Reliable Assessment of Insulin Resistance in Children

Abstract

Obesity is on the rise in the pediatric population. Obese individuals are at risk for insulin resistance, and this in combination with impaired β-cell function can result in type 2 diabetes. The diagnosis of insulin resistance is made in clinical practice by using estimates of insulin sensitivity. Scientific evidence regarding the evaluation of insulin resistance in children has revealed conflicting conclusions.

Introduction

Rates of obesity in children are rising at an alarming rate in the United States and have tripled since 1980 [1]. Data based on the National Health and Nutrition Examination Survey (NHANES) reveal that of children aged 2-19 years, nearly 17% had a body mass index (BMI) at the 95^th percentile or greater for age, and nearly 32% at the 85^th percentile or greater for age [1]. These trends in obesity are especially concerning given its association with insulin resistance and possible progression to type 2 diabetes.

The pancreatic beta cells produce insulin. Insulin resistance refers to a loss of sensitivity of the peripheral tissues to insulin. Individuals with insulin resistance who have sufficient beta cell function can compensate with higher insulin levels to maintain euglycemia. However if beta cell function declines in the setting of insulin resistance, glucose tolerance becomes impaired, putting the individual at risk for the development of type 2 diabetes [2-5]. β-cell dysfunction has been studied in obese adolescents, and is related to glucose tolerance. During an oral glucose tolerance test (OGTT), of those obese adolescents whose 2 hour glucose levels were in the normal range, progressively higher 2 hour glucose values (though normal) were associated with lower insulin sensitivity and decreased insulin secretion, indicating an early beta cell dysfunction [2]. Worsening β-cell function in comparison to insulin sensitivity has also been demonstrated in overweight children with normal glucose tolerance whose 2 hour glucose values were in the higher part of the normal range [3]. In addition, obese adolescents with the higher (though normal) 2 hour glucose values were more likely to develop impaired glucose tolerance at a 2 year follow up [2].

Insulin resistance is one of the key features of the metabolic syndrome. Additional features of the metabolic syndrome include obesity, hypertension, and dyslipidemia [4, 5]. The metabolic syndrome is associated with an increased risk of type 2 diabetes and heart disease [5]. While insulin resistance is usually found in individuals who are obese, not all obese individuals have insulin resistance. There are factors beyond simply being obese that determine whether an individual will develop insulin resistance. Those with other features of the metabolic syndrome are more likely to have insulin resistance [6].

Pathogenesis

There are multiple factors in addition to obesity that are currently believed to contribute to developing insulin resistance. The distribution of fat seems to play an important role in developing insulin resistance. Individuals with a higher proportion of visceral fat tend to have poorer glucose tolerance [6]. One current theory is the expandability hypothesis, which describes the subcutaneous tissue as having a maximum storage capacity, and when this is reached excess fat is deposited in other areas [7]. Based on this theory, individuals have differing storage capacities in the subcutaneous fat, which affects how the stored fat is distributed. Those with more capacity to store fat in the subcutaneous tissue tend to have a more favorable metabolic profile [7].

In obese adolescents, those with more visceral fat and less subcutaneous fat were found to have fewer large adipose cells, these cells overall had a larger diameter, and they had decreased expression of markers of adipogenesis/lipogenesis [8]. This type of profile has been linked to insulin resistance and developing type 2 diabetes [8].

As a result of this abnormal tissue expansion, there seems to be resulting inflammation [9]. The inflammatory cells present in the adipose tissue due to this disturbance will release inflammatory markers, which in turn have detrimental effects on insulin action [9]. These include IL-6, TNF-alpha, and more recently the role of Gas6 has been described [9].

Genetic predisposition plays a significant role in developing insulin resistance. Genetic studies are ongoing to identify the genes which predispose to insulin resistance and type 2 diabetes. Those genes with higher odds ratios of developing type 2 diabetes include TCF7L2 and PTPRD [10].

Identifying those at risk

While criteria to screen adolescents for insulin resistance are lacking, criteria to screen for type 2 diabetes are clear. The American Diabetes Association (ADA) recommends screening for type 2 diabetes in children whose BMI or weight for height is greater than the 85^th percentile, or in children whose weight is significantly higher than the ideal weight for height. Children should be screened for type 2 diabetes if they meet the aforementioned criteria and additionally have two risk factors; 1^st or 2^nd degree relative with type 2 diabetes, high risk race or ethnicity, physical signs of insulin resistance, or exposure to gestational diabetes [11]. A diagnosis of type 2 diabetes is rare in children less than 10 years of age, and the ADA recommends screening in children who are 10 years of age or older, or at puberty if before age ten [11]. Puberty is a time of physiologic insulin resistance, thus children with insulin resistance may have worsening of this condition during puberty [4].

At the time of the most recent consensus statement in 2010, there was no evidence to justify the screening of lean or obese children for insulin resistance [12]. For both obesity and insulin resistance the treatments are the same - diet and exercise. Thus an obese child should be treated regardless of whether he or she has insulin resistance. Furthermore general screening is not feasible as standards to diagnose insulin resistance are lacking [12].

