Sleep disorders in Wilson disease: a systematic review and meta-analysis

Abstract

Study Objectives:

Wilson disease (WD) is an autosomal recessive inherited disorder of copper metabolism resulting in pathologic accumulation of copper in many organs and tissues. Sleep disorders are highly prevalent in patients with WD. However, both prevalence rates and severity of different sleep disorders in patients with WD vary widely. The aims of the current study were to systematically review and perform a meta-analysis of the association between WD and prevalent sleep disorders, including insomnia, rapid eye movement (REM) sleep behavior disorder (RBD), excessive daytime sleepiness (EDS), sleep-disordered breathing (SDB), restless legs syndrome (RLS), periodic limb movement in sleep (PLM), cataplexy-like episodes (CLEs) and sleep paralysis, and objective sleep characteristics.

Methods:

We performed a systematic search of PubMed, EMBase, the Cochrane Library, PsycINFO and ISI Web of Science for case-control studies. A total of 7 studies with 501 participants were included.

Results:

We found that 54.1% of patients with WD experience sleep disorders and up to 7.65-fold higher odds compared to control patients. Specifically, patients with WD had higher rates of RBD, insomnia, and EDS based on self-reported questionnaires. No differences were observed in terms of RLS, PLM, or SDB between patients with WD and control patients. Furthermore, objective sleep disruptions based on polysomnographic studies included prolonged sleep onset latency and REM sleep onset latency, reduced total sleep time and sleep efficiency, higher percentage of stage N1 sleep and lower percentage of stage N2 sleep were observed in patients with WD.

Conclusions:

Our study indicates that sleep disorders are frequent in patients with WD. Future studies should examine the longitudinal association of WD with sleep disturbances.

Citation:

Xu J, Qingqing D, Qingsong Q, Vgontzas AN, Basta M, Xie C, Li Y. Sleep disorders in Wilson disease: a systematic review and meta-analysis. J Clin Sleep Med. 2020;16(2):219–230.

BRIEF SUMMARY

Current Knowledge/Study Rationale: Sleep disorders are highly prevalent in patients with Wilson disease (WD). However, both prevalence and severity of different sleep disorders in patients with WD vary widely. This is the first meta-analysis to examine the association of WD with sleep disorders.

Study Impact: Our findings indicate that WD is significantly associated with both self-reported sleep complaints and objective evidence of sleep disruptions, including rapid eye movement sleep behavior disorder, insomnia, excessive daytime sleepiness, prolonged sleep onset latency, and rapid eye movement sleep onset latency, reduced total sleep time and sleep efficiency, higher percentage of stage N1 sleep and lower percentage of stage N2 sleep. The results of the current meta-analysis provide important insight into the association of WD with self-reported and objective sleep disturbances.

INTRODUCTION

Wilson disease (WD) is an autosomal recessive inherited disorder of copper metabolism resulting in pathological copper accumulation in liver, cornea, brain, and other organs.¹ The estimated prevalence of WD worldwide is between 1/30,000 and 1/100,000 individuals.² The deficiency of a copper-transporting P-type adenosine triphosphatase (ATPase),^3,4 associated with the mutation of ATP7B gene (located on chromosome 13 band q14.3), results in accumulation of unbound copper in the liver and release in a free form into the bloodstream and in urinary rather than biliary elimination of copper.⁵

Copper accumulation results in a variety of symptoms ranging from neurologic and psychiatric disturbances to acute or chronic liver disease. It has been reported that approximately 40% to 50% of patients with WD present with neurologic symptoms and psychiatric disturbances, including dystonia, ataxia, tremor, choreoathetosis and parkinsonian-like extrapyramidal signs, depression, anxiety, and sleep disorders.^2,6–8 Although several studies have reported that 42% to 80% of patients with WD have sleep complaints, including insomnia,⁹ poor sleep quality,^10–14 daytime sleepiness,^10,11 rapid eye movement sleep behavior disorder (RBD),^15,16 cataplexy-like episodes (CLEs),^10,11 and sleep paralysis,^10,11 others failed to reveal such associations.^11–13

