UBC Faculty Research and Publications

Neighborhood walkability, physical activity, and walking for transportation: A cross-sectional study… Chudyk, Anna M.; McKay, Heather A; Winters, Meghan; Sims-Gould, Joanie; Ashe, Maureen C Apr 10, 2017

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
52383-12877_2017_Article_469.pdf [ 472.37kB ]
Metadata
JSON: 52383-1.0362100.json
JSON-LD: 52383-1.0362100-ld.json
RDF/XML (Pretty): 52383-1.0362100-rdf.xml
RDF/JSON: 52383-1.0362100-rdf.json
Turtle: 52383-1.0362100-turtle.txt
N-Triples: 52383-1.0362100-rdf-ntriples.txt
Original Record: 52383-1.0362100-source.json
Full Text
52383-1.0362100-fulltext.txt
Citation
52383-1.0362100.ris

Full Text

RESEARCH ARTICLE Open AccessNeighborhood walkability, physical activity,and walking for transportation: A cross-sectional study of older adults living on lowincomeAnna M. Chudyk1,2*, Heather A. McKay1,2, Meghan Winters1,3, Joanie Sims-Gould1,2 and Maureen C. Ashe1,2AbstractBackground: Walking, and in particular, outdoor walking, is the most common form of physical activity for olderadults. To date, no study investigated the association between the neighborhood built environment and physicalactivity habits of older adults of low SES. Thus, our overarching aim was to examine the association between theneighborhood built environment and the spectrum of physical activity and walking for transportation in olderadults of low socioeconomic status.Methods: Cross-sectional data were from the Walk the Talk Study, collected in 2012. Participants (n = 161, meanage = 74 years) were in receipt of a rental subsidy for low income individuals and resided in neighbourhoods acrossMetro Vancouver, Canada. We used the Street Smart Walk Score to objectively characterize the built environmentmain effect (walkability), accelerometry for objective physical activity, and the Community Healthy Activities ModelProgram for Seniors (CHAMPS) questionnaire to measure walking for transportation. We used regression analyses toexamine associations of objectively measured physical activity [total volume, light intensity and moderate intensityphysical activity (MVPA)] and self-reported walking for transportation (any, frequency, duration) with walkability. Weadjusted analyses for person- and environment-level factors associated with older adult physical activity.Results: Neighbourhood walkability was not associated with physical activity volume or intensity and self-reportedwalking for transportation, with one exception. Each 10-point increase in Street Smart Walk Score was associatedwith a 45% greater odds of any walking for transportation (compared with none; OR = 1.45, 95% confidenceinterval = 1.18, 1.78). Sociodemographic, physical function and attitudinal factors were significant predictors ofphysical activity across our models.Conclusions: The lack of associations between most of the explored outcomes may be due to the complexity ofthe relation between the person and environment. Given that this is the first study to explore these associationsspecifically in older adults living on low income, this study should be replicated in other settings.Keywords: Built environment, Walkability, Walk Score, Physical activity, Walking, Walking for transportation* Correspondence: anna.chudyk@hiphealth.ca1Centre for Hip Health and Mobility, 7th floor—2635 Laurel Street,Vancouver, BC V5Z 1M9, Canada2Department of Family Practice, University of British Columbia, 3rd Floor -5950 University Boulevard, Vancouver, BC V6T 1Z3, CanadaFull list of author information is available at the end of the article© The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Chudyk et al. BMC Geriatrics  (2017) 17:82 DOI 10.1186/s12877-017-0469-5BackgroundDespite the many benefits of a physically active lifestyle[1], adults aged ≥ 60 years represent the least active agegroup [2, 3]; only 13% of older adults in Canada [2] at-tain sufficient physical activity to meet public healthguidelines of engaging in ≥ 150 min of moderate-to-vig-orous physical activity (MVPA) per week [4, 5]. Bar-riers to engaging in physical activity in older adultsinclude poor health, unsupportive built environments(e.g., no sidewalks, parks or recreation centres), lackof knowledge about the relationship between physicalactivity and health and negative experiences with exerciseearlier in life [6]. Further, given the broad spectrum ofmobility-disability in the older adult population, achievingguideline levels of higher (moderate) intensity physical ac-tivity may not be possible. Although often overlooked,physical activity at levels below guidelines is important tothe general health, mobility and community engagementof older people. For example, total physical activity vol-ume may have stronger associations with cardiometabolicbiomarkers than MVPA accumulated in bouts [7]. Light-intensity physical activity was associated with older adults’physical health and well-being, independent of MVPA [8].It is also plausible that individuals who are more physicallyactive outdoors have more opportunities for communityengagement. Thus, although current physical activityguidelines are important for health, encouraging any phys-ical activity, including light physical activity, is increasinglyrecognized as important [9, 10]. Moreover, given thebroad range of mobility limitations for some older adults,this may be more a more realistic public health goal forolder adults.Walking, and in particular, outdoor walking, is themost common form of physical activity for older adults[11]. Walking requires minimal equipment, and intensityis dictated by the individual. Further, within a supportiveoutdoor environment, walking can be incorporated rela-tively easily into daily life routines as either structuredor incidental activity. The built environment, defined asurban design, land use, and transportation systems [12],can be an important facilitator or a barrier to outdoorwalking. For example, a built environment rich with desti-nations relevant to older adults provides an opportunityto walk for daily travel [13–16]. Yet findings are mixed re-garding specific built environment features associated witholder adult walking, and physical activity in general[17–20]. Built environment features most consistentlyassociated with older adult walking and physical activityinclude street connectivity, access to destinations (e.g.,shops, restaurants) and features related to perceivedsafety (e.g., good lighting, absence of crime, presence ofcrosswalks) [18, 20]. The extent to which an individualsuccessfully navigates his/her environment is a result ofthe match between the pressures exerted by theenvironment (e.g., features of the built environment) andthe competence (e.g., capacity) of the individual [21].Person-level factors that contribute to older adults’ cap-acity to be active in their neighbourhood include cogni-tive, physical, psychosocial, and financial domains; thesocial environment also plays an important role [22, 23].Importantly, older adults span diverse physical, psycho-social and cognitive abilities and person-level factors playa key role in the person-environment interaction. Thus, itis relevant to focus on distinct subgroups within theolder adult population [e.g., those with mobility limita-tions or of low socioeconomic status (SES)] to betterunderstand the association between the built environ-ment and older adult physical activity.Older adults of low SES are understudied in physicalactivity and aging research [24]. The built environmentmay more strongly influence the physical activity habitsof this population specifically, as they have less dispos-able income and as a result may rely more upon un-structured (and free) physical activities, such as outdoorwalking. Further, older adults of low SES are more likelyto walk or take public transit, instead of drive, as theirmain form of transportation [13, 25]. In doing so theymay accrue incidental