... to identify barriers to. PA among Brazilian elderly women from different SES. ..... American College of Sports Medicine and the American. Heart Association.
Journal of Physical Activity and Health, 2011, 8, 126 -132 © 2011 Human Kinetics, Inc.
Barriers to Physical Activity Among Brazilian Elderly Women From Different Socioeconomic Status: A Focus-Group Study Ana Carina Naldino Cassou, Rogerio Fermino, Ciro Romélio Rodriguez Añez, Mariana Silva Santos, Marlos R. Domingues, and Rodrigo S. Reis Background: The aim of this study was to identify barriers to physical activity among elderly Brazilian women of different socioeconomic status (SES). Methods: A focus-group approach was employed. Subjects were aged, on average, 69.9 years (±6.9; n = 25). SES was measured based on a structured interview and women were grouped according to SES classification. Content analysis was used to categorize mentions of barriers to physical activities followed by descriptive analysis of absolute and relative frequencies of similar reports. Results: Most common barriers among high-SES elderly women were those within “psychological, cognitive, and emotional” dimensions (33.8%) and “environmental” (29.2%). Among women from lower SES, barriers were inversely ranked, the highest prevalence was verified for environmental (33.8%) and “psychological, cognitive, and emotional” dimensions (25%). Conclusions: The results highlight that barriers perception varies according to women’s SES, indicating that physical activity promotion strategies must address such differences. Keywords: physical activity, focus groups, aging The proportion of the population who are elderly is increasing significantly. In Brazil, nearly 15 million people are older than 60 and it is expected that in the year 2030 this figure will reach 25 million people.1 Changes verified in the population pyramid age group play a major role in public health. A higher proportion of old people brings higher costs to the health system, especially with aging-related conditions, such as chronic diseases and hospital admissions.2 Regular physical activity (PA) is associated with chronic degenerative disease prevention and lower mortality/morbidity rates among the elderly.3,4 However, despite these benefits, PA levels are likely to decrease with age.5–7 Data from VIGITEL7 (a telephone-based surveillance system for chronic diseases risk factor monitoring) showed that physical inactivity among Brazilian people is around 29.5% and 23.5% in men and women, respectively. In individuals older than 65, prevalence of inactivity rises to 51.7% (men) and 53.2% (women). PA engagement may be influenced (negatively or positively) by many factors. When a factor hinders PA participation (negative factor) is also known as a barrier; and positive factors are those that make PA engagement easier. Sallis & Owen8 categorize these determinants into 6 dimensions (demographic and biologic; psychological, Cassou, Fermino, Reis, Rodriguez Añez, and Santos are with the Dept of Physical Education, Pontific Catholic University of Paraná, Curitiba, Brazil. Domingues is with the Sports Dept of the Postgraduate Program in Physical Education, Federal University of Pelotas, Pelotas, Brazil.
126
cognitive, and emotional; cultural and social; environmental; PA characteristics; and behavioral aspects), revealing the complexity and multiplicity of the aspects behind PA engagement. Some studies carried out with elderly people have presented a subset of PA correlates, in distinct settings, that might be associated to PA engagement.9–11 In Australia, Booth et al9 found that the main reported barriers were injuries or physical incapacity, poor health, and feeling too old. In Brazil, a population-based study in Pelotas (South region) by Reichert et al10 found as main barriers to PA the lack of money, feeling too tired, lack of company and time. Among individuals aged 60 to 69, the most frequently reported barriers were lack of money, injuries or a disabling condition, feeling too tired, and fear of injury. Moreover, high-SES people reported not participating in physical activities because they felt too tired. However, low-SES individuals mostly cited lack of money as a barrier to PA. The understanding of the factors influencing Brazilian women’s PA engagement is quite incomplete, in part because representative studies have not been accomplished, but also due to the nonexistence of accurate instruments to assess these factors. Published works have used instruments developed elsewhere, which may not be adequate for local conditions. To deeply understand the elements influencing PA, can be extremely helpful in the development of population-specific instruments. In addition to that, the results may be used to guide planning of public health strategies and interventions to increase PA. Therefore, the aim of this study was to identify barriers to PA among Brazilian elderly women from different SES.
