Вы находитесь на странице: 1из 18

Psychology of Sport and Exercise 4 (2003) 724 www.elsevier.


Toward a comprehensive model of physical activity

John C. Spence a,, Rebecca E. Lee b

University of Alberta, Faculty of Physical Education, E-424 Van Vliet Centre, T6G 2H9 Edmonton, Alberta, Canada b University of Kansas, Department of Preventive Medicine, School of Medicine, Kansas City, USA Received 22 March 2002; received in revised form 22 May 2002; accepted 24 May 2002

Abstract Background and purpose. Despite the widely acknowledged public health burden and years of individually based intervention approaches, physical inactivity remains a growing concern among industrialized nations. Interventions aimed at changing individual dispositions that increase physical activity generally result in small changes in behavior that dissipate within weeks. Correlational research testing theories and models focusing on these same individual dispositions explain, at best, 2040% of the variance in physical activity. As a result, recent calls have been made for consideration of broader, multilevel, ecological approaches to physical activity promotion. The purpose of this article is to dene a comprehensive model for understanding physical activity and consider future directions for research. Methods. Relevant literature is reviewed within each of the areas being discussed. Results and conclusions. Ecological models incorporate both intra- and extra-individual inuences that may impact on individual physical activity. However, the role of extra-individual factors has not been clearly dened in current ecological models of physical activity. We present the theoretical background of ecological models of health behavior, and dene an ecological model for physical activity promotion. This model portrays physical activity behavior as being inuenced by interplay between environmental settings and biological and psychological factors. Further testing of this and existing ecological models of physical activity is recommended. 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Exercise; Ecological model; Behavior; Theory

Tel.: +1-780-492-1379; fax: +1-780-492-2364. E-mail address: jc.spence@ualberta.ca (J.C. Spence).

1469-0292/03/$ - see front matter 2002 Elsevier Science Ltd. All rights reserved. doi:10.1016/S1469-0292(02)00014-6

J.C. Spence, R.E. Lee / Psychology of Sport and Exercise 4 (2003) 724

Toward a comprehensive model of physical activity behavior Human anatomy and physiology has remained relatively unchanged over the past 40,000 years (Astrand, 1994; Cordain, Gotshall, Eaton, & Eaton, 1998). The interrelationship between our energy intake, energy expenditure, and physical activity requirements are essentially the same as our Stone Age forebears. But, in terms of energy expenditure resulting from physical activity, Cordain and colleagues estimate that contemporary Westerners expend approximately 38% of the energy that their Paleolithic ancestors did. To equal the total energy expenditure of modern huntergatherers (e.g., the Kung! bushmen), a typical 70 kg North American male would have to walk approximately 19 kms more per day in addition to usual activities (Cordain, Gotschall, Eaton, & Eaton, 1998). Thus, while human bodies are biologically endowed for high levels of energy expenditure, the luxuries of modernization facilitate a sedentary lifestyle. Physical inactivity is currently acknowledged as a serious public health burden throughout the industrialized world (US Department of Health & Human Services, 1996; Villeneuve, Morrison, Craig, & Schaubel, 1998; World Health Organization, 1997). Traditional physical activity promotion interventions that employ individual approaches have demonstrated limited success in promoting long-term maintenance (Dishman & Buckworth, 1996; Marcus & Forsyth, 1999). Although these interventions often alter individual behavior initially, it is likely that enduring social and environmental factors shape return to previous sedentary behavior once the intervention is over (Gauvin, Levesque, & Richard, 2001; Laitakari, Vuori, & Oja, 1996). Correlational research testing theories and models that focus on the individual explain, at best, 2040% of the variance in physical activity (see Baranowski, Anderson, & Carmack, 1998; Godin & Kok, 1996; McAuley & Mihalko, 1998; Spence, Courneya, Blanchard, & Wilson, 2000). One alternative to testing only individual inuences has been to focus on broader and multiple factors that may inuence behavior. Ecological models of health behavior focus on individual inuences as well as on social and environmental factors that may facilitate or inhibit individual behavior (Sallis & Owen, 1997). Ecological models posit that multiple levels of inuence determine individual behavior. In an effort to classify the multiple levels of intervening inuences within, between, and external to individuals that can change human development, Bronfenbrenner (1977) differentiated the external inuences into levels of settings with the most proximal setting being the microsystem and the most distal setting to the individual being the macrosystem. For the purposes of discussion, levels of inuence and intervention in ecological models can be broadly divided into intra-individual (person) and extra-individual (environment). Intra-individual inuences might include individual attributes, beliefs, attitudes, and behaviors, while extra-individual inuences might include environmental topography, social and cultural context and policies. For example, change at an intra individual level of inuence might include improving attitudes toward physical activity, thereby increasing the probability that physical activity behavior might occur. Change at an extraindividual level of inuence might include providing safe spaces to do physical activity, also increasing the probability that physical activity might occur. Working within an ecological model requires that measurement and assessment take place at more than one of these levels. In the research dening ecological models of health behavior, a number of typologies of ecological variables have been identied (e.g., Cohen, Scribner, & Farley, 2000; McLeroy, Bibeau, Steckler, & Glanz, 1988; Richard, Potvin, Kishchuk, Prlic, & Green, 1996; Stokols, 1992). How-

