Indoor Built Environment and Older Adults’ Activity: A Systematic Review

Abstract Although the physical environment can influence people’s activity, there are few knowledge syntheses for indoor environments and older adults’ daily life routines. Therefore, we conducted a systematic review of peer-reviewed evidence to inform future research and practice. Inclusion criteria were studies with any research designs, across all years and languages focused on older adults 60 years of age or more, on physical activity/sedentary behaviour and the indoor environment. After searching five databases, two authors completed title/abstract and full-text screening. The last search was on December 19, 2020. We screened 1,367 citations, and included 23 studies situated in private or collective dwellings (e.g., assisted living). We identified physical activity-supportive indoor features across three domains: campus (e.g., amenities, pathways), building (e.g., area, floor level), and fixtures (e.g., elevators, hallways). Knowledge of indoor environmental factors for older adults’ engagement in daily activities can guide future research and policy on housing design.

with housing, aging in place is also facilitated by how older adults care for themselves through engagement in physical activity (Tao, Zhang, Gou, Jiang, & Qi, 2021;Van Holle et al., 2016;Yen, Michael, & Perdue, 2009). Furthermore, older adults often participate in physical activity at home, rather than in other locations (Chaudhury, Campo, Michael, & Mahmood, 2016). Physical activity can improve physical function, and reduce the risk of chronic diseases and adverse events, such as falls (World Health Organization, 2020b). In contrast, prolonged sedentary behaviour can increase the risk of chronic conditions (e.g., type-2 diabetes) and lead to poor health outcomes (World Health Organization, 2020b).
There are systematic reviews on the effect of the built environment on physical activity for various demographics, focusing on neighborhood, active transport, and outdoor environment design (Smith et al., 2017;Tcymbal et al., 2020;Thornton et al., 2017;Yen et al., 2009). Studies have frequently explored the effect of the built environment on older adults and their physical activity (Cerin, Nathan, van Cauwenberg, Barnett, & Barnett, 2017;Tao et al., 2021;Van Holle et al., 2014. Despite the increasing evidence base for the outdoor environment, less emphasis has been placed on the indoor physical features needed to support active living for older adults (Ahrentzen & Tural, 2015;Annear et al., 2014;Ashe, 2018) even though most people spend most of the day inside.
Therefore, the aim of this systematic review was to synthesize available evidence across all study designs to describe features of the indoor environment and physical activity and/or sedentary behaviour for older adults 60 years of age and older. For this synthesis, we define the indoor built environment as internal space(s) in which older adults reside and engage, and adjacent spaces (e.g., backyard, porch, driveway) (Peel et al., 2005).

Methods
We used the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines to conduct this systematic review (Moher, Liberati, Tetzlaff, & Altman, 2009), and registered it with PROSPERO (Registration No. CRD42018095359).

Inclusion and Exclusion Criteria
We included studies using any study designs, years, and languages. The study population included older adults 60 years of age and older, or the mean age of the study sample was over 60 years of age. Studies included outcomes related to physical activity and/or sedentary behaviour of older adults. The studies also reported features of the indoor environment. We excluded studies if they focused on populations younger than 60 years, or on people with dementia and falls-related research. A "one size fits all" approach does not apply to all individuals in a population, and this is especially true for people with dementia or falls-related research. We also excluded studies focused on ambient indoor temperature, because it is not part of the built environment. We excluded studies without physical activity or sedentary behaviour outcomes.

Information Sources and Search Strategy
We searched the following databases: MEDLINE ® and Embase, EBSCO Databases, PubMed, and Google Scholar (title only). We used the following headings to guide our search: population (older adults), exposure (indoor environment), outcome (physical activity or sedentary behaviour). The search strategy for MEDLINE is provided in Table 1. We conducted forward and backward citation searches for included studies. We completed our last search on December 19, 2020.

Study Selection
Two authors (FA, MA) completed title and abstract screening (Level 1) based on a priori criteria. For Level 2 screening, the full texts of all included studies were reviewed by the same two authors ( Figure 1); they resolved discrepancies through discussion and consensus.