physical activity, as well as engagewith the built environment and other people. Conversely,individuals of low SES are at increased risk of poor healthoutcomes (e.g., morbidity, physical impairment) that de-crease their capacity to be active [26–28]. In sum, olderadults of low SES may be more likely to be active in thebuilt environment and may also have health concerns thatrequire their physical activities take place in walkableenvironments.Our overarching aim was to examine the associationbetween the neighbourhood built environment andphysical activity of older adults living on low incomeacross a spectrum of physical activity.MethodsAimsOur primary aim was to study the association betweenthe built environment and the total physical activityvolume (as measured by accelerometry) of older adultsliving on low income, including: i) total activity counts(TAC) and ii) steps. Our secondary aims were to deter-mine the association between the built environmentand specific intensities and domains of physical activityincluding: i) light physical activity, ii) MVPA, and iii)self-reported walking for transportation.Design and study sampleWe conducted a cross-sectional study of older adults whowere participants in Walk the Talk, a larger study that in-vestigated the association between the built environmentand mobility and health of older adults living on lowChudyk et al. BMC Geriatrics  (2017) 17:82 Page 2 of 14income. We provide a detailed description of Walk theTalk study methods, including recruitment, outcomes anddata collection, elsewhere [15]. Briefly, we identified olderadults in receipt of a rental subsidy (Shelter Aid for ElderlyRenters, SAFER) through a provincial crown organization(BC Housing). In January—February 2012, we recruitedolder adults (aged ≥ 65 years) using a stratified design,randomly selecting households (ntotal = 2000) in theirstudy area (Metro Vancouver) across strata (deciles) ofWalk Score® (www.walkscore.com). Upper cut-points(deciles) were 100(1), 93(2), 87(3), 78(4), 72(5), 67(6),60(7), 52(8), 43(9), and 32(10). Walk Score is a publiclyavailable index that measures the walkability of a streetaddress based on its distance to pre-defined destinationcategories (e.g., grocery stores, etc.). We excluded individ-uals who: self-reported a medical diagnosis of dementia,did not understand or speak English, stated that they lefttheir home to go into their community less than once in atypical week, stated that they were unable to walk ≥ 10-mwith or without a mobility aid (e.g., cane, walker), and/orwere unable to participate in a mobility assessment thatinvolved a 4-m walk. A total of 161 older adults volun-teered to be measured in March-May 2012.Measures and instrumentsOutcome measuresPhysical activity We used ActiGraph GT3X+ (LLC,Pensacola, FL) tri-axial accelerometers to objectively as-sess participants’ patterns of physical activity. During thein-person measurement sessions, we instructed partici-pants on accelerometer use, including proper placement(e.g., just above right hip and in line with the middle ofthe right hip, underneath or on top of clothing as longas it fit snugly on the body), wear period (during wakinghours, for seven consecutive days), and to remove theaccelerometer during water-based activity. We also pro-vided participants with paper copies of these instructions(with photos) to take home. We requested that partici-pants wear their accelerometers on their right hip duringwaking hours, in the week following their in-person as-sessment. We collected data continuously (at 30 Hz) andthen reintegrated the data to 60-s epochs; we consideredmore than 60 min of continuous zeroes as non-weartime. To be as inclusive as possible, we chose an 8-h/daywear time criteria for accelerometry data [29]. As thisdecision had the potential to influence our findings, weconducted sensitivity analyses and found that estimatesof our main effect remained stable when we used a moreconservative 10-h valid day wear time criterion (data notshown). We excluded from our analyses participantswith less than three valid wear days. We used cut-pointsproposed by Freedson and colleagues to classify timespent in light physical activity (100–1951 counts perminute) and MVPA ( ≥ 1952 counts per minute) [30].We measured participants’ total volume of physical ac-tivity per day using TAC [31] and steps. We calculatedTAC (n/day), steps (n/day), light physical activity (min/day), and MVPA (min/day) as total amount of activityaccumulated during valid days divided by number ofvalid days. We processed accelerometry data usingActiLife software version 6.5.4 (LLC, Pensacola, FL).Self-reported walking for transportation We assessedself-reported walking for transportation [yes/no; frequency(ntrips/wk) and duration (hr/wk)] using a single item fromthe Community Healthy Activities Model Program for Se-niors (CHAMPS) survey [32]. The item asked participantswhether in a typical week in the last 4 weeks they hadwalked to do errands such as going to/from a store ortaking children to school (walked for transportation).Participants that reported having walked for transporta-tion were also asked to indicate the frequency (ntrips/wk)and duration (hr/wk) spent walking for transportation.Response options for the duration component of thequestion were < 1 h, 1–2.5 h, 3–4.5 h, 5–6.5 h, 7–8.5 hand ≥ 9 h. We used the midpoint of each response op-tion and recoded values to derive duration of walkingfor transportation (hr/wk); a value of 9.75 hr/wk repre-sented the highest possible duration of walking fortransportation [32].Independent variablesWe organize independent variables by domains adaptedfrom Webber and colleague’s framework of older adultmobility [23]. This includes a neighbourhood social envir-onment domain to control for (i) neighbourhood socialcohesion (e.g., shared beliefs and expectations) and (ii)neighbourhood physical and social disorder that influencephysical activity [22].Built environment domain We used the Street SmartWalk Score® as an objective measure of walkability forparticipants’ neighbourhood built environment. This wasour main effect of interest. The Street Smart Walk Scoreis a revised version of the Walk Score that uses an up-dated algorithm to measure the walkability of an address.The algorithm assigns an address a score of 0 to 100 basedon network distances from the address to nine differentamenity (destination) categories (e.g., grocery stores,restaurants, shopping). Different weights are assignedto different categories based on importance to walkabil-ity. Multiple destinations within each category counttoward the score in order to reflect depth of choice.Destinations located within ≤ 0.25 miles are assignedmaximum scores and those located > 1.5 miles are notfactored into the score. The score is penalized (maximumpenalty of 10% of total score) for street networkChudyk et al. BMC Geriatrics  (2017) 17:82 Page 3 of 14characteristics (intersection density and block length) thatdo not support pedestrian friendliness. The validity ofStreet Smart Walk Score was established in community-dwelling older adults who reside in the USA [33], in Can-adian communities that span a rural-urban continuum[34], and using common measures of the built environ-ment across different buffer sizes [35].We assessed participants’ perceptions of neighbourhoodaesthetics and safety (traffic, crime) using a modifiedversion of the Neighbourhood Environment WalkabilityScale—abbreviated (NEWS-A) [36]. Scale scores rangefrom 1 to 4 (strongly disagree, somewhat disagree,somewhat agree, strongly agree). We recoded someitems so that higher scores signify higher walkability forall three NEWS-A subscales.Neighbourhood social environment domain We mea-sured participants’ perceptions of neighbourhood socialcohesion and trust using a five-item measure (scalerange 1–5) [37]. We used a five-item measure drawnfrom the Project on Human Development in