Content uploaded by Kenji Obayashi
Author content
All content in this area was uploaded by Kenji Obayashi on Sep 24, 2015
Content may be subject to copyright.
Chronobiology International, 2014; 31(9): 976–982
!Informa Healthcare USA, Inc.
ISSN: 0742-0528 print / 1525-6073 online
DOI: 10.3109/07420528.2014.937491
ORIGINAL ARTICLE
Association between light exposure at night and insomnia in the
general elderly population: The HEIJO-KYO cohort
Kenji Obayashi, Keigo Saeki, and Norio Kurumatani
Department of Community Health and Epidemiology, Nara Medical University School of Medicine, Nara, Japan
Chronic circadian misalignment between the internal and environmental rhythms, which is typically related to
night-shift work and clock-gene variants, is associated with disruption of suprachiasmatic nucleus function and
increased risk of insomnia. Under controlled laboratory conditions, light at night (LAN) suppresses melatonin
secretion, delays the internal biological rhythm, and reduces sleepiness. Therefore, LAN exposure may cause circadian
misalignment and insomnia, though it remains unclear in real-life situations whether LAN exposure is associated with
insomnia. To evaluate an association between LAN exposure and sleep quality in home settings, we conducted a
cross-sectional community-based study in 857 elderly individuals (mean age, 72.2 years). We evaluated bedroom light
intensity using a light meter and subjectively and objectively measured sleep quality using the Pittsburgh Sleep
Quality Index and an actigraph, respectively, along with urinary 6-sulfatoxymelatonin excretion. Compared with the
lowest quartile group of LAN intensity, the highest quartile group revealed a significantly higher odds ratio (OR) for
subjective insomnia in a multivariate model adjusted for age, gender, body mass index, daytime physical activity,
urinary 6-sulfatoxymelatonin excretion, bedtime, rising time, and day length (adjusted OR, 1.61, 95% confidence
interval, 1.05–2.45, p¼0.029). In addition, higher OR for subjective insomnia was significantly associated with the
increase in quartiles of LAN intensity (p
trend
¼0.043). Consistently, we observed significant association trends between
the increase in quartiles of LAN intensity and poorer actigraphic sleep quality, including decreased sleep efficiency,
prolonged sleep-onset latency, increased wake-after-sleep onset, shortened total sleep time, and delayed sleep-mid
time in multivariate models adjusted for the covariates mentioned above (all p
trend
50.001). In conclusion, we
demonstrated that LAN exposure in home settings is significantly associated with both subjectively and objectively
measured sleep quality in a community-based elderly population.
Keywords: Actigraphy, circadian rhythms, insomnia, light at night, melatonin, sleep quality
INTRODUCTION
Epidemiological studies have demonstrated a higher
prevalence of insomnia among the elderly than the
younger population as well as a steady increase in the
prevalence of insomnia in recent decades (Calem et al.,
2012; Ford & Kamerow, 1989; Prinz et al., 1990). Previous
studies involving a large elderly population showed that
approximately 40% of elderly individuals reported any
type of insomnia such as difficulty initiating sleep,
difficulty maintaining sleep, early morning awakening,
and non-restorative sleep (Calem et al., 2012; Walsh
et al., 2011). Insomnia among the elderly is one of the
most important public health issues because of its high
prevalence and its association with increased risk of
psychiatric and neurodegenerative disorders such as
depression and dementia, cardiovascular diseases, and
mortality (Cricco et al., 2001; Dew et al., 2003; Eaker
et al., 1992; Yokoyama et al., 2010).
The solar 24 h cycle has led to the evolution of the
human circadian rhythms. The suprachiasmatic nucleus
(SCN) of the hypothalamus, which is an essential
component of the master biological clock, synchronizes
the internal biological rhythm to the environmental
rhythm. Physiologically, light exposure is the most
important environmental entraining cue for SCN func-
tion. Chronic circadian misalignment between the
internal and environmental rhythms, which is typically
related to night-shift work and clock-gene variants, is
associated with disruption of SCN function and
increased risk of insomnia (Allebrandt et al., 2010;
Boudreau et al., 2013; Wyatt et al., 1999). Melatonin, a
pineal gland hormone, is hypothesized to be a major
Correspondence: Kenji Obayashi, Department of Community Health and Epidemiology, Nara Medical University School of
Medicine, 840 Shijocho, Kashiharashi, Nara, 634-8521, Japan. Tel: +81-744-22-3051. Fax: +81-744-25-7657. E-mail: obayashi@
naramed-u.ac.jp
Submitted January 19, 2014, Returned for revision May 31, 2014, Accepted June 18, 2014
976
Downloaded by [Kenji Obayashi] at 14:28 11 August 2015
internal contributor to the association between circa-
dian misalignment and insomnia because endogenous
melatonin levels are closely associated with both light
exposure and sleepiness (Dijk & Cajochen, 1997;
Zeitzer et al., 2000).
