Why do some studies find lower endometrial-cancer risk among people who smoke, without justifying smoking?

01

The cancer site and subgroup matter

NCI summarises lower observed incidence in some smoking groups, with results concentrated in current-smoking or postmenopausal strata rather than a universal pattern.

Endometrial cancer is not cervical, ovarian or every uterine cancer, and incidence is not the same endpoint as mortality or treatment outcome.

[1]

02

A proposed hormone pathway is not an intervention

Hormonal hypotheses can motivate research, but a proposed oestrogen-related pathway does not demonstrate a safe or preventive intervention.

Menopause, body composition, reproductive and hormone history, competing mortality and smoking classification can reshape subgroup estimates.

[1]

03

What Mendelian randomization adds—and does not add

NCI contrasts conventional observational findings with analyses using genetic proxies for smoking. Those analyses did not support a causal reduction in endometrial-cancer risk in the cited cohorts. Genetic-proxy analysis is not a trial assigning people to smoke, and it has its own assumptions. The useful conclusion is that the inverse association does not establish a smoking prevention strategy, not that a genetic result predicts someone’s risk or explains all subgroups.

[1]

04

Do not turn stopping into a cancer warning

These findings do not show that quitting causes a predictable rise in personal cancer risk. Do not start smoking, continue it, use nicotine, or change hormones to reproduce an observational association. Questions about bleeding, diagnosis or care belong with qualified local clinical services.

[1][2]