Observations on the mechanism of increased weight loss during metformin administration in obesity

In two hospitalized obese, non-diabetic women on a 1250 calorie diet, the administration of metformin increased the rate of weight loss. A rebound effect was observed on withdrawal.

These phenomena could not be correlated to changes in the total weight, dry weight or fat content of faeces or to variations in fluid intake and diuresis. Although faecal losses of fat and dry matter increased with the dose of metformin, the degree of malabsorption was too small to account for any significant part of the weight loss.

The administration and the withdrawal of metformin produced changes in body weight similar to those provoked by restriction of calories, or selective carbohydrate restriction and re-alimentation, respectively.

Since such rapid shifts in weight are largely due to changes in water-balance, it is highly probable that metformin reduces the amount of body water. The mechanism involved may concern a decrease in carbohydrate metabolism as well as an increase in that of fat.

Insulin and glucose in plasma from umbilical vein and heel blood of newborn infants of diabetic women

The concentrations of insulin and glucose were estimated simultaneously in plasma from umbilical vein and from heel-stab blood of newborn infants of diabetic and non-diabetic women in the fasting state and after glucose injection into the umbilical vein.

In infants of diabetic mothers a rapid and high increase in insulin concentration was seen in plasma from the umbilical vein and also, although to a lesser extent, from the peripheral blood. In infants of non-diabetic mothers a slow rise was seen in umbilical vein plasma, but no rise at all in the peripheral blood.

The plasma glucose concentration in umbilical vein as well as in heelstab blood rose and fell without any difference being found between the two groups of infants, in contrast to the great differences in their insulin concentrations.

Insulin, insulin antibody and glucose in plasma of newborn infants of diabetic women

The venous plasma insulin and glucose concentration was investigated in 15 diabetic and 13 non-diabetic women at birth; in the umbilical cord blood of their infants at birth; 1.5 to 6 hours after birth in the fasting state; and 5 and 30 minutes after a glucose load injected through a catheter inserted in the umbilical vein. Insulin antibodies were estimated at birth.

Antibodies were found in the maternal and the umbilical cord blood in 13 diabetic mothers who had had insulin treatment for 2–20 years, but not in 2 diabetics with, respectively, no insulin treatment, and treatment for a few days only.

In infants of non-diabetic mothers, the insulin concentration at birth was lower than that of their mothers. In infants of diabetic women the insulin concentration was much higher than that of the control infants. After glucose injection a rapid increase of the insulin concentration in the umbilical vein was demonstrated in infants of diabetic women, whereas the rise was slow in infants of non-diabetics, thus indicating a greater reactivity of insulin secretion or release in infants of diabetic women.

After glucose injection, the plasma glucose concentration in the umbilical vein rose and fell rapidly without any difference being found between the two groups, in contrast to the great differences in their insulin concentration.

Foetal mortality in pregnancy of diabetics. Treatment by one team during pregnancy

A series comprising 90 diabetic women with 99 pregnancies and 102 infants, from the 19 year period 1946–1964 has been analysed. The series includes all women whose first attendance in the department was at least 150 (150–250) days before the calculated term and who continued their pregnancy till the infants weighed 1000 g or more. There was a preponderance of severe cases.

The perinatal mortality was 14 deaths or 14 per cent; however, in 12 class-F cases there were 8 deaths, and in 90 non-F cases 6 deaths or 7 per cent. By comparison with the results of previous pregnancies in these women the series shows a very favourable progress in non-F cases. The new PBSP system of classification according to prognostically bad signs during pregnancy has been used and is of value. The PBSP complications seen were toxaemia and hyperpyretic pyelitis only.

In the 69 non-F cases also without any PBSP complication the foetal mortality was 4 per cent.

The investigation and treatment of latent and manifest urinary tract infection should be reinforced.

At present the therapeutic result in F cases is so poor that induced abortion and sterilisation should be advised more often.

Prognosis of the outcome of pregnancies in diabetics. A new classification

In order to improve the possibilities of predicting the outcome of pregnancy in diabetics, we analysed a consecutive series of 304 pregnancies in 263 diabetic women in the Royal Maternity Department B, Rigshospitalet, Copenhagen, during the 5-year period 1959-1963. The period of supervision and treatment during pregnancy varied greatly. The perinatal mortality in the 306 infants was 17.9%.

According to the results of this analysis, patients with a poor prognosis were divided into the following four groups: Pregnant women who developed (1) hyperpyretic pyelitis, (2) pre-coma or severe acidosis, (3) toxaemia, or patients who could be so described, (4) »neglectors« These four groups, and the classification, are designated PBSP (Prognostically Bad Signs during Pregnancy).

The mothers of 130 infants belonged to one or more of the four groups of PBSP, and among the infants of these mothers the perinatal mortality was 31.5% as compared with 7.9% in the group of 176 infants without PBSP during pregnancy.

The poor prognosis for the pregnancies with PBSP applies to all foetal weight groups. In addition, these pregnancies terminate in premature delivery twice as often as the others. It is demonstrated that from the prognostic point of view, nothing is gained by including hydramnios in PBSP.

The risk involved by a PBSP complication to the foetus depends on the White (1949) class in which it occurs. A combination of White's classification of pregnant diabetics with regard to foetal prognosis used together with the present classification improves the possibility of predicting the foetal prognosis in a series of pregnant diabetics which is mixed as regards the length of treatment during pregnancy.