This is a sel -a chi ed e sion o an o iginal a icle. This e sion
may di e om he o iginal in pagina ion and ypog aphic de ails.
Au ho (s):
Ti le:
Yea :
Ve sion:
Copy igh :
Righ s:
Righ s u l:
Please ci e he o iginal e sion:
CC BY 4.0
h ps://c ea i ecommons.o g/licenses/by/4.0/
Associa ions o En i onmen al Fea u es Wi h Ou doo Physical Ac i i y on Weekdays
and Weekend Days : A C oss-Sec ional S udy Among Olde People
© 2020 Keskinen, Gao, Ran akokko, Ran anen and Po egijs.
Published e sion
Keskinen, Ki si E.; Gao, Ying; Ran akokko, Me ja; Ran anen, Taina; Po egijs, E ja
Keskinen, K. E., Gao, Y., Ran akokko, M., Ran anen, T., & Po egijs, E. (2020). Associa ions o
En i onmen al Fea u es Wi h Ou doo Physical Ac i i y on Weekdays and Weekend Days : A
C oss-Sec ional S udy Among Olde People. F on ie s in Public Heal h, 8, A icle 578275.
h ps://doi.o g/10.3389/ pubh.2020.578275
2020
ORIGINAL RESEARCH
published: 30 Oc obe 2020
doi: 10.3389/ pubh.2020.578275
F on ie s in Public Heal h | www. on ie sin.o g 1Oc obe 2020 | Volume 8 | A icle 578275
Edi ed by:
Maw Pin Tan,
Uni e si y o Malaya, Malaysia
Re iewed by:
Lené Le y-S o ms,
UCLA Luskin School o Public A ai s,
Uni ed S a es
Pa icia M. Al ,
Towson Uni e si y, Uni ed S a es
*Co espondence:
Ki si E. Keskinen
[email p o ec ed]
Special y sec ion:
This a icle was submi ed o
Aging and Public Heal h,
a sec ion o he jou nal
F on ie s in Public Heal h
Recei ed: 30 June 2020
Accep ed: 05 Oc obe 2020
Published: 30 Oc obe 2020
Ci a ion:
Keskinen KE, Gao Y, Ran akokko M,
Ran anen T and Po egijs E (2020)
Associa ions o En i onmen al
Fea u es Wi h Ou doo Physical
Ac i i y on Weekdays and Weekend
Days: A C oss-Sec ional S udy
Among Olde People.
F on . Public Heal h 8:578275.
doi: 10.3389/ pubh.2020.578275
Associa ions o En i onmen al
Fea u es Wi h Ou doo Physical
Ac i i y on Weekdays and Weekend
Days: A C oss-Sec ional S udy
Among Olde People
Ki si E. Keskinen1,2*, Ying Gao1,3, Me ja Ran akokko4, Taina Ran anen1and E ja Po egijs1
1Facul y o Spo and Heal h Sciences and Ge on ology Resea ch Cen e , Uni e si y o Jy äskylä, Jy äskylä, Finland,
2School o Resou ce Wisdom, Uni e si y o Jy äskylä, Jy äskylä, Finland, 3Depa men o Spo s Science, College o
Educa ion, Zhejiang Uni e si y, Hangzhou, China, 4School o Heal h and Social S udies, JAMK Uni e si y o Applied
Sciences, Jy äskylä, Finland
Backg ound: Physical ac i i y (PA) o highe in ensi y and longe du a ion mainly
accumula es om olde adul s’ ou -o -home ac i i ies. Ou doo PA is in luenced by
en i onmen al ea u es; howe e , he day- o-day a iabili y o PA and i s associa ions
wi h en i onmen al ea u es ha e no been widely s udied. This s udy ocused on he
associa ions o en i onmen al ea u es wi h accele ome e -measu ed PA in olde people
on weekdays and weekend days.
Me hods: The s udy popula ion comp ised 167 communi y-dwelling olde people
aged 75–90 yea s. Accele ome e s we e wo n on 7 consecu i e days and a s uc u ed
in e iew on physical unc ioning, heal h, and socioeconomic ac o s was adminis e ed.
A geog aphic in o ma ion sys em (GIS) was used o assess en i onmen al ea u es wi hin
a dis ance o 500 (numbe o land ypes, oad ne wo k slope, in e sec ion, and esiden ial
densi ies) o 1,000 m (habi a di e si y wi hin na u al and g een a eas) om pa icipan s’
homes. Accele ome e -based PA [numbe o PA bou s >10 min and minu es o mode a e
o igo ous physical ac i i y (MVPA)] was analyzed o weekdays and weekend days
sepa a ely. Associa ions be ween en i onmen al ea u es and PA we e analyzed using
linea eg ession models.
Resul s: Pa icipan s accumula ed on a e age 0.60 PA bou s and 34.2 MVPA minu es
on weekdays and 0.50 PA bou s and 31.5 MVPA minu es on weekend days. Especially
pa icipan s wi h low o e all PA we e less ac i e a weekends. Habi a di e si y in na u al
and g een a eas, in e sec ion densi y, and esiden ial densi y we e posi i ely associa ed
wi h numbe s o PA bou s and MVPA minu es on weekdays. Mo eo e , mo e di e si y
in na u al and g een a eas was associa ed wi h mo e MVPA minu es on weekend days.
A highe oad ne wo k slope was nega i ely associa ed wi h he numbe o PA bou s
h oughou he week and wi h MVPA minu es on weekend days.
