No, SARS-CoV-2 is not reducing our intelligence
Why this myth refuses to die
Over the past four years, a peculiar claim has repeatedly resurfaced in media headlines, social channels, and “doomscaping” subcultures: the idea that SARS-CoV-2 infections, or worse, reinfections, are collectively reducing human intelligence. According to this narrative, each viral exposure acts as a cumulative neurological insult, silently chipping away at our cognitive abilities and steering societies toward an inevitable decline.
It’s a powerful story. It is also deeply misleading.
There is no robust evidence that COVID-19 has caused a population-level reduction in intelligence, nor that reinfections progressively erode cognitive function. On the contrary: as larger, better-designed studies have accumulated, including randomised cohorts, long-term population datasets, and pre-/post-infection cognitive testing, the pattern has become remarkably consistent. Short-term cognitive symptoms are common during acute illness and recovery, but persistent, measurable cognitive impairment is uncommon, and large-scale indicators of intelligence show no decline.
Yet the myth persists, fuelled by sensational reporting, preprints with methodological problems, misinterpretation of small MRI findings, and online communities that reward catastrophising. Because the claim touches on something fundamental, the fear of losing our minds, it spreads easily, even when unsupported.
This Substack is not a defence of SARS-CoV-2; it is a defence of scientific reasoning. Extraordinary claims require extraordinary evidence, and the “COVID is lowering global IQ” narrative does not come close to meeting that threshold. By walking through the evidence, explaining where flawed claims arise, and presenting what high-quality studies actually show, we can replace fear with clarity, and keep the conversation anchored in reality rather than rumour.
The flawed “IQ Drop” claim
The strongest promoters of the “COVID lowers intelligence” narrative lean heavily on a narrow type of evidence: observational, retrospective association studies. These papers examine groups of people who had COVID-19 and compare their performance on cognitive tasks with those who did not. At first glance, this can look compelling. But the detail is in the methodology, and once you examine these, the conclusions are weak at best.
Confounding dominates these studies. People who get infected differ in countless ways from those who do not. These differences are not trivial; they map directly onto cognitive performance: education levels; type of employment; socioeconomic status; lifestyle and stress exposure; pre-existing health conditions; access to healthcare. and mental health burden during the pandemic.
No statistical model can fully disentangle these. Even “fully adjusted” analyses leave enormous residual confounding, meaning any observed difference could have existed long before the infection.
Lack of pre-infection baseline data. Most of these studies only measure cognition after COVID. Without pre-infection scores, any “decline” may simply reflect: natural variation, regression to the mean, or pre-existing cognitive differences between groups.
It is impossible to claim “COVID lowered cognition” if you never measured cognition before the infection.
Tiny effect sizes inflated into dramatic claims. Many of the widely shared “IQ drop” stories rest on minute differences in app-based testing, differences far smaller than what would meaningfully register in IQ scoring or daily functioning. A 1–3% change on a touchscreen task does not equal brain damage.
But because “COVID makes people slightly slower at a reaction-time task” is boring, the media converts these marginal findings into “COVID lowers your IQ.”
Self-selection and measurement bias. Most app-based testing studies recruit: people with health anxiety, people experiencing symptoms, people already concerned about COVID, and people motivated to seek testing.
This produces a biased sample that is guaranteed to exaggerate associations.
Overinterpretation of associative statistics. Retrospective datasets can detect correlation, but they cannot determine cause. Many COVID-related risk factors (e.g., lower socioeconomic status, chronic illness, higher stress burden) are independent predictors of poorer cognitive performance. Yet the headline becomes: “COVID causes cognitive decline.”
The leap from correlation to “IQ loss”. The term “IQ” is almost never used in the actual papers. Instead: reaction time, memory precision, attention tasks, or self-reported “brain fog” are stretched far beyond their meaning. These metrics do not map cleanly onto standardized IQ scales.
In other words: Saying “COVID lowers your intelligence” is a category error.
The problem with digital test platforms. Some of the most-cited studies use quick online or app-based games that: are not validated for clinical diagnosis, have large day-to-day variability, are affected by sleep, stress, and device type, and cannot determine long-term change.
These platforms are useful for population-level trends, not diagnosing neurological decline.
The core claim that “COVID lowers intelligence” rests on loose associations derived from biased datasets, using tools that don’t measure intelligence, in populations that differ in almost every meaningful way. This is not evidence of global IQ decline; it’s evidence of poor epidemiology.
Reinfections are misrepresented
A central pillar of the “COVID lowers global intelligence” narrative is the claim that each reinfection adds another layer of neurological injury, supposedly causing a stepwise decline in cognitive function. This idea is intuitively frightening, and that is precisely why it spreads so easily.
