summarize and appraise a quantitative study to determine its potential usefulness to inform nursing practice

R E S E A R CH A R T I C L E

Effects of sedentary behavior and physical activity on sleep quality in older people: A cross-sectional study

Jaehoon Seol MS1 | Takumi Abe PhD2 | Yuya Fujii MS1 | Kaya Joho MS3 |

Tomohiro Okura PhD4

1Graduate School of Comprehensive Human

Sciences, Doctoral Program in Physical

Education, Health and Sport Sciences,

University of Tsukuba, Tsukuba, Japan

2Research on Healthy Aging and Community

Health, Tokyo Metropolitan Institute of

Gerontology, Tokyo, Japan

3Graduate School of Comprehensive Human

Sciences, Doctoral Program in Human Care

Sciences, University of Tsukuba, Tsukuba,

Japan

4Faculty of Health and Sport Sciences,

University of Tsukuba, Tsukuba, Japan

Correspondence

Jaehoon Seol, Graduate School of

Comprehensive Human Sciences, Physical

Education, Health and Sport Sciences,

University of Tsukuba, 1-1-1 Tennodai,

Tsukuba, Ibaraki 305-8574 Japan.

Email: [email protected]

Funding information

Otsuka Toshimi Scholarship Foundation

(Osaka, Japan)

Abstract

The aim of this cross-sectional study was to investigate the influence of replacing

sedentary time with time engaged in one of two levels of physical activity on sleep

quality using an isotemporal substitution model. The participants were 70 commu-

nity-dwelling older Japanese adults (approximately 70% female). Physical activity

types were measured using a triaxial accelerometer and categorized based on inten-

sity as sedentary, light-intensity, and vigorous-intensity. The Pittsburgh Sleep Quality

Index assessed subjective sleep quality. Objective sleep parameters were assessed

using an actigraph. A series of multi-linear regression models analyzed the statistical

relationships. Our findings showed that replacing 30 min of sedentary activity per

day with an equal period of light-intensity physical activity significantly influenced

sleep quality parameters. However, there was no significant difference in sleep qual-

ity when light-intensity activity was replaced with vigorous-intensity activity. Engag-

ing in one activity type means less available time for other types of activity; habitual

replacement of sedentary activity with light-intensity physical activity might have

long-term benefits on the sleep quality of older people.

K E YWORD S

insomnia, Japan, older people, physical activity, sedentary behavior

1 | INTRODUCTION

About 36.2% of Japanese older people suffer from sleep disorders

linked to adverse health outcomes (Hishikawa et al., 2017), such as

mortality, falls, cognitive impairment, and depressive symptoms

(Okajima, Komada, Nomura, Nakashima, & Inoue, 2012; Stone et al.,

2008; Yaffe, Falvey, & Hoang, 2014). To reduce adverse outcomes, it

is important to develop evidence-based public health strategies to

decrease sleep disorder symptoms. Sleeping pills are often prescribed

to relieve sleep disorder symptoms in older people. However, several

studies have found that, despite their effectiveness for some people,

there are also side-effects, such as decreased cognitive functioning

(Brzecka et al., 2018) and increased risk of falls (Stone et al., 2008),

which often become chronic problems (Youngstedt, O'Connor,

Crabbe, & Dishman, 2000). Moreover, diminished sleep quality is a

standard feature of aging and often leads to sleep disorders (Mander,

Winer, & Walker, 2017); it is therefore important to alleviate the

symptoms of poor sleep quality before they result in sleep disorders.

Recently, the American Geriatrics Society, referring to the Beers

Criteria, strongly recommended that older people avoid sleeping pills

because of the harmful side-effects (American Geriatrics Society

Beers Criteria Update Expert Panel, 2019). In line with this recom-

mendation, research has increasingly focused on the improvement of

sleep quality using non-pharmacological methods, usually involving

cognitive behavioral therapy, timed exposure to bright light therapy,

and exercise (Riemann et al., 2017). However, with the exception of

exercise, these therapies tend to be prescribed for people with insom-

nia only (Bloom et al., 2009).

Received: 12 April 2019 Revised: 7 August 2019 Accepted: 13 August 2019

DOI: 10.1111/nhs.12647

64 © 2019 John Wiley & Sons Australia, Ltd Nurs Health Sci. 2020;22:64–71.wileyonlinelibrary.com/journal/nhs

1.1 | Literature review

A previous systematic review concluded that adequate exercise suffi-

ciently improved the sleep quality of older people (Buman & King,

2010). Findings from epidemiological studies have further suggested

that a lack of regular physical activity (PA) predicts insomnia (Morgan,

2003; Tsunoda et al., 2015). Compared to the use of hypnotic medica-

tions, PA provides a free or low-cost way to promote health benefits,

with relatively low risk of adverse side-effects (Driver & Taylor, 2000).

Consequently, engaging in PA is a useful, non-pharmacological way to

ensure high-quality sleep.

As people age, their physical activities tend to decrease, while sed-

entary activities (i.e. watching television, reading books, and using the

computer) increase (Harvey, Chastin, & Skelton, 2013). An increasing

number of studies have noted the adverse health effects of sedentary

activities for older people (e.g. de Rezende, 2014), including poor sleep

quality (Madden, Ashe, Lockhart, & Chase, 2014), although the precise

mechanisms underlying this effect are unclear. Previous studies agree

that PA and sedentary activity relate to sleep quality in older people

(Buman & King, 2010; Madden et al., 2014), but these studies have sev-

eral limitations. The current study aimed to address these limitations.

