Air pollution is not a single substance but a mixture of particles from different sources. A new study published in the Journal of Environmental Psychology examined how six such sources are linked to features of brain function in children. It turned out that exhaust, coal dust and fertilizers have their own, albeit weak, patterns of association with neural networks. But the authors stress that these are statistical associations, not a proven cause.
The focus is on fine particulate matter PM2.5 (from the English particulate matter), that is, particles suspended in the air smaller than 2.5 micrometers. They are so small that they penetrate deep into the lungs and can enter the bloodstream, and from there the brain.
The brain coordinates its work through large functional networks - groups of distant regions whose activity changes synchronously. This synchrony is measured by scanning: a person lies without a task while the device tracks changes in blood flow. In childhood and adolescence, such networks are reorganized as thinking, sensory processing and emotional control develop.
The study relies on data from the Adolescent Brain Cognitive Development Study, a nationwide program that follows thousands of American children. The analysis included scans of 6,291 children aged 9-11, as well as 1,972 participants rescanned about two years later, at ages 11-13. The scientists assessed pollution exposure by each child's home address for 2016 using air quality models. Then, using machine learning, they looked for links between pollution source and the activity of 12 brain networks.
The greatest exposure to exhaust was associated with weaker connectivity of the visual and motor networks, as well as the auditory network and the cognitive control network (which helps keep attention on a task). Exposure to ammonium sulfate (mainly from coal burning), by contrast, was associated with stronger connectivity of sensory systems with higher-order networks, including the default mode network (which is active during daydreaming and self-reflection).
Over two years of observation, stable patterns of change emerged for three sources. Exhaust was associated with increased connectivity between the hand movement and attention networks, but decreased connectivity between the auditory and control networks. Smoke from fires and biomass burning was associated with strengthened connectivity within attention networks. Ammonium sulfate was associated with weakened connectivity between the attention network and the hand motor network.
However, the effect size turned out to be modest. Even the identified patterns explained less than 1% of the differences in exposure levels at ages 9-11 and no more than 2% in the dynamics. When tested on children from other regions, the models did not predict better, and often predicted worse, than a simple mean. The largest errors were in the western United States and where pollution levels varied greatly between children.
The authors conclude that after adjusting for other factors, pollution sources probably play only a small role in differences in brain connectivity between children, and the results may not transfer to other regions. The patterns for biomass and ammonium sulfate proved most stable, while data on exhaust and soil dust should be treated more cautiously.
The study is observational: it does not prove that inhaling particles directly causes changes in the brain. In addition, the scientists did not adjust for multiple comparisons, so some of the associations found may be false positives. Exposure was assessed only by address for one year, without accounting for filters, ventilation and time spent outdoors. The sample was somewhat more affluent and white than the US population on average, and average pollution levels were below federal standards.