Data Extraction
We extracted the following information for each study: title, first author, year, location, conflicts of interest, purpose, study design, funding resources, participants, indoor and/or housing and campus features, and physical activity and sedentary behaviour outcomes. One author extracted data (F.A.), and a second author (M.A.) checked 10 per cent of entries for accuracy. The same two authors (F.A., M.A.) checked the data again during the synthesis process. We used Covidence by Veritas Health Innovation (Melbourne, Australia) to conduct this review.

Physical Activity and Sedentary Behaviour
Physical activity is energy expenditure produced by skeletal muscles during movement (Caspersen, Powell, & Christenson, 1985;World Health Organization, 2020b) and it includes various household activities (e.g., activities of daily living [ADLs]), sports,exercise (defined as planned and repetitive movement; e.g., swimming), and other activities (e.g., work-related) (Caspersen et al., 1985). Sedentary behaviour is defined as "any waking behaviour characterized by an energy expenditure ≤1.5 metabolic equivalents of task (METs), while in a sitting, reclining, or lying posture" (Sedentary Behaviour Research Network, 2012, p. 1). In our synthesis, we use the terms physical activity or sedentary behaviour (as defined here), or specify the type of physical activity (e.g., ADLs, exercise, walking).

Private and Collective Dwellings
We defined private dwellings (e.g., houses, apartments, town homes) as residences where older adults resided in their own or rented property with access to a private entrance (Statistics Canada, 2017b) and in a community of their choice (U.S. Department of Health and Human Services, n.d.). Collective dwellings (e.g., independent living, assisted living, and retirement villages) are residences where older adults either have private or shared units, live amongst a collective of older adults, and receive a range of services from meal preparation to bathing, as required (BC Seniors Living Association, 2021; Province of British Columbia, 2021; Statistics Canada, 2017a). We define "campus" as the site or property of collective dwellings and campus features, including available resources (e.g., pools, gyms) and destinations (e.g., clubhouses, gardens, and shops).

Indoor Environments
We extracted data on features of the indoor and adjacent environments. The indoor environment was within the residential unit, whereas features of the adjacent spaces were immediately outside the residential unit (Peel et al., 2005). In private dwellings, residential units included the house or apartment, and adjacent spaces included indoor hallways in apartments, backyards, gardens, and front lawns (e.g., distance between the house and garbage disposal or mailbox) (Peel et al., 2005). For collective dwellings, there were living units, resources (e.g., indoor, and outdoor pools), destinations (e.g., clubhouses, shops) and other features (e.g., indoor hallways and outdoor paths) located on the campus. the coded data and noted any discrepancies or questions to discuss at the next meeting. During the final meeting, authors confirmed the findings and created a visual representation of the synthesis ( Figure 2). We present the mean and standard deviations, if available, in the tables. If this information was not available, we included other data (e.g., median, interquartile range).
The remaining studies received high quality scores ranged from 0.85 to 1.00.

244
Farah T. Azim et al.  participants in private dwellings was from 25 to 65 years, and for people in collective dwellings it was from 65 to 97 years.
in another study (Pettigrew et al., 2020). Studies reported that older adults in private dwelling spent more time participating in light (Park et al., 2019) and moderate to vigorous physical activity (Park et al., 2019;Pettigrew et al., 2020); and had greater step counts (Cress et al., 2011) than their collective dwelling counterparts (Table 4).
After moving into a collective dwelling, for some participants, the percent of daytime physical activity decreased (13.3-10.2%) and sedentary behaviour increased (86.7-89.7%) (Regan et al., 2016). Prior to home modifications, older adults spent 5.5 hours performing leisure activities, which increased to 7.5 hours after renovations (Niva & Skär, 2006). The study also found that time spent at rest decreased from 11.5 hours to 9.0 hours (Niva & Skär, 2006).

Indoor Features by Domains
We categorized indoor environment features into the following three domains: campus, building, and fixtures. Figure 2 provides a summary of data synthesis. The campus domain encompassed features of the collective dwelling incorporating aesthetics and visibility, amenities and recreation, destinations, and outdoor pathways. The building domain included the following indoor features: area, floor level, and type of dwelling. The fixture domain consisted of the following features: elevators; indoor hallways; and stairs and  (Kmet et al., 2004). The minimum score is 0.00 and the maximum score is 1.00. a Study type designates whether the study design was qualitative, quantitative, or mixed methods. Qualsyst has different evaluation criteria for each study type. If the study is mixed methods, the study must be evaluated using both qualitative and quantitative criteria, resulting in two scores.

