Association between tracking of extracurricular sports practice and weight status during childhood: a prospective cohort study

ABSTRACT BACKGROUND: Overweight and obesity have reached epidemic prevalences. Obesity control involves many factors and needs to begin early in childhood. OBJECTIVES: To ascertain the association between tracked extracurricular sports practice and weight status; and to analyze tracking of overweight and obesity among school-aged children. DESIGN AND SETTING: Prospective cohort study conducted in 13 public schools in Cianorte, Paraná, in 2012-2016. METHODS: The sample comprised 2459 schoolchildren in Cianorte, of mean age 6.3 years at baseline and 9.4 years at follow-up. Body mass index was calculated from body mass and height measurements. The children were grouped as normal weight, overweight or obese. Information on extracurricular sports practice was collected through the dichotomous question “Do you participate in any extracurricular sports?” (“yes” or “no”). RESULTS: Tracking of weight status showed that 75.5% maintained this, with kappa of 0.530. Tracking of extracurricular sports practice showed that 80.9% maintained this, with low concordance (kappa of 0.054). Weight status correlation between baseline and follow-up showed that overweight or obese individuals were 4.65 times (CI: 4.05-5.34) more likely to maintain the same classification or move from overweight to obese at follow-up. Correlation of extracurricular sports practice with overweight or obesity at follow-up was not significant. CONCLUSIONS: These results demonstrated that overweight or obese children were at higher risk of gaining weight than were normal-weight children. In addition, the proportion of these children who maintained extracurricular sports practices over the years was low. Maintenance of this variable was not associated with weight status.


INTRODUCTION
Elevation of body mass index (BMI) levels has reached an alarming status within public health, both in developed and in developing countries. This has been seen as a gradual rise in the prevalence of overweight and obesity in different age groups. [1][2][3][4][5] In a systematic review performed by Ng et al., 6 the prevalence of overweight or obesity was reported to be approximately 47% among children worldwide. Among the different factors that have contributed to this increase is a decline in physical activity levels.
In Brazil, it has been estimated that 269.6 million dollars are spent annually on combating obesity and related complications. 7 A survey conducted in state capitals in 2014 showed that the prevalence of overweight in the Brazilian adult population was more than 51%. 8 Among children and adolescents, the prevalence of overweight was 22% and 24%, respectively. 9 More than 70% of subjects who presented higher adiposity in childhood remained under the same condition during adolescence. 10 This is worrying, as the adolescent population presents a higher risk of maintaining this indicator in adulthood. 11,12 The Bogalusa Heart Study, conducted between the 1970s and 1990s, indicated that obesity in children and adolescents is linked, respectively, to 2.4 and 7.1 times higher risk of presenting high cholesterol and triglycerides, in comparison with the population with normal body mass index. 12 This increase in body mass index is associated with different risk factors for cardiovascular diseases, for example: high blood pressure, dyslipidemias, fasting insulin and risk score. 13 Sports practice, predominantly performed in schools, has been considered to be one of the main forms of elevating daily energy expenditure among school-age children, and this is associated with weight control. 15 However, little is known about the longitudinal impacts of sports practice on prevention of overweight in childhood, since the majority of studies have attempted to observe this occurrence using cross-sectional designs. [15][16][17] Such studies do not longitudinally evaluate changes in or maintenance of weight status, or sports practice, in this population.

OBJECTIVE
The aim of this study was to ascertain the association between tracked extracurricular sports practice and weight status; and to analyze the tracking of overweight and obesity among schoolchildren during childhood.

Study design and sample selection
The data used in this study formed part of a database named: The sample was composed of all students enrolled in the municipality's 13 public schools, aged 5-8 years. All students were invited to participate in the research, and only those who were absent on the day of the test were not included in the sample. For the present study, only the data from the students who underwent both baseline and follow-up measurements (follow-up three years later) were used in the analysis, totaling 2,459 subjects in the sample. The total number of children who dropped out from the research was 981 individuals. A comparison between the children who dropped out and those who completed the follow-up is presented in Table 1.

Weight status classification
Body mass was assessed using a platform digital scale (model CA8000; G-life, São Paulo, Brazil), with precision of 0.1 kg.
Height was measured using a portable stadiometer (Alturexata, Belo Horizonte, Brazil), with accuracy of 0.1 cm. The protocols followed the procedures described by Gordon et al. 18 Body mass index (BMI) was determined from the body mass/height 2 quotient, in which body mass was expressed in kilograms (kg) and height in meters (m).
The BMI cutoff points used followed the recommendations of the World Health Organization, 19 in agreement with age and sex.
Body mass index values were classified as underweight, normal weight, overweight or obese. For the purposes of this analysis, subjects classified as underweight and normal weight were grouped as normal weight (NW), while overweight (OW) and obesity (OB) were maintained as separate categories.

