Purpose
To investigate risk factors for primary open-angle glaucoma (POAG) and pseudoexfoliative glaucoma (PEXG) in the Thessaloniki Eye Study.
Design
Cross-sectional, population-based study.
Methods
Randomly selected subjects 60 years of age and older (n = 2554) participated in the Thessaloniki Eye Study. Glaucomatous damage and presence of pseudoexfoliation (PEX) were defined according to specific criteria. Open-angle glaucoma (OAG) subjects were compared with nonglaucoma subjects (controls), POAG subjects were compared with controls without PEX, and PEXG subjects were compared with controls with PEX for demographic, lifestyle, ophthalmic, and systemic factors. Factors with P < .2 in the univariate analysis were retained for multivariate analyses.
Results
In multivariate analysis restricted to persons who participated in clinic visits and who had at least 1 phakic eye (n = 2078), intraocular pressure (odds ratio [OR], 1.21 per 1 mm Hg; P < .001), PEX (OR, 2.81; P < .001), history of coronary artery bypass or vascular surgery (OR, 1.95; P = .017), and moderate-to-high myopia (≥ −3 diopters; OR, 2.40; P = .009) were associated with higher odds for OAG. In analysis including all clinic visits (n = 2261), age became significantly associated (OR, 1.05; P = .004). In multivariate analysis for POAG (n = 1840), associations were found for age (OR, 1.04 per year; P = .048), IOP (OR, 1.19 per 1 mm Hg; P < .001), history of coronary artery bypass or vascular surgery (OR, 2.09; P = .01), and history of diabetes treated with insulin (OR, 3.05; P = .045). In multivariate analysis for PEXG (n = 238), the latter was associated with increased IOP (OR, 1.25 per 1 mm Hg; P < .001).
Conclusions
IOP was the only factor associated with both POAG and PEXG, whereas moderate-to-high myopia showed borderline significance in both. Vascular systemic diseases and their treatment were associated only with POAG. The implications of these differences for the pathogenesis between the 2 common types of OAG should be explored further.
Recently, an increasing amount of data has been reported from population-based studies and randomized clinical trials on risk factors for open-angle glaucoma (OAG). Identifying risk factors for a disease contributes substantially to our knowledge because it provides clues to the cause and enables the identification of groups at increased risk, which then can be targeted for early detection. To date, although some risk factors for OAG have been established, the results have been inconsistent. This may be explained partly by differences in study designs, the variety of definitions used for glaucoma, and the fact that few variables have been included consistently in risk factors analyses across studies. However, these discrepancies also could reflect differences among populations in terms of genetic predisposition and exposure to environmental factors. Therefore, assessing risk factors in different populations is important, not only to verify previously published data, but also to identify new factors associated with increased OAG risk.
Further, there is limited understanding of factors predisposing to the development of pseudoexfoliative glaucoma (PEXG), which represents one of the more common types of secondary OAG. Although it has been reported that pseudoexfoliation (PEX) is the most common identifiable cause of OAG, most subjects with PEX do not have glaucoma. Also, based on a recent study by Arnarsson and associates, only a small proportion of subjects with PEX progressed to glaucoma over a 5-year period. The known association of PEX with increased intraocular pressure (IOP) may suggest that the increased risk of glaucoma in subjects with PEX is mediated through elevated IOP. However, based on a previous Thessaloniki Eye Study report, subjects with PEX had a much higher risk of glaucoma compared with subjects without PEX for the same level of screening IOP, an association found for IOP higher than 20 mm Hg. The above suggests that there may be factors other than IOP contributing to increased vulnerability of the optic disc in the presence of PEX. Prevalence studies thus far have not provided data specifically on factors associated with the development of glaucoma among subjects with PEX. Investigating risk factors specifically for PEXG may increase our knowledge on the pathogenesis of this type of glaucoma. In addition, it could provide insight into whether PEXG is a secondary type of OAG or whether it represents a variant of primary open-angle glaucoma (POAG), as has been hypothesized by some investigators.
