Associations between obstructive sleep apnea and comorbidities in a large clinical biobank
Cade, B. E.; Li, L.; Duff, M.; Yang, C.; Hassan, S. M.; Yu, X.; Goodman, M. O.; Alex, R. M.; Azarbarzin, A.; Batool-Anwar, S.; Nezami, F. R.; Ramezani, A.; Sands, S. A.; Wang, H.; Kirchner, H. L.; Shah, N. A.; Epstein, L. J.; Pavlova, M. K.; Redline, S.
Show abstract
BackgroundObstructive sleep apnea (OSA) is associated with a wide range of comorbidities, but large-scale phenome-wide analyses in clinical biobanks remain under-reported. In this study, we identified common comorbidities enriched in patients with OSA, tested the temporality of these associations, and analyzed relevant associations with summary sleep recording data. Methods48,251 participants with OSA in the Mass General Brigham healthcare system were identified using a natural language processing phenotyping algorithm and/or evidence of an elevated apnea-hypopnea index (AHI). Controls were matched (2:1) on demographics, body mass index (BMI), and healthcare utilization. Associations with 358 incident and 563 cross-sectional diseases were tested using Modified Poisson regression, adjusting for covariates. Sensitivity analyses examined timing by binning data in years relative to the first OSA diagnosis. Selected laboratory results were obtained based on associated diseases. Associated diseases were tested with sleep recording statistics (n [≤]18,348). Findings179 incident and 421 cross-sectional diseases were associated with OSA at Bonferroni significance. 37 diseases had Bonferroni-significant sex interactions. Several associations were significant years before the first recorded OSA diagnosis. Four red blood cell laboratory measures were significant ten years prior to the first diagnosis. One incident and 47 cross-sectional diseases were associated with the AHI and/or chronic hypoxemia. InterpretationObstructive sleep apnea is associated with enrichment of hundreds of diseases, several of which are supported by orthogonal polysomnographic evidence. Leveraging early signs of OSA in clinical data may help to identify at-risk patients. FundingNational Institutes of Health and the American Academy of Sleep Medicine Foundation.
Matching journals
The top 8 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Methodological approach to sleep state misperception in insomnia disorder: comparison between multiple nights of actigraphy recordings and a single night of polysomnography recording 93%
- Effects of cognitive behavioral therapy for insomnia on subjective and objective measures of sleep and cognition 92%
- Impact of OSA Treatment Success on Changes in Hypertension and Obesity: A Retrospective Cohort Study 91%
Similar papers in this journal
- Development and Validation of a Metabolite Index for Obstructive Sleep Apnea across Race/Ethnicities 94%
- Novel Digital Markers of Sleep Dynamics: A Causal Inference Approach Revealing Age and Gender Phenotypes in Obstructive Sleep Apnea 92%
- Sleep Regularity Index as a Novel Indicator of Sleep Disturbance in Stroke Survivors: A Secondary Data Analysis 91%
Similar papers in this journal
- Obstructive sleep apnea is associated with specific gut microbiota species and functions in the population-based Swedish CardioPulmonary bioImage Study (SCAPIS) 92%
- Genetic associations and architecture of asthma-chronic obstructive pulmonary disease overlap 88%
- Rethinking Blood Eosinophils for Assessing ICS Response in COPD: A Post-Hoc Analysis from FLAME 87%
Similar papers in this journal
- Bidirectional relationship between sleep problems and long COVID: a longitudinal analysis of data from the COVIDENCE UK study 94%
- Clustering-based COPD Subtypes Have Distinct Longitudinal Outcomes and Multi-omics Biomarkers 88%
- Risk factors for asthma-related hospital and intensive care admissions in children, adolescents, and adults: a cohort study using primary and secondary care data 87%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.