By Colin Kenny, Editor – Diabetes Distilled
Investigators wanted to examine whether variation in HbA1c, the degree of variability and how HbA1c varies over time predict hospitalisation and mortality in individuals with type 2 diabetes. They performed a retrospective analysis of a large cohort of patients from UK primary care and found that HbA1c variability was strongly associated with overall mortality and emergency hospitalisation. This was not explained by average HbA1c or hypoglycaemic episodes. Findings were similar for emergency hospitalisations; however, increasing average HbA1c rather than variability was predictive of emergency hospitalisation for coronary artery disease and ischaemic stroke. Investigators suggest that clinical targets should focus on both the stability and absolute level of HbA1c.
Most diabetes management guidelines focus on target HbA1c levels. HbA1c often varies over time. Long-term variability in HbA1c may be predictive of hospitalisation or mortality, but its importance at different average levels or trajectories is unclear. In this study, investigators searched Clinical Practice Research Datalink and identified 58,832 patients with type 2 diabetes who had sufficient HbA1c data over a 3-year period (average HbA1c, variability and estimated trajectory) for inclusion in the study. They adjusted the resultant emergency hospitalisation and mortality hazard ratios for age, sex, smoking status, body mass index, duration of diabetes and deprivation.
Investigators discovered a consistent dose–response relationship between HbA1c variability and mortality. Trajectory had no effect on mortality. Average HbA1c level was important among patients who had the most unstable HbA1c levels (the top 10%); mortality was almost twice as high as in patients with the most stable HbA1c. Similar trends were identified for emergency hospitalisations; however, increasing average HbA1c rather than variability was predictive of emergency hospitalisation for coronary artery disease and ischaemic stroke. Investigators concluded that HbA1c variability was strongly associated with poor outcomes and was not explained by average HbA1c or hypoglycaemic episodes.
To access the publication, click here