Which physical and social characteristics of Brazilian cities contribute to physical activity of adolescents?

Quais características físicas e sociais das cidades brasileiras contribuem para atividade física de adolescentes?

Abstract

This study aimed to analyze the association between adolescents’ physical activity and the Brazilian capitals’ built and social environment. The units of analysis of this ecological study were the 26 capitals and the Federal District, with data from the National Adolescent Health Survey (2012). The outcome variable was the reported regular physical activity (PA) of ninth graders in Brazilian schools. Exposure variables included characteristics of the natural environment, socioeconomic and educational indicators, urban infrastructure, urban violence, and sociocultural factors retrieved from several secondary sources of Brazilian databases. We adopted multiple linear regression to verify the association between PA and exposure variables. The percentage of active adolescents was 33.0% (95%CI: 32.1; 33.9). In the final model, higher PA was associated with lower temperature, higher Primary Education Development Index, the higher percentage of ramps for wheelchair users, and a higher percentage of leisure-time active adults. The data show that climatic and educational factors, the infrastructure, and the social environment of the capitals can contribute to Brazilian adolescents complying with the recommended weekly PA levels.

Key words:
Physical activity; Adolescent; Infrastructure; Social environment; Temperature

Resumo

Este estudo teve como objetivo analisar a associação entre atividade física em adolescentes e ambientes construído e social das capitais brasileiras. Trata-se de um estudo ecológico cujas unidades de análise foram as 26 capitais e o Distrito Federal, com dados da Pesquisa Nacional de Saúde do Adolescentes (2012). A variável de desfecho foi a prática regular de atividade física (AF) relatada pelos escolares do 9º ano de escolas brasileiras. As variáveis de exposição incluíram características do ambiente natural, indicadores socioeconômicos e educacionais, de infraestrutura, violência urbana e fatores socioculturais, obtidas de diferentes bases de dados de fontes secundárias brasileiras. Adotou-se a regressão linear múltipla para verificar a associação entre AF e variáveis de exposição. O percentual de adolescentes ativos foi de 33,0% (IC95%: 32,1; 33,9). No modelo final, foram associados à maior prática de AF: menor temperatura, maior Índice de Desenvolvimento da Educação Básica, maior percentual de rampa para cadeirante e maior percentual de adultos ativos no lazer. Os dados mostram que fatores climáticos e educacionais, a infraestrutura e o ambiente social das capitais podem contribuir para que os adolescentes brasileiros cumpram os níveis semanais recomendados de AF.

Palavras-chave:
Atividade física; Adolescente; Infraestrutura; Ambiente social; Temperatura

Introduction

Living in cities is a significant demographic shift in recent decades for most of the global population. Estimates reveal that more than half of the world’s population lives in urban areas and could reach 70% by 2050. In Latin America, more than 80% of the region’s population resides in cities¹.

The growing urbanization process includes a dense and complex intersection of natural, built, and social environments in the cities². In this context, the physical and social environment have been increasingly investigated as determinants of individuals’ habits and behaviors, including physical activity^3,4.

Ecological models propose that the planning and implementing more effective actions to face physical inactivity depend on understanding its determinants at several influencing levels^5,6. This approach involves exploring the urban environment to understand its role in access, conditions, and safety to make physical activity part of everyday life for city dwellers⁷. It includes land use, security, transport, economy, housing, infrastructure and urban design, and availability of leisure facilities^5,7.

Research on this topic, especially adolescents, has increased in recent years, primarily in high-income countries.⁵ Studies have identified that increased levels of physical activity among adolescents were influenced by climate^8,9, air pollution^9,10, rainfall⁸, socioeconomic level^9,11, traffic violence and crime^12,13, infrastructure and urban design^14-16, afforestation, access to places for regular physical activity^12,17, education¹⁸ and sociocultural support^19,20.

However, it is observed from systematic reviews that evidence is still needed to help understand the relationship between physical activity in adolescents and the natural, built, and social environment, especially in low- and middle-income countries^{5,17,18,21,22}. Precisely in Brazil, the scientific production that addresses the topic consists of studies with local samples, the vast majority evaluating the physical activity associated with the perceived environment, with results that are still not consistent^12,15,23-26.

Thus, this study aimed to analyze the association between adolescents’ physical activity and the Brazilian capitals’ built and social environment. We hypothesized that mild temperatures, flat surface (slope < 30 degrees), better socioeconomic and educational and infrastructure indicators, lower violence rate, and greater sociocultural support for physical activity would be positively associated with the percentage of active adolescents living in urban areas.

Methods

Study design and sample

This ecological study employed data from the National School Health Survey (PeNSE), a nationwide school-based survey carried out in 2012, with representatives of ninth-graders from Brazilian public and private schools, from the North, Northeast, Southeast, South, and Midwest, and the 27 capital cities (26 states and the Federal District),²⁷ besides different secondary data sources on the natural environment, socioeconomic and educational indicators, infrastructure, violence, and sociocultural factors.

The PeNSE sample consisted of 27 geographic strata corresponding to capital cities and the Federal District. A sample of clusters was selected in two stages for each stratum. First, a sample of schools (primary sampling units) was selected. Subsequently, these schools were visited, and their ninth-grade classes were randomly listed, composing the secondary sampling units. All students from the selected classes, present on data collection, were invited to answer the survey questionnaire²⁷.

Of the total number of students who regularly attended classes (132,123), in the selected classes, only 110,873 attended classes on the data collection date. However, 1,651 did not wish to participate and 118 did not inform gender or age, being excluded from the database. Thus, the final sample of PENSE, 2012 edition, was composed of 109,104 respondents²⁷. For the analyses of this article, 61,145 adolescents residing in the state capitals were evaluated.

All adolescents included in the study signed an informed consent form. The student’s information is confidential. The PeNSE, carried in 2012, was approved by the National Research Ethics Commission. A detailed description of the sampling process and other information about PENSE can be obtained in the research report, previously published by the Brazilian Institute of Geography and Statistics²⁷.