J.C. Spence, R.E. Lee / Psychology of Sport and Exercise 4 (2003) 724

ever, until recently, little discussion existed on how ecological models could apply specically to physical activity (see Dzewaltowski, 1997; Gauvin, Levesque, & Richard, 2001; Owen, Leslie, Salmon, & Fotheringham, 2000; Sallis, Bauman, & Pratt, 1998; Welk, 1999). One challenging issue for employing ecological models to the study of physical activity is the denition of a theoretical and conceptual framework to study the putative relationships between extra-individual causal mechanisms and behavior (Sallis, Bauman, & Pratt, 1998). Further, of the ecological models that have either been applied (Gauvin et al., 2001) or developed (Sallis, Bauman, & Pratt, 1998; Welk, 1999) in the physical activity domain, little consideration has been given to the role that biological factors may play in determining behavior. A primary goal of this paper is to address these issues and discuss future directions in the development and application of ecological models. First, we will briey review the theoretical underpinnings of ecological models. Theoretical underpinnings of ecological models A growing body of rhetoric and research argues that there may be synergy between individuals and environments that may exert inuence on individuals beyond individual characteristics (Susser & Susser, 1996). Ecological models posit that this synergy is born out of a good t between the individual and environment (Kelly, 1990). A good t refers to a matching of intraindividual attributes with environmental (extra-individual) attributes that produces positive human health behavior beyond the summation of the intra- plus extra-individual environment. Similarly, a poor individualenvironment t is a mismatch between individuals and environments that leads to poor health and disease. A good individualenvironment t can be improved by enhancing either side, the individual or the environment. In the case of a poor t, ecological models posit that it is more efcient to enhance environment rather than change individuals, because enhancing one environment can have implications for many individuals. Because improvements in the individualenvironment t can be achieved from changes at either the intra-individual or extra-individual level, there are multiple levels of inuence and points of intervention. Ecological models have been described as typologies of individual, social, and environmental features (Sallis & Owen, 1997), reecting and building upon the work of Bronfenbrenner (1977). Others have depicted aggregate, contagion, environmental and structural features (Blakely & Woodward, 2000) that operationalize extra-individual inuences in terms of strategies of measurement and methods of impact on human health. Others classify ecological inuences in terms of resource availability, physical structures, social structures and cultural and media messages (Cohen, Scribner, & Farley, 2000). While many typologies of ecological models are discussed, they all share common features. Regardless of which typology is employed, ecological models posit that constituents (intra- and extra-individual) are interdependent (Kelly, 1990) and can exert direct effects on each other. If a change is made at one level of inuence, all other levels may be affected. For example, a national public health campaign to promote physical activity could stimulate a municipality to build a park in a low-SES neighborhood, which, in turn, may encourage the local residents to be more active. The increase in resident activity might feed back to the municipal leaders via community empowerment, spawning more community resources. Since all levels of ecological models are interdependent, it follows that available resources cycle among the levels of inuence (Kelly, 1990). If the resources that are available at one level


J.C. Spence, R.E. Lee / Psychology of Sport and Exercise 4 (2003) 724

of inuence change, then the other levels will have to compensate for this change. For instance, if the extra-individual-level inuence of the park in the previous example is removed, then residents will have to change their intra-individual-level behavior to accommodate the extra-individual-level change in resources. Thus, removing the park may mean that children will relocate to the street, families will not have a convenient spot to toss a tennis ball to their dog, and neighbors will not be able to socialize in a space free from passing car exhaust. The interdependence of levels in the ecological system provides health promoters with novel points of intervention (see Gauvin, Levesque et al., 2001; Richard, Potvin, Kishchuk, Prlic & Green, 1996; Stokols, 1996). Interventions made at either intra- or extra-individual points will have implications for all other parts of the system. Therefore, extra-individual (i.e., environment change) interventions could inuence the individual behavior of all members of that environment. Because persons are interdependent with their environments, extra-individual inuences can directly constrain some behaviors and facilitate others. Barker (1968) proposed that behaviors occur in consistent patterns of regularly encountered environments that are called behavior settings. In behaviorist terms, the behavior setting represents the discriminate stimulus (Skinner, 1954) that elicits predictable human behavior. The behavior of the environment that mandates specic human behavior is independent from the people in the environment in the sense that settings (environments) themselves generate forces necessary for their own maintenance and survival. However, individuals have the capacity to change their environments, leading back to the interdependence of intra-and extraindividual levels. Similar to behavioral denitions (Skinner, 1954), ecological models posit that individuals adapt, or vary their behaviors or characteristics in response to available, changing resources in the extraindividual environment. Change in the environment can create conditions that are more favorable for one group of individuals and less favorable for others. The group with the more favorable conditions is likely to dominate in terms of power and inuence and may show more favorable health behaviors, For instance, despite individual demographic characteristics (e.g., age, gender, social class, income) merely residing in a more afuent neighborhood will likely increase individual physical activity (Ecob & Macintyre, 2000; Ellaway & Macintyre, 1996; Gauvin, Richard et al., 2001; Yen & Kaplan, 1998) and decrease smoking (Diez-Roux et al., 1997). These studies show that something about the neighborhood of residence inuences individual behavior beyond individual demographic characteristics. Sampson, Raudenbush and Earls (1997) note that social and organizational characteristics of neighborhoods are associated with variations in individual behavior that are not explained by individual characteristics alone. Perhaps extra-individual environments (e.g., neighborhoods) generate common values and qualities that inuence residents beyond individual characteristics (Sampson et al., 1997). While not well documented or explained, ecological models also include temporal features, acknowledging that subtle relationships in ecological systems can take a long time to manifest themselves (Moore & Carpenter, 1999). Levine, Toro and Perkins (1993) posit that all ecological perspectives must be longitudinal. However, ecological models may not necessarily foster linear temporal relationships because of the dynamic and unpredictable nature of human environments. A critical implication of the temporal nature of ecological inuences is that environmental effects may be subtle and only observable after years of exposure, while at the same time, effects may be dramatic, resulting in rapid behavior changes (e.g., smoking bans and reductions in smoking rates).

J.C. Spence, R.E. Lee / Psychology of Sport and Exercise 4 (2003) 724


The need for an ecological approach to physical activity research and promotion From a historical perspective, the current physical inactivity levels evident in most industrialized nations are extreme. If we are to understand the physical activity choices of contemporary humans, then we should not ignore changing cultural and environmental forces (Frankish, Milligan, & Reid, 1998; Powell, Kreuter, Stephens, Marti, & Heineman, 1991; Siedentop, 1996; Treloar et al., 1999). The Inuit peoples of northern Canada provide a vivid example of the corrupting inuence of modernization on physical activity and tness. As a result of acculturation (e.g., introduction of motor boats, televisions, motorized ski mobiles), these people are experiencing dramatic decreases in physical tness (Rode & Shephard, 1994) as well as generational shifts in preferences from traditional cooperative games and activities to more competitive pursuits (Glassford, 1976). To appropriately assess these inuences on physical activity, conceptual models describing how extra-individual factors relate to behavior and interact with one another are required. Dzewaltowski (1997) and Sallis and Owen (1997) were among the rst to discuss the need for an ecological approach to physical activity research and promotion. Welk (1999) followed with the Youth Physical Activity Promotion (YPAP) Model, which he termed a heuristic model for bridging the gap between theory and practice. Its goal was to serve as a guide for physical activity promotion programs. Couched within a social ecological framework (McLeroy, Bibeau, Steckler & Glanz, 1988) and the PrecedeProceed Model (Green & Kreuter, 1991), the YPAP model recognizes the inuence of both intra- and extra-individual factors on childrens physical activity participation. At the individual cognitive level, Welk (1999) makes a useful distinction between am I able? (e.g., perceptions of competence, self-efcacy) and is it worth it? (e.g., enjoyment, beliefs, attitudes) variables. This model could be adopted for research purposes but to date we have seen no evidence of it being used. Recognizing the lack of conceptual models to guide environmental and policy interventions, Sallis, Bauman, and Pratt (1998) developed a model for conceptualizing the steps that may be required to implement macro-level interventions (p. 388). The model acknowledges the role of advocacy or coalition groups (e.g., public health, sport, tness industry), agencies (e.g., park and recreation departments, schools, media, medical), policies, and environments (supportive settings, facilities, and programs). However, because of its macro focus, this model does not suggest how the home environment may inuence physical activity behavior. Moreover, it implies that environmental and policy interventions only have direct effects on physical activity, and it avoids discussion of interpretation of these inuences on the part of the individual. Last, Sallis et al. (1998) have given little consideration to how cultural factors should be included in macro-level interventions. A structural model of the environment Based upon ecological systems theory (Bronfenbrenner, 1989), Wachs (1992) has outlined a model for understanding the role of the environment in childrens development. We believe his structural model of the environment (SME) is readily adaptable for describing environmental


