 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
PEDIATRICS Vol. 107 No. 1 January 2001, p. e13
,
, and
From the * Institute for Research in Extramural Medicine,
Faculty of Medicine, Vrije Universiteit, Amsterdam, The Netherlands;
Epidemiology, Demography, and Biometry Program, National Institute
on Aging, National Institutes of Health, Bethesda, Maryland; § National
Center for Health Statistics, Centers for Disease Control and
Prevention, Hyattsville, Maryland; and the Departments of Laboratory
Medicine and Medicine, University of Washington, Seattle, Washington.
 |
ABSTRACT |
|---|
Top
Abstract
Methods
Results
Discussion
Conclusion
References
|
|---|
Objective. Human adipose tissue expresses and releases the proinflammatory cytokine interleukin-6, potentially inducing low-grade systemic inflammation in persons with excess body fat. To limit potential confounding by inflammation-related diseases and subclinical cardiovascular disease, we tested the hypothesis that overweight is associated with low-grade systemic inflammation in children.
Design and Setting. The third National Health and Nutrition Examination Survey, 1988-1994, a representative sample of the US population.
Participants. A total of 3512 children 8 to 16 years of age.
Outcome Measures. Elevated serum C-reactive protein
concentration (CRP; 
.22 mg/dL) and white blood cell count
(109 cells/L).
Results. Elevated CRP was present in 7.1% of the boys and 6.1% of the girls. Overweight children (defined as having a body mass index or a sum of 3 skinfolds (triceps, subscapula, and supra-iliac) above the gender-specific 85th percentile) were more likely to have elevated CRP than were their normal-weight counterparts. After adjustment for potential confounders, including smoking and health status, the odds ratio (OR) was 3.74 (95% confidence interval [CI]: 1.66-8.43) for overweight boys and the OR was 3.17 (95% CI: 1.60-6.28) for overweight girls, based on the body mass index. Based on the sum of 3 skinfolds, these ORs were 5.11 (95% CI: 2.36-11.06) and 2.89 (95% CI: 1.49-5.59) for boys and girls, respectively. Overweight was also associated with statistically significant higher white blood cell counts. The results were similar when restricted to healthy, nonsmoking, nonestrogen-using children.
Conclusions. In children 8 to 16 years of age, overweight is associated with higher CRP concentrations and higher white blood cell counts. These findings suggest a state of low-grade systemic inflammation in overweight children. inflammation, obesity, children.
C-reactive protein (CRP) is an acute-phase protein and a
sensitive marker for systemic inflammation. In a recent meta-analysis of 7 prospective studies, elevated serum CRP concentration has been
shown to predict future risk of coronary heart disease.1
CRP concentrations well below the conventional clinical upper limit of
normal of 1 mg/dL have been associated with a twofold to threefold
increase in risk of myocardial infarction, ischemic stroke, peripheral
arterial disease, and coronary heart disease mortality in healthy men
and women.2-6 These findings demonstrate the potential
detrimental consequences of elevated CRP concentrations on health.
Several factors are known to increase CRP concentrations.
Smoking7-9 and hormone replacement
therapy10,11 have been associated with elevated CRP
concentration in middle-aged and elderly persons. In addition, several
inflammation-related diseases, such as respiratory
disease,8 rheumatoid arthritis,12 diabetes
mellitus,9,13,14 and (subclinical) cardiovascular
disease,9 have been associated with elevated CRP
concentrations. Moreover, recent studies have reported a positive
relationship between body mass index (BMI) and CRP
concentrations.5,8,9,15-18
The elevated CRP concentrations in overweight persons might be
explained by the expression of the cytokine interleukin-6 in adipose
tissue19-21 and its release into the
circulation.21,22 Interleukin-6 is a proinflammatory
cytokine that stimulates the production of acute-phase proteins,
including CRP, in the liver.23,24 Higher adipose tissue
content of interleukin-6 has been associated with higher serum CRP
concentrations in obese persons.25 The release of
interleukin-6 from adipose tissue may induce elevated CRP
concentrations in persons with excess body fat.
