Correlation of ocular surface function with sleep quality, anxiety, and depression in patients with dry eye disease

Abstract

BACKGROUND

Dry eye disease (DED) is a multifactorial ocular surface disorder with rising prevalence. It is closely related to systemic health and psychological factors, such as sleep and mood disorders, which significantly impact the quality of life of patients.

AIM

To explore the correlations between ocular surface function, sleep quality, and anxiety/depression in patients with DED.

METHODS

This was a cross-sectional investigative study that included 358 patients with DED between January 2022 and January 2025. Ocular surface was assessed using the ocular surface disease index (OSDI), tear film break-up time, fluorescein staining score, and Schirmer I test. The Pittsburgh Sleep Quality Index (PSQI), Self-Rating Anxiety Scale (SAS), and Self-Rating Depression Scale (SDS) were used to evaluate sleep quality and anxiety/depression levels. Correlation and linear regression analyses were used to explore the relationships.

RESULTS

The mean PSQI score of the patients was 9.94 ± 2.18; the mean SAS score was 47.30 ± 4.90, and the mean SDS score was 50.08 ± 5.52. These suggested a prevalence of sleep and psychological abnormalities. There was a significant correlation between the indicators of ocular surface function (OSDI, tear film break-up time, fluorescein staining, and Schirmer I test) and PSQI, SAS, and SDS scores (P < 0.05). Moreover, multiple regression revealed that age ≥ 50 years (β = 1.55, P = 0.029), PSQI scores (β = 0.58, P < 0.001), SAS scores (β = 0.17, P = 0.017), and SDS scores (β = 0.15, P = 0.019) were independent predictors of the OSDI scores.

CONCLUSION

Ocular surface function in patients with DED is closely related to sleep quality and anxiety/depression, emphasizing the need for holistic clinical management.

Key Words: Dry eye disease; Ocular surface function; Sleep quality; Anxiety; Depression

Core Tip: This study reveals significant correlations between ocular surface dysfunction and poor sleep quality and anxiety/depression in patients with dry eye disease. Sleep and mood disorders were prevalent among these patients, and factors such as age, sleep quality, and psychological scores independently predicted symptom severity. These findings underscore the importance of integrated, multidisciplinary management for dry eye disease, addressing both ocular and systemic health.

INTRODUCTION

Dry eye disease (DED) is a multifactorial-mediated ocular surface disease whose pathological mechanism involves abnormal tear secretion, decreased tear film stability, and ocular surface inflammatory response[1]. Recently, with the extensive use of electronic products and an increase in environmental pollution, the incidence of DED has shown a significant upward trend[2,3]. Epidemiological surveys have shown that the global prevalence of DED is approximately 5%-50%, with significant variations across regions and populations[4]. This range not only reflects the high prevalence of the disease but also highlights the serious threat it poses to public ocular health[5]. Notably, numerous studies have shown that DED is not only a localized ocular surface disease but is also closely related to systemic health conditions and psychological factors, especially the co-morbidities of sleep disorders and mood disorders (e.g., anxiety, depression), which are of increasing concern[6,7].

As an important interface between the human body and the external environment, the maintenance of the functional state of the ocular surface relies on the synergistic effects of the lacrimal glands, eyelids, neuromodulation, and systemic factors[8]. In recent years, an increasing number of studies have revealed that ocular surface dysfunction does not exist in isolation but is closely associated with the states of multiple systems throughout the body[9]. As an important physiological process in the human body, the quality of sleep directly affects the body’s neuroendocrine regulation, immune function, and metabolic balance[10]. Clinical observations have shown that patients with DED often experience reduced sleep quality, including difficulty in falling asleep, frequent nighttime awakenings, and shortened sleep duration[11]. Sleep disturbances may also further exacerbate the signs and symptoms of DED by affecting the rhythmicity of tear secretion, the stability of tear film components, and the repair process of ocular surface tissues[12]. This mechanism of bidirectional influence between ocular surface function and sleep quality provides new perspectives for a deeper understanding of the development of DED.

