Clinical characteristics and aetiological classification of insulin-dependent diabetes in the elderly.

In a prospective study of 195 newly-diagnosed diabetic patients aged 65 years or over, 80 (41.0 per cent) were treated initially by diet, 89 (45.6 per cent) by diet and oral hypoglycaemic agents, and 26 (13.3 per cent) by diet and insulin. Fifteen patients (7.7 per cent) died within a year of diagnosis. Of 26 patients treated with insulin, six died in the first year, 14 were successfully transferred to diet and oral agent treatment and six continued on insulin--two of whom failed to a trial of oral agents, two showed only a temporary response and two received no trial. A further nine patients were taking insulin 12 months after diagnosis because of no response (eight patients) or a transient response (one patient) only to oral agents. Age, percentage ideal body weight, history of acute onset, blood glucose, glycosylated haemoglobin, and random C-peptide concentration at diagnosis did not discriminate between patients requiring insulin at 12 months and those successfully treated without insulin. Patients who were insulin-dependent 12 months after diagnosis had an increased frequency of ketonuria at diagnosis and a previous medical history of endocrine disease. In insulin-dependent patients there was an increased frequency of HLA DR3 but not DR4 and an increased frequency of thyroid microsomal and gastric parietal cell antibodies but not islet cell antibodies. It is concluded that elderly newly-diagnosed diabetic patients who are treated at diagnosis with insulin are not necessarily insulin dependent and can be given a trial of oral agents with safety.(ABSTRACT TRUNCATED AT 250 WORDS)