Functional tic-like presentations differ strikingly from Provisional Tic Disorder

Background: Recent years have seen a dramatic increase in new “tic” cases in teens and young adults. These individuals often present with fulminant onset of symptoms not commonly seen in Tourette syndrome (TS) and are often diagnosed with Functional Neurological Symptom Disorder (FND-tic). However, some authors have questioned whether this illness truly differs from typical Provisional Tic Disorder (PTD) and TS. Previous studies have compared FND-tic, usually a few months after symptom onset, to patients with TS, usually years after symptom onset. We sought to test whether the presenting symptoms of FND-tic differ substantially from those in patients at a similar duration of symptoms who are later diagnosed with TS. Methods: This comparative study examines clinical features summarized from published reports of FND-tic with novel data from a longitudinal study of PTD. This study came from a referral center for TS and tic disorders and included 89 children with tics whose first tic occurred a median of 3.6 months earlier, nearly all of whom were diagnosed with a chronic tic disorder at follow-up. Specifically, we examine clinical features identified in a recent literature review as supporting a diagnosis of FND-tic, including symptom characteristics, course, severity and comorbidity. Results: Several clinical features dramatically distinguish the patients diagnosed with FND-tic from those diagnosed with typical PTD. For example, coprophenomena are reported at or shortly after symptom onset in over half of FND-tic patients, whereas even several months after onset, coprophenomena had occurred in only 1 of 89 children with PTD. Six clinical features each have a positive predictive value over 90% for FND-tic diagnosis if prior probability is 50%. Conclusions: These new data provide strong evidence supporting the diagnostic validity of FND-tic as distinct from TS.


Introduction
Recent years have seen a dramatic increase in new "tic" cases in teens and young adults. 1,2 These individuals often present with fulminant onset of symptoms not commonly seen in Tourette syndrome (TS), but often similar to those found in videos tagged as "Tourette" on social media. 3 For instance, echopraxia and coprolalia occur in more than half of these individuals at symptom onset. 3 The nature and characteristics of these symptoms, and the onset age and course of illness have led experienced clinicians to differentiate these cases from TS and to diagnose instead Functional Neurological Symptom Disorder (FND). FND with tic-like symptoms (hereinafter "FND-tic") has been reported previously, but prior to 2019 was considered rare, occurring in < 2% of tic or tic-like cases at five major referral centers from three continents. 4 Some experts have expressed skepticism as to whether a new diagnosis (FND) is needed for these patients, positing that the previous understanding of TS may have been too narrow. 5 Perhaps, for instance, echopraxia is present early in the course of TS much more often than has been recognized. We concur with these authors that clearly differentiating a new diagnosis from existing diagnoses is a key component of diagnostic validity. 6 Substantial published data describe typical clinical features of chronic tic disorders, but almost no prospective data have been published on symptoms in the first year after tic onset. 7,8 This evidence gap is crucial because most FND-tic patients at clinical centers have had symptoms for only a few months (mean 0.4 years in one study 4 ). Here, we directly address this concern by providing for the first time substantial data on relevant clinical features during the first few months after symptom onset in children ultimately diagnosed with TS.