An individual's discretionary time is limited to 24 h per day, mean-

ing that some activities, such as sedentary activities, light-intensity PA

(LPA), and moderate- to vigorous-intensity PA (MVPA), are mutually

exclusive (Yasunaga et al., 2017). That is, engaging in one type of

activity (e.g. sedentary) necessarily precludes engagement in a differ-

ent type (e.g. MVPA). Studies have investigated the comprehensive or

independent associations of PA or sedentary activities with sleep pat-

terns, but did not consider this mutual exclusivity. Recent technologi-

cal advances have resulted in the development of wearable devices,

such as FitBit and Apple Watch, for tracking an individual's fitness

parameters, including sleep, in a natural environment. These devices

can measure the intensity of an activity and distinguish between sed-

entary and more active behavior as being mutually exclusive. It seems

reasonable to predict that sleep quality might improve when PA dis-

places sedentary activity, as measured by a wearable fitness device.

However, to date, no research has investigated the influences of

replacing one type of activity with another on the sleep parameters of

older people. Moreover, few studies have investigated the influences

of PA intensity on sleep quality.

Our study provides useful information for the improvement of

sleep quality among health professionals and older people. Health

professionals might use the results to enhance the prevention and

wellness aspects of their treatment plans for older adults, and it might

enable caregivers and older people to increase their daily

PA. Therefore, this study supports the development of public health

strategies for older people with or at risk of developing sleep

disorders.

1.2 | Study purpose

The aim of this study was to examine the influences of replacing time

spent on sedentary activity with time spent on LPA or MVPA, on the

subjective and objective sleep quality of Japanese older adults. The

study took an empirical approach by using accelerometers, such as

the isotemporal substitution model (Yasunaga et al., 2017), and a

novel statistical method.

2 | METHODS

2.1 | Design and sample

This cross-sectional study was conducted from May to September

2017. Participants were recruited in May 2017 through an advertise-

ment in a newspaper distributed in Tsukubamirai city, Ibaraki prefec-

ture. Inclusion criteria were: (i) age >65 years; (ii) not using any sleep

medication; and (iii) not restricted from exercising by a physician. Sev-

enty four volunteers agreed to participate. The exclusion criteria

were: (i) a current diagnosis of sleep apnea syndrome (n = 1); or

(ii) <4 days of accelerometer data (n = 3). Finally, daily data were on

70 participants were collected from May to July 2017, in their usual

environments.

2.2 | Ethical approval and consent to participate

The ethics committee of the University of Tsukuba (ref no. Tai 28-68)

approved the study. All participants received oral and written informa-

tion on the research, and provided signed informed consent forms

prior to participation.

2.3 | Variables and measures

2.3.1 | Physical and sedentary activities

PA and sedentary activity times were measured using a triaxial accel-

erometer. Detailed algorithms and validation of this instrument for

evaluating PA and sedentary activities have previously been reported

(Kurita et al., 2017; Ohkawara et al., 2011; Oshima et al., 2010). The

accelerometer was programmed to record 60 s intervals, and partici-

pants wore it on the left side of their waists for 1 week. Daily PA was

measured as number of minutes per day by dividing all activities into

three types: (i) sedentary (1.0–1.5 metabolic equivalents); (ii) LPA

(1.6–2.9 metabolic equivalents); and (iii) MVPA (≥3.0 metabolic equiv-

alents) (Owen, Healy, Matthews, & Dunstan, 2010). The three activity

types were then scored into units of 30 min per day to facilitate inter-

pretation of the results.

The participants were required to wear the device during all wak-

ing hours, except when they engaged in water-based activities, such

as swimming or bathing. The time when they were not wearing the

device (non-wearing time) was measured in intervals of at least

60 consecutive minutes of zero counts (e.g. activity less than the

detection threshold), and the total wearing time was computed as

24 h minus non-wearing hours. During the study period, days with

more than 10 h of wearing time were suitable for analysis, and partici-

pants with four or more valid days were included in the analysis

(Mâsse et al., 2005).

SEOL ET AL. 65

2.3.2 | Sleep parameters

Sleep parameters were measured using an actigraph, worn on the

wrist of the non-dominant hand for 1 week. The sleep data were col-

lected at a sampling rate of 30 Hz and aggregated to 60 s periods rep-

resenting activity counts per minute. The data were analyzed using

ActiLife (version 6.13.3) software and the Cole–Kripke algorithm

(Cole, Kripke, Mullaney, & Gillin, 1992). Using the data of the vertical

axis, the results were scored in 60 s periods as either “awake” or

“asleep”, by measuring movement during each minute and movement

immediately before and after each period. The participants were

instructed to record a sleep log of their bedtimes and wake times, and

the sleep parameters were calculated using these data.

The five sleep parameters in the analysis were sleep onset latency,

sleep efficiency, minutes of wake after sleep onset, number of awak-

enings, and sleep fragmentation index. We considered a positive cor-

relation between the amount of time in bed with waking after sleep

onset and number of awakenings per sleep. Therefore, we measured

the amount of time in bed using the actigraph sleep data. Sleep data

on days where no activity data were measured by the accelerometer

were excluded from the analysis.

2.3.3 | Sleep quality

Sleep quality was subjectively measured using a derivation of the Pitts-

burgh Sleep Quality Index (PSQI) (Buysse, Reynolds, Monk, Berman, &

Kupfer, 1989), a self-rated questionnaire to assess an individual's sleep

quality over 1 month. The PSQI instrument has been validated as valid

and reliable for the identification of sleeping disorders in clinical and

research settings (Doi et al., 2000), and we obtained written permission

for its use. The questionnaire comprised seven items scored by the par-

ticipants (response range: 0–3) as follows: subjective sleep quality, sleep

onset latency, sleep duration, habitual sleep efficiency, sleep distur-

bances, use of sleep medication, and daytime dysfunction. The responses

on the items were summed into a composite score (range: 0–21).

