Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Mar 15.
Published in final edited form as: Environ Res. 2023 Jan 18;221:115295. doi: 10.1016/j.envres.2023.115295

Making the invisible visible: Using a qualitative system dynamics model to map disparities in cumulative environmental stressors and children’s neurodevelopment

Devon C Payne-Sturges 1, Ellis Ballard 2, Deborah A Cory-Slechta 3, Stephen B Thomas 4, Peter Hovmand 5
PMCID: PMC9957960  NIHMSID: NIHMS1867519  PMID: 36681143

Abstract

Background:

The combined effects of multiple environmental toxicants and social stressor exposures are widely recognized as important public health problems, likely contributing to health inequities. However, US policy makers at state and federal levels typically focus on one stressor exposure at a time and have failed to develop comprehensive strategies to reduce multiple co-occurring exposures, mitigate cumulative risks and prevent harm. This research aimed to move from considering disparate environmental stressors in isolation to mapping the links between environmental, economic, social and health outcomes as a dynamic complex system using children’s exposure to neurodevelopmental toxicants as an illustrative example. Such a model can be used to support a broad range of child developmental and environmental health policy stakeholders in improving their understanding of cumulative effects of multiple chemical, physical, biological and social environmental stressors as a complex system through a collaborative learning process.

Methods:

We used system dynamics (SD) group model building to develop a qualitative causal theory linking multiple interacting streams of social stressors and environmental neurotoxicants impacting children’s neurodevelopment. A 2 1/2-day interactive system dynamics workshop involving experts across multiple disciplines was convened to develop the model followed by qualitative survey on system insights.

Results:

The SD causal map covered seven interconnected themes: environmental exposures, social environment, health status, education, employment, housing and advocacy. Potential high leverage intervention points for reducing disparities in children’s cumulative neurotoxicant exposures and effects were identified. Workshop participants developed deeper level of understanding about the complexity of cumulative environmental health risks, increased their agreement about underlying causes, and enhanced their capabilities for integrating diverse forms of knowledge about the complex multi-level problem of cumulative chemical and non-chemical exposures.

Conclusion:

Group model building using SD can lead to important insights to into the sociological, policy, and institutional mechanisms through which disparities in cumulative impacts are transmitted, resisted, and understood.

Keywords: Neurodevelopment, Cumulative risk, Environmental exposures, Chemicals, System Dynamics, Group model building

Introduction/ Background

The combined effects of multiple environmental toxicants and social stressor exposures are widely recognized as important public health problems, likely contributing to health inequities [1, 2]. Significant research investments have been made to develop methods to assess and document the combined effects of multiple chemical exposures, and literature on the cumulative health effects of joint exposure to chemical and social stressors is growing [35]. Far too many neighborhoods in the US are located near current and former industrial sites, congested highways, industrial agricultural operations and excessive noise. Unfortunately, such neighborhoods are characterized as racially segregated enclaves with concentrated poverty and residents suffering from co-morbid chronic diseases. African Americans, Latinos, immigrants and other marginalized groups also must overcome the challenge of limited access to health care, life in a food desert, dilapidated housing, excessive police surveillance, and predatory lenders and institutional racism. Taken together these factors are associated with the social determinants of health inequities that interact with human biology to ultimately worsen quality of life and length of life for our most vulnerable residents. For years community advocates, stakeholders, and independent advisory bodies have urged the U.S. Environmental Protection Agency (EPA) and state environmental agencies to improve risk assessment and risk management practices to better account for multiple physical, chemical, biological, and psychosocial stressors that cumulatively impact community and population health.[2, 6] Recommendations focus on the need for more holistic and transdisciplinary approaches. However, policy actions on cumulative environmental health risk and impacts have not progressed. Environmental health policy continues to be based on single chemical risk assessment without taking the social and economic risk contributors or modifiers/enhancers into consideration.

An area for which the limitations of current environmental health risk assessment practice are well illustrated is children’s neurodevelopment which exhibits dynamic complexity, intergenerational effects and interdependent and causally linked nature of multiple factors [7]. Developmental disabilities affect 1 in 6 children in the US and their prevalence has increased over time; these include learning disabilities, sensory deficits, developmental delays, cerebral palsy, and autism, attention deficit and hyperactivity disorders (ADHD) [8]. Neurotoxic chemicals and pollutants are widespread and can be found in indoor and outdoor air, water, soil, food, household dust and consumer products. Prime examples of toxic chemicals that can contribute to learning, behavioral, or intellectual impairment, as well as specific neurodevelopmental disorders such as ADHD or autism spectrum disorder include lead, organophosphate (OP) pesticides, mercury, combustion-related air pollutants, PBDE flame retardants and PCBs [9], among others. These pollutant exposures may disproportionately influence the cognitive development of children in disadvantaged environments. Well known disparities in environmental exposures important to children’s health track along socioeconomic lines. [1028] In addition to independently influencing children’s cognitive development, social and neighborhood conditions can modify associations between environmental contaminant exposures and neurodevelopment. [29] For example poverty, maternal material hardship and poor diet have been shown to heighten the toxic effects of air pollutant and other chemical exposures on cognitive functioning of children.[5, 3035] Longitudinal studies of lead exposure show synergistic effects with lower socioeconomic status (SES), thus demonstrating that higher levels of lead exposure were associated with worse cognitive outcomes among children of low SES families [36, 37]. The traditional risk assessment approach is not a holistic approach; it does not account for multiple chemical exposures, underlying dynamic complexity and the interdependent and causally linked nature of socio-structural and interpersonal factors influencing neurodevelopmental risks over time. As a result, environmental health policy decisions based on traditional single chemical risk assessments may continue to contribute to or exacerbate social disparities in exposures as well as health effects [33, 38, 39].

Attention has been paid to the technical challenge of developing measurement tools that address multiple exposures over time. Yet the challenge goes deeper than technical tools to challenge the underlying paradigms through which public health regulatory agencies operate [40, 41]. The key to uncovering solutions to persistent problems, such as cumulative environmental health risk and health disparities, is to move away from looking at isolated events and their causes, and start to look at the underlying system (system structure), made up of interacting parts, whose behavior creates these disparities. Addressing complex and dynamic problems such as environmental health inequities from a systems perspective is consistent with calls by many scholars who focus on structural racism and environmental justice[4245]. Systems theory focuses on relationships and processes and not on singular, linear causation. Systems science methods enable investigators to examine the dynamic relationships of variables at multiple levels of analysis simultaneously, while also studying the impact of the non-linear behavior of the system as a whole over time[46]. Systems tools are used in conjunction with traditional empirical methods to integrate diverse sources of data, to generate and test theory, to assess potential policy outcomes, and to engage in the knowledge to policy translation process among other purposes[4750].

System dynamics (SD), a systems science method, uses informal maps and formal models with computer simulation to elucidate how counterintuitive behaviors of systems and unanticipated consequences of policy and program interventions may be explained by system structure, including resource accumulation and information delays, and formal and informal institutional rules [51]. System dynamics is formulated on the premise that the structure of the system - a network of social and biological causal mechanisms forming a directed cyclic graph or feedback mechanisms coupling a set of state variables and their rates of change (i.e., stocks and flows – governs system behavior [52]. A hallmark of the SD approach is the integration of multiple sources of structural data to represent complex systems through the use of causal loop diagrams (CLDs) and other qualitative mapping approaches as well as mathematical models. SD is increasingly being recognized in the health sector for its utility in mapping and understanding complex health problems, operationalizing research evidence and systematically analyzing ranges of intervention and policy solutions[47, 53, 54]. The application of SD to environmental health decision-making is limited, although there is growing interest [7, 55, 56].

