Environmental changes are expected to lead to decreased time in sedentary behavior and to increased levels of physical activity in populations. Past research has emphasized psychosocial determinants of physical activity. Progress in the field will require more focus on understanding sedentary behaviors and the role of environmental determinants.
Physically inactive lifestyles increase the risk of several chronic diseases.
Physical inactivity is prevalent among young adults and is more prevalent in older age groups and among the socially disadvantaged. More than half of the adult populations of industrialized nations can be classified as insufficiently active for health benefits (6,12,15). Recent public health strategies for chronic disease prevention are now focusing on regular, moderate-intensity physical activity and drawing scientific attention to habitual behaviors that are integral to the context of people’s everyday lives. This new public policy and programmatic context, along with the scientific evidence on which it is based, presents some new research opportunities and challenges for the exercise and sport sciences.
The overall challenge is to better understand the modifiable determinants of physical activity and to translate that knowledge into practical actions for public health benefit. Known determinants of physical activity include genetic, physiological, musculoskeletal, psychological, social, environmental, and other factors (12, 15). Of the determinants of physical activity, environmental influences are the least understood but arguably the most important class of determinant that should be targeted through public health interventions (9,12). A more complete understanding of how environmental factors act to determine physical activity habits, and understanding how they interact with genetic and other influences, involves a research effort of considerable scope and complexity (3, 12). Identifying, in a public health context, relevant physical activity behaviors and their potential environmental determinants, will help in starting to map the relevant territory.
Here, we argue that understanding and influencing physical activity habits in whole populations requires a strong focus on the environmental determinants of inactive lifestyles. As an initial conceptual step, we apply the concept of ‘behavioral settings’ as a guide to classifying ways in which environmental influences may act to shape physically active and sedentary behavior choices. We also argue that there should be a research focus on sedentary behavior as a distinct class of variables. Relevant theories and conceptual models and recent evidence on environmental determinants of physical activity are described. In doing so, we restrict our focus to studies of adults in industrialized countries.
Since the early 1990s, public health strategy and associated research has sharpened its focus on regular, moderate-intensity physical activity. Physical activity now has a central role in health system and societal efforts that are directed at the prevention of cardiovascular disease, diabetes, several cancers, and other chronic disease outcomes.
The supporting evidence, the research needs, and the bases for public policy and programmatic initiatives have been articulated in the 1996 U.S. Surgeon General’s Report on Physical Activity and Health (15) and have been strongly emphasized by other scientific and health system consensus groups. Recent national policy documents, along with physical activity and health guidelines from several countries, have articulated and disseminated information on this newer focus (12). A key fact is that more than half of the adult populations of industrialized nations are insufficiently active in leisure time to meet minimal health-related physical activity criteria (6, 12, 15).
Objectives that are either explicit or implicit in these new policy documents relate to the maintenance of a regular pattern of being physically active, emphasizing ‘lifestyle’ activity (particularly walking). They point to the types of physical activity patterns that can be sustainable habitual behaviors that are integral to the contexts of people’s everyday lives (12, 15).
Within the exercise and sport science research community and the related professions, the case for a more explicit population health and public health focus has been widely advocated (9, 12, 15). Although purposeful physical training, exercising, developing strength and skill, and focusing on physical fitness and functional performance attributes remain important, the new public health focus places its primary emphasis on increasing population-wide energy expenditure through moderate-intensity physical activities. In practice, recreational and transport-related walking are being emphasized, as these are available to the whole population. This new context requires innovative research strategies and some reorientation of focus on the part of the exercise and sport sciences. It will also require greater involvement in physical; activity research by adjunct disciplines and subdisciplines. Behavioral epidemiology, health psychology, transportation research, urban design and planning, and architectural studies will have particularly useful contributions to make (9). In the behavioral science domain, the main focus is now shifting from individually focused studies of motivation and persistence in exercise and sport to activities that realistically can impact on the overall energy expenditure of large population groups of sedentary adults (9, 12, 15).
The exercise and sports sciences are dealing with a new context and a new set of research challenges. Relevant concepts and research methods are needed to account for physical activity in the contexts of everyday life—in the workforce, in educational institutions, in domestic and neighborhood environments, in commuting settings, and in recreational and community environments. In doing so, our research must identify largescale, modifiable environmental determinants and evidence-based intervention strategies for whole populations.
In several domains of public health research, in behavioral epidemiology, and in health psychology, there is a recent and growing interest in how health-related behaviors can be determined by their environmental contexts (9,12). Higher rates of disease, disability and premature death among the socially disadvantaged are likely to be attributable to objectively observable differences in the material conditions of people’s lives (5).
Important determinants of these socioeconomic differentials in health outcomes include aspects of people’s everyday environments that can act as barriers to or facilitators of physical activity. Participation in physical activity shows clear social class differentials. These parallel the social differentials in risk of several chronic diseases (15). Understanding and influencing physical activity levels in whole populations by focusing on environmental determinants poses some new public policy, public health, and research challenges. These will require new concepts and research methods, along with stronger linkages between currently separate disciplines (3). The ultimate test will be whether such approaches actually influence the physical activity levels of communities and populations.
