Insomnia Symptoms and Daytime Function in Stable Heart Failure

Abstract

Objectives:

To evaluate insomnia symptoms and the extent to which they are associated with clinical and demographic patient characteristics, daytime symptoms, and functional performance in patients with stable heart failure (HF).

Design:

Cross-sectional, observational.

Setting:

Five structured HF disease management programs in the Northeastern U.S.

Participants:

173 stable chronic HF patients

Interventions:

N/A

Measurements and Results:

Full polysomnography was obtained for one night in participants' homes. Participants completed the six-minute walk test, Medical Outcomes Study SF-36, Epworth Sleepiness Scale, Pittsburgh Sleep Quality Index, Multi-Dimensional Assessment of Fatigue Scale, Centers for the Epidemiological Studies of Depression Scale, and questionnaire items eliciting insomnia symptoms (self-reported difficulty initiating and maintaining sleep and waking too early in the morning). Over half of HF patients reported insomnia symptoms. These were associated with increased daytime symptoms (depression, fatigue), excessive daytime sleepiness, and functional performance in models that statistically controlled for clinical and demographic covariates. These relationships were not explained by sleep disordered breathing.

Conclusions:

Insomnia symptoms are common in patients with stable heart failure and are associated with daytime symptoms and decrements in functional performance.

Citation:

Redeker NS; Jeon S; Muench U; Campbell D; Walsleben J; Rapoport DM. Insomnia symptoms and daytime function in stable heart failure. SLEEP 2010;33(9):1210-1216.

HEART FAILURE (HF) IS A DISABLING CHRONIC CONDITION THAT AFFLICTS OVER 5 MILLION AMERICANS¹ AND IS ASSOCIATED WITH DECREMENTS IN functional performance, depression, excessive daytime sleepiness, and fatigue. As many as 73% of people with chronic heart failure (HF) report poor sleep quality,^2–9 and many have poor sleep continuity.^10–13 Insomnia symptoms are also common.^7–9 Understanding the nature of insomnia symptoms and their associations with daytime symptoms and functional performance is necessary to guide sleep disorders treatment for HF patients.

People with HF report insomnia symptoms, including difficulty maintaining sleep (34% to 43% of sampled patients),^7–9 falling asleep (23% to 47%),^7–9 and waking too early in the morning (35% to 39%)^7–9 in studies of HF patients awaiting transplant HF,⁸ hospitalized HF patients,⁹ and a small group of patients in a structured HF management program (n = 59).⁷ HF patients were consistently more likely to report difficulty initiating and maintaining sleep and to have objective evidence of prolonged sleep latency and poor sleep continuity than a comparison group recruited from the same community,⁷ and rates^8,9 are higher than those reported by adults responding to the Sleep in America Poll.¹⁴ HF was associated with more than a two-fold increased risk of self-reported insomnia in a large community study.¹⁵ Although sleep disordered breathing (SDB) occurs in approximately 50% of HF patients^{10,11,16–26} and is often associated with difficulty maintaining sleep, few, if any studies have addressed the potential associations of SDB or polysomnographically recorded sleep characteristics with insomnia symptoms among HF patients.

Insomnia symptoms are associated with impaired functional performance, including decrements in exercise capacity,⁵ 6-minute walk test distance (6MWT),²⁷ and self-reported physical function,^9,27,28 but others have not found relationships in people with HF.^8,29 To our knowledge, the extent to which clinical and demographic variables contribute to the relationships between insomnia symptoms, daytime symptoms, and functional performance has not been addressed.

The objectives of this study were to examine the extent to which: (1) patients with stable HF report insomnia symptoms (Difficulty Initiating and Maintaining Sleep - DIMS); (2) demographic and clinical patient characteristics and SDB explain DIMS; and (3) self-reported DIMS is associated with daytime symptoms (depression, fatigue, excessive daytime sleepiness) and functional performance (6MWT and self-reported physical function), while accounting for SDB and potentially relevant clinical and demographic characteristics.

METHODS

The study was cross-sectional. The investigators obtained institutional review board approval from the university in which the first author was employed and all sites in which participants were recruited. Participants provided written informed consent. Details of the overall study design and methods were previously reported,³⁰ but are summarized here for clarity.

