eTable 1. Distribution of Quantitative Sensory Testing (QST) Metrics Tested on the Right Forearm in the Study Population
eTable 2. Associations Between Demographics and Quantitative Sensory Testing (QST) Metrics
eTable 3. Associations Between Mental Health Indices and Medications With Quantitative Sensory Testing (QST) Metrics
Customize your JAMA Network experience by selecting one or more topics from the list below.
Galor A, Levitt RC, McManus KT, et al. Assessment of Somatosensory Function in Patients With Idiopathic Dry Eye Symptoms. JAMA Ophthalmol. 2016;134(11):1290–1298. doi:10.1001/jamaophthalmol.2016.3642
Do patients with dry eye (DE) and ocular pain symptoms have increased sensitivity to stimuli outside the trigeminal system, including measures specific for central sensitization?
Data from a prospective, cross-sectional study demonstrate that individuals with neuropathic-like DE pain symptoms have increased pain sensitivity at a site remote from the eye (forearm). This increased sensitivity included enhanced temporal summation, which is indicative of central sensitization.
The findings of this study suggest that DE symptoms are not only manifestations of a local disorder but also involve somatosensory dysfunction beyond the trigeminal system.
Somatosensory dysfunction likely underlies dry eye (DE) symptoms in many individuals yet remains an understudied component of the disease. Its presence has important diagnostic and therapeutic implications.
To assess the integrity of nociceptive system processes in persons with DE and ocular pain using quantitative sensory testing (QST) techniques applied at a site remote from the eye.
Design, Setting, and Participants
A cross-sectional study conducted at Miami Veterans Affairs Hospital included 118 individuals with a wide variety of DE symptoms and signs. The study was conducted from October 31, 2013, to January 28, 2016.
Individuals completed questionnaires regarding ocular symptoms (5-Item Dry Eye Questionnaire [DEQ5], Ocular Surface Disease Index [OSDI], and Neuropathic Pain Symptom Inventory modified for the eye [NPSI-E]), psychological status, and medication use and underwent an ocular surface examination. The QST metrics included measures of vibratory and thermal thresholds and cold and hot pain temporal summation (surrogate measures of central sensitization) on the forearm.
Main Outcomes and Measures
Correlations among DE and ocular pain symptom severity with QST metrics measured on the forearm. The OSDI score ranges from 0 to 100, with 100 indicating the most severe DE symptoms. The DEQ5 score ranges from 0 to 22, with the highest score indicating the most severe symptoms, and the NPSI-E score ranges from 0 to 100, with the highest score indicating the most severe symptoms. Psychological state was measured with the 9-item Patient Health Questionnaire, the PTSD Checklist–Military Version for PTSD, and the Symptom Checklist–90 for anxiety.
Of the 118 patients who participated in the study, 105 (88.9%) were men (mean [SD] age, 60  years), and a mean of 41% had PTSD, 10% depression, and 0.93% anxiety. Using stepwise linear regression analyses, significant associations were identified between overall DE symptom severity and posttraumatic stress disorder scores and tear breakup time (DEQ5 model: R = 0.54; OSDI model: R = 0.61, P < .001). All other variables (ie, demographics, comorbidities, medications, tear film factors, and QST metrics) dropped out of these models. When specifically considering neuropathic-like qualities of DE pain, however, anxiety and hot pain temporal summation at the forearm explained 17% of the variability in ocular burning (R = 0.41; P < .001), and PTSD score, tear breakup time, and hot pain temporal summation at the forearm explained 25% of the variability in sensitivity to wind (R = 0.50; P < .001) and 30% of the variability in total NPSI-E scores (R = 0.55; P < .001).
Conclusions and Relevance
Our findings demonstrate that neuropathic-like DE pain symptom severity correlates with quantitative measures of pain sensitivity at a site remote from the eye. This result provides additional evidence that DE symptoms are not only manifestations of a local disorder but also involve somatosensory dysfunction beyond the trigeminal system.
