SLEEPJ, 2019, 1–8

doi: 10.1093/sleep/zsz160
Advance Access Publication Date: 15 July 2019
Original article

Original article
Dry eye and sleep quality: a large community-based study in

Downloaded from https://academic.oup.com/sleep/article/42/11/zsz160/5532656 by guest on 19 April 2024

Hangzhou
Xiaoning Yu1,2,3,†, Huilan Guo2,3,4,†, Xin Liu1,2,3, Guowei Wang2,3,4, Yan Min5, ,
Shih-Hua Sarah Chen5, Summer S. Han6, , Robert T. Chang7, , Xueyin Zhao2,3,4, Ann Hsing5, ,
Shankuan Zhu2,3,4,*, and Ke Yao1,*
1
Eye Center, The Second Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, China, 2Chronic Disease Research
Institute, School of Public Health, School of Medicine, Zhejiang University, Hangzhou, China, 3Women’s Hospital, School of Medicine,
Zhejiang University, Hangzhou, China, 4Department of Nutrition and Food Hygiene, School of Public Health, School of Medicine,
Zhejiang University, Hangzhou, China, 5Stanford Prevention Research Center, Department of Medicine, Stanford School of Medicine,
Stanford University, Stanford, CA, 6Department of Neurosurgery, Stanford School of Medicine, Stanford University, Stanford, CA,
7
Department of Ophthalmology, Stanford School of Medicine, Stanford University, Stanford, CA

*Corresponding authors. Ke Yao, Eye Center, The Second Affiliated Hospital, School of Medicine, Zhejiang University, Jiefang Road, Hangzhou 310058,
Zhejiang, China. Email: xlren@zju.edu.cn; Shankuan Zhu, Department of Nutrition and Food Hygiene, School of Public Health, School of Medicine,
Zhejiang University, 866 Yu-hang-tang Road, Hangzhou 310058, Zhejiang, China. Email: zsk@zju.edu.cn.

†
These authors contributed equally to this work.

Abstract
Study Objectives: To investigate the relationship between dry eye and sleep quality in a large community-based Chinese population.

Methods: A total of 3,070 participants aged 18–80 were recruited from a community-based study in Hangzhou, China during 2016–2017. Sleep quality was evaluated
using the Chinese version of the Pittsburgh Sleep Quality Index (CPSQI), and dry eye was evaluated using the Ocular Surface Disease Index (OSDI) questionnaire.
Multivariable linear regression and logistic regression models were used to investigate the associations, adjusting for age, smoking, drinking, season, and other
potential confounders.

Results: Overall, CPSQI score and sleep dysfunction were significantly associated with mild, moderate, and severe dry eye (ORs for CPSQI score: 1.07, 1.13, 1.14,
all p < 0.001; for sleep dysfunction: 1.31, 1.73, 1.66, all p < 0.05). Furthermore, worse OSDI score was presented in participants with worse CPSQI score or sleep
dysfunction (CPSQI score > 7) (β: 0.13, 0.54; all p < 0.001). In addition, six of the seven components of CPSQI showed significant associations with dry eye (all p < 0.001),
except for the component of sleep medication use. Moreover, we observed significant associations of dry eye in all three subscales of OSDI with CPSQI score and
sleep dysfunction.

Conclusion: Our large, community-based study showed a strong association between poor sleep quality and an increased severity of dry eye, suggesting that
preventing either one of the discomforts might alleviate the other.

Statement of Significance
Dry eye and sleep dysfunction draw global public health concerns with their high prevalence and extensive adverse effects. Our study dis-
covered a strong association between these two conditions. This is the first population-based study to evaluate the association of dry eye
and sleep quality using previously validated tools, the Ocular Surface Disease Index (OSDI) and the Chinese version of the Pittsburgh Sleep
Quality Index (CPSQI), respectively. Results indicated a strong positive association between poor sleep quality and higher severity of dry eye.
It is plausible to suggest that improvement of sleep quality would alleviate the syndromes of dry eye, and vice versa.

Key words: dry eye; sleep quality; the Ocular Disease Index; the Pittsburgh Sleep Quality Index

Submitted: 29 October, 2018; Revised: 16 May, 2019

© Sleep Research Society 2019. Published by Oxford University Press on behalf of the Sleep Research Society.
All rights reserved. For permissions, please e-mail journals.permissions@oup.com.

