Morphologic Predictive Factors for Development of Optic Disc Hemorrhages in Glaucoma

Abstract

purpose. To evaluate which optic disc parameters are predictive factors for the development of disc hemorrhages in chronic open-angle glaucoma.

methods. The prospective comparative clinical observational study included 432 eyes of 281 white patients with chronic open-angle glaucoma. Mean follow-up time was 38.8 months (median, 31.5). Eyes in the whole study group were divided into those with an optic disc hemorrhage during the follow-up period (hemorrhagic group; n = 38; 8.8%), those without disc hemorrhages and with neuroretinal rim loss as sign of progression of glaucoma (rim loss group; n = 42; 9.7%), and those with neither disc hemorrhages nor neuroretinal rim loss (stable group; n = 352; 81.5%). Color stereo optic disc photographs were obtained repeatedly in all patients and subjected to qualitative and morphometric evaluation.

results. At baseline, neuroretinal rim area was significantly (P < 0.03) smaller and the beta zone of parapapillary atrophy (temporal lower sector) was significantly (P < 0.03) larger in the hemorrhagic group than in the stable group. Both study groups did not vary significantly (P > 0.05) in optic disc size and shape, optic cup depth, alpha zone of parapapillary atrophy, and retinal vessel diameter. In multivariate analysis, the neuroretinal rim area was the only significant predictor of hemorrhages. The hemorrhagic group and the rim loss group did not differ significantly (P > 0.05) in any optic disc parameter measured.

conclusions. In chronic open-angle glaucoma, morphologic predictive factors for the development of disc hemorrhages are small size of neuroretinal rim and, possibly, a large parapapillary beta zone. Development of disc hemorrhages is independent of optic disc size and shape, size of alpha zone of parapapillary atrophy, retinal vessel diameter, and optic cup depth. Optic nerve heads in eyes with eventual development of disc hemorrhages and in eyes with eventual progressive rim loss without observed disc hemorrhages do not differ markedly in appearance.

Flame-shaped or splinterlike hemorrhages at the optic disc border are a characteristic hallmark of glaucomatous optic nerve damage. ^{1 2 3 4 5 6 7} Rarely found in normal eyes, ^{8 9 10 11} disc hemorrhages are detected in approximately 4% to 7% of eyes with glaucoma. ⁹ In early glaucoma, they are usually located in the inferotemporal or superotemporal disc regions. They are associated with localized retinal nerve fiber layer defects, neuroretinal rim notches, and circumscribed perimetric loss. ^{3 6 12 13} The diagnostic importance of disc hemorrhages is based on the findings that they are only rarely found in normal eyes; that they usually indicate the presence of glaucomatous optic nerve damage, even if the visual field is unremarkable ^{14 15 16 17 18} ; and that they suggest progression of glaucoma. ^{6 13 14 15 17 19 20 21 22} In view of the diagnostic and pathogenic importance disc hemorrhages have for the diagnosis of glaucoma, the purpose of the present study was to evaluate which morphologic features of the optic nerve head predispose the development of optic disc hemorrhages in the follow-up examination of patients with chronic open-angle glaucoma. 

Methods

The prospective clinical observational study included 432 eyes (215 right eyes, 217 left eyes) of 281 white patients (134 women, 147 men) with chronic open-angle glaucoma. In 130 patients, only one eye per patient was included in the study, because either the contralateral eye not included in the study did not show glaucomatous abnormalities of the optic nerve head or visual field or the quality of the optic disc photograph was not sufficient for inclusion in the study. All patients had an open anterior chamber angle and were prospectively and consecutively evaluated. After enrollment in the study, the patients were reexamined, usually within 6 or 12 months. Some patients had a shorter follow-up period. Baseline examination and follow-up examination consisted of a routine ophthalmologic examination including refractometry, slit lamp biomicroscopy of the anterior and posterior segment of the eye, gonioscopy, and tonometry. In addition, the intraocular pressure was measured in day-and-night profiles, computerized white-on-white perimetry was performed (Octopus perimeter program G1; Interzeag, Schlieren, Switzerland), color stereo optic disc photographs were taken, and electrophysiologic examination were performed, such as measurements of visual evoked potentials of the blue-sensitive pathway. The methods applied in the study adhered to the tenets of the Declaration of Helsinki for the use of human subjects in biomedical research. Informed consent was obtained from each subject before enrollment. Institutional review board and ethics committee approvals were not required for this study. The patients were part of an ongoing prospective study on the progression of glaucoma (Erlangen Glaucoma Register). 

