The diagnosis restless legs syndrome (RLS) in children depends on the history told by the child and his parents. The description of symptoms given by the child himor herself is most important. Additional criteria are, among others, the results of polysomnography (PSG). Description of the presenting symptoms is the aim of the study
Survey in two European pediatric sleep centers of presenting symptoms in children who after a detailed work-up proved to have RLS.
Fifty-two percent of the 31 children presented with symptoms similar to those mentioned in the “four questions” relevant for the diagnosis of RLS. In the other patients the description included often very colourful wordings, such as “ants or spiders in the legs, legs want to kick, need to stretch.” All children were tired or sleepy during daytime and nearly all reported an urge to move. Insomnia was mentioned by 61% of the patients. The presenting symptoms did not differ significantly from those mentioned in a previous study in the US, but did so when compared to adults with RLS. PSG revealed an abnormal periodic limb movement index in 81% of the children.
The description of presenting symptoms in children with the final diagnosis of RLS differs from that in adults. The results of this European study corroborate those from the US.
de Weerd A; Aricò I; Silvestri R. Presenting symptoms in pediatric restless legs syndrome patients. J Clin Sleep Med 2013;9(10):1077-1080.
The diagnostic criteria for restless legs syndrome (RLS) in adults are well known.1,2 As an extension to the adult criteria, the criteria for pediatric RLS were coined in 2003 and published in 2005, see Table 1.2,3 Theoretically, the criteria for RLS allow for a final diagnosis of RLS that is based on the history of the patient and an examination to exclude so called mimics. In many cases, particularly in children, the patient history and/or examination do not provide clear data for the four essential questions or for the exclusion of mimics, which often leads to additional tests, such as polysomnography (PSG) and—in case a peripheral nervous disorder is suspected—EMG and neurography.
Diagnostic criteria for restless legs syndrome in pediatric patients
Diagnostic criteria for restless legs syndrome in pediatric patients
The primary aim of this study was to describe the presenting symptoms (i.e., what the patients and their parents told during the first visit to the outpatient clinic) in a group of children who were ultimately diagnosed as having definite RLS. A 2008 study from the US examined the presenting symptoms of 18 pediatric RLS cases,4 and the present European results will be compared to this earlier work in an attempt to generate a consensus on the presenting symptoms of RLS patients.
Current Knowledge/Study Rationale: In adults the history told by the patient is pivotal in the diagnosis of RLS. Obviously, this is more difficult in (young) children.
Study Impact: This European study describes the complaints mentioned by the children themselves in their work-up to a diagnosis of RLS. The wording used by these children is different from that of adults, but is for many aspects similar to that described for children in the US.
PATIENTS AND METHODS
The present study was conducted on patients from two large sleep centers (i.e., Zwolle and Messina) between 2007 and 2010. Both of the sleep centers had a special interest in pediatric sleep disorders, particularly RLS. A pilot study was presented in 2008.5 Before the start of the study, the senior investigators, who both have longtime experience in child neurology and sleep, agreed on the data that needed to be collected: A detailed history based on a standardized questionnaire partially derived from the pediatric RLS guidelines of 2005, (demographic data, comorbidity, non-RLS medication, final diagnosis [RLS, RLS+ periodic limb movements in sleep, PLMS > 5 h], age at start symptoms, RLS in the family, presenting symptoms, abnormalities at physical examination, therapy [which medication, lifestyle changes], effect of therapy), to be filled out by the authors, if applicable in cooperation with the referring physician. A physical examination and tests (hemoglobin, hematocrit, creatinine, thyroid stimulating hormone, and ferritin; EMG and nerve conduction studies, imaging) tailored to each patient to exclude mimics. If behavioral or psychological disorders were suspected, the child and his or her parents were seen by a psychologist or psychiatrist. A PSG in the sleep clinic was recorded and assessed according to the AASM rules6 and our own normal values7 for the children seen between 2008 and 2010 and the Rechtschaffen and Kales guidelines for children seen in 2007. The PSG helped in excluding other sleep disorders and—if showing periodic limb movements (PLMS)—were evidence in favor of RLS
Based on the results, the patient was diagnosed as either definitely having RLS or not having RLS. Finally, only 2 patients had to be excluded because of significant apneas during PSG. This resulted in a group of 31 patients to be included in the present study (25 boys) from Italy and The Netherlands. The median age of the patients at the time of the first visit to a sleep center was 10 years (range 5-12 years). The mean interval between the initial visit and the first manifestations of RLS was estimated to be 3 years (range 1-5 years). In all of the patients, the initial symptoms were still present at the first visit to the outpatient clinic.
