Key Points
Question Among a cohort of US frontline and essential workers infected with the original strain or the Delta or Omicron variants of SARS-CoV-2, is there a difference in COVID-19 symptoms or viral RNA load among those receiving mRNA vaccines compared with being unvaccinated?
Findings In this prospective cohort study that included 1199 participants with SARS-CoV-2 infection, receipt of 2 or 3 mRNA vaccine doses before Delta infections and 3 mRNA vaccine doses before Omicron infections was significantly associated with milder COVID-19 (less frequently symptomatic, febrile, or medically attended or shorter duration of illness) compared with being unvaccinated. Receipt of 2 mRNA vaccine doses 14 to 149 days prior to either Delta or Omicron infection was significantly associated with lower viral RNA load.
Meaning Among a cohort of US frontline and essential workers, recent vaccination with 2 or 3 mRNA vaccine doses, compared with being unvaccinated, was associated with attenuated COVID-19 symptoms and lower viral RNA load for Delta and Omicron variants of SARS-CoV-2 in some comparisons.
Abstract
Importance Data on the epidemiology of mild to moderately severe COVID-19 are needed to inform public health guidance.
Objective To evaluate associations between 2 or 3 doses of mRNA COVID-19 vaccine and attenuation of symptoms and viral RNA load across SARS-CoV-2 viral lineages.
Design, Setting, and Participants A prospective cohort study of essential and frontline workers in Arizona, Florida, Minnesota, Oregon, Texas, and Utah with COVID-19 infection confirmed by reverse transcriptase–polymerase chain reaction testing and lineage classified by whole genome sequencing of specimens self-collected weekly and at COVID-19 illness symptom onset. This analysis was conducted among 1199 participants with SARS-CoV-2 from December 14, 2020, to April 19, 2022, with follow-up until May 9, 2022, reported.
Exposures SARS-CoV-2 lineage (origin strain, Delta variant, Omicron variant) and COVID-19 vaccination status.
Main Outcomes and Measures Clinical outcomes included presence of symptoms, specific symptoms (including fever or chills), illness duration, and medical care seeking. Virologic outcomes included viral load by quantitative reverse transcriptase–polymerase chain reaction testing along with viral viability.
Results Among 1199 participants with COVID-19 infection (714 [59.5%] women; median age, 41 years), 14.0% were infected with the origin strain, 24.0% with the Delta variant, and 62.0% with the Omicron variant. Participants vaccinated with the second vaccine dose 14 to 149 days before Delta infection were significantly less likely to be symptomatic compared with unvaccinated participants (21/27 [77.8%] vs 74/77 [96.1%]; OR, 0.13 [95% CI, 0-0.6]) and, when symptomatic, those vaccinated with the third dose 7 to 149 days before infection were significantly less likely to report fever or chills (5/13 [38.5%] vs 62/73 [84.9%]; OR, 0.07 [95% CI, 0.0-0.3]) and reported significantly fewer days of symptoms (10.2 vs 16.4; difference, −6.1 [95% CI, −11.8 to −0.4] days). Among those with Omicron infection, the risk of symptomatic infection did not differ significantly for the 2-dose vaccination status vs unvaccinated status and was significantly higher for the 3-dose recipients vs those who were unvaccinated (327/370 [88.4%] vs 85/107 [79.4%]; OR, 2.0 [95% CI, 1.1-3.5]). Among symptomatic Omicron infections, those vaccinated with the third dose 7 to 149 days before infection compared with those who were unvaccinated were significantly less likely to report fever or chills (160/311 [51.5%] vs 64/81 [79.0%]; OR, 0.25 [95% CI, 0.1-0.5]) or seek medical care (45/308 [14.6%] vs 20/81 [24.7%]; OR, 0.45 [95% CI, 0.2-0.9]). Participants with Delta and Omicron infections who received the second dose 14 to 149 days before infection had a significantly lower mean viral load compared with unvaccinated participants (3 vs 4.1 log 10 copies/μL; difference, −1.0 [95% CI, −1.7 to −0.2] for Delta and 2.8 vs 3.5 log 10 copies/μL, difference, −1.0 [95% CI, −1.7 to −0.3] for Omicron).
