Search Results

Our database searches returned 15,522 publications, with 13,615 unique citations. A total of 2677 (not counting 128 duplicates) were added after the updated search in August 2018. After screening and full-text review, 34 articles were included (Figure 1).

Figure 1 Overview of Article Selection Process

Summary of Study Characteristics

Table 2 provides an overview of included articles’ study characteristics. Most studies (44%) were conducted it the United States, followed by Australia (21%), Canada (12%), and Europe (12%). Two studies (6%) were conducted each in Israel, China, and Taiwan. More than half of articles (56%) were published between 2007 and 2012, with only 12 per cent having been published before 2006. The majority of studies (65%) were published in journals that were indexed for MEDLINE. Articles were most frequently published in the Journal of Applied Gerontology (15%) and Journal of Housing for the Elderly (12%). The former is indexed for MEDLINE but the latter is not.

Table 2 Study characteristics

Note. ADL = activities of daily living; IADL = instrumental activities of daily living.

The most frequently used descriptor for supportive housing was retirement living, with 44 per cent of studies using “retirement” in their definition of their setting, followed by “assisted living” (21%). Settings that included “retirement” were continuing care retirement communities, retirement villages, and retirement communities. Remaining articles (35%) used a variety of descriptors, such as senior housing/houses (including congregated senior housing and housing for seniors), supportive housing, and government-subsidized senior citizen apartment buildings (Table 2).

The push and pull framework was the most commonly applied framework (27%); half of studies used another conceptual approach, and 24 per cent used none. Almost all 34 articles reported participants’ age and gender, and 65 per cent reported at least one measure of health and/or functioning.

Study Designs

Studies used various designs to explore older adults’ factors for relocating to supportive housing. A qualitative approach was applied in half of the studies, while quantitative approaches were used in 47 per cent (Table 2). Only one study used mixed methods, in which they conducted interviews and applied quantitative instruments (Ewen & Chahal, Reference Ewen and Chahal2013).

Qualitative approaches consisted of interviews with older adults and/or their families who were planning to move to, or already resided in, supportive housing. Most authors analyzed qualitative data using thematic or content analysis (Table 3). A few studies also collected data via participant observation. These studies were guided by a phenomenological approach (Jungers, Reference Jungers2010), a grounded theory approach (Ball, Perkins, Hollingsworth, Whittington, & King, Reference Ball, Perkins, Hollingsworth, Whittington and King2009), and a thematic analysis through micro, meso, and macro lenses (Portacolone & Halpern, Reference Portacolone and Halpern2016).

Table 3 Summary table of reviewed studies

ADL = activities of daily living; IADL = instrumental activities of daily living; SD = standard deviation.

Quantitative approaches largely entailed the use of surveys and questionnaires developed by authors for the purposes of the study (Table 3). The majority of studies used a cross-sectional design to look at relationships between factors surrounding the transition, but a few studies were longitudinal in nature. A pair of studies, for example, administered a survey at two time points, one year apart, to investigate older adults’ relocation outcomes after moving to supportive housing (Smith & Sylvestre, Reference Smith and Sylvestre2008; Sylvestre & Smith, Reference Sylvestre and Smith2009). Two studies used longitudinal data from existing cohorts: the Longitudinal Study on Aging II in the United States (Hong & Chen, Reference Hong and Chen2009) and the ENABLE-AGE Project in Europe (Granbom et al., Reference Granbom, Lofqvist, Horstmann, Haak and Iwarsson2014). Finally, one study used online vignettes to present different scenarios to prospective older adults and their adult children to explore decision-making surrounding relocation (Caro et al., Reference Caro, Yee, Levien, Gottlieb, Winter and McFadden2012).

Guiding Theoretical Frameworks

The majority of studies (77%) used a theoretical framework to guide their inquiry into older adults’ relocation to supportive housing (Table 2). The most frequently used framework was the push and pull framework, which was applied explicitly in 27 per cent of studies, followed by the ecological theory of ageing, which was used in 18 per cent of studies. Different frameworks were used in 32 per cent of articles. The rest of the articles did not report the use of a framework. Table 4 lists these theoretical frameworks.