Previous studies investigating the association between body fatness and
CRP were primarily conducted in middle-aged and elderly adults in whom
the observed association may have been confounded by disease.
Rheumatoid arthritis, diabetes mellitus, and cardiovascular disease are
prevalent diseases in older adults and are clearly associated with both
obesity26-28 and increased CRP
concentrations.1,8,12,13,16 To limit the potential
confounding by inflammation-related diseases and subclinical
cardiovascular disease, we investigated the association between
overweight and systemic inflammation in children.
This study tested the hypothesis that overweight is associated with
low-grade systemic inflammation as measured by serum CRP concentration
and white blood cell count. The study population included 3512 children
8 to 16 years of age who were participants of the third National Health
and Nutrition Examination Survey (NHANES III), 1988-1994, a
representative sample of the US population.
Survey Design
NHANES III was conducted by the National Center for Health
Statistics of the Centers for Disease Control and
Prevention.29 The survey had a complex, stratified,
multistage, probability-cluster design for selecting a sample of
~40 000 persons representative of the noninstitutionalized civilian
population of the United States. Children <5 years of age, persons 60 years of age and older, Mexican Americans, and non-Hispanic blacks were
sampled at higher rates than were other persons. Eighty-seven percent of all eligible children 8 to 16 years of age were interviewed in their
household, of whom 5065 (81%) were subsequently examined in a mobile
examination center (n = 5052) or in their homes
(n = 13). Of the 4018 children who had complete data on
anthropometry, 506 children were excluded from the statistical analyses
because of missing data on serum CRP concentration or white blood cell count. A total of 3512 children (1725 boys and 1787 girls) were available for the statistical analyses.
Anthropometry
Body weight and height were measured using standardized
procedures previously described.30 BMI was calculated as
weight in kilograms divided by height in meters squared and used as an
indicator of total body fat.31-33 Skinfolds were measured
on the right side of the body using a Holtain T/W skinfold
caliper (Holtain Ltd., Crymych, UK) and recorded to the nearest
.1 mm. The triceps skinfold, subscapular skinfold, and the supra-iliac
skinfold were measured using standardized procedures and
locations.30 The sum of 3 skinfolds was calculated and
used as an indicator of subcutaneous body fat.
Children were considered overweight when they had a BMI or a sum of 3 skinfolds above the gender-specific 85th percentile, as proposed by a
consensus conference.34 The cutpoints for overweight were
created based on the 85th percentile of the total population of boys
and girls 8 to 16 years of age included in the NHANES III study
(n = 4220 for BMI and n = 4042 for the
sum of 3 skinfolds). The cutpoints for the BMI were >23.66 kg/m2 for boys and >24.52
kg/m2 for girls. The cutpoints for the sum of 3 skinfolds were >56.90 mm for boys and >68.27 mm for girls.
Inflammation Markers
Serum CRP
Serum specimens for the measurement of CRP were shipped on dry
ice to the laboratory, stored at -70°C, and analyzed within 2 months
after phlebotomy. CRP was analyzed using a modification of the Behring
Latex-Enhanced CRP assay on the Behring Nephelometer Analyzer System
(Behring Diagnostics, Westwood MA).35 Both within-assay
and between-assay quality control procedures were used and the
coefficient of variation of the method was 3.2% to 16.1% through the
period of data collection. The assay was designed primarily to detect
inflammation in patients, and it was included as part of the NHANES III
cohort originally to help detect inflammation as a confounding variable
for interpretation of nutrition markers. The assay could detect a
minimal concentration of .22 mg/dL, and values below this level were
classified as undetectable. Because a majority of individuals had
values less than the minimal detectable concentration, in this analysis
CRP is treated as a categorical rather than as a continuous variable.
