Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jul 29.
Published in final edited form as: J Pain. 2021 Dec 8;23(5):784–795. doi: 10.1016/j.jpain.2021.11.010

Self-Report of Severity of Ocular Pain Due to Light as a Predictor of Altered Central Nociceptive System Processing in Individuals With Symptoms of Dry Eye Disease

Daniel A Rodriguez *,†, Anat Galor *,†, Elizabeth R Felix ‡,§
PMCID: PMC13409074  NIHMSID: NIHMS2194677  PMID: 34890797

Abstract

Dry eye disease (DED) is a diagnosis given to individuals with a heterogeneous combination of symptoms and/or signs, including spontaneous and evoked ocular pain. Our current study evaluated whether and which ocular pain assessments could serve as screening tools for central sensitization in individuals with DED. A cohort of individuals with DED symptoms (n = 235) were evaluated for ocular pain, DED signs (tear production, evaporation), evoked sensitivity to mechanical stimulation at the cornea, and evidence of central sensitization. Central sensitization was defined for this study as the presence of pain 30 seconds after termination of a thermal noxious temporal summation protocol (ie, aftersensations) presented at a site remote from the eye (ventral forearm). We found that combining ratings of average intensity of ocular pain, ratings of average intensity of pain due to light, response to topical anesthetic eye drops, and corneal mechanical pain thresholds produced the best predictive model for central sensitization (area under the curve of .73). When examining ratings of intensity of ocular pain due to light alone (0–10 numerical rating), a cutoff score of 2 maximized sensitivity (85%) and specificity (48%) for the presence of painful aftersensations at the forearm. Self-reported rating of pain sensitivity to light may serve as a quick screening tool indicating the involvement of central nociceptive system dysfunction in individuals with DED.

Perspective:

This study reveals that clinically-relevant variables, including a simple 0 to 10 rating of ocular pain due to light, can be used to predict the contribution of central sensitization mechanisms in a subgroup of individuals with DED symptoms. These findings can potentially improve patient stratification and management for this complex and painful disease.

Keywords: Dry eye disease, ocular pain, central sensitization, quantitative sensory testing, aftersensations, photophobia, photoallodynia

Dry Eye Disease (DED), as defined by in the International Dry Eye Workshop II (DEWS II) report, “is a multifactorial disease… characterized by a loss of homeostasis of the tear film, and accompanied by ocular symptoms, in which tear film instability and hyperosmolarity, ocular surface inflammation and damage, and neurosensory abnormalities play etiological roles”.9 DED has been reported to affect between 5 and 50% of the population, and has been shown to severely affect quality of life, including impairments in driving ability, use of a computer, and emotional well-being.36,38,44

The wide range in estimates of DED prevalence can be explained by many factors, including varying disease definitions (some papers defined the disease by the presence of symptoms alone, others required decreased tear production, and yet others examined tear quality irrespective of tear production) and variation by population studied, with DED frequencies being higher in females, older individuals, and Asian populations.50

Common presenting symptoms of DED include spontaneous and evoked sensations of eye dryness, discomfort, or pain, and although the term “dry eye” suggests that tear insufficiency underlies these symptoms, there is discordance between the severity of these symptoms and the severity of DED signs/tear parameters.24,43 In fact, treatments aimed at improving tear insufficiency are variable in their effect on patient-reported symptoms:22,29 In 1 study, 20% of individuals had no improvement in symptoms with use of artificial tears, and 62% had only partial improvement.22 “Neurosensory abnormalities,” as stated within the DEWS II definition of DED, may be the likely explanation for ocular pain symptoms that are present in some patients with a DED diagnosis but who have normal tear parameters or who do not respond to typical treatments.3 Neurosensory dysfunction leading to ocular pain can occur peripherally, at the level of corneal nerves, as has been shown by findings of abnormal corneal nerve density and anatomy.26,58 Other neural mechanisms leading to chronic ocular pain may occur centrally, in higher-order pathways, as is often seen in individuals with comorbid migraine20 or traumatic brain injury.31 Stratifying patients based on the involvement of nociceptive system mechanisms at the peripheral and central levels can assist with the delivery of personalized solutions to improve DED morbidity and quality of life.

Unfortunately, however, diagnostic tests for detecting neurosensory dysfunction related to symptoms of ocular pain are limited to imaging of corneal nerves,13 measurements of corneal surface sensitivity to mechanical stimulation,49 and testing for persistence of ocular pain after application of topical anesthetic eye drops.10 Due to time constraints and availability issues, some of these tests are not easily integrated into clinical practice. Thus, there is a need for a quick clinical assessment that can screen for the possible involvement of nociceptive system processing abnormalities in individuals with DED symptoms. One potential screening tool for the presence of central sensitization in individuals with DED is the self-report of pain evoked by light, a symptom that is often comorbid with DED diagnosis14 and whose presence indicates dysfunction within the nociceptive system.14,15,30

The first goal of this study was to assess which combination of measures, including ocular symptom report, ocular surface metrics, sensitivity of the cornea to mechanical stimulation, and response to peripheral anesthetic, would maximize the sensitivity and specificity of detecting central sensitization in individuals with DED symptoms. The second goal of the study was to determine whether a self-report rating of intensity of pain due to light could serve as a singular, clinically-feasible screening tool for central sensitization, given that ratings of pain due to light are quick to obtain, do not require specialized equipment, and have been associated with central neuroplasticity.14 The presence of prolonged aftersensations to repetitive thermal noxious stimulation delivered at a site remote from the eye (ie, the forearm) was used as our gold standard indicating central sensitization.57

Methods

Study Population

Potential participants were prospectively recruited from the eye clinic at the Miami Veterans Affairs (VA) Hospital between August 21, 2014, and September 28, 2017. This population includes patients seen at the eye clinic for regular, symptom-free check-ups, for follow-up for an established condition, or for new ocular symptoms. Inclusion criteria for the present analyses required a positive screen for DED symptoms, defined as a score of ≥6 on the 5-Item Dry Eye Questionnaire (DEQ-5).6 We excluded individuals if they had comorbid conditions that could account for their DED symptoms, including abnormal anatomy (pterygium, corneal edema, iris defects), contact lens wear, the use of medications beyond artificial tears (glaucoma medications), an active external ocular process, cataract surgery within the past 6 months, any refractive, glaucoma, or retinal surgery, documented human immunodeficiency virus, sarcoidosis, graft-versus-host disease, or collagen vascular disease.

