Corneal Thickness Measurements and VisualFunction Abnormalities in Ocular

Hypertensive Patients

FELIPE A. MEDEIROS, MD, PAMELA A. SAMPLE, PHD, AND ROBERT N. WEINREB, MD

● PURPOSE: It has been suggested that a considerablesubset of patients currently classified as having ocularhypertension may have thicker than average corneas thatresult in an overestimation of the true intraocular pres-sure (IOP). As a consequence, ocular hypertensionpatients with greater corneal thickness may be at a lowerrisk for functional damage, such as that detected byshort-wavelength automated perimetry (SWAP). Thepurpose of this study was to evaluate the frequency ofSWAP deficits in ocular hypertension patients and tocorrelate these findings with corneal thickness measure-ments in the same patients.● DESIGN: Cross-sectional observational study.● METHODS: Sixty-eight ocular hypertension patientswith normal optic disks and 63 normal subjects wereincluded in the study. All participants underwent stan-dard automated perimetry (SAP), SWAP, and centralcorneal thickness measurements using ultrasoundpachymeter. Central corneal thickness measurements inocular hypertension patients with abnormal visual fieldtest results were compared with central corneal thicknessmeasurements in ocular hypertension patients with nor-mal visual field results. In addition, central cornealthickness measurements in ocular hypertension patientswere compared with central corneal thickness measure-ments in normal subjects.● RESULTS: Sixteen of 68 patients with ocular hyperten-sion (24%) demonstrated SWAP abnormalities, whereasfour of 68 (6%) showed a deficit on SAP. The meancentral corneal thickness in ocular hypertension patientswith abnormal SWAP results was significantly lowerthan the mean central corneal thickness in ocular hyper-tension patients with normal SWAP results (545 � 25

�m vs 572 � 35 �m; P � .006). The mean centralcorneal thickness in the normal group was 557 � 33 �m.The mean central corneal thickness in ocular hyperten-sion patients with normal SWAP results was significantlyhigher than in normal subjects (P � .02). There was nosignificant difference between mean central cornealthickness in normal subjects and in ocular hypertensionpatients with abnormal SWAP results (P � .19).● CONCLUSIONS: The patients classified as having ocularhypertension but with visual field loss detected by SWAPhad significantly lower central corneal thickness mea-surements than the ocular hypertension patients withnormal visual field results. These results suggest thatcentral corneal thickness should be taken into accountwhen assessing risk for the development of glaucomatousdamage among ocular hypertension patients. (Am JOphthalmol 2003;135:131–137. © 2003 by ElsevierScience Inc. All rights reserved.)

T HE INFLUENCE OF CENTRAL CORNEAL THICKNESS ON

intraocular pressure (IOP) measurements usingGoldmann applanation tonometry is well recog-

nized.1–3 Intraocular pressure may be overestimated orunderestimated in thick or thin corneas, respectively. Inrecent years, there have been several reports in theliterature of increased central corneal thickness in patientsdiagnosed with ocular hypertension.4–10 Argus5 examined36 ocular hypertensive patients and compared their centralcorneal thickness measurements with 29 normal subjectsand 31 glaucomatous patients. He found a significantlygreater corneal thickness in the patients with ocularhypertension compared with the other groups. In a recentreport of the Ocular Hypertension Treatment Study(OHTS) central corneal thickness measurements werereported for 1,301 ocular hypertension subjects.9 Themean central corneal thickness was 573 � 39 �m, and24% of the patients had central corneal thickness greaterthan 600 �m. Although this report did not includemeasurements of normal subjects, a recent meta-analysis ofthe literature found that the mean corneal thickness ofeyes reported as normal was 544 �m, and less than 5% of

Accepted for publication Aug 29, 2002.From the Hamilton Glaucoma Center and Visual Function Laboratory,

Department of Ophthalmology, University of California, San Diego, LaJolla, California.

This study was supported in part by National Eye Institute (NEI) EY08208.

