A Virus Most Adults Already Carry Awakens With COVID and Tracks With Long COVID

A family of latent human viruses that most adults already carry, called Anelloviridae, flares when COVID-19 strikes. A study of 1,154 hospitalized patients at 20 U.S. hospitals, the largest multiomic COVID-19 study to date, found that reactivation tracks most strongly with long-term physical disability and a long COVID diagnosis. The link is an association, not a cause, but it gives researchers a measurable, blood-detectable signal they have been missing.

The finding comes from a Nature paper by Boston Children's Hospital's Precision Vaccines Program and 15 collaborating U.S. institutions. Within the first 40 days of admission, the study detected 11 reactivated viruses. The most frequently observed were the familiar reactivation suspects: Epstein-Barr, herpes simplex 1, and cytomegalovirus, all herpesviruses that scientists have long known can resurface under immune stress. The standout for long-term outcomes was Anelloviridae, a family of small, circular DNA viruses latent in roughly 90% of people and so poorly understood that even its basic biology remains a research question.

The work grew out of IMPACC, the Immunophenotyping Assessment in a COVID-19 Cohort, a multiomic effort designed to map the immune and molecular signals that separate mild disease, severe disease, and recovery. Researchers tracked more than 200,000 samples and over a billion data points for more than a year, according to Joann Diray Arce, PhD, who leads the data management and analysis core and the IMPACC Clinical and Data Coordinating Center. She called the resulting data set "the largest and most comprehensive biomarker study of COVID-19."

The cohort design gives the Anelloviridae signal its weight. Patients were enrolled at hospital admission and sampled repeatedly over the following year, generating a longitudinal record that links viral reactivation to specific immune, inflammatory, and clinical trajectories. That structure lets the researchers ask not just whether a virus showed up, but when it showed up and what happened next. The Anelloviridae association with long-term physical disability emerged from that temporal resolution, and it is the kind of result a smaller or cross-sectional study would not have caught.

The Anelloviridae signal appeared later, in patients whose long COVID symptoms persisted. The reactivation could point the way to better understanding and ultimately better diagnostics and treatments, according to Ofer Levy, MD, PhD, the Precision Vaccines Program's director. Long COVID has had many candidate biomarkers, but few that show up this consistently across a cohort this large, persist beyond the acute phase, and correlate with a measurable outcome like physical disability.

Independent coverage of the Nature paper has reinforced the headline finding. Science magazine, MedicalXpress, and the University of Texas at Austin Dell Medical School, a co-author institution, have all reported the same reactivation pattern. The result lands in a research community that has been searching for a molecular foothold in long COVID for years.

Anelloviridae was first described in the 1990s and is now understood to infect most humans, often starting in childhood, without causing obvious disease. The family has no approved antivirals, no widely used clinical tests, and no clear connection to any specific illness. That makes the long COVID association striking for what it is not: a confirmation of an existing mechanism. It is closer to a clue, the kind of signal that lets researchers ask better questions about what long COVID actually does to the body.

The gap between association and mechanism is wide. Anelloviridae reactivation could be a marker of immune disturbance rather than a driver of symptoms, and the result is a lead to investigate, not a target to treat. Still, a measurable signal in the blood is the kind of foothold researchers have been missing. The IMPACC data set is now positioned to answer whether that foothold points to a treatment, a diagnostic, or simply a better way to group patients for clinical trials. The public data release that accompanies the paper will let other labs test the same idea in their own cohorts.