Conclusions: En i onmen al ea u es close o home, especially PA-suppo i e
in as uc u al ea u es and se ices, we e mo e s ongly associa ed wi h weekday han
Keskinen e al. Tempo al Va ia ion in En i onmen -PA Associa ions
weekend PA. This sugges s ha olde people’s ou -o -home ac i i ies, ypically
conduc ed on weekdays, a e ela ed o se ice use. Howe e , g ea e di e si y o na u al
a eas close o home seemed o mo i a e olde adul s o engage in highe MVPA
h oughou he week.
Keywo ds: aging, walking, mobili y, GIS, day- o-day a iabili y
INTRODUCTION
Olde adul s a e ecommended o engage in mode a e o
igo ous in ensi y (MVPA) o a leas 150 min a week (1).
Fo olde people, a la ge p opo ion o hei physical ac i i y
(PA) accumula es du ing daily ac i i ies, such as walking o
anspo , and is no necessa ily exe cise- ela ed (2). Then again,
anspo a ion walking may be a o m o daily exe cise o
some olde adul s o be combined wi h walking o leisu e,
which makes ca ego izing o olde adul s’ PA challenging. Ei he
way, ou -o -home ac i i ies a e associa ed wi h highe PA,
especially when mo ing h ough g ea e li e-space a eas (3). A
p e ious s udy in es iga ing he day- o-day a iabili y o PA
ound signi ican di e ences be ween weekdays and weekend
days in ime spen on PA and concluded ha hese di e ences
in habi ual PA we e p obably explained by daily ou ines and
p ac ices (4). I is no clea whe he en i onmen al ea u es in
he home neighbo hood ela e o PA accumula ion simila ly on
weekdays and weekend days among olde adul s.
Fo Finnish olde adul s, shopping, walking o exe cise, social
isi s, and unning e ands a e among he commones easons o
going ou side he home (5). The ex en o which neighbo hoods
and ci ies o e such des ina ions and a e conduci e o PA
a ies g ea ly (6) and hus is no he same o all olde adul s.
Based on a la ge in e na ional s udy, he di e ence be ween
he leas and he mos ac i i y-suppo ing u ban en i onmen
could mean a di e ence o mo e han 60 min in weekly MVPA
among adul s (6). Recen me a-analyses ha e showed posi i e
associa ions be ween mul iple en i onmen al ea u es and PA (7),
walking o anspo (8), and leisu e- ime PA (9). I has also been
obse ed ha physical unc ioning (10) and socioeconomic (11)
s a us may be in e wined wi h associa ions be ween objec i ely
de ined ea u es o he en i onmen and PA. Howe e , gi en he
a iabili y in daily ou ines and in he a ailabili y o se ices
by he day o he week, en i onmen al ea u es associa ed wi h
PA may di e ac oss days o he week. To lea n mo e abou
age- iendly en i onmen s calls o in o ma ion on indi iduals’
heal h beha io in space and ime (12).
Fea u es such as s ee connec i i y, esiden ial densi y, and
mixed land use, whe he as sepa a e en i onmen al ea u es o in
combina ion o o m a walkabili y index, a e indica i e o se ice
a ailabili y in he en i onmen , and ha e all been posi i ely
associa ed wi h ime spen on MVPA (2,10,11,13,14), al hough
no consis en ly (15–18). MVPA has shown posi i e associa ions
wi h closeness o pa ks (19) and densi y o ec ea ion acili ies
(8), ye he associa ions ha e also appea ed as non-signi ican
(10,13). Fu he mo e, esea ch indica es ha o olde adul s,
walking o a daily des ina ion ypically akes a leas 10 min (20).
Thus, o include habi ual ou doo ac i i ies, i may be necessa y
o cap u e con inuous bou s o PA las ing a leas 10 min. Bou s
o a leas 10 min ha e o a long ime conside ed bene icial o
heal h; howe e , cu en PA guidelines acknowledge he bene i
o any ac i i y and any b eaks in seden a y ime, ega dless o
hei du a ion (1).
Wi h declining unc ion, olde adul s become mo e ulne able
o en i onmen al ba ie s (21). Consequen ly, olde adul s may
modi y hei beha io , e.g., by es ing in he middle o a walk
wi h s eep slopes in he immedia e home en i onmen (22)—
in o he wo ds, by sho ening hei ac i i y bou s. Nega i e
associa ions be ween hilly e ain and walking (23), o al PA
(24), and ec ea ional PA (25) ha e been epo ed, al hough
non-signi ican ela ionships wi h leisu e- ime PA ha e also been
ound (26). Ne e heless, o he bes o ou knowledge, he
associa ions be ween hilly e ain and PA bou s and MVPA in
olde adul s ha e no been s udied.
The pu pose o his s udy was o gain mo e unde s anding
on a iabili y in PA le els be ween weekdays and weekend days
and, especially, on which en i onmen al ea u es may suppo
habi ual PA o olde adul s and when. Fo ha , we examined
he PA le els on weekdays and weekend days and explo ed
he associa ions o en i onmen al ea u es wi h he numbe o
PA bou s and MVPA minu es on weekdays and weekend days
in olde adul s. In addi ion, by applying a me hod simila o
ha used by Sallis e al. (6), we aimed o es ima e whe he
di e ences in he ex en o en i onmen al ea u es suppo ing
PA in a neighbo hood would show p ac ical ele ance o olde
adul s’ PA le els es ima ed as he numbe o PA bou s and
MVPA minu es. We included en i onmen al ea u es conduci e
o PA ha a e ela ed o pe o ming daily e ands (in e sec ion
densi y and esiden ial densi y) and engaging in ec ea ional
ac i i ies (numbe o land ypes, habi a di e si y in g een a eas),
and ea u es hinde ing PA (hilly e ain). Land ype, habi a
di e si y, and slope a e also among he na u al elemen s in
a neighbo hood. In e sec ion and esiden ial densi y, in u n,
a e ea u es o walkabili y, and hus indica e he amoun o
in as uc u e suppo ing ou doo mobili y (27).