Inquiries to Robert N. Weinreb, MD, Hamilton Glaucoma Center,University of California, San Diego, 9500 Gilman Dr, La Jolla, CA92093-0946; fax: (619) 534-1625; e-mail: weinreb@eyecenter.ucsd.edu

© 2003 BY ELSEVIER SCIENCE INC. ALL RIGHTS RESERVED.0002-9394/03/$30.00 131PII S0002-9394(02)01886-X

normal subjects are expected to show corneal thicknessgreater than 600 �m.10

These studies suggest that a considerable subset ofpatients currently classified as having ocular hypertensionmay simply have thicker than average corneas, whichresults in an overestimation of what is likely a normal, trueIOP. As a consequence, ocular hypertension patients withgreater corneal thickness may be at a much lower risk forglaucoma development than previously recognized.

In contrast, the development of more sensitive methodsfor detection of structural and functional damage inglaucoma has provided evidence that some patients previ-ously classified as having ocular hypertension have, in fact,initial visual field damage not detected by conventionalmethods. We know that visual field testing with standardautomated perimetry (SAP) is not selective for a particularganglion cell type. Further, newer tests that attempt toisolate specific subpopulations of ganglion cells by evalu-ating specific visual functions have shown promise fordetecting early glaucomatous functional deficits. One suchtest, short-wavelength automated perimetry (SWAP),which targets the small bistratified ganglion cells, has beenshown to be able to detect visual field loss in glaucomapatients 3 to 5 years sooner than SAP.11–22 Publishedstudies have reported that 8% to 30% of ocular hyperten-sion patients with normal SAP fields have visual fielddefects when examined by SWAP.23,24

This evidence collectively suggests that patients cur-rently diagnosed as having ocular hypertension may in-clude some in whom intraocular pressure has beenoverestimated as well as some with early glaucomatousdamage. Hence, we may speculate that ocular hypertensionpatients with thicker corneas (and overestimated IOP) aremore likely to show normal results on functional tests.

The purpose of this study is to evaluate the frequency ofSWAP deficits in ocular hypertension patients and tocorrelate these findings with corneal thickness measure-ments in the same patients.

METHODS

ALL PATIENTS WERE EVALUATED AT THE HAMILTON GLAU-

coma Center, University of California, San Diego, andretrospectively selected from our research database. Thesepatients were part of a prospective longitudinal studydesigned to evaluate optic nerve structure and visualfunction in glaucoma. All patients who met the inclusioncriteria described were enrolled in the current study.Informed consent was obtained from all participants.

The design of the current study was a cross-sectionalobservational one. Sixty-three normal subjects and 68ocular hypertension patients meeting entry criteria wereenrolled. Mean age � SD for normal subjects and ocularhypertension patients was 53 � 15 years and 64 � 11years, respectively (P � .001; Student t test). Among the

ocular hypertension patients, 63 were Caucasian, one wasAfrican American, and three were Hispanic; 30 were maleand 38 were female. Among the normal subjects, 53 wereCaucasian, three were African American, five were His-panic, and two were Asian; 22 were male and 41 werefemale. One eye of each patient was selected randomly forexamination. Each subject underwent a comprehensiveophthalmic examination, including review of medicalhistory, best-corrected visual acuity, slit-lamp biomicros-copy, Goldmann applanation tonometry, gonioscopy, di-lated fundoscopic examination using a 78-diopter lens, andstereoscopic optic disk photography.

Two perimetric procedures, SAP and SWAP, were usedto test visual function. The tests were all measured withinan interval not longer than 6 months. Stereoscopic opticdisk photographs were obtained within plus or minus 6months of visual function tests. All patients had centralcorneal thickness measurements using ultrasound pachym-etry (Pachette GDH 500; DGH Technology, Philadelphia,Pennsylvania, USA).

To be included, all subjects had to have best-correctedvisual acuity of 20/40 or better, spherical refraction within� 5.0 diopters, and cylinder correction within � 3.0diopters and open angles on gonioscopy. Evaluation ofstructural damage to the optic disk was based on assess-ment of simultaneous stereoscopic optic disk photographs(TRC-SS; Topcon Instrument Corp of America, Paramus,New Jersey, USA). The photographs were graded by twoexperienced graders, and each grader was masked to thesubject’s identity and to the other test results. All includedphotographs were judged to be of adequate quality orbetter.

Ocular hypertension patients met the following criteria:intraocular pressure 23 mm Hg or greater and less than 32mm Hg on at least two occasions and normal-appearingoptic disks and retinal nerve fiber layer (no diffuse or focalrim thinning, hemorrhage, cupping, or nerve fiber layerdefects indicative of glaucoma or other ocular patholo-gies). For this study, visual field results were not consideredas part of the definition of ocular hypertension.