2.3.4 | Potential control variables

Based on previous studies (e.g. Ancoli-Israel, 2005), we included age,

sex, body mass index, a medical history of hypertension (“yes” or “no”),

alcohol consumption (“drinker” or “non-drinker”), tobacco smoking sta-

tus (“current” or “previous/never”), and depressive symptoms, assessed

by the Japanese version of the 15 item Geriatric Depression Scale

(Niino, Imaizumi, & Kawakami, 1991), which is free to use.

2.4 | Analysis

Multi-linear correlation regression analyses were conducted to inves-

tigate the relationships of activity type (sedentary, LPA, and MVPA) to

the sleep parameters and estimated unstandardized regression coeffi-

cients (B) with 95% confidence intervals (CI). The activity type was

analyzed in 30 min units. The potential control variables were used in

all the statistical models. When the measures of “wake after sleep

onset” and “number of awakenings” were added to the model, we

included the measure of “time in bed.”

Multi-collinearity was evaluated using variance inflation factors,

which in all of the models, were <2.6, indicating acceptably weak

multi-collinearity. We confirmed the reliability of all subordinate items

of the PSQI by computing the Cronbach's alpha coefficient (.719, with

the exception of the use of sleep medication, where Cronbach's

α = .739); indicating that the data had acceptable internal consistency.

All analyses were performed using SPSS version 24.0. Statistical sig-

nificance was accepted at P < .05 (two tailed) or better in all models.

2.4.1 | Single-factor and partition models

Three single-factor regression models estimated the relationships

between the three activity type variables and sleep parameters, con-

trolling for the effects of possible confounding factors and total wear-

ing time. The sedentary activity model was estimated using equation

1 as follows:

Sleep parameters variable = sedentary activity + total wear time

+ confounding factors ð1Þ

The partitioned model examined the net influence of each activity

type variable on sleep quality, controlling for the effects of all the

other activity type variables and the control variables on the sleep

parameters expressed as equation 2 as follows:

Sleep parameters variable = sedentary activity + LPA+MVPA

+ confounding factors ð2Þ

2.4.2 | Isotemporal substitution models

The isotemporal substitution model estimated the effects of replacing

time spent on one activity type with equal time spent on a different

activity type (e.g. displacing 30 min per day of sedentary activity with

30 min per day of LPA) on the sleep parameters. The control variables,

time spent on the activity type (e.g. LPA or MVPA) instead of the ini-

tial behavior (e.g. sedentary activity), and total wearing time (i.e. the

sum of time spent in sedentary activity, LPA, and MVPA), were simul-

taneously entered into the model. The details of this statistical model

are described by Yasunaga et al. (2017). The isotemporal substitution

model (in the case of omitting sedentary activity) was expressed as

equation 3 as follows:

Sleep parameters variable = LPA+MVPA+ total wear time

+ confounding factors ð3Þ

3 | RESULTS

3.1 | Descriptive statistics

Approximately 77.1% of the sample was female (n = 54) and the mean

age was 70.7 (±4.4 years) (Table 1). Regarding the daily behavioral

66 SEOL ET AL.

variables, total sedentary, LPA, and MVPA were 468.8 min (48.5% of

total time), 433.7 min (44.9% of total time), and 63.3 min (6.6% of

total time), respectively.

3.2 | Single-factor and partitioned model results

The results of the single-factor model estimating the influences of

each behavior on the sleep variables are shown in Table 2. LPA posi-

tively related to sleep efficiency (B = .770, 95% CI: .21, 1.33), and

wake after sleep onset (B = −2.675, 95% CI: −5.06, −.29), number of

awakenings (B = −.654, 95% CI: −1.21, −.10), sleep fragmentation

index (B = −1.554, 95% CI: −2.54, −.57), and PSQI global score

(B = −.342, 95% CI: −.65, −.03). Sedentary activity had negative

effects on sleep efficiency (B = −.886, 95% CI: −1.42, −.36), wake

after sleep onset (B = 3.218, 95% CI: .96, 5.48), number of awaken-

ings (B = .863, 95% CI: .35, 1.38), and sleep fragmentation index

(B = 1.530, 95% CI: .58, 2.48). Additionally, MVPA had negative

associations with number of awakenings (B = 1.232, 95% CI:

.02, 2.45).

The results of the partitioned model estimating the net relation-

ships of each activity type with the sleep variables are shown in

Table 3. LPA positively related to sleep efficiency (B = .945, 95%

CI: .37, 1.52), wake after sleep onset (B = −3.023, 95% CI: −5.75,

−.30), number of awakenings (B = −.637, 95% CI: −1.25, −.02),

sleep fragmentation index (B = −1.601, 95% CI: −2.64, −.56), and

PSQI global score (B = −.384, 95% CI: −.71, −.05). MVPA was nega-

tively associated with number of awakenings (B = 1.375, 95% CI:

.09, 2.66). Sedentary activity was not related to any of the sleep

parameters.

3.3 | Isotemporal substitution model

Replacing 30 min per day of sedentary activity with LPA positively

influenced sleep efficiency (B = .856, 95% CI: .30, 1.41), wake after

sleep onset (B = −3.020, 95% CI: −5.40, −.65), number of awakenings

(B = −.768, 95% CI: −1.30, −.23), sleep fragmentation index

(B = −1.624, 95% CI: −2.62, −.63), and the PSQI global score

(B = −.338, 95% CI: −.66, −.02). Replacing 30 min per day of seden-

tary activity with MVPA influenced the number of awakenings, which

was negative (B = 1.506, 95% CI: .34, 2.67). Replacing LPA with

MVPA did not influence any of the sleep parameters Table 4.

4 | DISCUSSION

This is one of the first studies to investigate the inter-relationships

among activity types (sedentary, LPA, and MVPA) on subjective and

objective sleep quality among Japanese older people. The results

partially supported the hypothesis. Engaging in LPA, particularly

when it replaces sedentary activity, would improve sleep efficiency,

wake after sleep onset, number of awakenings, sleep fragmentation

index, and subjective sleep quality. This was in contrast to the detri-

mental effect of substituting sedentary activity with MVPA on num-

ber of awakenings. In a previous systematic review, PA sufficiently

improved sleep quality (Buman & King, 2010); LPA was particularly

effective according to self-rated surveys (Stevenson & Topp, 1990).

