ORIGINAL ARTICLE

Central corneal thickness and its relationship to Parkinson’sdisease severityDurdane Aksoy, MD,* Huseyin Ortak, MD,† Semiha Kurt, MD,* Emre Cevik, MD,* Betul Cevik, MD*

ABSTRACT ● RÉSUMÉ

Objective: To investigate the effect of Parkinson’s disease (PD) on blink rate (BR), tear breakup time test (TBUT), Schirmer’s test,and corneal thickness, and the relationship of these effects with disease severity.

Design: Prospective controlled study.Participants: Fifty-five eyes from 55 patients with PD and 40 eyes from 40 healthy subjects were analyzed in the study.Methods: The patients were divided into 2 groups according to their Hoehn–Yahr (H-Y) scores; patients classified as H-Y 1–2 were

designated as the mild group, and those classified as H-Y 3–5 were designated as the moderate group. Subjects were screenedfor BR, TBUT, and Schirmer’s test, and the central corneal thickness (CCT) was measured.

Results: The BR, Schirmer’s test, TBUT, and CCT values of the patient group were significantly lower than those of the controlgroup. The BR and TBUT of the mild group were significantly lower than those of the control group, but the decreases in theSchirmer’s test values and CCT were not statistically significant. In addition, significant decreases in the BR, TBUT, Schirmer’stest scores, and CCT were observed in the patient group as the H-Y score increased.

Conclusions: A reduced BR and poor tear quality in the early stages of PD, as well as decreased tear production as the diseaseprogresses, can result in reduced CCT. The possibility of a thin cornea should be taken into consideration while measuring theintraocular pressure in patients with severe PD.

Objet : Examen des effets de la maladie de Parkinson sur le taux de clignotement, le temps de la rupture lacrymale, le test deSchirmer, ainsi que l'épaisseur de la cornée et la relation de ces effets avec la gravité de la maladie.

Nature : Étude prospective contrôlée.Participants : Cinquante-cinq yeux de 55 patients atteints de la maladie de Parkinson et 40 sujets en santé ont été analysés.Méthodes : Les patients ont été répartis en deux groupes selon leurs données Hoehn-Yahr; ceux qui étaient classés Hoehn-Yahr 1-

2 étaient considérés comme0 groupe léger T et ceux classés Hoehn-Yahr 3-5, comme0 groupe modéréT. Les sujets ont subides tests de fréquence de clignotement et de temps de rupture lacrymale et le test Schirmer, et l'on a mesuré l'épaisseur ducentre de la cornée.

Résultats : Le taux de clignotement, le test Schirmer, le temps de rupture lacrymale et l'épaisseur du centre de la cornée du groupede patients étaient significativement inférieurs à ceux du groupe témoin. Les taux de clignotement et de le temps de la rupturelacrymale du groupe léger étaient significativement inférieurs à ceux du groupe témoin, mais la baisse des valeurs du testSchirmer et de l'épaisseur du centre de la cornée n'étaient pas importante statistiquement. En outre, on a observé une baissesignificative du taux de clignotement, du temps de la rupture lacrymale, des résultats du test Schirmer et de l'épaisseur du centrede la cornée alors que s'accroissaient les données Hoehn-Yahr du groupe de patients.

Conclusions : Une réduction du taux de clignotement et une piètre qualité lacrymale dans les premières étapes de la maladie deParkinson, ainsi qu'une production réduite de larmes alors que la maladie progresse, peut réduire l'épaisseur du centre de lacornée. L'on devrait aussi tenir compte de la possibilité d'amincissement de la cornée en mesurant la pression intraoculaire despatients qui ont une grave maladie de Parkinson.

