ABSTRACT
Background The frequency of persistent symptoms after coronavirus disease 2019 (COVID-19) in adults varies from 4.5% to 87%. Pulmonary function can also show long-term impairment in adults: 10% of hospitalised adults had reduced spirometry values, and 24% had decreased diffusion capacity. To date, only preliminary evidence is available on persistent respiratory sequelae in children and adolescents, therefore our objective was to examine the long-term effects of COVID-19 on pulmonary function in this age group.
Methods Multiple-breath washout, body plethysmography, and diffusion capacity testing were performed after an average of 2.6 months (range 0.4–6.0) following COVID-19 in 73 children and adolescents (age 5–18 years) with different disease severity. Cases were compared to 45 controls with and without infection within six months prior to assessment after exclusion of severe acute respiratory coronavirus-2 infection (SARS-CoV-2).
Results Of the 19 patients (27.1%) who complained about persistent or newly emerged symptoms since COVID-19, 8 (11.4%) reported respiratory symptoms. Comparing patients with COVID-19 to controls, no significant differences were detected in frequency of abnormal pulmonary function (COVID-19: 12, 16.4%; controls: 12, 27.7%; OR 0.54, 95% CI 0.22–1.34). Only two patients with persistent respiratory symptoms showed abnormal pulmonary function. Multivariate analysis revealed reduced forced vital capacity (p=0.045) in patients with severe infection regardless of SARS-CoV-2 infection.
Discussion Pulmonary function is rarely impaired in children and adolescents after COVID-19, except of those with severe infection. The discrepancy between persistent respiratory symptoms and normal pulmonary function suggests a different underlying pathology such as dysfunctional breathing.
INTRODUCTION
The clinical course of children and adolescents with SARS-CoV-2 infection is mostly less severe than in adults [1].
Long-COVID sums up all symptoms that persist or newly emerge four weeks to beyond twelve weeks after acute SARS-CoV-2 infection [2].
In one study, up to 87% of adults hospitalised due to COVID-19 infection reported persistent symptoms (for example fatigue, dyspnoea, cough, gastrointestinal symptoms) about two months after COVID-19 infection [3]. In another cohort with in- and outpatients in the same age range, only 4.5% were still symptomatic with similar symptoms after eight weeks [1]. There is only preliminary evidence of persistent symptoms in children: Buonsenso et al. described that 52.7% of their study population reported at least one persistent symptom equal to or more than four months after SARS-CoV-2 infection, while 14.7% stated they suffered persistent respiratory problems [4].
Long-lasting deterioration in pulmonary function has been described in adults [5–7]. Lerum et al. reported that about 10% of adults hospitalised with COVID-19 had reduced values for forced vital capacity (FVC) and forced expiratory volume in one second (FEV1), and 24% showed significantly reduced diffusion capacity at a three-month follow-up visit [5]. However, frequency and extent of long-term changes in pulmonary function in children and adolescents following COVID-19 infection are barely examined. Bottino et al performed spirometry and diffusion capacity testing in seven children about two months after recovery from mild SARS-CoV-2 infection and did not find any abnormalities [8].
To evaluate the impact of respiratory sequelae in children is important given the fact that most children worldwide will potentially get infected with SARS-CoV-2 as long as vaccines are predominantly reserved for adults and high-risk groups.
We analysed pulmonary function of children and adolescents after COVID-19 and compared their results to a group of controls after exclusion of SARS-CoV-2 infection.
PATIENTS AND METHODS
Study design and population
We conducted a single-centre, cross-sectional, prospective study to assess pulmonary function after COVID-19 in children and adolescents from 5 to 18 years of age. This study was part of a still ongoing study at the University Children’s Hospital of Bochum, Germany, assessing rates of SARS-CoV-2 seroconversion in children and adolescents in western Germany (CorKid).
We recruited from August 2020 to March 2021 and included subjects with positive SARS-CoV-2 antibodies tested in the CorKid study, as well as in- and outpatients who tested positive for SARS-CoV-2 using polymerase chain reaction (PCR) at the University Children’s Hospital of Bochum or in an outpatient practice in the region. Children with negative antibodies for SARS-CoV-2 and no other evidence of SARS-CoV-2 infection from the CorKid study served as controls. Exclusion criteria were inability to perform lung function testing and missing written consent.
The subjects and/or their guardians answered questionnaires regarding their medical history, especially acute and chronical pulmonary diseases, as well as current and previous medications. Any infections within six months prior to the assessment and their date, duration and symptoms, as well as PCR test results for SARS-CoV-2 were recorded. We also documented persistent symptoms after COVID-19 infection in the COVID-19 group.
