Causal Associations Between Sarcopenia and Gestational Diabetes Mellitus

Introduction

Gestational diabetes mellitus (GDM) is a common metabolic disorder that occurs during pregnancy and its incidence is increasing. It is associated with a range of complications affecting both the mother and fetus during pregnancy and childbirth, leading to extensive health and economic burden.¹ Hence, there is increasing importance in recognizing women at risk of GDM. Sarcopenia, a condition characterized by the loss of muscle mass and function, has traditionally been linked to the aging process in the elderly,² but it is now acknowledged that the onset of sarcopenia can occur at an earlier stage in life.³ Sarcopenia is driven by biological changes in muscle structure, hormonal imbalances, and inadequate energy intake.⁴

Muscles play a crucial role in the overall metabolic health, including energy expenditure, insulin sensitivity, and blood glucose stability. Consequently, sarcopenia can have adverse effects on metabolic health, leading to insulin resistance and abnormalities in glucose metabolism. Multiple studies have indicated that sarcopenia is associated with higher fasting blood glucose,⁵ insulin resistance,⁶ hyperlipidemia⁷ and type 2 diabetes mellitus.⁸ Type 2 diabetes mellitus is characterized by insulin resistance and abnormal lipid metabolism and may promote the development of sarcopenia. A Mendelian randomization (MR) study explored the causal relationship between sarcopenia and type 2 diabetes mellitus, revealing a bi-directional causal relationship between handgrip strength and type 2 diabetes mellitus as well as between walking pace and type 2 diabetes mellitus. Sarcopenia and type 2 diabetes mellitus may reciprocally exert a substantial causal influence on each other.⁹ Another meta-analysis showed an elevated prevalence of sarcopenia in individuals with diabetes. The development of sarcopenia is increased by risk factors such as glycated hemoglobin, diabetic nephropathy, visceral fat area, duration of diabetes, and C-reactive protein.¹⁰

GDM and sarcopenia have common risk factors. However, the potential causal relationship between sarcopenia and GDM remains unknown. Observational studies have the limitations such as difficulty in establishing causation, selection bias, less control over variables and retrospective data collection challenges.¹¹ In recent years, MR has emerged as a powerful tool for assessing causal relationships in observational studies.¹² MR leverages genetic variations found in the natural population as instrumental variables to mimic the conditions of a randomized experiment, providing valuable insights into causality.

Therefore, this study sought to employ MR methodology to investigate the causal relationship between genetic variants associated with sarcopenia and GDM. Through this approach, we aimed to enhance our understanding of how sarcopenia may influence the risk of GDM, thereby providing valuable insights into the development of strategies to improve maternal health and pregnancy outcomes through early prevention and intervention.

Materials and Methods

Study Design

In this study, we utilized a bi-directional MR study design, employing two-sample MR methodologies and various genome-wide association study (GWAS) summary level datasets to elucidate the causal relationship between sarcopenia-related traits and GDM risk. Sarcopenia was measured using appendicular lean mass (ALM), right-hand grip strength, left-hand grip strength, and usual walking pace.^2,13

This study was conducted in two stages to explore the causal association between sarcopenia and GDM. In the first stage, exposure was related to sarcopenia-related traits, with GDM considered as the outcome. In the second stage, GDM was investigated as an exposure, while sarcopenia-related traits were the outcome. A detailed illustration of the study design is shown (Figure 1).

Figure 1 Overall design of the MR analysis in the present study. Abbreviations: GDM, gestational diabetes mellitus; IVW, inverse-variance weighted; MR, Mendelian randomization; SNPs, single-nucleotide polymorphisms.

