A spatial evaluation of socio demographics surrounding National Priorities List sites in Florida using a distance-based approach

Quantifying Disparities in Per- and Polyfluoroalkyl Substances (PFAS) Levels in Drinking Water from Overburdened Communities in New Jersey, 2019-2021

BACKGROUND

Policymakers have become increasingly concerned regarding the widespread exposure and toxicity of per- and polyfluoroalkyl substances (PFAS). While concerns exist about unequal distribution of PFAS contamination in drinking water, research is lacking.

OBJECTIVES

We assess the scope of PFAS contamination in drinking water in New Jersey (NJ), the first US state to develop regulatory levels for PFAS in drinking water. We test for inequities in PFAS concentrations by community sociodemographic characteristics.

METHODS

We use PFAS testing data for community water systems (CWS) (n = 491 ) from the NJ Department of Environmental Protection (NJDEP) from 2019 to 2021 and demographic data at the block group level from the US Census to estimate the demographics of the NJ population served by CWS. We use difference in means tests to determine whether CWSs serving "overburdened communities" (OBCs) have a statistically significant difference in likelihood of PFAS detections. OBCs are defined by the NJDEP to be census block groups in which: a) at least 35% of the households qualify as low-income, b) at least 40% of the residents identify as people of color, or c) at least 40% of the households have limited English proficiency. We calculate statewide summary statistics to approximate the relative proportions of sociodemographic groups that are served by CWSs with PFAS detections.

RESULTS

We find that 63% of all CWSs tested by NJDEP from 2019 to 2021 had PFAS detections in public drinking water, collectively serving 84% of NJ's population receiving water from CWSs. Additionally, CWSs serving OBCs had a statistically significant higher likelihood of PFAS detection and a higher likelihood of exposure above state MCLs. We also find that a larger proportion of people of color lived in CWS service areas with PFAS detections compared to the non-Hispanic white population.

DISCUSSION

These findings quantitatively identify disparities in PFAS contamination of drinking water by CWS service area and highlight the extent of PFAS drinking water contamination and the importance of PFAS remediation efforts for protecting environmental health and justice. https://doi.org/10.1289/EHP12787.

Geographic Distribution of Accredited Paramedic Education Programs in the United States

Background: The geographic distribution and access to paramedic education programs is unclear but often cited as a reason for emergency medical services (EMS) workforce shortages. Our aims were: 1) to examine the spatial distribution of accredited paramedic programs and 2) to compare characteristics of communities with and without existing programs. Methods: We performed a cross-sectional study of US paramedic education programs accredited by the Commission on Accreditation of Allied Health Education Programs as of April 2020. Program locations were geocoded to county, and population estimates from the US Census Bureau were used to determine the adult population within the program's potential catchment area (30, 50, and 100 miles). Clustering of programs was examined using Moran's I. We compared community characteristics obtained from the 2018 American Community Survey, 2018-2019 Area Health Resources Files, and 2018 National Emergency Department Inventory between counties with and without programs. Logistic regression models were used to determine associations of community characteristics and existence of a paramedic program, controlling for urbanicity. Results: There were 790 paramedic program locations in the US, located in 596/3142 (19%) counties. Every state, except Rhode Island and Washington, DC, had at least one paramedic program site. The population within potential catchment areas ranged from 182 million (30 miles) to 248 million (100 miles), representing 73% to 99% of the US adult population, respectively. However, among counties classified as rural (n = 644), this decreased to 22% (30 miles) to 95% (100 miles). There was significant clustering of programs (p < 0.001). There were significantly higher odds of having a paramedic program for counties classified as metro compared to non-metro (OR 4.42, 95% CI 3.60-5.42) and with the presence of healthcare resources (e.g., emergency department in the county: OR 2.42, 95% CI 1.87-3.14). Conclusions: Approximately 73% of the US adult population lives within 30 miles of an existing paramedic education program; however, this decreases to 22% in rural areas. Geographic barriers to accessing paramedic education remain a challenge for ongoing efforts to address the rural EMS workforce shortage.

Spatial Associations Between Contaminated Land and Socio Demographics in Ghana

Associations between contaminated land and socio demographics are well documented in high-income countries. In low- and middle-income countries, however, little is known about the extent of contaminated land and possible demographic correlations. This is an important yet sparsely researched topic with potentially significant public health implications as exposure to pollution remains a leading source of morbidity and mortality in low-income countries. In this study, we review the associations between several socio demographic factors (population, population density, unemployment, education, and literacy) and contaminated sites in Ghana. Within this context, both correlation and association intend to show the relationship between two variables, namely contaminated sites and socio demographics. Aggregated district level 2010 census data from Ghana Statistical Service and contaminated site location data from Pure Earth’s Toxic Sites Identification Program (TSIP) were spatially evaluated using the number of sites per kilometer squared within districts as the unit of measurement. We found a low to medium positive correlation (ρ range: 0.285 to 0.478) between contaminated sites and the following socio demographics: higher population density, higher unemployment, greater education, and higher literacy rate. These results support previous studies and suggest that several socio demographic factors may be reasonably accurate predictors of contaminated site locations. More research and targeted data collection is needed to better understand these associations with the ultimate goal of developing a predictive model.