Conclusions and Relevance In a cohort of US essential and frontline workers with SARS-CoV-2 infections, recent vaccination with 2 or 3 mRNA vaccine doses less than 150 days before infection with Delta or Omicron variants, compared with being unvaccinated, was associated with attenuated symptoms, duration of illness, medical care seeking, or viral load for some comparisons, although the precision and statistical significance of specific estimates varied.
Introduction
There are limited data on the epidemiology of asymptomatic SARS-CoV-2 infections or mild to moderately severe COVID-19 disease, which constitute most SARS-CoV-2 infections and are believed to trigger and fuel SARS-CoV-2 outbreaks.1-4 Although early reports from prospective community cohorts observed that COVID-19 vaccines attenuated COVID-19 symptoms, duration, and viral RNA shedding,5-8 results have been mixed in recent studies,9-11 which may be due to the waning of vaccine-associated immunity,12,13 changes in virologic features and heightened immune evasion by the Delta14 and Omicron variants,15 or a combination of factors. Data are also limited from studies that conduct routine weekly testing for SARS-CoV-2 in the community or compare multiple SARS-CoV-2 strains within the same population.
This study describes a prospective cohort of essential and frontline workers in 6 US states with SARS-CoV-2 infection identified through weekly surveillance testing irrespective of symptoms and has 3 objectives. First, the study described COVID-19 by characterizing clinical and virologic features of infections with the original strains (origin), B.1.617.2 and all AY lineages (Delta), and B.1.1.529 and BA1 lineages (Omicron) and assessing associations between clinical characteristics and viral RNA load. Second, the study assessed the association of second and third doses of mRNA COVID-19 vaccines with the occurrence of symptomatic disease, symptoms, and viral load. Third, the study examined the potential transmissibility of COVID-19 by clinical presentation and vaccination status assessing viral viability by culture for the origin and Delta variant viruses.
Methods
Study Design and Participants
The HEROES-RECOVER network consists of large prospective cohorts of health care personnel, first responders, and other essential and frontline workers, defined as those who work at least 20 hours per week in an occupation involving regular contact within 3 feet of others, such as workers in the following sectors: education, agriculture, food processing, transportation, solid waste collection, utilities, government services, childcare, environmental services, and hospitality. The study was conducted in 6 US states (Arizona, Florida, Minnesota, Oregon, Texas, and Utah) with participants recruited through ongoing occupational cohort studies, health management organization membership, and university or hospital system employment using a stratified enrollment approach according to age and occupational category (eMethods in Supplement 1).16,17 All participants provided written informed consent, and their common protocols and procedures were approved by the institutional review boards at participating sites. The data for these analyses were collected from December 14, 2020, (after authorization of mRNA vaccines) to April 19, 2022, with follow-up until May 9, 2022, reported. The analysis was conducted among participants who tested positive for SARS-CoV-2, excluding those with SARS-CoV-2 infections prior to the study start date, infected prior to completion of a primary mRNA COVID-19 vaccine series (<14 days after dose 2), infected either less than 7 days or at greater than or equal to 150 days after receipt of dose 3, or who received the Ad26.COV2.S (Janssen/Johnson & Johnson) vaccine (eFigure 1 in Supplement 1).
Participants reported sociodemographic information, including self-reported race and ethnicity, history of SARS-CoV-2 infection, and chronic medical conditions (using a combination of fixed categories and open-ended text). Race and ethnicity were assessed due to the potential for differential vaccine receipt by race and ethnicity. Active surveillance for symptoms associated with COVID-19 (defined as presence of 1 or more of the following: fever, chills, cough, shortness of breath, sore throat, diarrhea, muscle aches, or a change in smell or taste) was conducted through weekly text messages, emails, and reports obtained directly from the participant.
