Socio-demographic characteristics of UK families using pesticides and weed-killers

Selection bias in case–control studies on household exposure to pesticides and childhood acute leukemia

“…Yet, residual confounding could occur after adjusting for some markers of socioeconomic status. On the basis of a relatively scarce literature, it is noteworthy that groups using more pesticides (Steer and Grey, 2006) appear to have the same characteristics as people more willing to participate in epidemiologic studies (Richiardi et al, 2002;Galea and Tracy, 2007;Mezei et al, 2008;Shen et al, 2008) such as being of rather higher socioeconomic status and living in more rural areas. If this pattern applied to controls in the reviewed studies, users of pesticides would have tended to be more willing to participate than non-users, and odds ratios would have been somewhat underestimated as compared to the true odds ratios (assuming the participation in exposed cases was similar to that in unexposed cases).…”

“…We assumed that the odds ratio for association between household exposure to pesticides and participation status could vary from 2.0 to 0.5, for both cases and controls. This range corresponds to the magnitude of the associations observed between household exposure to pesticides and sociodemographic factors (reduced to dichotomous variables) in the literature (Steer and Grey, 2006). An odds ratio for association between household exposure to pesticides and participation status greater than one means that participation in exposed subjects is higher than among unexposed subjects.…”

“…Two CAL case-control studies Rudant et al, 2007) reported higher rates of use in rural areas than urban areas, in controls. The proportions of pesticide use also differed across surveys (Steer and Grey, 2006) and between control groups in the different CAL casecontrol studies, indicating differences in use between countries, regions, and calendar periods.…”

“…Socio-demographic differences between the controls providing data and targeted controls for selection in the study base, can lead to selection distortion assuming these factors are associated with exposure, as was observed for household pesticides. A large survey recently carried out in the UK on 13,391 pregnant women investigated household exposure to pesticides during pregnancy and the four following years (Steer and Grey, 2006). They observed that the independent factors related to pesticide use were an older age at the time of pregnancy, a higher parental educational level, owning the housing tenure, and being white.…”

Impact of population characteristics on Illinois childcare adoption of integrated pest management (IPM)

“…Higher socioeconomic status (home ownership, higher income and education, non-manual social class) and age of mother correlated with high pesticide use in UK homes, but was a poor predictor (weak and inconsistent correlations) of pesticide storage and use in a US study of Minnesota families with children (Adgate et al 2000;Grey et al 2005;Steer et al 2006). High levels of pests and pesticide use are associated with gardens, older housing, home disrepair and routine extermination practices (Whyatt et al 2002;Steer et al 2006;Jacobs et al 2009). Household socio-demographic characteristics explained little of the variability in pesticide storage and use patterns.…”