Abstracts: ISEE 21st Annual Conference, Dublin, Ireland, August 25-29, 2009: Poster Presentations
Background and Objectives:
Ambient air pollution has been associated with preterm delivery and restricted fetal growth but relatively little attention has been focused on intermediate endpoints that may reflect the underlying biologic mechanisms. Particulate matter (PM) exposure has been associated with hypertension and cardiovascular events. Preeclampsia is a serious hypertensive disorder of pregnancy that can occur on the pathway to preterm birth. This analysis evaluates the acute impact of PM exposure on preeclampsia complicating delivery in the United States (US).
EPA air quality data has been linked to the National Hospital Discharge Survey (NHDS) for 1999–2005. The NHDS is a national probability survey representing inpatient hospitalizations from non-Federal short-stay hospitals in the US. Approximately 33% of delivery hospitalizations (unweighted 83,164/251,957) have air monitoring data for at least one pollutant within a five-mile radius based on zip code of residence. For the acute time-period before delivery, we calculated both a seven-day and six-week average PM10 and PM2.5 exposure for each delivery admission. Quartiles of exposure were based on the distribution for deliveries without preeclampsia.
No significant associations were observed between 6-week average PM10 or PM2.5 and risk of preeclampsia. No dose-response was observed but third quartile 7-day average exposure was significantly elevated for both PM10 (Odds Ratio [OR] = 1.16, 95% Confidence Interval [CI]: 1.03–1.30) and PM2.5 (OR = 1.15, CI: 1.04–1.28). These results were essentially unchanged after adjustment for maternal age, race and region of the country.
Preeclampsia risk may be increased by ambient PM exposure in the week preceding delivery but the lack of a dose-response suggests other unmeasured factors may be important. The NHDS has limited demographic data so we plan to evaluate this further by linking to contextual data on income and other factors that may vary with geography.