Exposure to light at night (LAN) is increasing globally,
not only among night-shift workers but also among the
general population because of the increased use of
artificial lighting in modern society (Navara & Nelson,
2007). The nocturnal input of non-visual light informa-
tion into the SCN through the photosensitive retinal
ganglion cells (pRGCs) is associated with a range of
neurobiological effects, including increased core body
temperature, suppression of melatonin secretion, and
stimulation of brain activity. Under controlled labora-
tory conditions, LAN suppresses melatonin secretion,
delays the internal biological rhythm, and reduces
sleepiness (Cajochen et al., 2000, 2005; Czeisler et al.,
1986; Lockley et al., 2006). Thus, LAN exposure may
cause circadian misalignment and insomnia, though it
remains unclear in real-life situations whether LAN
exposure is associated with insomnia.
In this cross-sectional study on 857 elderly individ-
uals, we examined the association between LAN in
home settings and sleep quality. We evaluated subject-
ively and objectively measured sleep quality using the
Pittsburgh Sleep Quality Index (PSQI) and an actigraph,
respectively, which are the two principal methods used
to measure sleep quality in field studies. We also
measured bedroom light intensity at night using light
meters as well as overnight urinary 6-sulfatoxymelato-
nin excretion (UME), the major melatonin metabolite,
as an index of melatonin secretion.
METHODS
Participants and study protocol
Between September 2010 and March 2013, 880 commu-
nity-based elderly subjects voluntarily enrolled in a
study titled ‘‘Housing Environments and Health
Investigation among Japanese Older People in Nara,
Kansai Region: a prospective community-based cohort
(HEIJO-KYO) study.’’ Of these, 857 home-dwelling
participants met the inclusion criteria of age 60 years
and completion of the PSQI questionnaire. All partici-
pants provided written informed consent, and the study
protocol was performed in accordance with the ethics
committee of Nara Medical University and the ethical
standards of the Journal (Portaluppi et al., 2010).
The protocols were described in our previous study
(Obayashi et al., 2012). In brief, after collecting demo-
graphic and medical information using a standardized
questionnaire, we initiated measurements of LAN
exposure and actigraphic parameters in two consecutive
days. Subsequently, we instructed the participants to
collect their urine the following night and to maintain a
standardized sleep diary by logging their bedtime and
rising time.
Measurements of subjective and objective
sleep quality
Subjective sleep quality was measured using the PSQI
questionnaire, in which sleep quality over the previous
month was asked using seven subscales measuring
different components of sleep, including sleep quality,
sleep latency, sleep duration, sleep efficiency, sleep
disturbance, use of sleep medication, and daytime
dysfunction. The score of each component ranges
from 0 to 3, with 3 indicating the worst sleep quality
(Buysse et al., 1989). A cut-off value of 6 was used for the
detection of sleep disturbance. Insomnia was defined as
individuals who were previously diagnosed of insomnia
and currently received a sleep medication and/or
individuals who had PSQI score of 6.
Objective sleep quality was measured at 1 min inter-
vals using an actigraph (Actiwatch 2; Respironics Inc.,
Murrysville, PA) that was worn on the non-dominant
wrist. Sleep/awake status at each epoch, sleep onset,
and sleep offset were automatically detected by the
Actiware version 5.5 (Respironics Inc.) with the default
algorithm (Philips Respironics Actiware Tutorials, 2013).
Epochs with higher activity counts than moderate
threshold (40 counts/min) were treated as awake.
Sleep onset was the first minute followed by 10-min
immobility period comprising not more than one epoch
with any motion count, and sleep offset was the last
minute following the 10 min period of immobility.
Five actigraphic sleep parameters were determined
using objective data (awake/sleep status and sleep
onset/offset) and self-reported data (bedtime and
rising time) as follows: (1) total sleep time (TST), the
time spent with sleep (below the activity threshold of
40 counts/min) between sleep onset and sleep offset;
(2) sleep efficiency (SE), the percentage calculated from
TST divided by the time between bedtime and rising
time; (3) wake after sleep onset (WASO), the time spent
with awake (above the activity threshold of 40 counts/
min) between sleep onset and rising time; (4) sleep
onset latency (SOL), the time between bedtime and
sleep onset; and (5) sleep-mid time, the mid-time
between sleep onset and sleep offset.
Measurement of LAN exposure
LAN exposure was measured at 1-min intervals using a
light meter (LX-28SD; Sato Shouji Inc., Kanagawa,
Japan) with the sensor fixed at 60 cm above the floor,
near the head of the bed, and facing the ceiling. The
optical sensor had spectral sensitivity that approximated
that of the human eye, the illuminance sensitivity that
ranged from 0 to 100 000 lux. The internal clocks of the
light meter and the actigraph were synchronized. We
used the average light intensity between bedtime and
rising time for a parameter of LAN exposure. We divided
the participants into quartile groups according to the
intensity of LAN exposure. Among the 841 participants,
the day-to-day reproducibility of average LAN intensity
Real-life link between LAN and insomnia 977
!Informa Healthcare USA, Inc.
