Objective measures of sleepiness and cognitive function in different symptom subtypes of OSA

ABSTRACT Obstructive sleep apnea is an extremely common condition that contributes independently to multiple adverse outcomes, including cardiovascular disease, diabetes, and neurodegeneration. It is a major public health problem given its high prevalence. OSA is heterogeneous, with established differences in symptoms and outcomes in individuals with comparable disease severity based on traditional metrics. We were the first to show that there are different symptom subtypes of OSA using unsupervised clustering analysis, including groups defined by disturbed sleep with symptoms of insomnia, minimal symptoms, and marked excessive sleepiness. The existence of these subtypes has since been replicated around the world in multiple population-based and clinical cohorts. We have also shown that there are differences in outcomes between the these subtypes, particularly for cardiovascular events and mortality, with increased risk in the excessively sleepy subtype. However, there are still many unanswered questions. To begin to address these unanswered questions, we propose to use data from the STAGES study, which recruited patients undergoing sleep studies across six sleep centers in the United States and Canada. Participants had in-depth phenotyping that included: detailed questionnaires about their sleep and health, depression and anxiety scores, actigraphy to assess sleep duration over a 2-week period, psychomotor vigilance test to assess sleepiness objectively, and a comprehensive battery of tests to assess cognition. In this project, we will first examine if there are differences in objectively-measured sleepiness among the symptom subtypes, i.e., are individuals who complain of sleepiness also objectively sleepy? We will also assess what role chronic insufficient sleep (based on actigraphy) and poor sleep quality (based on a new measure of sleep depth—the Odds Ratio Product) play in determination of symptom subtypes. Next, we will expand knowledge regarding the clinical relevance of these subtypes by evaluating differences in neurocognitive outcomes. We hypothesize that the excessively sleepy group will have worse cognition than the other groups, after controlling for key covariates. Finally, we hypothesize that objective data on sleepiness, sleep duration, and sleep quality can be used to better understand additional heterogeneity among subjects with OSA. Specifically, we will perform unsupervised clustering based on both subjective symptoms and objective measures of sleepiness and sleep behavior (e.g., sleep duration and quality). We predict that we will identify new subgroups of subjects and will assess whether there are larger differences in neurocognitive outcomes among these subtypes when compared to subtypes defined only by subjective symptoms. The rich phenotype data in STAGES allow us to address all of these important questions. Overall, this project is aimed at further defining the heterogeneity of obstructive sleep apnea, an essential step in development of more personalized approaches to diagnosis and management. Future directions for this study will include follow-up analyses in other cohorts to validate the results and assess impact on other health outcomes.