To examine the association between cigarette smoking and the risk of gastric cancer, we conducted a pooled analysis of 2 population-based prospective cohort studies in rural northern Japan. Cohort 1 included 9,980 men (>40 years old) and Cohort 2 included 19,412 men (40 -64 years old). The subjects completed a self-administered questionnaire on cigarette smoking and other health habits. We identified 228 cases of gastric cancer among Cohort 1 subjects (9 years of follow-up with 74,073 person-years) and 223 among Cohort 2 subjects (7 years of follow-up with 141,675 person-years). From each cohort, we computed the relative risk (RR) and 95% confidence interval (CI) of gastric cancer associated with smoking using a Cox regression analysis and pooled these estimates to obtain summary measures. The pooled multivariate RRs (95% CIs) for current smokers and past smokers compared to subjects who had never smoked were 1. An increased risk of gastric cancer associated with tobacco smoking has long been controversial. 1 Although the International Agency for Research on Cancer Working Group concluded in 2002 that there was "sufficient" evidence of causality, 2 it remains to be clarified if the effects of smoking differ by anatomic subsite or histologic subtype of gastric cancer. Furthermore, the majority of available evidence on smoking and gastric cancer are based on case-control studies 3-75 that are susceptible to selection and recall biases, and most prospective studies have methodologic limitations including the use of mortality rather than incidence as an endpoint, 76,78,79,81,82,84,85,[87][88][89] the use of a relatively small number (Ͻ300) of gastric cancer cases 77,79 -83,85-87,89 or no adjustment for dietary variables as potential confounders. 78 -89 To further examine the association between cigarette smoking and the risk of gastric cancer, we conducted a pooled analysis of 2 population-based prospective cohort studies in rural northern Japan, a high-risk area for gastric cancer. Our study involved 451 incidence cases of gastric cancer, and we adjusted for various potential confounders, including dietary variables. We also examined the association according to the anatomic subsite and histologic subtype of gastric cancer.
MATERIAL AND METHODS
Study cohortsThe study designs for the 2 cohort studies have been described in detail elsewhere. 90 -92 Briefly, Cohort 1 was started in January 1984, when we delivered a self-administered questionnaire to 33,453 men and women (40 years of age or older) in 3 municipalities of Miyagi Prefecture. Usable questionnaires were returned from 31,345 subjects (93.7%). For Cohort 2, we delivered a self-administered questionnaire from June to August 1990 to 51,921 men and women (40 -64 years of age) in 14 municipalities of the Prefecture. Usable questionnaires were returned from 47,605 subjects (91.7%). Study protocols for the 2 cohorts were approved by the institutional review board of the Tohoku University Graduate School of Medicine. We considered the return of self-administered...