Environmental correlates of physical activity in youth – a review and update

Authors


J Brug, Department of Public Health, Erasmus University Medical Center, PO Box 1738, 3000 DR Rotterdam, the Netherlands. E-mail: j.brug@erasmusmc.nl

Summary

Obesogenic environments are thought to underlie the increased obesity prevalence observed in youth during the past decades. Understanding the environmental factors that are associated with physical activity (PA) in youth is needed to better inform the development of effective intervention strategies attempting to halt the obesity epidemic. We conducted a systematic semi-quantitative review of 150 studies on environmental correlates of youth PA published in the past 25 years. The ANalysis Grid for Environments Linked to Obesity (ANGELO) framework was used to classify the environmental correlates studied. Most studies retrieved used cross-sectional designs and subjective measures of environmental factors and PA. Variables of the home and school environments were especially associated with children’s PA. Most consistent positive correlates of PA were father’s PA, time spent outdoors and school PA-related policies (in children), and support from significant others, mother’s education level, family income, and non-vocational school attendance (in adolescents). Low crime incidence (in adolescents) was characteristic of the neighbourhood environment associated with higher PA. Convincing evidence of an important role for many other environmental factors was, however, not found. Further research should aim at longitudinal and intervention studies, and use more objective measures of PA and its potential (environmental) determinants.

Physical activity (PA) is a health enhancing behaviour: when practised regularly, PA reduces the risk for a range of chronic disease (1–4). Also among the young, current and future health benefits can be obtained through engaging in physically active lifestyles (5): it helps building strong bones, healthy joints, a strong heart, a good mental health and prevents today’s major public health concern – obesity (6–9). Despite these health benefits, many young people are not engaging in recommended levels of PA (10–12). In addition, longitudinal studies have shown that a steep decrease in PA levels occurs during adolescence (13–15) and that PA levels established in youth tend to track into adulthood (16–20); PA promotion in youth is thought to facilitate a carryover of healthful habits into adulthood and a lifelong protection from other risk factors, and is therefore a priority in current public health policies (4,21).

Given the short-time frame in which the obesity prevalence has increased to epidemic scales, many scientists postulate that this is more likely due to changes in environments than in biology (22–26). In this vein, recent studies have indeed demonstrated associations between childhood obesity and environmental features, namely at the home and neighbourhood level (27–32). Consequently, it is important to understand, measure and alter environments that promote or hinder obesity-inducing behaviours, such as low PA (7,33–38). Environmental influences can be especially relevant to children and adolescents because they have less autonomy in their behavioural choices (39). Specific recommendations for research on the determinants of PA in youth have emphasized the need to examine environmental influences on youth PA at different levels (e.g. home, neighbourhood, school) (40–42) to better inform the development of interventions attempting to improve PA levels (43,44).

Now that an increasing number of studies focus on potential environmental influences on children’s and adolescents’ PA behaviour, it is important to get a detailed overview of the evidence these studies have provided so far, to define a research agenda in this area. In the year 2000, a comprehensive review of personal and environmental correlates of PA in children and adolescents (45) identified several variables which were consistently associated with children/adolescent’s PA. These included social and physical environmental factors such as direct help and support from parents and significant others, access to programmes/facilities, opportunities to be active and time spent outdoors. We now update the review of evidence provided by Sallis et al. (45) focusing specifically on, and characterizing into more detail, the environmental correlates of PA in children and adolescents.

Methods

Search strategies and procedures

Relevant studies were located from two main sources. First, the computerized literature databases MedLine (PubMed), PsycInfo, Web of Science, EMBASE and SportDiscus were searched. The following keyword combinations were used: physical activity, physical active lifestyle, vigorous activity, leisure activities, recreation, exercise, sport(s), motor activity, physical education, walking, running (bi)cycling, commuting, determinants, correlates, influences, associations, environment, physical environment, built environment, psychosocial determinants, social environment, social norms, socio-economic status, socio-cultural environment, parents, peers, neighbourhood, school, facilities, recreation, equipment, safety. These searches were restricted to studies performed in humans aged up to 18 years, and published between January 1980 and December 2004. After excluding duplicate studies, over 5000 articles were hereby identified. Two independent reviewers (IF, KvdH) screened and selected the articles retrieved whenever it could be ascertained first, from the title (304 articles), second from the abstract (88 articles), and finally from the full text (84 articles), that the selection criteria (see below) were met. These stepwise analyses were performed separately by each reviewer, and at each step, an article was kept whenever selected by at least one of the reviewers.

Second, manual searches using the reference of the previous systematic review from Sallis et al. (45), primary studies located from the previous source and our personal databases were performed and cross-checked with the 84 articles initially selected. This led to the inclusion of 66 additional articles. Together, these search strategies resulted in a total of 150 articles, which are reviewed herein.

Inclusion/exclusion criteria

Types of studies

The present review was concerned with PA levels occurring ‘naturally’ in populations of children and adolescents. Therefore, only observational studies (either cross-sectional or longitudinal) were included, whereas studies investigating PA-related interventions or with a quasi-experimental design were excluded (with exception of studies reporting on baseline data from intervention studies). Qualitative studies or studies that were solely descriptive in nature (i.e. reporting only frequency data), abstracts, case reports, expert opinions, dissertations and unpublished data were also excluded.

Participants and country

Subjects (or the majority of the participants) had to be in the age range of 3–18 years old; similarly to the review by Sallis et al. (45) we have divided studies among children (i.e. 3–12 years old) and adolescents (>12–18 years old). Studies on children and adolescents with chronic diseases (that may affect PA levels) or children participating in top-level competitive sports were not included. Only studies from samples drawn in countries with established market economies (as defined by the World Bank) and published in English as papers in international peer-reviewed journals were included.

The dependent variable(s) – physical activity

The dependent variable was any measure of (overall) PA of various types (i.e. play, games, sports, work, transportation, recreation, physical education [PE), or planned exercise) performed in the context of home/family, school and community, and expressed in terms of duration (e.g. in minutes), or frequency (e.g. times per week), or intensity (e.g. vigorous) or a combination of these, i.e. in terms of volume (e.g. METs or kcal) (46). When studies had multiple dependent measures of PA the correlates of mutually exclusive outcomes (e.g. habitual levels of moderate- and vigorous-intensity PA) were investigated and reported separately.

Studies in which the dependent variable was aerobic fitness, intention, self-efficacy, or other intermediate (non-behavioural) measures were not included; physical inactivity/sedentary behaviour was not considered as outcome because physical activity and inactivity are distinct behaviours, often unrelated and with distinct determinants (47–51). In addition, although we acknowledge physical inactivity as an important heath-impairing behaviour, a systematic review of its determinants among youth has been recently published (52).

The predictor variable(s) – environmental characteristics

Environmental variables were broadly defined as ‘anything outside the individual that can affect his/her PA behaviour’. To structure our review we were in need of a conceptual framework to categorize the various environmental factors studied. Different classifications of possible environmental determinants of health behaviours have been proposed (36,38,47,53,54), all of them showing great overlap and similarities. In the present review we have adopted The ANalysis Grid for Environments Linked to Obesity (ANGELO) conceptual framework (55) to classify potential environmental determinants of PA in children and adolescents. This framework was specifically developed to conceptualize ‘obesogenic’ environments (i.e. those that promote excessive energy intake and low PA), enabling the identification of specific areas and settings to be targeted by intervention programmes. Specifically, environmental variables can be distinguished within two ‘sizes’ (micro and macro) and four types (physical, socio-cultural, economic and political, the latter hereafter referred as ‘policies’). Micro-environments are defined as environmental settings where groups of people meet and gather. Such settings are often geographically distinct and allow direct mutual influences between individuals and the environment. Examples of micro-environmental settings are homes, schools and neighbourhoods. Macro-environments, on the other hand, include the broader, more anonymous infrastructure that may support or hinder health behaviours. Examples of macro-environments are the town planning, the transport infrastructure, the media and the healthcare system. All studies reviewed herein were required to examine at least one environmental variable (the independent variable), and this variable needed to be tested for its association with a measure of PA (the dependent variable), obtained at the individual level.

Data analyses

Because of the great variety of variables and methods drawn from diverse samples, a meta-analytical review was not possible to conduct. We have therefore adopted the same semi-quantitative approach outlined by Sallis et al. (45) recently also used by Gorely et al. (52) in a review of the correlates of television viewing among youth. An independent sample, i.e. the smallest independent sub-sample (based on age and gender) for which relevant data were reported (e.g. studies reporting findings for boys [M] and girls [F] separately, provide two independent samples) was used as the unit of analyses (56).

Study characteristics

The relevant characteristics from all the selected publications listed in the Bibliography section were retrieved and registered in detailed tables (which are available upon request from the Correspondence), according to current review guidelines (57,58). This extensive information was then summarized in one background table (Table 1).

Table 1.  Child and adolescents studies categorized by sample size, sex, study design, physical activity measurement issues, and country
 Children (3–12 years)Adolescents (>12–18 years)
Bibliography no.♯ Individual samples%Bibliography no.♯ Individual samples%
  • *

    These two studies report on the exact same dataset and were therefore considered as one individual sample only (hereafter coded as 10/31).

  • F, girls only; M/F, boys and girls analysed separately; I, II, III, IV, V, VI, data reported for different age sub-groups separately.

