Effect of the Coronavirus Disease 2019 Pandemic on Retinal Artery Occlusion Authors: Tetsuya Muto, Shigeki Machida, Shinichiro Imaizumi, Tetsuju Sekiryu, Deepali Published: 2026 ABSTRACT AbstractPurposeTo determine the prevalence and clinical characteristics of retinal artery occlusion (RAO), including branch RAO (BRAO) and central RAO (CRAO), among new patients before and after the declaration of a state of emergency due to coronavirus disease 2019 (April 7, 2020) in Japan.MethodsNew patients and patients with newly diagnosed RAO were categorized into “Before” and “After” groups based on the initial visit. The prevalence, sex ratio, and age of patients newly diagnosed with RAO were compared between the groups. The rates of patients newly diagnosed with rhegmatogenous retinal detachment (RRD) and pseudoexfoliation glaucoma (PEX) were also examined in the Before and After groups.ResultsThis study included 53 and 31 patients with BRAO and 55 and 40 patients with CRAO in the Before and After groups, respectively. There were no significant differences in prevalence of RAOs among newly diagnosed patients (BRAO: p = 0.449; CRAO: p = 0.060), patients with RRD (BRAO: p = 0.425; CRAO: p = 0.082), or patients with PEX (BRAO: p = 0.317; CRAO: p = 0.072), as well as in sex ratio and age. Although initial best‐corrected visual acuity (BRAO: p = 0.303; CRAO: p = 0.291) of RAOs showed no significant differences, the final best‐corrected visual acuity in BRAO significantly worsened in the After group (p = 0.002).ConclusionsAlthough the coronavirus disease 2019 pandemic did not affect the prevalence of RAOs among new patients compared with that of RRD and PEX, it may have affected its pathophysiology.Keywords: best-corrected visual acuity, branch retinal artery occlusion, central retinal artery occlusion, coronavirus disease 2019, prevalence 1. Introduction The coronavirus disease 2019 (COVID‐19) caused by the severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) is one of the worst pandemics of humankind [1], which resulted in confusion and major changes worldwide. The World Health Organization declared a public health emergency of international concern for COVID‐19 on January 30, 2020, and recommended a policy to prevent the transmission of infection [2]. SARS‐CoV‐2 is an enveloped, plus‐stranded RNA virus with a single‐stranded RNA genome of approximately 30,000 nucleotides [3]. Initially, it was speculated that SARS‐CoV‐2 caused only respiratory functional disorder. However, several reports have confirmed that COVID‐19 caused thrombosis during the course of infection [4–7]. Vaccine development had progressed rapidly to prevent additional COVID‐19 expansion. Between the end of 2020 and 2021, both mRNA vaccines produced by Pfizer/BioNTech and Moderna and adenovirus vector vaccines produced by AstraZeneca and Janssen Pharmaceuticals had been used worldwide; to date, more than 100 vaccines have been developed [8]. In February 2021, mRNA vaccination was initiated in Japan, with cases of COVID‐19 vaccine‐induced thrombosis being reported thereafter [9, 10]. SARS‐CoV‐2 binds to angiotensin‐converting enzyme 2 (ACE2) receptors, which are highly expressed on endothelial cells. In addition, severe COVID‐19 is associated with hyperinflammatory responses. Pathogenesis may lead to an increase in vascular disease clinically [11]. The association of COVID‐19 with vasculitis may cause vascular events in the patients in the ophthalmology clinic. Both retinal vein occlusion (RVO) and retinal artery occlusion (RAO) are common thrombotic diseases in the ophthalmological setting [12, 13]. The occurrence rate of RVO and RAO might have increased because COVID‐19 itself and its vaccine occasionally lead to the development of thrombosis. Surgery for rhegmatogenous retinal detachment (RRD) and pseudoexfoliation glaucoma (PEX) was selected as a control for comparison. To the best of our knowledge, no reports have indicated that the occurrence rates of surgery for these two diseases were affected by the COVID‐19 pandemic. The objective of this study was to investigate the occurrence rate and pathophysiology of branch RAO (BRAO) and central RAO (CRAO) before and after the declaration of a state of emergency (April 7, 2020) in Japan. 