National Academies Press: OpenBook

Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop (2021)

Chapter: 2 Overview of Systems Science Theories, Approaches, and Applications

« Previous: 1 Introduction
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×

2

Overview of Systems Science Theories, Approaches, and Applications

Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×

Part I of the workshop was an introductory session, moderated by Daniel E. Rivera, professor of chemical engineering and program director of the Control Systems Engineering Laboratory at Arizona State University. Speakers in this session introduced the concepts of complex systems and systems thinking as they apply to population health issues and broadly described systems science approaches (i.e., methodologies or tools) that could be applied to obesity solutions. (These key terms, defined in Box 2-1, were used throughout the workshop. See Appendix D for a full glossary of terms used by workshop speakers.) Three speakers delivered presentations, engaged in a panel discussion, and answered questions from workshop participants.

OVERVIEW AND HISTORY OF SYSTEMS SCIENCE

Ross Hammond, the Betty Bofinger Brown Associate Professor at Washington University in St. Louis and director of the Center on Social Dynamics and Policy and senior fellow in economics at the Brookings

Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×

Institution, presented a historical overview of systems science, including its advantages and potential applications to public health and obesity research and intervention. The multifaceted nature of many public health problems—including the obesity epidemic—makes them challenging for scientists and policy makers to address, he stated, adding that such intricate complexity makes these problems well suited to examination with systems science approaches.

Hammond described four characteristics of complex systems such as those that drive obesity. First, he began, many different factors interact across multiple sectors and scales to affect relevant behaviors and outcomes, the result of which is a deeply interconnected system. As an example, he showed a diagram constructed by the United Kingdom. Foresight Group to map measurable factors that drive obesity (see Figure 2-1). He pointed out the interconnected and interdependent system illustrated by this map, noting that an isolated focus on a single part of such a system risks missing many other important factors and linkages.

The second characteristic, Hammond continued, is that complex systems include multiple heterogeneous actors who may affect obesity outcomes, such as individual consumers, retailers, schools, health care providers, and community coalition leaders. These actors have different incentives, information, and network connections, he observed, explaining that their interconnections are important because an intervention may unintentionally affect certain actors, with implications for their connections with other actors. Actors are also adaptive, Hammond added, elaborating that differences may exist between their short- and long-term behavioral responses to environmental and policy changes.

Hammond described a third characteristic of complex systems by stressing the importance of detailed information with regard to the context and timing of various exposures as people move through their environments. To illustrate this point, he showed a map of average exposure to fast food outlets across five boroughs of New York City, pointing out the richness of information it reflected. If the map were collapsed into zip code averages for density of fast food outlets, he said, it would conceal important details about how people move through a space and experience its food environment. The structure of people’s lived experiences matters, he underscored, noting that social networks are another example of a structure that affects the development of obesity.

Lastly, Hammond highlighted the dynamic nature of complex systems. He stressed the potential importance of the timing and sequence of exposures and interventions as related to the development of obesity and other chronic disease outcomes across the lifespan.

Turning to the implications of applying systems science approaches to obesity, Hammond emphasized that searching for a single cause of a

Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×

problem driven by a complex system is often misleading. Furthermore, he added, standard tools focused on single drivers may have limited ability to capture obesity’s complexity. Thus, he said, a growing consensus supports solutions that are broad enough to address a variety of contributing factors while also being sufficiently tailored to specific contexts and adaptive over time.

To support these points, Hammond referenced a series of reports endorsing the use of complex systems thinking, approaches, and models to inform solutions for obesity and other public health issues. These include several publications of the Institute of Medicine and the National Academies, particularly a 2012 report titled Accelerating Progress in Obesity Prevention (IOM, 2012) and a report on the use of systems science approaches in tobacco regulation (IOM, 2015), as well as a Healthy People 2030 report from the Department of Health and Human Services (Secretary’s Advisory Committee for Healthy People 2030, 2018) (see also IOM, 2012, 2015; IOM and NRC, 2015; NASEM, 2016; United Kingdom Government Office for Science, 2007). Hammond also mentioned that the National Institutes of Health (NIH) has invested in at least five networks of scientists who have applied systems science approaches to various issues.

Hammond proposed that the interest in systems science approaches is driven by a desire to support the selection of policies and solutions that leverage or at least address the complexity of the problems they seek to address. He elaborated that these approaches can help tackle the challenges of coordinating policy and action across sectors, cultivating connections among scientific disciplines, and navigating heterogeneity across settings.

Hammond drew a distinction between the terms “systems science” and “systems thinking” as applied to public health. He described the latter term as an approach to thinking about the world as opposed to the use of quantitative or qualitative methods. Stakeholders applying systems thinking, he elaborated, examine complex systems from a holistic perspective and search for points of leverage and coordinated solutions beyond traditional arenas of health. Public health has advanced from using systems thinking, he explained, to adopting specific systems science tools, both qualitative and quantitative.