J.C. Spence, R.E. Lee / Psychology of Sport and Exercise 4 (2003) 724

inuences on physical activity behavior (see Fig. 1)1. According to the SME, environmental factors are categorized in a hierarchical, multilevel, and multidimensional fashion. This model describes a dynamic system that operates across space and time. At the most proximal level in the SME is the microsystem. This is the immediate setting within which individuals interact. It can consist of both social (e.g., verbal support) and physical (e.g.,

Fig. 1. A structural model of environmental inuences on behavior (adapted from Wachs, 1992).
Garcia Bengoechea and Johnson (2000) have recently used a version of Bronfenbrenners (1977, 1989) ecological systems theory to understand the determinants of sport participation in children.

J.C. Spence, R.E. Lee / Psychology of Sport and Exercise 4 (2003) 724


the presence of a safe playground) characteristics. Examples of microsystems are workplaces, schools, homes, and parks. It is likely that more than one microsystem plays a role in understanding physical activity. Encompassing the microsystem is the mesosystem, in which two or more microsystems may interact to exert inuence on physical activity behavior. The most important features at this and all subsequent levels are the linkages and processes occurring between settings. Thus, it is not just the presence of positive facets (e.g., walking trail) in the setting but also the quality of the interaction that takes place within and between settings. For instance, a childs physical activity level may be inuenced by both the verbal support received in the home microenvironment and the physical and social microenvironments at school. Other microsystems, external to the individual, also exert inuence on behavior. Exosystem dimensions are composed of the linkages and processes between two or more microsystems, at least one of which does not include the physically active/inactive person. An example of an exosystem dimension is the relationship between a parents workplace (e.g., existence of a health promotion program) and a childs physical activity in school. The fourth, and most distal, level consists of macrosystem dimensions. The macrosystem is the larger sociocultural context (e.g., social class structure, cultural values, community) in which the person resides. It encompasses the micro, meso, and exosystems with particular reference to the developmentally instigated belief systems, resources, hazards, lifestyles, opportunity structures, life course options and patterns of social interchange that are embedded in each of these systems. (Bronfenbrenner, 1989, p. 228) The nding that childhood SES is strongly related to subsequent adult participation in physical activity (Lynch, Kaplan, & Salonen, 1997) could be construed as an example of a macrosystem dimension. Other constructs included in the SME are physical ecology, pressure for macrosystem change, and higher level mediators. Physical ecology (e.g., climate, topology) and pressure for macrosystem change are thought to inuence behavior through the macrosystem. For instance, the climate and topology have dictated to a great degree the types of activities and games that are part of the Inuit culture (e.g., static indoor games, cooperative hunting activities). However, the Inuit have experienced a change in their cultural norms because of modernization (Glassford, 1976). Wachs (1992) does not explain what the higher level mediators are in his model (see Fig. 1). We assume that they are the individual dispositions of the developing person. Therefore, the environment in the SME inuences behavior indirectly through these higher level mediators. Nature of the structural model of the environment The more distal the level in the SME from the individual, the broader the inuence that the dimensions can exert over behavior. Therefore, upstream interventions (Orleans, 2000) that target population-level changes in behavior incorporate macrosystem dimensions, whereas downstream interventions target the mesosystem or microsystem levels. Distal processes can modify the pattern of relationships between environment and individual behavior occurring at levels that are more proximal. For example, the provision of quality physical education in the school (mesosystem


J.C. Spence, R.E. Lee / Psychology of Sport and Exercise 4 (2003) 724

inuence) can strengthen or reduce the effect of parental support for a childs participation in physical activity at home (microsystem). Because of the bidirectional nature of the SME, processes at the more proximal levels of the environment can inuence distal levels. For example, if nursing home residents demand the provision of quality exercise programs and equipment, it could lead to the development of certication guidelines for older adult activity instructors and policy for such programs. Wachs (1999) maintains that the more proximal the dimension to the individual, the more inuential it will be to immediate behavior. Thus, supportive proximal environments, such as parental support for physical activity and active school programs, can buffer detrimental macrosystem inuences (e.g., low SES). Because it is possible for both distal and proximal processes to serve as moderators2 of each other, the question for physical activity researchers becomes under what conditions is higher order environmental moderation of more fundamental proximal processes most or least likely to occur (Wachs, 1999, p. 362). Thus, it is important for researchers to develop hypotheses about the linkages between the various levels of the environment as they relate to physical activity behavior. The ecological model of physical activity To understand physical activity participation, we believe that four further additions need to be made to the SME (see Fig. 2) that focus on clarifying the roles of biological processes, higherlevel mediators, and physical ecology, and illuminating direct versus indirect roles of the environment. First, there should be little doubt that biological processes can inuence behavior (Bouchard & Shephard, 1994). Evidence suggests that energy exertion is probably controlled to some extent by an activity-stat or biological set-point (see Rowland, 1998) that controls the amount of activity in which an individual will be capable of participating. For instance, Goran, Gower, Nagy and Johnson (1998) found that as girls begin to experience puberty they maintain their resting metabolic rate while demonstrating a reduction in energy expenditure which is explained by a 50% reduction in physical activity. Goran et al. interpret this as being evidence for an energy-conserving mechanism that prepares girls for the demands of growth and maturation that take place in puberty. If such a mechanism is in place, then the interpretation of gender differences in adolescent physical activity (Garcia et al., 1995) needs to include more than the usual sociological (e.g., roles) or cultural explanations. Further, research shows that many facets of physical tness (e.g., body composition, maximal oxygen uptake, exibility, speed) are partially genetically determined (Bouchard, Malina, & Perusse, 1997). In fact, it is possible that there is heritability of behavior (see McGue & Bouchard, 1998), including physical activity behavior (e.g., Beunen & Thomas, 1999; Bouchard & Perusse, 1994). Second, as mentioned previously, the higher-level mediators in the Wachs (1992) model were probably supposed to consist of both biological and psychological factors. However, we believe

2 Throughout this manuscript we use the terms mediator and moderator to describe the roles of variables in various theories and models. When considering the relationship between three or more variables (X, Y, M,), and where X is assumed to affect another variable Y, mediation is said to take place when X no longer affects Y after M has been controlled (Baron & Kenny, 1986). Further, for the XY relationship to be mediated by M, it is required that X temporally precede M and Y (Kraemer, Stice, Kazdin, Offord, & Kupfer, 2001). A variable is considered a moderator if it inuences the association between X and Y.

