Current recommendations for physical activity (1) emphasize “lifestyle physical activity … the daily accumulation of 30 minutes of self-selected activities, which includes all leisure, occupational, or household activities that are at least moderate to vigorous in their intensity, and could be planned or unplanned activities, that are a part of daily life” (2). For persons with arthritis, specific physical activity objectives in the National Arthritis Action Plan advocate increasing the proportion who engage in any leisure time physical activity, sustained physical activity for at least 30 minutes daily, and/or vigorous physical activity on 3 or more days per week (3).
In light of the focus on lifestyle physical activity, a recent paradigm shift broadens individualized exercise prescription to encompass the community-based promotion of moderate daily activities. Physical activity promotion in the general population, and among persons with arthritis, has been dominated by a person-focused rather than a community-focused approach; however, recent attention has been directed toward developing community- based interventions that include attention to policy and environmental factors. This expanded, public health approach facilitates meeting the physical activity recommendations of the Centers for Disease Control and Prevention and the American College of Sports Medicine (4, 5). The logic behind this approach, predicated on other public health successes, is that larger numbers of people can be reached in the community, as compared with health care settings, including the most sedentary (6). Additional support for the need to create supportive community environments is found in the fact that many adults prefer to engage in physical activity outside of structured settings (5) and therefore may not seek out aerobics classes, sports teams, or exercise facilities.
Drawing conclusions concerning the effectiveness of community-based physical activity interventions is hampered by the wide array of approaches, settings, formats, and measures in the published literature. Most of the research on increasing physical activity comes from person- focused projects targeting determinants of physical activity/exercise adoption, including social support, past experiences with exercise, self efficacy beliefs, perceived lack of time, perceived benefits and barriers, intention to exercise, and enjoyment (6, 7). Applied research in communities to test a broader, public health approach that also includes attention to policy and environmental factors is in the early stages (8). Evidence of successful components of community-based physical activity programs among diverse groups may have applicability to program planning for persons with arthritis.
Definition of community and community-based intervention
“Community” and “community-based intervention” have many definitions (9). For the practical purpose of program implementation, community is often defined as a geographic area, such as a county, city, or census tract with clear geopolitical boundaries. “Community” may also describe a set of relationships among people who do not necessarily reside in close proximity but share a common identity, goal, or set of values. The label of “community- based intervention” has been attached to diverse programs that vary in terms of citizen ownership and involvement, change strategies, objectives, and focus. The term “community based” should be distinguished from “population based.” A community-based intervention may be provided to an entire population or to a smaller, defined group of participants (e.g., members of a senior center) in a community setting. Thus, community-based refers to the context or setting whereas population based refers to the universe of participants. Additionally, community-based interventions can encompass all levels of prevention—primary, secondary, and tertiary.
Review of the literature
Interventions can be categorized along an array of dimensions describing context, format, participants, and process. To facilitate the summary of a vast literature, four frameworks or “ideal types” are described below, along with evidence for program effectiveness. These frameworks, however, are best viewed heuristically because many interventions blend strategies from more than one framework.
Community-based, person focused interventions.
Some community-based programs use methods and strategies employed in health care settings for patient education and counseling (such as one-to-one counseling, classroom instruction, self paced instruction, cognitive–behavioral strategies, skills building, and print materials), but base these services in nonclinical settings, such as recreation departments, churches, senior centers, and shopping malls. As with patient education and rehabilitation in clinical settings, the intervention's aims are to increase knowledge, change attitudes, build skills, and influence behavior change; thus the “community” aspect refers mainly to the setting and broader availability. Although these techniques predominate person-focused interventions, they may also be combined with other broader based methods in multicomponent programs (i.e., socioecologic programs and risk factor reduction programs as described below).
