Sleep oxygen desaturation predicts survival in idiopathic pulmonary fibrosis

Abstract

Background: Recent studies suggest poor sleep quality in patients with idiopathic pulmonary fibrosis (IPF). However, so far, the impact of IPF-related sleep breathing disorders (SBDs) on survival has not been extensively studied.

Methods: In a cohort of 31 (24 males) treatment-naïve, newly diagnosed consecutive IPF patients, we prospectively investigated the relationship of SBD parameters such as apnea-hypopnea index (AHI), maximal difference in oxygen saturation between wakefulness and sleep (maxdiff SpO2), and lowest sleep oxygen saturation (lowest SpO2) with clinical (survival, dyspnea, daytime sleepiness), pulmonary function, submaximal (6-min walk test [6MWT]) and maximal exercise variables (cardiopulmonary exercise test [CPET]), and right ventricular systolic pressure (RVSP).

Results: Sleep oxygen desaturation exceeded significantly that of maximal exercise (p < 0.001). Maxdiff SpO2 was inversely related to survival, DLCO%, and SpO2 after 6MWT, and directly with dyspnea, AHI, and RVSP. The lowest SpO2 was directly related to survival and to functional (TLC%, DLCO%) as well as submaximal and maximal exercise variables (6MWT distance, SpO2 after 6MWT, peak oxygen consumption/kg, SpO2 at peak exercise), while an inverse association with dyspnea score, AHI, and RVSP was observed.

Conclusions: Our findings provide evidence that intermittent sleep oxygen desaturation significantly exceeds that of maximal exercise and is associated with survival in IPF patients. Furthermore, they imply the existence of a link between lung damage and apnea events resulting to the induction and severity of intermittent sleep oxygen desaturation that aggravate pulmonary arterial hypertension and influence IPF survival.

Keywords: Idiopathic pulmonary fibrosis; exercise limitation and testing; sleep disordered breathing; sleep oxygen desaturation; survival.

Figure 1. Survival of 31 patients with usual interstitial pneumonia/idiopathic pulmonary fibrosis (UIP/IPF)

All patients were followed until death (uncensored n = 10) or until reporting of the study (censored n = 21). Shown are cumulative Kaplan-Meier survival plot, sample size, and survival (median survival = 525 days).

Figure 2. Maximal fall in saturation of oxygen during sleep

Kaplan-Meier survival curves for various values of maximal fall in saturation of oxygen (maxdiff SpO₂) during sleep, as predicted by the respective statistically significant Cox proportional hazards model for the entire study population (n = 31). Correlation is significant at p ≤ 0.05.

Figure 3. Lowest oxygen saturation during sleep

Kaplan-Meier survival curves for various values of lowest saturation during sleep (lowest SpO₂), as predicted by the respective statistically significant Cox proportional hazards model for the entire study population (n = 31). Correlation is significant at p ≤ 0.05.

Figure 4. Apnea-hypopnea index

Kaplan-Meier survival curves for various values of apnea-hypopnea index (AHI) as predicted by the respective statistically significant Cox proportional hazards model for IPF patients of the study population not receiving treatment with continuous positive airway pressure (CPAP) (n = 25). Correlation is significant at p ≤ 0.05.