2. Materials and Methods In this study, we retrospectively evaluated 95 patients with naïve CRAO and 84 patients with naïve BRAO diagnosed at Dokkyo Medical University Saitama Medical Center (Koshigaya, Japan) and Fukushima Medical University (Fukushima, Japan) between April 2015 and December 2022. During the same period, 44,967 new patients, 264 with RRD and 116 with PEX, were also evaluated. Patients with symptoms lasting less than 1 month before examination and a history of more than two medical examinations after treatment were included. The CRAO and BRAO diagnosis was based on fundus examination and fluorescein angiographic findings, determined by one of the investigators (Tetsuya Muto). RRD cases were limited to those undergoing scleral buckling surgery as a first‐time procedure. Similarly, PEX cases were limited to those undergoing trabeculectomy or receiving Ex‐Press® (Alcon Laboratories, Fort Worth, TX, USA) implant as a first‐time surgery. New patients and those with newly diagnosed CRAO, BRAO, RRD, and PEX were categorized into “Before” and “After” groups. CRAO and BRAO were categorized based on the initial diagnosis and the declaration of the state of emergency, while RRD and PEX were categorized based on the initial surgery and the state of emergency. Prevalence rates were compared between the two groups. We excluded patients with undetermined symptom onset, symptom duration of more than 1 month before examination, senile dementia, complication of RVO, and dense cataracts resulting in poor image quality. Patients with a history of vitrectomy were also excluded. However, patients with pseudophakia who had undergone uncomplicated cataract surgery more than 3 months prior were included. This study adhered to the tenets of the Declaration of Helsinki. The Institutional Review Board of Dokkyo Medical University Saitama Medical Center and Fukushima Medical University approved the study protocol. Due to the retrospective nature of the study, the need for informed patient consent was waived by the Ethics Committee of both institutions. At the initial visit, all patients underwent comprehensive ophthalmologic evaluations, including standardized refraction, measurement of best‐corrected visual acuity (BCVA) using a Landolt ring, slit‐lamp biomicroscopy, optical coherence tomography, and color fundus photography. BCVA was expressed as the logarithm of the minimal angle of resolution (logMAR). Patients with BCVA of counting fingers were assigned a logMAR value of 2.6, hand motion value of 2.7, light perception value of 2.8, and no light perception value of 2.9 [14]. Ophthalmologic evaluations, including BCVA, slit‐lamp biomicroscopy, and OCT, were performed, with fluorescein angiography conducted as required. Retinal photocoagulation was performed in cases of vast nonperfusion areas observed on fluorescein angiography. All data are expressed as the mean ± standard deviation. A stepwise multivariable analysis with a p value threshold of 0.2 was used to identify the factors associated with the final BCVA. Factors included sex, age, laterality, initial BCVA, duration from onset to initial visit, ocular massage, oral administration of kallidinogenase, eye solution for glaucoma, combination of neovascular glaucoma (NVG), intravenous drip of vasodilator or thrombolytic agent, and oral administration of anticoagulant. Statistical software included StatMate® version V for Macintosh (ATMS, Tokyo, Japan) for unpaired t tests and chi‐squared tests and JMP 16 Pro (SAS Institute, Inc., Cary, NC, USA) for multivariable analysis. The chi‐squared test was used to determine the statistical significance of the prevalence rates, sex ratio, laterality ratio, NVG, blindness rates, and better visual prognosis rates. An unpaired t‐test was used to compare the groups before and after the state of emergency. Differences with a p value < 0.05 were considered statistically significant. 3. Results Table 1 shows patient demographics of BRAO and CRAO in the Before and After groups. No significant differences were found in prevalence rates, sex ratio and age between the Before and After groups of RAOs among newly diagnosed patients, patients with RRD, or those with PEX. TABLE 1Patient demographics of BRAO and CRAO in the Before and After groups. Before groupAfter group p valuePatients newly diagnosed with BRAO/new patients53/3012231/148450.449Patients newly diagnosed with CRAO/new patients55/3012240/148450.060Patients newly diagnosed with BRAO/patients newly diagnosed with RRD53/17931/850.425Patients newly diagnosed with CRAO/patients newly diagnosed with RRD55/17940/850.082Patients newly diagnosed with BRAO/patients newly diagnosed with PEX53/8131/350.317Patients newly diagnosed with CRAO/patients newly diagnosed with PEX55/8140/350.072Sex of patients with BRAO (male/female)33/2017/140.503Sex of patients with CRAO (male/female)30/2529/110.075Age of patients with BRAO (years)69.45 ± 13.5167.87 ± 19.160.687Age of patients with CRAO (years)69.73 ± 14.8669.53 ± 14.650.948Abbreviations: BRAO, branch retinal artery occlusion; CRAO, central retinal artery occlusion; PEX, pseudoexfoliation glaucoma; RRD, rhegmatogenous retinal detachment. Patient demographics of BRAO and CRAO in the Before and After groups. Abbreviations: BRAO, branch retinal artery occlusion; CRAO, central retinal artery occlusion; PEX, pseudoexfoliation glaucoma; RRD, rhegmatogenous retinal detachment. Table 2 shows the characteristics of 84 eyes of BRAO and 95 eyes of CRAO in the Before and After groups. No significant differences were observed in all factors except the time between the initial visit and symptom onset (p = 0.033) and final BCVA (p = 0.002) in BRAO. Time was shorter and the final BCVA was poorer in the After group compared with the Before group. TABLE 2Characteristics of BRAO (84 eyes) and CRAO (95 eyes) in the Before and After groups. Before groupAfter group p valueTime between initial visit and symptom onset in BRAO (days)4.91 ± 6.452.68 ± 2.900.033Time between initial visit and symptom onset in CRAO (days)3.07 ± 5.853.10 ± 6.500.983Initial BCVA in BRAO0.49 ± 0.940.85 ± 0.330.303Initial BCVA in CRAO1.80 ± 1.031.99 ± 0.890.291Laterality in BRAO (right/left)34/1915/160.157Laterality in CRAO (right/left)24/3119/210.709Outbreak of NVG in BRAO (yes/no)1/520/310.442Outbreak of NVG in CRAO (yes/no)2/532/380.744NLP/more than LP after treatment in BRAO4/490/310.117NLP/more than LP after treatment in CRAO10/453/370.233Final BCVA in BRAO0.68 ± 1.110.80 ± 0.390.002Final BCVA in CRAO1.54 ± 1.051.65 ± 1.020.647Final BCVA less than zero in BRAO (yes/no)29/2419/120.557Final BCVA less than zero in CRAO (yes/no)7/483/370.630Abbreviations: BCVA, best‐corrected visual acuity; BRAO, branch retinal artery occlusion; CRAO, central retinal artery occlusion; LP, light perception; NLP, nonlight perception; NVG, neovascular glaucoma. Characteristics of BRAO (84 eyes) and CRAO (95 eyes) in the Before and After groups. Abbreviations: BCVA, best‐corrected visual acuity; BRAO, branch retinal artery occlusion; CRAO, central retinal artery occlusion; LP, light perception; NLP, nonlight perception; NVG, neovascular glaucoma. Table 3 shows the factors associated with final BCVA using multivariable analysis of 53 eyes in the Before group and 31 eyes in the After group of BRAO. Initial BCVA (p < 0.001) and combination of NVG (p = 0.011) were associated with final BCVA in the Before group. TABLE 3Associated factors with final BCVA of BRAO in the Before and After groups using the multiple logistic regression model with a backward stepwise selection method.Factors95% CI p value Before group (n = 53) Sex−0.377 to 0.3210.871Age−0.011 to 0.0160.731Laterality−0.330 to 0.3970.852Initial BCVA0.593 to 0.976< 0.001Time between initial visit and symptom onset (days)−0.032 to 0.0220.717Ocular