Parkinson’s disease (PD) is a neurodegenerative disorderaccompanied by both motor and nonmotor symptoms.Nonmotor symptoms, including visual problems, auto-nomic dysfunction, gastrointestinal symptoms, andimpairments in sleep and cognitive function, are common,yet they are often under-recognized in clinical practice.1–3

Visual symptoms, such as colour discrimination, visuospa-tial deficit, reduced blink rate (BR), dry eye, transientdiplopia, vergence dysfunction, hypometric saccades,impaired smooth pursuit with catch-up saccades, mildimpairment of upgaze, convergence insufficiency, visualhallucinations, ocular discomfort, and pain, are reported inpatients with PD, and these problems negatively impact

From the Departments of *Neurology and †Ophthalmology,Gaziosmanpasa University Faculty of Medicine, Tokat, Turkey

Originally received Aug. 22, 2013. Final revision Dec. 22, 2013.Accepted Jan. 3, 2014

Correspondence to Durdane Aksoy, Gaziosmanpasa University Faculty ofMedicine, Department of Neurology, Kaleardi mah, Muhittin Fusunuglucad. Tokat 60100, Turkey; dbekar@yahoo.com

152 CAN J OPHTHALMOL—VOL. 49, NO. 2, APRIL 2014

activities of daily living such as ambulation, use of stairs,and driving.3–6

Corneal thickness of the central and midperipheralcornea is significantly decreased in individuals with dryeyes.7 A reduced BR and dry eye syndrome are welldescribed in the PD literature.3,5,8–10 However, the effectsof PD and the disease severity on corneal thickness areunknown. In this study, the BR, tear breakup time test(TBUT), Schirmer’s test score, and central corneal thick-ness (CCT) were analyzed in a group of patients with PDand compared with healthy individuals. The effects of PDseverity on the BR, tear tests, and corneal thickness werealso investigated.

Can J Ophthalmol 2014;49:152–1560008-4182/14/$-see front matter & 2014 Canadian OphthalmologicalSociety. Published by Elsevier Inc. All rights reserved.http://dx.doi.org/10.1016/j.jcjo.2013.12.010

Table 1—Comparison of age, blink rate, tear breakup time test,Schirmer, and central corneal thickness mean values with ttest for independent samples for patients with PD andcontrol groups

Group Mean SD t z p

Age PD 63.44 10.10Control 61.05 7.87 1.24 — 0.22

Blink rate PD 6.27 3.61Control 11.55 4.39 — –2.69 0.007*

TBUT PD 6.49 2.36Control 10.65 4.27 — –4.77 0.0001*

Schirmer PD 8.95 3.68Control 11.28 3.41 3.13 — 0.002

CCT PD 541.89 12.81Control 558.92 12.25 3.63 — 0.0001

TBUT, tear breakup time test; CCT, central corneal thickness.

*Because data do not follow normal distribution, Mann–Whitney U test was applied.

Central corneal thickness in Parkinson’s disease—Aksoy et al.

METHODS

Fifty-five patients with PD and 40 healthy subjects wereenrolled in this study. Patients with PD met the Parkin-son’s Disease Society Brain Bank (London, U.K.) clinicaldiagnostic criteria and were diagnosed by an experiencedneurologist. Patients diagnosed with diabetes mellitus,glaucoma, keratoconus, corneal dystrophy and keratitis,active uveitis, systemic therapy with pharmaceutical drugswith known corneal toxicity were excluded from the study,as were contact lens users, patients receiving topicaltherapy with antiglaucoma drugs, and patients who hada previous history of ophthalmic surgery. Patients treatedwith drugs other than L-dopa and dopamine agonists werealso excluded. The hospital ethics committee approved thestudy, and patients provided written informed consentafter the nature and purpose of the study were fullyexplained. All experiments were performed in accordancewith the Declaration of Helsinki.

A detailed neurologic examination was performed onpatients with PD, their Hoehn–Yahr (H-Y) scores werecalculated, and sex, age, disease duration, and duration ofL-dopa and/or dopamine agonist therapy were recorded.11

The patients were divided into 2 groups according to theirH-Y scores; patients who scored H-Y 1–2 were designatedas the mild group, and those scored H-Y 3–5 weredesignated as the moderate group.

Fifty-five eyes from 55 patients with PD (32 male, 23female) and 40 eyes from 40 healthy individuals (23 male,17 female), consistent in age and sex with the patients,were examined. A full ophthalmologic evaluation was per-formed on each subject. All patients were screened for BR,TBUT, Schirmer test under topical anaesthesia, and CCT;CCT was measured using an ultrasonic pachymeter.

BR was determined by 1 examiner recording the averagenumber of blinks per minute during a 3-minute restingperiod. A blink was defined as bilateral paroxysmal closureof the eyelids in the absence of external stimulus.

