Guideline Development Process
Background
The Health Implications of Social Connection
Decades of research show that the quality of our social lives are critical to our health, happiness, and wellbeing. Social connection is a basic human need. When we lack meaningful and fulfilling social connections, we experience loneliness. Similar to hunger and thirst, loneliness is our body's way of telling us what we need. When we fail to meet these needs, we experience severe negative impacts on our mental and physical health. Indeed, loneliness and social isolation can lead to stress, depression, and a weakened immune system. Studies show that the effects of being disconnected often outweigh the harmful effects of obesity, sedentary living, and even smoking. As such, building and maintaining strong social connections, through close relationships and everyday interactions, is essential for enhancing our overall wellbeing and quality of life.
Mechanisms Linking Social, Mental, and Physical Health
The connection between our social, mental, and physical health is intricate and influenced by multiple biological, psychological, social, and structural processes. Several of these processes are now well documented: First, loneliness and social isolation are distressing experiences that trigger the body’s stress response. Over time, the physical effects of stress lead to wear and tear on the body, ultimately resulting in poorer health. Second, as social beings, humans depend on each other for emotional regulation and support. This support helps buffer against stress and ensures our basic psychological needs are met. Without this social support, individuals become more vulnerable to health issues because they don’t have the social resources needed to buffer against the stresses they face. Third, people are seen to cope with loneliness in a variety of ways, some of which are not healthy. For example, we might eat unhealthy foods to feel better when we’re down, we might avoid physical activity in order to save our energy for other priorities, or we might consume alcohol to distract from feelings of loneliness. In addition to those direct pathways, poor social connections often stem from other health and societal challenges, like chronic illness or disability, financial insecurity, our environment (e.g., lack of housing, safe public space, lack of usable green space), lack of digital access, or lack of transport. These conditions contribute to loneliness because they can limit our ability to be with others. Thus, individuals with concurrent challenges experience a feedback cycle wherein poor health and social disconnection reinforce each other. In each of these ways, our social, mental, and physical health are linked. We can only be physically and mentally healthy when leading a healthy social life.
Societal Impacts of Social Disconnection
Social isolation and loneliness, apart from impacting our health and wellbeing, also have significant impacts on our society. Indeed, when people lack strong social connections, communities become less cohesive. Less cohesive communities can lead to weaker social bonds and a reduced sense of trust and cooperation, both horizontally (between people) and vertically (between institutions and individuals). This can result in a range of negative outcomes in a community, such as increased polarization, where people become more divided and less able to understand different perspectives. Social disconnection can also contribute to the rise of extremism and radicalization because isolated individuals may be more vulnerable to harmful ideologies held by radical groups. In contrast, when communities foster strong social connections, they experience greater social cohesion, which contributes to social stability, better political collaboration, and stronger economic growth. Therefore, the health of our society is deeply linked to the strength of the connections between its members.
Public Health Efforts to Improve Social Wellbeing
Recognizing these benefits, existing efforts to improve social wellbeing have led to a range of initiatives aimed at fostering stronger social connections, particularly among certain groups, such as socially isolated seniors. In some countries, such as the United Kingdom and Japan, national strategies have been developed to create more inclusive communities and improve social infrastructure. These strategies have included awareness campaigns highlighting the importance of social connections, programmes to address digital inclusivity for seniors and programs encouraging social engagement among older adults and intergenerational relationships. Policies have also been introduced to support activities that build community ties. At the individual level, interventions have been designed to help people develop the skills and resources necessary for maintaining healthy relationships. However, challenges remain in scaling successful programs across all age groups, integrating social connection initiatives into broader public health, health care, welfare and social services, meeting needs of diverse populations, and securing ongoing funding and policy support. To address those challenges, further efforts are needed to make social connection a policy and practice priority.