1
 2 | SLEEPJ, 2019, Vol. 42, No. 11
Introduction
In recent years, sleep dysfunction has become a major social
concern throughout the world [1, 2]. It is established that sleep
quality is vital for physical and psychological health, and that
sleep dysfunction impairs autonomic and endocrine functions,
resulting in extensive changes in many aspects of the body
system [2]. Previous studies have shown that sleep dysfunction
is associated with an increased risk of diabetes, hypertension,
and depression [3, 4]. However, little is known about its relation-
ship with the ocular surface diseases [5–7].
Dry eye, a multifactorial chronic disease of the ocular sur-
face, raises significant public health concerns worldwide [5–7].
The prevalence of dry eye varies from 5% to 30% according to
several epidemiologic studies in various ethnic groups using
different diagnostic methods [5, 6]. Dry eye is triggered by a
reduced tear production and/or increased tear evaporation,
causing the ocular surface inflammation and eventually, leading
to the destabilization of the cornea–tear interface [7]. Dry eye
not only causes significant ocular surface discomfort (e.g. pain,
irritation, foreign body sensation, and photophobia) and visual
disturbance but also interferes with daily activities and affects
the sufferer’s quality of life [8, 9]. Current therapies for dry eye
are unsatisfactory, as most therapies target inflammation, tear-
film instability, and aqueous deficiency to alleviate local symp-
toms and fail to address the underlying causes, which makes
the effects of treatment unpredictable, especially for more se-
vere cases [10, 11]. Understanding the causes and risk factors of
dry eye would help alleviate the problem.
Because the prevalence of dry eye and poor sleep is quite
high in the general population, the association between dry eye
and sleep quality has recently attracted more attention [12]. It
is estimated that more than 40% of people with dry eye suffer
from poor sleep quality [13, 14]. Few epidemiologic studies have
investigated the association between dry eye and sleep quality,
with most are hospital-based studies conducted in the United
States, South Korea, Japan, and Turkey [12, 14, 15], focusing on
special populations, such as obstructive sleep apnea syndrome
patients, veterans, or users of visual-display technologies [13, 16,
17], and evaluating only part of sleep quality (such as sleep dur-
ation) [18], using relatively small sample sizes [13, 14, 16, 19–21].
Thus, to further investigate the association between dry eye and
poor sleep quality on a larger scale study, we screened dry eye
and sleep quality in a community-based study of more than
3,000 individuals in the urban area of Hangzhou, China.
Methods
Study population
We recruit a total of 3,070 participants aged from 18 to 80 were
recruited from Xihu District, Hangzhou, China, over an 8-month
period (November 2016 to July 2017). The selected age range
accounted for approximately 80% of the total population with
690,935 residents in Xihu District. Within Xihu district, there are
two administration levels: subdistrict and community. To ensure
representativeness of the study subjects and cover the spectrum
across age groups, all subdistricts and the communities under
each subdistrict were sampled, and quota sampling was applied
to recruit each individual. The stratifications were based on sex
and age. The selected samples were equally distributed across
all subdistricts, and proportional to the population size of each
community. With the response rate of 78.9%, a total of 3,070 in-
dividuals were included in this cohort. All measurements and
questionnaire data of each participant were collected on the
same day within approximately 8 h. To avoid the confounding
conditions that might affect sleep quality and dry eye, we ex-
cluded 136 subjects, including 83 with thyroid diseases (48 with
hyperthyroidism, 33 with hypothyroidism, and 2 with other
thyroid diseases), 38 with mental disease (6 with anxiety dis-
order, 14 with depressive disorder, 5 with anxiety disorder and
depressive disorder, 4 with schizophrenia disorder, 1 with anx-
iety disorder and schizophrenia disorder, 1 with cognitive dis-
order patient, and 7 with other mental diseases), and 15 with
neurologic diseases (2 with Parkinsonism, 1 with myasthenia
gravis, 1 with optic atrophy, 2 with pediatric epilepsy, 2 with
Downloaded from https://academic.oup.com/sleep/article/42/11/zsz160/5532656 by guest on 19 April 2024
migraine, 2 with neuralgia, 1 with Meniere’s syndrome, 1 with
epilepsy, and 3 with other neurologic diseases). We further ex-
cluded 104 subjects due to missing data on the Ocular Surface
Disease Index (OSDI) score (n = 91), the Pittsburgh Sleep Quality
Index (CPSQI) score (n = 1), body mass index (BMI, n = 9), smoking
(n = 3), leaving 2,830 participants for final analysis. We found no
significant difference in age and gender between the 240 ex-
cluded and 2,830 included in analysis. The Institutional Review
Board (IRB)/Ethics Committee at Zhejiang University School of
Public Health, Hangzhou, China and Stanford University ap-
proved the study. Written informed consent was obtained from
each participant. The present research adhered to the tenets of
the Declaration of Helsinki.
Data collection
Baseline demographic characteristics, behavioral risk factors
(smoking and alcohol drinking), medical history (thyroid dis-
eases, autoimmune disease, mental disorders, neurologic dis-
eases, diabetes, and hypertension), and menopausal status
in women were collected through face-to-face interviews.
Overnight fasting blood was collected. Clinical markers,
including comprehensive metabolic panel, lipid panel, and liver
panel were obtained. Blood pressure was measured for each par-
ticipant after 5–10 min of rest for three times with an interval of
5 min (Omron HBP-9020, Omron, Kyoto, Japan), and the average of
the three measurements was used. Hypertension was defined as
systolic blood pressure ≥140 mm Hg or diastolic blood pressure
≥90 mm Hg, or self-reported clinically diagnosed hypertension.
Diabetes was defined as fasting plasma glucose ≥7.0 mmol/L
or self-reported clinically diagnosed diabetes. Physical activity
was assessed using a short form of the International Physical
Activity Questionnaire, and was classified into low, moderate,
and high levels based on the tertiles of metabolic equivalence
tasks. Psychological status were evaluated by the World Health
Organization 5-item well-being index (WHO5), which is widely
used to assess subjective psychological well-being, and screen
for depression [22]. Based on the prevalent meteorological fac-
tors (wind, humidity, temperature, etc.) in the study area, the
seasons in which our evaluation took place (during November
2016 and July 2017) were classified into four categories: au-
tumn (November 1, 2016–November 23, 2016), winter (November
24, 2016–March 19, 2017), spring (March 20, 2017–May 9, 2017),
and summer (May 9, 2017—July 31, 2017). Height was measured
without shoes to the precision of 0.1 cm. Weight was measured