Hammond briefly mentioned causal loop diagrams and systems maps as examples of qualitative tools, and then reviewed quantitative tools in more detail. He highlighted the existence of a great diversity of quantitative tools and implementation methodologies, referencing a listing of some of these methods, along with their potential uses, in the Accelerating Progress in Obesity Prevention report (IOM, 2012). All of these tools, he explained, embrace complexity and seek to understand the mechanisms that drive outcomes, but they differ in several respects. One difference is the perspective from which they seek to understand processes and outcomes, which he said

Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×

may be from the top down, marked by interest in the system’s structure, or from the bottom up, marked by interest in individual actors in that system and the fine details of how they interact. He identified as another difference the formalisms used, which include differential equations, a focus on network ties, and portrayals of individuals referred to as agents. Tools also differ in the relative emphasis they place on data inputs versus theoretical inputs, Hammond added, as well as their field of origin and the training required to use them. He noted that NIH’s initial investment and training in systems science approaches focused on system dynamics modeling, social network models, and agent-based modeling, which remain among the most common types of methods used in public health and obesity.

Hammond briefly recapped the history of complex systems science, pointing out that many of its methods date back to the 1950s or earlier (with the exception, for example, of agent-based modeling, which originated in the 1990s). These methods emerged from such fields as biology, social science, and engineering, he said, where many of them established strong track records as policy- or decision-making support tools under-girded by empirical experiments across a wide variety of topics. He added that a catalyst in merging these fields and methods was the creation of the independent, nonprofit Santa Fe Institute in the 1980s, which is dedicated to the multidisciplinary study of fundamental principles of complex adaptive systems.

Systems science approaches were first used in public health to inform control and management of infectious diseases, Hammond observed, activity that accelerated after the September 11, 2001, terrorist attacks amid concerns about bioterrorism and pandemics. Further traction was gained with an investment by NIH that resulted in the creation of the MIDAS (Models of Infectious Disease Agent Study) network in 2003, which Hammond said has had important scientific and policy impacts. MIDAS makes extensive use of agent-based computational modeling, he noted, a systems science method that plays a key role in informing policy options related to the coronavirus disease 2019 (COVID-19) pandemic.

The models used by MIDAS were first summoned to inform policy solutions during the H1N1 crisis in 2009 and 2010, Hammond recounted, by which time they had undergone several years of development and improvement since the launch of MIDAS. He explained that this chronology suggests the time horizon for which these models might be expected to yield actionable results in the context of obesity solutions.

According to Hammond, systems science approaches have been applied to public health issues for three key purposes, which he said could guide thinking about the application of these approaches to obesity issues. One purpose is etiological, which, he explained, includes generating hypotheses and building theories to help uncover potentially unobservable mechanisms

Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×

that drive behaviors or outcomes in the real world, informing future intervention targets, timing, dosage, and data collection needs and methods. The second purpose, Hammond continued, is to deduce retrospectively the key influences on an intervention’s outcomes so as to better understand why it failed or succeeded, and to inform future data collection needs for evaluation and intervention design. Models used for this purpose, he added, can also inform the replication of successful intervention elements in different settings or over longer time horizons. A third purpose is prospective modeling, which Hammond described as helping to forecast potential outcomes of different policy and intervention options. He suggested that prospective modeling is a particularly attractive method for informing obesity solutions because it turns implicit mental models into explicit scenarios that harness a problem’s complexity and help account for heterogeneity across individuals and settings. Furthermore, he continued, this virtual modeling may be more efficient, ethical, and/or cost-effective than conducting large-scale, real-world experiments.

Hammond shifted his exposition of systems science history to describe how obesity research has been informed by systems science approaches. Publications first appeared in this area around 2009, he remarked, and these initial publications explained why systems science approaches are apt for studying obesity and designing interventions to address it, reviewed potential applications of specific systems science methods and tools to obesity, and suggested potential data requirements and obstacles (Hammond, 2009). Within a short time, he said, an editorial in a special issue of the American Journal of Public Health (AJPH) argued that systems science approaches were a revolution for public health policy research (Mabry et al., 2010). Next, NIH funded a network of researchers who were using systems science approaches to examine the etiology of obesity; the results of this investment were reported in another special issue of AJPH in 2014 (Mabry and Bures, 2014). The policy implications of some of these models were reported in an obesity-focused issue of The Lancet, Hammond added, and the NIH-funded work on etiology has continued for obesity and other topics.

Hammond reported that the concept of complex systems modeling for obesity solutions entered mainstream consciousness in 2012 with the publication of the Accelerating Progress in Obesity Prevention report (IOM, 2012). He mentioned specific applications of this concept, including COMPACT (Childhood Obesity Modeling for Prevention and Community Transformation) (see Chapter 4); a paper advocating for combining systems science approaches that have complementary strengths (Hennessy et al., 2020); and two reviews describing current uses of complex systems modeling for efforts addressing obesity, diet, and food systems (Langellier et al., 2019; Morshed et al., 2019).

Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×

Years of investment, training, and collaboration have contributed to the evolution of systems science approaches in the context of obesity, Hammond maintained, noting that in the future, these approaches can inform the sustainable, effective implementation of multifaceted or whole-of-community interventions. Moreover, he argued, these approaches can enable disease prevention efforts that are as precise and tailored as disease treatments informed by precision medicine (Gillman and Hammond, 2016). Hammond concluded his presentation with the following quote:

Furthermore, there is a need to train scientists in academia, the private sector, and government agencies in all aspects of complex systems approaches—including systems research design, data collection, and analytical methodologies—and the use of models appealing for private sector actors and government agencies to leverage systems science approaches. (IOM and NRC, 2015)

SYSTEMS THINKING TO UNDERSTAND AND IMPROVE POPULATION HEALTH

Sandro Galea, dean and the Robert A. Knox Professor at the Boston University School of Public Health, discussed using systems thinking and systems science approaches to understand and improve population health. A fundamental point, he asserted, is that population health and its complex challenges, such as obesity, cannot be broadly understood without taking a systems science approach.