Barriers to Physical Activity
Methods Participants Twenty-five women (mean age = 69.9 ±6.9) took part of the study and were selected by convenience from communitarian groups gathering elderly people (over 60) from Curitiba. Curitiba city is the capital of Paraná state located in the South region of Brazil with a population of approximately 1.797.408 inhabitants and 84.8% of its population is white.12 Curitiba has been considered a model of health promotion and in quality of life, especially because of green areas and sustainable development strategies. Women belonged to singing, arts, recreation, or PA (gymnastics, pool, bowling, etc.) groups, and met once a week. Individuals who agreed to participate in the research responded to a questionnaire to check characteristics to define their eligibility for participation in the focus-group interviews. To be included, women should have autonomy, functional independence, and good health. To check those characteristics, a daily-life activities questionnaire was administered.13 Health problems perception and Brazilian economic criteria classification were also assessed.14 After this stage, they were invited to join focus groups. Only 16% reported a negative health self-perception and 32% had existing health condition (mostly chronic degenerative diseases). Women were informed about the research procedure and agreed to voluntarily participate by signing a consent form according to the Brazilian ethics guidelines for human research. The study protocol was approved by the Ethics in Research Committee of the Pontiff Catholic University of Parana (process number 687).
Focus Groups The number of participants in each group was chosen as proposed by Stewart & Shamdasani.15 The suggestion is that the groups should be homogeneous and including, at most, 12 people. Larger groups impair interview management. The number of choice for the study we aimed to form groups of 6 people. To assure homogeneity of characteristics within groups, participants were selected according to their SES,14 identified in the beginning of the data collection, and later categorized by education of head of the family and an assets score, into 2 levels: high (A and B classes—48%) and low (C, D, and E classes—52%). Therefore, 4 groups were arranged: 2 high-SES groups (6 participants each) and 2 low-SES groups (6 and 7 participants). Subjects were informed and invited to come for interview at specific place and time.
Focus-Group Interviews Interviews followed a previously set protocol, tested during a pilot-study. Attention was dedicated to ethical aspects, place of interviews, moderator experience, schedules, and interview length as well as information
127
recording. Groups were interviewed in pleasant environments, quiet and familiar to participants, in an attempt to avoid potential biases. The participation was voluntary and no reward was provided. Groups were coordinated by a single researcher, previously trained. Interviews were based on a set of open-ended questions and topics concerning individual, social, and environmental PA determinants. The script included 13 topics divided in to 3 subsets. First, the moderator was introduced, the aims of the study were discussed briefly and each participant introduced herself to promote integration within the group. The second part comprised an encouraged discussion. Images displaying exercising people were shown, and each woman was asked to report what that image represented to her. The last part covered PA questions (what, how, when, where, and why they were practicing PA—or not) highlighting individual, environmental, social, and psychological aspects. Questions were arranged in a way to not embarrass participants, trying to create a friendly environment to promote everybody’s involvement. Interviews took 50 to 70 minutes; minimum estimated time was 50 minutes and could not last more than 90 minutes. Answers and discussions were recorded by an assistant who could not intervene in the discussion. By the end of the interviews, a written summary of important points and an integral audio version of the conversation were obtained. Recording was done only with the consent of participants and was finished by the end of discussion. Later, audio files were entirely transcribed and a code was assigned to each participant (P1, P2, P3, etc.), to assure confidentiality.
Data Analysis Data analysis followed a quali-quantitative approach. First, focus-group information was analyzed in a qualitative manner, using a content analysis. Information concerning negative influences on PA engagement were highlighted and classified as barriers. Then, data were categorized in determinants, to be later grouped according to dimensions as suggested by Sallis & Owen.8 The following examples were extracted from the original reports and provide more detailed information to clarify the approach we used in our analysis. For example, one subject said “streets near my house are dark and not safe to walk alone” and that sentence was considered a report of “lack of safety,” which was considered as “environmental characteristics,” and finally included in the “environmental” dimension. The sentence “I don’t feel that I have skills to practice some physical activities or sports” was classified as “lack of self-efficacy” and included in the “psychology, cognitive, and emotional” dimension. Another subject said “when I feel pain in my knee and back I stop all practice,” this statement was considered as “physical limitations” and included in the “demographic and biological” dimension. All the quotations that we could not appropriate classify according to the strategy employed were considered “barriers to
128
Cassou et al
exercise.” For instance, the following sentence “I feel embarrassed while I am exercising with other people around” was classified as “barrier to exercise.” After content analysis and categorization that followed, descriptive statistics were carried out to obtain relative and absolute frequencies of similar reports. Before analysis, reports were divided according to PA barriers and dimensions, considering each group’s SES (high or low). Figure 1 displays the data analysis chart. All analyses were carried out with SPSS 11.0.