An attractive aspect of system dynamics modeling is the opportunity to engage with community and policy stakeholders throughout the process from problem formulation to model analysis and scenario planning [57, 58]. This stakeholder engagement creates opportunities for the generation of a range of insights about the nature of dynamic problems, structural insights, and paradigmatic insights in addition to insights about the relationship between system structure and behavior. To illustrate this continuum of insights and relationships between them, system dynamicists K. Stave et al. offers the following typology (see Figure 1)[59]. For example, problem-related insights are gained by learning to see dynamic behavior (trends) rather than discrete events and seeing a graph of a system indicator (variable) fluctuating over time as the problem space [59]. Structural insights include seeing how a variable can be both cause and effect (feedback relationship), beginning to understand what system structure is, and recognizing the importance of accumulations and delays on system behavior. Dynamic insights represent the overlap of the previous two categories where the relationship between feedback loops and behavior is understood and dynamic hypotheses can be generated. Lastly, paradigmatic insights is a restructuring of how one sees the world as a system, with a causal feedback structure that endogenously generates dynamic behavior [59].

Figure 1.

Figure 1

Relationships between different categories of insights (with permission)[59]

In this paper we describe formative work that was undertaken to formally describe the multiple interacting streams of social stressors and environmental neurotoxicants through the use of qualitative system maps within the system dynamics framework. To accomplish our goals, we held a 2 ½ -day SD group model building workshop from December 2–4, 2019 hosted by the Social System Design Lab at Washington University in St. Louis, Missouri. The purpose of this paper is to present the output from the workshop and convey the problem-related, structural, dynamic and paradigmatic insights workshop participants gained through the collaborative model building process.

Methodology

Addressing environmental health disparities requires input from different scientific domains/disciplines. We used a group model building (GMB) approach[60, 61] that included a multidisciplinary panel of experts to elucidate the pathways through which environmental neurotoxicants and social stressors effect children’s neurodevelopmental disorder disparities. Specifically GMB “focuses on building SD models with teams/stakeholders in order to enhance team learning, to foster consensus and to create commitment with a resulting decision”[60]. Such a participatory approach to modeling building enables practitioners and researchers to learn together about problems of mutual interest in a way that provides reciprocal benefits such as building new concepts, insights, and/or practical innovations that they could not produce alone[57]. Our workshop participants did not have significant prior experience with SD and one of or our objectives was to build their SD capacity. We took community-based system dynamics (CBSD) approach to GMB[57], in order to build capabilities with systems thinking methods among the stakeholders while facilitating the substantive model-based discussion.

Formation of Core-Modeling Team and Identification of Workshop Participants

A four-member design team, or “core modeling team” (CMT) consisted of faculty from University of Maryland School of Public Health (first author DP-S), from Rochester Medical Center (co-author DC-S) and from the Social System Design Lab at Washington University (second author EB and senior author PH) were responsible for planning and facilitating the GMB workshop as well as identifying workshop participants. The CMT developed a set of workshop objectives to inform the design of the workshop:

  • Integrate diverse forms of knowledge about children’s exposures to environmental neurotoxicants, social stressors and cumulative risk into a preliminary system dynamics model on disparities in children’s neurodevelopment

  • Convene diverse researchers (neurotoxicologists, exposure scientists, public health researchers, air pollution epidemiologists, health disparities, environmental justice, psychologists, education researchers) into a collaborative research effort around these issues

  • Expose research team members to concepts of system dynamics modeling

  • Develop action steps to advance a research agenda around cumulative exposures

Potential participants were identified by first identifying needed disciplinary perspectives not represented by the CMT. Experts in the fields of developmental psychology, education policy, children’s environmental health advocacy, environmental law, local public health, state environmental regulations, environmental epidemiology and pediatrics were then identified through published literature and professional networks. DP-S contacted each potential participant by email and followed-up with in person or phone conversation providing more detail on the objectives and how their expertise would be informative for a systems modeling workshop. Thirteen stakeholders participated in our workshop representing these diverse disciplines in addition to the CMT. Members of the CMT served in facilitation roles, with DP-S functioning as a community facilitator, EB & PH serving as modeler facilitators and reflectors, and graduate students in system dynamics serving roles as runners, wall-builders, and note-takers.

Procedure

The 2 ½ -day workshop consisted of a set of framing presentations to orient participants to key concepts and objectives of the workshop, and a sequence of group model building “scripts”[62, 63], or structured modeling activities, designed to elicit participant’s contributions to the development and refinement of a qualitative system dynamics model. The workshops ran from 9:00am to 4:00pm on the first two days, and from 9:00am to 12:00pm on the third day. Modeling scripts were interspersed with breaks, presentations, and debrief discussions to balance the multiple objectives of substantive insight, learning, and community building. The workshop sequence started by establishing group expectations and norms through a “Hopes & Fears” activity, then rapidly elicited variables to inform the development of the first round of modeling through a variable elicitation activity called “Behavior Over Time Graphs” responding to the prompt: “What are factors related to differential exposure to neurodevelopmental stressors in children?” These factors, clustered into emergent themes would be the starting point from which a set of small group and large group activities for model formulation, review, and critique were employed to develop a synthesis stock and flow causal diagram. The third script was the Causal loop diagram or CLD script where small groups worked together to develop preliminary theories of feedback structure of factors hypothesized to be driving differential exposure to environmental toxicants in children. Each small group’s CLD were synthesized into one CLD in real time by a member of the CMT who was listening in to group work and small group presentations. This rough synthesis was presented for feedback at the close of the small group CLD session. The synthesized CLD was then translated into a “stock and Flow” diagram overnight by one modeler facilitator (EB) and was presented the next day as part of a model review session to be critiqued and revised live in session. This output was then used as a tool to generate potential strategies through an “Action Ideas” activity. Table 1 presents an overview of the group model building scripts employed in this workshop, their instrumental purpose in the workshop design, and their outcomes.

Table 1:

Group Model building “Script”, purpose, and outcomes

Script & Timing Purpose Outcomes
Hopes and Fears
(30 minutes)
To elicit expectations for the workshop sessions, and to air anxieties, uncertainties, or concerns to be addressed in the workshop design Collection of participant hopes & fears, clustered into emergent themes
Behavior Over Time Graphs (BOTGs)
(30 minutes)
To elicit a set of dynamic factors from workshop participants to be inputs into future modeling activities; to introduce the framing of problems as dynamic, or changing over time, rather than static. Collection of graphs of different variables related to cumulative environmental exposures, clustered into emergent themes.
Causal-loop diagram (CLD)
(50 minutes)
To develop preliminary theories of feedback structure of factors hypothesized to be driving differential exposure to environmental toxicants in children; to provide input to the development of a synthesis diagram Four causal loop diagrams reflecting the interconnections of factors identified in the Behavior Over Time Graphs activity, and the feedback relationships hypothesized to be driving differential exposure.
Model Synthesis
(Conducted in two stages: first during the CLD exercise & second between sessions)
To develop a synthesis diagram that integrates structures and factors presented in previous modeling activities as well as critiques from previous model reviews A synthesis diagram representing key variables, interconnections, feedback relationships, and areas of uncertainty based on prior modeling activities. The first draft of this synthesis was conducted during the session, and the more robust revision was conducted overnight.
Model Review
(Conducted in two stages: 30 minutes following the first CLD exercise; 1 hour on day two.)
To present a synthesis causal maps for review and critique by workshop participants; to integrate this feedback in real time A revised causal map that more accurately reflects the major feedback loops, narratives, themes, and group understanding of the correspondence between model structure and system behavior. This model review activity was conducted first using a draft synthesis directly following the CLD activity, and again repeated on the second day to review a synthesis developed over night.
Action ideas
(40 minutes)
To generate a list of intervention ideas based on the causal structure from the synthesis causal map A list of intervention ideas including a description of how those intervention ideas may act on the system, including barriers and facilitators to impact; a version of the causal map that maps on intervention points to describe proposed mechanisms of action.

