Researchers in Singapore detected a marked spike in autoantibodies associated with brain inflammation during the first year of the COVID-19 pandemic. The incidence rates later declined to pre-pandemic levels, a trend the scientists attribute to widespread vaccination efforts and the arrival of less severe virus variants. The findings were published in the journal Brain and Behavior.
When a person catches a virus, their immune system generates targeted proteins called antibodies to fight off the invader. After the infection clears, the immune response usually winds down. In rare instances, this system makes an error. The body produces autoantibodies that mistakenly attack the person’s own healthy tissues.
When these errant antibodies target the brain, the resulting condition is known as autoimmune encephalitis. Patients with this disorder experience severe inflammation in their central nervous system. This can lead to a wide array of devastating symptoms, including memory loss, sudden alterations in personality, hallucinations, and seizures.
Medical professionals have recognized for years that certain viral infections can act as a trigger for autoimmune encephalitis. A classic example is the herpes simplex virus, which can prompt the immune system to attack specialized brain receptors long after the initial infection has passed. During the early days of the COVID-19 crisis, doctors began reporting isolated cases of autoantibody attacks in patients who had recently recovered from the novel coronavirus.
To understand if this was a widespread phenomenon, researchers needed to look at population-level laboratory data. Neurologist Tianrong Yeo and his colleagues at the National Neuroscience Institute in Singapore designed a study to evaluate regional testing trends. Their goal was to see if the incidence of specific brain attacking antibodies increased alongside the emergence of the coronavirus.
The researchers also decided to look at a second category of diseases known as paraneoplastic neurological syndromes. These conditions produce similar symptoms and mimic viral induced autoimmune encephalitis. However, they are triggered by an immune response to an underlying cancerous tumor, not an infectious disease. The researchers hypothesized that if a virus was causing a spike in autoimmune brain inflammation, the rates of cancer-linked inflammatory conditions should remain completely unaffected.
To test this idea, the team conducted a retrospective analysis of laboratory tests performed at their institute. The National Neuroscience Institute handles diagnostic samples from medical facilities that serve about three-quarters of the population of Singapore. This provided a massive dataset to examine long-term trends before and during the pandemic.
The researchers gathered data on blood serum and spinal fluid samples tested between 2017 and 2023. They looked for the presence of specific autoantibodies known to target the outer surface of brain cells. They also looked for a different set of autoantibodies known to attack the interior of cells, which are the primary hallmark of the cancer-linked neurological syndromes.
To eliminate false positive results, the laboratory team utilized incredibly strict conditions for inclusion. For certain autoantibodies, a patient’s sample was only counted as positive if the markers were found directly in the cerebrospinal fluid, the distinct liquid that bathes and protects the brain. For others, the sample had to test positive across two entirely separate types of laboratory tests. If a patient provided multiple samples over time, only their first positive result was recorded to prevent duplicate counting.
The team then calculated the incidence rate using a metric known as person-years. This is a statistical measurement that accounts for both the number of people observed and the amount of time they were monitored. It allows scientists to accurately compare how often a disease occurs across different time periods and shifting population sizes.
Between 2017 and 2023, the laboratory tested over four thousand individual samples for autoantibodies targeting brain surface proteins. Out of this large pool, 87 samples met the strict threshold for a positive result. The researchers noticed a distinct pattern when charting these positive results chronologically.
During the three years preceding the pandemic, the incidence of these autoantibodies hovered at around 2.44 cases per one million person-years. In 2020, the first year of the global viral outbreak, the rate shot up to 4.92 cases per one million person-years. The most common type of autoantibody found during this spike targeted a specific protein structure in the brain called the NMDA receptor, which handles memory and learning functions.
The research team views this sudden doubling as evidence of a biological link between the novel coronavirus and autoimmune brain inflammation. They propose two primary mechanisms that might explain the connection. The first is molecular mimicry, a scenario where the proteins on the outside of the coronavirus look so much like the proteins on the outside of human brain cells that the immune system gets fundamentally confused.
The second proposed mechanism relates to widespread internal inflammation. A severe coronavirus infection can cause a massive release of inflammatory molecules throughout the entire body. These molecules might damage the protective boundary between the bloodstream and the brain, allowing rogue immune cells to slip inside and wreak havoc.
Following the initial spike in 2020, the incidence of these autoantibodies dropped. Between 2021 and 2023, the rate hovered at roughly 2.74 cases per one million person-years. The researchers noted that while this drop was visually apparent on their charts, the difference between the 2020 peak and the later pandemic years was not statistically significant.
The researchers suspect the successful rollout of vaccination programs played a major role in bringing the numbers back down. Vaccines train the immune system in a safer environment and limit the severity of future infections. By late 2021, the vast majority of the population in Singapore had received at least two vaccine doses. With lower viral loads in the community, the chances of the virus triggering an extreme autoimmune response likely plummeted.
The natural evolution of the virus might have also contributed to the decline. As the pandemic progressed, the original strain of the coronavirus was replaced by variants that spread faster but generally caused milder illness. These newer iterations of the virus may have possessed a lower capacity to provoke an autoimmune malfunction.
Meanwhile, the data for the cancer-linked neurological syndromes behaved exactly as the researchers predicted. The team found 29 positive samples out of more than three thousand tests. Rather than spiking during the pandemic, the incidence of these specific autoantibodies climbed slowly and steadily in a straight upward trajectory from 2019 to 2023.
The steady rise in cancer-linked autoantibodies likely reflects a growing awareness among doctors, who are now ordering the specialized tests more frequently than they did in the past. Above all, the complete lack of a sudden jump in 2020 reinforces the idea that these particular syndromes operate independently of viral outbreaks.
While the data reveals an important trend, the study has several limitations. The researchers conducted an observational analysis based purely on laboratory records. They did not have access to complete clinical histories or direct physical evaluations for the patients whose samples were tested.
Because of this physical barrier, the team could not confirm if the patients with positive autoantibody tests had suffered a corroborated coronavirus infection just prior to their specific neurological symptoms. They also could not completely rule out the possibility that a few of the rapid onset cases in 2020 were actually triggered by undiscovered tumors. The researchers had to rely on generalized group trends rather than case-by-case proof.
Despite the strict laboratory criteria, there is always a tiny residual risk of false positive results in chemical testing. The team noted that their testing methods remained completely uniform throughout the entire study period. Any sudden change in the number of positive samples is highly unlikely to be the result of a sudden spike in testing errors.
The primary researchers hope their findings encourage medical professionals to remain vigilant about the secondary consequences of viral outbreaks. The human immune system is incredibly powerful, and its biological mistakes can result in lasting neurological damage. As the virus that causes COVID-19 becomes a permanent part of the respiratory illness landscape, keeping an eye on unusual neurological symptoms will remain a high priority in hospital care.
The study, “Impact of the COVID-19 Outbreak on the Incidence of Autoimmune Encephalitis and Paraneoplastic Neurological Syndromes Associated Antibodies in Singapore,” was authored by Rui Ling Rena Lau, Karine Su Shan Tay, Seyed Ehsan Saffari, Patricia Yut Wan Wong, Mei Ting Lim, Angelia Swee Hoon Koe, Jeanne May May Tan, Kok Pin Yong, Kevin Tan, Josiah Yui Huei Chai, and Tianrong Yeo.