Results During analysis of the interviews, 198 barriers were identified; 130 (65.7%) reported by high-SES and 68 (34.3%) by low-SES women. Among high-SES elderly, the most frequent reasons to not exercise were (14 reports—10.8%), lack of social support and everyday obstacles, or barriers faced during their daily life, (10 reports each—7.7%) and weather, social isolation and health conditions (8 reports each— 6.2%). Among low-SES women, most cited barriers were cost, everyday obstacles, and household chores (7 reports each—10.3%), lack of time (6 reports—8.8%), barriers to exercise and lack of safety (5 reports each—7.4%). Remaining reported barriers, frequencies and dimensions are presented in Table 1. Figure 2 presents frequencies of barriers grouped according to dimensions. Among elderly women belonging to high SES the negative influence of psychological, cognitive, and emotional dimension was cited 44 times
Figure 1 — Data analysis chart.
(33.8%) while environmental factors were cited 38 times (29.2%). Elderly from low SES reported as negative factors to engaging in PA the environment-related dimension 23 times (33.8%) and 17 times (25%) the psychological, cognitive and emotional dimension.
Discussion The current study aimed to explore, identify and describe the PA barriers among elderly Brazilian women from different SES. Report analysis allowed observing differences in PA perceived barriers in relation to SES. For women from high SES, psychological, cognitive, and emotional factors were more important, while low-SES women pointed to environmental aspects as more important. These factors were inversely ranked by group. Although this study did not aim to quantify the number of barriers reported by different SES groups, we found more barriers among high SES group (n = 130) than in low SES (n = 68). It is hypothesized that this difference is due to a higher education level, which could affect one’s opinion of tangible and psychological costs of certain behavior. This assumption is supported by health behavior theories such as the Health Belief Model.16 However, is not possible to further test this finding because of the exploratory nature of the study. In the higher SES group, psychological, cognitive, and emotional aspects prevail in the literature as potential explanation for a physically inactive behavior. These results are somewhat similar to those reported by Brownson et al,17 whose work evaluated 1818 people by telephone interviews, and have found that exercising at work, lack of time, feeling tired, and lack of motivation were the most reported barriers, regardless of economic status. In another study, Booth et al9 observed that feeling too old, lack of time, feeling not sporty, lack of motivation, need for rest, lack of persistence, and fear of injury are also important barriers to PA, changing according to age. Lees et al18 carried out a focus-group research and concluded that laziness, fear of injury, and lack of time were highly reported among people older than 65. Among low-SES subjects, most reported barriers were environment-related. Some studies have shown that such factors influence physically active behaviors significantly,17,19–23 Dawson et al21 identified that, for the elderly people, car traffic, poor conditions of sidewalks, unavailability of facilities or public spaces, and lack of safety, are among main barriers. Galea et al23 carried out a focus-group study and revealed that uneven surfaces were highly cited by elderly with intermittent claudication. Other environment-related features like lack of equipment/gear, activity participation costs, and lack of physical structures or facilities were also frequently reported.24 The existence of such barriers can partially be explained by the lower access to public PA places among low-SES people. In fact, Reichert et al10 identified that in individuals older than 60, lack of money was an important barrier to PA participation. On the other hand,
Table 1 Barriers to Physical Activity Reported by Elderly Women According to Socioeconomic Status Determinant Barrier to exercising
High (n = 12) AF 14
RF 10.8
Low (n = 13) AF RF 7 10.3
Lack of social support Routine obstacles Weather
10 10 8
7.7 7.7 6.2
Dimension Psy, Cog and Emo Cult and Soc Environ Environ
Determinant Costs Routine obstacles Household chores Lack of time
7 7 6
10.3 10.3 8.8
Dimension Environ
Social isolation
8
6.2
Cult and Soc
Barrier to exercising
5
7.4
