Secondary Logo

Journal Logo

ORIGINAL ARTICLES

Sleep Characteristics, Mood, Somatic Symptoms, and Self-Care Among People With Heart Failure and Insomnia

Breazeale, Stephen; Jeon, Sangchoon; Hwang, Youri; O’Connell, Meghan; Nwanaji-Enwerem, Uzoji; Linsky, Sarah; Yaggi, H. Klar; Jacoby, Daniel L.; Conley, Samantha; Redeker, Nancy S.

Author Information
doi: 10.1097/NNR.0000000000000585

Abstract

Chronic heart failure (HF) affects approximately 6.5 million Americans and results in $108 billion in direct and indirect costs annually (Centers for Disease Control and Prevention, 2022; Lesyuk et al., 2018). The prevalence and costs of HF are expected to increase because of an aging population and treatments that increase life expectancy (Lesyuk et al., 2018). For people with HF, managing day-to-day healthcare needs through self-care is critical to physical and psychological well-being (Riegel et al., 2016). Better self-care is associated with important outcomes for people with HF, including reduced hospitalizations (Ma, 2019) and better emotional and physical health-related quality of life (Riegel et al., 2009).

Self-care is a “naturalistic decision-making process that influences actions that maintain physiologic stability, facilitate the perception of symptoms, and direct the management of those symptoms” (Riegel et al., 2016, p. 226). Self-care in people with HF occurs through three interrelated processes: maintenance (the degree to which a person adheres to a care plan and exhibits healthy behaviors), symptom perception (an individual’s ability to identify and apply meaning to physical sensations), and management (an individual’s response to symptoms when present; Riegel et al., 2016). Self-care confidence, or the perceived ability to engage in self-care, influences each self-care process (Vellone et al., 2013). Examples of self-care activities that people with HF need to perform include taking prescribed medications, maintaining a low salt diet, titrating diuretic dosages, and noticing changes in HF symptom severity (Riegel et al., 2016).

People with HF often report disturbed mood (e.g., anxiety, depression, and stress) and somatic symptoms such as dyspnea, fatigue, and sleepiness (Gaffey et al., 2021; Park et al., 2019; Rasmussen et al., 2020). Abnormal sleep characteristics, including sleep disturbance; self-reported sleep quality; short, fragmented, and irregular sleep; and specific sleep disorders such as chronic insomnia and sleep-disordered breathing are also common (Redeker et al., 2010; Spedale et al., 2021). Approximately 50% of people with HF have chronic comorbid insomnia (Redeker et al., 2010), marked by difficulty initiating or maintaining sleep, and up to 80% have sleep-disordered breathing (Mentz & Fiuzat, 2014), including obstructive and central sleep apnea. Sleep characteristics, mood, and somatic symptoms are also closely related and may collectively influence self-care among people with HF (Jeon & Redeker, 2016).

Among people with HF, daytime sleepiness is associated with decrements in self-care. In contrast, poorer self-reported global sleep quality is associated with poor medication adherence (Spedale et al., 2021)—an essential component of self-care maintenance (Vellone et al., 2013). Mood disturbance and somatic symptoms are associated with rehospitalizations and poorer self-care (Chuang et al., 2019; Ma, 2019). However, most research in this area focused only on the associations of single symptoms with self-care. These approaches fail to account for the potential synergistic, multivariate associations of self-care with sleep characteristics and mood and somatic symptoms that better reflect the symptom burden that people with HF often report (Jeon & Redeker, 2016; Park et al., 2019).

The aims of this study were to (a) evaluate the levels of self-care maintenance and self-care confidence among people with stable HF and chronic insomnia; (b) identify the clinical and demographic correlates of self-care maintenance and confidence among people with stable HF and chronic insomnia; and (c) identify the associations between sleep characteristics, mood and somatic symptoms, and self-care maintenance and confidence among people with stable HF and chronic insomnia.

METHODS

Study Design

This cross-sectional study includes baseline data from a randomized controlled trial of cognitive behavioral therapy for insomnia in people with stable HF (ClinicalTrials.gov Identifier NCT02827799). This study is a post hoc analysis of data from the parent study (Redeker et al., 2017). Participants were randomized in groups after baseline assessment to 8 weeks of cognitive behavioral therapy for insomnia or HF self-management education provided in a group format. The Human Research Protection Program approved the study at Yale University, and all participants provided written informed consent.

Sample and Setting

Participants were recruited from March 2016 to September 2019 through the Yale New Haven Hospital Advanced Heart Failure Program and the Veterans Affairs Connecticut Health Care System. We identified eligible participants by reviewing medical records, face-to-face recruitment in the HF disease management clinic, advertisements via posters, messages in the electronic medical record, or postings on the research center’s website (Conley et al., 2020). We conducted a final screening and obtained written informed consent in person, after which we administered the study questionnaires. Participants completed the questionnaires at home and returned them in person or via mail with prepaid postage.

Participants were eligible for inclusion if they met the following criteria: (a) aged 18 years or older; (b) New York Heart Association (NYHA) Functional Classes I–IV; (c) cognitively intact; (d) spoke and understood English; (e) concerned about sleep for at least 1 month; (f) reported Insomnia Severity Index (ISI) scores of ≥8, suggesting at least mild insomnia; (g) diagnosed with HF with preserved or reduced systolic function; (h) had Patient Health Questionnaire-9 scores of <20, indicating no more than moderately severe depression; and (i) either had an Apnea–Hypopnea Index of <15 within the last year or had significant sleep-disordered breathing and were adherent to continuous positive airway pressure (CPAP) therapy for at least 6 hours per night for at least 3 months before enrollment.

