Vehof J, Kozareva D, Hysi PG, Harris J, Nessa A, Williams FK, Bennett DLH, McMahon SB, Fahy SJ, Direk K, Spector TD, Hammond CJ. Relationship Between Dry Eye Symptoms and Pain Sensitivity. JAMA Ophthalmol. 2013;131(10):1304-1308. doi:10.1001/jamaophthalmol.2013.4399
Dry eye disease (DED) is common, but little is known about factors contributing to symptoms of dry eye, given the poor correlation between these symptoms and objective signs at the ocular surface.
To explore whether pain sensitivity plays a role in patients’ experience of DED symptoms.
Design, Setting, and Participants
A population-based cross-sectional study of 1635 female twin volunteers, aged 20 to 83 years, from the TwinsUK adult registry.
Main Outcomes and Measures
Dry eye disease was diagnosed if participants had at least 1 of the following: (1) a diagnosis of DED by a clinician, (2) the prescription of artificial tears, and/or (3) symptoms of dry eyes for at least 3 months. A subset of 689 women completed the Ocular Surface Disease Index (OSDI) questionnaire. Quantitative sensory testing using heat stimulus on the forearm was used to assess pain sensitivity (heat pain threshold [HPT]) and pain tolerance (heat pain suprathreshold [HPST]).
Of the 1622 participants included, 438 (27.0%) were categorized as having DED. Women with DED showed a significantly lower HPT (P = .03) and HPST (P = .003)—and hence had higher pain sensitivity—than those without DED. A strong significant association between the presence of pain symptoms on the OSDI and the HPT and HPST was found (P = .008 for the HPT and P = .003 for the HPST). In addition, participants with an HPT below the median had DED pain symptoms almost twice as often as those with an HPT above the median (31.2% vs 20.5%; odds ratio, 1.76; 95% CI, 1.15-2.71; P = .01).
Conclusions and Relevance
High pain sensitivity and low pain tolerance are associated with symptoms of DED, adding to previous associations of the severity of tear insufficiency, cell damage, and psychological factors. Management of DED symptoms is complex, and physicians need to consider the holistic picture, rather than simply treating ocular signs.
Dry eye disease (DED) is a multifactorial disease of the tears that can potentially damage the ocular surface.1 It causes considerable morbidity, with symptoms of pain, irritation, and interference with normal visual tasks.2 Dry eye disease is a common problem, affecting 5% to 35% of middle-aged and elderly individuals, depending on the definition used.3- 7 Women are at increased risk compared with men.5,6,8,9
Studies of DED and the management of patients with DED in practice are complicated by the lack of an association with objective diagnostic signs of the ocular surface.10,11 For example, patients without any signs of DED may have symptoms, while others with extensive signs of DED may have few symptoms. Quiz Ref IDSo, besides the integrity of the optical tear layer, the corneal pain system has been proposed as a central pathogenic feature of dry eye.12,13 Moreover, the cornea is one of the most sensitive parts of the human body, since it is more densely innervated with sensory nerve fibers than any other organ.14- 16
We hypothesized that (generalized) pain sensitivity might play an important role in the experience of DED symptoms. Therefore, we investigated the association between DED symptoms and pain sensitivity (the heat pain threshold [HPT]) and pain tolerance (the heat pain suprathreshold [HPST]) in a large group of healthy volunteers.
The local Research Ethics Committee at St Thomas’ Hospital approved all procedures, and all participants provided informed consent. The experiment was conducted on 1635 white female volunteers from 1025 families (610 complete twin pairs and 415 unrelated twin participants), aged 20 to 83 years, from the TwinsUK adult registry of identical and nonidentical twins at St Thomas’ Hospital, London. This cohort of mostly female twins has been recruited from the general population via successive local and national media campaigns.17 The TwinsUK cohort is, we believe, representative of the general population and has been comparable to a UK singleton female cohort for a wide range of diseases and traits.18 It has been used in many epidemiologic studies of healthy aging.19 Although all participants were part of a twin pair, we did not make use of the twin relationship in this study. All participants completed extensive questionnaires gathering demographic information, clinical history, and current medications. Participants were asked not to take analgesic medication within 12 hours of the study visit, if possible, and were excluded if they did. Individuals with impaired arm function such as that caused by repetitive strain injury, neuropathy, or chemotherapy were also excluded.
All 1635 participants were asked the following 3 questions as a proxy for having DED, which have been used separately in other population-based epidemiologic studies5,20: (1) “Have you ever been diagnosed (by a clinician) as having dry eye syndrome?” (2) “Do you currently use artificial tear eyedrops or gel?” and (3) “For the past three months or longer, have you had dry eyes? (This is described as a foreign body sensation with itching and burning, sandy feeling, not related to allergy).” If a participant answered yes to any of these questions, she was assigned as having DED.
