Sleep Breath DOI 10.1007/s11325-014-1006-9
ORIGINAL ARTICLE
Sleep disordered breathing, hypoxia and inflammation: associations with sickness behaviour in community dwelling elderly with and without cardiovascular disease Peter Johansson & Erland Svensson & Urban Alehagen & Ulf Dahlström & Tiny Jaarsma & Anders Broström
Received: 21 February 2014 / Revised: 5 May 2014 / Accepted: 9 May 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Background Inflammation can induce a cluster of symptoms, referred to as sickness behaviour (e.g., depressive symptoms, sleep disturbances, pain and fatigue). Cardiovascular disease (CVD) and sleep disordered breathing (SDB) are common in older adults. CVD is associated with an increased inflammatory activity and in SDB, hypoxia can also increase inflammation. The purpose of this study is to explore if SDB-related hypoxia is associated differently with inflammation and the presence of sickness behaviour in older adults with and without CVD. Methods Three hundred and thirty-one older adults, whose mean age is 78 years, underwent one-night polygraphic P. Johansson : U. Alehagen : U. Dahlström Department of Cardiology, Linköping University, Linköping, Sweden P. Johansson : U. Alehagen : U. Dahlström Department of Medical and Health Sciences, Linköping University, Linköping, Sweden E. Svensson Swedish Defence Research Agency, 58111 Linköping, Sweden T. Jaarsma Department of Social and Welfare Studies, Linköping University, 60174 Norrköping, Sweden A. Broström Department of Nursing Science, School of Health Sciences, Jönköping University, 55185 Jönköping, Sweden A. Broström Department of Clinical Neurophysiology, County Council of Östergötland, 58185 Linköping, Sweden P. Johansson (*) Department of Cardiology, County Council of Östergötland, 58185 Linköping, Sweden e-mail:
[email protected]
recording to measure SDB and hypoxia. CVD was established by a clinical investigation. Questionnaires were used to measure sickness behaviour and depressive symptoms. High sensitivity C-reactive protein was used as a marker of inflammation. Results Structural Equation Modelling showed that SDBrelated hypoxia was associated with inflammation (β>0.40) which mediated indirect associations with sickness behaviour (β=0.19) and depressive symptoms (β=0.11), but only in those with CVD (n=119). In this model, inflammation had a direct effect on sickness behaviour (β=0.43) and an indirect effect on depressive symptoms (β=0.24). Hypoxia had the strongest effect (i.e., β=0.41; significant) on inflammation, whereas the AHI or ODI had weak and non-significant effects (β=0.03 and β=0.15). Conclusions Older adults with CVD and SDB are at a particular risk of developing sickness behaviour and depressive symptoms. The effect of SDB was mainly caused by hypoxia, suggesting that hypoxia is an important marker of SDB severity in older adults with CVD. Keywords Sleep disordered breathing . Hypoxia . Inflammation . Cardiovascular disease . Sickness behaviour
Introduction Sleep disordered breathing (SDB), including obstructive sleep apnoea and central sleep apnoea, is a sleep disturbance characterized by repeated cessations in breathing, decreases in oxygen saturation and increased sympathetic activity during sleep, which might contribute to cardiovascular stress [1, 2]. At least one fifth of older adults have been found to have moderate to severe SDB [3], and the occurrence is more frequent among those with cardiovascular disease (CVD). Being older and having CVD almost doubles the odds of
Sleep Breath
having a moderate to severe SDB after adjustment for body mass index and male sex [4]. Reports also suggest that symptoms of insomnia (i.e., difficulties initiating sleep [DIS], difficulties maintaining sleep [DMS] or non-restorative sleep [NRS]) are more frequent among those with CVD [5–7]. Pain and fatigue are other symptoms found to be common among older adults [8, 9] and are often seen in association with chronic diseases, such as CVD [10]. Pain and fatigue often coexist with insomnia symptoms and depression [10–12]. In middle-aged populations SDB has been associated with disturbed sleep, depression and poor physical and mental quality of life (QoL) [13–15]. In older adults, the impact of SDB on these complaints remains unclear [16–19]. However, the number of breathing events (i.e., the apnoea–hypopnea index (AHI)) may not to be the best quantification of SDB in the elderly. Recent studies performed on older adults have shown the frequency of hypoxia to be more important than the AHI. Yaffe et al. [20] reported that measures of hypoxia, having oxygen desaturation index >15 (ODI; the number of desaturation per hour of sleep) or having >7 % of sleep time in apnoea or hypopnea with drops in oxygen saturation >3 %, in older women after 3 years of follow-up increased the risk of cognitive dysfunction or dementia [20]. In another small study on geriatric patients, hypoxia, indicated by ODI>5, was found to be associated with poorer performance in activities of daily living and sleepiness [21]. Hypoxia-induced inflammation may be a possible explanation for these findings. The hypoxia inducible factor, the major cellular regulator of oxygen homeostasis, is up-regulated by chronic intermittent hypoxia as in SDB. This up-regulation induces a complex cascade of physiological processes including increased oxidative stress, inflammation and sympathetic activity [22, 23]. Inflammation has been suggested to be an underlying mechanism behind a cluster of symptoms, including sleep disturbances, pain, fatigue and depressive symptoms [24–28]. This symptom cluster is referred to as sickness behaviour and can arise in conditions associated with a persistent low grade inflammatory state [29]. CVD is associated with an increased inflammatory activity [30] as well as a higher general sympathetic tone [31]. In a study of older adults, hypoxia, as indicated by the ODI and the percentage of sleep time with SaO2 below 90 % (T90), was associated with poorer cognitive performance (i.e., another potential symptom of sickness behaviour), but only in those with CVD [32]. Hence, older adults with CVD seem to be particularly vulnerable to the negative effects of hypoxia, and thus are more likely to develop sickness behaviour. To our knowledge, no prior studies on older adults have explored if the pathways between SDB-related hypoxia, inflammation and the presence of sickness behaviour are unique in CVD. We therefore sought to explore if SDB-related hypoxia is associated differently with inflammation and the presence of sickness behaviour in older adults with and without CVD.