Downloaded by [Kenji Obayashi] at 14:28 11 August 2015
in two consecutive days was moderately high
(the Spearman rank correlation coefficient ¼0.66).
UME measurement
The overnight urine collection protocol involved dis-
carding the last void at bedtime and collecting each
subsequent void, including the first morning void.
Samples were stored in a dark bottle with a cold pack
at room temperature, the total volume was measured,
and the samples were stored at 20 C until the assay.
Urinary 6-sulfatoxymelatonin concentrations were
measured using a highly sensitive enzyme-linked
immunosorbent assay kit (RE54031; IBL International,
Hamburg, Germany). UME was calculated as
follows: UME (mg) ¼6-sulfatoxymelatonin concentration
(mg/mL) total overnight urine volume (mL). UME was
used as an index of melatonin secretion because there is
evidence that UME correlates closely with the secreted
levels of this hormone (Baskett et al., 1998).
Other measurements
Body mass index (BMI) was calculated as weight (kg)/
height (m
2
). Current smoking status, habitual alcohol
consumption, and sleep medication use were evaluated
by administering a questionnaire to participants.
The estimated glomerular filtration rate (eGFR) was
calculated using the formula from the Japanese Society
of Nephrology-Chronic Kidney Disease Practice
Guide: eGFR (mL/min/1.73 m
2
)¼194 [serum creatin-
ine (mg/dl)]
1.094
[age (years)]
0.287
. The result
was multiplied by a correction factor of 0.739 for
female. Diabetes mellitus was diagnosed on the basis
of the following assessments: medical history, current
antidiabetic therapy, fasting plasma glucose levels
7.0 mmol/L, and glycated hemoglobin levels 6.5% of
the National Glycohemoglobin Standardization Program
value. Daytime physical activity was the average of all
valid physical activity counts evaluated using the
actigraph (Actiwatch 2) between rising time and bed-
time. Data regarding the day length from sunrise to
sunset in Nara (latitude 34N) on the measurement days
were obtained from the webpage of the National
Astronomical Observatory of Japan (National
Astronomical Observatory of Japan, 2013).
Statistical analyses
Variables with a normal distribution were expressed as
the mean and standard deviation (SD), whereas asym-
metrically distributed variables were reported as the
median and interquartile range (IQR). Means and
medians were compared between the subjective insom-
nia and non-insomnia groups using the unpaired t-test
and the Mann–Whitney Utest, respectively. The chi-
square test was used for comparing categorical data.
Comparisons of adjusted means were conducted using
analysis of covariance (ANCOVA). Variables including
age, gender, BMI, current smoking status, alcohol
consumption habit, diabetes, eGFR, daytime physical
activity, UME, bedtime, rising time, day length, and
actigraphic sleep parameters were compared between
the subjective insomnia and non-insomnia groups. The
average values of LAN exposure, daytime physical
activity, UME, bedtime, rising time, day length, and
actigraphic sleep parameters on two consecutive days
were used for further analyses. UME and SOL with a
skewed distribution were naturally log-transformed for
analyses. Odds ratios (ORs) for subjective insomnia and
means of actigraphic sleep parameters were simultan-
eously adjusted for variables that were marginally to
significantly associated with subjective insomnia
(p50.20) in the univariate comparisons (Table 1).
Trends in the association of quartiles of LAN intensity
with ORs for subjective insomnia and means of
actigraphic sleep parameters were evaluated using
linear regression models for trends. Statistical analyses
were performed using SPSS version 19.0 for Windows
(IBM SPSS Inc., Chicago, IL). A two-sided pvalue50.05
was considered significant.
RESULTS
The mean age of the 857 participants was 72.2 ± 7.1
years, and 432 (50.4%) individuals were female. The
median intensity of LAN was 0.8 lux (IQR, 0.1–3.4 lux).
There were 310 individuals with subjective insomnia. Of
these, 84 participants were diagnosed of insomnia and
currently received a sleep medication, and 294 partici-
pants had PSQI score of 6. The subjective insomnia
group (n¼310) showed significantly higher age, more
female, less daytime physical activity, earlier bedtime,
and later rising time than the non-insomnia group
(n¼547, Table 1). The subjective insomnia group
showed marginally but insignificantly lower BMI
(p¼0.10) and UME (p¼0.11) and shorter day length
on measurement day (p¼0.09) than the non-insomnia
group.
