Stroke risk in patients with hypertensive retinopathy: a systematic review and meta-analysis Authors: Jing Zhang, Xiaorong Zhang, Yong Liu, Haiping Lian, Haipeng Tang, Xiaodong Wang, Ying Wang Published: 2026 ABSTRACT AbstractBackgroundHypertensive retinopathy (HTRP) is a complication of hypertension and is recognised as hypertension-mediated target-organ damage. Hypertension is also the leading risk factor for stroke. This systematic review and meta-analysis evaluated the stroke risk in patients with HTRP to determine whether HTRP is associated with stroke.MethodsPubMed, Embase, Web of Science and the Cochrane Library were searched from inception to 13 December 2025. Pooled risk ratios with 95% CIs were calculated. For the proportions of HTRP and stroke among patients with hypertension, risk differences and standard errors were calculated. Subgroup analyses were conducted according to HTRP severity (mild and severe) and stroke type (ischaemic and haemorrhagic).ResultsThis meta-analysis included 10 studies involving 26 776 participants. The proportions of HTRP and stroke among all the participants were 0.42 (95% CI 0.23 to 0.61) and 0.15 (0.11 to 0.18), respectively. The proportion of stroke in participants with HTRP was 0.20 (0.14 to 0.27). The pooled risk ratio for stroke in patients with HTRP compared with those without HTRP was 1.87 (1.42 to 2.45). Compared with patients with mild HTRP and those without HTRP, patients with severe HTRP had higher risk of stroke, with risk ratios of 1.88 (1.51 to 2.34) and 3.10 (1.91 to 5.05), respectively. The risk ratios for ischaemic and haemorrhagic stroke in patients with HTRP compared with those without HTRP were 1.79 (1.20 to 2.67) and 1.39 (0.86 to 2.26), respectively.ConclusionsHTRP is associated with an increased stroke risk, particularly ischaemic stroke. These findings suggest that HTRP can serve as an important biomarker for stroke-risk stratification, and that periodic retinal examinations can contribute to stroke prevention.Keywords: STROKE WHAT IS ALREADY KNOWN ON THIS TOPIC Hypertensive retinopathy (HTRP) is recognised as hypertension-mediated target-organ damage. Hypertension is the leading risk factor for stroke. Hypertensive retinopathy (HTRP) is recognised as hypertension-mediated target-organ damage. Hypertension is the leading risk factor for stroke. WHAT THIS STUDY ADDS HTRP is associated with an increased stroke risk. Mild and severe HTRP were associated with 1.80-fold and 3.10-fold higher stroke risks than no retinopathy, respectively. Severe HTRP was associated with a 1.88-fold higher stroke risk than mild HTRP. HTRP is associated with an increased stroke risk. Mild and severe HTRP were associated with 1.80-fold and 3.10-fold higher stroke risks than no retinopathy, respectively. Severe HTRP was associated with a 1.88-fold higher stroke risk than mild HTRP. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY HTRP can serve as an important biomarker for stroke-risk stratification, and that periodic retinal examinations can contribute to stroke prevention. HTRP can serve as an important biomarker for stroke-risk stratification, and that periodic retinal examinations can contribute to stroke prevention. Introduction Hypertensive retinopathy (HTRP) is a complication of hypertension and is recognised as hypertension-mediated target-organ damage.1 HTRP occurs when retinal vessels are damaged due to elevated blood pressure and comprises the following three pathophysiological stages: (1) the early constrictive stage, characterised by generalised narrowing of arteriolar diameter; (2) the intermediate stage, involving focal and generalised arteriolar narrowing due to intimal thickening and medial hyperplasia, resulting in arteriovenous nicking and altered arteriolar light reflexes; and (3) the exudative stage, characterised by isolated or multiple changes, including microaneurysms, flame haemorrhages, hard exudates, cotton-wool spots and papilloedema.2 3 Vision loss in HTRP is attributed to either secondary optic atrophy after prolonged papilloedema or retinal pigmentary changes after exudative retinal detachment. HTRP is more prevalent in men than in women among individuals aged <45 years, whereas among those aged >65 years, it is more prevalent in women than in men.4 Erden et al have shown that the incidence of HTRP is 66.3%, and that increasing incidence of retinopathy is related to both the severity and duration