Introduction
Previous research has identified a link between sleep disturbances and cognitive impairment; however, no study has examined this relationship using a national United States sample. This study examines how multiple longitudinal measures of sleep disturbances (sleep-initiation insomnia, sleep-maintenance insomnia, sleep-medication usage) are associated with dementia risk.
Methods
Ten annual waves (2011–2020) of prospective cohort data from a nationally representative U.S. sample of adults aged ≥65 years were analyzed from the National Health and Aging Trends Study. Sleep disturbances were converted into a longitudinal score and measured as sleep-initiation insomnia (trouble falling asleep in 30 minutes), sleep-maintenance insomnia (trouble falling asleep after waking up early), and sleep-medication usage (taking medication to help sleep). Cox regression models analyzed time to dementia diagnosis for a sample of 6,284 respondents.
Results
In the unadjusted model, sleep-initiation insomnia was significantly associated with a 51% increased dementia risk (hazard ratio=1.51, 95% CI=1.19, 1.90). Adjusted for sociodemographics, sleep-medication usage was significantly associated with a 30% increased dementia risk (adjusted hazard ratio=1.30, 95% CI=1.08–1.56). Adjusted for sociodemographics and health, sleep-maintenance insomnia was significantly associated with a 40% decreased dementia risk (adjusted hazard ratio=0.60, 95% CI=0.46, 0.77).
Conclusions
These findings suggest that sleep-initiation insomnia and sleep-medication usage may elevate dementia risk. On the basis of the current evidence, sleep disturbances should be considered when assessing the risk profile for dementia. Future research is needed to examine other sleep disturbance measures and to explore the mechanisms for decreased dementia risk among older adults with sleep-maintenance insomnia.
INTRODUCTION
Sleep disturbances are a common issue for persons with dementia.
1
These can include insomnia, shortened nocturnal sleep, increased nocturnal awakenings, increased daytime sleep, and reduced rapid eye movement (REM) sleep.2
UP to 44% of patients with Alzheimer's disease (AD) have sleep disturbances, and that number rises to 90% for those with dementia with Lewy bodies or Parkinson disease dementia.1
, 2
, 3
, 4
Other evidence suggests even higher rates among patients with vascular dementia, who have 2.5 times more frequent sleep disturbances than individuals with AD or frontotemporal dementia.1
These sleep disturbances cause problems related to irritability, attention, motivation, and cognition, which impact the quality of life of both the patients and their caregivers. Moreover, sleep debt is associated with an increased risk of falls and increased sympathovagal imbalance, tying this issue to serious morbidity and mortality among people with dementia.1
Evidence has grown over the past several years that sleep disturbances are a risk factor for dementia.
1
,5
, 6
, 7
, 8
For example, idiopathic REM sleep behavior disorder often predicts the onset of neurodegenerative disease in ages 10–15 years.1
Only 5 hours of sleep per night can increase dementia risk by 30% for adults aged ≥50 years,7
and sleep-related movement disorders, one of the classes of sleep disorders, increases the risk of all-cause dementia by 3.95 times for middle-aged and older adults.6
It is likely that sleep deprivation and insomnia increase amyloid-β levels in plasma and the brain—a major clinical indicator of and pathological agent in AD.4
,5
,9
,10
Sleep fragmentation may also exacerbate the effect of the APOEε4 gene on cognitive decline, and sleep restriction can impact the permeability of the blood‒brain barrier and oxidative stress in the brain.11
More generally, poor sleep impedes prefrontal cortex function and hinders sleep's ability to maintain cognitive health.5
Other research suggests that the association could be mediated by benzodiazepine and z-hypnotic sleep medication: regular use of short-acting benzodiazepines nearly doubles the risk of dementia in older adults, and the use of z-hypnotics increases the risk of cognitive decline by 33% and all-cause dementia by 79%.8
,12
Despite all this evidence, a review article examining the association between sleep disturbance and dementia has shown that many of these studies are cross-sectional in nature,
2
and interventions to improve cognitive function in older adults through sleep quality have had limited or mixed results.5
,13
Long-term cohort studies, by contrast, not only better establish temporal order but have also had more dramatic findings so far.6
,7
At this time, no research has examined the relationship between longitudinal sleep disturbance measures and dementia using a national U.S. cohort,2
,14
which enables these findings to be more generalizable to the whole population. Therefore, the purpose of this study was to analyze how multiple longitudinal measures of sleep disturbances (sleep-initiation insomnia, sleep-maintenance insomnia, sleep-medication usage) are associated with dementia risk in a nationally representative sample of U.S. older adults. Given the positive associations between sleep disturbances and cognitive impairment in previous research, this study hypothesized that higher frequencies for each of the sleep disturbance measures would be associated with increased dementia risk.METHODS
Study Sample
This study used 10 annual waves (2011–2020) of prospective data from the National Health and Aging Trends Study (NHATS), a longitudinal panel study that surveys a nationally representative sample of Medicare beneficiaries aged ≥65 years within the contiguous U.S. This study excluded older adults with a pre-existing dementia diagnosis or were living in residential care. Thus, the sample was restricted to 6,284 community-dwelling older adults who were dementia free at baseline in 2011. This study was approved by the SUNY Upstate IRB for the Protection of Human Subjects.