Associations between blood lead level and substance use and sexually transmitted infection risk among adults in the United States

The effects of low-level lead exposure on neuropsychological status in the United States (US) general adult population have been reported, and the relationship between neuropsychiatric dysfunction and health risk behaviors including substance use and sexual risk taking is well established. However, the potential influence of lead exposure on risk-taking behavior has received little attention. Using the National Health and Nutrition Examination Survey (NHANES) 2005-2010, we estimated multivariable logistic regression models to measure odds ratios (ORs) and 95% confidence intervals (CIs) for the cross-sectional associations between blood lead level and risk behaviors including binge drinking, drug use, and indicator of sexually transmitted infection (STI) risk. STI indicators included past 12 month sexual risk behaviors (age mixing with partners who were at least five years younger or older and multiple partnerships), self-reported STI, and biologically-confirmed herpes simplex virus type 2 (HSV-2) infection. Dose-response like relationships were observed between blood lead and substance use, age mixing with younger and older partners, self-reported STI, and HSV-2. In addition, participants with lead levels in highest quartile versus those with levels in the lowest quartile had over three times the odds of binge drinking and over twice the odds of injection drug or cocaine use in the past 12 months, while being in one of the top two quartiles was significantly associated with 30-70% increased odds of multiple partnerships, sex with older partners, and self-reported and biologically confirmed STI. Results from this study suggested that lead exposure may contribute to substance use, sexual risk-taking, and STI. However, given limitations inherent in the cross-sectional nature of the study, additional studies that use longitudinal data and measure detailed temporal information are warranted.

Uneven magnitude of disparities in cancer risks from air toxics

This study examines race- and income-based disparities in cancer risks from air toxics in Cancer Alley, LA, USA. Risk estimates were obtained from the 2005 National Air Toxics Assessment and socioeconomic and race data from the 2005 American Community Survey, both at the census tract level. Disparities were assessed using spatially weighted ordinary least squares (OLS) regression and quantile regression (QR) for five major air toxics, each with cancer risk greater than 10⁻⁶. Spatial OLS results showed that disparities in cancer risks were significant: People in low-income tracts bore a cumulative risk 12% more than those in high-income tracts (p < 0.05), and those in black-dominant areas 16% more than in white-dominant areas (p < 0.01). Formaldehyde and benzene were the two largest contributors to the disparities. Contributions from emission sources to disparities varied by compound. Spatial QR analyses showed that magnitude of disparity became larger at the high end of exposure range, indicating worsened disparity in the poorest and most highly concentrated black areas. Cancer risk of air toxics not only disproportionately affects socioeconomically disadvantaged and racial minority communities, but there is a gradient effect within these groups with poorer and higher minority concentrated segments being more affected than their counterparts. Risk reduction strategies should target emission sources, risk driver chemicals, and especially the disadvantaged neighborhoods.

A Nationwide Comparison of Driving Distance Versus Straight-Line Distance to Hospitals

Many geographic studies use distance as a simple measure of accessibility, risk, or disparity. Straight-line (Euclidean) distance is most often used because of the ease of its calculation. Actual travel distance over a road network is a superior alternative, although historically an expensive and labor-intensive undertaking. This is no longer true, as travel distance and travel time can be calculated directly from commercial Web sites, without the need to own or purchase specialized geographic information system software or street files. Taking advantage of this feature, we compare straight-line and travel distance and travel time to community hospitals from a representative sample of more than 66,000 locations in the fifty states of the United States, the District of Columbia, and Puerto Rico. The measures are very highly correlated (r² > 0.9), but important local exceptions can be found near shorelines and other physical barriers. We conclude that for nonemergency travel to hospitals, the added precision offered by the substitution of travel distance, travel time, or both for straight-line distance is largely inconsequential.

