This article was downloaded by: [Virginia Commonwealth University Libraries] On: 12 March 2015, At: 07:32 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Journal of Sports Sciences Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/rjsp20
Physical activity and social support in adolescents: analysis of different types and sources of social support ac
Gerfeson Mendonça
& José Cazuza de Farias Júnior
abc
a
Post-Graduation Associated Program in Physical Education – UPE/UFPB, Brazil
b
Federal University of Paraíba – UFPB, Brazil
c
Research and Study Group in Epidemiology of Physical Activity – GEPEAF, Brazil Published online: 09 Mar 2015.
Click for updates To cite this article: Gerfeson Mendonça & José Cazuza de Farias Júnior (2015): Physical activity and social support in adolescents: analysis of different types and sources of social support, Journal of Sports Sciences, DOI: 10.1080/02640414.2015.1020842 To link to this article: http://dx.doi.org/10.1080/02640414.2015.1020842
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Journal of Sports Sciences, 2015 http://dx.doi.org/10.1080/02640414.2015.1020842
Physical activity and social support in adolescents: analysis of different types and sources of social support
GERFESON MENDONÇA1,3 & JOSÉ CAZUZA DE FARIAS JÚNIOR1,2,3 Post-Graduation Associated Program in Physical Education – UPE/UFPB, Brazil, 2Federal University of Paraíba – UFPB, Brazil and 3Research and Study Group in Epidemiology of Physical Activity – GEPEAF, Brazil
Downloaded by [Virginia Commonwealth University Libraries] at 07:32 12 March 2015
1
(Accepted 16 February 2015)
Abstract Little is known about the influence of different types and sources of social support on physical activity in adolescents. The aim of this study was to analyse the association between physical activity and different types and sources of social support in adolescents. The sample consisted of 2,859 adolescents between 14–19 years of age in the city of João Pessoa, in Northeastern Brazil. Physical activity was measured with a questionnaire and social support from parents and friends using a 10-item scale five for each group (type of support: encouragement, joint participation, watching, inviting, positive comments and transportation). Multivariable analysis showed that the types of support provided by parents associated with physical activity in adolescents were encouragement for females (P < 0.001) and adolescents between 14–16 years of age (P = 0.003), and transportation (P = 0.014) and comments (P = 0.037) for males. The types of social support provided by friends were: joint participation in male adolescents (P < 0.001) and in these 17–19-year-olds (P < 0.001), and comments in both genders (males: P = 0.009; females: P < 0.001) and 14–16-year-olds (P < 0.001). We conclude that the type of social support associated with physical activity varies according to its source, as well as the gender and age of the adolescents. Keywords: adolescence, motor activity, associated factors, social influence
Introduction Social support is one of the main constructs of theories (Planned Behaviour and Sociocognitive Theories) and models (Health Belief and Socioecological Models) that have been used to explain physical activity behaviour (Glanz, Rimer, & Viswanath, 2008; Seabra, Mendonça, Thomis, Anjos, & Maia, 2008). It is defined as an intentional act that occurs through different forms (instrumental, psychological and instructional) in order to assist a person in adopting and/or maintaining behaviour that may originate from different social groups, e.g., parents, friends, brothers, uncles and teachers (Beets, Vogel, Forlaw, Pitetti, & Cardinal, 2006; Duncan, Duncan, & Strycker, 2005). Social support for physical activity can be provided in the form of equipment and sporting goods, incentives, positive feedback or comments, joint participation and transportation to practice sites (Beets et al., 2006; Duncan et al., 2005), and is one of the most investigated correlates of physical activity factors (Bauman et al., 2012; Craggs, Corder, van Sluijs, & Griffin, 2011; Gustafson & Rhodes, 2006; van der Horst, Paw, Twisk, & Mechelen, 2007). Cross-
sectional (Gustafson & Rhodes, 2006; Seabra et al., 2008; van der Horst et al., 2007) and longitudinal (Craggs et al., 2011; Dishman, Saunders, Motl, Dowda, & Pate, 2009; Dowda, Dishman, Pfeiffer, & Pate, 2007) studies have identified a positive relationship between the provision of social support and participation in physical activities among adolescents. Social support has also been identified as an important factor for maintaining physical activity levels, as well as a major mediator of the effects of interventions on these levels in adolescents (Bauman et al., 2012; Craggs et al., 2011). However, the majority of studies that evaluated the association between social support and physical activity were conducted in adolescents from high-income countries (Duncan et al., 2005; Lubans, Sylva, & Morgan, 2007), in non-representative samples (Beets, Pitetti, & Forlaw, 2007; Beets et al., 2006; Duncan et al., 2005; Okun et al., 2003) and restricted age groups (Beets et al., 2007; Duncan et al., 2005). In addition, previous studies rarely analysed the association between different types and sources of social support and physical activity
Correspondence: Gerfeson Mendonça, Federal University of Paraíba, Department of Physical Education, Campus I, Cidade Universitária, João Pessoa, PB 58059-900, Brazil. E-mail: [email protected] © 2015 Taylor & Francis
Downloaded by [Virginia Commonwealth University Libraries] at 07:32 12 March 2015
