Abstract
Background
Vision loss may be a risk factor for hallucinations, but this has not been studied at the population level.
Methods
To determine the association between self-reported vision loss and hallucinations in a large community-based sample of older adults, we performed a cross-sectional and longitudinal analysis of two large, nationally representative US health surveys: the National Health and Aging Trends Study (NHATS) and the Health and Retirement Study (HRS). Visual impairment and hallucinations were self- or proxy-reported. Multivariate single and mixed effects logistic regression models were built to examine whether visual impairment and history of cataract surgery were associated with hallucinations.
Results
In NHATS (n = 1520), hallucinations were more prevalent in those who reported difficulty reading newspaper print (OR 1.77, 95% confidence interval (CI): 1.32–2.39) or recognising someone across the street (OR 2.48, 95% CI: 1.86–3.31) after adjusting for confounders. In HRS (n = 3682), a similar association was observed for overall (OR 1.32, 95% CI: 1.08–1.60), distance (OR 1.61, 95% CI: 1.32–1.96) and near eyesight difficulties (OR 1.52, 95% CI: 1.25–1.85). In neither sample was there a significant association between cataract surgery and hallucinations after adjusting for covariates.
Conclusions
Visual dysfunction is associated with increased odds of hallucinations in the older US adult population. This suggests that the prevention and treatment of vision loss may potentially reduce the prevalence of hallucinations in older adults.
Keywords: blindness, vision loss, hallucinations, older people
Key points
Hallucinations are common in older adults with dementia.
Visual dysfunction is associated with an increased odds of hallucinations after adjusting for confounders.
The prevention and treatment of vision loss may also reduce the prevalence of hallucinations in the older adult population.
Introduction
Hallucinations are common in the older adult population, especially among those with neurodegenerative disorders, such as Alzheimer disease [1] or Parkinson disease [2]. Hallucinations are not only distressing for patients and their caregivers, but they are also a risk factor for negative health outcomes, such as falls, hip fracture and death [3,4]. The neurobiological underpinnings of visual hallucinations are incompletely understood, but a commonly accepted hypothesis is that visual hallucinations represent an imbalance between internally and externally generated percepts. In neurologic or psychiatric disease, an increase in internally generated images, through a number of mechanisms, results in visual hallucinations. However, a decrease in external stimulation due to vision loss may also cause perceptual imbalance and lead to hallucinations. In both Parkinson disease [5–8] and Alzheimer disease [9], ocular pathology and reduced visual acuity have been identified as risk factors for hallucinations in several small academic centre cohort studies. If this was true at the overall population level, it is possible that public health interventions to reduce the burden of vision loss might also prevent hallucinations and their associated consequences. However, cross-sectional population-based studies are needed to confirm the association between vision loss and hallucinations before future prospective studies can address the question of directionality and causal inference.
The objective of this study was to determine the cross-sectional and longitudinal association between visual impairment and hallucinations in two large, population-based longitudinal health surveys. We hypothesised that self-reported visual impairment would be associated with hallucinations after adjusting for confounders and that a history of cataract surgery would be associated with a decreased risk of hallucinations.
Methods
Study population
We analysed data from two longitudinal health surveys of the older US adult population: NHATS and HRS. As this was an analysis of fully de-identified publicly available data, the study was approved for exemption by the University of Pennsylvania Institutional Review Board. The National Health and Aging Trends Study (NHATS) is a nationally representative sample of Medicare beneficiaries aged 65 or older who have been surveyed annually since 2011 with replenishment of the sample in 2015 [10]. The Health and Retirement Study (HRS) is a large, nationally representative longitudinal survey of US adults over the age of 50 [11]. It is the result of two surveys from 1992 to 1996, HRS and Asset and Health Dynamics among the Oldest Old (AHEAD), which were merged in 1998. Participants have been re-surveyed biennially with replenishment in 1998, 2004, 2010 and 2016. For this study, we used data from all seven waves of NHATS and from seven waves of HRS (2002–2014). We limited HRS data to subjects 65 years of age or older to match the NHATS population.
