Abstract

There is clear overlap in the clinical symptoms of obsessive-compulsive disorder (OCD) and Tourette syndrome (TS). As a result, OCD (with or without tics) and TS (with or without obsessive-compulsive symptoms [OCS]) have been conceptualized to form a disorder spectrum–an overlapping set of phenotypes reflecting presumed commonality at the level of the underlying genetics and neuropathology. We identified the characteristics of a research sample of patients with OCD or TS, based on information obtained in semistructured clinical interviews, to examine the similarities and differences in the clinical symptoms across this spectrum. This sample conformed to known age-of-onset and sex distribution patterns for OCD and TS. Previously reported patterns of predominant aggressive and sexual obsessions and touching compulsions were observed in subjects with tic-related OCD, compared with non-tic-related OCD (ie, OCD alone). The majority of patients with tic-related OCD experienced horrifically violent obsessions that were less common in OCD alone and much less common in TS. Nonetheless, symptomatic and functional impairment in TS subjects was clearly related to the intensity of their OCS. The specific obsessions and compulsions associated with clinical impairment in TS differed from those associated with impairment in OCD. These results suggest that, despite the many overlapping dimensions of these disorders, the symptoms and associated impairment in "pure" OCD, tic-related OCD, and TS do not form a simple continuous spectrum.

Introduction

Studies of the neurobiology of obsessive-compulsive disorder (OCD) and Tourette syndrome (TS) have been catalysts for hypotheses about brain circuitry, the genetics of mental illness, neuroimmune processes, and mind-brain interactions. While the conceptual linkage of OCD and TS has clearly stimulated models for understanding the brain, the specific relationship between these disorders at the clinical, neurobiological, and genetic levels remains somewhat elusive. Family studies have provided the greatest support for the notion that OCD and TS are fundamentally linked, perhaps even different phenotypic expressions of the same genetic diathesis.^1,2 These studies have shown increased rates of TS among OCD probands and family members, and vice versa. In fact, if we believe that the prevalence of OCD is 2%³ and the prevalence of comorbid TS among these individuals is approximately 7%,⁴ then we are faced with the complex notion that the number of individuals with comorbid TS and OCD exceeds the total number of individuals with TS of any form, which is estimated to be approximately 1 in 1,000.^5,6

Intuitive evidence for an OCD-TS linkage comes from the perceived interchangeability of sensory, cognitive, and motor elements of these disorders.⁷ Indeed, this clinical impression of phenomenologic fluidity has driven many of our models for an OCD-TS spectrum.⁸ Within these models, tic-related OCD (ie, OCD accompanied at some point by tics) has received attention as a potential intermediate phenotype between OCD and TS.^9,10 This possible hybrid form of OCD shares some features of TS (ie, tics); may respond optimally to regimens that include medications that are effective for TS, but not OCD (eg, dopamine antagonists)¹¹; and may have epidemiologic features (eg, age of onset, sex distribution) that are midway between OCD and TS.¹² Other observations suggest that tic-related OCD is symptomatically distinct from "pure" OCD. Individuals with tic-related OCD are reported to be more likely to experience aggressive, religious, and sexual obsessions, as well as checking, counting, hoarding, and touching compulsions, compared with people with pure OCD.^13-16

If there is a continuous OCD-TS phenotypic spectrum, there are obvious implications for genetic models and for the logistics of selection criteria for genetic and other studies of these disorders. Furthermore, continuity at a clinical and genetic level might be accompanied by important similarities at the level of the neurobiology of these disorders. In this case, the observed neurobiologic differences between OCD and TS might reflect artifacts of their phenotypes (ie, neurochemical consequences of anxiety secondary to obsessions, or of movements involved in rituals versus tics), while common features in their neurobiology may be more associated with a primary common basis for this spectrum.

We compared clinical profiles among patients with OCD (pure and tic-related) or TS (with and without comorbid OCD) who participated in ongoing studies in our facility. This review yielded support for findings from previous studies regarding specific symptoms associated with pure vs tic-related OCD. Overall, there was substantial evidence for discontinuities, in addition to continuities, among the clinical profiles of patients with OCD or TS. While these findings underscore areas of clinical overlap between these disorders, they do not support the model of a simple clinical spectrum relationship between OCD and TS.