For each individual, demographic information (age, sex, race (white, black, or other), ethnicity), past ocular and medical history (depression, arthritis, sleep apnea, benign prostatic hyperplasia, diabetes, hypertension, migraine, and headaches), and medication information (antidepressants, analgesics, antianxiety, and antihistamines) were collected from electronic medical records. Individuals were asked to rate the average intensity of pain in nonocular regions during the previous 3 months, using a numerical rating scale (NRS) ranging from 0 (“no pain”) to 10 (“most intense pain imaginable”). Individuals were also asked to provide the total number of nonocular chronic pain problems they currently had. The Miami VA Hospital’s institutional review board approval was obtained and allowed for the prospective evaluation of subjects. The study was conducted in accordance with the principles of the Declaration of Helsinki. Written informed consent was obtained from all participants.

Metrics Collected

Measures used in the study are detailed below, grouped based on construct assessed (clinical symptoms, nociceptive system function, tear function). The order of collection of these measures during the study session was: 1) ocular symptom assessment; 2) mechanical sensitivity of the cornea; 3) presence of thermal pain aftersensations on the forearm; 4) ocular surface examination; and 5) topical anesthetic challenge for persistent ocular pain.

Ocular Symptoms

Dry Eye Symptoms

Participants completed the DEQ-5,6 a validated questionnaire regarding the frequency and severity of DE symptoms (eye discomfort, eye dryness, and watery eyes), with scores ranging from 0 to 22.

Self-Report Ratings of Ocular Pain and Pain Due to Light

Individuals were asked to rate their typical level of eye pain during the past week, using an NRS ranging from 0 (“no pain”) to 10 (“most intense pain imaginable”). They were also asked to rate the intensity of eye pain specifically evoked by light during the past 24 hours, on an NRS anchored at 0 for “no pain associated with light” and 10 for “the most severe pain associated with light”.21

Nociceptive System Function

Mechanical Sensitivity of the Cornea

A modified Belmonte noncontact aesthesiometer was used to assess mechanical detection and mechanical pain thresholds at the central cornea.49 Stimulation consisted of room-temperature pulses of air, ranging between 10 and 400 mL/min in 10mL/min steps. Air puffs were applied through the tip of the aesthesiometer which was perpendicular to, and 4 mm from, the surface of the cornea of the right eye. The method of limits was used to determine detection threshold, by taking the average of 2 ascending series, asking the participant to report when he/she first felt a slight pressure sensation on the eye. Pain threshold at the cornea was similarly measured, averaging 2 ascending series of stimulus presentations, asking the participant to report when he/she first felt the stimulus was painful. For each mechanical pain trial, the participant was also asked to rate the intensity of the pain at threshold, using a 0 (= “no pain”) to 10 (= “the most intense pain imaginable”) NRS. The average of these 2 ratings was calculated for further analyses.

Topical Anesthetic Challenge for Persistent Ocular Pain

Individuals were asked to rate the current intensity of their eye pain using the 0 to 10 NRS, after which a drop of anesthetic (proparacaine .5%, Akron) was applied to each eye. After 30 seconds, individuals were asked to rate their current eye pain intensity again. Individuals were grouped by pain responses into 3 categories: 1) pain before anesthesia that completely resolved; 2) pain before anesthesia that did not completely resolve; and 3) no pain present before anesthesia. Grouping participant responses was done in this way as it is indicative of potential mechanisms underlying ocular pain symptoms: 1) peripheral mechanisms underlying eye pain; 2) at least some contribution of central mechanisms underlying eye pain; and 3) indeterminate due to lack of current eye pain, respectively.10

Thermal Pain Aftersensations on the Forearm

Testing was performed on the right ventral forearm, at the midpoint between the wrist and cubital fossa, in order to study somatosensory function at a commonly-used test site remote from the eye.32 Painful aftersensations were separately measured after repeated presentation of a cold stimulus, set at 6.3°C (2°C below the average cold pain threshold for 20 participants in a cohort group from the same patient population), and after repeated presentation of a hot stimulus, set at 45.5°C (1°C above the average hot pain threshold from this same group of 20 participants). Stimuli were delivered using the Medoc TSA-II device, with a 30mm by 30mm square thermal contact probe. For each stimulus series, the thermode was set to the prescribed temperature and 10 one-second stimulus presentations were manually delivered to the skin at a rate of .5 Hz. At the end of the last stimulus presentation, a timer was started to mark the 30-second poststimulus time point. At this time point, the participant was asked to rate the intensity of the pain at the site of testing on the forearm at the present moment. The cold pain aftersensation trial was conducted first, with a minimum of 3-minute rest before the first stimulus of the hot pain aftersensation trial began. The presence of any 30-second aftersensations (rating > 0) on the forearm reported after the termination of either hot or cold trains of thermal noxious stimuli was used as an indicator of heightened central sensitization.32,54

Ocular Surface Examination Parameters

Standard techniques were used to evaluate tear function and ocular surface health, as put forth by the Tear Film and Ocular Surface DEWS II.56 This included measurements of: 1) tear osmolarity (TearLAB, San Diego, CA), with a value ≥326 mOsm/L in either eye or a difference of ≥8 mOsm/L between eyes considered abnormal; 2) tear stability measured via tear breakup time (TBUT) (measured as time to first observed black spot in the tear film after placement of fluorescein). The test is repeated 3 times per eye and results averaged. A more rapid TBUT is indicative of an unstable tear film, with values <5 seconds considered abnormal; 3) corneal epithelial cell disruption graded to the National Eye Institute scale,56 where 5 sections of the cornea are examined (range = 0–3) and a total score generated by summing the section scores (range 0–15). A higher number indicates more severe epithelial disruption, with a score ≥2 being considered abnormal; 4) tear production via Schirmer’s strips, which are placed in the corner of the eye and left in place for 5 minutes after anesthetic application. Lower scores indicate lower tear production, with values <5 mm considered abnormal; 5) meibum quality on a scale of 0 to 4 (0 = clear; 1 = cloudy; 2 = granular; 3 = toothpaste; 4 = no meibum extracted). Higher values indicate more severe meibomian gland abnormalities with a value ≥2 considered abnormal. Values for each of these measures were captured for both eyes and the more abnormal value used for the analyses. These ocular surface metrics were used to subgroup patients based on the presence or absence of DED signs. Using these subgroups, we evaluate the robustness of our ocular pain-central sensitization model’s performance in the setting of various tear and ocular surface abnormalities.

