It is unclear whether body mass index (BMI) and physical activity are associated with the risk of pancreatic cancer in Asian populations. We examined these associations in the Japanese Collaborative Cohort Study for Evaluation of Cancer Risk. Our cohort study included 110,792 Japanese men and women at enrollment (1988)(1989)(1990). Data on height, body weight (at baseline and at age 20 years) and physical activity were obtained from a questionnaire. Cox proportional hazards models were used to estimate the relative risks of pancreatic cancer mortality. We observed a total of 402 pancreatic cancer deaths during the follow-up period. Men with a BMI of 30 or more at age 20 years had a 3.5-fold greater risk compared with men with a normal BMI. Women with a BMI of 27.5-29.9 at baseline had 60% increased risk compared with women with a BMI of 20.0-22.4. In men, weight loss of 5 kg or more between 20 years of age and baseline age was associated with an increased risk of pancreatic cancer death. In contrast, women with weight loss of 5 kg or more over the same period had a decreased risk. Physical activity was not associated with pancreatic cancer risk in either men or women. Obesity in young adulthood may be associated with an increased risk of death from pancreatic cancer in Japanese men. The risk of pancreatic cancer in relation to BMI seems to differ according to sex and the period over which BMI was measured. ' 2007 Wiley-Liss, Inc.Key words: body mass index; physical activity; pancreatic cancer; obesity; relative risk; cohort study Pancreatic cancer is the fifth leading cause of cancer-related death in Japan, accounting for 20,000 deaths each year. 1 The etiology of pancreatic cancer remains largely unknown. Although epidemiological studies have reported that many environmental factors may contribute to the development of pancreatic cancer, only age and cigarette smoking have been established as consistent risk factors for the disease. 2 The association between diabetes and pancreatic cancer is less consistent, although epidemiological evidence strongly suggests that long-standing diabetes is also a risk factor for pancreatic cancer. [3][4][5][6][7][8][9] Recently, several studies have shown a positive association between post-load plasma glucose concentration and pancreatic cancer risk, [10][11][12] suggesting that impaired glucose tolerance and insulin resistance may have a role in the etiology of pancreatic cancer. As there is an established link between obesity and significant metabolic abnormalities such as glucose intolerance, 13,14 a positive association between obesity and pancreatic cancer seems possible. However, there are conflicting epidemiological findings on the link between obesity and risk of pancreatic cancer. Several large prospective studies carried out in the U.S.A. have demonstrated that overweight and obese subjects have an increased risk of pancreatic cancer compared with normal weight subjects, 15-17 whereas other studies did not observe this association. [18][19][20][21] A meta-analysis c...