Primary outcomes
The primary outcomes were self-reported visual function and proxy-reported hallucinations. This combination of self-reported and proxy-reported data reflects the questionnaire design of these health surveys. Both surveys included questions about self-reported visual function at each visit (Supplementary Material A1) and an objective cognitive assessment. NHATS assessed distance and near vision, the use of magnifiers and blindness using yes-or-no questions. HRS assessed overall eyesight, distance vision and near vision using 1–5 scales, and we defined a score of 4 or 5 (fair or poor) as having impaired vision, consistent with prior studies [12]. Legal blindness was recorded as a score of 6 for overall eyesight. All subjects answered the questions about vision. However, if a subject was unwilling or unable to complete the cognitive assessment, questions pertaining to his or her cognitive function were directed to the subject’s proxy (spouse, child or caregiver) when available. Of 12,427 NHATS and 37,498 age-eligible HRS participants, 1520 (12.2%) and 3682 (9.8%) completed proxy cognitive questionnaires. The proxy cognitive questionnaire included a question about hallucinations: ‘Does he or she ever see or hear things that are not there?’ Hallucinations were the primary outcome of interest, and thus only subjects who completed a proxy cognitive questionnaire were included in the analysis. Subjects were designated as having hallucinations if the proxy ever reported observing hallucinations at any study visit, and subjects whose proxy denied any history of hallucinations at every study visit were defined as not having hallucinations.
Covariates
NHATS and HRS include information on age, race and ethnicity, gender, smoking, level of education, annual household income, place of residence (e.g. nursing home), physical functional impairment and self-reported diagnoses of hypertension, diabetes, stroke, cataract surgery and hearing loss. HRS additionally contains information on self-reported glaucoma diagnosis. Depression and anxiety were assessed in NHATS using the PHQ2 and GAD2 questionnaires [13]. The predicted probability of dementia was determined in NHATS using a previously validated algorithm that incorporates self or proxy-reported diagnoses and performance on several cognitive function tests [14]. For HRS, several dementia prediction algorithms have been developed [15], but all rely heavily on self-reported measures of cognition and are, thus, unavailable for proxy-only respondents, so we were unable to evaluate for dementia in this dataset.
Cross-sectional and longitudinal study timepoints
We analysed NHATS and HRS data as both cross-sectional and longitudinal studies. In cross-sectional analyses, self-reported visual impairment and proxy-reported hallucinations were ascertained at the same visit, and each subject contributed one timepoint to the analysis. For subjects who had ever reported hallucinations that timepoint represented the first visit at which hallucinations were reported. For subjects who never reported hallucinations that timepoint was the baseline study visit (Supplementary Material A3). In longitudinal analyses, we incorporated all data on self-reported visual impairment and proxy-reported hallucinations across multiple visits per individual using mixed effects modelling. This allows self-reported vision and hallucinations to vary over time within a given subject and incorporates both current and lagged exposure information at each outcome risk assessment.
Statistical methods
Statistical analysis was performed using SAS v9.4 (Cary, NC), and statistical significance was defined at the P < 0.05 level. Data from NHATS and HRS were analysed separately. Descriptive statistics for hallucinators and non-hallucinators were summarised using cross-sectional data and analysed using t-tests, chi-square tests or ANOVA where appropriate. We used logistic regression to determine the cross-sectional association between self-reported vision and hallucinations adjusting for age and other potential confounders. Missing categorical variables were imputed to the mode. We analysed the longitudinal association between vision and hallucinations using mixed effects logistic regression modelling of longitudinal data (adjusted for age and other confounders) to allow multiple observations per individual. We also used mixed effects logistic regression to determine the relationship between cataract surgery and prevalent hallucinations. Of note, both NHATS and HRS are weighted population samples, and sample weights are therefore available for analysis in order to provide population level estimates of prevalence and association. However, because information on hallucinations was only available for the minority of subjects who had proxy respondents, and the use of a proxy is almost certainly non-random, these weights no longer apply to our study sample and were therefore not implemented.
Results
We included data from 1520 NHATS respondents and 3682 age-eligible HRS respondents who completed proxy cognition questionnaires. The prevalence of hallucinations and baseline characteristics for the NHATS and HRS cohorts are shown in Table 1. In both populations, hallucinations were associated with older age, female gender, race/ethnicity, hypertension, never smoking and hearing loss. Diabetes and lower annual household income were positively associated with hallucinations in HRS but not NHATS. History of stroke was more frequently reported by hallucinators in NHATS (30.6% versus 21.7% in non-hallucinators) and in HRS (39.5% versus 17.0% in non-hallucinators). There was no association between level of education and hallucinations in either sample (Supplementary Material A2).