Addressing physical activity in a public health framework requires an understanding of modifiable determinants (12,15). Although there has been extensive study of beliefs, attitudes, values, and other individual attributes as factors influencing activity, much less is known about potentially modifiable environmental characteristics and how these might best be measured (9,12). Despite the modest body of research on which to build evidence-based strategies, environmental changes, and a range of structural and social innovations are now being widely advocated and implemented (9). These aim to provide more opportunities for people to be physically active. Such initiates are unlikely to do harm. Ver6y probably they will do much that is good. Their proliferation emphasizes the urgent need for research to better understand how environmental factors might act to determine people’s apparent ‘choices’ to be physically active.
Studies of physical activity determinants have tended to focus on biological, psychological, and social influence. Sallis and Hovell’s (10) specific theoretical model of physical activity behavior builds on Social Learning Theory and incorporates explicit environmental constructs. In the environmental realm, social aspects have been most thoroughly studied, particularly social modeling and vicarious learning (10, 12). “Ecological” models that aim to account for the complex, multilevel determinants of health-related behaviors have been broadly influential (11). Theories that aim to account for stability and change in health-related behaviors acknowledge the importance of the internal and social environments, but some specify more explicitly the role of the physical environment (9,11,12). Behavioral Choice Theory (BCT) describes interactions between individual persons and the environment (8). BCT evolved from early laboratory based studies of animal and human choice behavior. These studies have viewed organisms (including humans) as existing in an indeterminate environment, choosing between uncertain outcomes. In the late 1980s, Rachlin synthesized the cognitive and behavioral approaches to the study of decision making and choice and described the interaction between individual person and their environment as a reciprocal relationship (8). In BCT, this relationship consists of external processes (such as feedback functions, which are properties of the environment) and internal processes (which are the value or worth of the alternatives) (8). Assessment of the properties of the environment that constrain or deter some behaviors (i.e. physical activity) and that facilitate or encourage others (i.e. sedentary behaviors) may be important in understanding why some behaviors are chosen over others. Epstein has applied an adaptation of BCT – behavioral economics – to study physical activity and sedentary behavior choices among overweight children (3).
In laboratory-based studies of obese children, Epstein has found that by reducing access to sedentary behaviors and increasing reinforcement of physical activity behaviors, the amount of time spent in physical activity increases (3). These studies have shown that in the laboratory, activity choice can be influenced by the relative reinforcing values, the costs and delayed access to alternative activities. However, the challenge is to apply the BCT/behavioral economics model and other environment-focused accounts of behavior more broadly to physical activity and in population based studies.
Behavior settings are those social and physical situations in which behaviors take place, by promoting and sometimes demanding certain actions and by discouraging or prohibiting others. The concept of behavior settings, originally proposed in work on ecological psychology in the 1970s (11), is a concept that may have considerable utility. Together with constructs based on the Sallis and Hovell (10) social learning model and suing Sallis and Owen’s studies building on that model (11, 12), the behavior settings construct may help in understanding the associations of environmental factors with physical activity. Sallis et al. (9) elaborated on these earlier models and built conceptual links to research on transport and urban design, focusing on how factors identified within these domains can influence commuting behaviors and incidental physical activity.
Based on this approach, we have developed a schematic representation of the influence of ‘behavior settings’ on people’s apparent choices to engage in either physically active or in sedentary behaviors, within a range of contexts (Fig. 1). Conceiving of environmental locations as behavior settings may be useful. Physical structures may be common across different behavior settings, so it is important to determine when or to what purpose the context might be influential in the shaping of physically active or sedentary behavior choices.
Fig 1 Environmental Determinants of Physical Activity and Sedentary Behavior. Potential influences of behavior settings and their functions on examples of physically active and sedentary choices.
Within Figure 1, we do not emphasize physical activity behaviors in themselves. Our main objective is to show how environmental contexts can promote or discourage a range of behaviors that do have important consequences for overall energy expenditure. With the examples in the Figure 1, we aim also to emphasize a nuance of the behavior settings construct – environmental influences on physical activity should not be approached simplistically and only in terms of their intuitively relevant physical attributes. In addition, behavior settings should be defined by the social or goal-related functions that they serve. For example, it is quite apparent that some environmental settings specifically promote physical activity (for example, sports fields, gymnasiums, bicycle trails). Others have functional attributes that act to discourage, restrict, or prohibit activity (for example, computer workstations, highways, domestic living areas). There are also less obvious functional and socially defined attributes that influence habitual physical activity – for example, access to stairs and the time use demands of occupational settings. In this context, consider also what might shape the need to walk to and from public transport. Another pertinent functional behavior settings example would be law enforcement arrangements that enable safe and pleasant local walking and cycling opportunities related to errands or recreation in urban areas.
There is a modest body of relevant evidence on environmental determinants of physical activity that has been reviewed elsewhere (9,11,12). Some environmental variables are associated with reported physical activity; however, the associations have at times been stronger with objectively assessed environmental variables that with selfreports. In a cross-sectional study, for example, an index of self-reported environmental barriers and facilitators of physical activity in local communities was not related to physical activity (11). Number of convenient facilities predicted increases in walking by both men and women (11). Time of year (mainly because of the prevailing weather conditions in the different seasons) embodies an obvious set of environmental influences on physical activity. The colder, wetter seasons in Scotland have been found to be associated with lower levels of leisure-time participation in physical activity – 32% participation rates were found in summer versus 23% in winter (14).
