Physical Activity and Health self-monitoring group had significantly better adherence over 12 weeks than those in the self- monitoring plus attention or control groups; how- ever, adherence over the last 6 weeks of the study was significantly better in the self-monitoring plus atten- tion group. Actual differences were not large, amount- ing to 4 to 5 days of gym attendance over 3 weeks, compared with about 3 days among controls. In all three groups, adherence dropped off most sharply during the first 6 weeks of the study. Classes, health clubs, and fitness centers are resources to promote physical activity, and numer- ous studies have been undertaken to improve atten- dance (Table 6-2). However, many people prefer to exercise on their own. Several studies have used behavioral management techniques to encourage people to do so on their own (Table 6-2). In some studies, training in behavioral management tech- niques has occurred in a group setting before the participants began exercising on their own; in oth- ers, information has been provided by mail. Results have been equivocal. King, Haskell, and colleagues (1995) assigned 50- through 6%year-old partici- pants to one of three conditions: a vigorous, group- based program (three 60-minute sessions); a vigorous, home-based program (three 60-minute sessions); and a moderate, home-based program (five 30-minute sessions). At 1 year, a'dherence was significantly greater in both home-based programs than in the group-based program. At 2 years, how- ever, the vigorous, home-based program had higher adherence than the other two programs. Research- ers hypothesize that it was more difficult for the moderate group to schedule 5 days of weekly physi- cal activity than for the vigorous group to schedule 3 days. Another study encouraged self-monitoring and social support (walking with a partner) and also tested a schedule of calling participants to prompt them to walk. Frequent calls (once a week) resulted in three times the number of reported episodes of activity than resulted from calling every 3 weeks (Lombard, Lombard, Winett 1995). Cardi- nal and Sachs (1995) randomly assigned 133 women to receive one of the three packets of information promoting physical activity: self-instructional pack- ages that were based on stage of change and that provided tailored feedback; a packet containing a standard exercise prescription; and a packet pro- vidingminimal information about health status and exercise status. No significant differences were observed among the three groups at baseline, 1 month, or 7 months. The advent of interactive expert-system com- puter technologies has allowed for increased indi- vidualization of mailed feedback and other types of printed materials for health promotion (Skinner, Strecher, Hospers 1994). Whether these technolo- gies can be shown to be effective in promoting physical activity at low cost is yet to be determined. Insummary, behavioral management approaches have been employed with mixed results. Where an effect has been demonstrated, it has often been small. Evidence of the effectiveness of techniques like self- monitoring, frequent follow-up telephone calls, and incentives appear to be generally positive over the short run, but not over longer intervals. Evidence on the relative effectiveness of interventions on adher- ence to moderate or vigorous activity is limited and unclear. Because of the small number of studies, the variety of outcome measures employed, and the di- versity of settings examined, it is not clear under what circumstances behavioral management approaches work best. In a number of studies, methodological issues, such as high attrition rates, short follow-up, small sample sizes, lack of control or comparison groups, incomplete reporting of data, or lack of clarity about how theoretical constructs were operationalized. also make it difficult to determine the effectiveness of behavioral management approaches or to general- ize results to other settings or population groups. Stages of change theory suggests that people move back and forth across stages before they become able to sustain a behavior such as physical activity. The relatively short time frame of many studies and the use of outcome measures that are not sensitive to stages of change may have limited the ability to determine if and to what extent possessing behav- ioral management skills is useful in the maintenance of regular physical activity. Interventions in Health Care Settings Health care settings offer an opportunity to indi- vidually counsel adults and young people about physical activity as well as other healthful behaviors, such as dietary practices (U.S. Preventive Services Task Force 1996). Approximately 80 percent of the 226 U.S. population see a physician during a l-year period (National Center for Health Statistics 1991), but the extent to which physicians counsel their patients to be physically active is unclear. One survey of physicians found 92 percent reporting that they or someone in their practice counseled patients about exercise (Mullen and Tabak 1989), but in a more rcccnt study, only 49 percent of primary care physi- cians stated they believed that regular daily physical nctivity was very important for the average patient c\Vechsler et al. 1996). Counseling is likely to be brief, often less than 2 minutes (Wells et al. 1986), .tnd ineffective counseling approaches are often cmpfoyed (Orleans et al. 1985). Physicians may be \css likely to counsel patients about health habits if their own health habits are poor (Wells et al. 1984). Only three studies attempting to improve the physical activity counseling skills of primary care physicians have been reported in the literature; the rcsufts suggest small but generally positive effects on patients, with from 7 to 10 percent of sedentary l,crsons starting to be physically active (Table 6-2). ( )nc feasibility trial ofmultiple risk factorreduction- IIK Industrywide Network for Social, Urban, and Rural Efforts (INSURE) Project-indicates that con- ttnuing medical education seminars, combined with rclmbursement for prevention counseling and re- rntndcrs to providers, can increase the.percentage of I lrcsc physicians' patients who subsequently start exer- t 151ng (Logsdon, Lazaro, Meir 1989). The Physician- `j.t~d Assessment and Counseling for Exercise (PACE) l)rogram incorporated social cognitive theory and the trnnstheoretical model to individualize brief (2-5 Inlnutcs) counseling messages for patients. Com- p;u-cd with patients who did not receive the program c,ounseling, those who did had significantly greater Improvements at 4-6 weeks in their reported stage of i~h!sical activity readiness, their reported amount of : ,rlking for exercise, and their scores from an activity ,n0nitor (Calfas et al. in press). The Canadian Task Force on the Periodic Health Examination (1994) cited insufficient evidence as the reason for not making a recommendation regard- ing physical activity counseling. However, several other professional organizations have recently rec- ommended routine physical activity counseling. The `~mcrican Heart Association (Fletcher et al. 1992), `Jc American Academy of Pediatrics (1994), the .`mcrican Medical Association (1994)) the President's 227 Understanding and Promoting Physical Activity Council on Physical Fitness and Sports (1992), and the U.S. Preventive Services Task Force (1989,1996) all recommend including physical activity counsel- ing as part of routine clinical preventive services for both adults and young people. In summary, many providers do not believe that physical activity is an important topic to discuss with their patients, and many lack effective counseling skills. The studies that have attempted to increase provider counseling for physical activity demonstrate that providers can be effective in increasing physical activity among their patients. It is not known what alternative approaches to provider counseling can be used effectively in health care settings, although the work of Mayer and colleagues (1994) suggests that well-trained counselors conducting health education classes with patients may help older adults make changes in their stage of physical activity. Community Approaches Communitywide prevention programs have evolved from the concept that a population, rather than an individual, approach is required to achieve primary prevention of disease through risk factor reduction (Luepker et al. 1994). Behaviors and lifestyle choices that contribute to an individual's risk profile are influenced by personal, cultural, and environmental factors (Bandura 1977b). Much of the current knowl- edge regarding community-based prevention strate- gies has been gained over the past 20 years from three U.S. research field trials for community-based health promotion-including physical activity promotion- to reduce cardiovascular disease (Table 6-2). These three trials, which were funded by the National Heart, Lung, and Blood Institute during the 198Os, were the Minnesota Heart Health Program (MHHP) (Luepker et al. 1994), the Pawtucket Heart Health Program (PHHP) (Carleton et al. 1995), and the Stanford Five-City Project (SFCP) (Farquhar et al. 1990). The MHHP advocated regular physical activity as part of its broad effort to reduce risk for CHD in whole communities in the upper Midwest (Crow et al. 1986; Mittelmark et al. 1986). Three intervention communities received a 5- to 6-year program designed to reduce smoking, serum cholesterol, and blood pressure and to increase physical activity; three other communities served as comparison sites. Mass media were used to educate the public about the relationship Physical Activity and Health between regular physical activity and reduced risk for CHD and to increase opportunities for physical activ- ity. Health professionals promoted physical activity through their local organizations, through their advi- sot-y committees on preventive practice. and through serving as role models and opinion leaders. Systematic risk factor screening and education provided on-site measurement, education, and counseling aimed in part at increasing to 60 percent the prevalence of physical activity among the residents in the three intervention communities. The adult education com- ponent made available personal, intensive, and mul- tiple-contact programs to increase physical activity; this strategy focused on self-management and in- eluded changes in existing behaviors, in the meaning of those behaviors, andin the environmental cues that supported them. Direct education programs forschool- aged children promoted physical activity in young people and their parents. The MHHP investigators reported small but significant effects for physical activity in the first 3 years among people in the cross- sectional study group; that effect disappeared with an increasing secular trend in physical activity in the comparison groups. The cohort group (followed over time) showed no intervention effect until the last follow-up survey (Figure 6-l). . Figure 6-l. Results of the Minnesota Heart Health Program on physical activity. Graph compares the percentage of respondents reporting regular physical activity in intervention cities and the secular trend estimated from control cities I I I I I -4 -2 0 2 4 6 8 Education year 75 I I 1 Panel 1: Cross-sectional Fitted secular trend and MHHP education program effect estimates m-v- Education effect I 95% Confidence bounds - Secular trend I 45 - I I Baseline I I Follow-up I 35 I I I I I -4 -2 0 2 4 6 8 Education year Panel 2: Cohort Fitted cohort levels and MHHP education program effect estimates ---- Education effect I 95% Confidence bounds - Cohort levels Sours e: Luepker RV et Ji. American ]ourna/ oi PuOlic Health 1994 (reprinted with permission). Note: Adjusted for age, sex. and education. 