Persistent symptoms and clinical findings in adults with post-acute sequelae of COVID-19/post-COVID-19 syndrome in the second year after acute infection: A population-based, nested case-control study

Abstract

Background: Self-reported health problems following severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection are common and often include relatively non-specific complaints such as fatigue, exertional dyspnoea, concentration or memory disturbance and sleep problems. The long-term prognosis of such post-acute sequelae of COVID-19/post-COVID-19 syndrome (PCS) is unknown, and data finding and correlating organ dysfunction and pathology with self-reported symptoms in patients with non-recovery from PCS is scarce. We wanted to describe clinical characteristics and diagnostic findings among patients with PCS persisting for >1 year and assessed risk factors for PCS persistence versus improvement.

Methods and findings: This nested population-based case-control study included subjects with PCS aged 18-65 years with (n = 982) and age- and sex-matched control subjects without PCS (n = 576) according to an earlier population-based questionnaire study (6-12 months after acute infection, phase 1) consenting to provide follow-up information and to undergo comprehensive outpatient assessment, including neurocognitive, cardiopulmonary exercise, and laboratory testing in four university health centres in southwestern Germany (phase 2, another 8.5 months [median, range 3-14 months] after phase 1). The mean age of the participants was 48 years, and 65% were female. At phase 2, 67.6% of the patients with PCS at phase 1 developed persistent PCS, whereas 78.5% of the recovered participants remained free of health problems related to PCS. Improvement among patients with earlier PCS was associated with mild acute index infection, previous full-time employment, educational status, and no specialist consultation and not attending a rehabilitation programme. The development of new symptoms related to PCS among participants initially recovered was associated with an intercurrent secondary SARS-CoV-2 infection and educational status. Patients with persistent PCS were less frequently never smokers (61.2% versus 75.7%), more often obese (30.2% versus 12.4%) with higher mean values for body mass index (BMI) and body fat, and had lower educational status (university entrance qualification 38.7% versus 61.5%) than participants with continued recovery. Fatigue/exhaustion, neurocognitive disturbance, chest symptoms/breathlessness and anxiety/depression/sleep problems remained the predominant symptom clusters. Exercise intolerance with post-exertional malaise (PEM) for >14 h and symptoms compatible with myalgic encephalomyelitis/chronic fatigue syndrome were reported by 35.6% and 11.6% of participants with persistent PCS patients, respectively. In analyses adjusted for sex-age class combinations, study centre and university entrance qualification, significant differences between participants with persistent PCS versus those with continued recovery were observed for performance in three different neurocognitive tests, scores for perceived stress, subjective cognitive disturbances, dysautonomia, depression and anxiety, sleep quality, fatigue and quality of life. In persistent PCS, handgrip strength (40.2 [95% confidence interval (CI) [39.4, 41.1]] versus 42.5 [95% CI [41.5, 43.6]] kg), maximal oxygen consumption (27.9 [95% CI [27.3, 28.4]] versus 31.0 [95% CI [30.3, 31.6]] ml/min/kg body weight) and ventilatory efficiency (minute ventilation/carbon dioxide production slope, 28.8 [95% CI [28.3, 29.2]] versus 27.1 [95% CI [26.6, 27.7]]) were significantly reduced relative to the control group of participants with continued recovery after adjustment for sex-age class combinations, study centre, education, BMI, smoking status and use of beta blocking agents. There were no differences in measures of systolic and diastolic cardiac function at rest, in the level of N-terminal brain natriuretic peptide blood levels or other laboratory measurements (including complement activity, markers of Epstein-Barr virus [EBV] reactivation, inflammatory and coagulation markers, serum levels of cortisol, adrenocorticotropic hormone and dehydroepiandrosterone sulfate). Screening for viral persistence (PCR in stool samples and SARS-CoV-2 spike antigen levels in plasma) in a subgroup of the patients with persistent PCS was negative. Sensitivity analyses (pre-existing illness/comorbidity, obesity, medical care of the index acute infection) revealed similar findings. Patients with persistent PCS and PEM reported more pain symptoms and had worse results in almost all tests. A limitation was that we had no objective information on exercise capacity and cognition before acute infection. In addition, we did not include patients unable to attend the outpatient clinic for whatever reason including severe illness, immobility or social deprivation or exclusion.