According to the presence or absence of glaucomatous visual field defects, the eyes of the study population were divided into eyes with glaucomatous abnormalities of the optic nerve head and normal white-on-white visual fields (n = 227; 52.5%), ^{23 24} and eyes with chronic open-angle glaucoma with glaucomatous visual field defects (n = 205; 47.5%). The definition of glaucomatous changes of the optic nerve head included an unusually small neuroretinal rim area in relation to the optic disc size, according to the physiologic correlation between disc size and neuroretinal rim area ²⁵ ; abnormal shape of the neuroretinal rim, which was not markedly broader in the inferior and superior disc region compared with the temporal disc region; cup-to-disc diameter ratios being higher vertically than horizontally; and localized or diffuse retinal nerve fiber layer defects. ¹ A glaucomatous visual field defect was defined as (1) a visual field (Octopus G1) with at least three adjacent test points having a deviation of equal to or greater than 5 dB and with one test point with a deviation of more than 10 dB lower, (2) at least two adjacent test points with a deviation equal to or greater than 10 dB, (3) at least three adjacent test points with a deviation equal to or greater than 5 dB abutting the nasal horizontal meridian, or (4) a mean visual field defect of more than 2 dB. The rates of false-positive and false-negative answers each had to be equal to or less than 15%. The 205 eyes with chronic open-angle glaucoma consisted of eyes with primary open-angle glaucoma (n = 88 eyes), eyes with secondary open-angle glaucoma due to conditions such as pseudoexfoliation or primary melanin pigment dispersion syndrome (n = 37 eyes), and eyes with normal-pressure glaucoma (n = 80 eyes). In the eyes affected by primary open-angle glaucoma, no obvious reason for the elevated intraocular pressure could be detected. Criteria for the diagnosis of normal-pressure glaucoma were maximum intraocular pressure readings equal to or less than 21 mm Hg in at least two 24-hour pressure profiles obtained by slit lamp applanation tonometry, with measurements at 5 PM, 9 PM, 12 AM, 7 AM, and 12 PM. Ophthalmoscopy, medical history, and neuroradiologic, neurologic, and medical examinations did not reveal any reason for optic nerve damage (such as intrasellar or suprasellar tumors, retinal vessel occlusions, optic disc drusen, or nonarteritic anterior ischemic optic neuropathy) other than glaucoma. 

For all eyes, 15° color stereo optic disc transparencies had been taken with a telecentric fundus camera (30° fundus camera, equipped with a 15° converter; Carl Zeiss, Oberkochen, Germany). The disc slides were projected in a scale of 1 to 15. The outlines of the optic cup, optic disc, peripapillary scleral ring, and alpha and beta zones of parapapillary atrophy were plotted on paper and morphometrically analyzed. To obtain values in absolute size units (i.e., millimeter or square millimeter), the ocular and photographic magnification was corrected by the Littmann method. ²⁶ The optic cup was defined on the basis of contour and not of pallor. The border of the optic disc was identical with the inner side of the peripapillary scleral ring. Parapapillary atrophy was differentiated into a peripheral alpha zone, with irregular pigmentation, and a central beta zone, with visible sclera and visible large choroidal vessels. The depth of the optic cup was scaled into degrees ranging from 0 for no cupping to 5 for very deep cupping. The reproducibility of the assessment of the optic cup’s depth had been determined in a previous study and was 2.0% and 8.9% for two examiners. ²⁷ The diameters of the retinal arterioles were measured at the optic disc border in the inferotemporal, superotemporal, superonasal, and inferonasal regions. The method has already been described in detail, ²⁸ including the comparison between planimetry of optic disc photographs and postmortem direct measurements of the optic disc in unfixed specimens, ²⁹ as well as the comparison between the planimetric method and confocal laser scanning tomography of the optic nerve head. ³⁰  

