Association of pseudoexfoliation syndrome with cardiovascular and cerebrovascular disease: a systematic review and meta-analysis Helen Chung, MD,* Sourabh Arora, MD, † Karim F. Damji, MD, † Ezekiel Weis, MD, MPH* ,† ABSTRACT ● Objective: Pseudoexfoliation syndrome (PEX) is a systemic disease, but evidence of its association with cardiovascular disease (CVD) and cerebrovascular disease (CVA) is controversial. A quantitative systematic review will provide an accurate summary of the current body of the literature. Design: Meta-analysis. Participants: Not applicable. Methods: A comprehensive literature search of published and unpublished English-language studies was performed. Summary statistics were calculated using inverse variance weighting and are presented in forest plots. Sources of variance were evaluated statistically. Results: After screening 4547 studies, 47 articles were reviewed, and 25 eligible studies were selected that reported patients from around the world. Twenty studies enrolling 9583 individuals with PEX evaluated CVD, providing a summary odds ratio (OR) of 1.61 (95% CI 1.37–1.90). Eleven studies, enrolling 1308 PEX patients, evaluated CVA and generated a summary OR of 1.76 (1.40–2.22). For any vascular event (AVE) using all 25 studies, there were 9716 PEX patients and 363,312 control patients, yielding a summary OR of 1.64 (95% CI 1.39–1.92). Analysis for publication bias with the Egger’ s test was not significant for studies reporting CVD and AVE (p ¼ 0.92 and 0.64, respectively) but was significant for CVA (p ¼ 0.03). Asymmetry of Begg’ s funnel plot was noted for the CVA and AVE analyses. Multiple sensitivity analyses were performed, including assessment of study quality; the OR for all 3 outcomes varied minimally and remained significant in all analyses. Conclusion: There is strong evidence that PEX is significantly associated with both CVD and CVA. Pseudoexfoliation syndrome (PEX) is a common age- related systemic disease affecting the elastin microfibrillar system. It is characterized by white-grey extracellular flaky material within the anterior segment 1 and was first described 100 years ago by Dr. John Lindberg. 2 Increas- ingly, there is evidence of the effect of PEX on the systemic vascular system. 3–8 Population-based and hospital-based studies asses- sing the association between PEX and cardiovascular disease (CVD) 9–28 and cerebrovascular disease (CVA) 6,9,11,13,18,22,24,26,29–31 report conflicting findings. Eight studies note a statistically significant association with CVD, 9,12,14,16,21–24 whereas 10 studies report no association. 11,13,15,17–20,26–28 Similarly, 3 studies report a statistically significant association with CVA, 11,18,22 whereas 6 studies report no association. 6,9,13,24,26,30 A meta-analysis by Siordia et al. 32 found that PEX is statistically significantly associated with ischemic heart disease; however, stringent search criteria limited the included studies. Another meta-analysis by Wang et al. 33 found positive associations between PEX and CVD and CVA, yet their methodology did not allow for accurate assessment of critical methodo- logical issues such as double-counting study participants, and several new studies have been published since. Given limitations in prior quantitative reviews, we believed that an updated meta-analysis was needed. The purpose of our study was to quantitatively summarize the current body of literature on this topic, which almost doubles the number of articles included in the previously published meta-analyses. METHODS Literature Search and Eligibility Criteria Detailed methodology used has been previously described by Shah et al. 34 In summary, PubMed, EMBASE, and the Cochrane Database were searched using key terms “pseudoexfoliation” and “exfoliation syndrome” until May 2016. ProQuest, Theses Canada Portal, CARL Harvester, DART Europe E-theses, NDLTD, EThOS, and Center for Research Libraries were also searched. A hand search of bibliographies was performed. Corresponding authors of included studies were contacted about relevant unpublished data. To maximize sample size, inclusion criteria were broad. Studies were excluded if the original article was not in English, if the authors did not follow standard epidemio- logic methodology allowing for comparison between cases & 2017 Canadian Ophthalmological Society. Published by Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.jcjo.2017.10.039 ISSN 0008-4182/17 CAN J OPHTHALMOL —VOL. ], NO. ], ] 2017 1