Typically, researchers or other recreation professionals determine the objectives, content, and format of person-focused programs, often with guidance from published literature or consumer needs assessments. Professionals, paraprofessionals, or lay persons provide the interventions, which may focus on persons with a particular characteristic, illness, or risk factor. The format often conforms to a one-to-one or small group health care delivery model, even though the services are being provided outside of the clinic (6). These interventions may include primary or secondary prevention, but typically focus on physical activity as means of chronic disease self management and disability prevention (tertiary prevention or downstream strategies) (10). The majority of physical activity interventions in the community for persons with arthritis and other rheumatic conditions have been of this type.
The theories frequently applied to developing person-focused interventions address the motivations and perceptions of individuals, with some attention to the influence of the social environment on an individual's beliefs and attitudes. Two approaches that have shown some success in increasing physical activity frequency are the Transtheoretical Model, which presents 5 stages of psychological readiness for behavior change (precontemplation, contemplation, preparation, action, and maintenance) for matching strategies to the individual's stage of readiness (11); and Social Cognitive Theory, which describes personal, behavioral, and environmental factors as reciprocally interacting determinants of the initiation and maintenance of a behavior. The focus is on a person's belief in his or her ability to perform a behavior (self efficacy) and intervening influences (e.g., the influence of social models, the reinforcement from initial success with the behavior) (12). Intervention strategies suggested by Social Cognitive Theory (generically called cognitive–behavioral strategies) include goal setting, self monitoring (i.e., logs and diaries), stimulus control strategies (i.e., strategic management of stimuli), social modeling (i.e., successful role models perceived to be similar to the observer), and reinforcement (i.e., incentives/rewards/recognition of success). Other theories of individual motivation, including the theory of reasoned action, locus of control, and the health belief model have also been applied without compelling evidence of success (13).
Person-focused interventions of relatively short duration (e.g., 12 weeks), have been found to increase physical activity frequency 10–75% compared with control groups; however, the impact on duration and intensity is unknown (13). A number of studies have successfully applied Social Cognitive Theory or specific components of it to physical activity promotion (14–17); sometimes combined with relapse prevention strategies (13). Application of the Transtheoretical Model to physical activity has shown that matching intervention strategies to stage of readiness for change is also superior to a no-intervention control condition (18–24). Increases in physical activity that result from person-focused interventions are typically not sustained over time (13, 25); however, some home based interventions with followup contacts have shown good maintenance of physical activity levels at 6 months, 1 year, or 2 years (26–29).
A review of 29 community-based, person-focused interventions (at home or community centers) for older adults revealed higher exercise participation rates than are typically found for younger persons (mean, 75% participation) (30). The most successful studies used goal setting, self monitoring, feedback, support, and relapse prevention training. Programs that used supervised, home based formats, alone or in addition to group formats, showed adherence levels equal to or exceeding adherence to class or group formats alone (30).
Worksite based physical activity promotion programs, a subset of community-based programs, usually combine education and counseling, exercise classes, and risk factor screening, sometimes including broader changes such as organizational policies (e.g., flex time, incentives) and building of exercise facilities. Some worksite programs have shown at least short-term effectiveness in increasing employees' physical activity levels (31, 32); however, too few comprehensive evaluations of worksite-based programs have been conducted to conclude what components are most effective or how long-term maintenance of physical activity is affected (5, 10).
Most physical activity promotion programs among youth have involved physical education classes in schools. Few community-based programs have been evaluated and have shown little significant improvements in physical activity or physical fitness, although some improvements in attitudes and intentions have been found (33, 34).
Taylor et al (35) reviewed the literature on interventions that included physical activity alone or in combination with other health-promoting behaviors (such as a healthy diet) with high-risk populations, including ethnic minorities. The interventions included home-based strategies with telephone and mail contact, and family-oriented, church-based, and community-wide formats, and were typically of short duration (less than 15 weeks). Most assessed the population's needs and preferences and had a community advisory panel. Community members delivered the intervention in some studies. Two studies found no improvements in physical activity (36, 37), 4 studies reported mixed results (38, 39) or changes only in specific subgroups (40, 41). In 1 study, better adherence and higher physical activity levels were found in communities with strong organization and committed local leaders (41). Two interventions reporting the most consistent, positive results for physical activity were weight loss interventions for minority women that included physical activity promotion (42, 43). Although it was not possible to conclude what factors contributed to success, the reviewers identified 2 factors with promise: a thorough assessment of needs and preferences prior to the program's implementation, and meaningful community participation and involvement.