ChicagoNeighbourhoods to measure participants’ perceptions ofneighbourhood physical and social disorder (scale range1–4) [38]. Since we adapted this measure, we providedetails about the questions for reproducibility. The itemsaddress: how much i) broken glass or trash participants seeon neighbourhood sidewalks and streets, and ii) graffiti par-ticipants see on neighbourhood buildings and walls, iii)how many vacant/deserted houses or storefront partici-pants see in their neighbourhood; how often iv) participantssee people drinking in public places in their neighbour-hood, and v) participants see unsupervised children hang-ing out on the street in their neighbourhood. We reversecoded items. Thus, higher scores indicate more positiveperceptions (less disorder).Physical domain We used a TANITA Electronic ScaleModel BWB-800 and Seca Stadiometer Model 242 tomeasure participants’ weight (kg) and height (cm), respect-ively; we used these data to calculate body mass index(BMI; kg/m2). We used the Functional Comorbidity Indexto measure self-reported number of comorbidities associ-ated with physical function (scale range 0–18) [39]. Finally,we calculated participants’ gait speed (m/s) as part of the4-m walk (usual pace) component of the Short PhysicalPerformance Battery [40].Psychosocial domain We measured how much partici-pants like to walk outside using a five-point scale (not atall, not much, neutral, somewhat, very much). We dichot-omized (very much vs. other) responses as a majority ofresponses were in the “very much” category. We usedthe Ambulatory Self-Confidence Questionnaire to meas-ure participants’ perceived self-efficacy to walk in 22different home and community environments (scalerange 1–10) [41].Sociodemographic factors We used a self-report ques-tionnaire to determine participants’ age, gender, maritalstatus, living arrangement, vehicle access in the last7 days (yes/no), and dog ownership (yes/no).AnalysisWe summarized continuous data using means andstandard deviations (SD) and categorical data usingcounts and percentages. We present summaries by gen-der, as it is a well-established determinant of older adultphysical activity [42].We fitted multivariable models (linear regression,logistic regression, Poisson regression, described in detailbelow) as per the type of dependent variable (e.g., con-tinuous, binary, count). For self-reported frequency(ntrips/wk) and duration (hr/wk) of walking for transpor-tation outcomes, we limit analyses to participants thatreported ≥ 1 walking for transportation trip (n = 124,77% of participants) in order to improve model fit.To examine the association between Street SmartWalk Score and TAC (n/day) we used linear regression.We first fitted a crude model to estimate the main effectof Street Smart Walk Score on TAC, with Street SmartWalk Score as the only independent variable. We thenfitted a second model identical to the first but control-ling for the effects of age and gender. Finally, we fitted athird model identical to the second but with all inde-pendent variables associated with TAC at p ≤ 0.20 in bi-variate analyses. We selected independent variables forbivariate analyses based on their known associationswith older adult physical activity and/or walking [18, 20,25, 42–44]. The independent variables spanned per-ceived built environment, neighbourhood social environ-ment, physical, psychosocial and sociodemographicdomains (described in measures).We followed the same procedure (above) for the othercontinuous dependent variables [steps (n/day), light inten-sity physical activity (min/day), MVPA (min/day), and dur-ation of walking for transportation (hr/wk)]. We applied alog transformation for TAC and MVPA as residuals werehighly skewed. For these models we present exponentiatedregression coefficients to interpret them in the originalunit of measurement (n/day and min/day). These expo-nentiated coefficients are interpreted as fold-change in thedependent variable.Using logistic regression and a truncated Poisson regres-sion model, we examined the association between StreetSmart Walk Score and: i) odds of any (compared withnone) walking for transportation, and ii) frequency of walk-ing for transportation (ntrip/wk), respectively. These ana-lyses followed the same procedure as for our continuousChudyk et al. BMC Geriatrics  (2017) 17:82 Page 4 of 14dependent variables. We report truncated Poisson modelcoefficients and associated confidence intervals transformedto incidence-rate ratios (IRRs), calculated as eβi.For each outcome, we used Akaike’s information criter-ion to guide selection between crude and adjusted models.We assessed the adequacy of the fitted normal linear re-gression models with residual plots. We assessed the ad-equacy of the fitted logistic regression models by plots ofobserved versus estimated probabilities grouped into dec-iles of estimated probability and with the Hosmer-Lemeshow goodness of fit test. We assessed the adequacyof the fitted truncated Poisson regression models with thelikelihood ratio test and comparison of standard errorsand point estimates between truncated Poisson modelsfitted with robust standard errors and Poisson models,respectively. For each of the fully adjusted models, wecalculated variance inflation factors; a variance inflationfactor > 10 was regarded as indicating serious multicol-linearity that warranted changes to the model. Finally,we also investigated outliers with the dfbeta commandin Stata.We considered p < 0.05 to be statistically significant inmultivariable analyses. We conducted all analyses usingStata version 13.0 (Stata Corp, TX).ResultsPreviously we described flow of participants into the study[15]. Briefly, we randomly sampled 2000 households fromour source population of 5806 households. After exclusionof five households due to prior attempted recruitment intoour pilot study, we contacted 1995 individuals (from 1995households) for study participation. All 161 individualsthat consented to participate completed an in-personmeasurement session. All but two participants (who de-clined) wore an accelerometer for 1 week to capture pat-terns of physical activity. One hundred and fifty eightparticipants returned accelerometers; 141 had ≥ 3 days ofvalid data. Participants that provided valid accelerometrydata wore accelerometers for a mean (SD) of 7 (1) daysand a mean (SD) of 784 (105) minutes/day.We present select characteristics of participants, bygender, in Table 1. Participant characteristics did notvary between those with vs. without ≥ three valid days ofwear time (data not shown). Participants’ mean age was74 years, 65% were women, and more than 3/4 (81%)lived alone. Approximately half of participants reporteda vehicle at their disposal in the 7 days prior to studyparticipation. Participants were overweight (BMI range25.0–29.9), had a mean of three comorbidities, and hada gait speed that was consistent with community walking( ≥ 0.8 m/s) [43], on average.Table 2 describes participants’ physical activity andwalking for transportation. Participants engaged in ap-proximately 240 min of physical activity/day (on average),of which 220 min was light and 20 min was moderate-to-vigorous intensity. Of note, participants obtained the vastmajority of their MVPA through moderate intensity phys-ical activity; on average, they spent less than 1 min/day invigorous physical activity (data not shown). One hundredand twenty-four participants (77%) reported any walkingfor transportation in the assessment week.Tables 3 and 4 display crude and adjusted linear re-gression analyses for physical activity volume (TAC andsteps) and intensity (light physical activity and MVPA).Street Smart Walk Score was not associated with any ofthese physical activity outcomes in crude or adjustedmodels. Among covariates, BMI was associated with allfour outcomes in fully adjusted models. Each unit in-crease in BMI was associated with a 3% (95% CI = -4, -1)decrease in TAC, 162 (95% CI = -245, -79) fewer steps,2.97 (95% CI = -5.38, -0.57) minutes less of light physicalactivity, and a 7% (95% CI = -11, -3) decrease in MVPA.Age and self-reported walking enjoyment (very muchlike to walk) were also associated with all physical activityoutcomes except light physical activity in fully adjustedmodels. Each 10-year increase in age was associatedwith a 19% (95% CI = -30, -5) decrease in TAC, 903(95% CI = -1642, -164) fewer steps, and a 34% (95% CI= -53, -7) decrease in MVPA. Very much liking to walk(vs. less than very much liking to walk) was associatedwith a 32% (95% CI = 7, 63) increase in TAC, taking1342 (95% CI = 337, 2346) more steps, and a 100%(95% CI = 25, 221) increase in MVPA. Women engagedin 34.09 (95% CI = 5.67, 62.50) more minutes of lightphysical activity and 47% (95% CI = -66, -16) less MVPAcompared with men (fully adjusted models). Finally,MVPA increased by 193% (95% CI = 25, 221) for each unitincrease in gait speed (fully adjusted models).Table 5 highlights the crude and adjusted results oflogistic regression models for any walking for transpor-tation. In the fully adjusted model, the odds of any walkingfor transportation was 1.45 (95% CI = 1.18, 1.78) timesgreater for each 10-point increase in Street Smart WalkScore. Further, in this model, the odds of any walking fortransportation were 14.29 (95% CI = 3.33, 50.00) timeshigher among participants who did not have a vehicleavailable and 5.60 (95% CI = 1.68, 18.65) higher for thosewho very much liked to walk. No other variables wereassociated with any walking for transportation in fullyadjusted models.Table 6 shows the crude and adjusted results of trun-cated Poisson regression models and linear regressionmodels for frequency (ntrips/wk) and duration (hr/wk) ofwalking for transportation, respectively. These modelsinclude only participants who reported any walking fortransportation (n = 124). Although Street Smart WalkScore was associated with frequency of walking for trans-portation in the crude model (IRR = 1.06, 95% CI = 1.01,Chudyk et al. BMC Geriatrics  (2017) 17:82 Page 5 of 141.10) and the model adjusted for age and gender (IRR =1.06, 95% CI = 1.01, 1.11), it was no longer significant inthe fully adjusted model (IRR = 1.03, 95% CI = 0.98, 1.08).Street Smart Walk Score was not associated with durationof walking for transportation in any model. In fully ad-justed models, very much like to walk (vs. less than verymuch like to walk) increased the incidence rate of fre-quency of walking for transportation by 1.52 (95% CI =1.15, 2.01) and duration of walking for transportationby 1.44 (95% CI = 0.15, 2.73). Further, each unit increasein BMI was associated with a 0.11 (95% CI = -0.21, -0.01)decrease in duration of walking for transportation.DiscussionWe noted that a walkable neighbourhood (as mea-sured by Street Smart Walk Score) was associated withsignificantly higher odds of engaging in any self-reportedwalking for transportation, but not with the volume orintensity of physical activity (as measured by accelero-metry). Further, among those who walked forTable 1 Descriptive statistics for select characteristics, by genderCharacteristic Men Women Totaln mean (SD) n mean (SD) n mean (SD)SOCIODEMOGRAPHICSAge (yrs) 59 74.2 (6.3) 102 74.4 (6.2) 161 74.3 (6.2)Married (%) 59 102 161No 81 97 91Yes 19 3 9Living arrangement (%) 59 102 161Lives alone 68 88 81Lives with others 32 12 19Had vehicle at disposal in last 7 days (%) 59 102 161No 41 50 47Yes 59 50 53Owns a dog (%) 59 102 161No 92 88 89Yes 8 12 11BUILT ENVIRONMENTStreet Smart Walk Score (/100) 59 71.3 (27.7) 102 71.8 (24.0) 161 71.6 (25.3)NEWS-Aa Subscale F: Aesthetics (/4) 58 2.9 (0.8) 102 3.3 (0.6) 160 3.2 (0.7)NEWS-Aa Subscale G: Traffic hazards (/4) 57 2.4 0.6) 98 2.4 (0.6) 155 2.4 (0.6)NEWS-Aa Subscale H: Crime (/4) 56 3.3 (0.7) 96 3.3 (0.7) 152 3.3 (0.7)PHYSICALBody mass index (kg/m2) 59 26.9 (4.6) 102 27.0 (5.7) 161 27.0 (5.3)Number of comorbiditiesb 57 2.8 (2.0) 101 3.0 (2.2) 158 2.9 (2.1)Gait speed (m/s)c 59 1.0 (0.2) 102 1.0 (0.3) 161 1.0 (0.3)PSYCHOSOCIALLikes to walk outside… (%) 59 102 161Less than very much (1-4 on a 5-point scale) 44 25 31Very much (5 on a 5-point scale) 66 75 69Ambulatory self-confidence questionnaire (/10) 59 8.6 (1.4) 102 8.2 (1.8) 161 8.4 (1.7)SOCIAL ENVIRONMENTNeighbourhood social cohesion and trustd (/5) 56 3.3 (0.8) 101 3.5 (0.7) 157 3.4 (0.7)Neighbourhood physical and social disordere (/4) 57 3.4 (0.6) 102 3.5 (0.4) 159 3.5 (0.5)aNEWS-A = Neighbourhood Environment Walkability Scale – abbreviated; some scales reverse coded so that higher score indicates better walkabilitybTotal number; measured with the Functional Comorbidity IndexcAssessed as part of the 4-m walk (usual pace) component of the Short Physical Performance Batteryd5-item measure of social cohesion and truste5-item measure of neighbourhood physical and social disorder; reverse coded so that higher score indicates better walkability (less disorder)Chudyk et al. BMC Geriatrics  (2017) 17:82 Page 6 of 14transportation, there was no association between neigh-bourhood walkability and frequency or duration ofwalking for transportation. As we found no publishedstudies of older adults living on low income, we compareour findings to studies conducted in a general older adultpopulation; that is, community-dwelling older adults whowere not recruited in reference to a specific characteristic(like, disease status, SES, etc.). Importantly, we note thetremendous diversity of outcomes reported and instru-ments used to measure physical activity and the builtenvironment.Seven studies used objective measures of the built envir-onment and assessed physical activity objectively (by accel-erometry) in older adults. Five of these [45–50] used acomposite index for neighbourhood walkability and all wereobservational trials. Buman and colleagues reported the im-portance of light intensity physical activity to the physicalhealth and psychosocial well-being of older adults [8]. Des-pite this, the current literature (and recommended guide-lines) tend to focus on cardiovascular benefits of physicalactivity and thus the association between objectively mea-sured features of the built environment and time spent inMVPA; findings were mixed [45–47, 49, 50]. Only onestudy reported older adults’ light intensity physical activity,divided into low-light and high-light physical activity inten-sity [50]. They reported a significant negative associationbetween low-light physical activity and objectively mea-sured walkability and no association between walkabilityand high-light physical activity [50]. Other studies investi-gated the association between total physical activity volume(by accelerometry) and objectively measured walkability[47, 48], also with mixed results. Collectively, these findingsspeak to a complex association between the built environ-ment and older adults’ time spent in physical activity. Giventhe many factors that contribute to older adults’ physicalactivity, finding no clear association between the built en-vironment and physical activity is not uncommon.There are a few potential explanations for the lack ofsignificant associations between neighbourhood walkabilityand physical activity volume and intensity. First, an individ-ual’s activity is a product of the dynamic interplay betweencharacteristics of the individual and features of the envir-onment [21]. Thus the association between the built envir-onment and older adult physical activity may bemoderated by person-level variables. For example, Dingand colleagues found that time spent in MVPA was signifi-cantly associated with walkability among drivers, but notnon-drivers [47, 50]. Van Holle and colleagues found thattime spent in MVPA was only associated with walkabilityin high walkability/low neighbourhood income areas[47, 50]. Second, physical activity is a broad constructthat encompasses four domains—leisure