Diet, reproductive factors and lung cancer risk among Chinese women in Singapore: Evidence for a protective effect of soy in nonsmokers

The factors associated with risk of lung cancer among nonsmokers have not been fully elucidated, but dietary factors have consistently been shown to play a role. Chinese women are unique in having a high incidence of lung cancer despite a low smoking prevalence. This population is also known to have a high intake of soy, a dietary source of phytoestrogens. We conducted a hospital-based case-control study among Singapore Chinese women, comprising 303 cases and 765 age-matched controls, of whom 176 cases and 663 controls were lifetime nonsmokers. Data on demographic background, reproductive factors and dietary intake of fruit, vegetables and soy foods were obtained by in-person interview. We observed an inverse association between intake of total, cruciferous and non-cruciferous vegetables and risk of lung cancer among smokers. Although smokers in the highest tertile of fruit intake also had a lower risk, this was not statistically significant. Higher intake of soy foods significantly reduced risk of lung cancer among lifetime nonsmokers, but not among smokers. When soy isoflavonoid intake in mg/week was computed based on frequency and portion size of intake of eight common local soy foods, the adjusted OR among nonsmokers for the highest tertile compared to the lowest was 0.56, 95% Cl 0.37-0.85 (p for trend <0.01). Fruit intake was also significantly associated with reduced lung cancer risk among nonsmokers, but the effect was not significant after adjustment for soy intake. On the other hand, soy intake remained an independent predictor of risk after controlling for fruit intake. Reproductive effects were also primarily confined to lifetime nonsmokers, among whom having 3 or more livebirths (adjusted OR 0.65, 0.44-0.96) and a menstrual cycle length of more than 30 days (OR 0.46, 0.25-0.84) accorded a significantly reduced risk of lung cancer. Place of birth was significantly associated with risk among nonsmokers (OR 2.6, 1.7-3.9 for China-born vs. local born) but not among smokers. When analysis was restricted to nonsmokers with adenocarcinomas, the dietary effects were consistent or enhanced. On stepwise regression, soy intake and cycle length emerged as the independent dietary and reproductive predictors of lung cancer risk in nonsmokers. These findings are consistent with other evidence suggesting an involvement of estrogen-related pathways in lung cancer among non-smoking women. (C) 2002 Wiley-Liss, Inc.