The relationship between population characteristics, low birth weight, cigarette smoking in pregnancy, and public health messages discouraging their use: An historical perspective

This research was designed to add to the understanding of the contribution of maternal characteristics and public health messages, (specifically, the Surgeon General's Warning of 1969 and the Surgeon General's Report of 1973) to cigarette smoking in pregnancy and low birth weight. It was designed to provide an historical perspective on these issues as the 1960s and 1970s were the years when Surgeon General Reports were beginning to be released.;The study involved a secondary analysis of data from 2,937 women and infants in the Atlanta Birth Defects Case Control Study (1968--1980). The prevalence and magnitude of smoking and the prevalence of low birth weight were modeled as a function of time using a logistic regression with linear splines. This model allowed for changes in the log odds of prevalence and magnitude of smoking and prevalence of low birth weight over time. Logistic regression tests were conducted to determine if the changes in the model were statistically significant, especially in the selected time periods (1968 to 1969 and 1972 to 1973). Likelihood ratio tests were used to determine if there were statistically significant differences in the patterns of smoking prevalence and low birth weight comparing women by race and age groups.;The results showed statistically significant reductions in the study population's prevalence of cigarette smoking during pregnancy associated with the 1969 and 1973 public health messages (36% and 40%, respectively). There were different patterns of changes in behavior, depending on group membership. Main decreases occurred in Black women <25 years of age and White women ≥25 years of age. The majority of women who smoked were heavy smokers and there was little change in the magnitude of smoking over time. There were no statistically significant reductions in low birth weight associated with the selected public health messages (1% and 0%, respectively). The changes in cigarette smoking behavior and low birth weight were not sensitive to adjustments for education, number of pregnancies, or family income.;This study showed that the prevalence of cigarette smoking in pregnancy may be influenced by some public health messages. Some population sub-groups may be more influenced by these messages than other groups. Better mechanisms to study the effectiveness of public health messages in various population groups need to be developed so that these messages can be better tailored to affect all women in their reproducing years.