788 Is Dysphagia a Marker for Obstructive Sleep Apnea (OSA) in Traumatic Brain Injury (TBI) Patients: A VA TBI Model System Study

Abstract

Listen

Translate article

Abstract Introduction Recent work has highlighted prevalent obstructive not central sleep apnea following traumatic brain injury (TBI). Treatment of comorbid OSA may facilitate neurologic recovery but widespread screening is limited. Mixed support exists for the presence of dysphagia as a biomarker of OSA in the general population and stroke patients. Dysphagia is common following TBI; however, no study has examined the relation between OSA and dysphagia in this cohort. Leveraging data from a recent six-center clinical trial of OSA and TBI during inpatient rehabilitation, this secondary analysis examined the association between OSA severity indices and proxy measures of dysphagia. Methods Level 1 polysomnography (PSG) was used to assess OSA (AHI ≥ 5 and ≥ 15) during inpatient rehabilitation for the overall sample (N=248; 203 male; 60.6% severe injury) evaluated at a median of 120.6 days post-TBI and subset ≤ 60 days post-injury. Dysphagia was approximated as the presence of a PEG tube and/or a modified texture diet (MTD) on the day of PSG. Chi square and Fisher’s Exact tests were utilized for group comparisons. Results As previously reported, OSA in this cohort was prevalent (68.2% (n=169) at AHI ≥ 5 and 33.5% (n=83) AHI ≥ 15) with predominantly obstructive events. 27.4% (n=68) met criteria for dysphagia combining proxy measures (34 peg; 49 MTD). No significant difference was found for presence of dysphagia across OSA severity cutoffs (AHI ≥ 5 & 15; p=0.1029 & 0.5959). When examining OSA across the individual proxy measures, persons without a peg tube were significantly more likely to have OSA at AHI ≥ 5 (62.5% vs 5.65%; p=0.0003) and AHI ≥ 15 (31.05% vs 2.42%; p=0.0353). When examining participants less than 60 days post-TBI, the group differences remained. Conclusion The incidence of dysphagia in TBI patients, as indexed by a modified diet or presence of a feeding tube, was not elevated in those with OSA. Sample bias (for undergoing Level 1 PSG and improvement facilitating inpatient rehabilitation admission) may have contributed to findings. Finally, future work with more sensitive indices of dysphagia is needed to accurately evaluate this association. Support (if any) PCORI (CER-1511–33005), NIIDLRR (90DPTB0004)