Introduction

Tic disorders (TD), including Tourette syndrome and chronic motor or vocal tic disorder, are consistently found to be linked to obsessive–compulsive disorder (OCD) [1, 2] and to attention deficit hyperactivity disorder (ADHD) [3–5] with significant bi-directional comorbidity. Indeed, the Diagnostic and Statistical Manual, fifth edition will likely include “tic-related” OCD as a subtype specific given the strength of the clinical, genetic, neurobiological and treatment response data [6]. Brain imaging studies of TD, ADHD and childhood-onset OCD consistently point to involvement of cortico-striato-thalamocortico circuits in all three disorders [7–11]. Family genetic studies suggest there are likely shared genetic vulnerabilities between OCD and TS [12–17]. For example, O’Rourke et al. [18] reported that the presence of OCD or subclinical OCD in TS proband was associated with significantly increased risk of comorbid ADHD and TD in their first-degree relatives.

Some studies have addressed the clinical implications of having comorbid OCD or ADHD and compared tic severity and related psychopathology to cases of only TD. In the case of OCD, the results have been somewhat inconclusive. For example, one study of 61 adults with TD, OCD or TD + OCD demonstrated elevated levels of co-occurring psychopathologies in the TD + OCD condition [19]. A recently published study reporting on over 5,000 children with TD also found that the presence of comorbid OCD was associated with higher rates of additional comorbid diagnoses [20]. In contrast, a study of 306 youth did not find TD + OCD to have elevated rates of comorbid diagnoses [21], when compared with TD or OCD alone. Some studies have found greater tic severity in TD + OCD [22], supporting the hypothesis that the comorbid condition represents a more severe subtype of TD.

With regard to ADHD, the evidence so far points more toward the importance of the comorbid condition for disruptive and aggressive behaviors than for tic severity. Several studies have shown that TD youth without ADHD are less prone to angry outbursts, disruptive behaviors and social maladaptation than those with both conditions [23, 24]. Although comorbid ADHD seems to explain or in some way trigger many of the disruptive behaviors commonly reported in TD, strengthening the case for a related etiology, a study of the course of the combined condition over 4 years showed that the tic symptoms were considerably more likely to remit than symptoms of ADHD [25]. A large-scale study of children with TD also pointed to comorbid ADHD being associated with higher likelihood of internalizing disorders, particularly in younger children [26]. Tic severity has generally been found to be similar [23, 27, 28] across TD only and TD with ADHD, although at least one study found more severe phonic tics in youth with the comorbid condition [24].

The purpose of the current report is to present data from a large sample of youth with TD, with and without OCD and ADHD, and to identify associations between the presence of the comorbid disorders and the severity of tics and other symptoms, additional comorbidities, level of functioning and degree of psychosocial stress. Our a priori hypothesis was that individuals who had comorbid OCD or ADHD would present with more severe tics and greater psychosocial stress than those who had TD alone.

Method

Subjects

We present data from two multisite, prospective longitudinal studies undertaken by researchers affiliated with the Tourette Syndrome Study Group [29, 30]. In brief, both studies looked at longitudinal data collected over 25 months on children with TD. The aim of the original studies was to compare the course of illness of children who met the published diagnostic criteria for pediatric autoimmune neuropsychiatric disorders associated with streptococcal infections (PANDAS) with those who did not [31]. Neither study provided evidence that exacerbations in tic or OC symptoms were linked with antecedent group A beta hemolytic streptococcal infections (GABHS) [29, 30].

In Study 1, a total of 40 case–control pairs matched for PANDAS status were recruited from ten sites across the USA [29]. In Study 2, 31 PANDAS cases and 53 non-PANDAS comparison cases were recruited from six sites, all of which also participated in Study 1 [30]. For the purpose of this analysis, only those individuals with a chronic TD were included. Inclusion criteria were as follows: (1) presence of OCD and/or a chronic tic disorder by DSM-IV criteria [32]; (2) age at onset between 3 and 14 years; (3) currently <16 years of age; and (4) informed consent provided by parent or guardian and assent provided by the subject. In both studies, PANDAS subjects were required to meet all five of the published diagnostic criteria for PANDAS [31].

Exclusion criteria included the following: (1) possible Sydenham chorea; (2) presence of mental retardation, pervasive developmental disorder, or a secondary tic disorder other than PANDAS (e.g., drug induced or neuroacanthocytosis); and (3) a psychiatric disorder other than OCD or chronic tics as the primary focus of treatment.