massage−0.028 to 0.8420.066Oral administration of kallidinogenase−0.335 to 0.5100.678Eye solution for glaucoma−0.243 to 0.6890.340Combination of NVG−2.771 to −0.3820.011Intravenous drip of vasodilator or thrombolytic agent−0.510 to 0.4410.884Oral administration of anticoagulant−0.152 to 0.6320.223 After group (n = 31) Sex−0.092 to 0.4720.175Age0.000 to 0.0150.041Laterality−0.088 to 0.6200.133Initial BCVA0.078 to 0.5790.013Time between initial visit and symptom onset (days)−0.054 to 0.0550.987Ocular massage−0.804 to 0.3980.489Oral administration of kallidinogenase−0.372 to 0.2770.763Eye solution for glaucoma−0.281 to 0.5820.476Combination of NVG0.018 to 1.2660.044Intravenous drip of vasodilator or thrombolytic agentn/an/aOral administration of anticoagulant−0.521 to 0.1180.203Abbreviations: BCVA, best‐corrected visual acuity; BRAO, branch retinal artery occlusion; CI, confidence interval; n/a, not available; NVG, neovascular glaucoma. Associated factors with final BCVA of BRAO in the Before and After groups using the multiple logistic regression model with a backward stepwise selection method. Abbreviations: BCVA, best‐corrected visual acuity; BRAO, branch retinal artery occlusion; CI, confidence interval; n/a, not available; NVG, neovascular glaucoma. However, age (p = 0.041), initial BCVA (p = 0.013), and intravenous drip of the vasodilator or thrombolytic agent (p = 0.044) was significantly associated in the After group. Table 4 shows the factors associated with final BCVA using multivariable analysis of 54 eyes in the Before group and 40 eyes in the After group of CRAO. Initial BCVA (p < 0.001) was significantly associated in both groups. TABLE 4Associated factors with final BCVA of CRAO in the Before and After groups using the multiple logistic regression model with a backward stepwise selection method.Factors95% CI p value Before group (n = 54) Sex−0.692 to 0.5570.828Age−0.015 to 0.0280.520Laterality−0.801 to 0.5430.700Time between initial visit and symptom onset (days)−0.088 to 0.0560.654Ocular massage−0.722 to 0.9910.753Oral administration of kallidinogenase−0.037 to 1.7790.060Eye solution for glaucoma−0.893 to 1.1450.804Combination of NVG−3.385 to 1.8610.561Oral administration of anticoagulant−0.382 to 1.1810.308 After group (n = 40) Sex−0.631 to 1.1270.569Age−0.025 to 0.0240.994Laterality−0.800 to 0.6630.849Time between initial visit and symptom onset (days)−0.022 to 0.1010.198Ocular massage−1.816 to 0.2660.010Oral administration of kallidinogenase−0.918 to 0.8640.951Eye solution for glaucoma−1.295 to 0.3060.217Combination of NVG−2.607 to 1.1500.435Oral administration of anticoagulant−1.400 to 0.2160.145Abbreviations: BCVA, best‐corrected visual acuity; CI, confidence interval; CRAO, central retinal artery occlusion; NVG, neovascular glaucoma. Associated factors with final BCVA of CRAO in the Before and After groups using the multiple logistic regression model with a backward stepwise selection method. Abbreviations: BCVA, best‐corrected visual acuity; CI, confidence interval; CRAO, central retinal artery occlusion; NVG, neovascular glaucoma. 4. Discussion According to our results, the occurrence rate of RAOs did not change between the two groups. Table 5 presents a summary of the findings from large‐scale data analyses and systematic reviews on this topic [15–24]. Al‐Moujahed et al. reported an initial increase in CRAO cases during the early months of the pandemic [15]. However, RAO diagnostic rates as a proportion of all new retinal clinic diagnoses remained stable throughout most of the pandemic [15]. Cho et al. reported that the overall incidence of RAO after the COVID‐19 pandemic did not increase; in fact, it exhibited a slight decrease compared with that before the outbreak [16]. Modjtahedi et al. observed no significant change in the RAO incidence related to COVID‐19 [17]. Park et al. found no significant relationship between COVID‐19 infection [18, 19] or vaccination [18] and the incidence of RAO or RVO. Au et al. concluded COVID‐19‐related CRAO cases