[Parents' education and infant mortality 1967-1998].

Background: We have examined the association between socioeconomic status and risk of infant death in Norway between 1967 and 1998.

Methods: Information from the Medical Birth Registry of Norway on all live births and infant deaths was linked to information from Statistics Norway on parents' education. There were 1,777,364 eligible live births and 15,517 infant deaths. Differences between educational-attainment groups were estimated as risk difference, relative risk, population-attributable fraction, and index of inequality ratio.

Results: The risk of infant death decreased in all educational-attainment groups and the level of education increased over time. For neonatal (0-27 days of life) death the risk difference between infants whose mothers had high or low education was reduced from 3.5/1000 in the 1970s to 0.9/1000 in the 1990s. The inequality ratio declined from 1.72 to 1.32 and the population-attributable fraction from 22.3 to 8.4. For risk of postneonatal (28-364 days of life) death, the difference between infants whose mothers were in high or low education brackets increased from 0.7/1000 in the 1970s to 2.0/1000 in the 1990s. The inequality ratio went up from 1.31 to 4.00 and the population-attributable fraction from 9.7 to 39.5.

Interpretation: Since the late 1960s, infant mortality has decreased and parental educational levels have risen. There is a higher degree of social equality with regard to risk of neonatal death, while the opposite holds for postneonatal death.