Master’s Thesis

The influence of the built environment on walking among urban, community-dwelling older
adults in the United States: A systematic review and thematic synthesis

Emily Winer
Certificate: Health Promotion Research & Practice
Thesis Type: Systematic Literature Review
Sponsor: Dr. Angela Aidala

Department of Sociomedical Sciences

Columbia Mailman School of Public Health
In partial fulfillment of MPH degree requirements for graduation May 2017
 ABSTRACT
Background and Purpose: Driven by changing demographics, cultural shifts, and
improvements in healthcare, the older adult population (≥65 years) in the United States (U.S.) is
growing at a rapid and unprecedented rate. Simultaneously, the U.S, is becoming increasingly
urbanized, with projections indicating that 87% of the U.S. population will live in urban centers
by 2050. The convergence of an urbanizing and aging population necessitates a focus on aligning
the urban environment with the needs of older adults. Among these needs is continued physical
activity (PA) throughout older adulthood, yet despite the numerous physical, mental, and social
benefits, the majority of older adults fail to meet recommended PA guidelines.
Objectives: To investigate the influence of the outdoor built environment on walking behaviors
among urban, community-dwelling older adults (≥65 years) in the U.S. and to analyze results
with reference to implications for designing and improving communities to support active living
for this population.
Methods: A systematic review was conducted following PRISMA guidelines. Eight databases
were searched for articles published 2011-2016 using search terms related to features of built
environment, older adults, and walking. A total of 3254 abstracts were reviewed for eligibility, of
which 17 studies fit all inclusion criteria. The final study set underwent quality appraisal,
followed by data extraction, analysis, and thematic synthesis.
Results: Findings suggest that older adults engage in PA for two primary purposes, leisure or
transport, and that unique aspects of the built environment support or act as a barrier to each type
of walking. Macroscale components of walkability (e.g., land use mix, street connectivity, and
population density) were consistently and positively associated with transport walking, whereas
microscale elements, such as aesthetics, pedestrian infrastructure, and crossing characteristics,
were more commonly associated with leisure walking.
Conclusions: Results support multiple associations between the built environment and walking
behavior among older adults. This review furthers the current evidence regarding how macro and
microscale features of the built environment can act upon older adults to support or discourage
walking. While this area of research is growing, numerous gaps in the literature were identified.
Future research is needed to improve the generalizability of findings to better inform future
interventions and policies that support this population.

1
 TABLE OF CONTENTS
Introduction ....................................................................................................................................4
Statement of the Problem ..........................................................................................................4
Background and Significance ...................................................................................................5
Physical Activity and Older Adults .........................................................................................5
Walking and Older Adults .......................................................................................................6
The Built Environment and Older Adults ................................................................................7
Review Aims ...................................................................................................................................8
Methods...........................................................................................................................................9
Query Development and Search Strategy................................................................................9
Literature Review ....................................................................................................................10
Phase One...............................................................................................................................10
Phase Two ..............................................................................................................................10
Quality Assurance ....................................................................................................................11
Data Extraction and Analysis .................................................................................................12
Results ...........................................................................................................................................13
Topics, Settings, and Methodologies of Reviewed Literature ..............................................13
General characteristics ...........................................................................................................13
Methodological characteristics of reviewed quantitative studies ..........................................13
Methodological characteristics of reviewed qualitative studies ............................................16
Built Environment Features Identified in the Reviewed Literature ...................................17
Residential and Population Density .......................................................................................21
Walkability ...............................................................................................................................22
Effect on type of walking .......................................................................................................22
Effect on those with mobility limitations...............................................................................24
Street Network Composition ...................................................................................................25
Neighborhood Land Composition ..........................................................................................25
Land use and access to daily destinations ..............................................................................26
Parks and outdoor recreational spaces ...................................................................................27
Public and motorized transport ..............................................................................................28
Pedestrian Infrastructure ........................................................................................................29
Sidewalk presence, quality, and design .................................................................................29
Street crossing amenities and support ....................................................................................30
Presence and quality of street lighting ...................................................................................31
Wayfinding strategies, aids, and cues ....................................................................................31
Safety .........................................................................................................................................32
Street safety............................................................................................................................32
Perceived safety and crime ....................................................................................................33
Environmental Conditions and Appeal .................................................................................33
Aesthetics ...............................................................................................................................33
Weather and environmental conditions .................................................................................34
Discussion .....................................................................................................................................35
Synthesis of Results ..................................................................................................................35
Gaps in the Literature and Directions for Future Research................................................40

2
 Participant diversity ...............................................................................................................40
Geographic diversity ..............................................................................................................41
Age representation .................................................................................................................43
Wayfinding ............................................................................................................................43
Climate change.......................................................................................................................44
Naturally occurring retirement communities .........................................................................44
Study Limitations .....................................................................................................................45
Conclusion ....................................................................................................................................46
References .....................................................................................................................................47
Appendices ....................................................................................................................................51
Appendix I. Adapted Newcastle Ottawa Quality Assessment Scale for Observational
Research ....................................................................................................................................51
Appendix II. Results of Adapted Newcastle Ottawa Quality Assessment Scale ................52
Appendix III. Mays and Pope Scale for Quality Assessment of Qualitative Research and
Results .......................................................................................................................................53
Appendix IV. Characteristics and Results of Quantitative Studies ....................................54
Appendix V. Characteristics and Results of Qualitative Studies ........................................60

3
 INTRODUCTION
Statement of the Problem
Driven by changing demographics, cultural shifts, and improvements in healthcare, the
older adult population (≥65 years) in the United States (U.S.) is growing at a rapid and
unprecedented rate.1 Estimates project a doubling of the U.S. older adult population from 2010-
2050, rising from 13% (40.2 million) to 20% (88.5 million) of the population by 2050.2
Additionally, age demographics within the population of older adults are undergoing a
substantial shift. Most notably, the population categorized as oldest-old (≥85 years), is expected
to increase from 14% of the population over 65 to over 21% by 2050.2 Simultaneously, the U.S.
is becoming increasing urbanized, with projections indicating that 87% of the U.S. population
will live in urban centers by 2050 compared to 81% in 2014.3 The convergence of an urbanizing
and aging population necessitates a focus on aligning the urban environment with the specific
health needs and challenges faced by older adults.4
Among these needs is continued physical activity (PA) throughout the life course.
Despite physical activity’s profound impact on health and wellbeing, the majority of older adults
fail to meet the U.S. Department of Health and Human Services (DHHS) recommended
guidelines for daily PA.5 This lack in activity has serious health and social consequences, as
sedentary older adults are significantly more likely to experience adverse, long-term health
problems, reduced quality of life, and earlier loss of functional independence.6,7 Further
complicating and underscoring the significance of this problem are the unique physical
challenges and changes older adults face as they age: 42% of adults over 65 currently report
having a health condition or disability, and 28.3% of those 65-74 years and 47% of those 75 and
over report limitations in physical functioning.7,8 These numbers are only expected to increase in
the coming years as the current population of middle-age adults shift into older adulthood.8 Due
to the prevalence of chronic health issues, sedentary behavior, and the continued growth of this
population, the health care costs associated with older adults are projected to escalate at an
alarming and unsustainable rate.6,7
Given these rising concerns, it is imperative to better understand the features of the built
environment that influence PA behaviors among older adults. Prior research studying the
relationship between the built environment and older adult PA indicates that a range of
characteristics, such as population density, land use mix, pedestrian streetscape features, and

4
perceptions of neighborhood quality and safety, can significantly facilitate or act as a barrier to
PA.9 By identifying and further understanding the diverse array of features that shape PA
behavior, communities can be better designed to support the unique needs of older adults, and in
doing so, the built environment can be used as a key resource to help support successful aging,
long-term health, and functional independence among this growing and vulnerable population.
Although past reviews have explored this topic, no known work has systematically investigated
the recently published (2011-2016) quantitative and qualitative literature on the relationship
between the built environment and PA among urban, community-dwelling older adults in the
U.S. The purpose of this paper is to systematically review the latest evidence on this topic and
contribute to the greater understanding of the diverse environmental factors that influence PA
among the urban, community-dwelling older adult population.

Background
Physical Activity and Older Adults
Engagement in regular PA is an essential component of healthy aging. PA confers a
range of valuable health benefits for older adults, including: prevention of obesity, type 2
diabetes, cardiovascular disease, osteoporosis, chronic obstructive pulmonary disease, high
cholesterol, high blood pressure, and certain cancers;6 reduction in the rates of functional
decline; preservation of muscle, bone mass, and healthy body weight; improvements in glucose
control, cardiovascular health, balance, and stability;7 and treatment of certain forms of chronic
pain.6 Beyond these physical health effects, exercise provides numerous mental and cognitive
health benefits, such as reduced rates of depression and anxiety, higher ratings of quality of life,
improved cognitive functioning, and reduced Alzheimer’s risk.5,6 Additionally, PA is associated
with decreased risk of falls, maintenance of physical functioning, and recovery from functional
limitations, all of which are important factors that contribute to continued independent living. In
order to experience these and other substantial health benefits associated with PA, the U.S.
DHHS recommends that older adults engage in 150 minutes of moderate-intensity aerobic
activity or 75 minutes of vigorous-intensity aerobic activity per week, complemented by muscle-
strengthening activities.10
Among community-dwelling older adults, engagement in PA is a critical facilitator of
successful aging-in-place and reduces the likelihood of experiencing the costly health

5
complications linked to sedentary living.6 Furthermore, maintenance of physical ability and
independence helps prevent the associated financial and emotional challenges of relocation to a
nursing home, assisted living, or other long-term care facility.11 Yet with increasing age, the
percentage of older adults who are inactive gradually rises: 35% of adults 65-74 years, 47% of
those 75-84 years, and 64% of those 85 and above report recent PA.12 Simultaneously, the
percentage of older adults meeting DHHS guidelines steadily declines with age: only 42% of
those 65-74 years, 31% of those 75-84 years, and 18% of those 85 years and older meet the
recommended PA levels.12 The majority of older adults not only fail to meet the DHHS
guidelines, but they are largely sedentary, with adults over 70 years averaging less than 10
minutes of accelerometer derived moderate-to-vigorous physical activity (MVPA) per day.7 This
decrease in PA, rise in sedentary behavior, and the associated negative health effects highlights a
major challenge facing the older adult population and the U.S. health system.
Walking and Older Adults
Among the many types of physical activity, walking is cited as a highly beneficial,
accessible, low-risk, and low-cost activity that most older adults can engage in year-round and
across multiple settings.10 For adults over 65 years, walking for a period of at least 10 minutes
counts towards daily MVPA.10 Walking serves diverse purposes, such as leisure-time exercise or
transportation. Even among older adults with mobility limitations or disabilities, walking is
possible with the aid of assistive devices, such as walkers or canes.13 Due to these many factors,
outdoor walking is the primary source of PA among older adults, and is the most frequently cited
activity among those who meet the DHHS PA guidelines.13,14
Further supporting the low-risk nature of this exercise, past research examining rates of
injury among older walkers and joggers found that only 5% of walkers, as compared to 57% of
joggers, experienced a lower extremity injury over a 12 week period.15 Studies examining
walking specifically have identified the clear health benefits of walking among this population,
such as decreased risk of depression, osteoarthritis, colon cancer, hypertension, and dementia.10
Jointly, this research indicates that regular engagement in walking reduces risk for a myriad of
diseases, improves strength and flexibility, and provides relief from arthritis related discomfort.10
As such, walking serves as a preventative measure to keep older adults within their community
and outside of an institutionalized environment.10 However, similar to PA engagement, older
adults engaging in walking as a form of PA decreases with increasing age: 59% among those 65-

6
74 years, 49% of those 75-84 years, and 41% of those 85 years or older report recent walking.12
Understanding the environmental factors that contribute to the marked decrease in walking is
therefore essential in developing communities that encourage and facilitate this health-promoting
behavior.
The Built Environment and Older Adults
Although the causes of this widespread trend of physical inactivity are multifaceted, a
key factor contributing to sedentary behavior is the physical environment in which older adults
live and spend their time.7 Due to lifestyle changes and transitions, older adults are highly
susceptible to the influence of the built environment.16 Notably, this period of the life course is
marked by departure from the labor force, with 73.8% of adults 65-74 years and 92% of adults
over 75 years no longer participating in the workforce.17 Upon retirement, individuals are more
likely to spend the majority of their time in the proximal area surrounding their place of
residence, and as a result, the neighborhood environment has greater opportunities to exert
influence on behavior.16
Older adults experience many concerns in regards to ability to age-in-place and maintain
residence in their communities safely, comfortably, and independently throughout older
adulthood. Among these concerns is the fear of isolation or inability to function independently
due to loss of driving status, especially for older adults residing in auto-dependent settings.6
Many communities lack the features that support mobility and independent community living
among non-driving older adults, thus creating barriers to active living and the maintenance of
functional independence. Additionally, despite its inherently low risk nature and copious
benefits, numerous age-related changes increase both the perceived and realized risks of walking
for older adults. For example, deterioration of visual acuity, which effects the sharpness with
which an individual perceives objects at a distance, puts older adults at higher risk of
misinterpreting the distance between themselves and oncoming traffic.18 When compared to
younger pedestrians, older adults are significantly more likely to accept a smaller gap in traffic
when crossing the road.18 Furthermore, due to reduced mobility, older pedestrians are limited in
their ability to react quickly to avoid immediate danger of oncoming vehicles, cyclists, or other
rapidly-moving objects.18 This reduced mobility impacts the speed at which older adults are able
to cross the road, putting them at increased risk even at timed-intersections, as automated timers
may not provide sufficient time for those with slower walking gaits.18

7
 These vulnerabilities, as well the high likelihood of pre-existing, underlying health
conditions or frailty, contribute to older adults being at increased risk of experiencing severe
pedestrian injuries or fatalities.18,19 The U.S. fatality rate for pedestrians over 75 years is higher
than any other age group, at 2.28 per 100,000.18 Regardless of intersection type (i.e., signalized
or unsignalized), older pedestrians have a higher probability of severe injuries.19 Unsignalized
intersections, however, are particularly problematic. In 62% of pedestrian crashes at unsignalized
intersections, older pedestrians were associated with a higher probability of severe injuries
compared to all other age groups.19 Additionally, past research has identified higher risks of
pedestrian crashes in areas with a larger proportion of older adult residents.20
Considering the growth of the aging population and its particular safety needs and
concerns, communities need to be planned with both older adult safety and accessibility in mind.
Features of the built environment, such as sidewalk quality, street crossing amenities, and
perceptions of traffic safety, may significantly impact the ability of older adults to interact with
their surrounding community. Despite the numerous health benefits conferred through PA
engagement, the sustainable, affordable, and safe nature of walking, and the high need for older
adults to maintain mobility and independence as they age, the built environment factors that
encourage and shape walking behavior among this population are not fully understood. It is
essential, therefore, to understand these variables, as they can be used to inform urban planners,
developers, city transportation officials, and public health departments how to design healthy
communities and implement effective interventions to the built environment that encourage
active living throughout the lifespan.

REVIEW AIMS

The objective of this review is to investigate the influence of the neighborhood built
environment on walking behaviors among urban, community-dwelling older adults living in the
U.S. While the built environment can also refer to the indoor home environment, this review will
focus exclusively on the neighborhood features and qualities that surrounding an individual’s
residence, with a specific focus on macroscale (e.g., land use mix, street connectivity, population
density) and microscale (e.g., sidewalk design, street crossing features, aesthetics) environmental
features. The specific aims of this review are to:

8
 1.! Identify recurring patterns and features of the built environment that influence walking
behavior among urban, community-dwelling older adults in the U.S. (≥65 years).
2.! Compare findings across studies and identify features of built environment that are most
influential on the walking behaviors of this population.
3.! Identify gaps in the literature, directions for future research, and areas for intervention
that would hold the greatest promise to influence this population’s walking behaviors.

