Definition of congenital malformations and detection of associations with maternal factors.

Data collected from a cohort of 2913 pregnancies were analyzed to determine whether the varying definitions of congenital malformations influence the statistical result in the search for possible etiological factors. According to three different definitions of congenital malformations, three study groups were formed, i.e. structural malformations, all congenital defects, and all disorders or abnormalities with possible prenatal etiology. These consisted of 93, 197 and 334 cases, respectively. The control group consisted of 2579 cases. Positive associations were noted between the study groups and some factors concerning pregnancy, and the social history of the parents. The associations were strongest in the group of structural malformations and became weaker when this group was 'diluted' by other defects with questionable prenatal origin. Hence, when comparisons for evaluating the causality of significant associations in different materials are made, the definition of the concept of congenital malformations should be taken into consideration.