The population was divided into 2 categories based on CRP
concentration: undetectable (<.22 mg/dL) and elevated ( White Blood Cell Count
White blood cell count was assessed using a quantitative,
automated hematology analyzer (Coulter Counter Model S-PLUS JR
[Beckman Coulter Inc., Fullerton, CA]).36 Lower
detection limit was .4 (109 cells/L). Both
within-assay and between-assay quality control procedures were used,
and the coefficient of variation of the method was <3.0% through the
period of data collection. White blood cell count was used as a
continuous variable in the analyses.
Potential Confounders and Effect Modifiers
Race and disease prevalence were based on proxy report, usually
by the mother or father of the child (95.1%). Race was defined as
non-Hispanic white, non-Hispanic black, Mexican American, or other.
Respiratory disease prevalence was determined through report of
physician-diagnosed chronic bronchitis or asthma or report of having a
cold in the past few days. Other diseases included physician-diagnosed
cardiovascular disease including hypertension, high cholesterol or
rheumatic heart disease, or diabetes mellitus defined as current use of
blood glucose regulators. Smoking status was based on self-report and
categorized as never and former/current smoking. In children 12 years
of age and older, serum cotinine concentration was measured by
high-performance liquid chromatography and atmospheric pressure
chemical ionization tandem mass spectroscopy.37 Children
with a serum cotinine concentration >10 ng/mL38 were
categorized as former/current smokers, regardless of self-report. Estrogen use was based on self-report and included oral contraceptive medications and implants. The stage of sexual maturation was assessed during the physical examination using the criteria of
Tanner.39 Children were categorized into prepubertal
(Tanner stage < 5) and postpubertal (Tanner stage = 5).
Statistical Analyses
Two outcome variables were defined: elevated CRP concentration
( Elevated CRP ( TABLE 1
METHODS
Top
Abstract
Methods
Results
Discussion
Conclusion
References
.22 mg/dL).
.22 mg/dL), which was contrasted with undetectable CRP, and white
blood cell count, which was used as a continuous variable. Within each
gender, the relationship between overweight and CRP concentration
category was examined by means of multiple logistic regression
analysis. We calculated odds ratios (ORs) and 95% confidence intervals
(CIs) for the BMI as a categorical variable with normal weight (BMI
equal or below gender-specific 85th percentile) as the reference
category and for BMI as a continuous variable, expressed per 4 kg/m2 (~1 standard deviation [SD]) increment.
Similar analyses were performed for the sum of 3 skinfolds using an
increment of 23.0 mm, corresponding to the SD. Within each gender the
relationship between overweight and white blood cell count was examined
using linear regression analyses, with BMI or the sum of 3 skinfolds as
a categorical variable (1 = above gender-specific 85th percentile, 2 = equal or below gender-specific 85th percentile) and as a
continuous variable, expressed per SD increment. Adjustments were made
for potential confounders, including age, race, smoking status, sexual maturation stage, estrogen use (girls only), respiratory disease, and
other diseases shown to be associated with low-grade inflammation in
adults. Racial differences in the association between overweight and
inflammation status were assessed in analyses stratified by gender and
race and were tested by using product terms. To assess potential effect
modification by smoking status, disease status, or estrogen use, the
analyses were repeated restricted to healthy, never smokers among boys
and girls, with an additional exclusion of estrogen users among girls.
Analyses were performed using SAS (SAS Institute, Inc, Cary, NC) and
SUDAAN (Research Triangle Institute, Research Triangle Park, NC) and
incorporated sampling weights to account for oversampling and
nonresponse to the household interview and examination.36
Variance estimates were calculated with SUDAAN, incorporating the
complex sampling design of NHANES III.36
RESULTS
Top
Abstract
Methods
Results
Discussion
Conclusion
References
.22 mg/dL) was present in 7.6% of the boys and
6.1% of the girls. Mean white blood cell count was 7.1 × 109/L (standard error [SE]: .1) for boys
and 7.3 × 109/L (SE: .1) for girls. Other
characteristics of the study population are shown in Table
1.
