Content uploaded by Phyllis C Zee
Author content
All content in this area was uploaded by Phyllis C Zee on Feb 28, 2021
Content may be subject to copyright.
603 Journal of Clinical Sleep Medicine, Vol. 10, No. 6, 2014
Study Objective: This research examined the impact of
daylight exposure on the health of of ce workers from the
perspective of subjective well-being and sleep quality as well
as actigraphy measures of light exposure, activity, and sleep-
wake patterns.
Methods: Participants (N = 49) included 27 workers working
in windowless environments and 22 comparable workers
in workplaces with signi cantly more daylight. Windowless
environment is de ned as one without any windows or one
where workstations were far away from windows and without
any exposure to daylight. Well-being of the of ce workers
was measured by Short Form-36 (SF-36), while sleep quality
was measured by Pittsburgh Sleep Quality Index (PSQI).
In addition, a subset of participants (N = 21; 10 workers in
windowless environments and 11 workers in workplaces with
windows) had actigraphy recordings to measure light exposure,
activity, and sleep-wake patterns.
Results: Workers in windowless environments reported poorer
scores than their counterparts on two SF-36 dimensions—role
limitation due to physical problems and vitality—as well as
poorer overall sleep quality from the global PSQI score and
the sleep disturbances component of the PSQI. Compared
to the group without windows, workers with windows at the
workplace had more light exposure during the workweek, a
trend toward more physical activity, and longer sleep duration
as measured by actigraphy.
Conclusions: We suggest that architectural design of of ce
environments should place more emphasis on suf cient
daylight exposure of the workers in order to promote of ce
workers’ health and well-being.
Keywords: light exposure, sleep quality, quality of life,
architectural design, of ce environment
Citation: Boubekri M, Cheung IN, Reid KJ, Wang CH, Zee
PC. Impact of windows and daylight exposure on overall
health and sleep quality of of ce workers: a case-control pilot
study. J Clin Sleep Med 2014;10(6):603-611.
http://dx.doi.org/10.5664/jcsm.3780
SCIENTIFIC INVESTIGATIONS
Since the sick building syndrome of the 1970s and the World
Health Organization’s Declaration on Occupational Health
for All in 1994,1 occupational health has become a salient
issue among health professionals and architects alike. With
the increased interest today in green architecture, daylighting
is becoming an important design consideration. Typically,
daylighting recommendations are made in the form of daylight
factor levels ranging between 2% to 6% depending on building
types and activities. A daylight factor is a percentage of indoor
illuminance compared to the outdoor illuminance on a hori-
zontal surface. The daylight factor principle is valid for stable
overcast sky conditions only; sunny conditions are too dynamic
and changing to be considered.
Although there are many studies that have explored the
relationship between daylighting, psychological well-being,
and workers’ productivity or school children’s performance,2-4
few have addressed the impact of daylight at the workplace
on sleep, quality of life, and overall health. Exposure to light-
dark patterns is one of the main environmental cues for circa-
dian rhythms that in uence approximately 24-hour biological,
mental, and behavioral patterns such as sleep and activity.5 The
timing of light exposure is very in uential on these rhythms,
and previous research has shown that of ce environment
Impact of Windows and Daylight Exposure on
Overall Health and Sleep Quality of Of ce Workers:
A Case-Control Pilot Study
Mohamed Boubekri, Ph.D.1; Ivy N. Cheung, B.A.2; Kathryn J. Reid, Ph.D.2; Chia-Hui Wang1,3; Phyllis C. Zee, M.D., Ph.D., F.A.A.S.M.2
1School of Architecture, University of Illinois at Urbana-Champaign, Champaign, IL; 2Department of Neurology, Northwestern
University, Chicago, IL; 3Department of Architecture, Hwa-Hsia Institute of Technology, Taipei, Taiwan
BRIEF SUMMARY
Current Knowledge/Study Rationale: Both the amount and timing
of light exposure is important for physical and mental health. While re-
search indicates possible links between light exposure in workplaces
and workers’ productivity and performance, less is known about the role
of workplace light exposure on workers’ quality of life and sleep quality.
Study Impact: Of ce workers with more light exposure at the workplace
tended to have longer sleep duration, better sleep quality, more physical
activity, and better quality of life compared to of ce workers with less
light exposure at the workplace. Of ce workers’ physical and mental
well-being may be improved via enhanced indoor lighting for those with
insuf cient daylight in current of ces as well as increased emphasis on
light exposure in the design of future of ces.
lighting during work hours can act as a regulator of circadian
physiology and behavior, with blue-enriched arti cial lighting
even competing with natural light as an entrainer.6 Given that
of ce hours occur during biologically natural daylight hours,
we posit that light exposure in the of ce environment will have
effects on sleep, and via sleep and other in uences also have
effects on physical and mental health.
There is much evidence that links insuf cient sleep and/
or reduced sleep quality to a range of signi cant short-term
impairments such as memory loss, slower psychomotor
Downloaded from jcsm.aasm.org by 67.175.33.213 on February 28, 2021. For personal use only. No other uses without permission.
Copyright 2021 American Academy of Sleep Medicine. All rights reserved.
604
Journal of Clinical Sleep Medicine, Vol. 10, No. 6, 2014
M Boubekri, IN Cheung, KJ Reid et al.
reexes, and diminished attention.7-9 If windowless environ-
ments or lack of daylight affect ofce workers’ sleep quality,
there will be subsequent effects not only individually but also
on a societal level, leading to more accidents, workplace errors,
and decreased productivity. Sleep quality is also an important
health indicator that may have effects on, and interactions with
mood, cognitive performance, and health outcomes such as
diabetes and other illnesses.10-13 Therefore, it is crucial to inves-
tigate the effects of daylight as it may provide a profound way
to improve ofce workers’ productivity and health as well as
the safety of the community they work and live in. Deprivation
to light damages monoamine neurons and produces a depres-
sive behavioral phenotype in rats.14 In humans, a direct correla-
tion between the severity level of seasonal affective disorder
and exposure to natural light is well documented.15-17 Results
of several studies suggest that both natural and articial bright
light, particularly in the morning, can improve signicantly
health outcomes such as depression, agitation, sleep, circadian
rest-activity, and seasonal affective disorder.18-26
These effects of light exposure, or the lack thereof, illus-
trate the importance of proper light exposure for physical
well-being and mental health. In our modern society, many
responsibilities at the workplace and at home dictate self-
imposed alterations and/or loss of daylight in our daily lives.