According to these data, we subdivided the COVID-19 group into those who were symptomatic and those who were asymptomatic, and showed signs of respiratory tract infections (if cough, rhinitis, sore throat and/or dyspnoea was confirmed) and other types. Severity of infection was subclassified into severe and non-severe. An infection was considered severe if it included dyspnoea, fever >38.5°C for >5 days, bronchitis or pneumoniae, and/or hospitalisation >1 day.
Patients were classified as having chronic lung disease if they had a known medically diagnosed pulmonary disease like bronchial asthma or if they had one of the following: recurrent wheeze, pneumonia in the year before measurement, previous or current use of inhaled steroids for at least 4 weeks or long-lasting productive cough (≥8 weeks).
We also divided the subjects into two groups (follow-up 0–3 and 4–6 months) depending on the interval between infection and assessment.
Pulmonary function testing
Body plethysmography and diffusing capacity testing for carbon monoxide (CO) were performed using MasterScreen Body/Diff (Vyaire, Hoechberg, Germany) according to American Thoracic Society/European Respiratory Society guidelines [9, 10]. We measured FEV1, FVC, mean expiratory flow at 75% (MEF75) and MEF25, as well as diffusing capacity of the lung for carbon monoxide (DLCO) and diffusing capacity divided by the alveolar volume (DLCO/VA) standardised to Z-scores using Global Lung Initiative reference values [11, 12]. DLCO and DLCO/VA were adjusted for haemoglobin levels. N2 multiple-breath washout was measured using Exhalyzer D (EcoMedics AG, Duernten, Switzerland) according to current guidelines [13]. We reported the Lung Clearance Index at 2.5% of starting concentration (LCI2.5%) as recommended.
All pulmonary function tests were done by specially trained staff and assessed by two paediatric pneumologists blinded to the patient’s SARS-CoV-2 status.
Pulmonary function was defined as abnormal, if at least one measured parameter was pathological (LCI2.5% >7.9, FVC Z-score, FEV1 Z-score, MEF75 Z-score, MEF25 Z-score, DLCO Z-score, DLCO/VA Z-score <-1.96) or there was at least one borderline parameter (LCI2.5% >7.0, FVC Z-score, FEV1 Z-score, MEF75 Z-score, MEF25 Z-score, DLCO Z-score, DLCO/VA Z-score <–1 ≥–1.96) for two different pulmonary function tests (N2 multiple-breath washout, body plethysmography, diffusion capacity).
Imaging of the lungs was performed only when clinically necessary.
Statistics
Descriptive statistics: We calculated mean, median, range and standard deviation for continuous variables. Categorical values were expressed as absolute values and percentages.
Inferential statistics: Odds Ratios (ORs) and 95% confidence intervals (95% CIs) were calculated for categorical values. Normal distribution was proven for all pulmonary function parameters either using the Komolgorov-Smirnov test or the Levene test. Independent t-test or analysis of variance (ANOVA) was used for between-group comparisons. Post-hoc analysis was performed using least significant difference. All statistical analysis were performed using SPSS version 27 (SPSS for Windows, SPSS Inc., Chicago, IL, USA). A p-value of 0.05 or less was considered statistically significant.
Ethical approval
The ethics committee of the Ruhr-University, Bochum, Germany, approved the project (register number: 20-6927). All participants and/or their parents were informed about the study and provided written informed consent.
RESULTS
We enrolled 73 subjects with confirmed SARS-CoV-2 infection, 26 (35.6%) of whom were symptomatic during the acute phase. Of the 45 subjects who served as controls, 14 (31.1%) had any symptomatic infection other than COVID-19 within six months before the assessment. Excluding six COVID-19 patients without PCR, the mean interval between infection and date of testing was similar in both groups (COVID-19: 2.59 months (range 0.4–6.0); controls: 3.43 months (range 1.03–6.3)).
The epidemiological data were similar in both groups (see Table 1a).
Disease severity and type of infection did not show significant differences between the two groups. However, 70.4% (n = 19) of children and adolescents with symptomatic SARS-CoV-2 infection did not have respiratory symptoms, whereas 92.9% (n = 13) of infections in the control group involved the respiratory tract. Rhinitis and cough were significant more often in the control group than in the COVID-19 group (rhinitis: OR 0.14, 95% CI 0.03–0.62; cough: OR 0.24, 95% CI 0.06–0.95). Other assessed symptoms did not show significance between the two groups (see Table 1b for more information).
Long-term complaints
Nineteen (27.1%, data were only available for 70 subjects) children and adolescents of the COVID-19 group reported persistent or newly emerged symptoms after SARS-CoV-2 infection. Eight (11.4%) reported at least one respiratory symptom, six (8.6%) of whom suffered ongoing breathing problems and two (2.9%) persistent cough. The mean age of children and adolescents with respiratory long-term symptoms was similar to all patients with SARS-CoV-2 infection (all COVID-19: 10.82 ± 3.25, patients with respiratory long-term symptoms: 11.75 ± 3.92 [mean ± standard deviation in years]). In two patients, these respiratory problems newly emerged after asymptomatic SARS-CoV-2 infection. Ten (14.3%) suffered from fatigue; five patients (7.1%) had both persistent respiratory symptoms and fatigue syndrome.