2
G. Mendonça & J. C. Farias Júnior
level and variations by gender (Beets et al., 2006; Duncan et al., 2005; Ricardo, Rombaldi, Otte, Perez, & Azevedo, 2013) and age of the adolescents. Despite scarce evidence, it has been observed that the types of social support associated with physical activity can vary according to the characteristics of adolescents (age and gender) and the sources of social support, particularly from parents and friends (Mendonça, Cheng, Melo, & Farias Júnior, 2014), because these social groups interact most with teenagers (Seabra et al., 2008). Therefore, the differences in types of social support associated with physical activity between genders and adolescents’ age need to be further investigated (Mendonça et al., 2014), since the social influence of parents and friends (Seabra et al., 2008) and patterns of physical activity in teens (Azevedo, Araújo, & Pereira, 2006) can change during adolescence. It is also important to note that teenagers are often involved in different types of physical activities; thus, it is assumed they must perceive the simultaneous provision of various types of support in order to increase the likelihood of being physically active (Beets et al., 2006; Mendonça et al., 2014). Studies analysing the associations between physical activity and the different types of social support, and their variation according to support source, adolescents’ gender and age, may help to determine whether some types of social support are more important than others in influencing the choices of adolescents in relation to engaging in physical activity, and if there are variations in these associations according to gender, age and source of social support for adolescents. This evidence may help establish which types of social support should be emphasised in physical activity intervention programmes, taking into account the traits of adolescents and the source providing social support. This study examined the association between different types (encouragement, joint participation, watching, inviting, positive comments and transportation) and sources (parents and friends) of social support and physical activity in adolescents. Methods This cross-sectional epidemiological school-based study, part of the research project entitled Physical activity and associated factors among high school students in the city of João Pessoa, PB: an ecological approach, was conducted in 2009 with adolescents aged 14–19 years, from public and private schools in the city of João Pessoa, Northeastern Brazil. The study was approved by the Human Research Ethics Committee of the Federal University of Paraiba (protocol n. 0062/09). All adolescents under 18 years of age were authorised by their parents or
legal guardians to take part in the study and the others gave their informed consent. Sample The sample size was determined by considering the multiple objectives of the main study and according to the following parameters: estimated prevalence of 50% (300 min or more of physical activity per week), which resulted in the highest estimated sample size for the same standard error, 95% confidence interval, maximum tolerable error of three percentage points, design effect of two, 30% increase in the minimum sample size to compensate for possible losses and refusals. Thus, the sample size was estimated at 2,686 adolescents. The sample was selected using two-stage cluster sampling. In the first stage, 30 schools were systematically selected and distributed proportionally by type (public and private), size (number of students) and geographic region (north, south, east and west). In the second, 135 classes were randomly selected and distributed proportionately by school shift (day [morning + afternoon] and evening) and high school level (1st, 2nd and 3rd grades). In both stages, distributions were proportional to those of the target population. The number of classes to be drawn in each school was determined by dividing the number of students to be recruited from each grade by 20 (the mean number of students expected in each class). Data collection, which occurred between May and September 2009, was carried out by a previously trained team composed of six students from the Physical Education course, supervised by the researcher in charge. A pilot study was conducted under the same conditions as the main study. The questionnaire was completed by the adolescents in the classroom during the regular class schedule in accordance with instructions provided by one of the researchers. Measures Sociodemographic variables such as gender, age, father’s and mother’s education, economic class and adolescents’ nutritional status were used to characterise the sample and as potential confounding factors. Age was determined by the difference between date of birth and date of data collection, and further categorised into two groups: “14–16” and “17–19 years of age”. Parents’ schooling was measured and categorised as follows: “elementary school graduate”, “incomplete elementary school”, “high school graduate” and “college graduate”. To determine economic class the methodology used was that suggested by the Brazilian Association of
Downloaded by [Virginia Commonwealth University Libraries] at 07:32 12 March 2015
