Older American Indians’ Perspectives on Health, Arthritis, and Physical Activity: Implications for Adapting Evidence-Based Interventions, Oregon, 2013

Introduction Despite the high prevalence of arthritis and physical disability among older American Indians, few evidence-based interventions that improve arthritis self-management via physical activity have been adapted for use in this population. The purpose of this study was to identify beliefs about health, arthritis, and physical activity among older American Indians living in a rural area in Oregon to help select and adapt an arthritis self-management program. Methods In partnership with a tribal health program, we conducted surveys, a focus group, and individual interviews with older American Indians with arthritis. Our sample comprised 6 focus group participants and 18 interviewees. The 24 participants were aged 48 to 82 years, of whom 67% were women. Forms B and C of the Multidimensional Health Locus of Control (MHLC) instrument, modified for arthritis, measured MHLC. Results The concepts of health, arthritis, and physical activity overlapped in that health was a holistic concept informed by cultural teachings that included living a healthy lifestyle, socializing, and being functionally independent. Arthritis inhibited health and healthy behaviors. Participants identified barriers such as unreliable transportation and recruiting challenges that would make existing interventions challenging to implement in this setting. The Doctor subscale had the highest MHLC (mean = 4.4 [standard deviation (SD), 1.0]), followed by the Internal subscale (3.9 [SD, 1.4]) and the Other People subscale (2.8 [SD, 1.1]). Conclusions Existing evidence-based programs for arthritis should be adapted to address implementation factors, such as access to transportation, and incorporate cultural values that emphasize holistic wellness and social interconnectedness. Culturally sensitive programs that build on indigenous values and practices to promote active coping strategies for older American Indians with arthritis are needed.


Introduction
American Indians have the highest prevalence of arthritis of any population in the United States (1). Although physical activity is considered a central component of nonpharmacologic arthritis management (2), American Indians are less likely to engage in physical activity than are people from other racial/ethnic groups (3). The Centers for Disease Control and Prevention (CDC) publishes a list of recommended evidence-based physical activity programs for arthritis (4). However, to our knowledge, no studies have described successful adaptations of these programs for use among American Indians, or in rural settings, where as many as 48% of older American Indians reside (5). These statistics illustrate the need for culturally tailored physical activity programs to manage arthritis for American Indians. For evidence-based programs to be effective, adaptations may be needed to ensure that programs align with the cultural values of the new target population and context of the new setting (6). Understanding lay health beliefs can provide insight into motivations that underlie individuals' health behaviors and coping strategies and inform relevant adaptations to existing programs (7). The purpose of this study was to identify beliefs about health, arthritis, and physical activity among older American Indians to help select and adapt an arthritis self-management and physical activity program for a tribe in rural Oregon.

Methods
In 2013, we collaborated with a federally recognized tribe's health program in rural Oregon to interview older, tribal-dwelling adults about their meaning of health, experience of arthritis, and interest in a group-based physical activity program. The study protocol was approved by the Oregon State University institutional review board and the tribe's Health Board and Tribal Council. Partners from the tribal health clinic and the study team recruited participants by word-of-mouth and by distributing flyers at local events and facilities. We used purposive sampling to target older adults with arthritis and ensure that men were sufficiently represented. Enrolled tribal members who had self-reported arthritis and identified as older adults were eligible to participate. The tribe did not want to exclude people who might benefit from a program for older adults, so people who identified as older adults self-selected into the study. Previous studies indicate that reservation-dwelling American Indians aged 45 years have health status similar to that of non-American Indians aged 65 years (8).
We scheduled several 90-minute focus groups at various times during February through September 2013. Only 6 people attended the first focus group, and subsequent groups were canceled because of insufficient registration. Therefore, we offered community members the option of participating in 30-to 45-minute indepth face-to-face interviews. Our final sample of 24 participants included 6 focus group attendees and 18 interviewees. The focus group and interviews were audio-recorded and transcribed verbatim. Participants completed a self-administered survey and received $50 for their participation.
The self-administered survey collected data on age, sex, education level, and marital status. Health was measured with a self-reported 6-item response (range, 1 = very poor, 2 = poor, 3 = fair, 4 = good, 5 = very good, 6 = excellent) (9). Participants ranked arthritis interference with daily activities in the past 6 months on a 10point scale where 0 indicated no interference and 10 indicated unable to carry on any activities (10). Locus of control, or the extent to which people believe their health is within their control, due to chance, or in the hands of others, is predictive of health behaviors (11), so we included 2 forms of the Multidimensional Health Locus of Control (MHLC) instrument in the questionnaire (12). We used the MHLC Form B to assess locus of control associated with general health and Form C to assess locus of control associated with arthritis. Form B, previously validated among American Indians (13,14), assesses general health locus of control via three 6-item subscales (Internal, Powerful Others, and Chance). The condition-specific Form C, modified for arthritis, subdivides the Powerful Others scale into two 3-item subscales: Doctors and Other People.