Normal controls eyes had intraocular pressures of 22 mmHg or less with no history of increased IOP and normaloptic disks based on grading of stereoscopic optic diskphotographs. Normal subjects also had normal SAP andSWAP visual fields according to the criteria provided.Average mean deviation (MD) and pattern standard de-viation (PSD) of SAP fields in normal subjects were 0.28� 1.14 dB and 1.58 � 0.21 dB, respectively. Average MDand PSD of SWAP fields were 0.52 � 2.25 dB and 2.36 �0.41 dB, respectively.

To be included, all subjects had to have reliable visualfield results on all tests. This was defined as 25% or fewerfalse-positive results, false-negative results, and fixationlosses. Normal and ocular hypertension subjects wereexcluded if they had a history of intraocular surgery(except for uncomplicated cataract surgery). Patients with

AMERICAN JOURNAL OF OPHTHALMOLOGY132 FEBRUARY 2003

secondary causes of elevated intraocular pressure (forexample, iridocyclitis, trauma), other intraocular eye dis-ease, other diseases affecting visual field (for example,demyelinating diseases, pituitary lesions, diabetic retinop-athy) were excluded.

Standard achromatic automated perimetry testing wasperformed using program 24-2 (Zeiss-Humphrey, San Le-andro, California, USA). The two locations just above andbelow the blind spot were not included in the analysis.Short-wavelength automated perimetry testing was per-formed as described elsewhere.15 In brief, SWAP is amodification of SAP in which a 440-nm, narrow-band,1.8-degree target at 200 ms duration is projected onto a100-cd/m2 yellow adaptation background field to selec-tively stress short-wavelength cones and their connections.

Risk factors for glaucoma, including highest intraocularpressure, African American race, family history of glau-coma in a primary relative, older age, and presence ofsystemic hypertension, coronary artery or vascular disease,or migraine were assessed in each patient.

Standard automated perimetry and SWAP abnormali-ties were determined by comparison with our normativedatabase (n � 348) and using Statpac-like Analysis forGlaucoma Evaluation (SAGE) criteria.25 SAGE criteriafor abnormal are a PSD worse than 1% probability level, aglaucoma hemifield test (GHT) result outside normallimits, one hemifield cluster worse than the 1% probabilitylevel, two hemifield clusters worse than 5% probabilitylevel, or four points worse than 5% probability level onpattern deviation. The hemifield clusters are the onesdefined in GHT analysis. All abnormalities must be con-firmed on a subsequent test. These criteria were derivedfrom an independent study that evaluated the performanceof a wide variety of criteria for detecting early glaucoma-tous visual field progression for SAP and SWAP.25 In thatstudy, three populations were evaluated: (1) 348 normalsubjects, tested to develop normative databases and statis-tical analysis packages for SAP and SWAP; (2) an inde-

pendent group of 47 normal subjects with longitudinalfollow-up, to determine specificity of different criteria; and(3) 295 patients with elevated IOP and normal baselineSAP to determine the sensitivity of different criteria fordetecting early glaucomatous visual field loss. The abovecriteria demonstrated high specificity for correctly identi-fying eyes with normal visual fields (98% to 100%) forboth SAP and SWAP and also provided the highestsensitivity for detection of conversion from a normal to aglaucomatous visual field in patients with elevated IOP.

Central corneal thickness measurements in ocular hy-pertension patients with abnormal visual field test resultswere compared with central corneal thickness measure-ments in ocular hypertension patients with normal visualfield results. Age, highest recorded IOP, visual field indices,and central corneal thickness measurements were com-pared using the Student t test. For proportions, such as thepresence of risk factors and male sex, statistical significancewas determined using the Fisher exact test. A P value lessthan .05 was considered statistically significant.

RESULTS

ACCORDING TO THE PERIMETRIC CRITERIA DESCRIBED, 16

of 68 patients with ocular hypertension (24%) demon-strated SWAP abnormalities, whereas four of 68 (6%)showed a deficit on SAP.