Galea asserted that systems science approaches are well suited to application to population health, which has been defined as “the health outcomes of a group of individuals, including the distribution of such outcomes within the group” (Kindig and Stoddart, 2003). Such approaches offer unique insights into the overall health of populations and the health inequities that exist within them, he elaborated, adding that these systemic insights are critical for informing interventions to improve public health.

Galea shared an illustration that he said represents typical, deterministic approaches to population health. He explained that such approaches do not apply complex systems lenses, and assume that all individuals are relatively interchangeable and lack specific network structures and interconnectivity. Furthermore, he added, these approaches are broadly predicated on an assumption of a linear relationship between exposure and outcome functions.

In reality, Galea pointed out, substituting one person for another could have vastly different effects in a given situation because populations are heterogeneous. They comprise markedly diverse individuals who have complex contact structures, social networks, and connections, he elaborated, and the same inputs do not always produce the same outcomes because

Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×

populations evolve and adapt over time. He added that real populations also display emergent properties—properties that emerge only when individuals interact in a broader population and that are distinct from the properties of each individual or the aggregate of individuals.

Galea described the COVID-19 pandemic’s reflection of the true characteristics of populations, pointing out such features as a tremendous diversity in global behaviors, randomness in the location of outbreaks in some urban areas but not others, and the essential role of contact structures and networks in mapping disease transmission. It is truly a picture of a complex system, he observed, emphasizing the importance of addressing such complex public health issues with approaches that go beyond linear, deterministic frameworks.

Turning to discuss obesity, Galea indicated that its exhibiting of classic epidemic behaviors makes it well suited to examination and intervention using systems science approaches. A vast range of inputs contribute to the determination of obesity, he pointed out, recalling the UK Foresight Group’s map that Hammond had shared. According to Galea, the map makes it abundantly clear that the only rational way to address obesity is to approach it as a complex system.

Galea shifted to expound on the compatibility of population health with systems thinking by reviewing three of nine principles that have been advanced as foundations of population health science (Keyes and Galea, 2016). The first, he began, is that population health manifests as a continuum. He explained that although this appears obvious, the lack of a simple dichotomy of healthy and unhealthy leads to recognition of the importance of understanding the full spectrum of health within and across populations. As an example, he presented a normal curve distribution of body mass index (BMI) in a population, pointing out that simply providing the proportion of the population that is above the cutoffs for overweight (BMI ≥25 kg/m2) and obesity (BMI ≥30 kg/m2) masks the full scale of distribution of weights and BMIs throughout the population. He asserted that this distribution and the dynamics of its shifts are as or even more important than those proportions. He then showed two overlaid curves illustrating changes in the distribution of BMI between 1976–1980 and 2005–2006 (see Figure 2-2).

Figure 2-2 depicts a more complex shift than that of individuals having (or not having) obesity, Galea emphasized; it demonstrates that the nature of the population curve has changed, which he said requires more sophisticated thinking about the underlying characteristics of the population. He presented another graphic (see Figure 2-3) illustrating the distribution of serum cholesterol levels in men who did or did not develop coronary heart disease in the Framingham Heart Study. Although there was a cutoff cholesterol level above which risk for developing the disease increased,

Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×

Galea noted that the curves were quite similar to that point, indicating the importance of understanding the broad factors driving the full shape of the population curve.

Galea moved on to discuss a second foundational principle of population health science: that small changes in ubiquitous causes may result in more substantial change in the health of populations relative to larger changes in rarer causes. He used the metaphor of a goldfish in a fishbowl to illustrate this point. If the goldfish wanted to be healthy, he said, it might receive a variety of behavioral suggestions, such as swimming daily laps and regulating its intake of fish food. But if the water in the bowl—which is ubiquitous around the goldfish—were not changed, such behavioral changes would matter little because the unclean water would hasten its death.

The public tends to think about improving health in a linear, deterministic way, Galea continued, referencing a series of consumer health and fitness books that promote individual behavior changes. He maintained that these books fail to sufficiently consider broader, ubiquitous environmental factors, such as increasing portion sizes of packaged and restaurant foods, which influence and can undermine individual behavior changes. Differences in food environments can help explain heterogeneity in the prevalence of obesity across adjoining communities, he added, and he called for the application of systems science approaches to understand both the individual drivers and the ubiquitous forces that affect population health.

Galea highlighted a third principle: the magnitude of an exposure’s effect on disease is dependent on the prevalence of the factors that interact with that exposure. To illustrate this principle, Galea asked workshop participants what percentage of obesity risk is determined by one’s genes (the exposure), based on the assumption that only two factors—genes and environment—matter for the outcome. He shared a series of images to simulate two scenarios, one in which the obesogenic environment (i.e., the interacting factor) pervades the entire population and therefore results in obesity among all of the population’s genetically predisposed individuals. In this scenario, Galea explained, 100 percent of obesity risk is determined by genes. He described a second scenario with the same population, in which the obesogenic environment affects only a segment of the population. In this scenario, he pointed out, the only genetically predisposed individuals who will develop obesity are those in the segment affected by the obesogenic environment, which in this example is equal to 40 percent of obesity risk determined by genes.

