Trajectories of prenatal alcohol exposure and behavioral outcomes: Findings from a community-based sample.

Abstract

To characterize patterns of prenatal alcohol exposure (PAE), and determine whether PAE trajectories were associated with behavior from a community-based sample of first-grade children. Using data collected as part of the Collaboration of Fetal Alcohol Spectrum Disorders Prevalence study (n = 1663), we performed longitudinal cluster analysis on prenatal alcohol use reported for four time points around conception and pregnancy. From the sample, 638 respondents reported any alcohol use in pregnancy and were included in trajectories for average daily and maximum drinks per drinking day (max DDD). We then estimated the association with behavioral problems measured by the Child Behavior Checklist (CBCL) and Teacher Report Form (TRF) with multivariable linear regression. The reference group had 1025 children with no reported PAE. Five trajectories were selected to describe max DDD patterns: very low/discontinuing (n = 186), low/discontinuing (n = 111), very low/continuing (n = 47), med/high (n = 245), and high (n = 49). Six trajectories best described average daily alcohol use: very low/discontinuing (n = 378), very low/continuing (n = 98), low/continuing (n = 56), low/discontinuing (n = 37), medium/high (n = 35), and high (n = 31). When assessing max DDD trajectories for both the CBCL and TRF, individuals with PAE in the two highest trajectories and the very low/continuing trajectory had more behavioral problems relative to children with no PAE, although confidence intervals for most estimates included the null. PAE modeled as average drinks per day did not predict behavior in any consistent pattern. In this community-based sample, select PAE trajectories were associated with behavior, even at relatively low levels of PAE that continued later in gestation.