Physical activity and social support Research Companies – ABEP (2009), which takes into account the presence of material goods and the number of salaried employees in the residence, as well as the head of the household’s schooling level. Based on the scores produced from the sum score of these items, adolescents were grouped into economic classes A1, A2, B1, B2, C1, C2, D and E, and regrouped as follows: A (best condition), B, C and D/E (worst condition). Nutritional status was assessed by body mass index (BMI = body weight [kg]/height [m]2), based on self-reported measures. Nutritional classification followed the criteria suggested by Cole, Bellizzi, Flegal, and Dietz (2000), with adolescents divided into “Not overweight” (underweight and normal weight) and “Overweight” (overweight and obesity). Physical activity was measured using pretested selfreport methods – Physical Activity Questionnaire for Adolescents – PAQ-A (reliability – Intraclass Correlation Coefficient – ICC = 0.88, 95%CI: 0.84– 0.91; validity: questionnaire measures were compared with the measures of four 24-h recalls – Spearman’s rho = 0.62, P < 0.001, kappa = 0.59) – (Farias Júnior et al., 2012). Adolescents reported the frequency (days/week) and duration (min/day) of physical activities that lasted at least 10 min in the week before data collection, considering a list of 24 moderate- and vigorous-intensity activities that include active commuting, sports within and outside the school setting, open-air activities (recreational) and physical exercises. The level of physical activity was determined from the sum of the product of time spent on activities and their frequency, resulting in a score in minutes per week. Adolescents who reported physical activity equal to or greater than 300 min per week were classified as “physically active” and others as “physically inactive”. Social support from parents and friends for physical activity was assessed by a 10-item scale, five for each group. This scale was previously tested and showed satisfactory levels of internal consistency (parents α = 0.81; friends α = 0.90) and test–retest reliability (parents – ICC = 0.91; friends – ICC = 0.89) – (Farias Júnior et al., 2011). Adolescents reported the frequency (never, seldom, often and always) with which their parents and friends provided some form of social support (encouragement, joint participation, watching, inviting, positive comments and transportation) for physical activity during a typical week. Social support scores of parents and friends were obtained from the sum of the values assigned to each response to the items of each scale. Consequently, tertiles of these scores were determined by adolescents’ gender and age. The first tertile concentrated adolescents with the lowest social support and the third those with the highest.
3
For analysis of the different types (encouragement, joint participation, watching, inviting, comments and transportation) of social support for each source (parents and friends), responses were grouped into “without social support” (never and seldom) and “with social support” (often and always). Analyses were stratified by adolescents’ gender and age. Analysis of simultaneous exposure to different types of social support was also conducted. Therefore, the sum of the responses was determined from their source, considering the classification of responses (“without social support” and “with social support”). Categories were then generated according to the number of social support types provided: “unsupported”, “one type of support”; “two types of support”; “three or more types”, separated by source (parents and friends). Statistical analysis Descriptive statistics included the calculation of ratios and 95% confidence intervals (95%CI) for categorical variables, and means and standard deviations for numerical variables. Chi-square tests for heterogeneity and linear trend (χ2) with its degrees of freedom (df) were used to compare the proportion of students classified as physically active according to social support score tertiles. A binary logistic regression was used to determine possible associations between physical activity and the different types and sources of social support. To that end, three models of analysis were developed, all of which had physical activity as dependent variable (physically active = 1 and physically inactive = 0) and social support as independent variable. The variables were analysed according to the following models: model 1 – tertiles of social support source scores (parents and friends); model 2 – different types of social support (encouragement, joint participation, watching, inviting, comments and transportation) provided by parents and friends (no social support and social support); model 3 – number of different types of social support provided simultaneously by parents and friends (0, 1, 2, 3 or more forms of social support). In multivariable analysis, all variables were entered into the model at the same level and maintained independently from the P-value (significance level). Hosmer–Lemeshow tests were conducted to assess the quality of the model fit, comparing the real value of the model with the predicted value (non-significant value indicates good model quality), deviance statistics (D) and Bayesian information criterion (BIC), which aims to compare concurrent models, where the best model is assumed to be that which represents the lowest values of these statistics. Interaction analysis was performed between sex and
4
G. Mendonça & J. C. Farias Júnior
age and scores of the different types of social support, stratified by source (parents and friends). All analyses were adjusted for parents’ schooling, economic class and adolescents’ nutritional status. These were carried out using Stata 11.0, considering the sample selection process (cluster sampling). The significance level was 5% for the two-tailed tests.