MATERIALS AND METHODS
S udy Design
This s udy is pa o he p ojec “Geog aphic cha ac e is ics,
ou doo mobili y and physical ac i i y in old age” (GEOage)
(28). In he p ojec , da a on pa icipan s’ PA we e combined
wi h da a on he en i onmen al cha ac e is ics o hei home
su oundings. The pa icipan da a had ea lie been collec ed in
he p ojec “Li e-space mobili y in old age” (LISPE), which has
F on ie s in Public Heal h | www. on ie sin.o g 2Oc obe 2020 | Volume 8 | A icle 578275
Keskinen e al. Tempo al Va ia ion in En i onmen -PA Associa ions
been desc ibed in de ail elsewhe e (29). Fo he LISPE s udy,
a andom sample o 2,550 communi y-dwelling olde people
aged 75–90 li ing in he neighbo ing municipali ies o Jy äskylä
and Muu ame in Cen al Finland, which he eby o med he
s udy a ea, was d awn om he popula ion egis e in he
win e o 2011/2012. O hese, 848 ul illed he eligibili y c i e ia
(li ing independen ly in he ec ui men a ea, being able o
communica e, and willing o pa icipa e in he s udy) and we e
in e iewed in hei homes, using a s uc u ed ques ionnai e,
in sp ing 2012. Pa icipan s signed a w i en in o med consen
be o e he home in e iew. The e hical commi ee o he
Uni e si y o Jy äskylä, Finland app o ed he LISPE and
GEOage p ojec s.
Resea ch s a ook an accele ome e o he home in e iew
whene e one was a ailable. A e comple ion o he home
in e iew, he in e iewe asked e bally whe he he pa icipan
was willing o pa icipa e in he accele ome e sub-s udy. Based
on pa icipan willingness and accele ome e a ailabili y, a
sub-sample o 190 pa icipan s was assigned o wea a i-
axial accele ome e (Hookie AM20 Ac i i y Me e ; Hookie
Technologies L d, Espoo, Finland) o 7 consecu i e days
ollowing he home in e iew. The pa icipan s we e ins uc ed
o wea he accele ome e (size 6.6 ×2.7 ×1.3 cm, mass 15 g),
which was a ached o an elas ic bel , on hei igh hip du ing
waking hou s. They we e old o ake o he accele ome e
only when engaging in ac i i ies, in which he accele ome e
would ge in o con ac wi h wa e . A e he measu emen
pe iod, pa icipan s e u ned he accele ome e in a p epaid
en elope by mail, o in some cases, he accele ome e was
picked up om hei home. Da a om 16 pa icipan s we e
excluded due o echnical p oblems ending he accele ome e
eco ding ab up ly (n=3), he accele ome e being los in
he mail (n=1), accele ome e wea ime no mee ing he
c i e ion o a leas 10 h pe day (n=11), and in e missions
o mo e han 1 day be ween consecu i e measu emen days
(n=1). O he emaining pa icipan s, only hose wi h alid
measu emen da a o a leas 2 weekdays (Monday–F iday) and
a leas one weekend day (Sa u day–Sunday) we e included in
he s udy (excluded n=7). This esul ed in a inal sample
o 167. On a e age, hose in he accele ome e sub-sample less
equen ly epo ed di icul ies in walking 500 m (15% o sub-
sample pa icipan s) han all he LISPE pa icipan s combined
[26%, χ2
(1) =8.09, p<0.05]. O he wise no di e ences we e
obse ed be ween he wo samples in mean age, p opo ion o
women, yea s o educa ion, o numbe o ch onic condi ions
( o all p≥0.269). Those who ag eed o pa icipa e in he
LISPE s udy we e younge , mo e o en li ed alone, pe cei ed hei
heal h as a leas mode a e, pe cei ed ewe di icul ies in ou doo
mobili y, and mo e equen ly mo ed ou side daily han hose
who declined o pa icipa e (29).
In he GEOage p ojec , pa icipan s’ homes, add esses o
which we e e ie ed om he popula ion egis e , we e loca ed
on a map (30) using a geog aphic in o ma ion sys em (GIS)
(A cMap e sion 10.3; Es i, Redlands, CA, USA). Openly
a ailable geospa ial da ase s we e impo ed o he GIS o
cha ac e ize he en i onmen in he s udy a ea and wi hin 500
o 1,000 m om pa icipan s’ homes. In gene al, he s udy a ea is
cha ac e ized by lakes, o es , low hills, and ela i ely con inuous
a eas o buil en i onmen su ounded by spa sely popula ed
a eas. In he yea 2012, he popula ion o he s udy a ea was
143,000 inhabi an s (31), wi h he majo i y concen a ed in and
a ound he cen e a eas.
S udy Measu es
Physical Ac i i y
Pa icipan s’ PA was objec i ely assessed by an accele ome e
(Hookie AM20 Ac i i y Me e ; Hookie Technologies L d, Espoo,
Finland). The accele ome e eco ds accele a ions along h ee
axes, x, y, and z, i.e., e ical, ho izon al, and pe pendicula ,
espec i ely, and has a dynamic ange o ±16 g, 13 bi s a 100 Hz.
