Sleep disordered breathing, insomnia, and health related quality of life — A comparison between age and gender matched elderly with heart failure or without cardiovascular disease

1. Introduction 

Sleep disordered breathing (i.e., obstructive sleep apnoea and central sleep apnoea with Cheyne–Stokes respiration) is frequent in the elderly [1], [2]. In patients with heart failure (HF) [3], [4] the prevalence varies between 17% and 61%, with higher rates related to milder severity [5], [6], [7], [8], [9], [10], [11], [12]. Insomnia, often secondary to sleep disordered breathing and/or HF, is another prevalent condition that occurs in approximately 20% of patients with HF [13]. Insomnia is defined as difficulty initiating sleep, difficulty maintaining sleep, non restorative sleep, or early morning awakenings [14], [15]. Sleep disordered breathing and insomnia could worsen the life situation for those suffering from HF [16], [17].

Obstructive sleep apnoea is defined as a repetitive reduction of airflow during sleep, designated as apnoeas or hypopnoeas, accompanied by respiratory movements due to partial or complete upper airway obstruction. It has in middle aged populations been associated with hypertension [18], left ventricular hypertrophy [19], as well as a risk factor for the development of cardiovascular morbidity and mortality [20]. Central sleep apnoea and Cheyne–Stokes respiration, in which central apnoeic/hypopnoeic episodes alternate regularly with hyperventilation, is generally absent in middle aged populations. The condition is assumed to occur as a consequence of cardiovascular diseases (CVD), such as HF or stroke [21], [22]. In patients with HF younger than <70years studies have found sleep disordered breathing associated to mortality [23], [24]. In elderly people >70years with or without HF this association is still unclear.

One recent study focusing elderly (mean age 78years) living in the community found mild sleep disordered breathing among more than half of the studied, while almost one fourth had moderate/severe sleep disordered breathing [25]. Since insomnia also is common among the elderly [16], this leads to the question whether these conditions are more common in elderly patients with HF compared to elderly people without HF/CVD. With regard to sleep disordered breathing and insomnia, no study have to the best of our knowledge used a design that allow a comparison between community dwelling elderly, with HF or without HF/CVD of the same age and living in the same geographical area. Furthermore, no studies have focused the associations between sleep disordered breathing and insomnia, or their impact on Hr-QoL in community dwelling elderly patients with as well as without HF/CVD. The effects of sleep disordered breathing symptoms on Hr-QoL have been studied in patients younger than 70years, but are inconclusive [5], [7], [26], [27]. One small study reported that HF patients with sleep disordered breathing rated poorer subjective sleep quality and Hr-QoL compared to those without sleep disordered breathing [26]. Another study comparing different types of sleep disordered breathing reported that those suffering from central sleep apnoea with Cheyne–Stokes respiration scored worse physical Hr-QoL compared to those with obstructive sleep apnoea [5]. Others have not found any associations between sleep disordered breathing and Hr-QoL [7], [27].

The mean age of HF patients living in the community is today ∼75years. Improved or maintained Hr-QoL is an important goal in the management of elderly HF patients [28]. As a help to achieve these goals today, there is clearly a need of more knowledge regarding the importance of sleep disordered breathing and insomnia. The aim of the present study was therefore two folded, (I) to compare the prevalence of sleep disordered breathing and insomnia between elderly people with HF and age and gender matched elderly without CVD, (II) as well as to examine the association between HF, sleep disordered breathing and insomnia, as well as their impact on Hr-QoL.