COVID-19 vaccination status was reported by participants via online surveys and verified through direct upload of images of vaccination cards at all sites. In addition, electronic medical records, occupational health records, or state immunization registries were reviewed by site staff in Minnesota, Oregon, Texas, and Utah.
Laboratory Testing
Participants self-collected midturbinate nasal swabs, irrespective of symptoms, each week and at the onset of COVID-19 symptoms. Specimens were shipped on cold packs and tested by qualitative reverse transcriptase–polymerase chain reaction (RT-PCR) assay at the Marshfield Clinic Research Institute. Virus lineage was determined by whole genome sequencing18 or imputed as the predominant circulating virus (eMethods in Supplement 1). To estimate viral shedding, viral RNA load (hereafter referred to as viral load) was assessed by quantitative RT-PCR assays conducted at the Wisconsin State Laboratory of Hygiene.
To assess virus viability, infectious virus particles were quantified by determination of plaque-forming units (PFU) per milliliter on Vero cells ectopically expressing TMPRSS3 and human ACE219 at St Jude Children’s Research Hospital. Infected cell monolayers were stained and PFUs were counted visually. Specimens were eligible for PFU assessment if the RT-PCR cycle threshold was less than 30, virus lineage had been confirmed, and the viral load was greater than 1 log 10 RNA copies/μL (eMethods in Supplement 1). Attempts to culture Omicron variant viruses were not successful.
Exposures
SARS-CoV-2 infection lineage and vaccination status were the primary exposures of interest. Virus lineage was categorized as origin (defined as viruses circulating prior to recognized variants), Delta variant, or Omicron variant. The infection index date was defined as the date of illness onset or specimen collection if asymptomatic. Vaccination status was categorized by doses received relative to the infection index date: 2 doses received within 14 to 149 days prior to infection, 2 doses received at least 150 days prior to infection, and 3 doses received within 7 to 149 days prior to infection. The comparator group was unvaccinated.
Outcomes
Clinical outcomes were presence of symptomatic COVID-19 and, among cases of symptomatic COVID-19, the number of symptoms (0 to 19), presence of fever or chills, illness duration, number of days sick in bed for at least half the day,20 hours of work missed due to illness, and medical care seeking. Clinical outcomes were self-reported via online surveys sent to participants on reporting of COVID-19 symptoms. Additional online surveys were sent after 7 days if symptoms were ongoing and on illness resolution; asymptomatic participants did not receive illness surveys but confirmed absence of symptoms at the time RT-PCR test results were communicated.
Virologic outcomes were viral load and PFU counts from self-collected specimens. When viral load was available from more than 1 specimen per infection, the primary analyses referred to the specimen with the highest viral load. Viral load is reported as overall viral load, peak viral load, time until peak, and time until load was less than 1. Viral load range was 10 to 6 million U/μL. PFU range was 20 to 11 000 U/mL. Both outcomes were analyzed and reported as log 10 units.
Statistical Analysis
For COVID-19 clinical outcomes, odds ratios (ORs) were estimated with logistic regression. Mean differences for clinical and virologic outcomes were estimated with linear regression. All models were adjusted for race, ethnicity, and number of daily prescription medications for potential association with vaccination status and outcomes (eTables 3 and 4 in Supplement 1). Occupation was included in adjusted models for the association between virus lineage and missed work, but could not be included in models assessing the association of vaccination status and missed work due to collinearity with vaccine exposure. Analyses of viral load and PFU counts were adjusted for days between index and specimen collection dates. Model assumptions, including linearity, collinearity, and equal variance for linear models, were assessed and met.
Similar to previously published methods,21 viral load by days since onset was estimated using bayesian linear hierarchical models with virus lineage or presence of symptomatic illness as exposures. Time since index date was fitted as a 3-knot natural spline. Models were fit among participants with at least 3 positive RT-PCR specimens and at least 1 specimen with quantifiable load. Viral load was predicted for positive specimens with too little virus to quantify using the linear relationship between the qualitative RT-PCR cycle threshold value and viral load. A sensitivity analysis required at least 2 quantifiable specimens (eMethods in Supplement 1). Models converged with R-hat less than 1.05 and effective sample size greater than 100.
