Aerobic fitness and biological and sociodemographic indicators in female school children

Aerobic fitness is an important health-related fitness component. The objective of this study was to analyze the influence of age, growth, body composition, sexual maturation and socioeconomic status on the aerobic fitness of female school children attending public schools in Cascavel, Paraná, Brazil. A cross-sectional epidemiological study was conducted in 2006 on 1,317 school children aged 8 to 17 years. Body mass, height and triceps and medial calf skinfold thickness were measured and relative body fat (%BF) was calculated. Sexual maturation was self-assessed using a breast development scale (M1 to M5). The questionnaire proposed by the Brazilian Association of Research Companies was used to assess socioeconomic status. Aerobic fitness was estimated by the 20-m shuttle run test. The results were analyzed using two-way ANOVA, Tukey post hoc test, Spearman and Pearson correlation coefficients, and multiple linear regression with 95% confidence intervals. Three variables (age, %BF and socioeconomic status) showed an interaction (p<0.05) in the prediction of VO2peakrel (R 2=0.73). All independent variables (body mass, height, age, %BF, socioeconomic status, and sexual maturation) were significant predictors (p<0.05) of VO2picoabs (R 2=0.88). In conclusion, chronological age and sexual maturation were the best predictors of VO2peakabs (positive) and VO2peakrel (negative).


INTRODUCTION
Studies on children and adolescents have emphasized the importance of the assessment and intervention in health-related components of physical fitness 1,2 for the adoption of a healthy lifestyle and the prevention of diseases such as obesity, diabetes and hypertension 3,4 .Cardiorespiratory or aerobic fitness is an important health-related component since it describes the capacity of the cardiovascular and respiratory systems to supply oxygen to muscles during continuous effort 5 .This component shows a strong association with health indicators 6 .
Peak oxygen uptake (VO 2peak ), i.e., the maximal oxygen consumption during an exercise test to exhaustion, has been shown to be the best indicator of aerobic fitness in children and adolescents 7 .This component of physical fitness is influenced by gender, body composition, sexual maturation, age, and socioeconomic status [8][9][10][11] .Boys present a higher VO 2peak than girls and this difference between genders seems to be related to the higher accumulation of body fat in girls and the increase of lean mass in boys during puberty 8,9 .Investigators have shown a decrease in aerobic fitness with increasing age 11 , which is due to the increase of body mass during adolescence 9,10 .In addition, school children (10 to 15 years) of low socioeconomic status exhibit lower levels of aerobic fitness 10 .
Within this context, knowledge of factors that influence aerobic fitness in girls during childhood and adolescence contributes to data interpretation and to the establishment of prevention and intervention measures in this group, with benefits during the present and subsequent phases of life.Therefore, the objective of the present study was to evaluate the influence of age, growth variables, body composition, sexual maturation and socioeconomic status on the aerobic fitness of female school children aged 8 to 17 years from public schools in Cascavel, Paraná, Brazil.

METHODOLOGICAL PROCEDURES
This study is part of a larger cross-sectional epidemiological study entitled "Anthropometry, body composition, motor performance and sexual maturation of school children of different socioeconomic classes from the municipality of Cascavel, Paraná".The study was approved by the Ethics Committee on Human Research of Universidade Federal de Santa Catarina (UFSC) (Permit No. 131/2006).
The municipality of Cascavel is located on the Third Plateau in the western region of the State of Paraná, southern Brazil.The municipality has a population of 286,205 inhabitants, 94.02% of them living in the urban area 11 .
The target population of the present study consisted of female school children ranging in age from 8 to 17 years and living in the urban area of the municipality.The students attended public schools in Cascavel and were enrolled in the 3 rd to 8 th grade of elementary school and in the 1 st to 3 rd year of high school in 2006.According to the report of the Regional Education Center and the City Department of Education (2006), the general population of school children enrolled in elementary and high schools was approximately 34,720 individuals attending 89 municipal and state public schools.Since the target population of the study included only female school children, a gender distribution of 50/50 was assumed, corresponding to 17,360 female students.
