Central corneal thickness and normal tension glaucoma: A cross-sectional study

Transcription

1 Optometry (2006) 77, Central corneal thickness and normal tension glaucoma: A cross-sectional study Michael Sullivan-Mee, O.D., a Kathy D. Halverson, O.D., a Mollie C. Saxon, O.D., a Glenn B. Saxon, O.D., a and Clifford Qualls, Ph.D. a,b a Albuquerque VA Medical Center, Albuquerque, New Mexico and b University of New Mexico, Albuquerque, New Mexico. KEY WORDS: Normal-tension glaucoma; Intraocular pressure; Tonometry; Cornea; Central corneal thickness BACKGROUND: Recently published evidence has identified thinner central corneal thickness (CCT) as a strong predictive factor for the conversion from ocular hypertension (OHT) to primary open-angle glaucoma (POAG). The association between CCT and development of normal-tension glaucoma (NTG), however, is less clear. Accordingly, we designed this cross-sectional study to further explore the relationship between CCT and NTG. PATIENTS AND METHODS: All patients with a clinical diagnosis of NTG and NTG suspect (NTGS) who were seen from September 2002 through May 2003 at the Albuquerque VA Medical Center eye clinic were identified retrospectively. After eligible subjects were categorized into no, mild, moderate, and advanced visual field loss groups, analysis of variance (ANOVA) and regression analyses were used to determine group differences for several IOP variables, several systemic variables, and CCT. Additional analyses were completed after eligible subjects were recategorized into thin, intermediate, and thick CCT groups. RESULTS: Eighty-four eyes in 84 NTGS subjects and 56 eyes in 56 NTG subjects were studied. Mean CCT was significantly thicker in the no field loss group (NTGS) when compared with all 3 groups with glaucomatous visual field loss (NTG). In multivariate regression analysis, the association between CCT and the presence of NTG-related visual field loss was robust and independent. Conversely, no relationship was found between CCT and severity of NTG-related visual field loss. CONCLUSIONS: In eyes characterized by statistically normal intraocular pressure (IOP) measurements as measured by Goldmann applanation tonometry, we found a significant relationship between CCT and the presence, but not severity, of glaucomatous visual field loss. A prospective study is required to further explore and confirm these relationships. Optometry 2006;77: Corresponding author: Michael Sullivan-Mee, O.D., Director, Optometric Education, Albuquerque VA Medical Center, Eye Clinic 112 A, 1501 San Pedro SE, Albuquerque, New Mexico Michael.sullivan-mee@med.va.gov While the Ocular Hypertension Treatment Study (OHTS) generated convincing evidence that thinner central corneal thickness (CCT) is a strong predictive factor for the conversion from ocular hypertension (OHT) to primary-open angle glaucoma (POAG), 1 the association between CCT and development of normal-tension glaucoma (NTG) is less clear. Although several investigators have reported that patients with NTG show thinner CCT compared with patients with OHT, 2-8 POAG, 2-5,9 chronic closed-angle glaucoma, 4,8 and normal eyes, 2-4,8 these reports have focused primarily on comparisons of mean CCT among these different groups. To date, no investigations have been published comparing CCT in patients with NTG versus patients who are NTG suspects. Accordingly, we designed this cross-sectional study to explore relationships between CCT and the presence and severity of NTG-related /06/$ -see front matter 2006 American Optometric Association. All rights reserved. doi: /j.optm

2 Sullivan-Mee et al Clinical Research 135 glaucomatous visual field loss through evaluation of patients with NTG and patients who are NTG suspects. Methods After Institutional Review Board approval from the Albuquerque VA Medical Center and the University of New Mexico School of Medicine, patient logs were used to identify every patient seen from September 2002 through May 2003 at the Albuquerque VA Medical Center eye clinic with a clinical diagnosis of either normal-tension glaucoma (NTG) or normal-tension glaucoma suspect (NTGS). Once potential subjects were identified, retrospective chart review was utilized to determine definitive eligibility for study inclusion. For study purposes, diagnosis of NTG required medical record documentation of optic nerve morphology consistent with glaucomatous optic neuropathy 10 (e.g., focal erosion or thinning, vertical elongation of the cup, nerve fiber layer defect) as well as a corresponding glaucomatous visual field defect detectable on the most recent, reliable 24-2 Humphrey threshold visual field test (e.g., arcuate scotoma, nasal step defect, paracentral defect). Subjects showing diffuse visual field loss only and subjects showing field defects that did not correspond to the optic neuropathy were excluded. Diagnosis of NTGS required normal visual field results in association with suspicious but nondefinitive glaucomatous optic