Cigarettes, alcohol and marijuana: varying associations with birthweight.

The relations of cigarette smoking, alcohol drinking and marijuana use during pregnancy to birthweight were examined in two prospectively studied pregnancy cohorts (Phases I and II). After analytic adjustment in ordinary least squares regressions for other factors that influence birthweight, cigarette smoking during at least half the pregnancy was associated with a significant decrease in mean birthweight (159 grams in Phase I, 202 grams in Phase II). In Phase II only, beer drinking was associated with a significant decrement of 8.4 grams in estimated birthweight per ounce of absolute alcohol per month. Neither wine nor liquor drinking in the Phase II data nor any of the three beverages in the Phase I data was associated with significant decrements in predicted birthweight. Furthermore, with one exception (drinking once a week in Phase II only), alcohol drinking, defined as the number of occasions per month on which any alcoholic beverage was consumed, was not associated with a change in birthweight. Regarding marijuana use, the data are not consistent between the two phases. In the Phase I data, no coherent trend in association with birthweight was observed. In the Phase II data, marijuana use 2-3 times per week, 4-6 times per week and daily was associated with increasing decrements in estimated birthweight: 127 g, 143 g and 230 g respectively. The inconsistent findings for alcohol drinking and marijuana use between phases stand in marked contrast to the consistent findings for cigarette smoking.