Central corneal thickness in highmyopia

Lene Pedersen, Jesper Hjortdal and Niels Ehlers

Department of Ophthalmology, Aarhus University Hospital, Arhus Sygehus, Arhus,

Denmark

ABSTRACT.

Purpose: To examine if central corneal thickness (CCT) is different in emme-

tropia and high myopia.

Methods: 57 emmetropic subjects (0 to + 1.5 D) and 48 high myopes (all more

than – 6 D in spherical equivalent refraction) were studied. CCT was measured

by a Haag–Streit Optical Low-Coherence Reflectometry (OLCR) pachymeter,

a recently developed high precision pachymeter with a standard deviation (SD)

for repeated measurements of 1 mm.

Results: Mean CCT for the emmetropic group was 538.6 mm (SD = 32.1), and for

the myopic group 527.7 mm (SD = 35.0). Neither the mean CCT nor the variance

from the two groups showed a statistically significant different (p > 0.05).

Conclusion: CCT is not systematically altered in myopia. The process by which

the myopia progresses does not to a measurable degree influence the central

cornea.

Key words: central corneal thickness – high myopia – optical low - coherence reflectometry –

pachymeter

Acta Ophthalmol. Scand. 2005: 83: 539–542Copyright # Acta Ophthalmol Scand 2005.

doi: 10.1111/j.1600-0420.2005.00498.x

Introduction

Although the true aetiology of myopiais still unknown (Goldschmidt 2003),the cornea is responsible for approxi-mately two-thirds of optical refractionand its role in myopia has consequentlybeen studied intensely over the years.Known possible changes in the highlymyopic eye are all located in the pos-terior segment: staphyloma, myopicconus, choroidal atrophy, thinning ofthe retina and sclera. Changes in theanterior segment associated with myo-pia are still under debate. Carney et al.(1997), among others, found that themyopic cornea has a steeper central

corneal curvature, while Chang et al.(2001) found no correlation betweencorneal curvature and central cornealthickness (CCT). The myopic eye isknown to be longer than the normalemmetropic eye (see, e.g. Touzeauet al. 2003). If this is the result of gen-eral growth, one might expect the cor-nea to have grown to be thicker than isnormal, in which case a correlationwith body mass index (BMI) mightexist. If instead the myopic eye is largerdue to a mechanism similar to that of aballoon being inflated, one wouldexpect the cornea to be thinner thannormal, according to a simple

‘stretching theory’. An emmetropiceye could then be compared to asphere, and a myopic eye to a prolatespheroid.

von Bahr (1956), Price et al. (1999)and Touzeau et al. (2003), amongothers, looked for a connection betweenCCT and myopia. Measurements weretaken with different types of pachy-meters and with different set-ups, andwith inconclusive results (Table 1).

Today, there are many methods ofmeasuring CCT (Ehlers & Hjortdal2004), primarily based upon opticaland ultrasound principles.

The precision (standard deviation,SD) of manual optical pachymetry is8 mm, with intraobserver errors of5–6 mm and interobserver errors of20 mm (Olsen et al. 1980). Ultrasoundpachymetry has better precision (Tam& Rootman 2003), but results mayvary due to applanation force and dif-ferences between types of instruments(Salz et al. 1983). The Orbscan system() is an automatic and non-contactoptical pachymeter and topographwith a precision similar to that of man-ual optical pachymetry (Yaylali et al.1997).

The apparatus used in this study wasan optical low-coherence reflectometry(OLCR) pachymeter. (Bohnke et al.(1999) reported a precision of about1 mm (SD), an intrasession reproducibil-ity around 0.9–1.2 mm and a high inter-session reproducibility. Other studieshave compared the OLCR pachymeterwith the ultrasound pachymeter. Walti(1998) found the OLCR pachymeter tobe faster and more precise (lower SD)than an ultrasound pachymeter. TheCCT values were found to be

ACTA OPHTHALMOLOGICA SCANDINAVICA 2005

539

significantly higher when measuredwith an ultrasound pachymeter thanwhen measured with the OLCR pachy-meter (Ventura et al. 2001; Genth et al.2002). Today, the OLCR seems to bethe most precise instrument to measureCCT.