Measures
Dementia diagnosis was derived from an NHATS algorithm
15
that uses 3 cognitive measures: (1) AD8 Dementia Screening Interview16
that assesses memory, temporal orientation, judgment, and function; (2) cognitive tests that evaluate the respondents’ memory (e.g., immediate 10-word recall), orientation (e.g., date), and executive function (e.g., clock drawing test); and (3) self-report of an AD or dementia diagnosis by a doctor. From these cognitive measures, respondents were categorized into 3 groups: no dementia, possible dementia, or probable dementia. In accordance with NHATS guidelines, this study collapsed these categories into a binary variable for a narrow definition of dementia (probable diagnosis) versus no dementia (no dementia or possible dementia), which the NHATS has tested to have a reasonably good sensitivity of 65.7% and high specificity of 87.2% compared with dementia diagnoses in a consensus expert panel from the Aging, Demographics and Memory Study.15
For each wave from 2011 to 2020, there were 3 measures for sleep disturbances. The first is sleep-initiation insomnia, in which respondents were asked, In the last month how often did it take you more than 30 minutes to fall asleep?. The second is sleep-maintenance insomnia, in which respondents were asked, In the last month on nights when you woke up before you wanted to get up, how often did you have trouble falling back asleep?. The third is sleep-medication usage, in which respondents were asked, In the last month how often did you take medication to help you sleep?.
For each of the 3 sleep disturbance measures, respondents could answer never/not a problem, rarely (once a week or less), some nights (2–4 nights a week), most nights (5–6 nights a week), or every night (7 nights a week). As the first step to create longitudinal sleep disturbance measures, this study converted each of the 3 sleep disturbances into a binary variable: no (never or rarely) or yes (some nights, most nights, or every night). For the second step, each binary sleep disturbance variable from all waves was converted into a longitudinal score, which measures the proportion of years with each respective sleep disturbance before dementia diagnosis or being censored. This longitudinal score can have a range from 0 to 1 (or 0%‒100%). For example, a respondent who entered the study in 2011 and died in 2015 may have reported sleep-initiation insomnia at 2 annual waves during the 4-year window they were alive from 2011 to 2014, resulting in a longitudinal sleep-initiation insomnia score that is 2/4 or 0.50. This approach was repeated for sleep-maintenance insomnia and sleep-medication usage. These scores therefore represent a longitudinal measure for each sleep disturbance between the first wave (2011) and up until 2020 or earlier depending on the year respondents were diagnosed with dementia or were censored.