Statistical Approach

All statistical analyses were performed using R (3.3.1). Descriptive analyses were performed to describe the characteristics of our study sample. Assessment of the impact of potential confounding variables (presence of comorbidities, use of medications, number of chronic pain conditions, and average intensity of nonocular pain conditions) on the presence of aftersensations was examined using Fisher’s exact tests and t-tests.

Machine Learning Modeling

To address the first goal of the study, identifying which measures of pain and nociceptive system function may be good indicators of central sensitization, we investigated the relationships between our “gold standard” indicator of central sensitization (presence of 30-second aftersensations evoked via repeated brief noxious thermal stimulation), and relevant measures of ocular pain, including: 1) rating of severity of eye pain during the past week (range 0–10), 2) self-report rating of severity of eye pain due to light (range 0–10), 3) grouping based on topical anesthetic challenge (eye pain before anesthesia that completely resolved, pain before anesthesia that did not completely resolve, and no pain before anesthesia), 4) corneal mechanical detection threshold (0–400 mL/min), 5) corneal mechanical pain threshold (0–400 mL/min), 6) pain intensity rating at corneal mechanical pain threshold (range 0–10), and 7) difference between mechanical detection threshold and pain threshold at the cornea.

We used a machine learning approach, utilizing the Random Forest classifier implemented in Python 3 SK-Learn (v. 0.18.1) package, to combine and test multiple predictive variables together, incorporating predictor interactions, and providing an opportunity for improving the accuracy of predicting outcomes and disease presence.8 The random forest is a model made up of many decision trees.28,39 Decision trees use the data to generate a set of rules which are used to predict an outcome. In the random forest algorithm, each tree is trained on a slightly different set of inputs, splitting nodes in each tree considering a limited number of features, then averaging the predictions of each tree. This improves the predictive accuracy, reduces the noise generated by the variance in the data, and attempts to control for over-fitting. 1000 estimators/trees and a max depth of 3 nodes in each tree were used. The mean accuracy score for the model was obtained through a threefold cross validation scheme. Training and learning sets were composed of a random selection of our cohort. We then calculated receiver operating characteristic (ROC) curves for all 127 possible combinations of the 7 variables tested. The top-ranked model was that which produced the highest value when assessing the mean area under the curve (AUC) of each combination.

Due to the heterogenous nature of DED signs, we additionally sought to assess the stability of our top-ranked model with respect to different possible DED presentations. We reran the top ranked predictive model, splitting individuals into groups based on the presence or absence of various DE signs. Specifically, each tear parameter was tested separately, splitting individuals into normal versus abnormal values for: tear osmolarity, TBUT, corneal epithelial cell disruption, Schirmer’s test wetting, and meibomian gland quality, using the cutoffs described above. Sample sizes for TBUT <5 seconds (n = 34) and Schirmer’s test <5mm wetting (n = 21) were low, therefore we used a relaxed twofold cross validation for these groups instead of the threefold cross described previously. This approach allowed us to test whether ocular surface parameters affected the robustness of our model for predicting central sensitization. Any patients with missing data were excluded from the analysis.

Finally, as the secondary goal of the study was to examine the utility of a simple self-reported rating of eye pain due to light, the most efficient and potentially meaningful assessment to apply in the clinical setting, we used this measure alone to predict the presence of central sensitization (ie, prolonged aftersensations after repeated noxious stimulation) using the random forest model, with severity of pain due to light rating as the only independent variable. Then, we generated an empirical ROC curve, using the R package ROCit, and estimated critical cutoffs for ratings of severity of eye pain due to light (range: 0–10) for predicting the presence of aftersensations measured at the remote site.

Results

Participant Characteristics

A cohort of 235 individuals with DE symptoms (DEQ-5 scores ≥6) and for whom QST metrics were obtained, were included in the present analyses. A flow diagram of recruitment and enrollment is presented in Fig 1. The mean age of participants was 59.7 years old, and the majority of participants were male (88.5%), Black (61.7%), and non-Hispanic (73.6). The most common comorbidities were hypertension (67.7%) and depression (67.2%), and the majority of participants were taking analgesic medications (68.9%), antianxiety medications (56.2%), or antidepressants (54.4%). Details of demographic, comorbidity, and medication information, QST measures, and DE characteristics (symptoms and signs) of the study sample are presented in Table 1. The distribution of ratings for eye pain due to light (our primary predictor variable of interest) is presented in Fig 2.

Figure 1.

Figure 1.

Open in a new tab

Cohort recruitment and inclusion flow chart. Diagram depicting the number of patients recruited and retained for inclusion in the present study.

Table 1.

Demographics, Comorbidities, DE Symptoms, Ocular Surface Examination, and QST Measures (n = 235 Except Where Otherwise Noted)

