Abstract
Background
Although research into the system dynamics underlying mental wellbeing has increased over recent years, it often remains rooted in positivist approaches that overlook lived experiences and dynamic system behaviour. This paper examines the system dynamics shaping preadolescent mental wellbeing by incorporating the perspectives of diverse stakeholders and analysing the system’s overall behaviour.
Design and methods
This qualitative study builds on earlier literature and scientific expert focus group data by incorporating the perspectives of preadolescents, caretakers, and professionals in Amsterdam, which were collected through a cocreated conversational tool and semistructured interviews. These insights informed a causal loop diagram (CLD), which was analysed using systems archetypes.
Results
Multi-actor input from preadolescents, caretakers, and professionals expanded a science-based CLD to 62 factors, 50 identified feedback loops, and eleven subsystems. It revealed how social norms, institutional structures, and local contexts interact to shape preadolescent mental wellbeing. Seven systems archetypes were identified – mainly ‘Success to the Successful’, ‘Shifting the Burden’, and ‘Fixes that Fail’ – offering an understanding of recurring patterns that sustain (poor) mental wellbeing, from fragmented poverty support to changing societal norms around smartphone use.
Conclusions
These findings underscore that complex public health issues have no quick or simple fixes. Instead, they require multilevel strategies that target various structural drivers while taking into account archetypal patterns. By mapping the system’s structure and behaviour, this study hopes to encourage policymakers, educators, and communities to adopt a systems (archetype) lens to understand preadolescent mental wellbeing and act accordingly.
Preadolescent mental wellbeing is shaped by a wide range of interacting dynamics, from livelihood insecurity affecting caretakers’ stress and availability, to social comparison norms that pressure young people to present a ‘perfect’ image. Societal norms such as rising individualism and constant online connectivity act as powerful drivers within these dynamics. By combining system structure (via a science-based Causal Loop Diagram) with system behaviour (via system archetype analysis), this study uncovers deeper, often hidden mechanisms, including context-dependent pathways and macro-level societal and political forces. These insights highlight an urgent need for preventive, cross-sector, and equity-focused public-health strategies that address the systemic conditions shaping children’s wellbeing, rather than relying on short-term or symptom-focused solutions.Significance for public health
Introduction
The mental wellbeing of preadolescents – children aged nine to twelve – has been under pressure in recent years. For example, the prevalence of emotional problems among 11- to 12-year-old girls has more than doubled, rising from 14% to 33% between 2017 and 2021. 1 These patterns are particularly concerning given that many adult mental disorders originate during (early) adolescence, a critical neurodevelopmental period. This underscores the need for early action strategies to support mental wellbeing.2,3 To inform the design and prioritization of such action strategies, insight into the factors that shape preadolescent mental wellbeing is needed. 4
Previous research has focused on isolated factors affecting preadolescents’ mental wellbeing, such as the quality of friendships or the impact of poverty.5–7 While these studies identified relevant associations, no single factor can fully explain the prevalence of preadolescent mental wellbeing. 8 Indeed, the interdependence between individuals and their environment has long been emphasized in psychological theory. For example, Bandura’s Social Cognitive Theory conceptualizes this through reciprocal determinism, in which personal factors, behaviour, and environmental context continuously interact.9,10 Similarly, Bronfenbrenner’s ecological systems theory posits that child development is shaped by interconnected environmental systems, ranging from the immediate microsystem (family) to the broader macrosystem (societal and cultural contexts).7,11 However, system dynamics theory takes this a step further, by examining the emerging dynamic interactions over time. 12
The use of approaches based on system dynamics theory to understand mental health has increased, primarily in clinical samples. Applications range from suicidal behaviour 13 to treatment access for individuals with eating disorders 14 and depression. 15 In a recent study, 16 we explored the system dynamics underlying preadolescent mental wellbeing from a scientific perspective on the basis of an umbrella review and focus groups with scientists. This study revealed the interplay between the individual, familial and societal levels. The identified dynamics revolved around, for instance, societal pressure to perform, identity formation, parenting stress, the norm to be always online, and neighbourhood constraints, each embedded within broader systemic forces beyond individual control.