Data Sources for GDM and Sarcopenia

The open GWAS database established by the MRC Integrated Epidemiology Unit and UK Biobank are expansive biomedical databases and research resources.^14,15 The summary-level GWAS data correlated with GDM were obtained from the FinnGen datasets, and the GWAS ID of GDM was finngen_R10_GEST_DIABETES. This dataset included 14,718 GDM patients and 215,592 controls. The GWAS summary data for ALM were derived from a GWAS with 244,730 UK Biobank participants (dataset ID: ebi-a-GCST90000027), with no difference in genetic effects between sexes or among different age strata.¹⁶ We also collected GWAS summary data from the UK Biobank for hand grip strength (right) in 461,089 people (dataset ID: ukb-b-10215), hand grip strength (left) in 461,026 people (dataset ID: ukb-b-7478), and usual walking pace in 459,915 people (dataset ID: ukb-b-4711) (Table 1). The participants of this study were exclusively of European origin.

Table 1 Phenotype Source and Description

Selection of Genetic Instruments

To ascertain the causal connection between sarcopenia and GDM, we employed four distinct sets of genetic instruments representing traits of sarcopenia when sarcopenia was present: a) index single-nucleotide polymorphisms (SNPs) representing ALM, b) index SNPs representing right-hand grip strength, c) index SNPs representing left-hand grip strength, and d) index SNPs representing usual walking pace. We also explored the genetic instruments representing GDM when GDM was studied as exposure in the second stage of MR analysis.

Genetic instruments were selected according to the following criteria: a) GWAS-correlated P-value ≥5×10⁸ as the threshold for SNPs, b) linkage disequilibrium r² < 0.001, and c) clumping distance < 10 MB as the cut-off value to assess the independent SNPs in linkage equilibrium. Furthermore, we also computed the F value and considered an F value greater than 10 to be sufficiently robust to address any potential issues related to weak instrument bias, ensuring the reliability of the selected SNPs.

In accordance with these recommendations, we initially selected a set of independent SNPs that were strongly associated with each exposure variable before matching in the outcome database. Outcome-related SNPs were removed during harmonization. We utilized the harmonized data function provided by TwoSampleMR (version 0.5.6) to ensure that the effect allele of each SNP was aligned with the corresponding allele of the exposure.¹⁷ Subsequently, using the procedure of MR-PRESSO (version 1.0, 1000 repeated settings), any outlier SNPs were excluded, enabling us to proceed with the MR analysis.

Mendelian Randomization Analysis

We used five distinct MR methods, including random-effect inverse-variance weighted (IVW), MR Egger, weighted median, simple mode, and weighted mode, to account for variant heterogeneity and investigate the potential impact of pleiotropy.

The primary outcome was assessed using the IVW method, whereas the other four methods were used to improve the IVW estimates as they could provide more robust estimates in a broader set of scenarios, despite being less efficient.¹⁸ MR Egger permits all genetic variants to exhibit pleiotropic effects, but it necessitates that these pleiotropic effects are unrelated to the association between the variant and the exposure.¹⁹ The weighted median approach permits the inclusion of potentially flawed instruments, assuming that a minimum of half of the instruments utilized in the MR analysis are reliable.²⁰ If the direction of β-value in the other four MR methods was inconsistent with the IVW method, the MR results were considered insignificant.

Analysis of Horizontal Pleiotropy and Heterogeneity

To further ensure the reliability of MR causal impact estimates, we performed pleiotropy and heterogeneity analyses. The pleiotropy test function in the two-sample MR package was used to determine whether the SNPs were suitable instrumental variables. Horizontal pleiotropy was evaluated using the MR-Egger intercept test and MR-PRESSO global test.²¹ If pleiotropy was not significant (P > 0.05), instrumental variables were used. A funnel plot was used to evaluate potential directional pleiotropy. Cochran’s Q test was used to assess heterogeneity among the selected instrumental variables.

Statistics

Statistical analyses were conducted using the R software (version 4.3.0), and statistical significance was determined using a two-tailed P-value < 0.05.

Results

Forward MR Analysis

In the first stage of MR analysis, following the removal of outlier SNPs through MR-PRESSO, 358 index SNPs were selected to genetically predict ALM, 146 index SNPs were used to genetically predict right-hand grip strength, 128 index SNPs were used to genetically predict left-hand grip strength, and 50 SNPs were used to predict usual walking pace (Table S1 - S4).