of hypertension.5 Traditionally, HTRP has been diagnosed via funduscopic examination; however, signs of HTRP are now more objectively and reliably documented using digital retinal fundus photography. Several grading systems have been proposed to classify the severity of HTRP. The most widely cited grading system is the Keith-Wagener-Barker classification.6 In this system, mild or moderate generalised retinal arteriolar narrowing is classified as grade 1; definite focal narrowing and arteriovenous nipping as grade 2; the features of grade 2 retinopathy together with retinal haemorrhages, exudates and cotton-wool spots as grade 3; and the features of grade 3 retinopathy with additional papilloedema or retinal oedema as grade 4. Using the Keith-Wagener-Barker classification, Del Brutto et al have reported that grade 1 HTRP is present in 37% and grade 2 in 17% of patients with hypertension.7 Stroke is another complication of hypertension. Hypertension is the leading risk factor for cardiovascular disease and mortality worldwide.8 Hypertension-mediated organ damage, comprising structural and functional changes in arteries or end organs, is a marker of cardiovascular disease, including stroke.9 Stroke includes subarachnoid haemorrhage and haemorrhagic or ischaemic lesions of the brain. In a study involving 21 243 Chinese adults aged ≥35 years with hypertension, the weighted prevalence of stroke was estimated as 1.8%.10 Several studies have also suggested a significant positive association between HTRP and stroke.11–13 The retinal vasculature is an extension of the cerebral vasculature, sharing embryological, anatomical and physiological features.14 Accordingly, HTRP is closely associated with the risk of cerebrovascular diseases, such as stroke. Specifically, patients with moderate HTRP compared with those without HTRP are more likely to experience incident stroke (hazard ratio (HR) 2.37, 95% confidence interval (CI) 1.39 to 4.02).13 Even among patients with hypertension who have achieved good blood pressure control, HTRP remains associated with an increased risk of cerebral infarction.13 Recently, a meta-analysis including only six cohort studies has shown that individuals with HTRP exhibit a significantly higher overall stroke risk than those without HTRP (risk ratio (RR), 1.46; 95% CI 1.29 to 1.65).15 However, several relevant studies were not included in the quantitative synthesis, warranting a comprehensive systematic review and meta-analysis to clarify the association between HTRP and stroke.11 12 16–19 Therefore, the present study aimed to determine the stroke risk in HTRP and the potential association between HTRP and stroke via a comprehensive systematic review and meta-analysis. Materials and methods The systematic review and data collection were performed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2020 guidelines.20 The review was registered with the National Institute for Health and Care Research, with the registration identifier CRD42024537811. The protocol is available at https://www.crd.york.ac.uk/PROSPERO/view/CRD42024537811. Ethical approval was waived by the Ethics Committee of Honghui Hospital, because the study was a review. Search strategy The PICO strategy was as follows: (1) population—hypertension; (2) intervention—blood-pressure control; (3) comparison—HTRP and no HTRP; and (4) outcome—stroke. A comprehensive literature search of PubMed, Embase, Web of Science and the Cochrane Library was conducted from inception to 13 December 2025, using a combination of Medical Subject Headings terms and free-text terms (online supplemental table 1). The following types of studies were excluded: duplicate publications, letters, case reports, reviews, conference abstracts without raw data and articles with incomplete data or those for which the full text could not be retrieved. Potential additional studies were identified through manual searching of the reference lists of selected articles. Data unavailable in the most recent publications were obtained from earlier reports. The detailed search strategy is presented in online supplemental table 1. Study selection Studies were included according to the following criteria: (1) language—English; (2) study design—retrospective or prospective study; (3) participants—patients with hypertension; (4) exposure—HTRP; and (5) outcome—stroke events. Studies that did not meet all the inclusion criteria were excluded. Two