Much of the expanding literature on preadolescent mental wellbeing has adopted positivist, measurement-driven designs, whereas interpretive approaches that capture stakeholders’ lived experiences and contextual mechanisms are underrepresented. Indeed, gathering the perspectives of relevant actors is deemed crucial to better understand how the system operates and how local social, political, and cultural dynamics shape mental wellbeing.17–19 As illustrated by recent research on preadolescents’ online experiences, such qualitative methods can uncover nuanced, lived perspectives, such as context-dependent online challenges that rapidly change and are often overlooked in quantitative studies. 20
Moreover, in current research, system-based analyses have focused on identifying factors, feedback loops, and subsystems, such as the structure of a system. However, ultimately, it is also of interest to understand the behaviour of the system, which refers to how the system acts over time and which dynamic patterns (“systems archetypes”) emerge from the system structure. 21 The aim of this paper is therefore to gain a richer understanding of the system dynamics driving preadolescent mental wellbeing by (a) including the perspectives of diverse stakeholders and (b) analysing the behaviour of the system as a whole.
Methods
Study design
This qualitative study expanded upon previously collected data from the scientific literature and focus groups with scientific experts 16 with the perspectives of preadolescents, caretakers and professionals living (or working) in Amsterdam. Perspectives of preadolescents and their caretakers were gathered through a cocreated conversational tool, whereas professionals’ perspectives were collected via individual semistructured interviews. These perspectives informed the development of a causal loop diagram (CLD), which was subsequently analysed via systems archetypes to identify the underlying system dynamics.
Ethical approval for this study was granted by the VU University Medical Ethical Committee (study protocol 2023.0781). Professionals who were interviewed received an information letter and provided written informed consent before participation. No personal or sensitive data were collected from caretakers or preadolescents; therefore, no informed consent was required from them.
Study setting
This study focused on three neighbourhoods within Amsterdam that prioritized youth mental health within their local policy agendas. Beginning in early 2024, these neighbourhoods committed to a six-year trajectory aimed at progressing from systems understanding to systems change. The present study represents the initial phase of this process. The neighbourhoods – Indische Buurt, Gaasperdam, and Oud-Noord – are characterized by relatively high proportions of residents with lower socioeconomic positions, with populations of 42,611, 32,593, and 32,250, respectively, in 2024.
Data collection and participants
Preadolescents: Conversations using “the tree”
To explore how preadolescents experience and interpret their mental wellbeing, we cocreated a conversational method called The Tree together with youth professionals from a local youth organization (Figure 1). One youth professional took the lead in creating the idea and prototype, and one interactive meeting was held to gather the ideas of six other professionals. All professionals were embedded in the local context, which ensured that this method was easy for both local preadolescents and professionals to use. The Tree is a 180 cm artificial tree featuring four brightly coloured birdhouses representing four key environments: home, school, online, and leisure. For each birdhouse, there was a ‘wellbeing card’ in the same colour as the birdhouse, containing an emotion thermometer and prompt. Preadolescents could fill in this card together with a trusted professional, and place it in the corresponding birdhouse. The trusted professionals (such as youth and child workers who are already familiar with the children) used predetermined prompts to spark brief, playful conversations that focus on the underlying causes of their state of mental wellbeing, as depicted on the card in that specific environment. A picture made during data collection with The Tree.
This concept draws on the tree metaphor previously used in system dynamics literature, which illustrates the interconnectedness between different elements: leaves represent the thoughts and behaviours of preadolescents; branches symbolize the underlying patterns and trends that support these thoughts and behaviours; the trunk represents the structures and drivers; and the roots represent the deeper social paradigms. 22 The Tree was designed as a child-focused, age-appropriate conversational tool. Prior research on interviewing young children shows that visual or projective techniques can help create a low-threshold setting, comfort children by doing something while talking, and thereby support children in sharing their experiences more freely than direct questioning alone. 23
We implemented The Tree in partnership with at least one local organization per neighbourhood – including youth work groups, playground-based child services, teen coaching services, and a chess school – each using the method for at least three weeks during both regular and special activities. The professionals used convenience sampling to include a broad range of preadolescents in the study. All children aged 9-12 years were eligible to participate anonymously, as long as they lived within one of the three participating neighbourhoods. Preadolescents filled in between one and four wellbeing cards. The professionals were encouraged to collect cards across all four birdhouses to ensure that underlying causes were captured for each environment. Before the implementation, professionals received one hour of training on the method. Afterwards, professionals were invited for a 1.5-hour semistructured interview with the research team (EM or NvdP), during which they were asked to share insights from the conversations. The interviewers asked follow-up questions to help distinguish between what professionals heard from preadolescents and their own interpretations.
