Non-zero risk of adverse psychological reactions induced by experimental dyspnoea studies in healthy volunteers: possible ethical implications.

Background: Experimentally inducing dyspnoea in healthy volunteers is a powerful research tool. Even though this approach is not considered to carry particular risks, dyspnoea occurs in a general context of stress. We hypothesised that experimental dyspnoea could be traumatic in some individuals, with immediate or delayed adverse psychological reactions.

Methods: 40 participants (median (interquartile range) age 38.7 (29.7-49.1) years; 20 women) were exposed, in random order, to inspiratory threshold loading (ITL; "excessive breathing effort"; unpleasantness rating 6 (4-7) on a 0-10 scale) and carbon dioxide stimulation with restricted ventilatory response (CO₂-rv; "air hunger"; unpleasantness rating 7 (7-8)). Participants were evaluated immediately after induced dyspnoea, at 7 days and after a further 3 months.

Results: Peritraumatic Distress Inventory (PDI) scores were higher immediately after CO₂-rv than after ITL (3 (1-6.5) versus 2 (0-3); p<0.0001). After 7 days, six participants (15%) had an Impact of Event Scale-Revised (IES-R) score ≥12 (high risk of post-traumatic stress disorder (PTSD)). They had significantly higher baseline anxiety trait indicators than the participants with an IES-R score <12 (baseline Anxiety Sensitivity Index 27.5 (14.75-61.25) versus 10 (3-19); p=0.0087). At 3 months, no participant had a PTSD Checklist for DSM-5 (PCL-5) score suggestive of PTSD (1 (0-4), range 0-23, threshold 33) but the six participants with IES-R ≥12 on day 7 had significantly higher PCL-5 scores (13.5 (1-21.5) versus 0 (0-4); p=0.0014).

Conclusions: Although no severe situations were encountered, this study highlights the need for caution when designing experimental dyspnoea studies, particularly when involving individuals with high trait anxiety.