Correlates of Sleep Disturbance Experienced by Informal Caregivers of Persons Living with Dementia: A Systematic Review

Abstract

Objectives:

This study aims to comprehensively review and update the literature concerning the correlates of sleep disturbance among caregivers of persons living with Alzheimer’s disease and related dementias to identify gaps in the literature and antecedent targets for interventions.

Methods:

We searched PubMed, CINAHL, PsycINFO, and Embase using terms related to “sleep,” “caregiver,” and “dementia.”

Results:

Thirty-six articles were included in this review. Based on the antecedents within the 3P model of insomnia, predisposing factors associated with caregiver sleep included caregiver demographics, and physiological factors like genotype and biomarkers. Precipitating factors related to caregiver sleep included caregiving status and responsibilities, and person living with dementia factors.

Conclusions:

Sleep disturbance is a significant issue for caregivers of persons living with dementia. However, this review has identified multiple precipitating factors that are modifiable targets for interventions to improve or enhance caregiver sleep.

Clinical Implications:

Numerous predisposing and precipitating factors contribute to caregivers of persons living with dementia being susceptible to sleep disturbance. Healthcare providers should ask patients about their caregiving status during annual visits. Healthcare providers should also evaluate caregivers’ sleep patterns, and the predisposing and precipitating factors of sleep disturbance, with a focus on the modifiable factors, to enable timely intervention.

The prevalence of sleep disturbance is almost universal in caregivers of persons living with dementia (Beaudreau et al., 2008; Fonareva et al., 2011). Sleep disturbance in caregivers is characterized by longer sleep onset latency, longer wake-after-sleep onset and multiple nightly awakenings compared to normative recommendations and non-caregivers (Castro et al., 2009; Fonareva et al., 2011; H. L. Peng et al., 2019; Vitaliano et al., 2011). Caregivers also report poor sleep quality (Fonareva & Oken, 2014; H. L. Peng et al., 2019; Wilcox & King, 1999), sleep less than the recommended duration (Gao et al., 2019), and spend less time in the restorative stages of sleep (Castro et al., 2009; Fonareva et al., 2011; McKibbin et al., 2005). Since sleep disturbance is so pervasive in caregivers of persons living with dementia, it is important to identify antecedents of caregiver sleep disturbance to target for intervention development.

While there are multiple theoretical frameworks such as the sociocultural stress and coping model (Knight & Sayegh, 2010), or the diathesis-stress model (Baharudin et al., 2019; Russo et al., 1995) to explain the development of sleep disturbance in caregivers of persons living with dementia, for this review, we used the Spielman 3P model which classifies factors that contribute to insomnia into predisposing, precipitating, and perpetuating factors (Spielman et al., 1987). For caregivers, the 3Ps can be both related and unrelated to the caregiving responsibilities. Predisposing factors are present before the symptoms of insomnia begins (McCurry et al., 2015; Perlis et al., 2017) and can include biological factors (gender, age, genetic predisposition, altered metabolism, hyper-reactivity, or dysregulated neurotransmitter systems), preexisting psychological diagnoses, or chronic health conditions (McCurry et al., 2015; Perlis et al., 2017). For example, between 45–60% of caregivers have one or more chronic conditions like hypertension, diabetes, and heart disease (Polenick et al., 2018) which often are present before they become caregivers.

Precipitating events trigger initial insomnia symptoms (Spielman et al., 1987). The most common precipitating factors of insomnia are related to family and health. McCurry et al. (2007) extends Spielman’s model by recognizing that the caregiving situation is unique because the actions of the person living with dementia may contribute to the caregiver’s sleep disturbance and point out that a change in the nighttime behaviors of the person living with dementia is a significant precipitating factor. Other factors that can also precipitate insomnia include menopause transition, or stressful events like illness, divorce, or death of a friend or family member (Bastien et al., 2004; Brewster et al., 2022; McCurry et al., 2007, 2015; Perlis et al., 2017). Over 60% of caregivers report that persons living with dementia have disturbed sleep (Creese et al., 2008; Gehrman et al., 2018). Specifically, 47% of caregivers noted that persons living with dementia had at least two nightly awakenings per week (Gehrman et al., 2018). McCurry et al. (2007) also point out that declining health and caregiver burden are contributors to caregivers’ poor sleep. Pertaining to caregiver health, almost 25% of spousal caregivers report their overall health as fair or poor and their mental health ratings are significantly poorer than age-matched norms (Creese et al., 2008). Caregivers also have greater rates of fatigue, depression, and anxiety (Chang et al., 2020; Vitaliano et al., 2016, 2003; Von Känel et al., 2010). Finally, perpetuating factors are behaviors caregivers implement to compensate for sleep-lessness. These behaviors may include staying in bed for longer hours to get the sleep they lost, napping during the day, or using aids to remain awake (caffeine) or fall asleep (alcohol; Perlis et al., 2017).

Poor sleep in caregivers is associated with many negative physical and psychological outcomes (Creese et al., 2008; Cupidi et al., 2012; Rongve et al., 2010) and earlier residential placement of the person living with dementia (Wulff et al., 2010). Therefore, effective interventions are essential to improve caregiver sleep. A recent review of caregiver sleep interventions found that most intervention effects on sleep outcomes were non-significant (Pignatiello et al., 2021). We surmise that intervening to manage antecedent factors would prevent or reduce the occurrence of caregiver sleep disturbance. While there are some reviews on factors associated with caregiver sleep (Gao et al., 2019; Peng & Chang, 2013), our updated review will specifically examine and identify antecedent correlates of sleep disturbance using the predisposing and precipitating factors of the 3P model.

Methods

This systematic review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Moher et al., 2015).

Data collection

Four electronic databases, MEDLINE/PubMed, CINAHL, PsycINFO, and Embase were reviewed to find literature related to sleep disturbances of informal caregivers of persons living with mild cognitive impairment, memory loss and/or Alzheimer’s disease and related dementias (ADRD). Relevant citations from the reference lists of retrieved systematic reviews were also extracted. Date ranges were applied to each database search to include articles up to and including August 31, 2022, with all ages and races included. Search terms were derived from discussions with an information specialist: sleep, insomnia, restless, fatigue, and circadian rhythm, combined with dementia, Alzheimer, Parkinson, neurocognit*, cognit*, neurodegen*, and memory loss, combined with caregiv*, caregiver*, caretak*, carer*, care partner*, homecar*. We initially identified 1823 studies after removing duplicates.

Study selection

Covidence, an online systematic review management software program, was used to evaluate the studies (Veritas Health Innovation, n.d.). Titles and abstracts of all studies were divided and screened by three reviewers (G.B., I.Y., and D.W.). One hundred and seventy-seven relevant articles were eligible for full-text screening. Eligible full text articles were then independently reviewed by two reviewers, with discrepancies resolved by the third reviewer. After full-text review, we excluded 142 studies. One additional study was identified through a search of the reference sections; thus a total of 36 studies were included in the review. Data from three studies were used to publish two manuscripts each (Brummett et al., 2006, 2007; Simpson & Carter, 2013a; Simpson & Carter, 2015; Von Känel et al., 2012, 2014) See PRISMA flow chart (Figure 1).

Figure 1.

PRISMA diagram.

Inclusion criteria were: (1) studies published in peer-reviewed journals and written in English; (2) PLwD’s medical condition needed to include cognitive impairment, memory loss and/or ADRD; (3) study participants identified themselves to be current informal caregivers of a person living with cognitive impairment, memory loss and/or ADRD; and (4) sleep parameters were an outcome. Studies were excluded if: (1) study participants were professional or trained caregivers; (2) study participants were former caregivers (e.g., PLwD passed away before the study was conducted); (3) they focused solely on sleep disturbance experienced by the PLwD; (4) they focused solely on fatigue and tiredness of caregivers, psychological burnout (e.g., stress) of caregivers; (6) they focused solely on caregivers of persons with other conditions (e.g., Parkinson’s disease); (7) they were intervention studies; and (8) they were solely qualitative, conference proceedings, meeting abstracts, editorials, clinical guidelines, or systematic reviews.

