Oxidative Stress and Adverse Adipokine Profile Characterize the Metabolic Syndrome in Children

Authors

  • Aaron S. Kelly PhD,

    1. From the Departments of Pediatrics,1Medicine,2 and Kinesiology,3University of Minnesota, Minneapolis, MN; and the Department of Research, St Paul Heart Clinic, St Paul, MN4
    Search for more papers by this author
  • 1,4 Julia Steinberger MD, MS,

    1. From the Departments of Pediatrics,1Medicine,2 and Kinesiology,3University of Minnesota, Minneapolis, MN; and the Department of Research, St Paul Heart Clinic, St Paul, MN4
    Search for more papers by this author
  • 1 Daniel R. Kaiser PhD,

    1. From the Departments of Pediatrics,1Medicine,2 and Kinesiology,3University of Minnesota, Minneapolis, MN; and the Department of Research, St Paul Heart Clinic, St Paul, MN4
    Search for more papers by this author
  • 2 Thomas P. Olson PhD,

    1. From the Departments of Pediatrics,1Medicine,2 and Kinesiology,3University of Minnesota, Minneapolis, MN; and the Department of Research, St Paul Heart Clinic, St Paul, MN4
    Search for more papers by this author
  • 3 Alan J. Bank MD,

    1. From the Departments of Pediatrics,1Medicine,2 and Kinesiology,3University of Minnesota, Minneapolis, MN; and the Department of Research, St Paul Heart Clinic, St Paul, MN4
    Search for more papers by this author
  • and 2,4 Donald R. Dengel PhD 3

    1. From the Departments of Pediatrics,1Medicine,2 and Kinesiology,3University of Minnesota, Minneapolis, MN; and the Department of Research, St Paul Heart Clinic, St Paul, MN4
    Search for more papers by this author

Aaron S. Kelly, PhD, St Paul Heart Clinic, 225 Smith Avenue North, Suite 400, St Paul, MN 55102
E-mail: kelly105@umn.edu

Abstract

Thirty-four children were assessed for body composition, blood pressure, lipids, glucose tolerance, markers of insulin resistance, oxidative stress, and adipokines. Children were divided into 3 groups: (1) normal weight, (2) overweight but otherwise healthy, and (3) overweight with the metabolic syndrome. There were no differences among any of the groups for age or Tanner stage, and anthropometric variables were similar between the overweight and the overweight with the metabolic syndrome groups. Differences across groups were found for high-density lipoprotein cholesterol (P<.001), triglycerides (P<.01), fasting insulin (P<.001), homeostasis model assessment (P<.01), adiponectin (P<.05), leptin (P<.0001), C-reactive protein (P<.0001), interleukin 6 (P<.0001), and 8-isoprostane (P<.001). In children, oxidative stress and adipokine levels worsen throughout the continuum of obesity and especially in the presence of components of the metabolic syndrome. Overweight children with components of the metabolic syndrome may be at elevated risk for future cardiovascular disease.

In adults, the metabolic syndrome (MetS) is associated with myocardial infarction and stroke1 and is an independent predictor of cardiovascular events.2 The MetS is prevalent in adults and, although no formal definition for the syndrome currently exists for children, recent evidence suggests that many components of the MetS are common in youngsters.3–5 In adults, the MetS is associated with a chronic inflammatory state6–13 and may be related to increased systemic oxidative stress.14,15 It is unclear, however, how early these changes in inflammation and oxidative stress occur in overweight children with and without components of the MetS.

The atherosclerotic process begins in childhood,16 and oxidative stress may be an initial trigger, since it can decrease the production and bioavailability of nitric oxide (NO), a key vascular protective molecule produced by endothelial cells.17 Similarly, adipokines, which are secreted from and regulated by fat cells, can act directly on the vascular wall to promote atherosclerosis.18–21 These hormones often work in concert with oxidative stress to induce endothelial dysfunction. Previous studies have shown that overweight and obese children have elevated levels of subclinical inflammation and abnormal adipokine levels compared with normal-weight controls.22 Less is known, however, about the additive cardiovascular risk associated with the presence of components of the MetS in overweight children, particularly in relation to levels of oxidative stress and adipokines. Therefore, we sought to compare the levels of oxidative stress and adipokines in healthy normal-weight children and in overweight children with and without components of the MetS.

