Third, results have been remarkably consistent despite major differences between samples in the methods used to diagnose alcohol problems, ranging from hospitalization or other treatment records and Swedish Temperance Board registrations to direct interview assessments, which, in community samples, include a relatively high proportion of “mild” cases. This pattern would only be expected if the same risk factors, genetic or environmental, operate across the entire spectrum of alcohol problems, from mild to severe. Kendler and colleagues (1992) noted this finding in women.
