The study results are a mix of good and bad news, said Paul Harrison, professor of psychiatry at the University of Oxford and senior author of the study. Among the reassuring aspects was the rapid resolution of symptoms such as depression and anxiety.
“I was surprised and relieved at how quickly the psychiatric effects subsided,” Harrison said.
David Putrino, director of rehabilitation innovation at Mount Sinai Health System in New York, who has been studying the long-term effects of the coronavirus since the start of the pandemic, said the study reveals some very troubling results.
“This allows us to see without a doubt the occurrence of significant neuropsychiatric sequelae in individuals who have had covid and much more often than those who have not,” he said.
Because it only focused on the neurological and psychiatric effects of the coronavirus, the study’s authors and others emphasized that this was not a rigorous long-term study of Covid.
How long covid can change the way we think about disability
“It would be excessive and unscientific to make the immediate assumption that all in [study] cohort had a long covid,” Putrino said. But the study, he said, “informs long-term Covid research.”
Between 7 million and 23 million people in the United States, according to recent government estimates, have persistent Covid, a catch-all term for a wide range of symptoms, including fatigue, shortness of breath and anxiety, that persist for weeks and months after the acute infection subsides. These numbers are expected to rise as the coronavirus becomes established as an endemic disease.
The study was led by Maxim Take, a senior research fellow at the University of Oxford who specializes in using big data to shed light on mental disorders.
The researchers matched almost 1.3 million patients diagnosed with covid-19 between January 20, 2020, and April 13, 2022, with an equal number of patients who had other respiratory illnesses during the pandemic. The data provided by the TriNetX electronic health record network comes primarily from the United States, but also includes data from Australia, Great Britain, Spain, Bulgaria, India, Malaysia and Taiwan.
The study group, which included 185,000 children and 242,000 older adults, found that risks differed by age, with people 65 and older at greatest risk of lasting neuropsychiatric effects.
For people between the ages of 18 and 64, the particularly significant increased risk was of persistent brain fog, affecting 6.4 percent of people who had Covid, compared with 5.5 percent in the control group.
Six months after infection, the children were not found to be at increased risk of mood disorders, although they remained at higher risk of brain fog, insomnia, stroke and epilepsy. None of these effects are permanent for children. In epilepsy, which is extremely rare, the increased risk is greater.
The study found that 4.5 percent of the elderly developed dementia in the two years after infection, compared with 3.3 percent of the control group. That 1.2-point increase in a diagnosis as damaging as dementia is particularly worrisome, the researchers said.
The study’s reliance on a set of de-identified electronic health data raised some cautions, especially given the tumultuous timing of the pandemic. Tracking long-term outcomes can be difficult when patients may have sought care through many different health systems, including some outside the TriNetX network.
“I personally find it impossible to judge the validity of the data or the conclusions when the source of the data is shrouded in mystery and the sources of the data are kept secret by legal agreement,” said Harlan Krumholz, a Yale scholar who has developed an online platform where patients can enter their own health data.
Take said the researchers used several ways to evaluate the data, including making sure it reflected what was already known about the pandemic, such as the drop in death rates during the omicron wave.
Also, Take said, “the validity of the data will not be better than the validity of the diagnosis. If clinicians make mistakes, we will make the same mistakes.
The study follows earlier research by the same group, which reported last year that a third of Covid patients had mood disorders, strokes or dementia six months after infection.
Although they caution that it is impossible to make full comparisons between the effects of recent variants, including omicron and its subvariants that are currently causing infections, and those that were prevalent a year or more ago, the researchers outlined some initial findings: Although omicron caused less severe immediate symptoms, long-term neurological and psychiatric outcomes appeared similar to delta waves, indicating that the burden on global health systems may persist even with less severe variants.
Hannah Davies, co-founder of the Patient-Led Research Collaborative, which studies prolonged covid, said the finding was significant. “This goes against the narrative that omicron is less severe for long-term covid, which is not based on science,” Davis said.
“We see this all the time,” Putrino said. “The general conversation continues to miss the long covid. The severity of the initial infection doesn’t matter when we’re talking about long-term consequences that ruin people’s lives.
Dan Keating contributed to this report.