The high prevalence of sleep disorders among patients with WD may be related to pathologic accumulation of copper in sleep/wake pathways in the brain,¹⁷ WD treatment,¹⁸ and other WD- related symptoms (ie, motor, dysautonomic, psychopathologic, and/or metabolic disorders).² It has been reported that diffusion of neuronal lesions in the brain of patients with WD is widespread^5,19 and involves most of the sleep-wake regulation pathways, including brainstem nuclei, dentate nucleus, pons, thalamus, basal ganglia, external capsule, claustrum, and frontal lobes as well as a certain degree of diffuse brain atrophy.^19,20 Impairments of the brainstem levels and the monoaminergic system (rapid eye movement [REM]-off neurons) could explain REM sleep-related disturbances (ie, REM sleep behavior disorder [RBD]) in patients with WD. REM sleep is generated in the brainstem by acetylocholinergic neurons that belonging to the cholinergic system (REM-on neurons). Furthermore, abnormal metabolism of neurotransmitters, such as increased noradrenalin and decreased 5-hydroxyindole acetic acid (a 5-hydroxytryptamine metabolite) and homovanillic acid (a metabolite of dopamine)²¹ have been observed in patients with WD. Neuroimaging findings showed a complex pathogenesis involving both afferent^19,20 and efferent nigrostriatal dopaminergic projections^22,23 in WD. It appears that the abnormal metabolism of dopamine may relate to the presence of restless legs syndrome (RLS) and parkinsonian syndromes in patients with WD. Parkinson disease is known to be strongly associated with REM sleep disturbances including RBD.^24,25 Furthermore, reduced glucose metabolism was found in all regions of the brain except the thalamus.²⁶ All the aforementioned findings suggest that dysregulation of sleep-wake related neurotransmitters and/or lesions of sleep-wake pathways may be one of the underlying mechanisms that link WD to sleep disturbances. However, other studies showed that patients with WD with isolated liver disease and normal magnetic resonance imaging studies of the brain may also have sleep disturbances. Moreover, medications treatments for WD or WD-related symptoms may have effects on sleep. For example, de-coppering therapy had significant effects on prolonged REM sleep onset latency.¹⁸ Dopamine mimic drugs such as levodopa dose-dependent effects on sleep, with lower doses improving sleep quality while higher doses decreased sleep efficiency.²⁷ Antiepileptic drugs have different effects on sleep architecture, which can be beneficial or otherwise.²⁸

Though sleep disorders appear to be highly prevalent in patients with WD, no study has examined in a comprehensive way the prevalence and severity of sleep disorders in this group of patients. Given the variability in rates and severity of different sleep disturbances in patients with WD, the goals of this meta-analysis were to: (1) evaluate the prevalence rates and severity of insomnia, RBD, excessive daytime sleepiness (EDS), RLS, sleep-disordered breathing (SDB), CLEs, and sleep paralysis in patients with WD; and (2) examine objective sleep characteristics in patients with WD. The current meta-analysis would provide clinical physicians with comprehensive information of sleep disturbances in patients with WD.

METHODS

To examine the association of WD with sleep disorders, a systematic review and meta-analysis of articles searched from five public databases was conducted according to PRISMA statement guidelines.²⁹ This review and meta-analysis was registered with the PROSPERO database (registration number: CRD42018108674).

Search strategy and selection criteria

We performed a systematic literature search of electronic bibliographic databases, including PubMed, EMBase, the Cochrane Library, PsycINFO, and ISI Web of Science. The literature search was limited to articles in the English language and was updated on September 2018. We also included reference lists and review articles for additional studies that met the inclusion criteria. In this study, the following terms were used for searching: “hepatolenticular degeneration” OR “degeneration, hepatolenticular” OR “pseudosclerosis” OR “Wilson disease” OR “Wilson's disease” OR “cerebral pseudosclerosis” OR “hepatolenticular degeneration syndrome” OR “hepato-neurologic Wilson disease” OR “hepato neurologic Wilson disease” OR “copper storage disease” OR “neurohepatic degeneration” and “sleep” OR “sleep disorders.”

