BACKGROUND: International ecological studies have shown a positive association between infant mortality as a proxy for low birth weight and cardiovascular disease mortality in adult life. METHODS: Mortality rates due to Cardiovascular Diseases (CVDMR) standardised by age in adults between 45 and 69 years of age and by place of birth (pob) and residence (res) were related to Infant Mortality Rates (IMR) in the Brazilian 1935 birth cohort. RESULTS: Two relationship patterns were noted between IMR and CVDMR: for the Southeast, South and Centre-West group of regions (r(pob)=0.46; r(res)=0.29) and for the North and Northeast group of regions (r(pob)=0.21; r(res)=0.33). For the latter pattern, two states were identified (Rio Grande do Norte and Paraíba) as atypical areas, whose exclusion strengthened the association (r(pob)=0.73; r(res)=0.91). CONCLUSIONS: The direction of the associations changed after the analysis by group of Brazilian regions (indirect control of socio-economic levels, coverage and quality of the information). There is a positive, although weak association between IMR and CVDMR. Attempts to control or minimise the interference of migratory movements, cohort effects and socio-economic levels represented methodological progress in ecological analyses of foetal programming in Brazil.
BACKGROUND: International ecological studies have shown a positive association between infant mortality as a proxy for low birth weight and cardiovascular disease mortality in adult life. METHODS: Mortality rates due to Cardiovascular Diseases (CVDMR) standardised by age in adults between 45 and 69 years of age and by place of birth (pob) and residence (res) were related to Infant Mortality Rates (IMR) in the Brazilian 1935 birth cohort. RESULTS: Two relationship patterns were noted between IMR and CVDMR: for the Southeast, South and Centre-West group of regions (r(pob)=0.46; r(res)=0.29) and for the North and Northeast group of regions (r(pob)=0.21; r(res)=0.33). For the latter pattern, two states were identified (Rio Grande do Norte and Paraíba) as atypical areas, whose exclusion strengthened the association (r(pob)=0.73; r(res)=0.91). CONCLUSIONS: The direction of the associations changed after the analysis by group of Brazilian regions (indirect control of socio-economic levels, coverage and quality of the information). There is a positive, although weak association between IMR and CVDMR. Attempts to control or minimise the interference of migratory movements, cohort effects and socio-economic levels represented methodological progress in ecological analyses of foetal programming in Brazil.