Efficacy of parent-delivered behavioral therapy for primary complex motor stereotypies

Abstract

Aim

‘Primary’ complex motor stereotypies (CMS) are persistent, patterned, repetitive, rhythmic movements in young people with typical development. This study evaluated the efficacy of an instructional DVD as a home-based, parent-administered, behavioral therapy for primary CMS.

Method

Eighty-one children with primary CMS were enrolled. Primary outcome measures included the Stereotypy Severity Scale (SSS) – Motor and Impairment scores, and Stereotypy Linear Analog Scale (SLAS). Mean CMS onset was 13.4 months (SD 13.1). Eligibility required observed CMS. Psychiatric disorders were not exclusionary and a stable medication regimen was required. Intellectual disability, neurological disorder, autism spectrum disorder, and tics were exclusionary. Initial assessments were completed via REDCap before receipt of the DVD. Fifty-four of the 81 children (34 male, 20 female; mean age 8y 2mo, SD 1.42, range 7–14y) completed assessments at 1, 2, or 3 months after receiving the DVD.

Results

Reductions (baseline to last assessment) in SSS Motor, SSS Impairment, and SLAS scores (all p<0.001) represented change ratios of –15%, –24%, and a –20%, respectively. Greatest relative treatment benefit was observed by younger children (ages 7–8y), and by one month after receipt of DVD, while a parent global assessment scale showed progressive improvement throughout the study.

Interpretation

An instructional DVD for parent-delivered behavioral therapy was a safe, effective intervention for primary CMS.

Motor stereotypies are ‘repetitive, rhythmic movements that have a predictable, fixed pattern (fashion, form, amplitude, and location), appear purposeful (but serve no obvious function), tend to be prolonged, and stop with distraction’.¹ Previously thought to occur only in the context of intellectual disability, sensory deprivation, or autism spectrum disorders, complex motor stereotypies (i.e., bilateral flapping/waving, rotating hands, or fluttering fingers in front of one's face) are now recognized to occur in a ‘primary’ form in young people with otherwise typical development.² The precise prevalence is unknown, although it has been suggested to affect up to 3% to 4% of young children.³ Primary complex motor stereotypies (CMS) typically appear before 3 years of age and have a persistent course.^2,4 Episodes last for seconds to minutes, occur in clusters, and appear multiple times per day. Primary CMS are associated with periods of excitement, engrossment, stress, fatigue, and boredom, and are readily suppressed by distractions. Parents express considerable concern regarding disruptions and stigmatization.

Although no formal pharmacological treatment studies have been performed in children with primary CMS, anecdotal data suggest this provides little benefit.^4,5 The only study to demonstrate efficacy for primary CMS treated children used a behavioral treatment consisting of awareness training and differential reinforcement of periods whereby the behavior was successfully inhibited.⁶ In this psychologist-directed treatment, therapy included two sessions of 1 hour, with additional follow-up as needed over a course of 4 months. Stereotypies were significantly reduced by approximately 14% (motor) and 33% (impairment), with a direct relationship between number of practice sessions and improvement for both scores. Recognizing the limited number of readily available practitioners to provide direct behavioral therapy and its potential cost,⁷ a home-based educational DVD for parents was developed, which mimicked the methodology utilized in the published study.

This investigation evaluated the efficacy of the instructional DVD as a home-based, parent-administered, behavioral therapy for reducing stereotypies in children with primary CMS. If effective, the therapy could potentially address the unmet needs of multiple affected children with limited access to in-person treatment. The study also sought to identify parent perceptions of their child's improvement, effects of treatment age, durability of response, and the potential effect of existing comorbidities.

Methods

The hospital institutional review board approved this protocol. Individuals with primary CMS, aged 7 to 17 years, were recruited from either the Johns Hopkins Outpatient Pediatric Neurology Clinic (Harvey S. Singer, Director), or a linked webpage (www.hopkinsmedicine.org/neuro/stereotypy). Participants were verbally consented for participation via telephone and were mailed the consent form, signed electronically via REDCap, a secure web-based application used to support the submittal of online questionnaires and data capture.

Participants

Eligibility criteria required the presence of observed (directly or via video) CMS. Participants were excluded if there was a history of intellectual disability, known neurological disorder, autism spectrum disorder, or motor/vocal tics; however, participants were not excluded if they had co-occurring disorders including attention-deficit–hyperactivity disorder (ADHD) or obsessive compulsive disorder. Participants receiving medication for coexisting problems continued their medication at a stable dose for the duration of the study.

