Vegetables, fruit, and lung cancer in the Iowa Women's Health Study.Cancer Res 1993; 53(3):536-43CR
Previous epidemiological studies have shown an inverse association between vegetable and fruit consumption and lung cancer risk; few of these studies have been prospective or have focused upon women. In 1986, we assessed food intake in 41,837 Iowa women, aged 55 to 69 yr, with a mailed 127-item food frequency questionnaire. After 4 years of follow-up, 179 incident cases of lung cancer were identified via the Iowa Surveillance, Epidemiology, and End Results cancer registry. After specific exclusion criteria were applied, a nested case-control comparison of 138 cases with 2,814 randomly selected noncases was undertaken. Intakes, in the upper-most quartile, of 11 vegetable and fruit groups, as well as of the nutrients beta-carotene and vitamin C, were explored. High intakes of all vegetables and fruit, all vegetables, and green leafy vegetables were each associated with an approximate halving of risk: age-, smoking-, and energy-adjusted odds ratios (ORs) were 0.49 (95% confidence interval, 0.28-0.86), 0.50 (95% confidence interval, 0.29-0.87), and 0.45 (95% confidence interval, 0.26-0.76), respectively. A lower lung cancer risk was also seen for all fruit (adjusted OR = 0.75 for high consumption), high vitamin C vegetables and fruit (OR = 0.75), carrots (OR = 0.71), and brocolli (OR = 0.72) and for the nutrients beta-carotene (OR = 0.81) and vitamin C (OR = 0.81) (all 95% confidence intervals included 1.0). Lung cancer risk was unrelated to consumption of the three food groups defined as "high-carotenoid" (beta-carotene, lutein, and lycopene) and tomatoes. In an analysis stratified by histological type of lung cancer, the strongest inverse associations for vegetables and fruit were seen for large cell carcinoma. Analysis by smoking status showed the inverse associations for most vegetable and fruit groups with lung cancer risk to be stronger for exsmokers than for current smokers. Results from the stratified analyses must be interpreted with caution because of the small number of cases in each stratum.