Time to Dementia Diagnosis Among Veterans with Comorbid Insomnia and Depressive Episodes

Abstract

Background:

Older adults with heart failure are at elevated risk of Alzheimer’s disease and related dementias (AD/ADRD). Research suggests that insomnia and depressive episodes contribute somewhat dissociable impacts on risk for AD/ADRD in this patient population, although the temporal ordering of effects is unknown.

Objective:

This study examined time to dementia diagnosis among patients with comorbid insomnia and/or depressive episodes in an epidemiological sample.

Methods:

Secondary data analyses were conducted using a cohort study of 203,819 Veterans with a primary admission diagnosis of heart failure in 129 VA Medical Centers.

Results:

Patients with diagnoses of both insomnia and depressive episodes had the shortest time to a dementia diagnosis at both 1-year (Hazard ratio = 1.43, 95% CI [1.36, 1.51]) and 3-year follow-up time points (Hazard ratio = 1.40, 95% CI [1.34, 1.47]) versus patients with one or neither comorbidity.

Conclusions:

Individuals with both comorbidities had the shortest time to dementia onset. Screening for these comorbidities may help to identify patients at elevated risk of dementia who could benefit from enhanced monitoring or early intervention strategies for more rapid detection and management of dementia symptoms.

Keywords

Alzheimer’s disease dementia depression heart failure sleep disorders Veterans

INTRODUCTION

Alzheimer’s disease and Alzheimer’s disease related dementias (AD/ADRD) are highly prevalent among older adults [1]. Patients with heart failure are at elevated risk of AD/ADRD [2], in part likely due to complications associated with heart failure as well as shared vascular comorbidities that impact both brain (i.e., decreased cerebral blood flow [3, 4]) and cardiac function. Unfortunately, experiencing heart failure together with AD/ADRD complicates the medical management of both conditions [5], and patients with both comorbidities typically experience a worse clinical prognosis [6, 7]. Identifying potentially modifiable factors that may increase risk of AD/ADRD among patients with heart failure is thus a critically-needed first step towards informing early AD/ADRD intervention efforts, and ultimately improving clinical outcomes in this important at-risk group.

Patients with heart failure who experience depressive and/or insomnia (i.e., difficulties falling asleep, staying asleep, or waking up too early [8]) episodes may be particularly at-risk for being diagnosed with AD/ADRD. Research shows that patients with versus without heart failure [9, 10] and patients with versus without AD/ADRD [11] show significantly higher rates of depressive and insomnia episodes. Few studies have examined insomnia and depressive episode overlap among patients with heart failure in AD/ADRD risk. However, a recent epidemiological study found that heart failure patients with one or both of these comorbidities were significantly more likely to also be diagnosed with AD/ADRD than patients with neither depressive nor insomnia episodes [12], underscoring the potential relevance of these conditions as markers of risk for AD/ADRD among patients with heart failure.

Critically, since depressive and insomnia episodes can impact cognitive functioning via somewhat disparate pathways, these comorbidities may prospectively increase risk for AD/ADRD. Indeed, both insomnia and depressive episodes predict subsequent decline in key aspects of cognition (i.e., attention, memory [13, 14]) implicated in AD/ADRD. The precise mechanisms by which these impacts occur are numerous, and may overlap to an extent (i.e., impacts on neuroinflammation, factors impacting neuroplasticity [15 –17]). However, research shows that insomnia symptoms worsen the impact of depressive episodes on function, and vice versa [18, 19], suggesting that these symptom profiles may provide unique contributions to cognitive decline as well. Indeed, epidemiological research found that patients with both comorbidities had greater prevalence of AD/ADRD than in patients with insomnia or depressive episodes alone, or neither condition [12]. These findings preliminarily suggest that insomnia and depressive episodes contribute somewhat dissociable impacts on risk for AD/ADRD in this patient population, although the temporal ordering of effects has yet to be examined in the literature.

Thus, the present study examined whether depressive and insomnia episodes differentially and additively contributed to time to AD/ADRD onset in a large, epidemiological sample of patients with heart failure. Based on prior research, we expected that patients with both depressive and insomnia episodes would show the shortest time to AD/ADRD onset.

