Key Points
Question
What is the association between objectively measured visual impairment (VI) and dementia among older US adults?
Findings
In this survey study, based on objective assessments of visual function with habitual correction in the 2021 nationally representative National Health and Aging Trends Study, all types of VI (distance VI, near VI, and contrast sensitivity impairment) were associated with a higher dementia prevalence. Having multiple VIs was more strongly associated with dementia than having a single VI.
Meaning
In this study, distance acuity, near acuity, and contrast sensitivity impairment are each associated with a higher prevalence of dementia among older US adults.
This survey study estimates the association between visual impairment and dementia in older US adults based on objective visual and cognitive function testing.
Abstract
Importance
Estimates of the association between visual impairment (VI) and dementia in the US population are based on self-reported survey data or measures of visual function that are at least 15 years old. Older adults are at high risk of VI and dementia so there is a need for up-to-date national estimates based on objective assessments.
Objective
To estimate the association between VI and dementia in older US adults based on objective visual and cognitive function testing.
Design, Setting, and Participants
This secondary analysis of the 2021 National Health and Aging Trends Study (NHATS), a population-based, nationally representative panel study, included 3817 respondents 71 years and older. Data were analyzed from January to March 2023.
Intervention
In 2021, NHATS incorporated tablet-based tests of distance and near visual acuity and contrast sensitivity (CS) with habitual correction.
Main Outcomes and Measures
VI was defined as distance visual acuity more than 0.30 logMAR, near visual acuity more than 0.30 logMAR, and CS more than 1 SD below the sample mean. Dementia was defined as scoring 1.5 SDs or more below the mean in 1 or more cognitive domains, an AD8 Dementia Screening Interview Score indicating probable dementia, or diagnosed dementia. Poisson regression estimated dementia prevalence ratios adjusted for covariates.
Results
Of 2967 included participants, 1707 (weighted percentage, 55.3%) were female, and the median (IQR) age was 76.9 (77-86) years. The weighted prevalence of dementia was 12.3% (95% CI, 10.9-13.7) and increased with near VI (21.5%; 95% CI, 17.7-25.3), distance VI (mild: 19.1%; 95% CI, 13.0-25.2; moderate, severe, or blind: 32.9%; 95% CI, 24.1- 41.8), and CS impairment (25.9%; 95% CI, 20.5-31.3). Dementia prevalence was higher among participants with near VI and CS impairment than those without (near VI prevalence ratio: 1.40; 95% CI, 1.16-1.69; CS impairment prevalence ratio: 1.31; 95% CI, 1.04-1.66) and among participants with moderate to severe distance VI or blindness (prevalence ratio: 1.72; 95% CI, 1.26-2.35) after adjustment for covariates.
Conclusions and Relevance
In this survey study, all types of objectively measured VI were associated with a higher dementia prevalence. As most VI is preventable, prioritizing vision health may be important for optimizing cognitive function.
Introduction
Visual impairment (VI) is strongly associated with dementia.1 Yet most of VI in older adults is preventable or has simply not yet been addressed with interventions like eyeglasses or cataract surgery. Cataract surgery reduces the risk of dementia,1,2 and optimizing vision may prevent up to 100 000 prevalent cases of dementia in the US.3
To our knowledge, there are no contemporary nationally representative estimates of the association between objectively measured visual function and dementia. Prior national estimates are based on data that are at least 15 years old or used self-reported vision,4,5 which may underestimate VI prevalence.6 Accordingly, we used newly available data from the National Health and Aging Trends Study (NHATS) to estimate the cross-sectional association of objectively measured VI and dementia in a nationally representative sample of adults 71 years and older.
Methods
Data are from round 11 (2021) of NHATS, a nationally representative panel study of Medicare beneficiaries 65 years and older. Data are collected annually through in-home interviews. Sample replenishment occurred in 2015, so participants were 71 years and older in 2021. Participants missing any cognitive or vision data were excluded; all others were included.
