Smoking as an independent variable in a multiple regression analysis upon birth weight and gestation.

Abstract

In most of the published reports relating to smoking and birth weight, the analysis of data consisted mainly of a comparison of the mean birth weight of infants of smokers, specific for one, or possible 2 variables, such as race, maternal age, parity, height, or social class. Using multiple regression techniques, the present paper investigates the relationship of smoking and birth weight, and smoking and gestation, after simultaneous adjustment for the effects of 15 concomitant variables relating to the mother and infant. The Baltimore Study data were used in the analyses. Smoking was found significant in the analysis of birth weight, the predicted birth weight of infants of smokers being 129 g less than for infants of nonsmokers, after adjustment for the concomitant variables. Smoking was found non-significant in a similar analysis of gestation. These findings derived after due consideration of multiple characteristics of the mother confirm those reported in the original analysis of the Baltimore Study. The relationship of the concomitant variables of birth weight and gestation is presented and implications for future research are discussed.