METHODS
Query Development and Search Strategy
A systematic literature review was conducted guided by the principles of the Preferred
Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.21 The
following electronic databases were searched for relevant literature: AgeLine, Abstracts in Social
Gerontology, Public Administration Abstracts, PubMed, Social Sciences Full Text, Social Work
Abstracts, The Journal of Planning Literature, and Urban Studies Abstracts.
Search terms and search query were developed from a scoping review of existing
empirical studies and theoretical commentary resulting in: (1) a list of relevant terminology and
Medical Subject Headings (MeSH terms) used in past reference databases; (2) synonyms for
terminology (e.g., population density and residential density); and (3) greater depth of
knowledge regarding how relevant topics have been addressed in the literature. Additionally,
past systematic reviews relating to this topic helped inform the search query development for this
review. Identified terms were searched in PubMed to determine how they are used and defined in
the literature, thus allowing the list of terms to be refined further. Once the final list of terms was
determined, the search query was built out methodically to ensure it captured the full range of
studies relevant to addressing the aims of this review.
Selected databases were searched using the following search terms: (“land use mix” OR
“street connectivity” OR “residential density” OR “street scale” OR “built environment” OR
“walkability” OR “population density” OR “environmental design”) AND (“walking” OR
“walk”) AND (“aged” OR “older adult”). The search was then restricted to research published
within the last five years (July 2011- July 2016) in order to capture the most recent body of
research and build upon previously published reviews that reported on literature from 2011 and
earlier (e.g., Kerr et al22, Moran et al5, and Yen et al23).

9
Literature Review
Phase One
A total of 3254 records were identified through database searching. All titles and
abstracts were reviewed in Phase One for eligibility based on the following predetermined
inclusion and exclusion criteria: (1) Participants were community-dwelling older adults (mean
age !65 years) living in urban areas (central city and the surrounding metropolitan region) in the
U.S.; (2) The study explored the outdoor built environment (exposure) and its impact on
participants’ PA behavior specific to walking (outcome); (3) Study population was not focused
on older adults with a specified pre-existing health condition (e.g., diabetes, obesity, mobility
disability); (4) Study was published in the past 5 years (July 2011-July 2016); and (5) Study was
available in English. Studies using qualitative as well as quantitative methods were included.
Based on the information available from the title and abstract, 65 studies met eligibility criteria
and moved on to Phase Two of the literature review.
Phase Two
The selected 65 studies underwent full-text evaluation in Phase Two. Seventeen (17)
studies met all inclusion criteria and 48 were excluded. Common reasons for removal included:
insufficient fit with this review’s exposure and outcome of interest; study location outside of the
U.S. or in a non-urban location; study focus on older adults with mobility disabilities or other
health conditions (e.g., diabetes, obesity); and inappropriate age or lack of transparency
regarding participant age distribution. For example, some articles focused on older adults !60
years without providing mean age, thus preventing verification as to whether mean age was
greater than or under 65 years. Commonly, studies recruited a range of adult age categories
(including !65), yet provided only aggregated results rather than a breakdown by age. Other
studies were excluded as they focused on indoor built environments, such as malls, or on the
built environment of adults living in long-term care, such as nursing homes or assisted living
facilities. Studies exploring mixed ages or neighborhood types (i.e., not exclusively urban) were
included only if results were clearly differentiated between groups. The seventeen studies that
passed Phase Two review were reexamined to ensure fit, all seventeen were approved for
inclusion in the final set of literature.

10
Table 1: PRISMA Flow Diagram

Quality Assurance
To evaluate the methodological rigor of the included research and whether it may have
shaped findings, a quality assurance analysis was performed on both the quantitative and
qualitative literature. The final 13 quantitative studies were carefully reviewed using an adapted
Newcastle-Ottawa Quality Assessment Scale (NOS) for use with observational research. This
adaptation was developed by Herzog et al24 for use in a systematic review comprised of
primarily cross-sectional literature, and has been adopted successfully by others.25 Using the
adapted NOS, the following criteria were assessed for each of the thirteen (13) quantitative
studies: (1) Selection: representativeness of the sample, sample size, non-respondents,
ascertainment of the exposure; (2) Comparability: controlling for confounding variables; (3)
Outcome: assessment of the outcome; use of statistical tests. No eligible articles were excluded
from content analysis due to methodological issues, though some were identified as weaker than

11
others. See Appendix I for an outline of the NOS protocol and Appendix II for the quality
assurance rating of all quantitative studies included in this review.
The four qualitative studies were appraised using a quality assurance checklist developed
by Mays and Pope.26 While there is less consensus and considerable debate on how methods and
quality can be assessed in qualitative research,26 this checklist was selected due to its use in prior
syntheses of qualitative research27-29 and its endorsement from the Cochrane Collaboration
Qualitative Methods Group.30 A primary concern in quality assessment of qualitative research is
that it is not possible or suitable to judge qualitative research using conventional criteria such as
reliability, validity, and generalizability.26,29 Mays and Pope26 attest that quality can be assessed
using these broad concepts, however, it is necessary to operationalize these terms differently for
the distinct goals and approaches of qualitative research. Past research has noted that the Mays
and Pope checklist is suitable for qualitative research assessment due to its clarity, simplicity,
and thorough approach to critical appraisal without being excessively prescriptive.29 These
considerations were deemed appropriate for this review. The checklist is comprised of questions
regarding relevance, context clarity, design, sampling, data collection, data analysis, and
reflexivity.29 While certain studies were more transparent about methodological processes or
results, none were deemed uniformly poor or excluded based on critical appraisal using the Mays
Pope criteria. See Appendix III for further details regarding the Mays and Pope checklist and
assessment of the four qualitative studies used in this review.

Data Extraction and Analysis

Data extraction was completed systematically in order to obtain an overview of each
study’s methodological characteristics and results. General characteristics and methods were
extracted first, including: study design, purpose, setting (city, state, and region), sample size,
sample age, exposure and outcome measurement, and variables that were controlled for in
analysis (e.g., socioeconomic status (SES), gender, or marital status). Next, relevant results were
extracted through multiple, careful examinations of the Tables, Results, and Discussion sections
of each article. Extracted findings were then grouped together based on similarity, from which
themes and subthemes emerged. Within each theme and subtheme, findings were grouped based
on quantitative or qualitative methodology. Study results that were not relevant to this review’s
research question were not included in the data extraction or analysis.

12
 RESULTS
Topics, Settings, and Methodologies of Reviewed Literature
General characteristics
In total, 17 studies met inclusion criteria: 13 quantitative and 4 qualitative. The majority
of quantitative studies were conducted in either the Northwest (n=8) or Mid-Atlantic (n=8)
regions of the U.S., followed by the West (n=4), Northeast (n=3), and Midwest (n=1), and eight
studies occurred across either two (n=5) or three (n=3) states. The four qualitative studies were
each conducted in a different region: Northwest, West, Midwest, and Northeast. No studies were
conducted in the Southwest, Southeast, or the non-contiguous states (Alaska and Hawaii). Most
studies included both men and women (n=15), with the exception of three quantitative studies
that focused on women only.
Methodological characteristics of reviewed quantitative studies
All quantiative studies used a cross-sectional (n=11) or longitudinal approach (n=2). The
two longitudinal studies (Ding et al31, King et al32) involved measurement of study variables at
two time points taken six months apart. The reviewed literature recruited a range of sample sizes
(≤100 (n=1), 100-500 (n=3), 500-1000 (n=5), 1000-5000 (n=2), >5000 (n=2)) and there were
varied approaches to participant recruitment and data collection. Studies collected data for both
neighborhood features and PA outcomes using either a combination of objective measurement
and survey or interview self-report (n=9), or self-report only (n=4). Table 2 includes a summary
of the methodological characteristics of all reviewed quantitative studies.
The majority of quantitative studies used Geographic Information System (GIS) software to
analyze and objectively measure features of the built environment known to be related to PA
behavior (n=8). For example, four studies used GIS to develop a walkability index for a 500
meter buffer around each participant's home based on residential density, retail floor area ratio,
intersection density, and land use mix.33-36 In a comparable approach, two studies explored
population, intersection, and facility density within 800 and 1200 meters of the participant’s
home.37,38 King et al32 used GIS-measured built environment variables (residential density, retail
floor area ratio, intersection density, and land use mix) as well as income data to categorize
neighborhoods based on walkability and income. Similarly, Siu et al39 used GIS to conduct a
cluster analysis of geographic units in Portland, OR. Objective measures of transit access, land
use mix, intersection density, and population density were then aggregated to develop six unique

13
urban forms (central city, city periphery, suburb, urban fringe with poor commercial access,
urban fringe with poor park access, and satellite city) which were used to define different regions
of the metro area.39

Number of Studies Reference Number

GENERAL
U.S. Region
Northwest 8 16,31-35,39,40
Mid-Atlantic 8 16,31-35,37,38
West 4 35,37,38,41
Northeast 3 37,38,42
Midwest 1 43
Southwest or Southeast 0
Gender
Men and women 10 16,31-35,40-43
Women only 3 37-39
METHODS
Study Design
Cross-sectional 12 16,33-35,37-43
Longitudinal 2 31,32
Sample size
n≤100 1 42
100<n≤500 3 35,40,43
500<n≤1000 5 16,31-34
1000<n≤5000 2 39,41
n>5000 2 37,38
Exposure Data Collection
Geographic Information 4 32,37-39
Systems (GIS)
Self-report (SR) 4 16,41-43
Environmental audit (EA) 1 40
GIS and SR 3 31,33,34
GIS and EA 1 35
Outcome Data Collection
Accelerometer and SR 7 16,31-35,40
SR only 6 37-39,41-43
All Data Collection
Mixed (Objective and SR) 10 16,31-35,37-40
SR Only 3 41-43
Objective only 0

Table 2: General and methodological characteristics of quantitative literature

14
 Four studies combined GIS-derived measurement with a non-objective measure, such as
self-reported perceptions of the neighborhood (n=3) or an environmental audit (n=1).31,33-35 Self-
reported perceptions were measured using the Neighborhood Environment Walkability Scale
(NEWS), a validated scale that assesses participant perceptions of a neighborhood’s urban form
and recreation-oriented variables.16 The environmental audit was conducted using the Microscale
Audit of Pedestrian Streetscapes (MAPS) tool, which measures streetscape characteristics (e.g.,
street design, transit stops, sidewalk qualities, street crossing amenities, and features related to
aesthetics) for a 0.25 mile route from the participant's home towards pre-selected non-residential
destinations.35 All studies using only self-report methods focused on measuring perceptions of
the built environment using NEWS (n=3),16,42,43 with the exception of the work from Li and
colleagues41, which did not specify the measure used for assessing the built environment. Lastly,
Sallis et al40 conducted the only study in this review that employed environmental audit alone,
using MAPS and its shortened version, MAPS-Mini.
To measure PA, seven studies combined objective, accelerometer measured PA with self-
report. All studies used Actigraph accelerometers, which are the most widely used brand in PA
research44 and are reliable and valid measures of PA in the older adult population.33
Accelerometers are used to measure overall moderate or vigorous physical activity (MVPA),
which among older adults is predominantly made up of walking for transport or leisure.44 They
are considered the ideal measurement tool for PA in older adults, as they are simple to use,
require no input from the participant during the collection period, and eliminate bias associated
with subjective recall of past PA via self-report questionnaires, a task that can be particularly
problematic for older adults due to declines in memory recall.44 Within the studies included in
this review, accelerometers were worn for a range of 5-7 days, with a valid day serving as a
minimum of 8-10 hours of continuous wear.33 Self-reported PA was assessed using the
Community Healthy Activities Model Program for Seniors (CHAMPS) survey, a validated
measure among older adults. CHAMPS tasks participants with recalling activities performed
over the previous four weeks, such as walking for errands, leisure, or transport.16
Combining these two measurement strategies is ideal for capturing the range of walking
behavior among older adults. Prior literature has identified that older adults engage in two
primary types of PA, walking for transportation (i.e., to a specified location) and leisure (i.e., for
recreation or relaxation), and that unique factors of the built environment shape engagement in

15
each type of walking.45 By measuring overall moderate-to-vigorous PA, accelerometers are
instrumental in gathering an accurate, general understanding of PA engagement. However, they
are not able to capture the type, purpose, or motivation behind walking behavior, thereby
highlighting the need to compliment objective measurements with self-report data.45
Among the seven studies that measured PA using self-report, two used previously
validated measures that sought to determine walking over the previous seven days. Gallagher and
colleagues43 used items from the Neighborhood Physical Activity Questionnaire, whereas
Maisel42 used selections of the International Physical Activity Questionnaire. Three studies did
not report use of a specific measurement tool, however, they asked participants standard
questions about recent PA engagement (e.g., total number of blocks walked daily;
frequency/total duration of walking in the past week). Alternatively, two studies sought to
determine if participants met the U.S. DHHS recommendation of 150 min/week of walking by
measuring the average time per week participants engaged in difference types of PA over the
past year and assigning a metabolic equivalence task (MET) value for walking based on
pace.37,38
Methodological characteristics of reviewed qualitative studies
Three of the four qualitative studies used exclusively qualitative methods, including
photovoice (n=2)46,47 and semi-structured interviews (n=1)14, and one study used mixed-
methods. 48 All were cross-sectional designs. The photovoice studies involved participants
describing photographs they took for the purposes of depicting perceived facilitators or barriers
to PA engagement.46,47 The one semi-structured interview study looked at built environment
factors that influence older adults’ fear of falling, as well as their ability and motivation to
engage in outdoor PA.14 The fourth qualitative study used a mixed-methods approach by
combining quantitative neighborhood audits with qualitative in-person interviews and walk-
alongs (i.e., an interview conducted while the participant walked through the neighborhood on a
pre-determined route).48 While all four studies differed slightly in study aims, each produced
data that described environmental facilitators and barriers to PA. Table 3 provides a summary of
all general and methodological characteristics of the four qualitative studies included in this
review.

16
 Number of Studies Reference Number
GENERAL
U.S. Region
Northwest 1 47
West 1 46
Northeast 1 14
Midwest 1 48
Mid-Atlantic, Southwest, Southeast 0
METHODS
Study Design
Qualitative 3 14,47,48
Mixed-Methods 1 48
Data Collection Method
Semi-structured interviews 2 14,48
Photovoice 2 46,47
Walk-along 1 48
Sample Size
n≤20 1 14
20<n≤30 1 46
30<n≤40 2 47,48

Table 3: General and methodological characteristics of qualitative literature

Built Environment Features Identified in the Reviewed Literature

The following sections outline the seven primary constructs that emerged from the data:
(1) Residential and Population Density, (2) Walkability, (3) Neighborhood Land Composition,
(4) Street Network Composition, (5) Pedestrian Infrastructure, (6) Safety, and (7) Environmental
Appeal and Conditions. Both the quantitative and qualitative studies covered environmental
variables within two broad categories, macroscale features and microscale features. Macroscale
features consist of the broader structural features that comprise an environment and its overall
walkability index, such as population and residential density, street connectivity, and land use
mix.35 Conversely, microscale features make up the finer details that more directly shape the
pedestrian experience and environment, such as aesthetics, sidewalk quality and design, and
street crossing amenities.35 Macroscale-oriented themes are described first, followed by
microscale themes, however, many themes included cross-over between the two general
categories. Table 4 summarizes all the broad built environment constructs identified within the
body of reviewed literature, the specific built environment features examined, and the findings
regarding the association between these characteristics and walking PA among older adults.

