Listening to Indigenous Elders: An Evaluation of the Cultural Health Stations in Taiwan

Abstract

The Council of Indigenous Peoples in Taiwan has established tribal-based Cultural Health Stations offering basic health services and social activities for Indigenous communities. However, little is known about the factors influencing participation. This pilot study examined the reasons for attendance among older adults from five Paiwan communities in eastern Taiwan. A structured questionnaire, informed by Andersen’s Behavioral Model of Health Services Use and shaped through community consensus, was administered to 211 elders, 77 of whom were attending Cultural Health Stations. Participants who attended were generally older, female, living alone, not engaged in paid work or farming, and reported positive perceptions of the activities. Common reasons for non-attendance included time constraints, preference for alternative social settings, and dissatisfaction with the programs. Multivariable logistic regression analysis revealed that being aged 75 years or older, living alone, and reporting a higher self-rated ability to manage health issues were positively associated with attendance. In contrast, having paid employment was negatively associated. Findings from the open-ended questions suggested attendees appreciated services and activities but requested more cultural content. Non-attendees expressed that activities should be inclusive of different age groups and ability levels, culturally connected to local communities, and mindful that many elders still need to work. These findings highlight the importance of incorporating elders’ perspectives in the planning and evaluation of community-based services and may inform policy improvements for Indigenous elder care in Taiwan. However, it is also necessary to consider the economic circumstances of elders, design culturally centered activities, and respond to their interest in courses related to cultural transmission to ensure services are both accessible and meaningful.

What We Already Know

• More Indigenous elders in Taiwan are living into old age, and they experience significant health and economic disparities compared to their urban, non-indigenous counterparts. Therefore, it is essential to develop culturally appropriate long-term care services for Indigenous elders in rural regions.

• Community-based Cultural Health Stations are part of Taiwan’s long-term care policy aimed at achieving aging in place for Indigenous elders. However, Indigenous elders and their viewpoints have been excluded from the evaluation of these stations.

• Evaluations of long-term care policies and programs should incorporate the authentic opinions and voices of Indigenous elders to guide policy direction to prioritize culturally grounded, value-driven services.

What This Article Adds

• This study is the first to evaluate the community-based Cultural Health Station program in Paiwan communities using a culturally grounded assessment tool developed in collaboration with Indigenous elders and including both participating and non-participating elders’ opinions.

• Analysis of the demographic variables showed that being aged 75 years or older and living alone significantly increased the odds of attendance, while having paid work was negatively associated.

• Regardless of their participation in current daycare activities, elders appreciate the health check-ups and many of the activities. However, it is necessary to consider economic factors for the elders, design culturally centered activities, and address their interest in courses related to cultural transmission.

Introduction

Taiwan’s strategic position in East Asia has made it a contested territory, leading to over 400 years of colonization experienced by the Indigenous Peoples of the island. The earliest known migrants were Han Chinese from coastal China. Han Chinese people now account for over 95% of the population, contributing to long-standing power imbalances in land ownership and political representation. The current regime, the Republic of China (ROC), was founded by the Kuomintang (KMT) in 1911. After Japan’s defeat in World War II, the KMT took control of Taiwan in 1945 and, following its loss in the Chinese Civil War, retreated to the island in 1949. Under martial law, the KMT established an authoritarian regime that continued the colonial legacy of previous rulers, and its legacy continues to affect Indigenous rights and wellbeing nowadays, despite democratization in the 1980s.

The Han Chinese–dominated ROC regime currently recognizes 16 distinct Indigenous groups in Taiwan, most of which have ancestral roots in rural and mountainous regions. The Paiwan, the second-largest Indigenous group and the focus of this study, primarily reside in the south. Since the 1950s, rapid industrialization and urbanization have prompted many young Indigenous people to migrate to urban areas in search of work, education, and social services. At the same time, elders have remained mainly in their home communities. This ongoing urban–rural imbalance has resulted in a near-equal distribution of the overall Indigenous population between urban and rural areas, but rural communities now face a disproportionately aging population. Jinfong Township, located in southeastern Taiwan and predominantly inhabited by ethnic Paiwan people, has one of the highest proportions of residents aged 65 and older among all township-level regions, based on actual residency rather than registered household data.

