Optimal Hemoglobin A1c Levels for Screening of Diabetes and Prediabetes in the Japanese Population

Abstract

The aim of this study was to evaluate the utility of hemoglobin A1c (HbA1c) to identify individuals with diabetes and prediabetes in the Japanese population. A total of 1372 individuals without known diabetes were selected for this study. A 75 g oral glucose tolerance test (OGTT) was used to diagnose diabetes and prediabetes. The ability of HbA1c to detect diabetes and prediabetes was investigated using receiver operating characteristic (ROC) analysis. The kappa (κ) coefficient was used to test the agreement between HbA1c categorization and OGTT-based diagnosis. ROC analysis demonstrated that HbA1c was a good test to identify diabetes and prediabetes, with areas under the curve of 0.918 and 0.714, respectively. Optimal HbA1c cutoffs for diagnosing diabetes and prediabetes were 6.0% (sensitivity 83.7%, specificity 87.6%) and 5.7% (sensitivity 60.6%, specificity 72.1%), respectively, although the cutoff for prediabetes showed low accuracy (67.6%) and a high false-negative rate (39.4%). Agreement between HbA1c categorization and OGTT-based diagnosis was low in diabetes (κ = 0.399) and prediabetes (κ = 0.324). In Japanese subjects, the HbA1c cutoff of 6.0% had appropriate sensitivity and specificity for diabetes screening, whereas the cutoff of 5.7% had modest sensitivity and specificity in identifying prediabetes. Thus, HbA1c may be inadequate as a screening tool for prediabetes.

Re‐examining the sensitivity of HbA1c to screen for diabetes mellitus

“…In the present study, the sensitivity of HbA1c as a screening test for DM was significantly higher than that found in earlier studies . One possible reason could be the lack of repeat testing to confirm the diagnosis of DM in many of the earlier studies (as recommended by ADA guidelines), resulting in an overdiagnosis of DM.…”

“…Using an HbA1c cut‐off point of 6.5% (48 mmol/mol), the pooled sensitivity and specificity from 12 studies in Eastern countries were 58.7% and 98.4%, respectively, whereas the pooled sensitivity and specificity from 13 studies in Western countries were 65.5% and 98.1%, respectively . In most of these studies, the diagnosis of DM was made using either a single positive FPG (≥126 mg/dL [7.0 mmol/L]), a single positive 2‐h PG (≥200 mg/dL [11.1 mmol/L]), or any one of the two tests . However, in clinical practice, American Diabetes Association (ADA) guidelines actually require that the diagnosis of DM be confirmed by repeat testing unless there is unequivocal hyperglycemia (i.e.…”

“…7 In most of these studies, the diagnosis of DM was made using either a single positive FPG (≥126 mg/dL [7.0 mmol/L]), [8][9][10][11] a single positive 2-h PG (≥200 mg/dL [11.1 mmol/L]), [12][13][14][15][16] or any one of the two tests. 5,[17][18][19][20][21][22][23][24][25][26][27][28][29][30][31] However, in clinical practice, American Diabetes Association (ADA) guidelines actually require that the diagnosis of DM be confirmed by repeat testing unless there is unequivocal hyperglycemia (i.e. the person is in hyperglycemic crisis or has classical symptoms of DM).…”

Prevalence of moderately increased albuminuria among individuals with normal HbA1c level but impaired glucose tolerance: Results from the LIFE‐Adult‐Study

“…6,7 Some guidelines, including those of the German Diabetes Association, recommend a stepwise diagnostic approach using HbA1c as a first screening test with two different cut-off values for the diagnostic decision, of which the lower cutoff is used to rule out diabetes and the higher cut-off is used to diagnose diabetes without measuring FPG and/or two-hour postload glucose levels (2 h-PG). [11][12][13][14][15][16][17][18][19][20][21][22][23][24] Importantly, stepwise diabetes screening strategies using only HbA1c as first-instance diagnostic parameter may cause failure to detect individuals with abnormal glucose regulation and possible signs of microvascular complications despite HbA1c levels below the "rule-out" cut-off. [8][9][10] Despite the advantages of HbA1c-based diabetes screening, there is an ongoing debate about the usefulness of this marker to detect individuals with diabetes and prediabetes.…”

“…A large number of studies have addressed this question in different populations, however resulting in ambiguous results and conclusions particularly with regard to optimal cut-off values for diagnostic decisions. [11][12][13][14][15][16][17][18][19][20][21][22][23][24] Importantly, stepwise diabetes screening strategies using only HbA1c as first-instance diagnostic parameter may cause failure to detect individuals with abnormal glucose regulation and possible signs of microvascular complications despite HbA1c levels below the "rule-out" cut-off. In this study, we analysed the diagnostic performance of HbA1c compared to FPG and 2 h-PG in a representative sample of adult individuals taken from an urban general population, who were not yet diagnosed with diabetes.…”