Dry eye (DE) is a heterogeneous disease that can include symptoms of ocular pain, visual disturbances, and various signs (eg, decreased tear production and increased evaporation).1 Although it would seem logical that patient-reported symptoms would correlate with objective ocular surface findings, research2,3 has demonstrated that DE symptoms correlate poorly with signs of ocular surface disease. One potential explanation for this dissociation is that, in some patients, symptoms may manifest as a result of or in conjunction with dysfunction within somatosensory pathways, including the nociceptive system. This dysfunction can occur anywhere along the corneal somatosensory pathway, including in primary neurons at the surface of the cornea, secondary neural networks within the trigeminal system, or within higher-order brain areas.
Recently, DE symptoms have been described4-6 as being more common in individuals with a greater number of chronic pain conditions. This finding is in line with the concept that pain does not exist in isolation and that individuals experiencing one form of chronic pain often have other chronic pain conditions,7-9 a concept termed chronic overlapping pain conditions (COPCs). Central sensitization is thought to partially underlie these findings,7 and DE symptoms have much in common with symptoms of other neuropathic pain disorders that are due to central sensitization. The common factors include characterizing ocular pain as burning10 and self-reporting hyperalgesia and allodynia (manifesting in the eye as increased sensitivity to wind and light).2,10 Similar to other COPCs, DE is also strongly associated with depression, posttraumatic stress disorder (PTSD), and anxiety, providing further evidence of a centralized pain disorder.11,12
Quantitative sensory testing (QST) has been used extensively to characterize somatosensory function in populations with chronic nonocular pain,13-15 but it has been infrequently applied to DE. Central pain syndromes are often associated14,16 with altered sensation as detected using QST techniques. Regarding DE, QST has primarily been limited to testing corneal sensitivity with the Cochet-Bonnet17 and modified Belmonte aesthesiometer,18 although 1 study19 compared pain sensitivity at the forearm between individuals with and those without DE symptoms. To further evaluate the details of somatosensory system dysfunction as a contributing mechanism to DE and ocular pain symptoms, we used an expanded set of QST techniques tested at the forearm in the present study. Decreased thresholds for pain in an area remote from the clinical pain site (forearm), as well as facilitated temporal summation (TS) of pain and enhanced or prolonged painful aftersensations, are indications of central sensitization.19 Therefore, we included these specific QST measures in the present study to provide new information regarding the contribution of central somatosensory system dysfunction to DE symptoms and related ocular pain.
Quiz Ref IDPatients with a wide variety of DE symptoms and signs (none to severe) and no overt eyelid or corneal abnormalities were prospectively recruited from the Miami Veterans Affairs Hospital between October 31, 2013, and January 28, 2016. We excluded patients if they wore contact lenses, used ocular medications other than artificial tears, had an active external ocular process, or had undergone cataract surgery within the past 6 months, refractive surgery, or any glaucoma or retinal surgery. We also excluded patients with human immunodeficiency virus, sarcoidosis, graft-vs-host disease, or collagen vascular disease. The Miami Veterans Affairs Hospital institutional review board approval allowed the prospective evaluation of patients. The study was conducted in accordance with the principles of the Declaration of Helsinki.20 Written informed consent was obtained from all participants; financial compensation was provided.
Demographics, medical history, and medication information was collected via self-report. Once collected, the data were verified by medical records.
Patients completed standardized questionnaires regarding DE symptoms (5-Item Dry Eye Questionnaire [DEQ5], with scores ranging from 0 to 22, with the highest score indicating the most severe symptoms,21 and Ocular Surface Disease Index [OSDI]); the OSDI score ranges from 0 to 100, with 100 indicating the most severe symptoms.22 In addition, the Neuropathic Pain Symptom Inventory,23 modified for the eye (NPSI-E), was administered to evaluate the severity of ocular neuropathic pain–like symptoms; on a scale of 0 to 100, with the highest score indicating the most severe level of the symptoms. In our modified version, we replaced 3 original items with items specific to ocular hyperalgesia and allodynia (eye pain evoked or worsened by wind, light, and heat or cold).10 Psychological status was assessed using the 9-item Patient Health Questionnaire (PHQ-9) for depression,24 PTSD Checklist–Military Version for PTSD,25 and the Symptom Checklist–90 (SCL-90) for anxiety.26 The PHQ-9 score ranges from 0 to 27, with the highest score indicating the most severe level of depression; the PTSD Checklist–Military Version score ranges from 17 to 84, with the highest score indicating the most severe level of PTSD; and the SCL-90 score ranges from 0 to 4, with the highest score indicating the most severe level of anxiety.
