Background
The aim of the study was to investigate whether findings in cardiorespiratory polygraphy had an association with stroke mortality or ischemic event recurrence after ischemic stroke.

Methods
We prospectively studied 204 ischemic stroke patients who underwent cardiorespiratory polygraphy within the first 48 h after the symptom onset. We followed all these patients for a median of 6.2 years. We evaluated mortality, time of survival, causes of death and new ischemic events.

Results
Of 204 ischemic stroke patients, 43 died and 48 had a new ischemic event during the follow-up. The lowest arterial oxyhemoglobin saturation (min SaO₂) (P = 0.007) was lower, the percentage of time spent below arterial oxyhemoglobin saturation less than 90% (T90) (P = 0.005) was higher, and central apnea index per hour (CAI/h) (P = 0.04) was higher among the deceased. Male gender, older age, diabetes mellitus, elevated modified Rankin scale (mRS) score, lower Glasgow Coma Scale (GCS) score and CAI/h independently predicted higher mortality. Peripheral arterial disease (PAD) and higher National Institutes of Health Stroke Scale (NIHSS) score were independent predictors for a recurrent ischemic event. Among those having respiratory event index (REI) at least 30, older age and lower GCS score independently predicted higher mortality. Only 21 stroke patients initiated continuous positive airway pressure (CPAP) treatment; of those, only one had a new ischemic event.

Conclusions
The non-survivors had more severe nocturnal hypoxemia and more central apneas than survivors. Among patients with REI at least 30/h, increased CAI predicted higher mortality, but not independently.

Clinical trial registration
URL:http://www.clinicaltrials.gov. Unique identifier: NCT01861275