Tic disorders and ADHD: answers from a world-wide clinical dataset on Tourette syndrome
- First Online:
- Cite this article as:
- Freeman, R.D. & Tourette Syndrome International Database Consortium Eur Child Adolesc Psychiatry (2007) 16: 15. doi:10.1007/s00787-007-1003-7
- 718 Views
Tourette syndrome (TS) is a neurodevelopmental disorder with frequent comorbidity with Attention- deficit-Hyperactivity disorder (ADHD). The impact of this association is still a matter of debate.
Using the TIC database containing 6,805 cases, the clinical differences were ascertained between subjects with and without ADHD.
The reported prevalence of ADHD in TS was 55%, within the range of many other reports. If the proband was diagnosed with ADHD, a family history of ADHD was much more likely. ADHD was associated with earlier diagnosis of TS and a much higher rate of anger control problems, sleep problems, specific learning disability, OCD, Oppositional-defiant disorder, mood disorder, social skill deficits, sexually inappropriate behaviour, and self-injurious behaviour. Subjects with seizures and with Developmental Coordination Disorder also had high rates of ADHD. Anxiety disorder, however, was not more frequent. Preliminary data suggest that most behavioural difficulties in ADHD are associated with the Combined or Hyperactive-Impulsive Subtypes of ADHD. Every large site (>200 cases) had a significantly increased rate of anger control problems in cases with ADHD.
Subjects with TS have high rates of ADHD and complex associations with other disorders. Clinically the findings confirm other research indicating the importance of ADHD in understanding the behavioural problems often associated with the diagnosis of TS. Additional ADHD comorbidity should be taken into account in diagnosis, management, and training.
Keywordscomorbidityattention-deficit/hyperactivity disorderobsessive-compulsive disorderoppositional-defiant disordertourette syndrome
What is the answer? [I was silent.] In that case, what is the question? Gertrude Stein 
Tourette syndrome (TS) is a usually complex neurodevelopmental disorder with onset of multiple tics in childhood and gradual improvement in adolescence and adult life . Attention-deficit-Hyperactivity disorder (ADHD) is generally acknowledged to be common, important clinically, and often comorbid with Tourette syndrome [4, 34], typically developing before the onset of tics . A high rate of comorbidity in TS was reported in a previous publication ; ADHD was the most common , and may not be due to referral bias alone. The consequences of ADHD (especially with associated OCD and Specific Learning Disability) may overshadow tics, whose effect on impairment is frequently small [10, 36, 37].
As will be seen, the dataset from which the following findings are derived is very suitable for approaching answers to a few questions, capable of provoking useful questions about others, and inappropriate for others.
Geographical distribution of cases (TOTAL: 6805)
The project originated from a concern by several clinician-researchers in Western Canada and North Dakota, about reports with widely different results, emanating from sites with relatively few cases, each of which had its own unstated or unknown selection biases. Since at that time the prevalence of TS was thought to be quite uncommon (perhaps 1:10,000) it seemed unreasonable to expect studies to be performed on the general population; the next best approach would be multiple series of consecutive cases from many sites which could then be compared or aggregated. A secondary goal was to be able to move toward research projects on small subsets of patients with TS who satisfied certain criteria: to locate them for more in-depth research would be difficult or impossible without collecting a very large series of patients. For this secondary purpose it would not be necessary to impose rigorous interrater reliability training and procedures, because the selection of subsets for research would incorporate those at a later stage.
We have found that this arrangement, which is designed to allow sites to submit cases without any funding or elaborate procedures, has been very successful in recruiting a huge dataset, yet it can be misunderstood. It is essential, in reading the presentation of data that follows in this paper, to realize that the figures represent clinical reports to a registry by experienced clinicians, but with undetermined ascertainment and referral biases. This enables us to make some descriptive statements about clinical samples with confidence, but others with less. We were not able to detect diagnostic error, and by agreement we do not identify specific sites. Nevertheless, if trends or associations are found within the data from multiple large sites, greater confidence may be extended to those findings in clinical cases, and they might be a stimulus for replication studies using different methodologies. These findings cannot be extrapolated to non-referred persons (community samples).