Participants with unknown sleep apnea status were screened using the Apnea Risk Evaluation System (Popovic et al., 2009), a forehead-worn Level III sleep monitoring device, for two consecutive nights in their homes. Participants with an Apnea–Hypopnea Index of ≥15 (moderate or severe sleep apnea) were excluded from the study and referred to a sleep medicine specialist for evaluation.

Variables and Measures

Clinical and Demographic Factors

We collected demographic information, including age, biological gender, body mass index (BMI), and self-reported race and ethnicity. We conducted a medical record review to elicit comorbidity, medications, and characteristics of HF, including left ventricular ejection fraction (LVEF), NYHA functional classification, prescribed beta-blockers, angiotensin-converting enzyme inhibitors, angiotensin II receptor blockers, diuretics, and/or statin medications. LVEF was recorded from participants’ most recent echocardiograms and further classified as HF with reduced ejection fraction (HFrEF; LVEF < 45%) or HF with preserved ejection fraction (HFpEF; LVEF ≥ 45%). NYHA functional class was either recorded directly from clinical documentation or calculated from reports of participant function in the medical record. The Charlson Comorbidity Index (CCI) was calculated via chart review. The CCI is a valid instrument calculated by assessing 16 conditions (e.g., acquired immunodeficiency syndrome and chronic obstructive pulmonary disease), each of which is given a weighted score (Charlson et al., 1987). Higher CCI scores indicate higher comorbidity and estimate 10-year survival probability (Charlson et al., 1987).

Sleep Characteristics

We assessed participants’ sleep characteristics, including insomnia severity, sleep quality, and sleep disturbance. Higher scores on each instrument described below indicate worse insomnia severity, sleep quality, and sleep disturbance (Bastien et al., 2001; Buysse et al., 1989; HealthMeasures, 2018).

We assessed insomnia severity with the seven-item ISI (Bastien et al., 2001). The ISI uses a recall period of 14 days and is valid for research in individuals with sleep disorders (Bastien et al., 2001). The ISI was reliable in our sample (Cronbach’s α = .74).

We assessed sleep quality with the single-item Component 1 of the Pittsburgh Sleep Quality Index (PSQI) entitled “Subjective Sleep Quality” (Buysse et al., 1989). The PSQI uses a 30-day recall period (Buysse et al., 1989). The PSQI is valid for research among individuals with sleep disorders (Buysse et al., 1989). Reliability calculations are not needed with a single-item instrument such as Component 1 of the PSQI.

We assessed sleep disturbance with the eight-item Patient-Reported Outcomes Measurement Information System (PROMIS) Sleep Disturbance Short Form 8a (HealthMeasures, n.d.). The PROMIS Sleep Disturbance instrument elicits individuals’ perceived difficulties with initiating and maintaining sleep, as well as their satisfaction with their sleep (HealthMeasures, 2018). The PROMIS Sleep Disturbance Short Form 8a uses a 7-day recall period (HealthMeasures, 2018). The PROMIS Sleep Disturbance instrument is valid for use in research among individuals with sleep disorders (HealthMeasures, 2018) and was highly reliable in our sample (Cronbach’s α = .87).

Mood and Somatic Symptoms

We used a series of patient-reported outcome measures to assess the severity of mood and somatic symptoms, including anxiety, depression, dyspnea, fatigue, perceived stress, and sleep impairment. Higher scores on each instrument indicate more severe symptoms (Cohen et al., 1983; HealthMeasures, n.d.; Redeker, 2006).

Perceived stress was assessed using the 10-item Perceived Stress Scale (PSS; Cohen et al., 1983). The PSS uses a 30-day recall period and is valid for research among individuals with insomnia (Cohen et al., 1983). The PSS was reliable in our sample (Cronbach’s α = .91).

Dyspnea was assessed using the 16-item Multidimensional Assessment of Dyspnea Scale (MADS; Redeker, 2006). The MADS uses a recall period of 7 days and is valid among individuals with insomnia (Redeker, 2006). The MADS was highly reliable in our sample (Cronbach’s α = .96).

Symptoms of anxiety, depression, fatigue, and sleep impairment were assessed using the corresponding eight-item PROMIS instruments (HealthMeasures, n.d.). The PROMIS Sleep-Related Impairment instrument assesses individuals’ perceived daytime functional impairments related to sleep problems (HealthMeasures, 2020). The PROMIS instruments use a 7-day recall period and are valid for use among individuals with insomnia (HealthMeasures, n.d.). The PROMIS anxiety (Cronbach’s α = .94), depression (Cronbach’s α = .94), fatigue (Cronbach’s α = .95), and sleep impairment (Cronbach’s α = .89) instruments were highly reliable in our sample.

Self-Care

We assessed self-care with the Self-Care of Heart Failure Index (SCHFI) Version 6.2 (Vellone et al., 2013). The SCHFI is a 22-item self-report instrument that uses a 30-day recall period to assess three domains of self-care in HF: maintenance, management, and confidence (Vellone et al., 2013). We assessed participants’ self-care maintenance, management, and confidence in this study but excluded self-care management from our analyses because we wanted to include the entire cohort in our study, and not all participants responded to the self-care management subscale because they did not experience HF symptoms that required utilizing the management strategies included in this measure. The developers of the SCHFI recommend the use of each subscale of the SCHFI independent of the others, which allowed us to exclude the self-care management subscale from our analysis without affecting the analysis or interpretation of the self-care maintenance and confidence subscales (Vellone et al., 2013).