Table 4 Theoretical frameworks

Most frameworks applied in the studies reviewed were directly related to ageing and relocation: the push and pull framework, the ecological theory of ageing, or frameworks describing different types of movers (e.g. Litwak and Longino Jr’s (Reference Litwak and Longino1987) and Gardner, Browning, and Kendig’s (Reference Gardner, Browning and Kendig2005) models, and concepts designed for examining person–environment interactions (e.g. complementary/congruence model of wellbeing). However, some researchers generalised concepts that are non-specific to older adults or relocation to study this phenomenon, including Rosenbaum’s (Reference Rosenbaum and Rosenbaum1990) theory of learned resourcefulness, grief, ecological system theory (Portacolone & Halpern, Reference Portacolone and Halpern2016), and the theory of planned behaviour (Ajzen, Reference Ajzen, Kuhl and Beckman1985).

Push and Pull Factors Affecting Relocation to Supportive Housing

The articles that applied the push and pull framework revealed several factors involved in older adults’ relocation to supportive housing (Table 2). Push factors for relocation included individuals’ or spouses’ health challenges, increasing social isolation, fear of burdening family, inadequate living arrangements, necessary maintenance of property, and aiming to achieve control over one’s future. The most frequently cited push factor in the studies were older adult’s or their spouse’s declining health (Bekhet, Zauszniewski, & Nakhla, Reference Bekhet, Zauszniewski and Nakhla2009; Crisp, Windsor, Anstey, & Butterworth, Reference Crisp, Windsor, Anstey and Butterworth2013; Ewen & Chahal, Reference Ewen and Chahal2013; Groger & Kinney, Reference Groger and Kinney2001; Stimson & McCrea, Reference Stimson and McCrea2004; Tyvimaa & Kemp, Reference Tyvimaa and Kemp2011). Articles that did not explicitly use the push and pull framework also showed that older adults who were relocated experienced increasing physical decline (Svidén, Wikström, & Hjortsjö-Norberg, Reference Svidén, Wikström and Hjortsjö-Norberg2002), falls (Castle & Sonon, Reference Castle and Sonon2007; Saunders & Heliker, Reference Saunders and Heliker2008), cognitive impairment (Rockwood et al., Reference Rockwood, Richard, Garden, Hominick, Mitnitski and Rockwood2014), and/or functional deficits (Ewen & Chahal, Reference Ewen and Chahal2013; Granbom, Lofqvist, Horstmann, Haak, & Iwarsson, Reference Granbom, Lofqvist, Horstmann, Haak and Iwarsson2014).

Pull factors for relocation were related to one’s lifestyle, community and social amenities, the prospect of receiving care, and affordability (Table 5). Pull factors generally involved the availability of amenities and care that enabled older adults to maintain an existing lifestyle (Stimson & McCrea, Reference Stimson and McCrea2004). Articles that did not apply the push and pull framework also suggested that reasons for relocation related to the maintenance of older adults’ current lifestyle. For example, Kemp (Reference Kemp2008) found that couples who moved to assisted living homes did so because of their desire to continue living together after a spouse’s major health transition. The push and pull factors are described in Table 5.

Table 5 Push and pull factors affecting relocation

Other Factors Influencing Relocation

Articles that used other models or no explicit guiding conceptual framework described additional factors, which may or may not be related to push and pull factors, influencing older adults’ relocation to supportive housing. An article that used the ecological theory of ageing examined how different dimensions affected relocation: functional status, features of current housing, social networks, features of retirement communities, and finances (Caro et al., Reference Caro, Yee, Levien, Gottlieb, Winter and McFadden2012; Sergeant & Ekerdt, Reference Sergeant and Ekerdt2008). Another article described how increasing dependence results in changes in the person–environment fit, which may precipitate the move (Granbom et al., Reference Granbom, Lofqvist, Horstmann, Haak and Iwarsson2014). Using the ecological theory of ageing, Koss and Ekerdt (Reference Koss and Ekerdt2016) categorised older adults’ reasoning for relocation as pre-emptive, where participants believed that their current homes would be suitable in the future, or contingent, where they have anticipated having the need to relocate.