Methods
The New Tics study is a prospective, longitudinal study that enrolled 89 children ages 5-10 years whose first tic occurred in the past 9 months (median 3.6 months). 9 Children are assessed using multiple informants (child, parent, trained interviewer, and observation by an experienced clinician for more than an hour, including by video while the child is alone). The diagnosis in this situation is Provisional Tic Disorder (PTD), and nearly all these children (77 of 79) were diagnosed with TS (70) or a persistent tic disorder (7) when they returned at the one-year anniversary of their first tic. 10 Here we report the prevalence and timing from the New Tics sample of various features that occur commonly in FND-tic patients.
The feature list was drawn from a recent review of FND-tic, a narrative review that included all primary data publications on FND-tic known to its authors as of August, 2022. 3 The comparison data for FND-tic patients comes from 26 published reports, with pertinent data in 17 reports 4,11-26 describing a total of 336 patients (data file available as Underlying data 27 ). For quantitative variables, the weighted mean is provided (weighted by N in each report), along with the median and range of the summary values reported in each relevant publication.
Most of these features thought to suggest FND-tic were recorded prospectively in the New Tics study, including age, sex, premonitory urges, tic suppression, coprophenomena, and family history. However, since the New Tics sample was enrolled almost entirely before the recent FND-tic upsurge, some of these clinical features were recorded indirectly. For instance, to match "severe symptoms at onset," defined differently in various reports on FND-tic, we conservatively chose from the New Tics sample all patients with emergency department visits or disability prior to the screening visit, or REVISED Amendments from Version 1 In this revision, we address the thoughtful comments of the reviewers. Specifically: * In the discussion, we expand on the limitations of this study, including likely differences between groups in ascertainment bias.
* We add statistical tests to Table 1 and add sensitivity and specificity as new columns in Table 2. * We supply new, individual participant data for the PTD sample as supplementary files added to the same dataset referenced in version 1 (doi 10.17605/OSF.IO/XZ328).
* We add context for the test statistics including the reliance of PPV on prior probability. * We discuss the applicability of our results to the situation in which a child presents with both typical-appearing tics and apparent functional tic-like symptoms.
* We note that features other than those studied here are likely important to diagnosing FND-tic.
Any further responses from the reviewers can be found at the end of the article a high score on the Yale Global Tic Severity Scale's 28 impairment item at screening. Details on other such choices are given in footnotes to Table 1.

Data analysis
Fisher's exact test was used to find the probability of differences in frequency of features between the two populations (fisher_exact from SciPy (RRID:SCR_008058) 1.9.1). 29

Ethical considerations
The New Tics study was approved by the Washington University Human Research Protection Office (IRB, protocol numbers 201109157 and 201707059), all participants assented to participation, and a parent or other legal guardian provided written documentation of informed consent.

Results
Stark differences in presentation distinguish the FND-tic patients from typical PTD (Table 1). For example, coprophenomena are reported at or shortly after symptom onset in 59% of FND-tic patients. By contrast, coprophenomena had occurred in only 1 of 89 children with PTD at an average of 3.6 months after tic onset. Similarly, the TS International Database Consortium found that only 2% of TS patients in tertiary centers retrospectively reported coprophenomena at symptom onset, and only 20% ever manifested coprophenomena by an average of 5 years after tic onset. 30 Movements or vocalizations that were dramatically worse in the presence of others versus when alone occurred in 47% of FND-tic patients, but in none of the New Tics PTD sample. Symptoms dramatically and persistently disrupted intended actions in 65% of FND-tic patients, but in only 2% of PTD. The prevalence of prolonged tic attacks was 65% in FND-tic, but 0-2% in PTD. Other features that differed substantially include lack of premonitory urges (53% vs. 4%) and severe symptoms at onset (48% vs. 3-6%). Table 1 provides details on these comparisons and includes statistics on a dozen more clinical features of FND-tic that differ from the New Tics PTD sample.   Table 2 illustrates the clinical value of these comparisons, viewing each clinical feature as a "test" for FND-tic, with its associated sensitivity and specificity. We also provide the positive predictive value (PPV) for each feature thought to indicate FND-tic. The PPV represents how confident one can be of a FND-tic diagnosis given the presence of the listed feature. The PPV depends on the prior probability of the FND-tic diagnosis, or its likelihood before one knows whether a given patient has that feature. We provide PPV for each feature based on two prior probabilities, 2% and 50%. The first, 2%, represents the approximate rate before the pandemic of functional tic-like symptoms at clinical referral centers. 4 Clinical equipoise about a given participant's diagnosis is represented by a prior probability of 50%, for instance if one knows that a referring clinician is ambivalent about whether the patient has PTD or FND-tic, but one has not yet read the chart nor seen the patient. Features thought to be less common in FND-tic are listed with a negative predictive value (NPV), equivalent to PPV for the absence of the given feature.
In a patient with recent onset of tics, coprophenomena at onset, or any one of the other features named above, raises the probability of a non-TS diagnosis from 50% (as when the clinician is ambivalent about the diagnosis prior to considering this feature) to over 90% (Table 2). Other features differ significantly but are less useful diagnostically. For instance, obsessive compulsive disorder (OCD) is more than three times less common in FND-tic than in PTD (p<.0001), but its absence only raises the probability of FND-tic from 50% to 57%.

