Adapting the Social Exposome Model to Analyze Causal Mechanisms in Loneliness Epidemics: An Interdisciplinary Review

1. Introduction

Since 2023 the World Health Organization has classified loneliness as a global health threat, with over fifty countries and local governments creating policies to address loneliness epidemics [1,2,3]. A history of treating loneliness with therapy techniques or small group interventions has produced significant but minor effect sizes, that calls are being made to shift to systems-level and ecological interventions in public health to address loneliness epidemics [4,5,6,7]. Underlying this shifting approach is classifying social connection as a human need equivalent to water, food, and shelter [8]. Social connection captures the biological dependence of people to live in large groups, believed to be a critical part of our evolution [4,8,9,10]. Similar to public health strategies for basic needs, the safety, quantity, and quality of social connection can be assessed for detecting deprivation in communities, with loneliness signaling an overarching problem with quality in social networks [8,11]. International efforts are underway to collect evidence on how public health can address the quality of social connection in communities [12].

The evidence, however, is primarily coming from various branches of psychology and clinical sciences that inform individual-level and small group interventions [12,13,14]. The knowledge base leans toward reductionist and specialized approaches that are unique to the individuals receiving care [15]. In trying to scale this evidence to prevent loneliness in populations, a loneliness researcher has stated, “It is virtually impossible to identify the broad set of determinants of loneliness: the relationships are mostly of a reciprocal nature and mechanisms behind interrelationships are difficult to disentangle” [16] (p. 75). Recent publications trying to compile evidence about loneliness from multidisciplinary teams of experts have produced long lists of knowledge gaps because they are trying to understand the relationship of any aspect of human society to the loneliness of its residents [8,17].

In contrast, public health has relied on ecology and environmental health as systems sciences to address basic needs at national and global levels to impact population health outcomes [18,19]. Systems sciences are capable of understanding regional and global problems because they identify a limited number of factors and relationships that are most critical to explaining an outcome [20,21]. Three frameworks were identified in public health that already integrate social and environmental knowledge for system-level interventions that can address social connection as a human need: the social determinants of health framework, the social-ecological model, and the social exposome model [11].

The social determinants of health framework integrates knowledge about human society and social life with environmental knowledge about access to basic needs and adverse environmental conditions, to explain unfair differences in population health outcomes [19,22]. The social-ecological model integrates psychology and ecology knowledge to organize society into different levels that explain how larger social processes are connected to patterns in social experiences that affect developmental and health outcomes [23]. The social exposome model was developed in environmental health and social epidemiology to understand causal pathways from well-known social problems, such as socio-economic inequality, to health outcomes, by measuring total and lifetime social exposures [24].

Studies that integrated and applied these models to understand how social environments can produce loneliness across populations, found a limited number of patterns in social interactions that explain why adults feel lonely [25,26,27]. Patterns, such as being rejected and losing relationships, were not new to loneliness research but have consistently been documented across a diversity of populations [17,28]. Larger social processes that underly and influence these patterns have also been examined for several decades in public health, such as the process of marginalization [22]. It raised concerns that time and funding could be wasted on disciplines discovering concepts and theory from other disciplines, because we assume that loneliness is new to public health.

Instead, it seems that public health already has knowledge that is critical for addressing loneliness epidemics but may need to undergo a massive re-interpretation of treating our social lives as a basic need. While systematic literature reviews are rigorous approaches to determine the level of evidence on important health issues, they do not support knowledge integration that is needed to understand the complexity of interdisciplinary topics like population loneliness [29,30]. To limit author bias, systematic literature reviews rely on pre-defined search strategies using specific terms and definitions that are unique to disciplines, producing disciplinary bias that is already a problem in loneliness research [13,14,29].

What is currently needed are interdisciplinary methodologies that are designed to collect and integrate knowledge across multiple disciplines so that complex population health topics can be understood as a cohesive whole [29,31]. Interdisciplinary reviews are designed to be flexible in language, identifying conceptual common ground, or knowledge that is already agreed upon and shared across disciplines even if the same words are not being used [29]. Results are limited to evidence-based knowledge from various disciplines, but focus on the rigor of interpreting and integrating the knowledge rather than the level of evidence. Results help health researchers collaborate on testing interdisciplinary theory that includes their respective disciplines but explains how they fit into a larger understanding about a health topic [31].

