The contribution of maternal working conditions to socio-economic inequalities in birth outcome

The aims of this study were to examine the association between maternal working conditions and birth outcomes, and to determine the extent to which these contributed to class inequalities in six birth outcomes. We used an existing job exposure matrix developed from survey data collected in 1977 and 1979 to apply occupational-level information on working conditions to the national Swedish Registry, including approximately 280,000 mothers and 360,000 births during the period 1980-1985. Data were analysed using multivariate logistic regressions. Low levels of job control, high levels of physical demands and job hazards were more common in manual compared to non-manual classes. The self-employed had intermediate levels of such exposures. Job exposures, particularly low levels of job control, were generally and significantly associated with higher risks for low birthweight, very low birthweight, small for gestational age, all preterm, very preterm and extremely preterm births, but not with mortality. Compared to middle non-manuals (the reference group), lower non-manual and manual classes had higher risks for all birth outcomes, and these risks were nearly all significant. The highest odds ratios were found for skilled and unskilled manual workers in the manufacturing sector, with ratios between 1.35 and 2.66 (all significant). Job control explained a considerable proportion of inequalities in all birth outcomes. Job hazards contributed particularly to very low birthweight and extremely preterm birth, and physical demands to low birthweight and all preterm births. In conclusion, class differences in maternal working conditions clearly contributed to class differences in low birthweight (explained fraction 14-38%), all preterm births (20-46%), very (14-46%) and extremely (12-100%) preterm births. For very low birthweight and small for gestational age, there was a similar contribution in the manufacturing sector only. For all birth outcomes, class differences could still be detected after working conditions were taken into consideration. (c) 2007 Elsevier Ltd. All rights reserved.