However we believe that screening for insulin resistance can be a helpful tool in adolescents. We recommend screening those obese adolescents who have a risk factor in addition to obesity (see Table 1). Detection of insulin resistance can allow for more rigorous intervention to halt the progression to type 2 diabetes. While the current standard of care treatment for insulin resistance is a combination of diet and exercise, a more rigorous lifestyle intervention (enrollment in a structured program) or glugophage may be used by some providers. In those cases identification of insulin resistant adolescents can be worthwhile.

Table 1.

Consider evaluating for insulin resistance in obese adolescents with ≥ 1 risk factor

History	GDM, SGA, family history of type 2 diabetes or prediabetes
Ethnicity	Consider in all ethnicities, especially Native American, African American, Latino, Asian America, and Pacific Islander
Exam	Acanthosis nigricans
Comorbidities	Hypertension, PCOS, high triglycerides

Multiple clinical characteristics can help to identify those at highest risk of insulin resistance, and thus type 2 diabetes. A family history of type 2 diabetes or exposure to gestational diabetes (GDM) increases the risk of type 2 diabetes [11]. There is an association of insulin resistance in children with polycystic ovary syndrome (PCOS), dyslipidemia, and hypertension, and in those were born small for gestational age (SGA). Populations at highest risk for type 2 diabetes include Native American, African American, Latino, Asian America, and Pacific Islander children [11]. Acanthosis nigricans is a physical exam finding associated with high levels of insulin, and manifests as velvety and hyperpigmented skin, in areas of skin folds, often most prominent on the neck and in the axillae [13].

Testing

The gold standard to assess insulin sensitivity is a euglycemic hyperinsulinemic clamp [12]. This technique was first described by DeFronzo in 1979. A continuous insulin infusion is used to maintain the plasma insulin concentration at a fixed, elevated level. Glucose is infused simultaneously, and the rate is adjusted to maintain a fixed blood sugar. The amount of glucose required to maintain a steady blood sugar reflects how sensitive the peripheral tissues are to the insulin infusion [14].

The frequently sampled IV glucose tolerance test (IVGTT) is an additional method to measure insulin sensitivity [12]. The IVGTT involves administration of a glucose load via IV, and frequent sampling of blood sugars for a specified amount of time afterwards. These tests are reserved for the research setting, as they require much time, multiple blood samples, and have a high cost [12, 15]. For these reasons additional, simpler, methods to measure insulin resistance have been explored, including fasting labs and sensitivity indices. Screening tests have been more extensively studied in adults. There are limited studies in the pediatric population.

Hemoglobin A1c

Hemoglobin A1c is used to indicate prediabetes if the value is 5.7-6.4 percent [11]. This reflects overall blood sugar, and is not indicative of insulin levels. The TODAY study evaluated adolescents with type 2 diabetes, and found that the hemoglobin A1c correlated well with β-cell function but did not correlate with the measures of insulin sensitivity [16]. Overweight/obese adolescents with hemoglobin A1c in the prediabetes range were found to have decreased β-cell function in comparison to insulin sensitivity [17].

Fasting labs

Fasting blood work would be an ideal way to evaluate insulin sensitivity, as it requires a single lab draw. This is more convenient for the patient and is less costly than clamps or tolerance testing. Various methods have been developed to determine insulin sensitivity based on a fasting blood sample alone, including measurements of fasting glucose, fasting insulin, and a fasting glucose to insulin ratio. However these methods lack ability to see elevations in insulin in response to a stimulus, as is seen in obese individuals with peripheral resistance to insulin [18].

Fasting insulin and glucose are commonly used to evaluate children and adolescents for impaired fasting glucose and fasting hyperinsulinemia. Gungor evaluated 156 healthy obese and nonobese children (25 with PCOS, 10 of these with impaired glucose tolerance) for insulin sensitivity using a euglycemic hyperinsulinemic clamp. Fasting insulin (r² = 0.84, p < 0.001) and the ratio of fasting glucose to fasting insulin (r² = 0.84, p < 0.001) were both well correlated with the clamp [15]. Similar results were seen by George in 188 overweight and obese pubertal subjects, including those with normal glucose tolerance, impaired glucose tolerance, type 2 diabetes, and type 1 diabetes (obese). These individuals had a 2 hour OGTT and a euglycemic hyperinsulinemic clamp. Correlations with the clamp were found in 1/fasting insulin (r = 0.82 and p < 0.001) and fasting insulin (r = 0.78, p < 0.05) [19]. Evaluation of 1/fasting insulin was found to be congruent with insulin sensitivity in 146 overweight/obese youth who had a clamp [17].

By far the largest clamp study evaluating surrogate markers of insulin sensitivity in children, Schwartz and associates evaluated 323 pubertal adolescents who underwent a euglycemic hyperinsulinemic clamp. They found that both the fasting insulin and fasting glucose to insulin ratio were not well correlated with insulin sensitivity as measured by the clamp [20].

