Sleep Disturbances and Dry Eye Symptoms and Signs, Systemic Conditions, and Inflammatory Markers in the Dry Eye Assessment and Management Study

Abstract

Purpose:

To investigate the association between sleep disturbances, dry eye disease (DED) severity, and DED risk factors.

Methods:

We conducted a secondary analysis of the Dry Eye Assessment and Management (DREAM) study, a randomized trial of 535 patients with moderate-to-severe DED. Participants self-reported sleep disturbances at baseline. DED symptoms were assessed using the Ocular Surface Disease Index (OSDI) and Brief Ocular Discomfort Index (BODI). DED signs were assessed using tear film break-up time, Schirmer test, corneal and conjunctival staining, tear osmolarity, and meibomian gland dysfunction assessment. Immune cells were assessed using conjunctival impression cytology. Outcomes were compared between participants with ongoing and no history of sleep disturbances, adjusting for confounders.

Results:

Mean age was 58 years, with 81% female participants. Ongoing sleep disturbances (n=113) were significantly associated with older age (mean 60.5 vs. 57.4; p=0.03), smoking (39.8% vs. 28.7%; p=0.02), and DED treatment usage. Systemic conditions associated with sleep disturbances included thyroid dysfunction (26.5% vs.15.7% p=0.007), irritable bowel (18.6% vs. 5.4%; p<.0001), osteoarthritis (38.9% vs. 21.6%; p=0.0007), and depression (30.1% vs. 11.5%; p<.0001). Sleep disturbances were associated with lower conjunctival staining scores (2.46 vs. 2.77; p=0.03) and higher concentrations of dendritic cells in ocular surface white blood cells (12.0% vs. 8.1%; p=0.01).

Conclusions:

Sleep disturbances were associated with older age, smoking history, systemic diseases, and higher dendritic cell concentration among patients with moderate-to-severe DED. These findings support further investigation of sleep disturbances as a comorbidity of patients with DED. Additional research is needed to understand the mechanisms behind this relationship.

Introduction

Dry eye disease (DED) is a major clinical and public health problem with an estimated global prevalence ranging from 5% to 50%^1,2. As a condition with multiple etiologies, DED is generally characterized by inflammation of the ocular surface, causing pain and visual disturbances³. These symptoms create negative burdens on both individual quality of life and economic systems due to decreased work productivity^2,4. The estimated economic burden of DED in the United States exceeds $55 billion annually due to direct costs (e.g., treatments, physician visits) and indirect costs from loss of productivity⁵.

Previous studies have found that sleep disturbances are more prevalent among patients with DED, but the mechanisms underlying this relationship are not well understood^6–9. Some studies have found correlations between ocular pain and insomnia, suggesting that discomfort due to DED symptoms may be a factor in sleep disturbances^10,11. Other studies have found associations between DED and mental health issues such as depression and anxiety, suggesting another potential axis by which sleep disturbances and DED are related^9,12. Sleep disorders also create significant burdens on society, causing an estimated $95 billion in incremental health care costs in the United States each year¹³. Overall, the underlying mechanisms and directionality between sleep disturbances and DED remain poorly understood despite the widespread impact of both conditions. Additionally, previous studies have not evaluated whether sleep disturbances are associated with more severe DED signs and symptoms.

To investigate the relationship between sleep disturbances and DED severity, as well as potentially related systemic conditions (e.g., irritable bowel, arthritis) and ocular inflammatory markers, we conducted a secondary analysis of rich data from the Dry Eye Assessment and Management (DREAM) study¹⁴. The well-characterized DREAM study cohort presents a unique opportunity to evaluate comorbid health conditions among patients with moderate-to-severe DED.

Materials and Methods

Specific methods used in the DREAM study have been described in previous literature^14,15. Only relevant features from the DREAM study are reported below.

