Skip to main content

Study shows use of SSRIs may protect against long COVID

Nearly five years after COVID-19 first appeared, the post-acute chronic condition known as long COVID remains poorly understood. Symptoms cross a range of biological domains, causing fatigue, heart disease, brain fog, and other debilitating consequences and can generally last three months or more, according to the CDC.

But in October, researchers from UC Berkeley School of Public Health published a study in BMC Medicine that added to the current understanding of the condition: The use of selective serotonin reuptake inhibitors (SSRIs)—a type of antidepressant medication—during acute COVID-19 may protect against long COVID.

Their findings support recent studies that have suggested that lingering SARS-CoV-2 virus particles in the gut may impede serotonin production—and that low serotonin may drive many long COVID symptoms. If this proves to be accurate, SSRIs—which increase synaptic serotonin availability—may be used to prevent or treat long COVID.

“Patients who were using SSRIs at the time they had COVID had about an 8% reduced risk of long COVID,” said Zachary Butzin-Dozier, a postdoctoral scholar in the Division of Biostatistics at UC Berkeley School of Public Health, who led the project. “It’s not a silver bullet to prevent long COVID, but it merits further investigation and provides modest support for Dr. Andrea Wong’s Cell study that serotonin may be a key mechanistic marker of long COVID.”

Butzin-Dozier and his fellow researchers analyzed electronic health records for more than 300,000 American patients who were diagnosed with a depressive condition before contracting COVID. One-third of those patients were using SSRIs.

Butzin-Dozier believes long COVID cases are underreported, and that depressed patients, in particular, might not have their health complaints taken seriously.

“There’s a huge gap between the number of people diagnosed with long COVID and the number who seem to be experiencing it,” he said. “That raises the red flag of diagnostic bias.”

Butzin-Dozier works with UC Berkeley public health professors Alan Hubbard, Jack Colford, and Mark van der Laan, who co-authored the paper. He also leads a research team that placed third in the NIH Long COVID Computational Challenge for building an ensemble machine learning model that predicted individual risk of long COVID diagnosis.

His team’s future studies will examine the relationship between use of SSRIs and specific long COVID symptoms. For example, SSRIs might be associated with a large reduction in neurocognitive symptoms, but not with respiratory symptoms.


Additional authors: Yunwen Ji, Sarang Deshpande, Jeremy Coyle, Junming Shi, and Andrew Mertens, UC Berkeley School Public Health; Eric Hurwitz, University of North Carolina at Chapel Hill; A. Jerrod Anbzalone, University of Nebraska Medical Center; Rena C. Patel, University of Alabama at Birmingham; and the National COVID Cohort Collaborative (N3C) Consortium.

Funding: This research was financially supported by a global development grant from the Bill & Melinda Gates Foundation to the University of California, Berkeley.