Returning to his question, Galea explained that it is impossible to determine the percentage of obesity risk that is determined by one’s genes unless the environment that creates the conditions for the disease is also known. This matters, he continued, because under a plausible assumption

Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×

of co-occurring causes, the relationship between one cause and the outcome can be understood only if the other cause is also understood. This principle is difficult to grasp, he acknowledged, but he asserted that it is the most fundamental argument for the application of systems thinking to obesity etiology because it accepts that multiple co-occurring factors determine obesity.

Galea moved on to the final portion of his presentation, a discussion of complex systems and counterfactual thinking. He described the concept of a counterfactual by comparing two universes: an observed universe and a parallel universe that is identical to the observed one except for a single variable and the outcome. The observation of a different outcome based on the manipulation of only one variable allows a researcher to determine that the manipulated variable has a causal effect on the outcome, Galea explained, and instills confidence in selecting that variable as the intervention target. He noted that counterfactual theories underlie all of modern causal thinking, but that it is difficult to simulate such parallel universes.

To overcome this difficulty, Galea suggested, systems science approaches can complement observational and experimental techniques to simulate the counterfactual, and although all approaches have imperfections, together they can contribute to a better understanding of strategies for improving population health (Marshall and Galea, 2015). He cautioned against elevating a single approach as the “gold standard.” This designation is commonly given to randomized controlled trials (RCTs), he observed, but he asserted that no such “gold standard” exists. Rather, he maintained, the gold standard is a well-designed study that uses appropriate methodologies for the question of interest, articulates its limitations, and offers findings and conclusions that can advance improvement in population health.

Finally, Galea reiterated that principles of population health science illuminate the utility of applying systems thinking to population health issues, but he also urged workshop participants to approach these issues with the mindset of “as simple as possible, but not simpler,” a quote attributed to Albert Einstein. In the context of this quote, he recounted a story about John Snow, a pioneer in the field of epidemiology, who convinced local officials to remove the handle of a water pump that he suspected to be the source of a cholera outbreak in London in the mid-1800s. Snow’s discovery made a seminal contribution to the understanding of disease transmission, Galea affirmed, but he suggested that this episode in the history of public health has led the public health workforce to a belief in the existence of simple solutions. He contended that it is ineffective to try to simplify solutions by searching for a single cause on which to focus intervention.

In completing his presentation, Galea suggested that the United States has been overly simplistic in its approach to population health, noting its trend of greater health expenditures yet lower life expectancy relative to other high-income countries (Roser, 2017). Population health epidemics

Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×

such as obesity are driven by a complex set of forces, he stressed. Thus, he called for systems thinking to identify key levers for intervention that will achieve maximal impact to improve overall population health and narrow health inequities.

APPLICATIONS OF SYSTEMS SCIENCE: CONTEXT, CAUSALITY, AND COMMUNITIES

Douglas Luke, professor and director of the Ph.D. program in public health sciences at Washington University in St. Louis, provided an overview of applications of systems science. Systems science approaches are beneficial for chronic disease prevention and policy implementation, he began, echoing Hammond’s and Galea’s statements about the complexity of public health problems. He noted that this complexity has led to the coining of the term “wicked problems,” which refers to complex problems that resist resolution. He listed characteristics of such problems, including the involvement of multiple actors and sectors, high economic and/or political stakes, interconnectivity with other problems, and lack of agreement or clarity regarding solutions.

Luke shared a causal loop diagram of the complex tobacco landscape (see Figure 2-4), pointing out its heterogeneous, interconnected elements and actors. This complex system cannot be understood by studying its individual parts, he maintained, adding that the important behavior of the system emerges over time when it is examined as a whole. This reality leads to a set of assumptions about the features of complex systems, he explained, including nonlinearity, non-normality, heterogeneity, multiple levels of analysis, and dynamism with feedback. Luke emphasized the interaction of actors in complex systems, and noted that many of the study designs and analytical tools that are typically part of the training of most researchers tend to be poorly suited to addressing this multilevel interaction and heterogeneity.

Luke next described insights related to context, causality, and communities that have stemmed from his group’s application of systems science approaches to the study of chronic disease. The first insight, he began, is that these approaches facilitate exploration of the role of context in chronic disease processes. He stated that chronic disease development and interventions are shaped by numerous contextual factors that traditional methods often ignore, yet such problems as obesity cannot be truly understood without considering context. Luke listed important layers of context, including social (both individual and organizational), economic, physical, temporal, and political/historical. It is important to use frameworks or models that capture this rich context, he stressed, noting that the focus is often on specific elements and outcomes of the program, policy, or practice being implemented.

Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Image
FIGURE 2-4 The complex tobacco landscape.
SOURCES: Presented by Douglas Luke, April 6, 2020. IOM, 2015.
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×

To develop more effective interventions, he suggested, one must understand how they are embedded within higher- and lower-level contexts.

To illustrate this point about the importance of considering context, Luke described a type of traditional epidemiological model that predicts and models the time course of an infectious disease outbreak. An underlying assumption in such a model, he explained, is the average number of contacts per person per unit time, which assumes random mixing and ignores social structures. Nonetheless, this traditional model with its simplifying assumption worked until the 1980s, Luke recounted, when the AIDS epidemic elevated the role of social context in disease transmission. The traditional models were then replaced by network graphics, he continued, which illustrate the nonrandom, social structure of disease transmission (see Figure 2-5). To provide a recent example, Luke described features of a contact tracing network graphic of COVID-19 outbreak clusters in Singapore.1 He emphasized the importance of physical and social context, such as attending the same church service or traveling on the same airplane flight, in understanding the dynamics of disease transmission.