Testing was performed on the ventral right forearm, using the skin overlying the midpoint between the wrist and cubital fossa. We chose the forearm because we wished to study somatosensory function in a site remote from the eye and the forearm is a site frequently tested in other QST studies,14,27,28 facilitating comparison.
Vibration detection thresholds at 100 Hz were measured with the vibratory sensory analyzer (VSA-3000; Medoc Ltd), using the handheld VSA component (1.22-cm2 circular contactor tip). The amplitude of the vibratory stimulus increased at a rate of 5 µm/s until the participants pressed the response button to indicate that they felt the stimulus. Three trials, separated by 15 seconds, were performed, and the mean was used to determine the vibrotactile threshold at the forearm.
A TSA II (Thermal Sensory Analyzer; Medoc Ltd) machine was used to assess cool and warm detection and cold and hot pain thresholds. A square thermode (9-cm2 surface area), set at a starting temperature of 32°C, was placed on the skin of the right forearm and cooled (or heated) using software accompanying the machine. Per the ascending method of limits, the probe temperature gradually decreased (for cool and cold pain thresholds at a rate of 1°C/s and 2°C/s, respectively) or increased (for warm and hot pain thresholds, at a rate of 1°C/s and 2°C/s, respectively) until the participants pressed a button (placed in the left hand) to indicate the first moment that they perceived the sensation or the cutoff temperature was reached (0°C for cold trials, 50°C for heat trials). Three trials for each threshold were performed, with 15 seconds between detection threshold trials and 45 seconds between pain threshold trials. The mean of the results of the 3 trials for each modality was determined. All thermal thresholds are reported as the absolute change from baseline temperature (32°C). Thus, higher values for both heat and cold indicate that a greater change in temperature was needed for the participant to report detection or pain. In addition, ratings of pain intensity at threshold for cold pain and hot pain were recorded using a 0- to 100-point numerical rating scale, with 0 indicating no pain and 100 indicating the most intense pain imaginable.
Temporal summation is used to gain insight regarding central sensitization of pain since persons with certain chronic pain conditions express greater summation of pain due to repetitive presentations of a noxious stimulus compared with persons who do not have chronic pain.29-31 The TS protocol was repeated twice, first using a cold pain stimulus, next using a hot pain stimulus, with at least 90 seconds between the 2-stimulus series. We first presented a single, 1-second stimulus on the forearm and asked the participant to rate the evoked pain intensity on a 0- to 100-point numeric rating scale. We then presented a train of 10 stimuli (1 second on, 1 second off, repeated 10 times) and asked the individuals to rate the peak pain intensity reached during the train of stimuli. The temperatures of the stimuli were based on the mean cold and hot pain thresholds measured for the first 20 participants in our sample so that the test stimuli were set at 2°C below (for cold pain, 6.3°C) and above (for hot pain, 45.5°C) the mean pain threshold at the forearm in this sample. Measures of TS were obtained by subtracting the rating of the first, singular stimulus from the rating of peak pain during the repetitive stimulus series.
Ratings of pain aftersensations (continued feelings of pain that last beyond the noxious stimulus presentation) were also obtained. These data were collected 15 seconds after the termination of the last stimulus in the TS protocol.
All patients underwent tear film assessment of both eyes, including measurement of (1) tear osmolarity (TearLab), (2) InflammaDry testing (Rapid Pathogen Screening Inc), (3) tear breakup time (TBUT), (4) corneal staining,18 (5) Schirmer strips with anesthesia, and (6) meibomian gland assessment. Eyelid vascularity was graded on a scale of 0 to 3 (0, none; 1, mild; 2, moderate; and 3, severe engorgement) and meibum quality on a scale of 0 to 4 (0, clear; 1, cloudy; 2, granular; 3, toothpaste; and 4, no meibum extracted).