2. Methods 

2.1. Design and population 

All subjects in this cross sectional cohort study were elderly persons who lived in a rural community with 10,300 inhabitants in the southeast of Sweden. The participants were primarily included in an earlier study that took place in the years 1998–2000. The design of that study has been described earlier [25], [29]. Briefly, in that study all inhabitants between 65 and 82years were invited to meet a cardiologist for a clinical and echocardiographic examination. Individuals suffering from another life threatening disease with short survival, a diagnosis of a serious psychiatric disease, dementia, communication problems or inability to read and speak Swedish were excluded. Of the total 1130 individuals contacted, 876 fulfilled the inclusion criteria and agreed to participate (participation rate 78%). Between the years 2003 to 2005 the cohort examined in 1998–2000 was contacted again and invited to another clinical and echocardiographic examination. Out of 876 that were possible to contact, 675 subjects choose to participate. Reasons for not participating were death (n=102), had moved to nursing homes or left the area (n=29), declined (n=61) or not showing up (n=9). Out of these 675 subjects, 346 (participation rate 51%) also accepted to have their respiration during sleep recorded, and of these 331 had a valid sleep recording. There were no differences regarding age, gender, co morbidities and insomnia, between those who participated in the sleep study and those who did not. To have a fair test of differences of the studied parameters we created one group with “HF” and one without CVD. All subjects who had a least mild impaired systolic function (left ventricular ejection fraction (LVEF)≤49%) with either dyspnoea and/or fatigue, and/or peripheral oedema were considered to have HF (n=41, 8 women and 33 men). Those found with normal systolic function (LVEF≥50) and not suffering from dyspnoea, fatigue, peripheral oedema, IHD, TIA/stroke, arrhythmias/atrial fibrillation or pacemakers were included in the comparison group free of HF or CVD (No CVD) (n=58, 30 women and 28 men). Out of these 99 participants two age and gender matched groups (i.e., HF vs. No CVD) were created, consisting of 36 individuals (8 women and 28 men) each. No differences between the groups were found regarding presence of respiratory- or other serious diseases, as well as prescriptions of hypnotics. The design of the study is described in Fig. 1. All participants provided informed written consent. The study protocol was approved by the ethics committee at the Faculty of Health Sciences, University of Linköping, Sweden, and is in accordance with the provisions of the Helsinki declaration.

View Large Image Download to PowerPoint

Fig. 1 Figure describing the design of the study.

2.5. Sleep-breathing measurement 

Sleep disordered breathing was recorded unattended for one night in the participants' home using the Embletta Portable Diagnostic System (Flaga Medical, Reykjavik, Iceland). The Embletta has earlier been used in HF studies [6], [25] and records airflow, body position, abdominal and thoracic movements, as well as oxygen saturation and pulse. Sleep disordered breathing was scored according to the American Academy of Sleep Medicine Task Force guideline [21]. An apnoea was defined as a cessation of nasal air flow for at least 10s. A more than 50% reduction in airflow and/or respiratory movements for at least 10s accompanied by a ≥4% fall in Sa02 was defined as a hypopnoea. An apnoea with maintained, paradoxical or increased thoracic/abdominal respiratory movements was characterised as obstructive sleep apnoea, while apnoeas with highly reduced or absent respiratory effort were defined as central sleep apnoea [21]. The total number of apnoeas and hypopnoeas were summed and divided by the hours of total sleep time to the apnoea hypopnoea index. Obstructive sleep apnoea, central sleep apnoea and oxygen desaturations were also calculated as indexes (i.e., obstructive sleep apnoea index, central sleep apnoea index, and oxygen desaturation index). The sleep recording was performed over a median of 12days (IQR=12) after the clinical and echocardiographic examination. Out of 346 sleep recordings, 15 were lost due to technical failure. The characteristics of 331 persons who had a valid sleep recording are described in Table 1. As the Embletta does not record EEG, sleep onset and end of the study period was estimated through the combination of breathing pattern, and movements and also from subject's self-estimated sleep onset and morning awakening. All recordings were scored by the main researcher (PJ).

Table 1. Background characteristics of the SDB screened population and individuals with heart failure (HF) and without cardiovascular diseases (No CVD).

