Retinopathy commonly refers to a spectrum of signs on the fundus (e.g. microaneurysms, soft/hard exudates, and/or retinal hemorrhage) that are common in the elderly, even in those without diabetes (1). A recent pooled analysis of 22,896 individuals with diabetes reported that nearly 35% had retinopathy (2). Furthermore, previous evidence suggested that up to 15% of non-diabetic patients exhibited clinical signs of retinopathy (3–7).

Previous literature has consistently described significant associations between retinopathy and systemic vascular comorbidities, including diabetes mellitus (DM) (5), high blood pressure (HBP) (8), chronic kidney disease (CKD) (6, 9–11) and cardiovascular disease (CVD) (12–15). The presence of retinopathy has been reported to be an independent risk factor for mortality (16–27). A recent meta-analysis of 20 observational studies reported that individuals with retinopathy had a two- to four-fold increase in their risks of all-cause mortality independent of other potential risk factors (28).

To date, only a few studies have investigated the joint effects of retinopathy and systemic vascular comorbidities [such as DM (16), CKD (16, 29–31), and clinical cerebrovascular disease (16)] on mortality. These previous studies were subject to selection bias, underestimation of retinopathy, short duration of follow-up and a lack of investigation of the effects of concomitant macrovascular disorders. Given that the retina is readily viewable via non-invasive retinal imaging and that there is a well-established association between retinopathy and systemic vascular comorbidities, clarifying the precise effects of the co-occurrence of retinopathy and systemic vascular comorbidities on the risk of mortality is of great significance.

The National Health and Nutrition Examination Survey (NHANES) is a continuous, national population-based study of elderly participants in the U.S., which provides a unique opportunity to investigate the impact of retinopathy and systemic vascular comorbidities on mortality.

In total, there were 6,797 individuals aged 40 years and older from the NHANES 2005-2008. Of these participants, 5,703 were included in the current analysis due to missing information on the grading of retinopathy for both eyes in 1,093 participants and missing data on vital status in 1 participant. Participants with missing data were more likely to be older, black, poorly educated, unmarried or other, have a lower level of household income and be unhealthier in terms of systemic comorbidities when compared to those with complete data. The characteristics of the excluded and the included participants are shown in Supplement Table 1 .

A total of 710 participants were identified as having retinopathy (weighted prevalence: 9.75%). The mean age of these participants was 56.5 ± 0.38 years (SE), and 52.6% were women. The baseline characteristics of participants by retinopathy status are presented in Table 1 . Participants identified as having retinopathy were more likely to be older, male, black, poorly educated, and lifetime abstainers/former drinkers and to have higher BMI, proportion of walking disability, poor/fair self-rated health status, and concomitant micro- and macrovascular pathologies (DM, HBP, CKD and CVD) compared to those without retinopathy. No significant difference in other characteristics was observed between these two groups.

The median duration of follow-up was 8.33 years (interquartile range [IQR]: 7.50-9.67 years). There were 949 (11.8%) deaths from all causes. Participants with retinopathy had higher all-cause mortality than those without (23.0% versus 10.5%, t-test P < 0.001). The baseline characteristics, including demographic factors, health-related behaviors and general health characteristics for participants by survival status, are presented in Table 2 . Several covariates, including age, sex, ethnicity, educational attainment, types of marital status, household income level, smoking status, alcohol consumption, the level of CRP, self-rated health status, walking disability, and micro- and macrovascular disorders (DM, CKD and history of CVD), were strongly associated with all-cause mortality in age- and sex-adjusted models. Multivariate Cox regression models suggested that the presence of any degree of retinopathy at baseline was associated with higher all-cause mortality (HR, 1.41; 95% CI, 1.08-1.83; P=0.012).

Interactions between retinopathy and each systemic vascular disorder were assessed, and no significant interaction for DM, CKD, or CVD was found. There was a significant interaction between retinopathy and HBP in the model (P=0.041). Kaplan-Meier curves for all-cause mortality according to retinopathy and systemic vascular conditions are shown in Figure 1 . Table 3 presents the synergistic impact of each vascular disorder together with retinopathy on mortality. All-cause mortality for participants with retinopathy and comorbidities such as DM (HR, 1.72; 95% CI, 1.21-2.43; P=0.003), HBP (HR, 1.47; 95% CI, 1.03-2.10; P=0.034), CKD (HR, 1.73; 95% CI, 1.26-2.39; P=0.002) and CVD (HR, 1.92; 95% CI, 1.21-3.04; P=0.007) was significantly higher than for those without any condition.

The results stratified by diabetes and hypertension status are shown in Supplement Table 2 . Among non-diabetic participants, all-cause mortality was higher in those with retinopathy and CKD (HR, 1.55; 95% CI, 1.07-2.25; P=0.023). Among diabetic participants, the co-occurrence of retinopathy and CVD (HR, 2.52; 95% CI, 1.20-5.29; P=0.016) further increased all-cause mortality. For participants without hypertension, the co-occurrence of retinopathy and CVD posed a higher risk of death (HR, 3.32; 95% CI, 1.32-8.35; P=0.012). In an analysis limited to participants with hypertension, the joint effect of retinopathy and CVD (HR, 1.64; 95% CI, 1.01-2.67; P=0.047) or CKD (HR, 1.85; 95% CI, 1.35-2.55; P<0.001) predicted higher all-cause mortality than that of those without either condition.

