Usefulness of population controls in retrospective studies of alcohol consumption. Experience from a case--control study of esophageal cancer in Ille-et-Vilaine, France.

Abstract

As part of a study on the etiological factors in esophageal cancer in northwestern France, information on alcohol consumption was obtained from 200 men with esophageal cancer, interviewed in medical facilities; 778 men from the general population of the Departement of Ille-et-Vilaine, interviewed in their homes; and 3 hospitalized age-matched control groups, which consisted of 200 patients with cancer of a site other than the esophagus or gastrointestinal tract, 200 patients from a medical ward, having no alcohol-related disease, and 200 patients from the surgical ward (excluding victims of traffic accidents). When hospital controls with alcohol-associated cancers (buccal cavity and pharynx) and residents of other departments were excluded, distribution of alcohol use was identical in the hospital control and population control samples. The setting of the interviews did not influence the subjects'' responses, and data on alcohol consumption obtained from hospital populations are comparable with those from general population samples.