Sample size 91100 134100
<1005F, 15M/F, 25, 32, 39, 66, 71, 82, 89M/F, 111, 113MI, V/FI, II, V, 129, 134M1920.920I, 23M/F, 24M/F, 26M/F, 30F, 50M, 68, 69, 90M/F, 106F, 113MIV,VI/FVI, 123, 126M, 133M, 138I,II2216.4
100–19910*, 21, 22F, 31*, 42, 46, 51, 58, 75M, 76, 103M/F, 108M/F, 113MII, 124M/F, 131MI/FI, 134F, 1502022.014, 44M/F, 45M/F, 48II, 49, 50F, 101, 102F, 110II, 113MIII/FIII,IV, 126F, 128, 133F1712.7
200–29913F, 28, 37, 38, 41, 63M/F, 70, 75F, 96, 109, 110I, 112MI/FI, 1181516.512, 16M/F, 20II, 93F, 102M, 112MII, III/FII, III, 125, 148MIII128.9
300–49927M/F, 56F, 81, 97, 107M/F, 115, 131MII/FII, 132, 148MI/FI1314.38M/F, 34M/F, 40F, 43, 67M, 79M/F, 105, 116F, 135, 147, 148MII/FII, III1611.9
500–99956M, 64, 84M/F, 100, 104, 119, 14488.84M/F, 7M/F, 17F, 18F, 29, 33, 36, 40M, 47, 48I, 61M/F, 62, 67F, 87F, 99, 145, 149M/F2115.7
1000–299911M/F, 57, 73, 85, 86M/F, 88, 95M/F, 120, 122, 137, 143M/F1516.51M/F, 3M/F, 6, 35F, 54, 65M/F, 80M/F, 87M, 91F, 94I,II, 114F, 117M/F, 136, 139, 141, 142M/F, 146M/F2518.7
3000–4999   2M/F, 72, 78, 83, 121, 14075.2
≥50001911.09, 52, 53, 55, 59M/F, 60, 74, 77, 92, 98, 127M/F, 1301410.5
Sex
Girls only5F, 13F, 2233.317, 18, 30, 35, 91, 93, 106, 114, 11696.7
Boys and girls combined10/31, 19, 21, 25, 28, 32, 37, 38, 39, 41, 42, 46, 51, 57, 58, 64, 66, 70, 71, 73, 76, 81, 82, 85, 88, 96, 97, 100, 104, 109, 110I, 111, 115, 118, 119, 120, 122, 129, 132, 137, 144, 1504246.26, 9, 12, 14, 20I,II, 29, 33, 36, 43, 47, 48I/II, 49, 52, 53, 54, 55, 60, 62, 68, 69, 72, 74, 77, 78, 83, 92, 94I,II, 98, 99, 101, 105, 110II, 121, 123, 125, 128, 130, 135, 136, 138, 139, 140, 141, 145, 1474936.6
Boys and girls separately11, 15, 27, 56, 63, 75, 84, 86, 89, 95, 103, 107, 108, 112I, 113I,II,V, 124, 131I,II, 134, 143, 148I4650.51, 2, 3, 4, 7, 8, 16, 23, 24, 26, 34, 40, 44, 45, 50, 59, 61, 65, 67, 79, 80, 87, 90, 102, 112II,III, 113III,IV,VI, 117, 126, 127, 133, 142, 146, 148II,III, 1497656.7
Study design
Cross-sectional5F, 10/31, 13F, 15M/F, 19, 21, 22F, 25, 27M/F, 28, 32, 37, 38, 39, 41, 46, 51, 56M/F, 57, 58, 63M/F, 64, 66, 70, 71, 73, 75M/F, 76, 81, 84M/F, 85, 86M/F, 88, 89M/F, 95M/F, 96, 97, 100, 103M/F, 104, 108M/F, 109, 110I, 111, 112MI/FI, 113MI,II,V/FI,II,V, 115, 119, 120, 122, 124M/F, 129, 131MI,II/FI,II, 132, 134M/F, 143M/F, 144, 1508189.01M/F, 3M/F, 4M/F, 6, 7M/F, 8M/F, 9, 12, 14, 16M/F, 17F, 18F, 23M/F, 24M/F, 29, 30F, 33, 34M/F, 35F, 36, 43, 44M/F, 47, 48I/II, 49, 50M/F, 52, 53, 54, 55, 59M/F, 60, 61M/F, 62, 65M/F, 67M/F, 68, 69, 72, 74, 77, 78, 79M/F, 80M/F, 83, 87M/F, 90M/F, 91F, 92, 94I,II, 98, 99, 101, 105, 106F, 110II, 112MII,III/FII,III, 113MIII,IV,VI/FIII,IV,VI, 114F, 116F, 117M/F, 121, 123, 125, 126M/F, 127M/F, 128, 130, 135, 136, 138I,II, 139, 140, 141, 142M/F, 145, 146M/F, 147, 149M/F11585.8
Longitudinal (length of study)11M/F (1 year), 42 (1 year), 82 (8 weeks), 107M/F (2 years), 118 (1 year), 137 (1 year), 148MI/FI (3 years)1011.02M/F (2.5 years), 20I (1 week), II (9 months), 26M/F (3 years), 40M/F (1 year), 45M/F (3 years), 93F (8 months), 102M/F (4 months), 133M/F (1 year), 148MII,III/FII,III (3 years)1914.2
Assessment of physical activity
Collection method
 Self-report11M/F, 25, 27M/F, 38, 41, 42, 46, 56M/F, 57, 58, 64, 75M/F, 84M/F, 85, 86M/F, 95M/F, 97, 103M/F, 104, 110I, 115, 118, 119, 120, 124M/F, 132, 137, 143M/F, 144, 148MI/FI4145.11M/F, 2M/F, 3M/F, 4M/F, 6, 7M/F, 8M/F, 9, 14, 16M/F, 17F, 18F, 20I,II, 26M/F, 30F, 33, 34M/F, 35F, 36, 40M/F, 43, 44M/F, 45M/F, 47, 48I/II, 49, 50M/F, 52, 53, 54, 55, 59M/F, 60, 61M/F, 62, 65M/F, 67M/F, 68, 69, 72, 74, 77, 78, 79M/F, 80M/F, 83, 87M/F, 91F, 92, 93F, 94I,II, 98, 99, 102M/F, 105, 106F, 110II, 114F, 116F, 117M/F, 121, 125, 126M/F, 127M/F, 128, 130, 133M/F, 135, 136, 139, 140, 141, 142M/F, 145, 146M/F, 147, 148MII,III/FII,III, 149M/F112 83.6
 Parent report15M/F, 21, 66, 73, 88, 100, 113MI,II/FI,II, 122, 129, 131MI,II/FI,II, 1501819.8   
 Composite: self- & parent report19, 76, 112MI/FI44.412, 29, 112MII,III/FII,III, 113MIII,IV/FIII,IV107.5
 Accelerometer5F, 13F, 37, 63M/F, 89M/F, 96, 113MV/FV, 134M/F1213.2113MVI,FVI, 12332.2
 Direct observation10/31, 28, 70, 71, 81, 82, 109, 11189.0   
 Doubly labelled water5111.1   
 Self-report & accelerometer/hear rate monitor32, 108M/F33.423M/F, 24M/F, 90M/F, 101, 138I,II96.7
 Parent-report & accelerometer3911.1   
 Composite: self- & parent-report & accelerometer107M/F22.3   
 Composite: two self-reports + fitness test22F11.1   
Reliability/validity of self- and parent reported methods
 Poor or unknown19, 21, 22F, 27M/F, 32, 38, 46, 56M/F, 57, 58, 66, 73, 88, 100, 103M/F, 118, 122, 129, 131MI,II/FI,II, 143M/F, 148MI/FI, 1503042.91M/F, 2M/F, 3M/F, 6, 7M/F, 8M/F, 9, 12, 14, 18F, 23M/F, 24M/F, 29, 34M/F, 36, 40M/F, 43, 44M/F, 45M/F, 54, 55, 60, 62, 65M/F, 67M/F, 68, 69, 72, 74, 77, 78, 87M/F, 92, 94I,II, 98, 105, 106F, 116F, 121, 125, 126M/F, 128, 130, 136, 139, 140, 141, 145, 148MII,III/FII,III 6851.9
 Acceptable11M/F, 15M/F, 25, 39, 41, 42, 64, 75M/F, 76, 84M/F, 85, 86M/F, 95M/F, 97, 104, 107M/F, 108M/F, 110I, 112MI/FI, 113MI,II/FI,II, 115, 119, 120, 124M/F, 132, 137, 1444057.14M/F, 16M/F, 17F, 20I,II, 26M/F, 30F, 33, 35F, 47, 48I,II, 49, 50M/F, 52, 53, 59M/F, 61M/F, 79M/F, 80M/F, 83, 90M/F, 91F, 93F, 99, 101, 102M/F, 110II, 112MII,III/FII,III, 113MIII,IV/FIII,IV, 114F, 117M/F, 127M/F, 133M/F, 135, 138I,II, 142M/F, 146M/F, 147, 149M/F 6348.1
Country
North America5F, 10/31, 11M/F, 13F, 21, 22F, 25, 28, 32, 37, 39, 41, 42, 46, 51, 57, 63M/F, 70, 71, 75M/F, 76, 81, 82, 84M/F, 85, 86M/F, 88, 89M/F, 95M/F, 96, 97, 100, 103M/F, 104, 107M/F, 108M/F, 109, 110I, 111, 112MI/FI, 113MI,II,V/FI,II,V, 115, 118, 120, 122, 124M/F, 132, 134M/F, 137, 1446874.74M/F, 6, 9, 16M/F, 17F, 18F, 20I,II, 26M/F, 29, 30F, 33, 34M/F, 35F, 36, 47, 48I/II, 49, 50M/F, 52, 53, 54, 55, 59M/F, 60, 61M/F, 62, 68, 69, 74, 77, 78, 79M/F, 80M/F, 87M/F, 90M/F, 91F, 92, 93F, 98, 99, 101, 102M/F, 106F, 110II, 112MII,III/FII,III, 113MIII,IV, VI/FIII,IV,VI, 114F, 116F, 117M/F, 121, 123, 125, 126M/F, 127M/F, 128, 133M/F, 135, 149M/F8563.4
Europe38, 56M/F, 64, 66, 73, 119, 129, 143M/F, 148MI/FI1213.51M/F, 2M/F, 3M/F, 7M/F, 8M/F, 12, 14, 23M/F, 24M/F, 40M/F, 44M/F, 45M/F, 65M/F, 67M/F, 72, 94I,II, 105, 130, 136, 138I,II, 139, 140, 141, 142M/F, 145, 146M/F, 147, 148MII,III/FII,III 4735.1
Oceania15M/F, 19, 27M/F, 58, 131MI,II/FI,II, 1501112.443, 832 1.5

Categorization of variables

Correlates of PA investigated in the studies reviewed were categorized in the ANGELO grid, i.e. were grouped in four environment types (physical, socio-cultural, economic, and policies) for each environmental setting (micro and macro) with a further distinction in specific levels (home, educational institution, neighbourhood, city/municipality, region). These data were then summarized in two tables providing an overview of the potential determinants of PA of children and adolescents separately (Tables 2 and 3 respectively).

Table 2.  Summary of correlates of physical activity among children (3–12 year olds)
CorrelateRelated to PA Biblio. no.Assoc. (+ or –)Unrelated to PA Biblio. no.♯ samplesSummary (n)
+0Assoc.
  1. Biblio. no., reference number under the Bibliography section; Assoc., association; +, positive; –, negative; 0, no relation; ?, indeterminate; N/A, summary code not applicable because the number of independent samples investigating the relationship is below 3; PA, physical activity; M, boys only; F, girls only; SES, socioeconomic status; studies with prospective study designs are highlighted in bold.