Two researchers (Xu JY and Deng QQ) independently assessed the articles for eligibility for inclusion. Any disagreements were resolved by the senior author (Li Y). The following inclusion criteria were set: (1) case-control studies with the WD group and control group; (2) WD diagnosis based on international diagnostic criteria and actual medical records; (3) sleep disorders were defined based on self-reported symptoms, questionnaires, clinical diagnosis, self-reported or objective sleep measures; (4) manuscripts written in English. We excluded studies if they were case reports, editorials, or review articles. If the same population was used in more than one publication and assessed the same sleep disorders, the larger sample was used for further analysis to avoid data duplication. If the same population was used in more than one publication and assessed different sleep disorders, the findings of those studies were included in this meta-analysis, and the larger sample was used to calculate the total sample.

Data extraction

Data were independently extracted by the first author (Xu JY) and the second author (Deng QQ) to confirm the accuracy. The following variables were manually extracted from all included studies: first author, publication year, research site (country), number of participants, number of females/males, mean age and numbers of participants with sleep disorders (ie, insomnia, EDS, RLS, RBD, SDB, CLEs and sleep paralysis) in the WD group and control group. Moreover, sleep disorders assessments and diagnostic criteria for WD were also included (Table 1). Objective sleep characteristics of each group assessed by polysomnography (PSG) are shown in Table 2.

Table 1.

Characteristics of included studies.

Table 2.

Polysomnography characteristics of patients with Wilson disease and control patients.

Definitions

We examined the association of WD with sleep disorders. Self-reported, clinical diagnosis, and objective measures of sleep disorders were used to verify the presence of sleep disturbances. The diagnosis of WD was based on genetic testing, clinical symptoms, and biochemical/histological tests. Pittsburgh Sleep Quality Index (PSQI)³⁰ was used to assess self-reported sleep quality and insomnia. In our study a total score of PSQI > 5 was considered insomnia,³⁰ the higher mean values of PSQI indicated the more severe insomnia. Epworth Sleepiness Scale (ESS)³¹ and Multiple Sleep Latency Test (MSLT)³² are the most widely used measurements to assess self-reported and objective EDS, respectively³³ in clinical settings. In this study, EDS was defined with ESS > 10³¹ or MSLT ≤ 8 minutes,³² the higher mean values of ESS and shorter latency of MSLT indicated the more severe self-reported and objective EDS, respectively. RBD was defined based on Questionnaire-Hong Kong (RBDQ-HK),³⁴ RBD screening questionnaire (RBD-SQ),³⁵ or International Classification of Sleep Disorders, Third Edition (ICSD-3) criteria,⁹ and the higher mean values of RBDQ-HK and RBD-SQ indicated the more severe RBD. RLS was defined based on the revised five criteria of the International Restless Legs Syndrome Study Group.³⁶ Both CLE³⁷ and sleep paralysis were defined by Uppsala Sleep Inventory. SDB was defined by AHI ≥ 5 events/h based on PSG study,⁹ the higher mean values of AHI indicated the more severe SDB. Overall sleep disturbances were defined as any sleep disorders, including insomnia, EDS, RBD, RLS, SDB, CLEs, and sleep paralysis. PSG-recorded sleep parameters were used to assess objective sleep characteristics.