J.C. Spence, R.E. Lee / Psychology of Sport and Exercise 4 (2003) 724


Fig. 2. The ecological model of physical activity (EMPA).

biological factors are more likely to serve as moderators than mediators between extra-individual factors and behavior. That is, biological factors (e.g., body composition, physical tness, pubertal status) are likely to inuence the type and extent of activity but not serve as the reasons for being physically active. The more likely mediators of extra-individual factors on behavior will be psychological factors such as cognitive and, possibly, personality constructs. Sufcient evidence exists documenting relationships between physical activity and constructs such as selfefcacy, perceived barriers, perceived benets, enjoyment of activity, and social support (Sallis & Owen, 1999). We suggest that the constructs from a cognitive- or personality-oriented theory with clearly identied mediators, such as social cognitive theory (Bandura, 1986) or theory of planned behavior (Ajzen, 1991), or possibly even the exercise self-esteem model (Sonstroem & Morgan, 1989), be used test the relationship between environment and behavior.


J.C. Spence, R.E. Lee / Psychology of Sport and Exercise 4 (2003) 724

Third, physical ecology should be recognized as having a direct inuence on both biological and psychological factors that may then moderate or mediate the physical ecologyphysical activity relationship. For instance, heat stress (Wenger, 1999) and air pollution (Linn & Gong, 1999) can compromise the biological capacity of the body to participate in physical activity. Further, physical ecology may inuence psychological factors (e.g., attitudes) that may also limit physical activity participation. For example, severe climate conditions, such as cold and ice, are often perceived as barriers to participation in outdoor activities by older adults (see OBrien Cousins, 1998) that may, in turn, reduce physical activity participation. Finally, in keeping with theoretical underpinnings of ecological models (see Sallis & Owen, 1997), we propose that environment has both direct and indirect effects on physical activity. For example, if stairs are the only route to ones ofce, then this environment will increase stairclimbing participation. If the stairs are nicely appointed with attractive oor and wall preparations, the environment may also inuence psychological factors that will increase favorable attitudes toward stair climbing that may generalize to other environments where stairs may be an option. We acknowledge that the notion that the environment has direct, unmediated, effects on behavior is contentious (cf. Bandura, 1996), and we will discuss it further in a latter section of this paper. We call our model the Ecological Model of Physical Activity (EMPA). In presenting the EMPA, we are not suggesting that researchers working within an ecological framework address each dimension and factor in the model. Rather, we hope that we have presented a conceptual framework that can guide our work whether we choose to use the complete model or only selected parts. The important point is that researchers have a clear understanding of the rationale behind collecting data from multiple sources and at multiple levels. At present, the EMPA will be most useful in identifying ecological determinants and correlates of physical activity in descriptive research. While this model could be used to guide intervention research, the paucity of evidence for the relationship between some aspects of the environment and physical activity may limit its effectiveness. Once more evidence becomes available, we should be able to identify specic characteristics of the relationships among factors in the model. We propose the following hypotheses to generate data to guide the future development of the model: Hypothesis 1. The inuence of more distal processes on physical activity should be buffered by proximal factors. For instance, low household SES should not inuence a childs physical activity if there is a quality physical education program in their school and support for physical activity at home. Similarly, children living in high SES households will be inactive if there is no support for physical active at home or in the school setting. Hypothesis 2. The environment exerts a direct effect on physical activity. This is a key proposition of ecological models. Thus, once individual dispositions such as cognitive constructs are accounted for, a relationship should still exist between extra-individual factors and physical activity. Hypothesis 3. Psychological factors mediate most of the relationship between extra-individual factors and physical activity. Hypothesis 4. Biological and genetic factors inuence physical activity participation. Hypothesis 5. The inuence of biological and genetic factors on physical activity are moderated by extra-individual factors such as physical ecology.

J.C. Spence, R.E. Lee / Psychology of Sport and Exercise 4 (2003) 724


Methodological considerations The EMPA should include data from both objective and subjective assessments. Objective data can be derived from both contextual (e.g., aggregated group data, neighborhood SES) and observational (e.g., number of joggers using a running path) sources. However, few tools for conducting objective assessments of the physical and social environments in relation to physical activity are available. Some scales have been developed to assess home (McKenzie et al., 1991), school (Sallis, Johnson, Calfas, Caparosa, & Nichols, 1997), and workplace settings (Plotnikoff, Fein, & Milton, 2001), but there is a need for more work in this area. Subjective data would include any assessments of individual dispositions (e.g., attitudes, perceived norms, self-efcacy) along with measures of perceived environment. The latter is important because it is possible that individual behavior patterns may be determined more by the perception of the environment than the environment itself (see Stokols, 1992; Wachs, 1999). Observational and subjective assessment tools are needed to determine the quality of interactions and support within and between settings described in the EMPA, because of the emphasis on the linkages and processes within and between settings. Collecting both subjective and objective data at both intra- and extra-individual levels may ameliorate some of the pitfalls observed in the public health arena. Group-level data analysis has been avoided in this area (see Macintyre & Ellaway, 2000), because of concerns surrounding the ecological fallacy (i.e., inferring individual-level relationships from relationships observed at the aggregate level). However, if individual-level data (i.e., physical activity, household income, educational level) is collected along with group-level data, as suggested by the EMPA, then ecological inference is not a problem (King, 1997). Because extra-individual effects may manifest themselves subtly and only be detectable after a long period of exposure, research is needed that assesses the longitudinal aspects of ecological models. These strategies must be sensitive to the multiple levels of inuence to avoid the trap of the ecological fallacy, and determine true ecological exposure effects. Spatial analyses using geographic information systems (GIS) show promise as a multidimensional venue for integrating the multiple levels (intra- and extra-individual) of ecological analyses (Moore & Carpenter, 1999). GIS allows data sources at different levels of analysis to be layered and examined simultaneously. There will need to be ongoing development of strategies and techniques to understand the dynamic and multilevel nature of ecological data. Future directions More than fty years ago, Lewin (195l/l997) proposed his idea of lifespace and that behavior was a function of the person and his or her environment. His work inuenced much of the thinking in environmental psychology (e.g., Barker, 1968) and led to the development of social ecological models for understanding behavior (e.g., Bronfenbrenner, 1977; McLeroy, Bibeau, Steckler & Glanz, 1988). In the health arena, researchers and practitioners have recognized a critical need to develop ecological strategies that include inuences beyond the individual (Green, Richard, & Potvin, 1996). The issue for exercise and health psychologists, who attempt to understand physical activity, is whether such approaches will be adopted in the eld. If researchers and practitioners in the area are to embrace ecological models, the following issues need to be addressed.


