It seems quite likely that the normal process of fetal lung biochemical maturation is delayed by maternal diabetes and that abnormalities in the pulmonary surfactant system are involved. The appearance of PG in amniotic fluid and possibly in fetal lung is impaired or at least delayed. The same is possibly true for DSPC, the main constituent of surfactant, but recent discrepant data call for further clarification of this specific point. Careful determination of the fetal lung phospholipid profile by amniotic fluid analysis helps predict and prevent RDS in IDM, along with a careful control of the maternal diabetic condition. A study of alveolar surfactant at birth, if it could be performed in addition to amniotic fluid analysis, would help to better characterize surfactant deficiency in IDM. On the basis of both in vivo and in vitro experimental approaches, it seems clear that hyperglycemia and fetal reactional hyperinsulinism are both involved in the processes delaying fetal lung maturation. Further advances in the understanding of cellular and molecular mechanisms leading to this delay will be conditional on the availability of animal models reproducing the features of the metabolic and hormonal environment of human fetuses in diabetic pregnancies. The appropriateness of in vivo models needs to be defined by two kinds of criteria: 1) presence of simultaneous hyperglycemia and hyperinsulinemia in the fetus; 2) the presence of delayed fetal lung maturation as judged by morphology and morphometry of epithelial lung cells, by physiological assessment of surfactant, and by the phospholipid composition of the lung (and including lung tissue per se, bronchoalveolar lavage fluid, lamellar bodies, and/or isolated surfactant fractions). Therefore, future studies must necessarily be comprehensive in scope and include information indicating that fetal growth, blood glucose, and circulating insulin are all increased. Such models already exist in rats and rabbits. Rat models are possibly not the best because of the high basal level of fetal blood insulin in this species and the relatively rapid rate of lung maturation that is not analogous to the human. Monkey models are of interest, because of their close relationship with the human pregnancy, and need to be studied further. They are particularly attractive also because primary fetal hyperinsulinism can be studied (268), as well as the combination of hyperglycemia and hyperinsulinemia in pregnancies of STZ-treated monkeys (152). An appropriate model of the diabetic pregnancy could provide answers to the following questions. Are the biosynthetic pathways of surfactant phospholipids directly impaired?(ABSTRACT TRUNCATED AT 400 WORDS)