Although previous studies have examined the influences of vari-

ous intensities of PA on older adults' sleep, some longitudinal studies

have demonstrated that light-to-moderate PA is better than MVPA

for slowing or halting sleep-quality degradation (Tsunoda et al., 2015).

However, a cross-sectional observational study found that regular

daily walking was associated with a lower prevalence of difficulty with

staying asleep, and regarding older women, vigorous PA with partici-

pation in an exercise program related to a higher prevalence of diffi-

culty staying asleep (Sherrill, Kotchou, & Quan, 1998). That is, PA of

TABLE 1 Participant characteristics (n = 70)

Variables Total mean (SD) Minimum MaximumDemographic

Age, years 70.4 (4.4) 65.0 84.0

Women, N (%) 54 (77.1)

BMI, kg/m2 23.0 (3.1) 16.8 30.2

Alcohol consumption

(drinker), N (%)

27 (38.6)

Smoking history, N (%) 12 (17.1)

High blood pressure

history, N (%)

20 (28.6)

GDS score, points 3.0 (2.5) .0 10.0

Both accelerometers

valid days, days

5.4 (.8) 4.0 6.0

Physical activity and

sedentary activity

Wear time, min/days 965.8 (128.7) 695.4 1307.0

Sedentary, min/days 468.8 (96.3) 284.0 729.7

LPA, min/days 433.7 (82.3) 289.3 626.5

MVPA, min/days 63.3 (35.8) 9.2 162.3

Objective sleep

parameters

Sleep onset latency, min 7.8 (2.3) 5.2 17.2

Sleep efficiency, % 86.6 (6.1) 66.7 95.0

Time in bed, min 431.6 (49.5) 317.5 546.7

Total sleep time, min 373.5 (47.0) 225.3 487.4

Wake after sleep onset,

min

50.3 (27.0) 9.7 149.6

Number of awakenings,

count

15.7 (6.7) 5.2 32.7

Sleep fragmentation

index, index

29.8 (11.6) 10.5 59.8

Subjective sleep

parameter

PSQI global score, point 5.9 (3.1) 1.0 14.0

BMI = body mass index; GDS = Geriatric Depression Scale; LPA = low-

intensity physical activity; MVPA = moderate-vigorous-intensity physical

activity; PSQI = Pittsburgh Sleep Quality Index; SD = standard deviation.

SEOL ET AL. 67

T A B L E 2

Si ng

le -v ar ia bl e m o de

lf o r as so ci at io ns

o f ea

ch ac ti vi ty

w it h sl ee

p pa

ra m et er s (n

= 7 0 )

U ns ta nd

ar di ze d re gr es si o n co

ef fi ci en

ts ,B

(9 5 %

C I)

V ar ia bl es

Sl ee

p o ns et

la te nc

y (m

in )

Sl ee

p ef fi ci en

cy (%

) W

ak e af te r sl ee

p o ns et

(m in )

N o .a w ak

en in gs

(c o un

t) Sl ee

p fr ag

m en

ta ti o n

in d ex

(in d ex

) P SQ

Ig lo b al

sc o re

(p o in t)

Se de

nt ar y (3 0 m in /d ay )

.2 0 1 (− .0 5 ,. 4 9 )

− .8 8 6 (− 1 .4 2 ,−

.3 6 )*

3 .2 1 8 (.9

6 ,5

.4 8 )*

.8 6 3 (.3

5 ,1

.3 8 )*

1 .5 3 (.5

8 ,2

.4 8 )*

.2 9 0 (− .0 2 ,. 5 9 )

LP A (3 0 m in /d ay )

− .1 9 5 (− .4 5 ,. 0 6 )

.7 7 0 (.2

4 ,1

.3 3 )*

− 2 .6 7 5 (− 5 .0 6 ,−

.2 9 )*

− .6 5 4 (− 1 .2 1 ,−

.1 0 )*

− 1 .5 5 4 (− 2 .5 4 ,−

.5 7 )*

− .3 4 2 (− .6 5 ,−

.0 3 )*

M V P A (3 0 m in /d ay )

.0 8 0 (− .4 8 ,. 6 4 )

− .7 7 2 (− 2 .0 4 ,. 5 0 )

3 .4 8 7 (− 1 .8 4 ,8

.8 1 )

1 .2 3 2 (.0

2 ,2

.4 5 )*

.2 8 4 (− 2 .0 1 ,2

.5 8 )

− .1 6 9 (− .8 7 ,. 5 3 )

*P < .0 5 .; R eg

re ss io n co

ef fi ci en

ts (9 5 %

C I) w er e ad

ju st ed

fo r ag e,

se x, bo

dy m as s in de

x, al co

ho lc o ns um

pt io n,

sm o ki ng

st at us ,h

yp er te ns io n m ed

ic al h is to ry ,G

er ia tr ic D ep

re ss io n Sc

al e sc o re ,a n d ac ce le ro m et er

w ea

r ti m e.

W he

n w ak e af te r sl ee

p o ns et

an d nu

m be

r o f aw

ak en

in gs

w er e in cl ud

ed in

th e an

al ys is ,w

e al so

in cl ud

ed ti m e in

be d.

Si ng

le -v ar ia bl e m o d el

as se ss ed

ea ch

ac ti vi ty

co m p o n en

t se p ar at el y.

C I=

co nf id en

ce in te rv al ;L

P A = lo w -i nt en

si ty

ph ys ic al ac ti vi ty ;M

V P A = m o de

ra te -v ig o ro us -i nt en

si ty

ph ys ic al ac ti vi ty ;P

SQ I=

P it ts bu

rg h Sl ee

p Q u al it y In d ex

.