Once or twice a year, all patients included in the study underwent white-on-white perimetry and stereo photography of the optic nerve head. Mean follow-up time was 38.8 months (median, 31.5; range, 1.9–98.1). Of the 432 eyes included in the study, 38 (8.8%) showed an optic disc hemorrhage on the fundus transparencies, 42 (9.7%) showed progressive loss of neuroretinal rim, and 352 (81.5%) remained stable (Table 1) . Progression of glaucoma was defined as a loss of neuroretinal rim that could be accompanied by an increase in parapapillary atrophy, a decrease in the visibility of the retinal nerve fiber layer, or optic disc hemorrhages. Perimetric results were not used as criteria for classification. To evaluate the progression of glaucoma, the first slide and the most recent slide of the optic nerve of the same eye were taken, mixed, and simultaneously projected in an unknown order. Two examiners (JBJ, WMB) jointly assessed qualitatively whether the two photographs differed. If a difference was detected, and if the slide with the more marked optic nerve damage was the more recent photograph, the eye was considered to have progressive glaucomatous optic nerve damage. Of all pairs of optic disc photographs that were judged to differ from each other, approximately 5% to 10% were considered to show a smaller neuroretinal rim in the earlier photograph. These eyes were then classified as not showing progression of glaucomatous optic nerve damage. 

Appearance of changes in the optic nerve head was statistically evaluated using methods for censored data. Descriptive analysis included Kaplan-Meier curves. Statistical differences between groups were assessed by using the log rank test. Continuous covariates were analyzed by bivariate and multivariate Cox regression analysis with forward variable selection. For descriptive purposes only, frequencies and mean values are additionally given, referring to the groups with or without change, ignoring time to change. Some of the potential morphologic predictors (e.g., area of parapapillary atrophy) showed extremely skewed distributions. Accordingly, all these variables were categorized into groups of small, medium, and high values, according to the 33% and 67% percentiles in the sample. Statistical dependency of data from both eyes of the same patient was adjusted for by correction of standard errors and χ² statistics according to the number of patients. With 34 compared with 39 individuals showing hemorrhages and progression, respectively, hazard rates of approximately 2.6 and 2.5 were detectable (assuming a level of significance of 0.05, two-sided, and a power of 0.8). ³¹ Statistical analyses were performed by using a commercially available statistical software package (SPSS for Windows, ver. 9.0; SPSS Science, Chicago, IL). The level of significance was 0.05 (two-sided) in all statistical testings. 

Results

In the entire sample, 38 hemorrhages were observed (Table 1 , Fig. 1 ). The patient’s age was a significant prognostic factor for disc hemorrhages (P = 0.01, Table 1 , Fig. 2 ). At baseline, the neuroretinal rim area was significantly (P < 0.03) smaller, and the beta zone of parapapillary atrophy (temporal inferior sector, P < 0.03) was significantly larger in the group with disc hemorrhages compared with the stable group (Tables 2 3 ; Figs. 3 4 ). The group of eyes with disc hemorrhages and the stable group did not vary significantly (P > 0.05) in size and shape of the optic disc, depth of the optic cup, size of alpha zone of parapapillary atrophy, and diameter of the retinal arteries and veins at the optic disc border (Tables 2 3 4) . Similar results were observed for the comparison between the stable group and the rim loss group. 