To improve guidance in public health on designing interventions for population loneliness, the goal of this interdisciplinary review is to identify common ground about causal mechanisms capable of producing epidemic-levels of loneliness. Causal mechanisms are defined by ecology and critical realism as explaining the structures or processes that tend to produce an outcome [32,33], helping to identify points of intervention to change it. This review examines common ground between loneliness theory produced in sociology, psychology, and neuroscience, as well as three public health frameworks or models on human social life, including the social determinants of health framework, the social-ecological model, and the social exposome model. Given the focus on causal mechanisms, results are used to update the social exposome model with the intention of informing social environmental approaches in public health that address social connection as a basic need.

2. Materials and Methods

This interdisciplinary review followed five phases that fulfill the initial goals of knowledge integration before conducting inter- or transdisciplinary research: 1) clarify the research question and its scope, 2) develop a search strategy to identify key literature in relevant disciplines that best answer the research question, 3) deeply read literature to understand the context, or disciplinary bias, in which this knowledge was produced, 4) identify gaps in knowledge that can be filled with new disciplines and literature until the research question has been sufficiently answered, 5) compare the interpretation of knowledge for common ground and integrate it into a cohesive whole [29,31].

These five phases are similar to the five phases common in scoping reviews, except they do not pre-define criteria for selecting studies and literature [34]. Instead, general guidelines are initially given that interdisciplinary reviews should identify more than one relevant discipline that can fully answer a research question, identify the work of authoritative experts about a topic, and identify knowledge that is generally accepted within a given discipline [29]. Developing a rigorous search strategy is an iterative, or cyclical, process that may be unique to the research question. In this way, interdisciplinary reviews have post-defined criteria that require justification.

The following details how these five phases were implemented in this review.

2.1. Research Question

This interdisciplinary review wants to understand how to holistically analyze the causal mechanisms that produce loneliness epidemics to inform system-level interventions in public health.

2.2. Search Strategy

Only recently has loneliness research expanded across health-related disciplines to respond to national awareness campaigns on loneliness and to understand the impact of social distancing policies during the COVID-19 pandemic [1,35]. Prior to this, loneliness research has primarily been funded to understand the social experiences of aging populations or to address loneliness when it coincided with mental health outcomes [8,36,37]. Historically, loneliness has been a relatively small area of research with few leading experts [17,38].

For this reason, the search strategy of this review focuses on identifying the work of authoritative experts in loneliness research to understand their theories on loneliness and underlying causes.

As the United Kingdom was the first country to develop a government loneliness strategy in 2018 and the United States of America was the first country to declare a loneliness epidemic in 2023, their government reports were skimmed to identify theory on loneliness and the experts who produced them [8,37]. To have a more international perspective, the World Health Organization report on loneliness was also skimmed to identify experts [3]. In reading the body of work of loneliness experts, more experts were identified due to frequent collaboration or discussion about competing theories. Table 1 describes final criteria for identifying authoritative experts on loneliness.

A complete list of each expert’s publications was found through the online search engine Google Scholar and the science networking platform ResearchGate. These platforms were used instead of scientific reference databases, such as PubMed, because full explanations of loneliness theory are not often found in word-limited journal articles that focus on individual research studies. From Google Scholar and ResearchGate, expert publication lists included books, book chapters, professional reports, and research articles. Publication lists of each expert were scanned for addressing loneliness in their titles or summaries. They were already organized by how frequently they were cited in other literature. Final criteria for how literature was selected is described in Table 1.

On the public health side, literature was collected on the social determinants of health framework, the social-ecological model, and the social exposome model. Criteria for how this literature was selected is explained in Table 1. Textbooks, recent reviews, and government agency reports were prioritized to meet all criteria.

