A Yale-led study has found that antibodies from some long COVID patients attacked brain and nerve tissues, adding fresh evidence that autoimmunity may be driving the illness in at least a subset of cases. The findings, published in the journal CELL, point to a possible path toward long Covid treatment strategies built around therapies already used for other autoimmune diseases.
The researchers compared blood samples from people with long COVID, healthy volunteers and people who had recovered from COVID without lingering symptoms. They found many long COVID patients carried autoantibodies aimed at parts of the brain and nervous system, including tissues tied to pain signaling, memory, balance, sensory processing and autonomic nervous system control. That pattern lined up with symptoms reported by patients, including fatigue, brain fog, weakness, dizziness, headaches, burning pain and numbness.
The work matters because it moves the discussion of long COVID from broad speculation to a specific immune mechanism. Since the condition first emerged after the pandemic began in 2020, researchers have been trying to explain why some people never fully recover after SARS-CoV-2 infection. The Yale team, which worked with a Mount Sinai Health System group led by David Putrino, said the data provide strong evidence that the immune system can launch an autoimmune attack against the body’s own cells in at least some patients.
Akiko Iwasaki said the findings could eventually point to treatments already used in other autoimmune diseases if they are validated, but she also cautioned that the study does not account for every case of long COVID. In her view, the result is important without being complete. The illness remains a broad, chronic condition with symptoms that vary widely from person to person, and the new data suggest autoimmunity is one important factor rather than the whole explanation.
That gap is the unresolved part of the story. The study does not yet identify which patients would benefit from autoimmune-targeted treatment, and long COVID is still difficult to diagnose because it is based entirely on symptoms. The CDC says there is no single test to determine whether a patient has long COVID or has recovered from it, which leaves doctors treating fatigue, brain fog and other complaints without a clear biological marker in many cases.
The broader scale of the problem keeps growing in view. A separate analysis reported nearly half a million medical records were studied with AI, estimating that 16% of people who got COVID developed long-term symptoms. Other estimates cited in June 2026 put the number of Americans living with long COVID at 20 million, with another 10 million possibly affected without realizing it. For now, the Yale findings narrow one promising path, but they also show how much of long COVID still needs to be sorted out before treatment can be matched cleanly to cause.