Findings from the previously discussed research suggest that
the light exposure determined by our daily schedules will have
subsequent consequences on our mood, cognitive performance,
and overall well-being. However, studies exploring the impact
of daylight exposure, or the lack thereof, on the health of ofce
workers are very scarce. Therefore, the aim of this study was
to examine the inuence of light exposure at the workplace,
through the existence or absence of windows and of daylight,
on ofce workers’ sleep patterns, physical activity, and quality
of life via actigraphy and subjective measures. In our study we
compared two groups of ofce workers—those with windows
and abundant levels of daylight and those without windows
and with no direct contact with daylight at their workstations—
in terms of overall health and well-being and subjective sleep
quality using well-validated scales, and objective measures of
sleep, activity levels, and light exposure via actigraphy. We
hypothesized that ofce workers with windows in the work-
place would have more light exposure, better sleep quality,
more physical activity, and higher quality of life ratings
compared to ofce workers without windows in the workplace.
METHODS
Participants
A total of 49 participants were recruited, including 27 day-
shift workers in windowless workplaces and 22 comparable
day-shift workers in workplaces with windows. Workers were
selected from volunteers within administrative support staff
and other ofce workers on the campus of the University of Illi-
nois at Urbana-Champaign (UIUC) whose work schedule was
from 08:00 to 17:00. The typical recruitment process was done
by contacting an ofce manager, who in turn provided names
of volunteers from his/her group. The participants were not told
about the specic objectives of the study but were informed
that the study was about the impact of workplace physical and
social conditions on productivity and well-being.
In addition, a subset of the participants had actigraphy
recordings to measure light exposure, activity, and sleep. A
total of 21 participants had actigraphy recordings, including 10
ofce workers in windowless workplaces and 11 ofce workers
in workplaces with windows. Participants were selected for
actigraphy based on a convenience sample with volunteers
from ofce locations with and without windows.
Once the volunteers were identied, daylight factors at their
workstations were measured. Only daylight factors > 2% were
kept in the study for workers in workplaces with windows.
Generally, daylight factors < 2% are deemed not useful for task
performance illumination. In this study, we dene a window-
less workplace as one without any windows or one where
workstations were far away from windows and therefore had
no exposure to daylight and no views to the outside world.
The Institutional Review Board of the University of Illinois
at Urbana-Champaign (UIUC) approved the research study,
and all volunteers gave informed written consent as required
by UIUC regulations and standards. The cities of Urbana-
Champaign are relatively small, and the commute for most
participants is generally less than 15 minutes by car. Nearly all
participants drove individual cars to work.
Measures - Questionnaires
Ofce workers’ health related quality of life was measured
by Short Form 36 (SF-36), a questionnaire with 36 items related
to the physical and psychosocial domains of health inuenced
by a person’s experiences, beliefs, and perceptions of health.
The SF-36 survey is a well-validated health status question-
naire that measures an individual’s physical functioning, bodily
pain, and perception of the ability to perform physical, social,
and emotional role functions.27
The Pittsburgh Sleep Quality Index (PSQI) was utilized to
evaluate subjective sleep quality of the participants. This self-
rated questionnaire assesses sleep quality and disturbances
over a 1-month time interval.28 The PSQI is composed of 19
self-rated questions and 5 questions rated by a bed partner or
roommate. Only the self-rated items were used in scoring the
scale. The 19 questions generate 7 component scores: subjec-
tive sleep quality, sleep latency, sleep duration, habitual sleep
efciency, sleep disturbances, use of sleeping medication, and
daytime dysfunction. Each component score ranges from 0 (no
difculty) to 3 (severe difculty). The component scores are
summed to produce a global score with a range of 0–21. A higher
score indicates lower sleep quality. A PSQI global score > 5 is
considered suggestive of signicant sleep disturbance.
A daylight deprivation survey was administered that includes
questions pertaining to demographic characteristics (age,
gender, race, and working experience) and behavioral char-
acteristics (self-reported amount of exposure to daylight on a
scale of 1-10 [with 1 being always exposed and 10 being never
exposed], hours of outdoor activities per day, eating behavior
prior going to bed, and duration of current light exposure level).
Measures - Actigraphy
Participants wore an Actiwatch-L (Minimitter) on their
non-dominant wrist. An actiwatch device is an ambulatory
Downloaded from jcsm.aasm.org by 67.175.33.213 on February 28, 2021. For personal use only. No other uses without permission.
Copyright 2021 American Academy of Sleep Medicine. All rights reserved.
605 Journal of Clinical Sleep Medicine, Vol. 10, No. 6, 2014
Windows, Daylight Exposure and Overall Health of Ofce Workers
physiological data logger often used in research and clinical
settings to detect and record motion during wake and sleep. The
Actiwatch-L has an accelerometer sensitivity of 0.05 g-force
and is equipped with a photodiode for measuring amount and
duration of light illuminance. Participants were instructed to
continuously wear these actiwatches for 2 weeks without
removing them (except for bathing) during the period of time
they were answering the questionnaires. Participants were also
instructed to leave the actiwatches exposed to the environment
at all times and to avoid covering them with clothing. The ques-
tionnaires and actiwatches were administered during late spring
and summer seasons.