Distance and near visual acuity and contrast sensitivity (CS) were measured in participants’ homes wearing habitual correction using validated tablet-based tests (Ridgevue Publishing); participants who did not wear habitual correction underwent measurement without correction to obtain a free-living measure of visual function.6 For distance visual acuity and CS testing, the tablet was placed 59 inches from the respondent; for near visual acuity, it was held at the participant’s preferred reading distance. Distance VI was defined using World Health Organization definitions (mild VI, more than 0.30 to 0.48 logMAR; Snellen equivalent, worse than 20/40 to 20/60; moderate VI, more than 0.48 to 1.00 logMAR; Snellen equivalent, worse than 20/60 to 20/200; severe VI, more than 1.00 to 1.30 logMAR; Snellen equivalent, worse than 20/200 to 20/400; blind, more than 1.30 logMAR; Snellen equivalent, approximately worse than 20/400).7 Near VI was defined as more than 0.30 logMAR (Snellen equivalent, worse than 20/40).7 CS impairment was defined as more than 1 SD below the sample mean.8
NHATS defines participants as having dementia if they scored 1.5 SDs or more below the mean of self-respondents in 1 or more cognitive domains (memory, orientation, and executive function), had an AD8 Dementia Screening Interview Score indicating probable dementia, or had previously received a medical diagnosis of dementia.9 Self-reported covariates included age, sex, education, race and ethnicity, smoking, and number of chronic conditions (diabetes, hypertension, stroke, myocardial infarction, heart disease, lung disease, or cancer).
NHATS was approved by The Johns Hopkins University Institutional Review Board, and participants provided written informed consent. This analysis of publicly available secondary data was approved by the University of Michigan Institutional Review Board. This report adheres to the American Association for Public Opinion Research (AAPOR) reporting guidelines. Poisson regression was used to estimate dementia prevalence ratios adjusting for covariates. Analyses were weighted and accounted for the complex NHATS survey design. Analyses were conducted using SAS version 9.4 (SAS Institute) and STATA version 16 (StataCorp).
Results
There were 3817 respondents in the 2021 NHATS sample. After excluding 850 respondents missing vision or dementia data, there were 2967 respondents included in this analysis (eTable in Supplement 1). Of these, 1707 (weighted percentage, 55.3%) were female, and the median (IQR) age was 76.9 (77-86) years. The weighted prevalence of dementia was 12.3% (95% CI, 10.9-13.7) and was higher among those with near VI (21.5%; 95% CI, 17.7-25.3), distance VI (mild, 19.1%; 95% CI, 13.0-25.2; moderate, severe, or blind, 32.9%; 95% CI, 24.1- 41.8), and CS impairment (25.9%; 95% CI, 20.5-31.3) (Table 1). Dementia prevalence, adjusted for covariates, was higher among participants with near VI and CS impairment than those without (near VI prevalence ratio, 1.40; 95% CI, 1.16-1.69; CS impairment prevalence ratio, 1.31; 95% CI, 1.04-1.66) (Table 2). Adjusted dementia prevalence was also higher among participants with moderate to severe distance VI and blindness (prevalence ratio, 1.72; 95% CI, 1.26-2.35) (Table 2). Additionally, dementia prevalence was higher among participants with multiple VIs compared with those with a single type of VI (prevalence ratio, 1.35; 95% CI, 1.03-1.76) (Table 2).
Table 1. Weighted Characteristics of Participants by Visual Impairment (VI), 2021 National Health and Aging Trends Study, Round 11.