All statistical analyses were adjusted for age, which was a significant predictor of hemorrhages (odds ratio, 1.48 for 10-year increase in age, P = 0.01, Cox regression model). Presence or absence of medication was not a significant predictor of hemorrhages (P > 0.1), nor was increased intraocular pressure. Instead, normal-pressure glaucoma showed a 3.6-fold increased risk for hemorrhages. Neuroretinal rim area was a predictive factor independent of intraocular pressure (rim area: P = 0.014, after inclusion of intraocular pressure). 

Comparison of the group of eyes with disc hemorrhages and the rim loss group revealed no statistically significant differences in any optic disc parameter measured (Tables 2 3 4) . Ten (26%) of 38 eyes with hemorrhages showed a progressive loss of neuroretinal rim compared with 32 (8%) of 352 eyes without hemorrhages—a statistically significant difference (P < 0.01). 

Multivariate Cox regression analysis revealed that, besides age, only the area of the neuroretinal rim was a significant predictor of hemorrhages. 

Discussion

The results suggest that the development of optic disc hemorrhages in patients with chronic open-angle glaucoma is associated with a small neuroretinal rim. Eyes in which an optic disc hemorrhage was detected during the follow-up period had a significantly smaller neuroretinal rim at baseline than eyes in which disc hemorrhages were not observed and in which the neuroretinal rim did not change. It suggests that disc hemorrhages are more common in patients with advanced glaucoma than in patients with early glaucoma. The findings are in agreement with a previous study in which the frequency of disc hemorrhages increased from an early stage of glaucoma to a medium advanced stage and decreased again toward the prefinal stage of glaucoma, until, in eyes with absolute glaucoma, disc hemorrhages were no longer detected. ⁹ The reason for the dependency of disc hemorrhages on the stage of glaucoma remains unclear. Assuming that disc hemorrhages indicate progression of glaucoma, ^{6 13 14 15 17 19 20 21} one may infer that progression in eyes with advanced glaucoma occurs more frequently than in eyes with early glaucoma. Clinically, it suggests that glaucoma should be treated as early as possible. 

The present study also showed that the development of disc hemorrhages was significantly associated with a large beta zone of parapapillary atrophy at baseline. Eyes with disc hemorrhages had a significantly larger beta zone than eyes without eventual disc hemorrhages. This finding was not confirmed in a multivariate Cox regression analysis, in which the prognostic value of parapapillary atrophy was not significant. However, the probability of parapapillary atrophy was 0.07 in this analysis, which means that significance was nearly reached. Thus, taking into account the relatively small number of 38 hemorrhages, no definite conclusion can be drawn about whether parapapillary atrophy might be an independent prognostic factor for the development of hemorrhages. In a previous cross-sectional investigation the occurrence of disc hemorrhages was significantly associated with a large beta zone of parapapillary atrophy. ⁹ The question arises of why eyes with a large beta zone of parapapillary atrophy compared with eyes with a small beta zone may more frequently show optic disc hemorrhages and, presumably, progressive glaucomatous damage of the optic nerve. The vascular bed supplying the region where parapapillary atrophy takes place is different from the vascular bed of the retinal layer in which disc hemorrhages are found. 

The result of the present study agrees with a recent investigation by Law et al., ³² who evaluated qualitatively the structural characteristics and the associated features that antedated the occurrence of a disc hemorrhage in patients with glaucoma. Examining 4018 pairs of stereoscopic optic disc images obtained during 15 years, they found that the optic disc characteristics that most antedated the disc hemorrhage were parapapillary atrophy, superior–inferior asymmetry in the neuroretinal rim, and thin sloping of the rim. 