The study flow is outlined in Figure 1. The study coordinator (TK) screened all participants for eligibility via structured telephone interview including the Autism Spectrum Screening Questionnaire. After screening and consent, parents were required to complete baseline assessment questionnaires online via REDCap before the instructional DVD was provided. Parents also received a written instructional form and a log-sheet for tracking the behavioral therapy. Post-DVD assessments were completed at 1, 2, and 3 months in a similar fashion. Participants received weekly reminders from the study coordinator.

Figure 1.

Study flow. ASSQ, Autism Spectrum Screening Questionnaire; SSS, Stereotypy Severity Scale; SLAS, Stereotypy Linear Analog Scale; MASC, Multidimensional Anxiety Scale for Children; ADHD-RS, Attention-Deficit–Hyperactivity Disorder Rating Scale-IV; CY-BOCS, Children's Yale–Brown Obsessive Compulsive Scale; PGI, Patient Global Impression of Improvement.

Assessment measures

Baseline assessments (all previously validated in this age group) included primary outcome measures (Stereotypy Severity Scale [SSS], Motor and Impairment scores; Stereotypy Linear Analog Scale [SLAS]), a secondary outcome measure (Patient Global Impression of Improvement [PGI-I]), and comorbidity measures (Multidimensional Anxiety Scale for Children-2 [MASC-2]; ADHD Rating Scale-IV; Children's Yale–Brown Obsessive Compulsive Scale).

The SSS is a five-item caregiver questionnaire consisting of two components (Motor and Impairment) for the ranking of motor stereotypy severity.^6,8 The SSS Motor score (range 0–18) quantifies motor severity and rates movements along four discriminate dimensions: number (0–3), frequency (0–5), intensity (0–5), and interference (0–5). The SSS Impairment score (range 0–50) is an independent rating of difficulties in self-esteem, family, school, or social acceptance caused by the movements.

The SLAS is a 100mm continuous line on which the parent places a mark indicating their child's stereotypy activity, ranging from 0 (best it has ever been) to 100 (worst it has ever been), considering the dimensions number, frequency, intensity, and interference.

The PGI-I is a seven-item scale that asks the parent to rate the relative improvement of CMS experienced by the patient since the beginning of the study. Ratings range from 1 (very much better) to 7 (very much worse).⁶

The Autism Spectrum Screening Questionnaire is a 27-item caregiver questionnaire addressing symptoms of autism spectrum disorder.⁹ Other parent completed measures included the MASC-2 (assessing symptoms of anxiety),¹⁰ the ADHD Rating Scale-IV (assessing symptoms of ADHD),¹¹ and the Children's Yale–Brown Obsessive Compulsive Scale (assessing symptoms of obsessive compulsive disorder).¹²

Parent-directed video-based behavioral therapy

The instructional DVD (44 minutes) included a 10-minute primary CMS didactic. The remainder of the video included instructions provided by a licensed psychologist (RW) as well as instructional vignettes, described below. Parents were instructed to implement awareness training during the first week, and to add data collection and reinforced suppression in week 2 and beyond, adapting the reinforcement to the individual progress of the child.

Awareness training

The goal of this training was to highlight for the patient when the behavior was occurring, by alternating between intentionally performing the behavior and then refraining from the behavior. The therapist first operationalized the movements and then the program taught the child to become more aware of his/her movements through a process of mimicking the behavior. Children were shown a brief video of themselves doing the movement before training. The child was requested to perform the behavior for 30 seconds, followed by 60 seconds of rest (total of five repetitions, carried out twice daily in front of a full-length mirror, if possible). The parent was requested to provide positive, verbal reinforcement when the movements were reproduced correctly, and to correct the child when the repetitions did not closely resemble the stereotypy. During a rest period, the parent verbally reinforced the child for not performing the movement. Parents were advised to withhold comment regarding movements outside of the training sessions.

Data collection

Parents were asked to maintain daily logs of the movements (i.e. situations when they occur, time of day, mood of the child, and duration).

Reinforced CMS suppression

The child was asked to practice suppressing the motor stereotypy three to five times daily during situations in which they usually occurred. Initially, sessions were 3 to 5 minutes, but were gradually increased to 25 to 30 minutes. Once there was a 90% to 95% success rate of inhibition, suppression time increased by 2 to 3 minutes each week. The parents were requested to provide verbal praise approximately every 30 seconds when no movement was observed. Reinforcement, such as points for prizes or money, contingent on successful suppression of the movement during practice sessions, was encouraged.

