Predictive value of lung function results for sleep-disordered breathing in children with sickle cell disease: A multicenter cross-sectional study Free

Background: Lung function abnormalities and sleep-disordered breathing are more prevalent in children with sickle cell disease (SCD) than in the pediatric general population. Sleep-disordered breathing (SDB) has been associated with adverse health outcomes. Spirometry is the most common lung function test performed and is increasingly used in children with SCD, especially those with respiratory symptoms. Previous studies have detailed abnormal lung function tests in this population; however, they have not investigated the association between these abnormalities and SDB. In this multicenter study, we aimed to investigate whether the spirometry values could predict abnormalities detected in studies in children with SCD.

Methods: This multicenter, cross-sectional study included children (0-18 years) with SCD who underwent both polysomnography (PSG) and spirometry at baseline status between 2012 and 2022 at 3 pediatric pulmonary centers in the southeastern United States (University of Alabama at Birmingham, University of Florida, and Duke University Hospital). Demographic and clinical data were collected, including age, sex, SCD genotype, hemoglobin levels, use of hydroxyurea, and asthma diagnosis. PSG metrics, such as obstructive apnea-hypopnea index (oAHI), oxygen saturation, carbon dioxide retention, periodic limb movements, and arousal index, were analyzed alongside spirometry parameters, including forced vital capacity (FVC), forced expiratory volume in one second (FEV1), FEV1/FVC ratio, and forced expiratory flow (FEF 25–75%). Descriptive analyses were used to describe the study cohort. T-tests, Wilcoxon rank-sum, chi-square, Fisher's exact tests, as appropriate, were used to compare spirometry predictors between groups. Linear regression models were performed to assess associations between spirometry and PSG results. For the dichotomous outcome, abnormal PSG, logistic regression models were used to calculate crude and adjusted odds ratios, accounting for demographic and clinical covariates. Receiver operating characteristic curves were generated to evaluate the predictive value of spirometry on sleep study outcomes. All tests were two-sided, and p-values ≤0.05 were considered statistically significant. All analyses were conducted using GraphPad Prism and SAS version 9.4 (SAS Institute, Cary, NC).

Results: A total of 109 children with SCD from the 3 centers were included in the study. The average age at the time of the sleep study was 10.7 years, and spirometry was typically performed approximately 15 months later, at an average age of 11.9 years. Most participants were male (52%) and had the severe SCD genotype (88%). Hydroxyurea was used by 74% of the cohort, and 39% of the participants had a diagnosis of asthma. Pulmonary function abnormalities were common: over 20% had abnormal FVC, more than 25% had abnormal FEV1, and FEF 25–75%, while only 12% had an abnormal FEV1/FVC ratio. SDB was also prevalent, with over 50% showing elevated oAHI, and significant proportions had abnormal oxygen saturation metrics and arousal indices. The study found a strong association between spirometry results and markers of nocturnal hypoxemia, particularly mean oxygen saturation (SpO2) and the percentage of total sleep time with oxygen saturation below 90% (TST90). FEV1/FVC is also associated with indicators of obstructive sleep apnea and sleep fragmentation. Univariate regression analysis revealed that age and hemoglobin levels significantly influenced spirometry outcomes, especially FVC and FEV1. Multiple linear regression models adjusting for these covariates confirmed that FEV1 remained a significant predictor of nocturnal hypoxemia across all models. FVC was consistently associated with mean SpO2 but only predicted TST90 when adjusted for age alone, not hemoglobin or both factors. Logistic regression analysis further demonstrated the predictive value of spirometry for sleep-related hypoxemia. For every 5% increase in predicted FVC, the odds of abnormal mean SpO2 and TST90 decreased significantly (ORs of 0.83 and 0.82, respectively). Similarly, a 5% increase in predicted FEV1 was associated with reduced odds of abnormal mean SpO2 and TST90 (ORs of 0.83 and 0.77, respectively).

Conclusions: These findings suggest that spirometry, particularly FVC and FEV1, may serve as useful indicators for identifying children with SCD at risk for SDB, highlighting the importance of integrated pulmonary and sleep assessments in this population.