The sample was stratified according to educational center and class cluster.Three educational centers were formed based on the distribution of the schools across the different geographic regions of the municipality.The geographic division proposed by the Regional Education Center of Cascavel was used to ensure the representativeness of the sample.The proportion was 35.8% school children in center I, 33.1% in center II, and 31.2% in center III.In the first phase, four schools per center, two municipal and two state schools, were selected by drawing lots.Next, the schools that would participate in the study were chosen by drawing lots based on a list provided by the institutions that contained the students' age.In the second phase, simple random selection of the classes was performed in such a way that the sample was representative of the target population.All adolescent girls aged 8 to 17 years, who were present in the classroom on the day of data collection, were invited to participate in the study.
Since this study was part of a larger project, different sample sizes were calculated.In the present study, an unknown prevalence of the outcome (50%), a tolerable error rate of 5%, 95% confidence interval, and design effect of 2.5 were adopted, thus estimating a sample of 920 school children.An extra 20% were added for possible losses and refusals and the final sample consisted of 1,152 school children.In view of the characteristics of the sampling process that involved all individuals belonging to the clusters, 1,317 school children were included in the sample.
Eligibility criteria were enrollment in a municipal or state public school, presence in the classroom on the day of data collection, and an age of 8 to 17 years.Criteria for exclusion were (a) the lack of written informed consent by the responsible person; (b) refusal to participate in the study, and (c) an incomplete socioeconomic questionnaire.
The chronological age of the students was determined based on the difference between the date of assessment and the date of birth, which was converted into decimal age as described by Ross and Marfell-Jones 12 .Body mass was measured on a bioimpedance scale (Tanita ® ) to the nearest 0.1 kg, with the subject being barefoot and wearing light clothing.Height was measured with a vertical stadiometer (Seca ® ) to the nearest 0.1 cm according to the procedures proposed by Gordon et al. 13 .Triceps (TSF) and medial calf (MCS) skinfold thickness was measured in the right hemibody to the nearest 1 mm using a Lange skinfold caliper (Cambridge Scientific Industries, Inc.), as described by Harrison et al. 14 .Body fat percentage (%BF) was estimated using the equation proposed by Slaughter et al. 15 by the sum of TSF and MCS: %F = 0.610*(TSF + MCS) + 5.1.VO 2peak was determined indirectly using the 20-meter shuttle run test proposed by Léger and Boucher 16 .The test consists of running a distance of 20 meters between two lines, accompanying the sound signal that indicates an increase in speed.The initial speed of the test is 8.5 km/h, with progressive increases of 0.5 km/h at intervals of one minute.The test is finished when the subject can no longer follow the pace and the speed of the last completed stage is recorded.VO 2peak relative (VO 2peak rel) (mL.kg -1 .min - ) was estimated using the equation: VO 2peak rel = 31.025+ 3.238 * V -3.248 * A + 0.1536 * V * I, where V = speed in km/h (in the stage achieved) and A = age in years.These results were also transformed into VO 2peak absolute (VO 2pe- ak abs) (L.min -1 ) 17 .This test has shown good validity (r=0.71) and reproducibility (r=0.89) in children and adolescents 18 and has also been used in test batteries from other countries such as FITNESSGRAM and EUROFIT 19 .
Socioeconomic status was defined using the questionnaire and classification of the Brazilian Association of Research Companies (ABEP) 20 , which estimates the purchasing power of families and classifies it from high to low into classes A, B, C, D and E. The students were assigned to high (A+B), medium (C+D), and low socioeconomic classes (E).
Sexual maturation was evaluated by self-assessment using the breast stages proposed Marshall and Tanner 21 (M1 to M5).The instructions were provided by a female examiner.
Descriptive statistics including means and standard deviations were used for statistical analysis.Body mass, height, %BF, VO 2peak rel and VO 2pe- ak abs were compared between ages and sexual maturation stages by two-way analysis of variance (ANOVA), followed by the Tukey multiple comparisons test to identify possible differences.Pearson's simple correlation coefficient was calculated to evaluate the correlation of VO 2peak rel and VO 2peak abs with age, anthropometric variables and body composition.Spearman's coefficient was used to determine the correlation with socioeconomic status and sexual maturation stage.Next, stepwise multiple linear regression analysis was performed to determine the effects of each independent variable (age, body mass, height, %BF, socioeconomic status, and maturation stage) on the dependent variables (VO 2peak rel and VO 2peak abs).Dummy variables were used to introduce the qualitative variables socioeconomic status and maturation stage in the multiple linear regression model.The data were stored and analyzed using SPSS ® 13.0 for Windows (Statistical Package for the Social Sciences).A level of significance of 5% was adopted for all variables.