The purpose of this study was todetermine whether CCT differs inemmetropia and high myopia usingthe OLCR pachymeter. Documenteddifferences between these groupsmight contribute to the ongoing discus-sion on the aetiology and pathogenesisof myopia.

Material and Methods

Subjects

All subjects were white and agedbetween 18 and 55 years. Subjectswith previous eye surgery, glaucoma,diabetes mellitus or other acute orchronic diseases possibly affecting thecorneal thickness were excluded.

The emmetropic group included 57subjects, all with normal visual acuityand refraction (self-reported to be from0 toþ 1.5 D spheq).Thegroupconsistedof volunteers sourced from hospital staff,students and patients’ relatives.

The myopic group was recruited fromsubjects referred for refractive surgeryfor high myopia. Only eyes with a mini-mum of � 6 D in spherical equivalent

refraction and a maximum of 2 D cor-neal astigmatism were accepted.

The refractive criteria were set toensure that only highly myopic eyeswere included in the group and thatcases with subclinical keratoconuswere excluded. On this basis, the studyincluded 48 highly myopic subjects withan average subjective refraction of �8.93 D (ranging from � 6 to � 15.75D, SD ¼ 2.02 D) in spherical equiva-lent and an average astigmatism of �0.81 D cylinder.

Those who normally used contactlenses (24 subjects) did not use them forat least 48 hours before examination.

The subjects in both groups answeredquestions on height and weight, fromwhich BMI was calculated.

Central corneal thickness and axiallength were measured in both eyes,without touching or medicating theeyes, but data from only one eye fromeach subject was included. The datapertained to the right eye in 99 subjects,but to the left eye in six subjects. Thiswas due to previous retinal detachmentin the right eye in one case, a previoussevere trauma in the right eye in onecase, the fact that the right eye did notfulfil the refractive criteria in threecases, and because the right eye wasunder topical anaesthetic in one case.

Table 2 summarizes data character-izing the two groups, the only statistic-ally significant difference being the

axial length, which was 3 mm longerin the myopic group.

Equipment

The OLCR pachymeter was attachedto a BC 900 slit-lamp (both fromHaag-Streit,) (Ballif et al. 1997; Genthet al. 2002). The OLCR pachymeterwas calibrated and used as describedin the instruction manual. A cornealrefractive index of n ¼ 1.376 wasassumed. The OLCR pachymeter isan automatic non-contact pachymeterbased on interferometry. The pachymeteris easy to use and an average CCTmeasurement (based on 20 scans) takes28 seconds including adjustment of theapparatus (Walti 1998).

All measurements were performedby the same examiner (LP). EachOLCR measurement is the result of 20consecutive scans, where the five upperand five lower scan values are deleted.The OLCR pachymeter calculates amean CCT and a standard deviationfrom the remaining 10 scans.

To confirm the reported precision ofthe OLCR pachymeter, the right eyesof five normal subjects were measured10 times. Each subject was repositionedbetween each measurement. The mea-surements of one subject were com-pleted within 30 mins. The averageintersession precision (SD) for five sub-jects was 1.0 mm corresponding with

Table 1. Overview of previously published papers with information on myopia and central corneal thickness.

Authors and year Country Equipment No of subjects Refractive range Results

CCT and myopia

Total Myopic

Kunert et al. 2003 India Ultrasound/

Orbscan

615 615 Up to � 20 D sph

(< 3 D cyl)

Thicker CCT in high myopic

Touzeau et al. 2003 France Orbscan 95 95 eyes þ 9.16 to

�19.23 D spheq

Thinner CCT when myopic

Srivannaboon 2002 Thailand Orbscan 280 280 � 0.5 to � 18 D sph Thinner CCT in high myopic

Chang et al. 2001 Taiwan Ultrasound 216 ? Roughly þ 7

to � 22 D spheq

Thinner in high myopic

Liu & Pflugfelder 2000 China Orbscan 30 30 � 0.75 to � 10.25 D sph No correlation

Cho & Lam 1999 China Ultrasound 151 ? þ 1.63 to � 13.50 D spheq No correlation (including high myopic)

Price et al. 1999 USA Ultrasound 450 ? 0 to � 30 D sph No correlation

Tanaka et al. 1996 Brazil Ultrasound 70 25 þ 3.2 to � 25.5 D spheq No correlation

