The horizontal line in the middle of the box indicates median; top and bottom borders of the boxes, 75th and 25th percentiles; error bars, range of data points within 1.5 times the interquartile range; and data points shown as circles, outliers.
Vardhan S A, Haripriya A, Ratukondla B, Ramulu P, Shivakumar C, Nath M, Vijayaraghavan P, Robin AL. Association of Pseudoexfoliation With Systemic Vascular Diseases in a South Indian Population. JAMA Ophthalmol. 2017;135(4):348–354. doi:10.1001/jamaophthalmol.2017.0064
Are patients with pseudoexfoliation more likely to demonstrate cardiovascular risk factors compared with persons without pseudoexfoliation?
In this cross-sectional postoperative study of 1406 individuals in India, 930 patients with pseudoexfoliation demonstrated higher systolic blood pressure and a greater likelihood of electrocardiographic abnormalities compared with 476 control patients. No association was noted between pseudoexfoliation and other risk factors, including blood glucose, serum cholesterol, and serum homocysteine levels.
Indian patients with pseudoexfoliation presenting for cataract surgery show higher blood pressures and a slightly higher frequency of cardiac morbidity than do patients without pseudoexfoliation presenting for cataract surgery.
Ocular pseudoexfoliation (PEX) syndrome may be associated with systemic vascular diseases, which might suggest a broader health significance of PEX, although previous reports are conflicting.
To determine whether prespecified vascular risk factors and cardiac abnormalities are more common among patients with PEX than among control individuals without PEX.
Design, Setting, and Participants
This cross-sectional analysis of patients recruited into the Aravind Pseudoexfoliation study included South Indian patients older than 40 years with or without PEX who required cataract surgery. Surgical procedures were performed at 4 tertiary Aravind eye hospitals in Tamil Nadu, India, from December 2, 2010, through March 26, 2012. Nine hundred thirty patients with PEX and 476 non-PEX controls underwent detailed ocular examinations, including specific ocular features reflecting PEX. Patients also underwent evaluation for multiple systemic potential cardiovascular diseases and their risk factors. The data collection for this analysis on systemic vascular diseases started on December 2, 2010, and ended on April 30, 2014. This study analysis was specified in the study protocol.
Main Outcomes and Measures
Five cardiovascular outcomes compared between patients with and without PEX included blood glucose, cholesterol, and homocysteine levels; blood pressure; and cardiac morbidity (defined by electrocardiographic [ECG] abnormalities).
The study analysis included 930 patients in the PEX group and 476 in the non-PEX group. The mean (SD) ages of patients in the PEX and non-PEX groups were 64.8 (6.8) and 59.9 (7.3) years (P < .001), respectively. More patients in the PEX group were men (470 [50.5%] vs 460 women [49.5%]) than in the non-PEX group (201 [42.2%] vs 275 women [57.8%]; P < .001). In multivariable analyses adjusting for age and sex, higher systolic blood pressure values were noted for the PEX group (difference [Δ], 4.0 mm Hg; 95% CI, 1.7-6.2 mm Hg; P = .001). Also, patients in the PEX group were more likely to demonstrate an ECG abnormality than in the non-PEX group (odds ratio, 1.64; 95% CI, 1.04-2.60; P = .03). Pseudoexfoliation was not observed to be associated with a higher level of blood glucose (Δ, 6.2 mg/dL; 95% CI, −2.0 to 14.3 mg/dL; P = .14), serum cholesterol (Δ, −0.6 mg/dL; 95% CI, −5.1 to 4.0 mg/dL; P = .81), or serum homocysteine level (Δ, 0.004 mg/L; 95% CI, −0.12 to 0.14; P = .96).
Conclusions and Relevance
Of the 5 cardiovascular outcomes examined in South Indian patients requiring cataract surgery, PEX was associated with higher systolic blood pressure and more frequent ECG abnormalities but not with higher blood glucose, serum cholesterol, or serum homocysteine levels. Patients with PEX are more likely to require attention to blood pressure and cardiac morbidity.
Pseudoexfoliation (PEX) syndrome is a systemic disorder characterized by the deposition of a fibrillar extracellular material in intraocular and extraocular tissues.1 Pseudoexfoliation is more commonly found in older individuals, and although it can be seen in all countries, its prevalence varies by geographic region, ethnicity, and race.2- 6 Pseudoexfoliation is a known risk factor for the development of cataract and glaucoma.7- 10 Its presence may increase the risk for complications during and after cataract surgery and for late dislocation of intraocular lenses.11- 13
Several reports have suggested that patients with PEX have a higher risk for several systemic diseases. Furthermore, PEX material has been identified in many visceral organs, such as the skin, heart, blood vessels, brain, lung, and kidney, making such a hypothesis plausible.14,15 The range of systemic diseases associated with PEX is broad and includes several vascular,16,17 cardiovascular,18- 27 and cerebrovascular28- 30 conditions.
