Effect of 4 years subcutaneous insulin infusion treatment on albuminuria, kidney function and HbA1c compared with multiple daily injections: a longitudinal follow-up study

AIM: The effect of insulin pump [continuous subcutaneous insulin infusion (CSII)] treatment on diabetes complications in a modern clinical setting is largely unknown. We investigated the effect of 4 years CSII treatment on HbA1c, albuminuria and kidney function compared with multiple daily injections (MDI) in a single-centre clinical setting.

METHODS: All patients initiating CSII treatment from 2004 to 2010 and followed for at least 4 years were included in the study: 193 people with Type 1 diabetes were matched (1 : 2) with 386 patients treated with MDI in the same period. Matching was based on diabetes duration, gender, HbA1c and normo-, micro- or macroalbuminuria at baseline. Urinary albumin/creatinine ratio (UACR) was measured yearly and annual change assessed from linear regression.

RESULTS: CSII- vs. MDI-treated patients were comparable at baseline. After 4 years, HbA1c was 62 ± 11 vs. 68 ± 11 mmol/mol (7.8 ± 1.0 vs. 8.4 ± 1.0 %) (P < 0.001). Annual UACR change in CSII- vs. MDI-treated patients was [mean (95% confidence interval)] -10.1 (-13.3; -6.8) vs. -1.2 (-3.6; 0.9)% (P < 0.001). Reduction in UACR was significantly associated with CSII treatment after adjustment for age, gender, diabetes duration, estimated GFR, UACR, mean arterial pressure, HbA1c , cholesterol, renin-angiotensin-aldosterone system inhibition, anti-hypertensive treatment and smoking (P < 0.001). This remained significant (P < 0.001) when only including patients on stable renin-angiotensin-aldosterone system inhibition during follow-up (n = 465).

CONCLUSIONS: Treatment with CSII over 4 years independently reduced HbA1c and UACR compared with MDI. Reduced UACR may be due to less glycaemic variability because the effect of CSII on HbA1c could only partially explain the effect. This needs confirmation in randomized controlled trials. This article is protected by copyright. All rights reserved.