248
Farah T. Azim et al.  Moving out of the home daily 709/848 participants move out of the home daily. 127/848 participants move out of the home less than daily.
Note. Only mean (SD) values were included unless specified. a Quantitative outcomes for physical activity and/or sedentary behaviour were classified as either objective or self-report measures. Objective measures use tools (e.g., accelerometers) to report physical activity and/or sedentary behaviour, and can include accelerometry, step count, and energy expenditure. Self-report measures are when participants are asked to report their own physical activity and/or sedentary behaviour through questionnaires or surveys. b We stratified studies by private, collective, or transition to collective dwelling. Private dwellings are either rented or owned property in the community (Statistics Canada, 2017b; U.S. Department of Health and Human Services, n.d). Collective dwellings include units where one lives amongst a collective of other older adults who receive support services (BC Seniors Living Association, 2021; Province of British Columbia, 2021; Statistics Canada, 2017a). If the study were observing older adults when they lived in a private dwelling and after they moved to a collective dwelling, we classified it as transition to collective dwelling. c If specific mobility instruments were not used (e.g., SPPB), we reported available data on whether participants used mobility aids and on the prevalence of chronic conditions according to their study groups. d Only quantitative physical activity and sedentary behaviour data were included. Specific instruments (e.g., CHAMPS), accelerometry, step count, energy expenditure, and study/investigator-developed outcomes were reported only. We defined "physical activity" as energy expenditure by skeletal muscles (World Health Organization, 2020b) through various types of activities (e.g., household tasks, sports, exercise) (Caspersen et al., 1985). Sedentary behaviour was defined as "energy expenditure ≤ 1.5 metabolic equivalents of task (METs), while in a sitting, reclining or lying posture" ( ramps. Tables 6 and 7 describe how features were associated with physical activity and/or sedentary behavior.

Aesthetics and visibility
One study observed that older adults who perceived the retirement village campus as "more aesthetically pleasing features (i.e., having more trees, greenery, and pleasant natural features) were more likely to engage in more leisure walking" (Nathan et al., 2014b, p. 10), whereas another study reported that the visibility of outdoor features (e.g., courtyards) was associated with uptake of physical activity (e.g., walking in the courtyard) (Joseph et al., 2005).

Amenities and recreation
Facilities offered amenities, recreational resources and/or activities (e.g., pools) (Haney et al., 2018;Joseph et al., 2005), exercise classes (Haney et al., 2018;Regan et al., 2016), trips and scavenger hunts (Voss et al., 2020), and support for older adults to participate in physical activity (Haney et   Note. Only mean (SD) values were included unless specified. a Quantitative outcomes for physical activity and/or sedentary behaviour were classified as either objective or self-report measures. Objective measures use tools (e.g., accelerometers) to report physical activity and/or sedentary behaviour, and can include accelerometry, step count, and energy expenditure. Self-report measures are when participants are asked to report their own physical activity and/or sedentary behaviour through questionnaires or surveys. b We stratified studies by private, collective, or transition to collective dwelling. Private dwellings are either rented or owned property in the community (Statistics Canada, 2017b; U.S. Department of Health and Human Services, n.d). Collective dwellings include units where one lives amongst a collective of other older adults who receive support services (BC Seniors Living Association, 2021; Province of British Columbia, 2021; Statistics Canada, 2017a). If the study were observing older adults when they lived in a private dwelling and after they moved to a collective dwelling, we classified it as transition to collective dwelling. c If specific mobility instruments were not used (e.g., SPPB), we reported available data on whether participants used mobility aids and on the prevalence of chronic conditions according to their study groups. d Only quantitative physical activity and sedentary behaviour data were included. Specific instruments (e.g., CHAMPS), accelerometry, step count, energy expenditure, and study/investigator developed outcomes were reported only. We defined "physical activity" as energy expenditure by skeletal muscles (World Health Organization, 2020b) through various types of activities (e.g., household tasks, sports, exercise) (Caspersen et al., 1985). Sedentary behaviour was defined as "energy expenditure ≤ 1.5 metabolic equivalents of task (METs), while in a sitting, reclining, or lying posture" (Sedentary Behaviour Research Network, 2012, p. 1). e Baecke, et al., 1982;Emplaincourt et al., 1997;Voorrips et al., 1991. f Stewart et al., 2001. g Kielhofner 2002. SD = standard deviation; SPPB = Short Physical Performance Battery; CHAMPS = Community Healthy Activities Model Program for Seniors.
Zimring, 2007; Kotlarczyk et al., 2020;Nathan et al., 2014a,b;Regan et al., 2016;Voss et al., 2020). Two studies demonstrated that a lack of amenities or activities limited older adults' participation in physical activity (Joseph et al., 2005;Voss et al., 2020). In contrast, access to amenities resulted in less opportunity to engage in ADLs, which may reduce light activities and increase sedentary behaviors (Kotlarczyk et al., 2020).

