Literature Review Resources

Psychosocial risk factors and Alzheimer’s disease: the associative effect of depression, sleep disturbance, and anxiety

Shanna L. Burke a, Tamara Cadet b, Amary Alcidea, Janice O’Driscolla and Peter Maramaldic,d,e

aSchool of Social Work, Robert Stempel College of Public Health and Social Work, Florida International University, Miami, FL, USA; bOral Health Policy and Epidemiology, Simmons College School of Social Work, Harvard School of Dental Medicine, Boston, MA, USA; cHartford Faculty Scholar & National Mentor, Simmons College School of Social Work, Boston, MA, USA; dOral Health Policy and Epidemiology, Harvard School of Dental Medicine, Boston, MA, USA; eDepartment of Social and Behavioral Sciences, Harvard T.H. Chan School of Public Health, Boston, MA, USA

ARTICLE HISTORY Received 6 June 2016 Accepted 12 September 2017

ABSTRACT Objectives: Alzheimer’s disease (AD) dementia is a neurodegenerative condition, which leads to impairments in memory. This study predicted that sleep disturbance, depression, and anxiety increase the hazard of AD, independently and as comorbid conditions. Methods: Data from the National Alzheimer’s Coordinating Center was used to analyze evaluations of 12,083 cognitively asymptomatic participants. Survival analysis was used to explore the longitudinal effect of depression, sleep disturbance, and anxiety as predictors of AD. The comorbid risk posed by depression in the last two years coupled with sleep disturbance, lifetime depression and sleep disturbance, clinician-verified depression and sleep disturbance, sleep disturbance and anxiety, depression in the last two years and anxiety, lifetime depression and anxiety, and clinician-verified depression and anxiety were also analyzed as predictors of AD through main effects and additive models. Results: Main effects models demonstrated a strong hazard of AD development for those reporting depression, sleep disturbance, and anxiety as independent symptoms. The additive effect remained significant among comorbid presentations. Conclusion: Findings suggest that sleep disturbance, depression, and anxiety are associated with AD development among cognitively asymptomatic participants. Decreasing the threat posed by psychological symptoms may be one avenue for possibly delaying onset of AD.

KEYWORDS Alzheimer’s disease; dementia; anxiety; depression; sleep disturbance

Introduction

Alzheimer’s disease (AD) research has identified several risk factors that may be associated with to the development of AD including advancing age, the apolipoprotein (APOE) e4 gene, female sex, family history of AD, mild cognitive impairment, smoking, obesity at midlife, diabetes, midlife hypertension, midlife high cholesterol, moderate and severe traumatic brain injury, fewer years of formal education, and lower levels of social and cognitive engagement (Albert et al., 2011; Alzheimer’s Association, 2015a; Di Marco et al., 2014; Green et al., 2002; He, Zhang, & Zhang, 2000 Lautensclager et al., 1996; Loy, Schofield, Turner, & Kwok, 2014; Sando et al., 2008; Vincent & Velkof, 2010; Wang, Xu, & Pei, 2012; Wilson et al., 2013). Cardiovascular factors, such as high cholesterol, high blood pressure, and heart disease, are known to contrib- ute to dementia risk (Diniz, Butters, Albert, Dew, & Reynolds, 2013). However, what is less established in the current litera- ture are definitive conclusions on the relationships between the contribution of other factors such as depression, anxiety, and sleep disturbance, and eventual AD development. There is also limited information on whether these factors are symp- toms of AD or if they contribute directly to the development of AD.

To date, depression has not been causally connected to AD, though many studies have yielded results that demon- strate a positive relationship between the two conditions (Andersen, Lolk, Kragh-Sorensen, Petersen, & Green, 2005;

Burke et al., 2016a, 2016b; Chen, Ganguli, Mulsant, & DeKosky, 1999; Gracia-Garcia et al., 2015; Li, Meyer, & Thornby, 2001; Steffens et al., 2004). Contradictory evidence about the role of depression remains. Findings from Ownby, Crocco, Acevedo, John, and Loewenstein (2006) suggest depression may be a prodromal symptom to AD, as opposed to a suggested cause. Furthermore, findings from Gatz, Tyas, St. John, and Mont- gomery (2005) indicate that depression may act as a predictor of AD development among individuals over the age of 65. Psychosocial factors, including depression symptoms, have been linked to future cognitive decline and development of AD (Burke et al., 2016; Green et al., 2003; Kessing & Nilsson, 2003; Speck et al., 1995; Yaffe et al., 1999).

Some studies have suggested amyloid deposits and atro- phy in frontal brain regions are related and contribute to AD development in a linear or even synchronous fashion (Becker et al., 2011; Ch�etelat et al., 2010; Oh, Habeck, Madison, & Jag- ust, 2014). Very low levels of dysphoria, apathy, and anhedo- nia may be indicators to neurodegeneration in areas of the brain that are associated with AD, but this may be indepen- dent of amyloid burden (Donovan et al., 2015). Further, patients in a three-year prospective cohort study experiencing mild cognitive impairment and depression were at twice the risk of developing dementia (Alzheimer’s type) than those without depression. Of the depressed patients studied, 85% developed dementia as opposed to 32% of the non- depressed patients (Modrego & Ferr�andez, 2004).

CONTACT Shanna L. Burke [email protected]

© 2017 Informa UK Limited, trading as Taylor & Francis Group

https://doi.org/10.1080/13607863.2017.1387760

AGING & MENTAL HEALTH, 2018 VOL. 22, NO. 12, 1577–1584

Psychological distress, such as that experienced by individ- uals with anxiety, may also be associated with an increased risk of AD development (Russ, Hamer, Stamatakis, Starr, & Batty, 2011). Russ et al. (2011) indicated that higher levels of psychological distress may be correlated with AD-related deaths. The study results remained the same even when other risk factors were taken into account, such as age, diabetes, smoking, and cardiovascular disease. Recently, Pietrzak et al. (2015) found that elevated anxiety symptoms (measured by the Hospital Anxiety and Depression Scale) moderated the effect of amyloid-b on cognitive decline in a preclinical AD study sample (n = 333). That is, the presence of anxiety symp- toms led to more rapid decline in verbal memory, language, and executive function. These domains suggest that impairment first occurred in the temporal and prefrontal corti- cal regions.

In addition, poor sleeping habits may also be a contribut- ing factor to the risk of AD development (Burke et al., 2016; Xie et al., 2013). This risk may be further increased by homozy- gous or heterozygous APOE-e4 carrier status, which is known to increase risk of AD in a dose-dependent fashion (Yoshizawa et al., 1994). Xie et al. (2013) found that cerebrospinal and interstitial fluid exchange increases substantially during sleep, which aids in b-amyloid protein removal. However, when sleep disturbance occurs (defined by less than five hours of sleep or frequent night fits), Ab-amyloid deposition may increase, which may exacerbate the potential to develop AD (Spira et al., 2013). Lim et al. (2013) indicated that APOE-e4 carriers with sleep disturbance may also have an increased chance of developing AD compared to e4 carriers who sleep for the recommended duration of 7–8 hours a night. It has not been definitively concluded that the sleep disturbance instigates AD development, but the aforementioned results provide a platform for further investigation into this correla- tion. Given the continued debate in the scientific literature, further investigation of the role of psychosocial risk factors in predicting AD risk in a cognitively normal population is warranted.