	Variables	SDB screened population n=331	HF n=41	No CVD n=58
	Male/Female, n (%)	164 (49)/167 (51)	33 (80)/8 (20)	30 (52)/28 (48)*
	Age, mean SD	78±3	79±3.4	78±3*
	Living alone, n (%)	117 (35)	10 (24)	21 (36)
	NYHA:
	I, n (%)	180 (54)	12 (29)	55 (95)*
	II, n (%)	102 (31)	15 (37)	1 (2)
	III, n (%)	49 (15)	14 (34)	2 (3)
	Body mass index, mean (SD)	28 (4)	27.4±4.2	26.8±3.9
	Systolic blood pressure, mean (SD)	148±19	146±22	151±18
	Diastolic blood pressure, mean (SD)	74±9	74±10	77±10
	Left venticular ejection fraction:
	≥50%, n (%)	276 (84)		58 (100)
	>49%, n (%)	33 (10)	33 (80)
	<40%, n (%)	22 (6)	22 (20)
	Arrhythmia/atrial fibrillation, n (%)	101 (30)	18 (44)
	Ischemic heart disease, n (%)	85 (27)	23 (56)
	Pacemaker, n (%)	7(2)	2 (5)
	TIA/stroke, n (%)	30 (9)	5 (12)
	Dyspnoea, n (%)	186 (56)	36 (88)
	Fatigue, n (%)	102 (31)	16 (39)
	Leg oedema, n (%)	27 (8)	5 (12)
	NT-proBNP, mean (SD)	532 (840)	1426 (1764)	269 (223)*

*p<0.05 compared to the HF group. Note: CVD — Cardiovascular disease; NYHA — New York Heart Association Functional Classification; NT-proBNP — N-terminal fragment brain natriuretic peptide; SDB — Sleep disordered breathing; SBP — Systolic blood pressure; TIA/stroke — Transient ischemic attack/stroke.

3. Results 

3.1. Comparison of sleep disordered breathing and insomnia 

Characteristics for the entire population (n=331) and the two created groups consisting of individuals with HF (n=41) and those without CVD (n=58) are shown in Table 1. Differences in relation to sleep disordered breathing, insomnia and Hr-QoL between the two age and gender matched groups with HF and without CVD are shown in Table 2, Table 3. Sleep disordered breathing were more common in those with HF. In the unmatched analysis those with HF had almost twice as high mean apnoea–hypopnoea index compared to people with no CVD (p<0.001). Those with HF had higher mean obstructive sleep apnoea index (p<0.02) as well as mean central sleep apnoea index (p<0.001). However, there was no difference between the groups in relation to the number of patients with obstructive sleep apnoea index≥5. In people with no CVD almost one third had at least mild sleep disordered breathing (apnoea–hypopnoea index≥5), significantly lesser (p=0.001) compared to the 80% in those with HF. Moderate/severe sleep disordered breathing (apnoea–hypopnoea index≥15) was found in 39% of those with HF, this was significantly more (p=0.008) than the 16% found in those with no CVD. Matching the groups according to age and gender did not change the results (Table 2).

Table 2. Comparison of sleep disorder breathing variables between elderly community-living people with heart failure (HF) or without cardiovascular diseases (No CVD).

	Sleep disordered breathing variables	Unmatched analysis		p	Age and gender matched analysis		p
		HF (n=41, 8f/33m)	No CVD (n=58, 30f/28m)		HF (n=36, 8f/28m)	No CVD (n=36, 8f/28m)
	Apnoea–hypopnoea index, mean (SD)	16.8 (13.6)	8.5 (10)	<0.001	17.6 (14)	6.3 (6.5)	<0.001
	Oxygen desaturation index, mean (SD)	14.9 (12.9)	7 (7.4)	<0.001	15.5 (13.4)	5.3 (5.3)	<0.001
	Obstructive apnoea index, mean (SD)	3.4 (4.7)	1.9 (3.1)	0.02	3.8 (5)	1.2 (2.2)	0.004
	Central apnoea index, mean (SD)	4.8 (1.9)	1.9 (6.1)	<0.001	5.2 (8.9)	1.5 (3.7)	0.01
	Mean Sa02, (SD)	94	95	0.45	94	95	0.43
	Nadir Sa02, (SD)	83	85	0.11	83	86	0.01
	Apnoea–hypopnoea index≥5, % (n)	80 (33)	28 (48)	0.001	81 (26)	47 (23)	0.003
	Apnoea–hypopnoea index ≥10, % (n)	56 (23)	33 (19)	0.02	56 (20)	25 (9)	0.008
	Apnoea–hypopnoea index ≥15, % (n)	39 (16)	16 (9)	0.008	42 (15)	8 (3)	0.001
	Obstructive apnoea index >5	39 (16)	34% (20)	0.64	42 (15)	29 (10)	0.25
	Central apnoea index >5	39 (16)	12 (7)	0.002	36 (13)	14 (5)	0.04