Downloaded by [Virginia Commonwealth University Libraries] at 07:32 12 March 2015
Results The final sample consisted of 2,859 adolescents (losses and refusals amounted to 17.8% of the cases), with a mean age of 16.5 years (s = 1.2). A posteriori calculations demonstrated that this sample had 80% power to identify odds ratios greater than or equal to 1.40 as being significant. Descriptive analysis found that most of the students were females (57.8%), aged 14–16 years (67.9%) and belonged to medium-to-low economic classes (C, D and E: 54.2%). The prevalence of adolescents classified as physically active was 50.3% (95%CI: 47.3 to 53.1), being higher in males (66.3%) than in females (38.5%, χ2 (1) = 216.4818; P < 0.001) – Table I. Chi-square results demonstrated a trend towards a linear increase in the proportion of adolescents classified as physically active, with an increase in the social support score tertiles of
parents (χ2 (2) = 159.7805; P < 0.001) and friends (χ2 (2) = 280.9142; P < 0.001) – data not shown in the tables. In adjusted logistic regression analysis for model 1, the social support score tertiles of parents and friends were positively and significantly associated with physical activity in adolescents (Table II). The magnitude of the association was significantly higher (without the intervention of 95%CI values) for the social support score of friends (OR = 4.92, 95%CI: 3.33 to 7.28) compared to that of parents (OR = 2.00, 95%CI: 1.37 to 2.90), but only in males (data not shown in the tables). Analysis of the interactions between social support score tertiles showed no significant interaction between gender and parents’ social support (P = 0.952) nor between age and social support in both groups (parents: P = 0.987 and friends: P = 0.778). However, there was significant interaction between gender and social support from friends (P = 0.022), demonstrating that the association between physical activity and social support from friends was twice as high in males compared to females (OR = 5.00, 95%CI: 3.39 to 7.40 vs. OR = 2.36, 95% CI: 1.68 to 3.30) – data not shown in the tables. In unadjusted logistic regression analysis (model 2), all types of social support were associated with physical activity in adolescents of both genders and
Table I. Characteristics of the sample of adolescents in the city of João Pessoa, Brazil, 2009. Male Variable Gender Male Female Age 14–16 years 17–19 years Economic class A (best condition) B C D/E (worst condition) Father’s schooling Incomplete elementary school Elementary school graduate High school graduate College graduate Mother’s schooling Incomplete elementary school Elementary school graduate High school graduate College graduate Nutritional status Not overweight Overweight Physical activity Physically inactive Physically active
n
Female %
n
All %
n
%
1206 1653
42.2 57.8
795 411
65.9 34.1
1146 507
69.3 30.7
1941 918
67.9 32.1
119 416 446 76
11.3 39.4 42.1 7.2
117 506 717 127
8,0 34.4 48.9 8.7
237 924 1167 205
9.3 36.5 46.1 8.1
309 233 402 229
26.3 19.9 34.3 19.5
563 279 491 266
35.2 17.5 30.7 16.6
877 514 895 497
31.5 18.5 32.2 17.8
309 221 399 266
25.8 18.5 33.4 22.3
538 315 502 283
32.9 19.2 30.6 17.3
853 538 906 550
30.0 18.9 31.8 19.3
927 207
81.7 18.3
1394 146
90.5 9.5
2321 353
86.8 13.2
406 800
33.7 66.3
1017 636
61.5 38.5
1423 1436
49.7 50.3
5
Physical activity and social support
Table II. Binary logistic regression analysis for the association between physical activity and scores of social support provided by parents and friends, Joao Pessoa, Brazil, 2009. Unadjusted analysis Sources of social support
Downloaded by [Virginia Commonwealth University Libraries] at 07:32 12 March 2015
Parents 1st Tertile 2nd Tertile 3rd Tertile Friends 1st Tertile 2nd Tertile 3rd Tertile
%
OR
37.6 53.3 65.6
1 1.90 3.16
32.2 56.2 70.7
1 2.71 5.09
95%CI
Adjusted analysis* P†
OR
95%CI
Journal of Sports Sciences Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/rjsp20
Physical activity and social support in adolescents: analysis of different types and sources of social support ac
Gerfeson Mendonça
& José Cazuza de Farias Júnior
abc
a
Post-Graduation Associated Program in Physical Education – UPE/UFPB, Brazil
b
Federal University of Paraíba – UFPB, Brazil
c
Research and Study Group in Epidemiology of Physical Activity – GEPEAF, Brazil Published online: 09 Mar 2015.