J.C. Spence, R.E. Lee / Psychology of Sport and Exercise 4 (2003) 724

While Sallis, Owen and Fotheringham (2000) dene behavioral science as an inclusive term that can encompass the activities of researchers in psychology, sociology, communication, and other elds (pp. 297298), the intervention literature in behavioral science has been dominated by psychological approaches (Green, 1984). The most popular health behavior theories, such as social cognitive theory (Bandura, 1986), theory of planned behavior (Ajzen, 1991), and transtheoretical model (Prochaska & Velicer, 1997), are psychological in nature. Throughout this manuscript we have maintained that to understand physical activity, more attention needs to be focused on physical and social environmental inuences that can affect behavior. In addition, the role of biological processes that interact with those extra-individual factors to moderate our behavior must also receive more attention. For instance, providing walking trails in a community populated predominantly with older adults may not be effective until the joint pain that many of the residents may experience is overcome. If, as McLeroy, Bibeau, Steckler, and Glanz (1988) maintain, the purpose of an ecological model is to focus attention on the environmental causes of behavior and to identify environmental interventions (p. 366), then perhaps we need to revise our denitions. A model such as McLeroy et al. described should be called an environmental model whereas one that situates the individual (including biological and psychological factors) within the broader environmental context should be called an ecological model. Further, researchers will need to spend time and energy fostering cross-disciplinary relationships, because conducting physical activity research with an ecological framework requires expertise that usually derives from more than one discipline (e.g., exercise science, public health, urban planning, geography, sociology). Researchers must overcome the parochial nature of disciplines and recognize that disciplines by denition usually focus on only one level within ecological frameworks (Cacioppo, Bernston, Sheridan, & McClintock, 2000). For instance, sociologists are concerned with macrosystem factors such as socioeconomic status and culture, whereas physiologists focus their work at the cellular level. Another challenge for ecological models will be to demonstrate direct, unmediated effects of the environment on physical activity. One of the holdovers from its behaviorist roots is the claim that, within an ecological framework, the environment exerts a direct effect on behavior (Barker, 1968; Sallis & Owen, 1997). This feature is, according to Sallis and Owen what distinguishes ecological models from intrapersonal theories, which sometimes hypothesize that selected environmental inuences are mediated through psychological processes (p. 412). An example of an intrapersonal theory that closely approximates an ecological approach is social cognitive theory (Bandura, 1986). Social cognitive theory recognizes personenvironmentbehavior interactions that are mediated by cognitive processes (i.e., efcacy beliefs). Further, similar to the biological factors identied in the EMPA, social cognitive theory hypothesizes that efcacy beliefs are driven in part by physiological cues. Bandura (1996) gives little credence to the idea that external inuences on behavior are unmediated by cognitive processes. Thus, the main difference between Banduras perspective on behavioral determinants and ecological approaches is the direct inuence of environment on behavior. Unless shown to be otherwise, theorists can claim that ecological models are just social cognitive theory incorporating multiple levels of environmental inuences (e.g., Ewart, 1991, 1993). In the general psychology literature, there is some evidence to indicate that unmediated inuences of external stimuli can take place on behavior (see Bargh & Ferguson, 2000), however this needs to be substantiated in the physical activity research domain. In fact much of the research that has collected data on environmental factors and potential cognitive

J.C. Spence, R.E. Lee / Psychology of Sport and Exercise 4 (2003) 724


mediators (e.g., self-efcacy) of physical activity has excluded these mediators from the nal analysis because they account for too much of the variance and overshadow the contribution of the environmental variables (e.g., Booth, Owen, Bauman, Clavisi, & Leslie, 2000; Leslie et al., 1999; Sallis et al., 1989). Informing intervention and policy Measurement of community and environmental indicators (Baker, Brennan, Brownson, & Houseman, 2000) will be important for informing policy and decision makers. One possible implication of ecological approaches is that neighborhood planners should consider creating spaces for physical activity. Despite the dearth of ecological data, this has already begun to happen with such policies as Californias Safe Routes to School Initiative (J. Corless, Surface Transportation Policy Project, Northern California, personal communication, January 19, 2001) that aims to increase the number of safe walking and cycling routes for children during to and from school travel. New communities may be designed and older communities modied to increase walking and cycling spaces for adults. Worksite policies may be implemented that include time and facilities at the worksite to increase physical activity. Further, neighborhood social context may need to be changed, for example, communities and policy makers may need to enhance neighborhood social capital and safety. Ethical considerations Numerous ethical considerations are implicated in ecological models that consider multiple levels of analysis. The tradition of health educators and outreach to individuals that has been employed by western medicine has always fostered great freedom and choice for individuals. Such traditional, individually based approaches rely on individual informed consent for participation. Research involving communities and environments does not have a forum for traditional participation consent procedures. Community and environmental approaches to changing individual behaviors have historically been the purview of public health and policy makers. These strategies are driven by the will of the majority rather than individual choice. Public health campaigns are typically implemented by government agencies that are comprised of decision makers that represent the good of the public. Policies are made via elected ofcials who represent the public good or via majority-voted referenda. Still, despite the implicit consent of the public majority, interventions that endeavor to manipulate social policy and human environments merit ethical concern. Changing policies and environments may not always produce the desired result. For example, installing neighborhood parks may give people a place to engage in physical activities. However, the same park may also give social miscreants a place to engage in illicit behaviors. Social policies and environmental changes must be considered carefully to ensure that they improve human health rather than endanger it. A second ethical issue lies in the sanctity of identifying information. In research on individuals, all identifying information is kept condential, to avoid stigmatization. Multilevel data that use geographic information from individual addresses must be kept condential, as must information about geographic spaces, or communities. Identifying information about communities must be considered with the same kind of care that information about individuals is given to avoid stigmat-