The development of disc hemorrhages during the follow-up period of the patients included in the present study was independent of size of the optic disc. It concurs with a previous cross-sectional study in which the prevalence of disc hemorrhages was statistically independent of the optic disc size, ⁹ and it agrees with other studies suggesting that, despite the marked interindividual variability in optic disc area, ³³ the disc size is probably not correlated with the susceptibility for glaucomatous optic nerve damage. ¹ It suggests clinically that neither a large nor a small optic disc is reason to intensify antiglaucoma treatment in an effort to decrease the risk of eventual disc hemorrhages that may indicate progression of glaucoma. It can also be inferred that, the variance in optic disc size among blacks, Asians, Hispanics, and whites may not be a reason to assume that the various ethnic groups differ in the risk for the development of disc hemorrhages and, presumably, the risk for progression of glaucoma. 

In the present study, progression of glaucoma was statistically independent of the shape of the optic disc as measured by the ratio of the horizontal-to-vertical disc diameter and by the ratio of the minimum to maximum disc diameter (Table 2) . It agrees with previous investigations on primary open-angle glaucoma in which the shape of the optic disc was not significantly correlated with neuroretinal rim area and mean perimetric defect, neither interindividually nor in an intraindividual bilateral comparison. ³⁴  

In the present study, eyes with eventual disc hemorrhages and eyes with eventual loss of neuroretinal rim did not vary in any optic disc parameter measured at baseline of the study (Table 2) . In agreement with preceding investigations, ^{6 13 14 15 17 19 20 21} it suggests that disc hemorrhages may be taken as indicators of progression of glaucomatous optic neuropathy. The advantage of disc hemorrhages compared with loss of neuroretinal rim as sign of progression of glaucoma is that a previous examination with or without photographic documentation of the appearance of the optic nerve head is not necessary for the detection of disc hemorrhages. 

The present study has limitations. Because its purpose was to evaluate morphologic features of the optic disc in their importance in eventual development of optic disc hemorrhages, other factors, such as the level of intraocular pressure and the type of chronic open-angle glaucoma, were not primarily taken into account. Also, because of the composition of the study population, race was not evaluated in the present study, although it may be a predictive factor in glaucoma. ³⁵ The normal time for optic disc hemorrhages to disappear is at approximately 2 months. ¹² We cannot, therefore, know whether the disc hemorrhages detected on the fundus transparencies of the patients included in the study were the only disc hemorrhages that developed in the patients during the follow-up period. The number of patients in whom disc hemorrhages developed may, therefore, be higher than the data found in the present investigation show. Because it was the purpose of the study, however, to evaluate the predictive factors for the development of the optic disc hemorrhages and not to measure the incidence of disc hemorrhages during the follow-up period, this limitation of the study may not have markedly influenced the results of the study. If it is assumed that overlooking of disc hemorrhages was independent of predictive factors, statistical effects were underestimated and statistical power was decreased by overlooking the disc hemorrhages. Overlooking the disc hemorrhages, however, caused no erroneously significant results to appear. Because the frequency of reexamination varied between the study participants, there is a questions of how the fact that some eyes were seen twice a year and others once a year influenced the results. It can be speculated that some eyes were seen more often because the clinician was suspicious of progression or because of a viable hemorrhage. Because the purpose of the study was to examine the predictive factors leading to a disc hemorrhage, a shorter time interval between further reexamination for eyes with an already detected disc hemorrhage than for eyes with a stable optic nerve head would not have influenced the results of the study. The eye with the disc hemorrhage would already have been classified as a disc bleeder at the first detection of the disc hemorrhage, independent of the time interval to the next reexamination. Another limitation of the study may be the relatively short minimum follow-up of 1.9 months. Again, the main goal of the study was to evaluate predictive factors for the development of disc hemorrhages. Disc hemorrhages are ophthalmoscopically detectable for approximately 2 months so that a follow-up period of 1.9 months may be, more or less, the shortest interval to detect disc hemorrhages without having an unnecessarily short follow-up period. 