Ten items were used to elicit the frequency with which respondents completed activities necessary for maintaining HF-related health (e.g., weighing oneself and using a pillbox; Vellone et al., 2013). Self-care management was assessed via six items that assessed respondents’ abilities to recognize and treat HF symptoms when present (e.g., taking an extra water pill and contacting a clinician; Vellone et al., 2013). Self-care confidence was assessed with six items that assessed respondents’ confidence in engaging in self-care processes (e.g., following treatment advice and recognizing changes in health; Vellone et al., 2013). Each item is scored on a 1–4 Likert scale, with higher scores indicating better self-care in each domain (Vellone et al., 2013). Scores for each domain are standardized to range 0–100, with scores of ≥70 indicating adequate self-care (Vellone et al., 2013). The SCHFI is valid for HF research (Vellone et al., 2013). Reliabilities of the SCHFI subscales in our study included self-care maintenance (Cronbach’s α = .61), self-care management (Cronbach’s α = 0.57), and self-care confidence (Cronbach’s α = .86).

Data Analysis

We used IBM SPSS Statistics (Version 27) and SAS (Version 9.4) for the analyses. We calculated descriptive statistics for the demographic and clinical characteristics, including means and standard deviations for continuous variables and frequencies and percentages for categorical variables. We examined the histograms, kurtosis, and skewness of the continuous variables to assess whether such variables were normally distributed. We calculated bivariate correlations between the demographic and clinical characteristics, mood and somatic symptoms, and self-care maintenance and confidence using Pearson’s correlation coefficient for normally distributed variables and Spearman’s rank correlation coefficient for nonnormally distributed variables. Significance was determined using p < .05.

Identifying Latent Factors

Before running the multivariate analyses, we assessed whether the data fit a multivariate normal distribution by inspecting a qq plot of the squared Mahalanobis distances of the data. The squared Mahalanobis distances of the data are plotted against values that would be expected in a multivariate normal distribution (Pituch & Stevens, 2016). If the data fit a multivariate normal distribution, the distribution of the actual squared Mahalanobis distances will be approximately linear (Pituch & Stevens, 2016).

We performed exploratory factor analysis (EFA) with IBM SPSS Statistics (Version 27) to identify the underlying factor structure of the sleep characteristics and mood and somatic symptoms. Extraction was accomplished using the full information maximum likelihood method, which handled missing data on the indicator variables. We assessed sampling adequacy using the Kaiser–Meyer–Olkin Measure of Sampling Adequacy. We identified the model by requiring the eigenvalue of an included factor to be >1 and examining the scree plot of the eigenvalues. We performed a varimax rotation of the final factor matrix to ensure that each indicator variable loaded strongly on one factor only. Model fit was tested using the goodness-of-fit test using p < .05. We labeled each identified factor qualitatively, a standard approach in EFA (Pituch & Stevens, 2016).

Examining the Associations of Latent Factors on Self-Care

Next, we performed confirmatory factor analysis (CFA) to check whether the data fit the measurement model identified in EFA analysis. The convergent validity of the CFA was assessed using factor loadings, average variance extracted (AVE) > 0.5, and composite reliability (CR) > .7. The discriminant validity was assessed by comparing the correlation coefficients (r) of the latent factors with AVE and was deemed acceptable if r2 < AVE. The CFA model was identified using model fit indices, including standardized chi-square (χ2/df) < 2, standardized root-mean-square residual (SRMR) < .1, root-mean-square error of approximation (RMSEA) < .8, and the comparative fit index (CFI) > .9 (Pituch & Stevens, 2016).

We performed structural equation modeling to examine the associations of the identified latent factors on self-care maintenance and confidence using the PROC CALIS procedure in SAS (Version 9.4). First, we estimated the unadjusted associations of each latent factor on self-care maintenance and confidence while considering a correlation between the two self-care outcomes, with a significance level of p < .05. We used the significant unadjusted relationships to test the final model. We examined the adjusted relationships of the latent factors, selected demographic and clinical characteristics, and the two self-care outcomes. We selected the demographic and clinical characteristics for the final model by running multiple general linear models in SAS (Version 9.4), regressing the self-care outcomes on the demographic and clinical characteristics. Significance was determined using p < .05. The final model was identified using the fit indices as follows: standardized chi-square (χ2/df) < 2, the adjusted goodness of fit index (aGFI) > .90, RMSEA < .05, and the CFI > .90 (Pituch & Stevens, 2016).

RESULTS

We obtained baseline data from 195 participants. Their demographic and clinical characteristics appear in Table 1 and were published previously (Conley et al., 2020). The sample was predominantly male (56.9%), with 17.9% of participants self-identified as Black race, 74.9% self-identified as White, and 7.2% were of other racial backgrounds. Most were obese (54.6%) and had HFpEF (66.5%). Most of the participants (56%) had clinical insomnia (ISI score ≥ 15), whereas 44.1% had mild insomnia (ISI score = 8–14). Over half (52.3%) of the sample had sleep-disordered breathing and adhered to CPAP therapy.

TABLE 1 - Demographic and Clinical Characteristics (N = 195)
Variable Mean (SD)/n (%)
Age (years) 63.0 (12.8)
Biological gender
 Female 84 (43.1%)
 Male 111 (56.9%)
Race
 White 146 (74.9%)
 African American 35 (17.9%)
 Native American 1 (0.5%)
 Asian 1 (0.5%)
 Other 12 (6.2%)
Body mass index
 <18.5 3 (1.6%)
 18.5 to <25 37 (20.0%)
 25 to <30 44 (23.8%)
 30+ 101 (54.6%)
New York Heart Association functional classification
 I 60 (31.1%)
 II 88 (45.6%)
 III 40 (20.7%)
 IV 5 (2.6%)
Ejection fraction (%)
 <45 64 (33.5%)
 ≥45 127 (66.5%)
Charlson Comorbidity Index 2.8 (1.9)
Health history
 Diabetes 68 (35.1%)
 Hypertension 123 (63.4%)
 COPD 45 (23.2%)
 Peripheral vascular disease 30 (15.5%)
 Myocardial infarction 54 (27.8%)
 Sleep apnea/CPAP use 102 (52.3%)
 Pacemaker 50 (25.8%)
 LVAD 3 (1.6%)
Heart failure medications
 ACE-I or ARB 101 (49.2%)
 Beta blocker 129 (66.2%)
 Statin 118 (60.5%)
 HCTZ 8 (5.7%)
 Loop diuretic 126 (72.0%)
Insomnia Severity Index 15.2 (4.6)
 Mild (8–14) 86 (44.1%)
 Clinical insomnia (15–28) 109 (56.0%)
Note. COPD = chronic obstructive pulmonary disease; CPAP = continuous positive airway pressure; LVAD = left ventricular assist device; ACE-I = angiotensin-converting enzyme inhibitor; ARB = angiotensin II receptor blocker; HCTZ = hydrochlorothiazide.

