Most people get over an infection in days or weeks. Some don't. A proportion stay disabled for months or years, long after the germ can be found. This is one of the least understood parts of medicine, so this page tries to be clear about what's actually known.
It's not rare and not limited to one germ
The best early evidence comes from rural Australia. The Dubbo study followed 253 people from the moment of an infection with Epstein-Barr virus, Q fever (a bacterium) or Ross River virus. At six months, 12% had a prolonged illness with disabling fatigue, pain, difficulty thinking and mood disturbance, and 11% met the criteria for chronic fatigue syndrome. Three important findings:
- It looked much the same whichever germ had caused it.
- It happened at about the same rate after each.
- The severity of the first illness predicted it best, more than psychological or demographic factors.
Similar syndromes have followed Giardia outbreaks, SARS, Ebola and, on a huge scale, COVID-19, at rates that vary from study to study.
ME/CFS
Myalgic encephalomyelitis / chronic fatigue syndrome is the established name for the most severe, long-lasting version. Current definitions focus on:
- Substantial loss of function lasting at least 6 months (3 months in the UK's guideline)
- Post-exertional malaise: worsening of symptoms, often delayed by a day or more, after physical or mental effort that used to be fine. This is the hallmark, and sets it apart from ordinary tiredness.
- Unrefreshing sleep
- Cognitive problems or being unable to stay upright for long
There's no diagnostic blood test. Doctors first look for other treatable causes.
What the research has found
Two of the most detailed immune studies so far are small or cross-sectional, so they're leads rather than conclusions.
Walitt 2024, NIH. 17 people with post-infectious ME/CFS were compared with 21 healthy volunteers, in great depth. It reported changes suggesting long-term immune stimulation, with more naive and fewer switched-memory B cells, and T cell changes, plus differences by sex. It also reported altered autonomic function. Its description of a change in "effort preference" attracted a lot of criticism from patients and some scientists for how it was worded, and for the small sample.
Klein 2023, Yale. In 275 people, those with long COVID had different circulating immune cell populations, higher antibodies against SARS-CoV-2 and against other viruses, particularly EBV, and lower cortisol. Levels of these overlap between people, so they're group differences and not a test.
DecodeME, 2025. The largest genetic study of ME/CFS, still a preprint, compared up to 15,579 cases with about 260,000 controls. It found eight regions of DNA linked with the illness, three of them near genes involved in the response to infection. That points to a biological basis, though it doesn't yet show how the illness works.
Each of these ideas has support. None has been proven as the cause, and they may all be partly right in different people.
- Persisting virus or viral pieces, found in tissues in some long COVID studies
- Reactivation of latent viruses, such as EBV. See Viruses that stay.
- Exhausted or misdirected T cells, unable to clear a remaining trigger
- Autoantibodies, which have been found in some patients, against receptors or other targets. Trials of removing them have so far been mixed.
- Damage to small blood vessels and clotting
- Autonomic dysfunction, such as POTS, a racing heartbeat on standing
- Changes in metabolism or in cells' energy production
- A sickness response that doesn't switch off, see The brain
- A changed gut microbiome
What it isn't
- It isn't simply a "weak immune system". People aren't generally more likely to get serious infections, and the findings suggest something more like a dysregulated system than a weak one.
- It isn't a matter of just being deconditioned or depressed. Depression can come with any long illness, but post-exertional worsening isn't explained by either.
- It can't be fixed by "boosting" immunity. See Boosting.
What's recommended
The UK's NICE guideline from 2021 is the most influential recent review of the evidence. It no longer recommends graded exercise therapy, on the grounds that the evidence for it was weak and that patients reported harm. It does recommend personalised energy management, also known as pacing: staying within your limits and avoiding the crashes. CBT may be offered for coping, not as a cure. Other than that, treatment focuses on specific symptoms, such as sleep, pain and POTS.
There's no approved disease-modifying treatment yet. Several drugs are being tested, including antivirals and immune drugs, much of it driven by long COVID funding. Low-dose naltrexone has been widely used on the strength of early reports, but a 160-person placebo-controlled trial in post-COVID fatigue, still a preprint, found no benefit. See Low-dose naltrexone: the idea.
Prospects vary. Many people improve over the first couple of years, particularly with shorter illness, but complete recovery is less common in long-standing ME/CFS. A doctor can help rule out other things: anaemia, thyroid problems, coeliac disease, sleep apnoea, inflammatory diseases and more.
In the best early study, about one person in ten had a prolonged illness six months after a serious infection, at similar rates across three very different germs. Immune differences have been found, but no cause or test is established. Pacing is the current advice, and post-exertional worsening is the feature to watch for.