It is important to not only view the concepts of depression, anxiety, and sleep disturbance as distinct and independent risk factors for AD, but to consider their interacting effect in AD development. For this purpose, the authors selected the syndemic perspective as the theoretical framework guiding the design and analysis of this study. While comorbidity is defined as the presence of multiple, but usually independent, diseases or disorders, a syndemic perspective involves an interaction between two or more diseases or disorders that result in additional negative health consequences (Singer, 2009). However, the syndemic perspective reaches beyond a person’s biology, and takes account of stress, inequality, the community, and the environment, all over time, as potential cofactors in the exacerbation of illnesses (Singer, 2009). The syndemic perspective has often been used in the study of medical problems, such as HIV/AIDS and its co-infections like tuberculosis or malaria. A search of the Medline and PsycInfo databases provided no evidence that the syndemic has been used to explain AD development, suggesting that this may be a novel application of the syndemic perspective.

This study hypothesized that the presence of sleep distur- bance, depression, and anxiety as individual factors increases the likelihood of meeting the criteria for AD diagnosis. It was also hypothesized that the synergistic effect (examined through additive effects) of two psychological factors, that

induce stress, will increase the hazard of AD development beyond the hazard of possessing the presence of one psycho- social effect alone.

Methods

This study examined psychosocial risk factors for late-onset AD through a secondary data analysis of the National Alz- heimer’s Coordinating Center’s (NACC) Uniform Data Set (UDS). A prospective cohort design was utilized with individu- als who were cognitively asymptomatic at the time of their initial visit (n = 12,053). The analysis sought to determine who among those initially unaffected demonstrated clinical signs of AD dementia by their last visit.

Participants

The participants volunteered from one of the 34 AD Centers (ADCs) in the United States. Data was collected between Sep- tember 2005 and September 2015. Participants with demen- tia, mild cognitive impairment, and cognitively intact individuals were recruited by individual ADCs throughout the United States (n = 33,610). The UDS is not a nationally repre- sentative sample of the United States population with respect to AD or dementia, as these participants self-select and volun- tarily present for an examination at one of the participating NIH/NIA sponsored ADCs. Written informed consent was obtained from all participants and/or self-designated inform- ants by the individual ADC site (NACC, 2010). Participants eli- gible for this study were cognitively asymptomatic at visit one (n = 12,053). Of this initial sample, 9,184 participants were eli- gible for analysis as a result of their participation in the origi- nal examination and at least one follow-up visit. Therefore, all participants had at least two total visits to the ADC. Partici- pants with more than one dementia-type diagnosis were excluded. This study received approval from the [blinded] University Institutional Review Board.

Measures

Data are collected particiants or provided by a trusted infor- mant by trained clinicians. These interviews acquire a wide range of demographic information; family history; medica- tions used; health history; a physical; and imaging and labs. Participants respond to several neuropsychological rating scales, functional assessments, and from these measures a diagnosis regarding dementia status is determined through the Clinical Rating Scale (Morris, 1993). Probable AD was diag- nosed within the UDS using criteria set forth by the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer’s Disease and Related Disorders Association (McKhann et al., 1984, 2011). Diagnoses arising from these interviews are assigned either by a consensus team or the examining physician (NACC, 2010). Individuals that began the study as cognitively normal, but were diag- nosed with AD at a later visit are defined in this study as ‘con- verters’ to AD; their counterparts that did not develop AD during the course of the study were defined as ‘non-convert- ers’ (Table 2).

The variables utilized for this study include self-reported depression in the last two years, other episodes of depression before the two-year timeframe, clinician-verified depression at the time of visit, self-reported sleep disturbance, and self-

1578 S. L. BURKE ET AL.

reported anxiety. Sporadic late-onset AD (probable AD) is the outcome of interest. Three depression measurements were used in the study. The first variable included active depression in the last two years self-reported by the participant. This includes depressive disorders for which a clinician was con- sulted, even if treatment or medication was not received. Depression includes major depressive disorder, situational depression, bipolar disorders, dysthymic disorders, and other mood disorders. The second measure of depression, depres- sion: other episodes, includes self-reported depression epi- sodes prior to the last two years and is considered to be the measure of lifetime depression. The third depression variable is the clinician’s judgment of symptoms and measures the presence or absence of depression at the time of the visit. The Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association, 2013) was utilized to inform the clinician’s diagnosis of depression.

Sleep disturbance was measured by the self-reported pres- ence or absence of nighttime behaviors. These behaviors include awakening during the night, rising too early in the morning, or taking excessive naps during the day. If any of the three symptoms were reported to occur in the past 30 days, participants were considered to have a sleep distur- bance. Anxiety was also measured by self-report, specifically whether the person in question was ‘very nervous, worried, or frightened for no apparent reason’ and seemed ‘very tense or fidgety’ or is ‘afraid to be apart from’ the informant (Cum- mings, 2009, p. 12). The variables of anxiety and sleep distur- bance were measured utilizing different questions of the Neuropsychiatric Inventory Questionnaire (NPI-Q), which measures the presence or absence of each individual psychi- atric symptom (Cummings, 1977/2009). The NPI-Q has dem- onstrated adequate test–retest reliability and convergent validity for the individual symptomology with which the cur- rent study is concerned (Kaufer et al., 2000).

Survival analysis was used to estimate and interpret hazard functions (Kleinbaum & Klein, 2012). The primary goal was to assess the effect of psychosocial predictor variables on the eventual diagnosis of AD dementia. An event (outcome vari- able) was defined as a diagnosis of probable AD among indi- viduals with normal cognition at baseline; a diagnosis of probable AD was the failure point and the participant’s remaining observations were removed from the study once AD diagnosis was delivered. Right censoring (Kleinbaum & Klein, 2012) was utilized to account for the fact that a partici- pant was not necessarily diagnosed with AD prior to their last observation. As a result, true survival time is unobserved unless a participant is diagnosed with AD on or prior to their last recorded observation. In order to exclude missing values from the analysis, but retain all other information from a par- ticipant’s visit, missing information was recoded and excluded in a listwise fashion.

The statistical software, STATA (StataCorp, Release 14, 2015), was utilized for the analyses, and a p-value of <0.05 was considered statistically significant for the analysis. Partici- pants’ observations ranged from 1 to 10 visits. Observation intervals were measured in days to account for the staggered nature of the visits.

