ABSTRACT
This chapter reviews several approaches for deriving evidence from regional observational studies. Observational studies have the advantage of reflecting realistic exposure conditions, but analyses may be hampered by high natural variability and the influence of confounding factors. The chapter focuses on the use of observational studies to provide evidence that the proximate cause and the biological effect co-occur or that the level of exposure to the stressor is sufficient to induce the effect. Stressor levels associated with biologic effects in the case can be compared with site groups categorized by different stressor levels or different effect levels. Categorical comparisons can be made graphically or in tabular form. A typical stressor–response model quantifies a change in a biological variable with changing (increasing or decreasing) exposure to a single stressor or a set of stressors that consistently and strongly covary.