MICRO-ENVIRONMENT
Home/household
Physical
 ♯ cars in household19, 131FI131MI,II/FII5 –230
 Access/availability of exercise equipment30F, 124F, 134F+30F, 81, 97, 109, 124M, 132, 134F, 134M/F, 123 –900
Socio-cultural
 Single-parent family103F, 108M, 113MI+95M/F, 95M/F, 103M, 107M/F, 108F, 108M/F, 112MI/FI, 113MII,V/FI,II,V203 –1700
 ♯ household residents/children  38, 95M/F, 95M/F, 113MI,II,V/FI,II,V11 – –1100
 Acculturation (language spoken at home; lifetime in the county; index)11M, 95M/F
19, 137
+
11F, 11M/F, 13F, 95M/F, 137123270
 Dog ownership  131MI,II/FI,II4 – –40
 Parents’ PA32, 63M, 89M/F, 100, 107M, 111, 112MI, 144, 150,
124F
+
11M/F, 25, 32, 63F, 107F, 108M/F, 108M/F, 112FI, 113MI,II,V/FI,II,V, 124M291011800
 Father’s PA22F, 38, 39, 39, 46, 89M/F, 95M, 119M/F, 134M, 148MI/FI, 148MI/FI+15M/F, 15M/F, 84M/F, 95M/F, 95F, 97, 110I, 134M/F, 134F2915 –14+?
 Mother’s PA15F, 38, 39, 39, 95F, 110I, 124F, 134M, 148FI, 148FI+15M/F, 15M, 22F, 84M/F, 89M/F, 95M, 95M/F, 97, 109, 119M/F, 124M, 134M/F, 134F, 148MI, 148MI3110 –2100
 Sibling’s PA110I+ 11 – –N/A
 Friend’s PA46+97, 134M/F, 134M/F61 –50
 PA from significant others (parents, siblings, friends)  411 – –1N/A
 Encouragement from parents71, 82, 95F, 95F, 107M, 144+11M/F, 63M/F, 70, 95M, 95M, 95M/F, 95M/F, 107F, 108M/F, 108M/F 226 –1600
 Support (logistic) from parents (transports child to play, plays with child, pays fees)5F, 22F, 107M, 107M, 107M, 108M/F, 108M, 144, 144+5F, 22F, 63M/F, 63M/F, 70, 107F, 107F, 107F, 108F, 108M/F, 108M/F, 112MI/FI27101700
 Support/encouragement from significant others (family, peers, teachers)112MI/FI, 113MI/FI,V, 115, 115, 120, 124M+41, 97, 97, 109, 113MI,II,V/FI,II,V, 113MII,V/FII, 124F249 –150?
 Social norms (value/enjoyment of PA of significant others – parents, siblings, peers)25, 75M/F, 75M/F, 134M, 134M, 150+25, 25, 41, 84M/F, 97, 100, 112MI/FI, 112MI/FI, 124M/F, 129, 129, 134F, 134F258 –1700
Economic
 Parental SES27F, 27M/F, 32, 72, 88, 95F, 122,
58
+
27M, 27M/F, 32, 71, 72, 95M, 95M/F, 103, 109, 119, 1372281130?
 Parental occupational status148FI,
19
+
123, 148MI, 148MI/FI61140
 Father occupational status  11M/F, 95M/F, 95M/F6 – –60
 Mother occupational status11F, 56M/F+11M, 95M/F, 95M/F72 –50
 Parental education63M, 112MI
108F
+
37, 37, 37, 38, 46, 63F, 96, 103, 107M/F, 108M, 108M/F, 112FI, 113MI,II,V/FI,II,V, 13724212100
 Father’s education level56M/F, 95F, 148MI, 148MI,
57
+
95M, 95M/F, 148FI, 148FI11515??
 Mother’s education level148MI+19, 46, 95M/F, 19, 46, 95M/F, 131MI,II/FI,II, 148FI12 1 –1100
 ♯ hours parents work  95M/F, 108M/F, 108M/F, 1507 – –70
 House owned  191 – –1N/A
Policy
 Time spent outdoors10&31, 70, 81, 109, 109+ 55 – – +
 Parenting styles (PA rules, control)109, 10943, 109, 112MI/FI6 240
Educational institutions (schools,…)
Physical
 Distance (from home)150 1 –1 –N/A
 Availability of PA equipment  811 – –1N/A
Socio-cultural
 Teacher’s PA  1001 – –1N/A
 Teacher’s attitudes toward PA  1001 – –1N/A
 Teachers specific education level28, 100+ 22 – –N/A
Economic
 School type attended (public vs. private; nursery vs. day care)19, 100+ 22 –N/A
Policy
 Support from community PA organizations  281 – –1N/A
 PA-related policies (e.g. time allowed for free play/spent outside, ♯ field trips)28, 81, 96+28, 2853 –2 +
 Class size28+ 11 – –N/A
 School quality  281 – –1N/A
Neighbourhood
Physical
 Distance to destinations58 1 –1 –N/A
 Access/availability to PA facilities/programmes41, 109, 131FII+5F, 5F, 30F, 30F, 131MI,II/FI,II, 131MI,II/FI, 113MI,II,V/FI,II,V203 –1700
 Available shelters/foot path conditions  150, 1502 – –2N/A
 Neighbourhood hazards (e.g. many roads/no lights crossings; heavy traffic; physical disorder; pollution)123,
58, 131MII, 131MII, 131MI
+
58, 88, 113MI,II,V/FI,II,V, 131MI/FI,II, 131MI/FI,II 131MII/FI,II, 150, 15024141900
 Neighbourhood physical disorder  881 – –1N/A
 Limited public transport131FI,II131MI,II4 22?
Social
 Neighbourhood social disorder88 11 – –N/A
 Involvement in community PA organizations132, 134M, 134M+11M/F, 132, 134F, 134F83 –50
 Length of residence in community30F+30F21 –1N/A
 Safety885F, 5F, 107M/F, 113MI,II,V/FI,II,V, 131MI,II/FI,II, 15016 –11500
Economic
 Neighbourhood SES/education level143F, 64, 64+30F, 30F, 88, 143M,61230
      
MACRO-ENVIRONMENT
City/municipality/regions
Physical
 Urban vs. suburban2166211 –N/A
 Urban vs. rural27M/F, 56M/F, 66, 72, 72, 118
27M/F, 27M/F
+
46, 57, 85, 86M/F17845??
 Coastal vs. mountains46 11 – –N/A
 Season (spring, summer)
10&31, 118
42, 51, 100, 118+
21, 37, 37, 3710424??
Table 3.  Summary of correlates of physical activity among adolescents (13–18 years olds)
CorrelateRelated to PA Biblio. no.Assoc. (+ or –)Unrelated to PA Biblio. no.♯ samplesSummary (n)
+0Assoc.
  1. Biblio. no., reference number under the Bibliography section; Assoc., association; +, positive; –, negative; 0, no relation; ?, indeterminate; N/A, summary code not applicable because the number of independent samples investigating the relationship is below 3; PA, physical activity; M, boys only; F, girls only; SES, socioeconomic status; PE, physical education; studies with prospective study designs are highlighted in bold.

MICRO-ENVIRONMENT
Home/household
Physical
 Access/availability of PA equipment18F, 18F, 33,
24F
+
18F, 23M/F, 23M/F, 24M, 24M/F,
26M/F, 93F, 93F, 133M/F, 133M/F
20311600
Socio-cultural
 Single-parent family29, 76, 113MIV,
76, 130
+
45M/F, 61M/F, 67M/F, 76, 112MII,III/FII,III,
113MIII,VI/FIII,IV,VI, 128, 142M/F
24321900
 ♯ household residents/children  61M/F, 113MIII,IV,VI/FIII,IV,VI, 142M/F, 149M/F12 – –1200
 Acculturation (adolescent/parent born abroad; generation of residence in country)45F, 52+45M, 52, 53, 116F62 –4 0
 Parents’ PA33, 54, 98, 99, 142M/F+17F, 26M/F, 68, 79M/F, 79M/F, 90M/F, 90M/F, 112MII,III/FII,III, 113MIII,IV,VI/FIII,IV,VI, 135, 149M/F 316 –2500
 Father’s PA23M, 24F, 48I, 49, 98, 105, 110II, 140, 140, 141, 142M/F, 148MII,III+3M/F, 23F, 23M/F, 24M, 24M/F, 48II, 48II, 49, 133M/F, 133M/F, 148FII,III3114 –170?
 Mother’s PA3F, 23F, 48I, 49, 98, 106F, 110II, 133F, 142M/F, 148FII,III+3M, 23M, 23M/F, 24M/F, 24M/F, 26M/F, 48II, 48II, 49, 105, 133M, 133M/F, 140, 141, 148MII,III3312 –2100
 Sibling’s activity3M/F, 98, 99, 110II, 141+23M/F, 23M/F, 24M/F, 24M/F, 110II, 140, 140, 141186 –1200
 Friend’s PA24M, 33, 116F, 140, 140+17F, 23M/F, 23M/F, 24F, 24M/F, 133M/F, 133M/F, 141, 149M/F205 –1500
 PA from significant others (parents, friends, other adults)8M/F, 9, 18F, 24M, 102F, 126M+14, 18F, 18F, 24M/F, 24F, 102M, 126F, 141167 –90?
 Support/encouragement from parents8M/F, 18F, 18F, 29, 44M, 61M/F, 68, 79M/F, 79M/F, 90F, 112MII,III/FII,III, 112MII,III, 113FIII, 114F, 114F, 135, 139, 149F+17F, 18F, 18F, 18F, 18F, 44F, 45M/F, 90M, 90M/F, 90M/F, 90M/F, 101, 101, 101, 113MIII,IV,VI/FIV,VI, 112FII,III, 149M5226 –26??
 Support/encouragement from friends44F, 83, 101, 113MIII,VI, 149M+17F, 44M, 101, 113MIV/FIII,IV,VI, 139, 149F156 –900
 Support/encouragement from significant others8M/F, 12, 14, 18F, 18F, 24F, 24F, 44M/F, 93F, 114F, 114F+18F, 24M, 24M, 60, 93F, 133M/F, 133M/F, 1392313 –10+?
 Social norms (value/enjoyment of PA of significant  others – parents, siblings, peers)9, 26F, 47, 47, 48I/II, 80M/F, 80M/F, 87M, 91F, 91F, 112FIII, 123, 127M/F+8M/F, 16M/F, 16M/F, 17F, 26M, 47, 68, 68, 69, 79M/F, 87F, 112MII,III/FII, 112MII,III/FII,III, 114F, 114F, 1234217 –250?
Economic
 Parental SES9, 12, 18F,18F, 121, 145, 147+4M/F, 7M/F, 7M/F, 17F, 18F, 48II, 76, 76, 76, 128207 –1300
 Occupational status of household head45F, 65M/F, 73, 140+45M, 67M/F, 140, 141105 –5??
 Father’s occupational status54, 94I, 136+2M/F, 94II, 148MII,III/FII,III, 148MII/FII123 –900
 Mother’s occupational status2M, 94I+2F, 94II, 13652 –30
 Parents’ educational level74, 77, 112MIII, 117F, 142M/F+61M/F, 112MII/FI,II,III, 113MIII,IV,VI/FIII,IV,VI, 117M196 –1300
 Father’s education level136+48II, 148MII,III/FII,III, 148MII/FII81 –70
 Mother’s education level53, 92, 136+48II, 116F53 –2 +
 Family (per capita) income29, 53, 74, 77, 142M/F+50M/F, 60, 73106 –4++
 ♯ parents working full time117M+117F21 –1N/A
 Adolescent’s paid work/pocket money34M/F, 83, 141+53, 116F, 125, 149M/F94 –50?
Policy
 Parenting styles (authoritative; PA rules)87M, 90M, 117M87F, 90F, 90M/F, 112MII,III/FII,III, 117F12 3900
Education institutions (childcare, schools)
Physical
 School facilities/resources33+ 11 – –N/A
Socio-cultural
 Main teacher’s/coach PA90M+90M/F, 90F, 140, 140, 141, 149M/F91 –80
 Support from teacher/coach45M, 83, 149F+45F, 149M/F, 149M73 –40?
 Classmates problems/teasing45F, 45F45M, 45M4 22?
 School support  44M/F, 833 – –30
 Relationship with PE teacher  331 – –1N/A
Economic
 Public vs. private school34M/F 2 – 2 –N/A
Policy
 School type (high school vs. vocational/alternative)1M/F, 2M/F, 7M/F, 55, 55, 55, 55
40M
+
7M/F, 40F, 551510 14++
 School provide (special) PE programme/sport teams53+36, 133M/F, 133M/F61 –50
 Instruction on sport/health benefits140+140, 141, 14141 –30
Neighbourhood
Physical
 Distance to PA facilities50M50F2 11N/A
 Access/availability to PA equipment/facilities/programmes33, 44M/F, 61M/F, 61M/F, 61F, 61M/F, 113FVI, 29, 29+
17F, 23M/F, 23M/F, 24M/F, 24M/F, 61M/F, 90M/F, 112MII,III/FII,III, 113MIII,IV,VI/FIII,IV,VI, 113MIII,IV,VI/FIII,IV, 125, 149M/F, 61M451123200
 Level of urbanization  741 – –10
 Dogs unattended  90M/F, 90M/F4 – –40
Socio-cultural
 % married couples  67M/F2 – –2N/A
 % youth  67M/F2 – –2N/A
 Neighbourhood exercisers  90M/F, 90M/F, 149M/F6 – –60
 Social disorganization  741 – –1N/A
 Ethnic minority concentration  741 – –1N/A
 Crime incidence50F, 5350M3 –21 –
 Safety50F50M, 90M/F, 90M/F, 113MIII,IV,VI/FIII,IV,VI, 149M/F14 –11300
Economic
 SES  741 – –1N/A
 % upper occupational status  67M/F2 – –2N/A
 % owner occupied housing67F+67M21 –1N/A
 % dwellings provided by employer  67M/F2 – –2N/A
 % unemployment among residents67F67M2 – 11N/A
 Length of unemployment   67M/F2 – –2N/A
MACRO-ENVIRONMENT
City/municipality/region
Physical
 Urban vs. suburban67F67M2 11N/A
 Town size73+ 11 – –N/A
 Urban vs. rural140+35F, 53, 140, 14151 –40
 Season (Spring, Summer)20II, 138I+53, 138II42 –2?
 Unsuitable weather20I+12521 –1N/A
Socio-cultural
 Exposure to/interest in sports media62, 62+17F, 26M/F52 –30
 Wanting to look like media figures127M/F+ 22 – –N/A