time physical ac-tivity, occupational physical activity, household physicalactivity and transportation-related physical activity [42].It could be that older adults who live in low walkableneighbourhoods supplement their physical activity withactivities that take place outside of an undesirableneighbourhood built environment. Van Holle and col-leagues suggested that older adults who live in less walk-able neighbourhoods may spend more time engaged inindoor activities, such as housework [50]. Finally, olderadults who live in more walkable neighbourhoods maymake shorter trips to nearby, accessible destinations.Therefore, they accrue less physical activity in their dailytravel than counterparts who live in less walkable neigh-bourhoods and thus travel further to reach amenities. Thisis consistent with a previous study from our group thatdemonstrated neighbourhood walkability was associatedwith smaller activity spaces [51].Importantly, living in a more walkable neighbourhoodwas associated with greater odds of older adults doing anywalking for transportation. Thus, low walkable neighbour-hoods may act as a barrier to older adults’ decision towalk for transportation. Alternatively, more walkableTable 2 Physical activity and walking for transportation outcomes, by genderOutcome Men Women Totaln mean (SD) n mean (SD) n mean (SD)PHYSICAL ACTIVITYa 49 92 141TACc (n/day) 175889.5 (97991.6) 170473.3 (103190.6) 172355.5 (101095.7)Steps (n/day) 5113 (2572) 5175 (3165) 5153 (2963)Light physical activity (min/day) 193.1 (66.8) 234.1 (79.9) 219.8 (77.9)MVPAb (min/day) 23.5 (20.5) 17.8 (20.4) 19.8 (20.6)WALKING FOR TRANSPORTATIONd 59 65 124Frequency (ntrips/wk) 4.4 (2.2) 4.0 (2.0) 4.2 (2.1)Duration (hr/wk) 4.1 (3.0) 3.4 (2.6) 3.7 (2.8)aAs measured by accelerometry (ActiGraph GT3X+, 60 s epochs), based on ≥ 3 days with ≥ 480 min/day valid weartimebMVPA =moderate-to-vigorous physical activitycTAC = total activity countsdAs measured by the Community Healthy Activities Model Program for Seniors survey; only includes participants that reported making ≥ 1 walking fortransportation trip (n = 124)Chudyk et al. BMC Geriatrics  (2017) 17:82 Page 7 of 14Table3Estimatesfromlinearregressionanalysesforphysicalactivityvolume[totalactivitycounts(TAC,number/day)aandsteps(number/day)]PredictorTAC(n/day)Steps(n/day)CrudeAdjustedCrudeAdjustedModel2cModel3dModel4cModel5e(n=141)b(n=141)bfitonaline(n=131)(n=141)b(n=141)(n=138)β(95%CI)Pβ(95%CI)Pβ(95%CI)Pβ(95%CI)Pβ(95%CI)Pβ(95%CI)PStreetSmartWalkScore(10-pointchange)1.01(0.97,1.05)0.5461.02(0.99,1.07)0.2211.00(0.96,1.04)0.99151(−148,250)0.612102(−93,297)0.304−23(−207,160)0.801Women0.98(0.8,1.21)0.8760.98(0.8,1.19)0.8160.90(0.74,1.10)0.30861(−978,1101)0.90731(−975,1038)0.951−347(−1309,616)0.478Age(10-yearchange)0.74(0.63,0.87)*<0.0010.73(0.62,0.85)*<0.0010.81(0.70,0.95)*0.011−1298(−2091,−505)*0.002−1362(−2168,−557)*0.001−903(−1642,−164)*0.017Vehicleavailable0.86(0.71,1.06)0.156--0.94(0.79,1.12)0.495−695(−1687,297)0.168--−293(−1150,565)0.500Aestheticsf1.23(1.07,1.42)*0.004--1.08(0.94,1.23)0.272879(178,1580)*0.014--233(−415,881)0.478Crimeg1.11(0.96,1.28)0.173--1.07(0.95,1.22)0.272------Bodymassindex(kg/m2 )0.96(0.94,0.97)*<0.001--0.97(0.96,0.99)*0.004−234(−316,−152)*<0.001--−162(−245,−79)*<0.001Comorbiditiesh0.92(0.87,0.96)*<0.001--0.97(0.92,1.02)0.180−399(−625,−173)*0.001--−120(−336,97)0.276Gaitspeed(m/s)i2.79(1.94,4.02)*<0.001--1.40(0.90,2.18)0.1404816(2997,6635)*<0.001--1738(−397,3873)0.110Verymuchliketowalkj1.67(1.37,2.04)*<0.001--1.32(1.07,1.63)*0.0102495(1519,3471)*<0.001--1342(337,2346)*0.009AmbulatoryConfidencek1.13(1.06,1.19)*<0.001--1.01(0.95,1.09)0.690546(259,834)*<0.001--98(223,418)0.547a TAC(number/day)presentedasexponentiatedregressioncoefficientsbn vehicleavailable=140;n aesthetics=140;n traffichazards=135;n comorbidities=139c AdjustedforStreetSmartWalkScore,gender,agedAdjustedforallpredictorvariableslistedinthistablee Adjustedforallpredictorvariableslistedinthistablewiththeexceptionofcrime,sincecrimewasnotassociatedwithsteps(n/day)atp≤0.2inbivariateanalysesfNeighbourhoodEnvironmentWalkabilityScale—abbreviated(NEWS-A)SubscaleF:Aesthetics(four-pointscale);reversecodedsothathigherscoreindicatesbetterwalkabilitygNEWS-ASubscaleH:Crime(four-pointscale);reversecodedsothathigherscoreindicatesbetterwalkabilityhTotalnumber;measuredwiththeFunctionalComorbidityIndexi Assessedaspartofthe4-mwalk(usualpace)componentoftheShortPhysicalPerformanceBatteryj Verymuchliketowalk(5ona5-pointscale)vs.lessthanverymuchlikingtowalk(1–4ona5-pointscale)k AssessedbytheAmbulatorySelf-ConfidenceQuestionnaire*p<0.05Chudyk et al. BMC Geriatrics  (2017) 17:82 Page 8 of 14Table4Estimatesfromlinearregressionanalysesforphysicalactivityintensity[lightphysicalactivity(minutes/day)andmoderate-to-vigorousphysicalactivity(MVPA,minutes/day)a ]Lightphysicalactivity(min/day)MVPA(min/day)PredictorCrudeAdjustedCrudeAdjustedModel2cModel3dModel4cModel5e(n=141)b(n=141)(n=134)(n=141)b(n=141)(n=138)β(95%CI)Pβ(95%CI)Pβ(95%CI)Pβ(95%CI)Pβ(95%CI)Pβ(95%CI)PStreetSmartWalkScore(10-pointchange)−3.48(−8.69,1.73)0.189−2.73(−7.80,2.33)0.288−5.22(−10.83,0.39)0.0681.04(0.95,1.15)0.3951.07(0.97,1.18)0.1641.00(0.92,1.09)0.950Women40.93(14.48,67.38)*0.00340.55(14.44,66.66)*0.00334.09(5.67,62.50)*0.0190.70(0.42,1.17)0.1750.69(0.42,1.13)0.1440.53(0.34,0.84)*0.007Age(10-yearchange)−22.97(−44.24,−1.71)*0.034−20.72(−41.61,0.18)0.052−6.20(−29.20,16.81)0.5950.54(0.36,0.79)*0.0020.51(0.34,0.76)*0.0010.66(0.47,0.93)*0.019Vehicleavailable------0.61(0.37,0.99)*0.045--0.71(0.48,1.06)0.097Aestheticsf19.58(1.00,38.15)*0.039--6.26(−13.03,25.55)0.5221.56(1.10,2.20)*0.012--1.18(0.87,1.60)0.275Traffichazardsg17.82(−5.43,41.07)0.132--11.33(−10.79,33.45)0.313-----Bodymassindex(kg/m2 )−3.59(−5.90,−1.28)*0.003--−2.97(−5.38,−0.57)*0.0160.89(0.85,0.93)*<0.001--0.93(0.89,0.97)*<0.001Comorbiditiesh−6.44(−12.55,−0.34)*0.039--−1.97(−8.46,4.52)0.5490.80(0.71,0.89)*<0.001--0.92(0.83,1.02)0.108Gaitspeed(m/s)i101.82(52.37,151.27)*<0.001--59.43(−2.77,121.63)0.06111.97(4.91,29.17)*<0.001--2.93(1.08,7.98)*0.035Verymuchliketowalkj36.86(9.63,64.08)*0.008--10.59(−19.24,40.43)0.4833.57(2.21,5.76)*<0.001--2.00(1.25,3.21)*0.004Ambulatoryconfidencek6.83(−1.02,14.69)0.088--0.78(−8.74,10.31)0.8711.33(1.16,1.53)*<0.001--1.02(0.87,1.18)0.844a MVPA(min/day)presentedasexponentiatedregressioncoefficientsbn vehicleavailable=140;n aesthetics=140;n traffichazards=135;n comorbidities=139c adjustedforStreetSmartWalkScore,gender,andagedadjustedforallvariableslistedinthistablewiththeexceptionofvehicleavailability,sincevehicleavailabilitywasnotassociatedwithtimespentinlightphysicalactivity(min/day)atp≤0.2inbivariateanalysese adjustedforallvariableslistedinthistablewiththeexceptionoftraffic,sincetrafficwasnotassociatedwithtimespentinMVPA(min/day)atp≤0.2inbivariateanalysesf NEWS-A(NeighbourhoodEnvironmentWalkabilityScale—abbreviated)SubscaleF:Aesthetics(four-pointscale)gNEWS-ASubscaleG:Traffichazards(four-pointscale);reversecodedsothathigherscoreindicatesbetterwalkabilityhTotalnumber;measuredwiththeFunctionalComorbidityIndexi Assessedaspartofthe4-mwalk(usualpace)componentoftheShortPhysicalPerformanceBatteryj Verymuchliketowalk(5ona5-pointscale)vs.lessthanverymuchlikingtowalk(1–4ona5-pointscale)k AssessedbytheAmbulatorySelf-ConfidenceQuestionnaire*p<0.05Chudyk et al. BMC Geriatrics  (2017) 17:82 Page 9 of 14neighbourhoods may enable older adults to integrateactive transportation into their daily life routines. How-ever, we did not find a dose-response among those whomade transportation-related walking trips. Previous stud-ies reported positive associations between older adults’self-reported time spent walking for transportation andobjectively measured walkability [14, 45, 46, 49, 50].Unique to our study, we included only participants livingon low income who made ≥ one walking for transporta-tion trip. Previously, we reported an association betweenobjectively measured walkability and frequency of