17
Table 4: General and Specific Built Environment Features Studied and Associations with Physical Activity Findings
General Built Specific Built
Environment Environment Physical Activity Related Findings Relevant Studies
Characteristic! Features
High population or residential density consistently associated with increased Adams16, Gallagher43, Siu39,
Residential and
Residential and likelihood of overall, leisure, and transport walking. Positive effect of Tamura37, Troped38
Population
population residential density on walking particularly strong when combined with other
Density
density! walkability features (e.g., land use mix, street connectivity).
Walkability High walkability emerged as a key facilitator of overall PA. High walkability Adams16, Carlson34, Ding31,
index ! particularly supportive for older adults with mobility limitations; supportive for King32
those who lack or have given up driving-status; supportive in helping older
adults overcome perceived barriers to PA.
Walkability Walkability positively associated with walking for both transport and leisure.
Walkability had stronger association with walking for transport compared to
walking for leisure. Interactions affecting leisure walking more related to
walking infrastructure (e.g., access to facilities, aesthetics) as opposed to
macroscale walkability index.
Street Overall street connectivity (e.g., limited cul-de-sacs, short block lengths, Maisel42, Siu39, Tamura37,
Street Network connectivity and presence of four-way intersections) and greater intersection density positively Troped38
Composition intersection associated with walking.
density !
Land use and Increased walking associated with heterogeneous neighborhood destinations Adams16, Bracy33, Buman46,
access to daily (restaurants, entertainment, shops, and services), especially those with Cain35, Carlson34,
destinations! utilitarian, recreational, or social component, such as services (e.g., post office), Chaudhury47,
shops (e.g., grocery store), facilities (e.g., library), and PA facilities (e.g., Chippendale14, Ding31,
recreation center, gym). Gallagher43, King32, Siu39,
Neighborhood
Tamura37, Troped38
Land
Composition
Parks and Abundant opportunities to rest along walking trails or paths helped facilitate Buman46, Chaudhury47,
outdoor park use. Barriers to park use include concerns about lack of safe paths, public Chippendale14! Li41,!Siu39
recreational restrooms, or places to rest; fear of being too isolated or vulnerable to crime.
spaces! Lack of overall perceptions of safety in parks as compared to indoor
recreational facilities or commercial areas.

18
 General Built Specific Built
Environment Environment Physical Activity Related Findings Relevant Studies
Characteristic! Features
Public and Availability of convenient, consistent public transit system encouraged Buman46, Chaudhury47,
motorized neighborhood walking. Presence or availability of motor vehicle parking not Chippendale14!
transport! influential on active living.

Sidewalk Facilitators of PA included sidewalk continuity; presence of smooth, wide, Buman46, Chaudhury47,
presence, spacious, and well-maintained sidewalks; buffer between the sidewalk and Chippendale14,!Gallagher43,
quality, and street; presence of curbs cuts and benches. Sallis40
design! Barriers to PA included uneven or slippery surfaces (e.g., cobblestone, brick,
tile, grates), cracks, obstructions, lack of ramps, insufficient maintenance or
snow/ice removal.
Crosswalk Features supporting PA included crosswalk markings, wide crossings, ramped Buman46, Cain35,
presence, curbs, crossing signals (e.g., walk, countdown, and/or audible signals), and Chaudhury47,
quality, and other pedestrian protections or aids (e.g., protected refuge islands, curb Chippendale14,!Sallis40
Pedestrian design! extensions).
Infrastructure Barriers to PA included inoperable/broken crosswalks and lack of midblock
crossing opportunities.
Presence and Presence of street lights provided support for PA by increasing perceptions of Buman46, Chaudhury47,
quality of street safety and security, improving visibility of sidewalk obstructions, vehicles, Chippendale14,!Sallis40
lighting! cyclists and other non-motorized transport, and reducing fear of falling.
Wayfinding Presence of clear street signs, numbers of buildings, block numbers, advance Marquez48 !
strategies, aids, street signs, clear and logical street labeling conventions (e.g., numbered grid
and cues! system), transit stops, railroad tracks, landmarks, and other individuals to ask
for help were all supportive of engagement in walking in both familiar and
unfamiliar neighborhoods. !
Street safety! Increased perceptions of street safety encouraged PA. Facilitators of street Buman46, Chaudhury47,
safety included elevators for subway access, benches for resting, kneeling Chippendale14,!Li41,!
buses, fences/railings to hold for support, and timed crossings. Maisel42
Safety and
Fear of traffic related injuries was a barrier to PA. Fears regarding street safety
Comfort
were enhanced by busy streets, high traffic volume and speed, unsafe
intersections and crosswalks, dangerous or impatient drivers, poor visibility,
and being struck by non-motorized transport (e.g., cyclists, skateboarders).

19
 General Built Specific Built
Environment Environment Physical Activity Related Findings Relevant Studies
Characteristic! Features
Perceived safety Perception of safety from crime encouraged walking within the neighborhood. Buman46, Chaudhury47,
and crime! Safety perceptions improved by presence of other people/pedestrians. Chippendale14,!Gallagher43,!
Perception of unsafe environment, fears of personal safety, and difficulty with Li41,!Maisel42, Marquez48
wayfinding discouraged engagement in PA and contributed to concerns about
leaving home, especially at night. Perceptions of unsafe environment enhanced
by presence of crime, graffiti, and vandalism.
Aesthetics! Positive aesthetics facilitated engagement in PA, including presence of street Buman46, Carlson34,
trees and well-maintained green spaces; plantings, flowers, or gardens; Chippendale14,!Maisel42,
attractive scenery, views, landscaping, or architecture; settings that promoted Sallis40
Environmental
connectivity to nature; and absence of broken glass, litter, graffiti.
Appeal and
Weather! Pleasant, temperate weather motivated walking. Walking was discouraged by Chippendale14, Gallagher43
Conditions
unpleasant weather or environmental conditions, such as high heat, humidity,
cold, snow, or pollution, as well as seasonal changes in daylight (e.g., early
darkness in winter).
Note: Italics indicate qualitative study.

20
Residential and Population Density
Residential and population density are macroscale measures of the built environment that
emerged as influential variables within the literature. Typically, residential density measures the
concentration dwelling units per square kilometer (or other unit of measurement), whereas
population density captures the average number of persons per square kilometer of an area.
While eight studies in the final set objectively measured either neighborhood residential density
(n=5) or population density (n=3) using GIS methods to analyze census and land use data, three
articles omitted explicit discussion of results relating to this variable. This lack of reporting was
likely due to population/residential density’s role as a component of the overall walkability index
in an area. As such, some studies reported only the impact of walkability on PA behavior, rather
than for each subcomponent of the walkability index. However, five studies did explicitly
address the effect of population or residential density on PA. Adams and colleagues17 found that
residentially dense neighborhoods, when combined with other walkability features (e.g., land use
mix diversity, access to amenities, and street connectivity), were associated with the best overall
PA outcomes, including highest engagement in transport and leisure walking compared to all
neighborhood profiles examined. Conversely, neighborhoods characterized by low residential
density (among other walkability variables), were associated with the poorest PA performance.16
A number of studies looked at gender specific effects of residential density. Gallagher et
al43 examined gender differences in walking behavior among older adults, finding that walking
for transport and total neighborhood walking were influenced by neighborhood density for men,
yet not for women. Despite disparities emerging in walking behaviors by gender in areas of
varying residential densities, this research found no significant gender differences in perception
of neighborhood density.43 Research examining the effects of population density on older women
found that participants were more likely to walk for exercise and transport in neighborhoods
characterized by high population density compared to women living in lower population density
areas.39 Additionally, living in an area of higher population density was significantly associated
with older women who met the DHHS recommended minimum PA levels via walking.38 Tamura
et al’s37 study exploring spatial clusters of PA in Massachusetts and California, found that
population density was significantly associated with areas of high physical activity. Women
living in higher density neighborhoods, especially those in the 90.1-95th and 95.1-100th
percentiles of population density, had significantly greater odds of meeting the DHHS

21
recommended minimum PA levels via walking.37 Among women in these areas of higher
population density, living in a neighborhood that also had high facility density (in particular,
services, cultural and educational, and recreational facilities) resulted in an even higher
likelihood of meeting PA recommendations.37 These results are explained in part by Siu et al’s39
findings that areas of high population density were often supported by characteristics associated
with high walkability, such as high street connectivity and convenient access to amenities,
especially transit and commercial areas. These findings did not hold true, however, in areas with
lower population density and high density of facilities, indicating that a relatively high level of
population density may be needed for facility access to support walking behavior among older
women, as these environments may be low walkable environments lacking in key supportive
infrastructure.37

Walkability
A neighborhood’s walkability is an aggregate of macroscale variables that are commonly
associated with a greater prevalence of PA, such as high residential density, mixed land use, and
a well-connected street pattern (i.e., street connectivity).16 Eight studies used GIS analysis of
census, land use, transportation, and other publicly available place-based data, to develop an
objective score of overall walkability surrounding participants’ place of residence. Other studies
relied on participant self-report of walkability using the Neighborhood Environment Walkability
Scale (NEWS), a validated survey based measure that was adapted for an older adult population
and explores the urban landscape and variables that effect recreation in an individual’s
environment.16 While some research broke down walkability and explored the relationships
between its subcomponents and PA, five research articles analyzed walkability as a summary
measure. Within these findings, two patterns emerged in regard to walkability’s effect on unique
types of walking behavior, as well as on older adults with different limitations: 1. Effect on type
of walking; 2. Effect on those with mobility limitations.
1.! Effect on type of walking
Neighborhood walkability emerged as a key driver of overall PA.16,32,34 Research by
Carlson et al34 exploring GIS-derived walkability indices found positive associations between
neighborhood walkability and accelerometer measured MVPA (p<.01). Additionally, older
adults residing in objectively measured higher walkability areas experienced fewer perceived
barriers to PA engagement.34 Using comparable methods, King et al32 identified similar findings,

22
with older adults living in higher walkability neighborhoods averaging 69.4 minutes/week of
MVPA, compared to 52.2 minutes/week in lower walkability neighborhoods. Although these
numbers are only 17.2 minutes/week apart, they reveal a 33% disparity in PA minutes/week
between high and low walkability neighborhoods. Considering the already low levels of PA
engagement among older adults, this difference is especially concerning, as overtime it could
result in significant disparities in the maintenance of physical abilities, functional independence,
and quality of life.32
King and colleagues32 also found interactions between walkability and neighborhood
income levels. Despite most older adults failing to meet the DHHS recommended >150
minutes/week of MVPA regardless of neighborhood, those residing in higher walkability/higher
income neighborhoods had a significantly greater likelihood of meeting recommendations
(17.8%), as to compared high walkability/low income neighborhoods (8.6%).32 However, when
comparing neighborhoods of similar income levels, researchers found significant differences
between high and low walkability. Among high income neighborhoods, high walkability was
associated with a 6.5% higher rate of residents meeting PA guidelines compared to residents in
high income but low walkability neighborhoods (17.8% of residents compared to 11.2%).32 In
lower income neighborhoods a similar effect was observed, with 4.7% of residents meeting
guidelines in high compared to low walkability neighborhoods (8.6% of residents compared to
3.9%).32 Regardless of these disparities, most older adults engaged in insufficient amounts
exercise. Even in high walkability areas, participants averaged less than 20 minutes/week,
highlighting that walkability alone is not a sufficiently powerful motivator for PA among this
population.32
In addition to having an impact on walking and overall MVPA, neighborhoods
characterized by high walkability were significantly associated with the most favorable outcomes
in terms of transport and leisure PA.16 For example, Carlson et al33 examined associations
between objectively measured walkability and subjectively defined barriers to PA, separately and
jointly. They found that individuals living in high walkability neighborhoods who had few
perceived environmental barriers to PA engaged in 58 minutes/week of transport walking.34 By
comparison, those living in neighborhoods of low walkability with high perceived barriers to PA
spent on average almost 50% fewer minutes walking for transport, with only 28 minutes/week.34
This finding was corroborated by King and colleagues, who found a significant, positive

23
relationship (p<.001) between objectively measured neighborhood walkability and the average
number of minutes/week spent walking for transport.32 Notably, interactions related to walking
for transport were more related to walkability, whereas interactions related to leisure walking
tended to be related to walking infrastructure, such as sidewalk quality, access to parks, or
neighborhood aesthetics.34 Despite this finding, research exploring GIS-measured walkability
found a positive association between neighborhood walkability and walking for leisure (p<.05),
highlighting the positive influence walkability can have across all types of walking behavior.34
2.! Effect on older adults with mobility limitations
The effect of neighborhood walkability held its influence even among older adults with
different types of limitations, such as physical impairment or driving status.31,32 In a multi-city
study, King and colleagues31 found that regardless of physical limitations, older adults living in
lower walkability neighborhoods had low rates of walking for transport (e.g. to complete errands
or travel outside the home). However, in comparison, older adults across all physical ability
levels exhibited a greater range of PA behavior in high walkability neighborhoods. Notably, the
most mobility impaired older adults living in high walkability neighborhoods reported transport
activity similar to those who were less impaired but living in a low walkable area.32 These
findings suggest that the incorporation of walkable design may allow residents across the
spectrum of mobility impairment to optimize their ability to participate in walking transport
activity.32 This finding underscores the ways in which the built environment can support all older
adults (regardless of mobility level) in increasing PA. Even brief amounts PA, such as short
errands within the neighborhood are significant, as they helps facilitate and support the continued
independence necessary for successful aging-in-place and maintained quality of life.32
Ding et al35 examined neighborhood walkability effects on PA by driving status. Among
older adults who differed based on driving status (i.e., current driver or non-driver), attributes of
the neighborhood environment exhibited similar influence on transport walking.36 This finding
highlights the importance of walkable, activity-friendly neighborhoods to encourage PA and
facilitate walking among all older adults.36 An interesting distinction arose in regards to where
driving and non-driving older adults chose to live: non-driving older adults were significantly
more likely to live in neighborhoods with activity-friendly environmental features compared to
their driving counterparts.36 However, because this finding arose from cross-sectional research,
temporal order cannot be determined between driving status and residing in a highly walkable

24
neighborhood. It is not known, therefore, whether older adults gave up driving and subsequently
relocated to a walkable neighborhoods, or if due to living in a pedestrian-friendly, non-
automobile reliant environment, they stopped driving.36

Street Network Composition

The street network composition, and in particular, street connectivity and intersection
density, emerged as a link between the neighborhood environment and PA behavior among older
adults. Street connectivity describes the ease of travel between different points based on street
design, such as limited cul-de-sacs, short block lengths, or the presence of four-way
intersections.42 Notably, street connectivity is among the variables that commonly make up GIS-
measured walkability. Similar to other variables that comprised walkability in this review (e.g.,
residential/population density, land use mix), street connectivity was not consistently discussed
as a unique variable, but rather, was analyzed as part of overall walkability.
Despite this limitation, street connectivity was discussed in the set of reviewed articles.
Maisel42 explored the relationship between perceptions of street connectivity and PA, finding a
small, positive correlation (r=.25; p<.01) between older adults with positive perceptions of street
connectivity and job walking (i.e., walking that occurs while engaging in paid or unpaid work,
such as volunteering), transportation walking, and total weekly walking. Additionally, research
looking at the influence of objectively-measured street connectivity on PA behavior of older
women found that those living in neighborhoods characterized by high street connectivity were
more likely to walk for leisure and transport.39 Similar results were found in a study by Tamura
et al37 comparing measurements of intersection density with walking behavior. Study results
indicate that that women living in neighborhoods with a density in the range of >2-11
intersections/km had an 18% greater odds of meeting DHHS daily recommendations for PA
(>150 minutes/week) compared to neighborhoods with <2 intersections/km.37

Neighborhood Land Composition

Neighborhood land use, and in particular, the presence of diverse destinations and amenities, was
referenced throughout the studies reviewed as a driver of PA. Three specific features of land use
composition emerged as important: 1. Land use and access to daily destinations; 2. Parks and
outdoor recreational spaces; 3. Public transit and motorized transport.