Coding and summarizing associations with physical activity

A variety of statistical techniques (e.g. correlations, t-tests, linear or logistic regression analyses, anova and structured equation models) were used to evaluate the associations. Most studies not only reported univariate but also multivariate analyses (e.g. with adjustment for demographic and/or other potential correlates investigated); whenever possible findings reported here were those from the fully adjusted models. As with regard to prospective studies, the associations found within the shortest follow-up period (where the likelihood of associations was higher) were the ones considered, and the cross-sectional findings embedded within these studies were disregarded. Studies reporting positive (coded as ‘+’) or inverse (coded as ‘–’) association(s) between the independent variable and PA were registered under the column ‘related to PA’; non-significant associations were coded under the column ‘unrelated to PA’ (coded as ‘0’). Findings for each independent variable were summarized by adding the number of associations in a given direction (+, –, 0); keeping in line with the earlier reviews (45,52), a final summary association code for each correlate examined was derived as follows: ≥60% of the associations in any direction was considered evidence for a positive (summary code ‘+’), negative (summary code ‘–’) or non-association (summary code ‘0’); a mixed pattern of associations <60% (but above 50%) was considered evidence for probable but inconsistent association (summary code ‘+?’ or ‘–?’ or ‘0?’); a variable that has been frequently studied (i.e. in ≥10 independent samples) but with considerable lack of consistence was attributed a summary code of two questions marks (??); where findings were consistent, the codes ‘++’, ‘––’ or ‘00’ were attributed. Final summary codes were only computed for variables that have been studied in at least three independent samples; otherwise a ‘non-applicable’ (N/A) summary code was attributed.

Results

General characteristics of the studies reviewed (Table 1)

We have identified a total of 150 publications that presented an empirical association between PA and at least one environmental correlate. The vast majority of studies (71.3%) were published in the last decade (Fig. 1) and a steep, almost threefold, increase in adolescent studies was noticed in the last 5 years. The overall studies reported data on 225 independent samples. Sixty-six studies (91 independent samples) of children were reviewed, representing 40.4% of the total independent samples; only 16 (17.6%) of those independent samples included more than 1000 subjects. Eighty-four studies of adolescents (134 independent samples; 59.6%) were reviewed (four of which provided also data on children); about one-third included more than 1000 subjects. In both children and adolescents, the vast majority of the studies used a cross-sectional design, reported results for boys and girls separately, relied on child and/or parental self reports as method of PA data collection (about half of which with acceptable reliability/validity), and were mostly conducted in North America. Studies that used objective methods of PA assessment were in the great majority restricted to studies among children; direct observation and doubly labelled water assessment were never used in studies of adolescents.

Figure 1.

Distribution of the 150 publications by year of publication (1980–2004).

Potential environmental determinants of children’s physical activity (Table 2)

Potential determinants at the home level

We have identified a total of 17 independent samples investigating the associations between variables of the home physical environment, namely the amount of cars in the family and the availability and access of exercise equipment (e.g. PA promoting toys), and PA levels of children. Both variables were unrelated to children’s PA. Socio-cultural environmental correlates of children’s PA at the home/family level were the most frequently investigated. Family structure variables such as single-parent family, household size or number of children in the family, dog ownership and level of acculturation to the country of residence, were unrelated to children’s PA. Modelling of PA from parents, siblings and friends were extensively examined (96 independent samples in total). Studies that have examined the relationship between children’s PA levels and those of their parents, not separating those of the father from those of the mother, as well as those from significant others (e.g. parents, siblings or friends), found no relevant associations. However, in studies where father’s and mother’s PA levels were separated from each other, father’s PA levels emerged as a probable positive correlate (in 52% of the cases), whereas mother’s PA levels were mostly unrelated to children’s PA. Studies investigating potential familial influences other than modelling, namely support, encouragement and PA-related social norms of parents, friends and significant others, have also been numerous (a total of 98 independent samples). These variables were generally unrelated to children’s PA. The economic environment of children’s home/family in relation to their PA levels was studied in 102 independent samples. Different estimates of family/parental socioeconomic status (SES) were generally unrelated to children’s PA. Finally, and within the household’s PA-related ‘policy environment, time spent outdoors was consistently associated with higher PA levels of children, whereas parenting styles were unrelated to children’s PA.

Potential determinants at the school level

Aspects of the school environment were studied seldom (most of them only once or twice, which has not enabled us to calculate a summary association). Only one aspect of the school policy environment– PA policies (i.e. time allowed from free play, time spent outdoors, and number of field trips) – was investigated in three or more independent samples, with 60% of the cases showing a positive association with children’s PA levels.

Potential determinants at the neighbourhood level

A total of 90 independent samples have examined associations between environmental characteristics at the neighbourhood levels and PA levels of young children. We have identified a total of six potential correlates of PA at the neighbourhood physical environment, half of which studied more than three times. Among these, availability and accessibility of PA programmes or facilities, neighbourhood safety and neighbourhood hazards (e.g. many roads, no lights crossings, heavy traffic, physical disorder and pollution – estimated as perceived by parents in almost all studies) were consistently unrelated to children’s PA. Aspects of the social and economic environments were unrelated to children’s PA.

Potential determinants at the city/municipality and region/country level

Only few studies have investigated differences in PA levels between children living in urban vs. suburban (only examined twice) and coastal vs. mountainous locations (only examined once). Whether residence in urban vs. rural regions is associated with children’s PA levels was undetermined by the available studies. Seasonal ‘effects’ on children’s PA were also undetermined by the available literature.

Potential environmental determinants of adolescents’ physical activity (Table 3)

Potential determinants at the home level

We have identified a total of 20 independent samples investigating the associations between variables of the home physical environment, namely the availability and accessibility of exercise equipment, and PA levels of adolescents; these variables were mostly unrelated to adolescents’ PA. Socio-cultural environmental correlates of adolescents’ PA at the home/family level were the most frequently investigated. Family structure variables such as single-parent family and household size or number of children in the family were unrelated to adolescents’ PA as were indicators of acculturation. Modelling of PA from parents, siblings and friends were extensively examined (in 149 independent samples). Overall, all these studies found no relevant associations. However, this lack of associations was somewhat undetermined with regard to father’s PA levels and those from significant others, because they were observed in less than 60% of the cases. Studies investigating potential familial influences other than modelling were also numerous (a total of 132 independent samples) but mostly unrelated to adolescents’ PA. However, a trend towards a positive association was found with regard to general support from significant others. The relationship between the economic environment of adolescents’ home/family and their PA levels was examined in 100 independent samples. Studies in which parental SES was defined as a composite of parent’s education and income levels/occupational status were generally unrelated to children’s PA. However, studies in which the specific association between parent’s education levels was analysed separately from parent’s occupational status or income level revealed that higher mother’s education levels and family (per capita) income were positively associated with PA; occupational status of the household’s head emerged as an undetermined correlate of PA. With regard to the policy environment, parenting styles were unrelated to adolescents’ PA.

Potential determinants at the school level

Similarly to what we have described in children, aspects of the school physical, socio-cultural, economic or policy environment were studied relatively seldom in adolescents. Regarding the socio-cultural environment, role modelling and support from teachers were generally unrelated to adolescents’ PA, whereas the existence of problems with (or teasing from) classmates was undetermined. Finally, the type of school attended, namely high vs. vocational school, was positively associated, whereas the provision of instruction on PA or sport-related health benefits and special PE programmes and/or school sports, were unrelated to adolescents’ PA.

Potential determinants at the neighbourhood level

A total of 92 independent samples have examined associations between environmental characteristics at the neighbourhood level and PA levels of adolescents. Although we have identified a wide range of potential correlates at the physical, socio-cultural and economical level, only few were examined in more than three independent samples. Among these, and within the physical environment, the availability and/or accessibility of PA equipment or facilities, was unrelated to PA. Within the socio-cultural environment, crime incidence (measured objectively) was inversely associated with adolescents’ PA in two out of the three studies available, a finding that was at odds with the lack of association between adolescents’ PA and neighbourhood safety estimates perceived by them.

Potential determinants at the city/municipality and region/country level

Only few studies have investigated differences in PA levels between adolescents’ residence location. Residence in urban vs. rural regions was not associated with adolescents’ PA. Seasonal ‘effects’ on adolescents’ PA were undetermined and exposure to or interest in sports media was not associated with adolescents’ PA.

Discussion

Overall, the current review of the literature on environmental correlates of PA in children and adolescents provided us with a broader and more detailed overview of the specific research performed through the course of the past 25 years. In the past 5 years in particular an increased attention to this field was observed, which may reflect a paradigm shift from intra-personal to ecological conceptual models in the study of health-related behaviours such as PA.

Updating the earlier review: current vs. previous findings

We have updated the review of Sallis et al. (45) by merging 51 of its original studies (those reporting on environmental potential determinants of PA, as defined in the present study) with 99 additional publications; 23 of the 99 additional studies had not been included in the previous review although they were published within the same period covered by it (1970–1998); interestingly half of those studies (12 out of 23) were performed in Europe, a region that may have thus been under-represented in the earlier review. With regard to the main findings, a comparative summary between the two reviews is presented on Table 4. In children, time spent outdoors remained a main correlate of children’s PA, although this was due to the fact that no additional studies in this regard were included in the present review. In addition, time spent outdoors may reflect (although not necessarily) a PA-related behaviour, which explains the associations found. The correlates of children and adolescents’ PA that have emerged in the present review differ considerably from those in the previous review. Overall, we can argue that the additional publications, of which 76 were published in the last 5 years, have thus contributed significantly to a better understanding of factors associated with the PA behaviours of children and adolescents, and have led to the identification and addition of new potential determinants to the body of knowledge in the field. However, the fact that the associations coded and summarized in our review were those derived, whenever possible, from multivariate rather than from univariate analyses may also have contributed to the differences between the two reviews. The previous review, which drew exclusively from univariate models, may have thus been somewhat inflated (as significant correlates are generally more abundant in univariate analyses).

Table 4.  Comparative summary of the main environmental correlates of physical activity in children and adolescents: earlier vs. current review
ChildrenAdolescents
Previous reviewCurrent reviewPrevious reviewCurrent review
  1. PA, physical activity; +, positive association; –, inverse association; ?, indeterminate.

Programme/facility access (+)Father’s PA (+?)Support from significant others (++)Support from significant others (+?)
Time spent outdoors (+)School PA-related policies (+)Parent support (++)Mother’s education level (+)
Time spent outdoors (+)Sibling PA (++)Family income (++)
Direct help from parents (+)Non-vocational school (++)
Opportunities to exercise (+)Neighbourhood crime incidence (–)

Home/family correlates of children’s and adolescents’ physical activity

Characteristics of the home environment, particularly those related to parental influences, were by far the most explored in the literature, in both children and adolescents.