Health conditions Environment characteristics
8 7
6.2 5.4
Demo and Biol Environ
5 4
7.4 5.8
Lack of motivation
7
5.4
4
5.8
Demo and Biol
Exercise dislike
6
4.6
Occupation
4
5.8
Demo and Biol
Localization Household chores
6 5
4.6 3.8
Psy, Cog and Emo Psy, Cog and Emo Environ Cult and Soc
Lack of safety Lack of social network Physical restrictions
Environ Cult and Soc Psy, Cog and Emo Psy, Cog and Emo Environ Cult and Soc
Weather Lack of motivation
3 3
4.4 4.4
Lack of safety Lack of time
5 5
3.8 3.8
Health conditions Social isolation
3 2
4.4 2.9
Taking care of children
5
3.8
Environ Psy, Cog and Emo Cult and Soc
Environ Psy, Cog and Emo Demo and Biol Cult and Soc
Low self-efficacy
1
1.5
Occupation
4
3.1
Demo and Biol
1
1.5
States of change
3
2.3
1
1.5
Equipment/gear Lack of knowledge on health and exercise No intention to exercise
2 2
1.5 1.5
Age Localization
1 1
1.5 1.5
Demo and Biol Environ
2
1.5
Other problems
1
1.5
Fear injury
2
1.5
Physical problem
1
1.5
Psy, Cog and Emo Demo and Biol
Social network
2
1.5
Psy, Cog and Emo Environ Psy, Cog and Emo Psy, Cog and Emo Psy, Cog and Emo Cult and Soc
Excessive exercise control Perceived effort
Psy, Cog and Emo Psy, Cog and Emo PA Charac
Taking care of children
1
1.5
Cult and Soc
Perceived fitness and health
2
1.5
Perceived access to facilities
1
0.8
Class’ characteristics Teacher’s characteristics
1 1
0.8 0.8
PA Charac
Perceived effort
1
0.8
PA Charac
Lack of group cohesion
1
0.8
Cult and Soc
Absence of fun
1
0.8
Age
1
0.8
Psy, Cog and Emo Demo and Biol
Total
130
100
68
100
Psy, Cog and Emo Environ Cult and Soc
Abbreviations: AF, absolute frequency; RF, relative frequency; Psy, Cog, and Emo, psychological, cognitive, and emotional; Environ, environmental; Cult and Soc, cultural and social; Demo and Biol, demographic and biologic; PA Char, physical activity characteristics.
129
130
Cassou et al
Figure 2 — Percent of reports categorized in dimensions according to socioeconomic status (Psy, Cog, and Emo: psychological, cognitive, and emotional; Environ: environmental; Cult and Soc: cultural and social; Demo and Biol: demographic and biologic; PA Char: physical activity characteristics).
according to Fisher et al19 people living in low–SES areas walk more than their counterparts living in richer areas, however, the current study did not consider commuting physical activities, not allowing comparison. The third most reported dimension by both groups was the cultural/social, especially among those from higher SES. That dimension is connected to interpersonal aspects, including social support, attitudes, incentives, and praises, that result in higher chance of an active lifestyle.8 These results are similar to previous studies. Eyler et al25 have found that women with little or no social support from the family or friends had a higher chance of being inactive. Reichert et al,10 Lim et al26 and Crombie et al27 also showed that lack of partners, disliking exercising or walking alone and the absence of social support are important barriers to PA. Among women from low SES, domestic chores were considered a barrier to PA. Although household tasks have decreased during the twentieth century,28 in Brazil this kind of work is mostly done by women and more hours dedicated to these activities are observed in women with respect to household chores (27.2 hours/week for women versus 10.7 for men).29 Thus, the buildup of tasks may decrease availability of time and willingness for other activities. However, no evidence was found in the literature that could explain such association. Another cultural barrier reported by elderly from high and low SES is the time spent taking care of (grand) children. In Brazil, many couples leave children with grandparents while out for work. Booth et al9 showed that people older than 60 report the factor “taking care of children” as a barrier to PA. Burton et al30 have also described that barrier. Baker & Silverstein31 showed that grandparents taking care of grandchildren are less concerned about their own health. The authors suppose