In conclusion, features of the optic disc that predict eventual development of optic disc hemorrhages in patients with chronic open-angle glaucoma are a small neuroretinal rim and, possibly, a large beta zone of parapapillary atrophy. Development of optic disc hemorrhages is independent of size and shape of the optic disc, size of the alpha zone of parapapillary atrophy, diameter of the retinal vessels, and depth of the optic cup. These findings may have implications in the decision of whether a patient needs a more intensive or aggressive antiglaucomatous treatment than others in an attempt to reduce the risk of eventual development of optic disc hemorrhages and further progression of glaucomatous optic neuropathy. 

 Supported by Grant SFB 539 from the German Research Foundation.

 Submitted for publication January 30, 2002; revised May 9, 2002; accepted May 20, 2002.

 Commercial relationships policy: N.

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Eyes with Disc Hemorrhage, No Progression, and Progression of Glaucoma

	Disc Hemorrhage	P *	Stable	Rim Loss	P , †
Eyes (n)	38		352	42
Subjects (n)	34		222	39
Females/males	12/22	0.057	105/117	17/22	0.55
Age (y)	54.7 ± 8.5	0.01	49.6 ± 12.5	56.2 ± 9.6	<0.001
Median	54		52	57
Range	37–75		15–75	31–75
Refractive error (D)	−0.84 ± 2.99	0.15	−1.08 ± 2.52	−1.24 ± 2.81	0.79
Median	0.0		−0.5	−0.13
Range	−7.75–4.75		−10.50–4.75	−12.0–2.0

 Data are expressed as the mean ± SD.

*  Disc hemorrhage group versus stable group.

†  Neuroretinal rim loss group versus stable group.

Figure 1.

 

In 390 of 256 patients, there were 38 optic disc hemorrhages. Upper and lower confidence limits are shown.

Figure 2.

 

Frequency of optic disc hemorrhage according to patient’s age. The sample was divided into three groups of equal size, according to the observed distribution of age: group 1: 15 to 46 years, group 2: 47 to 56 years, group 3: 57 to 75 years.

Table 2.

 

Optic Disc and Neuroretinal Rim Areas of Eyes with Disc Hemorrhage, with No Progression, and with Loss of Neuroretinal Rim

	Disc Hemorrhage	P *	Stable	Rim Loss	P , †
Eyes (n)	38		352	42
Optic disc area (mm2)	2.82 ± 0.64	0.66	2.76 ± 0.70	2.71 ± 0.68	0.53
Median	2.70		2.64	2.54
Range	1.75–4.16		1.51–6.43	1.52–4.98
Optic disc shape
Horizontal/vertical diameter	0.94 ± 0.063	0.84	0.94 ± 0.074	0.95 ± 0.069	0.16
Median	0.95		0.94	0.96
Range	0.78–1.06		0.75–0.97	0.76–1.15
Minimum/maximum diameter	0.90 ± 0.049	0.53	0.89 ± 0.047	0.89 ± 0.047	0.46
Median	0.91		0.90	0.90
Range	0.77–0.96		0.75–0.97	0.74–0.96
Neuroretinal rim area (mm2)
Total	0.90 ± 0.33	0.028	1.08 ± 0.41	0.90 ± 0.35	0.027
Median	0.90		1.10	0.80
Temporal horizontal	0.12 ± 0.068	0.055	0.16 ± 0.077	0.12 ± 0.066	0.003
Median	0.12		0.15	0.11
Temporal inferior	0.19 ± 0.11	0.057	0.28 ± 0.14	0.19 ± 0.11	0.004
Median	0.18		0.29	0.18
Temporal superior	0.21 ± 0.11	0.16	0.27 ± 0.12	0.23 ± 0.12	0.009
Median	0.20		0.26	0.20
Nasal	0.38 ± 0.11	0.086	0.42 ± 0.15	0.38 ± 0.14	0.054
Median	0.35		0.40	0.37
Optic cup depth (0–5)	3.16 ± 0.68	0.42	2.77 ± 0.96	2.80 ± 0.87	0.44
Median	3		3	3

 The data are expressed as the mean ± SD. The group with disc hemorrhages and the group with rim loss group did not differ significantly in any optic disc parameter measured.

