Results of studies of sons of alcoholics (SOAs) may vary depending on
the density of the family history of alcoholism of the SOA subjects se
lected. To assess how reliably SOA subjects report a history of alcoho
lism in first- and second-degree relatives, we compared family histori
es obtained from 20 SOAs with those provided by their fathers. In all
20 cases, SOAs and their fathers agreed that the father met criteria f
or alcohol dependence but for no other primary Axis I psychiatric diso
rder. However, agreement was not as good for whether other paternal re
latives in addition to the father were affected (kappa = 0.36) or for
whether the family history was unigenerational or multigenerational fo
r alcoholism (kappa = 0.44). Whether this discrepancy was due to under
-reporting by the son or over-reporting by the father is unclear. Thes
e results suggest that 18-25 year old SOAs are reliable sources regard
ing alcohol dependence in their fathers, but that for determination of
density of the family history of alcoholism, other relatives should a
lso be interviewed directly.