The majority had inadequate self-care maintenance, confidence, and management based on normative data (SCHFI subscale score < 70). Only 36.4% (n = 71), 44.6% (n = 87), and 28.7% (n = 56) had adequate self-care maintenance, confidence, and management, respectively. Descriptive statistics of the individual SCHFI items are presented in Supplemental Digital Content (SDC) 1 (https://links.lww.com/NRES/A428), and the SCHFI subscale scores are presented in SDC 2 (https://links.lww.com/NRES/A429).

Male gender (r = −.15, p < .05), higher BMI (r = −.16, p < .05), and HFpEF (r = −.17, p < .05) were negatively correlated with self-care maintenance, while being of White race compared to being of other racial backgrounds (r = .22, p < .01) was associated with higher levels of self-care maintenance. Age (r = −.22, p < .01) and HFpEF (r = −.16, p < .05) were negatively correlated with self-care confidence.

In the general linear models, adjusted for the demographic and clinical characteristics, HFrEF (β = 0.396, p < .01), and self-reported White race (β = 0.545, p < .001) were positively associated with self-care maintenance, whereas age (β = −0.194, p < .01) was negatively associated with self-care confidence. BMI, NYHA functional class, CCI scores, and sleep apnea/CPAP use were not associated with self-care maintenance or confidence.

Insomnia severity (r = −.19, p < .01), sleep disturbance (r = −.20, p < .01), and poorer sleep quality (r = −.14, p < .05) were negatively correlated with self-care maintenance. Anxiety (r = −.18, p < .05), depression (r = −.23, p < .01), dyspnea (r = −.18, p < .05), fatigue (r = −.24, p < .01), and sleep impairment (r = −.22, p < .01) were negatively correlated with self-care confidence. Perceived stress was negatively correlated with both self-care maintenance (r = −.18, p < .05) and self-care confidence (r = −.26, p < .01). The bivariate correlations between sleep characteristics, mood and somatic symptoms, and self-care maintenance and confidence are presented in Table 2.

TABLE 2 - Bivariate Correlations Between Symptoms and Sleep Characteristics and Self-Care Outcomes
Variable Self-care maintenance Self-care confidence
Insomnia severity −.19** −.04
Sleep disturbance −.20** −.07
Sleep qualitya −.14* −.12
Anxiety −.06 −.18*
Depression −.11 −.23**
Dyspnea −.02 −.18*
Fatigue −.06 −.24**
Perceived stress −.18* −.26**
Sleep impairment −.12 −.22**
a Spearman’s rank-order correlation was used.
*p < .05.
** p < .01.

We identified the latent factors that represented the sleep, mood, and somatic symptom variables. We assessed the qq plot of the squared Mahalanobis distances. We observed that the data distribution was approximately linear, indicating that the data fit a multivariate normal distribution and were appropriate for the subsequent analyses. The Kaiser–Meyer–Olkin measure of sampling adequacy was .805, indicating the data were appropriate for factor analysis. We identified a three-factor solution as the best fit because the eigenvalues of the first three factors were 4.14, 1.5, and 1.05, respectively, and the scree plot revealed an elbow and leveling out after the third factor. The three-factor solution explained 74.34% of the variance in the data and was a good fit as the goodness-of-fit test was insignificant (p = .744). Factor 1, labeled mood, represented perceived stress, anxiety, and depression. Factor 2, labeled sleep characteristics, was defined by insomnia severity, sleep disturbance, and sleep quality. Factor 3, labeled somatic symptoms, was represented by sleep impairment, dyspnea, and fatigue. The rotated factor matrix is presented in Table 3.

TABLE 3 - Rotated Factor Matrix of Sleep and Symptom Variables
Variable Factor
1 2 3
Insomnia severity .224 .486a .198
Sleep disturbance .096 .787a .201
Sleep quality .104 .838a .182
Anxiety .824a .175 .170
Depression .812a .127 .284
Dyspnea .103 .189 .481a
Fatigue .289 .153 .945a
Perceived stress .801a .144 .195
Sleep impairment .293 .309 .614a
Note. Extraction completed using maximum likelihood. Rotation completed using varimax technique with Kaiser normalization.
a Indicates strongest loading of a given variable.

CFA indicated that the latent factors were valid constructs and represented the data well. The average variances extracted and composite reliabilities for the latent factors mood (AVE = 0.73, CR = .89), sleep characteristics (AVE = 0.57, CR = .79), and somatic symptoms (AVE = 0.58, CR = .80) indicated acceptable convergent validity. The model had acceptable discriminant validity because the squared correlation coefficients of the latent factors were less than their corresponding AVEs. The latent factor mood significantly correlated with somatic symptoms (r = .601, p < .05), somatic symptoms significantly correlated with sleep characteristics (r = .505, p < .05), and sleep characteristics significantly correlated with mood (r = .364, p < .05). Model fit indices, including standardized χ22/df = 1.38), SRMR (.05), RMSEA (.05), and the CFI (.99) all indicated that the three-factor solution was a good representation of the data (Pituch & Stevens, 2016). The full CFA is presented in Figure 1.

