Article Text


Research paper
Risk factors for spinal cord lesions in dystonic cerebral palsy and generalised dystonia
  1. Emilie Guettard1,2,
  2. Damien Ricard3,
  3. Emmanuel Roze1,2,
  4. Alexis Elbaz4,
  5. Mathieu Anheim1,2,5,
  6. Stéphane Thobois6,
  7. Jean-Francois Lepeintre7,
  8. Damien Galanaud8,
  9. Christian Mazel9,
  10. Marie Vidailhet1
  1. 1AP-HP, Department of Neurology, Groupe Hospitalier Pitié-Salpêtrière, Paris, France
  2. 2Centre de Recherche de l'Institut du Cerveau et de la Moelle épinière (CRICM), CNRS UMR 7225, UPMC Université Paris 6/Inserm UMR_S 975, AP-HP, Paris, France
  3. 3Service de Neurologie, Hôpital d'Instruction des Armées du Val-de-Grâce, Paris, France
  4. 4INSERM, U708, Neuroepidemiology, Paris, France
  5. 5Department of Medical Genetics, Groupe Hospitalier Pitié-Salpêtrière, Paris, France
  6. 6Université Lyon I, Hospices Civils de Lyon, Hôpital Neurologique Pierre Wertheimer, Service de Neurologie C, Lyon, France
  7. 7Neurosurgery Department, Hôpital FOCH, Suresnes Cedex, France
  8. 8Department of Neuroradiology, Groupe Hospitalier Pitié-Salpêtrière, Paris, France
  9. 9Université Paris XIII, Institut Mutualiste Montsouris, Département de Chirurgie Orthopédique, 42 Boulevard Jourdan, Paris, France
  1. Correspondence to Professor Marie Vidailhet, Department of Neurology, CRICM UPMC/Inserm UMR_S 975, CNRS UMR 7225, Pitié-Salpêtrière Hospital, Paris, France; marie.vidailhet{at}


Background Cervical myelopathy (CM) in patients with cerebral palsy (CP) is underdiagnosed as symptoms of spinal cord lesions, being similar to those due to dystonia, may be overlooked or identified late. The aim of this study is to identify the risk factors and clinical characteristics of CM in patients with generalised dystonia, including dystonic CP.

Methods The authors conducted a case–control study to identify early clinical signs of CM in consecutive patients with generalised dystonia. The authors compared the clinical characteristics and symptoms of those who developed CM (cases) and those who did not (controls). The same clinical information on possible neurological manifestations of CM was collected for cases and controls at the date of the last visit.

Results Out of 54 patients, 17 (31%) developed symptomatic CM during the study period. In all cases, CM occurred after the age of 36 years. 81% of cases and 35% of controls had a Burke–Fahn–Marsden movement subscore for the neck >4. Age (OR per 10 years=2.3, 95% CI 1.4 to 4.2, p=0.006) and severity of neck dystonia (OR=7.7, 95% CI 1.7 to 49.6, p=0.005) were the main risk factors of CM. Gait disorders and falls, wasting of hand muscles and bladder disorders were the best clinical clues of CM.

Conclusions As severity of cervical dystonia and age are the major risk factors for spinal cord lesions, dystonic patients, including patients with dystonic CP, should be screened for CM from the third decade of life onwards. Early recognition of CM is crucial for functional prognosis and impact on autonomy.

  • Cervical myelopathy
  • cerebral palsy
  • cervical dystonia
  • generalised dystonia
  • spinal cord lesion
  • neurooncology, dystonia
  • chorea
  • paediatric movement disorders
  • Gilles de la Tourette
  • Tourette syndrome
  • cerebellar ataxia
  • cerebellar degeneration
  • heredit spastic paraplegia
  • Parkinson's disease
  • cerebellar disease
  • clinical neurology
  • Parkinson
  • functional imaging
  • movement disorders
  • pet
  • motor control
  • *new 1998

Statistics from


Generalised dystonia is a severe motor disorder characterised by sustained muscle contractions causing twisting and repetitive movements or abnormal postures. Cervical myelopathy (CM) is a complication of generalised dystonia that has been reported in various causes of the disease, including the most frequent one, dystonic cerebral palsy (CP).1–6