Sample

We recruited patients with stable chronic HF from 5 HF disease management programs in the Northeastern United States. Cardiologists and nurse practitioners working in these programs referred potential participants who met the study criteria for participation in the study. The inclusion criteria included New York Heart Association Functional Classification (NYHA) II-IV HF and age ≥ 18 years of age. Exclusion criteria included pregnancy; alcohol and drug abuse; Parkinson's disease; obstructive valvular, hypertrophic, or surgically correctable heart disease; previously diagnosed sleep disorders; and renal failure requiring dialysis. Temporary exclusion criteria included unstable medical or psychiatric conditions, active titration of vasoactive medications, and hospitalization or emergency department visits within the past month.

Variables and Instruments

Functional performance

Functional performance is the “day to day corporeal activities people do in the normal course of their lives to meet basic needs, fulfill usual roles, and maintain health and wellbeing,”³¹ We evaluated its objective (six-minute walk test [6MWT]) and subjective (physical function scale of the Medical Outcomes Study SF-36) characteristics.

The 6MWT,³² a measure of distance walked in 6 minutes in the clinic setting, is correlated with oxygen consumption during treadmill testing,³³ cycle ergometry, and self-reported functional status.³² It was conducted using standard methods.³⁴

The Medical Outcomes Study SF36 V2^35,36 physical function sub-scale (PF )was used to elicit self-reported physical function. Cronbach's α of the PF subscale was 0.90 in the current study.

Daytime symptoms

The Multi-Dimensional Assessment of Fatigue Scale (MAF), developed from the Piper Fatigue Scale,³⁷ was used to measure fatigue. For the present analysis, we computed a score for “sensory fatigue:” the average of the numeric rating scales (0-10) for degree, severity, and distress associated with fatigue. Cronbach's α was 0.91 for the sensory fatigue scale in the current study.

The Center for Epidemiological Studies Depression Scale (CESD)^38,39 was used to measure depressive symptoms. The CESD has good reliability⁴⁰ and adequate sensitivity and specificity.⁴¹ It had a Cronbach's α of 0.89 in the current study.⁴² A cut-off (CESD ≥ 16), suggestive of clinical depression, was used in the logistic analyses.

The Epworth Sleepiness Scale (ESS)⁴³ was used to measure self-reported excessive daytime sleepiness.⁴⁴ The ESS has well-documented reliability and validity in a variety of populations. Cronbach's α was 0.78⁴² in the data obtained in the current study. For the logistic analyses, we dichotomized the score (ESS ≥ 11), consistent with the literature.³

Insomnia symptoms

Insomnia symptoms (DIMS) were evaluated with questions from the Sleep Habits Questionnaire (SHQ) used in the Sleep Heart Health Study.⁴⁵ The 3 items were rated on a 5-item scale to indicate the frequency with which respondents experienced difficulty initiating and maintaining sleep and awakening too early in the morning (DIMS). These symptoms were rated as “never” –“almost always.” Using the methods from the SHHS study,⁴⁵ the presence of insomnia symptoms was defined as a response of often or almost always on one or more of these items. Cronbach's α was 0.83 for this scale.

Self-reported quantitative sleep characteristics

We evaluated respondents' perceptions of quantitative sleep characteristics (sleep duration, latency, time in bed) with items from the Pittsburgh Sleep Quality Index ⁴⁶ and calculated sleep efficiency as follows: [sleep duration/time in bed] × 100.

Polysomnographic characteristics of sleep

Full unattended nocturnal polysomnography was conducted in participants' homes, with the Safiro (Compumedics, Inc.) sleep recorder for one night. Full details of this method were previously reported.³⁰ Sleep data were scored with Rechtschaffen and Kales criteria.⁴⁷ Respiratory events were defined according to standard criteria.^48,42 In the current analysis, we utilized the respiratory disturbance index ([RDI] sum of apneas, hypopneas, and respiratory event related arousals)/hour of sleep as the primary measure of SDB. We also evaluated the percentage of sleep time during which participants had oxygen saturation ≤ 90% and calculated the arousal^49,48 and periodic limb movement indices. The proportion of participants who had central vs. obstructive apnea were also calculated, using methods reported in our earlier paper.³⁰

Participants were recruited at the time of a routine visit to the HF program. A research assistant explained the study, obtained informed consent, reviewed medical records, and performed the 6MWT. The study coordinator scheduled the polysomnography. Certified polysomnographic technicians made home visits to set up the PSG equipment and sensors. Unattended studies were recorded throughout the night, and the equipment was retrieved the following morning. Participants completed the study questionnaires while at home.

