The prevalence of obesity among US children was recorded to be at 18.5% as per the National Survey of Children’s Health (NSCH) in 2016 versus 10% in the 2003 survey (1). Total body adiposity, especially abdominal fat, increases risk factors for type 2 diabetes, insulin resistance, cardiovascular disease, dyslipidemia, non-alcoholic steatohepatitis (NASH), and circulating concentrations of pro-inflammatory cytokines (1–8).

Fasting hyperinsulinemia is associated with insulin resistance (9, 10), but intravenous and oral glucose tolerance tests (IVGTT and OGTT) predict better insulin resistance in obese children despite having similar fasting serum insulin levels (11–13). A number of indices of insulin sensitivity such as fasting insulin, quantitative insulin sensitivity check index (QUICKI), homeostatic model assessment for insulin resistance (HOMA), acute insulin response (AIR), and glucose disposal index (GDI) have variable correlations with the so-called gold standard of a euglycemic clamp study and with body fatness. Minimal model IVGTT (a 180-min test with frequent blood sampling) and euglycemic–hyperinsulinemic clamps are a more accurate method of quantifying insulin sensitivity (Si), GDI, and AIR (14, 15). Euglycemic clamp studies, though considered the gold standard in measuring insulin sensitivity, are impractical in a pediatrician’s office as they required serial measures during several hours and require complicated calculations of insulin sensitivity indices (16). There is a need to identify an early, convenient, and simple marker for assessing insulin resistance in children instead of serial laboratory values and complex formulas.

Insulin-like growth factor binding protein-1 (IGFBP-1) is a 28-kD protein, mainly produced in the liver, and its secretion is inhibited by insulin (17, 18). IGFBP-1 was shown to be useful in the assessment of whole-body insulin sensitivity and is an adiposity-independent liver-specific circulating marker of insulin sensitivity in adults and adolescents (4, 19–21). IGFBP-1 may constitute as an independent risk factor for the development of insulin resistance, increased risk of type 2 diabetes, and other adiposity-related comorbidities in children, even at a young age. Circulating concentrations of IGFBP-1, coupled with fasting levels of insulin, have been proposed as a more specific screening test for insulin resistance in children (5, 22, 23). We propose that IGFBP-1 should be considered as marker of insulin resistance in children, and it is easier to obtain than other commonly used indices of insulin sensitivity.

The mean BMI of the study group was 22.7 ± 4.9 kg/m² (range: 13.8–36.8 kg/m²). Eighteen students (18%, 6 females, 12 males) had BMI more than 2 SDS for age and sex (Table 1). Total cholesterol (172 ± 34, vs. 159 ± 27, p < 0.05) and adiponectin (13.4 ± 4.8 vs. 11.1 ± 4.6, p < 0.01) were significantly higher in males. CRP was significantly higher in females (4.8 ± 5.8 vs. 2.8 ± 3.4 ng/ml, p < 0.05).

Correlational analyses were performed relating circulating concentrations of IGFBP-1, anthropometric data, and indices of insulin sensitivity and dyslipidemia (Table 2). Circulating concentrations of IGFBP-1 were significantly and inversely correlated with age and all indices of body fatness such as body fat %, BMI (shown in Figure 2), and waist circumference (shown in Figure 3). IGFBP-1 correlated more strongly with waist circumference (Table 2, r = -0.62; p-value <0.001). Inverse correlation was seen with fasting insulin (shown in Figure 4), AIR, HOMA, TG, LDL, Chol/HDL, and TG/HDL ratios. In addition, circulating IGFBP1 concentrations were positively correlated with adiponectin, QUICKI, and HDL cholesterol (Table 2). The ROC curve for IGFBP-1 was constructed, and the AUC was high at 0.92. The cutoff value of IGFBP-1 levels of <5 had a sensitivity of 94% and a specificity of 67% to identify insulin resistance in children (Figure 5). There were no differences in the fasting insulin and IGFBP1 levels between the various racial/ethnic groups.

Peripheral serum insulin levels are not a true reflection of secreted insulin, as almost 50% of secreted insulin is degraded passing through the portal circulation (14). Elevated insulin levels in the portal system are the initial manifestation of insulin resistance, and later peripheral insulin levels increase. IGFBP-1 levels are regulated via an insulin-response element in the IGFBP-1 promoter region, which explains the mechanism of insulin inhibition of IGFBP1 expression. Portal insulin directly inhibits the transcription of hepatic IGFBP-1 (18, 30, 31). In experiments, insulin infusion results in a suppression of plasma IGFBP-1 concentrations (32). Infants and children with hypoglycemia due to congenital hyperinsulinism have significantly lower IGFBP-1 (33). Low fasting serum IGFBP-1 levels reflect portal hyperinsulinemia even before peripheral hyperinsulinemia is detected (14). IGFBP-1 is higher in individuals with type 1 diabetes (hypoinsulinemic) (34), decreased in those with type 2 diabetes (hyperinsulinemic), and significantly diminished by insulin administration (35). In type 1 diabetes with insulinopenia, the IGFBP-1 levels are high because of low levels of portal insulin. The data in our study have confirmed a negative correlation between fasting IGFBP-1 and fasting serum insulin (r = -0.51; Table 2 and Figure 4) and AIR (r = -0.39; Table 1). Travers et al. showed that insulin sensitivity derived from rapidly sampled IVGTT correlated strongly with IGFBP-1 levels in children (13). In our study, new findings were that IGFBP-1 independent of BMI negatively correlated with AIR and HOMA and positively correlated with adiponectin and TG/HDL ratio (Table 3 and Figure 2). Motaghedi et al. in their study of obese children also revealed that insulin sensitivity reported from rapidly sampled IVGTT data correlated more strongly with IGFBP-1 than other measures of insulin sensitivity (5). A low level of IGFBP-1 is a marker for metabolic syndrome risk in both adults and children (36, 37).

