Association between dry eye disease and poor sleep quality in the elderly

Abstract

Purpose

The elderly are at higher risk for sleep disorders and dry eye. This study explored correlations between dry eye and sleep quality in elderly individuals.

Methods

58 elderly dry eye patients and 31 controls were enrolled. Sleep quality was evaluated using Pittsburgh Sleep Quality Index (PSQI) and wearable intelligent devices. Dry eye symptoms and signs were assessed via Ocular Surface Disease Index (OSDI), Fluorescein tear breakup time (FBUT), Corneal fluorescein staining score (CFS) and Schirmer I test.

Results

The dry eye group showed significantly worse OSDI, FBUT, CFS, Schirmer I, and PSQI scores versus controls (all P < 0.01), with reduced deep sleep duration (P = 0.02) and deep sleep percentage (P < 0.01). Among dry eye patients, 69.0% had poor sleep, and 94.8% slept less than 8 h (mean: 6.51 ± 0.83 h). Subjective sleep quality correlated with FBUT and CFS (P < 0.05). A correlation was found between deep sleep percentage (%) and FBUT (r = 0.296, P < 0.05) and CFS (r=-0.418, p < 0.01). Prolonged video terminal use reduced deep sleep percentage (%) (F = 5.18, P < 0.01) and worsened FBUT (P < 0.01).

Conclusions

Elderly dry eye patients exhibit poorer sleep quality. Reduced deep sleep percentage (%) in elderly correlates with dry eye, and excessive video terminal use may contribute.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12886-026-05075-2.

Introduction

Dry eye disease (DED) is a common and multifactorial disorder of the ocular surface in which disruption of tear film homeostasis, together with ocular surface inflammation, epithelial damage, and neurosensory abnormalities, contributes to both symptoms and clinical signs [1]. DED can cause ocular discomfort, visual fluctuation, photophobia, burning, and pain [2], thereby reducing quality of life and increasing the social and economic burden associated with ocular surface disease [3]. Older age, inflammatory status, environmental exposure, and unfavorable visual habits have all been associated with an increased likelihood of DED [4]. Sleep disturbance has also been increasingly recognized as a potential contributor to dry eye [5–7].

Sleep is a fundamental physiological process that occupies approximately one third of human life. Sleep is broadly divided into rapid eye movement (REM) sleep and non-rapid eye movement (NREM) sleep [8]. NREM sleep includes N1, N2, and N3 stages, of which N1 and N2 represent lighter sleep and N3 represents deep sleep [9]. Aging is accompanied by well-recognized changes in sleep architecture, including reduced sleep duration and a lower proportion of deep sleep [10–12]. Sleep problems are highly prevalent in the general population and are particularly important in older adults because of their substantial impact on daily functioning and quality of life [13–14].

An increasing number of studies have suggested that sleep disturbance is associated with both the occurrence and severity of dry eye [15]. In a large community-based study from China, poorer sleep quality and multiple sleep-related complaints were associated with more severe dry eye manifestations [6]. Several aspects of sleep, including subjective sleep quality, sleep latency, sleep duration, and sleep efficiency, have been reported to be associated with dry eye severity. A study in Japan has also shown that sleep deprivation and poor sleep quality were significantly correlated with DED [16]. It has been found that sleep quality was associated with severe meibomian gland disruption in dry eye [17]. Because older adults are vulnerable to both sleep problems and ocular surface dysfunction, this population may be particularly susceptible to the combined effects of these conditions. However, the relationship between objective sleep structure and dry eye in elderly individuals remains insufficiently characterized. Therefore, this study evaluated sleep quality, sleep architecture, and ocular surface characteristics in elderly patients with DED to further explore the relationship between dry eye and sleep.

