Sleep Disorders, Sleep Medication Use, and Predictors of Sleep Disturbance in Children with Persistent Tic Disorders

Abstract

The present study examined rates of sleep disorders and sleep medication use, and predictors of sleep disturbance in children with persistent tic disorders (PTD). Sixty-three parents of children aged 10 to 17 years with PTDs completed an internet survey evaluating sleep patterns and clinical symptoms. Insomnia (19.4%), nightmares (16.1%), and bruxism (13.1%) were the most commonly reported lifetime sleep disorders. Fifty-two percent endorsed current sleep medication use. Higher ADHD severity, overall life impairment, and female sex predicted greater sleep disturbance. Findings suggest the utility of clinical management of co-occurring ADHD and impairment to mitigate sleep disturbance in children with PTDs.

Persistent tic disorders (PTDs), including Tourette’s disorder (TD), are neurological conditions involving the presence of motor and/or vocal tics enduring beyond one year, with onset before age 18 years (American Psychiatric Association, 2022). Tics emerge between 6 and 7 years, and disproportionately affect males (Freeman et al., 2000). Tics fluctuate in topography and severity over their course, with symptoms peaking in preadolescence, and declining during adolescence in many cases (Bloch & Leckman, 2009). PTDs are associated with significant social, emotional, family, academic, and physical impairment (Conelea et al., 2013). Psychiatric comorbidity is high, presenting at a rate of 85% in children with PTDs (Wolicki et al., 2019). Attention deficit/hyperactivity disorder (ADHD), obsessive-compulsive disorder (OCD), and anxiety disorders are particularly common in youth with PTDs (Wolicki et al., 2019).

Sleep disorders and disturbance are also prevalent in children with PTDs, with rates as high as 85% found (Młodzikowska-Albrecht et al., 2010). Insomnia disorder, or difficulties with sleep initiation and maintenance are the most common forms of sleep disturbance present in children with PTDs (Ghosh et al., 2014; Modafferi et al., 2016). Parasomnias, including sleep walking, night terrors, nightmares, and nocturnal enuresis, are also highly endorsed in children with TD (Młodzikowska-Albrecht et al., 2010; Modafferi et al., 2016). Increased motor activity and snoring during sleep also frequently present in this population (Modafferi et al., 2016). Findings regarding rates of sleep disturbance in children with TD from existing studies are informative. However, there is much variation in the type of sleep disturbances assessed, making it challenging to estimate relative rates across a number of specific sleep disturbances. Further, some prior samples are mixed in age group, making it challenging to draw conclusions about children with PTDs. Therefore, a more inclusive list of sleep disorders is needed to facilitate knowledge regarding the type of sleep disorders commonly present in children with PTDs, in order to best characterize this clinical issue and provide recommendations for future research and intervention.

The presence of sleep disruption in children with PTDs may prompt parents to seek prescription or over-the-counter sleep medication for their children. General rates of sleep medication use in children with PTDs are unknown. However, adults with PTDs endorse current prescription or over-the-counter sleep medication use at a rate of 40%, with melatonin and diphenhydramine (antihistamine) most commonly reported (Ricketts et al., 2022a). Studies evaluating outpatient physician practices for treating sleep disturbance in children show that antihistamines are most commonly prescribed or recommended, followed by alpha-2-agonists (also commonly prescribed for tics) and melatonin (Owens et al., 2003). Research suggests use of over-the-counter sleep medications, such as melatonin in children, is steadily rising among the population (Panjwani et al., 2021). Parents have reported a melatonin use rate of 14% in children with PTDs (Smith & Ludlow, 2021). Though the general prevalence of sleep medication use in children with PTDs is unknown, the higher relative prescribing/recommendation rates for sleep medication in children with neurodevelopmental disorders (e.g., ADHD) (Owens et al., 2003) suggests use may be high in children with PTDs. Due to the unknown efficacy of sleep medications in PTDs and their potential for adverse side effects (Owens et al., 2003), it remains important to investigate sleep medication use in youth with PTDs to understand patterns of use.

