± 3.8; F1,155 = 17.5, p < 0.0001, es = 10.1), and a trend toward greater alcohol doses (199.9 ± 156 vs. 157.3 ± 122 drinks/mo; F1,155 = 3.2, p = 0.078, es = 2.0). The imaging studies on Oahu were carried out on a 3.0T magnet vs. a 1.5T magnet in California, with 5 year newer software and hardware in Oahu. LTAA in Oahu were not as healthy as those in California (much higher BMI) and were much less highly educated than those in California, possibly indicating less brain functional reserve capacity — both of these factors may have limited their capacity for recovery from the brain morbidity of alcoholism (thus evidencing lower than NSAC subcortical volumes even after multi-year abstinence). Additionally, their greater alcohol use may have given them a greater alcohol brain morbidity burden to recover from. Finally, the improved imaging protocol may have been more sensitive in measuring the effects of alcoholism on subcortical structures. We believe the major import of this result is that one cannot generalize from one study of LTAA to the population of LTAA. LTAA cohorts may differ in alcoholism severity and in their ability to recover from such burden. Finally, given
