Author: Braud, Troylyn
Date published: January 1, 2011
The environment has been directly linked to health outcomes since the 18th century. While we are sometimes aware of the immediate threat imposed by the environment that we inhabit, long-term implications are sometimes difficult to measure. "The public's health, particularly its environmental health, depends on the interaction of many factors; in order to provide a healthy environment within the nation's communities, the places people spend the most time - their homes, schools, and offices - must be considered (Healthy People 2010, 2008)." According to the World Health Organization, "Environmental health comprises those aspects of human health, including quality of life, that are determined by physical, chemical, biological, social, and psychological factors in the environment; it also refers to the theory and practice of assessing, correcting, controlling, and preventing those factors in the environment that can potentially affect adversely the health of present and future generations (World Health Organization, 2004)." Infant mortality, which is impacted by prematurity (<37 weeks gestation) and low birth weight (<2,500 grams) (LBW), has been deemed a health burden for Memphis and Shelby County, Tennessee.
Memphis has one of the highest rates of infant mortality among the nation's largest cities, with a steadily climbing rate over the past few years. Memphis' average mortality rate of 14.1 babies per 1,000 live births is more than twice the national average of 6.9 babies per 1,000 live births, with the highest rate revealed for African-Americans (Tennessee Department of Health, 2006). African-Americans are twice as likely to experience the aforementioned outcome when compared with their same white counterparts. In a previous study of residential proximity to polychlorinated biphenyl (PCB)-contaminated waste sites, researchers noted that living in a zip code near a PCB-contaminated site posed risks of exposure and of giving birth to an infant of low birth weight (Baibergenova, Kudyakov, Zdeb, & Carpenter, 2003). Berry and Bove (1997) found an increase in the proportion of low birth weight babies and a lower than average birth weight in the population living closest to a landfill during a critical exposure period. Another study found after evaluating the effect of air pollution on birth outcomes in Shelby County that ambient concentrations of criterion air pollutants (e.g., lead and sulfur dioxide) along with economic characteristics of the mother's residential census tract demonstrated a sizeable negative effect on birth weight (Williams, Pennock-Roman, Suan, Magsumbol, & Ozdenerol, 2006). Shelby County is one of the top-ranking counties in the state with releases of known substances significant to developmental disorders (U.S. Environmental Protection Agency [U.S. EPA], 2007). Much of what we know is limited; however, we do realize that the unborn are susceptible to environmental toxins. Hence, the objective of our study was to evaluate the impact of residential proximity to toxic release sites (TRS) and potential implications on low birth weight and premature delivery in Shelby County women.
Toxic Release Sites
Memphis has over 100 sources of pollution dispersed via land, water, or air. This information is available at the U.S. Environmental Protection Agency's Web site in the 2005 Toxic Release Inventory Report, which furnishes identifiable markers such as company name, address, geographical coordinates, and chemical emissions. The 2005 Toxic Release Inventory Report, which was the most current at the time our analysis began, was comprised of data collected from 2005 to 2006 and released to the public in 2007 (U.S. EPA, 2007). Using the list in that report, we identified the top 10 polluters in Memphis and subsequently classified them by total emissions. Arbitrary names were assigned for each of the 10 TRS, i.e., Sites 1-10. Maternal proximity to each TRS was classified into three groups (<5 miles, 6-10 miles, and >10 miles). Maternal residences of more than 10 miles were used as the reference group.
Recruitment of Subjects
The data along with potential confounders and effect modifiers were obtained from the database of the Blues Project, which is a prospective cohort designed to positively impact health and birth outcomes through pre- and postnatal education, individualized case management, social support, and community referrals (all participants received the intervention). Eligible participants were <29 weeks gestation at enrollment in the Blues Project and volunteered to receive prenatal care and follow-up postpartum and pediatric care services until the baby's second birthday. The target population included 382 women who enrolled in Phase II of the Blues Project and resided in Shelby County during pregnancy and delivery The participants were recruited from August 2007 to December 2008 from three community health clinics located in areas that were identified by zip code as having the highest rates of infant deaths in Memphis. The final sample (N = 369) included all singleton births. All multiple deliveries, observations with missing address information, and outliers were excluded from the sample.