In collaboration with tribal partners, we developed a semistructured guide adapted for both the focus group and interviews. The open-ended questions reflected 4 domains: the participants' definition of health, effect of arthritis on peoples' lives, barriers to physical activities, and preferences for a physical activity program.
For the qualitative data, we used inductive content analysis to analyze the focus group and in-depth interview data (15). Initially, the 3 researchers separately reviewed and coded a transcript to develop initial codes. The initial codes were reviewed and consolidated to produce a codebook that the researchers used to code the remaining transcripts. Each transcript was independently and iteratively coded by at least 2 of the researchers, with new codes added as needed. We grouped similar codes into domains on the basis of the interview guide and created new categories as needed, then examined the codes for overarching themes. The researchers met throughout the process to discuss coding, resolve disagreements, and identify exemplar quotes for major themes. We conducted a subgroup analysis of the focus group to identify differences between the modes of data collection. NVivo qualitative software (QSR International) was used to organize the data.

Results
The full sample consisted of 16 women and 8 men, and the median age was 65 years (range, 48-82 y) ( Table 2). Approximately 71% of participants had some college education, and 33% were married. Mean self-rated health was 3.5 (SD, 1.1), and arthritis interference was 5.2 (SD, 2.5). For the MHLC Form B, the Internal scale ranked highest at 4.4 (SD, 1.2), followed by Powerful Others at 3.5 (SD, 1.1) and Chance at 2.7 (SD, 1.0) ( Table 1). For the MHLC Arthritis-Specific Form C, the Doctor scale had the highest mean score (4.4; SD, 1.0), followed by Internal (3.9; SD, 1.4) then Other People (2.8; SD, 1.1). We did not find significant differences between focus group and in-depth interview participants on any measures.

Qualitative findings
The qualitative findings and representative quotes are organized by the domains outlined in the interview guide (Table 3). Major themes were generally consistent across focus groups and interviews; however, focus group participants reported being more physically active and more thoroughly discussed barriers to accessing medical care.
Participants attributed their definitions of health to cultural values taught to them via their upbringing, traditions, and oral teachings from elders and family members. Health was a holistic concept that involved the mind, body, spirit, and caring for others. Living a healthy lifestyle that emphasized all these dimensions was considered central to health. Most participants described their upbringing as "healthy" in that they ate native foods (eg, salmon, roots, and berries) and participated in cultural activities that were physically active (eg, collecting roots and berries, fishing, hiking, preparing sweat houses, and riding horses). Several participants were concerned that there was a shift away from community events that promoted social engagement, creating fewer opportunities for the older adults to engage with youth and to pass down teachings about health.
All participants included functional ability in their definition of health. Functional ability was described as "being active." For example, one participant said, "being able to move every day, able to do something." Similarly, health involved independence and the ability to care for oneself, which enabled individuals to care for others. Illness and pain were seen as barriers to being active and healthy. Despite drawing the connection between healthy lifestyles as a way to facilitate health, participants struggled to maintain healthy behaviors and attributed these difficulties to changes in their local environment, in which availability of fast foods has increased and opportunities for physical activity have decreased.
The effect of arthritis on participants' lives ranged from minimal and sporadic to overpowering and constant. Arthritis pain affected participants' ability to complete activities of daily living, engage with family, take part in leisure activities, and get around. Many participants reported barriers to accessing medical care for their arthritis because of lack of local specialists and high staff turnover at the local clinic. Additionally, focus group participants discussed that clinicians did not understand or care about their pain or treat them respectfully. Several participants were unsure what arthritis was and confused arthritis pain and its treatment with their other conditions, such as spinal injuries, shingles, neuropathy, diabetes, and asthma.
Mechanisms for coping with arthritis pain were active and passive strategies that included using prescription and over-the-counter medications, nonpharmaceutical approaches, and physical movement. For some, medications were helpful in reducing pain but others experienced only minimal relief. Although over half of participants reported taking medications for arthritis, many preferred nonmedical interventions for arthritis and a few completely avoided clinical care. Nonpharmaceutical pain management included using heat or ice and rest, and traditional tribal remedies, including making teas from local plants and using sweat lodges and mineral baths. Many participants reported using mental strategies such as distraction and ignoring the pain.