Patients and methods

Survey procedure and data collection

This study is a cross-sectional study of elderly patients with dry eye. The overall study workflow is shown in Fig. 1. Data collection started in March 2022 and ended in September 2023. Eligible participants were aged 50 years or older [15], had dry eye, had no other ocular disease, and voluntarily agreed to participate in the study. The diagnosis of DED was based on the TFOS DEWS II definition and classification report [1]. In the present study, participants were classified as having DED when they had relevant dry eye symptoms, defined as an Ocular Surface Disease Index (OSDI) score of ≥ 13, together with at least one objective sign indicating loss of tear film homeostasis, including a fluorescein tear breakup time (FBUT) of < 10 s and/or positive corneal fluorescein staining. Schirmer I test was additionally performed to assess aqueous tear production as a complementary ocular surface parameter. Control participants were required to have no clinical diagnosis of DED and no abnormal findings suggestive of tear film instability or ocular surface damage on the study examinations. Exclusion criteria included (1) any active external ocular processes, (2) history of ocular surgery, (3) history of ocular trauma, (4) other eye diseases, (5) long term or frequent contact lens use, and (6) cognitive dysfunction. 58 elderly patients who were diagnosed with dry eye (dry eye group) and 31 normal participants (control group) were recruited. The study was approved by the Ethical Committee of the Beijing Tongren Hospital and was conducted in accordance with the tenets of the Declaration of Helsinki. Informed consent was obtained from all the participants. Participants were stratified into three groups according to daily video terminal use time (> 6 h, 3–6 h, and < 3 h). Baseline demographic and clinical data collected for each participant included age, sex, daily video terminal use time, subjective sleep quality, objective sleep parameters, and ocular surface examination results.

Fig. 1.

Study workflow and assessment framework. The diagram illustrates participant enrollment and grouping, subjective sleep quality assessment using the Pittsburgh Sleep Quality Index (PSQI), objective sleep structure monitoring using the Xiaomi Mi Band 5 wristband, Bluetooth-based transfer of sleep data to the mobile application, and ocular surface examinations including OSDI, FBUT, corneal fluorescein staining, and Schirmer I test. The sleep quality domain included subjective sleep quality, sleep latency, sleep duration, sleep efficiency, sleep disturbance, sleep medication use, daytime dysfunction, and total PSQI score. The sleep structure domain included total sleep time, light sleep, deep sleep, and rapid eye movement sleep, expressed as both duration and proportion where applicable

Dry eye symptoms

Dry eye symptoms were assessed using the Ocular Surface Disease Index (OSDI), a validated 12-item questionnaire that evaluates symptom severity and its effect on visual function over the previous week. It provides a rapid assessment of eye discomfort symptoms consistent with dry eye and their effects on visual function. The OSDI consists of 12 items, each scored from 0 to 4. The total score of the OSDI is calculated as the sum of scores of all questions answered multiplied by 25 and then divided by the total number of questions answered [18]. OSDI < 13 is considered normal, and OSDI ≥ 13 is considered abnormal.

Fluorescein tear breakup time (FBUT)

1–2% fluorescein sodium dye was placed into the conjunctival sac of patient’s lower eyelid. Participants were then asked to blink several times to ensure even distribution of fluorescein over the tear film and corneal surface. The tear film was examined under cobalt-blue illumination using a slit-lamp microscope. FBUT was defined as the interval between the last complete blink and the first appearance of a dark spot in the stained tear film. Each eye was measured three times with a stopwatch, and the mean value was calculated. For statistical analysis, the right-eye FBUT value was used consistently for all participants to avoid inter-eye correlation [19]. A schematic illustration of the FBUT procedure is provided in Supplementary Figure S1.

Corneal fluorescein staining score (CFS)

A sterile fluorescein strip moistened with normal saline was gently applied to the lower conjunctival sac. After several blinks, the cornea was examined under cobalt-blue illumination with a slit-lamp microscope. The Punctate epithelial erosions (PEEs) were counted and scored according to the Oxford scoring standard revised by Sjögren’s International Collaborative Clinical Alliance (SICCA) as follows: 0 = absent, 1 = 1 to 5 PEEs, 2 = 6 to 30 PEEs, and 3 = more than 30 PEEs. An additional point is added when PEEs occur in the central 4 mm diameter of the cornea, when any filaments are seen on the cornea or when any patches of confluent staining including linear stains are found anywhere on the cornea. The maximum total score is 6 [20].

