Introduction

The phenomenon of population ageing is occurring in both high-income countries (HIC) as well as low and middle-income countries (LMIC) [1]. By 2050, estimates suggest that older persons will constitute 21.5% of the global population, doubling from 12.3% in 2015. In the near future, four out of every five older people globally will reside in LMIC [2]. By 2030, older persons in the African region will rise to 105 million and about 220 million by 2050 [2]. On the continent, East Africa ranks first among the regions with the highest population of older persons (18.9 million), followed by North Africa (17.9 million) and West Africa (16 million) [1].

Nigeria has the highest African population and ranks seventh globally [2]. In Nigeria, older persons aged 60 years and above constitute about 5.0% of its over 200 million population [2]. Older persons were estimated to number about 8.2 million in 2015 and will rise to 12 million by 2030. This number will increase further and reach about 25.2 million by 2050 [2].

The ongoing demographic transition will likely result in an increased prevalence of chronic diseases, functional decline, and increased reliance of older persons on others for their long term care (LTC) needs [3]. According to the World Health Organization (WHO), LTC is ‘the series of activities undertaken by others to ensure that people at risk of significant loss or with ongoing loss of intrinsic capacity can maintain a level of functionality consistent with their basic rights, fundamental freedom and human dignity’ [4]. In its meeting in 2016, the World Health Assembly resolved that all countries must make provisions for LTC for their older population. Although the 194 member states agreed with the recommendation, few have made provisions and measures to support the older population segment.

Long-term care provisions may be provided formally or informally within the home, community, or dedicated institutions [5, 6]. Common LTC settings include senior centres, assisted living facilities (ALF), and in-house services, referred to as ‘Ageing in Place’. Senior centres provide day-care and community interaction services in a non-residential setting. The ALF offer older persons needed support but not strictly in the health facility. Furthermore, these facilities provide individualised services to suit the preferences and needs of the client. Services provided include support, social, recreational and health-related services [7].

Ageing in place is defined as: ‘living at home or in the community with some level of independence, privacy, safety and control over one’s environment’ [8]. The concept aims to support older persons in their preferred accommodation and communities where they reside for as long as possible despite actual or potential changes in health and functioning in old age [9]. According to Rosenwohl and colleagues, the process of ageing in place may be explained by three important constructs, namely identity, connectedness, and sense of place. In addition, the experience of ageing is a balance between threats and agency. On the one hand, agency is the ability of older persons to make decisions and the capacity to respond to threats. Also, agency assists in maintaining important elements of their daily lives while ageing in place [10]. As such, agency depends on factors such as financial stability, social connections and available services. On the other hand, threats include changes in health functions and needing assistance with activities of daily living [10].

Ageing in place as a model is increasingly gaining recognition in many HIC, especially with the increased burden of chronic disease and disability due to longevity [11–13]. The decision to AIP benefits older persons’ health and overall quality of life (QoL) [14, 15]. Furthermore, accepting the approach in many HIC is based on its potential for reduced cost of care compared with institutionalised care [16, 17]. Existing research, mainly from HIC, shows that older persons prefer to AIP interminably regardless of their functional capacity [18].

Several socio-economic and demographic factors are documented to stimulate the expectations and decisions of an older person regarding the setting for LTC. A study conducted among older persons in Detroit, USA, established that living alone, home-ownership, and independent transportation mode through driving influenced expectations to AIP [19]. Likewise, Golant identified other factors, including the environmental characteristics and the availability of dedicated caregivers to provide appropriate care [20]. Additionally, Shiamberg and McKinney [21], documented factors that influenced the older person’s decision to AIP, including individual factors such as financial capacity, socio-economic characteristics, and healthcare needs [21]. Other factors that influenced decision-making in the study were contextual and included safety, economic independence, and availability of jobs and health care [21].