Luke suggested that this type of analysis is a particularly helpful tool for describing the individual and social contexts for obesity. As an example of social network analysis, he referenced a publication examining the role of peer group structure in smoking initiation among adolescents, highlighting its use of social network analysis to illustrate patterns of peer influence that a regression model would be unlikely to capture (Ennett and Bauman, 1993). As a second example, he cited a series of publications describing the clustering of obesity in personal networks, a phenomenon he said suggests that even noncontagious diseases can be spread through the influence of close social networks and shared environments (Christakis and Fowler, 2007).

Luke went on to observe that network analysis can be extended to map organizational systems, a point he illustrated by referencing a map of a national network of agencies that collaborate to provide tobacco control services and resources for LGBTQ communities (see Figure 2-6). This map illustrates that many of the agencies are connected only through the lead agency (i.e., direct, separate lines from each agency to the lead agency) he pointed out, which he said alerted the funder that additional interorganizational ties would make the system more robust and resistant to network collapse if the lead agency were compromised. Organizational and social systems mapping is also useful for designing interventions or implementing new best practices in clinical settings, Luke added, settings in which interpersonal interactions are important.

Luke then discussed a second insight—that systems science approaches facilitate the exploration of underlying causal mechanisms. He explained

___________________

1 See https://www.againstcovid19.com/singapore/cases/search (accessed September 16, 2020).

Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Image
FIGURE 2-5 The first AIDS network graphic.
SOURCES: Presented by Douglas Luke, April 6, 2020. Auerbach et al., 1984; Luke and Stamatakis, 2012.

that although traditional methods such as RCTs are useful for establishing causality—that is, whether something works—they are less useful for determining how or why something works. He referenced a seminal paper describing numerous uses of computational systems models in public health and societal contexts beyond predicting disease mortality, such as illuminating dynamics central to the system at hand and suggesting analogous dynamics within other systems (Epstein, 2008).

To illustrate how complex systems models can help identify the inner workings of underlying causal mechanisms, Luke described Tobacco Town, a modeling effort from his policy research on tobacco control. Tobacco Town, which uses the systems science approach of agent-based modeling, is a policy laboratory used to explore potential impacts of various retail policies across contexts and populations. Luke explained that Tobacco Town is based on evidence indicating that people who live in neighborhoods with a high density of tobacco retailers are more likely to start smoking and

Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Image
FIGURE 2-6 Example of an organizational systems map.
SOURCE: Presented by Douglas Luke, April 6, 2020.

have difficulty quitting. The logical policy response would be to reduce the density of tobacco retailers, he continued, but such reduction is difficult to study in the real world. Different communities take different actions at different times, he explained, and clarity is lacking about how density reduction would work in practice.

Luke’s research team developed a “mechanism metaphor,” he recounted, starting with the assumption of a simple linear relationship whereby a high density of tobacco retailers lowers the time and monetary costs of obtaining tobacco products (see Figure 2-7). He described a mental model in which a sample tobacco user lives in a dense, urban environment. This person does not have to travel far to purchase tobacco, he pointed out, and her behavior is unlikely to change if only one or even a few retailers in a given

Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Image
FIGURE 2-7 Assumed linear relationship between density of tobacco retailers and time and monetary costs of procuring tobacco products.
SOURCE: Presented by Douglas Luke, April 6, 2020.

intersection are removed. Only when she has to travel farther to purchase tobacco is she likely to notice that the density of retailers has been reduced.

Luke identified as one implication of this mental model that the assumption of linearity is likely false; instead, there is likely a nonlinear or threshold effect. Another implication, he continued, is that tobacco-seeking behaviors and any interventions designed to influence them are environmentally dependent, so Tobacco Town entails building computer-based, virtual towns as simulation models to identify interactions between the retail tobacco environment and purchase and use behaviors. Researchers collaborate with community stakeholders to tailor models to specific communities, he elaborated, so they can test the impact of policies prioritized by community members and disseminate the results.

Luke reviewed examples of Tobacco Town in action to show that it confirmed the nonlinearity between retailer density and the costs of procuring tobacco products. Reduction of retailer density may need to reach a threshold, he explained, before behavior effects are observed. The patterns of these effects differ based on the urbanicity and income levels of modeled towns, he added, an observation that implies that policies have different potential for affecting disparities and behaviors in different types of communities.

Luke shared a heat map visualization of the potential effects of two types of tobacco retailer policies as an example of how agent-based

Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×

modeling can facilitate the exploration of underlying mechanisms. One policy requires a 600-meter buffer between retailers, while the other requires a 600-meter buffer between retailers and schools. Luke explained that, when applied to the city of Philadelphia, the first policy reduced retailer density from 4.5 to 0.6 retailers per square kilometer and increased the average distance between residents and retailers from 200 to 480 kilometers. By contrast, he said, the second policy reduced retailer density from 4.5 to 0.76 retailers per square kilometer and increased the average distance between residents and retailers from 200 to 730 kilometers. Luke pointed out that although the resulting retailer density metric looks similar for both policies, the second policy has a greater impact on proximity because it creates a much higher average distance between residents and retailers. According to Luke, the underlying mechanism of proximity in this example has implications for community policy development and intervention.

