Abstract

Washington State birth certificates were examined for associations between adverse pregnancy outcomes and paternal employment in solvent exposed occupations. Four cohorts defined by live, singleton births to fathers usually employed as auto body shop workers, painters (construction and maintenance), printers, or fibreglass workers were compared retrospectively with both a systematically selected control cohort and a low solvent exposed, occupationally defined control cohort (paternal electricians). The effects of maternal race and medical illness were controlled by sample restriction; maternal age and gravidity by stratified analysis. There was evidence of increased risk of low birth weight for infants born to fathers employed as body shop workers (relative risk = 1.6; 95% confidence interval = 1.1-2.4) or painters (RR = 1.4; CI = 0.9-2.1) when compared with the systematically selected controls but not with the electrician controls. The excess risk appeared stronger when only term infants were analysed suggesting a mechanism of growth retardation rather than prematurity. There was no evidence of increased risk among the other exposed cohorts or for other adverse pregnancy outcomes. Several design features limit the interpretation of the findings and confirmation by other studies is needed.