Various pathogenetic mechanisms have been postulated to explain the higher risk for vascular disease in patients with PEX, including the deposition of PEX material in blood vessels, increasing vascular resistance, and vascular dysregulation. Insults to the vascular endothelium and smooth muscle function by PEX deposits have also been postulated as a mechanism for atherosclerosis and thrombus formation, which could lead to cardiac events.21,31- 33 Higher levels of coronary risk markers, such as lipoprotein A, apolipoprotein A, and homocysteine, have also been previously reported in patients with PEX compared with control individuals without PEX, mainly as parts of relatively small clinical studies.34 Higher levels of systemic homocysteine have been suggested as a possible reason for an increase in cardiovascular disease given that homocysteine,35- 40 an amino acid synthesized during methionine metabolism, is an independent risk factor for cardiovascular disease in some studies.41
Many questions remain regarding the association between PEX and cardiovascular disease. For example, some studies18,20,22,23,27 report a positive association of PEX with ischemic heart diseases and blood pressure, whereas others42,43 did not find significant associations. In addition, many prior studies were limited by small sample sizes and/or a reliance on hospital health records to assess systemic diseases.
Herein we evaluated potential vascular risk factors and cardiac morbidity (as determined by electrocardiographic [ECG] abnormalities) in a large sample of patients with PEX and non-PEX controls undergoing cataract surgery as part of the Aravind Pseudoexfoliation study. The purposes of the present study are (1) to determine whether specific vascular risk factors are more common in patients with PEX and (2) whether patients with PEX more frequently have cardiac damage (as judged by ECG abnormalities).
The present study is part of a cross-sectional analysis nested within the Aravind Pseudoexfoliation long-term prospective clinical study, which was designed to evaluate the long-term ocular, systemic, and cataract surgery–related complications associated with PEX. This study followed the tenets of the Declaration of Helsinki,44,45 was approved by the Aravind institutional review board, and was registered at clinicaltrials.gov.46 Written informed consent was obtained from all participants after explanation of the nature and possible consequences of the study.
Patients with or without PEX who were older than 40 years and had consented to undergo cataract surgery at 1 of 4 Aravind eye hospitals in Tamil Nadu, India, from December 2, 2010, through March 26, 2012, were eligible to participate. Patients who had preoperative phacodonesis,47 patients with elevated intraocular pressure requiring trabeculectomy, or those with any prior intraocular surgery in the eye planned for cataract surgery were excluded. Before surgery, a detailed dilated ocular examination to identify specific ocular features reflecting PEX was performed. Participants were classified as having PEX if any PEX material was present in at least 1 eye. The presence of systemic disease was not a contraindication for study inclusion. The present analysis represents a secondary aim of the study, with the primary aim being a comparison of intraoperative and postoperative cataract surgery outcomes in the PEX and non-PEX groups.
We examined all participants for 5 prespecified cardiovascular variables, including blood pressure, blood glucose level, cardiac disease (defined by ECG abnormalities), serum cholesterol level, and serum homocysteine level. All participants were asked about any history of treatment for types 1 and 2 diabetes, hypertension, ischemic heart disease, and dyslipidemia. All blood pressure measurements were performed by one of us (B.R.) using a manual sphygmomanometer after the patient had been seated for 5 minutes. If the first blood pressure measurement was 140/90 mm Hg or higher, the patient was allowed to sit still for another 10 minutes, a second blood pressure measurement was taken, and the mean of the 2 readings was calculated as the final measurement. Blood pressure was analyzed as a continuous measure, and hypertension was diagnosed if the systolic blood pressure was 140 mm Hg or higher and/or diastolic blood pressure was 90 mm Hg or higher or if patients were already receiving treatment for hypertension.
Venous blood was drawn to measure random blood glucose, total serum cholesterol, and serum homocysteine levels. The glucose oxidase–peroxidase enzymatic and cholesterol oxidase–phenol 4-aminoantipyrine methods were used to measure the plasma glucose and total cholesterol levels, respectively. To determine homocysteine levels, a separate 4-mL blood sample was collected, centrifuged for 10 to 15 minutes, and stored at −20°C. Homocysteine was analyzed by the enzymatic recycling method. All samples were analyzed in an automated biochemistry analyzer (Selectra E; Vital Scientific NV), and all tests were performed with the same brand of reagents and analyzer (Agappe Diagnostics for glucose levels, Spinreact for total cholesterol levels, and Dialab for homocysteine levels).
