Endocrine Abstracts

Introduction: The strong correlation between HbA1c and blood glucose (BG) has been recognised in many studies. We investigated this relationship using BG data from 119 children with diabetes, to better understand factors affecting HbA1c and characteristics of children with good versus poor control.

Methods: BG data was obtained on 119 children over a 1-month period and on a subset of 43 children over three consecutive months using the Diasend System (Aidera, Sweden). HbA1c was obtained at the beginning and end of the 3-month period. Linear regression was used to assess the relationship between HbA1c and BG and any additional effects of BG variability or age.

Results: Our model confirmed the linear relationship between HbA1c and BG. Each additional 1 mmol BG corresponded to an increase of 0.35 (95% CI 0.3–0.4) HbA1c%. Age had a significant effect on HbA1c after adjusting for average BG (P=0.003) – for an average BG of 10 mmol, predicted HbA1c measurements were 7.8% for age eight and 8.1% for age fifteen. The 3-month data showed a significant relationship (P=0.027) between absolute change in HbA1c and BG variability (S.D.). Children with more variable BG were more likely to have a larger change in HbA1c. Children with poor control (HbA1c >9.5%) tended to take fewer BG measurements, have more variable measurements, and have a peak measurement between 1800–2000 h compared with those with good control (<7.5%).

Conclusion: Routinely downloading BG meter readings has been well accepted by parents and children and has become an essential part of clinical consultation. It allows HbA1c to be expressed as the more easily understood average BG, enables children to visualise and discuss their data, and identifies areas in which to target improvements. We recommend that downloads and assessment of average BG becomes routine in clinic.