METHODS

Participants and procedures

Secondary data analyses were conducted on a cohort study of 373,897 Veterans with a primary admission diagnosis of heart failure that visited one of 129 Department of Veterans Affairs (VA) Medical Centers between October 1, 2011, and September 30, 2020. In instances where patients had multiple heart failure-related admissions, one hospitalization was randomly selected for analysis to minimize bias towards earlier or more advanced disease progression. We excluded any patients with an existing AD/ADRD diagnosis, with a lookback time of 3 years prior to the index admission (n = 77,499), as well as any patients with a death recorded prior to the index admission and a length of stay greater than 90 days for the index admission (n = 299), bringing the final sample size to N = 203,819. See Fig. 1 for a flow chart describing the sample. Veterans Health Administration (VHA) electronic records were used to collect information about demographics, medical and psychiatric comorbidity, prior health care utilization and mortality. All study procedures were approved by the Institutional Review Board (IRB) at the Providence VA Medical Center and were conducted in accordance with the ethical standards from the 1964 Declaration of Helsinki.

Fig. 1

Flow chart describing the cohort selection process.

Measures

Exposures. Presence of insomnia (i.e., primary insomnia, organic insomnia, insomnia due to a medical or mental health condition) and depressive episodes (i.e., major depressive disorder single or recurrent episode, Bipolar I disorder most recent episode depressed, adjustment disorder) were obtained from the VHA electronic health record using International Classification of Disease (ICD) codes for the year prior to hospital admission. ICD-9 and ICD-10 Codes were selected for insomnia (307.42, 327.00, 327.09, 327.01, G47.01, 327.02, F51.05, 780.51, 780.52, G47.00, F51.01, F51.03, F51.04, F51.09, and G47.09), depressive episodes (296.2, 296.3, F33, 296.5, F31.3, F31.4, F31.4, 300.4, F34.1, 309, 311, F20.4, F32, F41.2, and F43.2), and ADRD (331.0, G30.0, G30.1, G30.8, or G30.9, 331.11, G31.01, 331.19, G31.09, 331, 290.0, 331.2, G31.1, 290.10, 290.11, 290.12, or 290.13, 290.20, 290.13, 290.21, 290.3, 290.40, 290.41, 290.42, F01.51, 290.43, F01.50, 294.0, F04, 294.10, F02.80, 294.11, F03.90, 294.20, F03.90, F03.91, 294.8, F06.8, 797, G13.8, F05, F06.1, G31.2, G94, R41.81, and R54).

Outcomes. AD/ADRD (i.e., Alzheimer’s disease, Pick’s disease, presenile or vascular or frontotemporal dementia, cerebral or senile degeneration, dementia) was obtained for the year prior to hospital admission from the VHA electronic health record.

Covariates. Demographic variables (e.g., age, gender, race, ethnicity) were obtained from the VHA electronic health record. All-cause mortality within 30 days post admission and within 365 days post admission was captured from federal data stored in the VHA Vital Status File.

Data analysis

Baseline characteristics across the four diagnostic groups (insomnia only, depressive episodes only, insomnia and depressive episodes, and neither insomnia nor depressive episodes) were compared using analysis of variance (ANOVA) for continuous variables and χ² tests for categorical variables to identify differences between groups. The main analysis was examining if the diagnostic group predicted a new dementia diagnosis within 1 or 3 years of the index hospitalization using a Cox proportional hazards model, interpreting the results as a weighted average under the assumption the proportional hazards model would not be met [20]. For both outcomes, we computed adjusted models. The adjusted models control for age, sex, race/ethnicity, Elixhauser comorbidity index, stroke diagnosis, myocardial infarction, coronary artery bypass surgery, atrial fibrillation, rurality, smoking status, length of stay for index admission, ejection fraction, number of emergency department visits in year prior to admission, inpatient admissions and health care costs in the year prior, length of stay for heart failure admissions in the year prior, in addition to VA medical centers. These covariates were chosen based on the nature of the sample (a heart failure cohort) and known associations between these covariates, insomnia, depression, and dementia. The main analysis used the neither insomnia nor depressive episodes group as the reference group, and we planned a sensitivity analysis where insomnia and depressive episodes group was the reference group. This sensitivity analysis allows us to see if the group with both comorbidities was different from the group with either insomnia or depressive episodes alone. Analyses were conducted using SAS 9.4 (SAS Institute Inc., Cary, NC), using two-tailed tests and a p value of < 0.05. These data were obtained via a data use agreement with the VA and therefore cannot be publicly shared.

RESULTS

A description of the entire cohort appears in Table 1. The sample was predominantly comprised of older White men. The mean (SD) number of Elixhauser comorbidities was 5.4 (2.8). Among the 143,024 (70%) patients whose ejection fraction (EF) was known, it was reduced in 47,120 (33%). Current or former smoking was present in 85% of the sample, atrial fibrillation in 30%, and stroke in 12%. The majority of patients did not have insomnia or a recent depressive episode.