Pulmonary function testing
Comparing the two groups of children and adolescents with and without SARS-CoV-2 infection, there were no significant differences in LCI2.5% (absolute) and age-related Z-scores of FVC, FEV1, MEF75, MEF25, DLCO and DLCO/VA. Details are shown in Table 2 and Figure 1.
Depiction of pulmonary function parameters for COVID-19 and controls. Boxplots show medians, quartiles, minimum and maximum values. The dots represent the individual values of participants. Participants with severe infection within the last six months are marked as red squares. LCI2.5%, Lung Clearance Index at 2.5% of starting concentration; FVC, Forced Vital Capacity; FEV1, Forced Expiratory Volume in the first Second; DLCO, Diffusion Capacity of the lungs for Carbon Monoxide
Subjects suffering from severe infection within six months prior to the assessment had significant reduction in FVC (p = 0.045) and MEF75 (p = 0.002) compared to those with non-severe infection and asymptomatic infection. Neither site of infection nor COVID-19 status influenced pulmonary function (see Figure 2). MEF25, DLCO, DLCO/VA and LCI2.5% did not show any significant differences between children and adolescents with severe, non-severe or asymptomatic infection six months prior to the assessment.
Depiction of A FVC Z-score and B MEF75 Z-score of all participants (COVID-19 and controls) divided into severe, non-severe and asymptomatic infection, respectively no infection in the control group. Boxplots show medians, quartiles, minimum and maximum values. The dots represent the individual values of participants.
FVC, Forced Vital Capacity; MEF75, Mean Expiratory Flow at 75%
Overall, 16.4% (n = 12) of the children and adolescents after COVID-19 infection had abnormal pulmonary function compared to 27.7% (n = 12) without SARS-CoV-2 infection (OR 0.54, 95% CI 0.22–1.34). Comparing all subjects with and without abnormal pulmonary function, no influence of SARS-CoV-2 status, previous or current pulmonary disease, symptomatic infection or severity of infection was found. Most abnormal pulmonary functions occurred in the 5–8-years age group (see Table 3). This effect was mainly visible in the measurements of LCI2.5% and FVC (see Figure 3). Two of the eight patients reporting persistent respiratory complaints, and three of the ten suffering fatigue syndrome showed abnormal pulmonary function.
Scatter plot with A LCI2.5% (absolute) and B FVC Z-score shown on the x-axis and age in years depicted on the y-axis. Patients after COVID-19 infection are depicted as blue dots, controls are marked as red triangles.
A The dotted lines mark LCI2.5% of 7.0 and 7.9. Values >7.0 are borderline. Values >7.9 are pathological.
B The dotted lines mark FVC Z-score of –1.96 and –1. Values <–1 are borderline. Values <–1.96 are pathological.
LCI2.5%, Lung Clearance Index at 2.5% of starting concentration; FVC, Forced Vital Capacity
Only one child with pneumonia underwent pulmonary imaging during acute COVID-19 infection. No other child or adolescent required imaging of the lungs based on medical assessment.
DISCUSSION
Pulmonary involvement, including severe interstitial pneumonia and acute respiratory distress syndrome (ARDS), in SARS-CoV-2 infections has been the cause of high mortality and long-lasting morbidity in adults [7, 14]. In children and adolescents, severe pulmonary complications are rare [15], even though 11.1% of our patients with symptomatic SARS-CoV-2 infection described dyspnoea with acute COVID-19.
Most SARS-CoV-2 infections in our cohort, however, were asymptomatic during the acute phase. Only 35.6% of patients in our COVID-19 group stated to having had any symptoms during acute infection; respiratory symptoms were reported in 26%. In an Italian cohort with children and adolescents (n = 16), 87% showed symptoms at the time of diagnosis of SARS-CoV-2, 19% presenting with respiratory symptoms [8].
Long-term pulmonary symptoms persistent or emerging more than four weeks after the acute infection such as cough, dyspnoea or hyperreactive airways were described by 11.4% of our patients. This is similar to the findings of an Italian cohort study, in which 14.7% of the children had persistent respiratory problems for several months after COVID-19 infection [4]. In adults, frequency of persistent dyspnoea three months after discharge from hospital varies from 9% [16] to 54% [5]. Persistent pulmonary problems after viral infections are a well-known phenomenon in children: for example, children with respiratory syncytial virus (RSV) bronchiolitis show more wheezing episodes in the following 12 months [17] and even up to the age of six years, children with lower respiratory tract RSV infection in early childhood suffer more frequent wheeze (up to 4.3 times) than children without lower respiratory tract infection [18].
