Abstract
Purpose.:
To evaluate whether veterans with posttraumatic stress disorder (PTSD) or depression have differences in dry eye symptoms and signs compared to a population without these conditions.
Methods.:
Male patients aged ≥50 years with normal eyelid, conjunctival, and corneal anatomy were recruited from the Miami Veterans Affairs Eye Clinic (N = 248). We compared dry eye symptoms (determined by the Dry Eye Questionnaire 5 [DEQ5] score) to tear film indicators obtained by clinical examination (i.e., tear osmolarity, corneal staining, tear breakup time, Schirmer's, meibomian gland quality, orifice plugging, lid vascularity) between patients with PTSD or depression and those without these conditions. Student's t-tests, χ2 analyses, and linear and logistic regressions were used to assess differences between the groups.
Results.:
DEQ5 scores were higher in the PTSD (mean = 13.4; standard error [SE] = 1.1; n = 22) and depression (mean = 12.0; SE = 0.8; n = 40) groups compared to the group without these conditions (mean = 9.8; SE = 0.4; n = 186; P < 0.01 and P = 0.02, respectively). More patients in the PTSD and depression groups had severe dry eye symptoms, defined as a DEQ5 score ≥ 12 (77% and 63% vs. 41%; P < 0.01 and P = 0.02, respectively). No significant differences in tear film indicators were found among the three groups. Multivariable logistic regression indicated that a PTSD diagnosis (odds ratio [OR] = 4.08; 95% confidence interval [CI] = 1.10–15.14) and use of selective serotonin reuptake inhibitors (OR = 2.66; 95% CI = 1.01–7.00) were significantly associated with severe symptoms.
Conclusions.:
Patients with PTSD have ocular surface symptoms that are not solely explained by tear indicators. Identifying underlying conditions associated with ocular discomfort is essential to better understand the mechanisms behind ocular pain in dry eye syndrome.
Dry eye syndrome (DES) is one of the most common ocular conditions in the United States, with approximately 1.7 million men aged 50 and older affected.
1,2 This number is projected to increase to over 2.8 million by 2030.
2 The most common symptoms of DES include irritated, gritty, scratchy, or burning eyes; sensation of something in the eyes; watery eyes; and blurred vision.
3 Because of these symptoms, DES has been shown to negatively impact one's quality of life and emotional well-being.
4 DES may also pose difficulties in daily tasks, such as reading, working, using a computer, or driving.
5 These physical, mental, and social limitations due to DES have ignited research regarding the associations between DES and specific mental health diagnoses.
Preliminary studies investigating the association between mental health and DES in Western populations have used administrative databases to define DES.
6,7 However, recent studies on Asian populations have used symptoms in addition to clinical indicators to measure DES, with mixed results. For example, Li et al. found that Chinese patients with depression and anxiety reported more dry eye symptoms, but did not differ in clinical measurements (i.e., Schirmer test 1, tear breakup time [TBUT], and corneal fluorescein staining) from a control group.
8 Similarly, Kim et al. documented that depression scores in a geriatric Korean population were correlated with dry eye symptoms, but not with TBUT or Schirmer test score.
9 Conversely, Wen et al. reported a greater prevalence of dry eye symptoms and clinical measurements (i.e., Schirmer test 1, TBUT, and corneal fluorescein staining) in Chinese individuals with depression and anxiety disorders.
10
To the authors' knowledge, no studies to date have investigated the relationship between DES, as defined by objective measures, and depression or anxiety in a Western population. More specifically, the association between DES and posttraumatic stress disorder (PTSD) has never been examined. PTSD is defined briefly as experiencing the following symptoms, lasting for at least 1 month: experience of a traumatic event; intrusive recollection of the event; avoidance/numbing; and hyperarousal.
11 In addition, significant emotional distress and functional impairment must be present. Even though PTSD is categorized as an “Anxiety Disorder” in the
Diagnostic and Statistical Manual for Mental Disorders, it is a unique disorder with regard to its clinical symptomatology, presentation, and treatment protocol.
12 Furthermore, there is scant research evaluating potential links between psychiatric disorders and DES.
6 As both PTSD and depression have been associated with systemic inflammation,
13,14 and DES with local inflammation,
15–17 it is possible that inflammation may serve as a link between these entities.
Veterans are a unique population with distinctive physical and psychiatric comorbidities, including a higher prevalence of PTSD and depression compared to the general population.
18–20 Using administrative databases, we found that male and female US veterans had a DES prevalence of 19% and 22%, respectively, and those with PTSD or depression had a higher likelihood of having a dry eye diagnosis.
6,7 The objective of the current study was to augment the existing DES literature by evaluating associations between a symptoms-based DES diagnosis and objective parameters of DES among male veterans with and without PTSD or a depression diagnosis. Additionally, we explored whether elevated C-reactive protein (CRP) levels were associated with depression or PTSD among male veterans with and without DES to study whether systemic inflammation could be a possible link between psychiatric disorders and DES.