Toddlers who were small for gestational age at birth have an increased lifetime risk of development of type 2 diabetes, and they also have significantly lower IGFBP-1 levels even before manifesting insulin resistance (23). Reinehr et al. and others have reported significant positive correlations of IGFBP-1 with various indices of insulin sensitivity in children with obesity (36, 38–40). IGFBP-1 has also been negatively correlated with the slope of trajectories of body weight gain across a spectrum of adiposity in adults and children (41). Katz et al. reported positive significant correlations of IGFBP-1 with adiponectin even controlling for insulin sensitivity, fat mass, body fat, and waist circumference. In this study, the correlation of IGFBP-1 with waist circumference was stronger than that compared to BMI and percent body fat (Table 2). The results in our study are in agreement with other similar studies in children (13, 36).

The major finding of this study was that the circulating concentration of IGFBP-1 can identify the presence of insulin resistance in the group with BMI 1 to 2 SDS, even before the known markers of insulin resistance such as elevated TG and even before decreased HDL and adiponectin levels are identified. There were no differences between groups with more than 2 SDS and 1 to 2 SDS in waist circumference. This demonstrates that decreased IGFBP-1 in obese/overweight children is simply not a consequence of increased fatness/adiposity, but it may reflect other independent metabolic processes conveying higher risk for dyslipidemia and type 2 diabetes. IGFBP-1 in this cohort of school-aged subjects is potentially an independent risk factor for these comorbidity risks. There was no difference seen in insulin and IGFBP1 levels in the various racial groups, however. The unique aspect of this study is to report the relation of IGFBP-1 in school-age children from very diverse racial and ethnic backgrounds.

One of the limitations of the study was that pubertal exam in the subjects was not performed due to the nature of the setting and insulin sensitivity is known to decrease in the mid-puberty age group (42).

The TG/HDL ratio was first proposed by Reaven et al. as another marker of insulin resistance, supported by other studies (43, 44). Multiple studies have shown the inverse correlation of IGFBP-1 levels with a TG/HDL ratio in adults (2, 3, 45). A lifestyle interventional study in obese subjects also revealed a negative correlation between IGFBP-1 and serum TG levels (36). The TG/HDL ratio in the present study was significantly and inversely correlated with IGFBP-1 levels (r = -0.26, p < 0.01; Table 2). This reinforces the fact that IGFBP-1 can be used as a marker of insulin resistance in children at high risk.

Adiponectin, an adipocyte-derived cytokine, reduces levels of free fatty acids in blood and has been associated with improved lipid profiles, better glycemic control, and reduced inflammation in diabetic patients (46). Low levels of adiponectin play a role in the development of insulin resistance and subsequent type 2 diabetes (47). Adiponectin levels have been shown to have a significant positive correlation with IGFBP-1 in adult studies (48). The present study also showed a positive correlation between IGFBP-1 and adiponectin levels (Table 2 r = 0.28, p < 0.005).

In summary, low IGFBP-1 can be detected during childhood with a high risk for the development of insulin resistance. This cross-sectional study revealed the role of IGFBP-1 as an independent marker of insulin resistance and adiposity in this cohort of multiethnic diverse groups of schoolchildren. Although this is only a cross-sectional pilot study, it is necessary to perform similar studies on a larger and prospective scale including lean and obese children and children with metabolic syndrome. Our data indicate that by measuring IGFBP-1, the practitioners will have a reliable marker of insulin resistance.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The clinical protocol was approved by the Institutional Review Board for each participating hospital, and for the New York City Department of Education. Written informed consent to participate in this study was provided by the participants’ legal guardian/next of kin.

Authors AB and RG contributed equally to the preparation of this manuscript the same as the first author. Authors MR and ST contributed equally to the preparation of this manuscript as a last author. RG, AB, and ST were involved in clinical care of the participants and their follow-up. Authors SS, DC, SA, LA, BL, and WR were involved in editing and revision of the manuscript. AB, ST, MR, PS, RR, and IF did a critical analysis of the manuscript and its editing. All authors contributed to the article and approved the submitted version.

The study was supported by AMDeC, the Starr Foundation, and NIH grant numbers UL1RR0023568 and UL1RR0024156.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The reviewer EG declared a shared affiliation with the author RR to the handling editor at the time of review.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.