All scripted activities (Table 1 and Figure 2) were designed to explore interaction and interdependencies between social and environmental factors affecting children neurodevelopment and disparities in those outcomes. The workshop closed with a review of the synthesis model and a discussion of a set of calls to action to make movement on substantive insights and research gaps identified over the course of the workshop. Model review involved discussing each proposed causal link to ensure it had face validity and was supported by the literature from the respective disciplines of workshop participants. The “calls to action” were consolidated into a set of commitments by workshop participants to follow up activities.

Figure 2.

Figure 2.

Group Model Building Activities

Evaluation Survey

Participants were surveyed six months following the workshop and one year later regarding the degree to which the workshop contributed to new insights and how these insights may have influenced their work. This evaluation step builds upon best practices for participatory model building[57]. Workshops participants were provided a copy of the final causal map (“stock and flow” diagram) we developed together and were asked “To what degree did the workshop contribute to any new insights about: 1) cumulative environmental exposures as a systems issue; 2) the multiple factors related to cumulative environmental exposures; 3) how those factors are related through feedback loops; 4) how different stakeholders and disciplines think about the cumulative environmental exposures; 5) how accumulations in the system affect neurological development (stocks and flows)?” Participants were given answer options on a 5-point Likert scale (0=not at all, 1=a little, 2=a moderate amount, 3=a lot, 4= a great deal). Finally, participants answered two open-ended questions to elicit specific examples of new insights about addressing cumulative environmental exposures and to describe major takeaways or insights and how these have influenced their work (e.g., new research questions, incorporation of new methods, engagement with new stakeholders or partners; translation of insights to policy action, etc.).

Data Analysis

Analysis of the model (CLD) was built into the sequence of structured activities described in Table 1. This enabled the CMT to evaluate the structure of each iteration of the model for face validity and to ensure participant ideas and constructs were accurately reflected. Throughout the workshop multiple iterations of the model were documented through photographs and notes. We used STELLA Architect[64] to adapt the synthesized CLD in real-time to incorporate emergent ideas or proposed refinements. Examples of these revisions included reflections of how climate change amplifies exposures through multiple pathways on the map, creation of explicit linkages between exposures to adverse childhood experiences and neurological functioning, and highlighting the ways that differential experiences of institutional discrimination may act on the map at multiple points. All revisions were grounded in the model itself and based on notes taken by CMT during the GMB sessions. Frequencies of responses to survey questions (6 months after workshop and 1 year follow-up) were summarized and described. Open-ended answers from workshop evaluation surveys were analyzed qualitatively using Stave’s insights typology (Figure 1). Authors DPS and EB coded the responses independently according to Stave’s typology and then conferred to reach consensus. Example quotes to illustrate Stave’s typology are reported here. Analysis of workshop outputs and evaluation responses were considered not to be subject to IRB review for human subjects research as they collected survey data on public behavior.

Results

Below we present the system dynamics causal map produced by workshop participants and describe systems insights workshop participants gained through the collaborative model building process.

Causal Map outputs

Outputs from the GMB sessions were synthesized into a large system dynamics stock and flow diagram (Figure 3) by the CMT representing a dynamic hypothesis of the system structure of cumulative environmental and social stressors creating disparate developmental outcomes for children. The stock and flow (S&F) diagram included 7 subsystems or sectors: environmental pollutant (neurotoxicants) exposures, social environment, housing environment, health status, education environment, employment and asset (wealth) development and advocacy. The subsystems are causally linked and include feedback loops, which provide explanations for trends in social disparities in exposures and outcomes among children. Importantly, this diagram is an artifact of the workshop, meaning it is a product of the specific configuration of prompts, participants, and moment in time of this workshop. However, the diagram is also a dense integration of the professional expertise and lived experience of a group of substantive experts and system science practitioners. It is meant to be an explicit visual representation of a negotiated collective mental model of the system that serves as a boundary object [65]. Boundary objects are, in their essence, intended to be challenged, modified or transformed.

Figure 3:

Figure 3:

Figure 3:

Synthesized Stock and Flow Diagram

The diagram uses a visual vocabulary to represent accumulations, causal links, and feedback loops. Stocks, depicted by boxes in the diagram, are representations of accumulations within a system. These accumulations can be tangible, such as student academic achievement or people living in houses with lead hazards. They can also be intangible, such as in the case of political and social capital for environmental justice, representing the multiple definitions of power accumulation and political pressure to advance environmental justice policy aims. The S&F can be read using the direction of arrows or links, which represent hypothesized causal relationships. The plus symbol indicates a relationship in the same direction, while minus signs indicate an inverse relationship. Typically, S&F diagrams contains two major types of feedback loops: reinforcing (R) and balancing (B) loops. Reinforcing feedback loops amplify or accelerate the rate of change, while balancing feedbacks counteract and oppose change.

In the S&F diagram created by the workshop participants, multiple neurotoxic chemical exposures are denoted by arrays of stocks with multiple inflows. Arrays are employed here to explicitly highlight how a child may accumulate multiple neurotoxic exposures that occur at different rates and at different times. This diagramming convention was used to account for discussions and debates within the workshop group about multiple sources and pathways of neurotoxic chemical exposures (e.g. lead in drinking water, particulate matter air pollution, phthalates in consumer products). An alternative might be to individually represent the multiple exposures, but that detail complexity was judged to obscure the overall feedback narrative we sought to illuminate. This accumulation of exposures during childhood increases exposure burden during preconception and prenatal exposures later in life which increases child exposures which is represented by feedback relationships between “being exposed” during early life and accumulation through adulthood with delay leading to prenatal exposures. Given the abundance of research on long-term impacts of adverse childhood experiences [66] and the synergistic effects of non-chemical and psychosocial stressors and environmental pollutant exposures [30, 32, 33, 6769], reinforcing relationships stemming from social environment are depicted in the S&F diagram. Adverse childhood experience was also identified as an accumulation in the CLD that related to a multigenerational exposures reinforcing loop. Additional causal loops were added for discrimination in health care, education and housing. To counter the reinforcing feedbacks, balancing feedback relationships are included related to education and environmental advocacy. The dynamic hypothesis is that quality learning environments increase academic achievement over time which increases individual and community wealth which increases political and social capital to advocate for environmental justice which reduces exposures to neurotoxicants Each of these elements represents stocks that can be tracked over time.