We analyzed blood glucose level as a continuous outcome and defined diabetes as a random blood glucose level of at least 200 mg/dL (to convert to millimoles per liter, multiply by 0.0555) or currently receiving treatment for diabetes. We characterized participants with serum total cholesterol values of at least 210 mg/dL (to convert to millimoles per liter, multiply by 0.0259) as having a high cholesterol level. Homocysteine values exceeding 2.03 mg/L (to convert to micromoles per liter, multiply by 7.397) were considered elevated. Cholesterol and homocysteine levels were also analyzed as continuous variables.
Twelve-lead ECGs (MAC 1200 ST digital system; GE Medical Systems) were measured by a trained nurse and read by a masked internist (B.R.). All ECG abnormalities were further categorized into ischemia, conduction defects, sinus tachycardia or bradycardia, left ventricular hypertrophy (LVH), and other nonspecific changes. More specifically, the presence of ischemic heart disease was noted if the patient had a history of anticoagulation treatment for myocardial infarction, if the patient had undergone coronary artery bypass surgery or angioplasty with stenting in the past, or if the current ECG showed definite abnormalities, such as pathologic Q waves in the ECG or ST-segment elevation or depression, as suggested by joint European Society of Cardiology and American College of Cardiology guidelines.48 Sinus bradycardia was diagnosed if the heart rate was less than 60 beats/min with normal sinus rhythm, and sinus tachycardia was diagnosed if the heart rate was greater than 100 beats/min with normal sinus rhythm. Conduction defects were diagnosed in patients who presented with ECG findings suggestive of right or left bundle branch block, left anterior fascicular block, or left posterior fascicular block based on published guidelines.49 We diagnosed LVH based on an R wave in V5 or V6 exceeding 25 mm, an S wave in V1 or V2 exceeding 25 mm, or the sum of the R wave in V5 or V6 plus the S wave in V1 or V2 exceeding 35 mm.50 Asymmetric T-wave inversion and ST depression of less than 1 mm were characterized as other ECG changes.
The data collection for this analysis on systemic vascular diseases started on December 2, 2010, and ended on April 30, 2014. This study analysis was specified in the study protocol. We used χ2 tests to assess differences between groups for categorical variables such as sex, diagnosis of diabetes, hypertension, hypercholesterolemia, hyperhomocysteinemia, and ECG abnormalities. We used a 2-sample t test to evaluate group differences in continuous variables, including age, blood glucose level, blood pressure, serum cholesterol level, and serum homocysteine level. Multivariable linear regression and logistic regression analyses were performed to adjust for relevant confounders, with age and sex chosen as potential confounders in all multivariable models. In models in which ECG abnormalities (cardiac morbidity) were considered as the dependent variable, additional confounders were added, including diabetes, hypertension, and cholesterol levels. All data analyses were performed with STATA software (version 14; StataCorp).
A total of 930 patients with PEX and 476 non-PEX controls were enrolled and completed the study protocol. Patients with PEX were older than the controls (mean [SD] age, 64.8 [6.8] vs 59.9 [7.3] years; P < .001) (Table 1). The PEX group also had a higher percentage of men (470 [50.5%] vs 460 women [49.5%]) than the control group (201 [42.2%] vs 275 women [57.8%]; P < .001). All participants were of South Indian ethnicity.
In unadjusted analyses, patients with PEX had a higher mean (SD) systolic blood pressure (131.8 [19.2] vs 127.7 [16.6] mm Hg; P = .002) and mean (SD) arterial pressure (96.1 [10.9] vs 94.3 [10.2] mm Hg; P = .008) compared with non-PEX controls (Figure). The prevalence of hypertension was also higher in the PEX group compared with the non-PEX group (375 [40.3%] vs 163 [34.2%]; P = .04) (Table 1).
In multivariable analyses adjusting for age and sex, higher systolic blood pressure values were noted for patients with PEX compared with non-PEX controls (difference [Δ], 4.0 mm Hg; 95% CI, 1.7 to 6.2 mm Hg; P = .001), but there were no group differences with respect to diastolic BP (Δ, 0.4 mm Hg; 95% CI, −0.7 to 1.5 mm Hg; P = .49) (Table 2). Of note, women had higher systolic blood pressure (Δ, 2.8 mm Hg; 95% CI, 0.77-4.78; P = .01) compared with men. In multiple logistic regression analysis having systemic hypertension as the outcome variable (0 indicates absent; 1, present) and including study groups, age and sex as the prognostic variables showed that the odds of having systemic hypertension in PEX group over the controls was 1.12 (95% CI, 0.87-1.46; P = .38).
