Aging is associated with increased HbA1c levels, independently of glucose levels and insulin resistance, and also with decreased HbA1c diagnostic specificity

N Dubowitz,W Xue,Q Long,J G Ownby,D E Olson,D Barb,M K Rhee,A V Mohan,P I Watson-Williams,S L Jackson,A M Tomolo,T M Johnson 2nd,L S Phillips

DOI: https://doi.org/10.1111/dme.12459

Abstract:Aim: To determine whether using HbA1c for screening and management could be confounded by age differences, whether age effects can be explained by unrecognized diabetes and prediabetes, insulin resistance or postprandial hyperglycaemia, and whether the effects of aging have an impact on diagnostic accuracy. Methods: We conducted a cross-sectional analysis in adults without known diabetes in the Screening for Impaired Glucose Tolerance (SIGT) study 2005-2008 (n=1573) and the National Health and Nutrition Examination Survey (NHANES) 2005-2006 (n=1184). Results: Both glucose intolerance and HbA(1c) levels increased with age. In univariate analyses including all subjects, HbA(1c) levels increased by 0.93 mmol/mol (0.085%) per 10 years of age in the SIGT study and by 1.03 mmol/mol (0.094%) per 10 years in the NHANES; in both datasets, the HbA(1c) increase was 0.87 mmol/mol (0.08%) per 10 years in subjects without diabetes, and 0.76 mmol/mol (0.07%) per 10 years in subjects with normal glucose tolerance, all P<0.001. In multivariate analyses of subjects with normal glucose tolerance, the relationship between age and HbA(1c) remained significant (P<0.001) after adjustment for covariates including race, BMI, waist circumference, sagittal abdominal diameter, triglyceride/HDL ratio, and fasting and 2-h plasma glucose and other glucose levels, as assessed by an oral glucose tolerance test. In both datasets, the HbA(1c) of an 80-year-old individual with normal glucose tolerance would be 3.82 mmol/mol (0.35%) greater than that of a 30-year-old with normal glucose tolerance, a difference that is clinically significant. Moreover, the specificity of HbA(1c) -based diagnostic criteria for prediabetes decreased substantially with increasing age (P<0.0001). Conclusions: In two large datasets, using different methods to measure HbA(1c), the association of age with higher HbA(1c) levels: was consistent and similar; was both statistically and clinically significant; was unexplained by features of aging; and reduced diagnostic specificity. Age should be taken into consideration when using HbA(1c) for the diagnosis and management of diabetes and prediabetes.