Luke then outlined three advantages of such mechanism-based modeling. First, understanding how a policy works opens the door to designing more effective policies or interventions with less guesswork; second, it helps stakeholders advocate more persuasively for programs to be maintained or terminated; and third, it offers a new way to understand health disparities by focusing on the processes through which those disparities arise, rather than simply documenting them.

Luke’s third insight was that systems science approaches can produce tools that are highly relevant and useful for community stakeholders and policy makers. He revisited the paper he had mentioned previously to list three additional uses of computational systems models: training practitioners, disciplining the policy dialogue, and educating the public (Epstein, 2008). To illustrate this third insight, he recalled an application of the Tobacco Town model to communities in Minnesota where the effects of policies to reduce the density of tobacco retailers depended on context. Baseline retailer density varied by community urbanicity (rural, suburban, or urban) and income level (high or low), he elaborated, and the Tobacco Town dashboard illustrated how various policies would potentially affect retailer density across these different types of communities. He noted that as policies became stronger, to include restricting tobacco sales at pharmacies and requiring buffers between retailers, retailer density dropped considerably in all communities, and disparities in density diminished across communities of varying urbanicity and income levels.

In a final example of disseminating systems science tools to community stakeholders, Luke showed workshop participants an interactive Tobacco Swamp dashboard his group is developing to help policy makers explore the effects—based on underlying computational models—of different retailer-focused tobacco control policies in different cities. He loaded Washington,

Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×

DC, into the dashboard2 as an example, and observed that nearly 100 percent of its residents are within a 10-minute walk of a tobacco retailer. As he selected various policy interventions in the dashboard, he pointed out how the different options, such as school buffers, could be expected to affect retailer density in the city.

PANEL DISCUSSION

Rivera moderated a panel discussion with the three session speakers following their presentations. Topics addressed included the importance of applying systems science approaches to obesity solutions and potential metrics of success; examples of communicating systems science concepts to academics, policy makers, and the public; the role of leadership and the alignment of multisector action in systems science approaches; the types of training and collaboration that can help prepare stakeholders to undertake systems modeling efforts; and funding opportunities for systems science approaches.

Importance of Applying Systems Science Approaches and Metrics of Success

Rivera began by asking the speakers to explain why it is critical to apply a systems science approach to obesity solutions and to describe how they would measure success in doing so during the next 5 years. Hammond reiterated the importance of implementing packages of interventions to address the obesity epidemic, an approach he said is inherently systems-focused because it calls on stakeholders to assemble different packages for different settings and to coordinate efforts into a coherent set of solutions. He voiced his 5-year vision for progress in blending systems science tools with conventional tools to create a package of actions and processes that communities can leverage to tackle obesity as they apply this package to their specific context and evidence base.

Galea agreed that obesity solutions are not one size fits all, and emphasized the importance of identifying the highest-priority factors to address in a particular place at a given point in time. Different contexts call for different sets of solutions, he argued, cautioning against deterministic perspectives that simplify problems and apply the same approach across the board. He suggested that an indication of success would be the use of sophisticated thinking to assemble a suite of context- and time-specific approaches to obesity solutions in communities.

___________________

2 See https://aspirecenter.org/tobacco-swamps (accessed October 20, 2020).

Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×

Luke commented on the difference between efficacy and effectiveness, noting that although an intervention’s success may be predicted under ideal conditions, its real-world outcome can be quite different. The difference is often systemic in nature, he suggested, giving the example of a real-world tobacco retail policy that generates responses from industry and retailers. Such effects are sometimes called unintended consequences, Luke added, but he noted that some systems science models reject this term and simply call them consequences.

Communicating Systems Science Concepts to Academics, Policy Makers, and the Public

Rivera next asked the speakers to share examples of an effective explanation of systems science concepts to academics, policy makers, and the public. Hammond acknowledged that different strategies for communicating with each of these audiences are often warranted, but stated that data visualization tools are useful for explaining systems science models to both policy makers and the public. Visualizations can convey sophisticated dynamics without complex data tables and equations, he elaborated, and can communicate powerfully as they portray location- and context-specific scenarios. As for academics and scientists, Hammond suggested that they are interested in whether systems science models yield insights and results that might have been unattainable with other techniques.

Galea suggested a paradox in that scientists may find it more difficult to grasp systems science approaches because academic training tends to guide doctoral students to narrow their expertise to a particular exposure and outcome. Conversely, he said, systems science approaches have intuitive appeal for policy makers and the public, although he warned that such methods as agent-based modeling can lead policy makers to have false confidence in the ability to predict future outcomes. The communication challenge for modelers, he argued, is to explain the utility of these models alongside their caveats and limitations. Luke concurred, urging researchers to be “appropriately skeptical” about a model’s capabilities when they discuss using modeled results to plan action with community stakeholders.

The Role of Leadership and Alignment of Multisector Action

Rivera asked the speakers to discuss the role of leadership and the alignment of multisector action based on systems science approaches. Galea referenced epidemiologist Geoffrey Rose’s declaration that it is not population health scientists’ role to determine action; rather, their role is to present options and let policy makers balance various inputs to determine the appropriate action. As an example, he noted that many infectious disease

Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×

models publicized during the COVID-19 pandemic have been used to determine blunt policy approaches. This is inappropriate, Galea asserted, and he urged that leaders avoid “model absolutism” by considering a full spectrum of potential consequences—including those forecast by models and those not modeled—in light of various societal values and the moral and philosophical worldviews that underlie them. Models are “vastly imperfect,” he stressed, a limitation that he said elevates the role of leadership in balancing both modeled and nonmodeled inputs.