Several clinical characteristics are associated with sleep disturbance in children with PTDs. Some research suggests that higher tic severity is associated with greater sleep disturbance, though findings remain mixed (Modafferi et al., 2016; Ricketts et al., 2022b; Sambrani et al., 2016). Additionally, individuals with PTDs with psychiatric comorbidity are more likely than individuals with PTDs who are free of comorbidity to have sleep problems (Freeman et al., 2000; Ricketts et al., 2022b; Sambrani et al., 2016). Co-occurring ADHD symptoms have consistently been associated with greater sleep disturbance in youth PTDs (Hibberd et al., 2020; Ricketts et al., 2022b). Further, anxiety and poor quality of life are also related to greater sleep disruption in children with TD (Modafferi et al., 2016; Storch et al., 2009). Demographic factors, including sex and age, have also been associated with sleep disturbance in youth with TD, wherein younger female adolescents are more likely to have poor sleep sufficiency relative to older male youth (Ricketts et al., 2018). Despite a number of clinical and demographic factors found to be implicated in sleep disturbance in children with PTDs, their relative importance is unclear. Therefore, a predictive model examining the clinical and demographic factors implicated in sleep disturbance in children with PTDs would help to identify the characteristics most strongly associated with sleep disturbance.

Given the aforementioned literature on sleep disturbance and sleep medication use in youth with PTDs, the present study aimed to assess patterns of sleep disorders and sleep medication use, and the degree to which relevant clinical and demographic characteristics, including ADHD severity, anxiety disorders, tic severity, overall life impairment, age, and sex, predict parent-reported sleep disturbance in children with PTDs. Such information would have implications for the relevance and scope of sleep intervention in this clinical population.

Method

Participants

Sixty-three parents of children aged 10 to 17 years (M = 13.3, SD = 2.2) with TD (n = 52, 82.5%) or persistent motor tic disorder (n = 11, 17.5%) completed an internet survey examining sleep patterns and clinical symptoms in youth with PTDs. Participants were recruited through social media and support group websites and Chapters of the Tourette Association of America. The survey was advertised as seeking children aged 10 to 17 years with TD or other chronic tic disorders and their parents for an internet survey on sleep and wake habits, with questions assessing demographics, medical/psychiatric history, tic symptoms, and emotional and behavioral functioning. The sample was compromised of 64.5% (n = 40) male and 83.6% (n = 51) white children. See Table 1 for sample demographics.

Table 1.

Parent-Reported Sample Demographic and Clinical Characteristics

Characteristic	n	%
Sex
Female	22	35.5%
Male	40	64.5%
Race
White	51	83.6%
Asian	5	8.2%
Black/African-American	1	1.6%
American Indian or Alaskan Native	2	3.3%
Biracial	2	3.3%
Ethnicity
Hispanic	14	22.6%
Non-Hispanic	48	77.4%
Highest Biological Parent Educational Level
Some high school	1	1.6%
High school diploma or equivalent	4	6.3%
Technical/trade school or some college	12	19.0%
Junior/community college	6	9.5%
College graduate or equivalent	17	27.0%
Postgraduate/professional degree	23	36.5%
Highest Biological Parent Employment
Full time	51	82.3%
Part-time	4	6.5%
Not employed	6	9.7%
Retired	1	1.6%
Marital status of biological/primary caregivers
Married	45	72.6%
Divorced	7	11.3%
Separated	2	3.2%
Single	5	8.1%
Widowed	1	1.6%
Other	2	3.2%
Household income
$0 - $29,999	2	3.4%
$30,000 - $59,999	6	10.2%
$60,000 - $89,999	9	15.3%
$90,000 - $119,999	7	11.9%
$120,000 - $149,999	18	30.5%
$150,000 - $170,000	6	10.2%
$180,000 - $209,999	3	5.1%
$210,000 or more	8	13.6%
Clinical Measures	M	SD
PTQ	38.18	29.07
SNAP-IV	24.97	12.99
CSHQ	53.94	10.65
CSDS-P	20.75	14.67

Note. Data are presented for available cases. Valid percentages are reported. PTQ = Parent Tic Questionnaire; SNAP-IV = Swanson Nolan and Pelham-IV Teacher and Parent 18-Item Rating Scale; CSHQ = Children’s Sleep Habits Questionnaire; CSDS-P = Child Sheehan Disability Scale - Parent.

