Back Home  Contact
Proceedings: 2003 2004 2005 2006  Next Conference
Contact Information
Eileen Wollburg, cand.-psych.

Walton Roth, MD

Department of Psychiatry and Behavioral Sciences,
Stanford University School of Medicine and
VA Palo Alto Health Care System,
Stanford University, United States

Psychophysiological Responses of Panic and Tense Patients to a Hyperventilation Test

Eileen Wollburg, Walton T. Roth, Ansgar Conrad, Anett Mueller, Sigrun Doberenz, Alicia E. Meuret, Sunyoung Kim

We assessed psychological and physiological data of three kinds of people during a voluntary hyperventilation (HV) test: those with panic disorder (PD), those without a PD but complaining of chronic mental tension, and psychiatrically healthy controls. Although chronic mental tension is a common complaint, it has the status of a symptom rather than a disorder in the DSM-IV.

Hyperventilation has a long history of being linked to anxiety, especially PD. Several studies have observed lower pCO2 levels in PD than in comparison groups (Wilhelm, Trabert, Roth, 2001; Wilhelm, Gerlach, Roth, 2001). Furthermore, PD patients have repeatedly been observed to recover more slowly from voluntary hyperventilation (e.g. Wilhelm, Trabert, Roth, 2001). Previous studies had PD patients hyperventilate to 20 mm Hg, while the current study tested whether panic patients would recover more slowly after hyperventilation to 25 mm Hg, and whether slower recovery was diagnostically specific to PD.

Participants were instructed to hyperventilate for three minutes at end-tidal pCO2 levels of 25 mm Hg, which were achieved with the help of experimenter feedback, and then to sit quietly and breathe normally for eight minutes. Continuous measures of EMG, heart rate, respiration rate, tidal volume, end-tidal pCO2 and skin conductance were recorded. After the procedure participants filled out a number of questionnaires, including the Anxiety Sensitivity Index (ASI) and the Tension and Relaxation Questionnaire (TRQ). Immediately before and after each task they rated their feelings and sensations.

Preliminary results for 12 PD patients, 16 tense patients, and 22 controls show a group effect for respiratory sinus arrhythmia (RSA, transfer function) and non-specific fluctuations (NSF) during the recovery period. RSA was lowest in the tense people. PD and tense people had higher NSF than controls. Groups did not differ in pCO2 level, respiratory rate, tidal volume, tidal volume instability, heart rate, skin conductance, the left and right gastrocnemius, left and right forearm, the upper trapezius, or the lateralis frontalis EMG. We found one time x group interaction in that groups recovered differently in activity of the left gastrocnemius muscle. The control group showed elevated muscle activity throughout the recovery period, but higher at the end, whereas the PD patients and tense people had more constant muscle activity over time. On psychological variables, PD patients and tense people rated themselves as more anxious and tense during HV than the controls. They also experienced more symptoms than the controls during HV, but not during recovery. Panic patients and tense people had the same total score on the ASI and rated themselves as equally tense on the TRQ.

Contrary to many previous findings, the three study groups reached the same recovery values in most variables. The PD group did not show the expected parasympathetic deactivation (RSA), but the tense group did. The PD and tense group were not different in number of NSFs. Panic patients and tense people only differed from the control group in the number of items endorsed during hyperventilation itself, not during recovery. During the laboratory assessment PD patients reported higher tension than controls and tense people, but both patient groups reported higher anxiety. On physiological variables, tense people differed only in two measures from the other groups. RSA was lower compared to panic patients at two minutes and compared to the control group at one minute. Panic patients and tense people had higher NSFs than controls and were equal to each other. On all other variables tense people reacted similarly to controls and PD patients on the VHT. We failed to replicate slower recovery from HV in panic patients.

We tentatively conclude that panic patients react to a milder hyperventilation of 25 mm Hg differently than they do to 20. We conclude that a pCO2 target level of 25 mm Hg is too high to bring out the group differences. We cannot definitely assert that the reaction of PD patients to voluntary hyperventilation is specific to this diagnosis.