scale epidemiological studies focusing on alcohol use, although this area has received some research attention. Nonetheless, reviews of lesbian health research consistently include alcohol abuse as a problem for which lesbians appear to be at greater risk than heterosexual women (Cassidy and Hughes, 1997; Eliason, 1996; Haas, 1994; O'Hanlan, 1995; Rosser, 1993; Skinner and Otis, 1996), and alcohol abuse has been widely viewed as a prevalent and serious problem among lesbians (Cabaj, 1992, 1996; Finnegan and McNally, 1990; Glaus, 1989; Hall, 1993; NGLTF, 1993; Skinner, 1994).
Data across a wide range of non-probability small-sample studies suggest that about 30% of lesbians may have alcohol problems (Bloomfield, 1993; Hall, 1993). However, this estimate may be inflated since these studies have generally had a number of methodological problems, including the fact that subjects have often been recruited using convenience sampling from settings in which alcohol consumption is likely to occur (e.g., bars). Further, it has been suggested that contemporary patterns of alcohol use among lesbians may be lower because bars have become a less important component of the lesbian culture as other options for social gathering have become increasingly available (Hall, 1993).
Despite methodological limitations of the research on alcohol use among lesbians (e.g., the use of samples from bars or other settings likely to attract people who consume higher levels of alcohol), some tentative patterns emerge across studies: