Introduction
Tic disorders (TD) are complex neuropsychiatric disorders with onset before the age of 18 years, characterized by the presence of repetitive, involuntary, nonrhythmic, sudden movements, or vocalizations that can involve discrete muscle groups [
1,
2]. According to the DSM-5 [
3], TD encompass provisional tic disorder (PTD), chronic motor or vocal tic disorder (CTD), and Tourette syndrome (TS). PTD is considered when individuals have at least one motor tic and/or vocal tic, with a disease duration of less than one year. CTD is diagnosed when individuals have single or multiple motor or vocal tics, but not both appear at the same time during the course of the disease, with a duration of longer than one year. TS is diagnosed when individuals have multiple motor tics and one or more vocal tics, which may not appear at the same time, with a disease duration of more than one year. Tics tend to follow an unpredictable waxing and waning pattern over time, and could persist into adulthood [
4]. Although the typical natural history is of improvement or remission over time, TD are frequently associated with a variety of comorbid problems whose negative effects may exceed those of the tics themselves [
2,
5]. About 85–88% of individuals with TS have been reported at least one comorbid disorder, with the most common being attention-deficit/hyperactivity disorder (ADHD) and/or obsessive–compulsive disorder (OCD) [
6‐
8]. Other disorders such as anxiety/depression disorders, learning disorder, oppositional defiant disorder, disruptive behavior disorders, externalizing disorders and autism spectrum disorders are also observed [
9,
10]. Among these, the presence of comorbid behavioral problems is very common and can cause significant adverse effects on quality of life, and should therefore be considered in such patients [
11,
12].
Behavioral problems include, but are not restricted to, withdrawal, hyperactivity, aggression, disruptive behavior, depression, and schizoid [
11]. TD and comorbidities typically present deficits in inhibition, characterized by inattention, hyperactivity, aggression, obsessive–compulsive and other behavioral problems [
13,
14]. Biological, psychological, and socio-environmental factors can contribute to the occurrence and persistence of TD and comorbidities. Previous studies have confirmed several factors, including genetic factors, parental psychiatric disorders, prenatal and perinatal epigenetic factors, family structure, poor parental relationships, and abnormal immune responses, are linked to TD development [
9,
15‐
17]. However, most of those published studies were limited to investigating one factor or one category of factors, and there is lack of a predictive model specifically designed for behavioral problems in TD patients and the ability to draw conclusions about the most important predictors in an inclusive model and the combined impact of these predictors. Furthermore, to date, no published studies have yet developed practical predictive tools to examine the impact of sociodemographic and clinical characteristics. Knowledge of the impact of sociodemographic and clinical characteristics on behavioral problems would be helpful to clinicians in tailoring treatment interventions, and ultimately improve the quality of care for children with TD. Nomograms have been used to provide individualized evaluation of the clinical event incidence on many occasions and as a reliable statistical tool to create a simple intuitive graph to quantify the risk of a clinical event [
18‐
20]. It is typically constructed based on multivariate regression analysis and transforms complex regression equations into visual graphs, to exhibit the combined impact of variables in the prediction model. To develop a model for predicting the risk of behavioral problems in TD patients and to provide a quantitative predictive tool for early clinical screening of individualized risk of TD with behavioral problems have high clinical application value.
In this study, we used a hospital-based database to explore the sociodemographic and clinical characteristics of children with TD and behavioral problems under real-world conditions, relative to children with TD without behavioral problems. We aimed to develop a nomogram prediction model based on independent predictors to examine the impact of sociodemographic and clinical characteristics on behavioral problems in children with TD.
Discussion
Here, we describe the sociodemographic and clinical characteristics of children with TD and behavioral problems in a hospital-based population of children aged 4–16 years. Real-world studies are required to achieve a comprehensive understanding of the trajectories and management of this long-term condition. Herein, we developed a model to predict the individualized risk of behavioral problems among children with TD based on sociodemographic and clinical characteristics, and applied a nomogram to exhibit the combined impact of variables in the prediction model. The identified predictors were as follows: age 12–16 years, abnormal birth history, parenting pattern of indulgence, parent/close relatives with psychiatric disorders, CTD/TS and moderate/severe tic severity. The model showed good calibration and discrimination and was eligible for clinical practice according to the satisfactory results of the C-index, AUC of the ROC curve, DCA and CIC.
Behavioral problems were identified in 30.32% of the TD cases using the CBCL scale in this study. This rate was much higher than the prevalence of behavioral problems among primary and middle school students in Hubei province [
27]. The relationships between individual factors, family factors, clinical characteristics and behavioral problems in children with different medical conditions were inconsistent [
28‐
30]. However, sex effect was more exactly in surveys [
31,
32]. This study sample contained a much higher proportion of males than females (male-to-female ratio close to 4:1 in our study), which is consistent with previous studies [
1]. The gender differences have also been well documented in neurodevelopmental disorders in previous studies [
33,
34]. In this study, male patients had a higher rate of behavioral problems than female patients, and the rate gradually significantly increased with age in the male patients, whereas the opposite trend was observed in the female patients. This may be related to sexual dimorphism in the maturation of neural networks. Previous studies have also reported that elevations in emotional reactivity and reward processing follow an inverted U-shape in terms of onset and remission, with a peak occurring during adolescence. Furthermore, sex-dimorphic activation patterns of enhanced left fronto-striatal activation in females and enhanced right parietal activation in males during motor inhibition appear to be the result of underlying gender differences in the functional maturation of these brain regions [
35,
36].
