COVID-19 in non-hospitalised adults caused by either SARS-CoV-2 sub-variants Omicron BA.1, BA.2, BA.4/5 or Delta associates with similar illness duration, symptom severity and viral kinetics, irrespective of vaccination history

Abstract

Background: SARS-CoV-2 variant Omicron rapidly evolved over 2022, causing three waves of infection due to sub-variants BA.1, BA.2 and BA.4/5. We sought to characterise symptoms and viral loads over the course of COVID-19 infection with these sub-variants in otherwise-healthy, vaccinated, non-hospitalised adults, and compared data to infections with the preceding Delta variant of concern (VOC).

Methods: In a prospective, observational cohort study, healthy vaccinated UK adults who reported a positive polymerase chain reaction (PCR) or lateral flow test, self-swabbed on alternate weekdays until day 10. We compared participant-reported symptoms and viral load trajectories between infections caused by VOCs Delta and Omicron (sub-variants BA.1, BA.2 or BA.4/5), and tested for relationships between vaccine dose, symptoms and PCR cycle threshold (Ct) as a proxy for viral load using Chi-squared (χ2) and Wilcoxon tests.

Results: 563 infection episodes were reported among 491 participants. Across infection episodes, there was little variation in symptom burden (4 [IQR 3-5] symptoms) and duration (8 [IQR 6-11] days). Whilst symptom profiles differed among infections caused by Delta compared to Omicron sub-variants, symptom profiles were similar between Omicron sub-variants. Anosmia was reported more frequently in Delta infections after 2 doses compared with Omicron sub-variant infections after 3 doses, for example: 42% (25/60) of participants with Delta infection compared to 9% (6/67) with Omicron BA.4/5 (χ2 P < 0.001; OR 7.3 [95% CI 2.7-19.4]). Fever was less common with Delta (20/60 participants; 33%) than Omicron BA.4/5 (39/67; 58%; χ2 P = 0.008; OR 0.4 [CI 0.2-0.7]). Amongst infections with an Omicron sub-variants, symptoms of coryza, fatigue, cough and myalgia predominated. Viral load trajectories and peaks did not differ between Delta, and Omicron, irrespective of symptom severity (including asymptomatic participants), VOC or vaccination status. PCR Ct values were negatively associated with time since vaccination in participants infected with BA.1 (β = -0.05 (CI -0.10-0.01); P = 0.031); however, this trend was not observed in BA.2 or BA.4/5 infections.

Conclusion: Our study emphasises both the changing symptom profile of COVID-19 infections in the Omicron era, and ongoing transmission risk of Omicron sub-variants in vaccinated adults.

Trial registration: NCT04750356.

Fig 1. CONSORT diagram.

Fig 2. Symptoms of COVID-19 are an interaction between prevailing variants and vaccinations.

(A) Time since vaccine dose (2 or 3) in days before the start of an infection episode for each variant of concern (VOC). (B) Proportion of participants reporting no, mild, moderate, or severe symptoms during their infection episode. (C) Duration of each infection episode in days stratified by the VOC and number of doses received prior to infection. (D) Number of symptoms reported by participants, stratified by VOC and number of doses received prior to infection (E) Percentage of individuals reporting each symptom is shown as a heatmap. Percentage shown in each tile, with the tiles shaded to reflect that percentage. The denominator used is all infection episodes of the corresponding VOC and number of doses. (F) Heatmap showing negative decimal logarithms of P values from χ2 tests comparing the presence/absence of a symptom between 3d-BA.4/5 (ref, reference) and the indicated infection episodes. Symptoms are ordered as in Fig 2E. Significant comparisons are marked as follows: P < 0.001 with ***; P < 0.01 with ** and P<0.05 with *.

Fig 3. Hierarchical clustering of symptom patterns in infection episodes occurring after dose 2.

All symptomatic episodes from the Legacy cohort clustered by individual (columns) and symptoms (rows) using Jaccard distances. The presence of a symptom is indicated by rose shading and the absence of a symptom by grey shading. Above each individual episode (each column) the colour bar indicates the severity, assigned VOC of that infection episode, and the number of doses of vaccine received before that infection. An individual may be present >1 if they experienced more than one infection episodes. Symptomatic episodes depicted across all VOCs in (A), all infection episodes, (B) Delta, (C) Omicron BA.1 (D) Omicron BA.2, (E) Omicron BA.4/5. Asymptomatic infection episodes are not shown.

Fig 4. Peak viral load from symptomatic infection episodes in triple-vaccinated participants, compared to days since vaccination.

(A) Viral load (Ct) trajectories (day 0 = symptom onset), plotted separately for each variant and stratified by the number of preceding vaccinations. Smoothed spline fits are shown. (B) Peak viral load on days 1–4 following symptom onset from Delta, Omicron BA.1, BA.2 and BA.4/5 infection episodes by vaccine dose number. (C) Viral load (Ct) trajectories for symptomatic BA.1 or BA.2 infections with febrile and afebrile infection episodes in dark or light blue respectively. Smoothed spline fits are shown. (D) Peak viral load on days 1–4 in participants with either BA.1 or BA.2 reported by symptom severity grade. (E) Peak Ct value across Delta, BA.1, BA.2 and BA.4/5 plotted against the time in days since last vaccine dose after either two or three vaccine doses. Lines from linear regression are shown with a coefficient (β), 95% confidence interval and P value for significant (P < 0.05) estimates.