Little is known about patterns of change in alcohol consumption and predictors of these patterns over the prenatal to postnatal period.

To determine trajectories of maternal alcohol consumption before and after pregnancy and predictors of these trajectories.

Materials and Methods
A total of 6597 Australian women were sampled from a longitudinal study. Group‐based trajectory modelling was applied to determine drinking trajectories from prepregnancy, early pregnancy, late pregnancy and 6 months after the birth. Predictors associated with drinking trajectories were examined using multinomial logistic regression.

Three trajectories of maternal alcohol consumption were identified: abstainers/minimal consumption (53.2%), light consumption (39.4%) and heavy consumption (7.4%). The heavy consumption group substantially reduced their consumption in pregnancy but increased their consumption once the baby was born. Some 80.0% of this group were breastfeeding their babies. The light consumption group had only minor changes in their drinking pattern. Lower family income, being married, high frequency of church attendance, low level of adversity, poor health lifestyle, remaining married to original partner and having many children predicted membership of the abstaining/minimal consumption trajectory. Being unmarried, having only one child, having unhealthy health lifestyle and never going to church predicted membership of the heavy consumption group.

Women who consume higher levels of alcohol prior to their pregnancy reduce their consumption once pregnant, but tend to increase their alcohol consumption shortly after the birth. A public health campaign dealing with predictors associated with heavier alcohol consumption and safe breastfeeding targetted at these women is needed.