Three studies in Scandinavia have matched twin registries to national databases containing hospital discharge data. In Finland, Koskenvuo and colleagues (1984) conducted such a match using only an alcoholism discharge code and found a significantly higher risk ratio for male MZ than for male DZ twins of males hospitalized for alcoholism (i.e., 11.8 versus 5.5). Romanov and colleagues (1991) conducted a further followup of a subsample of the same twin cohort using a broader definition of alcoholism that included alcohol abuse-related discharge codes (e.g., alcoholic liver cirrhosis and alcoholic psychosis) and found a nonsignificant trend in the same direction (i.e., risk ratios of 8.8 versus 4.6, respectively). In Sweden, Allgulander and colleagues (1991, 1992) performed a similar match of the Swedish twin registry to alcohol-related discharge codes; these data again showed an elevated risk ratio in MZ, compared with DZ, male twins (i.e., 7.9 versus 5.3), although again the difference was not significant. For women, Koskenvuo and colleagues (1984) found no same-sex twin pairs in which both twins had an alcoholism discharge code, making it impossible to estimate a risk
