Complex Systems Science

Health disparities involve a tangled web of influences, defined in terms of populations of individuals who evolve and coevolve over time in varied and dynamic contexts. Multiple causal pathways may interact over time, and in ways that differ substantially across population groups. How health disparities accumulate over the life course and are distributed over geographic space are important factors in identifying, understanding, and addressing them. Complex systems science allows researchers ways to move beyond simply identifying associations between determinants and observed levels of minority health and health disparities outcomes, and instead directly characterize and study heterogeneous, dynamic, and interdependent relationships. In this chapter, we describe how the development and use of complex adaptive systems models can help explain system-level nonlinear behaviors; incorporate consideration of spatial effects, social network structures, and process timing; consider heterogeneity in effect pathways; and act as "policy laboratories," allowing researchers to test interventions that cannot be feasibly explored in the real world due to time, cost, or ethical constraints. We then identify and give illustrative examples of three common model categories: etiological (i.e., ones that are primarily intended to explore how pathways operate within a given system to drive key outcomes), retrospective (i.e., ones intended to study why specific policies, interventions, and/or natural experiments had observed effects), and prospective (which explore the potential impacts of policies and interventions that are under consideration but not yet be implemented). Finally, we provide guidance on best practices in model design, development, use, and dissemination of findings.