Diabetes mellitus is an increasingly prevalent complica-tion of cystic fibrosis (CF) as a result of the increased life expectancy resulting from advances in CF treatment. The management of cystic fibrosis related diabetes (CFRD) has previously focused on symptomatic relief under the assumption that, with limited life expectancy, aggressive treatment was unnecessary. However, there is now evidence of a clear association between CFRD and increased morbidity and mortality. 1 2 Lung function and clinical status deteriorate up to 2–4 years before a diagnosis of CFRD based on the oral glucose tolerance test (OGTT), 2 3 suggesting that insulin deficiency may have adverse clinical effects prior to the development of sufficient hyperglycaemia to make a diagnosis of diabetes on conventional OGTT criteria. The introduction of insulin treatment can produce clinical improvement in weight and lung function as well as glycaemia. 4 5 We noticed that in CF patients with diabetes the initiation of insulin treatment was associated with a notable clinical improvement that could not be explained by the treatment of their relatively mild hyperglycaemia. In addition glucose values on home or ward monitoring were often higher than those seen during the OGTT. We therefore gave insulin treatment to four patients with long standing CF, who had weight loss and deteriorating lung function without a clear cause, and had high random glucose values even when their OGTT showed normal glucose tolerance. We report improvement in all four patients in objective markers of clinical status.

CASE 1 Case 1 was a 20 year old male, homozygous for DF508, diagnosed at 6 years of age with failure to thrive and a positive sweat test, with pancreatic insufficiency and stable cirrhosis. From the age of 19 years his lung function and weight had been deteriorating despite trials of corticosteroids, DNase, intravenous and nebulised antibiotics, and nutritional supplements. An HbA1c (4.6%) and OGTT (75 g glucose, baseline 3.8 mmol/l, 60 min 10 mmol/l, 120 min 4.6 mmol/l) were normal. He was given a glucometer to monitor his random postprandial blood glucose (RBG) concentrations, revealing 11 (out of 30) values above 11.1 mmol/l (6.0–24.0 mmol/l) during a two week period. He was commenced on a trial of Insulatard. Three months later on a dose of 8 units daily, without symptoms or glucose values suggestive of hypoglycaemia, his weight (43.7 kg to 49.4 kg) and spirometry (1.48/3.69 l to 1.78/3.85 l; see fig 1) had increased.