The HPA axis has also been studied in patients with anxiety disorders, particularly panic disorder, with and without comorbid major depression. Both the cortisol response to dexamethasone and the response to CRH have been examined in pure panic disorder without comorbid depression. The earliest study with dexamethasone demonstrated a 15% nonsuppression rate in panic disorder (Curtis et al. 1982). A number of other studies have since been conducted, and the overall incidence of cortisol nonsuppression is 17% in panic disorder (13 studies), while the incidence for major depression is 50% (Heninger 1990). Grunhaus et al. (1987) compared patients with major depression to those with major depression with panic disorder and found a similar rate of cortisol nonsuppression following dexamethasone administration (approximately 50%) in the two populations, suggesting that the presence of comorbid panic disorder had little impact beyond that of depression on dexamethasone nonsuppression. In CRH challenge tests, panic disorder patients have demonstrated a decreased integrated ACTH response relative to control subjects in some studies (Holsboer et al. 1987; Roy-Byrne et al. 1986) but a normal response in others (Brambilla et al. 1992). Similar to findings in depression, "baseline" plasma cortisol was increased in panic patients who showed blunted CRH responses. A study of CRH challenge in panic disorder patients (Curtis et al. 1997) demonstrated a normal response to CRH challenge.
Studies of the HPA axis in social phobia have not found evidence of baseline hyperactivity by urinary free cortisol (Uhde et al. 1994), although few challenges other than a social speaking task have been used. Public speaking challenges in anxiety disorders do not support an exaggerated ACTH or cortisol response to this stressor (Gerra et al. 2000; Levin et al. 1993; Martel et al. 1999). A few studies in children with anxiety disorders have also examined the HPA axis. Children with anxiety disorders coming in for a CO2 challenge demonstrated elevated "basal" cortisol in those who panicked in response to CO2, suggesting that increased reactivity to a threatening situation (i.e., anticipation of a procedure that would cause discomfort) was linked to activation of the HPA axis (Coplan et al. 2002). This interpretation is further supported by an extremely large study of basal 24-hour cortisol in normal children and children with either anxiety disorders or major depression, which found lower nighttime cortisol and a sluggish morning rise in cortisol in children with an anxiety disorder. This suggests that anxiety disorders lead to stress hyperreactivity (in the case of anxious children, in the context of a threatening experimental procedure of CO2) with compensatory decreased basal cortisol secretion (24-hour study) (Feder et al. 2004).
Overall, the studies to date do not suggest HPA axis hyperactivity in anxiety disorders to the same extent as shown in depression (Abelson and Curtis 1996). Feedback elements are generally normal, and the abnormalities that do exist may reflect "extrinsic" factors that contribute to heightened reactivity within activational elements of the system. The question of whether stress-activated HPA axis elements are abnormal in comorbid major depression and anxiety has not yet been well studied.
Was this article helpful?