Abstract
Lung cancer among nonsmokers has emerged as a distinct clinicopathologic entity for which the etiology is still poorly understood, but which accounts for a significant proportion of the lung cancers among women. Although estrogens have been shown to have mitogenic effects in lung cells and interact with growth factor pathways in tumorigenesis, epidemiologic evidence for a link between reproductive hormones and lung cancer is sparse and inconsistent. We examined the effect of parity, age at menarche/menopause, cycle length and use of exogenous hormones, and dietary soy and soy isoflavonoid intake on lung cancer risk in a prospective cohort of middle-aged and elderly Chinese women in Singapore among whom 91% were lifetime nonsmokers. Among 35,298 women (mean follow-up time, 9.6 years), 298 cases of incident lung cancer were recorded, of which 189 (63.4%) occurred in nonsmokers. Compared with nulliparous women, those with one to two, three to four, and more than five livebirths had relative risks of between 0.49 and 0.59 (P for trend<0.01) for all lung cancers, and between 0.32 and 0.42 (P for trend<0.001) for adenocarcinomas. This relationship was observed in both smokers and nonsmokers. Age at menarche and menopause did not seem to influence risk. Dietary soy isoflavonoid intake was associated with a statistically significant inverse trend among nonsmokers only (relative risks, 0.59 for highest versus lower quartile; P for trend, 0.021). These findings add support for the role of hormonal factors in the etiology of lung cancer among nonsmoking women, and are consistent with emerging experimental evidence in this regard.