Participant System Insights

We noted insights throughout the workshop, as well as afterwards when we surveyed participants about system insights & new ideas generated from the workshop. A majority of respondents at 6 months and at 1 year reported favorably (“a great deal” or “a lot”) on the degree to which the workshop contributed new insights (see Table 2). The longitudinal follow up suggests a general persistence of insight and attribution of that insight to the workshop, which is elaborated below in qualitative responses.

Table 2.

Percent reporting the degree to which the workshop contributed to new systems insights

% Rating cumulative environmental exposures as a systems issue the multiple factors related to cumulative environment al exposures how those factors are related through feedback loops how different stakeholders and disciplines think about the cumulative environmental exposures how accumulations in the system affect neurological
6 mths 1 year 6 mths 1 year 6 mths 1 year 6 mths 1 year 6 mths 1 year
n = 7 n=12 n = 7 n=12 n = 7 n=12 n = 7 n=12 n = 7 n=12
A great Deal or A lot 6 (86%) 9 (75%) 6 (86%) 9 (75%) 6 (86%0 8 (67%) 6 (86%) 11 (92%) 5 (71%) 9 (75%)
A moderate amount 1 (14%) 3 (25%) 0 (0%) 2 (17%) 0 (0%) 3 (25%) 1 (14%) 1 (8%) 1 (14.5%) 2 (17%)
A little 0 (0%) 0 (0%) 1 (14%) 1 (8%) 1 (14%) 1 (8%) 0 (0%) 0 (0%) 1 (14.5%) 1 (8%)
None at all 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%)

Notes: Response rates were 53% at 6 months and 92% at 1 year.

Responses to open ended questions were analyzed according to Stave et al.’s (2016) typology. Recurring or key concepts relevant to our topic of children’s cumulative exposures and neurodevelopment expressed by workshop participants are summarized in Table 3 corresponding to Stave’s typology and further described below.

Table 3.

Overview of Insights from Workshop Participants According to Stave’s Typology

Insight Typology Key Concepts from Workshop Participants Example correspondence to the S&F diagram
Problem Related Insights Intergenerational aspects Shift in the motivating problem definition away from health consequences from individual exposures towards “cumulative environmental exposure stock” and away from only considering effects on current populations towards exposures generating future risk.
Structural Insights Need for effective communication between stakeholders who make up the system; seeing implications of accumulations of chemical exposures and adverse childhood experiences, the role inherent government agency silos; importance of working across environmental media for better environmental health protections Identifying hypothesized interacting and compounding links between exposures to health outcomes (e.g. health status, neurological functioning, and aberrant behavior combining to influence academic achievement); conceptualizing environmental exposures as arrayed accumulations following similar mechanisms rather than unique processes; how adverse health effects caused by pollutant exposure can spill over into other domains such as education and employment
Dynamic Insights Potential points of intervention and the downstream consequences of actions Mentally simulating consequences of housing interventions on health outcomes, ACES, and future education & employment outcomes.
Paradigmatic Insights Fight reductionism; engage with the public; force experts to get out of silos Seeing where existing actions are focused on the diagram, and where there are significant gaps; identifying the potential for systems maps and visualization for cross-silo dialogue.

Problem related insights means seeing the problem space as trends over time, rather than discrete events. So instead of focusing on cross-sectional relationships, one shifts to framing neurodevelopmental disparities in terms of the observed time trends in exposure to multiple environmental neurotoxicants, social stressors and neurodevelopmental outcomes and how they might be related to one another. One quote from a participant exemplified consensus among all workshop participants that defining the problem in terms of time trends must include intergenerational aspects of neurotoxicant exposures and how the social and environmental experience work together to create an intergenerational set of risks.

I think the multigenerational nature of the model was the most compelling part of the “model construction” discussion. We tend to view adverse childhood events as specific to a child. What really happens is that when such an event occurs, it is like when a ball is kicked and starts to roll down a slope. It has the potential to “snowball”, increase in velocity or stop, depending upon which “track” it is placed. Over time, this ball rolls to the next generation, who must contend with the “accumulation” that may predate his or her existence.

[Doctoral student]

Structural insight is beginning to understand what system structure is, seeing causal connections, specific points in the system (self, others) and where things accumulate, and understanding how links and polarity work (CLD drawing conventions). For our specific topic, structural insights would recognize a set of processes and feedback mechanisms that may generate racial and class disparities in children’s neurodevelopmental outcomes without the need to include individual level characteristics like genetics. Deeper structural insights would lead to developing a rigorous scientific understanding as to why a given system model structure can/cannot generate the dynamic patterns of interest. Workshop participants shared a number of insights that focused mostly on seeing specific points in the system and the need for better communication across these points.

System Dynamic Models visualization reflects the importance of working across environmental media. It is much easier to point out where there are breaks in the regulatory system and why that is important. Regulatory authority to reduce toxic exposures generally does not address cross media contamination. This leads to a lack in agency structure to address these issues.

[State environmental regulatory agency scientist]

It allows for child development researchers who typically work on the individual-child level, to think about how broader aspects of the environment affect that child’s development. Those who work on larger environmental issues relevant to child development, can see how that may affect different individual children (not just averages).

[Academic researcher]

The importance of communication and collaboration among the different stakeholders to identify-short and long-term solutions.

[Pediatrician]

System dynamic models visualization reflects the importance of working across environmental media. It is much easier to point out where there are breaks in the regulatory system and why that is important. Regulatory authority to reduce toxic exposures generally does not address cross media contamination. This leads to a lack in agency structure to address these issues.

[State environmental regulatory agency scientist]

Dynamic insights advance the understanding of relationships between structure and behavior, principles of accumulations, feedback loops and behavior, effects of delays, and policy insights. These set of insights facilitate generation of hypotheses about what is causing the dynamic behavior of the system. For example, while many of the policy actions identified during the workshop focused on specific sectors (e.g., tighten fuel efficiency standards; creation of a Superfund to clean up lead in housing; regulate toxic chemicals by class; trauma and pollutant informed health care; increase investment in early education), participants noted the how SD modeling offered the opportunity to evaluate the impacts of “bundling” the actions together in a coordinated manner. We highlight a few quotes from participants which indicated that understanding the potential points of intervention and the downstream consequences of actions were the most salient dynamic insights for the group.

Factors that we looked at have a far greater and extended impact than people usually think of. That provides many points of potential intervention, but the consequences of a potential intervention may also require understanding to a greater extent what the consequences of an intervention would be.

[Academic researcher in neurotoxicology]

Systems mapping helps reveal that the most effective sites for intervention may not be where the interventions are actually occurring.

[State health department epidemiologist]

It is also essential to consider how time delays affect accumulating exposures, and system models can help address these questions of timing.

[State health department epidemiologist]

Further, learning the conventions of causal loop diagraming aided in these dynamic insights as illustrated with these two example quotes. However, our quantitative survey results seem to signal that understanding feedback relationships may remain a challenge if this skill is not continuously practiced.

As the synthesis of all the teams’ causal loop diagrams took place, it became evident how system dynamics modelling can be a very effective tool to discover and explore the effect of intervention ideas.

[Pediatrician]

I have continued to reflect on the multidisciplinary approach required to effectively break cycles or get the feedback loops going in the right direction.