A descriptive analysis of the baseline sample was con- ducted and included distributions across predictor variables, percentages for categorical variables, and means and stan- dard deviation for continuous variables. Baseline survival function was determined prior to the addition of predictors

and covariates (age, race, education, e4 carrier status, the presence of hypertension and hypercholesterolemia) in the model. We also controlled for APOE as 30% of the sample were e4 carriers. Log-rank tests for equality were used to test for significant differences in survival curves between the vari- ous response categories within each predictor variable (i.e. those reporting ‘yes’ to anxiety versus those reporting ‘no’ to the same question). Chi-square and t-tests were conducted to assess whether there was a statistically significant difference between those who eventually developed AD and those who did not with respect to demographic and predictor variables.

The relationships of certain predictor variables were exam- ined relative to the outcome variable using the Cox propor- tional hazards model (Cox, 1972). The time variable used for the cox analysis was number of days from the first visit until the first occasion of AD diagnosis. Regression modeling included simultaneous control of multiple predictors and covariates. Four models were developed to explore the main effects of the predictor variables. In the first model, unad- justed main effects of each individual predictor were exam- ined. In the second model, the covariates, sex, age, education, and race were controlled. In the third model, probable AD was examined in relation to the aforementioned controlled variables with the addition of APOE genotypes. In the final model, all previous covariates were controlled, including the presence of hypertension and hypercholesterolemia. A similar structure was applied to the exploration of synergistic effects. The assumption of proportionality was examined through inspection of Schoenfeld residuals in order to determine whether the proportional hazards assumption had been met.

Results

The minimum amount of time participants were under obser- vation was 208 days until the first occasion of AD diagnosis. The maximum observation interval was 3458 days (M = 1549.3; Mdn: 1456 days). The mean number of visits was 3.26 (SD: 2.12), with a range of 1 to 10 visits. There were 361 diag- noses of AD dementia by the end of the observation period among older adults who presented for at least an initial visit and at least one follow-up visit. The mean age of subjects with normal cognition at visit one was 71 years (SD: 10.86 years). At visit one, 80.5% of the sample population were White, 13.55% were African American, and 5.95% were from other ethnic groups. Almost 6% of the sample reported His- panic origin. Approximately 35% of subjects reported that their mother had been diagnosed with dementia, while 16.45% reported that their father had been diagnosed with dementia. Almost 18% of subjects reported depression in the last two years or lifetime depression, 10.05% of subjects were diagnosed with clinician-verified depression, 10.55% reported a sleep disturbance, and 8.69% reported the presence of anxi- ety. Percentages, means, and standard deviations (where applicable) are displayed in Table 1.

Preliminary log-rank tests for equality of survivor functions revealed that those who reported depression in the last two years, clinician verified depression at the time of the visit, sleep disturbance, and anxiety expressed statistically signifi- cant (p < 0.001) different survival curves than those who did not. The same was true for those reporting lifetime depression versus those who did not (p < 0.05). Chi-square and t-tests examined differences between participants who ultimately were diagnosed with AD and those who did not demonstrate

AGING & MENTAL HEALTH 1579

a statistically significant difference on all psychosocial predic- tors and most demographic factors. There was not a signifi- cant difference between converters and non-converters with regard to race.

There was a significant association (p< 0.001) between individuals reporting depression in the last two years and the occurrence (diagnosis) of AD. Specifically, in model one, those who self-reported depression in the last two years experi- enced a significantly higher risk of AD dementia diagnosis (HR = 2.32 [95% CI, 1.87–2.88]) compared to those who had not. The hazard remained stable even when adjusted for covariates in model two, adjusting further for APOE-e4 carrier status, and finally adjusting for high blood pressure and cho- lesterol in the fourth model.

The main effect of lifetime depression episodes, occurring more than two years earlier, presented a significant risk of AD diagnosis within the follow-up period compared to those who did not report such episodes (HR = 1.32 [1.04–1.68], p < .05). Similarly, depression verified by a clinician was signifi- cantly associated with the diagnosis of AD during the follow- up period as compared to those without verified depression symptoms in model one (HR = 2.72 [2.15–3.43], p < .001), and remained relatively similar when the effect of APOE-e4 carrier status was controlled as well as demographic confounders (HR = 2.89 [2.24–3.72], p < .001).

The presence of sleep disturbance was also significantly associated with the eventual diagnosis of AD. The unadjusted hazard (HR = 2.86 [2.25–3.63], p < .001) was similar to that of clinician-verified depression (HR = 2.72 [2.15–3.43], p < .001).

Finally, anxiety symptoms were significantly associated with an increased hazard of eventual AD development (HR = 3.50 [2.77–4.44], p < .001), presenting the strongest associa- tion out of the aforementioned psychosocial predictors. Main effects for all primary predictors are displayed in Table 3.

Additive interactions

The hazard of eventual AD diagnosis for those experiencing both recent depression symptoms and sleep disturbance was statistically significant (HR = 4.95 [95% CI 3.53–6.94], p < .001), as compared to those experiencing neither symptom. The effect of lifetime depression and sleep disturbance in rela- tion to AD indicated a strong positive relationship (HR = 3.26Ta

bl e 1.

D em

og ra ph

ic ov er vi ew

of sa m pl e an d pr ed ic to rv ar ia bl es .

D ep re ss io n – la st tw o ye ar s

D ep re ss io n – lif et im e

D ep re ss io n – cl in ic ia n ve rifi

ed Sl ee p di st ur ba nc e

An xi et y

At vi si to

ne – ba se lin e

Pr es en t

Ab se nt

Pr es en t

Ab se nt

Pr es en t

Ab se nt

Pr es en t

Ab se nt

Pr es en t

Ab se nt

N or m al co gn

iti on

12 ,0 83

21 63

(1 7. 97 % )

98 71

(8 2. 03 % )

21 60

(1 8. 09 % )

97 82

(8 1. 91 % )

12 14

(1 0. 05 % )

10 ,8 69

(8 9. 95 % )

11 80

(1 0. 55 % )

10 ,0 00

(8 9. 45 % )

97 1 (8 .6 9%

) 10 ,2 09

(9 1. 31 % )

Fe m al e

7, 86 5 (6 5. 09 % )

15 86

(2 0. 17 % )

62 45

(7 9. 40 % )

15 90

(2 0. 22 % )

61 83

(7 8. 61 % )

90 8 (1 1. 54 % )

69 57

(8 8. 46 % )

73 4 (9 .3 3%

) 65 42

(8 3. 18 % )

64 4 (8 .1 9%

) 66 32 (8 4. 32 % )

Ag e

X = 71 .0 5 (S D = 10 .8 6)

X = 68 .6 5 (S D = 10 .7 8)

X = 68 .1 4 (S D = 10 .5 9)

X = 68 .5 9 (S D = 10 .7 9)

X = 72 .1 3 (S D = 11 .1 8)

X = 72 .2 4 (S D = 11 .4 3)

Ra ce W hi te

96 53

(8 0. 5%

) 18 65

(1 9. 32 % )

77 51

(8 0. 30 % )

18 46

(1 9. 12 % )

76 99

(7 9. 76 % )

10 01

(1 0. 37 % )

86 52

(8 9. 63 % )

