Sleep architecture in patients with interstitial lung disease with and without pulmonary hypertension

Abstract

Sleep disturbance is common in patients with advanced interstitial lung disease (ILD) often complicated by pulmonary hypertension (PH) and may contribute to poor quality of life. The etiology of sleep disturbance and the relationship between PH and sleep architecture in patients with ILD remains unknown. We performed a retrospective cohort study comparing sleep architecture on polysomnography in patients with ILD with and without PH, defined as mean pulmonary artery pressure on right heart catheterization ≥ 20mmHg. We tested the hypothesis that patients with ILD and PH would have increased wake time after sleep onset (WASO) compared to patients with ILD without PH using univariate analysis and multivariable linear regression. In our cohort of patients with ILD who underwent polysomnography (N = 49), patients with PH had lower total diffusion capacity for carbon monoxide (DLCO) (9.28 vs. 12.87ml/min/mmHg, P = 0.04) and percent DLCO (39% vs. 53%, P = 0.03). On polysomnography, patients with PH had increased percentage of total sleep time with saturation < 90% (T90) (17% vs. 6%, P = 0.03), decreased N2 sleep (181 vs. 233min, P = 0.03), decreased %N2 sleep (59% vs. 66%, P = 0.04), increased %N1 sleep (22% vs. 14%, P = 0.02), decreased sleep efficiency (62% vs. 72%, P = 0.03), and increased WASO (133 vs. 84min, P = 0.01). In multivariable analysis, PH was associated with a 43-min increase in WASO (95% CI 6.2-80.2, P = 0.02). Patients with ILD and PH have decreased total and %N2 sleep, increased %N1 sleep, decreased sleep efficiency, and increased WASO, likely indicating increased sleep fragmentation.