Medically unexplained dyspnea: partly moderated by dysfunctional (thoracic dominant) breathing pattern

Dysfunctional breathing (DB) may contribute to disproportionate dyspnea and other medically unexplained symptoms. The extent of dysfunctional breathing is often evaluated using the Nijmegen Questionnaire (NQ) or by the presence of abnormal breathing patterns. The NQ was originally devised to evaluate one form of dysfunctional breathing - hyperventilation syndrome. However, the symptoms identified by the NQ are not primarily due to hypocapnia and may be due to other causes including breathing pattern dysfunction.

The relationships between breathing pattern abnormalities and the various categories of NQ symptoms including respiratory or dyspnea symptoms have not been investigated. This study investigates these relationships.

62 patients with medically unexplained complaints, that seemed to be associated with tension and breathing dysfunction, were referred, or self-referred, for breathing and relaxation therapy. Dysfunctional breathing symptoms and breathing patterns were assessed at the beginning and end of treatments using the NQ for assessment of DB symptoms, and the Manual Assessment of Respiratory Motion (MARM) to quantify the extent of thoracic dominant breathing. Subscales for the NQ were created in 4 categories, tension, central neurovascular, peripheral neurovascular and dyspnea. Relationships between the NQ (sum scores and subscales) and the MARM were explored.

Mean NQ scores were elevated and mean MARM values for thoracic breathing were also elevated. There was a small correlation pre-treatment between MARM and NQ (r=0.26, p<0.05), but classification of subjects as normal/abnormal on both measurements agreed in 74% (p < 0.001) of patients. From the sub scores of NQ only the respiratory or 'dyspnea' items correlated with the MARM values. Dyspnea was only elevated for subjects with abnormal MARM. After treatment, both MARM and NQ returned to normal values (p< 0.0001). Changes in NQ were largest for subjects with abnormal MARM pre-treatment. There was a large interaction between the change in the NQ sub score dyspnea and initial MARM values. (p<0.001).