Home / Resources / Articles / International Committee Advises: Don’t Measure Fasting Insulin in Children

International Committee Advises: Don’t Measure Fasting Insulin in Children

Sep 28, 2009

There is no reason to test overweight or obese children for insulin resistance, according to an international committee of experts in pediatric endocrinology and diabetes.

The panel’s conclusions were summarized by five of its members at a joint meeting of the Lawson Wilkins Pediatric Endocrine Society and the European Society for Pediatric Endocrinology.


The evidence-based document, to be submitted for publication in October, will address the definition, measurement, risk assessment, treatment, and prevention of insulin resistance in children. It has been endorsed by seven specialty societies and is financially supported by the Institut National de la Santé et de la Recherche Médicale (INSERM) and an unrestricted educational grant from the French pharmaceutical company Ipsen.

Although the document is subject to revision, it is expected to recommend against the use of fasting insulin levels — or any laboratory test — to screen for insulin resistance in children, and against the use of medication to treat children with insulin resistance in the absence of specific diagnoses such as Type 2 diabetes or polycystic ovarian syndrome.

In adults, insulin resistance has been strongly linked to obesity, Type 2 diabetes, and cardiovascular disease, and there is also some evidence linking it with a risk for those conditions among children, said Dr. Franco Chiarelli, panel cochair.

“But unfortunately for us pediatricians, there is a lack of clarity as to what insulin resistance means in childhood, how it is best assessed, what clinical disorders occur, and its consequences. And, there is debate on how to treat and possibly prevent insulin resistance in children,” said Chiarelli, professor and head of pediatrics at the University of Chieti (Italy).

Dr. Claire Levy-Marchal, another panel cochair, said that population data on the distribution of normal insulin levels is fairly well characterized in adults but not in children, in whom fasting insulin levels vary by weight, nutrition, activity, gender, developmental stage, ethnicity, and other factors. Thus, there is no clear cutoff between normal and abnormal. Moreover, there is no standardized methodology for measuring fasting insulin, nor for the conditions under which it should be measured.

“Fasting insulin is not an optimal tool for the individual assessment of peripheral insulin sensitivity…. The current level of evidence does not support the development of a definition of insulin resistance by fasting insulin,” said Dr. Levy-Marchal, of Robert Debre Hospital, Paris.

The committee indicated the need for a strong stance against testing for insulin resistance because the practice is common, Dr. Silva Arslanian, a panel member, said. “We get a lot of referrals of children with a ‘high insulin level’ and meanwhile the child is obese and the parent was never told that the child is obese…. That’s why the insulin level is high.”

Measuring insulin levels is an unnecessary health care expenditure, added Dr. Arslanian, the Richard L. Day Endowed Professor of Pediatrics at the University of Pittsburgh. “We’re already talking about how expensive health care costs are in the United States. Why do that when your eyes can tell you — or the body mass index can tell you. If you’re obese, the insulin level will be higher. You treat the obesity and the insulin comes down. You don’t treat the insulin.”

Dr. Alan Sinaikoof the University of Minnesota, Minneapolis, described the various methods used to measure insulin sensitivity/resistance. The hyperinsulinemic euglycemic clamp, which assesses the amount of glucose taken up by the tissues in response to infused insulin, remains the standard method. The frequently sampled intravenous glucose tolerance test (known as the “minimal model”) has the advantage of also assessing beta-cell function.

A third measure, the steady-state plasma glucose, uses a steady glucose infusion — rather than insulin, as the clamp does — to see how much remains in the blood after uptake by the tissues. And, a modified oral glucose tolerance test called the “whole-body insulin sensitivity index,” uses more glucose samples than does the usual OGTT.

All of these tests require research-like settings. They’re time consuming — typically 2-3 hours — expensive, and not covered by insurance. “It’s impractical to do this type of testing in any kind of clinic or general practice,” Dr. Sinaiko said.

Several attempts have been made to develop more practical tests for insulin sensitivity. In addition to simply measuring fasting insulin levels, others include the homeostasis model assessment (HOMA), the quantitative insulin-sensitivity check index (QUICKI), the fasting glucose:insulin ratio, and the MacCaulay method.

None of these formulas are any better than fasting insulin alone, however, because all of them incorporate that number into the calculation, and data from adults and children have demonstrated a wide range of fasting insulin levels for any given insulin sensitivity as determined by the clamp. The lack of advantage of any of these surrogate tests over either fasting insulin alone or the clamp has been shown in numerous adult studies and in a study Dr. Sinaiko and his associates conducted in children (Diabetes Care 2008;31:783-8).

The draft document states that fasting insulin is a weak measure of insulin resistance, and its measurement should be restricted to large cohorts of children, not individuals, he said.

Dr. Arslanian reviewed the literature regarding risk factors for insulin resistance in children, including obesity, high body-mass index, and high waist circumference. African American children are at greater risk, as are those entering puberty, when insulin sensitivity declines an average of 30%. Polycystic ovary syndrome also confers an increased risk as does intrauterine exposure to a mother’s diabetes during pregnancy. Nonalcoholic fatty liver disease is also strongly associated with insulin resistance in children, she said.

Dr. Wayne Cutfield, of the University of Auckland, New Zealand, summarized the literature regarding treatment, which provides strong evidence for the use of lifestyle modification — comprising nutrition therapy, physical activity, and education about both — as first-line therapy for obesity/insulin resistance.

Low-calorie diets have proven benefit in improving insulin sensitivity in children via weight loss, and appear to be superior to either low-fat or low-glycemic index diets in children, although there are some data supporting the latter. Fiber has been shown to improve insulin sensitivity in adults but not enough data are available in children, Dr. Cutfield said.

Weight-reducing drugs — specifically sibutramine and orlistat — do improve insulin sensitivity in children based on five randomized trials, but the effect is thought to be mediated through weight loss rather than a direct effect.

Metformin improves insulin sensitivity in adolescents with Type 2 diabetes and in those with polycystic ovary syndrome, but it is unclear whether metformin improves insulin sensitivity in obese adolescents, and no study has shown that it can prevent or delay the development of glucose intolerance in that population. Therefore, metformin cannot be recommended for the treatment of insulin resistance or obesity in children and adolescents, Dr. Cutfield said.

Dr. Chiarelli concluded the session with a summary of the group’s recommendation for prevention of insulin resistance in children, including efforts to reduce maternal and childhood obesity, and the promotion of breast-feeding as a means of reducing obesity for the child later in life. The breast-feeding recommendation sparked some debate and received only a “C” level of evidence, but there are data to support it, he said.