Schirmer I test

Standardized Schirmer I test strips were used to assess aqueous tear production. Without topical anesthesia, the strips were placed simultaneously at the junction of the middle and lateral thirds of the inferior palpebral conjunctiva in both eyes and left in place for 5 min with the eyes gently closed. The length of strip wetting was recorded in mm/5 min. A value of < 5 mm/5 min was considered abnormal, whereas values between 5 and 10 mm/5 min were considered borderline and values ≥ 10 mm/5 min were considered normal for clinical interpretation. For statistical analysis, Schirmer I test results were analyzed as continuous variables, and the right-eye value was used consistently across participants [21].

Sleep quality

The Pittsburgh sleep quality questionnaire (PSQI) is the most commonly used effective questionnaire for evaluation of sleep quality in clinical and research settings [22]. It examines 7 components of sleep quality over the past month including subjective sleep quality, sleep latency, sleep duration, sleep efficiency, sleep disturbance, sleep medication, and daytime dysfunction. The score of each of the seven components is 0–3. The sum generates the total score, ranging from 0 to 21. Higher scores represent worse sleep quality. Participants were categorized as having good sleep quality (PSQI 0–5) or poor sleep quality (PSQI > 5) [23].

Sleep structure detection

A Xiaomi series of wearable wristband devices have been widely used in the study of sleep time and sleep structure [24–26]. Xiaomi Mi band 5 was used to detect patients’ sleep time and sleep structure in this study. This newly developed wearable device is suitable for older adults [27, 28]. The device can determine the stage of sleep through a heart rate (HR) sensor, which measures HR and detects movements through an accelerometer. It identifies user’s sleep state as light sleep (N1 + N2 sleep stages), deep sleep (N3 sleep stages), and rapid eye movement sleep (REM sleep stages) [28]. Sleep data were transferred from the wristband to a smartphone application via Bluetooth. The mean values of sleep parameters of the Participants during 14 consecutive days were used as the final outcomes.

Statistical analysis

For statistical analyses, OSDI, FBUT, CFS, and Schirmer I test results were treated as continuous variables. Ocular surface examination data were entered using the right-eye value for each participant and were handled consistently across all analyses. Continuous-variable analysis was chosen to preserve the quantitative information contained in the measurements, whereas clinical cutoffs were used only for descriptive interpretation. SPSS version 25.0 (SPSS Inc., Chicago, IL, USA) was used for statistical analysis. Continuous variables are presented as mean ± standard deviation (SD), and categorical variables are presented as number (percentage). The characteristics of the study population were summarized using descriptive statistics. The intra-class correlation coefficient (ICC) was used to assess the agreement between subjective sleep duration in the PSQI questionnaire and objective sleep duration monitored by wearable devices. ICC ≥ 0.75 was considered to indicate good reliability. Normality of continuous variables was assessed using the Shapiro-Wilk test and visual inspection of Q-Q plots, and linearity was evaluated using scatterplots. Homogeneity of variance was assessed using Levene’s test before parametric group comparisons. Pearson’s correlation coefficient was applied only to approximately normally distributed continuous variables with linear relationships and without extreme outliers. Correlations between ordinal PSQI component scores and continuous ocular surface parameters were assessed using Kendall’s tau-b correlation coefficient. For Table 1, continuous variables were compared between the dry eye and control groups using the independent-samples t test when the assumptions for parametric analysis were satisfied; otherwise, non-parametric methods were considered as appropriate. Categorical variables were compared using the chi-square test or Fisher’s exact test, as appropriate. For Table 2, associations between PSQI component scores and ocular surface parameters were assessed using Kendall’s tau-b correlation coefficient. For Table 3, associations between continuous objective sleep parameters and continuous ocular surface parameters were assessed using Pearson’s correlation coefficient. For Table 4, comparisons across the three video terminal use groups were performed using one-way ANOVA when the assumptions for parametric analysis were satisfied, followed by LSD post hoc testing when homogeneity of variance was met and Tamhane’s T2 testing when it was not. A two-sided P < 0.05 was considered statistically significant.