Another factor reported to influence the preferred setting for the care of older individuals is the availability of informal support. For instance, in a qualitative study among older persons in seven naturally occurring retirement communities, Greenfield noted that informal support from neighbours contributed to promoting ageing in place [22]. Likewise, Gardner reported that relationships with neighbours positively influenced the decision to AIP [23]. Furthermore, Ball and colleagues conducted a qualitative study investigating ALF [24]. The authors demonstrated that managing functional decline was the key to ageing in such facilities. Therefore, no matter where the older person is ageing, either in the home community or in a specially designated facility, provisions must be put in place to manage functional decline [24].

Ageing in place has been shown to include benefits to the older person, family, communities, and government. Benefits to the individual include health and well-being, promoting attachment and familiarity with the environment [25]. Research shows that the inability of older persons to remain in a familiar environment leads to loneliness, depression, difficulty adjusting and functional deterioration [26]. Furthermore, ageing in place has economic benefits as home care is far cheaper than residential care [11]. Other benefits of ageing in place include social inclusion, social connection, autonomy, and companionship [27].

However, the picture in many LMICs remains largely unchartered as few studies have established an accurate picture. Nevertheless, the ability to AIP seems to exist and is purported to be the preferred choice among individuals in low-income settings. However, there may be evidence that ageing in place in such environments may be more an arrangement of convenience rather than choice. For instance, in their study among low-income Chinese city dwellers, Lum and colleagues set out to establish whether individuals were ‘ageing in place’ or ‘stuck in place’ when the capacity to relocate was unavailable [28].

Furthermore, Golant pointed out the flaws of studies that suggest that older persons desire to AIP and criticised the failure of such studies to distinguish preferences among older persons with varying health needs and disabilities [20]. Similarly, Lehning, Smith, and Dunkle examined the association between age-friendly community characteristics and the desire to AIP [19]. Although the authors found no differences across income levels, it was unclear whether the increased expectations to AIP were an actual choice among low-income earners or a lack of any other affordable alternative.

Most research carried out to date on preferred care settings for older persons, including ageing in place has been conducted in HIC [13, 29–31]. The studies show that older persons prefer to AIP despite their functional capacity. There is a narrow focus on LMIC, such as in Nigeria, particularly emphasising the preference of older persons. Rural and urban communities in these areas have differences in structure, services, and amenities. Considering the poor policy and support systems for older persons in LMIC, there are concerns that older persons may choose to AIP not because they choose to but because they must. Also, vulnerable adults such as the poor, more senior, poorly educated, unmarried and females may not have any option and may be trapped.

Recent estimates suggest that many older persons in SSA require LTC at different times across their life span [4]. This need is bolstered by the WHO Global Strategy and Action Plan, which calls for all countries to face the challenges of LTC provisions for their ageing population [32]. Likewise, one of the action areas for the United Nations (UN) Decade of Healthy Ageing (2021–2030) is the provision of LTC for older persons [33]. This call for multidisciplinary efforts to achieve healthy ageing necessitates an enquiry into the preferences of the most critical stakeholders (older persons). This study compared, using quantitative measures, the preferred care setting for ageing and factors associated with the decision among rural and urban community-dwelling older persons aged 60 years and above in Oyo State, Nigeria. Also, the study explored the views about their choices using qualitative measures.

Materials and methods

This paper describes part of a more extensive study which compared the preferred setting for care, quality of life, attitude to ageing and disability among older persons in a rural and an urban community in Oyo State, south-western Nigeria. The first study highlighted the rural–urban differences in quality of life and associated factors among community-dwelling older persons in Oyo state, South-Western Nigeria [34]. The second study described the populations attitude towards ageing and perceived health status [35].

The study presented in this paper used a mixed research method which has been described previously [34, 35]. A mixed-methods approach was employed because of the exploratory nature of this study. The quantitative measures were to substantiate the preferred settings for care by the participants and associated factors. The qualitative aspect of the study explored the views regarding their preferred care setting. Triangulating quantitative and qualitative data collection techniques enabled a richer and deeper understanding of the research findings [36].

