Characterizing seasonal trend in lung function in individuals with chronic lung disease may lead to timelier treatment of acute respiratory symptoms and more precise distinction between seasonal exposures and variability. Limited research has been conducted to assess localized seasonal fluctuation in lung function decline in individuals with cystic fibrosis (CF) in context with routinely collected demographic and clinical data. We conducted a longitudinal cohort study of 253 individuals aged 6-22 years with CF receiving care at a pediatric Midwestern US CF center with median (range) of follow-up time of 4.7 (0-9.95) years, implementing two distinct models to estimate seasonality effects. The outcome, lung function, was measured as percent-predicted of forced expiratory volume in 1 second (FEV1). Both models showed that older age, being male, using Medicaid insurance and having Pseudomonas aeruginosa infection corresponded to accelerated FEV1 decline. A sine wave model for seasonality had better fit to the data, compared to a linear model with categories for seasonality. Compared to international cohorts, seasonal fluctuations occurred earlier and with greater volatility, even after adjustment for ambient temperature. Average lung function peaked in February and dipped in August, and FEV1 fluctuation was 0.81 % predicted (95% CI: 0.52 to 1.1). Adjusting for temperature shifted the peak and dip to March and September, respectively, and decreased FEV1 fluctuation to 0.45 % predicted (95% CI: 0.08 to 0.82). Understanding localized seasonal variation and its impact on lung function may allow researchers to perform precision public health for lung diseases and disorders at the point-of-care level.
Keywords: lung disease; medical monitoring; nonlinear; seasonality; sine wave; temperature.