The proportion of older individuals is rising rapidly, particularly in developed countries. Globally, ~2.1 billion people are estimated to be aged 60 and older in 2050 (United Nations Department of Economic Social Affairs Population Division, 2017). Aging causes a variety of changes to the brain and eye, making an older person more prone to Alzheimer's disease (AD) and visual impairment. Both AD (Wood et al., 2016) and eye diseases (Knudtson et al., 2005) decrease quality of life and worsen active aging. The major age-related eye diseases, including age-related macular degeneration (AMD) and diabetic retinopathy (DR), are the leading causes of visual impairment and blindness worldwide (Flaxman et al., 2017). In the light of the rising incidence of both AD (GBD 2016 Dementia Collaborators, 2019, 2021) and age-related eye diseases (Tham et al., 2014; Wong et al., 2014; Teo et al., 2021), knowledge on the pattern of age-related eye diseases in individuals with AD and cognitive impairment, no dementia (CIND) (Jacova et al., 2008) is critical and useful in the context of screening, early detection, and management of such ophthalmic conditions in cognitively impaired individuals.

In recent years, several studies have investigated the influence of neurological diagnosis on AMD but with inconclusive results (Baker et al., 2009; Keenan et al., 2014; Nolan et al., 2014; Williams et al., 2014; Frost et al., 2016). While Frost et al. found an increased prevalence of early AMD in patients with AD compared to cognitively normal controls (Frost et al., 2016), other researchers reported no association between AD and AMD after adjusting for potential confounders (Baker et al., 2009; Nolan et al., 2014; Williams et al., 2014). Contrary to previous studies which were cross-sectional in study design, a longitudinal study by Keenan et al. (2014) found that dementia and AD were associated with decreased incidence of AMD. However, their study was derived from patients seeking tertiary eye care services, which may represent the more severe spectrum of the disease, admission to hospital for AMD treatments, i.e., intravitreal antivascular endothelial growth factor (VEGF) therapy. Overall, past studies focused overwhelmingly on western populations, reflecting a paucity of data on other races/ethnicities. While several studies have examined the impact of having DR on cognitive impairment (Cheng et al., 2021), no studies have examined the impact of having cognitive impairment on DR. Finally, very limited data are available on the impact of CIND on eye diseases, a condition with established risk of progression to AD and other forms of dementia (Petersen, 2016).

In this study, we examined the associations of neurological diagnosis with age-related eye diseases, namely, AMD and DR, in an Asian setting. We hypothesized that prevalence of eye diseases will be higher in persons with cognitive impairment than NCI.

A total of 656 participants were recruited between September 2009 to September 2020, 592 (90.2%) of whom had gradable fundus photographs. A comparison of characteristics between persons with and without gradable fundus photographs is shown in Supplementary Table 1. Persons who had ungradable photographs were more likely to be older and have AD and diabetes (p ≤ 0.034). A total of 124 NCI participants, 252 CIND participants, and 216 AD participants with gradable fundus images were available for analysis.

Table 1 presents the characteristics of the participants with gradable fundus photographs by their neurological diagnosis. The mean ± standard deviation (SD) of age was 68.3 ± 7.7 years for NCI, 72.8 ± 7.7 years for CIND, and 74.7 ± 7.4 years for AD. AD participants were likely to be older and of Malay ethnicity, had lower educational level, more likely to be divorced and widowed, more likely to be living with children, relatives, or friends, higher prevalence of hypertension and diabetes, and higher systolic blood pressure (p ≤ 0.007). There was no significant difference in sex, smoking status, hyperlipidemia, BMI, and diastolic blood pressure among the groups.

Of the 592 participants with gradable photographs, 396 (66.9%) had at least one eye disease, with AMD being the most common (60.1%; n = 356), followed by DR (8.4%; n = 50). Table 2 shows the associations of neurological diagnosis with any eye diseases, AMD, and DR. Only AD participants were more likely to have higher odds of having DR (OR = 2.95, 95% CI = 1.15–7.60) in the multivariable model. However, this association was no longer significant after further adjusting for the presence of diabetes (OR = 1.83, 95% CI = 0.66–5.04). We next analyzed the association of DR in AD in persons with vascular dementia and report a stronger association between DR and AD (OR = 5.22, 95%CI = 1.74–15.67; Table 2). Neurological diagnosis was not significantly associated with an increased odds of having any AMD (OR = 0.75, 95%CI = 0.45–1.24; Table 2). Figure 2 further demonstrates that the presence of AMD was similar across the neurological groups whereas DR status was significantly higher in AD participants. We performed a separate analysis and report that both CIND without MCI (OR = 1.07, 95% CI = 0.29–3.97) and CIND with MCI (OR = 0.99, 95% CI = 0.35–2.82) were not independently associated with having DR (Supplementary Table 2). We also performed a subgroup analysis and report that both CIND and AD were not independently associated with either early AMD (OR = 0.83, 95% CI = 0.44–1.58) or intermediate AMD or worse (OR = 0.68, 95% CI = 0.38–1.22; Supplementary Table 3).

In a multiethnic Asian population, our data showed that the odds are three times higher that an patient with AD will get moderate DR or worse compared to a participant with NCI whereas neurological diagnosis was not associated with AMD, after controlling for age, sex, race, educational level, and marital status. There is less available data comparing the odds of DR between cognitive groups. Further statistical adjustment for the presence of diabetes demonstrates that the association between DR and AD is due to higher numbers of diabetic individuals in the AD cohort compared with the CIND and NCI cohorts. A meta-analysis pooled a total of 17 studies involving 1,746,777 individuals reported that participants with diabetes had significant higher incidence of AD than those without diabetes (Zhang et al., 2017).