228 The PHHP fostered communiry involvement in heart healthy behavior changes in Pawtucket, Rhode \s\and(Carletonetal. 1995).Thefocuswasongrassroots organizing, volunteer delivery, and partnerships with existing organizations rather than on using electronic media (Lasater et al. 1986). In the area of physical .Ictlvity promotion, the emphasis was on environmen- t,ll and policy change through partnerships with city oo\,ernment and others. Working with the Department I- of Parks and Recreation, the PHHP was instrumental in establishing cardiovascular fitness trails in both of the c~t).`sparks. Earlyin itsexistence, the PHHPalso helped that department place on the ballot and pass a large bond issue in return for renovations (e.g., lights, fenc- ing to keep stray dogs out, resurfacing) to an existing c,ciarter-mile track for walking. The Pawtucket 6- Lvcck Imagine Action Program, designed around the stages of change model, enrolled more than 600 participants, who subsequently reported being more .IC`~IVC as a result of the program (Marcus, Banspach, et ,11. 1992). Results of this uncontrolled study suggest I 11`11 ;I stage-based approach may be effective in moving people toward regular physical activity. The SFCP included two intervention and two comparison communities in northern California (only morbidity and mortality data were monitored in the hfth city, and those results were not reported in this ytucly). This project was designed to increase physical ~cr~viry and weight control and to reduce plasma cho- Icstcrol levels, cigarette use, and blood pressure l'nrquhar et al. 1990). Greater emphasis was placed on rlutrition, weight control, and blood pressure than on physical activity. The program used concepts from \ocial learning theory, diffusion theory, community organization, and social marketing in combination \vlth a communication and behavior change model (Flora, Maccoby, Farquhar 1989). The program relied heavily on the use of electronic and print media for the Jchvery of health education information. General edu- .ition was supplemented by four to five annual educa- IIon campaigns targeting specific risk factors. Direct face-to-face activities included classes, contests, and school-based programs (Farquhar et al. 1990). Overall, Ihe educational intervention had no significant im- pact on physical activity levels, knowledge, self- efficacy, or attitudes toward physical activity (Young ct al. T in press). In the cross-sectional sample, men in `he experimental communities were significantly more Understanding and Promoting Physical Activity likely than those in the control communities to engage in at least one vigorous activity. For women in both the cross-sectional and cohort studies, a small but signifi- cant increase was observed in the number of moderate activities engaged in (Young et al., in press). Among smaller-scale community studies, the re- sults of efforts to promote physical activity have been mixed (Table 6-2). One exception was the community- based cardiovascular disease prevention program aimed at black residents in rural communities in the Missouri "Bootheel" (Brownson et al. 1996). In this 5- year, low-cost intervention project, educational ef- forts were combined w,ith environmental changes. Local coalitions formed walking clubs, built walking trails, started exercise classes in churches, and orga- nized special events to promote both physical activity and good nutrition. Although no difference in levels of physical inactivity was observed between the Bootheel and the rest of the state at follow-up, physical inactiv- ity declined an average of 3 percent in Bootheel communities that had coalitions and increased an average of 3.8 percent in those without, for a net improvement of 6.8 percent. In summary, results of community-based inter- ventions to increase physical activity have been generally disappointing. Measurement of physical activity has varied across studies, making compari- sons difficult. The presence of active communiry coalitions, widespread community involvement, and well-organized community efforts appear to be im- portant, however, in increasing physical activity levels. Worksite Programs Physical activity programs conducted on the worksite have the potential to reach a large percentage of the U.S. population (Bezold, Carlson, Peck 1986; Na- tional Center for Health Statistics 1987). As settings for physical activity promotion, manyworksites have easy access to employees and supportive social net- works and can make changes in the environment to help convey physical activity as an organizational norm (Shephard, in press). The proportion of worksites offering physical activity and fitness programs has grown in recent years, from 22 percent in 1985 to 42 percent in 1992 (Table 6-3). For two groups of employers, those with 50-99 employees and those with loo-249 229 Physical Activity and Health Table 6-3. Summary of progress toward Healthy People 2000 objective 1 .lO "Increase the proportion of worksites offering employer-sponsored physical activity and fitness programs as follows:" Year 2000 objective 1985 1992 Year 2000 target Physical activity and fitness worksites with: SO-99 employees 14% 33% 20% loo-249 employees 23% 47% 35% 250-749 employees 32% 66% 50% 750+ employees 54% 83% 80% Source: U.S. Department of Health and Human Services, 1992 National Survey of Worksite Health Promotion Activities, 1993. r employees, the percentage with exercise programs more than doubled over that time period. In each worksitesize category, the percentage with exercise programs had already (i.e., in 1992) exceeded the year 2000 national objective for worksite health promotion listed in Healtlly People 2000 (USDHHS 1993). Generally, the extent of participation, effec- tiveness, and quality of those programs is unknown, for only a few worksite physical activity programs have been evaluated (Table 6-2). In the Johnson 6;r Johnson Live for Life program (Wilbur 1983), employees at four experimental sites participated in lifestyle seminars, contests, and exer- cise programs and received newsletters on health issues and other health communications. Experi- mental and control sites both received an annual health assessment. Overall, at the end of 2 years, 20 percent of women and 30 percent of men in the experimental sites reported beginning a vigorous exercise program; the prevalence at three compari- son sites was 7 percent for women and 19 percent for men (Blair et al. 1986). Fries and associates (1993) evaluated the effec- tiveness of a health promotion program that included physical activity for Bank of America retirees. In one intervention group, each participant paid $30 for a personalized. mail-delivered program that included a health risk appraisal and behavioral management books and other materials. A second group received a risk appraisal and nothing else for the first 12 months, after which it received the full intervention. A control group was monitored for claims data only. The first intervention group did not differ from the second in self-reported physical activity at the end of year 1 but was significantly different in year 2. Worksite programs less often attract sedentary, blue-collar, or less-educated employees, but interven- tions that are tailored to these persons' needs and interests (King, Carl, etal. 1988) and provide counsel- ing and peer support (Heirich et al. 1993) show promise. In a controlled study, Heirich and colleagues (1993) compared different programs at four automo- tive manufacturing plants of like size and employee populations. The three approaches tested were 1) a staffed physical fitness facility, 2) one-to-one counsel- ingandoutreachwith high-riskemployees (i.e., those who had hypertension, were overweight, or smoked cigarettes), and 3) one-to-one counseling and out- reach to all employees, peer support, and organiza- tional change (e.g., the institution of nonsmoking areas). The fourth site, which served as a control, offered health education classes and special events. After 3 years, exercise prevalence at the four sites was lowest at the plant with the exercise facility _ In the two counseling and outreach sites, nearly half of the employees reported exercising 3 times a week. In summary, considerable progress has been made in meeting the Healthy People 2000 goals for worksite physical activity programs. Too few studies exist to clearly determine what elements are required for physical activity programs at work to be effective in increasing physical activity levels among all em- ployees, attracting diverse employee groups (such as blue-collar workers), or maintaining exercise levels over time. However, the limited research available suggests that widespread employee involvement and support coupled with organizational commitment evidenced by the presence of policies and programs may be important factors in increasing levels of physical activity. Existing controlled studies have been done in larger worksites; studies have not yet shown what might work in smaller worksites and in diverse worksites (e.g., where many employees travel or facilities may not exist). Communications Strategies Communications strategies, both electronic and print, \lave the potential for reaching individuals and com- munities with a rapidity unmatched by other inter- ifcntion strategies. For the general population, media can play several roles: to increase the perceived importance of physical activity as a health issue, to communicate the health and other benefits of physi- cal activity, to generate interest in physical activity ;md awareness about available programs, to provide role models for physically active lifestyles, and to provide cues to action, such as getting people to request further information on physical activity, \,isit an exercise site, or begin exercising (Donovan and Owen 1994). The effectiveness of different forms of media alone, including broadcast and print media, for promoting either initial adoption or subsequent maintenance of physical activity remains unclear hccause the few systematically evaluated interven- I [(Ins employing communications strategies have \hown mixed results (Osler and Jespersen 1993; liooth et al. 1992; Owen et al. 1995; Luepker et al. 1994; Farquhar et al. 1990). The SFCP, discussed cnrlier, resulted in small increases in the number of moderate activities engaged in by women and vig- orous activity engaged in by men. Two national mass media campaigns to increase physical activity, i~nrticularly walking, to prevent cardiovascular dis- I"KC were conducted in Australia in 1990 and 1991 (Booth et al. 1992). Drawing on social marketing `lnd social learning theories, both campaigns in- cluded paid advertisements on national television, Pllblic service announcements on radio, scripted episodes on two nationally broadcast television clrarnas, posters and leaflets, stickers, T-shirts and -\\,catshirts, magazine articles, distribution of a Understanding and Promoting Physical Activity professional article, soap operas specially scripted to feature physical activity, and publicity tours by two experts in heart health. The budgets and paid television coverage for the 1990 and 1991 cam- paigns were similar. Both campaigns were evalu- ated by one-on-one, home-based interviews with structured cross-sectional random samples of ap- proximately 2,500 people 2 weeks before and 3 to 4 weeks after each campaign. Both campaigns re- sulted in significant differences in message aware- ness (46 percent vs. 71 percent in 1990; 63 percent vs. 74 percent in 1991). The 1990 postcampaign survey revealed significant increases in walking for exercise (p < 0.01) compared with the precampaign period, although the actual percentage increase was small (73.9 percent vs. 70.1 percent). In particular, adults over 50 years of age were nearly two times more likely to report walking at follow-up than before the campaign. The 1991 campaign produced different results. Evaluation showed that the per- centage of persons reporting walking in the previ- ous 2 weeks declined from precampaign levels among all adult age groups except people over 60 years of age. Intention to become more active also declined overall, from 26.3 percent to 24.8 percent (Owen et al. 1995). Communications intended to serve as cues to action have been tested at places where people can choose whether to walk or ride. This approach in- volves placing signs to use the stairs near escalators in public places like train and bus stations or shop- ping malls (Brownell, Stunkard, Albaum 1980; Blarney, Mutrie, Aicchison 1995). For example, signs that said "Stay Healthy, Save Time, Use the Stairs" increased the percentage of people using stairs in- stead of an adjacent escalator from 8 percent to 1% 17 percent (Blarney, Mutrie, Aitchison 1995). Twelve weeks after the sign was removed, the increase in stair use remained significant but showed a trend toward baseline. In summary, communications strategies have had limited impact. It is not clear if communications approaches would be more effective in getting people to be regularly active if they were linked with oppor- tunities to act on messages or if messages were tailored to stages of change or to the needs of sub- groups in the population (Carleton et al. 1995; Donovan and Owen 1994; Young et al. in press). 231 physical Activity and Health Appropriately placed communications that serve as cues to action appear to increase the decision to use the stairs instead of ride the escalator. Special Population Programs Racial and Ethnic Minorities The few interventions studies that have been con- ducted with racial and ethnic minorities have pro- duced mixed results. The Bootheel Project referred to earlier in this chapter found increased levels of physical activity in black communities with coali- tions. The Physical Activity for Risk Reduction project (Lewis et al. 1993) was undertaken in black commu- nities in Birmingham, Alabama, using a combination of behavioral management and community organi- zation approaches. In the intervention groups, com- munity members played roles in defining needs, identifying strategies, and conducting interventions. In those communities where strong organization, leadership, and commitment to the project were observed, statistically significant increases in physi- cal activity were also noted. Results of two family-based health promotion programs that used behavioral management ap- proaches to promote physical activity showed no greater increase in physical activity among those participating in the programs than among those in a control group. Nader and colleagues (1989) con- ducted a nutrition and physical activity program for Anglo-American and Mexican American families with children in fifth and sixth grades; the program im- proved dietary habits but did not succeed in increas- ing physical activity levels, although participation in the program was high. Another family-based pro- gram, a 14week intervention for African American families that included educational sessions and twice- weekly fitness center activities, had low attendance and did not increase physical activity (Baranowski et al. 1990). The Indian Health Service undertook the community-based Zuni Diabetes Project to increase physical activity and decrease body weight among Zuni Indians in New Mexico who had non-insulin- dependent diabetes mellitus (NIDDM) (Leonard, Leonard, Wilson 1986). The exercise program con- sisted of several l-hour aerobic sessions offered during the week. Zuni Indians who were trained in exercise and group leadership methods helped coordinate the program and build community ownership. After par- ticipating in aerobic sessions through the program, 43 percent of the participants began and maintained an at-home exercise program, whereas only 18 percent of a comparison group of previously sedentary non- participants with NIDDM did so (Heath et al. 1987). People Who Are Overweight Being overweight increases the risk of developing chronic diseases (see Chapter 4). Results of inter- ventions to promote physical activity for weight loss have been mixed (Perri et al. 1988; Jeffery 1995; King et al. 1989). The MHHP , one of the large community inter- vention trials discussed earlier in this chapter (Luepker et al. 1994), developed a series of compo- nent programs containing strategies to increase physi- cal activity for losing weight or preventing weight gain Ueffery 1995). The Building Your Fitness Fu- tures program was a 4-week adult education class that focused on how to develop a regular exercise program. The Wise Weighs programs was an g-week adult education class that emphasized weight man- agement strategies related to diet and exercise. The third MHHP intervention, a correspondence course, addressed diet and exercise through monthly news- letters and tested two levels of financial contract incentives ($5 and $60 dollars). Each of these pro- grams was evaluated in the MHHP randomized trial. The Building Your Fitness Futures and the Wise Weighs programs resulted in only small weight loss that was not significant after 1 year. The correspon- dence course resulted in significantly greater weight loss among participants with $60 incentives than among those with $5 incentives. Preventing weight gain may be easier than promot- ing weight loss. Wing (1995) suggests that there are three time periods during which interventions to pre- vent weight gain might be most effective: in the years between ages 25 and 35 years, in the peri-menopausal period for women, and in the year following successful weight loss. A fourth MHHP program that addressed physical activity, the Weight Gain Prevention Pro- gram, was a randomized trial of 211 communityvolun- teers. The participants (approximately two-thirds women) were randomly assigned to either the inter- vention group (n = 103) or the no-contact control group (n = 108). This program was for normal-weight adults and included monthly newsletters and four classes emphasizing diet and regular exercise as well as a financial incentive component linked to weight maintenance. The intervention group lost 2 pounds on average over the course of the year and were significantly less likely to gain weight than the control group (82 percent vs. 56 percent) (Jeffery 1995). Older Adults \lany of the diseases and disabling conditions asso- elated with aging can be prevented, postponed, or ameliorated with regular physicalactivity (see Chap- tcr 4). The few interventions that have been tested to Increase physical activity levels among older adults ,how generally positive results. The 1990 Australian ticart Week campaign reviewed earlier resulted in a t\vofold increase in walking among adults over 50 \`ears of age (Owen et al. 1995). Retirees in the study by Fries (1993), also discussed earlier, showed sig- nlficantly greater improvements in physical activity In year 2 than did persons in the control group. Pnrricipants in a longitudinal study of Medicare rcclpients (n = 1,800) who belonged to a health maintenance organization were randomly assigned 10 a preventive care or a control group (Mayer et al. 1094). The intervention employed information and hchavior modification approaches. Participants re- ccived recommended immunizations, completed a health risk appraisal, received face-tq-face counsel- ~ng that included goal setting, received follow-up rclcphone counseling, and participated in educa- ~lonal sessions on health promotion topics. A focus ~III physical activity was a priority in goal-setting dIscussions; 42 percent of participants selected in- creasing physical activity as their goal. Members of both groups were largely white, well educated, and ~~enerally had above-average incomes. The preva- r cnce of physical activity was high in both groups at llaseline; approximately 60 percent reported get- llng regular exercise. At 1 year, the intervention croup showed a significant 7 percent increase in yell-reported physical activity. Much of the published research on physical activ- I$ describes researcher-initiated interventions. How- ever, individuals and small groups of people often Imtiate physical activity on their own, independent of nny formal program. A qualitative research study by Duncan, Travis, and McAuley (1995) used observa- `ions and in-depth interviews to examine motivation Understanding and Promoting Physical Activity for initiating and maintaining mall walking by older persons in rural West Virginia. Most participants in this study reported becoming physically active at the urging of their physicians; several others were moti- vated by personal interest in health maintenance, and some were encouraged by family members. Mall walk- ers maintained a regular routine, showing up at the same time each day, walking in pairs or small groups, and then adjourning to a mall eatery for coffee or breakfast. Interviews revealed that participants per- ceived mall walking as meaningful "work" to be doing during retirement. A need for socializing with others, a sense of belonging to a community of mall walkers, and the safe environment of the mall were other factors contributing to adherence. Study researchers recommended that community-based physical activ- ity programs try to replicate various aspects of work, such as keeping attendance records and providing occasional recognition or acknowledgment of a job well done (such as pins, certificates, or celebrations). People with Disabilities People with disabilities have similar health promo- tion and disease prevention needs as persons with- out disabilities. Interventions to promote physical activity for risk reduction among persons with mo- bility, visual, hearing, mental, or emotional impair- ments are largely absent from the literature. Physical activity interventions for managing chronic condi- tions, on the other hand, have led to enhanced cardiorespiratory fitness and improved skeletal muscle function in persons with multiple sclerosis (Ponichtera-Mulcare 1993), increased walking ca- pacity and reduction in pain for patients with low back pain (Frost et al. 1995), and improvements in endurance among patients with chronic obstructive pulmonary disease (Atkins and Robert 1984). In summary, interventions that have been suc- cessful in increasing