Wi h accele ome e aw da a a ailable o ou use, he esul an
accele a ion o each eco ded sample was calcula ed and used
in all u he analyses. Mean ampli ude de ia ion (MAD) (32)
was calcula ed in non-o e lapping 5-s epochs and subsequen ly
a e aged in 1-min epochs, using a cus om-w i en Ma lab sc ip
(R2015b, Ma hwo ks, Inc., Na ick, MA, USA). The p e-p ocessed
1-min da a we e di ided in o 24-h segmen s om midnigh
o midnigh , and u he p ocessing was done in hose 24-h
segmen s. Non-wea ime was de ined as any con inuous epochs
las ing a leas 1 h wi h all he 1-m MADs below 0.024 g. This non-
wea algo i hm p oduced esul s cong uen wi h sel - epo ed
accele ome e non-wea ime.
PA was assessed om he 1-m MAD epoch alues o each o
he 24-h segmen s. The 1-m alues we e classi ied in o seden a y
(<0.0167 g), ligh PA (0.0167 o <0.091 g), mode a e PA (0.091
o <0.414 g), o igo ous PA (≥0.414 g) a e excluding all
non-wea minu es. The in ensi y cu -o s we e based on he
op imal classi ica ion o ligh PA (0.0167 g) (32), and a MADs
co esponding o 3 me abolic equi alen s (MET, 0.091 g), and 6
METs (0.414 g) o mode a e PA and igo ous PA, espec i ely
(33). MVPA minu es was he sum o he minu es spen in
mode a e PA and igo ous PA. The accumula ion o PA bou s
was assessed based on he 1-m epochs o ligh PA and MVPA
(34) and all ac i e bou s las ing >10 min we e coun ed (35).
F om he numbe o PA bou s las ing >10 min and o al MVPA
minu es in each 24-h segmen , o e all alues we e calcula ed
as he mean o all he 24-h segmen s, weekday alues as he
mean o he 24-h segmen s om Monday o F iday, and weekend
day alues as he mean o he 24-h segmen s om Sa u day o
Sunday o bo h he numbe o PA bou s and MVPA minu es.
Simila ly, accele ome e wea ime o e all, on weekdays, and on
weekend days was calcula ed as he mean alue o he espec i e
24-h segmen s.
En i onmen al Fea u es
Numbe o land ypes [n] was eco ded as coun s o di e en
land ypes wi hin a 500-m adius o he pa icipan ’s home
(36). The o iginal 48 land ype classes in he Co ine Land
Co e da ase (37) we e eclassi ied in o h ee buil and 10
na u al en i onmen land ypes (38). Thus, he alue o he
a iable e lec s he a ia ion p esen especially in he na u al
en i onmen su ounding he pa icipan s’ homes.
Habi a di e si y in na u al and g een a eas [index, ange
0.. .10] was de ined as he highes no malized alue o he
F on ie s in Public Heal h | www. on ie sin.o g 3Oc obe 2020 | Volume 8 | A icle 578275
Keskinen e al. Tempo al Va ia ion in En i onmen -PA Associa ions
Shannon’s Di e si y Index (SHDI) (36) among na u al and g een
a eas, which we e o a minimum size o 10 hec a es and loca ed
a leas pa ly wi hin 1,000 m om pa icipan s’ homes. We
chose a 1,000-m adius o his a iable, as p e ious s udies ha e
shown ha g eenness (39,40) and a ac i e des ina ions (41)
a longe dis ances om he home may be posi i ely associa ed
wi h pa ame e s o heal h and PA. To cap u e he di e si y
o he pa icipan s’ na u al en i onmen s, SHDI alues we e
calcula ed only conside ing he nine na u al en i onmen land
ypes, excluding wa e , included in he eclassi ied Co ine Land
Co e da a (37). To enable meaning ul in e p e a ion o he
esul s, he inal SHDI alue was calcula ed by mul iplying he
o iginal index alues ( ange 0–1) by 10.
Road ne wo k slope was de ined as he a e age slope [% ise]
(whe e 1% poin equals a g adien o 0.45 deg ees) in he 500-
m oad ne wo k (42) o each pa icipan . We used he Digi al
Ele a ion Model a ailable in he 2 m ×2 m as e da ase (43)
o calcula e slope alues o e e y 2-m sec ion o he oads in he
s udy a ea. A pa icipan ’s oad ne wo k slope was he mean o
he oad sec ion slope alues in he 500-m oad ne wo k.
In e sec ion densi y [10 in e sec ions/km2] was calcula ed
as he numbe o in e sec ions wi hin a 500-m adius o he
pa icipan ’s home di ided by he su ace a ea o his zone. To
enable meaning ul in e p e a ion o he esul s, his numbe was
subsequen ly di ided by 10. An in e sec ion was de ined as he
junc ion o a minimum o h ee oads, and in e sec ions wi hin
a dis ance o 10 m om one ano he we e me ged. In calcula ing
his a iable, we used he oad da a in he Topog aphic Da abase
2013 (42).
Residen ial densi y [1,000 esiden s/km2] was de ined as
a e age esiden ial densi y wi hin a 500-m adius o he
pa icipan ’s home. Fo meaning ul in e p e a ion, his numbe
was di ided by 1,000. To calcula e mean alues, we i s
ans o med he o iginal 1 km ×1 km g id d awn om he
Popula ion g id da a 2012 (44) in o ine cells o 100 m ×100 m.
Co a ia es
To accoun o socio-demog aphic di e ences (7,45), age, sex,
and yea s o educa ion we e used as co a ia es in he analyses.
Age and sex we e e ie ed om he na ional popula ion egis e .
Comple ed yea s o educa ion was asce ained du ing he home
in e iew. No impu a ion was made o he h ee missing cases.
