From City- to Health-Scapes: Multiscale Design for Population Health
Reconciling the growing proportion of the global population that lives in urban centers with the goal of creating healthy cities for all poses one of the major public health challenges of the 21st century. Genetics has accounted for only 10 % of diseases, and the remainder appears to be from the interaction of multiple socio-environmental causes that potentially determine epigenetic changes leading to diseases. Therefore, quantifying the dynamics of socio-environmental factors and the environment-disease linkages is extremely important for understanding, preventing and managing multiple diseases simultaneously considering population and individual biological information of exposed and non-exposed individuals. This is particularly important for the aim of reprogramming health-trajectories of populations and developing/managing cities with a quantitative health-based design. Here we show how complex systems models, and specifically, dynamic network factor analysis (DNF), and global sensitivity and uncertainty analysis can map the exposome-genome-diseasome network (i.e., the macrointeractome), determine network factor metrics useful for urban design, and assess probability distribution of comorbidities conditional to exposure in space and time, respectively. These probabilities are useful to make syndemic predictions by for design of socio-technical and ecological systems and intervention strategies in existing cities. As a case study, we use the SHIELD study in Minneapolis focused on measuring children’s exposures to multiple environmental stressors and related effects on respiratory health and learning outcomes. Results show the very high degree of directional interaction among exposure factors and their spatial heterogeneity coupled to bi-directionally interacting diseases. We find non-linear conditional probabilities of disease co-occurrence and context-dependent dose-response curves that manifest large health disparities in populations. We show that macro socio-environmental features are much more important than biomarkers in predicting disease patterns with a particular focus on respiratory diseases and learning outcomes. Urban texture results as the most important factors, thus, such metric should be clearly considered in the design of socio-environmental systems via a minimization of the systemic health risk.
The developed probabilistic models are extremely flexible for the analysis of big data, city health-scape predictions, and optimal management of communicable and non-communicable diseases in socio-ecological systems via systems design. The understanding of linkages between structural, architectural, social, and environmental factors at the population scale will allow designers, architects, engineers, and scientists to design communities—from the material to the city scale—in which population health is the central objective of the design process.