Glucose-6-Phosphate Dehydrogenase Deficiency in Retinal Vein Occlusion

Abstract

purpose. Glucose-6-phosphate dehydrogenase (G6PD) deficiency is one of the most common human genetic abnormalities, with a high prevalence in Sardinia, Italy. Evidence indicates that G6PD-deficient patients are protected against ischemic heart and cerebrovascular disease. The purpose of this study was to assess the frequency of G6PD deficiency in Sardinian patients with retinal vein occlusion (RVO) and to ascertain whether the deficiency may offer protection against RVO.

methods. Erythrocyte G6PD levels were measured in 448 consecutive RVO patients: 194 with central RVO (CRVO) and 254 with branch RVO (BRVO). Age- and sex-matched subjects (n = 896) who were undergoing cataract surgery and had no history of RVO served as the control. Multiple logistic regression models were used to investigate the association between G6PD deficiency and RVO, CRVO, or BRVO.

results. G6PD deficiency was found in 21 (4.7%) patients with RVO, 7 (3.6%) with CRVO, 14 (5.5%) with BRVO, and 107 (11.9%) control subjects. Differences between cases and controls were statistically significant (P < 0.005). Multiple conditional logistic regression analysis, including as covariates G6PD deficiency, hypertension, diabetes, and hypercholesterolemia, revealed that G6PD deficiency was significantly associated with decreased risk of development of RVO, CRVO, or BRVO. After adjustment for hypertension, diabetes, and hypercholesterolemia, the association between G6PD deficiency and RVO, CRVO, or BRVO remained statistically significant. Similar results were obtained after adjustment for systolic or diastolic blood pressure, plasma glucose, and cholesterol levels. However, when the patients with CRVO or BRVO were categorized by gender, a significant association was found only in the women.

conclusions. The frequency of G6PD deficiency in patients with RVO was significantly lower than expected. The results suggest that G6PD-deficient patients have a significantly decreased risk of development of RVO in the Sardinian population.

Retinal vein occlusion (RVO) is a major cause of visual loss. ^{1 2} This condition can involve the central trunk or branches of the retinal venous circulation (CRVO and BRVO, respectively). RVO, cardiovascular disease, and cerebrovascular disease have several risk factors in common, including systemic hypertension, diabetes mellitus, atherosclerosis, and smoking. ³ Recent studies have shown that patients with G6PD deficiency are protected against ischemic heart and cerebrovascular disease. ⁴  

G6PD is a cytoplasmic enzyme that is present in all cells, where it plays the key role in regulating carbon flow through the pentose phosphate pathway. Specifically, the enzyme affects the production of the reduced form of the extramitochondrial nicotine adenosine dinucleotide phosphate (NADPH) coenzyme by controlling the conversion from glucose-6-phosphate to 6-phosphogluconate in the pentose phosphate pathway (Fig. 1) . In red blood cells, defense against oxidative damage is heavily dependent on G6PD activity, which is the only source of NADPH. ⁵ The gene encoding G6PD is located in the distal long arm of the X chromosome (band Xq28). More than 300 alleles with missense point mutation in the G6PD gene sequence have been identified. ⁶ The G6PD-Mediterranean allele, associated with levels of enzyme activity undetectable with routine methods (World Health Organization [WHO] class II), is common on the island of Sardinia, Italy, where the reported prevalence of G6PD deficiency ranges from 10% to 15%. ^{4 7 8 9} This condition is a public health issue in Sardinia, because of the seasonal occurrence of favism, a hemolytic anemia induced by ingestion of the broad bean (Vicia faba) in subjects expressing the deficient phenotype. 

The present study was designed to assess the frequency of G6PD deficiency in Sardinian patients with RVO and to ascertain whether G6PD deficiency may have a protective effect against this retinal vascular disorder. 

