Article Text

Download PDFPDF

Dry eye disease, dry eye symptoms and depression: the Beijing Eye Study
  1. Antoine Labbé1,2,3,
  2. Ya Xing Wang1,
  3. Ying Jie4,
  4. Christophe Baudouin2,3,
  5. Jost B Jonas1,5,
  6. Liang Xu1
  1. 1Beijing Institute of Ophthalmology, Beijing Tongren Hospital, Capital Medical University, Beijing, China
  2. 2Quinze-Vingts National Ophthalmology Hospital, Paris and Versailles Saint-Quentin-en-Yvelines University, Versailles, France
  3. 3Clinical Investigation Center (CIC) 503, Quinze-Vingts National Ophthalmology Hospital, Paris, France
  4. 4Department of Ophthalmology, Beijing Tongren Eye Center, Beijing Tongren Hospital, Capital Medical University, Beijing, China
  5. 5Department of Ophthalmology, Medical Faculty Mannheim of the Ruprecht-Karls-University Heidelberg, Heidelberg, Germany
  1. Correspondence to Dr Liang Xu, Beijing Institute of Ophthalmology, Beijing Tongren Hospital, Capital Medical University, 17 Hougou Lane, Chong Wen Men, Beijing 100005, China; xlbio1{at}


Aim To investigate the association between dry eye symptoms and depression in an adult population.

Methods In this population-based cross-sectional study, a random sample of 1957 subjects from the Beijing Eye Study was examined for dry eye disease (DED) in 2006. All patients completed an interviewer-assisted questionnaire on dry eye symptoms and underwent measurement of tear break-up time (TBUT), slit-lamp evaluation of corneal staining and meibomian gland dysfunction (MGD), and the Schirmer test. In 2011, 1456 subjects from this sample were evaluated for depression using a depression scale. The association between depression symptoms and dry eye clinical tests was evaluated.

Results Definite depression was more prevalent in patients with DED than in subjects without DED (13.7±0.4% vs 8.6±0.3%, p=0.02). The depression score was correlated with dry eye symptoms (correlation coefficient r=0.07; p=0.013) but not with TBUT (p=0.18), the Schirmer test (p=0.37), corneal staining (p=0.30) and MGD evaluation (p=0.93). In multivariate regression analysis, the risk of definite depression remained significantly associated with dry eye symptoms (p=0.028) after adjusting for lower cognitive status (p=0.01), rural region of habitation (p=0.023) and lower body weight (p=0.05).

Conclusions In an older population from Beijing, depression was associated with DED and in particular with dry eye symptoms.

  • Ocular surface
  • Epidemiology

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


According to the 2007 Dry Eye Workshop, dry eye disease (DED) is a multifactorial disease of the tears and ocular surface that results in symptoms of discomfort, visual disturbance and tear film instability.1 Typical symptoms of dry eye include burning or stinging, foreign body sensation, photophobia and blurred vision. Although these symptoms are rarely severe, the chronic ocular surface irritation and visual disturbance observed in DED directly decrease the quality of life of those with the condition,2–4 and interfere with the ability to carry out daily functions.3

The chronic discomfort observed in DED may also have a negative impact on other aspects of patient health, such as cognitive processes, mood and mental health.5 The association between DED and psychiatric disorders has recently been investigated in large retrospective epidemiological studies analysing the comorbidities associated with dry eye.6–8 These studies, based on medical record diagnostic classification and treatments, found a positive association between DED and depression. Nevertheless, no detailed evaluation of this association, in particular with regard to objective DED parameters or symptoms and depression, has been conducted. Additionally, few studies to date have further analysed the relationship between DED parameters and depression,2 ,9 ,10 and only one population-based study evaluated an older population.11 Thus, the aim of the present study was to investigate the association between depression symptoms and results from objective and subjective tests for DED in an adult population-based study.