Public Health Guidelines for Social Connection
One strategy to prioritize the importance of social connection is to develop national public health guidelines for social connection. Indeed, researchers have suggested that creating public health guidelines for social connection could help raise the status of social connection as a determinant of health, just as food and physical activity guidelines have done for food and exercise. Indeed, such guidelines could help raise awareness about the critical role social connections play in overall health, encouraging healthier social behaviours and promoting positive social development. Social connection guidelines would also provide clear, measurable targets for social connection, which could be used to tailor care and welfare, health promotion and prevention strategies, to evaluate programs, monitor public health, and guide academic research, and contribute to policy agenda setting. Additionally, these guidelines could drive further rigorous research on the relationship between social connection and health, ensuring that policies and programs are informed by solid evidence. Ultimately, such guidelines would support the creation and implementation of initiatives that help individuals and communities strengthen their social ties, leading to better health outcomes and a more socially cohesive society.
Methodology
Project Aim
The aim of this project was to develop recommended public health guidelines for social connection based on available evidence, expert perspectives, and community consultations.
Project Design
To accomplish our aim, this project employed a multi-step approach to develop public health guidelines for social connection, integrating both qualitative and quantitative research techniques to ensure the guidelines have been rigorously evaluated. The core component of this project was a Delphi study, which involves iterative rounds of surveys and feedback from a global panel of social connection experts to systematically draft our recommended public health guidelines. This approach was complemented by a series of focus group interviews with key populations, evidence reviews of existing literature, and rapid surveys designed to provide empirical insights and preliminarily assess the feasibility and acceptability of the proposed guidelines. By combining such methods, we were able to capture and integrate diverse views and perspectives.
Below we describe in more details each of the components mentioned above:
Delphi Study
Participant Recruitment.
We recruited expert consultants using several methods. First, we directly contacted frequently cited authors in the fields of loneliness, isolation, and social connection through the email addresses of corresponding authors. Second, we distributed invitations via email list servers of professional societies. Third, we employed snowball sampling by encouraging participants to nominate peers or forward our invitation email to others in their networks. In total, we extended between 500 and 700 direct invitations. Additionally, our invitation was shared through four email list servers associated with prominent social psychology professional organizations and networks. While we cannot precisely quantify the number of participants reached through chain referral, this approach allowed us to broaden our recruitment reach significantly.
Round 1 (n = 95)
In the first round, we used open-ended questions to gather initial ideas about several key areas: (a) the foundational principles for developing and implementing public health guidelines for social connection, (b) potential guidelines tailored to individuals, (c) potential guidelines aimed at collective entities such as organizations, communities, and governments, and (d) factors that influence social connection and wellbeing. The responses from those open-ended questions were analyzed thematically, allowing us to identify and synthesize the key concepts within each area, which then informed the development and categorization of potential guidelines for the next round.
Round 2 (n = 92)
In the second round, we presented results of the thematic analyses as a series of statements that participants were asked to rate (a) their level of agreement with the guiding principles for guideline development (using a scale from Strongly Agree to Strongly Disagree), (b) the importance of each guideline (ranging from Absolutely Essential to Should NOT Be Included), and (c) the significance of each factor related to social connection (from Extremely Important to Not Important At All). Participants were also given the opportunity to provide additional feedback through open-ended comments in each section. Descriptive statistics were used to quantify the levels of support, and thematic analysis was employed to interpret the open-text responses. These data informed the development of an initial draft set of 12 guidelines.
Round 3 (n = 82)
In the third round, we presented participants with the draft guidelines, developed from the results of Round 2. For each guideline, they were asked to rate their level of expertise on the content (High to None), indicate whether they supported the inclusion of the guideline (Yes, No, I prefer not to vote), and provide feedback. We established specific criteria for guideline approval, requiring at least 80% endorsement from those who voted "Yes" or "No," 75% support from all participants (including those who chose not to vote), and 85% approval from those with "High" expertise in the relevant domain. The 80% threshold is a common standard in Delphi studies, but we included more stringent criteria (75% of all participants and 85% of domain experts) to ensure that guidelines were strongly supported, especially by those with relevant expertise.