Table 1 .
Prevalence of hallucinations, baseline demographics and medical and neuropsychiatric comorbidities in the NHATS and HRS samples
| NHATS | HRS | |||||
|---|---|---|---|---|---|---|
| No hallucinations (no, %) | Hallucinations (no, %) | P-value | No hallucinations (no, %) | Hallucinations (no, %) | P-value | |
| Total population | 921 | 599 | 2659 | 1023 | ||
| Age | <0.001 | <0.001 | ||||
| 65–74 | 173 (18.8) | 48 (8.0) | 1445 (54.3) | 180 (17.6) | ||
| 75–84 | 346 (37.6) | 188 (31.4) | 786 (29.6) | 357 (34.9) | ||
| 85+ | 402 (43.6) | 363 (60.6) | 428 (16.1) | 486 (47.5) | ||
| Sex | <0.001 | <0.001 | ||||
| Male | 413 (44.8) | 188 (31.4) | 1450 (54.5) | 326 (31.9) | ||
| Female | 508 (55.2) | 411 (68.6) | 1209 (45.5) | 697 (68.1) | ||
| Race/ethnicity | 0.01 | 0.02 | ||||
| White | 507 (55.1) | 347 (57.9) | 2087 (78.6) | 813 (79.5) | ||
| Black | 244 (26.5) | 154 (25.7) | 443 (16.7) | 182 (17.8) | ||
| Hispanic | 75 (8.1) | 63 (10.5) | 286 (10.8) | 114 (11.2) | ||
| Other | 95 (10.3) | 35 (5.8) | 127 (4.8) | 28 (2.7) | ||
| Hypertension | 0.02 | <0.001 | ||||
| Hypertension | 639 (69.5) | 447 (75.0) | 1590 (60.0) | 783 (76.8) | ||
| No hypertension | 281 (30.5) | 149 (25.0) | 1059 (40.0) | 236 (23.2) | ||
| Diabetes | 0.50 | 0.003 | ||||
| Diabetes | 267 (29.0) | 183 (30.7) | 594 (22.4) | 227 (27.1) | ||
| No diabetes | 653 (71.0) | 414 (69.3) | 2061 (77.6) | 745 (72.9) | ||
| Smoking | 0.03 | <0.001 | ||||
| Ever | 400 (43.7) | 225 (37.9) | 1508 (57.9) | 502 (49.4) | ||
| Never | 515 (56.3) | 368 (62.1) | 1098 (42.1) | 515 (50.6) | ||
| Stroke | <0.001 | <0.001 | ||||
| History of stroke | 199 (21.7) | 183 (30.6) | 451 (17.0) | 403 (39.5) | ||
| No history of stroke | 720 (78.3) | 416 (69.4) | 2203 (83.0) | 617 (60.5) | ||
| Hearing loss | <0.001 | <0.001 | ||||
| Hearing aid | N/A | N/A | 407 (15.4) | 240 (23.6) | ||
| Hearing loss | 241 (26.2) | 258 (43.1) | 610 (23.0) | 344 (33.8) | ||
| No hearing loss | 680 (73.8) | 341 (56.9) | 1634 (61.6) | 435 (42.7) | ||
| Depression (mean PHQ-2) | 3.52 (1.75) | 4.76 (1.92) | <0.001 | N/A | N/A | |
| Anxiety (mean GAD-2) | 3.25 (1.70) | 4.40 (2.13) | <0.001 | N/A | N/A | |
| Dementia | <0.001 | N/A | N/A | |||
| Probable | 467 (50.7) | 511 (91.3) | ||||
| Possible | 140 (15.2) | 4 (0.7) | ||||
| None | 314 (34.1) | 45 (8.0) | ||||
Information on depression, anxiety and dementia was unavailable from the HRS sample. In NHATS, depression, anxiety and dementia were all strongly associated with hallucinations; 91.3% of subjects with hallucinations had probable dementia compared to 50.7% of those without hallucinations (P < 0.001). In the HRS sample, hallucinations were associated with physical functional impairment and nursing home placement, both of which may be proxies for dementia (Supplementary Material A2).
