Background

Dystonia is a neurological condition that can cause abnormal and repetitive muscle contractions and spasms [1]. It is the third most commonly diagnosed movement disorder after Parkinson’s disease and essential tremor [2] with a prevalence variously estimated to be in range of 30–7320 per million [2]. The wide variations on prevalence estimates reflect the different approaches and definitions used by different authors.

Idiopathic adult onset dystonias affecting one body part such as the hand (writer’s cramp) or eyes (blepharospasm) are known as focal dystonia and usually affect adults [1]. In contrast, generalised dystonia typically occurs during childhood and causes spasms and cramping in two or more affected body regions [1]. There may be genetic associations with some idiopathic adult onset dystonias; however, in a majority of cases the cause of disease is unknown [1].

Treatments for generalised and focal dystonia, including pharmaceutical or surgical options, focus predominately on the motor symptoms of the condition. Conservative therapies that have been suggested for helping people with idiopathic adult onset dystonia include: acupuncture, biofeedback, chiropractic manipulation, osteopathy, physiotherapy, speech therapy and transcutaneous electrical nerve stimulation (TENS) [3–6].

However People living with idiopathic adult onset dystonia can experience low self-efficacy, social withdrawal, a negative self-concept and subsequently, poor health outcomes and quality of life [7–10].

More specifically non-motor symptoms of dystonia can include impact on mood (low mood including depression) and emotions, cognitive functioning (such as memory) and impaired sleep [11] Similarly as with other conditions such as Parkinson’s disease, the non-motor symptoms dominate clinical impact and can lead to severe disability and an impaired quality of life [12].

A systematic review investigating the relationship between mental health and ‘primary and other genetic forms of dystonia’ found a relationship between dystonia and emotional disturbance, dystonia and psychiatric disorders and dystonia and cognitive difficulties [11].

People with cervical dystonia for example have a 91.4 % chance of meeting the criteria for a psychiatric illness compared to 35 % in the general population [13]. Other psychiatric illnesses include; major depression, social phobia, panic disorder and for these with focal dystonia a higher risk of obsessive compulsive disorder and anxiety disorder [14]. There is some evidence that cognitive behavioural therapy is effective in the use of the management of anxious thoughts and mood related to muscle tension which could lead to better patient outcomes, however the evidence is limited [15].

We report a systematic review of the evidence for all behavioural interventions for adults living with idiopathic adult onset dystonia.

Methods

The focus of this review is adults living with primary dystonia (idiopathic adult onset dystonia). We excluded studies of people with acquired dystonia to ensure we were reporting dystonia specific effects rather than the combined symptoms from any acquired disorder (e.g., psychiatric, infectious, toxic or neoplastic) [1].

When we developed the protocol for this review we used the terms then current and set our parameter for study entry to be a primary dystonia; this was the terminology used in our PRoSEPRO registration. Since then the Movement Disorder Society‘s new classification has been published. We did not exclude any studies on the basis of clinical features; axis one of the Movement Disorder Society (MDS) classification. In axis two of the MDS classification our entry criteria map onto the idiopathic domain which includes both sporadic and familial cases. We excluded studies of people with known nervous system pathology, known degenerate causes, and those with acquired causes from axis 2 of the MDS classification.

We have post-hoc classified the included studies according to the new MDS criteria. We have, however, kept our original terminology elsewhere in the paper.

Nearly all of the studies we identified had substantial methodological limitations including poor study designs, small samples, and inappropriate statistical tests. Therefore, we were unable to conduct a rigorous meta-analysis.

Search strategy

Between December 2014 and January 2015 we searched: Medline, PsycINFO, AHMED and CINAHL. No date restrictions were applied. Primary dystonia is a broad category encapsulating several different types of medical conditions but studies on dystonia are inconsistently coded in publication databases. For this review, we constructed a broad (sensitive) search strategy to identify relevant studies (please refer to Fig. 1).

Fig. 1
figure 1

List of review search terms

Full size image

Selection criteria

Relevant outcome measures included in this review are: wellbeing and/or quality of life and/or psychological functioning and/or sleep and/or other clinical measures such as dystonia severity and/or pain. We included studies of adults (≥18) reporting on individuals living with primary dystonia that examined the effect of behavioural interventions on relevant outcomes. Behavioural interventions are defined as interventions where patient behaviour is the key and the main aim to change that behaviour and can be applied at an individual level, community level or international level [16] (and can include highlighting thoughts processes, beliefs and appraisals and learning to challenge and modify these to improve and change behaviour [15].