Comparisons of actigraphic sleep parameters after
adjustment for age and gender (Table 2) revealed
significantly lower SE and longer SOL, WASO, and TST
in the subjective insomnia group than the non-insomnia
group (SE, 83.6 versus 85.0%, p¼0.019; log-transformed
SOL, 3.1 versus 2.9 log min, p¼0.035; WASO, 85.6 versus
74.8 min, p¼0.001; TST, 429.2 versus 417.9 min,
p¼0.020).
Compared with the lowest quartile group of LAN
intensity (median, 0 lux), the unadjusted OR for sub-
jective insomnia in the highest quartile group (median,
9.7 lux) was 1.82 [95% confidence interval (CI), 1.22–
2.72, p¼0.004; Table 3]. Consistently, in the multivariate
logistic model, the adjusted OR for subjective insomnia
in the highest quartile group was significantly higher
than that in the lowest quartile group even after
adjusting for age, gender, BMI, daytime physical activity,
log-transformed UME, bedtime, rising time, and day
length (adjusted OR, 1.61; 95% CI, 1.05–2.45; p¼0.029).
In addition, the higher adjusted ORs for subjective
978 K. Obayashi et al.
Chronobiology International
Downloaded by [Kenji Obayashi] at 14:28 11 August 2015
insomnia were significantly associated with the
increase in quartiles of LAN intensity (p
trend
¼0.043).
Consistently, after adjustment for age and gender, the
higher adjusted ORs for subjective insomnia defined
only by PSQI score were significantly associated with the
increase in quartiles of LAN intensity (p
trend
¼0.016).
In contrast, UME did not significantly differ across
quartiles of LAN intensity (p
trend
¼0.60).
Regarding the actigraphic sleep parameters, the
increase in quartiles of LAN intensity were significantly
associated with poorer sleep quality, including
decreased SE (both p
trend
50.001), prolonged SOL
(both p
trend
50.001) and WASO (both p
trend
50.001),
and delayed sleep-mid time (unadjusted p
trend
¼0.003,
adjusted p
trend
50.001) in both unadjusted and adjusted
models (Table 4). Although prolonged TST was margin-
ally associated with the increase in quartiles of LAN
intensity in an unadjusted model (p
trend
¼0.07), this
parameter was significantly shortened with quartiles of
LAN intensity increasing in an adjusted model
(p
trend
50.001).
Furthermore, we have conducted additional analyses
with regard to the association between LAN and
actigraphic sleep parameters in the actual sleep period
defined by actigraphy. Consistently, after adjustment for
age, gender, bedtime, and rising time, an increase in
quartiles of LAN intensity was significantly associated
with poorer sleep quality, including decreased SE (Q1,
88.8%; Q2, 88.6%; Q3, 88.4%; Q4, 87.4%; p
trend
¼0.041),
TABLE 2. Comparisons of actigraphic sleep parameters between subjective insomnia and non-insomnia.
Adjusted mean* (95% CI)
Insomnia Non-insomnia
Characteristics (n¼310) (n¼547) p
PSQI score 8.3 (8.1–8.5) 3.0 (2.8–3.1) 50.001
Actigraphic sleep parameters
SE, % 83.6 (82.7–84.5) 85.0 (84.3–85.6) 0.019
Log-transformed SOL, log min 3.1 (3.0–3.2) 2.9 (2.8–3.0) 0.035
WASO, min 85.6 (80.7–90.5) 74.8 (71.1–78.5) 0.001
TST, min 429.2 (421.6–436.8) 417.9 (412.2–423.5) 0.020
Sleep-mid time, clock time 02:37 (02:31–02:43) 02:38 (02:33–02:43) 0.85
PSQI, Pittsburgh Sleep Quality Index; SE, sleep efficiency; SOL, sleep-onset latency; WASO, wake after
sleep-onset; TST, total sleep time; CI, confidence interval. *Adjusted for age and gender using analysis of
covariance.
TABLE 1. Basic and clinical characteristics between subjective
insomnia and non-insomnia.
Insomnia
Non-ins
omnia
Characteristics (n¼310) (n¼547) p
Basic parameters
Age, mean (SD), years 73.6 (6.7) 71.4 (7.2) 50.001
Gender, female,
number (%)
180 (58.1) 252 (46.1) 0.001
Body mass index,
mean (SD), kg/m
2
22.9 (3.0) 23.2 (3.0) 0.10
Current smoker,
number (%)
15 (4.8) 29 (5.3) 0.77
Alcohol consumption
(430 g/day), number (%)
39 (12.6) 78 (14.3) 0.49
Clinical parameters
Diabetes, number (%) 33 (10.6) 65 (11.9) 0.58
eGFR, mean (SD),
mL/min/1.73m
2
71.1 (16.0) 72.5 (14.1) 0.23
Daytime physical
activity, mean (SD),
count/min
286.7 (107.8) 305.4 (105.1) 0.014
UME, median
(IQR), (mg)
6.2 (3.8–9.1) 6.9 (4.1–10.3) 0.11
Bedtime, mean (SD),
clock time
22:16 (1:14) 22:33 (1:06) 0.001
Rising time, mean (SD),
clock time
6:55 (0:52) 6:42 (0:58) 0.003
Day length,
median (IQR), min
642 (606–681) 650 (610–690) 0.09
eGFR, estimated glomerular filtration rate; UME, urinary
6-sulfatoxymelatonin excretion; IQR, interquartile range.
