Reduced Scd36 Following Weight Loss Corresponds to Improved Insulin Sensitivity, Dyslipidemia and Liver Fat in Obese Children

ORIGINAL ARTICLE Reduced sCD36 following weight loss corresponds to improved insulin sensitivity, dyslipidemia and liver fat in obese children

L Knøsgaard1, K Kazankov2, NH Birkebæk3, P Holland-Fischer3, A Lange3, J Solvig4, A Hørlyck4, K Kristensen3, S Rittig3, H Vilstrup2, H Grønbæk2 and A Handberg1,5

BACKGROUND/OBJECTIVES: Childhood obesity is a major health problem with serious long-term metabolic consequences. CD36 is important for the development of obesity-related complications among adults. We aimed to investigate circulating sCD36 during weight loss in childhood obesity and its associations with insulin resistance, dyslipidemia, hepatic fat accumulation and low-grade inﬂammation. SUBJECTS/METHODS: The impact of a 10-week weight loss camp for obese children (N = 113) on plasma sCD36 and further after a 12-month follow-up (N = 68) was investigated. Clinical and biochemical data were collected, and sCD36 was measured by an in- house assay. Liver fat was estimated by ultrasonography and insulin resistance by the homeostasis model assessment (HOMA-IR). RESULTS: Along with marked weight loss, sCD36 was reduced by 21% (P = 0.0013) following lifestyle intervention, and individual sCD36 reductions were signiﬁcantly associated with the corresponding decreases in HOMA-IR, triglycerides and total cholesterol. The largest sCD36 decrease occurred among children who reduced HOMA-IR and liver fat. After 12 months of follow-up, sCD36 was increased (P = 0.014) and the metabolic improvements were largely lost. CONCLUSIONS: Weight-loss-induced sCD36 reduction, coincident with improved insulin resistance, circulating lipids and hepatic fat accumulation, proposes that sCD36 may be an early marker of long-term health risk associated with obesity-related complications. European Journal of Clinical Nutrition (2016) 70, 1073–1077; doi:10.1038/ejcn.2016.88; published online 8 June 2016

INTRODUCTION both adipocytes and macrophages in diet-induced obesity.16 Physical inactivity and unhealthy diet are major causes of Recently, a circulating non-cell-bound form of CD36 (sCD36) was 17,18 obesity.1,2 The threat to public health is particularly related to identified and proposed to reflect tissue CD36 expression. obesity-associated insulin resistance preceding type 2 diabetes Associations of sCD36 with insulin resistance, body mass index mellitus, as well as cardiovascular diseases, hepatic steatosis and (BMI), hepatic fat accumulation, obesity-driven low-grade inflam- certain cancers.2,3 Childhood obesity is a serious public health mation and bariatric surgery-induced weight loss and metabolic – problem, as fat accumulation in children and adolescents carries improvements have previously been reported.17,19 23 Further- an increased risk of premature metabolic complications.2–4 Intra- more, sCD36 is suggested to be a marker of lipid accumulation in abdominal visceral fat is a major contributor to obesity-related arterial walls.24 metabolic complications. High lipolytic activity results in increased Weight reduction, particularly if due to physical activity and portal-free fatty acid flow and possibly contributes to increased caloric restriction, improves insulin sensitivity, and thus reduces 25–27 ectopic hepatic fat accumulation.2,5 A state of low-grade metabolic complications of obesity in adults and children. As inflammation has an important role in obesity-related metabolic childhood obesity is a major health problem, leading to serious complications.6,7 A well-known hypothesis is that the enlargement long-term metabolic consequences, all aspects of these inap- of fat cells causes hypoxia, which leads to a pro-inflammatory propriate metabolic conditions need attention. In the present response, and thus may contribute to the low-grade inflammatory study, we aimed to investigate sCD36 in childhood obesity during condition related to obesity and its complications.8 lifestyle intervention and at follow-up, and in particular, to study The multifunctional membrane receptor CD36 is involved in sCD36 as a marker of obesity-induced insulin resistance, many biological processes.9 Besides scavenging of oxidized low- dyslipidemia, hepatic fat accumulation and low-grade inflamma- density lipoproteins (oxLDLs),9–11 CD36 has an important function tion. We hypothesized that sCD36 in childhood obesity is in the facilitation of fatty acid uptake. In case of fatty acid associated with measures of insulin resistance, dyslipidemia, overload, this uptake contributes to ectopic fat accumulation and hepatic fat accumulation and low-grade inflammation, and that insulin resistance, in muscle, fat and liver cells, respectively.12–15 weight loss along with metabolic improvements causes a In addition, CD36 is involved in the inflammatory responses in reduction in the circulating levels of sCD36.