Click for updates To cite this article: Gerfeson Mendonça & José Cazuza de Farias Júnior (2015): Physical activity and social support in adolescents: analysis of different types and sources of social support, Journal of Sports Sciences, DOI: 10.1080/02640414.2015.1020842 To link to this article: http://dx.doi.org/10.1080/02640414.2015.1020842
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Journal of Sports Sciences, 2015 http://dx.doi.org/10.1080/02640414.2015.1020842
Physical activity and social support in adolescents: analysis of different types and sources of social support
GERFESON MENDONÇA1,3 & JOSÉ CAZUZA DE FARIAS JÚNIOR1,2,3 Post-Graduation Associated Program in Physical Education – UPE/UFPB, Brazil, 2Federal University of Paraíba – UFPB, Brazil and 3Research and Study Group in Epidemiology of Physical Activity – GEPEAF, Brazil
Downloaded by [Virginia Commonwealth University Libraries] at 07:32 12 March 2015
1
(Accepted 16 February 2015)
Abstract Little is known about the influence of different types and sources of social support on physical activity in adolescents. The aim of this study was to analyse the association between physical activity and different types and sources of social support in adolescents. The sample consisted of 2,859 adolescents between 14–19 years of age in the city of João Pessoa, in Northeastern Brazil. Physical activity was measured with a questionnaire and social support from parents and friends using a 10-item scale five for each group (type of support: encouragement, joint participation, watching, inviting, positive comments and transportation). Multivariable analysis showed that the types of support provided by parents associated with physical activity in adolescents were encouragement for females (P < 0.001) and adolescents between 14–16 years of age (P = 0.003), and transportation (P = 0.014) and comments (P = 0.037) for males. The types of social support provided by friends were: joint participation in male adolescents (P < 0.001) and in these 17–19-year-olds (P < 0.001), and comments in both genders (males: P = 0.009; females: P < 0.001) and 14–16-year-olds (P < 0.001). We conclude that the type of social support associated with physical activity varies according to its source, as well as the gender and age of the adolescents. Keywords: adolescence, motor activity, associated factors, social influence
Introduction Social support is one of the main constructs of theories (Planned Behaviour and Sociocognitive Theories) and models (Health Belief and Socioecological Models) that have been used to explain physical activity behaviour (Glanz, Rimer, & Viswanath, 2008; Seabra, Mendonça, Thomis, Anjos, & Maia, 2008). It is defined as an intentional act that occurs through different forms (instrumental, psychological and instructional) in order to assist a person in adopting and/or maintaining behaviour that may originate from different social groups, e.g., parents, friends, brothers, uncles and teachers (Beets, Vogel, Forlaw, Pitetti, & Cardinal, 2006; Duncan, Duncan, & Strycker, 2005). Social support for physical activity can be provided in the form of equipment and sporting goods, incentives, positive feedback or comments, joint participation and transportation to practice sites (Beets et al., 2006; Duncan et al., 2005), and is one of the most investigated correlates of physical activity factors (Bauman et al., 2012; Craggs, Corder, van Sluijs, & Griffin, 2011; Gustafson & Rhodes, 2006; van der Horst, Paw, Twisk, & Mechelen, 2007). Cross-
sectional (Gustafson & Rhodes, 2006; Seabra et al., 2008; van der Horst et al., 2007) and longitudinal (Craggs et al., 2011; Dishman, Saunders, Motl, Dowda, & Pate, 2009; Dowda, Dishman, Pfeiffer, & Pate, 2007) studies have identified a positive relationship between the provision of social support and participation in physical activities among adolescents. Social support has also been identified as an important factor for maintaining physical activity levels, as well as a major mediator of the effects of interventions on these levels in adolescents (Bauman et al., 2012; Craggs et al., 2011). However, the majority of studies that evaluated the association between social support and physical activity were conducted in adolescents from high-income countries (Duncan et al., 2005; Lubans, Sylva, & Morgan, 2007), in non-representative samples (Beets, Pitetti, & Forlaw, 2007; Beets et al., 2006; Duncan et al., 2005; Okun et al., 2003) and restricted age groups (Beets et al., 2007; Duncan et al., 2005). In addition, previous studies rarely analysed the association between different types and sources of social support and physical activity
Correspondence: Gerfeson Mendonça, Federal University of Paraíba, Department of Physical Education, Campus I, Cidade Universitária, João Pessoa, PB 58059-900, Brazil. E-mail: [email protected] © 2015 Taylor & Francis
Downloaded by [Virginia Commonwealth University Libraries] at 07:32 12 March 2015
2
G. Mendonça & J. C. Farias Júnior