Statistical analysis

A series of random effects meta-analysis were conducted using Comprehensive Meta-Analysis Software (Stata 14.0, StataCorp LLC, College Station, Texas). Number of events, including insomnia, EDS, RBD, RLS, SDB, CLEs, and sleep paralysis were used to calculate (1) the prevalence of overall sleep disturbances; (2) the prevalence of each sleep disorder in patients with WD; and (3) and odds ratio (OR) and 95% confidence interval (CI) of overall sleep disturbances in patients with WD compared to control patients. Mean values of PSQI, ESS, RBDQ-HK, RBD-SQ, and PSG sleep variables [ie, sleep onset latency (SOL), total sleep time (TST), sleep efficiency (SE), percentage of stage N1 sleep, stage N2 sleep, percentage of stage N3 sleep, percentage of REM sleep, AHI, and periodic leg movement index (PLMI)] and corresponding standard deviations (SD) in each group were entered into meta-analysis to determine weighted mean difference (WMD) of PSQI, ESS, objective sleep variables, and standard mean difference (SMD) of scores of RBD-related questionnaires (RBDQ-HK and RBD-SQ) between patients with WD and control patients. As both RBDQ-HK and RBD-SQ are used to identify RBD, we calculated Hedges g SMD of each questionnaire for further meta-analysis. We used I² index to test heterogeneity. An I² ≥ 50 indicates moderate to high heterogeneity. Furthermore, Egger test was used to test the potential publication bias.³⁸ A value of P < .05 was considered statistically significant.

RESULTS

The study was conducted following the workflow shown in Figure 1. A total of 1,454 articles and abstracts were identified in the search strategy. Screening of titles and abstracts resulted in 18 articles of which the full text was evaluated. Upon removal of ineligible studies, seven studies were included in this meta-analysis.

Figure 1. Search flow diagram.

Study characteristics

All included studies were case-control studies.³⁹ Among those, three were conducted in Brazil,^13,14,16 two were conducted in India,^12,18 one was conducted in Sweden,¹⁰ and one was conducted in the Czech Republic.¹¹ Because two studies from Brazil^14,16 used the same population but different assessments for sleep, we only included the largest sample from the study by Tribl et al¹⁴ to calculate sample size. Out of the seven studies, four studies^10–12,14 with 335 participants contributed data to prevalent rate of insomnia in WD; three studies^11,14,16 with 271 participants contributed data to prevalent rate of RBD; three studies^10–12 with 256 participants contributed data to prevalent rate of EDS; two^11,13 studies with 194 participants contributed data to prevalent rate of RLS; two studies^14,18 with 129 participants contributed data to prevalent rate of SDB; two studies^10,11 with 206 participants contributed data to prevalent rate of CLEs, and two studies^10,11 with 206 participants contributed data to prevalent rate of sleep paralysis. Furthermore, three studies^12,14,16 with a total of 220 participants contributed data to the mean PSQI, three studies^11,14,16 with 235 participants contributed data to the SMD of RBD questionnaire, and three studies^11,12,14 with 242 participants contributed data to the mean ESS. Four studies^11,13,14,18 used an overnight PSG recording to assess objective sleep characteristics: four studies^11,13,14,18 with 249 participants contributed data to the mean TST and SE, percentage of stage N1 sleep, percentage of stage N2 sleep, percentage of stage N3 sleep, and percentage of REM sleep; three studies^13,14,18 with 202 participants contributed data to the mean SOL, REM latency and AHI; two studies^13,14 with 153 participants contributed data to the mean PLMI. All of the patients with WD were recruited from inpatient or outpatient settings (hospitals or clinics), whereas control patients were recruited from community and/or staff and researchers of the hospital/clinic.

WD and overall sleep disturbances

When sleep disorders were combined, we found that 54.1% [95% CI (42.1% to 57.9%)] of patients with WD had overall sleep disturbances. As shown in Figure 2, patients with WD had 7.65-fold higher odds of overall sleep disturbances compared to control patients (OR = 7.65, 95% CI [2.02 to 28.93]; I² = 90.2%, P < .001).

Figure 2. Forest plot for meta-analysis of studies estimating the odds of patients with Wilson disease with overall sleep disturbances compare to control patients.

CI = confidence interval, OR = odds ratio.

WD and RBD

Random-effects meta-analysis of prevalence of RBD in patients with WD and SMD of RBDQ-HK and RBD-SQ were conducted between patients with WD and control patients. Among patients with WD, 11.2% (95% CI [1.00 to 17.48]; I² = 0%) had RBD based on the questionnaire and ICSD-3 diagnostic criteria. Furthermore, patients with WD had significantly higher SMD of RBD questionnaires (SMD = 0.68, 95% CI [0.26 to 1.10]; I² = 72.8%, P = .012), which suggested patients with WD had more severe RBD symptoms compared to control patients (Figure 3). No publication bias was observed (P = .818). In a sensitivity analysis, after excluding one study¹¹ that used questionnaire to define RBD, the prevalence rate of RBD in patients with WD decreased to 6.72%.

Figure 3. Forest plot for meta-analysis of studies estimating the standard mean differences of RBD scores in patients with WD and control patients.

RBD = rapid eye movement sleep behavior disorder, WD = Wilson disorder.

WD and insomnia

Random effects meta-analysis of prevalence of insomnia in WD and WMD of PSQI between patients with WD and control patients were conducted. Among patients with WD, 34.5% (95% CI [0.85–1.94]; I² = 54.8%) had insomnia. Furthermore, patients with WD had significantly increased scores of PSQI (WMD = 2.43, 95% CI [1.48 to 3.37]; I² = 0%, P = .551), which suggested patients with WD had more severe insomnia symptoms compared to control patients (Figure 4). No publication bias was observed (P = .175).

Figure 4. Forest plot for meta-analysis of studies estimating the weighted mean differences of PSQI in patients with WD and control patients.

CI = confidence interval, PSQI = Pittsburgh Sleep Quality Index, WD = Wilson disorder, WMD = weighed mean difference.

WD and EDS

Random effects meta-analysis of prevalence of EDS in patients with WD and WMD values of ESS between patients with WD and control patients were conducted. Among patients with WD, 19.2% (95% CI [0.95–4.61]; I² = 0%) had EDS. Furthermore, patients with WD had significantly higher ESS (WMD = 1.23, 95% CI [0.17 to 2.29]; I² = 0%, P = .383), which suggested patients with WD had higher levels of EDS compared to control patients (Figure 5). No publication bias was observed (P = 0.622). However, no original study examines the association between WD and hypersomnia, such as narcolepsy and idiopathic hypersomnia. Firneisz et al⁴⁰ reported idiopathic hypersomnia occurred in a patient with WD.

Figure 5. Forest plot for meta-analysis of studies estimating the weighted mean differences of ESS in patients with WD and control patients.

CI = confidence interval, ESS = Epworth Sleepiness Scale, WD = Wilson disorder, WMD = weighed mean difference.

WD, RLS, and periodic limb movements in sleep

Random-effects meta-analysis of prevalence of RLS in patients with WD and WMD of PSG-recorded PLMI between patients with WD and control patients were conducted. Among patients with WD, 28.9% (95% CI [0.23 to 101.14]; I² = 76.8%) had RLS. PSG studies indicated that patients with WD have similar PLMI with controls (WMD = 2.20, 95% CI [−0.20 to 4.60]; I² = 0%, P = > .99, Figure 6).

Figure 6. Forest plot for meta-analysis of studies estimating the weighted mean differences of PSG variables in patients with WD and control patients.

CI = confidence interval, PSG = polysomnography, WD = Wilson disorder, WMD = weighed mean difference.

WD and SDB

Random-effects meta-analysis of prevalence of SDB in patients with WD and WMD of AHI between patients with WD and control patients were conducted. Among patients with WD, 3.17% (95% CI [0.02 to 34.30]; I² = 68.6%) had SDB. No significant difference in terms of the severity of SDB has been observed between patients with WD and control patients based on PSG studies (AHI, WMD = −0.63, 95% CI [−1.67 to 0.41]; I² = 0%, P = .449, Figure 6). Publication bias was found in this association (P = .036). However, after assessing trim-and-fill analysis results remained unchanged.

Other sleep problems in patients with WD

Random-effects meta-analysis of prevalence of CLEs and sleep paralysis in patients with WD were conducted. Among patients with WD, 25.3% (95% CI [0.06 to 11.84]; I² = 60.4%) and 24.1% (95% CI [0.75 to 16.44]; I² = 66.4%) had CLEs and sleep paralysis, respectively.

