Kirkbride et al¹ provide a comprehensive overview of the social determinants of mental health. Their paper reviews the evidence for the causal influence of those determinants on population mental health and demonstrates the potential for prevention interventions that address those determinants across the life course. They argue convincingly that we stand at the threshold of a new era in prevention interventions for mental health globally – namely, those that focus on the social determinants of mental health.

Among the many contributions of their paper, several aspects stand out. First, the authors place a strong emphasis on a social justice framework when characterizing social determinants. As they point out, these are fundamentally a product of inequitable social and economic systems, which concentrate power and privilege in the hands of a few. Inequities in the distribution of mental health in populations are a product of experiences of exclusion and discrimination brought about by fundamentally unjust social systems. Second, the authors provide compelling evidence of causal links between social determinants and mental health outcomes, at both the individual and the wider social levels. These are documented with a strong emphasis on marginalized groups, which are frequently exposed to intersecting social determinants. Third, their review of the observational and intervention research strongly emphasizes a life course approach, demonstrating how early exposure to adversity carries lifelong mental health consequences, and why early intervention is so important. Fourth, they carefully document the evidence for social interventions that span the continuum of universal, selective and indicated prevention. Finally, their review demonstrates the modifiability of many social determinants, and the need to integrate a social determinants framework into existing, largely individually focused clinical treatments.

There are three key areas for future development of research on social determinants of mental health, which Kirkbride et al mention, but are worth highlighting here. The first is the need for more longitudinal observational research. Currently there is limited evidence on causal pathways linking social determinants to the mental health outcomes of populations. A recent study commissioned by the Wellcome Trust involves landscaping of longitudinal mental health datasets around the world and is a key step forward in advancing the field². This study has compiled more than 3,000 longitudinal datasets from 146 countries, improving their accessibility and opening possibilities for further analysis and enrichment.

The second area for future development is the evaluation of prevention interventions that address the social determinants of mental health. Three key steps are necessary if we are to prevent mental illness by addressing its social determinants. First, we need to build more robust theoretical models, mapping out the pathways by which social interventions yield mental health improvements. These may include distal socioeconomic mechanisms (for example, the mediating role of income instability in the association between economic recessions and the incidence of anxiety disorders) and more proximal neuropsychological mechanisms (such as the mediating role of self-regulation in the relationship between multi-dimensional poverty and adolescent depression). Second, we need to design studies that can test these mechanisms, for example by conducting randomized controlled trials that include analysis of key mediators in our hypothesized causal models. In order to demonstrate that a mediator is a causal factor, there must be a temporal relationship between that mediator and the outcome, a dose-response association, evidence that no third variable causes changes in the mediator and the outcome, robust experimental research and a strong theoretical framework³. Third, it is vitally important that we share data across diverse settings, because context really does matter when it comes to addressing social determinants. For example, specific experiences of multi-dimensional poverty or humanitarian emergencies brought about by climate change will vary substantially by context and will require diverse measurement and intervention approaches. There are also likely to be diverse mediators which may serve as targets for interventions. All of this requires an inter-disciplinary effort, bringing together economists, epidemiologists, mental health specialists, neuroscientists and people with lived experience, to develop shared approaches to these complex challenges.

As an example of this effort, in the Improving adolescent mental health by reducing the impact of poverty (ALIVE) study, we are designing and evaluating a selective prevention intervention to reduce the incidence of depression and anxiety among adolescents living in urban poverty in Colombia, Nepal and South Africa⁴. Our hypothesis is that multi-dimensional poverty increases risk for depression and anxiety among adolescents both directly and through its negative impact on self-regulation. By self-regulation we mean the capacity to set goals and maintain goal-directed behaviour, despite emotionally salient and challenging environments⁵. Our four-arm pilot trial includes an economic intervention (cash transfers, financial literacy, negotiation skills, and information about returns to education); an intervention designed to strengthen self-regulation; an intervention that combines economic and self-regulation components; and a control arm. The study includes detailed cultural adaptation and validation of key measures, and strong involvement of adolescents in the design and delivery of the research in each country site.

The third key area for future development is research on the social determinants of mental health in low- and middle-income countries (LMICs). As Kirkbride et al point out, most of the evidence on the social determinants of mental health (including observational and intervention research) originates from the Global North. It is vital that this trend is reversed. Most of the world's poor and vulnerable populations live in LMICs. The world's children and adolescents are concentrated in these countries (90% of the world's 1.2 billion adolescents live in LMICs⁶), making the argument for early life course interventions even more cogent. Although LMICs are highly diverse, they share a heightened vulnerability to looming climate change, conflict, and food insecurity. If we are to take seriously Kirkbride et al's call for a social justice approach to the social determinants of mental health, and develop population level interventions that have the potential to globally prevent mental health conditions such as depression, anxiety and psychosis, it is essential that greater research funding and policy attention is allocated to LMICs.

Kirkbride et al's paper is a landmark contribution that signals a growing community of practice across low-, middle- and high-income countries. Crucial for the future of this field is more robust engagement with policy makers and implementers in national governments and international aid agencies – such as multilateral development banks – to facilitate partnerships in funding, scaling up and evaluating the population level impact of interventions that address the social determinants of mental health.