We examined the relation between gallstones, cholecystectomy, and the development of pancreatic cancer in the Nurses' Health Study and the Health Professionals Follow-up Study. Among 104 856 women and 48 928 men without cancer at baseline, we documented 349 cases of pancreatic cancer during up to 16 years of follow-up. Participants were classified according to a history of gallstones or cholecystectomy. The age-adjusted relative risk of pancreatic cancer following cholecystectomy or diagnosis of gallstones was 1.31 (95% CI, 0.93 -1.83). However, adjustment for other pancreatic cancer risk factors attenuated the association (RR=1.11, 95% CI, 0.78 -1.56); this risk did not increase with increasing time following cholecystectomy or gallstones. Gallstones or cholecystectomy do not appear to be significant risk factors for pancreatic cancer. Pancreatic cancer is the fifth leading cause of cancer-related mortality in the US (Greenlee et al, 2001). New environmental risk factors have consistently been linked to the risk of pancreatic cancer. Animal models suggest an important role for cytokines in experimental pancreatic carcinogenesis (Watanapa and Williamson, 1993). Cholecystokinin (CCK) has been shown to stimulate both gallbladder contraction and pancreatic enzyme secretion (Marx et al, 1987;Smith et al, 1990) and to have a trophic effect on pancreatic acinar cells (Rivard et al, 1991). The presumed increased release of CCK following cholecystectomy (Hyvarinen and Partanen, 1987) might contribute to an increased risk for pancreatic cancer risk following gallbladder removal. In addition, cholecystectomy may suppress the normal inhibitory effect of CCK on the Sphincter of Oddi (Luman et al, 1997). The presence of gallstones, on the other hand, appears to be associated with chronic pancreatitis (Hardt et al, 2001), but whether chronic pancreatitis predisposes to pancreas cancer remains controversial (Lowenfels et al, 1993;Ekbom et al, 1994). Some of the earlier studies reported relative risks of pancreatic cancer of 1.2 -2.0 among individuals who had undergone a cholecystectomy previously (Haines et al, 1982;Shibata et al, 1994;Ekbom et al, 1996;Chow et al, 1999;Gullo, 1999;Silverman et al, 1999;Coughlin et al, 2000); however, the strength of the association remains uncertain because of the retrospective design of most analyses and the reliance on nextof-kin respondents. Prospective studies can overcome these limitations. We examined the relation between cholecystectomy or gallstones and pancreatic cancer risk in two large cohorts, the Nurses' Health Study (NHS) and the Health Professionals Follow-up Study (HPFS).
MATERIALS AND METHODS
Study cohortsThe Nurses' Health Study (NHS) enrolled 121 700 female, registered nurses aged 30 -55 years in 1976, and information on their health status, medical history, and known and suspected risk factors for cancer and coronary heart disease was gathered through mailed questionnaires. The Health Professionals Follow-up Study (HPFS) began in 1986, when 51 529 American male dentists, o...