Discussion
We demonstrate conclusively that patients with functional tic-like symptoms differ notably from typical tic patients at the same stage of the disorder, namely in the first few months after symptom onset. Previous reports have compared FND-tic to TS, 3,31 but not to a large PTD sample. We also provide for the first time quantitative estimates of the diagnostic significance of individual clinical features previously suggested to indicate FND-tic. This approach addresses a current debate 5,32 by showing that some features are significantly more common in patients diagnosed with FND-tic, yet individually do not substantially raise the likelihood of an FND-tic diagnosis.
The primary concern with the validity of these conclusions arises from the different potential sources of ascertainment bias in the two groups. Fortunately, these differences may not be as problematic as one might suppose. The FND-tic group is older, and if one's information were limited to this study alone, one might posit that the natural history of tic disorder included different early symptoms at different ages of tic onset. However, the literature includes decades of previous clinical information on typical tic disorders. Retrospective studies of TS and a prospective study of PTD in siblings of TS probands all found peak tic onset before age 10 years old; adult onset of tics is quite uncommon. 7,8,33 The PTD cases were volunteers willing to participate in a rather intensive research study, and were ascertained by a variety of methods (advertising and referrals from clinicians being the most common) rather than by clinical care-seeking. However, at the screening visit for the study, over 60% of them had sought clinical care for the tics or were planning to. The two FND-tic reports from Calgary, Alberta (child and young adult) were from a prospective registry of all patients seen for tics at their center, which is the only specialty center in their region. FND-tic cases were defined by "rapid onset of complex tic-like behaviors, with escalation to peak severity within hours to days." In other words, these reports are selected only for being seen at a specialty center for tic-like phenomena. Yet their clinical features are similar to those of the FND-tic literature overall (or even more different from the PTD group). For instance, 95% of the Calgary FND-tic child patients were of female sex, only 25% had ADHD, and only 5% had OCD. We can conclude that potential bias from selecting more interesting, severe or classic cases for published case series does not substantially alter our results. Perhaps most importantly, there are no larger published sources for data about FND-tic nor about PTD. Nevertheless, ideally the results presented here should be confirmed in a new, independent sample.
We note that 22% of the patients reported in the FND-tic papers in fact had a prior history of typical tics, and many of them exhibited both tics and functional tic-like symptoms at the time they presented due to the latter. We have structured our results around whether FND-tic can be diagnosed, not whether PTD can also be diagnosed. In other words, when a patient presents with a typical history for TS but also new symptoms, the features above would allow more confidence in diagnosing FND-tic in addition to TS.
Other features in addition to those studied here may also be important for diagnosis. For instance, exposure to tic-like symptoms on social media was a common feature discussed in the papers discussing FND-tic in the past few years. 3 Unfortunately, we have no prospectively collected data from the PTD group on exposure to others with similar symptoms.
The data presented here do not prove the etiology of the tic-like symptoms diagnosed in the cited reports; hence the limited claim that these symptoms represent a different illness than PTD/TS. However, the marked difference in presentation these data demonstrate is an important argument adduced in the cited reports to support the diagnosis of functional neurological symptom disorder. Diagnosing FND-tic is important, since to the extent of our current knowledge, its prognosis and optimal treatment differ from those of TS. 3 In conclusion, these new clinical data about the first few months after tic onset prior to diagnosis of TS provide strong evidence supporting the diagnostic validity of functional tic-like symptoms as distinct from PTD and TS.