Mechanical detection and pain thresholds of the right central cornea were assessed with a modified Belmonte noncontact aesthesiometer using methods previously described.18 The tip of the aesthesiometer (0.5-mm diameter) was placed perpendicular to and 4 mm from the surface of the cornea of the right eye. Stimulation consisted of pulses of air at room temperature applied to the corneal surface. The method of limits, using ascending series only, was used to measure both mechanical detection and mechanical pain thresholds.
Descriptive statistics were used to summarize patient demographic and clinical information. The distributional spread of each QST metric was assessed. Similar to the methods in other QST studies,13,32 log transformation of vibration, cool, and warm detection thresholds was required for normalization. Cold and hot pain thresholds and all TS variables were not normally distributed in raw form or as log-transformed values. We present the results for both parametric and nonparametric tests given the differences in distributional properties across variables. For correlations, Pearson r and Spearman ρ values are presented. Linear regression analyses with forward selection were used to evaluate the contribution of the different variables on DE symptoms and ocular pain. For the multivariable models, we present the statistic for the full model (R). We inspected residuals from the linear regression analysis for departures from normality and heterogeneity. We opted to give information on all variables compared rather than correcting the P value since the latter method has limitations.33 Statistical analysis was performed using SPSS, version 22.0 (SPSS Inc).
A total of 118 patients participated in the study (mean [SD] age, 60  years). Of these, 105 were men (88.9%), 52 were white (44.1%), and 30 were Hispanic (25.4%) (Table 1).
The QST measures obtained at the forearm are presented in eTable 1 in the Supplement, and associations between QST measures and demographics and health variables are presented in eTable 2 in the Supplement. Age was positively correlated with vibration and thermal detection thresholds (r = 0.42 for vibration, r = 0.19 for cool, r = 0.23 for warm; P < .05 for all). Women had lower vibration detection thresholds than men and higher pain intensity ratings at pain threshold (mean [SD] cold pain intensity ratings: 53.9 [22.1] vs 37.0 [27.6], P = .04; hot pain intensity ratings: 62.6 [26.5] vs 45.3 [27.6], P = .03). A similar pattern was seen with regard to race, with black individuals having lower vibration thresholds and higher pain intensity ratings at threshold compared with white participants (cold pain intensity ratings: 44.0 [28.4] vs 32.3 [25.0]; P = .02; hot pain intensity ratings: 51.7 [26.6] vs 41.4 [28.8]; P = .05).
Higher depression (measured with PHQ9), PTSD, and anxiety scores (measured with SCL-90) were associated with lower thermal pain thresholds and higher hot pain aftersensation intensity ratings (eTable 3 in the Supplement). Individuals receiving antidepressants similarly had lower thermal pain thresholds compared with those not receiving these medications.
Although traditional measures of DE symptom severity (DEQ5 and OSDI) were not correlated with detection (vibration, cool, and warm) and pain (hot pain, cold pain) thresholds, the measures were correlated with ratings of painful aftersensations evoked by hot pain TS, and OSDI total scores were additionally positively correlated with ratings of pain intensity at both hot and cold pain thresholds (Table 2). A similar pattern emerged when considering self-reported measures of neuropathic-like qualities of eye pain (ie, severity of hot, burning pain; sensitivity to wind or light stimuli; and total NPSI-Eye score), wherein threshold measures were not generally related to neuropathic-like eye pain, but ratings of painful sensations, particularly hot pain, were related. Associations were found between (1) burning eye pain and cold pain threshold, hot pain TS, and aftersensations evoked by hot pain TS; and (2) total scores on the NPSI-Eye and ratings of pain intensity at cold pain and hot pain thresholds, hot pain TS, and aftersensations evoked by hot pain TS.
