Total hemoglobin count has significant impact on A1C - Data from National Health and Nutrition Examination Survey 1999-2014.

BACKGROUND: Previous study showed A1C was affected by hemoglobin and gender in non-anemic Koreans. However, there was no other data to support those important findings.

METHODS: Participants from National Health and Nutrition Examination Survey (NHANES) - 1999 to 2014 with age over 12 years were examined. After excluding participants with hypoglycemia and moderate to severe anemia, 22,974 participants were included in analysis. Both female and male were divided into three groups based on their hemoglobin level. Linear regression and multivariate regression analyses were conducted to assess the relationship between A1C and hemoglobin in three fasting plasma glucose (FPG) intervals.

RESULTS: For both female and male, A1C was significantly negatively correlated with hemoglobin in the FPG interval range between 80 to 126mg/dl, p<0.0001. A1C was not significantly associated with hemoglobin over the other two intervals (70≤FPG<80mg/dl and FPG≥126mg/dl, p>0.05). In addition, over FPG interval between 80 to 126mg/dl, after adjusting age and fasting glucose level, A1C still significantly negatively correlated with hemoglobin (regression coefficient=-0.052, p<0.0001 in females, regression coefficient=-0.050, p<0.0001 in males).

CONCLUSIONS: A1C would decrease by 0.05% (about 0.5mmol/mol) as a result of each 1g/dl increase in total hemoglobin count when plasma glucose level in normal or prediabetes range. The mechanism behind this finding is unclear. Further study should be conducted and hemoglobin count may need be taken into consideration when applying A1C for the diagnosis of diabetes or to screen for prediabetes.