The infections belong to the group paediatricians call TORCH, an acronym for maternal infections that can reach and harm the fetus. The study covers four, cytomegalovirus, Toxoplasma gondii, rubella virus and the herpes simplex viruses. These are infections diagnosed by a doctor and entered in a register. Infections that went unnoticed, at birth or later, are not counted.
The study was published in JAMA Pediatrics on 21 September 2026. The exposed were followed for an average of 18.7 years, the others for 20.5, until death, emigration or the end of 2023. Results are given as hazard ratios, the ratio between the rates at which a diagnosis appears in two groups followed over time. A value of 1 means no difference.
For autism the hazard ratio is 3.10, with a 95 per cent confidence interval from 2.55 to 3.76. That is the «triple» in the ScienceDaily headline. The figure splits into two different parts. For autism with intellectual disability the hazard ratio rises to 6.23. For autism without it, it falls to 1.97. The link is strongest where there is also cognitive damage.
For intellectual disability the overall hazard ratio is 7.22. It grows with severity, from 3.01 for mild forms to 23.51 for severe to profound ones, with an interval from 18.06 to 30.60. The Karolinska press release speaks of a risk up to 30 times higher. Thirty is the upper edge of the interval, not the estimate.

Hazard ratios, with 95 per cent confidence intervals, for autism and intellectual disability in people born in Sweden between 1987 and 2021 with a diagnosed congenital TORCH infection (975) compared with the unexposed; orange marks sibling comparisons. Log scale, the vertical line at 1 marks no difference. Chart produced by ScienceOnline from data in Sjöqvist H. et al., JAMA Pediatrics 2026, doi 10.1001/jamapediatrics.2026.4229 (data from the publisher’s abstract). Original newsroom chart.

The authors repeated the analysis among full siblings, one exposed and one not. The comparison removes what a family shares, part of the genes, the home, income, habits. Autism stays at 3.19, with a wide interval from 1.97 to 5.15. For intellectual disability it rises to 11.28, from 5.97 to 21.33. Wide intervals are the mark of few cases.
For obsessive-compulsive disorder the study finds no consistent association. For attention deficit hyperactivity disorder an association appears in the population and weakens in the sibling comparison. The press release sums up both as no clear link. At 16, exposed children had school grades 1.5 points lower on the Swedish standardised scale.
The abstract does not say how many of the 975 children were diagnosed with autism. Without that number the absolute risk is missing from the study as read. An external figure gives a sense of scale. In Italy, according to the Istituto Superiore di Sanità, the national health institute, autism affects one child in 77. Three times that would be roughly one in 25. This is a newsroom calculation, made on another country and treating a hazard ratio as a risk ratio, and it only gives a scale.

The full text was behind a paywall. The figures in this article come from the structured abstract the publisher deposited in bibliographic registries. Funding, conflicts of interest, figures for each pathogen and the limitations stated by the authors were not read.
The authors point to preventing infections passed on in pregnancy. The abstract does not break the results down by pathogen, and the weight of each remains to be read in the full text. For rubella there is a vaccine. The press release notes that these infections explain a small share of autism cases. The children with a diagnosed TORCH infection, among 3,666,002 births, numbered 975.

References
Sjöqvist H., Dalman C., Mataix-Cols D., Gardner R.M., Karlsson H., «Congenital TORCH Infections and Neurodevelopmental Outcomes», JAMA Pediatrics, published online 21 September 2026 (volume, issue and pages not given in the metadata consulted; licence not stated). doi: 10.1001/jamapediatrics.2026.4229
Istituto Superiore di Sanità, National Observatory for the Monitoring of Autism Spectrum Disorders, institutional page (figure «one child in 77 in Italy», ISS press release of 1 April 2019), osservatorionazionaleautismo.iss.it.
Karolinska Institutet, press release, republished by ScienceDaily on 23 September 2026 as «Rare pregnancy infections may triple the risk of autism» (secondary source).