Methods

Study Participants and Interviews

Subjects with OCD (n=103) or TS (n=61) were recruited from local advertisements to participate in research studies at the University of California, San Diego, Medical Center. Adult (>18 years of age) OCD subjects (n=103) were administered subsections of the Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental Disorders, Third Edition-Revised,¹⁷ by two trained interviewers (N.S., H.H.) to assess the presence of psychotic disorders and to confirm a diagnosis of OCD. OCD subjects were also administered the Yale-Brown Obsessive Compulsive Scale (YBOCS)¹⁸ to determine the type and severity of obsessions and compulsions. For all YBOCS interviews, instructions were reviewed with the subject, the subject completed the YBOCS checklist, each checklist item was then reviewed with the subject by the interviewer, and when appropriate, the content of the symptom was documented on the checklist via the use of direct quotes from the subject. Adult TS subjects (n=50) were administered the YBOCS and Yale Global Tic Severity Scale (YGTSS)¹⁹ by trained interviewers (N.S., S.Z., R.F.) to characterize the type and severity of their tic symptoms, as well as their global impairment level. Clinical information from TS children (n=11) was obtained via the age-appropriate versions of the Children's YBOCS (C-YBOCS) and the Children's YGTSS (C-YGTSS). Diagnostic assessment in children was also aided by the use of the Diagnostic Interview Schedule for Children-Revised (DISC-R, version 2.3, modules A-E).²⁰

Subjects were excluded from participation at the point of initial communication (screening phone interview) or upon questioning on the day of interview for a history of substance abuse or dependence, recreational drug use within 1 month prior to testing, schizophrenia in the subject or in a first-degree relative, head injury with loss of consciousness, severe medical illness, electroconvulsive therapy (ECT), or pregnancy. Most (84%) OCD and TS subjects underwent urine toxicology screening on the day of the interview, and results verified no recreational drug use. (In the remaining 16% of subjects, urine testing was not completed, often due to contamination obsessions; their results were examined separately and found to be representative of the urine-tested subjects.) Upon completion of the above clinical assessments, subjects participated in one of several psychophysiologic studies. Data from these measures have been reported elsewhere,^21-25 but no report of the detailed clinical assessments of these participants has been published.

Using information obtained from the YBOCS interview, subjects were categorized based on the lifetime presence or absence of symptoms in any one of the major YBOCS subgroups of obsessions (aggressive, contamination, sexual, hoarding, religious, symmetry, and miscellaneous) or compulsions (cleaning/washing, checking, repeating, counting, ordering, hoarding, and miscellaneous). Based on published reports of the clinical importance of specific symptoms,^13-16 subgroups were analyzed in greater detail by categorizing obsessions or compulsions based on the presence or absence of specific content (eg, aggressive symptoms with specific violent content, aggressive images, nonviolent intrusive images, touch/tap/rub compulsions, etc). In a small number of cases, clear categorization of symptom content could not be made (ie, type of violent act was not clearly identified); therefore, data were omitted from subsequent analyses (ie, not recorded as either present or absent). Subjects were also categorized on the basis of specific clinical variables, such as self-reported age at the time of first obsessive-compulsive symptoms (OCS) or TS symptoms, and a positive or negative history of tics among OCD subjects.

Among the 103 subjects with OCD, 18 were identified as having a clear history of tics (chronic motor tics, chronic vocal tics, or both), but did not meet diagnostic criteria for TS (first tics after age 21 years, or only one type of tic [motor or vocal]), and were categorized as "OCD + tic." In 39 subjects, a tic history was categorized as "equivocal" (subject unable to recall, or description of symptoms uncharacteristic of tics), and these subjects were categorized as "OCD ? tic." The remaining 46 OCD subjects stated unequivocally that they had never experienced tics, and were categorized as "OCD alone." Two methods were used to categorize TS patients. Subjects with a diagnosis of OCD (obsessions and/or compulsions, and scores >1 on YBOCS items 1, 2, 3, 6, 7, or 8) were categorized as "TS + OCD" (n=29). Those with significant OCS (C-YBOCST or YBOCST>=15) were categorized as "TS + OCS" (n=13), vs those without significant OCS (C-YBOCST or YBOCST<15), who were categorized as "TS - OCS" (n=19). Based on this information and the semistructured interview, subject subgroups included tic-related OCD (OCD + tic; n=18), non-tic-related OCD (OCD alone; n=46), OCD without certain tic history (OCD ? tic; n=39), TS + OCD (n=29), and TS alone (n=32).



