Long COVID after breakthrough SARS-CoV-2 infection

Abstract

The post-acute sequelae of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection-also referred to as Long COVID-have been described, but whether breakthrough SARS-CoV-2 infection (BTI) in vaccinated people results in post-acute sequelae is not clear. In this study, we used the US Department of Veterans Affairs national healthcare databases to build a cohort of 33,940 individuals with BTI and several controls of people without evidence of SARS-CoV-2 infection, including contemporary (n = 4,983,491), historical (n = 5,785,273) and vaccinated (n = 2,566,369) controls. At 6 months after infection, we show that, beyond the first 30 days of illness, compared to contemporary controls, people with BTI exhibited a higher risk of death (hazard ratio (HR) = 1.75, 95% confidence interval (CI): 1.59, 1.93) and incident post-acute sequelae (HR = 1.50, 95% CI: 1.46, 1.54), including cardiovascular, coagulation and hematologic, gastrointestinal, kidney, mental health, metabolic, musculoskeletal and neurologic disorders. The results were consistent in comparisons versus the historical and vaccinated controls. Compared to people with SARS-CoV-2 infection who were not previously vaccinated (n = 113,474), people with BTI exhibited lower risks of death (HR = 0.66, 95% CI: 0.58, 0.74) and incident post-acute sequelae (HR = 0.85, 95% CI: 0.82, 0.89). Altogether, the findings suggest that vaccination before infection confers only partial protection in the post-acute phase of the disease; hence, reliance on it as a sole mitigation strategy may not optimally reduce long-term health consequences of SARS-CoV-2 infection. The findings emphasize the need for continued optimization of strategies for primary prevention of BTI and will guide development of post-acute care pathways for people with BTI.

Fig. 1. Risk and 6-month excess burden of post-acute sequelae in people with BTI compared to the contemporary control group.

Risk and 6-month excess burden of death, at least one post-acute sequela and post-acute sequelae by organ system are plotted. Incident outcomes were assessed from 30 days after the positive SARS-CoV-2 test to the end of follow-up. Results are in comparison of BTI (n = 33,940) to the contemporary control group that consisted of those with no record of a positive SARS-CoV-2 test (n = 4,983,491). Adjusted HRs (dots) and 95% CIs (error bars) are presented, as are estimated excess burden (bars) and 95% CIs (error bars). Burdens are presented per 1,000 persons at 6 months of follow-up.

Fig. 2. Risk and 6-month excess burden of post-acute sequelae in those with BTI by acute phase care setting.

Risk and 6-month excess burden of death, at least one post-acute sequela and post-acute sequelae by organ system are plotted by care setting of the acute phase of the disease (not hospitalized, hospitalized and admitted to ICU). Incident outcomes were assessed from 30 days after the positive SARS-CoV-2 test to the end of follow-up. Results are in comparison of BTI (non-hospitalized n = 30,273; hospitalized n = 3,667; admitted to ICU n = 811) to the contemporary control group with no record of a positive SARS-CoV-2 test (n = 4,983,491). Adjusted HRs (dots) and 95% CIs (error bars) are presented, as are estimated excess burden (bars) and 95% CIs (error bars). Burdens are presented per 1,000 persons at 6 months of follow-up.

Fig. 3. Risk and 6-month excess burden of post-acute sequelae in people with BTI compared to those with SARS-CoV-2 infection without prior vaccination.

Risk and 6-month excess burden of death, at least one post-acute sequela and post-acute sequelae by organ system are plotted. Incident outcomes were assessed from 30 days after the positive SARS-CoV-2 infection test to the end of follow-up. Results are in comparison of BTI (n = 33,940) to those with SARS-CoV-2 infection without prior vaccination (n = 113,474). Adjusted HRs (dots) and 95% CIs (error bars) are presented, as are estimated excess burden (bars) and 95% CIs (error bars). Burdens are presented per 1,000 persons at 6 months of follow-up.

Fig. 4. Risk and 6-month excess burden of post-acute sequelae in those with BTI compared to those with SARS-CoV-2 infection without prior vaccination by acute phase care setting.

Risk and 6-month excess burden of death, at least one post-acute sequela and post-acute sequelae by organ system are plotted by care setting of the acute phase of the disease (not hospitalized, hospitalized and admitted to ICU). Incident outcomes were assessed from 30 days after the positive SARS-CoV-2 test to the end of follow-up. Results for a given care setting are in comparison of BTI (non-hospitalized n = 30,273; hospitalized n = 3,667; and admitted to ICU n = 811) to the SARS-CoV-2 infection without prior vaccination group (non-hospitalized n = 100,700; hospitalized n = 12,774; and admitted to ICU n = 2,982) with the same care setting during the acute phase of the disease. Adjusted HRs (dots) and 95% CIs (error bars) are presented, as are estimated excess burden (bars) and 95% CIs (error bars). Burdens are presented per 1,000 persons at 6 months of follow-up.

Extended Data Fig. 1. Risk of post-acute sequelae in people with breakthrough SARS-CoV-2 infection compared to the contemporary control group.