There are also potential geographical location effects on physical activity. IN a recent Australian study, coastal place of residence was found to be associated with being more likely to be physically active, after controlling for socioeconomic states and other potential confounding variables (1). Although city living may limit some forms of activity, some studies have found higher urban population density to be associated with less motor vehicle use and more incidental physical activity (2). Street layouts that allow walking and cycling for local errands, effects of traffic volume on perceived safety, distances to travel in spread-out suburbs, and several other sets of environmental factors can act to influence habitual physical activity.
Although our focus is on the specific study of environment-behavior linkages, other important domains of evidence can shed light on the broader issue. Of particular relevance is the argument that can be made using ecological evidence on energy intakes and average weight gains in populations. Using such data, Prentice and Jebb (7) have argued that environmental influences are strongly related to physical inactivity, with such factors as increasing rates of television viewing and car ownership underlying the recent increases in the population prevalence of obesity. They argue that modern lifestyles, associated with increased affluence, require less energy expenditure as energy saving domestic devices and motorized equipment displace physical tasks. Such knowledge and argument can help to link our environmental approach to a quite compelling body of relevant evidence.
Observational and descriptive epidemiology studies of average weight gains in whole populations have identified physical inactivity as a central influence (7). It has been argued that time spent in sedentary behaviors may result in decreased overall energy expenditure and increase people’s risk of overweight and obesity and associated health outcomes (7). Sedentary behaviors can be clearly identified by the amount of energy required to perform them. Behaviors that involve low levels of energy expenditure may be defined by their allocated MET values [the ratio of physical activity metabolic rate to resting metabolic rate (RMR); (12,15)]. Where an RMR is equivalent to one MET, sedentary behavior may be defined as having a MET value between one and 1.5 (for example, equivalent to sitting or laying down).
Study of sedentary behavior may be as important as the study of physical activity, if we are ultimately to find ways to influence important health outcomes. Thus, it may be helpful to explore sedentary behavior as a unique attribute in its own right and to examine what is known about some of its outcomes.
Although sedentary behavior may arguably be conceptualized as no more than the other side of the physical activity coin, we see it as a class of behaviors that can coexist with and also compete with physical activity. One difficulty involved n measuring sedentary behaviors is that the repetitive, often non-interactive nature of these behaviors makes recall more difficult. One of the most common leisure-time sedentary behaviors, television viewing, has been studied extensively in adults and children. Findings suggest that this behavior may have detrimental effects on overweight and obesity that can be independent of leisure-time physical activity level. These television-viewing studies emphasize the relevance of studying sedentary behavior as a distinct entity.
Data on strong positive relationships between hours of television viewing and body mass index suggest that prolonged periods of sedentary behavior can be associated with being more likely to be overweight. Such observations are key elements in the argument that several aspects of life in the urban environments of industrialized countries can act to discourage physical activity and to promote sedentary behavior. For example, a recent cross-sectional study of television viewing in a large representative sample of Australian adults (Fig2) found associations between hours of television viewing and body mass index (13).
An important inference from this study is that physical activity may not be protective against prolonged periods of sedentary behavior and associated weight gain. Even those who were highly active in their leisure time but reported watching more than 4 h of television per day were twice as likely to be overweight as were those who watched less than 1 h of television per day. Longitudinal studies of the associations between body mass index and television viewing have also found positive relationships which can be independent of physical activity level.
With the rapid development of information technologies for communication and entertainment, the emergence of ‘information environments’ as a sedentary behavior setting is of increasing relevance. Our recent research has identified computer use as playing a significant role in the discretionary time use of young adults, negatively associated with physical activity participation (4). A sample of young adults completed a self-report survey assessing physical activity (2 wk physical activity recall) as well as computer use and preferences for computer use. To assess computer use, we summed time spent using computers of study or course work, for paid employment, for recreational use of the Internet, for playing computer games, and for non-recreational, non-study purposes, such as paying bills and gathering information (4). Participants’ reported level of overall computer use varied in relation to their physical activity level. Higher levels of computer use were associated with increased likelihood of physical inactivity (Fig 3).
Computer use was reported as a barrier to physical activity ‘often’ or ‘very often’ by 15% of respondents, ‘sometimes’ by 28%, and ‘rarely’ or ‘never’ by 57%. Figure 3 also shows participants’ reports of computer use as preventing physical activity, in relation to levels of computer use. Participants who reported the highest levels of computer use were more likely to report computer use to be a barrier to activity than those who reported lower levels of computer use (4).
Being a group with a high rate of computer use, the young adults who participated may provide early indication of what will be the case for the wider population as personal computers and Internet access become ubiquitous. As information environments for commercial, social, educational, and health-related transaction become increasingly Internet-based, large segments of the population will spend increasing amounts of occupational, domestic maintenance, and recreational time in a context that in essence mandates sedentary behavior.