The DREAM study was a multicenter, randomized clinical trial of 535 participants conducted between October 2014 and July 2016 to evaluate the efficacy of ω-3 fatty acid supplements in treating DED compared to placebo. The DREAM study is registered with ClinicalTrials.gov (NCT02128763). Participants were enrolled from 27 clinical centers throughout the United States. Each center obtained its respective Institutional Review Board approval and abided by the Health Insurance Portability and Accountability Act. The DREAM study was conducted under the US Food and Drug Administration as an investigational new drug application (IND 106,387).

Participant Selection

The trial was designed to include patients with moderate-to-severe DED. Eligibility criteria for the study were age 18 years or older, DED symptoms for at least 6 months, artificial tear use at least twice a day for 2 weeks before screening, and an Ocular Surface Disease Index (OSDI) score of 25–80 at the screening visit and 21–80 at the eligibility-confirmation visit. Additionally, participants must have had at least two of the following four signs of DED in the same eye at both screening and eligibility-confirmation visits: conjunctival staining score 1 or more, corneal staining score 4 or more, tear film break-up time (TBUT) of 7 seconds or less, and Schirmer’s test of 1 to 7 mm per 5 minutes. Participants were evaluated for DED symptoms and signs, as well as ocular surface inflammatory markers and depression, at baseline, 6-month, and 12-month follow-up visits.

Informed consent was collected from all participants and the study adhered to principles of the Declaration of Helsinki.

Assessment of Baseline Participant Characteristics and Health Conditions

At the screening visit, participants were verbally asked about their medical history for 58 systemic or extraocular diagnoses, including sleep disturbances, as told to them by a doctor or other health professional within the past two years. Possible answers were no history, past history, or ongoing disease. This study analyzes participant responses to 10 systemic conditions reported in previous literature as potential risk factors for DED, including rosacea, Sjogren’s Syndrome, thyroid dysfunction, hypertension, diabetes, rheumatoid arthritis, irritable bowel, osteoarthritis, hypercholesterolemia, and depression¹⁶. Additionally, participants were asked at the baseline visit whether they had ever smoked cigarettes daily and whether they currently smoke. The Medical Outcomes Study 36-Item Short Form Health Survey (SF-36) version 2.0 was also administered to each participant to generate Physical Component Summary (PCS) and Mental Component Summary (MCS) scores ranging from 0 to 100, with higher scores indicating better physical or mental health¹⁷.

Assessment of DED Symptoms and Signs

At baseline, 6 months, and 12 months, DED symptoms were evaluated by the widely used Ocular Surface Disease Index (OSDI) and the Brief Ocular Discomfort Inventory (BODI; a modification of the Brief Pain Inventory), both of which had scores ranging from 0 to 100 (higher scores indicating more severe DED symptoms)^18,19. We also investigated the BODI #3 pain score from the same inventory.

DED signs were assessed in both eyes for the following six signs: corneal fluorescein staining (range of 0–15; 15 indicating greater abnormality), conjunctival lissamine green staining (range 0–6, with 6 indicating greatest abnormality), TBUT (lower scores indicating greater abnormality), Schirmer’s test (lower scores indicating greater abnormality), tear osmolarity (higher scores indicating greater abnormality), and meibomian gland dysfunction evaluating plugging and lid secretion (rated by clinicians on a scale of 0–3 each with total score 0–6; higher scores indicating greater dysfunction). Additionally, a composite score for the six DED signs was computed by taking the mean of the six signs transformed to a common 0 to 1 score using methods from past studies^15,16,20.

Assessment of Ocular Surface Inflammatory Markers

Detailed methods for ocular surface inflammatory marker collection and analysis in the DREAM study have been previously described²¹. Conjunctival impression cytology samples were collected from 527 participants (98.5%) and assayed for white blood cells (WBC) including T helper cells, T regulatory cells, Th1, Th17, and dendritic cells.