Objective sleep characteristics in patients with WD

Random-effects meta-analysis of WMD of sleep variables between patients with WD and control patients were conducted. Compared to control patients, patients with WD had significantly longer SOL (WMD = 14.35, 95% CI [7.00 to 21.69]; I² = 0%, P = .572) and REM sleep latency (WMD = 27.40, 95% CI [12.03 to 42.76]; I² = 0%, P = .659), decreased TST (WMD = −30.26, 95% CI [−43.17 to −17.36]; I² = 2.4%, P = .381) and SE (WMD = −11.41, 95% CI [−14.16 to −8.66]; I² = 0%, P = .712), increased percentage of stage N1 sleep (WMD = 2.68, 95% CI [1.17 to 4.20]; I² = 0%, P = .682) and decreased percentage of stage N2 sleep (WMD = −6.09, 95% CI [−8.47 to −3.71]; I² = 0%, P = .473). No significant differences of percentages of stage N3 sleep and REM, AHI, and PLMI were observed. No publication bias was observed in TST, SE, percentage of stage N1 sleep, percentage of stage N2 sleep, percentage of REM sleep and PLMI (P > .05), whereas publication bias was found in SOL (P = .021), REM latency (P = .038), percentage of stage N3 sleep (P = .016) and AHI (P = .036). However, after assessing trim-and-fill analyses data remained unchanged.

DISCUSSION

This systematic review and meta-analysis indicate that sleep disturbances are highly prevalent in patients with WD, because 54.1% of patients with WD reported sleep abnormalities, presenting a 7.7-fold higher odds of sleep disturbances compared to control patients. Specifically, this is the first meta-analysis to report association between WD and RBD, whereas associations between insomnia and EDS based on self-reported scales/questionnaires were also found. Furthermore, PSG data suggest objective sleep impairment in patients with WD. These findings provide a comprehensive profile of the association between WD and sleep disorders, which may have implications for clinical management of WD.

Rapid eye movement sleep behavior disorder

Our findings suggested that patients with WD have higher prevalence of RBD and have more severe RBD symptoms in comparison with control patients. RBD is characterized by abnormal maintenance of muscle tone during REM sleep and is associated with a lesion or a dysfunction of the descending pathways (ie, locus subcoeruleus). WD appears to be associated with a dopaminergic deficit¹¹ and has been classified as a secondary parkinsonian syndrome.¹¹ It is well known that RBD is associated with the development of Parkinson disease, and parkinsonism is one of the most common manifestations in the neurologic form of WD. Thus, it is not surprising that patients with WD have higher rates of RBD compared to control patients. RBD and WD may share similar impaired regions of the brain.¹⁴ Tribl et al reported that four WD cases with RBD all had mesencephalic tegmental/tectal sonographic hyperechogenicities.¹⁴ In addition to the lesions and dysfunction of descending pathways, other factors, such as treatment with D-penicillamine, comorbidity of psychiatric problems, young age of onset of WD, and use of antidepressants and antipsychotics may also be associated with RBD in WD.^14,16 Moreover, a recent study showed that RBD could appear before any other symptom of WD.¹⁶ Another case reported progressive disappearance of RBD symptoms with improvement of WD after trientine treatment.³⁰ Furthermore, consistent with symptoms improvement, repeated magnetic resonance imaging studies showed progressive disappearance of pontine and mesencephalic tegmental lesions with WD treatment. However, in this case, it is not clear if the disappearance of the tegmental lesions was spontaneous or an effect of WD treatment. Future studies of recently diagnosed cases of RBD in WD are needed to provide further insight into the chronologic development and causal relationship of clinical manifestations, morphologic lesions, and treatments effects on RBD. In the sensitivity analysis, after excluding a single study that used questionnaire to define RBD, the prevalent rate of RBD in WD decreased from 11.2% to 6.72%. However, care should be taken to interpret the change of the prevalence rate of RBD. First, limited studies examined the association between WD and RBD. Among the three included studies, one used a questionnaire to define RBD and the other two used ICSD-3 criteria to define RBD. Second, it has been reported that the sensitivity and specificity of the RBD-SQ in patients with WD were 0.92 and 0.56, respectively.³⁷ The high sensitivity and relatively low specificity of RBD-SQ may overestimate the presence of RBD in patients with WD. More studies based on ICSD and Diagnostic and Statistical Manual of Mental Disorders diagnostic criteria need to be performed to examine the association between WD and RBD.