level and variations by gender (Beets et al., 2006; Duncan et al., 2005; Ricardo, Rombaldi, Otte, Perez, & Azevedo, 2013) and age of the adolescents. Despite scarce evidence, it has been observed that the types of social support associated with physical activity can vary according to the characteristics of adolescents (age and gender) and the sources of social support, particularly from parents and friends (Mendonça, Cheng, Melo, & Farias Júnior, 2014), because these social groups interact most with teenagers (Seabra et al., 2008). Therefore, the differences in types of social support associated with physical activity between genders and adolescents’ age need to be further investigated (Mendonça et al., 2014), since the social influence of parents and friends (Seabra et al., 2008) and patterns of physical activity in teens (Azevedo, Araújo, & Pereira, 2006) can change during adolescence. It is also important to note that teenagers are often involved in different types of physical activities; thus, it is assumed they must perceive the simultaneous provision of various types of support in order to increase the likelihood of being physically active (Beets et al., 2006; Mendonça et al., 2014). Studies analysing the associations between physical activity and the different types of social support, and their variation according to support source, adolescents’ gender and age, may help to determine whether some types of social support are more important than others in influencing the choices of adolescents in relation to engaging in physical activity, and if there are variations in these associations according to gender, age and source of social support for adolescents. This evidence may help establish which types of social support should be emphasised in physical activity intervention programmes, taking into account the traits of adolescents and the source providing social support. This study examined the association between different types (encouragement, joint participation, watching, inviting, positive comments and transportation) and sources (parents and friends) of social support and physical activity in adolescents. Methods This cross-sectional epidemiological school-based study, part of the research project entitled Physical activity and associated factors among high school students in the city of João Pessoa, PB: an ecological approach, was conducted in 2009 with adolescents aged 14–19 years, from public and private schools in the city of João Pessoa, Northeastern Brazil. The study was approved by the Human Research Ethics Committee of the Federal University of Paraiba (protocol n. 0062/09). All adolescents under 18 years of age were authorised by their parents or
legal guardians to take part in the study and the others gave their informed consent. Sample The sample size was determined by considering the multiple objectives of the main study and according to the following parameters: estimated prevalence of 50% (300 min or more of physical activity per week), which resulted in the highest estimated sample size for the same standard error, 95% confidence interval, maximum tolerable error of three percentage points, design effect of two, 30% increase in the minimum sample size to compensate for possible losses and refusals. Thus, the sample size was estimated at 2,686 adolescents. The sample was selected using two-stage cluster sampling. In the first stage, 30 schools were systematically selected and distributed proportionally by type (public and private), size (number of students) and geographic region (north, south, east and west). In the second, 135 classes were randomly selected and distributed proportionately by school shift (day [morning + afternoon] and evening) and high school level (1st, 2nd and 3rd grades). In both stages, distributions were proportional to those of the target population. The number of classes to be drawn in each school was determined by dividing the number of students to be recruited from each grade by 20 (the mean number of students expected in each class). Data collection, which occurred between May and September 2009, was carried out by a previously trained team composed of six students from the Physical Education course, supervised by the researcher in charge. A pilot study was conducted under the same conditions as the main study. The questionnaire was completed by the adolescents in the classroom during the regular class schedule in accordance with instructions provided by one of the researchers. Measures Sociodemographic variables such as gender, age, father’s and mother’s education, economic class and adolescents’ nutritional status were used to characterise the sample and as potential confounding factors. Age was determined by the difference between date of birth and date of data collection, and further categorised into two groups: “14–16” and “17–19 years of age”. Parents’ schooling was measured and categorised as follows: “elementary school graduate”, “incomplete elementary school”, “high school graduate” and “college graduate”. To determine economic class the methodology used was that suggested by the Brazilian Association of
Downloaded by [Virginia Commonwealth University Libraries] at 07:32 12 March 2015
