Sampling
This study is part of longitudinal research entitled Analysis of Behaviors of Children During Growth (ABCD – Growth Study). This is an on-going study designed to identify the impact of sports participation on different health aspects among adolescents, including metabolic syndrome components. The study is being carried out in the city of Presidente Prudente (200,000 inhabitants and human development index of 0.806, the western state of Sao Paulo, Brazil). Researchers and staff members of the Laboratory of InVestigation in Exercise (LIVE) (Sao Paulo State University - UNESP) were responsible for the 2017 baseline and follow-up data collection. The ethics committee of the UNESP campus of Presidente Prudente approved the study (process number 1.677.938/2016). All parents, coaches, and adolescents signed a written consent form.
The sampling process has been described elsewhere [25–27]. Briefly, after authorization of local authorities, school units and sports clubs were contacted to explain the research aims and methods of the ABCD – Growth Study. In the eleven facilities (spread out in the metropolitan area of the city) that were accepted to host the research, adolescents and their parents/legal guardians were contacted, and written consent forms were delivered. The only adolescent who fulfills all inclusion criteria were measured/interviewed: (i) aged between 11 and 18 years; (ii) absence of any known diseases previously diagnosed; (iii) no regular medicine use related to blood pressure or lipid metabolism; (iv) if athletes, at least one year of training experience; if control group (schoolchildren), at least one year without any regular engagement in organized sports or physical exercise routines; (v) written parental consent and adolescents’ assent, both signed.
At baseline, 285 adolescents agreed to participate and take part in the baseline data collection. Twenty-six adolescents had any missing data at baseline and were excluded from this manuscript (n= 259). Moreover, after 12 months of follow-up, there 75 dropouts due to fear about blood collection, moved to another city, lack of time to participate in data collection, or gave up participating (n= 184). Additionally, 13 adolescents were excluded due to any missing in the components of MetS on follow-up data collection. This sample size of 171 participants allowed the statistical power (80%) to detect significant (Z= 1.96) coefficients of correlation (standardized score [r]) ≥0.213 in our statistical analysis.
Metabolic syndrome components
Blood samples were collected in a private laboratory following at least twelve hours of fasting. High-density lipoprotein-cholesterol (HDL-c [mg/dL]), triacylglycerol (TG [mg/dL]), and glucose (mg/dL) were analyzed by the colorimetric method of dry chemistry and processed in a biochemical Autohumalyzer (Vitros, model 250) obtained from the Ortho Clinical Diagnostics company, Rochester, New York, USA. Systolic (SBP) and diastolic blood pressure (DBP) were assessed using an automatic device (Omron Healthcare, Inc., Intellisense, model HEM 742 INT, Bannockburn, Illinois, USA), validated for adolescents [28]. Mean blood pressure (MBP) was calculated using SBP and DBP [MBP = 1/3 (SBP – DBP) + DBP] [29]. Measurements were performed after 10 minutes at rest and three measurements were obtained with an interval of one minute between them and the mean of the three measurements was considered [30]. Body fatness (BF in percentage values [%]) was estimated using a densitometry scanner (General Electrics; model Lunar – DPX-NT, General Electric Healthcare, Little Chalfont, Buckinghamshire, United Kingdom) equipped with the software GE Medical System Lunar (version 4.7).
To obtain a MetS Z score considering all components, absolute changes ([Δ] subtracting baseline value from follow-up value) were converted to standardized z scores ([individual value Δ – group average Δ] / group standard deviation). A standardized Z score was created summing TG, HDL-c, glucose, MBP, and BF (MetS Z score). To obtain the same meaning of other variables, HDL-c Z score was multiplied by -1 (high values of HDL-c are beneficial to health).
Sport participation
Adolescents were engaged in nine sports (judo [n= 4], karate [n= 14], kung fu [n= 13], gymnastics [n= 10], baseball [n= 10], basketball [n= 16], swimming [n= 24], tennis [n=15] and track & field [n= 8]), while non-sport group was composed of schoolchildren (n= 57). In general, sports with typical physical training involving aerobic fitness (e.g. running activities), usually associated with greater improvements in terms of CRF and intensity above six metabolic equivalents (METs) [24, 31] were classified as high CRF sports (track & field, basketball, swimming and tennis [High CRF Sport; n= 64]). Sports involving less aerobic activities and in which their modality is classified as below six METs [24], those adolescents were classified as low CRF sports (judo, karate, kung fu, gymnastics and baseball [Sport Low-CRF; n= 50]).
Coaches keep contact with researchers during the entire follow-up period and only adolescents who frequently participated in the training sections were reassessed in the follow-up moment. At baseline, it was reported the previous time of engagement in the current sport (in months), the number of days/week training, time spent in the exercise sections (in minutes). Therefore, it also calculated the overall time per week of training (minutes/week).
Additionally, two non-consecutive training sections in the same week were entirely monitored by the researchers. Heart rate (HR) was monitored during the training sections using a Polar V800 HR monitor (a plastic strap was placed on the trunk to measure HR via telemetry [Polar Electro®, Finland]) and the average values of the two sections were adopted [32]. The average percentage of maximum HR reached during the two sections has been multiplied by the number of training per week of each athlete in order to create a proxy of training load.
Covariates
Ethnicity (caucasian or others [black, Asian, Native American and others]), chronological age (difference between birthday and baseline measurements), and sex (boy/girl) were self-reported during a face-to-face interview. Bodyweight (electronic scale [Filizzola PL 150, model Filizzola Ltda, Brazil]) and height (wall-mounted stadiometer [Sanny, model American Medical of the Brazil Ltda, Brazil]) were measured according to standardized procedures [33], while the biological maturation was estimated through the peak height velocity (PHV) proposed by MIRWALD et al., (2002) (time in years before [negative score] and after [positive score] the moment of maximum gain of height).
Statistical analysis
Descriptive data were presented by means and 95% confidence intervals (95%CI). Absolute change (Δ) was calculated using the data of 1-year of follow-up. Analysis of covariance (ANCOVA) was used to compare the components of MetS according to the engagement in sports adjusted by covariates (sex, ethnicity, PHV [baseline], and body fatness [baseline]). Levene’s test assessed the assumption of homogeneity of variances in the models, while Bonferroni’s post hoc test was used when necessary. Effect size was expressed as eta-squared (ES-r) values and classified as follow: (i) from 0.010 to 0.059 [small]; (ii) 0.060 to 0.139 [moderate]; (iii) ≥0.140 [elevated] [35]. The relationship between parameters of sports participation and changes on MetS components was assessed using partial correlation, adjusting by sex, ethnicity, PHV (baseline), and body fatness (baseline). Statistical significance was set at 5% (p < 0.05). Analyzes were performed using the software BioEstat (version 5.0).