T A B L E 3

P ar ti ti o n m o de

lf o r as so ci at io ns

o f ea

ch ac ti vi ty

w it h sl ee

p pa

ra m et er s (n

= 7 0 )

U ns ta nd

ar di ze d re gr es si o n co

ef fi ci en

ts B (9 5 %

C I)

V ar ia bl es

Sl ee

p o ns et

la te nc

y (m

in )

Sl ee

p ef fi ci en

cy (%

) W

ak e af te r sl ee

p o ns et

(m in )

N o .a w ak

en in gs

(c o un

t) Sl ee

p fr ag

m en

ta ti o n

in d ex

(in d ex

) P SQ

Ig lo b al

sc o re

(p o in t)

Se de

nt ar y (3 0 m in /d ay )

− .0 2 2 (− .2 4 ,. 2 0 )

.0 8 8 (− .3 8 ,. 5 6 )

.0 0 4 (− 2 .2 5 ,2

.2 5 )

− .1 3 2 (− .3 8 ,. 6 4 )

.0 2 3 (− .8 2 ,. 8 7 )

− .0 4 6 (− .3 1 ,. 2 2 )

LP A (3 0 m in /d ay )

− .2 2 9 (− .5 0 ,. 0 4 )

.9 4 5 (.3

7 ,1

.5 2 )*

− 3 .0 2 3 (− 5 .7 5 ,−

.3 0 )*

− .6 3 7 (− 1 .2 5 ,−

.0 2 )*

− 1 .6 0 1 (− 2 .6 4 ,−

.5 6 )*

− .3 8 4 (− .7 1 ,−

.0 5 )*

M V P A (3 0 m in /d ay )

.1 7 9 (− .4 4 ,. 8 0 )

− 1 .1 7 9 (− 2 .5 1 ,. 1 6 )

4 .5 7 1 (− 1 .1 4 ,1

0 .2 8 )

1 .3 7 5 (.0

9 ,2

.6 6 )*

.8 6 5 (− 1 .5 3 ,3

.2 6 )

.0 0 2 (− .7 6 ,. 7 6 )

*P < .0 5 .; R eg

re ss io n co

ef fi ci en

ts (9 5 %

C I) w er e ad

ju st ed

fo r ag e,

se x, bo

dy m as s in de

x, al co

ho lc o ns um

pt io n,

sm o ki ng

st at us ,h

yp er te ns io n m ed

ic al h is to ry ,a n d G er ia tr ic D ep

re ss io n Sc

al e sc o re .W

h en

w ak e

af te r sl ee

p o ns et

an d nu

m be

r o f aw

ak en

in gs

w er e in cl ud

ed in

th e an

al ys is .W

e al so

in cl ud

ed ti m e in

be d.

P ar ti ti o n m o de

le xa m in ed

al lt he

be ha

vi o rs

si m u lt an

eo u sl y, w it h o u t ad

ju st in g fo r ac ce le ro m et er

w ea

r ti m e.

C I=

co nf id en

ce in te rv al ;L

P A = lo w -i nt en

si ty

ph ys ic al ac ti vi ty ;M

V P A = m o de

ra te -v ig o ro us -i nt en

si ty

ph ys ic al ac ti vi ty ;P

SQ I=

P it ts bu

rg h Sl ee

p Q u al it y In d ex

.

68 SEOL ET AL.

relatively low intensity was positively related to improved subjective

sleep quality in older adults. Our findings partially support these

results and suggest that engaging in LPA, particularly when it replaces

sedentary activity, might improve objective and subjective sleep qual-

ity in older adults.

Although we could not identify the underlying causes of the

results because of the cross-sectional study design, there are several

possible reasons for our findings, both in terms of the beneficial influ-

ence of LPA on sleep and the detrimental association of MVPA to

sleep. For example, sedentary activity tends to include relatively more

screen time (i.e. watching television or using a cellular phone) than PA,

which might relate to poor sleep quality when engaged in close to

bedtime (Madden et al., 2014). However, LPA accounted for a consid-

erable quantity of the participants' daily PA (44.9%). It could be that

increasing daily PA would influence energy conservation or body res-

toration, such that it influences elders' nocturnal sleep patterns

(Driver & Taylor, 2000).

Sleep quality is strongly linked to mental health (Hisler & Brenner,

2019). Our findings were similar to those of a previous study, where

substituting 30 min per day of sedentary activity with LPA positively

influenced psychosocial well-being (Buman et al., 2010). Further, anti-

anxiety and antidepressant effects are a reported cause of PA-induced

sleep changes (Sciolino & Holmes, 2012). Thus, these results support

our findings.

Some studies have reported that an exercise intervention of

60–85% HRmax (>40 min over 2 days, every week) positively

influenced older people's subjective and objective sleep quality (King

et al., 2008; King, Oman, Brassington, Bliwise, & Haskell, 1997). How-

ever, we found that replacing 30 min per day of sedentary activity

with MVPA did not influence number of awakenings. This inconsis-

tency might be explained by the differences in participants' character-

istics (King et al., 1997; 2008) and the definitions of PA intensity

(Buman, Phillips, Youngstedt, Kline, & Hirshkowitz, 2014) between

earlier studies and the current study. For example, participants in pre-

vious studies were younger than the participants in our study. If the

effects of exercise on sleep depend in part on age, and LPA is effec-

tive for older people whose physical fitness has significantly declined

(Tsunoda et al., 2015), the results might vary because of participant

characteristics. Another age-related reason might be that the higher

body temperatures and hyperarousal caused by engaging in vigorous-

intensity exercise at night increase sleep fragmentation for older peo-

ple more than younger people (Edinger et al., 1993). Unfortunately,

we did not track the time of day participants engaged in PA, and it is

possible that nighttime MVPA caused an unfavorable association with

number of awakenings in our study.