nerve morphology. All optic nerves were evaluated initially by optometric students and residents, followed by confirmatory assessments by an attending provider/investigator (M.S.M., K.H., M.S., G.S.) All optic nerve assessments were done using a 60 diopter (D) Volk lens (clinic protocol required measurement of disc size and rim tissue thickness in each quadrant, along with evaluation of peripapillary tissue, nerve fiber layer, and evidence of disc hemorrhage). Suspicious optic nerve morphology was considered present when rim configuration deviated from the normal rim thickness pattern as described by Jonas, 10 or when vertical cup-to-disc ratio was larger than expected when accounting for vertical disc size (as measured with a 60 D Volk indirect lens during Haag-Streit or Marco slit lamp funduscopy). Larger than expected cup-to-disc ratios were determined based on data published by Litwak 11 and Lim et al. 12 Specifically, when the coordinates for cup-todisc ratio versus disc size were plotted on or above the trend line shown in Figure 1, that cup-to-disc ratio was considered larger than expected. Any medical record documentation of definitive signs of glaucomatous neuropathy such as focal erosion, obvious nerve fiber layer defects, or optic disc hemorrhages resulted in exclusion of that subject from the study (when no corresponding, repeatable field defect was present). In addition to optic nerve and visual field requirements, both NTG and NTGS diagnoses also required that no pretreatment intraocular pressure (IOP) measurements, via Figure 1 Threshold for suspicious cup size in relation to vertical disc size. *All coordinates that plot above trend line are considered a sign of abnormally large cups in relation to disc size. Figure derived from data modified from Litwak 11 and Lim et al. 12 predilation Goldmann applanation tonometry, were statistically elevated (22 mmhg or higher). Subjects were also excluded if any treated IOP readings were elevated. Each subject was also required to have best-corrected visual acuity of at least 20/50; dilated, stereoscopic optic nerve evaluations using a 60-D lens; open angles by 4-mirror gonioscopy (subjects were excluded from this study if visibility of the scleral spur was obscured in any quadrant); reliable ultrasound pachymetry measurements that were not influenced by corneal pathology; and reliable, achromatic Humphrey 24-2 SITA-standard threshold visual field testing results. Exclusion criteria included secondary forms of open angle glaucoma, corneal pathology that could influence pachymetry or intraocular pressure measurements, and history of refractive, corneal, or glaucoma surgery. All subjects completed automated achromatic Humphrey 24-2 SITA-Standard threshold visual field testing within the prior 12 months. Although all prior fields were reviewed, the most recent, reliable test was used for analysis. Visual fields with greater than 20% fixation losses, 33% falsepositive responses, or 33% false-negative responses were excluded. Subjects with visual field defects due to nonglaucomatous entities such as retinal pathology, nonglaucomatous optic nerve pathology, and visual tract compromise were also excluded. Although we did not require repeatable visual field loss, we did require that subjects have field defects that were typical for glaucoma and that all field defects corresponded to optic nerve morphology. By using these criteria, the optic nerve status was indirectly included in confirming the diagnostic staging of glaucoma and in confirming the field results. To determine presence and severity of glaucomatous visual field loss, 2 masked investigators (K.H., M.S.) independently graded visual fields using Advanced Glaucoma Intervention Study (AGIS) numerical scoring criteria. 13 When a discrepancy in the calculation of an AGIS score arose between investigators, the field was rescored a final time by the principal investigator (M.S.M.), and a final score was assigned. The AGIS visual field scoring procedure systematically generates a numerical score of 0 to 20

3 136 Optometry, Vol 77, No 3, March 2006 where 0 represents no visual field loss, 1 through 5 represent mild loss, 6 through 11 represent moderate loss, 12 through 17 represent severe loss, and 18 through 20 represent end-stage visual field loss. To facilitate statistical analysis and because severe and end-stage glaucomatous visual field loss often are managed similarly, we combined the severe and end-stage field loss groups into 1 category. This resulted in 4 visual field loss groups (none, mild, moderate, and advanced). Measurement of central corneal thickness was completed on all subjects between September 2002 and May 2003 with a calibrated DGH 500 ultrasonic pachymeter (Exton, Pennsylvania). Five readings, with a divergence no greater than 15 m, were obtained for each eye, and the mean of these 5 readings was used as the central corneal thickness for that eye. Mean values for pretreatment and post-treatment