2.3. Extracting Knowledge

An initial selection of literature from each loneliness expert was skimmed for relevant knowledge to answer the research question. If relevant, literature was deeply read to summarize knowledge about causal mechanisms in loneliness theory and re-read to compare similarities and differences across literature [29]. Knowledge was summarized in a table, including the citation, disciplinary context, type of literature, and key contributions to results [29]. Once literature from a loneliness expert no longer provided substantial and new information about their loneliness theory, no more literature was included by that expert.

Similarly, literature on the three frameworks or models in public health was deeply read to summarize knowledge about causal mechanisms that connected human society and human social life to population health outcomes. Knowledge was summarized in a table, including the citation, the discipline for which the literature was produced, the type of literature, and a summary of its contribution to explaining causal mechanisms.

2.4. Comparing Knowledge and Filling Knowledge Gaps

When comparing knowledge between loneliness theory and public health theory on causal mechanisms that connect human social life to population health outcomes, knowledge gaps were identified. Instead of accepting these knowledge gaps, an important difference in interdisciplinary reviews is that additional disciplines and theory are identified that can fill these knowledge gaps until the research question has been answered [29]. It contributes to the rigor of explaining a health topic in full, extending beyond the bounds and bias of a specific set of disciplines. Literature from additional disciplines had to meet the same criteria as literature included on public health frameworks and models (Table 1).

Results were shared in a working paper and presentations to an interdisciplinary group of 20 PhD-level experts at the U.S. Department of Veterans Affairs. The group included a variety of medical providers, public health experts, and social scientists who are experts in gerontology, neuroscience, biology and clinical science, medical anthropology, psychology, community and behavioral health, and health services research. The group provided feedback on additional literature to consider, disciplinary context, and clarity of how results are presented. The study author also completed a fellowship for Applied Critical Realism in an academic year of 2025 – 2026. The study author has undergraduate-level training in ecology research and PhD-level training in public health research.

2.5. Final Figure and Explanation

First, common ground across loneliness research and public health frameworks were identified based on overlapping theory about causal mechanisms connecting human social life to population health outcomes. Second, knowledge was integrated into an adapted and unified social exposome model by combining points of common ground into a cohesive whole. A final summary of this knowledge is provided in the results.

Before being integrated, evidence-based knowledge adheres to disciplinary interpretations and results can be assessed for how accurately they reflect the interpretations of that discipline [29] In the results section, it will explain the contributions of each authoritative loneliness expert and each public health model or framework before integrating their knowledge.

Knowledge was integrated into a final figure to fully describe how population loneliness is theorized to occur based on literature included in the review. These results are still based on evidence-based knowledge, but knowledge integration between disciplines can change their original interpretation [29]. However, this review aimed to maintain the original interpretation of loneliness experts, documenting any time interpretations were changed and the logic for doing so. The final model is intended to be a communication tool to improve collaboration across disciplines by identifying the knowledge they share in common, to test and improve in future research studies [29].

Results can be evaluated based on explanatory power. Explanatory power describes the ability of theory to more fully explain existing evidence on a topic to the point that it is an improved explanation over existing theory [39]. For this review, explanatory power is judged on the ability of the final model to answer the original research question and the rigor of how knowledge was integrated from literature to develop results into a cohesive whole [29]. Results are intended to produce plausible explanations in a real world, minimizing changes to how knowledge is integrated from each discipline. It is not intended to be exhaustive, only focusing on points of common ground and including literature until a sufficient explanation can be produced [29].

3. Results

The body of research from thirteen authoritative experts on loneliness was included in this review (Table 2). When combined, these experts have conducted loneliness research from 1955 to present, within the United States of America, the United Kingdom, or the Netherlands. Many have worked together and with consistent collaborators. Each expert more accurately represents a clustering of several researchers who are not individually named in the results. They also represent broader global trends in health research priorities and advancements. Experts are primarily trained in the disciplines of sociology, psychology, and neuroscience, often integrating these disciplines, such as social psychology or social neuroscience.