Community-based physical activity interventions created specifically for persons with arthritis have included various person-focused approaches. A home-based exercise therapy program resulted in improved functional outcomes but no significant differences in overall physical activity level between treatment versus control participants (44). Community-based land and aquatic programs developed by the Arthritis Foundation (AF) have resulted in improvements in functional and psychosocial outcomes. See Boutaugh's article in this issue for a review of AF programs (45). State health departments are currently pilot testing community-based physical activity interventions for persons with arthritis based on a team concept. See Brady and Sniezek's article in this issue for a description of the It's Your Life, Join the Movement program (46).
Physical activity is included in the multicomponent Arthritis Self-Management Program (ASMP), a widely disseminated, volunteer-led program adopted by the AF. Although the ASMP has resulted in improvements in other health outcomes, 4-month changes in aerobic exercise were not significant for 3-week and 6-week versions of the program. Four-month changes in stretching exercise were significant for the 6-week program (47).
Some clinic-based, supervised programs for persons with arthritis have included educational components designed to encourage self-directed physical activity in the home/community after several weeks of intervention (48, 49). One-year followup after a clinic-based, supervised fitness walking program did not reveal long-term adherence (50). Persons with arthritis who make “an uninterrupted transition from supervised to self-directed” physical activity appear to be more successful at long-term maintenance (49).
Community-based risk factor reduction.
Community- based risk factor reduction interventions are comprehensive in scope. These interventions (referred to as midstream interventions) (10) involve a variety of methods and settings and often tailor intervention components to the special needs of subgroups, such as children, older adults, and ethnic groups. In addition to face-to-face instruction and educational materials, these interventions employ mass media and social marketing techniques, risk factor screening, improved access to programs and services, community improvements (such as the addition of parks and walking paths), and supportive components such as walking clubs and worksite incentive programs. Activities take place in multiple settings (places of worship, schools, worksites, civic organizations, shopping malls); thus, the purview for intervention expands into the social context. In these programs, theories addressing individual motivations often have been adapted or combined with macrolevel perspectives, drawing from diverse disciplines (6, 51). What characterizes this approach is the goal of identifying elevated risk (e.g., health screening).
Promoting physical activity through media messages is a common technique. Message development may be based on communication theories, including factors most likely to affect the success of the health message, such as characteristics of the message source and receiver, the format and content of the message itself, and characteristics of the available media channels (newspaper, television, radio, etc.) (52). Communication theory may be combined with social marketing principles to develop messages that are tailored to consumer preferences. The Social Marketing approach to physical activity promotion uses principles derived from basic product marketing (53). Programs are developed to address the needs and preferences of a consumer of physical activity behaviors. Use of communication theory and social marketing principles are not limited to those community-based interventions categorized here as community risk factor reduction.
A review of 28 mass media campaigns to increase physical activity and information technology found that although recall of campaign messages is often very high (70% recall), these efforts have had very little impact on physical activity behavior (54). The most effective campaigns were of relatively smaller scale, included frequent contact with the audience, and used messages tailored to the perceptions and needs of specific subgroups. Mass media campaigns have appeal for conveying physical activity recommendations to the public on a large scale; however, they are most effectively used within a broader approach to create an identity or umbrella for comprehensive community interventions (55). As stand alone initiatives, their impact is not likely to be cost effective in physical activity promotion. Mass media campaigns are difficult to rigorously evaluate, and many empirical questions remain concerning their development and application (54). For a description of a campaign targeting physical activity for persons with arthritis, see Brady and Sniezek's article in this issue (46).
