Gestational diabetes and impaired glucose tolerance in pregnancy were found to be important teratogenetic risk factors for the development of diabetes in the offspring. Mechanisms of action and prevention of maternofetal transmission of teratogenetic susceptibility to diabetes are presented. Gestational diabetes induced in the F0 generation produced the following effects in the F1 and/or F2 generation: Early postnatal hyperinsulinaemia, decreased noradrenaline and serotonin and increased endorphin concentrations in specific brain regions, permanent hypoplasia of the hypothalamic ventromedial nuclei, decreased insulin responsiveness to glucose, impaired glucose tolerance and increased diabetes susceptibility.

In view of experimental and clinical findings, it was predicted (Dörner, 1973 and 1976; Dörner and Mohnike, 1973 and 1977) that a preventive therapy of diabetes mellitus may be possible by preventing hyperinsulinism in perinatal life by means of prevention of hyperglycaemia in pregnant women and overnutrition in newborns. Meanwhile, this prediction appears to have been realized. Thus, the prevalences of diabetes mellitus in children, who were born in Berlin/GDR over the past decade, were found to be significantly decreased as compared to those born between 1962 and 1972. On the other hand, the children born in Berlin/GDR between 1962 and 1972 displayed a significantly higher prevalence of diabetes mellitus as compared to those born between 1957 and 1961. The decreasing prevalence of childhood-onset diabetes over the past decade has been apparently achieved by systematic prevention of hyperglycaemia and impaired glucose tolerance in pregnant women and overnutrition in newborns. These findings suggest that a preventive therapy of diabetes mellitus--even of insulin-dependent childhood-onset diabetes--is possible by preventing hyperinsulinism in the foetus and newborn during differentiation and maturation of the neuroendocrine central nervous-pancreatic system.

A highly significantly decreased prevalence of insulin-dependent childhood-onset diabetes (less than 1/3 of the initial prevalence rate) could be achieved in Berlin/GDR since 1973 by improving systematically diagnostic and therapeutic measures for pregnant diabetics, particularly for non-insulin-dependent gestational diabetics. In addition, a highly significantly increased incidence rate of diagnosed, diet-treated and delivered non-insulin-dependent pregnant diabetics was found between 1979 and 1983 in Berlin/GDR, Halle and Leipzig as compared to the other districts of the GDR. Simultaneously, a highly significantly decreased prevalence rate of diabetic children (less than 1/3), who were born during this period, was found in 1983 for Berlin, Halle and Leipzig as compared to the other districts of the GDR. Finally, a highly significant inverse correlation could be demonstrated for the 15 districts of the GDR between the incidence rates of diagnosed, diet-treated and delivered non-insulin-dependent pregnant diabetics and the prevalence rates of diabetic children who were born during this period (1979-1983). In view of these findings, an interruption and even a reversal of the continued dramatic increase of idiopathic insulin-dependent diabetes mellitus appears to be possible in the developed countries by preventing maternal hyperglycaemia during pregnancy and hence hyperinsulinism in the foetuses and newborns.

We studied glucose concentration of breast milk of nursing diabetic mothers and its possible relationship to the quality of metabolic control. Eleven Type 1 (insulin-dependent) diabetic mothers and 11 age-matched control subjects were included in the study. Although a near-normoglycemic control of diabetic mothers was accomplished by intensified insulin treatment, the HbA1 value was significantly higher in comparison to non-diabetic mothers (8.1 +/- 0.9% versus 6.2 +/- 0.5%; p less than 0.01). Regardless of this, the glucose concentration of breast milk did not differ between diabetic mothers and that of non-diabetic women (0.68 +/- 0.50 mmol/l versus 0.66 +/- 0.55 mmol/l). No correlation exists between glucose concentration of breast milk and relevant blood glucose concentration as well as glycosylated haemoglobin A1 of the mother. In conclusion, our data support the view that breast-feeding of infants of diabetic mothers is not associated with an increased offer of glucose and thus not being of importance as a possible mechanism to sustain a hyperinsulinemic state in the newborns.