Older age for menopause raises risk for lung cancer

Publish date: February 28, 2022

This study was published on Medrxiv.org as a preprint and has not yet been peer reviewed.

Key takeaways

in analyses of more than 100,000 women that used Mendelian randomization (MR) as a tool to reduce residual confounding.
The MR analyses showed no significant association between ANM and breast cancer, endometrial cancer, ovarian cancer, coronary heart disease, ischemic stroke, and Alzheimer’s disease.
The clear lack of a causal effect of ANM on the outcomes of coronary heart disease and ischemic stroke in the MR analyses despite a strong inverse association seen in the observational data of this study (without MR) suggests residual confounding plays a substantial role in driving the observed outcomes.

The authors said that, to their knowledge, this is the first study that has shown a causal association between older ANM and higher risk of postmenopausal lung cancer.
This finding was directionally opposite to the significant protective effect of increased ANM documented in an observational analysis of roughly the same data as well as prior reports that did not use MR. This “notable inconsistency” suggests very substantial residual confounding without MR that could be driven by factors such as smoking, diet, and exercise.
If these results are replicated in additional datasets, it would highlight a need for randomized, controlled trials of antiestrogen therapies in postmenopausal women for the prevention or treatment of lung cancer.

The study included data from 106,853 postmenopausal women enrolled in the Women’s Health Initiative (WHI) and 95,464 women who were 37-73 years old included in the UK Biobank (UKB). Analyses for each outcome also included data from smaller numbers of women obtained from several additional datasets.
The MR analysis used up to 55 single-nucleotide polymorphisms previously discovered through a genome-wide association study of about 70,000 women of European ancestry and independent of all datasets analyzed in the current study. The authors included all single-nucleotide polymorphisms with a consistent direction of effect on ANM.
The MR analysis for lung cancer included 113,371 women from the two primary datasets and an additional 3012 women from six additional datasets.
The MR analysis for bone fracture involved 113,239 women from the WHI and UKB only. The MR analysis for osteoporosis involved 137,080 women from the WHI, UKB, and one additional external dataset.

Results from a meta-analysis of the MR results using data from the WHI, UKB, and the additional datasets showed ANM was causally associated with an increased risk of lung cancer by an odds ratio of 1.35 for each 5-year increase in ANM. In contrast, the adjusted observational analysis of data just from the WHI and UKB showed a significant 11% relative risk reduction in the incidence of lung cancer for each 5-year increase in ANM.
The MR results also showed causally protective effects for fracture, with a 24% relative risk reduction, and for osteoporosis, with a 19% relative risk reduction for each 5-year increase in ANM.
The MR analyses showed no significant association between AMN and outcome for breast cancer, endometrial cancer, ovarian cancer, coronary heart disease, ischemic stroke, and Alzheimer’s disease.