[Advocacy non-profit organization director]

Seeing the world in SD terms, restructuring one’s abstract images and ideas into SD terms, and large shifts in one’s conceptualizations signal paradigmatic insights. For our participants this shift in thinking revolved around using SD to counter reductionism inherent in the bio-medical paradigm, to be more effective in engaging with the public and impacted communities, and help force experts to get out of their silos and advance multidisciplinary scientific research to solve problems.

I am excited about this because it may give us researchers an approach to think about child development as a complex system across multiple levels of analysis.

[Academic research in education]

The complexity issue is one that many scientists simply don’t want to deal with. On the basic side of science, the reductionism is rampant; on the epi side, interactions are always sample size-dependent. It raises the question of how we begin to make the community understand the critical nature of dealing with this complexity.

[Academic researcher in neurotoxicology]

What I realized through the systems science workshop is that the many facets of this childhood exposure to neurotoxicants problem can be understood as by-products of greater societal phenomena. As such, we can better understand how seemingly unrelated domains of this problem can interact synergistically.

[PhD student]

Doing this work well requires multidisciplinary teams whose members are willing and able to do the work of not only examining other points of view and methodological approaches as well as actively questioning their own.

[Academic researcher in psychology and child development]

I have used the model image (the synthesized S&F as shown in Figure 3) in multiple presentations to our commissioner’s office as we drafted potential cumulative impacts/environmental justice legislative language.

[State environmental regulatory agency scientist]

These insights informed my recognition that we needed to start engaging with and listening to environmental justice BIPOC leaders if we hope to help reduce neurotoxic exposures to children (and adults) disproportionately impacted.

[Advocacy non-profit organization director]

Discussion

This study presents insights various stakeholders gained through a novel application of GMB using SD to understand the social mechanisms that create disparities in cumulative exposures to neurodevelopmental toxicants. The workshop described above was a formative investment in both conceptualizing the social structure of cumulative exposures and risk beyond individual level biological mechanisms, and in developing a transdisciplinary cohort of researchers, policy stakeholders, and other experts working to bridge disciplinary and linguistic divides. Group model building using SD can lead to important insights to into the sociological, policy, and institutional mechanisms through which disparities in cumulative impacts are transmitted, resisted, and understood. Thus the insights described here suggest a number of opportunities, new ideas, and challenges for integrating participatory systems science methods to support environmental health research and policy.

Challenges & opportunities to operationalizing cumulative exposures

Motivations for this work stem from the lack of progress in addressing cumulative environmental health risk in decision-making by policy makers at state and federal levels. The current approach in environmental policy making is to regulate risk from single contaminant /pollutant exposure determined in the absence of any social/physical context, likely underestimating true health risks. The limitations of the current approach are particularly glaring in the case of children’s exposures to neurodevelopmental toxicants. Children are often exposed prenatally and in early childhood to multiple chemicals and stressors that can adversely affect their cognitive abilities, academic performance and their consequent educational trajectories, adult health, wealth and social status [13, 16, 20, 27, 7076]. The extent and breadth of such exposures is influenced by race/ethnicity and socioeconomic status (SES), that may contribute to pronounced health disparities for multiple outcomes including school readiness and cognitive delay [7783]. The systems model developed through GMB identified seven subsystems showing interconnectedness among variables and feedbacks in children’s social and physical environments impacting neurodevelopment with intergenerational spillover effects and broader societal implications. We can think about this qualitative map (the S&F Diagram in Figure 3) as a road map to reflect an argument of the potential causal pathways that connect subsystems in order to generate hypotheses as well as to inform data collection priorities for further quantitative and mixed methods research. These hypotheses may invite researchers from multiple domains to drill into the sociological, policy, and institutional mechanisms through which risk is transmitted, resisted, and understood.

Conceptualizing the role of structural racism in cumulative exposures

Research and discourse on structural racism has highlighted the importance of environmental exposures as a key site of differential harm based on race. It is now well established that racism is more than interpersonal animus [43]. Scholarship has demonstrated the structural nature of racism and discrimination and its effects on health [43, 44, 84]. From the beginning of the workshop, participants expressed hopes and fears around “focusing on social structures that underlie racial health disparities” and “seeing how far back in time the challenges of cumulative environmental stressors have been happening.” All workshop participants agreed with statement by several participants who noted that racism is the “canvas upon which the causal diagram was drawn”. However, during initial review of synthesized S&F we began to specify how racism is impacting feedback loops and segments of the S&F. Additional causal loops were added for discrimination in health care, education and housing. In follow-up surveys, participants stated that the workshop provided a language they could use to better communicate with colleagues and the public about the consequences of children’s health disparities, especially multigenerational effects. For example, one workshop participant from a state environmental regulatory agency noted that regulatory governmental agencies rarely analyze how their decisions can have multigenerational effects or unintended consequences. As result, ostensibly neutral environmental regulations may create or exacerbate existing disparities. An additional well known benefit of using SD is the identification of potential points of intervention and the downstream consequences of policy actions [85]. These insights are consistent with emerging applications of system dynamics modeling as a method to explicitly explore and represent social forces, institutions, ideologies, and processes interact with one another to generate and reinforce inequities among racial/ethnic groups [8688].

The promise of participatory systems science methods for trans-disciplinary research & policy translation

This modeling approach illustrates opportunities for environmental and public health policies to consider cumulative exposures impacting children, but also spotlights the necessity trans-disciplinary and cross-institutional collaboration. In this paper we present one approach to employing participatory systems science methods to support collective mapping of the multiple pathways and interconnections that comprise the complex system of environmental exposure risks and harm. The visualization of the system developed through these methods (via the CLD, the S&F), as noted by workshop participants, may allow state agencies and organizations to identify other agencies and organizations with which they should be collaborating to bring about improvements in child neurodevelopment; no one agency or academic department contains the expertise or community and industry connections to address these systems alone. With Stave’s typology, we illustrate how even qualitative, non-mathematical applications of systems methods can illuminate a diverse array of relevant policy insights by shifting the predominant “mental models” of environmental regulators and other stakeholders towards “seeing” the system structures that create and maintain disparities in cumulative chemical and non-chemical exposures and poor health outcomes. Beyond the system insights categorized according to Stave’s model, participants also noted how SD and GMB can aid in the integration of multidisciplinary evidence, identify knowledge gaps and allow for the discussion of racism and its explicit mechanisms to make the invisible more visible. Even though workshop participants came from diverse expertise backgrounds, they all could effectively communicate ideas and incorporate them into the first causal loop map. Participants increased their appreciation for the need to work across disciplines. These outcomes are consistent with prior work examining the potential of participatory systems mapping and modeling tools for integrating diverse sources of expertise mobilizing policy action to address complex issues [8991].

Potential Limitations

As discussed previously, the workshop and system dynamics causal map presented here are artifacts of a specific set of participant perspectives at a single point in time. In this spirit, the output of the causal map is not intended to be a rigorous, complete example of qualitative system dynamics modeling. The model as presented includes significant reinforcing behavior, but the sessions were not built to meaningfully engage in the dynamics of resisting or balancing this intergenerational transmission of exposure and harm. To this end, the product and insights are not intended to be generalizable theory, but rather a provocation for future engagement, research, and collaboration that cuts across disciplinary siloes. Participatory approaches to system dynamics modeling are gaining traction in a variety of fields, from natural resources management to public health. While they provide a means to develop shared language and capacity for thinking about complex topics, participation alone does not equate to empirical validity or analytic rigor. Further efforts would be required to build confidence in this map.