The triglyceride-to-HDL cholesterol (TG-HDL-C) ratio is another fasting measure that has been studied in adults and is related to insulin resistance. In applying the TG-HDL-C to adolescent subjects, a higher ratio was associated with lower Whole Body Insulin Sensitivity Index (WBISI), and predictive of insulin resistance in white adolescents [21]. Significance was not reached in other ethnic groups, thus additional studies are needed to identify ethnicity-specific limits [21].

HOMA

The Homeostatic Model Assessment (HOMA) was developed using fasting insulin and glucose measurements to evaluate β-cell function and approximate insulin resistance, and is calculated as $\frac{fastinginsulinxfastingglucose}{22.5}$, using μU/ml for insulin and mmol/l for glucose [22]. When initially evaluated in comparison to a euglycemic hyperinsulinemic clamp in 12 healthy adults and 11 diabetic adults, HOMA correlated better with the clamp in the diabetic individuals (r = 0.92, p < 0.0001) than in the healthy adults (r = 0.83, p < 0.01) . HOMA-IS (insulin sensitivity) is the reciprocal of HOMA-IR (insulin resistance) [19], and both of these measures have been used in pediatric studies to evaluate insulin sensitivity.

In Gungor’s study of obese and nonobese children noted above, the HOMA-IS and euglycemic hyperinsulinemic clamp were correlated (r² = 0.82, p < 0.001) [18]. Similar results were obtained by George in overweight and obese pubertal children, where the HOMA-IS was correlated with the euglycemic hyperinsulinemic clamp (r = 0.81, p < 0.05) [21].

Yeckel previously studied 26 children aged 8-18 years with normal glucose tolerance and 12 children with impaired glucose tolerance. Data from the euglycemic hyperinsulinemic clamp indicates that unlike the results above, there was less concordance between the HOMA-IR (r = −0.57, p < 0.005) and clamp study [18]. Schwartz and colleagues also found that HOMA was not well correlated with insulin sensitivity [20].

OGTT: WBISI

The OGTT is commonly used to further evaluate a child who is found on fasting blood work to have elevated glucose or insulin. Results obtained during an OGTT can be used to diagnose diabetes (fasting blood sugar ≧ 126 or 2 hour plasma glucose ≧ 200) or prediabetes (fasting blood sugar 100-125 or 2 hour plasma glucose 140-199) [11]. Serum insulin measurements are needed during the OGTT to evaluate for insulin resistance using the following surrogate indices.

The WBISI or Matsuda Index, uses glucose tolerance test results to measure insulin sensitivity and is calculated as $\frac{10,000}{\sqrt{(fastingglucosexfastinginsulin)(meanglucosexmeaninsulin)}}$, using μU/ml for insulin and mg/dl for glucose [18]. In the initial study comparing data obtained from adults with normal and impaired glucose tolerance and type 2 diabetes, WBISI correlated with the euglycemic hyperinsulinemic clamp (r = 0.73, p < 0.0001) [23]. In the overweight and obese pubertal subjects studied by George, the euglycemic hyperinsulinemic clamp and WBISI were correlated (r = 0.77, p < 0.05) [19]. Similar results were seen in Yeckel’s evaluation of children with normal glucose tolerance and impaired glucose tolerance WBISI (r = 0.78, p < 0.005) [18].

DI

The disposition index (DI) can be calculated during an OGTT or a clamp, and is calculated by multiplying the insulin sensitivity and β-cell function [24]. The exact calculations vary depending on the test performed. The insulin sensitivity measurement is based the fasting insulin value, and the β-cell function component evaluates changes in insulin and glucose during the first phase insulin response. The DI obtained during a euglycemic hyperinsulinemic clamp correlated with that obtained during an OGTT in 185 overweight/obese adolescents, who were categorized by varying degrees of glucose tolerance [25].

Conclusions

Hemoglobin A1c can be used to screen for type 2 diabetes or prediabetes, but should not the solitary investigation in pediatric patients. There are conflicting results in the pediatric population regarding the use of fasting insulin and fasting glucose alone or as a ratio to screen for insulin resistance. However given that these measures require only one blood draw they may be preferable to undergoing evaluation via OGTT. HOMA-IR also has conflicting results, however it may be an acceptable option given the ease and low cost of obtaining fasting blood work. WBISI and the DI are better measures of insulin resistance and would be the best choices to diagnose insulin resistance if the patient can be evaluated with an OGTT.

Surrogate Marker	Ease of Use	Reliability
Hemoglobin A1C	Simple	Not useful
Fasting insulin	Moderate	Good
Fasting glucose	Moderate	Good
Fasting TG-HDL-C ratio	Moderate	Poor
HOMA	Moderate	Poor
WBISI	Difficult	Best
DI	Difficult	Best

Insulin resistance is a significant finding in adolescents as it can be the first indication that the individual is on track to develop type 2 diabetes. Intervention at this very early stage is crucial in halting the progression to type 2 diabetes. The importance of identifying insulin resistance and beta cell dysfunction early and engaging the adolescent in treatment is emphasized by the TODAY study, which has shown that type 2 diabetes is extremely difficult to treat in this population [26].