IOP were calculated, utilizing all available undilated Goldmann applanation tonometry measurements. The highest pretreatment and post-treatment IOP values were also recorded and both the mean and high (pretreatment and post-treatment) IOP values were used within data analysis. Mean IOP reduction and percentage IOP reduction were also calculated for each subject by subtracting mean post-treatment IOP from highest pretreatment IOP. Thirty-six of 56 (64%) NTG subjects and 12 of 84 (14%) NTGS subjects were treated and had treated IOP data available for analysis. All subjects also had at least 3 pretreatment applanation tonometry measurements on separate days at different times of the day. The SAS/STAT statistical program (Version 8.2, 2001; SAS Institute Inc., Cary, North Carolina) was utilized for all statistical analyses, and P values less than 0.05 were considered statistically significant in all analyses. One eye of each subject was used for data analysis to avoid intrasubject correlation bias. 14 When both eyes were categorized into the same AGIS visual field category, the right eye was utilized for data analysis by default. When a subject had 1 eye that fell into a higher AGIS visual field loss category than the fellow eye, the eye with the greater visual field loss was utilized for data analysis. For each of the 4 visual field loss groups (per AGIS classification), ranges, means, and standard deviations were calculated for the following variables: highest and mean pretreatment and treated IOP values, mean IOP reduction, percentage IOP reduction, CCT, AGIS score, and age. Comparison of all calculated mean values between the 4 visual field groups was accomplished via 1-way analysis of variance using Fisher s least significant difference method for post-hoc, pairwise comparisons. Analysis of variance results were verified by the nonparametric Kruskal-Wallis test. Univariate and multivariate logistic regression analysis utilizing stepwise and all subsets modeling was then used to determine associations with the presence of glaucomatous visual field loss. Continuous variables in this analysis included highest and mean pretreatment IOP, CCT, and age. Categorical variables included self-reported family history of glaucoma, prior diagnoses of diabetes, sleep apnea, hypothyroidism, or hypertension as well as concurrent use of systemic calcium channel blockers, -blockers, ACE inhibitors, -agonists, and diuretics. For this analysis, NTGS subjects (AGIS 0) were compared with NTG subjects (AGIS 0). Adjusted Wald odds ratios with 95% confidence intervals were also calculated, using a step size of 2 standard deviations. Univariate and multivariate linear regression analyses were then used to determine associations with the severity of glaucomatous visual field loss. In addition to the variables listed previously, these analyses also included the following variables: mean and high post-treatment IOP values, mean IOP reduction, and percentage IOP reduction. Only subjects with measurable visual field loss (AGIS 0) were included in these analyses, and AGIS score was used as the dependent variable. Similar to the method used to analyze CCT in the OHTS, the database was then divided into 3 equal segments according to CCT, resulting in a thin, intermediate, and thick cornea group. Within these groups, ranges, means, and standard deviations were calculated for AGIS visual field score, age, central corneal thickness, highest pretreatment IOP, mean pretreatment IOP, highest post-treatment IOP, mean post-treatment IOP, mean IOP reduction, and percentage IOP reduction. Mean and percentage IOP reduction variables were derived by subtracting mean post-treatment IOP from highest pretreatment IOP for each subject. These values were compared within and across each group via 1-way analysis of variance using pairwise comparisons within Fisher s least significant difference method. In addition, frequencies of mild, moderate, and advanced visual field loss were calculated and compared for differences within the thin, intermediate, and thick CCT groups using the Jonckheere-Terpstra test. This test evaluates trends in visual field severity within and across CCT groups. Because ethnic differences in CCT have been reported, we obtained ethnic data for our subjects via self-reported information in the medical record. Categories that were utilized included Caucasian, African American, Hispanic, Native American and Unknown. Chi-square tests were utilized to determine if any racial differences were present when comparing the 4 visual field loss groups and when comparing the 3 CCT groups. Results One hundred forty eyes in 140 subjects were included in data analysis. Of these, 84 (60%) had no visual field loss, and 56 (40%) had visual field loss. Of the 56 subjects with visual field loss, 25 (16.9%) had mild visual field loss, 16 (10.8%) had moderate visual field loss, and 15 (10.1%) had advanced visual field loss per AGIS visual field criteria (see Table 1).

