Impact of COVID-19 Infection on Health-Related Quality of Life, Work Productivity and Activity Impairment by Symptom-Based Long COVID Status and Age in the US

COVID-19 infection adversely impacts patients’ wellbeing and daily lives. This survey-based study examined differences in patient-reported COVID-19 symptoms, Health-Related Quality of Life (HRQoL) and Work Productivity and Activity Impairment (WPAI) among groups of patients defined based on age and symptom-based long COVID status. Symptomatic, COVID-19-positive US outpatients were recruited from 31 January–30 April 2022. Outcomes were collected via validated instruments at pre-COVID, Day 3, Week 1, Week 4, Month 3 and Month 6 following infection, with changes assessed from pre-COVID and between groups, adjusting for covariates. EQ-5D-5L HRQoL and WPAI scores declined in all groups, especially during the first week. Long COVID patients reported significantly higher symptoms burden and larger drops in HRQoL and WPAI scores than patients without long COVID. Their HRQoL and WPAI scores did not return to levels comparable to pre-COVID through Month 6, except for absenteeism. Patients without long COVID generally recovered between Week 4 and Month 3. Older (>50) and younger adults generally reported comparable symptoms burden and drops in HRQoL and WPAI scores. During the first week of infection, COVID-19-related health issues caused loss of 14 to 26 work hours across the groups. These data further knowledge regarding the differential impacts of COVID-19 on clinically relevant patient groups.


COVID-19 causes a wide variety of acute and long-term symptoms and multi-organ health problems that can persist for weeks or months following infection [1].
There is a rapidly growing body of evidence suggesting that persistent COVID-19 symptoms can negatively impact patients' Health-Related Quality of Life (HRQoL), productivity and activity levels [2][3][4][5][6].To date, however, few studies have provided a comprehensive characterization of the impact of COVID-19 on these patient outcomes during the entire course of illness and on the extent to which these differ according to relevant clinical risk factors and patient characteristics.
Studies and literature reviews investigating determinants of HRQoL in adults with COVID-19 have associated certain baseline characteristics with HRQoL, measured using patient-reported EQ-5D tools [7][8][9][10][11].These include female gender, pre-existing comorbidity and unemployment status.Results for age are mixed, calling for further research: some studies associated younger age with lower HRQoL [7,8], while others reported older age as a determinant of poor HRQoL [9][10][11].
Features of COVID-19 disease have also been associated with diminished HRQoL, including acute disease severity (e.g., ICU admission), not being vaccinated and symptoms burden at the time of infection [7][8][9][10][11].Although numerous studies have explored the burden of COVID-19 symptoms on HRQoL, most assessed the burden separately for the acute and long COVID phases and compared the HRQoL in long COVID patients with limited control groups [12][13][14][15].
Better understanding of signs, symptoms and magnitude of effects of COVID-19 on productivity, daily activities and HRQoL by patient characteristics and infection features can help healthcare professionals guide prevention and management efforts.Moreover, the data could inform estimates of indirect costs and health utilities in economic evaluations to better understand the economic burden of COVID-19 and the broad value of primary and secondary prevention.
In a prior study, we assessed differences in patient-reported symptoms, HRQoL and Work Productivity and Activity Impairment (WPAI) by vaccination status among 328 adult symptomatic outpatients testing positive for COVID-19 at US community pharmacies and reporting persisting symptoms four weeks following infection.Consistent with the existing body of evidence, we found that pre-infection vaccination status was associated with a lower risk and burden of long COVID symptoms, resulting in better HRQoL and lower WPAI compared to unvaccinated participants [6].As a continuation of our research efforts on the impact of acute and long COVID on patients' daily lives, we used this previously described cohort [6] to assess differences in patient-reported outcomes according to age and long COVID status up to 6 months after infection.
The objectives of this study were threefold: (1) to describe characteristics and symptoms of patients by long COVID status and age; (2) to evaluate changes in work productivity, activity levels and HRQoL by long COVID status and age; and (3) to supplement prior research with new quantitative measures of work productivity, expressed as work hours lost and actual hours worked.