J.C. Spence, R.E. Lee / Psychology of Sport and Exercise 4 (2003) 724

ization and ill affects from residence in or near these communities. In addition to increasing stigma, research that identies communities as unsafe and impoverished may also increase problems with securing resources (e.g., adequate housing, Troutt, 1993), creating a cycle of disadvantage. Another potential ethical hazard of health research and promotion is the tendency for Blaming the Victim (Ryan, 1976). Victim blaming holds individuals to task for behavior that is largely dictated by environmental circumstances. For example, the failure of an intervention that implemented a walking program among individuals who lived in unsafe areas might result in part from the lack of safety, rather than aspects of the individual. Ecological models that specify the environment, as well as individual facilitators and barriers to physical activity participation, have a great opportunity to reduce this hazard. Ecological models propel physical activity researchers and promoters to examine both physical as well as social structures that may impact on physical activity participation. Culture that encompasses socially transmitted behaviors, beliefs, and institutions may be a determinant of physical activity participation (Siedentop, 1996). Culture stems from national, ethnic, and social backgrounds and will need to be dened and included in ecological models. At the same time, health researchers, promoters and policy makers will need to be responsive and sensitive to cultural traditions. Partnerships with community advisory boards may be one way to address this concern (Strauss et al., 2001). It is important to incorporate cultural traditions while maintaining respect for these traditions. Conclusions Physical inactivity is a major public health burden with estimates that upwards of two-thirds of the industrialized world does not achieve minimum physical activity guidelines (Craig, Russell, Cameron and Beaulieu, 1999; Oja, 1995; US Department of Health & Human Services, 1996). Based upon population-attributable risk estimates, approximately 20% of premature mortality could be avoided if everyone became physically active (Katzymaryk, Gledhill, & Shephard, 2000). This epidemic of inactivity has led theorists, researchers and practitioners to seek innovative models and methods to understand and increase physical activity. Traditional individually based approaches remain important and efcacious (Marcus & Forsyth, 1999); however, they are limited by their scope (Glasgow, Vogt & Boles, 1999). We have presented the theoretical background of ecological models of health behavior, and dened an ecological model for physical activity promotion. This model portrays physical activity behavior as being inuenced by interplay between environmental settings and biological and psychological factors. Further testing of this and existing ecological models of physical activity is recommended. Development of other conceptual models of physical activity is a next step, and this will include implementing innovative methodological strategies to evaluate ecological models. Ultimately, as Lewin (1951/1997) would concur, we need to get practical and start testing some ecological theories. Acknowledgements This research was supported in part by a grant from the Alberta Sport, Recreation, Parks and Wildlife Foundation. We thank Lise Gauvin and James Sallis for thoughtful comments on a previous version of this manuscript.

J.C. Spence, R.E. Lee / Psychology of Sport and Exercise 4 (2003) 724


Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179211. Astrand, P. O. (1994). Physical activity and tness: Evolutionary perspective and trends for the future. In C. Bouchard, R. J. Shephard, & T. Stephens (Eds.), Physical activity, tness, and health: International proceedings and consensus statement (pp. 98105). Champaign, IL: Human Kinetics. Baker, E. A., Brennan, L. K., Brownson, R., & Houseman, R. A. (2000). Measuring the determinants of physical activity in the community: Current and future directions. Research Quarterly for Exercise and Sport, 71(2 (Suppl)), S146158. Bandura, A. (1986). The explanatory and predictive scope of self-efcacy theory. Journal of Social and Clinical Psychology, 4, 359373. Bandura, A. (1996). Ontological and epistemological terrains revisited. Journal of Behavior Therapy and Experimental Psychiatry, 27, 323345. Baranowski, T., Anderson, C., & Carmack, C. (1998). Mediating variable framework in physical activity interventions: How are we doing? How might we do better? American Journal of Preventive Medicine, 15(4), 266297. Bargh, J. A., & Ferguson, M. J. (2000). Beyond behaviorism: On the automaticity of higher mental processes. Psychological Bulletin, 126, 925945. Barker, R. J. (1968). Ecological psychology. Stanford, CA: Stanford University Press. Baron, R. M., & Kenny, D. A. (1986). The moderatormediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 11731182. Beunen, G., & Thomas, M. (1999). Genetic determinants of sports participation and daily physical activity. International Journal of Obesity and Related Metabolic Disorders, 23(Suppl 3), S55S63. Blakely, T. A., & Woodward, A. J. (2000). Ecological effects in multi-level studies. Journal of Epidemiology and Community Health, 54, 367374. Booth, M. L., Owen, N., Bauman, A., Clavisi, O., & Leslie, E. (2000). Socialcognitive and perceived environment inuences associated with physical activity in older Australians. Preventive Medicine, 31, 1522. russe, L. (1997). Genetics of tness and physical performance. Champaign, IL: Bouchard, C., Malina, R. M., & Pe Human Kinetics. Bouchard, C., & Perusse, L. (1994). Heredity, activity level and tness, and health. In C. Bouchard, R. J. Shephard, & T. Stephens (Eds.), Physical activity, tness, and health: International proceedings and consensus statement (pp. 106118). Champaign, IL: Human Kinetics. Bouchard, C., & Shephard, R. J. (1994). Physical activity, tness, and health: The model and key concepts. In C. Bouchard, R. J. Shephard, & T. Stephens (Eds.), Physical activity, tness, and health: International proceedings and consensus statement (pp. 7788). Champaign, IL: Human Kinetics. Bronfenbrenner, U. (1977). Toward an experimental ecology of human development. American Psychologist, 32, 513531. Bronfenbrenner, U. (1989). Ecological systems theory. Annals of Child Development, 22, 723742. Cacioppo, J. T., Bernston, G. G., Sheridan, J. F., & McClintock, M. K. (2000). Multilevel integrative analyses of human behavior: Social neuroscience and the complementing nature of social and biological approaches. Psychological Bulletin, 126, 829843. Cohen, D. A., Scribner, R. A., & Farley, T. A. (2000). A structural model of health behavior: A pragmatic approach to explain and inuence health behaviors at the population level. Preventive Medicine, 30, 146154. Cordain, L., Gotschall, R. W., Eaton, S. B., & Eaton, S. B. (1998). Physical activity, energy expenditure and tness: An evolutionary perspective. International Journal of Sports Medicine, 19, 328335. Craig, C. L., Russell, S. J., Cameron, C., & Beaulieu, A. (1999). Foundations for joint action: Reducing physical inactivity. Ottawa, ON: Canadian Fitness and Lifestyle Research Institute. Diez-Roux, A. V., Nieto, J., Muntaner, C., Tyroler, H. A., Comstock, G. W., Shahar, E., Cooper, L. S., Watson, R. L., & Szklo, M. (1997). Neighborhood environments and coronary heart disease: A multilevel analysis. American Journal of Epidemiology, 146, 4863. Dishman, R. K., & Buckworth, J. (1996). Increasing physical activity: A quantitative synthesis. Medicine and Science in Sports and Exercise, 28, 706719.