Data collection and analysis

Quantitative data.

Sample size calculation. The sample size for the quantitative study was determined using the sample size calculator in Stata Version 12 and the formula for comparing two independent proportions assuming a difference of 10%, power of 90%, and α of 5% [41, 42]. The minimum calculated sample size was 522. However, in total, 1,180 respondents (592 urban and 588 rural) were selected for the survey.

Sampling technique. A comparative cross-sectional survey was conducted in the purposively selected LGA. A multi-stage sampling procedure was utilised for the participant selection to select the LGA, enumeration areas (EA), ward and households. The purposive selection of the LGA was based on characteristics earlier highlighted (previous mapping and listing). In the LGA selected, the EA chosen included the ward of interest. The ward was also chosen because it has been listed and divided using a standard approach into areas. In each selected ward, four areas were selected through balloting without replacement. The available database from the Department of Community Medicine census was probed to identify households with the desired target population. The selected households were then visited. If an intended respondent was unavailable due to relocation or demise, the field workers were trained to move to the next listed household. Similar to previous studies, if a selected respondent was not met at home, the house would be visited thrice, after which the field workers would move on to the next house on the list [43, 44]. Also, if more than one eligible participant was in a particular household, balloting was carried out to select a participant.

Data collection. Quantitative data were collected over two months in July 2018 using a semi-structured questionnaire. Trained research assistants (RA) administered the questionnaires after informed consent had been obtained. The survey was deployed using the Research Electronic Data Capturing (REDCap) on mobile devices (tablets).

A total of 10 RA with a minimum educational attainment of at least a secondary school leaving certificate were trained in the proper approach to administering the questionnaires. The training was done over three days before a pre-test and subsequent commencement of fieldwork. Also, two community inhabitants domiciled in the respective LGA were recruited to guide the research team on the field. The field guides assisted with community entry and identification of selected households. The RA visited each selected household, and face-to-face interviews were conducted.

The developed instrument reflected the study objectives and consisted of seven sections including socio-demographic characteristics, family dynamics, current health status and selected health risk behaviours. Also, the instrument assessed the participant’s attitude to ageing, quality of life, self-reported functioning, and disability. Lastly, the preferred care setting was determined. Socio-demographic variables obtained included (age, sex, employment status, marital status, living arrangement, and socio-economic status). Respondents’ economic status was determined using the equity tool index based on the wealth indices generated from Demographic and Health Surveys (DHS) and other similar surveys. The surveys include questions about the respondent’s household characteristics and owned assets [45]. Wealth status was categorised into quintiles by relating each respondent’s total possessions to the median number of possessions of the sample. Participants’ attitude to ageing was assessed using the attitude to ageing questionnaire (AAQ) [46]. The instrument measured the attitude to ageing in psychosocial, physical change and psychological domains.

Participants were asked to rate their quality of life on a five-point Likert scale ranging from very satisfied to very bad. The responses were categorised as having good QoL (if their response to the question of Qol were “good” or “very good”). Likewise, respondents were said to have poor QoL if they reported these to be “bad”, “very bad”, or “neither bad nor good”. Participants were also asked to rate their health on a five-point Likert scale ranging from very good to very bad. The responses were categorised as having good self-rated health (SRH) (if their response were “good” or “very good”. Likewise, respondents were said to have poor SRH if they reported their health as “bad”, “very bad”, or “neither bad nor good”. The respondent’s level of functioning was measured using the 12-item World Health Organization Disability Assessment Scale (WHODAS), Version 2.0 [47]. The domains evaluated the level of disability among respondents in six domains. Other variables measured include living dynamics such as living arrangements, having assistance and home-ownership. Ownership of housing was categorised as owning the homes in which they reside or not. Participants were also asked if they had assistance at home, to which the response was Yes or No.