Alsbirk 1978 Greenland Optical 325 ? ? Thinner CCT when myopic

Ehlers & Hansen 1976 Denmark Optical 101 ? ? No correlation

Hansen 1971 Denmark Optical 113 ? þ 3.0 up to � 5 D No correlation

Martola & Baum 1968 USA Optical 121 ? Up to more than � 6 D No correlation

(Boston) (tendency to be thicker)

von Bahr 1956 Sweden Optical 125 12 eyes More than þ 3 to more

than � 4 D

Thinner CCT when myopic (> � 4 D)

Blix 1880 Sweden Optical 8 2 Hypermetropic to myopic No difference

ACTA OPHTHALMOLOGICA SCANDINAVICA 2005

540

previous findings (Bohnke et al. 1999).High and low CCT values were deter-mined with the same SD.

Ethical approval

The study was conducted in accord-ance with a protocol approved by thelocal ethics committee of Arhus inaccordance with the Helsinki DeclarationII. Informed consent was obtainedfrom all persons included in the study.

Statistics

The material was tested by kurtosis andskewness tests and was found not todeviate from a normal distribution.Paired t-test was used to establishwhether there was any differencebetween right and left eyes. Student’st-tests and F-tests were used to com-pare mean values and variancesbetween the two groups.

Results

The mean CCT for the emmetropicgroup was 538.6 mm (SD ¼ 32.1 mm,range 459.9–606.0 mm), and for the myo-pic group 527.7 mm (SD ¼ 35.0 mm,range 452.2–599.5 mm). The differenceof 10.9 mm was not statistically signifi-cant different from zero (p> 0.05). TheF-test showed no statistical differencebetween the variances from the twogroups.

The mean CCT for the 24 myopicsubjects wearing glasses was 523.4 mm(SD ¼ 36.19 mm) and the mean CCTfor the 24 myopic subjects normallywearing contact lenses was 531.9 mm(SD ¼ 33.93 mm). This difference of8.5 mm was not significantly differentfrom zero (p > 0.05).

There was no statistically significantdifference between the CCT values ofthe right (534.3 mm, SD ¼ 33.6 mm)

and left eyes (534.4 mm, SD ¼33.3 mm) nor was there any statisticallysignificant difference between the CCTvalues from men (534.9 mm, SD ¼30.9 mm) and women (532.0 mm,SD ¼ 37.4 mm).

Discussion

This study showed no statistically sig-nificant difference between the meanCCT of the myopic subjects and thatof the emmetropic subjects. This resultis in agreement with the majority ofprevious studies (Table 1). The mini-mal detectable difference in central cor-neal thickness (MIREDIF) can becalculated to 24.2 mm (for n ¼ 52 ineach group, SD ¼ 33.46 mm, a ¼ 5%[type 1 error] and b ¼ 95% [statisticalpower]) (Sokal & Rohlf 2000). Thus,the present sample sizes should be suf-ficiently large to identify meaningfulreal differences in corneal thickness inemmetropia and myopia.

The results from this study showed adifference of 10.9 mm in mean CCT(the myopic being the thinner), this isaround a quarter of what would beexpected from pure stretching of theocular tunics (see appendix for mathe-matical reflections). If a stretchingmechanism really is active, the thinningseems to be confined to the sclera.

This study included 24 myopic sub-jects who normally wore contact lensesand 24 myopic subjects who normallywore glasses. The mean CCT of thetwo subgroups was not significantlydifferent.

The inconclusive results of previousstudies (Table 1) might be explained byany of the following: pachymeterswith low reproducibility; inexperiencedobservers; no consideration of diurnalvariation; the influence of contactlenses; genetic difference in CCT;different criteria for exclusion; lack of

highly myopic subjects, and too small asample size.

The advantages of this study are:its use of high precision apparatuscompared to traditional ultrasoundand optical pachymeters; the fact thatonly highly myopic subjects with lim-ited astigmatism were included, and theminimized influence of confounders.

Previous studies indicate that theOLCR pachymeter is the most precise(lowest SD) clinically availablepachymeter. It remains unknownwhich pachymeter in general is clo-sest to the real CCT in terms of accu-racy. In the present study, precisionwas considered more important thanaccuracy.