Methods

Subjects. Thirty-four healthy normal-weight and overweight children between the ages of 8 and 14 years were recruited from the Minneapolis and St Paul, MN, metropolitan area to participate in the study. Testing occurred in the University of Minnesota General Clinical Research Center. All children and parents/guardians provided verbal and written informed consent. The study protocol was approved by the University of Minnesota Institutional Review Board. The study was conducted according to institutional and Health Insurance Portability and Accountability Act of 1996 (HIPAA) guidelines.

Classification and Definition of Overweight and the Metabolic Syndrome. Children were divided into 3 groups: (1) normal weight and healthy (n=11; 7 girls), (2) overweight, but otherwise healthy (n=13; 6 girls), and (3) overweight with the MetS (overweight + MetS) (n=10; 5 girls). Children were classified as overweight if body mass index exceeded the 85th percentile for age and sex. The MetS was defined using previously published criteria modified for adolescents (triglycerides ≥110 mg/dL, high-density lipoprotein [HDL] cholesterol ≤40 mg/dL, waist circumference ≥90th percentile, fasting glucose ≥110 mg/dL, blood pressure ≥90th percentile).25 Any combination of 3 or more components constituted the MetS.

Laboratory Measurements. Fasting blood was collected for the measurement of lipids, glucose, insulin, C-reactive protein (CRP), interleukin 6 (IL-6), tumor necrosis factor α (TNF-α), adiponectin, leptin, resistin, and 8-isoprostane. Homeostasis model assessment for insulin resistance was calculated as fasting glucose × fasting insulin/22.5, as described by Matthews et al.26 All subjects had been free from illness and injury in the previous 2 weeks. Assays for lipids, glucose, insulin, and CRP were conducted at Fairview Diagnostic Laboratories, Fairview University Medical Center, Minneapolis, MN. Cholesterol, triglycerides, and glucose were analyzed by colorimetric reflectance spectrophotometry. Ultrasensitive CRP was analyzed via rate nephelometry. Insulin was measured by chemiluminescent immunoassay. IL-6, TNF-α, adiponectin, leptin, resistin, and 8-isoprostane were analyzed in the University of Minnesota Cytokine Reference Laboratory by standard techniques with enzyme-linked immunosorbent assay (ELISA). Inter- and intra-assay coefficients of variation for the adipokines and 8-isoprostane were as follows (presented as inter- and intra-assay coefficients of variation, respectively): IL-6 (3.3–6.4; 1.6–4.2), TNF-α (10.8–16.7; 5.3–8.8), adiponectin (5.8–6.9; 2.5–4.7), leptin (3.5–5.4; 3.0–3.3), resistin (7.8–9.2; 3.8–5.3), 8-isoprostane (6–22; 4–35).

Blood pressure was measured in the supine position with a calibrated radial artery tonometer (Vasotrac APM205A, Medwave, St Paul, MN) after at least 10 minutes of quiet rest. Standard 2-hour oral glucose tolerance tests were performed to assess glucose and insulin response. Body composition was determined by dual-energy x-ray absorptiometry. Tanner stage for classifying pubertal development was determined by a trained pediatrician.

Statistical Analyses. Data are expressed as mean ± SEM. Triglycerides, CRP, and IL-6 were log-transformed for analysis. Comparisons of variables among the different groups were conducted by analysis of variance (ANOVA) with Bonferroni posttests. ANOVA with posttests for linear trend were used to address differences across the groups for variables of interest. An α of .05 was preidentified to indicate statistical significance. All analyses were conducted with GraphPad Prism version 4.0 (GraphPad Software, Inc, San Diego, CA).

Results

There were no significant differences among any of the groups for age or Tanner stage (Table I). By design, compared with the normal-weight group, both the overweight and the overweight + MetS groups had significantly higher body mass index (P<.001), waist circumference (P<.001), total percentage body fat (P<.001), and percentage trunk fat values (P<.001) (Table I); however, there were no significant differences in any of these anthropometric variables between the overweight and overweight + MetS groups. There was a trend (P=.06 for ANOVA main effect) toward higher systolic blood pressure values across groups, but no statistically significant difference for diastolic blood pressure values (Table I).

Table I. Descriptive Variables
Variable, mean± SEMNormal WeightOverweightOverweight+ MetS
Age, y11.1±0.611.2±0.611.1±0.6
Tanner stage1.7±0.52.4±0.42.0±0.4
Body mass index, kg/m218.1±0.7*29.1 ± 1.431.9±2.2
Waist circumference, cm66.5±1.8*92.8±2.598.8±5.5
Total body fat, %26.1±2.2*43.1±1.545.8±1.3
Trunk fat, %24.3±2.3*43.9±1.747.6±1.6
Systolic blood pressure, mm Hg113±3119±2127±6
Diastolic blood pressure, mm Hg60±264±268±3
*P<.001 for normal weight vs overweight and overweight with the metabolic syndrome (MetS) groups. P values are from analysis of variance with Bonferroni posttests.

In the entire sample, all children had normal fasting glucose levels. Of the remaining components of the MetS, systolic blood pressure, HDL cholesterol, and triglycerides were the only distinguishing features separating the 2 overweight groups. In the overweight + MetS group, the 2 most common components of the MetS were low HDL cholesterol and increased waist circumference. These 2 components of the MetS were present in all 10 of the children classified in this category.

There were no significant differences among the groups for total cholesterol or low-density lipoprotein cholesterol; however, HDL cholesterol was significantly lower (P<.001) across the groups and triglycerides were significantly higher (P<.01) across the groups (Table II). Fasting insulin (P<.001) and homeostasis model assessment of insulin resistance (P<.01) were significantly higher across the groups, and there was a trend toward statistical significance (P=.06) for elevated 2-hour oral glucose tolerance test insulin levels across the groups. No differences were observed for fasting glucose or 2-hour oral glucose tolerance test glucose (Table II).

Table II. Lipids and Glucose Tolerance
Variable, mean± SEMNormal WeightOverweightOverweight± MetS
Total cholesterol, mmol/L3.75±0.263.98±0.214.14±0.28
mg/dL145±10154±8160±11
LDL cholesterol, mmol/L2.17±0.182.52±0.182.60±0.25
mg/dL83.8±6.897.5±7.0100.5±9.7
HDL cholesterol, mmol/L*1.25±0.091.11±0.040.94±0.02
mg/dL48.2±3.643.0±1.436.2±0.8
Triglycerides, mmol/L†0.71±0.120.79±0.091.33±0.21
mg/dL63.2±10.470.0±8.3118.2±18.6
Fasting glucose, mmol/L4.74±0.074.79±0.094.60±0.07
mg/dL85.4±1.286.3±1.782.9±1.2
Fasting insulin, μU/mL*6.0±0.78.2±1.013.8±2.4
Homeostasis model assessment†1.26±0.151.76±0.232.83±0.49
2-h OGTT glucose, mmol/L6.00±0.336.38±0.446.22±0.33
mg/dL108±6115±8112±6
2-h OGTT insulin, μU/mL38.2±8.450.0±11.595.6±26.5
*P<.001. †P<.01. P values are from analysis of variance for linear trend across groups. MetS indicates metabolic syndrome; LDL, low-density lipoprotein; HDL, high-density lipoprotein; and OGTT, oral glucose tolerance test.

There were no statistically significant differences across groups for resistin (normal weight, 14.6±1.4 ng/mL; overweight, 15.9+1.7 ng/mL; overweight + MetS, 19.8±2.6 ng/mL; P=.17). There was a nonsignificant trend toward elevated TNF-α levels across the groups (normal weight, 0.82±0.08 pg/mL; overweight, 0.96±0.13 pg/mL; overweight + MetS, 1.36±0.25 pg/mL; P=.08). Significant differences across groups were observed for CRP (P<.0001) (Figure 1A), IL-6 (P<.0001) (Figure 1B), adiponectin (P<.05) (Figure 1C), leptin (P<.0001) (Figure 1D), and 8-isoprostane (P<.001) (Figure 2).

Figure 1.

Figure 1.

Differences in variables across the groups for C-reactive protein (A), interleukin 6 (B), adiponectin (C), and leptin (D). ANOVA indicates analysis of variance; NW, normal weight; OW, overweight; and w/ MetS, with the metabolic syndrome.

Figure 2.

Figure 2.