Methods

The present study was of a case–control design, with 448 patients with RVO (CRVO or BRVO) and 896 control subjects recruited between January 1996 and December 2005. All patients and control subjects were of Sardinian ancestry. Sample size was computed before the survey, with a 95% confidence level (two-tailed test) and 80% statistical power to detect a relative risk of 0.5, assuming a G6PD deficiency prevalence rate of 12%, as reported previously. ^{4 7 8} The case–control ratio was 1:2. 

All patients with RVO underwent a full ophthalmic evaluation, including best corrected visual acuity (BCVA), slit lamp examination, applanation tonometry, fundus biomicroscopy, and fluorescein angiography. Medical conditions, including systemic hypertension, diabetes mellitus, hypercholesterolemia, and cardio- and cerebrovascular status (presence of angina, myocardial infarction, transient ischemic attacks, and stroke) were also recorded. Exclusion criteria included age <18 years and non-Sardinian ancestry. 

Two age- and sex-matched control subjects per case were randomly selected from the cataract register. Exclusion criteria included age <18 years, non-Sardinian ancestry, and previous history of RVO. All control subjects underwent standard ophthalmic evaluation, including BCVA, slit lamp examination, applanation tonometry, and fundus examination. Medical conditions, including systemic hypertension, diabetes mellitus, hypercholesterolemia, and cardio- and cerebrovascular status (presence of angina, myocardial infarction, transient ischemic attacks, or stroke) were also recorded. The control subjects were recruited concurrently with the patients with RVO during the recruitment period. 

Subjects were considered to have hypertension if they were receiving treatment with antihypertension drugs or if their blood pressure was >140 mm Hg systolic or >90 mm Hg diastolic (as defined by the WHO/International Society of Hypertension). Subjects were classified as diabetic if they were in treatment for insulin- or non–insulin-dependent diabetes mellitus or if they had a fasting plasma glucose level of ≥126 mg/dL and/or a plasma glucose level of ≥200 mg/dL 2 hours after a 75-g oral glucose load in a glucose tolerance test (as defined by the WHO). Hypercholesterolemia was defined by a fasting plasma cholesterol level of >220 mg/dL or the intake of lipid-lowering drugs. 

Institutional ethics review board approval was obtained, and the study was conducted in full accord with the tenets of the Declaration of Helsinki. Each participant received detailed information and provided informed consent before inclusion. 

Red blood cell G6PD activity was determined with a quantitative assay (G6PD/6PGD; Biomedic snc, Sassari, Italy) based on the method described by Beutler et al. ¹⁰ Quantitative testing for G6PD deficiency is routinely performed in all patients admitted to our hospital. Whole blood was collected in test tubes containing EDTA and analyzed within 4 hours. Enzyme activity was measured as instructed by the manufacturer. In addition, in a subgroup of G6PD-deficient subjects, the G6PD genotype was determined by PCR-RFLP (restriction fragment length polymorphism) methodology. ¹¹ In particular, the C→T mutation at nucleotide (nt) 563 in exon 6 (G6PD-Mediterranean), which is the most common Italian mutation included in WHO class II, was tested by MboII digestion. ^{6 9 11}  

Categorical values were compared by χ² test. The differences between cases and controls for quantitative variables were analyzed by Student’s t-test. Multivariate logistic regression models, including as covariates age, gender, and known risk factors for RVO, such as systemic hypertension, diabetes mellitus, and hypercholesterolemia, were used to determine the significance of the association between G6PD deficiency and RVO, CRVO, or BRVO. ¹² Multivariate analyses with adjustment for systolic or diastolic blood pressure, plasma glucose, and cholesterol levels were also performed. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated. P ≤ 0.05 was considered to be statistically significant. Statistical analysis was performed with commercial software (Stata ver. 9.0; StataCorp, College Station, TX). 

In addition, plasma homocysteine levels were measured with a capillary electrophoresis–laser-induced fluorescence (CE-LIF) method in 75 patients with RVO (33 with CRVO and 42 with BRVO) and in 72 apparently healthy age- and sex-matched control subjects. These results have been published elsewhere. ¹³  

Two percent of the RVO cases and 3% of the control subjects who were eligible for the study declined to participate. The major reason was “not interested.” 