Physical activity and social support Research Companies – ABEP (2009), which takes into account the presence of material goods and the number of salaried employees in the residence, as well as the head of the household’s schooling level. Based on the scores produced from the sum score of these items, adolescents were grouped into economic classes A1, A2, B1, B2, C1, C2, D and E, and regrouped as follows: A (best condition), B, C and D/E (worst condition). Nutritional status was assessed by body mass index (BMI = body weight [kg]/height [m]2), based on self-reported measures. Nutritional classification followed the criteria suggested by Cole, Bellizzi, Flegal, and Dietz (2000), with adolescents divided into “Not overweight” (underweight and normal weight) and “Overweight” (overweight and obesity). Physical activity was measured using pretested selfreport methods – Physical Activity Questionnaire for Adolescents – PAQ-A (reliability – Intraclass Correlation Coefficient – ICC = 0.88, 95%CI: 0.84– 0.91; validity: questionnaire measures were compared with the measures of four 24-h recalls – Spearman’s rho = 0.62, P < 0.001, kappa = 0.59) – (Farias Júnior et al., 2012). Adolescents reported the frequency (days/week) and duration (min/day) of physical activities that lasted at least 10 min in the week before data collection, considering a list of 24 moderate- and vigorous-intensity activities that include active commuting, sports within and outside the school setting, open-air activities (recreational) and physical exercises. The level of physical activity was determined from the sum of the product of time spent on activities and their frequency, resulting in a score in minutes per week. Adolescents who reported physical activity equal to or greater than 300 min per week were classified as “physically active” and others as “physically inactive”. Social support from parents and friends for physical activity was assessed by a 10-item scale, five for each group. This scale was previously tested and showed satisfactory levels of internal consistency (parents α = 0.81; friends α = 0.90) and test–retest reliability (parents – ICC = 0.91; friends – ICC = 0.89) – (Farias Júnior et al., 2011). Adolescents reported the frequency (never, seldom, often and always) with which their parents and friends provided some form of social support (encouragement, joint participation, watching, inviting, positive comments and transportation) for physical activity during a typical week. Social support scores of parents and friends were obtained from the sum of the values assigned to each response to the items of each scale. Consequently, tertiles of these scores were determined by adolescents’ gender and age. The first tertile concentrated adolescents with the lowest social support and the third those with the highest.
3
For analysis of the different types (encouragement, joint participation, watching, inviting, comments and transportation) of social support for each source (parents and friends), responses were grouped into “without social support” (never and seldom) and “with social support” (often and always). Analyses were stratified by adolescents’ gender and age. Analysis of simultaneous exposure to different types of social support was also conducted. Therefore, the sum of the responses was determined from their source, considering the classification of responses (“without social support” and “with social support”). Categories were then generated according to the number of social support types provided: “unsupported”, “one type of support”; “two types of support”; “three or more types”, separated by source (parents and friends). Statistical analysis Descriptive statistics included the calculation of ratios and 95% confidence intervals (95%CI) for categorical variables, and means and standard deviations for numerical variables. Chi-square tests for heterogeneity and linear trend (χ2) with its degrees of freedom (df) were used to compare the proportion of students classified as physically active according to social support score tertiles. A binary logistic regression was used to determine possible associations between physical activity and the different types and sources of social support. To that end, three models of analysis were developed, all of which had physical activity as dependent variable (physically active = 1 and physically inactive = 0) and social support as independent variable. The variables were analysed according to the following models: model 1 – tertiles of social support source scores (parents and friends); model 2 – different types of social support (encouragement, joint participation, watching, inviting, comments and transportation) provided by parents and friends (no social support and social support); model 3 – number of different types of social support provided simultaneously by parents and friends (0, 1, 2, 3 or more forms of social support). In multivariable analysis, all variables were entered into the model at the same level and maintained independently from the P-value (significance level). Hosmer–Lemeshow tests were conducted to assess the quality of the model fit, comparing the real value of the model with the predicted value (non-significant value indicates good model quality), deviance statistics (D) and Bayesian information criterion (BIC), which aims to compare concurrent models, where the best model is assumed to be that which represents the lowest values of these statistics. Interaction analysis was performed between sex and
4
G. Mendonça & J. C. Farias Júnior
age and scores of the different types of social support, stratified by source (parents and friends). All analyses were adjusted for parents’ schooling, economic class and adolescents’ nutritional status. These were carried out using Stata 11.0, considering the sample selection process (cluster sampling). The significance level was 5% for the two-tailed tests.