physical activity among minori- ties have employed community organization strategies, such as coalition building and community engagement at all levels. Family-oriented interven- tions in community centers that have employed behavioral management approaches have not re- sulted in increases in physical activity. Physical activity interventions incorporating incentives show promise for promoting weight loss or preventing weight gain. Although there are a limited number of 233 Physical Activity and Health studies, positive effects have been shown for inter- ventions and communications strategies promoting physical activity in older adult populations, at least among older white adults with moderate incomes and education levels. What is not well known is what interventions may be effective with racial or ethnic minority older adults who may face barriers such as language, transportation, income, education, or dis- ability. It is not clear what interventions might be effective to promote physical activity. other than for disease management, among people with disabili- ties, or what strategies might assist with the manage- ment of pain, periods of illness, environmental barriers, or other circumstances to improve adher- ence with physical activity recommendations. Summary The review of adult intervention research literature provides limited evidence that interventions to pro- mote physical activity can be effective in a variety of settings using a variety of strategies. Controlled interventions that have been effective at the work- place, in health care settings, and in communities have resulted in increased physical activity, although effects have tended to be small, in the range of 5-10 percent, and short-lived. Multiple interventions con- ducted over time may need to be employed to sustain physical activity behavior. Most experimental and quasi-experimental intervention research has been theory-based, much if not most relying largely on behavioral management strategies, often in combi- nation with other approaches, such as communica- tions and social support. Mixed results have made it impossible to determine what theory or theories alone or in combination have most relevance to physical activity. Research strategies that appear promising include the tailoring of interventions to people's needs, experiences, and stages of change; the timing of intervention strategies to reinforce new behaviors and prevent relapse (such as through frequent follow-up telephone calls); peer involve- ment and support; and an engaged community at all levels. It is not known if interventions could be strengthened by combining them with policy ap- proaches (Luepker 1994; Winkleby 1994). Intervention studies with adults were often con- ducted over a brief period of time, had little or no follow-up, and focused on the endpoint of specified vigorous physical activity rather than on moderate- 234 intensity physical activity or total amount of activity. Studies used different endpoints, such as class par- ticipation versus specified changes in behavior, mak- ing them difficult to compare. Because physical activity interventions were often only one compo- nent of an intervention to reduce multiple risk fac- tors, they may not have been robust enough to result in much or any increase in physical activity. Few if any studies compared their results to a standard of effectiveness, such as recommended frequency or duration of moderate or vigorous physical activity, or clearly stated the extent of stage-based change. Behavioral Research on Physical Activity among Children and Adolescents Behavioral research in this area includes studies on the factors influencing physical activity among young people as well as studies examining the effec- tiveness of interventions to increase this behavior. This research, however, is more limited than the determinants and interventions literature for adults. Factors Influencing Physical Activity among Children and Adolescents The emphasis in this section is on factors that influ- ence unstructured physical activity during free time among youths rather than on supervised physical activity, such as physical education classes. Studies of organized youth sports have also been excluded. Only studies with some measure of physical activity as the outcome, however, are included in this re- view. For example, studies that investigated atti- tudes toward physical activity and did not relate those to a measure of physical activity were ex- cluded. As was the case in the adult section, this section focuses on studies that address modifiable determinants of physical activity, such as self- efficacy, rather than on studies that examine factors that cannot be altered to influence participation in physical activity, such as age, sex, and racejethnicity. Modifiable Determinants The modifiable determinants of youth physical ac- tivity include personal, interpersonal, and environ- mental factors (Table 6-l). Self-efficacy, a construct from social cognitive theory, has been positively associated with physical activity among older chil- dren and adolescents (Reynolds et al. 1990; Trost et al. 1996; Zakarian et al. 1994). Similarly, perceptions of physical or sports competence (Biddle and Armstrong 1992;BiddleandGoudas 1996; Dempsey, Kimiecik, Horn 1993; Ferguson et al. 1989; Tappe, Duda, Menges-Ehrnwald 1990) also have been posi- tively associated with physical activity among older children and adolescents. Expectations about the outcomes of physical activity are associated with physical activity among preadolescents and adolescents. Perceived benefits have been positively associated (Ferguson et al. 1989; Tappe, Duda, Menges-Ehrnwald 1990; Zakarian et al. 1994), whereas perceived barriers have been negatively associated (Stucky-Ropp and DiLorenzo 1993; Tappe, Duda, Menges-Ehrnwald 1990; Zakarian et al. 1994). Intention to be active, a construct from the theory of reasoned action and the theory of planned behavior, has been consistently und positively related to physical activity among older children and adolescents (Biddle and Goudas 1996; Ferguson et al. 1989; Godin and Shephard 1986; Reynolds et al. 1990) . Enjoyment, the major reason young people en- gage in physical activity (Borra et al. 1995), has been positively associated with physical activity among both children and adolescents (Stucky-Ropp and DiLorenzo 1993; Tinsley et al. 1995). Favorable ,Ittitudes toward physical education also have been llositively related to adolescent participation in physical activity (Ferguson et al. 1989; Zakarian et 111. 1994). Social influences-such as physically active role models and support for physical activity-are im- portant determinants of physical activity among \`onng people (Tinsley et al. 1995). Parental activity i Moore et al. 1991; Poest et al. 1989; Sallis, Patterson, `ilcKenzie et al. 1988) is positively related to physical .\ctivity among preschoolers. Studies reveal no rela- [ionship between parental physical activity and physi- cal activity among elementary school children (McMurray et al. 1993; Sallis, Alcaraz, et al. 1992), and either no relationship (Biddle and Goudas 1996; Garcia et al. 1995; Stucky-Ropp and DiLorenzo 1993; jallis, Patterson, Buono, et al. 1988) or positive :efationships (Anderssen and Wold 1992; Butcher Understanding and Promoting Physical Activity 1985; Gottlieb and Chen 1985; Stucky-Ropp and DiLorenzo 1993; Sallis, Patterson, Buono, et al. 1988) to the physical activity of middle school stu- dents (grades 5-8). Parental physical activity is posi- tively related to physical activity among older adolescents (Reynolds et al. 1990; Zakarian et al. 1994). The physical activity of friends (Anderssen and Wold 1992; Stucky-Ropp and DiLorenzo 1993; Zakarian et al. 1994) and siblings (Perusse et al. 1989; Salhs, Patterson, Buono, et al. 1988) also is positively associated with physical activity among older children and adolescents. Parental encouragement is positively related to physical activity among*preschoolers (McKenzie, Sallis, et al. 1991; Klesges et al. 1984, 1986; Sallis et al. 1993), and parental or adult support for physical activity is positively associated with physical activity among adolescents (Anderssen and Wold 1992; Biddle and Goudas 1996; Butcher 1985; Zakarian et al. 1994). Friends' support for physical activity (Anderssen and Wold 1992; Zakarian et al. 1994) also is positively related to physical activity among adolescents. Direct help from parents, such as organizing exercise activities (Anderssen and Wold 1992) or providing transportation (Sallis, Alacraz, et al. 1992>, is positively related to physical activity among older children and younger adolescents. Access to play spaces and facilities (Garcia et al. 1995; Sallis et al. 1993; Zakarian et al. 1994) is positively related to physical activity among youths of all ages. The avail- ability of equipment has been positively related to physical activity among preadolescent and adoles- cent girls (Butcher 1985; Stucky-Ropp and DiLorenzo 1993). Further, two studies of young children have demonstrated that time spent outdoors is a positive correlate of physical activity level (Klesges et al. 1990; Sallis et al. 1993). Determinants for Population Subgroups Among the limited number of subgroup-specific determinants studies, sex-specific differences are investigated most frequently. In two studies of ado- lescents (Kelder et al. 1995; Tappe, Duda, Menges- Ehrnwald 199O), competition motivated boys more than girls, and weight management motivated girls more than boys. Additionally, boys have higher levels of self-efficacy than girls (Trost et al. 1996) 235 physical Activity and Health and higher levels of perceived competence (Tappe, Duda, Menges-Ehrnwald 1990) for physical activity. Summary Few studies of the factors that influence physical activity among children and adolescents have applied the theories and models of behavioral and social science. The research reviewed in this section, how- ever, has revealed that many of the factors that influ- ence physical activity among adults are also determinants of physical activity among children and adolescents. Older children's and adolescents' inten- tions to engage in physical activity, as well as their perceptions of their ability to engage in such activity (i.e., self-efficacy and perceived competence), are positively related to their participation in physical activity. Social influences, such as parental and peer engagement in, and support for, physical activity, also are positively related to physical activity among young people. Further, exercise enjoyment and positive atti- tudes toward physical education have been positively associated with physical activity among older chil- dren and adolescents. Research is limited, however, on patterns of determinants for population subgroups, such as girls, ethnic minorities, and children with disabilitiesorchronic healthconditions (e.g.,asthma). Interventions to Promote Physical Activity among Children and Adolescents The most extensive and promising research on in- terventions for promoting physical activity among young people has been conducted with students in schools, primarily at the elementary school level. Although many school-based studies have focused on short-term results, a few studies have also exam- ined long-term behavioral outcomes. There is