Evaluation of metabolic score for insulin resistance index as a sensitive marker of insulin resistance and its ability to predict hepatic steatosis in obese children and adolescents. A pilot study

INTRODUCTION

Insulin resistance (IR) is a common health issue in medical practice. We can define IR as decreased tissue sensitivity to insulin. Obesity is the primary risk factor for IR. It is well known that IR is predisposed to developing several metabolic disorders, such as hyperglycaemia, hepatic steatosis, dyslipidaemia, and hypertension in obese individuals. Insulin resistance is a state of dysfunction as there is decreased sensitivity and responsiveness to insulin despite a normal or elevated serum insulin concentration [1]. Not all obese children present IR, while recent evidence supports that insulin resistance could be met even in individuals with normal weight [2, 3]. The causal directions of these relationships remain mostly untested. Fat distribution is a more important factor for IR than total weight. Hence, children with normal weight who have fat around their middle and excess visceral fat can develop IR and hepatic steatosis. For obese children, hyperinsulinaemia is a compensatory mechanism to allow insulin action in mild to moderate IR. Such a defect in insulin signalling not only impairs the utilisation of glucose and lipids, but also causes organ complications like metabolic associated fatty liver disease [1].
The homeostatic model assessment (HOMA) is the most commonly used parameter to evaluate IR. It has the disadvantage that an abnormal HOMA-IR result may be false, especially in the absence of other metabolic disturbance components [4, 5]. To overcome this limitation, the triglyceride-glucose (TyG) index has been developed [6]. The triglyceride-glucose index is considered a reliable alternative biomarker of IR [6]. The triglyceride-glucose index is a very simple screening method to evaluate insulin resistance. It requires an evaluation of fasting triglycerides and fasting plasma glucose (FPG). It is used to predict cardiovascular risks in individuals [7, 8]. Recently, the use of the TyG index, as a predictor of cardiometabolic risk factors in childhood and adolescence, has also increased [9, 10]. Several studies considered the TyG index as a marker of IR to identify adults at risk for non-alcoholic fatty liver disease (NAFLD) [11]. Central IR is located in the liver. Hepatic insulin resistance refers to the impaired suppression of glucose production by insulin in hepatocytes. This usually manifests as impaired fasting glucose and NAFLD [4, 12, 13].
Increasing evidence suggests that the metabolic score for insulin resistance (METS-IR) is a novel indicator to replace the traditional indexes to estimate the IR level. The metabolic score for insulin resistance involved more lipid types, so it evaluated more metabolic status and could represent metabolic status and IR status [14–16]. Hepatic steatosis is characterised by fat accumulation in more than 5% of hepatocytes [17], without excessive alcohol consumption.
The aim of this study was to investigate the association between the HOMA-IR, TyG, and METS-IR indexes as markers of insulin resistance and the ability of METS-IR to predict hepatic steatosis in obese children and adolescents.

MATERIAL AND METHODS

STUDY POPULATION

We performed a cross-sectional study. The study protocol was approved by the Human Ethics Committee of the National Research Centre, and written informed consent was obtained from all children and their parents. We enrolled 190 subjects: 50 normal-weight children 50 children with overweight and 90 with obesity. This study is part of project no. 10010324 carried out in the National Research Centre, Cairo, Egypt. They were recruited from the Paediatrics Clinic at the NRC. All non-obese volunteers were age-matched healthy subjects in apparent good health and taking no medications. A child was obese if their body mass index (BMI) was > 95% percentile for age and gender percentile curves of growth for our population [18]. The studied participants were classified into 3 groups according to body mass index: normal weight (BMI 18.5–24.9), overweight (BMI 25–29.9), and obesity (BMI ≥ 30 kg = m²).
Exclusion criteria: patients with any of the following criteria were excluded from the study: hepatobiliary diseases, chronic liver diseases including viral hepatitis malignancies, ascites, medications known to cause hepatic steatosis (such as oestrogens, corticosteroids, amiodarone, and valproate; at present or within the last 2 years), anti-inflammatory drugs, inflammatory bowel disease, human immunodeficiency virus, renal diseases, hypothyroidism, Cushing syndrome, or Turner syndrome, as well as obesity with mental retardation, such as Prader-Willi syndrome, Laurence-Moon-Biedl, and Cohen syndrome.

CLINICAL EXAMINATION

The following were performed on all studied groups:
1) Full history taking through clinical examination, with emphasis on any complications or medications;
2) Blood pressure measured according to American Heart Association guidelines;
3) Anthropometric indices: body weight measured to the nearest 0.1 kg with a balance scale and height measured to the nearest 0.1 cm.
Body mass index was calculated as weight divided by height squared (kg/m²). Waist circumference was measured at the level midway between the lowest rib margin and the iliac crest. Hip circumference was measured at the widest level over the greater trochanters in a standing position by the same examiner; then the waist to hip ratio (WHR) and waist to height ratio (WHTR) were calculated. Percentage body fat (BF %) was calculated by the equation: child BF % = (1.51 × BMI) − (0.70 × age) − (3.6 × gender) + 1.4 where (female = 0 and male = 1) [19].