Quality assessment

Three reviewers (G.B., I.Y., and D.W.) used Hawker’s checklist, a validated appraisal tool designed to systematically evaluate and assess methodological rigor (Drewniak et al., 2020; Hawker et al., 2002). Reviewers evaluated nine domains: 1) abstract/title; 2) introduction and aims; 3) method and data; 4) sampling; 5) data analysis; 6) ethics and bias; 7) results; 8) transferability and generalizability and 9) implications and usefulness. Each study was appraised and rated from 1 (very poor) to 4 (good); maximum score is 36 (Hawker et al., 2002). Reviewers used quality scores to evaluate the generalizability and credibility of the results, but studies were included regardless of quality. The scores of the studies ranged from 27 to 36. Studies received lower ratings due to limited generalizability of the results because they had a mostly homogenous sample (White participants), a small sample size with no sample size justification or power calculation, and/or not providing reliability or validity for the measures used in the study.

Data extraction and synthesis

Three reviewers (G.B., I.Y., and D.W.) divided the manuscripts, extracted data independently into a table, and compared results to ensure reliability. The reviewers individually re-read the included full-text publications and identified the specific factors that were associated with caregiver sleep outcomes. The team initially categorized these factors (e.g., psychological factors), then held ongoing discussions to finalize the overarching categories.

Results

Description of study characteristics

The final sample included 32 cross-sectional studies, 3 longitudinal studies, 1 case-control study (Table 1). The number of participants in the samples ranged from 15 to 669. Most of the caregiver participants were female spouses with ages ranging from 56 to 75 years. For the studies that provided race, most of the participants identified as White. Five studies evaluated sleep with the PSG in home or ambulatory (for 1 to 3 days) and 8 studies used actigraphy to evaluate sleep. 50% of the studies used the full-version or a few questions from the PSQI. Six studies assessed sleep with one item or an investigator-developed measure.

Table 1.

Nonintervention studies (n = 36).