Insomnia

It has been reported that insomnia, which can be the first symptom of WD, worsens during the process of WD, and may also appear during voluntary WD-specific medications withdrawal.¹⁴ Results of the current meta-analysis indicate that WD is associated with insomnia, with rates as high as twofold compared to control patients. Moreover, our findings suggest that patients with WD have more severe insomnia symptoms compared to control patients. Consistent with self reports, PSG data also confirmed this association. The high prevalence of insomnia may be related to lesions in the sleep-wake regulation systems,⁵ iatrogenic effects (ie, levodopa)⁴¹ and comorbidity of mood disorders (ie, depression and anxiety)⁴² in patients with WD. Furthermore, use of medications, such as dopaminergic, antiepileptic, and antidepressant agents have different effects on sleep architecture.^18,43,44 Because all the individuals included in the current meta-analysis are medication treated, the independent effects of WD on sleep are not clear. Future PSG studies should examine the association between sleep and WD in drug-naïve patients. Cross-sectional findings cannot support a causal association between insomnia and WD, neither possible effects of insomnia in the course of WD. Further longitudinal studies should address these questions.

Excessive daytime sleepiness

Previous studies that assessed EDS in patients with WD were inconsistent.^10–12 Two studies reported higher ESS scores in patients with WD, whereas another study did not observe such an association. Only one study measured objective EDS and reported 14% of the patients with WD had objective EDS as defined by MSLT ≤ 8 minutes.¹¹ Findings of the current meta-analysis indicate that almost 20% of patients with WD have EDS and have more severe levels of EDS compared to control patients based on ESS scores. It has been reported that EDS is more prevalent in the neurologic form of WD,¹¹ indicating that EDS in WD may be related to the lesions of sleep-wake pathways. Furthermore, a case reported that EDS disappeared after 14 months of d-penicillamine treatment for WD.¹² However, another case reported no EDS improvement after treatment of WD.⁴⁵ Finally, it is difficult to rule out the effects of iatrogenic effects of symptomatic treatments on EDS, such as use of sedative medications, antidepressants, antiepileptic medications, and dopaminergic medications in patients with WD.

RLS and periodic limb movements in sleep

Data on RLS and periodic limb movements in sleep (PLMS) in WD are scarce. Only two studies examined the association of WD with RLS^13,16 and PLMS,^13,16 and the findings were inconsistent. Although the abnormalities of iron and dopamine metabolism and therapeutic use of dopaminergic, antidepressants, and antiepileptic agents⁴⁶ for reducing symptoms in patients with WD may suggest a link between WD and RLS/PLMS, the current meta-analysis did not observe significant association of WD with RLS/PLMS. Because we only included two studies in the meta-analysis of RLS/PLDM in WD, as well as the small sample size of each of the included studies, the negative findings of the association between WD and RLS/PLDM need to be interpreted very carefully. Further studies with larger sample size are needed to examine the association of WD with RLS and PLDS.

Sleep-disordered breathing

Based on the current meta-analysis, patients with WD did not have higher prevalence of SDB and mean AHI compared to control patients. Because we only included two studies for calculating pooled prevalence rate of SDB and three studies for calculating pooled AHI of WD in this meta-analysis, as well as the small sample size of each of the included studies, the negative findings of the association between WD and SDB needs to be interpreted very carefully. It has been reported that SDB was found in obese patients with WD¹⁶ and the patients on de-coppering therapy had more mixed apneas compared to other patients.¹⁸ Future studies should examine the association between SDB and different forms of WD, as well as the association between different forms of sleep apneas and WD.