Statistical Analysis

We compared participants with self-reported, ongoing sleep disturbances against participants without history of sleep disturbances for scores of DED symptoms (OSDI, BODI) and 6 signs, the presence of ongoing systemic conditions, and ocular surface inflammatory markers. The comparisons of DED symptoms and signs across sleep disturbance groups were made through univariable and multivariable models that were adjusted for age, gender, smoking status, Sjögren’s Syndrome, facial rosacea, rheumatoid arthritis, and depression defined by Mental Component Summary score of 42 or below, since these factors were significantly associated with the severity of DED symptoms and signs in the DREAM study^16,22.

Because the DREAM study did not find a significant effect of ω-3 supplementation on DED outcome measures¹⁴, and because ω-3 supplementation has not been found to have an effect on sleep disturbance in adults²³, analyses were conducted using the data from the 2 treatment groups combined. Additionally, we performed analyses using combined data collected from baseline, 6-month, and 12-month visits to improve statistical power while accounting for the correlation of repeated measures using generalized estimating equations.

Statistical analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC), and two-sided P<0.05 (without adjustment for multiple comparisons) was considered statistically significant.

Results

Participant Characteristics

Among the 535 participants enrolled in the DREAM study, the mean age was 58 (SD=13.2) years. 434 (81%) were female and 398 (74.4%) were white. 168 (31.4%) reported a history of smoking (Table 1).

Table 1:

Patient characteristics and systemic conditions by status of self-reported sleep disturbances

	History of Sleep Disturbances
	No History (N=407)	Ongoing (N=113)	P-value
Age (Years): Mean (SD)	57.4 (13.39)	60.5 (12.46)	0.03
Race			0.54
White	306 (75.2%)	86 (76.1%)
Black/African American	44 (10.8%)	15 (13.3%)
Other	57 (14.0%)	12 (10.6%)
Gender: Female (%)	324 (79.6%)	97 (85.8%)	0.14
Smoking History: Yes (%)	117 (28.7%)	45 (39.8%)	0.02
Treatments used for DED: Yes (%)
Artificial tears or gel	312 (76.7%)	104 (92.0%)	0.0003
Cyclosporine drops	153 (37.6%)	49 (43.4%)	0.27
Warm lid soaks	80 (19.7%)	30 (26.5%)	0.11
Baby Shampoo	19 (4.7%)	12 (10.6%)	0.02
Lid Scrubs	39 (9.6%)	22 (19.5%)	0.004
Self-reported Ongoing Systemic Conditions: Yes (%)
Rosacea (facial)	76 (18.7%)	28 (24.8%)	0.15
Sjogren’s Syndrome:	39 (9.6%)	17 (15.0%)	0.10
Thyroid dysfunction	64 (15.7%)	30 (26.5%)	0.007
Hypertension	107 (26.3%)	35 (31.0%)	0.25
Diabetes	39 (9.6%)	14 (12.4%)	0.67
Rheumatoid arthritis	32 (7.9%)	12 (10.6%)	0.60
Irritable Bowel	22 (5.4%)	21 (18.6%)	<.0001
Osteoarthritis	88 (21.6%)	44 (38.9%)	0.0007
Hypercholesterolemia	128 (31.4%)	36 (31.9%)	0.48
Depression	47 (11.5%)	34 (30.1%)	<.0001
SF-36 score: Mean (SD)
Physical Component Summary (PCS)	48.6 (9.38)	43.4 (9.84)	<.0001
Mental Component Summary (MCS)	53.4 (8.49)	49.1 (11.09)	<.0001
Depression defined as MCS <=42: Yes (%)	47 (11.5%)	32 (28.3%)	<.0001

At baseline, 113 (21.1%) reported ongoing sleep disturbances and 407 (76%) reported no history of sleep disturbances (Table 1). 15 (2.8%) participants reported past history of sleep disturbances; these patients were excluded from statistical analyses due to the small sample size and unavailability of sleep disturbance history prior to the screening visit.

Compared to participants without a history of sleep disturbance, participants who reported ongoing sleep disturbances tended to be older (mean age 60.5 vs. 57.4 years; p=0.03), with a higher percentage of reporting smoking history (39.8% vs. 28.7%; p=0.02). Participants who reported ongoing sleep disturbances were similar in race and gender to those who reported no history of sleep disturbances (Table 1).