Parents as role models

Research findings regarding the relationship between PA levels of parents and those of their children have been mixed. Most of the studies have in fact failed to find any association. Nevertheless, fathers appear to be more important role models as compared with mothers, especially in childhood (references no. 22,38,39,46,89,95, 119,134 and 148 of the bibliography section); fathers’ PA may be related to their child’s PA regardless of their gender, whereas mothers’ PA appears to be more often associated with girls’ (references no. 3,15,23,95,106,124,133 and 148) rather than boys’ PA; however, parents’ PA has been generally unrelated to children’s future PA levels (as could be ascertained by the few prospective studies examining this issue).

In samples of children, parental PA levels were almost always assessed by the parents themselves (self-reports) whereas in the adolescent samples they were assessed by both adolescents’ reports (‘perceived’ parental PA levels) and parents’ self-reports. It is thus possible that differences in the agent reporting on parental PA levels (parent vs. offspring) may explain some of the lack of associations found. Indeed, there is some evidence that a low agreement exists between parents and children reports with regard to the levels of parental PA (59), and we have noticed that associations between PA of mothers or fathers and those of their offspring (adolescents) tended to be more often positive when the mothers or fathers reported their own level of PA (Table 5).

Table 5.  Analyses of the review findings regarding the association between physical activity (PA) levels of parents and their offspring (adolescents) according to the agent reporting on parental PA levels
 AssociationChi-squared (P value)
+0
  1. Data are number of independent samples (bibligraphy ♯).

(a) studies examining parental associations (total of 31 independent samples)
Assessment of parents’ PA
Parent self-report4 (98, 99, 142M/F)16 (68, 79 M/F, 79M/F, 112 MII,III/FII,III, 113MIII,IV,VI/FII,III)0.02 (0.90)
Perceived by the child2 (33, 54)9 (17F, 26M/F, 90M/F, 90M/F, 149M/F) 
(b) studies examining paternal associations (total of 31 independent samples)
Assessment of father’s PA
Father’s self-report7 (98, 105, 110II, 142M/F, 148 MII,III)4 (3M/F, 148 FII,III)2.35 (0.13)
Perceived by the child7 (23M, 24F, 48I, 49, 140, 140, 141)13 (23F, 23M/F, 24M, 24M/F, 48II, 48II, 49, 133M/F, 133M/F) 
(c) studies examining maternal associations (total of 33 independent samples)
Assessment of mothers’ PA
Mother’s self-report7 (3F, 98, 110II, 142M/F, 148FII,III)6 (3M, 26M/F, 105, 148MII,III)2.83 (0.09)
Perceived by the child5 (23F, 48I, 49, 106F, 133F)15 (23M, 23M/F, 24M/F, 24M/F, 48II, 48II, 49, 133M, 133M/F, 140, 141) 

Parental support, encouragement and beliefs

It has been hypothesized that the support and encouragement parents provide, rather than their own PA behaviour, may influence the PA behaviour of their offspring. In the present review, and with regard to children, these potential influences were reflected by the beneficial impact the time parents allow their children to spend outdoors seems to have on their PA, despite such beneficial impact was not found in relation to parental support and encouragement. However, as many studies have shown parental support to be positively or not to be associated with PA levels of adolescents. Taken together, these findings lend some support to the view that parents may need to be more than just active role models if their child is to lead a physical active lifestyle (32,60). This is supported by several (school-based) risk-reduction programmes that have included and evaluated (generally positively) parental involvement as a means to enhance programme effectiveness [e.g. The San Diego Family Heart Project (61); the Children and Adolescent Trial for Cardiovascular Health (CATCH I and II) (62,63); The Minessota Home Team (64,65)].

Parental socioeconomic status

Parental/family SES is associated with a wide array of health, cognitive and socio-emotional outcomes in children, throughout their development from (even before) birth to adulthood (66,67). In the studies reviewed herein, several measures of SES have been used, most including some quantification of family income, parental education and occupational status (or a combination of these). Mother’s education level and family income emerged as independent correlates of adolescents’ (but not children’s) PA levels. These findings not only emphasize the need to separate such aspects as education, occupational status and income levels, but also suggest that on reaching adolescence and young adulthood, those who have lower income may be more restricted in their PA choices and opportunities. In younger children, PA is mostly of informal nature, and may therefore not involve much extra financial cost. Possibly, with increasing age participation in physical activities becomes more elaborate and financial costly (e.g. sport clubs fees), which may reduce the likelihood of PA in adolescents from lower income families (68). This needs further investigation.

School influences on children and adolescents’ physical activity

Schools offer many opportunities for young people to engage in physical activities, such as PE classes, recess periods, extracurricular sports or PA programmes, leisure time free use of its playing fields and playgrounds. Schools have also the personnel who, with sufficient training and commitment, can define and deliver PA programmes and policies that support the adoption of healthy lifestyles. The literature showing that well-designed and well-implemented school-based programmes can improve PA of young people is overwhelming (69–71), and guidelines for school programmes to promote lifelong PA actually exist (72–74). Despite this, little research has investigated specific features of the school environment that impact on youth PA. Indeed, although most studies reviewed herein have recruited their target populations from school settings, aspects of the school physical, socio-cultural, economic or political environment, remained relatively unexplored. Most of the characteristics of the school environment identified were almost never tested in more than 10 and often in less than three independent samples. Despite this, the present review has identified ‘school policies related to PA’ to be positively associated with children’s PA and ‘school type’ (i.e. attending high rather than vocational schools) to be a positive correlate of adolescent’s PA.

Additionally, we have identified an interesting set of studies that have investigated PA levels of classes/groups of youngsters in the context of PE lessons or recess time. One study found that classes of children taught by PE specialists (as compared with generalists) received longer as well as more very active lessons, leading to higher energy expenditure rates. Additionally, outdoor lessons generated more time spend in vigorous activities and higher total energy expenditures than indoor classes (75). In another study, school size, length of recess and the availability of balls in the playground were identified as correlates of higher engagement in PA by children (76). In adolescents, teacher’s speciality and gender were not associated with classes PA levels, neither was the location where the lessons were taught; the only significant correlates were class size and lesson-specific context (fitness activities, free play, game play and skill drills, management time, and knowledge) (inversely associated with class PA) (77). Another study found that, despite the availability of the PA facilities, they were used by very few students during their leisure time at school (i.e. before and after school classes, and during the lunch break) (78). These findings were then further explored and followed by the observation that not only the availability of PA facilities, but its size and state of conservation, and particularly the existence of supervision/organized activities, were decisive of adolescents’ engagement in PA during their leisure time at school (79). These findings and those of the present review, together with the observation that many schools are not providing enough time for physical activities (80,81), emphasize the important role school’s environments may play on children’s and adolescents’ PA levels (74,82). Further, school-based PA may represent an important equalizing factor for PA opportunities in children and adolescents of different SES backgrounds (83).

Neighbourhood influences on children and adolescents’ physical activity

Recently, the importance of neighbourhood physical and socio-cultural characteristics in shaping PA of individuals has been increasingly investigated, but relatively few studies in the current review have already addressed these possible associations. Among these studies, features of the physical environment (also commonly referred in the literature as the ‘built environment’), in particular the availability and accessibility to PA equipment, facilities or programmes were investigated more often, but were generally unrelated to youth PA. The present review did not identify any feature of the neighbourhood environment to be associated with children’s activity levels. In adolescents, crime incidence, as measured through objective police reports, was inversely associated with adolescents’ PA levels, a finding that apparently contrasted with the lack of association between perceived neighbourhood safety levels and adolescents’ PA. This contradiction suggests that the differential associations with youth PA may depend on the assessment method (perceived vs. objective) of environmental characteristics. Which features are more important remains unknown, an issue that therefore deserves further investigation (see Methodological considerations).

The importance of understanding neighbourhood effects on health-related behaviours relies on their potential to influence large populations (84,85). Although researchers are starting to address the potential effects of communities and neighbourhoods in individuals’ PA behaviour, few empirical studies have determined, using appropriate multilevel statistical techniques, whether relations between the environment and PA actually exist at the neighbourhood rather than the individual level (86,87).

Methodological considerations

Measurement of physical activity and environmental characteristics

The selection of an appropriate measurement instrument depends on the specific research question(s) to be addressed and on an ‘accuracy-practicality trade-off’ (88–90). The majority of the research on the potential determinants of PA reviewed herein relied on (parental or child/adolescent) self-reports, which included diaries and recall instruments; these methods may pose serious limitations because they provide less accurate estimates of PA levels than those obtained by more objective methods such as direct observation, motion sensors, heart rate monitors, and doubly labelled water (91). In addition, because the degree of the relationship between objectively and self-report measures of PA is only moderate, notably among self-report methods with ‘acceptable’ validity (92), there may be a substantial amount of variance not shared by the two methods; in other words, different instruments (objective vs. self-report) may measure different aspects of the PA behaviour, and therefore those measures are not interchangeable. As such, the correlates of PA may also differ as a function of the method used to measure the behaviour, thereby impairing the generalization of the findings obtained with the use of one or the other method (93). In the present review we were able to identify seven publications (10 independent samples – three in children and seven in adolescents, all with a cross-sectional design) which enable a more close examination of this issue, by providing self-report and objective data in the same samples (Table 6). In these studies, the magnitude of the associations between the two measures of PA was at the most moderate. Furthermore, clear discrepancies between correlates of objectively measured and self-reported PA levels were found. Several factors may explain these discrepancies. The proposed correlates investigated in each study may have more explanatory power for self-reported measures (e.g. number of vigorous activities) than for total PA levels (mostly computed by the objective measures). In addition, accelerometers, the most frequently used objective measure, are unable to access common activities such as bicycle riding and swimming that could have been (self-) reported, but captured incidental PA throughout the day, which in turn could have been forgotten on self-reports that usually refer specifically to intentional PA. Finally, there may be a shared method variance between self-reported PA and self-reported potential determinants, which then leads to an inflated association between the two.

Table 6.  Determinants of objective vs. self-reported physical activity – summary of findings
Bibliography no.Method of physical activity (PA) assessmentEnvironmental correlates of PA*
ObjectiveSelf-reportCorrelation between PA assessed by the two methodsObjectively measured PASelf-reported PA
  • *

    Only the environmental variables that were correlated with physical activity levels measured either by one or the other method are reported (i.e. listed variables do not cover all the variables investigated in each study).

  • PA, physical activity; M, boys only; F, girls only; SES, socioeconomic status.

