Exercise ventilatory irregularity can be quantified by approximate entropy to detect breathing pattern disorder.

BACKGROUND: Breathing pattern disorder (BPD) is a prevalent cause of exertional dyspnea and yet there is currently no reliable objective measure for its diagnosis. We propose that statistical analysis of ventilatory irregularity, quantified by approximate entropy (ApEn), could be used to detect BPD when applied to cardiopulmonary exercise test (CPET) data. We hypothesised that ApEn of ventilatory variables (tidal volume (VT), breathing frequency (Bf), minute ventilation (VE)) would be greater, i.e. more irregular, in patients with BPD than healthy controls. METHODS: We evaluated ventilatory ApEn in 20 adults (14 female) with exertional dyspnoea, undergoing CPET and independently diagnosed with BPD by a specialist respiratory physiotherapist. Data were compared with 15 age- gender- and BMI-matched controls. ApEn for VT, Bf and VE were calculated for an incremental cycle exercise test. RESULTS: Patients with BPD more frequently rated breathlessness as the reason for exercise limitation and had a lower mean (SD) peak oxygen uptake compared with controls: 80 (18) vs. 124 (27) % predicted (P < 0.001). ApEn was significantly greater for VT (p = .006) and VE (p = .002) in BPD than controls. ApEn VE was inversely related (r2 = 0.24, p = .03) to peak oxygen uptake in BPD but not controls. ROC analysis revealed that ApEn VE > 0.88, conferred a sensitivity and specificity of 70% and 87% respectively, for detection of BPD. CONCLUSIONS: Non-linear statistical interrogation of CPET-acquired ventilatory data has utility in the detection of BPD. A simple calculation of approximate entropy of ventilation, during an incremental cardiopulmonary exercise test, provides a quantitative method to detect BPD.