Objective: To examine the effects of change in glycaemic control on clinical outcomes in patients with type 2 diabetes.

Research Design and Methods: Data were obtained from 9981 participants of the ADVANCE trial and ADVANCE-ON follow-up. Participants were categorized into 3 groups by glycated haemoglobin (HbA1c) change from baseline to 6 months, >0.25% decrease, >0.25% increase and minimal change (±0.25%). Adjusted Cox and logistic regression analyses examined relationships between HbA1c change and all-cause mortality, macrovascular events, microvascular events and severe hypoglycaemia.

Results: Mean (±SD) age was 65.8±6.4 years, diabetes duration 7.9±6.4 years and HbA1c 7.5±1.6% at baseline. In the first 6 months, 48%, 22% and 30% of participants had an HbA1c decrease, minimal change or increase respectively. Median (IQR) follow-up was 9.4 (5.1-10.4) years. Compared to those with minimal HbA1c change at 6 months, participants with an HbA1c increase at 6 months had greater risk of risk of all-cause mortality [1.26 (1.11-1.43)], macrovascular events [1.26 (1.10-1.44)] and microvascular disease [1.36 (1.09-1.70) whereas participants with an HbA1c decrease did not have greater risk of all-cause mortality [0.99 (0.87-1.12)], macrovascular events [1.06 (0.93-1.21)] or microvascular disease [0.97 (0.77-1.21)]. Neither HbA1c increase [1.05 (0.83-1.33)] or decrease [1.16 (0.91-1.47)] was associated with greater risk of severe hypoglycaemia. These effects were similar when HbA1c change at 3 months was considered and in subgroup analyses by age, gender, diabetes duration, or baseline HbA1c.

Conclusions: Increased rather than decreased HbA1c at 6 months was associated with increased risks of death and vascular events in type 2 diabetes.

Disclosure

N. Nanayakkara: None. T. Ohkuma: None. M. Woodward: Consultant; Self; Amgen Inc., Kirin. J. Chalmers: Research Support; Self; Idorsia. S. Zoungas: None.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at http://www.diabetesjournals.org/content/license.