4 Sullivan-Mee et al Clinical Research 137 Table 1 Category of visual field loss (per AGIS classification) None (0) Mild (1) Moderate (2) Advanced (3) Number treated 84 (12) 25 (8) 16 (15) 15 (15) AGIS score Age (yr) CCT ( m) * Highest pretreatment IOP, mmhg (n) (84) (25) (16) (15) Mean pretreatment IOP, mmhg (n) (84) (25) (16) (15) Highest post-treatment IOP, mmhg (n) (12) (8) (15) (15) Mean post-treatment IOP, mmhg (n) (12) (8) (15) (15) Overall IOP reduction, mmhg (n) (12) (8) (15) (15) Overall IOP percentage reduction (n) t(12) (8) (15) (15) Note: All figures expressed as mean values SD. IOP reduction variables calculated by subtracting mean post-treatment IOP from highest pretreatment IOP. * Significantly higher at 0.05 level compared to category 1, 2, and 3. Significantly higher at 0.05 level compared with category 2. Significantly lower at 0.05 level compared with category 0 and category 2. Mean CCT was found to be 557 m (range, 485 to 639 m) for the no visual field loss group, and 522 m (range, 433 to 627 m) for the glaucomatous visual field loss group, a result that compares favorably with previously reported mean CCT in NTG. 2-5 Within the glaucomatous field loss group, mean CCT was 529 m (range, 448 to 579 m) for the mild visual field loss group, 508 m (range, 433 to 562 m) for the moderate visual field loss group, and 521 m (range, 464 to 627 m) for the advanced visual field loss group. Using Fisher s least significance differences method, mean CCT in the no visual field loss group was found to be significantly thicker (P 0.001) than each of the groups with field loss (mild, moderate, and advanced). When comparing the mild, moderate, and advanced visual field loss groups, mean CCT in the moderate group was found to be significantly thinner than mean CCT in the mild group, but no other significant differences were found. Using a mean of 3.9 pretreatment IOP measurements per subject (n 140), and a mean of 6.7 post-treatment IOP measurements per treated subject (12 of 84, 8 of 25, 15 of 16, 15 of 15 in the no, mild, moderate, and advanced field loss groups were treated, respectively), no clinically significant IOP relationships were evident (see Table 1). In univariate logistic regression comparing NTG subjects with NTGS subjects, CCT (P 0.001) and use of systemic -agonist medication (P 0.01) were found to be significantly associated with glaucomatous field loss. In multivariate analysis, CCT (P 0.001) and hypothyroidism (P 0.04) were found to be the only independent variables associated with field loss. Calculated odds ratios (OR) suggested that CCT is inversely associated with field loss (OR, 0.06) and the relatively tight 95% confidence intervals (CI, 0.02 to 0.18) indicate that this relationship is strong statistically. Concurrent diagnosis of hypothyroidism was directly associated with presence of field loss (OR, 2.31), but the wide confidence intervals that approach unity (1.03 to 5.18) suggest that this relationship is less robust (see Table 2). In linear regression analysis, evaluating factors associated with severity of NTG-related field loss, none of the studied variables were found to be statistically associated with AGIS score. After the full sample was divided into 3 equal groups according to CCT, the range for the thin group was 433 to 525 m (n 46), the range for the intermediate group was 526 to 556 m (n 47), and the range for the thick group was 559 to 639 m (n 47). Frequency of measurable Table 2 Univariate and Multivariate Adjusted Odds Ratios*(with 95% CIs) for normal tension glaucoma risk factors (n 140) Variables Univariate Multivariate CCT 0.08 ( ) 0.06 ( ) Age 0.62 ( ) Mean pretreatment 0.51 ( ) IOP Highest pretreatment 0.75 ( ) IOP Family history 0.69 ( ) Diabetes 0.88 ( ) Sleep apnea 1.38 ( ) Hypothyroidism 1.39 ( ) 2.31 ( ) Hypertension 1.35 ( ) Calcium channel 1.27 ( ) blockers Ace Inhibitors 1.22 ( ) Diuretics 0.84 ( ) Alpha agonists 2.54 ( ) -blockers 0.54 ( ) Significant values (P 0.05) are in bold letters. Note: Adjusted odds ratios below 1.0 signify an inverse relationship between visual field loss and the risk factor, and ratios above 1.0 signify a direct relationship between visual field loss and the risk factor. Thus, a Wald odds ratio of 0.10 implies a 10x (1/0.1) increase in odds with an inverse relationship between the factors, and a Wald odds ratio of 2.0 implies a 2 (2/1) increase in odds with a direct relationship between the factors. * Based on step-size of 2 standard deviations.