Participants reported that "moving" was beneficial to reducing arthritis pain, but when probed about types of movements, they answered using generalized statements such as "I move around so I won't get stiff;" or "I stretch . . . my legs and arms." Few participants noted that physical activity, though initially painful, helped with their arthritis symptoms. Although almost all of the focus group participants engaged in some physical activity, no interview participants engaged in physical activity on a regular basis. Overall, the most common type of physical activity was walking.
Barriers to physical activity included the following themes: health status and pain, environmental and community factors, and lack of motivation. Many participants reported that pain caused by arthritis, chronic conditions, or injuries interfered with their ability to be active. Environmental and community barriers included lack of transportation, unsafe walking locations, and the cost of joining an exercise facility. Personal motivation was identified as both a barrier and a facilitator to physical activity. Some participants discussed how by motivating themselves, they were able to overcome arthritis pain to be active. They cited previous examples of using self-motivation to do healthy behaviors -including walking and losing weight -but described motivation as dwindling or insufficient to maintain these behaviors. Consequently, more than half of the sample identified lack of motivation as a barrier to currently being physically active. Those who wanted to be active reported being motivated by family, wanting to be healthy, and wanting to avoid negative outcomes that they had witnessed family members and friends experience.
Many participants had questions about arthritis, its causes and treatments, and about safe ways to exercise. Despite seeking information, participants voiced mixed interest in participating in an ongoing educational and physical activity program. About half of the sample was interested in a group exercise class, but some said they would not participate at all. Those interested in a program raised concerns that it would be difficult to find a suitable location, the classes might not be well-attended, and reliable transportation would be an issue. Given the transportation difficulties, a twice-or thrice-a-week class was not preferred. Participants suggested that engaging younger adults with families and youth in a health and PREVENTING CHRONIC DISEASE www.cdc.gov/pcd/issues/2016/16_0098.htm • Centers for Disease Control and Prevention physical activity program would be beneficial for the community. Other suggestions included providing transportation and incentives to participate, and recruiting through word-of-mouth, through locals or the leader providing personal outreach and encouragement, and by sharing success stories of local people who had improved through a similar program.

Discussion
Our study examined older American Indians' perceptions of health, arthritis, and physical activity to aid in identifying or adapting a suitable evidence-based program for a rural tribal community. Arthritis posed barriers to health and healthy behaviors, including physical activity. Meanings of health were couched in participants' native values, upbringing, and traditions. Participants discussed the importance of being functionally independent in relationship to caring for themselves and others and to participating in cultural activities. This finding corroborates those of other studies that found that rural lay definitions of health are highly contextual, informed by rural lifestyles and local traditions (16). Similar to other American Indian populations, our sample lacked knowledge about arthritis and physical activity (17). Yet, participants were aware that movement was important for alleviating stiffness and immobility. Furthermore, there was a clear interest in learning about nonpharmaceutical coping strategies from both physically active and inactive participants. Together, these findings suggest that a culturally relevant physical activity program is desirable in this community. However, participants also identified critical barriers that should be addressed before such a program could be successfully implemented in their community.
Our findings suggest that only a few of the CDC-recommended physical activity programs for arthritis might be adaptable for this community. From an implementation perspective, most of the CDC-recommended programs require special facilities (eg, fitness facility or pool) and license fees that would be prohibitive in rural communities (see reference no. 4 for a detailed description of CDC-approved physical activity programs for arthritis). All of the programs are group-based and require regular meetings over an extended period that may pose challenges given the transportation issues identified. A possible adaptation to improve implementation could involve a peer-to-peer program conducted in-home or via telephone instead of in a group setting. Telephone-based coaching programs have shown promising results in promoting healthy behaviors (18). Furthermore, peer education programs are as effective as professionally delivered programs and are a viable option in communities where resources are limited (19). They also provide added benefits of localized knowledge and social support that that can support the health of rural older American Indians (20). Although several peer-led evidence-based arthritis programs exist (4), none are designed for in-home delivery or have been tested for use in rural settings.
In addition to the feasibility issues, existing programs should be adapted to ensure that messages and activities align with the values, contexts, and traditions unique to this tribe. Furthermore, any adaptations must align with the programs' theoretical basis. Theory-based programs are not only more effective than those without (21), but also provide a framework for explaining the link between program components and resulting behavioral changes. Adaptations that significantly drift from the program's theory-based components may result in decreased effectiveness or other undesirable outcomes (22). Attitudes, beliefs, and values, for example, are culturally informed constructs that are common to many behavioral change theories (23). Understanding the target populations' beliefs and values about the program topic has implications for program adaptations. In our study, for example, the value that health is holistic indicates that a program focused solely on physical activity and arthritis might be too limited, and information about other aspects of healthy lifestyles, including nutrition, social relationships, and spiritual well-being, should be included to maximize relevance.