Material and methods Design and population The patient population in this study was derived from the CoroKind study. The study design has been described previously [3, 33]. In brief, the major aim was to evaluate the prevalence of heart failure among community-dwelling elderly. Initially the study took place from 1998 to 2000. In that study, all those aged 65–82 years who lived in a rural community with 10,300 inhabitants in the southeast of Sweden were invited to clinical and echocardiography examinations. Of 1,130 individuals, 876 agreed to participate (participation rate, 78 %). Between the years 2003 and 2005, the cohort examined in 1998–2000 was contacted again and invited to another clinical and echocardiography examination. Out of 876 that were contacted, 675 subjects chose to participate. Reasons for not participating were death (n=102), having moved to nursing homes or left the area (n=29), declined (n=61) or did not show up (n=9). Out of these 675 subjects, 346 (participation rate 51 %) also agreed to have their respiration recorded during sleep, and of these, 331 had a valid sleep recording. There were no differences regarding age, sex, co-morbidities, or declared insomnia, between those who participated in the sleep study and those who did not. All participants provided informed written consent. The study protocol was approved by the ethics committee of the Faculty of Health Sciences, University of Linköping, Sweden and conforms to the principles outlined in the Declaration of Helsinki. Cardiovascular disease In the present study a cardiologist examined all the participants, taking the patient history and giving a clinical examination. All the participants with ischemic heart disease, heart failure, and/or transient ischemic attack/stroke (TIA/stroke) were amalgamated into one group labelled the CVD group (n=119). Ischemic heart disease was defined as a history of angina pectoris, and/or previous myocardial infarction. Heart failure was present when the subject had at least mild impaired systolic function (LVEF, 5, n (%) AHI>10, n (%) AHI>15, n (%) Hypoxia T90, median (25th;75th quart) Inflammation CRP, median (25th;75th quart) Fatigue Vitality scale,a mean (SD)
5.2 (1.8; 11.6) 183 (55 %) 119 (36 %) 76 (23 %)
4.4 (1.4; 9.5) 104 (49 %) 65 (31 %) 38 (18 %)
7.8 (2.2; 15.8) 79 (66 %) 54 (45 %) 38 (32 %)
0.001 0.02 0.007 0.004
0.4 (0.1:2)
0.4 (0.1;1.7)
0.5 (0.1;2.3)
0.16
2.4 (1.1;4.75)
2.2 (1.1;4.8)
2.7 (1.2;4.7)
0.37
Fatigue item,b median (25th; 75th quart) Pain, mean (SD) Bodily pain scale Depressive symptoms, mean (SD) HAD-D scale
15 (15 34 (29 41 (34 84 (71 85 (71 23 (19 29 (24
%) %) %) %) %) %) %)
0.11 0.09 0.001 0.49 0.001 0.27 0.001
41 (34 %) 69 (58 %) 47 (40 %)
0.03 0.001 0.20
61.4 (22.8) 2 (2; 3)
63.2 (22.4) 2 (2; 3)
58 (23) 2 (2; 3)
0.04 0.62
64 (25)
63.9 (25.4)
64.1 (24.4)
0.44
4.1 (2.7)
3.9 (2.6)
4.4 (2.9)
0.17
ACE/ARB angiotensin converting enzyme inhibitor/angiotensin receptor blockade, AHI apnea–hypopnea index, BMI body mass index, CRP C-reactive protein, GFR glomerular filtration rate, HAD-D Hospital Anxiety Depression scale, NT-proBNP N-terminal fragment brain natriuretic peptide, ODI oxygen desaturation index, T90 percentage of sleep time with SaO2