Durable Interferon-Linked Blood Signatures During COVID-19 Convalescence

Thrombotic risk in COVID-19 extends beyond acute illness, including in nonhospitalized individuals, suggesting that thrombo-inflammatory biology may persist during recovery. Longitudinal sampling with pre-infection baselines is needed to distinguish infection-associated signals from inter-individual variability and time-related drift in blood gene expression. To define coordinated, time-resolved whole-blood transcriptomic programs during SARS-CoV-2 convalescence using longitudinal within-person comparisons before and after infection, alongside COVID-19-naïve controls. Adults were sampled longitudinally with specimens collected before SARS-CoV-2 infection (T0) and at ~3 months (T1) and ~6 months (T2) post-infection; COVID-19-naïve controls were sampled across matched timepoints. COVID-19-naïve status was confirmed using a multiplex anti-nucleocapsid IgG assay spanning multiple variant antigens. Whole-blood RNA-seq (PAXgene) was analyzed with DESeq. 2 using models that incorporated time point and participant. To prioritize infection-linked changes, genes differentially expressed over time in controls were identified and excluded from comparisons between COVID-19 time points. Pathways were assessed by pre-ranked GSEA (MSigDB) and clustered with aPEAR. After excluding control-associated genes, COVID-19 convalescence remained associated with marked transcriptional remodeling (T1 vs. T0: 782 genes; T2 vs. T0: 655 genes; p < 0.05) and persistent pathway-level changes. Interferon-α and interferon-γ signatures were enriched at both T1 and T2, and a core set of interferon-associated genes remained elevated across convalescence. SARS-CoV-2 infection is followed by durable whole-blood immune-defense programs, including persistent interferon signaling, detectable up to 6 months post-infection. These sustained signatures support a model of prolonged post-infectious immune activation that may contribute to extended thromboinflammatory risk.