Data availability
Underlying data Information on and individual subject data from 'The New Tics Study: A Novel Approach to Pathophysiology and Cause of Tic Disorders' can be found at NIH RePORTER (Project Number 1R01MH104030-01A1) and at the NIMH Data Archive. Open

Open Peer Review
The authors, all highly experienced 'Tourettologists', have approached this topic in an interesting and novel manner to try and provide some evidence for the ability to differentiate functional tic like behaviours from those seen in provisional tic disorders, usually related to Tourettes at follow up. This is a complex area, with some researchers proposing some shared mechanisms and overlapping neurobiology in the presentations. This paper uses some usefully collected detailed (although some retrospective) clinical data of children with PTD within a few months of presentation to compare characteristics demonstrating some useful distinguishing features. The presentation of the statistics could benefit from a little more expansion and explanation, especially for the non-expert, as it seems an important aspect of this paper. For example, table 2 is not that clearly explained and could do with more detail.
There are some limitations, but these have been clearly highlighted by the authors, such as the difference in ages of the two groups. I therefore would support this paper as it represents a useful contribution to the ongoing exploration of the functional tic-like phenomenon. I do wonder whether the dichotomy of subjects is a problem and would welcome wider discussions by the authors, on the situations seen commonly in the clinic in which young people, demonstrate both Tourette-related tics and functional tic-like behaviours together. This paper raised a question in my mind, as to whether the authors could have included distinguishing factors for the concept of provisional or transient functional disorder, especially when appearing in the context of exposure to social media or influencers.
Thank you for an interesting contribution.

If applicable, is the statistical analysis and its interpretation appropriate? Partly
Are all the source data underlying the results available to ensure full reproducibility? Yes Are the conclusions drawn adequately supported by the results? Partly 2. I do wonder whether the dichotomy of subjects is a problem and would welcome wider discussions by the authors, on the situations seen commonly in the clinic in which young people, demonstrate both Tourette-related tics and functional tic-like behaviours together.
We agree the discussion did not address this issue previously. The discussion now includes this additional paragraph: "We note that 22% of the patients reported in the FND-tic papers in fact had a prior history of typical tics, and many of them exhibited both tics and functional tic-like symptoms at the time they presented due to the latter. We have structured our results around whether FNDtic can be diagnosed, not whether PTD can also be diagnosed. In other words, when a patient presents with a typical history for TS but also new symptoms, the features above would allow more confidence in diagnosing FND-tic in addition to TS." 3. This paper raised a question in my mind, as to whether the authors could have included distinguishing factors for the concept of provisional or transient functional disorder, especially when appearing in the context of exposure to social media or influencers.
The reviewer raises a very interesting question that appears to include two concepts: diagnostic utility of exposure to social media and nosology.  )). Unfortunately, we have no prospectively collected data on this question from the PTD group. Anecdotally, I can report that although over a third of our New Tics study participants had a parent or sibling with tics, only once or twice was there any suggestion that the child's first tic may have appeared to be prompted by the family member's similar tic. We have added to the discussion the following paragraph: "Other features in addition to those studied here may also be important for diagnosis. For instance, exposure to tic-like symptoms on social media was a common feature discussed in the papers discussing FND-tic in the past few years. 3 Unfortunately, we have no prospectively collected data from the PTD group on exposure to others with similar symptoms." b) Traditionally, functional neurological symptoms have not been diagnosed differently based on duration of the defining symptoms. The closest analogy I can think of is Briquet's syndrome (and its dramatically modified descendants such as DSM-IV somatization disorder), which required not only atypical, medically unexplained symptoms starting before midlife but also numerous symptoms over time, affecting a variety of organ systems. Personally I would not favor creating a new diagnosis to represent "presumed functional symptoms in the first year since onset". Done.
3. I think in the conclusion the authors should discuss the importance of testing the potentially meaningful differentiators in an independent dataset.
We agree. The discussion now includes: "ideally the results presented here should be confirmed in a new, independent sample."

It would also be very worthwhile in further studies to use individual participant data to determine the PPV of multiple predictors to engage in a diagnostic distinction.
That's an excellent idea. We have supplied individual participant data for the New Tics group as a supplementary file.

5.
If possible, a histogram exploring age and gender differences between the two groups would be helpful.
We do not have individual participant data for the FND-tic group (which came from 17 different publications), so we can't do this.

Competing Interests:
No competing interests were disclosed.