Highlights
- •
Increased adiposity has been directly associated with insulin resistance (IR), and cytokines released by adipose tissue seem to link adiposity to IR in youth.
- •
Recent progress in “omics” technologies has provided new opportunities for the development of diagnostic and therapeutic algorithms for patients.
- •
Differential abundance of identified cytokines was validated using individual-level data on clinical and metabolic characteristics by insulin status.
- •
Serum levels of VEGF, FGF-9, IL-15, NT-4, and IP-10 proteins were significantly decreased, while HGF and leptin cytokines were significantly increased, in overweight/obese adolescents according to insulin status.
- •
These data may provide novel insights into the pathogenic mechanisms underlying insulin resistance in youth, offering new targets for prevention.
Abstract
Background and aims
Increased adiposity has been directly associated with insulin resistance (IR), and cytokines released by adipose tissue seem to link adiposity to IR in youth. We used an antibody-based array to investigate the differential levels of serum cytokines according to insulin status in a cohort of overweight/obese and inactive adolescents and evaluated their potential associations with clinical and metabolic characteristics.
Methods and results
We performed a cross-sectional data analysis from 122 adolescents (11–17 years of age). We assessed body composition, cardiometabolic risk factors, biochemical variables, and physical fitness. The concentration of 55 cytokines was quantified in blood samples. The homeostasis model assessment insulin resistance (HOMA‐IR) and AST/ALT and TG/HDL ratios were calculated. IR adolescents as defined as HOMA-IR >2.5. The number of adolescents with IR in the study was 91 (66 % girls). In the IS group, after controlling for confounders, higher IL-15 levels were significantly associated with higher alanine aminotransferase levels and lower AST/ALT ratio, respectively ( Ps <0.05). In the same line, there were significantly higher alanine aminotransferase levels and lower AST/ALT ratio, respectively, with FGF-9 ( Ps <0.05). Likewise, higher alanine aminotransferase levels were significantly associated positively with HGF ( p =0.045). Additionally, leptin levels are associated with six adiposity indexes (i.e., fat mass/height index, body fat, body mass index, android fat mass and gynoid fat mass) in overweight/obese adolescents with IR ( Ps <0.05).
Conclusions
These data may provide novel insights into the pathogenic mechanisms underlying IR in youth, offering new targets for prevention.
Introduction
Insulin resistance (IR), defined as impaired insulin action, shares many pathological features with other disorders of metabolism, including hypertension, dyslipidemia, type 2 diabetes mellitus (T2DM), metabolic syndrome (MetSynd) and non-alcoholic fatty liver disease (NAFLD) . Unsurprisingly, a strong association has been observed between IR and the prevalence of the components of the MetSynd in obese youth, and with a heightened risk of the MetSynd .
IR is characterized by a state of low-grade tissue-specific inflammation induced by pro-inflammatory and/or oxidative stress mediators, notably pro-inflammatory cytokines, has and is a well-recognized sequela of obesity . Indeed, adipose tissue is a major insulin-responsive organ and produces adipose-derived mediators such as free fatty acids that regulate insulin sensitivity and energy metabolism . In this line, subcutaneous and visceral adipose tissue secretes free fatty acids, and their elevated levels in plasma seem to be associated with IR.
The prevalence of impaired glucose tolerance in children and adolescents can reach levels as high , and there seems to be a direct correlation between IR prevalence and the degree of obesity . It is therefore 25 %, even with normal fasting glucose blood levels important to recognize IR in the pre-disease stage when therapeutic intervention is likely to be more successful than in manifest disease, particularly in youth with excess adiposity.
Recent progress in “omics” technologies has provided new opportunities for developing diagnostic and therapeutic algorithms for patients. High-performance platforms can screen large numbers of molecules (including proteins) with the potential to serve as disease biomarkers or as indicators of therapeutic effectiveness and can be targeted directly . Accordingly, screening inflammatory markers in early-age groups at the highest risk of developing MetSynd, T2DM, or NAFLD (i.e., adolescents with excess adiposity) could be substantial in the prediction of disease onset. It may have utility in the future design of multiparametric models of IR risk assessment.
Given that increased adiposity is directly associated with IR, and that adipose-derived mediators link adiposity to IR in youth , we aimed to explore the differential expression of serum cytokines according to insulin status in a cohort of overweight/obese and inactive adolescents, and to examine their potential associations with clinical and metabolic characteristics.