reviewers (YL and YW) independently screened titles and abstracts according to the eligibility criteria, and studies were categorised into three groups (included, excluded and unsure) at this stage. Subsequently, two reviewers (JZ and HT) independently examined the full texts of potentially eligible and unclear studies to reach a final decision on inclusion or exclusion. Any disagreements between reviewers were resolved through discussion or consultation with a third reviewer (XW). The assessment criteria for retinopathy and stroke are presented in online supplemental table 2. Grades 1 and 2 of HTRP were classified as mild, and grades 3 and 4 were classified as severe. Data extraction To minimise errors, two reviewers (HL and XZ) independently extracted relevant data using a standardised, predefined data collection form. The extracted data were reviewed and cross-checked by both reviewers. Any disagreements were resolved through discussion or consultation with a third reviewer (XW). For duplicated study populations, the study with the longest follow-up or the largest sample size was selected. Corresponding authors were contacted when the required data could not be extracted from the published articles. The following study characteristics were extracted from each included article: (1) general study characteristics (author, publication year and country of study); (2) study design; (3) sample size; (4) age; (5) classification of HTRP; (6) duration of follow-up; (7) type of stroke; (8) conclusions and (9) numbers of HTRP and stroke cases. Risk-of-bias assessment Two reviewers (HT and YL) independently evaluated the risk of bias of the included studies using the Agency for Healthcare Research and Quality (AHRQ) assessment tool.21 Any disagreements regarding the quality assessment were resolved through discussion or consultation with a third reviewer (YW). The AHRQ tool assesses the risk of bias of individual studies in systematic reviews across the following five domains: selection bias, performance bias, detection bias, attrition bias and reporting bias. For each domain, the risk of bias was judged as low, high or unclear. Each item was scored as 3 points for low risk, 2 points for unclear risk and 1 point for high risk. The maximum total score was 27, with higher scores indicating a lower risk of bias (online supplemental table 3). Statistical analysis Data were entered into Review Manager (RevMan, V.5.3; The Nordic Cochrane Centre, The Cochrane Collaboration, 2014, Copenhagen, Denmark). For the incidence of HTRP and stroke, the risk difference and standard error (SE) were calculated. Pooled RRs with 95% CIs were calculated for stroke in individuals with HTRP compared with those without HTRP. Heterogeneity among studies was assessed using forest plots and tested using the χ² test and the I² statistic. If significant statistical heterogeneity was detected (p<0.1 and I²≥50%), the meta-analysis was performed using the Mantel-Haenszel method with a random-effects model.22 Otherwise, a fixed-effects model was applied. Subgroup analyses were conducted according to the severity of HTRP (mild and severe) and the stroke type (ischaemic and haemorrhagic). Sensitivity analysis was performed by excluding studies with a high risk of bias and those with small sample sizes to assess whether any single study influenced the overall results. Publication bias was assessed using funnel plots. A p value <0.05 was considered to indicate statistical significance. Results Search results The flowchart of the study selection process is presented in figure 1. A total of 466 records were identified, and 441 studies were excluded based on titles and abstracts. Subsequently, 21 potentially relevant articles were identified for full-text review. After detailed assessment according to the eligibility criteria, 11 studies were included in the systematic review. Finally, 10 studies involving a total of 26 776 patients were included in the meta-analysis (table 1). The risk of bias is summarised in figure 2, and the risk-of-bias scores of the 10 included studies are presented in online supplemental table 3. Additionally, eight studies had at least six domains rated as low risk, whereas two studies exhibited a high risk of bias. PRISMA flowchart of the literature search process. PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses. Table 1The characteristics of the studies included in the