Statistics

Comparisons were designed to assess the relationship between OCD symptoms, their intensity, functional impact, and content across the specific clinical subgroups. The hypothesis was that a continuous disorder spectrum would be reflected by continuities (ie, conserved patterns) in the clinical features of OCD across these subgroups. The frequency of specific OCD symptom categories and symptom content were compared across the four diagnostic subgroups using chi-squared tests. Continuous variables, including YBOCS and YGTSS scores and subscores and age at symptom onset, were compared across the diagnostic subgroups and across groups characterized by the presence or absence of specific OCD symptoms using an analysis of variance. Simple regression analyses were used for correlations among continuous variables. The a level was 0.05; thus, P<=0.05 was considered significant.

Results

Demographics

FIGURE 1. Yale-Brown Obsessive Compulsive scale (YBOCS) scores (mean ± standared error of the mean [SEM]) in obsessive-compulsive disorder (OCD) subgroups (OCD alone, OCD ? tic, and OCD + tic) and Yale Global Tic Severity scale (YGTSS) scores (mean ± SEM) in Tourette syndrome (TS) specific obsessions and compulsions. Statistical differences are indicated at right, based on one-factor analysis of variance.

Age, sex, age at first symptoms, handedness, and employment status are shown in Table 1. Among OCD subjects overall, the equal representation of males and females is consistent with previous reports.²⁶ TS subjects overall were predominantly male; the ratio of 3.07 males to 1 female with TS (M:F overall TS group=46:15) is also comparable with that reported elsewhere.27-30 Age at symptom onset in both OCD and TS groups was also consistent with our understanding of these disorders.29,30 Both the sex ratio (M:F=2.6:1) and the mean age at first OCD symptoms (9.39 years) for the OCD + tic group were intermediate values between those determined for OCD alone and TS groups. Interestingly, the latter comparison primarily reflected the early age of first symptoms in females among the OCD + tic group (6.40 ± 1.95 mean ± standard deviation [SD]). Neither sex ratios nor age at first symptoms distinguished TS subjects with OCD (or OCS) from those without OCD (or OCS). Unemployment tended to be greater in OCD groups vs TS groups; among TS subjects, unemployment was more than 2.5 times higher among subjects with comorbid OCD compared with those without OCD.

Clinical Ratings and Impairment

Clinical rating-scale scores are seen in Table 2. Because both YBOCS and YGTSS ratings are disorder-specific by design, YBOCS data for participants with TS alone are reported only as a point of reference. Among OCD subjects, no significant differences in YBOCS total or subscale scores were observed between the OCD alone vs the OCD + tic groups. TS + OCD subjects had lower YBOCS scores compared with all other OCD groups. Compared with TS alone subjects, TS + OCD subjects scored significantly higher on the YGTSS total scale (F=12.74; df 1,56; P<0.001), accounted for largely by greater YGTSS impairment scores in these subjects (F=14.39; df 1,56; P<0.0005), with only trends toward greater scores in motor (F=2.87; df 1,56; P<0.10) and phonic (F=3.50; df 1,56; P<0.07) subscales. Similar patterns were evident when TS subjects were categorized on the basis of YBOCS >=15. A 3 x 4 regression matrix yielded significant correlations between YGTSS total score and YGTSS impairment score with YBOCS total score, obsession subscale score, and compulsion subscale score among individuals with TS (Table 3). Interestingly, the relationship between YGTSS impairment scores and these measures of OCS severity was generally as robust as the relationship between YGTSS impairment and YGTSS motor (R=0.46), phonic (R=0.36), or total tic (R=0.50) scores. Thus, while the presence of tics was not associated with a more "malignant" form of OCD, based either on symptom ratings (YBOCS) or one functional measure (employment), the presence of obsessions or compulsions in TS was associated with a more severe condition with clear impairment in both symptom and functional domains.