Anxiety and depression, as common psycho-emotional disorders, are significantly more prevalent in patients with DED than in the general population[13]. It is important to note that anxiety and depression are not only concomitant symptoms of DED but may also contribute to lacrimal gland dysfunction and activation of ocular surface inflammatory responses by affecting autonomic function. This creates a vicious cycle of ‘ocular surface dysfunction - psycho-emotional abnormality - aggravation of ocular surface symptoms’[14].

However, there is a lack of systematic exploration regarding the association of ocular surface functional status with sleep quality and anxiety/depression levels in patients with DED. In view of this, this study aimed to systematically assess ocular surface function, sleep quality, and anxiety/depression levels in patients with DED through a cross-sectional survey, and to explore the correlations among them. The conduct of this study has important clinical significance and scientific value. From a clinical practice perspective, clarifying the correlations among ocular surface function, sleep quality, and psychological state can help clinicians move beyond focusing solely on local ocular symptoms to also include the regulation of sleep quality and psychological interventions in the diagnosis and treatment of DED, thereby improving overall treatment outcomes and patients’ quality of life. From a basic research perspective, this study will provide a direction for exploring the neuromodulator mechanisms and molecular pathways underlying the ocular surface-sleep-psychological relationship, thereby advancing understanding of DED pathogenesis.

MATERIALS AND METHODS

Research participants

This was a cross-sectional investigative study involving patients with DED who were treated at Longyan First Affiliated Hospital of Fujian Medical University between January 2022 and January 2025. A total of 390 questionnaires were distributed in this study, and 358 were validly collected with a recovery rate of 91.79%. The inclusion criteria for the patients were as follows: (1) Patients who met the diagnostic criteria for DED in the “China Dry Eye Expert Consensus: Definition and Classification (2020)”[15]; (2) Age ≥ 18 years, with no sex restrictions; (3) Voluntary participation in the study with signed informed consent; and (4) Complete clinical information. The exclusion criteria were as follows: (1) Presence of other serious eye diseases (such as keratitis, acute glaucoma attack, or uveitis); (2) Eye surgery (e.g., cataract, refractive surgery) or eye trauma within the last 1 month; (3) Long-term use of medications that may affect tear production (e.g., antihistamines, antidepressants, hormonal drugs); (4) Presence of cognitive impairment or history of mental illness; and (5) Pregnant or lactating women. All patients in this study provided informed consent. The study was approved by the Medical Ethics Committee of Longyan First Affiliated Hospital of Fujian Medical University (Approval No. LTREC2022-K016-01). The research process is illustrated in Figure 1.

Figure 1  Research flowchart.

Data collection

General information collection: The general information of patients was collected through the electronic medical record system and self-made questionnaires, including age, sex, education level, chronic diseases (hypertension, diabetes), and duration of illness.

Clinical data collection: A standardized Chinese version of the ocular surface disease index (OSDI) scale, comprising 12 items, was used to assess the severity of subjective symptoms across three dimensions: Visual function, ocular symptoms, and limitation of daily activities. The total score ranged from 0 to 100, with higher scores indicating more severe dry eye symptoms[16,17]. The tear film break-up time (TBUT) was measured using a slit lamp microscope with cobalt blue light. The patients were asked to blink naturally and keep their eyes open, while the time from the last blink to the appearance of the first dry spot on the corneal surface was recorded. This measurement was repeated thrice, and the average value was taken[18]. Regarding the fluorescein staining (FL) score, 1% fluorescein sodium solution was administered as eye drops. Corneal epithelial defects were observed under cobalt blue light and scored on a 4-point scale from 0 to 12 (0-3 points per quadrant), with higher total scores indicating more severe corneal damage[19]. The Schirmer I test (SIT) was conducted without surface anesthesia by placing a Schirmer test strip on the conjunctiva of the lower eyelid to stimulate tear production. The length of the strip that was wetted after 5 minutes was measured[20].