Sociodemographic and health covariates were included in the regression models. All covariates were selected on the basis of previous research on dementia risk factors
17
and a comparable study using this data set for dementia research.18
Sociodemographic covariates included age, gender (male or female), race and ethnicity (non-Hispanic White, non-Hispanic Black, Hispanic, Asian, or Other), the highest level of education (less than high school, high school, or college), total household income, marital status (married or unmarried), and metropolitan residence (metropolitan or nonmetropolitan). Health covariates included self-rated overall health condition (poor, fair, good, very good, or excellent), BMI, activities of daily living (ADLs) (no ADL limitations or at least 1 ADL limitation), proxy respondent, major depressive disorder, generalized anxiety disorder, history of heart attack, history of hypertension, and history of diabetes.Statistical Analysis
The number of years from dementia free at baseline (2011) to dementia diagnosis was analyzed using a Cox proportional hazards model with survey sampling weights applied and Stata subpop command to generate results for only older adults who met eligibility criteria. Respondents were censored if they died, dropped out, or did not have a dementia diagnosis by the end of the study window. To maximize the full number of respondents in the data set and minimize bias because of missing data (6.4%), multiple imputation by chained equations generated 100 imputed data files with 10 iterations each for regression analyses. Other than the outcome, all variables were imputed. All statistical analyses were performed in Stata statistical software, Version 17 (StataCorp LLC, College Station, TX) at a 0.05 significance level.
RESULTS
After applying survey sampling weights, the 6,284 NHATS respondents were estimated to represent 30,378,878 older adults in the U.S. population. About 13.6% had an incident dementia diagnosis between 2011 and 2020, with an average incident dementia diagnosis of 2.9 years. The average sample age was 76.3 years, and more than half (55.5%) were female at baseline (Table 1). Racial and ethnic composition for the sample was primarily non-Hispanic White (81.7%), followed by non-Hispanic Black (7.8%), Hispanic (6.3%), Asian (2.2%), and other (2.1%). The majority were college educated (64.0%), with an average household income of about $62,000. Most were married (60.5%) and were residing in a metropolitan area (81.6%).
Table 1Weighted Sample Characteristics at Baseline
| Characteristics | Whole sample (N=30.4) | No dementia (n=26.3) | Dementia (n=4.1) | Bivariate test |
|---|---|---|---|---|
| Age (in years) | 76.28 (6.7) | 73.65 (6.3) | 78.28 (8.0) | t= −17.33, p<0.001 |
| Female (%, n) | 55.46 (16.8) | 54.80 (14.4) | 59.68 (2.5) | F=8.39, p<0.01 |
| Race and ethnicity (%, n) | F=7.86, p<0.001 | |||
| White, non-Hispanic | 81.66 (24.8) | 82.84 (21.8) | 74.12 (3.1) | |
| Black, non-Hispanic | 7.76 (2.4) | 7.32 (1.9) | 10.56 (0.4) | |
| Hispanic | 6.26 (1.9) | 5.60 (1.5) | 10.49 (0.4) | |
| Asian | 2.20 (0.7) | 2.14 (5.6) | 2.62 (0.1) | |
| Other | 2.12 (0.6) | 2.10 (5.5) | 2.20 (0.1) | |
| Highest level of education (%, n) | F=239.45, p<0.001 | |||
| Less than high school | 9.22 (2.8) | 6.83 (1.8) | 24.44 (1.0) | |
| High-school degree | 26.74 (8.1) | 25.14 (6.6) | 36.95 (1.5) | |
| College degree | 64.04 (19.5) | 68.03 (17.9) | 38.61 (1.6) | |
| Income (thousands of U.S. dollars) | 62.24 (194.1) | 65.57 (201.8) | 41.04 (85.9) | t=5.28, p<0.001 |
| Married (%, n) | 60.52 (18.4) | 62.36 (16.4) | 48.83 (2.0) | F=42.69, p<0.001 |
| Metropolitan residence (%, n) | 81.61 (24.8) | 82.04 (21.6) | 78.87 (3.3) | F=3.00, p=0.09 |
| Self-rated health (0–4; poor‒excellent) | 2.36 (1.10) | 2.41 (1.06) | 2.04 (1.28) | t=7.16, p<0.001 |
| BMI | 27.82 (5.53) | 27.90 (5.41) | 27.28 (6.23) | t=3.07, p<0.01 |
| ADL limitations (%, n) | F=101.44, p<0.001 | |||
| None | 91.34 (27.8) | 92.77 (24.4) | 82.22 (3.4) | |
| At least 1 | 8.66 (2.6) | 7.23 (1.9) | 17.78 (0.7) | |
| Proxy respondent (%, n) | 1.66 (0.5) | 1.43 (0.4) | 3.13 (0.1) | F=10.27, p<0.01 |
| Depression (%, n) | 12.0 (3.6) | 10.73 (2.8) | 20.05 (0.8) | F=57.77, p<0.001 |
| Anxiety (%, n) | 10.44 (3.2) | 9.35 (2.4) | 17.36 (0.7) | F=52.88, p<0.001 |
| History of heart attack (%, n) | 13.13 (4.0) | 12.4 (3.3) | 17.73 (0.7) | F=15.98, p<0.001 |
| History of hypertension (%, n) | 63.55 (19.3) | 62.95 (16.6) | 67.35 (2.8) | F=4.56, p<0.05 |
| History of diabetes (%, n) | 23.03 (7.0) | 22.21 (5.8) | 28.26 (1.2) | F=15.28, p<0.001 |
Note: Boldface indicates statistical significance (p<0.05).