Number	Author (Year)	Purpose and Design	Characteristics of Sample and Setting	Sleep Measurement	Domain (Factor: Measure)	Results	Quality of Evidence
1	Beaudreau et al. (2008)	Determine whether distress and burden were associated with objective measures of sleep disturbance in dementia caregivers Cross-sectional	N = 60 Age: 64.8 (±12.5) years Race: Caucasian (100%) Gender: Female (100%) Relationship: Spouse (53.3%), Adult child (38.4%) USA	Actigraphy	Demographics (Age) questionnaire) Psychological factors (Depressive symptoms: Beck Depression Inventory)	Older age was associated with increased time in bed. Greater severity of caregiver depressive symptoms was significantly related to poorer sleep efficiency but not time in bed or sleep onset latency.	31.5
2	Brummett et al. (2006)	Examine associations among negative affect, perceptions of social support, and quality of sleep among caregivers and non-caregivers Cross-sectional	N = 344 Caregivers (n = 175) Age: 61(±13.5) years Gender: Female (74.3%) Race: Not reported Relationship: Spouse (48%), Relative (52%) USA	PSQI	Social factors (Social support: Interpersonal support evaluation list) Psychological factors (Negative Affect: Center for Epidemiologic Studies Depression Scale, Feelings of Hopelessness, Perceived Stress Scale, Spielberger Trait Anxiety Scale)	The path model suggested that negative affect mediated that relationship between caregiver status and sleep quality. The indirect effect of caregiver status on social support, negative affect, and sleep quality was also significant. The relationship between caregiving and sleep quality may be mediated by negative affect and social support.	36
3	Brummett et al. (2007)	Examine the effects of the 5-HTTLPR genotype on sleep quality as moderated by a major life stressor – caregiving for a close relative with Alzheimer’s disease or other dementia. Cross-sectional	N = 288 Caregivers (n =142) Age: 60.9 (±13.9) years Race: 30% African American Gender: Female (74.6%) Relationship: Not reported USA	PSQI	Physiological factors (5-HTTLPR – genotype from peripheral blood samples) Demographics (Age)	Homozygosity for the S allele, in combination with the chronic stress of caregiving, was associated with poorer sleep quality. The relationship between the caregiver status and genotype interaction term was significantly associated with sleep latency, and sleep disturbances; and marginally associated with sleep efficiency and problems with daytime functioning. Caregivers with the S/S allele had poorer sleep quality compared with caregivers with the L allele and all non-caregivers.	36
4	Castro et al. (2009)	Describe subjective and objective sleep characteristics in women dementia family caregivers compared to women non-caregivers. Cross-sectional	N = 43 Caregivers (n = 9) Age: 62.6 (±9.7) years Race: White (67%), Black (11%), Asian (22%) Gender: Female (100%) Relationship: Spouse (66.7%), Parent (33.3%) USA	Home PSG Actigraphy PSQI Sleep diary ESS	Demographics (Age) PLwD factors (Daytime and nighttime behavior of PLwD: Revised Memory and Behavior Problem Checklist) Psychological factors (Depressive symptoms: CES-D Scale)	In the older age group (ages 60 + years), caregivers reported more subjective sleep disturbance and longer sleep latency, but PSG showed minimal differences compared to non-caregivers. Depressive symptoms were positively correlated with PSG-measured sleep latency. Caregiver’s sleep quantity was highly correlated with the sleep quantity of their care recipient.	33
5	Chen et al. (2019)	Examine features of sleep disturbance in Alzheimer’s disease caregivers To explore whether functional abilities and psychiatric symptoms of PLwD were significantly linked to caregivers’ sleep quality Cross-sectional	N = 98 Age = 74.7 (± 4.9) years Race: Not Reported Gender: Female (58.2%) Relationship: Spouse (100%) China	PSQI	PLwD factors(Neuropsychiatric symptoms: Neuropsychiatric Inventory (NPI)) Caregiver Responsibilities (Activities of Daily Living (ADL): Basic Activities of Daily Living) Psychological factors (Depressive symptoms: Hamilton Depression Rating Scale)	Compared with non-caregivers, caregivers had a significantly increased severity of sleep disturbance. There was a statistically positive correlation between the PSQI score of caregivers and NPI scores of PLwD, ADL scores of PLwD, and caregivers’ depressive symptoms.	27.5
6	Chiu et al. (2014)	Test a stress process model of multilevel stressors on sleep disturbance for family caregivers of PLwD Cross-sectional	N = 180 Age: 56.0 (±13.8) years Race: Not reported Gender: Female (65%) Relationship: Adult Children (55.6%), Spouse (40.6%) Taiwan, China	GSDS	Psychological factors (Depression: Center for Epidemiological Studies Depression Scale – Short Form) PLwD factors (Neuropsychiatric symptoms: Chinese neuropsychiatric inventory) Physical factors (Physical fatigue: Visual Analogue for Fatigue Scale)	Sleep disturbance was positively correlated with the PLwD neuropsychiatric symptoms and depression/bad mood. The predictors of sleep disturbance were caregivers’ physical fatigue and depressive symptoms, and the synergistic effect of depressive symptoms and physical fatigue.	33.5
7	Creese et al. (2008)	Investigate the sleep characteristics of a sample of spousal caregivers residing with PLwD Determine the relationship between caregiver sleep variables and health status Cross-sectional	N = 60 Age: 73.65 (±9.26) years Race: Not reported Gender: Female (68%) Relationship: Spouse (100%) Canada	Author developed sleep quality measure	PLwD factors (Nighttime behaviors: Author developed questionnaire) Psychological factors (Depressive symptoms: Center for Epidemiological Studies Depression Scale; Caregiver Responsibilities Role Burden: Zarit Burden Inventory)	63% of caregivers reported their sleep was disturbed by their spouse’s nighttime behavior. Poorer ratings of current sleep quality were significantly associated with a decline in sleep quality over the past year, and more frequent nocturnal disruptions by the PLwD. Poorer caregiver sleep quality was significantly associated with needing to use the bathroom and wandering. No significant differences between caregivers who shared a room with the PLwD and those who slept in separate rooms on sleep quality nor frequency of nocturnal disruptions.	32.5
8	Flaskerud et al. (2000)	Describe distressing emotions and sleep problems in caregivers of people with AIDS, age-related dementias, and advanced-stage cancers Cross-sectional	N = 117 Caregiver (n = 40) Age = 70.6 (±6.2) years Gender: Female (100%) Race: White (70%) Relationship: Not reported USA	Three items from the PSQI	Psychological factors (Depressive symptoms: CES-D) PLwD factors (Caregiver reported wandering)	Alzheimer’s disease caregivers: 60% reported trouble falling asleep, 44% reported restless sleep, 61% reported trouble staying asleep. Sleep problems were related to depressive mood.	32
9	Fonareva et al. (2011)	Compare sleep quality between two groups of generally healthy older adults: dementia caregivers and age/gender-matched non-caregivers Assess relationships between sleep architecture and pro-inflammatory and pro-coagulant activity reportedly increased in caregivers and linked to sleep Cross-sectional	N = 20 Age: 64.50 (±7.13) years Gender: Female (90%) Race: White (97.5%) Relationship: Spouse (70%), Children (30%) USA	Portable PSG ESS PSQI	Demographics (Age) Physiological factors (Salivary cortisol, C-reactive protein – Blood assay) Psychological factors (Depressive symptoms: CESD-10; Perceived Stress: Perceived Stress Scale)	The caregiver group had more problems with sleep on PSQI compared to the non-caregiver group. Caregivers took longer to fall asleep, spent greater proportions of their sleep in Stage N1, smaller proportions of their sleep in Stage R than non-caregivers. Perceived stress was related to percent TST in Stage N2, Stage N3, and Stage R. Awakening cortisol level was related to percent TST in Stage N1, Stage N3, and Stage R. C-reactive protein levels were associated with percent TST in Stage 2, and Stage N3. Depressive symptoms were positively related to overall PSQI scores and negatively related to percent TST in Stage R. Age, group membership, and depressive symptoms were significant predictors of percent of TST in Stage N1 and percent TST in Stage R.	32.5
10	Gibson and Gander (2020)	Identify the prevalence of sleep disturbances and the factors associated with poor sleep among family caregivers Cross-sectional	N = 526 Age: 72.2 (±10.2) years Gender: Female (66%) Ethnicity: NZ European ethnicity (92%), Maori (7%) Relationship: Spouse/partner (87%), Adult child (11%), Other family member (2%) New Zealand	ISI	Demographic (Sex) PLwD factors (Nighttime behavior: Sleep Disorders Inventory) Physical factors (General Health: Short Form Health Survey- SF-12) Caregiver Responsibilities (Timing of care provided)	Greater sleep disturbance was more likely among female carers, those providing care overnight or across 24 hours compared to just during the day, sleep symptoms of the PLwD, and those who rated their health as fair-poor or good as opposed to very good-excellent.	35
11	Lee et al. (2014)	Examine characteristics of sleep disturbances and sleep patterns in caregivers of PLwD with amnestic mild cognitive impairment (aMCI) and dementia to determine whether there is any relationship between BPSD in PLwD with dementia and the quality of sleep experienced by their caregivers Cross-sectional	N = 129 (58 aMCI, 71 dementia) Race: 100% Korean aMCI caregivers Age: 58.0 (±14.26) years Gender: Female (65.3%) Relationship: Spouse group (n = 30); Offspring group (n = 28) Dementia caregiver Age: 57.87 (±7.89) years Gender: Female (60.7%) Relationship: Spouse group (n = 27), Offspring group (n = 44) South Korea	PSQI (Korean Version) ESS ISI Berlin Questionnaire	Demographics (Relationship with PLwD) PLwD factors (Cognition: MOCA, MMSE) Psychological factors (Depressive symptoms: Geriatric Depression Scale-Korean Version)	Compared to non-caregivers, dementia caregivers had higher ISI scores. aMCI caregivers exhibited lower sleep efficiency than non-caregivers. The PSQI global score was significantly higher for both the aMCI and dementia caregivers than non-caregivers, and in the aMCI spouse group than in the offspring group, but this was not the case for the dementia spouse group. ISI and GDS-S scores predicted higher PSQI global scores in aMCI and dementia caregivers.	30
12	Leggett et al. (2018)	Examine caregiver and PLwD characteristics that may be associated with caregivers’ nighttime awakenings. Cross-sectional	N = 451 Age: 59.1 (±1.5) years Race: Not reported Gender: Female (61.1%) Relationship: Child (40.6%); Spouse (35.1%); Other (24.3%) USA	Single-scaled item	PLwD factors (Fall history) Caregiver responsibilities (Time when care is provided) Psychological factors (Emotional difficulty in the care situation) Physical factors (Number of chronic conditions)	Occurrence of caregiver nighttime awakenings: 7% reported every night, 10% reported it happened most nights, and 31% reported at least some nights. Caring for a PLwD who had fallen in the last month was significantly associated with caregivers’ nighttime awakenings. For caregivers, more chronic medical conditions, and higher reports of the emotional difficulty within the care situation were significantly associated with more nighttime awakenings.	34.5