This study makes an important contribution to the field, as most

other studies examined the effects of PA intensity on sleep only in an

experimental environment (Akbari Kamrani, Shams, Shamsipour

Dehkordi, & Mohajeri, 2014; King et al., 1997; King et al., 2008; Ste-

venson & Topp, 1990) or using self-report data (Morgan, 2003;

Sherrill et al., 1998; Tsunoda et al., 2015). The subjective evaluation

of PA and/or sleep used in most previous observational studies might

be a crucial limitation, as they likely include some extent of recall biasT A B L E 4

Is o te m po

ra ls ub

st it ut io n m o de

lf o r as so ci at io ns

o f re al lo ca te d ti m e in

o ne

ac ti vi ty

to an

o th er

ac ti vi ty

w it h sl ee

p pa

ra m et er s (n

= 7 0 )

U ns ta nd

ar di ze d re gr es si o n co

ef fi ci en

ts B (9 5 %

C I)

V ar ib le s

Sl ee

p o ns et

la te nc

y (m

in )

Sl ee

p ef fi ci en

cy (%

) W

ak e af te r sl ee

p o ns et

(m in )

N um

be r o f aw

ak en

in gs

(c o un

t) Sl ee

p fr ag

m en

ta ti o n

in d ex

(in d ex

) P SQ

Ig lo b al

sc o re

(p o in t)

Se de

nt ar y to

LP A

(3 0 m in /d ay )

− 0 .2 0 7 (− 0 .4 7 ,0

.0 5 )

0 .8 5 6 (0 .3 0 ,1

.4 1 )*

− 3 .0 2 0 (− 5 .4 0 ,−

0 .6 5 )*

− 0 .7 6 8 (− 1 .3 0 ,−

0 .2 3 )*

− 1 .6 2 4 (− 2 .6 2 ,−

0 .6 3 )*

− 0 .3 3 8 (− 0 .6 6 ,−

0 .0 2 )*

Se de

nt ar y to

M V P A

(3 0 m in /d ay )

0 .1 5 7 (− 0 .4 1 ,0

.7 2

− 1 .0 9 0 (− 2 .3 0 ,0

.1 2 )

4 .5 6 7 (− 0 .6 0 ,9

.7 3 )

1 .5 0 6 (0 .3 4 ,2

.6 7 )*

0 .8 8 8 (− 1 .2 8 ,3

.0 5 )

− 0 .0 4 4 (− 0 .7 3 ,0

.6 4 )

LP A to

M V P A (3 0 m in /d ay )

− 0 .0 5 0 (− 0 .6 3 ,0

.5 3 )

− 0 .2 3 4 (− 1 .4 7 ,1

.0 0 )

1 .5 4 7 (− 3 .7 7 ,6

.8 7 )

0 .7 3 8 (− 0 .4 6 ,1

.9 4 )

− 0 .7 3 6 (− 2 .9 6 ,1

.4 9 )

− 0 .3 8 2 (− 1 .0 9 ,0

.3 2 )

*P < 0 .0 .

R eg

re ss io n co

ef fi ci en

ts (9 5 %

C I) w er e ad

ju st ed

fo r ag e,

se x, B M I, al co

ho lc o ns um

pt io n,

sm o ki ng

st at us ,h

yp er te ns io n m ed

ic al hi st o ry ,G

D S sc o re ,a n d ac ce le ro m et er

w ea

r ti m e.

W h en

w ak e af te r sl ee

p o n se t an

d

nu m be

r o f aw

ak en

in gs

w er e in cl ud

ed in

th e an

al ys is ,w

e al so

in cl u de

d ti m e in

be d.

In is o te m po

ra ls ub

st it ut io n m o de

l, th e co

ef fi ci en

t fo r o ne

ty pe

o f ac ti vi ty

re p re se n ts

th e ef fe ct

o f in cr ea

si n g th is ty p e o f

ac ti vi ty

w hi le

ho ld in g th e o th er

ac ti vi ti es

co ns ta nt .

LP A ,L o w -i nt en

si ty

ph ys ic al ac ti vi ty ;M

V P A ,M

o de

ra te -v ig o ro us -i nt en

si ty

ph ys ic al ac ti vi ty ;P

SQ I, P it ts bu

rg h Sl ee

p Q ua

lit y In de

x.

SEOL ET AL. 69

(Paul et al., 2018), which would limit the generalizability of their find-

ings. It is significant that we objectively measured activities to over-

come the limitations of previous studies, using a high-quality hip

accelerometer to measure PA levels and sleep in the participants' nat-

ural environments.

Another strength is this study's novel use of the isotemporal sub-

stitution methodology to examine the effects of substituting behav-

iors on sleep in older adults. Because individual discretionary time is

limited to 24 h a day, engaging in one behavior means time spent on

another behavior must be reduced. Previous studies focused on the

influences of each activity type, but our study obtained meaningful

knowledge on the effects of each activity as the substitution of

another.

These results could help develop effective public health strategies

to treat older people's sleep problems. We found that replacing

30 min per day of sedentary activity with 30 min per day of LPA

increased sleep efficiency by .9%, and decreased wake after sleep

onset by 3.020 min per night, number of awakenings by .8 incidents

per night, sleep fragmentation index by 1.624 index per night, and

PSQI global score by .3 points. Health professionals and older people

might consider increasing LPA in place of sedentary activity to

improve overall sleep quality, and the positive effects might increase

if the replacement becomes a long-term habit.