Unless otherwise indicated, data are expressed as mean (SD). All frequencies are in millions.
a t test for continuous variables and chi-square test for categorical variables.ADL, activity of daily living.
In terms of health characteristics, the average health was 2.4, which is between good (2) and very good (3) health. The average BMI was 27.8, suggesting that most of the sample was overweight. The majority (91.3%) did not have any ADL limitations or need for a proxy respondent (1.7%). The most prevalent health conditions within the sample were a history of hypertension (63.6%) and diabetes (23.0%).
Average longitudinal scores measuring the frequency for each type of sleep disturbance are reported in Table 2. On average, respondents experienced sleep-initiation insomnia about 43.0% of the time, sleep-maintenance insomnia was present approximately 42.3% of the time, and sleep-medication usage occurred 21.7% of the time. There were no statistically significant differences in average scores for all the 3 sleep disturbance measures between those with and without dementia. On average, older adults with dementia had more frequent sleep-initiation insomnia and sleep-medication usage but less frequent sleep-maintenance insomnia.
Table 2Weighted Average Frequency of Sleep Disturbances by Dementia Status
| Sleep disturbance type | Whole sample, % | No dementia, % | Dementia, % | t test result |
|---|---|---|---|---|
| Sleep-initiation insomnia | 42.96 | 42.68 | 44.75 | t= −1.23, p=0.22 |
| Sleep-maintenance insomnia | 42.31 | 42.67 | 40.04 | t=1.55, p=0.13 |
| Sleep-medication usage | 21.71 | 21.39 | 23.84 | t= −1.90, p=0.06 |
Note: This longitudinal score can have a range from 0 to 100%, which measures the proportion of years with each respective sleep disturbance before dementia diagnosis or being censored.
For sleep-initiation insomnia, the unadjusted crude model (Model A) in Table 3 indicates that sleep-initiation insomnia is significantly associated with about a 51% increased risk for dementia (hazard ratio [HR]=1.51, 95% CI=1.19, 1.90, p<0.01). This association is elevated but no longer significant in subsequent models after adjusting for sociodemographics (Model B) and further adjusting for health (Model C).
Table 3Unadjusted and adjusted hazard of dementia by sleep disturbance type
| Sleep disturbance type | Model A HR (95% CI), p-value | Model B aHR (95% CI), p-value | Model C aHR (95% CI), p-value |
|---|---|---|---|
| Sleep-initiation insomnia | 1.51 (1.19, 1.90), <0.01 | 1.12 (0.89, 1.42), 0.33 | 1.00 (0.79, 1.25), 0.97 |
| Sleep-maintenance insomnia | 0.66 (0.50, 0.86), <0.01 | 0.66 (0.51–0.85), <0.01 | 0.60 (0.46, 0.77), <0.001 |
| Sleep-medication usage | 1.15 (0.95, 1.38), 0.15 | 1.30 (1.08, 1.56), <0.01 | 1.08 (0.90, 1.28), 0.40 |
| Model significance | F(3, 54)=6.12, p<0.01 | F(14, 54)=68.94, p<0.001 | F(23, 54)=51.53, p<0.001 |
Note: Boldface indicates statistical significance (p<0.05).
Model A is an unadjusted crude model, Model B is adjusted for sociodemographics, and Model C is adjusted for sociodemographics and health.
aHR, adjusted hazard ratio; HR, hazard ratio.
