Stream Condition and Infant Mortality in U.S. Mid-Atlantic States

ABSTRACT

Infant mortality rate (IMR) serves as a summary measure of the health of a nation's population. The U.S. IMR has declined over the past several decades (to 6.85 per 1000 in 2003), but remains high compared with other developed countries. We hypothesized a relationship between IMR and poor water quality at a local scale. We used degraded stream condition, represented by a multimetric index of biotic condition, as a metric for poor water quality. Using conditional probability analysis on county-aggregated data for the state of Maryland, we show that there is a significant association between the extent of a county's stream miles in poor ecological condition and the probability of a county's IMR exceeding the national norm. The overall relationship appears to be robust, as similar associations were also found for the states of West Virginia and Pennsylvania. We are not implying that IMR and degraded stream condition are directly causally related, but rather that there may be common factors that affect both of them. We hypothesize that attributes and activities at the county scale (e.g., socioeconomic conditions, land use, human activities) may contribute to both stream degradation and increased IMR.

Key Words:

• stream condition
• infant mortality
• conditional probability
• public health
• ecological condition
• statistical analysis

ACKNOWLEDGMENTS

Thanks go to Jennifer Orme-Zavaleta, Dina Schreinemachers, Hillel Koren, Patricia Bradley, Richard Wiggins, Linda Birnbaum, Ron Klauda, and four anonymous reviewers for the critical and constructive reviews that they provided and to John Barton for editorial assistance. Special thanks go to the Maryland Department of Natural Resources and their MBSS, the USEPA Environmental Monitoring and Assessment Program, and the CDC Wide-Ranging Online Data for Epidemiologic Research system for generating high quality data and making these data available to all.

The research described in this article was funded by the USEPA. This article has not been subjected to Agency review; therefore, it does not necessarily reflect the views of the Agency. Mention of trade names or commercial products does not constitute endorsement or recommendation for use. The National Health and Environmental Effects Research Laboratory Human Research Protocol Office has confirmed that the research in this study did not meet the definition of human study research as defined in 40 CFR 26.102.