Demographics and risk factors of ocular hypertensionpatients with normal and abnormal SWAP results arepresented in Table 1. There was no statistically significantdifference in age between ocular hypertension patientswith normal and abnormal SWAP results. The highestrecorded IOP (mean � SD) of 68 ocular hypertensionpatients was 28 � 3.7 mm Hg. There was no statisticallysignificant difference between the highest recorded IOP inpatients with abnormal or normal SWAP results (Table 1).There was no statistically significant difference in the

TABLE 1. Demographics and Risk Factors of 68 Ocular Hypertensive Patients With Normaland Abnormal Short-wavelength Automated Perimetry Results*

Variable

Normal

(n � 52 eyes)

Abnormal

(n � 16 eyes) P Value

Age (years) 65 � 12 64 � 11 .87

Male 24 (46) 9 (56) .57

Highest intraocular pressure

(mean � SD, mm Hg)

27.8 � 3.9 28.6 � 3.3 .76

African American 1 (2) 0 .99

Family history 30 (57) 11 (69) .56

Systemic hypertension 13 (25) 3 (19) .74

Coronary disease/vascular disease 9 (17) 3 (19) .99

Migraine 4 (8) 1 (6) .99

*Values are expressed as a number (percentage) unless otherwise specified.

CORNEAL THICKNESS IN OCULAR HYPERTENSIONVOL. 135, NO. 2 133

number of patients of African American race, familyhistory of glaucoma, systemic hypertension, or coronaryartery disease or vascular disease between patients with anormal or abnormal SWAP result. However, given thesmall prevalence of risk factors, the power to detect adifference in a demographic or risk factor was less than0.30 in all analyses.

Visual field indices and central corneal thickness mea-surements are provided in Table 2. The mean � SD of theMD and PSD of ocular hypertension patients with abnor-mal SWAP results were �2.53 � 3.94 dB and 3.40 � 0.86dB, respectively. The corresponding values for ocularhypertension patients with normal SWAP results were0.06 � 2.94 dB and 2.64 � 0.60 dB. The mean MD andPSD values for the ocular hypertension patients withabnormal SWAP results were statistically significant dif-ferent than the values for the ocular hypertension patientswith normal SWAP results.

Mean central corneal thickness was 565 � 35 �mamong all 68 ocular hypertension patients. Mean centralcorneal thickness in ocular hypertension patients withabnormal SWAP results was significantly lower than themean central corneal thickness in ocular hypertensionpatients with normal SWAP results (545 � 25 �m vs 572� 35 �m; P � .006). The mean central corneal thicknessin the normal group was 557 � 33 �m. The mean centralcorneal thickness in ocular hypertension patients withnormal SWAP results was significantly higher than innormal subjects (P � .020). There was no statisticallysignificant difference between mean central corneal thick-ness in normal subjects and in ocular hypertension patientswith abnormal SWAP results (P � .19). Figure 1 shows thedistribution of central corneal thickness measurements inocular hypertension and normal subjects. Thirteen of 52ocular hypertension patients (25%) with normal SWAPresults had central corneal thickness measurements greaterthan 600 �m, whereas none of the ocular hypertensionpatients with abnormal SWAP results had central cornealthickness greater than 600 �m.

As only four eyes presented with abnormalities in SAP,we did not perform an analysis comparing ocular hyper-

tension eyes with normal and abnormal SAP results owingto the small sample size in the abnormal subgroup. Threeof the four eyes with SAP abnormalities also had abnor-malities on SWAP fields. The exclusion of these foursubjects from the analyses comparing central cornealthickness measurements in ocular hypertension eyes withnormal and abnormal SWAP results did not change thestatistical significance of the results.

DISCUSSION

THERE IS NOW CONSIDERABLE EVIDENCE TO INDICATE

that SWAP is a more sensitive test than SAP for detectionof early functional deficits produced by glaucoma. Short-wavelength automated perimetry deficits occur severalyears before the development of abnormality on standardperimetry and are predictive of the onset and location offuture SAP defects.11–22

In our study, we have found that, among the ocularhypertension patients, 24% already presented with func-tional deficits detected by SWAP. In contrast, only 6% ofthese patients had defects detected by SAP. This is inaccordance with previous studies in which the prevalenceof SWAP deficits in ocular hypertension eyes has beengenerally reported to be between 8% and 30%, dependingon the criteria used and the specific risk characteris-tics.22–24,26 Our inclusion criteria were different from thatof some previous reports as the SAP visual fields were notpart of the definition of the ocular hypertension group.There has been a change in the understanding of the bestway to evaluate new techniques for measuring visualfunction in glaucoma. Earlier, each new method wasevaluated relative to the gold standard, SAP. However, thefinding that some of the new procedures, like SWAP, aremore sensitive than SAP for identifying glaucomatousoptic neuropathy has made us shift the emphasis to selecta structural measure of glaucomatous optic disk damage asour standard when evaluating results of psychophysicaltests of vision. In this study, we used stereophotographs ofthe optic disk and IOP because they are the currentclinical standard.