The cross-sectional association between self-reported vision and hallucinations in NHATS is presented in Table 2. Self-reported inability to read newspaper print was associated with an increased odds of hallucinations after adjusting for multiple confounders (OR 1.77, 95% confidence interval (CI): 1.32–2.39). This effect size was more pronounced for distance vision as defined by self-reported difficulty recognising someone across the street (OR 2.48, 95% CI: 1.86–3.31). Blindness was rare (only 20 subjects with hallucinations and 13 without hallucinations) and was associated with hallucinations in unadjusted (OR 2.41, 95% CI: 1.19–4.89) but not adjusted (OR 2.05, 95% CI: 0.88–4.78) analyses. The use of magnifiers or other visual aids was associated with decreased odds of hallucinations in unadjusted analysis (OR 0.64, 95% CI: 48–0.86), but this did not remain statistically significant after adjusting for confounders (OR 0.71, 95% CI: 0.50–1.03).
Table 2 .
Cross-sectional association between self-reported visual function and hallucinations in the NHATS and HRS samples
| NHATS | HRS | |||||||
|---|---|---|---|---|---|---|---|---|
| Hallucinations | Hallucinations | |||||||
| No (no, %) | Yes (no, %) | OR (95% CI) | AOR (95% CI)a | No (no, %) | Yes (no, %) | OR (95% CI) | AOR (95% CI)b | |
| Impaired overall eyesight | ||||||||
| Yes | 791 (30.03) | 546 (55.10) | 2.86 (2.46–3.32) | 1.32 (1.08–1.60) | ||||
| No | 1843 (69.97) | 445 (44.90) | REF | REF | ||||
| Impaired near vision | ||||||||
| Yes | 130 (14.4) | 178 (31.6) | 2.74 (2.12–3.54) | 1.77 (1.32–2.39) | 637 (24.33) | 499 (51.18) | 3.26 (2.80–3.80) | 1.52 (1.25–1.85) |
| No | 771 (85.6) | 386 (68.4) | REF | REF | 1981 (75.67) | 476 (48.82) | REF | REF |
| Impaired distance vision | ||||||||
| Yes | 134 (15.1) | 227 (41.2) | 3.95 (3.07–5.07) | 2.48 (1.86–3.31) | 594 (22.63) | 521 (53.38) | 3.92 (3.35–4.57) | 1.61 (1.32–1.96) |
| No | 755 (84.9) | 324 (58.8) | REF | REF | 2031 (77.37) | 455 (46.62) | REF | REF |
| Requires magnifiers or other visual aids | ||||||||
| Yes | 152 (20.0) | 80 (13.8) | 0.64 (0.48–0.86) | 0.71 (0.50–1.03) | ||||
| No | 608 (80.0) | 499 (86.2) | REF | REF | ||||
| Blind | ||||||||
| Yes | 13 (1.4) | 20 (3.3) | 2.41 (1.19–4.89) | 2.05 (0.88–4.78) | 17 (0.64) | 27 (2.64) | 4.21 (2.29–7.76) | 1.99 (0.94–4.19) |
| No | 908 (98.6) | 579 (96.7) | REF | REF | 2642 (99.36) | 996 (97.36) | REF | REF |
Adjusted for age, race, sex, income, hypertension, diabetes, smoking, stroke, education, depression, anxiety, dementia and hearing loss.
Adjusted for age, race, sex, income, hypertension, diabetes, smoking, stroke, education, nursing home status, physical functional impairment and hearing loss.
In HRS, poor self-reported, overall, distance or near eyesight were all associated with hallucinations in both unadjusted and adjusted analyses (Table 2). Similar to NHATS, blindness was rare and was associated with hallucinations in unadjusted (OR 2.61, 95% CI: 1.34–5.08) but not adjusted (OR 1.99, 95% CI: 0.94–4.19) analyses. Using a longitudinal approach with mixed effects modelling, the association between blindness and hallucinations in HRS persisted after adjusting for confounding (OR 1.79, 95% CI: 1.10–2.92). Otherwise, longitudinal analysis yielded similar results to cross-sectional analysis in both cohorts (Table 3).
Table 3 .