1Department of Clinical Biochemistry, Aalborg University Hospital, Aalborg, Denmark; 2Department of Hepatology and Gastroenterology, Aarhus University Hospital, Aarhus, Denmark; 3Department of Pediatrics, Aarhus University Hospital, Aarhus, Denmark; 4Department of Radiology, Aarhus University Hospital, Aarhus, Denmark and 5Department of Clinical Medicine, Faculty of Medicine, Aalborg University, Aalborg, Denmark. Correspondence: Professor A Handberg, Department of Clinical Biochemistry, Aalborg University Hospital, Hobrovej 18-22, Aalborg, DK-9000, Denmark. E-mail: [email protected] Received 23 November 2015; revised 24 February 2016; accepted 24 April 2016; published online 8 June 2016 sCD36 is reduced by weight loss in obese children L Knøsgaard et al 1074 MATERIALS AND METHODS nonparametric Spearman's rank correlations test, and described by Study population and clinical examinations Spearman’s rho (ρ) and P-values. The change in sCD36 during intervention was related to corresponding changes in various parameters by repeated One hundred and thirteen children from a weight loss camp at the Danish measurements, determined as differences or ratios depending on visual Christmas Seal Home (Julemærkehjemmet, Hobro, Denmark) were testing with scatter plots and Bland–Altman plots. To assess differences in included in this study. The Danish Christmas Seal Home is an opportunity proportions, the χ2-test or Fisher’s exact test was used. Statistical analyses for overweight school children between 7 and 14 years of age, and the were two-sided and P-values ⩽ 0.05 considered statistically significant. All children are referred by their school physician and offered a 10-week stay analyses were performed using the statistical software package STATA free of charge. The main purpose of the program was weight loss. The version 12.0 (StataCorp LP, College Station, TX, USA). children attended regular school classes, were physically active for at least 1 h every day and followed a fixed diet plan with focus on reduced intake 28–30 of calories. This study population was previously described in details. RESULTS Of the original 117 participants, baseline plasma for sCD36 measurement was available in 113 participants. Anthropometrical and biochemical characteristics of the study Clinical examinations and fasting blood samples were obtained at population baseline (N = 113), after the 10-week stay at the Danish Christmas Seal The anthropometrical and biochemical data of the participating Home (N = 113) and after 12 months of follow-up (N = 68). Body mass index children are presented in Table 1. In brief, after 10 weeks of (BMI) and BMI-standard deviation score (BMI-SDS) were calculated. lifestyle intervention, the children obtained a significant weight Multifrequency electrical bioimpedance analysis (Quadscan 4000, Bodystat Ltd, Isle of Man, UK) was used to estimate body composition, particularly loss. They reduced BMI by 11%, BMI-SDS by 21%, body content of the amount of body fat. fat by 23% and waist circumference by 11%. Furthermore, insulin The study was approved by the Ethics Committee of Region Midtjylland resistance, estimated by HOMA-IR, decreased by 33%, and (1–10–72–114–15). Prior to inclusion, the children and their parents were dyslipidemia improved. At the follow-up visit after 12 months, informed about the purpose, risks and drawbacks of the study and related the children had in average gained 10.8 kg and increased their procedures. Subsequently, they all gave informed written declaration of BMI-SDS by 9%; however, 24% had maintained or even further consent. reduced their BMI-SDS. Mean body content of fat and HOMA-IR increased, and the dyslipidemic state was aggravated. Biochemical analyses Alanine aminotransferase, gamma-glutamyltransferase, triglycerides, total Circulating sCD36 decreased during lifestyle intervention and cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein increased during follow-up cholesterol, glucose, leukocytes and high sensitivity C-reactive protein were