Characterization of painful Restless Legs Syndrome sensations in an English-speaking South African population

1 Introduction

Restless Legs Syndrome (RLS) is a condition characterised by an urge to move in response to unusual sensations normally experienced in the legs [1]. Patients express difficulty in describing the unusual sensations of RLS, however the diagnosis of RLS is primarily based on the subjective descriptions of sensations. Currently, the diagnostic criteria indicates that the urge to move is in response to “uncomfortable” and “unpleasant” sensations [1] however this may not take the presence of pain into consideration.

Between 21.4% and 61% of RLS patients have described their symptoms as painful [2–5]. The alleviation of RLS symptoms with analgesic medications, amongst other treatments, indicates that pain pathways may be involved in the sensations associated with RLS [6]. RLS patients have also been shown to have amplified nociceptive processing and increased pain sensitivity [7], [8]. The McGill Pain Questionnaire (MPQ) has been used to quantify and qualify RLS sensations [9], [10]. Previous research characterising RLS symptoms has shown that patients with painful RLS symptoms had greater sleep disturbances and more severe RLS symptoms and the distribution of terms (self-selected and from the MPQ) used to describe their sensations were more intense and pain-related [2], [5].

RLS is diagnosed based on the patient’s subjective, self-reported symptoms however the choice of descriptors used by patients with RLS may be affected by pain, different languages and cultural influences which in turn may also affect and confound diagnosis and prevalence estimations. The objective of this study was to characterise the descriptors used for painful RLS to determine if the presence of pain changed the preferred description of RLS sensations in an English speaking South African sample of patients with RLS.

2 Methods

3 Results

4 Discussion

In this group of English speaking South African RLS patients the presence of painful RLS symptoms were associated with a greater RLS severity however no other relationships with any other RLS variables/clinical features were noted. Although there were limited differences in clinical characteristics between the two groups there were distinct differences in the symptom descriptors chosen by patients with painful and non-painful RLS symptoms which may confound diagnostic accuracy.

Painful RLS symptoms were reported by 35% of the patients, which was confirmed by a higher PRI score in this group of patients. The prevalence of painful RLS symptoms reported in previous studies ranges from 21.4% to 61% of RLS patients [2–5]. This South African cohort had a slightly higher prevalence than the figures reported for cohorts in South Korea (23.8%) [2] and North America/Europe (21.4%) [3] however was lower than another North American/European study (59.4%) [4] and a French study (55–61%) [5]. The differences in pain prevalence could be attributed to differences in determinants of painful symptoms between studies. Some studies relied on patients to classify whether their symptoms were painful or not [5], others simply reported whether pain was present [3], [4], and some classified patients based on the descriptive terms that they utilised [2].

Predictably, the most frequently chosen words in both groups were “restless” and “uncomfortable”. These common terms may imply that the presence of pain may not influence the diagnosis of RLS based on a predetermined list of words. However, participants with painful RLS were also more likely to choose pain-related terms such as “cramping” and “painful” compared to the participants with non-painful RLS. The choice of these pain-related terms may result in patients with painful RLS being misdiagnosed with a primary pain condition. Despite dividing the participants according to the presence or absence of pain, the most frequently selected word from the literature terms in the painful RLS group was not in fact “painful”. Winkelman et al. [15] suggested that the description of painful RLS sensations by patients may be confounded by the manner in which their symptoms are interrogated [15].

Restricting patients to selecting words within subclasses such as on the MPQ may help differentiate characteristics of painful RLS symptoms. Results from the MPQ demonstrated that patients with painful RLS symptoms select significantly more words from both sensory and affective subclasses. Additionally, the word distribution was biased towards more intense descriptors within subclasses for patients with painful RLS symptoms with “cramping” and “unbearable” characterising the painful RLS symptoms. Interestingly, Cho et al. [2] classified patients who used the word “cramping” into their non-painful group [2]. Previous research by Karroum et al. [5], using the MPQ to describe RLS sensations, identified “burning” as the most frequent descriptor of painful RLS in a French speaking population. The patients with painful RLS in the current study did not select “burning” from the MPQ more frequently to describe their sensation however did select “burning” more often from the literature terms. Whether the discrepancy within our study, and between the two studies, is a reflection of the limitation of selecting a single word from each subclass, is a language difference or indicates underlying phenotypical differences, requires further investigation. Importantly, the MPQ used in the French study was a validated, modified version of the MPQ, the Questionnaire Douleur de Saint-Antoine (QDSA) [16], and does not contain all the words of the MPQ [14]. The term “burning” is however, common to both whereas “cramping” only appears on the original MPQ.

The current study demonstrated clear differences in the choice of words between RLS patients with painful and non-painful sensations. The words “cramping” and “painful” were favoured by the patients with painful RLS. Although the authors were cautious to rule out RLS mimics, the high prevalence of the choice of the descriptor “cramping” may represent a false positive. In a clinical setting, a patient complaining of cramping and painful sensations in their legs may be diagnosed with a condition that mimics RLS such as leg cramps or peripheral neuropathy, resulting in a marked under-diagnosis of RLS [3]. Coupling these words (painful and cramping) with the other commonly selected word “restless”, may assist in clearly defining the RLS diagnosis. These results are consistent with Ekbom’s original proposal that there may be two main forms of RLS, one characterised by painful symptoms and the other by non-painful paraesthesia [17]. Different descriptors may indicate a difference in aetiology as occurs in postoperative pain [18]. This may imply that the presence or absence of pain indicates two phenotypes of RLS. The potential impact of this finding on the aetiology, treatment, course and measurement of RLS requires further study.

Patients with painful RLS symptoms in the current study reported greater RLS severity as compared to patients with non-painful RLS. The presence of pain, therefore appears to play a role in the severity of RLS symptoms despite the IRLS severity scale not asking specific pain-related questions and measuring the intensity of the actual sensations. Previous research has shown a moderate correlation between the IRLS severity score and the MPQ pain rating index as well as the number of words chosen from the MPQ [5], [10]. Additionally, Karroum et al. [5] showed that more patients with painful RLS were on current dopaminergic treatment in their larger patient group however in their clinical series, neither current dopaminergic treatment nor opioid treatment was different between painful and non-painful RLS patients [5]. There is a marked under treatment of RLS in the South Africa population and the same comparison of dopaminergic and opioid therapy related to painful symptoms could not reliably be made in the current study. Whether pain presents as part of the dopaminergic aggravated augmentation of RLS warrants further investigation.

The participants in this study represent a small sample of English speaking South Africans with RLS and the findings may differ in other cultures and language groups. The ethnicity breakdown is not a true reflection of the South African population demographics thus the requirement of proficiency in English may have biased the study to a predominantly Caucasian sample. Given that South Africa has 11 official languages, it was necessary for this preliminary study to be restricted to a single language especially for international comparisons. Further research investigating different languages may yield different results. The majority of our participants suffered from mild to moderate RLS while patients with very severe RLS may choose different words. There may be benefit in looking at other clinical features in future studies.

In conclusion, the word choice in patients with RLS appears to vary according to the presence or absence of painful RLS sensations which may support the idea of RLS as a disorder with multiple phenotypes. Future research should take cognisance of patients presenting with painful RLS as this may be an important confounding factor. More research is needed to tease out possible relationships between RLS and pain.