Study Design and Participants
The study design has been previously described (clinicaltrials.govNCT05160636) [6].Briefly, this was a nationwide prospective Patient-Reported Outcomes (PROs) survey-based study targeting adults at least 18 years of age who tested positive for COVID-19 using Reverse Transcription Polymerase Chain Reaction (RT-PCR) at one of over ~5000 CVS Health test sites across the United States and had self-reported at least one symptom at the time of testing.Patients were eligible for these study analyses if they had symptoms lasting more than four weeks post-acute infection.A total of 328 patients meeting the inclusion and exclusion criteria were selected for this study from a previously described cohort [6].Recruitment took place between 31 January 2022 and 30 April 2022, with follow-up through 30 October 2022.

Baseline Characteristics and Acute Symptoms
The baseline characteristics of participants were acquired using the CVS Health pretest screening questionnaire which comprised self-reported information regarding demographics, comorbidities, COVID-19 vaccination history and social determinants of health, including the Social Vulnerability Index (SVI), and work or residency in a high-risk or healthcare setting.The acute COVID-19 symptoms list from the CDC was utilized to record participant symptoms at the time of testing [1].

Long-Term COVID-19 Symptoms
The study assessed the presence of long COVID symptoms via a questionnaire that included 20 symptoms based on the CDC long COVID symptom list updated in 2022 [1].The questionnaire was administered beginning at four weeks post-enrollment, with followup questionnaires at Months 3 and 6 post-enrollment.In alignment with the CDC definition of long COVID, we considered Week 4 as the start of long-term symptoms [1].The list of symptoms included general symptoms (tiredness/fatigue, symptoms exacerbated by physical or mental activities, fever, general pain/discomfort), respiratory and cardiac symptoms (difficulty breathing or shortness of breath, cough, chest or stomach pain, heart palpitations), neurological symptoms (change in smell or taste, headache, lightheadedness, "brain fog," numbness, sleep problems, mood changes, memory loss) and other symptoms (rash, diarrhea, joint or muscle pain, menstrual cycle irregularities).

Exposure Groups
The previously described cohort of 328 participants reporting symptoms at Week 4 following infection [6] was categorized and pre-specified into exposure groups based on their age and long COVID status.They were classified as older adults if they reported being 50 years of age and older at the time of testing and were classified as younger adults if younger than 50 years of age.
Existing studies describing the prevalence of long COVID have employed different thresholds for duration and intensity of symptoms, as well as differing sets of symptoms [16][17][18].As previously described [6], our study leveraged the clinical case definition of long COVID from the CDC-funded INSPIRE registry, which, as previously reported, employed a similar list and number of symptoms [19,20].In our main analyses (base case), a patient was classified as having long COVID if reporting ≥3 symptoms at long COVID start (Week 4).In sensitivity analyses, the study used an alternative cutoff threshold of ≥2 symptoms at Week 4.

Health-Related Quality of Life (HRQoL)
Our study assessed HRQoL via the EQ-5D-5L questionnaire [21] that subjects were asked to complete at enrollment, then 1, 3 and 6 months post-enrollment [22].Five dimensions of EQ-5D-5L at each time point were converted into the Utility Index (UI) using the US-based weights established by Pickard et al. [23].Lower scores for both EQ VAS and UI correspond to lower overall self-reported health-related quality of life.UI and Visual Analogue Scale (VAS) scores were compared among cohorts and across assessment times [21].

Work Productivity and Activity Impairment
The Work Productivity and Activity Impairment General Health v2.0 (WPAI:GH) measure was used to measure impairments in both paid and unpaid work [24,25].Participants were asked to complete the survey seven days after presenting for testing and again at Months 1, 3 and 6 post-enrollment.Higher scores correspond to greater activity impairment and work productivity loss.Only employed participants were included for work productivity analyses.WPAI results were compared across cohorts and assessment times.