In this large sample of middle-aged and older adults, participants with retinopathy had a higher all-cause mortality rate. Moreover, the co-occurrence of retinopathy and systemic vascular conditions (DM, HBP, CKD, and history of CVD) further increased all-cause mortality.

Our findings are in line with previous results that reported a significant association between retinopathy and mortality (17–26). Contrary to the EURODIAB Prospective Complications Study, which found that the relationship between retinopathy and mortality can be largely explained by the presence of coexisting cardiovascular risk factors (40), we report that retinopathy was an independent risk factor for mortality after comprehensive adjustment of covariates. We postulate that the observed difference might be a result of missing fundus photographs (31%) and morbidity data (14%) in the EURODIAB Prospective Complications Study, which may have led to the underestimation of the association between retinopathy and mortality.

Only a limited number of studies have attempted to explore the joint effects of retinopathy and systemic vascular comorbidities [such as DM (16), CKD (16, 29–31), and clinical cerebrovascular disease (16)] on mortality. A more detailed investigation of systemic vascular comorbidities (DM, HBP, CKD, and CVD) in our study provides further insights into the association between retinopathy and mortality. That is, our results not only support previous findings of a higher mortality risk in individuals with retinopathy but also suggest that retinopathy in the presence of concomitant DM, HBP, CKD or CVD increases the risk of mortality even further.

Mechanisms underlying the association between retinopathy and mortality remain unknown. It has been proposed that retinopathy may be an indicator of the burden of CVD risk factors, including dyslipidemia, obesity, hypertension, and smoking (17). However, our results do not support this theory. Although we found that the distribution of CVD risk factors differed between non-retinopathy and retinopathy groups, retinopathy was found to be an independent risk factor for mortality in the fully adjusted model. Another potential explanation is that retinopathy, as a marker of abnormal microcirculation in the retina, may be an important indicator of systemic micro- and macro-vascular abnormalities (24). These micro- and macrovascular disorders may share similar pathophysiological processes with retinopathy, including endothelial dysfunction, inflammation, apoptosis, and neovascularization (41, 42). In addition, the co-occurrence of retinopathy and systemic vascular disorders may reflect a greater burden and/or severity of vascular pathologies, which may explain the further increase in mortality among the participants who have both retinopathy and systemic vascular disorders, and the marginally interaction between retinopathy and HBP.

Given that retinopathy can be readily detected via non-invasive means, our findings may have several practical implications. Firstly, our finding of a statistically significant association between retinopathy and systemic vascular abnormalities indicates that individuals with retinopathy may benefit from a comprehensive vascular assessment and should be closely monitored. Secondly, the current study and previous studies find that the co-occurrence of retinopathy and micro- or macrovascular disorders poses a further increase in all-cause mortality. Therefore, intensive vascular risk reduction may be warranted in the management of these patients.

The strengths of this analysis include the use of a large-scale sample with national representativeness, the use of a standardized objective grading protocol to assess retinopathy, access to death records, long-term follow-up and inclusion of a comprehensive range of covariates. However, several potential limitations should be considered. Firstly, retinopathy status and potential confounders adjusted for in our analysis were assessed on a single occasion. During the follow-up period, the behaviors and/or retinopathy status of the patients may have changed, which may have a direct impact on the outcome. Secondly, discrepancies may exist between the self-reported data and clinically measured data. This may cause underreporting of milder cases and subsequently lead to a bias in the analysis. Finally, participants with missing data tended to be in poorer health status, which may have resulted in selection bias.

In summary, our findings suggest that middle-aged and elderly people with retinopathy have increased all-cause mortality. Furthermore, the joint effects of retinopathy and major systemic vascular comorbidities increase the all-cause mortality further. Our results indicate that more extensive risk factor assessment and management of individuals with retinopathy may be beneficial to reduce their mortality rate, especially in patients who have both retinopathy and systemic vascular disorders.

Publicly available datasets were analyzed in this study. This data can be found here: https://www.cdc.gov/nchs/nhanes/index.htm.

Study concept and design: ZZ, MH, and XY. Acquisition, analysis, or interpretation: All authors. Drafting of the manuscript: ZZ and XS. Critical revision of the manuscript for important intellectual content: WW, JH, YC, MH, and XY. Statistical analysis: ZZ, XS, and WW. Obtained funding: MH. Administrative, technical, or material support: ZZ, XS, WW, MH, and XY. Study supervision: MH and XY. All authors contributed to the article and approved the submitted version.

The present work was supported by Fundamental Research Funds of the State Key Laboratory of Ophthalmology (303060202400362), National Natural Science Foundation of China (82000901, 82101173), Project of Special Research on Cardiovascular Diseases (2020XXG007), and Research Foundation of Medical Science and Technology of Guangdong Province (B2021237). MH receives support from the University of Melbourne at Research Accelerator Program and the CERA Foundation. The Centre for Eye Research Australia receives Operational Infrastructure Support from the Victorian State Government. The sponsor or funding organization had no role in the design or conduct of this research.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.