As the work-up of all patients was similar to the normal clinical routine, the ethics committee had no objections to the study design.
One of the major problems of a history-based study with children is the communication skills of the children in the study. The ability to effectively communicate was a bigger factor than age in determining the participants of the present study. Indeed, even two 5-year-old children were able to participate in the present study because they could effectively communicate. Teenagers older than 14 years at the first visit and children suspected of secondary RLS were excluded from the study. Although parental descriptions were taken into account, the data provided by the children was deemed more important.
Descriptive statistics (the mean and standard deviation or the median, range and interquartile range [IQR]) were used to present the data, and comparisons with previous data were performed using Student t test or the Mann-Whitney U test. The significance level for these comparisons was set at p < 0.05. Tests were chosen based on the parametric or nonparametric characteristics of the data.
An urge to move and sensory symptoms were mentioned by all 31 patients. The sensory symptoms were described as ants in the legs, itchy, hurts, deep ache, spiders in or on the legs, just need to move, funny, too much energy, legs want to kick, need to stretch, or were just called growing pains (Table 2). These descriptions were given by the patients themselves and confirmed by the parents. For 16 patients, the description was compatible with the adult definite RLS criteria, but in the other 15 the description did not point directly to RLS, for example ants or spiders in the legs, legs want to kick, funny, and growing pains. Furthermore, the patients complained about insomnia, tiredness, or even excessive sleepiness during the day (EDS). Table 2 gives the prevalence of these features. All children had their symptoms when at rest. They were worse during the evening and night and improved with movement. Thus, already at the first visit, 16 of the 31 patients (53%) presented with a history compatible with the diagnosis of definite RLS. The other 15 patients, who described the symptoms unconventionally, had insomnia (N = 14), a parent with RLS (N = 3), PLMS > 5/h of sleep (all 15). All 15 patients had ≥ 2 of these features and met current criteria for childhood RLS as well (Table 1). Figure 1 depicts the differences between the presenting symptoms in our study and those from a previous study by Picchietti's group. Although our patients had a history of less insomnia and EDS, they had more complaints of being tired, which was not mentioned in the study of Picchietti et al.
Comparison of our data with the data from the Picchietti et al.3
Comparison of our data with the data from the Picchietti et al.3
Twenty-seven of the 31 patients suffered from a second disorder which was diagnosed before entering the study in 17 patients and during the work-up after the first interview in the other 10 patients. Remarkable differences between the present study and the Picchietti study were observed for epilepsy (19% vs. 0%, respectively), ADHD (42% vs. 72%, respectively), parasomnia (7% vs. 39%, respectively), anxiety (10% vs. 33%, respectively), and depression (0% vs. 28%, respectively). Comorbid ADHD was always the hyperactive or combined type. Furthermore, the genetic component appeared to be less prevalent in our study (19% vs.72%).
The patients in the present study showed abnormal leg movements during sleep and abnormal scores on the PLM index (PLMI). In the present study, the PLMI (median 12; range: 7-21, IQR 7-19), was abnormal in 81% of the patients (N = 25). In the Picchietti study, these figures were respectively 65%, 13, and IQR: 5.5-13.
The present study provided data on the presenting symptoms of pediatric RLS patients from two different regions of Europe. Independent of the location, a variety of sensorimotor symptoms were described by the children. Complaints of insomnia, tiredness during the daytime and some overt EDS were primarily mentioned by the parents. The study group may be compared with the group of 18 patients described by Picchietti et al.3 The subjective measures of sleep and daytime performance suggested that there was less insomnia and EDS in the patients in the present study; however, the children in our group were more tired during the day. Because an MSLT may not give reliable information in children, we had to rely on answers obtained from the parents. Indeed, the sensation of being tired during the day and EDS may be difficult for children to differentiate.