Statistical analyses

The total sample of 81 participants completed the baseline assessment and received the training DVD. This aggregate population was then subdivided into the (1) Intent-to-treat group (ITT; n=54), which included those who completed at least one assessment 1, 2, or 3 months after receiving the DVD; and (2) the Lost-to-follow-up group (LTF; n=27), which included those who completed no post-DVD receipt assessments. Parents of seven of the 27 participants in the LTF group withdrew their child from the study after receiving the DVD, citing one of the following reasons: mild symptoms, associated time commitment, extenuating family circumstances, or child anxiety about the treatment. One parent gave no reason. The study coordinator withdrew 19 of 27 participants because of failure to respond to multiple weekly telephone prompts.

Group differences between participants who continued with the study (n=54) and those who discontinued after receiving the DVD (n=27) were evaluated for demographic variables (age, age of stereotypy onset, sex, race/ethnicity, prior treatment, current medication status, stereotypy severity ratings, and scores on comorbidity measures) using t-tests for continuous variables, and χ² analyses for categorical variables. For group comparisons, assumptions for parametric analyses were assessed using Kolmogorov–Smirnov tests for normality of distributions, with non-parametric analyses used as indicated.

Within the ITT group, paired-tests between the baseline and last assessment were performed to evaluate the level of fold change. Two repeated measures models using generalized estimating equations (GEE) were performed to examine stereotypy severity (SSS Motor, SSS Impairment, and SLAS total scores): (1) with time (baseline, 1 month, 2 month, 3 month); and (2) time and age (at start of treatment, dichotomized at the median) and time-by-age interaction. The two-factor model controlled for age of symptom onset, sex, ADHD symptoms (ADHD RS-IV Total), anxiety (MASC-2 Total), and symptoms of obsessive compulsive disorder (Children's Yale–Brown Obsessive Compulsive Scale Total). Post-hoc tests (comparing between-assessment times: baseline–1 month, 1–2 months, 2–3 months) employed p-values adjusted by Bonferroni correction (0.05/3=0.0167). Significant time-by-age interactions were explored, comparing effects of time in younger (age 7y–8y) and older (age 9y–14y) children.

Results

Group comparisons at baseline

Baseline characteristics were compared between the ITT and LFT groups to determine whether there were meaningful differences. The ITT group had significantly higher (p=0.017) SSS Motor scores (mean 11.70, SD 1.88) than the LTF group (mean 10.63, SD 1.86). Otherwise, there were no significant group differences in terms of baseline age (p=0.341), age of stereotypy onset (p=0.408), sex (p=1.000), race (p=0.613), prior treatment (p=0.711), current medication status (p=0.318), SSS Impairment (p=0.883), SLAS (p=0.590), Children's Yale–Brown Obsessive Compulsive Scale (p=0.081), MASC-2 (p=0.854), or ADHD Rating Scale-IV (p=0.959).

Intent-to-treat group analysis

The ITT group contained 54 participants: mean age (SD) was 8 years and 2 months (1.42, range 7–14y) and included 34 males and 20 females. The sample was 94% Caucasian, 4% Hispanic, and 2% African-American. Mean age at onset for stereotypies was 12.3 months (13.9mo, range <1–60mo). Six participants (11%) reported use of active psychotropic medications – none prescribed for motor stereotypies. Medications included guanfacine (n=1), lamotrigine (n=1), methylphenidate (n=2), and dextroamphetamine (n=2). No participants reported concurrent non-pharmacological treatment for CMS; however, 10 reported prior non-pharmacological treatment for stereotypies, including psychoeducation (n=3) or supportive psychotherapy (n=1), behavioral therapy (n=4), habit reversal training (n=1), or social skills training (n=1).

Within the ITT group, neither age at study entry, nor age of stereotypy onset was significantly associated with any of the stereotypy ratings (all p>0.348). At baseline, higher ratings of ADHD were significantly associated with greater SSS Impairment (r=0.43, p=0.001), and SLAS ratings (r=0.35, p=0.009), but not with SSS Motor. Higher MASC-2 scores were associated with greater SSS Impairment (r=0.47, p<0.001), but not with SSS Motor or SLAS ratings. Obsessive-Compulsive ratings were not significantly associated with any stereotypy ratings.

(a) Stereotypy severity

Primary outcomes were assessed by comparing baseline scores to the last recorded assessment values for the SSS and SLAS scores using paired t-tests. Compared with baseline, there were significant reductions in SSS Motor (p<0.001), SSS Impairment (p<0.001), and SLAS scores (p<0.001). Calculated change ratios revealed a 15% reduction in SSS Motor, 24% reduction in SSS Impairment, and 20% reduction in SLAS scores (Table I).