Despite ongoing aging, elderly and long-term care services remain unequally distributed, with significant gaps in Indigenous regions. To address this, Taiwan’s Council of Indigenous Peoples (CIP) initiated the development of community-based Cultural Health Stations (文化健康站). These stations have been shaped by key legislative milestones, including the Long-Term Care Services Act (2015) and the Indigenous Peoples’ Health Act, ¹ which emphasize community-based, culturally responsive elder care and support for aging in place. Funded by the National Long-term Care Fund since 2016, the program expanded from 121 stations in 2018 to approximately 519 stations by 2025, serving Indigenous communities across Taiwan. ²

Cultural Health Stations provide medical support and health-promoting activities. Staff monitor blood pressure and blood sugar, refer elders to health providers, and follow-up on missed appointments. Stations also offer weekday lunches, health education, networking, gardening, cultural and art programs, and spaces for religious sharing. To date, the CIP has demonstrated the policy success of these stations using audit-type reports that focus on station numbers, elder demographics, and participation rates. ³

Several qualitative studies have provided valuable insights into the experiences of caregivers and staff at Cultural Health Stations. One such study emphasized the need for future research to “incorporate quantitative methods to broadly and deeply investigate issues and needs.” National evaluations have also highlighted disparities in elder participation across communities and have called for the integration of utilization metrics into station operations. Together, these findings highlight a clear gap in the systematic and measurable evaluation of how effectively these services address community needs, underscoring the need for the development of appropriate quantitative assessment tools for Cultural Health Stations.^3-5

Drawing on international experiences, more Indigenous communities are seeking to design interventions that integrate and affirm Indigenous culture, such as land-based and language-oriented approaches rooted in Indigenous traditions. ⁶ Studies from various Indigenous communities around the world have demonstrated how meaningful partnerships can incorporate Indigenous culture, knowledge, values, and language into both interventions and evaluation frameworks.^7-10 However, such participatory evaluation methods have not been widely adopted in Taiwan. Therefore, this exploratory study aimed to identify the reasons why Paiwan elders choose to participate or not participate in Cultural Health Stations, using research methods and tools developed in collaboration with elders in Paiwan communities, to inform future policies and programs that are responsive to the needs and values of Indigenous communities.

Before presenting the data collection, it is important to clarify the theoretical rationale guiding this study. We drew upon Aday and Andersen’s ¹¹ Behavioral Model of Health Services Use, which conceptualizes health service utilization as shaped by predisposing, enabling, and need factors. This framework, widely applied in studies of vulnerable populations, was particularly suitable because Cultural Health Stations are free of charge, making participation more dependent on social, cultural, and individual determinants. Accordingly, we hypothesized that elders’ demographic and cultural characteristics, accessibility of resources, and perceived health needs would influence their participation in the program.

Methods

Study Design

The Cultural Health Stations in Jingfong Township were evaluated through a survey of Paiwan Indigenous elders to understand why some elders choose not to participate and to explore factors influencing participation among others. The first author, a local Paiwan community member fluent in English and Mandarin and proficient in Pinaiwanan (the Paiwan language), led recruitment and data collection. A Mandarin- and Pinaiwanan-speaking community liaison provided support as needed.

Ethical Considerations and Community Approval

Ethical approval for this study involving human participants was obtained from the Institutional Review Board (IRB) at the University of Hawaiʻi at Mānoa. Informed consent was obtained using forms provided in both English and Mandarin.

Before the study began, meetings were held with community leaders, including mamazangiljan (traditional tribal leaders), community officials, and religious leaders, to introduce the research. These sessions outlined the study’s purpose, anticipated findings, potential implications, plans of community dissemination, and details regarding participant compensation. The leaders gave their approval to proceed. In addition, formal approval was obtained from the Office of Indigenous Peoples’ Social Welfare, Department of Indigenous Peoples Affairs, Taitung County Government, which oversees the operation of the Cultural Health Stations within Taitung County.