to the precision of 0.1 kg (SECA704, Hamburg, Germany). BMI was
calculated as weight (kg)/height (m2). Smoking was categorized
into three groups: never, former, and current smokers. Alcohol
drinking was categorized into four groups: never, former, current
light drinkers (defined as the participants who have drunk less
than 12 times per year), and current heavy drinkers (defined as
the participants who have drunk more than 12 times per year).
Education level was categorized as illiteracy or primary school,
middle or high school, and college or above. Menopause status
in women was divided into pre- and postmenopausal status.
Postmenopause was defined as a complete natural cessation of
menses for more than 12 months. Self-reported used of medica-
tions was elicited by face-to-face interviews and classified into
four categories: cardio-cerebrovascular diseases medications,
endocrine or metabolic diseases medications, osteoarticular
diseases medications, and other medications (including medi-
cations for cancer medications, allergic diseases, and renal, re-
spiratory and digestive system diseases).
The Ocular Disease Index questionnaire
We used the Ocular Disease Index (OSDI) questionnaire was
used to assess self-reported severity of dry eye. OSDI consists of
12 questions covering three subscales: typical symptom (three
questions), vision function (six questions), and environmental
triggers (three questions). Based on the total OSDI score, partici-
pants were classified into different severities of dry eye: normal
(score 0–12), mild (score 13–22), moderate (score 23–32), and se-
vere (score 33–100) [23].
The Chinese Version of Pittsburgh Sleep Quality
Index questionnaire
The Chinese Pittsburgh Sleep Quality Index (CPSQI) was ad-
ministered to measure sleep quality of the participants. The
Pittsburgh Sleep Quality Index (PSQI) questionnaire, an effective
and useful tool widely used in large-scale epidemiologic studies
for measuring subjective sleep quality, was customized and
validated in the Chinese population [24]. The CPSQI showed
an overall reliability coefficient of 0.82–0.83 and test–retest re-
liability of 0.77–0.85 for Chinese adults [25]. In addition, a CPSQI
score over 7, which had a diagnostic of 98.3% and specificity
of 90.2% in distinguishing normal subjects from patients with
sleep quality problems, has been recommended in Chinese
clinical practice and research [26–29].The CPSQI questionnaire
consists of 18 items that generated seven components of sleep
assessment, including subjective sleep quality, sleep latency,
sleep duration, habitual sleep efficiency, sleep disturbance, sleep
medication use, and daytime dysfunction. The score for each
component ranges from 0 to 3, totaling to a CPSQI score from
0 to 21 [21, 23]. A higher score of CPSQI indicates poorer sleep,
and a total score greater than 7 was defined as sleep dysfunction
[23, 24]. In the current study, both total CPSQI score (continuous
variable) and sleep dysfunction (binary variable) were used to
evaluate sleep quality.
Statistical analysis
We used mean and standard deviation to describe continuous
variables and percentage (%) to describe categorical variables.
Yu et al. | 3
To assess statistical differences, we used ANOVA test for con-
tinuous variables and chi-square test for categorical vari-
ables. To investigate the associations between dry eye severity
(normal, mild, moderate, and severe) and CPSQI score/sleep dys-
function, we used multinomial logistic regression models with
the normal group as the reference. Multivariable linear regres-
sion analysis was used to assess the associations of CPSQI score
and OSDI score. The effect sizes were evaluated by calculating
R2 (coefficient of determination) and Cohen’s f2, which are com-
monly employed indices of effect size [30]. All models were ad-
justed for sex, age, BMI, smoking, alcohol drinking, education,
physical activity, diabetes, hypertension, WHO5 scores, season,
medications use, and menopausal status in women. In add-
ition, to exclude the effect of medications use, we did sensitivity
analysis by excluding the participants taking medications. We
Downloaded from https://academic.oup.com/sleep/article/42/11/zsz160/5532656 by guest on 19 April 2024
log-transformed the OSDI score and its subscale scores due to
data skewness. We further tested the interactions between sex
and CPSQI in the multivariable logistic regression models. No
significant interactions were found. Therefore, men and women
were analyzed together throughout the study. All statistical ana-
lyses were conducted using Stata version 13 (Stata Corporation,
College Station, TX). A two-tailed p value <0.05 was considered
statistically significant, otherwise indicated.
Results
The prevalence of dry eye evaluated by typical symptoms and
demographic characteristics of the 2,830 participants were sum-
marized in Table 1. Overall, 39.8% of the subjects reported some
signs of dry eye. Dry eye was classified into four categories based
on OSDI scores: normal (60.2%), mild (20.9%), moderate (8.8%),
and severe (10.0%). Compared with participants without typical
dry eye symptoms, those with dry eye tended to drink less, be
female, and of older age. Furthermore, dry eye was positively as-
sociated to hypertension, education level, season, medications
use, WHO5 scores and CPSQI score.
Table 2 showed the odds ratios (ORs) for dry eye in relation to
CPSQI and sleep dysfunction from multinomial logistic regres-
sion analyses. As shown, a poorer total CPSQI score was signifi-
cantly associated with a higher severity of dry eye. Specifically,
compared with the normal group, participants with sleep dys-
function had an increased risk of dry eye severity, with an OR
of 1.31 (95% CI 1.03–1.67), 1.73 (95% CI 1.25–2.38), and 1.66 (95%
CI 1.22–2.26) for the mild, moderate, and severe group, respect-
ively (the results for unadjusted models were also calculated,
and attached in the Supplementary Table S1). In addition, poor
subjective sleep quality, long sleep latency, short sleep duration,
poor habitual sleep efficiency, sleep disturbance, and daytime
dysfunction were significantly associated with higher risk of dry
eye. These associations remained significant after Bonferroni
corrections at the alpha level of 0.0056. And the results remained
similar after excluding patients taking medications.
As presented in Table 3, after adjusting for age, sex BMI,
smoking, drinking, physical activity, education, diabetes, hyper-
tension, WHO5 scores, season, and menopausal status in women,
participants with sleep dysfunction (CPSQI score > 7) had a
worse OSDI score (p < 0.001). (Supplementary Table S2 presents
the results in unadjusted model). Six of the seven CPSQI com-
ponents, with the exception of sleep medication use, showed
significant associations with OSDI score after Bonferroni correc-
tions (p < 0.05/9 = 0.0056). Besides, the indices of R2 and Cohen’s
 4 | SLEEPJ, 2019, Vol. 42, No. 11