One eye of each patient was chosen at random forstatistical analysis. The patient’s cornea was anaesthetizedwith topical proparacaine hydrochloride 0.5% (Alcon-Couvreur, Puurs, Belgium), and CCT measurements wereobtained by a single ophthalmologist using an ultrasonicpachymeter (Optikon 2000 SPA; Pacline, Rome, Italy).The TBUT test was performed using a sterile fluoresceinstrip that was placed in the wet surface of the lower fornixof the conjunctiva. The patient was asked to blink twiceand then look straight ahead without blinking. The timebefore the first defect appeared in the stained tear film wasrecorded as the tear film TBUT. The average value of3 consecutive TBUT tests was obtained.

The Schirmer test was performed under anaesthesia.Three minutes after 2 proparacaine hydrochloride 0.5%eye drops were instilled, the lid margin was dried withcotton. The Schirmer’s test strip was placed in the lowertemporal fornix at the junction of the middle and lateral

thirds of the eyelid. The patient was asked to keep botheyes immobile and 5 minutes later, the strip was removedand the amount of wetting was recorded in millimeters.All dry eye tests were performed during the same dailytime interval (from 10:00 AM to 2:00 PM) at a temperaturebetween 201C and 251C and relative humidity between35% and 45%.

The Kolmogorov–Smirnov normality test was used totest the normal distribution of continuous variables.Continuous variables were compared using Student t testfor independent samples and the Mann–Whitney U test.For the comparison of the mild, moderate, and controlgroups, 1-way ANOVA and Kruskal–Wallis varianceanalysis were used. Tukey honestly significant difference(HSD) and Mann–Whitney U tests were used for post hoctests. The correlation between the test parameters and H-Yscores in the patient group was evaluated using thePearson correlation coefficient. The p values less than0.05 were considered statistically significant. Calculationswere made using statistical analysis software (IBM SPSSStatistics 19; SPSS Inc, Chicago, Ill.).

RESULTS

The average ages of the patient group and the controlgroup were (mean � SD) 63.44 � 10.10 and 61.05 �7.87, respectively. There was no significant age differencebetween the 2 groups (Table 1). The average duration ofdisease in the PD group was 4.81 years, and the averageH-Y score was 2.17 � 0.95. The BR, Schirmer’s test,TBUT, and CCT values of the PD group were signifi-cantly lower than those of the control group (Table 1).Patients with PD were subdivided into 2 groups (mild andmoderate). When the 3 groups (mild, moderate, andcontrol) were compared using 1-way ANOVA and Krus-kal–Wallis variance analysis, a statistically significantdifference in the BR, TBUT, Schirmer’s test, and CCTvalues was observed (Table 2).

Tukey test analysis revealed that the BR and TBUTvalues were significantly lower in the mild group than in

CAN J OPHTHALMOL—VOL. 49, NO. 2, APRIL 2014 153

Table 2—Comparison of mild, moderate, and control groups in terms of blink rate, Schirmer’s test, tear breakup time test, andcentral corneal thickness values (1-way ANOVA/Kruskal–Wallis variance analysis)

Control (mean � SD) Mild (mean � SD) Moderate (mean � SD) F χ² p

Blink rate 11.55 � 4.39 7.09 � 4.04 5.05 � 2.43 — 31.15 0.0001*TBUT 10.65 � 4.27 7.00 � 2.19 5.73 � 2.45 — 25.02 0.0001*Schirmer’s test 11.28 � 3.41 10.88 � 2.95 6.05 � 2.64 22.63 — 0.0001CCT 558.92 � 12.25 549.36 � 12.77 530.68 � 10.14 12.17 — 0.0001

TBUT, tear breakup time test; CCT, central corneal thickness.

*Because data do not follow normal distribution, Kruskal–Wallis variance analysis was used.

Central corneal thickness in Parkinson’s disease—Aksoy et al.

the control group. The Schirmer’s and CCT values werealso lower in the mild group, but the differences were notstatistically significant (Table 3).

The Pearson correlation coefficient was used to deter-mine the correlation between the BR, tear test, and CCTvalues and diseases severity (H-Y scores). A strong andsignificant negative correlation was detected between theH-Y scores and CCT values, as well as between the H-Yscores and Schirmer’s test scores. Analysis also revealed aweak but significant negative correlation between the H-Yscores and BR and TBUT values (Table 4). These resultssuggest that in patients with PD, the BR, TBUT,Schirmer’s test, and CCT values significantly decrease asthe H-Y score increases.