99 4 (1 0. 30 % )

79 44

(8 2. 30 % )

82 1 (8 .5 1%

) 81 17

(8 4. 09 % )

Af ric an

Am er ic an

16 25

(1 3. 55 % )

17 1 (1 0. 52 % )

14 88

(9 1. 57 % )

18 5 (1 0. 02 % )

14 18

(8 7. 26 % )

11 1 (6 .8 3%

) 15 14

(9 3. 17 % )

10 3 (6 .3 4%

) 13 97

(8 5. 97 % )

80 (4 .9 2%

) 14 20

(8 7. 38 % )

H is pa ni c

71 4 (5 .9 3%

) 16 6 (2 3. 24 % )

54 5 (7 6. 33 % )

13 5 (1 9. 12 % )

57 1 (8 0. 88 % )

12 8 (1 7. 93 % )

58 6 (8 2. 07 % )

11 2 (1 7. 5%

) 52 8 (8 2. 5%

) 66

(1 8. 86 % )

28 4 (8 1. 14 % )

O th er

71 3 (5 .9 5%

) 10 4 (1 4. 58 % )

60 4 (8 4. 71 % )

10 2 (1 4. 31 % )

60 2 (8 4. 43 % )

77 (1 0. 80 % )

63 6 (8 9. 20 % )

75 (1 0. 52 % )

59 1 (8 2. 89 % )

59 (8 .2 7%

) 60 7 (8 5. 13 % )

N ot e: Ro w pe rc en ta ge s ar e di sp la ye d.

Table 2. Demographic overview of converters and non-converters. Converters to AD

(n = 712) Non-Converters (n = 1039) t or x2 Statistica

Age (years) 84.64 (SD:8.49)b 81.93 (SD: 9.44) –6.18, df = 1763, p = 0.00

Female 476 (62.58%)c 642 (57.42%) 4.27, df = 1, p = 0.039 Education (years) 16.36 (SD: 4.66) 15.62 (SD: 9.84) –2.12, df = 1763, p =

0.03 Race 1.77, df = 2, p = 0.412

White 614 (86.24%) 872 (83.93%) African-American 75 (10.53%) 129 (12.42%) Other 23 (3.23%) 38 (3.66%) Hispanic 41 (5.75%) 28 (2.69%) 10.42, df = 1, p = 0.001

e4 Carrier 277 (38.90%) 289 (27.82%) 25.69, df = 1, p = 0.00 Depressed 2 Years 277 (38.90%) 275 (26.47%) 31.09, df = 1, p = 0.00 Depression – Lifetime

206 (28.93%) 238 (22.91%) 8.68, df = 1, p = 0.003

Clinician-Verified Depression

205 (28.79%) 210 (20.21%) 16.86, df = 1, p = 0.00

Anxiety 196 (27.53%) 163 (15.69%) 38.64, df = 1, p = 0.00 Sleep Disturbance 185 (25.98%) 163 (15.69%) 30.00, df = 1, p = 0.00 a x2 test statistics are displayed for categorical variables, t test statistics for continuous variables.

b Continuous variables are described with mean and standard deviation. c Categorical variables are described with sample size and percentage.

1580 S. L. BURKE ET AL.

[2.24–4.75] p < 0.001), compared to those who did not endorse these symptoms. Clinician-verified depression and sleep disturbance was also strongly correlated (p < 0.001) with eventual AD diagnosis; yielding a hazard three times greater than those without these symptoms (HR = 3.82 [2.47– 5.90]). The findings for additive effects of those reporting anxi- ety symptoms are similar to participants with recent depres- sion (HR = 5.09 [3.68–7.03], p < .001), lifetime depression (HR = 3.41 [2.30–5.05], p < .001), and clinician-verified depression (HR = 4.14 [2.86–6.01], p < .001) as compared to those who do not report anxiety nor depression symptoms. The additive effect of sleep disturbance and anxiety was also statistically significant (HR = 4.67 [3.23–6.75], p < .001), compared to those without sleep disturbance nor anxiety. These figures are displayed in Table 4.

Discussion

This study examined correlations between depression, sleep disturbance, and anxiety with eventual AD dementia develop- ment, both as individual factors as well as comorbid condi- tions. Previous empirical literature on the association between mental health symptoms and AD and/or dementia focuses on the synchronous relationship between the two. Specifically, studies have examined the correlation between the level of psychiatric symptoms and dementia stage (Lopez et al., 2003; Verkaik, Nuyen, Schellevis, & Francke, 2007), the incidence of psychiatric comorbidity (Steinburg et al., 2008), and psychiatric symptoms impacting dementia caregiver bur- den and the risk of subsequent nursing home placement (Buhr, Kuchibhatla, & Clipp, 2006).

Few studies have focused on psychological symptoms, such as depression (Richard et al., 2013), anxiety, and sleep disturbance as independent and combined predictors of AD dementia. Similar studies often sample or examine an already impaired population, tracking the progression from MCI to dementia to understand the association between psychiatric

symptoms and progression rate (Gabryelewicz et al., 2007). In these studies, the assumptions and hypotheses indicate that psychosocial or psychological symptoms are behavioral mani- festations or challenging behaviors associated with dementia as opposed to independent mental health symptoms. Symp- toms which precede dementia diagnosis are often assumed to be prodromal symptoms and perhaps a signal of advancing AD pathology (Jorm, 2001).

An estimated 50–70 million Americans suffer chronically from a sleep disorder, impacting their health across daily functioning and lifetime longevity. The Centers for Disease Control and Prevention (2015) reports that 10% of the US population reports chronic insomnia, which is just one facet of the sleep disturbance described in the current study. There are approximately 90 unique sleep disorders, many of which share the following symptoms ‘excessive daytime sleepiness, difficulty initiating or maintaining sleep, and abnormal events occurring during sleep’ (Naismith et al., 2011). In the fifth edi- tion of the Diagnostic and Statistical Manual (DSM-V), the char- acteristics of various sleep disorders were taken into consideration and organized into 11 sleep–wake disorder diagnostic groups (American Psychiatric Association, 2013). Likewise, diagnostic indicators of depression as outlined in the DSM-V include the following: depressed mood daily or nearly daily, decreased interest in activities most of the day, significant weight and appetite change, change in sleep, change in activity, fatigue, guilt/worthlessness, diminished capacity for concentration, and thoughts of death or suicide (American Psychiatric Association, 2013). Sleep disturbance among the aging may be a prognostic indicator for decline in cognitive functioning. Among older adults, sleep disturbance may predict depression. Sleep disturbance in depressed patients can represent the presence of a residual mood disor- der. Its presence can linger and may be a precursor for a later depressive episode (Lee et al., 2013).

A recent study of Manhattan residents (n = 2160) over the age of 65 (Richard et al., 2013) found depression to be

Table 3. Main effects (Probable Alzheimer’s Disease Dementia as Outcome Variable).

