Exposure assessment in cohort studies of childhood asthma

The environment is suspected to play an important role in the development of childhood asthma. Cohort studies are a powerful observational design for studying exposure-response relationships, but their power depends in part upon the accuracy of the exposure assessment.

The purpose of this paper is to summarize and discuss issues that make accurate exposure assessment a challenge and to suggest strategies for improving exposure assessment in longitudinal cohort studies of childhood asthma and allergies.

Exposures of interest need to be prioritized, because a single study cannot measure all potentially relevant exposures. Hypotheses need to be based on proposed mechanisms, critical time windows for effects, prior knowledge of physical, physiologic, and immunologic development, as well as genetic pathways potentially influenced by the exposures. Modifiable exposures are most important from the public health perspective. Given the interest in evaluating gene-environment interactions, large cohort sizes are required, and planning for data pooling across independent studies is critical. Collection of additional samples, possibly through subject participation, will permit secondary analyses. Models combining air quality, environmental, and dose data provide exposure estimates across large cohorts but can still be improved.

Exposure is best characterized through a combination of information sources. Improving exposure assessment is critical for reducing measurement error and increasing power, which increase confidence in characterization of children at risk, leading to improved health outcomes.