The Beijing Eye Study is a population-based cross-sectional study in Northern China. First conducted in the year 2001, the study was repeated in 2006 and 2011. The only eligibility criterion for inclusion in the study was age 40+years. The Medical Ethics Committee of the Beijing Tongren Hospital approved the study protocol and all participants gave informed consent. At baseline and follow-up examinations, all study participants underwent an interview with standardised questions and a complete ophthalmic examination. The study has been described in detail previously.12

Of the 4439 subjects participating in the Beijing Eye Study 2001, in 2006 a random sample of 1957 subjects underwent a complete evaluation of the ocular surface in the following order: dry eye symptom analysis, measurement of the tear-film break-up time (TBUT), assessment of fluorescein staining of the cornea, Schirmer test with anaesthesia and evaluation of meibomian gland dysfunction (MGD).13 The subjective symptoms of dry eye were evaluated using a simple questionnaire composed of three questions: ‘Do your eyes ever feel dry?’, ‘Do you ever feel a gritty or sandy sensation in your eyes?’ and ‘Do your eyes ever have a burning sensation?’. Possible answers to the questions were no (0), less than once a month (1), once or twice a week (2), at least once every day (3) or all the time (4). The presence of dry eye symptoms was defined as having one or more symptoms at least once every day (score of 3 or 4 for any question). A quantitative grading score of subjective dry eye symptoms was obtained by summing the answers to the different questions (score of 0–12). TBUT was measured by instilling fluorescein into the inferior cul-de-sac and calculating the average of three consecutive tear breakup times. Fluorescein staining of the cornea was graded 0 for no staining, 1 for mild staining with a few disseminated stains and limited to less than one-third of the cornea, 2 for moderate staining with a severity between grades 1 and 3, or 3 for severe staining with confluent stains and occupying half or more of the cornea. The Schirmer test was performed 1 min after instillation of a drop of 0.5% proparacaine, and after 5 min the amount of wetting (mm) was recorded. MGD was assessed by evaluating the presence of plugging and by expression of the meibomian glands. MGD was graded 0 for no obstruction, grade 1 for plugged with translucent serous secretion when the lid margin was compressed, grade 2 for plugged with viscous or waxy white secretion when the lid margin was compressed and grade 3 for plugged with no secretion when the lid margin was compressed.13 For each study participant, both eyes were tested and the data of the worse eye, based on the clinical judgment of the examiner, were used for analysis. DED was defined as the presence of dry eye symptoms (one or more symptoms of dry eye at least every day) associated with a Schirmer test <5 mm and/or BUT<10 s.14 During the 2011 survey, a simple question aiming to evaluate subjective symptoms of dry eye (‘Do your eyes ever feel dry?’) was also asked, and graded as previously described.

Depressive symptoms were evaluated in the survey conducted in 2011 using a Chinese depression scale adapted from the Zung self-rated depression scale.12 The total score of depression symptoms was converted to a 100-point scale. Definite depression was defined as having a depression score of 40 or higher. Cognitive function was also assessed using the Mini Mental Status Examination (MMSE) scale.12

Statistical analysis was performed using SPSS for Windows, V.20.0 (SPSS, Chicago, Illinois, USA). The data were given as mean values±SDs. The Gaussian distribution of the parameters was tested using the Kolmogorov–Smirnov test. Continuous variables were compared using the Student t test and the χ2 test was used to compare proportions. The association between variables was examined applying the Pearson correlation test. Multivariate logistic regression analysis was used to investigate the effects of various parameters on depression. ORs and 95% CIs were presented. All p values were two-sided and were considered statistically significant when the values were less than 0.05.


From the 2006 random sample (1957) that had an evaluation of the ocular surface, 1456 subjects (74.4%) underwent depressive symptom evaluation in 2011. The subjects evaluated in 2011 and the 2006 sample did not differ significantly in terms of gender (p=0.49), frequency of DED (p=0.58), best corrected visual acuity (BCVA) (p=0.12), rural versus urban region of habitation (p=0.66), level of education (p=0.25) and gross income (p=0.55). Moreover, for the 1456 participants, the individual scores of subjective dry eye symptoms obtained in 2006 were not different to the results obtained in 2011 (paired-samples t test, p=0.732).

Out of the 1456 participants examined in 2011, 241 (16.6%) subjects had DED and 138 (9.5%) subjects had definite depression. The mean age of participants was 64.8±9.5 years (range: 50–91) and 867 were women (59.5%). Patients with DED were older (p<0.001), more frequently women (p=0.002), more often living in an urban area (p<0.001), had a higher income (p<0.001) and had a lower BCVA (p<0.001) as compared to patients without DED (table 1). Patients with depression were more frequently women (p=0.009) and living in a rural region (p<0.001), had a lower level of education (p=0.005) and income (p<0.001), and a lower cognitive function score (p=0.007) as compared to patients without depression (table 1).