In unadjusted analyses and in multivariable analyses adjusting for age and sex, no statistically significant differences were seen between the PEX and non-PEX groups in random mean (SD) blood glucose (125.2 [68.3] vs 119.2 [51.6] mg/dL; Δ, 6.2 mg/dL; 95% CI, −2.0 to 14.3 mg/dL; P = .14) or serum cholesterol (178.3 [35.3] vs 181.2 [37.1] mg/dL; Δ, −0.6 mg/dL; 95% CI, −5.1 to 4.0 mg/dL; P = .81) levels (Table 1 and Table 2). Women had higher serum cholesterol levels (Δ, 16.3 mg/dL; 95% CI, 12.3-20.4 mg/dL; P < .001) compared with men independent of PEX status.
When we classified diabetes and hypercholesterolemia as present or absent, the prevalence of diabetes was higher in the PEX group compared with the non-PEX group (162 [17.4%] vs 62 [13.0%]; P = .007) in unadjusted analyses (Table 1). However, when adjusted for age and sex, this association with diabetes was no longer observed (odds ratio [OR], 1.37; 95% CI, 0.94-2.00; P = .10).
The mean levels of serum homocysteine did not differ between the 2 groups (2.42 [1.12] vs 2.38 [1.08] mg/L; P = .51) (Table 1). In multivariable models adjusting for age and sex, PEX was not associated with higher serum homocysteine levels compared with no PEX (Δ, 0.004 mg/L; 95% CI, −0.12 to 0.14 mg/L; P = .96) (Table 2). Higher homocysteine levels were found in men than in women (Δ, 0.51 mg/L; 95% CI, 0.41-0.64; P < .001), but no association was observed between age and homocysteine levels (Δ, 0 mg/L; 95% CI, −0.01 to 0.01 mg/L; P = .64).
The total number of participants with abnormal ECG findings was 105 in the PEX group and 39 in the control group. In logistic regression models adjusting for age and sex, patients with PEX were more likely to have an ECG abnormality compared with non-PEX controls (OR, 1.55; 95% CI, 1.02-2.35; P = .04) (Table 3). Being male was independently associated with a higher rate of ECG abnormalities in this multivariable logistic regression model (OR, 1.60; 95% CI, 1.11-2.31; P = .01). Additional post hoc analyses evaluated whether PEX was associated with specific types of ECG abnormalities (Table 3). We found that the incidence of LVH was higher in the PEX group than in the control group (OR, 3.81; 95% CI, 1.27-11.5; P = .02). Because cardiac morbidity (ECG abnormalities) has other potential confounders, separate regression models were performed to account for confounders, including diabetes, systemic hypertension, and cholesterol levels. In these models, the PEX group had more ECG abnormalities than did the non-PEX group (OR, 1.64; 95% CI, 1.04-2.60; P = .03). Male sex (OR, 1.63; 95% CI, 1.07-2.47; P = .02) and hypertension (OR, 1.67; 95% CI, 1.12-2.51; P = .01) were other factors associated with a higher likelihood of ECG abnormalities in these regression analyses.
In our sample of South Indian participants scheduled for cataract surgery, patients with PEX had higher systolic blood pressure and a greater likelihood of ECG abnormalities compared with controls without PEX. No association between PEX and other cardiovascular risk factors, such as diabetes, hypercholesterolemia, and hyperhomocysteinemia, was noted. To our knowledge, these findings represent the largest prospectively evaluated cohort in which these associations have been assessed and provide significant insight into the possible systemic implications of PEX.
At least 2 previous population-based studies26,27 reported a higher rate of systemic hypertension among patients with PEX, although multiple other studies20,22,23,42,43 have found no such association. We found that mean systolic blood pressure was higher in the PEX group, although the frequency of hypertension was not significantly different between the PEX and non-PEX groups. This difference likely reflects the fact that PEX is associated with higher blood pressure but not necessarily blood pressure high enough to be classified as clinically significant systemic hypertension. Previous studies19,20,22,26,51 were limited in that they did not analyze blood pressure as a continuous variable and had inconsistent diagnostic criteria for hypertension. One challenge in analyzing blood pressure through continuous measures is the need to account for persons using antihypertensive therapy. In the present study, we found higher systolic blood pressure in patients with PEX despite a greater percentage of these patients having previously diagnosed hypertension and using antihypertensive treatment compared with non-PEX controls (116 [12.5%] vs 42 [8.8%]; P = .05), suggesting that the true difference in untreated blood pressure may actually be larger than what we report herein.
