Haemoglobin A1c and 5-year all-cause mortality in French type 2 diabetic patients aged 70 years and older: The GERODIAB observational cohort

Abstract

Aim: The role of glycaemic control in the mortality of elderly diabetic patients remains uncertain. GERODIAB is the first multi-centre, prospective, observational study that aims to describe the link between HbA1c and 5-year mortality in French, type 2 diabetic patients aged ≥70 years.

Methods: Consecutive patients (n=987; mean age 77 years) were included from 56 diabetes centres and followed for five years. Individual histories, risk factors, standard diabetes parameters and geriatric evaluations were regularly recorded. Survival was studied using the Kaplan-Meier method. Multivariable analyses used Cox regression.

Results: Twenty-one percent of the patients died, 13% were lost during follow-up. Patients with a 5-year mean HbA1c in the range [40-50) mmol/mol ([5.8-6.7) %) had the highest survival (84%); those in the range [50-70) mmol/mol ([6.7-8.6) %) or <40mmol/mol (<5.8%) an intermediary survival rate (79%); patients with HbA1c ≥70mmol/mol (≥8.6%) the worst survival (71%). Patients with mean HbA1c ≥70mmol/mol (≥8.6%) had a significantly higher mortality than those with lower HbA1c (P=0.011), and HbA1c remained a significant predictor of mortality after adjusting for individual, diabetic and geriatric factors (hazards ratio [95%CI]: 1.76 [1.21 to 2.57], P=0.0033). Survival was also significantly associated with both HbA1c variability and with the frequency of HbA1c determinations.

Conclusion: In this large sample of elderly French type 2 diabetic patients, an HbA1c level <70mmol/mol (<8.6%) was associated with lower mortality. The range [40-50) mmol/mol ([5.8-6.7) %) could be an acceptable target provided patients are not exposed to hypoglycaemia.

Trial registration: ClinicalTrials.gov NCT01282060.

Keywords: Elderly; GERODIAB; HbA1c; Mortality; Type 2 diabetes.