Much of what is known about this approach comes from large-scale cardiovascular risk reduction studies, such as the Minnesota Heart Health Project (56), the Stanford Five-City Project (57), and the Pawtucket Heart Health Project (58). In these studies (midstream interventions) physical activity is usually one of several promoted behaviors. Researchers or other health professionals typically determine the goals, objectives, content, and format of the intervention, often with input from a local advisory group of professionals and citizens from the community.
The largest and most comprehensively evaluated community-based interventions to promote physical activity have been cardiovascular disease risk reduction projects. During 6 years of the project in the intervention and comparison communities, the Stanford Five-City Project used print and electronic media, informational presentations, neighborhood-based walking events and lay-led walking groups, and worksite and school-based promotions. Treatment effects were modest at best. The researchers concluded that “… the educational intervention had little, if any, impact on physical activity” (57). Lack of a behavioral impact was attributed to a weak intervention (insufficient contact with residents over the various intervention channels, inadequate segmentation of population subgroups with appropriate tailoring of messages) and/or insensitive measures of physical activity. Of note is that physical activity–related messages comprised only about 8% of the total educational content provided (57).
The Pawtucket Heart Health program provided 7 years of programming for cardiovascular risk reduction (58). Physical activity programming involved partnerships with 500 local organizations, including schools, civic groups, government, churches, and worksites. Educational programs involved exercise offerings, such as walking clubs and fitness trails; a 6-week exercise campaign at worksites; and a campaign targeting sedentary individuals, based on the Transtheoretical Model of behavior change. No differences in attempts to increase physical activity or in the prevalence of inactivity were found between the intervention and comparison communities. The researchers suggested that modest penetration of the intervention activities led to the null findings (58).
The Minnesota Heart Health Program (55) advocated physical activity along with other cardiovascular risk reduction activities in 3 intervention communities paired with 3 comparison communities. Methods included individual, group, and community-wide strategies, such as mass media and persuasive communication, community planning and participation, risk factor screening, school-based education, face-to-face adult education and counseling, and the use of influential role models and opinion leaders. Cohort data over the 6 communities revealed an increase in physical activity in all of the communities, with the intervention communities slightly exceeding the comparison communities at the last followup survey. A small increase in daily kilocalories expended was found in the early years, which decreased in later years. The amount of time spent in vigorous activities decreased slightly in the intervention communities. The researchers noted positive secular trends during the study period for physical activity and concluded that “the exposure data suggest that the Minnesota Heart Health Program may not have added a great deal to the level of risk reduction activity that would have been expected without the program” (56).
Community planning for policy and environmental changes.
An emerging, public health/urban planning hybrid framework for promoting physical activity in the community has as its focus changes in the community environment itself (59). The community planning approach (an upstream intervention) (10) seeks to reduce barriers and create supports for physical activity through changes in public policy and the environment. The impetus behind this approach is an increasingly sedentary and automobile dependent populace, arising from a transportation environment that is friendly to automobiles but hostile to pedestrians and cyclists and a land use model that isolates people in homes distant from work and services, necessitating multiple trips by car for work, recreation, shopping, and school (59). This state of affairs is thought to be especially detrimental to public health given the broad appeal and potential for walking as a form of physical activity. Walking is commonly found to be the most popular form of physical activity for people of all socioeconomic statuses (60).
Sallis et al (61) have suggested 4 categories for attention in policy and environmentally focused community interventions: 1) the natural environment, such as weather and terrain; 2) the constructed environment, such as buildings, trails, parks, and worksite amenities; 3) policies related to incentives, such as subsidized health club memberships or incentives for bicycle commuting; and 4) policies related to resources and infrastructure, such as accessible stairwells, funding for trails and recreation areas, and sidewalk ordinances.
In addition to land use and transportation planning, intervention methods to change policies and environments may include coalition building, policy and media advocacy, community revitalization efforts, and crime prevention. The goals and objectives of policy and environmentally focused interventions may be professionally initiated; however, macrolevel efforts typically require a coalition of both professionals from the public and private sectors and concerned citizens working in partnership to engage the political process for community improvements. Physical activity promotion may be one objective, along with other complementary objectives such as air quality and urban sprawl control, all subsumed under an umbrella of enhanced quality of life and sustainable community development.
