Autoantibodies in Diabetes Testing

In order to diagnose diabetes, technologists measure glucose or hemoglobin A1C concentrations or perform an oral glucose tolerance test. That being the case, what are autoantibody tests ordered on patients with diabetes or pre-diabetes, or those suspected of having diabetes?

Diabetes is classified as type 1, type 2, gestational and “other specific types”, with about 40 different types of diabetes known. Type 1 and type 2 account for the majority of the diabetes cases, with type 1 accounting for roughly 10% of all diabetes and type 2 accounting for roughly 90%. Type 1 diabetes (sometimes referred to as autoimmune diabetes) is caused by the autoimmune destruction of the pancreatic beta cells. This results in an absolute insulin deficiency and requires insulin to treat. Type 2 diabetes is associated with obesity and insulin resistance and is usually treated initially with diet, exercise and weight loss. Several drugs are currently available to treat hyperglycemia in type 2 diabetes if behavioral modifications do not succeed in lowering glucose levels.

The diagnosis of diabetes is made based on the glucose level and/or hgbA1c concentration, however, not all cases are clear-cut or follow the classic presentations or indications to allow classification as type 1 or 2. Yet determining whether diabetes is type 1 or 2 is important since the treatments are different. Also, persons with type 1 diabetes are prone to other autoimmune disorders, such as autoimmune thyroid disease. In cases that are less than clear-cut, measuring autoantibodies can provide useful information. The main autoantibodies, along with their utilities include:

ICA – Islet cell cytoplasmic autoantibodies are rare in the general population but are present in 70-80% of new-onset type 1 diabetics. They are also present before the onset of type 1 diabetes, so the presence of ICA in a non-diabetic is indicative of a markedly increased risk of type 1 diabetes. If ICA are detected in a type 2 diabetic, it suggests the slowly progressive autoimmune destruction of the beta cells, and that the person actually has a form of diabetes known as latent autoimmune diabetes of adulthood (LADA). Although present in 70-80% of newly diagnosed type 1 diabetics, ICA decline to a frequency of about 5 – 10 % by 10 years post diagnosis.

GAD65/GAD67/GADA – these are the 65 KD and 67KD forms of glutamic acid decarboxylase autoantibodies. GADA like ICA are present in roughly 70-80% of new-diagnosis type 1 diabetes. They also are indicative of progression to type 1 diabetes if found in non-diabetics or type 2 diabetics. However GADA are also found in up to 3% of the general population, so ICA are more specific for type 1 diabetes.

IA2 or IAA – insulin autoantibodies are the least common type of autoantibody present at onset of type 1 diabetes, only occurring in 50-60% of children, and occurring uncommonly in adults with new onset type 1 diabetes. They are also the least disease-specific of the autoantibodies. In addition assays to measure IA2 do not distinguish between antibodies to insulin that may be produced against insulin being administered, versus insulin autoantibodies.

IA-2A – insulinoma-associated-2 autoantibodes are present in roughly 60% of new-onset type 1 diabetes and generally more prevalent then IAA.

ZnT8A – This is the newest autoantibody discovered and it is raised against the transporter that moves zinc from the cytoplasm to the insulin-containing secretory granules in the beta cells. ZnT8A are present in 60-70 % of new onset type 1 under the age of 20, and about 40% after 20 years old. They are also present in 14% of cases which are negative for GADA, IAA and IA-2A. They are also common in LADA patients so can be useful in that diagnosis.

The autoantibodies are especially useful when the diagnosis of the type of diabetes is unclear, and when there is some suspicion that a person with type 2 diabetes may in fact have autoimmune diabetes. Because none of these autoantibodies are present in greater than 80% of type 1 diabetics, measuring several of them is sometimes necessary in order to sort out the diagnosis.



-Patti Jones PhD, DABCC, FACB, is the Clinical Director of the Chemistry and Metabolic Disease Laboratories at Children’s Medical Center in Dallas, TX and a Professor of Pathology at University of Texas Southwestern Medical Center in Dallas.


Estimated Average Glucose

Most people in the clinical lab and among the Diabetes population are aware that there is not a good correlation between a blood glucose level and a hemoglobin A1C (HbA1c) level. This only makes sense. A blood glucose is essentially a snapshot of what the glucose concentration is in your body at that particular point in time when the blood sample is collected. HbA1c is a measure of the percent of your hemoglobin that has glucose attached to it. The longer your glucose is high and the higher it is, the more glucose will be attached to hemoglobin and the higher your percent HbA1c.  The hemoglobin with glucose attached stays around for the life of the red blood cell that holds it, 120 days.  Therefore, a HbA1c level is an indication of what your blood glucose has been averaging for the last 4 months.