Table I. Intent-to-treat analysis: comparison of primary outcome measures at baseline and last follow-up assessment.

	Baseline assessment	Last post-DVD assessment		Change ratio
	(n=54a)	(n=54)		(n=54)
	Mean (SD)	Mean (SD)	p	Mean (SD)
SSS Motor	11.70 (1.88)	9.87 (2.65)	<0.001	–0.15 (.21)
SSS Impairment	20.00 (10.46)	14.85 (9.26)	<0.001	–0.24 (.48)
SLAS Score	57.80 (20.43)	41.80 (23.54)	<0.001	–0.20 (.57)

^aParticipants with at least one follow-up assessment provided by parents at month 1, 2, or 3.

SSS, Stereotypy Severity Scale; SLAS, Stereotypy Linear Analog Scale; Change Ratio, mean of (follow-up score – baseline score)/baseline score.

(b) Effects of time (duration of participation) and age at time of treatment

For SSS Motor, GEE analysis revealed a significant effect for time (p<0.001), indicating an overall reduction in symptoms over time (Table II). In addition, there was also a significant time-by-age interaction (p<0.001). Follow-up examination of the interaction revealed that there was a significant reduction in symptoms over time in both the younger group (p<0.001) and the older group (p=0.003). In both the younger and older groups, change was driven primarily by significant baseline versus 1 month differences (both p<0.001), but not differences between 1 and 2 month, or 2 and 3 month reports.

Table II. Parent reports of stereotypy severity according to length of study participation.

	Baseline		1 month		2 month		3 month
	n=81		n=54		n=38		n=29
	Mean	SD	Mean	SD	Mean	SD	Mean	SD
SSS Motor	11.35	1.93	10.57	1.99	10.24	2.38	9.79	2.53
SSS Impairment	2.41	1.01	2.13	0.87	2.08	1.10	1.97	0.82
SLAS Score	56.94	19.94	46.83	21.11	48.39	25.61	40.86	23.76

SSS, Stereotypy Severity Scale; SLAS, Stereotypy Linear Analog Scale.

For SSS Impairment scores, there was a significant effect for time (p<0.001), again showing an overall reduction in reported impairment over time (Table II). There was also a significant time-by-age interaction (p=0.0001). Follow-up examination of the interaction revealed that there was a significant reduction in symptoms over time in the younger group (p<0.001), but not in the older group (p=0.357), with change in the younger group driven primarily by significant baseline versus 1 month differences (p<0.001), but not differences between 1 and 2 month, or 2 and 3 month reports.

For SLAS scores, there was a significant effect for time (p=0.0001), showing an overall reduction in reported symptoms over time (Table II). There was also a significant time-by-age interaction (p<0.0001). Follow-up examination of the interaction revealed that there was a significant reduction in symptoms over time in the younger group (p<0.001), but not in the older group (p=0.077), with change in the younger group driven primarily by significant baseline versus 1 month differences (p<0.001), but not differences between 1 and 2 month, or 2 and 3 month reports.

(c) Secondary outcomes-(PGI-I scores)

The following descriptive statistics pertain to subsets of the 54 participants in the ITT group whose parents completed PGI-I assessments at 1 month (n=44), 2 month (n=45), and 3 month (n=37) follow-up time points. Participants were classified based on parent ratings of perceived change: very much/much improved (PGI-I score 1–2), improved (PGI-I score 3), no change (PGI score 4), and minimally worse/much worse (PGI-I score 5–6). The proportions of parent ratings at the 1, 2, and 3 month reports, respectively, were as follows: very much/much improved 6.8%, 24.4%, and 37.8%; improved 50.0%, 51.1%, and 48.6%; no change 38.6%, 15.6%, and 16.2%; and minimally/much worse 4.5%, 8.8%, and 2.7%. For the 37 participants who remained in the study and completed follow-up PGI-I reports at all three time points, repeated measures analysis of variance (ANOVA) revealed a significant overall effect for treatment duration (p=0.001; mean PGI-I 3.37, SD 0.81; 3.03, SD 0.94; and 2.83, SD 0.89; at 1, 2, and 3 month follow-up points, respectively).

