TL;DR
Does a history of the "common cold" protect you from COVID-19? This retrospective study of over 2,700 patients suggests that while seasonal coronaviruses may leave behind cross-reactive T cells, they do not provide clinical protection. Having a documented prior seasonal coronavirus infection neither reduces the risk of catching SARS-CoV-2 nor lessens the severity of the disease.
What Problem Does This Paper Address?
Early in the COVID-19 pandemic, immunologists discovered a curious phenomenon: roughly 20% to 50% of people who had never been exposed to SARS-CoV-2 already possessed "cross-reactive" memory T cells that could recognize parts of the virus. These cells were likely generated during previous infections with seasonal coronaviruses (the "common cold" varieties like OC43, HKU1, 229E, and NL63).
The big question for public health was whether this "pre-existing immunity" actually mattered in the real world. If these T cells provided a head start for the immune system, they might explain why some people have asymptomatic cases while others become critically ill. This paper seeks to determine if a documented history of seasonal coronavirus infection correlates with lower infection rates or milder COVID-19 outcomes.
Methods
The researchers conducted a retrospective cohort study using Electronic Health Records (EHR). They looked for a specific sequence of events:
- Pre-Pandemic Phase: A patient received an RT-PCR test for seasonal coronaviruses between January 1, 2016, and February 1, 2020.
- Pandemic Phase: That same patient received an RT-PCR test for SARS-CoV-2 after February 1, 2020.
By comparing patients who tested positive for a seasonal coronavirus in the past against those who tested negative, the team could isolate the impact of prior infection. They used Chi-Square tests to compare the rates of subsequent SARS-CoV-2 infection and categorized COVID-19 severity into "Mild" (not hospitalized) versus "Moderate/Severe" (hospitalized or intubated).
Datasets
The study leveraged massive EHR datasets from two major academic medical centers:
- Stanford Health Care: Records from 2.9 million patients.
- University of Washington: Records from 5.4 million patients.
From this pool of over 8 million people, they identified 2,768 individuals who met the inclusion criteria (1,419 from California and 1,281 from Washington). This represents a robust, multi-site dataset that reflects real-world clinical testing patterns.
Key Insights and Ideas
- Cross-Reactivity $\neq$ Protection: The most significant takeaway is the decoupling of cellular biology from clinical outcomes. While other studies have proven that cross-reactive T cells exist, this study shows they don't necessarily work to prevent or mitigate this specific disease.
- No "Age Shield": The researchers specifically checked if older patients (≥65), who generally have more lifetime exposure to seasonal viruses, saw any benefit. They found no higher protection in this demographic.
- Consistency Across Regions: Despite different patient populations in Northern California and Washington, the results were remarkably consistent, strengthening the conclusion that seasonal coronavirus history is not a significant factor in COVID-19 susceptibility.
Key Results
- Infection Rates: Among the 302 patients with a history of seasonal coronavirus, 3.3% eventually tested positive for SARS-CoV-2. Among the 2,466 patients with no such history, 3.1% tested positive. This difference was statistically insignificant ($P = 0.86$).
- Disease Severity: For those who did catch COVID-19, the severity was comparable between the two groups. 40.0% of those with a prior seasonal infection had moderate-to-severe cases, compared to 48.1% in the control group ($P = 0.63$).
- Recency Matters Not: Analyzing the data by year to see if a recent seasonal infection (e.g., in 2019) provided better protection than an older one (e.g., in 2016) also yielded no significant differences.
Limitations and Future Work
The authors acknowledge several limitations:
- Small Positive Cohort: While the total dataset was large, only 10 individuals in the "prior seasonal infection" group actually tested positive for SARS-CoV-2. This small sample size for the primary outcome limits the statistical power of the severity analysis.
- Testing Bias: The study relies on RT-PCR records. Many people catch seasonal coronaviruses but never go to a doctor to get tested, meaning the "negative history" group likely contained people who actually had prior infections.
- Inference vs. Measurement: The study infers the presence of cross-reactive T cells based on infection history rather than measuring the T cells directly.
Future Work: The authors suggest that prospective studies—where researchers measure actual T cell levels in individuals before they are exposed to SARS-CoV-2—are necessary to definitively confirm whether specific types of cross-reactive T cells might offer subtle benefits that this EHR-based study couldn't detect.
Summary updated