Tear osmolarity, TBUT, and corneal staining did not correlate with QST metrics (Table 3). In contrast, individuals with ocular surface inflammation (via InflammaDry) had higher mean (SD) hot-pain ratings at threshold compared with InflammaDry-negative individuals (56.1 [27.2] vs 41.7 [27.5]; P = .008), demonstrating increased pain sensitivity on the forearm in those with ocular surface inflammation. Schirmer scores negatively correlated (r = −0.29, P = .002) and eyelid vascularity and meibum quality positively correlated (r = 0.24, P = .009; r = 0.17, P = .06, respectively) with vibration thresholds, demonstrating deceased sensitivity to vibration in individuals with less healthy ocular surface factors. When we reexamined this association adjusting for age and reported as standardized coefficient (β), only tear production remained negatively correlated with vibration detection thresholds constant: t = −1.22, P = .23; age: β = 0.43, t = 5.55, P < .001; Schirmer score: β = −0.29, t = −3.75, P < .001).
Mechanical detection and pain thresholds on the cornea positively correlated with vibration thresholds at the forearm and negatively correlated with the degree of hot pain TS and TS-evoked aftersensations. These findings demonstrate that greater corneal sensitivity was associated with more vibrotactile and hot pain sensitivity at a remote site (Table 4).
Quiz Ref IDTo test the robustness of the association between QST metrics and DE symptoms, we performed forward stepwise linear regression analyses that controlled for demographics, medication (antidepressants and anxiolytics), mental health (PTSD, depression, and anxiety), ocular surface signs, and corneal sensitivity. When controlling for these covariates, QST variables did not remain associated with overall DE symptoms. The PTSD scores and TBUT indicated the probability of symptoms in these models (DEQ5 model: PTSD β = 0.51; P < .001; TBUT β = 0.29; P = .001; R = 0.54; OSDI model: PTSD β = 0.62; P < .001; TBUT β = −0.18; P = .03; R = 0.61). When focusing on neuropathic-like qualities of eye pain, anxiety scores and HPTS explained 17% of the variability in reported severity of ocular burning (total model: R = 0.41, P < .001; HPTS alone: R = 0.25, P = .007); PTSD scores, TBUT, and hot pain TS explained 25% of the variability in sensitivity to wind (total model: R = 0.50, P < .001; HPTS alone: R = 0.21, P = .02) and 30% of the variability in total NPSI-E scores (total model: R = 0.55, P < .001; HPTS alone: R = 0.26, P = .004). Results of these omnibus analyses suggest that neuropathic pain–like eye symptoms are most associated with signs of central sensitization (hot pain TS), even after controlling for characteristics that affect pain report (PTSD and anxiety) and local tear disruption (TBUT).
Quiz Ref IDIn our sample of veterans, most of whom were men, we found that QST measures at the forearm were not related to general DE symptom severity (DEQ5 and OSDI scores) when other associated factors were considered. However, QST measures indicative of central sensitization (hot pain TS) were associated with the severity of symptoms characteristic of neuropathic ocular pain (total NPSI-E score, burning ocular pain, and sensitivity to wind). This association between sensitivity to evoked pain at a site remote from the eye and pain-specific ocular symptoms suggests that central neuropathy, beyond that of the trigeminal system, may be driving ocular pain in some patients with idiopathic DE.
Measures of hot pain TS and ratings of 15-second aftersensations on the forearm provided the most robust findings with regard to positive correlations between DE symptom report, including neuropathic-like qualities of eye pain and pain sensitivity at the cornea. Increased TS of pain is an indication of pain facilitation thought to reflect central sensitization within the somatosensory system.34 Similar findings, including greater TS and prolonged painful aftersensations, have been reported31 in persons with other diagnosed chronic pain syndromes compared with healthy individuals. Thus, the enhanced summation of hot pain and prolonged aftersensations on the forearm found in our participants with more severe DE symptoms and neuropathic-like ocular pain suggest that these eye symptoms may be linked to central somatosensory dysfunctions, including within ocular pain processing pathways.