walkingtrips [15]. This study extends our previous study as we usea different instrument (self-report physical activityquestionnaire vs. travel diaries) and focus our analyseson participants who engaged in any walking fortransportation.The interaction between a person and their environ-ment is dynamic and together they influence [walking]behavior [21]. Our findings at the person-level underscorethis dynamism. More specifically, sociodemographic,physical function and attitudinal factors are establishedpredictors of older adult physical activity [42–44]. All sur-faced as significant predictors of physical activity and/orwalking for transportation in our regression models.As per other reports [52, 53], our results were gender-specific in that men spent more time engaged inMVPA and less time in light intensity physical activitycompared with women. Age was inversely associatedwith all levels of physical activity (by accelerometry)except that of light intensity. This speaks to an olderadult’s ability to sustain light activities over time whereasthe ability to engage in more intense activities may de-cline. Body mass index was inversely associated with allfour physical activity outcomes and duration of walkingfor transportation. Gait speed was positively associatedwith MVPA. This speaks to the close link between phys-ical capacity and mobility and one’s engagement in higherintensity physical activities. Finally, an attitudinal factor(how much participants liked to walk) was associated withall outcomes except for light physical activity.We note that our study has some strengths. These in-clude: i) our focus on an understudied and potentially“at risk” population (older adults living on low income),ii) reporting outcomes across a spectrum of physicalactivity – the importance of light activity and physicalactivity volume are, in our view, currently understudied,iii) the selection of independent variables for our modelsbased on a theoretical framework [23], iv) a robust sam-pling frame that included stratification across deciles ofneighbourhood walkability to ensure variability in urbanform, v) objective measures of physical activity volumeand intensity in real-time (by accelerometry), vi)Table 5 Estimates from logistic regression analyses for making any walking for transportation trip/wk (vs. none)Crude AdjustedModel 2b Model 3c(n = 161)aOR (95% CI)P (n = 161)OR (95% CI)P (n = 151)OR (95% CI)PStreet Smart Walk Score(10-point change)1.37 (1.18, 1.59)* <0.001 1.37 (1.18, 1.60)* <0.001 1.45 (1.18, 1.78)* <0.001Women 1.24 (0.58, 2.63) 0.576 1.21 (0.54, 2.72) 0.846 0.97 (0.32, 2.97) 0.571Age (10-year change) 1.09 (0.60, 1.98) 0.772 0.94 (0.49, 1.80) 0.648 1.29 (0.54, 3.07) 0.963Vehicle available 0.09 (0.03, 0.28)* <0.001 - - 0.07 (0.02, 0.30)* <0.001Aestheticsd 1.46 (0.89, 2.40) 0.136 - - 1.15 (0.50, 2.61) 0.746Comorbiditiese 0.85 (0.72, 1.01) 0.068 - - 0.88 (0.68, 1.14) 0.335Gait speed (m/s)f 3.61 (0.79, 16.51) 0.098 - - 1.71 (0.10, 29.34) 0.713Very much like to walkg 4.30 (1.99, 9.30)* <0.001 - - 5.60 (1.68, 18.65)* 0.005Ambulatory confidenceh 1.17 (0.95, 1.45) 0.134 - - 0.97 (0.68, 1.40) 0.885Social cohesioni 0.60 (0.34, 1.03) 0.065 - - 0.49 (0.23, 1.05) 0.067Disorderj 0.36 (0.14, 0.89)* 0.028 - - 0.67 (0.16, 2.73) 0.572anvehicle available = 159; naesthetics = 160; ncomorbidities = 158; nsocial cohesion = 157; ndisorder = 159badjusted for Street Smart Walk Score, gender, and agecadjusted for all predictor variables listed in this tabledNeighbourhood Environment Walkability Scale—abbreviated (NEWS-A) Subscale F: Aesthetics (four-point scale); reverse coded so that higher score indicatesbetter walkabilityeTotal number; measured with the Functional Comorbidity IndexfAssessed as part of the 4-m walk (usual pace) component of the Short Physical Performance BatterygVery much like to walk (5 on a 5-point scale) vs. less than very much liking to walk (1–4 on a 5-point scale)hAssessed by the Ambulatory Self-Confidence Questionnairei5-item measure of social cohesion and trustj5-item measure of neighbourhood physical and social disorder; reverse coded so that higher score indicates better walkability (less disorder)*p < 0.05Chudyk et al. BMC Geriatrics  (2017) 17:82 Page 10 of 14Table6a Estimatesfromregressionanalysesforfrequency(ntrips/wk)bandduration(hr/wk)cofwalkingfortransportationFrequency(ntrips/wk)Duration(hr/wk)CrudeAdjustedCrudeAdjustedModel2eModel3fModel4eModel5hPredictor(n=124)dIRR(95%CI)P(n=124)IRR(95%CI)P(n=121)IRR(95%CI)P(n=124)gβ(95%CI)P(n=124)β(95%CI)P(n=112)β(95%CI)PStreetSmartWalkScore(10-pointchange)1.06(1.01,1.10)*0.0131.06(1.01,1.11)*0.0111.03(0.98,1.08)0.206−0.01(−0.23,0.21)0.920−0.01(−0.23,0.21)0.956−0.01(−0.27,0.25)0.935Women0.89(0.74,1.07)0.2280.89(0.74,1.08)0.2370.87(0.72,1.06)0.181−0.69(−1.71,0.34)0.187−0.68(−1.71,0.35)0.196−0.81(−1.89,0.28)0.145Age(10-yearchange)0.94(0.81,1.09)0.4370.93(0.80,1.08)0.3541.04(0.89,1.23)0.608−0.21(−1.02,0.60)0.613−0.19(−1.00,0.62)0.6430.12(−0.76,1.00)0.786Vehicleavailable------−0.98(−1.97,0.01)0.053--−0.68(−1.78,0.41)0.217Crimei------−0.79(−1.52,−0.06)*0.034--−0.76(−1.62,0.09)0.079Bodymassindex(kg/m2 )0.98(0.96,1.00)*0.041--0.99(0.97,1.01)0.420−0.09(−0.19,0.00)0.058--−0.11(−0.21,−0.01)*0.046Comorbiditiesj0.96(0.92,1.01)0.097--1.00(0.95,1.06)0.989------Verymuchliketowalkk1.66(1.28,2.14)*<0.001--1.52(1.15,2.01)*0.0041.29(0.15,2.44)*0.027--1.44(0.15,2.73)*0.029Ambulatoryconfidencel1.09(1.02,1.16)*0.010--1.05(0.98,1.13)0.184------SocialCohesionm------−0.81(−1.55,−0.06)*0.034--−0.65(−1.45,0.15)0.109Disordern0.80(0.68,0.95)*0.009--1.19(0.99,1.43)0.057−0.71(−1.68,0.25)0.144--0.23(−0.96,1.43)0.699a Theseanalysesonlyincludeparticipants(n=124)thatself-reported≥1walkingfortransportationtrip[asmeasuredbytheCommunityHealthyActivitiesModelProgramforSeniors(CHAMPS)survey]bAnalysedusingtruncatedpoissonregressionmodels.Dataarepresentedasincidentrateratios(IRRs)c Analysedusinglinearregressionmodelsdn crime=117;n comorbidities=121;n disorder=123e adjustedforStreetSmartWalkScore,gender,andagef adjustedforallpredictorvariableslistedinthistablewiththeexceptionofvehicleavailability,crime,andsocialcohesion,sincethesethreevariableswerenotassociatedwithfrequencyofwalkingfortransportation(ntrips/wk)atp≤0.2inbivariateanalysesgn vehicleavailable=123;n crime=117;n socialcohesion=120;n disorder=123hadjustedforallpredictorvariableslistedinthistablewiththeexceptionofcomorbiditiesandambulatoryconfidence,sincethesetwovariableswerenotassociatedwithdurationofwalkingfortransportation(hr/wk)atp≤0.2inbivariateanalysesi NeighbourhoodEnvironmentWalkabilityScale—abbreviated(NEWS-A)SubscaleH:Crime(four-pointscale);reversecodedsothathigherscoreindicatesbetterwalkabilityj Totalnumber;measuredwiththeFunctionalComorbidityIndexk Verymuchliketowalk(5ona5-pointscale)vs.lessthanverymuchlikingtowalk(1–4ona5-pointscale)l AssessedbytheAmbulatorySelf-ConfidenceQuestionnairem5-itemmeasureofsocialcohesionandtrustn5-itemmeasureofneighbourhoodphysicalandsocialdisorder;reversecodedsothathigherscoreindicatesbetterwalkability(lessdisorder)*p<0.05Chudyk et al. BMC Geriatrics  (2017) 17:82 Page 11 of 14complementing objective measures of physical activitywith self-reported data in order to capture domain-specific physical activity (walking for transportation), asthis may be more closely associated with the built envir-onment [54], and vii) including perceived measures of thebuilt environment to complement objective measures.We acknowledge that our study also had a number oflimitations. Our recruitment rate was only 8%. This rela-tively low response rate likely reflects our focus upon alow SES population, a group that tends to not participatein research [55]. Self-selection bias may exist if active,healthy participants were more likely to participate. Thatsaid, key determinants of physical activity (e.g., age andgender) in study participants were similar to those of thesource population (SAFER recipients). Finally, in orderto