Valid data were recorded for a range of 6 to 10 workdays
and 2 to 4 free days in participants, with the average partici-
pant yielding 8.4 workdays and 3.4 free days of actigraphy data
meeting inclusion criteria for analysis, as determined by < 4
h off-wrist time per day. Analysis was conducted on Actiware
software version 5 (Philips Respironics) with 30-sec sampling
epochs and wake threshold value of 40 activity counts. Sleep
start was dened as the rst 10-min period in which no more
than one epoch was scored as mobile. Sleep end was dened
as the last 10-min period in which no more than one epoch was
scored as immobile. Wake threshold selection was set at medium.
Actigraphy measures were calculated as the average of each
participant’s valid workdays (split into wake time to 08:00 for
workday mornings, 08:00 to 17:00 for work hours, and 17:00
to sleep start for workday evenings) and valid free days for
activity and light exposure variables, and for nighttime hours
following workdays and free days for sleep variables. Actig-
raphy variables analyzed include total activity counts (sum of
all valid physical activity counts for all epochs in the active
period from wake time to 08:00 for workday mornings, 08:00
to 17:00 on workdays for work hours, 17:00 to sleep start for
workday evenings, and for wake periods during free days),
sleep onset time (clock time of sleep start on nights following
workdays and free days), sleep onset latency (time elapsed
between the start time of a given rest interval and the following
sleep start time on nights following workdays and free days),
sleep efciency (the percentage of scored total sleep time to
interval duration minus total invalid time for the given rest
period on nights following workdays and free days), wake after
sleep onset (total minutes between the start time and end time
of a given sleep interval scored as wake on nights following
workdays and free days), sleep time (total minutes between the
start time and end time of a given interval scored as sleep on
nights following workdays and free days), sleep fragmentation
(sum of percent mobile and percent immobile bouts < 1 min
duration to the number of immobile bouts for the given interval
on nights following workdays and free days), and average light
exposure (sum of all valid illuminance data in lux on a loga-
rithmic scale for all epochs from the start time to the end time
of a given interval multiplied by the epoch length in minutes
from wake time to 08:00 for workday mornings, 08:00 to 17:00
on workdays for work hours, 17:00 to sleep start for workday
evenings, and for wake periods during free days).
Statistical Methods
First, we performed a χ2 test (homogeneity for proportions)
to compare distributions of the demographics and behavioral
characteristics as measured by the daylight deprivation survey
(age, race, gender, working experience, self-reported amount of
exposure to daylight, hours of outdoor activities per day, eating
behavior prior to going to bed, and duration of current light
level exposure) between participants working in workplaces
without windows and participants working in workplaces
with windows. Secondly, we performed t-tests to determine
any statistical difference between the two groups in terms of
ofce workers’ health related quality of life and sleep quality as
measured on the SF-36 and PSQI.
For the subset of participants with actigraphy recording,
distributions of the demographics and behavioral characteris-
tics as measured by the daylight deprivation survey between
workers in workplaces with no windows and workers in work-
places with windows were compared to distributions in the
overall group. T-tests were then utilized to gauge differences
between the two groups in terms of the following previously
dened actigraphy measures: total activity counts, sleep onset
time, sleep onset latency, sleep efciency, wake after sleep
onset, sleep time, fragmentation index, and light exposure.
Pearson bivariate correlations were run between work hour
light exposure as measured by actigraphy and subjective ques-
tionnaires and other actigraphy variables.
RESULTS
Demographics and Behavioral Characteristics of the
Two Groups of Workers
Results of the χ2 test show no signicant differences between
these two groups in terms of distributions of age, race, gender,
working experience, hours of outdoor activities per day, eating
behavior prior to going to bed, and duration of current light
level exposure (Table 1). Therefore, these two groups were
comparable except in their amount of self-reported amount of
exposure to daylight (Table 2).
For the subset of participants with actigraphy recording,
distributions of the demographic and behavioral characteris-
tics as measured by the daylight deprivation survey between
workers in workplaces with no windows and workers in work-
places with windows are comparable to respective distributions
in the overall group, again with no signicant differences in
these distributions between groups except in their amount of
self-reported amount of exposure to daylight.
Light Exposure of the Two Groups of Workers
The self-reported amount of exposure to daylight scale shows
ofce workers in workplaces without windows perceived they
had signicantly less exposure to daylight than ofce workers in
workplaces with windows, as expected (Table 2). Results from
actigraphy conrm average light exposure differences during
work hours for the two groups, with workers in workplaces
with windows receiving more light exposure than workers in
workplaces without windows (Table 3 and Figure 1A; 3.00 log
lux versus 2.58 log lux; p = 0.02). There was no signicant
difference in light exposure from wake time to start of the work
period (Table 3; 2.57 log lux versus 2.38 log lux; p = 0.32);
however, workers with windows in the workplace had more
light exposure during workday evenings (Table 3; 2.50 log lux
Downloaded from jcsm.aasm.org by 67.175.33.213 on February 28, 2021. For personal use only. No other uses without permission.
Copyright 2021 American Academy of Sleep Medicine. All rights reserved.
606
Journal of Clinical Sleep Medicine, Vol. 10, No. 6, 2014
M Boubekri, IN Cheung, KJ Reid et al.