The Miami Veterans Affairs (VA) Institutional Review Board reviewed and approved the prospective examination of patients for this study, which was conducted in accordance with the principles of the Declaration of Helsinki. Patients were recruited from the Miami VA Eye Clinic (October 2010–December 2011). Patients were seen in the eye clinic by an ophthalmologist or optometrist for a variety of concerns, including refractive issues, cataract evaluation, and retinal pathology. Inclusion criteria included having normal eyelid, conjunctival, and corneal anatomy. Patients were not eligible to participate if they were female; were under 50 years of age; used contact lenses; used any ocular medication (with the exception of artificial tears/topical cyclosporine); or had human immunodeficiency virus, sarcoidosis, graft-versus-host disease, a collagen vascular disease, an active external ocular process (e.g., keratitis), or any history of refractive surgery or cataract surgery within the last 3 months. Patients were prescreened by eye care practitioners, and eligible subjects were informed about an opportunity to participate in a 1-day research study with the purpose of evaluating tear film function. Potential subjects were told that the goal of the study was to understand why some individuals had tear dysfunction while others had healthy tears. This script was drafted specifically to mitigate possible selection bias and recruit patients with and without DES. Interested patients were scheduled for a research visit, at which time informed consent was obtained.
The sample was divided into three groups: depression only, PTSD only, and those without either diagnosis. We excluded participants who had a diagnosis of both depression and PTSD (n = 15) in order to study each diagnosis separately.
For each individual, demographic information, medical history, and psychiatric and medication information were collected. Demographic information, health status, medical history, psychiatric diagnoses (e.g., PTSD and depression), and medication information were gathered via a chart review of the VA administrative database. Specifically, lifetime and current PTSD and depression diagnoses were coded using the International Classification of Disease-9 (ICD-9) by the patient's treating physician. Current antidepressant and antianxiety medication use was categorized by drug category. Antidepressants were categorized as selective serotonin reuptake inhibitors (SSRIs), other, or none. This categorization was used due to small sample sizes in the non-SSRI categories; antidepressant medications included in the “other” group included serotonin and norepinephrine reuptake inhibitors (SNRIs), aminoketones, and piperazinoazepines. Even though SSRIs and SNRIs are similar in their inhibition of reuptake of serotonin, SNRIs are a distinct therapeutic class because these agents also block the reuptake of norepinephrine, and were thus classified in the “other” group. Similarly, antianxiety medication use was categorized as benzodiazepines, other, and none. Other antianxiety medications included atypical anxiolytics and antihistamines (i.e., only when used for anxiety treatment purposes).
The validated five-item Dry Eye Questionnaire (DEQ5) was administered to all patients.
21 The DEQ5 comprises questions regarding the frequency and intensity of dry eye symptoms, such as eye discomfort, eye dryness, and watery eyes. The total score ranges from 0 to 22, with a score of zero indicating no dry eye symptoms and a score of 22 indicating that the subject experiences the most frequent and intense symptoms. Based on previous validation studies, the presence of mild symptoms was defined as a score ≥ 6, and the presence of severe symptoms score was defined as a score ≥ 12.
21
The ocular surface examination, in the order performed, consisted of tear osmolarity (measured once in each eye; TearLAB, San Diego, CA), TBUT (measured twice in each eye and averaged per eye), conjunctival and corneal staining (punctuate epithelial erosions [PEE]; range, 0–5),
22,23 Schirmer's strips with anesthesia,
23 and morphologic and qualitative eyelid and meibomian gland information. For the TBUT, a fluorescein strip (Fluorets; Laboratoire Chauvin, Aubenas, France) was wetted with the application of one drop of nonpreserved saline to the lower one-fourth of the strip. Excess fluid was gently removed such that the saturated tip delivered approximately 3 to 5 μL liquid sodium fluoride (NaFl). With the patient gazing up, the examiner introduced the NaFl into the lower fornix. Starting with the right eye, timing was stopped upon visualization of the first break (one or more black [dry] spots) appearing in the precorneal tear film or after 15 seconds had elapsed. The procedure was then repeated for the left eye. Morphologic information collected included the degree of eyelid vascularity (0 = none; 1 = mild engorgement; 2 = moderate engorgement; 3 = severe engorgement)
24 and the presence of inferior eyelid meibomian orifice plugging (0 = none; 1 = less than one-third lid involvement; 2 = between one-third and two-thirds involvement; 3 = greater than two-thirds lid involvement). Meibum quality was graded on a scale of 0 to 4 (0 = clear; 1 = cloudy; 2 = granular; 3 = toothpaste; 4 = no meibum extracted).