Methods
Study design
A baseline analysis of the clinical trial Exercise Training and Hepatic Metabolism in Overweight/Obese Adolescent (HEPAFIT) – ClinicalTrials.gov Identifier NCT02753231 was carried out between October 2017 and January 2018. We performed a cross-sectional data analysis from a sample of 127 adolescents (70 % girls) between 11 and 17 years of age. For this analysis we used data from 122 participants as a subsample with complete biochemical profile. The study received ethical approval from the Medical Research Ethics Committee of the University of Rosario (N° UR-21042016). All participants were informed of the study’s goals, and written informed consent was obtained from participants and their parents or legal guardians. Details of the background and design methods of the HEPAFIT Study have been previously described .
The following inclusion criteria were applied: primary overweight/obese status, defined according to the International Obesity Task Force , and inactivity (no participation in exercise more than once a week in the previous six months). The exclusion criteria included having a clinical diagnosis of cardiovascular disease, having type 1 or T2DM, pregnancy, using alcohol or drugs, and not having lived in Bogotá (Colombia) for at least one school year. Adolescents with other causes of liver disease presenting elevated liver-enzyme levels were excluded.
Clinical parameters
Variables were collected at the same time in the morning (between 7:00 and 10:00 a.m.) following an overnight fast. Body mass (kg) was measured using an electric scale (Model Tanita® BC-418® Tokyo, Japan) with a range of 0–200 kg and accurate to within 100 g. Height was measured with a portable stadiometer with a precision of 0.1 mm and a range of 0–2.50 m (Seca® 206, Hamburg, Germany). Body mass index (BMI) was calculated as weight (kg)/height (m 2 ), and BMI z-score was calculated using WHO Anthro Plus software v1.0.4. Waist circumference (WC) was measured as the narrowest point between the lower costal border and the iliac crest using a metal tape measure (Lufkin W606PM®, Parsippany, NJ). The waist-to-height ratio was calculated as the ratio of WC to height in accordance with the International Society for the Advancement of Kinanthropometry guidelines . Somatic maturity was estimated by peak height velocity as proposed by Mirwald et al. . The technical error of measurement was less than 2 % for all anthropometric variables.
Body composition parameters (fat mass, body fat, android fat mass, gynoid fat mass, android/gynoid ratio, visceral adipose tissue, fat mass/height and appendicular lean mass), were measured by dual-energy x-ray absorptiometry using the Hologic Horizon DXA System® (Hologic Inc., Bedford, MA). Scans and analyses were completed using a standardized protocol by the same researcher and the equipment was calibrated daily using a known calibration standard.
Blood pressure (BP) and heart rate were measured from the right arm with an Omron® HEM 705 CP (Omron® Healthcare Europe B.V., Hoofddorp, The Netherlands) with an appropriately sized cuff, with subjects sitting still after a 5-min rest. Mean arterial blood pressure (MAP) was calculated as MAP = diastolic BP + (0.333 * [systolic BP × 2 diastolic BP]).
Transient elastography (TE; liver stiffness) and controlled attenuation parameter (CAP) were determined with a FibroScan® 502 Touch device (Echosens, Paris, France). All patients were measured with the 3.5-MHz standard “M” or “XL” probe at a depth of 25 to 65 mm, according to the manufacturer’s specifications. The technical background and reference values in the paediatric population (1 to 18 years) have been recently described . Only liver stiffness measurements with 10 validated measurements were considered reliable .
Metabolic parameters
Blood samples were drawn from the antecubital vein after a fasting period of 10–12 hours Serum lipids, including high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), total cholesterol (TC) and total triglycerides (TG), were measured by using an automatic biochemical analyser (Deyi Biomedical Technology Co., Ltd., Beijing, China). A cardiometabolic z-score MetSynd was created from the sum of systolic blood pressure (SBP), TG, WC, HDL-C *(-1), and fasting glucose z-score . Individuals with a cardiometabolic risk z-score + 1 SD above the mean were identified as having increased cardiometabolic risk.