We previously presented the Thessaloniki Eye Study results on the prevalences of OAG, POAG, and PEXG in a well-defined population, in Thessaloniki, Greece. The aim of this study was to investigate risk factors associated with POAG and PEXG, as identified in the same population-based setting. To facilitate comparisons with previous studies, we also present risk factors associated with OAG in our population.
Methods
The Thessaloniki Eye Study is a cross-sectional population-based study of chronic eye diseases in the population of Thessaloniki, which is the major urban center in Northern Greece. The study was approved by the Aristotle University Hospital Ethics Committee and the Institutional Review Board of the University of California, Los Angeles. All study procedures adhered to the principles outlined in the Declaration of Helsinki for research involving human subjects, and all participants gave written informed consent before their participation.
Details about the recruitment process have been described previously. Briefly, the initial recruitment frame of the Thessaloniki Eye Study consisted of 5000 people 60 years of age or older who were selected randomly in February 1999 from approximately 321 000 persons registered in the municipality registers of the city of Thessaloniki. Randomization was provided by the municipality statistical service. A certificate of registration in the municipality registers is obligatory and is required by the Thessaloniki Police Department for issuance of a Greek citizen photograph identification card. A similar certificate is required by Greek authorities for obtaining a passport. Thus, every Greek citizen residing in the municipality possesses a mandatory and unique identifier, making municipality registers a very reliable source of data for the general population of Greece. Subjects from the Thessaloniki Eye Study recruitment group were contacted by phone or mail to ascertain their willingness to participate in the study. Subjects who agreed to participate were invited to the Thessaloniki Eye Study center at the Aristotle University of Thessaloniki for an extensive ophthalmic screening examination. A home visit eye examination was arranged for persons unable to visit the study examination center because of illness or major disability. Among the 3617 eligible subjects, 2554 participated in the study (participation rate, 71%); of these, 2261 (89%) had the clinic visit examination and 293 (11%) had the home visit examination. To minimize misclassification bias, only clinic visit participants, who had uniformly collected data, were included in the present analyses.
All clinic visit participants were interviewed for demographic data (age, sex), ophthalmic and systemic diseases (hypertension, diabetes, cardiovascular disease, history of heart attack, coronary artery bypass or vascular surgery, and migraines), systemic medications (use of antihypertensive and diabetes treatment), and lifestyle (smoking, alcohol consumption, diet, and hours of sleep per day). Visual acuity was measured with the Early Treatment Diabetic Retinopathy Study chart and habitual correction was recorded. If visual acuity was less than 20/30 with habitual correction, full refraction was performed and best-corrected visual acuity was measured. Ocular measurements also included Humphrey automated perimetry (Carl Zeiss Meditec, Dublin, California, USA), slit-lamp examination, applanation tonometry, gonioscopy, and dilated fundus examination. Details of observation procedures are described elsewhere. Specifically, all subjects underwent 76-suprathreshold visual field testing, followed by 30-2 full-threshold or Swedish interactive threshold algorithm-standard C-30 visual field testing in case of unreliable or abnormal results. Also, IOP was measured using a calibrated Goldmann applanation tonometer (Haag-Streit, Bern, Switzerland). The mean IOP of 3 readings in each eye was defined as the pressure for that eye. Gonioscopy was performed in all study participants using a 4-mirror Sussmann lens in dim, ambient illumination, with a shortened slit that does not fall on the pupil. The angle was graded according to the Spaeth gonioscopic grading system. An angle was considered occludable if the pigmented trabecular meshwork was not visible in more than 180 of angle in static gonioscopy. In addition, dilation was conducted in all study participants; those with an occludable angle underwent laser peripheral iridotomy and were examined subsequently under pupil dilation. Blood pressure (BP) was considered as the average of 2 readings taken with an automated sphygmomanometer (OMRON 705 CP; OMRON Corporation, Kyoto, Japan) at least five minutes apart in the same arm, with the cuff approximately level with the heart. Readings were obtained before instillation of mydriatic drops and after the participant was seated for 10 minutes. Also, weight and height were measured.