Table 1—Demographic and behavioral characteristics of the two groups
Variables
Work place without windows
(N = 27)
Work place with windows
(N = 22)
All
(N = 49) p value
Demographic Characteristics
Gender
Males 44.4% (12) 31.8% (7) 38.78% (19) 0.37
Females 55.6% (15) 68.2% (15) 61.22% (30)
Age (years)
19-30 11.1% (3) 18.2% (4) 14.29% (7)
0.65
31-45 40.7% (11) 27.3% (6) 34.69% (17)
46-59 44.4% (12) 45.5% (10) 44.90% (22)
60+ 3.7% (1) 9.1% (2) 6.12% (3)
Race
Black/African-American 0 9.1% (2) 4.08% (2)
0.25
American Indian/Alaskan Native 0 4.5% (1) 2.04% (1)
White/Non-Hispanic 92.6% (25) 86.4% (19) 89.80% (44)
Asian/Pacic Islander 0 0 0
Latino/Hispanic 3.7% (1) 0 2.04% (1)
Other 3.7% (1) 0 2.04% (1)
Working experience (years)
0-1 7.4% (2) 4.5% (1) 6.12% (3)
0.79
2-4 18.5% (5) 22.7% (5) 20.41% (10)
5-7 25.9% (7) 18.2% (4) 26.83% (11)
8-10 18.5% (5) 31.8% (7) 24.49% (12)
> 11 29.6% (8) 22.7% (5) 26.53% (13)
Behavioral Characteristics
Outdoor activities (hours per day)
0-1 81.5% (22) 68.2% (15) 75.51% (37)
0.282-4 18.5% (5) 31.8% (7) 24.49% (12)
4-6 0 0 0
Years at current light exposure level
0-1 7.4% (2) 9.1% (2) 8.16% (4)
2-4 25.9% (7) 31.8% (7) 28.57% (14)
5-7 25.9% (7) 27.3% (6) 26.53% (13) 0.98
8-10 18.5% (5) 13.6% (3) 16.33% (8)
> 11 22.2% (6) 18.2% (4) 20.41% (10)
Eating behavior prior going to bed
Eating directly prior going to bed 25.9% (7) 13.6% (3) 20.41% (10) 0.29
No eating prior going to bed 74.1% (20) 86.4% (19) 79.59% (39)
Table 2—Self-reported amount of exposure to daylight between the two groups
Levels of exposure to daylight
Work place without windows
(N = 27)
Work place with windows
(N = 22) All (N = 49) p value
1 Always Exposed 0 18.2% (4) 8.16% (4)
0.02*
23.7% (1) 27.3% (6) 14.29% (7)
3 3.7% (1) 4.5% (1) 4.08% (2)
40 9.1% (2) 4.08% (2)
5 Sometimes Exposed 3.7% (1) 4.5% (1) 4.08% (2)
6 7.4% (2) 9.1% (2) 8.16% (4)
714.8% (4) 9.1% (2) 12.24% (6)
8 33.3% (9) 13.6% (3) 24.49% (12)
918.5% (5) 4.5% (1) 12.24% (6)
10 Never Exposed 14.8% (4) 0 8.16% (4)
* p ≤ 0.05.
Downloaded from jcsm.aasm.org by 67.175.33.213 on February 28, 2021. For personal use only. No other uses without permission.
Copyright 2021 American Academy of Sleep Medicine. All rights reserved.
607 Journal of Clinical Sleep Medicine, Vol. 10, No. 6, 2014
Windows, Daylight Exposure and Overall Health of Ofce Workers
Table 3—Results of t-tests for actigraphy measures between the two groups
Mean ± SD
p valueWork place without windows (N = 10) Work place with windows (N = 11)
Workdays
Mornings
Total activity counts (arbitrary units) 36,274 ± 48,654 135,071 ± 163,184 0.07†
Average light exposure (log lux-min) 2.38 ± 0.51 2.57 ± 0.36 0.32
Work hours
Total activity counts (arbitrary units) 115,208 ± 172,793 476,290 ± 523,782 0.06†
Average light exposure (log lux-min) 2.58 ± 0.55 3.00 ± 0.16 0.02*
Evenings
Total activity counts (arbitrary units) 69,083 ± 96,477 295,188 ± 412,374 0.09†
Average light exposure (log lux-min) 1.93 ± 0.51 2.50 ± 0.36 0.008**
Sleep onset time (hour: minute) 22:04 ± 1:34 21:46 ± 0:48 0.58
Sleep onset latency (min) 19.16 ± 38.88 9.61 ± 7.15 0.43
Sleep efciency (%) 89.35 ± 4.22 91.24 ± 3.29 0.26
Wake after sleep onset (min) 37.25 ± 13.38 30.10 ± 14.87 0.26
Sleep time (min) 429.65 ± 39.84 476.31 ± 45.23 0.02*
Sleep fragmentation 22.23 ± 11.06 18.84 ± 5.81 0.38
Free days
Total activity counts (arbitrary units) 224,696 ± 262,373 839,780 ± 1,113,613 0.12
Average light exposure (log lux-min) 2.37 ± 0.55 3.03 ± 0.32 0.003**
Sleep onset time (hour: minute) 22:48 ± 1:48 22:06 ± 1:08 0.29
Sleep onset latency (min) 19.56 ± 50.04 15.03 ± 17.97 0.78
Sleep efciency (%) 90.13 ± 4.46 90.82 ± 6.02 0.77
Wake after sleep onset (min) 36.38 ± 17.53 31.13 ± 19.00 0.52
Sleep time (min) 413.67 ± 71.45 506.17 ± 62.86 0.005**
Sleep fragmentation 21.55 ± 9.11 20.27 ± 8.30 0.74
† p ≤ 0.10, * p ≤ 0.05, ** p ≤ 0.01. Workday mornings refer to wake time to 08:00 period on workdays; Workday work hours refers to 08:00 to 17:00 work period
on workdays; Workday evenings refers to 17:00 to sleep onset period for activity and light measures and refers to the sleep period following a workday for the
sleep measures; Free days refer to days spent away from the ofce environment without work hours.
Figure 1—Actigraphy measures of light exposure, total activity, and sleep time between workers in workplaces with windows
(N = 11) and without windows (N = 10).
Actigraphy data collected in a subset of the ofce workers show that those with windows in the workplace had higher light exposure (A), more total activity
(B), and longer sleep time (C) than workers without windows in the workplace. * p < 0.05, † p < 0.10.
versus 1.93 log lux; p = 0.008) and during free days (Table 3;
3.30 log lux versus 2.37 log lux; p = 0.003) than workers without
windows in the workplace. While we cannot say from our data
collection whether this difference is from natural daylight
or articial lighting in the ofce building, workers without
windows at the workplace had signicantly lower average light
Downloaded from jcsm.aasm.org by 67.175.33.213 on February 28, 2021. For personal use only. No other uses without permission.