23Heart-rate monitoring
(1 week; time spent in moderate-to-vigorous PA, i.e. >140 beats min−1)
Recall of PA and sport participations
(1 week; hours)
Not associated
(estimate not reported)
Father’s PA (M)Mother’s PA (F)
24Heart-rate monitoring
(1 week; time spent in moderate-to-vigorous PA, i.e. >140 beats min−1)
Recall of PA and sport participations
(1 week; hours)
Not associated
(estimate not reported)
Father’s PA
(F) Friends’ PA (M)
Parental encouragement (F)
Parental support (F)
Home equipment (F)
32Accelerometer
(2 week days + 1 weekend day; METs)
Frequency, duration and type of PA
(2 week days + 1 weekend day; METs)
r = 0.46Parental PA
Parental SES
39Accelerometer
(2 week days + 1 weekend day; counts d−1)
Frequency, duration and type of PA
(2 week days + 1 weekend day; METs)
r = 0.39 (light PA)
r = 0.35 (moderate-to-high intensity PA)
Father’s PA
Mother’s PA
Father’s PA
Mother’s PA
90Accelerometer
(up to 8 d; counts h−1)
PA record of hard and very hard intensity PA
(7 d; h week−1)
Not associated 
(estimate not reported)
Teacher’s PA (M)
PA rules (M)
Parent transports child to PA location (F)
101Accelerometer
(5-day period; min d−1)
Participation in PA for ≥60 min (PACE+)
(past week; d week−1)
r = 0.46Parent support
Peer support
110Accelerometer (1 week day + 2 weekend days; score)Recall checklist of PA performed for at least 15 min (1 week day + 1 weekend day; score)?
(Not reported)
Parental education (F)
Single-parent status (M)
Parent transports child to PA location (F)
Parent plays with child (M)

Furthermore, self-reports of environmental factors represent perceived rather than ‘real’ features of the physical, socio-cultural, economic and political environments. Little is know about the accuracy of such perceived features (94). In adults, some studies have shown objective environmental measures to be associated with PA whereas the same features measured through self-reports were not (95,96).

Limitations of study design and data analyses methodologies

The studies incorporated in the present review had mostly a cross-sectional design and therefore their findings were limited in that only association could be established and not prediction or ‘causation’. Nevertheless, all those studies have interpreted the results as if ‘causality’ existed and to be unidirectional (e.g. parents may influence their children). It is of course possible that reverse or reciprocal influences are operative as well (e.g. children influence their parents), an issue that needs to be further explored.

In an attempt to disentangle the information provided by prospective from cross-sectional studies we have highlighted those studies in Tables 2 and 3. However, their low number does not enable solid conclusions with regard to the potential environmental predictors of PA change.

The main question of how environmental features influence youth PA remained further largely unanswered because of the data analytical methods used. Conceptually, environmental influences can play a direct role in shaping PA behaviour or can be mediated by cognitive processes (97–99). In order to understand these mechanistic processes better data analytical methods (and study designs) are needed [for details see Bauman et al. (100)]. The majority of the findings reviewed herein were those that resulted from adjusted models (most often, for potential confounders such as age, sex and ethnicity, but in many studies for potential mediators such as self-efficacy and attitudes), and thus concern the ‘independent’ contribution of environmental characteristics in the explanation of PA behaviour.

Further, although most of the data included in the present review have an intrinsic multilevel structure, they were most frequently analysed as obtained in simple random samples of a single population. As such, the potential interdependence within clusters (e.g. classes within schools) has been ignored, which can have led to inflated estimation of the associations; multilevel or hierarchical analytic approaches are thus needed.

Limitations of the present review

We acknowledge several limitations of our current review. First, the search terms used to retrieve studies from existing databases may have not been sensitive enough. This is sustained by the fact that almost half of the studies included in this review were found through the literature sections of articles primarily retrieved in those databases. This may have been due to the fact that some articles included are simply not registered within those databases, and/or in many articles retrieved, environmental correlates of children/adolescents’ PA were not the primarily research goal but were embedded within a broader (i.e. health-enhancing behaviours in general) or related research question. Nevertheless, better search terms may still need to be defined. However, the vast amount of studies included suggests we have covered the existing literature in a quite satisfactory way. Second, the use of only English published data and the exclusion of studies from non-market economies may have discarded some studies that could have added relevant information into the field. Third, the main outcome was any form of PA. In most studies this was measured across several settings (e.g. the total amount of moderate-to-vigorous PA, performed at school and during leisure time – either at home or in the neighbourhood, or in sport clubs, accumulated throughout the day or the past week), not enabling us to determine the specific environmental correlates of specific physical activities. Fourth, the conceptual framework we have used may have led to disputable categorizations of the correlates of PA investigated.

Conclusions, implications and recommendations

Clearly, many factors influence the complex behaviour of youth PA. We have identified father’s PA habits, school PA-related policies and time spent outdoors as potential determinants of PA in children. In adolescents, support from significant others, mother’s education levels, family income, attendance of a non-vocational school and low neighbourhood crime incidence emerged as potential determinants of adolescents’ PA. These variables need to be targeted by multilevel interventions aiming at the increase of youth PA. The other variables, however, should not be discarded without further investigation, namely those whose associations with PA were undetermined or not possible to infer from the limited number of existing studies (particularly those at the neighbourhood and school settings as well as at the macro-environment level). In the light of a solution-orientated research paradigm for the prevention of childhood obesity the lack of clear information regarding the influence of schools and neighbourhoods or the media on PA should not refrain PA-promoting environmental policies and interventions to actually take place, because waiting for such evidence to emerge before action is taken may simply represent a delay, with foreseen costs, to the solution of the obesity epidemic (101). Future studies that use prospective or intervention designs enabling the analyses of whether the environment and PA behaviours of children and adolescents associations are casual and which (if any) cognitive processes may mediate or contextual variables may moderate such associations, are in great need. In addition, it is important to conduct future research with clear, possibly standardized, definitions and objective methods of environmental attributes and PA behaviour assessment, within the strongest study design possible.

Conflict of interest statement

No conflict of interest was declared.

Bibliography

* indicates studies included in the previous review of Sallis et al. (2000).

1.* Aarnio M, Kujala UM, Kaprio J. Associations of health-related behaviors, school type and health status to physical activity patterns in 16 year old boys and girls. Scand J Soc Med 1997; 25: 156–167.

2. Aarnio M, Winter T, Kujala U, Kaprio J. Associations of health related behaviour, social relationships, and health status with persistent physical activity and inactivity: a study of Finnish adolescent twins. Br J Sports Med 2002; 36: 360–364.

3. Aarnio M, Winter T, Kujala UM, Kaprio J. Familial aggregation of leisure-time physical activity – a three generation study. Int J Sports Med 1997; 18: 549–556.

4.* Aaron DJ, Kriska AM, Dearwater SR, Anderson RL, Olsen TL, Cauley JA, Laporte RE. The epidemiology of leisure physical activity in an adolescent population. Med Sci Sports Exerc 1993; 25: 847–853.

5. Adkins S, Sherwood NE, Story M, Davis M. Physical activity among African-American girls: the role of parents and the home environment. Obes Res 2004; 12: S38–S45.

6. Allison KR, Dwyer JJ, Makin S. Perceived barriers to physical activity among high school students. Prev Med 1999; 28: 608–615.

7. Andersen LB, Schelin B. Physical activity and performance in a random sample of adolescents attending school in Denmark. Scand J Med Sci Sports 1994; 4: 13–18.

8.* Anderssen N, Wold B. Parental and peer influences on leisure-time physical activity in young adolescents. Res Q Exerc Sport 1992; 63: 341–348.

9. Anthsel KM, Anderman EM. Social influences on sports participation during adolescence. J Res Dev Educ 2000; 33: 85–94.

10.* Baranowski T, Thompson WO, DuRant RH, Baranowski J, Puhl J. Observations on physical activity in physical locations: age, gender, ethnicity, and month effects. Res Q Exerc Sport 1993; 64: 127–133.

11. Barnett TA, O’Loughlin J, Paradis G. One- and two-year predictors of decline in physical activity among inner-city schoolchildren. Am J Prev Med 2002; 23: 121–128.

12. Baxter-Jones AD, Maffulli N. Parental influence on sport participation in elite young athletes. J Sports Med Phys Fitness 2003; 43: 250–255.

13. Beech BM, Kumanyika SK, Baranowski T, Davis M, Robinson TN, Sherwood NE, Taylor WC, Relyea G, Zhou A, Pratt C, Owens A, Thompson NS. Parental cultural perspectives in relation to weight-related behaviors and concerns of African-American girls. Obes Res 2004; 12: S7–S19.

14.* Biddle S, Goudas M. Analysis of children’s physical activity and its association with adult encouragement and social cognitive variables. J Sch Health 1996; 66: 75–78.

15. Bogaert N, Steinbeck KS, Baur LA, Brock K, Bermingham MA. Food, activity and family – environmental vs. biochemical predictors of weight gain in children. Eur J Clin Nutr 2003; 57: 1242–1249.

16. Bungum T, Dowda M, Weston A, Trost SG, Pate RR. Correlates of physical activity in male and female youth. Pediatr Exerc Sci 2000; 12: 71–79.

17.* Bungum TJ, Vincent ML. Determinants of physical activity among female adolescents. Am J Prev Med 1997; 13: 115–122.

18.* Butcher J. Socialization of adolescent girls into physical activity. Adolescence 1983; 18: 753–766.

19. Carlin JB, Stevenson MR, Roberts I, Bennett CM, Gelman A, Nolan T. Walking to school and traffic exposure in Australian children. Aust N Z J Public Health 1997; 21: 286–292.

20.* Crocker PR, Bailey DA, Faulkner RA, Kowalski KC, McGrath R. Measuring general levels of physical activity: preliminary evidence for the Physical Activity Questionnaire for Older Children. Med Sci Sports Exerc 1997; 29: 1344–1349.

21. Damore DT. Preschool and school age activities: comparison of urban and suburban populations. J Community Health 2002; 27: 203–211.

22. Davison KK, Cutting TM, Birch LL. Parents’ activity-related parenting practices predict girls’ physical activity. Med Sci Sports Exerc 2003; 35: 1589–1595.

23. Deflandre A, Lorant J, Gavarry O, Falgairette G. Determinants of physical activity and physical and sports activities in French school children. Percept Mot Skills 2001; 92: 399–414.

24. Deflandre A, Lorant J, Gavarry O, Falgairette G. Physical activity and sport involvement in French high school students. Percept Mot Skills 2001; 92: 107–120.

25.* Dempsey JM, Kimiecik JC, Horn TS. Parental influence on children’s moderate to vigorous physical activity participation: an expectancy-value approach. Pediatr Exerc Sci 1993; 5: 151–167.

26.* DiLorenzo TM, Stucky-Ropp RC, Vander Wal JS, Gotham HJ. Determinants of exercise among children. II. A longitudinal analysis. Prev Med 1998; 27: 470–477.

27. Dollman J, Norton K, Tucker G. Anthropometry, fitness and physical activity of urban and rural south Australian children. Pediatr Exerc Sci 2002; 14: 297–312.

28. Dowda M, Pate RR, Trost SG, Almeida M, Sirard JR. Influences of preschool policies and practices on children’s physical activity. J Community Health 2004; 29: 183–196.

29. Duncan SC, Duncan TE, Strycker LA, Chaumeton NR. A multilevel analysis of sibling physical activity. J Sport Exerc Psychol 2004; 26: 57–68.

30. Dunton GF, Jamner MS, Cooper DM. Assessing the perceived environment among minimally active adolescent girls: validity and relations to physical activity outcomes. Am J Health Promot 2003; 18: 70–73.

31.* DuRant RH, Baranowski T, Johnson M, Thompson WO. The relationship among television watching, physical activity, and body composition of young children. Pediatrics 1994; 94: 449–455.

32.* Epstein LH, Paluch RA, Coleman KJ, Vito D, Anderson K. Determinants of physical activity in obese children assessed by accelerometer and self-report. Med Sci Sports Exerc 1996; 28: 1157–1164.

33. Fein AJ, Plotnikoff RC, Wild T, Spence JC. Perceived environment and physical activity in youth. Int J Behav Med 2004; 11: 135–142.

34. Feldman DE, Barnett T, Shrier I, Rossignol M, Abenhaim L. Is physical activity differentially associated with different types of sedentary pursuits? Arch Pediatr Adolesc Med 2003; 157: 797–802.