Building confidence in this map can take multiple forms. One approach is to expand engagement to include further audiences, including policymakers, regulators, industry representatives, and community residents to refine and challenge causal links. Efforts to meaningfully advance this line of inquiry may require development of developing formal computer simulation models to understand and analyze systems for identifying the origins of system behavior and identify potential high leverage policy interventions. The resources, time and skills to do this may be limited, especially when considering the adequacy and availability of empirical estimates for model parameters and initial conditions. Moreover, failing to move toward these deeper insights in a timely way may limit the perceived value of a new approach and thereby increase the effort to maintain the connections over time, and/or limit the growth of a network or participants to include affected communities, policy makers, and other stakeholders critical to developing and implementing relevant policies in EH decision making. As a next step, we intend to reengage with workshop participants to refine a proof-of-concept quantitative simulation model [7] based on the synthesized S&F model.

Conclusion

It is well documented that fetuses, infants and children are more susceptible and vulnerable to environmental risks than adults due to their physiological, developmental and psychosocial characteristics [92]. Environmental exposures to chemicals early in life are important preventable and significant causes of disease in children including neurodevelopmental disorders [93]. Environmental risks associated with include toxic chemicals and pollutant exposures are compounded by exposure to poverty, social stressors, natural disasters, and adverse consequences related to climate change. Yet existing approaches to conceptualizing these systems of effects, remain constrained by siloed approaches to research and policy making. Therefore, it is of extreme importance to focus efforts in understanding how these complex environmental and social stressor exposures are taking place and how they are affecting communities in different ways to be able to better protect our children’s health preventing life-long consequences. As a foundational step to exploring alternative ways of conceptualizing this complex system, we engaged a diverse group child health and environmental health experts and advocates in applying SD group model building to uncover insights about how disparities in children’s neurodevelopment may arise as a result of the combined exposure to environmental chemical contaminants and social stressors. A central tenet of SD is that the complex behaviors of organizational and social systems are the result of ongoing accumulations over time embedded within a set of feedback mechanisms or loops [94]. The combined effects of multiple environmental toxicants and social stressor exposures are widely recognized as important public health problems, likely contributing to health inequities. However cumulative environmental health risk and impacts have received little attention by U.S. policy makers at state and federal levels to develop comprehensive strategies to reduce these exposures, mitigate cumulative risks/impacts and prevent harm. This topic necessitates an interdisciplinary approach. Group model building using SD can lead to important insights to into the sociological, policy, and institutional mechanisms through which disparities in cumulative risks and impacts are transmitted, resisted, and understood.

Highlights.

Children are exposed to multiple chemical and other stressors that are neurotoxic Chemical by chemical approach to environmental policy leaves children underprotected Multiple chemical and nonchemical exposures represent a complex system System dynamics (SD) examines complex behaviors of organizational and social systems Group model building using SD can better inform policies to protect children’s health

Acknowledgements

We thank all workshop participants for their invaluable insights.

Funding

This work was supported by the National Institute of Environmental Health Sciences (Award Number K01ES028266). All authors read and approved the final manuscript. The content of this article is solely the responsibility of the authors and does not necessarily represent the official views of the funding agencies.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Competing interests

The authors declare they have no actual or potential competing financial interests.

