Sleep Apnoea Variability in Pacemaker Patients: A Women-Predominant Phenotype

Single-night polysomnography assumes stable sleep apnoea (SA) severity; night-to-night variability may identify clinically relevant phenotypes, particularly in women. Sex/gender-specific longitudinal phenotypes were characterised using device-detected respiratory disturbance index (RDI) and an instability metric independent of mean severity.

Pacemaker patients in the prospective ACaSA study with at least 180 valid nightly RDI measurements after a 30-day post-implant period were analysed. Severity was summarised by mean RDI. Variability was summarised by an Instability Index (residual from the log-log regression of RDI standard deviation on mean RDI). Phenotypes were defined by mean RDI (< 20 vs. ≥ 20 events/h) and instability (above vs. below median).

About 37,260 nights from 207 patients (median age 76 years, IQR [69-81]) were analysed. Four phenotypes were identified: Mild-Stable (n = 68), Mild-Unstable (n = 67), Severe-Stable (n = 35), and Severe-Unstable (n = 37). Women comprised 37.7% of the total cohort and they were disproportionately overrepresented in the Mild-Unstable versus in the Mild-Stable phenotype (61.2% vs. 33.8%; p = 0.002). Among patients with mean RDI < 20 (n = 135), female sex/gender (adjusted OR 3.38, 95% CI 1.60-7.40; p = 0.002) and obesity (BMI ≥ 30 kg/m²; adjusted OR 3.80, 95% CI 1.31-12.53; p = 0.019) were independent predictors of the Mild-Unstable phenotype. Mild-Unstable had more severe-range nights (RDI ≥ 20/h) than Mild-Stable (7.8% vs. 2.2%; p < 0.001). Severe-Unstable recorded RDI < 20/h on 33.9% of nights versus 13.9% in Severe-Stable (p < 0.001).

Night-to-night variability identifies longitudinal SA phenotypes not captured by mean RDI. A female- and obesity-predominant Mild-Unstable phenotype shows intermittent severe-range nights despite low average burden, while instability in severe SA increases the likelihood of underestimating severity when classification is based on a single-night test.