The studies described above suggest that prolonged screen-based, keyboard, mouse and joystick-oriented behaviors, along with high rates of habitual television viewing and Internet use are important to study in their own right. Other common sedentary behaviors, such as prolonged periods of automobile travel and regular automobile commuting or long periods of time spent impassive recreation and entertainment, may also be important to include in studies.
As we have argued earlier in this article, it might seem that studies of physical activity and sedentary behavior are doing no more than looking at two sides of the same coin. We argue that this is not the case. As we have demonstrated above, there are some new research findings, from which it seems reasonable to infer that there may be both independent and interactive effects of physical activity and sedentary behavior. The studies that we have described suggest that there may be scientific utility in differentiating physical activity and sedentary behavior. Each may be usefully addressed as a problem for research and for public health action in its own right.
Understanding the macro-level determinants of physical activity and sedentary behavior in populations presents new areas of research opportunity for the exercise and sport sciences. Many informative findings will soon emerge, for example, from further studies on the associations of land use and transportation arrangements with walking and bicycle use. They will come also from studies of the impact of television, computer and Internet use and particularly from studies of objectively observed physical activity environments. New approaches to measurement that emphasize direct observation of environmental attributes and behaviors and that use miniaturized electronic monitors to directory measure activity in specific environmental settings will be particularly informative.
Studying applications of BCT within the framework of the model that we have proposed in Figure 1 may provide useful insights. It may be particularly informative to analyze the determinants of sedentary versus physically active behavior choices in particular settings by applying constructs from BCT. Such studies might include, for example, point-of-choice studies using direct observation methodologies, where the probability of sedentary or active behavior choices are examined as a function of the reinforcing value of relevant behaviors. The influence of naturally occurring or experimentally manipulated environmental barriers could be assessed through direct observation, self-report, or both. I this field, applications of BCT and behavioral economics (3,8) hold promise, particularly in generating a theoretically coherent focus for field-based studies of the determinants of sedentary versus physically active behavior choices.
The exercise and sport sciences add several strong domains of scientific discipline to the study of environmental determinants of physical activity and sedentary behavior. However, such studies will need also knowledge from several disciplines beyond the central biological and behavioral disciplines of mainstream exercise and sport science. Studies from epidemiology, transportation science, urban design, architecture, economics, political science and others will be particularly relevant. Such research will lead to a better understanding of the basic determinants of the low levels of habitual energy expenditure that are now characteristic of human populations living and working in environments that make them extremely technology-dependent. It can provide an evidence base on which to build relevant and effective environmental-change initiatives. We hope that politicians and policy makers will come to see that broad social and environmental changes are needed. The health of many millions of humans will benefit from regaining lost opportunities to be physically active within the texture of their daily lives.
1. Bauman, A., b. Smith, L. Stoker, B. Bellew, M. Booth. Geographical influences upon physical activity participation: evidence of a ‘coastal’ effect. Aust. NZJ. Public Health 23:322-325, 1999.
2. Cervero, R., and R. Gorham. Commuting in transit versus automobile neighborhoods. J. Am. Planning Assoc. 61:210-225, 1995.
3. Epstein, L.H. Integrting theoretical approaches to promote physical activity. Am. J. Prev Med. 15:257-265, 1998.
4. Fotheringham, M.J., R.L. Wonnacott, and N. Owen. Computer use and physical inactivity in young adults: Public health perils and potentials of new information technologies. Ann. Behav. Med., in press.
5. Lynch, J.W., G.D. Smith. Kaplan, J.S. House. Income inequality and health: importance to health of individual income, psychosocial environment, or material conditions. BMJ 2000;320:1200-1204.
6. Owen, N. and A. Bauman. The descriptive epidemiology of a sedentary lifestyle in adult Australians. Int. J. Epidemiol. 21:305-310, 1992.
7. Prentice, A.M., and S.A. Jebb. Obesity in Britain: gluttony or sloth? BMJ 311:437-439, 1995.
8. Rachlin, H. Judgement, decision, and choice: a cognitive/behavioral synthesis. New York: W. H. Freeman, 1989.
9. Sallis, J.F., A. Bauman, M. Pratt. Environmental and policy interventions to promote physical activity. Am. J. Prev. Med. 15:379-397, 1998.
10. Sallis, J.F., and M.F. Hovell. Determinants of exercise behavior. Exerc. Sport Sci. Rev. 18:307-330, 1990.
11. Sallis, J.F. and N. Owen. Ecological models. In: Health Behavior and Health Education: Theory, Research and Practice 2nd ed., edited by K. Glanz, F.M. Lewis, and B.K. Rimer. San Francisco: Jossey-Bass, p. 403-424, 1997.
12. Sallis, J.F. and N. Owen. Physical Activity and Behavioral Medicine. Thousand Oaks, CA: Sage, 1999.
13. Salmon, J., A. Bauman, D. Crawford, A. Timperio, N. Owen. The association between television viewing and overweight among Australian adults participating in varying levels of leisure-time physical activity. Int. J. Obesity 24:600-606, 2000.
14. Uienbroek, D.G. Seasonal variation in leisure time physical activity. Med. Sci. Sports Exerc. 25:755-760, 1993.
15. United States Department of Health and Human Services. Physical Activity and Health: A Report of the Surgeon General. Atlanta, GA: Centers for Disease Control, 1996.
