Environmental Burden and School Readiness in an Urban County: Implications for Communities to Promote Healthy Child Development

Abstract

Geographic disparities threaten equitable access for children to health-promoting safe green spaces, and quality early education in the communities in which they live and grow. To address gaps in the field, we integrated the fields of developmental psychology, public health, and environmental science to examine, at the population level, associations between the environmental burden, socioeconomic vulnerability, and kindergarten readiness in a diverse urban county. Three administrative datasets were integrated through an early childhood data sharing research partnership in Miami-Dade County. The Bruner Child Raising Vulnerability Index, the five domains of the Environmental Burden module from the Environmental Justice Index, and public school kindergarten readiness scores were aggregated at the census tract level. Analysis of variance and multiple regression analyses found associations between socioeconomic vulnerability and race/ethnicity. The socioeconomic vulnerability levels were highest in census tracts with a higher percentage of Black residents, compared to all other races/ethnicities. Areas of greater social vulnerability had lower kindergarten readiness and a higher environmental burden. A higher environmental burden predicted lower kindergarten readiness scores above and beyond race/ethnicity and socioeconomic vulnerability. The findings advance our understanding of global challenges to sustainable healthy child development, such as the persistence of a disproportionate environmental burden and inequitable access to resources such as green spaces and early education programs. The present study results can inform community health improvement plans to reduce risk exposures and promote greater access to positive environmental and educational resources for all children.

1. Introduction

Equitable access to health care, quality early education programs, and safe green communities are key contributors to sustainable community health and child well-being. Environmental toxic exposure in children’s homes and neighborhoods is a critical challenge to the United Nations 2030 Sustainable Development Goals that call for policies that ensure healthy lives and promote well-being for all ages and make cities and human settlements inclusive, safe, resilient, and sustainable for all residents [1]. High exposure to toxic chemicals, pollution, and other environmental risks in children’s communities negatively influence their physical health, cognitive and social-emotional development, and behavior, which are the foundation of early learning and school readiness [2]. From a public health perspective, it is critical to identify where high levels of environmental risk exposure affect our youngest citizens during their most formative developmental periods. Additionally, geographic inequities across communities—in the distribution of toxic exposures and access to promotive factors such as health care, high-quality education, and safe green open spaces—disproportionately and directly affect opportunities for low-income children to develop positive health and well-being [2,3]. In fact, it has been demonstrated that where you live during critical points in your life strongly influences your future health, educational attainment, and economic well-being [4].

The accumulation of risks associated with adverse environmental exposures within communities has been conceptualized as an “environmental burden” within the environmental justice framework [3]. On average, children and families are exposed to some level of environmental risks in their community; however, some communities bear a greater burden of environmental risks and related health sequelae than others. This inequitable distribution of risk exposure has been termed an “environmental burden.” To date, few studies have examined the relationship between a neighborhood’s environmental burden and children’s developmental skills at kindergarten entry. “Kindergarten readiness” or “school readiness” is defined as a multidimensional set of developmental skills that children acquire prior to kindergarten from birth to five years through early learning experiences. These developmental skills support a successful transition to formal schooling both academically and socially [5]. Skills considered integral to school readiness include cognitive skills such as language development and higher order thinking and problem solving skills, early literacy and numeracy skills, social–emotional skills, approaches to learning, and physical development and health [6].

School readiness skills rapidly develop during the early childhood period. Resources present within children’s family, community, and early care and education settings provide learning opportunities and have the potential to set a strong foundation for future academic and social–emotional success [7]. During this critical period of development, children are most vulnerable to the environmental risk and protective factors that they are exposed to in their residential neighborhood [8]. Much prior research has identified associations between poverty and reduced opportunities or “opportunity gaps” for children [9,10,11]. Poverty is a global problem that affects every aspect of children’s lives. More work is needed to understand the dual role that poverty may play in terms of both the environmental burden and disparities in access to resources in the neighborhood community for children growing up in poverty, both of which may be most evident in our nation’s urban centers [12].

Thus, more research within urban communities is needed to identify geographic disparities in the environmental burden and to examine whether environmental factors are associated with children’s school readiness skills. Informed by a bioecological developmental model and the related environmental justice framework, the present study advances developmental science by responding to the recent call for integrating the fields of environmental science with developmental psychology [13]. We conducted a population-level study for an entire cohort of kindergarten children entering a large public school system in a large urban area. We examined the environmental burden, associated neighborhood sociodemographic variables, and relationships to school readiness at the census tract level. This set of analyses provided a descriptive picture of geographic disparities in sociodemographic and environmental factors associated with school readiness at the community level. Our study goal was to inform targeted community-based strategies that address one of the critical challenges to children’s healthy development globally within urban environments—the promotion and sustainability of healthy, safe community environments across the lifespan.

1.1. Environmental Burden and Community Demographic Variables

Geographic disparities in terms of the environmental burden are well-documented across race, ethnicity, and poverty at the community level in the U.S. Ethnically and racially minoritized children are disproportionately more likely to live in poverty compared to their non-minoritized peers [14,15]. Ethnicity and race are correlated with residential location, with minorities and whites often segregated, and minority groups are more likely to live in less advantaged neighborhoods [16]. Minoritized children, thus, are more likely to live in lower-opportunity neighborhoods with higher environmental burdens due to systemic structural racism, racial segregation, and historic redlining practices [4,17].

Minoritized communities experience greater exposure to environmental hazards, such as air pollution, toxic sites, and water pollution [18] and a greater likelihood of living in proximity to hazardous waste and environmental pollutants [19,20]. Air pollutants often come from traffic and highway sources. Due to historical precedents—such as redlining, federal highways placed through minoritized communities, and residential segregation, Black families living in low-income areas are more likely to live where there is higher air pollution [21,22]. In urban areas, racial and ethnic minority communities are exposed to pollution from nearby railroads, highways, industries, municipal incinerators, and other toxic facilities or waste sites. Some of these communities also lack adequate infrastructure, including aging infrastructure that threatens sewage and safe drinking water [23]. Jbaily and colleagues (2022), using national data from 2004 to 2016, found that communities with a higher proportion of white and higher income populations were consistently exposed to lower fine particulate matter PM2.5 levels than areas with higher Black, Asian, and Hispanic populations [24]. In addition, low-income communities were more likely to be exposed to higher PM2.5 levels than areas with higher income groups [25]. In addition, Geron and colleagues (2022) found that Black non-Hispanic, Black Hispanic, and White Hispanic pregnant women had significantly higher levels of toxic metal traces in their urine than White non-Hispanic women [26]. Higher levels of metal traces were also associated with residential areas higher in crime, minority populations, and poverty.

1.2. Environmental Burden, Sociodemographic Variables, and School Readiness

Several studies document associations between a community’s environmental burden and child health outcomes; however, few studies examine associations between a community’s environmental risks and school readiness. For example, neighborhood levels of air pollution such as fine particulate matter (PM), volatile organic compounds (VOCs) and polycyclic aromatic hydrocarbons (PAH) are found to be associated with poor asthma morbidity and the development of respiratory disease in children [27,28,29]. Other studies have found that higher home exposure to air pollutants is associated with poorer neurodevelopmental outcomes for children [30] and that pesticide exposure is associated with internalizing behaviors in children [31].

In addition, exposure to air pollutants during pregnancy and after birth has been linked to cognitive impairments in school-aged children, such as attention and memory deficits and executive functioning challenges [25,32,33]. Traffic-related air pollution has also been shown to be associated with reduced attention and inhibitory control, which can negatively affect academic performance [32,34]. Entering school with respiratory conditions like asthma, which are influenced by environmental contaminants, has shown to be a significant predictor of lower academic achievement [35], lower attendance, and grade retention risk [36].