In addition, to investigate pain symptoms in further detail, we asked a consecutive subset of 689 participants (from 394 families) attending for quantitative sensory testing to complete the Ocular Surface Disease Index (OSDI). The OSDI, developed by the Outcomes Research Group at Allergan, Inc, is a 12-item questionnaire designed to provide a rapid assessment of the symptoms of ocular irritation consistent with DED and their effect on vision-related functioning.21 The presence of symptoms during the past week was rated on a 5-point scale from none of the time to all of the time. The OSDI score (range, 0-100) can be calculated with a formula using the sum score of all questions. An OSDI score of 15 or higher was used to define those with DED.21,22 Since this study is about pain sensitivity, we were particularly interested in the presence of pain symptoms. The first 3 questions of the OSDI investigate pain and discomfort symptoms by asking whether participants have experienced (1) eyes that are sensitive to light, (2) eyes that feel gritty, and (3) painful or sore eyes. Other questions relate to problems with vision and limitation of performance and/or discomfort in certain situations such as watching television and windy conditions. Pain symptoms of DED were evaluated by scoring these first 3 questions of the OSDI. An individual with a sum score of 3 or higher on those questions was regarded as having DED pain symptoms. Those who scored 0 on all these questions were regarded as having no DED pain symptoms. In addition, this subset of participants also was asked whether they had been through menopause and, if yes, if they had used hormone therapy.
Quantitative sensory testing using heat as the stimulus was used to capture the variation in participants’ perception of pain (thus, endogenous pain sensitivity) by measuring pain threshold and tolerance.23 The volar surface of the forearm was inspected for possible confounding factors (eg, cuts, bruises, burns, or skin irritation). The participant, while seated, placed her arm on the table in front of her. A 25 × 50-mm probe connected to a Modular Sensory Analyzer (Somedic) was placed on the volar surface of the upper central forearm and secured gently with a fabric-covered band. The starting arm for quantitative sensory testing was alternated, depending on the day of the week, and recorded. The HPT, a measure of pain sensitivity, was measured by slowly heating the probe from an adaptation temperature of 32°C at a rate of 0.5°C per second until the participant perceived the stimulus as changing from hot to painful and stopped the experiment by pressing a button, at which point the temperature (the HPT) was logged, with the probe temperature then quickly returning to 32°C. If the probe reached 50°C, the machine automatically returned to the adaptation temperature to prevent thermal burning. All participants were given standardized instructions before having a practice run at measuring the HPT. Following the HPT assessment, the probe was removed and replaced at the same position on the opposite arm. The HPST, a measure of pain tolerance, was ascertained by heating the probe from an adaptation temperature of 32°C at a rate of 1°C per second until the participant perceived the stimulus as changing from painful to unbearable and stopped the experiment by pressing a button. At this point, the temperature (the HPST) was logged, and the probe temperature quickly returned to 32°C. If the probe reached 50°C, the machine automatically returned to the adaptation temperature to prevent thermal burning. All participants were given standardized instructions. The HPST was assessed without a practice run.
Data were analyzed with the SPSS statistical package (version 17.0; SPSS, Inc). The Kolmogorov-Smirnov test was used to analyze the distribution normality of the variables (P > .05). First, the demographics of the study population and the prevalence of the different DED criteria were calculated. Participants were divided into having high (below median value HPT) or low (above median value HPT) pain sensitivity. We analyzed the relationship of potential confounders—age and menopausal status with DED and high pain sensitivity—by using logistic regression. Next, mean HPT and HPST values were calculated per group using several DED criteria. To compare differences in the HPT and HPST between participants with and without DED, we used the Mann-Whitney test. Finally, logistic regression was used to analyze whether women with high pain sensitivity had more DED pain symptoms than those with low pain sensitivity. These results were corrected for potential confounders. P < .05 was considered statistically significant in all analyses.
The median age of the 1635 participants was 60 years (first and third quartiles, 54 and 66 years, respectively). All participants were female. Depending on the question, up to 1.6% of participants answered “not sure” and were excluded from the analysis of that item from the questionnaire. Of the included participants, 438 (27.0%) were categorized as having DED from questionnaire answers, with 354 (21.6%) having dry eye symptoms in the past 3 months, 260 (16.2%) using artificial tears, and 218 (13.2%) being diagnosed with DED by a physician.