measured on automated analysis platforms in the routine laboratory Circulating sCD36 decreased by 21% after 10 weeks of lifestyle at Aarhus University Hospital. intervention (P = 0.0013) and was increased by 32% (P = 0.014) Circulating sCD36 in plasma was measured by an in-house enzyme- during the 12 months of follow-up (Figure 1). No gender linked immunosorbent assay as previously described,17 with an interassay differences were observed (P = 0.69). sCD36 was not associated coefficient of variation of 16.4%. Serum insulin (N = 112) was measured with age at baseline or subsequent visits. by a two-side immunospecific enzyme-linked immunosorbent assay (DakoCytomation, Copenhagen, Denmark); (interassay coefficient of variation of 7.5%), tumor necrosis factor-α and interleukin-6 by an enzyme-linked immunosorbent assay from R&D Systems (Mineapolis, Table 1. Clinical and biochemical characteristics of the children at MN, USA); (interassay coefficient of variation of 10.6 and 9.6%), and oxLDL baseline, after 10 weeks and 12 months of follow-up particles by an enzyme-linked immunosorbent assay from Mercodia (Uppsala, Sweden); (interassay coefficient of variation of 8.5%). All analyses Baseline 10 weeks 12 months were performed on anonymized samples by a lab technician. Age (years) 12.2 ± 1.4 12.4 ± 1.4 13.4 ± 1.3 Insulin resistance Gender (boys:girls) 47:66 47:66 28:40 Weight (kg) 70.3 ± 13.5 63.3 ± 12.0a 74.1 ± 16.1a,b Insulin resistance was estimated by the homeostasis model assessment BMI (kg/m2) 28.0 ± 3.6 24.8 ± 3.2a 27.2 ± 4.4b of insulin resistance (HOMA-IR) (HOMA-IR = (fasting serum insulin BMI-SDS 2.9 ± 0.5 2.3 ± 0.6a 2.5 ± 0.8a,b (μIU/l) × fasting glucose (mmol/l))/22.5). The diagnostic threshold cutoff Fat mass (kg) 23.5 ± 8.3 18.2 ± 6.3a 20.2 ± 8.7a,b 31 point for insulin resistance among adults has been defined as 2.5. No Waist circumference (cm) 94.7 ± 10.1 84.4 ± 9.3a 95.1 ± 16.4b unambiguous similar cutoff point has been defined for children and ALT (IU/l) 30.4 ± 19.8 27.2 ± 18.2a 26.5 ± 15.6 32–35 adolescents, but it tends to exceed 2.5. GGT (IU/l) 19.9 ± 6.2 16.3 ± 5.2a 19.0 ± 9.9a Triglycerides (mmol/l) 0.92 ± 0.48 0.81 ± 0.44a 1.05 ± 0.49b ± ± a ± b Liver fat accumulation Total cholesterol (mmol/l) 3.8 0.8 3.3 0.5 4.0 0.7 HDL (mmol/l) 1.29 ± 0.27 1.24 ± 0.28a 1.29 ± 0.30 To determine the presence of fatty liver, three scoring systems were used: LDL (mmol/l) 2.11 ± 0.73 1.70 ± 0.53a 2.19 ± 0.56b echogenicity of the liver compared with the right kidney, changes in the ± ± ± a,b 29,30 Glucose (mmol/l) 4.8 0.5 4.8 0.4 5.1 0.3 liver texture and plump liver signs, all measured by ultrasonography. Insulin (μIU/l) 8.0 ± 5.4 5.4 ± 2.9a 10.8 ± 5.3a,b Total liver scores were determined by adding up the three scores. Fatty HOMA-IR 1.75 ± 1.23 1.18 ± 0.69a 2.46 ± 1.26a,b fi liver was de ned as an echogenicity score of 1 or above. Leukocytes (x109/l) 7.0 ± 1.8 6.5 ± 1.7a 6.5 ± 1.5a hs-CRP (mg/l) 1.74 ± 2.70 1.63 ± 4.68a 1.40 ± 1.74b Statistics TNFα (ng/l) 1.48 ± 0.68 1.48 ± 0.77 1.43 ± 1.02 IL-6 (ng/l) 1.62 ± 1.07 1.47 ± 0.81 1.23 ± 0.84a,b Continuous variables are presented as mean ± s.d. Linear mixed model for ± ± a ± b the repeated measurements test was used to assess the differences in oxLDL (IU/l) 47.7 14.1 42.6 11.1 47.9 11.3 variables at baseline, after 10 weeks and after 12 months. For post hoc Abbreviations: ALT, alanine aminotransferase; BMI, body mass index; BMI- pairwise comparisons of baseline with 10 weeks, baseline with 12 months, SDS, BMI-standard deviation score; GGT, gamma-glutamyltransferase; HDL, and 10 weeks with 12 months, a paired t-test and the Wilcoxon signed high-density lipoprotein cholesterol; HOMA-IR, homeostasis model assess- rank test were used for normally and non-normally distributed variables, ment of insulin resistance; hs-CRP, high-sensitivity C-reactive protein; IL-6, respectively. To compare normally distributed continuous variables in interleukin-6; LDL, low-density lipoprotein cholesterol; oxLDL, oxidized multiple groups, one-way analysis of variance was used, with Student’s low-density lipoprotein; TNFα, tumor necrosis factor α. aSignificantly t-test for pairwise comparisons. For non-normally distributed variables, we different from baseline values, Po0.05. bSignificantly different from 10- used Kruskal–Wallis and Mann–Whitney tests, respectively. The relation- week values, Po0.05. Values are presented as mean ± s.d. ships between sCD36 and other continuous variables were analyzed by

