Objective: To prospectively examine the incidence and risk factors for sleep apnea in consecutive brain injury rehabilitation admissions.
Setting: Inpatient neurorehabilitation hospital.
Participants: Participants (n = 86) were consecutive neurorehabilitation admissions.
Design: Retrospective analysis of prospectively collected data.
Main measures: Polysomnography.
Results: Half (49%) of the sample was diagnosed with sleep apnea. For the full sample, univariate logistic regression revealed age (odds ratio: 1.08; 95% confidence interval: 1.04-1.11) and hypertension (odds ratio: 7.77; 95% confidence interval: 2.81-21.47) as significant predictors of sleep apnea diagnosis. Results of logistic regression conducted within the traumatic brain injury group revealed age (odds ratio: 1.07; 95% confidence interval: 1.02-1.13) as the only significant predictor of apnea diagnosis after adjustment for other variables. Hierarchical generalized linear regression models for the prediction of apnea severity (ie, apnea-hypopnea index found that Functional Independence Measure Cognition Score (P = .01) and age (P < .01) were significant predictors. Following adjustment for all other terms, only age (P < .01) remained significant.
Conclusion: Sleep apnea is prevalent in acute neurorehabilitation admissions and traditional risk profiles for sleep apnea may not effectively screen for the disorder. Given the progressive nature of obstructive sleep apnea and morbidity associated with even mild obstructive sleep apnea, early identification and intervention may address comorbidities influencing acute and long-term outcome.