Ambient air pollution exposure and full-term birth weight in California

Abstract

Background: Studies have identified relationships between air pollution and birth weight, but have been inconsistent in identifying individual pollutants inversely associated with birth weight or elucidating susceptibility of the fetus by trimester of exposure. We examined effects of prenatal ambient pollution exposure on average birth weight and risk of low birth weight in full-term births.

Methods: We estimated average ambient air pollutant concentrations throughout pregnancy in the neighborhoods of women who delivered term singleton live births between 1996 and 2006 in California. We adjusted effect estimates of air pollutants on birth weight for infant characteristics, maternal characteristics, neighborhood socioeconomic factors, and year and season of birth.

Results: 3,545,177 singleton births had monitoring for at least one air pollutant within a 10 km radius of the tract or ZIP Code of the mother's residence. In multivariate models, pollutants were associated with decreased birth weight; -5.4 grams (95% confidence interval -6.8 g, -4.1 g) per ppm carbon monoxide, -9.0 g (-9.6 g, -8.4 g) per pphm nitrogen dioxide, -5.7 g (-6.6 g, -4.9 g) per pphm ozone, -7.7 g (-7.9 g, -6.6 g) per 10 microg/m3 particulate matter under 10 microm, -12.8 g (-14.3 g, -11.3 g) per 10 microg/m3 particulate matter under 2.5 microm, and -9.3 g (-10.7 g, -7.9 g) per 10 microg/m3 of coarse particulate matter. With the exception of carbon monoxide, estimates were largely unchanged after controlling for co-pollutants. Effect estimates for the third trimester largely reflect the results seen from full pregnancy exposure estimates; greater variation in results is seen in effect estimates specific to the first and second trimesters.

Conclusions: This study indicates that maternal exposure to ambient air pollution results in modestly lower infant birth weight. A small decline in birth weight is unlikely to have clinical relevance for individual infants, and there is debate about whether a small shift in the population distribution of birth weight has broader health implications. However, the ubiquity of air pollution exposures, the responsiveness of pollutant levels to regulation, and the fact that the highest pollution levels in California are lower than those regularly experienced in other countries suggest that precautionary efforts to reduce pollutants may be beneficial for infant health from a population perspective.

Figure 1

Difference in birth weight in grams associated with full pregnancy gaseous pollutant exposures for births within 10 km monitor distance, single and two-pollutant linear models (95% confidence interval). Results displayed in the figures are controlled for infant's sex, gestational age, calendar year of birth, season, maternal educational attainment, age, marital status, race/ethnicity, country of birth and parity, adequacy of prenatal care, an indicator variable reflecting common medical risk factors, and neighborhood poverty rate, owner occupancy, low education rate, and unemployment rate.

Figure 2

Difference in birth weight in grams associated with full pregnancy particulate matter exposures for births within 10 km monitor distance, single and two-pollutant linear models (95% confidence interval). Adjustments as in Figure 1.

Figure 3

Difference in birth weight in grams associated with full pregnancy gaseous pollutant exposures for births within 10 km monitor distance, stratified by neighborhood level poverty rate (95% confidence interval). Adjustments as in Figure 1.

Figure 4

Difference in birth weight in grams associated with full pregnancy particulate matter exposures for births within 10 km monitor distance, stratified by neighborhood level poverty rate (95% confidence interval). Adjustments as in Figure 1.

Figure 5

Difference in birth weight in grams associated with full pregnancy gaseous pollutant exposures for births within 10 km monitor distance, stratified by maternal race and ethnicity (95% confidence interval). Adjustments as in Figure 1.

Figure 6

Difference in birth weight in grams associated with full pregnancy particulate matter exposures for births within 10 km monitor distance, stratified by maternal race and ethnicity (95% confidence interval). Adjustments as in Figure 1.