Assessment of sleep quality and psychological state: The Pittsburgh Sleep Quality Index (PSQI) was assessed using 19 self-reported items (observer-rated items were not included in this study) covering seven dimensions: Subjective sleep quality, sleep duration, sleep latency, sleep efficiency, sleep disturbances, hypnotic medication use, and daytime dysfunction. The total score ranged from 0 to 21, with a score of > 7 indicating sleep quality problems (Kappa = 0.89)[21]. The standardized Self-Rating Anxiety Scale (SAS) and Self-Rating Depression Scale (SDS), developed by Zung[22,23], were used to assess anxiety and depression levels, respectively. Each scale comprised 20 items rated on a 4-point scale (1 = rarely or not at all, 4 = most or all of the time). The scores were converted using a specific formula, with a standard score of ≥ 50 on the SAS indicating anxiety. The status of depression was determined by the depression severity index: The depression severity index = SDS score/80. A depression severity index of ≥ 0.50 is considered to indicate the presence of depressive symptoms[22,23].

Quality control

The questionnaires were administered by trained researchers to ensure that they were completed accurately. During the survey process, the researchers thoroughly explained the questionnaire content and completion method to the patients to ensure that they understood and could respond accurately. For missing or inconsistent responses in the questionnaire, timely communication and verification with patients were made to ensure the completeness of the data. All questionnaires were completed on the day of the patient's visit to minimize recall bias. All examinations were performed by strictly trained ophthalmic professionals to ensure consistency in examination methods and standards. Examination equipment was calibrated and maintained before each examination to ensure accuracy and reliability. All data were double-entered, and strict quality checks were performed during the entry process to ensure accuracy and consistency.

Statistical analysis

Data were analyzed using SPSS version 23.0 (IBM Corp., Armonk, NY, United States). Measurement data were tested for normality using the Kolmogorov-Smirnov test, and those conforming to normal distribution are expressed as mean ± SD. Non-normally distributed data are expressed as median (interquartile range), and count data are expressed as n (%). Comparisons between groups were made using the independent samples t-test or the Mann-Whitney U test. Comparisons of categorical variables were made using the χ² test or the Fisher’s exact probability method. Correlation analyses were performed using Pearson’s or Spearman’s methods, depending on the type of data distribution. Linear regression analysis was performed using the OSDI scores as the dependent variable. Variables with P < 0.05 in the one-way linear regression analysis were included in the multiple linear regression model. The final model was used to assess the contribution of each variable using standardized regression coefficients (β), and 95% confidence intervals (CIs) were calculated. The variance inflation factor was calculated to diagnose collinearity in the model. The goodness of fit of the model was evaluated using the determination coefficient (R²), and the overall significance of the model was evaluated using the F-test. The Durbin-Watson value was used to evaluate the autocorrelation of residuals. All statistical tests were two-sided, and differences were considered statistically significant at P < 0.05.

RESULTS

Baseline characteristics and sleep quality of the patients

A total of 358 patients with DED were included in this study, of which 67.88% (243/358) were females. The mean PSQI score of the patients was 9.94 ± 2.18, suggesting poor overall sleep quality (Table 1). Age-stratified analysis revealed that the PSQI scores were significantly higher in the ≥ 50 years group than in the < 50 years group (10.27 ± 2.35 vs 9.71 ± 2.03, P = 0.020). In addition, the PSQI scores were higher in the less educated group (10.07 ± 2.11) compared to those with a college education or higher (9.45 ± 2.37, P = 0.028). However, gender, hypertension, diabetes mellitus, and duration of illness had no significant effect on sleep quality (P > 0.05).

Table 1 Distribution of sleep quality scores among the patients, n (%)/mean ± SD.

Variable	Number	PSQI	Statistic	P value
Total	358 (100)	9.94 ± 2.18
Age (years)			t = -2.34	0.020
< 50	213 (59.50)	9.71 ± 2.03
≥ 50	145 (40.50)	10.27 ± 2.35
Gender			t = -0.28	0.780
Female	243 (67.88)	9.91 ± 2.25
Male	115 (32.12)	9.98 ± 2.02
Degree of education			t = -2.20	0.028
College or higher	77 (21.51)	9.45 ± 2.37
Below college	281 (78.49)	10.07 ± 2.11
Hypertension			t = -1.48	0.140
No	239 (66.76)	9.82 ± 2.18
Yes	119 (33.24)	10.18 ± 2.17
Diabetes			t = -1.72	0.086
No	283 (79.05)	9.83 ± 2.16
Yes	75 (20.95)	10.32 ± 2.20
Duration of illness			t = -1.51	0.133
< 12 months	188 (52.51)	9.77 ± 2.19
≥ 12 months	170 (47.49)	10.12 ± 2.16

PSQI: Pittsburgh Sleep Quality Index.