European Journal of Clinical Nutrition (2016) 1073 – 1077 © 2016 Macmillan Publishers Limited, part of Springer Nature. sCD36 is reduced by weight loss in obese children L Knøsgaard et al 1075 Circulating sCD36 in relation to inflammation and oxidative stress Significant decreases in high-sensitivity C-reactive protein (6%) and leukocyte count (7%) were observed after lifestyle intervention (Table 1), whereas tumor necrosis factor-α and interleukin-6 were unaffected. No correlations were observed between sCD36 and either high-sensitivity C-reactive protein, leukocyte count or pro-inflammatory cytokines. The atherogenic marker of oxidative stress (oxLDL) decreased by 11% during the intervention period and increased by 12% at 12 months of follow-up. oxLDL did not correlate with sCD36 at baseline, and the changes in oxLDL showed no association with the corresponding changes in sCD36.

DISCUSSION In the present study, we aimed to investigate the role of sCD36 in childhood obesity, particularly its relation to insulin resistance, dyslipidemia, hepatic fat accumulation and low-grade inflamma- Figure 1. Circulating levels of sCD36 at baseline, after 10 weeks of tion. Relations between childhood obesity and expression of the lifestyle intervention and 12 months follow-up. Boxes represent medians and the interquartile ranges of sCD36, and whiskers membrane receptor CD36 have been sparsely investigated, and represent the variability outside the upper and lower quartiles. we are not aware of any previous studies that have examined the Outliers are not included in the graphical presentation. One-way effect of lifestyle intervention and weight loss on CD36 expression analysis of variance: Po0.01. Post hoc comparisons: * denotes or circulating sCD36 among obese children. P = 0.0013 compared with baseline and ** denotes P = 0.014 Our main finding was the marked decrease in circulating sCD36 compared with 12 months (unadjusted). after the 10-week lifestyle intervention, which was associated with reduced insulin resistance and decreased levels of triglycerides and cholesterol. Furthermore, it was clear that the largest decrease Circulating sCD36 in relation to weight, BMI, body composition in sCD36 occurred among the children who improved insulin and circulating lipids sensitivity and resolved fatty liver. These findings support the ρ presumed relationship between CD36 and ectopic fat accumula- At baseline, sCD36 correlated with BMI-SDS ( = 0.17, P = 0.008), – BMI (ρ = 0.21, P = 0.025), body weight (ρ = 0.21, P = 0.024) and fat tion, as well as the development of insulin resistance,12 15 and mass (ρ = 0.20, P = 0.03). The changes in sCD36 and fat mass thus the importance of CD36 in obesity-related complications.9,18 during the intervention period tended to correlate (ρ = 0.17, The fact that sCD36 decreased with weight loss and improvement P = 0.08). In children who maintained or further reduced their in metabolic aberrations is in accordance with a study of weight BMI-SDS during the follow-up period after lifestyle intervention, loss induced by bariatric surgery in adults.23 In addition, medical sCD36 at 12 months tended to be lower compared with those interventions were reported to affect sCD36 levels.22,36,37 The who increased their BMI-SDS (0.45 (interquartile range 0.34–0.63) reversion of the children’s metabolic improvements during follow- vs 0.57 (interquartile range 0.43–1.06) arbitrary units, P = 0.08). up, with a concomitant increase in circulating sCD36, for the first At baseline, sCD36 was not associated with triglycerides, total time