The research instrument was translated into the local Yoruba language for easy communication. It was back translated to English to ensure the original meanings were maintained before administration. A pre-test was conducted among the older persons in a different LGA from those selected for the study. Following the pre-test, ambiguous questions were rephrased or removed where necessary. Each of the interviews lasted between 30 minutes to 45 minutes.

Data analysis. The primary outcome variable for the study is the participants preferred care setting. The options included staying within their homes or community (AIP). Quantitative data were analysed using Stata version 14 [42] and presented using frequency tables and charts. Summarisation was done using proportion and means. Associations were measured between independent and dependent variables using the Student t-test, Chi-square test and logistic regression. Variables significant at 10% on bivariate analysis were included in the regression model and fitted for the rural, urban, and total population at a 5% level of statistical significance.

Qualitative data collection and analysis. The qualitative aspect of the study included the conduct of Focus Group Discussions (FGD). The FGD explored in depth the participants’ experiences and views on preferred care setting in their respective communities across the spectrum of functionality. Also, the dynamics surrounding their choices were investigated. Each focus group comprised eight to ten people (60 years and above) and the research team. The research team comprised of a moderator/ facilitator, an observer, and a recorder. Efforts were made to ensure that the groups were as homogenous as possible regarding socio-demographic characteristics such as gender (male, female), age stratification (60–69 years, 70 years and above) and location (rural, urban). The FGD participants were from the selected communities but did not necessarily partake in the questionnaire survey.

Participants were recruited purposively through the help of field scouts and relevant gatekeepers in the community, like the Chairman of the Ward Development Committee in the urban area and a notable and well-respected pastor in the rural community. The discussions in the urban area were held in a centrally located comprehensive primary health care centre. The rural discussions took place in the house of a community leader.

The author and two other assistants (one male and one female) who were experts in qualitative research facilitated the FGD. The three facilitators were bilingual and obtained informed consent from each participant before the discussion. The discussions were facilitated with a semi-structured interview guide with suitable probes to ensure uniformity in questions asked for each group. The sessions were recorded using a tape recorder. A designated note-taker took detailed notes, which aided transcription. The duration of each discussion was between 60 to 90 minutes. A debriefing among the team was held after each FGD to discuss emerging themes. Respondents were provided with refreshments and reimbursed for their travel-related expenses.

Data saturation was attained after 22 FGD had been conducted. The FGD were conducted in Yoruba, transcribed to English by the RA who assisted or did the facilitation and peer-reviewed among the facilitators to confirm translation accuracy. The principal investigator approached a subset of the participants for member checking. The aim was to verify the results obtained as a true reflection of the discussion. The research team also had frequent peer debriefings to ensure that the accounts of the discussion were accurate. Clarifications were sought and discussed extensively, potential biases were probed, and the audit trail was consulted as needed.

The research team included the first author and two research assistants with expertise in qualitative research who served as the note-taker and recorder/ observer. The three members of the team took turns facilitating the group discussions. The facilitator was responsible for the dynamics of the group. Prior to the commencement of the discussion, ground rules were discussed. These included allowing everyone the opportunity to talk one after the other, and there were no right or wrong answers. The facilitator for each discussion ensured compliance.

Qualitative data analysis. Data collected were transcribed and analysed using Atlas ti. Version 8 and the thematic framework approach [48]. The transcripts were reviewed several times to familiarise with the content before labelling the key words and phrases. The data analysis started with line-by-line open coding, breaking segments of the data into corresponding categories, such as preferred care setting, determinants of the choices, benefits of AIP and views about institutional care (Table 1). Apart form ageing in place within the home or community, other options revealed by the FGD included moving in with in-law, first-degree relative/child, church establishment or institutionalised care.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Older persons preferred setting for care and its determinants.

https://doi.org/10.1371/journal.pone.0292939.t001

Next, emergent themes that represented the preferred care settings and issues raised based on study objectives were sorted and coded. The themes were categorised and compared before making a decision on the dominant themes to report. In cases of conflicts, a discussion between the lead author and the research ensured a consensus before finalising the various categories for each theme. In the end, the generated themes were analysed. Relevant quotes were identified and used verbatim to buttress the explanation of the findings. Overall,172 older persons participated in the 22 FGD. The participants were mostly Yoruba with a mean age of 70.9 (±2.9) years.