Disorders other than RLS or PLMS occurred in 27 of the 31 patients—epilepsy, ADHD, and anxiety were the most common disorders. The relatively high prevalence of epilepsy observed in the present study may be due to the character of the participating centers, which specialize in both sleep and epilepsy. One can only speculate whether this influences the results of the study. As there is up to now no known interaction between epilepsy and RLS, this is probably of little importance. Cognitive deficits that are sensitive to sleep loss have been ascertained in adults with RLS.8 Although it is impossible to run the same tests in children, it is interesting to hypothesize that the daytime hyperactivity or ADHD that was observed in many of our children could be akin to the cognitive deficits observed in adults. In addition, the children in the present study were less likely to fall asleep. Interestingly, they showed hyperactivity features akin to the patients in the group described by Gamaldo et al.9 In particular, the Messina Sleep Center found that the percentage of children diagnosed with RLS and ADHD after the total work-up was nearly double the number of children diagnosed after the interviews with the children and their parents.10
The present study was originally started as part of a survey of many aspects of pediatric RLS in Europe. Most of our colleagues in child neurology and/or sleep who were approached to join the study told us that they only sporadically see children suspected of RLS (if at all). Despite many oral and written contacts with major pediatric sleep centers in the UK, France, Germany, Middle Europe, Italy, and The Netherlands, only a limited number of patients could be included in the present study. Although the present study was not designed to estimate the RLS prevalence in children, the difficulty in recruiting patients casts doubts on the frequent occurrence of pediatric RLS (estimated at 1% to 2%) mentioned in a previous publication.3 On the other hand, we suspect that in Europe, many children who in fact have RLS are diagnosed as having ADHD.
A limitation of the present study was in the long interval between the first manifestations of RLS and the visit to the outpatient clinics, which implies that the descriptions of the presenting symptoms were partly based on retrospective data. This potential bias is inevitable, but it may be mitigated by the persistence of the symptoms during the period between the first manifestations of RLS and the initial visit to the outpatient clinic (all of the parents mentioned that the symptoms persisted during this interval). We believe that the strong point of this study consisted in the consequent and comprehensive work-up for each patient, which led to a homogeneous group of children.
A recent paper from Arbuckle et al.11 explored the feasibility of a multidimensional, self-administered Pediatric RLS Severity Scale (P-RLS-SS) to measure the outcome and the impact of pediatric RLS symptoms. Symptoms were subdivided into four major symptom domains (RLS sensations, move/rub legs due to RLS, relief of RLS from moving/rubbing legs, RLS hurt/pain) and four impact domains (RLS impact on sleep, RLS impact on awake activities, RLS impact on emotions, RLS impact on tiredness). The scale would provide scores for the symptom and impact domains and a total RLS score. Interestingly, data from the 33 children and adolescents with definite RLS included in the Arbuckle study confirmed that there are no gender- or ADHD-comorbidity-based differences in RLS symptoms or impact. Age, however, played a major role in the different terms used for the descriptions. For example, “kick” and “wiggly” were used more often by children younger than 11, whereas “tingling” was used more often by adolescents. Subjective impairment of sleep was rated by 88% of the patients across all age groups, and impact on tiredness was mentioned by 70%. The impact of RLS on emotions and awake activities was less consistent despite parents' reports on these measures. Although the P-RLS-SS requires further validation, the scale certainly represents an innovation in the RLS field and may be potentially used just like the validated scale for adults (International RLS Score, IRLS). Furthermore, the P-RLS-SS constitutes a constructive assessment instrument for future pediatric RLS trials, which will assist in the collection of additional data on pediatric RLS features and symptoms.
Recently, a proposal for more detailed criteria for pediatric RLS has been approved by the International RLS Study Group (IRLSSG). One of us (AW) was part of the committee that prepared the proposal, which is not yet formally published. However, all children in the present study would have met these criteria as well.
The most important finding of the present study was the similarity of the presenting symptoms of RLS among our patients. Although there were variations in the descriptions of symptoms, abnormal sensations in the legs and an urge to move were the major presenting symptoms in our study. The descriptions of abnormal sensations and urges to move were inherently supportive for the diagnosis of RLS at the first visit in 16 of the 31 cases in our study. In the other half of the patients, the history had to be verified through additional examinations and/ or the course of the disorder over time. The results of the present study were in broad lines similar to studies by Picchietti et al.3,4
In our opinion, the findings in Picchietti's and our studies allow for the statement that many cases of pediatric RLS can be detected by collecting a detailed history from the parents and carefully listening to the description of the symptoms offered by the children. Similar to adults, sensorimotor symptoms prevail in children and are the cornerstone of the diagnosis. However, more anamnestic pitfalls occur in children. When doubt remains after a detailed history, the clinical course and results of additional tests (e.g., exclusion of mimics, PSG) should be taken into account before the final diagnosis of pediatric RLS is made.
This was not an industry supported study. The authors have indicated no financial conflicts of interest.