This mapping exercise confirmed that a wide range of community-based interventions targeting improved public mental health are currently being delivered across England, showing some evidence of being tailored to key drivers of PMH locally. Assessing this was challenging, and determining effectiveness was not possible, given we were unable to consistently collect information of an evaluative nature, including: measurement of outcomes; costs; funding; participant numbers; and how long interventions had been running. Across localities there was strongest provision for social isolation and loneliness, most commonly through social activities and/or befriending services, yet broader interventions focusing on wider structural determinants are uncommon. There appeared to be some tailoring of services according to the relative size of population sub-groups.
Determinants of public mental health: do the services available respond to drivers of PMH?
When compared with our conceptual work exploring the determinants of both poor and positive public mental health across the life course  this mapping exercise suggests that provision of public mental health interventions is currently limited to addressing a small number of these determinants. Across the five localities, the provision of services focused on addressing only a small number of the risk and protective factors, principally, the individual and their social environment. We found very few interventions focused on other broader determinants of mental health and wellbeing, including issues surrounding housing, retirement, and family circumstances. We did not identify interventions aimed at structural and environmental determinants, such as air and water quality, population density, walkability of local environment and urban decay, economic recession, climate change, natural disasters, media and advertising, the welfare system and political structures.
One explanation for the disparity between PMH determinants and service provision is that some risk or protective factors may be considered by local authorities and third sector organisations to be more important than others, therefore there may not be a requirement to have specific interventions or strategies targeting every determinant. For example, if communities can simultaneously tackle social isolation, loneliness and financial stress, then this may create a solid foundation for individuals to build resilience to other problems. Evidence suggests that social support may moderate environmental vulnerabilities and confer resilience to stress [60, 61]. However, the evidence base is insufficient to support a hierarchical approach to tackling risk and protective factors. Furthermore, there is likely to be coalescence, for instance, between loneliness and other aspects of deprivation.
The high number of services captured offering social activities and/or befriending may reflect the recent growth and prominence of social isolation and loneliness in UK government campaigns and high profile third sector organisations making it a priority for decision-makers (e.g., Let’s Talk Loneliness Campaign; Campaign to End Loneliness; Co-op Loneliness Campaign; Country Living Loneliness Campaign; Age UK campaigns). For example, the England pilot region for the Campaign to End Loneliness was Cambridgeshire and Peterborough where the highest number of social isolation and loneliness interventions were identified in the mapping exercise. This suggests that the risk factors/protective factors and subpopulations ‘targeted’ by these interventions may reflect political will and council priorities . It is also plausible that social activities and befriending services were so common across all five localities because of their ease of provision and relatively low cost of delivery compared to more complex and expensive interventions that require a skilled workforce such as advice services, advocacy and legal support or education and training interventions.
We consider it likely that this study did not map interventions aimed at structural and environmental determinants because they did not explicitly state that they targeted mental health outcomes. This may indicate that providers of these broader services do not view themselves as playing a PMH role and/or are not involved with strategic discussions around PMH. It is also possible that we did not find evidence of these determinants of PMH being addressed at the community level because local authorities and third sector organisations do not currently have the resources, authority, influence, or power that is required to deliver such services. If this is the case, then consideration should be given to expanding the scope and abilities of community-led responses as evidence suggests that community approaches play an important role in increasing people’s self-efficacy, confidence, helping them to develop a sense of control over their own lives, reducing health inequalities, improving health outcomes and increasing resident’s sense of wellbeing [63, 64]. Therefore, there may be benefits to developing community-based interventions to tackle as many determinants of PMH as possible. If drivers associated with structural and environmental determinants are deemed important to address at the community level, then greater integration of policy and practice may be required, and reflected in engagement of a wider range of stakeholders from strategy to delivery and receipt of support.
Universal and targeted interventions: gaps in provision of targeted interventions
The mapping exercise identified universal interventions that are open to all residents, as well as interventions targeted at specific groups in the community. For optimal improvements in PMH it is important to find the right balance between these two approaches . Universal interventions have the advantages of being less stigmatising, less vulnerable to funding cuts and potentially easier to deliver than targeted interventions. However, they lack focus on the individualised needs of high-risk groups. Targeted interventions on the other hand can be tailored to the precise needs of vulnerable groups, which means fewer resources deployed on people who are unlikely to develop poor mental health  and therefore avoid further enabling advantage and increasing inequalities.
We found that the highest proportion of targeted interventions were for older adults. These interventions predominantly aimed to tackle social isolation and loneliness, especially in the more rural areas of Redcar and Cleveland and Cambridgeshire. We know that older adults and people living in rural areas have an increased risk of experiencing social isolation and loneliness . However, we also know that older adults have needs other than social isolation and loneliness such as financial stress , physical activity for wellbeing  and bereavement . Whilst targeted interventions for these challenges were identified in one area, in others they were not.