Variables

The outcome variable of this study was the proportion of active adolescents per capital, that is, those who reported practicing 300 minutes or more of accumulated physical activity per week²⁸. The variable was obtained from the PENSE questionnaire (physical activity block) and created from the combination of six questions that consider information on frequency and time spent on physical activity practices in the last seven days prior to the survey. Three different domains were included²⁷: 1) commuting to school (in the last seven days, on how many days did you walk or walk to or cycle to school?; when do you go to or from school walking or cycling, how much time do you spend?); 2) physical education at school (in the last seven days, how many days did you have physical education classes at school?; in the last seven days, how much time per day did you practice physical activity or sport during physical education classes at school?); 3) other extracurricular physical activities (in the last seven days, not counting physical education classes at school, how many days did you practice physical activity such as sports, dance, gymnastics, bodybuilding, fights, or other exercises?; how long do these activities (such as sports, dancing, gymnastics, bodybuilding, fights, or other activity) that you do usually last?). To calculate the total accumulated physical activity practice time, the frequency (days) was multiplied by the duration (time of daily practice) of the referred activity in each domain, followed by the sum of the three domains resulting in the total time of practice of physical activity.

Additionally, the variables proportion of male adolescents and mean age per city were retrieved from PeNSE (general information block). The explanatory variables were obtained from several secondary data sources and measured at the city level. These variables were divided into five domains described below, with their respective sources.

Natural environment: Consisting of climatic and geographic variables. The former was collected from the database available on the website of the National Institute of Meteorology²⁹ website, considering a six-month period from the beginning to the end of the PeNSE 2012 data collection (April 1^st, 2012 - September 30, 2012). The following measurements were created: mean maximum temperature, mean minimum temperature, mean compensated temperature (in degrees Celsius (ºC)), number of rainfall days (number), mean total rainfall (mm), and mean relative air humidity (%)²⁹. Climate data of the cities of Porto Velho and Campo Grande were not available at the INMET website, and they were obtained from spreadsheets sent by the Institute. Moreover, there were two months of relative air humidity data missing from the city of Vitoria.

Vis-à-vis the geographic variables, the mean slope values of Brazilian capitals were obtained through the geoprocessing software (SRTM-ArcGis Desktop version 10.3.1 and the Spatial Analyst and 3D Analyst extensions), which allows measuring the slope of each geographic area in degrees horizontally (0°-90°), or in percentage, where a slope of 45 degrees is equivalent to 100% slope.

Socioeconomic and educational indicators: Composed of the Municipal Human Development Index (MHDI)³⁰; per capita income³¹; Gini coefficient³², both from the year 2010, and Primary Education Development Index (IDEB)³³. The IDEB is a national indicator of educational quality that combines information on performance in standardized Portuguese and Mathematics exams (Prova Brasil or Primary Education Assessment System (SAEB)), achieved by students at the end of elementary school stages (4^th-5^th grades - initial year; 8^th-9^th grades - final year) with data on school performance (approval/pass grade)^33,34. In this study, the IDEB mean was used for each capital from four measures referring to state schools (initial year + final year) and municipal schools (initial year + final year) of the year 2013. There are only state schools in Brasília. Five capitals did not present one of the four IDEB measurements for 2013, hence data available from the previous year was used. The cities of Maceio, Boa Vista and Rio Branco did not have the final grade IDEB for 2013 (9^th grade) for municipal schools; therefore, data for 2011, 2007 and 2005 were used. Sao Paulo and Vitoria, in turn, did not have municipal and state initial-year (5^th grade) IDEB for 2013, and the values for 2011 and 2009, respectively, were used. Besides these four indicators are the proportion of illiterates of the year 2010³⁵ and the mean objective test score in the National High School Exam of the year 2012³⁶.

Infrastructure: Composed of the percentage of the following characteristics in each capital: households with street identification; street lighting; paving; curb/guide; sidewalk; wheelchair ramp; afforestation; storm drain/mouth; open sewer; garbage accumulated in public places. All variables were obtained by the last Brazilian Institute of Geography and Statistics census^37,38.

Urban violence: The following mortality rates were considered for this domain according to the capital: Specific Mortality Rates (SMR) for homicides for all age groups³⁹, SMR for homicides in the 15-19 age group³⁹, and SMR for land transport accidents for all age groups³⁹, all of the year 2012³⁹.

Sociocultural support for physical activity: In this domain, we evaluated the percentage of respondents (18 years of age or older) of the National Health Survey (carried out in 2013)⁴⁰ on: 1) engaged in the recommended leisure physical activity level (> 150 minutes/week of mild or moderate physical activities, or > 75 minutes/week of vigorous physical activity); 2) engaged in work-related physical activity (walking, heavy cleaning, carrying heavy loads, or another activity that requires intense physical effort) for 150 minutes or more/week; 3) physically active when commuting to their usual activities (work, school), requiring at least 30 minutes a day to commute to/from activities such as walking or cycling; 4) engaged in physical activity for 150 minutes or more while performing household activities (heavy cleaning or activities that require intense physical effort)²⁸. The percentage of students with two or more days of physical education school classes was also included in this domain and obtained from PeNSE carried out in 2012²⁷.

Data analysis

We performed simple and multiple linear regressions to verify the physical activity-associated factors, with estimated regression coefficient and their respective confidence intervals. The explanatory variables were standardized due to their different scales.

All explanatory variables with p-value < 0.20 in the univariate analysis were considered suitable for entering the multiple models, and those with p-value < 0.05 remained in the final model. In the multiple analysis, we adopted the hierarchical entry of variables in blocks: natural environment, socioeconomic and educational indicators, infrastructure, urban violence, and sociocultural support for physical activity. The most distal variables remained as an adjustment for those in the more proximal blocks. The proportion of male adolescents and mean age were included as an adjustment in all models.

We evaluated collinearity separately in each domain, considering Spearman’s correlation coefficient (Supplementary Materials, available at: https://doi.org/10.48331/scielodata.AAEDDM). When two variables were collinear, the one with the highest coefficient of determination (R²) in the univariate analysis was maintained. Therefore, the following were included in the multiple analysis: the mean minimum temperature and relative air humidity (natural environment domain); IDEB and per capita income (socioeconomic and educational indicators domain); households with a ramp for wheelchair users and open sewage (infrastructure domain); SMR for homicides for all age groups (urban violence domain); leisure-time active adults and schoolchildren with two or more days of physical education classes at school (sociocultural support domain for physical activity).