Sports practice
Extracurricular sports practice was observed during data collection every year throughout the study, using the following dichotomous question "Do you participate in any extracurricular sports?" with answer options of "yes" or "no". In the analysis in which extracurricular sports practice was tracked, the subjects were grouped into the following categories: practice to practice (P-P); practice to no practice (P-NP); no practice to practice (NP-P); or no practice to no practice (NP-NP).

Statistical analysis
The sample were characterized in terms of medians and interquartile ranges. Data normality was tested using the Kolmogorov-Smirnov test, and comparisons were then made using the Wilcoxon test. BMI was tracked in relation to extracurricular sports practice and was evaluated through the percentage of subjects who remained in the same classification over the years. The agreement between baseline and follow-up was calculated by means of the kappa index, which takes into account the proportion of observed (Po) and expected (Pe) agreement (kappa: Po -Pe / 1 -Pe); analyses were performed stratified according to sex and group. Associations were analyzed using hazard ratios (HR) and assessed using the Cox multivariate regression model. All analyses were adjusted for sex, age, extracurricular sports practice and BMI at baseline. The software used was SPSS 25.0 (IBM, New York, United States) and values were considered significant at P < 0.05.  Table 2. It can be seen that 74.4% of the boys presented maintenance of weight status classification over the three years, with kappa classified as moderate (K = 0.523; P < 0.01), while 76.8% of the girls maintained this, with moderate kappa index (K = 0.537; P > 0.01). In contrast, approximately 25% of the subjects changed their weight status classification. This occurred mostly among individuals in the normal weight group, who increased their BMI. Only a small proportion of the overweight or obese children at the baseline changed their weight status to normal weight. Table 3 shows that the children tended to remain within the same sports practice classification, since tracking values of 76.5% and 84.2% were reported for the boys and girls, respectively. The incidence of extracurricular sports practice was extremely low

DISCUSSION
The aim of the present study was to ascertain the association between tracked extracurricular sports practice and weight status, and to track overweight and obesity among schoolchildren during three years of follow-up. The main finding was that children with excess weight maintained their classification after three years of evaluation. In addition, no association was found between tracked extracurricular sports practice and weight status. Only a low number of children participated in sports outside school.
Although no direct association was found between extracurricular sports practice and the prevalence of obesity at the time of the follow-up, it was observed that a large number of children did not practice sports frequently: approximately 85%. This is worrying, since sports practice may be directly associated with levels of physical activity and cardiorespiratory fitness during adolescence.
Werneck et al. 20 showed that sports practice was indirectly associated with metabolic risks and demonstrated that there was an association between sports practice during childhood and physical activity during adolescence.
In a systematic review, Hallal et al. 21 confirmed the hypothesis that sports practice during childhood was directly linked to physical activity levels during adolescence. In addition, a study in Finland on 7,794 individuals found that sports practice during childhood led to greater likelihood that these children would become physically active adults. 22 Moreover, maintenance of sports practice during adolescence is an essential indicator of weight status in adulthood; maintenance of sports practice during adolescence plays an important role in controlling BMI in adulthood. 23 One possible explanation for the present results, in which no association was found between extracurricular sports practice and weight status, is that the sports practice tracked tended to be significant only for late adolescence and adulthood. Tracking of sports practice from early adolescence to adulthood shows that this practice seems to be not significant, 24 thus suggesting that there is a specific point within adolescence (late adolescence) that influences individuals' involvement in sports practice.
Moreover, sports practice has been correlated with daily calorie expenditure, 25 which seems to be influenced by the intensity, duration, and frequency of sports. This is another possible factor affecting the association between sports practice and weight status, In other words, they consume larger amounts of sugary foods and spend more time watching television. 30 However, these hypotheses could not be verified in the present study.
Likewise, a study in which body adiposity was tracked among children, based on skinfold measurements (subscapular and triceps) and controlled in relation to baseline body mass index, showed that 70% of the subjects remained in the highest tercile.
It was also found that approximately 29% of the subjects classified in the low and medium terciles presented increased adiposity and were reclassified into the highest tercile over the three-year period. 10 The results from that study were similar to those of the present study, in which 12.0% of the subjects changed from normal weight to overweight or obesity.