Predictor Model 1 Hazard Ratio (95% CI) Model 2 Hazard Ratio (95% CI) Model 3Hazard Ratio (95% CI) Model 4Hazard Ratio (95% CI)

Depression – last 2 years 2.32 (1.87–2.88)** 2.51 (2.02–3.12)** 2.53 (2.00–3.21)** 2.53 (1.99–3.21)** Depression – lifetime 1.32 (1.04–1.68)* 1.50 (1.18–1.90)** 1.53 (1.19–1.98)** 1.53 (1.19–1.98)** Depression – clinician verified 2.72 (2.15–3.43)** 2.67 (2.10–3.37)** 2.89 (2.24–3.72)** 2.88 (2.23–3.71)** Sleep disturbance 2.86 (2.25–3.63)** 2.57 (2.01–3.27)** 2.57 (1.98–3.34)** 2.56 (1.97–3.33)** Anxiety 3.50 (2.77–4.44)** 3.25 (2.56–4.12)** 3.17 (2.45–4.10)** 3.16 (2.44–4.09)**

Model 1: Main effect unadjusted. Model 2: Main effects adjusted for sex, education, age, and race. Model 3: Main effects adjusted for adjusted for sex, age, education, race, e4 carrier status. Model 4: Main effects adjusted for adjusted for sex, age, education, race, e4 carrier status, the presence of hypertension and hypercholesterolemia. * indicates statistical significance at p < 0.05, ** indicates p < 0.001.

Table 4. Additive effects among psychosocial predictor variables (Probable Alzheimer’s disease as outcome variable).

Predictor Model 1Hazard ratio

(95% CI) Model 2 Hazard ratio

(95% CI) Model 3Hazard ratio

(95% CI) Model 4Hazard ratio

(95% CI)

Depressed 2 years + sleep disturbance 4.95 (3.53–6.94)** 4.81 (3.42–6.75)** 4.66 (3.22–6.75)** 4.65 (3.21–6.74)** Depression lifetime + sleep disturbance 3.26 (2.24–4.75)** 3.24 (2.22–4.74)** 3.32 (2.20–5.00)** 3.31 (2.19–5.00)** Clinician-verified depression + sleep disturbance 3.82 (2.47–5.90)** 3.76 (2.43–5.81)** 4.13 (2.58–6.59)** 4.10 (2.56–6.55)** Sleep Disturbance + anxiety 4.67 (3.23–6.75)** 4.27 (2.94–6.21)** 4.21 (2.79–6.35)** 4.19 (2.77–6.32)** Depressed 2 years + anxiety 5.09 (3.68–7.03)** 5.09 (3.67–7.05)** 4.75 (3.31–6.80)** 4.75 (3.31–6.82)** Depression lifetime + anxiety 3.41 (2.30–5.05)** 3.47 (2.34–5.15)** 3.26 (2.10–5.07)** 3.26 (2.10–5.08)** Clinician-verified depression + anxiety 4.14 (2.86–6.01)** 4.03 (2.78–5.85)** 4.28 (2.88–6.38)** 4.26 (2.86–6.35)**

Model 1: Main effect unadjusted Model 2: Main effects adjusted for sex, education, age, and race. Model 3: Main effects adjusted for adjusted for sex, education, age, race, and e4 carrier status. Model 4: Main effects adjusted for adjusted for sex, education, age, race, e4 carrier status, the presence of hypertension and hypercholesterolemia. * indicates statistical significance at p < 0.05, ** indicates p < 0.001.

AGING & MENTAL HEALTH 1581

associated with an increased risk of dementia diagnosis, but that depression did not precede dementia development. Alternatively, a study focusing on diabetes and depression, both as independent and combined factors, found depression to be independently associated with risk of dementia devel- opment; the hazard of which was further increased for people diagnosed with diabetes. The study included a wide age range (30–75) and a large sample (n = 19,239). These investi- gations suggest that ongoing research is examining the rela- tionships between depression and AD but not yet conclusive on how depression may act as an independent risk factor of AD. Similarly, anxiety was found to be a risk factor dementia in men (n = 1481) 17 years after baseline anxiety assessments (Gallacher et al., 2009).

Furthermore, a recent study found that daytime sleepiness was associated with increased incidence of dementia (Tsapa- nou, Gu, Scarmeas, & Stern, 2015). In a study of Swedish men followed from age 50 to 88 between 1970 and 2010 for the Uppsala Longitudinal Study of Adult Men, researchers found self-reported sleep disturbance (as evidenced by difficulty fall- ing asleep, inability to fall back asleep if waking too early, or the use of sleeping pills three or more times per week) increased the risk of dementia development (+33%) and AD development (+51%; Benedict et al., 2015). Using data from the survey of Health, Aging and Retirement in Europe, researchers found that self-reported sleep disturbance measured as ‘sleeping problems in the past 6 months’, ‘recent trouble sleeping or change in pat- tern’, and ‘restless sleep’ were also significantly related to the development of dementia (Theou, Rusak, & Rockwood, 2004). Finally, Jorm’s (2001) comprehensive meta-analysis indicate that affective symptomology, such as depression, anxiety, and sleep disturbance are worthy risk factors for exploration in rela- tion to AD risk and development.

The findings from the current investigation support the hypothesis that the presences of psychosocial factors of sleep disturbance, depression, and anxiety as individual factors (which may induce stress), increasing the likelihood ofmeeting the crite- ria for AD diagnosis. When these factors were combined to cre- ate hypothetical scenarios in which a participant experienced two of these symptoms simultaneously, the synergistic effect remained significant, growing beyond the simple sum of the main effects when covariates were taken into account. These finding provide partial support for the study’s second hypothesis; that the co-occurrence of two psychosocial factors will increase the hazard of AD development beyond the hazard of possessing the presence of one psychosocial effect alone. The mechanism behind such increasing risk, however, is unknown. This, in turn, provides evidence of a syndemic framework or perspective in which the interaction between some of the factors of interest in this study exacerbates AD development.

The strength of this novel study is the inclusion of psycho- social symptoms and cognitive status related to the onset and progression of AD. All participants in this study were determined by a physician or a consensus team to possess normal cognition at visit one. The symptoms explored in this study were reported at baseline.

This study has several limitations. It cannot be confirmed without neuroimaging results that the subjects did not pos- sess underlying AD pathology. It is unlikely that they would be diagnosed with ‘normal cognition’ and not some form of cognitive impairment if this was the case, however, previous studies have shown that up to a third of subjects categorized as not-demented (Storandt, Grant, Miller, & Morris, 2006) or

with pre-MCI (Storandt et al., 2006) during their lifetimes dem- onstrate AD pathology at autopsy (Storandt et al., 2006; Jack et al., 2002). In addition, we now know that AD pathology likely begins developing 10–20 years prior to any observable manifestations (Sperling et al., 2011). Future studies of this nature would benefit from confirmation of normal cognition through neuroimaging at baseline. Also, the sample was not randomly selected, but it was comprised of individuals volun- teering themselves to at least two visits: the initial visit plus one follow-up. Therefore, the results of this study cannot be generalized to the aging population at large, though it pro- vides preliminary evidence that could inform future research with a nationally representative sample. This study also did not compare cognitively normal participants with at least two visits to those with less than two visits (also known as those not eligible for survival analysis).