4.1 | Limitations

Our study has some limitations. First, the cross-sectional design lim-

ited our conclusions because we could not determine causality. How-

ever, some previous studies found bidirectional relationships between

sleep quality parameters and PA (Master et al., 2019). Second, the iso-

temporal substitution model yielded statistical estimates that did not

reflect the actual replacement of one activity with another. Thus,

future studies should consider prospective cohort or interventional

designs. Third, data collection was not randomized, and along with the

small sample size, the results are unlikely to apply to all Japanese older

people. For example, our participants reported moderate sleep com-

plaints, with a PSQI global score of 5.9 ± 3.1 points (Doi et al., 2000).

Therefore, the results should be interpreted with caution. Future stud-

ies should aim for a larger, representative sample of Japanese older

adults recruited from the general population.

4.2 Conclusion

In this study, we explored the net relationships of sedentary activity,

LPA, and MVPA on subjective and objective sleep parameters in older

Japanese adults. Our main finding was that replacing 30 min per day

of sedentary activity with LPA might contribute to improving and

maintaining sleep quality. Further prospective and interventional stud-

ies are needed to clarify these effects on sleep in older adults.

AUTHOR CONTRIBUTIONS

Study design: J.S. and T.O.

Data collection: J.S., T.A., Y.F., and K.J.

Data analysis: J.S., Y.F., and K.J.

Manuscript writing and revisions for important intellectual content:

J.S. and T.A.

ACKNOWLEDGMENTS

We would like to thank Masaru Aizawa and Taiichiro Yosida for their

help with data collection. The authors would like to thank Naruki

Kitano for helping interpretation of data for the work.

ORCID

Jaehoon Seol https://orcid.org/0000-0003-3441-4191

REFERENCES

Akbari Kamrani, A. A., Shams, A., Shamsipour Dehkordi, P., & Mohajeri, R.

(2014). The effect of low and moderate intensity aerobic exercises on

sleep quality in men older adults. Pakistan Journal of Medical Sciences,

30(2), 417–421. Ancoli-Israel, S. (2005). Sleep and aging: Prevalence of disturbed sleep and

treatment considerations in older adults. Journal of Clinical Psychiatry,

66(Suppl. 9), 24–30. Bloom, H. G., Ahmed, I., Alessi, C. A., Ancoli-Israel, S., Buysse, D. J.,

Kryger, M. H., … Zee, P. C. (2009). Evidence-based recommendations

for the assessment and management of sleep disorders in older per-

sons. Journal of the American Geriatrics Society, 57(5), 761–789. Brzecka, A., Leszek, J., Ashraf, G. M., Ejma, M., �Avila-Rodriguez, M. F.,

Yarla, N. S., … Aliev, G. (2018). Sleep disorders associated with

Alzheimer's disease: A perspective. Frontiers in Neuroscience, 12, 1–10. Buman, M. P., Hekler, E. B., Haskell, W. L., Pruitt, L., Conway, T. L.,

Cain, K. L., … King, A. C. (2010). Objective light-intensity physical

activity associations with rated health in older adults. American Journal

of Epidemiology, 172(10), 1155–1165. Buman, M. P., & King, A. C. (2010). Exercise as a treatment to enhance

sleep. American Journal of Lifestyle Medicine, 4(6), 500–514. Buman, M. P., Phillips, B. A., Youngstedt, S. D., Kline, C. E., &

Hirshkowitz, M. (2014). Does nighttime exercise really disturb sleep?

Results from the 2013 National Sleep Foundation Sleep in America

poll. Sleep Medicine, 15(7), 755–761. Buysse, D. J., Reynolds, C. F., Monk, T. H., Berman, S. R., & Kupfer, D. J.

(1989). The Pittsburgh sleep quality index: A new instrument for psy-

chiatric practice and research. Psychiatry Research, 28(2), 193–213. Cole, R. J., Kripke, D. F., Mullaney, D. J., & Gillin, J. C. (1992). Automatic

sleep/wake identification from wrist activity. Sleep, 15(5), 461–469. de Rezende, L. F. M. (2014). Sedentary behavior and health outcomes

among older adults: A systematic review. BMC Public Health, 14, 1–9. Doi, Y., Minowa, M., Uchiyama, M., Okawa, M., Kim, K., Shibui, K., &

Kamei, Y. (2000). Psychometric assessment of subjective sleep quality

using the Japanese version of the Pittsburgh sleep quality index

(PSQI-J) in psychiatric disordered and control subjects. Psychiatry

Research, 97(2–3), 165–172. Driver, H. S., & Taylor, S. R. (2000). Exercise and sleep. Sleep Medicine

Reviews, 4(4), 387–402. Edinger, J. D., Morey, M. C., Sullivan, R. J., Higginbotham, M. B.,

Marsh, G. R., Dailey, D. S., & McCall, W. V. (1993). Aerobic fitness,

acute exercise and sleep in older men. Sleep, 16(4), 351–359. Harvey, J. A., Chastin, S. F. M., & Skelton, D. A. (2013). Prevalence of sed-

entary behavior in older adults: A systematic review. International Jour-

nal of Environmental Research and Public Health, 10(12), 6645–6661.

70 SEOL ET AL.

Hishikawa, N., Fukui, Y., Sato, K., Ohta, Y., Yamashita, T., & Abe, K. (2017).

Cognitive and affective functions associated with insomnia: A

population-based study. Neurological Research, 39(4), 331–336. Hisler, G. C., & Brenner, R. E. (2019). Does sleep partially mediate the

effect of everyday discrimination on future mental and physical

health? Social Science & Medicine, 221, 115–123. King, A. C., Oman, R. F., Brassington, G. S., Bliwise, D. L., & Haskell, W. L.

(1997). Moderate-intensity exercise and self-rated quality of sleep in

older adults. A randomized controlled trial. JAMA, 277(1), 32–37. King, A. C., Pruitt, L. A., Woo, S., Castro, C. M., Ahn, D. K., Vitiello, M. V., …

Bliwise, D. L. (2008). Effects of moderate-intensity exercise on polys-

omnographic and subjective sleep quality in older adults with mild to

moderate sleep complaints. The Journals of Gerontology. Series A, Bio-

logical Sciences and Medical Sciences, 63(9), 997–1004. Kurita, S., Yano, S., Ishii, K., Shibata, A., Sasai, H., Nakata, Y., … Oka, K.