The mode a ing e ec s o he pa icipan s’ physical unc ioning
on he associa ion be ween he buil en i onmen and physical
ac i i y (46–48) we e e alua ed wi h he a iables walking
di icul ies and numbe o ch onic condi ions. Di icul ies in
walking 500 m we e asce ained du ing he in e iew, and hose
epo ing a leas some di icul ies we e assigned o he ca ego y
o pe cei ed walking di icul ies ( s. no walking di icul ies).
Sel - epo ed numbe o ch onic condi ions was summed based
on a lis o 22 physician-diagnosed ch onic condi ions and an
addi ional open-ended ques ion (29).
Analysis
Pa icipan s and hei en i onmen s we e cha ac e ized by
means, s anda d de ia ions and anges, by medians and
in e qua ile anges (IQR), o by pe cen ages depending on
a iable dis ibu ion. Dis ibu ions o PA a iable alues be ween
weekdays and weekend days we e compa ed using he Wilcoxon
signed- ank es and Pea son chi-squa e es . Bland-Al man plo s
wi h 95% con idence in e als o di e ences in mean alues
be ween indi iduals’ weekday and weekend day PA, including
eg ession lines (49), we e c ea ed o PA bou s and MVPA ime.
Bland-Al man plo s a e especially sui ed o isually de ec ing
di e ences in co esponding alues om epea ed measu emen s
o om wo measu emen me hods along he measu emen
scale (50).
The PA a iables showed igh -skewed dis ibu ions and,
excep o mean MVPA on weekdays, included a subs an ial
p opo ion o ze o alues. Hence, o he u he analysis, he
PA a iable alues we e ans o med using a na u al loga i hm
a e adding he alue o one o emo e ze os. Associa ions o
each en i onmen al a iable wi h each o he log- ans o med PA
a iables we e s udied using linea eg ession. Associa ions we e
adjus ed i s o age, sex, and accele ome e wea ime (Model
1), and hen—one a a ime due o he ela i ely small sample
size— o pe cei ed di icul ies in walking 500 m (Model 2), yea s
o educa ion (Model 3), and ch onic condi ions (Model 4). To
in e p e he esul s, he alues o coe icien s and con idence
in e als om he log- ans o med linea eg ession analyses
we e exponen ia ed (ma ked as expβ). In he esul ables, he
epo ed expβ alue equals he p opo ional change in he
ou come a iable alue ob ained om a one-uni inc ease in he
p edic o alue. Addi ional sensi i i y analyses we e pe o med
by e unning he analyses wi h a sample es ic ed o pa icipan s
wi h alid da a o 3 o mo e weekdays and bo h weekend days
(n=139).
To es ima e he magni ude o en i onmen al e ec s on PA, we
pe o med calcula ions using he esul s o he Model 2 linea
eg ession analyses, he high-low di e ence in en i onmen al
a iable alues wi hin he s udy a ea, and he median alues o
he numbe o PA bou s and MVPA minu es o e all. Coe icien s
and con idence in e als o en i onmen al a iables showing
s a is ically signi ican ly associa ions wi h PA we e mul iplied by
he alue o he high-low di e ence, which was de ined as he
di e ence be ween he means o he highes and lowes 10%
o he en i onmen al a iable alues. This p oduc was hen
exponen ia ed o show he p opo ional e ec on PA o he
high-low di e ence o he en i onmen al a iable in ques ion.
Fu he mo e, o exp ess he absolu e e ec s o he high-low
di e ence o he en i onmen al a iables in he numbe o PA
bou s and MVPA minu es, he p opo ional e ec alues we e
mul iplied by he espec i e PA median alues.
Da a analyses we e pe o med wi h IBM SPSS S a is ics o
Windows ( e sion 24.0; IBM Co p., A monk, NY, USA) and R
so wa e e sion 3.5.2 (51). S a is ical signi icance was se a p<
0.05 in all es s.
RESULTS
Pa icipan and neighbo hood cha ac e is ics a e p esen ed in
Table 1. Pa icipan mean age was 80.3 yea s, 65% we e women,
and 15% pe cei ed walking di icul ies (Table 1).
F on ie s in Public Heal h | www. on ie sin.o g 4Oc obe 2020 | Volume 8 | A icle 578275

Keskinen e al. Tempo al Va ia ion in En i onmen -PA Associa ions
Va iabili y in Physical Ac i i y Le els
Be ween Weekdays and Weekend Days
The median numbe o accumula ed PA bou s on weekdays and
weekend days we e simila (p=0.646), wi h alues below 1
indica ing ha he majo i y o he pa icipan s did no engage
in a 10-min bou o con inuous mo emen e e y day (Table 2).
The median ime spen in MVPA was abou hal an hou and did
no di e be ween weekdays and weekend days (p=0.125). On
a e age, pa icipan s wo e he accele ome e o a longe ime
on weekdays han weekend days (p<0.001). The numbe o
pa icipan s accumula ing ze o PA bou s was highe on weekend
days (72 pa icipan s) han weekdays (43 pa icipan s) [χ2
(1) =
48.366, p<0.001].
The Bland-Al man plo da a on indi iduals showed ha he
pa icipan s accumula ed on a e age 0.06 mo e PA bou s [ (166)
=1.015, p=0.311] and 1.64 ewe MVPA minu es [ (166) =
−0.972, p=0.332] on weekend days compa ed o weekdays;
howe e , hese mean di e ences did no s a is ically signi ican ly
di e om ze o. Howe e , pa icipan s wi h lowe PA in gene al
o en exhibi ed nega i e di e ences, indica ing g ea e ac i i y on
weekdays compa ed o weekend days (Figu es 1,2).