| Characteristic | Total, No. | Weighted % (95% CI) | |||||
|---|---|---|---|---|---|---|---|
| Total | Normal visiona | Near VIb | Distance VI | CS impairmentc | |||
| Mild VIb | Moderate to severe VI or blindb | ||||||
| Overall | 2967 | 100 | 72.3 (70.0-74.5) | 22.2 (20.2-24.1) | 5.6 (4.5-6.6) | 4.8 (3.9-5.7) | 9.9 (8.4-11.4) |
| Age, y | |||||||
| 71-74 | 412 | 29.6 (27.3-31.9) | 33.0 (30.1-35.9) | 21.5 (17.0-26.0) | 14.2 (6.5-21.9) | 10.3 (3.4-17.2) | 11.5 (6.1-16.8) |
| 75-79 | 887 | 32.7 (30.3-35.0) | 34.5 (31.8-37.3) | 27.2 (23.1-31.4) | 30.1 (23.9-36.2) | 24.4 (15.8-33.0) | 27.4 (20.2-34.6) |
| 80-84 | 757 | 20.1 (18.6-21.7) | 19.4 (17.8-21.0) | 21.6 (18.7-24.5) | 21.0 (14.4-27.5) | 25.6 (16.5-34.6) | 21.8 (16.4-27.1) |
| 85-89 | 547 | 11.6 (10.7-12.6) | 9.5 (8.4-10.6) | 17.3 (14.6-19.9) | 18.2 (11.4-25.0) | 19.0 (11.7-26.4) | 21.0 (16.4-25.6) |
| ≥90 | 364 | 6.0 (5.2-6.7) | 3.6 (2.9-4.2) | 12.4 (10.2-14.7) | 16.6 (11.5-21.8) | 20.7 (15.1-26.2) | 18.4 (14.0-22.8) |
| Sex | |||||||
| Female | 1707 | 55.3 (52.8-57.7) | 55.5 (52.7-58.2) | 56.2 (51.5-60.9) | 56.7 (46.4-67.0) | 54.2 (43.2-65.2) | 52.1 (44.6-59.6) |
| Male | 1260 | 44.7 (42.3-47.2) | 44.5 (41.8-47.3) | 43.8 (39.1-48.5) | 43.3 (33.0-53.6) | 45.8 (34.8-56.8) | 47.9 (40.4-55.4) |
| Race and ethnicityd | |||||||
| Non-Hispanic Black | 591 | 7.9 (6.8-9.1) | 6.4 (5.3-7.5) | 11.3 (8.4-14.1) | 14.2 (10.2-18.1) | 10.3 (6.2-14.3) | 11.2 (8.3-14.1) |
| Hispanic | 136 | 6.9 (5.0-8.9) | 5.5 (4.0-7.0) | 10.0 (5.4-14.5) | 9.8 (3.0-16.7) | 5.9 (0-12.5) | 12.2 (6.1-18.3) |
| Non-Hispanic White | 2174 | 81.8 (79.2-84.5) | 85.0 (82.6-87.4) | 74.8 (69.8-79.8) | 71.0 (62.8-79.3) | 80.3 (73.9-86.6) | 73.4 (66.2-80.5) |
| Other race | 66 | 3.3 (2.1-4.5) | 3.0 (1.8-4.3) | 3.9 (2.3-5.6) | 5.0 (0-10.4) | 3.6 (0-8.1) | 3.3 (0-6.6) |
| Education | |||||||
| Less than high school | 453 | 12.6 (10.9-14.3) | 9.3 (7.6-11.0) | 21.3 (17.5-25.1) | 17.3 (11.4-23.3) | 27.7 (19.0-36.4) | 25.4 (19.3-31.5) |
| High school graduate | 758 | 24.3 (21.8-26.9) | 24.1 (21.1-27.0) | 24.8 (20.8-28.9) | 25.9 (19.4-32.5) | 23.6 (15.4-31.9) | 26.3 (20.7-31.9) |
| Some college, no degree | 651 | 23.0 (20.8-25.1) | 23.9 (21.3-26.4) | 22.3 (18.1-26.6) | 24.5 (17.2-31.7) | 15.3 (7.6-22.9) | 18.4 (12.5-24.4) |
| College graduate or more | 1105 | 40.1 (36.5-43.6) | 42.7 (38.8-46.7) | 31.5 (26.3-36.8) | 32.3 (23.2-41.3) | 33.5 (24.1-42.8) | 29.9 (23.7-36.0) |
| Income, $ | |||||||
| <21 000 | 719 | 19.7 (17.6-21.8) | 15.5 (13.4-17.7) | 30.9 (26.4-35.4) | 23.5 (17.1-29.9) | 39.3 (30.3-48.3) | 37.3 (30.7-44.0) |
