Article Text


Sydenham’s chorea may be a risk factor for drug induced parkinsonism
  1. A L Teixeira,
  2. F Cardoso,
  3. D P Maia,
  4. M C Cunningham
  1. Movement Disorders Clinic, The Federal University of Minas Gerais, Av Pasteur 89/1107, 30150-290 Belo Horizonte MG, Brazil
  1. Correspondence to:
 Professor Francisco Cardoso; 

Statistics from

Sydenham’s chorea, the most common cause of acquired chorea in childhood, is a delayed complication of group A β-haemolytic streptococcal infection.1 It is thought to be caused by antibodies induced by streptococci which cross react with basal ganglia antigens.2 Despite the decrease in Sydenham’s chorea in developed countries, there is a renewed interest in this condition because of the hypothesis that a similar mechanism may play a role in the pathogenesis of a subset of patients with tics and other neuropsychiatric disorders.3

The treatment of Sydenham’s chorea is based on the combination of penicillin and antichoreic drugs (valproic acid and/or dopamine antagonists). At the movement disorders clinic of the Federal University of Minas Gerais (MDC-UFMG), located in an area where Sydenham’s chorea remains endemic, we have been struck by the occurrence of drug induced parkinsonism among patients with Sydenham’s chorea. We therefore decided to investigate the incidence of this complication in a cohort of patients with the condition, as well as in a group of subjects with Tourette’s syndrome. The latter is an interesting choice for a control group as Tourette’s syndrome and Sydenham’s chorea share several clinical features—for example, childhood onset, the constellation of motor and behavioural disturbances, and the response to neuroleptic agents.4


In the first part of the study we undertook a retrospective review of the case records of patients with Sydenham’s chorea and Tourette’s syndrome followed up at the MDC-UFMG from July 1993 to October 2002, looking for drug induced parkinsonism. We then compared the chlorpromazine equivalent dose used in patients with Tourette’s syndrome, Sydenham’s chorea, and drug induced parkinsonism. For each patient with Sydenham’s chorea we randomly selected two age matched subjects with Tourette’s syndrome. Sydenham’s chorea was diagnosed according to a modified Jones criteria,1 and the Diagnostic and Statistical Manual of Mental Diseases, fourth edition (DSM-IV) criteria were used to identify patients with Tourette’s syndrome. Drug induced parkinsonism was diagnosed when patients exposed to neuroleptics were found to have bradykinesia and at least one of the following: rigidity, tremor, postural instability. All patients were seen by one of us (FC). Differences were considered statistically significant at a probability (p) value of < 0.05.


Sydenham’s chorea was diagnosed in 91 patients and Tourette’s syndrome in 97 during the study period. Five patients (5.5%) with Sydenham’s chorea (mean (SD) age, 13.4 (2.1) years), of whom four were female, developed drug induced parkinsonism, whereas this complication was not identified in the Tourette’s syndrome group (p = 0.03, Fisher’s exact test). Parkinsonism was characterised by the presence of bradykinesia and rigidity in all five subjects, tremor was identified in three, and postural instability was not observed. The mean cumulative chlorpromazine equivalent dose used in the patients with Sydenham’s chorea when they developed parkinsonism was 16 518 (6254) mg. The onset of drug induced parkinsonism occurred after a mean exposure of 88.2 (65.8) days.

The cumulative chlorpromazine equivalent dose in the 10 Tourette’s syndrome patients (two female; mean age 13.5 (1.1) years (p = 0.26 v Sydenham’s chorea)) during a three month period was 19 575 (6329) mg (p = 0.76 v Sydenham’s chorea). At the time of the onset of drug induced parkinsonism the mean chlorpromazine equivalent dose in the Sydenham’s chorea patients was 176.6 (95.5) mg; in the Tourette’s syndrome group the figure was 217.5 (220.0) mg (p = 0.05, paired t test).


We showed that 5.5% of our patients with Sydenham’s chorea developed drug induced parkinsonism during treatment with neuroleptics, while this complication was not observed in a cohort of Tourette’s syndrome patients of the same body weight. There are several possible explanations for this. First, the patients with Sydenham’s chorea might have received a higher dose of neuroleptics. This hypothesis is ruled out by the finding that 10 randomly selected, age matched Tourette’s syndrome patients and the five Sydenham’s chorea patients with drug induced parkinsonism were treated with a similar cumulative chlorpromazine equivalent dose during the three month period when the latter developed parkinsonism. One may argue that although both groups received a similar cumulative dose of neuroleptics in the time between the start of treatment and the onset of drug induced parkinsonism, the patients with Sydenham’s chorea could have been exposed to a higher dose of dopamine receptor blocking drugs at the time when they developed parkinsonism. However, this was not the case because if anything the mean dose of neuroleptic in the Sydenham’s chorea subjects at the onset of drug induced parkinsonism was lower than in the Tourette’s syndrome patients, though the difference failed to reach statistical significance. A second explanation for our findings is the concomitant use of valproic acid in three of our Sydenham’s chorea group. Although this drug has been implicated in the development of drug induced parkinsonism, this complication has only been described in adults treated for a period of 12 months or more.5 A third and also unlikely hypothesis is the overrepresentation of female patients in the Sydenham’s chorea group; however, recent studies have consistently failed to identify sex as a risk factor for drug induced parkinsonism.6 Our results thus support the conclusion that in comparison with Tourette’s syndrome, patients with Sydenham’s chorea are at greater risk of developing drug induced parkinsonism.

Our study has limitations. First, it is a retrospective investigation of patients seen at a tertiary referral centre. This approach can lead to false positive associations, particularly in studies of disease clustering.7 However, this limitation is minimised by the use of a control group of Tourette’s syndrome patients referred to the MDC-UFMG by the same process. Second, we were not blinded to the clinical status of the patients. Although this would have been ideal, to have remained blinded to the diagnosis would have been impossible because of the obvious phenomenological differences between Sydenham’s chorea and Tourette’s syndrome.

We hypothesise that the increased susceptibility of patients with Sydenham’s chorea to develop drug induced parkinsonism reflects an underlying nigro-striatal dysfunction produced by cross reactive streptococcus induced antibodies. Recent studies suggest that the antibodies circulating in the sera of Sydenham’s chorea patients not only recognise antigens of the basal ganglia2 but also CNS myelin, causing acute disseminated encephalomyelitis.8 It is thus possible that the anti-basal ganglia antibodies also cross react with neurones of the substantia nigra.

The results of our study have two implications. First, there is a need for caution when treating patients with Sydenham’s chorea with dopamine receptor blocking drugs. Second, as patients with Sydenham’s chorea and Tourette’s syndrome respond differently to neuroleptics, this weakens the hypothesis that similar mechanisms are involved in the pathogenesis of these conditions.


View Abstract

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.