Is long COVID as common and as long-lasting as earlier research suggested? A Stanford University study published online on 27 October 2026 in the Proceedings of the National Academy of Sciences (PNAS) is forcing a second look. Using 14.4 billion insurance claims from about 245 million Americans, the researchers found that most long-term effects once blamed on the virus likely reflect a flaw in how patients were compared. They are careful to say this does not mean long COVID is not real. Here is what the study shows, what it does not show, and which long COVID tests a doctor may order when symptoms linger after an infection.
What the Stanford study found
Many large studies compared people who tested positive for COVID-19 with people who had no known infection, whether or not they were ever tested. The problem: people who get tested are not like people who do not. They tend to see doctors more often and may already have health problems. That can make conditions unrelated to the virus look like after-effects of COVID-19.
The team, which included Hamed Nilforoshan, Jure Leskovec, Stefan Wager and Steven Goodman, used a different comparison. They looked only at people getting tested, and compared those whose test came back positive with those whose test came back negative. Across 614 health outcomes followed for two years, here is how many could be attributed to the virus with each method:
| Time after infection | Usual method (compared with untested people) | New method (test-positive vs test-negative) |
|---|---|---|
| 1 to 4 months | 262 health problems | 43 (83.5% fewer) |
| 4 to 12 months | 286 | 5 (98.2% fewer) |
| 1 to 2 years | 369 | 0 |
To check their approach, the researchers also used “negative controls”, events the virus cannot cause, such as lightning strikes, dog bites and gunshot wounds. The usual method linked some of these to COVID-19; the new one did not. Among the top effects the study did confirm were muscle disease (myopathy), pneumonia not caused by tuberculosis and other breathing problems, including a collapsed lung (pneumothorax).
What the study does not say
“This doesn’t mean long COVID is 98 percent less common,” Leskovec said in Stanford’s announcement. The study shows that the way a comparison is built changes what can be confidently blamed on the virus. It does not dismiss what patients go through.
- Insurance claims capture fatigue and “brain fog” poorly, because these complaints are often not diagnosed or coded. Yet they are among the most common long COVID symptoms.
- The authors acknowledge their method may modestly underestimate some effects.
- Symptoms beyond one year can be genuine; what remains uncertain is whether they are directly caused by the virus.
- Epidemiologist Steven Goodman suggests long COVID may look more like other post-viral syndromes than previously thought.
How long COVID is defined in the US
In 2024, the National Academies of Sciences, Engineering, and Medicine (NASEM) proposed a common US definition: long COVID is an infection-associated chronic condition that occurs after SARS-CoV-2 infection and is present for at least three months, as a continuous, relapsing or progressive disease affecting one or more organ systems. The definition does not require a positive test, and there is no single lab test that confirms it. In practice, doctors first look for other explanations for the symptoms, and research programmes such as the NIH-funded RECOVER Initiative are still studying who develops it and why.
Which tests can monitor or screen for this?
No blood test can diagnose long COVID. Lab work is used to look for another cause of the symptoms, which is exactly the central message of the Stanford study. Depending on your symptoms, a clinician may start with:
- a full blood count (FBC), to look for anaemia or abnormal white cells;
- a comprehensive metabolic panel, which covers kidney function, electrolytes and liver tests;
- C-reactive protein (CRP), a marker of inflammation;
- ferritin, which reflects iron stores;
- fasting glucose or HbA1c, to check for diabetes;
- TSH, when symptoms suggest a thyroid problem.
Other tests depend on the symptoms, for example creatine kinase (CK) for muscle pain or weakness, or a D-dimer test if a blood clot is suspected. For lasting tiredness, our guide to blood tests for fatigue walks through the usual work-up.
What this means for you
If you still have symptoms after COVID-19, this study does not mean they are imagined. It makes two useful points. First, any symptom that lingers deserves a work-up, because it may have another, treatable cause, such as anaemia, diabetes, a thyroid disorder or low iron. Second, care for long COVID remains mostly about managing symptoms: gradual return to activity, breathing or smell rehabilitation and mental health support, coordinated by your GP.