Common risk factors for LBW and prematurity were examined in an effort to identify potential confounders. The following information was available from Blues Project data: maternal address at delivery, birth weight, gestational age approximated by the last menstrual period, maternal age, marital status, education, employment status, race, sexually transmitted disease, and substance use (drugs, alcohol, and tobacco use). Maternal age was classified into two groups, <19 years and >19 years (>19 years = reference group). Race was divided into the following four categories: African-American, white, Latino, or other (white = reference group). Marital status, employment, sexually transmitted diseases, and substance use (drugs, alcohol, and tobacco) were classified as yes or no (yes = reference group). Education was categorized as less than a high school education and at least a high school education (at least a high school education = reference group).
Frequency tables were generated to evaluate the occurrence of LBW and prematurity within the sample. The LBW and premature cases were evaluated with respect to the TRS zip codes. Logistic regression was used to evaluate predictive properties for each of the TRS. Crude and adjusted odds ratios (OR), along with the respective 95% confidence intervals (Ci), were calculated for the outcomes of interest. Data were analyzed using SAS version 9.1. Race was not included in the final model because all prematurity and LBW cases were within the same race category (Table 1).
ArcGIS version 9.2 was used to map the mother's residence with respect to each TRS. The most current street and county shape files (Tiger Line 2000) were downloaded from the U.S. Census Bureau Web site. The shape files were used to geocode both participant and release site addresses. Seventy-five percent of the residents' addresses were geocoded successfully with a match score of 80100, while 60% of the toxic release sites were successfully geocoded with a match score of 80-100. Coordinates for the remaining 25% and 40%, respectively, were generated using online software available at www.batchgeocode.com. The aforementioned Web site was also used to calculate straight-line distance from each TRS to each residence.
A total of 382 women were enrolled and 369 (97%) remained in the study through delivery. Participants with multiple births or missing data and one outlier were excluded from this analysis (n = 13). Of the 369 deliveries, 8.67% were preterm and 9.76% were LBW The mean birth weight was 3116.59 ± 500.37 grams, and the mean gestational age was 39.05 ± 1.699 weeks. Nineteen percent reported sexually transmitted infections. Most were African-American (96%), unmarried (96%), unemployed (71%), had less than a high school education (50%), and were >19 years of age (88%). The mean maternal age was 22.73 ± 5.00 years. Few participants reported use of tobacco, alcohol, or illegal drugs during pregnancy. The distribution of birth outcomes and potential risk factors for LBW and premature birth are presented in Table 1. With the exception of Site 6 and Site 7, all of the premature and LBW cases were within a 10-mile radius of all pollution sites. No statistically significant association existed between TRS proximity and LBW outcomes after adjusting for employment, education, maternal age, marital status, substance use, and sexually transmitted diseases (Table 2). Proximity to Site 8 (OR = 4.018, Cl = 1.103-14.643) and Site 10 (OR = 2.667, Cl = 1.036-6.862), however, put mothers at increased risk for preterm births. None of the remaining eight TRS posed a risk for the outcomes of interest (Table 3).
Production information for each of the TRS is presented in Table 4. Figure 1 shows the geographic distribution of LBW cases contrasted with all normal babies within the sample. Figure 2 projects the geographic distribution of premature delivery contrasted with all normal babies within the sample.
Proximity to TRS did not significantly impact LBW outcomes. When the direction of the odds ratios is evaluated, however, it is evident that risk increases as an individual approaches Site 8, Site 9, and Site 10, but risk decreases with increasing proximity to the remaining seven sites (Table 2). Individuals within five miles of Site 8 were four times more likely to have premature babies after adjustment for confounders. Additionally, those who lived within five miles of Site 10 were three times more likely to have premature babies after adjustment for confounders.
