- Background: Up to 40% of older surgical patients develop postoperative delirium (POD) and/or postoperative cognitive dysfunction (POCD)–objective cognitive deficits occurring 1-12 months after surgery. Although POCD and POD are both associated with decreased quality of life, increased mortality, and long-term cognitive decline, there are few interventions to prevent them. One potentially modifiable POD and POCD risk factor is poor sleep quality, due to obstructive sleep apnea (OSA), a frequently undiagnosed disorder characterized by repetitive breathing interruptions and hypoxia during sleep. Although OSA is associated with cognitive impairment and earlier dementia onset, and conflicting studies suggest a possible role in POD, it is unknown whether untreated OSA is associated with POCD. In this study, older non-cardiac surgery patients underwent home sleep apnea testing to diagnose preoperative OSA, in addition to pre- and postoperative cognitive testing and delirium assessment to determine the relationship of OSA with POD and POCD.
Methods: The SANDMAN (Sleep Apnea, Neuroinflammation, and Cognitive Dysfunction Manifesting after Non-cardiac surgery) study is an IRB-approved sub-study of the NIH-funded INTUIT study. In INTUIT, patients age >60 undergoing non-cardiac surgery were assessed for delirium twice daily postoperatively with 3D-CAM, and for POCD with a cognitive testing battery performed before and 6 weeks after surgery. SANDMAN patients additionally underwent preoperative home sleep apnea testing (HSAT) to measure the apnea-hypopnea index (AHI), an index of OSA severity. Patients also reported their subjective sleepiness level preoperatively using the Epworth Sleepiness Scale (ESS). Multivariable linear regression models compared ESS, OSA, delirium, and POCD severity, while controlling for age, sex, and baseline cognition.
Results: 145 INTUIT patients were screened for SANDMAN inclusion, and 101 eligible patients underwent preoperative HSAT. Of the 96 subjects with valid HSAT data, 74 had complete POCD/delirium cognitive data, including 28 with mild OSA, 7 with moderate OSA, and 8 with severe OSA; 9 patients had postoperative delirium. There was no significant relationship between OSA diagnosis category with POCD severity at 6 weeks (p=0.893) and no significant relationship between AHI and POCD severity in our multivariate model adjusting age, sex, and baseline cognition (effect estimate -0.01, p=0.288). There was no significant relationship between OSA severity (AHI) and delirium incidence (OR 0.99; 95% CI 0.93-1.04, p=0.97) or severity (OR 0.98, 95% CI 0.95-1.01, p=0.32). ESS was not associated with delirium incidence, but moderate-severe sleepiness (ESS≥13) was associated with increased delirium severity (OR 3.98; 95% CI 1.10-14.37, p=0.035).
Conclusion: These data suggest that preoperative OSA status is not associated with POCD or increased delirium incidence or severity; however, increased preoperative subjective sleepiness is associated with postoperative delirium severity. This suggests that a contributor to sleepiness outside of OSA, such as sleep deficits, could play a role in delirium.