Diabetes Mellitus is rare in pregnancy, but when it does occur the disease is fatal in three-fourths of the children, and it hastens the death of the woman, according to the common opinion of obstetricians, but this opinion is disputed. In making the diagnosis we must exclude lactosuria and other pseudodiabetic conditions. A sugar reaction which is often mistaken for the glycosuria of true diabetes is from lactose in excessive milk secretion. This lactosuria is harmless. Again, when women are taking tonics or cough mixtures containing derivatives of wild cherry their urine may give a sugar reaction from the phloridzin of the wild cherry. The phloridzin so acts on the epithelium of the kidneys that it lets the blood-sugar escape into the urine. Medical writers who report diabetes in large numbers of pregnant women mistake these reactions for the reaction of true glycosuria.

In the genuine diabetes of pregnancy there is a high mortality. Offergeld,[200] in sixty cases, found that the women died within two and a half years, and that 76 per cent. of the children were lost. Diabetics commonly are sterile from atrophy of the uterus and ovaries: in a series of 114 diabetic married women, LacorchÉ found only seven pregnancies. In a third of such as do become pregnant abortion or premature labor occurs. Coma happens in 30 per cent. of these pregnancies, and it is almost always fatal. Delivery frequently causes collapse, coma, or sudden death. The liver in any gestation has more work than it has in the unimpregnated state, but a diabetic liver is unfit for almost any normal function. If albuminuria is found the prognosis becomes very bad. Diabetic women have poor resistance against a tubercular infection. Half their children are still-born, and 10 per cent. more die within a few days after birth (many of these children are diabetic).

There is some evidence of heredity in diabetes—it is likely "to run in a family." Heiberg[201] reported one family in which five of thirteen children had diabetes; in another, four of eight children, the mother, two of the mother's brothers, and the maternal grandfather had diabetes. In another, two brothers, the father, and grandfather died of it. I know of a case where the only two sons and the father in a family died of it. Heiberg did not find any essential difference in the histology between the hereditary cases and those which were not hereditary.

Joslin[202] reported seven cases of diabetes associated with pregnancy. Four of the seven are now dead, one by suicide, one from uremia, one from coma, one from tuberculosis. Of the three living one is in good condition, one is not well and she has lost two of three children, and one is very ill with diabetes. In persons beyond middle age diabetes with proper treatment may go on for from ten to fifteen years before it is fatal, but it quickly kills young patients. A young woman at the marriageable age who has diabetes will die anyhow in two or three years, and if she marries and becomes pregnant she will die very probably in her first pregnancy.

When the child is viable, and the diabetic mother shows albuminuria, progressive weakness, or diacetic acid in marked quantity, it may be necessary to perform therapeutic abortion; but if this is done no anesthetic may be used, and great precautions should be taken to avert physical and mental shock. Even ergot acts badly with these cases. The last sacraments should be given in good time, especially if coma threatens. When labor begins in a diabetic and everything appears to be normal the sacraments should be given, because there is always danger of sudden collapse and death.