Demographics
Age, mean (SD) [min-max] 59.7 (9.5) [27–87]
Gender, male, n (%) 208 (88.5%)
Race, black, n (%) 145 (61.7%)
Ethnicity, Hispanic, n (%) 62 (26.4%)
Comorbidities n (%)
Depression 158 (67.2%)
Arthritis 135 (57.4%)
Sleep apnea 69 (29.3%)
Benign prostatic hyperplasia 43 (18.3%)
Diabetes 66 (28.1%)
Hypertension 159 (67.7%)
Migraine 44 (18.7%)
Headache 61 (26.0%)
Total number of other nonocular chronic pain problems, Mean (SD) [min-max] 2.73 (1.5) [0–5]
Average rating of nonocular pain over the past 3 months (0–10), Mean (SD) [min-max] 5.38 (2.8) [0–10]
Medications n (%)
Antidepressant 128 (54.4%)
Antianxiety 132 (56.2%)
Antihistamine 59 (25.1%)
Analgesics 162 (68.9%)
Dry eye symptoms Mean (SD)
Dry Eye Questionnaire - 5 (0–22) [min-max] 13.0 (3.5) [6–22]
Ocular pain and photophobia
Average eye pain rating during the past week (0–10), Mean (SD) [min-max] 3.8 (2.5) [0–10]
Response to topical anesthetic (proparacaine .5% eye drops):
 pain before anesthesia that completely resolved, n (%) 19 (8.1%)
 pain before anesthesia that did not completely resolve, n(%) 33 (14.0%)
 no pain before anesthesia, n (%) 183 (77.5%)
Intensity rating of pain due to light (0–10), Mean (SD) [min-max] 4.0 (3.3) [0–10]
Corneal sensitivity (Belmonte) Mean (SD)
Corneal detection threshold, mL/min (n = 232) [min-max] 80.7 (38.2) [10–195]
Corneal pain threshold, mL/min (n = 232) [min-max] 240.4 (122.3) [25–410]
Corneal pain intensity rating (0–10) at threshold (n = 226) [min-max] 2.9 (2.7) [0–10]
Corneal threshold difference*, mL/min (n = 232) [min-max] 160.0 (113.2) [0–390]
Central sensitization probe n (%)
Presence of 30 s aftersensation to either noxious cold or heat stimuli 41 (17.4%)
Presence of 30 s aftersensation to noxious cold stimulation 32 (13.6%)
Presence of 30 s aftersensation to noxious heat stimulation 27 (11.5%)
Ocular surface findings† Abnormal n (%)
Tear osmolarity (n = 214) 94 (43.9%)
Tear film breakup time (n = 232) 33 (14.2%)
Corneal staining (n = 234) 97 (41.5%)
Schirmer’s test (n = 233) 21 (9.0%)
Meibum quality (n = 231) 128 (55.4%)
Open in a new tab
*

(Corneal pain threshold value – corneal detection threshold value).

†

Represents value from more severely affected eye.

Figure 2.

Figure 2.

Open in a new tab

Distribution of ratings of the intensity of eye pain due to light. The height of each bar corresponds to the number of participants who reported that rating for the intensity of pain due to light.

Associations Among Aftersensations and Comorbidities and Medications

The presence of aftersensations/central sensitization was not different between individuals reporting comorbidities of arthritis, migraine, nonmigraine headaches, depression, Benign prostatic hyperplasia, hypertension, or diabetes (Fisher’s exact tests P = .63, P = .79, P = .87, P = .63, P = .19, P = .32, and P = .65, respectively). However, individuals who exhibited central sensitization via aftersensations did have a significantly greater number of chronic pain comorbidities and reported greater nonocular average pain ratings than individuals who did not exhibit central sensitization (two-sided t-test P = .002, P = .02, respectively). Individuals taking analgesics, antidepressants, antianxiety, or antihistamine medication did not differ in regards to the presence of aftersensations compared to those who were not using these medications (Fischer’s exact tests P = .53, P = .93, P = .69, and P = .68, respectively).

Machine Learning Model of Central Sensitization

As detailed above, the variables of interest as possible indicators of central sensitization were: self-reported rating of severity of eye pain during the past week; self-reported ratings of intensity of pain evoked by light; grouping based on response to topical anesthetic; mechanical detection thresholds at the cornea; mechanical pain thresholds at the cornea; pain intensity rating at corneal mechanical pain threshold; and the difference between mechanical detection and pain thresholds at the cornea. To test the predictive value of the association between aftersensations (indicator of central sensitization) and our diagnostic variables of interest, we assembled a machine learning model using a supervised random forests approach. After running all combinations of variables, the top-performing model had a mean area under the curve (AUC) of .73 § .03 (Fig 3) and included self-reported rating of eye pain intensity, self-reported rating of intensity of pain due to light, grouping based on response to topical anesthetic, and corneal mechanical pain threshold. When these 4 metrics were analyzed together, the machine learning algorithm was able to correctly classify 68% of individuals into two groups – those with 30 second aftersensations, and those without 30 second aftersensations due to repetitive noxious stimulation.

Figure 3.

Figure 3.

Open in a new tab

Receiver operating characteristic (ROC) curve of best performing model. ROC curve for 4-metric model (self-reported rating of eye pain intensity, self-reported rating of intensity of pain due to light, grouping based on anesthetic challenge, and corneal mechanical pain threshold) predictability of 30 second aftersensations due to repetitive noxious stimulation. Threefold cross validation scheme generated 3 ROC curves for each fold, 0 to 2, and the blue line indicates the mean ROC curve. The red dotted line indicates a nondiscriminatory test where the number of true positives is equivalent to the number of false positives.

Evaluating Our Model in Relation to Dry Eye Disease Signs

We next tested the robustness of the top-performing model by evaluating its performance on classifying individuals with or without specific DED signs (Table 2). For 4 of the 5 DED signs examined (tear osmolarity, TBUT, corneal staining, and meibum quality), the model performance was similar between those with and without abnormal tear parameters. However, for 1 DED sign (Schirmer), the AUC associated with the model did differ, with the AUC being higher in those with normal tear production.

Table 2.

Top Performing Model Performance on Classifying Individuals With or Without Specific DE Signs

DE sign AUC ± SD (normal for sign)* AUC ± SD (abnormal for sign)*
Tear osmolarity .72 ± .06 .69 ± .09
Tearfilm break up time (TBUT) .72 ± .01 .66 ± .08
Epithelial disruption (corneal staining) .67 ± .03 .68 ± .06
Schirmer .71 ± .04 .40 ± .05
Meibum quality .65 ± .03 .62 ± .05
Open in a new tab
*

Normal or abnormal measures of each DE sign as described in the methods.

Prediction of Central Sensitization by Ratings of Ocular Pain Due to Light

Based on the clinical feasibility and ease of capturing self-reported ratings of eye pain due to light, and prior data suggesting that the presence of pain due to light may be indicative of central neuroplasticity,14 we applied our machine learning approach using self-reported rating of pain sensitivity to light (NRS 0–10) as the only independent variable. This model was found to be less robust compared to the full, 4-factor model but still explained up to 60% of variability in aftersensation presence, with an AUC of .61 ± .04. (Fig 4). We then constructed a logistic regression-based ROC curve using ratings of the severity of eye pain due to light to predict the presence of painful aftersensations/central sensitization. Consistent with the results from the machine learning algorithm, our empirical ROC resulted in an AUC of .65. From the curve we obtained the Youden index point, which indicates the cutoff point for optimal classification. The optimal cutoff was at a rating of pain intensity due to light of 2, where the ROC curve has a sensitivity of .85 and specificity of .48.