Our results suggest that a subgroup of patients classifiedas ocular hypertensive based on a normal optic diskappearance have in fact early glaucomatous functionaldamage. However, the majority of ocular hypertensionpatients had no evidence of structural or functional dam-age and a large number of them probably will neverdevelop such deficits, even during lengthy observation. Arecent study evaluating a group of 500 eyes of 250 patientswith ocular hypertension longitudinally showed that theincidence of newly developed visual field defects usingSWAP was 1.23% annually.23

Clinical experience with ocular hypertension patientshas shown that it is usually difficult to accurately predictwhich patients will eventually develop field loss or glau-

TABLE 2. Visual Field Indices*

Variable

Normal

(n � 52 eyes)

Abnormal

(n � 16 eyes) P Value†

SWAP MD (dB) 0.06 � 2.94 �2.53 � 3.94 .006

SWAP PSD (dB) 2.64 � 0.60 3.40 � 0.86 �.001

CCT (�m) 572 � 35 545 � 25 .006

*Mean deviation (MD) and pattern standard deviation (PSD)

and central corneal thickness measurements (CCT) of 68 ocular

hypertensive patients with normal and abnormal short-wave-

length automated perimetry (SWAP) results.†Student’s t test.

AMERICAN JOURNAL OF OPHTHALMOLOGY134 FEBRUARY 2003

comatous optic neuropathy. Several reports have shownthat some patients with ocular hypertension have centralcorneal thicknesses greater than normal subjects or pa-tients with glaucoma.4–8 Because there is an overestima-tion of the IOP measured by applanation tonometry insubjects with thicker corneas, some of the subjects cur-rently classified as having ocular hypertension may havetrue IOP in the normal range. Consequently, these subjectsmay be at a much lower risk for development of structuraland functional damage from glaucoma. In our study, weshowed that patients currently classified as having ocularhypertension, but with visual field loss detected by SWAP,had significantly lower central corneal thickness measure-ments than the ocular hypertension patients with normalvisual field results. The mean central corneal thickness ofour ocular hypertension patients with abnormal SWAPresults was not statistically significantly different from themean central corneal thickness of the normal controlgroup, and it was similar to previous reports in theliterature of normal and glaucomatous populations.4–8

The mean central corneal thickness of the ocularhypertension patients with normal visual field results was572 �m, similar to a previous report of the OHTS.9 In thatstudy, the mean central corneal thickness of 912 whitesubjects with high IOP and normal fields was reported tobe 579 �m. These values are higher than mean centralcorneal thickness measurements in normal white subjects.

In a recent report by La Rosa and associates,27 the meancentral corneal thickness of 51 normal Caucasian subjectswas 556 �m, virtually identical to the value found in ournormal controls. The mean central corneal thickness ofwhite subjects has been reported to be higher than inAfrican Americans,9,27 however, this should not haveinfluenced the results of our study, as our ocular hyperten-sion population was composed almost exclusively of whitesubjects (94%) and all patients identified with visual fieldloss were white.

The Goldmann applanation tonometer is theoreticallyonly accurate for central corneal thickness at 520 �m, andseveral algorithms to correct IOP on the basis of centralcorneal thickness have been reported.28–30 In a recentmeta-analysis study of the corneal thickness literature,Doughty and Zaman10 derived a correction factor of 2.5mm Hg for each 50-�m change in the central cornealthickness. In another study, Ehlers and associates28 cannu-lated 29 eyes at the time of cataract surgery and deter-mined that a 70-�m change in central corneal thicknesscorresponded to approximately 5 mm Hg of IOP differ-ence. Using this method of correction, Copt and associ-ates4 reported that 56% of patients who were diagnosedinitially with ocular hypertension actually had IOPs of 21mm Hg or lower. Using Ehlers’ correction factor, 23 of the68 ocular hypertension patients (34%) in our study had a“corrected” IOP less than or equal to 21 mm Hg. It is

FIGURE 1. Distribution of central corneal thickness among the ocular hypertensive (OHT) patients with normal and abnormalshort-wavelength automated perimetry (SWAP) results and in the control subjects.