There was a significant association between sleep-maintenance insomnia and dementia across all the models, in which the HR remained fairly consistent in the unadjusted versus in fully adjusted model. In the final model (Model C), sleep-maintenance insomnia was significantly associated with decreased dementia risk by 40% (adjusted HR [aHR]=0.60, 95% CI=0.46, 0.77, p<0.001).
Finally, there was initially no association between sleep-medication usage and dementia in the unadjusted model (aHR=1.15, 95% CI=0.95, 1.38, p=0.15). Once adjusted for sociodemographic covariates, sleep-medication usage was significantly associated with increased dementia risk by 30% (aHR=1.30, 95% CI=1.08, 1.56, p<0.01). This association remained elevated but no longer significant after further adjusting for health (aHR=1.08, 95% CI=0.90, 1.28, p=0.40).
DISCUSSION
This study analyzed the association between sleep disturbances and dementia risk. There were statistically significant differences for a large number of sociodemographic and health variables by dementia status, which are well documented in the literature,
17
providing the rationale to adjust for them in the multivariable models. Although the bivariate results generally suggested that older adults with dementia did not have significantly higher frequencies of sleep disturbances, the unadjusted and adjusted regression models indicated significant associations.Sleep-initiation insomnia was only significantly associated with dementia in the unadjusted Cox regression model. Previous research has associated longer sleep latency with higher odds of incident cognitive impairment,
19
with another study similarly finding a significant association between sleep latency >30 minutes and dementia incidence.20
In both studies, the relationship between sleep latency and cognitive impairment/dementia withstood controlling for sociodemographic and health variables, whereas this study did not. However, a different paper did not see any significant HRs between difficulty in initiating sleep and incidence in all-cause dementia,21
potentially putting doubt on the association similar to this study. Different results across the literature may have come from differences in covariates used, especially given that the significance of these results disappeared when sociodemographic variables were added to the model.The most consistent and most surprising results come from the sleep-maintenance insomnia variable. The bivariate association between this form of insomnia and dementia was statistically significant, and all Cox regression models were statistically significant as well. Although despite the initial hypotheses, respondents with sleep-maintenance insomnia were less likely to develop dementia over the course of the decade than respondents who did not report insomnia. This negative association falls in an interesting spot in the literature: whereas some studies have found positive associations,
11
,22
others have similarly found negative ones23
or no association at all.21
,24
These mixed results deserve further research for their own clarity, but if the negative association is correct, more investigation should be done on how these conditions can connect counterintuitively. One possible explanation relates to the Cognitive Reserve Theory.25
Whereas insomnia may relate to dementia incidence, overall sleep time does not.19
,26
If sleep disturbances do not negatively affect sleep architecture and if sleep efficiency remains high, self-reported sleep interruptions might not then meaningfully reduce the quality of sleep among older adults, regardless of their expectations. More time awake, even at night, may allow older adults to spend more time enjoying social and physical activities that can replenish or maintain their cognitive reserve.25
This extra productivity may be the mechanism by which those with sleep-maintenance insomnia reduce their own dementia risk over the years.23
Finally, sleep-medication usage was significantly associated with dementia incidence in only the second of the Cox regression models. This finding was partially expected: many older adults take sleep medication owing to their health conditions,
27
and controlling for health variables would have thus attenuated the association between sleep medication and dementia. Depression in particular can have a large overlap with both dementia and insomnia because some medication is used to treat both depression and poor sleep.- Full KM
- Pusalavidyasagar S
- Palta P
- et al.
Associations of late-life sleep medication use with incident dementia in the atherosclerosis risk in communities (ARIC) study.
J Gerontol A Biol Sci Med Sci. 2022; (In press. Online April 14)https://doi.org/10.1093/gerona/glac088
27
In addition, some sleep medications such as long-acting benzodiazepines may be used less often in older adults, especially those with multiple comorbidities, than in others because of extant concerns that they may cause cognitive decline and other issues such as falls.- Full KM
- Pusalavidyasagar S
- Palta P
- et al.
Associations of late-life sleep medication use with incident dementia in the atherosclerosis risk in communities (ARIC) study.