That’s where an estimated average glucose (eAG) comes in. eAG is a value that’s calculated from the HbA1c, so it is also an indicator of what your blood glucose has averaged over the last 120 days rather than being a snapshot of your current blood glucose. Like HbA1c, its utility lies in that a person may not have been in control of their glucose for the last 4 months, but they are careful the day they come in to have their glucose checked. Their snapshot blood glucose may be 130 mg/dL or close to the normal range, but their eAG would still be 200 mg/dL or more, indicating what they’ve been averaging the last 4 months.

The formula for calculating eAG was developed by Nathan et al (1). Their study in which they derived it is impressive.  They collected roughly 2700 separate glucose measurements on each of 507 study subjects. The study cohort contained 268 persons with type 1 diabetes, 159 persons with type 2 diabetes and 80 normal controls. All participants in their study had their glucose under control for the 3 month run of the study. With so many individual measurements on each study subject, the authors were able to determine an average glucose for each subject and correlate it with the subject’s HbA1c. The formula they derived is:

      eAG = 28.7 X HgbA1C – 46.7 (for US units of mg/dL)

Interestingly, HbA1c gives you essentially the same information as eAG: an indication of what a person’s average glucose has run over the last 120 days. The difference is this: most people monitoring blood glucose know what a glucose value means and when their glucose value is too high. Thus an eAG of 250 mg/dL may make more sense to them than an HbA1c of 10.3 %.  They now know that they have been averaging a glucose of 250 mg/dL, even if today’s glucose was 126 mg/dL.  Whether it’s true or not that eAG is easier to understand than HbA1c, HbA1c has become widely used and eAG has not. Despite that, it does have the potential to be a useful calculation.

1. Nathan DM, Kuenen J, Borg R, Zheng H, Schoenfeld D, Heine JR. Translating the A1C assay into estimated average glucose values. Diabetes Care, 31(8):1473-1478. 2008.


-Patti Jones PhD, DABCC, FACB, is the Clinical Director of the Chemistry and Metabolic Disease Laboratories at Children’s Medical Center in Dallas, TX and a Professor of Pathology at University of Texas Southwestern Medical Center in Dallas.

Cystic Fibrosis Related Diabetes

Cystic fibrosis-related diabetes (CFRD) is a type of diabetes that affects individuals who have Cystic Fibrosis. CFRD is an entity unto itself, having several aspects that make it different from other forms of diabetes.

Cystic Fibrosis (CF) is one of the most common genetic defects among the Caucasian population, and it is a devastating, systemic disease. When CF was first being diagnosed, children with this disorder rarely lived to reach their teens]; now the average life expectancy  of an individual with CF is around 36 years. Still horrifically short, but better. The fact that people with CF are living longer means they acquire other disorders, including a type of diabetes. It has been shown that with increasing age in the CF population there is increasing incidence of diabetes mellitus.  Roughly 20% of adolescents with CF have diabetes and about 50% of adults with CF have CFRD (1).

CFRD is not as straight-forward to diagnose as type 1 and type 2 diabetes, so it’s important for laboratory professionals to be aware of this disease. People with CF who have diabetes may not always have hyperglycemia. Also hemoglobin A1c (Hgb A1c) values, which is a test recommended by the ADA for diagnosing diabetes, may not be elevated in these patients. The oral glucose tolerance test (OGTT) is recommended for diagnosis of CFRD, and yet even these results may be equivocal in CFRD patients (2). Nonetheless, the ADA/CFF guidelines suggest that all CF patients over 10 years of age should be screened yearly for CFRD using the OGTT. In addition, at least one study in the literature has found that when performing an OGTT on CFRD patients, a glucose level at the 1 hr time point correlates best with the patient’s lung function (3). Thus, if your lab performs OGTT on individuals with suspected CFRD, the physician requesting the test may want the glucose value on a one hour time point as well as the standard 2 hour OGTT.

Individuals with CF who get CFRD tend to have weight loss, protein catabolism, worsened lung function and significantly increased mortality compared to CF individuals without diabetes. The increased mortality is directly related to decreased pulmonary function, rather than to the atherosclerotic vascular disease seen in other types of diabetes. Insulin therapy is the recommended therapy for CFRD.

-Patti Jones


  1. Moran A, Brunzell C, Cohen RC, Katz M, Marshall BC, Onady G, Robinson KA, Sabadosa KA, Stecenko A, Slovis B. Clinical care guidelines for cystic fibrosis-related diabetes.  Diabetes care 33(12):2697-2708. 2010.
  2. Rana M, Munns CF, Selvadurai H, Donaghue KC, Craig ME. Cystic fibrosis-related diabetes in children – gaps in the evidence? Nature Reviews: Endocrinology, 6:371-378. July 2010.
  3. Brodsky J, Dougherty S, Makani R, Rubenstein RC, Kelly A. Elevation of 1-hour plasma glucose during oral glucose tolerance testing is associated with worse pulmonary function in cystic fibrosis. Diabetes Care, 34:292-205. 2011.