Table 1 Methods used, reasons for using complex systems methods, participatory methods and research aims

System mapping (n = 11)

Group model building (n = 5)

Causal loop diagram and behaviour over time graph

Brennan (2015)

Understand complex system (understand insights from multiple communities regarding behaviours of systems affecting health); identify causal relationships among variables

Yes

To report on an evaluation of a multi-community healthy eating and active living community-based initiative; integrated study

Causal loop diagram

Guariguata (2021)

Understand complex problem; identify intervention points and causal pathways

Yes

To report on factors that influence population physical activity and identify potential areas for intervention; sole focus on PA

Causal loop diagram and behaviour over time graph

Keane (2015)

Understand complex problem (identify perceived influences on active living and healthy eating and how those influences change over time); identify feedback loops

Yes

To report on the evaluation of a healthy eating and active living initiative in one community; integrated study

Behaviour over time graph

Hoehner (2015)

Understand complex system (study system behaviour related to policies, environments, collaborations and social determinants, and compare trends among multiple communities); create a visual of how a variable changes over time

Yes

To report on perceived trends in system behaviour regarding past, current and future changes over time related to policy, systems and environment in terms of healthy eating, active living and childhood obesity; integrated study

Causal loop diagram

Waterlander (2021)

Understand complex problem (increase understanding of the complexity of obesity-related behaviours); Vvsualize elements in a holistic system and their causal relationships

Yes

To study the complexity of obesity-related behaviours in youth (diet, physical activity, sedentary behaviour, sleep); integrated study

Other mapping methods (n = 6)

Conceptual map (with policy audit)

Bellew (2020)

Understand complex problem (understand obesity and physical inactivity); identify the pathways to possible solutions

No (but group sense making after maps created)

To describe a national project in terms of a high-level conceptual systems map including physical activity influences, governance, translation, advocacy mechanisms and system intervention points for policies and programmes; sole focus on PA

Causal pathways diagram (with surveying and multilevel modelling)

Carlson (2012)

Understand complex system (understand the relationships between the built environment and physical health); create causal pathways and feedback loops

No

To describe how the built environment affects behaviour and health by conceptualizing how physical and perceived barriers impact the relationship between destination walking and self-reported health status; integrated study

Conceptual map

Cavill (2020)

Understand complex system (portray various interactions that constitute the system); promote cross-sectoral systems thinking; support future planning and implementation of actions

No (but group sense making after maps created)

To report on the value of system mapping in a city-wide population physical activity promotion programme; sole focus on PA

Concept map

Holdsworth (2017)

Understand complex problem (visualize relationships between concepts such as clusters of factors for dietary and physical activity behaviour)

Yes

To report on factors that influence dietary and physical activity behaviour in ethnic populations; integrated study

Mapping policy and practice

Murphy (2020)

Understand complex system (visualize nonlinearity in systems); ensure cross sectoral collaboration, coproduction, knowledge sharing, and defining roles; identify intervention design

Yes

To describe the process and results of a systems approach to enhance multisectoral communication and identify current good practices and future action for promoting population physical activity using the Global Action Plan on Physical Activity (GAPA) framework; sole focus on PA

System map

Signal (2012)

Understand complex system; co-create illustrations of systems /control parameters; recommend interventions

Yes

To identify possible public policy interventions targeting food security and physical activity and illustrate systems, control parameters, and interventions; integrated study

Simulation modelling (n = 10)

Agent-based modelling (n = 5)

Almagor (2021)

Understand complex system (simulate interactions of actors with one another and with the environment); explore potential impact of interventions in several domains

No

To simulate the impact of various physical activity interventions (active travel, outdoor play, school physical education and their combination) on children’s daily activities in an urban environment; sole focus on PA

Frerichs (2020)

Understand complex system (create prototype that emulates a system); develop models to influence decision making; pilot participatory approach to increase participants’ understanding of elements of an agent-based model

Yes

To describe participatory methods for creating a basic structure of a model regarding physical activity; sole focus on PA

Garcia (2018)

Understand complex problem (study patterns of leisure time physical activity considering the interactions among individual psychological attributes and built and social environments; understand interrelations and impacts of factors at different levels); identify interventions

Yes

To simulate population patterns of leisure time physical activity among adults taking into consideration the interaction between individuals’ psychological attributes and the built and social environments in which they live; focused solely on PA

Orr (2016)

Understand complex problem (study characteristics of the obesity system through multilevel analysis); identify policy options in terms of impact on body mass index of Black and White people

No

To simulate the impact of policies with respect to PA, neighbourhood food and educational environments on Black/White disparities in body mass index; integrated study

Salvo (2021)

Understand complex system [identify sustainable development goals (SDG) that may benefit from PA strategies and the multiple sectors and systems at play]; simulate the effects of interventions or strategies on recreational- and transportation-based PA and six SDG-related outcomes

No

To simulate the impacts of physical activity promotion strategies on SDG-related outcomes across high-, middle- and low-income country city types, and provide recommendations for future research, policy and practice; sole focus on PA

System dynamics modelling (n = 4)

With group model building and causal loop diagram

MacMillan (2014)

Understand complex problem (compare policies, incorporating feedback effects, nonlinear relationships and time delays between variables); involve policy, community and academic stakeholders in participatory modelling

Yes

To simulate the effects of five active transportation policy scenarios for commuter bicycling on injury, physical activity, fuel costs, air pollution and carbon emissions outcomes; integrated study

With behaviour change over time graphs and prevalence/incidence diagrams

Powell (2017)

Understand complex problem; demonstrate how different models impact childhood obesity over time; identify policy options

No

To simulate the impact of various policies on the future prevalence of childhood obesity; integrated study

With causal loop diagram

Soler (2016)

Understand complex problem; forecast the impact of various policies on outcomes (estimate premature deaths and medical and productivity costs)

No

To report on (1) short-term benefits of interventions targeted to decrease obesity by increasing physical activity, improving nutrition, decreasing tobacco use or decreasing exposure to second-hand smoke; and (2) long-term benefits in terms of health outcomes; integrated study

With causal loop diagram

Yang (2019)

Understand complex problem (understand system structure and the dynamic interaction of multiple variables)

No

To simulate how multilevel factors (i.e. urban design, urban sprawl, economic development, crime) influence and impact active travel to school and health behaviours of children; integrated study

Cross impact balance (n = 1)

With group modelling and causal loop diagram

Stankov (2021)

Understand complex problem (understanding of urban health issues); examine interactions among system elements and future scenarios

Yes

To study the strength and nature of relationships among factors that influence the transportation and food systems and identify future scenarios; integrated study

Network analysis (n = 4)

Social network analysis (n = 3)

With system inventory of actions

Blackford (2021)

Understand complex system (examine types of relationships within a network and increase understanding of network operations and the roles of key actors or organizations); identify strategies to strengthen networks; inventory or audit the physical activity, nutrition and obesity prevention initiatives taking place

No

To inventory current actions, examine networks and identify potential strategies for improving the obesity prevention system; integrated study

With system inventory of actions

Jancey (2021)

Understand complex system; inform policy and practice to improve obesity prevention interventions through increased understanding of network relationships

No

To report on network relationships to inform policy and practice regarding (physical activity and nutrition) obesity prevention; integrated study

Marks (2018)

Understand complex system (describe network structures as a measure of community capacity to implement and sustain interventions)

No

To report on the structure of professional obesity prevention networks as a measure of potential capacity to implement interventions; integrated study

Comparative network analysis (n = 1)

Network analysis

McGlashan (2018)

Understand complex system (compare thematic clusters identified by expert-driven and community-created system maps as to their size and strength of causal relationships)

No

To compare the expert-driven foresight obesity system map with community-based causal loop diagrams to determine similarities and differences; integrated study