There is now substantial evidence that an abnormal threshold for suffo
cation alarm underlies panic disorder. Because this disorder is highly
familial, evidence of an abnormal suffocation threshold may be appare
nt in high-risk individuals before they develop clinical illness. To e
xplore this possibility, we used a single inhalation of 35% CO2 vs. ai
r to evaluate II subjects who had at least one first-degree relative w
ith DSM-III-R panic disorder, 13 who had at least two relatives treate
d for mania or for depression (HR-AD), and 15 low-risk controls who ha
d no family history of panic disorder, affective disorder, or alcoholi
sm (LR-C). All were aged 18-34 and had no history of panics or of any
Research Diagnostic Criteria disorder. Five (45.5%) of the subjects at
high risk for panic disorder, but none of the LR-C subjects (p = .007
), nor any of the HR-AD subjects (p = .011), developed a panic attack
following inhalation of the CO2 mixture. (C) 1997 Society of Biologica
l Psychiatry.