Statistical Methods
To summarize participant characteristics at baseline and outcomes at follow-up, means and standard deviations for continuous variables and frequency and percentages for categorical variables were used.For between-group differences, t-tests and chi-square tests were used to test continuous variables and categorical variables, respectively.When an expected cell frequency was less than 5, Fisher's exact tests were used for 2-by-2 tables and Fisher-Freeman-Halton tests for r-by-c tables [26,27].p-values were all two-sided.
To estimate the impact of long COVID or age (≥50 vs. <50 years old) on HRQoL and WPAI over time, Mixed Models for Repeated Measures (MMRM) were used [28] with an unstructured covariance matrix for categorical assessment time.For each time point of assessment, Least Squares (LS) mean and standard errors of PRO scores were calculated for each cohort and their difference.The EQ-5D-5L UI and WPAI scores were calculated based on their tool guidelines [21,25]; no imputation was made for missing data and the analyses were based on all available data.
Cohen's d, or a variation of it, was calculated to examine the difference in pre-COVID scores among patients with or without long COVID or age ≥ 50 or <50 years old, the magnitude of score change from pre-COVID at Week 4, Month 3 and Month 6 within each cohort, as well as the differences between cohorts (with vs. without long COVID; age ≥ 50 vs.<50 years old) [29,30].Specifically, within-cohort Effect Size (ES) from pre-COVID to follow-up was calculated as mean change scores divided by the standard deviation of change scores [22].Between-cohort ES was calculated as the difference between cohort means divided by the pooled standard deviation for either pre-COVID scores or change scores from pre-COVID to follow-up [22].Values of 0.2, 0.5 and 0.8 Standard Deviation (SD) units represent, respectively, "small," "medium" and "large" effect sizes [22,29].
The study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline [31].SAS Version 9.4 (SAS Institute, Cary, NC, USA) was used to conduct all analyses.

EQ-5D-5L
Participants with long COVID reported lower pre-COVID mean EQ VAS and Utility Index (UI) compared with individuals without COVID-19 (84.9 and 0.88 versus 88.6 and 0.94, respectively; Table 3), in line with the observed differences in baseline comorbid status between the two groups.COVID-19 had a detrimental impact on the quality of life of both groups, with the largest ESs for the mean changes from pre-COVID versus soon after infection, at Day 3. At Week 4, patients without long COVID had model-based EQ VAS and UI scores that were numerically lower although comparable to pre-COVID baseline (85.8 and 0.91 versus 88.8 and 0.94, respectively; Table 3).On the other hand, quality of life scores for patients with long COVID did not return to levels comparable to pre-COVID at any point in time.At Month 6, their EQ VAS and UI scores were still significantly lower than pre-COVID (79.3 and 0.80 versus 84.9 and 0.88, respectively; Table 3).The model-based EQ VAS and UI scores in patients with long COVID were significantly lower versus those without long COVID across all time points of the acute (Day 3, Week 4) and long-term (Month 3 and Month 6) survey phases, with medium-to-large ESs (Table 3).

Work Productivity and Activity Impairment
A total of 245 participants (75%) reported being employed at baseline and were eligible to complete the work productivity questions.Of those, 94 (38%) had long COVID and 151 (62%) did not have long COVID.
At enrollment, participants reported pre-COVID mean values for absenteeism, work productivity, missed work hours and actual hours worked that were not significantly different between the two groups.Instead, participants without long COVID reported significantly lower presenteeism than those with long COVID.
COVID-19 had a large impact on all WPAI scores at Week 1 (Table 3).In both groups, the absenteeism levels returned to levels comparable to baseline after Week 4. The presenteeism levels in participants with long COVID continued to be impacted and, at Month 6, did not yet return to levels comparable to baseline.On the other hand, presenteeism levels of participants without long COVID returned to levels comparable to baseline at Week 4.
At enrollment, participants with and without long COVID reported similar pre-COVID baseline mean values of 37 to 38 actual hours worked per week and 4 to 5 missed work hours per week.COVID-19-related health issues were associated with, respectively, 22.3 and 14.3 work hours lost during Week 1 (Table 3).Study participants reported being able to work for, respectively, 15.6 and 21.6 hours during Week 1.The mean work productivity loss at Week 1 was, respectively, 68.0% and 55.0%, corresponding to large ESs in the mean change from pre-COVID baseline and medium ES in the mean changes between the two groups.At Week 4 and Month 6, participants with long COVID did not return to work productivity levels comparable to baseline.
The participants with long COVID reported a pre-COVID mean Activity Impairment (AI) score of 21.3%, significantly higher than those without long COVID (11.1%) (Table 3).
At Week 1, the model-based AI increased to 57.3% and 40.7% in subjects with and without long COVID, corresponding to large ESs in the mean change from pre-COVID baseline and medium ES in the mean changes between the two groups.The AI scores returned to levels comparable to baseline from Week 4 in participants without long COVID.The model-based AI scores for the subjects with long COVID were numerically similar between Week 4 (27.8%),Month 3 (22.5%)and Month 6 (27.9%) and did not return to pre-COVID levels at any point in time.A summary of observed EQ-5D-5L and WPAI results for those with and without long COVID, with long COVID defined as reporting three or more symptoms, is presented in Supplemental Table S2.
A sensitivity analysis was carried out with long COVID defined as reporting two or more symptoms at Week 4. The results with this alternative definition of long COVID were consistent with the results reported in the main analysis.A summary of observed EQ-5D-5L and WPAI results using the alternative definition is presented in Supplemental Table S3.Model-based LSE estimates are presented in Supplemental Table S4.