meta-analysisAuthor, countryStudy designParticipantSample (female)Age, yearClassification of HTRPStrokeFollow-upConclusionChen et al,China11ProspectiveHypertension9793(6066)R: 45–75Mean: 63.3KWBFirst non-fatal/fatal stroke; first ischaemic stroke/haemorrhagic stroke; excluding subarachnoid haemorrhage and silent strokeMedian: over 4.4 yearsHTRP is a predictor of the strokeChopra et al, India17ProspectiveHypertension80 (42)R: 30–70KWBNot described1 yearHTRP increases the chance of strokeDel Brutto et al,Ecuador16ProspectiveArterial pressure ≥120/≥80 mm Hg236(138)≥ 60Mean: 69.3KWBNon-lacunar strokes, incident strokeMean:5.2 yearsHTRP predicts mortality in older adultsHiremath et alIndia12ProspectiveHypertension240 (61)>18KWBIschaemic and haemorrhagic stroke, TIA1 yearSignificant association between HTRP and strokeKabedi et al,Congo18Cross-sectionalHypertension159 (86)≥ 18Mean: 57.9Scheie classificationNot describedNoNo association is found between HTRP and strokeKwon et al,Korea19Cross-sectionalHypertension550(175)R: 25–83Mean: 59.3KWBSilent brain infarctionNoHTRP is associated with the silent brain infarctionLi et al,JPN23ProspectiveUrban population7027(3768)R: 30–79Mean: 54.3KWBIntracerebral/haemorrhage, cerebral infarctionMedian: 16.7 yearsMild HTRP is positively associated with strokeOng et al,AUS13ProspectiveHypertension2907(1733)R: 50–73Mean 60.4None, mild, moderate and malignantCerebral infarction (thrombotic/embolic) or haemorrhagic stroke (subarachnoid/intracerebral)Mean:13 yearsHTRP predicts the long-term risk of strokeSuri et al,USA24ProspectiveResidents4753(2883)R: 25–74Mean 49.8NoIncident ischaemic stroke20 yearsThis study confirms the association between HTRP and strokeHarbaoui et al,France25ProspectiveHypertension1031(394)Mean 44KWBFatal stroke30 yearsRetinopathy is a predictive variable of fatal strokeHTRP, hypertensive retinopathy; KWB, Keith-Wagener-Barker classification; R, range; TIA, transient ischaemic attack. The characteristics of the studies included in the meta-analysis HTRP, hypertensive retinopathy; KWB, Keith-Wagener-Barker classification; R, range; TIA, transient ischaemic attack. Risk ratios of stroke in patients with HTRP. (a) Risk ratio of stroke in HTRP compared with no retinopathy. (b and c) Risk ratio of stroke in mild (grades 1 and 2) (b) and severe (grades 3 and 4) (c) retinopathy compared with no retinopathy. (d) Risk ratio of stroke in severe retinopathy (grades 3 and 4) compared with mild retinopathy. HTRP, hypertensive retinopathy; M-H, Mantel–Haenszel. Study characteristics The characteristics of the 10 included studies are shown in table 1. Among them, five studies were conducted in Asia,11 12 17 19 23 one in North America,24 one in South America16 one in Australia,13 one in Africa18 and one in Europe.25 No randomised controlled trials were identified. There were two cross-sectional studies18 19 and eight prospective case-series studies.11–13 16–19 23–25 HTRP was classified according to the Keith-Wagener-Barker classification in seven studies,11 12 16 19 23 25 two studies used alternative classification systems,13 17 18 and one study did not report a classification system.24 Additionally, four studies reported both ischaemic and haemorrhagic stroke,11–13 23 whereas two studies did not describe the type of stroke.16–18 25 Study outcome The RRs of stroke in HTRP ranged from 1.17 to 8.23 (figure 2a). Substantial heterogeneity was observed across the studies (I²=87%, p<0.01); therefore, a random-effects model was applied. The pooled RR of stroke in patients with HTRP was 1.87 (95% CI 1.42 to 2.45), compared with those without HTRP (figure 2a). The RR of stroke in mild HTRP compared with no HTRP was 1.80 (95% CI 1.42 to 2.29) (figure 2b). The RR of stroke in severe HTRP compared with no HTRP was 3.10 (95% CI 1.91 to 5.05) (figure 2c). The RR of stroke in severe HTRP compared with mild HTRP was 1.88 (95% CI 1.51 to 2.34) (figure 2d). Figure 3 shows the proportions of HTRP and stroke in patients with hypertension. The proportion of HTRP ranged from 0.14 to 0.84 across the 10 studies (figure 3a). Marked heterogeneity was detected (I²=100%, p<0.01); therefore, a random-effects model was used. The pooled rate of HTRP in all the patients was 0.42 (95% CI 0.23 to 0.61). The proportion of stroke in patients with hypertension ranged from 0.03 to 0.82 (figure 3b). High heterogeneity was also observed (I²=99%, p<0.01); accordingly, a random-effects model was