Tourette Syndrome and Risk of Depression: A Population-Based Cohort Study in Taiwan
Tourette syndrome (TS) is a developmental neuropsychiatric disorder characterized by multiple, brief, stereotypical
nonrhythmic movements and phonations (called
tics) lasting at least 1 year.1 The onset of tic symptoms in TS typically occurs during early childhood, and the disorder affects 0.15% to 3.8% of Western2 and 0.56% of
Taiwanese3 school-age children. Children with TS often function poorly in numerous psychosocial domains4 and have high rates of psychiatric comorbidity.5 An epidemiological study in an elementary school with 2000 Taiwanese
children aged from 6 to 12 years demonstrated that the prevalence TS was approximately 0.56%. The ratio of males to females with TS was 9:2. Comorbid rates were 36% for attention-deficit hyperactivity disorder (ADHD),
27% for self-injurious behaviors, and 18% for obsessivecompulsive disorder (OCD) and learning difficulties.3
A recent study showed that children with TS experience more frequently with comorbid depression and
anxiety disorders in adolescence and early adulthood than the comparison group. The severity of tics in TS children usually declines during adolescence, but comorbidities may persist and often cause more functional impairment.6 It is unclear, however, whether psychosocial functioning, such as tics, improves by this age. Previous longitudinal studies have found that patients with TS functioned reasonably well in late adolescence or early adulthood, but these studies were limited by the absence of a control group.7–9 By contrast, controlled cross-sectional studies have reported that adults with TS had a poorer quality of life10 and more psychopathology.11 A critical limitation to these studies was that the participants were not necessarily identified as having TS in childhood and attended
specialty clinics as adults. Consequently, the poor outcomes described may apply mainly to the minority of TS
patients who continue to have prominent tics and pursue treatment during adulthood.
This study aims to determine whether individuals with TS are more likely to be diagnosed with depression than those without TS. A representative data set was used to perform a population-based retrospective cohort study to determine the risk of depression for TS patients.
The database was obtained from the National Health Insurance (NHI) system in Taiwan.
METHODS Data Source
The Taiwan National Health Insurance (NHI) program has provided universal health insurance since 1996 and
covered almost 99% of the Taiwanese population in 1998. The NHI Research Database (NHIRD) includes
annual reimbursement claims data from the Taiwan NHI program. The National Health Research Institute (NHRI) established and manages the database and encrypts all identification information to protect patient privacy before the data are released for research.
This research uses the Longitudinal Health Insurance Database (LHID), a subset of the NHIRD. The LHID includes historical claims data for 1 million people randomly sampled from the whole insured population from
1996 to 2000. The NHRI created an anonymous identification number to link each claimant’s demographic
information (including sex, birth date, occupation, and residential area) with their registry of medical services. This research used the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) to define disease diagnoses from outpatient and inpatient data.
All the included study subjects with TS diagnoses and outcome measure of depression diagnoses were already confirmed by accurate ICD-9 coding according to the study subject’s medical records in the NHIRD. In addition, TS and depression diagnoses in NHIRD were
already evaluated based on validated definitions.12 Study Population
This research is a population-based retrospective cohort study. The Tourette syndrome (TS) cohort included patients younger than 18 years with TS (ICD-9- CM 307.2) diagnosed between 2000 and 2010. The baseline was set as the date of TS diagnosis. A comparative
cohort was randomly selected from rest of children in the same years frequency-matched individuals without TS by sex, age (in intervals of 3 years), urbanization of residence area, parental occupation, and baseline year. Newly
occurring depression (ICD-9-CM 296.2-296.3, 300.4, and 311) was noted in the 2 cohorts. For each child with TS, 10 comparison children were selected to increase the statistical power and avoid small number of depression in the
stratified analysis. Children with depression diagnosed at baseline were excluded from the study cohorts. Follow-up was continued until developing depression, insurance was withdrawn, or December 31, 2010.
Parental occupation was classified into 3 groups (white collar, blue collar, and other). Residential area was grouped into 4 urbanization levels based on an index that included population density (people/km2), population ratio of educational level, population ratio of older adults, population
ratio of agricultural workers, and the number of physicians per 100,000 people.13 Level 1 reflects the highest degree of urbanization and Level 4 reflects the lowest. We also collected the children comorbidity history before the end
of follow-up. The comorbidity included attention-deficithyperactivity disorder (ADHD, ICD-9-CM 314) and
obsessive-compulsive disorder (OCD, ICD-9-CM 300.3).
Statistical Analysis
The means and standard deviations (SDs) for continuous variables and counts and percentages for categorical variables were used to demonstrate the baseline distribution of the Tourette syndrome (TS) and comparison
cohorts. The t test and x2 tests were used for continuous and categorical variables, respectively, to assess the difference between the cohorts. The overall depression
incidence and demographic-specific depression incidence
rate were calculated by number of patients per 10,000 person-years. The Kaplan-Meier method was used to depict cumulative depression incidence for the TS and comparison cohorts. A log-rank test was used to test the difference between the curves. Cox’s proportional hazards regression model was used to estimate the hazard
ratio (HR) and confidence interval (CI) for the TS cohort compared with the comparison cohort, after controlling for potential confounding factors. Because the proportions of comorbidity in study population were small,
the comorbidities were excluded from the adjusted model. Demographic-specific HRs were estimated to compare patients with TS with patients without TS by demographic characteristics. We grouped Levels 1 and 2 of urbanization into the urban group and grouped Levels 3
and 41 of urbanization into the rural group for estimating demographic-specific HRs. Further data analysis was
conducted to observe how the risk of depression changed during the follow-up period.
Data management and analysis were performed using SAS 9.1 software (SAS Institute, Cary, NC), and cumulative incidence curves were drawn using R software (R Foundation for Statistical Computing, Vienna, Austria). The significance level was set at p , .05 for two-side testing.