Patient Characteristics and Symptoms
Of 328 study participants, 95 were 50 years or older (older adults) and 233 were younger than 50 (younger adults).At enrollment, compared with younger adults, older patients did not differ by vaccination status, race, geography, SVI, prior infection status and number of acute COVID-19 symptoms at the time of infection (Table 4).However, the older adult group was characterized by a higher proportion of males (33.7% vs. 23.2%)and higher comorbidity burden (41.1% vs. 20.6% with at least 1 comorbidity; 0.57 vs. 0.26 mean number of comorbidities).Diabetes and heart conditions or hypertension were significantly more prevalent among older adults.A lower proportion of them reported working in healthcare (5.3% vs. 13.7%)and, in general, in high-risk settings (4.2% vs. 12.4%) than younger adults.Older adults reported a similar number of acute symptoms as younger adults with a mean of 5.2 vs. 5.5, but significantly less fatigue (48.4% vs. 67.8%)(Table 4).At Week 4, older and younger adults reported a similar mean of, respectively, 3.0 and 3.1 symptoms.The prevalence of symptoms was generally similar across the two age groups, with fatigue being the most prevalent in both groups.Compared with younger adults, older adults experienced less diarrhea (1.2% vs. 8.1%), more joint or muscle pain (31.6% vs. 15.5%) and more cough (37.9% vs. 21.5%)(Table 5).
At Month 3, 81 older adults and 211 younger adults completed the surveys.Both groups reported a mean of 2.7 symptoms, with fatigue persisting as the most prevalent.Compared with younger adults, older adults experienced less diarrhea (1.1% vs. 9.4%) (Table 5).

EQ-5D-5L
The older and younger adults reported similar pre-COVID mean EQ VAS (86.1 and 87.6, respectively) and mean UI scores (0.92 for both) (Table 6).COVID-19 was associated with a decline in HRQoL in both groups, with the largest ESs for the mean changes between Day 3 and pre-COVID (−12.6 and −14.3 for EQ VAS and −0.11 and −0.14 for UI, respectively).In both age groups, the quality of life scores at all time points were significantly lower than at pre-COVID.

Work Productivity and Activity Impairment
Of the 245 participants that reported being employed at baseline and were eligible to complete the work productivity questions, 55 (22%) were older adults and 195 (78%) were younger adults.
At enrollment, participants reported pre-COVID mean values for absenteeism, presenteeism, work productivity loss, missed work hours and actual hours worked that were not significantly different by age.Both older and younger adults experienced the largest issues with work activities at Week 1.Compared to pre-COVID baseline, they reported, respectively, that 55.3% and 47.5% of their working time was missed (absenteeism), 36.8% and 35.4% was impaired (presenteeism) and the total work productivity loss was, respectively, 52.2% and 50.0%.Older and younger adults reported, respectively, a total of 25.6 and 22.1 lost work hours during Week 1 of infection, corresponding to a total loss of 20.2 and 18.2 work hours compared to pre-COVID baseline.These work impairments were comparable between the two age groups (Table 6).
Younger adults experienced more long-term impact than older adults, with percentages of work time impaired and total work productivity loss between Week 4 and Month 6 that were generally higher compared to pre-COVID baseline (Table 6).
COVID-19 was associated with large activity impairments during Week 1 across both age groups.The AI scores returned to pre-COVID levels at Month 6 in older adults and at Month 3 among younger adults (Table 6).
A summary of observed EQ-5D-5L and WPAI results by age is presented in Supplemental Table S5.The sensitivity analysis results using the alternative definition of long COVID (two or more symptoms at Week 4) were consistent with the results reported in the base case main analysis.