| 21 000 to <40 000 | 700 | 22.2 (19.9-24.5) | 21.4 (18.9-23.8) | 25.4 (21.3-29.5) | 29.0 (22.2-35.9) | 17.6 (11.1-24.1) | 22.8 (16.9-28.6) |
| 40 000 to <75 000 | 751 | 27.0 (24.6-29.4) | 28.7 (25.9-31.5) | 22.3 (18.6-26.0) | 26.6 (18.2-35.1) | 18.5 (11.9-25.0) | 19.5 (14.2-24.9) |
| ≥75 000 | 797 | 31.1 (27.3-34.9) | 34.4 (30.4-38.4) | 21.4 (16.3-26.6) | 20.9 (11.4-30.4) | 24.7 (15.9-33.5) | 20.3 (14.5-26.2) |
| No. of health conditionse | |||||||
| 0 | 401 | 15.9 (14.4-17.5) | 16.7 (14.8-18.6) | 12.5 (8.9-16.1) | 13.4 (5.6-21.2) | 14.7 (8.9-20.5) | 12.6 (7.6-17.6) |
| 1 | 979 | 34.7 (32.5-37.0) | 35.7 (32.9-38.5) | 33.9 (29.7-38.2) | 31.5 (24.4-38.6) | 30.6 (22.5-38.8) | 29.6 (23.8-35.4) |
| 2 | 931 | 28.9 (26.8-30.9) | 29.4 (26.9-31.9) | 25.7 (21.2-30.3) | 32.0 (23.3-40.6) | 28.9 (21.4-36.3) | 32.3 (25.7-38.8) |
| 3 | 482 | 15.3 (13.7-16.8) | 13.6 (11.9-15.3) | 20.6 (16.8-24.4) | 18.5 (11.3-25.8) | 19.0 (10.9-27.0) | 19.2 (13.6-24.9) |
| ≥4 | 174 | 5.2 (4.4-6.0) | 4.6 (3.6-5.7) | 7.2 (5.3-9.2) | 4.6 (1.5-7.6) | 6.9 (1.8-12.0) | 6.3 (3.0-9.5) |
| Ever smoked | 1444 | 49.7 (47.7-51.8) | 49.6 (47.3-51.9) | 50.6 (46.9-54.4) | 49.3 (41.5-57.2) | 49.1 (41.4-56.8) | 48.6 (42.1-55.2) |
| Dementia | 497 | 12.3 (10.9-13.7) | 9.2 (7.8-10.6) | 21.5 (17.7-25.3) | 19.1 (13.0-25.2) | 32.9 (24.1-41.8) | 25.9 (20.5-31.3) |
Abbreviation: CS, contrast sensitivity.
Normal vision was defined as no near VI, distance VI, or contrast sensitivity impairment.
Mild distance VI was defined as more than 0.30 to 0.48 logMAR (Snellen equivalent, worse than 20/40 to 20/60); moderate distance VI, more than 0.48 to 1.00 logMAR (Snellen equivalent, worse than 20/60 to 20/200); severe distance VI, more than 1.00-1.30 logMAR (Snellen equivalent, worse than 20/200 to 20/400); blind, logMAR greater than 1.30 (Snellen equivalent, approximately worse than 20/400); and near VI, more than 0.30 logMAR (Snellen equivalent, worse than 20/40).
CS impairment defined as more than 1 SD below the mean. Data are based on the distribution of contrast sensitivity in the National Health and Aging Trends Study sample.
Race and ethnicity were self-reported and were included in the analysis because of previously identified racial and ethnic differences in the prevalence of VI and dementia. The other race category includes participants who identified as American Indian or Alaska Native, Asian, Native Hawaiian or Other Pacific Islander, other specified race, or more than 1 race.
