Impaired glucose tolerance and cardiovascular risk factors in relation to infertility: a Mendelian randomization analysis in the Norwegian Mother, Father, and Child Cohort Study
Hernáez, Álvaro; Lee, Yunsung; Page, Christian Magnus; Skåra, Karoline Hansen; Håberg, Siri Eldevik; Magnus, Per Minor; Njølstad, Pål Rasmus; Andreassen, Ole; Corfield, Elizabeth Claire; Havdahl, Alexandra; Fraser, Abigail; Burgess, Stephen; Lawlor, Deborah A.; Magnus, Maria Christine
Journal article, Peer reviewed
Published version
View/ Open
Date
2024Metadata
Show full item recordCollections
- Department of Clinical Science [2448]
- Registrations from Cristin [10818]
Abstract
Study question: Are impaired glucose tolerance (as measured by fasting glucose, glycated hemoglobin, and fasting insulin) and cardiovascular disease risk (as measured by low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglycerides, systolic blood pressure, and diastolic blood pressure) causally related to infertility?
Summary answer: Genetic instruments suggest that higher fasting insulin may increase infertility in women.
What is known already: Observational evidence suggests a shared etiology between impaired glucose tolerance, cardiovascular risk, and fertility problems.
Study design, size, duration: This study included two-sample Mendelian randomization (MR) analyses, in which we used genome-wide association summary data that were publicly available for the biomarkers of impaired glucose tolerance and cardiovascular disease, and sex-specific genome-wide association studies (GWASs) of infertility conducted in the Norwegian Mother, Father, and Child Cohort Study.
Participants/materials, setting, methods: There were 68 882 women (average age 30, involved in 81 682 pregnancies) and 47 474 of their male partners (average age 33, 55 744 pregnancies) who had available genotype data and who provided self-reported information on time-to-pregnancy and use of ARTs. Of couples, 12% were infertile (having tried to conceive for ≥12 months or used ARTs to conceive). We applied the inverse variance weighted method with random effects to pool data across variants and a series of sensitivity analyses to explore genetic instrument validity. (We checked the robustness of genetic instruments and the lack of unbalanced horizontal pleiotropy, and we used methods that are robust to population stratification.) Findings were corrected for multiple comparisons by the Bonferroni method (eight exposures: P-value < 0.00625).
Main results and the role of chance: In women, increases in genetically determined fasting insulin levels were associated with greater odds of infertility (+1 log(pmol/l): odds ratio 1.60, 95% CI 1.17 to 2.18, P-value = 0.003). The results were robust in the sensitivity analyses exploring the validity of MR assumptions and the role of pleiotropy of other cardiometabolic risk factors. There was also evidence of higher glucose and glycated hemoglobin causing infertility in women, but the findings were imprecise and did not pass our P-value threshold for multiple testing. The results for lipids and blood pressure were close to the null, suggesting that these did not cause infertility.
Limitations, reasons for caution: We did not know if underlying causes of infertility were in the woman, man, or both. Our analyses only involved couples who had conceived. We did not have data on circulating levels of cardiometabolic risk factors, and we opted to conduct an MR analysis using GWAS summary statistics. No sex-specific genetic instruments on cardiometabolic risk factors were available. Our results may be affected by selection and misclassification bias. Finally, the characteristics of our study sample limit the generalizability of our results to populations of non-European ancestry.
Wider implications of the findings: Treatments for lower fasting insulin levels may reduce the risk of infertility in women.