F1
FIGURE 1:
Confirmatory factor analysis. CFI = comparative fit index; RMSEA = root-mean-square error of approximation; SRMR = standardized root-mean-square residual. *p < .05.

In the unadjusted models, the latent factor sleep characteristics was negatively associated with self-care maintenance (r = −.183, p < .05), but the latent factors mood and somatic symptoms were not. On the other hand, the latent factors mood (r = −.259, p < .05) and somatic symptoms (r = −.243, p < .05), but not the latent factor sleep characteristics, were associated with self-care confidence. Self-care maintenance and self-care confidence significantly covaried in each of these unadjusted models. The unadjusted standardized coefficients of the latent factors on the self-care outcomes are presented in SDC 3 (https://links.lww.com/NRES/A430).

The structural equation modeling revealed that self-care maintenance and confidence (r = .24, p < .05) remained significantly correlated with one another. The relationships between the demographic and clinical characteristics and the self-care components remained significant and unchanged. The latent factor sleep characteristics (r = −.194, p < .01) was negatively associated with self-care maintenance. However, the relationships between the latent factors mood (r = −.172, p = .07) and somatic symptoms (r = −.159, p = .1) and self-care confidence were no longer statistically significant. In the final model, the latent factor sleep characteristics, HFrEF, and self-identifying as White race were significantly positively associated with self-care maintenance, whereas increasing age alone was significantly correlated with self-care confidence. Model fit indices including standardized χ22/df = 1.21), aGFI (.91), RMSEA (.03), and the CFI (.99), all indicated that the final model represented the data well. The full structural equation model is presented in Figure 2.

F2
FIGURE 2:
Structural equation model. aGFI = adjusted goodness of fit index; CFI = comparative fit index; LVEF = left ventricular ejection fraction; RMSEA = root-mean-square error of approximation. *p < .05.

DISCUSSION

This study expands evidence about the associations between sleep, mood, symptoms, and self-care among people with HF by considering the multidimensional aspects of these constructs. Our most notable findings were the negative associations of poor sleep with self-care maintenance, but not self-care confidence; the separate associations between the latent factors somatic symptoms and mood, and self-care confidence; and the use of a multivariate approach that considered the simultaneous contributions of multiple aspects of sleep, somatic symptoms, and mood. Our multivariate approach better reflects the relationships among these phenomena than most previous studies that primarily employed bivariate approaches with single-symptom, mood, and sleep variables.

The correlations between the latent factor sleep characteristics and self-care maintenance suggest that poor sleep may have a negative influence on the ability to complete health behaviors needed to manage HF and extends studies documenting associations between poor sleep and medication adherence (Spedale et al., 2021), sleep disorders, and self-care (Kamrani et al., 2014). The magnitude of the relationship between sleep and self-care was similar to findings from a Korean study that used a different HF self-care measure (Ryou et al., 2021). However, a third study found no relationship between sleep and HF self-care (Kessing et al., 2016). To our knowledge, no previous studies of self-care in HF characterized the presence of sleep apnea in the participants. None considered the associations of insomnia—another common sleep disorder known to contribute to daytime dysfunction. Although our findings suggest that sleep contributes to self-care, it is also possible that the inability to manage health behaviors, such as sleep, contributes to poorer health and, in turn, poorer self-care.

Unlike the latent factor sleep characteristics, mood and somatic symptoms were negatively associated with self-care confidence. This construct reflects a person’s perceived ability to perform self-care behaviors in the bivariate analyses. The reasons for this are not completely clear. Negative affectivity may influence a person’s perceptions of their health and abilities and thus their self-care confidence. Conversely, low self-care confidence may contribute to stress and negative affectivity. The associations between mood and somatic symptoms, and self-care confidence were similar to previous work in which depressive symptoms were negatively associated with self-care (Chuang et al., 2019). Our work expands these findings by documenting the negative associations between multiple moods, somatic symptoms, and self-care confidence. Although the relationships between the latent factors mood and somatic symptoms, and self-care confidence were no longer statistically significant in the structural equation model, the magnitude of the associations was similar to that of the relationship between sleep characteristics and self-care maintenance.

The relationships found in this study both support and challenge the relationships posited in the conceptual model of Riegel and Weaver (2009). The unadjusted negative association of the latent factor mood, which includes depressive symptoms, with self-care confidence supports the posited association of depression with poor self-care (Riegel & Weaver, 2009). However, Riegel and Weaver do not include a direct association between poor sleep quality and poor self-care in their conceptual model. Instead, they suggested that depression, poor cognition, and excessive daytime sleepiness mediate the relationship between sleep disturbance and self-care. The association of the latent factor sleep characteristics and self-care maintenance in our structural equation model shows that sleep and self-care are directly associated with one another. Furthermore, the lack of associations between the latent factors mood and somatic symptoms with self-care maintenance and confidence in the structural equation model in our study does not support the idea that depression and daytime sleepiness mediate the relationship between sleep and HF self-care because such a relationship must exist if mediation is present.

Race, age, and LVEF were associated with self-care in our study. The negative association between age and self-care confidence is similar to a previous Italian study (Cocchieri et al., 2015). Age was negatively associated with self-care maintenance and confidence. Still, it contrasts with a Korean study in which younger age was associated with poorer self-care (Ryou et al., 2021), although the Korean researchers used a different measure of HF self-care. Older individuals may experience cognitive impairment that limits their ability to perform self-care (Cocchieri et al., 2015). Still, the ability of younger adults with HF to perform self-care may be hindered by their need to simultaneously manage their professional and self-care needs (Ryou et al., 2021).