Table 3. Comparison of subjective sleep complaints and Hr-QoL between elderly community-living people with heart failure (HF) or without cardiovascular diseases (No CVD).

	Variables	Unmatched analysis		p	Age and gender matched analysis		p
		HF (n=41, 8f/33m)	No CVD (n=58, 30f/28m)		HF (n=36, 8f/28m)	No CVD (n=36, 8f/28m)
	Insomnia:
	Difficulties maintaining sleep % (n)	73 (30)	52 (30)	0.03	72 (26)	50 (18)	0.05
	Difficulties initiating sleep, % (n)	42 (17)	47 (27)	0.61	42 (15)	42 (15)	1.0
	Non restorative sleep, % (n)	44 (18)	31 (18)	0.19	47 (17)	31 (11)	0.15
	Early morning awakenings, % (n)	39 (16)	31 (18)	0.41	36 (13)	25 (9)	0.31
	Excessive daytime sleepiness:
	ESS, m (SD)	7 (3.9)	5.6 (3)	0.09	7.2 (4)	6.1 (2.9)	0.41
	EDS, % (n) (ESS≥10)	24 (10)	9 (5)	0.03	25 (9)	8 (3)	0.05
	Hr-QoL (SF-36):
	PF, mean (SD)	59.6 (22.8)	78.5 (19.4)	<0.001	60.1 (23.2)	78.6 (20.3)	<0.001
	RP, mean (SD)	44 (41.4)	77.2 (34.1)	<0.001	43 (40.8)	76.4 (37.8)	<0.001
	BP, mean (SD)	61.1 (22)	74.1 (22)	0.004	61.7 (22)	75.2 (20.3)	0.07
	GH, mean (SD)	56.6 (16.2)	73.4 (15.6)	<0.001	57.6 (15.7)	73.5 (15.4)	<0.001
	VT, mean (SD)	56.1(22.4)	76 (17.3)	<0.001	56.1 (22)	74.7 (17.2)	<0.001
	SF, mean (SD)	87.2(19)	91 (18.1)	0.24	88.2 (17.4)	92 (18.9)	0.23
	RE, mean (SD)	66.7 (38)	85.6 (28)	0.006	66.7 (39)	88.9 (23.9)	0.007
	MH, mean (SD)	80.8 (15)	85.6 (13.9)	0.10	79.8 (14.9)	86 (14)	0.05
	PCS, mean (SD)	36.3 (9.5)	46.6 (9)	<0.001	36.3 (9.4)	46 (10)	<0.001
	MCS, mean (SD)	51.2 (7.5)	54.7 (7.6)	0.11	51.2 (9)	54.4 (7)	0.08

Note: BP — Bodily pain; CVD — Cardiovascular disease; ESS — Epworth sleepiness scale; GH — General health; MH — Mental health; MCS — Mental component score; PCS — Physical component score; PF — Physical functioning; RE — Role limitations due to emotional health problems; RP — Role limitations due to physical health problems; VT — Vitality; SF — Social functioning.

Insomnia with regard to difficulty initiating sleep, non restorative sleep, or early morning awakenings did not differ between the groups, irrespective of matching to age and gender (Table 3). However, those with HF reported more problems with difficulty maintaining sleep (p=0.03), as well as more excessive daytime sleepiness (p=0.03). These differences remained unchanged after the matching procedure (Table 3).

Those with HF rated poorer Hr-QoL. Both the unmatched and matched analysis showed that people with HF scored significantly worse in six of the eight domains of the SF-36 (Table 3).