Hammond emphasized that models are just one of many inputs for decision making and do not replace the need for judgment or eliminate uncertainty. He suggested that models could be more useful for decision making if stakeholders and decision makers were engaged early in their development and use. He also pointed to MIDAS as a network including multiple models that are developed by different groups and can be communicated as an ensemble to policy makers. He shared a lesson from the 2009–2010 H1N1 pandemic: the importance of having a “translator” who understands the abilities and limitations of models and is conversant in the vernacular of both modelers and policy makers. This role can bridge the two worlds, he explained, and help prevent problems that could otherwise arise.

Luke concurred in urging that systems modeling efforts be multidisciplinary and transdisciplinary from the beginning, including modelers, subject-matter experts, translators, and community stakeholders and decision makers. He recounted past efforts that failed to incorporate these perspectives, which he said turned out to be a poor use of resources.

Hammond highlighted the importance of training, explaining that systems models are relatively easy to use poorly if not applied and interpreted properly. The skills of modeling and computer programming are distinct, he said, explaining that the former involves discerning what variables and assumptions to include and how to engage with the right stakeholders.

Training and Collaboration

Rivera asked the speakers to elaborate on the types of training and collaboration that can help prepare stakeholders to undertake systems modeling efforts. Hammond underscored that substantial training and experience are essential to the appropriate use of systems science tools. He called for training that provides learners with a sophisticated understanding of the tools’ capabilities and limitations so they can participate effectively in teams that include modeling experts. He added that to become a full-fledged modeling practitioner, a multiyear investment is expected, typically in the context of a doctoral program or apprenticeship.

Luke remarked that the environment for training opportunities has improved over the past few years. He mentioned a summer institute that he

Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×

conducts with Hammond at Washington University in St. Louis to provide an introduction to predominant systems science approaches that enhance the social impact of health and social science research. An impetus for the institute, he explained, was that many of their research partners lacked access to full doctoral-level classes in systems science approaches.

Galea agreed that it takes dedication to develop the expertise necessary to use systems science approaches effectively. He also emphasized the importance of participating in transdisciplinary teams so that population health stakeholders can engage with expert modelers. Luke concurred with the value of a team-based approach, noting that all members need not be expert modelers. He added that systems science training often emphasizes systems science methodologies, but not systems thinking. Galea agreed, and cautioned against blindly using a particular method without contemplating whether it is being used to ask the right questions.

Rivera commented that transdisciplinary efforts had become more common over the course of his career. When problems are discussed among transdisciplinary collaborators, he observed, their diverse mindsets can initially breed tension, but as they are exposed to different ways of thinking about the problem at hand, these initial mindsets evolve and a shared vocabulary develops, allowing everyone to understand the problem better. Hammond described systems modeling as a journey rather than a destination and provided two examples to support his point. First, he explained that some of the most valuable contributions to modeling come from the process of documenting mental models. This process includes providing data that support the model’s assumptions, Hammond elaborated, and is aided by qualitative and quantitative tools that help one think like a modeler. Second, he continued, models are often iterative. The success of such groups as MIDAS, he observed, is based on sustained investment that allowed such iteration to occur and the questions, data, and assumptions that feed into models to be refined, making them better.

Funding Opportunities

Rivera invited the speakers to share funding opportunities for the use of systems science approaches. Galea and Luke reported a gradually improving landscape with regard to funders’ recognition of the value of incorporating these approaches into federally funded research on public health topics. Luke observed that NIH’s review committees are increasingly including members with modeling expertise, and Hammond suggested that increased training will grow the pool of experts who can fill reviewer roles. Funding opportunities for systems science approaches have been limited, Hammond suggested, by the relatively small set of experts who are qualified to review proposals that include these approaches. Another limitation,

Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×

he said, is the relatively long time horizon required to iterate and fine-tune systems efforts, which has implications for the types of funding vehicles and teams that are assembled.

AUDIENCE DISCUSSION

Following the panel discussion, speakers addressed questions from workshop participants about mapping systems components, building community capacity to use models and propel policy change, lessons learned from the transportation sector, funding mechanisms to support systems science approaches, and training physicians in health systems science.

Approaches to Mapping Systems Components

In response to a question about approaches to mapping complex systems in order to identify relevant features and processes, Galea replied that when stakeholders review a base of literature as part of the modeling process, they identify parameters that can inform models, as well as areas in which literature does not exist to inform other parameters. This identification of gaps leads to proposing studies that can fill these gaps, he observed, or to articulating the ranges of assumptions that will be included in a model.

According to Hammond, participatory group model building is a qualitative systems science approach that can engage community members and stakeholders in articulating their lived experiences in systems. Tapping into community members’ perspectives and experiences can help researchers better understand and visualize systems structures, he explained, adding that the translation of these qualitative inputs into quantitative models must be done carefully.

Luke noted that systems science approaches can be used to inform theories and meet community needs. He pointed out that systems mapping, an approach that uses network methods to examine the organizational interconnections in a community, helps provide a springboard for a community’s future research, evaluation, and planning.

Building Community Capacity to Use Models and Propel Policy Change

Luke emphasized that it is important for researchers to understand community partners’ perspectives on the research. They must explain clearly why community partners should care about the work, he maintained, and this can be accomplished by involving them from the outset. He referenced his colleague Ross Brownson’s concept of “designing for dissemination,” which entails thinking about dissemination at the start of a project.

Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×

In response to a question about how to use systems maps to drive policy change, Luke replied that resources are available online with which communities can test different scenarios. Researchers may get face time with community partners, he added, but virtual technologies provide access to partners around the world. Most online resources are open-access and easily available, Luke continued, and some sustainability assessment tools have been released under the Creative Commons license. It takes expertise to develop dashboards that are user-friendly in the way they present results and allow stakeholder interaction, he noted, calling out R Shiny products as helpful for that purpose.

Luke explained that his group developed its Tobacco Town models in close consultation with a variety of community and policy partners to increase the models’ potential for community impact. He highlighted the value of working with legal experts to help shape how potential tobacco control policies are presented, given that they may be contested in court over First Amendment or other constitutional challenges.

Lessons Learned from the Transportation Sector

Models have been used to inform urban development, Luke said, such as by incorporating data on travel patterns to aid in improving traffic congestion. Rivera observed that transportation modeling efforts tap into a plethora of GPS and other contextual data (e.g., from satellite radio) to generate their results. He suggested that using these sources of contextual data would improve understanding of obesity as well. Hammond pointed out that in any sector, policy decision making involves trade-offs between various outcomes of interest, and that models can help optimize different policy objectives.

Funding Mechanisms to Support Systems Science Approaches

Luke observed that some systems science approaches, such as group model building and network mapping, require less time and money than others. Therefore, he reasoned, it would be useful to have a variety of funding opportunities to support different types of systems science approaches. In the context of obesity, Hammond suggested that funding opportunities incentivize the formation of interdisciplinary teams with heterogeneous expertise in methods and content, ideally across institutions and geographies. According to Hammond, some grant development mechanisms, such as the NIH process, are generally not well suited to this objective, and a funding mechanism to ensure such interdisciplinary collaboration would need to be carefully crafted. Galea advocated for including junior investigators in these collaborations to create a pipeline of scholars who naturally engage in systems thinking.

Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×

Training Physicians in Health Systems Science

A workshop participant stated that health systems science is now considered the third pillar of medicine, joining basic and clinical science, and asked the speakers how future physicians could be trained to understand the role of human factors, systems engineering, leadership, and patient improvement strategies so as to transform health care and ensure greater patient safety. Galea said he was happy to hear that systems thinking is entering medical school curricula, noting that it is important for physicians to recognize that a complex set of determinants drives patients’ clinical outcomes. He cautioned against charging physicians with solving the complex systems problems that ultimately shape population health, noting that their primary role is to provide clinical care, which is only one component of broader systems. Hammond agreed, and added that systems science approaches can be used to study and make changes in the health care system.

CLOSING REMARKS FOR PART I

Following the discussion, Christina Economos, co-founder and director of ChildObesity180 and professor and New Balance chair in childhood nutrition at the Friedman School of Nutrition Science, Tufts University, recapped key points from Part I of the workshop:

  • Systems science approaches and systems thinking can be used in concert with traditional research designs and approaches. This implies that there is benefit in interdisciplinary collaboration involving trained teams of community stakeholders, translators, modelers, and scientists.
  • Contextual effects are important and can be measured if they are considered in analyses and captured with dynamic, real-time data.
  • Real-world experiments are costly and time-intensive, but systems science approaches using quantitative and qualitative methods allow for tailoring and testing of multilevel solutions and policies—such as those that address health inequities—in specific communities, populations, and contexts.
  • Advances in data visualization have helped communicate the outputs of systems science approaches, demonstrating the value of these approaches and making them more accessible to scientists, practitioners, policy makers, trainees, and the public.
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 5
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 6
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 7
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 8
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 9
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 10
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 11
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 12
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 13
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 14
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 15
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 16
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 17
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 18
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 19
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 20
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 21
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 22
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 23
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 24
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 25
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 26
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 27
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 28
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 29
Suggested Citation:"2 Overview of Systems Science Theories, Approaches, and Applications." National Academies of Sciences, Engineering, and Medicine. 2021. Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25766.
×
Page 30
Next: 3 Complex Systems in Society and the Context for Obesity »
Integrating Systems and Sectors Toward Obesity Solutions: Proceedings of a Workshop Get This Book
×
Buy Paperback | $60.00 Buy Ebook | $48.99
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

A virtual workshop titled Integrating Systems and Sectors Toward Obesity Solutions, held April 6, 2020 (Part I), and June 30, 2020 (Part II), was convened by the Roundtable on Obesity Solutions, Health and Medicine Division, National Academies of Sciences, Engineering, and Medicine. The workshop introduced the concept of complex systems and the field of systems science, and explored systems science approaches to obesity solutions. Speakers provided an overview of systems science theories, approaches, and applications, highlighting examples from within and outside the obesity field. Presentations and discussions examined complex systems in society that have the potential to shape public health and well-being, and considered opportunities for systems change as they relate to obesity solutions. Specifically, the workshop explored factors that can influence obesity - such as (in)equity, relationships, connections, networks, capacity, power dynamics, social determinants, and political will - and how these factors can impact communications and cross-sector collaboration to address obesity. This publication summarizes the presentations and discussion of the workshop.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    Switch between the Original Pages, where you can read the report as it appeared in print, and Text Pages for the web version, where you can highlight and search the text.

    « Back Next »
  6. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  7. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  8. ×

    View our suggested citation for this chapter.

    « Back Next »
  9. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!