Data Analysis

Data were double-entered into an SPSS database, corrected for errors, and examined for the extent to which they met assumptions for parametric analysis. Descriptive and bivariate analyses were conducted to evaluate the associations between demographic and clinical patient characteristics, sleep disordered breathing, sleep characteristics, and DIMS.

Confirmatory factor analysis (CFA) was performed using M-plus version 5.2. (www.statmodel.com) to evaluate the effects of insomnia symptoms on the intercorrelated dependent variables (physical function, 6MWT, sensory fatigue, depressive symptoms, and excessive daytime sleepiness) using the latent variable model. We used CFA to test for latent variables including daytime symptoms (sensory fatigue, depression, excessive daytime sleepiness) and functional performance (6MWT and physical function).

We tested the effects of the insomnia symptoms and the demographic and clinical variables on the dependent variables and the interaction effects between insomnia symptoms (DIMS) and each of the clinical (NYHA classification, left ventricular ejection fraction, comorbidity, RDI) and demographic (age, gender) covariates. Goodness of fit was examined using the comparative fit index (CFI) and the root mean square error of approximation (RMSEA). The values of CFI > 0.9 and RMSEA < 0.08 indicate acceptable model fit.^50,51

To confirm our findings from the CFA, we compared the least square means and the 95% confidence intervals of the original outcome variables (physical function, 6MWT, sleepiness, fatigue, and depression) from the generalized linear regression model after controlling for the same covariates used in the previous model. Additionally, we estimated the odds ratios (OR) for daytime sleepiness and depressive symptom, using clinical thresholds (ESS ≥ 11 and CES-D ≥ 16) from the logistic regression models.

RESULTS

Sample

We approached 324 patients, of whom 41 were ineligible upon further screening, 233 consented, and 35 declined to complete the PSG. Other reasons for incomplete data were death or deteriorating health (n = 4), intolerance of PSG monitoring, technical problems (n = 5), and loss to follow-up (n = 11). Full details have previously been published.³⁰ Demographic and clinical patient characteristics and self-reported and PSG-measured sleep characteristics for the overall sample and for participants with and without insomnia symptoms are in Table 1. The sample consisted of 35% women and 35% (n = 61) minority group members, of whom the majority were black (n = 50). The age range of the sample was 21-28 years.

Table 1.

Comparison of participants with and without insomnia symptoms (DIMS) on demographic, clinical, and sleep variables