Round 4 (n = 35)
In the fourth round, we held a series of focus group interviews to refine the wording of the guidelines. Although the guidelines had already received consensus approval in Round 3, our goal was to ensure that they were communicated effectively to public audiences while staying true to their original intent. Throughout this process, we iteratively revised the guidelines, with a total of seven revisions made between focus group sessions. The final guidelines and accompanying considerations were carefully reviewed and approved by our internal research team.
Appendix 1 contains a comprehensive report of the Delphi study results.
Complementary Research Activities
In addition to the Delphi Study described above, we also undertook complementary research activities in order to inform the development of our guidelines. These complementary activities included (1) rapid evidence reviews, (2) key population focus group interviews, and (3) quantitative surveys assessing preliminary acceptability and feasibility of the guidelines:
Rapid Evidence Reviews
The first was the development of a series of evidence briefs to inform our understanding of the evidence base relating to our developed guidelines. In total, 49 such evidence briefs were developed. In addition to reviewing published literature, the evidence briefs also included relevant quantitative or qualitative analyses from the Canadian Social Connection Study, an annual serial cross-sectional survey with a longitudinal sub-cohort. These data analyses were used to conduct informal tests of salient hypotheses (e.g., examine the empirical relationship between number of friends and various indicators of mental and social wellbeing).
Key Population Focus Group Interviews
Additionally, we conducted focus group interviews with individuals of diverse ages from five key populations:
- Two-spirit, lesbian, gay, bisexual, trans, queer, intersex, asexual and other sexual and gender minorities;
- Racialized peoples;
- Indigenous peoples;
- People living with disabilities; and
- Migrants, immigrants, and refugees.
These communities were selected in recognition of the distinct challenges that individuals from these communities sometimes face in seeking equity and inclusion.
During the focus group interviews, participants were presented with the guidelines as they were developed after Round 3. These were the same guidelines presented to our experts in the third round of the Delphi study. In presenting the guidelines to these audiences, we were interested in the perceived feasibility and acceptability of these guidelines and how the unique lived experiences of individuals from these communities might shape their understanding of and ability to comply with the guidelines. These focus groups were conducted between Rounds 3 and 4 of the guideline development process and the considerations from these guidelines were incorporated along with the expert feedback from Round 4.
Appendix 2 includes a detailed report of results from the key population focus group interviews.
Quantitative Surveys Assessing Acceptability and Feasibility of Guidelines
Finally, we leveraged an online survey to provide preliminary insight into whether the guidelines we developed were feasible and acceptable to the general population. Participants for the online survey were predominately recruited through paid advertisements on Facebook and Instagram. The guideline-related questions were administered to a subset of participants. These questions assessed whether participants (1) recognized the importance of social connection to health; (2) agreed that the guidelines were (a) achievable for a typical person, (b) achievable for them personally, (c) important to leading a healthy social life, (d) whether they were currently living in accordance with the guideline, (e) whether they believed the guideline would benefit those who followed it, and (f) whether the guideline was already common knowledge. Additionally, participants could provide open-text reactions to these guidelines.
Results from our surveys are detailed in Appendix 3.
Grading of Developed Guidelines
After each of the study components described above were completed, we undertook a grading exercise to systematically evaluate each guideline's strength and the quality of evidence supporting it. Due to the subject matter, breadth of our social connection guidelines, and nature of the supporting evidence upon which the guidelines are based, we decided that a formal Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach was not feasible. As such, our grading approach was informed by the GRADE process, but adapted to meet the specific needs of this project. In doing so, our approach provided a structured framework to assess the balance of benefits and harms, values and preferences, feasibility, and overall certainty of the recommendations.
Evidence Collection and Synthesis
The evidence used in the grading evaluation was drawn from the Delphi study, evidence reviews, key population focus group interviews, and quantitative surveys. This multi-source approach ensured that the guidelines were based on a comprehensive and robust evidence base, incorporating both expert consensus and empirical data.