demonstrate reversibility of circulating sCD36 levels asso- cholesterol, low-density lipoprotein cholesterol or high-density ciated with changes in metabolic improvements and aberrations. lipoprotein cholesterol. However, the decrease in circulating Together, these findings suggest that sCD36 may be considered as sCD36 during the 10 weeks of lifestyle intervention correlated a dynamic and a modifiable biomarker of obesity-related significantly with the corresponding decreases in triglycerides comorbidity. (ρ = 0.30, P = 0.001) and total cholesterol (ρ = 0.23, P = 0.01). In a large European nondiabetic healthy adult population,19 sCD36 was significantly associated with insulin resistance, Circulating sCD36 in relation to insulin resistance measured by the hyperinsulinemic-euglycemic clamp technique. ρ In the present study, mean decreases in sCD36 and HOMA-IR At baseline, sCD36 correlated with fasting glucose ( = 0.19, during the intervention period correlated significantly, and the P = 0.04). Following the lifestyle intervention, the decrease in largest decrease in sCD36 was observed among children with circulating sCD36 correlated with the corresponding decreases in improved insulin sensitivity. Furthermore, the mean values of fasting glucose (ρ = 0.27, P = 0.004), insulin (ρ = 0.27, P = 0.003) and ρ HOMA-IR in our participants did not exceed the diagnostic in HOMA-IR ( = 0.29, P = 0.002). Children showing a reduction in threshold of insulin resistance at any time point, which suggests HOMA-IR at the end of the weight loss camp (n = 75) had a fi that the majority of the children could not be diagnosed as decrease in sCD36, which was signi cantly higher than that in insulin-resistant throughout the study. Together, these findings children with an increase in HOMA-IR during the camp (n = 37) support the idea of sCD36 as a marker of insulin resistance, even (30% (95% confidence interval: 19–40%) vs 2% (95% confidence – at the early stages of impaired insulin sensitivity. interval: decrease of 16 increase of 23%), P = 0.003). Hepatic CD36 mRNA expression levels were higher among patient groups with hepatic fat accumulation compared with Circulating sCD36 in relation to liver fat subjects without liver fat.38 In addition, sCD36 was significantly The children with fatty liver at baseline did not have higher levels associated with steatosis and its severity in patients with chronic of sCD36 compared with those with normal ultrasonography, and viral hepatitis C.20 These studies support the role of CD36 in the changes in sCD36 and ultrasonographic liver scores were not facilitating fatty acid uptake in hepatocytes, and that circulating associated. However, the decrease in sCD36 during the lifestyle levels of sCD36 may reflect these cellular circumstances. In our intervention was significantly larger among children with ultra- present study, no general associations between sCD36 and sonographic resolution of fatty liver (n = 17) compared with those degrees of hepatic fat were observed. However, the fact that who did not (n = 29); (43% (95% confidence interval: 19–59%) vs the children who showed ultrasonographic resolution of fatty liver 12% (95% confidence interval: decrease of 32–increase of 13%), had a significantly larger decrease in circulating sCD36 than those P = 0.04). who did not contribute to the notion of sCD36 as a marker of

© 2016 Macmillan Publishers Limited, part of Springer Nature. European Journal of Clinical Nutrition (2016) 1073 – 1077 sCD36 is reduced by weight loss in obese children L Knøsgaard et al 1076 obesity-related ectopic hepatic fat accumulation, even among University Hospital, Denmark, for skillful laboratory assistance; Morten H Nielsen, children. Department of Clinical Biochemistry, Aalborg University Hospital, for assistance in Because of the apparent pathophysiological importance preparation of figures, as well as the staff of Julemærkehjemmet, Hobro, Denmark, for of low-grade inflammation related to obesity, the children’s their collaboration during the