Because the research has just begun in recent years, the evidence for policy and environment focused interventions is scant (61). Only recently have experts begun the process of identifying which aspects of the physical environment merit intervention. For example, the Active Community Environments initiative at the Centers for Disease Control and Prevention has sponsored a working paper that reviews the impact of land use and transportation planning on public health, describing how these factors affect the convenience, safety, and viability of pedestrian and bicycle use both for recreation and travel utility (60). The weight of the available evidence suggests that aspects of land development patterns and transportation systems do influence physical activity behaviors; however, establishing cause and effect associations, disentangling the effects of single features versus sets of features, and controlling for a variety of competing influences has proven complex and challenging (60). For example, mixed land use, higher neighborhood density (the “traditional” urban neighborhood), mass transit availability, proximity of exercise facilities, sidewalks, and other pedestrian-friendly amenities have been associated with higher rates of walking (61, 62). Data from 5 states reveal that the perceived safety of one's neighborhood is associated with physical activity level (63). “Inherent difficulties of evaluation” have hampered the research on community-based interventions with a policy/environment focus (61). Environmental changes on a small scale, such as posting signs to encourage stair use instead of the elevator, have resulted in increases of 5–18% while the sign was posted (64, 65). Larger-scale interventions focused on environmental changes have provided suggestive evidence of increases in physical activity behaviors, weaknesses in the research designs, however, limit their interpretation (66–68).
A model that has garnered much attention in the last decade in the community health promotion and disease prevention literature, and more recently in the area of physical activity promotion, is the socioecologic model. The socioecologic model conceptualizes health behavior as being affected by multiple levels of influence (69, 70): intrapersonal factors, such as knowledge, attitudes, and skills of the individual; interpersonal factors, such as social support systems of family, friends, and peers at the workplace; institutional factors, including organizational characteristics and rules and norms for conduct; community factors, such as the nature of relationships among multiple sectors of community life; and public policy and law at the local, state, and national levels. Community-based interventions are assumed to be most successful if changes are promoted at multiple levels of this system, from person oriented interventions to public policy (69). A panel of nutrition and physical activity experts recently delineated a similar type of framework specifically for interventions targeting physical activity and food choices, incorporating multiple layers of leverage points for intervention (71).
For such massive efforts to succeed, citizen participation and ownership at the local level are essential. Community-based interventions following a socioecologic approach are most likely to be characterized by a partnership between professionals and local leaders (72).
For the ON THE MOVE! Project, the state of California funded 9 agencies to promote physical activity among multiethnic communities using a socioecologic approach (73). Local ON THE MOVE! programs sponsored projects based on local needs and interests, including ethnically based aerobic dance, walking clubs and walkathons, community gardens, walk-and-talks with community leaders, and worksite programs, community health advocates, lay exercise leaders, and local government changes. Communities were required to use a “spectrum of prevention” approach addressing individual knowledge and skills development, community education, service provider education, coalition and network fostering, changes in organizational practices and policy, and legislation influence. Local coalitions tailored the intervention to fit the community's characteristics (73). Although a rigorous evaluation of impacts has not been published for these projects, detailed case studies and process evaluation appear in a March/April 1999 supplement to the Journal of Health Education.
The Sumter County Active Lifestyles project is a physical activity intervention in progress, based on the socioecologic model, which is underway in South Carolina (72). A multisectorial community coalition, in partnership with a county health department and Prevention Research Center, has conducted a needs and assets assessment, held a community visioning workshop, created objectives, and begun implementation of multiple strategies. The project focuses on multiple levels of influence, including policy and environment. Strategies include walking groups and walking promotional materials, sidewalk audits, park revitalization projects, worksite program and policy development, community education, media advocacy, and policy advocacy. The project has a quasiexperimental design and a qualitative process evaluation.
The following general conclusions can be drawn from this overview of community-based interventions.