Table 1

Cohort characteristics

Variable	Overall	Insomnia only	Depressive episodes only	Insomnia and depressive episodes	Neither insomnia nor depressive episodes	p
	(N = 203,817)	(n = 7,308)	(n = 49,001)	(n = 8,056)	(n = 139,454)
Age at index submission	70.5 (11.0)	72.2 (10.9)	67.4 (10.1)	67.6 (10.7)	71.6 (11.0)	< 0.001
Age group						< 0.001
18–54	11,872 (5.8%)	343 (4.7%)	3,784 (7.7%)	673 (8.4%)	7,072 (5.1%)
55–64	50,097 (24.6%)	1,412 (19.3%)	16,000 (32.7%)	2,462 (30.6%)	30,223 (21.7%)
65–74	73,816 (36.2%)	2,690 (36.8%)	19,031 (38.8%)	3,104 (38.5%)	48,991 (35.1%)
75–84	42,298 (20.8%)	1,731 (23.7%)	6,636 (13.5%)	1,151 (14.3%)	32,780 (23.5%)
85+	25,734 (12.6%)	1,132 (15.5%)	3,549 (7.2%)	666 (8.3%)	20,387 (14.6%)
Sex						< 0.001
Female	5,161 (2.5%)	175 (2.4%)	1,943 (4.0%)	335 (4.2%)	2,708 (1.9%)
Male	198,650 (97.5%)	7,133 (97.6%)	47,056 (96.0%)	7,721 (95.8%)	136,740 (98.1%)
Missing	6 (0.0%)		1 (0.0%)		5 (0.0%)
Race and ethnicity						< 0.001
White	149,026 (73.1%)	5,668 (77.6%)	35,847 (73.2%)	6,115 (75.9%)	101,396 (72.7%)
Black	39,403 (19.3%)	1,133 (15.5%)	9,555 (19.5%)	1,342 (16.7%)	27,373 (19.6%)
Other	279 (0.1%)	10 (0.1%)	54 (0.1%)	9 (0.1%)	206 (0.1%)
Hispanic	14,499 (7.1%)	479 (6.6%)	3,443 (7.0%)	572 (7.1%)	10,005 (7.2%)
Missing	610 (0.3%)	18 (0.2%)	101 (0.2%)	18 (0.2%)	473 (0.3%)
Rural	82,693 (40.6%)	3,108 (42.5%)	19,634 (40.1%)	3,306 (41.0%)	56,645 (40.6%)	0.001
Missing	539 (0.3%)	14 (0.2%)	134 (0.3%)	17 (0.2%)	374 (0.3%)
Elixhauser comorbidity index (sum score)	5.4 (2.8)	5.9 (2.7)	6.8 (2.7)	7.5 (2.9)	4.8 (2.7)	< 0.001
Stroke diagnosis	23,780 (11.7%)	951 (13.0%)	6,490 (13.2%)	1,176 (14.6%)	15,163 (10.9%)	< 0.001
Myocardial infarction diagnosis	29,234 (14.3%)	1,257 (17.2%)	8,408 (17.2%)	1,638 (20.3%)	17,931 (12.9%)	< 0.001
Coronary artery bypass surgery	1,457 (0.7%)	62 (0.8%)	488 (1.0%)	90 (1.1%)	817 (0.6%)	< 0.001
Atrial fibrillation diagnosis	60,213 (29.5%)	2,570 (35.2%)	13,911 (28.4%)	2,550 (31.7%)	41,182 (29.5%)	< 0.001
Ejection fraction
≤35%	47,120 (23.1%)	1,875 (25.7%)	11,567 (23.6%)	1,987 (24.7%)	31,691 (22.7%)	< 0.001
> 35%	95,904 (47.1%)	3,764 (51.5%)	25,797 (52.6%)	4,458 (55.3%)	61,885 (44.4%)	< 0.001
Missing	60,793 (29.8%)	1,669 (22.8%)	11,636 (23.7%)	1,611 (20.0%)	45,877 (32.9%)	< 0.001
Number of emergency room visits in the past year	2.4 (3.5)	3.2 (4.2)	3.2 (4.2)	4.2 (5.2)	2.0 (3.0)	< 0.001
Length of stay of index hospitalization	5.8 (7.3)	5.5 (6.7)	5.8 (7.3)	5.7 (6.8)	5.9 (7.3)	0.001
Missing	6 (0.0%)	1 (0.0%)	3 (0.0%)	1 (0.0%)	1 (0.0%)	< 0.001
Number of acute inpatient hospitalizations in the past year	1.9 (1.4)	2.2 (1.6)	2.1 (1.6)	2.5 (2.0)	1.7 (1.2)	< 0.001
Length of stay for acute inpatient hospitalizations in past year	5.2 (9.2)	6.7 (10.2)	7.0 (11.5)	9.1 (13.7)	4.3 (7.6)	< 0.001
Number of heart failure hospitalizations in the past year	0.3 (0.7)	0.4 (0.8)	0.4 (0.7)	0.4 (0.8)	0.3 (0.6)	< 0.001
Length of stay for heart failure hospitalizations in past year	0.9 (3.3)	1.2 (4.0)	1.1 (4.0)	1.4 (5.0)	0.8 (2.8)	< 0.001
Total health care costs in past year	41,046.4 (66,972.7)	54,543.2 (14,3103.9)	55,969.9 (75,023.4)	73,039.7 (86,451.2)	33,247.2 (53,529.2)	< 0.001
Smoking status
Smoker or former smoker	172,165 (84.5%)	6,333 (86.7%)	43,537 (88.9%)	7,215 (89.6%)	115,080 (82.5%)	< 0.001
Never smoked	23,028 (11.3%)	784 (10.7%)	4,377 (8.9%)	665 (8.3%)	17,202 (12.3%)
Missing	8,624 (4.2%)	191 (2.6%)	1,086 (2.2%)	176 (2.2%)	7,171 (5.1%)
New ADRD diagnosis within 1 year	33,773 (16.6%)	1,351 (18.5%)	8,400 (17.1%)	1,572 (19.5%)	22,450 (16.1%)	< 0.001
New ADRD diagnosis within 3 years	51,413 (25.2%)	1,956 (26.8%)	13,136 (26.8%)	2,354 (29.2%)	33,967 (24.4%)	< 0.001