Additionally, the measures used in the UDS measure depression in three ways – past two years, depression in the lifetime outside of the past two years, and clinician verified depression at the time of visit. Lifetime depression is difficult to interpret without additional treatment data, and it does not distinguish between a chronic mood disorder and an acute episode of depression. The measures also do not define depression by more explicit DSM diagnoses; for example bipolar depressive episodes are not distinguished from clini- cal depression or dysthymia. Finally, while this study gives insight into the relationships between self-reported depres- sion, lifetime depression, clinician verified depression at the time of visit, anxiety, sleep disturbance, and AD development, more research is needed to refine the causal pathways. Specif- ically, utilizing the syndemic perspective requires further investigation with societal, political, and economic inequality factors that may be relevant to survival modeling.

The total estimated worldwide costs of Alzheimer’s and dementia in 2015 was $818 billion in US dollars (Prince et al., 2015). The World Alzheimer’s Institute estimates that total dementia costs will reach $1 trillion worldwide per year by 2018 and $2 trillion per year worldwide by 2030 (Prince et al., 2015). In the United States, total health care, long-term care, and hospice care costs for AD in 2016 are estimated to be $236 billion, with Medicare and Medicaid paying for approxi- mated 68% of these costs (Alzheimer’s Association, 2016). Without any changes, Alzheimer’s costs are expected to sur- pass $1 trillion in the United States alone by 2050, presenting a nearly 500% increase in Medicare and Medicaid spending (Alzheimer’s Association, 2016).

The Alzheimer’s Association (2015b) estimates that if treat- ment providing a delay in AD diagnosis becomes available by 2025, costs of the disease have the potential to drastically decrease over the subsequent five years. Within the first year alone, savings could amount to about $3 billion in Medicare costs and by 2035 this amount could increase to $67 billion (Alzheimer’s Association, 2015b). Even further, they predict that Medicaid savings could amount to $1 billion and $38 billion by 2035 (Alzheimer’s Association, 2015b). Cumulatively, these gov- ernment-funded programs could have a total savings of $535 billion over a 10-year period and $935 billion for payer sources that are not government funded (Alzheimer’s Association, 2015b). Additionally, the out-of-pocket expenditures for patients and their families have the potential to decrease by $2 billion by 2026 (Alzheimer’s Association, 2015b).

Given the current state of knowledge regarding possible risks of AD development, it is essential to identify and develop

1582 S. L. BURKE ET AL.

ways to reduce occurrence before the pathophysiological dis- ease progression begins. This may be beneficial to those at risk but also to the national economy and overall healthcare costs. The Centers for Disease Control and Prevention’s Syn- demics Prevention Network suggest the use of system dynamics models to identify the course and development of health ailments, then to use those models to create a more balanced system of health protection (Milstein, 2008). In this scenario, screening and treatment for depression, sleep dis- turbance and anxiety are relatively low-cost and potentially preventive measures to be considered in a more holistic approach to AD prevention. This point of view for treating AD broadens the focus from the individual’s biology at the moment of AD diagnosis to the individual within his or her environment over time before AD development. Seeing these factors as modifiable ailment ties in a syndemic system, versus comorbidities or prodromal symptoms of AD, may provide an avenue of treatment that delays the onset of AD development.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Shanna L. Burke http://orcid.org/0000-0001-6969-3536 Tamara Cadet http://orcid.org/0000-0002-4660-8009

References

Albert, M. S., DeKosky, S. T., Dickson, D., Dubois, B., Feldman, H. H., Fox, N., … Phelps, C. H. (2011). The diagnosis of mild cognitive impairment due to Alzheimer’s disease: Recommendations from the National Insti- tute on Aging–Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimer’s & Dementia, 7(3), 270– 279.

Alzheimer’s Association. (2015a). 2015 Alzheimer’s disease facts and fig- ures. Retrieved from https://www.alz.org/facts/downloads/facts_fig ures_2015.pdf.

Alzheimer’s Association. (2015b). Changing the trajectory of Alzheimer’s disease: How a treatment by 2025 saves lives and dollars. Retrieved from https://www.alz.org/documents_custom/trajectory.pdf.

Alzheimer’s Association. (2016). 2016 Alzheimer’s disease facts and fig- ures. Retrieved from http://www.alz.org/documents_custom/2016- facts-and-figures.pdf.

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Washington, DC: Author.

Andersen, K., Lolk, A., Kragh-Sørensen, P., Petersen, N. E., & Green, A. (2005). Depression and the risk of Alzheimer disease. Epidemiology, 16 (2), 233–238.

Becker, J. A., Hedden, T., Carmasin, J., Maye, J., Rentz, D. M., Putcha, D., … Johnson, K. A. (2011). Amyloid-b associated cortical thinning in clini- cally normal elderly. Annals of Neurology, 69(6), 1032–1042.

Benedict, C., Byberg, L., Cedernaes, J., Hogenkamp, P. S., Giedratis, V., Kilander, L., … Schioth, H. B. (2015). Self-reported sleep disturbance is associated with Alzheimer’s disease risk in men. Alzheimer’s & Demen- tia, 11, 1090–1097.

Buhr, G. T., Kuchibhatla, M., & Clipp, E. C. (2006). Caregivers’ reasons for nursing home placement: Clues for improving discussions with fami- lies prior to the transition. The Gerontologist, 46(1), 52–61.

Burke, S. L., Maramaldi, P., Cadet, T., & Kukull, W. (2016a). Associations between depression, sleep disturbance, and apolipoprotein E in the development of Alzheimer's disease: dementia. International Psycho- geriatrics, 28(9), 1409–1424.

Burke, S. L., Maramaldi, P., Cadet, T., & Kukull, W. (2016b). Neuropsychiatric symptoms and Apolipoprotein E: Associations with eventual Alzheim- er's disease development. Archives of gerontology and geriatrics, 65, 231–238.

Centers for Disease Control and Prevention. (2015). Sleep and sleep disor- ders. Retrieved from http://www.cdc.gov/sleep/index.html.

Chen, P., Ganguli, M., Mulsant, B. H., & DeKosky, S. T. (1999). The temporal relationship between depressive symptoms and dementia: A commu- nity-based prospective study. Archives of General Psychiatry, 56, 261– 266.

Ch�etelat, G., Villemagne, V. L., Bourgeat, P., Pike, K. E., Jones, G., Ames, D., … Rowe, C. C. (2010). Relationship between atrophy and beta-amyloid deposition in Alzheimer disease. Annals of Neurology, 67(3), 317–324.

Cox, D. R. (1972). Regression models and life-tables. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 342: 187–220.

Cummings, J. L. (1977). The Neuropsychiatric Inventory: Assessing psycho- pathology in dementia patients. Neurology, 48, S10–S16.