Results

The study group consisted of 448 patients with RVO (208 men, 240 women; mean age, 65 ± 11.8 years). One hundred ninety-four patients (106 men, 88 women; mean age, 65.2 ± 13.1 years) had CRVO and 254 (102 men, 152 women; mean age, 64.9 ± 10.6 years) had BRVO. Of the 254 eyes with BRVO, the occlusion involved the superior temporal branch in 166 (65.3%) cases, the inferior temporal branch in 84 (33.1%), and one of the nasal branches in 4 (1.6%). Details about the affected eye, visual acuity, intraocular pressure, and glaucoma history of the study group subjects are shown in Table 1 . 

The patients’ and the controls’ systemic characteristics are reported in Table 2 . Both groups had similar levels of plasma glucose and similar rates of diabetes and TIA/stroke (P > 0.05). In contrast, patients with RVO had significantly higher levels of systolic and diastolic blood pressure and plasma cholesterol, a significantly higher frequency of systemic hypertension and hypercholesterolemia, and a significantly lower frequency of G6PD deficiency than the control subjects. Significant associations were also found when the patients were categorized by type of vein occlusion. Those with RVO also had a significantly higher frequency of angina and myocardial infarction (P = 0.04); however, this significance was lost when the patients with CRVO or BRVO were analyzed separately. 

G6PD deficiency was found in 21 (4.7%) of 448 patients with RVO, 7 (3.6%) of 194 patients with CRVO, 14 (5.5%) of 254 patients with BRVO, and 107 (11.9%) of 896 control subjects. Enzyme deficiency was total in men (hemizygotes) and partial in women (homozygotes). None of the patients with G6PD deficiency had shown clinical manifestations of favism or drug-induced hemolysis in the 2 years before their enrollment in this study. 

Similar to other studies analyzing the relationship between certain risk–protective factors of vascular disease and RVO, the control group was selected from patients undergoing cataract surgery. ^{13 14} It is highly unlikely that this strategy introduced a selection bias, because the frequency (11.9%) of G6PD deficiency found in the control group was the same as in the general Sardinian population. ^{4 7 8} This result, along with the finding that patients with cataract with either normal or deficient G6PD activity had similar mean ages (64.9 ± 11.7 and 66.4 ± 11.5 years, respectively), rules out the hypothesis that G6PD deficiency may result in increased susceptibility to cataract, in full agreement with Meloni et al., ¹⁵ who showed that G6PD-deficient patients do not have a higher risk of cataract. 

Genotyping was performed in 28 patients (11 men, 17 women) with G6PD deficiency; 9 (3 men, 6 women) had RVO, and 19 (8 men, 11 women) were control subjects. The G6PD Mediterranean mutation was found in 8 (72.7%) men and 11 (64.7%) women. Overall, the frequency of the G6PD Mediterranean mutation was 67.9%, a result consistent with that reported in another study. ⁹  

Multiple logistic regression results are reported in Tables 3 4 5to 6 . In multiple conditional analyses, including G6PD deficiency, systemic hypertension, diabetes, and hypercholesterolemia, the ORs for RVO, CRVO, and BRVO revealed that G6PD deficiency was significantly associated with a decreased risk of development of the vascular disorder (Table 3) . In this model, hypercholesterolemia was significantly associated with increased risk for RVO, CRVO, and BRVO. Hypertension was significantly associated with increased risk for RVO and BRVO, whereas the OR for CRVO fell just short of statistical significance (P = 0.06). 

Even after adjustment for possible confounding factors (i.e., hypertension, diabetes, and hypercholesterolemia), the association between G6PD deficiency and RVO or CRVO remained statistically significant (Table 4) . As for BRVO, a significantly decreased risk of development of the disease was found when the patients were analyzed as a single group; however, when they were categorized by gender, a significant association between G6PD deficiency and BRVO was found in the women but not in the men. 