Longitudinal association between self-reported visual function and hallucinations in the NHATS and HRS samples
| NHATS | HRS | |||
|---|---|---|---|---|
| OR (95% CI), raw | OR (95% CI), adjusteda | OR (95% CI), raw | OR (95% CI), adjustedb | |
| Impaired overall eyesight | N/A | N/A | 1.96 (1.74–2.21) | 1.40 (1.22–1.60) |
| Impaired near vision | 1.93 (1.58–2.36) | 1.77 (1.43–2.17) | 2.00 (1.77–2.26) | 1.42 (1.23–1.63) |
| Impaired distance vision | 2.02 (1.68–2.44) | 1.74 (1.43–2.11) | 2.37 (2.10–2.69) | 1.63 (1.41–1.87) |
| Requires magnifiers or other visual aids | 0.85 (0.66–1.11) | 0.81 (0.62–1.07) | N/A | N/A |
| Blind | 1.75 (1.05–2.92) | 1.62 (0.98–2.68) | 2.16 (1.37–3.40) | 1.79 (1.10–2.92) |
| Cataract surgery | 1.32 (1.09–1.60) | 1.10 (0.90–1.35) | 1.54 (1.35–1.75) | 0.89 (0.76–1.05)c |
Adjusted for age, race, sex, income, hypertension, diabetes, smoking, stroke, education, depression, anxiety, dementia and hearing loss.
Adjusted for age, race, sex, income, hypertension, diabetes, smoking, stroke, education, nursing home status, physical functional impairment and hearing loss.
Additionally adjusted for glaucoma.
Exploratory analyses examined the association between history of cataract surgery and hallucinations (Table 3). Cataract surgery was associated with increased odds of hain unadjusted models for NHATS (OR 1.32, 1.09–1.60) and for HRS (1.54, 1.35–1.75). Adjustment for sociodemographic, clinical and functional variables attenuated and rendered these associations non-significant (NHATS OR 1.11, 95% CI: 0.91–1.37; HRS OR 0.92, 95% CI: 0.79, 1.08).
Discussion
In this large population-based study of older adults, we demonstrate a robust association between self-reported visual dysfunction and hallucinations in two nationally representative cohorts. With data on over 5200 subjects, this is the among the largest studies to date of visual hallucinations in older community-dwelling adults. The vast majority of hallucinations occurred in the setting of dementia, and many of the associations we observed with hallucinations (e.g. age, history of stroke, hypertension and other vascular risk factors, hearing loss) are also risk factors for cognitive impairment and dementia. Because the majority of NHATS and HRS participants did not complete the questionnaire that included hallucinations, we are unable to estimate the absolute prevalence of hallucinations in the general US population, especially, among non-demented individuals, though other population-based studies suggest that this is very low [16,17].
The association between self-reported visual dysfunction and hallucinations was consistent across multiple measures of visual impairment including overall, distance and near eyesight and several composite measures of self-reported vision and was also replicated in two independent cohorts. With the exception of blindness, which had limited statistical power due to a very small number of blind subjects, all associations between self-reported visual function and hallucinations were robust to adjustment for multiple confounders. The association between vision and hallucinations in these two large population-based samples validates findings from smaller studies of Parkinson disease [5–8] and Alzheimer disease [9] and is consistent with the general hypothesis that external visual stimulation is important for suppressing internally generated visual images and thus that vision loss is a risk factor for hallucinations. Since many causes of vision loss are preventable or treatable, it is possible that the early detection and treatment of ophthalmic disease in the older adult population could also prevent hallucinations and its negative consequences. However, it is difficult to draw definitive conclusions from cross-sectional data, and further studies are needed to determine the etiology and severity of vision loss compared to other risk factors for hallucinations in the older adult population in order to assess the population attributable fraction of vision loss and the extent to which hallucinations could actually be prevented.
In both NHATS and HRS, a history of cataract surgery was associated with an increased risk of hallucinations in unadjusted analyses, contrary to our hypothesis that cataract surgery would be protective against hallucinations. However, the positive association between cataract surgery and hallucinations attenuated and was no longer statistically significant after adjusting for covariates. Given the inherent limitations of an observational study and the lack of information on referral patterns and visual outcomes from cataract surgery, there is not adequate evidence to suggest that cataract surgery is protective against hallucinations at this time.