J Gerontol A Biol Sci Med Sci. 2022; (In press. Online April 14)https://doi.org/10.1093/gerona/glac088
8
Overall, the literature on potential associations between sleep medication and dementia is mixed, and both the complex relationship between the 2 and a lack of long-term studies have impeded attempts to determine causality in either direction.12
Future research is needed in particular to assess whether this association may hold depending on the type of sleep medication, which was not collected in this study.Limitations
There were 4 limitations in this study. First, the NHATS collects information for only 3 types of sleep disturbances annually. However, previous research indicates other forms of sleep disturbances, such as REM sleep behavior disorder, are linked to cognitive impairment.
1
,3
Second, a binary cutoff was used for each sleep disturbance variable to create longitudinal sleep disturbance scores, and these scores do not measure the timing of when the sleep disturbances may have started or ended. Third, the dependent variable was all-cause dementia; however, previous research indicates that the association between sleep disturbance and dementia may depend on dementia subtype. For instance, 1 meta-analysis found that insomnia significantly increased the risk for AD but not for vascular dementia.14
Fourth, censoring individuals who have died during the study is standard practice for Cox models; however, there may be biased risk estimates because death is a competing risk. Despite the limitations, these findings offer an important contribution to the field because they present the first examination analyzing the link between multiple longitudinal measures of sleep disturbances and dementia risk. Core strengths of this study include using a large, nationally representative older adult sample with longitudinal sleep measures, which have been limited in previous research on this topic.CONCLUSIONS
This study identified that sleep-initiation insomnia and sleep-medication usage are associated with increased dementia risk, whereas sleep-maintenance insomnia is associated with decreased dementia risk. Given the high prevalence of sleep disturbances among older adults compared with that among other age groups,
1
these findings highlight the importance of considering sleep disturbance history when assessing the dementia risk profile for older adults. Future research is needed to examine other sleep disturbance measures using a national longitudinal sample, whether these sleep-dementia findings hold true for specific dementia subtypes, and how certain characteristics (e.g., age, education, race‒ethnicity) may moderate the association between sleep and dementia.ACKNOWLEDGMENTS
Neither the National Health and Aging Trends Study team nor the National Institute on Aging had a role in the design, methods, analysis, or preparation of this paper.
National Health and Aging Trends Study data are sponsored by the National Institute on Aging (Grant Number U01AG32947) and were conducted by Johns Hopkins University. This study was approved by the SUNY Upstate IRB for the Protection of Human Subjects (Number 1826969-1).
No financial disclosures were reported by the authors of this paper.
CRediT AUTHOR STATEMENT
Roger Wong: Conceptualization, Formal analysis, Methodology, Supervision, Writing - review & editing. Margaret Lovier: Formal analysis, Writing - original draft.
REFERENCES
- Sleep disturbances and dementia.Psychogeriatrics. 2015; 15: 65-74https://doi.org/10.1111/psyg.12069
- Is sleep disruption a risk factor for Alzheimer's disease?.J Alzheimers Dis. 2017; 58: 993-1002https://doi.org/10.3233/JAD-161287
- Sleep disorders in aging and dementia.J Nutr Health Aging. 2010; 14: 212-217https://doi.org/10.1007/s12603-010-0052-7
- Sleep deprivation induced plasma amyloid-β transport disturbance in healthy young adults.J Alzheimers Dis. 2017; 57: 899-906https://doi.org/10.3233/JAD-161213
- The independent associations of physical activity and sleep with cognitive function in older adults.J Alzheimers Dis. 2018; 63: 1469-1484https://doi.org/10.3233/JAD-170936
- Increased risk of dementia among sleep-related movement disorders: a population-based longitudinal study in Taiwan.Medicine (Baltimore). 2015; 94: e2331https://doi.org/10.1097/MD.0000000000002331