Early detection of spinal cord lesions due to CM is crucial for the functional prognosis of these patients. However, the symptoms of CM are similar to those of dystonia and, therefore, CM may be overlooked or diagnosed too late,6–8 leading to an irreversible dependent state with tetraparesis.5 As dystonia is rare, it may appear as an unfrequent disorder. However, this becomes a larger health problem when one considers that this also concerns dystonic CP which is a frequent condition (2–3 per 1000 live births).1–6 Overall, CM appears to be underdiagnosed as demonstrated by the relative scarcity and heterogeneity of the literature.

To date, case reports and small groups of patients with primary4 5 8–13 or secondary dystonias (mainly CP)1 2 6 14–21 have been reported, with various types of neuro-orthopaedic complications,6 11 18 multiple diagnostic approaches4 18 and different assessments of clinical outcome.9 22 Overall, the data are too heterogeneous to draw conclusions and frame guidelines.

Identification of the most reliable clinical factors that confirm the occurrence of spinal cord lesions is important from a public health perspective in order to increase the awareness of the medical community, help avoid irreversible damage5 and preserve the remaining autonomy. Early identification of CM may lead to proposing a timely treatment; in the era of deep brain stimulation for dystonia, improvement of neck movement disorders can be expected with bilateral pallidal stimulation mainly in primary dystonia,23 24 whereas the benefit is smaller and less predictable in dystonic CP.25

Among patients with generalised dystonia followed in our department between 1997 and 2007, we compared the clinical characteristics of those who developed CM (cases) and those who did not (controls). The aim of this case–control study was to identify factors associated with CM in order to allow an early diagnosis of this complication.

Patients and methods

Among 485 consecutive patients with various forms of dystonia managed in our movement disorders clinic between 1997 and 2007, we included those with generalised dystonia and neck involvement. Additional inclusion criteria were: (1) no history of traumatic brain injury or neurometabolic disorders, (2) no severe psychiatric disorders and little or no cognitive impairment, (3) little or no spasticity and no motor disturbances other than dystonia at the first visit, (4) normal brain MRI or slight abnormalities on T1-weighted MRI (eg, decreased grey–white matter contrast with partial disappearance of the basal ganglia and minimal atrophy of the pallidum or putamen) and (5) availability of videos of the patients taken during routine visits in stable state: all patients were regularly followed and examined; video recordings were taken to characterise the symptoms of generalised dystonia and to assess the severity of cervical and trunk involvement; in those who subsequently developed CM, at least one video had been taken prior to CM onset. The aetiology and age at onset of dystonia were available for all participants.


Cases were patients who developed signs of CM during the study period. CM was defined on the basis of clinical findings and cervical spine MRI. Clinical signs of CM included recent development or worsening (in CP) of upper or lower limb weakness, wasting of the hand musculature, sensory abnormalities, hyperactive deep tendon reflexes, recent emergence or worsening (in CP) of spasticity and Lhermitte's sign. MRI signs of myelopathy included spinal cord compression, intramedullary T2 signal hyperintensity and spinal cord oedema (figure 1). MRIs were performed either under sedation or general anaesthesia. Age at onset of CM was recorded.

Figure 1

Spinal cord lesions in patients with cervical myelopathy Sagittal (A1, B1, C1) and axial (A2, B2, C2) T2-weighted MR images of the cervical spine in three patients with severe neck dystonia (maximal BFM movement subscore for the neck=8). (A) 25-Year-old woman, BFM neck subscore=6. Hypersignal of the posterior column at level C3-C5 in the absence of discal protrusion or cervical canal stenosis (patient with severe hyperkinetic, ballistic movements with continuous, violent flexion/extension and torsion of the neck). (B) 43-Year-old man, BFM neck subscore=8. Centromedullar high-intensity lesions with anterior and posterior spinal cord compression at level C5-C6. C5-C6 discal protrusion and posterior osteophytes associated with cervical canal stenosis and degenerative vertebral lesions. (C) 60-Year-old man, BFM neck subscore=8. Intramedullar high-intensity lesions at levels C3-C4 and C5-C6, with spinal cord atrophy and severe C5-C6 discal protrusion. BFM, Burke–Fahn–Marsden dystonia scale.