Two of the three emissions from Site 8, which imposed the greatest risk among the top 10 sites, included toluene and xylene, which are known developmental toxicants. Additionally, methyl isobutyl ketone was emitted by Site 8, but not enough literature exists to support the theory that it is a developmental toxicant. Some of the chemicals emitted into the water or air via Site 9, which may affect development and reproductive health, include hydrogen fluoride, benzene, dioxins, ethylbenzene, lead, mercury, N-hexane, naphthalene, styrene, phenol, carbon disulfides, glycol ethers, toluene, and xylene. Site 10 emitted acetaldehyde, which is also a developmental toxicant, into the air.
The cohort was comprised of women who received the same quality of prenatal care and the Blues Project intervention, which is designed to counteract the common risk factors. The data were obtained from an active study; therefore, the limitations identified throughout the research process can be addressed and subsequently used to strengthen future research projects. Residential address was used to calculate Euclidian distance, thus providing a means for unambiguous assessment of each participant from each of the 10 TRS.
Gestational age was calculated from the last menstrual period reported by the participants. Because the majority of the premature babies had a gestational age of 36 weeks, actual cases of prematurity may not be precisely represented. Sram and co-authors (2005) suggest that the first month of gestation is the most vulnerable period in terms of air pollution. As many as 25% of women change residence between conception and the end of the first trimester (Baibergenova et al., 2003). Length of residence and wind direction were not accounted for; therefore, direct exposures were not assessed. The information obtained from the Toxic Release Inventory Report was compiled from 2005 to 2006 and made available to the public in 2007; therefore, it is assumed that emissions were the same for the study period. The sample was taken from a cohort aimed at reducing prematurity and LBW outcomes; thus, it is not representative. Nevertheless, findings from this study are useful to inform future research and provide a direction for further investigation.
The purpose of our study was to evaluate birth outcomes, LBW and prematurity, and the effect of TRS proximity. Our study provided some insight regarding residential location to hazardous sites and the potential consequence of living in close proximity. Additionally, it provided a platform for future research regarding toxic release substances and external factors that may negatively impact human health.
Acknowledgments: The authors wish to acknowledge Pam Connor, PhD, Esra Ozdenerol, PhD, and Elizabeth Webb for research support and mentoring.
Baibergenova, A., Kudyakov, R., Zdeb, M., & Carpenter, D.O. (2003). Low birth weight: Proximity to PCB-contaminated waste sites. Environmental Health Perspectives, 111(10), 1352-1357.
Berry, M., & Bove, F (1997). Birth weight reduction associated with residence near a hazardous waste landfill. Environmental Health Perspectives, 105(8), 2.
Healthy People 2010. (2008). Environmental health. Retrieved December 5, 2008, from www.healthypeople.gov/document/html/ volume1/08environmental.htm
Sram, R., Binkova, B., Dejmek, J., & Bobak, M. (2005). Ambient air pollution and pregnancy outcomes: A review of the literature. Environmental Health Perspectives, 113(4), 375-382.
Tennessee Department of Health. (2006). Racial disparity in infant mortality in Tennessee. Nashville: Tennessee Department of Health Office of Policy, Planning, and Assessment.
U.S. Environmental Protection Agency. (2005). Toxic release inventory explorer. Retrieved November 2008, from www.epa.gov/triexplorer
U.S. Environmental Protection Agency. (2007). Toxic release inventory explorer. Retrieved January 2009, from www.epa.gov/triexplorer
Williams, B., Pennock-Roman, M., Suan, K.H., Magsumbol, M., & Ozdenerol, E. (2007). Assessing the impact of the local environment on birth outcomes: A case for HLM. fournal of Exposure Science & Environmental Epidemiology, 17(5), 445-457.
World Health Organization. (2004). Environmental health. Retrieved October 5, 2010, from http://www.who.int/topics/environmental_health/en/
Troylyn Braud, MS
Simonne Nouer, MD, PhD
Kimberly Lamar, MPH, PhD
Corresponding Author. Troylyn Braud, Health Educator, University of Tennessee Health Science Center, Department of Preventive Medicine 600 Jefferson Ave., 3rd floor, Memphis, TN 38105. E-mail: email@example.com.