Science1 publisher3 min readPublished Updated
Congenital TORCH infections triple autism risk across 3.7 million Swedish birth records
Karolinska researchers traced 975 congenital infections through a cohort born from 1987 to 2021 and found a thirtyfold risk of severe intellectual disability, alongside an estimated 1.2 percent share of all such cases in Sweden.
The Scientist · Science desk

What happened
- Researchers at Karolinska Institutet examined health and education records for 3.7 million people born in Sweden between 1987 and 2021, publishing the results in JAMA Pediatrics.
- Within that cohort, 975 people had a diagnosed congenital TORCH infection, and participants were followed for as long as three decades.
- The infected group was about three times more likely to be diagnosed with autism, more than seven times more likely to have an intellectual disability, and up to 30 times more likely to have severe to profound intellectual disability.
- Comparing infected children with their own uninfected siblings left the pattern similar, which the authors read as evidence that shared family factors alone do not account for it.
Compiled by The ScientistSomething wrong?How this is made
Why it matters
- constraint With 975 exposed people spread over 35 birth years, the study can size the risk for the TORCH group as a whole, but it cannot tell a clinician whether congenital cytomegalovirus carries the same risk as toxoplasma.
- capability The sibling design narrows what an explanation can be. Whatever drives the autism association has to be specific to the affected pregnancy; shared family background no longer covers it.
- decision A health ministry weighing screening reads 1.2 percent of severe intellectual disability cases. The parent of an affected newborn reads a one-in-five chance of an autism diagnosis. Both come out of the same dataset.
- exposure Lower average grades turn up even in infected children who were never diagnosed with autism or an intellectual disability, a group that falls outside the categories which usually trigger educational support.
975 diagnosed congenital infections in a cohort of 3.7 million works out to about one in 3,800 births [18]. Every number in the study is anchored to those 975 people, spread across 35 birth years of Swedish health and education records [2] [22]. The TORCH label covers cytomegalovirus, rubella, toxoplasma and herpesvirus, pathogens that can sometimes cross the placenta and infect the fetus [7].
The absolute risk is the figure most useful to a parent. The researchers estimate that roughly one in five children born with a TORCH infection may later develop autism [10]. Set that against the threefold relative risk [4] and the comparison group's rate comes out near one in fifteen [19]. That 6.7 percent comes from dividing two published numbers. The Karolinska release does not give the figure itself. One in five of 975 is also about 195 children [20], the numerator behind the autism estimate. The release also leaves out confidence intervals and any breakdown of the 975 cases by individual pathogen.
The Karolinska team also compared infected children with their own siblings who had not been infected, and the pattern held. The authors take that as evidence that shared family characteristics alone are unlikely to explain the associations [11].
That comparison strips out the genetics and the household environment two siblings share. It does not strip out whatever else was particular to the affected pregnancy, and it does not strip out the extra developmental follow-up a newborn identified with a congenital infection receives. Here the null result helps: the team found no clear link between congenital TORCH infection and ADHD or obsessive-compulsive disorder [12]. Closer surveillance that inflated every diagnosis would be expected to lift ADHD as well.
At the population level the numbers are small. The researchers estimate congenital TORCH infections are associated with about 1.2 percent of severe intellectual disability cases in Sweden, and about 0.034 percent of autism cases [8] [9]. Renee Gardner, a researcher at the Department of Global Public Health at Karolinska Institutet [16], said: "It is important to emphasize that it is unusual for these infections to be transmitted from mother to child. They account for a very small proportion of all cases of autism in the population, but for those children who are actually affected, we see a clearly elevated risk of both autism and intellectual disability" [14].
Hugo Sjoqvist, a PhD student and the study's lead author [17], said the autism side of the question had been the hard part: "It has long been known that these infections can cause intellectual disability. However, the link to autism has been less clear in previous research, which has often been based on small patient groups. This study is the largest to date in this field and is based on national register data covering almost the entire population of Sweden" [15].
The school records carry a separate finding. Among children with a congenital TORCH infection who were not diagnosed with autism or an intellectual disability, average grades were lower than in children without an infection [13].
What to watch
- The confidence intervals in the JAMA Pediatrics paper, and whether the 30-fold severe intellectual disability estimate rests on enough exposed cases to be stable.
- Any breakdown by pathogen, since cytomegalovirus, rubella, toxoplasma and herpesvirus are pooled under one exposure label.
- Replication in another national register with different congenital infection screening practice, which would test whether the autism association tracks diagnosis intensity.