Significant elevation across the groups for 8-isoprostane, a marker of oxidative stress. ANOVA indicates analysis of variance; NW, normal weight; OW, overweight; and w/ MetS, with the metabolic syndrome.

Discussion

The novel finding of the current study is that compared with normal-weight children and with overweight children without the MetS, overweight children with the MetS have the highest levels of systemic oxidative stress and most abnormal adipokine profiles. These abnormalities occur at an early age, well before the onset of clinical cardiovascular disease. The presence of 3 or more components of the MetS, in addition to obesity, may further increase the level of cardiovascular risk in children. Since not all obese children are at heightened cardiovascular risk, the presence of the MetS may serve as a valuable means to stratify children based on risk and identify those individuals most in need of early and aggressive intervention to decrease future cardiovascular morbidity and mortality.

Little is known about oxidative stress in overweight children. The present study suggests that oxidative stress is increased at an early age in overweight children and especially in overweight children who have components of the MetS. Molnar et al27 reported that children and adolescents with the MetS have decreased antioxidant levels in the blood compared with obese children and adolescents without the MetS and healthy controls. Sinaiko et al28 showed that 8-isoprostaglandin F levels were elevated with increasing insulin resistance and body mass index in adolescents (mean age, 15 years). Our data extend these findings in the context of obesity and the MetS in a younger group of children. Increased levels of oxidative stress may have deleterious effects on the vascular wall as it interferes with the production and bioavailability of NO, which has many antiatherogenic properties.17

Previous work has demonstrated that increased cardiovascular risk is associated with obesity and the MetS in childhood.29–32 Recently, Iannuzzi et al32 showed that obese children with the MetS had stiffer carotid arteries than obese children without the MetS. Similarly, Whincup et al33 reported a significant inverse relationship between the number of components of the MetS and arterial distensibility in a large cohort of children and adolescents. Our data are in agreement with these studies and suggest that early vascular derangements are occurring in children with the MetS. In fact, increased oxidative stress and adverse adipokine profiles may be potential mechanisms explaining the increased arterial stiffness. Evidence exists that CRP directly quenches NO by interfering with endothelial NO synthase messenger RNA stability.21 Leptin has been shown to promote vascular injury and induce atherosclerosis and thrombosis in mice.18 Furthermore, adiponectin19 and resistin20 are associated with coronary artery calcification in humans. In our study, the worsening of adipokines and oxidative stress with the MetS does not seem to be mediated by increased total body fat or trunk fat since the overweight and overweight + MetS groups did not differ in these categories. This finding suggests that the MetS confers an independent and additive cardiovascular risk in overweight children.

A limitation of the current study is the small sample size, especially as it relates to measures of body fat. Although there were no statistically significant differences in any of the anthropometric measurements between the overweight and the overweight + MetS groups, a larger sample may have detected significant differences in these variables. This is important because body fat drives the production of many of the adipokines. Another limitation is the lack of a standardized definition of the MetS in children. Using different cutoffs for the components of the MetS may have altered the current results. Finally, we did not have a direct measure of insulin resistance in these children; we used fasting insulin and homeostasis model assessment as surrogate markers. Insulin resistance likely plays a prominent role in the complex relationships among obesity, the MetS, and the expression of adipokines. Despite these limitations, the current data support a likely biologic basis for the increased vascular risk present in overweight children, which is further worsened with the MetS.

Conclusions

The findings of the current study suggest that compared with normal-weight children and overweight children without the MetS, overweight children with the MetS have the highest levels of systemic oxidative stress and the most abnormal adipokine profiles. The presence of components of the MetS may serve as a valuable means to stratify overweight children based on risk and identify those individuals most in need of early and aggressive intervention to decrease future cardiovascular morbidity and mortality. Studies are needed to further examine the early vascular consequences of the metabolic syndrome in children and to assess the effectiveness and long-term benefits of addressing the accompanying risk factors at a young age.

Disclosure: This work was supported by Minnesota Obesity Center Grant No. 1P30DK50456-08 (D. R.K.), American Heart Association Pre-Doctoral Grant Nos. 0315213Z (A.S.K.) and 0410034Z (T.P.O), and General Clinical Research Center, No. M01-RR00400, General Clinical Research Center Program, National Center for Research Resources/National Institutes of Health.

Ancillary