Incident outcomes were assessed from 30 days after the positive SARS-CoV-2 test to end of follow-up. Results are in comparison of breakthrough SARS-CoV-2 infection (n = 33,940) to the contemporary control group that consisted of those with no record of a positive SARS-CoV-2 test (n = 4,983,491). Adjusted hazard ratios (dots) and 95% confidence intervals (error bars) are presented. Myocarditis (hazard ratio: 1.99; 95% confidence intervals: 0.47, 8.40) was not plotted to remove excess x-axis scaling.

Extended Data Fig. 2. Risk and 6-month excess burden of post-acute sequelae in people with breakthrough SARS-CoV-2 infection compared to the historical control group.

Risk and 6-month excess burden of death, at least one post-acute sequela, and post-acute sequelae by organ system are plotted. Incident outcomes were assessed from 30 days after the positive SARS-CoV-2 test to end of follow-up. Results are in comparison of breakthrough SARS-CoV-2 infection (n = 33,940) to the historical control group that consisted of those with no record of a positive SARS-CoV-2 test (n = 5,785,273). Adjusted hazard ratios (dots) and 95% confidence intervals (error bars) are presented, as are estimated excess burden (bars) and 95% confidence intervals (error bars). Burdens are presented per 1000 persons at 6 months of follow-up.

Extended Data Fig. 3. Risk and 6-month excess burden post-acute in people with breakthrough SARS-CoV-2 infection compared to the vaccinated control group.

Risk and 6-month excess burden of death, at least one post-acute sequela, and post-acute sequelae by organ system are plotted. Incident outcomes were assessed from 30 days after the positive SARS-CoV-2 test to end of follow-up. Results are in comparison of breakthrough SARS-CoV-2 infection (n = 33,940) to the vaccinated control group that consisted of those with no record of a positive SARS-CoV-2 test (n = 2,566,369). Adjusted hazard ratios (dots) and 95% confidence intervals (error bars) are presented, as are estimated excess burden (bars) and 95% confidence intervals (error bars). Burdens are presented per 1000 persons at 6 months of follow-up.

Extended Data Fig. 4. Risk and 6-month excess burden of post-acute sequelae in people with breakthrough SARS-CoV-2 infection by immunocompromised status.

(A) Breakthrough SARS-CoV-2 infection compared to the contemporary control (n=4,983,491) in non-immunocompromised (n=27,431) and immunocompromised (n=6,509) individuals. (B) A within breakthrough SARS-CoV-2 infection comparison between immunocompromised (n=6,509) and non-immunocompromised (n=27,431) individuals. Incident outcomes were assessed from 30 days after the positive SARS-CoV-2 test to end of follow-up. Adjusted hazard ratios (dots) and 95% confidence intervals (error bars) are presented, as are estimated excess burden (bars) and 95% confidence intervals (error bars). Burdens are presented per 1000 persons at 6 months of follow-up.

Extended Data Fig. 5. Risk of post-acute sequelae in people with breakthrough SARS-CoV-2 infection compared to the contemporary control group by acute phase care setting.

Risks are plotted by care setting of the acute phase of the disease (not hospitalized, hospitalized, and admitted to the intensive care unit). Incident outcomes were assessed from 30 days after the positive SARS-CoV-2 test to end of follow-up. Results are in comparison of breakthrough SARS-CoV-2 infection (non-hospitalized n=30,273; hospitalized n=3,667; admitted to ICU n=811) to the contemporary control group that consisted of those with no record of a positive SARS-CoV-2 test (n=4,983,491). Adjusted hazard ratios (dots) and 95% confidence intervals (error bars) are presented. Outcomes that occurred in less than 10 participants in a group were excluded.

Extended Data Fig. 6. Risk of post-acute sequelae in people with breakthrough SARS-CoV-2 infection compared to those with SARS-CoV-2 infection without prior vaccination.

Incident outcomes were assessed from 30 days after the positive SARS-CoV-2 test to end of follow-up. Results are in comparison of breakthrough SARS-CoV-2 infection (n=33,940) to those with SARS-CoV-2 infection without prior vaccination (n=113,474). Adjusted hazard ratios (dots) and 95% confidence intervals (error bars) are presented. Myocarditis (hazard ratio: 0.05; 95% confidence intervals: 0.01, 0.38) was not plotted to remove excess x-axis scaling.

Extended Data Fig. 7. Risk and 6-month excess burden of post-acute sequelae in those hospitalized during the acute phase with breakthrough SARS-CoV-2 infection compared to seasonal influenza.

Risk and 6-month excess burden of death, at least one post-acute sequela, and post-acute sequelae by organ system are plotted. Incident outcomes were assessed from 30 days after the positive SARS-CoV-2 test to end of follow-up. Results are in comparison of those hospitalized with breakthrough SARS-CoV-2 infection (n = 3,667) to those who were hospitalized with seasonal influenza (n = 14,337). Adjusted hazard ratios (dots) and 95% confidence intervals (error bars) are presented, as are estimated excess burden (bars) and 95% confidence intervals (error bars). Burdens are presented per 1000 persons at 6 months of follow-up.