CORNEAL THICKNESS IN OCULAR HYPERTENSIONVOL. 135, NO. 2 135

noteworthy that only one of these patients had visual fielddefects as detected by SWAP or SAP. However, thecorrection of the IOP with such linear correction factorsprobably represents an oversimplification of a complexrelationship between corneal thickness and true IOP.More interestingly, a simple analysis of the distribution ofcorneal thickness values in the ocular hypertension pa-tients reveals that none of the patients with abnormalSWAP results had a central corneal thickness greater than600 �m.

Although it is evident that corneal thickness plays animportant role in the measurement of intraocular pressure,it should be considered in the context of the absolute levelof the measured IOP and also of other risk factors for thedevelopment of glaucomatous damage. A recent reportfrom the OHTS evaluating baseline demographic andclinical factors that predict the development of primaryopen-angle glaucoma among ocular hypertension subjectsfound that a thinner central corneal measurement was apowerful predictor of the onset of glaucoma.31 Amongocular hypertension participants who developed primaryopen-angle glaucoma, the mean central corneal thicknesswas 553 � 38.8 �m compared with 574 � 37.8 �m amongthose who did not develop primary open-angle glaucoma.The mean central corneal thickness measurement of ocularhypertension subjects with normal SWAP results in ourstudy was very similar to the mean central corneal thick-ness measurement from OHTS patients who did notdevelop glaucoma. Also, the mean central corneal thick-ness of our ocular hypertension subjects with SWAPabnormalities was comparable to the mean central cornealthickness of subjects in the OHTS study who developedglaucoma.

In conclusion, the results of our study add to theevidence that central corneal thickness should be takeninto account when assessing risk for the development ofglaucomatous damage among ocular hypertension subjects.

REFERENCES

1. Whitacre MM, Stein R. Sources of error with the Gold-mann-type tonometers. Surv Ophthalmol 1993;38:1–30.

2. Kruse Hansen F. A clinical study of the normal humancentral corneal thickness. Acta Ophthalmol 1971;49:82–89.

3. Kruse Hansen F, Ehlers N. Elevated tonometer readingscaused by a thicker cornea. Acta Ophthalmol 1971;49:775–778.

4. Copt R, Thomas R, Mermoud A. Corneal thickness in ocularhypertension, primary open-angle glaucoma, and normaltension glaucoma. Arch Ophthalmol 1999;117:14–16.

5. Argus WA. Ocular hypertension and central corneal thick-ness. Ophthalmology 1995;102:1810–1812.

6. Johnson M, Kass MA, Moses R, Grodzki WJ. Increasedcorneal thickness simulating elevated intraocular pressure.Arch Ophthalmol 1978;96:664–665.

7. Herman DC, Hodge DO, Bourne WM. Increased corneal

thickness in patients with ocular hypertension. Arch Oph-thalmol 2001;119:334–336.

8. Herndon LW, Choudri SA, Cox T, et al. Central cornealthickness in normal, glaucomatous and ocular hypertensiveeyes. Arch Ophthalmol 1997;115:1137–1141.

9. Brandt JD, Beiser JA, Kass MA, Gordon MO. Centralcorneal thickness in the Ocular Hypertension TreatmentStudy (OHTS). Ophthalmology 2001;108:1779–1788.

10. Doughty MJ, Zaman ML. Human corneal thickness and itsimpact on intraocular pressure measurement: a review andmeta-analysis. Surv Ophthalmol 2000;44:367–408.

11. Sample PA, Weinreb RN. Color perimetry for assessment ofprimary open-angle glaucoma. Invest Ophthalmol Vis Sci1990;31:1869–1875.

12. Sample PA, Taylor JDN, Martinez G, Lusky M, WeinrebRN. Short-wavelength color visual fields in glaucoma sus-pects at risk. Am J Ophthalmol 1993;115:225–233.