Recruitment and Eligibility of Participants

The target population for this evaluation was Indigenous elders residing in Jingfong Township, which has an estimated population of 3659. Approximately 1004 (28%) are aged 55 and older. In Taiwan, older adulthood is defined as 55 years and older for Indigenous Peoples, compared to 65 years and older for non-Indigenous. Individuals in this age group are eligible to access the services provided by the Cultural Health Station. ¹²

We aimed to oversample elders that formerly or never attended the Cultural Health Station to make sure to identify ways to improve services. We hypothesized that 60% of those attending would agree or strongly agree that activities provided at the Cultural Health Station were enjoyable compared to only 40% of those not attending. This corresponds to a medium effect size (Cohen’s h = 0.41). A sample size of 206 (77 elders who attended vs 129 elders that did not) would have 80% power to detect this difference in proportions at a two-sided 5% level of significance. ¹³

The recruitment and survey administration were performed from September to November, 2022. Eligible participants were those who self-identified as Paiwan, were aged 55 or older, and resided in Jinfong Township. This included current and former participants of the Cultural Health Stations, as well as elders who had never participated in the program. Recruitment was conducted in person. The first author visited Cultural Health Stations to introduce the study and invite their participation. Meetings were held to discuss preferences for completing the survey—either individually or in groups—and whether translation support was required. The community liaison was available to provide additional language assistance when required.

Lists of former attendees were provided by Cultural Health Station staff, and home visits were made to reach them. For elders who had never participated, mamazangiljan and community officials helped identify eligible individuals, and the first author conducted outreach through visits. Surveys were either completed independently by participants or administered through guided, face-to-face interviews, based on the participants’ preference. Translation support was also available during interviews upon request. All participants were encouraged to ask questions if any items were unclear. Upon completion, participants received a supermarket gift voucher valued at 100 NTD (approximately 3 USD) as a token of appreciation.

Measures

The questionnaire was developed with reference to Andersen’s Behavioral Model of Health Services Use to ensure theoretical coherence and contextual relevance. Guided by the model, items were organized into three domains: predisposing factors (e.g., age, gender, cultural responsibilities, religious affiliation), enabling factors (e.g., living arrangements, family support, accessibility), and need factors (e.g., chronic illness, perceived care needs). To further reflect the realities of Cultural Health Stations in Indigenous communities, the design incorporated barriers reported by non-participating elders and feedback from service providers in a preceding study. ¹⁴ Demographic items included age, gender, living arrangements, employment and pension status, cultural responsibilities (e.g., farming, childcare for grandchildren), religious affiliation, and chronic illness conditions. This theoretically and contextually grounded structure provided a systematic rationale for examining hypothesized relationships and assessing determinants of Paiwan elders’ participation in Cultural Health Stations. The full questionnaire is available in Supplementary Material 1.

Participants were asked about their participation status: whether they were currently attending, had previously attended, or had never attended a Cultural Health Station. Based on this, they were directed to different sections of the questionnaire. Current participants completed section 1, which asked them to rate their agreement (1 = strongly disagree to 4 = strongly agree) with statements about their motivations for participation, such as having sufficient time, preserving cultural practices, and receiving regular health monitoring from station staff. They also responded to four open-ended questions about when they began attending, their favorite and least favorite activities, and their suggestions for improving services.

Participants who had never attended or had dropped out completed section 2, which included scaled items on reasons for non-participation (e.g., limited mobility, lack of time, irrelevance of services) and one open-ended question about what changes might make the stations more appealing to them. All participants proceeded to section 3, which included items on health-related behaviors and beliefs adapted from a validated tool by DiLorenzo et al. ⁹ Some items were further modified or excluded based on the tribal context and feedback from elders, following Elder and Kersten’s ¹⁰ emphasis on culturally relevant adaptation. Example items included “Sometimes I don’t take my medication the way my doctor wants me to” and “If I need care, my family provides it.” Responses were measured using a four-point Likert scale, ranging from 1 (strongly disagree) to 4 (strongly agree). ⁹

The original survey was developed in English and subsequently translated into Mandarin. A Paiwan elder oversaw the translation of questions into Pinaiwanan to ensure that the questions were both grammatically accurate and contextually appropriate for use in the survey. It was pretested with two eligible Paiwan elders—one current participant and one non-participant. Feedback led to revisions for clarity and conciseness of the item. A second round of pretesting with two additional elders confirmed that no further changes were needed. The final survey took approximately 20 minutes to complete in one-on-one interviews.

Data Analysis

To ensure confidentiality, each survey was assigned a unique identifier and de-identified before analysis. Responses were entered into Microsoft Excel and analyzed using R version 4.5.0 (R Core Team, 2025) with the following packages: tidyverse, psych, corrplot, performance, gtsummary, epitools, dplyr, and ltm.