We conducted a comprehensive MR study to investigate genetic predictors of sarcopenia in relation to GDM. The IVW results of ALM (OR= 1.2182, 95% CI: 1.1397–1.3021, P<0.0001), right-hand grip strength (OR= 1.4194, 95% CI: 1.0773–1.8701, P= 0.0128), left-hand grip strength (OR= 1.6064, 95% CI: 1.2829–2.0115, P<0.0001) and usual walking pace (OR= 3.3676, 95% CI: 1.8769–6.0423, P<0.0001) were found to increase the risk of GDM. The other four MR methods yielded consistent results (Figures 2 and 3A - D). Sarcopenia had a causal effect on GDM.

Figure 2 Multivariate MR analyses of sarcopenia and GDM. P<0.05:*, P<0.01:**, P<0.001:***, P<0.0001:****. Abbreviations: GDM, gestational diabetes mellitus; MR, Mendelian randomization; SNPs, single-nucleotide polymorphisms.

Figure 3 Scatter plot of the causality of sarcopenia on GDM. (A) The causality of appendicular lean mass on GDM. (B) The causality of grip strength (right) on GDM. (C) The causality of grip strength (left) on GDM. (D) The causality of usual walking pace on GDM. Abbreviations: GDM, gestational diabetes mellitus; MR, Mendelian randomization; SNPs, single-nucleotide polymorphisms.

To evaluate the robustness and reliability of the MR results, Cochran’s Q test, leave-one-out analysis, the MR-PRESSO global test, and the MR-Egger intercept test were employed for statistically significant estimates (Table 2 and Figure S1A - D). The results of the MR-PRESSO global test indicated the presence of horizontal pleiotropy in ALM, right-hand grip strength, and usual walking pace. However, it is important to note that all MR-Egger intercept test P-values were > 0.05, suggesting the absence of horizontal pleiotropy. The primary method for evaluating horizontal pleiotropy was the MR-Egger intercept. No outliers were detected in the leave-one-out plot (Figure S2A - D). Heterogeneity was observed in ALM, right-hand grip strength, and usual walking pace, as indicated by Cochran’s Q test (P < 0.05), except for left-hand grip strength. Because we employed random-effects IVW as our primary result, the presence of heterogeneity was considered acceptable.

Table 2 Heterogeneity and Pleiotropy Testing of Sarcopenia on Genetic Prediction of GDM

Reverse MR Analysis

In the second stage, we explored the effect of GDM on sarcopenia after removing outlier SNPs using MR-PRESSO. We identified nine index SNPs of GDM on genetic prediction of ALM, eight index SNPs of GDM on genetic prediction of right-hand grip strength, nine index SNPs of GDM on genetic prediction of left-hand grip strength, and ten index SNPs of GDM on genetic prediction of usual walking pace (Tables S5 - S8). The IVW analysis showed that GDM had no causal effect on ALM (β= −0.0057, 95% CI: −0.0395–0.0282, P= 0.7424), right-hand grip strength (β= 0.0098, 95% CI: −0.0036–0.0231, P= 0.1509), left-hand grip strength (β= 0.0125, 95% CI: 0.0018–0.0232, P= 0.0225), and usual walking pace (β= −0.0158, 95% CI: −0.0282 - −0.0035, P= 0.0119) (Table 3).

Table 3 MR Results From GDM on Genetic Prediction of Appendicular Lean Mass, Grip Strength (Right and Left) and Usual Walking Pace

The Cochran’s Q test, MR-Egger intercept test, MR-PRESSO global test and leave-one-out analysis were performed to verify the reliability of the reverse MR analysis results (Table 4). The MR-PRESSO global test and MR-Egger intercept test results had P-values of > 0.05, indicating that no horizontal pleiotropy was detected. According to the leave-one-out analysis, eliminating one SNP did not change the directionality of the results, except for the GDM estimates in ALM (Figure S3A - D). All Cochran’s Q tests had P-values > 0.05, except for ALM and left-hand grip strength, which had P-values < 0.05.

