This fact sheet compares the following tests:

• A1C test
• fasting plasma glucose (FPG) test
• oral glucose tolerance test (OGTT)
• random plasma glucose (RPG) test

Confirming Diagnosis of Type 2 Diabetes and Prediabetes

Diagnosis must be confirmed unless symptoms are present. Repeat the test using one of the following methods:

• Repeat the same test on a different day—preferred.
• If two different tests are used—e.g., FPG and A1C—and both indicate diabetes, consider the diagnosis confirmed.
• If the two different tests are discordant, repeat the test that is above the diagnostic cut point.

If diagnosis cannot be confirmed using the results of two tests, but at least one test indicates high risk, health care providers may wish to follow the patient closely and retest in 3 to 6 months.¹

Interpreting Laboratory Results

When interpreting laboratory results health care providers should

• consider that all laboratory test results represent a range, rather than an exact number²
• be informed about the A1C assay methods used by their laboratory²
• send blood samples for diagnosis to a laboratory that uses an NGSP-certified method for A1C analysis to ensure the results are standardized³
• consider the possibility of interference in the A1C test when a result is above 15% or is at odds with other diabetes test results¹
• consider each patient’s profile, including risk factors and history, and individualize diagnosis and treatment decisions in discussion with the patient¹

Comparing Diabetes Blood Tests^‡

View the content below as a table in the PDF Version (460 KB) .

A1C Test

Uses:

• Screening and diagnosis of prediabetes
  • 5.7–6.4%^†
• Screening and diagnosis of type 2 diabetes
  • ≥ 6.5%^†
  • repeat for confirmation of diagnosis
• Monitoring of diabetes

Technical Features:

• Diagnosis requires a laboratory test certified by the NGSP, not meter—point-of-care A1C tests are only suitable for monitoring
• Sample any time of day, no fasting
• Sample: anticoagulated whole blood
• Sample stability: superior
• Sensitivity: less than the FPG test and the OGTT
• Coefficient of variation: assay variability, see www.ngsp.org

With a coefficient of variation as large as that allowed by the National Glycohemoglobin Standardization Program, a reported A1C result of 7.0% could indicate a true A1C of anywhere from ~6.5 to 7.5%.

FPG Test

Uses:

• Screening and diagnosis of prediabetes or impaired fasting glucose (IFG)
  • 100–125 mg/dL^†
• Screening and diagnosis of diabetes
  • ≥126 mg/dL^†
  • repeat for confirmation of diagnosis

Technical Features:

• Diagnosis requires a laboratory test, not meter
• Sample in morning, after 8-hour fast
• Sample: sodium fluoride plasma preferred
• Sample stability: low—requires processing within 30 minutes
• Sensitivity: greater than the A1C test, less than the OGTT
• Coefficient of variation: assay variability:

With a coefficient of variation 5.7% (typical biological variation within the same person), an FPG test result of 126 mg/dL could indicate a true FPG of anywhere from ~110 to 142 mg/dL.

OGTT

Uses:

• Screening and diagnosis of prediabetes or impaired glucose tolerance (IGT)
  • 140–199 mg/dL at 2 hr.^†
• Screening and diagnosis of diabetes
  • ≥200 mg/dL at 2 hr.^†
  • repeat for confirmation of diagnosis
• Screening and diagnosis of gestational diabetes mellitus (GDM)*

Technical Features:

• Sample in morning, after 8-hr. fast and 2 hrs. after glucose load
• Sample stability: low—requires processing within 30 minutes
• Patients should ingest at least 150 g/day of carbohydrates for 3 days prior
• Sensitivity: greater than the A1C or the FPG tests
• Range of variability: 16.7%

RPG Test

Uses:

• Diagnosis of diabetes—used only with classic symptoms of hyperglycemia or hyperglycemic crisis:
  • polyuria, polydypsia, and unexplained weight loss
  • 200 mg/dL^†

Technical Features:

• Sample any time, no fasting
• Sample stability: low—requires processing in fewer than 2 hours

Comparing Diagnoses

In some people, a blood glucose test may indicate a diagnosis of diabetes even though an A1C test does not.

The reverse can also occur—an A1C test may indicate a diagnosis of diabetes even though a blood glucose test does not.

Because of these variations in test results, health care providers should repeat tests before making a diagnosis. People with differing test results may be in an early stage of the disease, where blood glucose levels have not risen high enough to show on every test.

References