Characterization of a 7% carbon dioxide (CO2) inhalation paradigm to evoke anxiety symptoms in healthy subjects.

The present study is aimed at characterizing the carbon dioxide (CO2) procedure in healthy subjects to achieve reliable provocation of anxiety symptoms. Thirty healthy subjects inhaled in single-blind both compressed air and 7% CO2 mixture. Panic Symptom List (PSLIII-R), Visual Analogue Scale-Anxiety (VAS-A), State Anxiety Inventory (STAI-Y/1), respiratory parameters and skin conductance were measured. 'Responders' were classified depending on PSLIII-R scores after CO2. Twelve out of the 21 'responders' performed a second test to assess test-retest repeatability. In 21 subjects Delta%VAS-A (45.4 +/- 32.1) and PSLIII-R (pre-test 2.3 +/-2.1, post-test 17.5 +/- 8.2) but not STAI-Y/1, significantly increased during CO2 inhalation. Respiratory Rate, Minute Volume, end-Tidal CO2 and skin conductance rose in 'responders'. Repeatability was studied with Bland-Altman plots, revealing mean difference between tests close to 0 for both Delta%VAS-A and PSLIII-R. Among physiologic parameters, end-Tidal CO2 and Respiratory Rate showed good repeatability, with a within-subject CV of 9.2% and 6%, respectively. The challenge produced measurable response in healthy subjects. Good test-retest repeatability was observed in 'responders'. These data indicate that the test can be suitable for testing putative anti-panic or anxiolytic drugs in clinical studies using a within subject, crossover design.