DISCUSSION

Ocular motility impairments, dry eye, and ocular sur-face irritation are common clinical ophthalmologic find-ings in patients with PD.3,6,8 Reduced BR in patients withPD has been reported in several studies, and reduced BR isconsidered a symptom that supports a PD diagno-sis.3,5,6,12,13 Reduced BR in patients with PD is associatedwith hypokinesia resulting from reduced dopaminergicactivity.3 In a study that evaluated the BR in patients withPD in a variety of circumstances (while being interviewed,watching a video, and reading from a book), the BR wassignificantly lower in the patient group than in the controlgroup in all situations; however, no correlation was foundbetween BR and disease severity.10 Other studies havereported a negative correlation between the H-Y scoresand BR in patients with PD.8,13 The results of our studyare consistent with previous reports, suggesting a statisti-cally significant negative correlation between the H-Yscores and BR.

Table 3—Comparison of mild group and control group interms of blink rate, tear breakup time test, Schirmer’s test, andcentral corneal thickness values using Tukey test

Mild Group(mean � SD)

Control Group(mean � SD)

p

Age (y) 63.21 � 9.97 61.05 � 7.87 0.30Blink rate 7.09 � 4.04 11.55 � 4.39 0.001TBUT 7.00 � 2.19 10.65 � 4.27 0.001Schirmer’s test 10.88 � 2.95 11.28 � 3.41 0.59CCT 549.36 � 12.77 558.92 � 12.25 0.07

TBUT, tear breakup time test; CCT, central corneal thickness.

154 CAN J OPHTHALMOL—VOL. 49, NO. 2, APRIL 2014

Abnormalities in tear secretion are frequently reportedamong patients with PD.3,7–9 In a study of patients withearly-stage PD who had not received any treatment, dry eyesyndrome was detected in almost two thirds of the patients.Moreover, when tear tests were evaluated separately, theTBUT values of early-stage patients were lower than thoseof the control group. In addition, increased ocular surfacedisease and tear film abnormalities in patients with PDwere highlighted in a recent study, whereas noting thatthese patients could be asymptomatic because of reducedcorneal sensitivity.9 These studies all indicate poor tear filmquality in patients with PD. Another study, evaluating theSchirmer’s test, BR, TBUT and Rose bengal stain, andphenol red thread tests, reported that patients with PD hadabnormalities in at least 1 of the tear tests.8 In our study,patients with early-stage PD had significantly lower BRsand TBUT values than subjects in the control group. TheSchirmer’s test and CCT values were also, although notsignificantly, lower. A reduced spontaneous BR in patientswith PD may contribute to ocular discomfort caused bydry eye syndrome. Blinking is necessary to heal the tear–cornea optical interface and maintain a smooth and regulartear film.3,14–16 Our findings and those of prior studies inthe field suggest that impairments in the BR and tearquality may occur even in the early stages of the disease.3,8

A significant difference was detected in the BR, TBUT,Schirmer’s test, and CCT values between the mild,moderate, and control groups in this study. The BR andTBUTs were significantly lower in the mild group than inthe control group, whereas the Schirmer’s test values andCCT were slightly lower. Furthermore, the same valueshad significant negative correlation with the H-Y scores.As a result, we report that the BR and TBUT aredecreased during the mild stages of PD and, as the diseaseprogresses, the BR, TBUT, Schirmer’s test, and CCTvalues decrease in correlation with increasing H-Y scores.

Table 4—Evaluation of relation between Hoehn–Yahr scoresand central corneal thickness, Schirmer’s test, tear breakuptime test, and blink rate using simple correlation analysis byPearson correlation coefficient test

r p

Hoehn–Yahr/blink rate –0.412 0.002Hoehn–Yahr/TBUT –0.316 0.019Hoehn–Yahr/Schirmer –0.581 0.0001Hoehn–Yahr/CCT –0.535 0.0001

TBUT, tear breakup time test; CCT, central corneal thickness.

Central corneal thickness in Parkinson’s disease—Aksoy et al.