25
1.! Land use and access to daily destinations
Neighborhood land use and the availability of non-residential, daily destinations emerged
as a key facilitator of older adults’ PA behavior. Land use mix, a GIS-measured component of an
area’s overall walkability index, was frequently references in the literature, however, similar to
other walkability measures it was less often analyzed as a unique variable.16,31-35,39 Cain and
colleagues35 were among the few studies that described land use mix as a stand-alone variable,
finding that it significantly influenced older adults’ engagement in objectively measured
MVPA.35 More commonly, however, the studies in this review focused on the specific types and
density of destinations in a neighborhood, both of which played a key role as facilitators of PA.
Cain et al35 found a positive relationship between that the presence of heterogeneous
neighborhood destinations and walking behavior, especially restaurants, entertainment, shops,
and services (e.g., banks). It is likely that these destinations elicited positive results due to their
role in providing a specific motivation to leave one’s residence and thus offering a context for
PA engagement.23,35
Multiple studies explored the effects of destinations on women only. A study from Siu
and colleagues39 looked at the relationship between different urban forms and walking among
older women.39 Neighborhoods characterized as urban fringe with poor commercial access
ranked lowest among six different types of urban forms in terms of median number of daily
blocks walked.39 In contrast, older women living in high walkable neighborhoods with
convenient access to amenities (especially public transit and commercial areas), were most likely
to walk for both exercise and transport.39 Across all neighborhood types, access to any
destination, especially commercial areas, was beneficial in terms of PA engagement.39
Additional research conducted specifically with older women supports the exercise-promoting
nature of facilities in the neighborhood. Troped et al38 found that facility density was
significantly associated with greater odds of meeting DHHS PA recommendations for walking,
even when adjusted for the participant’s age, race, ethnicity, education, BMI, walking
limitations, smoking status, preference for staying indoors, and number of years at current
address (an indication of neighborhood familiarity). Among the eight facility types examined
(retail, services, cultural/education, physical activity, restaurants, fast-food, grocery stores,
convenience stores), the presence of service (e.g., post office) and/or physical activity (e.g., gym
or recreation center) facilities was most associated with meeting PA recommendations.38

26
Services may provide a particular motivation for PA, as each additional service facility/km of
road was associated with a 53% higher odds of meeting PA recommendations.38 Density of PA
facilities was associated with a 91% greater odds of meeting walking recommendations when
adjusted for age, however, this finding was greatly attenuated when adjusted for all previously
listed potential confounders.38 This finding is likely due to the financial barriers associated with
access or membership to PA facilities.
Other research supports the value of facility density for older women: Tamura et al36
found that among women living in areas with high population density, density of services, as
well as cultural, educational, and PA facilities had the strongest, positive relationship with
walking outcomes.37 Additional findings compliment this research. Using the NEWS self-report
measure of neighborhood environment, Gallagher and colleagues40 investigated the impact of
destinations on walking behavior by gender.43 For women, reporting destinations within walking
distance was associated with both increased walking for transportation and total neighborhood
walking, with a friend’s house cited as the most common destination.43 Men, however, were
slightly more likely to report the presence of destinations within walking distance compared to
women.43
Findings from the qualitative research reviewed provides additional evidence for the PA
facilitating role of amenities and destinations. Buman et al’s study46 of individuals across five
low-income housing sites found supermajority consensus (≥67% of residents referring to a built
environment variable) regarding the importance of attractive amenities and destinations (e.g.,
shops, restaurants, and public services) in supporting active living. Using photovoice methods to
document neighborhood-based facilitators and barriers to PA, Chaudhury and colleagues47 found
that accessible amenities with either a utilitarian, recreational, or social component (e.g., bank,
grocery store, post office, mall, library, gym, recreation center) were most supportive of older
adult PA.47 Other qualitative research using semi-structured interviews found that stores for
browsing or shopping, street fairs, and farmers markets were all activity promoting
destinations.14 Additionally, for some older adults, the incentive of potential social interaction
(either informal or formal) at a destination served as a motivator for PA.14
2.! Parks and outdoor recreational spaces
The quantitative studies reviewed presented mixed support for park access as a facilitator
of PA. Both Maisel41 and Siu et al39 found that access to parks can encourage PA among older

27
adults. However, the same study from Siu and colleagues39 found that park access was not as
impactful on PA behavior compared to other amenities. For example, in urban fringe areas, daily
walking was more positively impacted by access to commercial areas compared to parks,
suggesting that living close to a commercial area might play a stronger role in promoting or
encouraging walking compared to proximity to a park.39
Across the qualitative studies reviewed, green spaces, public gardens, urban trails, and
walking paths emerged as a PA promoting destinations.14,46 Older adults described suitable, safe
walking paths,14,46 walking trails along blue spaces (e.g., river, lake),14 and ample opportunities
to rest14 as key facilitators to PA. A cited barrier to park utilization was poor access, and older
adults in higher density areas may rely on other neighborhood-based outlets for PA engagement
instead of seeking out parks.47 Interestingly, those residing in low density areas more commonly
referenced seeking out parks or beaches to engage in PA compared to higher density areas,
possibly due to the less supportive walking infrastructure associated with low density
environments.47
3.! Public and motorized transit
The specific issue of public transportation was explicitly discussed only by the qualitative
studies. Within the qualitative studies reviewed, public transportation emerged as a present but
not commonly cited variable influencing PA engagement as compared to shops, services, parks,
or other amenities. Overall, residing in higher density neighborhoods was associated with a more
positive outlook on the public transit system, and when perceived as accessible and convenient,
public transit was described as a PA motivating amenity in the community.47 Older adults
reported that the availability of public transit as an alternative form of mobility encouraged PA,
as it served a backup/safety-net in case of becoming tired on longer walks.14 The same study
reported that some older adults found unreliable public transit to also be a facilitator of PA, as
lack of reliability encouraged walking to a destination rather than riding transit.14 While the
presence of transit generally encouraged PA, features of the transit stations themselves were
cited as barriers, such as steps down to the subway, the absence of an elevator or escalator, and
concerns about wet floors, open grates, or tiled walkways.14
Older adults living in more residentially dense neighborhoods discussed that it was often
more convenient to walk or use public transit rather than drive, as compared to lower density
areas where car ownership was both more common and less burdensome.47 Similarly,

28
participants across three different neighborhoods in Buman et al’s46 study agreed that access to
parking was not an influential variable on active living, underscoring older adults’ preference for
engaging in physical activity within the confines of their neighborhood, rather than needing to
drive elsewhere to do so.

Pedestrian Infrastructure
Pedestrian infrastructure, a microscale characteristic of the built environment, played a
key role in encouraging PA across studies, with four specific features emerging: 1. Sidewalk
presence, quality, and design; 2. Street crossing amenities and support; 3. Presence and quality
of street lighting; 4. Wayfinding strategies, aids, and cues.
1.! Sidewalk presence, quality, and design
Sidewalk presence, quality, and design addresses the presence, continuity, and
maintenance of the sidewalk network throughout a neighborhood, as well as more granular
aspects of sidewalk design, such as ramped curbs, surface quality, and width. Across multiple
cross-sectional studies, significant associations were found between walking for transport or
leisure and different sidewalk variables, including: sidewalk presence or absence,40,43,47 specific
design elements such as presence of a buffer between the sidewalk and street, curbs cuts, and
benches,40 and sufficient sidewalk width and continuity (i.e., lack of abrupt endings).47 Among
older women, Gallagher and colleagues43 found that the perception of one’s ability to overcome
barriers to PA (including the ability to navigate obstacles on the sidewalk and maintain balance)
significantly influenced confidence in the ability to walk throughout the neighborhood for
sustained periods of time.
Sidewalk presence emerged as a common theme within the qualitative studies reviewed,
and there were multiple areas of convergence between the qualitative and quantitative findings.
Poor sidewalk quality and design was repeatedly referred to as a barrier to PA, and common
barriers included uneven surfaces or slippery (e.g., cobblestones, brick), cracks in the sidewalk,
and other obstructions that could be tripping hazards.14,46,47 A lack of suitable sidewalk
infrastructure for those with a wheelchair, walker, or other mobility disability (e.g., curb cuts or
ramps) was raised as a concern.46 Chippendale and Boltz14 examined the relationship between
the built environment, fear of falling (FOF), and PA behavior in qualitative interviews with older
adults in three urban neighborhoods. Participants identified numerous structural factors that

29
impacted FOF, which in turn, shaped PA engagement, including: uneven walking surfaces on
sidewalks (e.g., uneven grates, brick surfaces, embedded decorative stones), inadequate
maintenance (e.g., cracked sidewalks, potholes, pools of rainwater, delayed snow and ice
clearance), presence of curbs, and sidewalk obstructions.14
In contrast, several studies identified facilitators to PA, such as paved, flat, smooth, wide,
continuous, and well-maintained walking surfaces, as well as curb cuts, convenient routes, and
benches spaced consistently throughout the neighborhood.14,46 In particular, seating or benches
were viewed as a valuable but often limited resource. Having a safe place to rest was referred to
as a key feature that helped create a more comfortable and supportive environment, especially
given the mobility needs and limitations of many older adults.47,14 Other sidewalk features that
were emphasized as facilitators included the importance of railings, handrails, ramps, safe stairs,
and water fountains, and that blocked access to handrails created a barrier to safely navigating
stairs or steep areas in parks or other areas of neighborhood.14,47

2.! Street crossing amenities and support

Due to age-related changes in physical, auditory, and visual acuity, older adults are more
likely to experience difficulty in safely navigating street crossings.18 This review revealed that
the presence of high-quality street crossings serves as an important environmental facilitator of
PA.35,40 Cain et al’s35 cross-sectional research identified a significant relationship between
objectively measured PA and positive crossing characteristics, such as crosswalk markings, wide
crossings, curb cuts, crossing signals (e.g., walk, countdown, and/or audible signals), and other
pedestrian protections or aids (e.g., protected refuge islands, curb extensions). These findings
were supported by research from Sallis and colleagues40, which identified a significant
relationship between active transport (i.e., walking as a form of transportation) and the presence
of microscale crossing amenities, such as crosswalks, curb cuts, and crossing signals.
Qualitatively, older adults reported crosswalk limitations, such as inoperable or broken
signals and motorists not stopping at the crosswalk, to be a major barrier to PA engagement.46
Additionally, on longer blocks, older adults reported a lack of midblock crossing opportunities to
be a barrier to engaging in PA within the neighborhood.14 Conversely, the presence of safe
crosswalks was reported as a common facilitator of PA.46 Concerns about the ability to safely
cross streets emerged repeatedly across both the quantitative and qualitative literature,14,35,40,46,47
and the findings in this review suggest that negative or a lack of crossing amenities can have

30
serious and detrimental effects on PA engagement. Conversely, the presence of positive crossing
characteristics may provide relief from crossing anxieties or fears, and as such encourage and
facilitate increased engagement in PA.
3.! Presence and quality of street lighting
Street lights play a critical role in cultivating feelings of safety and security during
activity that takes place at dawn, dusk, or nighttime.40 Lighting was not commonly addressed
across the quantitative literature, with the exception of Sallis and colleagues40, who found
significant associations between walking for transport and the presence of street lights.
Qualitatively, however, the presence of lighting mitigated major concerns affecting older adult
PA engagement, including reducing fears of falling and increasing visibility of cyclists and
sidewalk obstructions.14,47 An exception to these findings was one qualitative study that did not
identify lighting conditions as an important facilitator to PA.46 However, this may have been a
reflection of study methods, as data collection using photovoice was conducted exclusively in
daytime, thus potentially biasing responses.46 The researchers do suggest that older adults may
simply avoid going out in the dark, and that well-lit streets might not be a powerful enough
facilitator to encourage evening PA.46
4.! Wayfinding strategies, aids, and cues
Supportive wayfinding is key to enabling older adults with a range of functional,
cognitive, and sensory abilities to remain mobile and engaged in the community. Despite its
value, wayfinding was explicitly addressed by only one study in this review.48 Using exploratory
mixed-methods, Marquez et al48 looked specifically at the wayfinding strategies older adults
employ when walking in their neighborhood, and the ways in which the presence of wayfinding
aids impact PA engagement. Infrastructure that was identified as a wayfinding aid or cue
included the presence of clear street signs, numbers on buildings, block numbers, advance street
signs, clear and logical street labeling conventions (e.g., numbered grid system), transit stops,
railroad tracks, and landmarks.48 Landmarks, such as a cathedral or prominent community
building, were the most frequently cited facilitator of wayfinding, as they provided orientation
and a frame of reference.48 In terms of determining a safe walking route and general wayfinding
support, older adults commonly employed the help of known individuals such as friends, as well
as highly regarded trusted officials, such as transit workers or police officers.48 Major barriers to
wayfinding included a lack of landmarks, missing street signs, confusing street naming systems

31
or alignments, and distractions (e.g., acoustic distractions, such as loud train noises). Notably,
few older adults identified mobile phones, GPS, or other technological resources as sources of
wayfinding information.48 This finding underscores the need for supportive infrastructure to help
non-technology oriented older adults engage in active living by navigating the community with
comfort and ease.

