"Many studies in São Paulo have already shown that air pollution could affect adult and child health," researcher Nelson Gouveia told environmentalresearchweb. "Traffic is the main source of air pollution in our city. The idea of this study was to show the same association but using an indicator of traffic density."

Gouveia and colleagues examined data for 14 districts of São Paulo, accounting for around 23% of the city's population and 44% of its area over a six-month period. They analysed factors for 313 babies who survived, 162 babies who died during pregnancy and 146 babies who died within a week of being born.

The team found that the risk of early death went up as exposure to traffic-related air pollution increased. Infants from mothers exposed to traffic levels in the top 25% of measurements had a 50% higher risk of dieing within their first week. The researchers made allowances for other factors that may affect mortality, such as socioeconomic conditions. The link between traffic density and foetal mortality was less clear-cut.

Earlier studies have found a link between infant health and pollution measured by air quality monitoring stations. But according to Gouveia, such stations collect pollution data over a relatively large area and do not take into account exposure to roads with heavy traffic. People living closer to busy roads may have greater exposure to toxic compounds released in vehicle exhaust or formed near the road. With this in mind, the researchers calculated the distance-weighted traffic density near the mothers' homes, based on their address, distance to surrounding streets up to a radius of 750 feet (228.6 m) and the traffic flow in those streets.

Air pollution could be affecting infant health by causing an effect on the placenta, embryo, mothers' immunologic system, ovarian-hypothalamic axis and/or the induction of intrauterine growth restriction (IUGR), which can lead to a more vulnerable foetus, say the researchers. Some toxic compounds in exhaust fumes, such as polycyclic aromatic hydrocarbons (PAHs) may be responsible, as may air particles, which can change plasma viscosity, the inflammation process and blood pressure, and carbon monoxide, which can cross the placental barrier and reduce oxygen transportation in the blood.

The team says that although it could not provide strong evidence of association between early infant death and exposure to pollution from traffic, the consistent literature and the biologic plausibility indicate that motor vehicle exhaust exposures may be important.

"These results were discussed with policy makers and will possibly be used in planning actions to abate air pollution in our cities," said Gouveia. "I'm now starting a similar study looking not at deaths but at other pregnancy outcomes such as birthweight, prematurity and perhaps birth defects in relation to traffic density, since these conditions are also related to maternal exposure to air pollution."

The researchers reported their results in Environmental Health Perspectives.