Destinations
Destinations, including on-site gardens (Joseph et al., 2005), local shops, and clubhouses (Nathan et al., 2014a,b) promoted physical activity. Two studies reported that dining halls or similar destinations for eating meals encouraged walking (Kotlarczyk et al., 2020;Voss et al., 2020), but also resulted in longer periods of sedentary time while older adults were served meals (Kotlarczyk et al., 2020). Note. This table provides further information on how the indoor features within the fixtures domain affects physical activity. This domain applies to both private and collective dwellings. a "Activity" is defined as physical activity, which is energy expenditure by skeletal muscles (World Health Organization, 2020b) through various types of activities (e.g., household tasks, sports, exercise) (Caspersen et al., 1985). b Adjacent environment includes spaces immediately outside the residential area, such as hallways, gardens, backyards, and front lawns (Peel et al., 2005).

Outdoor pathways
Older adults observed that well-planned and connected paths influenced walking (Joseph & Zimring, 2007). Longer path lengths, steep paths, and campuses with hills were associated with increased participation in recreational walking (Joseph & Zimring, 2007).

Building Area
Two studies observed that smaller spaces limited physical activity (Reid, 2004;Voss et al., 2020), and that renovations (e.g., bathroom expansions in private dwellings) increased engagement in ADLs (e.g. showering) (Thordardottir et al., 2020). However, older adults experienced difficulties navigating and engaging in physical activity when retirement villages were too large (Nathan et al., 2014a,b).

Floor level
In private dwellings, high rise urban areas restricted physical activity (Portegijs et al., 2017), and older adults had trouble moving between apartment floors (Lilja & Borell, 1997), whereas one-story, ground level houses were associated with increased physical activity (van den Hombergh et al., 1995). Another study found no observed evidence between the apartment floor level and physical activity of older adults residing in high-rise retirement communities (Cheng et al., 2014).

Type of dwelling
There were differences in physical activity engagement across different dwellings. For example, older adults living in private dwellings participated in more physical activity (Harrison et al., 2010;Park et al., 2019;Pettigrew et al., 2020),whereas some older adults in collective dwellings participated in less physical activity (Cress et al., 2011;Harrison et al., 2010;Pettigrew et al., 2020). After relocating to a collective dwelling, older adults reported an increase in sedentary behaviours (Kotlarczyk et al., 2020).

Fixtures Elevators
Lack of access to an elevator restricted physical activity (Perez-Hernandez et al., 2018), whereas another study found that elevators limited physical activity (van den Hombergh et al., 1995).

Indoor hallways
In collective dwellings, indoor hallways promoted walking (Voss et al., 2020), and older adults preferred to have "signs indicating the length of hallways as a way for residents to track their progress while walking" (Kotlarczyk et al., 2020, p. 8).