Cummings, J. L. (2009). The neuropsychiatric inventory: Comprehensive assessment of psychopathology in patients with dementia. Retrieved from http://www.dementia-assessment.com.au/behavioural/npi.pdf

Di Marco, L. Y., Marzo, A., Mu~noz-Ruiz, M., Ikram, M. A., Kivipelto, M., Ruefe- nacht, D., … Frangi, A. F. (2014). Modifiable lifestyle factors in demen- tia: A systematic review of longitudinal observational cohort studies. Journal of Alzheimer’s Disease, 42(1), 119–135.

Diniz, B. S., Butters, M. A., Albert, S. M., Dew, M. A., & Reynolds, C. F. (2013). Late-life depression and risk of vascular dementia and Alzheimer’s dis- ease: Systematic review and meta-analysis of community-based cohort studies. The British Journal of Psychiatry, 202(5), 329–335.

Donovan, N. J., Hsu, D. C., Dagley, A. S., Schultz, A. P., Amariglio, R. E., Mor- mino, E. C., … Marshall, G. A. (2015). Depressive symptoms and bio- markers of Alzheimer’s disease in cognitively normal older adults. Journal of Alzheimer’s Disease, 46(1), 63–73.

Gabryelewicz, T., Styczynska, M., Luczywek, E., Barczak, A., Pfeffer, A., Androsiuk, W., … Barcikowska, M. (2007). The rate of conversion of mild cognitive impairment to dementia: Predictive role of depression. International Journal of Geriatric Psychiatry, 22(6), 563–567.

Gallacher, J., Bayer, A., Fish, M., Pickering, J., Pedro, S., Dunstan, F., … Ben- Shlomo, Y. (2009). Does anxiety affect risk of dementia? Findings from the Caerphilly Prospective Study. Psychosomatic Medicine, 71(6), 659–666.

Gatz, J. L., Tyas, S. L., St. John, P., & Montgomery, P. (2005). Do depressive symptoms predict Alzheimer’s disease and dementia ? Journal of Ger- ontology, 60(6), 744–747.

Gracia-Garcia, P., de la-Camara, C., Santabarbara, J., Lopez-Anton, R., Quin- tanilla, M. A., Ventura, T., … Lobo, A. (2015). Depression and incident Alzheimer disease: The impact of disease severity. The American Jour- nal of Geriatric Psychiatry, 23(2), 119–129.

Green, R. C., Cupples, L. A., Go, R., Benke, K. S., Edeki, T., Griffith, P. A., … Ferrer, L. A. (2002). Risk of dementia among white and African Ameri- can relatives of patients with Alzheimer disease. JAMA, 287(3), 329– 336.

Green, R. C., Cupples, L. A., Kurz, A., Auerbach, S., Go, R., Sandovnick, D., … Farrer, L. (2003). Depression as a risk factor for Alzheimer disease: The MIRAGE study. Archives of Neurology, 60, 753–759.

He, Y. L., Zhang, X. K., & Zhang, M. Y. (2000). Psychosocial risk factors for Alzheimer's disease. Hong Kong Journal of Psychiatry, 10(2), 2–7.

Jack, C. R., Dickson, D. W., Parisi, J. E., Xu, Y. C., Cha, R. H., O’brien, P. C., … Kokmen, E. (2002). Antemortem MRI findings correlate with hippo- campal neuropathology in typical aging and dementia. Neurology, 58 (5), 750–757.

Jorm, A. F. (2001). History of depression as a risk factor for dementia: An updated review. Australian and New Zealand Journal of Psychiatry, 35 (6), 776–781.

Kaufer, D. I., Cummings, J. L., Ketchel, P., Smith, V., MacMillan, A., Shelley, T., … DeKosky, S. T. (2000). Validation of the NPI-Q: A brief clinical form of the neuropsychiatric inventory. The Journal of Neuropsychiatry and Clinical Neurosciences, 12(2), 233–239.

Kessing, L. V., & Nilsson, F. M. (2003). Increased risk of developing demen- tia in patients with major affective disorders compared to patients with other medical illnesses. Journal of Affect Disorders, 73, 261–269.

Kleinbaum, D. G. & Klein, M. (2012). Survival analysis: A self-learning text 3rd ed. New York: Springer.

Lautenschlager, N. T., Cupples, L. A., Rao, V. S., Auerbach, S. A., Becker, R., Burke, J., … & Green, R. C. (1996). Risk of dementia among relatives of Alzheimer's disease patients in the MIRAGE study What is in store for the oldest old?. Neurology, 46(3), 641–650.

Lee, E., Cho, H. J., Olmstead, R., Levin, M. J., Oxman, M. N., & Irwin, M. R. (2013). Persistent sleep disturbance: A risk factor for recurrent depres- sion in community-dwelling older adults. Sleep, 36(11), 1685–1691. doi:10.5665/sleep.3128

AGING & MENTAL HEALTH 1583

Li, Y. S., Meyer, J. S., & Thornby, J. (2001). Longitudinal follow-up of depres- sive symptoms among normal versus cognitively impaired elderly. International Journal of Geriatric Psychiatry, 16, 718–727.

Lim, A. S., Yu, L., Kowgier, M., Schneider, J. A., Buchman, A. S., & Bennett, D. A. (2013). Modification of the relationship of the apolipoprotein E e4 allele to the risk of Alzheimer disease and neurofibrillary tangle den- sity by sleep. JAMA Neurology, 70(12), 1544–1551. doi:10.1001/ jamaneurol.2013.4215

Lopez, O. L., Becker, J. T., Sweet, R. A., Klunk, W., Kaufer, D. I., Saxton, J., … DeKosky, S. T. (2003). Psychiatric symptoms vary with the severity of dementia in probable Alzheimer’s disease. The Journal of Neuropsychi- atry and Clinical Neurosciences, 15(3), 346–353.

Loy, C. T., Schofield, P. R., Turner, A. M., & Kwok, J. J. (2014). Genetics of dementia. Lancet, 383, 828–840.

McKhann, G., Drachman, D., Folstein, M., Katzman, R., Price, D., & Stadlan, E. M. (1984). Clinical diagnosis of Alzheimer’s disease: Report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer’s disease. Neurol- ogy, 34(7), 939–944.

McKhann, G., Knopman, D. S., Chertkow, H., Hyman, B. T., Jack Jr C. R., Kawas, C. H., … Phelps, C. H. (2011). The diagnosis of dementia due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimer’s & Dementia, 7(3), 263–269. doi:10.1016/j.jalz.2011.03.005

Milstein, B. (2008). Hygeia’s constellation: Navigating health futures in a dynamic and democratic world. Atlanta, GA: The Centers for Disease Control and Prevention, Syndemics Prevention Network.

Modrego, P. J., & Ferr�andez, J. (2004). Depression in patients with mild cognitive impairment increases the risk of developing dementia of Alzheimer type: A prospective cohort study. Archives of Neurology, 61 (8), 1290–1293.