Caretakers: Conversations using “the tree”
The Tree was also used to explore caretakers’ perspectives. Researchers (EM, NvdP) distributed wellbeing cards among caretakers based on convenience sampling, again featuring an emotion thermometer and prompts for each key environment (home, school, online, leisure) to spark conversations about what influences their child’s mental wellbeing. These researcher-led conversations took place in familiar community settings such as school playgrounds and neighbourhood events. Inclusion criteria for caretakers were that they took care for at least one child between nine and twelve years old and lived within one of the three participating neighbourhoods. The conversations lasted approximately 15-30 minutes, were anonymous and not audio-recorded. Researchers documented key insights in short summaries immediately after the conversations. For these conversations with caretakers, data collection continued until no substantially new underlying causes of mental wellbeing across the four environments came to light. 24
Professionals: Interviews
Professionals working with preadolescents (e.g., in education, health care, or sports) were invited via email for one-hour, semistructured interviews conducted either in person or online (with EM or NvdP). Interviews were audio-recorded. Invitations were distributed through existing professional networks of collaborating colleagues working within the municipality, and snowball sampling was used to further include professionals. Professionals were eligible if they worked with children aged 9-12 years in one of the three participating neighbourhoods. The interviews followed a structured protocol for interview-based methods for mapping mental models 25 and explored professionals’ perspectives on the factors influencing preadolescents’ mental wellbeing. Guiding questions focused on potential feedback loops and perceived underlying mechanisms, using follow-up probes (e.g., “Why is that?”, “What is the consequence…?”) to explore underlying dynamics. All the interviews concluded with the following question: “How is it that, despite all efforts, the mental wellbeing of preadolescents remains relatively low in this neighbourhood?” Data collection continued until thematic saturation was reached within each neighbourhood, meaning that no new issues or insights emerged. 24
Data analysis
Qualitative analysis
Data from preadolescents and caretakers were transcribed and coded in MS Excel. The interview data of professionals were transcribed and coded in MAXQDA 2018. First, parts of the interview (i.e., one or multiple sentences) in which professionals spoke about factors or relations impacting preadolescent mental wellbeing were highlighted in MAXQDA. These were also given an associated code corresponding to the causal structure (factor/link/loop). Then, quotation comments were made that reflected the causal structures as described in the highlighted text in more detail (e.g., factor A → + factor B → - factor C). To do so, a codebook including causal structures, which is based on the previously developed science-based CLD, 16 was used to assess whether the factors, links or loops mentioned in the data were already represented. If so, the same terminology was applied (deductive coding). If a new factor, link, or loop was identified, it was inductively added to the codebook and marked as a novel element. One researcher (EM) conducted the initial coding, which was reviewed by a second researcher (NvdP). Disagreements were discussed within the broader research team.
The coding process resulted in a query report per neighbourhood containing quotations, associated codes, and quotation comments containing causal structures in causal-loop diagram notations. This report was used for a brief thematic analysis of the novel elements, guiding the identification of mechanisms (see “System-based analysis: Mechanisms and archetypes”).
Development of a causal loop diagram
The CLD functions as a meta-level framework that synthesizes diverse individual trajectories into a model and thus does not apply uniformly for all preadolescents or their families. A CLD visualizes interconnected factors linked by positive or negative causal relationships. Some relationships involve a time gap between a change in one element and its impact on another. Together, some relationships form reinforcing and/or balancing feedback loops. In the reinforcing loops, changing the value of a variable within the loop in either direction is amplified by the action of the loop. In contrast, in a balancing loop, changing the value of a variable in either direction is counteracted or offset by the action in the loop. This is shown via a ‘delay’ symbol, represented by two short parallel lines. Altogether, the loops help to uncover the underlying mechanisms that shape system behaviour over time.
