Background Few studies have assessed the role of lung diffusing capacity (DLCO%) in bronchiectasis. We sought to examine the relationship between DLCO% and clinical and radiological variables in a well-defined population of bronchiectasis patients to determine its potential prognostic significance and compare with FEV₁%.

Methods Of 312 consecutive bronchiectasis patients attending our institution over a 3-year period, 204 patients were suitable for study inclusion. Exclusion criteria consisted of patients with cystic fibrosis or traction bronchiectasis, patients with missing data and patients with absent radiological evidence of bronchiectasis on independent expert thoracic radiologist review. Univariate analyses was performed using Pearson’s correlation. Backwards stepwise logistic regression analysis was subsequently performed to determine independent associations of DLCO% and FEV1%.

Results DLCO% strongly correlated with all measured lung function parameters FEV₁%, FVC%, FEF 25–75% and FEV₁/FVC ratio (all p < 0.001). Negative correlations were noted with age at diagnosis (p = 0.047), body mass index (p < 0.001) and number of comorbidities (p = 0.002). Significant symptom associations included SOB (p = 0.001) and fatigue (p < 0.001). Reduced DLCO% was associated with higher MRCD scores (p < 0.001), higher number of hospitalisations on follow up (p = 0.015), higher BSI scores (p < 0.001) and increased mortality (p = 0.028). No correlations were noted with gender, smoking history, aetiology or bacterial colonisation of any form. Reduced DLCO% was associated with increased number of lobes (p = 0.004), and higher total Reiff and modified Bhalla HRCT scores (p = 0.001 and p < 0.001 respectively). The modified Bhalla was excluded from stepwise regression as unavailable in most clinical settings. Backwards elimination showed DLCO% to be significant in predicting BSI (p < 0.001), number of lobes (0.017) and mortality (p = 0.028). Comparatively, backwards stepwise regression of FEV₁% showed significance in predicting BSI (p < 0.001), number of hospitalisations on follow up (p = 0.008), and mortality (p = 0.024). Separate regression models of DLCO% and FEV₁% using modified Bhalla components as cofactors showed DLCO% to be associated with disease extent (p = 0.002), bronchial wall thickness (p < 0.001), bronchial wall dilatation (p < 0.001) and reduced parenchymal attenuation (p = 0.015); FEV₁% was associated with bronchial wall dilatation only (p < 0.001).

Conclusion DLCO% predicts radiological disease, disease severity and mortality in bronchiectasis, independently of aetiology, and may identify patients with advanced disease who could benefit from intensive management.