Age was categorized into three groups: 55–64, 65–74, and 75 years and older. Variables such as living arrangement, employment status, social responsibility, religious affiliation, and chronic illness were recoded into binary formats (e.g., living alone vs not alone). Cases with missing values in the recoded demographic variables were excluded from the analysis. Variable selection procedures were considered; ¹⁵ however, all demographic variables were retained based on Andersen’s Behavioral Model of Health Services Use and consensus with the community, ensuring theoretical and contextual relevance. Multivariable logistic regression was performed to assess the association between attendance and demographic variables. Reference categories were: female, age <66, not living alone, not currently working with income, not currently working, not receiving pension, no family or community duties, no religious affiliation, and no chronic disease. Interactions were evaluated for variables showing strong associations with gender (employment with income, P = .012; pension, P < .001). Interactions between age group and employment variables were tested but excluded due to sparse cell counts that produced unstable estimates.

Items from sections 1 and 2, which explored reasons for attending or not attending the Cultural Health Stations, were aligned based on thematic equivalence (e.g., “geographically convenient” vs “geographically inconvenient”; see Table 3). To enhance clarity and reduce redundancy, the item “We can share our religious beliefs” was excluded from the analysis. For attendees, the responses “I can share traditional knowledge” and “I want to pass on Paiwan traditions” were combined into a single item: “I want to share and pass on traditional knowledge.” For non-attendees, “Staff don’t understand or reflect Paiwan culture” and “Activities don’t reflect Paiwan spirituality” were merged into: “Activities do not reflect Paiwan culture.” Responses across all three sections were dichotomized into two groups: agreement (strongly agree and agree) and disagreement (disagree and strongly disagree). The section for non-attendees had two additional questions: “I am unclear what the station offers” and “I don’t identify myself as an older person.” Frequencies were reported, and group differences were examined using the chi-square test.

Table 3.

Level of Agreement With Reasons Elders for Attending/Might Not Attend the Cultural Health Stations.

Level of agreement with reasons for elders attending the cultural health stations (N = 77)		Level of agreement with reasons for elders not attending the cultural health stations (N = 134)
Questions	Agree or strongly agree. N (%)	Questions	Agree or strongly agree. N (%)
1A. The location is convenient	76 (99)	2A. The location is too far or inconvenient	15 (11.2)
1B. I can discuss community affairs	75 (97)	2B. Prefer to discuss community issues elsewhere	124 (92.5)
1C. Staff help me find needed information	76 (99)	2C. I already know how to find health information	119 (88.8)
1D. I have a free and open schedule	75 (97)	2D. Busy with work, caregiving, etc	107 (79.9)
1E. I can see friends and family	74 (96)	2E. Don’t need to attend just to see friends	102 (76.1)
1F. Staff check my blood pressure/sugar	76 (99)	2F. Don’t need health checks like blood pressure/sugar	87 (64.9)
1G. Staff remind me to take medication	76 (99)	2G. Don’t need reminders for medications	85 (63.4)
1H. Staff understand my needs	76 (99)	2H. Staff don’t understand my needs	85 (63.4)
1I. I enjoy the activities	74 (96)	2I. I don’t enjoy the activities	70 (52.2)
1K. I make time to attend	76 (99)	2K. Station hours don’t fit my schedule	47 (35.1)
1L. I want to share and pass on traditional knowledge	76 (99)	2L. Activities do not reflect Paiwan culture	54 (40.3)
1M. I am physically able to attend	74 (96)	2M. I am physically unable to attend	27 (20.1)
1N. My family encourages me to attend	56 (73)	2N. My family discourages me from going	21 (15.7)
		2O. Unclear what the station offers	44 (32.8)
		2P. I don’t identify myself as an older person	30 (22.4)

Data from the open-ended questions were analyzed using thematic content analysis. Two independent researchers (the first and third authors, who are trained in qualitative methods) repeatedly read and coded the responses, consolidating similar answers into distinct thematic groups. Discrepancies were resolved through discussion, resulting in an intercoder agreement rate of over 85%. Excel software was used to manage and organize these data.

