Abstract
We did not find convincing evidence of a relationship between baseline hematological parameters and susceptibility to SARS-CoV-2 infection or COVID-19 severity, though this relationship should be re-examined as larger and better-powered genetic analyses of SARS-CoV-2 infection and severe COVID-19 become available.
Introduction
Since the onset of the COVID-19 pandemic, clinical observations have linked hospitalized patients with specific hematological signatures, such as lymphopenia (low lymphocyte counts) and thrombocytopenia (low platelet counts). These markers are often utilized for prognosis. However, it remains unclear whether an individual's natural, pre-infection blood profile influences their susceptibility to SARS-CoV-2 or the risk of severe disease. Distinguishing between pre-existing risk factors and the consequences of acute infection is challenging in observational studies. This research aims to untangle these relationships using baseline data and genetic tools.
Methods
The study utilized a multi-pronged approach to investigate the link between 15 measured blood traits and COVID-19 outcomes:
- Observational Analysis: Testing associations in the UK Biobank (UKB, ~423,000 participants) and the Vanderbilt University Medical Center Synthetic Derivative (VUMC SD, ~1 million participants).
- Genetic Correlation: Assessing overlapping genetic architecture using LD Score Regression.
- Mendelian Randomization (MR): Using genetic variants as instrumental variables to test for causal influences of blood levels on COVID-19 outcomes, bypassing traditional confounding factors.
- Coincident Loci Analysis: Identifying specific genomic regions where signals for blood traits and COVID-19 risk overlap.
Data were sourced from the UKB, VUMC SD, the COVID-19 Host Genetics Initiative (HGI) Round 5, and recent large-scale blood trait GWAS.
Results
The Replication Gap and Observational Findings
A significant finding was the lack of consistency between datasets. For instance, associations observed in the UK Biobank, such as basophil percentage and infection risk, failed to replicate in the VUMC cohort. This discrepancy likely stems from differences in participant recruitment and measurement methods (e.g., single timepoint vs. median values).
Meta-analysis suggested that lower total white blood cell counts were associated with higher odds of infection, though this was primarily driven by the VUMC cohort and may reflect clinical ascertainment bias.
Genetic Overlaps and Causal Analysis
Mendelian randomization revealed no significant causal relationships between baseline hematological indices and COVID-19 infection or severity after adjusting for multiple testing. However, "coincident signals" were identified at specific loci. Notably, variants at the TYK2 locus were associated with both COVID-19 severity and lymphocyte/platelet counts, suggesting pleiotropic roles in both immunity and hematopoiesis.
Nominal associations (p < 0.05) were noted for increased mean platelet volume with higher hospitalization risk, and higher red blood cell counts with lower reported infection risk.
Discussion
The study highlights the challenge of low heritability for COVID-19 outcomes (0.1% to 0.3%) compared to the high heritability of blood traits (up to 31%), which complicates the detection of shared genetic signals. Limitations include potential underpowering for small causal effects, broad phenotype definitions for "infection," and a primary focus on individuals of European ancestry.
Future research should leverage larger, more diverse genetic datasets and more granular clinical phenotypes to further explore these potential connections.
Summary updated