Of the consecutive subset of 689 participants (median age, 60 years; first and third quartiles, 52 and 65 years, respectively) who completed the OSDI questionnaire, 217 (31.6%) were diagnosed with DED by having an OSDI score higher than 15. In total, 13.7% were diagnosed by both the OSDI and the 3 questions, 17.8% by the OSDI but not by any of the 3 DED questions, and 14.1% by any of the 3 DED questions but not by the OSDI. Of the participants, 118 (17.1%) had an OSDI pain sum score of 3 or higher, while 338 (48.8%) had no DED pain symptoms. Although age was associated with having DED symptoms in the past 3 months (odds ratio [OR], 1.014 per year; P = .03, whole sample), it was not associated with having DED pain symptoms on the OSDI (OR, 1.001 per year; P = .94, random subset). Menopausal status or hormone therapy showed no statistically significant association with DED pain symptoms in the subset.
The mean (SD) HPT and HPST of all participants were 45.4°C (2.5°C) and 47.1°C (1.8°C), respectively. The median HPT was 46.0°C, so those with a lower value were regarded as having high pain sensitivity. Quiz Ref IDAge showed a strong association with high pain sensitivity (OR, 1.03 per year; P < .001). Menopausal status showed a statistically significant association with high pain sensitivity as well (OR, 1.55 if postmenopausal; P = .02), which was independent of age (OR, 1.72; P = .04, when corrected for age). The use of hormone therapy did not alter the association between menopause and pain sensitivity. The mean HPTs among those with and without DED, using several criteria, are presented in the Table. No significant differences in the HPT or HPST were found between participants with and without a diagnosis of DED by a physician or those who did and did not use artificial tears. However, a significantly lower HPST was observed in women who had dry eye symptoms in the past 3 months compared with those who had not (P = .01). Combining all 3 DED questions, persons with DED showed a significantly lower HPT (P = .03) and HPST (P = .003) than those without DED.
In the random subset that completed the OSDI questionnaire, the HPT and HPST were lower in women who reached the cutoff OSDI score of 15 for DED than in those who did not, although this difference was not significant in this smaller sample size (P = .12 and P = .15, respectively; Table). A strong significant association between the presence of pain symptoms and high pain sensitivity was found (P = .008 for the HPT and P = .003 for the HPST). In addition, participants with a high pain sensitivity had DED pain symptoms approximately 75% more often than those with a low pain sensitivity (31.2% vs 20.5%; OR, 1.76; 95% CI, 1.15-2.71; P = .01). The strength of the association increased after correction for age as a possible confounder (OR, 1.95; 95% CI, 1.24-3.06; P = .004). Including menopausal status as well in this regression model did not alter the association (OR, 1.89; 95% CI, 1.20-2.98; P = .006).
The current data provide the first empirical evidence that individuals with DED diagnosed by population-based questionnaires show altered pain sensitivity. Specifically, this study demonstrates that those with DED pain and discomfort symptoms have a lower pain threshold and pain tolerance of a heat-based stimulus than those without DED. Similarly, participants with HPTs below the median were almost twice as likely to report DED pain symptoms as those above the median. These findings support the hypothesis that a subset of persons with DED is more sensitive to pain.
In future studies, it would be interesting to investigate whether patients without signs but with symptoms of DED particularly show higher pain sensitivity compared with those with both signs and symptoms, or whether minor signs of DED provoke pain in patients with low pain thresholds. Similarly, given the generally poor relationship between symptoms and signs in DED, it would be useful to examine whether any particular clinical findings in DED populations (such as corneal staining, reduced tear breakup time, or reduced Schirmer values) relate to pain sensitivity and pain symptoms.
Pain is a complex, multidimensional sensory experience that is difficult to represent as a single parameter. Quantitative sensory testing has been applied to assess pain, since it is a sensitive and quantitative tool, but the stimulus modalities applied and perceptual responses vary widely.24 Our group elected to use the HPT tests because we previously have shown them to be reliable and heritable.25 More recently, we have found that the HPST is more reliable and reproducible than the HPT in our cohort (J.H., A.N., F.K.W., D.H.L.B., and S.B.M., unpublished data, 2007), and indeed the present study has discovered stronger associations using this measure than the HPT (Table). Results vary with HPT tests in other chronic pain conditions: some studies on chronic back pain report higher and others report lower pain thresholds, which may reflect differing testing protocols, definitions, and patient populations.26 Decreased heat pain detection threshold (ie, increased pain sensitivity) has been reported in burning mouth syndrome, another chronic pain condition that may reflect subclinical neuropathic pain or central pain.27
Quiz Ref IDConsistent evidence demonstrates that psychological factors influence the perception of pain and pain sensitivity, exerting a significant influence on the development and maintenance of chronic pain conditions.28,29 Similarly, anxiety and depression are more common among patients with DED,30 and while this may be secondary to reduced quality of life caused by the symptoms of disease, psychological distress may be associated with increased sensitivity and more symptoms. Moreover, Kim et al31 recently have shown that depression is associated with DED only in patients who have a normal to mildly reduced tear production. This association between depression and DED was not found in patients with a Schirmer score of 5 mm or lower. Patients with a chronic pain condition also appear to have morphometric changes on brain magnetic resonance imaging, particularly in the cingulate cortex, insula, and temporal lobe, which all deal with supraspinal nociceptive processing.32 It remains to be proven whether this is a cause or a consequence of the pain, but hope remains that some way of modulating brain processing may break the cycle of chronic pain.