But it is not supported by high-quality evidence.
Reinfections are consistently milder physiologically. Across large datasets, including UK, US, Danish, Qatari, and multiple Asian cohorts, reinfections almost universally show: lower inflammatory markers, shorter symptom duration, fewer complications, lower hospitalization rates and lower risk of longCOVID
If inflammation-driven cognitive symptoms were cumulative, reinfections should produce worse outcomes. Instead, they produce less inflammation and better recovery profiles.
No prospective evidence of cognitive decline after reinfection. Studies with repeated cognitive testing over time do not show stepwise deterioration. Instead, scores remain stable or improve, or vary within the same small range seen in non-infected controls.
If reinfections caused ongoing brain injury, this pattern would be impossible.
Associational studies exaggerate reinfection risk. The few studies claiming cognitive issues after reinfection typically suffer from: self-reported symptoms (not objective testing), unclear infection verification, unverified reinfection classification, short time windows after reinfection, biased recruitment (“self-selected symptom reporters”)
Symptoms reported weeks after reinfection are often acute-phase effects, temporary mood or sleep disruption, or misattributed viral illness (many “reinfections” in surveys are not even SARS-CoV-2).
Time bias is ignored. Reinfections usually occur more recently than first infections. That means: people with a single early infection may have recovered for years and those with reinfections are closer to their most recent acute illness
Comparing these groups without accounting for time since infection is fundamentally flawed. Yet this mistake appears repeatedly in alarmist interpretations.
Reinfections do not imply cumulative brain injury. There is no biological mechanism by which repeated mild viral infections, especially when systemic inflammation is lower each time, would sequentially degrade human intelligence. If such a mechanism existed, seasonal flu, RSV, adenovirus, and common coronaviruses would have caused measurable declines long before 2020.
They have not.
Headlines misrepresent what the data show. Media coverage of reinfection studies often introduces claims not made in the papers themselves:
“Reinfection causes brain damage”
“Cognitive deficits stack with each case”
“Each COVID infection makes you less sharp”
The actual papers typically describe, self-reported fatigue, slower reaction times, acute concentration issues, and temporarily reduced task accuracy. These are not measures of intelligence, nor do they demonstrate cumulative decline.
What the best evidence indicates. The highest-quality studies converge on a simple conclusion. Reinfections do not produce additive, long-term cognitive impairment.
If anything, the physiological response becomes milder over time. The claim of cumulative neurological damage rests almost entirely on misinterpreted associations and social-media amplification, not on rigorous science.
The MRI study myth
One of the most persistent arguments behind the “COVID lowers intelligence” narrative comes from the widely cited UK Biobank MRI study, which compared brain scans taken before and after infection. It has been repeatedly misrepresented as proof of “brain shrinkage” or permanent neurological injury.
The reality is far more modest. The changes were tiny. The structural differences detected were very small, often within the range of natural variation seen with, ageing, temporary inflammation, dehydration, or loss of smell.
These changes do not equate to cognitive impairment. Effects were concentrated in older adults. Younger and middle-aged adults showed little to no measurable structural change. Most findings came from participants over 60, where brain volume naturally fluctuates.
Olfactory areas respond to smell loss. Changes in regions connected to smell processing are expected during and after anosmia. This happens with other respiratory viruses too. It is not evidence of intelligence loss.
Authors explicitly cautioned against overinterpretation. The study’s own authors stressed that; MRI variation ≠ cognitive decline, structural differences ≠ brain injury, clinical significance was uncertain.
These caveats rarely make it into headlines. There is no corresponding drop in cognitive performance. Follow-up cognitive tests showed very small differences that did not translate into IQ changes or functional impairment.
The MRI study showed small, mostly reversible changes associated with smell pathways in older adults, not global brain shrinkage and not loss of intelligence.
The main concerns in many studies:
A large fraction of alarming claims arises not from strong evidence, but from recurring weaknesses in study design. These limitations don’t make the research useless, but they make sweeping conclusions about “IQ decline” completely unjustified.
Reliance on self-reported symptoms. Studies often measure “cognitive impairment” through: self-reported brain fog, self-reported memory problems, and subjective questionnaires.
These reflect perception, not actual cognitive performance. Fatigue, stress, depression, poor sleep, and anxiety all inflate self-reported deficits.
No pre-infection baseline data. Without knowing what a person’s cognitive abilities were before infection, researchers cannot determine whether any measured deficit is new. Yet many widely shared studies only assess people after COVID.
Poor infection classification. Some datasets include people who were: never tested, self-diagnosed, misclassified, infected with other viruses, unsure of dates or number of infections.
This introduces noise that can easily look like “cognitive effects.”