were just coincidental [20]. Both Yeung et al. and Singh et al. did not mention the correlation between RAO and COVID‐19 vaccination in their reviews [21, 22]. In contrast, Li et al. and Feltgen et al. reported a significant increase in RAO and RVO after COVID‐19 vaccination [23, 24]. TABLE 5Association between RAOs and COVID‐19 pandemic from large‐scale data analyses and systematic reviews.No.AuthorsNumber of patientsConclusions1Al‐Moujahed et al. [15]442,186 patientsThe number of patients with RAO increased during the first few months of the COVID‐19 pandemic.2Cho et al. [16]32,028 patients diagnosed with RAOThe incidence of RAO did not increase during the COVID‐19 outbreak.3Modjtahedi et al. [17]432,515 patients diagnosed with COVID‐19 infectionThere was a smaller increase in the incidence of RAOs after COVID‐19 diagnosis4Park et al. [18]8,418,590 patientsNo increase in the hazard ratio of RAO relative to COVID‐19 or COVID‐19 vaccination, except for a possible increase in the hazard ratio of RAO among women vaccinated with mRNA‐1273.5Park et al. [19]21,696 patients with RAOSARS‐CoV‐2 infection did not significantly increase the incidence of RAO.6Au and Ko [20]15 patients with CRAOCOVID‐19 related CRAO cases were just coincident.7Yeung et al. [21]15 patients with RAOCorrelation between RAO and COVID‐19 vaccination was not mentioned.8Singh et al. [22]433 patients with RAOCorrelation between RAO and COVID‐19 vaccination was not mentioned.9Li et al. [23]1,478,132 patientsThe risk of RAO and retinal vein occlusion significantly increased during the first 2 weeks after vaccination.10Feltgen et al. [24]421 patients with RAO and retinal vein occlusionNo evidence of any association was found between SARS‐CoV‐2 vaccination and a higher risk for RAO and retinal vein occlusion.Abbreviations: COVID‐19, coronavirus disease 2019; CRAO, central retinal artery occlusion; RAO, retinal artery occlusion; RVO, retinal vein occlusion; SARS‐CoV‐2, severe acute respiratory syndrome coronavirus 2. Association between RAOs and COVID‐19 pandemic from large‐scale data analyses and systematic reviews. Abbreviations: COVID‐19, coronavirus disease 2019; CRAO, central retinal artery occlusion; RAO, retinal artery occlusion; RVO, retinal vein occlusion; SARS‐CoV‐2, severe acute respiratory syndrome coronavirus 2. Although the occurrence rate and male rate of CRAO tended to increase in the After group, the differences were barely not significant. The incidence of several diseases, such as acute primary angle closure [25], Vogt–Koyanagi–Harada disease [26], optic neuritis [27], dry eye [28], and infectious endophthalmitis after vitrectomy [29], increased after the COVID‐19 pandemic. Rubino et al. reported that the prevalence of thromboembolic events was significantly higher in children with multisystem inflammatory syndrome after the COVID‐19 pandemic [30]. Theoretically, it is no wonder that RAOs increase like the above diseases [25–30]. Time between initial visit and symptom onset in BRAO was shorter in the After group compared with the Before group, and we speculated that remote work had enabled workers to visit the hospital. Because workers were easy to use time freely in such situations. However, it had not a significant difference in CRAO, and the reason for this observation was undetermined. Although the time between initial visit and symptom onset was shorter in the After group compared with the Before group, the final BCVA of BRAO was poorer in the former group. Shah et al. reported that a shorter symptom duration before treatment did not influence final visual acuity outcomes in CRAO [31]; this result might be proper. Initial BCVA and NVG of BRAO were associated with final BCVA in the Before group, while age, initial BCVA, and intravenous drip of vasodilator or thrombolytic agent were associated in the After group. There were no NVG cases observed in the After group; thus, it was not possible to perform statistical analysis. In general, NVG rarely develops in patients with BRAO [32, 