Towards a complex systems model of evidence for public health

There has been growing interest in the adoption of complex systems approaches to tackle major public health challenges such as obesity. This perspective redirects focus from individually oriented and narrowly focused interventions towards strategies that reshape structural drivers of health and disease across multiple levels, offering novel avenues for enhancing population health and tackling health inequalities. Despite growing consensus on the use of a complex systems model of evidence to support these kinds of approaches, there remains limited agreement on the types of evidence required both to understand and to tackle complex public health problems. In this paper, we propose a complex systems model of evidence that combines three types of evidence-causal, intervention, implementation-across three dimensions of complex systems-mechanisms, dynamics and patterns. This model thus identifies nine categories of evidence: causal evidence, which explains the mechanisms behind public health problems, the dynamics driving changes therein and emerging health patterns; intervention evidence, which focuses on the set of actions that can modify these mechanisms and dynamics, including their intended and unintended consequences on health outcomes and implementation evidence, which addresses what is needed to implement systems change effectively, how systems adapt to these changes and how systems changes contribute to changes in health patterns. This complex systems model of evidence may serve as a guide to researchers and decision-makers when designing research programmes and evidence-based policies in response to complex public health problems.