A key limitation of this study is the fact that the majority of respondents in the NHATS and HRS cohorts did not complete proxy questionnaires. Proxy questionnaires are typically completed when the subject is unable to do so independently, primarily due to cognitive impairment. This is reflected by the very high prevalence of dementia among those with hallucinations. However, a large proportion of those without hallucinations also had dementia, so while hallucinations may be associated with dementia severity, they are not a marker of dementia per se. In addition to potential selection bias, dementia is also a confounder in this study, as it shares many common risk factors with visual impairment, most notably age. This is reflected in the difference between unadjusted and adjusted point estimates, especially for our analyses of cataract surgery (in unadjusted analyses, a history of cataract surgery was associated with an increased risk of hallucinations, but this attenuated after adjusting for covariates, which suggests confounding due to shared risk factors for both cataracts and hallucinations). Dementia is also associated with patterns in health-seeking behaviour and healthcare delivery that can affect visual function. For example, a recent analysis of NHATS data also found that individuals with cognitive impairment and dementia were less likely to receive cataract surgery [18]. Thus, subjects who underwent cataract surgery may have had less cognitive impairment or been otherwise systematically different from those who did not receive cataract surgery, resulting in confounding.
Other limitations of this study include the self-reporting of most covariates including vision, which may not necessarily correlate with objective measures of visual acuity [19]. In the future, we plan to address this using data from the Aging, Demographics and Memory Study, an HRS substudy that contains near acuity measurements in addition to standard questionnaires, to assess the correlation between self-reported and objective measures of visual function and their association with hallucinations. We were also unable to ascertain the cause of a subject’s visual impairment and thus whether it was actually reversible or not. Hallucinations were also proxy-reported, which has the potential to be misclassified at several levels: first, subjects must verbalise complaints or demonstrate behaviours suggestive of hallucinations, and second, caregivers must notice these and disentangle them from other visual (e.g. photopsias and metamorphopsia) and cognitive (e.g. agitation) symptoms. As hallucinations are a purely internal phenomenon, they are unfortunately difficult to confirm or validate using objective measures (unlike visual acuity, for example), and the NHATS and HRS questionnaires did not distinguish between visual and auditory hallucinations. Thus, combining both visual and auditory hallucinations might have reduced the observed association between visual impairment and visual hallucinations. In our cross-sectional analyses, there was a bias towards hallucinators being older than non-hallucinators because baseline or follow-up visit data were used for hallucinators whereas only baseline visit data were used for non-hallucinators. However, we adjusted for age in our cross-sectional logistic regression models, and mixed effects logistic regression models which incorporated baseline and follow-up visit data for all subjects confirmed our findings.
In this study, we found a replicable association between visual impairment and hallucinations in the general US population. Further research on the prevalence and severity of specific eye diseases in the older adult population with and without hallucinations is needed to determine whether hallucinations could be prevented with improved detection and treatment of vision loss.
Supplementary Material
Acknowledgements
The NHATS is sponsored by the National Institute on Aging (grant number NIA U01AG032947) through a cooperative agreement with the Johns Hopkins Bloomberg School of Public Health. The HRS is supported by the National Institute on Aging, supplemented by the Social Security Agency, and operated from the Institute for Social Research (ISR) at the University of Michigan. This analysis uses data or information from the Harmonised HRS dataset and Codebook, Version B as of October 2018 developed by the Gateway to Global Aging Data. The development of the Harmonised HRS was funded by the National Institute on Aging (R01 AG030153, RC2 AG036619, 1R03AG043052). For more information, refer to www.g2aging.org.
Declaration of Conflicts of Interestac
None.
Declaration of Sources of Funding
This work was supported by the National Institute of Neurological Disorders and Stroke (grant number T32 NS061779–10 to AGH, grant number R01 NS099129-01A1 to AWW.
