Liver disease severity is poorly related to the presence of restless leg syndrome in patients with cirrhosis
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.263171
Source of Support: None, Conflict of Interest: None
Keywords: Chronic liver disease, cirrhosis, quality of life, sleep disorders, sleep-related movement disorders
Cirrhosis is the end result of long-standing liver injury. It is characterized by replacement of normal liver tissue by thick strands of fibrosis and alteration in the liver architecture by regenerating nodules. Patients with cirrhosis frequently suffer major life-threatening complications such as variceal bleeding, ascites, spontaneous bacterial peritonitis, hepatic encephalopathy and hepatocellular carcinoma. In addition, the condition is associated with reduction in quality of life, impaired capacity to work, easy fatigability, and occurrence of minimal hepatic encephalopathy, sexual dysfunction and sleep disturbances. Recently, restless leg syndrome (RLS) has been identified as a common additional manifestation among patients with cirrhosis.
RLS, a common sleep disorder, is characterized by an irresistible urge to move limbs, which improves at least partially with leg movements and occurs particularly at bed time. Though often ignored by physicians as a relatively-benign clinical condition, RLS is an important determinant of quality of life. Large population-based studies have shown that RLS affects up to 3-10% of healthy individuals. It is frequently observed in persons with peripheral neuropathy, and its prevalence is increased in several systemic illnesses, such as chronic kidney disease (CKD), hypothyroidism, diabetes mellitus and iron deficiency anaemia.
The prevalence of RLS has been shown to be higher in patients with cirrhosis than in healthy groups in USA, Japan  and India. In India, RLS has been well described in several similar disease groups, such as patients with chronic kidney disease  and iron deficiency anaemia. Further, in a previous study, we found that the prevalence of RLS among Indian patients with cirrhosis was 6.6%, i.e., higher than that in healthy persons (0.8%). However, in that study, most of the patients had mild liver disease, and the patients with advanced liver disease were unrepresented, making it difficult to assess the frequency of RLS in patients with advanced cirrhosis and whether the occurrence of RLS was related to the severity of liver disease. In the current study, we therefore assessed the frequency of RLS in patients with different severities of liver dysfunction.
The current study enrolled adult patients with cirrhosis who attended our department's outpatient services between August 2016 and July 2018. The diagnosis of cirrhosis was based on the presence of typical clinical, laboratory, radiologic and/or endoscopic features, and on liver biopsy where available. For each subject, relevant medical history was reviewed, a physical examination was performed, and relevant clinical and laboratory test findings were recorded. Patients were excluded if they had evidence of (i) deterioration of clinical condition, as indicated by fresh appearance or worsening of jaundice, ascites or encephalopathy, a reduction in urine output, gastrointestinal bleeding or hospitalization, in the last 4 weeks; (ii) peripheral neuropathy or any other neurological illness; or (iii) pregnancy.
The severity of liver disease was assessed and graded as Child-Turcotte-Pugh (CTP) class A, B or C; using Model for end-stage liver disease (MELD) score, and as a binary variable, i.e., as compensated or decompensated liver disease. Decompensated liver disease was defined using the APASL criteria, i.e., presence, currently or in past, of any one or more of the following: ascites, hepatic encephalopathy, variceal bleed, serum bilirubin >2.5 times the upper limit of normal) and prolonged prothrombin time (prolonged by >3 s or international normalized ratio (INR) >1.5), or any combination of them.
Each subject was screened for possible RLS by a gastroenterologist (AG or VJH), using a previously validated tool for the detection of this condition. It consists of a set of five questions administered by a physician in Hindi, the local language. There were: (i) an urge to move the legs, usually accompanied by uncomfortable and unpleasant sensations in the legs; (ii) which begins or worsens during periods of rest or inactivity; (iii) occurs only or is worse in the evening or night than during the day; (iv) is partially or totally relieved by repeated leg movements, and (v) the occurrence of above features is not solely accounted for by another medical or behavioural condition. If the person provided a positive response to none or only one of the above questions, he was considered as not having RLS. Persons with affirmative responses to two or more of these five questions were then evaluated by a neurologist (VKP) on the same day. If the neurologist obtained positive responses to all the five questions, the person was diagnosed to have RLS.
In the subjects with RLS, the severity of symptoms was assessed using a validated Hindi translation of International RLS severity (IRLS) scoring system. This score consists of a set of 10 self-administered questions, each of which is scored on a scale extending from 0 to 4. The scores obtained for individual questions are then aggregated to yield a total score ranging from 0 to 40. Based on the IRLS score, RLS was graded as mild (0-10), moderate (11-20), severe (21-30) or very severe (31-40). All the subjects diagnosed with RLS were provided the standard-of-care management by the neurologist (VKP) in the research team.
Data for categorical variables were expressed as proportions and compared using chi-squared test, and those for numerical variables were expressed as median and range, and compared using Wilcoxon rank sum test. The tests were done using Statistical Package for the Social Sciences (SPSS) software, version 17.0.
The sample size required to compare the prevalence rates of RLS in compensated and decompensated CLD patient groups was calculated assuming the prevalence of RLS in compensated LC as 6% (from the previous study by our group), and that in decompensated CLD patients as 17%, from a previous study of predominantly decompensated CLD in Japan. Using the level of significance of P < 0.05 and study power at 80%, the sample size was calculated to be 150 participants in each group.
The study had been approved by our institution's Ethics Committee. Each study participant provided a written, informed consent. IRLS scoring tool was used after obtaining permission from the MAPI Trust.
A total of 356 patients with cirrhosis (median age [range] 48 (18-83) years; 241 [67.7%] male) were enrolled and evaluated for RLS. Data on clinical characteristics, causes of liver disease, laboratory findings, liver disease severity scores and associated conditions of these subjects are summarized in [Table 1]. None of the patients was receiving or had received a psychotropic agent or a dopamine agonist in the last 6 weeks.
Of these 356 subjects, 76 had at least one of the five RLS criteria, including 48 (13.5%) who had two or more RLS criteria. These 48 subjects underwent re-evaluation by the neurologist and RLS was confirmed in 36 (10.1%) patients. The clinical characteristics, liver disease severity scores and the frequency of associated conditions in those with or without RLS are summarized in [Table 2].
Of the 36 persons with cirrhosis and RLS, 8 (22.2%) had diabetes mellitus, 2 (5.6%) had hypothyroidism and none had chronic kidney disease. The cause of cirrhosis in these 36 patients with RLS was: HCV infection in 17 (47.2%), cryptogenic in 9 (25.0%), hepatitis B virus infection in 6 (16.7%), alcohol in 3 (8.3%) and hepatic venous outflow obstruction in 1 (2.8%). The severity of RLS was found to be mild in one patient (2.8%), moderate in 15 patients (41.7%), severe in 19 patients (52.8%) or very severe in one (2.8%) patient. The causes of cirrhosis and presence of associated conditions were no different in persons with and without RLS.
Distributions of two of the three measures of severity of liver disease, namely the CTP score/class and the presence of decompensation, were similar among persons with and without RLS. By comparison, the MELD score was marginally lower among persons with RLS than in those without RLS.
A comparison of RLS prevalence and severity between those with compensated or decompensated cirrhosis is shown in [Table 3].
This prospective study assessed a large cohort of patients with cirrhosis of different severity and causes for the presence and severity of RLS. The overall prevalence of RLS was found to be around 10%. Most of the patients with RLS had moderate or severe symptoms, with very few having mild or very severe symptoms. The prevalence and severity of RLS had only a weak association, if any, with the severity of liver disease.
RLS is a type of sleep-related movement disorder, which is poorly recognised and hence has not received adequate attention. Though the exact pathogenesis of RLS is not known,, it is associated with several common systemic conditions, such as iron-deficiency anaemia, diabetes mellitus, chronic kidney disease, pregnancy and hypothyroidism. In most people, RLS runs a relatively benign course without causing much disability or functional impairment. Recognition and management of RLS is important because its presence adversely affects the quality of life (QoL) of the affected persons.,
In Caucasian populations, the prevalence of RLS among healthy persons has ranged  from 5% to 15%. Data from Singapore, Japan,, and Turkey  indicate that RLS is less common in Asian populations. Also, in the only previous population-based study of RLS from India, which included 1266 adults from Bangalore, the prevalence of RLS was quite low, being only 2.1%. The reason for this lower prevalence of RLS in Asia is unclear.
Data on the prevalence of RLS in persons with cirrhosis are quite limited. The current study extends our previous work on the prevalence of RLS in patients with cirrhosis, by studying a larger number of patients, and, in particular, those with advanced cirrhosis. It identified RLS in 10% of patients with cirrhosis, a rate similar to the 6.6% prevalence in our previous study. By contrast, the prevalence of RLS in patients with cirrhosis in other studies from different countries , has been reported to vary from 17% to 62%. Thus, the current study confirms our previous finding of a relatively lower prevalence of RLS in Indian patients with cirrhosis than in those of such patients in the Western populations.
The lower prevalence of RLS in our patients may have several explanations. First, the prevalence of RLS is lower in the Indian healthy population than in Europe and North America, and in Indian patients with kidney disease and maintenance hemodialysis, a well-recognised risk factor for RLS., Second, our criteria for diagnosis of RLS were quite stringent, and included a thorough evaluation by a trained neurologist, whereas other studies used either telephonic surveys or self-administered questionnaires;, the latter may increase the risk of false positive diagnosis of RLS. In fact, in our study, only about 60% of those with two or more features of RLS at the screening evaluation were finally diagnosed with RLS.
Our starting hypothesis was that RLS would be more common in persons with more advanced liver disease. However, we did not find much association between the prevalence of RLS and severity of liver disease. The two previous studies had also failed to find such an association,, though those were not done specifically to study this relationship. Our study was specifically aimed to study this relationship and assessed the severity of liver disease using three different systems. Two of these, i.e., the CTP class and the APASL criteria for the diagnosis of decompensated disease, largely depend on clinical variables and showed no relation with the occurrence of RLS. The third system for the assessment of severity of liver injury, i.e., MELD, a laboratory test-based score, somewhat counter-intuitively showed that patients with RLS had a slightly milder liver disease. We believe that our failure to find RLS more frequently in persons with more severe liver disease may be related to the symptoms of RLS getting overshadowed by other more important symptoms of advanced liver disease and morbidities associated with it. It is also possible that the RLS symptoms improve with worsening of liver function.
The dopaminergic system is believed to play a role in the pathophysiology of RLS.,,, Cirrhosis is associated with derangements in the dopaminergic pathways., Thus, basal ganglia, in particular the dopaminergic tracts originating from A11-A14 nuclei, have been proposed as a likely site for the origin of RLS., These tracts lie in close vicinity of the brain centre that controls the sleep-wake cycle; this may explain the worsening of RLS with sleep.
Occurrence of RLS in cirrhosis could have other explanations. RLS is known to be associated with iron deficiency, which is quite common in patients with cirrhosis. However, in our study, the patients with and without RLS had comparable haemoglobin levels. Alternatively, RLS in cirrhosis could be related to an increased body fluid volume (as in the case of chronic kidney disease, which too is associated with increased RLS), or to electrolyte disturbances such as diuretic-induced hypokalemia, dilutional or diuretic-induced hyponatremia, hypocalcemia or hypomagnesemia. Alternatively, vitamin D deficiency, reduced physical activity and muscular strength, and increased serum levels of endotoxins and inflammatory cytokines, all of which are common in cirrhosis, could also play a part.
Most of our patients had moderately severe RLS. Though the prevalence of RLS was much higher in a study from USA than in our patients, the mean RLS severity in that study too was moderate. It is possible that the presence of other symptoms in patients with liver disease masks the symptoms of RLS when the latter is mild.
Our study is limited by the fact that some patients had other risk factors for RLS, such as anaemia and diabetes. However, we believe that this is unlikely to have altered our conclusions much since the prevalence of RLS was similar in persons with and without these associated conditions. Iron deficiency anaemia, a common cause for RLS, is commonly seen among patients with cirrhosis. Majority of our study participants, in both groups, had anaemia but we did not test them for serum iron, serum ferritin, and serum total iron binding capacity to evaluate their body iron stores.
Also, we did not do nerve conduction studies to exclude a co-existing neuropathy, which could have occurred due to alcohol intake or diabetes, and may have caused RLS. However, only a few of our patients had alcohol-related cirrhosis, and RLS was equally common in persons with cirrhosis related to alcohol or other causes.
In conclusion, RLS is present in about 10% of the Indian patients with cirrhosis, and very severe RLS is quite uncommon among these patients. Furthermore, the presence of RLS appeared to be independent of the severity of liver disease. Thus, RLS is unlikely to be a major concern in Indian patients with chronic liver disease.
The authors acknowledge and thank the MAPI trust for permission to use the Hindi translation of IRLS Severity Scoring system without any charge.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Table 1], [Table 2], [Table 3]