Table 1 shows the mean and standard deviation of body mass, height %BF, VO 2peak rel, and VO 2peak abs.Comparison of these parameters to the subsequent age showed a significant increase in body from 11 to 14 years, in height from 10 to 13 years, in %BF only at 12 years, and in VO 2peak abs from 11 to 13 years, with a significant reduction at 17 years.Regarding VO 2peak rel, significant reductions were observed from 10 to 17 years.Values are reported as the mean (standard deviation); *p<0.05≠ compared to the subsequent year.
A significant increase in body mass, height, %BF, VO 2peak rel and VO-2peak abs was observed when compared to the more advanced maturation stage.No significant differences were found for %BF between M1 and M2 and for height between M4 and M5 (Table 2).Table 3 shows the correlation of VO 2peak rel and VO 2peak abs with age, body mass, height, %BF, maturation stage, and socioeconomic status.For VO 2peak rel, moderate and high negative correlations were observed with all independent variables, except for socioeconomic status which showed a weak positive correlation.VO 2peak abs showed moderate and high positive correlations with all independent variables, except for age and socioeconomic status.
Multiple regression analysis revealed a significant interaction of the three independent variables age, %BF and socioeconomic status in predicting VO 2peak rel (R 2 =0.73), and of all independent variables (body mass, age, height, %BF, socioeconomic status and maturation stage) in predicting VO 2peak abs (R 2 =0.88).The independent variables explained 73% of the variability in VO 2peak rel and 88% of the variability in VO 2peak abs.Thus, the models [VO 2peak rel = 63.073+ (-1.668 x age) -(0.0126 x %BF) + (0.335 x socioeconomic status) and VO 2peak abs = 0.546 + (0.035 x body mass) + (-0.082 x age) + (0.004 x height) + (-0.003 x %BF) + (0.025 x socioeconomic status) + (0.015 x maturation stage)] can be accepted since analysis of variance indicated that these variables significantly reduce the variation in VO 2peak rel and VO 2peak abs (F=1154.671and p=0.001)Values are reported as the mean (standard deviation).* p ≤ 0.05 ≠ compared to the more advanced maturation stage.The correlation of VO 2peak rel and VO 2peak abs obtained in the exercise test and the values calculated with the multiple linear regression model showed high correlations for the two variables (VO 2peak rel: r=0.84 and p<0.001;VO 2peak abs: r=0.94 and p<0.001) (Figure 1).

DISCUSSION
In general, the present results showed that in this group of school children body mass, height and VO 2peak abs began to increase at about 10 years of age and stabilized from 14 to 17 years, except for VO 2peak abs which decreased at 17 years.With respect to maturation stage, body mass, height, VO 2peak rel and VO 2peak abs also increased gradually when compared to the more advanced stage.For %BF, the differences occurred at 12 years of age when compared to age 13 and between pubertal stages (M2, M3, and M4).
In contrast, VO 2peak rel gradually decreased from 10 to 17 years of age and advancing stage of maturation.Furthermore, VO 2peak rel was negatively correlated with all independent variables, except for socioeconomic status, whereas VO 2peak abs showed moderate or high positive correlations with all variables, except for age and socioeconomic status.Multiple linear regression analysis revealed that age, %BF and socioeconomic status explained 73% of the variability in VO 2peak rel (R 2 =0.73), and age, body mass, height, %BF, socioeconomic status and sexual maturation explained 88% of the variability in VO 2peak abs (R 2 =0.88).
Similarly, studies indicate an almost linear increase of VO 2peak abs in girls up to age 13, which tends to stabilize at about 14 years of age 8 .Using the same test for the determination of aerobic fitness, Vasques et al. 10 observed a decline in VO 2peak rel from 10 to 15 years and concluded that this decrease is mainly a consequence of increasing body mass during this phase.Similar to the present findings, an increase of absolute values and a decrease of relative values have been reported for North American girls aged 8 to 16 years, with the latter being related to the increase in body fat 9 .In the girls studied here, the increase of body fat, which is accompanied by a significant increase of body mass, was more apparent with advancing sexual maturation.In this respect, %BF was negatively correlated with VO 2peak rel, in agreement with other studies that used the same test to estimate VO 2peak 23,24 .According to Armstrong 5 , VO 2peak values are conventionally expressed relative to body mass to control for the effect of growth in children and adolescents.However, the use of relative values results in a different scenario than the use of absolute values as demonstrated in the present study.This author also found that VO 2peak abs values increase by approximately 80% in girls from 8 to 16 years of age and that VO 2peak is correlated with body mass.In the present study, considering mean percentages, the increase of body mass in girls aged 8 to 17 years was approximately 47%, whereas the increase of VO 2peak abs was about 32%, a fact leading to a significant reduction of VO 2peak rel of approximately 34%.