lim- ited evidence concerning the effectiveness of school-community programs, interventions in health care settings, family programs, and programs for special populations. In this section, the emphasis is on interventions designed to promote both unstruc- tured physical activity during free time and super- vised physical activity, such as physical education classes. Interventions designed to increase participa- tion in, or adherence to, organized youth sports have been excluded from this review. The review places special emphasis on experimental studies, which feature random assignment of individuals or groups to intervention (experimental) or comparison (con- trol) conditions, or quasi-experimental studies, which feature intervention and comparison groups. School Programs Because most young people between the ages of 6 and 16 years attend school, schools offer an almost populationwide setting for promoting physical activ- ity to young people, primarily through classroom curricula for physical education and health educa- tion. The CDC (in press) recommends that compre- hensive school and community health programs promoting physical activity among children and ado- lescents be developed to increase knowledge about physical activity and exercise, develop behavioral and motor skills promoting lifelong physical activity, fos- ter positive attitudes toward physical activity, and encourage physical activity outside ofphysical educa- tion classes. CDC's 1994 School Health Policies and Programs Study (Kann et al. 1995) examined the current nationwide status of policies and programs for multiple components of a school health program. The study examined kindergarten through 12th-grade health education and physical education at state, district, school, and classroom levels (Errecart et al. 1995). Results from the health education component of this study revealed that physical activity and fitness instruction were required in 65 percent of states and 82 percent of districts and were included in a required health education course in 78 percent of schools. Only 41 percent of health education teachers pro- vided more than one class period of instruction on these physical activity topics during the school year (Collins et al. 1995). Results from the physical education component of the School Health Policies and Program Study revealed that physical education instruction is re- quired by most states (94 percent) and school dis- tricts (95 percent) (Pate, Small, et al. 1995). These policies, however, do not require students to take physicaleducationeveryyear. Forinstance,although most middle and junior high schools (92 percent) and most senior high schools (93 percent) require at least one physical education course, only half of these middle and junior high schools and only 26 percent of these senior high schools require the equivalent of at least 3 years of physical education. Additionally, only 26 percent of all states require 236 Understanding and Promoting Physical Activity schools to offer a course at rhe senior high school level in lifetime physical activity (i.e., physical activ- it!, that can be practiced throughout one's lifetime) (Pate. Small, et al. 1995). The School Health Policies .Ind Programs Study also revealed that instructional practices in physical education often do not reflect the emphasis on lifetime physical activity that is recommended in the national objectives in Healrl~y Pcoplc 2000 (USDHHS 1990), in the National Physi- cal Education Standards (National Association for Sport and Physical Education 1995), and in the CDC's Guiddincs for Schwa and Community Health Proyrams to Promote Physical Activity Among Youth (In press). More than half of physical education teachers devoted multiple class periods to traditional bports activities, such as basketball (87 percent), volleyball (82 percent), and baseball/softball (82 percent), whereas much smaller proportionsof teach- ers devoted multiple class periods to lifetime physi- cal activities, such as jogging (47 percent), aerobic dance (30 percent), and swimming (14 percent) (Pate, Small, et al. 1995) (Table 6-4). Additionally, only 15 percent of all physical education teachers required students to develop individualized fitness programs (Pate, Smail, et al. 1995). Despite current guidelines' emphasis on lifetime physical activity, during the 2 years preceding the study only 22 percent of physical education teachers received in- service training on developing individualized fitness programs, and only 13 percent received training on increasing students' physical activity outside of physical education class (Pate, Small, et al. 1995). Detailed findings from the School Health Poli- cies and Programs Study are important because school-based physical education may be the most widely available resource for promoting physical activity among young people in the United States. For physical education to meet public health goals, it should provide all students with recommended amounts of weekly physical activity (USDHHS 1990). Table 6-4. Percentage of all physical education courses in which more than one class period was devoted to each activity, by activity, School Health Policies and Programs Study, 1994 Activity Percentage of all courses - Uasketball 86.8 Volleyball 82.3 lL~seball/soitball ai .5 Flag/touch football 68.5 Soccer 65.2 logging 46.5* Weight lifting or training 37.3* Tennis 30.3* Aerobic dance 29.6* LYalking quickly 14.7* Swimming 13.6* Handball 13.2* Racquetball 4.9+ HikinGackpacking 3.0' Bicycling 1.3* hxe: Adapted from Pate RP et al. School physical education. /ournal oikhool Healfh 1 Y95 (reprinted with permission). `LIfetIme physical activities 237 Physical Activity and Health Ironically, observations of physical education classes indicate that insufficient class time is spent actually engaging in physical activity (McKenzie et al. 1995; McKenzie et al., in press; Simons-Morton et al. 1991, 1993, 1994). The School Health Policies and Programs Study provided a national overview of the status of school health programs (Kann et al. 1995). Intervention research has been reported from several studies (Table 6-5). Most of the early research in schools focused on knowledge-based health education class- room lessons; these studies generally reported posi- tive changes in knowledge and attitudes but not in behaviors. Summarized in review articles (Sallis, Simons-Morton, et al. 1992; Simons-Morton, Parcel, O'Hara et al. 1988), these studies suffered from methodological problems, such as small samples and measurement limitations. Contemporary programs emphasize the importance of multicomponent inter- ventions that address both the individual and the environmental level to support engagement in physi- cal activity among youths (Kelder, Perry, Klepp 1993; Luepker et al. 1996; McKenzie et al., in press; Perry et al. 1990, 1992; Simons-Morton, Parcel, O'Hara 1988; Stone et al. 1995). The Know Your Body (KYB) program (Williams, Carter, Eng 1980) has been the focus of three school- based cardiovascular risk reduction studies (Bush, Zuckerman, Taggart, et al. 1989; Bush, Zuckerman, Theiss, etal. 1989; Resnicowetal. 1992; Walter 1989). This program includes health screening, behavior- oriented health education curricula, and special interventions for students with one or more cardio- vascular disease risk factors (e.g., hypercholester- olemia, hypertension, obesity, lack of exercise, cigarette smoking) (Williams, Carter, Eng 1980). Although this program was designed to improve students' knowledge, attitudes, and behaviors re- lated to physical activity, nutrition, and cigarette use, the measurement and reporting of physical activity behavior has been inconsistent among the three studies. In the first study, the measure for self- reported physical activity was found to be unreliable, and the results related to this measure were not reported (Walter 1989). In the second KYB study, students' physical activity behavior was not assessed (Resnicow et al. 1992). The third study was a 5-year, randomized cardiovascular risk reduction trial among 1,234 African American students in grades four through six from nine schools stratified for socio- economic status (Bush, Zuckerman, Taggart, et al. 1989; Bush, Zuckerman, Theiss, et al. 1989). This project included the KYB health education curricu- lum, health screening, parent education, and KYB advisory boards for parents, community members, students, and physicians. After 4 years, students from both the intervention and control schools had significant increases in health knowledge at posttest, and intervention students had significantly better gains in health knowledge (Bush, Zuckerman, Theiss, et al. 1989). Physical activity, however, decreased significantly amongstudents f&m both the interven- tion and control schools, and there was no difference in physical activity between the intervention and control schools. The Stanford Adolescent Heart Health Program (Killen et al. 1988) was a classroom-based random- ized cardiovascular disease risk reduction trial for 1,447 tenth graders from four matched high schools within two school districts. One school within each district was designated at random to receive a 20- week risk reduction intervention, and the other school served as the control. The classroom-based intervention focused on three cardiovascular risk factors, including physical activity. At the 2-month follow-up, students from the intervention schools had significantly higher gains in knowledge about physical activity than did students in the control schools. Among students not regularly exercising at baseline, those in the intervention schools had sig- nificantly greater increases in physical activity than did those in control schools. Additionally, students who received the intervention had significantly lower resting heart rates and subscapular and triceps skinfold measures. The long-term effectiveness of this program was not reported. An Australian study (Dwyer et al. 1979, 1983) was one of the first randomized trials that investi- gated the effects of daily physical activity on the health of elementary school students. The study included 513 fifth-grade students from seven Adelaide metropolitan schools. Three classes from each school participated in the study and were ran- domly assigned to one of three conditions: fitness, skills, or control. Students in the control condition received the usual three 30-minute physical educa- tion classes per week. The students in both interven- tion conditions received 75 minutes of daily physical 238 Understanding and Promoting Physical Activity education: one condition emphasized fitness activi- ties featuring high levels of physical activity, and the other emphasized skill development activities with- out special emphasis on the intensity or duration of physical activity. With the class as the unit of analy- sis, the fitness condition led to significantly greater Increases in endurance fitness and decreases in skinfold measurements. Although this study did not evaluate the impact of increased physical education on students' engagement in physical activity outside of class, it showed that academic test scores did not differ between the intervention and control groups, despite the additional 275 minutes of class time the intervention groups spent on physical education rather than on traditional