Reviewed articles also explored the impact of adult children (Castle & Sonon, Reference Castle and Sonon2007; Sylvestre & Smith, Reference Sylvestre and Smith2009), older adults’ subjective interpretations of the new residential setting (Smith & Sylvestre, Reference Smith and Sylvestre2008), socio-economic status and race (Ball et al., Reference Ball, Perkins, Hollingsworth, Whittington and King2009), learned resourcefulness (Bekhet, Zauszniewski, & Wykle, Reference Bekhet, Zauszniewski and Wykle2008), grief (Ayalon & Green, Reference Ayalon and Green2012), and the larger cultural and political context (Portacolone & Halpern, Reference Portacolone and Halpern2016; Sergeant & Ekerdt, Reference Sergeant and Ekerdt2008) as factors for relocation.

Study Participants

All articles reported participants’ gender, and all but one (97%) reported participants’ age (Table 2). With the exception of one study that included only women (Saunders & Heliker, Reference Saunders and Heliker2008), 60–70 per cent of participants were women (Table 3). All mean and median ages were greater than 60 years (Table 3). Younger participants (with a mean age of 65 years old) tended to be community-dwelling residents who may have been relocating to supportive housing (Crisp, Windsor, Anstey, & Butterworth,, Reference Crisp, Windsor, Anstey and Butterworth2013; Weeks, Keefe, & Macdonald, Reference Weeks, Keefe and Macdonald2012). In contrast, in articles with participants who were already living in supportive housing, the participants were 70–80 years old (Bäumker et al., Reference Bäumker, Callaghan, Darton, Holder, Netten and Towers2012).

Participants were often described as healthy (Walker & McNamara, Reference Walker and McNamara2013) and/or cognitively unimpaired (Bekhet et al., Reference Bekhet, Zauszniewski and Wykle2008); no studies focused on older adults with significant physical and/or cognitive impairments. Approximately two thirds of studies used at least one measure of health or functioning (Table 6). Up to 62 per cent assessed general health, with self-rated health being the most frequently used instrument. Activities of daily living (ADLs) and instrumental activities of daily living (IADLs) were the second most frequently assessed aspect of health and functioning, with 32 per cent of studies applying an instrument to measure them. Other standardized instruments, such as the Center for Epidemiologic Studies Depression Scale (CES-D) and the Minimum Data Set Cognitive Performance Scale (MDS-CPS), were used to measure depression/mood and cognition, respectively (Table 6). A total of 12 per cent of studies collected information related to specific health conditions from patients and/or their medical records (Ball et al., Reference Ball, Perkins, Hollingsworth, Whittington and King2009; Ewen & Chahal, Reference Ewen and Chahal2013; Hong & Chen, Reference Hong and Chen2009; Sergeant & Ekerdt, Reference Sergeant and Ekerdt2008). Cardiovascular disease and hypertension were the most commonly reported health conditions (Ball et al., Reference Ball, Perkins, Hollingsworth, Whittington and King2009; Ewen & Chahal, Reference Ewen and Chahal2013; Sergeant & Ekerdt, Reference Sergeant and Ekerdt2008).

Table 6 Instruments to assess health and/or function

Articles that commented on participants’ health or functioning generally stated that participants were in good physical health with only minor problems. For example, Groger and Kinney (Reference Groger and Kinney2001) reported that participants had high levels of well-being, with the exception of a few reporting minor “forgetfulness” problems. Studies that used self-rated health as a measure of general health commonly reported that participants were in “fair” or “excellent” health (Bekhet et al., Reference Bekhet, Zauszniewski and Nakhla2009; Huang, Reference Huang2012; Weeks et al., Reference Weeks, Keefe and Macdonald2012). Some of the articles suggest some deficits in ADLs/IADLs among study participants. One study reported average scores of 7.2/10 and 5.5/10 on the Older Americans Resources and Services ADL and IADL Scales (Castle & Sonon, Reference Castle and Sonon2007), whereas another found that only 31 per cent of 215 retirement home residents were independent with two or more IADLs (Caro et al., Reference Caro, Yee, Levien, Gottlieb, Winter and McFadden2012). The three articles that examined how health and functioning impacted relocation found that worse health, dependence with IADLs, cognitive deficits, and accessibility problems were associated with moving to supportive housing (Granbom et al., Reference Granbom, Lofqvist, Horstmann, Haak and Iwarsson2014; Hong & Chen, Reference Hong and Chen2009; Rockwood et al., Reference Rockwood, Richard, Garden, Hominick, Mitnitski and Rockwood2014).