Table 1.

Demographic, sleep-related, ocular surface, and video terminal use characteristics of the study population

	Dry eye group (n = 58) N (%), mean (SD)	Control group (n = 31) N (%), mean (SD)	p
Sociodemographic data
Women	33(56.9%)	16(51.6%)
Men	25(43.1%)	15(48.4%)
Age (years)	59.17 ± 5.40	59.29 ± 5.63	0.923
PSQI score, mean ± SD	7.52 ± 3.07	3.65 ± 2.33	< 0.01**
Poor sleep quality (PSQI > 5)	40(69.0%)	7(22.6%)	—
Sleep parameters
Total sleep time, hours	6.51 ± 0.83	6.61 ± 0.40	0.54
<8 h	55(94.8%)	31(100%)	—
Light sleep
hours	4.22 ± 0.65	4.22 ± 0.41	0.95
%	64.8 ± 5.2%	63.95 ± 6.2%	0.50
Deep sleep
hours	1.28 ± 0.22	1.48 ± 0.38	0.02*
%	19.7 ± 2.7%	22.25 ± 5.1%	< 0.01**
Rapid eye movement sleep
hours	1.00 ± 0.32	0.91 ± 0.15	0.13
%	15.5 ± 4.8%	13.80 ± 2.1%	0.03*
Ophthalmic examination parameters
OSDI	37.5 ± 18.8	6.02 ± 4.68	< 0.01**
FBUT	3.52 ± 0.68	7.10 ± 2.3	< 0.01**
CFS	3.03 ± 1.14	0 ± 0	< 0.01**
Schirmer I test	5.52 ± 2.12	11.21 ± 8.26	< 0.01**
Use time of video terminal
<3 h	9(15.5%)	7(22.6%)	—
3–6 h	37(63.8%)	15(48.4%)	—
>6 h	12(20.7%)	9(29.0%)	—

Continuous variables were compared using the independent-samples t test; categorical variables were compared using the chi-square test or Fisher’s exact test, as appropriate

Table 2.

Associations between PSQI component scores and ocular surface parameters

PSQI component	Ophthalmic examination parameters
	OSDI	FBUT	CFS	Schirmer I test
Subjective sleep quality	0.220	-0.356**	0.421**	-0.183
Sleep latency	-0.071	0.058	0.093	-0.127
Sleep duration	0.099	0.163	0.073	0.170
Sleep efficiency	-0.288*	-0.002	0.061	0.219
Sleep disturbance	-0.087	-0.374**	0.328*	-0.215
Sleep medication	0.254	-0.093	0.078	-0.136
Daytime dysfunction	0.048	-0.572**	0.318*	-0.079
Total score	0.011	-0.353**	0.381**	-0.063

Values are Kendall’s tau-b correlation coefficients

* P < 0.05

** P < 0.01

Table 3.

Correlations between objective sleep parameters and ocular surface parameters (N = 58)

Sleep parameters	Ophthalmic examination parameters					Sleep quality
	OSDI	FBUT	CFS	Schirmer I test		PSQI
Total sleep time, hours	-0.040	0.015	0.099	0.018		-0.119
Light sleep
hours	0.028	0.025	0.151	-0.066		0.02
%	0.091	0.033	0.147	-0.146		0.222
Deep sleep
hours	-0.218	0.237	-0.225	0.169		-0.438**
%	-0.191	0.296*	-0.418**	0.208		-0.470**
REM sleep
hours	-0.01	-0.175	0.104	0.063		-0.045
%	0.008	-0.232	0.074	0.057		0.025

Values are Pearson correlation coefficients

* P < 0.05

** P < 0.01

Table 4.

