Epidemiologic survey on lung cancer with respect to cigarette smoking and plant diet.

This case-control study of lung cancer was based on a cross-sectional questionnaire survey of inpatients at 5 general hospitals in Okinawa, Japan, from 1982 to 1987. The purpose of the study was to clarify the relations of lung cancer to cigarette smoking and plant diet. Ingestion frequencies of 17 major dietary plants and/or herbs were obtained by means of a questionnaire interview. As eligible subjects for a case-control analysis, there were 673 respondents aged over 30 years with clear smoking history, age, sex and diagnosis. Psychiatric patients were excluded. Odds ratios of newly diagnosed lung cancer were calculated by the Mantel-Haenszel procedure. A pair consisted of a case and two controls which were selected randomly by using multivariate caliper matching. Sixty-four pairs matched for age (+/- 5) and sex showed a significantly high odds ratio of 2.9 (P less than 0.0005). However, three male groups who were categorized by the number of cigarettes smoked did not exhibit dose-dependency of lung cancer on smoking. Lung cancer was more prevalent in ex-smokers than in current smokers. Case-control analyses by male generations revealed that lung cancer incidence was age-dependent, and there was a clear dose-response relationship between smoking and lung cancer in males in their sixties. A case-control analysis of each of 17 edible plants based on 44 pairs who were matched for age (+/- 5), sex and smoking history demonstrated that the odds ratio of aloe (Aloe arborescens Mill var. natalensis Berger) was 0.5 (P less than 0.1), suggesting that the aloe may prevent human carcinogenesis at various sites.