Additional evidence for a relationship between YBOCS elevation and functional impairment in TS subjects is the comparison of YBOCS scores among TS adults who were employed vs unemployed (mean [standard error of the mean (SEM)] YBOCST employed vs unemployed=9.15 [1.45] vs 18.75 [5.30]; F=5.79; df 1,23; P<0.025). Elevations in both obsession and compulsion YBOCS subscales were associated with unemployment in TS subjects, but interestingly, neither motor, phonic, nor total tic subscores of the YGTSS differed significantly between employed vs unemployed TS subjects (F=1.83, 2.05, and 2.38, respectively; P=not significant). In other words, as measured by the YGTSS impairment subscale and by a critical functional measure (employment), impairment in TS subjects was closely linked with the intensity of OCS. This relationship was at least as strong, and perhaps stronger, than the link between functional impairment and tics in these subjects.

Symptom Features Across Subgroups

The percentage of subjects from each major diagnostic subgroup who endorsed the lifetime presence of specific obsessions and compulsions is described in Table 4. Compared with OCD alone subjects, more OCD + tic subjects endorsed sexual obsessions ([chi]²=5.07; P<0.025) and touch/tap/rub compulsions ([chi]²=15.13; P<0.0001). The prevalence of aggressive obsessions overall did not differ statistically among the OCD alone (86%) and OCD + tic groups (100%) ([chi]²=2.80; P<0.10). This likely reflected a "ceiling effect," as analysis of the relatively more severe aggressive obsessions revealed these to be much more common in OCD + tic subjects.

To increase the sensitivity of this comparison, aggressive obsessions were categorized according to their specific content in a manner that formed an intensity gradient with the following obsessions: (1) any aggressive obsession; (2) aggressive obsession other than embarrassment/blurting concerns; (3) a specific violent act; and (4) "worst-case scenario" (eg, stabbing, maiming, or killing, etc). Intrusive violent images were also categorized separately, but their content severity ranged from mild (eg, image of hitting mother) to severe (eg, mutilation scene). When assessed along this intensity gradient, OCD + tic subjects were significantly more likely to have experienced aggressive obsessions other than embarrassment/blurting concerns ([chi]²=6.24; P<0.015) and aggressive obsessions in which a violent act was involved ([chi]²=7.70; P<0.006), particularly when it included the worst-case scenario ([chi]²=12.42; P<0.0005). OCD + tic subjects were also marginally more likely than OCD alone subjects to report intrusive violent images ([chi]²=4.95; P<0.03), and trends were observed for OCD + tic subjects to report more nonaggressive images, as well as blink/stare compulsions, consistent with previous reports.^13,14 OCD alone and OCD + tic groups did not differ significantly in the likelihood of experiencing any other obsessions or compulsions (see Table 4).

To examine the other side of the OCD-TS spectrum, obsessions and compulsions were compared among the OCD + tic group and the TS + OCD and TS + OCS groups (see Table 4). Compared with TS + OCD subjects, OCD + tic subjects were more likely to have experienced any aggressive symptom ([chi]²=7.55; P<0.01), at any level of severity: aggressive other than embarrassment/blurting ([chi]²=13.53; P<0.0005), aggressive + violent act ([chi]²=11.85; P<0.001), and aggressive + stab/maim/kill ([chi]²=13.46; P<0.0005). These groups did not differ in the prevalence of intrusive violent images ([chi]²<1). Similar patterns were evident in the comparison between OCD + tic and TS + OCS groups (not shown). Examples of aggressive obsessions in OCD + tic and TS + OCD subjects are seen in Table 5. Compared with TS + OCD subjects, OCD + tic subjects were also more likely to have endorsed religious obsessions ([chi]²=4.04; P<0.05), cleaning/washing compulsions ([chi]²=6.53; P<0.02), counting compulsions ([chi]²=6.52; P<0.02), intrusive nonviolent sounds or images ([chi]²=5.54; P<0.02), and touch/tap/rub compulsions ([chi]²=3.82; P=0.05).

