It has been suggested that oxidative stress and inflammation are associated with the pathophysiology of sleep disorders in end-stage renal disease (ESRD) patients. We examined the relationship of the sleep variables reflecting sleep breathing disorder and limb movements during sleep with the clinical variables reflecting the metabolic abnormalities in ESRD patients receiving hemodialysis.
Nocturnal polysomnography was conducted in 30 ESRD patients (21 men, 9 women), who were receiving hemodialysis. Blood was sampled before hemodialysis for each patient in order to measure uric acid (UA), C-reactive protein (CRP), and interleukin-6 (IL-6).
UA was correlated positively with the total sleep time (TST)(r=0.407) and negatively with the apnea-hypopnea index (AHI) and oxygen desaturation index (ODI)(r=-0.377, -405).
CRP was positively correlated with the limb movement index (LMI)(r=0.401). Our study showed that increased UA was related to decreased respiratory disturbance during sleep in ESRD patients receiving hemodialysis.
Sleep problems are very common in patients with end-stage renal disease (ESRD), as more than 80% of them complain of difficulty in maintaining sleep, restless legs, and daytime sleepiness.
Recently, the increased prevalence of oxidative stress and inflammation in patients with moderate to severe chronic kidney disease was reported.
Similarly, SA has been recognized as an oxidative stress disorder. It has also been reported that patients with SA have elevated inflammatory markers such as interleukin-6 (IL-6), tumor necrosis factor (TNF)-α and C-reactive protein (CRP), and have a higher risk of cardiovascular disease than the general population.
We aimed to examine the relationship of the nocturnal sleep characteristics linked to sleep breathing disorder (SBD) or limb movements during sleep with the levels of UA, IL-6 and CRP, which act as metabolic markers reflecting the antioxidant capacity or inflammation in ESRD patients receiving hemodialysis.
We studied ESRD patients who were receiving hemodialysis at Kangwon National University Hospital from October 2004 until December 2005.
According to the guidelines of the National Foundation-Dialysis Outcome Quality Initiative (NKF-DOQI) in United States,
For 30 ESRD patients, the Epworth Sleepiness Scale (ESS)
All subjects were studied by one-night laboratory-based polysomnography (PSG). Considering the effect of hemodialysis on sleep, the PSG (Embla S7000, Medicaresystem, New York, USA) was conducted according to each subject's habitual sleep time one day after hemodialysis. PSG included electroencephalogram (C3-A2, C4-A1, O1-A2, O2-A2 by international 10-20 system), chin electromyography (EMG), electrooculogram, electrocardiogram, snoring, respiratory effort using piezoelectric belts over the chest and abdomen, and airflow through the nose and mouth using a thermistor. We also recorded bilateral surface EMGs on the legs (with electrodes placed over the anterior tibialis muscles). We monitored oxygen saturation by pulse oximetry. A polysomnographic technologist, who had an interrater reliability of more than 90% with a psychiatrist certified by the American Board of Sleep Medicine, manually scored all of the recordings for the sleep stages, limb movements, and respiratory events by standard techniques.
Blood was sampled for each subject before hemodialysis to measure the UA, CRP, and IL-6 levels. UA was measured by the enzymatic method (Denka, Tokyo, Japan) with a TBA-200FR Chemical analyzer. CRP was analyzed by the latex agglutination quantitative method (Denka, Tokyo, Japan) with a TBA-200FR device (Toshiba, Tokyo, Japan). IL-6 was measured by enzyme-linked immunosorbent assay (ELISA)(Endogen, Woburn, USA). The assay ranged from 15-2,450 pg/mL; the intraassay and interassay coefficients of variation were below 10%. The detection limit of the IL-6 ELISA was 15-20 pg/mL. The indirect indices of protein intake, serum albumin, and normalized protein catabolic rate (nPCR){normalized to body weight derived from the urea distribution space (Vurea/ 0.58)} were obtained. For all variables, the values were the means of 3 measurements, except for IL-6, which was measured once. The IL-6 level was obtained for only 23 of 30 ESRD patients, because it was not detected in the other 7 patients.
Spearman correlation analyses were used to examine the relationship between the sleep variables and clinical variables of UA, IL-6, and CRP. A Pearson correlation analysis on nPCR and UA was conducted, considering that nPCR is a marker reflecting the nutrition state in ESRD patients. All statistical analyses were performed using Statistical Package for the Social Sciences (SPSS) version 11.5, and p values below 0.05 indicated statistical significance.
