Risk of Macrovascular and Microvascular Disease in Diabetes Diagnosed Using Oral Glucose Tolerance Test With and Without Confirmation by Hemoglobin A1c: The Whitehall II Cohort Study

Abstract

Background: It is unclear whether replacing oral glucose tolerance test (OGTT) with hemoglobin A1c (HbA1c) measurement for diagnosing diabetes is justified. We aimed to assess the proportion of OGTT-diagnosed diabetes cases that can be confirmed by HbA1c and to examine whether individuals with OGTT diagnosis but nondiagnostic HbA1c are at higher risk of macrovascular and microvascular disease.

Methods: Participants were 5773 men and women from the population-based Whitehall II prospective cohort study in the United Kingdom. New OGTT diabetes cases diagnosed in clinical examinations in 2002 to 2004 and 2007 to 2009 were assessed for HbA1c confirmation (≥6.5%) in these and subsequent clinical examinations in 2012 to 2013 and 2015 to 2016. All participants were followed up for major cardiovascular events through linkage to electronic health records until 2017 and for incident chronic kidney disease (estimated glomerular filtration rate <60 mL·min^-1·1.73 m^-2) until the last clinical examination. In analysis of vascular disease risk, new OGTT-diagnosed diabetes cases with and without diagnostic HbA1c and preexisting diabetes cases were compared with diabetes-free participants.

Results: Of the 378 (59.3%) participants with OGTT-diagnosed diabetes, 224 were confirmed by HbA1c during 4.1 years (SD, 4.1 years) of follow-up. We recorded 942 cardiovascular events over 12.1 years. After adjustment for nonmodifiable risk factors and compared with the 4997 diabetes-free participants, 371 participants with new HbA1c-confirmed diabetes and 405 participants with preexisting diabetes had increased risk of cardiovascular disease (hazard ratio, 1.53 [95% CI, 1.12-2.10] and 1.85 [95% CI, 1.50-2.28], respectively). The corresponding hazard ratios in the analysis of incident chronic kidney disease (487 cases; follow-up, 6.6 years) were 1.69 (95% CI, 1.09-2.62) for 282 participants with new HbA1c-confirmed diabetes and 1.67 (95% CI, 1.22-2.28) for 276 participants with preexisting diabetes. In both analyses, OGTT cases with nondiagnostic HbA1c (n=149 and 107) had a risk (hazard ratio, 0.99-1.07) similar to that of the diabetes-free population.

Conclusions: More than 40% of OGTT-diagnosed diabetes cases were not confirmed by HbA1c during an extended follow-up. However, because these individuals have a risk of cardiovascular disease and chronic kidney disease similar to that of the diabetes-free population, replacement of OGTT with HbA1c-based diagnosis appears justified.

Keywords: cardiovascular diseases; cohort studies; diabetes mellitus, type 2; glucose tolerance test; glycated hemoglobin A; renal insufficiency, chronic.

Figure 1.

Study design including study-related procedures and follow-ups for different analyses. In analyses of macrovascular and microvascular risk, no diabetes refers to participants without diabetes at baseline and follow-up; known diabetes refers to diabetes cases who had diabetes diagnosed before attending baseline clinical examination; confirmed diabetes refers to oral glucose tolerance test (OGTT)–diagnosed diabetes in study clinic that was confirmed by hemoglobin A1c (HbA1c) test in the same or subsequent clinical examination; and unconfirmed diabetes refers to OGTT-diagnosed diabetes with normal HbA1c at baseline and follow-up. CHD indicates coronary heart disease; CVD, cardiovascular disease; and eGFR, estimated glomerular filtration rate.

Figure 2.

Flowchart for selection of participants for analyses of HbA1c confirmation (A), macrovascular disease (B), and microvascular disease (C). Covariates for analyses of hemoglobin A1c (HbA1c) confirmation (A): ethnicity, body mass index, high-density lipoprotein cholesterol (HDL-C), and HbA1c. Covariates for analyses of macrovascular and microvascular disease (B and C): ethnicity, social position, smoking, total cholesterol, HDL-C, systolic blood pressure, antihypertensive and lipid-lowering medication, and diabetes status. CKD indicates chronic kidney disease; CVD, cardiovascular disease; eGFR; estimated glomerular filtration rate; and OGTT, oral glucose tolerance test.

Figure 3.

Kaplan-Meier estimates for HbA1c confirmation of OGTT-diagnosed diabetes by type of abnormal glucose value (fasting vs 2-hour vs both). HbA1c indicates hemoglobin A1c.

Figure 4.

HRs for new-onset CVD and incident CKD according to diabetes status at baseline. *No diabetes refers to participants without diabetes at baseline and follow-up. Known diabetes refers to diabetes cases who had diabetes diagnosed before attending baseline clinical examination. Oral glucose tolerance test (OGTT)–diagnosed diabetes refers to OGTT-diagnosed diabetes at study clinic either confirmed or not confirmed by hemoglobin A1C test in the same or subsequent clinical examination. †Model 1: adjusted for age, sex, ethnicity, occupational position, and prevalent cardiovascular disease (CVD) at baseline. ‡Model 2: as model 1 but additionally adjusted for smoking, total cholesterol, high-density lipoprotein cholesterol, systolic blood pressure, and use of antihypertensive and lipid-lowering medication. ¶Excess risk mediated by modifiable risk factors listed in Model 2. CKD indicates chronic kidney disease; HR, hazard ratio; and NA, not applicable.