An investigation of excess pancreatic cancer mortality among workers in a chemical manufacturing plant failed to find an association with alcohol consumption or any dietary factor, other than an inconsistent association with decaffeinated coffee (108). However, other studies have identified possible diet-pancreatic cancer relationships (109-113). Case- control studies were conducted in Adelaide, Australia; Montreal and Toronto, Canada; Utrecht, The Netherland; and Opole, Poland (109). These five populations have moderate-to-high rates of pancreatic cancer and distinctive dietary practices (not specified). Comprehensive diet histories were compiled using a common protocol and questionnaire. Positive dose-related associations were consistently found for intake of carbohydrates and cholesterol, and negative associations for intake of dietary fiber and vitamin C. There was no association with weight or body mass index. In Adelaide, case-subjects consumed more boiled eggs, omelettes, and sweet and fatty food items than controls (110). They consumed less of certain vegetables, fruits, and several nutrients derived principally from plant foods. Pan creatic cancer was not associated with coffee drinking. In Opole, the association between pancreatic cancer and cholesterol intake (but not serum cholesterol) was particularly strong (111). There was a strong inverse association with vitamin C intake and weak inverse associations with retinol and fiber. The data suggested an inverse association between pancreatic cancer and intake of PUFAs and positive associations with carbohydrate and perhaps protein. In Utrecht, after controlling for confounding factors, significant inverse dose-related effects were seen for vegetables, particularly cruciferous vegetables (112). Positive dose-response gradients were found for consumption of eggs and fish. Low-fiber vegetables were more strongly protective than high-fiber vegetables in directly interviewed subjects, pointing to protective elements in addition to fiber. Vitamin C was protective in women but not in men. Consumption of beer, spirits, red wine, and fortified wine was not related to risk for pancreatic cancer in Utrecht; there was an inverse relationship for white wine consumption, but only a small number of subjects were involved (113). There was no association with lifetime consumption of tea or instant coffee, caffeinated or decaffeinated; an inverse relationship with lifetime total coffee consumption failed to reach significance (p<.06).