Among 30 patients, 15 (50.0%) had diabetes, 8 (26.7%) glomerulonephritis, 3 (10.0%) hypertension, and 4 (13.3%) unknown diseases as the main causes of ESRD. There was no significant difference in any of the sleep variables among the various groups of patients with different causes of ESRD (Kruskal-Wallis test, p>0.05). The prescribed drugs were antilipemic agents for 7 patients (23.3%), ACE inhibitors for one (0.03%), beta-blockers for 13 (43.3%), intravenous iron for 6 (20.0%), and statins for 7 (23.3%). The accompanying illnesses were ischemic heart disease in 2 patients (6.6%) and left-ventricle hypertrophy in 19 (63.3%), as well as diabetes in 15 (50%). The demographic/ clinical characteristics and Nocturnal polysomnography (NPSG) results are presented in
Among 30 patients with ESRD, 28 (93%) were diagnosed as having SAS or PLMS; 25 (83.3%) had a diagnosis of SAS, 18 (60%) OSAHS, and 18 (60%) PLMS. Fifteen patients (50%) had both SAS and PLMS, 13 patients (43.3%) had both OSAHS and PLMS, and 2 patients had diagnoses other than the primary sleep disorder. Upon reviewing the questions related to RLS in the SDQ (Q12. restless legs in falling asleep, Q31. restless legs disturb sleep) for 16 patients who completed it, RLS was diagnosed in 5 patients (31%).
Among the metabolic markers in 30 HD patients with ESRD, UA was positively correlated with the total sleep time (TST)(r=0.407)(
In this study, the proportions of patients with sleep disorders and SAS among 30 ESRD patients were 93.3% and 83.3% respectively, which were higher than those of previous studies (50-70%).
Increased UA was related to increased total sleep time (TST) and to decreased AHI and ODI in our study. Hsu et al.
Both OSA
As the CRP level increased, the LMI increased in our study, and there was no sleep parameter correlated with IL-6. In ESRD patients, although it was reported that sleepiness was induced by the abnormal production of TNF, a substance with somnogenic properties,
The limitations of our study are as follows. First, comorbid medical diseases such as hypertension and diabetes mellitus were not excluded, since these are relatively common in ESRD patients. Since it was also very difficult to exclude patients taking some form of medication (except hypnotics), these substances may have affected nocturnal sleep in our study. Second, our results are not conclusive, since confounding variables such as age, sex and BMI, which affect AHI and LMI, were not controlled. Third, we did not conduct multiple sleep studies to reduce the first-night effect and ESRD patients with possible primary sleep disorders were not excluded from our study, resulting in the presence of extreme values of the nocturnal sleep variables.
However, a limited number of studies on the relationship between sleep disorder and metabolic markers in ESRD patients have been reported. In Asian countries, there has only been one study using the sleep questionnaire without conducting NPSG. Considering the recent findings that ESRD is recognized as an oxidative stress disease or an inflammatory disease, it is meaningful that UA was significantly related to decreased severity of SBD, and CRP was related to increased severity of LM in our study. These findings suggest that UA or CRP levels reflect certain sleep disorders in ESRD patients. Further studies on the relationship of these markers to the pathophysiology of sleep disorders are necessary.
This work was supported by the Fund of
Demographic and clinical characteristics of end-stage renal disease (ESRD) patients receiving hemodialysis (N=30)
BMI: body mass index, ESS: Epworth Sleepiness Scale, HD: hemodialysis, CRP: C-reactive protein, nPCR: normalized protein catabolic rate
Nocturnal sleep characteristics in ESRD patients receiving hemodialysis (N=30)
ESRD: end-stage renal disease, TST: total sleep time (min), SE: sleep efficiency (%), SL: sleep latency (min), S1: stage 1 sleep (%), SWS: slow wave sleep (%), REMS: rapid eye movement sleep (%), AHI: apnea-hypopnea index (n/hr), LOS: lowest oxygen saturation (%), ODI: oxygen desaturation index (n/hr), LMI: limb movement index (n/hr)
Correlations of sleep architecture variables with metabolic markers ESRD patients receiving hemodialysis (N=30)
*p<0.05 (Spearman correlation analysis). ESRD: end-stage renal disease, TST: total sleep time (min), SE: sleep efficiency (%), SL: sleep latency (min), S1: stage 1 sleep (%), SWS: slow wave sleep (%), REMS: rapid eye movement sleep (%), UA: uric acid, IL-6: interleukin-6, CRP: C-reactive protein
Correlations between the variables related with sleep breathing disorder and limb movements and metabolic markers in ESRD patients receiving hemodialysis (N=30)
*p<0.05 (Spearman correlation analysis). AHI: apnea-hypopnea index (n/hr), LOS: lowest oxygen saturation (%), ODI: oxygen desaturation index (n/hr), LMI: limb movement index (n/hr), UA: uric acid, IL-6: interleukin-6, CRP: C-reactive protein