Other research has examined individual social and health risk factors influencing children’s neurocognitive, health, and educational development [37,38,39,40]. Social vulnerabilities (e.g., racial and ethnic minority status, poverty) are related to lower phonological skills [37], lower executive functioning skills [38], fewer language skills [39], and lower rates of school readiness [41]. Similarly, health vulnerabilities (e.g., pre-existing chronic diseases such as asthma, diabetes) have associations with child maladjustment [40], social and behavioral problems, decreased cognitive scores, and lower verbal IQ [2]. All of these factors may affect a child’s readiness for school by the time they enter kindergarten [42].

1.3. Theoretical Framework

Two distinct but somewhat related theoretical frameworks can help to inform our understanding of the complex inter-relationships between health disparities, environmental burden, and children’s neurobehavioral outcomes such as kindergarten readiness. Both the bioecological [43] and environmental justice models [44] support the critical need to measure and examine the direct and interactive influence of multiple factors on the developing child. To begin with, the bioecological model identifies several domains of influence on young children’s outcomes, including biological, behavioral, physical/built environment, and socioeconomic environment influences that operate at multiple ecological levels. Key to the ecological model is the interactive dynamic nature of environmental influences on the developing child, during a time in when children are particularly vulnerable to environmental exposure, as their positive and healthy development is highly dependent upon the resources available within their immediate contexts, such as the family, school, and community [43].

A related framework, the environmental justice (EJ) framework, also has implications for children’s developmental outcomes. Through an EJ lens, children and families are seen as contending with constellations of correlated environmental, health, and socioeconomic risks in combination rather than isolated instances of adverse circumstances [45]. The EJ framework conceptualizes that community risks and resources influence children’s development and health, within the macro context of structural racism and poverty. In historically marginalized areas, risks may be disproportionately present, contributing to a higher likelihood that children will experience higher exposure to toxins in their environment and less exposure to protective factors that promote healthy development over their lifetime [2,44]. These risks are unequally distributed, thus disproportionately affecting racially marginalized, politically disenfranchised, and socioeconomically disadvantaged populations [46].

Both the bioecological and environmental justice models suggest that children who experience greater risks relative to single risk exposures are likely to have worse developmental and health outcomes [13,47,48]. More specifically, the environmental justice (EJ) framework recognizes an unequal distribution of environmental risk disproportionately affecting racially marginalized, politically disenfranchised, and socioeconomically disadvantaged populations [46].

Over time, the meaning, scope, and focus of EJ work have evolved, reflecting a growing recognition of how the environments where we “live, work, and play” [49] affect human health and well-being [44]. Despite this evolution, the mandate to understand and redress inequities in the distribution of hazardous environmental exposures that may lead to negative physical and social health outcomes remains. By integrating environmental science, social science, public health, and urban planning, EJ acknowledges that science cannot exist in isolation from the social context in which it is conducted and promotes interdisciplinary work to examine disparities [49].

In the present study, we employed both a bioecological and EJ lens to investigate how environmental and social determinants intersect to influence kindergarten readiness, a key predictor of children’s short- and long-term academic success [50,51]. Our study examines environmental burden, a critical challenge to the development of healthy community environments, within the broader context of neighborhood-level socioeconomic and demographic characteristics in an urban setting.

1.4. Study Purpose

Although direct associations are well-documented between community environmental risk exposures and children’s health outcomes (e.g., asthma, neurodevelopmental disorders, and school attendance), less is known about the relationship between a neighborhood’s environmental burden and the school readiness skills of young children. Similarly, although many prior studies focus on the effects of air pollution, lead exposure, and other environmental toxins on the respiratory health and neurodevelopmental skills of children, few studies have examined other environmental variables that contribute to healthy (or unhealthy) neighborhoods in which children live, learn, and grow.

To address these gaps in the literature, the present study used population-level data for an entire cohort of children entering kindergarten in a large urban county to examine three research questions: (a) At the population level, what is the descriptive picture of environmental burden across the county in areas where children reside?; (b) What neighborhood sociodemographic variables are associated with the environmental burden geographically, such as race, ethnicity, and poverty?; and (c) What is the association between neighborhood environmental burden and kindergarten readiness, controlling for neighborhood sociodemographic variables?

To answer these research questions, we assessed the environmental burden using the Environmental Burden module from the Environmental Justice index (EJI), a nationally representative dataset that measures five key domains at the census tract level: exposure to hazardous and toxic sites, air and water pollution, built environment, and transportation infrastructure [3].

1.5. Hypotheses

We expected to find geographic variation in the environmental burden across communities, with socioeconomic vulnerability correlated with a higher environmental burden. In addition, based on prior research, we hypothesized that lower-resourced areas of the county, with a higher percentage of minoritized residents, would be associated with a higher socioeconomic and environmental burden. Finally, we expected that socioeconomic and environmental risk would be negatively associated with kindergarten readiness at the community level. This important descriptive analysis provides a county-wide understanding of the environmental landscape to identify areas of higher exposure, as risks vary geographically and across other neighborhood characteristics.

2. Materials and Methods

2.1. Study Setting and Participants

The participants included an entire cohort of 19,373 five-year-old children who entered Miami-Dade County public school kindergarten in the fall of 2021 and who took the statewide kindergarten readiness test (Florida Kindergarten Readiness Screener; FLKRS) [52]. Of this cohort of students, 73% were Hispanic, 18% Black non-Hispanic, 6% White non-Hispanic, 1% Asian, and 2% multiethnic or other. Of the students, 49% were female, and approximately 72% qualified for free or reduced lunch. In terms of home language, student records indicated 47% spoke English, 46% spoke Spanish, 2% spoke Haitian Creole, with the remainder speaking one of the 56 other languages of enrolled students. The demographics reflect those of the Miami-Dade County Public Schools.

County-wide data were included for all 707 census tracts (based on the 2020 U.S. Census). Miami-Dade County is the seventh largest U.S. county by population, with 2.7 million residents and spans a relatively large geographic area (5135 square kilometers in land area) [53]. The county consists of areas of varying urbanicity, including urban, suburban, and semi-rural areas. The publicly available data used in this study came from the American Community Survey [54] and the Environmental Justice Index [3]. The kindergarten readiness data were collected from 19,373 kindergarteners in Miami-Dade County Public Schools for the fall 2021. The children’s scores were aggregated to their home census tracts for analysis.

2.2. Procedures

Prior to the start of the study, approval for the research project was obtained through the university Institutional Review Board. Through a formal data sharing early childhood research partnership between the University and four major early childhood programs in the County (including the local school district), an integrated data system was used to link four administrative datasets for the present study. The data were shared and integrated under the auspices of a formal Data Sharing and Collaboration Agreement that provided guidelines for data sharing and use. Student-level files were shared with university data scientists through an encrypted secure data transfer procedure and housed in a secure protected network server. University data scientists geocoded student-level data at the census tract and de-identified files before sharing with the research team for analyses. A series of data quality checking steps was conducted manually to cross-check and ensure the accuracy of linkage, reliability, and validity of the whole dataset.

The data sources included: (a) individual public school student records for a cohort of children enrolled during the 2021–22 school year that included child and family demographic variables, including the home address, (b) individual student statewide kindergarten assessment scores, (c) 11 variables from the American Community Survey [54], and (d) the Environmental Burden module from the Environmental Justice Index [3]. The student-level data were linked across datasets by census tract.

2.3. Datasets and Measures

2.3.1. Community Race and Ethnicity

Percent White, percent Black, and percent Hispanic at the census tract level were obtained through the U.S. Census website, for the 2017–2021 American Community Survey [54]. These variables report the percent of White, Black, and Hispanic residents in each of the census tracts in Miami-Dade County. Race and ethnicity variables were categorized to be able to easily compare across census tracts with differing racial and ethnic characteristics. For ethnicity, census tracts were categorized in four groups: <25% Hispanic, 25–49% Hispanic, 50–74% Hispanic, and >75% Hispanic. For race, census tracts were categorized into seven groups: <25% Black non-Hispanic, 25–49% Black non-Hispanic, 50–74% Black non-Hispanic, >75% Black non-Hispanic, <25% White non-Hispanic, 25–49% White non-Hispanic, and >50% White non-Hispanic. Because there was only one census tract with greater than 75% White non-Hispanic residents, we collapsed the 50–74% and >75% categories to be able to perform post hoc analyses.

