The outcome of brittle type 1 diabetes-a 20 year study

Methods: The cohort studied was a group of brittle diabetic patients from various parts of UK originally identified between 1979 and 1985. Patients were traced, where possible, via their diabetic clinics and/or general practitioners. Data on survival or otherwise were obtained from hospital case notes and information from diabetes care team members. For survivors, clinical and demographic information obtained included complication status and whether they still had brittle characteristics. They were also compared with a matched case-control group of type 1 patients with no history of brittle behaviour. Results: The original cohort comprised 33 patients- all female and mean +/- SD, aged 18 +/- 5 years and diabetes duration 8 +/- 4 years. Thirteen were not traceable and 10 of the remaining 20 (50%) had died during the mean 22 years of follow-up. Deaths occurred evenly throughout the period, and causes were chronic renal failure (3), DKA (3), hypoglycaemia (2), subarachnoid haemorrhage (1) and uncertain (1). Age at death ranged from 27 to 45 years. Of the 10 survivors, none remained brittle, but they had a substantial burden of complications. Compared with the non-brittle control group, there was a significant excess of nephropathy and autonomic neuropathy. Conclusions: We conclude that brittle diabetes characterized by recurrent DKA has high long-term outcome mortality. These deaths were premature and almost all diabetes related. Those who survived had resolution of brittleness, but suffered a significant complication burden.