study. The NOVO Nordisk Foundation supported inflammatory state was investigated. In contrast to previous the study. studies in overweight glucose-intolerant men,21 and fructose fed 18 rats, we found no associations between sCD36 and any of the REFERENCES indicators of low-grade inflammation. This inconsistency may fl 1 Berghofer A, Pischon T, Reinhold T, Apovian CM, Sharma AM, Willich SN. Obesity re ect an age-related difference in mechanisms of fat tissue prevalence from a European perspective: a systematic review. BMC Public Health expansion. The numbers of adipocytes are mainly set by a 2008; 8: 200. hyperplastic fat cell expansion during childhood and 2 World Health Organization. Obesity: preventing and managing the global epi- 39 adolescence, followed by a hypertrophic volume expansion of demic. Report of a WHO consultation. World Health Organ Tech Rep Ser 2000; 894: pre-existing adipocytes, which appears to be the predominant 1–253. mechanism among adults. The latter mechanism may cause a 3 Han JC, Lawlor DA, Kimm SY. Childhood obesity. Lancet 2010; 375:1737–1748. state of low-grade inflammation due to local hypoxia proposed to 4 Nader PR, O'Brien M, Houts R, Bradley R, Belsky J, Crosnoe R et al. Identifying risk 118 – induce a pro-inflammatory response.40 for obesity in early childhood. Pediatrics 2006; : e594 e601. Oxidative stress increases the long-term risk of atherosclerosis, a 5 Capurso C, Capurso A. From excess adiposity to insulin resistance: the role of free fatty acids. Vascul Pharmacol 2012; 57:91–97. major complication of obesity. Possible relations between sCD36 6 Hotamisligil GS, Shargill NS, Spiegelman BM. Adipose expression of tumor and oxLDL among children have, however, never been investi- necrosis factor-alpha: direct role in obesity-linked insulin resistance. Science 1993; gated. Although both CD36 and oxLDL may be attributed a role in 259:87–91. the pathogenesis of atherosclerosis, and we observed simulta- 7 Shoelson SE, Lee J, Goldfine AB. Inflammation and insulin resistance. J Clin Invest neous changes in their blood levels, we found no direct 2006; 116: 1793–1801. associations between sCD36 and oxLDL, which is consistent with 8 Wood IS, de Heredia FP, Wang B, Trayhurn P. Cellular hypoxia and adipose tissue previous results in adults.22 The changes in oxLDL coincident with dysfunction in obesity. Proc Nutr Soc 2009; 68:370–377. weight changes indicate that, even in childhood, the adverse 9 Love-Gregory L, Abumrad NA. CD36 genetics and the metabolic complications of obesity. Curr Opin Clin Nutr Metab Care 2011; 14:527–534. effects of obesity may be present, and that it is possible to partly 10 Su X, Abumrad NA. Cellular fatty acid uptake: a pathway under construction. eliminate this risk factor if weight loss and a healthy lifestyle are Trends Endocrinol Metab 2009; 20:72–77. maintained. 11 Febbraio M, Hajjar DP, Silverstein RL. CD36: a class B scavenger receptor involved The strength of our study is the high number of children in angiogenesis, atherosclerosis, inflammation, and lipid metabolism. J Clin Invest attending a standardized lifestyle intervention program with a 2001; 108:785–791. long-term follow-up period, during which the effects of decreased 12 Zhou D, Samovski D, Okunade AL, Stahl PD, Abumrad NA, Su X. CD36 level and body weight and healthy lifestyle on the level of sCD36 were trafficking are determinants of lipolysis in adipocytes. FASEB J 2012; 26: – analyzed. A number of children did not attend the follow-up visit 4733 4742. 