The Effect of Short and Intensive Exposure to PM10 on Birth Outcomes in Montevideo

Background: Several studies have found associations between exposure to particulate matter during pregnancy and low birth weight. Most research, however, has been conducted in developed countries. The effects may be stronger in less developed countries, where higher fractions of the population face nutrition and health limitations, as well as higher barriers to health care services. Moreover, most investigations have been conducted in highly contaminated cities. Studying short and intensive exposures to particulate matter in less contaminated places may shed light on a barely studied effect of climate change: the effect of intense accumulation of respirable particles during heat waves on human health.

Objective: This study explores the impact of short and intense exposures to air pollution on adverse birth outcomes in Montevideo, Uruguay. We focus on the effect of respirable suspended particles with diameter of 10 micrometres or less per cubic meter (PM10) on the likelihood of a premature birth and low birth weight.  We exploit the fact that during discrete intervals of time in 2011 the ashes and dust resulting from the eruption of the Puyehue volcano in Chile increased substantially the exposure to PM10 in Montevideo.

Data: We obtain data on prenatal and birth characteristics for all pregnancies that took place in Montevideo between 2009 and 2012 from the Perinatal Information System (Uruguayan Ministry of Public Health).  The air quality data come from the Environmental Quality Lab of the Municipal Government of Montevideo.

Methods: We estimate the effects of mother’s exposure to PM10 in each trimester of the pregnancy on perinatal outcomes (low birth weight and preterm birth) adjusting for mothers’ socio-demographic and behavioral characteristics, pregnancy-specific conditions, prenatal care center fixed effects, time trends, and weather conditions. We consider exposure to average levels of PM10 as well as exposure to thresholds above 50 µg/m³ and above 100 µg/m³.

Findings: We find statistically significant adverse effects of high PM10 concentration on perinatal outcomes. Our estimates suggest that the Puyehue episode increased the likelihood of low birth weight by 4.6 percentage points and the likelihood of preterm birth by 5.2 percentage points. These effects are not heterogeneous across women of different education levels and do not depend on the type of health coverage available (public vs. private).

Conclusions: Our study provides new evidence of an association between PM10 and fetal development in a developing country. Unlike many previous studies, our results show effects of relevant magnitude on low birth weight and prematurity, even after controlling for a rich set of covariates. In addition, our estimates are unlikely to capture delayed or chronic cumulative effects of pollution on health. In this sense, they can be good baseline estimates to project the potential effects of climate change on health through the retention of particles during episodes of heat waves.