Adjusted Cox proportional hazard ratio models are shown in Table 2, with the neither insomnia nor depressive episodes group as the reference. In the 1-year follow-up model, the hazard ratio (HR) of the insomnia only, depressive episodes only, and insomnia and depressive groups were significantly higher than the referent group. The insomnia and depressive episodes group had the highest HR, suggesting the quickest time to a dementia diagnosis, HR = 1.43, 95% CI (1.36, 1.51). Similar results were seen in the 3-year model.

Table 2

Proportional hazards models for ADRD diagnosis within 1- and 3-years of index hospitalization

Predictors	Hazard ratio	95% CI
1-year model
Adjusted model
Insomnia only	1.17	1.10, 1.23
Depressive episodes only	1.28	1.24, 1.31
Insomnia and depressive episodes	1.43	1.36, 1.51
Neither insomnia or depressive episodes	Reference	–
3-year model
Adjusted model
Insomnia only	1.12	1.07, 1.17
Depressive episodes only	1.29	1.26, 1.31
Insomnia and depressive episodes	1.40	1.34, 1.47
Neither insomnia or depressive episodes	Reference	–

Models control for age, sex, race/ethnicity, Elixhauser comorbidity index, stroke diagnosis, myocardial infarction, coronary artery bypass surgery, atrial fibrillation, rurality, smoking status, length of stay for index admission, ejection fraction, number of emergency department visits in year prior to admission, inpatient admissions and health care costs in the year prior, length of stay for heart failure admissions in the year prior, and VA medical centers.

We conducted a sensitivity analysis, where we made insomnia and depressive episodes the reference group to test for differences between the insomnia only and depressive episodes only group and the group with both comorbidities. These results are in Table 3 and show that there were differences between the insomnia and depressive episodes groups and the insomnia only and depressive episodes only groups. In the 1-year model the insomnia only and depressive episodes only groups had a shorter time to a dementia diagnosis (HR < 1) compared to the group with both comorbidities. In the 3-year model, the insomnia only and depressive episodes only groups had a longer time until a dementia diagnosis than the group with both comorbidities.

Table 3

Sensitivity analysis for ADRD diagnosis within 1- and 3-years of index hospitalization

Predictors	Hazard ratio	95% CI
1-year model
Insomnia only	0.81	0.76, 0.88
Depressive episodes only	0.89	0.84, 0.94
Insomnia and depressive episodes	Reference	–
Neither insomnia or depressive episodes	0.70	0.66, 0.74
3-year model
Insomnia only	0.80	0.75, 0.85
Depressive episodes only	0.92	0.88, 0.96
Insomnia and depressive episodes	Reference	–
Neither insomnia or depressive episodes	0.71	0.68, 0.74

DISCUSSION

This study examined whether insomnia and/or depressive episodes predicted a subsequent AD/ADRD diagnosis 1- and 3-years post-hospitalization for heart failure among an epidemiological cohort of Veterans with heart failure. At both follow-ups, patient groups with one or both comorbidities were more likely to develop AD/ADRD in comparison to the group with neither insomnia or depressive episodes. The comorbid insomnia and depression group had the shortest time to AD/ADRD onset. These findings suggest that these comorbidities are relevant to understanding risk of AD/ADRD following a heart failure-related hospitalization, and have important implications for research and clinical practice aimed at the early detection and prevention of AD/ADRD.