Tables 5 and 6show the results of analyses adjusted for systolic or diastolic blood pressure, plasma glucose, and cholesterol levels. Adjustment for these factors as continuous variables produced results similar to those obtained by adjustment for hypertension, diabetes, and hypercholesterolemia. When patients with CRVO and BRVO were categorized by gender, a lack of association between G6PD deficiency and the retinal vascular disease was found in both the male BRVO and CRVO groups. 

Discussion

RVO is a significant cause of ocular morbidity. After diabetic retinopathy, RVO (including both CRVO and BRVO) is probably the most common vascular disorder affecting the eye. Systemic hypertension, diabetes mellitus, and arteriosclerosis are important risk factors for RVO. ^{1 2 13 14} There is still no general agreement on the role of elevated plasma homocysteine in RVO. Some peer-reviewed studies have shown plasma homocysteine to be significantly higher in patients than in controls ³ ; other case–control studies have failed to demonstrate an association. ^{16 17} We have recently shown that patients with CRVO have increased plasma cysteine, a cardiovascular risk factor. ¹⁶ As with cardiovascular disease, the risk of RVO has been found to decrease with increasing levels of physical activity and alcohol consumption. ^{13 14} Furthermore, in postmenopausal women, the use of exogenous estrogens is associated with a decreased risk of CRVO. ¹³  

Most patients in whom CRVO develops are older than 50 years. Histologic studies suggest that most, if not all, forms of CRVO have a common mechanism: thrombosis of the central retinal vein at and posterior to the level of the lamina cribrosa. It has been suggested that, in many instances, an atherosclerotic central retinal artery impinges on the central retinal vein, causing turbulence, endothelial damage, and thrombus formation. ¹ BRVO occurs most frequently between the ages of 60 and 70 years. Histologic studies suggest that a common adventitia binds the artery and vein together and that atherosclerotic thickening of the arterial wall compresses the vein, resulting in turbulence of flow, endothelial cell damage, and thrombotic occlusion. ^{2 18} Actually, CRVO and BRVO are two separate clinical entities sharing some common risk factors and a similar pathogenetic mechanism, but with different natural history and prognosis. 

In this case-control study, patients with RVO had a significantly higher frequency of systemic hypertension, hypercholesterolemia, and angina and myocardial infarction. Our results are consistent with earlier studies, suggesting a cardiovascular risk profile in patients with RVO. ^{13 14} Unlike other reports, we found that both patients and control subjects had similar rates of diabetes. ^{1 2 13 14} This can be explained by several reasons, including the high incidence of diabetes in Sardinia and the increased susceptibility to cataract in persons with diabetes. ^{19 20}  

G6PD deficiency is one of the most common human genetic abnormalities, affecting an estimated 400 million people worldwide. ⁶ The disorder is found mainly in the tropical and subtropical regions of the world, with the highest rates, usually 5% to 30%, in Africa, Asia, the Middle East, the Mediterranean, and Papua, New Guinea. ^{21 22 23} Worldwide, the frequency figures range from 62% in Kurdish Jews to 0.1% in Japan and Northern Europe. ²¹ In the United States, black males are commonly affected, with a prevalence of approximately 10%. ²¹ Sardinia is one of the areas with the highest prevalence, with reported rates ranging from 10% to 15%. ^{4 7 8} The high prevalence of G6PD deficiency in Sardinia has been related to malaria, which was endemic in this part of Italy until the 1950s. ²³ Several studies have suggested that the geographic distribution of G6PD deficiency, which correlates highly with the distribution of current or past malaria endemicity, is the result of a balanced polymorphism that confers resistance to infection with falciparum malaria. ^{23 24}  

A large body of experimental evidence linking G6PD activity, cholesterol synthesis, and cell growth has accumulated in recent years. ^{7 25} However, despite their genetic condition, G6PD-deficient individuals grow normally. ²⁶ It is likely that alternative sources of NADPH, such as the extramitochondrial isocitrate dehydrogenase enzyme and the malic enzyme, provide enough NADPH to support the endogenous cholesterol synthesis required for normal cell replication. ²⁵  