Copyright 2021 American Academy of Sleep Medicine. All rights reserved.
608
Journal of Clinical Sleep Medicine, Vol. 10, No. 6, 2014
M Boubekri, IN Cheung, KJ Reid et al.
exposure than workers with windows during workday work
hours and evenings as well as during free days.
Physical and Mental Conditions of the Two Groups of
Workers
Workers in workplaces without windows had signicantly
worse scores on two of the SF-36 dimensions—role limita-
tion due to physical problems (RP) and vitality (VT)—than
workers in workplaces with windows (Figure 2; p = 0.001 and
p = 0.004, respectively). There was also a positive correlation
between light exposure during work hours and role limitation
due to physical problems (R = 0.503, p = 0.02). Overall, both
the physical component summary (PCS) (p = 0.09) and mental
component summary (MCS) (p = 0.11) scores of those in
workplaces without windows were lower than scores of those
working in workplaces with windows (Table 4). Participants
in workplaces without windows reported poorer scores on all
eight dimensions of the SF-36 than participants in workplaces
with windows.
In addition, actigraphy monitoring indicated that workers
with windows had more than four times as much activity on
average during work hours than workers without windows,
although this difference did not reach statistically signicance
(Table 3 and Figure 1B; 476,290 activity counts versus 115,280
activity counts; p = 0.06). There was also a trend for workers
with windows to have more physical activity during workday
mornings (Table 3; 135,071 activity counts versus 36,274
activity counts; p = 0.07) and workday evenings (Table 3;
295,188 activity counts versus 69,083 activity counts; p = 0.09)
than workers without windows; however, there was no signi-
cant statistical difference during free days (Table 3; 839,780
activity counts versus 224,696 activity counts; p = 0.12). There
was little correlation between activity and light exposure levels
during work hours (R = -0.075, p = 0.75), workday evenings
(R = -0.025, p = 0.91), and free days (R = -0.138, p = 0.55).
Sleep Quality of the Two Groups of Workers
Workers without windows had a tendency toward poorer
scores on overall sleep quality from the global PSQI score
than workers with windows (Table 5 and Figure 3; p = 0.05),
although we did note that the global PSQI score in both groups
was high, as a score > 5 is considered suggestive of poor sleep
quality. The signicant difference in global score may be
attributed mainly to sleep disturbance, which was found to
be different between the two groups (Table 5 and Figure 3;
p = 0.02), while differences in daytime dysfunction and sleep
efciency components contributed only moderately to poorer
global PSQI scores for workers without windows than workers
with windows (Table 5 and Figure 3; p = 0.08 and p = 0.07,
respectively). Other PSQI subscores did not differ signicantly
between the two groups.
Analysis of rest and activity patterns from actigraphy data
showed workers with windows at the workplace slept an average
of 46 minutes more per night during the workweek than workers
without windows at the workplace (Table 3 and Figure 1C;
476 min versus 430 min; p = 0.02). There was also a positive
correlation between light exposure during work hours and sleep
Figure 2—Short Form 36 (SF-36) measures of vitality and
role limitation due to physical problems between workers
in workplaces with windows (N = 22) and without windows
(N = 27).
Workers with windows in the workplace reported better scores on
vitality (A) and role limitation due to physical problems (B) on the SF-36
compared to workers with no windows in the workplace. * p < 0.05.
Table 4—Results of t-tests for Short Form-36 between the two groups
Mean ± SD
p value
Work place without
windows (N = 27)
Work place with windows
(N = 22)
Norms of USA general
population
PCS (physical component summary) 50.09 ± 7.83 53.57 ± 5.86 50.00 ± 10 0.09†
MCS (mental component summary) 44.47 ± 10.71 49.51 ± 10.86 50.00 ± 10 0.11
Physical Function (PF) 89.07 ± 13.45 91.36 ± 10.49 82.29 ± 23.76 0.52
Role limitation due to physical problems (RP) 67.59 ± 37.86 96.59 ± 8.78 82.51 ± 25.52 0.001***
Bodily Pain (BP) 74.81 ± 19.67 78.32 ± 19.79 71.33 ± 23.66 0.54
General Health (GH) 67.59 ± 20.40 75.91 ± 19.50 70.85 ± 20.98 0.15
Vitality (VT) 45.56 ± 21.27 61.82 ± 15.32 58.31 ± 20.02 0.004**
Social Function (SF) 79.63 ± 21.13 88.07 ± 18.29 84.30 ± 22.92 0.15
Role limitation due to emotional problems (RE) 69.14 ± 42.29 81.82 ± 36.70 87.40 ± 21.44 0.27
Mental Health (MH) 68.15 ± 15.59 75.64 ± 16.37 74.99 ± 17.76 0.10†
† p ≤ 0.10, * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001.
Downloaded from jcsm.aasm.org by 67.175.33.213 on February 28, 2021. For personal use only. No other uses without permission.
Copyright 2021 American Academy of Sleep Medicine. All rights reserved.
609 Journal of Clinical Sleep Medicine, Vol. 10, No. 6, 2014
Windows, Daylight Exposure and Overall Health of Ofce Workers
time on workday nights (R = 0.483, p = 0.03). While there were
no signicant differences between workers with windows and
workers without windows in sleep onset time (21:46 versus
22:04), sleep onset latency (10 min vs 19 min), sleep efciency
(91% vs 89%), wake after sleep onset (30 min vs 37 min), and
sleep fragmentation (19 vs 22) on workday nights, the aver-
ages point toward better measures of sleep quality for workers
with windows at the workplace than workers without windows
at the workplace during the workweek. Similarly, workers
with windows at the workplace slept more than their counter-
parts on free day nights (506 min vs 389 min; p = 0.005), and
although there were no differences in sleep onset time (22:06 vs
22:48), sleep onset latency (15 min vs 20 min), sleep efciency
(91% vs 90%), wake after sleep onset (31 min vs 36 min), and
sleep fragmentation (20 vs 22) on free day nights, the averages
point toward better measures of sleep quality for workers with
windows at the workplace than workers without windows at the
workplace during free day nights.