academic subjects. Go For Health (GFH) was a 3-year school health project designed to promote healthful diet and exer- cise behaviors among elementary school students (Parcel et al. 1987; Simons-Morton, Parcel, O'Hara 1988; Simons-Morton et al. 1991). This project in- volved four elementary schools (kindergarten through fourth grade) from the Texas City Indepen- clcnt School District. Two schools were assigned to serve as controls, and the other two were designated JS GFH intervention schools. The intervention was hased on social cognitive theory and included a GFH health education curriculum, physical education classes that focused on vigorous physical activity, and lower-fat school lunches. The physical activity results revealed a significant increase from pretest to posttest (2 years) in the percentage of physical edu- cation class time that students in the intervention schools were engaged in moderate-to-vigorous physi- cal activity. Additionally, posttest values were sig- nificantly greater than those for the control schools (Simons-Morton et al. 1991). Although this study did not examine changes in physical activity outside of physical education classes, it highlighted the im- portance of organizational changes to promote physi- cal activity among students. The Sports, Play, and Active Recreation for Kids (SPARK) study, conducted in San Diego, California, tested the effects ofcombininga health-related physi- cal education curriculum and in-service programs on the quantity and quality of physical education classes in elementary schools (McKenzie et al. 1993). In a single school district, 28 fourth-grade classes in jeven schools were randomly assigned to one of three conditions: 10 classes were taught in their usual manner by classroom teachers (control group); 10 classes were taught the SPARK program by cfass- room teachers who had received in-service training and follow-up consultations; and 8 classes were taught the SPARK program by physical education specialists hired by the research project. Direct ob- servation found that students assigned to either of the two intervention groups engaged in significantly more weekly physical activity during physical edu- cation classes than did controls. Teachers who re- ceived the new physical education curriculum and in-service training provided significantly higher- quality instruction than did teachers in the control group, although the trained classroom teachers' in- struction did not match the quality of the instruction provided by the physical education specialists. This study demonstrated that an improved physical edu- cation curriculum, combined with well-designed training for physical education specialists and class- room teachers, can substantially increase the amount of physical activity children receive in school (McKenzie et al. 1993) and can help ensure that the resulting physical education classes are enjoyable (McKenzie et al. 1994). The Child and Adolescent Trial for Cardiovas- cular Health (CATCH) study was a multicenter, randomized trial to test the effectiveness of a cardio- vascular health promotion program in 96 schools in four states (Luepker et al. 1996; Perry et al. 1990, 1992; Stone 1994). A major component of CATCH was an innovative health-related physical education program, beginning at the third grade, for elemen- tary school students. For 2.5 years, randomly assigned schools received a standardized physical education intervention, including new curriculum, staff de- velopment, and follow-up consultations. In these intervention schools, observed participation in moderate-to-vigorous activity during physical education classes increased from 37.4 percent of class time at baseline to 51.9 percent (Luepker et al. 1996). This increase represented an average of 12 more minutes of daily vigorous physical activity in physical education classes than was observed among children in control schools (Luepker et al. 1996; McKenzie et al. 1995). Figure 6-2 shows the effect of CATCH on physical activity during physical educa- tion class. The CATCH study showed that children's 239 Physical Activity and Health Table 6-5. Studies of interventions to increase physical activity among children and adolescents --- study Design Theoretical approach Population School programs Bush, Zuckerman, Taggart, et al. (1989), Bush, Zuckerman, Theiss, et al. (1989) (Know Your Body) 4 year experimental Social learning theory 1,234 students initially in grades 4-6, follow-up in grades 7-9 Killen et al. (1988) 7 week (Stanford Adolescent experimental Heart Health Program) 2-month follow-up Dwyer et al. (1983) 14 week experimental Social cognitive theory 1,447 students in grade 10 None 513 students in grade 5 Simons-Morton et al. (1991) (Go For Health) 3 year quasi-experimental Social cognitive theory 409 grades 3 and 4 PE classes McKenzie et al. (1993) (SPARK) Luepker et al. (1996); McKenzie (in press); Edmundson et al. (1996) (CATCH) 8 month experimental 3 year experimental 112 PE lessons Social cognitive theory 96 schools; 3,239 students initially in grade 3, follow-up at grade 5 School-community programs Kelder, Perry, Klepp (1993) (Minnesota Heart Health Program: Class of 1989 Study) 7 year quasi-experimental Social learning theory Students in grade 6 from 2 Minnesota Heart Health Program communities I = tnterventlon; C = control or comparison; HE = health education; PE = physical education. 240 Understanding and Promoting Physical Activity Intervention .---- Findings and comments t-1 : -I5 minutes, 2 times/week, Know Your Body HE curriculum; health screening and results t-2: -t5 minutes, 2 times/week, Know Your Body HE curriculum; health screening C: Health screening I: 20 classroom PE sessions, 50 minutes each, 3 times/week, HE risk reduction curriculum (1: No intervention I-I : 75 minutes daily PE, fitness curriculum Physical activity not assessed; no differences in academic I-?: 75 minutes daily PE, skill curriculum achievement between intervention and control C: 30 minutes PE 3 times/week, standard groups despite additional 275 minutes of time curriculum spent in PE by intervention groups I: 6 behaviorally based HE modules; five 6- to 8-week modules of PE, children's active PE curriculum; reduced fat and sodium school lunch C: No intervention I-1 : PE provided by PE specialists I-J: PE provided by "specially trained" classroom teachers (1: PE provided by classroom teachers I-1 : HE curricula; PE featuring enjoyable moderate-to-vigorous physical activity; EAT SMART school food service intervention I-2: Same as I-1 with family involvement (1: No intervention I: Peer-led physical activity challenge at grade 8; 10 lesson Slice of Life HE curriculum at grade 10 (1: No intervention Decrease in physical activity for both groups between pretest and follow-up. No difference in physical activity between groups at posttest. Increase in posttest knowledge by each group. Great increases in knowledge by intervention groups at posttest, 18% response rate at 4-year follow-up Intervention groups compared with control had a higher proportion of nonexercisers at baseline exercising at follow-up Increase from pretest to posttest in the percent of PE class time intervention school students spent in moderate-to- vigorous physical activity; higher percentage of PE class time spent in moderate-to-vigorous physical activity by intervention schools compared with controls in posttest At posttest PE specialists spent more minutes per lesson on very active physical activity and fitness activities than specially trained classroom teachers and classroom teachers; specially trained classroom teachers spent more minutes per lesson on very active physical activity and fitness activities than classroom teachers Intervention schools compared to control schools provided a greater percentage of PE time spent in moderate to vigorous physical activity at posttest; family involvement had no effect on physical activity and psychosocial outcomes; data from the intervention groups combined for comparison with the control groups; intervention students were not different from control students in total daily physical activity at posttest; intervention students spent 12 more minutes per day engaged in vigorous physical activity than controls; pretest-to-posttest increases in students' perceptions of self-efficacy for exercise and positive social reinforcement for exercise among both intervention and control students; intervention students' posttest scores on these and other psychosocial measures were not different from those of control students At 7-year follow-up students from schools in intervention community had higher levels of physical activity than students from schools in control community, particularly among girls; 45% response rate at 7-year follow-up 241 Physical Activity and Health Figure 6-2. Moderate-to-vigorous and vigorous physical activity observed during Child and Adolescent Trial for Cardiovascular Health (CATCH) physical education classes 60 Intervention Intervention --- -__L -L------/H Vigorous lo- Ol I I I I I 2 3 4 5 6 Semester Source: Luepker RV et al. journal of the American Medical Association 1996 (reprinted with permission). Note: Observed at six time points, 1991 through 1994. The CATCH intervention, introduced during semester 2. increased the percentage of time spent in moderate-to-vigorous and vigorotis activity JS measured by the System for Observtng Fitness Instruction Time classroom observation system. Intervention and control curves diverged significantly according to repeated-measures analysis of variance with the class session as the unit of analysis: ior moderate- to-vigorous activity, P = 2.17, df = 5, 1979, P = .02; for vigorous activity, F = 2.95, df = 5, 1979, P = 04. Analysts controlled for CATCH site, the location of the lesson, the specialty of the teacher, and random variation among schools and weeks of observation. physical activity can be increased by a standardized intervention applied to existing physical education programs in four geographically and ethnically di- verse regions. Although the intervention students showed significant pretest to follow-up increases in their perceptions of positive social reinforcement and self-efficacy for exercise (Edmundson et al. 1996), these psychosocial determinants were not significantly more prevalent than those observed among the control groups at follow-up (Luepker et al. 1996). Although the family intervention compo- nent produced no additional increase in physical activity among students (Luepker et al, 1996), the CATCH physical education and classroom programs successfully increased moderate-to-vigorous physi- cal activity in physical education class and increased students' daily participation in vigorous physical activity. School-Community Programs The Class of 1989 Study (Kelder, Perry, Klepp 1993; Kelder et al. 1995), an ancillary study of the MHHP (Luepker et al. 1994), tested the efficacy of a school- based health promotion program. One of three MHHP intervention communities and its matched pair were involved in the Class of 1989 Study. The intervention cities were engaged in an extensive communitywide intervention program designed to improve eating, exercise, and smoking patterns for the entire population. The physical activity intervention included a peer-led physical activity challenge, in which students were