Discussion
Evidence on the impacts of a COVID-19 diagnosis and associated health problems on patients' daily lives is still scarce, especially regarding their work experience.Using a previously described cohort of mild symptomatic patients [6], this study aimed to characterize the trajectory of symptoms and impacts of a COVID-19 diagnosis on HRQoL, activity and work outcomes during the progression of illness up to 6 months after infection, by long COVID status and age.
Our study found that all the four patient groups analyzed (<50 years old, >50 years old, with long COVID, without long COVID) experienced a wide range of symptoms and significant declines in HRQoL.These effects were found to be sustained up to 6 months after infection, regardless of long COVID status and age, as none of the groups returned to levels of wellbeing comparable to pre-COVID, except for participants without on-going long COVID symptoms.
We found that participants with long COVID experienced a significantly greater level of burden than those without long COVID symptoms.Compared with individuals without long COVID, they consistently reported a higher mean number of acute and long-term symptoms throughout the 6-month follow-up period.At every time point, they reported a higher prevalence of tiredness or fatigue, difficulty thinking or concentrating, sleep problems, headache and joint or muscle pain compared to participants without long COVID.Participants with long COVID also consistently reported significantly lower mean EQ-5D VAS and utility values throughout the follow-up.
While comparisons versus existing studies are impaired by differences in study design and methods, we found several similarities and consistencies with the long COVID literature.First, our long COVID cohort had traits that were consistent with studies that investigated predictors of long COVID: female gender, pre-existing comorbidity burden and acute symptoms were significantly more prevalent in the long COVID cohort than in the cohort of participants without long COVID [1-5,32].Second, our symptoms burden results are in line with prior research on long COVID that reported fatigue as the most prevalent and pervasive symptom [1-5,7,13,33].Third, our comparative results for symp-toms burden and HRQoL detriments are directionally in line with a study conducted in Japan [15] that employed a relatively similar study design.The author team compared symptoms burden and HRQoL (measured with EQ-5D-3L) between long COVID patients and patients with a history of COVID-19 and no on-going symptoms.Consistent with our study, the authors reported that participants with long COVID reported more prolonged symptoms and an overall greater symptoms burden that was expressed with significantly lower mean EQ VAS and utility values.Additional studies assessing HRQoL detriments in long COVID patients compared outcomes versus normative or no controls, hindering side-by-side comparisons [12][13][14][15].
Literature investigating determinants of HRQoL in adults with COVID-19 using patient-reported EQ-5D tools have reported mixed results for age [7][8][9][10][11].Some studies associated younger age with lower HRQoL [7,8], while others reported older age as a determinant of low HRQoL [9][10][11].In contrast with existing literature, in our study, younger and older adults reported comparable symptoms burden.The prevalence of symptoms was generally similar between older and younger adults, with fatigue being the most prevalent in both groups and with older adults experiencing less diarrhea and more joint or muscle pain over the follow-up period.The two age groups also reported similar declines in HRQoL.While symptom prevalence and HRQoL estimates vary across studies due to heterogeneous definitions, settings and methods, we observed that the magnitude of decline in HRQoL scores during the acute phase was similar to existing research [33], although the evolution over time was found to differ, most likely related to differences in patient characteristics and study design.A longitudinal cohort study in the Belgian adult population testing positive for COVID-19 reported a mean UI before infection of 0.92 (95% CI = 0.917; 0.923) regardless of long COVID status [33].The mean UI was 0.80 (95% CI = 0.795; 0.805) at the time of infection, with the drop greater than the minimal detectable change threshold and considered clinically meaningful [33].In our study, the pre-COVID UI was 0.92 in the two age groups (regardless of long COVID status) and dropped to 0.80 in older adults and 0.78 in younger adults.However, in the Belgian population the UI increased to 0.91 (95% CI = 0.905; 0.911) after 3 months [33], while at Month 3 in our study the UI increased less prominently to 0.82 in older adults and 0.87 in younger adults and was still significantly lower than pre-COVID.
