High-sensitivity C-reactive protein is a good marker of cardiovascular risk in obese children and adolescents.

doi:10.1530/EJE-08-0212

HOME

HELP

FEEDBACK

SUBSCRIPTIONS

High-sensitivity C-reactive protein is a good marker of cardiovascular risk in obese children and adolescents.

Leandro Soriano-Guillén¹^,2, Bárbara Hernández-García¹, Jimena Pita², Nieves Domínguez-Garrido¹, Genoveva Del Río-Camacho¹ and Adela Rovira²^,3

^¹ Deparment of Pediatrics, Fundación Jimenez Diaz, University Autonoma of Madrid, Avenida Reyes Catolicos no 2, 28040 Madrid, Spain^² Laboratory of Endocrinology and ^³ Deparment of Endocrinology. Fundacion Jimenez Diaz. University Autonoma of Madrid,, Madrid, Spain

(Correspondence should be addressed to L Soriano-Guillén; Email: lsoriano{at}fjd.es) (leansor4{at}hotmail.com)

Abstract

Top
Abstract
Introduction
Subjects and methods
Results
Conclusion
Disclosure
References

Objective: We intend to assess the utility of the high-sensitivity C-reactiveprotein (hs-CRP) as a marker of cardiovascular risk in obesechildren and adolescents.

Methods: The study included children and adolescents between 6 and 18years of age with a body mass index (BMI) higher than 2 SDS.All the patients had their blood pressure taken and hs-CRP,hepatic function, lipid profile and uric acid were determinedafter 12 h of fasting. Likewise, an oral glucose tolerance testwas performed, determining basal glucose and insulin levels,and after stimulus. We considered the presence of metabolicsyndrome when the obese children and teenagers showed at leasttwo of the following conditions: decreased high density lipoprotein(HDL)-cholesterol, hypertriglyceridemia, hypertension or alterationin glucose metabolism.

Results: Out of the 115 obese children studied, 24% showed signs of metabolicsyndrome. Those with metabolic syndrome presented higher levelsof hs-CRP (mean: 3.8 mg/l; 95% CI: 2.8–4.8) in comparisonwith the obese patients who did not show signs of metabolicsyndrome (mean: 2 mg/l; 95% CI: 1.5–2.5). After a multivariateanalysis, the variables that appear to influence the changesin hs-CRP were BMI, triglycerides and HDL-cholesterol levels.

Conclusion: The hs-CRP is a useful tool for early diagnosis of cardiovascularrisk in obese children and teenagers.

Introduction

Top
Abstract
Introduction
Subjects and methods
Results
Conclusion
Disclosure
References

The rise in C-reactive protein (CRP) as a response to the increasein the secretion of cytokines of adipose origin detected inobese individuals (1) has been used as a marker of cardiovascularrisk and diabetes in adults (2, 3). When measured with new high-sensitivityassays, the levels of CRP have proven to predict future cardiovascularrisk. Among apparently healthy men and women, the levels ofhigh-sensitivity-CRP (hs-CRP) of <1, 1–3 and >3mg/l distinguish between those at low, moderate and high riskof future cardiovascular disease (4).

Later studies have shown high values of hs-CRP among obese childrenand adolescents, indicating that in early stages there alreadyexists a certain level of inflammation in relation to the non-obesepopulation (5, 6, 7, 8). However, not many data are availablethat guarantee its utility as a marker of cardiovascular riskand diabetes in childhood (7, 8). Therefore, our intention isto study the levels of hs-CRP in obese children and adolescentswith and without metabolic syndrome to evaluate the utilityof this parameter as a method of early diagnosis of cardiovascularrisk in early stages of life and to analyse whether there arevariables associated with higher levels of CRP.

Subjects and methods

Top
Abstract
Introduction
Subjects and methods
Results
Conclusion
Disclosure
References

This is a cross-sectional study of obese children and adolescents.From all children and adolescents between 6 and 18 years ofage, who came to our Unit (Pediatric Endocrinology Unit of HospitalFundacion Jimenez Díaz, Madrid) and who were suspiciouslyobese, we recruited those who had body mass index (BMI)>2SDS for their age and sex according to Spanish BMI data andwho did not suffer any chronic illness or acute infectious disease.

Considering that the average level of hs-CRP in the normal populationis around 1 mg/l, to achieve a 50% difference in hs-CRP levelsamong obese individuals with and without metabolic syndrome,it is necessary to obtain a minimum sample size of 90 patientswith a statistical power of 90% and a confidence level of 95%.

These children and adolescents had their weight (kg), height(cm) and blood pressure (mmHg) taken thrice during a periodof time of 10 min (we calculated the mean systolic and diastolicblood pressure). After 12 h of night fasting, blood sampleswere taken to measure hs-CRP, lipid profile, hepatic function,basal glucose and insulin levels. Later, a glucose tolerancetest was carried out (1.7 gxkg, maximum=75 g) and then, 120min later, glucose and insulin levels were measured again. Takinginto account that all our patients met one criterion of metabolicsyndrome (being obese), metabolic syndrome was considered whenat least two other of the following criteria were present (9,10): HDL cholesterol below 5 percentile and/or triglyceridesabove 95 percentile for the age and sex, diastolic and/or systolicblood pressure higher than 95 percentile for the age, sex andheight (11), and alteration in glucose metabolism accordingto criteria of the American Society of Diabetes (12). Otherstudy variables taken into account include age, sex, race andTanner stages of pubertal development.

This project was approved by the Ethics Committee of our hospitaland it was compulsory to sign the corresponding informed consentin order to take part.

Student's t-test was used to compare independent data of normaldistribution, applying the Mann–Whitney U test when thevariables did not follow a normal distribution. Finally, inorder to analyse the correlation between the levels of hs-CRPand the study variables, a univariate analysis was carried outand later a multivariate linear regression was performed. Thedata were analysed using the SAS software package (SAS InstituteInc., Cary, NC, USA).

Results

Top
Abstract
Introduction
Subjects and methods
Results
Conclusion
Disclosure
References

The data have been expressed as mean and in 95% confidence interval.A total of 115 obese children and adolescents were studied,of whom 28 met metabolic syndrome criteria (24%). The childrenand adolescents with metabolic syndrome presented significantlyhigher levels of hs-CRP (mean: 3.8 mg/l; 95% CI: 2.8–4.8)in comparison with those who did not present metabolic syndrome(mean: 2 mg/l; 95% CI: 1.5–2.5) (Table 1). There werealso significant differences between these two groups when analysingthe levels of basal insulin and 120 min later, triglycerides,HDL-cholesterol and uric acid. By contrast, no significant differencesbetween the two groups were observed regarding age and BMI.Finally, there were no differences in the levels of hs-CRP whencomparing the different pubertal stages, race and sex.

View this table:
[in this window]
[in a new window]

Table 1 Comparison of clinical and biochemical data among obese children with or without metabolic syndrome.

After performing a simple linear regression between the differentquantitative variables that could affect the hs-CRP, the onlyones that showed a linear relationship were BMI (r=0.35, P<0.01),basal insulin (r=0.3, P=0.01), triglycerides (r=0.33, P<0.01)and HDL-cholesterol (r=–0.3, P<0.01) (Fig. 1), althoughwith a remarkable dispersion. Later on, a multivariate linearregression model was built considering hs-CRP as a dependentvariable and BMI, basal insulin, triglycerides and HDL-cholesterolas independent variables. In this model, the only significantvariables were BMI, triglycerides and HDL-cholesterol, witha determination coefficient of R²=0.25.

View larger version (18K):
[in this window]
[in a new window]

Figure 1 Linear regression analysis among hs-CRP and BMI, triglycerides, HDL-cholesterol and basal insulin.

	Conclusion

Top Abstract Introduction Subjects and methods Results Conclusion Disclosure References

The hs-CRP is significantly increased in obese children andadolescents with metabolic syndrome in comparison with the groupwithout metabolic syndrome. With this study, we have shown thatBMI is not the only variable that can affect the increase inhs-CRP, unlike what has been described in previous studies (5,6). Therefore, a possible interaction is found between hs-CRPand the levels of triglycerides and HDL-cholesterol, suggestingits utility as a metabolic risk marker (7, 8). This interactionbetween hs-CRP, triglycerides and HDL-cholesterol could be mediatedby different cytokines of adipose origin that in some geneticallypredisposed individuals are elevated in relation to higher abdominaladiposity (1). However, more prospective studies will be necessaryin order to clarify this relationship.

The absence of influence of age or Tanner pubertal stage inhs-CRP levels shows the utility of this parameter as a metabolicrisk marker in early stages of life (8).

A limitation of our study is that with the multivariate modelbuilt we only partially explain the changes observed in hs-CRP.Therefore, it is necessary to carry out more studies that takeinto consideration other variables such as adiponectin, interleukin-6and tumor necrosis factor- ${alpha}$ , considered by other authors andwhich in this case have not been studied (13).

In short, hs-CRP seems to be an excellent marker in order todistinguish the presence of metabolic syndrome among obese childrenand adolescents and could be a useful tool for the early detectionof cardiovascular risk factors among this population.

Disclosure

Top
Abstract
Introduction
Subjects and methods
Results
Conclusion
Disclosure
References

The work presented does not have any source of funding. Thereis no conflict of interest that would prejudice its impartiality.