encouraged to engage in out-of-school exercise activities. The program's assessment included annual measurements collected from a large number of students (baseline n = 2,376) for 7 years, beginning in the sixth grade. Throughout most of the follow-up period, physical activity levels were significantly higher among female students in the intervention community than among those in the control community. For male students, the levels did not differ significantly between the communities. Results suggest that at least among female students, a multicomponent intervention that includes peer-led behavioral education in schools and complementary communitywide strategies can increase levels of regular physical activity (Kelder, Perry, Klepp 1993; Kelder et al. 1995). Interventions in Health Care Settings Health professionals also have a potential role in promoting physical activity, healthy eating, and other health behaviors among children and adolescents (American Medical Association 1994; U.S. Preventive Services Task Force 1996). Results of a national sur- vey of pediatricians showed that one-half of respon- dents believed that regular exercise during childhood is important in preventing cardiovascular disease in adulthood (Nader et al. 1987). However, only one- fourth believed they would be effective in counseling their young patients to get regular vigorous exercise. The American Medical Association's Guidelines for Adolescent Preventive Health Services (1994) is one 242 example of practical counseling recommendations that have been developed for those who provide health services to adolescents. Special Population Programs Physical activity can assist in the treatment or reha- bilitation of several diseases that occur during youth (Rowland 1990; Greenan-Fowler 1987); however, relatively few interventions have been conducted to examine how to promote physical activity among young people with special needs. The most extensive study is a series of randomized investigations of children who are overweight (Epstein, Wing, Valoski 1985; Epstein, McCurley, et al. 1990; Epstein, Valoski, et al. 1990; Epstein et al. 1994). In this series, family- based treatments of 5- to 12-year-old obese children incorporated both physical activity and nutrition in- terventions, and the programs were based specifically on principles of behavior modification. Parents were trained to improve their children's physical activity by setting behavioral change goals with their children, by identifying effective reinforcers (e.g., spending time with parents), and by reinforcing children when goals were met. Ten-year follow-ups of children in these four randomized studies revealed that 30 percent of children receiving family-based interventions were no longer obese, and 20 percent had decreased their percentage overweight by 20 percent or more (Epstein ct al. 1994). The lo-year follow-up investigation also revealed that the percentage of overweight children in each study decreased most when the intervention involved both the parent and the child or when a change in lifestyle exercise was emphasized. Epstein and colleagues (1994) also compared the effective- ness of three forms of physical activity interventions: lifestyle physical activity, in which activity was incor- porated into daily routines; structured aerobic exer- cise; and calisthenics. At the lo-year follow-up, the lifestyle group had lost the most weight, and both the lifestyle group and the aerobic exercise group had greater weight-loss results than the calisthenics group (Epstein et al. 1994). Summary The preceding review of the research literature on interventions among young people reveals that school-based approaches have had consistently strong effects on increasing physical activity in Understanding and Promoting Physical Activity elementary school students when the intervention orients the physical education program toward de- livering moderate-to-vigorous physical activity. Further, social learning theory appears to have had the widest application to this interventions re- search. Much research has taken place at the el- ementary school level; very little is known about increasing children's physical activity in middle and high school physical education classes or in settings other than school physical education classes. It seems likely that these interventions would be strengthened by designing programs that combine school and community policy with health educa- tion and physical education. Data are lacking on ways to tailor interventions to the needs and inter- ests of young people and to prevent the rapid decline in physical activity that occurs during late childhood and adolescence, especially among girls. Additionally, few physical activity interventions and research studies encompass populations par- ticularly characterized by race/ethnicity, socioeco- nomic status, risk factor status, disabilities, or geographic location. Promising Approaches, Barriers, and Resources Many questions remain about how best to pro- mote physical activity in the general population of young people and adults, as well as in clinical populations and other subgroups. Policy initia- tives, the provision of more physical activity facili- ties and programs, and media campaigns are promising, but studies testing their effects are limited. The following two sections describe exist- ing policy and program approaches' that have the potential to increase population levels of physical activity but have received little or no evaluation. They are reviewed separately from the previously discussed, better-documented research studies. `Descriptions of specific physical activity programs across the United States can be found in the Combined Health Information Database, a computerized bibliographic database of health infor- mation and health promotion resources developed and managed by several federal agencies, including the CDC, the National Institutes of Health, the Department of Veterans Affairs, and the Health Resources and Services Administration. Intended for all health professionals who need to locate health information for themselves or their clients, this resource is available in many libraries, state agencies, and federal agencies. 243 Physical Activity and Health Environmental and Policy Approaches Most interventions that have been evaluated in re- search studies are discrete programs targeting popu- lation subgroups (e.g., employees, schoolchildren) or communities. Interventions have shown some success in promoting physical activity, but their results have been inconsistent. A possible reason for limited results is a lack of concomitant support from the larger environment within which such interven- tions take place. Many physical activity researchers believe that environmental and policy interventions must occur to complement interventions that focus on behavior change among individuals or small groups. This larger perspective recognizes the pow- erful moderating effect that environment has on individual volition. As King, Jeffery, and colleagues (1995) observe, "Environmental and policy inter- ventions are based on the recognition that people's health is integrally connected to their physical and social environments" (p. 501). Two premises underlie environmental and policy approaches. First, interventions addressing chronic disease risk factors, such as physical inactivity, re- quire comprehensive, population-based approaches that incorporate both individual and societal-level strategies (Green and Simons-Morton 1996; Schmid, Pratt, Howze 1995). Second, strategies should not rely solely on active approaches requiring individual initiative, such as enrolling in exercise classes, but should also incorporate passive approaches, such as providing walking trails or policies that permit em- ployees to exercise during work hours (Schmid, Pratt, Howze 1995). An example of intervention elements combining passive and active approaches is a school board policy that permits school facilities to remain open before and after school for commu- nity use, together with health communications that make citizens aware of these facilities and encour- age their use. As presented previously, ecological models of health behavior (McLeroy et al. 1988; CDC 1988; Stokols 1992) provide frameworks for conceptualiz- ing what the role of policy approaches is to health promotion and how individuals interact with their social, institutional, cultural, and physical environ- ments. The concept of the health-promoting envi- ronment suggests that communities and other settings can facilitate healthy behaviors by providing envi- ronmental inducements to be active, such as by 244 offering safe, accessible, and attractive trails for walking and biking. National objectives and recommendations have encouraged the development of policies, programs, and surveillance strategies that would help create an environment that promotes physical activity (USDHHS 1990; Pate, Pratt, et al. 1995; National Association for Sport and Physical Education 1995; U.S. Department OfTransportation [USDOT] 1994). Increasing national levels of physical activity and of cardiorespiratory fitness has also been targeted as a priority health objective in Healthy People 2000 (USDHHS 1990) and the Dietary Guidelines far Americans (U.S. Department of Agriculture and USDHHS 1995). Many efforts to raise public awareness and pro- mote physical activity are under way. In 1994, the American Heart Association, the American College of Sports Medicine, and the American Alliance for Health, Physical Education, Recreation and Dance formed a National Coalition for Promoting Physical Activity. The coalition's goals are to increase public awareness of the benefits of physical activity, pro- vide an opportunity for forming effective partner- ships, and enhance delivery of consistent messages about physical activity (National Coalition for Pro- moting Physical Activity 1995). The CDC has estab- lished guidelines for promoting physical activity and healthy eating among young people (CDC 1996; CDC in press) and has initiated a public education effort to encourage active lifestyles and healthy eat- ing among Americans. The National Institutes of Health (NIH) has used national campaigns to pro- mote messages to both the general public and pa- tients on the importance of physical activity and a heart healthy diet. The NIH also sponsors research on physical activity in special populations, includ- ing women from diverse economic backgrounds, and in various settings, such as worksites, schools, and health care institutions. In 1995, the NIH spon- sored the Consensus Development Conference on Physical Activity and Cardiovascular Health, which recommended regular physical activity for most persons aged 2 years and older (see Appendix B in Chapter 2). The President's Council on Physical Fitness and Sports works with a broad range of partners in private industry, voluntary organiza- tions, and the media to promote physical activity, fitness, and sports participation by Americans of all ages. As part of the midcourse review of the physical activity and fitness objectives of Healtlry People 2000, the council presented a synopsis of ongoing grassroots activities by Healthy People 2000 Consortium mem- bers in support of increasing participation in physi- cal activity and improvement in fitness (USDHHS 