Objective: Documentation of hypothalamic-pituitary-adrenal (HPA) axis distu
rbance in panic disorder has been inconsistent. Increased cortisol levels h
ave been associated with altered HPA function due to stress. The authors ex
amined salivary cortisol levels in spontaneously occurring, unprovoked pani
c attacks. Method: Patients with panic disorder (N = 25) collected saliva s
amples when panic attacks occurred. Levels of cortisol in the saliva sample
s were determined and were compared with levels in comparison samples of sa
liva obtained 24 hours after the panic attack occurred. Results: During spo
ntaneous panic attacks there was a subtle but significant elevation of cort
isol levels, compared with levels obtained 24 hours later. No significant c
orrelations were found between the cortisol elevations during panic attacks
and the severity of the attack as measured by using the Acute Panic Invent
ory or the severity of illness as measured by using the Panic and Agoraphob
ia Scale. Conclusions: Saliva sampling may be a useful method for investiga
ting neuroendocrine parameters during spontaneously occurring panic attacks
.