Pulmonary function in hospitalised adults is frequently impaired on discharge: Mo et al. described that 47% of their patients had reduced DLCO, total lung capacity (TLC) was decreased in 25%, 9.1% had impaired FVC and 13.6% showed reduced FEV1 [6]. But even three to four months after discharge, pulmonary function is still impaired in adults [5, 7, 15, 19, 20]. In a Norwegian cohort, FVC and FEV1 were reduced in about 10% and diffusion capacity in 24% of hospitalised adults three months after discharge [5]. Qin et al. detected significant differences in DLCO levels between discharged patients with severe and non-severe COVID-19 infection at a three-month follow-up [16]. Even after six months, lower DLCO and TLC were detected in patients with critical disease [19]. A Mexican study investigating the correlation between persistent dyspnoea and pulmonary function reported that patients with persistent dyspnoea had significantly lower FVC, FEV1 and DLCO values than those without [21].
Pulmonary function impairment in children after SARS-CoV-2 infection is rare, as we could show in our cohort. LCI2.5% was abnormal in 4.4% of children and adolescents with COVID-19, 7% had reduced FVC and 5.3% showed impaired DLCO. No patient had pathological FEV1 values after SARS-CoV-2 infection. Even in children and adolescents with persistent respiratory symptoms, pulmonary function was normal in 75% of the subjects. Neither LCI2.5%, FEV1, DLCO or DLCO/VA showed any significant differences in patients and controls. These results support the preliminary findings of Bottino et al, in which seven children did not show any abnormalities in spirometry and diffusion capacity about two months after their mild SARS-CoV-2 infection [8]. In our cohort, minimal changes persisted for up to six months only in children and adolescents with more severe infection (SARS-CoV-2 or other infections). Similarly, other studies in children with viral infections, such as RSV or rhinovirus, detected acute and long-term loss of pulmonary function [17, 22, 23]. Other known risk factors for reduced pulmonary capacity such as preterm birth [24] or underlying pulmonary disease had no effects in our cohort.
We did not perform pulmonary imaging in all our patients, but only in the ones with severe pulmonary impairment. One had pneumonia at the time of initial diagnosis consistent with several studies showing radiological changes in children with acute COVID-19 [25, 15]. Fortunately, we did not detect any permanent changes on follow-up imaging.
Limitations of this single-centre design with a rather small sample size include age-dependent inconsistencies in pulmonary function performance, patient-reported symptoms and lack of information on long-term outcome of symptoms in controls. Further, our cohort did not include children and adolescents with critical respiratory involvement during acute COVID-19.
To our knowledge, our study is the first to compare pulmonary function in symptomatic and asymptomatic children and adolescents with and without evidence of SARS-CoV-2 infection; no difference between these two groups was observed. Even most patients with persistent respiratory symptoms did not show impaired lung capacity. Severity of infection proved to be the only predictor for mild pulmonary function changes. To conclude, our findings suggest that children and adolescents after SARS-CoV-2 infection do not suffer from persistent deterioration of respiratory function, including body plethysmography, multiple-breath washout and diffusion capacity testing.
Further studies with a larger, more representative cohort (including patients with critical respiratory involvement) and over a longer period are needed for better understanding of the respiratory long-term impairment after SARS-CoV-2 infection in children and adolescents. The discrepancy between subjective persistent respiratory complaints and normal pulmonary function might be caused by functional respiratory disorders, for example hyperventilation, as already described in adults [26, 27].
Thus, further studies including treadmill testing would be useful and were already initiated in our cohort.
Data Availability
Data are available upon reasonable request.
The RECORD statement – checklist of items, extended from the STROBE statement, that should be reported in observational studies using routinely collected health data.
Acknowledgements
We are grateful to the healthcare team at the Children’s University Hospital Bochum, especially Chantal Matenar whose efforts were fundamental to this work. We would like to thank all the patients and families involved.
Footnotes
↵3 This work is part of the thesis of Leona Knoke.
Funding This study has partially been funded by the Federal Ministry of Education and Research (Bundesministerium für Bildung und Forschung [BMBF], trial registration 01KI20173).
Clinical trial ID DRKS00022434, German clinical trial register
Competing interests None declared
Patient consent for publication Not required.
Ethics approval Ethical approval was provided by the Ethics Committee of the Ruhr-University Bochum, Germany (No. 20-6927).
Checklist for the appropriate study type STROBE and RECORD checklists are included as supplementary appendices.
Data availability statement Data are available upon reasonable request.
REFERENCES
