Chronic exposure to the intermittent roar of airliners can lead to deficits in children's reading skills and long-term memory, according to a study published in the September issue of Psychological Science (Vol. 13, No. 9).

Previous cross-sectional studies have linked aircraft noise to poor classroom performance, but until now there have been no prospective longitudinal studies on the problem. The current study provides some of the first strong evidence that chronic noise is a cause, not just a correlate, of impairments in reading and memory.

To conduct the study, Staffan Hygge, PhD, of the University of Gävle in Sweden, and his co-authors took advantage of the closing of an old Munich airport and the opening of a new one. Six months before, 12 months after and 18 months after the switch, they administered tests of reading, memory, attention and hearing to children ages 8 to 12 who lived near the airports.

They found that the reading skills and long-term memory of children near the old airport improved once air traffic moved to the new airport, while the performance of children near the new airport declined. Children chronically exposed to the noise also suffered from higher blood pressure, higher levels of stress hormones and more psychological stress than children in quieter environments.

Hygge and colleagues also found that noise-exposed children are relatively insensitive to speech, even though their hearing is unimpaired. However, contrary to some previous research, they found no connection between speech perception and reading: After the old airport closed, children living near it showed marked improvements in reading and memory, but their speech perception remained impaired.

According to Cornell University psychologist Gary Evans, PhD, one of the study's authors, the failure of speech perception to mediate the effect of noise on reading is intriguing, but it requires further research.

"Language acquisition is a pretty complex and well-developed area, and we just took one measure from it," he says. "One possibility is that speech might not, in fact, be the mediator, but another possibility is that the measure we used just wasn't sensitive."