DISCUSSION
These results demonstrate a relationship between work-
place light exposure and ofce workers’ sleep quality, activity
patterns, and quality of life. Workers in workplaces with
windows not only had signicantly more light exposure during
work hours but also slept an average of 46 minutes more per
night during the workweek than workers in workplaces without
windows. Workers with windows in the workplace also had
more light exposure during the workday evenings and during
free days, as well as longer sleep time compared to workers
without windows in the workplace. However, there were no
differences in light exposure in the mornings before the work
period. Workers without windows also reported poorer scores
than their counterparts on the global PSQI score and the PSQI
component score for sleep disturbances. None of the other
component scores of the PSQI were signicantly different
between groups, nor were actigraphy sleep variables other than
sleep time different between the groups.
These ndings suggest that light exposure, or the lack thereof,
during work hours may have effects beyond the workplace that
impact sleep duration and quality, which may then have further
effects on other health factors. Research indicates that insuf-
cient sleep and reduced sleep quality have myriad health and
safety consequences. For example, insufcient sleep and reduced
sleep quality have been associated with higher evening levels of
cortisol, impaired glucose metabolism, increases in appetite via
decreased leptin and increased ghrelin levels, and higher body
mass index, as well as increased fatigue and deterioration of
performance, alertness, and mental concentration, which can
lead to increased error rates and subsequent risk of injury.7-9,29-32
These health and performance consequences may affect
perceived health related quality of life, as measured by the
Table 5—Results of t-tests for Pittsburgh Sleep Quality Index between the two groups
Mean ± SD
p valueWork places without windows (N = 27) Work places with windows (N = 22)
Component 1: Subjective sleep quality 1.11 ± 0.64 1.00 ± 0.76 0.58
Component 2: Sleep latency 1.00 ± 1.07 0.73 ± 0.88 0.34
Component 3: Sleep duration 1.48 ± 0.94 1.14 ± 0.89 0.29
Component 4: Sleep efciency 0.74 ± 1.16 0.27 ± 0.55 0.07†
Component 5: Sleep disturbance 1.31 ± 0.67 0.95 ± 0.38 0.02*
Component 6: Use of sleep medication 0.42 ± 1.00 0.14 ± 0.64 0.23
Component 7: Daytime dysfunction 1.12 ± 0.51 0.82 ± 0.66 0.08†
Global PSQI Score 7.23 ± 4.21 5.05 ± 3.17 0.05*
† p ≤ 0.10, * p ≤ 0.05.
Figure 3—Pittsburgh Sleep Quality Index (PSQI) measures
between workers in workplaces with windows (N = 22) and
without windows (N = 27).
Workers with windows in the workplace reported better overall global
score on the PSQI (A) compared to workers with no windows in the
workplace. The difference in global score is made up mainly of differences
in sleep disturbances (B), daytime dysfunction (C), and sleep efciency
(D), with workers without windows reporting poorer scores than workers
with windows on all three PSQI subscores. * p < 0.05, † p < 0.10.
Downloaded from jcsm.aasm.org by 67.175.33.213 on February 28, 2021. For personal use only. No other uses without permission.
Copyright 2021 American Academy of Sleep Medicine. All rights reserved.
610
Journal of Clinical Sleep Medicine, Vol. 10, No. 6, 2014
M Boubekri, IN Cheung, KJ Reid et al.
SF-36. Our results from the SF-36 show workplaces without
windows have signicantly negative impact on workers’ role
limitation due to physical problems (RP) and vitality (VT), as
well as a marginal negative impact on workers’ mental health
compared to workplaces with windows. These results are
similar to the ndings of a study that examined ve dimensions
(GH, V, SF, RE, and MH) of the SF-36 and found that the scores
of vitality (VT), social functioning (SF), and mental health
(MH) for those working in dark ofces are lower than scores
for those working in ofces with more lighting.33 Another study
focusing on predictors of burnout among nurses found that
exposure to at least three hours of daylight per day resulted in
less stress and higher satisfaction at work.34 While those with
more daylight in the workplace also have higher daily physical
activity during work hours and workday evenings, our anal-
ysis cannot determine whether the workers get more activity
because of the daylight or whether they have more daylight
exposure due to activity. There was no difference in physical
activity between the two groups during free days despite differ-
ences in light exposure during free days, and correlations
between physical activity levels and light exposure during work
hours, workday evenings, and free days did not suggest a strong
relationship. Nonetheless, it remains a possibility that differ-
ences in activity level may inuence light exposure and also
sleep, yet the tendency towards higher activity levels indicates
workers with more daylight exposure may have fewer physical
problems or complaints regarding vitality in parallel with our
ndings on subjective measures of the SF-36.
Prior to this study, little was known about how architectural
features such as windows impact light exposure and subsequent
effects on physical and mental factors. Via examination of the
inuence of ofce settings with and without windows on ofce
workers’ light exposure, sleep, physical activity, and quality of
life via actigraphy and subjective measures, this research study
shows ofce workers in workplaces with windows may have
more light exposure, better sleep quality, more physical activity,
and higher quality of life ratings than ofce workers in work-
places without windows.
This study has some limitations that could be addressed in
future work. For example, the small sample size and sampling
methodology could be addressed in a larger study. Participants
for this study were volunteers based on a convenience sample,
which may have introduced bias. The amount of light in an
ofce may be associated with position or level of experience
in the workplace; however, we found no differences in age,
race, gender, years at current job, and duration of working in
current light levels between workers in ofce settings with and
without windows. We also do not have data from the partici-
pants on caffeine use, measurements of stress levels, and chro-
notype, which is of interest given the outcome measures of this
study. Although we observed no differences in sleep onset time
between the two groups of workers on workday nights and free
day nights, the possibility remains that chronotype, circadian
timing, or other behavioral measures may be responsible for
some of the differences observed in the two groups of workers.
