Congenital TORCH infection and autism: rare association, careful counselling

Most autism is not caused by an infection in pregnancy. A large Swedish study published in September 2026 does, though, matter for a small group of children: those born with a confirmed congenital infection. It also shows how to talk about risk without overstating cause.

What TORCH means

TORCH is a short label for infections that can pass from a pregnant woman to her baby before birth. The letters stand for Toxoplasma gondii, other infections such as syphilis, rubella, cytomegalovirus (CMV), and herpes simplex virus (HSV).

These infections are uncommon in newborns. When they do reach the fetus, they can affect how the brain develops. The link with intellectual disability has been known for a long time. The link with autism has been less clear, because earlier studies were small.

What the study did

Hugo Sjöqvist and colleagues at Karolinska Institutet used Swedish national health registers (1). They included 3,666,002 people born in Sweden between 1987 and 2021, followed up to the end of 2023. Of these, 975 had a recorded diagnosis of congenital TORCH infection.

The researchers compared these children with the rest of the population. They also compared them with their own brothers and sisters who did not have the infection. A sibling comparison helps to check whether things shared within a family, such as genes, home environment, or income, could explain the findings on their own.

What they found

Compared with children without a recorded congenital TORCH infection:

Autism was diagnosed about three times as often (hazard ratio 3.10). The sibling comparison gave a similar figure (3.19).
Intellectual disability was diagnosed about seven times as often (hazard ratio 7.22). In the sibling comparison the figure was higher (11.28).
The more severe the intellectual disability, the stronger the link. It was about three times as common for mild intellectual disability, and more than 20 times as common for severe to profound intellectual disability.
The link was much stronger for autism with intellectual disability (hazard ratio 6.23) than for autism without intellectual disability (hazard ratio 1.97).
A small link with ADHD (attention deficit hyperactivity disorder) disappeared in the sibling comparison. There was no consistent link with obsessive-compulsive disorder (OCD).
Teenagers who had the infection but no autism or intellectual disability diagnosis had slightly lower school grades at age 16, on average.

A hazard ratio compares how often something happens in one group with how often it happens in another. A hazard ratio of 3 means about three times as often. It does not tell you how many children in each group are affected.

CMV made up about half of the recorded TORCH cases. It had a strong link with autism (hazard ratio 4.57) and with intellectual disability. Rubella has almost disappeared in Sweden since vaccination began, so there were few rubella cases in the recent birth years.

The number that matters for counselling

A three-fold or seven-fold risk sounds large. But congenital TORCH infection is rare, so it explains very few cases of autism across the whole population.

The researchers estimated that congenital TORCH infection accounts for about 0.034% of autism cases in Sweden. That is roughly 1 in 3,000 autism cases. It accounts for about 1.2% of cases of severe intellectual disability. This figure is called the population attributable fraction: the share of all cases in a population that can be linked to one exposure.

They also estimated that about one in five children born with a congenital TORCH infection may later receive an autism diagnosis. So, for a child who actually had the infection, the extra risk is real. For explaining autism in general, these infections explain almost none of it.

Association is not the same as cause

This is an observational study. It shows that two things happen together more often than expected. It cannot prove that the infection caused the autism in any one child.

The sibling comparison makes a shared family explanation less likely. That fits with what we already know: these infections can damage the developing brain directly. But there are limits:

The study only counted infections that were diagnosed and recorded. Many congenital CMV infections are never tested for, so milder cases will have been missed.
The number of exposed children was small (975), so some estimates, especially in the sibling comparison, are imprecise.
Sweden is not the UK. Health services, screening, and vaccination rates differ.
A raised risk is not a certainty. Most children with a congenital TORCH infection in this study did not have an autism diagnosis. Most autistic children have never had a congenital TORCH infection.

What this means in clinic

For a child with a confirmed congenital infection, such as congenital CMV:

Be honest that the chance of developmental difficulties is higher than for other children. This includes intellectual disability, and to a lesser degree autism.
Plan developmental follow-up on clinical grounds: hearing, vision, motor skills, language, and learning. Do not wait for an autism diagnosis to offer support.
If autism is diagnosed, the infection may well be part of the explanation. It is still worth considering the usual assessment, including genetic testing where the child meets NHS criteria.

For families of autistic children without a known congenital infection:

This study does not suggest that a hidden infection in pregnancy explains most autism. The numbers point the other way.
Parents sometimes ask whether an illness in pregnancy caused their child’s autism. For common infections, a large earlier Swedish study found that the link with autism largely disappeared when siblings were compared (2). The rare infections in this new study are different, because they infect the baby directly.
Nothing in this study suggests that a mother did something wrong.

Prevention

Some of these infections can be prevented. This is a public health point. It is not a claim that infection is a major cause of autism.

Rubella: the MMR (measles, mumps and rubella) vaccine before pregnancy. Congenital rubella is now very rare in the UK because of vaccination.
Syphilis: screening is offered to every pregnant woman in the NHS, and treatment with antibiotics protects the baby.
CMV: there is no vaccine, and the NHS does not screen for it routinely in pregnancy. Hand washing, and avoiding contact with young children’s saliva and urine (for example, not sharing cups or cutlery), lowers the risk.
Toxoplasmosis: avoid raw or undercooked meat, wash fruit and vegetables, and wear gloves for gardening or cat litter.
Herpes simplex: tell the midwife or doctor about any history of genital herpes. Treatment in late pregnancy, and sometimes a caesarean section, lowers the risk to the baby.

What this means for families

Congenital TORCH infection is uncommon. When it is confirmed, there is a clearly higher risk of intellectual disability and a higher risk of autism, especially autism with learning disability. The child should have careful developmental follow-up.

At the same time, these infections account for only a tiny fraction of autism overall. Counsel the family in front of you about their child. Do not use one rare exposure to explain autism in general.

References

Sjöqvist H, Dalman C, Mataix-Cols D, Gardner RM, Karlsson H. Congenital TORCH infections and neurodevelopmental outcomes. JAMA Pediatrics. Published online 21 September 2026. doi:10.1001/jamapediatrics.2026.4229
Brynge M, Sjöqvist H, Gardner RM, et al. Maternal infection during pregnancy and likelihood of autism and intellectual disability in children in Sweden: a negative control and sibling comparison cohort study. Lancet Psychiatry. 2022;9(10):782-791.

Dr Odet Aszkenasy is a Consultant Community Paediatrician and the author of The Genetics of Autism: A Guide for Parents and Professionals.