Conclusions: In this study, we observed that the majority of working age patients with PCS did not recover in the second year of their illness. Patterns of reported symptoms remained essentially similar, non-specific and dominated by fatigue, exercise intolerance and cognitive complaints. Despite objective signs of cognitive deficits and reduced exercise capacity, there was no major pathology in laboratory investigations, and our findings do not support viral persistence, EBV reactivation, adrenal insufficiency or increased complement turnover as pathophysiologically relevant for persistent PCS. A history of PEM was associated with more severe symptoms and more objective signs of disease and might help stratify cases for disease severity.

Fig 1. Flow-chart covering EPILOC study phases 1 and 2.

NB * before (at phase 1) and **after clinical examination (in phase 2). The time from PCR-confirmed acute SARS-CoV-2 infection to phase 1 was 8.7 months (median), the time from phase 1 participation until clinical examination in phase 2 was 8.5 months (median), and the median time between acute infection and phase 2 was 17.2 months, ranging from 9.2 to 24.4 months.

Fig 2. Change in case-control status of study participants (N = 1,558) between initial questionnaire survey (phase 1) and clinical examination (phase 2).

The time from phase 1 participation until clinical examination in phase 2 was 8.5 months (median). Factors associated with improvement of patients with PCS in phase 1 and with worsening among recovered participants in phase 1 were assessed for significance after calculation of ORs with mutual adjustment for the following variables: sex, age, university entrance qualification, marital status, medical treatment of acute infection, obesity (BMI ≥  30 kg/m²), full-time employment (phase 1), time between phases 1 and 2 (per month), secondary SARS-CoV-2 infection since phase 1, two or more vaccine doses, (any) specialist consultation in the last 6 months, participation in a post-COVID-rehabilitation program (see S2 Table).

Fig 3. Means (geometric mean for COMPASS-31 and TMT-B) of self-reported health outcomes and neurocognitive tests (with 95% CI) by case-control status at clinical examination in phase 2, adjusted for sex-age class combinations, study centre and university entrance qualification.

The reported area under the curve (AUC) for persistent PCS vs. continued recovery by the respective instrument also includes sex-age class combinations and university entrance qualification. The AUC for sex-age class combinations, study centre and university entrance qualification alone was 0.64. For comparability, the x-axis is scaled from mean −1 SD to mean + 1 SD for all panels. Abbreviations: PSQI, Pittsburgh Sleep Quality Index; CFQ-11, Chalder Fatigue Scale; SF-12, Short Form-12 Health Survey; PHQ-9, Patient Health Questionnaire 9; GAD-7, Generalised Anxiety Disorder 7; PSS-10, Perceived Stress Scale; Flei, “Fragebogen zur geistigen Leistungsfähigkeit” (subjective mental performance questionnaire); COMPASS-31, Composite Autonomic Symptom Score 31; MoCA, Montreal cognitive assessment scale (points); SDMT, Symbol Digit Modalities Test (number of correct symbols); TMT-B, Trail making test part B (time in seconds).

Fig 4. Cardiopulmonary function indicators and grip strength (means with 95% CI) by case-control status at the clinical examination in phase 2, adjusted for sex-age class combinations, study centre, university entrance qualification, BMI, smoking status and use of beta blocking agents.

Cardiopulmonary exercise testing could be completed in 1,331 participants (87.2% of participants with continued recovery; 83.7% of patients with persistent PCS). For comparability, the x-axis is scaled from mean −1 SD to mean + 1 SD for all panels. Abbreviations: pro-BNP, N-terminal brain natriuretic peptide; LV-EF, left ventricular volume and ejection fraction; LV-E/eʹ, ratio between early mitral inflow and mitral annular early diastolic velocities; LV-E/A, ratio of maximal early to late diastolic transmitral flow velocity; FEV1, forced expiratory volume in one second; FVC, forced vital capacity; SpO₂, peripheral oxygen saturation; HR, heart rate; VO_2max, oxygen uptake; BR, breathing reserve; RER, respiratory exchange ratio; VE/VCO₂ slope, slope of minute ventilation to carbon dioxide production.