Patients with tear dysfunction typically report irritationsymptoms that include foreign body sensation, burning, anddryness, as well as vision-related symptoms such as photo-phobia and blurred and fluctuating vision.17,18 The aqueouslayer of the tear film is normally isotonic or slightly hypotonic.In all forms of dry eye syndrome, however, there is an increasein the osmolarity of the tear film, and hypertonic tear filmcauses a decrease in the corneal thickness.18,19 A chronic stateof desiccation and immune activation in dry eye syndrome maycontribute to the observed corneal thinning.17,20 A reduced BRand reduced tear quality in the early stages of the disease, andthe dry eye symptoms that result from impairments in tearproduction as the disease manifests, may have contributed to adecrease in the CCT in our patients with PD.

In addition, corneal epithelial disruptions, changes incorneal thickness, corneal endothelial damage, and cornealedema are reported in patients with PD who are on aman-tadine therapy for a long period.21–23 Patients on amanta-dine or other drugs that are known to affect the cornea wereexcluded from our study. Only patients using L-dopa anddopamine agonists, which have no significant effect on thelacrimal glands, were enrolled in the study.8

A high incidence of glaucomatous visual field defects andenlarged cup-to-disk ratios has been reported in patients withAlzheimer’s disease and PD. In addition, the identification ofprobable glaucoma in 24.5% of patients with Alzheimer’sdisease and 23.7% of patients with PD was significantlyhigher than in the normal population.24,25 CCT has animportant impact on the intraocular pressure (IOP) measure-ment when diagnosing and treating glaucoma, and eyes witha greater mean CCT tend to have a higher IOP. Because anincreased CCT may lead to falsely high IOP readings, andthin corneas may lead to falsely low IOP readings, thereshould be a correlation between the CCT and baselineIOP.26–29 Accumulating evidence suggests that the CCT is acrucial ocular parameter in glaucoma that should be meas-ured in patients of advanced age.30

The relationship between the CCT and PD severity hasnot yet been established. Our findings suggest that BR andtear quality are impaired even in the early stages of PD.The Schirmer’s test and CCT values also decrease as thedisease becomes more severe. We found a significantnegative correlation between the H-Y scores and theCCT values. PD and glaucoma have a higher prevalencewith increasing age, and glaucoma tends to occur morefrequently in patients with PD. Therefore, the possibilityof a thin cornea should be taken into consideration whenmeasuring the IOP in patients with severe PD.

Disclosure: The authors have no proprietary or commercialinterest in any materials discussed in this article.

Acknowledgements: The authors are grateful to Dr. IlkerEtikan, Biostatistics Department of Gaziosmanpasa University,for his contribution on the statistical analysis.

REFERENCES

1. Witjas T, Kaphan E, Azulay JP, et al. Nonmotor fluctuationsin Parkinson’s disease: frequent and disabling. Neurology. 2002;59:408-13.

2. Bonnet AM, Jutras MF, Czernecki V, et al. Nonmotor symptoms inParkinson’s disease in 2012: relevant clinical aspects. Parkinson’s Dis.2012;2012:1-15.

3. Biousse V, Skibell BC, Watts RL, et al. Ophthalmologic features ofParkinson’s disease. Neurology. 2004;62:177-80.

4. Zesiewicz TA, Cimino CR, Malek AR, et al. Driving safety inParkinson’s disease. Neurology. 2002;59:1787-8.

5. Clark D, Eggenberger E. Neuro-ophthalmology of movementdisorders. Curr Opin Ophthalmol. 2012;23:491-6.

6. Pieri V, Diederich NJ, Raman R, Goetz CG. Decreased colordiscrimination and contrast sensitivity in Parkinson’s disease.J Neurol Sci. 2000;172:7-11.

7. Liu Z, Pflugfelder SC. Corneal thickness is reduced in dry eye.Cornea. 1999;18:403-7.

8. Tamer C, Melek IM, Duman T, et al. Tear film tests in Parkinson’sdisease patients. Ophthalmology. 2005;112:1795.

9. Reddy VC, Patel SV, Hodge DO, Leavitt JA. Corneal sensitivity,blink rate, and corneal nerve density in progressive supranuclearpalsy and Parkinson disease. Cornea. 2013;32:631-5.

10. Fitzpatrick E, Hohl N, Silburn P, O’Gorman C, Broadley SA. Case-control study of blink rate in Parkinson’s disease under differentconditions. J Neurol. 2012;259:739-44.