Two-sided 95% CIs for ORs that did not include 1.0 and mean differences that did not include 0 were considered statistically significant. All analyses were conducted with SAS software, version 9.4 (SAS Institute), and R software, version 4.0.2 (R Foundation for Statistical Computing). Because of the potential for type I error due to multiple comparisons, findings should be interpreted as exploratory.
Results
Cohort Characteristics
Of 7223 participants enrolled in active surveillance, 1710 were excluded due to prior SARS-CoV-2 infections; of the remaining 5513 participants, 997 (18.1%) were primary health care personnel (defined as physicians, physician assistants, nurse practitioners, and dentists), 1785 (32.4%) were nurses and other allied health professionals, 1218 (22.1%) were first responders, and 1513 (27.4%) were other essential/frontline workers. Participants completed a median (IQR) of 90.5% (78.1%-97.0%) of weekly surveillance reporting and specimen collection. Ninety-five percent of illness surveys were completed, with a range of 94.0% to 98.5% completion across the vaccination status groups. A total of 1493 infections were identified by RT-PCR testing. After exclusion criteria (eFigure 1 in Supplement 1), the study sample consisted of 1199 unique infections confirmed by RT-PCR testing during the 70-week study period (Table 1); these consisted of 14.0% origin, 24.0% Delta, and 62.0% Omicron infections. Among the 1199 infected participants, the median (IQR) age was 41 (35.0-50.5) years, 714 (59.5%) were women, 232 (19.3%) were Hispanic, 22 (1.8%) were non-Hispanic Asian, 35 (2.9) were non-Hispanic Black, and 871 (72.6%) were non-Hispanic White (eTable 2 in Supplement 1). At least 1 chronic medical condition was reported for 27.4% of infected participants, and 27.7% took at least 2 medications per day. Characteristics of the uninfected participants not included in this study are shown in Table 1.
Of the 1199 participants with SARS-CoV-2 infections, 352 (29.4%) were unvaccinated, 72 (6.0%) received the second dose 14 to 149 days before infection, 374 (31.2%) received the second dose at least 150 days before infection, and 401 (33.4%) received the third dose 7 to 149 days before infection (Table 2; eTable 1 in Supplement 1). Of those who received the third dose, 11 (2.7%) were immunocompromised by self-report; dose 3 was considered primary series vaccination for these individuals. All participants received the same vaccine product for dose 1 and dose 2: 280 (62.3%) received BNT162b2 (Pfizer-BioNTech) and 166 (37.2%) received mRNA-1273 (Moderna) vaccines. For dose 3, a total of 22 of 401 individuals (5.5%) received a different mRNA vaccine product than their primary series.
Clinical Characteristics of COVID-19 Among Unvaccinated Participants
To describe the clinical characteristics of infection by each viral lineage, the 352 episodes of infection that occurred among unvaccinated individuals (among the 1199 total infection episodes) were examined (Table 3); of these, 44 (12.5%) were asymptomatic and 24 (6.8%) had symptoms not in the COVID-19 case definition. Asymptomatic infections were significantly more commonly associated with Omicron compared with Delta variants (OR, 5.6 [95% CI, 1.6-19.6]). When symptomatic, 64.8% of participants with COVID-19 reported fever or chills; the percentage was significantly higher for Delta compared with origin infection (84.9% vs 71.8%; OR, 2.2 [95% CI, 1.1-4.8]). Medical care seeking was significantly less common among Omicron infections (24.7%) compared with Delta (47.3%) infections (OR, 0.39 [95% CI, 0.2-0.8]), but was not significantly different among Omicron vs origin (36.9%) infections (OR, 1.8 [95% CI, (1.0-3.2)]).