Table 1

Demographic characteristics of study subjects: depression and dry eye parameters

Definite depression was more prevalent in patients diagnosed as having DED than in subjects without DED (13.7±0.4% vs 8.6±0.3%, OR: 1.68; 95% CI 1.10 to 2.55; p=0.021) (table 1). The proportion of subjects with a definite depression was also significantly higher in the group of patients experiencing dry eye symptoms as compared to patients without symptoms (13.8±0.4% vs 8.3±0.3%; OR: 1.77; 95% CI 1.19 to 2.63; p=0.006). These results were similar when using the dry eye symptoms evaluated during the 2011 survey (13.5±0.3% vs 8.6±0.3% OR: 1.65; 95% CI 1.11 to 2.47; p=0.013). The mean depression score was also significantly higher in subjects with dry eye symptoms than in subjects without symptoms (32.0±7.7 vs 30.9±7.1, p=0.016, using the 2006 survey data; 32.3±7.2 vs 30.8±7.3, p=0.004, using the 2011 survey data). Correspondingly, patients with depression experienced more dry eye symptoms than patients without depression (p=0.02) (table 1) and the depression score was significantly correlated to the dry eye symptoms (r=0.07; p=0.013) (table 2). Nevertheless, the depression score was not significantly associated with the TBUT (p=0.18), the Schirmer test (p=0.37), corneal staining (p=0.30) or MGD evaluation (p=0.93). In a parallel manner, the dry eye symptoms were not correlated to any of the ocular surface objective parameters (table 2).

Table 2

Results of correlation between dry eye parameters and depression symptoms

In univariate analysis, definite depression was associated with dry eye symptoms (r=0.08; p=0.003), DED (r=0.06; p=0.014), BCVA (r=−0.07; p=0.046), female gender (r=0.07; p=0.012), gross income (r=−0.11; p<0.001), level of educational background (r=−0.08; p=0.003), rural residence (r=−0.09; p<0.001), cognitive evaluation (r=−0.1; p<0.001), body height (r=−0.07; p=0.004) and weight (r=−0.06; p=0.024), but not with age (r=−0.02; p=0.34), or marital status (r=−0.01; p=0.75) (table 3). The multivariate logistic regression analysis included presence or absence of depression as dependent parameter and the variables that were significantly associated with depression in the univariate analysis as independent parameters. After stepwise reduction of the list of independent parameters by dropping those parameters with a p value>0.05, depression was eventually associated with dry eye symptoms (p=0.028), lower cognitive status (p=0.01), rural residence (p=0.023) and lower body weight (p=0.05) (table 4).

Table 3

Univariate analysis of associations between definite depression, demographics and dry eye parameters

Table 4

Multivariate logistic regression analysis between definite depression, demographics and dry eye parameters


In an adult population from Beijing, depression was associated with DED and in particular with dry eye symptoms. These results are in accordance with previous retrospective epidemiological studies that showed a positive association between DED and depression.6–8 ,10 In a veterans population study, Galor et al,7 evaluated the relationship between DED and psychiatric disorders and observed that the presence of depression induced an approximately twofold increased risk of having DED. Similarly, Fernandez et al,10 recently reported more dry eye symptoms in veterans with depression and post-traumatic stress disorders. Nevertheless, these studies evaluated only veteran populations and did not use dimensional measures to evaluate depression symptoms and to categorise patients with depression. Kim et al11 evaluated DED and depression symptoms and their relation to objective tests on DED in a population-based cross-sectional study. As observed in the present study, these authors found a positive correlation between depression and dry eye symptoms. Similarly, there was no association between objective dry eye parameters (TBUT and Schirmer test) and the depression score.11 These results were also in agreement with two smaller-scale case-control studies showing that depression and anxiety scores were correlated to the Ocular Surface Disease Index (OSDI).2 ,9 In contrast, another study evaluating the burden of DED in daily life reported that mental health was unaffected by dry eye symptoms regardless of the severity level or diagnosis.4 A reason for the discrepancy between the studies may have been that the general health status questionnaire (Short Form 36 (SF-36)) used in that study was not specific for measuring the psychological status of the subjects.