Downloaded by [Virginia Commonwealth University Libraries] at 07:32 12 March 2015
Results The final sample consisted of 2,859 adolescents (losses and refusals amounted to 17.8% of the cases), with a mean age of 16.5 years (s = 1.2). A posteriori calculations demonstrated that this sample had 80% power to identify odds ratios greater than or equal to 1.40 as being significant. Descriptive analysis found that most of the students were females (57.8%), aged 14–16 years (67.9%) and belonged to medium-to-low economic classes (C, D and E: 54.2%). The prevalence of adolescents classified as physically active was 50.3% (95%CI: 47.3 to 53.1), being higher in males (66.3%) than in females (38.5%, χ2 (1) = 216.4818; P < 0.001) – Table I. Chi-square results demonstrated a trend towards a linear increase in the proportion of adolescents classified as physically active, with an increase in the social support score tertiles of
parents (χ2 (2) = 159.7805; P < 0.001) and friends (χ2 (2) = 280.9142; P < 0.001) – data not shown in the tables. In adjusted logistic regression analysis for model 1, the social support score tertiles of parents and friends were positively and significantly associated with physical activity in adolescents (Table II). The magnitude of the association was significantly higher (without the intervention of 95%CI values) for the social support score of friends (OR = 4.92, 95%CI: 3.33 to 7.28) compared to that of parents (OR = 2.00, 95%CI: 1.37 to 2.90), but only in males (data not shown in the tables). Analysis of the interactions between social support score tertiles showed no significant interaction between gender and parents’ social support (P = 0.952) nor between age and social support in both groups (parents: P = 0.987 and friends: P = 0.778). However, there was significant interaction between gender and social support from friends (P = 0.022), demonstrating that the association between physical activity and social support from friends was twice as high in males compared to females (OR = 5.00, 95%CI: 3.39 to 7.40 vs. OR = 2.36, 95% CI: 1.68 to 3.30) – data not shown in the tables. In unadjusted logistic regression analysis (model 2), all types of social support were associated with physical activity in adolescents of both genders and
Table I. Characteristics of the sample of adolescents in the city of João Pessoa, Brazil, 2009. Male Variable Gender Male Female Age 14–16 years 17–19 years Economic class A (best condition) B C D/E (worst condition) Father’s schooling Incomplete elementary school Elementary school graduate High school graduate College graduate Mother’s schooling Incomplete elementary school Elementary school graduate High school graduate College graduate Nutritional status Not overweight Overweight Physical activity Physically inactive Physically active
n
Female %
n
All %
n
%
1206 1653
42.2 57.8
795 411
65.9 34.1
1146 507
69.3 30.7
1941 918
67.9 32.1
119 416 446 76
11.3 39.4 42.1 7.2
117 506 717 127
8,0 34.4 48.9 8.7
237 924 1167 205
9.3 36.5 46.1 8.1
309 233 402 229
26.3 19.9 34.3 19.5
563 279 491 266
35.2 17.5 30.7 16.6
877 514 895 497
31.5 18.5 32.2 17.8
309 221 399 266
25.8 18.5 33.4 22.3
538 315 502 283
32.9 19.2 30.6 17.3
853 538 906 550
30.0 18.9 31.8 19.3
927 207
81.7 18.3
1394 146
90.5 9.5
2321 353
86.8 13.2
406 800
33.7 66.3
1017 636
61.5 38.5
1423 1436
49.7 50.3
5
Physical activity and social support
Table II. Binary logistic regression analysis for the association between physical activity and scores of social support provided by parents and friends, Joao Pessoa, Brazil, 2009. Unadjusted analysis Sources of social support
Downloaded by [Virginia Commonwealth University Libraries] at 07:32 12 March 2015
Parents 1st Tertile 2nd Tertile 3rd Tertile Friends 1st Tertile 2nd Tertile 3rd Tertile
%
OR
37.6 53.3 65.6
1 1.90 3.16
32.2 56.2 70.7
1 2.71 5.09
95%CI
Adjusted analysis* P†
OR
95%CI