3.2. Sleep disordered breathing and insomnias association to health related quality of life 

To examine sleep disordered breathing and insomnia associations to Hr-QoL correlational, as well as multiple regression analysis were performed, using the two higher order components of the SF-36, PCS and MCS as the dependent variables. Correlation analysis, when using demographic, clinical, as well as sleep variables are shown in Table 4. Correlates to PCS were age, NT-proBNP, arrhythmias/atrial fibrillation, diastolic blood pressure, NYHA, difficulty initiating sleep, difficulty maintaining sleep, non restorative sleep, and early morning awakenings and ESS. Correlates to MCS were NT-proBNP, non restorative sleep and ESS. Variables that had a p<0.10 were chosen to be included as independent variables in the multiple linear regression analysis. The bivariate analysis showed that variables measuring sleep disordered breathing had no association to either the physical or the mental aspect of Hr-QoL. Independent associates in the multiple linear regression analysis to PCS were age, NYHA II, III and difficulty maintaining sleep, explaining 36% of the variance. Only non restorative sleep was independently associated to the MCS and explained 6% of the variance (Table 5). Sleep disordered breathing may have a possible association to Hr-QoL, by an association with insomnia or excessive daytime sleepiness. However, our bivariate analysis did not support this assumption since sleep disordered breathing had no significant correlations to difficulty initiating sleep, difficulty maintaining sleep, non restorative sleep, early morning awakenings or ESS (data not shown).

Table 4. Correlations between demographic variables, clinical variables, sleep variables, and health related quality of life (n=99).

	Variables	PCS	MCS
	Demographic variables:
	Age	−0.44*	−0.14
	Gender	−0.20	−0.11
	Living alone	−0.03	−0.02
	Clinical variables:
	Body mass index	−0.15	−0.15
	NT-proBNP	−0.38*	−0.20**
	Arrhythmia/atrial fibrillation	−0.31*	−0.15
	Systolic blood pressure	0.11	0.14
	Diastolic blood pressure	0.24*	0.05
	NYHA	−0.48*	−0.10
	Sleep variables:
	Apnoea–hypopnoea index	−0.12	−0.23
	Oxygen desaturation index	−0.17**	−0.06
	Obstructive apnoea index	−0.14	−0.08
	Central apnoea index	−0.08	−0.08
	Mean Sa02	−0.14	0.04
	Difficulties maintaining sleep	−031*	−0.12
	Difficulties initiating sleep	−0.28*	−0.16
	Early morning awakening	−0.27*	−0.06
	Non restorative sleep	−0.20*	−0.24*
	Epworth sleepiness scale	−0.19**	−0.18**

*p<0.05. **p<0.10, variable forced into the regression analysis. Note: NYHA — New York Heart Association Functional Classification; MCS — Mental component score; NT-proBNP — N-terminal fragment brain natriuretic peptide ;PCS — Physical component score.

Table 5. Final multiple linear regression analyses, controlled for age and gender, of variables associated to health related quality of life (n=99).

	Variables	PCS
		B	p	Model statistics
	Age	−0.35	<0.001	F=12.2, p<0.001, r2=0.36
	Male	0.04	0.603
	NYHA IIa	−0.26	0.002
	NYHA IIIa	−0.32	0.001
	Difficulties maintaining sleep	−0.21	0.012
		MCS
		B	p	Model statistics
	Age	−0.12	0.250	F=3.0, p=0.037, r2=0.06
	Male	0.08	0.447
	Non restorative sleep	−0.23	0.022

Note: MCS — Mental component score; NYHA — New York Heart Association Functional Classification; PCS — Physical component score.

a Dummy variables with NYHA I as reference.

4. Discussion 

This is to our knowledge the first study that has examined a cohort of community-living elderly persons and compared the prevalence of objectively recorded sleep disordered breathing and subjectively reported insomnia in two age and gender matched groups with HF (established with Doppler echocardiography) or without CVD. Moreover, no present studies have examined the effect of sleep disordered breathing and insomnia on Hr-QoL in such a population.