Distribution characteristics of anxiety

The mean SAS score of the patients was 47.30 ± 4.90, indicating severe levels of anxiety (Table 2). The anxiety level of patients with an illness course of ≥ 12 months was significantly higher than that of those with an illness course of < 12 months (47.84 ± 4.61 vs 46.81 ± 5.11, P = 0.047). No statistical differences were found in other characteristics.

Table 2 Distribution of anxiety scores among the patients, n (%)/mean ± SD.

Variable	Number	SAS	Statistic	P value
Total	358 (100)	47.30 ± 4.90
Age (years)			t = -1.83	0.068
< 50	213 (59.50)	46.91 ± 4.92
≥ 50	145 (40.50)	47.87 ± 4.83
Gender			t = -0.55	0.581
Female	243 (67.88)	47.20 ± 4.89
Male	115 (32.12)	47.50 ± 4.92
Degree of education			t = 0.01	0.996
College or higher	77 (21.51)	47.30 ± 4.80
Below college	281 (78.49)	47.30 ± 4.93
Hypertension			t = -0.98	0.328
No	239 (66.76)	47.12 ± 4.78
Yes	119 (33.24)	47.66 ± 5.13
Diabetes			t = -0.44	0.657
No	283 (79.05)	47.24 ± 4.93
Yes	75 (20.95)	47.52 ± 4.81
Duration of illness			t = -1.99	0.047
< 12 months	188 (52.51)	46.81 ± 5.11
≥ 12 months	170 (47.49)	47.84 ± 4.61

SAS: Self-Rating Anxiety Scale.

Distribution characteristics of depression

The mean SDS score of the patients was 50.08 ± 5.52, indicating severe levels of depression (Table 3). Patients aged ≥ 50 years [51.16 ± 5.38 vs 49.34 ± 5.51 (those aged < 50 years), P = 0.002], those with diabetes mellitus [51.35 ± 5.23 vs 49.74 ± 5.56 (those without diabetes mellitus), P = 0.025], and those with an illness course of ≥ 12 months [51.09 ± 5.73 vs 49.16 ± 5.17 (those with an illness course of < 12 months), P < 0.001] had significantly higher levels of depression.

Table 3 Distribution of depression scores among the patients, n (%)/mean ± SD.

Variable	Number	SDS	Statistic	P value
Total	358 (100)	50.08 ± 5.52
Age (years)			t = -3.10	0.002
< 50	213 (59.50)	49.34 ± 5.51
≥ 50	145 (40.50)	51.16 ± 5.38
Gender			t = -1.71	0.088
Female	243 (67.88)	49.73 ± 5.62
Male	115 (32.12)	50.80 ± 5.25
Degree of education			t = -0.76	0.445
College or higher	77 (21.51)	49.65 ± 5.27
Below college	281 (78.49)	50.19 ± 5.59
Hypertension			t = -1.57	0.117
No	239 (66.76)	49.73 ± 5.14
Yes	119 (33.24)	50.76 ± 6.18
Diabetes			t = -2.26	0.025
No	283 (79.05)	49.74 ± 5.56
Yes	75 (20.95)	51.35 ± 5.23
Duration of illness			t = -3.35	< 0.001
< 12 months	188 (52.51)	49.16 ± 5.17
≥ 12 months	170 (47.49)	51.09 ± 5.73

SDS: Self-Rating Depression Scale.

Relationship between ocular surface function and demographic characteristics

The mean OSDI score was 32.66 ± 6.79, and the mean TBUT was 4.69 ± 1.18 s. Moreover, the mean FL score was 2.77 ± 0.52, and the mean SIT score was 4.84 ± 1.20 mm/5 minutes (Table 3). Patients aged ≥ 50 years had a higher OSDI score (34.04 ± 7.27 vs 31.73 ± 6.29, P < 0.01) and a shorter TBUT (4.52 ± 1.09 vs 4.81 ± 1.23, P < 0.05) compared to those aged < 50 years. Patients with diabetes exhibited more severe corneal damage (FL score: 2.97 ± 0.48 vs 2.72 ± 0.51, P < 0.01) and lower tear production (SIT score: 4.68 ± 1.10 vs 4.88 ± 1.22, P < 0.05) than those without diabetes. Moreover, those with a disease duration ≥ 12 months had higher corneal staining scores (2.86 ± 0.50 vs 2.69 ± 0.52, P < 0.01) and lower tear production (4.60 ± 1.15 vs 5.05 ± 1.21, P < 0.01) than those with a disease duration < 12 months (Table 3).