2.3.2. Neighborhood Socioeconomic Risk

The Bruner Child Raising Vulnerability Index was used to assess neighborhood socioeconomic risk [55] based on the 2017–2021 American Community Survey. The Index includes 11 indicators comprising four domains of social, educational, and economic vulnerability. The social domain included 3 indicators: the percentage of households with children under 18 with single parents, the percentage of the adult population with limited English proficiency, and the percentage of 16- through 19-year-olds not in school or working. The educational domain included 3 indicators: the percent of adults 25 years or older without a high school diploma, the percent of adults 25 years or older with at least a college degree (reverse coded), and the percent of three to four-year-olds enrolled in school (reverse coded). The economic domain included 3 indicators: the percentage of households without wage income, the percent of families with children living in poverty, and the percent of households on public assistance. Finally, the wealth domain included 2 indicators: the percent of owner-occupied housing (reverse coded) and the percent of households with interest, rent, or dividend income (reverse coded). Using Bruner’s methodology, census tracts that were at least one standard deviation away from the U.S. national mean were classified as “vulnerable” on that indicator and received a score of 1. An overall total socioeconomic vulnerability index score was calculated by summing across the indicators (scores could range from 0 to 11). A categorical variable of socioeconomic vulnerability was also calculated for each census tract with four categories: no vulnerabilities, 1–2 vulnerabilities, 3–5 vulnerabilities, and 6 or more vulnerabilities.

2.3.3. Environmental Burden

The environmental burden was measured using the Environmental Burden Module of the Environmental Justice Index [3]. The Environmental Justice Index (EJI) was recently created to measure the environmental burden at the community level. It is a national place-based tool used to measure neighborhood environmental burden and social and health vulnerability for each census tract in the U.S [3]. The most recent available EJI data were accessed and downloaded through the U.S. Centers for Disease Control website on 7 December 2024.

For the current study, we focused on the EJI Environmental Burden Module, which consists of five domains: air pollution, potentially hazardous and toxic sites, water pollution, built environment, and transportation infrastructure [56]. These five domains were chosen by the Centers for Disease Control (CDC) to capture factors that “contribute either negatively or positively to human health and well-being” Data from the EJI Environmental Burden module are pulled from several data sources, including the U.S. Environmental Protection Agency’s (EPA) Air Quality System (AQS), National AirToxScreen, National Walkability Index (NWI), Watershed Index Online (WSIO), and the Facility Registry Services (FRS); the U.S. Mine Safety and Health Administration’s Mine Data Retrieval System; the U.S. Geospatial Survey’s Protected Areas Database of the United States (PAD-US) 4.0; the U.S. Census Bureau’s American Community Survey (ACS); the U.S. Department of Transportation’s National Highway System (NHS) and National Transportation Atlas Database [3]. Further definition and description of each of these domains are provided below.

The air pollution domain includes the following indicators: the mean number of days above the regulatory standard for ozone, the mean number of days above the regulatory standard for particulate matter 2.5, the ambient concentrations of diesel particulate matter, and the likelihood of developing cancer from air toxins throughout a lifetime, assuming continuous exposure.

The potentially hazardous and toxic sites domain includes the following indicators: the proportion of the census tract that is within a 1-mile buffer of an EPA National Priority List (NPL) site, the proportion of the census tract that is within a 1-mile buffer of an EPA Toxic Release Inventory (TRI) site, the proportion of the census tract that is within a 1-mile buffer of an EPA Treatment, Storage, and Disposal Facility (TSDF), the proportion of the census tract that is within a 1-mile buffer of an EPA Risk Management Plan (RMP) site, the proportion of the census tract that is within a 1-mile buffer of a coal mine, and the proportion of the census tract that is within a 1-mile buffer of a lead mine.

The built environment domain includes the following indicators: the lack of access to green space measured by the proportion of the census tract that is within a 1-mile buffer of a park or greenspace (reverse coded), the percentage of houses built before 1980 to estimate potential exposure to lead, and the lack of walkability measured by the National Walkability Index (NWI) rank (reverse coded).

The transportation infrastructure domain includes indicators measuring the proximity to different modes of transportation that are considered a significant source of noise and air pollution and safety risk; these include the following: the proportion of the census tract that is within a 1-mile buffer of a railway, the proportion of the census tract that is within a 1-mile buffer of a high-volume roadway or highway, and the proportion of the census tract that is within a 1-mile buffer of an airport.

The water pollution domain includes the following indicator: the percentage of the census tract that intersects an impaired or impacted watershed at the Hydrologic Unit Code (HUC)-12 level. The HUC was developed by the US Geological Survey (USGS) to map and manage water resources in the United States [57]. Each census tract received a percentile rank score for each indicator and domain that were included in the analyses.

2.3.4. School Readiness

The Star Early Literacy Florida Kindergarten Readiness Screener (FLKRS) [58] was administered to all kindergarten children entering Miami-Dade County Public Schools. During the fall of 2021, the kindergarten screener was the STAR Early Literacy assessment designed to measure early literacy skills and numeracy of children aged 3 to 9 years. The test assessed the following domains: word knowledge and skills, comprehension strategies and constructing meaning, and numbers and operations [52]. The present study aggregated children’s individual FLKRS total scores to the census tract level using their home address.

2.4. Data Analysis

All analyses were conducted using aggregate data at the census tract level. Therefore, the census tract was the unit of analysis. Out of 707 census tracts, 12 census tracts (1.5%) were missing kindergarten readiness scores, likely because kindergarten-age students did not reside in these census tracts. Because the missingness was less than 5%, we dealt with this using pairwise deletion in analyses as per the recommendations of Schafer and Graham [59].

To examine research question 1 (a descriptive picture of the environmental burden exposure across Miami-Dade County), our study employed descriptive statistics including the mean, standard deviation, median, and range. The number of “high burden” census tracts for the Environmental Burden Module, all five environmental domains, and all environmental burden indicators in Miami-Dade County was also calculated.

To examine research question 2 (associations between neighborhood demographic variables and environmental burden), our study employed four sets of Analysis of Variance (ANOVAs) with demographics and socioeconomic vulnerability categories as the grouping variables and the environmental burden domains as the dependent variable. Black and White groups were tested in two separate ANOVAs. Results of the omnibus F tests and Games-Howell post hoc comparisons were conducted to examine mean differences between groups.

To examine research question 3 (associations between a neighborhood’s environmental burden and kindergarten readiness), our study employed a hierarchical multiple linear regression model to assess the extent to which race and ethnicity makeup, exposure to socioeconomic vulnerability, and exposure to environmental burdens predicted kindergarten readiness scores. More specifically, race and ethnicity variables were entered in the first step to control for their effects on kindergarten readiness scores. Child Raising Vulnerability Index scores were entered in the second step to determine the predictive power of exposure to socioeconomic vulnerability above and beyond race and ethnicity alone. Finally, the five domain scores of the Environmental Burden Module of the EJI were entered in the third step, to determine their ability to predict kindergarten readiness above and beyond race, ethnicity and exposure to socioeconomic vulnerability experiences. The changes in r² values and significance level were evaluated between each block entered across the three distinct regression models.

All assumptions were checked and met except for the homogeneity of variance for some of the ANOVAs. To account for this violation and the increased risk of Type I error, we used the Games–Howell test to perform the post hoc analyses [60]. All analyses were conducted using IBM SPSS Statistics (Version 29) [61].

3. Results

Descriptive results are reported, followed by the results separated by research questions.

