Does obstructive sleep apnea increase the risk of post-acute sequelae of SARS-CoV-2?

In a recent article published in the journal Sleep, researchers generated a harmonized analysis investigating the impact of preexisting obstructive sleep apnea (OSA) as a risk factor for post-acute sequelae of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) [PASC] in children and adults. They used electronic health record (EHR) data from three research networks within the REsearching COVID to Enhance Recovery (RECOVER) initiative funded by the National Institutes of Health (NIH).

Study: Risk of post-acute sequelae of SARS-CoV-2 infection associated with pre-coronavirus disease obstructive sleep apnea diagnoses: an electronic health record-based analysis from the RECOVER initiative. Image Credit: p.ill.i / Shutterstock

Background

Previous studies have identified a positive relationship between OSA and acute coronavirus disease 2019 (COVID-19) outcomes. OSA, characterized by repeated obstruction of airways during sleep, is highly prevalent in the United States (US), affecting nearly 20% of adults; thus, this warrants further research as a potential risk factor for PASC. Per recent estimates, PASC affects 7% to 54% of COVID-19 patients even after complete recovery.

PASC risk varies by gender, age, and specific preexisting health conditions like hypertension and diabetes, raising the possibility that OSA could be a risk factor for PASC. Nevertheless, studies have barely investigated and elucidated the impact of preexisting conditions like OSA on the risk of developing PASC. In addition, studies should look beyond acute outcomes among COVID-19 survivors having preexisting comorbid OSA.

About the study

The present study is the first real-world data analysis conducted between March 1, 2018, and March 1, 2020, across multiple data sources, using different Long COVID definitions and employing varying approaches to identify COVID-19 patients at a higher risk of developing PASC due to preexisting OSA. Specifically, the researchers considered evidence of OSA diagnosed within two years before the study duration.

The National COVID Cohort Collaborative (N3C) and the Patient-Centered Clinical Research Network (PCORnet) encompassed adult populations aged ≥18 years; the latter's PEDSnet arm also covered the pediatric population. N3C analyzed data of >15 million patients from 77 sites, while PCORnet drew analyses from 11M patients from 19 sites. Likewise, PEDSnet selected 8.5M patients from a network of eight pediatric health systems.

A clinical science core (CSC) at New York University Langone Health coordinated all three RECOVER research networks, albeit each network created distinct diagnosis-based computable phenotype (CP) definitions to find probable PASC patients on its own. Notably, algorithms designed to identify patients with an OSA diagnosis from EHRs have excellent validity. Also, N3C and PCORnet limited their analyses to adults aged ≥21 years, while PEDSnet to children below 21 years as it facilitated delineating results for adults and children.

The team trained all CPs on patients who visited a Long COVID clinic and definitions rooted in rules involving labs, clinical diagnoses, and medications. The eligibility criteria for inclusion was that a patient showed proof of a COVID-19 infection between March 1, 2020, and February 28, 2022, usually a positive SARS-CoV-2 reverse transcription-polymerase chain reaction (RT-PCR) or antigen test.

Further, the researchers used the International Statistical Classification of Diseases and Related Health Problems – Clinical Modification (ICD-10-CM) and ICD-9-CM diagnostic codes or the Systematized Nomenclature of Medicine – Clinical Terms (SNOMED-CT) codes to identify patients with preexisting OSA.

Lastly, the researchers used logistic regression models to estimate unadjusted and adjusted odds ratios (ORs) with 95% confidence intervals (CIs) for the association between a preexisting OSA diagnosis and the likelihood of developing PASC.

Study findings

Regardless of the approach used to identify probable PASC patients, the current analysis showed preexisting OSA increased risk of PASC-like conditions among adult patients. Even after adjusting for other comorbidities, this positive association remained significant though it attenuated slightly. Sensitivity analysis adjusting for preexisting hypertension and diabetes in place of comorbidity score did not alter the findings for adults.

Conversely, the apparent positive associations between preexisting OSA and probable PASC among children became insignificant after adjusting for comorbidities. Additionally, sensitivity analyses adjusting for asthma,  hypertension, and diabetes for children altered the observed associations and fetched different effect estimates.

Obesity was most prevalent in PCORnet and accounted for some visible differences in OSA-related PASC outcomes across all three networks. In part, obesity and similar comorbidities confounded and diminished the strength of association with PASC.

Conclusions

PASC is not a cohesive condition and has no accepted case definition yet, and its patients have highly heterogeneous symptoms. So, the study authors used a range of PASC definitions to overcome these challenges while examining associations between OSA and PASC risk among adults and children. Finally, they found positive associations between OSA and PASC risk consistently, irrespective of the data source, approach, and PASC definition applied.

However, the authors did not explore symptoms most prevalent in OSA patients at high risk of probable PASC. Thus, future research should examine the association of OSA and other preexisting conditions with specific PASC variations and the trajectory of impacted COVID-19 patients.

Written by

Neha Mathur

Neha is a digital marketing professional based in Gurugram, India. She has a Master’s degree from the University of Rajasthan with a specialization in Biotechnology in 2008. She has experience in pre-clinical research as part of her research project in The Department of Toxicology at the prestigious Central Drug Research Institute (CDRI), Lucknow, India. She also holds a certification in C++ programming.