Definitions
Details about glaucoma definitions were described in a previous report. In summary, a 2-scale definition of glaucoma was used to avoid omitting subjects with mildly atypical findings or those with some missing data. Definition 1 was based on conservative and strict criteria that required the presence of both structural and functional damage, regardless of IOP. Specifically, structural damage required the presence of thinning or notching or cup-to-disc ratio asymmetry of 0.2, and functional damage required a confirmed threshold glaucomatous visual field defect. In addition, subjects were classified as having glaucoma when the clinical judgment was strongly in favor of the presence of glaucoma, even though the strict criteria (requiring both visual field defect and optic disc abnormality) were not fulfilled (definition 2). This was applied in (1) cases with missing data (unable to perform visual field testing or unreliable visual field test results secondary to low vision), (2) cases with only visual field damage presenting typical characteristics of glaucomatous visual field defect, (3) cases with only optic disc damage (thinning or notching of the optic disc rim combined with matching asymmetry of more than 0.2 cup-to-disc ratio), or (4) cases with high IOP or a history of high IOP combined with optic disc findings (thinning or notching of the optic disc rim or asymmetry between the 2 eyes of more than 0.2 cup-to-disc ratio). Three independent ophthalmologist graders were responsible for the assessment of the presence of glaucomatous appearance of the optic disc (thinning or notching). A consensus agreement between at least 2 of them was required to assign the diagnosis of glaucoma. When disagreement between the graders existed, an open discussion for final classification and diagnosis was carried out. The principal investigator (F.T.) examined all study participants and was responsible for the final adjudication of diagnosis.
Subjects were classified as having POAG if they had glaucoma and an open, normal-appearing anterior chamber angle and the absence of other secondary causes of glaucoma in either eye. Subjects were classified as having PEXG if they had glaucoma and PEX in either eye. OAG was considered to be the combination of POAG and PEXG. Subjects were classified as controls in the absence of glaucoma.
The definition of PEX has been described previously. In summary, PEX was defined as the presence of pseudoexfoliative material in either eye at the pupil margin, on the lens capsule, or both. Before pupil dilation, a detailed high-magnification slit-lamp assessment of the pupil margin was performed. After pupil dilation, the anterior lens surface from each eye was scanned from left to right using a narrow slit-lamp beam and then was examined using a vertical broad slit-lamp beam, looking specifically for early signs of PEX, including pregranular radial lines, as well as established granular deposits. The detection of PEX required consensus agreement between at least 2 of the 3 independent ophthalmologist graders.
Statistical Analyses
Subjects with OAG were compared with controls in a univariate model with regard to the following demographic, lifestyle, ophthalmic, and systemic variables: age, sex, smoking status, alcohol consumption, consumption of leafy green vegetables, body mass index (BMI), number of hours of sleep per day, higher IOP between the 2 eyes, BP status (systolic BP [SBP] and diastolic BP [DBP] within normal limits, SBP and DBP within normal limits with antihypertensive treatment, and high BP), diabetes, diabetes treatment, diabetes status (no diabetes, diabetes without treatment, and diabetes with treatment), diabetes treated with insulin, history of cardiovascular disease, history of heart attack, history of coronary artery bypass or vascular surgery, history of migraines, PEX, and myopia.
