Age negatively affects diagnostic efficacy of HbA1c for diabetes
The diagnostic accuracy of HbA1c for diabetes decreases with age, with declining red blood cell (RBC) counts mediating this effect, reveals a new study on Chinese middle-aged and elderly adults.
In the study sample of 4,235 participants (mean age 63.1 years; 42.1 percent male), the sensitivity and specificity of the 6.5 percent HbA1c threshold for diagnosing diabetes were 35.6 and 98.9 percent, respectively. None of the participants had previously known diabetes.
Both sensitivity and specificity declined with increasing age, with the oldest age group (≥75 years) having 33.6 and 98.4 percent sensitivity and specificity, respectively. The negative trend was statistically significant for specificity (p=0.013) but not sensitivity (p=0.477).
Receiver operating characteristic curves were plotted for each age group and revealed that the area under the curve (AUC) values were 0.878, 0.840, 0.838 and 0.755 for the 40 to 54, 55 to 64, 65 to 74 and ≥75 years age groups, respectively.
Compared to the 40 to 54 age group, the diagnostic efficacy of HbA1c ≥6.5 percent was significantly lower in the ≥75 age group (AUC, 0.755 vs 0.878; p<0.001).
Following successive adjustments for various factors that might influence the relationship between HbA1c and age, multiple regression models revealed that estimated glomerular filtration rate and haemoglobin were significant interacting factors, yielding statistically insignificant β values of -0.006 (p=0.359) and -0.005 (p=0.399).
Interestingly, inclusion of RBC count into the multiple regression model not only eliminated the significant interaction between age and HbA1c, but also removed its negative nature, yielding a β value of 0.001 (p=0.856).