13	Liang et al. (2020)	Explore association between role overload and sleep maintenance insomnia, and moderating effects of social support and social engagement Cross-sectional	N = 669 Age: 60.37 (±13.9) years Race: White (55.95%) Gender: Female (70.23%) Relationship: Adult children (66.22%), Other (18.53%), Spouse (15.25%) USA	Single-scaled item	Demographics (Relationship with PLwD) Social factors (Social support: use of organized support services) Caregiver responsibilities (Instrumental support: daily activities and caregiving tasks questionnaires; Role overload: four item questionnaire)	Role overload of adult dementia caregivers is positively associated with sleep maintenance insomnia Interaction of instrumental support and role overload was related to caregiver sleep disturbance among adult children caregivers, not spousal caregivers. Service use moderated the association between role overload and sleep maintenance insomnia in spousal but not adult children caregivers, not spousal caregivers.	32.5
14	McCurry and Teri (1995)	Determine the kinds of sleep problems reported by a cross-section of elderly caregivers of PLwD Determine how caregiver characteristics (age, gender, depression level, residential status) relate to sleep quality in this population Determine how PLwD characteristics (level of cognitive impairment, depression, and disruptive behaviors) relate to caregiver sleep Cross-sectional	N = 136 Age: 66.0 (±10.3) years Race/ethnicity: Not reported Gender: Female (55%) Relationship: Spouse (59%), Child (32%), Sibling (2%), Other (7%) USA	Investigator-developed sleep questionnaire	Demographics (Age, residential status) PLwD factors (Depression: Hamilton Depression Rating Scale; Progress and Severity of Dementia: Blessed Dementia Rating Scale; PLwDs’ behaviors and Caregivers’ reactions: Revised Memory and Behavior Problem Checklist) Psychological factors (Depression: CES-D)	68% caregivers endorsed some type of sleep problem ≥3 nights/week during past month. Older caregivers were more likely to experience frequency of mid-night awakenings and daytime naps. Increased caregiver sleep disturbance was related to caregiver and PLwD depression, and increased PLwD functional impairment. Caregiver and PLwD depression, caregiver age, residential status and PLwD function predicted caregiver sleep.	32
15	McCurry et al. (2008)	Describe the co-occurrence of sleep problems in PLwD-caregiver dyads, variability of sleep problems over time, and covariates that contribute to these patterns Longitudinal	N = 44 Age: 64.6 (±15.2) years Race: White (86%) Gender: Female (66%) Relationship: Spouse (57%) USA	Actigraphy	Demographics (Gender, Age) Psychological factors (Caregiver Management Strategies: Dementia Management Strategies Scale)	Male caregivers had less total sleep/night and lower sleep efficiency. Older caregivers spent less time in bed. For caregivers in the concordant group, poor caregiver sleep was associated with poor sleep the previous night, being male, and higher use of encouragement as a behavior management strategy. For caregivers in the disconcordant group, poor caregiver sleep was associated with all 3 subjective styles of managing PLwD behaviors (criticism, encouragement, active).	34.5
16	McKibbin et al. (2005)	Examine differences in objective and subjective measures of sleep between spousal caregivers and non-caregivers Cross-sectional	N = 113 Caregivers (n = 73) Caregiver of spouse with moderate or severe AD Age: 73.5 ± 9.4 years Race: White (90.8%) Gender: Female (66.7%) Caregiver of spouse with mild AD Age: 71.1 ± 8.3 years Race: White (95%) Gender: Female (75%) Relationship: Spouse (100%) USA	Home PSG PSQI Functional Outcomes of Sleep Questionnaire	Demographics (Age)	Caregivers of PLwD with moderate to severe Alzheimer’s disease reported significantly more sleep problems than non-caregivers. Objectively, older caregivers of those with moderate to severe Alzheimer’s disease slept less than older non-caregivers. There was a main effect for age, with older caregivers and older non-caregivers having lower sleep efficiency, less slow-wave sleep, and more Stage 1 sleep than younger caregivers and non-caregivers.	33.5
17	Merrilees et al. (2014)	Characterize sleep in PLwD with mild to moderate behavioral variant frontotemporal dementia (bvFTD) and semantic dementia (SD) and their primary family caregivers and to compare PLwD and caregiver data. Cross-sectional	N = 22 (13 FTD, 9 semantic dementia) Age: FTD-59.9 (±9.1) years; SD- 63.0 (±10.8) years Gender: Female (63.6%) Race: Not reported Relationship: Spouse (100%) USA	Actigraphy PSQI	Demographics (Gender) PLwD Factors (Neuropsychiatric behaviors: Neuropsychiatric Inventory questionnaire)	The average PSQI global score was 7.8 for the bvFTD caregivers and 4.9 for the SD caregivers. Female caregivers reported poorer sleep quality than male caregivers. Nighttime disruption was reported more often in PLwDs with bvFTD than SD. There was a strong positive correlation between nighttime behaviors in PLwDs with bvFTD and caregiver distress related to sleep disruption, but not in the SD group.	29.5
18	Mills et al. (2009)	Examine effect of caregiver gender and spouse dementia severity on sleep, coagulation, and inflammation in caregiver Cross-sectional	N = 122 Caregivers (n = 81) Age: 70.2 years (mean age for all participants) Race: White (88.5%) Gender: Female (72.1%) Relationship: Spouse (100%) USA	Home PSG	Demographics (Gender) Physiological factors (Biomarkers: IL-6 and D-dimer)	Males caring for spouse with moderate to severe dementia had more time awake after sleep onset than females. Males had worse sleep apnea than females Males had less slow-wave sleep than females Male caregivers caring for spouses with worse dementia had significantly higher circulating levels of D-dimer than female counterparts.	32
19	Moore et al. (2011)	Investigate the role of pleasant events and activity restriction in determining sleep-related outcomes among dementia caregivers Cross-sectional	N = 125 Age: 73.7 (±7.99) years Race: White (88.5%) Gender: Female (72.1%) Relationship: Spouse (100%) USA	PSQI	Psychological factors (Depression: CES-D) Social factors (Caregiver engagement in pleasant events: Modified version of Pleasant Events Schedule; Caregiver activity restriction: Activity Restriction Scale)	Caregivers with low pleasant events and high activity restriction had significantly worse overall sleep quality than caregivers with high pleasant events and low activity restrictions The correlation between depressive symptoms and sleep quality was highly significant	34
20	Naruse et al. (2012)	Use a data mining technique (chi-square automatic interaction detection) to segment caregivers into homogeneous subgroups based on their sleep dissatisfaction rates Cross-sectional	N = 280 Age: Less than 65 years (48.9%) Race: Not reported Gender: Female (77.1%) Relationship: Daughters (41.4%), Wives (32.9%), Husband (12.2%), Son (8.2%) Japan	SDI	PLwD factors (Severity of dementia symptoms: Japanese index of independence/dementia symptoms; Difficulties experienced by care recipients in activities of daily living: Katz index of independence)	Caregivers caring for people who had difficulty with incontinence and eating and severe dementia tended to be dissatisfied with their sleep. Care recipients’ severity of illness index was associated with caregivers’ sleep dissatisfaction in two categories: (a) unstable and bedridden and (b) stable or non-bedridden.	31.5
21	H. L. Peng et al. (2019)	Examine the association among caregivers’ depression, burden, health status, sleep hygiene, care-recipients’ sleep, and objectively measured caregivers’ sleep Cross-sectional	N = 43 Age: 65 (±9.84) years Gender: Female (93%) Race: White (93%) Relationship: Spouse (69.8%) USA	Actigraphy Sleep diary PSQI SHI	PlwD factors (Care Recipients’ sleep: Sleep Disorders Inventory) Psychological factors (Depression: CES-D) Caregiver responsibilities (Burden: Caregiver Burden Inventory)	PLwD sleep problems predicted caregivers’ total sleep time and sleep efficiency. Caregiver burden predicted their wake after sleep onset. Caregivers’ depression predicted caregivers’ overall sleep quality and daytime function.	34
22	L. M. Peng et al. (2018)	Examine relationships between PLwD risky wandering behaviors and four caregiver’s health outcomes, while controlling for the characteristics of PLwD and caregivers Cross-sectional	N = 180 Age: 56.01 (±13.18) years Gender: Female (65%) Relationship: Child or child-in-law (55.6%), Spouse (40.6%), Other (3.4%) China	GSDS (Chinese version)	Demographics (Age) PLwD factors (Risky Behaviors: Chinese Revised Algase Wandering Scale-Community Version)	Age of the caregiver significantly predicted sleep disturbance; younger caregivers experienced higher levels of sleep disturbance. Eloping behavior was a significant predictor of sleep disturbance.	35.5
23	Polenick et al. (2018)	Examine how medical/nursing tasks and care-related sleep disturbances are associated among spousal caregivers of a PLwD Cross-sectional	N = 104 Age: 75.57 (±1.15) years Gender: Female (59%) Race: Not reported Relationship: Spouse (100%) USA	One question about frequency of sleep disruption in the last month due to caregiving	Caregiver responsibilities (Medical/Nursing Skills performed: Caregivers reported tasks they completed; Activities of Daily Living: Severity of dementia care and ADL assistance provided; Role overload: 3-item questionnaire about caregiver status) Psychological factors (Emotional difficulties: emotional caregiving difficulties and the extent of these difficulties; Negative caregiving relationship quality: 1-item questionnaire)	Significant positive correlations were found between caregivers’ sleep disturbance and activities of daily living assistance, emotional caregiving difficulties, role overload, and negative caregiving relationship quality. More frequent care-related sleep disturbances were also significantly correlated with caregivers’ greater nighttime awakenings. Performing a higher total number of medical/nursing tasks was significantly linked to more frequent care-related sleep disturbances. Caring for wounds/sores was significantly associated with more care-related sleep disturbances.	31.5
24	Pollak and Stokes (1997)	Assess the daily rest-activity cycles of older people living with dementia and older people without dementia and their caregivers Cross-sectional	N = 43 Caregivers, N = 25 Age: 67.3 (±10.2) years Gender: Not reported Race: Not reported Relationship: Children (53%), Spouse (39%) USA	Actigraphy	PLwD factors (Sleep: Actigraphy)	Older adults with dementia were not significantly more active than their caregivers. No significant effect of cohabitation was found for daytime activity, nighttime activity, or activity over the 24-hour day. Significant, positive correlations were found between the median lights-off and rise of older people with dementia and their caregivers but not between those who were not demented and their caregivers.	29