Safety
Across both the quantitative and qualitative literature, perceptions of safety within the
community arose as a common theme that shaped walking behavior among older adults, with
two main safety-related constructs: 1. Street safety; 2. Perceived safety and crime.
1.! Street safety
Concerns about and perceptions of street or traffic safety emerged as variables that
influenced PA behavior. Quantitatively, Maisel’s42 cross-sectional study found that perceptions
of traffic safety were correlated with both job walking and total weekly walking, and that greater
safety concerns were associated with reduced PA engagement. The impact of street safety
concerns on PA engagement surfaced more frequently across the qualitative literature, including
personal safety fears (e.g., being hit by a vehicle), busy streets, high traffic volume and speed,
unsafe intersections and crosswalks, dangerous or impatient drivers, drivers ignoring traffic laws,
and poor visibility.47,46 Additionally, difficulty or dangers securing a taxi at night was discussed
as a concern that prevented older adults from wanting to go out in the evening.14 Neighborhoods
with higher population density elicited a greater volume of concerns, an expected finding given
the heavier traffic patterns found in dense neighborhoods.47 Outside of cars, getting struck by
cyclists, skateboarders, and other forms of non-motorized transport was a concern among older
adults, with some expressing more worry about cyclists than motorists.14 Certain aspects of the
social environment promoted feelings of increased safety, such as communication from drivers
that it was safe to cross (e.g., waving to cross the street) or the presence of crossing guards and
other officials.14 The presence of select built structures also increased feelings of safety,
including: elevators to access the subway, benches for resting, kneeling buses, fences or railings
to hold for support, timed crossings that indicated the amount of time remaining to cross, and
flashing crosswalks that calmed traffic and made older adults feel more secure in crossing streets
without signaled crossings.47,14

32
2.! Perceived safety and crime
The influence of perceptions of safety and crime was explored across both the
quantitative and qualitative studies. Maisel42 found that older adults with more positive
perceptions of crime safety reported greater total weekly walking compared to those with more
negative perceptions. Additionally, concerns about crime safety were correlated (however,
weakly) with both recreational and total weekly walking.42 Gallagher and colleagues43 found that
perception of neighborhood crime significantly influenced walking duration. Li et al’s41 study of
walking behaviors among older Asian adults found that perceptions of safety were significantly
related to walking, most notably among Filipino-American older adults.
Perceptions of safety were commonly discussed in the qualitative literature, with older
adults expressing a preference for crime free environments that cultivated a strong sense of
physical safety and psychological security.47,46 Higher density neighborhoods provoked greater
feelings of negativity in relation to safety and security, with older adults reporting issues such as
feeling unsafe, as well as concerned about the presence of crime, graffiti, and vandalism.47 Safety
and vulnerability also emerged as a factor influencing the ability to orient oneself and feel
comfortable walking around an unfamiliar place.48 Some older adults described increased
perceptions of safety corresponding to the presence of other people.14 For example, busy, rather
than isolated, neighborhoods encouraged PA by not only enhancing perceptions of safety, but by
also helping older adults feel more secure in navigating the environment, as it increased the
likelihood that trustworthy, helpful individuals may be able to assist them in case they became
lost or disoriented.48 Fear of personal safety was commonly reported as a barrier to leaving home
in the evening, as some older adults felt particularly vulnerable and feared they may be targeted
due to their older appearance.14

Environmental Conditions and Appeal

The sensory appeal, perceived attractiveness, and environmental conditions of the neighborhood
were commonly explored as variables influencing PA behavior. Two main constructs surfaced
within the reviewed studies, 1. Aesthetics; 2. Weather and environmental conditions.
1.! Aesthetics
Quantitatively, interactions emerged between objectively-measured PA, aesthetics (e.g.,
presence of street trees, attractive landscaping, views, buildings, or homes), barriers to PA (e.g.,
discomfort when walking, time constraints), and social support (e.g., having family or friends to

33
walk with or provide encouragement to do PA).34 Among participants with few self-reported
barriers to PA, high ratings of aesthetics accounted for 30 additional min/week of MVPA
compared to low ratings for aesthetics.34 For participants with more barriers, high compared to
low ratings of aesthetics accounted for 18 fewer min/week of MVPA.34 Additionally, the
combinations of low aesthetics/low social support and high esthetics/high social support were
found to be explanatory variable combinations for the minutes per week older adults spent
walking for leisure.34 Supporting this finding, Maisel42 found that aesthetics were significantly
correlated with recreation (i.e., leisure) walking, however, they were unrelated to job or
transportation walking. Similarly, Sallis et al40 concluded that both aesthetics and social
characteristics were largely unrelated to active transport in older adults. Aesthetics surfaced as an
influential variable in the qualitative literature as well. Positive aesthetics, such as the presence
of trees, flowers, and attractive scenery, were cited as common facilitators to PA.46 Similarly,
appealing architecture, well maintained green spaces, water views, and settings that promote
connectivity to nature were all mentioned as factors that help facilitate walking and exercise
among older adults.14
2.! Weather and environmental conditions
Weather as an environmental condition was mentioned as a negative influence and barrier
to engagement in PA. The perception of weather as a barrier to PA, as well as a variable shaping
an individual’s self-efficacy to engage in PA engagement, was found to be significant, especially
among older women.43 When faced with perceived environmental barriers, including inclement
weather, older women were significantly less confident in their ability to walk in their
neighborhood both overall and for increasing durations, as compared to older men.43 However,
older men with low self-efficacy for PA engagement were less likely to report walking in the
face of similar environmental barriers.43 Additionally, within the qualitative literature,
insufficient sidewalk and street maintenance related to weather (e.g., lack of snow, ice, or slush
removal, pools of rainwater on sidewalks or street corners) was cited as a key barrier to PA
engagement.14

34
 DISCUSSION

Synthesis of Results
The studies included in this review further the evidence regarding the ways in which
walking behaviors of older adults are shaped by the built environment and in particular, the
neighborhood surrounding one’s place of residence. The seventeen studies reviewed revealed
seven major built environment features associated with older adult’s physical activity: (1)
Residential and Population Density, (2) Walkability, (3) Street Network Composition, (4)
Neighborhood Land Composition, (5) Pedestrian Infrastructure, (6) Safety, and (7)
Environmental Appeal and Conditions. Within each major construct, various sub-constructs
provided more granular level details regarding the built environment and its impact on physical
activity. Overall, this review supports the previously identified relationship between the micro
and macroscale features that commonly comprise high walkable neighborhoods, such as high
population density, land use mix, street connectivity, aesthetics, and perceptions of safety, and
overall engagement in PA.49 All studies included in this review referred to all or a combination
of macro and microscale features, with findings suggesting that variables in both categories play
an influential role in shaping PA behavior.
Many findings from this review were corroborated by past reviews investigating the
relationship between older adult PA and the built environment. In regards to street connectivity,
findings are consistent with evidence summarized in a prior review by Haselwandter and
colleagues7, who found that street density and connectivity was significantly associated with the
number of blocks walked among women ≥65 years. Similarly, multiple reviews support findings
associated to land use mix. Quantitative studies included in reviews by Haselwander et al7 and
Yen et al23 indicate positive associations between PA and neighborhood access to commercial
and retail businesses,23 grocery stores,23 and recreation facilities (e.g., indoor gyms, facilities
with treadmills).7 Moran et al’s5 systematic review of qualitative literature supports these
findings as well, with access to daily destinations, such as shops (e.g., grocery store), services
(e.g., post office), and other facilities (e.g., library, senior center) emerging as a key theme across
literature reviewed from 1996-2012.
In regards to access to parks and outdoor recreational spaces, past reviews supports the
mixed findings in this present review. Earlier studies reviewed by Haselwandter et al7 provide
support for park proximity as a facilitator for PA. However, Kerr and colleagues6 note that park

35
proximity and/or density has shown an inconsistent association with PA in past literature, and of
the recent studies included in their review, two found positive associations, while four noted a
lack of association. A possible interpretation of these findings is that many parks may be
inappropriate or unappealing destinations for older adults due to a lack of safe paths, public
restrooms, or places for rest.6 Additionally, older adults may avoid going to parks that are not
consistently trafficked by others, as the isolation may increase perceptions of vulnerability to
crime or injury in the case of a fall.6 In comparison, recreation facilities were cited as a preferred,
safe, and more supportive environment for PA.6 The review by Moran et al5 supports these
mixed perspectives on parks, with findings indicating that although parks can serve as motivators
for PA, older adults were also averse to isolated parks or trails due to low visibility and increased
fear of crime, and generally preferred PA engagement in commercial areas or recreational
facilities.
The value of microscale features of pedestrian infrastructure in facilitating walking
among older adults is supported by earlier literature. Moran and colleagues5 found that the
presence, usability, and comfort (i.e., sheltered from inclement weather) of benches and seating
areas played a role in facilitating PA within qualitative studies reviewed.5 Additionally, Kerr et
al’s6 narrative review discussed the value provided by street crossing amenities, the challenges
posed from insufficient presence of crossing aids (e.g., unsignaled intersections, large crossing
distances), and the impact these features have on PA behavior. Notably, however, because of the
relatively new focus on microscale features in this area of research, past findings in regards to
the microscale environment were not commonly addressed in review articles of earlier literature.
Older adult concerns about safety, as well as the effect of these concerns on
neighborhood walking, were also substantiated in prior studies summarized in past reviews. In
terms of street safety, the presence of sidewalks and light rather than heavy traffic has been
previously associated with increased perceptions of safety and greater engagement in PA in older
adults.7 The association between perceived safety from crime and older adult mobility is also
supported by prior studies included in systematic reviews and summary discussions of the
relevant literature.5-7,23 Yen et al23 commented that while mixed land use may provide an
important contextual influence for PA, its presence alone may not be a sufficient motivator for
PA if the perception of safety and security within the environment is insufficient. Certain stores
such as retail liquor sales, as well as the presence (or absence) of other pedestrians may enhance

36
the perception of high neighborhood crime, and in turn, prevent older adults from engaging in
outdoor activity.23 Kerr and colleagues6 offered an interesting perspective: the authors surmise
that safety influences walking depending on the purpose of the activity. For example, regardless
of perceptions of safety, older adults may walk for transport due to necessity, cost, or
convenience.6 In contrast, perceived safety of the environment may significantly influence
leisure and recreational walking, thereby underscoring the role of perceptions of safety and
security as an influential mechanism in older adults’ decisions about PA engagement.6,23
This review’s findings regarding the importance of positive aesthetics, weather, and
environmental conditions were also consistent with findings based on earlier studies. Previous
research has found links between increased older adult mobility and the perceived attractiveness
of the neighborhood, an absence of negative features (e.g., litter, broken glass, graffiti) and an
abundance of positive characteristics, (e.g., trees, plantings, gardens, or other greenery, pleasant
scenery or views, and attractive architecture) – all of which were found to affect PA among older
adults in the more recent studies reviewed in the present paper.5-7,22,23 Prior studies also support
and complement findings regarding the challenges of weather, noting inclement (e.g., rain, snow)
or extreme (temperature, humidity, sun exposure) weather, seasonal challenges (e.g., early
darkness), and environmental quality (e.g., polluted areas, traffic exhaust and noise) as serious
barriers to PA engagement.5,7
Overall, the thirteen quantitative and four qualitative studies included in this review
offered unique contributions and findings. The qualitative research in particular added greater
depth to the research on microscale features, with each of the four included studies exploring the
subtle aspects of the built environment that shape older adult PA perceptions and behaviors.
Unearthing the nuances regarding how even small variables influence this population is
instrumental to effective design and implementation of built environment interventions. While
features such as well-maintained sidewalks or consistent curb cuts may seem minute, this
population experiences unique barriers to and fears regarding PA, and features that encourage
even incremental improvements in regular walking can have major benefits for aging
individuals, especially considering the widespread inactivity among this population.42 Benches,
for example, need to be not only present, but consistently spaced, abundant (especially in areas
with hills), and usable for older adults (e.g., easy to sit on, sheltered).5 Additionally, many
microscale characteristics can be modified at lower costs and in a shorter time-frame (e.g.,

37
repairing sidewalks, improving street crossing design) as compared to the major costs and
barriers associated with reconfiguration of macroscale features.35
Within the quantitative literature, the greater proportion of studies focused on the
macroscale-oriented features of the built environment, such as walkability, land use mix, and
street connectivity. The lack of quantitative focus on microscale features was likely due to three
factors: (1) limitations in the ability of GIS to measure microscale features, (2) the high cost of
quantitatively assessing microscale features of the built environment through environmental
audits, and (3) the relatively new nature of microscale-focused research, as it has not been
studied to the same depth as macroscale features. For example, the Microscale Audit of
Pedestrian Streetscapes (MAPS), a leading in-person audit tool designed for fine-grained
examination of the built environment, was released only in 2012.50 Two studies included in this
review made use of the MAPS tool (Cain et al; Sallis et al),35,40 both of which provided
quantitative substantiation to the extensive discussion of microscale features apparent in the
qualitative literature. The addition of the MAPS tool also highlights a promising methodological
finding from this review. Previous reviews on this topic have suggested that a limitation of this
area of research was the lack of consistent and validated measures that allowed for cross-study
comparisons.51 Findings from this review suggests that researchers in this field are beginning to
employ comparable, validated built environment and PA measurement scales and tools, such as
MAPS35,40, NEWS16,42,43, CHAMPS16,31-35,40, GIS32,37-39, and Actigraph accelerometers.16,31-35,40
Research conducted prior to this review suggests that older adults engage in two primary
types of PA, walking for transportation (e.g., for errands) and leisure (e.g., for recreation), and
that distinct factors of the built environment shape engagement in each type of walking.45 The
findings of this review further support this claim, with the reviewed literature revealing unique
environmental qualities that encourage each type of walking. The variable most prominently
impacting transport walking was overall walkability, suggesting that the basic necessities of
walkable environments may be sufficient in encouraging active transport among older adults. For
example, the well-connected streets, mixed land use, and high residential density needed to
support a walkable neighborhood may cultivate an environment in which it is possible or even
more convenient to engage in active transport. These findings were supported by a recent
systematic review and meta-analysis of studies in multiple countries exploring neighborhood
correlates with active transport.52 From this review, Cerin and colleagues52 concluded that there

38
was very strong and consistent evidence supporting a positive association between objectively
measured walkability and walking for transport.
Leisure walking, however, was most influenced by aesthetics, amenities, safety, and
access to facilities. This contrast between leisure and transport walking underscores the influence
of a pleasant and comfortable atmosphere on older adult decisions to engage in PA for pleasure
rather than necessity or errands. Even if a neighborhood is highly walkable in the conventional,
macroscale-defined sense, older adults may avoid non-necessary walking if the environment
makes them feel unsafe or vulnerable (i.e., high perceived crime, broken sidewalks, insufficient
crossing amenities, inadequate snow or ice removal), or is aesthetically unappealing (i.e., litter,
graffiti, lack of street trees or plantings). Kerr and colleagues6 postulate that this is likely because
transport walking frequently occurs out of convenience, cost, or necessity, whereas leisure
walking is typical carried out as pleasurable activity. Despite these findings, walkable
environments that encourage transport-oriented PA are still of high value, as they may help
facilitate the daily, routine forms of walking that typically require less of the planning,
scheduling, and motivation required for leisure activity engagement.32
Overall, the findings from recent studies included in the present review, as well as results
from earlier studies conducted in the U.S. and other countries, suggest that although numerous
areas of possible intervention may support older adult PA, in order to achieve the greatest effect
across all domains of PA, improvements need to be made to both macro- and micro-oriented
features. For example, the presence of amenities alone is not sufficient to encourage PA, as it is
critical to also consider how older adults can actively and safely travel to these destinations (e.g.,
are the sidewalks high quality and continuous? Can older adults cross the street safely? Are the
streets well connected and easy to navigate?). The results of this review also highlight how
environmental interventions should focus on addressing the unique needs of older adults, and
that changes to the built environment need to focus on variables that will most encourage and
motivate PA given these particular needs. For example, prioritizing safety concerns within the
neighborhood through traffic calming measures (e.g., reducing speed limits in pedestrian areas),
increasing time allocated to pedestrians at signalized crossings, installing pedestrian refuge
islands on wide streets, and repairing broken curbs, may cumulatively have a stronger effect on
supporting older adult PA compared to adding bike paths or providing parking at a local
park.18,42

39
Gaps in the Literature and Directions for Future Research
Based on this review’s assessment of the past and current state of the literature, it is clear
that the study of the relationship between the physical environment and PA in older adults is on
the rise. The first review on this topic, published in 2004, included only six studies all based on
self-report data.53 Over the past decade, the research on this topic has become increasingly robust
and sophisticated as PA tracking technology through accelerometers and geospatial mapping
programs have advanced. For example, a 2012 review of studies published from 2000-2010
found that only 2/33 articles (7%) used accelerometers.6 By comparison, 7/17 studies (41%) in
this review of studies published 2011-2016 used accelerometers, highlighting how tracking
technology has become more integrated into this area of public health research. Despite these
advances, this review identified a few noteworthy gaps in the literature and identified areas in
need of future study: (1) Participant diversity; (2) Geographic diversity; (3) Age representation;
(4) Wayfinding; (5) Climate change; (6) Naturally occurring retirement communities.
1. Participant diversity
On average, study participants in this review represents a relatively narrow demographic
profile of majority white and higher-resourced older adults. Notably, seven studies obtained their
sample from the existing Senior Neighborhood Quality of Life Study (SNQLS), an observational
study that explored the relationship between neighborhood environments and health outcomes
among ambulatory, community-dwelling older adults (≥66 years). Participants in SNQLS were
sampled from neighborhoods stratified based on GIS-measured walkability (i.e., residential
density, land use mix, intersection density, and retail floor area ratio) and median household
income in two regions (Baltimore, MD-Washington D.C. and Seattle-King County, WA).16
Although each study using the SNQLS sample had unique objectives, research aims, and
contributions to this review, the use of the same sample and study locations may have skewed
this review’s findings. While SNQLS intended to recruit a socio-economically diverse audience,
it was comprised of a sample that was both less diverse and more educated than the U.S. older
adult population. This overall lack of sample diversity indicates a need for adjustments in the
recruitment techniques of future research, and efforts should be made to prioritize recruitment or
oversample underrepresented groups. Additionally, racial, ethnic, and socioeconomic disparities
seen in PA engagement necessitates further exploration of populations at higher risk for

