This study was done to determine the strength of association between substa
nce abuse and panic states, including subsyndromal panic, its temporal rela
tionship, and self-medication for panic using abusable substances. A commun
ity-based sample was screened for panic using DSM-III-R criteria. Panic and
matched control groups participated in a structured interview concerning t
he presence of substance abuse, use of substances to treat panic symptoms,
and the age-of-onset of panic and substance abuse. Of 97 individuals with p
anic, 39% had abused at least one substance. None of the panic disorder-sub
syndromal panic differences reached significance. Only 10% of subjects repo
rted using alcohol and 6% reported ever using illicit drugs to treat their
panic. The majority (63%) of those abusing alcohol reported that alcohol us
e began prior to onset of panic, and the majority (59%) of those abusing il
licit drugs reported that drug use began first. This study documents the pa
nic-substance abuse relationship even in those with subsyndromal panic. Sub
stance abuse began prior to onset of panic and substances were used to self
-medicate for panic attacks by only a few subjects. (C) 1999 Elsevier Scien
ce Ltd.