25 For each participant, data from the worst eye were used.
Blood samples were obtained from each patient and analyzed for plasma concentrations of high-sensitivity CRP. For the CRP analysis, we excluded participants with CRP levels greater than 10 (
n = 22; 9%), as this level indicates an acute rather than a chronic inflammatory process.
26
The main outcome measure was the comparison of dry eye symptoms, as measured by the DEQ5, and clinical signs between those with PTSD or depression and a group without these comorbidities.
Statistical analyses were performed using SPSS 19.0 (SPSS, Inc., Chicago, IL) and SAS 9.3 (SAS Institute, Inc., Cary, NC). T-tests and χ2 analyses were used to characterize the demographic data. Linear models (generalized for binary data and general for continuous data) were used to analyze the tear film indicators obtained by clinical examination. Logistic regression analysis was used to evaluate demographic, medical, and psychiatric factors associated with severe symptoms. Statistical significance was defined as P < 0.05.
A total of 248 veteran men were included in this study. The mean age was 69 years old, with the majority of the sample being white (70%,
n = 172) non-Hispanic (73%,
n = 180) former smokers (55%,
n = 126) and in self-reported good health (56%,
n = 137). Demographic characteristics of the study population are reported in
Table 1.
The PTSD (mean = 13.4, standard error [SE] = 1.1) and depression (mean = 12.0, SE = 0.8) groups had significantly higher mean DEQ5 scores compared to the group without these diagnoses (mean = 9.8, SE = 0.4;
P < 0.01 and
P = 0.02, respectively). These two groups also had a higher percent of severe symptoms (as measured by a DEQ score ≥ 12) compared to the group without these diagnoses (
P < 0.01 and
P = 0.02, respectively). Results are reported in
Table 2.
Linear models (generalized for binary data and general for continuous data) were used to analyze the tear film indicators obtained by clinical examination. There were no significant differences among groups with respect to tear osmolarity, corneal staining, TBUT, meibomian gland quality, lid vascularity, or meibomian gland orifice plugging in either mean scores or cutoff definitions (
Table 2). Several different cutoffs for tear osmolarity were evaluated (308, 312, 325), but were not found to influence the results (results shown only for 325).
27
After controlling for age, race, ethnicity, self-reported smoking status, self-reported health status, current use of antidepressants, current use of antianxiety medications, CRP, and psychiatric diagnosis (e.g., PTSD or depression), PTSD diagnosis (odds ratio [OR] = 4.02; 95% confidence interval [CI] = 1.10–15.14), SSRI use (OR = 2.66; 95% CI = 1.01–7.00), and never smoker (versus current smoker; OR = 3.11; 95% CI = 1.11–8.75) remained significant predictors of a symptoms-based DES diagnosis. Results are illustrated in
Table 3.
No significant differences in CRP levels were found between the group without any psychiatric diagnoses (mean = 2.8; SE = 0.2) and the depression group (mean = 2.5; SE = 0.4;
P = 0.45) or the PTSD group (mean = 2.4; SE = 0.5;
P = 0.45), nor was there a significant difference between the PTSD and depression groups (
P = 0.88). The inclusion of CRP levels in the models shown in
Table 2 did not alter the significance of the results (data not shown). Additionally, CRP levels were not significantly associated with severe dry eye symptoms in our multivariable model (
Table 3).
Our results demonstrate that within our population, a diagnosis of PTSD or depression was associated with DES severity when measured using a symptom-based instrument. However, measurements of objective tear film parameters could not explain the degree of symptoms. These findings are similar to those reported by Li et al.
8 and Kim et al.,
9 who also found no differences in tear parameters in those with depression or anxiety. Understanding mechanisms behind ocular discomfort is a vital first step in alleviating morbidity for DES suffers. While abnormal tear film indicators are important in the pathway of ocular pain, our study and others
28,29 confirm that other factors must be important as well, based on the poor correlations between the symptoms and signs of DES.
30 Our study suggests that having a psychiatric disorder independently affects the symptoms of DES. However, while we postulated that inflammation may be a potential link between the two entities, we were not able to demonstrate differences in CRP levels between the groups.
There are several potential reasons that may explain why veterans with depression report more dry eye symptoms, but do not have measurable tear film disturbances greater than those of veterans without depression. First, it has been well established that depression and pain often coexist.
31 For example, the Medical Outcomes Study found that patients with depression tended to have greater bodily pain, worse physical functioning, and worse perceived current health compared to patients without depression.
31 Similarly, Katon et al. found that those with depression or anxiety, in addition to a chronic medical illness, reported significantly higher numbers of medical symptoms compared to those with a medical illness only.
32 In addition, studies using a variable pressure dolorimeter found that individuals who are depressed have the same pain sensitivity as control subjects but report higher levels of pain.
33 However, it is important to note that in our multivariable model, depression did not remain an independent predictor of dry eye symptoms.