Table 1. Descriptive statistics for key variables.

Variable	N (Census Tracts)	Mean (SD)	Range
Percent White	698	57.8 (20)	0–100
Percent Black	698	15.3 (23.6)	0–95.7
Percent Hispanic	698	68.1 (24.6)	4.3–100
Child Raising Vulnerability Index	707	2.32 (1.96)	0–9
Kindergarten Readiness Screener (FLKRS)	695	522.6 (51.6)	388–877

Note. Out of 707 total census tracts, 695 had available kindergarten readiness data.

presents the descriptive statistics for race and ethnicity, socioeconomic vulnerability, and the FLKRS scores by census tract.

Table 2. Correlations between key variables.

	1	2	3	4	5	6	7	8	9	10
1. Percent White	-
2. Percent Black	−.29 **	-
3. Percent Hispanic	.32 **	−.74 **	-
4. Child Raising Vulnerability Index	.01	.33**	.08 *	-
5. Air Pollution	−.10 *	.20 **	.08*	.34**	-
6. Potentially Hazardous and Toxic Sites	−.03	.22 **	−.10 **	.24**	.44 **	-
7. Built Environment	−.08 *	.20 **	−.13 **	.14 **	.16 **	.07	-
8. Transportation Infrastructure	−.01	.09 *	−.02	.13 **	.34 **	.52 **	.11 **	-
9. Water Pollution	−.02	−.32**	.12**	−.15**	−.14**	.00	−.09 *	.10 **	-
10. Florida Kindergarten Readiness Screener (FLKRS)	−.03	−.27**	−.01	−.47**	−.24**	−.25 **	−.08 *	−.12 **	.22 **	-

Note: * p < .05, ** p < .01, *** p < .001.

presents the bivariate correlations between all the study variables. The FLKRS scores were significantly correlated with percent Black (r = −.27, p < .01), the Bruner Child Raising Vulnerability Index (r = −.47, p < .01), the air pollution domain (r = −.24, p < .01), the potentially hazardous and toxic sites domain (r = −.25, p < .01), the built environment domain (r = −.08, p < .05), the transportation infrastructure domain (r = −.12, p < .01), and the water pollution domain (r = .22, p < .01).

3.1. Descriptive Picture of Environmental Burden

Descriptive statistics for the EJI Environmental Burden Module domains and indicators are presented in

Table 3. Descriptive statistics for Environmental Burden Module domains and indicators.

Domains and Indicators	N	Mean (SD)	Median	Min	Max	# High Burden
Air Pollution	704	.39 (.17)	.41	.02	.61	0
Ozone	706	.20 (.15)	.32	0	.32	0
Particulate Matter (PM2.5)	706	.00 (.00)	0	0	0	0
Diesel Particulate Matter	704	.82 (.16)	.87	.08	.98	544
Probability of Developing Cancer	704	.19 (.21)	.04	0	.50	0
Potentially Hazardous and Toxic Sites	706	.38 (.33)	.34	0	1	126
National Priority List Site	706	.15 (.35)	0	0	1	109
Toxic Release Inventory Site	706	.35 (.36)	.31	0	.86	166
Treatment, Storage, and Disposal Facility	706	.31 (.36)	0	0	.84	139
Risk Management Plan Site	706	.16 (.34)	0	0	.98	105
Coal Mine	706	.00 (.00)	0	0	0	0
Lead Mine	706	.00 (.00)	0	0	0	0
Built Environment	697	.22 (.18)	.18	0	.80	2
Lack of Park or Greenspace	706	.56 (.05)	.55	.55	.92	15
Exposure to Lead	697	.50 (.32)	.55	0	1	212
Lack of Walkability	706	.19 (.16)	.15	0	1	6
Transportation Infrastructure	706	.49 (.28)	.54	0	1	76
Railway	706	.41 (.36)	.49	0	.81	234
High Volume Roadway or Highway	706	.54 (.16)	.61	0	.61	0
Airport	706	.11 (.31)	0	0	.99	85
Water Pollution	706	.67 (.36)	.87	0	.89	499
Impaired or Impacted Watershed	706	.67 (.36)	.87	0	.89	499
Environmental Burden Total Score	697	.36 (.28)	.30	0	1.00	90

Note: N = census tracts in Miami-Dade County. Census tracts are considered High Burden if they score at or above the 75th percentile compared nationally to all other census tracts in the U.S.

. Because the Environmental Burden Module scores are percentile ranks normed using national data, the means and standard deviations reported give a picture of how Miami-Dade County compared to the nation overall. On average, Miami-Dade County scored below the 50th percentile on the following domains: air pollution, potentially hazardous and toxic sites, built environment, and transportation infrastructure, compared to the rest of the United States. Miami-Dade County also scored below the 50th percentile on the overall total Environmental Burden Module score compared to the rest of the United States.

When examining each domain score individually, Miami-Dade County scored at or above the 50th percentile on the water pollution domain and the following indicators within that domain: amount of diesel particulate matter, lack of park or greenspace, exposure to lead, proximity to a high-volume roadway or highway, and proximity to an impaired or impacted watershed compared to the rest of the United States. Additionally, the number of high-burden census tracts was calculated by counting the number of census tracts that were at or above the 75th percentile. More than a quarter of the census tracts in Miami-Dade County were categorized as high-burden census tracts for the following indicators: amount of diesel particulate matter, exposure to lead, proximity to a railway, and proximity to an impaired or impacted watershed.

3.2. Associations Between Environmental Burden and Neighborhood Demographic Variables

Table 4. Mean (standard deviation) of Environmental Burden Domain scores by census tract demographic characteristics.

	Air Pollution	Hazardous and Toxic Sites	Built Environment	Transportation Infrastructure	Water Pollution	Total Environmental Burden
	M (SD)	M (SD)	M (SD)	M (SD)	M (SD)	M (SD)
Ethnicity
<25% Hispanic (n = 48)	.51 (.10) a	.59 (.32) a	.32 (.18) a	.54 (.21)	.45 (.41) a	.53 (.27) a
25–49% Hispanic (n = 118)	.39 (.14) b	.40 (.32) b	.25 (.18) a	.50 (.25)	.66 (.36) b	.37 (.26) b
50–74% Hispanic (n = 196)	.29 (.15) c	.34 (.28) b	.20 (.18) a,b	.48 (.28)	.70 (.33) b	.29 (.24) c
>75% Hispanic (n = 336)	.43 (.17) d	.36 (.35) b	.21(.18) b	.49 (.29)	.67 (.36) b	.38 (.30) b
Race
<25% Black non-Hispanic (n = 557)	.38 (.17) a	.34 (.32) a	.20 (.18) a	.48 (.28) a	.71 (.33) a	.34 (.27) a
25–49% Black non-Hispanic (n = 55)	.36 (.19) a	.47(.31) b	.22 (.17) a	.54 (.28) a	.55 (.37) b	.40 (.30) a
50–74% Black non-Hispanic (n = 58)	.48 (.11) b	.48 (.37) b	.31 (.18) b	.56 (.23) b	.49 (.40) b	.47 (.30) b
>75% Black non-Hispanic (n = 28)	.57 (.07) c	.65 (.31) b	.36 (.14) b	.54 (.22) a	.34 (.41) b	.58 (.29) b
<25% White non-Hispanic (n = 557)	.42 (.18) a	.39(.34) a	.22 (.18)	.50 (.28) a	.62 (.38) a	.38 (.30) a
25–49% White non-Hispanic (n = 111)	.29 (.12) b	.36 (.27) a	.21 (.18)	.51 (.25) a	.83 (.20) b	.32 (.20) b
>50% White non-Hispanic (n = 30)	.32 (.32) b	.26 (.24) b	.21 (.18)	.37 (.24) b	.85 (.16) b	.25 (.16) b
Child Raising Vulnerability
No vulnerabilities (n = 118)	.30 (.13) a	.31 (.29) a	.21 (.18) a	.49 (.27) a	.77 (.28) a	.30 (.21) a
1–2 vulnerabilities (n = 311)	.37 (.17) b	.32 (.31) a	.19 (.18) a	.46 (.28) a	.67 (.36) b	.31 (.26) a
3–5 vulnerabilities (n = 211)	.45 (.17) c	.43 (.35) b	.25 (.19) c	.51 (.27) a,b	.62 (.37) b	.43 (.31) b
6 or more vulnerabilities (n = 64)	.49 (.15) c	.56 (.35) c	.26 (.15) a,b	.61 (.26) c	.60 (.38) b	.54 (.30) c

