Impulse oscillometry-derived respiratory mechanical abnormalities and fibrotic disease progression in interstitial lung disease: a single-center cohort study

Objective To examine whether continuous impulse oscillometry (IOS)-derived respiratory mechanical measures are associated with subsequent fibrotic disease progression in interstitial lung disease (ILD). Methods In this single-center ambispective cohort study, 255 adults with fibrotic ILD underwent baseline spirometry, diffusing capacity testing and IOS. Fibrotic disease progression was defined using a PPF-like composite requiring at least two of three domains (worsening respiratory symptoms, physiological progression and radiological progression) within a 9–15-month assessment window. All IOS parameters were analyzed continuously using Firth penalized logistic regression, adjusted for age, sex, height and baseline FVC % predicted. X5 was additionally assessed using continuous scaling and restricted cubic splines. ROC analyses, non-IPF sensitivity analyses, and bootstrap validation were exploratory. Results Fibrotic disease progression occurred in 41 of 255 patients (16.1%). A 1-SD worsening in X5 was associated with higher odds of progression (OR 8.89, 95% CI 4.66–19.47; P<0.001), as were AX (OR 1.67, 95% CI 1.22–2.38; P = 0.001) and Fres (OR 1.61, 95% CI 1.11–2.36; P = 0.012). Each 0.01 kPa·s/L increase in X5 was associated with lower odds of progression (adjusted OR 0.67, 95% CI 0.59–0.76; P<0.001), and the X5–progression association was nonlinear (P for nonlinearity<0.001). After exclusion of IPF, X5 and AX remained associated with progression. Bootstrap validation showed improved discrimination after adding AX or X5, but did not constitute external validation. Conclusions Continuous IOS measures, particularly X5 and AX, were associated with subsequent fibrotic disease progression in this single-center ILD cohort, with a weaker signal for Fres. The X5 association was nonlinear and remained evident after exclusion of IPF. ROC-derived thresholds were unstable and should be considered exploratory. These findings support further evaluation of oscillometric respiratory-mechanical abnormalities in larger externally validated cohorts but do not establish clinically applicable predictive thresholds.