Analysis of sleep and ocular surface parameters according to daily video terminal use

Time of video terminal use	<3 h		3–6 h		>6 h		F	P
	Mean	SD	Mean	SD	Mean	SD
PSQI score, mean ± SD	6.22	2.82	7.65	3.08	8.08	3.20	1.04	0.36
Total sleep time, hours	6.13	1.05	6.54	0.73	6.72	0.94	1.36	0.27
Light sleep
hours	3.88	0.82	4.21	0.53	4.55	0.77	2.90	0.06
%	62.67	4.18	64.38	4.73	67.67	7.14	2.67	0.08
Deep sleep
hours	1.30	0.25	1.30	0.20	1.19	0.27	1.18	0.32
%	21.22	0.97	19.92	2.36	17.83	3.49	5.18	<0.01
REM sleep
hours	0.95	0.20	1.03	0.34	0.99	0.37	0.24	0.79
%	15.67	4.39	15.68	4.63	14.67	5.85	0.20	0.82
OSDI	31.97	17.50	38.77	18.73	37.73	20.45	0.47	0.63
FBUT	4.00	0.61	3.55	0.64	3.09	0.59	5.42	<0.01
CFS	2.56	1.13	3.00	1.18	3.50	0.90	1.87	0.16
Schirmer I test	5.78	2.44	5.76	2.14	4.58	1.68	1.49	0.23

Comparisons were performed using one-way ANOVA; post hoc analyses used LSD or Tamhane’s T2 test as appropriate

Results

Cohort characteristics

The demographic, sleep-related, ocular surface, and video terminal use characteristics of the study population are summarized in Table 1. Compared with the control group, the dry eye group showed significantly worse OSDI, FBUT, CFS, Schirmer I test, and PSQI scores (all P < 0.01), together with a significantly shorter duration and lower proportion of deep sleep. Within the dry eye group, 40/58 participants (69.0%) had poor sleep quality (PSQI > 5), 55/58 (94.8%) slept less than 8 h per day, and most participants reported a daily mean video terminal use time of 3–6 h. In addition, 12/58 participants (20.7%) in the dry eye group reported daily video terminal use greater than 6 h.

Relationship between dry eye and subjective sleep quality

The correlation analysis between the components of the PSQI questionnaire and OSDI, FBUT, CFS, Schirmer I test was summarized in Table 2. A higher score (0–3) in each component of the PSQI questionnaire indicated a worse sleep quality of the patient on the respective aspect. It was found that sleep efficiency was significantly correlated with OSDI. Subjective sleep quality, sleep disturbance, daytime dysfunction, and total score of PSQI were significantly correlated with FBUT and CFS (Table 2).

Correlation analysis between sleep architecture and ophthalmic parameters

The correlation analysis between the objective sleep parameters measured by wearable devices and dry eye parameters was summarized in Table 3. The duration of deep sleep of patients was significantly correlated with PSQI scores. Deep sleep occupancy (%) was significantly correlated with FBUT, CFS, and PSQI (Fig. 2). The remaining sleep parameters were not significantly associated with dry eye (Table 3)

Fig. 2.

Correlation analysis between sleep parameters and ophthalmic parameters. (A) Deep sleep duration and PSQI scores were negatively correlated. (B) Deep sleep occupancy and FBUT were positively correlated. (C) Deep sleep occupancy and CFS were negatively correlated. (D) Deep sleep occupancy and PSQI scores were negatively correlated

Analysis of dry eye and sleep under the influence of video terminal use

Patients were divided into three groups (< 3 h, 3–6 h, > 6 h) according to the duration of video terminal use. As shown in Table 4, deep sleep occupancy decreased significantly with increasing duration of video terminal use (F = 5.18, P < 0.01) (Fig. 3). The quality of sleep tended to decrease with extended time of video terminal use, although the difference was not statistically significant. FBUT decreased significantly across the three groups (F = 5.42, P < 0.01) (Fig. 3) (Table 4).

Fig. 3.