*  Disc hemorrhage group versus stable group.

†  Neuroretinal rim loss group versus stable group.

Table 3.

 

Parapapillary Atrophy of Eyes with Disc Hemorrhage, with No Progression, and with Loss of Neuroretinal Rim

	Disc Hemorrhage	P*	Stable	Rim Loss	P†
Alpha zone (mm2)
Total	0.91 ± 0.46	0.89	0.88 ± 0.72	0.82 ± 0.76	0.17
Median	0.87		0.77	0.79
Temporal horizontal	0.37 ± 0.16	0.40	0.33 ± 0.26	0.29 ± 0.29	0.058
Median	0.37		0.31	0.23
Temporal inferior	0.22 ± 0.14	0.96	0.20 ± 0.18	0.20 ± 0.20	0.15
Median	0.21		0.18	0.17
Temporal superior	0.19 ± 0.14	0.86	0.20 ± 0.20	0.19 ± 0.17	0.18
Median	0.19		0.17	0.17
Nasal	0.13 ± 0.23	0.44	0.15 ± 0.24	0.13 ± 0.23	0.42
Median	0		0	0
Beta zone (mm2)
Total	0.42 ± 0.36	0.072	0.41 ± 0.60	0.84 ± 1.76	0.039
Median	0.40		0.18	0.39
Temporal horizontal	0.18 ± 0.18	0.21	0.16 ± 0.22	0.21 ± 0.28	0.23
Median	0.14		0.07	0.13
Temporal inferior	0.14 ± 0.14	0.027	0.11 ± 0.18	0.25 ± 0.44	0.020
Median	0.12		0	0.12
Temporal superior	0.07 ± 0.096	0.21	0.08 ± 0.15	0.15 ± 0.38	0.27
Median	0.003		0	0.002
Nasal	0.03 ± 0.08	0.89	0.06 ± 0.21	0.22 ± 0.75	0.15
Median	0		0	0

 Data and statistical significance are as described in Table 2 .

Figure 3.

 

Frequency of optic disc hemorrhages according to the area of the neuroretinal rim. The sample was divided into three groups of equal size, according to the observed distribution of neuroretinal rim area: group 1: 0–0.89 mm², group 2: 0.90–1.22 mm², group 3: 1.23–2.66 mm².

Figure 4.

 

Frequency of optic disc hemorrhage according to parapapillary atrophy (temporal inferior sector). The sample was divided into three groups of equal size, according to the observed distribution of parapapillary atrophy: group 1: 0.0 mm², group 2: 0.01 to 0.11 mm², group 3: 0.12 to 1.09 mm².

Table 4.

 

Retinal Artery Diameter of Eyes with Progression and Those with No Progression of Glaucoma

	Disc Hemorrhage	P*	Stable	Rim Loss	P†
Arteries (mm)
Temporal inferior	0.10 ± 0.47	0.66	0.097 ± 0.052	0.091 ± 0.027	1.0
Temporal superior	0.095 ± 0.065	0.29	0.090 ± 0.036	0.086 ± 0.066	0.13
Nasal superior	0.082 ± 0.072	0.11	0.080 ± 0.045	0.075 ± 0.037	0.38
Nasal inferior	0.084 ± 0.093	0.98	0.089 ± 0.066	0.089 ± 0.084	0.25
Veins (mm)
Temporal inferior	0.14 ± 0.052	0.42	0.13 ± 0.051	0.14 ± 0.066	0.38
Temporal superior	0.13 ± 0.13	0.93	0.13 ± 0.050	0.18 ± 0.18	0.87
Nasal superior	0.14 ± 0.13	0.37	0.12 ± 0.64	0.14 ± 0.12	0.94
Nasal inferior	0.11 ± 0.091	0.12	0.11 ± 0.047	0.14 ± 0.12	0.29

 Data and statistical significance are as described in Table 2 .