Stairs and ramps
Older adults reported restrictions when participating in physical activity because of narrow stairs without hand rails, high steps, steep ramps (Reid, 2004), and high threshold or step difference (Benzinger et al., 2014). These features also restricted access to spaces (e.g., waste disposal bins) (Benzinger et al., 2014). Conversely, ramps enabled physical activity (Thordardottir et al., 2020).  Niva and Skär (2006) observed that the following modifications increased accessibility: removal of thresholds, new taps in the bathroom and kitchen, and wider doorways. Some housing features limited accessibility: narrow doors or doorways (Benzinger et al., 2014;Niva & Skär, 2006;Reid, 2004), heavy doors (Reid, 2004), and thresholds and room design (Niva & Skär, 2006). One study also discussed mobility limitations from insufficient "places to grab onto to help [older adults] through the entrance" (Reid, 2004, p. 206).

Safety and environmental hazards
Safety and environmental hazards were identified as limiting physical activity for older adults. Kotlarczyk et al. (2020) reported that pebbles and uneven sidewalks prevented older adults from walking outside in collective dwellings. In private dwellings, Reid (2004) found that the following safety and environmental hazards limited physical activity: uneven flagstones, cement, and floor; narrow stairs without handrails; long driveways or steep ramps; heavy doors; and water on balconies. Two studies also reported that safety affected participation in ADLs (Petersson et al., 2012;Thordardottir et al., 2020); and that improving safety through house modifications could enable older adults to engage in ADLs (Thordardottir et al., 2020).