Comparison of deep sleep occupancy (%) and FBUT under different durations of video terminal use. (A) Deep sleep occupancy decreased as the duration of video terminal use increased. (B) FBUT decreased as the duration of video terminal use increased

Discussion

In this study, we evaluated both subjective sleep quality and objective sleep structure in elderly individuals with DED and compared them with normal controls. The dry eye group showed poorer sleep quality, shorter deep sleep duration, and a lower proportion of deep sleep. We further found that worse ocular surface findings were associated with less favorable sleep-related parameters. In addition, longer video terminal use was associated with reduced deep sleep proportion and poorer FBUT. These findings suggest that sleep disturbance, especially reduced deep sleep, may be clinically relevant to ocular surface instability in older adults. Wearable monitoring has become increasingly relevant in clinical research because it allows repeated assessment under real-life conditions [29]. Wearable smart devices increase the likelihood of improving patient monitoring situations. Polysomnography (PSG) is the gold standard for objective assessment of sleep characteristics. However, it has limitations under certain circumstances. For example, patients cannot be tested by PSG in a home environment, and PSG can only assess patients’ sleep for one night [30]. In contrast, wearable devices permit continuous monitoring across multiple nights and may better reflect habitual sleep patterns in daily life [31]. For the present study, this approach provided a practical way to characterize sleep structure in elderly participants outside the laboratory environment. A large population-based study reported that shorter sleep duration was associated with a higher prevalence of dry eye symptoms [16]. However, that study included a broader age range and younger participants than those in our elderly cohort. Therefore, it should be regarded as supportive background evidence for the general association between insufficient sleep and dry eye, rather than as a directly comparable study. In contrast, our study specifically focused on older adults and additionally incorporated objective sleep structure measurements. Age is an important factor affecting sleep duration, which tends to decrease with advancing age [32]. Reduced sleep duration may therefore be an important contributor to dry eye in older adults.

Previous work has suggested that shorter sleep duration may be associated with certain dry eye subtypes, including aqueous-deficient dry eye [33]. However, our study did not classify dry eye into subtypes, and Schirmer I test values were analyzed as continuous variables rather than as subtype-defining thresholds. Therefore, that study is not directly comparable to our results and is discussed here only as background evidence suggesting that sleep disturbance may influence some aspects of tear dysfunction. The elderly have a reduced proportion of deep sleep and are more easily aroused from sleep by auditory stimuli, causing relative sleep deprivation [11]. Sleep deprivation causes a reduction in the duration of sleep and can also lead to tear hyperosmolarity, shortened TBUT, and decreased tear secretion, aggravating signs and symptoms of dry eye [34]. Sleep deprivation can cause imbalance of circulating leptin and ghrelin levels [35], thereby increasing intake of appetite and calorie and reducing energy expenditure. These can further lead to insulin resistance, poor glycemic control, and obesity [36]. Poor glycemic control is associated with reduced corneal sensitivity, shorter BUT, decreased Schirmer test, epithelial metaplasia, squamous metaplasia, goblet cell loss, and changes in tear proteins [37, 38]. Hanyuda A [39]. have found that both higher frequency of having difficulty falling asleep and waking tiredness were significantly associated with increased prevalence of DED (P < 0.001). Our studies also found that a decrease in sleep quality correlated with a worsening of dry eye. Older adults typically show a lower proportion of deep sleep [40]. Our studies revealed a correlation between deep sleep occupancy (%) and both FBUT(r = 0.296, p < 0.05) and CFS (r=-0.418, p < 0.01). This suggests that the reduction of deep sleep occupancy (%) will aggravate the signs of dry eye.

In the present study, Schirmer I test values differed significantly between the dry eye and control groups, but were not significantly associated with PSQI component scores or objective sleep architecture parameters. This pattern may suggest that, in this elderly cohort, sleep disturbance was more closely related to tear film instability and ocular surface epithelial damage, reflected by FBUT and CFS, than to aqueous tear volume alone. In addition, the Schirmer I test is known to show considerable variability and may be influenced by reflex tearing and patient discomfort, which may reduce its sensitivity in correlation analyses. Therefore, the absence of a significant association with Schirmer I values should not be interpreted as evidence of no relationship between sleep disturbance and dry eye, but rather as an indication that the association may be stronger for tear film stability and ocular surface integrity than for tear secretion quantity alone.