13 Bastie CC, Hajri T, Drover VA, Grimaldi PA, Abumrad NA. CD36 in myocytes at 12 months, which could be a confounding factor, but the fi channels fatty acids to a lipase-accessible triglyceride pool that is related to cell children who did and did not attend the nal follow-up visit were lipid and insulin responsiveness. Diabetes 2004; 53:2209–2216. not significantly different (data not presented). In addition, the 14 Koonen DP, Jacobs RL, Febbraio M, Young ME, Soltys CL, Ong H et al. Increased children were studied just before and during puberty. The hepatic CD36 expression contributes to dyslipidemia associated with diet- distribution of sCD36 may reflect this inhomogeneity and be a induced obesity. Diabetes 2007; 56: 2863–2871. complicating factor for future use of sCD36 as a risk biomarker in 15 Chabowski A, Zendzian-Piotrowska M, Konstantynowicz K, Pankiewicz W, Mikłosz A, adolescents. Puberty is characterized by hormonal and bodily Łukaszuk B et al. Fatty acid transporters involved in the palmitate and oleate fl induced insulin resistance in primary rat hepatocytes. Acta Physiol (Oxf) 2013; 207: changes, which may have in uenced our results, and to evade – those circumstances, delta values were statistically analyzed. 346 357. 16 Cai L, Wang Z, Ji A, Meyer JM, van der Westhuyzen DR. Scavenger receptor CD36 Furthermore, the study has a limited external validity restricted expression contributes to adipose tissue inflammation and cell death in diet- to children who meet the criteria to be offered a stay at the induced obesity. PLoS One 2012; 7: e36785. Christmas Seal Homes. 17 Handberg A, Levin K, Hojlund K, Beck-Nielsen H. Identification of the oxidized low- In conclusion, we found clear indications of circulating sCD36 as density lipoprotein scavenger receptor CD36 in plasma: a novel marker of insulin a dynamic biomarker closely associated with obesity-related resistance. Circulation 2006; 114: 1169–1176. dysmetabolic features such as insulin resistance, dyslipidemia 18 Koonen DP, Jensen MK, Handberg A. Soluble CD36- a marker of the (pathophy- and fatty liver in obese children. The reversibility of sCD36 being siological) role of CD36 in the metabolic syndrome? Arch Physiol Biochem 2011; 117 – coincident with the changes in insulin resistance, circulating lipids :57 63. 19 Handberg A, Hojlund K, Gastaldelli A, Flyvbjerg A, Dekker JM, Petrie J et al. and hepatic fat accumulation, proposes that sCD36 may be an Plasma sCD36 is associated with markers of atherosclerosis, insulin resistance and early marker of a long-term health risk associated with obesity- fatty liver in a nondiabetic healthy population. J Intern Med 2012; 271:294–304. related complications. However, our results must be verified in 20 Petta S, Handberg A, Marchesini G, Cammà C, Di Marco V, Cabibi D et al. High larger and more generalizable study populations among children sCD36 plasma level is associated with steatosis and its severity in patients with as well as adults, and further experimental studies are needed to genotype 1 chronic hepatitis C. J Viral Hepat 2013; 20: 174–182. determine the mechanisms and the role of sCD36 in obesity and 21 Handberg A, Lopez-Bermejo A, Bassols J, Vendrell J, Ricart W, Fernandez-Real JM. its metabolic complications. Circulating soluble CD36 is associated with glucose metabolism and interleukin-6 in glucose-intolerant men. Diab Vasc Dis Res 2009; 6:15–20. 22 Glintborg D, Hojlund K, Andersen M, Henriksen JE, Beck-Nielsen H, Handberg A. CONFLICT OF INTEREST Soluble CD36 and risk markers of insulin resistance and atherosclerosis are elevated in polycystic ovary syndrome and significantly reduced during pioglitazone A Handberg is the inventor of two patent applications on sCD36 as a biomarker of treatment. Diabetes Care 2008; 31:328–334. the metabolic syndrome. The Ideas Clinic of Aalborg University Hospital owns the 23 Knosgaard L, Thomsen SB, Stockel M, Vestergaard H, Handberg A. Circulating patent IP rights. sCD36 is associated with unhealthy fat distribution and elevated circulating triglycerides in morbidly obese individuals. Nutr Diabetes 2014; 4:e114. ACKNOWLEDGEMENTS 24 Handberg A, Skjelland M, Michelsen AE, Sagen EL, Krohg-Sørensen K, Russell D et al. Soluble CD36 in plasma is increased in patients with symptomatic athero- We thank Jane Hansen, Aarhus University Hospital, Skejby, Denmark, for the sclerotic carotid plaques and is related to plaque instability. Stroke 2008; 39: collection and processing of blood samples; Lone Larsen, Clinical Institute, Aarhus 3092–3095.

