RESUMO
In order to compare pregnancy smoking habits in two generations, data on first generation deliveries in 1964-67 were obtained from prospective questionnaires. By record linkage with the Swedish Medical Birth Registry, data on daughters giving second generation births (n = 1659) were obtained, including prospective smoking information. The catchment area was that of Helsingborg County Hospital (population 145,000). Second generation births until December 31, 1993 were identified in all of Sweden. The odds ratio of maternal smoking in the first generation for pregnancy smoking among the daughters was significantly increased and approximately doubled. Adjustment for confounders was made using the Mantel-Haenszel method. Similar odds ratios remained after stratification either for a social environment index of the first generation or the educational level of the second generation. Thus, our data suggest a biological association between pregnancy smoking habits over generations. Research on causative mechanisms could help design more efficient intervention programs. In a separate analysis of two parturient cohorts in Helsingborg 30 y apart, we found that the prevalence of maternal smoking had decreased comparatively more among younger than older women.