Compared to participants without a history of sleep disturbance, participants who reported ongoing sleep disturbances were more likely to have self-reported treatments for DED, specifically using artificial tears or gel (92.0% vs. 76.7%; p=0.0003), baby shampoo (10.6% vs. 4.7%; p=0.02), and lid scrubs (19.5% vs. 9.6%; p=0.004) (Table 1).

Sleep Disturbances and Systemic Conditions

Participants who reported ongoing sleep disturbances also reported some ongoing systemic conditions at higher percentages compared to participants who reported no history of sleep disturbances. These included thyroid dysfunction (26.5% vs. 15.7%; p=0.007), irritable bowel (18.6% vs. 5.4%; p<.0001), osteoarthritis (38.9% vs. 21.6%; p=0.0007), and self-reported depression (30.1% vs. 11.5%; p<.0001) (Table 1). Participants with and without sleep disturbances had similar rates of self-reported facial rosacea, Sjogren’s Syndrome, hypertension, diabetes, rheumatoid arthritis, and hypercholesterolemia (all p≥0.10, Table 1).

Additionally, participants with ongoing sleep disturbances had worse quality of life than those with no history of sleep disturbances in terms of lower mean SF-36 PCS scores (43.4 vs. 48.6; p<.0001) and SF-36 MCS scores (49.1 vs. 53.4; p<.0001). Participants with ongoing sleep disturbances also had higher rates of depression defined as MCS <=42 (28.3% vs. 11.5%; p<.0001) (Table 1).

Sleep Disturbances and DED Symptoms and Signs

In the univariate analysis for associations between sleep disturbances and DED symptoms and signs, participants with ongoing sleep disturbances had higher mean BODI scores (e.g., worse symptoms) than those with no history of sleep disturbances (28.4 vs. 25.2; p=0.005) and higher BODI #3 pain scores (39.7 vs. 36.5; p=0.02). All other DED symptoms and signs were similar in both groups (Table 2).

Table 2:

Univariate analysis of the comparison of DED symptoms and signs combining baseline, 6 months, and 12 months across sleep disturbance groups among DREAM participants

	Sleep Disturbance
	No History (N=407)	Ongoing (N=113)
	Unadjusted mean (SE)	Unadjusted mean (SE)	Mean Difference (95% CI)	P-value
Patient-level Dry Eye Symptoms
OSDI total score	34.95 (0.55)	36.31 (1.03)	1.35 (−0.92, 3.63)	0.24
BODI score	25.21 (0.53)	28.41 (1.00)	3.20 (0.98, 5.41)	0.005
BODI #3 pain score	36.52 (0.63)	39.69 (1.18)	3.17 (0.54, 5.80)	0.02
Eye-level Dry Eye Signs *
Tear Break-up Time (sec)	3.54 (0.09)	3.51 (0.16)	−0.03 (−0.39, 0.34)	0.88
Schirmer test (mm in 5 minutes)	10.00 (0.29)	9.24 (0.48)	−0.76 (−1.87, 0.35)	0.18
Corneal staining score	3.35 (0.13)	3.70 (0.22)	0.35 (−0.16, 0.85)	0.18
Conjunctival staining score	2.74 (0.07)	2.54 (0.12)	−0.20 (−0.47, 0.07)	0.15
Meibomian gland abnormality	2.94 (0.08)	2.91 (0.13)	−0.02 ( −0.33, 0.28)	0.88
Composite dry eye severity score based on Signs	0.42 (0.01)	0.42 (0.01)	0.00 (−0.03, 0.02)	0.93
Tear osmolarity (mOsms/L)	302.76 (0.64)	304.06 (1.35)	1.30 (−1.62, 4.22)	0.38

^*Generalized Estimating Equations were used to account for inter-eye correlation for the comparison of signs.

95% CI: 95% confidence interval.