Looking at personal factors, first, we found that children aged 12–16 years were more likely to have behavioral problems, and the overall age differences in the adjusted logistic regression analysis were nearly significant (
P = 0.070). Although studies have shown that frequency and tic severity decline with age, social, peer and family relationships, abilities, and school/work impairment caused by tic decrease at follow-up do not completely improve [
37]. Second, our study reported that TD cases with an abnormal birth history had a risk effect on behavioral problems when compared to TD patients with a normal birth history. The association between prenatal and perinatal epigenetic factors and TD has been reported in previous studies [
9], indicating that prenatal and perinatal factors of abnormal birth history should be considered in the clinical spectrum of TD and comorbidities, which may share a common etio-pathogenetic basis.
Considering parental factors, it seemed that higher paternal education level was a protective factor against behavioral problems among TD cases, although the overall association between parental education level and behavioral problems was not significant in our study. This outcome is inconsistent with previous findings. Cui et al. previously investigated the risk factors of comorbid ADHD in children with TS, reporting that low family education and lower cultural levels of parents were key risk factors for the co-occurrence of TS and ADHD [
38]. Hosokawa et al. found that lower maternal education level predicted externalized problems and behavioral problems, while paternal education level did not predict any clinically significant behavioral problems [
39]. The lack of significance of the association in our study may be due to different sample populations, or affected by other factors which may act as confounders in the described relation between parental education level and behavioral problems. Because literature on this relation in children with TD is rare, further research mapping out the associations is needed.
Regarding family-environmental factors, as expected, the current study showed a positive association between the parenting pattern of indulgence and behavioral problems in children with TD. This result is consistent with those of previous studies [
40,
41]. The epidemiological findings of children’s behavioral problems also indicated that lack of supervision, limited conversation time and parent–child interactions may contribute to problematic behaviors [
42,
43]. Negative parenting styles foster a hostile and neglectful environment for children and inhibit their ability to appropriately self-regulate behavioral problems [
43]. Therefore, it would be helpful to minimize the occurrence and development of behavioral problems in children with TD via providing proper suggestions for parents on parenting patterns in clinical practice. In addition, the findings in our study revealed the role of family history in TD and behavioral problems. TD are considered to be one category among the most heritable neuropsychiatric conditions. Comorbid symptoms such as ADHD, OCD, and depression persist into adulthood and require close monitoring for its heritability [
9]. These findings should encourage clinicians and child-health practitioners to pay more attention to high-risk individuals.
The relationship between behavioral problems and clinical characteristics in children with TD was significant, which is consistent with our expectations. First, having moderate/severe tic severity increases the risk of behavioral problems. Children with CTD/TS also have an increased risk of behavioral problems. Previous studies have also reported the effect of clinical characteristics on TD with behavioral problems [
11,
44]. This independent association could be explained in several possible ways: First, the more severe, complex or persistent the tics, the more obvious the symptoms and the greater the functional impairment will present [
45]. Second, these children showed more problems in their peer relationships and were perceived as withdrawn and unpopular by their peers [
10]. Third, such children nearly always present with multiple psychiatric comorbidities, which are often more impairing than the tics themselves [
7,
46]. Complex tics were statistically significant in the unadjusted logistic analysis, but lost this association in the adjusted logistic analysis. Further studies are required to determine whether these findings can be replicated using larger datasets.
Moreover, this study developed a prediction model for the individualized risk of behavioral problems in children with TD, and a nomogram was plotted for the prediction model. To date, the application of a nomogram to predict the risk of TD with behavioral problems is lacking, although nomograms have been widely used as a reliable clinical tool to create a simple intuitive graph to quantify the risk of a clinical event of interest in other diseases [
19,
20]. In the present study, the model based on age, abnormal birth history, parenting pattern, family history, TD type and tic severity had a significant predictive performance for behavioral problems in children with TD. This nomogram was feasible for making beneficial decisions in clinical practice, according to the satisfactory results of the DCA and CIC. This showed that a family history of parent/close relatives with psychiatric disorders accounted for the largest contribution to the risk of TD with behavioral problems, which further emphasizes the heritability of TD development. The contributions of the factors including age 12–16 years, CTD/TS, parenting pattern of indulgence, moderate/severe tic severity and abnormal birth history, though less than that of family history, indicate that clinical characteristics and other sociodemographic characteristics also contribute to the occurrence of TD with behavioral problems. This nomogram would allow clinicians to rapidly identify patients with a higher risk of behavioral problems and tailor necessary interventions as early as possible to improve clinical outcomes.
Although this study has some advantages, it also has several limitations. First, this was a hospital-based cross-sectional study, and all participants were recruited from one hospital, potentially leading to admission bias and therefore a relatively lower quality of evidence. Second, behavioral problems are complex and involve multiple factors. Residual confounding factors may have affected the associations, although we assessed numerous factors. Third, the outcome variable was integrated, and non-specific psychometrics was used in this study. Therefore, the clinical utility of distinguishing between specific psychopathologies is limited. Finally, the prediction model was internally validated in our study, and external validation is lacking. Future research involving large-scale, multicenter settings is required to further validate our findings of the study.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.