Chronic health conditions included diabetes, hypertension, stroke, myocardial infarction, heart disease, lung disease, or cancer.
Table 2. Multivariable-Adjusted Association Between Visual Impairment (VI) and Dementia, 2021 National Health and Aging Trends Study, Round 11.
| Variable | Total, No. | Weighted dementia prevalence (95% CI) | Prevalence ratio (95% CI)a | P value |
|---|---|---|---|---|
| Near VIb | ||||
| No near VI | 2178 | 9.7 (8.3-11.1) | 1 [Reference] | NA |
| Any near VI | 789 | 21.5 (17.7-25.3) | 1.40 (1.16-1.69) | .001 |
| Distance VIb | ||||
| No distance VI | 2564 | 10.8 (9.4-12.2) | 1 [Reference] | NA |
| Mild distance VI | 218 | 19.1 (13.0-25.2) | 1.17 (0.85-1.61) | .34 |
| Moderate to severe distance VI or blind | 185 | 32.9 (24.1-41.8) | 1.72 (1.26-2.35) | .001 |
| CS impairmentc | ||||
| No CS impairment | 2575 | 10.8 (9.4-12.2) | 1 [Reference] | NA |
| CS impairment | 392 | 25.9 (20.5-31.3) | 1.31 (1.04-1.66) | .02 |
| Single VI vs multiple VI | ||||
| No VI | 1974 | 9.2 (7.8-10.6) | 1 [Reference] | NA |
| Single vs no VI | 586 | 16.1 (12.4-19.9) | 1.20 (0.95-1.51) | .13 |
| Multiple vs no VI | 407 | 27.9 (22.7-33.0) | 1.61 (1.31-1.98) | <.001 |
| Multiple vs single VI | NA | NA | 1.35 (1.03-1.76) | .03 |
Abbreviations: CS, contrast sensitivity; NA, not applicable.
Models adjusted for age, sex, race and ethnicity, education, income, smoking history, and number of chronic health conditions.
Mild distance VI was defined as more than 0.30 to 0.48 logMAR (Snellen equivalent, worse than 20/40 to 20/60); moderate distance VI, more than 0.48 to 1.00 logMAR (Snellen equivalent, worse than 20/60 to 20/200); severe distance VI, more than 1.00-1.30 logMAR (Snellen equivalent, worse than 20/200 to 20/400); blind, logMAR greater than 1.30 (Snellen equivalent, approximately worse than 20/400); and near VI, more than 0.30 logMAR (Snellen equivalent, worse than 20/40).
CS impairment defined as more than 1 SD below the mean. Data are based on the distribution of contrast sensitivity in the National Health and Aging Trends Study sample.
Discussion
In a nationally representative sample of older US adults, all types of objectively measured VI were associated with a higher dementia prevalence, and multiple VIs were more strongly associated with dementia than having a single VI. Prior studies investigating the association between visual function and dementia are based on data that is now at least 15 years old or used self-reported vision, which may underestimate the prevalence of VI.6
A cross-sectional analysis of the 1999 to 2002 National Health and Nutrition Examination Survey found that objectively measured distance VI was associated with worse cognitive function.5 Studies using self-reported visual function data from the 2011 to 2018 rounds of NHATS found that functional self-reported visual difficulty was associated with a higher cross-sectional hazard of dementia and that participants with baseline visual difficulty had a higher likelihood of developing dementia over time.4,10 The present study expands on prior work by investigating the association between objectively measured VI and dementia in a contemporary, nationally representative sample and by presenting data on the association between CS and dementia.