33]. However, the development of NVG is associated with poor visual prognosis [34]. Mason et al. reported that visual prognosis after BRAO seems to be correlated to presenting visual acuity [35]. Yuzurihara and Iijima reported that the visual acuity of patients with CRAO is poor at presentation, and the prognosis is generally poor, with a few exceptions [36]. In contrast, the visual acuity of patients with BRAO is markedly better both at presentation and the final visit [36]. In general, arteriosclerosis of young age is milder than in old age, and reopening of the artery occlusion tends to occur. Although we could not find reports regarding the use of an intravenous drip of a vasodilator or thrombolytic agent in BRAO, several studies reported the use of such therapy in CRAO [37–39]. Schrag et al. showed that systemic fibrinolysis is significantly effective only within the first 4.5 h of symptom onset [37]. Janská et al. stated that intravenous thrombolytic therapy in CRAO appears to be effective and safe [38]. Suzuki et al. showed that visual acuity and retinal vein constriction improved after liposomal prostaglandin E1 therapy [39]. Although Yasuda et al. reported that female sex (odds ratio = 15.63, p = 0.032) was independently associated with worse visual acuity in BRAO [40], our results did not reveal associations with sex. Analysis of additional data are required to confirm these findings. In this study, we selected RRD and PEX as control diseases. The prevalence rates of several ocular diseases were affected by COVID‐19 [25–30]. Nevertheless, there were no reports found concerning the change in the prevalence rates of RRD and PEX according to the results of a search in the PubMed database (keywords: “COVID‐19, RRD” or “COVID‐19, buckling surgery” or “COVID‐19, pseudoexfoliation syndrome” or “COVID‐19, glaucoma surgery”; search years: 2000–2024). Particularly, PEX was considered the best control disease because the age of onset of RAOs and the age of PEX that required trabeculectomy or use of an Ex‐Press® implant as a first‐time surgery were almost the same. This study had several limitations. First, the tendency for an increase in the occurrence and male rates of CRAO in the After group was barely not statistically significant. Analysis of a larger sample could have yielded different results. Second, we only investigated ocular massage in cases with a description on the medical record. It is possible that ocular massage was performed without description in the record. Third, the number of individuals who had COVID‐19 or received vaccination was unclear due to the retrospective nature of the study. Fourth, the detailed mechanism by which these factors might influence the clinical features of RAOs remains unclear. 5. Conclusions Our results show that the prevalence of RAOs did not increase during the COVID‐19 pandemic. However, the clinical features of RAOs may have been influenced mildly by the pandemic. Funding The authors declare that they have received no funding. Conflicts of Interest The authors declare no conflicts of interest. Acknowledgments The authors wish to thank Enago for English editing of the manuscript. The authors also extend their gratitude to Yasuo Haruyama of the Department of Public Health, School of Medicine, Dokkyo Medical University, for supporting the statistical analyses. Muto, Tetsuya , Machida, Shigeki , Imaizumi, Shinichiro , Sekiryu, Tetsuju , Effect of the Coronavirus Disease 2019 Pandemic on Retinal Artery Occlusion, Journal of Ophthalmology, 2026, 5740618, 7 pages, 2026. 10.1155/joph/5740618 Academic Editor: Deepali Contributor Information Tetsuya Muto, Email: ueda.castle@gmail.com. Deepali, Email: dedeepali@wiley.com. Data Availability Statement The datasets used and/or analyzed during this study are available from the corresponding author upon reasonable request. References Associated Data Data Availability Statement The datasets used and/or analyzed during this study are available from the corresponding author upon reasonable request.