a higher threshold for alcohol related problems before receiving treatment. Indeed, Dawson (1996) found that, at the most common levels of alcohol dependence severity (5 to 10 symptoms), men were about 50% more likely than women to have received treatment, while at the most severely dependent (20 or more symptoms), they were equally likely. This gender by severity interaction may lead to over-representation of the most severely affected women in treatment samples. In contrast, samples of men and women drawn from treatment-naïve alcoholic populations may not have this bias and, hence, may be more likely to have comparable preservation/impairment of cognitive function across gender. In fact, after controlling for lifetime alcohol consumption, Pfefferbaum and colleagues (2001) did not find gender differences in alcoholics in cortical gray and white matter, and found less abnormality in cortical sulcal volumes for alcoholic women than for alcoholic men. Moreover, Yonker and colleagues (2005) reported that moderate to heavy drinking women performed somewhat better on episodic memory and spatial visualization tasks than did non/light drinking women, a pattern not found in men. Finally, Sullivan and colleagues (2002) have found that, while cognitive deficits in abstinent alcoholic men are most often reported in visuospatial and executive