J.C. Spence, R.E. Lee / Psychology of Sport and Exercise 4 (2003) 724

Dzewaltowski, D. A. (1997). The ecology of physical activity and sport: Merging science and practice. Journal of Applied Sport Psychology, 9, 254276. Ecob, R., & Macintyre, S. (2000). Small area variations in health related behaviours: Do these depend on the behaviour itself, its measurement, or on personal characteristics? Health & Place, 6, 261274. Ellaway, A., & Macintyre, S. (1996). Does where you live predict health related behaviours?: A case study in Glasgow. Health Bulletin (Edinb), 54, 443446. Ewart, C. K. (1991). Social action theory for public health psychology. American Psychologist, 46, 931946. Ewart, C. K. (1993). Health promotion and disease prevention: A social action conception of compliance behavior. In N. A. Krasnegor, L. H. Epstein, S. B. Johnson, & S. J. Yaffe (Eds.), Developmental aspects of health compliance behavior (pp. 251280). Hillsdale, NJ: Lawrence Erlbaum. Frankish, C. J., Milligan, C. D., & Reid, C. (1998). A review of relationships between active living and determinants of health. Social Science & Medicine, 47, 287301. Garcia, A. W., Broda, M. A. N., Frenn, M., Coviak, C., Pender, N. J., & Ronis, D. L. (1995). Gender and developmental differences in exercise beliefs among youth and prediction of their exercise behavior. Journal of School Health, 65, 213219. Garcia Bengoechea, E., & Johnson, G. (2000). Ecological systems theory and childrens development in sport: Toward a process-person-context-time research paradigm. Avante, 7, 2031. Gauvin, L., Levesque, L., & Richard, L. (2001). Helping people initiate and maintain a more active lifestyle: A public health framework for physical activity promotion research. In R. N. Singer, H. A. Hausenblas, & C. N. Janelle (Eds.), Handbook of sport psychology (2nd ed.) (pp. 718739). New York: Wiley. Gauvin, L., Richard, L., Potvin, L., Craig, C. L. Spence, J. C., Quinney, H. A., & Dassa, C. (2001, March). Multilevel determinants of physical activity: Individual- and neighbourhood-level inuences. Poster presented at the 22nd Society for Behavioral Medicine meeting, Seattle, WA. Glasgow, R. E., Vogt, T. M., & Boles, S. M. (1999). Evaluating the public health impact of health promotion interventions: The REAIM framework. American Journal of Public Health, 89, 13221327. Glassford, R. G. (1976). Application of a theory of games to the transitional Eskimo culture. New York: Arno Press. Godin, G., & Kok, G. (1996). The theory of planned behavior: A review of its applications to health-related behaviors. American Journal of Health Promotion, 11, 8798. Goran, M. I., Gower, B. A., Nagy, T. R., & Johnson, R. (1998). Developmental changes in energy expenditure and physical activity in children: Evidence for a decline in physical activity in girls before puberty. Pediatrics, 101, 887891. Green, L. W. (1984). Modifying and developing health behavior. Annual Review of Public Health, 5, 215236. Green, L. W., & Kreuter, M. (1991). Health promotion planning: An educational and environmental approach. Mountain View, CA: Mayeld. Green, L. W., Richard, L., & Potvin, L. (1996). Ecological foundations of health promotion. American Journal of Health Promotion, 10, 270281. Katzymaryk, P. T., Gledhill, N., & Shephard, R. J. (2000). The economic burden of physical inactivity in Canada. Canadian Medical Association Journal, 163, 14351440. Kelly, J. G. (1990). Changing contexts and the eld of community psychology. American Journal of Community Psychology, 18, 769792. King, G. (1997). A solution to the ecological inference problem: Reconstructing individual behavior from aggregate data. Princeton, NJ: Princeton University Press. Kraemer, H. C., Stice, E., Kazdin, A., Offord, D., & Kupfer, D. (2001). How do risk factors work together? Mediators, moderators, and independent, overlapping, and proxy risk factors. American Journal of Psychiatry, 158, 848856. Laitakari, J., Vuori, I., & Oja, P. (1996). Is long-term maintenance of health-related physical activity possible? An analysis of concepts and evidence. Health Education Research, 11, 463477. Leslie, E., Owen, N., Salmon, J., Bauman, A., Sallis, J. F., & Lo, S. K. (1999). Insufciently active Australian college students: Perceived personal, social, and environmental inuences. Preventive Medicine, 28, 2027. Levine, M., Toro, P. A., & Perkins, D. V. (1993). Social and community interventions. Annual Review of Psychology, 44, 528558. Lewin, K. (1997). Field theory in social science: Selected theoretical papers. Washington, DC: American Psychological Association. (Original work published in 1951).