Findings from both quantitative and qualitative data were integrated during the data-interpretation stage. The goal was to obtain different but complementary data that validated the overall results about the participant’s preferred care setting and associated factors.

Results

Quantitative findings

Overall, 1,180 respondents participated in the survey (592 in the urban area and 588 in the rural are). Participants were mostly female in the rural (406, 69.0%) and urban (417, 70.0%) communities. Rural participants compared to urban participants were older (74.2 ± 9.5 vs 72.3 ± 8.9 years, p = 0.001), homeowners (67.0% vs 61.6%, p = 0.06) had home assistance (79.8% vs 74.7%, p = 0.04) and disability (53.1% vs 50.7%, p = 0.41). More participants in the rural community desired to AIP 362 (61.6%) compared with those in the urban setting 232 (39.2%), (p<0.05). Among the participants in the rural community, older participants, 277 (69.4%), were more likely to desire to AIP. Likewise, a higher proportion of males, 125 (68.7%), were likely to want to AIP compared with females 237 (58.4%) (p<0.05). Also, more unmarried participants preferred to AIP 204 (66.5%) compared with those who were married 158 (56.2%). Likewise, a significantly higher proportion of participants who did not have formal education, 309 (64.9%), were more likely to desire ageing in place compared with those who had formal education 53 (47.3%). Among the participants in the urban community, the desire to AIP was associated with older age (≥70 years) and unemployment (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Selected demographic factors associated with the desire to age in place.

https://doi.org/10.1371/journal.pone.0292939.t002

As shown in Table 3, more rural participants who reported having assistance at home desired to AIP 301 (64.2%) compared with those who had no help at home 61 (51.3%), (p<0.05). Other factors associated with a desire to AIP among the rural participants were having owned homes, and good self-rated quality of life (SRQoL). Among the participants in the urban community, factors associated with the desire to AIP were living in an owned home and having good SRQoL (p<0.05).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Other factors associated with the desire to age in place.

https://doi.org/10.1371/journal.pone.0292939.t003

Table 4 shows details of the most significant factors associated with the desire to AIP for rural participants were older age (≥70 years) [OR: 2.07 (95% CI: 1.37–3.14)], being male [OR: 2.41 (95% CI: 1.53–3.81)] having home assistance [OR: 1.79 (95% CI: 1.15–2.79)] and living in an owned home [OR: 2.47 (95% CI: 1.69–3.61)]. However, compared with those who were currently unmarried, participants who were currently married were less likely to desire AIP [OR: 0.61(95% CI: 0.40–0.92). Also, compared with those with low educational attainment, participants who had higher education were less likely to desire to AIP [OR: 0.49 (95% CI: 0.30–0.81)]. However, the most significant factors associated with the desire to AIP for urban participants were older age (≥70 years) [OR: 1.54 (95% CI: 1.03–2.31)] and home-ownership [OR: 5.83 (95% CI: 3.82–8.91)].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. The most significant factors associated with the preference to age in place by location.

https://doi.org/10.1371/journal.pone.0292939.t004

Regardless of their desire to AIP or not, participants were asked why they may relocate from their current location. Their responses are shown in Fig 1. The most frequently reported factor that may influence the desire to relocate in the rural community included a desire to be close to family members 211 (35.9%), a need for a manageable size of the housing 187 (31.8%), the need for independence 147 (25.0%) and access to health services 114 (19.4%). The participants less frequently mentioned other reasons such as concerns for safety, house maintenance costs and transport difficulties. Similarly, in the urban area, the most reported reasons to consider relocation include a need for a manageable size of the housing 298 (50.3%), the desire to be close to the family 299 (50.5%) and to facilitate their independence 348 (29.5%).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Respondents’ reasons for considering relocation by location.