Figure 4.

Figure 4.

Open in a new tab

Receiver operating characteristic (ROC) curve of rating of intensity of pain due to light as predictor of evoked painful aftersensations. ROC curve for prediction of the presence 30 second aftersensations due to repetitive noxious stimulation using only ratings of intensity of pain due to light. Threefold cross validation scheme generated 3 ROC curves for each fold, 0 to 2, and the blue line indicates the mean ROC curve. The red dotted line indicates a nondiscriminatory test where the number of true positives is equivalent to the number of false positives.

Discussion

This study evaluated the utility of several metrics for predicting central sensitization [as indicated by the presence of lingering painful sensations 30 seconds after a repeated noxious stimulus was presented to an unaffected area (forearm)] in individuals with DE symptoms. We found that, when taken together, rating of average intensity of ocular pain, rating of intensity of pain due to light, corneal mechanical pain threshold, and response to topical anesthetic on the cornea provided the most robust predictive model for the presence of aftersensations on the forearm. Additionally, the model largely held when tested on individuals grouped by abnormal versus normal DED signs, highlighting the independence between measures of tear parameters and measures of nociceptive system (dys)function related to ocular pain.26,24 Some of the assessed metrics can be utilized in clinic (pain assessments, response to anesthetic) while others are more appropriate for research (aesthesiometry). In order to maximize clinical feasibility, we then tested the utility of the easiest metric to capture in clinic, a 0 to 10 rating of average eye pain severity due to light, as an indicator of central sensitization. A cutoff value of ≥2 for this single metric produced good sensitivity (.85) for indicating central nociceptive system dysfunction in patients with DED. Although this model produced low specificity (.48), the intent of the rating of pain intensity due to light question is to serve as a clinically feasible screening tool, to indicate those who may have central neuropathic mechanisms involved in their DED diagnosis. Thus, using ratings of intensity of pain due to light of 2 or greater would capture the large majority of individuals with indications of central sensitivity, warranting further assessment by the clinician.

Our study builds on previous work that has described an association between DED symptoms and abnormal nerve function.5,12,49 We previously found that individuals with persistent ocular pain after topical anesthetic (indicative of central or nonocular mechanisms underlying eye symptoms) self-reported higher ratings of ocular pain evoked by wind and light, ocular equivalents of the hyperalgesia and allodynia frequently seen in individuals with neuropathic pain in areas outside of the eye.10 We also previously demonstrated a positive relationship between painful DED symptoms and evoked pain sensitivity at a site remote from the eye (forearm),25 suggesting systemic pain dysregulation. In the present study, we add data on relationships between specific symptoms of ocular pain (ie, pain due to light) and central pain processing abnormalities revealed via prolonged aftersensations from noxious stimulation at a site remote from the eye. We also show that the presence of aftersensations was not associated with the use of medications such as analgesics or antidepressants, or due to other, nonpain comorbid conditions.

We did, however, find an association between the total number and intensity of nonocular chronic pain problems and the presence of aftersensations. This result is consistent with our prior studies, in which we demonstrated that individuals with overlapping chronic pain conditions reported more severe ocular pain symptoms compared to individuals without multiple chronic pain conditions.11,23 Furthermore, we found that individuals who reported persistent pain after anesthesia (a proposed marker of a central contribution to ocular pain), reported greater sensitivity to wind and light (ie, ocular hyperalgesia and allodynia) and to noxious heat stimulation at the forehead and forearm, compared to individuals who did not report persistent ocular pain after anesthesia.10

There is biologic plausibility that nerve lesions or dysfunction contribute to DE symptoms in some individuals. Pathologic neuroplasticity associated with DE can occur at multiple levels within the corneal pathway, including at the level of the primary afferents, trigeminal nucleus caudalis, thalamus, and cortex.26 The rich density and superficial location of the corneal afferents, integrated between epithelial cells, makes them vulnerable to injury in the setting of trauma, ocular surface stress, and disease.27 Episodic or ongoing damage to corneal nerves through tear evaporation, inflammation, hyperosmolarity, or mechanical stimulation (eg, contact lens wear) may result in altered neuronal healing, maladaptive neuroplasticity, and prolonged hypersensitivity to normally non-noxious stimuli (such as wind and light). Previous studies have reported alterations in nerve function occurring after eye stressors, both peripherally, demonstrated by lowered threshold and enhanced peripheral nerve responsiveness due to sensitization by inflammatory mediators,2 and centrally, demonstrated by signs of increased excitability and receptive field area enlargement for neurons in the caudal trigeminal brainstem after removal of the lacrimal gland (causing decreased tear production).40 In fact, we have shown that therapies that attempt to normalize nerve function can improve painful DED symptoms and severity of pain due to light, including oral treatment with a2d ligands (eg, gabapentin), injection of anesthetics,48 or electrical stimulation,47,59 further adding plausibility that nerve dysfunction drives painful DE symptoms in some individuals.

Evidence from animal models supports the role of central nociceptive pathways in mediating nocifensive behaviors evoked by light as well. Behavioral,55 transgenic,33,60 and pharmacologically-induced35 mouse models of pain due to light have been developed. Utilizing these techniques, intrinsically photosensitive retinal ganglion cell-dependent and independent pathways have been reported to be responsible for photoallodynia.35 Furthermore, the neuropeptide calcitonin gene-related peptide induces light aversion in transgenic mice, acting on both peripheral and central mechanisms,34 where intracerebroventricular injection of calcitonin gene-related peptide caused sensitivity to low light. This peptide plays an important role in migraine16 and neurovascular headaches, which exhibit disruptions in nociceptive processing in higher-order pathways, similar to those found in DE.37 Understanding the overlap between these pathologies may provide insight into therapeutic options for individuals with central sensitization and painful DE symptoms refractory to topical medication (eg, artificial tears).

