Genetically predicted 17beta-estradiol and cardiovascular risk factors in women: a Mendelian randomization analysis using young women in Hong Kong and older women in the Guangzhou Biobank Cohort Study

Research output: Contribution to journalArticlepeer-review


  • Shiu Lun Au Yeung
  • Jie Zhao
  • Weisen Zhang
  • Chaoqiang Jiang
  • Tai Hing Lam
  • Gabriel Matthew Leung
  • C Mary Schooling

Colleges, School and Institutes


PURPOSE: The role of estrogen in cardiovascular health remains contested with discrepancies between findings from randomized controlled trials and observational studies. Mendelian randomization, which assesses the effect of lifelong endogenous exposure, may help elucidate these discrepancies.

METHODS: We used separate sample instrumental variable analysis to estimate the association of log 17β-estradiol with factors related to cardiovascular disease risk (systolic and diastolic blood pressure, lipids, fasting glucose, body mass index, waist hip ratio, and waist circumference) and Framingham score, a predictor of 10-year risk of ischemic heart disease events, in older Chinese women from the Guangzhou Biobank Cohort Study (GBCS, n = 3092). The estimate was derived using the Wald estimator, that is, the ratio of the association of genetic determinants (rs1008805 and rs2175898) of log 17β-estradiol with cardiovascular disease risk factors and Framingham score in GBCS and the association of these genetic determinants with log 17β-estradiol in a sample of young women from Hong Kong (n = 236).

RESULTS: Genetically, higher 17β-estradiol was not associated with any cardiovascular disease-related risk factor or with Framingham score (-0.01, 95% confidence interval = -1.34 to 1.31).

CONCLUSIONS: Lifetime exposure to estrogen does not appear to be cardioprotective via the cardiovascular disease-related risk factors examined.


Original languageEnglish
JournalAnnals of Epidemiology
Early online date4 Feb 2016
Publication statusE-pub ahead of print - 4 Feb 2016


  • Estrogen, Cardiovascular risk factors, Mendelian randomization analysis, Chinese, Women