that grandparents devote more time to grandchildren in early years also because low-SES parents cannot afford babysitters. Factors that were classified as “demographic and biological” dimension are a set of aspects that may influence PA participation. However, many characteristics cannot be changed, such as innate aspects (sex, age), assets (SES), habits (occupation), and other circumstantial conditions (health issues) that depend on social contexts as well.8 Reichert et al10 showed that demographic and biological barriers are important for the Brazilian population, but methodological differences impair a direct comparison with the current study. Internationally, several studies have tried to investigate the factors that affect negatively PA.9,21,30,32 In spite of distinct methodologies, most point out the importance of “demographic and biological” factors. In Australia, Booth et al,9 carried out a population-based study and observed that injuries, poor health, lack of money and taking care of children were barriers to PA among older individuals (60 or older). Burton et al30 discuss an inverse an association between age/education and PA; as age and education increase, the chance of being physically active decreases. Der Ananian et al33 also mention physical limitations and pain as aspects influencing PA among elderly people. In a study by Galea et al,23 unspecific body discomfort and pain were among most cited barriers. The results of the current study seem to be consistent with literature even though the understanding of how individual aspects interact to influence of PA participation is very complex. For example, considering that high-SES individuals enjoy higher access, can easily afford PA participation and have better knowledge about the benefits, the perception of the barriers among them could be connected to unchangeable domains (advanced age, body pains, physical and chronic
Barriers to Physical Activity
conditions, etc.). Nevertheless, such a model of determination could be not be fully explored in the current study. An aspect to be considered in the current study is the number of participants. The sample encompassed 25 individuals, not representative of the original population. The exploratory nature of the current study provides in depth information, which are useful to generate hypotheses. In fact, larger samples or groups were not easily workable due to time needed for interviewing and transcriptions. Therefore, the results should be carefully considered and generalization is limited. The hypotheses obtained using this approach should be further tested in larger and representative samples from different populations, perhaps using distinct instruments and designs. Despite methodological characteristics, focus-group approach enables to clearly identify individual perceptions, encouraging discussions between participants and stimulating an enormous assortment of characteristics not identified by questionnaires. In conducting the literature review for the current study, the authors were unable to identify studies that relied on focus-groups methodology to assess barriers to PA among elderly people from developing countries. These results may assist future creation of more objective instruments, such as scales or questionnaires, to assess barriers reported by elderly in similar settings.
4. 5.
6. 7. 8. 9.
10.
11. 12.
Conclusion Elderly women belonging to distinct SES groups perceived the barriers to PA differently. Low-SES women are more concerned with environmental features while richer women are more affected by psychological, cognitive, and emotional aspects. According to the results, new strategies and policies in the PA area should focus on population groups bearing in mind the economic situation of target population. Practical implications may be suggested based in the results of this study. For instance, increase of availability and diversity of places for PA practice to the low-income population could help this group to overcome the most frequent barriers cited. For those people with high income, PA programs should include group activities and emphasize the enjoyment to address the barriers reported in this group. There is a need for future evaluations of barriers to PA using instruments that include multidimensional factors and also those aspects linked to the economic background of the population.
13. 14. 15. 16.
17.
18. 19. 20.
References 1. IBGE. Instituto Brasileiro de Geografia e Estatística. Pesquisa sobre expectativa de vida. http://ww.ibge.gov. br. Accessed June 2007. 2. Peixoto SV, Giatti L, Afradique ME, Lima-Costa MF. Cost of public hospitalization among elderly in Brazil’s Unified Health System. Epidem e Serv Saúde. 2004;13:239–246. 3. Nelson ME, Rejeski WJ, Blair SN, et al. Physical activity and public health in older adults: recommendation from the American College of Sports Medicine and the
21.
22.