applied. The pooled rate of any stroke in patients with hypertension was 0.15 (95% CI 0.11 to 0.18). The proportions of hypertensive retinopathy and stroke among patients with hypertension. Hypertensive retinopathy (a) and stroke (b). Figure 4 shows the proportion of stroke in patients with HTRP. The rate ranged from 0.07 to 0.93 (figure 4a). Substantial heterogeneity was present (I²=99%, p<0.01); therefore, a random-effects model was used. The pooled rate of any stroke in patients with HTRP was 0.20 (95% CI 0.14 to 0.27) (figure 4a). The estimated rates of stroke in mild and severe HTRP were 0.08 (95% CI 0.06 to 0.10) and 0.14 (95% CI 0.11 to 0.17), respectively (figure 4b,c). The proportion of stroke in patients with hypertensive retinopathy. (a) Overall rate of stroke in hypertensive retinopathy. Stroke rate in mild (b) and severe (c) hypertensive retinopathy. Figure 5 shows the RRs for different types of stroke in HTRP. The RR of ischaemic stroke in patients with HTRP was 1.79 (95% CI 1.20 to 2.67) (figure 5a), and the RR of haemorrhagic stroke was 1.39 (95% CI 0.86 to 2.26) (figure 5b). Risk ratios by stroke subtype in patients with hypertensive retinopathy. Risk ratios of ischaemic (a) and haemorrhagic (b) stroke in hypertensive retinopathy compared with no retinopathy. HTRP, hypertensive retinopathy; M-H, Mantel–Haenszel. Publication bias and sensitivity analyses Figure 6 presents the assessment of publication bias for the 10 studies included in the meta-analysis. Considerable heterogeneity was observed across the studies. Sensitivity analysis confirmed that HTRP remained associated with an increased stroke risk (figure 7). After excluding the study with the smallest sample size,17 the RR of stroke in patients with HTRP was 1.85 (95% CI 1.41 to 2.42) (figure 7a). After excluding the study with the highest risk of bias,12 the RR was 1.98 (95% CI 1.58 to 2.49) (figure 7b). After excluding both studies,12 17 the RR was 1.97 (95% CI 1.56 to 2.48) (figure 7c). Assessment of publication bias among the studies included in the meta-analysis. Funnel plots of the incidence (a) and risk ratio (b) of stroke in hypertensive retinopathy. Sensitivity analyses of the pooled risk ratio of stroke in hypertensive retinopathy. Exclusion of the study with the smallest sample size (a), the highest risk of bias (b) and both (c). HTRP, hypertensive retinopathy; M-H, Mantel–Haenszel. Discussion This study provides a comprehensive estimation of the prevalence and RR of stroke in HTRP based on 10 studies involving 26 776 participants. The results demonstrated that the presence of HTRP is associated with an increased stroke risk in patients with hypertension. The prevalence of stroke in patients with HTRP was estimated at 0.20, which was significantly higher than that observed in the overall hypertensive population. These findings indicate the importance of early screening and appropriate management of patients with HTRP to reduce the stroke risk. Hypertension is a well-established and leading risk factor for stroke.26 Notably, stroke is the second leading cause of death worldwide and the second most common cause of disability-adjusted life years.27 Persistently elevated blood pressure causes progressive arterial damage, which can lead to arterial obstruction and thrombus formation, thereby limiting blood supply to vital organs. In the absence of adequate oxygen and nutrients, tissue injury occurs. Without treatment, hypertension may progress and affect multiple organ systems, leading to coronary heart disease, heart failure, atrial fibrillation, dementia and chronic kidney disease. Effective blood pressure control is essential for preventing recurrent ischaemic stroke.28 HTRP is the most common pathological manifestation of hypertension. Because the retinal and cerebral arterioles share common anatomical and physiological characteristics, the presence of retinal microvascular disease suggests generalised microangiopathic processes affecting not only the eye but also other organs, such as the brain. Despite major advances in neuroimaging technologies, direct in vivo assessment of cerebral microcirculation remains challenging. However, the retina provides a unique window through which cerebral microvascular health can be assessed directly and non-invasively in vivo. Accordingly, HTRP may represent a risk factor for cerebral microvascular disease. In this regard, the association between HTRP and