Our findings need to be evaluated while considering the study limitations, which include a specific population (ie, veterans from 1 clinic location who were over-whelmingly male) and select measures for capturing DED and ocular pain symptoms and for assessing corneal and central somatosensory function. For example, we did not collect behavioral responses to noxious light stimulation, such as pupil dilation response, which has been shown to be correlated with pain intensity ratings across a variety of cutaneous noxious stimuli.7,18,19 Additionally, Belmonte aesthesiometry is not widely available and corneal sensitivity is more often qualitatively measured with a cotton tip or quantitatively measured with Cochet-Bonnet aesthesiometer in clinic. As we did not perform the cotton tip or Cochet-Bonnet tests, we could not include them in our analyses. Furthermore, the presence of aftersensations is not the only QST metric that has been used to identify central abnormalities in pain processing. Prolonged painful aftersensation was chosen as our “gold standard” because it has been repeatedly demonstrated to be significantly associated with ratings of fibromyalgia pain severity1,4,45,51-54 and to be a predictor of persistent postsurgical pain,17,41,46 as well as being a comparatively better predictor of clinical pain than other QST metrics, such as temporal summation.41,42,45,52,54

Despite these limitations, we present a robust predictive model for the contribution of central sensitization in a subgroup of individuals with DE symptoms. These results can aid ophthalmologists and pain clinicians in screening individuals who may have a component of central nervous system dysfunction as a driver of symptoms. This is important, as identifying the underlying cause(s) of symptoms in an individual will allow delivery of precision-based medicine that will lead to improvement in disease morbidity and quality of life. Future studies are needed to assess the utility of other clinically available tests in broader DED populations. Furthermore, research is needed to evaluate which therapies are optimal in treating DED symptoms in individuals with detected central sensitization.

Acknowledgments

We would like to recognize our clinical and research staff at the Veteran Affairs hospital in Miami. This work would not have been possible without the incredible volunteers who agreed to be a part of our study.

Supported by the Department of Veterans Affairs, Veterans Health Administration, Office of Research and Development, Clinical Sciences R&D (CSRD) I01 CX002015 (Dr. Galor) and Biomedical Laboratory R&D (BLRD) Service I01 BX004893 (Dr. Galor), Department of Defense Gulf War Illness Research Program (GWIRP) W81XWH-20-1-0579 (Dr. Galor) and Vision Research Program (VRP) W81XWH-20-1-0820 (Dr. Galor), National Eye Institute R01EY026174 (Dr. Galor) and R61EY032468 (Dr. Galor), NIH Center Core Grant P30EY014801 (institutional) and Research to Prevent Blindness Unrestricted Grant GR004596 (institutional).

Footnotes

The authors declare no conflicts of interest.