Although an association has been detected between depression and DED, it may be more likely that the symptoms of DED and not DED itself were associated with depression or depression symptoms. There may be several possible explanations for this relationship. First, dry eye symptoms may induce or increase anxiety and depression symptoms. Despite complex pathophysiology, chronic pain negatively impacts multiple aspects of patient health including cognitive processes and mental health.5 Moreover, DED negatively affects the patient's everyday living and quality of life, as a result of chronic pain and also through visual performance alteration and the perception of visual function.2–4 ,15 The need for frequent instillations of eye drops can also affect social interactions.16 In a recent study, Pouyeh et al17 showed that the presence of ocular surface symptoms was negatively correlated to the performance of daily activities, the capacity to work as well as emotional well-being. These chronic impacts of dry eye symptoms on many components of daily life might contribute to the development of depression in patients with DED. Second, depression and its medications might induce or aggravate DED. Although antidepressant medications are known risk factors for DED,7 ,18 it has been shown that depression itself is involved in the pathophysiology of DED and not just its treatments .7 Despite common risk factors such as female gender or age, biological studies showed a dysregulation of neuropeptides and an increased production of inflammatory cytokines in patients with depression,19–21 which are also mechanisms involved in DED.7 ,11 If depression itself is involved in the pathophysiology of DED, one can argue that other clinical parameters evaluating the tear film such as tear film break-up time, Schirmer test and corneal staining will be associated with depression scores. However, results from the present and previous studies found no significant correlation between depression and the results of clinical objective parameters of DED.9–11 In a parallel manner, the results of dry eye symptoms evaluation were not significantly associated with DED objective parameters in our study and in previous other investigations.16 One may also take into account that perception of dry eye symptoms can be influenced by many factors including depression and anxiety.22 ,23 A depressive mood may aggravate dry eye symptoms or lower the threshold for perceiving ocular surface discomfort.24 Moreover, changes in the psychological status of patients could also modify the reported intensity of symptoms.

Our study has limitations that need to be considered when interpreting the results. First, the DED and depression evaluations were performed with a 5-year interval (DED evaluation in 2006 and depression evaluation in 2011). Although DED is a chronic disease, some patients may have experienced relief from their dry eye symptoms and others may have newly developed DED or its symptoms may have worsened over time. The individual dry eye symptoms evaluated in 2011 were however not different than those evaluated in 2011, so that this limitation of the study may not have markedly influenced the results and conclusions of our study. Second, dry eye symptoms were evaluated using a simple scoring system made up of three questions. Questionnaires such as the OSDI (with 12 questions) or the Dry Eye Questionnaire (with 21 questions) would have delivered a more detailed analysis of dry eye symptoms.14 Nevertheless, questionnaires using only three questions have also been used in previous large epidemiological studies such as the Women's Health Study.25 Third, the questionnaire used to evaluate depression symptoms was adapted from the Zung self-rating depression scale, but was not self-administered. Since some study participants could not answer the different questions alone; consequently, these questions were asked by trained technicians for all subjects. This may explain the relatively low value of depressions score observed in the present study.9 ,26 Finally, several parameters that may influence the relationship between dry eye and depression, such as the presence of concomitant ocular surface disorders or other systemic diseases, the use of systemic (in particular antidepressant medications) or topical medications were not systematically evaluated in our study.

In conclusion, our study found an association between dry eye symptoms and depression. Due to the study design, this investigation could not address whether depression increased dry eye symptoms or vice versa or whether both conditions influenced each other. Clinicians should be aware of the frequent association between depression and DED in order to better understand and better evaluate patient issues in DED.


View Abstract


  • Contributors All authors included in the paper fulfil the criteria of authorship. No one else who fulfils the criteria has been excluded as an author.

  • Competing interests AL has received an unrestricted grant from ‘La Fondation de France’.

  • Ethics approval The protocol of the study was approved by the Medical Ethics Committee of the Beijing Tongren Hospital.

  • Provenance and peer review Not commissioned; externally peer reviewed.