European Journal of Clinical Nutrition (2016) 1073 – 1077 © 2016 Macmillan Publishers Limited, part of Springer Nature. sCD36 is reduced by weight loss in obese children L Knøsgaard et al 1077 25 Coker RH, Williams RH, Yeo SE, Kortebein PM, Bodenner DL, Kern PA et al. 33 Kurtoglu S, Hatipoglu N, Mazicioglu M, Kendirici M, Keskin M, Kondolot M. The impact of exercise training compared to caloric restriction on hepatic and Insulin resistance in obese children and adolescents: HOMA-IR cut-off levels in peripheral insulin resistance in obesity. J Clin Endocrinol Metab 2009; 94: the prepubertal and pubertal periods. J Clin Res Pediatr Endocrinol 2010; 2: 4258–4266. 100–106. 26 O'Gorman DJ, Krook A. Exercise and the treatment of diabetes and obesity. 34 Sahin NM, Kinik ST, Tekindal MA. OGTT results in obese adolescents with normal Endocrinol Metab Clin North Am 2008; 37:887–903. HOMA-IR values. J Pediatr Endocrinol Metab 2013; 26: 285–291. 27 Vitola BE, Deivanayagam S, Stein RI, Mohammed BS, Magkos F, Kirk EP et al. 35 Reinehr T, Andler W. Changes in the atherogenic risk factor profile according to Weight loss reduces liver fat and improves hepatic and skeletal muscle insulin degree of weight loss. Arch Dis Child 2004; 89:419–422. sensitivity in obese adolescents. Obesity (Silver Spring) 2009; 17: 1744–1748. 36 Liani R, Halvorsen B, Sestili S, Handberg A, Santilli F, Vazzana N et al. Plasma 28 Birkebaek NH, Lange A, Holland-Fischer P, Kristensen K, Rittig S, Vilstrup H et al. levels of soluble CD36, platelet activation, inflammation, and oxidative stress Effect of weight reduction on insulin sensitivity, sex hormone-binding globulin, are increased in type 2 diabetic patients. Free Radic Biol Med 2012; 52: sex hormones and gonadotrophins in obese children. Eur J Endocrinol 2010; 163: 1318–1324. 895–900. 37 Chmielewski M, Bragfors-Helin AC, Stenvinkel P, Lindholm B, Anderstam B. Serum 29 Gronbaek H, Lange A, Birkebaek NH, Holland-Fischer P, Solvig J, Hørlyck A et al. soluble CD36, assessed by a novel monoclonal antibody-based sandwich ELISA, Effect of a 10-week weight loss camp on fatty liver disease and insulin sensitivity predicts cardiovascular mortality in dialysis patients. Clin Chim Acta 2010; 411: in obese Danish children. J Pediatr Gastroenterol Nutr 2012; 54: 223–228. 2079–2082. 30 Kazankov K, Moller HJ, Lange A, Birkebaek NH, Holland-Fischer P, Solvig J et al. 38 Miquilena-Colina ME, Lima-Cabello E, Sanchez-Campos S, García-Mediavilla MV, The macrophage activation marker sCD163 is associated with changes in NAFLD Fernández-Bermejo M, Lozano-Rodríguez T et al. Hepatic fatty acid translocase and metabolic profile during lifestyle intervention in obese children. Pediatr Obes CD36 upregulation is associated with insulin resistance, hyperinsulinaemia and 2014; 10: 226–233. increased steatosis in non-alcoholic steatohepatitis and chronic hepatitis C. Gut 31 Kuk JL, Ardern CI. Are metabolically normal but obese individuals at lower risk for 2011; 60:1394–1402. all-cause mortality? Diabetes Care 2009; 32: 2297–2299. 39 Efrat M, Tepper S, Birk RZ. From fat cell biology to public health preventive 32 Keskin M, Kurtoglu S, Kendirci M, Atabek ME, Yazici C. Homeostasis model strategies - pinpointing the critical period for obesity prevention. J Pediatr assessment is more reliable than the fasting glucose/insulin ratio and quantitative Endocrinol Metab 2013; 26:197–209. insulin sensitivity check index for assessing insulin resistance among obese 40 Ye J. Emerging role of adipose tissue hypoxia in obesity and insulin resistance. Int children and adolescents. Pediatrics 2005; 115: e500–e503. J Obes (Lond) 2009; 33:54–66.