Latest scientific advances
Beyond the question of how common it is, research is trying to understand what happens in the bodies of people who stay sick. A Swiss study published in the journal Science in 2024 measured thousands of proteins in the blood of patients followed for a year. In people with active long COVID, part of the immune system called “complement”, which normally helps fight germs, stayed abnormally switched on. The team also found signs of small blood vessel damage and activation of platelets, the cells involved in clotting.
What this means for you: these markers are not routine tests yet, and none can diagnose long COVID in a lab today. But they point toward future blood tests and more targeted treatments. The Stanford study, for its part, shows that to know whether a problem is truly caused by the virus, how patients are compared matters as much as how many are studied. A related line of research looks at the reactivation of other viruses, such as Epstein-Barr virus, which we covered in our article on EBV reactivation after COVID.
When to see a doctor
- If symptoms last more than four weeks after COVID-19: fatigue, shortness of breath, palpitations, memory or concentration problems, loss of smell.
- If symptoms get worse after even light effort, or keep you from working.
- Right away, by calling 999, for chest pain, sudden shortness of breath or breathlessness at rest, fainting, or pain and swelling in one leg.
- If you notice persistent low mood or anxiety: mental health support is part of care.
Glossary
- Long COVID: a chronic condition that follows SARS-CoV-2 infection and lasts at least three months, according to the 2024 NASEM definition.
- Selection bias: an error that appears when the groups being compared differ from the start, here because people who get tested are not like people who do not.
- Negative control: an event the factor being studied cannot cause, used to check that a method does not find false links.
- Insurance claims data: records of billed medical care, useful for very large studies but poor at capturing symptoms that are not diagnosed.
- Complement: a group of blood proteins that helps the immune system destroy germs.
Frequently asked questions
Does the Stanford study mean long COVID does not exist?
No. It shows that many effects once attributed to the virus may have other causes. The authors state that symptoms beyond one year can be genuine, even if their link to the virus is uncertain.
Is there a blood test for long COVID?
No. Blood tests help rule out other causes, such as anaemia, diabetes, thyroid disease or low iron. Research markers like complement proteins are not routine tests.
Which tests should I ask about if I am still tired after COVID?
A doctor will often start with a full blood count (FBC), a metabolic panel, CRP, ferritin and blood sugar, and add TSH or other tests depending on your symptoms.
Why are fatigue and brain fog hard to capture in this kind of study?
Insurance records mostly contain coded diagnoses and procedures. Fatigue and concentration problems are often not formally diagnosed, so they may be missed.
Sources
- Nilforoshan H, Reisler J, Jahanparast E, et al. Revised estimates of types and duration of long COVID symptoms based on claims records from 245 million US patients. PNAS, 2026. DOI: 10.1073/pnas.2604414123
- Stanford University. Study refines our understanding of long COVID, 28 October 2026, via Mirage News
- Healthcare in Europe. Study refines what “long Covid” actually is, 29 October 2026
- Cervia-Hasler C, Brüningk SC, Hoch T, et al. Persistent complement dysregulation with signs of thromboinflammation in active Long Covid. Science, 2024. DOI: 10.1126/science.adg7942
- National Academies of Sciences, Engineering, and Medicine. A Long COVID Definition, 2024
- NIH RECOVER Initiative
Further reading
- Blood Tests for Fatigue: 1 in 6 US Adults Are Frequently Tired, CDC Says
- EBV Reactivation After COVID: What Your Blood Test Results Mean
- Full Blood Count (FBC): How to Read Your Results
- D-dimer blood test: what your results mean
Understand your own blood test results with AI DiagMe
Still feeling unwell after COVID and holding results you are not sure how to read? AI DiagMe explains each value of your blood test in plain language within minutes, so you can go into your next appointment with clear questions.