Results

Characteristics of Participants

Overall, 211 elders completed the survey; 77 elders attended the Cultural Health Stations, while 134 did not. The demographic characteristics of the participants are presented in Table 1. One missing value in the age group variable was identified. The sample included more females (68%) than males (32%). Approximately 43% of the sample was under the age of 65, while 36% were in the 65- to 75-year age group, and 21% were 75 years or older. The elders who participated mostly lived with other family members; those living alone comprised only 7% of the participants. Only 14% of the participants were retired with a pension, while approximately 61% reported currently working for wages. In addition, 83% of elders reported having unpaid social responsibilities, including farming (53%), caregiving (19%), and volunteering (12%). Most elders (88%) reported having religious beliefs, and about 83% reported having one or more chronic conditions.

Table 1.

Participant Characteristics by Attendance.

Variable	Overall N = 211 a	Attendance		P-value b
		No N = 134 a	Yes N = 77 a
Gender				.009
Female	144 (68%)	83 (62%)	61 (79%)
Male	67 (32%)	51 (38%)	16 (21%)
Age group (years)				<.001
<66	89 (42%)	67 (50%)	22 (29%)
66-75	76 (36%)	50 (37%)	26 (34%)
>75	45 (21%)	17 (13%)	28 (37%)
Missing values	1	0	1
Living alone				.010
No	188 (89%)	125 (93%)	63 (82%)
Yes	23 (11%)	9 (6.7%)	14 (18%)
Currently working with income				<.001
No	130 (62%)	61 (46%)	69 (90%)
Yes	81 (38%)	73 (54%)	8 (10%)
Currently working				<.001
No	109 (52%)	54 (40%)	55 (71%)
Yes	102 (48%)	80 (60%)	22 (29%)
Receiving pension				.006
No	182 (86%)	109 (81%)	73 (95%)
Yes	29 (14%)	25 (19%)	4 (5.2%)
Having family or community duties				.3
No	59 (28%)	41 (31%)	18 (23%)
Yes	152 (72%)	93 (69%)	59 (77%)
Having any religious beliefs				.12
No	23 (11%)	18 (13%)	5 (6.5%)
Yes	188 (89%)	116 (87%)	72 (94%)
Having any disease				.5
No	35 (17%)	24 (18%)	11 (14%)
Yes	176 (83%)	110 (82%)	66 (86%)

^aFrequency (%).

^bPearson’s chi-squared test.

Also shown in Table 1 are the relationships between attendance in Cultural Health Stations and demographic variables. The results show that the elders who attended were more likely to be female (P = .009), in the older age groups (P < .001), living alone (P = .01), not working for wages (P < .001) or working at all (P < .001), and not receiving a pension (P = .006).

Multivariable Logistic Analysis

Findings from the multivariable logistic regression examining the relationship between demographic variables and attendance are presented in Table 2. One case with missing age group data was excluded, resulting in a final sample size of 210. After adjusting for other variables, being aged 75 or older and living alone were associated with higher odds of attendance, while currently having paid work was associated with lower odds of attendance. There was no evidence of interaction between gender and either employment or pension status.

Table 2.

Predictors of Daycare Station Attendance (n = 210).

Variable	Crude OR [95% CI]	Adjusted OR [95% CI]	P-value
Female (ref)	—	—	—
Male	0.43 [0.22, 0.80]	0.55 [0.25, 1.18]	.1462
<66 years (ref)	—	—	—
66-75 years	1.58 [0.81, 3.13]	1.13 [0.51, 2.50]	.7672
>75 years	5.02 [2.35, 11.10]	3.10 [1.24, 8.03]	.0174
Not living alone (ref)	—	—	—
Living alone	3.09 [1.28, 7.78]	3.54 [1.23, 11.18]	.0232
Not working with income (ref)	—	—	—
Currently working with income	0.10 [0.04, 0.21]	0.07 [0.02, 0.27]	.00017
Not currently working (ref)	—	—	—
Currently working	0.27 [0.15, 0.49]	2.30 [0.81, 7.06]	.1285
Not receiving pension (ref)	—	—	—
Receiving pension	0.24 [0.07, 0.65]	0.45 [0.05, 2.42]	.3907
No family or community duties (ref)	—	—	—
Having family or community duties	1.45 [0.77, 2.80]	1.69 [0.78, 3.77]	.1909
No religious beliefs (ref)	—	—	—
Having any religious beliefs	2.23 [0.85, 7.00]	1.81 [0.56, 6.56]	.3369
No chronic disease (ref)	—	—	—
Having any disease	1.31 [0.62, 2.94]	0.87 [0.33, 2.37]	.7821
Interaction: male × currently working with income	—	1.42 [0.22, 8.29]	.6983
Interaction: male × receiving pension	—	0.97 [0.07, 13.37]	.9822

