A Prognostic Study of Insulin Treatment in Schizophrenia

From a follow-up, five to ten years after hospital admission, of about 500 schizophrenic patients, about half of whom had been treated by insulin coma therapy and the other half not so treated, the following conclusions emerged: 1. Patients with no family history of mental illness had a lower chance than others of a spontaneous remission; this disadvantage was nullified by insulin treatment, under which patients with all types of family history did almost equally well, and better than controls. 2. Spontaneous remission rates drop with increasing age of onset, while rates of remission with insulin treatment remain much more steady; as with type of family history, the prognostically least favourable group derives special benefit from insulin treatment. 3. Spontaneous remission rates are better in persons of pyknic and athletic physique than in leptosomes and dysplastics; insulin remission rates vary little with physique. Again it is the prognostically least favourable group which has most to gain from treatment. 4. Patients with a good personality did better than others, in both insulin and control groups; insulin treatment did not nullify the relative disadvantage of a poor personality. 5. Insulin treatment provided only a slight improvement on control expectations in cases with psychogenic precipitation, but a very marked improvement in cases with physical precipitating factors. 6. Insulin proved a rather more powerful aid to recovery in cases with a sudden onset than in those with an insidious onset, even though the former also did better than the latter in the control group. 7. Gain in weight under treatment was a prognostically favourable sign. It is held that insulin treatment should not be stopped while weight is still being gained, nor regarded as failed before at least 50 comas have been given. 8. The period of hospitalization was reduced by insulin treatment from 16.8 months to 5.4 months. In the control series the mortality over 10 years was 32 per cent., in the insulin group 13.5 per cent.; however this great improvement in mortality is not due to the insulin treatment alone. 9. It is held that the findings noted in paragraphs 1–7 above support the view that the action of insulin treatment is predominantly somatic, increasing powers of bodily resistance to the disease. 10. As a result of an analysis of inter-correlations, it was found that assigning weights to certain clinical features (+2 for duration under 1 year, + 1 for good personality, +1 for presence of precipitation, psychogenic or exogenic) provided a marking system which gave quite an efficient prognosis; thus 84 per cent. of those with a score of +4 had good results but only 13 per cent. of those with a score of 0.