HFrEF was positively associated with self-care maintenance but not with self-care confidence. Our findings contrast with findings that LVEF was positively associated with self-care confidence but not self-care maintenance (Zaharova et al., 2021). Despite conflicting results, our findings suggest differences in self-care between those with HFrEF and those with HFpEF. Previous researchers (Kalogirou et al., 2020) proposed that people with HFrEF reported higher self-care maintenance because of the abundance of treatment modalities available to them compared to those with HFpEF—a more heterogeneous condition. However, more research is necessary to evaluate this potential mechanism. It is also possible that there is a need for a self-care measure that focuses on the needs of people with HFpEF.

We found that self-reported race was associated with self-care maintenance, with participants of White race (compared with a group that predominantly reported being African American) reporting higher levels. These findings are consistent with a previous study in which self-identified race was associated with self-care maintenance (Baah et al., 2021) but contrasted with another study of self-care in a predominantly African American sample of men with comorbid Type II diabetes and HF (Aga et al., 2019) who had better self-care than others. These findings merit further consideration given evidence of greater risk for adverse HF outcomes and health equity concerns among underrepresented minority groups (Steinberg et al., 2021), including African Americans. Still, they should be interpreted with caution. Only 17.9% (n = 35) of the sample self-identified as African American, and limited available information on social determinants and access to resources may explain these findings in our study.

The results of this study have several important implications for clinical practice and future research. The significant association of poor sleep characteristics with poorer self-care maintenance in our fully adjusted model suggests the importance of understanding the contributions of multiple sleep characteristics to self-care. Sleep disturbances, including insomnia that was present in all study participants, and the presence of treated sleep-disordered breathing, present in about half, may contribute to subtle or more severe cognitive changes or lack of attention that may contribute to poor self-care—an essential focus of health teaching needed for HF disease management. Although further research is needed, improved assessment and treatment of sleep disturbance may improve self-care of both HF and other comorbid conditions that commonly occur in people with HF, such as diabetes. For example, researchers demonstrated that cognitive behavioral therapy for insomnia (CBT-I; Redeker et al., 2019) and brief behavioral interventions (Harris et al., 2019) improved sleep quality among individuals with HF, and a pilot randomized controlled trial of CBT-I in people with Type 2 diabetes (Alshehri et al., 2020) revealed that people who received CBT-I reported significantly greater improvement in insomnia symptoms, as well as self-care behavior for diabetes, than those in the control group.

Further exploration of the effects of sleep interventions on self-care in people with HF is warranted. Given the multimodal nature of CBT-I, it is possible that providing more efficacious ways of managing sleep-related cognitions improves self-efficacy and mood that in turn contribute to self-care. It is also possible that the effects of CBT-I are through improvement in sleep and cognitive function. However, further study of these mechanisms is needed.

The demographic and clinical characteristics of self-care represent modifiable and nonmodifiable risk factors that can be assessed in future clinical and scientific endeavors. The influences of age, race, and type of HF (HFrEF vs. HFpEF) will help clinicians identify those at highest risk for poor self-care and aid researchers in targeting and tailoring future interventions to improve self-care and long-term outcomes in individuals with both HF and insomnia. Identifying people at high risk for poor HF self-care and providing personalized interventions may increase their ability to perform HF self-care and, by extension, promote equitable HF outcomes such as lower hospitalizations and better health-related quality of life.

The three latent factors we identified support the presence of an underlying structure of the sleep and symptoms experiences people with HF and insomnia report that may be used to support future research and intervention focused on these clustered phenomena. Given that co-occurring symptoms are thought to share common biological mechanisms (Miaskowski & Aouizerat, 2007) and are associated with worse self-care and HF outcomes (Mosarla & Wood, 2021), future research should employ multivariate approaches that examine potential biological mechanisms for sleep, co-occurring symptoms, and self-care to inform future precision health interventions to improve outcomes among people with HF.

Limitations of this study include using a cross-sectional design that precludes causal inferences and its post hoc nature. There is a need for future longitudinal analyses to determine the temporal relationships and causal mechanisms underlying the associations between poor sleep characteristics, mood and somatic symptoms, and self-care among individuals with HF.

The sample size was calculated based upon the aims for the parent study (Redeker et al., 2017), and some of the correlations were small, although similar to those of other work (Cocchieri et al., 2015). The lack of statistical significance was likely due to inadequate statistical power. Further study with a larger sample and additional possible mediator variables, such as sensitive cognitive function measures or objective sleepiness measures, are needed.

Lastly, we excluded self-care management, an individual’s response to symptoms when present (Riegel et al., 2016), from our analyses because we included both symptomatic and nonsymptomatic individuals with HF. However, all study participants had insomnia symptoms. Thus, we were unable to assess the influences of the variables in our study on self-care management.

Conclusion

We identified three latent factors, which we entitled sleep characteristics, mood, and somatic symptoms, that show that multiple sleep, mood, and somatic symptoms co-occur in people with HF and insomnia. Our unadjusted models showed that poor sleep was negatively associated with self-care maintenance and that the latent factors mood and somatic symptoms were negatively associated with self-care confidence. Our fully adjusted model indicates that the latent factor sleep characteristics, race, and LVEF were associated with self-care maintenance and that age was negatively associated with self-care confidence. Our study findings provide an important foundation for future interventions to improve self-care in individuals with HF.