Appropriateness of questions
How common is the ADHD diagnosis in clinical samples worldwide?
With what other disorders, factors, and patterns of behaviour is ADHD associated?
Does the subtype of ADHD matter?
Are there significant sex differences?
ADHD in the literature
The ADHD prevalence in adults is about 4–8% in the 18–29-year age group [18, 19]. It was found to be significantly associated with Intermittent Explosive Disorder, anxiety disorders (29%), mood disorders (21%), and impulse control disorders (25%), and in general, ADHD is associated with comorbidity [3, 5, 15, 42]. In adolescence and adult life, ADHD symptoms persist in up to 65% [5, 22, 23, 43], but the symptom pattern may change. A community study by Kurlan et al.  found a rate of ADHD of 38.4% in those with tics. They reviewed previous studies, which reported a range from 40% to 70% for concurrent ADHD. Spencer et al. [36–38] showed that the presence of tics did not influence the course of ADHD, so that findings in ADHD samples might be applicable to ADHD in a TS group. Other researchers have come to the conclusion that TS-Only cases are close to normal controls in most respects, but that comorbid ADHD is associated with high rates of psychopathology [6, 8, 9, 11, 13, 14, 17, 20, 30, 33, 35–37, 39, 40].
Size of sitea
Specialist distribution of cases
The eight very large sites with over 200 cases each were subjected to separate analysis. This sub-sample consisted of 3,065 cases. After results were obtained on the pooled sample, the same procedures were run on the sub-sample to see whether the trends found were confirmed.
For the three ADHD Subtypes, only recent data on 153 consecutive cases was available. There were only 11 of the Hyperactive-Impulsive Subtype, thus too small a number for meaningful analysis. The preliminary comparisons made are therefore between the Predominantly Inattentive Subtype (N = 64) and the Combined Subtype (N = 78).
Statistical analyses were performed using SPSS (Statistical Package for the Social Sciences, version 15.0, SPSS Corp., 2006). Groups were compared using Pearson’s χ2-test for categorical variables with Fisher’s Exact Test modification, one-way ANOVA, or nonparametric correlations where appropriate. Variables with statistically significant differences between groups in the bivariate analysis were then entered into binary logistic regression analyses to control for the effect of multiple variables. For these analyses, the probability of F to enter the regression equations was set at ≤0.05 and the probability of F to remove was set at ≥0.10. Because of the very large dataset, differences of small or trivial magnitude could acquire statistical significance. The minimum p-value for serious consideration was therefore set at 0.01, not the more usual 0.05 level, except where specified.
General characteristics of dataset
TS/tic family history
Tic severity mild
Tic severity moderate
Tic severity severe
TS-only (no comorbidity)
Mood disorder comorbid
Anxiety disorder comorbid
Mental retardation comorbid
Anger control problems
Self-injurious behaviour (SIB)
Social skill deficits
Inappropriate sexual behaviour
Child (<18 years)
Medication for tics (ever)
Between-group comparisons in 6,805 people with Tourette syndrome and ADHD (TS+ADHD) and without ADHD (TS−ADHD)
Mood disorder (children)
Specific LD (children)
MR (intellectual disability)
DCD in children
Sleep (now) (children)
Anger (now) (children)
Self-injurious behaviour (children)
Social skills deficits (children)
Sexually inappropriate behaviour (children)
Medication for tics (children)
Overall, ADHD was diagnosed in 61% of children and 39% of adults. The range for large sites was 38–91%.
Male excess in TS was found at every site without exception (regions vary from 15% to 26% female cases); the mean for the full dataset is 19% female, or a 1:4 ratio. 59% of males and 40% of females were diagnosed with comorbid ADHD (55% and 36% in children, respectively). There were no significant sex effects between the Combined and Inattentive Subtypes.