A total of 85 books, book chapters, textbooks, review articles, research articles and government agency reports were included in this review (Supplementary Table S.1). Reading of literature occurred over several years. Literature primarily comes from the disciplines of sociology, psychology, social neuroscience, public health (general, social epidemiology, and environmental health), ecology, evolutionary biology, and philosophy of science. Table 3 summarizes information about causal mechanisms that are included in relevant public health frameworks or models, and the additional disciplines that were needed to fill in knowledge gaps before integrating knowledge together.

3.1. Knowledge Gaps

Three main knowledge gaps were found between loneliness theory and public health theory regarding causal mechanisms. One gap was the missingness from loneliness research in understanding social exposures and how they are influenced by underlying mechanisms, despite loneliness experts recognizing the critical role of social interactions in loneliness outcomes (Table 2). There seems to be a lack of awareness that it’s possible to analyze social interaction patterns and exposures that are critical to health and survival outcomes. The social-ecological model, the social exposome model, and ecology were useful for framing loneliness research to fill this gap.

A second gap is the common reference that loneliness researchers make about evolutionary theory that is completely missing from the three public health frameworks used in the review. References about evolutionary theory were added to fill this gap.

The third gap was how to make sense of the three different ways that sociology, psychology, and neuroscience define loneliness (See Section 3.5). These definitions are distinct and largely non-competing, meaning they don’t overlap. Critical realism and the three public health frameworks helped to fill this gap by recognizing separate layers of reality, helping to integrate these definitions into one to better reflect the complexity of loneliness.

3.2. Adding Critical Realism, Ecology & Evolutionary Theory

Critical realism is a philosophy of science that was incorporated into this review to help organize knowledge from various disciplines into a cohesive whole by acknowledging a complex and layered reality [33]. It is the only philosophy of science that can integrate natural and social science knowledge by organizing reality into three different layers: 1) experience (subjectivity), 2) events that can be observed (objectivity & outcomes), 3) underlying causal mechanisms that are not always visible and tend to produce events [39]. These layers can describe our external world (e.g. our environment) as well as our internal world (e.g. our bodies).

Critical realism underpins the final model. The model clarifies each layer of reality and how it’s being interpreted. Loneliness research that helps explain, for example, the subjective experience of loneliness, has its place in the model, while loneliness research that explains the biology of loneliness also has its place in the same model. These are not contradictory or competing theories, they are complementary once their role is understood within a complex world.

The final model reflects not only the social sciences, but also ecology. It is not like an ecological model; it is both an ecological and social model. In this way it overlaps a great deal with the social-ecological model in public health. However, it re-emphasizes ecology knowledge that may be unfamiliar to non-experts to better understand the direct social exposures that loneliness researchers associated with loneliness outcomes. Ecology, environmental health, social epidemiology, and social psychology all share an understanding that social interactions in people’s environments are a key causal pathway for how human health is impacted by our social lives (Appendix Table A1). They do not define interaction differently – their existing definitions conceptually overlap. What these disciplines are saying is that the interactions that people experience are objective, as they can be observed in the environment and measured in ways that are known to be critical to survival.

Ecology and evolutionary theory are heavily intertwined to explain networks of evolutionary relationships, the impact of interactions among living beings on survival outcomes, and the social behaviors of animals (Table 3). However, ecology has not included knowledge on human social life. In the reverse, public health frameworks do not include evolutionary theory, missing some key ideas about the role of evolutionary relationships in defining social connection across all species (Table 3). To fill this gap, evolutionary theory was incorporated into the model to explain current knowledge and how it can support our understanding of social connection as a basic need.

3.3. Addressing Conflicting Knowledge

Knowledge integrated into the final model largely retained its original interpretation from its respective disciplines. However, there were two instances in which conflict was identified between theory and one interpretation was selected over the other.

There is conflict in how psychologists make sense of information about social interactions compared to environmental and public health disciplines. When lonely individuals describe the interactions they have with other people, in psychology these interactions have been defined as subjective feelings about the universal experience of loneliness [49,51]. It conveys that interactions are an internal experience rather than objective events happening in the environment around us. Psychologists and social demographers have been testing these universal patterns in interactions in the very loneliness scales they have designed, including the UCLA Loneliness Scale and the De Jong Gierveld Scale [49,51] (Appendix Table A.1).