Table 4 Heterogeneity and Pleiotropy Testing of GDM on Genetic Prediction of Sarcopenia

Discussion

Our study provides compelling evidence to support a causal relationship between sarcopenia and GDM development. The MR analysis revealed a robust association between genetic variants linked to sarcopenia and increased risk of GDM, shedding light on the potential significance of muscle health during pregnancy. Our study investigated the previously unexplored causal relationship between sarcopenia and GDM. This novel approach stems from the growing recognition of the role of sarcopenia in metabolic health, and its relevance to pregnancy outcomes.

The observed causal link between sarcopenia and GDM prompted further exploration of the underlying mechanisms (Figure 4). The skeletal muscle, constituting the largest organ in terms of mass, serves as the primary regulator of glucose homeostasis, accounting for 80% of the uptake of postprandial blood glucose, playing a critical role in maintaining normal insulin sensitivity.²² Reduced muscle mass has been associated with decreased glucose uptake and utilization, potentially contributing to hyperglycemia. Insulin resistance is caused by reduced sensitivity of muscles to the insulin released by the pancreas, resulting in impaired glucose uptake and elevated blood glucose levels.²³ Park et al²⁴ carried out a cross-sectional study involving 372,399 Korean individuals to explored the relationship between skeletal muscle mass and diabetes incidence, insulin resistance, glycated hemoglobin. Their findings indicated that skeletal muscle mass was negatively associated with the incidence of diabetes, insulin resistance, and glycated hemoglobin. A prospective cohort study aimed to investigate the association between insulin resistance from early adulthood to middle age and ALM in individuals without diabetes found that a trajectory of high insulin resistance was associated with a reduction in ALM/BMI among middle-aged individuals (β= −0.043, 95% CI: −0.063 – −0.023).²⁵ Insulin resistance throughout early and middle adulthood has a long-lasting impact on the health of skeletal muscles, which may initiate a pathway of harm to muscular health by potentially disrupting the normal physiological environment. This disruption could involve damage to capillary ultrastructure, inflammation, and, over time, muscle atrophy.²⁶ Additionally, the depletion of muscle mass plays a role in the development of glucose intolerance and the promotion of gluconeogenesis, subsequently exacerbating insulin resistance. Consistent with prior research on sarcopenia and diabetes, our study confirmed that sarcopenia plays a role in the development of GDM. Sarcopenia can disrupt glucose metabolism and lead to insulin resistance,²⁷ elevating the risk of GDM.²⁸

Figure 4 The two-sample MR framework showed that sarcopenia potentially causes GDM. Abbreviations: GDM, gestational diabetes mellitus; MR, Mendelian randomization. Notes: Figure 4 was created in BioRender. žÄ, OE. (2025) http://BioRender.com/i60o427.).

Muscle tissue contributes to the regulation of inflammation and metabolism in the body.^29,30 In a typical physiological state, pro-inflammatory cytokines play a role in regulating the balance between skeletal muscle synthesis and metabolic breakdown. However, in instances of muscle atrophy, there is an upregulation in the expression of proinflammatory cytokines, resulting in increased skeletal muscle breakdown metabolism. Moreover, these cytokines can hinder protein synthesis within the skeletal muscle cells, ultimately compromising muscle function. Anti-inflammatory cytokines can enhance glucose metabolism in muscles, serve as recruiting factors for skeletal muscle cells during the muscle growth process, and promote muscle growth and regeneration. In a cohort of Chinese participants, elevated levels of the inflammatory cytokines TWEAK and TNF-α were linked to a higher likelihood of developing sarcopenia, whereas increased concentrations of the metabolic hormone insulin-like growth factor 1, insulin, and adiponectin were linked to a reduced risk of sarcopenia.³¹ A meta-analysis including 3072 participants with sarcopenia and 8177 controls found that sarcopenia was significantly associated with high levels of C-reactive protein, while serum IL6 and TNF-α levels were not significantly different compared with the control.³² Sarcopenia can lead to an increased release of inflammatory cytokines, which can disrupt normal insulin signaling, further exacerbating insulin resistance and high blood glucose.³³ Previous studies revealed that serum inflammatory cytokines in women with GDM increased during gestation,³⁴ and the high levels of proinflammatory cytokines were associated with insulin resistance. Therefore, we can infer that pregnant women diagnosed with sarcopenia experience enhanced expression and secretion of pro-inflammatory cytokines. This leads to an imbalance between proinflammatory and anti-inflammatory cytokines, ultimately impairing glucose homeostasis and promoting the occurrence and development of GDM.