References

  • 1.Anderson RJ, McCrae CS, Staud R, Berry RB, Robinson ME: Predictors of clinical pain in fibromyalgia: examining the role of sleep. J Pain 13:350–358, 2012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Belmonte C, Acosta MC, Gallar J: Neural basis of sensation in intact and injured corneas. Exp Eye Res 78:513–525, 2004 [DOI] [PubMed] [Google Scholar]
  • 3.Belmonte C, Nichols JJ, Cox SM, Brock JA, Begley CG, Bereiter DA, Dartt DA, Galor A, Hamrah P, Ivanusic JJ, Jacobs DS, McNamara NA, Rosenblatt MI, Stapleton F, Wolffsohn JS: TFOS DEWS II pain and sensation report. Ocular Surface 15:404–437, 2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Bosma RL, Mojarad EA, Leung L, Pukall C, Staud R, Stroman PW: FMRI of spinal and supra-spinal correlates of temporal pain summation in fibromyalgia patients. Hum Brain Mapp 37:1349–1360, 2016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Bourcier T, Acosta MC, Borderie V, Borras F, Gallar J, Bury T, Laroche L, Belmonte C: Decreased corneal sensitivity in patients with dry eye. Invest Ophthalmol Vis Sci 46:2341–2345, 2005 [DOI] [PubMed] [Google Scholar]
  • 6.Chalmers RL, Begley CG, Caffery B: Validation of the 5-Item Dry Eye Questionnaire (DEQ-5): Discrimination across self-assessed severity and aqueous tear deficient dry eye diagnoses. Contact Lens Anterio 33:55–60, 2010 [DOI] [PubMed] [Google Scholar]
  • 7.Chapman CR, Oka S, Bradshaw DH, Jacobson RC, Donaldson GW: Phasic pupil dilation response to noxious stimulation in normal volunteers: relationship to brain evoked potentials and pain report. Psychophysiology 36:44–52, 1999 [DOI] [PubMed] [Google Scholar]
  • 8.Chen JH, Asch SM: Machine learning and prediction in medicine - beyond the peak of inflated expectations. N Engl J Med 376:2507–2509, 2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Craig JP, Nichols KK, Akpek EK, Caffery B, Dua HS, Joo CK, Liu Z, Nelson JD, Nichols JJ, Tsubota K, Stapleton F: TFOS DEWS II definition and classification report. Ocul Surf 15:276–283, 2017 [DOI] [PubMed] [Google Scholar]
  • 10.Crane AM, Feuer W, Felix ER, Levitt RC, McClellan AL, Sarantopoulos KD, Galor A: Evidence of central sensitisation in those with dry eye symptoms and neuropathic-like ocular pain complaints: incomplete response to topical anaesthesia and generalised heightened sensitivity to evoked pain. Brit J Ophthalmol 101:1238–1243, 2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Crane AM, Levitt RC, Felix ER, Sarantopoulos KD, McClellan AL, Galor A: Patients with more severe symptoms of neuropathic ocular pain report more frequent and severe chronic overlapping pain conditions and psychiatric disease. Br J Ophthalmol 101:227–231, 2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.De Paiva CS, Pflugfelder SC: Corneal epitheliopathy of dry eye induces hyperesthesia to mechanical air jet stimulation. Am J Ophthalmol 137:109–115, 2004 [DOI] [PubMed] [Google Scholar]
  • 13.Dermer H, Hwang J, Mittal R, Cohen A, Galor A: Corneal sub-basal nerve plexus microneuromas in individuals with and without dry eye. Br J Ophthalmol, 2021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Diel RJ, Mehra D, Kardon R, Buse DC, Moulton E, Galor A: Photophobia: shared pathophysiology underlying dry eye disease, migraine and traumatic brain injury leading to central neuroplasticity of the trigeminothalamic pathway. Br J Ophthalmol, 2020 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Digre KB, Brennan KC: Shedding light on photophobia. J Neuroophthalmol 32:68–81, 2012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Edvinsson L, Haanes KA, Warfvinge K, Krause DN: CGRP as the target of new migraine therapies - successful translation from bench to clinic. Nat Rev Neurol 14:338–350, 2018 [DOI] [PubMed] [Google Scholar]
  • 17.Edwards RR, Mensing G, Cahalan C, Greenbaum S, Narang S, Belfer I, Schreiber KL, Campbell C, Wasan AD, Jamison RN: Alteration in pain modulation in women with persistent pain after lumpectomy: influence of catastrophizing. J Pain Symptom Manage 46:30–42, 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Eisenach JC, Curry R, Aschenbrenner CA, Coghill RC, Houle TT: Pupil responses and pain ratings to heat stimuli: reliability and effects of expectations and a conditioning pain stimulus. J Neurosci Methods 279:52–59, 2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Ellermeier W, Westphal W: Gender differences in pain ratings and pupil reactions to painful pressure stimuli. Pain 61:435–439, 1995 [DOI] [PubMed] [Google Scholar]
  • 20.Farhangi M, Diel RJ, Buse DC, Huang AM, Levitt RC, Sarantopoulos CD, Felix ER, Galor A: Individuals with migraine have a different dry eye symptom profile than individuals without migraine. Brit J Ophthalmol 104:260–264, 2020 [DOI] [PubMed] [Google Scholar]
  • 21.Farhangi M, Feuer W, Galor A, Bouhassira D, Levitt RC, Sarantopoulos CD, Felix ER: Modification of the Neuropathic Pain Symptom Inventory for use in eye pain (NPSI-Eye). Pain 160:1541–1550, 2019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Galor A, Batawi H, Felix ER, Margolis TP, Sarantopoulos KD, Martin ER, Levitt RC: Incomplete response to artificial tears is associated with features of neuropathic ocular pain. Brit J Ophthalmol 100:745–749, 2016 [DOI] [PubMed] [Google Scholar]
  • 23.Galor A, Covington D, Levitt AE, McManus KT, Seiden B, Felix ER, Kalangara J, Feuer W, Patin DJ, Martin ER, Sarantopoulos KD, Levitt RC: Neuropathic ocular pain due to dry eye is associated with multiple comorbid chronic pain syndromes. J Pain 17:310–318, 2016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Galor A, Feuer W, Lee DJ, Florez H, Venincasa VD, Perez VL: Ocular surface parameters in older male veterans. Invest Ophthalmol Vis Sci 54:1426–1433, 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Galor A, Levitt RC, McManus KT, Kalangara JP, Seiden BE, Park JJ, Covington DB, Sarantopoulos CD, Felix ER: Assessment of somatosensory function in patients with idiopathic dry eye symptoms. Jama Ophthalmol 134:1290–1298, 2016 [DOI] [PubMed] [Google Scholar]
  • 26.Galor A, Moein HR, Lee C, Rodriguez A, Felix ER, Sarantopoulos KD, Levitt RC: Neuropathic pain and dry eye. Ocul Surf 16:31–44, 2018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Guthoff RF, Wienss H, Hahnel C, Wree A: Epithelial innervation of human cornea: a three-dimensional study using confocal laser scanning fluorescence microscopy. Cornea 24:608–613, 2005 [DOI] [PubMed] [Google Scholar]
  • 28.Ho TK: Random decision forests. in In: Proceedings of the Third International Conference on Document Analysis and Recognition; 1995. p. 278–282 [Google Scholar]
  • 29.Jones L, Downie LE, Korb D, Benitez-Del-Castillo JM, Dana R, Deng SX, Dong PN, Geerling G, Hida RY, Liu Y, Seo KY, Tauber J, Wakamatsu TH, Xu J, Wolffsohn JS, Craig JP: TFOS DEWS II management and therapy report. Ocul Surf 15:575–628, 2017 [DOI] [PubMed] [Google Scholar]
  • 30.Katz BJ, Digre KB: Diagnosis, pathophysiology, and treatment of photophobia. Surv Ophthalmol 61:466–477, 2016 [DOI] [PubMed] [Google Scholar]
  • 31.Lee CJ, Felix ER, Levitt RC, Eddy C, Vanner EA, Feuer WJ, Sarantopoulos CD, Galor A: Traumatic brain injury, dry eye and comorbid pain diagnoses in US veterans. Brit J Ophthalmol 102:667–673, 2018 [DOI] [PubMed] [Google Scholar]