We are unaware of any previously reported study exploring the hypothesis of a decreased risk of RVO, CRVO, or BRVO in subjects expressing G6PD deficiency. In our study, we found that the frequency of G6PD deficiency in patients with RVO, CRVO, or BRVO was lower than expected. Multiple conditional logistic regression analysis revealed that G6PD deficiency was significantly associated with decreased risk of RVO. Significant associations were also found when the patients were categorized by type of vein occlusion (CRVO or BRVO). After adjustment for systemic hypertension, diabetes, and hypercholesterolemia, the association between G6PD deficiency and RVO, CRVO, or BRVO remained statistically significant. Similar results were obtained after adjustment for systolic or diastolic blood pressure and plasma glucose and cholesterol levels, thus demonstrating that the association with G6PD is indeed robust. Of note, when the patients were categorized by gender, a lack of association between G6PD deficiency and CRVO or BRVO was found in the men. 

Overall, our data suggest that G6PD deficiency may have a protective effect against both types of RVO, with an apparently greater protection against CRVO. Statistical analysis showed that both genders have a significantly decreased risk of RVO; however, this decreased risk appears to be more significant in women. The reason that a partial deficiency, found in heterozygous females, may offer greater protection against RVO than a total deficiency, found in hemizygous males, is unclear and needs further research. 

A clear limitation of this study is that it was restricted to a very limited group of patients (i.e., those of Sardinian ancestry). As a result, our findings may not be applicable to RVO patients of non-Sardinian ancestry. 

Our results are consistent with those reported by Cocco et al., ⁴ who found a decrease in mortality rates from cardiovascular and cerebrovascular disease in G6PD-deficient subjects and suggested that patients with G6PD deficiency are less susceptible to these vascular disorders. 

Batetta et al. ⁷ have recently shown that the Mediterranean variant of G6PD deficiency is characterized by peculiar alterations in plasma and intracellular cholesterol metabolism, such as reduced synthesis and esterification. Consequently, in G6PD-deficient subjects, the reduced ability to esterify and accumulate cholesterol in the arteries may account for a lower risk of atherosclerotic disease. These findings may explain our results, calling into question a mechanistic connection between G6PD deficiency and RVO. Theoretically, the slower progression of the atherosclerotic process involving the trunk of the central retinal artery may reduce the degree of compression of the central retinal vein at and posterior to the level of the lamina cribrosa and therefore may reduce the risk of CRVO. Similarly, this mechanism may also be effective to a lesser extent in BRVO, where the venous narrowing at the arteriovenous crossing may induce downstream hemodynamic changes predisposing to endothelial damage and thrombus generation. ¹⁸  

Balance between nitric oxide synthase (NOS) activity, ^{27 28} which is NAPDH dependent, and levels of glutathione (GSH), a physiological scavenger of NO, ²⁹ may also be an important factor in preventing the occurrence of cardiovascular disease, cerebrovascular disease, and RVO. It is unknown whether the two factors are out of balance in G6PD-deficient individuals. Both NO and its S-nitrosocysteine adducts are powerful vasodilators and act as scavengers of superoxide radicals, ^{30 31} thus abrogating their toxicity ^{32 33} ; preventing the oxidation of low-density lipoproteins (LDL) ³⁴ ; and inhibiting platelet aggregation, leukocyte adhesion, and vascular smooth muscle proliferation. ^{35 36}  

Overall, our study suggests that patients in the Sardinian population who have G6PD deficiency may have a significantly lower risk of RVO. However, further studies are necessary for a better understanding of the mechanism by which G6PD deficiency may offer protection against vascular disorders and to establish whether G6PD-deficient patients of non-Sardinian ancestry show an equally reduced susceptibility to RVO. 

 Submitted for publication September 7, 2006; revised November 14, 2006 and February 11, 2007; accepted April 10, 2007.