13. Johnson CA, Adams AJ, Casson EJ, Brandt JD. Blue-on-yellow perimetry can predict the development of glaucoma-tous visual field loss. Arch Ophthalmol 1993;111:645–650.

14. Johnson CA, Adams AJ, Casson EJ, Brandt JD. Progressionof early glaucomatous visual field loss as detected by blue-on-yellow and standard white-on-white automated perime-try. Arch Ophthalmology 1993;111:651–656.

15. Sample PA. Short-wavelength automated perimetry: its rolein the clinic and for understanding ganglion cell function.Prog Retin Eye Res 2000;19:369–383.

16. Girkin CA, Emdadi A, Sample PA, et al. Short-wavelengthautomated perimetry and standard perimetry in the detectionof progressive optic disc cupping. Arch Ophthalmol 2000;118:1231–1236.

17. De Jong LAMS, Snepvangers CEJ, van den Berg CT,Langerhorst CT. Blue-yellow perimetry in the detection ofearly glaucomatous damage. Doc Ophthalmol 1990;75:303–314.

18. Felius J, De Jong L, van den Berg T, et al. Functionalcharacteristics of blue-on-yellow perimetric thresholds inglaucoma. Invest Ophthalmol Vis Sci 1995;36:1665–1674.

19. Landers J, Goldberg I, Graham S. A comparison of shortwavelength automated perimetry with frequency doublingperimetry for the early detection of visual field loss in ocularhypertension. Clin Exper Ophthalmol 2000;28:248–252.

20. Bayer AU, Erb C. Short-wavelength automated perimetry,frequency doubling technology perimetry, and pattern elec-troretinography for prediction of progressive glaucomatousstandard visual field defects. Ophthalmology 2002;109:1009–1017.

21. Sample PA, Bosworth CF, Weinreb RN. Short-wavelengthautomated perimetry and motion automated perimetry inpatients with glaucoma. Arch Ophthalmol 1997;115:1129–1133.

22. Polo V, Larosa JM, Pinilla I, et al. Predictive value ofshort-wavelength automated perimetry. A 3-year follow-upstudy. Ophthalmology 2002;109:761–765.

23. Demirel S, Johnson CA. Incidence and prevalence of shortwavelength automated perimetry deficits in ocular hyperten-sive patients. Am J Ophthalmol 2001;131:709–715.

24. Sample PA, Bosworth CF, Blumenthal EZ, Girkin C, Wein-reb RN. Visual-function specific perimetry for indirect com-parison of different ganglion cell populations in glaucoma.Invest Ophthalmol Vis Sci 2000;41:1783–1790.

25. Johnson CA, Sample PA, Cioffi GA, Liebmann JR, WeinrebRN. Structure and function evaluation (SAFE): I. Criteriafor glaucomatous visual field loss using standard automatedperimetry (SAP) and short wavelength automated perimetry(SWAP). Am J Ophthalmol 2002;134:177–185.

AMERICAN JOURNAL OF OPHTHALMOLOGY136 FEBRUARY 2003

26. Johnson CA, Brandt JD, Khong AM, Adams AJ. Short-wavelength automated perimetry in low-, medium-, andhigh-risk ocular hypertensive eyes. Arch Ophthalmol 1995;113:70–76.

27. La Rosa FA, Gross RL, Orengo-Nania S. Central cornealthickness of caucasians and African Americans in glaucoma-tous and nonglaucomatous populations. Arch Ophthalmol2001;119:23–27.

28. Ehlers N, Bramsen T, Sperling S. Applanation tonometryand central corneal thickness. Acta Ophthalmol (Copenh)1975;53:34–43.

29. Whitacre MM, Stein RA, Hassanein K. The effect of cornealthickness on applanation tonometry. Am J Ophthalmol1993;115:592–596.

30. Wolfs RCW, Klaver CCW, Vingerling JR, et al. Distributionof central corneal thickness and its association with intraoc-ular pressure: the Rotterdam Study. Am J Ophthalmol1997;123:767–772.

31. Gordon MO, Beiser JA, Brandt JD, et al. The OcularHypertension Treatment Study: baseline factors that predictthe onset of primary open-angle glaucoma. Arch Ophthal-mol 2002;120:714–720.

CORNEAL THICKNESS IN OCULAR HYPERTENSIONVOL. 135, NO. 2 137