When adjusting for possible confounders (age, gender, smoking status, Sjogren’s Syndrome, facial rosacea, rheumatoid arthritis, and depression defined by MCS <=42) in the multivariate analysis, participants with ongoing sleep disturbances had lower conjunctival staining scores than those with no history of sleep disturbances (2.46 vs. 2.77; p=0.03). All other DED symptoms and signs were similar in both groups (Table 3).

Table 3:

Multivariate analysis of the comparison of DED symptoms and signs combining baseline, 6 months, and 12 months across sleep disturbance groups among DREAM participants

	Sleep Disturbance
	No History (N=407)	Ongoing (N=113)
	Adjusted mean (SE)**	Adjusted mean (SE)**	Adjusted Mean Difference (95% CI)	Adjusted P-value**
Patient-level Dry Eye Symptoms
OSDI total score	35.47 (0.54)	34.77 (1.03)	−0.69 (−3.00, 1.61)	0.56
BODI score	25.72 (0.52)	26.96 (1.00)	1.25 (−0.99, 3.48)	0.27
BODI #3 pain score	37.06 (0.63)	38.08 (1.19)	1.02 (−1.66, 3.69)	0.46
Eye-level Dry Eye Signs *
Tear Break-up Time (sec)	3.49 (0.08)	3.66 (0.17)	0.17 (−0.20, 0.53)	0.38
Schirmer test (mm in 5 minutes)	9.86 (0.28)	9.73 (0.48)	−0.14 (−1.23, 0.96)	0.80
Corneal staining score	3.42 (0.12)	3.44 (0.21)	0.02 (−0.46, 0.49)	0.94
Conjunctival staining score	2.77 (0.07)	2.46 (0.12)	−0.30 (−0.57, −0.04)	0.03
Meibomian gland abnormality	2.98 (0.08)	2.80 (0.14)	−0.18 (−0.50, 0.13)	0.26
Composite dry eye severity score based on Signs	0.42 (0.01)	0.40 (0.01)	−0.02 (−0.05, 0.00)	0.07
Tear osmolarity (mOsms/L)	302.91 (0.63)	303.67 (1.33)	0.76 (−2.12, 3.65)	0.60

^*Generalized Estimating Equations was used to account for inter-eye correlation

^**Adjusted by age, gender, smoking status, Sjögren syndrome, facial rosacea, rheumatoid arthritis, and depression defined by Sf-36 MCS <=42.

95% CI: 95% confidence interval.

Sleep Disturbances and Inflammatory Markers

Participants with ongoing sleep disturbances had higher median percentages of dendritic cells in white blood cells collected from ocular surface compared to participants with no history of sleep disturbances (12.0% vs. 8.1%; p=0.01). Total number of gated cells and all other ocular surface immune cell percentages were similar in both groups (Table 4).

Table 4:

Comparison of immune cells between sleep disturbance groups at all time points combined

Outcome	Sleep Disturbance	Median (Q1, Q3)
Number of Gated Cells	No history	14871 (8731, 22908)
	Ongoing	13846 (7929, 23739)
	P-value	0.76
Percentage of dendritic cells in white blood cells (WBC)	No history	8.10 (1.40, 25.50)
	Ongoing	12.00 (2.00, 29.60)
	P-value	0.01
Percentage of T Cells in WBC	No history	70.05 (43.90, 87.30)
	Ongoing	65.20 (42.40, 85.20)
	P-value	0.11
Percentage of cytotoxic T cells in WBC	No history	8.60 (3.40, 16.00)
	Ongoing	7.50 (2.40, 14.00)
	P-value	0.07
Percentage of T Helper Cells in WBC	No history	0.50 (0.00, 2.80)
	Ongoing	0.00 (0.00, 2.50)
	P-value	0.24
Percentage of regulatory T cells in WBC	No history	0.00 (0.00, 0.40)
	Ongoing	0.00 (0.00, 0.60)
	P-value	0.81
Percentage of Th1 Cells in WBC	No history	0.00 (0.00, 0.00)
	Ongoing	0.00 (0.00, 0.00)
	P-value	0.45
Percentage of Th17 Cells in WBC	No history	0.00 (0.00, 0.00)
	Ongoing	0.00 (0.00, 0.00)

