Table 11.2 is a listing, taken from the two references cited above, of those agents or conditions of lifestyle causally related to specific neoplasms in humans on the basis of epidemiological studies. The carcinogenicity of alcoholic beverages for human is related mostly to their action in associa- tion with another known carcinogenic agent. The combination of excessive alcohol ingestion and smoking markedly increases the risk of men and women for oral and laryngeal cancer compared with either of these two agents alone or in their absence (Herity et al., 1982; Day et al., 1993; Blot et al., 1994). The combination of ethanol ingestion in the form of alcoholic beverages and smoking also increased the risk of several cancers of the digestive tract, including the esophagus, rectum, and liver (Choi and Kahyo, 1991) in a cohort in Korea. Several studies have also impli- cated the combination in enhancing the development of premalignant lesions in the colon (Mar- tiñez et al., 1995; Boutron et al., 1995) and in the esophagus (Castelletto et al., 1992). Cancer of the liver is associated with excessive alcohol consumption (Tuyns, 1979; Adami et al., 1992), and data reported by Ohnishi et al. (1982) suggest that habitual alcohol intake can promote the devel- opment of hepatocellular carcinoma in patients infected with the hepatitis B virus, a known onco- genic virus for the human liver (see below). Similarly, excessive intake of alcoholic beverages enhances the carcinogenic effect of the hepatitis C virus, an RNA virus oncogenic for human liver (Donato et al., 1997). While excessive ingestion of alcoholic beverages has also been associ- ated with esophageal, rectal, and pancreatic cancer, a much more extensive series of investiga- tions has been concerned with the potential relationship of alcoholic beverages and breast cancer. Although a causal relationship between alcohol ingestion and breast cancer has been disputed by some (Schatzkin and Longnecker, 1994), there is substantial epidemiological evidence from sin- gle and multiple studies for such an association (Smith-Warner et al., 1998). In Figure 11.1 may be seen a regression curve for the relationship between total alcohol intake per day and the risk of developing breast cancer. This study involved a pooled analysis of six prospective studies involv-
Figure 11.1 Nonparametric regression curve for the relationship between total alcohol intake and breast cancer. (From Smith-Warner et al., 1998, with permission of the authors and publisher.)
ing more than 300,000 women evaluated for up to 11 years. While there does appear to be some form of a threshold below 10 g/day, the relationship of significant risk to daily alcohol intake of 30 g or more appears very solid. Interestingly, consumption of alcoholic beverages reportedly is protective against the development of endometrial (Webster et al., 1989) and ovarian cancer (Gwinn et al., 1986). Although a variety of alcoholic beverages contain known chemical carcin- ogens, including nitrosamines (Tuyns and Griciute, 1980), and different forms of alcoholic bev- erages may have slightly different influences on the incidence of certain types of human cancer, most evidence suggests that the principal effect is due to the alcohol itself and is largely independent of the form in which it is drunk (Doll and Peto, 1981).