Morris, J. C. (1993). The Clinical Dementia Rating (CDR): Current version and scoring rules. Neurology, 43(11), 2412–2414.

Naismith, S. L., Rogers, N. L., Lewis, S. J. G., Diamond, K., Terpening, Z., Nor- rie, L., & Hickie, I. B. (2011). Sleep disturbance in mild cognitive impairment: Differential effects of current and remitted depression. Acta Neuropsychiatrica, 23(4), 167–172. doi:10.1111/j.1601- 5215.2011.00555.x

National Alzheimer’s Coordinating Center. (2010). The UDS study popula- tion. Retrieved from https://www.alz.washington.edu/WEB/study_pop. html

Oh, H., Habeck, C., Madison, C., & Jagust, W. (2014). Covarying alterations in Ab deposition, glucose metabolism, and gray matter volume in cog- nitively normal elderly. Human Brain Mapping, 35(1), 297–308. doi:10.1002/hbm.22173

Ownby, R. L., Crocco, E., Acevedo, A., John, V., & Loewenstein, D. (2006). Depression and risk for Alzheimer disease: Systematic review, meta- analysis, and metaregression analysis. Archives of General Psychiatry, 63(5), 530–538.

Pietrzak, R. H., Lim, Y. Y., Neumeister, A., Ames, D., Ellis, K. A., Harrington, K., … Villemagne, V. L. (2015). Amyloid-b, anxiety, and cognitive decline in preclinical Alzheimer disease: A multicenter, prospective cohort study. JAMA Psychiatry, 72(3), 284–291.

Prince, M., Wimo, A., Guerchet, M., Ali, G., Wu, Y., & Prina, M. (2015). World Alzheimer report 2015 of the global impact of dementia: An analysis of prevalence, incidence, costs and trends. Retrieved from http:// www.alz.co.uk/research/WorldAlzheimerReport2015.pdf.

Richard, E., Reitz, C., Honig, L. H., Schupf, N., Tang, M. X., Manly, J. J., … Luchsinger, J. A. (2013). Late-life depression, mild cognitive impairment, and dementia. JAMA Neurology, 70(3), 383–389.

Russ, T. C., Hamer, M., Stamatakis, E., Starr, J. M., & Batty, D. (2011). Psycho- logical distress as a risk factor for dementia death. Archives of Internal Medicine, 171(20), 1859.

Sando, S. B., Melquist, S., Cannon, A., Hutton, M., Sletvold, O., Saltvedt, I.,… Aasly, J. (2008). Risk-reducing effect of education in Alzheimer’s dis- ease. International Journal of Geriatric Psychiatry, 23(11), 1156–1162.

Singer, M. (2009). Introduction to syndemics: A critical systems approach to public and community health. San Francisco, CA: Jossey-Bass.

Speck, C. E., Kukull, W. A., Brenner, D. E., Bowen, J. D., McCormick, W. C., Teri, L. P., … Larson, E. B. (1995). History of depression as a risk factor for Alzheimer’s disease. Epidemiology, 6, 366–369.

Sperling, R. A., Aisen, P. S., Beckett, L. A., Bennett, D. A., Craft, S., Fagan, A. M., … Phelps, C. H. (2011). Toward defining the preclinical stages of Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimer’s & Dementia, 7(3), 280–292. doi:10.1016/j.jalz.2011.03.003

Spira, A. P., Gamaldo, A. A., An, Y., Wu, M. N., Simonsick, E. M., Bilgel, M., … Resnick, S. M. (2013). Self-reported sleep and b-amyloid deposition in community-dwelling older adults. JAMA Neurology, 70(12), 1537–1543.

StataCorp. (2015). Stata statistical software: Release 14. College Station, TX: StataCorp LP.

Steffens, D. C., Welsh-Bohmer, K. A., Burke, J. R., Plassman, B. L., Beyer, J. L., Gersing, K. R., & Potter, G. G. (2004). Methodology and preliminary results from the neurocognitive outcomes of depression in the elderly study. Journal of Geriatric Psychiatry & Neurology, 17, 202–211.

Steinberg, M., Shao, H., Zandi, P., Lyketsos, C. G., Welsh-Bohmer, K. A., Nor- ton, M. C., … & Tschanz, J. T. (2008). Point and 5-year period preva- lence of neuropsychiatric symptoms in dementia: the Cache County Study. International journal of geriatric psychiatry, 23(2), 170–177.

Storandt, M., Grant, E. A., Miller, J. P., & Morris, J. C. (2006). Longitudinal course and neuropathologic outcomes in original vs revised MCI and in pre-MCI. Neurology, 67(3), 467–473. doi: 10.1212/01.wnl.0000228231.26111.6e

Theou, O., Rusak, B., & Rockwood, K. (2004). Sleep disturbance predicts incident dementia and mortality. Alzheimer’s & Dementia: The Journal of the Alzheimer’s Association, 9(4), 541.

Tsapanou, A., Gu, Y., Scarmeas, N., & Stern, Y. (2015). Self-reported sleep problems and incidence of dementia. Neurology, 84(14 Suppl), P2–173.

Verkaik, R., Nuyen, J., Schellevis, F., & Francke, A. (2007). The relationship between severity of Alzheimer’s disease and prevalence of comorbid depressive symptoms and depression: A systematic review. Interna- tional Journal of Geriatric Psychiatry, 22(11), 1063–1086.

Vincent, G. K., & Velkoff, V. A. (2010). The next four decades: The older popu- lation in the United States: 2010 to 2050 (No. 1138), Washington, D.C.: US Department of Commerce, Economics and Statistics Administra- tion, US Census Bureau.

Wang, H., Xu, W., & Pei, J. (2012). Leisure activities, cognition and demen- tia. BBA – Molecular Basis of Disease, 1822(3), 482–491.

Wilson, R. S., Boyle, P. A., Yu, L., Barnes, L. L., Schneider, J. A., & Bennett, D. A. (2013). Life-span cognitive activity, neuropathologic burden, and cognitive aging. Neurology, 81(4), 314–321.

Xie, L., Kang, H., Xu, Q., Chen, M. J., Liao, Y., Thiyagarajan, M., … Neder- gaard, M. (2013). Sleep drives metabolite clearance from the adult brain. Science, 42(6156), 373–377.

Yaffe, K., Blackwell, T., Gore, R., Sands, L., Reus, V., & Browner, W. S. (1999). Depressive symptoms and cognitive decline in nondemented elderly women: A prospective study. Archives of General Psychiatry, 56, 425–430.

Yoshizawa, T., Yamakawa�Kobayashi, K., Komatsuzaki, Y., Arinami, T., Oguni, E. I., Mizusawa, H., … Hamaguchi, H. (1994). Dose�dependent association of apolipoprotein e allele e4 with late -onset, sporadic Alz- heimer’s disease. Annals of Neurology, 36(4), 656–659.

1584 S. L. BURKE ET AL.

Copyright of Aging & Mental Health is the property of Routledge and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.