Millions of people have had long-term effects after COVID-19, but are these caused by something specific to SARS-CoV-2 – or rather by the body having experienced a severe infection? A large nationwide registry study shows that the risk rises especially after the most severe disease courses and that severe COVID-19 resembles other serious infections more than researchers had expected.
Lene, 54, and her father, Jørgen, 72, contract COVID-19 within a short time of each other. Lene recovers quickly, whereas Jørgen becomes so ill that he has to be hospitalised urgently. He gets over the infection itself, but shortly afterwards develops new respiratory symptoms that doctors believe may be a long-term effect of the severe infection.
COVID‑19 can therefore present very differently from one person to another. A recent study has estimated that at least 65 million people globally have long-term effects – involving organs, skin and the circulatory system as well as muscles and the brain. The question therefore extends far beyond a single virus: are these long-term effects something specific to SARS-CoV-2, or are they traces left by severe infection?
Researchers at Rigshospitalet, the Copenhagen Mental Health Centre and the University of Copenhagen have therefore delved into Denmark’s health registries. There, they tracked new diagnoses after COVID-19 across the population and compared the pattern with what occurs after other infections. Because the answer is not only about long COVID but about serious infections more broadly.
The study was led by Michael E. Benros, Professor and consultant at the Mental Health Centre Copenhagen, and Daniel Kondziella, Professor and consultant in neurology at Rigshospitalet. Clara S. Grønkjær, a statistician and PhD student on the project and first author, carried out the data analysis. The results have been published in Nature Communications.
“The conclusion is that the risk of long-term effects increases with the duration and severity of COVID-19, particularly among those who were so seriously ill that intensive care was needed. But the risk is no higher than the follow-on effects of other infections of comparable severity,” says Michael E. Benros.
The trail began in the brain – but did not end there
When Benros first became interested in COVID-19, the focus was the brain. This was shortly after the pandemic gathered pace in 2020, and he and his colleagues were already working on how infections and the immune system’s response can trigger the nervous system and mental disorders. Could COVID-19 also affect the brain – and if so, was it because of the virus itself or the body’s response to the infection?
The question did not come out of the blue. Some people with severe COVID-19 were clearly mentally affected both during and after their illness. They could be confused, have difficulty remembering things or struggle to concentrate.
This first led him and his colleagues to investigate how COVID-19 differed from influenza and later how severe COVID-19 courses affect the brain.
They examined the patients’ cognition and took samples of cerebrospinal fluid (the fluid that surrounds the brain and spinal cord). Here, they found that the virus very rarely entered the brain. The body’s immune response, by contrast, stood out more clearly.
Because the immune system affects the whole body through tissues and the bloodstream, the question moved from the brain to the rest of the body. If a severe infection in the lungs could send signals on to many organs, it might also leave long-term effects in several places.
If the long-term effects were the result of the body’s overall response, the researchers should also be able to see new diagnoses outside the brain. And if that idea held, the signal might not point to COVID-19 alone but to how hard the infection had affected the body.
Previous studies made the risk difficult to assess
The theory is not new. Internationally, many studies have examined long-term effects after COVID-19, but they have often been based on small or non-representative datasets and different patient groups. This has made the results difficult to compare.
The study in Denmark enabled the researchers to investigate the association on a much larger scale and from several angles at once.
Clara S. Grønkjær linked polymerase chain reaction (PCR) tests for SARS-CoV-2 of Danes with information on new hospital diagnoses, vaccinations, prescriptions for anti-infective medicines and hospitalisation for other infections. This enabled the researchers to examine whether the risk followed the test result, vaccination status or the severity of the illness – and whether the same picture appeared outside COVID-19.
First, the researchers looked at the COVID-19 course itself. They compared people who had tested positive for SARS-CoV-2 with people who had tested negative, had never been tested or had been prescribed medicine for other infections. This was to show whether the SARS-CoV-2 infection itself was followed by more new diagnoses – or whether the same picture could also be seen after other infectious diseases.
