Insulin dependence: problems with the classification of 100 consecutive patients.

We have evaluated the clinical and immunogenetic features of 100 consecutive patients presenting to an adult diabetic clinic who were judged clinically to need insulin therapy but were not sufficiently ill to be admitted to hospital. Over this same period 15 newly diagnosed patients (aged 13-70 years) were started on insulin as in-patients of whom ten were in ketoacidosis. The 100 out-patients, aged 11-75 years at the time of starting insulin, were followed for at least a year. Fifty-six had islet cell antibodies and/or were heterozygous for HLA DR3 and DR4 (Group A) whereas 44 had neither of these markers (Group B). Islet cell antibodies and/or DR3, DR4 heterozygosity were most common in the 70 patients diagnosed below the age of 40 years but were also found in older patients. Patients in Group A were significantly younger at diagnosis (29 vs. 43 years), had a shorter duration of symptoms (17 vs. 61 weeks), were more likely to have ketonuria, and had a lower random C-peptide level at diagnosis (0.2 vs. 0.31 nmol/l). The two groups could not be distinguished by weight, haemogloblin A1 or blood glucose at diagnosis or by diabetic control or insulin dose after one year. The National Diabetes Data Group (NDDG) definition of insulin dependence stresses the importance of HLA types and islet cell antibodies although we found their prevalence to be low in the 30 patients diagnosed over 40 years who clinically were indistinguishable from the younger patients.(ABSTRACT TRUNCATED AT 250 WORDS)