Rheumatic fever is an autoimmune disorder triggered by specific strains of β-hemolytic streptococci infections.1 Several studies have reported high frequencies of obsessive-compulsive symptoms, obsessive-compulsive disorder (OCD), tic disorders, and attention deficit hyperactivity disorder (ADHD) in rheumatic fever with and without Sydenham's chorea (the late CNS expression of rheumatic fever).2,3 These psychopathological manifestations have been described in acute phase rheumatic fever patients. The presence of OCD, tic disorders, and body dysmorphic disorder has been recently reported in another sample of rheumatic fever patients.4 The study presented here was designed to explore the frequency of specific neuropsychiatric disorders and symptoms in heart disease patients with and without history of rheumatic fever.

Subjects

All participants belonged to a heart disease outpatient clinic in an academic hospital (Clinical Hospital, University of Sao Paulo Medical School ) and were assessed between June 2000 and January 2003. Inclusion criteria: The rheumatic fever group was composed of 51 consecutive heart valve disease patients aged between 18 and 45 years old who had rheumatic fever in the past (seven of which had had Sydenham's chorea). The rheumatic fever diagnosis was made according to modified Jones Pharma, Inc. criteria5 by a pediatrician (MTM) or a cardiologist (GS) from the same academic hospital. The comparison group was initially designed to match for age and gender and included 46 consecutive adults who were followed in the same outpatient clinic and had not reported rheumatic fever, Sydenham's chorea, arthritis episodes, or long-term penicillin treatment. This group includes patients with nonrheumatic heart valve diseases (N=11), severe high blood pressure (N=11), coronary insufficiency (N=11), congenital heart disease (N=5), and arrhythmia (N=5). Exclusion criteria: age younger than 18 or older than 45, and the presence of neurological disorders with CNS manifestation (one case with major cognitive impairment attributed to a cerebrovascular disease was excluded). The Ethics Committee of the University of Sao Paulo approved the study, and written informed consent was obtained from all patients.

Psychiatric Evaluation

The diagnoses of OCD and tic disorder were made according to DSM-IV criteria.6 Tic disorder diagnosis included Tourette's disorder and chronic tic disorder. Participants were given the Yale-Brown Obsessive Compulsive Scale (Y-BOCS) checklist to assess obsessive-compulsive symptoms.7 The module for ADHD of the Kiddie-Schedule for Affective Disorders and Schizophrenia—Epidemiological Edition (K-SADS-E) was the instrument used to assess ADHD symptoms.8 The interviewers were blind to cardiologic and rheumatic fever status. Two senior psychiatrists (ECM, AGH), also blind to which group the patient belonged, performed Best Estimate Diagnosis9 for psychiatric disorders. Interreliability between these psychiatrists for the diagnosis of all these disorders ranged between 0.90 and 0.95. In two cases of disagreement, final diagnoses were obtained by consulting a third senior psychiatrist opinion (MTM).

Statistical Analysis

Comparisons of categorical variables among groups were performed using Fisher's exact and chi-square tests for 2×2 tables. Comparisons of continuous variables were carried out using analyses of variance (ANOVAs) (Students' t test for two independent groups). Values of p<0.05 (two-tailed) were considered significant. We used the SPSS 11.0 package (SPSS Inc., Chicago, Ill.) for the statistical analyses.

Primary Analysis

The total heart disease patient group comprised 97 individuals: 51 with rheumatic fever and 46 with no rheumatic fever history. The mean ages at the time of the interview of rheumatic fever and comparison groups were, respectively, 31.47 (SD=8.403) and 33.37 (SD=9.137, p=0.289). The mean age of first rheumatic fever episode found in the rheumatic fever group was 10.4 (SD=4.3). There were no significant differences between groups regarding gender, socioeconomic status (p=0.064) and total years of education (p=0.164), whether stratifying data by referral source or not.

The frequency of obsessive-compulsive symptoms increased in the rheumatic fever group, in contrast to the comparison subjects. The rheumatic fever group also showed a higher frequency of aggressive obsessions, total compulsions, and ordering-arranging compulsions (Table 1).