 Disclosure: A. Pinna, None; C. Carru, None; G. Solinas, None; A. Zinellu, None; F. Cart a, None

 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.

 

Figure 1.

 

Pentose phosphate pathway. G6PD deficiency may offer protection against RVO. G6PD deficiency is one of the most common human genetic abnormalities, affecting an estimated 400 million people worldwide. Evidence indicates that G6PD-deficient persons are protected against ischemic heart and cerebrovascular disease. In this study, the frequency of G6PD deficiency in patients with RVO, a significant cause of ocular morbidity, was lower than expected. Logistic regression analysis revealed that G6PD deficiency was significantly associated with decreased risk for RVO. Overall, these data suggest that G6PD deficiency may offer protection against RVO.

Table 1.

 

Data for Patients with RVO

	RVO (n = 448)	CRVO (n = 194)	BRVO (n = 254)
Affected eye, n (%)
Right eye	220 (49.1)	93 (47.9)	127 (50)
Left eye	228 (50.9)	101 (52.1)	127 (50)
Visual acuity (mean)
Affected eye	0.35	0.28	0.41
Fellow eye	0.84	0.83	0.85
Intraocular pressure, mean mm Hg ± SD
Affected eye	15.4 ± 2.9	15.8 ± 3.4	15.0 ± 2.4
Fellow eye	15.3 ± 2.7	15.6 ± 2.8	15.2 ± 2.3
Glaucoma history, n (%)
Yes	25 (5.6)	15 (7.7)	10 (3.9)
No	423 (94.4)	179 (92.3)	244 (96.1)

Table 2.

 

Systemic Characteristics of Patients and Control Subjects

	RVO (n = 448)	CRVO (n = 194)	BRVO (n = 254)	Control (n = 896)	P
										RVO vs. Control	CRVO vs. Control	BRVO vs. Control
Age, years (mean ± SD)	65.0 ± 11.7	65.2 ± 13.1	64.9 ± 10.6	65 ± 11.7	0.97	0.91	0.88
Systolic blood pressure, mm Hg (mean ± SD)	146.3 ± 17.4	146.2 ± 16.6	146.4 ± 18.0	134.7 ± 13.1	0.00001	0.00001	0.00001
Diastolic blood pressure, mm Hg (mean ± SD)	84.8 ± 9.9	84.7 ± 9.5	84.9 ± 10.2	80.8 ± 8.5	0.00001	0.00001	0.00001
Plasma glucose, mg/dL (mean ± SD)	102.7 ± 33.4	101.2 ± 33.9	103.8 ± 33.0	105.4 ± 29.3	0.12	0.07	0.45
Plasma cholesterol, mg/dL (mean ± SD)	207.4 ± 35.9	207.2 ± 31.1	207.5 ± 39.2	186.7 ± 27.4	0.00001	0.00001	0.00001
G6PD deficiency, n (%)
Yes	21 (4.7)	7 (3.6)	14 (5.5)	107 (11.9)	0.00001	0.0006	0.003
No	427 (95.3)	187 (96.4)	240 (94.5)	789 (88.1)
Systemic hypertension, n (%)*
Yes	251 (56)	99 (51)	152 (59.8)	397 (44.3)	0.0001	0.0088	0.0001
No	197 (44)	95 (49)	102 (40.2)	499 (55.7)
Diabetes mellitus, †
Yes	60 (13.4)	29 (15)	31 (12.2)	134 (15)	0.44	0.99	0.27
No	388 (86.6)	165 (85)	223 (87.8)	762 (85)
Hypercholesterolemia, ‡
Yes	160 (35.7)	70 (36.1)	90 (35.4)	146 (16.3)	0.0001	0.00001	0.00001
No	288 (64.3)	124 (63.9)	164 (64.6)	750 (83.7)
Cardiovascular history
Angina/MI	47 (10.5)	21 (10.8)	26 (10.2)	65 (7.3)	0.043	0.09	0.12
No	401 (89.5)	173 (89.2)	228 (89.8)	831 (92.7)
Cerebrovascular history
TIA/stroke	15 (3.3)	10 (5.2)	5 (2)	28 (3.1)	0.82	0.16	0.33
No	433 (96.7)	184 (94.8)	249 (98)	868 (96.9)

*  Blood pressure >140 mm Hg systolic or >90 mm Hg diastolic or taking antihypertension medication.

