American Association for Clinical Chemistry
Better health through laboratory medicine
NACB - Scientific Shorts
NACB - Scientific Shorts (formerly NACB Blog)
By William Winter, MD

I received a call from a local pathologist that their lab was receiving requests for 3-hour oral glucose tolerance tests (OGTTs) on children. In addition, insulin levels were to be measured at each of the 4 time points (zero, 1, 2 and 3 hours). This was a challenge to the patients as the children would then require 4 venipunctures.

Is a 3-h OGTT standard practice? Why might the insulin levels be requested? Are insulin levels diagnostic of diabetes?



3 hour OGTTs are not standard outside of pregnancy. Furthermore, in the 2011 American Diabetes Association (ADA) Clinical Practice Recommendations, a 75-g, 3-point (0, 1 and 2 h), 2 hour OGTT is recommended as the diagnostic procedure for gestational diabetes mellitus (GDM) (to replace the 10-g, 3-h OGTT) assuming that diabetes has not otherwise been diagnosed using the "non-pregnant" criteria.


A 3-h OGTT is not standard and is not the diagnostic tool of choice compared to the 2-h OGTT. Using current diagnostic guidelines, there is no additional diagnostic information that is derived from more time points between fasting and 2 hours, nor is more diagnostic information provided by time points beyond 2 hours. Of course in the absence of frank hyperglycemia with symptoms, diabetic ketoacidosis (DKA) or hyperglycemic, hyperosmolar state (HHS), initial testing for diabetes should usually begin with the measurement of the fasting plasma glucose (FPG) or hemoglobin A1c.
A research tool, the whole body insulin sensitivity index (WBISI), is a 2-h test of glucose tolerance that can be used to measure insulin sensitivity in adults in research settings. This requires 2 baseline samples (excluding the zero time sample) and then samples every 30 minutes for 2 hours (this includes a separate zero time glucose). This test was validated in adults as a measure of insulin sensitivity (Matsuda M, DeFronzo RA. Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp. Diabetes Care. 1999 Sep;22(9):1462-70). However this is a tool for research and is not a routine clinical tool.
A 3-h variation of the WBISI has been reported in children (Yeckel CW, Weiss R, Dziura J, Taksali SE, Dufour S, Burgert TS, Tamborlane WV, Caprio S. Validation of insulin sensitivity indices from oral glucose tolerance test parameters in obese children and adolescents. J Clin Endocrinol Metab. 2004 Mar;89(3):1096-101.). This requires 2 baseline samples (one of which is the zero time sample) and then samples every 30 minutes for 3 hours. However like adults, this should be considered a tool for research and is not a routine clinical tool. As well, this is different than measuring glucose and insulin every hour as noted in our case (and not every 30 minutes as noted in the paper by Yeckel et al.).
Whereas insulin resistance and hyperinsulinism are characteristic of type 2 diabetes, there is no recommendation from the ADA that insulin (or C-peptide) be measured in the diagnosis of any form of diabetes. Furthermore, hyperinsulinism may not persist in long-standing type 2 diabetes as progressive beta-cell failure can occur. As well, there are significant overlaps in insulin and C-peptide levels between patients with type 1 and type 2 diabetes. Because the measurement of insulin lacks diagnostic value, insulins should not be measured as a routine part of the OGTT. In addition, while there is usually reference range data for fasting insulin levels, there is little data on “normal” insulin responses to oral glucose.
Why measure insulin?


In patients suspected of having the metabolic syndrome (the insulin resistance syndrome, a.k.a.: dysmetabolic syndrome X, ICD-9: 277.7), physicians will sometimes measure a fasting insulin level. If the insulin level is elevated, the physician can provide this information to the patient as objective evidence that the patient is indeed insulin resistant. This practice is strictly at the discretion of the physician and is not advised by the ADA. While a grossly elevated insulin level associated with a normal blood glucose level is consistent with insulin resistance, by itself the insulin level provides no information that specifically guides patient management. Because of the cost of the insulin measurement and its lack of proven value in affecting patient outcome, insulin measurements can not be routinely recommended. Additionally excluding the research setting, insulin measurements following a glucose challenge should be discouraged because they are difficult to interpret. Possibly in the future as we learn more about insulin resistance, calculated measures of insulin sensitivity (e.g., HOMA-R, QUICKI, WBISI or the insulin/glucose ratio) may be valuable in routine practice. However that day is not yet here.