25	Riccio (1996)	Determine the perceived quality of sleep during the main sleep period of the 24-hour day Determine the levels of daytime sleepiness during the main wake period of the 24- hour day Determine whether there is a relationship between perceived sleep quality and levels of daytime sleepiness Determine whether sleep quality and sleepiness are related to other extraneous variables in female caregivers Cross-sectional	N = 15 Age: 72 (±8.9) years Gender: Female (100%) Race: Not reported Relationship: Spouse (93%), Daughter (7%) USA	LSS SSS VAS	Psychological factors (Anxiety: Spielberger State Anxiety Inventory)	Sleep quality scores correlated positively with the level of anxiety. Level of daytime sleepiness correlated negatively with the level of anxiety.	32.5
26	Rowe et al. (2008)	Describe sleep patterns in a subset of dementia caregivers who provide care during the night, and compare those patterns to non-caregivers Examine night-to-night variability in caregiver’s sleep patterns and compare to controls Explore correlation of subjective measures and objective measures of sleep between caregivers and control Cross-sectional	N = 31 Age: 70.7 (±7.8) years Gender: Female (74%) Race: White (97%) USA	Actigraphy Sleep diary	Demographics (Age) Psychological factors (Depressive symptoms: Center for Epidemiologic Studies Depression Scale)	ESS was higher for caregivers than controls. Caregivers had lower TST, SE, and night-to-night variability measured objectively and subjectively Caregivers had greater SOL measured objectively but not subjectively Caregivers had greater WASO measured subjectively but not objectively Older caregivers had worse objectively measured sleep compared to non-caregivers Higher depressive symptoms predictive of subjectively measured sleep (longer TST, greater sleep onset latency, and WASO).	33
27	Sakurai et al. (2015)	Assess autonomic nervous system activity and quality of sleep in family caregivers of people with ambulatory dementia Case-control	N = 20 Age: 60 (range: 56.0–65.8) years Gender: Female (80%) Ethnicity: Japanese (100%) Relationship: Not reported Japan	Actigraphy Sleep diary PSQI	Psychological factors (Stress: Perceived Stress visual analog scale) Physiological factors (Sympathetic nervous system activity: Heart rate variability spectral analysis)	Caregivers had worse PSQI score compared to non-caregivers. Caregivers exhibited higher levels of sympathetic nervous system activity during sleep than non-caregivers. Caregivers had greater high frequency heart rate variability compared to non-caregivers The low frequency/high frequency ratio for the total sleep period was significantly greater in the caregivers than in the non-caregivers.	32.5
28	Simpson and Carter (2013a)	Examine caregiver sleep quality, particularly in relation to the daytime and nighttime behavioral and psychological symptoms of dementia (BPSD) exhibited by the PLwD Cross-sectional	N = 80 Age: 63.3 (SD = 12.8) White: 90% Black: 10% Hispanic:13.8% Non-Hispanic: 86.3%, Female: 88% Spouse (48.4%), Adult child (41.3%) USA	PSQI Individual subscale for subjective sleep quality, duration, sleep onset latency, sleep efficiency	Demographics (Residential status) PLwD factors (Neuropsychiatric factors: Neuropsychiatric Inventory questionnaire)	Nonresidential caregivers were disrupted at night less often. Most common reasons for the disturbance were that the PLwD went to the bathroom and that the PLwD awoke and became confused or disoriented. The correlations between disruptions at night by the PLwD and the global sleep score, and the relationship between disturbances by PLwD and poorer appraisal of the caregivers’ sleep experience were significant. Agitation and the caregivers’ reports of greater frequency in PLwD apathy predicted caregivers’ appraisal of their sleep quality.	33.5
29	Simpson and Carter (2015)	Examine similarities and differences, with a focus on sleep quality, between distal caregivers and caregivers living with the PLwD Cross-sectional	N = 80 Cohabitating caregivers (n = 59) Age: 65.08 (±12.87) years Race: White (89.8%) Gender: Female (89.8%) Relationship: Spouse/partner (38%), Adult child (15%), Other (6%) Distal caregivers (n = 21) Age: 58.14 (±11.36) years Race: White (90.5%) Gender: Female (85.7%) Relationship: Spouse/partner (2%), Adult child (18%), Other (1%) USA	PSQI	Demographics (Residential location: Living with or distance caregiver)	Distal caregivers experienced the same level of poor sleep quality as co-habiting caregivers. Co-habiting caregivers reported more frequent disruptions by PLwD at night.	34
30	Taylor et al. (2015)	Assess whether approach and avoidant coping styles systematically differed between dementia caregivers with and without clinically low sleep efficiency Cross-sectional	N = 60 Age: 73.31(±7.05) years Gender: Female (81.7%) Race: White 96.7% Relationship: Spouse (100%) USA	Home PSG PSQI	Psychological factors (Coping: Brief Cope, Depressive symptoms: HAMD)	For each unit increase on the avoidant coping factor, participants were 4.9 times more likely to be classified in the low sleep efficiency group. Approach coping was unrelated to sleep efficiency. Sleep efficiency was associated with both approach and avoidant coping Avoidant coping was also associated with significantly longer sleep latency.	33
31	Teel and Press (1999)	Compare perceptions of fatigue among older adults caring for spouses with Alzheimer’s disease, Parkinson’s disease, or cancer with a control group of older adults whose spouses required no extra care Cross-sectional	N = 92 Caregivers of persons living with Alzheimer’s disease (n = 30) Age: 72 years Race: White (100%) Gender: Female (53.3%) Relationship: Spouse (100%) USA	VSH	PLwD factors (Diagnosis of the PLwD) Caregiver status	Caregiver subjective sleep ratings were poorer for caregivers of persons with Alzheimer’s disease compared to caregivers of persons with Parkinson’s disease Alzheimer’s disease caregiver group reported significantly greater sleep difficulty compared to the control.	30.5
32	Von Känel et al. (2012)	Examine subjective and objective sleep characteristics longitudinally using the PSQI and actigraphy in community-dwelling elderly spousal Alzheimer’s disease caregivers compared to non-caregiving control subjects, over a period of up to three years and adjusting for important covariates of sleep Investigate the effects of placement or death of the Alzheimer’s disease spouse on caregivers’ sleep Longitudinal	N = 109 Age 74.1 ± 8.1 Gender: Female (69.7%) Race: White (92%) Relationship: Spouses (100%) USA	Actigraphy PSQI	Demographics (Age) PLwD factors (death) Physical factors (Body Mass Index) Caregiver Responsibilities (Role Overload: 4-item scale) Psychological factors (Depressive symptoms: Center for Epidemiologic Studies-Depression scale) Physiological factors (cytokines: TNF-α)	Greater levels of role overload and depressive symptoms predicted greater PSQI scores. Age and BMI both predicted daytime TST. BMI predicted nighttime sleep percent. There was a significant interaction between caregiver status and depressive symptoms for nighttime TST. Death of AD spouse was associated with poor subjective sleep Spousal death increased caregivers’ nighttime wake after sleep onset (WASO) by 23 min (P = #@##.002) and daytime total sleep time (TST) by 29 min (P = #@##.003), while nighttime sleep percent decreased by 3.2% (P = #@##.009) There were significant interactions between caregiver status and TNF-α for nighttime WASO and nighttime sleep percent.	31.5
33	Von Känel et al. (2014)	Examine the longitudinal relation between positive affect (PA) and sleep in 126 spousal Alzheimer’s disease caregivers Longitudinal	N = 126 (Baseline) Age: 74.20 (±7.90) years Gender: Female (70.6%) Race: Not reported Relationship: Spouse (100%) USA	Actigraphy PSQI	Demographics (Gender) Caregiver responsibilities (Role overload: a four-item scale) PLwD factors (Transition in the relationship with spouse: placement or bereavement; PLwD behavior and Caregiver reactions: Problem Behaviors- Revised Memory and Behavior Problem Checklist) Psychological factors (Caregiver Affect: Positive and Negative Affect Scale)	Caregivers with higher mean levels of PA had significantly better subjective sleep. Greater mean level of NA was significantly associated with lower subjective sleep quality. A greater mean level of the positivity ratio was also associated with a significantly better subjective sleep quality. Elevated PSQI scores was significantly predicted by female gender, increased role overload, more problem behaviors, and deceased status of the Alzheimer’s disease spouse. Increases in PA from one assessment to another was associated with significant improvement in subjective sleep quality. NA was an independent predictor of poor sleep quality. Caregivers with an increase in the positivity ratio between two assessments also experienced improvement in subjective sleep quality during the same time interval.	33.5
34	Wilcox and King (1999)	Examine sleep complaints among older women caregivers Cross-sectional	N = 90 Age: 62.8 (±9.5) years Gender: Female (100%) Race: White (87.5%) Relationship: Spouse (53.3%), Other (46.7%) USA	PSQI	Demographics (Education) PLwD factors (nighttime disruptions: Survey asking how often caregiver’s nighttime sleep was interrupted by care recipient)	Caregivers with less education were more sleep impaired than caregivers with higher education (B = −0.34, t (76) = −2.40, p < #@##.02) 28% of caregivers reported care recipients woke them in past month; 60% of those stated that these disruptions occurred 3 or more times per week	29.5
35	Willette-Murphy et al. (2006)	Describe sleep and mental health outcomes of older wife caregivers for spouses living with dementia in comparison with age-matched non-caregiving spouses Describe the relationship of selected variables (PLwD behavioral problems, hours of caregiving time per day, and number of years spent caregiving) to wife caregiver appraisal of perceived stress, self-reported sleep disturbance, and mental health outcomes Cross-sectional	N = 37 Age: 74.1 (±7.6) years Gender: Female (100%) Relationship: Spouse (100%) Race: Not reported USA	MDSD	Demographics (Relationship to PLwD) Caregiver responsibilities (Appraisal of Burden: Zarit Burden Interview)	There were differences between caregivers and non-caregivers TST, SOL, awakenings, WASO, and sleep quality. Caregiving wives had sleep efficiency scores that were significantly lower. Caregiver appraisal of burden did not directly or indirectly predict caregiver sleep efficiency.	32
36	Wilson et al. (2019)	Examine dementia family caregiver risks for heightened pain, sleep problems, and depressive symptoms, in relation to the cycles of intrusive thoughts and avoidance that characterize caregiving-related distress Cross-sectional	N = 130 Age: 65 (± 12.5) years Gender: Female (75.4%) Race: White (82.3%), Black (14.6%), Multiracial (3.1%) Relationship: Spouses (55.4%), Adult children (44.6%) USA	PSQI	Physiological factors (genes: 5-HTTLPR and rs6295)	Among short/short (S/S) allele caregivers and short/long (S/L) allele caregivers, having more distress was associated with greater sleep problems; the effect was not significant among long/long allele caregivers. Among 1019C/1019C allele caregivers and 1019C/1019 G allele caregivers, having more distress was related to greater sleep problems; the effect was not significant among 1019 G/1019 G allele caregivers. C/C caregivers had a significantly stronger association between distress and sleep problems than G/G caregivers and C/G caregivers.	33

