Responders to insulin therapy at 18 months in adults with newly diagnosed diabetes: which insulin regimen?

Objectives: To describe baseline characteristics of responders to insulin therapy (HbA(1c) targets < 58 mmol/mol, 7.5%) at 18 months among adults with newly diagnosed diabetes.

Methods: A retrospective UK study derived from 479 general practices electronic dataset. We included all adults (age > 18 years) with newly diagnosed diabetes who required insulin therapy within 6 months of diagnosis. The data comprised insulin regimen (long-acting only; premixed insulin only; basal bolus insulin regimen), gender, Townsend quintile, baseline and an 18-month measurement of clinical and biochemical variables. Multiple imputations were undertaken and logistic regression used to assess the effect of covariates.

Results: A total of 1492 patients (aged 19-93 years) were analysed. Means (SD) baseline HbA(1c) and BMI were 10.3% (2.6%) and 29.6 (7.0%), respectively. Following multiple imputation for missing data, logistic regression analysis indicated important covariates to achieve HbA(1c) targets were baseline HbA(1c), lipid lowering therapy, gender and age. Including all covariates, those treated with premixed insulin were 47% more likely to achieve target HbA(1c) at 18 months than those treated with a basal-bolus regimes (adjusted OR 1.47; 95% CI 1.12-1.92, P = 0.006)) and 32% more likely than those treated with long-acting insulin was (adjusted OR 1.32; 95% CI 1.01-1.74, P = 0.044). Those with a higher baseline HbA(1c) level, on lipid-lowering therapy, women and younger patients had a lower response rate. Mean weight gain (SD) was 2.4 kg (8.5 kg) and was not influenced by treatment regimen.

Conclusion: The use of premixed insulin regimen among newly diagnosed patients with diabetes appears to be most effective in reaching HbA(1c) target values, independent of other confounders. The appropriate choice of insulin regimen at initiation should therefore take into account various metabolic and psychosocial factors.