  • 32.Maier C, Baron R, Tolle TR, Binder A, Birbaumer N, Birklein F, Gierthmuhlen J, Flor H, Geber C, Huge V, Krumova EK, Landwehrmeyer GB, Magerl W, Maihofner C, Richter H, Rolke R, Scherens A, Schwarz A, Sommer C, Tronnier V, Uceyler N, Valet M, Wasner G, Treede DR: Quantitative sensory testing in the German Research Network on Neuropathic Pain (DFNS): somatosensory abnormalities in 1236 patients with different neuropathic pain syndromes. Pain 150:439–450, 2010 [DOI] [PubMed] [Google Scholar]
  • 33.Marquez de Prado B, Hammond DL, Russo AF: Genetic enhancement of calcitonin gene-related Peptide-induced central sensitization to mechanical stimuli in mice. J Pain 10:992–1000, 2009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Mason BN, Kaiser EA, Kuburas A, Loomis MM, Latham JA, Garcia-Martinez LF, Russo AF: Induction of migraine-like photophobic behavior in mice by both peripheral and central CGRP mechanisms. J Neurosci 37:204–216, 2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Matynia A, Parikh S, Chen B, Kim P, McNeill DS, Nusinowitz S, Evans C, Gorin MB: Intrinsically photosensitive retinal ganglion cells are the primary but not exclusive circuit for light aversion. Exp Eye Res 105:60–69, 2012 [DOI] [PubMed] [Google Scholar]
  • 36.Miljanovic B, Dana R, Sullivan DA, Schaumberg DA: Impact of dry eye syndrome on vision-related quality of life. Am J Ophthalmol 143:409–415, 2007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Noseda R, Burstein R: Migraine pathophysiology: anatomy of the trigeminovascular pathway and associated neurological symptoms, CSD, sensitization and modulation of pain. Pain 154(Suppl 1), 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Pouyeh B, Viteri E, Feuer W, Lee DJ, Florez H, Fabian JA, Perez VL, Galor A: Impact of ocular surface symptoms on quality of life in a United States veterans affairs population. Am J Ophthalmol 153:1061–1066.e1063, 2012. [DOI] [PubMed] [Google Scholar]
  • 39.Quinlan J: Induction of decision trees. Mach Learn 1:81–106, 1986 [Google Scholar]
  • 40.Rahman M, Okamoto K, Thompson R, Katagiri A, Bereiter DA: Sensitization of trigeminal brainstem pathways in a model for tear deficient dry eye. Pain 156:942–950, 2015 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Rojas MJ, Felix E, Small L, Goldhardt R, Galor A: Autonomic function testing and after-sensations as predictors for post-operative dry eye symptoms after laser-assisted in situ keratomileusis. Invest Ophth Vis Sci 60:5202, 2019 [Google Scholar]
  • 42.Sato H, Saisu H, Muraoka W, Nakagawa T, Svensson P, Wajima K: Lack of temporal summation but distinct aftersensations to thermal stimulation in patients with combined tension-type headache and myofascial temporomandibular disorder. J Orofac Pain 26:288–295, 2012 [PubMed] [Google Scholar]
  • 43.Schein OD, Tielsch JM, Munoz B, Bandeen-Roche K, West S: Relation between signs and symptoms of dry eye in the elderly. A population-based perspective. Ophthalmology 104:1395–1401, 1997 [DOI] [PubMed] [Google Scholar]
  • 44.Schiffman RM, Walt JG, Jacobsen G, Doyle JJ, Lebovics G, Sumner W: Utility assessment among patients with dry eye disease. Ophthalmology 110:1412–1419, 2003 [DOI] [PubMed] [Google Scholar]
  • 45.Schreiber KL, Loggia ML, Kim J, Cahalan CM, Napadow V, Edwards RR: Painful after-sensations in fibromyalgia are linked to catastrophizing and differences in brain response in the medial temporal lobe. J Pain 18:855–867, 2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Schreiber KL, Martel MO, Shnol H, Shaffer JR, Greco C, Viray N, Taylor LN, McLaughlin M, Brufsky A, Ahrendt G, Bovbjerg D, Edwards RR, Belfer I: Persistent pain in postmastectomy patients: comparison of psychophysical, medical, surgical, and psychosocial characteristics between patients with and without pain. Pain 154:660–668, 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Sivanesan E, Levitt RC, Sarantopoulos CD, Patin D, Galor A: Noninvasive electrical stimulation for the treatment of chronic ocular pain and photophobia. Neuromodulation 21:727–734, 2018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Small LR, Galor A, Felix ER, Horn DB, Levitt RC, Sarantopoulos CD: Oral gabapentinoids and nerve blocks for the treatment of chronic ocular pain. Eye Contact Lens 46:174–181, 2020 [DOI] [PubMed] [Google Scholar]
  • 49.Spierer O, Felix ER, McClellan AL, Parel JM, Gonzalez A, Feuer WJ, Sarantopoulos CD, Levitt RC, Ehrmann K, Galor A: Corneal mechanical thresholds negatively associate with dry eye and ocular pain symptoms. Invest Ophth Vis Sci 57:617–625, 2016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Stapleton F, Alves M, Bunya VY, Jalbert I, Lekhanont K, Malet F, Na KS, Schaumberg D, Uchino M, Vehof J, Viso E, Vitale S, Jones L: TFOS DEWS II epidemiology report. Ocul Surf 15:334–365, 2017 [DOI] [PubMed] [Google Scholar]
  • 51.Staud R, Robinson ME, Price DD: Temporal summation of second pain and its maintenance are useful for characterizing widespread central sensitization of fibromyalgia patients. J Pain 8:893–901, 2007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Staud R, Robinson ME, Vierck CJ Jr, Cannon RC, Mauderli AP, Price DD: Ratings of experimental pain and pain-related negative affect predict clinical pain in patients with fibromyalgia syndrome. Pain 105:215–222, 2003 [DOI] [PubMed] [Google Scholar]
  • 53.Staud R, Vierck CJ, Cannon RL, Mauderli AP, Price DD: Abnormal sensitization and temporal summation of second pain (wind-up) in patients with fibromyalgia syndrome. Pain 91:165–175, 2001 [DOI] [PubMed] [Google Scholar]
  • 54.Staud R, Weyl EE, Riley JL, Fillingim RB: Slow temporal summation of pain for assessment of central pain sensitivity and clinical pain of fibromyalgia patients. Plos One 9, 2014 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Thiels E, Hoffman EK, Gorin MB: A reliable behavioral assay for the assessment of sustained photophobia in mice. Curr Eye Res 33:483–491, 2008 [DOI] [PubMed] [Google Scholar]
  • 56.Wolffsohn JS, Arita R, Chalmers R, Djalilian A, Dogru M, Dumbleton K, Gupta PK, Karpecki P, Lazreg S, Pult H, Sullivan BD, Tomlinson A, Tong L, Villani E, Yoon KC, Jones L, Craig JP: TFOS DEWS II diagnostic methodology report. Ocular Surface 15:539–574, 2017 [DOI] [PubMed] [Google Scholar]
  • 57.Woolf CJ: Central sensitization: Implications for the diagnosis and treatment of pain. Pain 152:S2–S15, 2011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Yang AY, Chow J, Liu J: Corneal innervation and sensation: the eye and beyond. Yale J Biol Med 91:13–21, 2018 [PMC free article] [PubMed] [Google Scholar]
  • 59.Zayan K, Aggarwal S, Felix E, Levitt R, Sarantopoulos K, Galor A: Transcutaneous electrical nerve stimulation for the long-term treatment of ocular pain. Neuromodulation 23:871–877, 2020 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60.Zhang Z, Winborn CS, Marquez de Prado B, Russo AF: Sensitization of calcitonin gene-related peptide receptors by receptor activity-modifying protein-1 in the trigeminal ganglion. J Neurosci 27:2693–2703, 2007 [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES