In this study sample, the mean age of population was 60.47 ± 13.92 years, similar to 60.11 ± 13.69 years in Taiwanese HD population in 2005 [5]. The proportion of female was 52.3% of patients, rather close to the proportion of female in HD patients in Taiwan [20]. After analyzed the age and gender distribution of sample with good of fit test, there was no significant difference.
Our study skips off the weakness of previous surveys on sleep disturbances in ESRD patients on maintenance HD in Taiwan. The patients assessed in this study are widely distributed in Taiwan and do not limit to centers in homogenous geographic district. The second shortcoming that we overcame is the sample size and patients were selected by systematic random selection in this study. Another weak point bypassed is the dialytic strategy by physicians setting, whereas patients were dialyzed in 63 different facilities. These facilities distributed from hospital-based to clinic-based units. This permits us to exclude the deviation of therapeutic policy by single center or multiple centers with similar physician training program from same health care system. Furthermore, the previous studies in Chinese ESRD patients with sleep disturbances have focused on CAPD or description of small sample size in HD without or with intervention [13, 19, 55, 77, 78]. Thus, we divided the HD patients into two groups and compared them, avoided the different therapeutic bias between groups.
After first report of sleep disorders in HD patients by Strub et al. [37] in 1982, our study still clearly presents high prevalence of sleep complaints in HD patients.
Our data revealed that age, family income, sleep medicine, RLS symptom, and depression are significantly independent variables related to QOS. Whereas gender, time on dialysis, sum of number of co-morbidity, education status, occupation status,
lifestyle behavior, and marital status are not the significant determinant of QOS.
The mean global PSQI scores was 8.49 ± 4.64, similar to 8.7 ± 4.5 in Canadian HD population reported by Iliescu et al. [15]. Compared with scores of normal healthy control group from Buysse et al. [66], 2.67 ± 1.7, the global PSQI scores revealed rather poor quality of sleep (QOS) in our HD patients. Another study [76]
in elder Taiwanese with mean age of 70.22 years, the global PSQI scores was 6.13
±1.1. The scores were still lower than our HD patients. The prevalence of sleep complaints in this study was 66%. Whereas Hui et al. [13] reported that the prevalence of sleep complaints in CAPD patients was 52% in Taiwan, but they definite the ‘poor sleepers’ when global PSQI scores over 8. The similar result that 61% of patients had self-reported sleep disorders was reported in Hong Kong [55].
This study is also similar to the 50-86% prevalence of sleep-wake complaints in Caucasian HD populations expressed in previous studies [12, 14, 15, 46, 49, 78].
The mean score for seven components of the PSQI in our HD patients, Canadian HD population [15], US healthy control [66], and elder Taiwanese [76] are shown in Figure 5-1. Figure 5-2 shows the mean score in the seven components of all HD subjects, ‘poor sleepers’, and ‘good sleepers’ in this study. The mean score in the seven components revealed the highest score in sleep latency. It meant that HD patients were difficult to initiate a falling asleep, similar to previous studies [15, 76].
The best performance in the seven components of sleep quality was daytime dysfunction. The mean scores of daytime dysfunction were lower in our HD patients than Canadian HD population (0.79 vs 1.08) [15] and elder Taiwanese (0.79 vs 1.08) [76]. The differences of daytime dysfunction between HD patients and elder Taiwanese might be due to the age. Another reason, the proportion of patients without work might be the contributing factor. The daily activity of patients is free from work and keeping up enough enthusiasm to get thing done is not so important
to patients. The differences of daytime dysfunction between Taiwanese and Canadian HD patients need to further study.
Both quality and efficiency of sleep were significantly worse in ‘poor sleeper’
than ‘good sleeper’. It was reasonable that there was significantly higher proportion of using sleep medicine in ‘poor sleeper’ than ‘good sleeper’. Contrarily, the proportion of using sleep medicine in ‘good sleeper’ was also 21.3%. In other words, the higher prevalence of HD patients with sleep disturbances may be found. An important and urgency message to health care system of the differences in QOS is the advanced evidence of a 48.66% prevalence of using sleep medicine in this HD population.
The age was significantly related to QOS in this study. Despite this study was similar to previous studies [14, 49, 59], it was different to the study of Chinese CAPD patients in Hong Kong [55] and study of US [12]. One of the most common complaints of elder is difficult to fall asleep. This problem results in daytime napping and poor sleep efficiency, poor concentration and memory, even increasing risk of accident. Aging is a nature process with change in the circadian rhythms. The decreased level with age in neurohormone melatonin had been found to play an important role in the circadian rhythms and quality of sleep in human survey [80]
and dove study-the animal with similar circadian rhythms to human [81]. However, a study found that the mean melatonin concentrations of HD patients were significantly higher compared with that of intact kidney controls [82]. Hence, the true reason of changing sleep quality of HD patients with aging is still unclear.
Furthermore, in our study, ‘poor sleepers’ had a greater proportion of female, but there was no significant relationship after adjusting the difference of the other variables. Contrarily, poor QOS was significantly related to male gender reported by another studies [14, 60]. Kimmel et al. [83] surveyed 26 HD patients with sleep
disturbances by polysomnography and found that there was no difference in gender.
High TG was found to associate with global PSQI by study in Taiwanese CAPD patients [13]. The other factor, phosphate, had also been reported that it was related with QOS [60]. There was also no significant difference of metabolic data between
‘good sleepers’ and ‘poor sleepers’ in our study. The urea reduction rate and Kt/V (the measurements of small solute clearance), both were not significant difference between ‘poor sleepers’ and ‘good sleepers’. Hanly et al. [62] found that sleep latency in 58% of 24 HD patients was associated with BUN, but the others [12, 14, 15, 54] reported that there was no relationship between sleep latency and BUN. A study reported that the Kt/V was not correlated to sleep behaviors [54]. There was also no relationship between Kt/V and sleep quality in Canadian HD patients [15].