Discussion
This systematic review synthesizes evidence for a relationship between the indoor environment and physical activity in older adults. We found limited evidence for sedentary behaviour, but identified features of the relationship between the indoor environment and physical activity across three domains: campus, building, and fixtures. Features which enabled physical activity in the campus domain were: aesthetics (Nathan et al., 2014a,b), outdoor features (Joseph et al., 2005), amenities and recreation (Haney et al., 2018;Joseph et al., 2005;Joseph & Zimring, 2007;Kotlarczyk et al., 2020;Nathan et al., 2014a,b;Regan et al., 2016;Voss et al., 2020), and destinations (Joseph et al., 2005;Kotlarczyk et al., 2020;Nathan et al., 2014a,b;Voss et al., 2020). Absence of amenities and recreational resources limited physical activity (Joseph et al., 2005;Voss et al., 2020). However, the presence of dining halls and some amenities promoted sedentary behaviour by reducing the opportunity to engage in ADLs (Kotlarczyk et al., 2020). For the building domain, greater area (Thordardottir et al., 2020), ground level housing (van den Hombergh et al., 1995), and private dwellings (Harrison et al., 2010;Park et al., 2019;Pettigrew et al., 2020) promoted physical activity. Sedentary behaviour reportedly increased after transition to a collective dwelling (Kotlarczyk et al., 2020). The following features hindered physical activity in the building domain: smaller area (Reid, 2004;Voss et al., 2020), larger retirement village campuses (Nathan et al., 2014a,b), highrise buildings (Lilja & Borell, 1997;Portegijs et al., 2017), and collective dwellings (Cress et al., 2011;Harrison et al., 2010;Pettigrew et al., 2020). In the last domain, fixtures which supported physical activity included indoor hallways (Kotlarczyk et al., 2020;Voss et al., 2020), and ramps (Thordardottir et al., 2020). Stairs which were narrow and without handrails, steep ramps (Reid, 2004), and high threshold or step differences (Benzinger et al., 2014) restricted physical activity. The presence of elevators (van den Hombergh et al., 1995), along with the lack of access to elevators (Perez-Hernandez et al., 2018), also limited physical activity. Indoor features related to safety and environmental hazards also impeded engagement in physical activity, such as uneven floors (Kotlarczyk et al., 2020;Reid, 2004). Our review observed that the availability of amenities and recreational resources, such as golf courses and pools, can increase participation in physical activity for older adults living in collective dwellings (Haney et al., 2018;Joseph et al., 2005;Joseph & Zimring, 2007;Kotlarczyk et al., 2020;Nathan et al., 2014a,b;Regan et al., 2016;Voss et al., 2020). Another systematic review highlighted how the type of facility can affect sedentary behaviour: amenities related to exercise were associated with lower sedentary behaviour, whereas socialization or educational activities (e.g., salons, music rooms) were associated with greater sedentary behaviour (Ahrentzen & Tural, 2015). Despite access to these resources, other work observed that only half of the amenities were used by older adults living in retirement villages (Holt, Lee, Jancey, Kerr, & Howat, 2016). This finding suggests that simply increasing the number of and/or access to amenities is not enough to increase uptake of physical activity. The Model of Human Occupation proposes that volition (characterized by people's values and interests) can affect engagement in activities (Kielhofner & Burke, 1980). Specifically, people are more likely to engage in activities that they find meaningful (Kielhofner & Burke, 1980). Behaviour strategies and identification of possible barriers to and facilitators of physical activity engagement should also be considered (Jancey et al., 2008). Evidence suggests that peer leaders, staff, or facilitators are more influential in the uptake of physical activity, than is access to recreational amenities (Ahrentzen & Tural, 2015;Dorgo et al., 2009;Jancey et al., 2008as cited in Holt et al., 2016). Therefore, it simply may not be enough to build facilities. The social environment and other behaviour strategies play a role in the adoption and maintenance of physical activity (Annear et al., 2014). Ahrentzen and Tural (2015) included people with dementia in their systematic review of active aging across dwellings. Similar to our findings, their review observed that steep ramps hindered physical activity (Ahrentzen & Tural, 2015). They also noted higher step counts were associated with larger areas and communal dwellings, whereas home modifications enabled participation in ADLs (Ahrentzen & Tural, 2015), which are consistent with our findings. The review (Ahrentzen & Tural, 2015) had more results for sedentary behaviour, reporting that the frequency of indoor hallways was associated with less sedentary behaviour (Kerr et al., 2011as cited in Ahrentzen & Tural, 2015. Another review found that the "smoothness" of paths, accessibility, and safety increased participation in physical activity (Annear et al., 2014). However, the review reported that "poor-quality" (p. 602) pathways served as a barrier to physical activity (Annear et al., 2014), which our review did not find. However, these differences may be because of the different populations under study: one review included people with dementia (Ahrentzen & Tural, 2015), and other reviews studied the implications of social and societal effects (e.g., relationships with staff working at collective dwellings, poverty) (Ahrentzen & Tural, 2015;Annear et al., 2014).
We observed several illustrations of the connection between the older adult and their environment. Different settings could potentially support specific types of physical activity: Older adults in collective dwellings often do not engage in as many ADLs because they receive services from the facility (e.g., prepared meals reduce the need to cook) (Kotlarczyk et al., 2020). Similarly, older adults can influence their environments. For example, when older adults move to collective dwellings because of functional decline (Crisp, Windsor, Butterworth, & Anstey, 2013), they often move into a space smaller than their previous private dwelling (Hansen & Gottschalk, 2006). Another study observed that hills promoted recreational physical activity (Joseph & Zimring, 2007), whereas some evidence demonstrates challenges with using hills on campus (Holt et al., 2016). Perez-Hernandez et al. (2018) noted that lack of access to elevators restricted physical activity (Perez-Hernandez et al., 2018), whereas van den Hombergh et al. (1995) found that elevators limited physical activity (van den Hombergh et al., 1995). This discrepancy may be explained in two ways: lack of access to elevators restricts the frequency of older adults leaving their homes, resulting in decreased physical activity; and/or access to elevators results in reduced use of stairs, which could also impact overall physical activity. Elevators and stairs are dependent on people's mobility. For those without restrictions, stairs can help maintain physical activity and function.
The intricate relationship between older adults and their environments can be understood through the Canadian Model of Occupational Performance and Engagement (CMOP-E) (Law et al., 1977;Townsend & Polatajko, 2007) and the personenvironment fit framework (Su, Murdock, & Rounds, 2015). The CMOP-E explains how the person, environment, and occupation (defined as a person's role in an environment [Warren, 2002]) interact when people are engaging in various behaviours (Law et al., 1977;Townsend & Polatajko, 2007). The person-environment fit framework suggests that "people shape their environments and environments shape people" (Rounds & Tracey, 1990 as cited in Su et al., 2015, p. 83). The CMOP-E and person-environment fit apply here, such as how functional decline can result in an older adult moving into a smaller home (Crisp et al., 2013;Hansen & Gottschalk, 2006), which may not provide as many opportunities for physical activity (Reid, 2004;Voss et al., 2020).
Accessibility and safety were cross-cutting domains because they can impact multiple levels (campus, building, and fixtures) of the indoor environment. Evidence suggests that limitations in accessibility are negatively correlated with physical activity in adults with disabilities (Saebu, 2010). Older adults in both private and collective dwellings reported that hazards such as uneven grounds or floors prevented participation in physical activity (Kotlarczyk et al., 2020;Reid, 2004). This finding is supported by other evidence which report that "obstructions on the pathway" (Holt et al., 2016, p. 408) were a barrier to walking in collective dwellings (Holt et al., 2016). Private dwellings often require modifications, such as the installation of ramps or grab bars, to increase both accessibility and safety (Niva & Skär, 2006;Thordardottir et al., 2020), demonstrating their synergistic relationship, and can encourage engagement in ADLs (Ahrentzen & Tural, 2015). Accessibility and safety are integral to Universal Design principles (Connell et al., 1997;Null, 2013a).
The findings from this synthesis align well with the seven principles of Universal Design (Null, 2013b): "products, environments, programmes and services to be usable by all people… without the need for adaptation or specialized design" (Office of the United Nations High Commissioner for Human Rights, 2021). Housing and campus design should consider that features such as safety and aesthetics (Equitable Use), how older adults choose to use outdoor campus pathways (Flexibility of Use), minimizing hazards such as uneven floors (Error Tolerance), managing heavy doors (Reid, 2004;Thordardottir et al., 2020) (Physical Effort), and living space (Size and Space for Approach and Use) may influence older adult's participation in physical activity (Nathan et al., 2014a, b;Reid, 2004;Thordardottir et al., 2020;Voss et al., 2020). Universal Design principles coincide with the Global Age-Friendly Cities Project (World Health Organization, 2010). Older adults sometimes reside in older housing which may require retrofitting according to Universal Design principles. Although designing activity-friendly housing from the beginning is ideal, initiatives such as Complete Streets (for outdoor environments) (Transport Canada, 2009, p. 1), provides an example of retrofitted infrastructure to accommodate all people, and could be considered within the housing sector to promote healthy, active aging in place.