The diurnal variation of cornealthickness has been studied with differ-ent techniques. The results are incon-clusive. Muller-Treiber et al. (2001)studied the diurnal changes with theOLCR pachymeter, and found the cor-nea to be thickest in the morning andaround 5 mm lower in the late afternoon.To minimize this confounder, earlymorning measurements were avoided.The two groups were measured atalmost the same average time of day.

In conclusion, the CCT of myopicand emmetropic eyes do not differ.The growth alterations in the oculartunics of myopic subjects do not toany measurable degree involve the cor-neal thickness.

Acknowledgement

This study was made possible by agrant from the Institute of ClinicalMedicine, Aarhus University, Arhus,Denmark.

ReferencesAlsbirk PH (1978): Corneal thickness. 1. Age

variation, sex difference and oculometric cor-

relations. Acta Ophthalmol Scand 56: 95–104.

von Bahr G (1956): Corneal thickness: its meas-

urement and changes. Am J Ophthalmol 42:

251–266.

Ballif J, Gianotti R, Chavanne P, Walti R &

Salathe RP (1997): Rapid and scalable scans

at 21 m/s in optical low-coherence reflecto-

metry. Optics Lett 22: 757–759.

Blix M (1880): Oftalmometriska studier.

Upsala Lakareforenings Forhandlingar 15:

349–xxx.

Bohnke M, Masters BR, Walti R, Ballif JJ,

Chavanne P, Gianotti R & Salathe RP

(1999): Precision and reproducibility of

Table 2. Data characterizing the two groups. Mean values are shown in the first five lines.

Myopic group Emmetropic group P

Age (years) 37 (SD ¼ 8.8) 36 (SD ¼ 8.6) > 0.05

Height (m) 1.74 (SD ¼ 9.6) 1.76 (SD ¼ 9.3) > 0.05

Weight (kg) 75 (SD ¼ 2.9) 75 (SD ¼ 3.1) > 0.05

BMI (kg/m2) 25 (SD ¼ 0.3) 24 (SD ¼ 0.2) > 0.05

Axial length (mm) 26.52 (SD ¼ 1.265) 23.52 (SD ¼ 0.775) < 0.01

Sex (male/female) 18/30 28/29 –

Examination time 10.18 a.m 11.46 a.m –

ACTA OPHTHALMOLOGICA SCANDINAVICA 2005

541

measurements of human corneal thick-

ness with rapid optical low-coherence

reflectometry (OLCR). J Biomed Opt 4:

152�156.

Carney LG, Mainstone JC & Henderson BA

(1997): Corneal topography and myopia. A

cross-sectional study. Invest Ophthalmol Vis

Sci 38: 311–320.

Chang SW, Tsai IL, Hu FR, Lin LL & Shih YF

(2001): The cornea in young myopic adults.

Br J Ophthalmol 85: 916–920.

Cho P & Lam C (1999): Factors affecting the

central corneal thickness of Hong Kong

Chinese. Curr Eye Res 18: 368–374.

Doughty MJ & Zaman ML (2000): Human

corneal thickness and its impact on intrao-

cular pressure measures: a review and meta-

analysis approach. Surv Ophthalmol 44:

367–408.

Ehlers N & Hansen FK (1976): Further data

on biometric correlations of central corneal

thickness. Acta Ophthalmol Scand 54:

774–778.

Ehlers N & Hjortdal J (2004): Corneal thick-

ness: measurement and implications. Exp

Eye Res 78: 543–548.

Genth U,MrochenM,Walti R, Salaheldine MM

& Seiler T (2002): Optical low coherence

reflectometry for non-contact measure-

ments of flap thickness during laser in situ

keratomileusis. Ophthalmology 109:

973–978.

Goldschmidt E (2003): The mystery of myo-

pia. Acta Ophthalmol Scand 81: 431–436.

Hansen FK (1971): Clinical study of normal

human central corneal thickness. Acta

Ophthalmol Scand 49: 82–89.

Kunert KS, Bhartiya P, Tandon R, Dada T,

Christian H & Vajpayee RB (2003): Central

corneal thickness in Indian patients under-

going LASIK for myopia. J Refract Surg 19:

378–379.

Liu Z & Pflugfelder SC (2000): The effects of

longterm contact lens wear on corneal thick-

ness, curvature and surface regularity.

Ophthalmology 107: 105–111.