THE LABORATORY MEASUREMENTS

Five millimetres of venous blood were withdrawn under complete aseptic precautions from fasting subjects (12–14 hours). Pertinent laboratory investigations were performed for all patients and controls including glycosylated haemoglobin (HbA1c) levels, lipid profile (cholesterol, triglyceride, high-density lipoprotein – HDL, and low-density lipoprotein), and liver enzymes including aspartate aminotransferase and alanine aminotransferase (ALT), using an Olympus AU 400 supplied from Olympus Life and Material Science (Europe GmbH, Wendenstraße, Hamburg, Germany). Viral Markers (HB s Ag, HCV Ab, and HIV Ab) were assessed to exclude viral hepatitis in our cases using the PRECHECK Kit (USA). FBS and lipid profile were assessed using an OLYMPUS AU 400 chemistry analyser. Insulin was estimated by enzyme immunoassay (ELISA).
Insulin resistance was calculated by the homeostasis model (HOMA-IR) using the following formula: HOMA-IR = fasting insulin (mU/l) × fasting glucose (mmol/l)/ 22.5 [4, 9]. Triglyceride-glucose index was calculated with the use of the mathematical formula: natural logarithm – Ln [fasting triglycerides (mg/dl) × FPG (mg/dl)]/2 [9].
The metabolic score for IR (METS-IR) = Ln [(2 × fas-ting blood glucose (mg/dl) + fasting triglyceride (mg/dl)] × BMI (kg/m²))/(Ln [HDLc (mg/dl)] [15].
Abdominal ultrasonography was performed, by the same expert, for liver size and echogenicity. Normal liver parenchyma has a homogeneous echotexture with echogenicity equal to or slightly greater than that of the renal cortex and spleen. The liver reveals echogenicity more than the kidney and spleen due to fatty infiltration [20]. Various (0–3) grades of steatosis have been proposed based on analysis of the intensity of the echogenicity [21]. The model of ultrasound apparatus is SA-R3 (No S06YM3 HDC00012F) SAMSUNG MEDISON Company – South Korea.

STATISTICAL ANALYIS

Each variable was assessed for normal distribution. The standard computer program SPSS for Windows, release 17.0 (SPSS Inc., USA) was used for data entry and analysis. All numeric variables were expressed as mean ± standard deviation. The intergroup comparisons were performed by using an independent-sample t test and a one-way analysis of variance followed by Scheffe’s post hoc test. Categorical variables are presented as numbers and percentages (%). The 2 groups were compared using the χ² test. Pearson’s correlation tests (r = correlation coefficient) were used for correlating normal. The potential of the IR indices and NAFLD was examined using receiver operator characteristic (ROC) curve analysis and analysis of variance. For all tests, a probability (P) less than (< 0.05) is considered significant, and p < 0.01 was considered highly significant.

STUDY LIMITATIONS

This study is a single-centre cross-sectional study, so these findings need to be further verified by a multicentre prospective cohort study. The small sample size of the included cases led to difficulty in getting solid conclusions. Also, physical activity was not evaluated. Despite these limitations, this study could have important clinical implications.

RESULTS

A total of 190 children were included in our study: 50 normal-weight children, 50 children with overweight, and 90 children with obesity. Our results showed that among the individuals with obesity, 62 (68.9%) were female and 28 (31.1%) were male. Among the individuals with overweight, 27 (54.0%) were female and 23 (46.0%) were male. Among the control individuals, 35 (70.0%) were female and 15 (30.0%) were male. There were no significant statistical differences between groups in terms of age or gender (p > 0.05). Descriptive data (clinical, anthropometric, and ultrasonography characteristics and laboratory parameters) of the studied groups are summarised in Tables 1 and 2.
Our data revealed that there were significant statistical differences between them as regards BMI, WHTR, WHR, % body fat, visceral fat thickness (VFT), and liver span (p = 0.000). Systolic and diastolic BP levels were significantly higher in overweight and obese children (p = 0.007 and 0.003, respectively).
Significantly greater values of HOMA-IR, TyG index, and METS-IR were found in overweight and obese cases (p = 0.000). Figure 1 shows the values of METS-IR index in normal weight, overweight, and obese cases.
Role of gender: no differences in VFT, TyG index, and METS-IR were observed between males and females in the 3 groups (p = 0.484, 0.430, and 0.940, respectively).
The studied individuals were classified into 2 groups without hepatic steatosis and with hepatic steatosis. The number of hepatic steatosis cases was 37 (19.5%) in all studied children. In children of normal weight only one (2.0%) had hepatic steatosis, in children with overweight 4 (8.0%) had hepatic steatosis, and 32 (90%) of the children with obesity. Hepatic steatosis was significantly higher in boys than girls [22 (59.5%), 15 (40.1%), respectively (p < 0.001)]. Table 3 shows the clinical and laboratory characteristics of the groups with hepatic steatosis and without hepatic steatosis. The results showed statistically significant differences in BMI, WHR, BF %, subcutaneous fat thickness, visceral fat, and ALT between the hepatic steatosis group and the group without hepatic steatosis. In our results disregarding ALT, there was a statistically significant difference between cases without hepatic steatosis and cases with hepatic steatosis (p = 0.013). There were statistically significant differences in the METS-IR) and the TyG index for NAFLD cases (p < 0.001 and < 0.005, respectively).
We found that the TyG index and METS-IR were significantly and positively correlated with ΗΟΜΑ-IR, VFT, fasting insulin, and BMI, (p = 0.000). The details are shown in Table 4.
The receiver operating characteristic curve was generated to detect the predictive capabilities of visceral fat, HOMA-IR, TyG index, and METS-IR for predicting hepatic steatosis (Figure 2). Details of the area under the ROC curve are shown in Table 5.