Evaluation Criteria
Each guideline was assessed across five key domains:
- The strength of the research base supporting the guideline, with a focus on the consistency, generalizability, and rigor of the studies reviewed.
- An assessment of whether the benefits of following the guideline outweighed potential harms or risks, taking into consideration any unintended consequences.
- The alignment of the guideline with the values and preferences of the target community or population, ensuring that the recommendations are culturally sensitive and widely acceptable.
- The practicality of implementing the guideline in real-world settings, considering resource availability, existing infrastructure, and potential barriers to adoption.
- The overall confidence in the recommendation based on the quality of evidence, strength of consensus, and expected impact.
Rating Process
The grading evaluation process involved synthesizing evidence and applying criteria to assess the strength of each guideline's recommendation and the quality of evidence supporting it.
Strength of Recommendation
The strength of a recommendation reflects the degree of confidence that the benefits of following the guideline outweigh any potential harms or risks. In determining the strength of each guideline, several factors were considered, including the balance of benefits and harms, the alignment with the values and preferences of the target population or community, and the feasibility of implementation in real-world settings.
Below is a detailed description of the criteria used to classify each guideline into one of three levels of recommendation: Strong, Moderate, or Conditional.
- Strong Recommendation: The benefits of following the guideline clearly outweigh the potential harms or risks. There is high alignment with the values and preferences of the target population or community, and the guideline is highly feasible to implement in real-world settings. This recommendation is highly encouraged for implementation.
- Moderate Recommendation: The benefits of following the guideline generally outweigh the potential harms or risks, though the balance may be less clear. The guideline aligns well with the values and preferences of the target population or community, but there may be some challenges in feasibility. This recommendation is encouraged, but some contextual factors may need to be considered.
- Conditional Recommendation: The benefits of following the guideline may be comparable to the potential harms or risks, with significant variability depending on context. The guideline may align with the values and preferences of some, but not all, target populations or communities, and feasibility may be limited by resource constraints or other barriers. This recommendation should be applied with caution, and further research or adaptation may be needed.
This classification helps stakeholders understand the level of urgency and priority that should be given to each guideline when implementing public health strategies for social connection.
Quality of Evidence
The quality of evidence indicates the degree of confidence in the research findings that support a particular guideline. It reflects the robustness, consistency, and applicability of the evidence base, as well as the likelihood that future research will alter the findings.
Below we outline the criteria used to evaluate the quality of evidence, categorizing it into four levels: High, Moderate, Low, and Very Low.
- High Quality: The evidence base is robust, with consistent findings across multiple high-quality studies (e.g., systematic reviews, meta-analyses, well-conducted cohort studies) that are directly applicable to the guideline's context. The results are unlikely to be significantly impacted by future research.
- Moderate Quality: The evidence base is solid, but there may be some inconsistencies or limitations in the available studies. The findings are generally applicable to the guideline's context, but there is some uncertainty due to variability in study design, population, or outcomes. Future research may impact the confidence in the results.
- Low Quality: The evidence base is limited or inconsistent, with significant variability in study findings or relevance to the guideline's context. The results are highly uncertain, and future research is likely to have a significant impact on the confidence in the findings.
- Very Low Quality: The evidence base is insufficient, with little to no relevant research available, or the studies that do exist are of poor quality and provide minimal guidance. The findings are highly uncertain, and further research is essential to draw any meaningful conclusions.
This categorization provides a clear understanding of how much trust can be placed in the evidence underpinning each guideline, which in turn informs decision-making processes related to public health interventions and policy development.