Among symptomatic participants, COVID-19 symptoms associated with infection due to the Omicron variant lasted a mean (SD) of 12.3 (9.1) days, which was 2.9 (95% CI, −5.5 to −0.3) days fewer than origin (mean of 15.6 days) infection and 3.6 (95% CI, −6.6 to −0.7) days fewer than Delta (mean of 16.4 days) infection. Participants with Omicron infection spent a mean (SD) of 2.6 (3.1) days sick in bed, which was a mean of 1.2 (95% CI, −2.5 to −0.0) days fewer than those with origin infection and 2.0 (95% CI, −3.4 to −0.6) days fewer than those with Delta infection. Further findings regarding the associations between clinical and participant characteristics are in Table 3 and eTable 3 in Supplement 1.
Comparison of Delta and Omicron Clinical Characteristics by Vaccination Status
Vaccinated and unvaccinated participants were examined to assess potential vaccine-associated attenuation of disease. Among participants with Delta infections, illness was milder among those who were vaccinated, although the precision of these estimates varied (Table 2). Participants who received the second dose 14 to 149 days before Delta infection compared with those unvaccinated were significantly less likely to be symptomatic (77.8% vs 96.1%; OR, 0.13 [95% CI, 0.0-0.6]) and spent significantly fewer days sick in bed (2.3 vs 4.9; difference, −2.3 [95% CI, −4.4 to −0.2] days). Symptomatic participants who received dose 2 at least 150 days prior to Delta infection compared with those who were unvaccinated were significantly less likely to report fever or chills (70.4% vs 84.9%; OR, 0.39 [95% CI, 0.2-0.9]) or seek medical care (31.5% vs 47.3%; OR, 0.51 [95% CI, 0.3-1.0]) and missed significantly fewer hours of work (47.1 vs 62.8; difference, −15.2 [95% CI, −26.4 to −4.1] hours). Participants who received the third vaccine dose 7 to 149 days prior to Delta infection compared with those who were unvaccinated were significantly less likely to have symptomatic illness (72.2% vs 77.8%; difference, 0.13% [95% CI, 0.0%-0.7%]) and, when symptomatic, were significantly less likely to report fever or chills (38.5% vs 84.9%; OR, 0.07 [95% CI, 0.0-0.3]) and spent significantly fewer days symptomatic (mean [SD] of 10.2 [5.4] days vs 16.4 [8.6]; difference, −6.1 [95% CI, −11.8 to −0.4] days).
Significant differences in clinical presentation between unvaccinated individuals and those who received 2 vaccine doses were not observed among Omicron variant infections. However, symptomatic participants who received the third vaccine dose 7 to 149 days prior to Omicron infection compared with those who were unvaccinated were significantly less likely to report fever or chills (51.5% vs 79.0%; OR, 0.25 [95% CI, 0.1-0.5]) or receive medical care (14.6% vs 24.7%; OR, 0.45 [95% CI, 0.2-0.9]) and missed significantly fewer hours of work (mean [SD] of 26.6 [25.3] vs 35.9 [31.5]; difference, −11.1 [95% CI, −18.4 to −3.9] hours), although guidance on postinfection isolation and returning to work changed during the Omicron wave. Participants who received the third vaccine dose 7 to 149 days prior to Omicron infection were more likely to be symptomatic than those who were unvaccinated (88.4% vs 79.4%; OR, 2.0 [95% CI, 1.1-3.5]).
Viral Load
For 886 participants (73.8%), the specimen with the highest viral load was their first positive specimen collected. Mean viral load did not differ by demographic characteristics (eTable 4 in Supplement 1). Symptomatic infections compared with asymptomatic infections had a significantly higher mean viral load (mean difference, 1.3 log 10 copies/μL [95% CI, 0.3-1.6]; Figure 1); this difference occurred consistently for Delta and Omicron infections (eTable 5 in Supplement 1). For all studied lineages, mean viral load was significantly higher for infections with fever or chills (3.8 vs 3.4 log 10 copies/μL; mean difference, 0.4 [95% CI, 0.2-0.7]), those with 8 or more symptoms (3.9 vs 3.0 log 10 ; mean difference, 0.9 [95% CI, 0.6-1.3] log 10 copies/μL), and those with symptoms lasting at least 14 days (4.0 vs 3.5; mean difference, 0.6 [95% CI, 0.3-0.9] log 10 copies/μL) (Figure 1 and eTable 5 in Supplement 1).
