The effect of income inequality and macro-level social policy on infant mortality and low birthweight in developed countries - a preliminary systematic review

Objective To carry out a preliminary systematic review of literature to address the question - among rich nations (or states within nations) what is the evidence that income inequality and differences in macro-level social policy affect rates of infant mortality and low birthweight (LBW)? Study design A systematic literature review. Search strategy Medline database (1968-August 2003) was searched for empirical studies of the relationship between macro-level economic and social policies in rich nations and rates of infant mortality (IMR) and LBW. Cross-national comparison of infant mortality and LBW that did not compare the effects of macro-level economic and social policies was excluded from the review as were studies including less developed countries. Keywords representing IMR and LBW were entered into Medline along with exposures related to international comparison and macro-level policy. Abstracts obtained from the initial search were reviewed for relevant studies. Full papers of potentially relevant studies were obtained and reviewed for inclusion. Secondary search of papers cited in included papers was undertaken. For this review, papers were not excluded on the basis of quality although methodological limitations were commented on and taken into account in interpreting the results. Summary statistics were not estimated. Results Twelve studies, fulfilling the inclusion criteria, were identified. Ten studies examined the association of IMR with income inequality, eight of which reported a statistically significant positive association with higher levels of inequality after adjustment for a range of variables. Six studies reported significant positive associations of IMR with other indicators of less re-distributive social and economic policy. Associations with LBW were reported in four studies; three showed significant positive associations with higher levels of income inequality and one showed no association with low levels of parental leave entitlement. Methodological differences, particularly the wide range of variables used to adjust for confounding, make interpretation of the findings difficult. Conclusions The results of this review represent a preliminary attempt to summarize the literature linking macro-level economic and social policies in rich nations with IMR and LBW. The findings, taking account of the methodological limitations of the review and of the included studies, suggest a statistically significant association between IMR and higher income inequality and other indicators of less re-distributive social policy. Only three studies examined the association of income inequality with LBW and, although they suggest a significant association, further studies will be needed to confirm this finding.