Environmental changes are expected to lead to decreased time in sedentary behavior and to increased levels of physical activity in populations. Past research has emphasized psychosocial determinants of physical activity. Progress in the field will require more focus on understanding sedentary behaviors and the role of environmental determinants.
Physically inactive lifestyles increase the risk of several chronic diseases.
Physical inactivity is prevalent among young adults and is more prevalent in older age groups and among the socially disadvantaged. More than half of the adult populations of industrialized nations can be classified as insufficiently active for health benefits (6,12,15). Recent public health strategies for chronic disease prevention are now focusing on regular, moderate-intensity physical activity and drawing scientific attention to habitual behaviors that are integral to the context of people’s everyday lives. This new public policy and programmatic context, along with the scientific evidence on which it is based, presents some new research opportunities and challenges for the exercise and sport sciences.
The overall challenge is to better understand the modifiable determinants of physical activity and to translate that knowledge into practical actions for public health benefit. Known determinants of physical activity include genetic, physiological, musculoskeletal, psychological, social, environmental, and other factors (12, 15). Of the determinants of physical activity, environmental influences are the least understood but arguably the most important class of determinant that should be targeted through public health interventions (9,12). A more complete understanding of how environmental factors act to determine physical activity habits, and understanding how they interact with genetic and other influences, involves a research effort of considerable scope and complexity (3, 12). Identifying, in a public health context, relevant physical activity behaviors and their potential environmental determinants, will help in starting to map the relevant territory.
Here, we argue that understanding and influencing physical activity habits in whole populations requires a strong focus on the environmental determinants of inactive lifestyles. As an initial conceptual step, we apply the concept of ‘behavioral settings’ as a guide to classifying ways in which environmental influences may act to shape physically active and sedentary behavior choices. We also argue that there should be a research focus on sedentary behavior as a distinct class of variables. Relevant theories and conceptual models and recent evidence on environmental determinants of physical activity are described. In doing so, we restrict our focus to studies of adults in industrialized countries.
Since the early 1990s, public health strategy and associated research has sharpened its focus on regular, moderate-intensity physical activity. Physical activity now has a central role in health system and societal efforts that are directed at the prevention of cardiovascular disease, diabetes, several cancers, and other chronic disease outcomes.
The supporting evidence, the research needs, and the bases for public policy and programmatic initiatives have been articulated in the 1996 U.S. Surgeon General’s Report on Physical Activity and Health (15) and have been strongly emphasized by other scientific and health system consensus groups. Recent national policy documents, along with physical activity and health guidelines from several countries, have articulated and disseminated information on this newer focus (12). A key fact is that more than half of the adult populations of industrialized nations are insufficiently active in leisure time to meet minimal health-related physical activity criteria (6, 12, 15).
Objectives that are either explicit or implicit in these new policy documents relate to the maintenance of a regular pattern of being physically active, emphasizing ‘lifestyle’ activity (particularly walking). They point to the types of physical activity patterns that can be sustainable habitual behaviors that are integral to the contexts of people’s everyday lives (12, 15).
Within the exercise and sport science research community and the related professions, the case for a more explicit population health and public health focus has been widely advocated (9, 12, 15). Although purposeful physical training, exercising, developing strength and skill, and focusing on physical fitness and functional performance attributes remain important, the new public health focus places its primary emphasis on increasing population-wide energy expenditure through moderate-intensity physical activities. In practice, recreational and transport-related walking are being emphasized, as these are available to the whole population. This new context requires innovative research strategies and some reorientation of focus on the part of the exercise and sport sciences. It will also require greater involvement in physical; activity research by adjunct disciplines and subdisciplines. Behavioral epidemiology, health psychology, transportation research, urban design and planning, and architectural studies will have particularly useful contributions to make (9). In the behavioral science domain, the main focus is now shifting from individually focused studies of motivation and persistence in exercise and sport to activities that realistically can impact on the overall energy expenditure of large population groups of sedentary adults (9, 12, 15).
The exercise and sports sciences are dealing with a new context and a new set of research challenges. Relevant concepts and research methods are needed to account for physical activity in the contexts of everyday life—in the workforce, in educational institutions, in domestic and neighborhood environments, in commuting settings, and in recreational and community environments. In doing so, our research must identify largescale, modifiable environmental determinants and evidence-based intervention strategies for whole populations.
In several domains of public health research, in behavioral epidemiology, and in health psychology, there is a recent and growing interest in how health-related behaviors can be determined by their environmental contexts (9,12). Higher rates of disease, disability and premature death among the socially disadvantaged are likely to be attributable to objectively observable differences in the material conditions of people’s lives (5).
Important determinants of these socioeconomic differentials in health outcomes include aspects of people’s everyday environments that can act as barriers to or facilitators of physical activity. Participation in physical activity shows clear social class differentials. These parallel the social differentials in risk of several chronic diseases (15). Understanding and influencing physical activity levels in whole populations by focusing on environmental determinants poses some new public policy, public health, and research challenges. These will require new concepts and research methods, along with stronger linkages between currently separate disciplines (3). The ultimate test will be whether such approaches actually influence the physical activity levels of communities and populations.