References
- 1. Zhao Q-F, Tan L, Wang H-Fet al. The prevalence of neuropsychiatric symptoms in Alzheimer’s disease: systematic review and meta-analysis. J Affect Disord 2016; 190: 264–71. [DOI] [PubMed] [Google Scholar]
- 2. Barnes J, David AS. Visual hallucinations in Parkinson’s disease: a review and phenomenological survey. J Neurol Neurosurg Psychiatry 2001; 70: 727–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Wilson RS, Krueger KR, Kamenetsky JMet al. Hallucinations and mortality in Alzheimer disease. Am J Geriatr Psychiatry 2005; 13: 984–90. [DOI] [PubMed] [Google Scholar]
- 4. Scarmeas N, Brandt J, Albert Met al. Delusions and hallucinations are associated with worse outcome in Alzheimer disease. Arch Neurol 2005; 62: 1601–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Holroyd S, Currie L, Wooten GF. Prospective study of hallucinations and delusions in Parkinson’s disease. J Neurol Neurosurg Psychiatry 2001; 70: 734–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. de Maindreville AD, Fénelon G, Mahieux F. Hallucinations in Parkinson’s disease: a follow-up study. Mov Disord 2005; 20: 212–7. [DOI] [PubMed] [Google Scholar]
- 7. Matsui H, Udaka F, Tamura Aet al. Impaired visual acuity as a risk factor for visual hallucinations in Parkinson’s disease. J Geriatr Psychiatry Neurol 2006; 19: 36–40. [DOI] [PubMed] [Google Scholar]
- 8. Archibald NK, Clarke MP, Mosimann UP, Burn DJ. Visual symptoms in Parkinson’s disease and Parkinson’s disease dementia. Mov Disord 2011; 26: 2387–95. [DOI] [PubMed] [Google Scholar]
- 9. Chapman FM, Dickinson J, McKeith I, Ballard C. Association among visual hallucinations, visual acuity, and specific eye pathologies in Alzheimer’s disease: treatment implications. Am J Psychiatry 1999; 156: 1983–5. [DOI] [PubMed] [Google Scholar]
- 10. NHATS Public Use Data . Rounds 1-7. Sponsored by the National Institute on Aging (grant number NIA U01AG032947) through a cooperative agreement with the Johns Hopkins Bloomberg School of Public Health .https://www.nhats.org (accessed 15 February 2019).
- 11. Harnonized HRS Dataset and Codebook Version B, October 2018 . Developed by the Gateway to Globa Aging Data, Funded by the National Institute on Aging (R01 AG030153 RC2 AG036619, 1R03Ag043052) .https://www.G2aging.org (accessed 15 February 2019).
- 12. Sloan FA, Ostermann J, Brown DS, Lee PP. Effects of changes in self-reported vision on cognitive, affective, and functional status and living arrangements among the elderly. Am J Ophthalmol 2005; 140: 618–27. [DOI] [PubMed] [Google Scholar]
- 13. Davydow DS, Zivin K, Langa KM. Hospitalization, depression and dementia in community-dwelling older Americans: findings from the national health and aging trends study. Gen Hosp Psychiatry 2014; 36: 135–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Kasper J, Freedman V, Spillman B. Classification of Persons by Dementia Status in the National Health and Aging Trends Study. Technical Pper #5. Baltimore, Maryland, USA: Baltim Johns Hopkins Bloom School of Public Health, 2013. [Google Scholar]
- 15. Gianattasio KZ, Wu Q, Glymour MM, Power MC. Comparison of Methods for algorithmic classification of dementia status in the health and retirement study. Epidemiol Camb Mass 2019; 30: 291–302. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Geda YE, Roberts RO, Knopman DSet al. Prevalence of neuropsychiatric symptoms in mild cognitive impairment and normal cognitive aging: population-based study. Arch Gen Psychiatry 2008; 65: 1193–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Lyketsos CG, Steinberg M, Tschanz JT, Norton MC, Steffens DC, Breitner JC. Mental and behavioral disturbances in dementia: findings from the Cache County study on memory in aging. Am J Psychiatry 2000; 157: 708–14. [DOI] [PubMed] [Google Scholar]
- 18. Stagg BC, Ehrlich JR, Choi H, Levine DA. Association of Cognitive Impairment and Dementia with receipt of cataract surgery among community-dwelling Medicare beneficiaries. JAMA Ophthalmol 2018. doi: 10.1001/jamaophthalmol.2018.5138. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Zhang X, Bullard KM, Cotch MFet al. Association between depression and functional vision loss in persons 20 years of age or older in the United States, NHANES 2005-2008. JAMA Ophthalmol 2013; 131: 573–81. [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.