Three categories were used to describe smoking status: nonsmoker (has never smoked at least once daily for 1 year), current smoker (has smoked at least once daily for 1 year and has smoked within the past year), or exsmoker (has smoked at least once daily for 1 year but has not smoked within the past year). Regular alcohol consumption was considered to be consumption of any type of alcohol with a frequency of once weekly or more. BMI was calculated as weight divided by the square of height. Based on BMI, subjects were classified as normal (<25), overweight (≥25 and <30), or obese (≥30). SBP and DBP were defined as within normal limits when less than 140 mm Hg and less than 90 mm Hg, respectively. High BP was defined as SBP of 140 mm Hg or more, DBP of 90 mm Hg or more, or both. These cutoffs are in accordance with the seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. Diabetes was defined as a self-reported history of diabetes, diabetes treatment, findings of diabetic retinopathy in at least 1 eye during dilated fundus examination, or a combination thereof. Diabetes treatment included both oral medications and insulin. The spherical equivalent of refractive error was calculated as the spherical value plus half of the cylindrical value. Myopia was analyzed among subjects with at least 1 phakic eye and was defined as mild when −1 to −3 diopters (D) and moderate to high when more than −3 D. Myopia for each subject was based on the eye with the highest myopia.
In univariate analysis, the t test was used for continuous variables and chi-square tests were used for categorical variables. P values were considered statistically significant when less than .05. Factors with a P value less than .2 in the univariate analysis were selected for the regression model. Because pseudophakic subjects could be misclassified as non-PEX, as nonmyopic, or both, only subjects with at least 1 phakic eye were included in the multivariate analysis. The multivariate analysis was also run to include all clinic visit participants.
To identify risk factors separately for POAG and PEXG, the above analyses were conducted separately for each 1 of the 2 types of OAG. Specifically, in univariate models, subjects with POAG were compared with controls who did not have PEX (non-PEX controls), and subjects with PEXG were compared with controls who had PEX (PEX controls).
Results
Risk Factors for Open-Angle Glaucoma
Among clinic visit participants, 135 OAG subjects and 2126 controls were identified and included in the analysis. In the univariate analysis, age (mean ± standard deviation [SD], 73.2 ± 6.0 years vs 70.7 ± 5.7 years; P < .001) and higher IOP between the 2 eyes (mean ± SD, 19.8 ± 5.9 mm Hg vs 15.7 ± 3.7 mm Hg; P < .001) were statistically significantly higher in OAG patients compared with controls. With regard to systemic diseases, diabetes (32 [24%] vs 339 [16%]; P = .017), treatment for diabetes (25 [19%] vs 254 [12%]; P = .023), diabetes treated with insulin (10 [7%] vs 50 [2%]; P = .002), and history of coronary artery bypass or vascular surgery (23 [17%] vs 216 [10%]; P = .012) were more frequent in OAG patients compared with controls. Also, there was a borderline significance for different distribution of diabetes status ( P = .051) and BP status ( P = .063) between the groups. In addition, PEX in at least 1 eye was more frequent in OAG patients compared with controls (41 [30%] vs 229 [11%]; P < .001), and the distribution of myopia among subjects with at least 1 phakic eye was statistically significantly different between the groups ( P < .001).
In the multivariate analysis restricted to those with at least 1 phakic eye ( Table 1 ), increased IOP (odds ratio [OR], 1.21 per 1 mm Hg; 95% confidence interval [CI], 1.16 to 1.27; P < .001) and a history of coronary artery bypass or vascular surgery (OR, 1.95; 95% CI, 1.13 to 3.37; P = .017) were associated with higher odds for OAG. Also, the presence of PEX in at least 1 eye (OR, 2.81; 95% CI, 1.76 to 4.49; P < .001) and moderate to high myopia (≥ −3 D; OR, 2.4; 95% CI, 1.24 to 4.64; P = .009) were associated with increased risk for OAG. When the multivariate analysis was run to include all clinic visit participants, increased age was an additional factor associated with higher odds for OAG (OR, 1.05 per year; 95% CI, 1.01 to 1.08; P = .004).
