Long COVID in a prospective cohort of home-isolated patients

Abstract

Long-term complications after coronavirus disease 2019 (COVID-19) are common in hospitalized patients, but the spectrum of symptoms in milder cases needs further investigation. We conducted a long-term follow-up in a prospective cohort study of 312 patients-247 home-isolated and 65 hospitalized-comprising 82% of total cases in Bergen during the first pandemic wave in Norway. At 6 months, 61% (189/312) of all patients had persistent symptoms, which were independently associated with severity of initial illness, increased convalescent antibody titers and pre-existing chronic lung disease. We found that 52% (32/61) of home-isolated young adults, aged 16-30 years, had symptoms at 6 months, including loss of taste and/or smell (28%, 17/61), fatigue (21%, 13/61), dyspnea (13%, 8/61), impaired concentration (13%, 8/61) and memory problems (11%, 7/61). Our findings that young, home-isolated adults with mild COVID-19 are at risk of long-lasting dyspnea and cognitive symptoms highlight the importance of infection control measures, such as vaccination.

Conflict of interest statement

The authors declare no competing interests.

© 2021. The Author(s).

Figures

Fig. 1. Study population flow chart.

Number of participants eligible for inclusion at baseline, blood sampling at 2 months and assessment and fatigue evaluation at 6 months. BMEC, Bergen Municipality Emergency Clinic; LTF, lost to follow-up.

Fig. 2. Six-month follow-up of patients with COVID-19 in Bergen, Norway.

The relationship of severity of initial COVID-19 illness and of age with anti-SARS-CoV-2 spike (a and b) and microneutralizing antibody titers (e and f) at 2 months. The relationship of antibody titers at 2 months with number of persistent symptoms (c and g) and total fatigue score according to the Chalder scale (d and h) at 6-month follow-up. The 13 symptoms used are listed in Table 2. The severity of illness was as follows: 1—asymptomatic (n = 5); 2—home-isolated with symptoms (n = 242); 3—hospitalized without medical needs (n = 0); 4—hospitalized with medical needs (n = 31); 5—hospitalized needing O2 (n = 24); 6—hospitalized needing non-invasive ventilation (n = 4); and 7—hospitalized needing respirator (n = 6). The cohort was divided into 15-year age groups: 0–15 (n = 16), 16–30 (n = 65), 31–45 (n = 69), 46–60 (n = 90) and >60 (n = 72). a, b, e and f show median spike IgG and microneutralizing antibody titers (horizontal line), 25% and 75% quantiles (box), plus 95% confidence intervals (lines) with outliers (dots). c and d are violin plots of number of symptoms (up to 13) or fatigue score (values 0–33) divided into four categories of spike IgG: <150 (n = 21), 150–4999 (n = 96), 5,000–20,000 (n = 103) and >20,000 (n = 92). 5k, 5,000; 20k, 20,000. g and h are violin plots of four categories of microneutralizing titers: <20 (neg n = 42), 20–79 (n = 107), 80–320 (n = 92) and >320 (n = 71). O2, supplemental oxygen.

Extended Data Fig. 1. Correlations between severity of acute illness, antibody titres at 2 months and number of symptoms at 6 months follow-up.

Panel a is a scatter plot showing the correlation between severity of initial illness and the Spike IgG antibody titres (log 10) at 2 months follow-up. Panel b shows the correlation between Spike IgG antibody titres (log 10) at 2 months follow-up and the number of symptoms at 6 months follow-up, stratified by severity of initial illness, that is home-isolated versus hospitalised patients. A regression line calculated by linear model is shown in each graph with standard error represented by the shaded area.