In February 2024, the New England Journal of Medicine published a landmark study from Imperial College London that assessed cognitive function in more than 112,000 people across England who had recovered from COVID-19. The findings were stark. Even people whose COVID-19 symptoms had fully resolved within four to twelve weeks showed cognitive deficits equivalent to a 3-point drop in IQ compared to those who had never been infected. Those with persistent long COVID symptoms — defined as lasting beyond twelve weeks — showed the equivalent of a 6-point IQ loss. For patients whose illness had been severe enough to require intensive care, the deficit reached 9 points, with particular impairments in memory, reasoning, and executive function.
These are not small numbers. A 3-point IQ difference is the kind of effect that shows up measurably in professional performance, academic outcomes, and daily decision-making across a population. Multiplied across hundreds of millions of COVID-19 survivors worldwide, the cognitive burden represents what researchers are increasingly describing as a slow-moving public health emergency — one that has received a fraction of the attention devoted to the acute phase of the pandemic, and one that carries specific and underexamined implications for countries like Bangladesh and its South Asian neighbours.
What Is Happening in the Brain
The science of long COVID's neurological effects has advanced considerably since the early patient-reported descriptions of "brain fog" in 2020 and 2021. Researchers now have a substantially clearer picture of the biological mechanisms producing these effects, even as significant questions about treatment remain unanswered.
A 2024 study published in Nature Neuroscience by researchers at Trinity College Dublin identified one central mechanism: disruption to the blood-brain barrier. In patients with long COVID and cognitive impairment, the blood vessels supplying the brain showed measurable "leakiness" — a breakdown in the normally tight barrier that separates circulating blood from brain tissue. This barrier disruption allows inflammatory molecules, cytokines, and potentially viral particles to penetrate the brain, triggering neuroinflammation that impairs the neural circuits responsible for memory, attention, and executive function. The finding was important precisely because it provided an objective, measurable biomarker that could distinguish patients with cognitive impairment from those with long COVID but without brain fog.
A 2025 study from Yokohama City University in Japan, published in Brain Communications, identified another specific mechanism: abnormal increases in AMPA receptor density across multiple brain regions. AMPA receptors are essential to synaptic plasticity — the process by which brain connections strengthen or weaken in response to experience, which is fundamental to learning and memory. The widespread elevation of AMPA receptor activity observed in long COVID patients with brain fog suggests a state of dysregulated neuronal excitability that interferes with normal cognitive processing. The researchers framed this as establishing brain fog as a "measurable, biological condition" — an important clarification in contexts where the symptom has sometimes been dismissed as psychological.
The systematic review published in the Archives of Clinical Neuropsychology in February 2025, which analysed 36 studies across the peer-reviewed literature, found that the cognitive domains most consistently impaired in long COVID are executive function, memory, attention, and processing speed. A separate meta-analysis of over four million COVID-19 patients published in BMC Neurology found that approximately 34 percent of survivors reported cognitive deficits lasting beyond six months post-infection. A Mount Sinai longitudinal study that followed patients for 42 months found that while cognitive measures generally improved over time, processing speed and executive functioning remained below the normative mean even at three and a half years post-infection.
The South Asian Dimension
Bangladesh reported over 2 million confirmed COVID-19 cases by the end of the pandemic's main waves, with the true infection burden substantially higher given the limitations of testing infrastructure documented extensively in public health analyses of the country's pandemic response. The WHO's 2024 assessment of Bangladesh's health security noted the need for continued development of diagnostic and surveillance capacity — a gap that means the population of Bangladeshis living with long COVID cognitive effects is essentially unmeasured.
This unmeasured burden is not evenly distributed. The academic literature on long COVID consistently identifies several risk factors that are disproportionately prevalent in South Asian populations: diabetes, obesity, hypertension, more severe initial infection, and hospital or ICU admission. Bangladesh has a documented and growing burden of non-communicable disease — according to WHO data, diabetes affects approximately 14 percent of the adult population, hypertension prevalence is rising, and the health system has historically been better equipped to address infectious disease than chronic conditions. These comorbidities, the literature makes clear, are associated with both worse acute COVID-19 outcomes and greater likelihood of persistent long COVID symptoms including cognitive impairment.
The University of Miami research published in PLOS One in 2024, which followed COVID-19 patients for two years, specifically noted that its findings about persistent brain fog, word-finding difficulties, and working memory deficits could be "especially concerning for Hispanic, Black and other underrepresented communities, which are already at higher risk for neurocognitive disorders." The underlying point — that populations with fewer diagnostic resources, higher comorbidity burdens, and less access to specialist neurological care are likely to experience the cognitive effects of long COVID more severely and to receive less appropriate treatment — applies directly to Bangladesh and South Asia.
The research published under the Brain Communications systematic review on persistent neuropsychiatric symptoms after COVID-19 included a reference to a Bangladeshi study: "Treatment, persistent symptoms, and depression in people infected with COVID-19 in Bangladesh" — evidence that Bangladeshi researchers have begun to engage with this question, though the volume of research on the South Asian long COVID experience remains small relative to the scale of the affected population.