Reasons Elders Attend the Cultural Health Station and Suggestions for Improvement

Table 3 presents the level of agreement among elders regarding reasons for attending or not attending Cultural Health Stations. Notably, 96% to 99% of attendees agreed with all reasons except “my family wants me to attend.” They found the stations convenient, had time to attend, enjoyed meeting friends and activities, and felt that the stations facilitated cultural transmission and discussions on community affairs. They also believed staff understood their needs, monitored their health, and helped them find information.

Responses to open-ended questions from attendees are detailed in Supplementary Material 2. Of 76 attendees who responded, 52.2% enjoyed cultural activities the most, while 33.7% preferred exercise, and 14% liked outdoor activities. Around 62% did not identify any least favorite activities, 13% expressed disinterest in drawing, and 12% disliked indoor activities. Less than 10% disliked handicrafts, exercise, and singing. When asked about additional activities, 44% had no suggestions, while 22% wanted more outdoor activities and 16% desired more cultural classes.

Reasons Elders Do Not Attend the Cultural Health Station and Suggestions for Improvement

In response to personal reasons, approximately 22% of the sample agreed that they did not consider themselves an older person, 80% agreed that they were too busy with their work or caring for grandchildren, and 76% agreed that they could meet their friends without needing to come to the stations (Table 3).

Two questions about Cultural Health Station staff revealed that 64% of respondents felt workers did not understand their needs, and 40% believed that activities do not reflect Paiwan culture. Regarding the activities, 67% knew what was offered, but 52% expressed dislike for them. Almost all (93%) preferred discussing community affairs outside the stations. For health services, 89% knew how to find health information elsewhere, 63% did not require medication reminders, and 65% did not need regular blood pressure or blood sugar checks.

Fifty-six elders who did not attend Cultural Health Stations responded to open-ended questions, grouped into four areas (Supplementary Material 2). About 39.3% focused on activity design, suggesting that activities should consider different ages, connect with communities, and include technology lessons. Around 26.8% wanted more cultural classes, such as learning from older elders and practicing the Paiwan language. Stations should also recognize that many elders need to continue working, as 28.6% cited personal reasons such as work, farming, and caring for grandchildren. Finally, 5.4% noted that the stations lacked physical therapy and respite services for family caregivers and did not accommodate individuals who were bedridden.

Differences in Health Concerns Between Elders Who Do and Do Not Attend the Cultural Health Stations

Supplementary Material 3 shows the health-related behaviors of the elders who attended and those who did not attend the Cultural Health Stations. Although most respondents could manage their health care independently, elders who attended the Cultural Health Stations were more likely to seek medical attention when needed and were more likely to reach out to the Cultural Health Stations for health care assistance. However, elders who did not attend Cultural Health Stations were more likely to obtain prescription medicine easily.

Discussion

Quantitative data are often more persuasive to policymakers and can promote social justice agendas more effectively than qualitative data, particularly in Indigenous health contexts.^16,17 This study is the first quantitative investigation examining Cultural Health Stations in Paiwan communities, including both users and non-users in Jingfong Township (N = 211). Approximately one third of the participants attended the stations, while two thirds did not. Overall, attendees expressed satisfaction with the services and activities, but requested more cultural content, such as Paiwan language classes, cultural knowledge-sharing sessions led by elders, and activities related to community affairs.

Elders who attended the Cultural Health Stations were generally older, female, and more available during the day. In contrast, non-attendees were often younger, male, and occupied with work or caregiving responsibilities. After adjusting for other demographic variables, being aged 75 or older and living alone were significantly associated with higher odds of attendance, while having paid employment was associated with lower odds. These findings highlight the increased care needs of older and isolated elders and the exclusionary impact of financial responsibilities. This aligns with the findings of Yang et al, ¹⁸ who reported that advancing age and Indigenous status are associated with higher overall long-term care needs. In addition, the study by Tsai and Shih ¹⁹ indicated that the economic conditions of Indigenous elders in later life significantly influence their access to long-term care services.