40
sedentary behavior, such as non-Hispanic black and Hispanic older adults, and older adults with
less than a college education.13
A few studies in this review did deviate from the typical sample profile, such as Li and
colleague’s41 exploration of correlates of the neighborhood environment with walking among
older Asian Americans. Their findings suggest that older Asian Americans not only walk more
than their white counterparts, but that neighborhood effects varied significantly across different
Asian subgroups (e.g., Chinese, Korean, Filipino).41 Given these findings, greater consideration
needs to be taken regarding the influence of older adults’ cultural, racial, or ethnic background
on PA behavior, especially among non-white and more socioeconomically diverse populations.
In regards to gender, three of the seventeen studies included in the final study set focused
on women only,37-39 one explicitly explored gender effects,43 and none examined men only.
While many studies had approximately equal distributions by gender, proportions tended to skew
more heavily towards women. The emphasis on older women was not surprising, as past research
has tended to focus more on women due to their longer life expectancy, greater likelihood of
living alone and independently, and increased sensitivity and vulnerability to neighborhood
environment features.43 Future research is needed, however, to better tease out gender-specific
differences and patterns in walking behavior. For example, Gallagher et al43 identified gender
differences between older men and women in regards to self-efficacy and perceived barriers for
neighborhood walking, however, a clearer understanding is needed in order to effectively tailor
physical activity interventions by gender.
2. Geographic diversity
The majority of research included in this review was located in the Northeast, Mid-
Atlantic, West Coast, or Northwest regions of the U.S, with two studies conducted in the
Midwest as well (Chicago, Illinois and Ann Arbor, Michigan). As such, generalizability of this
review’s findings may be limited to older adults residing exclusively in these regions. While
studies with multiple study sites found similar results across different states (e.g., California,
Massachusetts, and Pennsylvania), these effects may not hold true over different geographies,
and generalizability cannot be assumed.38 This finding reveals a large and problematic gap in the
literature regarding how the built environment impacts the walking behavior of older adults in
other regions of the U.S. This gap is especially important as many cities in the Southeast,
Midwest, and Southwest are characterized by urban sprawl, a pattern of development in which

41
large percentages of the population live in lower density residential areas across a metropolitan
area.54 Sprawl is associated with increased reliance on automobiles, lack of sidewalks, decreased
ability to walk to destinations, and a higher risk for physical inactivity and obesity, all of which
are serious concerns among older adults.54
The association between sprawl and reliance on automobiles is particularly problematic
for older adults, as age-related reductions in physical, cognitive, and visual functioning
commonly diminishes the ability to operate a vehicle overtime.55 The adverse effects of driving
cessation may be intensified in automobile-dependent sprawling cities, especially those without a
convenient, accessible transportation alternative. These adverse outcomes, including decreases in
social integration and activities outside the home, increases in depression and anxiety symptoms,
and an increased risk of nursing home placement, may adversely impact the ability of older
adults to maintain functional independence and age-in-place.56,55 In particular, loss of driving
status is of great concern for older women: compared to men, older women are more likely to
report driving cessation at a younger age,55 have lower self-efficacy for walking,43 and lower
overall engagement in physical activity.39 Findings suggest, however, that both men and women
in lower density urban areas engage in less physical activity,16,37-39 and more research is needed
in sprawling, automobile-dependent cities to determine the ways in which communities can
better support and encourage walking. Future research is especially critical for non-driving
populations, as their lack of driving status indicates a heightened risk for the physical, cognitive,
or visual limitations that also create barriers to PA engagement.55
Regional differences in walking behavior and disparities in population rates of older adult
disability further underscore the need for geographical diversification of research. In regards to
walking, only 57% and 60% of adults in the South and Midwest reported recent walking,
compared to 66% in the Northeast and West.13 These disparities are likely reflected across older
adult populations across these regions, as adults who do not exercise when younger are even less
likely to engage in PA with older age due to the associated losses in physical ability, higher
prevalence of disability, and reduced exercise-related self-efficacy, a key determinant of PA
behavior.57 Adding to these region-specific challenges to older adult walking is the high
proportion of older adult disability in the Southeast (40%), with states in this region containing
the highest proportion of older adults living with at least one disability, including Mississippi
(48%), Alabama (45%), Texas (42%), and Georgia (41%).58 By comparison, older adults living

42
with disabilities in the states included in this review are significantly lower, such as 39% in
Washington D.C., 35% in Maryland.58 Although many of these higher disability Southern states
have large proportions of the population living in rural areas (and thus not within the scope of
this review), small, mid-sized, and large cities dot these regions and are in need of further study
to better understand how place shapes PA engagement in areas with high proportions of
vulnerable and more physically limited older adults. Although disability issues were not widely
discussed within the studies included in this review, they warrant discussion considering the
large percentage of older adults who have, or will develop, a mobility limitation or disability.8

3. Age representation
The studies included in this review lack full age representation of the entire older adult
population, and in particular, older adults in the middle-old (75-84 years) oldest-old category
(≥85 years). As the baby boomer population shifts into older age categories, both populations are
projected to grow significantly over the next few decades, with the middle-old age group rising
from 4.2% to 7% and the oldest old rising from 1.9% to 4.5% by 2050.59 Currently, only 14% of
the older adult population is over 85 years, however, this proportion is expected to increase to
over 21% by 2050.2 While studies in this review included participants in the middle-old and
oldest-old age groups, most mean ages were in the upper 60s and low 70s, and the highest mean
age reported was 75.8 years, further underscoring insufficient representation of the older age
categories in this review. Additionally, by defining older adults as individuals ≥65, this review
was forced to reject numerous studies that defined older adulthood as those who are >55 or 60
years. As older adults experience longer functional independence and life expectancy, defining
this population by a cut-point of 55 and 60 years is likely to increasingly misrepresent the needs
of the population, indicating the need for an update to the definition of older adulthood,
especiallly to facilitate cross-study comparisons.

4. Wayfinding
Wayfinding was among the topics that emerged from this review, yet was insufficiently
addressed across the studies. Only one exploratory study from Marquez and colleagues48
discussed wayfinding as a component of older adult engagement in PA. The ability to safely
navigate one’s neighborhood and the surrounding area is key in overcoming barriers to and
enhancing self-efficacy for engaging in PA, especially in an environment an older adult may be
less familiar with on a day-to-day basis (e.g., neighborhood near one’s doctor or dentist, area

43
surrounding a friend or family member’s home). Wayfinding is especially important for the
many older adults who are unfamiliar with or lack the ability to use mobile GPS technology.48
This topic was not discussed in prior reviews on this topic, further emphasizing the need for
additional attention and exploration.

5. Climate change
Older adults are particular vulnerability to extreme temperatures and weather patterns.60
For example, extreme heat exposure increases risk of illness, hospitalization, and death among
older adults, especially those with chronic health conditions such as diabetes, congestive heart
failures, or lung conditions.60 Given the changing climate, future research needs to focus on
better understanding how older adults respond and adapt to weather, as well as identifying the
ways in which the built environment can support PA despite these changes. While the research
included in this review does mention older adult concerns about unpleasant or extreme weather,
it was not done so in great depth and focused more on concerns regarding cold weather (e.g.,
snow and ice removal, slush, darkness) or rain.15 Older adults did cite particular concerns about
their ability to engage in PA in the face of severe or unpleasant weather, such as high heat and
humidity, snow and ice, and extreme cold, yet no solutions were offered in how they manage
such challenges beyond staying inside.14,43 Additionally, the multistate research included in this
review offered no comparisons in how older adults in different climates respond to or manage
weather (e.g., California compared to New York). Future research is needed to explore these
gaps in greater detail, as this issue is projected to only increase in the coming decades.60

6. Naturally occurring retirement communities

Naturally occurring retirement communities (NORCs) are defined by the residential density
of large numbers of community-dwelling seniors and tend to be characterized by a range of
healthy living built environment qualities, such as close access to destinations, services, and
facilities, safe and well-maintained walking paths, low crime and high perceived safety, social
support of seeing other peers be active, and having local government that prioritizes the needs of
older adults.6 Future research needs to better explore qualities of these communities, as they
commonly demonstrate the features older adults need to successfully age-in-place. Additionally,
because these communities are not intentionally planned, studying their qualities allows for a
better understanding of the environmental variables that older adults seek out when choosing a
place to live. Better understanding of NORCs may provide valuable information regarding key

44
PA facilitating features, and thus help planners, developers, and public health workers determine
which improvements to the built environment will best encourage PA.6

Study Limitations
This systematic review is limited by a few methodological shortcomings. In regards to
search processes and query development, it is possible that search terms and criteria included in
this review may have excluded relevant research. For example, the query development process
revealed that the MeSH term “aged,” refers to adults over the age of 65. However, the MeSH
definition of “aged” is not consistently applied across databases, and in an effort to achieve more
narrow, targeted results, the term “older adult” was added to the search query. While “older
adult” is the commonly used term in current research on this population, it is possible that
additional studies would be identified by including other synonyms, such as “seniors” or
“elderly.” Research may also have been missed due to a lack of searching all possible health-
related databases and by the lack of robust searches of the grey literature. Additionally, this
review included studies published in English only. However, due to this review’s focus on
research in the U.S. only, it is unlikely that studies were missed due to this limitation.
All literature searches and evaluation of selected studies were conducted by the author of
this review. Due to the large number of abstracts reviewed in Phase One (n=3254), it is possible
that eligible studies were inadvertent excluded. This risk was mitigated by limiting the number of
studies reviewed in one sitting throughout Phase One. Additionally, in the case of uncertainty
regarding the fit of a study with inclusion and exclusion criteria, an eligibility decision was made
without consensus from other researchers. As a way of managing this limitation, all articles that
elicited uncertainty were flagged and reviewed a second time at a later date. An additional
limitation of a single-researcher approach is that the risk of bias is higher and more likely to be
introduced. As a way of managing this risk, exclusion and inclusion criteria were specifically
defined and were applied fastidiously.
Lastly, due to an exclusive focus on urban, community-dwelling, older adults (≥65 years)
in the U.S., this review is limited in generalizability to older adult populations outside of these
criteria. Furthermore, this review excluded research that focused exclusively on older adult
populations with a pre-existing health condition, such as diabetes, obesity, or physical
limitations. Due to the prevalence and rise of chronic illness among older adults, and in
particular, the increase in physical limitations with age, this exclusion criterion may have

45
resulted in omission of key and informative literature. Despite this potential limitation, most
studies included in this review did not explicitly exclude older adults with physical limitations,
including those who use assistive devices, from the study population. Due to the prevalence of
assistive device use among older adults included in this review, findings from this study may be
applicable to community-dwelling older adults living with some degree of mobility disability or
limitation.

CONCLUSION

This review presents evidence documenting the role the built environment plays in
encouraging or preventing PA among older adults, as well as gaps in the current literature and
areas of future research. The research synthesized in this review adds to the body of knowledge
on how the built environment shapes older adult PA, and should be used to complement previous
evidence, programs, and case studies of successful communities to help develop interventions for
this population. From this review, it is evident that the creation and preservation of walkable
communities is instrumental in supporting active living among older adults, which thereby helps
reduce, prevent, and manage the many chronic illnesses associated with sedentary living. Even in
neighborhoods with high walkable indices, the majority of older adults fail to meet PA
recommendations, further emphasizing the need to diversify the scope of built environment
features that are designed to encourage and motivate PA among this population.32
When optimized for the particular needs of older adults, the compendium of features that
make up the built environment can serve as key support structures to allow older adults to remain
active, independent, and successfully age-in-place. Over the coming decades, the older adult
population will continue to grow and experience increasingly longer life expectancy. This major
shift in demographics, combined with the social, health, and fiscal costs associated with
sedentary living and loss of functional independence, underscores the need for design solutions
tailored to an older population. By intervening at key leverage points and prioritizing the
development of thoughtfully designed urban environments, communities can become essential
drivers of active living and improved longevity among this vulnerable and growing segment of
the U.S. population.

46
References
1. World Health Organization. Global Health and Aging. Geneva, Switzerland: World
Health Organization;2014.
2. Vincent GK, Velkoff, VA. The Next Four Decades: The Older Population in the United
States: 2010 to 2050. Washington, DC: U.S. Department of Commerce, Economics and
Statistics Administration;2010.
3. United Nations, Department of Economic and Social Affairs, Population Division. World
Urbanization Prospects: The 2014 Revision, Highlights New York, NY: United
Nations;2014.
4. World Health Organization. Global Report on Urban Health. Geneva, Switzerland:
World Health Organization;2016.
5. Moran M, Van Cauwenberg J, Hercky-Linnewiel R, Cerin E, Deforche B, Plaut P.
Understanding the relationships between the physical environment and physical activity
in older adults: a systematic review of qualitative studies. The international journal of
behavioral nutrition and physical activity. 2014;11:79.
6. Kerr J, Rosenberg D, Frank L. The Role of the Built Environment in Healthy Aging:
Community Design, Physical Activity, and Health among Older Adults. Journal of
Planning Literature. 2012;27(1):43-60.
7. Haselwandter EM, Corcoran MP, Folta SC, Hyatt R, Fenton M, Nelson ME. The Built
Environment, Physical Activity, and Aging in the United States: A State of the Science
Review. Journal of Aging & Physical Activity. 2015;23(2):323-329.
8. National Center for Health Statistics. Summary Health Statistics, National Health
Interview Survey. Atlanta, GA: Centers for Disease Control and Prevention, U.S.
Department of Health and Human Services;2014.
9. Brookfield K, Thompson CW, Scott I. The Uncommon Impact of Common
Environmental Details on Walking in Older Adults. International journal of
environmental research and public health. 2017;14(190).
10. U.S. Department of Health and Human Services. 2008 Physical Activity Guidelines for
Americans. Washington, DC: U.S. Department of Health and Human Services;2008.
11. Choi NG, Ransom S, Wyllie RJ. Depression in older nursing home residents: The
influence of nursing home environmental stressors, coping, and acceptance of group and
individual therapy. Aging and Mental Health. 2008;12(5):536-547.
12. U.S. Department of Health and Human Services. Step It Up! The Surgeon General’s Call
to Action to Promote Walking and Walkable Communities. Washington, DC: U.S. Dept
of Health and Human Services, Office of the Surgeon General;2015.
13. Centers for Disease Control and Prevention. Vital signs: walking among adults--United
States, 2005 and 2010. MMWR Morbidity and mortality weekly report. 2012;61(31):595-
601.
14. Chippendale T, Boltz M. The Neighborhood Environment: Perceived Fall Risk,
Resources, and Strategies for Fall Prevention. The Gerontologist. 2015;55(4):575-583.
15. Pollock ML, Carroll JF, Graves JE, Leggett SH, Braith RW, Limacher M, Hagberg JM.
Injuries and adherence to walk/jog and resistance training programs in the elderly.
Medicine and science in sports and exercise. 1991;23(10):1192-1200.
16. Adams MA, Sallis JF, Conway TL, et al. Neighborhood environment profiles for physical
activity among older adults. American journal of health behavior. 2012;36(6):757-769.