Studies using regression models have demonstrated a positive effect of maturation stage on VO 2peak abs in children and adolescents 7,25 .In addition, this type of model permits to show the effects of age, gender and sexual maturation on VO 2peak , irrespective of the body size of boys and girls 25 .In a study using this type of model, a positive and significant effect of matura-tion stage on VO 2peak was observed when this parameter was included in the model 7 .Therefore, like chronological age, sexual maturation has also been recognized as a variable that can explain the results, irrespective of body mass and fat in boys and girls aged 11 to 17 years 7 .Still with respect to sexual maturation, as observed in the present study, other investigators also reported that advanced maturation was accompanied by an increase in VO 2peak abs 5 .However, in a study involving young female athletes of different modalities aged 10 to 17 years, analysis of variance and multiple regression analysis revealed no significant effect of sexual maturation on aerobic fitness 26 .
As demonstrated in the present investigation and in other studies, aerobic fitness of children and adolescents is influenced by age, growth, body composition and sexual maturation 7,9,[23][24][25] .As a consequence, the understanding of aerobic fitness is compromised if these factors are not taken into account at the time of data interpretation 5 .The use of VO 2peak expressed relative to body mass to control for differences and to permit comparison of the results is a widely accepted approach.However, data expressed in this manner should be interpreted with caution since the hypothesis that VO 2peak rel increases in direct proportion to body mass has been questioned and inadequate control of this parameter may impair the understanding of aerobic fitness during growth and sexual maturation 5 .
One factor that has been little explored, particularly in international studies, and that is relevant because of the major distortions observed in Brazil is socioeconomic status.In the present study, multiple linear regression analysis showed an influence of socioeconomic status on both VO 2pe- ak rel and VO 2peak abs.A study investigating the influence of socioeconomic status of adolescents from Florianópolis, Santa Catarina, Brazil, reported lower VO 2peak rel values in less privileged classes and, consequently, a higher proportion of adolescents who did not meet recommended health criteria 10 .In contrast, a study involving Danish school children found no differences in cardiorespiratory fitness between different socioeconomic classes 27 .
The motivation of adolescents to undergo the aerobic fitness test, a variable not investigated here, may have affected the results and is a limitation of the present study.Furthermore, the cross-sectional design of the study does not permit to establish causal relationships between the variables analyzed or to infer an increase or decrease of aerobic fitness as a function of age and sexual maturation stage.It should be noted that sexual maturation was obtained by self-assessment, an approach that may have been influenced by cultural issues.However, according to the literature, this method is reliable to estimate sexual maturation stage 28 .Clinical evaluation by a pediatrician may prevent such errors and permit to obtain more precise results.
The strengths of the study include the representativeness of the population studied, which permits inferences for the population of adolescent school children from Cascavel, Paraná.In addition, the test used for the evaluation of aerobic fitness in school children shows good validity 29 .The equations used for the calculation of VO 2peak rel and VO 2peak abs should not be applied indiscriminately since the models were based on a general population of female school children from Cascavel (Paraná, Brazil).These results are not intended to replace physical tests, but rather to identify variables that, alone or in combination, reveal variations in VO 2peak rel and VO 2peak abs during growth and therefore contribute to the understanding of aerobic fitness.

CONCLUSION
The present results showed that age and sexual maturation stage interfered positively and negatively with VO 2peak abs and VO 2peak rel, respectively.In addition, VO 2peak abs was influenced by all variables studied, whereas VO-2peak rel was influenced mainly by age, %BF and socioeconomic status.These variables should therefore be taken into account in the analysis of aerobic fitness in female children and adolescents.

Figure 1 .
Figure 1.Mean relative (VO 2peak rel) and absolute (VO 2peak abs) peak oxygen uptake obtained in the exercise test and by multiple regression according to age.Cascavel, PR, Brazil, 2006.