Martola EL & Baum JL (1968): Central and

peripheral corneal thickness – a clinical

study. Arch Ophthalmol 79: 28–xx.

Muller-Treiber A, Joos RE, Bussacker H,

Schmid P & Trachsel M (2001): Veranderung

der Corneadicke unter Silicon-Hydrogellinsen

gemessen mit dem Breitbandreflektometer.

Die Kontaktlinse 7–8: 11–16.

Olsen T, Nielsen CB & Ehlers N (1980): On the

optical measurement of corneal thickness. 1.

Optical-principle and sources of error. Acta

Ophthalmol Scand 58: 760–766.

Price FW Jr, Koller DL & Price MO (1999):

Central corneal pachymetry in patients

undergoing laser in situ keratomileusis.

Ophthalmology 106: 2216–2220.

Salz JJ, Azen SP, Berstein J, Caroline P,

Villasenor RA & Schanzlin DJ (1983):

Evaluation and comparison of sources of

variability in the measurement of corneal

thickness with ultrasonic and optical pachy-

meters. Ophthalmic Surg Lasers 14: 750–754.

Sokal RR & Rohlf FJ (2000): Biometry. 3rd

edn. New York: WH Freeman & Co.

Srivannaboon S (2002): Relationship between

corneal thickness and level of myopia.

J Med Assoc Thai 85: 162–166.

Tam ES & Rootman DS (2003): Comparison

of central corneal thickness measurements

by specular microscopy, ultrasound pachy-

metry and ultrasound biomicroscopy.

J Cataract Refract Surg 29: 1179–1184.

Tanaka HM, Mori ES, Maia N, Freitas D,

Campos M & Chamon W (1996): Corneal

thickness in high myopes. Invest Ophthalmol

Vis Sci 37: 2566–2566.

Touzeau O, Allouch C, Borderie V, Kopito R

& Laroche L (2003): Correlation entre la

refraction et la biometrie oculaire. J Fr

Ophtalmol 26: 355–363.

Ventura AC, Bohnke M & Mojon DS (2001):

Central corneal thickness measurements in

patients with normal tension glaucoma, pri-

mary open-angle glaucoma, pseudoexfolia-

tion glaucoma or ocular hypertension. Br J

Ophthalmol 85: 792–795.

Walti R (1998): Rapid and precise in vivo

measurement of human corneal thickness

with optical low-coherence reflectometry in

normal human eyes. J Biomed Opt 3: 253–258.

Yaylali V, Kaufman SC & Thompson HW

(1997): Corneal thickness measurements

with the Orbscan topography system and

ultrasonic pachymetry. J Cataract Refract

Surg 23: 1345–1350.

Received on November 9th, 2004.

Accepted on April 6th, 2005.

Correspondence:

Lene Pedersen

Department of Ophthalmology

Aarhus University Hospital

Arhus Sygehus

Nørrebrogade 44

8000 Arhus

Denmark

Tel: þ 45 89 49 32 22

Fax: þ 45 86 12 16 53

Email: lenep74@hotmail.com

Appendix

An emmetropic eye is compared to asphere (surface area [Semme] ¼ 4pr2),and a myopic eye to a prolatespheroid (surface area [Smyop] seewww.mathworld.wolfram.com). If thevolume (V) of the cornea þ sclera isconstant:

V ¼ Smyop � CCTmyop

¼ Semme � CCTemme

ðIÞ

V ¼ 2pa2 þ 2pac2ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffic2 � a2

p sin�1

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffic2 � a2

p

c

!

� CCTmyop ¼ 4pr2 � CCTemme

ðIIÞ

The parameters ‘r’ (radius) and ‘a’(equatorial radius) are estimated to bealike and are calculated as half meanaxial length in an emmetropic group(23.24 mm/2) (Touzeau et al. 2003).The parameter ‘c’ (polar radius) is cal-culated as half of the mean axiallength in a highly myopic group(26.36 mm/2) (Touzeau et al. 2003).A meta-analysis finds the averageCCT in normal populations to be535 mm (SD ¼ 31 mm) (Doughty &Zaman 2000). If the parameters men-tioned above are inserted in equationII, CCTmyop is calculated to be490.5 mm or 44.5 mm (8%) thinnerthan CCTemme.

ACTA OPHTHALMOLOGICA SCANDINAVICA 2005

542