Characteristics of the Study Population: NHANES III,
1988-1994
The relationship of BMI category or skinfolds category with the prevalence of elevated CRP concentration is shown in Fig 1. A higher prevalence of elevated CRP concentration was observed in boys and girls with a BMI or sum of 3 skinfolds above the gender-specific 85th percentile, the proposed cutpoint for overweight in children. Based on BMI, an elevated CRP concentration was observed among 20.6% and 18.7% of the overweight boys and girls, respectively. Based on the sum of 3 skinfolds, these percentages were 19.8% and 16.1%, respectively. The prevalence of elevated serum CRP concentration among overweight children (category 5, Fig 1) was higher compared with the prevalence at all other categories, the only exception being the difference between category 5 and category 4 of the sum of 3 skinfolds in girls (P = .2). In boys and girls, the prevalence of elevated serum CRP concentration did not differ among categories 1 to 4. These results could be interpreted as a threshold effect in the association between overweight and elevated serum CRP concentration.
|
Based on BMI
after adjustment for potential confounders including age,
race, smoking, respiratory and cardiovascular disease, diabetes
mellitus, sexual maturation stage, and estrogen use (girls only)overweight boys were 3.74 and overweight girls were 3.17 times
more likely to have elevated CRP, compared with their normal weight
counterparts (Table 2). Based on the sum
of 3 skinfolds these numbers were 5.11 and 2.89 for boys and girls, respectively. Per 1 SD increase in BMI, boys were 1.65 and girls were
1.60 times more likely to have elevated CRP. Per 1 SD in the sum of 3 skinfolds, these numbers were 1.68 and 1.61. In addition, overweight
boys and girls had higher white blood cell counts than did normal
weight children (Table 2). No effect modification by race was observed
(P > .12).
|
To avoid any potential effect modification by disease, smoking, or estrogen use, the analyses were repeated restricted to 2419 healthy, never-smoking, nonestrogen-using children. The positive association between overweight and elevated CRP remained statistically significant after adjustment for age, race, and sexual maturation stage (Table 3). Similar results were observed for white blood cell count.
|
| |
DISCUSSION |
|---|
|
Top
Abstract
Methods
Results
Discussion
Conclusion
References
|
|---|
In this study we observed a higher prevalence of elevated CRP concentration in overweight children compared with normal weight children, even after carefully controlling for disease and other factors known to influence CRP concentrations. Being overweight was also associated with a higher white blood cell count, confirming the presence of low-grade systemic inflammation. A positive association between BMI and CRP concentration has been repeatedly observed in adults.5,8,9,15-18 Our study extends these important findings to children in whom the prevalence of any confounding subclinical disease is very low.
Overweight at young age is associated with dyslipidemia42,43 and insulin resistance.44,45 Prospective studies have shown that overweight in childhood is an important determinant of overweight in adulthood.46-48 Moreover, childhood overweight is associated with the metabolic syndrome in adulthood, independent of adult weight,49 and is a more powerful predictor of cardiovascular morbidity and mortality than is overweight in adulthood.50 The prevention and management of childhood overweight is important to reduce these potential health risks.
To our knowledge, this is the first study reporting an association between childhood overweight and inflammation. Although the health effects of low-grade systemic inflammation in children are unknown, in healthy adults it has been shown to increase the risk for cardiovascular disease and diabetes mellitus.2-6,51 Moreover, CRP induces the production of tissue factor, a potent procoagulant, in monocytes.52 Because of the reported adverse health effects of systemic inflammation in adults, the inflammation observed in overweight children may be an additional risk factor for future disease. Whether the low-grade systemic inflammation in overweight children might partly explain their increased risk for cardiovascular disease and diabetes mellitus in adulthood is unknown. More information is needed about the long-term health impact of inflammation and other adipose tissue-related factors, such as plasma levels of plasminogen activator inhibitor type 1, fibrinogen, and factor VII53 in overweight children.
The BMI is a clinical indicator of overweight in adults.40 Its use as an indicator of overweight in children is still being discussed.41 Therefore, we used 2 anthropometric measures of body fat in the study: the BMI as an indicator of overall body fat, including the visceral fat depots, and the sum of 3 skinfolds as an indicator of subcutaneous body fat. Classification of overweight using the BMI or using the sum of 3 skinfolds consistently showed a higher prevalence of low-grade systemic inflammation in overweight children.
Two potential limitations regarding the assessment of elevated serum CRP concentration should be discussed. First, in this study we used a single CRP measurement, which may not accurately reflect long-term inflammation status. The biological variability of CRP is substantial, with reported values ranging between 10.6% and 63.0%.54-57 However, because random misclassification caused by biological variability will lead to underestimation of true associations, this limitation is unlikely to explain the study findings. Second, the definition of elevated serum CRP concentration was based on the detection level of the CRP assay. The conventional cutpoint for elevated CRP concentration (a concentration >1 mg/dL) was not used because the prevalence of elevated serum CRP concentration using this criteria was too low (only 1.6% in boys and 1.8% in girls) to be used as the study outcome. However, when the analyses were repeated using the 95th percentile of serum CRP concentration as the cutpoint for elevated CRP (4-11 years of age: >.37 mg/dL for boys and >.68 mg/dL for girls, and for 12-19 years of age: >.65 mg/dL for boys and >.67 mg/dL for girls),35 similar results were obtained. For example, among healthy, nonsmoking, nonestrogen-using children, overweight boys were 6.12 (95% CI: 1.23-30.52) and 7.11 (95% CI: 2.52-20.06) times more likely to have an elevated CRP concentration based on the BMI and the sum of 3 skinfolds, respectively. For overweight girls these numbers were 5.59 (95% CI: 2.20-14.22) and 3.77 (95% CI: 1.42-9.99), respectively. Thus, using a more extreme cutpoint to define elevated CRP in children did not change the conclusions of the study.
Measurements of the serum concentration of interleukin-6 were not available in the present study. Although the results support the hypothesis that interleukin-6 produced by adipocytes increases CRP concentration, direct assessment of interleukin-6 concentration is needed in future studies to further test this hypothesis.
| |
CONCLUSION |
|---|
|
Top
Abstract
Methods
Results
Discussion
Conclusion
References
|
|---|
The results of this large-scale cross-sectional study show that overweight is associated with higher CRP concentrations and higher white blood cells counts in children, which could not be explained by disease or other factors associated with inflammation. In children, subclinical disease is unlikely to explain these findings. These data suggest a state of low-grade systemic inflammation in overweight children.
| |
FOOTNOTES |
|---|
Received for publication May 3, 2000; accepted Aug 16, 2000.
Reprint requests to (M.V.) Institute for Research in Extramural Medicine, Faculty of Medicine, Vrije Universiteit, Van der Boechorststraat 7, 1081 BT, Amsterdam, The Netherlands. E-mail: m.visser.emgo{at}med.vu.nl
| |
ABBREVIATIONS |
|---|
CRP, C-reactive protein; BMI, body mass index; NHANES III, third National Health and Nutrition Examination Survey; OR, odds ratio; CI, confidence interval; SD, standard deviation; SE, standard error.
| |
REFERENCES |
|---|
|
Top
Abstract
Methods
Results
Discussion
Conclusion
References
|
|---|
This article has been cited by other articles:
 |
|
 |
|
  J J Reilly, E Methven, Z C McDowell, B Hacking, D Alexander, L Stewart, and C J H Kelnar Health consequences of obesity Arch. Dis. Child., September 1, 2003; 88(9): 748 - 752. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2001 American Academy of Pediatrics. All rights reserved. |
||
|
||
P < .05 versus lowest category.