We included studies published in English, French, German, Italian and Punjabi. Studies were excluded if they conducted research with individuals under the age of 18 years, animals and secondary types of dystonia or dystonia with another primary condition. Studies of non-behavioural - interventions e.g., medical treatments and psychodynamic therapy were excluded.

We included all primary research reporting quantitative outcome; irrespective of study design. Non-primary articles including secondary analyses, review papers, guidelines, statements, meeting summaries or comments were excluded.

Two independent reviewers (DE and CB) screened the studies’ titles and abstracts for eligibility. For potentially eligible studies’, full-texts were then screened by the rest of the team with the exception of MU who acted as an independent adjudicator, making a final decision on any disputed texts.

Assessing methodological quality

Assessments on all identified studies were performed using a risk of bias rating table that was adapted from a recent systematic review on sleep in primary dystonia [17]. The criteria of risk were based on: diagnosis process, sample size, use of a control group, relevant and valid outcome measures, sufficient description of the intervention, medication and statistics. Risk of bias for each study was rated according to a ‘low’, ‘moderate’, ‘high’ or ‘unclear’ score.

Data extraction

Socio-demographic and clinical details of study participants were extracted by CB and reviewed by the rest of the team for accuracy. Information about the studies’ research question(s), setting(s), behavioural intervention(s), method(s) and outcome measure(s) were also recorded using a pro-forma. The results of the studies were extracted to evaluate psychological (e.g., quality of life, well-being, sleep) and clinical (e.g., pain, dystonia severity, functioning) outcome measures in relation to the reported behavioural intervention(s). A narrative synthesis of the results is reported.

Results

The searches retrieved from the databases yielded 2133 hits. Two studies from one included article’s [18] reference list were identified and the abstracts reviewed; one was potentially eligible but the authors were unable to locate the full-text article and so it has been excluded [19]. We examined 14 full papers and included nine: (Fig. 2 [20]).

Fig. 2
figure 2

Flow diagram of screening and identification process [20]

Full size image

We found limited evidence of the following behavioural interventions for different idiopathic adult-onset focal dystonia types: Solution-Focused Guided Imagery (SFGI) for the yips [18, 21, 22], exercise and relaxation practice [23] and Cognitive Behavioural Therapy (CBT) [24] for cervical dystonia, relaxation, systematic desensitisation (SD), assertiveness training and/or Electromyography (EMG) feedback [25], habit reversal with in vivo exposure, awareness training and re-education [26] and relaxation or habit reversal therapy [27] for writer’s cramp and conditioning therapy and relaxation practice for blepharospasm [28]. No studies reported on generalised dystonia. Of the nine studies we identified, two were Randomised Controlled Trials (RCTs) [23, 27] utilising small sample sizes. One RCT found a significant finding for improvement in handwriting quality in both the intervention and control groups at follow-up (week 20) [27]. While all studies reported improvements in dystonia severity and/or emotional distress at weekly, monthly and yearly follow-ups, the poor methodological quality of the data means that firm conclusions cannot be drawn about the effectiveness of behavioural therapies for people living with idiopathic adult onset dystonia.

Quality assessment

Table 1 provides an overview of the findings. Please see web Appendix for the criteria by which the studies were judged (adapted from Hertenstein et al. [17]).

Table 1 Quality assessment summary
Full size table

We found two RCTs (N = 20) [23] and (N = 23) [27], two are described as using multiple baseline across subjects design (N = 4) (which is when the same treatment is staggered across participants and multiple data points are obtained from the baseline, intervention and follow-up phases) [21, 22], and two case studies (N = 1) [24, 28]. The remaining three were described as being: a case study using a multiple baseline across situations design (N = 1) [18] (which includes studying the same behaviour across different circumstances, with an aim to test the relationship between treatment and behavioural change in a single subject), an uncontrolled multimodal study (drawing on different modalities including relaxation, systematic desensitization and assertiveness training and biofeedback which was tailored to participants), (N = 15) [25], and a clinical study (reporting the clinical treatment of four cases, delivered by a trained therapist) (N = 4) [26] (Table 2).

Table 2 Included studies basic characteristics and sample
Full size table

The studies originated from Australia [17], England [24, 26, 27] and the United States of America (USA) [21, 22]; Bell and Thompson [18], Cottraux et al. [25] and Sharpe [28] do not specify the country where their studies were conducted (Table 2).

Boyce et al. [23], Greenberg [26], Sharpe [28] and Wieck et al. [27], conducted the research in hospitals. Bell and Thompson [18] and Bell et al. [21, 22] observed putting on golf courses. Cottraux et al. [25] described their research setting as a clinical study using participants who were referred to their department. Faircloth and Reid [24] did not provide definite details of where their research was conducted (Table 2).

