The optimal glycemic target for elderly patients with diabetes has not been established. The purpose of this study was to elucidate relationship between HbA1c and mortality in elderly patients with diabetes.

Three hundred consecutive elderly (≥65yrs) patients with type 2 diabetes mellitus admitted for control of hyperglycemia between 2002 and 2010 were registered. Upon mortality survey at the end of 2012, 201 (70%) of them were traceable (men/women 121/80, mean age 71yrs, duration of diabetes 11yrs and HbA1c 9.9%). The analysis took account of the following baseline information: gender, age, duration of diabetes, HbA1c, body mass index, systolic blood pressure, eGFR, urinary albumin excretion, serum lipid levels and use of insulin and oral hypoglycemic agents. The follow-up HbA1c was also recorded.

The mean follow-up period was 5.7yrs and 45 of the patients have died. The mortality hazard as a function of the baseline HbA1c quartile was significantly V-shaped with the nadir in quartile 2 (HbA1c 8.5–9.4%) (P=0.02), and this relationship remained significant after adjustment for the confounders such as estimated glomerular filtration rate and insulin use. The follow-up HbA1c was 7.7±1.6% and not significantly related to mortality.

There was a V-shaped relationship between baseline HbA1c and all-cause mortality in elderly patients with insufficiently controlled glycemia. The nadir was in Q2 in which the HbA1c value was 8.5–9.4%. No significant relationship was found between the follow-up HbA1c and mortality. Further studies are needed to clarify the relationship between HbA1c and mortality in the elderly.