DISCUSSION

Insulin resistance is an increasing problem due to its association with obesity. To the best of our knowledge, this is the first study to investigate the association between the HOMA-IR, TyG index, and METS-IR insulin resistance indices in obese, normal-weight, and overweight children and adolescents from Egypt. Obese children and adolescents may be at high risk of developing hepatic steatosis if they have insulin resistance. Evidence supports that insulin resistance can cause hepatic steatosis even in individuals who are phenotypically normal weight [2–4].
Concerning hepatic steatosis, it is well known that this disease exhibits age and gender differences in both prevalence and severity. It was significantly higher in boys than girls [22 (59.5%), 15 (40.1%), respectively (p < 0.001)]. This result confirms the existence of a gender difference in the prevalence of hepatic steatosis, which has also been shown in previous studies [22–25]. Jeffrey et al. and Imanzadeh et al. speculate that obese adolescent boys have an increased prevalence of fatty liver compared with obese adolescent girls [22, 23]. Tominaga et al. reported that the prevalence of hepatic steatosis was 6.6% in males and 2.0% in females [24], while Gupta et al. stated that hepatic steatosis prevalence was 9.8% in girls and 22% in boys [25]. The higher prevalence of hepatic steatosis in boys may be due to excess abdominal fat and the effects of sex hormones.
There have been many hypotheses for the pathogenesis of obesity-associated IR – hyperinsulinaemia and lipotoxicity have been the major concepts. The question of whether insulin resistance or hyperinsulinaemia is the primary disorder of hepatic steatosis in children remains unclear. Our results confirm that there is a strong association between insulin and hepatic steatosis [26, 27]. We found that fasting blood insulin was significantly higher in paediatric hepatic steatosis. Kawasaki et al. [28] suggested that hyperinsulinaemia in obese children is the most important predictor of hepatic steatosis.
Our study has shown that significantly greater values of HOMA-IR TyG index were found in overweight and obese cases (p = 0.000). The homeostatic model assessment for insulin resistance had a positive correlation with the TyG index (r = 0.305, p < 0.000); similar correlation rates reported by Locateli et al. [29]. The triglyceride-glucose index level has been proposed as an alternative marker for identifying IR, when compared with the (HOMA -IR) index. Furthermore, a recent study provided evidence that an elevated TyG index is associated with the presence of cardiovascular risk factors in apparently healthy, normal-weight children and adolescents [30].
The metabolic score for insulin resistance calculation depends on FPG, triglyceride, BMI, and HDL-C, so it can better reflect the metabolic risk of patients with obesity. In this study, we showed that METS-IR is a novel marker for assessing the risk of IR. The metabolic score for insulin resistance is a score that is easy to obtain and calculate in the clinic. Our results indicate that obese children had a significantly higher mean METS-IR level [51.8571 ±7.72254, (p = 0.000)]. Furthermore, correlation with HOMA-IR demonstrated significant positive correlations with METS-IR. (r = 406, p < 0.000). Similar correlation rates have been reported by others [31].
Moreover, hepatic steatosis is expected to become a more serious public health issue because of the increasing prevalence of obesity. Obesity plays a significant role in the development of NAFLD, but the traditional IR indexes did not consider obesity, so they may not be suitable for evaluation in obese people. The metabolic score for insulin resistance combined BMI and IR level, on the other hand, in our study demonstrates that METS-IR was the best independent predictor for hepatic steatosis in obese children and adolescents. Similar results were reported by others [31, 32], also in adults [33, 34].

CONCLUSIONS

This study provided evidence that METS-IR is a sensitive marker of IR among obese children and adolescents, and it has a close relationship with hepatic steatosis. The metabolic score for insulin resistance index has significant and positive correlations with ΗΟΜΑ-IR, TyG index, VFT, fasting insulin, and BMI. Children with liver steatosis and elevated liver enzymes are usually asymptomatic, so the screening test for hepatic steatosis among obese children is very important. The metabolic score for insulin resistance could be used by primary care physicians as a screening tool for hepatic steatosis. However, further multicentre prospective cohort studies in children and adolescents with overweight, obesity, as well as normal weight are required to develop cutoff values.

DISCLOSURES

Institutional review board statement: Not applicable.
We would thank the National Research Centre (in-house office for research projects) for the research grants supporting this work. This study was a part of a project No. 10010324. granted by the National Research Centre Egypt.
Financial support and sponsorship: None.
Conflicts of interest: None.

REFERENCES