5 138 Optometry, Vol 77, No 3, March 2006 Table 3 Comparison of AGIS score, age, and IOP variables per CCT category Thin CCT Category Intermediate Number CCT range/mean ( m) (502) (540) (584) AGIS score (all subjects) 5.3 ( 5.8)* 2.7 ( 5.3) 0.8 ( 2.6) AGIS score (NTG subjects only) 7.8 ( 5.4) 6.6 ( 6.6) 6.2 ( 4.9) Number of NTG subjects (%) 31 (67)* 19 (40)* 6 (13)* Age (yr) 68 ( 11.1) 67 ( 12.6) 67 ( 11.4) Highest pretreatment IOP (mmhg) 17.7 ( 2.2) 17.1 ( 2.7) 17.6 ( 2.2) Mean pretreatment IOP (mmhg) 16.0 ( 2.1) 15.6 ( 2.5) 16.6 ( 2.0) Treated subjects (%) 26 (57) 13 (28) 11 (23) Mean post-treatment IOP (mmhg) 13.1 ( 1.9) 13.5 ( 2.0) 13.8 ( 2.5) Mean post-treatment IOP reduction (mmhg) 3.4 ( 1.4) 2.5 ( 1.8) 3.3 ( 2.4) Mean post-treatment IOP reduction (%) 0.20 ( 0.08) 0.15 ( 0.10) 0.19 ( 0.14) * Significantly different at 0.05 level compared with the other 2 categories. Significantly different at 0.05 level compared with category 2. Thick visual field loss was 31 of 46 (67.4%) in the thin CCT group, 19 of 47 (40.4%) in the intermediate CCT group, and 6 of 47 (12.8%) in the thick CCT group. These frequencies were significantly different ( , df 2, P 0.001), and post-hoc pairwise comparisons showed that these frequencies all significantly differed from one another (all P 0.01). Mean AGIS scores for the thin, intermediate, and thick CCT groups were calculated as 5.3, 2.7, and 0.8, respectively. Although no significant difference was evident between the medium and thick CCT groups, mean AGIS score within the thin CCT group was significantly higher (P 0.001) than both the medium and thick CCT groups (see Table 3). When mean AGIS score per CCT category was recalculated, using only subjects with measurable glaucomatous visual field loss (AGIS 0), mean AGIS scores were 7.8, 6.6, and 6.2 for the thin, intermediate and thick cornea groups, respectively. These means were not significantly different from one another, suggesting that the significantly higher mean AGIS score in the all-subject thin cornea group is explained by increased frequency of visual field loss, not increased severity of field loss (see Table 3). Within the CCT groups, no significant differences were found for any of the posttreatment IOP variables. This result contrasts with findings from a post-hoc analysis from the OHTS, 15 which reported that patients with thicker CCT show reduced IOP responses to antiglaucoma medication compared with patients with thinner CCT. The reason for these conflicting results may be due to differences in study populations. When comparing frequency of each stage of measurable visual field loss within each CCT group using the Jonckheere-Terpstra test, no overall intracategory differences emerged (P 0.17). Thus, within the thin, intermediate, and thick CCT categories, no apparent trend for worse or better visual field loss was detected. This result implies that subjects with thinner CCT did not disproportionately suffer greater visual field loss in this sample (see Figure 2). Discussion Although it has become rather well accepted that thinner CCT is a strong risk factor for development of POAG, no studies have been published that identify CCT as a risk factor for development of NTG. Although this study s cross-sectional, retrospective design precludes specific declarations about CCT as a risk factor, the strong relationship between CCT and NTG-related visual field loss found in this study is noteworthy. Not only was CCT strongly associated with field loss, it was independently associated per multivariate logistic regression analysis results comparing NTG subjects with NTG suspects. Interestingly, these findings correspond with prior reports describing the relationship between CCT and POAG. 1,16 In addition to mean CCT findings and multivariate analysis that support the relationship between CCT and NTG, we also found that 67% of subjects within the thin CCT category versus 40% within the intermediate CCT category versus 13% within the thick CCT category showed glauco- Figure 2 Degree of visual field loss per CCT category.