Age at onset, diagnosis, and when registered
These differ when ADHD has been diagnosed: age at onset in TS+ADHD is 6.1 years, 6.8 years in TS−ADHD (p < 0.001, SD 2.5, 3.2 respectively). Diagnosis was earlier in TS+ADHD: 11.6 years vs. 15.1 years (p < 0.001, SD 7.8 and 11.1). The patient was first seen by the TS specialist at 13 years in TS+ADHD, 17.0 years in TS-ADHD (p < 0.001, SD 8.7 and 12.1). The delay between tic onset and diagnosis was 5.4 years in TS+ADHD, 7.8 years in TS-ADHD (p < 0.001, SD respectively 7.4 and 9.8 years). Cases of the Combined Subtype were seen earlier by reporting clinicians (9.4 years, SD 2.7) than Inattentive cases (11 years, SD 4.1).
Are there major differences in the rate of ADHD diagnosed by different specialists? Psychiatrists diagnose it in 62% of children and 42% of adults; neurologists diagnose it in 59% and 35% respectively, and pediatricians in 55% in children.
Subtypes of ADHD
This was not tracked until recently, so these findings are preliminary. Of 153 sequential cases, 11 (7%) fit the Hyperactive-Impulsive Subtype, 78 (51%) the Combined Subtype, and 56 (37%) the Predominantly Inattentive Subtype.
Peak tic severity
There were no significant differences in TS+ADHD vs. TS−ADHD.
Family history of tics and OCD
Rates did not differ because of the diagnosis of ADHD (51% of TS+ADHD had a positive family history of tics, and 53% in TS−ADHD); for OCD it was 21% in TS+ADHD, 22% in TS−ADHD.
Family history of ADHD
Rates did differ: 34% had a positive family history of ADHD if the patient had ADHD, but only 13% if he or she did not. The Subtype of ADHD made no difference. In 4% bilineal ADHD was ascertained.
Looking at Table 5, all categories are significantly increased in the TS+ADHD group on 2 × 2 cross-tabulations, except anxiety disorder. TS-associated “pure” ADHD is uncommon, representing only 18% of those diagnosed. One comorbid disorder was reported in 34%, 2 in 28%, and 3 in 13%.
Diagnosed ADHD gradually declines from 65% at age 4 through 48% at age 18 to 32% at age 28. Among the TS+ADHD group, the following significantly decline with age (percentages are, in order, children below 18, then adults 18+): Sleep problems (22 → 18%), anger control problems (39 → 32%), specific LD (33 → 22%), and ODD/CD (21 → 12%). The following increase significantly with age: OCD (21 → 37%), Mood disorder (14 → 35%), Anxiety disorders (16 → 21%), coprophenomena (13 → 22%), SIB (14 → 27%), and Mental Retardation (3.2 → 7.8%). PDD rates undergo no significant changes with age.
(a) Changes in prevalence with age in TS+ADHD (b) Changes in prevalence with age in TS−ADHD
Comorbidity with OCD
It is of interest that OCD is more common in the Combined Subtype (21%) than in the Inattentive Subtype (11%), although this does not reach statistical significance. For OCB (subclinical obsessive-compulsive symptoms), there are no differences between Subtypes to report. About 60% of cases with OCD had a diagnosis of ADHD.
Comorbidity with ODD/CD
Of those diagnosed with ODD or CD, 87% were also diagnosed with ADHD. The Combined Subtype was much more frequent (16% vs. 3%) than the Inattentive Subtype (p = 0.02).
Comorbidity with anxiety disorder
There were no between-subtype differences. There was no increase in the TS+ADHD group. About 56% of anxiety disorder cases were diagnosed with ADHD.
Comorbidity with mood disorder
There was a significant increase in children and a lesser increase in adults.
Comorbidity with LD, MR, PDD, and DCD
There were no differences between subtypes.
There are no between-subtype differences. About 66% have ADHD.
Although differences do not reach statistical significance, rates of PDD were 13% in the Combined Subtype and 5% in the Inattentive Subtype. About 67% of the PDD group were also diagnosed with ADHD.
There are too few adults with this diagnosis and too few among the Subtype subset to report.