Exploratory Analyses

Regarding gender, we found a higher frequency of obsessive-compulsive symptoms in men rheumatic fever than in comparison men (p=0.038). No differences were observed regarding obsessive-compulsive symptoms comparing women with rheumatic fever and comparison women (p=0.211). The presence of tic disorder in the whole sample was associated with attention deficit symptoms in both rheumatic fever (p=0.027) and comparison subjects (p=0.027). Furthermore, seven female subjects of the rheumatic fever group had had Sydenham's chorea. Regarding all the psychiatric symptoms studied, this chorea subgroup did not differ when compared with those with rheumatic fever without chorea (p=0.19).

This is the first study comparing adults with a previous history of rheumatic fever to comparison subjects that reports a higher frequency of obsessive-compulsive symptoms in nonactive rheumatic fever patients. The mean age of the rheumatic fever first episode found in our sample was similar to prior studies with children.2,4

These findings may suggest that rheumatic fever activity1 is not a necessary condition for the expression of neuropsychiatric symptoms. In acute phase rheumatic fever patients, these manifestations could be the result of an abnormal immune process related to the activity of the disease. In the nonacute phase rheumatic fever patients,1 it cannot be excluded that acute changes might have persisted or might have triggered other psychopathological changes that increase the susceptibility to neuropsychiatric disorders. It is well established that carditis continues to evolve even in the absence of rheumatic activity and also that may worsen with reinfections.10 It is also known that pregnant women with past Sydenham's chorea are prone to develop chorea gravidarum, even without the recurrence of strep infections.11 Furthermore, basal ganglia have been implicated in OCD genesis2,3,4 and persistent changes in the caudate of Sydenham's chorea patients have been detected by MRI.12 Thus, it is possible that strep infections and/or rheumatic fever may trigger obsessive-compulsive symptoms, which may persist through adulthood independently of new rheumatic fever episodes. Another possibility is that obsessive-compulsive symptoms and OCD share a common genetic vulnerability with rheumatic fever. Patients bearing susceptibility for obsessive-compulsive symptoms and OCD could have symptoms triggered by a rheumatic fever episode. Interestingly, the association between tic disorder and ADHD symptoms which was previously reported by our group in children2 was found in our adult sample.

The most striking finding in the present study was the occurrence of obsessive-compulsive symptoms, mainly aggressive obsessions and ordering-arranging compulsions in a significant rate of the rheumatic fever group. These phenomenological characteristics have been reported as being part of the early onset OCD subtype and the tic-related OCD phenotype.13 It is worth noting that according to our findings in previous studies that assessed children with streptococcal infections, men outnumbered women regarding neuropsychiatric disorders.3 It is possible that significant obsessive-compulsive symptom rates in women with rheumatic fever would be expected in larger samples. OCD, tic disorder, and ADHD symptoms were more frequent (though not significant) in the rheumatic fever group (Table 1). This may be explained by the small sample sizes.

There are some limitations to this investigation, such as the small sample sizes and the difficulty in rigorously matching groups for gender. Unfortunately, it was not possible to obtain reliable information regarding the onset of the obsessive-compulsive symptoms. Patients often did not remember the time of onset of psychiatric symptoms. The interval between the first episode of rheumatic fever and the beginning of the neuropsychiatric symptoms would bring relevant data. Although our study was not designed to assess severity of symptoms, doing so could have brought additional data. Moreover, structural brain imaging studies have not been performed. Finally, the high level of psychopathology found in the rheumatic fever group could also be secondary to the chronic heart valve disease. We hope to have controlled for these variables by using the comparison group and blind interviewers.

Obsessive-compulsive symptoms, especially aggressive obsessions and ordering-arranging compulsions, are more frequent in heart disease patients with a history of rheumatic fever when compared to heart disease comparison patients. This phenotype resembles the early onset OCD subtype. These results should encourage clinicians to evaluate rheumatic fever patients for obsessive-compulsive symptoms. Further studies are needed to understand how rheumatic fever may be related to an increased risk of obsessive compulsive symptoms and related disorders.

This work was presented, in part, at the Fourth International OCD Conference, St. Thomas-U.S. Virgin Islands, February 2000, and at the American Psychiatric Association Annual Meeting, Washington, DC, May 1999.

This project was sponsored by grants 98/09895-9 and 99/08560-6 from the Research Foundation of the State of São Paulo (FAPESP). Also supported by Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPQ), Brazil (521369/96-7).

The authors thank James F. Leckman, Max Grimberg, and Mariana Cury.