What consumers eat 9 (Mensink and Katan, 1987). Refined carbohydrates substantially affect post- prandial hyperglycemia and they appear to accentuate insulin resistance. With respect to other nutrients, there is converging, but not yet conclusive, evidence that moderate alcohol intake, vitamin E and folic acid are inversely associated with the risk of coronary heart disease(CHD)( Gaziano et al, 1993; Stampfer et al, 1993; Robinson et al, 1998). Salt intake, on the contrary, contributes to the elevation of blood pressure levels in susceptible individuals and thus to the crease of CVD risk(Beilin et al, 1999). The evidence on the role of specific dietary factors in cancer aetiology has en critically summarised in recent reviews(Willett and Trichopoulos, 1996 Willett, 2000). With respect to food groups, vegetable consumption, and perhaps less definitely fruit consumption, have a beneficial effect on a broad spectrum of human cancer types. Among macronutrients, animal protein intake has been reported to increase the risk for colorectal cancer, while intake of saturated fat is positively associated with endometrial, prostate, colorectal, lung and kidney cancer. Although the percentage of calories from dietary lipids does not appear related to colon cancer, greater risks have been seen with higher consumption of red meat, possibly suggesting that factors other than dietary lipids per se may be important. Fibre intake, on the contrary, appears to protect against cancer of the pancreas and the large bowel. There are also indications of a protective role of monounsaturated lipids against breast cancer(Trichopoulou, 1995). Concerning micronutrients the evidence is largely insufficient. Recent studies indicate an inverse association of lycopene(Gann et al, 1999), selenium(Yoshizawa et al, 1998)and vitamin E(Tzonou et al, 1999)with prostate cancer, folic acid in rela- tion to colon and breast cancers( Giovannucci et al, 1998); while beta-carotene supplements have been found to be ineffective against lung cancer risk (Hennekens et al, 1996) Consumption of large quantities of alcoholic beverages, particularly in con- junction with tobacco smoking, has been reported to increase the risk of cancer in the upper respiratory and digestive tract, whereas alcoholic cirrhosis frequently leads to liver cancer. There are also data suggesting that intake of smaller quan- tities of alcohol may be linked to the occurrence of breast and colorectal cancer. Among added substances, only salt appears to be an important contributor to stomach cancer. Moreover, intake of salty fish very early in life has been linked to the occurrence of nasopharyngeal cancer in Southern Asia. Finally, in Central Asia and Southern America the intake of very hot drinks has been found to increase the risk of esophageal cancer(Kinjo et al, 1998) Many of the early efforts have been focused on identifying specific dietary components that may be responsible for effects on people's health. Evaluating the effects of specific foods and nutrients, rather than integral dietary patterns, on disease illustrates how shifting from the empirical evidence may increase uncer- tainty. Dietary exposures are unusually complex and strongly intercorrelated Current data suggest that apparently favourable effects cannot be exclusively attributed to specific components and in several instances these components may act synergistically( Gerber et al, 2000). Consequently, instead of focusing only(Mensink and Katan, 1987). Refined carbohydrates substantially affect post￾prandial hyperglycemia and they appear to accentuate insulin resistance. With respect to other nutrients, there is converging, but not yet conclusive, evidence that moderate alcohol intake, vitamin E and folic acid are inversely associated with the risk of coronary heart disease (CHD) (Gaziano et al, 1993; Stampfer et al, 1993; Robinson et al, 1998). Salt intake, on the contrary, contributes to the elevation of blood pressure levels in susceptible individuals and thus to the increase of CVD risk (Beilin et al, 1999). The evidence on the role of specific dietary factors in cancer aetiology has been critically summarised in recent reviews (Willett and Trichopoulos, 1996; Willett, 2000). With respect to food groups, vegetable consumption, and perhaps less definitely fruit consumption, have a beneficial effect on a broad spectrum of human cancer types. Among macronutrients, animal protein intake has been reported to increase the risk for colorectal cancer, while intake of saturated fat is positively associated with endometrial, prostate, colorectal, lung and kidney cancer. Although the percentage of calories from dietary lipids does not appear related to colon cancer, greater risks have been seen with higher consumption of red meat, possibly suggesting that factors other than dietary lipids per se may be important. Fibre intake, on the contrary, appears to protect against cancer of the pancreas and the large bowel. There are also indications of a protective role of monounsaturated lipids against breast cancer (Trichopoulou, 1995). Concerning micronutrients the evidence is largely insufficient. Recent studies indicate an inverse association of lycopene (Gann et al, 1999), selenium (Yoshizawa et al, 1998) and vitamin E (Tzonou et al, 1999) with prostate cancer, folic acid in rela￾tion to colon and breast cancers (Giovannucci et al, 1998); while beta-carotene supplements have been found to be ineffective against lung cancer risk (Hennekens et al, 1996). Consumption of large quantities of alcoholic beverages, particularly in con￾junction with tobacco smoking, has been reported to increase the risk of cancer in the upper respiratory and digestive tract, whereas alcoholic cirrhosis frequently leads to liver cancer. There are also data suggesting that intake of smaller quan￾tities of alcohol may be linked to the occurrence of breast and colorectal cancer. Among added substances, only salt appears to be an important contributor to stomach cancer. Moreover, intake of salty fish very early in life has been linked to the occurrence of nasopharyngeal cancer in Southern Asia. Finally, in Central Asia and Southern America the intake of very hot drinks has been found to increase the risk of esophageal cancer (Kinjo et al, 1998). Many of the early efforts have been focused on identifying specific dietary components that may be responsible for effects on people’s health. Evaluating the effects of specific foods and nutrients, rather than integral dietary patterns, on disease illustrates how shifting from the empirical evidence may increase uncer￾tainty. Dietary exposures are unusually complex and strongly intercorrelated. Current data suggest that apparently favourable effects cannot be exclusively attributed to specific components and in several instances these components may act synergistically (Gerber et al, 2000). Consequently, instead of focusing only What consumers eat 9