To our knowledge, this study was the first to examine whether time to an AD/ADRD diagnosis differed as a function of insomnia and/or depressive episodes. As expected, we found that patients with one or both comorbidities experienced a shorter time to an AD/ADRD diagnosis. These results are consistent with a broader body of literature suggesting insomnia [21], or late-life depressive episodes [22] often precede, and may even be prodromal to, AD/ADRD. However, that patients with both insomnia and depressive episodes showed a more rapid progression to AD/ADRD diagnosis was a novel and critical finding of this study. Associations between insomnia and depressive episodes with AD/ADRD risk are usually studied separately in the literature, and there can be diagnostic overlap between these comorbidities [8] that makes understanding precisely how these diagnoses contribute to an AD/ADRD prognosis challenging. Nevertheless, that hazard ratios were larger among patients with both insomnia and depressive episodes versus either comorbidity alone indicates that this is a patient group with faster time to cognitive decline that may be in particular need of early identification and intervention to slow the progression of AD/ADRD symptoms. These findings are particularly noteworthy among patients with heart failure, who are at increased risk of experiencing comorbid insomnia and/or depressive episodes.

These findings may also suggest that insomnia and depressive episodes have combinatorial effects on AD/ADRD onset, although potential explanations for these effects are not yet clear from the literature. One possibility is that these comorbidities have unique functional (i.e., functional decline, reduced treatment adherence in depressive episodes [23, 24]) and/or neurological sequelae (i.e., disruption of neural repair and consolidation, increased aggregation of amyloid beta peptides and tau proteins in insomnia [25]) that in combination worsen patient prognosis, increasing AD/ADRD risk. Alternatively, since insomnia symptoms can be a symptom of depressive episodes [8, 15], the presence of both diagnoses in the medical record may demarcate more severe manifestations of both symptom profiles, which could worsen risk. It is also possible that insomnia and depressive episodes appear to be related to AD/ADRD because they share risk factors (i.e., neuroinflammation, life stressors) with AD/ADRD [26 –31], and that, rather than insomnia and depressive episodes causing AD/ADRD, instead the presence of these comorbid symptoms compounds the impact of these risk processes or reduces resilience to AD/ADRD decline. Indeed, these comorbid symptoms could have mechanistic pathways that explain these findings. One possibility is a disruption of neuroprotective symptoms (e.g. brain derived neurotrophic factor (BDNF) or the BDNF signaling pathway) that may lead to neurodegeneration [32], which is the hallmark symptom of AD/ADRD. Another possible pathway is that the physiological dysregulation and/or hyperarousal features of insomnia, exacerbated through depression or other depressive symptoms, could drive AD/ADRD pathophysiology [33]. Another possible explanation of our findings may be instead that comorbid insomnia and depressive symptoms later in life, rather than truly representing symptomatology that is independent of AD/ADRD, instead represent a cluster of important prodromal symptoms for AD/ADRD [34, 35]. Although this study relied on a retrospective data analysis, which limits the ability to establish temporal relationships between insomnia and depressive episodes and the onset of AD/ADRD, our results underscore the need to investigate potential mechanisms by which these comorbidities contribute to AD/ADRD risk. Information in this vein could potentially inform strategies to prevent or slow the onset of AD/ADRD symptoms, especially among patients with heart failure.

Taken together, these findings have important implications for clinical efforts to improve early identification and intervention of AD/ADRD. Screening for insomnia and depressive episodes in medical settings, particularly amongst older adults at-risk for AD/ADRD, may help providers identify which patients may benefit from enhanced monitoring for earlier detection of AD/ADRD symptom progression. Moreover, since both comorbidities have been found to have causal impacts on neurological factors implicated in cognitive decline [25], providing timely interventions to ameliorate insomnia or depressive symptoms may be one means of slowing the progression of AD/ADRD symptoms, and ultimately may even reduce the risk of a subsequent AD/ADRD diagnosis. Other important avenues of early intervention may include possible upstream factors that can be related to AD/ADRD risk and contribute to insomnia and depressive episodes (e.g., heart failure, poorly controlled medical conditions).