En i onmen al Fea u es Associa ed Wi h
Physical Ac i i y on Weekdays and
Weekend Days
The linea eg ession analyses adjus ed o age, sex, a e age
accele ome e wea ime, and walking di icul ies yielded
TABLE 1 | Cha ac e is ics o pa icipan s and hei neighbo hood en i onmen s
(n=167).
Mean (S anda d Range
de ia ion) Min–Max
Pa icipan cha ac e is ics
Age [yea s] 80.3 (4.2) 74.3–89.3
Educa ion [yea s] 9.9 (4.0) 2–25
Ch onic condi ions [n] 4.3 (2.3) 0–11
Women [%] 65
Walking di icul ies [%] 15
Neighbo hood cha ac e is ics
Land ypes [n] 6.1 (1.2) 4–9
Habi a di e si y [10*SHDI] 4.3 (1.2) 1.3–7.1
Slope [% ise] 2.1 (0.6) 1.1–3.9
In e sec ion densi y [10 c ossings/km2] 5.9 (2.4) 0.5–10.9
Residen ial densi y [1,000 esiden s/km2] 1.8 (1.5) 0.002–5.0
di e en en i onmen -PA associa ions be ween weekdays and
weekend days (Table 3). On weekdays, habi a di e si y in na u al
and g een a eas, in e sec ion densi y, and esiden ial densi y
we e posi i ely associa ed wi h bo h he numbe o PA bou s
and MVPA minu es. On weekend days, only habi a di e si y
was posi i ely associa ed wi h MVPA and had a sligh ly lowe
coe icien alue han on weekdays. Road ne wo k slope showed
a nega i e associa ion wi h PA bou s on weekdays and weekend
days and wi h MVPA on weekend days only. Numbe o land
ypes was no associa ed wi h PA in any o he analyzed models.
Addi ional File 1 also shows he esul s o he models in which
associa ions we e adjus ed o age, sex, and accele ome e wea
ime (Model 1), and also o yea s o educa ion (Model 3),
and ch onic condi ions (Model 4). No subs an ial di e ences
in esul s we e obse ed be ween he models, excep o slope,
which was associa ed wi h MVPA only in he model adjus ed
o walking di icul ies (Model 2). The esul s o e unning he
models in he sensi i i y analyses by including pa icipan s wi h
alid accele ome e da a o a leas 5 days la gely esembled
hose o he ull sample, wi h some associa ions in he ull
sample wi h p<0.100 eaching s a is ical signi icance and some
associa ions u ning s a is ically non-signi ican (p- alue be ween
0.050 and 0.100) (see Addi ional File 2).
E ec s o High-Low Di e ences in
En i onmen al Fea u es on Physical
Ac i i y
The p opo ional and absolu e e ec s on PA o he high-low
di e ences in en i onmen al ea u es in he s udy a ea a e shown
in Table 4. The e ec s we e es ima ed in hose en i onmen al
ea u es, which we e s a is ically signi ican ly associa ed wi h
numbe o PA bou s and MVPA minu es o e all on all days. The
highes e ec s we e de ec ed in esiden ial densi y, in which he
high-low di e ence o 4.55 housand people pe km2 esul ed in
an inc ease o 36% in he numbe o PA bou s and 74% in MVPA
minu es o e all (p opo ional e ec s), which co esponded o
0.21 mo e PA bou s and 23.1 mo e MVPA minu es daily (absolu e
e ec s). Al oge he , he p opo ional e ec s o he en i onmen al
a iables posi i ely associa ed wi h PA a ied be ween 21 and
36% o PA bou s and be ween 57 and 74% o MVPA minu es.
Depending on he high-low di e ence in he en i onmen al
ea u e in ques ion, hese esul ed in absolu e e ec s be ween 0.12
and 0.21 mo e PA bou s and be ween 18.0 and 23.1 mo e MVPA
minu es daily. Slope was he only en i onmen al ea u e showing
a nega i e e ec , wi h he high-low di e ence esul ing in a 26%
TABLE 2 | Accele ome e -de i ed PA and wea ime o e all (all days) and on weekdays and weekend days (n=167).
O e all Weekdays Weekend days Wilcoxon signed- ank es a
Median (IQR) Median (IQR) Median (IQR) p- alue
PA bou s [n/day] 0.57 (0.14, 1.33) 0.60 (0.00, 1.33) 0.50 (0.00, 1.50) 0.646
MVPA [min/day] 31.3 (16.7, 54.4) 34.2 (17.8, 56.4) 31.5 (10.5, 50.0) 0.125
Accele ome e wea ime [h/day] 13.5 (12.7, 14.3) 13.7 (12.7, 14.5) 13.2 (12.3, 14.2) <0.001
aCompa ison be ween weekday and weekend day alues. Values in bold; p <0.05. IQR, In e qua ile ange (25%, 75%).
F on ie s in Public Heal h | www. on ie sin.o g 5Oc obe 2020 | Volume 8 | A icle 578275
Keskinen e al. Tempo al Va ia ion in En i onmen -PA Associa ions
FIGURE 1 | Bland-Al man plo o numbe o PA bou s on weekend days s. weekdays (n=167). Mean di e ence wi h 95% con idence in e als and eg ession line
(R2=0.036, p=0.014 o β).
dec ease in PA bou s and 33% in MVPA, equi alen o 0.15 ewe
PA bou s and 10.3 ewe MVPA minu es daily.
DISCUSSION
Ou esul s show ha se e al en i onmen a iables we e
associa ed wi h PA on weekdays and ewe on weekend days.
The posi i e associa ions o in e sec ion and esiden ial densi ies
wi h numbe o PA bou s and wi h MVPA minu es ound
only o weekdays indica e ha li ing in en i onmen s wi h a
highe amoun o in as uc u e suppo ing ou doo mobili y
and wi h he close p oximi y o se ice des ina ions is especially
conduci e o PA on weekdays bu o less ele ance a weekends.