N = 707. Note: Means with different superscripts are significantly different from each other, p < .05. n’s represent the number of census tracts.

presents the differences across the Environmental Burden Domain scores by race, ethnicity, and socioeconomic vulnerability. The air pollution scores differed by percent of Hispanic residents living in the census tract (F(3, 694) = 43.2, p < .001, $\eta$² = .16), percent of Black non-Hispanic people living in the census tract (F(3, 694) = 18.7, p < .001, $\eta$² = .08), percent of White non-Hispanic people living in the census tract (F(2, 695) = 27.9, p < .001, $\eta$² = .07), and by socioeconomic vulnerability level (F(3, 700) = 29.1, p < .001, $\eta$² = .11). Census tracts with a lower percent of Hispanic residents and a lower percent of White non-Hispanic people had higher mean air pollution scores. Census tracts with a higher percentage of Black non-Hispanic people had a higher mean air pollution score. Census tracts with more socioeconomic vulnerabilities had a higher mean air pollution score.

The potentially hazardous and toxic sites scores differed by percent of Hispanic people living in the census tract (F(3, 694) = 8.50, p < .001, $\eta$² = .04), percent of Black non-Hispanic people living in the census tract (F(3, 694) = 12.1, p < .001, $\eta$² = .05), and socioeconomic vulnerability level (F(3, 702) = 13.9, p < .001, $\eta$² = .06). Census tracts with a lower percentage of Hispanic people had higher mean potentially hazardous and toxic sites scores. Census tracts with a lower percentage of Black non-Hispanic people had a lower mean potentially hazardous and toxic sites score. Census tracts with a higher percentage of White non-Hispanic residents had a lower mean potentially hazardous and toxic sites scores. Census tracts with more socioeconomic vulnerabilities had a higher mean potentially hazardous and toxic sites score.

The built environment scores differed by percent of Hispanic people living in the census tract (F(3, 693) = 7.79, p < .001, $\eta$² = .03), percent of Black non-Hispanic people living in the census tract (F(3, 693) = 11.8, p < .001, $\eta$² = .05), and socioeconomic vulnerability level (F(3, 693) = 4.99, p = .002, $\eta$² = .02). Census tracts with a lower percent of Hispanic people had a higher mean built environment score. Census tracts with a lower percentage of Black non-Hispanic people had a lower mean built environment score. Census tracts with more socioeconomic vulnerabilities had a higher mean built environment score.

The transportation infrastructure scores differed by percent of White non-Hispanic people living in the census tract (F(2, 695) = 3.25, p = .04, $\eta$² = .01) and socioeconomic vulnerability level (F(3, 702) = 6.30, p < .001, $\eta$² = .03). Census tracts with a lower percent of White non-Hispanic people had higher mean transportation infrastructure scores. Census tracts with more socioeconomic vulnerabilities had a higher mean transportation infrastructure score. Although the omnibus test of the differences in transportation infrastructure scores by the percent of Black non-Hispanic people living in the census tract was not significant (F(3, 694) = 2.43, p = .06, $\eta$² = .05), post hoc analyses revealed that the mean for census tracts comprised of 50–74% Hispanic residents was significantly different from the means of all other groups.

The water pollution scores differed by the percent of Hispanic people living in the census tract (F(3, 694) = 6.53, p < .001, $\eta$² = .03), percent of Black non-Hispanic people living in the census tract (F(3, 694) = 18.5, p < .001, $\eta$² = .07), percent of White non-Hispanic people living in the census tract (F(2, 695) = 21.4, p < .001, $\eta$² = .06), and socioeconomic vulnerability level (F(3, 702) = 5.52, p < .001, $\eta$² = .02). Census tracts with a lower percent of Hispanic people and census tracts with a lower percent of White non-Hispanic people had lower mean water pollution scores, whereas census tracts with a lower percent of Black non-Hispanic people had a higher mean water pollution score. Census tracts with fewer socioeconomic vulnerabilities had a higher mean water pollution score.

The Total Environmental Burden scores differed by the percentage of Hispanic people living in the census tract (F(3, 693) = 10.8, p < .001, $\eta$² = .05), percent of Black non-Hispanic people living in the census tract (F(3, 693) = 10.5, p < .001, $\eta$² = .04), percent of White non-Hispanic people living in the census tract (F(2, 694) = 4.59, p = .01, $\eta$² = .01), and by socioeconomic vulnerability level (F(3, 693) = 19.5, p < .001, $\eta$² = .08). Census tracts with a lower percentage of Hispanic people and census tracts with a lower percentage of White non-Hispanic people had higher mean Total Environmental Burden scores. Census tracts with a lower percentage of Black non-Hispanic people had a lower mean Total Environmental Burden score. Census tracts with more socioeconomic vulnerabilities had a higher mean Total Environmental Burden score. All of the aforementioned effect sizes ranged from medium to large, suggesting a significant practical effect [62].

3.3. Associations Between Environmental Burden and School Readiness

The overall hierarchical regression model was statistically significant, $r^2=0.28,F\left(\mathrm{11,405}\right)=4.65,p<.001$, wherein race, ethnicity, socioeconomic vulnerability, and the domain scores of the Environmental Burden Module explained 28% of the variance in the census tract level FLKRS scores, which is a moderate effect size and is consistent with expectations, especially in the context of social science research looking at neighborhood-level influences on children [4,62,63,64,65].

In step one, all the race and ethnicity variables were statistically significant predictors of FLKRS scores, $r^2=0.17,F\left(6, 410\right)=48.3,p<.001$. Specifically, census tracts with a higher percentage of White residents had significantly higher FLKRS scores ($B=.55,p<.01$). Census tracts with a higher percentage of Black residents had significantly lower FLKRS scores ($B=-.82,p<.001$). Census tracts with a higher percentage of Hispanic residents had significantly lower FLKRS scores ($B=-.84,p<.001$).

Including the Bruner Child Raising Vulnerability Index in the model in step two yielded a statistically significant increase in r² of 0.08, ($\mathsf{\Delta }F\left(3, 407\right)=72.3,p<.001),$ meaning that the socioeconomic vulnerability index score significantly predicted FLKRS scores above and beyond race and ethnicity. Within this model, census tracts with a higher percentage of White residents and with a higher percentage of Hispanic residents remained as a significant predictor ($B=.51,p<.01; B=-.30,p<.05;$ respectively). In addition, the Child Raising Vulnerability Index scores were significantly predictive of FLKRS scores ($B=-9.24,p<.001$). Specifically, children living in census tracts with higher levels of socioeconomic vulnerability had lower average FLKRS scores.