131
American Heart Association. Med Sci Sports Exerc. 2007;39:1435–1445. Hollmann W, Struder HK, Tagarakis CV, King G. Physical activity and the elderly. Eur J Cardiovasc Prev Rehabil. 2007;14:730–739. Haskell WL, Lee IM, Pate RR, et al. Physical activity and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart Association. Circulation. 2007;116:1081–1093. CDC. Prevalence of regular physical activity among adults—United States, 2001 and 2005. MMWR Morb Mortal Wkly Rep. 2007;56:1209–1212. VIGITEL BRASIL 2008. Vigilância de fatores de risco e proteção para doenças crônicas por inquérito telefônico. http://bvsms.saude.gov.br Accessed May 2009. Sallis JF, Owen N. Physical activity & behavioral medicine. California: SAGE Publications. Behavioral Medicine and Health Psychology Series; 1999. Booth ML, Bauman A, Owen N, Gore CJ. Physical activity preferences, preferred sources of assistance, and perceived barriers to increased activity among physically inactive Australians. Prev Med. 1997;26:131–137. Reichert FF, Barros AJ, Domingues MR, Hallal PC. The role of perceived personal barriers to engagement in leisure-time physical activity. Am J Public Health. 2007;97:515–519. Schutzer KA, Graves BS. Barriers and motivations to exercise in older adults. Prev Med. 2004;39:1056–1061. IBGE. Contagem da população 2007. Rio de Janeiro, RJ: Instituto Brasileiro de Geografia e Estatística; 2008. Benedetti TR, Petroski EL, Gonçalves LH. Perfil do idoso do município de Florianópolis, SC: relatório final da pesquisa. Florianópolis – SC: NUCIDH/UFSC; 2004. ABEP. Associação Brasileira de Empresas de Pesquisa. Critério de Classificação Econômica do Brasil. http:// www.abep.org. Accessed June 2007. Stewart DW, Shamdasani PN. Focus groups: theory and practice. California: Sage; 1990. Champion VL, Skinner CS. The health belief model In: Glanz K, Rimer BK, Viswanath K, editors. Health behavior and health education: theory, research, and practice. San Francisco; 2008 pgs. 45–66. Brownson RC, Baker EA, Housemann RA, Brennan LK, Bacak SJ. Environmental and policy determinants of physical activity in the United States. Am J Public Health. 2001;91:1995–2003. Lees FD, Clarkr PG, Nigg CR, Newman P. Barriers to exercise behavior among older adults: a focus-group study. J Aging Phys Act. 2005;13:23–33. Fisher KJ, Li F, Michael Y, Cleveland M. Neighborhoodlevel influences on physical activity among older adults: a multilevel analysis. J Aging Phys Act. 2004;12:45–63. Kamphuis CB, van Lenthe FJ, Giskes K, Brug J, Mackenbach JP. Perceived environmental determinants of physical activity and fruit and vegetable consumption among high and low socioeconomic groups in the Netherlands. Health Place. 2007;13:493–503. Dawson J, Hillsdon M, Boller I, Foster C. Perceived barriers to walking in the neighborhood environment: a survey of middle-aged and older adults. J Aging Phys Act. 2007;15:318–335. Parks SE, Housemann RA, Brownson RC. Differential correlates of physical activity in urban and rural adults of various socioeconomic backgrounds in the United States. J Epidemiol Community Health. 2003;57:29–35.
132
Cassou et al
23. Galea MN, Bray M, Sr. GKA. Barriers and facilitators for walking in individuals with intermittent claudication. J Aging Phys Act. 2008;16:69–84. 24. Rasinaho M, Hirvensalo M, Leinonen R, Lintunen T, Rantanen T. Motives for and barriers to physical activity among older adults with mobility limitations. J Aging Phys Act. 2007;15:90–102. 25. Eyler AA, Brownson RC, Donatelle RJ, King AC, Brown D, Sallis JF. Physical activity social support and middleand older-aged minority women: results from a US survey. Soc Sci Med. 1999;49:781–789. 26. Lim K, Taylor L. Factors associated with physical activity among older people—a population-based study. Prev Med. 2005;40:33–40. 27. Crombie IK, Irvine L, Williams B, et al. Why older people do not participate in leisure time physical activity: a survey of activity levels, beliefs and deterrents. Age Ageing. 2004;33:287–292.
28. Cutler DM, Glaeser EL, Shapiro JM. Why have Americans become more obese? J Econ Perspect. 2003;17:93–118. 29. Pinheiro L, Fontoura NO, Querino AC, Bonetti A, Rosa W. Retrato das desigualdades de gênero e raça. http://www. ipea.gov.br/default.jsp. Accessed December 2008. 30. Burton LC, Shapiro S, German PS. Determinants of physical activity initiation and maintenance among communitydwelling older persons. Prev Med. 1999;29:422–430. 31. Baker LA, Silverstein M. Preventive health behaviors among grandmothers raising grandchildren. J Gerontol B Psychol Sci Soc Sci. 2008;63:S304–S311. 32. Kaplan MS, Newsom JT, McFarland BH, Lu L. Demographic and psychosocial correlates of physical activity in late life. Am J Prev Med. 2001;21:306–312. 33. Der Ananian C, Wilcox S, Watkins K, Saunders R, Evans AE. Factors associated with exercise participation in adults with arthritis. J Aging Phys Act. 2008;16:125–143.
