Maternal risk factors predicting child physical characteristics and dysmorphology in fetal alcohol syndrome and partial fetal alcohol syndrome

Abstract

Background: Previous research in South Africa revealed very high rates of fetal alcohol syndrome (FAS), of 46-89 per 1000 among young children. Maternal and child data from studies in this community summarize the multiple predictors of FAS and partial fetal alcohol syndrome (PFAS).

Method: Sequential regression was employed to examine influences on child physical characteristics and dysmorphology from four categories of maternal traits: physical, demographic, childbearing, and drinking. Then, a structural equation model (SEM) was constructed to predict influences on child physical characteristics.

Results: Individual sequential regressions revealed that maternal drinking measures were the most powerful predictors of a child's physical anomalies (R² = .30, p < .001), followed by maternal demographics (R² = .24, p < .001), maternal physical characteristics (R²=.15, p < .001), and childbearing variables (R² = .06, p < .001). The SEM utilized both individual variables and the four composite categories of maternal traits to predict a set of child physical characteristics, including a total dysmorphology score. As predicted, drinking behavior is a relatively strong predictor of child physical characteristics (β = 0.61, p < .001), even when all other maternal risk variables are included; higher levels of drinking predict child physical anomalies.

Conclusions: Overall, the SEM model explains 62% of the variance in child physical anomalies. As expected, drinking variables explain the most variance. But this highly controlled estimation of multiple effects also reveals a significant contribution played by maternal demographics and, to a lesser degree, maternal physical and childbearing variables.

Figure 1

Hypothesized model of children’s characteristics predicted from maternal characteristics. Asterisks are parameters to be estimated. D* indicates disturbance (error) for the dependent latent variable. All residual variances for measured variables (E*) also are to be estimated (not shown).

Figure 2

Final model of children’s characteristics predicted from maternal characteristics. Unstandardized and standardized (in parentheses) parameter estimates. Residuals for measured variables are not shown.