Start here: what an average deficit means
When researchers say COVID is linked to a cognitive deficit, they mean the average score of a large group sat a little below a comparison group. That is different from saying any given person is impaired. Inside that average are people who scored higher than before, people unchanged, and a smaller group with a clear drop.
Think of it like average height in two towns. One town can be a centimetre shorter on average while still containing the tallest person in the region. A population signal is real and worth understanding, and it still tells you little about where you personally land.
What the largest study actually found
In a study that invited 800,000 adults in England to a structured online assessment, people who had COVID and recovered showed a small global deficit of roughly a fifth of a standard deviation against those never infected. People whose symptoms had not resolved showed a larger gap, around two-fifths of a standard deviation, with memory, reasoning, and executive tasks most affected.
Translated very roughly, the smaller end corresponds to a few IQ-equivalent points and the larger end to more. Deficits were bigger after early-pandemic variants and after hospital or intensive care. The authors reported no adverse events and were explicit that this is observational.
high certainty for the effect sizewell-founded associationobservational, not proof of cause
Why measured and felt cognition do not always match
The same study found that objective task scores correlated only weakly with how people rated their own memory and brain fog. That mismatch is important and often misread. A weak correlation does not mean your experience is invented. It means the brief tasks and your daily demands are measuring different things.
Real-world thinking runs under fatigue, time pressure, interruption, and post-exertional malaise. A ten-minute test in a quiet moment can miss exactly the conditions where your difficulty shows up. Both the number and the lived report can be true at once.
What might drive the signal
Several mechanisms are under study rather than decided. Proposed routes include neuroinflammation and microglial reactivity, small-vessel and clotting changes that reduce delivery to brain tissue, and low peripheral serotonin affecting vagal and hippocampal signalling. These are candidate explanations, and the page on each treatment reflects how thin or strong its evidence is.
What the cognition data cannot do is prove which route operates in any one person. That is why this page stays with the observation, a measurable average shift, and links out to the mechanism pages rather than asserting a single cause.
What this does and does not predict
A population average does not forecast your trajectory. Many people improve over months, and the study found that those whose symptoms had resolved scored similarly to people with only short illness. Persistence of symptoms, not infection alone, tracked the larger deficits.
So the useful read is conditional. If your symptoms are easing, the data are mildly reassuring. If they persist, you belong to the group worth taking seriously, documenting, and supporting, not dismissing as stress or age.
How to use this if it is you
Objective testing can validate and track what you feel, but interpret it with the mismatch in mind. A normal screen does not erase a real struggle, and a low score is a starting point, not a label. Ask for testing that probes memory, reasoning, and executive function rather than a single quick screen.
Pair any assessment with the conditions of your real life. Note when difficulty spikes, especially after exertion, poor sleep, or orthostatic intolerance, because those patterns point toward drivers that have their own management.
What we don't know
Honest about the edges of the evidence. These are open questions, not settled answers.
- Whether the average cognitive deficit fully reverses over years, or leaves a small residue in a subgroup.
- Which biological route dominates in people with persistent measurable deficits.
- Why objective scores and self-reported brain fog correlate only weakly.
- How much of the population signal is driven by sleep disruption, mood, and orthostatic symptoms rather than direct brain injury.
- Whether later variants and vaccination meaningfully shrink the deficit.
- Which cognitive domains recover first, and whether targeted rehabilitation changes the curve.
What this means for you
A measurable average dip in the population is not a sentence for you. It is a reason to take cognitive symptoms seriously, to ask for proper testing, and to track your own pattern over time rather than accept a brush-off.
If your symptoms are improving, the data lean reassuring. If they persist, you are in the group the evidence says to investigate and support. Either way, the number describes a crowd, and you are a person inside it whose course is still being written.
References
Each reference links to the source on PubMed, PMC, or the publisher.