Howe e , en i onmen al ea u es pe aining o na u al elemen s
we e mo e consis en ly ela ed o PA i espec i e o he day
o he week. Highe habi a di e si y in na u al and g een
a eas was associa ed wi h mo e MVPA ime on bo h weekdays
and weekend days. The esul s sugges ha olde people may
engage in pa ially di e en ac i i ies du ing weekdays compa ed
o weekends.
P e ious s udies suppo ou esul showing a highe
likelihood o PA on weekdays in a eas wi h highe in e sec ion
and esiden ial densi ies. In hose, neighbo hood walkabili y has
shown a posi i e, nea ly s a is ically signi ican end in he
numbe o MVPA bou s las ing a leas 10 min (15). Walking
acili ies, in e sec ion densi y, mixed land use, and densi y o
ec ea ion cen e s we e posi i ely associa ed wi h walking o
e ands (19). Walking o anspo , ypically o des ina ions
a leas 10 min away, was posi i ely associa ed wi h numbe
o neighbo hood ameni ies and also, o a ce ain ex en , s ee
connec i i y (20). Thus, i seems ha o olde adul s, ha ing
des ina ions o daily e ands wi hin walking dis ance may play a
majo ole in accumula ing longe las ing and b isk PA, especially
on weekdays. Ye in en i onmen s p o iding mo e se ices,
oppo uni ies o pa icipa e in o he meaning ul ac i i ies such
as o ganized ac i i ies may be g ea e as well.
Habi a di e si y in na u al and g een a eas was posi i ely
associa ed wi h MVPA minu es on bo h weekdays and weekend
days as well as wi h PA bou s on weekdays. These esul s
a e in line wi h p e ious indings ha , in olde adul s,
p oximi y o o he a ailabili y o a pa k a e posi i ely
F on ie s in Public Heal h | www. on ie sin.o g 6Oc obe 2020 | Volume 8 | A icle 578275
Keskinen e al. Tempo al Va ia ion in En i onmen -PA Associa ions
FIGURE 2 | Bland-Al man plo o MVPA minu es on weekend days s. weekdays (n=167). Mean di e ence wi h 95% con idence in e als and eg ession line (R2=
0.052, p=0.003 o β).
associa ed wi h objec i ely measu ed daily MVPA (19) and
sel - epo ed leisu e- ime PA (9) and ha se e al ea u es o
ec ea ional des ina ions a e posi i ely associa ed wi h sel -
epo ed ec ea ional walking (52). Among olde adul s wi h
good walking capabili y, na u al a eas wi h highe di e si y ha e
also been ela ed o highe sel - epo ed PA and o pe cep ions
o nea by na u e as a mo i a o o go ou side he home (38).
Howe e , no conclusi e e idence on he associa ions be ween
he a ailabili y o pa ks and ec ea ion a eas and sel - epo ed
ac i e a el among olde adul s has p e iously been epo ed
(45). Thus, based on he p esen esul s, i seems ha a ac i e
na u e-based des ina ions close o home may be impo an
acili a o s o ou doo leisu e- ime and longe las ing and
highe in ensi y exe cise- ype PA h oughou he week among
olde adul s.
In con as , oad ne wo k slope was nega i ely associa ed wi h,
in pa icula , long-las ing PA h oughou he week and wi h
MVPA on weekend days in ou s udy. These esul s a e in line
wi h ea lie obse a ions (23,24). In addi ion, ou ea lie s udies
showed ha pe cei ing hilliness as an ou doo mobili y ba ie
p edic ed maladap i e walking modi ica ions, i.e., educing he
equency o walking o gi ing up walking (22) and ha s eepe
oads in he home neighbo hood p edic ed he de elopmen o
walking di icul ies (53). In he cu en s udy, he lowe numbe
o PA bou s and ewe MVPA minu es obse ed among olde
adul s who li e su ounded by a s eep oad ne wo k, suppo s
he ea lie inding ha s eep slopes hinde he daily walking o
olde adul s.
In he cu en s udy, he p opo ion o pa icipan s
wi h no 10-min PA bou s was highe du ing weekend
days han weekdays. Those who we e less ac i e o e all
especially ended o accumula e a lowe numbe o PA
bou s and ewe MVPA minu es on weekend days han
weekdays. These obse a ions suppo Ma shall e al. (54),
who ound, among hose who we e he mos seden a y
o e all, ha seden a y ime on weekend days was g ea e
han on an a e age weekday. In ou s udy, PA anges we e
na owe on weekdays han weekend days, which migh
sugges highe s abili y in PA beha io on weekdays han a
weekends. Highe a iabili y in weekend-day PA le els, as
F on ie s in Public Heal h | www. on ie sin.o g 7Oc obe 2020 | Volume 8 | A icle 578275
Keskinen e al. Tempo al Va ia ion in En i onmen -PA Associa ions
TABLE 3 | Associa ions o en i onmen al ea u es wi h PA bou s and MVPA on weekdays and weekend days (n=167).
