Reliability of lifetime residential history and activity measures as elements of cumulative ambient ozone exposure assessment.

UNLABELLED

Assessment of long-term effects of lifetime ambient oxidant pollution relies on ecological exposure assignment. Given the characteristics of ozone exposure profiles, with daily and seasonal peaks and clearly lower indoor concentrations, exposure misclassification is of considerable concern. This study considered residential location, time spent outdoors, and outdoors activity as major determinants of "effective" lifetime ozone exposure. Given the lack of a valid "gold standard" for these measures, we evaluated the repeatability of the retrospective assessment of these factors.

METHODS

A convenience sample of 175 college students raised in California participated in the study, designed as test-retest 5-7 days apart. Lifetime residential history and outdoors activity ("moderate" and "heavy") for each residence were extensively evaluated on both visits.

RESULTS

Although reliability of residential location decreased considerably with increasing number of lifetime residences, assigned lifetime ambient ozone concentrations from the nearest monitor yielded highly reliable cumulative values (intraclass correlation = 0.99). "Doers" of both "moderate" and "heavy" activity could be identified reliably (kappa = 0.83 and 0.93, respectively). Derived "hours per month" were reasonably correlated across visits (r = 0.73 for "moderate," 0.64 for "heavy") with 27% and 36% of the variance between visits, respectively. Reporting errors increased with increasing levels of activity.

CONCLUSIONS

Among students 17-21 years old, lifetime residential location could be repeatedly assessed. The overall performance of the set of activity questions was reasonably reliable and compares well with other measures used in epidemiology, such as blood pressure or dietary intake. Such questionnaires may be useful tools in reducing misclassification of lifetime exposure to ambient oxidants.