The study followed the standards of research involving human beings of the National Health Council (resolution number 466/2012), having been approved by the Research Ethics Committee of the Federal University of Paraná (CAAE: 98133218.8.0000.0102) and authorized by adolescents’ the parents or legal guardians by means of a free and informed consent form.
This is a cross-sectional study conducted in 2018 with a representative sample of adolescents aged 15 to 17 enrolled in high school in the public school.
For the preliminary sample calculation, an association of OR = 1.49 was considered between SS and PA24, a prevalence of being insufficiently active of 50%25, a confidence level of 95% (α = 0.05) with a power of 80% (β = 0.20), resulting in a minimum sample of 804 subjects. However, an increase of 30% was considered for possible losses and refusals, which estimated a minimum sample required for the study of 965 subjects. The sample was selected from the process of sampling by multiple stages, conglomerates, in three stages: 1st stage - all state schools were stratified according to each of the nine administrative regions of the city of Curitiba; 2nd stage - a draw was made among the nine administrative regions of the city, being two of them the one chosen; 3rd stage - a simple random selection of a class of each year was performed, according to the number of students, separated by sex, required for a given region. Finally, all students in each class were invited to participate in the study.
A total of 2506 adolescents were evaluated, but those who presented physical and/or cognitive limitations that could restrict the practice of PA (informed by the school) (n = 12) and were 18 years old (n = 125) were excluded. Adolescents who did not present the informed consent form signed by their parents or legal guardians (n = 104), those who refused to participate in the study, or missed the day of collection (n = 56), as well as those who answered the questionnaires incorrectly (n = 66) and those who did not answer the PA level (n = 753) when the questionnaire was sent home, were considered as sample loss. Thus, the analytical sample of the study was 1390 adolescents.
Statistical analyses carried out subsequently showed that this sample could identify statistically significant ratios above 1.36 and below 0.73, considering a power of 80% (β = 20%), a confidence level of 95% (α = 5%) and considering a prevalence of 34% of adolescents with low SS and who did not meet the minimum recommendations for PA.
Physical Activity - Parents / Legal Guardians and Adolescents
The MVPA level of parents or legal guardians was measured through the IPAQ questionnaire - short version (International Physical Activity Questionnaire), in which parents or legal guardians reported “how many days a week” and “how long a day” they practiced MVPA, over a normal week26. The time dedicated to each PA practice (leisure and displacement) per week was categorized according to the current recommendations of MVPA practice26.
The PA level of adolescents was evaluated by the Physical Activity Questionnaire for Adolescents (QAFA)27. This questionnaire was first developed in the checklist format by Sallis et al, 199628 for North American adolescents, translated and adapted for Brazilian adolescents by Farias-Júnior et al, 2012.The questionnaire consists of a list of 24 moderate to vigorous physical activities, with the possibility of adolescents adding activities beyond those listed. In Brazilian adolescents, the questionnaire showed good reproducibility (CCI = 0.88; CI95% = 0.84–0.91) and concurrent validity when compared to 24-hour recall (r = 0.62; p<0.001)27. To analyse the MVPA level, it was classified as “it meets the recommendations” (≥420 minutes per week of MVPA practice) and “it does not meet” (<420 minutes per week of MVPA practice).
SS received from parents or legal guardians for the practice of PA
The SS of the adolescents’ parents or legal guardians to practise PA was measured by the scale of 5 items Social Support for Physical Activity Practice for Adolescents (ASAFA), which has satisfactory consistency (SS of parents α ≥ 0.77 and IFC ≥ 0.83) 29. Adolescents reported the frequency (never = 1, rarely = 2, frequently = 3, always = 4) with which parents offered some type of SS for the practice of PA (encourage, practice, transport, assist, comment) during a typical or normal week29. For purposes of analysis, and with the aim of ensuring comparability with similar studies30, the response options “rarely” and “frequently” were grouped and classified as “sometimes”. Thus, the weekly frequency categories of SS were classified as never, sometimes and always. For the creation of the median variable, we considered the sum of the values assigned to each item (5 to 20 points) and computed an overall score of parents’ SS, categorized by the median: “low SS” and “high SS”.
Sociodemographic Information and Health
Gender was self-reported (“male”, “female”) and age was calculated from the date of birth (informed by the adolescent) subtracted from the date of data collection and divided by 365, resulting in: “15 years”, “16 years”, “17 years”. The nutritional status was calculated with the measured data on body mass and size (kg/m2) and adolescents were classified as “not excess of weight” and “excess of weight”31. The education level of the main family breadwinner and his/her socioeconomic status (SES) were evaluated with a standardized questionnaire32. The Education level was classified as: Elementary School, Secondary School or higher Education levels. SES was classified into three categories: “low” (class C+D), “medium” (class B1+B2) and “high” (class A1+A2) (ABEP, 2016)32.
Data Analysis
After having identified the data distribution by the Kolmogorov-Smirnov test, descriptive statistics were used to distribute absolute and relative frequencies and the chi-square test (χ2) was used to verify differences in proportions in relation to sex and recommended levels of MVPA in adolescents. Binary logistic regression was used to test, in an exploratory manner, the relationship between the study variables. Three blocks of analysis were established to this aim. Firstly, the sociodemographic variables and health were included: gender, age, BMI, the main breadwinner’s education level and SES. In the second block it was inserted the SS received from parents/legal guardians for the practice of PA, while in the third block the information on MVPA of parents/legal guardians was included. The variables that presented p<0.20 in this exploratory analysis were inserted and tested again in the model. The results were tested according to gender (male and female), using the command “divided file”. To avoid bias related to the complex sample selection process (cluster stratification), the association analyses were corrected by the complex design using the SPSS software version 23.0 “complex sample” command.