Procedures

At the initial visit, informed consent and/or assent were obtained. Recruitment rates for both studies were high, with over 90 % of patients queried agreeing to participate. A systematic clinical interview and examination were performed to ensure that the subject satisfied all of the enrollment criteria. In Study 1, a fully structured psychiatric interview, the Diagnostic Interview Schedule for Children-DSM-IV version (DISC-IV) [33], was used to verify that criteria for OCD or a chronic tic disorder were met. The DISC-IV was also used to determine the presence of any comorbid conditions. For eligible subjects, symptom severity was assessed with the following instruments: Yale Global Tic Severity Scale (YGTSS) for tics [34]; the Children’s Yale-Brown Obsessive–Compulsive Scale (CY-BOCS) for OCD [35]. Ratings were also collected for ADHD (Conners Abbreviated Symptom Questionnaire-Parent [ASQ-P]) [36], anxiety (Multidimensional Anxiety Scale for Children [MASC]) [37] and depression (Child Depression Inventory-Short Version [CDI SV]) [38]. In Study 2, the assessment methods were identical except that the diagnoses, including TD, OCD, and ADHD, were made after reviewing all available information including data from the Schedule for Affective Disorders and Schizophrenia for School-Age Children (SADS) [39].

In both studies, parents completed a rating of the level of perceived psychosocial stress at the baseline assessment (Perceived Stress Scale Parent—PSSP [40]). In both studies, the Child Global Assessment Scale for global functioning (C-GAS) was rated independently both by the site investigators and by the parents of each subject [41].

Data analysis

Analysis was carried out using PASW Statistics 18. A multivariate general linear model (MANCOVA) including both OCD and ADHD was used to test for their effects on tic severity (YGTSS), depression (CDI), anxiety (MASC), global functioning (C-GAS), and psychosocial stress (PSSP) and for potential interactions between the comorbid disorders. Meeting criteria for PANDAS and the study of origin were included in each analysis as a covariate. Binary logistic regression was used to model the effects of OCD and ADHD on the likelihood of additional comorbid diagnoses. All tests are two tailed with significance set at the 0.05 level.

Results

A total of 158 children, aged 6–14.5 years who met criteria for a chronic TD, were recruited from ten sites across the USA. The majority of these (N = 143, 90.5 %) met criteria for Tourette syndrome, and the minority for chronic motor (N = 10, 6.32 %) or vocal (N = 1, 0.6 %) tic disorders or tic disorder NOS (N = 5, 3.16 %). Eighty-five (53.8 %) of the 158 subjects with TD also met criteria for a diagnosis of OCD, and 61 (38.6 %) met criteria for ADHD. Thirty-eight subjects (24.1 %) met criteria for both OCD and ADHD and 50 (31.6 %) had neither ADHD nor OCD. These rates resemble those reported by others [42], although ADHD has usually been reported to be the more frequent comorbid condition [43]. Thirteen additional subjects did not meet criteria for a chronic TD and were not included in the current report. As presented in Table 1, the demographic and clinical characteristics were similar across the two studies. Subjects in Study 1 were slightly older and had somewhat higher rates of psychiatric diagnoses (including OCD and ADHD) than those in Study 2. Study 2 included more subjects who met the criteria for PANDAS.

Table 1 Demographic and clinical characteristics for 158 children with tic disorders by sample of origin
Full size table

Tic severity

Youth with comorbid OCD had significantly more severe tics when compared with those without OCD, as measured by the YGTSS (F (1) = 14.6, p < 0.001). Subjects with ADHD showed a non-significant trend to more severe tics compared with the non-ADHD subjects (F (1) = 1.5, p = 0.2). The analyses did not reveal a significant interaction between the presence of OCD and ADHD with regard to tic severity (p > 0.39).

Related psychopathology

Comorbid OCD was also significantly associated with all the secondary measures of psychopathology, which included anxiety (F (1) = 5.33, p < 0.05), depression (F (1) = 17.3, p < 0.001), and psychosocial stress (F (1) = 16.1, p < 0.001), as well a significant negative association with the overall global level of functioning as assessed by both parents (F (1) = 11.7, p < 0.005) and clinicians (F (1) = 37.4, p < 0.001). The presence of ADHD was significantly associated with the levels of stress (F (1) = 4.6, p < 0.05) and global functioning as reported by parents (F (1) = 8.1, p < 0.01) and clinicians (F (1) = 5.1, p < 0.05), but not with specific measures of anxiety or depression. Table 2 summarizes the measures of tic severity and related psychopathology across conditions. No significant interactions were found between the effects of OCD and ADHD for any of these measures.

Table 2 Tic severity and related psychopathology in 158 youth with tics stratified by the presence of comorbid OCD and/or ADHD and by PANDAS status
Full size table

To investigate the consequences of having both comorbid disorders, we also conducted a multi-group analysis of variance comparing TD only, to TD + OCD, to TD + ADHD, and to TD + ADHD + OCD. In this analysis, only comparisons to the doubly comorbid condition were significant, implying that while OCD alone was considerably more significant than ADHD alone, the presence of both conditions posed more heightened risk for increased tics and related psychopathology than either condition does alone.

Comorbid diagnoses

The results of the logistic regression analyses showed that youth with OCD and/or ADHD were more likely to meet criteria for additional diagnoses compared to those without OCD or ADHD. Specifically, comorbid OCD significantly increased the likelihood of another internalizing disorder such as depression or an anxiety disorder, while ADHD increased the likelihood of additional externalizing disorders such as oppositional defiant disorder. Table 3 summarizes the impact of OCD and ADHD as predictors of additional diagnoses.

Table 3 Binary logistic regression of anxiety disorders, oppositional defiant disorder, and depression in 158 youth with tic disorders as predicted by the presence of OCD or ADHD
Full size table

Impact of PANDAS

Meeting the published diagnostic criteria for PANDAS was included as a covariate in the GLM analyses of tic severity and related psychopathology. PANDAS status was not a significant variable in any of the analyses. PANDAS status was also not a significant predictor of comorbid conditions in the logistic regression models.

Discussion

This study reports on a large and well-characterized sample of pediatric patients diagnosed with a chronic TD and compares clinical features based on the presence or absence of a diagnosis of OCD or ADHD. The results highlight the impact that two frequently comorbid conditions can have on the severity and associated difficulties of youth with tic disorders. Consistent with our a priori hypothesis, pediatric patients with comorbid TD and OCD had more severe tics than children with TD who did not have OCD. This finding is in line with those reported by Coffey et al. [19] who found TD + OCD subjects to be characterized by more severe tics as measured by YGTSS. In contrast are the findings reported by Lewin et al. [21], who found no significant increase in tic severity in TD + OCD patients as measured by ADIS clinician severity rating (CSR). It is not certain to what extent this discrepancy may relate to the different assessment tool used in each study.

In addition to greater tic severity, subjects with comorbid OCD had greater severity on several measures of psychopathology including depression, anxiety, and levels of psychosocial stress as well as poorer global functioning. Interestingly, comorbid ADHD had a less pronounced impact on these variables. On average, youth with ADHD had only non-significantly more severe tics than those without ADHD, and the effect on other measures of psychopathology was limited to general measures such as psychosocial stress and global functioning but not tied to more severe specific symptoms such as anxiety or depression. These findings also reaffirm those reported by others who identified a greater effect of comorbid ADHD on stress and global functioning than on tic severity [28].

One way of understanding these results is to see OCD and ADHD as interacting in different ways with TD. Comorbid TD and OCD might represent a more severe subtype of TD, leading to increased symptoms relevant to both disorders. Comorbid ADHD may be better conceptualized as a separate problem complicating the presentation through added difficulties for the child and family. Earlier studies in adults have pointed to comorbid TD + OCD as a subtype of TD [19] and the relationship is likely to be recognized by the inclusion of a tic-related diagnosis of OCD in DSM V [6]. The current results within the pediatric population support this hypothesis. An alternative possibility that needs also to be considered is the potential for some compulsions to be characterized as tics, leading to greater tic severity ratings among youth with comorbid OCD. The similarities, and consequent ambiguous differences, between tics and compulsions, were described as long ago as by Pierre Janet [44] and more recently captured with the term “compultics” [45]. ADHD by comparison may be adding to the more general difficulty in functioning normally and adding considerable stress as demonstrated by the results of our investigation. The results of this study also support the strong link that has been reported elsewhere between disruptive behaviors in TD and the presence of ADHD [23, 24]. In fact, apart from psychosocial stress and global functioning, the impact of ADHD was most prominent in increasing the likelihood of meeting criteria for a disruptive behavior disorder.

The findings should be interpreted in light of the studies limitations including the absence of OCD-only, ADHD-only and healthy control groups. Inclusion of these contrast groups would present a clearer picture and allow for better interpretation. Second, we relied on patients who had been referred to specialty TS clinics. This may have created an ascertainment bias. A potential limitation of the present study lies in the fact that the data analyzed were collected as part of an effort to investigate the relationship of streptococcal infections to the occurrence or exacerbation of TD and OCD. As such, the subjects were recruited because they either fit the published criteria for PANDAS or were matched to others who did, creating a potential bias that could in principle limit the ability to generalize these findings. However, the fact is that neither of the two studies showed a significant link between the PANDAS criteria and symptom severity. To further address this issue in the present study, PANDAS status was included as a covariate in the analyses. This variable was not found to be a significant factor in any of the results. Finally, the present study focuses on cross-sectional comparisons between subjects with and without OCD or ADHD, but presents only limited data addressing each of those conditions as one variable in a multivariate approach. This also presents a limitation on the interpretation of causal links explaining the associations reported.

Conclusion

Chronic TD with comorbid OCD, on average, is a more severe form of TD associated with more severe tics and internalizing psychopathology and poorer global functioning compared to youth with chronic TD and no OCD. In contrast, chronic TD with comorbid ADHD is associated with higher levels of psychosocial stress, poorer functioning, and increased risk for externalizing behavioral problems. However, the interactions between the two comorbid conditions need more research and an even larger sample size is necessary to properly investigate this multi-group model. Research needs to further investigate the links between TD, OCD, and ADHD as well as the genetic, epigenetic, and neurobiological factors contributing to these associations.