https://doi.org/10.1371/journal.pone.0292939.g001

Discussion

This study determined factors associated with the preferred care setting among community-dwelling older persons in a rural and urban setting in Oyo State, south-western Nigeria and explored their views about their choices. The quantitative aspect of the study revealed that older persons in both settings prefer to AIP. Factors associated with the decision include rural setting, older age (≥ 70 years), being male, home-ownership and having assistance at home. The FGD corroborated the choice of community-dwelling older persons to AIP within the home or community mostly based on the traditional expectations for reciprocity of care. Factors associated with this preference include functional status, social connectedness, and community participation. Additionally, participants were ambivalent regarding institutional care.

This study corroborates findings from other studies across income settings showing that older persons mainly desire to AIP [21, 49–51]. A possible reason for this finding may be the expected reciprocity of care in traditional settings in Nigeria and the high expectation that the family is responsible for providing LTC [38, 52, 53]. As reflected in the FGD, there is a high expectation that children will care for their parents when they grow old. However, the ongoing societal changes, such as the high level of rural-urban migration and increased female participation in the workforce, threaten this arrangement’s viability. According to the focus group discussants, the best alternative was to explore other arrangements that allowed older persons to stay in an environment which facilitated peer and family interaction.

The decision to AIP may result from a strong attachment of older persons to their homes and neighbourhood in such environments [19, 49, 54]. In this study, the most frequently reported reason influencing the respondent’s desire to relocate in both communities was a desire to be close to family members. This finding may be because of the younger population’s high rate of rural-urban migration in search of better economic opportunities [50]. Similar to previous research, another reason the participants gave for considering relocation was the need to remain independent which may include relocating to a smaller apartment [49]. Additionally, the participants in this study frequently reported the desire for better access to health services. This finding may result from the functional and health challenges experienced with the ageing process [55]. Nevertheless, as documented in this study, other less frequently reported reasons for the desire to relocate include concerns for safety, house maintenance costs and transport difficulties.

The study revealed that more rural participants preferred to AIP in their homes or communities than urban participants. This finding is expected as many individuals have strong ties with their villages [6, 56, 57]. Also, a higher proportion of rural dwellers had always lived in the same community compared to urban dwellers. In many instances, after migrating to the urban area to work and spend their productive years, many older persons return to their villages, where they prefer to die in the hands of their loved ones. Also, if they die outside their homes, they often leave advanced directives, which likely stipulate that their remains should be buried in their towns of origin [58, 59]. Another explanation for the finding may be the increased social participation and intergenerational living arrangement in the rural area compared with the urban setting. In the FGD, participants discussed how neighbours filled the care void when family members could not offer care and support. Participants also spoke about their involvement in many community activities, including constructing boreholes and church activities. Also, some mentioned that they also cared for their grandchildren as well.

Among both the rural and urban participants, factors shown to predict a desire to AIP were older age (≥70 years) and living in owned housing, similar to research by Lofqvist and colleagues [49]. This finding may be due to the decline in health due to old age. The individual’s health has been copiously documented to influence the decision to relocate or AIP [60]. Apart from familiarity within the space, home-ownership facilitates necessary structural modifications to ensure person-environment fit [61]. However, factors peculiar to the rural participants and associated with a desire to AIP were the male gender and having assistance at home. The sex difference is expected as older men are often better off financially than older females. This finding may be because often, men may have had better educational and employment opportunities. Again, older females, particularly widows, have been documented to be more financially vulnerable [3]. Also, many participants preferred to AIP mostly in their homes with family members on hand to provide the necessary assistance, care and support. The FGD corroborates that care within family homes is common in many African communities.

