Citations

ImportancePersistent symptoms and disability following SARS-CoV-2 infection, known as post–COVID-19 condition or “long COVID,” are frequently reported and pose a substantial personal and societal burden.ObjectiveTo determine time to recovery following SARS-CoV-2 infection and identify factors associated with recovery by 90 days.Design, Setting, and ParticipantsFor this prospective cohort study, standardized ascertainment of SARS-CoV-2 infection was conducted starting in April 1, 2020, across 14 ongoing National Institutes of Health–funded cohorts that have enrolled and followed participants since 1971. This report includes data collected through February 28, 2023, on adults aged 18 years or older with self-reported SARS-CoV-2 infection.ExposurePreinfection health conditions and lifestyle factors assessed before and during the pandemic via prepandemic examinations and pandemic-era questionnaires.Main Outcomes and MeasuresProbability of nonrecovery by 90 days and restricted mean recovery times were estimated using Kaplan-Meier curves, and Cox proportional hazards regression was performed to assess multivariable-adjusted associations with recovery by 90 days.ResultsOf 4708 participants with self-reported SARS-CoV-2 infection (mean [SD] age, 61.3 [13.8] years; 2952 women [62.7%]), an estimated 22.5% (95% CI, 21.2%-23.7%) did not recover by 90 days post infection. Median (IQR) time to recovery was 20 (8-75) days. By 90 days post infection, there were significant differences in restricted mean recovery time according to sociodemographic, clinical, and lifestyle characteristics, particularly by acute infection severity (outpatient vs critical hospitalization, 32.9 days [95% CI, 31.9-33.9 days] vs 57.6 days [95% CI, 51.9-63.3 days]; log-rank P < .001). Recovery by 90 days post infection was associated with vaccination prior to infection (hazard ratio [HR], 1.30; 95% CI, 1.11-1.51) and infection during the sixth (Omicron variant) vs first wave (HR, 1.25; 95% CI, 1.06-1.49). These associations were mediated by reduced severity of acute infection (33.4% and 17.6%, respectively). Recovery was unfavorably associated with female sex (HR, 0.85; 95% CI, 0.79-0.92) and prepandemic clinical cardiovascular disease (HR, 0.84; 95% CI, 0.71-0.99). No significant multivariable-adjusted associations were observed for age, educational attainment, smoking history, obesity, diabetes, chronic kidney disease, asthma, chronic obstructive pulmonary disease, or elevated depressive symptoms. Results were similar for reinfections.Conclusions and RelevanceIn this cohort study, more than 1 in 5 adults did not recover within 3 months of SARS-CoV-2 infection. Recovery within 3 months was less likely in women and those with preexisting cardiovascular disease and more likely in those with COVID-19 vaccination or infection during the Omicron variant wave.

AbstractIntroductionPrevious bronchoalveolar lavage fluid (BALF) proteomic analysis has evaluated limited numbers of subjects for only a few proteins of interest, which may differ between asthma and normal controls. Our objective was to examine a more comprehensive inflammatory biomarker panel in quantitative proteomic analysis for a large asthma cohort to identify molecular phenotypes distinguishing severe from nonsevere asthma.MethodsBronchoalveolar lavage fluid from 48 severe and 77 nonsevere adult asthma subjects were assessed for 75 inflammatory proteins, normalized to BALF total protein concentration. Validation of BALF differences was sought through equivalent protein analysis of autologous sputum. Subjects' data, stratified by asthma severity, were analysed by standard statistical tests, principal component analysis and 5 machine learning algorithms.ResultsThe severe group had lower lung function and greater health care utilization. Significantly increased BALF proteins for severe asthma compared to nonsevere asthma were fibroblast growth factor 2 (FGF2), TGFα, IL1Ra, IL2, IL4, CCL8, CCL13 and CXCL7 and significantly decreased were platelet‐derived growth factor a‐a dimer (PDGFaa), vascular endothelial growth factor (VEGF), interleukin 5 (IL5), CCL17, CCL22, CXCL9 and CXCL10. Four protein differences were replicated in sputum. FGF2, PDGFaa and CXCL7 were independently identified by 5 machine learning algorithms as the most important variables for discriminating severe and nonsevere asthma. Increased and decreased proteins identified for the severe cluster showed significant protein–protein interactions for chemokine and cytokine signalling, growth factor activity, and eosinophil and neutrophil chemotaxis differing between subjects with severe and nonsevere asthma.ConclusionThese inflammatory protein results confirm altered airway remodelling and cytokine/chemokine activity recruiting leukocytes into the airways of severe compared to nonsevere asthma as important processes even in stable status.