Abstract Rationale The frequency of acute exacerbations in childhood interstitial lung disease (chILD) varied among children. This study aimed to evaluate the relationship between pulmonary function parameters and annual exacerbation rates in chILD. Methods The longitudinal study was conducted at a tertiary healthcare facility in India from 2023 to 2025, enrolling patients 18 years diagnosed with chILD. Clinico-demographic characteristics and severity of illness (score 1-5) were recorded at recruitment, and participants were prospectively followed for 12 months. The baseline study assessments in the participants, which included pulse oximetry (SpO2), spirometry (FEV₁ and FVC), and diffusing capacity of lungs for carbon monoxide (DLco; steady-state method), were repeated quarterly and during each exacerbation. Participants were grouped according to the total number of exacerbations recorded during the follow-up. Baseline differences between the groups were analyzed by using the Kruskal-Wallis test with post-hoc Holm-Sidak adjustment. After assessing the adequacy of the fit for count data by the chi-squared goodness-of-fit test and dispersion analysis, mixed-effects Poisson regression was applied to repeated measurements of SpO2, FVC, FEV₁, and DLco to determine their association with annual exacerbation rates. Results Out of the 105 enrolled patients, 93 (52 boys) completed follow-up. Median (IQR) age was 9.75 (5.5-12.7) years, and median duration of illness was 55.2 (23.0-79.4) months. The severity of illness score of ≥ 3 was noted in 45.2% of participants at baseline. A total of 117 exacerbations occurred, with a median (range) annual rate of 1 (0-6). The proportions of participants in the groups with annual exacerbation rates of 0, 1, and ≥2 were 36.6%, 29.0%, and 34.4%, respectively. Across the three groups, baseline median (IQR) SpO2 (%) was 98 (97-99), 97 (95-99), and 97 (93.5-98) (p = 0.06); FVC (Z-score) was -1.31 (-2.51, -0.80), -3.98 (-4.88, -2.13), and -4.07 (-5.65, -3.32) (p 0.001); FEV₁ (Z-score) was -1.61 (-2.54, -0.87), -4.17 (-4.80, -2.71), and -4.73 (-5.33, -3.40) (p 0.001); and DLco (% predicted) was 83 (74-112), 76 (48-96), and 58.5 (47.5-69.5) (p = 0.001). Multivariable analysis adjusted for baseline characteristics revealed reduction in FEV₁ during follow-up was strongly associated with higher exacerbation rates (IRR 0.59, 95% CI 0.43-0.81; p = 0.001) within the participants. The margins plots demonstrated an inverse relationship between FEV₁ (Z-score) and predicted exacerbation rates, as shown in Figure A. Conclusions In patients with chILD, declining FEV₁ over time independently predicted higher annual exacerbation rates, highlighting its potential as a marker for exacerbation risk stratification. This abstract is funded by: None
Nandy et al. (Fri,) studied this question.