The skeletal muscle is one of the primary energy-consuming organs. Reduced muscle mass may result in decreased basal metabolic rate and reduced energy expenditure. This can lead to energy accumulation within the body, exacerbating metabolic disturbances. Mitochondria within muscle cells are crucial for energy production. A decrease in muscle mass may be accompanied by mitochondrial dysfunction, including ATP synthesis and oxidative phosphorylation, leading to the disruption of energy metabolism. This can exacerbate insulin resistance and contribute to diabetes development. Crane et al³⁵ discovered that individuals with sarcopenia show increased lipid accumulation and a reduced number of mitochondria in muscle cells, resulting in reduced mitochondrial function and lipid metabolism. Lipid signaling pathways were impaired in adipocytes of sarcopenic muscle. The muscle tissue also plays a crucial role in hormonal regulation, including insulin, estrogen, and other hormones involved in pregnancy. Sarcopenia may disrupt the normal regulation of these hormones, contributing to metabolic disturbances.³⁶ Shao et al³⁷ through a biopsy of rectus abdominis muscle, observed that the level of membrane protein plasma cell membrane glycoprotein-1 increased by approximately 63% in the GDM group. Their findings indicate that heightened phosphorylation of serine/threonine residues in muscle insulin receptors results in reduced insulin receptor tyrosine kinase activity. These post-receptor defects in the insulin signaling pathway may play a role in the pathophysiology of GDM. Understanding these mechanistic pathways is essential to elucidate the precise physiological connections between sarcopenia and GDM.

The prevalence of sarcopenia was found 2–3 times higher in people with type 2 diabetes mellitus than in control.³⁸ Type 2 diabetes mellitus is associated with poorer skeletal muscle strength and quality and a greater loss of muscle mass. Diabetes and sarcopenia have a causal relationship. However, GDM was not identified as a risk factor for sarcopenia in our study. Our findings have several clinical implications. Recognizing sarcopenia as a potential risk factor for GDM can inform prenatal care and management strategies. Implementing interventions aimed at preserving or enhancing muscle mass and function during pregnancy may help mitigate the risk of GDM. Such interventions may include tailored exercise and nutritional support programs.

It is crucial to acknowledge the strengths and limitations of the study. MR analysis allowed us to establish a causal link between sarcopenia-related genetic variants and GDM risk, providing a robust foundation for our findings. However, selection bias may have been present because of the binary exposures used in MR analysis. Additionally, the homogeneity of the study population, which primarily consisted of individuals of European descent, limits the generalizability of our findings to diverse ethnic and racial groups. Future research should explore in greater detail the specific mechanisms linking sarcopenia and GDM. Longitudinal studies that track muscle health throughout pregnancy and its impact on glucose metabolism are warranted. We are currently conducting a prospective study to investigate whether pregnant women with sarcopenia are more likely to develop GDM than those without sarcopenia. Furthermore, investigating potential interventions to prevent or mitigate GDM risk in individuals at risk of sarcopenia could advance personalized prenatal care.

Conclusion

In conclusion, our study adds to the growing body of evidence suggesting that sarcopenia plays a causal role in the development of GDM, whereas GDM does not causally affect sarcopenia. This newfound understanding opens avenues for targeted interventions and personalized care to reduce the burden of GDM among pregnant individuals. Further research is needed to refine our understanding of the intricate physiological pathways involved, and to develop effective strategies for prevention and management.