OSA symptom subtypes and hypoxic burden independently predict distinct cardiovascular outcomes.

Study objectives: Studies on obstructive sleep apnoea (OSA) have identified clinically relevant symptom-based subtypes and novel OSA-specific nocturnal hypoxic measures. Both traits are individually associated with cardiovascular outcomes, but evidence about their independent or shared effects is unknown. This study investigated the simultaneous contributions of OSA symptom subtypes and hypoxic burden (HB) on incident cardiovascular outcomes.

Methods: Sleep Heart Health Study participants with high-quality oxygen saturation, apnoea-hypopnea index (AHI) and symptom data were included. Participants with OSA (AHI ≥5 events·h^-1) were grouped into symptom subtypes. HB was calculated from respiratory event-related hypoxia. Cox proportional hazards models assessed whether symptom subtypes and/or HB were independently associated with cardiovascular mortality and major adverse cardiovascular events (MACE).

Results: 4396 participants free of cardiovascular disease were analysed, with median follow-up >11 years. Higher HB was associated with worse cardiovascular mortality (HR (95% CI): 1.63 (1.13-2.35); p=0.009) independently of symptom subtypes. Compared to those without OSA, the excessively sleepy OSA subtype had higher risk of incident MACE (1.62 (1.23-2.15); p<0.001), independently of HB. Among participants with moderate-severe OSA (AHI ≥15 events·h^-1), excessively sleepy participants had higher risk of cardiovascular end-points compared to other subtypes, but HB was not associated with cardiovascular mortality or MACE risk.

Conclusion: OSA symptom subtypes and HB are independently associated with MACE and cardiovascular mortality, respectively. Thus, both are important for understanding OSA-related cardiovascular risk. Future studies using clinical samples including OSA therapy information that incorporate symptom subtypes and novel biomarkers, such as HB, could improve predictive models for cardiovascular disease risk.