It has been shown in human experiments that selective stimulation of corneal cold receptors by small decreases in temperature leads to awareness of cooling sensations that become increasingly unpleasant with lower temperature.33,34 If the same pain pathways are involved in the skin nerves, increased sensitivity of the skin could be associated with increased sensitivity of the corneal nerves, which provides a direct explanation for our results. However, we used increasing instead of decreasing temperature, but similar mechanisms might be involved. Belmonte35 first proposed that altered expression of membrane ion channels in the cornea could give rise to abnormal spontaneous and stimulus-evoked nerve firing in damaged nerves after photorefractive surgery. This study also speculated that these abnormal sensory discharges are read by the brain as ocular surface dryness. It may be that healthy variation in expression of these ion channels may cause variable pain sensitivity and symptoms of dry eye.
With regard to pain sensitivity, some authors have found corneal hypersensitivity in DED,36,37 whereas others have found hyposensitivity of the cornea.38- 41 It has been suggested that tear hyperosmolarity and its inflammatory consequences lead to epithelial injury in the cornea, which subsequently stimulates corneal nerve endings, leading to an increased sensory drive from the ocular surface.42 The reasons for these paradoxical results of hypersensitivity and hyposensitivity of the cornea in DED remain to be fully explained. Longitudinal studies may help in the future.42 For reasons of clarity, the present study is different from those studies, since we measured a “generalized” pain sensitivity via arm HPT and not (changes in) sensitivity in the affected organ (ie, the cornea), as done in the DED studies cited.
The TwinsUK cohort was chosen to study DED because it is an extensively phenotyped and genotyped large national volunteer cohort of perimenopausal women with an expected high prevalence of DED. While the results of this study are based on a twin cohort, random selection of 1 individual within each twin pair did not alter the effect size of the results (not shown). The age-related rise in the prevalence of DED in our cohort using questionnaire-based diagnostic criteria is similar to that found in epidemiologic studies of women of a similar age.9 As shown in our results, DED prevalence is highly variable when different criteria and questions are used. Finally, the results may not necessarily be generalizable to men.
Quiz Ref IDThis study has demonstrated that high pain sensitivity and low pain tolerance are associated with pain symptoms in DED, adding to previous associations among the severity of tear insufficiency, cell damage, and psychological factors. Management of DED symptoms is complex, and physicians need to consider the holistic picture, rather than simply treating ocular signs.
Corresponding Author: Jelle Vehof, PhD, Department of Ophthalmology & Epidemiology, University Medical Center Groningen, Hanzeplein 1, Postbus 30.001, Groningen, the Netherlands (email@example.com)
Submitted for Publication: January 21, 2013; final revision received March 11, 2013; accepted March 19, 2013.
Published Online: August 1, 2013. doi:10.1001/jamaophthalmol.2013.4399.
Author Contributions: Drs Vehof, Spector, and Hammond had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Vehof, Harris, Williams, McMahon, Spector, Hammond.
Acquisition of data: Vehof, Kozareva, Harris, Nessa, Williams, Fahy, Direk.
Analysis and interpretation of data: Vehof, Hysi, Bennett, McMahon, Hammond.
Drafting of the manuscript: Vehof, Harris, Fahy, Direk, Hammond.
Critical revision of the manuscript for important intellectual content: Vehof, Kozareva, Hysi, Nessa, Williams, Bennett, McMahon, Fahy, Spector.
Statistical analysis: Vehof, Hysi, Hammond.
Obtained funding: McMahon, Spector, Hammond.
Administrative, technical, and material support: Vehof, Kozareva, Harris, Nessa, Williams, Hammond.
Study supervision: Vehof, Williams, McMahon, Spector, Hammond.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported by the Wellcome Trust, the Department of Health via the National Institute for Health Research (NIHR) Clinical Research Facility at Guy’s & St Thomas’ NHS Foundation Trust, NIHR Biomedical Research Centre based at Guy’s and St Thomas’ NHS Foundation Trust and King’s College London, and Pfizer Pharmaceuticals, which paid for collection of pain phenotypes via an unrestricted grant. Dr Hammond is an NIHR Senior Research Fellow and Dr Spector is an NIHR Senior Investigator.