The second highest proportion of targeted interventions identified in the mapping exercise were aimed at people from a minority ethnic background. We know that those from minority ethnic groups are disproportionately affected by poor mental wellbeing, likely as a result of long-standing discrimination and disadvantage , therefore the existence of these interventions is encouraging as it indicates that attempts are being made to address the needs of these sub-populations. However, further targeting based on additional stressors was rare. There were very limited interventions found that specifically focused on men from minority ethnic backgrounds and these were only found in the study sites based in London. This is concerning given the evidence that men of minority ethnic background are less likely to seek help for common mental health problems [70, 71] and the evidence that culturally sensitive interventions are more likely to lead to positive outcomes .
With respect to other potentially underserved populations, there are a few notable gaps in provision. The mapping exercise revealed that there are a very limited number of interventions specifically aimed at LGBTQ+ people of minority ethnic background. This is concerning as people who belong to multiple minority groups are particularly vulnerable to developing poor mental health due to the cumulative effects of being exposed to experiences of stigmatisation, discrimination and fear of rejection from the wider population as well as others from each minority group [73,74,75]. Finally, the mapping exercise also revealed there to be a limited number of interventions targeted at carers. It is well known that caring can have an adverse impact on mental health which carers attribute to a lack of support . This may indicate a gap in provision for this high-risk group or that services for carers do not explicitly state that they aim to improve the mental health and wellbeing of the carer and therefore were not captured by this study.
COVID-19: implications for community PMH provision
The pandemic has likely impacted many determinants of PMH especially social isolation and loneliness, financial stress, job insecurity and unemployment, caring responsibilities, bereavement and gender-based violence [28,29,30,31]. The mapping exercise identified many interventions in place prior to March 2020 indicating the presence of expertise and infrastructure to respond to these challenges. However, the delivery of many public mental health support interventions was limited by lockdown measures , and likely affected the interventions mapped in this study. Digital technology enabled some interventions to continue to provide their services virtually during the pandemic , however, due to digital exclusion and inequality, access to such provision is unequal among residents [77, 78]. As this mapping exercise was undertaken prior to March 2020, we did not collect information regarding whether identified interventions have adapted or suspended their service provision.
The pandemic has not affected all communities equally, with high-risk sub-groups including: health and care workers; non-medical frontline workers (such as shop workers); members of Black, Asian and minority ethnic groups; people of lower or precarious incomes; people who have experienced COVID-19 or COVID-19-related bereavement; as well as older adults and people who have severe health conditions and had to “shield” for prolonged periods of time [28, 29, 31].
Given that almost half of the interventions identified in this mapping exercise were universal interventions, strategies need to be developed and implemented at national and local level to ensure the provision of timely and effective support to reduce or mitigate the risks of poor mental health among these higher risk groups [29,30,31]. Therefore, a shift toward targeting may urgently be needed for some part of the foreseeable future.
The primary limitation of the study is that it is difficult to assess how comprehensive the mapping was for each study site. It cannot be guaranteed that our community contacts were able to provide the most accurate and up-to-date information about interventions and services in their area and some local authority and third sector websites were out of date and incomplete; this is particularly pertinent given the fast-changing provision landscape resulting from the COVID-19 pandemic. Small community-based interventions (particularly those with a limited online presence – such as those organised by minority ethnic groups) may have been missed. Though we used multiple sources to identify and obtain information to minimize such limitations, detailed more community engaged exploration of community-funded organisations in particular is warranted in the future.
Only interventions and services that explicitly aimed to improve mental health or wellbeing were included, and we excluded wholly private-for-profit interventions. Work-based interventions and community focused initiatives by business are hence not captured. Many services positively impacting public mental health may not list mental health or wellbeing improvement as an explicit aim (e.g. a cycling group, a yoga course, a book club, choirs etc). This observation leads us to note that opportunities for collaborative and more effective working between mental health stakeholders and these services might currently be missed. We also excluded interventions aimed at children and adolescents, although we did include parenting interventions.
Inevitably, there was some subjectivity regarding how to define a “unit” of intervention, for example in the case of one small community group providing many different services. Equally, many small-scale social activity interventions might reach fewer people in total than one financial stress intervention reaching hundreds of people. As such, frequency data should be interpreted with some caution, and we suggest that patterns of relative provision are more insightful.
Lastly, the study was unable to consistently collect information on funding, programme numbers, and how long interventions had been running. This was potentially due to complexities surrounding the project-based nature of interventions where different funding streams begin and end at different stages within the life of the project. The study was also unable to collect information on the percentage of interventions that had been evaluated. This is because there were complexities within some organisations that delivered interventions. Either only part of a larger intervention had been evaluated or evaluations had focused on whether individuals had benefited from their overall involvement with an organisation, but each individual may have undertaken different individual interventions or a different combination of individual interventions within that organisation. It also should be considered that the existence of an intervention does not guarantee it is well designed or delivered, or effective at mitigating PMH risks. Indeed, the paucity of evaluation data for identified interventions means that this mapping study cannot determine their impact on PMH, only patterns of provision and localised prevalence of particular types. In future, policy-makers could address this heterogeneity of information about interventions and therefore enhance the availability of evaluation data, by making embedding standardised monitoring and evaluation in funding conditions, enabling more consistent or standardised funding streams, and/or providing better support for evaluations in the form of funding and/or expertise.