35. Felton GM, Dowda M, Ward DS, Dishman RK, Trost SG, Saunders R, Pate RR. Differences in physical activity between black and white girls living in rural and urban areas. J Sch Health 2002; 72: 250–255.

36.* Ferguson KJ, Yesalis CE, Pomrehn PR, Kirkpatrick MB. Attitudes, knowledge, and beliefs as predictors of exercise intent and behavior in schoolchildren. J Sch Health 1989; 59: 112–115.

37. Finn K, Johannsen N, Specker B. Factors associated with physical activity in preschool children. J Pediatr 2002; 140: 81–85.

38. Fogelholm M, Nuutinen O, Pasanen M, Myohanen E, Saatela T. Parent-child relationship of physical activity patterns and obesity. Int J Obes Relat Metab Disord 1999; 23: 1262–1268.

39.* Freedson PS, Evenson S. Familial aggregation in physical activity. Res Q Exerc Sport 1991; 62: 384–389.

40.* Fuchs R, Powell KE, Semmer NK, Dwyer JH, Lippert P, Hoffmeister H. Patterns of physical activity among German adolescents: the Berlin-Bremen Study. Prev Med 1988; 17: 746–763.

41.* Garcia AW, Broda MA, Frenn M, Coviak C, Pender NJ, Ronis DL. Gender and developmental differences in exercise beliefs among youth and prediction of their exercise behavior. J Sch Health 1995; 65: 213–219.

42.* Garcia AW, Pender NJ, Antonakos CL, Ronis DL. Changes in physical activity beliefs and behaviors of boys and girls across the transition to junior high school. J Adolesc Health 1998; 22: 394–402.

43. Garton AF, Harvey R, Price C. Influence of perceived family environment on adolescent leisure participation. Aust J Psychol 2004; 56: 18–24.

44.* Gentle P, Caves R, Armstrong N, Balding J, Kirby B. High and low exercisers among 14- and 15-year-old children. J Public Health Med 1994; 16: 186–194.

45. Gillander Gadin K, Hammarstrom A. Can school-related factors predict future health behaviour among young adolescents? Public Health 2002; 116: 22–29.

46. Gilmer MJ, Harrell JS, Miles MS, Hepworth JT. Youth characteristics and contextual variables influencing physical activity in young adolescents of parents with premature coronary heart disease. J Pediatr Nurs 2003; 18: 159–168.

47.* Godin G, Shephard RJ. Normative beliefs of school children concerning regular exercise. J Sch Health 1984; 54: 443–445.

48.* Godin G, Shephard RJ. Psychosocial factors influencing intentions to exercise of young students from grades 7 to 9. Res Q Exerc Sport 1986; 57: 41–52.

49.* Godin G, Shephard RJ, Colantonio A. Children’s perception of parental exercise: influence of sex and age. Percept Mot Skills 1986; 62: 511–516.

50. Gomez JE, Johnson BA, Selva M, Sallis JF. Violent crime and outdoor physical activity among inner-city youth. Prev Med 2004; 39: 876–881.

51. Goran MI, Nagy TR, Gower BA, Mazariegos M, Solomons N, Hood V, Johnson R. Influence of sex, seasonality, ethnicity, and geographic location on the components of total energy expenditure in young children: implications for energy requirements. Am J Clin Nutr 1998; 68: 675–682.

52. Gordon-Larsen P, Harris KM, Ward DS, Popkin BM. Acculturation and overweight-related behaviors among Hispanic immigrants to the US: the National Longitudinal Study of Adolescent Health. Soc Sci Med 2003; 57: 2023–2034.

53. Gordon-Larsen P, McMurray RG, Popkin BM. Determinants of adolescent physical activity and inactivity patterns. Pediatrics 2000; 105: e83.

54.* Gottlieb NH, Chen MS. Sociocultural correlates of childhood sporting activities: their implications for heart health. Soc Sci Med 1985; 21: 533–539.

55. Grunbaum JA, Lowry R, Kann L. Prevalence of health-related behaviors among alternative high school students as compared with students attending regular high schools. J Adolesc Health 2001; 29: 337–343.

56.* Guillaume M, Lapidus L, Bjorntorp P, Lambert A. Physical activity, obesity, and cardiovascular risk factors in children. The Belgian Luxembourg Child Study II. Obes Res 1997; 5: 549–556.

57. Harrell JS, Gansky SA, Bradley CB, McMurray RG. Leisure time activities of elementary school children. Nurs Res 1997; 46: 246–253.

58. Harten N, Olds T. Patterns of active transport in 11–12 year old Australian children. Aust N Z J Public Health 2004; 28: 167–172.

59. Heath GW, Pratt M, Warren CW, Kann L. Physical activity patterns in American high school students. Results from the 1990 Youth Risk Behavior Survey. Arch Pediatr Adolesc Med 1994; 148: 1131–1136.

60. Higgins JW, Gaul C, Gibbons S, Van Gyn G. Factors influencing physical activity levels among Canadian youth. Can J Public Health 2003; 94: 45–51.

61. Hoefer WR, McKenzie TL, Sallis JF, Marshall SJ, Conway TL. Parental provision of transportation for adolescent physical activity. Am J Prev Med 2001; 21: 48–51.

62. Hofstetter RC, Hovell MF, Sallis JF, Zakarian J, Beirich H, Mulvihill M, Emerson J. Exposure to sports mass media and physical activity characteristics among ethnically diverse adolescents. Med Exerc Nutr Health 1995; 4: 234–242.

63.* Hovell MF, Kolody B, Sallis JF, Black DR. Parent support, physical activity, and correlates of adiposity in nine year olds: an exploratory study. J Health Educ 1996; 27: 126–129.

64. Hussey J, Gormley J, Bell C. Physical activity in Dublin children aged 7–9 years. Br J Sports Med 2001; 35: 268–272.

65. Huurre T, Aro H, Rahkonen O. Well-being and health behaviour by parental socioeconomic status: a follow-up study of adolescents aged 16 until age 32 years. Soc Psychiatry Psychiatr Epidemiol 2003; 38: 249–255.

66. Johansson B, Drott P. Informal parental traffic education and children’s bicycling behaviour. Ups J Med Sci 2001; 106: 133–144.

67. Karvonen S, Rimpela AH. Urban small area variation in adolescents’ health behaviour. Soc Sci Med 1997; 45: 1089–1098.

68. Kimiecik JC, Horn TS. Parental beliefs and children’s moderate-to-vigorous physical activity. Res Q Exerc Sport 1998; 69: 163–175.

69.* Kimiecik JC, Horn TS, Shurin CS. Relationships among children’s beliefs, perceptions of their parents’ beliefs, and their moderate-to-vigorous physical activity. Res Q Exerc Sport 1996; 67: 324–336.

70.* Klesges RC, Eck LH, Hanson CL, Haddock C, Klesges LM. Effects of obesity, social interactions, and physical environment on physical activity in preschoolers. Health Psychol 1990; 9: 435–449.

71.* Klesges RC, Malott JM, Boschee PF, Weber JM. The effects of parental influences on children’s food intake, physical activity, and relative weight. Int J Eat Disord 1986; 5: 335–346.

72. Kristjansdottir G, Vilhjalmsson R. Sociodemographic differences in patterns of sedentary and physically active behavior in older children and adolescents. Acta Paediatr 2001; 90: 429–435.

73. Lasheras L, Aznar S, Merino B, Lopez EG. Factors associated with physical activity among Spanish youth through the National Health Survey. Prev Med 2001; 32: 455–464.

74. Lee RE, Cubbin C. Neighborhood context and youth cardiovascular health behaviors. Am J Public Health 2002; 92: 428–436.

75. Lewko JH, Ewing ME. Sex differences and parental influence in sport involvement of children. J Sport Psychol 1980; 2: 62–68.

76. Lindquist CH, Reynolds KD, Goran MI. Sociocultural determinants of physical activity among children. Prev Med 1999; 29: 305–312.

77.* Lowry R, Kann L, Collins JL, Kolbe LJ. The effect of socioeconomic status on chronic disease risk behaviors among US adolescents. JAMA 1996; 276: 792–797.

78. Macintosh D. Socio-economic, educational and status characteristics of Ontario interschool athletes. Can J Appl Sport Sci 1982; 7: 272–283.

79. McGuire MT, Hannan PJ, Neumark-Sztainer D, Cossrow NH, Story M. Parental correlates of physical activity in a racially/ethnically diverse adolescent sample. J Adolesc Health 2002; 30: 253–261.

80. McGuire MT, Neumark-Sztainer DR, Story M. Correlates of time spent in physical activity and television viewing in a multi-racial sample of adolescents. Pediatr Exerc Sci 2002; 14: 75–86.

81.* McKenzie TL, Sallis JF, Nader PR, Broyles SL, Nelson JA. Anglo- and Mexican-American preschoolers at home and at recess: activity patterns and environmental influences. J Dev Behav Pediatr 1992; 13: 173–180.

82.* McKenzie TL, Sallis JF, Nader PR, Patterson TL, Elder JP, Berry CC, Rupp JW, Atkins CJ, Buono MJ, Nelson JA. BEACHES: an observational system for assessing children’s eating and physical activity behaviors and associated events. J Appl Behav Anal 1991; 24: 141–151.

83. McLellan L, Rissel C, Donnelly N, Bauman A. Health behaviour and the school environment in New South Wales, Australia. Soc Sci Med 1999; 49: 611–619.

84.* McMurray RG, Bradley CB, Harrell JS, Bernthal PR, Frauman AC, Bangdiwala SI. Parental influences on childhood fitness and activity patterns. Res Q Exerc Sport 1993; 64: 249–255.

85. McMurray RG, Harrell JS, Bangdiwala SI, Deng S. Cardiovascular disease risk factors and obesity of rural and urban elementary school children. J Rural Health 1999; 15: 365–374.

86. McMurray RG, Harrell JS, Bangdiwala SI, Gansky SA. Biologic and environmental factors influencing the aerobic power of children. Med Exerc Nutr Health 1995; 4: 243–250.

87. Mellin AE, Neumark-Sztainer D, Story M, Ireland M, Resnick MD. Unhealthy behaviors and psychosocial difficulties among overweight adolescents: the potential impact of familial factors. J Adolesc Health 2002; 31: 145–153.

88. Molnar BE, Gortmaker SL, Bull FC, Buka SL. Unsafe to play? Neighborhood disorder and lack of safety predict reduced physical activity among urban children and adolescents. Am J Health Promot 2004; 18: 378–386.

89.* Moore LL, Lombardi DA, White MJ, Campbell JL, Oliveria SA, Ellison RC. Influence of parents’ physical activity levels on activity levels of young children. J Pediatr 1991; 118: 215–219.

90. Morgan CF, McKenzie TL, Sallis JF, Broyles SL, Zive MM, Nader PR. Personal, social, and environmental correlates of physical activity in a bi-ethnic sample of adolescents. Pediatr Exerc Sci 2003; 15: 288–301.

91. Motl RW, Dishman RK, Ward DS, Saunders RP, Dowda M, Felton G, Pate RR. Examining social-cognitive determinants of intention and physical activity among black and white adolescent girls using structural equation modeling. Health Psychol 2002; 21: 459–467.

92. Murphey DA, Lamonda KH, Carney JK, Duncan P. Relationships of a brief measure of youth assets to health-promoting and risk behaviors. J Adolesc Health 2004; 34: 184–191.

93. Neumark-Sztainer D, Story M, Hannan PJ, Tharp T, Rex J. Factors associated with changes in physical activity: a cohort study of inactive adolescent girls. Arch Pediatr Adolesc Med 2003; 157: 803–810.