There were gender differences with preference of setting to AIP. The FGD revealed that many women preferred to AIP in their husband’s homes. However, participants accepted that the place to stay as they aged was flexible and could change. This acceptance was amplified if there were health issues and the individuals could not care for themselves and required a lot of assistance. Older females were more receptive to being taken to their children’s place in the event of functional decline or ill health. However, a frequently mentioned problem was the possibility of isolation.

Furthermore, older women in the study were more open to care by employed formal caregivers, especially if the children needed to be away at work. However, many desired to return to their marital homes after their death. Likewise, as reflected in the FGD, older females were hesitant to stay with their sons as they were careful not to disturb their daughters-in-law. This finding is not unusual, as daughters-in-law have been previously documented as perpetrators of abuse of older women [62]. Also, educated rural participants were less likely to desire to AIP, which contradicts another research [63]. A possible reason may be that poverty levels are high in the typical rural setting and further complicated by a poor level of development and infrastructure. Both male and female participants noted a preference to live within communities where they had social ties, friends, and relationships.

However, it is surprising that many participants were not opposed to the notion of care beyond the family. Participants in this study also recognised that institutional care might be considered when there are extreme levels of disability. This change in attitude may be due to the country’s high poverty level. As such, participants were more open to any form of care irrespective of their deeply held traditions and norms so long as this option would provide for all their basic needs. Additionally, as emphasised by the older persons in the qualitative research, such arrangements were acceptable so long as their fundamental rights of freedom and dignity were not infringed upon.

This finding is contrary to previously documented research and anecdotal evidence. Van der Geest reported similar findings to this study in Ghana [53]. This finding may reflect the reality on the ground whereby older adults have seen the reduced viability of sole reliance on the family [53]. The added attraction, reasoned by the participants during the FGD, is that being in an institutional setting may provide an increasingly rare opportunity for older dwellers to interact with their peers. This study’s younger participants (<70 years) favoured institutional care. This finding may be because they were more likely to be educated and exposed to western culture. However, participants were more favourably disposed to institutional care if the facility was privately owned or managed by religious bodies. This finding may be because of the high level of religiosity in the country, and the trust older persons have for their religious leaders.

Participants who rejected institutionalisation as an option for LTC of older persons did so because they opined that it was based on a foreign culture. Additionally, there is a high expectancy of reciprocity in the African setting. As such, the expectation is that if the older person has laboured to train the child, the child would, in turn, do the same [38, 52, 57, 64]. Also, some participants who did not support institutional care opposed the arrangement because of the cost. As reflected in the FGD, participants believed that caring for older persons is expected to be by the children or family and not outsourced or paid for. Lehnert and colleagues, in their review, reported that married older persons mostly prefer informal LTC arrangements [65]. Also, older persons with higher incomes or who cared for an older relative were more likely to choose informal arrangements for their LTC. However, the review revealed that age, gender, and education yielded inconsistent effects [65].

Findings from the FGD revealed that participants in both locations accepted institution-based care arrangements, mainly if the facilities are provided by the government and quality is assured. There was, however, some scepticism as participants felt the government was not inclined to do such.

Conclusion

To our knowledge, this is the first study that explores the preferences for LTC of older persons in Nigeria. Family-based care is the predominant care system for older persons in many SSA, including Nigeria. However, the present form of family-based care cannot deliver care at the recommended quality of LTC. This study has shown that ageing in place was the choice of community-dwelling older persons in the selected communities, not because they lacked any other affordable alternative. The participants prefer to age in place even with noted functional decline. This study contributes to ongoing global efforts to improve the welfare of older persons, particularly in LMICs. Information obtained will assist in targeted action for providing community-based LTC for older persons in Nigeria and similar resource settings.

Acknowledgments

My gratitude goes to all my research assistants, ably coordinated by Miss Odunayo Ife Ajayi, for their efforts during data collection and my wonderful participants in Igboora and Yemetu Alaadorin.