Demographic Variables	Overall Sample (n = 173)	No DIMS (n = 84) M (SD)/N (%)	DIMS (n = 89) M (SD)/N (%)
Age (Years)	60.3 (16.1)	61.1 (16.6)	59.6 (15.8)
Gender
Male	113 (65.3%)	60 (53.1%)	53 (46.9%)
Female	60 (34.7%)	24 (40.0%)	36 (60.0%)
Race
White	110 (63.6%)	57 (51.8%)	53 (48.2%)
Minority	61 (46.8%)	26 (42.6%)	35 (57.4%)
Clinical Variables
Left ventricular ejection fraction (%)	32.6 (15.2)	32.9 (15.6)	32.4 (14.9)
NYHA classification	2.5 (0.7)	2.4 (0.6)	2.5 (0.7)
Body mass index	30.7 (8.0)	30.6 (8.7)	30.8 (7.4)
Charlson Comorbidity Index	2.4 (1.5)	2.3 (1.6)	2.6 (1.5)
Clinical depression	30 (17.3)	11 (13.1)	19 (21.3)
Self-Reported Sleep Characteristics
Time in bed (min)	487.2 (99.6)	494.2 (83.6)	480.6 (112.8)
Sleep latency (min)***	29.2 (30.5)	16.1 (12.1)	41.6 (36.8)
Sleep duration (min)***	387.1 (100.5)	428.0 (86.0)	348.5 (98.2)
Sleep efficiency***	80.0 (17.7)	86.6 (14.5)	73.8 (18.3)
Polysomnographic Sleep Characteristics
Sleep duration (min)**	323.7 (96.6)	343.5 (85.2)	304.9 (103.3)
Time in bed (min) **	423.5 (98.8)	444.6 (84.4)	403.7 (107.4)
Sleep efficiency*	71.0 (16.3)	73.7 (13.7)	68.5 (18.1)
Sleep latency (min)	30.2 (35.3)	26.3 (37.2)	33.8 (33.2)
Respiratory disturbance index (RDI)	24.3 (19.0)	26.5 (20.4)	22.2 (17.4)
RDI ≥ 15	103 (60.1%)	53 (63%)	50 (48.1)
Percent time at oxygen saturation < 90%	11.9 (18.8)	13.0 (19.9)	10.8 (17.9)
Predominant obstructive sleep apnea	38 (22.6)	21 (25.0)	17 (20.2)
Predominant central sleep apnea	16 (9.5)	8 (9.5)	8 (9.5)
EEG arousal index	21.36 (10.49)	22.26 (11.21)	20.49 (9.74)
Periodic limb movement index > 5	42 (24.7)	22 (26.2)	20 (23.3)

^***P < 0.001,

^**P < 0.01,

^*P < 0.05

Fifty-one percent of the sample reported difficulty initiating sleep, maintaining sleep, or waking too early in the morning often or almost always. The most common sleep symptom was trouble maintaining sleep (47%), followed by trouble falling asleep (42%), and waking too early in the morning (24%). Twenty-eight percent reported insufficient sleep often or almost always. As seen in Table 1, there were no statistically significant differences in insomnia symptoms based on age, race, body mass index, left ventricular ejection fraction, NYHA classification, comorbidity, history of clinical depression, RDI, predominant central or obstructive apnea, time at oxygen saturation < 90%, arousal index or periodic limb movement index > 5. Sixty percent of women compared with 47% of men reported insomnia symptoms (P = 0.102).

Self-reported, but not PSG-measured sleep latency, was significantly more prolonged in patients with insomnia symptoms, compared with those without. Patients with insomnia had significantly shorter sleep time and lower sleep efficiency measured by self-report and PSG. The insomnia-related differences in the self-reported sleep characteristics were larger than the insomnia-related differences in the PSG-measured sleep characteristics, with the exception of time in bed.

The latent variable “daytime symptoms” had high factor loadings on fatigue (0.724, SE = 0.056) and depression (0.716, SE = 0.057). The “functional performance” latent variable had factor loadings of 0.723 (SE = 0.057) and 0.651 (SE = 0 .060) on physical function and 6MWT. These two latent variables were highly correlated with one another (r = −0.631, SE = 0.198), but had lower correlations with excessive daytime sleepiness. Therefore, in subsequent analyses we separately considered the effects of insomnia symptoms on the latent variables daytime symptoms (sensory fatigue and depression), functional performance (self-reported physical function and 6MWT), and excessive daytime sleepiness.

We examined the effects of insomnia symptoms and the clinical and demographic covariates on the 2 latent variables and excessive daytime sleepiness (Table 2). The model was a good fit with the data as indicated by acceptable values of CFI and RMSEA. Participants with insomnia symptoms had significantly more daytime symptoms, more excessive daytime sleepiness, and lower functional performance.

Table 2.