Behaviour problems associated with ADHD
Anger control problems
For ADHD Subtypes, there is a predominance of subjects in the Combined (33%) vs. Inattentive (15%) Subtypes (p < 0.01). The 39% rate in TS+ADHD compared with 16% in TS−ADHD is highly significant.
These problems are more frequent in the Combined Subtype (37%) than in the Inattentive Subtype (24%). About 65% of patients with sleep problems were diagnosed with ADHD.
ADHD Subtype makes no significant difference. About 71% of children with coprophenomena have ADHD, and 48% of adults.
Age at onset of SIB in TS+ADHD is 7.4 years (SD 3.9) compared with 10 years (SD 5.0) in TS−ADHD. In OCD, by contrast, there is no difference. (These numbers are still small, only 79 cases from our recent and incomplete project.) SIB is somewhat increased in the Combined Subtype (30%) vs. Inattentive (20%), (p < 0.05). Overall, 62% of SIB cases had ADHD.
Social skill deficits in children
There was a significant and substantial increase in TS+ADHD cases. Deficits were significantly more frequent in the Combined Subtype (47% vs. Predominantly Inattentive Subtype (19%)) (p < 0.001). About 78% of children with this problem have ADHD.
Sexually inappropriate behaviour
There was a significant increase in TS+ADHD. These cases were mostly of the Combined Subtype (7 of 9). About 82% of this group have ADHD.
In trichotillomania (TTM), comparison figures are insignificant in children, and not convincingly different in adults (p < 0.05). However, of seven cases of TTM among the 153 sub-sample, six were in the Combined Subtype (p < 0.05). About 64% of this group have ADHD.
In stuttering (or similar dysfluencies), ADHD Subtype makes no significant difference. About 63% of stutterers had ADHD. There is a modest increase in the TS+ADHD group.
About 87% are right-handed, 10% are left-handed, and 3% are ambidextrous. In TS+ADHD left-handedness in 207 cases constitutes 11%, while in 142 cases of TS−ADHD 8% are left-handed (p < 0.01). There were no significant ADHD Subtype differences.
Binary logistic regression with ADHD as the dependent variable did not identify coprophenomena as significantly increased in TS+ADHD, when controlling for other factors, despite a significant 2 × 2 correlation. Variables retained in the equation were: age at onset and diagnosis of tics (earlier), LD, ODD, child, mood, anger control and social skill problems. The variance explained was 22% and correct classification was 68.0%. Specificity = 93.4%, sensitivity = 31.8%.
The prevalence of ADHD in TS is in line with other studies [20, 29]. Whether ADHD is in fact an “entity” and whether ADHD Subtypes are valid subcategories are beyond the scope of this paper . The relative proportions of the three ADHD Subtypes is consistent with other studies [26, 27]. The sex proportion (more common in males) has been replicated here. The expected shift to more females in the Inattentive Subtype was not found. ADHD comorbidity of 82% compares with 65% in the Goldman et al.  report.
The validity of the ADHD diagnosis was given support in a prospective study of at-risk young children with TS parents by McMahon et al. ; 29% of the children developed a tic disorder, and 41% had ADHD when followed up.
The presence of ADHD is associated with earlier diagnosis of TS by 3.5 years. This may be assumed to be due to the often-disruptive character of ADHD-related symptoms, and seems supported by the earlier diagnosis in the Combined Subtype than in the Inattentive Subtype, in which it occurs fully 2 years earlier (delay from onset of tics of 3.3 years vs. 5.3 years).
All comorbid disorders with the exception of Anxiety Disorder (17% in both groups) were significantly correlated with ADHD (there seems to be no ready explanation for the latter, but Roessner et al.  could show that emotional problems are present in both chronic tic disorders and ADHD, while externalizing behaviour is more closely related to ADHD). Mennin et al.  cited rates of 8–43% in the literature, and recommended using two or more anxiety disorder subtypes for further study of the relationship with ADHD. In particular, anger control problems, sometimes referred to as “rage,” have been shown to be related to comorbid ADHD, not to tics alone [7, 13].