We performed a residual analysis to verify normality, homoscedasticity, and linearity assumptions. Analyses adopted Stata software, version 16 (Stata Corporation, College Station, U.S.). We employed a 5% significance level.

Results

The final sample consisted of 61,145 adolescents (50.84% girls) from 2,219 classes in 1,469 schools. The rate of active adolescents was 33.0% (CI: 32.1-33.9) in the capital cities and the Federal District. The rates of active adolescents were higher in Brasília (39.9%), Campo Grande (38.5%), and Curitiba (38.3%), while Teresina (27.6%), São Luís (27.7%), and Maceió (27.9%) were at the opposite extreme (Figure 1).

Figure 1
Proportion of accumulated physical activity in adolescents according to the 27 Brazilian capitals. National Adolescent School-based Health Survey, Brazil, 2012.

Table 1 shows the mean of each context variable for the set of capitals and the Federal District. We observed a mean maximum temperature of 29.68ºC and a minimum of 20.11ºC. The mean MHDI was 0.78, and the mean IDEB was 4.30. We identified a low proportion of households with garbage accumulated in the streets (6.40%), with a ramp for wheelchair users (7.28%), and with open sewage (17.65%). On the other hand, more than 70% of the households had public lighting (96.11%), paving (82.35%), and curbs/guides (71.43%). The mean SMR for homicides in the 15-19 age group was high (91.60/100,000 inhabitants), while the SMR for land transport accidents for all age groups was 19.93/100,000 inhabitants. Approximately 50% of adolescents had two or more days of physical education at school (Table 1).

Table 1
Description of variables in the natural environment, socioeconomic and education indicators, infrastructure, urban violence and sociocultural support to practice physical activity. Brazil, 2010-2013.

Variables	Mean	Std. dev.	Min., max.
Natural environment domain
Mean maximum temperature (°C)	29.68	3.63	21.61, 35.48
Mean minimum temperature (°C)	20.11	3.54	11.82, 24.79
Mean compensated temperature (°C)	24.25	3.43	15.63, 28.09
Number of rainy days (amount)	67.48	31.67	19.00, 132.00
Mean total rainfall (mm)	109.78	59.41	27.48, 258.38
Mean relative air humidity (%)	73.94	9.57	52.95, 85.06
Maximum slope (%)	52.56	27.60	16.90, 100.00
Mean slope (%)	5.58	3.33	1.47, 14.36
Socioeconomic and education indicators domain
MHDI	0.78	0.04	0.72, 0.85
Per capita income (BRL/1000)	0.94	0.30	0.63, 1.57
IDEB	4.30	0.58	3.15, 5.40
Gini index	0.61	0.03	0.55, 0.69
Average of the objective test score on the ENEM	527.50	25.87	483.30, 568.52
Illiterate rate (%)	4.94	2.45	1.90, 11.30
Infrastructure domain
Households with street identification (%)	66.69	20.63	21.00, 94.10
Households with public lighting (%)	96.11	3.61	81.70, 99.60
Households with pavement (%)	82.35	13.77	49.70, 99.20
Households with curb (%)	71.43	21.48	16.00, 97.50
Households with sidewalk (%)	67.45	19.40	18.20, 94.00
Households with wheelchair ramp (%)	7.28	6.81	0.80, 24.80
Households with trees (%)	58.00	22.69	13.90, 96.40
Households with storm drain (%)	46.05	22.70	7.00, 90.00
Households with open air sewage (%)	17.65	18.89	0.50, 71.80
Households with garbage piled on streets (%)	6.40	3.44	0.80, 15.20
Urban violence domain
Mortality rate from homicide for all age groups (per 100,000)	41.33	16.78	14.30, 79.80
Mortality rate from homicide for 15-19-year-old group (per 100,000)	91.60	57.42	28.10, 223.60
Mortality rate from land transportation accidents for all age groups (per 100,000)	19.93	8.01	9.20, 44.30
Sociocultural support domain to practice physical activity
Active adults in leisure (%)	26.94	4.43	19.10, 35.70
Active adults at work (%)	10.33	2.24	7.30, 15.50
Active adults travelling (%)	28.40	6.23	18.10, 37.90
Active adults in domestic chores (%)	12.25	3.05	7.70, 20.30
Students who had two or more days of physical education (%)	48.97	21.13	14.30, 80.50

• MHDI: Municipal Human Development Index; IDEB: Basic Education Development Index; ENEM: National High School Test.
• Source: Authors.

The variables of the natural environment, socioeconomic and educational indicators, infrastructure, urban violence, and sociocultural support for physical activity were significantly associated with accumulated physical activity (Table 2). The mean minimum temperature, IDEB, households with a wheelchair ramp, and leisure-time active adults maintained statistical significance in the multiple analysis. The analysis indicated that increasing the minimum temperature by one standard deviation reduces the percentage of active adolescents (Model 1). The increase of one standard deviation of IDEB (Model 2), percentage of ramp for wheelchair users (Model 3), and percentage of leisure active adults (Model 4) increase the percentage of active adolescents (Table 3).

Table 2
Univariate association between accumulated physical activity in adolescents and variables in the natural environment, socioeconomic and education indicators, infrastructure, urban violence and sociocultural support to practice physical activity. Brazil, 2010-2013.

