Secondary Logo

Journal Logo

Institutional members access full text with Ovid®

Diagnosed Sleep Apnea and Cardiovascular Disease in Atrial Fibrillation Patients

The Role of Measurement Error from Administrative Data

Ogilvie, Rachel P.a,b; MacLehose, Richard F.b; Alonso, Alvaroc; Norby, Faye L.b; Lakshminarayan, Kamakshib; Iber, Conradd; Chen, Lin Y.e; Lutsey, Pamela L.b

doi: 10.1097/EDE.0000000000001049

Background: Atrial fibrillation and obstructive sleep apnea are common conditions, but little is known about obstructive sleep apnea and cardiovascular risk among atrial fibrillation patients.

Methods: Using the Truven Health MarketScan databases, we constructed a prospective cohort of atrial fibrillation patients from 2007 to 2014. Atrial fibrillation, obstructive sleep apnea, stroke, myocardial infarction, and confounders were defined using the International Classification of Disease-9-CM codes. We matched individuals with an obstructive sleep apnea diagnosis with up to five individuals without a diagnosis by age, sex, and enrollment date. Cox proportional hazards models adjusted for confounders and high-dimensional propensity scores. We included migraines as a control outcome. Bias analysis used published sensitivities and specificities to generate rate ratios adjusted for obstructive sleep apnea misclassification.

Results: We matched 56,969 individuals with an obstructive sleep apnea diagnosis to 323,246 without. During a mean follow-up of 16 months, 3234 incident strokes and 4639 incident myocardial infarctions occurred. After adjustment, obstructive sleep apnea diagnosis was strongly associated with reduced risk of incident stroke (hazard ratio = 0.48, 95% confidence interval = 0.43, 0.53) and myocardial infarction (0.40, [0.37, 0.44]) and a smaller reduced risk of migraines (0.82, [0.68, 0.99]). Bias analysis produced wide-ranging or inestimable rate ratios adjusted for misclassification of obstructive sleep apnea.

Conclusions: Obstructive sleep apnea diagnosis in atrial fibrillation patients was strongly associated with reduced risk of incident cardiovascular disease. We discuss misclassification, selection bias, and residual confounding as potential explanations.

From the aDepartment of Psychiatry, University of Pittsburgh, Pittsburgh, PA

bDivision of Epidemiology and Community Health, School of Public Health, University of Minnesota, Minneapolis, MN

cDepartment of Epidemiology, Rollins School of Public Health, Emory University, Atlanta, GA

dDepartment of Medicine, University of Minnesota Medical School, Minneapolis, MN

eCardiac Arrhythmia Center, Cardiovascular Division, Department of Medicine, University of Minnesota Medical School, Minneapolis, MN.

Submitted July 16, 2018; accepted May 24, 2019.

The research reported in this publication was supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health under Award Number R01HL122200 and the American Heart Association award 16EIA26410001 (A.A.). R.P.O. was also supported by provided by the NIH grants T32HL007779 and T32HL082610.

The authors report no conflicts of interest.

Supplemental digital content is available through direct URL citations in the HTML and PDF versions of this article (

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or the American Heart Association.

Due to copyright, the data for this article cannot be shared; however, it is commercially available. SAS Macros for the high dimensional propensity scores are available at Bias analysis spreadsheets are available at

Correspondence: Rachel P. Ogilvie, 3609 Forbes Avenue, Second Floor, University of Pittsburgh, Pittsburgh, PA 15213. E-mail:

Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.