Table 1. Characteristics of study participants by dry eye severity (n = 2,830)

Ocular Surface Disease Index (mean ± SD) or N (%)

Variables Total Normal Mild Moderate Severe P

N (%) 2,830 1,704 (60.2) 592 (20.9) 250 (8.8) 284 (10.0)
Age (mean and SD) 50.7 ± 13.9 48.8 ± 13.5 52.3 ± 13.8 54.2 ± 14.1 55.5 ± 14.3 <0.001
Gender (men %) 1,257 (44.4) 797 (46.7) 275 (46.5) 88 (35.2) 97 (34.2) <0.001
BMI (mean and SD) 23.5 ± 3.3 23.6 ± 3.4 23.6 ± 3.2 23.3 ± 3.2 23.1 ± 3.3 0.045
Drinking status 0.022
Nondrinkers 1,442 (50.95) 837 (49.1) 298 (50.3) 137 (54.80) 170 (59.9)
Occasional drinkers 675 (23.85) 417 (24.5) 141 (23.8) 58 (23.2) 59 (20.8)
Regular drinkers 646 (22.83) 414 (24.3) 136 (23.0) 51 (20.4) 45 (15.9)
Former drinkers 67 (2.37) 36 (2.11) 17 (2.87) 4 (1.6) 10 (3.5)
Smoking 0.123
Nonsmokers 2,010 (71.0) 1,201 (70.5) 409 (69.1) 188 (75.2) 212 (74.7)

Downloaded from https://academic.oup.com/sleep/article/42/11/zsz160/5532656 by guest on 19 April 2024

Current smokers 221 (7.8) 126 (7.4) 60 (10.1) 18 (7.2) 17 (6.0)
Former smokers 599 (21.2) 377 (22.1) 123 (20.8) 44 (17.6) 55 (19.4)
Physical activity 0.130
Low 660 (23.3) 379 (22.2) 142 (24) 56 (22.4) 83 (29.2)
Middle 1,780 (62.9) 1,078 (63.3) 366 (61.8) 56 (22.4) 172 (60.6)
High 390 (13.8) 247 (14.5) 84 (14.2) 30 (12.0) 29 (10.2)
Education <0.001
Illiteracy and primary school 743 (26.3) 394 (23.1) 176 (29.7) 61 (24.4) 112 (39.4)
Middle or high school 1,272 (45.0) 783 (46.0) 251 (42.4) 124 (49.6) 114 (40.1)
University and above 815 (28.8) 527 (30.9) 165 (27.9) 65 (26.0) 58 (20.4)
Diabetes 249 (8.8) 136 (8.0) 50 (8.5) 28 (11.2) 35 (12.3) 0.053
Hypertension 823 (29.1) 424 (24.9) 206 (34.8) 82 (32.4) 113 (39.4) <0.001
WHO-5 score 24.5 ± 0.8 24.7 ± 0.7 24.4 ± 0.8 24.3 ± 0.9 24.1 ± 1.1 <0.001
Season of year <0.001
Spring 331 (11.7) 205 (12.0) 59 (10) 24 (9.6) 43 (15.1)
Summer 1,776 (62.8) 1,134 (66.6) 351 (59.3) 156 (62.4) 135 (47.5)
Autumn 68 (2.4) 30 (1.8) 24 (4.1) 7 (2.8) 7 (2.5)
Winter 655 (23.1) 335 (19.7) 158 (26.7) 63 (25.2) 99 (34.9)
Medications use <0.001
Cardio-cerebrovascular diseases 434 (15.3) 228 (13.4) 114 (19.3) 44 (17.6) 48 (16.9)
Cancer 21 (0.7) 13 (0.8) 1 (0.2) 2 (0.80) 5 (1.8)
Endocrine and metabolic diseases 207 (7.3) 112 (6.6) 42 (7.1) 24 (9.6) 29 (10.2)
Osteoarticular diseases 86 (3.0) 34 (2.0) 23 (3.9) 15 (6.0) 14 (4.9)
Respiratory system diseases 32 (1.1) 18 (1.1) 10 (1.7) 10 (1.7) 3 (1.1)
Digestive system diseases 115 (4.1) 61 (3.6) 24 (4.1) 11 (4.4) 19 (6.7)
Renal diseases 31 (1.1) 13 (0.8) 3 (0.5) 2 (0.8) 13 (4.6)
Allergic diseases 38 (1.3) 20 (1.2) 6 (1.0) 6 (2.4) 6 (2.11)
Total CPSQI score 5.1 ± 3.0 4.7 ± 2.8 5.4 ± 3.1 6.1 ± 3.1 6.5 ± 3.4 <0.001
Components of CPSQI
Subjective sleep quality 1.0 ± 0.7 1.0 ± 0.7 1 ± 0.7 1.2 ± 0.8 1.2 ± 0.7 <0.001
Sleep latency 1.0 ± 0.9 0.9 ± 0.9 1 ± 0.9 1.2 ± 1 1.2 ± 1.0 <0.001
Sleep duration 0.8 ± 0.7 0.8 ± 0.7 0.9 ± 0.8 0.9 ± 0.8 1.0 ± 0.8 <0.001
Habitual sleep efficiency 0.8 ± 1.0 0.7 ± 1.0 0.9 ± 1.1 0.9 ± 1.1 1.1 ± 1.1 <0.001
Sleep disturbance 1.0 ± 0.5 0.9 ± 0.5 1.1 ± 0.5 1.2 ± 0.5 1.2 ± 0.5 <0.001
Sleep medication use 0.1 ± 0.4 0.1 ± 0.4 0.1 ± 0.5 0.1 ± 0.5 0.1 ± 0.6 0.090
Daytime dysfunction 0.4 ± 0.7 0.4 ± 0.6 0.5 ± 0.7 0.5 ± 0.7 0.6 ± 0.8 <0.001
Sleep dysfunction (CPSQI score > 7) 554 (19.6) 268 (15.7) 129 (21.8) 70 (28.0) 87 (30.6) <0.001

