One problem with drawing firm conclusions about safe levels of drinking from epidemiologic studies is that such studies depend upon self-reporting of alcohol consumption, which is known to underestimates true consumption. Indeed, the authors of a large and influential meta-analysis (Hamajima et al., 2002) stated this point clearly, writing that “…self-reported information on alcohol consumption is known to underestimate true consumption…” and that “systematic under-reporting of consumption by both cases and controls would result in an overestimation of the relative risk of breast cancer for a given level of alcohol consumption.” Of course, it is precisely this issue that is the most relevant from a public health perspective, and for providing guidance for safe levels of drinking. In addition, as recently noted (Jimenez et al., 2012) underreporting alcohol drinking could potentially lead to spurious associations at lower levels of consumption.