Depression and PTSD share many similarities, and the above mechanisms may play a role in the relationship between PTSD and ocular surface symptoms.
11,31–40 However, individuals with PTSD have unique attributes, such as amplified emotional reactions and arousal and increased attention when faced with pain-related or threatening stimuli,
11,41 which may lead to an increased focus on pain, sensitivity to painful stimuli, and amplification of the pain experience.
42,43
In our study, even though we did not find a difference in CRP levels between the three groups, inflammation is another potential mechanism that may link PTSD and depression to DES symptoms. Patients with PTSD and depression are known to have higher levels of inflammatory markers in the blood, including increased CD4:CD8 ratios and CRP, Intercellular Adhesion Molecule 1 (ICAM-1), TNFα, IL-1, and IL-6 levels.
13,14,44–46 Patients with DES have been shown to have elevated levels of these markers in the tears and conjunctivae, including infiltration of CD4+ lymphocytes into the conjunctival epithelium
15–17 and increased levels of IL-1β, IL-2, IL-6, IL-8, IFN-γ, and TNF-α in the tears and conjunctival epithelium.
47–51 It is not known, however, whether a higher burden of systemic inflammatory cells and soluble mediators translate into higher symptoms on the ocular surface. Given our negative findings regarding CRP levels, futures studies can consider exploring other proteins such as IL-1β or IL-6 as potential links between DES and mental illness.
Consistent with prior studies, SSRI use was significantly associated with DES.
10,52 The association between antidepressants and ocular discomfort is not clear, but there are a few potential explanations. The anticholinergic side effects of antidepressants, especially SSRIs, have long been acknowledged to exert ocular reactions, including dry eye symptoms.
10,53–56 In addition, serotonin receptors have been identified in the corneal and conjunctival epithelium,
17,57 and serotonin has been previously isolated from human tears.
58 It is possible that altered levels of serotonin due to antidepressant treatment can affect the sensitivity thresholds of corneal nerves, given that serotonin has been found to the affect the sensitivity, threshold, and function of peripheral nerves.
59,60
Inconsistent with previous cross-sectional epidemiologic studies, we found that participants who had never smoked reported more dry eye symptoms compared to current smokers.
61–63 However, two longitudinal incidence studies found no association between smoking and DES,
52,64 whereas one prospective case series found that smokers had more damage to their lipid layer of precorneal tear film compared to nonsmokers.
65 Therefore, it is possible that temporality issues intrinsic to cross-sectional designs, including the present study, may play a role in the conflicting results. However, it is also possible that nonsmokers are more health conscious, thus may be more aware of or more prone to report physical health symptoms. Future studies should utilize prospective studies and objective measurements, such as cotinine, to validate self-reported smoking status.
A major limitation of this study is that we did not use clinical interviews or other validated dimensional measures to categorize patients as having depression or PTSD, but instead relied upon documentation of these conditions in the medical record via ICD-9 codes. Due to the nature of this retrospective chart review, it is uncertain whether these diagnoses originally came from a clinician (e.g., psychiatrist) interview or a standardized measure. Future studies should use validated measures to assess for current mental illness. Similarly, as pharmacy information was collected retrospectively, we did not have detailed information about the length of time on medication to examine its dose–response effects on DES symptoms. Because we grouped patients who were taking psychotropic medications due to small sample sizes, we were unable to examine the effects of medications besides SSRIs and benzodiazepines on dry eye symptoms. Furthermore, the stability of the DES measurements could not be assessed in this study, as tear film indicators obtained by clinical examination were measured at only one time point.
Despite these limitations, this is the first study to confirm that while a symptom-based DES diagnosis was more prevalent in veterans with PTSD and depression seen in an eye clinic, clinical tear film measurements were not notably different from those without these diagnoses. This suggests that the etiology of DES symptoms extends beyond tear dysfunction. Since ocular discomfort is the main cause of DES morbidity, future studies using validated metrics for systemic abnormalities are needed to examine additional factors that drive ocular pain (e.g., local and/or systemic nerve dysfunction). This information may provide new directions in the evaluation and treatment of DES. In the meantime, given the known morbidity of DES symptoms with respect to physical and mental well-being, the presence of ocular surface discomfort should be evaluated in veterans with depression and PTSD. This is especially warranted due to the new influx of veterans returning from the war in Afghanistan, the Middle East, and elsewhere.
We thank Bozorgmehr Pouyeh, Gail Lewis, and Ashley Katsikos for their assistance with the implementation of this study.
Supported by a grant from the Veterans Affairs Medical Center (AG) and with unrestricted funds from Research to Prevent Blindness, New York, New York.
Disclosure: C.A. Fernandez, None; A. Galor, None; K.L. Arheart, None; D.L. Musselman, None; V.D. Venincasa, None; H.J. Florez, None; D.J. Lee, None