FU1

ORCID iDs

Stephen Breazeale https://orcid.org/0000-0002-2228-0993

Sangchoon Jeon https://orcid.org/0000-0003-2855-2053

Youri Hwang https://orcid.org/0000-0003-3943-5071

Meghan O’Connell https://orcid.org/0000-0002-8293-0150

Uzoji Nwanaji-Enwerem https://orcid.org/0000-0003-2391-7025

Sarah Linsky https://orcid.org/0000-0002-4204-7881

H. Klar Yaggi https://orcid.org/0000-0002-9888-2480

Daniel L. Jacoby https://orcid.org/0000-0002-9182-275X

Samantha Conley https://orcid.org/0000-0002-4501-5244

Nancy S. Redeker https://orcid.org/0000-0001-7817-2708

REFERENCES

Aga F. B., Dunbar S. B., Kebede T., Higgins M. K., Gary R. A. (2019). Correlates of self-care behaviors in adults with Type 2 diabetes and comorbid heart failure. Diabetes Educator, 45, 380–396. 10.1177/0145721719855752
Alshehri M. M., Alothman S. A., Alenazi A. M., Rucker J. L., Phadnis M. A., Miles J. M., Siengsukon C. F., Kluding P. M. (2020). The effects of cognitive behavioral therapy for insomnia in people with Type 2 diabetes mellitus, pilot RCT Part II: Diabetes health outcomes. BMC Endocrine Disorders, 20, 136. 10.1186/s12902-020-00612-6
Baah F. O., Chittams J., Carlson B., Sethares K. A., Daus M., Moser D. K., Riegel B. (2021). Sociodemographic indicators of social position and self-care maintenance in adults with heart failure. Clinical Nursing Research, 30, 847–854. 10.1177/1054773821995593
Bastien C. H., Vallières A., Morin C. M. (2001). Validation of the Insomnia Severity Index as an outcome measure for insomnia research. Sleep Medicine, 2, 297–307. 10.1016/S1389-9457(00)00065-4
Buysse D. J., Reynolds C. F. 3rd, Monk T. H., Berman S. R., Kupfer D. J. (1989). The Pittsburgh Sleep Quality Index: A new instrument for psychiatric practice and research. Psychiatry Research, 28, 193–213. 10.1016/0165-1781(89)90047-4
Centers for Disease Control and Prevention. (2022). Heart failure. https://www.cdc.gov/heartdisease/heart_failure.htm
Charlson M. E., Pompei P., Ales K. L., MacKenzie C. R. (1987). A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. Journal of Chronic Diseases, 40, 373–383. 10.1016/0021-9681(87)90171-8
Chuang H. W., Kao C. W., Lin W. S., Chang Y. C. (2019). Factors affecting self-care maintenance and management in patients with heart failure: Testing a path model. Journal of Cardiovascular Nursing, 34, 297–305. 10.1097/JCN.0000000000000575
Cocchieri A., Riegel B., D’Agostino F., Rocco G., Fida R., Alvaro R., Vellone E. (2015). Describing self-care in Italian adults with heart failure and identifying determinants of poor self-care. European Journal of Cardiovascular Nursing, 14, 126–136. 10.1177/1474515113518443
Cohen S., Kamarck T., Mermelstein R. (1983). A global measure of perceived stress. Journal of Health and Social Behavior, 24, 385–396. 10.2307/2136404
Conley S., O’Connell M., Linsky S., Moemeka L., Darden J. W. 4th, Gaiser E. C., Jacoby D., Yaggi H., Redeker N. S. (2020). Evaluating recruitment strategies for a randomized clinical trial with heart failure patients. Western Journal of Nursing Research, 43, 785–790. 10.1177/0193945920970229
Gaffey A. E., Jeon S., Conley S., Jacoby D., Ash G. I., Yaggi H. K., O’Connell M., Linsky S. J., Redeker N. S. (2021). Perceived stress, subjective, and objective symptoms of disturbed sleep in men and women with stable heart failure. Behavioral Sleep Medicine, 19, 363–377. 10.1080/15402002.2020.1762601
Harris K. M., Schiele S. E., Emery C. F. (2019). Pilot randomized trial of brief behavioral treatment for insomnia in patients with heart failure. Heart & Lung, 48, 373–380. 10.1016/j.hrtlng.2019.06.003
HealthMeasures. (n.d.). Intro to PROMIS®. https://www.healthmeasures.net/explore-measurement-systems/promis/intro-to-promis
HealthMeasures. (2018, July 13). Sleep disturbance: A brief guide to the PROMIS® Sleep Disturbance instruments. http://www.healthmeasures.net/administrator/components/com_instruments/uploads/PROMIS%20Sleep%20Disturbance%20Scoring%20Manual%20(2).pdf
HealthMeasures. (2020, April 14). Sleep-related impairment: A brief guide to the PROMIS® sleep-related impairment instruments. http://www.healthmeasures.net/images/promis/manuals/PROMIS_Sleep-Related_Impairment_Scoring_Manual.pdf
Jeon S., Redeker N. S. (2016). Sleep disturbance, daytime symptoms, and functional performance in patients with stable heart failure: A mediation analysis. Nursing Research, 65, 259–267. 10.1097/NNR.0000000000000169
Kalogirou F., Forsyth F., Kyriakou M., Mantle R., Deaton C. (2020). Heart failure disease management: A systematic review of effectiveness in heart failure with preserved ejection fraction. ESC Heart Failure, 7, 195–213. 10.1002/ehf2.12559
Kamrani A. A., Foroughan M., Taraghi Z., Yazdani J., Kaldi A. R., Ghanei N., Mokhtarpour R. (2014). Self-care behaviors among elderly with chronic heart failure and related factors. Pakistan Journal of Biological Sciences, 17, 1161–1169. 10.3923/pjbs.2014.1161.1169
Kessing D., Denollet J., Widdershoven J., Kupper N. (2016). Fatigue and self-care in patients with chronic heart failure. European Journal of Cardiovascular Nursing, 15, 337–344. 10.1177/1474515115575834