Of interest, hyperinsulinism in children can be treated with metformin. However there appears to be no data in children that this slows progression to type 2 diabetes. There is data in adults that progression to type 2 diabetes is averted by metformin therapy; however, metformin was only half as effective as diet and exercise in reducing the later development of type 2 diabetes (Knowler WC, Barrett-Connor E, Fowler SE, Hamman RF, Lachin JM, Walker EA, Nathan DM; Diabetes Prevention Program Research Group. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med. 2002 Feb 7;346(6):393-403.).
One last comment: Practitioners are often astute at recognizing insulin resistance from their physical examination and routine laboratory testing without the need for insulin measurements. Insulin resistant patients typically have various combinations of centripetal obesity, hypertension, hypertriglyceridemia, low HDL-C, hyperuricemia, acanthosis nigricans, and, in women, hyperandrogenism or polycystic ovary syndrome.
Note: The diagnostic procedure for a 2-h OGTT for diagnostic purposes outside of pregnancy is as follows:


Subjects must be on a stable diet, at a stable weight, with a stable level of exercise, without acute illness or recent hospitalization. Carbohydrate (100 - 150 gm/day) is taken as part of the diet for 3 days prior to the scheduled OGTT. The subject is to fast overnight usually for 8 to 14 hours and is npo except for water. No medications, caffeine or tobacco are to be taken until the completion of the test. Two samples are drawn: time zero and 2 hour samples. Thus the OGTT can be termed a “2 + 2" (2 samples - 2 hours). Time zero is when the subject begins to drink the sugar beverage. Adults are to receive 75 g of glucose whereas children receive 1.75 g per kilogram to a maximum of 75 g. The beverage should be consumed in 5 minutes or less. The maximum glucose concentration in the beverage is 25 g per 100 ml.




Please log in to leave a comment!

Posted by
On 11/6/2011

I would like to draw the attention to Dr. Winter's statement in response, "there is no additional diagnostic information that is derived from more time points between fasting and 2 hours, nor is more diagnostic information provided by time points beyond 2 hours". In practice, we come across OGTT results where both fasting as well as 2 hr post glucose glucose are well within the limits but only one are two results between zero and 2 hrs going above 140 mg/dL. In the classification well known before, it was termed impaired glucose tolerance which is a pre-diabetic state. Hence the points between zero and two hrs certainly are of value especially in patients who are having family history of diabetes or diabetic risk factors eg., central obesity but who are asymptomatic and normoglycemic during presentation to the clinic. It may be argued that HbA1c can be used to unmask such cases. Unless the role of HbA1c in detecting pre-diabetic state is well established, in my opinion it cannot be used for the said purpose. Dr. Venkatesh Dambala Consultant - Clinical Biochemistry Apollo Bramwell Hospital, Mauritius

Posted by
On 9/13/2011

HbA1c%, fructosamine and fasting glucose with triglyceride would be adequate for the work up for DM in children. Simultaneously measured plasma insulin level may help in diagnosisng insulin resistance and metabolic syndrome in the children who are no infrequently the victims of civilisation i.e metabolic syndrome/insulin resistance due to sedentary life with TV/computers and fast foods. This comment was approved by the NACBLOG editorial board. Please remember to add your name and affiliation!

Posted by
On 9/11/2011

ogtt This comment was approved by the NACBLOG editorial board. Please remember to add your name and affiliation!

Posted by
On 9/10/2011

Routine insulin measurements with the OGTT are not necessary nor recommended. Data obtained were sometimes used to differentiate (among other conditions) juvenile diabetics whose insulin fails to respond to glucose and obese maturity-onset diabetics who show a delayed response (1-2h) sometimes to high insulin values that fail to return to normal over some hours. Now, many endocrinologists are satisfied with and an OGTT with only 2 glucose measurements: fasting and at 2 hours after a drink of 75 g of glucose in 250–350 ml of water (in children: 1.75 g/kg up to maximum of 75 g). HbA1c has also a place in diagnosis of pediatric diabetes. Testing is indicated in children in whom diabetes is suspected but the classic symptoms and a casual plasma glucose ≥11.1 mmol/l are not found. This comment was approved by the NACBLOG editorial board. Please remember to add your name and affiliation!

Posted by
On 9/9/2011


About the Author
William Winter, MD
William Winter, MD