Numbe o PA bou s MVPA minu es
Weekdays Weekends Weekdays Weekends
expβ(95% CI) expβ(95% CI) expβ(95% CI) expβ(95% CI)
Land ypes [n] 1.00 (0.94–1.05) 1.04 (0.98–1.10) 0.99 (0.89–1.09) 1.05 (0.94–1.17)
Habi a di e si y [10*SHDI] 1.07 (1.01–1.12) 1.02 (0.97–1.08) 1.16 (1.06–1.28) 1.11 (1.00–1.23)
Slope [% ise] 0.86 (0.77–0.96) 0.85 (0.75–0.97) 0.82 (0.66–1.01) 0.78 (0.62–0.98)
In e sec ion densi y [10 c ossings/km2]1.03 (1.01–1.06) 1.01 (0.98–1.05) 1.07 (1.01–1.12) 1.03 (0.97–1.09)
Residen ial densi y [1,000 esiden s/km2]1.09 (1.04–1.13) 1.03 (0.98–1.08) 1.15 (1.06–1.25) 1.08 (0.99–1.18)
Resul s o Model 2; adjus ed o age, sex, accele ome e wea ime, and di icul ies in walking 500 m. Values in bold; p <0.05. An iloga i hm alues o uns anda dized eg ession
coe icien s (expβ) and hei 95% con idence in e als (CI) om uni a ia e linea eg ession models show a p opo ional e ec o a one-uni inc ease in he p edic o alue on he ou come
a iable alue (e.g., a one-uni inc ease in esiden ial densi y, equaling an inc ease o 1,000 esiden s in a 1-km2a ea, shows a 9% inc ease in he numbe o PA bou s and 15% inc ease
in MVPA minu es on weekdays).
TABLE 4 | P opo ional and absolu e e ec s o high-low di e ences in en i onmen al a iables on PA (n=167).
En i onmen al a iable High-low di e ence [in
en i onmen al a iable uni s]
expβ(95% CI)aP opo ional e ec
(95% CI)
Absolu e e ec [in
uni s o PA a iable]
Numbe o PA bou s o e all [n/day]
Land ypes [n] 4.12 1.01 (0.96–1.06)
Habi a di e si y [10*SHDI] 3.84 1.05 (1.00–1.10) 1.21 (1.01–1.45) 0.12
Slope [% ise] 1.92 0.86 (0.77–0.95) 0.74 (0.61–0.91) −0.15
In e sec ion densi y [10 c ossings/km2] 8.37 1.03 (1.00–1.05) 1.26 (1.02–1.56) 0.15
Residen ial densi y [1,000 esiden s/km2] 4.55 1.07 (1.03–1.11) 1.36 (1.13–1.64) 0.21
MVPA minu es o e all [min/day]
Land ypes [n] 4.12 1.00 (0.91–1.10)
Habi a di e si y [10*SHDI] 3.84 1.15 (1.05–1.26) 1.72 (1.22–2.41) 22.4
Slope [% ise] 1.92 0.81 (0.67–0.99) 0.67 (0.46–0.98) −10.3
In e sec ion densi y [10 c ossings/km2] 8.37 1.06 (1.01–1.11) 1.57 (1.05–2.36) 18.0
Residen ial densi y [1,000 esiden s/km2] 4.55 1.13 (1.05–1.22) 1.74 (1.23–2.47) 23.1
aAn iloga i hm alues o uns anda dized eg ession coe icien s (expβ) and hei 95% con idence in e als (CI) o uni a ia e linea eg ession Model 2 adjus ed o age, sex, accele ome e
wea ime, and pe cei ed di icul ies in walking 500 m. Values in bold; p <0.05.
also epo ed by Abel e al. (4), may pa ly explain why ewe
en i onmen al ea u es we e ela ed o PA le els on weekend
days han weekdays.
The es ima ed po en ial e ec s on PA le els o high-low
di e ences in he alues o he en i onmen al a iables showed
ha en i onmen al cha ac e is ics could ha e p ac ical ele ance
o olde adul s’ PA. Based on ou es ima ion, he numbe
o PA bou s was 36% g ea e o pa icipan s li ing in high
s. low esiden ial densi y a eas. Simila ly, 26% ewe PA
bou s we e es ima ed o hose li ing in high s. low slope
a eas. Mo eo e , o pa icipan s li ing in a high habi a
di e si y a ea we e calcula ed 72% mo e minu es o MVPA
han o hei coun e pa s li ing in a low habi a di e si y
a ea. Al hough conside ably highe , ou esul s pa allel hose
o Sallis e al. (6), who, in a s udy on adul s in 14 ci ies
wo ldwide, es ima ed om 14 o 21% mo e weekly MVPA
minu es o hose li ing in neighbo hoods in he highes
5% o PA-suppo i e en i onmen al ea u es (in e sec ion and
esiden ial densi ies, numbe o pa ks) compa ed o hose li ing
in a eas in he lowes 5% o hese ea u es. Howe e , he
la ge con idence in e als in ou hypo he ical es ima es a e
a eminde ha , a he han exac numbe s, he associa ions
and hei di ec ions mus be conside ed when assessing he
p ac ical implica ions o hese esul s. Fu he mo e, i is no
possible o gene alize hese esul s o he popula ion le el,
owing o he s udy a ea-speci ic anges in he en i onmen
a iables and a s udy sample consis ing o olde people
wi h be e han a e age unc ional capabili y. Howe e , his
es ima ion exe cise shows ha e en wi hin a ela i ely small
s udy a ea, such as he one s udied he e, he cha ac e is ics
o di e en neighbo hood en i onmen s can a y in ways
ha , depending on hei home loca ion, a o some adul s
mo e han o he s wi h espec o he ex en o which hey
suppo ou doo PA. Howe e , since indi idual heal h and
psychosocial ac o s seem o explain a la ge pa o olde
adul s’ PA, he con ibu ion o en i onmen al ac o s o PA
le els is necessa ily limi ed (19). Ne e heless, he en i onmen -
PA associa ions ha we ound we e no no ably a ec ed by
F on ie s in Public Heal h | www. on ie sin.o g 8Oc obe 2020 | Volume 8 | A icle 578275