†  Fasting plasma glucose ≥126 mg/dL and/or plasma glucose ≥200 mg/dL 2 hours after a 75-g oral glucose load or taking insulin or oral hypoglycemics.

‡  Fasting plasma cholesterol >220 mg/dL or taking lipid-lowering drugs.

Table 3.

 

Multiple-Condition Logistic Regression Analysis

Factor	RVO					CRVO					BRVO
		OR	95% CI	P	OR	95% CI	P	OR	95% CI	P
G6PD deficiency (yes/no)	0.39	0.24–0.64	0.0001	0.26	0.11–0.60	0.002	0.50	0.27–0.91	0.023
Hypertension (yes/no)	1.53	1.19–1.97	0.001	1.46	0.98–2.16	0.06	1.59	1.16–2.20	0.004
Diabetes (yes/no)	0.80	0.56–1.13	0.204	1.03	0.61–1.74	0.92	0.64	040–1.03	0.067
Hypercholesterolemia (yes/no)	3.01	2.25–4.03	0.0001	3.91	2.44–6.28	0.0001	2.37	1.67–3.37	0.0001

 Data show ORs for RVO (n = 448), CRVO (n = 194), and BRVO (n = 254). Number of controls: 896 (416 men, 480 women).

Table 4.

 

Logistic Regression Analysis Adjusted for Systemic Hypertension, Diabetes, and Hypercholesterolemia

	RVO					CRVO					BRVO
		OR	95% CI	P	OR	95% CI	P	OR	95% CI	P
All patients	0.38	0.24–0.63	0.0001	0.25	0.11–0.58	0.0005	0.47	0.27–0.83	0.008
Men	0.43	0.20–0.92	0.02	0.37	0.13–1.05	0.05	0.51	0.17–1.57	0.23
Women	0.35	0.18–0.66	0.0008	0.08	0.01–0.54	0.0009	0.50	0.25–1.03	0.05

 Data show ORs for RVO (n = 448: 208 men, 240 women), CRVO (n = 194: 106 men, 88 women), and BRVO (n = 254: 102 men, 152 women). Number of controls: 896 (416 men, 480 women).

Table 5.

 

Logistic Regression Analysis, Adjusted for Systolic Blood pressure, Plasma Glucose, and Cholesterol

	RVO					CRVO					BRVO
		OR	95% CI	P	OR	95% CI	P	OR	95% CI	P
All patients	0.39	0.23–0.67	0.001	0.29	0.12–0.72	0.008	0.48	0.25–0.91	0.024
Men	0.46	0.21–1.04	0.06	0.49	0.17–1.47	0.21	0.45	0.13–1.50	0.19
Women	0.35	0.17–0.69	0.003	0.11	0.02–0.59	0.01	0.49	0.22–1.05	0.07

 Data are as described in Table 4 .

Table 6.

 

Logistic Regression Analysis, Adjusted for Diastolic Blood Pressure, Plasma Glucose, and Cholesterol

	RVO					CRVO					BRVO
		OR	95% CI	P	OR	95% CI	P	OR	95% CI	P
All patients	0.38	0.23–0.63	0.0001	0.30	0.13–0.72	0.007	0.44	0.23–0.83	0.011
Men	0.44	0.19–0.98	0.045	0.55	0.19–1.63	0.29	0.37	0.11–1.25	0.11
Women	0.34	0.17–0.66	0.001	0.13	0.03–0.61	0.01	0.47	0.22–0.99	0.047

 Data are as described in Table 4 .