SMABS 2004 Jena University
SMABS 2004 Home Organization About Jena Sponsors Links Imprint / Contact SMABS Home

European Association of Methodology

Department of methodology and evaluation research

Jena University

Contributions: Abstract

Developmental outcomes for children born to mothers with PKU: Results from the international maternal PKU study

Keith Widaman
University of California at Davis
USA

Persons having the genetic defect underlying phenylketonuria (PKU) cannot metabolize phenylalanine (PHE), which leads to many negative developmental outcomes, including severe mental retardation. With appropriate diet, a person with PKU can avoid these negative outcomes. However, if a woman with PKU does not adhere to a strict diet when she is pregnant, her offspring can be negatively affected in many ways even if the offspring does not have PKU, because the maternal PHE crosses the placental barrier during gestation.

The International Collaborative Maternal PKU Study was undertaken to document the birth, infant, and child outcomes associated with prenatal exposure to phenylalanine (PHE). The participants in the study were 413 children and their mothers, selected because the mother had PKU. The analysis of individual intellectual outcomes (e.g., Bayley Mental Development Index scores at 1 and 2 years offspring age, scores on the Wechsler Intelligence Scale for Children at 7 years) required the use of two-piece linear splines, which fit the data consistently better than did linear models. These two-piece linear splines allowed the estimation of the level of exposure to PHE at which teratogenic effects occur.

The two-piece linear splines also had implications for the scoring of manifest variables included in longitudinal structural equation models of the influences among background, pregnancy-related, birth-related, and child outcome variables. The outcomes of these models suggest that pregnancy-related variables, primarily the average level of PHE in the mother's blood during pregnancy, have strong direct effects on child outcomes at 1, 2, 4, and 7 years of age, and that these variables mediate the relation of background variables (including mother's IQ) on the child outcome variables. Methodological implications of the results are stressed.