select search filters
briefings
roundups & rapid reactions
Fiona fox's blog
A study published in The Journal of Physiology looks at associations between prenatal air pollution exposure and neurodevelopmental outcomes in childhood.
Prof Roy Harrison FRS, Professor of Environmental Health, University of Birmingham, said:
“There is now very strong evidence that air pollution affects both short-term and long-term cognitive function across all age ranges. This is a well planned and executed study of almost 500 toddlers whose expectant mothers’ exposure to nitrogen dioxide and airborne particulate matter was evaluated together with measures of cognitive development in the children. It clearly demonstrates a negative association between air pollution exposure of the pregnant women during their first trimester and the cognitive development of the developing foetus, even after controlling for the many potentially confounding factors. This comes as no surprise as air pollution is already known to be associated with premature deliveries and low birth weight, and with cognitive deficiencies in other age ranges. Our own research has estimated that air pollution exposure is causing a collective loss of around 65 billion IQ points across the global population (1), providing further evidence of the massive benefits of air pollution abatement for public health.”
1 – Reframing air pollution as a cognitive and socioeconomic risk, T. Faherty, L.-J. A. Ellis-Bradford, H. Onyeaka, R.M. Harrison & F.D. Pope, npj Clean Air (2026) 2:18, https://doi.org/10.1038/s44407-026-00059-4
Prof Kevin McConway, Emeritus Professor of Applied Statistics, Open University, said:
“The results of this new study do give some cause for concern and add to the considerable amount of evidence on possible associations between air pollution and infant and child health. But the study has some important limitations, so it certainly can’t establish everything that might be going on.
“I’ll start off by emphasising the general importance of findings like this. If you read the research paper in detail, you’ll see that the researchers describe the size of the associations they found as ‘small’ or ‘very small’. This is based on a quite common statistical way of describing these sizes. But please don’t think this means that they do not matter. As the researchers make clear, everyone in an area of high air pollution is exposed to the pollution. Air pollution in London has decreased hugely since the smogs of the 1950s and indeed since the belching diesel fumes of a few decades ago. But in the short term, the only way to escape high air pollution entirely would be to move to somewhere much less polluted, and that’s impossible for most people. So even a ‘very small’ effect can add up over the whole population.
“It’s also revealing to look at the ranges of pollution levels, during pregnancy in the mothers of the toddlers in the study. Their ranges are given towards the end of Table 1 in the research paper. For PM2.5 (very fine particulate matter), the lowest levels recorded for any of the participants, across pregnancy, were about 8.5 units on the scale used (µg/m3). The World Health Organisation’s 2021 guideline for annual average PM2.5 is 5 on this scale. The current UK legal limit (annual average) for PM2.5 is 20, and the upper ends of the ranges in Table 1 indicate that some of the PM2.5 records for these mothers were above that. The legal target for annual average PM2.5 by 2040, set in the UK by the 2021 Environment Act, is 10 µg/m3, and most (though not all) of the PM2.5 levels for mothers in this study are well above that still.
“The position on larger particulates (PM10) isn’t quite as bad as for PM2.5, but still not good. And on nitrogen dioxide (NO2), where the WHO guideline is 10 µg/m3, every mother in this study was exposed above this level. Indeed (since the upper ends of the ranges for NO2 in the study are in the 60s and 70s), it’s clear than many were exposed above the current UK legal limit of 40 µg/m3.
“This isn’t the first study to have found a correlation between exposure to higher levels of air pollution before birth and measures of early childhood development. But previous studies didn’t all agree with one another on which were the critical time periods during pregnancy when air pollution was associated with development. One important feature of the new findings is that they throw further light on this question of timing.
“The conclusion was that the only clear statistical evidence of an association between language development in toddlers and their mothers’ exposure to high air pollution levels was for exposure during the first three months of pregnancy. There remains a certain amount of uncertainty about whether that association is one of cause and effect (see below).
“Another key feature is that the new study throws some light on whether the level of correlation between exposure to air pollution in the womb depends on whether the child is eventually born preterm (prematurely).
“The study found evidence that motor (movement) development at the age of eighteen months wasn’t correlated with air pollution levels in toddlers who were not born preterm. But in toddlers born preterm, the higher the level of air pollution that their mothers were exposed to (throughout pregnancy), the lower their motor development scores were. This became more apparent in toddlers who were born very prematurely. A difficulty with this finding, as I’ll explain further below, is that we can’t be entirely sure to what extent the differences in developmental outcomes are actually caused by the pollution, but it certainly remains possible that they are.