The time of using video terminals among people has been steadily increasing with increasing popularity of smartphones and computers in recent years. Exposure to blue light has now become more frequent than ever before. Ishizawa M et al. [41] have found that the proportion of deep sleep was significantly reduced in the blue light exposure group (P < 0.01), suggesting that blue light exposure affects sleep quality by decreasing the proportion of deep sleep. The use of video terminals affects the ocular surface microenvironment, inducing dry eye. Several studies have shown that prolonged video terminal use is associated with increased tear osmolarity, reduced tear meniscus height, and shortened tear breakup time, thereby contributing to or aggravating dry eye [42–44]. By analyzing 64 video terminal workers, Fenga C et al. [45] have found that tear osmolarity was directly correlated with corneal staining and ocular surface dysfunction, and negatively correlated with TBUT while no correlation was found when compared to OSDI. In our study, increasing daily video terminal use was associated with a lower proportion of deep sleep and a shorter FBUT. These findings are consistent with the possibility that extended screen exposure contributes to both impaired sleep quality and tear film instability. One potential explanation is that prolonged exposure to blue light and sustained visual tasking may influence sleep regulation while also promoting ocular surface stress through reduced blinking and tear film disruption.

In this study, we examined the sleep structure of patients by wearable electronic devices and simultaneously evaluated the sleep quality and structure of elderly patients with dry eye. We found that decreased deep sleep occupancy in elderly dry eye patients was correlated with increased severity of dry eye. Our studies suggest that increased duration of video terminal use may be an important risk factor linking dry eye and sleep disturbance. The current study had several limitations. First, PSG was not applied in this study, and the use of wearable devices instead of gold-standard polysomnography may have affected the precision of sleep architecture assessment. Second, the sample size was relatively small, which may limit the reliability and generalizability of the findings. Third, the characterization of the included participants was not comprehensive, and some potentially relevant demographic and clinical factors were not fully evaluated. Therefore, larger studies with more comprehensive participant characterization and multimodal sleep assessment are warranted to validate and extend the present findings.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Ao Li and Yubo Wu had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Ao Li and Yubo Wu have contributed equally to this work and share first authorship. Concept and design: Ao Li, Yubo Wu, Lei Tian, Ying Jie. Acquisition, analysis, or interpretation of data: Ao Li, Yubo Wu, Ziyang Yin. Drafting of the manuscript: Ao Li, Yubo Wu, Rongteng Lin. Critical revision of the manuscript for important intellectual content: All authors. Statistical analysis: Ao Li, Yubo Wu, Ying Jie. Administrative, technical, or material support: Lei Tian, Ying Jie. Supervision: Lei Tian, Ying Jie.

Funding

This research is supported by “Yicheng Outstanding Talents” in Beijing Economic-Technological Development Area; Talent Development Plan for High-level Public Health Technical Personnel Project, academic leaders, 02–22; Beijing Municipal Public Welfare Development and Reform Pilot Project for Medical Research Institutes (PWD&RPP-MRI, JYY2023-6); Capital Health Research and Development of Special (Grant No. 2022-1-1081 to Dr Jie); Beijing Hospitals Authority’s Ascent Plan(DFL20240202); The Youth Beijing Scholars program (NO.022).

Data availability

The authors declare that all data supporting he findings of this study are available within the paper. They are not publicly available, and restrictions apply to their use. All requests would require evaluation on an individual basis and can be made by contacting jie_yingcn@aliyun. com.

Declarations

Ethics approval and consent to participate

The Ethical Committee of the Beijing Tongren Hospital approved the study procedures. The study was conducted in accordance with the tenets of the Declaration of Helsinki. Informed consent was obtained from all the participants.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.