The final model selected the interpretation from environmental, public health, social psychology, and critical realism, that interactions are objective events happening in our environments (Appendix Table A.1).

The second conflict was between two different psychology interpretations about the relationship between an individual and society. Psychologists defining existential loneliness describe the mental mapping of an individual’s role in society that can reflect on their value [63]. However, this relationship is interpreted by one loneliness expert as a relationship with oneself and how an individual determines their own value or worth [63]. A second loneliness expert in cultural psychology, however, interprets this relationship as real in how society determines the value and worth of an individual within the context of cultural norms, values and beliefs [83].

Since the final model of this review integrates public health models with loneliness research, it selected for the interpretation in cultural psychology that existential loneliness reflects on an environmental relationship between an individual and the society in which they live.

3.4. Final Model and Causal Mechanisms

Based on conceptual common ground, Figure 1 was developed to integrate knowledge into a cohesive whole. Table 4 documents the reasoning for terms and causal mechanisms captured in Figure 1. To reiterate, causal mechanisms are being defined by ecology and critical realism to describe the structures or processes that explain how an outcome, specifically loneliness, tends to be produced (i.e. how does something work to produce an outcome) [32,33]. Critical realism describes how causal mechanisms apply in social systems [33].

This is not the same thing as causal effects [32]. Causal effects describe study designs and statistical approaches to determine if a caused b, or if an intervention caused an outcome. It does not describe, for example, how an intervention works.

A total of eight causal mechanisms were consistently identified in this interdisciplinary review to explain how mass loneliness can occur in populations: marginalization, intersectionality, social instability, agency to shape culture, non-conscious processes in social homeostasis, cognitive processes, aging and health, and historical evolution as a social species (Table 4). Each mechanism represents a larger underlying process that can be made up of multiple causal mechanisms that serve the same function of influencing loneliness outcomes. For example, evolution can include multiple causal mechanisms that contributed to the historical process of human evolution and our generational dependence on social connection (Table 3). Combined, these eight causal mechanisms interact as a system to produce observable changes in population-level loneliness.

Please see definitions in Table 4. Eight causal mechanisms are bolded that include intersectionality, social instability, marginalization, agency to shape culture, non-conscious processes of social homeostasis, cognitive processes, aging & health, and evolution.

3.5. Defining Loneliness

Authoritative loneliness experts consistently recognized three different definitions of loneliness that were developed from three different disciplines: social needs theory from sociology, cognitive discrepancy model from psychology, and evolutionary theory of loneliness from neuroscience.

Social Needs Theory: Loneliness is a response to the absence of an attachment figure or a response to the absence of a place in an accepting community, creating a deficit in social provisions [40,41]

Cognitive Discrepancy Model: Loneliness is a subjective feeling that occurs when there is a discrepancy between our social needs and desires and the quality of our actual relationships [8,44]

Evolutionary Theory of Loneliness: Loneliness is a biological warning system in social species that signals when salutary social relations are endangered or damaged [61]

These three definitions were all integrated into the final model as they describe a different aspect of loneliness and are non-competing. To better define the complexity of loneliness, results were developed into a unified, interdisciplinary definition of loneliness.

Interdisciplinary Definition of Loneliness: Loneliness is an emotional and protective state of distress, informed by both cognitive and non-conscious processes to make sense of the social world in which someone lives, signaling exposure to different forms of marginalization from society when failing to meet cultural expectations and motivating ways to find belonging.

4. Discussion

Loneliness research does not just explain loneliness as an outcome or experience but sheds light on how to understand social connection as a basic need. The most important word from this review is interaction. Various disciplines in this review identify interaction as the unit of measure or building block of social connection (Table 2 & 3). Despite this understanding, loneliness experts have captured interactions as subjective feelings in validated loneliness scales (Appendix Table A.1).