To examine the possibility that some of the differences in obsessive-compulsive (OC) symptoms between OCD + tic and TS + OCD subjects simply reflected the greater severity of OCD (as measured by higher YBOCS scores) in OCD + tic subjects, subgroups of 10 OCD + tic and 10 TS + OCD subjects were matched for YBOCST scores (mean [SEM] YBOCS: OCD + tic=20.3 [1.13]; TS + OCD=19.4 [1.41]) and the prevalence of obsessions and compulsions were compared between these subgroups. Despite the reduced sample sizes in this subgroup comparison, significant group differences persisted among the aggressive obsessions, which remained more common among OCD + tic subjects: aggressive other than embarrassment/blurting ([chi]²=4.92; P<0.03), aggressive + violent act ([chi]²=7.29; P<0.007), and aggressive + stab/maim/kill ([chi]²=8.56; P<0.005). All other symptom differences noted in the unmatched group comparisons either disappeared or were reduced to "trend" levels among these matched groups. Thus, while some differences in OCD symptoms between OCD + tic and TS + OCD groups may reflect the greater severity of OCD in OCD + tic subjects, the distinct profile of aggressive symptoms appears to reflect a categorical difference between these two types of illnesses.

The lifetime prevalence of specific obsessions and compulsions was compared in TS + OCD vs TS alone groups (see Table 4). Because these groups were distinguished by the presence of an OCD diagnosis, it was anticipated that most obsessions and compulsions would be relatively more common in the TS + OCD group. Surprisingly, only four OCD symptoms were significantly more common in the TS + OCD group: sexual obsessions ([chi]²=17.27; P<0.0001), checking compulsions ([chi]²=8.79; P<0.005), repeating compulsions ([chi]²=5.46; P<0.02), and touch/tap/rub compulsions ([chi]²=4.95; P<0.03). Similar patterns were evident when TS subjects were categorized on the basis of YBOCS >=15.

A rank order of most to least common obsessions and compulsions among the major diagnostic subgroups was tabulated. While aggressive obsessions of all severities were among the most common symptoms in OCD + tic subjects, the most common symptoms in TS + OCD subjects were checking, touch/tap/rub compulsions, and repeating compulsions, with the most extreme violent obsessions being the least common symptom in this group. Spearman Rank Order Correlation of symptoms revealed significant correlations in comparisons of OCD alone vs OCD + tic groups (R=0.63, P<0.01), and of TS + OCD vs TS alone groups (R=0.63, P<0.01), but not of OCD + tic vs TS + OCD (R=0.35, P=NS) (see Table 4).

Obsessions and compulsions associated with relatively more severe symptom rating scores among the major diagnostic subgroups are seen in Figure 1. Among OCD subjects as a whole, elevated YBOCS scores were seen in subjects endorsing the lifetime presence of symmetry obsessions, washing/cleaning compulsions, or repeating compulsions. Interestingly, aggressive obsessions of any severity were not associated with elevated YBOCS scores in OCD subjects. In sharp contrast, among TS subjects, elevated YGTSS scores were seen in subjects endorsing a lifetime presence of aggressive obsessions of any kind; YGTSS scores increased in proportion to the degree of aggressive content (see Figure 1). Elevated YGTSS scores in TS subjects were also associated with a history of sexual obsessions or repeating or counting compulsions. Intuitively, one might predict that increased YGTSS scores associated with violent or sexual obsessions might primarily reflect elevations on the impairment subscale, while YGTSS elevations associated with counting and repeating obsessions might reflect elevated scores on the phonic and motor subscales, respectively (ie, because of the phonic component of counting, and the motor component of repetitive movements). However, this was not the case. A history of aggressive or sexual obsessions in TS subjects was associated with elevated scores on all YGTSS subscales, including motor and phonic subscales (aggressive obsessions: motor–F=11.68; df 1,53; P<0.002, phonic–F=5.30; df 1,53; P<0.03; sexual obsessions: motor–F=5.77; df 1,56; P<0.02, phonic–F=4.23; df 1,56; P<0.05). A history of repeating compulsions was associated with elevated scores on YGTSS phonic (F=5.02; df 1,56; P<0.03) and impairment subscales (F=5.49, df 1,56; P<0.025), but not on the motor subscale, while a history of counting compulsions was associated only with elevations on the impairment subscale (F=6.74; df 1,56; P<0.015). Thus, there is little overlap between the types of obsessions and compulsions associated with the more clinically malignant forms of OCD and TS.