Lesyuk W., Kriza C., Kolominsky-Rabas P. (2018). Cost-of-illness studies in heart failure: A systematic review 2004–2016. BMC Cardiovascular Disorders, 18, 74. 10.1186/s12872-018-0815-3
Ma C. (2019). Rehospitalisation rates and associated factors within 6 months after hospital discharge for patients with chronic heart failure: A longitudinal observational study. Journal of Clinical Nursing, 28, 2526–2536. 10.1111/jocn.14830
Mentz R. J., Fiuzat M. (2014). Sleep-disordered breathing in patients with heart failure. Heart Failure Clinics, 10, 243–250. 10.1016/j.hfc.2013.10.001
Miaskowski C., Aouizerat B. E. (2007). Is there a biological basis for the clustering of symptoms?Seminars in Oncology Nursing, 23, 99–105. 10.1016/j.soncn.2007.01.008
Mosarla R. C., Wood M. (2021). The impact of depression and anxiety on cardiovascular disease risk and outcomes in women. Current Treatment Options in Cardiovascular Medicine, 23, 10. 10.1007/s11936-020-00889-9
Park J., Moser D. K., Griffith K., Harring J. R., Johantgen M. (2019). Exploring symptom clusters in people with heart failure. Clinical Nursing Research, 28, 165–181. 10.1177/1054773817729606
Pituch K. A., Stevens J. P. (2016). Applied multivariate statistics for the social sciences: Analyses with SAS and IBM’s SPSS (6th ed.). Routledge.
Popovic D., King C., Guerrero M., Levendowski D. J., Henninger D., Westbrook P. R. (2009). Validation of forehead venous pressure as a measure of respiratory effort for the diagnosis of sleep apnea. Journal of Clinical Monitoring and Computing, 23, 1–10. 10.1007/s10877-008-9154-8
Rasmussen A. A., Johnsen S. P., Berg S. K., Rasmussen T. B., Borregaard B., Thrysoee L., Thorup C. B., Mols R. E., Wiggers H., Larsen S. H. (2020). Predictors of patient-reported outcomes at discharge in patients with heart failure. European Journal of Cardiovascular Nursing, 19, 748–756. 10.1177/1474515120902390
Redeker N. S. (2006). Somatic symptoms explain differences in psychological distress in heart failure patients vs. a comparison group. Progress in Cardiovascular Nursing, 21, 182–189. 10.1111/j.0889-7204.2006.05643.x
Redeker N. S., Jeon S., Andews L., Cline J., Mohsenin V., Jacoby D. (2019). Effects of cognitive behavioral therapy for insomnia on sleep-related cognitions among patients with stable heart failure. Behavioral Sleep Medicine, 17, 342–354. 10.1080/15402002.2017.1357120
Redeker N. S., Jeon S., Muench U., Campbell D., Walsleben J., Rapoport D. M. (2010). Insomnia symptoms and daytime function in stable heart failure. Sleep, 33, 1210–1216. 10.1093/sleep/33.9.1210
Redeker N. S., Knies A. K., Hollenbeak C., Yaggi H. K., Cline J., Andrews L., Jacoby D., Sullivan A., O'Connell M., Iennaco J., Finoia L., Jeon S. (2017). Cognitive behavioral therapy for insomnia in stable heart failure: Protocol for a randomized controlled trial. Contemporary Clinical Trials, 55, 16–23. 10.1016/j.cct.2017.01.009
Riegel B., Dickson V. V., Faulkner K. M. (2016). The situation-specific theory of heart failure self-care: Revised and updated. Journal of Cardiovascular Nursing, 31, 226–235. 10.1097/jcn.0000000000000244
Riegel B., Moser D. K., Anker S. D., Appel L. J., Dunbar S. B., Grady K. L., Gurvitz M. Z., Havranek E. P., Lee C. S., Lindenfeld J., Peterson P. N., Pressler S. J., Schocken D. D., Whellan D. J. (2009). State of the science: Promoting self-care in persons with heart failure: A scientific statement from the American Heart Association. Circulation, 120, 1141–1163. 10.1161/circulationaha.109.192628
Riegel B., Weaver T. E. (2009). Poor sleep and impaired self-care: Towards a comprehensive model linking sleep, cognition, and heart failure outcomes. European Journal of Cardiovascular Nursing, 8, 337–344. 10.1016/j.ejcnurse.2009.06.003
Ryou C., Kang S.-M., Jang Y. (2021). Factors associated with self-care behaviours among Koreans with heart failure. European Journal of Cardiovascular Nursing, 20, 276–284. 10.1177/1474515120934060
Spedale V., Luciani M., Attanasio A., Di Mauro S., Alvaro R., Vellone E., Ausili D. (2021). Association between sleep quality and self-care in adults with heart failure: A systematic review. European Journal of Cardiovascular Nursing, 20, 192–201. 10.1177/1474515120941368
Steinberg R. S., Nayak A., Morris A. A. (2021). Racial differences in outcomes after receiving advanced heart failure therapies. Current Opinion in Cardiology, 36, 352–359. 10.1097/HCO.0000000000000854
Vellone E., Riegel B., Cocchieri A., Barbaranelli C., D’Agostino F., Antonetti G., Glaser D., Alvaro R. (2013). Psychometric testing of the Self-Care of Heart Failure Index Version 6.2. Research in Nursing & Health, 36, 500–511. 10.1002/nur.21554
Zaharova S., Litwack K., Gopalakrishnan S., Ellis J., Saltzberg M. T. (2021). Self-management in heart failure: The importance of self-regulation but not complexity of condition. Western Journal of Nursing Research, 44, 375–382. 10.1177/0193945921997428
Keywords:

chronic insomnia; heart failure; self-care; self-management; symptoms

Supplemental Digital Content

Copyright © 2022 Wolters Kluwer Health, Inc. All rights reserved.