Our study found that COVID-19 deeply affected patients' work experience during the course of illness and is among the first to provide quantitative measures of work hours lost.The largest work impairments were experienced at the time of infection during Week 1, with lingering long-term effects on presenteeism and work productivity in participants with long COVID and younger adults.Long COVID patients were estimated to miss 28.6 hours of work during Week 1 of infection, corresponding to a decline of 23.9 hours versus pre-COVID.Participants without long COVID were also affected, with 19.3 missed hours of work during Week 1, corresponding to a mean change of 15.4 hours versus pre-COVID.Older and younger adults were similarly impacted at Week 1 with a loss of 20.2 and 18.2 working hours, respectively.Further analyses of Week 1 work hours lost revealed that unvaccinated participants were the most impacted, with the highest number of work hours lost (mean: 28.5).Those boosted with BNT162b2 had the lowest work hours lost (mean: 15.9), solidifying prior evidence of broad benefits of COVID-19 vaccination with BNT162b2 on patient-centric outcomes (Supplemental Table S6) [6].
COVID-19 was associated with large activity impairments during Week 1 across all groups.The activity impairment scores did not return to pre-COVID levels at any point in time among long COVID patients and more slowly for older adults (Month 6) than younger adults (Month 3).
Although our results on work experience and activity levels are not directly comparable to prior survey-based studies due to the difference in study design, our data similarly suggest that COVID-19 can deeply affect paid and unpaid work.O'Mahoney et al. [8] found that a high proportion of long COVID patients reported moderate or severe limitations in their ability to carry out daily activities, with over a third of their sample reporting severe impairments in their ability to work.Similarly, Davis et al. reported that a quarter of the long COVID sample in their study were unable to work due to illness, and almost half required reduced work schedules compared to pre-COVID.Our results complement and supplement these findings by providing different quantitative measures of work productivity losses and activity impairments.Our study showed that a COVID-19 diagnosis had large impacts across age and long COVID status.Long COVID participants were significantly more impacted than those without long COVID, while older and younger adults generally experienced comparable burden.These data suggest that a significant number of people will experience prolonged symptoms that will impact quality of life and functional capacity.The findings provide further evidence that both COVID-19 and long COVID are important public health issues that could affect adults of all ages and may have longer-term effects even after recovery from the acute infection.
To our knowledge, our study is among the first to provide a detailed characterization of the impact of COVID-19 and its evolution throughout the duration of illness according to long COVID status and age [2][3][4][5][6].It is also among the first to report work hours lost and actual hours worked using validated patient-reported outcome measures, which allows further use to inform estimates of indirect costs.
The study is subject to several limitations.As previously described [6], the selfreported nature of the study means that data may be subject to missingness, errors, recall bias, social desirability bias and selection bias associated with attrition.By Month 6, 21% of the 328 participants were lost to follow-up, which could be due to response fatigue and/or survey burden.Our study population only included adults, was over-represented by females and the source population from which subjects were enrolled was relatively healthy.The WPAI analyses had a smaller eligible population and analyses were impacted by relatively small sample sizes, especially for older adults.
We used a long COVID definition based on presence of symptoms and did not assess severity of symptoms.A diverse set of descriptors exists based on number, type and duration of long COVID symptoms and conditions [18], and a universal definition of long COVID has not been established yet.While we conducted a sensitivity analysis using an alternative long COVID definition, our long COVID results may not be fully comparable with existing research.Despite adjusting for several covariates in the model, there is still a risk of residual confounding.These findings may not be generalizable to populations that were excluded from the study or to prior or future variants, other countries or time periods.We did not analyze outcomes based on different SARS-CoV-2 variants.
This study contributes to knowledge gaps related to patient-reported outcomes of COVID-19.Characterization of acute infection and long COVID continues to evolve, and future studies could corroborate these findings with different data collection methods and designs.Moreover, our analyses focused on two variables that have been shown to affect the evolution of COVID-19 patient outcomes (long COVID status and age).Stratification analyses by vaccination status were previously reported for this cohort [6] and further solidified the growing evidence that COVID-19 vaccines could alleviate the detrimental effects of COVID-19 analyzed in this study.Future research could assess these outcomes in patient cohorts defined by additional clinical risk factors, socio-demographic characteristics and treatment history.