ESS- Epworth Sleepiness Scale; GSAQ – Global Sleep Assessment Questionnaire; GSDS- General Sleep Disturbance Scale; IDQ – Investigator Developed Questionnaire; ISI- Insomnia Severity Index; LSS-Likert Sleep Scale; MDSD- Morin Daily Sleep Diary; MOS Sleep Scale- Medical Outcomes Study Sleep Scale; PHQ- Patient Health Questionnaire; PSG – Polysomnography; PSQI – Pittsburgh Sleep Quality Index; SD – Sleep Diary; SDI- Sleep Dissatisfaction Index; SE – Sleep efficiency; SHI- Sleep Hygiene Index; SMI – Sleep Maintenance Insomnia; SOL – Sleep onset latency; SQ – Sleep quality; SSS- Stanford Sleepiness Scale; REM- Rapid Eye Movement; TST – Total sleep time; VAS- Visual analogue scale; VSH- Verran and Snyder-Halpern Sleep Scale; WASO – wake after sleep onset

Predisposing factors

Demographic factors.

Age, sex, education, and cohabitation status were associated with caregiver sleep. Specifically, older age was associated with greater sleep disturbance (t = 3.13, P = .03) and longer sleep onset latency (SOL, t = 3.70, P = .002; Castro et al., 2009). Findings regarding how age was associated with time in bed differed across studies; one study reported older adult caregivers had increased time in bed (β = .32; P = .01; Beaudreau et al., 2008) and another reported that older age was associated with less time in bed (P = .04; McCurry et al., 2008). Being a female caregiver was associated with more symptoms of insomnia (P = .006; Gibson & Gander, 2020), whereas being a male caregiver was associated with shorter sleep duration (P = .003; McCurry et al., 2008) and lower sleep efficiency (P = .004; McCurry et al., 2008), less slow-wave sleep (P = .001; Mills et al., 2009), and longer wake after sleep onset (P = .011) (WASO, time awake in bed after initially falling asleep and before final awakening; Mills et al., 2009). Moreover, male caregivers had a 9-fold increased risk for poor sleep compared to female caregivers when the PLwD was sleeping poorly (relative risk = 8.76, P = .02; McCurry et al., 2008). As it pertains to education, caregivers with less education reported greater sleep disturbance than caregivers with higher levels of education (r = −.23, P < .05; Wilcox & King, 1999). Finally, caregivers cohabiting with the PLwD reported more frequent nightly disruptions compared to caregivers who lived apart from the PLwD (t = 5.0, P < .001; Simpson & Carter, 2015); however, caregiver’s sleep quality did not differ between distal caregivers and caregivers living with the PLwD (t = −.38, P = .71; Simpson & Carter, 2015).

Physiological factors.

TNF (tumor necrosis factor)-α, inflammatory biomarkers including C-reactive protein (CRP) and cortisol, and genes like genotype 5-HTTLPR (region of the gene that codes for serotonin transport), and genotype rs6295 (polymorphism of the receptor gene) were physiological factors associated with caregiver sleep. Relative to non-caregivers, caregivers reported longer nighttime WASO if TNF-α levels were high, but shorter WASO if TNF-α levels were low (P = .049 for the interaction; Von Känel et al., 2012). Additionally, caregivers had lower sleep efficiency if TNF-α levels were high and higher sleep efficiency if TNF-α levels were low (P = .025 for the interaction; Von Känel et al., 2012). Compared to non-caregivers, CRP levels, which were higher in caregivers than non-caregivers, increased as the percentage of total sleep time (TST) spent in Stage N2 increased (r = .38, P = .03) and increased as the percentage of TST spent in Stage N3 decreased (r = −.44, P = .01) (Fonareva et al., 2011). Similarly, cortisol awakening levels, also greater in caregivers compared to non-caregivers, increased as the percent TST in Stage N1 increased (r = .34, P = .04), and increased as the percent TST in Stage N3 (r = −.36, P = .03) and REM decreased (r = −.33, P = .05) (Fonareva et al., 2011).

There was an interaction between caregiver status and differences in the 5-HTTLPR genotype (P = .009) such that caregivers with homozygous short/short (S/S) allele had poorer sleep quality compared to caregivers with only one S allele (Brummett et al., 2007). Moreover, the difference in the Pittsburgh Sleep Quality Index (PSQI) component scores, including sleep latency (P =.04), and sleep disturbance (P = .05), was greater for caregivers with a homozygous 5-HTTLPR genotype compared to those with a heterozygous 5-HTTLPR genotype (Brummett et al., 2007). Wilson et al. (2019) reported a moderator effect of the S allele: caregivers with the S/S allele (estimate = 0.20, standard error = 0.04, P < .0001) and caregivers with the short/long (S/L) allele (estimate = 0.11, standard error = 0.04, P = .004) had more distress associated with greater sleep problems. Wilson et al. (2019) also found that for caregivers with the Cytosine/Cytosine (C/C) allele (estimate = .21, standard error = .05, P < .0001) and the Cytosine/Guanine (C/G) alleles (estimate = .06, SE = .03, P = .032), having more distress was related to greater sleep problems. The effect was not significant among caregivers with the G/G allele (P > .250), suggesting that the C allele may moderate sleep disturbance in caregivers (Wilson et al., 2019).

Precipitating factors

Caregiver status and responsibilities.