47
17. U.S. Bureau of Labor Statistics. Civilian labor force participation rate by age, gender,
race, and ethnicity. 2015; https://www.bls.gov/emp/ep_table_303.htm. Accessed
February 26, 2017.
18. World Health Organization. Pedestrian Safety: A Road Safety Manual for Decision-
Makers and Practitioners. Geneva, Switzerland: World Health Organization;2013.
19. Haleem K, Alluri P, Gan A. Analyzing pedestrian crash injury severity at signalized and
non-signalized locations. Accident; analysis and prevention. 2015;81:14-23.
20. Zegeer CV, Bushell M. Pedestrian crash trends and potential countermeasures from
around the world. Accident; analysis and prevention. 2012;44(1):3-11.
21. Moher D, Liberati A, Tetzlaff J, Altman G, The PRISMA Group. Preferred Reporting
Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. Journal of
Clinical Epidemiology,. 2009;62:1006-1012.
22. Kerr J, Rosenberg D, Frank L. The Role of the Built Environment in Healthy Aging
Community Design, Physical Activity, and Health among Older Adults. Journal of
Planning Literature. 2012;27(1):43-60.
23. Yen IH, Fandel Flood J, Thompson H, Anderson LA, Wong G. How design of places
promotes or inhibits mobility of older adults: realist synthesis of 20 years of research.
Journal of aging and health. 2014;26(8):1340-1372.
24. Herzog R, Álvarez-Pasquin MJ, Díaz C, Del Barrio JL, Estrada JM, Gil Á. Are healthcare
workers’ intentions to vaccinate related to their knowledge, beliefs and attitudes? a
systematic review. BMC public health. 2013;13:154-154.
25. Aidala AA, Wilson MG, Shubert V, et al. Housing Status, Medical Care, and Health
Outcomes Among People Living With HIV/AIDS: A Systematic Review. American
journal of public health. 2016;106(1):e1-e23.
26. Mays N, Pope C. Qualitative research in health care: Assessing quality in qualitative
research. BMJ (Clinical research ed). 2000;320:50-52.
27. Humphreys A, Johnson S, Richardson J, Stenhouse E, Watkins M. A systematic review
and meta-synthesis: evaluating the effectiveness of nurse, midwife/allied health
professional consultants. Journal of clinical nursing. 2007;16(10):1792-1808.
28. Metcalfe A, Coad J, Plumridge GM, Gill P, Farndon P. Family communication between
children and their parents about inherited genetic conditions: a meta-synthesis of the
research. European Journal of Human Genetics. 2008;16(10):1193-1200.
29. Robinson L, Spilsbury K. Systematic review of the perceptions and experiences of
accessing health services by adult victims of domestic violence. Health and Social Care
in the Community. 2008;16(1):16-30.
30. Hannes K. Critical appraisal of qualitative research. In: Noyes J, Booth A, Hannes K,
Harden A, Harris J, Lewin S, Lockwood C, ed. Supplementary Guidance for Inclusion of
Qualitative Research in Cochrane Systematic Reviews of Interventions. 1 ed: Cochrane
Collaboration Qualitative Methods Group; 2011.
31. Ding D, Sallis JF, Norman GJ, et al. Neighborhood environment and physical activity
among older adults: do the relationships differ by driving status? Journal of aging and
physical activity. 2014;22(3):421-431.
32. King AC, Sallis JF, Frank LD, et al. Aging in neighborhoods differing in walkability and
income: associations with physical activity and obesity in older adults. Social science &
medicine (1982). 2011;73(10):1525-1533.

48
33. Bracy NL, Millstein RA, Carlson JA, et al. Is the relationship between the built
environment and physical activity moderated by perceptions of crime and safety? The
international journal of behavioral nutrition and physical activity. 2014;11(1):24.
34. Carlson JA, Sallis JF, Conway TL, et al. Interactions between psychosocial and built
environment factors in explaining older adults' physical activity. Preventive medicine.
2012;54(1):68-73.
35. Cain KL, Millstein RA, Sallis JF, et al. Contribution of streetscape audits to explanation
of physical activity in four age groups based on the Microscale Audit of Pedestrian
Streetscapes (MAPS). Social science & medicine (1982). 2014;116:82-92.
36. Ding D, Sallis JF, Norman GJ, et al. Neighborhood Environment and Physical Activity
Among Older Adults: Do the Relationships Differ by Driving Status? Journal of aging
and physical activity. 2014;22(3):421-431.
37. Tamura K, Puett RC, Hart JE, Starnes HA, Laden F, Troped PJ. Spatial clustering of
physical activity and obesity in relation to built environment factors among older women
in three U.S. states. BMC public health. 2014;14:1322.
38. Troped PJ, Starnes HA, Puett RC, et al. Relationships between the built environment and
walking and weight status among older women in three U.S. States. Journal of aging and
physical activity. 2014;22(1):114-125.
39. Siu VW, Lambert WE, Fu R, Hillier TA, Bosworth M, Michael YL. Built environment
and its influences on walking among older women: use of standardized geographic units
to define urban forms. Journal of environmental and public health. 2012;2012:203141.
40. Sallis JF, Cain KL, Conway TL, et al. Is Your Neighborhood Designed to Support
Physical Activity? A Brief Streetscape Audit Tool. Preventing chronic disease.
2015;12:E141.
41. Li Y, Kao D, Dinh TQ. Correlates of neighborhood environment with walking among
older Asian Americans. Journal of aging and health. 2015;27(1):17-34.
42. Maisel JL. Impact of Older Adults' Neighborhood Perceptions on Walking Behavior.
Journal of Aging & Physical Activity. 2016;24(2):247-255.
43. Gallagher NA, Clarke PJ, Gretebeck KA. Gender differences in neighborhood walking in
older adults. Journal of aging and health. 2014;26(8):1280-1300.
44. Barnett A, van den Hoek D, Barnett D, Cerin E. Measuring moderate-intensity walking in
older adults using the ActiGraph accelerometer. BMC geriatrics. 2016;16(211):1-9.
45. Van Cauwenberg J, Clarys P, De Bourdeaudhuij I, et al. Physical environmental factors
related to walking and cycling in older adults: the Belgian aging studies. BMC public
health. 2012;12:142.
46. Buman MP, Winter SJ, Sheats JL, et al. The Stanford Healthy Neighborhood Discovery
Tool: a computerized tool to assess active living environments. American journal of
preventive medicine. 2013;44(4):e41-47.
47. Chaudhury H, Mahmood A, Michael YL, Campo M, Hay K. The influence of
neighborhood residential density, physical and social environments on older adults'
physical activity: An exploratory study in two metropolitan areas. Journal of aging
studies. 2012;26(1):35-43.
48. Marquez DX, Hunter RH, Griffith MH, Bryant LL, Janicek SJ, Atherly AJ. Older Adult
Strategies for Community Wayfinding. Journal of Applied Gerontology. 2017;36(2):213-
233.

49
49. Saelens BE, Sallis JF, Black J, Chen D. Neighborhood-Based Differences in Physical
Activity: An Environment Scale Evaluation. American journal of public health.
2011;93(9):1552-1558.
50. Cain K, Millstein R, Geremia C. Microscale Audit of Pedestrian Streetscapes (MAPS):
Data Collection & Scoring Manual. San Diego, CA: University of San Diego;2012.
51. Van Cauwenberg J, Bourdeaudhuiji I, De Meester F, Van Dyck D, Salmon J, Clarys P,
Deforche B. Relationship between the physical environment and physical activity in older
adults: A systematic review. Health & place. 2010;17:458-469.
52. Cerin E, Nathan A, Van Cauwenberg J, Bernett D, Barnett A. The neighbourhood
physical environment and active travel in older adults: a systematic review and meta-
analysis. International Journal of Behavioral Nutrition and Physical Activity.
2017;14(15):1-23.
53. Cunningham GO, Michael YL. Concepts guiding the study of the impact of the built
environment on physical activity for older adults: a review of the literature. American
Journal of Health Promotion. 2004;18(6):435-443.
54. Lopez R. Urban Sprawl and Risk for Being Overweight or Obese. American journal of
public health. 2004;94(9):1574-1579.
55. Foley DJ, Heimovitz HK, Guralnik JM, Brock DB. Driving Life Expectancy of Persons
Aged 70 Years and Older in the United States. American journal of public health.
2002;92(8):1284-1289.
56. Carr DB, Ott BR. The Older Adult Driver with Cognitive Impairment: “It’s a Very
Frustrating Life”. Journal of the American Medical Association. 2010;303(16):1632-
1641.
57. Neupert S, Lachman M, Whitbourne SB. Exercise Self-Efficacy and Control Beliefs
Predict Exercise Behavior After an Exercise Intervention for Older Adults. Journal of
Aging & Physical Activity. 2009;17(1):1-16.
58. He W, Larsen LJ. American Community Survey Reports, ACS-29, Older Americans With
a Disability: 2008–2012. Washington DC: U.S. Census Bureau;2014.
59. Ortman J, Velkoff V. An Aging Nation: The Older Population in the United States.
Washington, D.C.: U.S. Census Bureau;2014.
60. United States Environmental Protection Agency. Climate Change and the Health of Older
Adults. 2016; https://www.epa.gov/sites/production/files/2016-10/documents/older-
adults-health-climate-change.pdf. Accessed March 1, 2017.

50
APPENDIX I. Adapted Newcastle Ottawa Quality Assessment Scale for Observational Research

ASSESSMENT CRITERIA POINTS

1 SELECTION (Maximum 5 points)
1.1 Representativeness of the sample
Truly representative of the average in the target population (e.g., all subjects 1
or random sampling).
Somewhat representative of the average in the target population (e.g., non- 1
random sampling).
Selected group of users. 0
No description of the sampling strategy. 0
1.2 Sample Size
Justified and/or satisfactory 1
Not justified and/or unsatisfactory 0
1.3 Non-respondents
Comparability between respondent and non-respondent characteristics is 1
established, and the response rate is satisfactory.
The response rate is unsatisfactory, or the comparability between respondent 0
and non-respondent is unsatisfactory.
No description of the response rate or the characteristics of the responders and 0
the non-responders.
1.4 Ascertainment of the exposure (risk factor)
Validated measurement tool (e.g., Neighborhood Environmental Walkability 2
Scale).
Non-validated measurement tool, but the tool or method of data collection is 1
available or described.
No description of the measurement tool or ascertainment of exposure. 0
2 COMPARABILITY (Maximum 2 points)
2.1 Controls for confounding factors to ensure that subjects are comparable.
The study controls for important demographic confounders. 1
The study includes partial adjustment for important confounders. 1
The study lacks adjustment for potential confounders. 0
3 OUTCOME (Maximum 3 points)
3.1 Assessment of the outcome
Independent, objective assessment (e.g., Actigraph accelerometer). 2
Record linkage. 2
Self-report (e.g., CHAMPS Activities Questionnaire for Older Adults). 1
No description. 0
3.2 Statistical test
The statistical test used to analyze the data is clearly described and 1
appropriate, and the measurement of the association is presented, including
confidence intervals and the probability level (p value).
The statistical test is not appropriate, not described or incomplete. 0
Total possible points 10

51
APPENDIX II. Results of Adapted Newcastle Ottawa Quality Assessment Scale
Study Selection Comparability Outcome Total
First author, Study Design Representativ Sample Non- Ascertainme Controls Assessment Statistical Total
year e-ness of the Size respondents nt of for of outcome test Points (out
Sample exposure confoundin of 10)
g factors
Adams, 2012 Cross-sectional 1 1 0 2 1 2 1 8

Bracy, 2014 Cross-sectional 1 1 0 2 1 2 1 8

Cain, 2014 Cross-sectional 1 1 0 2 1 2 1 8

Carlson, 2012 Cross-sectional 1 1 0 2 1 2 1 8
Ding, 2014 Longitudinal 1 1 0 2 1 2 1 8
Gallagher, 2014 Cross-sectional 1 1 0 2 1 1 1 7
King, 2011 Longitudinal 1 1 0 2 1 2 1 8
Li, 2015 Cross-sectional 0 1 0 1 1 1 1 5
Maisel, 2016 Cross-sectional 1 0 0 2 1 2 1 7
Sallis, 2015 Cross-sectional 1 1 0 2 1 2 1 8
Siu, 2012 Cross-sectional 0 1 0 2 1 2 1 7
Tamura, 2014 Cross-sectional 0 1 0 2 1 2 1 7
Troped, 2014 Cross-sectional 0 1 0 2 1 2 1 7

52
APPENDIX III. Mays and Pope Scale for Quality Assessment Qualitative Research and Results
Key to Results
0 Low clarity and quality as assessed by the reviewer.
1 Reasonable clarity and quality as assessed by the reviewer.
2 Reflects a finding of high clarity and quality as assessed by the reviewer.
NC Not clear or not available from the paper
Buman, Chaudhury, Chippendale, Marquez,
2013 2012 2015 2015
1 Worth or Relevance
Was this piece of work worth doing at all? Has it contributed
2 2 2 2
usefully to knowledge?
2 Clarity of Research Question
Was the research question clear? 2 2 2 2
3 Appropriateness of the Design to the Question
Was an appropriate method used to answer the research question? 2 2 2 2
4 Context
Is the context or setting adequately described so that the reader could
2 2 1 2
relate the findings to other settings?
5 Sampling
Did the sample include the full range of possible cases or settings (so
2 2 1 1
that conceptual generalizations could be made)?
Were efforts made to obtain data that might contradict or modify the
2 2 0 1
analysis by extending or modifying the sample?
6 Data Collection and Analysis
Were the data collection and analysis procedures systematic? Was
an 'audit trail' provided, such that someone else could repeat each 2 2 2 2
stage, including the analysis?
How well did the analysis succeed in incorporating all the
2 2 2 2
observations?
Did the analysis develop concepts and categories capable of
2 2 2 2
explaining key processes (e.g., coding, thematic analysis)?
Was it possible to follow iteration between data and the explanations
2 0 NC 2
for the data?
Did the researcher search for or provide information about cases that
2 2 1 2
do not fit observed patterns)?
7 Reflexivity of the Account
Did the researcher assess the likely impact of the methods used on
2 1 1 1
the data obtained?
Were sufficient data included in the reports to provide sufficient
2 1 1 2
evidence for readers to assess whether analytical criteria were met?

53
APPENDIX IV. Characteristics and Results of Quantitative Studies

Key
sr: self-report
obj: objectively measured
ea: environmental audit
PA: physical activity
MVPA: moderate-to-vigorous physical activity
GIS: Global Information Systems
NEWS: Neighborhood Environment Walkability Scale
CHAMPS: Community Healthy Activities Model Program for Seniors Survey
MAPS: Microscale Audit of Pedestrian Streetscapes
MET: Metabolic equivalence task

First Focus of study Study Built Activity-related Findings

author, design, environment outcomes and
year sample size, features and measures
ages measures
Adams, To explore associations Cross- sr: perceived obj: MVPA
2012 between multivariate sectional neighborhood (Actigraph Physical activity differed significantly by differences
neighborhood profiles, n= 728 (NEWS) accelerometer in neighborhood profiles (based on assessment of
physical activity (PA), Age: 66-97 worn for 7 days) urban form and recreation environment variables).
and BMI. MVPA differed significantly by as much as 10
sr: PA (CHAMPS) minutes/day, 1.1 hours/week for walking for errands,
and almost 50 minutes/week for leisure PA. Low
social/recreational environment variables clustered
along with low walkable environment scores.