TABLE 3. Logistic regression analysis for the associations of exposure to LAN with subjective insomnia.
Quartiles of intensity of LAN (lux) [median, range]
0[50.1] 0.4 [0.1–0.8] 1.7 [0.8–3.4] 9.7 [43.5] p
trend
No. of participants (missing) 205 (5) 208 (2) 207 (4) 208 (2)
No. of cases 63 74 76 92
Unadjusted OR (95% CI) 1.00 (ref) 1.27 (0.84–1.91) 1.31 (0.87–1.98) 1.82 (1.22–2.72) 0.005
Age/gender-adjusted OR (95% CI) 1.00 (ref) 1.26 (0.83–1.92) 1.23 (0.81–1.87) 1.86 (1.23–2.82) 0.005
Fully-adjusted OR* (95% CI) 1.00 (ref) 1.25 (0.81–1.91) 1.19 (0.77–1.83) 1.61 (1.05–2.45) 0.043
LAN, light at night; OR, odds ratio; CI, confidence interval; UME, urinary 6-sulfatoxymelatonin excretion.
*Adjusted for age, gender, body mass index, daytime physical activity, log-transformed UME, bedtime, rising time, and day
length (per quartile increment).
Real-life link between LAN and insomnia 979
!Informa Healthcare USA, Inc.
Downloaded by [Kenji Obayashi] at 14:28 11 August 2015
prolonged WASO (Q1, 46.8 min; Q2, 49.4 min; Q3,
51.5 min; Q4, 56.4 min; p
trend
50.001), and delayed
sleep-mid time (Q1, 02:34; Q2, 02:35; Q3, 02:38; Q4,
02:43; p
trend
50.001). In contrast, there was no signifi-
cant association trend between LAN intensity and TST
(p
trend
¼0.79).
DISCUSSION
The present study demonstrated that LAN exposure in
home settings was significantly associated with both
subjective and objective sleep quality in a community-
based elderly population. The findings were evidenced
by the significant association trends of an increase in
LAN intensity with higher prevalence of subjective
insomnia and poorer actigraphic sleep measures,
independent of several potential confounding fac-
tors (p
trend
¼0.043 in subjective measurements and
p
trend
50.001 in all actigraphic measurements).
Furthermore, the OR for subjective insomnia was 61%
higher in the highest quartile group of LAN intensity
than that in the lowest quartile group (adjusted OR, 1.61;
95% CI, 1.05–2.45; p¼0.029).
The present study, to the best of our knowledge,
provided the first human evidence regarding the asso-
ciation between very low average intensity of LAN
exposure in home settings and risk of insomnia.
A previous well-controlled experimental study indicated
that LAN exposure can exert a dose-dependent alerting
effect, as assessed by subjective ratings, slow eye
movements, and electroencephalographic activity
(Cajochen et al., 2000); whereas, the effective LAN
intensity was higher than that of the highest LAN
quartile group in the present study (median, 9.7 lux).
However, LAN intensity in the present study can include
LAN with high intensity but short duration, because the
LAN intensity was the average bedroom light intensity
during the in-bed period. A recent experimental study
indicated that light exposure with high intensity
but short duration may effect on human circadian
physiology using the duration-response curve for the
association between light exposure and melatonin sup-
pression (Chang et al., 2012). In addition, human
circadian physiology is more closely correlated to
shorter wave length rather than intensity (Cajochen
et al., 2005). Thus, additional experimental researches
are needed to better understand the effect of LAN with
high intensity but short duration, or low intensity but
short wave length as well as very low intensity on sleep
quality in humans.
The clinical implications of sleep disturbances mea-
sured in the present study could be interpreted by
comparing with previously reported data of actigraphic
sleep measures in elderly individuals. A previous well-
designed study (Levenson et al., 2013), where partici-
pants’ age and actigraphic sleep measurement methods
were similar to those in our study, reported that SE was
the best predictor for insomnia because SE takes into
account TST, SOL, and WASO in its measurement, and
that SE was 81.3% and 83.7% in elderly individuals with
and without insomnia, respectively. These results were
also similar to our current data, although SE was 1–2%
higher and difference of SE between insomniacs and
non-insomniacs was 1% smaller in our study than that
in the previous study. On the other hand, in our study,
SE was 3.9% lower in the highest LAN quartile group
than that in the lowest quartile group (82.0% versus
85.9%, respectively). This difference of actigraphic SE
was larger than those observed between elderly individ-
uals with and without depressed mood in large-scale
TABLE 4. Actigraphic sleep parameters stratified by LAN exposure.