Cancer risk from exposure to hazardous air pollutants: spatial and social inequities in Tampa Bay, Florida

Recent environmental justice studies have emphasized the growing need to analyze the health impacts of disproportionate exposure to multiple pollution sources and incorporate geostatistical techniques that are suitable for analyzing spatial data. These objectives are addressed in a case study that evaluates spatial and social inequities in potential cancer risk from inhalation exposure to hazardous air pollutants (HAPs) from four types of emission sources in the Tampa Bay Metropolitan Statistical Area, Florida. This study utilizes modeled estimates of lifetime cancer risk from the 1999 National-Scale Air Toxics Assessment and socio-demographic data from the 2000 US Census. Statistical analyses are based on conventional multiple regression and locally derived spatial regression models that account for residual autocorrelation. Race, ethnicity, and home ownership are found to be significant predictors of cancer risk from ambient exposure to all four HAP source categories, after controlling for other relevant explanatory factors and spatial dependence in the data.

Disproportionate proximity to environmental health hazards: methods, models, and measurement

We sought to provide a historical overview of methods, models, and data used in the environmental justice (EJ) research literature to measure proximity to environmental hazards and potential exposure to their adverse health effects. We explored how the assessment of disproportionate proximity and exposure has evolved from comparing the prevalence of minority or low-income residents in geographic entities hosting pollution sources and discrete buffer zones to more refined techniques that use continuous distances, pollutant fate-and-transport models, and estimates of health risk from toxic exposure. We also reviewed analytical techniques used to determine the characteristics of people residing in areas potentially exposed to environmental hazards and emerging geostatistical techniques that are more appropriate for EJ analysis than conventional statistical methods. We concluded by providing several recommendations regarding future research and data needs for EJ assessment that would lead to more reliable results and policy solutions.

Estimation of exposure to toxic releases using spatial interaction modeling

BACKGROUND

The United States Environmental Protection Agency's Toxic Release Inventory (TRI) data are frequently used to estimate a community's exposure to pollution. However, this estimation process often uses underdeveloped geographic theory. Spatial interaction modeling provides a more realistic approach to this estimation process. This paper uses four sets of data: lung cancer age-adjusted mortality rates from the years 1990 through 2006 inclusive from the National Cancer Institute's Surveillance Epidemiology and End Results (SEER) database, TRI releases of carcinogens from 1987 to 1996, covariates associated with lung cancer, and the EPA's Risk-Screening Environmental Indicators (RSEI) model.

RESULTS

The impact of the volume of carcinogenic TRI releases on each county's lung cancer mortality rates was calculated using six spatial interaction functions (containment, buffer, power decay, exponential decay, quadratic decay, and RSEI estimates) and evaluated with four multivariate regression methods (linear, generalized linear, spatial lag, and spatial error). Akaike Information Criterion values and P values of spatial interaction terms were computed. The impacts calculated from the interaction models were also mapped. Buffer and quadratic interaction functions had the lowest AIC values (22298 and 22525 respectively), although the gains from including the spatial interaction terms were diminished with spatial error and spatial lag regression.

CONCLUSIONS

The use of different methods for estimating the spatial risk posed by pollution from TRI sites can give different results about the impact of those sites on health outcomes. The most reliable estimates did not always come from the most complex methods.

Evaluating geographic imputation approaches for zip code level data: an application to a study of pediatric diabetes

BACKGROUND

There is increasing interest in the study of place effects on health, facilitated in part by geographic information systems. Incomplete or missing address information reduces geocoding success. Several geographic imputation methods have been suggested to overcome this limitation. Accuracy evaluation of these methods can be focused at the level of individuals and at higher group-levels (e.g., spatial distribution).

METHODS

We evaluated the accuracy of eight geo-imputation methods for address allocation from ZIP codes to census tracts at the individual and group level. The spatial apportioning approaches underlying the imputation methods included four fixed (deterministic) and four random (stochastic) allocation methods using land area, total population, population under age 20, and race/ethnicity as weighting factors. Data included more than 2,000 geocoded cases of diabetes mellitus among youth aged 0-19 in four U.S. regions. The imputed distribution of cases across tracts was compared to the true distribution using a chi-squared statistic.

RESULTS

At the individual level, population-weighted (total or under age 20) fixed allocation showed the greatest level of accuracy, with correct census tract assignments averaging 30.01% across all regions, followed by the race/ethnicity-weighted random method (23.83%). The true distribution of cases across census tracts was that 58.2% of tracts exhibited no cases, 26.2% had one case, 9.5% had two cases, and less than 3% had three or more. This distribution was best captured by random allocation methods, with no significant differences (p-value > 0.90). However, significant differences in distributions based on fixed allocation methods were found (p-value < 0.0003).

CONCLUSION

Fixed imputation methods seemed to yield greatest accuracy at the individual level, suggesting use for studies on area-level environmental exposures. Fixed methods result in artificial clusters in single census tracts. For studies focusing on spatial distribution of disease, random methods seemed superior, as they most closely replicated the true spatial distribution. When selecting an imputation approach, researchers should consider carefully the study aims.