Diabetes Mellitus

A survey of the literature gives no convincing evidence that the disease is caused by bacteria. It seemed possible that it may be caused by a filterable virus.

Experiments were made on rabbits to determine whether it is possible to obtain evidence of the presence and action of a filterable virus. It seemed probable that the causative organism is thrown off from the patient's body in the urine. The urine of diabetic patients was therefore passed through Berkefeld filters, and the filtered urine was injected into the circulation of rabbits in doses of 2 cc. Two rabbits were injected with the urine of each patient, each animal receiving only one injection.

After the rabbits had been injected they were confined in metabolism cages which permitted the collection of the urine. Daily analysis of urine was carried out, using Benedict's quantitative reagent.

Diabetes mellitus is generally a slowly progressing disease and it was believed that if it is transmissible to rabbits evidences of infection would probably be delayed for some time.

Experiment 1. The first pair of rabbits was injected with the filtered urine of patient Q on January 12, 1926. Almost daily urine examination failed to show the presence of glucose until 22 days later, when a trace of sugar was found in the urine of rabbit No. 1 (female). Rabbit No. 2 (male) first showed glucose in the urine on the 42nd day after inoculation.

A peculiar phenomenon in the experimental glycosuria in rabbits has been the irregular periodicity of the occurrence of glucose in the urine. After glucose was demonstrated for several successive days it disappeared, but reappeared more or less irregularly. After glucose had been demonstrated for several months in the urine of these 2 rabbits they were taken out of the metabolism cages.