Correlation of ocular surface function with sleep quality and psychological factors

The PSQI scores were significantly positively correlated with the OSDI (r = 0.245, P < 0.001) and FL score (r = 0.206, P < 0.001), and negatively correlated with the TBUT (r = -0.177, P < 0.001) and SIT score (r = -0.116, P = 0.028), indicating that poorer sleep quality was associated with worse ocular surface symptoms and reduced tear film stability (Table 4). The SAS and SDS scores were significantly positively correlated with the OSDI (r = 0.183 and r = 0.200, P < 0.001) and FL score (r = 0.218 and r = 0.284, P < 0.001), and negatively correlated with the TBUT (r = -0.208 and r = -0.162, P < 0.001 or P = 0.002) and SIT score (r = -0.167 and r = -0.234, P < 0.001), indicating that higher levels of anxiety and depression were associated with more severe ocular surface dysfunction. The PSQI was significantly positively correlated with the SAS score (r = 0.157, P = 0.003) and SDS score (r = 0.207, P < 0.001), indicating a synergistic effect between sleep quality and anxiety/depression.

Table 4 Distribution of ocular surface functions among the patients, n (%)/mean ± SD.

Variable	Number	OSDI	TBUT	FL	SIT
Total	358 (100)	32.66 ± 6.79	4.69 ± 1.18	2.77 ± 0.52	4.84 ± 1.20
Age (years)
< 50	213 (59.50)	31.73 ± 6.29b	4.81 ± 1.23a	2.74 ± 0.53	4.92 ± 1.22
≥ 50	145 (40.50)	34.04 ± 7.27	4.52 ± 1.09	2.81 ± 0.49	4.71 ± 1.16
Gender
Female	243 (67.88)	32.69 ± 6.64	4.75 ± 1.22	2.77 ± 0.53	4.84 ± 1.18
Male	115 (32.12)	32.61 ± 7.13	4.57 ± 1.11	2.77 ± 0.50	4.82 ± 1.25
Degree of education
College or higher	77 (21.51)	32.23 ± 6.90	4.91 ± 1.09	2.72 ± 0.53	4.81 ± 1.24
Below college	281 (78.49)	32.78 ± 6.77	4.63 ± 1.20	2.78 ± 0.51	4.84 ± 1.19
Hypertension
No	239 (66.76)	32.20 ± 6.78	4.65 ± 1.17	2.73 ± 0.52	4.82 ± 1.24
Yes	119 (33.24)	33.60 ± 6.75	4.77 ± 1.22	2.83 ± 0.50	4.87 ± 1.12
Diabetes
No	283 (79.05)	32.51 ± 6.67	4.76 ± 1.18*	2.72 ± 0.51b	4.88 ± 1.22
Yes	75 (20.95)	33.24 ± 7.24	4.42 ± 1.17	2.97 ± 0.48	4.68 ± 1.10
Duration of illness
< 12 months	188 (52.51)	32.41 ± 6.65	4.76 ± 1.11	2.69 ± 0.52b	5.05 ± 1.21b
≥ 12 months	170 (47.49)	32.95 ± 6.95	4.61 ± 1.26	2.86 ± 0.50	4.60 ± 1.15

^aP < 0.05.

^bP < 0.01.

OSDI: Ocular surface disease index; TBUT: Tear film break-up time; FL: Fluorescein Staining Score; SIT: Schirmer I test.

