Newsletter | The Digital Determinants of Health
How do you begin to understand the way that pervasive trends have reshaped the experience of everyday life? And how do you mitigate – and forestall – the resulting harms to our health? According to a new series of papers in The Lancet, “digital systems are fundamentally reorganising the conditions and environments in which health is determined,” and “the safeguarding of health is lagging far behind the pace at which technology is woven into non-digital activities, institutions, and environments.”
The authors of the papers in the Lancet series try to wrap their arms around a transformation of the daily experience across the globe that has taken place over the past three decades – one that is currently accelerating with the emergence of AI – and understand it from the vantage point of public health. It’s a daunting task, but the three papers provide a helpful start, offering a framework for understanding the structures and forces that generate the ongoing digitalization, assessing the health impacts – particularly on children and youth – of the transformation, and recommendations for better aligning its impact with public health goals. It’s an interesting – and helpful – piece of work that connects nicely with the work that we’ve been doing to measure the health impacts of everyday products and services and to develop new mechanisms for accountability for these impacts.
The paper on health impact, by Louise Holly et al, is sobering, both the scale (and direction) of impact that it finds and for the inadequacy of the science at this point of the digital transformation. The authors look at a vast range of studies of digital media use (DMU) for children aged 0-10 and youth 10-24, finding that:
“Across 121 systematic reviews (including 52 meta-analyses) covering 2790 unique primary studies among nine million people aged 0–24 years, DMU is consistently associated with adverse mental health and wellbeing, behavioural, physical, cognitive, and social outcomes.”
These reviews found that for children, higher screen time was associated with negative emotions or conduct problems and common mental health problems and smartphone/ tablet use was associated with poorer sleep. For youth, “problematic DMU” was associated with negative emotions or conduct problems, common mental health problems,loneliness, poorer quality of life, self-harm or suicidal ideation or behaviour, sleep problems, reduced physical activity and alcohol use or other risk behaviours. Screen time for youth was associated with obesity and social media use with common mental health problems. All told, the review found that across the reviews, which represented nine million subjects, 71% of the findings for children and 80% for youth were negative.
Using a technique similar to what we employed in our recent work to quantify health impact, the authors went on to estimate the global health impact implied by the adverse impacts found in the systematic reviews, calculating that 20% of the cases of depression and 15% of the cases of obesity in children and youth are due to digital media use.
To be fair, the evidence is a bit ragged. The authors note many limitations in the collections of studies they examined: a lack of standardization of concepts and measures (e.g. there is a need for measures more sophisticated than “screen time”); causality and reverse causality issues; geographic skewing (mostly studies come from high-income western countries); and the problems that result from the near ubiquity of the technology. Fundamentally, the science of studying the impact of digital technologies is not developing as fast as the technologies, their application and adoption.
The vast extent to which digital systems have been adopted and adapted into our lives is a key reason to lean on the precautionary principle. The authors argue, “given that exposure is so broad, the recognition of risk – even without conclusive evidence – that protecting children's and adolescents’ health calls for a precautionary, structural response—one that moves beyond individual behaviour change to reshape the digital systems, incentives, and knowledge practices that determine how digital media affects health.”
Another paper in the series, by James Shaw et al, provides a language and a systems-level view for understanding the development, adoption and impact of the digital transformation. The authors see the “digital determinants of health” as the “structural conditions that affect health and wellbeing through digitally mediated infrastructures, business models, design choices, and information environments.” The digital determinants result from a deeper “digital sociotechnological order” they define as a set of coordinated dynamic structures that enforce a particular set of assumptions and practices of the digital era by producing new ways of participating in politics (political norms), conducting business (economic logics), building foundations for digitally mediated activities and networks (infrastructures), and relating to self and others (social and personal practices). The interplay of these concepts is shown in the figure below:
There’s a lot going on in this framework. Suffice it to say that the downstream effects we see from products and services like Insta and TikTok, like Netflix and DoorDash, have deep roots in shared, often unspoken and undebated assumptions such as the importance of data-driven knowledge, the prominence of individual choice and responsibility, and the moral imperative to pursue technological progress. What’s useful about this analysis is that it allows for looking forward as well – for understanding how AI is, and likely will, play out absent strong counter-action.
The series concludes with a paper on governance, by Barbara Prainsack and her colleagues. They note that the governance of digital determinants of health, as with many other determinants of health, lies largely outside the strict jurisdiction of public health – thus the need for a cross-cutting, "health in all policies” type approach that coordinates governance across the many agencies, at different levels. This challenge is complicated by the AI arms race between the superpowers that pushes back against regulation. The paper offers several recommendations, ranging from proactive, assertive engagement by public health agencies to using health impacts as a test case for digital governance approaches more broadly. As is often the case when addressing massive, sprawling, deeply embedded problems, these recommendations sound underpowered, but the paper’s conclusion is cogent:
“The DDoH are neither neutral nor immutable. They reflect political choices about power, responsibility, and values. A set of coordinated reforms, anchored in public health values and law, enforced through national, supranational, and international regulatory capacity, embedded in procurement and competition policy, and guided by child-centred and youth-centred protections, could substantially reduce harm as well as preserving space for beneficial innovation. The stakes are generational. The soft and hard digital infrastructures being consolidated today will shape the health trajectories of children and young people for decades. The governance of the DDoH must be understood as a primary instrument for aligning digital transformation with equity, prevention, and collective wellbeing.”
A handful of academic papers does not a revolution make, but this series from The Lancet is important nonetheless. From our very beginning, Building H has been arguing that the product environment, the set of everyday products and services with which we interact regularly, is a driver of health. That public health – and the industries that produce this environment – have to take it as more than a given and actually start to re-shape it in the direction of healthy. Seeing these calls echoed in prominent journals like The Lancet is progress. Such thoughtful, deliberate analyses represent progress. There is much, much work to be done, but much to give us hope.
Read the full newsletter.