f2 which represent the effect sizes of the linear regression were
also summarized in Table 3, where the effect size was medium
to high in R2, and was close to medium in Cohen’s f2 [30].
The associations of CPSQI score and sleep dysfunction
with OSDI subscales (symptom, vision function, and environ-
ment) are given in Table 4. Significant associations of dry eye
in all three subscales of OSDI with CPSQI score and sleep dys-
function (the results for unadjusted models were presented in
Supplementary Table S3). Moreover, all CPSQI components, with
the exception of sleep medication, had significant associations
with all three subscales of OSDI after Bonferroni correction
(p < 0.05/27 = 0.0019). And the results remained similar after
excluding patients taking medications.
Discussion
In this large community-based study, we showed a significant
association between poorer sleep quality and dry eye, because
worse CPSQI score was associated to higher OSDI severity and
worse OSDI score. In addition, of the seven CPSQI subscales, six
were significantly associated with OSDI results, with the excep-
tion of the subclass of sleep medication use.
 Yu et al. | 5

Table 2. Odds ratios for dry eye disease in relation to Chinese version of Pittsburgh sleep quality index (CPSQI) score with dry eye severity based
on Ocular Surface Disease Index (OSDI)†in 2,830 individuals in urban Hangzhou

Dry eye severity based on OSDI

Mild Moderate Severe

CPSQI scores Normal OR (95% CI) P OR (95% CI) P OR (95% CI) P

Total CPSQI score Ref 1.07 <0.001* 1.13 <0.001* 1.14 <0.001*
(1.04 to 1.11) (1.08 to 1.18) (1.10 to 1.19)
Sleep dysfunction (CPSQI score > 7) Ref 1.31 0.030 1.73 0.001* 1.66 0.002*
(1.03 to 1.67) (1.25 to 2.38) (1.22 to 2.26)
Components of CPSQI
Subjective sleep quality Ref 1.19 0.015 1.64 <0.001* 1.50 <0.001*
(1.03 to 1.36) (1.36 to 2.00) (1.24 to 1.80)
Sleep latency Ref 1.10 0.086 1.33 <0.001* 1.27 0.001*
(0.99 to 1.21) (1.16 to 1.53) (1.11 to 1.46)

Downloaded from https://academic.oup.com/sleep/article/42/11/zsz160/5532656 by guest on 19 April 2024

Sleep duration Ref 1.19 0.009 1.24 0.021 1.41 <0.001*
(1.05 to 1.35) (1.03 to 1.49) (1.19 to 1.67)
Habitual sleep efficiency Ref 1.10 0.044 1.17 0.020 1.26 <0.001*
(1.00 to 1.21) (1.02 to 1.33) (1.12 to 1.42)
Sleep disturbance Ref 1.36 0.004* 1.94 <0.001* 1.86 <0.001*
(1.10 to 1.67) (1.45 to 2.59) (1.41 to 2.45)
Sleep medication use Ref 1.08 0.485 0.97 0.855 1.00 0.973
(0.87 to 1.34) (0.73 to 1.31) (0.76 to 1.31)
Daytime dysfunction Ref 1.43 <0.001* 1.62 <0.001* 1.70 <0.001*
(1.23 to 1.66) (1.33 to 1.98) (1.41 to 2.06)

†
Adjusted for age, sex, BMI, smoking, drinking, physical activity, education, diabetes, hypertension, WHO5 scores, medications use, season, and menopausal status in
women.
*P < 0.0056 (adjusted by Bonferroni correction, 0.05/9 tests). Each variable was analyzed in separate regression model.

Table 3. Beta coefficients with 95% confidence interval and the effect sizes of Chinese version of Pittsburgh sleep quality index (CPSQI) score
with Ocular Surface Disease Index (OSDI)† (n = 2,830)

OSDI scores

CPSQI scores β 95% CI P R2 ‡ Cohen’s f2 ‡

Total CPSQI score 0.13 0.10 to 0.16 <0.001* 0.104 0.116

Sleep dysfunction (CPSQI score > 7) 0.54 0.30 to 0.77 <0.001* 0.090 0.099
Components of CPSQI
Subjective sleep quality 0.33 0.20 to 0.46 <0.001* 0.091 0.101
Sleep latency 0.25 0.15 to 0.35 <0.001* 0.092 0.101
Sleep duration 0.31 0.19 to 0.44 <0.001* 0.091 0.101
Habitual sleep efficiency 0.18 0.09 to 0.27 <0.001* 0.089 0.097
Sleep disturbance 0.63 0.43 to 0.82 <0.001* 0.096 0.106
Sleep medication use −0.03 −0.25 to 0.18 0.773 0.084 0.091
Daytime dysfunction 0.53 0.38 to 0.67 <0.001* 0.100 0.112

†
Adjusted for age, sex, BMI, smoking, drinking, physical activity, education, diabetes, hypertension, WHO5 scores, medications use, season, and menopausal status in
women.
R (the coefficient of determination) and Cohen’s f2: the indices of the effect size.
‡ 2

*P < 0.0056 (adjusted by Bonferroni correction, 0.05/9 tests). Each variable was analyzed in separate regression model.