94. Nutbeam D, Aar L, Catford J. Understanding childrens’ health behaviour: the implications for health promotion for young people. Soc Sci Med 1989; 29: 317–325.

95. O’Loughlin J, Paradis G, Kishchuk N, Barnett T, Renaud L. Prevalence and correlates of physical activity behaviors among elementary schoolchildren in multiethnic, low income, inner-city neighborhoods in Montreal, Canada. Ann Epidemiol 1999; 9: 397–407.

96. Pate RR, Pfeiffer KA, Trost SG, Ziegler P, Dowda M. Physical activity among children attending preschools. Pediatrics 2004; 114: 1258–1263.

97.* Pate RR, Trost SG, Felton GM, Ward DS, Dowda M, Saunders R. Correlates of physical activity behavior in rural youth. Res Q Exerc Sport 1997; 68: 241–248.

98.* Pérusse L, Leblanc C, Bouchard C. Familial resemblance in lifestyle components: results from the Canada Fitness Survey. Can J Public Health 1988; 79: 201–205.

99.* Pérusse L, Tremblay A, Leblanc C, Bouchard C. Genetic and environmental influences on level of habitual physical activity and exercise participation. Am J Epidemiol 1989; 129: 1012–1022.

100.* Poest CA, Williams JR, Witt DD, Atwood ME. Physical activity patterns of preschool children. Early Child Res Q 1989; 4: 367–376.

101. Prochaska JJ, Rodgers MW, Sallis JF. Association of parent and peer support with adolescent physical activity. Res Q Exerc Sport 2002; 73: 206–210.

102.* Reynolds KD, Killen JD, Bryson SW, Maron DJ, Taylor CB, Maccoby N, Farquhar JW. Psychosocial predictors of physical activity in adolescents. Prev Med 1990; 19: 541–551.

103. Robinson CH, Thomas SP. The Interaction Model of Client Health Behavior as a conceptual guide in the explanation of children’s health behaviors. Public Health Nurs 2004; 21: 73–84.

104. Romero AJ, Robinson TN, Kraemer HC, Erickson SJ, Haydel KF, Mendoza F, Killen JD. Are perceived neighborhood hazards a barrier to physical activity in children? Arch Pediatr Adolesc Med 2001; 155: 1143–1148.

105. Rossow I, Rise J. Concordance of parental and adolescent health behaviors. Soc Sci Med 1994; 38: 1299–1305.

106. Runyan SM, Stadler DD, Bainbridge CN, Miller SC, Moyer-Mileur LJ. Familial resemblance of bone mineralization, calcium intake, and physical activity in early-adolescent daughters, their mothers, and maternal grandmothers. J Am Diet Assoc 2003; 103: 1320–1325.

107.* Sallis JF, Alcaraz JE, McKenzie TL, Hovell MF. Predictors of change in children’s physical activity over 20 months: variations by gender and level of adiposity. Am J Prev Med 1999; 16: 222–229.

108.* Sallis JF, Alcaraz JE, McKenzie TL, Hovell MF, Kolody B, Nader PR. Parental behavior in relation to physical activity and fitness in 9-year-old children. Am J Dis Child 1992; 146: 1383–1388.

109.* Sallis JF, Nader PR, Broyles SL, Berry CC, Elder JP, McKenzie TL, Nelson JA. Correlates of physical activity at home in Mexican-American and Anglo-American preschool children. Health Psychol 1993; 12: 390–398.

110.* Sallis JF, Patterson TL, Buono MJ, Atkins CJ, Nader PR. Aggregation of physical activity habits in Mexican-American and Anglo families. J Behav Med 1988; 11: 31–41.

111.* Sallis JF, Patterson TL, McKenzie TL, Nader PR. Family variables and physical activity in preschool children. J Dev Behav Pediatr 1988; 9: 57–61.

112. Sallis JF, Prochaska JJ, Taylor WC, Hill JO, Geraci JC. Correlates of physical activity in a national sample of girls and boys in grades 4 through 12. Health Psychol 1999; 18: 410–415.

113. Sallis JF, Taylor WC, Dowda M, Freedson PS, Pate RR. Correlates of vigorous physical activity for children in grades 1 through 12: comparing parent-reported and objectively measured physical activity. Pediatr Exerc Sci 2002; 14: 30–44.

114. Saunders RP, Motl RW, Dowda M, Dishman RK, Pate RR. Comparison of social variables for understanding physical activity in adolescent girls. Am J Health Behav 2004; 28: 426–436.

115. Saunders RP, Pate RR, Felton G, Dowda M, Weinrich MC, Ward DS, Parsons MA, Baranowski T. Development of questionnaires to measure psychosocial influences on children’s physical activity. Prev Med 1997; 26: 241–247.

116. Saxena R, Borzekowski DL, Rickert VI. Physical activity levels among urban adolescent females. J Pediatr Adolesc Gynecol 2002; 15: 279–284.

117. Schmitz KH, Lytle LA, Phillips GA, Murray DM, Birnbaum AS, Kubik MY. Psychosocial correlates of physical activity and sedentary leisure habits in young adolescents: the Teens Eating for Energy and Nutrition at School study. Prev Med 2002; 34: 266–278.

118.* Shephard RJ, Jequier JC, Lavallee H, La Barre R, Rajic M. Habitual physical activity: effects of sex, milieu, season and required activity. J Sports Med Phys Fitness 1980; 20: 55–66.

119. Shropshire J, Carrol B. Family variables and children’s physical activity: influence of parental exercise and socio-economic status. Sport Educ Soc 1997; 2: 95–116.

120. Simons-Morton BG, McKenzie TJ, Stone E, Mitchell P, Osganian V, Strikmiller PK, Ehlinger S, Cribb P, Nader PR. Physical activity in a multiethnic population of third graders in four states. Am J Public Health 1997; 87: 45–50.

121. Starfield B, Riley AW, Witt WP, Robertson J. Social class gradients in health during adolescence. J Epidemiol Community Health 2002; 56: 354–361.

122. Starfield B, Robertson J, Riley AW. Social class gradients and health in childhood. Ambul Pediatr 2002; 2: 238–246.

123. Strauss RS, Rodzilsky D, Burack G, Colin M. Psychosocial correlates of physical activity in healthy children. Arch Pediatr Adolesc Med 2001; 155: 897–902.

124.* Stucky-Ropp RC, DiLorenzo TM. Determinants of exercise in children. Prev Med 1993; 22: 880–889.

125.* Tappe MK, Duda JL, Ehrnwald PM. Perceived barriers to exercise among adolescents. J Sch Health 1989; 59: 153–155.

126.* Tappe MK, Duda JL, Menges-Ehrnwald P. Personal investment predictors of adolescent motivational orientation toward exercise. Can J Sport Sci 1990; 15: 185–192.

127. Taveras EM, Rifas-Shiman SL, Field AE, Frazier AL, Colditz GA, Gillman MW. The influence of wanting to look like media figures on adolescent physical activity. J Adolesc Health 2004; 35: 41–50.

128.* Terre L, Ghiselli W, Taloney L, DeSouza E. Demographics, affect, and adolescents’ health behaviors. Adolescence 1992; 27: 12–24.

129.* Theodorakis Y, Doganis G, Bagiatis K, Gouthas M. Preliminary study of the ability of reasoned action model in predicting exercise behaviour of young children. Percept Mot Skills 1991; 72: 51–58.

130. Theodorakis Y, Papaioannou A, Karastogianidou K. Relations between family structure and students’ health-related attitudes and behaviors. Psychol Rep 2004; 95: 851–858.

131. Timperio A, Crawford D, Telford A, Salmon J. Perceptions about the local neighborhood and walking and cycling among children. Prev Med 2004; 38: 39–47.

132.* Trost SG, Pate RR, Dowda M, Saunders R, Ward DS, Felton G. Gender differences in physical activity and determinants of physical activity in rural fifth grade children. J Sch Health 1996; 66: 145–150.

133.* Trost SG, Pate RR, Saunders R, Ward DS, Dowda M, Felton G. A prospective study of the determinants of physical activity in rural fifth-grade children. Prev Med 1997; 26: 257–263.

134. Trost SG, Pate RR, Ward DS, Saunders R, Riner W. Correlates of objectively measured physical activity in preadolescent youth. Am J Prev Med 1999; 17: 120–126.

135. Trost SG, Sallis JF, Pate RR, Freedson PS, Taylor WC, Dowda M. Evaluating a model of parental influence on youth physical activity. Am J Prev Med 2003; 25: 277–282.

136. Tuinstra J, Groothoff JW, van den Heuvel WJ, Post D. Socio-economic differences in health risk behavior in adolescence: do they exist? Soc Sci Med 1998; 47: 67–74.

137. Unger JB, Reynolds K, Shakib S, Spruijt-Metz D, Sun P, Johnson CA. Acculturation, physical activity, and fast-food consumption among Asian-American and Hispanic adolescents. J Community Health 2004; 29: 467–481.

138. Vermorel M, Vernet J, Bitar A, Fellmann N, Coudert J. Daily energy expenditure, activity patterns, and energy costs of the various activities in French 12-16-y-old adolescents in free living conditions. Eur J Clin Nutr 2002; 56: 819–829.

139. Vilhjalmsson R. Effects of social support on self-assessed health in adolescence. J Adolesc Health 1994; 23: 437–452.

140. Vilhjalmsson R, Kristjansdottir G. Gender differences in physical activity in older children and adolescents: the central role of organized sport. Soc Sci Med 2003; 56: 363–374.

141. Vilhjalmsson R, Thorlindsson T. Factors related to physical activity: a study of adolescents. Soc Sci Med 1998; 47: 665–675.

142. Wagner A, Klein-Platat C, Arveiler D, Haan MC, Schlienger JL, Simon C. Parent-child physical activity relationships in 12-year old French students do not depend on family socioeconomic status. Diabetes Metab 2004; 30: 359–366.

143. Wardle J, Jarvis M, Steggles N, Sutton S, Williamson S, Farrimond H, Cartwright M, Simon AE. Socioeconomic disparities in cancer-risk behaviors in adolescence: baseline results from the Health and Behaviour in Teenagers Study (HABITS). Prev Med 2003; 36: 721–730.

144. Welk GJ, Wood K, Morss G. Parental influences on physical activity in children: an exploration of potential mechanisms. Pediatr Exerc Sci 2003; 15: 19–33.

145. Williams EA, Jenkins C, Nevill AM. Social area influences on leisure activity – an exploration of the ACORN classification with reference to sport. Leisure Stud 1988; 7: 81–94.

146. Wold B, Oygard L, Eder A, Smith C. Social reproduction of physical activity, Implications for health promotion in young people. Eur J Public Health 1994; 4: 163–168.

147. Woodfield L, Duncan M, Al-Nakeeb Y, Nevill A, Jenkins C. Sex, ethnic and socio-economic differences in children’s physical activity. Pediatr Exerc Sci 2002; 14: 277–285.

148. Yang X, Telama R, Laakso L. Parent’s physical activity, socioeconomic status and education as predictors of physical activity and sport among children and youths – a 12-year follow-up study. Int Rev Soc Sports 1996; 31: 273–291.

149.* Zakarian JM, Hovell MF, Hofstetter CR, Sallis JF, Keating KJ. Correlates of vigorous exercise in a predominantly low SES and minority high school population. Prev Med 1994; 23: 314–321.

150. Ziviani J, Scott J, Wadley D. Walking to school: incidental physical activity in the daily occupations of Australian children. Occup Ther Int 2004; 11: 1–11.

Ancillary