Participants with Delta and Omicron infections who received the second vaccine dose 14 to 149 days before infection had a significantly reduced mean viral load compared with unvaccinated participants (3 vs 4.1 log 10 copies/μL; mean difference, −1.0 [95% CI, −1.7 to −0.2] log 10 copies/μL for Delta; 2.8 vs 3.5 log 10 copies/μL; mean difference, −1.0 [95% CI, −1.7 to −0.3] log 10 copies/μL for Omicron) (Figure 1).
Viral Load Peak and Decline
Viral load over the course of infection was estimated using a bayesian hierarchical model among 407 vaccinated and unvaccinated participants who contributed at least 3 specimens positive for SARS-CoV-2 (eTable 6 in Supplement 1). The time from SARS-CoV-2 detection to mean peak viral load was 2.5 days for all 3 virus lineages. Overall mean viral load for origin and Delta infections were higher than for Omicron (mean difference, 0.2 [95% credible interval (CrI), 0.0-0.4] log 10 copies/μL for origin infection; mean difference, 0.2 [95% CrI, 0.1-0.5] log 10 copies/μL for Delta infection) (Figure 2A). During the estimated peak, viral load did not differ by viral lineage. The estimated time of decline in viral load from peak to a count of less than 1 log 10 copies/μL was 11 days for both origin and Delta infection and 10 days for Omicron infection (Figure 2A).
Across viral lineages, the mean viral load for asymptomatic infections was lower than for symptomatic COVID-19 (mean difference, 0.7 [95% CrI, 0.4-1.1] log 10 copies/μL) (Figure 2B). Estimated peak viral load was also lower for asymptomatic infections (3.8 [95% CrI, 3.4-4.2] log 10 copies/μL) than for symptomatic COVID-19 (4.5 [95% CrI, 4.4-4.6] log 10 copies/μL) (eTable 6 in Supplement 1). The estimated time from peak viral load to a count of less than 1 log 10 copies/μL was 8 days for asymptomatic infections and 10 days for symptomatic COVID-19 (Figure 2B).
Infectious SARS-CoV-2 Counts
Counts of viable viral particles were only examined for unvaccinated participants infected with origin viruses and among both vaccinated and unvaccinated participants infected with the Delta variant (Figure 1). Mean PFU was significantly higher for participants with symptomatic COVID-19 compared with those with asymptomatic infections (4.2 vs 3; mean difference, 1.9 [95% CI, 0.6-3.2] PFU/mL). Mean PFU was significantly lower among participants who received dose 2 at least 150 days before breakthrough infection or the third dose 7 to 149 days before breakthrough infection compared with unvaccinated participants (4.1 vs 4.8; mean difference, −0.8 [95% CI, −1.5 to −0.1] PFU/mL and 3.1 vs 4.8; mean difference, −1.6 [95% CI, −2.8 to −0.5] PFU/mL) (Figure 1).
Discussion
In this prospective cohort of health care personnel, first responders, and other essential and frontline workers followed up from December 14, 2020, to April 19, 2022, recent receipt of 2 or 3 doses of COVID-19 mRNA vaccine 14 to 149 days prior to Delta or Omicron BA.1 SARS-CoV-2 infection was variably associated with a milder clinical presentation, shorter illness duration, less medical care seeking, and lower viral load, although the precision of specific estimates varied.