6 Sullivan-Mee et al Clinical Research 139 matous visual field loss. Taken together, these findings substantiate the importance of CCT in normal tension glaucoma. Although it is unclear why CCT was found to be independently related to NTG development, there is speculation that a thinner CCT may be a marker for structural anomalies in the trabecular meshwork, lamina cribosa, or peripapillary scleral tissue. Sigal et al., 17 in a study of optic nerve biomechanics using a finite elements model, reported that IOP-induced lamina cribosa strain increased with thinner peripapillary scleral thickness and suggested that if a direct correlation between CCT and peripapillary scleral thickness exists, it may help explain why thinner CCT is associated with increased risk for glaucoma development. Considering that mean CCT is consistently reported as thinner in NTG compared with POAG, 2-5,9 and that IOP measurement error does not seem to explain the relationship between NTG and thinner CCT, 18 this hypothesis may hold promise in helping explain why some optic nerves are susceptible to glaucomatous damage at relatively low levels of IOP. Confirmation of such a relationship would likewise lend support to the notion that NTG and POAG are a continuum of the same disease process, with those that develop glaucoma at statistically normal IOP levels showing increased susceptibility to lower levels of IOP, in part, because of structural vulnerability related to a thinner CCT. It should be noted, however, that if such a relationship were proven, it would not discount the importance of pathologic contributions from other causes, including vascular abnormalities. 19 In addition to CCT, a concurrent diagnosis of hypothyroidism was also independently associated with NTG-related visual field loss within our multivariate analysis. Other studies have also identified hypothyroidism as a risk factor for glaucoma, including a recent study by Girkin et al. 24 that examined a large population of male VA Medical Center patients in Alabama. Our results, also using a population of male veteran patients, support the possibility of a relationship between hypothyroidism and open-angle glaucoma. Although thinner CCT was strongly associated with NTG in this study, CCT was not found to be associated with severity of NTG visual field loss. In linear regression analyses of NTG subjects using AGIS score as the dependent variable, CCT was not related to the severity of visual field loss. Furthermore, when comparing mean CCT among subjects with mild, moderate, and advanced visual field loss, we found no relationship between decreasing CCT and increasing field loss severity. Moreover, the Jonckheere- Terpstra test found no field loss trends among the thin, intermediate, and thick CCT groups as shown in Figure 2. Taken together, these findings suggest that CCT was unrelated to severity of NTG-related visual field loss in our sample. Other reports 18,25,26 have described similar results regarding the inability of CCT to predict severity of field loss in POAG. Ethnicity investigation showed the following profile: Hispanic, 37.9%; unknown, 29.3%; white, 22.1%; black, 6.4%; and Native American, 4.3%. Although 29.3% of the sample could not be ethnically identified utilizing medical record review, regional southwest U.S. demographics suggest that the majority of these unknown subjects are either white or Hispanic, with minimal black representation. Thus, it is unlikely that black ethnicity affected mean CCT in this study Regarding the large percentage of Hispanic ethnicity in this study, Shimmyo et al. 16 and Aghaian et al. 17 reported no differences in mean CCT between whites, Hispanics, and Asians. Conversely, Hahn et al. 31 reported that mean CCT in Latinos was lower than in whites but higher than in blacks. Thus, it is possible that because of our high percentage of Hispanic subjects, mean CCT values could have been reduced slightly compared with other studies. Using analysis of variance, however, no statistical relationships were found between race and AGIS score or mean CCT. Additionally, using 