Addressing physical activity in a public health framework requires an understanding of modifiable determinants (12,15). Although there has been extensive study of beliefs, attitudes, values, and other individual attributes as factors influencing activity, much less is known about potentially modifiable environmental characteristics and how these might best be measured (9,12). Despite the modest body of research on which to build evidence-based strategies, environmental changes, and a range of structural and social innovations are now being widely advocated and implemented (9). These aim to provide more opportunities for people to be physically active. Such initiates are unlikely to do harm. Ver6y probably they will do much that is good. Their proliferation emphasizes the urgent need for research to better understand how environmental factors might act to determine people’s apparent ‘choices’ to be physically active.
Studies of physical activity determinants have tended to focus on biological, psychological, and social influence. Sallis and Hovell’s (10) specific theoretical model of physical activity behavior builds on Social Learning Theory and incorporates explicit environmental constructs. In the environmental realm, social aspects have been most thoroughly studied, particularly social modeling and vicarious learning (10, 12). “Ecological” models that aim to account for the complex, multilevel determinants of health-related behaviors have been broadly influential (11). Theories that aim to account for stability and change in health-related behaviors acknowledge the importance of the internal and social environments, but some specify more explicitly the role of the physical environment (9,11,12). Behavioral Choice Theory (BCT) describes interactions between individual persons and the environment (8). BCT evolved from early laboratory based studies of animal and human choice behavior. These studies have viewed organisms (including humans) as existing in an indeterminate environment, choosing between uncertain outcomes. In the late 1980s, Rachlin synthesized the cognitive and behavioral approaches to the study of decision making and choice and described the interaction between individual person and their environment as a reciprocal relationship (8). In BCT, this relationship consists of external processes (such as feedback functions, which are properties of the environment) and internal processes (which are the value or worth of the alternatives) (8). Assessment of the properties of the environment that constrain or deter some behaviors (i.e. physical activity) and that facilitate or encourage others (i.e. sedentary behaviors) may be important in understanding why some behaviors are chosen over others. Epstein has applied an adaptation of BCT – behavioral economics – to study physical activity and sedentary behavior choices among overweight children (3).
In laboratory-based studies of obese children, Epstein has found that by reducing access to sedentary behaviors and increasing reinforcement of physical activity behaviors, the amount of time spent in physical activity increases (3). These studies have shown that in the laboratory, activity choice can be influenced by the relative reinforcing values, the costs and delayed access to alternative activities. However, the challenge is to apply the BCT/behavioral economics model and other environment-focused accounts of behavior more broadly to physical activity and in population based studies.
Behavior settings are those social and physical situations in which behaviors take place, by promoting and sometimes demanding certain actions and by discouraging or prohibiting others. The concept of behavior settings, originally proposed in work on ecological psychology in the 1970s (11), is a concept that may have considerable utility. Together with constructs based on the Sallis and Hovell (10) social learning model and suing Sallis and Owen’s studies building on that model (11, 12), the behavior settings construct may help in understanding the associations of environmental factors with physical activity. Sallis et al. (9) elaborated on these earlier models and built conceptual links to research on transport and urban design, focusing on how factors identified within these domains can influence commuting behaviors and incidental physical activity.
Based on this approach, we have developed a schematic representation of the influence of ‘behavior settings’ on people’s apparent choices to engage in either physically active or in sedentary behaviors, within a range of contexts (Fig. 1). Conceiving of environmental locations as behavior settings may be useful. Physical structures may be common across different behavior settings, so it is important to determine when or to what purpose the context might be influential in the shaping of physically active or sedentary behavior choices.
Fig 1 Environmental Determinants of Physical Activity and Sedentary Behavior. Potential influences of behavior settings and their functions on examples of physically active and sedentary choices.
Within Figure 1, we do not emphasize physical activity behaviors in themselves. Our main objective is to show how environmental contexts can promote or discourage a range of behaviors that do have important consequences for overall energy expenditure. With the examples in the Figure 1, we aim also to emphasize a nuance of the behavior settings construct – environmental influences on physical activity should not be approached simplistically and only in terms of their intuitively relevant physical attributes. In addition, behavior settings should be defined by the social or goal-related functions that they serve. For example, it is quite apparent that some environmental settings specifically promote physical activity (for example, sports fields, gymnasiums, bicycle trails). Others have functional attributes that act to discourage, restrict, or prohibit activity (for example, computer workstations, highways, domestic living areas). There are also less obvious functional and socially defined attributes that influence habitual physical activity – for example, access to stairs and the time use demands of occupational settings. In this context, consider also what might shape the need to walk to and from public transport. Another pertinent functional behavior settings example would be law enforcement arrangements that enable safe and pleasant local walking and cycling opportunities related to errands or recreation in urban areas.
There is a modest body of relevant evidence on environmental determinants of physical activity that has been reviewed elsewhere (9,11,12). Some environmental variables are associated with reported physical activity; however, the associations have at times been stronger with objectively assessed environmental variables that with selfreports. In a cross-sectional study, for example, an index of self-reported environmental barriers and facilitators of physical activity in local communities was not related to physical activity (11). Number of convenient facilities predicted increases in walking by both men and women (11). Time of year (mainly because of the prevailing weather conditions in the different seasons) embodies an obvious set of environmental influences on physical activity. The colder, wetter seasons in Scotland have been found to be associated with lower levels of leisure-time participation in physical activity – 32% participation rates were found in summer versus 23% in winter (14).