They then gradually turned up the severity of the illness in the analysis: from a positive test to hospitalisation and intensive care. They also considered vaccinations, virus variants and changing testing patterns during the pandemic. This showed whether new diagnoses mainly followed the infection itself or whether the risk only really rose in the severe COVID-19 courses.
The next step was to investigate whether the pattern was specific to COVID-19. Grønkjær therefore compared the disease with other infections at roughly the same level: milder COVID-19 courses were compared with infections treated with medicine, and severe courses were compared with infections that had also led to hospitalisation.
In this way, disease courses at roughly the same level were compared with one another. This gave a more accurate comparison than simply contrasting COVID‑19 with healthy people or those who had tested negative.
“It enabled us to see whether the pattern after COVID-19 differed from what we see after infections of comparable severity,” says Clara S. Grønkjær.
Not infection alone but severe illness drove up the risk
To compare the groups, the researchers used Cox analysis. This produces a hazard ratio, which shows whether a new diagnosis occurs more often in one group than in another over time. If the hazard ratio is around 1, the difference is small or absent; if it is above 1, the risk is higher. The analysis also considered factors such as age, sex, pre-existing illness, income, education and employment.
The figures first showed that a positive COVID-19 test in itself did not move the risk noticeably. For other physical diseases – what researchers call somatic diseases – the hazard ratio was 1.01 when the researchers compared those who tested positive with those who tested negative. For mental disorders, they also found no clinically relevant increased risk.
There were, however, more limited signals. For somatic diseases, the risk was elevated before Omicron and among people who had received fewer than three vaccine doses. Among unvaccinated people, the hazard ratio was 1.12; among those with one vaccine dose, it was 1.14; and among those with two doses, it was 1.05. Among people with three or more doses, the figure was 0.99 – again, around the same level as among those who tested negative. Here too, the figures pointed towards the phase of the pandemic when COVID-19 more often led to severe illness.
Here, the pattern became clear. The strongest dividing line was the severity of the illness. People who had been hospitalised with COVID-19 had a substantially higher risk of new diagnoses than people who had not been hospitalised. For mental disorders, the hazard ratio was 1.88; for somatic diseases, it was 2.53. Among people who had been in intensive care, the figures rose to 2.49 and 5.09, respectively.
But the crucial question was what happened when COVID-19 was compared not with healthy people or those who had tested negative but with other serious infections. Here, the picture changed: among hospitalised people, the increased risk was at the same level as after other infections – including serious lung infections.
The analysis pointed back to the question that had driven the study: severe COVID-19 may be followed by new diagnoses, but the pattern resembles what researchers see after other serious infections of comparable severity, says Grønkjær.
The findings may change who the healthcare system monitors after infection
The story ends in a different place from where it began. Across test results, vaccinations, hospitalisation and other infections, the study points away from COVID-19 as a special case and towards a broader problem: how people with serious infections can be treated and followed so that the acute illness does not initiate new problems afterwards.
When long-term effects emerge especially after the most severe courses, prevention becomes the obvious next step. The point is not that COVID-19 is harmless but that people with other serious infections may also need attention afterwards.
“One area that deserves greater focus and research is this general impact on organs after severe infections,” says Michael E. Benros, adding that the next question is whether these lasting consequences can be prevented.
The study also has implications for healthcare planning. More precise risk assessments can help the healthcare system understand who has the greatest risk of long-term effects, who should be followed more closely after discharge and how resources can be allocated in the event of a new pandemic.
In the longer term, the research may lead right back to the acute illness itself: can doctors intervene in the processes that damage the body – for example, by dampening an inappropriate immune response while the person is still being treated for the infection? The next trail is not only about following people afterwards but about understanding what happens while the body is still fighting.
“The registry study primarily shows associations. The next step is to understand the biological mechanisms behind the long-term effects – that is, what happens in the body after a serious infection – so that we can point more precisely to the molecular processes future treatments should target. With that knowledge, we can begin to investigate whether targeted treatment can prevent long-term effects across infections,” says Michael E. Benros.