1995). The President's Council on Physical Fitness and Sports is also an advisory body to the President and to the Secretary of the DHHS on matters involv- ing physical activity, fitness, and sports that enhance and improve health. Thirty-nine Governor's Coun- cils on Physical Fitness and Sports stimulate state ;rnd local activities and program development; these efforts target fitness promotion for school-aged ),ouths, older adults, working adults, and families (National Association of Governor's Councils on Physical Fitness and Sports 1996). Community-Based Approaches Community-based programs can be tailored to meet the needs of their specific populations. More col- laborative work is under way between state and local governments, community groups, and businesses to reduce risk factors among employees and residents. Two-year follow-up data from one such effort in Smyth County, Virginia, suggested that 40 percent of school system employees had increased their physi- cal activity participation during the program period (CDC 1992). Two large subpopulations may be especially im- portant to address in community-based programs: )`oung people and older adults. Communities will face. a growing need to provide a supportive environment for their children and adolescents. Between 1995 and 2020. the number of young people under 18 years old will increase by an estimated 13 percent, from 69 million to 78 million (Bureau of the Census 1996). The framework for community-level physical activity programs for young people is already in place: mil- lions of American youths participate in sports spon- sored by community leagues, religious organizations, social service organizations, and schools. In addition to organized sports, communities need to provide recreational programs and opportunities for all young people in a community, because such programs may encourage a lifetime habit of physical activity as well ~1s other immediate community benefits. According to The Trust for Public Land, arrests among young people in one community decreased by 28 percent Understanding and Promoting Physical Activity after the community instituted an academic and rec- reational support program for teenagers (National Park Service 1994). In another community, juvenile crime dropped 55 percent when community recre- ational facilities stayed open until 2 a.m. (National Park Service 1994). Communities will also need to meet the chal- lenges of a growing population of older adults. Be- tween 1995 and 2020, the number of people over the age of 60 will increase by 43 percent, from 44 to 63 million (Bureau of the Census 1996). Programs and facilities designed to meet the needs of aging baby boomers and older adultscan help ensure that these rapidly growing segments of the population obtain the health benefits of regular, moderate physical activity. In one community, 35 age-peer exercise instructors for older adults were recruitedand trained by a local university as volunteers to conduct age- appropriate physical activity programs on a regular basis at sites such as libraries, senior centers, and nursing homes in their neighborhoods. Because they were age peers, the instructors were sensitive to many of the concerns that older adults had about physical activity, such as fear of falling and fracturing a hip. Over the following year, instructors conducted more than 1,500 half-hour exercise programs for more than 500 older adults at 20 sites (DiGilio, Howze, Shack 1992 >. Places of worship represent a potentially effec- tive site for physical activity promotion programs in communities, since these settings can provide the impetus for starting-and the social support for maintaining-behavioral regimens (Eng, Hatch, Callan 1985; Eng and Hatch 1991) such as regular physical activity. Among the advantages of such settings are a history of participating in a range of community health and social projects; large mem- berships, including families; a presence in virtually every U.S. community; and connections to minority, and low-income communities typically underserved by health promotion programs (King 1991). The Fitness Through Churches Project promoted aero- bic exercise in conjunction with other health behav- iors to African American residents of Durham, North Carolina (Hatch et al. 1986). The results from this pilot program suggest that physical activity pro- grams offered at places of worship are feasible and attractive to clergy and their congregations. Another project, the Health and Religion Project (HARP) of 245 Physical Activity and Health Rhode Island (Lasater et al. 1986), found that volun- teets can be trained to provide heart health pro- grams, including physical activity, in church settings (DePue et al. 1990). Societal Barriers The major barrier to physical activity is the age in which we live. In the past, most activities of daily living involved significant expenditures of energy. In contrast, the overarching goal of modern technology has been to reduce this expenditure through the production of devices and services explicitly de- signed to obviate physical labor. From the days of hunting and gathering to turn-of-the-century farm- ing practices and early industrial labor, the process of earning a living was once a strenuous activity. Today, many Americans engage in little ot no physical activity in the course of a working day typically spent sitting at a desk or standing at a counter or cash register. A large part of many people's time is spent inside buildings where elevators or escalators are prominent features and stairs ate difficult to find and may seem unsafe. Motorized transportation carries millions of Americans to and from work and on almost every errand. These inactive daily expedi- tions occur virtually door-to-door, with the help of parking lots built as neat to destinations as possible to minimize walking and increase convenience and safety. Whereas older cities and towns, were built on the assumption that stores and services would be within walking distance of local residents, the design of most new residential areas reflects the supposition that people will drive from home to most destina- tions. Thus work, home, and shopping ate often separated by distances that not only discourage walk- ing but may even necessitate commuting by motor- ized transportation. Television viewing, video games, and computer use have contributed substantially to the amount of time people spend in sedentary pursuits (President's Council on Physical Fitness and Sports and Spotting Goods Manufacturers Association 1993). Next to sleeping, watching TV occupies the greatest amount of leisure time during childhood (Dietz 1990). Preschoolers exhibit the highest rate of TV watching (27-28 hours per week). By the time a person gradu- ates from high school, he ot she will likely have spent 15,000-18,000 hours in front of a television-and 12,000 hours in school (Strasburger 1992). In the face of these powerful societal induce- ments to be inactive, efforts must be made to encour- age physical activity within the course of the day and to create environments in communities, schools, and workplaces that afford maximum opportunity to be active. Policy interventions can address public concerns about safety, financial costs, and access to indoor and outdoor facilities. Such interventions also can address the concerns of employers and governments about liability in the event of injury. At the state and local level, governments determine building codes and public safety, traffic, and zoning statutes that have potential bearing on physical ac- tivity opportunities in communities. Concerns about crime can be a major barrier to physical activity for both adults and young people. In a national survey of patents, 46 percent believed their neighborhood was not very safe from crime for their children (Princeton Survey Research Associates 1994). Minority parents were about half as likely as white parents to report that their neighborhoods were safe. Successful implementation ofpolicy inter- ventions may help address such concerns. For ex- ample, decisions to put more police on a beat in a high-crime area may help residents feel safer going outside to walk. Similarly, neighborhood watch groups formed to increase safety and reduce crime may be a vehicle for promoting physical activity. Opening schools for community recreation and malls for walking can provide safe and all-weather venues that enable all members of the community to be active. Transportation, health, and community plan- ners as well as private citizens can help ensure that children living in areas near schools can safely walk or bike to school and that adults can walk or bike to work. Fear of traffic is one of the most frequently cited reasons for not bicycling (USDOT 1993). Adult pedestrians and bicyclists account for 14 percent of yearly traffic fatalities (USDOT 1994). In a survey of adults, those who rode a bicycle in the preceding year were asked whether they would commute to work by bicycle under specific conditions. Fifty-three per- cent said they would do so if safe, separate, desig- nated paths existed; 47 percent would if their 246 employer offered financial or other incentives; 46 percent would if safe bike lanes were available; and 45 percent would if their workplace had showers, lockers, and a secure area for bike storage (USDOT 1994). More than half the respondents indicated they would walk, or walk mote, if there were safe pathways (protected from automobile hazards) and if crime were not a consideration. A majority also wanted their local government to provide better opportunities to walk and bicycle. These percentages stand in sharp relief against current practice: only 4.5 percent of Americans com- mute to work by bicycle or on foot (USDOT 1994). Even in such comparatively small numbers, these people are estimated to save as much as 1.3 billion gallons of gasoline yearly and to prevent 16.3 million metric tons of exhaust emissions (USDOT 1994). Every mile walked or cycled for transportation saves 5 to 22 cents that would have been spent for a mile by automobile, including reduced cost from pollu- tion and oil imports (USDOT 1994). The Intermodal Surface Transportation Efficiency Act, passed in 1991, promotes alternatives to automobile use by making funds available for states to construct or improve bicycling facilities and pedestrian walkways (USDOT 1993). Decisions on how these funds are used ate made locally, and organizations such as local transportation, health, and parks departments can promote the use of these funds in ways that increase the prevalence of physical activity in their communities. In a growing number of communities, concerns nbout environmental quality have led to zoning restrictions that protect open spaces and other areas that can subsequently be used for recreational pur- suits. Such greenways, or linear open space, can connect neighborhoods and foster the use of bicy- cling and walking for transportation (Indianapolis I)cpartment of Parks and Recreation 1994). Societal Resources .Jlthough there is no comprehensive listing of physi- cal activity resources in the United States, such a document would be extensive. Millions of ,Jmericans have sports supplies, bicycles, and exercise lnachines in their homes or have access to public and private resources such as tennis courts, parks, play-