Variables	β	95%CI	p-value	R²
Natural environment domain
Mean maximum temperature (°C)	-1.46	-2.89, -0.03	0.046* a	0.15
Mean minimum temperature (°C)	-2.31	-3.53, -1.08	0.001*	0.37
Mean compensated temperature (°C)	-2.14	-3.42, -0.86	0.002* a	0.32
Number of rainy days (amount)	-1.24	-2.70, 0.23	0.095* a	0.11
Mean total rainfall (mm)	-0.50	-2.03, 1.04	0.512	0.02
Mean relative air humidity (%)	-1.68	-3.06, -0.29	0.020*	0.20
Maximum slope (%)	1.29	-0.16, 2.75	0.079* a	0.12
Mean slope (%)	0.76	-0.76, 2.28	0.311	0.04
Socioeconomic and education indicators domain
MHDI	2.36	1.15, 3.57	< 0.001* b	0.39
Per capita income (BRL/1000)	2.38	1.17, 3.58	< 0.001*	0.40
IDEB	2.58	1.45, 3.71	< 0.001*	0.47
Gini index	-2.05	-3.35, -0.74	< 0.001* b	0.29
Average of the objective test score on the ENEM	1.14	-0.34, 2.62	0.125* b	0.09
Illiterate rate (%)	-2.08	-3.38, -0.79	0.003* b	0.31
Infrastructure domain
Households with street identification (%)	2.16	0.89, 3.43	0.002* c	0.32
Households with public lighting (%)	0.94	-0.56, 2.44	0.209	0.06
Households with pavement (%)	1.07	-0.42, 2.55	0.152* c	0.08
Households with curb (%)	0.64	-0.89, 2.17	0.399	0.03
Households with sidewalk (%)	-0.09	-1.64, 1.46	0.906	0.01
Households with wheelchair ramp (%)	2.24	0.99, 3.49	0.001*	0.35
Households with trees (%)	1.06	-0.43, 2.54	0.157* c	0.08
Households with storm drain (%)	1.76	0.39, 3.13	0.014* c	0.22
Households with open air sewage (%)	-2.24	-3.49, -0.99	0.001*	0.35
Households with garbage piled on streets (%)	-1.68	-3.07, -0.30	0.019* c	0.29
Urban violence domain
Mortality rate from homicide for all age groups (per 100,000)	-2.29	-3.52, -1.06	0.001*	0.37
Mortality rate from homicide for 15-19-year-old group (per 100,000)	-1.56	-2.98, -0.15	0.031* d	0.17
Mortality rate from land transportation accidents for all age groups (per 100,000)	0.36	-1.19, 1.90	0.637	0.01
Sociocultural support domain to practice physical activity
Active adults in leisure (%)	2.57	1.44, 3.70	< 0.001*	0.47
Active adults at work (%)	1.31	-0.15, 2.76	0.076* e	0.12
Active adults travelling (%)	-0.28	-1.83, 1.26	0.709	0.01
Active adults in domestic chores (%)	0.72	-0.80, 2.24	0.338	0.04
Students who had two or more days of physical education (%)	2.41	1.22, 3.60	< 0.001*	0.41

• MHDI: Municipal Human Development Index; IDEB: Basic Education Development Index; ENEM: National High School Test. R²: determination coefficient. In bold, the variables that were considered for multiple linear regression. *P-value < 0.20. ^a Collinear variable with “Mean minimum temperature” not used in posterior analyses. ^b Collinear variable with “IDEB” not used in posterior analyses. ^c Collinear variable with “Percentage of households with open air sewage” not used in posterior analyses. ^d Collinear variable with “Mortality rate from homicide for all age groups” not used in posterior analyses. ^e Collinear variable with “Percentage of active adults in leisure” not used in posterior analyses.
• Source: Authors.

Table 3
Hierarchical model of association between accumulated physical activity in adolescents and variables in the natural environment, socioeconomic and education indicators, infrastructure, urban violence and sociocultural support to practice physical activity. Brazil, 2010-2013.

Variables	Model 1		Model 2		Model 3		Model 4
	β	95%CI	β	95%CI	β	95%CI	β	95%CI
Block 1 (natural environment) a
Mean minimum temperature (°C)	-1.63	-3.12, -0.14*	-1.13	-2.57, 0.30	0.09	-1.57, 1.75	-0.14	-1.52, 1.24
Block 2 (socioeconomic and education indicators) b
IDEB			1.76	0.21, 3.31*	2.17	0.73, 3.61	1.60	0.35, 2.85
Block 3 (infrastructure) c
Households with wheelchair ramp (%)					1.73	0.27, 3.20*	1.07	-0.21, 2.36
Block 4 (sociocultural support to practice physical activity) d
Active adults in leisure (%)							1.54	0.56, 2.51*
AIC	139.52		135.44		130.62		120.96
R²	0.37		0.47		0.57		0.71

• IDEB: Basic Education Development Index; AIC: Akaike information criteria; R²: determination coefficient. All models were adjusted per sex and age. ^a Block 1. ^b Controlled by block 1. ^c Controlled by blocks 1 and 2. ^d Controlled by blocks 1, 2 and 3. *P-value < 0.05.
• Source: Authors.

Discussion

This is the first Brazilian nationally representative study to evaluate the association of urban contextual variables and physical activity among adolescents in Brazilian capitals. The main findings of this study pointed to an association between lower average temperature, high educational quality, better urban infrastructure, and high percentage of leisure-time active adults and physical activity among adolescents.

In this study, a negative association was found between the mean minimum temperature and the percentage of active adolescents. However, this result differs from previous studies in European countries, which observed that higher temperatures were associated with higher levels of physical activity²². Recently, Zheng et al.²² conducted the first systematic review and meta-analysis on the associations between climatic conditions and physical activity. They observed a positive association between temperature and physical activity and identified that children and adolescents (aged 3-19 years, from Australia and European, North American, and Asian countries) were more active in favorable climatic conditions, such as temperature around 20°C and less rain. A study using the ecological approach evidenced that the proportion of active English adolescents and adults among those who lived in pleasant places (mild climate and low pollution) was higher than those who lived in hostile environments (exposed to cold weather, pollution, and living near factories)⁹. While this study objectively evaluated the measured temperature, research in Latin American countries, especially Brazil, has used only the perception of barriers, also showing that the unfavorable climate was associated with a higher prevalence of physical activity in adolescents^41,42.

The possible explanations for our result fall into two aspects. The first refers to a specific behavior in which too much heat or cold can directly reduce physical activity, primarily in open environments^6,43. It is noteworthy that increasing temperature in Brazilian capitals reduces thermal comfort⁴⁴ and, consequently, the proportion of active adolescents. Second, land-use policies can affect the availability of open spaces, and transport policies can affect air quality, causing considerable climate change (e.g., temperature), which could restrict physical activity and impact health⁷. Thus, public policies investing in the increase of green areas and including access to sports facilities in open spaces⁸, air protection and improvement, and the planning of transport infrastructure could promote the health of urban dwellers⁷.