Our findings expand on those reported by Vehof et al28 in a population-based female cohort. In that study, comparisons of heat pain thresholds and heat pain tolerance at the forearm between individuals with and without DE symptoms (via OSDI) revealed greater, but nonsignificant, sensitivity in those with OSDI scores of 15 or higher. These differences were more pronounced and significant when focusing on responses to the pain-specific questions within the OSDI. Our study, analyzing data from a primarily male veteran clinical population, has added to the evidence of generalized somatosensory dysfunction by revealing correlations between several pain sensitivity metrics (cold pain thresholds, ratings of pain intensity to threshold and suprathreshold thermal noxious stimuli, and TS of noxious heat) and measures of ocular pain and DE symptom severity. Thus, in many patients with DE, the disorder may not be limited to mechanisms within only the ocular surface, but can also include both local (corneal18) and widespread somatosensory dysfunction.
Our findings build on the idea that DE symptoms may represent a peripheral manifestation of a COPC.4,33,35 Many patients with COPCs have enhanced sensitivity to evoked pain, and the underlying mechanisms responsible for these COPCs involve both peripheral and central pathways.36 Thus, the results of the present study showing enhanced pain sensitivity at a site remote from the primary symptom of DE provide evidence of similarities between idiopathic DE and other COPCs. Genetics have been found to have an important role in the clinical variation of COPCs, including pain perception, processing, and response to therapies.37-39 Twin-based studies33,35 have shown DE to be moderately heritable and demonstrated a latent factor underlying DE and other COPCs (eg, chronic widespread pain, chronic pelvic pain, and irritable bowel syndrome) that have a strong heritable component as well.
Quiz Ref IDAs with all studies, the limitations of ours need to be considered, which include a unique DE population, specific metrics used to capture symptoms and somatosensory function, and many statistical tests performed, which increases the risk of type I error. In addition, the participants had a variety of localized and systemic comorbidities that may have affected their DE phenotype. We decided to adjust for these factors (eg, eyelid laxity, medications, and sleep apnea) rather than excluding patients with such abnormalities, which would have restricted our potential patient pool. Despite these limitations, we demonstrated that patients with neuropathic-like ocular pain symptoms and/or ocular surface inflammation have evidence of central sensitization. Thus, central sensitization is suggested as a potential mechanism underlying previous observations4,40 that DE symptom severity correlates with other COPCs, depression, and PTSD. Finally, these results provide further evidence that DE is a disorder not only localized to the ocular surface. An increased understanding of the underlying pathophysiologic and somatosensory dysfunction may help to explain the discordance between DE signs and symptoms and lead to more effective individualized preventive and treatment algorithms for this complex disorder.
Neuropathic-like DE pain symptom severity correlates with quantitative measures of pain sensitivity at a site remote from the eye. This finding provides evidence that DE symptoms are not only manifestations of a local disorder but also involve somatosensory dysfunction beyond the trigeminal system.
Corresponding Author: Anat Galor, MD, MSPH, Bascom Palmer Eye Institute, University of Miami, 900 NW 17th St, Miami, FL 33136 (firstname.lastname@example.org).
Accepted for Publication: August 14, 2016.
Published Online: September 29, 2016. doi:10.1001/jamaophthalmol.2016.3642
Author Contributions: Dr Galor had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Galor, Levitt, Park, Sarantopoulos, Felix.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Galor, Levitt, Seiden, Park, Sarantopoulos.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Galor, Park.
Obtained funding: Galor.
Administrative, technical, or material support: Levitt, McManus, Park, Covington.
Study supervision: Galor, Kalangara, Seiden, Park, Felix.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.
Funding/Support: Support for the study was provided by the Department of Veterans Affairs, Veterans Health Administration, Office of Research and Development, Clinical Sciences Research EPID-006-15S (Dr Galor), National Institutes of Health Center Core Grant P30EY014801 and Research to Prevent Blindness unrestricted grant.
Role of the Funder/Sponsor: The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Disclaimer: The contents of this study do not represent the views of the Department of Veterans Affairs or the US government. The authors alone are responsible for the content and writing of the article.
Create a personal account or sign in to: