C-reactive protein is inversely and independently associated with maximal oxygen uptake in adolescents with obesity

João Elias Dias NUNES*/** Heitor Santos CUNHA */** Renata Roland TEIXEIRA ** Foued S. ESPINDOLA ** Nádia Carla CHEIK */** * L a b o r a t o r y o f t h e P h y s i o l o g y o f P e r f o r m a n c e , Federal Univers i ty o f U b e r l â n d i a , Uberlândia, MG, Brazil. **Laboratory of the B iochem is t r y and Molecular Biology, Federal Univers i ty o f U b e r l â n d i a , Uberlândia. MG, Brazil. Abstract


Introduction
C-reactive protein is inversely and independently associated with maximal oxygen uptake in adolescents with obesity Th e increased prevalence of obesity in adolescence has become a public health problem 1 , because obesity in childhood and adolescence is strongly correlated with adult obesity, which reduces the quality of life of individuals and increases the economic costs of health care 2 .
Recent studies have demonstrated a systemic infl ammation in obese children and adolescents, characterized by high level of pro-infl ammatory cytokines and chemokines 3,4 . Another studies show a mediator role of systemic infl ammation in the development of chronic diseases such as cardiovascular diseases and type 2 diabetes [5][6][7] . Therefore, adolescents with obesity are more exposed to possible comorbid conditions.
A high cardiorespiratory fi tness is associated with a reduced systemic infl ammation 8 . Previous studies have shown an inverse association between the C-reactive protein (CRP) and cardiorespiratory fi tness [9][10][11] . However, research on the association between markers of infl ammation and cardiorespiratory fi tness were not performed with the obese adolescent, which would be important due to potential future prevention of comorbidities in this population.
In addition, the association between infl ammatory markers and a practical parameter of prescription of training, which can be used in the field, as the velocity associated with maximum oxygen consumption (vVO 2max ), has not yet been studied. The vVO 2max is defined as the speed which elicited the highest oxygen consumption, and was first examined by Cooper 12 , as an alternative method for predicting VO 2max in the effort to simplify the procedures of the tests and Methods reduce costs.
Thus, the purposes of this study were: 1) investigate whether different markers of infl ammation (CRP and TNF-) are correlated to vVO 2max and VO 2max in adolescents with obesity; 2) examine the association of these variables when adjusted by parameters of body composition.
One hundred and seventy-fi ve participants were initially recruited for participation in this study which were medically screened, their pubertal stage was assessed, and their anthropometric measures were recorded (ie, height, weight and BMI). Th e endocrinologist completed a clinical interview including questions to determine eligibility based on inclusion criteria: being post-pubescent, according to the classifi cation of Tanner and Whitehouse 13 and obese, BMI>95th percentile of the curve proposed by the Center for Diseases Control (CDC); and exclusion criteria: identifi cation of genetic, metabolic or endocrine diseases, and previous use of drugs. At the end of this screening were selected 57 individuals, 34 girls and 23 boys (TABLE 1). Th is study was approved by the Ethics Committee and Research in Human Beings of the Federal University of Uberlândia-Brazil and all the participants and their parents have signed a consent form.
Weight was measured by an electronic balance and height by a stadiometer. Body mass index (BMI) was calculated as kilograms per square meter. Th e abdominal circumference (AC) was assessed by perimeter of the abdomen at the height of the navel with a anthropometric tape (Sanny, SN-4010, Brasil). Body composition was estimated by analyzer of tetrapolar bioelectrical impedance (Biodynamics, BIA310e, USA). Visceral fat (VF) was estimated by ultrasonography (Mindray, DC-6, China), defi ned as the distance between the internal surface of the abdominal rectus muscle and the anterior wall of the aorta 14 . All abdominal ultrasonographic (US) procedures and the measurements of visceral fat tissue were performed by the same physician. The intraexamination coeffi cient of variation for US was 2.6%.
Blood samples were obtained by venipuncture after an overnight fast. Plasma TNF- was measured by a quantitative two-site high-sensitivity enzyme immunoassay (DRG Diagnostics, Germany), CRP by a polystyrene particle enhanced highsensitivity immunoturbidimetric assay (Aptec Diagnostics, Belgium) Graded maximal exercise testing (2min by stage, initial speed -3km.h -1 , increase of 1km.h -1 by stage and treadmill inclination fi xed at 1%) was performed on a motor-driven treadmill (Movement, RT 250pro, Brasil). A ergospirometer (Cosmed, FitMatePro, Italy) was used for analysis of the consumption of oxygen breathing by breathing. Th e maximum oxygen consumption (VO 2max ) was defi ned as the highest 30-s average during exercise. Th e velocity associated with the VO 2max (vVO 2max ) was obtained by the following formula: vVO 2max (km.h -1 ) = v + 1.0 x (t/120). Where v is the last velocity completed for 120 seconds; t the number of seconds the fi nal not completed velocity was sustained; 1.0 is the value of the increase in velocity in km.h-1 from the last stage and 120 is the number of seconds in the stage 15 .
Initially, we tested the normality of all variables (Shapiro Wilk's test). CRP, TNF-, relative VO 2max and vVO 2max were logarithmically transformed [log(x+1)] to reduce skewness. Descriptive statistics are given as mean (SD) (as well as range). The correlation between the parameters of cardiorespiratory fi tness (VO 2max , vVO 2max ) and the markers of systemic infl ammation (CPR and TNF-) was performed by Pearson's correlation coeffi cient. To assess whether infl ammatory markers were independently associated with cardiorespiratory fi tness after adjustment for potential confounders (body composition), was performed multivariable regression analyses. Predictive variables were included in models with a signifi cance level of p≤0,10. Statistical signifi cance was inferred at p≤0,05. Statistical analyses were performed with Statistica version 7.0. The participants characteristics are shown in TABLE 1. Significant correlations of CRP with VO 2max and vVO 2max are shown in FIGURE 1A and B.
No correlations were observed between TNFand VO2max e vVO2max ( figure 1C and D), the association of these factors adjusted for body composition did not change the results (TABLE 2).

Discussion
Th e main fi nding of the present study was that, CRP was inversely related to VO 2max and vVO 2max (FIGURE 1A and B), even after adjustment for potential confounders such as BMI, AC and VF (TABLE 2), in adolescents with obesity. Although other studies have demonstrated the inverse correlation between infl ammatory markers and cardiorespiratory fi tness [9][10][11] , this is the fi rst study to demonstrate: 1) this association in obese adolescents and 2) a signifi cant correlation between the CRP and vVO 2max .
It has been recommended for subjects with chronic diseases, including obese, exercise of moderate and vigorous intensities 16-18. In this sense, the vVO 2max can be used as a reference for prescription of training, as shown in some studies 19,20 . It is a parameter which can be obtained in practice: it is viable and not requiring equipment of high cost and specialized personnel. Adding to this, the fi ndings of this study, it becomes clear the usefulness of vVO 2max as a tool in the exercise evaluation for obese adolescents.
As vVO 2max and the VO 2max show similar correlations with inflammatory markers (FIGURE 1) and indicators of body composition (TABLE 3), we can infer that it can be reliably used for prescription of training in obese adolescents. Th e parameter most scientifi cally studied, VO 2max , is not commonly available at training centers, gyms, parks and schools, and thus rarely used in the training prescription. In addition, the exercise based on vVO 2max , reduces the empiricism and generalization, because suits the intensity to biological individuality of obese adolescents.
Th e levels of CRP were associated to the VO 2max and vVO 2max independently of measurements of body composition (TABLE 2). One study with 172 subjects aged 26-84 years also found a signifi cant correlation (r = -0.40) between CRP and VO 2max , which is in accordance with the data found in this study 11 . Th is fi nding is also consistent with the results of previous studies 10, 21 , however more studies are needed to clarify the reasons for this association.
The motives for the correlation between cardiorespiratory fi tness and levels of CRP are not well elucidated. However, hepatic CRP production is stimulated by interlukin-6 (IL-6) and, to a lesser extent, by iterleukin-1 (IL-1) and TNF- 22 . Individuals who are obese and/or hyperinsulinemic have increased adipocyte production of inflammatory markers, including IL-6 and TNF- [23][24][25] . Several studies suggested a close relationship between cardiorespiratory fi tness and infl ammatory status 26,27 . Although we have not examined the concentrations of IL-6 and IL-1, our study demonstrated a production of TNF- in obese adolescents (TABLE 1), similarly to other studies 28 . Whereas, the inverse correlation between the accumulation of body fat and cardiorespiratory fi tness is well established, therefore, the relationship between, CRP, VO 2max and vVO 2max should be mediated by the adipose tissue.
Association between the VO 2max , vVO 2max and the TNF- was not found (FIGURE 1C and D). Th e reason for increased circulating TNF- levels observed in obese people is not thought to be associated with over-production in the adipose tissue. It is hypothesized that systemic eff ects of leptin or other adipokines may induce TNF- secretion from macrophages and lymphocytes 29 . However, TNF- is involved in the pathophysiology of obesity-associated insulin resistance and atherosclerosis 30,31 , therefore, the study of its association with the variables related to obesity still deserves attention.
Limitations of the present study include the relatively small sample size and cross-sectional design. Future studies are needed to identify which infl ammatory pathway is associated with cardiorespiratory fi tness and determine the pathophysiological mechanisms involved in this association.
In conclusion, the CRP was inversely associated with direct measurement of VO 2max and the indirect variable of cardiorespiratory fitness vVO 2max , in obese adolescents, even after adjustments in body composition. No association was found between TNF- and the parameters of cardiorespiratory fi tness.