Finally, inclusion of the domains of the Environmental Burden Module in step three led to a statistically significant increase in r² of .03, $\mathsf{\Delta }F\left(2, 405\right)=4.65,p<.001$, meaning the domain scores of the Environmental Burden Modules significantly predicted FLKRS scores above and beyond race and ethnicity and the Child Raising Vulnerability Index. Within this model, the percentage White ($B=.44,p<.05$), the Child Raising Vulnerability Index ($B=-8.83,p<.001$) remained significant predictors of FLKRS scores. Additionally, the potentially hazardous and toxic sites and water pollution domains predicted FLKRS scores ($B=-20.8,p<.01; B=16.1,p<.01;$ respectively). Specifically, children living in census tracts near potentially hazardous and toxic sites had lower average FLKRS scores. Children living in census tracts close to impaired or impacted watersheds had higher average FLKRS scores. Post hoc analyses, including the individual potentially hazardous and toxic sites indicators, revealed proximity to treatment storage and disposal sites and proximity to risk management plan sites as the unique indicator predictors of FLKRS scores ($B=-.19,p<.01; B=-.11,p<.05;$ respectively). See

Table 5. Hierarchical regression results predicting the Kindergarten Readiness Screener (FLKRS) Score.

Predictor Variables	Florida Kindergarten Readiness Screener (FLRKS) Score
	B	SE B	$\mathit{\beta }$
Step 1: Demographic Characteristics
Percent White	.55 **	.20	.21
Percent Black	−.82 ***	.22	−.38
Percent Hispanic	−.84 ***	.12	−.40
R2	.17
F	48.3 ***
Step 2: Child Raising Vulnerability
Percent White	.51 **	.19	.20
Percent Black	−.18	.22	.08
Percent Hispanic	−.30 *	.13	−.14
Child Raising Vulnerability Count	−9.24 ***	1.09	−.35
R2	.25
F	58.0 ***
ΔR2	.08
ΔF	72.3 ***
Step 3: Environmental Burden Module Domains
Percent White	.44 *	.19	.17
Percent Black	−.06	.22	−.03
Percent Hispanic	−.24	.13	−.11
Child Raising Vulnerability Count	−8.83 ***	1.09	−.34
Air Pollution	.25	11.9	.00
Potentially Hazardous and Toxic Sites	−20.8 **	6.46	−.13
Built Environment	1.64	9.44	.01
Transportation Infrastructure	.00	7.28	.00
Water Pollution	16.1 **	5.16	.11
R2	.28
F	29.1 ***
ΔR2	.03
ΔF	4.65 ***

Note: * p < .05 ** p < .01 *** p < .001.

for all the regression results.

4. Discussion

This population-based study capitalized on integrated data at the census tract level to examine the environmental burden across Miami-Dade County and associations with demographic variables and kindergarten literacy and mathematics skills. Overall, we found a substantial variation in the environmental burden across the county [3], with a higher than previously documented number of census tracts experiencing air and water pollution, known predictors of poor health and developmental outcomes for children [29,32,66]. We also found that areas of the county with a higher percentage of minoritized and socioeconomically disadvantaged residents were more likely to experience higher levels of environmental risks such as air pollution, noise pollution from railroads, train tracks, and highways. Racial and ethnic minorities were also less likely to live near positive influences on community health, such as parks or green spaces. Below, we review and contextualize the findings as they contribute to sustainable strategies to reduce environmental risk exposures and promote healthier communities for the positive development of young children.

4.1. Descriptive Picture of Environmental Burden

Overall, at the population level, a surprisingly high number of census tracts experienced environmental burden relative to the state and nation. Consistent with our hypotheses, we identified geographic variation across the EJI indicators of environmental risk and indicators associated with community health, such as greenness and walkability. Some census tracts experienced high exposure to diesel particulate matter and proximity to railways, while others experienced high lead exposures or impaired watersheds. According to the EJI data, 25.7% of residents in Miami-Dade County lived in highly burdened areas, compared to 18.3% in Florida and 22.9% in the U.S. [3]. This percentage represents a sizable number of people living in the county, especially young children. In Miami-Dade County, in the year in which data for this study were collected, kindergarten students lived in 695 of the County’s 704 census tracts. These findings call on us as researchers to ask, what impact does the widely distributed risk exposure have on the 140,000 children [53] under 5 years of age who live and grow up in Miami-Dade County? What implications does this level of environmental risk have for the promotion of healthy communities where children live?

Of note, within Miami-Dade County, the air pollution levels were elevated, with an alarmingly high number of census tracts (544 out of 704) exposed to high levels of diesel particulate matter (measured by the mean number of days above the U.S. national regulatory standard for particulate matter 2.5 [3]. The median census tract exposure was at the 87th percentile rank compared to national norms, with census tracts ranging in exposure from the 8th percentile to 98th percentile rank. While we expected to find areas exposed to higher levels of environmental risk based on prior research [18,19], we did not expect to find so many census tracts exposed to such high levels of lead and diesel particulate matter compared to national norms.

Surprisingly, the average exposure to potentially hazardous and toxic sites in Miami-Dade County was lower than the national average. Fewer than a quarter of the census tracts were at or above the 75th percentile on any indicators for this EJI domain. However, two indicators in this domain had the greatest number of high-burden census tracts in the county: proximity to toxic release sites and proximity to treatment, storage, and disposal facilities. Proximity to toxic release inventory sites affected 166 census tracts, and 139 tracts were near treatment, storage, and disposal facilities. Toxic release inventory sites are facilities with high levels of the manufacture and/or usage of toxic chemicals. Treatment, storage, and disposal facilities are responsible for handling hazardous waste throughout collection, transfer, and ultimately their disposal. What this means is that children’s level of toxic exposures is concentrated in these census tracts and experienced by a more targeted group of children in the county. Children living in these high-burden census tracts facing multiple exposures over prolonged periods of time may have a higher lifetime chance of experiencing cardiovascular disease, reproductive birth defects, neurodevelopmental disorders, or cancers [67].

Concerning the built environment, on average, we found that most areas of the County had less access to green space when compared to the rest of the U.S. The proportion of census tracts not within a 1-mile buffer of a park, recreational area, or public forest fell at the 56th percentile compared to the nation, and ranged from the 55th to 92nd percentile. This aligns with prior research conducted by Brown et al. (2016) that found heterogeneity across the county in access to green space [68]. Although Miami is one of the major urban counties in the U.S., it is unexpectedly low in terms of the percentage of areas covered by an urban green canopy compared to other U.S. cities, and substantial areas remain with limited access to green spaces. In response, the Miami-Dade County Parks and Open Spaces Master Plan recently focused on the goal of creating a “seamless, sustainable system of parks, recreation and conservation open spaces for this and future generations.” One of the foci of the 50-year plan is to increase green coverage to 30%, which has been difficult given current policies and efforts to deregulate environmental protections [69].

In terms of exposure to lead, measured on the EJI by the percentage of occupied housing units built before 1980 in census tracts, on average, Miami-Dade County ranked at the 50th percentile compared to the national average. However, over a quarter of the census tracts (212 of 697) were considered high burden (͕>0.75). Studies have historically reported on lead prevalence rates and health effects in children using a lead cutoff of ≥10 μg/dl. However, inverse associations have been identified between even lower blood lead levels (<5 μg/dL) and neurological deficits such as poor executive functioning across childhood, confirming there is no safe blood lead concentration for children [70,71].

The transportation infrastructure domain on the EJI refers to hazardous exposures associated with transportation, such as car emissions, noise pollution, and safety risks from accidents and train derailments. Miami-Dade County census tracts ranked at the 41st percentile compared to the national average; however, almost a third of the census tracts (234 or 706) were high burden (>0.75) for exposure to railways. This high level of exposure to railways in certain areas is a concern because research finds negative health and developmental effects on children. Road-traffic and aircraft noise exposure is adversely associated with blood pressure (cardiovascular), stress hormones (neuroendocrine), memory and cognitive functioning (neurodevelopment), sleep quality, and behavior (mental health). Given the amount of time children spend indoors, other factors such as acoustics, noise frequency and duration, and child coping strategies should be considered [72]. Negative effects have been moderated by positive family relations, neighborhood quality, and closeness to green space [73].