The differences between Omicron and Delta infections in clinical and virologic characteristics observed may help explain the previously reported increased transmissibility of the Omicron variant.3,22,23 COVID-19 symptoms associated with Omicron appeared to be milder and of a shorter duration than COVID-19 associated with Delta by many indicators. Additionally, milder symptoms that lasted fewer days and shorter duration of illness was associated with lower viral RNA shedding. In this cohort, 20.2% of infections due to Omicron were asymptomatic, similar to another recent report.24 In contrast, only 3.9% of Delta infections were asymptomatic and were associated with a longer duration of illness. Across all viral lineages in this study, infected participants without symptoms shed viral RNA for about 12 days after peak viral load compared with symptomatic participants with symptoms, who shed viral RNA for about 18 days. During the period of detection, the viral load of Omicron BA.1 infections was similar to Delta infections and higher than infections with origin strains. Although viral RNA shedding cannot be directly attributable to transmission, the relatively high viral load of Omicron infections together with the higher frequency of asymptomatic infection supports previous studies suggesting an association with increased transmission,21,25 particularly during the first 3 to 5 days when viral load peaked.
A study from the HEROES-RECOVER cohort previously reported that participants with origin and other pre-Delta virus infections (from December 2020 to April 2021) after at least 1 dose of mRNA COVID-19 vaccine had lower viral load, were less likely to have febrile symptoms, and had shorter illnesses with fewer days spent sick in bed compared with unvaccinated participants.8 Similarly, in the current study, Delta-associated breakthrough infections 14 to 149 days after dose 2 had lower viral loads compared with those in unvaccinated participants; further, breakthrough infections 14 to 149 days after dose 2 or 7 to 149 days after dose 3 had shorter illnesses, with lower counts of viable and presumed infectious virus. For Omicron infections, only participants with a breakthrough infection after dose 3 showed consistent signs of vaccine attenuation of COVID-19, such as being less likely to have febrile symptoms or seek medical care. It is possible that the recall of immunologic memory that reduces viral replication and accelerates elimination of virally infected cells that may underlie vaccine attenuation of disease26 wanes with time, similar to waning in neutralizing antibodies27 and clinical effectiveness of mRNA vaccines.12
Limitations
This study has several limitations. First, although this was among the largest studies of its kind with routine surveillance testing for SARS-CoV-2, stratification by virus lineage and vaccine exposures resulted in relatively small subgroups, which reduced the precision of estimates. Second, although several relevant confounders were controlled for, unmeasured confounding in this observational study could have occurred. Limitations in sample size and the ability to adjust models for potential confounders made it particularly difficult to interpret unexpected findings, such as a higher percentage of individuals with symptomatic disease among those vaccinated with the third vaccine dose 14 to 149 days before Omicron infection compared with those who were unvaccinated. Third, findings regarding virologic features of COVID-19 are limited to virus present in the nasal cavity, self-collected with midturbinate nasal swabs. Therefore, possible differences between variants in viral RNA present in the full nasopharynx or lower respiratory system could not be examined. Fourth, due to methodological obstacles culturing Omicron variant viruses, virus viability could only be examined with origin and Delta viruses. Fifth, the detection of viral RNA and isolation of viable virus through virus culture are not direct measures of virus transmissibility, although evidence is emerging for SARS-CoV-225 and both are associated with transmission of other viral infections.28-30 Sixth, findings on missed work hours should be interpreted with caution because guidance on postinfection isolation and returning to work changed over time and work-related isolation protocols were not considered. Seventh, participant-reported symptoms and duration of illness may be subject to recall and confirmation biases. Eighth, these findings may not generalize to all population in the US, especially given the relatively low percentage of participants who were not White.
Conclusions
In a cohort of US essential and frontline workers with SARS-CoV-2 infections, recent vaccination with 2 or 3 mRNA vaccine doses less than 150 days before infection with Delta or Omicron variants, compared with being unvaccinated, was associated with attenuated symptoms, duration of illness, medical care seeking, and viral load for some comparisons, although the precision and statistical significance of specific estimates varied.
Back to top Article Information
Corresponding Author: Ashley L. Fowlkes, ScD, MPH, Influenza Division, Centers for Disease Control and Prevention, 1600 Clifton Rd, Atlanta, GA 30333 (ahl4@cdc.gov).