The participants in this study also emphasized the importance of health as a mechanism for caring for family members and for participating in social and cultural activities. They reflected on the teachings about health from their elders, and discussed their desire to engage with the youth in their community. This emphasis on the importance of social relationships has been reflected among other minority groups; other researchers suggest that common Western-based theoretical constructs, such as self-efficacy, that form the theoretical basis for interventions do not sufficiently represent the importance of relationships among minority groups in influencing health behavior (24). Interventions whose components draw from Western-based constructs may not be easily adapted to minority populations. Instead, interventions that attend to local contexts and values must be developed and tested to improve effectiveness, relevance, and acceptability.
Another construct in need of further attention is that of the health locus of control. Overall, participants had high Internal locus of control on the general health scale, a finding that has been linked to healthy lifestyles including regular exercise (25). On further examination, however, separating out Doctors from Powerful Others via the arthritis-specific scale evidenced that doctors are attributed a higher level of control, a finding that has elsewhere been associated with poor health (26). Given the desire of our participants to avoid medical treatments and medications for arthritis, this finding is potentially confusing. One reason for this disconnect may lie PREVENTING CHRONIC DISEASE in the qualitative findings that participants have difficulties accessing local and culturally sensitive health providers. O'Hea et al hypothesized that restricted access to health information could cause individuals to view their control of health as directly related to their ability to access health professionals (27). That is, people living in rural settings where transportation and access to health information is difficult may perceive that they are unable to take active control of their health given the difficulties in accessing resources. Developing and increasing accessibility to culturally appropriate information about arthritis self-management strategies has the potential to empower individuals to better manage their health.
Limitations to this study were the small sample size and the selfselection of participants, which should restrict generalizations of our findings to other American Indian groups. We had difficulty recruiting participants to focus groups, probably because of the transportation barriers identified by participants. Another limitation is that social desirability bias, especially in the focus group, may have affected responses by inflating reports of physical activity. Finally, the Cronbach α for the Other People subscale for Arthritis-Specific Form C was notably low, which suggests it had either poor internal consistency or too few items (only 3). Although the MHLC has been validated among American Indians, there are 566 distinct federally recognized tribes (28), each with its own values and traditions. Future studies are needed to further validate Form C among American Indians and to more thoroughly examine the applicability and implications of the health locus of control construct across tribes.
A tailored arthritis program that is relevant to local and cultural lifestyles, values, and beliefs of older, rural American Indians is needed to empower individuals to effectively manage their arthritis symptoms and reduce reliance on passive coping strategies. Adapting programs to improve implementation aspects for rural settings, such as delivering telephone or in-home programs, and incorporating culturally based values, such as integrating other aspects of health that acknowledge the holistic and social dimensions of health, are 2 potential considerations. A low-cost intervention strategy for older adults with arthritis that provides benefits of improved physical functioning and decreasing disability will be a significant contribution to mitigating the effect of arthritis among older American Indians.    Health is the ability to move, to be active, and to take care of self and others • "(Health is) where you can be more active at cooking . . . taking care of yourself. And take care of children and able to work." (Female participant, I5) • "(Health is) to be able to get up every day, get out of bed, feed myself, and dress myself." (Female participant, I11) Health is freedom from pain and illness • "(Health is) getting up without having to take . . . the 8 pills I take and not wait the hour for it to kick in." (Male participant, I12) The ability to live a healthy lifestyle has changed over time and affects the community • "Sometimes (moving) will hurt, but I notice it feels good afterwards. I won't be stiff and I'll get around better." (Female participant, I1) Access to medical care is challenging because of lack of local doctors, and the perception that Western doctors are not responsive • "We just don't have an arthritis doctor for us here on the reservation." (Female participant, FG6) • "(Clinicians) don't really take it to hear that you're suffering. . . . That's part of the reason why people don't go to the clinic. . . . I think it's natural for Native Americans to go back to old ways without any doctors and attention or somebody to help you out." (Male participant, FG1)  Lack motivation or confidence that physical activity will help • "I guess I just don't want to be really active." (Female participant, I4) • "Physical exertion and stretches. . . . I told myself they're not helping. I feel like they're making it worse." (Male participant, I13)