IDEB is an educational quality indicator that allows, for example, identifying schools with low incomes and increasing the transfer of financial resources to improve this result³⁴. To the best of our knowledge, this is the first time this indicator has been used to assess the association between the quality of education and the percentage of active adolescents using population-based data. In this study, IDEB was positively associated with the percentage of active adolescents, indicating that improving educational quality can increase the proportion of active students, precluding noncommunicable chronic diseases such as hypertension, diabetes, and other diseases²⁸.

This result may be partially justified because capitals with better education can promote awareness activities about healthy habits and behaviors⁴⁵. Guidance on the physical activity benefits is vital for adopting a more active lifestyle²⁸. The educational environment and health education through the dissemination of knowledge and skills development allow students to incorporate healthy behaviors and enable them to participate as subjects of their health process⁴⁵. Additionally, high IDEB values may be related to better conditions in the school environment, which range from pedagogical resources to physical infrastructure, such as the presence of sports equipment⁴⁶. In the latter case, schools that have more and/or better physical activity facilities can encourage the practice of physical activity inside and outside the school and, therefore, students may be more likely to be active enough^47,48. These findings suggest that the school environment also plays an important role in the health of students. Therefore, public policies aimed at education must guarantee, among other aspects, material resources and infrastructure in adequate conditions for the practice of physical activity^46,48.

Furthermore, we found a positive and significant correlation between IDEB and per capita income and MHDI, a proxy for the city’s socioeconomic level. It is essential to consider that the socioeconomic inequality level of a place can contribute to the unequal distribution of urban infrastructure⁵. Lower-income peripheral areas tend to have worse quality of features of the physical activity-related built environment, influencing the probability of being active¹¹. For example, American teenagers from low-income neighborhoods were less likely to be active in commuting due to their low ability to walk in the neighborhood⁴⁹. Furthermore, Brazilian adolescents residing in neighborhoods with better per capita household income were more likely to be active in leisure time¹⁵.

Among the indicators of the infrastructure domain, we observed a positive association between the percentage of households with a ramp for wheelchair users and physical activity. The ramp was defined as “a lowering of the sidewalk or curb/guide, usually near corners, specifically intended to give access to people using wheelchairs”³⁸. It is noteworthy that the installation of ramps must follow specific criteria established by the Brazilian Association of Technical Standards⁵⁰, such as slope not exceeding 8.33%, minimum width of 1.20 m, and non-slip floor. A city with a higher percentage of wheelchair ramps probably also has better quality sidewalks (fewer holes and more suitable walking floors) and a better infrastructure in general, which would facilitate leisure-time physical activity for children and adolescents, such as walking, cycling, and skateboarding⁵¹.

Studies investigating the association between city infrastructure and physical activity levels among adolescents have increased in recent decades. In these works, researchers have shown that other urban design attributes can influence the likelihood of more active adolescents, such as the availability of sidewalks¹⁴, location and material of sidewalks¹⁶, public lighting, paved streets¹⁵, bicycle lanes^15,17, and intersection density (street connectivity)^16,52. These results strengthen the findings of this study, indicating that interventions in the urban infrastructure can influence the physical activity levels in adolescents.

Regarding physical activity’s sociocultural support, we observed that a higher percentage of leisure-time active adults was associated with regular physical activity in adolescents, which is consistent with results obtained by other studies, showing that having active adults at home contributes to an increase in physical activity, especially among boys^19,20. In general, parents and relatives can play an essential role in increasing physical activity levels,⁵³ so that the higher the percentage of adults engaged in physical activity, the greater the incentive for adolescents to engage in physical activity²⁰. This association can be explained by the social learning theory, which considers that people learn by observing others. In the practice of physical activity, it can be expressed through the modeling of interests and skills and in the reinforcement of behaviors⁵⁴.

Some scholars argue that social support has been considered an essential factor for adopting a more active lifestyle²³, which can occur through emotional, financial, material⁵⁵, organizational, community, or political support¹⁸. Adolescents with more supportive parents, siblings, and friends tend to participate in a higher physical activity level^18,55. Moreover, we could consider that cities with more active adults have more physical activity structure, such as public spaces for leisure sports^56,57. While not evaluated in this study, the research results above reinforce the importance of promoting public policies to foster physical activity in adolescents with a socio-ecological approach.

Regarding the urban violence domain, only homicide rates were associated with the percentage of active adolescents in the crude analysis, which was not maintained in the adjusted analysis. As far as we know, in Brazil, only one study analyzed the association between physical inactivity and the homicide rate among adolescents living in 122 municipalities with over 100,000 inhabitants. It was part of the Study of Cardiovascular Risk Factors in Adolescents, in which researchers found no significant association⁵⁸. Previously published studies have investigated the relationship between physical activity and urban violence from adolescents’ perspectives. However, inconsistent results persist. While Mendonça et al.¹² showed that adolescents from João Pessoa who perceived that the neighborhood environment was not violent were more likely to report active commuting, other national studies have not found an association between urban violence and adolescent physical activity^23-25.

Significant associations were found between perceived level of safety in the neighborhood (walking/cycling infrastructure, traffic risks, and crime safety) and leisurely walking in Poland and the Czech Republic, which was stronger in girls and transport walking.¹³ These studies reveal the difficulty of comparing studies due to the different methodological approaches to assess urban violence, and few works specifically evaluated homicide vis-à-vis physical activity in adolescents.

This study has limitations inherent to its ecological design: variables that would contribute to physical activity among adolescents, such as the availability of public and private leisure facilities (e.g., gyms, courts, soccer fields, and parks), were not included. Additionally, there may be heterogeneous attributes within cities, and we were unable to assess the neighborhood of each adolescent. Although it is an index that only considers public schools, IDEB is representative of the outcome studied since 75% of the PeNSE sample (2012) in the capitals derives from public schools.

Furthermore, the same variables remained associated when only the percentage of active adolescents from public schools were considered, and the latter model had lower explanatory power than the one that evaluates students from public and private schools. Nevertheless, the major strength of this study was that it had a nationally representative sample of adolescents evaluating the association between several environmental factors (natural, constructed, and social) and regular physical activity, allowing the generalization of the results.