Previous studies indicated that poor sleep quality is more
common in individuals with dry eye rather than those with other
eye diseases [12]. Kawashima et al. [13] reported that the typical
dry eye symptoms and the Schirmer value were significantly as-
sociated with PSQI scores among the observed users of visual
display technologies. In a nationally representative, population-
based survey in Korea, people with sleep duration shorter than
5 h were found to be 20% more likely to suffer from dry eye, com-
pared with those with more than 6 h of sleep [18]. Galor et al.
[17] found that dry eye patients with severe ocular pain tend to
suffer from worse insomnia compared with those in mild and
moderate dry eye group. Ayaki et al. [20] reported that typical
treatment of dry eye would promote sleep quality evaluated by
PSQI scores. In addition, an intervention study conducted by Lee
et al. [31], also showed that sleep deprivation could induce tear
hyperosmolarity, shorten tear break-up time, and reduce tear
secretion, all of which would trigger the development of ocular
surface diseases such as dry eye. On the other hand, dry eye is
a classic symptom of primary Sjogren’s Syndrome (pSS) [32, 33].
Hackett et al. [33] reported that, compared with normal controls,
pSS patients were more likely to suffer from sleep dysfunction.
Our results not only consisted with the above studies, but also
 6 | SLEEPJ, 2019, Vol. 42, No. 11

Table 4. Association of the Chinese version of Pittsburgh sleep quality index (CPSQI) score with subscales of Ocular Surface Disease Index
(OSDI)† for dry eye in 2,830 study participants

OSDI

Symptom Vision function Environment

CPSQI scores β (95% CI) P β (95% CI) P β (95% CI) P

Total CPSQI score 0.14 (0.11 to 0.17) <0.001* 0.10 (0.07 to 0.14) <0.001* 0.12 (0.09 to 0.15) <0.001*
Sleep dysfunction (CPSQI score > 7) 0.70 (0.46 to 0.95) <0.001* 0.42 (0.17 to 0.68) 0.001* 0.56 (0.31 to 0.80) <0.001*
Components of CPSQI
Subjective sleep quality 0.40 (0.26 to 0.53) <0.001* 0.30 (0.16 to 0.44) <0.001* 0.30 (0.16 to 0.44) <0.001*
Sleep latency 0.25 (0.16 to 0.36) <0.001* 0.26 (0.16 to 0.37) <0.001* 0.24 (0.13 to 0.34) <0.001*
Sleep duration 0.38 (0.25 to 0.51) <0.001* 0.22 (0.08 to 0.35) 0.001* 0.24 (0.11 to 0.37) <0.001*
Habitual sleep efficiency 0.23 (0.14 to 0.33) <0.001* 0.09 (−0.01 to 0.19) 0.067 0.20 (0.10 to 0.30) <0.001*
Sleep disturbance 0.61 (0.41 to 0.82) <0.001* 0.54 (0.33 to 0.75) <0.001* 0.58 (0.37 to 0.78) <0.001*
Sleep medication use 0.02 (−0.20 to 0.24) 0. 843 −0.01 (−0.24 to 0.22) 0.913 −0.02 (−0.24 to 0.21) 0.893

Downloaded from https://academic.oup.com/sleep/article/42/11/zsz160/5532656 by guest on 19 April 2024

Daytime dysfunction 0.50 (0.35 to 0.64) <0.001* 0.44 (0.28 to 0.59) <0.001* 0.56 (0.41 to 0.71) <0.001*

†
Adjusted for age, sex, BMI, smoking, drinking, physical activity, education, diabetes, hypertension, WHO5 scores, medications use, season, and menopausal status in
women.
*P < 0.0019 (adjusted by Bonferroni correction, 0.05/27 tests). Each variable was analyzed in separate regression model.