Information on neurological manifestations suggestive of CM was collected, including cervical pain, hand clumsiness, recent occurrence or worsening of bladder disorders (voiding difficulties, incontinency), gait impairment or unexplained falls.


Controls were patients with generalised dystonia and neck involvement managed in our department who fulfilled the inclusion criteria and who had not developed signs of CM at the end of the study period (2007). The same clinical information as for the cases was collected during routine visits and at the date of last visit. Cervical MRIs were only performed when the patient developed signs suggestive of CP. As sedation was needed in order to obtain good quality images, MRI were not performed for a screening of asymptomatic lesions.

Dystonia assessment

Clinical assessment of dystonia was based on standardised video analysis using the Burke–Fahn–Marsden dystonia scale (BFM).26 Functional disability related to dystonia was assessed with the BFM disability subscale. Dystonia severity for the neck and trunk was rated by a movement-disorders expert (ER) blinded to the diagnosis of CM, using the BFM movement subscale. Movement scores are the product of a severity factor (ie, 0=no dystonia, to 4=maximal severity) and a provoking factor (ie, the circumstances in which dystonia appears: 1=task specific, to 4=dystonia that persists at rest, with weights of 0·5 for the neck and 1 for the trunk). The maximum score was 8 for the neck and 16 for the trunk.

Severity was assessed on videos taken within 1 year before CM onset in the cases, and on videos taken within 1 year before the last visit in the controls.

Statistical analysis

Descriptive statistics (medians for continuous variables and percentages for categorical variables) were used to describe the patients' characteristics. The non-parametric Wilcoxon rank sum test for continuous variables and Fisher's exact test for categorical variables were used to compare cases and controls. Exact logistic regression was used for multivariable analysis. SAS software V.9.2 (SAS Institute) was used for all analyses.


Fifty-four patients fulfilled the eligibility criteria and 17 (31%) of them developed symptomatic CM during the study period. Their main characteristics are shown in table 1. CM was not significantly associated with gender. Median age at onset of dystonia was slightly higher in controls but the difference was not statistically significant (p=0.07). Cases were significantly older than controls (p<0.0001). The OR per 10-year age increase was 3.8 (95% CI 1.9 to 9.0). CM occurred after a median of 42 years after the onset of dystonia. The duration of dystonia was significantly longer in cases than controls (p<0.0001). The OR per 10-year increase was 2.3 (95% CI 1.4 to 4.2). Age and the duration of dystonia were correlated (Spearman correlation coefficient=0.68, p<0.0001); when both age and disease duration were included in a logistic regression model the relation with age remained significant (OR per 10 years=2.9, 95% CI 1.3 to 7.5, p=0.006) while the relation with disease duration disappeared (OR per 10 years=1.4, 95% CI 0.7 to 2.7, p=0.35).

Table 1

Characteristics of the patients with generalised dystonia and cervical involvement according to the presence or absence of cervical myelopathy (CM)

Patients who developed CM had more severe neck dystonia (p=0.005): the BFM movement subscore for the neck was >4 in 81% of cases and 35% of controls (OR=7.7, 95% CI 1.7 to 49.6, p=0.005); this association became slightly weaker but remained significant after adjustment for age (OR=7.0, 95% CI 1.1 to 70.3, p=0.035) or duration of dystonia (OR=6.2, 95% CI 1.2 to 47.7, p=0.032). When the neck subscore was added to a model including age and duration of dystonia, the neck subscore and age remained associated with CM, but not duration of dystonia. Thus, both the severity of neck involvement and age appear to be significant risk factors for CM. Table 2 cross-tabulates age in tertiles and BFM neck subscores in the cases and controls. There were only three cases among persons with a BFM neck subscore ≤4 and none of the cases was younger than 36 years at CM onset. Most cases belonged to the older age groups and had BFM neck subscores above 4. Neither the severity of trunk dystonia nor global disability (assessed with the BFM disability scale) was associated with CM (table 1).