Objective sleep characteristics

Objective evidence of sleep impairment is observed in patients with WD based on PSG studies. Patients with WD have more fragmented and lighter sleep, presented as prolonged SOL, decreased TST and SE, increased percentage of stage N1 sleep and decreased percentage of stage N2 sleep compared to control patients. Furthermore, our findings indicate that patients with WD had lower pressure of REM sleep presented as prolonged REM sleep latency. The lesions and dysfunction of sleep/wake pathways in brain⁵ may be one of the underlying mechanisms linking WD to objective sleep disturbances, but it needs to be further studied in future. Prolonged REM latency may also associate with use of de-coppering therapy.¹⁸

Strengths and limitations

To our knowledge, this is the first meta-analysis to examine the association of WD with sleep disorders. Furthermore, the current study presents both self-reported sleep complaints and objective evidence of sleep disruptions in patients with WD. The results of the current meta-analysis provide important insight into the association of WD with self-reported and objective sleep disturbances. Several limitations need to be acknowledged. First, data on sleep in WD are scarce; only 7 studies and a total of 501 individuals were included in this meta-analysis. The relatively small sample size and limited studies may affect the statistical power of the current meta-analysis. Future studies including larger samples of patients with WD are required to confirm both our positive and negative findings. Second, no study reported data on the longitudinal association between WD and sleep. Longitudinal studies will help understand the causal relation between WD and sleep disturbances. Third, the limited studies and small sample size does not allow further examination of the association between different/rare forms of WD and sleep disturbances. Finally, objective PSG findings were based on a single-night PSG recording and the first-night effect may affect objective sleep parameters. However, a study by Gaines et al suggested that objective measures of sleep duration, such as TST and SE, in fixed-time recordings are relatively stable across consecutive nights.⁴⁶

CONCLUSIONS

Sleep disorders, as assessed by self report and objective measures, are highly prevalent in patients with WD. Future, large well-designed studies, both on case-control and prospective cohorts, are needed to confirm the current findings and provide more evidence of the association between WD and sleep. Findings of the current meta-analysis provide important evidence of early detection and intervention for sleep disturbances in WD patients.

Research agenda

1. Longitudinal studies that examine the association between WD and sleep disturbances are warranted in order to understand the causal relationship between WD and sleep disturbances.

2. Examination of the association between drug-naïve WD patients and sleep disturbances is needed.

3. Examination of the association between different forms of WD and sleep disturbances is needed.

4. Examination of the association between WD and RBD based on ICSD and Diagnostic and Statistical Manual of Mental Disorders diagnostic criteria of RBD is needed.

DISCLOSURE STATEMENT

All authors have read and approved the manuscript. This study was supported by National Natural Science Foundation of China (No. 81600068 and 81573510), the Young Elite Scientists Sponsorship Program by CAST (No. YESS20160072), Medical Science Foundation of Guangdong Provence (A2018296), Department of Education of Guangdong Province (2017KTSCX065) and Grant for Key Disciplinary Project of Clinical Medicine under the Guangdong High-level University Development Program. The authors report no conflicts of interest.

ABBREVIATIONS

AHI

apnea-hypopnea index

CI

confidence interval

CLEs

cataplexy-like episodes

EDS

excessive daytime sleepiness

ESS

Epworth Sleepiness Scale

MSLT

Multiple Sleep Latency Testing

NREM

non-rapid eye movement

OR

odds ratio

PLM

periodic limb movements in sleep

PLMI

periodic limb movement index

PSG

polysomnography

PSQI

Pittsburgh Sleep Quality Index

RBD

rapid eye movement sleep behavior disorder

RBDQ-HK

RBD Questionnaire-Hong Kong

RBD-SQ

RBD Screening Questionnaire

RLS

restless legs syndrome

SD

standard deviation

SDB

sleep-disordered breathing

SE

sleep efficiency

SMD

standard mean difference

SOL

sleep onset latency

TST

total sleep time

WD

Wilson disease

WMD

weighted mean difference