There are also potential geographical location effects on physical activity. IN a recent Australian study, coastal place of residence was found to be associated with being more likely to be physically active, after controlling for socioeconomic states and other potential confounding variables (1). Although city living may limit some forms of activity, some studies have found higher urban population density to be associated with less motor vehicle use and more incidental physical activity (2). Street layouts that allow walking and cycling for local errands, effects of traffic volume on perceived safety, distances to travel in spread-out suburbs, and several other sets of environmental factors can act to influence habitual physical activity.
Although our focus is on the specific study of environment-behavior linkages, other important domains of evidence can shed light on the broader issue. Of particular relevance is the argument that can be made using ecological evidence on energy intakes and average weight gains in populations. Using such data, Prentice and Jebb (7) have argued that environmental influences are strongly related to physical inactivity, with such factors as increasing rates of television viewing and car ownership underlying the recent increases in the population prevalence of obesity. They argue that modern lifestyles, associated with increased affluence, require less energy expenditure as energy saving domestic devices and motorized equipment displace physical tasks. Such knowledge and argument can help to link our environmental approach to a quite compelling body of relevant evidence.
Observational and descriptive epidemiology studies of average weight gains in whole populations have identified physical inactivity as a central influence (7). It has been argued that time spent in sedentary behaviors may result in decreased overall energy expenditure and increase people’s risk of overweight and obesity and associated health outcomes (7). Sedentary behaviors can be clearly identified by the amount of energy required to perform them. Behaviors that involve low levels of energy expenditure may be defined by their allocated MET values [the ratio of physical activity metabolic rate to resting metabolic rate (RMR); (12,15)]. Where an RMR is equivalent to one MET, sedentary behavior may be defined as having a MET value between one and 1.5 (for example, equivalent to sitting or laying down).
Study of sedentary behavior may be as important as the study of physical activity, if we are ultimately to find ways to influence important health outcomes. Thus, it may be helpful to explore sedentary behavior as a unique attribute in its own right and to examine what is known about some of its outcomes.
Although sedentary behavior may arguably be conceptualized as no more than the other side of the physical activity coin, we see it as a class of behaviors that can coexist with and also compete with physical activity. One difficulty involved n measuring sedentary behaviors is that the repetitive, often non-interactive nature of these behaviors makes recall more difficult. One of the most common leisure-time sedentary behaviors, television viewing, has been studied extensively in adults and children. Findings suggest that this behavior may have detrimental effects on overweight and obesity that can be independent of leisure-time physical activity level. These television-viewing studies emphasize the relevance of studying sedentary behavior as a distinct entity.
Data on strong positive relationships between hours of television viewing and body mass index suggest that prolonged periods of sedentary behavior can be associated with being more likely to be overweight. Such observations are key elements in the argument that several aspects of life in the urban environments of industrialized countries can act to discourage physical activity and to promote sedentary behavior. For example, a recent cross-sectional study of television viewing in a large representative sample of Australian adults (Fig2) found associations between hours of television viewing and body mass index (13).
An important inference from this study is that physical activity may not be protective against prolonged periods of sedentary behavior and associated weight gain. Even those who were highly active in their leisure time but reported watching more than 4 h of television per day were twice as likely to be overweight as were those who watched less than 1 h of television per day. Longitudinal studies of the associations between body mass index and television viewing have also found positive relationships which can be independent of physical activity level.
With the rapid development of information technologies for communication and entertainment, the emergence of ‘information environments’ as a sedentary behavior setting is of increasing relevance. Our recent research has identified computer use as playing a significant role in the discretionary time use of young adults, negatively associated with physical activity participation (4). A sample of young adults completed a self-report survey assessing physical activity (2 wk physical activity recall) as well as computer use and preferences for computer use. To assess computer use, we summed time spent using computers of study or course work, for paid employment, for recreational use of the Internet, for playing computer games, and for non-recreational, non-study purposes, such as paying bills and gathering information (4). Participants’ reported level of overall computer use varied in relation to their physical activity level. Higher levels of computer use were associated with increased likelihood of physical inactivity (Fig 3).
Computer use was reported as a barrier to physical activity ‘often’ or ‘very often’ by 15% of respondents, ‘sometimes’ by 28%, and ‘rarely’ or ‘never’ by 57%. Figure 3 also shows participants’ reports of computer use as preventing physical activity, in relation to levels of computer use. Participants who reported the highest levels of computer use were more likely to report computer use to be a barrier to activity than those who reported lower levels of computer use (4).
Being a group with a high rate of computer use, the young adults who participated may provide early indication of what will be the case for the wider population as personal computers and Internet access become ubiquitous. As information environments for commercial, social, educational, and health-related transaction become increasingly Internet-based, large segments of the population will spend increasing amounts of occupational, domestic maintenance, and recreational time in a context that in essence mandates sedentary behavior.