Limitations

Although a major strength of this study is the epidemiological nature of the sample, there are several specific limitations inherent to the use of a cohort of this size. First, the use of diagnoses gleaned from medical records relies on the accuracy of clinical record keeping, and Veterans are underdiagnosed for some conditions including insomnia [36] and AD/ADRD. Second, these analyses used broad phenotypes such as AD/ADRD, insomnia, and depressive episodes that combine disorders with different etiologies. For example, Alzheimer’s disease and vascular dementia are combined under AD/ADRD, major depression and bipolar disorder (with most recent episode being depressive) are similarly combined, and the term insomnia includes a number of conditions with different pathophysiology. Accordingly, additional research that parses these phenotypic categories in order to better understand the biological underpinnings of the putative relationships is needed. A third limitation is the challenge to examine sleep-related diagnoses in the context of diagnoses that include depressive episodes when there are bidirectional relationships between insomnia and mood disorders [37]. The medical records used for these analyses do not allow for temporal relationships between mood and sleep to be established. Prospective research efforts could improve upon this limitation by assessing the timing of these phenomena relative to one another. Fourth, the ICD billing codes for identification of insomnia and depressive episodes do not enable us to examine effects of symptom severity, which likely contributed to our pattern of findings. Other limitations include the relative under-representation of women and racial and ethnic individuals in the overall sample. However, the presence of almost 40,000 Black individuals in the current analyses is an important contribution to the literature given the historical underrepresentation of Black individuals in medical research [38]. It is possible that the results are unique to Veterans or to Veterans who use VHA for medical care, as Veterans who use the VA differ from those who do not use VA on important demographic variables and are more likely to report psychopathology [39]. Moreover, this study was conducted on a HF cohort and there may be unique AD/AD/ADRD risk factors in this population that would not generalize to individuals who are not living with HF. Future research should seek to replicate these findings in different samples that address these limitations.

Conclusions

Despite these limitations, this study was the first to examine time-to-AD/ADRD diagnosis across patient groups with insomnia and/or depressive episodes. We examined this important research question in an epidemiological sample of older Veterans with heart failure, a patient group that is at particularly heightened risk of AD/ADRD. Our findings indicate that patients experiencing both insomnia and depressive episodes experience a shorter time-to-diagnosis versus patients with either or neither comorbidity, demarcating them as a patient group at particularly elevated risk of AD/ADRD. Additional research is needed to identify the mechanisms that explain the worsened prognosis among patients with both comorbidities. Nevertheless, our findings suggest that screening for insomnia and depressive episodes among older adults in medical settings could facilitate early identification of adults at-risk of AD/ADRD, and potentially enable targeted interventions to slow disease progression.

AUTHOR CONTRIBUTIONS

Melanie Lauren Bozzay (Conceptualization; Methodology; Supervision; Writing – original draft; Writing – review & editing); Hannah E. Joyce (Writing – original draft); Lan Jiang (Data curation; Formal analysis; Writing – original draft); Alyssa N. De Vito (Methodology; Writing – original draft); Sheina Emrani (Conceptualization; Writing – original draft); Julia Browne (Conceptualization; Writing – review & editing); Thomas Bayer (Conceptualization; Writing – review & editing); McKenzie Quinn (Writing – review & editing); Jennifer Primack (Writing – original draft); Catherine M Kelso (Writing – review & editing); Wen-Chih Wu (Conceptualization; Writing – review & editing); James L Rudolph (Conceptualization; Data curation; Writing – review & editing); John E McGeary (Conceptualization; Methodology; Writing – original draft); Zachary J Kunicki (Conceptualization; Formal analysis; Methodology; Writing – original draft; Writing – review & editing).

Footnotes

ACKNOWLEDGMENTS

The authors have no acknowledgments to report.

FUNDING

This work was supported by VA Health Services Research Center of Innovation in Long Term Services and Supports (CIN 13-419). Dr. Browne is supported by a VA Rehabilitation Research and Development Career Development Award (IK1RX003904). Dr. Bayer is supported by a VA Health Services Research & Development Award (23-SWIFT-1). Dr. Wu is supported by a VA CSRD CARA-006-19F and NHLBI 1R01HL156518-01A1. The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the United States Government or Department of Veterans Affairs.

CONFLICT OF INTEREST

The authors have no conflict of interest to report.

DATA AVAILABILITY

The data supporting the findings of this study are not publicly available due to privacy or ethical restrictions.