Strengths and Limitations
Our study included all studies regardless of language, dwelling, or study design. Because our review only included studies from higher income countries, the results may not be representative of other regions. Only a limited number of studies reported data on race/ ethnicity; therefore, the findings may not be generalizable.
The inclusion of both qualitative and quantitative studies strengthened our review, as we could draw upon different data. Although only one author extracted data, we tried to mitigate risk by having a second author review and complete data extraction for 10 per cent of the studies. Our review did not exclude studies of low quality; however, only one study had a quality score under 0.85/1.00. Differences in physical activity engagement could arise from varying functional mobility of older adults across dwellings, and/or the level of care that older adults received in their residence (e.g., laundry services). Further, we located limited findings for indoor features and sedentary behaviour. Finally, we did not review the relationships of physical activity or sedentary behavior and (i) social environment or (i) health care costs.

Implications of the Main Findings
The findings of this synthesis could inform future housing policy. As the population of older people increases globally, governments may need to focus on providing activity-supportive housing and/or retrofitting pre-existing infrastructure to support accessibility and physical activity.

Future Research and Recommendations
Although our study explored the effects of indoor environment on physical activity and/or sedentary behaviour, future studies should take the social environment and health care costs into consideration. Future studies should investigate the relationship between the indoor environment and sedentary behavior. Understanding this relationship can assist with designing indoor environments, which may reduce prolonged periods of sedentary behaviour. We need more research to inform retrofitting existing infrastructure, physical activity, and Universal Design principles in an effective and efficient manner. In the long term, future research or policy could consider developing a rating system for evaluating physicalactivity-friendly buildings for aging in place. Additional research is needed for low-and middle-income countries to improve representation and generalizability, especially as the studies included in this review were from higher-income countries. Further, for studies conducted in higher-income countries, greater diversity of study participants should be included in future research. We also encourage researchers to report data for race/ethnicity, as only four studies included these data in this synthesis.
This systematic review was registered at PROSPERO: CRD42018095359.
Canadian Journal on Aging / La Revue canadienne du vieillissement 255