Standardized coefficients of insomnia symptoms (DIMS), demographic, and clinical variables from the latent variable model

	Daytime Symptoms**			Sleepiness			Functional Performance**
	Coeff.	SE	P	Coeff.	SE	P	Coeff.	SE	P
DIMS (yes vs. no)	0.98	0.29	0.001	0.18	0.07	0.013	−0.45	0.13	0.001
Demographic Variables
Age	−0.25	0.13	0.058	0.07	0.10	0.519	−0.28	0.10	0.006
Gender (male vs. female)	−0.38	0.08	< 0.001	−0.03	0.08	0.669	0.46	0.07	< 0.001
Race (white vs. minority)	0.13	0.09	0.158	−0.10	0.09	0.257	0.20	0.08	0.020
Clinical Variables
Left ventricular ej. fraction %	−0.02	0.08	0.788	0.02	0.08	0.791	−0.04	0.07	0.632
NYHA classification	0.19	0.08	0.022	0.08	0.08	0.323	−0.40	0.08	< 0.001
Body mass index	−0.19	0.09	0.043	0.03	0.09	0.719	< 0.01	0.08	0.994
Charlson Comorbidity Index	0.13	0.09	0.126	0.11	0.08	0.184	−0.46	0.10	< 0.001
Respiratory disturbance index	0.04	0.09	0.602	0.15	0.08	0.063	−0.08	0.08	0.341
%Time at O2 saturation < 90%	−0.06	0.08	0.420	0.06	0.08	0.462	−0.08	0.07	0.266
Interaction Effects
Age × DIMS	−0.60	0.29	0.041	-	-	-	-	-	-
Charlson Comorbidity Index × DIMS	-	-	-	-	-	-	0.33	0.15	0.025

^**indicates effects on the latent variables; Goodness of fit Test: CFI = 0.970/RMSEA = 0.041

There were significant associations between several demographic and clinical variables and the dependent variables (Table 2). Age was associated with lower levels of daytime symptoms, but poorer functional performance. Female gender was significantly associated with higher levels of daytime symptoms, more daytime sleepiness, and poorer functional performance. Poorer NYHA classification was significantly associated with higher daytime symptoms and impairment of functional performance, while lower BMI was associated with lower levels of daytime symptoms, and comorbidity was associated with poorer functional performance. There were no statistically significant associations between the latent variables and RDI, oxygen desaturation, arousal index or periodic limb movements. There was a statistically nonsignificant trend (P = 0.064), suggesting that insomnia symptoms were associated with a lower RDI.

The means of the dependent variables, adjusted for relevant covariates in the regression analyses, are presented in Table 3. On average, patients who reported insomnia symptoms (DIMS) had CESD scores 6 points higher than those who did not have DIMS, and this level was 4 points higher than the level of CESD suggestive of clinical depression (CESD ≥ 16). Participants with DIMS were more likely to be depressed (OR = 5.09; 95% CI = 2.41 − 10.75), after adjusting for clinical and demographic covariates. DIMS was not significantly associated with the dichotomized excessive daytime sleepiness score (OR = 1.47; 95% CI = 0.71 −3.05). However, the least-square mean of the ESS score, adjusted for the covariates, was significantly greater (about 2 points higher) in patients who had DIMS compared to those who did not have DIMS.

Table 3.

Least square means for fatigue, depression, sleepiness, physical function, and six-minute walk after adjusting for covariates in the general linear regression model

Symptoms	Insomnia Symptoms (DIMS)
	Yes	No
	Mean (95% CI)	Mean (95% CI)
Sensory fatigue	6.10 (5.59, 6.61)	4.88 (4.33, 5.43)
Depression	20.66 (18.45, 22.87)	13.84 (11.47, 16.21)
Sleepiness	9.55 (8.60, 10.51)	7.72 (6.68, 8.75)
Functional Performance
Physical function	17.10 (16.16, 18.04)	18.60 (17.60, 19.60)
Six-minute walk (feet)	875.3 (791.0, 959.6)	976.8 (882.9, 1070.7)

There was a statistically significant interaction effect between insomnia symptoms and age on daytime symptoms. Patients without insomnia symptoms reported fairly consistent levels of sensory fatigue across age levels, but age was inversely associated with fatigue in participants who had insomnia (Figure 1A). The change in fatigue for patients with DIMS was −0.07 (SE = 0.02) for each year of age. However, for those without DIMS it was −0.02 (SE = 0.02). There was a similar interaction effect between insomnia symptoms and age on depression (Figure 1B). The change in the CESD was −0.27 (SE = 0.08) for each year of age in patients with insomnia symptoms, while the change was only −0.10 (SE = 0.09) in those without.

