Aims: To examine the association of HbA1c and glucose levels with incident diabetic retinopathy according to black African or white European ancestry. Methods: In this retrospective cohort study of 202 500 US Veterans with diabetes (2000–2014), measures included HbA1c, outpatient random serum/plasma glucose, and incident retinopathy [conversion from negative to ≥2 positive evaluations (ICD-9 codes), without a subsequent negative]. Results: At baseline, the study population had a mean age of 59.3 years, their mean BMI was 31.9 kg/m2, HbA1c level was 57 mmol/mol (7.4%) and glucose level was 8.8 mmol/l, and 77% were of white European ancestry (white individuals) and 21% of black African ancestry (black individuals). HbA1c was 0.3% higher in black vs white individuals (P < 0.001), adjusting for baseline age, sex, BMI, estimated glomerular filtration rate (eGFR), haemoglobin, and average systolic blood pressure and glucose. Over 11 years, incident retinopathy occurred in 9% of black and 7% of white individuals, but black individuals had higher HbA1c, glucose, and systolic blood pressure (all P < 0.001); adjusted for these factors, incident retinopathy was reduced in black vs white individuals (P < 0.001). The population incidence of retinopathy (7%) was associated with higher mean baseline HbA1c in individuals with black vs white ancestry [63 mmol/mol (7.9%) vs 58 mmol/mol (7.5%); P < 0.001)], but with similar baseline glucose levels (9.0 vs 9.0 mmol/l; P = 0.660, all adjusted for baseline age, sex and BMI). Conclusions: Since retinopathy occurs at higher HbA1c levels in black people for a given level of average plasma glucose, strategies may be needed to individualize the interpretation of HbA1c measurements.
Bibliographical noteFunding Information:
The authors gratefully acknowledge the contributions of Ms Christine Jasien, who shared her expertise in Veterans Affairs data and informed the present work. The authors futher acknowledge the contributions of Dr. David Benkeser for statistical guidance.?L.R.S. and L.S.P. are the guarantors of the work and take full responsibility for the work as a whole, including the study design, access to data, and the decision to submit and publish the manuscript. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the US Centers for Disease Control and Prevention or the US Veterans Affairs Department.
PubMed: MeSH publication types
- Journal Article
- Research Support, Non-U.S. Gov't
- Research Support, N.I.H., Extramural
- Research Support, U.S. Gov't, Non-P.H.S.