Relation between HbA1c and incident cardiovascular disease over a period of 6 years in the Hong Kong population.

AIM: The current trend on diabetes management advocates replacing the paradigm from a uniform to an individualized patient-centered haemoglobin A1c (HbA1c) target, but there is no consensus on the optimal HbA1c level. The study aimed at examining the association between HbA1c and the risk of cardiovascular diseases (CVD) for diabetic patients with different characteristics, in order to identify patient-centered treatment targets. METHODS: A retrospective cohort study was conducted on 115,782 Chinese adult primary care patients with type 2 diabetes mellitus (DM) but no known CVD history, who were prescribed antidiabetic medications in 2010-2011. The cumulative mean HbA1c over a median follow-up period of 5.8 years was used to evaluate the relationship between HbA1c and CVD incidence using Cox analysis. Subgroup analyses were conducted by stratifying different baseline characteristics including gender, age, smoking status, diabetes duration, body mass index, Charlson's comorbidity index and DM treatment modalities. RESULTS: For patients with a DM duration of<2years, an exponential relationship between HbA1c and risk of CVD was identified, suggesting that there was no threshold HbA1c level for CVD risk. For other diabetic patients, an HbA1c level of 6.8-7.2% was associated with a minimum risk for CVD and a J-shaped curvilinear association between HbA1c. The risk of CVD increased in patients with HbA1c<6.5% or ≥7.5%. CONCLUSION: Among Chinese primary care patients at the early (<2years) disease stage, lower HbA1c targets (<6.5%) may be warranted to prevent CVD events whilst for all others, excessively lower HbA1c levels may not necessarily better and can potentially be harmful.