Conclusions
This study characterized the differential impacts of COVID-19 infection on patientreported symptoms, HRQoL and WPAI by age and long COVID status.All the patient groups analyzed experienced significant symptoms burden and declines in HRQoL and WPAI scores during the course of their infection.Participants with long COVID experienced a significantly greater level of burden than those with a history of COVID-19 infection and no on-going long COVID symptoms.However, younger and older adults generally reported comparable symptoms burden and drops in HRQoL and WPAI scores.During

Table 1 .
Patient Characteristics and acute symptoms experienced by patients with long COVID vs. those without long COVID, at long COVID start (Week 4).
cImmunocompromised conditions include compromised immune system (such as from immunocompromising drugs, solid organ or blood stem cell transplant, HIV or other conditions), conditions that result in a weakened immune system, including cancer treatment, and kidney failure or end stage renal disease; d COVID-19 test nasal swab day.

Table 2 .
Summary of Post-COVID symptoms in patients with long COVID vs. those without long COVID, at long COVID start (Week 4), at Month 3 and at Month 6.

Table 3 .
Least-Square Estimates of HRQoL and WPAI for patients with long COVID and those without long COVID a .

Table 3 .
Cont.Multivariate models include variables for number of symptoms ≥ 3 (yes, no), time and interaction of number of symptoms ≥ 3 by time, vaccination status and interaction of time by vaccination status, as well as covariates of participant pre-COVID-19 symptom onset score, sociodemographic characteristics (age, sex, regions, social vulnerability, race/ethnicity, high-risk occupations), previously tested positive for COVID-19, severity of acute illness (number of symptoms reported on index date) and immunocompromised status.b Long COVID was defined as ≥3 symptoms reported.c At pre-COVID, scores were summarized for observed measures with mean (SD).d Measures on Day 3 and Week 1 were summarized based on number of post-COVID symptoms at Week 4.

Table 4 .
Patient Characteristics and acute symptoms experienced by subjects aged 50 years or older or younger than 50, at long COVID start (Week 4).

Table 4 .
Cont. : standard deviation; a The Social Vulnerability Index uses 16 US census variables to help local officials identify communities that may need support before, during or after disasters; b Immunocompromised conditions include compromised immune system (such as from immunocompromising drugs, solid organ or blood stem cell transplant, HIV or other conditions), conditions that result in a weakened immune system, including cancer treatment, and kidney failure or end stage renal disease; c COVID-19 test nasal swab day. SD

Table 5 .
Summary of Post-COVID symptoms in patients aged 50 years or older or younger than 50, at long COVID start (Week 4), at Month 3 and at Month 6, n(%).
SD: standard deviation; a P values of t-test for number of symptoms, Chi-square tests or Fisher's exact tests when any one cell has an expected frequency less than 5 for individual symptoms and number of symptom category comparing age ≥ 50 and age < 50 years.

Table 6 .
Least-Square Estimates of HRQoL and WPAI for patients aged 50 years or older and younger than 50 a .

Table 6 .
Cont.Multivariate models include variables for number of symptoms ≥ 3, time and interaction of number of symptom ≥ 3 by time, vaccination status and interaction of time by vaccination status, as well as covariates of participant pre-COVID-19 symptom onset score, sociodemographic characteristics (age, sex, regions, social vulnerability, race/ethnicity, high-risk occupations), previously tested positive for COVID-19, severity of acute illness (number of symptoms reported on index date) and immunocompromised status.b Pre-COVID, scores were summarized for observed measures with mean (SD).c Measures on Day 3 and Week 1 were summarized based on number of post-COVID symptoms at Week 4. a