furthered the evaluation of sleep quality in multidimensions,
suggesting that the various components of sleep quality as-
sessment in CPSQI were also mostly associated with dry eye,
including subjective sleep quality, sleep duration, and sleep la-
tency, habitual sleep efficiency, sleep disturbance, and daytime
dysfunction.
It is biologically plausible that poorer sleep quality leads to
dry eye. Sleep disorders tend to be associated with autonomic
dysfunction [34, 35],which would affect the parasympathetic
fibers in the lacrimal glands, leading to reduced tear secretion
[31]. In addition, the activation of the hypothalamic–pituitary–
adrenal axis during sleep could result in a relatively dehydrated
state, thereby reducing tear secretion [36, 37]. Other mechanisms
have also been proposed. For example, sleep disorders, such as
obstructive sleep apnea (OSA), can cause alterations in ocular
cytokines (e.g. tumor necrosis factor alpha, interleukin-1, and
interleukin-6) that contribute to topical inflammation [15, 38].
Continuous ocular surface inflammation could cause damage to
lacrimal glands, Meibomian glands, and conjunctival epithelium
[6], and consequently, suppressing tear production and upset-
ting tear film stability.
It is also reasonable to speculate that dry eye can lead to
poorer sleep quality. Patients with dry eye experienced deteri-
orated quality of life, resulting in depression, anxiety, and sleep
disorders [21]. It has been shown that dry eye treatment can
improve sleep quality [20]. However, to date, there is no suffi-
cient evidence to support a causal relationship between dry eye
and poorer sleep quality. Prospective studies are needed to in-
terrogate the temporal relationship between sleep quality and
dry eye.
Our data showed that use of sleep medications was not sig-
nificantly associated with dry eye symptoms, which was con-
sistent with results from a study from Japan [19]. However, only
3.7% of the participants reported having used sleep medications
in our study, limiting the power of detection small effect. Sleep
medications mainly include benzodiazepines, hypnotic benzodi-
azepine receptor agonists, antidepressants, antihistamines, etc.
[39]. Former researchers indicated that antidepressants would
promote dry eye symptoms via the anticholinergic adverse ef-
fects (to suppress the cholinergic nerve fiber in Meibomian gland
and lacrimal gland) or other potential pharmacological mech-
anism that decreases tear secretion [40], while the association
between most other sleep medications and dry eye remains
unclear. Given the possibility that diverse sleep medicines may
have different influences on tear secretion, it is essential to dis-
tinguish the association among sleep medications of different
mechanism with dry eye symptoms. Above all, the association
needs to be evaluated in a larger study with a sufficient number
of sleep medication users and include a spectrum of sleep medi-
cations to help clarify pharmacological mechanism associated
to different sleep medications.
Our study has several notable strengths. To our knowledge,
this is the first large-scale, population-based study assessing
the association between dry eye and sleep quality as well as
various components of sleep quality. Our study is in a large
scale with a high response rate, and covered a wide range of
age spectrum and education groups, thereby providing rep-
resentative samples and data to enhance generalizability.
This study applied self-reported assessment of dry eye and
sleep quality through face-to-face interview using OSDI and
the CPSQI questionnaires. Considering the effect on life quality,
the self-assessment of symptoms and quality in our study
method might be complementary to, if not more important
than, objective examination results presented in other studies.
Besides, our results furthered the evaluation of sleep quality
in multidimensions with CPSQI subscales. As the complicated
process of sleep is composed of different steps and patterns,
our presentation of the details of CPSQI could be helpful in as-
sisting future research designs.
Our study also has several limitations. The cross-sectional
design of the study limits the assessment of the direction of the
sleep quality and dry eye association and the establishment of
a temporal relationship. Since the patterns of dry eye and sleep
quality were evaluated based on face-to-face interview, recall,
and misclassification biases are possible but may not be differ-
ential as the dry eye is assessed by questionnaire and the study
participants have not been told to have dry eye by their doctors
previously. Although the OSDI questionnaire is a validated and
widely used instrument to evaluate dry eye, it may not identify
those with mild dry eye disease but without typical symptoms.