Figure 1A.

Relationships between sensory fatigue and age in persons with insomnia symptoms and those without insomnia symptoms. Analysis is controlled for clinical and demographic covariates (gender, BMI, NYHA classification, left ventricular ejection fraction, comorbidity, respiratory disturbance index).

Figure 1B.

Relationships between CESD (depressive symptoms) and age in persons with insomnia symptoms and those without insomnia symptoms. Analysis is controlled for clinical and demographic covariates (gender, BMI, NYHA classification, left ventricular ejection fraction, comorbidity, respiratory disturbance index).

There was also a statistically significant interaction effect of insomnia symptoms and comorbidity on self-reported physical function. Comorbidity was significantly and negatively associated with physical function only in participants who did not report insomnia (Figure 2A). The change in physical function for patients with insomnia symptoms was −0.48 (SE = 0.34) for each one point increase in comorbidity, but the change in physical function for patients without insomnia symptoms was −1.19 (SE = 0.32). Similarly, the change in 6MWT distance for each one point increase in comorbidity was −17.3 feet (SE = 30.9) in patients with DIMS while the change was −74.7 feet in those without DIMS (Figure 2B). There were no other statistically significant interactions between DIMS and the clinical and demographic variables on the symptom or functional performance variables.

Figure 2A.

Relationships between self-reported physical function and comorbidity in persons with insomnia symptoms and those without insomnia symptoms. Analysis is controlled for clinical and demographic covariates (gender, age, BMI, NYHA classification, left ventricular ejection fraction, respiratory disturbance index).

Figure 2B.

Relationships between six-minute walk distance and comorbidity in persons with insomnia symptoms and those without insomnia symptoms. Analysis is controlled for clinical and demographic covariates (gender, age, BMI, NYHA classification, left ventricular ejection fraction, respiratory disturbance index).

DISCUSSION

Over half of patients with stable HF reported insomnia symptoms, with difficulty maintaining sleep the most common. Insomnia symptoms were more prevalent than in the Sleep Heart Health Study, a study of cardiovascular cohorts, in which 42.4% of women and 32.5% of men reported it, using the same items from the Sleep Habits Questionnaire.⁴⁵

Insomnia symptoms were associated with polysomnographic and self-reported sleep characteristics, including sleep efficiency and duration and self-reported sleep latency. However, there were differences in magnitude between the self-report and objective variables. These discrepancies may be due to misperceptions about objective sleep characteristics, as is common in people with insomnia, or differences in the time-frames tapped by these measures (one night for PSGs and one month for the PSQI). The low PSG-recorded sleep efficiency may also reflect the intrusive nature of the sleep recordings.

We found that insomnia symptoms were not explained by SDB, EEG arousals, or periodic limb movements. The absence of a relationship between insomnia symptoms and SDB was consistent with past studies of older community residents.^52,53 Our findings are also consistent with studies in which severity of HF (left ventricular ejection fraction, NYHA classification) did not explain insomnia symptoms.^8,27

The large association between DIMS and depressive symptoms and the 100 foot difference in 6MWT distance, an objective measure of functional performance, between those with and without insomnia symptoms underscore the potential clinical significance of these findings. However, the clinical significance of the small incremental increase in sleepiness associated with DIMS is not clear. Although there has been little previous focus on the association between DIMS and excessive daytime sleepiness in HF, previous studies, including our work using the current sample,³⁰ found no associations between SDB and ESS scores. Whether these findings are due to the lack of sensitivity of the ESS, a true lack of difference/association, or the alerting effects of increased sympathetic arousal associated with HF⁴² is not known. The extent to which insomnia symptoms are associated with objectively measured sleepiness in HF patients should be explored in a future study.

This study extended the findings of previous studies of HF populations^5,8,27,29 by incorporating subjective and objective measures of daytime function and multi-dimensional measures of sleep. Unlike most previous studies, our sample was racially diverse, with 35% minority participants, the majority of whom were black. Recruitment of stable patients from structured HF disease management programs where treatment was somewhat standardized may have reduced the potential impact of acute illness on the sleep, symptom, and functional variables.