Hanly and Pierratos found that the improvement of sleep apnea was significantly in patients with lower creatinine undergoing nocturnal HD than conventional HD [43].
A study with objective polysomnography measures of nocturnal sleep in the CKD patients and HD patients with the similar metabolic data in 2005 [48], the authors suggested that the differences of sleep disturbances in the chronic renal disease patients and HD patients may be due to different etiologies. It may not be possible to find the effect on sleep quality by the laboratory parameters going. Another word, we need a more delicate measurement to detect the difference (such as on line day-to-day measurements of solute removal) or measurements of another solute removal that is still unused. Another measurement to check a substance with larger molecular weight (like theβ2-microglobulin) is necessary for the evaluation of HD adequacy.
The socio-economic, psychological factors, and intrinsic sleep disruption (RLS, PLMS) may play a more important role in HD patients. It is well known that uremia and dialysis have medical, social, and economical consequences, both for the
individual with dialysis and for Bureau of National Health Insurance (BNHI). The proportion of the patients without occupation status in this HD population run up to 78%, three persons of 4 people are out of work. The lost of productivity makes patient sad, worry, even depression due to illness. The morbidity cost was inevitably high due to absence from work. Most of the family with a member of ESRD patient suffered from financial tension. The economic burden bothered the patient, and the poor quality of sleep was the result. A study on Canada general population revealed low family income was correlated to sleeping difficulties [84]. Our study also found that the HD patients with family income less than NT $ 60,000 per month in Taiwan was significantly inverse correlation to the quality of sleep. This result agrees with previous survey from the cohort study of general population with middle-aged in US [85]. That study found that annual family income below 50,000 dollars and depressive symptoms were predictors of insomnia complaints [85].
Although the incidence of restless legs symptom in our HD patients is 51.4%, it was much higher than that reported from other countries [56, 18, 79]. It was similar to that reported by Walker et al. (57.4%) [14], but it was still lower than that of HD patients in Hong Kong (70%) [78] and HD patients in US (83%) [12]. In spite of different estimation in prevalence of RLS was persistent in previous studies, it might be due to different race, culture, socio-economic supports, and therapeutic styles. The most important reason is that the diagnosis of restless legs by different criteria described. RLS represents a complex phenomenon that is difficult to measure objectively. However, the exact results may vary between individuals with ambiguously definite criteria. Furthermore, applying a uniform tool to investigate in RLS field is a good method. A convenient instrument to widely survey and easily compare with other studies is the consent. The IRLS is suitable to play the role for widely surveying in HD patients [75].
Our finding’s clinical implication of restless legs symptom is reflected on quality of sleep. This study revealed that the proportion of severe restless legs symptom by IRLS was 15.48 % of HD patients, similar to report of Unruh et al. [18].
They also found that the sleep-quality score was significantly lower in dialysis patients with severe restless legs symptom after adjusting for other factors. In addition, they reported that the worse health related quality of life and shorter survival time were found in dialysis patients with severe restless legs symptom. In US, a study presented the significantly decreased survival time in patients with RLS by polysomnography after taking age, gender, and time on dialysis into account [39].
The Framingham Heart Study [44] found right ventricular wall thickness was associated with sleep-disordered breath. In Italy, a study reported that sleep apnea inducing hypoxemia was associated with cardiovascular events [45]. All of these three studies found that the patients complained the fragmented sleep at night. More severe RLS that the patients complained more urge to move at night resulted the fragmented sleep. Poor QOS are the sequence of severe RLS.
Many studies reported that ESRD patients had psychologic problems, such as depression, sadness, anxiety, worry, and stress [12, 54, 56, 69, 73]. We found highly significant correlation between the PSQI scores and BDI scores. This was also similar to the previous study that reported the association between quality of sleep and depression in ESRD patients in Canada [15]. In our experience, the prevalence of depression in HD patients was higher than general population, which is one of the most common psychiatric problems in HD population. In this study, there were over 40 % of patients with mild to severe degree of depression symptom. Most of the depressive patients have never been referred to psychiatrist. Depression is one of diseases in HD patients that is easily underdiagnosed and ignored by health care system [86, 87]. In 2006, a study reported that the prevalence of depression in HD
patients was 38.7% and 41.9% measured by BDI-II ≧14 and nurses’ diagnosis in Canada [87]. A cross-sectional study in 12 countries with 9382 HD patients by random selection revealed that the prevalence of depression by Center for Epidemiological Studies Depression Sceening Index (CES-D) was 43 %, but only 2
% (Japan) to 21.7 % (United State) was diagnosed by physician [87]. There are a lot of patients with depression without adequate exploration. A convenient and valid instrument is needed to screen depression widely in HD patients. We also found that the levels of severity in depression were positively significant association with the poor quality of sleep. The psychosocial factors are highly associated with survival rate and hospitalization, especially depression [87]. Same as general population, HD patients have high suicide rate [88]. Early diagnosis of depression can help ESRD patients who are at higher risk of suicide. Psychologic interventions can release partial burden on patients, which can prevent the tragedy. In US, a study reported that higher levels of perceived social support were significantly inverse relationship to mortality risk [89]. They also reported that quantifying depressive symptom by BDI was a risk factor to survival.
Our study was multi-facility with large systematic random selected samples. The age and gender distribution is similar to national HD population [5, 20]. It may be possible to generalize to Taiwanese HD patients after increasing the sample size and conducting a long-term cohort study.