References

Beydoun

, Beydoun

, Gamaldo

, Rostant

, Dore

, Zonderman

, Eid

(2015) Nationwide inpatient prevalence, predictors, and outcomes of Alzheimer’s disease among older adults in the United States, 2002–2012. J Alzheimers Dis 48, 361–375.

Qiu

, Winblad

, Marengoni

, Klarin

, Fastbom

, Fratiglioni

(2006) Heart failure and risk of dementia and Alzheimer disease: A population-based cohort study. Arch Intern Med 166, 1003–1008.

Roher

, Debbins

, Malek-Ahmadi

, Chen

, Pipe

, Maze

, Belden

, Maarouf

, Thiyyagura

, Mo

, Hunter

, Kokjohn

, Walker

, Kruchowsky

, Belohlavek

, Sabbagh

, Beach

(2012) Cerebral blood flow in Alzheimer’s disease. Vasc Health Risk Manag 8, 599–611.

Hoth

, Poppas

, Moser

, Paul

, Cohen

(2008) Cardiac dysfunction and cognition in older adults with heart failure. Cogn Behav Neurol 21, 65–72.

Lin

, Fillit

, Cohen

, Neumann

(2013) Potentially avoidable hospitalizations among Medicare beneficiaries with Alzheimer’s disease and related disorders. Alzheimers Dement 9, 30–38.

Rattinger

, Dutcher

, Chhabra

, Franey

, Simoni-Wastila

, Gottlieb

, Stuart

, Zuckerman

(2012) The effect of dementia on medication use and adherence among Medicare beneficiaries with chronic heart failure. Am J Geriatr Pharmacother 10, 69–80.

Hoffman

, Maust

, Harris

, Ha

, Davis

(2022) Medicare spending associated with a dementia diagnosis among older adults. J Am Geriatr Soc 70, 2592–2601.

American Psychiatric Association (2013) Diagnostic and statistical manual of mental disorders: DSM-5, American Psychiatric Association, Washington, DC.

Laugsand

, Strand

, Platou

, Vatten

, Janszky

(2014) Insomnia and the risk of incident heart failure: A population study. Eur Heart J 35, 1382–1393.

10.

Celano

, Villegas

, Albanese

, Gaggin

, Huffman

(2018) Depression and anxiety in heart failure: A review. Harv Rev Psychiatry 26, 175–184.

11.

Snowden

, Atkins

, Steinman

, Bell

, Bryant

, Copeland

, Fitzpatrick

(2015) Longitudinal association of dementia and depression. Am J Geriatr Psychiatry 23, 897–905.

12.

Kunicki

, Frietchen

, McGeary

, Jiang

, Duprey

, Bayer

, Singh

, Primack

, Kelso

, Wu

, Rudolph

, Bozzay

(2023) Prevalence of comorbid depression and insomnia among veterans hospitalized for heart failure with Alzheimer disease and related disorders. Am J Geriatr Psychiatry 31, 428–437.

13.

Cross

, Carrier

, Postuma

, Gosselin

, Kakinami

, Thompson

, Chouchou

, Dang-Vu

(2019) Association between insomnia disorder and cognitive function in middle-aged and older adults: A cross-sectional analysis of the Canadian Longitudinal Study on Aging. Sleep 42, zsz114.

14.

Wei

, Ying

, Xie

, Chandrasekar

, Lu

, Wang

, Li

(2019) Late-life depression and cognitive function among older adults in the US: The National Health and Nutrition Examination Survey, 2011–2014. J Psychiatr Res 111, 30–35.

15.

Palagini

, Hertenstein

, Riemann

, Nissen

(2022) Sleep, insomnia and mental health. J Sleep Res 31, e13628.

16.

Zhang

, Li

, Zhang

, Yang

, Rao

, Xia

, Li

, Chen

, Wang

(2020) Astroglial mechanisms underlying chronic insomnia disorder: A Clinical Study. Nat Sci Sleep 8, 693–704.

17.

Gałecki

, Talarowska

, Anderson

, Berk

, Maes

(2015) Mechanisms underlying neurocognitive dysfunctions in recurrent major depression. Med Sci Monit 21, 1535–1547.

18.

Staner

(2010) Comorbidity of insomnia and depression. Sleep Med Rev 14, 35–46.

19.

Fang

, Tu

, Sheng

, Shao

(2019) Depression in sleep disturbance: A review on a bidirectional relationship, mechanisms and treatment. J Cell Mol Med 23, 2324–2332.

20.

Stensrud

, Hernán

(2020) Why test for proportional hazards? JAMA 323, 1401–1402.

21.

Resciniti

, Yelverton

, Kase

, Zhang

, Lohman

(2021) Time-varying insomnia symptoms and incidence of cognitive impairment and dementia among older US adults. Int J Environ Res Public Health 18, 351.