Responses to the items in sections 1 and 2 suggest that elders exercised personal agency when deciding whether to attend the Cultural Health Stations. Their decisions were influenced by factors such as their own perceived needs, ability to manage daily schedules, and preferences for assistance, rather than family influence. Notably, there was a divergence between attendees and non-attendees regarding the perceived cultural and community functions of the stations. Attendees were more likely to view the stations as spaces for discussing community affairs, while non-attendees did not share this view. This highlights the need for further exploration of the stations’ roles in supporting cultural and social engagement, beyond their health promotion or health care functions.

The open-ended responses revealed elders’ desire for more cultural and language classes led by Paiwan elders, additional opportunities for cultural knowledge transmission, more activities related to tribal affairs, and increased outdoor activities. Non-attendees reported similar preferences. These findings align with McKinley ²⁰ who noted that American Indian elders preferred engaging with nature and cultural activities to maintain health. The lack of tribal affairs integration echoes Gomes et al, ²¹ emphasizing the importance of involving Indigenous elders in community affairs, and Oosman et al ²² who advocated for high community engagement and intergenerational approaches in Indigenous health promotion.

Our findings underscore the need for program design and evaluation frameworks at Cultural Health Stations to more genuinely reflect Indigenous community agency and values. As Chiu et al ²³ have pointed out, formal evaluation systems often diverge from the realities of care workers’ culturally embedded practices, highlighting the importance of flexible, culturally grounded approaches. Likewise, Lenglengman Rovaniyaw ²⁴ argued that evaluations may emphasize symbolic representations, such as clothing or crafts, while failing to consider whether these elements hold genuine cultural meaning for the elders. Such superficial, short-term assessments risk missing the depth of authentic cultural care.

These insights suggest that current elder care policies and service designs in Taiwan are not rooted in Indigenous self-determination or cultural worldviews. Rather than simply modifying existing mainstream services, there is a pressing need to recognize and support locally developed, strength-based elder care models that are aligned with Indigenous values and intervention priorities. ²⁵ This calls for more than culturally appropriate content, demanding the creation of culturally driven, evidence-based long-term care frameworks. Ultimately, elder care services should be reimagined through the lens of Indigenous elders’ social and cultural roles, blending professional medical perspectives with the lived values of the community to build truly localized and empowering care systems. ¹⁴

Limitations and Strengths

A potential limitation in data collection was the risk of misinterpretation due to translation between English, Mandarin, and Pinaiwanan. To mitigate this, a community elder with deep knowledge of Pinaiwanan and local cultural context was engaged to oversee the translation. The primary researcher, a member of the Paiwan community, conducted interviews prioritizing the use of Pinaiwanan to support clear and culturally appropriate communication. In addition, a community liaison fluent in Pinaiwanan assisted during interviews by providing real-time translation and clarification, helping ensure participants’ understanding and comfort throughout the process.

Another potential limitation of this study is that the questionnaire was not subjected to a formal validation process before its use. This decision was mainly due to the small population size targeted. Conducting a pretest could have resulted in participant overlap, thereby reducing the study’s statistical power. Moreover, the primary aim of this research was exploratory, ie, to identify potential factors that could inform future study designs and pathways for community collaboration. The intention of the questions related to why elders used or did not attend the Cultural Health Stations was to capture diverse voices and detailed perspectives rather than to generate a single composite score.

Relying on Cultural Health Stations employees and tribal leaders to identify respondents introduced bias, as the sample selection was not random. However, we also want to stress that, in contrast to general surveys conducted in a larger, more heterogeneous population, our population of interest demonstrated higher homogeneity in demographic attributes due to shared cultural and socioeconomic contexts, as other surveys conducted in Indigenous communities around the world have suggested. Therefore, we considered our selection of participants, although not entirely unbiased under dominant statistical perspectives, to sufficiently represent our population of interest.

Conclusion

The most significant feature and applicable aspect of this study is its evaluation of the implementation of current Cultural Health Stations on a township-wide scale from the perspective of elders who used and did not use the stations. To ensure the project’s effectiveness, it is crucial to engage in further discussions about future participants and their needs. Moreover, policy design and implementation should consider and address the feedback from elders regarding the operation of tribal-based Cultural Health Stations and make necessary adjustments.