Multiple linear regression analysis of the factors influencing ocular surface function

The regression model with OSDI as the dependent variable revealed that age ≥ 50 years (β = 1.55, 95%CI: 0.16-2.93, P = 0.029), PSQI score (β = 0.58, 95%CI: 0.26-0.90, P < 0.001), anxiety (β = 0.17, 95%CI: 0.03-0.31, P = 0.017), and depression (β = 0.15, 95%CI: 0.03-0.28, P = 0.019) were all independent predictors of the exacerbation of ocular surface symptoms (Table 5). Moreover, the strength of the effect of sleep quality on ocular surface symptoms (β = 0.58) was greater than that of anxiety (β = 0.17) and depression (β = 0.15), suggesting that improved sleep quality may be more clinically significant in alleviating ocular surface symptoms.

Table 5 Correlation among the ocular surface status indicators, sleep quality, anxiety, and depression.

Variable		PSQI	SAS	SDS	OSDI	TBUT	FL	SIT
PSQI	r	1	0.1574	0.207	0.245	-0.177	0.206	-0.116
	P value		0.003	< 0.001	< 0.001	< 0.001	< 0.001	0.028
SAS	r	0.1574	1	0.162	0.183	-0.208	0.218	-0.167
	P value	0.003		0.002	< 0.001	< 0.001	< 0.001	0.001
SDS	r	0.207	0.162	1	0.200	-0.162	0.284	-0.234
	P value	< 0.001	0.002		< 0.001	0.002	< 0.001	< 0.001
OSDI	r	0.245	0.183	0.200	1	-0.111	0.213	-0.245
	P value	< 0.001	< 0.001	< 0.001		0.037	< 0.001	< 0.001
TBUT	r	-0.177	-0.208	-0.162	-0.111	1	-0.254	0.195
	P value	< 0.001	< 0.001	0.002	0.037		< 0.001	< 0.001
FL	r	0.206	0.218	0.284	0.213	-0.254	1	-0.280
	P value	< 0.001	< 0.001	< 0.001	< 0.001	< 0.001		< 0.001
SIT	r	-0.116	-0.167	-0.234	-0.245	0.195	-0.280
	P value	0.028	0.001	< 0.001	< 0.001	< 0.001	< 0.001

PSQI: Pittsburgh Sleep Quality Index; SAS: Self-Rating Anxiety Scale; SDS: Self-Rating Depression Scale; OSDI: Ocular surface disease index; TBUT: Tear film break-up time; FL: Fluorescein staining score; SIT: Schirmer I test.

Evaluation of the multiple linear regression model

The variance inflation factor values for the variables in the model were less than 5, indicating that there was no serious multicollinearity among the independent variables and confirming the model’s stability (Table 6). The model coefficient of determination (R²) was 0.112, indicating that age, PSQI, SAS score, and SDS score explained 11.2% of the variation in the OSDI. The F-test results showed that the model was significant as a whole [F(4, 353) = 11.094, P < 0.001], indicating that the combined explanation of the independent variables for the dependent variable was statistically significant. The Durbin-Watson value was 2.007, indicating no autocorrelation between the residuals of the model, which meets the assumptions of the multiple linear regression model.

Table 6 Multiple linear regression analysis.

Variables	β	SE	t	P value	β (95%CI)	VIF
Intercept	10.60	4.27	2.48	0.014	10.60 (2.23-18.97)
Age (years)
< 50					Reference1
≥ 50	1.55	0.71	2.19	0.029	1.55 (0.16-2.93)	1.040
PSQI	0.58	0.16	3.59	< 0.001	0.58 (0.26-0.90)	1.070
SAS	0.17	0.07	2.40	0.017	0.17 (0.03-0.31)	1.048
SDS	0.15	0.06	2.36	0.019	0.15 (0.03-0.28)	1.082
R2	0.112
F value	F(4, 353) = 11.094, P < 0.001
Durbin-Watson value	2.007

¹Reference group.

PSQI: Pittsburgh Sleep Quality Index; SAS: Self-Rating Anxiety Scale; SDS: Self-Rating Depression Scale; CI: Confidence interval; VIF: Variance inflation factor.