Measures

Demographics and Medical and Psychiatric History

Parents completed questionnaires regarding their child’s demographics and psychiatric and medical history. Parents also reported their marital status, highest educational attainment, household income, and current employment status. Presence of sleep disorders was evaluated using a checklist of 15 sleep disorders succeeding the question “Has your child ever been diagnosed with any of the following sleep disorders by a health professional?” Current sleep medication use was assessed using the question, “Is your child currently taking any of the following medications (prescribed or over-the-counter) to help him/her sleep?” A checklist of 15 medications was provided. Current tic medication use was evaluated through the question, “Is your child currently taking any of the following medications to treat tics?”, which was followed by a yes/no checklist of 44 tic medications. Presence of comorbid psychiatric disorders was similarly evaluated using a checklist of 16 psychiatric disorders following the question, “Has your child ever been diagnosed with any of the following by a health professional?” Participants were coded as having an anxiety disorder if they endorsed separation anxiety disorder, generalized anxiety disorder, social phobia, panic disorder, or specific phobia.

Parent Tic Questionnaire (PTQ) (Chang et al., 2009).

The PTQ is a parent-rated scale assessing frequency and intensity of child tic severity. Parents rate lists of 14 motor tics and 14 vocal tics in terms of their presence (yes/no), frequency, and intensity in the past week. Frequency is rated using a Likert scale ranging from 1 (weekly) to 4 (constantly). Intensity is rated on a 4-point Likert scale ranging from 0 to 4, with higher scores indicating higher intensity. Ratings are summed to yield a motor tic severity score (0-112), vocal tic severity score (0-112), and total score (0 to 224), with higher scores representing higher tic severity. The PTQ demonstrates excellent internal consistency, good discriminant validity, and excellent test-retest reliability (Chang et al., 2009).

Swanson Nolan And Pelham-IV Teacher and Parent 18-Item Rating Scale (Swanson et al., 2001).

The SNAP-IV 18-item rating scale is a brief parent-reported measure of ADHD symptom severity, with 9 items each assessing inattention and hyperactivity. Items are rated on a 4-point Likert scale ranging from 0 (not at all) to 3 (very much) and summed to produce a total score ranging from 0 to 54. Higher scores indicate increased ADHD symptom severity. The SNAP-IV demonstrates acceptable internal consistency and good validity (Bussing et al., 2008).

Children’s Sleep Habits Questionnaire (CSHQ) (Owens et al., 2000).

The CSHQ is a 33-item parent-report measure assessing sleep patterns across eight domains (i.e., bedtime resistance, sleep onset latency, sleep duration, sleep anxiety, nighttime awakenings, parasomnias, sleep-disordered breathing, and daytime sleepiness) in children in the past week. Sleep behaviors are rated on a 3-point scale ranging from 1 (rarely) to 3 (usually), with higher scores indicative of greater sleep disturbance. The CSHQ displays acceptable test-retest reliability, adequate internal consistency, and support for validity (Owens et al., 2000).

Child Sheehan Disability Scale - Parent (CSDS-P) (Whiteside, 2009).

The CSDS-P is a 5-item parent-report measure evaluating the degree to which the child’s symptoms impair various life domains. Items are rated using an 11-point Likert scale ranging from 0 (not at all) to 10 (very, very much), with higher scores indicative of greater impairment. Two items assess symptom-related impairment in the child’s school and social life, and three items assess symptom-related impairment in the parent’s work, social, and family life/home responsibilities. Items are summed to obtain child impairment, parent impairment, and total scores. The CSDS-P demonstrates adequate internal consistency and support for construct validity (Whiteside, 2009).

Procedure

The survey was completed using a link hosted by Research Electronic Data Capture (REDCap) from February 2019 to April 2021. Participants were shown a progress bar to indicate the percent of the survey completed on each survey page. The survey was setup such that participants were not required to respond to items. One hundred thirty-six parents accessed the survey link and 135 initiated the survey. Interested parents were presented with an institutional review board-approved consent form. They indicated whether they agreed to participate in the study by checking a yes or no box. Of the 135 participants who engaged with the survey, 120 consented to participate (88.9%) and 15 did not consent (11.1%). Of the 120 participants who consented, forty-seven (39.2%) did not continue completing the survey. Of the remaining 73 participants, 10 participants were excluded because either (1) the current age of the child was younger than 10 years or older than 17 years (n = 5), or (2) no endorsement of lifetime TD or PTD diagnosis (i.e., “Has your child ever been diagnosed with Tourette’s Disorder (Tourette Syndrome), Persistent (Chronic) Motor Tic Disorder or Persistent (Chronic) Vocal Tic Disorder by a health professional?”) (n = 5). The final sample included 63 participants. Upon survey completion, participants were directed to a separate link prompting them to submit their email address in order to be entered into a random drawing with 1 in 20 odds to win a $50 gift card. Neither survey participation, nor survey completion were required in order to enter the drawing.