On average, Miami-Dade County ranked at the 54th percentile compared to the national average on exposure to high-volume roadways or highways. High volume roadways are also sources of traffic-related air pollution (TRAP), which contain harmful emissions containing nitrogen oxide (NOx), particulate matter, and other toxic air pollutants. Traffic-related air pollution, combined with sunlight and heat, contributes to ground-level ozone (O3), which can damage lung tissue and weaken the immune system, compromising the ability for young children to fight infection and minimize inflammation [74]. Stowell et al. (2024) observed associations between higher ground-level ozone and emergency department visits for children in a national dataset [75]. Respiratory and ear disorders, allergic disease, and asthma rates were associated with increased ambient ground-level ozone levels and persisted at levels below the current U.S. National Ambient Air Quality Standards (NAASQ) (of 70 ppb) [76]. Our study findings suggest that in Miami-Dade a percentage of children in certain areas of higher burden are exposed to higher ground-level ozone (O3) and higher levels of air pollutants. This is of concern, given Stowell et al.’s (2024) recommendations that the NAASQ thresholds may not be protective for children and any level of exposure can be harmful [75].

In addition to air pollution, water pollution levels were elevated in the county. Water pollution is measured on the EJI by impaired surface water. Impaired surface water in rivers, lakes, or reservoirs can be contaminated by industrial chemical discharges, agricultural pesticides or runoff, sewage overflows, and waterborne pathogens [3]. Impaired surface water can lead to contaminated drinking water, declines in healthy fish population, and reduction in water recreational activities (EJI). On average Miami-Dade County ranked at the 67th percentile compared to the national average, and over 70% of the census tracts were considered high burden (>0.75). Many areas of Miami-Dade County are surrounded by water, and its drinking water is vulnerable because urban and agricultural freshwater supply relies on the Biscayne Aquifer, a porous limestone formation underlying South Florida. The aquifer is replenished by rain and flows from the Everglades. The porous aquifer is vulnerable to contamination and especially vulnerable to the effects of climate change, with saltwater intrusion due to sea level rise and other human development activities on land [77]. Extreme heat and flooding, increasingly experienced in Miami-Dade County, further contaminate surface water and disrupt the living environment and quality of life for vulnerable children living in higher burden tracts [63]. Everyday activities such as cooking, cleaning, and outside recreational play for children are compromised. Exposure to contaminated water can lead to gastrointestinal and neurological illnesses, skin problems, and bloodstream infections, particularly in children [67,78].

4.2. Environmental Burden and Neighborhood Demographics

Several neighborhood sociodemographic characteristics were associated with environmental burden, aligned with our hypotheses. We expected, given the historic racism, redlining, and marginalization of Black communities in Miami-Dade, that predominantly Black non-Hispanic census tracts would disproportionately experience an environmental burden. At the census tract level, the environmental burden scores were associated with race, ethnicity, and socioeconomic vulnerability as measured by the Bruner Child Raising Vulnerability Index [55]. Census tracts with higher numbers of Black non-Hispanic residents consistently experienced higher levels of air pollution, potentially hazardous and toxic sites, a hazardous built environment, and total environmental burden than census tracts with higher numbers of Hispanic and White non-Hispanic residents.

Our findings replicated prior work documenting inequitable distribution of the environmental burden, concentrated in areas with a higher percentage of Black residents, especially those living in poverty [18]. In Miami-Dade County, these findings reflect persistent effects of historic redlining practices resulting from the 1935–1940s Federal Homeowners’ Loan Corporation, which limited residential neighborhoods within Miami-Dade County for Black Miamians. During the 1935–1940s, mortgage lending practices drove investments and land development up in desirable predominantly White communities and disenfranchised Black communities. Public transportation practices and the installation of the I-95 interstate highway overpass further devastated the flourishing popular Black community of Overtown, resulting in the mass displacement of tens of thousands of residents [79]. The influx of Cuban immigrants in 1959 and again in a second wave in 1980 increased the Hispanic/Latine population in Miami-Dade, which settled and thrived in Miami Beach [80]. Our census tract data resemble the shift in race and ethnicity, which majoritizes the Hispanic population and minoritizes the Black population. The Jim Crow era and ensuing segregation historically explain the environmental racism seen today in the predominantly Black residential Miami-Dade census tracts. Similarly, census tracts with more socioeconomic vulnerabilities consistently experienced higher levels of air pollution, potentially hazardous and toxic sites, and total environmental burden than census tracts with fewer socioeconomic vulnerabilities. Children living in these census tracts are experiencing compounded risks to their development and learning, including socioeconomic and environmental.

Surprisingly, water pollution exposure did not follow this pattern. We found that the environmental risk experienced by proximity to polluted watersheds was highest in non-Hispanic White and Hispanic communities. Census tracts with a higher percentage of White non-Hispanic and Hispanic residents showed higher scores on the EJI water pollution domain. Census tracts with no or lower socioeconomic vulnerabilities on the Bruner index also showed higher scores on water pollution. Geographically, this pattern could reflect that non-Hispanic White and Hispanic communities may live in census tracts closer to bodies of water. South Florida water canals and aquifers have documented high pesticide residue, volatile organic compounds (VOCs) from personal care and home cleaning products, building materials, and nutrients from storm runoff affecting both surface and ground water for waterfront residents [81]. In Miami-Dade, many areas where individuals live are in proximity to high levels of water pollution. Higher resourced and more affluent areas tend to be more desirable to live in and are located near bodies of water in the county, such as the Atlantic Ocean, Biscayne Bay, or the Intracoastal Waterway [82]. However, affluent areas remain relatively insulated due to being highly resourced, receiving additional political backing for improvements and less overall economic burden [83].

4.3. Environmental Burden and School Readiness

Census tracts with a higher percentage of Black and Hispanic residents had lower kindergarten readiness scores than census tracts with a higher percentage of White non-Hispanic residents. This finding is consistent with the existing literature documenting a persistent school readiness gap at kindergarten entry, with Black and Hispanic students and children of lower socioeconomic status demonstrating lower early academic skills than White students and those of higher income [84].

Above and beyond sociodemographic vulnerability, several environmental risks were associated with kindergarten readiness scores at the census tract level. Environmental burden indicators measuring proximity to potentially hazardous and toxic sites predicted lower kindergarten readiness scores above and beyond socioeconomic vulnerability and demographics. The treatment, storage, and disposal facility (hazardous waste: volatile substances generated by waste may become aerosolized or migrate into soil and water) and risk management plan site (product use or storage of highly toxic or flammable chemicals) were uniquely negatively predictive of kindergarten readiness scores. [85].

Again, contrary to our expectations, census tracts close to polluted water were associated with higher kindergarten readiness scores. Like our findings above, related to socioeconomic vulnerability, it could be that higher levels of water pollution in Miami-Dade County are found closer to areas adjacent to waterways or oceans, which may be closer to wealthier neighborhoods. In Miami-Dade, desirable and more expensive real estate tends to be located near water. As we mentioned earlier, because risk and protective factors tend to correlate and interact with one another in predicting child outcomes [43,48], it is most likely that kindergarten children living in these wealthier census tracts had access to more early learning resources and thus tended to perform better academically than children in less advantaged neighborhoods [86].

Of note, while each of the three sets of variables (demographic, socioeconomic, and environmental risks) were entered into the final model, socioeconomic vulnerability (as measured by the Bruner Child Raising Vulnerability Index) continued to explain the most amount of variance in kindergarten scores. This means that socioeconomic vulnerability within children’s communities measured by economic, social, and educational risks such as a higher percentage of children living in poverty, on public assistance, single parents, or unemployed adults, and a lower percentage of adults with a high school diploma or college degree were the most powerful predictors of kindergarten readiness, above and beyond the racial and ethnic composition of the census tract and above and beyond the environmental burden. While these three sets of variables are correlated, the robust prediction of socioeconomic vulnerability indicators such as those measured by the Bruner Child Raising Vulnerability Index suggests a strong persistent effect at the community-level on children’s early academic skills above and beyond other factors. The findings align with research documenting consistent negative associations at the child level between poverty and lower socioeconomic status and kindergarten developmental skills, such as phonological skills [37], executive functioning skills [38], language skills [39], and overall measures of kindergarten readiness [41,50].

