reviewed by Federica Piani

Age at menarche could be a screening tool to identify females at increased risk for cardiometabolic diseases, including myocardial infarction (MI) (1). It has been previously described that the lowest risk of MI is observed in females whose menarche occurs around ages 13–14 (2).

The objective of the study “Reassessing the association between age at menarche and cardiovascular disease: observational and Mendelian randomization analyses” by Tschiderer and colleagues (3) was to examine the potential non-linear and causal relationship between age at menarche and later cardiovascular outcomes, specifically MI, ischemic stroke and haemorrhagic stroke, using both observational data and Mendelian randomization.

In this observational study, the authors included data from 283,210 females free of MI or stroke at baseline, drawn from the European Prospective Investigation into Cancer and Nutrition-Cardiovascular Disease (EPIC-CVD) and the UK Biobank.

The median age at menarche was 13 years (IQR: 12–14). The authors performed a Cox regression to explore non-linear associations (U-shaped), comparing age at menarche ≤12 and ≥16 to reference (13 years).

A Mendelian Randomization analysis was also performed with a Non-linear Mendelian Randomization via fractional polynomials and a Linear Mendelian Randomization via inverse-variance weighted regression.

The observational results showed a U-shaped association between age at menarche and MI and ischemic stroke, meaning that females with age at menarche ≤12 or ≥16 had higher risks compared to those with menarche at age 13.

The Mendelian Randomization analyses pointed out that:

• genetically proxied age at menarche showed no departure from linearity in its association with cardiovascular outcomes; 
• each one-year increase in genetically proxied age at menarche was associated with a 8% lower risk of MI (hazard ratio: 0.92; 95% CI: 0.86–0.99);
• no causal association was found between genetically determined age at menarche and either ischemic or haemorrhagic stroke. 

The authors concluded that the observational U-shaped pattern for MI and ischemic stroke may reflect confounding or other non-causal influences. Mendelian Randomization evidence suggests a causal protective effect of later menarche on MI risk—but no causal link to stroke outcomes. Finally, it is stated that higher genetically determined age at menarche reduces MI risk in a linear fashion, with no evidence of increased risk at later ages or a threshold.

The strength of this research lies in its large sample size (more than two hundred thousand females) its methodological rigor, leveraging both fractional polynomial models for non-linear Mendelian Randomization and inverse-variance weighted regression for linear mendelian randomization. These insights challenge previous assumptions, implicating that the observed U-shaped patterns may be due to confounding factors (such as BMI, socioeconomic status, or recall bias) rather than true biological effects of menarche timing. The potential for recall bias underscores the importance of treating age at menarche as a fundamental health datum, to be routinely captured in medical records alongside other key variables, rather than relying on retrospective self-report.

Limitations of the study also include generalizability (predominantly European cohorts) and typical mendelian randomization caveats (e.g., potential pleiotropy).

In summary, this article suggests that the relationship between age at menarche and cardiovascular disease risk is more complex than previously assumed, with Mendelian randomization analyses indicating a causal protective effect of later menarche on MI, but not on stroke outcomes. This study offers valuable clarification on inconsistencies in prior research, demonstrating the complexities of using age at menarche as a cardiovascular risk marker and underscoring the role of genetic methods in revealing causal relationships in women’s health research.