Data Analysis

SPSS 28.0 was used to perform analyses. Missing value analysis showed that 11.79% of values for the six variables (PTQ total, SNAP-IV total, CSDS-P total, anxiety disorder, sex, and age) used in multiple regression analysis; thus, multiple imputation was performed. The assumption of data missingness-at-random needed for multiple imputation was met per Little’s test of missing completely at random (MCAR; Χ²(21) = 25.22, p = .24). However, data did not meet the assumption of a parametric distribution (Kleinke, 2017). Therefore, predictive mean matching, which imputes an actual observed value selected from a sample of k > 1 near-neighbor values, and maintains the original data distribution and associations among variables, was performed (Kleinke, 2017). As recommended for default, a near-neighbor value of 5 was selected (Kleinke, 2017). Ten imputations were performed to produce sufficient missing data estimates (Van Buuren, 2018). Items were imputed at the total score level, rather than item level, as this is optimal for small sample sizes (Rombach et al., 2018). Addition of auxiliary variables (e.g., subtotal scores) in the imputation model can improve multiple imputation; however, data were imputed at the total score level, as inclusion of subtotal scores caused the ratio of variables to cases to fall below 1:3, which is not recommended in small samples (Hardt et al., 2012).

Descriptive statistics (see Tables 2 and 3) were performed on unimputed data to examine frequencies and means for demographics, lifetime psychiatric disorders, lifetime sleep disorders, and current use of prescription or over-the-counter sleep medication. Predictors were selected a priori based on the factors commonly reported in the literature to be associated with sleep disturbance in youth with TD. Multiple linear regression analysis was performed on imputed data to examine the degree to which tic severity (PTQ total), impairment (CSDS-P total), ADHD severity (SNAP-IV total), anxiety disorder (presence vs. absence), age, and sex predicted sleep disturbance (CSHQ total). Adjusted R² is also reported to establish the predictive power of the model, with adjustment for the number of predictors. Higher adjusted R² values, indicating higher predictive power of the model corrected for the number of predictors in the model, are reported (Austin & Steyerberg, 2015). To examine multicollinearity among predictors, the variance-inflation factor (VIF) is reported. A VIF factor below 10 indicates reduced collinearity among the variables within the model (Mela & Kopalle, 2002). Effect size was determined using Cohen’s f² and can be interpreted as follows: .02 (small), .15 (moderate), and .35 (large) (Cohen, 1998). Pooled regression statistics are reported. A sensitivity analysis was also performed using complete case (unimputed) data.

Table 2.

Parent Report of Lifetime Co-occurring Psychiatric and Sleep Disorders in Children with Persistent Tic Disorders

Lifetime Psychiatric Disorders	n	%	Total n	Lifetime Psychiatric Disorders	n	%	Total n
≥ 1 co-occurring psychiatric disorder	40	76.9%	52	≥ 1 co-occurring sleep disorder	27	45.0%	60
Attention-deficit/hyperactivity disorder (ADHD or ADD)	32	52.5%	61	Insomnia	12	19.4%	62
Obsessive-compulsive disorder	28	46.7%	60	Hypersomnia	1	1.6%	61
Trichotillomania (hair-pulling) disorder	6	10.3%	58	Bruxism	8	13.1%	61
Excoriation (skin-picking) disorder	10	17.5%	57	Obstructive sleep apnea	2	3.3%	60
Separation anxiety disorder	14	25.0%	56	Delayed sleep phase syndrome	2	3.3%	61
Generalized anxiety disorder	32	53.3%	60	Advanced sleep phase syndrome	1	1.6%	61
Social phobia (social anxiety disorder)	11	19.6%	56	Narcolepsy	1	1.6%	61
Panic disorder	7	12.3%	57	Sleep walking	6	9.8%	61
Specific phobia	6	10.7%	56	Nocturnal enuresis	3	4.9%	61
Any anxiety disorder*	35	58..3%	60	Night terror	3	4.9%	61
Major depressive disorder	8	14.3%	56	Nightmares	10	16.1%	62
Bipolar disorder	4	7.3%	55	Restless legs syndrome	6	9.8%	61
Oppositional defiant disorder	4	7.1%	56	Periodic limb movement disorder	7	11.5%	61
Posttraumatic stress disorder (PTSD)	3	5.4%	56	Rapid eye movement sleep behavior disorder	2	3.2%	61
Anorexia nervosa	1	1.8%	56
Bulimia nervosa	0	0%	56
Binge eating disorder	4	7.1%	56