ODD comorbidity here reported to be 19% is low compared with 40% reported by Goldman et al. . The increase in OCD and mood disorders in adult life is not unexpected; anxiety disorder also is substantially increased in females. The decrease in specific LD and ODD may be a function of subjects being out of school or adult psychiatrists and neurologists being less experienced with the diagnosis of these conditions.
Accumulation of additional cases over the next year should provide sufficient data for more robust information on ADHD Subtypes and for details of coprophenomena, SIB and other repetitive behaviour.
Due to lack of interrater reliability measures or operational definitions of behavioural categories and possible recall bias of parents and adult patients, the findings deserve varying levels of confidence. Robust findings are of associations with sex and age first seen.
At the time the project was conceived, diagnostic categories were being used in a somewhat different way. ADHD was included as a category without breakdown into Subtypes. Oppositional-Defiant Disorder was thought to usually lead to Conduct Disorder in adult life, so the two were combined . Anxiety Disorder was not subdivided. These subdivisions have now been added, but the additional information applies to only a subset of the recent data. Developmental Coordination Disorder as a diagnostic category was added later.
Medication for tics and for ADHD cannot be adequately addressed in this type of registry with one-time entry. Comparisons of children and adults may be confounded by the fact that different specialists may be seeing children and adults, with different training and experience.
Tic severity is another variable that has limited utility in this study, probably because it is typically so unstable in childhood, anchor points are not established, and it is likely to be confounded with awareness of behaviour problems and comorbid disorders.
As a registry, the database provides a snapshot view of persons with TS coming to specialist-clinicians, at varying ages. The figures cited in this paper are not lifetime diagnoses. Their absolute levels may not be dependable, but it is felt that patterns within the data, despite differences in sensitivity, may be useful for stimulating further research.
The diagnosis of ADHD is the most commonly made of all. It is significantly more likely in males and in this dataset accounts for much of the sleep and anger control problems that are often described as characteristic of TS. Disruptive behaviour is more common in the Combined Subtype. Coprophenomena, when controlled for other significant variables, are not associated with ADHD.
The lack of significant difference between psychiatrist and neurologist reporting of ADHD at least indicates comparable levels of awareness, though relative sensitivity of the diagnostic threshold cannot be determined from this data. The high comorbidity rate of ADHD (with the exception of anxiety disorders) is a strong indicator that training in recognition and management of these syndromes is essential for all clinicians coming into contact with them. For example, one-third of TS+ADHD cases were given a diagnosis of Learning Disability, 26% had social skill deficits, and 39% presented with problems controlling anger. All three of these should be seen clinically as markers for ADHD investigation. Overall, findings may be seen as providing a strong recommendation for clinicians to develop, or have available, psychoeducational as well as psychopharmacologic skills, and adequate assessment resources.
Reports from clinicians in 27 countries confirm that ADHD is the most common comorbidity in TS, approximating 60%. The expected male excess is also found beyond that already in TS itself. ADHD is itself comorbid. The reported presence of comorbid ADHD accounts for most of the sleep, anger control problems, and SIB that are often reported, other comorbidities adding little. However, this study is the first to report that, when controlled for other factors, ADHD is not a significant factor for coprophenomena, and for social skill deficits, ADHD is only one of several factors. High ODD/CD and LD rates in the TS+ADHD group corroborate previous reports. The ADHD Subtype data is preliminary, but as expected is indicative that the Combined Subtype is associated with more behavioural problems in the areas of anger management and social skills, yet not in sleep problems. Overall, the diagnosis of ADHD is significantly associated with a wide variety of other diagnoses and symptoms, only 18% not having further comorbidity.
It is therefore very clear that the recognition and management of ADHD and related symptoms are essential features of the overall management of TS . The world-wide patterns represented here for the most part confirm previous findings with different methodologies at single sites.
I wish to thank Larry Burd, Ph.D., for his frequent inspiration, and Boris Kuzeljevic (Statistician/Data Manager, BC Research Institute) for helping with logistic regression procedures and their interpretation.