The Workplace and Economic Consequences
Long COVID's cognitive effects are not an academic concern. The neurological burden of post-COVID syndrome has been estimated at approximately $3.7 trillion USD globally in associated healthcare costs and productivity loss. In the United States, more than one million working-age adults have withdrawn from the workforce as a direct consequence of long COVID symptoms, with brain fog and cognitive impairment among the primary reasons cited by those unable to return to full employment.
For Bangladesh, where approximately 65 percent of the population is of working age, where the garment industry depends on a large skilled and semi-skilled workforce, and where the IT and services sectors are growing rapidly as an economic priority, the workforce implications of a poorly documented cohort of COVID survivors experiencing cognitive impairment deserve serious attention. The cognitive domains most consistently affected — processing speed, executive function, attention, and working memory — are precisely those most relevant to complex manufacturing, quality control, software development, and service delivery work.
The effect on younger workers and students is a particular concern. A 2025 study published in Brain and Behavior Immunity, which examined COVID-19's cognitive impact in university undergraduates, found that 40 percent self-reported brain fog and 37 percent showed objective evidence of cognitive impairment on computerised testing up to 17 months after infection. The neuroimaging data showed that previously infected students exhibited brain activity patterns during cognitive tasks that resembled those of adults four decades older. Bangladesh has a large and growing university-age population that experienced significant COVID-19 exposure during the pandemic years and that represents the country's future skilled workforce.
Recovery: What the Evidence Shows
The trajectory of cognitive recovery in long COVID patients is, overall, encouraging — with important caveats. The Mount Sinai 42-month longitudinal study found that attention, working memory, verbal learning and memory, and processing speed all improved progressively over the follow-up period. However, processing speed and executive functioning remained below normative levels even at 42 months in many patients, suggesting that complete recovery is not guaranteed and that some patients face long-term functional deficits.
Several intervention approaches have shown promise in the research literature. A 2025 narrative review published in Biomedicines examined 12 studies of interventions for long COVID neurocognitive deficits and found that cognitive training, non-invasive brain stimulation therapy, graded exercise rehabilitation, and targeted pharmacological interventions all showed positive signals — though the evidence base for each remains limited and the optimal protocol for cognitive rehabilitation in this population has not been established. A randomised controlled trial of constraint-induced cognitive therapy published in 2025 found that participants who received the intervention adhered well to the protocol and reported high satisfaction with outcomes, providing a basis for larger trials.
Vaccination status has also been shown to matter for recovery. The Imperial College London NEJM study found that being vaccinated narrowed the cognitive performance gap between those who had COVID-19 and those who had not. This finding, combined with the evidence that the original SARS-CoV-2 strain and the Alpha variant produced worse cognitive outcomes than later variants, suggests that the Bangladeshi population — which experienced significant vaccination roll-out but with delayed and uneven access — may face a complex picture of variable cognitive risk depending on which variant cohort an individual was infected by and what their vaccination status was at time of infection.
The Diagnostic Gap and What Bangladesh Should Do
The most pressing issue facing Bangladesh with respect to long COVID cognitive effects is the absence of systematic measurement. Without knowing the scale of the affected population, the severity distribution, the comorbidity profile, or the relationship between acute infection severity and long COVID cognitive outcomes in Bangladeshi patients specifically, it is impossible to plan appropriate health system responses.
icddr,b — which has produced foundational research on infectious disease, nutrition, and maternal and child health that has shaped global health policy — has the scientific infrastructure to conduct rigorous long COVID cognitive outcome studies in Bangladesh. The institution's experience with population-based cohort studies, its existing relationships with clinical facilities across the country, and its analytical capacity position it to fill a research gap that matters both locally and for global understanding of long COVID's effects in lower-middle-income country populations. The tools exist: validated neuropsychological assessment batteries including the Montreal Cognitive Assessment, the Brief Memory and Executive Test, and computerised processing speed assessments have been deployed in long COVID research globally and can be adapted for use with Bangladeshi populations with appropriate cultural and linguistic calibration.
At the health system level, the recognition of long COVID cognitive symptoms as a legitimate clinical entity requiring specialist neurological assessment is a prerequisite for appropriate care. Bangladesh's neurological care capacity is concentrated in tertiary hospitals in Dhaka and a small number of divisional centres — a distribution that means patients in rural and peri-urban areas face near-insurmountable barriers to specialist assessment of cognitive symptoms that are subtle enough that many patients themselves may not recognise them as COVID-related sequelae years after their initial infection.
The global scientific understanding of long COVID's neurological mechanisms is advancing rapidly. The identification of blood-brain barrier disruption and AMPA receptor dysregulation as specific, measurable biological processes has opened pathways toward targeted diagnostic tests and pharmacological interventions that did not exist even two years ago. For Bangladesh and South Asia more broadly, the imperative now is to build the research infrastructure to understand the local burden, the clinical infrastructure to identify and support affected patients, and the awareness among clinicians and the public that the pandemic's cognitive consequences did not end when the acute emergency did.
WinTK covers global health innovation and its implications for Bangladesh and South Asian health systems. For more analysis on post-pandemic health challenges, explore our news and analysis section.