Note. Frequencies are presented for available cases. Valid percentages are reported. Our whole sample size (total n) was 63. The total n value represents the number of participants who responded to each item.

^*Participants were coded as having an anxiety disorder if they endorsed separation anxiety disorder, generalized anxiety disorder, social phobia, panic disorder, or specific phobia.

Table 3.

Parent-reported Sleep Medication Use in Children with Persistent Tic Disorders

Medication	n	%
≥ 1 Current prescribed or over-the-counter sleep medication	33	52.4%
> 1 prescribed or over-the-counter sleep medication	8	12.7%
Ambien (zolpidem)	3	5.2%
Belsomra (suvorexant)	2	3.4%
Desyrel (trazodone)	3	5.1%
Halcion (triazolam)	0	0.0%
Lunesta (eszopiclone)	0	0.0%
Restoril (temazepam)	1	1.7%
Rozerem (ramelteon)	0	0.0%
Sonata (zaleplon)	3	5.2%
Silenor (doxepine)	2	3.5%
Xanax (alprazolam)	2	3.4%
Benadryl, Unisom sleep gels, etc. (diphenhydramine)	5	8.3%
Unisom sleep tabs (doxylamine succinate)	3	5.2%
Melatonin	25	39.7%
Valerian	0	0.0%

Note. Frequencies are presented for available cases. Valid percentages are reported.

Results

Lifetime Sleep Disorders and Current Sleep Medication Use

Nearly half of the sample (n = 27, 45%) endorsed one or more lifetime sleep disorders (see Table 2). Insomnia was the most commonly endorsed sleep disorder (n = 12, 19.4%), with nightmares (n = 10, 16.1%), bruxism (n = 8, 13.1%), periodic limb movement disorder (n = 7, 11.5%), sleep walking (n = 6, 9.8%), and restless legs syndrome (n = 6, 9.8%) closely following. Over half of the sample reported currently using one or more prescribed or over-the-counter sleep medications (n = 33, 52.4%), with some using more than one medication (n = 8, 12.7%). Approximately 40% (n = 25, 39.7%) endorsed melatonin use, with a few using diphenhydramine products (n = 5, 8.3%) (see Table 3).

Multiple Linear Regression Analysis of Clinical Predictors of Sleep Disturbance

Multiple linear regression analysis was conducted to examine clinical and demographic predictors of parent-reported child sleep disturbance (see Table 4). SNAP-IV total (p = .006) and CSDS-P total (p = .013) significantly predicted sleep disturbance, such that higher SNAP-IV total and CSDS-P total scores were associated with higher sleep disturbance. Sex also predicted sleep disturbance (p = .013), with females more likely to have higher sleep disturbance than males. The model accounted for a substantial portion (39%) of the variance in sleep disruption. Tic severity, presence of an anxiety disorder, and age did not significantly predict sleep disturbance.

Table 4.