Serum insulin and glycated hemoglobin (HbA1c) concentrations were analyzed by electrochemiluminescence immunoassay kits (Roche Diagnostics GmbH, Mannheim, Germany). C-peptide, alanine aminotransferase, aspartate aminotransferase, γ-glutamyl transferase, alkaline phosphatase, creatinine, ureic nitrogen, serum iron, iron capacity, iron fixation capacity, and ferritin levels were determined by standard laboratory techniques (Roche Diagnostics GmbH, Mannheim, Germany). A homeostasis model assessment (HOMA)-IR >2.5 was accepted as the presence of IR , and this model has been previously validated against clamp measurements .
Cytokine concentrations
Blood samples were collected into 3.5-mL serum separating tubes and were allowed to clot for 1 h at room temperature. Cytokine/growth factor analysis was performed with the Abcam Human Cytokine Antibody Array Kit (ab133998; Abcam, Cambridge, UK), which allows for the detection of 80 human proteins. Detection was carried out through biotin-conjugated antibodies and horseradish peroxidase-conjugated streptavidin. Array evaluation was performed by densitometry with normalization to positive and negative controls, as well as to overall protein concentration of the sample. The normalization process led to the value of RDM (relative densitometric measurement). The array detects the following cytokines: ENA-78, GCSF, GM-CSF, GRO, GRO-alpha (CXCL2), I-309, IL-1alpha, IL-1beta, IL-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL-8 (CXCL8), IL-10, IL-12 p40/p70, IL-13, IL-15, IFN-gamma, MCP-1 (CCL2), MCP-2 (CCL8), MCP-3 (CC7), MCSF, MDC (CCL22), MIG (CXCL9), MIP-1beta (CCL4), MIP-1delta (CCL15), RANTES (CCL5), SCF, SDF-1, TARC (CCL17), TGF-beta1, TNF-alpha, TNF-beta, EGF, IGF-I, Angiogenin, Oncostatin M, Thrombopoietin, VEGF-A, PDGF-BB, Leptin, BDNF, BLC, Ckß8-1, Eotaxin (CCL11), Eotaxin-2 (CCL24), Eotaxin-3 (CCL26), FGF-4, FGF-6, FGF-7, FGF-9, Flt-3 Ligand, Fractalkine, GCP-2, GDNF, HGF, IGFBP-1, IGFBP-2, IGFBP-3, IGFBP-4, IL-16, IP-10 (CXCL10), LIF, LIGHT, MCP-4, MIF, MIP-3 alpha (CCL20), NAP-2 (CXCL7), NT-3, NT-4, Osteopontin, Osteoprotegerin, PARC (CCL18), PLGF, TGF-beta2, TGF-beta3, TIMP-1, TIMP-2. Values are presented in arbitrary units (a.u).
Statistical analysis
Both statistical (Kolmogorov-Smirnov test) and graphical methods (normal probability plots) were used to examine the fit to a normal distribution for each continuous variable. Descriptive data were assessed by t-test, Mann-Whitney U test or Chi square test by applying the HOMA-IR cut-point in insulin status. Adolescents’ characteristics were described as the mean, standard deviation (SD), or frequencies. Due to their skewed distribution, all dependent variables were log-transformed before inclusion in the models. To aid interpretation, data were back-transformed from the log scale for presentation in the results ( Table 1 ). Interactions by sex or age were explored including interaction terms in the models, as there were no significant interactions (Ps > 0.1). Therefore, all the analyses were performed jointly for the entire sample by insulin status group .