Accepted for Publication: September 20, 2022.
Author Contributions: Drs Thompson and Ms Grant had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Thompson, Yoon, Naleway, Meece, Fabrizio, Caban-Martinez, Burgess, Gaglani, Olsho, Grant, Phillips, Mak, Hegmann, Lutrick, Hunt, Fowlkes, Thiese, Herring, Respet, Bruner, Mayo Lamberte, Beitel, Webby, Tyner.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Thompson, Yoon, Caban-Martinez, Burgess, Grant, Lundgren, Fowlkes, Thiese, Jones, Dickerson, Mayo Lamberte.
Critical revision of the manuscript for important intellectual content: Thompson, Yoon, Naleway, Meece, Fabrizio, Caban-Martinez, Burgess, Gaglani, Olsho, Grant, Phillips, Groom, Bateman, Stefanski, Schaefer Solle, Ellingson, Dunnigan, Wesley, Mak, Hegmann, Kuntz, Guenther, Bissonnette, Lutrick, Rose, Morrill, Hunt, Fowlkes, Thiese, Hollister, Herring, Respet, Yoo, Bruner, Rivers, Fleary, Odean, Praggastis, Bedrick, Romine, Khan, Beitel, Webby, Tyner.
Statistical analysis: Thompson, Grant, Mak, Thiese, Yoo, Jones, Bedrick.
Obtained funding: Thompson, Yoon, Burgess, Olsho, Thiese.
Administrative, technical, or material support: Thompson, Yoon, Naleway, Meece, Fabrizio, Caban-Martinez, Burgess, Gaglani, Olsho, Phillips, Groom, Bateman, Stefanski, Schaefer Solle, Ellingson, Wesley, Lundgren, Hegmann, Kuntz, Bissonnette, Lutrick, Morrill, Hunt, Fowlkes, Thiese, Hollister, Herring, Respet, Bruner, Rivers, Fleary, Praggastis, Dickerson, Mayo Lamberte, Khan, Beitel, Webby, Tyner.
Supervision: Thompson, Yoon, Meece, Olsho, Lundgren, Hegmann, Kuntz, Bissonnette, Thiese, Bruner, Praggastis, Webby.
Conflict of Interest Disclosures: Dr Naleway reported receiving research funding from Pfizer and Vir Biotechnology outside the submitted work. Drs Webby and Fabrizio are funded by the National Institute of Allergy and Infectious Diseases, National Institutes of Health (grant no. HHSN2272201400006C), and ALSAC. Dr Gaglani reported receiving grants from Janssen and Pfizer and serving as co-chair of the Infectious Diseases and Immunization Committee, Texas Pediatric Society, Texas Chapter of the American Academy of Pediatrics. Dr Hegmann reported receiving personal fees from the American College of Occupational and Environmental Medicine/Reed Group for work as an editor of peer-reviewed guidelines, including COVID-19 guidelines, outside the submitted work. Dr Kuntz reported receiving research grant funding to the institution from Vir Biotechnology, Pfizer, and Novartis outside the submitted work. No other disclosures were reported.
Funding/Support: Funding was provided in whole or in part by federal funds from the National Center for Immunization and Respiratory Diseases, Centers for Disease Control and Prevention (CDC), under contract numbers 75D30120R68013 awarded to Marshfield Clinic Research Laboratory, 75D30120C08379 to University of Arizona, and 75D30120C08150 awarded to Abt Associates, Inc.
Role of the Funder/Sponsor: The CDC was involved in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, and approval of the manuscript; and decision to submit the manuscript for publication.
Disclaimer: The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
Data Sharing Statement: Data will be available by CDC when objectives of the research are complete.
Additional Information: HEROES-RECOVER Network includes Arizona Healthcare, Emergency Response and Other Essential Workers Surveillance Study (HEROES) and Research on the Epidemiology of SARS-CoV-2 in Essential Response Personnel (RECOVER).