Associations of Breathing Pattern Disorder and Nijmegen Score With Clinical Outcomes in Difficult-to-Treat Asthma

Abstract

BackgroundBreathing pattern disorder (BPD) reflects altered biomechanical patterns of breathing that drive breathing difficulty and commonly accompanies difficult-to-treat asthma. Diagnosis of BPD has no gold standard, but Nijmegen Questionnaire (NQ) >23 is commonly used. ObjectivesWe sought to advance clinical characterisation of BPD and better understand clinical utility of NQ in difficult asthma, in patients from the Wessex AsThma CoHort of difficult asthma (WATCH) study. MethodsAssociation between demographic and clinical factors in difficult asthma and BPD, ascertained by clinical diagnosis (yes/no, n=476), by NQ scores (≤23: normal (no suggestion of BPD) and >23: abnormal (suggested BPD), n=372, as well as the continuous raw NQ scores) were assessed in univariate models to identify significant risk factors associated with the three BPD outcomes. For the clinician-diagnosed and NQ-based BPD, associations of continuous factors were assessed using independent samples t-test or Mann-Whitney U test as appropriate for the data distribution or by Spearman correlation test. Dichotomous associations were evaluated using chi-squared tests. Multivariable logistic (dichotomous outcomes) and linear regression models (continuous outcomes) were developed to identify predictive factors associated with clinician-diagnosed and NQ-based BPD, dichotomous and continuous. Patients with data on NQ scores were grouped into NQ quartiles (low, moderate, high, and very high). The patterns of association of the quartiles with four health-related questionnaire outcomes were assessed using linear regression analyses. ResultsMultivariable regression identified that clinically diagnosed BPD was associated with female sex (OR 1.85; 95% CI 1.07, 3.20), comorbidities (rhinitis (OR 2.46; 95% CI 1.45, 4.17), GORD (OR 2.77; 95% CI 1.58, 4.84), ILO (OR 4.37; 95% CI 2.01, 9.50) and any psychological co-morbidity (OR 1.86; 95% CI 1.13, 3.07)) and healthcare usage (exacerbations (OR 1.07; 95% CI 1.003, 1.14) and previous ICU admissions (OR 2.03; 95% CI 1.18, 3.47)). Abnormal NQ-based BPD diagnosis was associated with history of eczema (OR 1.83; 95% CI 1.07, 3.14), GORD (OR 1.94; 95% CI 1.15, 3.27) or any psychological comorbidity (OR 4.29; 95% CI 2.64, 6.95) at multivariable regression. Differences between clinical and NQ-based BPD traits were also found with 42% discordance in BPD-state between these definitions. Multivariable linear regression analysis with NQ as a continuous outcome showed positive association with worse asthma outcomes (admission to ICU, p=0.037), different phenotypic traits (female sex p=0.001, ever smoker, p=0.025)) and greater multimorbidity (GORD, p=0.002, sleep apnoea, p=0.040, any psychological comorbidity, p<0.0001). ConclusionBPD is associated with worse health outcomes and negative health impacts in difficult asthma within a multimorbidity disease model. It therefore merits better recognition and prompt treatment. Clinical diagnosis and NQ offer different perspectives on BPD, so this goal may be best addressed by considering clinical features alongside magnitude of NQ.