2 analysis, we found no statistically significant relationships when comparing racial frequencies between visual field loss groups (none vs. mild vs. moderate vs. advanced) and between corneal thickness groups (thin vs. intermediate vs. thick). Strengths of this study include its multivariable analyses and utilization of masked visual field scoring with scores agreed on by 2 independent examiners. By reducing investigator subjectivity in determining degree of glaucoma, use of the systematic AGIS visual field scoring system may also be considered a strength. Study weaknesses include issues intrinsic to cross-sectional, retrospective research design, including the inability to perform analysis of the relationship between duration of disease and development and severity of visual field loss. Also, we did not require that patients show repeatable visual field loss for study inclusion, although we recognize that repeatable field loss is an important factor in glaucoma management. To address this concern, we did require that subjects show patterned glaucomatous visual field loss that was typical for glaucoma, and we required that all glaucomatous visual field defects matched compromised optic nerve rim tissue. Thus, if a subject s field did not match the nerve, the subject was not enrolled in the study. By using these criteria, the optic nerve was indirectly included in confirming the fields and in staging the degree of glaucoma. Our differentiation of suspicious, nonglaucomatous optic nerves from glaucomatous optic nerves is also a potential weakness owing to the subjective nature of such judgments, the multiple providers making those judgments, and the use of a retrospective strategy to identify subjects. To maximize our discriminatory ability in this area, we used published strategies for guidance, and we rigorously executed these strategies on a daily basis via uncompromising clinic protocols. Even with those strategies, however, subjectivity and the use of the medical record as opposed to photographic review might have affected study results. Finally, because our sample had no female subjects, gender bias may have affected results. The OHTS was the largest study to examine intergender CCT differences 27 and reported that women had significantly thicker mean central

7 140 Optometry, Vol 77, No 3, March 2006 corneal thickness than men, but the difference was less than 5 m (575.0 vs m.) This result agrees with findings from Doughty s meta-analysis, which found that the combination of average CCT values from several studies was slightly higher for women compared with men, although this relationship did not emerge in all studies. 32 These reports suggest that if CCT differs according to gender, the difference is likely very small. Our findings suggest that CCT is significantly thinner in patients with NTG compared with patients that are NTG suspects. These results imply that CCT may be a valuable tool for aiding determination of which patients might benefit from closer monitoring or earlier initiation of IOP-lowering treatment. Our results also suggest that CCT is independently related to development of glaucoma, and that CCT is not related to severity of NTG-related visual field loss. Taken together, these findings are intriguing, but prospective, longitudinal study is required to more fully define these relationships. References 1. Gordon MO, Beiser JA, Brandt JD, et al. The Ocular Hypertension Treatment Study: baseline factors that predict the onset of primary open-angle glaucoma. Arch Ophthalmol 2002;120(6): Copt RP, Thomas R, Mermoud A. Corneal thickness in ocular hypertension, primary open-angle glaucoma, and normal tension glaucoma. Arch Ophthalmol 1999;117(1): Emara BY, Tingey DP, Probst LE, et al. Central corneal thickness in low-tension glaucoma. Can J Ophthalmol 1999;34(6): Shah S, Chatterjee A, Mathai M, et al. Relationship between corneal thickness and measured intraocular pressure in a general ophthalmology clinic. Ophthalmology 1999;106(11): Singh RP, Goldberg I, Graham SL, et al. Central corneal thickness, tonometry, and ocular dimensions