The studies described above suggest that prolonged screen-based, keyboard, mouse and joystick-oriented behaviors, along with high rates of habitual television viewing and Internet use are important to study in their own right. Other common sedentary behaviors, such as prolonged periods of automobile travel and regular automobile commuting or long periods of time spent impassive recreation and entertainment, may also be important to include in studies.
As we have argued earlier in this article, it might seem that studies of physical activity and sedentary behavior are doing no more than looking at two sides of the same coin. We argue that this is not the case. As we have demonstrated above, there are some new research findings, from which it seems reasonable to infer that there may be both independent and interactive effects of physical activity and sedentary behavior. The studies that we have described suggest that there may be scientific utility in differentiating physical activity and sedentary behavior. Each may be usefully addressed as a problem for research and for public health action in its own right.
Understanding the macro-level determinants of physical activity and sedentary behavior in populations presents new areas of research opportunity for the exercise and sport sciences. Many informative findings will soon emerge, for example, from further studies on the associations of land use and transportation arrangements with walking and bicycle use. They will come also from studies of the impact of television, computer and Internet use and particularly from studies of objectively observed physical activity environments. New approaches to measurement that emphasize direct observation of environmental attributes and behaviors and that use miniaturized electronic monitors to directory measure activity in specific environmental settings will be particularly informative.
Studying applications of BCT within the framework of the model that we have proposed in Figure 1 may provide useful insights. It may be particularly informative to analyze the determinants of sedentary versus physically active behavior choices in particular settings by applying constructs from BCT. Such studies might include, for example, point-of-choice studies using direct observation methodologies, where the probability of sedentary or active behavior choices are examined as a function of the reinforcing value of relevant behaviors. The influence of naturally occurring or experimentally manipulated environmental barriers could be assessed through direct observation, self-report, or both. I this field, applications of BCT and behavioral economics (3,8) hold promise, particularly in generating a theoretically coherent focus for field-based studies of the determinants of sedentary versus physically active behavior choices.
The exercise and sport sciences add several strong domains of scientific discipline to the study of environmental determinants of physical activity and sedentary behavior. However, such studies will need also knowledge from several disciplines beyond the central biological and behavioral disciplines of mainstream exercise and sport science. Studies from epidemiology, transportation science, urban design, architecture, economics, political science and others will be particularly relevant. Such research will lead to a better understanding of the basic determinants of the low levels of habitual energy expenditure that are now characteristic of human populations living and working in environments that make them extremely technology-dependent. It can provide an evidence base on which to build relevant and effective environmental-change initiatives. We hope that politicians and policy makers will come to see that broad social and environmental changes are needed. The health of many millions of humans will benefit from regaining lost opportunities to be physically active within the texture of their daily lives.
1. Bauman, A., b. Smith, L. Stoker, B. Bellew, M. Booth. Geographical influences upon physical activity participation: evidence of a ‘coastal’ effect. Aust. NZJ. Public Health 23:322-325, 1999.
2. Cervero, R., and R. Gorham. Commuting in transit versus automobile neighborhoods. J. Am. Planning Assoc. 61:210-225, 1995.
3. Epstein, L.H. Integrting theoretical approaches to promote physical activity. Am. J. Prev Med. 15:257-265, 1998.
4. Fotheringham, M.J., R.L. Wonnacott, and N. Owen. Computer use and physical inactivity in young adults: Public health perils and potentials of new information technologies. Ann. Behav. Med., in press.
5. Lynch, J.W., G.D. Smith. Kaplan, J.S. House. Income inequality and health: importance to health of individual income, psychosocial environment, or material conditions. BMJ 2000;320:1200-1204.
6. Owen, N. and A. Bauman. The descriptive epidemiology of a sedentary lifestyle in adult Australians. Int. J. Epidemiol. 21:305-310, 1992.
7. Prentice, A.M., and S.A. Jebb. Obesity in Britain: gluttony or sloth? BMJ 311:437-439, 1995.
8. Rachlin, H. Judgement, decision, and choice: a cognitive/behavioral synthesis. New York: W. H. Freeman, 1989.
9. Sallis, J.F., A. Bauman, M. Pratt. Environmental and policy interventions to promote physical activity. Am. J. Prev. Med. 15:379-397, 1998.
10. Sallis, J.F., and M.F. Hovell. Determinants of exercise behavior. Exerc. Sport Sci. Rev. 18:307-330, 1990.
11. Sallis, J.F. and N. Owen. Ecological models. In: Health Behavior and Health Education: Theory, Research and Practice 2nd ed., edited by K. Glanz, F.M. Lewis, and B.K. Rimer. San Francisco: Jossey-Bass, p. 403-424, 1997.
12. Sallis, J.F. and N. Owen. Physical Activity and Behavioral Medicine. Thousand Oaks, CA: Sage, 1999.
13. Salmon, J., A. Bauman, D. Crawford, A. Timperio, N. Owen. The association between television viewing and overweight among Australian adults participating in varying levels of leisure-time physical activity. Int. J. Obesity 24:600-606, 2000.
14. Uienbroek, D.G. Seasonal variation in leisure time physical activity. Med. Sci. Sports Exerc. 25:755-760, 1993.
15. United States Department of Health and Human Services. Physical Activity and Health: A Report of the Surgeon General. Atlanta, GA: Centers for Disease Control, 1996.
联系客服