A first draft CLD was created for the first neighbourhood by expanding the existing science-based CLD 16 with newly identified causal structures (factors, connections, and loops). The mapping followed an iterative process across the three neighbourhoods, whereby each version of the CLD was built on and the previous one was refined. In the final version, duplicate connections representing both direct and indirect connections were removed to improve readability and usability. The CLD was supplemented with a table containing the identified feedback loops. The CLD comprises numerous interconnected feedback loops; the loops identified represent the most salient loops identified by the researchers during the mapping process. The CLD and table were subsequently validated by the research team, which comprised six scientific experts. The process resulted in the creation of a multi-actor CLD map of the system of mental wellbeing in preadolescents, reflecting the system’s structure.
System-based analysis: Mechanisms and archetypes
Systems archetypes and explanations.
As a first step in the system analysis, mechanisms were described. The mechanisms were not only based on the system’s structure as captured in the CLD but also on the query reports of the data derived from the conversations via The Tree and the interviews. These mechanisms typically consisted of a combination of reinforcing and balancing feedback loops and provided a qualitative description of the system behaviour. In a second step, the mechanisms were used to iteratively identify characteristic system behaviours explaining or underlying these mechanisms that aligned with the systems archetypes listed in Table 1. Once a potential archetype was identified, we first agreed upon its storyline and then developed the archetype CLDs. One researcher (EM) first identified the archetypes, which were reviewed by the other researchers of the team in an iterative process of multiple rounds. Where interpretations remained uncertain, we verified them with interviewed professionals or consulted external professionals. This review process was used to determine the appropriateness of the mechanisms to the specific archetypes and the terms used within the archetypes, resulting in a table containing mechanisms and archetypes.
Research team and qualitative rigor
The study team consisted of a PhD student in preadolescent mental wellbeing from a systems perspective, with a background in public administration and psychology (EM), a researcher with a background in sociology (NvdP), a professor with expertise in integrated community-based approaches to health promotion (CR), an associate professor specialised in system dynamics approaches (WW), an associate professor in mental wellbeing (MvS), and a researcher in the public health field with expertise in health behaviour and systems (VB). The combined experience of this multidisciplinary research team enabled us to triangulate our specific areas of expertise, discuss and identify potential blind spots and thereby reduce the likelihood that important dynamics would be missed. EM and NvdP conducted the interviews. Both were female researchers affiliated with the university, had completed training in qualitative research methods, and had prior experience conducting interviews. Interviewees were not known to the interviewers prior to the study, and the interviewers did not share any personal goals or opinions during the interviews.
Given the interpretative nature of the methods and the embedded role of the researchers in the collection and analysis of the data, we used a combination of triangulation, reflexive discussions and iterative analysis to enhance qualitative rigor. Multiple data sources were integrated to inform the CLD, including scientific evidence and the perspectives of preadolescents, caretakers and professionals from three different neighbourhoods. The interviews followed a structured protocol to improve rigor. 25 Cross-neighbourhood comparison of results supported analytical consistency while preserving contextual nuance. Coding, CLD development, and archetype identification were conducted iteratively within the multidisciplinary research team, going back and forth between the collected data and the CLD, allowing interpretations and underlying dynamics to be continuously discussed, challenged, and refined. In addition, targeted member checking with professionals for the archetype analysis was conducted.
Results
Participants
In total, 233 wellbeing cards from preadolescents and 83 wellbeing cards from caretakers were collected. The exact number of participating preadolescents was unclear, as each individual could submit up to four cards- one for each environment -if they chose to do so. Additionally, interviews were conducted with two groups of professionals: eleven professionals who acted as spokespersons for preadolescents they had spoken to using The Tree and 45 professionals who were interviewed on the basis of their own professional experiences. The data were evenly distributed across the three neighbourhoods.
Results from the conversations and interviews
All the factors in the science-based codebook were mentioned by the participants during the conversations and interviews. Multi-actor data collection, however, also revealed several new themes that were not included in the original science-based codebook. These themes included caretakers’ ambitions leading to greater performance pressure, digital influences and their effects on the desire for wealth and success, challenges around setting rules for screen use and caretakers’ digital literacy, lack of emotional rest at home due to housing issues, adultification and exposure to risky environments, cultural norms around privacy and help-seeking, the role of (positive) role models in identity formation, overprotective caretakers and (social norms regarding) outdoor play, caretakers’ distrust in institutions and barriers to the uptake of social services and care, discrimination and social exclusion, language barriers and social cohesion, and school programmes for wellbeing.