Although a few prior studies20,22,23,27 showed a positive association between cardiovascular disease and PEX, other studies42,51- 53 showed no such association. Previous literature on this topic is limited by the fact that most prior studies were retrospective. Cardiovascular disease includes a broad array of conditions; a variety of cardiac problems, including coronary artery disease, cardiomyopathy, arrhythmias, asymptomatic diastolic dysfunction, and heart failure,18,24,43,54 have been associated with PEX. In the present study, we aimed to identify cardiovascular disease objectively using ECG tracings read by an internist masked to the PEX status. Of note, ECG abnormalities were more common in the PEX group, largely because of the higher prevalence of LVH. High systolic blood pressure is known to lead to LVH,55 reinforcing our finding that PEX is associated with higher systolic blood pressure.
Many prior studies19,42,43 have demonstrated no difference in diabetes rates between patients with and without PEX, although a few studies26,51 have suggested that diabetes may be less common in patients with PEX. In the present study, we noted no association between PEX and blood glucose or serum cholesterol levels. Differences across studies may reflect the different genetic and cultural backgrounds in which these various studies were conducted or the variety of definitions used to define diabetes or hypercholesterolemia.
Elevated homocysteine levels are reported as a risk factor for cardiovascular diseases, and homocysteine levels have been reported to be elevated in patients with PEX in some studies35- 40 but not others.41 We found no differences in mean levels or the distribution of homocysteine levels when comparing the PEX and the control groups. This finding agrees with the results of a study in Turkish patients with PEX41 but conflicts with other reported studies,36- 39 possibly owing to the different populations studied or the specific type of patients with PEX studied (ie, any PEX vs PEX with cataract vs PEX glaucoma). Variations in the observed association of PEX with homocysteine levels in various study populations may also reflect differences in the dietary habits56and methylenetetrahydrofolate reductase (MTHFR) gene polymorphisms57 of the studied populations. The absence of an association between PEX and serum homocysteine levels in the present patient population suggests that homocysteine levels may not mediate the link between PEX and cardiovascular disease.
Strengths of the present study include the various cardiovascular variables assessed and the large sample size of the PEX and control groups. Limitations include the fact that this study is hospital based and that all participants had clinically significant cataracts. Our findings may not apply to those who are younger, do not yet have significant cataracts, have phacodonesis, or have glaucoma for which they previously underwent surgery. Blood tests were performed with nonfasting random samples. Because glucose levels were nonfasting, we set a high cutoff (≥200 mg/dL) for diagnosis of new diabetes that was suggested by prior research.58,59 Likewise, Fokkema et al60 have demonstrated that homocysteine values are valid with random blood samples, producing slightly higher but clinically insignificant changes compared with fasting values. Finally, the cardiac morbidity end point encompasses several cardiac diseases and may not reflect a single condition, although more specific conditions were explored in the post hoc analyses looking at specific cardiac morbidities. In addition, our population was from South India, perhaps limiting generalizability of the findings to persons of other ethnic backgrounds.
In a South Indian population with cataracts, PEX was associated with higher systolic blood pressure and more ECG abnormalities. Because many of the risk factors overlap for vascular and cardiovascular diseases, patients with PEX should undergo an examination for cardiovascular disease (ECG) and systemic hypertension (blood pressure measurement). Routine measurement of homocysteine levels is unlikely to be of diagnostic benefit in our population.
Corresponding Author: Ashok Vardhan S, MD, Aravind Eye Hospital 1, Anna Nagar, Madurai 625 020, Tamil Nadu State, India (email@example.com).
Accepted for Publication: January 3, 2017.
Published Online: March 9, 2017. doi:10.1001/jamaophthalmol.2017.0064
Author Contributions: Drs Vardhan S and Haripriya had full access to all the data and take responsibility for the integrity of the data and the accuracy of data analysis.
Study concept and design: Haripriya, Robin.
Acquisition, analysis, or interpretation of data: Vardhan S, Haripriya, Ratukondla, Ramulu, Shivakumar, Nath, Vijayaraghavan.
Drafting of the manuscript: Vardhan S, Ramulu, Robin.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Vardhan S, Ramulu.
Obtained funding: Haripriya.
Administrative, technical, or material support: Ratukondla, Vijayaraghavan.
Study supervision: Haripriya, Ramulu, Nath, Robin.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Ramulu reports receiving grants from the National Institutes of Health outside the submitted work. Dr Robin reports receiving a nonpaid position from Aerie Pharmaceuticals, DSM, Novaliq, Clearside, and Biolight outside the submitted work. No other disclosures were reported.
Funding/Support: This study was supported by donation of Capsular tension rings from Aurolab and intraocular lenses from Alcon Laboratories.
Role of the Funder/Sponsor: The supporting institutions had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.