Bracy, To tease out the Cross- sr: perceived obj: MVPA Total MVPA was associated with walkability and
2014 relationship between sectional traffic, pedestrian, (Actigraph number of parks. Leisure walking and walking for
safety concerns and n=718 and crime safety accelerometer transport associated with pedestrian safety.
PA, looking Age: ≥66 (modified NEWS) worn for 5 days) Relationship between crime, pedestrian, and traffic
specifically at crime, safety and PA remains elusive.
pedestrian, and traffic obj: neighborhood sr: PA (CHAMPS)
safety as moderators of environment
built environment (GIS walkability
associations with PA. index)

54
First Focus of study Study Built Activity-related Findings
author, design, environment outcomes and
year sample size, features and measures
ages measures
Cain, To identify associations Cross- ea: neighborhood obj: MVPA Both macro and microscale attributes of the
2014 of microscale sectional environment (Actigraph environment influenced older adult PA.
streetscape attributes n=367 (MAPS) accelerometer
with multiple PA Age: ≥66 worn for 5 days) Factors associated with significantly more MVPA:
measures across four obj: neighborhood residential mix, crossing and segment characteristics,
age groups, including environment (GIS sr: transport, and objective walkability index.
older adults. walkability index) leisure, and
neighborhood PA Factors related to increased transport activity:
(CHAMPS) presence of destinations, non-residential land use, and
positive land uses (e.g., restaurants, shops, service);
streetscape characteristics and street segment score
(e.g., design features, slope, sidewalk quality);
positive building height and set-backs (i.e., human-
scale building design); quality of street crossing (e.g.,
positive crossing and segment characteristics);
positive aesthetics/social features.

Carlson, To evaluate ecological Cross- obj: neighborhood obj: PA (Actigraph Walkability, social support, and self-efficacy were
2012 model predictions of sectional environment accelerometer consistently related to PA; combination of walking
cross-level interactions n=718 (GIS walkability worn for 5 days) partner and supportive environment may be
among psychosocial Age: ≥65 index; number of sr: PA (CHAMPS) particularly effective in facilitating older adult PA.
and environmental parks and private Significant interactions related to walking for transport
correlates of PA. recreation involved walkability, while significant interactions
facilities) related to walking for leisure involved walking
infrastructure.
sr: perceived
neighborhood
environment
(modified NEWS)

55
First Focus of study Study Built Activity-related Findings
author, design, environment outcomes and
year sample size, features and measures
ages measures

Ding, To determine whether Longitudinal obj: neighborhood obj: PA (Actigraph Driving status may moderate the association between
2014 associations between (two time environment accelerometer the neighborhood environment and leisure walking.
attributes of points) (GIS walkability worn for 7 days) Almost all environmental attributes showed positive,
neighborhood n=880 index; number of sr: walking significant interactions with PA among driving older
environments and PA Age: 66-97; parks and private (CHAMPS) adults, but not among non-driving older adults.
were moderated by M=75 recreation
driving status among facilities) Most attributes of neighborhood environments were
older adults. related to transport walking regardless of driving
sr: perceived status.
neighborhood
environment Non-driving older adults were more likely to live in
(modified NEWS) neighborhoods with activity-friendly environmental
features.
Gallagher To compare mobility, Cross- sr: perceived sr: neighborhood Built environment variables explained 32% of the
, 2014 self-efficacy, outcome sectional neighborhood walking variance in neighborhood walking in men (p<.001)
expectations, n=326 (106 environment (Neighborhood and 27% of the variance in women (p<.01).
neighborhood (density, men; 216 (modified NEWS) Physical Activity
destinations, and women) Questionnaire) For men, density and design characteristics,
design), and Age: 60-99; specifically sidewalks and perceived crime safety,
neighborhood walking M=76.78 were associated with walking, in addition to scores on
among older men and (men), 75.81 measures of walking self-efficacy.
women (women)
For women, built environment destinations were
association with neighborhood walking, as well as
walking self-efficacy scores. Regarding walking self-
efficacy subscales, women were significantly less
confident than men in their ability to walk for
increasing durations of time, walk in the face of
neighborhood or personal barriers. safely navigate
common obstacles, and maintain balance.

Walking was associated with self-efficacy for walking

despite individual barriers in women and
neighborhood barriers in men.
56
First Focus of study Study Built Activity-related Findings
author, design, environment outcomes and
year sample size, features and measures
ages measures
King, To uncover relations Longitudinal obj: neighborhood obj: MVPA Across regions, time, and neighborhood income, older
2011 among objectively (two time environment (Actigraph adults living in more walkable neighborhoods had
measured points) (GIS walkability accelerometer 400% more transport activity (p<.0001) relative to
neighborhood design, n=719 index) worn for 7 days) those living in less walkable neighborhoods. Residents
mobility impairment, Age:≥66 sr: transport and in high walkability neighborhoods were 33% more
PA, and body weight leisure walking active than those in low walkable neighborhoods. The
by comparing two (CHAMPS) most mobility-impaired adults living in more walkable
neighborhoods neighborhoods reported transport activity levels that
differing in walkability were similar to less mobility-impaired adults living in
and income levels. less walkable neighborhoods.

Li, 2015 To explore the Cross- sr: neighborhood sr: transport and Asian older adults walked significantly more than
relationship between sectional variables leisure walking White counterparts. There is heterogeneity among
neighborhood factors n=1,045 (social cohesion; (past week) Asian subgroups in terms of relationship of
and walking among Age: ≥65 availability of neighborhood factors with walking behavior. Social
older Asian Americans, (55-64 recreational cohesion was associated with increased minutes of
examining specific included for facilities, walking regardless of subgroup, though most notable
ethnic subgroups (e.g., comparison perceived among Chinese older adults. Access to
Chinese, Korean, purposes) neighborhood park/playground associated with increased amounts of
Filipino). safety) walking among older Chinese/Korean adults.
Neighborhood safety related to increased walking
among Filipino adults but no other subgroups.
Maisel, To examine the effect Cross- sr: perceived sr: job, Total weekly walking associated with overall
2016 of neighborhood sectional neighborhood transportation, neighborhood satisfaction and more positive
perceptions and self- n=32 urban environment recreational, and perceptions of street connectivity (r=.25; p<.01) and
reported walking older adults (modified NEWS) total neighborhood crime safety.
behavior for older adult (study total walking (past 7 Job walking and total weekly walking associated with
residents of urban, n=112) days, International perceptions of traffic safety. Recreational walking
suburban, and rural Age:≥65 Physical Activity associated with aesthetics and crime safety.
neighborhoods. Questionnaire) Transportation walking was moderately correlated to
recreation walking. Job walking, transportation
walking, and total weekly walking were associated
with perceptions of street connectivity.

57
First Focus of study Study Built Activity-related Findings
author, design, environment outcomes and
year sample size, features and measures
ages measures
Sallis, To evaluate the Cross- ea: neighborhood obj: MVPA Significant associations found for self-reported
2015 efficacy of a condensed sectional environment (Actigraph walking for transport and microscale features (p<.05):
15-item environmental n=367 (MAPS) accelerometer street lights, benches, curb cuts, presence of a
audit tool compared to Age: ≥66 worn for 7 days) sidewalk, buffers between streets, sidewalks, crossing
a full version. Tool was sr: active transport and intersection characteristics (crosswalk, curb cuts,
evaluated through and leisure PA crossing signal).
assessment of (CHAMPS)
microscale elements of Findings show a linear and positive relationship
neighborhood design between positive microscale BE features and walking
and their relationship for transport. No single attribute was dominant.
with PA. Aesthetics and social characteristics were largely
unrelated to active transport.

Lowest v. highest quintiles on neighborhood

microscale environment scores: 242% different in PA.

Siu, 2012 To explore Cross- obj: neighborhood sr: utilitarian and Urban areas with the best access to amenities (transit,
neighborhood attributes sectional environment (GIS leisure walking parks, and commercial areas), high street connectivity,
that may affect health n=2005 measures of (total number of and high population density were most likely to
outcomes, specifically women accessibility to blocks walked promote walking in older women. Characteristics of
evaluating the Age: ≥65 transit services, daily) the central city were associated with increased
relationship between land use mix, utilitarian and leisure walking as compared to the city
urban forms and street connectivity, periphery, the suburbs, or urban fringe with poor
walking behavior population commercial area or park access.
among older women. density)
Across all six neighborhood clusters, those who
walked for leisure purposes tended to walk more than
those who walked only for utilitarian purposes.

58
First Focus of study Study Built Activity-related Findings
author, design, environment outcomes and
year sample size, features and measures
ages measures
Tamura, To explore spatial Cross- obj: neighborhood sr: walking for Population density, intersection density, and
2014 clusters of PA, examine sectional environment (GIS exercise or diversity/density of facilities significantly greater in
whether the geographic n=22,599 measured transport (MET higher PA clusters compared to low PA clusters.
distribution of women population density, value assigned for
covariates affects Age: 57-85; intersection walking based on Variables associated with greater odds of meeting
clusters, and compare M=69.9 density, density of pace; dichotomous DHHS PA recommendations: higher population
built environment facilities) walking outcome density; intersection density, >2-11 intersections/km
characteristics inside categorized compared to <2 (18% greater odds); additional service
and outside clusters. participants as facility/km of road (53% greater odds)
meeting or not
meeting DHHS PA Significant interactions between population density,
recommendations) facility density variables, and odds of meeting PA
recommendations via walking mostly found among
women living in the 90.1-95th and 95.1-100th
percentiles of population density.

Troped, To examine the Cross- obj: neighborhood sr: walking for Population density (odds ratio(OR)=1.04 [1.02,1.07]),
2014 relationship between sectional environment (GIS exercise or intersection density (ORs=1.18-1.28), and facility
objective built n=22,599 measured transport (MET density (ORs=1.01-1.53) were positively associated
environment variables women population density, value assigned for with walking.
and the likelihood of Age: 57-85; intersection walking based on
meeting the U.S. M=70 density, density of pace; dichotomous Strongest associations between facility density
Department of Health facilities) walking outcome variables and walking found among women from
and Human Services categorized higher population density areas. Relationships
PA recommendations participants as between accessible facilities and walking may be most
via walking and weight meeting or not important in higher density areas. No clear pattern of
status, using individual meeting DHHS PA differences in associations across CA, MA, and PA.
residential buffers to recommendations)
define environmental
exposures.

59
APPENDIX V. Characteristics and Results of Qualitative Studies

Key
sr: self-report
ea: environmental audit
PA: physical activity

First author, Focus of study Study design, Built Activity- Findings

year sample size, ages environment related
features and outcome(s)
measures
Buman, 2013 To develop and Community- sr: neighborhood sr: PA 1. Common facilitators
evaluate the based environment (GPS (GPS Aesthetics (e.g., presence of trees, flowers);
utility of a participatory to recorded recorded parks/playgrounds (e.g., walking paths; public garden);
computerized, research common walking common amenities/destinations (e.g., shops, restaurants, public
tablet-based (photovoice and routes; geocoded walking services); personal safety (e.g., "crime free," "upscale
participatory audio narratives) audio narratives; routes; living"); sidewalk features (e.g., convenient, well-kempt
tool designed to n=27 photographs of geocoded routes)
identify Age:≥65 facilitators/barriers audio 2. Common barriers
neighborhood to active living) narratives) Negative sidewalk features (e.g., cracks, unevenness);
elements that personal safety issues (e.g., afraid of being hit by a vehicle);
affect active disability issues (e.g., street not suitable for wheelchair or
living. walker, lack of ramps); crosswalk limitations (e.g., cars do
not stop, signals inoperable); and road safety (e.g., speeding
cars, blind driveways)

Chaudhury, To explore the Photovoice sr: neighborhood sr: leisure 1. Safety and security
2012 influence of n=32 environment and Barriers: maintenance and upkeep of the physical
neighborhood Age: ≥65 (65-92; (Participants transport environment (e.g., uneven sidewalks, obstacles/barriers that
residential 62% from ≥75) instructed to PA were tripping hazards/made it unsafe to walk, absence of
density and photograph sidewalks, ending or narrow sidewalks; facilitators: paved,
physical/social facilitators or flat, smooth, and wide walking surfaces with good
environments barriers to PA) lighting/accessible seating, ramps); traffic hazards (e.g., busy
on physical streets, high traffic volume and speed, unsafe intersections
activity of older and crosswalks, dangerous/impatient drivers with little
adults. respect for rules of the road, poor visibility); neighborhood
atmosphere (e.g., feeling unsafe, crime, graffiti, vandalism)

60
First author, Focus of study Study design, Built Activity- Findings
year sample size, ages environment related
features and outcome(s)
measures
2. Accessibility
Facilitators: access to convenient public transportation;
access to neighborhood facilities or services (i.e., accessible
amenities with utilitarian purpose and recreational/social
component, such as bank, grocery store, post office, mall,
library, gym, recreation center)
Barrier: poor access to local parks
3. Comfort of movement
Facilitators: available seating, railings, handrails, ramps, safe
stairs, water fountains
4. Peer support
Facilitators: community gardens, spaces for socialization in
public areas (e.g., benches, picnic tables), formal social
support (e.g., planned activities such as walking groups,
community based programs), informal social support
(socializing and peer support after or during PA was
valuable, e.g., walking with others for exercise, walking to a
meeting spot for coffee, meeting or socializing while on a
walk)

Chippendale, To explore Semi-structured sr: neighborhood sr: 1. Barriers to PA

2015 neighborhood interview environment physical Built barriers: sidewalks (e.g., cobble stones, brick surfaces,
factors that n=14 (semi-structured activity, uneven grates, broken surfaces, curbs); insufficient street
influence Age: ≥65 interview) time spent crossing amenities; insufficient park benches; poor visibility
barriers to PA (M=75.86) outdoors or insufficient street lighting; barriers to transit (e.g., steps,
and safety per day wet subway floors, tiled walkways)
(including fear Natural/environmental barriers: wind, snow, ice; lack of or
of falling), delayed snow removal, pools of slush or rainwater
resources that Program/service specific barriers: lack of programmatic
support PA and accommodation for older adults at gyms/exercise facilities
safety, and Social barriers: presence of too many pedestrians, dog
motivators for walkers, cyclists/skateboarders; gait speed of other
PA. pedestrians; personal safety; vulnerability due to looking
older
61
First author, Focus of study Study design, Built Activity- Findings
year sample size, ages environment related
features and outcome(s)
measures
2. Resources for PA and safety
Built resources for PA: walking trails, benches for rest,
stores for browsing and shopping, senior centers, gyms and
recreation facilitates, low/no cost facilities, street fairs,
farmers markets, public transit as alternative form of
mobility in case of fatigue
Built structures safety: structures to hold for support (e.g.,
railings, fences), benches, timed pedestrian crossings,
kneeling buses, elevators in the subway,
Social environment: perception of safety (e.g., pedestrians to
help cross the street, crossing guards)
3. Motivators for PA
Environmental motivators: appealing architecture, well
maintained green spaces, water views, settings that promote
connectivity to nature
Built motivators: continuous and spacious walking paths and
benches
Social motivators: places that promote socialization serve as
walkable destinations (e.g., basketball courts,
playgrounds/parks that promote conversation and people
watching; dog parks; senior centers)
Marquez, To identify Mixed methods ea: neighborhood sr: transport Wayfinding facilitators and barriers
2015 features that (environmental environment and leisure Facilitators: landmarks, numbers on buildings, signage (e.g.,
facilitate or audit, in-person (CDC Healthy walking street signs, block number, advance street signs); land use
inhibit interviews, Aging Research sr: self- items (e.g., railroad tracks), transit stops; availability of
wayfinding in community Network efficacy for people to ask for help
outdoor settings walks; map Environmental walking and Barriers: places with no good landmarks or missing signs;
and better drawing exercise) Audit Tool) wayfinding distractions (e.g., noise of train); confusing street
understand how n=35 sr: neighborhood (Healthy alignments; lack of other pedestrians deemed trustworthy to
they impact ≥65 (M=70.6) environment Aging ask for help
mobility, PA, (interviews, Research Preferred cues: landmarks, smells, memorable features,
and community community walks, Network logical street labeling system
engagement. map drawing Protocol)
exercise
62