Moderate/severe sleep disordered breathing was found in approximately 40% of those with HF. This was significantly more than the 8% in age and gender matched individuals without CVD. Sleep disordered breathing has long been recognized as a frequent problem among patients with HF. Two previous studies have examined the prevalence of sleep disordered breathing in outpatients with HF younger than 70years, and found rates of 17%–25% [7], [11]. Direct comparisons with these studies are, however, problematic. First, Rao et al. [7] discussed that they, to some extent, may have underestimated the prevalence of SDB since their recording method did not allow them to discriminate apnoeas from hypopnoeas, as we did. Secondly, the mean age of the population in the study of Quintana-Gallego was 56years [11]. These studies did not compare data to a group without CVD of the same age and located in the same geographical area, as we did. Our design with an age and gender matched control group from the same community shows that sleep disordered breathing might be more prevalent in elderly people with HF. Another interesting aspect is that we found moderate to severe sleep disordered breathing in 16% of those considered to have no CVD. This is lower compared to the 23%–25% of moderate to severe sleep disordered breathing found in other population based studies including people older than 70years [2], [38]. This may be explained by the fact that these studies have not excluded those with HF, or other cardiovascular diseases. Although the cause and pathophysiology of Central sleep apnoea with Cheyne–Stokes respiration still is obscure, associations to decreased cardiac function [39], as well as higher age [40] have been made. Our study found higher central sleep apnoea index in those with HF, even after matching for age. However, it is difficult to say if this difference is of clinical significance. Christ et al. reported that high plasma values of BNP in a sample of patients with HF predicted the occurrence of central sleep apnoea with Cheyne–Stokes respiration [41]. Thus, in the elderly occurrence of sleep disordered breathing and especially central sleep apnoea with Cheyne–Stokes respiration could be of clinical interest since it may be seen as a marker of impairment of cardiac function or HF.

Insomnia is known to be a common problem in the elderly [1]. In the present study difficulty maintaining sleep, as well as excessive daytime sleepiness, were more common in people with HF compared those without CVD. Redeker and Stein [42] also found more of difficulty maintaining sleep, early morning awakenings and excessive daytime sleepiness in HF patients compared to a sample from the community without HF (n=63 in both groups). However, compared to the mean age of approximately 78years in our study their participants had a much lower mean age (56years). The high proportion of difficulties maintaining sleep in our group with HF may be reflected by the high prevalence of sleep disordered breathing. Strangely no correlations were found between sleep disordered breathing variables (i.e., apnoea–hypopnoea index, obstructive sleep apnoea index, central sleep apnoea index, oxygen desaturation index, difficulty maintaining sleep or excessive daytime sleepiness. Moreover, sleep disordered breathing did not correlate to the physical or mental aspect of Hr-QoL, and thus may not be a variable that can explain poorer Hr-QoL scorings in the HF group. In our study insomnia had a relationship to Hr-QoL. Difficulty maintaining sleep was associated to the physical whereas non restorative sleep to the mental aspects of Hr-QoL. The association of difficulty maintaining sleep to physical Hr-QoL may be a reflection of the severity of HF. Dyspnoea, coughing, dysrhythmia and nocturia are factors described by HF patients causing a disturbed sleep [43]. On the other hand we found that difficulty maintaining sleep was associated to Hr-QoL also after control for cardiac function (NT-proBNP) and NYHA class. Fatigue is a core symptom of depression and non restorative sleep was associated to mental Hr-QoL. It is plausible that non restorative sleep may be perceived as fatigue, and one interpretation could be that non restorative sleep can be seen as a sign of depression. Others have found that elderly HF patients who complained of a poor sleep were at a higher risk to suffer from depression [44]. Furthermore, depression has been shown to exert a poorer prognosis in elderly people with HF [45].

