.
4. Discussion
GDM is a complex metabolic disorder that develops during pregnancy and has gained increasing clinical attention due to its profound physiological and psychological implications. Our MR analysis demonstrated a positive causal relationship between genetically predicted GDM and perinatal mental disorders (Fig. 2).
This section examines the plausible biological and clinical pathways that may underlie this association.
A growing body of evidence suggests that GDM induces low-grade chronic inflammation, characterized by elevated chemokines such as CCL2, CXCL8, and CXCL12, which play regulatory roles in neuroimmune signaling relevant to affective disorders.[14] This inflammation may alter the hypothalamic-pituitary-adrenal axis and neuroendocrine balance, promoting susceptibility to depression and anxiety.[15] Longitudinal studies further indicate that women with GDM show heightened psychological distress – including depressive and anxiety symptoms – that correlate with metabolic and inflammatory markers.[16,17] Suboptimal glycemic control is inversely related to emotional stability, suggesting that impaired insulin signaling and neuroinflammation act as convergent pathways that affect mood regulation.[16] Taken together, these findings imply that the causal effect observed in our MR study reflects a multifaceted interaction between metabolic, inflammatory, and neuroendocrine disturbances.
Beyond inflammation, metabolic dysregulation in GDM also involves abnormalities in lipid metabolism and oxidative stress, both of which are implicated in the pathogenesis of mood disorders.[18] The shared lipidomic profiles between GDM and depressive phenotypes further support the overlapped biological pathways.[18] Behavioral factors, such as disrupted sleep and circadian rhythm misalignment, commonly observed in women with GDM, exacerbate emotional vulnerability by altering central nervous system homeostasis.[19] In particular, studies have shown that poor sleep quality during pregnancy predicts postpartum depression, suggesting a behavioral pathway linking metabolic imbalance and psychological distress.[19] Together, these mechanisms emphasize that biological and behavioral processes contribute to the psychiatric risk of GDM.
Our findings align with prior observational reports of increased depression and anxiety among women with GDM but extend beyond correlation to demonstrate causality. Traditional observational designs have struggled to unravel the bidirectional influence of mood and metabolic disorders.[20,21] By contrast, this MR study’s strengths include the use of large-scale GWAS datasets from FinnGen, a robust two-sample MR design minimizing confounding, and the application of multiple complementary analytical methods (IVW, MR-Egger, weighted median) to ensure consistency and robustness of causal estimates.
MR reduces confounding and reverse causation through the use of genetic instruments, offering a more robust inference. This approach clarifies the inconsistencies across previous studies where associations weakened after adjusting for socioeconomic and metabolic factors.[22]
Psychosocial evidence supports the direct psychological impact of GDM. Many women describe anxiety, guilt, and frustration related to lifestyle restrictions and clinical follow-up, which can worsen emotional distress and reduce their quality of life.[21,23] Structured counseling and psychosocial interventions have proven beneficial in improving coping strategies and emotional resilience.[24] Addressing mental health may therefore improve not only psychological outcomes but also metabolic control, reinforcing the importance of integrated care.
This study had several limitations. The relatively modest sample size and reliance on genetic proxies without direct clinical validation may have limited external validity. Although MR reduces confounding, the absence of experimental validation restricts mechanistic interpretation.
The heterogeneity of the population and the potential effects of the batch could introduce variability in the effect estimates, highlighting the need for replication between larger and more diverse cohorts (Table 4).
5. Conclusion and clinical implications
This study provides genetic evidence of a positive causal relationship between GDM and perinatal mental disorders, including depression, anxiety, and suicidality. These results underscore the need to incorporate mental health screening and psychological support into standard GDM management. The early identification of at-risk individuals and multidisciplinary collaboration between obstetric and psychiatric teams can help reduce the burden of perinatal mental illness.
From a broader perspective, the integration of metabolic and psychological care within maternal health programs could improve both physical and cognitive outcomes, contributing to a more comprehensive perinatal care model.
Table S1, Supplemental Digital Content 1 is available for this article.
Acknowledgments
The authors thank the FinnGen consortium for access to the GWAS data.
Author contributions
Data curation: Chunyan Yang, Kaosheng Lu, Wei Dang, Xiaoxue Cheng, Lili Liu.
Formal analysis: Chunyan Yang, Wenhui Song, Kaosheng Lu, Meijia Zhang, Xiaoxue Cheng.
Investigation: Chunyan Yang.
Visualization: Chunyan Yang, Wenhui Song.
Project administration: Xiaohua Wu.
Supervision: Xiaohua Wu.
Methodology: Kaosheng Lu.
Conceptualization: Wei Dang, Meijia Zhang, Lili Liu.
Writing – original draft: Chunyan Yang, Wenhui Song, Wei Dang, Xiaoxue Cheng, Lili Liu.
Writing – review & editing: Xiaohua Wu, Kaosheng Lu, Meijia Zhang.