Conclusions

The study observed that climatic and educational factors, infrastructure, and social context could increase the percentage of active Brazilian adolescents. The environment of cities impacts health, and there is a need for healthier and more sustainable cities. Urban planning policies and their interfaces must be multisectoral. They must integrate and prioritize transport, land use, the economy, housing, and infrastructure planning with urban design, which will protect citizens from traffic, environmental pollution, noise, crime, and violence and promote active lifestyles such as safe walking, cycling, and access to parks.

Acknowledgments

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) for the research productivity scholarship to the researcher WT Caiaffa and DC Malta. Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) for the doctoral scholarship to the PhD student JI Silva.

References

• ¹
  United Nations (UN). World Urbanization Prospects: the 2018 revision New York: UN; 2019. [cited 2021 jun 15]. Available from: https://population.un.org/wup/Publications/Files/WUP2018-Report.pdf
• ²
  Caiaffa WT, Ferreira FR, Ferreira AD, Oliveira CDL, Camargos VP, Proietti FA. Saúde urbana: "a cidade é uma estranha senhora, que hoje sorri e amanhã te devora." Cien Saude Colet 2008; 13(6):1785-1796.
• ³
  Diez Roux AV, Slesinski SC, Alazraqui M, Caiaffa WT, Frenz P, Fuchs RJ, Miranda JJ, Rodriguez DA, Dueñas OLS, Siri J, Vergara AV. A novel international partnership for actionable evidence on urban health in Latin America: LAC-Urban Health and SALURBAL. Global Challenges 2019; 3(4):1800013.
• ⁴
  Singh S, Beagley J. Health and the new urban agenda: a mandate for action. Lancet 2017; 389(10071):801-802.
• ⁵
  Bauman AE, Reis RS, Sallis JF, Wells JC, Loos RJF, Martin BW; Lancet Physical Activity Series Working Group. Correlates of physical activity: why are some people physically active and others not? Lancet 2012; 380(9838):258-271.
• ⁶
  Sallis JF, Cervero RB, Ascher W, Henderson KA, Kraft MK, Kerr J. An ecological approach to creating active living communities. Annu Rev Public Health 2006; 27:297-322.
• ⁷
  Giles-Corti B, Vernez-Moudon A, Reis R, Turrell G, Dannenberg AL, Badland H, Foster S, Lowe M, Sallis JF, Stevenson M, Owen N. City planning and population health: a global challenge. Lancet 2016; 388(10062):2912-2924.
• ⁸
  Beghin L, Vanhelst J, Drumez E, Migueles J, Manios Y, Moreno LA, Henauw S, Gottrand F. Influence of meteorological conditions on physical activity in adolescents. J Epidemiol Community Health 2020; 74(4):395-400.
• ⁹
  Rind E, Shortt N, Mitchell R, Richardson EA, Pearce J. Are income-related differences in active travel associated with physical environmental characteristics? A multi-level ecological approach. Int J Behav Nutr Phys Act 2015; 12:73.
• ¹⁰
  Dong J, Zhang S, Xia L, Yu Y, Hu S, Sun J, Zhou P, Chen P. Physical activity, a critical exposure factor of environmental pollution in children and adolescents health risk assessment. Int J Environ Res Public Health 2018; 15(2):176.
• ¹¹
  Santos DS, Hino AAF, Höfelmann DA. Iniquidades do ambiente construído relacionado à atividade física no entorno de escolas públicas de Curitiba, Paraná, Brasil. Cad Saude Publica 2019; 35(5):e00110218.
• ¹²
  Mendonça G, Florindo AA, Rech CR, Freitas DKS, Farias Júnior JC. Perceived neighborhood environmental characteristics and different types of physical activity among Brazilian adolescents. J Sports Sci 2018; 36(9):1068-1075.
• ¹³
  Mitáš J, Sas-Nowosielski K, Groffik D, Frömel K. The safety of the neighborhood environment and physical activity in Czech and Polish adolescents. Int J Environ Res Public Health 2018; 15(1):126.
• ¹⁴
  An R, Shen J, Yang Q, Yang Y. Impact of built environment on physical activity and obesity among children and adolescents in China: a narrative systematic review. J Sport Health Sci 2019; 8(2):153-169.
• ¹⁵
  Silva ICM, Hino AA, Lopes A, Ekelund U, Brage S, Gonçalves H, Menezes AB, Reis RS, Hallal PC. Built environment and physical activity: domain-and activity-specific associations among Brazilian adolescents. BMC Public Health 2017; 17(1):616.
• ¹⁶
  Smith M, Hosking J, Woodward A, Witten K, MacMillan A, Field A, Baas P, Mackie H. Systematic literature review of built environment effects on physical activity and active transport - an update and new findings on health equity. Int J Behav Nutr Phys Act 2017; 14:158.
• ¹⁷
  Van Hecke L, Ghekiere A, Veitch J, Dyck DV, Cauwenberg JV, Clarys P, Deforche B. Public open space characteristics influencing adolescents' use and physical activity: a systematic literature review of qualitative and quantitative studies. Health Place 2018; 51:158-173.
• ¹⁸
  Hu D, Zhou S, Crowley-Mchattan ZJ, Liu Z. Factors that influence participation in physical activity in school-aged children and adolescents: a systematic review from the social ecological model perspective. Int J Environ Res Public Health 2021; 18(6):3147.
• ¹⁹
  Dozier SGH, Schroeder K, Lee J, Fulkerson JA, Kubik MY. The association between parents and children meeting physical activity guidelines. J Pediatr Nurs 2020; 52:70-75.