in glaucoma and ocular hypertension. J Glaucoma 2001;10(3): Wu LL, Suzuki Y, Ideta R, et al. Central corneal thickness of normal tension glaucoma patients in Japan. Jpn J Ophthlamol 2000;44(6): Ventura AC, Bohnke M, Mojon DS. Central corneal thickness measurements in patients with normal tension glaucoma, primary open angle glaucoma, pseudoexfoliation glaucoma, or ocular hypertension. Br J Ophthalmol 2001;85(7): Bechmann M, Thiel MJ, Roesen B, et al. Central corneal thickness determined with optical coherence tomography in various types of glaucoma. Br J Ophthalmol 2000;84: Morad Y, Sharon E, Hefetz L, et al. Corneal thickness and curvature in normal-tension glaucoma. Am J Ophthalmol 1998;125: Jonas JB, Budde WM, Panda-Jonas S. Ophthalmoscopic evaluation of the optic nerve head. Surv Ophthalmol 1999;43(4): Litwak AB. Evaluation of the optic nerve. In: Glaucoma Handbook. Boston: Butterworth-Heinemann, 2001: Lim CS, O Brien C, Bolton NM. A simple clinical method to measure the optic disc size in glaucoma. J Glaucoma 1996;5: The Advanced Glaucoma Intervention Study Investigators. Advanced Glaucoma Intervention Study: 2. Visual field test scoring and reliability. Ophthalmology 1994;101(8): Ray WA, O Day DM. Statistical analysis of multi-eye data in ophthalmic research. Invest Ophthalmol Vis Sci 1985;26: Brandt JD, Beiser JA, Gordon MO, et al. Ocular Hypertension Treatment Study (OHTS) group. Central corneal thickness and measured IOP response to topical ocular hypotensive medication in the Ocular Hypertension Treatment Study. Am J Ophthalmol 2004;138: Herndon LW, Weizer JS, Stinnett SS. Central corneal thickness as a risk factor for advanced glaucoma damage. Arch Ophthalmol 2004; 122: Sigal IA, Flanagan JG, Tertinegg I, et al. Finite element modeling of optic nerve head biomechanics. Invest Ophthalmol Vis Sci 2004;45: Sullivan-Mee M, Halverson KD, Saxon GB, et al. The relationship between central corneal thickness-adjusted intraocular pressure and glaucomatous visual-field loss. Optometry 2005;76: Doyle A, Bensaid A, Lachkar Y. Central corneal thickness and vascular risk factors in normal tension glaucoma. Acta Ophthalmol Scand 2005;83: Karadimas P, Bouzas EA, Topouzis F, et al. Hypothyroidism and glaucoma. A study of 100 hypothyroid patients. Am J Ophthalmol 2001;131: Munoz-Negrete FJ, Rebolleda G, Almodovar F, et al. Hypothyroidism and primary open-angle glaucoma. Ophthalmologica 2000;214: Smith KD, Arthurs BP, Saheb N. An association between hypothyroidism and primary open-angle glaucoma. Ophthalmology 1993;100: Gillow JT, Shah P, O Neill EC. Primary open-angle glaucoma and hypothyroidism: chance or true association? Eye 1997;11: Girkin CA, McGwin G Jr, NcNeal SF, et al. Hypothyroidism and the development of open-angle glaucoma in a male population. Ophthalmology 2004;111: Jonas JB, Stroux A, Velten I, et al. Central corneal thickness correlated with glaucoma damage and rate of progression. Invest Ophthalmol Vis Sci 2005;46: Chauhan BC, Hutchison DM, LeBlanc RP, et al. Central corneal thickness and progression of the visual field and optic disc in glaucoma. Br J Ophthalmol 2005;89: Brandt JD, Beiser JA, Kass MA, et al. Central Corneal Thickness in the Ocular Hypertension Treatment Study (OHTS). Ophthalmology 2001;108(10): La Rosa FA, Gross RL, Orengo-Nania S. Central corneal thickness of Caucasians and African Americans in glaucomatous and nonglaucomatous populations. Arch Ophthalmol 2001;119: Shimmyo M, Ross AJ, Moy A, et al. Intraocular pressure, Goldmann applanation tension, corneal thickness, and corneal curvature in Caucasians, Asians, Hispanics, and African Americans. Am J Ophthalmol 2003;136(4): Aghaian E, Choe JE, Lin S, et al. Central corneal thickness of Caucasians, Chinese, Hispanics, Filipinos, African Americans, and Japanese in a glaucoma clinic. Ophthalmology 2004;111: Hahn S, Azen S, Ying-Lai M, et al. Central corneal thickness in Latinos. Invest Ophthalmol Vis Sci2003;44: Doughty MJ, Zaman ML. Human corneal thickness and its impact on intraocular pressure measures: a review and meta-analysis approach. Surv Opthalmol 2000;44(5):