This warrants further investigation. The objective measures of
wrist actigraphy support the subjective ndings; however, actig-
raphy data were collected for only 21 of the 49 total partici-
pants. Furthermore, although actigraphy has reasonable validity
and reliability and is often used as a sleep assessment tool in
sleep medicine, this methodology has some limitations. Sleep
diaries were not collected in this study, and therefore were
unavailable for the actigraphy analysis. For sleep-wake periods,
actigraphy has low specicity for detecting wakefulness within
sleep periods. Actigraphy is also neither sensitive to low light
levels nor calibrated for articial uorescent lighting. As such,
light exposure measurements for workers in ofce settings
without windows may be an underestimate. In addition, since
light exposure data are collected from the wrist, there is the
possibility that error may be introduced by covering of the acti-
watch, and therefore, reported values may not be fully represen-
tative of the light levels reaching the retina. Our data collection
methods also do not allow for differentiation between natural
daylight and articial lighting, and do not allow for analysis of
specic wavelengths of light exposure. Future studies would
benet from using devices that collect spectral distribution for
comparison between the two workplace groups. Lastly, addi-
tional benets of workplaces with windows, such as the roles of
views and other dimensions, were not taken into account in this
study. Views may bring some psychological dimension while
daylight may have physiological effects. Future research may
be able to dissociate the different roles of views and daylighting
of windows. This can be done, for example, by exploring the
differences between skylights that provide very limited views
to the sky only versus side windows. Despite these limitations,
signicant differences are seen with light exposure levels and
subsequent measures of sleep quality and physical and mental
well-being.
As emphasized in the World Health Organization’s Declara-
tion on Occupational Health for All,1 the focal point for prac-
tical occupational health activities is the workplace. Therefore,
employers have a social responsibility to plan and design a safe
and healthy working environment for their employees. Some
countries (such as Canada, Germany, and France) recommend
certain amounts of daylight in schools and ofces. Yet even in
these countries it is not a requirement. In the United States, the
national building code lists windows primarily as a means of
emergency escape and rescue as opposed to natural lighting.
Given the results of this study, we conclude that emphasizing
daylight exposure and lighting in the workplace may posi-
tively affect the well-being of people working in those spaces.
Lower amounts of light exposure in the workplace was associ-
ated with reduced sleep duration, poorer sleep quality, lower
activity levels, and reduced quality of life in this sample of
ofce workers. Light exposure in the workplace may therefore
have long-lasting and compounding effects on the physical and
mental health of the workers not only during but also beyond
work hours. Enhanced indoor lighting for those with insuf-
cient lighting in current ofces as well as increased emphasis
on light exposure in the architectural design of future ofce
environments is recommended to improve ofce workers’ sleep
quality and physical well-being. Workers with limited or no
access to windows in the workplace may increase their light
exposure during work hours in various ways. Taking a walk
during a break and enjoying lunch outdoors are simple ways
to increase daytime natural light exposure. Further research is
needed to determine what light exposure durations or intensi-
ties are sufcient or optimal for benets to well-being.
Downloaded from jcsm.aasm.org by 67.175.33.213 on February 28, 2021. For personal use only. No other uses without permission.
Copyright 2021 American Academy of Sleep Medicine. All rights reserved.
611 Journal of Clinical Sleep Medicine, Vol. 10, No. 6, 2014
Windows, Daylight Exposure and Overall Health of Ofce Workers
REFERENCES
1. WHO. Declaration on occupational health for all. Geneva, Switzerland: World
Health Organization Ofce of Occupational Health, 1994.
2. Haynes BP. The impact of ofce comfort on productivity. J Fac Management
2008;6:37-51.
3. Group HM. Windows and ofces: A study of ofce worker performance and the
indoor environment. California Energy Commission: P500-03-082-A-9, 2003.
4. Thayer BM. Daylighting and Productivity at Lockheed. Solar Today 1995;9:26-29.
5. Roenneberg T, Kantermann T, Juda M, Vetter C, Allebrandt KV. Light and the
human circadian clock. Handb Exp Pharmacol 2013:311-31.
6. Vetter C, Juda M, Lang D, Wojtysiak A, Roenneberg T. Blue-enriched ofce
light competes with natural light as a zeitgeber. Scand J Work Environ Health
2011;37:437-45.
7. Åkerstedt T, Czeisler C, Dinges D, Horne J. Accidents and sleepiness: A
consensus statement. At: International Conference on Work Hours, Sleep and
Accidents. Stockholm, Sweden: J Sleep Res 1994;3.
8. Leger D. The cost of sleep-related accidents: a report for the National
Commission on Sleep Disorders Research. Sleep 1994;17:84-93.
9. Mitler MM, Carskadon MA, Czeisler CA, Dement WC, Dinges DF, Graeber RC.
Catastrophes, sleep, and public policy: consensus report. Sleep 1988;11:100-9.
10. Bower B, Bylsma LM, Morris BH, Rottenberg J. Poor reported sleep quality
predicts low positive affect in daily life among healthy and mood-disordered
persons. J Sleep Res 2010;19:323-32.
11. Leppamaki S, Partonen T, Vakkuri O, Lonnqvist J, Partinen M, Laudon M. Effect
of controlled-release melatonin on sleep quality, mood, and quality of life in
subjects with seasonal or weather-associated changes in mood and behaviour.
Eur Neuropsychopharmacol 2003;13:137-45.
12. Nebes RD, Buysse DJ, Halligan EM, Houck PR, Monk TH. Self-reported sleep
quality predicts poor cognitive performance in healthy older adults. J Gerontol B
Psychol Sci Soc Sci 2009;64:180-7.