Pooled Regression Estimates for Clinical Predictors of Sleep Disturbance

Predictor	B	SE	β	t	p	VIF	Adj. R2	Cohen’s f2
							.39	.81
Constant	51.52	1.55		4.88	<.001
PTQ Total	.06	.05	.02	.13	.901	1.77
SNAP-IV Total	.29	.11	.35	2.74	.006	1.48
Anxiety Disorder	1.03	2.74	.05	.38	.706	1.30
CSDS-P Total	.23	.09	.31	2.48	.013	1.62
Sex	−6.54	2.60	−.29	−2.51	.013	1.16
Age	−.45	.66	−.09	−.69	.497	1.33

Note. Regression was conducted on imputed data. B = unstandardized beta; SE = standard error; β = standardized beta; VIF = variance inflation factor; Adj. = adjusted; PTQ = Parent Tic Questionnaire; SNAP-IV = Swanson Nolan and Pelham-IV Teacher and Parent 18-Item Rating Scale; CSDS-P = Child Sheehan Disability Scale - Parent.

Results of multiple linear regression analysis on complete case (unimputed) data are reported in Table S1. CSDS-P total (p = .035) significantly predicted sleep disturbance, such that higher CSDS-P total scores were associated with higher sleep disturbance. Sex also predicted sleep disturbance (p = .030), with females more likely to have higher sleep disturbance than males. SNAP-IV total did not significantly predict sleep disturbance, p = .074. The model accounted for a substantial portion (32%) of the variance in sleep disruption. Tic severity, presence of an anxiety disorder, and age remained non-significant predictors of sleep disturbance.

Discussion

The present study examined lifetime presence of sleep disorders, patterns of sleep medication use, and predictors of parent-reported sleep disturbance in youth with PTDs. Forty-five percent of parents reported their child had a lifetime sleep disorder diagnosis, with insomnia and nightmares most commonly endorsed. Just over half (52.4%) of parents endorsed child sleep medication use, with melatonin most commonly cited, followed by diphenhydramine. Finally, higher ADHD severity and overall life impairment in addition to female sex significantly predicted greater sleep disturbance. The present study updates the literature with more inclusive assessment of sleep disorders, provides estimates of sleep medication use patterns, and confirms the clinical predictors of sleep disturbance found in prior research.

Insomnia disorder was the most highly endorsed lifetime sleep disorder (19.4%). The ranking of insomnia as most common aligns with prior research findings (Ghosh et al., 2014), although our rate is lower than rates of current insomnia (32% - 42%) reported in a prior sample of children with PTDs (Ghosh et al., 2014). Reasons for the lower rate are unclear, but are likely due our limited sample size, and differences in study methods, as Ghosh et al. (2014) coded presence or absence of specific sleep disorders from a self-report of sleep disturbances. Though insomnia is often tied to comorbid ADHD, it is also highly endorsed in TD samples free of ADHD, suggesting it is a stand-alone comorbidity of TD (Ghosh et al., 2014). Nightmares, nocturnal bruxism, periodic limb movement disorder, sleep walking, and restless legs syndrome were also commonly endorsed in the stated order. Results are consistent with prior findings showing high endorsement of sleep walking, motor activity during sleep, nightmares, and nocturnal bruxism in children with TD (Modafferi et al., 2016), and align with research suggesting mechanistic links between restless legs syndrome and periodic limb movement disorder and TD (Rivière et al., 2009).

Interestingly, night terrors were endorsed at a lower rate (5%) than anticipated based on higher endorsement in children with TD in prior studies (Młodzikowska-Albrecht et al., 2010). The low endorsement rate in the present sample may be a limitation of our sample size or a function of a lack of formal diagnosis, despite the potential presence of night terrors in these children with PTDs. Nevertheless, together, the frequency of specific lifetime sleep disorders found in the current sample still indicates the common co-occurrence of insomnias, parasomnias, and sleep-related movement disorders in children with PTDs. Importantly, genetic and biochemical overlap may explain the link between these conditions. For example, variants in the BTBD9 gene, which underlies both restless legs syndrome and periodic limb movement disorder, are also associated with TD (Alonso-Navarro et al., 2019). Further, PTDs, restless legs syndrome, and ADHD have all been linked to iron deficiency (Alonso-Navarro et al., 2019). Research is needed to further evaluate the pathophysiological significance of the co-occurrence of PTDs with these sleep disorders and others in children with PTDs.