DISCUSSION

This study explored the correlation between ocular surface function and sleep quality, anxiety, and depression in patients with DED using a cross-sectional survey. The results revealed significant associations in various aspects. We found that patients with DED generally had reduced sleep quality (mean PSQI: 9.94 ± 2.18), and that sleep disturbances were independently and positively correlated with the severity of DED symptoms (OSDI). In addition, elevated levels of anxiety and depression were significantly associated with the indicators of ocular surface function (e.g., decreased tear film stability and increased corneal damage), with depression having a particularly prominent effect on corneal epithelial damage (r = 0.284). Multiple regression analyses further confirmed that age, sleep quality, and psychological state were independent predictors of DED symptom exacerbation. These findings not only support the complexity of DED as a multisystem interaction disease but also provide new targets for clinical intervention.

In this study, we found that patients with DED had poorer sleep quality and higher mean PSQI scores, indicating that sleep problems were prevalent among the patients. Based on a community survey, Yu et al[24] reported a strong association between poor sleep quality and the aggravation of DED, suggesting a potential vicious cycle between the two. From the perspective of pathophysiological mechanisms, sleep disturbances may exacerbate DED through multiple pathways. Sleep deprivation can lead to autonomic dysfunction, particularly sympathetic overactivation, which inhibits the parasympathetic innervation of the lacrimal glands, thereby reducing basal tear production[25]. Furthermore, Li et al[26] demonstrated through animal experimental models that sleep deprivation can lead to corneal epithelial cell defects, increased corneal sensitivity, and apoptosis, which may induce DED. This is consistent with the negative correlation between the PSQI and SIT scores in the present study. In addition, poor sleep quality may affect the ocular surface microenvironment by promoting a systemic inflammatory response[27]. Existing studies have shown that the decline in sleep quality can lead to an increase in the levels of various inflammatory factors (such as C-reactive protein and interleukin-6), and these inflammatory factors can induce the necrosis of corneal epithelial cells, thereby aggravating DED symptoms[28,29]. This might explain the negative correlation between the PSQI and TBUT. Furthermore, the circadian rhythm disorder caused by sleep disorders may interfere with the nocturnal repair process of the corneal epithelium[30]. Zeng et al[31] discovered through animal models that circadian rhythm disorders may inhibit the recovery of corneal epithelial cells and aggravate inflammatory responses by regulating the expression of related genes. This is consistent with the finding that the corneal staining score was higher in those with a longer disease course.

The results of this study also showed a significant association between anxiety/depression and ocular surface function. Previous studies have shown that the association between DED and anxiety/depression may involve multiple pathophysiological mechanisms[32]. Depressive states are often accompanied by hyperactivation of the hypothalamic-pituitary-adrenal axis, which sustains elevated cortisol levels[33]. Cortisol not only inhibits mucin secretion from the lacrimal gland but also exacerbates corneal epithelial barrier disruption by upregulating matrix metalloproteinase-9 expression, which may account for the high correlation between depression and FL scores[34,35]. A previous study revealed that patients with depression exhibit significantly lower blinking frequency[36]. This decrease in blinking, which is a key physiological mechanism for maintaining uniform tear film distribution, can directly lead to excessive tear evaporation[36]. This may partly explain the observed association between depression and reduced TBUT. Notably, FL demonstrated the strongest correlation with the SDS score among all ocular surface parameters (r = 0.284). This suggests that corneal epithelial damage may be particularly associated with depressive symptoms in patients with DED, potentially reflecting shared neuroinflammatory pathways[37].

In this study, corneal damage was most severe in patients with diabetes mellitus combined with depression, suggesting that there may be a synergistic effect between hyperglycemia and psychological stress[38]. Hyperglycemia induces a sustained immune-inflammatory response, which may be further amplified by the inflammation associated with depression, forming a vicious cycle involving metabolic, psychological, and ocular surface dysfunction. In addition, the role of age in the progression of dry eye cannot be ignored. Patients ≥ 50 years of age in this study had higher OSDI scores and poorer tear film stability, which may be related to age-related physiological changes. On the one hand, the gradual replacement of the parenchymal tissue of the lacrimal gland with fatty and fibrous tissue with age leads to a decrease in tear secretion. On the other hand, the decrease in corneal nerve density diminishes the transient reflex and further impairs tear film dynamics[39]. Still, the specific mechanisms involved need to be verified by further experimental studies.