While it may seem like a stretch to incorporate evolutionary theory into this review, both public health and loneliness experts seem unaware of advances in evolutionary theory that describe interaction and networks as critical ways to understand evolutionary relationships across all species [94,101,102]. Interaction and networks can occur from the cellular level all the way up to interactions and networks among living beings and their larger environment [101]. While it may be practically useful to distinguish between interaction versus social interaction to describe human relationships, this distinction may fail to understand the diversity in how interaction and networks have built up across nature to exist as social connection in people today. Evolutionary theory and social connection as a basic need seem to be inseparable.

If this is the case, why are interaction patterns, like rejection, being classified as subjective in loneliness research and why is loneliness commonly defined as a purely subjective experience [8,49,51]? This review identified that early loneliness research initially struggled to understand the difference between loneliness and social isolation [40]. Loneliness researchers recognize social isolation, or a quantitative lack of relationships and contact with other people, to be objective [8]. Since social isolation and loneliness are now known to be unrelated and loneliness is an emotion, experts separate these concepts by objectivity versus subjectivity [8,15,44]. While this distinction remains important, it may be undermining the complexity of what loneliness is.

For example, health researchers also know that loneliness is critical to survival outcomes [4,8,64]. It is hard to make sense of how a purely subjective experience can result in people dying. It comes across like entire populations are feeling their own way into death. Critical realism became necessary in this review to help organize knowledge from loneliness research that could make sense of its complexity, including both subjectivity and objectivity. Critical realism is a philosophy of science that bridges natural and social sciences by acknowledging three layers of reality: subjective experiences, objective events in the environment around us, and underlying causal mechanisms that tend to produce these objective events [33,98]. So, if lonely people around the world are observing rejection in their interactions, as described by validated loneliness scales, critical realism provides guidance on how to consider their flawed perceptions, while still recognizing that interactions are objectively occurring in their environment in which rejection is possible.

Currently, validated loneliness scales are claiming that interaction patterns like rejection, along with loneliness, are a purely subjective experience [46,49,51,62]. Since loneliness researchers do not make the case that reality cannot be determined and everything is subjective, what seems to be happening is that we have not figured out an objective way to measure patterns in interactions, like rejection. If our measures of rejection are considered subjective, such as listening to people’s stories about it, loneliness researchers have ended up classifying patterns in interactions as a part of the overall subjective experience of loneliness. As scientists we must ask ourselves if the way we measure the world should dictate reality or should reality dictate how we try to measure it, even if we fall short. Critical realism was used in this review to think through reality in loneliness research regardless of how we are currently measuring it.

While these points may sound theoretical and abstract, they make a substantial difference in public health practice. Results in this review identified that fifty years of loneliness research consistently described eight main mechanisms, or underlying causes, that can interact to explain loneliness outcomes. In comparing these mechanisms with public health knowledge, four of them fall within existing public health efforts to improve population health. They include intersectionality, social instability, marginalization, and agency to shape culture, that are often described within the social determinants of health [22]. However, trying to address these concerns in public health is controversial [103].

Part of the controversy that overlaps with loneliness research is how to address social problems that are experienced subjectively and shaped by perception. How do we determine social problems that warrant public health intervention and objective improvements in results to know that an intervention was effective? Public health has largely relied on identifying demographic risk factors for populations whose overall health is worse than their peers, known as health disparities, and classifying these populations as vulnerable [104]. For a social problem, like discrimination, decades of research have also uncovered pathways that link various forms of structural discrimination and adverse social experiences to population health disparities [105]. Public health uses approaches like Health In All Policies to consider structural changes that address large-scale social problems impacting a wide range of population health outcomes [106].

However, this approach can backfire by creating an identity of vulnerability for at-risk populations based primarily on demographic data, rather than the local context producing disparities [104]. This approach can also give the impression that some populations face marginalization, but not others, failing to explain how or why it is occurring at this time, in this place.

