A method to predict the metabolic effects of changes in insulin treatment in subgroups of a large population based patient cohort.

This case-control study was designed to analyse predictors of the effects on HbA1c levels in 4001 type 1 and type 2 diabetic patients after changing their insulin treatment. Patients from 15 outpatient diabetic clinics were treated with basal insulin and multiple injections of short-acting insulin. The effects on HbA1c of changing from NPH insulin to insulin glargine as basal insulin were studied, compared to patients continuing with NPH insulin. The following possible predictors were examined with multiple regression analysis: age, sex, type and duration of diabetes, smoking, metformin use, insulin requirement, number of basal doses per day, BMI and HbA1c at baseline. The difference between the two regression functions yielded the effect of switching treatment to insulin glargine compared to continuing with NPH insulin. Male gender, low BMI and high baseline HbA1c levels were significant predictors for a greater decrease in HbA1c when changing to insulin glargine. For example, for men with a BMI of 25 and an HbA1c of 8.0%, there was a calculated mean benefit in HbA1c of 0.26 percentage points by changing to insulin glargine, whereas women with a BMI 30 had no benefit of such a change. Thus, changing to insulin glargine had best effect in male patients with low BMI. This is one of the first studies designed to find responders to insulin treatment. Analyses of predictors may prove useful in order to tailor insulin treatment in diabetic patients in clinical practice. The clinical effects need to be confirmed in other studies and randomised controlled trials.