J.C. Spence, R.E. Lee / Psychology of Sport and Exercise 4 (2003) 724


Linn, W. S., & Gong, H. (1999). Air pollution, exercise, nutrition, and health. In J. M. Rippe (Ed.), Lifestyle medicine (pp. 12151226). Malden, MA: Blackwell. Lynch, J. W., Kaplan, G. A., & Salonen, J. T. (1997). Why do poor people behave poorly? Variation in adult health behaviours and psychosocial characteristics by stages of the socioeconomic lifecourse. Social Science and Medicine, 44, 809819. Macintyre, S., & Ellaway, A. (2000). Ecological approaches: Rediscovering the role of the physical and social environment. In L. F. Berkman, & I. Kawachi (Eds.), Social epidemiology (pp. 332348). New York: Oxford University Press. Marcus, B. H., & Forsyth, L. H. (1999). How are we doing with physical activity? American Journal of Health Promotion, 14, 118124. McAuley, E., & Mihalko, S. L. (1998). Measuring exercise-related self-efcacy. In J. L. Duda (Ed.), Advances in sport and exercise psychology measurement (pp. 370390). Morgantown, WV: Fitness Information Technology. McGue, M., & Bouchard, T. J. (1998). Genetic and environmental inuences on human behavioral differences. Annual Review of Neuroscience, 21, 124. McKenzie, T. L., Sallis, J. F., Nader, P. R., Patterson, T. L., Elder, J. P., Berry, C. C., Rupp, J. W., Atkins, C. J., Buono, M. J., & Nelson, J. A. (1991). BEACHES: An observational system for assessing childrens eating and physical activity behaviors and associated events. Journal of Applied Behavior Analysis, 24, 141151. McLeroy, K. R., Bibeau, D., Steckler, A., & Glanz, K. (1988). An ecological perspective on health promotion programs. Health Education Quarterly, 15, 351377. Moore, D. A., & Carpenter, T. E. (1999). Spatial analytical methods and geographic informations systems: Use in health research and epidemiology. Epidemiologic Reviews, 21, 143161. OBrien Cousins, S. (1998). Exercise, aging, and health: Overcoming barriers to an active old age. Bristol, PA: Brunner/Mazel. Oja, P. (1995). Descriptive epidemiology of health-related physical activity and tness. Research Quarterly for Exercise and Sport, 214, 303312. Orleans, T. C. (2000). Promoting the maintenance of health behavior change: Recommendations for the next generation of research and practice. Health Psychology, 19, 7683. Owen, N., Leslie, E., Salmon, J., & Fotheringham, M. J. (2000). Environmental determinants of physical activity and sedentary behavior. Exercise and Sport Sciences Reviews, 28, 153158. Plotnikoff, R. C., Fein, A., & Milton, L. (2001, December). An ecological perspective on workplace active living. Poster presented at the joint Centers for Disease Control and Health Canada conference, Communicating Physical Activity and Health Messages: Science into Practice, Whistler, BC, Canada. Powell, K. E., Kreuter, M. W., Stephens, T., Marti, B., & Heineman, L. (1991). The dimensions of health promotion applied to physical activity. Journal of Public Health Policy, 12, 492509. Prochaska, J. O., & Velicer, W. F. (1997). The Transtheoretical Model of health behavior change. American Journal of Health Promotion, 12, 3848. Richard, L., Potvin, L., Kishchuk, N., Prlic, H., & Green, L. W. (1996). Assessment of the integration of the ecological approach in health promotion programs. American Journal of Health Promotion, 10, 318328. Rode, A., & Shephard, R. J. (1994). Physiological consequences of accultration: A 20-year study of tness in an Inuit community. European Journal of Applied Physiology, 69, 516524. Rowland, T. W. (1998). The biological basis of physical activity. Medicine and Science in Sports and Exercise, 30(3), 392399. Ryan, W. (1976). Blaming the victim. New York: Vintage Books. Sallis, J. F., Bauman, A., & Pratt, M. (1998). Environmental and policy interventions to promote physical activity. American Journal of Preventive Medicine, 15, 379397. Sallis, J. F., Hovell, M. F., Hofstetter, C. R., Faucher, P., Elder, J. P., Blanchard, J., Caspersen, C. J., Powell, K. E., & Christenson, G. M. (1989). A multivariate study of determinants of vigorous exercise in a community sample. Preventive Medicine, 18, 2034. Sallis, J. F., Johnson, M. F., Calfas, K. J., Caparosa, S., & Nichols, J. F. (1997). Assessing perceived physical environmental variables that may inuence physical activity. Research Quarterly for Exercise and Sport, 68, 345351. Sallis, J. F., & Owen, N. (1997). Ecological models. In K. Glanz, F. M. Lewis, & B. K. Rimer (Eds.), Health behavior and health education: Theory, research, and practice (2nd ed.) (pp. 403424). San Francisco: Jossey-Bass.


J.C. Spence, R.E. Lee / Psychology of Sport and Exercise 4 (2003) 724

Sallis, J. F., & Owen, N. (1999). Physical activity and behavioral medicine. Thousand Oaks, CA: Sage Publications. Sallis, J. F., Owen, N., & Fotheringham, M. J. (2000). Behavioral epidemiology: A systematic framework to classify phases of research on health promotion and disease prevention. Annals of Behavioral Medicine, 22, 294298. Sampson, R. J., Raudenbush, S. W., & Earls, F. (1997). Neighborhoods and violent crime: A multilevel study of collective efcacy. Science, 277, 918924. Siedentop, D. (1996). Valuing the physically active life: Contemporary and future directions. Quest, 48, 275289. Skinner, B. F. (1954). Science and human behavior. New York: MacMillian. Sonstroem, R. J., & Morgan, W. P. (1989). Exercise and self-esteem: Rationale and model. Medicine and Science in Sports and Exercise, 21, 329337. Spence, J. C., Courneya, K. S., Blanchard, C., & Wilson, P. (2000). The theory of planned behavior and physical activity: A quantitative review of explained variation in intention and behavior [Abstract]. International Journal of Behavioral Medicine, 7, 146. Stokols, D. (1992). Establishing and maintaining healthy environments: Toward a social ecology of health promotion. American Psychologist, 47, 622. Stokols, D. (1996). Translating social ecological theory into guidelines for community health promotion. American Journal of Health Promotion, 10, 282298. Strauss, R. P., Sengupta, S., Quinn, S. C., Goeppinger, J., Spaulding, C., Kegeles, S. M., & Millett, G. (2001). The role of community advisory boards: Involving communities in the informed consent process. American Journal of Public Health, 91, 19381943. Susser, M., & Susser, E. (1996). Choosing a future for epidemiology: II. From black box to Chinese boxes and ecoepidemiology. American Journal of Public Health, 86, 674677. Treloar, C., Porteous, J., Hassan, F., Kasniyah, N., Lakshmanudu, M., Sama, M., Sjabani, M., & Heller, R. F. (1999). The cross cultural context of obesity: An INCLEN multicentre collaborative study. Health & Place, 5, 279286. Troutt, D. D. (1993). The thin red line: How the poor still pay more. San Francisco, CA: West Coast Regional Ofce of Consumers Union. U.S. Department of Health & Human Services (1996). Physical activity and health: A report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion. Villeneuve, P. J., Morrison, H. I., Craig, C. L., & Schaubel, D. E. (1998). Physical activity, physical tness, and risk of dying. Epidemiology, 9, 626631. Wachs, T. D. (1992). The nature of nurture. Newbury Park, CA: Sage. Wachs, T. D. (1999). Celebrating complexity: Conceptualization and assessment of the environment. In S. L. Friedman, & T. D. Wachs (Eds.), Measuring environment across the life span: Emerging methods and concepts (pp. 357392). Washington, DC: American Psychological Association. Welk, G. J. (1999). The youth physical activity promotion model: A conceptual bridge between theory and practice. Quest, 51, 523. Wenger, C. B. (1999). Physiologic and pathologic responses to heat stress. In J. M. Rippe (Ed.), Lifestyle medicine (pp. 11711185). Malden, MA: Blackwell. World Health Organization (1997). The Heidelberg guidelines for promoting physical activity among older persons. Journal of Aging and Physical Activity, 5, 18. Yen, I. H., & Kaplan, G. A. (1998). Poverty area residence and changes in physical activity level: Evidence from the Alameda County Study. American Journal of Public Health, 88, 17091712.