ARCHIVED - Early primary school outcomes associated with maternal use of alcohol and tobacco during pregnancy and with exposure to parent alcohol and tobacco use postnatally

1. Introduction

This secondary data analysis was commissioned by the Fetal Alcohol Spectrum Disorder Initiative, Public Health Agency of Canada. The data analyses used the Better Beginnings, Better Futures (BBBF) longitudinal database. The BBBF longitudinal study began in 1993. Over 500 children born in 1994 in six disadvantaged neighbourhoods across Ontario were recruited for the longitudinal study at birth; over 400 remained in the cohort at 8 years of age (Grade 3).

Children who are prenatally exposed to alcohol and tobacco have been found to be at risk for a range of adverse health and developmental outcomes from infancy into adulthood (Huizink & Mulder, 2006; Richter & Richter, 2001). The main purpose of the present study was to examine relationships between prenatal and postnatal exposure to alcohol and tobacco separately and in combination on developmental outcomes in young children over the first four years of primary school.

Measures in five domains of child development outcomes were analyzed, including general development, cognitive development/academic performance, social/emotional functioning, physical health, and externalizing and internalizing behaviour problems.

Two sets of analyses were used. First, analysis of covariance (ANCOVA) was used to determine whether prenatal exposure to alcohol and/or tobacco may have differential effects on these various aspects of children’s functioning during the early primary school years. Based on results from the ANCOVA, more complex statistical techniques (structural equation modelling [SEM]) were used to examine the pathways from prenatal and postnatal exposure to alcohol and tobacco smoke, to parent and teacher reports of children’s behaviour problems at age 8 (Grade 3). In this path analysis, we focused on externalizing and internalizing behaviour problems. Given the large array of variables available and the complexity of SEM, we needed to reduce the scope to make SEM feasible. We decided to begin with an array of 12 variables that allowed us to use existing research literature to develop a confirmatory model of the behaviour problem measures.

The BBBF longitudinal dataset was made up of over 400 children. These children and their families were recruited from disadvantaged Ontario communities at birth, and were followed prospectively at 33 and 48 months, and again at age 8. Thus, it was also possible to measure postnatal exposure to alcohol (i.e. maternal drinking) and tobacco (i.e. second-hand or environmental smoke), and to examine whether any negative effects of prenatal exposure to alcohol and tobacco on children’s developmental outcomes increased or decreased over a four-year period between 4 and 8 years of age.

Although previous studies have found evidence for an association between prenatal substance exposure and adverse effects on children’s development and functioning, many other studies have used clinical samples with very high prenatal substance exposure. In contrast, this is a large community-based sample of children who were followed prospectively from 3 months to 8 years of age. A strength of the prospective longitudinal study design is the ability to control for many familial and demographic factors.

Maternal alcohol use was assessed using the CAGE questionnaire (Ewing, 1984), while maternal tobacco use was assessed with questions from the National Longitudinal Survey of Children and Youth (NLSCY) and other population surveys.

Because associations between smoking during pregnancy and child outcomes may be due to more than just nicotine, for consistency, we use the term “prenatal tobacco exposure” in this report to refer to the effects on offspring of maternal smoking during pregnancy.

In the following section, we review some of the current research literature on the effects of prenatal exposure to alcohol and tobacco on children’s health.