RESULTS
The study includes a cohort of 1,337 patients with Tourette syndrome (TS) and a comparative cohort of 13,370 patients (Table 1). The mean age of the 2 cohorts was 9.3 years, and there were 3 times more boys than girls; 34% of the study population lived in highly urbanized areas and 63% of parents were engaged in whitecollar occupations; 3.7% of patients developed depression in the TS cohort compared with 0.8% in the comparative cohort. The mean age of depression diagnosis in the TS cohort was nearly 2 years less than in the comparison cohort (14.4 vs 16.5 years, p ,. 0001).
The depression incidence in the TS cohort was 66.47 per 10,000 person-years (Table 2), which is 4.6 times
higher than the comparison cohort (14.50 per 10,000 person-years). Figure 1 shows the cumulative TS incidence curve for the 2 cohorts and indicates that the TS incidence curve is significantly higher than the comparison cohort (p value for the log-rank test , .0001). After adjusting for potential confounding, the TS cohort is more than 4.85 times more likely to develop depression than the comparison cohort (hazards ratio [HR] 5 4.85, 95% confidence
interval [CI] 5 3.46–6.79).
Table 2 shows demographic-specific depression rates and the HRs for the 2 cohorts. The hazards of being diagnosed with depression were greater for boys than for girls (HR 5 5.76 vs 2.51) and greater for older children ($9 years) than for younger children (,9 years) (HR 5 5.16 vs 3.5). The depression rate in the TS cohort was higher for children living in urban areas than in rural areas and the lowest for children with blue-collar patents.
Further data analysis measured depression being diagnosed for both cohorts in the follow-up period (Table 3). The events of depression were on the wane over time in the TS cohort, declining from 101.2 per 10,000 person-years in Year 1 to near a half in Year 5 and longer.
On the other hand, the depression rate increased from 5.39
per 10,000 person-years to nearly five-fold higher, correspondingly, in the comparison children. The corresponding
HR for the TS cohort dwindles greatly to nearly nine-fold.
DISCUSSION
In this population-based cohort study, Tourette syndrome
(TS) was significantly associated with an increased rate of depression measured in the TS group than in the
comparison group, despite where children lived or what parental occupations were. The evidence shows that depression rate in the TS cohort was 4.6 times higher than that in the comparison cohort. The mean age of depression diagnosis in the TS cohort was less than that in the comparison cohort. This is consistent with previous studies stating that children with TS experience
comorbid depressive and anxiety disorders during adolescence
and early adulthood more frequently than those without TS.6,8 A Canadian study found approximately 40% of children with TS had experienced depression or a nonobsessive-compulsive disorder (OCD) anxiety disorder.
6 Besides, we found that males with TS were more likely to be diagnosed with depression later than females.
Although the cause of depression is multifactorial, and adult female is more vulnerable to depression, males with TS may be relatively at an elevated risk of depression. The important finding is that most depression cases in the TS cohort are identified in earlier follow-up period. It is possible that depression is a comorbidity being found because of the diagnosis and care of TS.
The reason for this association remains uncertain.
Comings and Comings14 have reported that depression is an integral part of TS and shares a common genetic basis, whereas others argue that the disorders are genetically unrelated and that TS is a risk factor for depression because of the emotional consequences of having
a chronic illness.15 The effects of tic severity, attentiondeficit hyperactivity disorder (ADHD), OCD, and psychosocial stress may account for the association of TS
with depression.16,17 Gorman et al recently found that a major depressive disorder was the only comorbid disorder not associated with a lifetime diagnosis of ADHD
or OCD, any tic, or ADHD or OCD severity. Unlike
learning and conduct disorders, the between-group differences for major depressive disorder rates remain significant
after controlling for ADHD diagnosis or severity.
Therefore, these disorders could be inherently linked, independent of the emotional effects of chronic illness.
The study was subject to certain limitations. First, the National Health Insurance Research Database (NHIRD) does not include detailed information on patients, such as physical activity, socioeconomic status, and family history of systemic or neuropsychiatric diseases. These are all risk factors that could be associated with depression. Second, evidence from a cohort study is usually less reliable than
that from randomized trials, because a cohort study design is subject to many biases related to confound adjustments.
Despite the meticulous study design with adequate control of confounding factors, bias could still remain because of possible unmeasured or unknown confounders. It is also possible that the children in the TS cohort might be more closely monitored for the development of psychiatric comorbidity and a higher rate of depression was
identified. Third, we were unable to probe for associations of depression with specific medications and psychotherapy because of the complicacies of data requiring
a larger sample size.
The data show that older adolescents with TS identified during childhood continue to experience considerable psychosocial impairment and they have high
lifetime rates of depressive disorders. Thus, although tics are likely to improve by the age of 18 years,3,4 outcomes for psychosocial functioning and comorbid conditions must be monitored. This study could provide an evidence to inform the prognosis for a child with TS. The mechanism of TS and the subsequent increased depression risk
requires further investigation.
CONCLUSIONS
This study has used a follow-up analysis to prove that children with Tourette syndrome are more than four-fold likely to be diagnosed with depression than general children. The hazard of depression wanes over time.
Whether or not depression is associated, medication for children deserves further investigation.
ACKNOWLEDGMENTS
This work was supported by the study project (DMR-101-36) in China Medical University Hospital; Taiwan Department of Health Clinical Trial and Research Center and for Excellence (DOH102-TD-B-111-004), Taiwan Department of Health Cancer Research Center for Excellence
(DOH102-TD-C-111-005); and International Research-Intensive Centers of Excellence in Taiwan (I-RiCE) (NSC101-2911-I-002-303).