Declaration of interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

  • 1.National Research Council, Science and Decisions: Advancing Risk Assessment. 2009, Washington, DC: The National Academies Press. 424. [PubMed] [Google Scholar]
  • 2.National Environmental Justice Advisory Council, Ensuring Risk Reduction in Communities with Multiple Stressors: Environmental Justice and Cumulative Risks/Impacts. 2004: Washington, D.C. [Google Scholar]
  • 3.Payne-Sturges DC, et al. , Methods for Evaluating the Combined Effects of Chemical and Nonchemical Exposures for Cumulative Environmental Health Risk Assessment. Int J Environ Res Public Health, 2018. 15(12). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Huang H, et al. , Cumulative Risk and Impact Modeling on Environmental Chemical and Social Stressors. Curr Environ Health Rep, 2018. 5(1): p. 88–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Appleton AA, Holdsworth EA, and Kubzansky LD, A Systematic Review of the Interplay Between Social Determinants and Environmental Exposures for Early-Life Outcomes. Curr Environ Health Rep, 2016. 3(3): p. 287–301. [DOI] [PubMed] [Google Scholar]
  • 6.National Research Council, Science and Decisions. 2009, National Academies of Sciences: Washington, D.C. [Google Scholar]
  • 7.Payne-Sturges DC, et al. , Defining and Intervening on Cumulative Environmental Neurodevelopmental Risks: Introducing a Complex Systems Approach. Environ Health Perspect, 2021. 129(3): p. 35001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Zablotsky B, et al. , Prevalence and Trends of Developmental Disabilities among Children in the United States: 2009–2017. Pediatrics, 2019. 144(4). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Bennett D, et al. , Project TENDR: Targeting Environmental Neuro-Developmental Risks The TENDR Consensus Statement. Environ Health Perspect, 2016. 124(7): p. A118–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Adamkiewicz G, et al. , Moving environmental justice indoors: understanding structural influences on residential exposure patterns in low-income communities. Am J Public Health, 2011. 101 Suppl 1: p. S238–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Brochu PJ, et al. , Particulate air pollution and socioeconomic position in rural and urban areas of the Northeastern United States. Am J Public Health, 2011. 101 Suppl 1: p. S224–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Fann N, et al. , Maximizing health benefits and minimizing inequality: incorporating local-scale data in the design and evaluation of air quality policies. Risk Anal, 2011. 31(6): p. 908–22. [DOI] [PubMed] [Google Scholar]
  • 13.Gray SC, Edwards SE, and Miranda ML, Race, socioeconomic status, and air pollution exposure in North Carolina. Environ Res, 2013. 126: p. 152–8. [DOI] [PubMed] [Google Scholar]
  • 14.Gray SC, et al. , Assessing the impact of race, social factors and air pollution on birth outcomes: a population-based study. Environ Health, 2014. 13(1): p. 4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Miranda ML, et al. , Making the environmental justice grade: the relative burden of air pollution exposure in the United States. Int J Environ Res Public Health, 2011. 8(6): p. 1755–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Morello-Frosch R and Shenassa ED, The environmental “riskscape” and social inequality: implications for explaining maternal and child health disparities. Environ Health Perspect, 2006. 114(8): p. 1150–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Morello-Frosch R, et al. , Understanding the cumulative impacts of inequalities in environmental health: implications for policy. Health Aff (Millwood), 2011. 30(5): p. 879–87. [DOI] [PubMed] [Google Scholar]
  • 18.Morello-Frosch R and Jesdale BM, Separate and unequal: residential segregation and estimated cancer risks associated with ambient air toxics in U.S. metropolitan areas. Environ Health Perspect, 2006. 114(3): p. 386–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Bell ML and Ebisu K, Environmental inequality in exposures to airborne particulate matter components in the United States. Environ Health Perspect, 2012. 120(12): p. 1699–704. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Landrigan PJ, Rauh VA, and Galvez MP, Environmental justice and the health of children. Mt Sinai J Med, 2010. 77(2): p. 178–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Hicken MT, et al. , A novel look at racial health disparities: the interaction between social disadvantage and environmental health. Am J Public Health, 2012. 102(12): p. 2344–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Hajat A, et al. , Air pollution and individual and neighborhood socioeconomic status: evidence from the Multi-Ethnic Study of Atherosclerosis (MESA). Environ Health Perspect, 2013. 121(11–12): p. 1325–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Hajat A, Hsia C, and O’Neill MS, Socioeconomic Disparities and Air Pollution Exposure: a Global Review. Curr Environ Health Rep, 2015. 2(4): p. 440–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Salihu HM, et al. , Particulate pollutants and racial/ethnic disparity in feto-infant morbidity outcomes. Matern Child Health J, 2012. 16(8): p. 1679–87. [DOI] [PubMed] [Google Scholar]
  • 25.O’Neill MS, et al. , Poverty, environment, and health: the role of environmental epidemiology and environmental epidemiologists. Epidemiology, 2007. 18(6): p. 664–8. [DOI] [PubMed] [Google Scholar]
  • 26.Mohai P, et al. , Racial and socioeconomic disparities in residential proximity to polluting industrial facilities: evidence from the Americans’ Changing Lives Study. Am J Public Health, 2009. 99 Suppl 3: p. S649–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Evans GW and English K, The environment of poverty: multiple stressor exposure, psychophysiological stress, and socioemotional adjustment. Child Dev, 2002. 73(4): p. 1238–48. [DOI] [PubMed] [Google Scholar]
  • 28.Bullard RD, et al. , Toxic Wastes and Race and Toxic Wastes and Race at Twenty. 2007, United Church of Christ: Cleveland, OH. [Google Scholar]
  • 29.Wright RJ, Moving towards making social toxins mainstream in children’s environmental health. Curr Opin Pediatr, 2009. 21(2): p. 222–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Rauh VA, et al. , Developmental effects of exposure to environmental tobacco smoke and material hardship among inner-city children. Neurotoxicol Teratol, 2004. 26(3): p. 373–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Guxens M, et al. , Prenatal exposure to residential air pollution and infant mental development: modulation by antioxidants and detoxification factors. Environ Health Perspect, 2012. 120(1): p. 144–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Vishnevetsky J, et al. , Combined effects of prenatal polycyclic aromatic hydrocarbons and material hardship on child IQ. Neurotoxicol Teratol, 2015. 49: p. 74–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Chari R, et al. , Integrating susceptibility into environmental policy: an analysis of the national ambient air quality standard for lead. Int J Environ Res Public Health, 2012. 9(4): p. 1077–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Hubbs-Tait L, et al. , Neurotoxicants, Micronutrients, and Social Environments: Individual and Combined Effects on Children’s Development. Psychol Sci Public Interest, 2005. 6(3): p. 57–121. [DOI] [PubMed] [Google Scholar]
  • 35.Lovasi GS, et al. , Neighborhood Social Context and Individual Polycyclic Aromatic Hydrocarbon Exposures Associated with Child Cognitive Test Scores. J Child Fam Stud, 2014. 23(5): p. 785–799. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Bellinger D, et al. , Low-level lead exposure, social class, and infant development. Neurotoxicol Teratol, 1988. 10(6): p. 497–503. [DOI] [PubMed] [Google Scholar]
  • 37.Bellinger D, Leviton A, and Sloman J, Antecedents and correlates of improved cognitive performance in children exposed in utero to low levels of lead. Environ Health Perspect, 1990. 89: p. 5–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Cory-Slechta DA, Studying toxicants as single chemicals: does this strategy adequately identify neurotoxic risk? Neurotoxicology, 2005. 26(4): p. 491–510. [DOI] [PubMed] [Google Scholar]
  • 39.Israel BD, Student risk assessment articles: An environmental justice crtique of risk assessment. N.Y.U Environmental Law Journal, 1994. 3: p. 470–522. [Google Scholar]
  • 40.Payne-Sturges DC, Sangaramoorthy T, and Mittmann H, Framing Environmental Health Decision-Making: The Struggle over Cumulative Impacts Policy. Int J Environ Res Public Health, 2021. 18(8). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Harrison JL, ‘We do ecology, not sociology’: interactions among bureaucrats and the undermining of regulatory agencies’ environmental justice efforts. Environmental Sociology, 2017. 3(3): p. 197–212. [Google Scholar]
  • 42.Pellow DN, Environmental Inequality Formation: Towards a Theory of Environmental Injustice. American Behavioral Scientist, 2000. 43(4): p. 581–601. [Google Scholar]
  • 43.Powell JA, Structural Racism: Building Upon the Insights of John Calmore. North Carolina Law Review, 2008. 86: p. 791–816. [Google Scholar]