(2017). Comparability of activity monitors used in Asian and Western-

country studies for assessing free-living sedentary behaviour. PLoS

One, 12(10), e0186523.

Madden, K. M., Ashe, M. C., Lockhart, C., & Chase, J. M. (2014). Sedentary

behavior and sleep efficiency in active community-dwelling older

adults. Sleep Science, 7(2), 82–88. Mander, B. A., Winer, J. R., & Walker, M. P. (2017). Sleep and human aging.

Neuron, 94(1), 19–36. Mâsse, L. C., Fuemmeler, B. F., Anderson, C. B., Matthews, C. E.,

Trost, S. G., Catellier, D. J., & Treuth, M. (2005). Accelerometer data

reduction: A comparison of four reduction algorithms on select out-

come variables. Medicine and Science in Sports and Exercise, 37(Suppl.

11, S544–S554. Master, L., Nye, R. T., Lee, S., Nahmod, N. G., Mariani, S., Hale, L., &

Buxton, O. M. (2019). Bidirectional, daily temporal associations between

sleep and physical activity in adolescents. Scientific Reports, 9(1), 1–14. Morgan, K. (2003). Daytime activity and risk factors for late-life insomnia.

Journal of Sleep Research, 12(3), 231–238. Niino, N., Imaizumi, T., & Kawakami, N. (1991). Japanese translation of the

geriatric depression scale. Clinical Gerontologist, 10, 85–86. Ohkawara, K., Oshima, Y., Hikihara, Y., Ishikawa-Takata, K., Tabata, I., &

Tanaka, S. (2011). Real-time estimation of daily physical activity inten-

sity by a triaxial accelerometer and a gravity-removal classification

algorithm. British Journal of Nutrition, 105(11), 1681–1691. Okajima, I., Komada, Y., Nomura, T., Nakashima, K., & Inoue, Y. (2012).

Insomnia as a risk for depression: A longitudinal epidemiologic study on

a Japanese rural cohort. Journal of Clinical Psychiatry, 73(3), 377–383. Oshima, Y., Kawaguchi, K., Tanaka, S., Ohkawara, K., Hikihara, Y.,

Ishikawa-Takata, K., & Tabata, I. (2010). Classifying household and

locomotive activities using a triaxial accelerometer. Gait and Posture,

31(3), 370–374. Owen, N., Healy, G. N., Matthews, C. E., & Dunstan, D. W. (2010). Too

much sitting: The population health science of sedentary behavior.

Exercise and Sport Sciences Reviews, 38(3), 105–113.

Paul, D. R., McGrath, R., Vella, C. A., Kramer, M., Baer, D. J., &

Moshfegh, A. J. (2018). Understanding the nature of measurement error

when estimating energy expenditure and physical activity via physical

activity recall. Journal of Physical Activity & Health, 15(7), 543–549. Riemann, D., Baglioni, C., Bassetti, C., Bjorvatn, B., Dolenc Groselj, L.,

Ellis, J. G., … Spiegelhalder, K. (2017). European guideline for the diag-

nosis and treatment of insomnia. Journal of Sleep Research, 26(6),

675–700. Sciolino, N. R., & Holmes, P. V. (2012). Exercise offers anxiolytic potential:

A role for stress and brain noradrenergic-galaninergic mechanisms.

Neuroscience and Biobehavioral Reviews, 36(9), 1965–1984. Sherrill, D. L., Kotchou, K., & Quan, S. F. (1998). Association of physical

activity and human sleep disorders. Archives of Internal Medicine, 158

(17), 1894–1898. Stevenson, J. S., & Topp, R. (1990). Effects of moderate and low intensity

long-term exercise by older adults. Research in Nursing & Health, 13(4),

209–218. Stone, K. L., Ancoli-Israel, S., Blackwell, T., Ensrud, K. E., Cauley, J. A.,

Redline, S., … Cummings, S. R. (2008). Actigraphy-measured sleep

characteristics and risk of falls in older women. Archives of Internal

Medicine, 168(16), 1768–1775.

The 2019 American Geriatrics Society Beers Criteria® Update Expert

Panel. (2019). American Geriatrics Society 2019 updated AGS beers

criteria® for potentially inappropriate medication use in older adults.

Journal of the American Geriatrics Society, 67, 674–694.

Tsunoda, K., Kitano, N., Kai, Y., Uchida, K., Kuchiki, T., Okura, T., &

Nagamatsu, T. (2015). Prospective study of physical activity and sleep

in middle-aged and older adults. American Journal of Preventive Medi-

cine, 48(6), 662–673. Yaffe, K., Falvey, C. M., & Hoang, T. (2014). Connections between

sleep and cognition in older adults. Lancet Neurology, 13(10),

1017–1028.

Yasunaga, A., Shibata, A., Ishii, K., Koohsari, M. J., Inoue, S., Sugiyama, T., … Oka, K. (2017). Associations of sedentary behavior and physical activ-

ity with older adults' physical function: An isotemporal substitution

approach. BMC Geriatrics, 17(1), 1–7. Youngstedt, S. D., O'Connor, P. J., Crabbe, J. B., & Dishman, R. K. (2000).

The influence of acute exercise on sleep following high caffeine intake.

Physiology & Behavior, 68(4), 563–570.

How to cite this article: Seol J, Abe T, Fujii Y, Joho K,

Okura T. Effects of sedentary behavior and physical activity

on sleep quality in older people: A cross-sectional study. Nurs

Health Sci. 2020;22:64–71. https://doi.org/10.1111/nhs.

12647

SEOL ET AL. 71