Loneliness research improves this conversation by considering the role of marginalization in an evolutionary context and ways we can better assess social environments to understand local problems. We can know that marginalization is a real phenomenon critical to survival because it is critical to the survival of many species [59,60]. However, every human being will likely experience the forces of marginalization in their lifetime, indicated by the universal feeling of loneliness. As described by John T. Cacioppo in social neuroscience, marginalization is so important to survival that we have developed a defense mechanism against it [60].

Instead of only relying on demographic information, public health can start to examine social exposures in local places and corroborate this information with social structures, such as local laws and policies. For example, even if we can only identify social exposures like rejection through people’s stories, we can compare this information to concrete data, such as recorded communication, written laws and policies, or documented events. The importance of evaluating social environments by the kinds of interactions occurring is that loneliness research has identified multiple forms of marginalization that may require different solutions. Within this review, five patterns of interactions were linked to loneliness, including rejection, loss of relationships, lack of reciprocity in excessive labor demands, a mismatch in having nothing in common, and an inability to socially participate often due to health problems (Appendix Table A.1).

These different kinds of problems will likely require different solutions. For example, if severe enough forms of marginalization are occurring in a place to impact population health outcomes, loneliness research suggests it could be due to having nothing in common, such as having different cultures that don’t understand each other and disengaging from interaction. A public health approach to this may be developing a bridge to build common ground among different cultures in the same area.

The Adapted Social Exposome Model on Loneliness in this review is not intended to be a definitive guide. Instead, it supports collaboration across disciplines studying loneliness to critique and improve this knowledge in a coherent way. Public health can also use this model to better evaluate and design interventions for their jurisdiction. Last, results provide an interdisciplinary definition of loneliness to clarify its complexity so that experts and novices alike can better understand the breadth of knowledge we currently have about loneliness.

Limitations

Currently, there are no standardized guidelines on how to conduct an interdisciplinary review. This review relied on inter- and trans-disciplinary methodologists who outline a process for integrating knowledge across multiple disciplines before collaborating on research [29,31]. Interdisciplinary reviews share a lot in common with both scoping reviews and narrative reviews [34,107]. These are not just different methods; each of them is underpinned by a different philosophy of science. Scoping reviews are shaped by positivism with concerns about following a systematic approach that can be replicated and limits author bias [34]. Narrative reviews are shaped by social constructivism and interpretivism that do not define reality, such as evidence, instead allowing for multiple perspectives, including non-experts, to synthesize their collective knowledge about a topic [107].

Methodologists have not yet explained the philosophy they are following to conduct an interdisciplinary review, as it requires including disciplines with different philosophies to address disciplinary bias [29,31]. However, this review is underpinned by critical realism. It is the only philosophy of science that bridges a divide between natural and social sciences, while still caring about reality and causation [33]. The rigor is not in systematic replication, but in the logic of integrating knowledge into a cohesive whole that describes overlapping evidence from multiple disciplines, or common ground. This approach defines grand challenges in the sciences to explain complex topics [29].

Regardless, this review is vulnerable to author bias [29]. To limit author bias, results relied on literature from authoritative loneliness experts that have been recognized by national governments and an international public health agency to incorporate their theories on loneliness. Twenty PhD-level experts across medical, basic science, public health, and social science disciplines also provided feedback on results. Only twice did the review intentionally select one theoretical perspective over another when they conflicted in the final model. Results detail why this selection occurred so that future researchers can reconsider these decisions.

5. Conclusions

An interdisciplinary review identified eight causal mechanisms consistently discussed by authoritative loneliness experts as contributing to population-level loneliness. They were compared with three common public health frameworks and models to identify the knowledge that public health already has to address loneliness. The final Adapted Social Exposome Model for Loneliness gives insight on how to assess social environments and their underlying causes for public health interventions, including marginalization, intersectionality, social instability, and agency to shape culture. This knowledge informs a new interdisciplinary definition of loneliness as an emotional and protective state of distress, informed by both cognitive and non-conscious processes to make sense of the social world in which someone lives, signaling exposure to different forms of marginalization from society when failing to meet cultural expectations and motivating ways to find belonging.