| Risk Factor | OR | 95% CI | P Value |
|---|---|---|---|
| Age (per year) | 1.03 | 1.00 to 1.07 | .084 |
| Higher IOP between 2 eyes (per 1 mm Hg) | 1.21 | 1.16 to 1.27 | <.001 |
| BP status | |||
| SBP/DBP WNL | Ref (1) | — | — |
| SBP/DBP WNL-Tx | 1.69 | 0.83 to 3.45 | .15 |
| High BP (high SBP and/or high DBP) | 1.32 | 0.74 to 2.35 | .35 |
| Diabetes status | |||
| No diabetes | Ref (1) | — | — |
| Diabetes without treatment | 0.98 | 0.39 to 2.46 | .96 |
| Diabetes with treatment | 1.07 | 0.56 to 2.04 | .85 |
| Diabetes treated with insulin, yes vs no | 2.46 | 0.87 to 6.97 | .089 |
| Coronary artery bypass or vascular surgery, yes vs no | 1.95 | 1.13 to 3.37 | .017 |
| PEX in at least 1 eye, yes vs no | 2.81 | 1.76 to 4.49 | <.001 |
| Myopia in at least 1 eye | |||
| No | Ref (1) | ||
| Mild (> −1 D and < −3 D) | 1.83 | 0.95 to 3.51 | .069 |
| Moderate to high (≥ −3 D) | 2.40 | 1.24 to 4.64 | .009 |
a Among 2261 subjects (135 OAG patients and 2126 controls), 2078 (120 OAG patients and 1958 non-OAG patients) were included in the regression analysis because of bilateral pseudophakia or aphakia, missing values, or both.
Risk Factors for Primary Open-Angle Glaucoma
To assess risk factors specifically for POAG, 94 POAG subjects and 1897 non-PEX controls were identified among clinic visit participants and were included in a separate analysis. In the univariate model ( Table 2 ), age (mean ± SD, 72.7 ± 6.1 years vs 70.3 ± 5.5 years; P < .001) and higher IOP between the 2 eyes (mean ± SD, 19.2 ± 5.2 mm Hg vs 15.7 ± 3.7 mm Hg; P < .001) were statistically significantly higher in POAG patients compared with non-PEX controls. With regard to systemic diseases, the distribution of BP status ( P = .033) and of diabetes status ( P = .01) was statistically significantly different between POAG patients and non-PEX controls. Also, diabetes (26 [28%] vs 305 [16%]; P = .003), treatment for diabetes (21 [23%] vs 228 [12%]; P = .003), diabetes treated with insulin (10 [11%] vs 46 [2%]; P < .001), and a history of coronary artery bypass or vascular surgery (19 [20%] vs 195 [10%]; P = .002) were more frequent in POAG patients compared with non-PEX controls. Among subjects with at least 1 phakic eye, the distribution of myopia was statistically significantly different in POAG patients compared with non-PEX controls ( P = .01).