Quartiles of intensity of LAN (lux) [median, range]
0[50.1] 0.4 [0.1–0.8] 1.7 [0.8–3.4] 9.7 [43.5]
No. of participants (missing) 205 (5) 208 (2) 207(4) 208 (2)
Unadjusted Mean (5%–95% range) p
trend
SE, % 86.3 (75.5–93.7) 85.3 (71.8–93.1) 84.5 (72.4–93.1) 81.8 (65.3–91.5) 50.001
Log-ransformed SOL, log min 2.6 (0.8–4.1) 2.8 (1.4–4.4) 3.1 (1.3–4.3) 3.4 (1.8–4.7) 50.001
WASO, min 66.0 (27.4–128.5) 73.0 (30.5–144.3) 79.5 (29.0–155.8) 96.4 (40.8–201.0) 50.001
TST, min 416.5 (312.0–530.8) 417.1 (314.9–523.3) 428.3 (313.0–556.5) 425.7 (314.3–542.0) 0.07
Sleep-mid time, clock time 02:26 (0:37–03:46) 02:40 (01:14–04:00) 02:43 (01:24–03:58) 02:42 (00:57–04:12) 0.003
Adjusted* Mean (95% CI) p
trend
SE, % 85.9 (84.9–87.0) 85.2 (84.1–86.2) 84.7 (83.7–85.7) 82.0 (81.0–83.0) 50.001
Log-transformed SOL, log min 2.7 (2.6–2.8) 2.8 (2.7–3.0) 3.0 (2.9–3.2) 3.3 (3.2–3.4) 50.001
WASO, min 71.1 (65.7–76.5) 75.5 (70.2–80.8) 77.3 (71.9–82.6) 92.1 (86.7–97.4) 50.001
TST, min 430.0 (424.6–435.4) 425.8 (420.5–431.1) 423.9 (418.6–429.3) 409.0 (403.7–414.3) 50.001
Sleep-mid time, clock time 02:34 (02:31–02:37) 02:34 (02:31–02:37) 02:39 (02:36–02:42) 02:43 (02:40–02:46) 50.001
LAN, light at night; SE, sleep efficiency; SOL, sleep-onset latency; WASO, wake after sleep-onset; TST, total sleep time; CI, confidence
interval; UME, urinary 6-sufatoxymelatonin excretion.
*Adjusted for age, gender, body mass index, daytime physical activity, log-transformed UME, bedtime, rising time, daylength (per quartile
increment) using ANCOVA.
980 K. Obayashi et al.
Chronobiology International
Downloaded by [Kenji Obayashi] at 14:28 11 August 2015
community-based studies (Maglione et al., 2012; Paudel
et al., 2008).
Causality of LAN exposure with regard to insomnia
could not be ascertained in the present study because it
was a cross-sectional design. Some participants with
poor sleep may turn on light in the bedrooms. In
addition, epidemiological data in elderly individuals
have demonstrated that poor sleep quality causes
psychiatric and neurodegenerative disorders such as
depression and dementia (Cricco et al., 2001; Yokoyama
et al., 2010), and recent experimental studies conducted
in mice showed that chronic exposure to dim LAN
(5 lux) causes depressed mood and impaired cognition
compared with complete darkness at night (Bedrosian
et al., 2013; Fonken et al., 2012). When considering
together, the present study may indicate that LAN
exposure possibly and partly explains the risk for
psychiatric and neurodegenerative disorders in poor
sleepers. Further research using a longitudinal design is
required to improve our understanding of the associ-
ation between LAN, sleep quality, and psychiatric and
neurodegenerative disorders.
Although the neurophysiologic mechanisms that
mediate LAN-induced insomnia are not fully under-
stood, possible mechanisms are suggested by
data regarding circadian phase delay. LAN is a well-
established suppressor of melatonin secretion, and
endogenous melatonin levels are closely associated
with sleepiness (Dijk & Cajochen, 1997; Zeitzer et al.,
2000). Therefore, melatonin is hypothesized to be one of
the major internal contributors to the association
between LAN exposure and insomnia, although the
present study showed no significant associations
between LAN exposure and UME in home settings.
In addition, we did not measure potential melatonin
suppression by LAN exposure. However, the present
study indicated a significant association between LAN
exposure and sleep disturbances even after adjusting
for UME, an index of endogenous melatonin amplitude.
Whereas, changes in circadian biological phase remain
to be a potential mediator to the association between
LAN exposure and insomnia. According to the circadian
phase–response curve to light (Khalsa et al., 2003), LAN
delays the subsequent circadian phase, which may be a
potential risk for circadian misalignment between the
internal and environmental rhythms. Chronic circadian
misalignment is associated with the disruption of SCN
function and increased risk of insomnia (Boudreau
et al., 2013; Wyatt et al., 1999). In the present study, a
modest but significant association trend was observed
between an increase in LAN intensity and a delay in
sleep-mid time, which is correlated with a marker of
circadian phase (Burgess et al., 2003).
