Reconsidering the HbA1c Cutoff for Diabetes Diagnosis Based on a Large Chinese Cohort

Jiying Qi,Yang Su,Qianqian Song,Zhaojun Ding,Min Cao,Bin Cui,Yan Qi

DOI: https://doi.org/10.1055/a-0833-8119

2021-01-01

Experimental and Clinical Endocrinology & Diabetes

Abstract:Introduction The HbA1c has been considered as the 'gold standard' in diabetes diagnosis and management, however, age, gender and body mass index (BMI) might have certain effects on HbA1c. We are aiming to further investigate the correlation between age and HbA1c, and whether it was affected by gender and BMI. Methods A cross-sectional survey including 135,893 nondiabetic individuals who took the physical examination between 2013 and 2017 was conducted. The subjects were grouped by gender, age and BMI, and the interactive and independent effects of the 3 factors on the HbA1c were detected. The median and 95 % confidence interval (CI) of HbA1c levels were calculated. Results The HbA1c levels gradually increased along with age, both in female and male, and there is a positive association between BMI and the HbA1c. The difference on HbA1c in gender was associated with both age and BMI, the age-related increase in HbA1c was accentuated in the subgroup with higher BMI, and there was a marked accentuation of the positive association between BMI and HbA1c as age increased. In almost all the young and middle-aged (aged 20-59) subgroups, the 97.5th percentiles of HbA1c levels were lower than 6.5 %, suggesting that the single HbA1c cutoff value is probably not applicable to the young and middle-aged population. Conclusions We recommend that the effects of age, gender and BMI should be taken into consideration when using HbA1c for the diagnosis and management of diabetes, especially in the young and middle-aged population.