BACKGROUND: The hemoglobin glycation index (HGI) assesses biological variation in A1c after accounting for the effect of mean blood glucose (MBG). Previous studies minimized analytical variation that could mask biological variation and showed that HGI was consistent within individuals over time and positively associated with risk for microvascular complications. We tested the hypothesis that biological variation in A1c can be assessed by HGI calculated using routine MBG and A1c data obtained from a typical diabetes clinic. METHODS: Self-monitored MBG and A1c were collected from charts of 202 pediatric type 1 diabetes patients attending 1612 clinic visits over 6 yr. Predicted A1c was calculated from the linear regression equation of A1c on MBG in the study population. HGI was calculated by subtracting predicted A1c from observed A1c. Patients were divided into low, moderate, and high HGI tertile groups. RESULTS: Patients used 12 models of glucose meters. Download protocols varied with clinical practice over time. A1c was measured by multiple assays and laboratories. Despite this analytical heterogeneity, HGI was significantly different between individuals and correlated within individuals. MBG (mean ± SD, mg/dL) was similar in the low (186 ± 31), moderate (195 ± 28), and high (199 ± 42) HGI groups. A1c (%) was significantly different (p < 0.0001) in the low (7.6 ± 0.7), moderate (8.4 ± 0.7), and high (9.6 ± 1.1) HGI groups. CONCLUSION: Biological variation in A1c is a robust quantitative trait that can be assessed using HGI calculated from routine clinic data. This suggests that HGI could be used clinically for more personalized assessment of complications risk.
BACKGROUND: The hemoglobin glycation index (HGI) assesses biological variation in A1c after accounting for the effect of mean blood glucose (MBG). Previous studies minimized analytical variation that could mask biological variation and showed that HGI was consistent within individuals over time and positively associated with risk for microvascular complications. We tested the hypothesis that biological variation in A1c can be assessed by HGI calculated using routine MBG and A1c data obtained from a typical diabetes clinic. METHODS: Self-monitored MBG and A1c were collected from charts of 202 pediatric type 1 diabetespatients attending 1612 clinic visits over 6 yr. Predicted A1c was calculated from the linear regression equation of A1c on MBG in the study population. HGI was calculated by subtracting predicted A1c from observed A1c. Patients were divided into low, moderate, and high HGI tertile groups. RESULTS:Patients used 12 models of glucose meters. Download protocols varied with clinical practice over time. A1c was measured by multiple assays and laboratories. Despite this analytical heterogeneity, HGI was significantly different between individuals and correlated within individuals. MBG (mean ± SD, mg/dL) was similar in the low (186 ± 31), moderate (195 ± 28), and high (199 ± 42) HGI groups. A1c (%) was significantly different (p < 0.0001) in the low (7.6 ± 0.7), moderate (8.4 ± 0.7), and high (9.6 ± 1.1) HGI groups. CONCLUSION: Biological variation in A1c is a robust quantitative trait that can be assessed using HGI calculated from routine clinic data. This suggests that HGI could be used clinically for more personalized assessment of complications risk.
Authors: Daniel S Hsia; Neda Rasouli; Anastassios G Pittas; Christine W Lary; Anne Peters; Michael R Lewis; Sangeeta R Kashyap; Karen C Johnson; Erin S LeBlanc; Lawrence S Phillips; James M Hempe; Cyrus V Desouza Journal: J Clin Endocrinol Metab Date: 2020-03-01 Impact factor: 5.958
Authors: James M Hempe; Shuqian Liu; Leann Myers; Robert J McCarter; John B Buse; Vivian Fonseca Journal: Diabetes Care Date: 2015-04-17 Impact factor: 19.112