Person-focused interventions of short duration are effective in increasing physical activity behaviors in the short term. Cognitive–behavioral strategies are most effective. Supervised, home-based formats are as effective as group formats.
Long-term maintenance of physical activity after most interventions is poor. The transition from clinic-based, supervised exercise for persons with arthritis to self- directed physical activity in the community frequently leads to a drop in adherence. Behavioral maintenance is enhanced by relapse prevention strategies and long-term followup contacts with participants.
Program success is enhanced by a thorough assessment of community needs, assets, and preferences and by meaningful community participation and involvement in program development and implementation. Local leadership affects program success.
The success of large-scale projects appears to hinge on long-term high dosage penetration of strategies over time.
Market segmentation and tailoring of program components to population subgroups is essential. Needs and preferences vary according to location, ethnicity, income, age, sex, and health status.
To date, evidence does not support the effectiveness of mass media campaigns; they may be more most effective when linked to comprehensive initiatives rather than used as stand alone interventions.
Large numbers of adults prefer unstructured physical activity over classes, teams, or exercise facilities. Although walking is the most popular form of physical activity overall, not all community environments are conducive to safe, pleasant walking.
Land development and transportation patterns, proximity of facilities, perceived safety, neighborhood density, and cyclist and pedestrian-friendly amenities and other aspects of the physical environment affect physical activity levels. Intervention research in this area is in its infancy.
The ideal combination of strategies and time frame for community-based interventions is unknown. Differences in community contexts combined with multiple levels of influence on physical activity behavior (from intrapersonal attitudes to national policies) suggest a need for long-term comprehensive initiatives.
The conclusions of this summary are in accord with a recent report from the Task Force on Community Preventive Services, which recommended or strongly recommended 6 types of interventions. They recommended large-scale, comprehensive community-wide campaigns; individually tailored health behavior change programs; socially supportive interventions in community settings; and enhanced access to places for physical activity. They also recommended school-based physical education and point-of-decision prompts to encourage stair use. The Task Force concluded that insufficient evidence exists to recommend mass media campaigns, classroom-based health education (elementary and college aged), and social support in family settings. Transportation, infrastructure, and land use planning approaches were noted as “pending” results in the Task Force report (74).
Implications and recommendations
Programs from within each of the 4 frameworks described above have value in promoting physical activity for primary, secondary, and tertiary prevention of arthritis and related disability. Each framework presents strengths and weaknesses. The person-focused approach can be successful in the short term, particularly for persons motivated by pain and dysfunction to initiate behavior changes; however, making physical activity a permanent lifestyle feature has proven challenging. Exercise prescription and individually oriented strategies are appropriate for rehabilitation efforts, and community-based classes are appropriate for persons attracted to structured exercise settings; however, physical activity levels decline at the end of short-term programs, and large numbers of people with arthritis or at risk for developing arthritis do not seek out structured programs at all. In addition to establishing skills for long-term maintenance, the importance of creating a supportive community environment that provides safe, accessible, and pleasant options cannot be underestimated.
Based on the health promotion literature, socioecologic approaches to physical activity promotion are currently in favor, perhaps in part because they subsume other approaches, addressing influences on behavior at multiple levels. Such community-based efforts are massive and not likely to be undertaken with a specific focus on arthritis. Therefore, there is a need for creating a linkage between successful person-focused, community-based programs for persons with arthritis and other broader-based community interventions targeting environmental and policy issues that ultimately affect quality of life for everyone. The National Arthritis Action Plan (3) suggests involving “new and innovative partners” in disseminating key messages about arthritis. The reverse is also true: there is a need for persons with arthritis and organizations that serve persons with arthritis to partner with local, state, and national initiatives and coalitions that address chronic disease prevention and management and those that specifically promote physical activity. On the research front, the call for “new and innovative partners” suggests a multidisciplinary approach and the realization that the prevention, treatment, and management of arthritis have medical, social, and public health influences and outcomes. These special articles in Arthritis Care and Research are an example of this emerging perspective.