Future Work
Guidelines are a starting point, not a finish line
The development and implementation of public health guidelines for social connection represent a pivotal step in addressing the profound levels of loneliness, social isolation, and disconnection experienced by people across the globe. However, we acknowledge that the guidelines presented in this report are only a starting point. Our recommended guidelines are intended to motivate and support behaviours, policies, and practices that will improve social connection. We acknowledge that guidelines alone will not solve the social connection crisis. In particular, guidelines are primarily a prevention tool and their efficacy for people who are already lonely or isolated may be limited.
Recognizing the need for further, high-quality research
In considering the evidence supporting our suggested guidelines, we acknowledge the strong reliance on limited observational data of varying degrees of quality. While grounded in a comprehensive review of the available literature, in consultation with more than 100 global experts and members of the public from diverse communities. Still, there will inevitably be opportunities to gather high quality observational data to further improve and refine potential guidelines.
Furthermore, we are aware that many of the samples of existing studies excluded key populations and may not be generalizable to all individuals and communities. As such, and just as with all other established public health guidelines, this first iteration will undoubtedly be benefitted by future research with more diverse populations. It is our hope that the guidelines can catalyze more applied health and social science research.
Recognizing inherent research limitations
In acknowledging deficiencies in existing studies used to inform our guidelines, we think it is also important to note that it would be inappropriate to judge our guidelines against the present day standards used in other guideline development processes. Indeed, while randomized controlled trials (RCTs) may be well suited for addressing research related to exercise and nutrition, it is relatively difficult to randomize individuals to a particular pattern of social life (e.g., such as the time they spend socializing, or the number of friends they have). Some RCTs certainly do exist, but they are often limited in scope and it is difficult to know whether experimental and real world conditions are similar enough. The major source of reliable data would thus come from observational studies and other methodological approaches and forms of analysis.
Longitudinal studies and replication
Given the difficulty of research in applied health and social sciences, it is especially important to invest in longitudinal observational studies that can better isolate the effects of social connection on health; that is crucialfor disaggregating these effects from person-level characteristics. Replication studies are also of critical importance to ensure that the findings of one study are generalizable. With sufficient investments, we will be able to increase the specificity of public health guidelines for social connection and provide greater confidence in developed guidelines.
The importance of studying individual and group differences
Another important issue worth considering with respect to the quality of evidence supporting our guidelines is whether there are important characteristics that might warrant guidelines specific to different demographic or otherwise-defined groups. Indeed, it remains plausible that individuals may benefit differently from different guidelines we’ve provided. However, in calling attention to the potential value of such analyses, we also think it is important that researchers pay close attention not only to the statistical significance of their findings, but also the effect sizes and meaningfulness of potential group differences. Qualitative approaches can also be used to understand the relative importance of various contributors to social wellbeing.
The importance of evaluating interventions
In addition to investments in observational research, it is also necessary to conduct studies that help evaluate the effectiveness of public health and medical approaches to improving social connection. Research is needed in order to demonstrate that change in social behaviour is feasible and clearly beneficial. Such studies can help improve the specificity of guidelines as well as identify potential subgroups that may require customized guidance.
The need for research frameworks and standards
To support future research, it may also be valuable to develop precise tools and measures that capture key constructs relevant to the guidelines, including measures of adherence. Such tools can help systematize measurement, and monitoring, thereby allowing for greater comparability across studies and across-implementation settings. This comparability might eventually allow for meta-analyses which can form the basis of future guidelines.
Ongoing refinement of guidelines
With the scientific advancements outlined above, we are hopeful that we cannot only refine specific guidelines, but also begin to understand the relative contributions of these guidelines to individual- and community-level social wellbeing.
Acknowledgements
The authors would like to thank all the participants who contributed to the process of developing our recommended guidelines, including (a) the international experts who participated in our Delphi study, (b) those who contributed evidence briefs for our consideration, and (c) the community-members who participated in our key population focus groups and surveys.
Funding
This project was funded by a Project Grant from the Canadian Institutes of Health Research. Dr. Kiffer G. Card received salary support from a Michael Smith Health Research BC Scholar Award, allowing him to undertake this work. We thank the funders for their generous support.