Being a caregiver was a significant predictor of poor sleep. Compared to non-caregivers, caregivers on average spent a greater proportion of their sleep in light sleep (F(1, 37) = 4.56, P = .04), a smaller proportion of their sleep in rapid-eye-movement (REM) sleep (F(1, 37) = 5.16, P = .03) and took significantly longer to fall asleep (F(1, 37) = 4.56, P = .04) (Fonareva et al., 2011). Moreover, during sleep, caregivers had higher levels of autonomic nervous system activity compared to non-caregivers (P = .048; Sakurai et al., 2015). Increased nursing tasks and role overload were associated with negative caregiver sleep outcomes. Caregivers who performed a higher number of medical/nursing tasks reported significantly more frequent care-related sleep disturbances (t =2.63, P =.01; Polenick et al., 2018). Wound care was independently associated with more frequent care-related sleep disturbances (t = 2.61, P = .01; Polenick et al., 2018). According to Von Känel et al. (2012), higher caregiver overload significantly predicted poorer subjective sleep quality (P < .001) but not objective sleep measures. Finally, Willette-Murphy et al. (2006) did not find a predictive relationship between caregiving and appraisal of burden on sleep efficiency (P =.22).

Social factors.

Instrumental support (caregivers’ acceptance of help with daily chores and caregiving tasks from friends/family), perceived social support (caregivers’ perception of another person’s availability to offer help, companionship, or instrumental support along with maintaining their self-esteem), and engagement in pleasant activities influenced caregiver sleep. Instrumental support moderated the relationship between role overload and sleep maintenance insomnia among the adult children caregivers of the PLwD (β = −.055, P = .027) but not among spousal caregivers (Liang et al., 2020). Poorer perceived social support with negative affect explained the relationship between caregiving and sleep quality (standardized structural coefficient = .12, P = .001; Brummett et al., 2006). Finally, there were group difference among the 3 groups of caregivers (1) high pleasant events and low activity restriction, 2) either high in pleasant events and high in activity restriction, or low in pleasant events and low in activity restriction, and 3) low pleasant events and high activity restriction caregivers) in subjective sleep quality (P = .010), sleep latency (P = .007), habitual sleep efficiency(P = .021), sleep disturbance (P = .001), and daytime dysfunction (P = .004; Moore et al., 2011).

Physical factors.

Poorer self-rated health (β = .29, p < .001, Gibson & Gander, 2020), (odds ratio = 1.33, P < .001, Leggett et al., 2018) and body mass index (P = .005; Von Känel et al., 2012) predicted greater sleep disturbance.

Psychological factors.

Depressive symptoms, affect, anxiety and coping styles all influenced caregiver sleep. Greater depressive symptoms were associated with poorer sleep quality (r = 0.52, P < .010, Chen et al., 2019), (r = .31, P = .04, Flaskerud et al., 2000), (r = 0.45, P < .001, Moore et al., 2011) and increased WASO (β = −.298, P = .021, Taylor et al., 2015). Further, the presence of caregiver depression predicted higher caregiver sleep disturbance (P < .001; McCurry & Teri, 1995), and higher depressive symptoms predicted poor sleep quality (P < .001; Von Känel et al., 2012), longer sleep duration (P < .01; Rowe et al., 2008), (P = .095; Von Känel et al., 2012), greater SOL (P < .05; Rowe et al., 2008), and greater WASO (P < .01; Rowe et al., 2008). Beaudreau et al. (2008) found that the greater severity of caregiver depressive symptoms significantly correlated to poorer sleep efficiency (β = −.27; P = .04), but not time in bed or SOL, while Castro et al. (2009) found that depressive symptoms were positively correlated with PSG-measured sleep latency (r = .67, P = .05). In addition to having an independent effect, the presence of depressive symptoms combined with physical fatigue predicted the frequency of caregivers’ sleep disturbance (P < .05; Chiu et al., 2014). Depressive symptoms also influenced the relationship between caregiving and poor subjective and objective sleep. Specifically, caregivers had longer nighttime TST if depressive symptom levels were high (P = .095) but shorter nighttime TST when depressive symptom levels were low (P = .73) compared to non-caregivers (Von Känel et al., 2012).

In a 4-year longitudinal study examining caregiver sleep, elevated levels of positive affect (p =.016) and a greater positivity ratio (P < .001), as well as annual increases in these two domains in caregivers’ sleep patterns (P < .001), significantly predicted better self-reported sleep (Von Känel et al., 2014). In contrast, a greater mean level of negative affect was significantly associated with lower subjective sleep quality (P = .001; Von Känel et al., 2014). Similarly, negative affect correlated to poorer sleep quality (P < .001); it also mediated the relationship between caregiving and sleep quality (standardized structural coefficient = .21, P = .001; Brummett et al., 2006). Greater psychological distress (r = .59, P < .01, Wilcox & King, 1999) and caregiver anxiety (Riccio, 1996) were related to poorer overall sleep quality.

Caregiver coping style was another psychological factor that impacted sleep: when caregivers practiced avoidant coping, they experienced lower sleep efficiency (β = −.268, P = .040) and higher SOL (β = .378, P = .003; Taylor et al., 2015). Specifically, for each unit increase on the avoidant coping composite factor, participants were about five times more likely to be classified in the poor sleep efficiency group, (B = 1.601, χ2(1) = 3.943, P = .047). For caregivers whose sleep was concordant with the PLwD, higher use of encouragement as a dementia behavior management strategy was associated with poor caregiver sleep (relative risk = 1.21, P = .04; McCurry et al., 2008).

PLwD-associated factors.

Caregiver sleep quality was affected by the severity and type of cognitive and functional impairment of PLwD. Specifically, caregiver sleep quality was related to the Mini-Mental Status Exam (r = .273, P = .035) and Montreal Cognitive Assessment scores (P < .05; Lee et al., 2014). Caregivers of patients with frontotemporal dementia compared to caregivers of patients with semantic dementia reported more frequent nighttime behaviors (means = 5.5 vs. 4.7), poorer sleep quality (PSQI mean = 7.8 vs. 4.9), and greater use of sleep medications (Merrilees et al., 2014). Further, caregiver subjective sleep ratings were poorer for caregivers of persons living with Alzheimer’s disease compared to caregivers of persons living with Parkinson’s disease or cancer, or control caregivers (Alzheimer’s disease: 4.48, Parkinson: 3.88, Cancer: 3.88, Controls: 3.02; Teel & Press, 1999). Caregivers of independent PLwD with severe dementia experienced greater sleep dissatisfaction (83.3%) compared to dependent PLwD with severe dementia (65.3%; Naruse et al., 2012). Finally, caregivers reported greater sleep disturbances when PLwD had higher functional impairment/more limited activities of daily living (r = 0.61, P < .010, Chen et al., 2019), (r = 0.29, P < .01, Chiu et al., 2014).

Behavioral and psychological symptoms of dementia were significantly associated with caregiver sleep quality. Caregiver sleep quality also significantly correlated with PLwD’s neuropsychiatric symptoms (r = 0.95, P < .010, Chen et al., 2019), (r = .32, P < .001, Chiu et al., 2014), depression/mood (r = 0.32, P < .01, Chiu et al., 2014), (r = 0.339, P < .01, Lee et al., 2014; r = .23, P ≤ .05, McCurry & Teri, 1995; t = 5.70, P < .001, Simpson & Carter, 2013a), agitation (r = .287, P < .05, Lee et al., 2014; r = .27, P < .01, Simpson & Carter, 2013), appetite change (r = .322, P < .01, Lee et al., 2014), and inappropriate behavior (r = .23, P < .05, Simpson & Carter, 2013a).