Our study findings have important implications for clinical practice, particularly in managing patients with DED. Clinicians should consider routinely assessing sleep quality (e.g., using brief questionnaires such as PSQI) and screening for anxiety/depression (e.g., using SAS/SDS) in patients with DED, especially in those who are older, have comorbid diabetes, or report longer disease duration. Special attention to comorbidities and targeted follow-up is essential in older patients. For patients with diabetes, blood sugar testing should be strengthened. Integrated intervention strategies could be proposed for patients with these issues. For instance, cognitive-behavioral therapy for insomnia or sleep hygiene education might be recommended to improve sleep quality, which we found had a stronger influence on ocular surface symptoms than psychological factors. Similarly, for those with significant anxiety or depression, collaboration with psychiatrists for appropriate management (e.g., anxiolytic medications and counseling) could improve their mental well-being and potentially alleviate DED symptoms.

Although this study provides valuable clinical evidence, it had some limitations. First, this study was conducted at a single center, and all participants were from the local area. Regional differences in environmental factors and medical service access may lead to selection bias. This emphasizes the necessity for conducting multi-center research. Second, the cross-sectional design was unable to establish a causal relationship between sleep disorders and DED, and future longitudinal studies or intervention trials are required. However, existing prospective studies and animal experiments have provided supportive evidence for the potential causal direction of this association. A large-scale prospective cohort study in Ningbo, China, followed 257932 participants for 8 years and found that patients with sleep disorders had a 3-fold higher incidence of DED than those without[40]. For psychological factors, a 1-year follow-up of the DREAM study showed that patients with depression had more severe DED symptoms and corneal staining signs, indicating that depressive states may exacerbate DED progression[14]. Despite these supportive findings, future longitudinal studies or intervention trials (e.g., assessing whether improving sleep quality reduces DED incidence) are still required to confirm the causal relationship. Third, owing to the limitations of the cross-sectional design, we did not conduct prior stratification based on the severity of DED nor carry out intervention subgroup analysis. Therefore, it was not possible to clarify whether the association strength between sleep/psychological factors and ocular surface damage varied with the disease stage, and it was also challenging to determine the precise intervention timing for each stage based on this. This further emphasizes the necessity of subsequent longitudinal intervention studies. Fourth, this study only used questionnaires and clinical examination indices to assess patients’ sleep quality and anxiety/depressive symptoms, lacking objective biological markers. Future studies can combine brain imaging and neurotransmitter testing to further explore the neurobiological mechanisms among DED, sleep quality, and anxiety/depressive symptoms. Finally, age, sleep, anxiety, and depression collectively explained 11.2% of the variability in ocular surface function in patients with DED. This indicates that although these factors have a certain impact on ocular surface function, there are still many variations that have not been explained by these variables. Nevertheless, the explanations of these non-disease factors still account for more than 10% and have certain clinical value. The development of DED is essentially the result of a combination of multi-systemic factors. The low R² underscores the need for a holistic assessment of patients with DED beyond the ocular surface parameters. Future studies need to further explore other potential influencing factors, such as environmental factors, digital screen time, systemic comorbidities, and psychosocial factors, in order to more fully understand the pathogenesis of DED.

CONCLUSION

This study revealed significant correlations among ocular surface function, sleep quality, and the level of anxiety and depression by systematically analyzing the clinical data of 358 patients with DED. The results showed that patients with poorer sleep quality and higher levels of anxiety and depression had higher OSDI scores, shorter TBUT, increased FL score, and reduced SIT score. Multiple linear regression analyses further showed that the sleep quality (PSQI) and anxiety (SAS) and depression (SDS) scores were independent predictors of ocular surface function, with sleep quality influencing ocular surface symptoms more strongly than psychological factors. Ocular surface dysfunction was more severe in patients aged ≥ 50 years, with comorbid diabetes and longer disease duration, suggesting that clinical attention needs to be paid to the comprehensive management of specific populations. The results of this study emphasize the importance of sleep quality and psychological status in the clinical management of DED. Moreover, improving ocular surface function and relieving psychological stress may be important for improving sleep quality and psychological well-being in patients with DED. Future studies should further explore the causal relationship and underlying mechanisms among the three factors and validate the generalizability of the present findings through multicenter, large-sample prospective studies.