Here's some info form the 2009 North American CF Conference. I'd look into insulin, now.

S19.1: Pts with CFRD have a 6-fold increase in morbidity and mortality. CFRD is usually asymptomatic and can remain undetected for up to 4 years prior to diagnosis. This team recommends Continuous Glucose Monitoring (CGM) in pts with normal and impaired glucose tolerance. Early insulin therapy in CF improves growth, lung function, and reduces the number of chest infections. Some say 5--75% of CF pts will be diagnosed with CFRD by age 30. HbA1c test should NOT be used as a screening test for CFRD, because it will look normal. The red blood cell turnover is less than 3 months in CF pts, especially those with poor pulmonary function. Normal random glucose levels do not exclude a diagnosis of CFRD, either. Fasting plasma glucose is also not reliable and has many false positives due to large individual variation. Nearly 2/3rds of CFRD adults do not have fasting hyperglycemia. Glycosuria is also a poor test in al already catabolic condition such as CF. Only 1/3 of CFRD pts had symptoms of polyuria or polydipsia at the time of diagnosis. Oral GLucose Tolerance Testing may NOT be the most accurate tool in diagnosing CFRD, since about 33% of diagnoses are missed when one relies on this test alone. The answer to all of these problems is CGM, which is a continuous measure of interstitial glucose every 5 minutes or 288 times per day via a subcutaneous sensor inserted for 3-6 days. The newest meters display real time BG concentrations and retrospective data. Sensors are wireless. Accuracy of CGM is not equal to BG testing but gives a continuous 24 hour picture like a video camera.

S19.2: CFRD results from reduced secretion of insulin secondary to pancreatic damage. The average age of onset of CFRD is 19-21 years. About 20% have CFRD. CFRD is more prevalent in females. The prevalence of CFRD increases with age. CFRD is tightly linked to poor lung function. In the US, average CFRD FEV1 was 55% compared to non-CFRD CF FEV1 of 67.5%. The rate of decline of FEV1 increases 2-4 years prior to diagnosis with CFRD. Insulin therapy restores FEV1 in CFRD pts to levels recorded at the onset. Fewer than 25% of CFRD pts survived to age 30, compared to 60% of non-CFRD pts. There needs to be annual surveillance in CFRD pts for retinopathy and neuropathy. So far, no one has found accelerated atherosclerotic macrovascular disease in CFRD pts. Pre-diabetic decline can be reversed with early insulin therapy, and it also reverse weight loss. Insulin therapy also improves lung function and BMI in people with normal GT but abnormal random BG concentrations and in people with impaired GT. Insuline therapy is the only recommended treatment for CFRD. CFRD pts require little basal insulin, but need insulin for meal coverage. Multiple daily injections of short-acting insulin before each meal with a small dose of long-acting insulin at night are recommended. CGM is also encouraged. In CFRD, caloric restriction is not appropriate, but some diet manipulation might be possible. Replace some refined sugary food, such as breakfast cereal, with a lower glycemic index food, such as porridge. CGM with a food diary has been helpful in highlighting problematic foods. Whle people with CF require a high fat intake, in time recommendations on fat intake may need to be adjusted as a small proportion of people with CF have hypertriglyceridemia and hypercholesterolemia, suggesting that macrovascular disease could develop as people live longer.