11. Goetz CG, Poewe W, Rascol O, et al. Movement Disorder SocietyTask Force on Rating Scales for Parkinson’s Disease. MovementDisorder Society Task Force report on the Hoehn and Yahr stagingscale: status and recommendations. Mov Disord. 2004;19:1020-8.

12. Agostino R, Bologna M, Dinapoli L, et al. Voluntary, spontaneous,and reflex blinking in Parkinson’s disease. Mov Disord. 2008;23:669-75.

13. Karson CN, LeWitt PA, Calne DB, et al. Blink rates in parkinson-ism. Ann Neurol. 1982;12:580-3.

14. Moss SE, Klein R, Klein BE. Prevalence of and risk factors for dryeye syndrome. Arch Ophthalmol. 2000;118:1264-8.

15. Huang FC, Tseng SH, Shih MH, et al. Effect of artificial tears oncorneal surface regularity, contrast sensitivity, and glare disability indry eyes. Ophthalmology. 2002;109:1934-40.

16. Bye L, Vamadevan P, Chaudhuri KR. Ophthalmological aspectsof Parkinson’s disease. In: Chaudhuri KR, Tolosa E, Schapira A,Poewe W, eds. Non-motor Symptoms of Parkinson’s Disease. NewYork: Oxford University Press; 2009. 297-308.

17. Stephen C. Pflugfelder tear dysfunction and the cornea: LXVII EdwardJackson Memorial Lecture. Am J Ophthalmol. 2011;152:900-9.

18. Horwath-Winter J, Berghold A, Schmut O, et al. Evaluation of the clinicalcourse of dry eye syndrome. Arch Ophthalmol. 2003;121:1364-8.

19. Karadayi K, Ciftci F, Akin T, et al. Increase in central cornealthickness in dry and normal eyes with application of artificial tears: anew diagnostic and follow-up criterion for dry eye. OphthalmicPhysiol Opt. 2005;25:485-91.

20. Dayanir V, Sakarya R, Ozcura F, et al. Effect of corneal drying oncentral corneal thickness. J Glaucoma. 2004;13:6-8.

21. Chang KC, Jeong JH, Kim MK, et al. The effect of amantadine oncorneal endothelium in subjects with Parkinson’s disease. Ophthal-mology. 2010;117:1214-9.

22. Kubo S, Iwatake A, Ebihara N, et al. Visual impairment inParkinson’s disease treated with amantadine: case report and reviewof the literature. Parkinsonism Relat Disord. 2008;14:166-9.

23. Dubow JS, Rezak M, Berman AA. Reversible corneal edema asso-ciated with amantadine use: an unrecognized problem. Mov Disord.2008;23:2096-7.

24. Bayer AU, Keller ON, Ferrari F, et al. Association of glaucomawith neurodegenerative diseases with apoptotic cell death: Alz-heimer’s disease and Parkinson’s disease. Am J Ophthalmol. 2002;133:135-7.

25. Yenice O, Onal S, Midi I, et al. Visual field analysis in patients withParkinson’s disease. Parkinsonism Relat Disord. 2008;14:193-8.

26. Su DH, Wong TY, Foster PJ, et al. Central corneal thickness and itsassociations with ocular and systemic factors: the Singapore MalayEye Study. Am J Ophthalmol. 2009;147:709-16.

CAN J OPHTHALMOL—VOL. 49, NO. 2, APRIL 2014 155

Central corneal thickness in Parkinson’s disease—Aksoy et al.

27. Doughty MJ, Zaman ML. Human corneal thickness and its impacton intraocular pressure measures: a review and meta-analysisapproach. Surv Ophthalmol. 2000;44:367-408.

28. Ozdamar Y, Cankaya B, Ozalp S, et al. Is there a correlation betweendiabetes mellitus and central corneal thickness? J Glaucoma. 2010;19:613-6.

156 CAN J OPHTHALMOL—VOL. 49, NO. 2, APRIL 2014

29. Shah S, Chatterjee A, Mathai M, et al. Relationship between cornealthickness and measured intraocular pressure in a general ophthal-mology clinic. Ophthalmology. 1999;106:2154-60.

30. Guedes G, Tsai JC, Loewen NA. Glaucoma and aging. Curr AgingSci. 2011;4:110-7.