System structure: Causal loop diagram
Following the integration of the multi-actor data with the science-based CLD, the final multi-actor CLD (Figure 2) consists of 62 factors, 50 identified feedback loops and eleven subsystems. Table 2 shows the identified feedback loops within each subsystem along with the overarching mechanisms that were identified through them. Mechanisms – or parts of mechanisms – based on the conversations and interviews were added in blue. Figure 2 displays only those connections that form part of a feedback loop or mechanism listed in Table 2; other connections were omitted to reduce visual complexity. See the supplementary files for figures per subsystem, which were produced to increase readability. The multi-actor CLD representing the system dynamics influencing preadolescents’ mental wellbeing. System dynamics within the multi-actor CLD: subsystems, feedback loops, and mechanisms. Note. The subsystems, feedback loops and associated descriptions that are newly based on the multi-actor perspective are presented in bold. The rest comes from the earlier published science-based CLD. Loops are either reinforcing (R) or balancing (B).
Together, the identified mechanisms illustrate how broad societal forces, such as individualism and digitalization, intersect with daily life to shape preadolescents’ wellbeing. They capture, for instance, the tension between universally high academic expectations and the very uneven home and school conditions that enable success; the pervasive pull of screen-based activities in a 24/7 connected world; the degree to which safe, welcoming neighborhoods foster or inhibit outdoor play and social cohesion; and families’ struggles to navigate complex social‐service systems when under financial stress.
System function: Archetypes
System-based analysis using archetypes.
For example, the analysis of the system’s structure (Table 2) revealed a mechanism around caretakers requiring sufficient mental headspace to access social services and preventive care. The archetypal pattern explaining why mental headspace is a prerequisite for seeking services is a “Fixes that Fail” for vulnerable families. That is, numerous short-term plans and regulations (‘fixes’) aim to improve livelihood security for vulnerable families, but their cumulative effect creates long-term complexity and reduced visibility. Consequently, these well-intentioned support measures can backfire: the resulting complexity discourages uptake, leaving vulnerable families without sufficient headspace unable to access aid and ultimately experiencing greater livelihood insecurity.
Discussion
This study aimed to gain a richer understanding of the system dynamics underlying preadolescent mental wellbeing. Unlike earlier positivist approaches that focused narrowly on a single relationship or determinant, this study did not provide an exhaustive account of detailed evidence, but rather aimed to explore both the interrelations between factors and feedback loops (the system’s structure) and the patterns that underlie these interactions (the system’s behaviour) at the population level.
System structure: Multi-actor CLD
The multi-actor CLD included 62 factors, 50 feedback loops, and 11 subsystems. The multi-actor perspective adds two subsystems, seven feedback loops, and nine factors to the original science-based CLD. By integrating the perspective of stakeholders with the scientific perspective, the CLD captures both proximal and distal factors and their interconnections, which were then used to identify larger mechanisms. Many of these new mechanisms illustrate indirect influences, such as caretakers not finding their way to preventive services and care or providing in-depth knowledge of a specific theme, such as adultification, which interferes with preadolescent identity development.
Moreover, the multi-actor perspective highlights that while the system’s structure captures dynamics across a broad range of themes, the specific expression of each mechanism depends on the social and environmental context. For example, neighbourhood quality and safety influence children’s outdoor play in different ways in different contexts: in one area through traffic or distant playgrounds or in another through (drug-related) crime. While the outcome – children spending too much time indoors – remains the same, the specific underlying mechanisms vary. Similarly, mechanisms differ across cultures. For example, in some collectivist cultures, mental health issues are seen as private family matters, with the home considered a closed space and discussing personal or emotional problems with outsiders viewed as inappropriate. These examples illustrate that while system-level mechanisms are broadly applicable, understanding the nuances of neighbourhood contexts and cultural values is essential for accurate interpretation. This interpretative nature is particularly important to consider when discussing the system’s structure with stakeholders.
System’s behaviour: Archetypes
Systems archetypes, although introduced by Senge in 1990, have been used mainly in empirical contexts such as biodiversity and agriculture.27–29 This study is among the first to use them to understand a complex social issue. Among the multitude of interconnected factors, loops, and mechanisms, seven archetypes were identified that together provide an understanding of the system’s behaviour.
The seven archetypes predominantly reflected macrolevel social or political phenomena. Interestingly, one archetype (‘Fixes that fail: Caretakers as role models’) was more at the micro level, describing how caretakers' modelling behaviour is key when setting rules for preadolescents’ smartphone usage. When thinking about actions that could disrupt the patterns described there, another archetype (‘Shifting the burden: Societal norms underlying preadolescent smartphone usage’) comes to mind. That is, actions targeting the role modelling behaviour of caretakers place responsibility at the individual level instead of at the societal or business level. Ultimately, combining these insights when developing actions is important. Actions focused on the societal or political level, such as shifting social norms regarding the age at which children receive their first smartphone or regulating the addictive algorithms used by tech companies, may have the greatest impact, but could be supplemented with subactions focused on the individual level, such as stimulating positive role model behaviour.
These identified patterns are supported by insights from previous literature. The first pattern (‘Unequal opportunities in performing well academically’) is consistent with research showing that socioeconomic status is positively associated with cognitive performance 30 and that small initial differences can amplify over time through cumulative advantage mechanisms. 31 For the second pattern (‘Societal norms underlying preadolescent smartphone usage’), research has shown that adolescent technology use is largely driven by social pressure, 32 indicating the need for coordinated societal efforts to establish new norms. 33 Third, previous research highlights the importance of parental modelling and suggests targeting both children’s and caretakers’ behaviour in interventions, 34 in line with the pattern of ‘Caretakers as role models’. The fourth pattern (‘Availability of playgrounds and sports facilities’) can be linked to research demonstrating that budget cuts in the Netherlands have led many municipal governments to scale down their investments in sport facilities, despite researchers calling for increasing the attractiveness of sport facilities. 35 With respect to the fifth pattern (‘Informal support’), Dutch youth policy tends to treat informal networks as add-on tasks rather than essential foundations of care, 36 despite evidence that such networks enhance mental wellbeing. 37 The sixth pattern (‘Fragmented support for poverty’) is consistent with Dutch studies reporting that fragmented rules, opaque eligibility, and short-term interventions hinder families’ access to effective support.38–40 Finally, policy reports have highlighted that schools with strong leadership, parental involvement, and stable staff are more likely to implement sustainable interventions, 41 reflecting the seventh pattern ‘Schools, goals and limited capacity’.
System structure versus behaviour: Findings and implications
Overall, the multi-actor CLD mapped the system’s structures that shape preadolescent mental wellbeing. It provides insight into the relationships between multiple factors, highlighting not only surface-level symptoms but also “causes behind the causes.” The systems archetype analysis then provided an extra layer of insight into the underlying patterns sustaining a certain mechanism, e.g., the fragmented landscape of rules and schemes for livelihood security support, explaining why caretakers need mental headspace to gain access.
This analysis of the system’s structure and behaviour reveals several insights. First, the analysis identified many distinct but interconnected layers of dynamics, ranging from immediate pressures in school or family life to broader societal norms and institutional structures. These insights reveal how there is no easy explanation or quick fix for this complex public health issue.
Second, a number of societal norms or paradigms that were embedded within system dynamics were identified. These include the belief that success is self-made and must be rewarded; the perception that children from higher socioeconomic backgrounds are more ‘talented’; the expectation of being constantly online and reachable; the emphasis on individual responsibility in dealing with technological challenges; and the persistent stigma around discussing mental health issues.