Our results shows that sleep disordered breathing not obviously is associated to subjective health complaints neither in elderly with HF or without CVD. In a study including patients with obstructive sleep apnoea above and below 65years the same experiences were found. This study reported that younger patients rated poorer Hr-QoL in all scales of the SF-36 except for the GH scale [46]. In contrast those above 65years only scored worse in the physical scales compared to a normal group of their age. Gooneratne et al. [47] found more insomnia in elderly (mean aged 72years) without sleep disordered breathing compared to those with sleep disordered breathing. In patients with HF younger than 70years, a number of studies report that sleep disordered breathing is not associated to Hr-QoL [5], [7], [27] or excessive daytime sleepiness [7]. It is possible that subjective reports of health, such as sleep, excessive daytime sleepiness or Hr-QoL, not are good tools to assess the impact of sleep disordered breathing in elderly. In a HF sample Hasting et al. [48] found no relations between sleep disordered breathing and subjectively reported excessive daytime sleepiness. However, they found sleep disordered breathing to be associated with reduced daytime activity and poorer sleep quality assessed with objective methods. Sleep disordered breathing has, however, known negative objective health effects. Community based studies including middle aged community-living people report that sleep disordered breathing is an independent predictor of mortality [49], [50]. This has not been shown in studies focusing on sleep disordered breathing and mortality in the elderly [51], [52]. Some studies have found sleep disordered breathing to be a predictor of mortality in patients with HF [23], [53], [54], however these results were based on younger patients (<70years) with predominately central sleep apnoea with Cheyne–Stokes respiration recruited at hospitals.

5. Limitations 

One limitation with this study is the small number of people considered to have HF (n=41) in the examined cohort, which limits the interpretations and the generalizability of the results. Performing studies in the community and focusing HF, as well as sleep disordered breathing is difficult and expensive .We initially used a cohort consisting of 675 elderly community dwelling people, among these 346 also accepted an examination of their breathing pattern during sleep. Out of these 346 sleep recordings, 331 participants had a valid recording and thus possible to be used when creating age and gender matched groups with HF and without CVD. In our study, sleep disordered breathing and echocardiographic examinations of approximately 1000 persons had been required to reach a sample of 100 individuals with HF. This was not possible with our design. However, this study is one of the largest studies that have investigated the prevalence of sleep disordered breathing in community-living elderly. The Sleep Heart Health Study is somewhat larger [2], but compared to this study, none of their participants were clinically examined by an experienced cardiologist or underwent Doppler echochardiography. The majority of the studies that have investigated the impact of sleep disordered breathing on HR-QoL on HF patients have included between 50 and 60 HF patients [5], [7], [26], [27]. Slightly higher compared to our study, however, none of these studies included an age and gender matched comparison group without HF and CVD who lived in the same geographical area. We consider our findings comparable to elderly populations in other countries similar to Sweden. Impaired systolic function was in our study established with echocardiography, which is a recommended method to assess systolic function [28]. The use of LVEF≤49% as a criterion to define impaired systolic function can be seen as a limitation. A number of studies examining sleep disordered breathing in patients with HF have, however, used the criterion LVEF<45% [5], [11], [23], [27], [55]. We could not use this criterion because systolic function was determined semiquantitatively and classified into the groups: LVEF≥50%, LVEF 40–49% and LVEF<40%. Since our criterion is not far from the criterion LVEF<45%, we believe that our results had remained more or less the same if the latter criterion had been used.

6. Conclusion 

In an elderly community population, 40% of those with HF had moderate/severe sleep disordered breathing, significantly more than the matched group without CVD. Those with HF also scored more difficulty maintaining sleep, excessive daytime sleepiness, as well as worse Hr-QoL. However, sleep disordered breathing had no association to any of the insomnia symptoms, excessive daytime sleepiness, or Hr-QoL. Difficulty maintaining sleep was independently associated to the physical domain, whereas non restorative sleep independently was associated to the mental domain of Hr-QoL. In recent years maintained or improved Hr-QoL has achieved great attention as an important outcome in health care. Elderly people who complain of having a sleeping problem must therefore be taken seriously and be carefully evaluated, but the findings in this study do not suggest that sleep disordered breathing is a major problem in such cases. The impact of sleep disordered breathing, in elderly community-living people with HF or without CVD, on objective parameters of health, such as mortality, however, still remains unclear and should therefore be further studied.