In addition to behavioral and psychological symptoms of dementia, caregivers reported poorer sleep quality when the PLwD under their care wandered (r (58) = .33, P = .009, Creese et al., 2008), (r = .33, P = .03, Flaskerud et al., 2000), (β = .209, P < .05, L. M. Peng et al., 2018) and/or had to go to the bathroom (r(58) = .34, P = .008, Creese et al., 2008). Similarly, Simpson and Carter (2013a) found that the PLwD going to the bathroom during the night was the most common reason reported for sleep disturbance. Also, caring for a PLwD who had fallen in the last month (odds ratio = 2.33, P < .01) was significantly associated with caregivers’ greater nighttime awakenings (Leggett et al., 2018). Between 28% and 67% (Creese et al., 2008; Simpson & Carter, 2013a; Wilcox & King, 1999) of caregivers reported being disturbed during the night by the PLwD, with approximately 60% of the awakenings occurring 3 or more times a week (Wilcox & King, 1999). Increased PLwD sleep disturbance correlated with poorer appraisal of caregivers’ sleep (r = .22, P = .05, Simpson & Carter, 2013a). Gibson and Gander (2020) also reported that PLwD sleep disturbance predicted caregivers’ poor sleep (β = .23, p = .043). Finally, death of the spouse who had dementia increased caregivers’ nighttime WASO (P = .002) and daytime TST (P = .003), while nighttime sleep percent decreased (P = .009, Von Känel et al., 2012) while the placement of the AD spouse in a long-term care facility was not significantly associated with changes in caregiver sleep.

Discussion

In this comprehensive systematic review, we found that multiple predisposing and precipitating factors were associated with caregiver sleep disturbance (Spielman et al., 1987), See, Figure 2. Many of these factors are modifiable antecedents to poor caregiver sleep and can thus be targeted for interventions to prevent or reduce caregiver sleep disturbance.

Figure 2.

Predisposing and precipitating correlates of sleep disturbance.

Predisposing demographic factors such as age, sex, education, and cohabitation status, and predisposing physiological factors like genetics and biomarkers were all correlated with some aspect of caregiver sleep disturbance. Results pertaining to gender and sleep in caregivers of persons living with dementia are similar to results seen in the general populations where women are at an increased risk of experiencing insomnia (Zhang & Wing, 2006). Moreover, women tend to have an advance in the diurnal and circadian rhythms of sleep and alertness compared to men, meaning women tend to have early evening bedtimes and morning awakenings (Boivin et al., 2016) and they usually experience a change in their sleep patterns when they are experiencing menopause (McCurry et al., 2015).

The results were equivocal regarding age and sleep in caregivers of PLwD. In general, sleep patterns change with age leading to greater wake after sleep onset, earlier awakenings and less time spent in REM sleep (Suzuki et al., 2017), so it is probable that the relationship between increased age in caregivers and greater sleep disturbance could be because caregivers are aging instead of their caregiving responsibilities (Ohayon, 2002). Older caregivers may spend more time awake in bed because of chronic conditions such as urinary frequency, pain or neurological changes associated with the sleep/wake regulation (Beaudreau et al., 2008). However, there is evidence that younger caregivers may have more sleep disturbance than older caregivers (Koyanagi et al., 2018). Cohabiting caregivers, regardless of whether or not they share a room, may report more sleep disturbance during the night because PLwD may awaken them multiple times as they are moving around in bed or the house, or to ask for assistance (Creese et al., 2008; Gehrman et al., 2018; Simpson & Carter, 2013b). In addition, caregivers may become hyper-vigilant and tend to be light sleepers in anticipation of meeting the PLwD nighttime needs (Arber & Venn, 2011; Gibson et al., 2014; Simpson & Carter, 2013c).

Precipitating factors like depressive symptoms, caregiver responsibilities, and behavioral and psychological symptoms of dementia are modifiable factors associated with caregiver sleep. Since there are numerous caregiver psychoeducational interventions which reduce depressive symptoms, caregiver sleep should be assessed before and after they complete the intervention to determine whether they affect sleep (Brewster et al., 2020; Cheng et al., 2019; Hepburn et al., 2021; Zhao et al., 2019). In addition, caregivers should be encouraged to engage in pleasant activities and taught how to perform nursing tasks or provided with a trained professional to assist with these tasks to reduce some of their caregiving responsibilities. Finally, interventions should include both the PLwD and caregiver dyads since they usually have an interdependent sleep pattern.

Current effective caregiver sleep interventions have attempted to target stress reduction/management, self-efficacy, social support, reduction of patient disruptive behaviors, assistance with locating community resources, and education on tailoring sleep hygiene and compression strategies (Lorig et al., 2019; McCurry et al., 2007, 1998, 2015). Thus, it may be prudent to simultaneously manage both the sleep disturbance and the antecedents of sleep disturbance. These interventions may also increase the time that caregivers’ spend in REM sleep; since, caregivers spend less time in REM sleep (Fonareva et al., 2011). REM sleep plays a role in memory consolidation and REM sleep deprivation impairs memory formation (Boyce et al., 2017; Peever & Fuller, 2017). An extended period of decreased REM may increase a person’s risk for cognitive impairment (Della Monica et al., 2018) thus making them more likely to have difficulty with the tasks associated with their role.

Future research should examine how factors like gender, race/ethnicity, culture, and geographical location influence caregiver sleep. For instance, gender may drive the incidence of sleep apnea and needs to be further examined as it pertains to insomnia in caregivers. In addition, there is a disparity in sleep outcomes for Black adults compared to White adults (Hoggard & Hill, 2018; Ownby et al., 2010). Future research can explore whether this disparity also exists when comparing caregivers of different races and ethnicities. Gaps also exist in the research about the impact of relationship dynamics and/or emotional attachment between caregivers and the PLwD on sleep outcomes. More qualitative or mixed-method studies are needed to explore these dynamics and how they relate to sleep. Finally, more research is needed about how distinct types of care-related tasks contribute to caregiver sleep disturbance.

Limitations

Some studies had sample sizes of less than 30 participants and having a small sample size limits the findings from being extrapolated to larger populations and reduces the chance of detecting a true effect (Button et al., 2013; Faber & Fonseca, 2014). Most of the studies were cross-sectional, which prevented the assessment of temporal relationships, as well as identification of causality or directionality (Wang & Cheng, 2020). Longitudinal studies are needed to monitor the course and progression of caregiver sleep disturbance. Generalizability is limited because most studies used very strict inclusion and exclusion criteria: most of the participants were female, Caucasian, and older adults.

Most of the studies used the PSQI, a subjective measure, to assess sleep quality. Since the PSQI asks participants to rate sleep over the past month, it does not capture day-to-day variability in sleep and often does not correlate with actigraphy data in caregivers (Creese et al., 2008; Merrilees et al., 2014; Von Känel et al., 2014). Future studies should also use sleep diaries in conjunction with actigraphy to capture day-to-day variability which is prevalent in caregivers (Brewster et al., 2015). Collecting objective and subjective data would also allow researchers to account for the sleep state misperception often reported by caregivers (D’Aoust et al., 2015).The 5 studies in this review which used polysomnography (PSG) captured the data in the participants’ homes, which is a more accurate reflection of reality than a single night in a laboratory (McKibbin et al., 2005). In-home PSG captures caregiving responsibilities and environmental influences, and should be recorded at least three nights to better capture day-to-day variability (Castro et al., 2009). Additional studies should also examine the relationship between the various stages of sleep and caregiving activity since only one study in this review investigated this relationship (Fonareva et al., 2011).

Conclusion

Sleep disturbance in caregivers of persons living with dementia are associated with multiple predisposing and precipitating factors. Identification and management of the modifiable predisposing and precipitating factors via multicomponent interventions is critical for caregivers at risk for sleep disturbance.

Clinical implications.

• Given the numerous caregiver risk factors and pervasive impact of poor sleep, healthcare providers should better identify those caregivers who currently have, or who are at risk for developing sleep disturbance.

• Caregiver sleep assessment should include questions about sleep patterns, and predisposing (e.g., age, gender) and precipitating (e.g., psychological symptoms, caregiver responsibilities, presence of behavioral and psychological symptoms of dementia) factors of poor sleep.

• Caregiver interventions like coping skills, pleasant activity scheduling, or activities to increase positive affect, which target the modifiable precipitating factors should be recommended to prevent or reduce the incidence of sleep disturbance.

Data Availability Statement

Data sharing is not applicable to this article as no new data were created or analyzed in this study.