Sleep medication use was frequently endorsed in our sample, with a higher rate found here (52.4%) relative to that (40.0%) found in adults with PTDs (Ricketts et al., 2022a). Approximately 13% of parents endorsed use of more than one sleep medication by their child, suggesting the relevance of polypharmacy for a small proportion of the sample. Increasing attention has been given to the potential for increased risk for adverse medication effects related to polypharmacy in children (Bogler et al., 2019), thus efforts should be made to circumvent polypharmacy when possible (see Limitations section). Rates of melatonin use were especially high (40%) in the current sample. A nutritional supplement survey completed by parents of youth with TD found a lower rate of melatonin use (14%) than that found in the present study, yet still higher relative to the 0% of typically developing children in their sample (Smith & Ludlow, 2021). As melatonin is a dietary supplement, it is not regulated by the FDA (Emens & Burgess, 2015). Its short-term use is associated with mild side effects (e.g., daytime sleepiness, dizziness), but there are some concerns for its safety (e.g., the potential for pubertal delays) over longer-term use, and there has been limited long-term safety testing of this supplement in children (Emens & Burgess, 2015). Further, the content of over-the-counter melatonin products shows high variability relative to label specifications (Erland & Saxena, 2017). Nonetheless, melatonin remains a popular choice for children among parents (Panjwani et al., 2021), and has shown to be effective in reducing sleep onset latency and shifting circadian phase depending on timing of administration (Emens & Burgess, 2015). Melatonin formulations that pass the United States Pharmacopeial Convention (USP) independent audit and display the (USP) label should contain an accurate dose with no impurities (Emens & Burgess, 2015).

Per parent report of youth, melatonin has been associated with improved sleep, and motor and vocal tics in children with PTDs (Smith & Ludlow, 2021). In light of its widespread use and aforementioned benefits and limitations, the present findings highlight the need for establishing the efficacy and side effect potential of melatonin use in children with PTDs through controlled trials. Antihistamine use was endorsed at a much lower rate of 8%. Antihistamine is among the most frequently recommended over-the-counter medications for pediatric insomnia (Owens et al., 2003), however, there are not yet studies examining its efficacy for sleep problems. Notably, a number of prescription sleep medications were endorsed, each at low rates. There are presently no FDA approved medications to treat sleep problems in children. Benzodiazepines and nonbenzodiazepine hypnotics, each endorsed at low rates, are less often prescribed by physicians to treat pediatric sleep problems than antihistamines, alpha-2-adrenergic agonists, melatonin and antidepressants due to adverse side effects (Owens et al., 2003).

With respect to clinical predictors of parent-reported sleep disturbance, in line with prior mixed findings (Modafferi et al., 2016; Sambrani et al., 2016), tic severity was not a significant predictor of sleep disturbance in the current sample. Our findings suggest that sleep disturbance in children with PTDs may be more readily impacted by factors which are commonly associated with tics (e.g., ADHD, impairment), rather than tic severity. Results are consistent with literature showing ADHD symptoms are frequently associated with sleep disturbance in both children with primary PTDs (Hibberd et al., 2020; Keenan et al., 2021) and children with primary ADHD (Lunsford-Avery et al., 2016). It is important to note that stimulant medication for ADHD, not assessed in this study, and psychiatric comorbidity may have also contributed to the association between ADHD symptoms and sleep disturbance. For example, stimulant medication and psychiatric comorbidities (e.g., anxiety, emotional symptoms, conduct problems, oppositional defiant disorder) have accounted for the relationship between ADHD symptoms and sleep disturbance in some studies (Frick et al., 2022; Mick et al., 2000). Further, it is important to note that numerous comorbid psychiatric conditions (e.g., ADHD, disruptive behavior disorders, anxiety, OCD) may have contributed to the positive association found between overall life impairment and sleep disturbance, as psychiatric comorbidities exacerbate impairment within TD (Ghanizadeh & Mosallaei, 2009; Lebowitz et al., 2012; Lewin et al., 2011) and are associated with sleep disturbance independently of TD (Chase & Pincus, 2011; Shanahan et al., 2014; Storch et al., 2008). This suggests the clinical significance of psychiatric comorbidities on sleep disturbance, highlighting the importance of addressing psychiatric complaints when treating sleep disturbance.