  • 44.Gee GC and Ford CL, STRUCTURAL RACISM AND HEALTH INEQUITIES: Old Issues, New Directions. Du Bois Rev, 2011. 8(1): p. 115–132. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Gee GC and Hicken MT, Structural Racism: The Rules and Relations of Inequity. Ethnicity & Disease, 2021. 31(Suppl): p. 293–300. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Sterman JD, Business Dynamics: Systems Thinking and Modeling for a Complex World. 2000, Boston, MA: McGraw-Hill. [Google Scholar]
  • 47.Luke DA and Stamatakis KA, Systems Science Methods in Public Health: Dynamics, Networks, and Agents. Annual Review of Public Health, 2012. 33(1): p. 357–376. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Rutter H, et al. , The need for a complex systems model of evidence for public health. Lancet, 2017. 390(10112): p. 2602–2604. [DOI] [PubMed] [Google Scholar]
  • 49.Morais LMO, et al. , Promoting knowledge to policy translation for urban health using community-based system dynamics in Brazil. Health Res Policy Syst, 2021. 19(1): p. 53. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Minyard K, et al. , Community and programmatic factors influencing effective use of system dynamic models. System Dynamics Review, 2018. 34(1–2): p. 154–171. [Google Scholar]
  • 51.Richardson GP, Reflections on the foundations of system dynamics. System Dynamics Review, 2011. 27(3): p. 219–243. [Google Scholar]
  • 52.Sterman J, System dynamics at sixty: the path forward. System Dynamics Review, 2018. 34(1–2): p. 5–47. [Google Scholar]
  • 53.Atkinson JA, et al. , A modelling tool for policy analysis to support the design of efficient and effective policy responses for complex public health problems. Implement Sci, 2015. 10: p. 26. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Calancie L, et al. , Using Behavior Over Time Graphs to Spur Systems Thinking Among Public Health Practitioners. Prev Chronic Dis, 2018. 15: p. E16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Currie DJ, Smith C, and Jagals P, The application of system dynamics modelling to environmental health decision-making and policy - a scoping review. BMC Public Health, 2018. 18(1): p. 402. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Prochaska JD, et al. , The utility of a system dynamics approach for understanding cumulative health risk from exposure to environmental hazards. Environ Res, 2019. 172: p. 462–469. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Hovmand PS, Community Based System Dynamics. 2014, New York: Springer. [Google Scholar]
  • 58.Stave KA, Using system dynamics to improve public participation in environmental decisions. System Dynamics Review, 2002. 18(2): p. 139–167. [Google Scholar]
  • 59.Stave K, Zimmermann N, and Kim H, Exploring the nature of insight in System Dynamics, in 34th Conference of the International System Dynamics Society 2016: Delft, Netherlands. [Google Scholar]
  • 60.Vennix J, Group Model Building. 1996, West Sussex, England: John Wiley & Sons Ltd. 297. [Google Scholar]
  • 61.Andersen D and Richardson GP, Scripts for group model building. System Dynamics Review, 1997. 13(2): p. 107–129. [Google Scholar]
  • 62.Hovmand PS, et al. , Group Model-Building ‘Scripts’ as a Collaborative Planning Tool. Systems Research and Behavioral Science, 2012. 29(2): p. 179–193. [Google Scholar]
  • 63.Hovmand PS, et al. Scriptapedia. 2015. [cited 2022 February 9]; Available from: https://en.wikibooks.org/wiki/Scriptapedia.
  • 64.ISEE Systems. STELLA Architect [computer software]. 2015; Available from: http://www.iseesystems.com/store/products/stella-architect.aspx.
  • 65.Black LJ and Andersen DF, Using Visual Representations as Boundary Objects to Resolve Conflict in Collaborative Model-Building Approaches. Systems Research and Behavioral Science, 2012. 29(2): p. 194–208. [Google Scholar]
  • 66.Hughes K, et al. , The effect of multiple adverse childhood experiences on health: a systematic review and meta-analysis. The Lancet Public Health, 2017. 2(8): p. e356–e366. [DOI] [PubMed] [Google Scholar]
  • 67.Padula AM, et al. , A review of maternal prenatal exposures to environmental chemicals and psychosocial stressors-implications for research on perinatal outcomes in the ECHO program. J Perinatol, 2020. 40(1): p. 10–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68.Clougherty JE, Shmool JLC, and Kubzansky LD, The Role of Non-Chemical Stressors in Mediating Socioeconomic Susceptibility to Environmental Chemicals. Current Environmental Health Reports, 2014. 1(4): p. 302–313. [Google Scholar]
  • 69.Stein LJ, et al. , Early childhood adversity potentiates the adverse association between prenatal organophosphate pesticide exposure and child IQ: The CHAMACOS cohort. Neurotoxicology, 2016. 56: p. 180–187. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 70.Halle T, et al. , Disparities in Early Learning and Development: Lessons from the Early Childhood Longitudinal Study – Birth Cohort (ECLS-B). 2009, Child Trends: Washington, D.C. [Google Scholar]
  • 71.Suor JH, et al. , Tracing Differential Pathways of Risk: Associations Among Family Adversity, Cortisol, and Cognitive Functioning in Childhood. Child Dev, 2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 72.Gee GC, Walsemann KM, and Brondolo E, A life course perspective on how racism may be related to health inequities. Am J Public Health, 2012. 102(5): p. 967–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 73.Halfon NL, Larson K, and Russ S, Why Social Determinants? Healthcare Quarterly, 2010. 14. [DOI] [PubMed] [Google Scholar]
  • 74.Evans GW and Kim P, Multiple risk exposure as a potential explanatory mechanism for the socioeconomic status-health gradient. Ann N Y Acad Sci, 2010. 1186: p. 174–89. [DOI] [PubMed] [Google Scholar]
  • 75.Tulve NS, et al. , Development of a Conceptual Framework Depicting a Child’s Total (Built, Natural, Social) Environment in Order to Optimize Health and Well-Being. J Environ Health Sci, 2016. 2(2): p. 1–8. [Google Scholar]
  • 76.Trentacosta CJ, et al. , Environmental Contaminants and Child Development. Child Development Perspectives, 2016. 10(4): p. 228–233. [Google Scholar]
  • 77.Hackman DA and Farah MJ, Socioeconomic status and the developing brain. Trends Cogn Sci, 2009. 13(2): p. 65–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 78.Hillemeier MM, et al. , Disparities in the prevalence of cognitive delay: how early do they appear? Paediatr Perinat Epidemiol, 2009. 23(3): p. 186–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 79.Brooks-Gunn J and Duncan GJ, The effects of poverty on children. Future Child, 1997. 7(2): p. 55–71. [PubMed] [Google Scholar]
  • 80.Entwisle DRA, Early schooling and social stratification, in The transition to kindergarten, Pianta MJ, Editor. 1999, Paul H. Brooks Publishing Co: Baltimore, MD. p. 13–38. [Google Scholar]
  • 81.Hair E, et al. , Children’s school readiness in the ECLS-K: Predictions to academic, health, and social outcomes in first grade. Early Childhood Research Quarterly, 2006. 21(4): p. 431–454. [Google Scholar]
  • 82.Noble KG, Norman MF, and Farah MJ, Neurocognitive correlates of socioeconomic status in kindergarten children. Dev Sci, 2005. 8(1): p. 74–87. [DOI] [PubMed] [Google Scholar]
  • 83.Burkam DT and Lee VE, Social and academic disadvantage as children enter kindergarten, in Inequality at the Starting Gate:. 2002, Economic Policy Institute: Washington, D.C. [Google Scholar]
  • 84.Bailey ZD, et al. , Structural racism and health inequities in the USA: evidence and interventions. The Lancet, 2017. 389(10077): p. 1453–1463. [DOI] [PubMed] [Google Scholar]
  • 85.Meadows DH, Leverage Points: Places to Intervene in a System. 1999, The Sustainability Institute: Hartland, VT. p. 21. [Google Scholar]
  • 86.Headen IB, M.; & Ballard E, System Dynamics and Structural Racism: A Scoping Review., in International System Dynamics Conference. 2021. [Google Scholar]
  • 87.Burrell M, et al. , Depicting “the system”: How structural racism and disenfranchisement in the United States can cause dynamics in community violence among males in urban black communities. Soc Sci Med, 2021. 272: p. 113469. [DOI] [PubMed] [Google Scholar]
  • 88.Johnson Iv JJ, Padilla JJ, and Diallo SY, Closing the academic achievement gap: A system dynamics study. Journal of Simulation, 2020: p. 1–25. [Google Scholar]
  • 89.Hogan M, et al. , Integrating Content Expertise and Methodological Expertise in Team-Based Settings to Address Complex Societal Issues-A Systems Perspective on Challenges. Systems Research and Behavioral Science, 2018. 35(6): p. 908–915. [Google Scholar]
  • 90.Bammer G, et al. , Expertise in research integration and implementation for tackling complex problems: when is it needed, where can it be found and how can it be strengthened? Palgrave Communications, 2020. 6(1). [Google Scholar]
  • 91.Luna‐Reyes LF, et al. , Modeling and Simulation as Boundary Objects to Facilitate Interdisciplinary Research. Systems Research and Behavioral Science, 2018. 36(4): p. 494–513. [Google Scholar]
  • 92.Etzel R and Balk SJ, eds. Pediatric Environmental Health. 4th ed., ed. American Academy of Pediatrics Council on Environmental Health. 2019, American Academy of Pediatrics: Itasca, IL. [Google Scholar]
  • 93.Landrigan PJ and Goldman LR, Children’s vulnerability to toxic chemicals: a challenge and opportunity to strengthen health and environmental policy. Health Aff (Millwood), 2011. 30(5): p. 842–50. [DOI] [PubMed] [Google Scholar]
  • 94.Homer JB and Hirsch GB, System dynamics modeling for public health: background and opportunities. Am J Public Health, 2006. 96(3): p. 452–8. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES