The present study addresses a number of problems that have been encountered in association studies on alcohol dependence. In many case–control studies, the data have been collected among patients who have sought or received treatment for alcohol addiction thus representing a selected subpopulation of alcohol-dependent individuals. Many studies are also hampered by significant phenotypic variation and heterogeneity as well as ethnic diversity and insufficient size of study populations. In the current study, all subjects were diagnosed using the same diagnostic DSM-IV codes. The population-specific dataset is large, with age- and gender-matched cases (n = 503) and controls (n = 506) allowing detection of relative risk >1.4. The Finnish population provides an excellent basis for genetic analyzes as it has remained genetically very homogeneous when compared with the majority of other populations of Caucasian origin. We aimed to further limit the ethnic variation by excluding persons with a mother tongue other than Finnish or Swedish (the two official languages in Finland). With respect to analyses on alcohol consumption, the populations were of substantial size thus giving enough power to detect even