Finally, in line with prior research (Sambrani et al., 2016), female sex was a predictor of sleep disturbance in the present study. The reasons for this are unclear but may relate with increased depression and anxiety found in females relative to males with PTDs (Garris & Quigg, 2021). Negative affect is commonly associated with increased sleep disturbance in adolescents (Alvaro et al., 2014; Alvaro et al., 2017). Further, sex steroids may also contribute to this sex difference in sleep disturbance, as ovarian hormones, estrogen and progestogen are known to impact sleep in females (Dorsey et al., 2021). Thus, these are areas worth investigating in future research.

The present study has the following limitations. The small sample size limits the ability to generalize patterns of sleep disorders and sleep medication use to the broader population of children with PTDs. The small sample also limited the number of predictors included in the model. For example, we did not prioritize inclusion of lifetime OCD diagnosis and current tic medication use, which have been linked to sleep disturbance in youth with PTDs (Ghosh et al., 2014; Hibberd et al., 2020; Sambrani et al., 2016). Also, parents were asked to endorse child sleep disorders as diagnosed by a health care professional; however, the sleep disorders assessed included problems (e.g., nightmares, sleep walking, nocturnal enuresis) that families may be less likely to seek treatment or assessment for, and thus less likely to receive an official diagnosis for. This may have influenced rates of endorsement. Additionally, though multiple imputation was used to account for missing data, the amount of missing data may have skewed the results. Although our results using unimputed data are similar to the imputed data analysis, ADHD symptom severity did not rise to statistical significance. Further, as this study is based on cross-sectional data, the directionality of the relationships remains unclear. For example, sleep disturbance may also adversely influence ADHD severity (Gosling et al., 2022) and overall life impairment. Moreover, as the survey was advertised as focusing on sleep, a selection bias may have occurred, whereby parents of children with sleep problems were more likely to partake in the study relative to those without such difficulties. Additionally, we lack confirmation of lifetime diagnoses by a health professional; and as lifetime diagnoses require retrospective report, they may be limited by recall bias. Finally, clonidine, an alpha-2-adrenergic agonist used to treat PTDs and often prescribed off-label to address sleep disturbance (Owens et al., 2003), was not included as a sleep medication option in the survey, which may have lowered the medication use rates. In the current sample, 13.8% (n = 8) endorsed current clonidine use to treat tics. Future investigations should seek to determine the rates of clonidine use for sleep and test its efficacy for reducing sleep disruption in children with PTDs, as this may have implications for reducing polypharmacy. Limitations notwithstanding, findings draw attention to the utility of controlled testing of melatonin and investigation of behavioral sleep intervention alternatives or supplements to medication in children with PTDs. Findings also suggest the importance of clinical management of ADHD symptoms and functional impairment to combat sleep problems in children with PTDs, and future research investigating the mechanisms underlying the association between female sex and sleep disturbance in youth with PTDs is warranted.

Implications for Practice

The common occurrence of sleep medication use found in the current sample, taken together with the frequent endorsement of insomnia, suggests the relevance of future research aimed at evaluating behavioral sleep interventions as an alternative or addition to sleep medication. Treatments such as cognitive-behavioral therapy (CBT) for insomnia (Dewald-Kaufmann et al., 2019), targeting enduring symptoms of insomnia, including anxiety related to sleep, sleep-interfering behaviors, and sleep drive dysregulation (Mitchell et al., 2012) have yet to be investigated in PTDs. This research highlights the importance of addressing functional impairment when treating sleep disruption in youth with PTDs. Treatment protocols, such as Living with Tics, aimed at directly targeting functional impairment, have been associated with reduced tic-related impairment and improved quality of life in youth with PTDs (McGuire et al., 2015). Further, treating ADHD symptoms, which are frequently associated with sleep disturbance in youth with PTDs (Modafferi et al., 2016), should take precedence. Treatments which blend behavior therapy for tics with behavioral interventions (e.g., parent training) for ADHD (Dale et al., 2021) should be evaluated to establish the degree to which treating co-occurring ADHD symptoms positively impacts sleep. One such intervention – modified comprehensive behavioral intervention for tics – combining CBIT with CBT for ADHD and components of Living with Tics (described above), has shown to be a promising and feasible intervention for addressing tics, co-occurring ADHD symptoms, and functional impairment in youth with PTDs (Greenberg et al., 2022). Further investigation of novel blended therapies is needed to address the burden of comorbidities and functional impairment on sleep in youth with PTDs.