Table 2

Cross-tabulation of age in tertiles and neck BFM subscores in cases and controls

The aetiology of dystonia included CP (n=23), primary dystonia (n=29), including six DYT1 (torsin A gene) mutation, and tardive (post-neuroleptic) dystonia (n=2). The distribution of aetiologies differed significantly between cases and controls (table 1, p=0.017), with a higher frequency of dystonic CP among cases, and a higher frequency of primary dystonia among controls. However, this difference was no longer significant after adjusting for age at the time of the study (p=0.20) or disease duration (p=0.57).

Neurological symptoms and signs of CM are listed in table 3. Hand clumsiness, gait impairment and falls were the most frequent disorders reported by the cases. On physical examination, lower limb or upper limb weakness, abnormally brisk reflexes, and spasticity were significantly more frequent in cases than controls. Wasting of hand muscles, upper limb sensory abnormalities, loss of reflexes, and bladder or genital dysfunction were relatively rare but were present only in patients with CM. Cervical pain was equally frequent in the two groups. Onset of neurological symptoms was insidious in 82% of cases. Only one case patient had isolated signs of radiculopathy with radicular pain.

Table 3

Neurological signs and symptoms of cervical myelopathy (CM)*


Age and severity of neck dystonia were the main risk factors for CM in this series of 54 adults with generalised dystonia with neck involvement. These findings have important implications for clinical practice as timely diagnosis and early treatment may prevent irreversible neurological injury and loss of autonomy in patients who are already severely disabled. Although primary generalised dystonia is rare,27 28 CP is a frequent condition (affecting 2–3 per 1000 live births)29 30 and dystonic CP is the second most frequent form of CP, representing 10% to 20% of all forms of CP.31 32 Although there are no precise data on the risk of CM in these patients, the number of patients with dystonic CP who are at risk of developing CM is not small and may represent an important public health problem.

Risk factors for CM

We found that median age at diagnosis of CM was 47 years and that CM did not occur before age 36 years. CM was more frequent in patients with dystonic CP than in those with other types of dystonia, although the difference was no longer significant after adjustment for age. Age appears to be an important risk factor for CM, regardless of the type of dystonia. In addition, age played a stronger role than the duration of dystonia. Dystonic patients with severe neck involvement should therefore be screened for CM, both clinically and by cervical MRI, as part of a systematic neurological evaluation from the third decade onwards. Our findings are in keeping with previous reports. In a small series of adult patients, age at onset of cervical CM was 44 years on average,22 and ranged from 32 to 49 years in both CP1 9 16 20 and primary cervical dystonia.8 13 33 34

The severity of neck dystonia remained a significant risk factor for CM after adjustment for age and duration of dystonia. Few authors1 18 35 have underlined the importance of movement-disorder severity: violent flexion-extension involuntary movements of the neck may lead to excessive mechanical stress and cervical spondylosis17 with ‘wear and tear’ of the spinal cord. Additional mechanical constraints may contribute to the onset of premature myelopathy.36 This has been mainly reported in CP17 with a high incidence of narrowing at the C4-C5 level, listhetic instability at the C3-C4, C4-C5 and C5-C6 levels, and advanced disc degeneration in over half of the patients older than 34 years (ie, 8 times the frequency in age-matched normal subjects).17 It has been suggested that these changes are due to the excessive mechanical stress caused by involuntary neck movements in dystonic CP patients, with an acceleration of the cervical disc degenerative changes that is observed in normal ageing after the fifth decade.17 Due to our relatively small number of patients with CP, we could not specifically review the cervical spine degenerative changes or compare them with descriptions based on cervical spine radiograph.17 36

Clinical clues for CM

Gait impairment (owing to recent-onset or worsening (in CP) of lower limb weakness) with falls, recent wasting of the hand musculature (associated with upper limb weakness) and bladder disorders were the most reliable clinical signs of spinal cord damage in patients with generalised dystonia and severe neck movement disorders. However, in patients who already have severe motor symptoms related to their movement disorders, late neurological deterioration, especially when insidious, may be overlooked or unduly attributed to dystonia. Indeed, it may be difficult to detect symptomatic CM in some patients who exhibit other pre-existent neurological signs such as mild spasticity. Fuji and colleagues16 18 pointed out the difficulty of linking spinal cord lesions to gait disorders1 11 12 16 18 and bladder dysfunction37 in CP patients, as both can occur in the course of the disease. Conversely, wasting of the hand muscles, a robust sign of CM associated with cervical disc disease in CP,1 18 is a late and inconsistent sign. There are fewer confounding factors in primary dystonia, as bladder dysfunction has not been reported in this context,38 and as the onset of gait disorders or bladder dysfunction is strongly suggestive of CM.1 16

In our study, cervical MRI was only performed in patients with generalised dystonia who developed signs suggestive of CM. We cannot rule out that, if cervical MRI had been systematically performed in cases and controls, we would have found abnormal images (eg, spinal cord swelling, hypersignal on T2-weighted images, significant compression) in asymptomatic patients. However, as good quality spinal cord images may not be obtained in the awake condition due to the severity of movement disorders, the risks of sedation or general anaesthesia in CM asymptomatic patients must be carefully considered. Based on our results, we stress the importance of performing cervical spine MRI from the third decade onwards in ‘at risk’ patients with severe cervical movement disorders. CM has also been reported during the course of severe cervical dystonia5 8 11 13 33 34 36 but, in this context, the signs related to CM can easily be distinguished from those of focal dystonia. Therefore, we focused on generalised dystonia in this study and propose to screen for CM from the third decade onwards as symptoms of spinal cord lesions, being similar to those due to dystonia, may be overlooked or identified late.

Early treatment of CM may prevent irreversible spinal cord injury. Important factors for treatment options are the severity of the CM signs, the duration of symptoms, the radiological evidence of myelomalacia, the severity of cervical spine lesions and the type of dystonia. Botulinum toxin (BT) injections, an effective treatment in cervical dystonia,39 40 can be used in some selected patients to decrease the severity of cervical dystonia, as an alternative treatment option to surgery, in order to prevent spinal cord complications. Our group of patients could not benefit from BT injections as the pattern of dystonia was too complex and widespread, Spinal surgery consists in decompression combined with segmental internal fixation.8 22 Perioperative use of BT may result in a partial control of the abnormal movements but requires vigilance to minimise complications.41 42 Bilateral pallidal neurostimulation has been proposed as a useful adjunct to spinal surgery and may help to reduce the severity of cervical movements disorders, one of the risk factors of CM.8 In patients with generalised primary23 24 43 or tardive44 dystonia and severe cervical involvement refractory to BT injections, one may consider bilateral pallidal stimulation before neurological or orthopaedic complications occur.45 In patients with dystonic CP, the therapeutic strategy should be individually tailored, since the benefit of BT injections or neurostimulation may be limited in this group.

In conclusion, early recognition of CM is crucial for the care of generalised dystonia and dystonic CP. As the severity of neck dystonia is a significant risk factor of CM, this diagnosis should be suspected in any dystonic patient with neck movement disorders, who was previously in a stable neurological condition, and who develops recent neurological deterioration including increased gait difficulties and unexplained falls. In addition, as age is an important risk factor of CM, we suggest that adult patients should be screened clinically and, if needed, by cervical MRI, from the third decade onwards. Besides BT injections for neck dystonia as an alternative treatment to prevent spinal cord complications, early surgery prior to the development of spinal cord lesions may be discussed in primary dystonia, as shorter disease duration tends to be associated with better general outcomes.27 In contrast, the therapeutic strategy is more complex with unpredictable outcome in dystonic CP.


We are grateful to Alliance France Dystonie, AMADYS, Fondation Motrice (patients associations) for their support. We thank Drs Sophie Sangla, Jean-Marc Trocello, Fréderic Bourdain, Valérie Cochen de Cock, Pascal Derkinderen and Gaël Gallouedec, from the Movement Disorders group, for referring the patients. We also thank David Young for editorial assistance and Floria Edouard for clinical record-keeping.


View Abstract


  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.