improve model fit, we limited analyses for self-reported frequency and duration of walking trip to par-ticipants that reported making ≥ 1 walking for transpor-tation trip. This limits the generalizability of the findingsof these two analyses to individuals who leave the home.ConclusionsA more walkable neighbourhood appears to be a worth-while investment as it encourages older adults living onlow income to walk for transportation. However, factorsbeyond the built environment alone appear to influenceduration and frequency of these trips and physical activ-ity in general. This is best investigated in future, throughbuilt environment studies with larger, diverse samples ofolder adults living on low income that might also providethe opportunity to investigate moderating relations.In our view, future studies that consider or incorporatethe following would be of great benefit: i) how the ratioof indoor and outdoor physical activity differs betweenresidents of highly walkable and low walkable neigh-bourhoods; ii) characterizing older adult movement andphysical activity using a combination of accelerometryand global positioning systems to investigate walking triplengths, as well as the association between walking fortransportation and physical activity, across a range ofwalkability; and iii) investigating the association betweenperson and environment-level variables on light physicalactivity as measured by accelerometry. We live in a timewhen the health and mobility of an aging demographicwill dictate the demands placed upon municipal andprovincial governments. Therefore it seems crucial toidentify aspects of the built environment that promotehealthy behaviours, like physical activity, across a broad(economic and health) spectrum of older adults as ameans to guide these decision makers.AbbreviationsBMI: Body mass index; CHAMPS: Community Healthy Activities Model Program forSeniors survey; CI: Confidence interval; IRR: Incident rate ratio; MVPA: Moderate-to-vigorous physical activity; NEWS-A: Neighbourhood Environment WalkabilityScale—abbreviated; SAFER: Shelter Aid for Elderly Renters; SD: Standard deviation;SES: Socioeconomic status; TAC: Total activity countsAcknowledgementsThe Walk the Talk team would like to gratefully acknowledge the importantcontributions of its participants, as well as of key community partners;namely BC Housing, the City of Vancouver, and the BC Ministry of Health.For statistical support, we thank Dr. Penny Brasher.FundingWe are most grateful to the Canadian Institutes of Health Research (CIHR,Mobility and Aging Team Grant Competition) for their support of the Walkthe Talk: Transforming the Built Environment to Enhance Mobility in SeniorsTeam (CIHR grant # 108607). Anna Chudyk was supported by a VanierCanada Graduate Scholarship from the CIHR. Drs. Ashe and Sims-Gould are sup-ported by career awards from the CIHR and the Michael Smith Foundation forHealth Research. These funding bodies were not involved in the design ofthe study, nor in the collection, analysis, and interpretation of data, nor inwriting the manuscript.Availability of data and materialThe datasets analysed during the current study are available from thecorresponding author on reasonable request.Authors’ contributionsAMC was responsible for all major areas of concept development, datacollection, analysis and presentation of findings, and manuscript writing.HAM and MCA were the supervisory authors and guided all aspects of theresearch, including contributing to concept development, design, approach,presentation of findings and edits to the manuscript. MW and JSG wereinvolved in concept development and contributed to manuscript edits.All authors read and approved the final manuscript.Competing interestsThe authors declare that they have no competing interests.Consent for publicationNot applicable.Ethics approval and consent to participateWe obtained written informed consent from all participants prior to studyparticipation. The University of British Columbia’s Clinical Research EthicsBoard approved the study (certificate: H10-02913).Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.Author details1Centre for Hip Health and Mobility, 7th floor—2635 Laurel Street,Vancouver, BC V5Z 1M9, Canada. 2Department of Family Practice, Universityof British Columbia, 3rd Floor - 5950 University Boulevard, Vancouver, BC V6T1Z3, Canada. 3Faculty of Health Sciences, Simon Fraser University, 11522 -8888 University Drive, Burnaby, BC V5A 1S6, Canada.Received: 29 September 2016 Accepted: 22 March 2017References1. Chodzko-Zajko WJ, Proctor DN, Fiatarone Singh MA, Minson CT, Nigg CR,Salem GJ, Skinner JS. American College of Sports Medicine position stand.Exercise and physical activity for older adults. Med Sci Sports Exerc.2009;41(7):1510–30.2. Colley RC, Garriguet D, Janssen I, Craig CL, Clarke J, Tremblay MS. Physicalactivity of Canadian adults: accelerometer results from the 2007 to 2009Canadian Health Measures Survey. Health Rep. 2011;22(1):7–14.3. Troiano RP, Berrigan D, Dodd KW, Masse LC, Tilert T, Mcdowell M. Physicalactivity in the United States measured by accelerometer. Med Sci SportsExerc. 2008;40(1):181–8.4. Chodzko-Zajko WJ, Proctor DN, Singh MAF, Minson CT, Nigg CR, Salem GJ, SkinnerJS. Exercise and physical activity for older adults. Med Sci Sports Exerc. 2009;41.Chudyk et al. BMC Geriatrics  (2017) 17:82 Page 12 of 145. Tremblay MS, Warburton DE, Janssen I, Paterson DH, Latimer AE, Rhodes RE,Kho ME, Hicks A, Leblanc AG, Zehr L, et al. New Canadian physical activityguidelines. Appl Physiol Nutr Metab. 2011;36(1):36–46.6. Schutzer KA, Graves BS. Barriers and motivations to exercise in older adults.Prev Med. 2004;39(5):1056–61.7. Wolff-Hughes DL, Fitzhugh EC, Bassett DR, Churilla JR. Total activity countsand bouted minutes of moderate-to-vigorous physical activity: Relationshipswith cardiometabolic biomarkers using 2003-2006 NHANES. J Phys Act Health.2015;12(5):694–700.8. Buman MP, Hekler EB, Haskell WL, Pruitt L, Conway TL, Cain KL, Sallis JF,Saelens BE, Frank LD, King AC. Objective light-intensity physical activityassociations with rated health in older adults. Am J Epidemiol.2010;172(10):1155–65.9. Lee I. How much physical activity is good enough? National Physical ActivityPlan Alliance Commentaries on Physical Activity and Health. National PhysicalPlan Alliance. 2015;1(2).10. Sparling PB, Howard BJ, Dunstan DW, Owen N. Recommendations forphysical activity in older adults. BMJ. 2015;350:h100.11. Dai S, Carroll DD, Watson KB, Paul P, Carlson SA, Fulton JE. Participation intypes of physical activities among US adults–National Health and NutritionExamination Survey 1999-2006. J Phys Act Health. 2015;12 Suppl 1:S128–140.12. Handy SL, Boarnet MG, Ewing R, Killingsworth RE. How the built environmentaffects physical activity - Views from urban planning. Am J Prev Med.2002;23(2):64–73.13. Cao XY, Mokhtarian PL, Handy SL. Neighborhood design and the accessibilityof the elderly: an empirical analysis in Northern California. Int J Sustain Transp.2010;4(6):347–71.14. Cerin E, Lee KY, Barnett A, Sit CH, Cheung MC, Chan WM, Johnston JM. Walking fortransportation in Hong Kong Chinese urban elders: a cross-sectional study on whatdestinations matter and when. Int J Behav Nutr Phys Act. 2013;10:78.15. Chudyk AM, Winters M, Moniruzzaman M, Ashe MC, Sims-Gould J, McKay HA.Destinations matter: the association between where older adults live and theirtravel behavior. J Transp Health. 