• ²⁰
  Ramos CGC, Andrade RG, Andrade ACS, Fernandes AP, Costa DAS, Xavier CC, Proietti FA, Caiaffa WT. Contexto familiar e atividade física de adolescentes: cotejando diferenças. Rev Bras Epidemiol 2017; 20(3):537-548.
• ²¹
  Kärmeniemi M, Lankila T, Ikäheimo T, Koivumaa-Honkanen H, Korpelainen R. The built environment as a determinant of physical activity: a systematic review of longitudinal studies and natural experiments. Ann Behav Med 2018; 52(3):239-251.
• ²²
  Zheng C, Feng J, Huang WY, Wong SHS. Associations between weather conditions and physical activity and sedentary time in children and adolescents: a systematic review and meta-analysis. Health Place 2021; 69:102546.
• ²³
  Farias Junior JC, Reis RS, Hallal PC. Physical activity, psychosocial and perceived environmental factors in adolescents from Northeast Brazil. Cad Saude Publica 2014; 30(5):941-951.
• ²⁴
  Lopes AAS, Lanzoni AN, Hino AAF, Rodriguez-Añez CR, Reis RS. Ambiente do bairro percebido e atividade física entre estudantes do ensino médio de Curitiba, Brasil. Rev Bras Epidemiol 2014; 17(4):938-953.
• ²⁵
  Dias AF, Brand C, Lemes VB, Gaya AR, Gaya ACA. Perception of square and park characteristics and physical activity practice among high school students. Rev Bras Ativ Fis Saude 2017; 22(2):155-164.
• ²⁶
  Dias AF, Gaya AR, Santos MP, Brand C, Pizarro AN, Fochesatto CF, Mendes TM, Mota J, Gaya ACA. Neighborhood environmental factors associated with leisure walking in adolescents. Rev Saude Publica 2020; 54:61.
• ²⁷
  Instituto Brasileiro de Geografia e Estatística (IBGE). Pesquisa Nacional de Saúde do Escolar 2012. Rio de Janeiro: IBGE; 2013. [acessado 2022 dez 11]. Disponível em: https://actbr.org.br/uploads/arquivos/PeNSE_2012.pdf
• ²⁸
  World Health Organization (WHO). Guidelines on physical activity and sedentary behaviour. Geneva: WHO.
• ²⁹
  Instituto Nacional de Meteorologia. Banco de dados climáticos disponível das cidades [Internet]. 2012. [cited 2016 Mar 15]. Available from: http://www.inmet.gov.br/portal/
  » http://www.inmet.gov.br/portal
• ³⁰
  Programa das Nações Unidas para o Desenvolvimento. Municipal Human Development Index [Internet]. 2010. [cited 2016 mar 25]. Available from: http://www.pnud.org.br/atlas/ranking/Ranking-IDHM-Municipios-2010.aspx
  » http://www.pnud.org.br/atlas/ranking/Ranking-IDHM-Municipios-2010.aspx
• ³¹
  Instituto Brasileiro de Geografia e Estatística (IBGE). Censo Demográfico 2010. Renda per capita [Internet]. 2010. [acessado 2022 mar 25]. Disponível em: https://sidra.ibge.gov.br/tabela/3563
  » https://sidra.ibge.gov.br/tabela/3563
• ³²
  Instituto Brasileiro de Geografia e Estatística (IBGE). Censo Demográfico 2010. Índice de Gini da renda domicilia per capita [Internet]. 2010. [acessado 2016 out 20]. Disponível em: http://tabnet.datasus.gov.br/cgi/idb2011/b09capc.htm
  » http://tabnet.datasus.gov.br/cgi/idb2011/b09capc.htm
• ³³
  Instituto Nacional de Estudos e Pesquisas Educacionais Anísio Teixeira (Inep). Índice de Desenvolvimento da Educação Básica por cidade e dependência administrativa [Internet]. 2013. [acessado 2016 jul 5]. Disponível em: http://ideb.inep.gov.br/resultado/home.seam?cid=4715572
  » http://ideb.inep.gov.br/resultado/home.seam?cid=4715572
• ³⁴
  Instituto Nacional de Estudos e Pesquisas Educacionais Anísio Teixeira (Inep). Nota técnica sobre o Índice de Desenvolvimento da Educação Básica [Internet]. [acessado 2016 jul 17]. Disponível em: http://download.inep.gov.br/educacao_basica/portal_ideb/o_que_e_o_ideb/Nota_Tecnica_n1_concepcaoIDEB.pdf
  » http://download.inep.gov.br/educacao_basica/portal_ideb/o_que_e_o_ideb/Nota_Tecnica_n1_concepcaoIDEB.pdf
• ³⁵
  Instituto Brasileiro de Geografia e Estatística (IBGE). Censo Demográfico 2010. Proporção de analfabetos por cidade [Internet]. 2010. [acessado 2016 out 22]. Disponível em: http://tabnet2.datasus.gov.br/cgi/tabcgi.exe?idb2013/b01a.def
  » http://tabnet2.datasus.gov.br/cgi/tabcgi.exe?idb2013/b01a.def
• ³⁶
  Instituto Nacional de Estudos e Pesquisas Educacionais Anísio Teixeira (Inep). Exame Nacional do Ensino Médio: nota por escola, estado, cidade [Internet]. 2012. [acessado 2016 mar 12]. Disponível em: http://portal.inep.gov.br/visualizar/-/asset_publisher/6AhJ/ontente/ministro-divulga-enem-por-escola
  » http://portal.inep.gov.br/visualizar/-/asset_publisher/6AhJ/ontente/ministro-divulga-enem-por-escola
• ³⁷
  Instituto Brasileiro de Geografia e Estatística (IBGE). Censo Demográfico 2010. Características urbanísticas do entorno dos domicílios [Internet]. 2010. [acessado 2016 jun 10]. Disponível em: http://www.ibge.gov.br/home/estatistica/populacao/censo2010/entorno/entorno_tab_municipios_zip_xls.shtm
  » http://www.ibge.gov.br/home/estatistica/populacao/censo2010/entorno/entorno_tab_municipios_zip_xls.shtm
• ³⁸
  Instituto Brasileiro de Geografia e Estatística (IBGE). Censo Demográfico 2010. Características urbanísticas do entorno dos domicílios [Internet]. 2010. [acessado 2016 jun 10]. Disponível em: https://biblioteca.ibge.gov.br/visualizacao/periodicos/96/cd_2010_entorno_domicilios.pdf