Discussion

In this study, we analyzed data from 535 participants with moderate-to-severe DED to evaluate the associations between self-reported sleep disturbances and participant characteristics, systemic conditions, DED symptoms and signs, and inflammatory markers. About 21% of participants reported ongoing sleep disturbances, which is in line with previous findings of sleep disturbances among DED patients, ranging from about 20–40%^10,24. We also found that participants who reported ongoing sleep disturbances tended to be older and more likely to have a history of smoking, which is consistent with studies of patients without DED,^25,26 likely explained by higher rates of illnesses associated with increasing age and smoking such as obstructive airway diseases²⁷.

Associations Between Systemic Conditions and Sleep Disturbances

Our findings indicate that participants with ongoing sleep disturbances are more likely to report thyroid dysfunction, irritable bowel syndrome, osteoarthritis, and depression (self-reported and as defined by MCS<=42). These associations suggest that sleep disturbances may exacerbate or be exacerbated by these systemic conditions, potentially through shared inflammatory pathways or the physical and psychological stress they impose. For instance, previous studies have found an association between thyroid dysfunction and sleep disorders, potentially through circadian control of thyroid hormone^28,29. Sleep disturbances are also well documented in people with irritable bowel syndrome³⁰ potentially due to increased sensitivity to visceral pain and psychiatric comorbidities³¹. Our finding of a significant association between sleep disturbances and osteoarthritis is also well supported by previous studies, likely through chronic pain and fatigue pathways³². The relationship between sleep disturbances and depression has also been thoroughly documented via biochemical, genetic, and clinical perspectives³³. Of note, depression has previously been found to be associated with more severe DED signs and symptoms among the DREAM study participants²². The shared physical or psychological burden of these systemic diseases and sleep disturbances is corroborated by our significant findings in quality of life measured via SF-36 PCS and MCS scores, which are composite scores for a patient’s overall physical and mental health, respectively¹⁷. Future studies should explore the bidirectional nature of these relationships and explore whether treating these systemic conditions can improve sleep or DED outcomes.

We did not find significant associations between sleep disturbances and rosacea, Sjogren’s Syndrome, hypertension, diabetes, rheumatoid arthritis, and hypercholesterolemia. Although previous studies have found independent associations between sleep disturbances and these systemic conditions, the evidence remains inconsistent and the mechanisms behind these associations have yet to be elucidated^34–38. Moreover, none of these associations with sleep have been studied among patients with DED, which may contribute to the manifestation of these conditions.

Associations Between Sleep Disturbances and DED Symptoms and Signs

In this study, ongoing sleep disturbances were significantly associated with severity of DED symptoms as measured by BODI and BODI pain scores, only without adjustment for known confounders for DED symptoms among this cohort (age, gender, smoking status, Sjogren’s Syndrome, facial rosacea, rheumatoid arthritis, and depression defined by MCS <=42). The discrepancies between the univariate and multivariate models suggest that the demographic factors and comorbidities controlled for in the multivariate models may explain the relationship between sleep disturbances and dry eye symptoms.

Interestingly, our multivariate analysis revealed that ongoing sleep disturbances were associated with lower conjunctival staining scores, indicating less severe DED among those with sleep disturbances. This finding could be a result of the higher use of DED treatments, such as artificial tears, baby shampoo, and lid scrubs, among those with sleep disturbances. These treatments are commonly used to treat DED, with baby shampoo and lid scrubs targeted toward eyelid findings such as meibomian gland dysfunction and blepharitis³⁹. It is thus possible that increased usage of these treatments among those with sleep disturbances contributed to the lower conjunctival staining in these participants. It should be noted, however, that the difference in mean staining score between patients with and without sleep disturbances was small and likely not clinically significant (adjusted mean difference of −0.30, 95% CI: −0.56, −0.04). When examining past literature, few studies have assessed the relationship between conjunctival staining and sleep disturbances^7,24. To our knowledge, only Rolando et al²⁴ has found a statistically significant association between insomnia severity and increased conjunctival staining among 270 patients. However, our study evaluated a larger population size, only included participants with moderate-to-severe DED, and used a broader definition of sleep disturbances for participants instead of insomnia specifically—all of which may contribute to our findings.