stroke has been investigated previously.11 23 In China, a stroke prevention trial including 9793 hypertensive participants with a median follow-up of 4.4 years demonstrated a significant positive association between HTRP and the risk of first stroke.11 In the Suita study involving 7027 participants, mild HTRP was positively associated with stroke risk (HR, 1.28; 95% CI 1.01 to 1.62).23 These findings suggest that retinal photography can be useful for cardiovascular-risk stratification in primary prevention.23 This meta-analysis found that HTRP severity is associated with a progressively increased stroke risk. Mild and severe retinopathy were found to be associated with 1.80-fold and 3.10-fold higher stroke risks than no retinopathy, respectively, and severe retinopathy was associated with a 1.88-fold higher stroke risk than mild retinopathy. Our findings are consistent with outcomes reported in previous studies.11 13 25 For instance, Chen et al have reported HRs for stroke of 1.26 (95% CI 1.01 to 1.58) for mild retinopathy versus no retinopathy and 2.40 (95% CI 1.49 to 3.84) for severe retinopathy versus no retinopathy. These results indicate that even mild retinopathy has significant predictive value for stroke in patients with hypertension. Despite adequate blood-pressure control, patients with HTRP remain at increased stroke risk.11 These findings indicate that blood-pressure monitoring alone may be insufficient for stroke prevention, and that retinal assessment may provide additional prognostic value. Ong et al have reported that individuals with moderate HTRP are more likely to experience stroke after multivariable adjustment.13 In their study, HTRP was also associated with an increased risk of cerebral infarction (mild retinopathy: HR, 1.96 (95% CI 1.09 to 3.55); moderate retinopathy: HR, 2.98 (95% CI 1.01 to 8.83)). Taken together with the present findings, these data indicate that increasing severity of HTRP corresponds to a higher stroke risk. This meta-analysis demonstrates that HTRP is significantly associated with stroke in patients with hypertension. As an important and readily accessible biomarker for stroke-risk prediction, HTRP supports the implementation of periodic retinal examinations. Notably, HTRP can be directly assessed using non-invasive retinal imaging techniques. The evaluation of HTRP may facilitate the identification of a specific subgroup of patients who would benefit from more intensive stroke monitoring. Regular retinal examinations may help identify patients at elevated stroke risk, enabling earlier implementation of targeted prevention strategies. The present findings provide the most comprehensive evidence to date regarding the association between HTRP and stroke. However, several limitations should be acknowledged. First, the number of studies included in this review was relatively small, and the geographical coverage was limited, with approximately half of the studies conducted in Asia, introducing the possibility of ethnic bias. Second, substantial heterogeneity was observed across studies, and such heterogeneity may affect the robustness of the pooled estimates. Third, publication bias was present, reflecting the tendency for studies with statistically significant results to be preferentially published. These limitations should be carefully considered when interpreting the results. In summary, HTRP is associated with an increased stroke risk in patients with hypertension and represents an important biomarker for the prediction of stroke in clinical practice. Periodic retinal examination should be recommended in patients with hypertension as part of routine stroke-risk assessment. Supplementary material Acknowledgements We would like to thank MogoEdit (https://www.mogoedit.com) for its English editing during the preparation of this manuscript. Footnotes Funding: This research was supported by the Key Research and Development Program of Shaanxi (Program No.S2025-YF-YBSF-1825) Provenance and peer review: Not commissioned; externally peer-reviewed. Patient consent for publication: Not applicable. Ethics approval: This study is exempted by the ethics committee (Honghui Hospital) because of the study design of systematic review and meta-analysis. Data availability statement Data are available upon reasonable request. References Associated Data Supplementary Materials Data Availability Statement Data are available upon reasonable request.