Comparison of Self-Reported Pest Treatments and Insecticide Concentrations in Carpet Dust in the Northern California Childhood Leukemia Study

Abstract

Background: Parental self-reported pesticide use has been the usual exposure assessment method in studies of pesticide exposures and childhood leukemia. This approach does not identify active ingredients and may suffer from recall bias. Measuring pesticides in carpet dust is an objective method for determining chemical concentrations in homes and may be a useful exposure indicator for children. Methods: In the Northern California Childhood Leukemia Study, parents of 277 cases and 306 controls were interviewed about household treatment (yes/no) in the prior 12 months for flying insects, fleas/ticks, termites/carpenter ants, lawn/garden insects, and professional indoor and outdoor insect treatments. Carpet dust samples were collected in 2001-2007 and analyzed for nine insecticides using gas chromatography/mass spectrometry. Using multivariable linear regression, we modeled the relationship between log-transformed concentrations of the insecticides and pest treatments, adjusting for home characteristics and other factors. Results: Among both cases and controls, reported flea/tick treatment was associated with 2 to 3- fold higher concentrations of cypermethrin, permethrin and piperonyl butoxide compared to households not treating fleas/ticks (p<0.05). In both groups, lawn/garden insect treatment increased chlorpyrifos concentrations by 1.6 times (p<0.05), and professional outdoor applications increased cyfluthrin 4-fold in cases and 10-fold in controls (p<0.05). For the most part, associations between reported treatments and insecticide measurements were consistent between cases and controls, suggesting minimal effect of recall bias. Conclusions: The statistically significant associations between self-reported pest treatments and insecticide concentrations in carpet dust, which were generally consistent between cases and controls, lend credibility to the use of both methods for characterizing residential exposure to pesticides.