The strength of the present study included a large
study sample size. This advantage enabled us to analysis
the associations between LAN intensity and outcomes
for actigraphic measurements, although our previous
study was limited to report the association between LAN
and SOL (Obayashi et al., 2014). In addition to its cross-
sectional design, the present study has three limitations.
The first was that the light meters used in our study were
not ambulatory devices; thus, LAN intensity may have
been underestimated because the light meters did not
record light exposure in rooms other than the bedroom
(e.g. during nocturia). Further research using ambula-
tory eye-level light meters is required to assess the
association between LAN exposure in home settings and
sleep quality. However, compared with a wrist light
meter, the fixed bedroom light meter used in the present
study presents the great advantage of never being
covered by bed linen or night clothes. Second, lighting
sources could not be distinguished, e.g. bedroom light
or morning sun light entering the bedroom, and we have
no information related to the location of the windows in
the bedroom. Therefore, interpretation of our results
would be limited to the association between total
amounts of LAN intensity and sleep quality. The third
limitation was non-random sampling because partici-
pants were recruited with the cooperation of local
resident associations and elderly-resident clubs, which
may have led to a selection bias. However, the gener-
alizability of our findings may be acceptable because
some basic data (e.g. BMI and eGFR) were consistent
with those of the National Health and Nutrition Survey
in Japan in 2010 (The National Health and Nutrition
Survey Japan, 2010).
In conclusion, the present study demonstrated that
LAN exposure in home settings was significantly
associated with both subjectively and actigraphically
measured sleep quality in a general elderly population.
ACKNOWLEDGMENTS
We are indebted to all participants of this study.
We would also like to thank Sachiko Uemura and
Naomi Takenaka for their valuable support during the
data collection.
DECLARATION OF INTEREST
All authors report no conflicts of interest. This work was
supported by Grants from the Department of Indoor
Environmental Medicine, Nara Medical University;
Scientific Research from the Ministry of Education,
Culture, Sports, Science and Technology; Mitsui
Sumitomo Insurance Welfare Foundation; Meiji Yasuda
Life Foundation of Health and Welfare; Osaka Gas Group
Welfare Foundation; Japan Diabetes Foundation; and the
Japan Science and Technology Agency.
REFERENCES
Allebrandt KV, Teder-Laving M, Akyol M, et al. (2010). CLOCK gene
variants associate with sleep duration in two independent
populations. Biol Psychiatry. 67:1040–7.
Real-life link between LAN and insomnia 981
!Informa Healthcare USA, Inc.
Downloaded by [Kenji Obayashi] at 14:28 11 August 2015
Baskett JJ, Cockrem JF, Antunovich TA. (1998).
Sulphatoxymelatonin excretion in older people: Relationship
to plasma melatonin and renal function. J Pineal Res. 24:58–61.
Bedrosian TA, Weil ZM, Nelson RJ. (2013). Chronic dim light at
night provokes reversible depression-like phenotype: Possible
role for TNF. Mol Psychiatry. 18:930–6.
Boudreau P, Dumont GA, Boivin DB. (2013). Circadian adaptation
to night shift work influences sleep, performance, mood and
the autonomic modulation of the heart. PLoS One. 8:e70813.
Burgess HJ, Savic N, Sletten T, et al. (2003). The relationship
between the dim light melatonin onset and sleep on a regular
schedule in young healthy adults. Behav Sleep Med. 1:102–14.
Buysse DJ, Reynolds 3rd CF, Monk TH, et al. (1989). The Pittsburgh
Sleep Quality Index: A new instrument for psychiatric practice
and research. Psychiatry Res. 28:193–213.
Cajochen C, Zeitzer JM, Czeisler CA, Dijk DJ. (2000). Dose-response
relationship for light intensity and ocular and electroencepha-
lographic correlates of human alertness. Behav Brain Res. 115:
75–83.
Cajochen C, Mu
¨nch M, Kobialka S, et al. (2005). High sensitivity of
human melatonin, alertness, thermoregulation, and heart rate
to short wavelength light. J Clin Endocrinol Metab. 90:1311–16.
Calem M, Bisla J, Begum A, et al. (2012). Increased prevalence of
insomnia and changes in hypnotics use in England over 15
years: Analysis of the 1993, 2000, and 2007 National Psychiatric
Morbidity Surveys. Sleep. 35:377–84.
Chang AM, Santhi St N, Hilaire M, et al. (2012). Human responses
to bright light of different durations. J Physiol. 590:3103–12.
Cricco M, Simonsick EM, Foley DJ. (2001). The impact of insomnia
on cognitive functioning in older adults. J Am Geriatr Soc. 49:
1185–9.
