Voluntary breath holding in panic and generalized anxiety disorders

Objective: Because breath holding causes arterial pCO(2) to increase, we used it to test the hypothesis that in panic disorder (PD) a biological suffocation monitor is pathologically sensitive. Method: Nineteen patients with PD, 17 with generalized anxiety disorder (GAD), and 22 normal controls took deep breaths on signal and held them until a release signal was given 30 seconds later. This was repeated 12 times separated by 60-second normal breathing periods. Results: PD patients reported having had in the past more symptoms of shortness of breath when anxious, and more frequent frightening suffocation experiences than the other groups. However, increases in self-rated anxiety between periods of normal breathing and periods of breath holding were similar in all three groups. Skin conductance, blood pressure, and T-wave amplitude reactions to breath holdings were also similar, but heart rate acceleration upon taking a deep breath was greater in GAD patients. Before and after individual breath holdings, end-tidal pCO(2) was lower in PD patients than in normal controls; GAD patients were intermediate. Inspiratory flow rate did not differ between groups. Conclusions: Our physiological results provide no direct support for an overly sensitive suffocation alarm system in PD. Lower pCO(2) may be due to anxiety causing hyperventilation in patients prone to panic.