22.

Byers

, Yaffe

(2011) Depression and risk of developing dementia. Nat Rev Neurol 7, 323–331.

23.

Espiritu

DAV

, Rashid

, Mast

, Fitzgerald

, Steinberg

, Lichtenberg

(2001) Depression, cognitive impairment and function in Alzheimer’s disease. Int J Geriatr Psychiatry 16, 1098–1103.

24.

Arlt

, Lindner

, Rösler

, von Renteln-Kruse

(2008) Adherence to medication in patients with dementia: Predictors and strategies for improvement. Drugs Aging 25, 1033–1047.

25.

Sadeghmousavi

, Eskian

, Rahmani

, Rezaei

(2020) The effect of insomnia on development of Alzheimer’s disease. J Neuroinflammation 17, 289.

26.

Troubat

, Barone

, Leman

, Desmidt

, Cressant

, Atanasova

, Brizard

, El Hage

, Surget

, Belzung

, Camus

(2021) Neuroinflammation and depression: A review. Eur J Neurosci 53, 151–171.

27.

Heneka

, Carson

, El Khoury

, Landreth

, Brosseron

, Feinstein

, Jacobs

, Wyss-Coray

, Vitorica

, Ransohoff

, Herrup

, Frautschy

, Finsen

, Brown

, Verkhratsky

, Yamanaka

, Koistinaho

, Latz

, Halle

, Petzold

, Town

, Morgan

, Shinohara

, Perry

, Holmes

, Bazan

, Brooks

, Hunot

, Joseph

, Deigendesch

, Garaschuk

, Boddeke

, Dinarello

, Breitner

, Cole

, Golenbock

, Kummer

(2015) Neuroinflammation in Alzheimer’s disease. Lancet Neurol 14, 388–405.

28.

Zielinski

, Gibbons

(2022) Neuroinflammation, sleep, and circadian rhythms. Front Cell Infect Microbiol 12, 853096.

29.

Hoeijmakers

, Lesuis

, Krugers

, Lucassen

, Korosi

(2018) A preclinical perspective on the enhanced vulnerability to Alzheimer’s disease after early-life stress. Neurobiol Stress 8, 172–185.

30.

Hammen

(2015) Stress and depression: Old questions, new approaches. Curr Opin Psychol 4, 80–85.

31.

Pillai

, Roth

, Mullins

, Drake

(2014) Moderators and mediators of the relationship between stress and insomnia: Stressor chronicity, cognitive intrusion, and coping. Sleep 37, 1199–1208.

32.

Zuccato

, Cattaneo

(2009) Brain-derived neurotrophic factor in neurodegenerative diseases. Nat Rev Neurol 5, 311–322.

33.

Chappel-Farley

, Lui

, Dave

, Chen

, Mander

(2020) Candidate mechanisms linking insomnia disorder to Alzheimer’s disease risk. Curr Opin Behav Sci 33, 92–98.

34.

Yao

, Pelletier

, Fereshtehnejad

, Cross

, Dang-Vu

, Postuma

(2022) Insomnia symptom subtypes and manifestations of prodromal neurodegeneration: A population-based study in the Canadian longitudinal study on aging. J Clin Sleep Med 18, 345–359.

35.

Amieva

, Le Goff

, Millet

, Orgogozo

, Pérès

, Barberger-Gateau

, Jacqmin-Gadda

, Dartigues

(2008) Prodromal Alzheimer’s disease: Successive emergence of the clinical symptoms. Ann Neurol 64, 492–498.

36.

Ryden

, Martin

, Matsuwaka

, Fung

, Dzierzewski

, Song

, Mitchell

, Fiorentino

, Josephson

, Jouldjian

, Alessi

(2019) Insomnia disorder among older veterans: Results of a postal survey. J Clin Sleep Med 15, 543–551.

37.

Carskadon

, Sharkey

, Knopik

, McGeary

(2012) Short sleep as an environmental exposure: A preliminary study associating 5-HTTLPR genotype to self-reported sleep duration and depressed mood in first-year university students. Sleep 35, 791–796.

38.

Sharma

, Palaniappan

(2021) Improving diversity in medical research. Nat Rev Dis Primers 7, 74.

39.

Meffert

, Morabito

, Sawicki

, Hausman

, Southwick

, Pietrzak

, Heinz

(2019) US veterans who do and do not utilize Veterans Affairs health care services: Demographic, military, medical, and psychosocial characteristics. Prim Care Companion CNS Disord 21, 18m02350.