- Association of sleep duration in middle and old age with incidence of dementia.Nat Commun. 2021; 12: 2289https://doi.org/10.1038/s41467-021-22354-2
- Benzodiazepines, z-hypnotics, and risk of dementia: special considerations of half-lives and concomitant use.Neurotherapeutics. 2020; 17: 156-164https://doi.org/10.1007/s13311-019-00801-9
- Increased cerebrospinal fluid amyloid-β during sleep deprivation in healthy middle-aged adults is not due to stress or circadian disruption.J Alzheimers Dis. 2020; 75: 471-482https://doi.org/10.3233/JAD-191122
- Cerebrospinal fluid amyloid-β levels are increased in patients with insomnia.J Alzheimers Dis. 2018; 61: 645-651https://doi.org/10.3233/JAD-170032
- Sleep fragmentation and the risk of incident Alzheimer's disease and cognitive decline in older persons.Sleep. 2013; 36: 1027-1032https://doi.org/10.5665/sleep.2802
- The frequency of sleep medication use and the risk of subjective cognitive decline (SCD) or SCD with functional difficulties in elderly individuals without dementia.J Gerontol A Biol Sci Med Sci. 2020; 75: 1693-1698https://doi.org/10.1093/gerona/glz269
- Effect of a multimodal lifestyle intervention on sleep and cognitive function in older adults with probable mild cognitive impairment and poor sleep: a randomized clinical trial.J Alzheimers Dis. 2020; 76: 179-193https://doi.org/10.3233/JAD-200383
- Sleep disturbances increase the risk of dementia: a systematic review and meta-analysis.Sleep Med Rev. 2018; 40: 4-16https://doi.org/10.1016/j.smrv.2017.06.010
- Classification of persons by dementia status in the National Health and Aging Trends Study. NHATS technical paper #5.Johns Hopkins University Bloomberg School of Public Health, Baltimore, MD2013 (Published JulyAccessed July 6, 2022)
- The AD8: a brief informant interview to detect dementia.Neurology. 2005; 65: 559-564https://doi.org/10.1212/01.wnl.0000172958.95282.2a
- 2022 Alzheimer's disease facts and figures.Alzheimers Dement. 2022; 18: 700-789https://doi.org/10.1002/alz.12638
- Role of neighborhood physical disorder and social cohesion on racial and ethnic disparities in dementia risk.J Aging Health. 2022; 34: 1178-1187https://doi.org/10.1177/08982643221101352
- Measures of sleep–wake patterns and risk of mild cognitive impairment or dementia in older women.Am J Geriatr Psychiatry. 2016; 24: 248-258https://doi.org/10.1016/j.jagp.2015.12.002
- Examining sleep deficiency and disturbance and their risk for incident dementia and all-cause mortality in older adults across 5 years in the United States.Aging (Albany NY). 2021; 13: 3254-3268https://doi.org/10.18632/aging.202591
- Complaints of daytime sleepiness, insomnia, hypnotic use, and risk of dementia: a prospective cohort study in the elderly.Alzheimers Res Ther. 2022; 14: 12https://doi.org/10.1186/s13195-021-00952-y
- Sleep in older adults: normative changes, sleep disorders, and treatment options.Clin Geriatr Med. 2014; 30: 591-627https://doi.org/10.1016/j.cger.2014.04.007
- 0272 Decreased risk of 2-year incidence of Alzheimer's disease among older adults who report sleep symptoms.Sleep. 2022; 45: A122-A123https://doi.org/10.1093/sleep/zsac079.270
- The association of self-reported sleep duration, difficulty sleeping, and snoring with cognitive function in older women.Alzheimer Dis Assoc Disord. 2006; 20: 41-48https://doi.org/10.1097/01.wad.0000201850.52707.80
- Whitepaper: defining and investigating cognitive reserve, brain reserve, and brain maintenance.Alzheimers Dement. 2020; 16: 1305-1311https://doi.org/10.1016/j.jalz.2018.07.219
- Association of sleep quality and dementia among long-lived Chinese older adults.Age (Dordr). 2013; 35: 1423-1432https://doi.org/10.1007/s11357-012-9432-8
- Associations of late-life sleep medication use with incident dementia in the atherosclerosis risk in communities (ARIC) study.J Gerontol A Biol Sci Med Sci. 2022; (In press. Online April 14)https://doi.org/10.1093/gerona/glac088
Article info
Publication history
Published online: January 25, 2023
Publication stage
In Press Journal Pre-ProofIdentification
Copyright
© 2023 American Journal of Preventive Medicine. Published by Elsevier Inc. All rights reserved.