2015;2(1):50–7.16. King D. Neighborhood and individual factors in activity in older adults:results from the neighborhood and senior health study. J Aging Phys Act.2008;16(2):144–70.17. Intersection between the built and social environments and older adult’smobility: an evidence review. http://www.ncceh.ca/sites/default/files/Built_and_Social_Environments_Older_Adults_Nov_2012.pdf. Accessed 1 Feb 2016.18. Rosso AL, Auchincloss AH, Michael YL. The urban built environment andmobility in older adults: a comprehensive review. J Aging Res. 2011;2011:816106.19. Van Cauwenberg J, De Bourdeaudhuij I, De Meester F, Van Dyck D, Salmon J,Clarys P, Deforche B. Relationship between the physical environmentand physical activity in older adults: A systematic review. Health Place.2011;17(2):458–69.20. Yen IH, Flood JF, Thompson H, Anderson LA, Wong G. How design ofplaces promotes or inhibits mobility of older adults: realist synthesis of20 years of research. J Aging Health. 2014;26(8):1340–72.21. Lawton MP. Environment proactivity and affect in older people. In:Spacapan S, Oskamp S, editors. The Social Psychology of Aging. California:SAGE Publications Ltd; 1989. p. 135–64.22. McNeill LH, Kreuter MW, Subramanian SV. Social environment and physicalactivity: A review of concepts and evidence. Soc Sci Med. 2006;63(4):1011–22.23. Webber SC, Porter MM, Menec VH. Mobility in older adults: A comprehensiveframework. Gerontologist. 2010;50(4):443–50.24. Hughes SL, Leith KH, Marquez DX, Moni G, Nguyen HQ, Desai P, Jones DL.Physical activity and older adults: expert consensus for a new researchagenda. Gerontologist. 2011;51(6):822–32.25. Turcotte M. Profile of seniors’ transportation habits. Can Soc Trends.2012;93:1–16.26. Koster A, Bosma H, van Groenou MIB, Kempen GIJM, Penninx BWJH, van Eijk JT,Deeg DJ. Explanations of socioeconomic differences in changes in physicalfunction in older adults: results from the Longitudinal Aging Study Amsterdam.BMC Public Health. 2006;6:244.27. Koster A, Penninx BWJH, Bosma H, Kempen GIJM, Harris TB, Newman AB,Rooks RN, Rubin SM, Simonsick EM, van Eijk JT, et al. Is there a biomedicalexplanation for socioeconomic differences in incident mobility limitation? JGerontol A Biol Sci Med Sci. 2005;60(8):1022–7.28. Institute of Medicine (US) Committee on Health and Behavior. 4. Social riskfactors. In: Health and Behavior: The Interplay of Biological, Behavioral, andSocietal Influences. Washington (DC): National Academic Press; 2001.29. Gorman E, Hanson HM, Yang PH, Khan KM, Liu-Ambrose T, Ashe MC.Accelerometry analysis of physical activity and sedentary behavior in olderadults: a systematic review and data analysis. Eur Rev Aging Phys Act.2014;11:35–49.30. Freedson PS, Melanson E, Sirard J. Calibration of the Computer Science andApplications, Inc. accelerometer. Med Sci Sports Exerc. 1998;30(5):777–81.31. Bassett DR, Troiano RP, McClain JJ, Wolff DL. Accelerometer-based physicalactivity: total volume per day and standardized measures. Med Sci SportsExerc. 2015;47(4):833–8.32. Stewart AL, Mills KM, King AC, Haskell WL, Gillis D, Ritter PL. CHAMPS physicalactivity questionnaire for older adults: outcomes for interventions. Med SciSports Exerc. 2001;33(7):1126–41.33. Enhancing Walk Score’s ability to predict physical activity and activetransportation. http://activelivingresearch.org/files/2013_Bike-WalkScore_Frank.pdf.Accessed 1 Feb 2016.34. Nykiforuk CIJ, McGetrick JA, Crick K, Johnson JA. Check the score: Fieldvalidation of Street Smart Walk Score in Alberta, Canada. Prev Med Rep.2016;4:532–9.35. Duncan DT, Aldstadt J, Whalen J, Melly SJ. Validation of Walk Scores andTransit Scores for estimating neighborhood walkability and transit availability:a small-area analysis. GeoJ. 2013;78(2):407–16.36. Cerin E, Saelens BE, Sallis JF, Frank LD. Neighborhood environmentwalkability scale: validity and development of a short form. Med Sci SportsExerc. 2006;38(9):1682–91.37. Sampson RJ, Raudenbush SW, Earls F. Neighborhoods and violent crime: amultilevel study of collective efficacy. Science. 1997;277(5328):918–24.38. Sampson RJ. Great American City: Chicago and the enduring neighbourhoodeffect. Chicago: The University of Chicago Press; 2012. p. 121–48.39. Groll DL, To T, Bombardier C, Wright JG. The development of a comorbidityindex with physical function as the outcome. J Clin Epidemiol.2005;58(6):595–602.40. Guralnik JM, Simonsick EM, Ferrucci L, Glynn RJ, Berkman LF, Blazer DG,Scherr PA, Wallace RB. A short physical performance battery assessing lowerextremity function: association with self-reported disability and prediction ofmortality and nursing home admission. J Gerontol. 1994;49(2):M85–94.41. Asano M, Miller WC, Eng JJ. Development and psychometric properties ofthe ambulatory self-confidence questionnaire. Gerontology. 2007;53(6):373–81.42. Sun F, Norman IJ, While AE. Physical activity in older people: a systematicreview. BMC Public Health. 2013;13:449.43. Fritz S, Lusardi M. White paper: “walking speed: the sixth vital sign”. J GeriatrPhys Ther. 2009;32(2):46–9.44. Jenkins KR, Fultz NH. The relationship of older adults’ activities and bodymass index. J Aging Health. 2008;20(2):217–34.45. Bracy NL, Millstein RA, Carlson JA, Conway TL, Sallis JF, Saelens BE, Kerr J,Cain KL, Frank LD, King AC. Is the relationship between the built environmentand physical activity moderated by perceptions of crime and safety? Int JBehav Nutr Phys Act. 2014;11(1):24.46. Carlson JA, Sallis JF, Conway TL, Saelens BE, Frank LD, Kerr J, Cain KL, King AC.Interactions between psychosocial and built environment factors in explainingolder adults’ physical activity. Prev Med. 2012;54(1):68–73.47. Ding D, Sallis JF, Norman GJ, Frank LD, Saelens BE, Kerr J, Conway TL, Cain KL,Hovell MF, Hofstetter CR, et al. Neighborhood environment and physicalactivity among older adults: do the relationships differ by driving status? JAging Phys Act. 2014;22(3):421–31.48. Hall KS, McAuley E. Individual, social environmental and physicalenvironmental barriers to achieving 10 000 steps per day among olderwomen. Health Educ Res. 2010;25(3):478–88.49. King AC, Sallis JF, Frank LD, Saelens BE, Cain KL, Conway TL, Chapman J,Ahn DK, Kerr J. Aging in neighborhoods differing in walkability and income:associations with physical activity and obesity in older adults. Soc Sci Med.2011;73(10):1525–33.50. Van Holle V, Van Cauwenberg J, Van Dyck D, Deforche B, Van de Weghe N, DeBourdeaudhuij I. Relationship between neighborhood walkability and olderadults’ physical activity: results from the Belgian Environmental Physical ActivityStudy in Seniors (BEPAS Seniors). Int J Behav Nutr Phys Act. 2014;11:110.51. Hirsch JA, Winters M, Ashe MC, Clarke P, McKay HA. Destinations that olderadults experience within their GPS activity Spaces relation to objectivelymeasured physical activity. Environ Behav. 2016;48(1):55–77.52. Arnardottir NY, Koster A, Van Domelen DR, Brychta RJ, Caserotti P,Eiriksdottir G, Sverrisdottir JE, Launer LJ, Gudnason V, Johannsson E, et al.Objective measurements of daily physical activity patterns and sedentaryChudyk et al. BMC Geriatrics  (2017) 17:82 Page 13 of 14behaviour in older adults: Age, Gene/Environment Susceptibility-ReykjavikStudy. Age Ageing. 2013;42(2):222–9.53. Davis MG, Fox KR, Hillsdon M, Sharp DJ, Coulson JC, Thompson JL. Objectivelymeasured physical activity in a diverse sample of older urban UK adults. MedSci Sports Exerc. 2011;43(4):647–54.54. Sallis JF, Cervero R, Ascher W, Henderson KA, Kraft MK, Kerr J. An ecologicalapproach to creating active living communities. Annu Rev Public Health.2006;27:297–322.55. Martinson BC, Crain AL, Sherwood NE, Hayes MG, Pronk NP, O’Connor PJ.Population reach and recruitment bias in a maintenance RCT in physicallyactive older adults. J Phys Act Health. 2010;7(1):127–35.•  We accept pre-submission inquiries •  Our selector tool helps you to find the most relevant journal•  We provide round the clock customer support •  Convenient online submission•  Thorough peer review•  Inclusion in PubMed and all major indexing services •  Maximum visibility for your researchSubmit your manuscript atwww.biomedcentral.com/submitSubmit your next manuscript to BioMed Central and we will help you at every step:Chudyk et al. BMC Geriatrics  (2017) 17:82 Page 14 of 14

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.52383.1-0362100/manifest

Comment

Related Items