To our knowledge, this is the first study investigating the risk of DR in subjects with clinically diagnosed CIND or AD (Table 3). Previous studies investigating the association between DR and cognitive function all focused on DR as the independent factor, i.e., as a potential marker for present or incident CIND/AD. To this end, they found that patients with DR had increased prevalence (Ong et al., 2012; Liao et al., 2017) and incidence (Gupta et al., 2019) of CIND, as well as increased incidence of AD (Chen et al., 2016; Lee et al., 2019; Pedersen et al., 2022). A recent systematic review and meta-analysis support these results by concluding that having moderate DR or worse was strongly associated with cognitive impairment (Cheng et al., 2021). Considered in tandem with the findings of our study, this suggests that the association between DR and cognitive impairment may be bidirectional. Although there is consistent evidence of shared risk factors between retinal and cerebral neurodegeneration (Jindal, 2015; Ong et al., 2015), including microvasculature-specific associations (Chua et al., 2020), the contribution of AD in the pathogenesis of DR has not been elucidated. In particular, it has been suggested that vascular lesions in eyes with DR may mirror similar pathological processes in the cerebral microcirculation (Cheung and Wong, 2008). Our finding of increased odds of DR in vascular dementia supports this proposition. Future longitudinal studies will need to be performed to determine the risk of DR in pre-diabetic AD and CIND cases.

There is currently no published guideline on the prevalence, management, and outcomes of DR in patients with AD, although there is a national diabetic retinal photograph screening for individuals with diabetes in Singapore (Goh et al., 2014). The reasons underpinning higher odds of DR in patients with AD are multifactorial. Apart from the higher prevalence of diabetes as pointed out earlier, individuals with AD may be less likely to attend annual DR screening program since they might be more reliant on caregivers to attend primary eyecare appointments and less likely to notice or report visual symptoms. Previous research reported that a high proportion of individuals with dementia are living with undiagnosed age-related eye diseases (Wong et al., 2015). A persistent lack of referral of moderate DR or worse can lead to worsening of retinopathy, visual impairment, or blindness, whereas timely delivery of sight-saving treatments can decrease the risk of blindness (Ting et al., 2017).

The finding that AD was not associated with AMD is in concordance with most studies on older White populations (Table 3) (Baker et al., 2009; Nolan et al., 2014; Williams et al., 2014). While an Australian study found increased prevalence of early AMD in subjects with AD (Frost et al., 2016), they did not find a significant association of AD diagnosis with intermediate and advanced AMD. It most likely indicates that AMD is associated with higher risks of cognitive dysfunction or AD, as consistently demonstrated in cross-sectional (Wong et al., 2002; Seden et al., 2015) and cohort (Klaver et al., 1999; Tsai et al., 2015; Chen et al., 2016; Lee et al., 2019; Choi et al., 2020) studies, but not vice versa. Another potential reason for the lack of association may have been consequent to the limited sample size and thus the lack of power in statistical analysis; to date, the odds of AMD in persons CIND or AD have not been investigated in a large-scale cohort study.

Notably, high prevalence of AMD (59.3%) was observed in adults aged 50 and older in this study. A previous population-based cross-sectional study of a multiethnic Asian cohort residing in Singapore found the prevalence of AMD (early and late AMD combined) in persons aged 80 and above to be 26.3% (Cheung et al., 2014). This difference may be accounted for partly by the different grading criteria used. The aforementioned population-based study used the modification Wisconsin Age-Related Maculopathy Grading System, which defines AMD as either soft indistinct or reticular drusen or both soft, distinct drusen plus retinal pigment epithelium abnormalities which is stricter than our study. Moreover, the AMD prevalence in our study of 59.3% is comparable to the 41.5% reported by Williams et al. (2014), similarly a clinic-based study in which disease prevalence would tend to be higher than population-based studies.

This cross-sectional study of older adults from a multiethnic Asian population demonstrated that having AD was associated with DR, for which there was limited prior data. These data suggest that AD with diabetes should be followed up more actively by their physicians, with interventions as appropriate, to prevent visual impairments. Potential barriers to care for these vulnerable individuals need to be investigated and addressed in future studies.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by National Healthcare Group Domain-Specific Review Board protocol number R1500/83/2017. The patients/participants provided their written informed consent to participate in this study.

JC, ZZ, CChen, and LS conceived and designed the study. JC, ZZ, DW, BiT, BK, CS, SH, NV, BoT, CCheu, GG, AP-C, TW, CChen, and LS analyzed and interpreted the data. JC, ZZ, and LS wrote the main manuscript text. All authors reviewed the manuscript.

This study was supported by the National Medical Research Council (CG/C010A/2017_SERI, OFIRG/0048/2017, OFLCG/004c/2018, TA/MOH-000249-00/2018, MOH-OFIRG20nov-0014, and NMRC/CG2/004b/2022-SERI), National Research Foundation Singapore (NRF2019-THE002-0006 and NRF-CRP24-2020-0001), A^*STAR (A20H4b0141), the Singapore Eye Research Institute and Nanyang Technological University [SERI-NTU Advanced Ocular Engineering (STANCE) Program], the Duke-NUS Medical School [Duke-NUS-KP(Coll)/2018/0009A], and the SERI-Lee Foundation (LF1019-1) Singapore. 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.