  » https://biblioteca.ibge.gov.br/visualizacao/periodicos/96/cd_2010_entorno_domicilios.pdf
• ³⁹
  Brasil. Ministério da Saúde (MS). Sistema de Informações sobre Mortalidade. Indicadores de mortalidade [Internet]. 2012. [acessado 2016 abr 19]. Disponível em: http://tabnet2.datasus.gov.br/cgi/tabcgi.exe?idb2013/ c09.def
  » http://tabnet2.datasus.gov.br/cgi/tabcgi.exe?idb2013/ c09.def
• ⁴⁰
  Instituto Brasileiro de Geografia e Estatística (IBGE). Pesquisa Nacional de Saúde 2013. Rio de Janeiro: IBGE; 2013.
• ⁴¹
  Dias DF, Loch MR, Ronque ERV. Barreiras percebidas à prática de atividades físicas no lazer e fatores associados em adolescentes. Cien Saude Colet 2015; 20(11):3339-3350.
• ⁴²
  Santos MS, Hino AAF, Reis RS, Rodriguez-Añez CR. Prevalência de barreiras para a prática de atividade física em adolescentes. Rev Bras Epidemiol 2010; 13(1):94-104.
• ⁴³
  Coutts C, Forkink A, Weiner J. The portrayal of natural environment in the evolution of the ecological public health paradigm. Int J Environ Res Public Health 2014; 11(1):1005-1019.
• ⁴⁴
  Associação Brasileira de Normas Técnicas (ABNT). NBR 16401-2. Instalações de ar-condicionado - sistemas centrais e unitários. Parte 2: Parâmetros de conforto térmico [Internet]. 2008. [acessado 2016 jul 18]. Disponível em: http://ftp.demec.ufpr.br/disciplinas/EngMec_NOTURNO/TM374/NBR_16401-2_2008.pdf
  » http://ftp.demec.ufpr.br/disciplinas/EngMec_NOTURNO/TM374/NBR_16401-2_2008.pdf
• ⁴⁵
  Faial LCM, Silva RMCRA, Pereira ER, Refrande SM, Souza LMC, Faial CSG. A escola como campo de promoção à saúde na adolescência: revisão literária. Pró-UniverSUS 2016; 7(2):22-29.
• ⁴⁶
  Reis ACA, França M, Costa ACM. Ambiente escolar e o Ideb em municípios do Pará, Rio Grande do Norte e Minas Gerais. Rev Educ Questão 2017; 55(45):254-280.
• ⁴⁷
  Rezende LFM, Azeredo CM, Silva KS, Claro RM, França-Junior I, Peres MFT, Luiz OC, Levy RB, Eluf-Neto J. The role of school environment in physical activity among Brazilian adolescents. PLoS ONE 2015; 10(6):e0131342.
• ⁴⁸
  Silva KS, Bandeira AS, Ravagnani FCP, Camargo EM, Tenório MC, Oliveira VJM, Santos PC, Ramires VV, Sandreschi PF, Hallal PC, Barbosa Filho VC. Educação física escolar: Guia de Atividade Física para a População Brasileira. Rev Bras Ativ Fis Saúde 2021; 26:e0219.
• ⁴⁹
  Sallis JF, Conway TL, Cain KL, Carlson JA, Frank LD, Kerr J, Glanz K, Chapmane JE, Saelens BE. Neighborhood built environment and socioeconomic status in relation to physical activity, sedentary behavior, and weight status of adolescents. Prev Med 2018; 110:47-54.
• ⁵⁰
  Associação Brasileira de Normas Técnicas (ABNT). NBR 9050:2004. Acessibilidade a edificações, mobiliário, espaços e equipamentos urbanos [Internet]. 2004. [acessado 2016 jul 15]. Disponível em: https://www.prefeitura.sp.gov.br/cidade/secretarias/upload/nbr_%2009050_acessibilidade%20-%202004%20-%20acessibilidade_a_edificacoes_mobiliario_1259175853.pdf
  » https://www.prefeitura.sp.gov.br/cidade/secretarias/upload/nbr_%2009050_acessibilidade%20-%202004%20-%20acessibilidade_a_edificacoes_mobiliario_1259175853.pdf
• ⁵¹
  Sallis JF, Cain KL, Conway TL, Gavand KA, Millstein RA, Geremia CM, Frank LD, Saelens BE, Glanz K, King AC. Is your neighborhood designed to support physical activity? A brief streetscape audit tool. Prev Chronic Dis 2015; 12:150098.
• ⁵²
  Queralt A, Molina-García J. Physical activity and active commuting in relation to objectively measured built-environment attributes among adolescents. J Phys Act Health 2019; 16(5):371-374.
• ⁵³
  Lawler M, Heary C, Nixon E. Peer support and role modelling predict physical activity change among adolescents over twelve months. J Youth Adolesc 2020; 49(7):1503-1516.
• ⁵⁴
  Seabra AF, Mendonça DM, Thomis MA, Anjos LA, Maia JÁ. Determinantes biológicos e sócio-culturais associados à prática de atividade física de adolescentes. Cad Saude Publica 2008; 24(4):721-736.
• ⁵⁵
  Ren Z, Hu L, Yu JJ, Yu Q, Chen S, Ma Y, Lin J, Yang L, Li X, Zou L. The influence of social support on physical activity in chinese adolescents: the mediating role of exercise self-efficacy. Children 2020; 7(3):23.
• ⁵⁶
  Hino AAF, Rech CR, Gonçalves PB, Reis RS. Acessibilidade a espaços públicos de lazer e atividade física em adultos de Curitiba, Paraná, Brasil. Cad Saude Publica 2019; 35(12):e00020719.
• ⁵⁷
  Manta SW, Del Duca GF, Silva KS, Rech CR, Gomes RS, Maximiano GP, Malta DC. Is the availability of open public spaces associated with leisure-time physical activity in Brazilian adults? Health Promot Int 2020; 35(1):e51-e58.
• ⁵⁸
  Rodrigues RRD, Szklo M, Souza AM. Association between homicide rates and prevalence of cardiovascular risk factors in the municipalities included in the Study of Cardiovascular Risk Factors in Adolescents. Public Health 2020; 187:103-108.

Chief editors:

Romeu Gomes, Antônio Augusto Moura da Silva

Publication Dates

• Publication in this collection
  10 Nov 2023
• Date of issue
  Nov 2023

History

• Received
  28 Sept 2022
• Accepted
  06 Feb 2023
• Published
  08 Feb 2023