Most VI is preventable or has yet to be addressed.1 Near VI can often be treated with reading glasses, which are inexpensive and widely available over the counter in the US. A large proportion of distance VI is due to uncorrected refractive error or cataracts, which can be treated with glasses and cataract surgery, respectively.1 Cataract surgery is associated with lower risk of dementia development, suggesting that interventions to improve vision in older adults may help to maximize cognitive function.2
Strengths and Limitations
Strengths of this study include, to our knowledge, the first objectively measured and nationally representative population data on visual function and dementia among older US adults since 2008. Objective visual function tests, such as those used in this study, are the criterion standard for estimating VI, and measuring visual function with habitual correction captures a free-living measure of visual function corresponding to participants’ day-to-day experience. Limitations include the cross-sectional design, inclusion of only participants 71 years and older, the potential for unmeasured confounding, and lack of data on causes of VI. It is also possible that those with dementia may have had greater difficulty taking vision tests. The median age was greater among respondents who were excluded due to missing data; since VI, dementia, and the competing risk of mortality are all associated with older age, the true association between VI and dementia may be even stronger.
Conclusions
In this study, all types of objectively measured VI were associated with a higher dementia prevalence. Thus, prioritizing vision health may be key to optimizing both sight and overall health and well-being. Randomized trials are warranted to determine whether optimizing vision is a viable strategy to slow cognitive decline and reduce dementia risk.
eTable. Comparison Between Respondents Included and Excluded Due to Missing Data
Data Sharing Statement
References
- 1.Burton MJ, Ramke J, Marques AP, et al. The Lancet Global Health Commission on Global Eye Health: vision beyond 2020. Lancet Glob Health. 2021;9(4):e489-e551. doi: 10.1016/S2214-109X(20)30488-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Lee CS, Gibbons LE, Lee AY, et al. Association between cataract extraction and development of dementia. JAMA Intern Med. 2022;182(2):134-141. doi: 10.1001/jamainternmed.2021.6990 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Ehrlich JR, Goldstein J, Swenor BK, Whitson H, Langa KM, Veliz P. Addition of vision impairment to a life-course model of potentially modifiable dementia risk factors in the US. JAMA Neurol. 2022;79(6):623-626. doi: 10.1001/jamaneurol.2022.0723 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Chen SP, Azad AD, Pershing S. Bidirectional association between visual impairment and dementia among older adults in the United States over time. Ophthalmology. 2021;128(9):1276-1283. doi: 10.1016/j.ophtha.2021.02.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Chen SP, Bhattacharya J, Pershing S. Association of vision loss with cognition in older adults. JAMA Ophthalmol. 2017;135(9):963-970. doi: 10.1001/jamaophthalmol.2017.2838 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Hu M, Freedman VA, Ehrlich JR, Reed NS, Billington C, Kasper JD. Collecting objective measures of visual and auditory function in a national in-home survey of older adults. J Surv Stat Methodol. 2021;9(2):309-334. doi: 10.1093/jssam/smaa044 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.ICD-11 for Mortality and Morbidity Statistics . 9D90 Vision impairment including blindness. Accessed August 1, 2022. https://icd.who.int/browse11/l-m/en#/http%3a%2f%2fid.who.int%2ficd%2fentity%2f1103667651
- 8.Killeen OJ, De Lott LB, Zhou Y, et al. Population prevalence of vision impairment in US adults 71 years and older: the National Health and Aging Trends Study. JAMA Ophthalmol. 2023;141(2):197-204. doi: 10.1001/jamaophthalmol.2022.5840 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Kasper JD, Freedman VA, Spillman BC. Classification of persons by dementia status in the National Health and Aging Trends Study. Accessed June 7, 2023. https://www.nhats.org/sites/default/files/inline-files/DementiaTechnicalPaperJuly_2_4_2013_10_23_15.pdf
- 10.Kuo PL, Huang AR, Ehrlich JR, et al. Prevalence of concurrent functional vision and hearing impairment and association with dementia in community-dwelling Medicare beneficiaries. JAMA Netw Open. 2021;4(3):e211558. doi: 10.1001/jamanetworkopen.2021.1558 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable. Comparison Between Respondents Included and Excluded Due to Missing Data
Data Sharing Statement