4.4. Application to Sustainable Community Development, Practice, and Policy

Our study is responsive to the urgent call from the United Nations for policies that ensure healthy lives and promote well-being for all ages, ensure inclusive and equitable quality education, and make cities and human settlements inclusive, safe, resilient, and sustainable for all residents [1]. Our results show that many children living in Miami-Dade County communities experience high exposure to toxic chemicals, pollution, and other environmental risks. From a public health perspective, it is critical to identify where high levels of environmental exposure affect our youngest citizens at their most formative times in development. Public health and environmental interventions that reduce toxicant exposure, increase greenness and access to safe open spaces, and reduce the impacts of climate change such as global warming can prevent and promote healthier human development, health, and learning—an ounce of prevention early in development is most cost-effective in the long run [87,88].

Consistent with prior research, we found that predominantly Black non-Hispanic census tracts had higher levels of adverse environmental exposure, higher socioeconomic vulnerability, and lower kindergarten readiness. This finding underscores persistent racial segregation, inequities in residential access to resources, and exposures that promote healthy development [18], particularly for Black children. Due to historical shifts in demographics since the 1960s, even though Miami-Dade County is now majority minority, Black communities continue to be more disadvantaged compared to Hispanic communities. More policy attention should be paid to these persistent inequities to address environmental exposures and infuse greater access to resources that support early learning and development. A specific strategy might be for researchers to collaborate with local community leaders and place-based initiatives (e.g., Overtown Children and Youth Coalition) to co-create research agendas that drive collective action and change.

Using a strength-based approach, government policymakers collaborating with community leaders, and researchers can identify assets that could be strengthened and more resources that could be infused in under-resourced communities. In addition, aligned with recommendations by Stingone et al. (2023), ways to create healthy communities that support positive child and family development must include multiple community-level approaches to systematically measure the totality of environmental exposures. Community voices must be well represented in the design and assessment of all procedures to ensure that any findings and future interventions are representative, equitable, and actionable [63].

Another practical application of our findings is that community health assessments need to include measures of children’s learning and development, as well as community access to quality early childhood education. Typically, health impact assessments focus on risks that interfere with children’s healthy outcomes, such as preventable deaths of newborns and children under five, respiratory or chronic illnesses, and access to medications and vaccines. When municipalities such as Miami-Dade County conduct environmental or community health assessments, they should include additional child factors like kindergarten outcomes or access to quality early learning programs [89]. Urban planners and public health professionals who integrate complex environmental, social, and health data can meet the needs of young learners while accounting for geographic and social variability. Additionally, utilizing large administrative datasets such as the EJI, as was used in the current study, municipalities can anticipate vulnerabilities during disasters to build community resilience and prioritize policy changes to reduce health disparities [90]. The U.S. Department of Health and Human Services, Office of Disease Prevention and Health Promotion, highlights in “Healthy People 2030” that a key objective to improve health and well-being over the next decade is to increase educational opportunities and help children and adolescents do well in school and, specifically, to increase the proportion of children who are developmentally ready for school [91].

Access to quality prekindergarten or “early learning” programs may help to narrow the racial and socioeconomic gap we observed in our study. Research documents the short- and long-term effects of high quality early learning programs on the academic and social-emotional skills of children 0 to 5 years old [92]. Research suggests that, compared to those children who do not attend high-quality early learning programs, those who attend early learning programs enter kindergarten with higher early literacy, mathematics, and social-emotional skills [93]. For children living in poverty, access to high quality programs is especially important, for whom additional benefits are observed [94,95]. However, disparities in access to quality programs in the U.S. and globally must be addressed for sustainable community promotion of early development and learning [96,97].

4.5. Strengths and Limitations

While our study findings advance our knowledge base in several ways, several limitations must be acknowledged. First, our findings are limited to Miami-Dade County, an ethnically and racially diverse southeastern urban community with a unique history of urbanization, structural racism, and immigration. While the diverse language, ethnic, and racial composition of Miami-Dade County may be a microcosm of the future demographics of U.S. cities, our findings are limited in generalizability to Miami-Dade County. The findings provide an initial descriptive picture of the level of environmental burden experienced across the entire population of Miami-Dade County and its disproportionate associations with sociodemographic variables and kindergarten readiness. Future research will need to replicate these findings in other samples or other cohorts of children in Miami-Dade County and other urban counties in the U.S.

Many risks co-occur within communities. It is difficult to disentangle the unique versus cumulative effects of environmental and socioeconomic risks on communities and children. In addition, our analysis was correlational and cross-sectional. We can only assume associations between our independent variables and kindergarten outcomes and cannot make causal inferences.

In addition, while we used psychometrically sound multidimensional measurement tools from multiple data sources in our statistical models, there are limitations to census tract measurement of our key independent variables in the Bruner and EJI. While U.S. census estimates are more reliable at the census tract than block level, there can be limitations in that actual data collected, especially in some underrepresented groups like newly immigrated families who may not feel comfortable responding to the census [98]. Another limitation is that our kindergarten readiness assessment included only early academic skills. The Florida Kindergarten Readiness Screener measures early mathematics, literacy, and language skills as mandated by the Florida Department of Education.

5. Future Directions

Future research should examine other variables that could be responsible for the associations we found with kindergarten readiness. For example, research suggests that other social environmental variables such as social capital, collective efficacy, and perceived neighborhood safety are associated with the environmental burden and children’s early learning and kindergarten readiness [4,99]. Future studies could collect questionnaires from neighborhood residents on these factors and examine whether variation in these factors is associated with variation in environmental risk exposures and resources available for children within communities. In addition, future research can apply mediation or moderation statistical models to tease out whether these social environmental factors are mechanisms that influence direct associations between environmental factors and children’s development. To inform strength-based community interventions, mediating models can identify potential mechanisms through which environmental or socioeconomic factors influence children’s outcomes [100].

Future research should extend the measurement of children’s kindergarten readiness beyond academic skills that were measured by the state assessment in the current study. Researchers should include more comprehensive measures of developmental skills to provide a holistic picture of the influence of environmental burden on other skills foundational to kindergarten readiness, such as executive functioning, social skills, and physical development [101].

Finally, to inform sustainable community improvement plans and interventions, given the complexity of factors associated with the healthy development of children within communities, multilevel models can be used to examine other individual child, family, and community factors that influence children’s outcomes. For example, studies could examine the role of social capital and community collective efficacy that is affected by environmental risk exposures, disenfranchisement, and structural racism. Guided by an ecological framework, it would be important to study factors within the child, family, school, and community that directly and interactively predict variability in kindergarten outcomes. To inform community interventions that can strengthen families, mediating models should also examine potential mechanisms through which environmental burden or socioeconomic risks influence children’s outcomes.

6. Summary and Conclusions

To inform sustainable community strategies to promote healthy child development, this population-based study combined several datasets at the census tract level to examine geographic variation in environmental burden, socioeconomic vulnerability, and kindergarten readiness. The findings identify several challenges in sustainable equitable access to a healthy community for all children, including exposure to high levels of lead, air pollution, socioeconomic vulnerability, and low access to green spaces and parks. While our findings contribute to advancing the knowledge base in several ways, we highlight that areas of the county where a higher concentration of Black children live disproportionately experience greater socioeconomic risk and higher environmental burden than children in non-Hispanic White and Hispanic areas of the county. In addition, above and beyond socioeconomic vulnerability, children living in predominantly Black census tracts are additionally experiencing risks from potentially hazardous and toxic sites that relate to lower kindergarten readiness. Follow-up research is warranted to explore other potential community-level factors that can provide a more holistic picture of children and promote healthy early learning and development.