| Characteristics | IS group (n = 31) | IR group (n = 91) | P-value | ||
|---|---|---|---|---|---|
| Chronological age, years | 13.8 | 1.8 | 13.3 | 1.5 | 0.125 |
| Peak height velocity | 0.2 | 1.2 | 0.4 | 1.4 | 0.469 |
| Boys/Girls, % a | 25.8 % | 74.2 % | 34.0 % | 66.0 % | 0.396 |
| Anthropometric parameters | |||||
| Weight, kg | 55.0 | 9.1 | 57.9 | 10.7 | 0.175 |
| Body mass index, kg/m 2 | 23.6 | 4.8 | 23.9 | 3.4 | 0.680 |
| Body mass index, z-score | 1.3 | 0.9 | 1.6 | 0.8 | 0.193 |
| Waist circumference, cm | 72.7 | 5.6 | 76.6 | 8.2 | 0.016 |
| Waist-to-height ratio | 0.47 | 0.03 | 0.49 | 0.05 | 0.020 |
| Body composition parameter | |||||
| Fat mass/height, kg/m 2 | 9.1 | 2.2 | 9.7 | 2.0 | 0.206 |
| Body fat, % | 39.2 | 5.0 | 40.0 | 4.3 | 0.382 |
| Android fat mass, % | 39.4 | 5.7 | 42.1 | 5.1 | 0.016 |
| Gynoid fat mass, % | 44.2 | 4.4 | 44.0 | 4.5 | 0.770 |
| Android/ Gynoid ratio, % | 0.89 | 0.08 | 0.96 | 0.09 | <0.001 |
| Visceral adipose tissue, cm 3 | 310.0 | 83.0 | 351.9 | 108.1 | 0.050 |
| Appen, Lean/height, kg/m 2 | 5.8 | 0.7 | 6.1 | 0.8 | 0.080 |
| Vibration controlled transient elastography | |||||
| Controlled attenuation parameter, dB/m | 197.9 | 29.6 | 234.2 | 41.5 | <0.001 |
| Liver stiffness, kPa | 3.7 | 0.7 | 3.8 | 0.8 | 0.351 |
| High NAFLD risk, % a | 5 | 16.1 % | 52 | 55.9 % | <0.001 |
| Blood pressure | |||||
| Systolic blood pressure, mmHg | 108.7 | 7.0 | 109.1 | 9.3 | 0.837 |
| Diastolic blood pressure, mmHg | 65.8 | 6.1 | 66.9 | 5.7 | 0.389 |
| Mean arterial pressure, mmHg | 79.9 | 5.1 | 80.7 | 5.9 | 0.490 |
| Heart rate, bpm | 77.7 | 10.6 | 79.3 | 13.0 | 0.557 |
| Metabolic parameters | |||||
| Total cholesterol, mg/dl | 154.9 | 30.4 | 160.3 | 22.9 | 0.295 |
| HDL-C, mg/dl | 45.2 | 8.6 | 43.2 | 8.4 | 0.258 |
| LDL-C, mg/dl | 98.6 | 25.0 | 103.2 | 20.7 | 0.306 |
| Triglycerides, mg/dl | 106.4 | 48.9 | 130.7 | 50.8 | 0.021 |
| Glucose, mg/dl | 84.9 | 14.7 | 99.2 | 14.6 | <0.001 |
| Insulin, uU/ml | 9.7 | 2.4 | 20.2 | 11.4 | <0.001 |
| Cardiometabolic z-score, SD | -1.58 | 2.01 | 0.49 | 2.45 | <0.001 |
| HbA1c, % | 5.1 | 0.4 | 5.0 | 0.4 | 0.400 |
| C-peptide, ng/ml | 1.9 | 0.6 | 2.6 | 0.9 | <0.001 |
| Alanine aminotransferase, U/l | 15.5 | 5.8 | 19.2 | 15.6 | 0.200 |
| Aspartate aminotransferase, U/l | 19.9 | 3.8 | 21.8 | 7.3 | 0.184 |
| γ-Glutamyl transferase, U/l | 13.3 | 4.6 | 20.8 | 14.3 | 0.005 |
| Alkaline phosphatase, U/l | 185.8 | 101.0 | 228.1 | 117.8 | 0.076 |
| Creatinine, mg/dl | 0.62 | 0.11 | 0.59 | 0.10 | 0.202 |
| Ureic nitrogen, mg/dl | 10.7 | 2.5 | 10.5 | 2.58 | 0.710 |
| Serum Iron, ug/dl | 98.2 | 32.5 | 89.4 | 29.7 | 0.167 |
| Iron capacity, ug/dl | 243.5 | 58.6 | 267.8 | 46.0 | 0.019 |
| Iron fixation capacity, ug/dl | 341.8 | 45.3 | 357.3 | 39.6 | 0.070 |
| Ferritin, ng/ml | 72.6 | 42.4 | 69.1 | 49.9 | 0.724 |
| TG/HDL ratio | 2.5 | 1.5 | 3.2 | 1.7 | 0.036 |
| AST/ALT ratio | 1.4 | 0.4 | 1.3 | 0.6 | 0.521 |
| HOMA-IR | 2.0 | 0.4 | 4.9 | 3.1 | <0.001 |
| High cardiometabolic z-score MetSynd (>1 SD), % a | 3 | 9.6 % | 35 | 38.4 % | 0.003 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