Different time intervals since infection. People with a single early infection may be years post-recovery, while those with reinfections may only be weeks out from an acute illness. Short-term symptoms then get misinterpreted as chronic cognitive decline.
Selection bias. Many studies recruit: people experiencing symptoms, people concerned about long COVID, people more likely to seek help or testing.
This creates a non-representative sample that exaggerates risk.
Overinterpretation of tiny effect sizes. Differences that are statistically significant in large samples can be functionally irrelevant. A 1–2% change in reaction time or working memory accuracy is not an IQ drop.
App-based and web-based cognitive tests. These tools are convenient but, are not validated for clinical diagnosis, are sensitive to device type, time of day, and fatigue, and have large within-person variability.
They are not robust measures of intelligence.
Most “IQ drop” claims trace back to studies that cannot distinguish true cognitive change from noise, bias, or confounding. These limitations explain why stronger, better-controlled studies consistently fail to find population-wide cognitive decline.
What leading studies actually show
ARIC / Collaborative Cohort (JAMA Network): Prospective testing in older adults found accelerated cognitive decline only in those hospitalized, with no significant change for mild or moderate, non-hospitalized cases.
COVID Symptom Study Biobank (UK): Those with ≥12 weeks of symptoms (“long COVID”) had slightly lower accuracy scores, but people who fully recovered showed no deficits. No evidence of continuing decline over time, especially outside of long COVID.
Cognivue Clarity Study: No overall decline in global cognition one year post-infection in middle-aged adults.
NIHR UK Large Sample: Over 110,000 participants. Small, measurable deficits in persistent-symptom cases (–0.42 SD), but smaller or absent deficits in those who recovered. Hospitalized or severe infections had stronger effects, but no indication of widespread IQ loss.
JAMA Network Survey: Among individuals self-reporting “long COVID” (~14,767 adults), cognitive complaints abound but are strongly linked to depression and daily function, not objective decline.
European Psychiatry Longitudinal Study: Some individuals with mild COVID had long-term cognitive symptoms, but impacts were moderate and not universal. Persistent fatigue and mental health were bigger impacts on daily life than “IQ drop”.
PHOSP-COVID (Oxford): Hospitalized patients showed attention and memory deficits equal, by the researchers’ interpretation, to “10 IQ points”, but applicability to non-hospitalized cases is unclear and effect sizes remain small.
Interventions Trial (JAMA Neurology): Treatment trials for long COVID-related cognitive symptoms failed to find significant improvement over short follow up, highlighting the challenge of post-infectious cognitive complaints.
Why the myth persists
The formula is simple: fear plus simplicity equals virality.
Journalists, often lacking scientific nuance, can twist correlation into causation.
Doom-focused audiences amplify the worst-case interpretations.
Fatigue and anxiety make people vulnerable to alarmist headlines, especially those suggesting irreversible brain damage.
Reality versus narrative
Despite headlines warning of “brain damage,” “shrinking intelligence,” or “stacking cognitive deficits,” the actual scientific evidence paints a far more grounded picture.
Cognitive symptoms do occur, but they are not IQ loss.
Fatigue, brain fog, and concentration problems are well-documented after many viral infections, including influenza, glandular fever, dengue, and others. COVID-19 is no exception. These symptoms: are typically transient, fluctuate with sleep, stress, and mental health, and do not equate to measurable reductions in intelligence.
Domain-specific issues, not global cognitive decline. Where studies do detect impairment, they usually affect: attention, memory precision, executive function, and largely in those with long COVID or severe acute illness. These are narrow effects, not the broad decline one would expect from “brain damage” or IQ loss.
Most people recover fully. The overwhelming pattern across longitudinal research is: stability or improvement over time, no continued decline, no progressive worsening with reinfections, and full return to baseline for the majority.
Long COVID exists, but it is not a collapse of global intelligence. A subset of people genuinely experience persistent cognitive symptoms. These deserve serious attention. But even in this group, deficits: are moderate, are not equivalent to IQ loss, and are heavily influenced by fatigue, mood, and sleep quality.
Intelligence metrics at the population level have not budged. If COVID truly caused a measurable fall in intelligence, we would see it in: educational outcomes, professional testing, cognitive trend data, national IQ assessments, longitudinal population cohorts.
We do not. Countries with millions of infections show no detectable drop in population-level intelligence.
The narrative persists because fear is contagious. Alarmist interpretations spread much faster than corrections. Online communities reward catastrophising. Headlines amplify the scariest framing. Yet none of this changes the underlying scientific reality.
Post-COVID cognitive symptoms are real for some individuals, but they are not evidence of a global drop in intelligence. The data consistently show transient, domain-specific effects, not the collapse of cognitive ability imagined in viral social-media narratives.
