Acute panic inventory symptoms during CO2 inhalation and room-air hyperventilation among panic disorder patients and normal controls

There is scant literature on anxiety symptoms induced during respiratory ch allenges developed to induce panic symptoms and attacks. Here we report on the prevalence of Acute Panic Inventory (API) symptoms during three consecu tive respiratory challenges to patients with panic disorder (PD) and normal controls (NC). The challenges performed using a closed canopy system inclu ded voluntary room air hyperventilation (RAM, inhalation of 5 % CO2, and 7 % CO2-enriched air The PD patients were 41 men and 53 women whose mean age was 33.4 (SD = 8.55). The normal comparison group consisted of 35 men and 2 7 women with a mean age of 31.3 (SD = 9.21). The diagnosis of panic disorde r was made using the Structured Clinical Interview for DSM-III-R. All poten tial normal controls underwent structured clinical interview using the Sche dule for Affective Disorders and Schizophrenia-Lifetime Version Modified fo r the Study of Anxiety Disorders (SADS-LA), and must have been free of a li fetime history of anxiety disorders, affective disorders, substance use dis orders, and schizophrenia. All participants also had a complete medical eva luation and were in good health. The experiment consisted of seven experime ntal epochs. three baseline/recovery periods each followed by a respiratory challenge, and then a final recovery epoch. The AN was administered at the end of each epoch. Clinical staff trained and experienced in rating panic attacks rated participants' response during each challenge as panic or no p anic. Three groups were defined for analysis: PD patients who panicked, PD patients who did not panic, and NC who did not panic. Staff ratings indicat ed that the 7 % CO2 challenge was the most panicogenic, followed by the 5 % CO2, and the RAH challenges. Conventional statistics (analysis of variance and partial correlations) indicated that many baseline symptoms as well as symptom increments differed across groups, and were associated with the ou tcome of panic/no panic during each challenge. However, logistic regression analysis indicated that only a few symptoms independently predicted the pa nic/no panic outcome because many symptoms were redundant. The symptom clus ter of fear in general, dizziness, difficulties with concentrating, and doi ng one's job pre-dicted panic to RAH. The cluster of fear in general, confu sion, dyspnea, and twitching/trembling predicted the response to 5 % CO2. F inally, fear in general, confusion, twitching/trembling and dizziness predi cted the response to 7 % CO2. While univariate analyses indicated that many symptoms distinguished between panic and no panic outcome, logistic regres sion revealed that group differences were subsumed under a few prominent sy mptoms, namely, fear in general, confusion, dizziness, twitching/trembling, and dyspnea. The results are discussed in the context of patient (having a diagnosis of PD) and panic effects (rated as panicking to a challenge). (C ) 2001 Wiley-Liss, Inc.