A new Dutch study in BMJ Open 2025 followed people 2 years after COVID-19.
Even among those never hospitalized, 43% still had persistent symptoms.
This is a massive population-level impact.🧵
Hospitalized patients had higher prevalence of post-COVID (65%).
But they are a fraction of all infections.
The real burden comes from the huge group of mild cases!
The CORFU study = Corona Follow Up.
4,200 participants, 7 patient cohorts + controls.
Follow up at 3, 6, 12, 18, and 24 months.
Groups.
non-hospitalized
hospitalized (ward)
ICU
controls (no COVID)
At 24 months - 901 patients, 434 controls.
Symptoms tracked - fatigue, sleep problems, brain fog, muscle weakness, shortness of breath, palpitations, loss of smell/taste, etc.
Quality of life measured by
EQ-5D-5L - 5 health domains
EQ-VAS - self-rated health (0–100)
HUS - health utility score (0–1)
Most common symptoms (non-hospitalized vs controls):
Fatigue 29% vs 19%
Sleep problems 18% vs 12%
Cognitive issues 14% vs 4% (sic!)
Shortness of breath 12% vs 4%
Muscle weakness/pain 16% vs 11%
Overall - 43% of non-hospitalized had ≥1 symptom after 2 years.
Quality of life at 2 years.
Controls - EQ-VAS 75.6 / HUS 0.84
Patients - EQ-VAS 70–72 / HUS 0.77–0.81
That drop is equivalent to living with a chronic illness! And importantly - seen across all groups, not just ICU.
So Long COVID ≠ only PICS (post-intensive care syndrome).
The impact is everywhere.
Work & school - fatigue, brain fog, poor sleep
Healthcare - repeated visits, rehab, mental health, crowded clinics
Society - lower quality of life, disability, early exits from workforce
Policy signal.
Even if only 40% of non-hospitalized are affected, they represent the largest absolute burden.
Long COVID is not rare. It is a population-wide phenomenon reshaping health, education, and the economy.
And this is the reality in many countries:
Hundreds of thousands living with fatigue, cognitive problems, sleep disruption, or breathlessness - often without adequate support.
The result? A damaged population.
The consequence? Enormous economic costs.
These national economic losses expose a fundamental failure of public health. And of governments. @szupraha @ZdravkoOnline
Klein at al., Two years and counting: a prospective cohort study on the scope and severity of post-COVID symptoms across diverse patient groups in the Netherlands - insights from the CORFU study. bmjopen.bmj.com/content/15/9/e…
Most infections in this study happened before vaccines and during the wild type/Alpha/Delta waves.
By 24 months, 89% had been vaccinated.
Later follow-ups did include some Omicron cases, but they were a minority.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
New mega–meta-analysis on #LongCOVID ever prevalence.
Long COVID has affected 36% of people after a COVID-19 infection worldwide.
South America: 51%
Europe: 39%
Asia: 35%
USA: 29%🧵
Important note.
This 36% is “ever prevalence” - the share of people who at any point during follow-up had symptoms lasting ≥2 months after infection.
It does not mean they all still have symptoms today. Current prevalence is lower (eg 9% of US adults).
Why 2 months?
Different studies used different definitions of long COVID (WHO = 3 months, NASEM = 3 months, others = 2 months). To include the widest possible dataset, this meta-analysis defined long COVID as ≥1 new or persistent symptom lasting ≥2 months after infection.
New study in COVID (MDPI) followed 297 people 3 years after a single COVID-19 infection.
Key message: cognitive performance did not return to normal. Even one infection left measurable deficits🧵
Participants.
Mild - 101
Moderate (hospitalized) - 101
Severe (ICU, hypoxia) - 95
These were infections from the era of the original virus & Alpha variant (up to 2021). Omicron was not yet circulating.
Scores after 3 years (higher = better).
Divided attention: 44.5 - 36.1 - 27.3
Working memory: 10.6 - 9.3 - 9.0
Executive control: 9.0 - 7.6 - 7.5
Verbal fluency: 14.6 - 14.3 - 12.2
Recognition memory: 23.9 - 23.2 - 21.6
Trend is clear: more severe illness = greater cognitive loss.
May a new NIH preprint rewrite the COVID textbooks?
Turns out SARS-CoV-2 doesn’t start with ACE2. It latches onto heparan sulfate clusters on the cell surface. ACE2 only acts later, inside endosomes. Robust methods. Still preprint.🧵
Why that matters - NIH labs have the time, tools, and independence to go deep. When they drop a preprint, it’s worth paying attention.
The tech here is next-level.
MINFLUX nanoscopy (1–3 nm precision)
STED super-resolution (60 nm)
electron microscopy
You can literally see individual virions docking.
Behind each number is a child. After COVID, kids had 2× higher risk of self-harm. Now, the same lead author Kim et al. 2025 shows: kids depression & anxiety diagnoses rise by 49% in the year after infection🧵
This was no small study. Researchers tracked 154,000 children aged 6–15 in Utah.
In 2021 -
9.2% of those with COVID developed new depression or anxiety vs 5.3% in children who never had COVID.
That’s almost a doubling of risk.
Severity mattered.
Mild COVID +40% risk.
Severe (hospitalization/ER) +59% risk.
The more severe the illness, the heavier the psychological toll.
A new preprint study shows that the SARS-CoV-2 protein ORF6 directly kills human neurons.
Not by accident, but through necroptosis - a kind of cell death where neurons burst and fuel inflammation.
This may underlie the long-term brain symptoms of Long COVID.🧵
Researchers tested 22 viral proteins. The most toxic? ORF6.
In neurons, it flipped on the necroptosis switch - RIPK3 and MLKL.
Biologically - these molecules punch holes in the cell membrane until the neuron literally ruptures.
In patient brain tissue, especially the hippocampus (memory center), they found the same pattern
dead neurons + necroptosis markers.
That’s a direct biological link between infection and symptoms like brain fog and memory loss.
Human data.
In a cohort of 44M+ people, those hospitalized with severe COVID-19 had a 19% higher risk of being diagnosed with lung cancer compared to uninfected controls!
Mild infections didn’t carry this risk - it’s the severity that matters.
Animal models.
Mice that had recovered from severe COVID-19 or flu showed faster tumor growth and worse survival after tumor initiation.
The effect persisted for weeks to months after the virus was cleared.