Czeisler CA, Allan JS, Strogatz SH, et al. (1986). Bright light resets
the human circadian pacemaker independent of the timing of
the sleep-wake cycle. Science. 233:667–71.
Dew MA, Hoch CC, Buysse DJ, et al. (2003). Healthy older adults’
sleep predicts all-cause mortality at 4 to 19 years of follow-up.
Psychosom Med. 65:63–73.
Dijk DJ, Cajochen C. (1997). Melatonin and the circadian regula-
tion of sleep initiation, consolidation, structure, and the sleep
EEG. J Biol Rhythms. 12:627–35.
Eaker, ED, Pinsky J, Castelli WP. (1992). Myocardial infarction and
coronary death among women: Psychosocial predictors from a
20-year follow-up of women in the Framingham Study. Am J
Epidemiol. 135:854–64.
Fonken LK, Kitsmiller E, Smale L, Nelson RJ. (2012). Dim nighttime
light impairs cognition and provokes depressive-like responses
in a diurnal rodent. J Biol Rhythms. 27:319–27.
Ford DE, Kamerow DB. (1989). Epidemiologic study of sleep
disturbances and psychiatric disorders: An opportunity for
prevention? JAMA. 262:1479–84.
Khalsa SB, Jewett ME, Cajochen C, Czeisler CA. (2003). A phase
response curve to single bright light pulses in human subjects.
J Physiol. 549:945–52.
Levenson JC, Troxel WM, Begley A, et al. (2013). A quantitative
approach to distinguishing older adults with insomnia from
good sleeper controls. J Clin Sleep Med. 9:125–31.
Lockley SW, Evans EE, Scheer FA, et al. (2006). Short-wavelength
sensitivity for the direct effects of light on alertness, vigilance,
and the waking electroencephalogram in humans. Sleep. 29:
161–8.
Maglione JE, Ancoli-Israel S, Peters KW, et al. (2012). Depressive
symptoms and subjective and objective sleep in community-
dwelling older women. J Am Geriatr Soc. 60:635–43.
National Astronomical Observatory of Japan. (2013). Available at:
http://eco.mtk.nao.ac.jp/koyomi/index.html.en (last accessed
on 30 December 2013).
Navara KJ, Nelson RJ. (2007). The dark side of light at night:
Physiological, epidemiological, and ecological consequences.
J Pineal Res. 43:215–24.
Obayashi K, Saeki K, Iwamoto J, et al. (2012). Positive effect of
daylight exposure on nocturnal urinary melatonin excretion in
the elderly: A cross-sectional analysis of the HEIJO-KYO study.
J Clin Endocrinol Metab. 97:4166–73.
Obayashi K, Saeki K, Iwamoto J, et al. (2014). Effect of exposure to
evening light on sleep initiation in the elderly: A longitudinal
analysis for repeated measurements in home settings.
Chronobiol Int. 31:461–7.
Paudel ML, Taylor BC, Diem SJ, et al. (2008). Association between
depressive symptoms and sleep disturbances in community-
dwelling older men. J Am Geriatr Soc. 56:1228–35.
Philips Respironics Actiware Tutorials. Available at: http://lear-
nactiware.com/tutorials/ (last accessed on 30 June 2014).
Portaluppi F, Smolensky MH, Touitou Y. (2010). Ethics and
methods for biological rhythm research on animals and
human beings. Chronobiol Int. 27:1911–29.
Prinz PN, Vitiello MV, Raskind MA, Thorpy MJ. (1990). Geriatrics:
Sleep disorders and aging. N Engl J Med. 323:520–6.
The National Health and Nutrition Survey Japan. (2010).
Available at: http://www.mhlw.go.jp/bunya/kenkou/eiyou/
h22-houkoku.html (Japanese; last accessed on 30 December
2013).
Walsh JK, Coulouvrat C, Hajak G, et al. (2011). Nighttime insomnia
symptoms and perceived health in the America Insomnia
Survey (AIS). Sleep. 34:997–1011.
Wyatt JK, Ritz-De Cecco A, Czeisler CA, Dijk DJ. (1999). Circadian
temperature and melatonin rhythms, sleep, and neurobeha-
vioral function in humans living on a 20-h day. Am J Physiol.
277:R1152–63.
Yokoyama E, Kaneita Y, Saito Y, et al. (2010). Association between
depression and insomnia subtypes: A longitudinal study on the
elderly in Japan. Sleep. 33:1693–702.
Zeitzer JM, Dijk D, Kronauer RE, et al. (2000). Sensitivity of
the human circadian pacemaker to nocturnal light:
Melatonin phase resetting and suppression. J Physiol. 526:
695–702.
982 K. Obayashi et al.
Chronobiology International
Downloaded by [Kenji Obayashi] at 14:28 11 August 2015
