Individual causal models and population system models in epidemiology.

Area of Science:

• Epidemiology
• Systems Science
• Public Health

Background:

• Traditional epidemiology views populations as collections of independent individuals.
• This perspective often overlooks how connections between individuals impact population health.
• A new theoretical structure is needed to address these interconnectedness dynamics.

Purpose of the Study:

• To propose a theoretical framework for "population system epidemiology."
• To integrate individual-level health determinants with population-level interaction patterns.
• To advance the understanding of how social and biological networks influence health outcomes.

Main Methods:

• Contrasting infection transmission models with sufficient-component cause models.
• Highlighting the need to integrate time, interactions, and joint exposure effects.
• Leveraging G-estimation and discrete individual transmission models for integration.

Main Results:

• Sufficient-component cause models focus on individual exposures and assume populations are sums of individuals.
• Transmission models simulate individual interactions over time, showing nonlinear population risks.
• Neither model alone fully captures the complexity of population health systems.

Conclusions:

• A "population system epidemiology" requires integrating individual-level exposures with interaction determinants.
• This integrated approach should model time-related processes and network effects.
• Advances in G-estimation and transmission modeling offer pathways for this integration.