Use of confirmatory factor analysis permitted us to address overlaps between depressive symptoms and fatigue—problems that are both highly prevalent in this population and difficult to disentangle, given that fatigue is a somatic component of depression.⁵⁴ Although there was no bivariate relationship between age and DIMS, the interaction effect between DIMS and age on fatigue and depression suggest that the impact of DIMS on these daytime symptoms is greater in young adults than older adults. Older adults had similar levels of fatigue and depression regardless of DIMS. The reasons for this are not clear, but older adults may become accustomed to insomnia symptoms, attribute them to aging, and/or underestimate them. It is also possible that the generally lower levels of fatigue and depression in the older adults are associated with adaptation of one's activities in response to long term experience with insomnia, fatigue, or depression.

The interaction effects between DIMS and comorbidity on self-reported and objective functional performance suggest that among people with DIMS, comorbidity has little effect on self-reported and objective functional performance. People who do not have DIMS have better functional performance at low levels of comorbidity than those with DIMS. However, this advantage does not exist at high levels of comorbidity. People with and without DIMS have similar levels of functional performance at higher levels of comorbidity.

Taken together, the interaction effects suggest that age and comorbidity may be important factors related to the impact of DIMS. These interactions and their implications for the diagnosis and treatment of insomnia should be examined in future studies.

Limitations

We did not utilize the full ICSD or DSM insomnia criteria that also include sleep-related dysfunction. Therefore, our data represent the proportion of the sample who reported selected insomnia symptoms rather than the proportion that have a clinical diagnosis of insomnia, and may overestimate the true prevalence of insomnia. The importance of these distinctions in obtaining accurate prevalence data was emphasized in a recent report.⁵⁵ The rationale for this choice was the need to quantify the unique variance in symptoms and functional performance associated with insomnia symptoms. It would have been tautological to include sleep-related daytime dysfunction as a component of insomnia as an explanatory variable for daytime symptoms and functional performance. In addition, the nearly universal nature of daytime symptoms and functional performance deficits among HF patients (especially those with NYHA class II-IV HF that constituted the sample) may make it difficult for them to accurately attribute these experiences to insomnia.

A second limitation is inability to quantify the extent to which participants who reported insomnia symptoms might be appropriate candidates for insomnia treatment, as indicated by the presence of full insomnia criteria. Given the virtually universal nature of daytime dysfunction in this population and the dearth of studies of insomnia treatment among HF patients, there is a need to further identify HF patients who may benefit from behavioral or pharmacological treatment of insomnia or its symptoms.

Due to the cross-sectional nature of this study, it is possible that insomnia symptoms, daytime function, and functional performance decrements all reflect the underlying pathophysiology of HF. However, the absence of associations between insomnia and measures of severity of illness, such as left ventricular ejection fraction, comorbidity, and NYHA classification, conflict with this explanation. The chronic nature of HF, insomnia symptoms, and related problems, such as clinical depression, suggest the difficulty in disentangling the temporal relationships among these comorbid problems. For example, depression may be a symptom, cause of, or result of insomnia. Although insomnia symptoms were associated with depressive symptoms, there was no association with clinical depression, as recorded in the medical record.

Conclusions and Implications

Insomnia symptoms are common among stable HF patients and are associated with important quality of life concerns, including symptoms of depression, fatigue, excessive daytime sleepiness, and functional performance. Age and comorbidity appear to be important factors associated with the differential impact of DIMS on symptoms and functional performance. The presence of insomnia symptoms, despite the stable condition of the HF patients who received evidence-based HF disease management, suggests that HF disease management alone is not sufficient to ameliorate insomnia symptoms. Future studies are needed to further evaluate the temporal relationships and bio-behavioral mechanisms underlying the associations between insomnia and daytime function in HF patients and to identify HF patients who may benefit from pharmacological and/or behavioral insomnia treatment.

DISCLOSURE STATEMENT

This was not an industry supported study. Dr. Walsleben has participated in speaking engagements for Takeda. The other authors have indicated no financial conflicts of interest.