In addition, it may also underestimate the dry eye symptoms
of patients with corneal sensation disorder (which might be
caused by older age, neurological problems, diabetes, etc.) [41].
Those patients usually could be diagnosed by clinically exam-
inations. However, since our study has more than 3,000 parti-
cipants, it was not feasible to provide such examination to all
subjects at baseline. Besides, when we focus on the effect on life
quality, as we did in this study, among our local citizen popula-
tion the self-complaints of the disease itself might weigh more
than objective examination.
Dry eye and poor sleep quality are important public health
problems. In our community-based study in Hangzhou, we
found that these two conditions are quite common and poor
sleep quality is associated with dry eye. Prospective studies are
needed to identify the incidence of dry eye in individuals with
poor sleep quality and to clarify the potential biological mech-
anism into the causal relationship for proposals in prevention
strategies.
Supplementary material
Supplementary data are available at SLEEP online.
Funding
Initial funding for the Stanford Wellness Living Laboratory
(WELL) was provided by Amway via an unrestricted gift through
the Nutrilite Health Institute Wellness Fund. This work was
also supported by grants from the Cyrus Tang Foundation (no.
419600-11102), the Zhejiang University Education Foundation
(no. 100000-11320/028), and the Zhejiang Province Key labora-
tory Fund of China (no. 2011E10006).
Acknowledgments
We acknowledge the Community Health Service Centers, CDC
and the Health Bureau of Xihu District, Hangzhou, China.
Authors’ Contributions: K.Y., S.Z., and X.Y. designed the study;
X.Y. drafted the manuscript; H.G. and G.W. analyzed the data;
X.Y., H.G., X.L., G.W., M.Y., and X.Z. collected the data; X.Y., H.G.,
X.L., G.W., M.Y., S.H.S.C., S.H., R.C., X.Z., A.H., S.Z., and K.Y. pro-
vided comments and revised the article. S.Z. is the Principal
Investigator (PI) of the Wellness Living Laboratory China project
in China and A.H. is the WELL China PI in United States. All au-
thors have approved the final version of the article.
Conflict of interest statement. None declared.
References
1. Kerkhof GA. Epidemiology of sleep and sleep disorders in
The Netherlands. Sleep Med. 2017;30:229–239.
2. Altevogt BM, et al. Sleep disorders and sleep deprivation:
an unmet public health problem. J Am Acad Child Adolesc
Psychiatry. 2008;47:473–474.
3. Koren D, et al. Role of sleep quality in the metabolic syn-
drome. Diabetes Metab Syndr Obes. 2016;9:281–310.
4. Luca A, et al. Sleep disorders and depression: brief review of
the literature, case report, and nonpharmacologic interven-
tions for depression. Clin Interv Aging. 2013;8:1033–1039.
Yu et al. | 7
5. Smith JA, Albeitz J, Begley C, et al. The epidemiology of dry
eye disease: report of the Epidemiology Subcommittee
of the International Dry Eye WorkShop. Ocul Surf.
2007;5:93–107.
6. Schaumberg DA, et al. Epidemiology of dry eye syndrome.
Adv Exp Med Biol. 2002;506(Pt B):989–998.
7. Gipson IK, et al. Research in dry eye: report of the research
subcommittee of the International Dry Eye Workshop
(2007). Ocul Surf. 2007;5(2):179–193.
8. Kawashima M. Systemic health and dry eye. Invest
Ophthalmol Vis Sci. 2018;59(14):DES138–DES142.
9. Vehof J, et al. clinical characteristics of dry eye patients with
chronic pain syndromes. Am J Ophthalmol. 2016;162:59.e2–
65.e2.
10. Jones L, et al. TFOS DEWS II management and therapy re-
port. Ocul Surf. 2017;15(3):575–628.
Downloaded from https://academic.oup.com/sleep/article/42/11/zsz160/5532656 by guest on 19 April 2024
11. Şimşek C, et al. Current management and treatment of dry
eye disease. Turk J Ophthalmol. 2018;48(6):309–313.
12. Ayaki M, et al. Sleep disorders are a prevalent and ser-
ious comorbidity in dry eye. Invest Ophthalmol Vis Sci.
2018;59:DES143–DES150.
13. Kawashima M, et al.The association of sleep quality with dry eye
disease: the Osaka study. Clin Ophthalmol. 2016;10:1015–1021.
14. Ayaki M, et al. High prevalence of sleep and mood disorders
in dry eye patients: survey of 1,000 eye clinic visitors.
Neuropsychiatr Dis Treat. 2015;11:889–894.
15. Karaca EE, et al. Evaluation of ocular surface health in
patients with obstructive sleep apnea syndrome. Turk J
Ophthalmol. 2016;46(3):104–108.
16. Acar M, et al. Ocular surface assessment in patients with
obstructive sleep apnea–hypopnea syndrome. Sleep Breath.
2013;17(2):583–588.
17. Galor A, et al. The association of dry eye symptom severity
and comorbid insomnia in US veterans. Eye Contact Lens.
2018;44 (Suppl. 1):S118–S124.
18. Lee W, et al. The association between sleep duration
and dry eye syndrome among Korean adults. Sleep Med.
2015;16(11):1327–1331.
19. Ayaki M, et al. Sleep and mood disorders in women with dry
eye disease. Sci Rep. 2016;6:35276.
20. Ayaki M, et al. Preliminary report of improved sleep quality
in patients with dry eye disease after initiation of topical
therapy. Neuropsychiatr Dis Treat. 2016;12:329–337.
21. Ayaki M, et al. Sleep and mood disorders in dry eye disease
and allied irritating ocular diseases. Sci Rep. 2016;6:22480.
22. Topp CW, et al. The WHO-5 Well-Being Index: a sys-
tematic review of the literature. Psychother Psychosom.
2015;84(3):167–176.
23. Schiffman RM, et al. Reliability and validity of the
ocular surface disease index. Arch Ophthalmol.
2000;118(5):615–621.
24. Liu XC, et al. Reliability and validity of the Pittsburgh
sleep quality index. Chinese Journal of Psychiatry.
1996;1994(350):1–9.
25. Tsai PS, et al. Psychometric evaluation of the Chinese
version of the Pittsburgh Sleep Quality Index (CPSQI)
in primary insomnia and control subjects. Qual Life Res.
2005;14(8):1943–1952.
26. Liao Y, et al. Sleeping problems among Chinese heroin-
dependent individuals. Am J Drug Alcohol Abuse.
2011;37(3):179–183.
27. Zhang HS, et al. Poor sleep quality is significantly associ-
ated with low sexual satisfaction in Chinese methadone-
maintained patients. Medicine (Baltimore). 2017;96(39):e8214.
 8 | SLEEPJ, 2019, Vol. 42, No. 11
28. Lou P, et al. Association of sleep quality and quality of life in
type 2 diabetes mellitus: a cross-sectional study in China.
Diabetes Res Clin Pract. 2015;107(1):69–76.
29. Wang SY, et al. Factors related to fatigue in Chinese pa-
tients with end-stage renal disease receiving maintenance
hemodialysis: a multi-center cross-sectional study. Ren Fail.
2016;38(3):442–450.
30. Cohen JE. Statistical Power Analysis for the Behavioral Sciences.
Hillsdale, NJ: Lawrence Erlbaum Associates, Inc; 1988: 283–286.
31. Lee YB, et al. Sleep deprivation reduces tear secretion and im-
pairs the tear film. Invest Ophthalmol Vis Sci. 2014;55(6): 3525–3531.
32. Fox RI. Sjögren’s syndrome. Lancet. 2005;366(9482):321–331.
33. Hackett KL, et al. An investigation into the prevalence of
sleep disturbances in primary Sjögren’s syndrome: a sys-
tematic review of the literature. Rheumatology (Oxford).
2017;56(4):570–580.
34. Chiaro G, et al. REM sleep behavior disorder, autonomic dys-
function and synuclein-related neurodegeneration: where
do we stand? Clin Auton Res. 2018;28(6):519–533.
35. Palma JA. Autonomic dysfunction in sleep disorders: intro-
duction to the series. Clin Auton Res. 2018;28(6):507–508.
36. Dartt DA. Regulation of tear secretion. Adv Exp Med Biol.
1994;350:1–9.
37. Dartt DA. Neural regulation of lacrimal gland secretory
processes: relevance in dry eye diseases. Prog Retin Eye Res.
2009;28(3):155–177.
38. Stern ME, et al. Conjunctival T-cell subpopulations in
Sjögren’s and non-Sjögren’s patients with dry eye. Invest
Ophthalmol Vis Sci. 2002;43(8):2609–2614.
39. Okun ML, et al. A review of sleep-promoting medi-
cations used in pregnancy. Am J Obstet Gynecol.
2015;212(4):428–441.
40. Koçer E, et al. Dry eye related to commonly used new anti-
depressants. J Clin Psychopharmacol. 2015;35(4):411–413.
41. Davidson EP, et al. Impaired corneal sensation and nerve
Downloaded from https://academic.oup.com/sleep/article/42/11/zsz160/5532656 by guest on 19 April 2024
loss in a type 2 rat model of chronic diabetes is reversible
with combination therapy of menhaden oil, α-lipoic acid,
and enalapril. Cornea. 2017;36(6):725–731.