We did not find significant associations between sleep disturbances and other DED signs and symptoms, with and without controlling for confounders. Previous studies have found significant associations between sleep disturbances and more severe DED signs as measured by the OSDI, but use different assessments of sleep disturbances^4,10,24. Moreover, the DREAM study cohort only included patients with moderate-to-severe DED as opposed to previous studies that included patients with mild to severe DED. Our narrower inclusion criteria likely resulted in a more homogenous patient population with smaller differences in DED signs and symptoms, thus yielding fewer statistically significant associations between outcomes.

Ocular Inflammatory Markers

We found higher percentages of dendritic cells in ocular surface WBC among participants with ongoing sleep disturbances, pointing to potential shared inflammatory pathways between DED and sleep disturbances. Previous studies have found evidence that dendritic cell density in ocular surfaces is influenced by sleep/wake cycles⁴⁰. Increased dendritic cells have also been found on the ocular surface of patients with DED^41,42. Our finding thus points to a potential role of sleep disturbances in DED pathology via the activation of inflammatory processes that upregulate ocular dendritic cells. These findings also support the measurement of dendritic cells as a potential biomarker for sleep disturbances and DED. Previous analysis of other ocular surface WBC collected from the DREAM study participants found few significant associations with DED signs and symptoms, which is in line with our lack of significant findings in other immune cell subtypes²¹. Future studies should investigate the causal pathways linking sleep disturbances and immune responses to assess whether interventions targeting sleep can reduce ocular surface inflammation.

Strengths and Limitations

Though studies have previously found associations between sleep disturbances and DED, few studies have been able to examine this relationship in the context of DED severity as assessed via ocular surface exams and symptomatology. The DREAM study represents one of the largest multicenter cohorts of patients with moderate-to-severe DED, with extensive medical history and ocular exam data spanning 12 months. It also is the first study of this magnitude to include evaluation of ocular surface immune cells. This allows us to evaluate sleep disturbances across systemic conditions, DED symptoms and signs, and inflammatory markers.

One key limitation of this study is that sleep disturbances were based on self-report instead of using standardized sleep dysfunction tools such as the Pittsburgh Sleep Quality Index (PSQI)⁴³. This prevented us from evaluating detailed aspects of sleep such as duration, latency, or quality, which may have provided more nuanced results compared to our dichotomous self-reported sleep disturbance metric. Self-reporting of other systemic conditions also prevented us from evaluating severity or duration of these diseases. Additionally, our findings are limited in generalizability because we only analyzed participants from the DREAM trial, all of whom met the inclusion and exclusion criteria of the trial and had moderate-to-severe DED.

In conclusion, our study evaluated sleep disturbances among 535 well-characterized participants with moderate-to-severe DED. We found significant associations of sleep disturbances with age, smoking, and four systemic conditions (thyroid dysfunction, irritable bowel, osteoarthritis, and depression), as well as SF-36 physical and mental component summary scores. DED symptoms were worse in those with sleep disturbances only when not adjusting for confounders including demographic and systemic conditions. Conjunctival staining was slightly lower in participants with sleep disorders, potentially because of increased usage of DED treatments, and the small differences were likely not clinically significant. Finally, ocular surface dendritic cells were found in higher concentrations among those with sleep disturbances. Taken together, these results further elucidate the complex relationship between sleep and DED, as well as certain associated comorbidities. Further studies are needed to understand the mechanisms and directionality behind these observed relationships.