What we know isn’t true
❌Infection with SARS-CoV-2 does not cause permanent, population-wide brain injury.
❌Reinfections do not accumulate cognitive harm.
❌Global intelligence is not in decline because of COVID-19.
❌Online surveys do not constitute reliable proof of neurological catastrophe.
❌Human Intelligence: Remarkably Resilient
Intelligence is shaped by decades of environment, education, and social factors. It does not crumble from short-term illnesses, even those as intense as COVID-19. If respiratory viruses could meaningfully lower population IQ, we would have seen it after generations of influenza, measles, or RSV, there’s simply no precedent.
Other studies:
COVID-19 and Cognitive Impairment: Systematic Review (2025)
This review synthesises the data from cohort, case-control, and longitudinal studies, highlighting that the most pronounced cognitive declines occur in individuals who experienced severe COVID-19; especially those hospitalised or in ICU. The review emphasises that mild and moderate cases generally do not show significant, persistent losses, and that deficits, when present, tend to affect attention and executive function rather than global intelligence.
Findings are domain-specific, often reversible, and not observed on a population scale. Severe COVID’s impact is the exception, not the rule.
Changes in Memory and Cognition During SARS-CoV-2 (2024)
This paper uses baseline-corrected cognitive scores to document subtle but statistically significant decreases in memory precision and executive functioning after infection. The authors note that detectable impairment is most common among those with longCOVID or severe acute illness, but these changes are small, often improve over time, and do not result in measurable drops in global IQ across the general population.
Most people recover full cognitive performance over time, and impairment is not widespread or durable enough to change national intelligence metrics
Cognitive Impairment 2 Years After Mild to Severe SARS-CoV-2 Infection (2025)
A cohort study found persistent symptoms in a subset of survivors, particularly among those who experienced severe infections, but most recovered with no substantial change in global cognitive scores at two-year follow-up.
Persistent effects are limited to survivors of severe disease; mild cases show no measurable drop in cognitive ability or IQ. Recovery is the rule.
COVID-19 Enduringly Impacts Cognitive Performance in Undergraduates (2025)
This study of university students found that while some had lingering attention and executive function deficits up to 17 months post-infection, these effects were neither universal nor substantial enough to affect general intelligence tests or academic achievement rates.
Small, domain-specific effects in select individuals do not amount to population-wide IQ decline.
Longitudinal Cognitive Decline in Mild Cases (2024)
In a study with pre-pandemic baseline cognitive data, mild COVID-19 cases showed a slight increase in odds of MOCA score decline, but most remained within normative ranges, and differences were not sufficient to change standardised intelligence metrics.
Subtle changes in cognitive screening scores do not equate to measurable, lasting intelligence loss.
Cognitive Impairment in Post-Acute COVID Phases (2024)
A meta-review confirmed domain-specific impairment, mainly in attention and processing speed, with risks tied to inflammation and acute illness rather than viral infection itself.
A small subset may experience persistent cognitive symptoms, but these are not widespread or severe enough to move population-wide IQ averages.
Impact of Pandemic Era on Cognitive Decline in Seniors (2025)
This large cohort study found that the pandemic period accelerated age-linked cognitive and structural brain changes, especially among elders and those infected, even with mild disease, but there was no evidence for a drop in national intelligence.
Effects are age-related and specific to cognitive domains, not global intelligence, and were evident even before the pandemic.
Physical, Cognitive, and Mental Health Impacts of Omicron Reinfection (BMC Medicine, 2025)
A large observational, with invited subjects based on unknown criteria, study from Wuhan found that people with both original and much more recent Omicron COVID-19 infections reported more physical, mental, and subjective cognitive symptoms than those with a single past infection or none.
Symptoms were self-reported and did not reflect measurable drops in population IQ or global cognitive ability.
Scientific consensus
SARS-CoV-2 is genuine and can cause lasting symptoms for a subset of sufferers, especially those with longCOVID or severe initial illness. But the narrative of a world quietly growing less intelligent is a fear story, not a scientific fact. The most rigorous studies repeatedly refute it.
While post-COVID cognitive symptoms, including fatigue, brain fog, and slower processing, are well documented, they are domain-specific, common to other viral and bacterial illnesses, and typically resolve over time.
Population-level intelligence (as measured by standardised tests and educational outcomes) has not dropped in countries with millions of SARSD-CoV-2 infections. Severe, persistent cognitive sequelae are rare and most prevalent among those with critical illness, and even then, do not amount to permanent losses in general intelligence or national IQ metrics.
SARS-CoV-2 is not causing humanity to lose intelligence. The myth persists due to sensationalism and selective citation, not rigorous science. What’s urgently needed is clarity, honesty, and the will to let scientific evidence, not panic, lead us forward.