Age at Menarche Mediating Visceral Adipose Tissue's Influence on Pre-Eclampsia

A Mendelian Randomization Study

Peizhi Deng; Qingwei Yu; Haibo Tang; Yao Lu; Yingdong He


J Clin Endocrinol Metab. 2023;108(2):405-413. 

In This Article

Abstract and Introduction


Context: The association between visceral adipose tissue (VAT) and pre-eclampsia (PE) shows inconsistent results and the underlying mediator remains unknown.

Objective: We aimed to explore the causal effect of VAT on PE risks and the mediation role of age at menarche (AAM) in explaining this relationship.

Methods: Summary data for PE were obtained from the FinnGen genome-wide association study (3556 cases and 114 735 controls). For exposure data, 70 genetic variants associated with the predicted VAT in 161 149 European women from UK Biobank were used as instrumental variables. Inverse variance weighted and multiple sensitivity analyses were applied. We also conducted multivariable Mendelian randomization (MR) analyses to test the association between VAT-associated single-nucleotide variations and PE. Next, mediation analyses were performed to study whether the association between VAT and PE was mediated via AAM.

Results: In univariable MR analysis, higher volume of VAT was associated with the advancement of AAM and increased PE risk (beta = −0.33; 95% CI, −0.49 to −0.16 for AAM; odds ratio 1.65, 95% CI, 1.23 to 2.20 for PE). After adjusting for waist circumference, waist to hip ratio, and hip circumference, the multivariable MR results presented the consistent positive causality of VAT on PE. Two-step MR analysis proved an estimated 14.3% of the positive effect of VAT on PE was mediated by AAM.

Conclusion: Our findings provided evidence of the causal relationship between VAT and PE and proved VAT could accelerate AAM and then contribute to the risk of incident PE.


Pre-eclampsia (PE), which is identified as new-onset hypertension and proteinuria developing combined with multisystem pregnancy disorder in the last 20 weeks of fetation and resolving after childbirth, is one of the leading causes of maternal and neonatal mortality and morbidity, with an estimated global incidence of up to 3% to 5% and causing at least 42 000 worldwide maternal deaths annually.[1–4] In low- and middle-income countries, PE even accounts for 30% of all maternal mortality.[5] Thus, exploring the precise risk factors and therapeutic approaches of PE is an active area that promotes clinical treatment.

Recently, a growing number of studies have shown that visceral adipose tissue (VAT) is a proven high-risk factor associated with the risk of incident PE.[6–9] Multiple thrombogenic and inflammatory factors secreted by VAT lead to the dysfunction of glucose metabolism and enhanced insulin resistance, contribute to metabolic syndrome, and even cause cardiovascular diseases.[10,11] Observational studies found that the activated macrophage content and inflammatory adipokine in VAT were higher in PE and the higher VAT thickness increased the risk of PE by 23%/cm.[6,12–14] Existing studies have also indicated that age at menarche (AAM), which could be influenced by VAT, might explain the different risks of PE, hinting at the intermediation of AAM.[15–21] However, there are some conflicts in the observational literature, meaning that more studies are required to obtain the true causal estimates.[22,23] These conventional epidemiological studies might be influenced by unknown confounding factors, measurement error, and reverse causal effects from the environmental data.

Mendelian randomization (MR) design can provide a method for overcoming above limitations by utilizing genetic variants related to the exposures of interest to determine the true causal relationship of exposure on the outcome. The reverse causality is averted due to the stability of instrumental variants (IVs), which ensures their impact on exposure occurs before the outcome. Currently, there is only one MR article that explored the effect of VAT on PE.[24] However, the study is limited in assessing the real association, due to its lower statistical power, unadjusted confounders for VAT and absence of mediation analysis for exploring the underlying mechanisms. In view of the remaining defects, we performed a two-step and multivariable MR design to determine the authentic role of VAT in PE risk as well as estimate the mediation role of AAM on the association between VAT and PE.

Leveraging the recent largest genome-wide association study (GWAS) data for VAT accumulation, we aim to: (i) explore the causal effect of VAT on PE; (ii) identify AAM as a mediator in the association between VAT and PE, and clarify the proportion mediated.