The suggestive evidence should encourage research into the associations to evaluate the association with stronger evidence.
A new study published in Ageing Research Reviews1 found an association between eye disease and dementia, along with dementia subtypes. The association, found in studies with weaker evidence, should encourage future studies into the area.
Cognitive impairment and dementia affect millions of people throughout the world, with this number projected to triple by 2050 as the population ages.2 Higher mortality rates and shorter life expectancy are both effects of cognitive impairment, which makes understanding the risk factors as soon as possible a necessity. Poor vision is one such risk factor, as it can lead to such adverse effects as social isolation, reduced physical activity, and depression. Visual impairment (VI) has been previously linked to cognitive function, but the link to dementia has been less clear. This meta-analysis aimed to collect data from previous studies to assess the relationships of VI and eye diseases with dementia and cognitive impairment.1
This review and meta-analysis aimed to collect data on outcomes that were associated with VI and eye disease and their effect on cognitive impairment and dementia. PubMed, MEDLINE (Ovid), EMBASE, Web of Science, Cochrane CENTRAL, and CDSR were searched for studies on the topic, with studies published through May 30, 2024, eligible for inclusion. Systematic reviews without a meta-analysis and individual studies were excluded.
Author, year of publication, number of studies included in the meta-analysis, design of the studies, vision assessments, VI characteristics, cognitive assessments, and the outcomes were extracted from the included studies. All associations were calculated by extracting the effect sizes of individual studies. Studies were classed as either convincing, highly suggestive, suggestive, weak, or no association based on the strength of the association between dementia and VI.
There were 13 publications included in this study, all of which were published between 2014 and 2024 and contained both cross-sectional and longitudinal designs. The quality of the studies was a mix between high (n = 1), moderate (n = 10), and low (n = 2) based on AMSTAR 2 measures. Associations reported included those between VI and cognitive impairment through different methods of assessing VI, including objective measures, cross-sectional studies, subjective measures, and longitudinal studies. Associations evaluated also included VI and dementia, age-related macular degeneration (AMD) and dementia, glaucoma and dementia, cataracts and dementia, and diabetic retinopathy and dementia.
There were 14 outcomes identified, of which 57.1% had unique significant events between VI and eye diseases and cognitive impairment. Small study effects were found in 57.1% of the studies and excess significance bias was found in 21.5%. None of the studies had convincing or highly suggestive levels of evidence, whereas 21.5% had suggestive evidence and 35.7% had weak evidence.
The pooled analysis found that there was suggestive evidence for the association between cataracts and dementia (equivalent OR [eOR], 1.20; 95% CI, 1.16-1.25). The association between diabetic retinopathy and dementia had only weak evidence (eOR, 1.33; 95% CI, 1.17-1.50), similar to the weak evidence surrounding the association between cataracts and vascular dementia (eOR, 1.26; 95% CI, 1.09-1.45). Cataracts and Alzheimer disease had a suggestive association (eOR, 1.21; 95% CI, 0.15-1.28). A suggestive association was also found between AMD and Alzheimer disease (eOR, 1.27; 95% CI, 1.27-1.27).
VI identified by objective measures (eOR, 1.56; 95% CI, 1.12-2.18) and VI identified by cross-sectional design (eOR, 2.37; 95% CI, 2.31-2.44) both had associations with cognitive impairment but were supported by weak evidence. The association between VI and dementia (eOR, 1.50; 95% CI, 1.23-1.84) was also supported by weak evidence. Statistical significance was not found in any other association.
There were some limitations to this study. Age-related response bias is possible in any of the studies. There was a lack of a clear definition for VI, which could affect the results. There was a lack of reporting on coexisting near VI. Definition criteria for AMD, cataract, glaucoma, and diabetic retinopathy were not clear. There was heterogeneity in the study population, which indicates that the patient population could have different life expectancy and income levels. Associations may also not indicate causality and could be due to common pathologies.
The researchers concluded that associations between VI and dementia exist but would need stronger evidence to confirm their association overall. Future research should focus on using stronger methodological approaches to corroborate the associations found in studies with weaker evidence.
References
Atsena Therapeutics and Nippon Shinyaku Enter Exclusive License Agreement for ATSN-101
November 13th 2024Atsena Therapeutics and Nippon Shinyaku have formed an exclusive licensing agreement for ATSN-101, a gene therapy for Leber congenital amaurosis (LCA1). Nippon Shinyaku will commercialize ATSN-101 in the US and Japan, while Atsena retains global rights outside these territories.
Read More
Economic Burden in Late-Stage AMD High in Patients and Caregivers in US
November 6th 2024Reduced well-being and loss of productivity were contributors to the high economic burden in patients with late-stage age-related macular degeneration (AMD) living in the US, Germany, and Bulgaria.
Read More