These systems understanding insights can inform meaningful action, empowering policymakers, educators, and communities to disrupt patterns (the identified archetypes) that negatively impact preadolescent mental wellbeing. That is, viewing complex public health issues through a systems lens can encourage practitioners and policymakers to (1) adopt a systems thinking perspective that moves beyond isolated symptoms to consider the underlying structures, goals, and belief systems sustaining these challenges 42 and (2) engage in adaptive action, embracing real-time learning and iterative responses that address deeper structural drivers. 43
Strengths, limitations, and future directions
A strength of this study lies in its novel approach to analyse both the system’s structure and behaviour, showing the interconnected layers of dynamics underlying preadolescent mental wellbeing, as well as the archetypes sustaining the system. Moreover, the use of a science-based CLD ensures both theoretical consistency and contextual relevance. In addition, the large volume of qualitative data, combined with team-based analysis, enhanced the credibility of the findings.
This study has several limitations. The methods used, such as The Tree, were cocreated with stakeholders and remain partially experimental. While we believe that this participatory approach enables high levels of engagement and inclusivity, it also introduces elements whose validity has yet to be thoroughly established. For example, to understand the perspective of preadolescents, we relied on trusted professionals’ interpretations. Although this ensured that these young teenagers spoke about potentially sensitive topics with a trusted professional, this added an extra layer of interpretation. Nevertheless, the consistency across neighbourhoods, shared perspectives among different respondent groups and alignment with existing scientific literature underscore the robustness of the findings. Moreover, the evolving nature of the codebook during analysis may have introduced an interpretive bias, although this bias was mitigated through interneighbourhood checks, systematic reflection and regular team discussions. A final limitation relates to the systems (archetype) analysis. These analyses are subject to the researchers’ interpretations and could therefore have been unintentionally influenced by the researchers (e.g., by their personal beliefs, presumptions, etc.). 44 We mitigated this through iterative, reflexive team discussions and, in cases of doubt, targeted member checking with interviewed or external professionals. Future systems research should continue recognizing these archetypal patterns to mitigate these risks of bias.
In the context of the current study, future research should focus on translating systems understanding into systems change. One promising avenue is to investigate whether disrupting the identified systems archetypes can lead to improvements in preadolescent mental wellbeing. In addition, further methodological advancements are needed to refine and validate participatory tools such as The Tree to investigate how these tools can be used to better understand system dynamics.
Conclusion
This study provides an understanding of the system dynamics underlying preadolescent mental wellbeing in Amsterdam. By integrating multiple perspectives and analysing the system as a whole, through developing a multi-actor CLD and conducting a systems archetype analysis, it becomes clear that the dynamics underlying preadolescent wellbeing are wide-ranging in terms of themes, spanning a broad range without being limited to a specific domain. In many dynamics, social norms revolving around, for instance, individualism or the 24/7 online world play a central role. Together, the findings highlight a critical need to shift from reactive, short-term solutions toward more cross-domain, preventive, structural strategies. These systems archetypes may serve as powerful tools to catalyse change. By viewing complex challenges through a ‘systems (archetype) lens’, policymakers, educators, and communities can better understand the bigger picture and work together to create more resilient and inclusive environments that support the mental wellbeing of preadolescents.
Supplemental material
Supplemental material - System dynamics of preadolescent mental wellbeing: A multi-actor perspective in Amsterdam using system archetypes
Supplemental material for System dynamics of preadolescent mental wellbeing: A multi-actor perspective in Amsterdam using systems archetypes by Eline M. Meuleman, Vincent Busch, Wilma E. Waterlander, Nanda E. van der Poel, Carry M. Renders, and Maartje M. van Stralen in Journal of Public Health Research.
Footnotes
Acknowledgements
We would like to acknowledge the preadolescents, caretakers and professionals who participated in this study.
Ethical considerations
This study received approval from the VU University Medical Ethical Committee (study protocol 2023.0781).
Consent to participate
No identifying participant information was collected for the purpose of this study, and informed consent was obtained for all (adult) scientific experts before study participation.
Author contributions
Conceptualization (EM, MvS, WW, CR, VB), investigation (EM, NvdP), supervision (MvS, VB), funding acquisition (MvS, VB), visualization (EM, NvdP), writing – original draft (EM), writing – review & editing (EM, MvS, WW, NvdP, CR)
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by a grant from The Netherlands Organisation for Health Research and Development (ZonMw). The funding agency had no role in the design of the study; in the collection, analysis, and interpretation of data; or in the writing of the manuscript. Grant number: 05550032110027.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