13. Knutson KL, Ryden AM, Mander BA, Van Cauter E. Role of sleep duration
and quality in the risk and severity of type 2 diabetes mellitus. Arch Intern Med
2006;166:1768-74.
14. Gonzalez MM, Aston-Jones G. Light deprivation damages monoamine neurons
and produces a depressive behavioral phenotype in rats. Proc Natl Acad Sci
U S A 2008;105:4898-903.
15. Graw P, Recker S, Sand L, Krauchi K, Wirz-Justice A. Winter and summer
outdoor light exposure in women with and without seasonal affective disorder.
J Affect Disord 1999;56:163-9.
16. Kripke D, Mullaney D, Savides TJ, Gillin JC. Phototherapy for nonseasonal
major depressive disorders. In: Rosenthal N, Blehar N, eds. Phototherapy
for nonseasonal major depressive disorders. New York: The Guilford Press,
1989:342-56.
17. Saarijarvi S, Lauerma H, Helenius H, Saarilehto S. Seasonal affective disorders
among rural Finns and Lapps. Acta Psychiatr Scand 1999;99:95-101.
18. Beauchemin KM, Hays P. Sunny hospital rooms expedite recovery from severe
and refractory depressions. J Affect Disord 1996;40:49-51.
19. Benedetti F, Colombo C, Barbini B, Campori E, Smeraldi E. Morning sunlight
reduces length of hospitalization in bipolar depression. J Affect Disord
2001;62:221-3.
20. Lewy AJ, Bauer VK, Cutler NL, et al. Morning vs evening light treatment of
patients with winter depression. Arch Gen Psychiatry 1998;55:890-6.
21. Lovell BB, Ancoli-Israel S, Gevirtz R. Effect of bright light treatment on agitated
behavior in institutionalized elderly subjects. Psychiatry Res 1995;57:7-12.
22. Rosenthal NE, Sack DA, Gillin JC, et al. Seasonal affective disorder. A
description of the syndrome and preliminary ndings with light therapy. Arch
Gen Psychiatry 1984;41:72-80.
23. Rosenthal NE, Sack DA, Carpenter CJ, Parry BL, Mendelson WB, Wehr TA.
Antidepressant effects of light in seasonal affective disorder. Am J Psychiatry
1985;142:163-70.
24. Terman JS, Terman M, Lo ES, Cooper TB. Circadian time of morning light
administration and therapeutic response in winter depression. Arch Gen
Psychiatry 2001;58:69-75.
25. Van Someren EJ, Kessler A, Mirmiran M, Swaab DF. Indirect bright light
improves circadian rest-activity rhythm disturbances in demented patients. Biol
Psychiatry 1997;41:955-63.
26. Wallace-Guy GM, Kripke DF, Jean-Louis G, Langer RD, Elliott JA, Tuunainen
A. Evening light exposure: implications for sleep and depression. J Am Geriatr
Soc 2002;50:738-9.
27. Ware JE Jr., Sherbourne CD. The MOS 36-item short-form health survey (SF-
36). I. Conceptual framework and item selection. Med Care 1992;30:473-83.
28. Buysse DJ, Reynolds CF, 3rd, Monk TH, Berman SR, Kupfer DJ. The Pittsburgh
Sleep Quality Index: a new instrument for psychiatric practice and research.
Psychiatry Res 1989;28:193-213.
29. Leproult R, Copinschi G, Buxton O, Van Cauter E. Sleep loss results in an
elevation of cortisol levels the next evening. Sleep 1997;20:865-70.
30. Spiegel K, Leproult R, Van Cauter E. Impact of sleep debt on metabolic and
endocrine function. Lancet 1999;354:1435-9.
31. Spiegel K, Leproult R, L’Hermite-Baleriaux M, Copinschi G, Penev PD, Van
Cauter E. Leptin levels are dependent on sleep duration: relationships with
sympathovagal balance, carbohydrate regulation, cortisol, and thyrotropin.
J Clin Endocrinol Metab 2004;89:5762-71.
32. Taheri S, Lin L, Austin D, Young T, Mignot E. Short sleep duration is associated
with reduced leptin, elevated ghrelin, and increased body mass index. PLoS
Med 2004;1:e62.
33. Mills PR, Tomkins SC, Schlangen LJ. The effect of high correlated colour
temperature ofce lighting on employee wellbeing and work performance.
J Circadian Rhythms 2007;5:2.
34. Alimoglu MK, Donmez L. Daylight exposure and the other predictors of burnout
among nurses in a University Hospital. Int J Nurs Stud 2005;42:549-55.
ACKNOWLEDGMENTS
Drs. Boubekri and Cheung contributed equally to this work. The authors thank the
subjects for their participation.
SUBMISSION & CORRESPONDENCE INFORMATION
Submitted for publication July, 2013
Submitted in nal revised form December, 2013
Accepted for publication February, 2014
Address correspondence to: Mohamed Boubekri, Ph.D., School of Architecture,
University of Illinois at Urbana-Champaign, 611 Taft Drive, Champaign, IL 61820; Tel:
(217) 333-2848; Fax: (217) 244-2900; E-mail: boubekri@illinois.edu
DISCLOSURE STATEMENT
This was not an industry supported study. This research was supported by the Il-
linois Campus Research Board of the University of Illinois at Urbana-Champaign and
NIH grants 5T32 HL790915 and P01 AG11412. Dr. Zee has a Philips/Respironics
Educational/Research Gift to Northwestern University, owns stock in Teva, and is a
consultant for Sano-Aventis, UCB, Johnson and Johnson, Merck and Co, Takeda,
Purdue, Philips, Jazz, Vanda, and Ferring. Dr. Reid has received research support
from Philips Consumer Lifestyles for research unrelated to the work reported in the
paper. The other authors have indicated no nancial conicts of interest.
Downloaded from jcsm.aasm.org by 67.175.33.213 on February 28, 2021. For personal use only. No other uses without permission.
Copyright 2021 American Academy of Sleep Medicine. All rights reserved.