Idiopathic adult onset focal dystonia

We found one study examining blepharospasm [28], two studies that investigated cervical dystonia [23, 24], three studies that conducted research with patients living with writer’s cramp [25–27] and three studies that tested an intervention for the yips [18, 21, 22] which is a type of dystonia that causes ‘jerks, tremors and spasms [and] predominately affects the distal upper extremity’ (p.424) [29]. We found the yips to occur in golfers [18, 21, 22]. Consequently, none of the identified studies included patients with generalised dystonia and therefore, the findings of this review can only report on the effectiveness of behavioural therapies for focal dystonia (Table 3).

Table 3 Included studies interventions
Full size table

Three studies provided a full report of participants’ ethnicity [18, 21, 22]. Greenberg [26] described the ethnicity of one out of his four participants. The mean age taken across the entire sample (N = 73) was 47 years (range 36–57.8 years) although it is unclear whether Wieck et al. [27] included the ages of three participants who withdrew (Table 2).

Of the 73 participants 27 (37 %) were female. For five studies 100 % of the sample were male [18, 21, 22, 24, 28] (Table 2). Adherence rates for all of the identified studies were at above 50 % although sample sizes for each study were very low (see Table 4). Two RCTs looking at cervical dystonia and writer’s cramp reported slightly higher levels of adherence among the control groups (91 % for Boyce et al. [23] and 92 % for Wieck et al. [27]) compared with the intervention groups (78 % for Boyce et al. [23] and 82 % for Wieck et al. [27]). Cottraux et al. [25] described an adherence rate of 60 % among people with writer’s cramp with two participants refusing treatment and four participants withdrawing during the intervention phase. Interestingly, Cottraux et al. [25] found that three participants, who had received treatment, later withdrew from the study despite reporting handwriting improvements.

Table 4 summary of outcome measures and findings for each identified study
Full size table

Three studies looking at writer’s cramp reported participants were living with various emotional disturbances [25–27]. Of these three studies, two [25, 27] examined patients’ physical and mental health. However, Cottraux et al. [25] only screened 10 patients (out of the 15 originally recruited into the study) using Multiple Personality Inventory (MMPI) referred to as MimiMult. While participants who were screened with MMPI were not reported as presenting with a consistent pathological pattern, Cottraux et al. [25] noted that seven individuals reported chronic emotional or sexual disturbances. Furthermore, all participants had previously used various treatments including tranquillizers, bromocriptine and psychotherapy [25]. Greenberg [26] reported that one participant out of four displayed general anxiety symptoms. Wieck et al. [27] described one participant in their study as anxious and displaying avoidance behaviour and two patients were reported with living with generalised anxiety.

Bell et al. [21, 22] did not specify how long their participants had been living with dystonia. Sharpe [28] was the only author who reported that the participant had been living with dystonia for under a year (10 months), stating that the participant had also experienced dystonia symptoms for four weeks 11 months earlier. The average duration of dystonia in the remaining six studies was seven years.

Interventions

All of the identified studies examined various behavioural therapies. These include SFGI protocol to individuals experiencing yips [18, 21, 22], which typically causes spasms in the upper limb(s) and can occur in various sports including golf [29], combined behavioural and physical practice including relaxation for patients with cervical dystonia [23], SD, assertiveness training and/or EMG feedback on patients with writer’s cramp [25] and CBT for cervical dystonia [24]. Greenberg [26] investigated the effectiveness of habit reversal therapy (which aims to reduce spasm and involuntary movements by delivering fine motor techniques that are opposite to the involuntary contraction) with awareness training, re-education, and ‘in vivo exposure in which a hierarchy of situations was constructed in which cramp was increasingly likely to occur.’(p.234) Sharpe [28] examined the effects of relaxation with conditioning therapy for one patient with blepharospasm. Finally, Wieck et al. [27] investigated writer’s cramp and assigned the control group individually tailored sessions of relaxation and the intervention group, habit reversal therapy. While the studies’ interventions were fairly diverse, they all focused on examining the effectiveness of enabling participants to self-manage their disability outside of the clinical setting [18, 21–28].

Outcome measures

Eight studies reported outcome measures utilising independently assessed and/or facilitator-led ratings and/or self-report un-validated and/or validated measures to evaluate physiological and/or psychological improvement [18, 21–27].

Two studies used the Beck Depression Inventory to assess depression scores [23, 24]. Faircloth and Reid [24] also used un-validated self-report measures to evaluate psychological wellbeing and an un-validated measure to investigate physical functioning. The authors did not objectively measure any clinical outcomes [24].

The studies examining writer’s cramp observed writing frequency among participants in and outside of the clinic [25–27]. However, Greenberg [26] was the only author examining writer’s cramp to not use an independent assessor to verify handwriting improvements. Instead, only patients and therapists involved in treatment delivery rated handwriting quality [26]. Cottraux et al. [25] utilised an independent evaluator to rate spasm frequency, handwriting quality and handwriting improvement on three separate scales that were combined to obtain the overall mean value of each participant’s handwriting improvement score. Since these data were only obtained in the clinic participants were also asked to self-evaluate their handwriting in a non-clinical setting [25]. However, data obtained from participants outside of the clinic were not systematically scored [25].

Having applied the SFGI protocol, Bell and Thompson [18] requested that the participant and a playing partner record yip behaviour during putting. Similarly, Bell et al. [21] used trained observers to watch putting behaviour among the participants. Video observations of putting behaviour were used by Bell et al. [22] to increase inter-reliability. Bell and Thompson [18] and Bell et al. [21, 22] also used the SFGI protocol to ask participants questions about the severity of their dystonia.

Sharpe [28] did not report using any outcome measures. The authors of this review were unable to contact Sharpe [28] given how long ago the study was published. Nonetheless, Sharpe [28] provides details about outcomes post-treatment and at the 9-month follow-up.

Discussion

We found very limited evidence for behavioural interventions in the treatment and management of idiopathic adult onset dystonia. All of the studies we identified were for the MDS idiopathic adult onset focal dystonias group. The findings therefore only apply to this group. There were some promising results from the RCTs and uncontrolled studies that behavioural therapies can improve peoples’ physical and emotional well-being. Some of the underlying mechanisms mapped onto the behavioural interventions specifically included action planning and problem solving to improve productivity and overall management, education and behavioural experiments to empower and help control symptoms and education and distraction to manage anxiety through cognitive restricting to lessen anxiety and also the severity of dystonia. However all of the identified studies used small samples and nearly all had substantial methodological weaknesses which precluded a rigorous meta-analysis. Therefore, none of the findings can be generalised beyond the study setting. The available evidence applies only to focal dystonias and we cannot make any comment based on possible effectiveness from empirical data for dystonias affecting more than one body segment. Nevertheless, the theoretical justification for behavioural interventions is equally applicable to more than just focal dystonias. This means that the very limited empirical data might be applicable to a wider population of those affected by dystonia. A possible reason for why there have been very few research studies examining the effectiveness of behavioural therapies for people with dystonia could be because idiopathic adult onset dystonia is often misdiagnosed as an emotional illness limiting acceptance of behavioural interventions [9].

There have however been other studies which have also shown that CBT could be beneficial with other movement disorders. For example in a pilot study, tailored CBT for this with depression and Parkinson disease showed a reduction in negative cognitions, reduced depressive symptoms and an increase in the sense of social support [30].

Strengths and limitations of the review

The authors of this review employed a rigorous search and screening strategy to identify and evaluate the relevant studies. It has highlighted an important gap in the literature and is one of the first reviews too examine behavioural interventions for the management of dystonia.

Since Sharpe’s [28] study was conducted some time ago the review authors were unable to obtain information about what outcome measures he utilised. Furthermore, the review found no studies examining the effects of behavioural therapies in people with generalised dystonia. Consequently, this review is limited in not being able to comment whether behavioural treatments would be effective for this patient group. Conducting a meta-analysis was not possible because of study heterogeneity.

Future research and practice

Future research is necessary to establish the effects, if any, of behavioural interventions in patients living with idiopathic adult onset dystonia. Consequently, researchers should consider RCTs with powered sample size, clearly specifying the diagnostic process involved and using physiological and psychological validated outcome measures to capture changes. Detailing the severity of patients’ dystonia could also be useful for comparing treatment effectiveness across different levels of dystonia.

Future research should also examine generalised dystonia in relation to behavioural therapies to address the current gap in the medical and psychology literature.

It is only by addressing the many challenges in designing robust studies of behavioural interventions for the dystonias that we can find out if behavioural treatments can help these stigmatising and disabling chronic disorders.

Conclusion

In conclusion, this review identified nine studies all with low methodological quality, making it difficult to establish the effectiveness of behavioural interventions in patients living with idiopathic adult onset dystonia. The limited data indicates, however, that behavioural therapies could have a beneficial effect for enabling individuals to manage their dystonia, particularly when combined with relaxation practice [23, 25, 27, 28]. Further high-quality research is needed to fully assess the effectiveness of holistic behavioural therapies in patients with idiopathic adult onset dystonia.