| Risk Factor | POAG (n = 94) | Controls without PEX (n = 1897) | P Value | PEXG (n = 41) | Controls with PEX (n = 229) | P Value |
|---|---|---|---|---|---|---|
| Age (years) | ||||||
| Mean ± SD | 72.7 ± 6.1 | 70.3 ± 5.5 | <.001 a | 74.5 ± 5.6 | 73.6 ± 6.4 | .41 a |
| Female gender, n (%) | 38 (40%) | 848 (45%) | .42 b | 20 (49%) | 115 (50%) | .87 b |
| Smoking status, n (%) | .74 b | .75 b | ||||
| Nonsmoker | 42 (45%) | 892 (47%) | 23 (56%) | 114 (50%) | ||
| Current smoker | 18 (19%) | 392 (21%) | 7 (17%) | 46 (20%) | ||
| Former smoker | 34 (36%) | 613 (32%) | 11 (27%) | 69 (30%) | ||
| Regular alcohol consumption, n (%) | 52 (55%) | 1052 (55%) | .98 b | 26 (63%) | 115 (50%) | .12 b |
| Leafy green vegetables, n (%) | .92 b | .87 b | ||||
| <1 time/week | 7 (7%) | 123 (6%) | 4 (10%) | 17 (7%) | ||
| 1 to 3 times/week | 51 (54%) | 1058 (56%) | 22 (54%) | 128 (56%) | ||
| ≥1 time/day | 36 (38%) | 716 (38%) | 15 (37%) | 84 (37%) | ||
| BMI, kg/m 2 | ||||||
| Mean ± SD | 28.6 ± 4.7 | 28.4 ± 4.3 | .57 b | 27.4 ± 3.9 | 28.0 ± 4.2 | .44 b |
| Normal (<25) c | 22 (23%) | 367 (19%) | .23 b | 10 (25%) | 52 (23%) | .60 b |
| Overweight (25 to 29.9) | 39 (41%) | 951 (51%) | 22 (55%) | 113 (50%) | ||
| Obese (≥30) | 33 (35%) | 565 (30%) | 8 (20%) | 63 (28%) | ||
| Sleep, n (%) | ||||||
| 8 hrs or more/day | 51 (54%) | 981 (52%) | .64 b | 21 (51%) | 112 (49%) | .79 b |
| In the afternoon | 72 (77%) | 1383 (73%) | .43 b | 30 (73%) | 147 (64%) | .27 b |
| Higher IOP between 2 eyes (mm Hg) | ||||||
| Mean ± SD | 19.2 ± 5.2 | 15.7 ± 3.7 | <.001 a | 21.3 ± 7.2 | 16.0 ± 3.5 | <.001 a |
| BP status, n (%) | .033 b | .42 b | ||||
| SBP/DBP WNL | 10 (11%) | 405 (21%) | 7 (17%) | 41 (18%) | ||
| SBP/DBP WNL-Tx | 19 (20%) | 290 (15%) | 4 (10%) | 42 (18%) | ||
| High BP (high SBP and/or high DBP) | 65 (69%) | 1199 (63%) | 30 (73%) | 146 (64%) | ||
| Diabetes, n (%) | 26 (28%) | 305 (16%) | .003 b | 6 (15%) | 34 (15%) | .97 b |
| Diabetes treatment, n (%) | 21 (23%) | 228 (12%) | .003 b | 4 (10%) | 26 (11%) | 1.00 b |
| Diabetes status, n (%) | .010 b | .86 b | ||||
| No diabetes | 67 (72%) | 1588 (84%) | 35 (85%) | 195 (85%) | ||
| Diabetes without treatment | 5 (5%) | 77 (4%) | 2 (5%) | 8 (3%) | ||
| Diabetes with treatment | 21 (23%) | 228 (12%) | 4 (10%) | 26 (11%) | ||
| Diabetes treated with insulin | 10 (11%) | 46 (2%) | <.001 b | 0 | 4 (2%) | 1.00 b |
| Cardiovascular disease, n (%) | 34 (36%) | 643 (34%) | .67 b | 16 (39%) | 79 (35%) | .60 b |
| History of heart attack, n (%) | 11 (12%) | 174 (9%) | .41 b | 5 (12%) | 18 (8%) | .36 b |
| Coronary artery bypass or vascular surgery, n (%) | 19 (20%) | 195 (10%) | .002 b | 4 (10%) | 21 (9%) | 1.00 b |
| Migraines, n (%) | 3 (3%) | 84 (4%) | .80 b | 0 | 10 (4%) | .37 b |
| Myopia in at least 1 eye among subjects with at least 1 phakic eye (n = 1850), n (%) | .010 b | .012 b | ||||
| No | 65 (77%) | 1564 (89%) | 26 (70%) | 169 (84%) | ||
| Mild (> −1 D and < −3 D) | 12 (14%) | 127 (7%) | 2 (5%) | 18 (9%) | ||
| Moderate to high (≥ −3 D) | 7 (8%) | 75 (4%) | 9 (24%) | 15 (7%) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