“Looking at preterm births is important in this context, because much previous research in many countries (including in London) has found associations between high air pollution levels during pregnancy and a higher-than-average risk of preterm birth. So, the question arises as to whether any association between development scores in toddlers and air pollution during pregnancy might arise because high air pollution is associated with more preterm births, and that preterm birth is in turn associated with differences in average development scores. The new findings throw some light on that question though not on all aspects of it.
“This new research, and the great majority of previous research on associations between air pollution and child health, is observational. It isn’t an experiment where mothers were assigned at random to live in areas with different pollution levels – that would be unethical and entirely impractical. But in all observational research of this kind, there remains some doubt as to whether any association between pollution and a health outcome is one of cause and effect, or whether other factors are the real cause.
“The researchers for this new study made a pretty good job of allowing for many possible other factors, as far as they could, and I think the statistical analysis is generally of high quality – but we still can’t be completely sure what’s causing what. In fact, the researchers specifically mention that they could not take into account all the many factors that are associated with preterm births.
“The researchers also list several other limitations to what their study can show.
“The air pollution levels during pregnancy, at the place where the mothers lived, was not measured directly but came from a statistical model of pollution levels in individual postcodes. The model used is highly developed, using detailed data on things like traffic levels, and its quality is very good – but it can’t include data on the mothers’ exposure to air pollution inside their homes, or in other places where they went apart from their home postcode, and it can’t fully take into account seasonal differences in pollution levels. (Some studies have used GPS tracking of where people go, and personal meters that record pollution levels where they are, but that’s expensive and complicated to do.)
“The new research used data from an existing study called dHCP (the Developing Human Connectome Project), which provided data on 498 toddlers, all from greater London. Of these toddlers about a quarter (125 of them) were born preterm. That’s far more than the number of preterm births that one would expect in a representative sample of 498 London toddlers, which would be roughly 40. “Making sure that the dHCP sample contained more preterm births than would represent the London population makes sense, since the data were collected to study development, which proceeds (on average) differently in preterm babies.
“But this does emphasise that the toddlers in the sample are not typical, on average, of London toddlers, to say nothing of toddlers elsewhere. It may very well be the case that similar results would have been found in a more representative sample of London mothers and toddlers, but we can’t be sure of that. The paper’s authors write, “Further multicentre studies, incorporating unprecedented sample sizes, are required to fully elucidate the impact of prenatal air pollutant exposure on neurodevelopmental outcomes.”
Prof Jonathan Grigg, Professor of Paediatric Respiratory and Environmental Medicine, Queen Mary University of London (QMUL), said:
“Previously, an association between maternal exposure to air pollution and reduced head circumference at birth has been reported, but whether this was associated with long-term changes in brain function was unclear. In the present study, researchers assessed both exposure of pregnant women using a robust pollution model and cognitive function in their offspring when aged 17 to 36 months. The association between increased exposure to air pollution during early pregnancy and lower language scores is indeed compatible with a long-term effect of fetal exposure to air pollution on subsequent cognitive function. Biological plausibility of this finding is provided by recent studies showing that inhaled pollutant particles both enter the blood stream and accumulate in placental cells.”
Prof Jill Belch, Professor of Vascular Medicine, University of Dundee, said:
“The findings from this group are consistent with what we already know about how fine particles can trigger inflammation and affect the developing foetus, and its blood supply through the placenta. This appears to be a well-conducted study using a large dataset and established methods. The conclusions are supported by the data, though the effect sizes are small.
“Importantly, even small shifts at a population level can translate into meaningful impacts across society. This study reinforces the need to reduce air pollution exposure, particularly for pregnant women and other vulnerable groups.”
‘Prenatal air pollution exposure is associated with altered neurodevelopmental outcomes in early childhood’ by Alexandra F. Bonthrone et al. was published in The Journal of Physiology at 00:01 UK time on Wednesday the 29th of April 2026.
DOI: 10.1113/JP290327
Declared interests
Prof Roy Harrison: “He is a member of the Defra Air Quality Expert Group and the DHSC Committee on the Medical Effects of Air Pollutants. He has research funding from UKRI, Defra and the EU Horizon Programme and none from industry.”
Prof Kevin McConway: “No conflicts of interest to declare”
Prof Jonathan Grigg: “No declarations of interest”
Prof Jill Belch: “Professor Belch’s work is locally funded by charities. She has previously published research that has found that air pollution prompts spikes in child hospital admissions.”