(d) Post-treatment questionnaire

Parents of 24 of 54 participants in the ITT group completed a post-treatment questionnaire. The majority of this group (n=21) had provided follow-up assessments at months 1, 2, and 3, before finishing the post-treatment questionnaire. Participants (n=24) indicated having watched the video multiple times (mean 4.04, SD 2.69). The majority (n=20) found the video ‘useful’ and (96%) would ‘recommend it to others’. Most (n=23) reported using the video without other concurrent behavioral therapy. A subset (8%) reported concurrently being followed by a psychologist (8%), and none reported current contact with a psychiatrist.

Discussion

The presence of motor stereotypies in an otherwise normal child disrupts routine activities and results in social stigmatization. Various behavioral interventions have been used to treat stereotypic behaviors in children with autism,^13,14 and a small published study has demonstrated benefit for children with primary CMS.⁶ Additional studies have also established the potential of family-implemented behavioral treatment for repetitive behaviors in autism.¹⁵ This report describes the preliminary efficacy of an instructional DVD designed for parents to provide home-based therapy for CMS. The selection of a therapy based on the combination of awareness training and the differential reinforcement of other behaviors was selected because of its successful use in face-to-face treatment.⁶

Results suggest that parent-administered therapy effectively reduces primary motor stereotypy severity. Support for this conclusion includes a 15% reduction (from baseline to last assessment) in SSS Motor, 24% reduction in SSS Impairment, and 20% reduction in SLAS scores, and consistently favorable responses on post-treatment questionnaires. These findings are consistent with a parallel in-person, psychologist-administered behavioral treatment for CMS, which showed reductions in SSS motor (14%), SSS impairment (33%), and SLAS (45%) scores.⁶ In terms of parent-rated improvement, 24% of our participants who remained in the study for 2 months were scored very much/much improved, only slightly less than the 30% reported in the largest randomized 8 week controlled trial of face-to-face behavioral treatment for another childhood movement disorder, Tourette syndrome.¹⁶ Baseline reports of anxiety and ADHD symptoms were associated with greater functional impairment from the stereotypies, but not with increased frequency or severity of movements. Participants who completed all assessments achieved significant reductions by the 1 month assessment, with greater relative response observed via SSS Impairment and SLAS ratings in the younger children (age 7–8y). Conversely, age at baseline assessment was not associated with greater stereotypy severity, study participation, or magnitude of improvement. Recognizing this study's reliance on parental report, it is acknowledged that improvement could reflect actual reduction in stereotypies, or potentially decreased parental concern because of increased understanding, or reduced attention to the movements at non-practice times.

Participants in the ITT group were largely indistinguishable from those in the LTF group, except for the former having a greater stereotypy score at baseline. This finding was not unexpected, recognizing that parents who have children with a greater stereotypy burden are more likely to participate in an active treatment process. Although some parents provided a rationale for not starting the home-based therapy (mild symptoms, time commitment, family circumstances, or child anxiety), for the majority the reason for dropping out of the study is unknown. The importance of a family's availability of resources (i.e., time, motivation, and willingness for treatment compliance) is likely a factor in determining who will be most likely to benefit from a DVD-based treatment; future research leading to a better predictive understanding of which families are more likely to complete a home-based therapeutic program would be of value.

Despite the apparent positive treatment response in the majority of active treatment participants, results must be considered in light of several limitations. First, of those participants completing the 1 month assessment, only 70% completed a second, and 51% a third evaluation. Design shortcomings include the lack of a randomly assigned control group and reliance on solely parent-based evaluations. For example, to better control for time actively engaged in treatment activities, as well as to isolate the components of effective treatment, future investigations should include the use of a ‘control’ DVD that contains only psychoeducational information, instructions to refrain from directly commenting on stereotypic movements, and support for patient/parent distress often associated with them. Second, evaluations were entirely parent-based and treatment was parent-delivered, with no objective confirmation by experienced investigators of symptom improvement or adherence to treatment. Future investigations should also include clinician-administered assessments, multiple respondents, and treatment fidelity checks. Lastly, an important next step is to determine if younger children (e.g. preschool children) are also able to benefit from clinic-based or home-based interventions to treat primary CMS. The latter step is essential for recognizing the disparity between time of symptom onset (mean 12.3mo) and age of the participants (mean 8y 2mo) in this study.

In conclusion, there are few treatment options for children with primary CMS. Face-to-face treatment shows promise. Telemedicine increases access and reduces costs, but does not address shortages of appropriately trained providers. The instructional DVD for parent-delivered behavioral therapy may therefore be a safe, effective front-line intervention for primary complex motor stereotypies in children.

What this paper adds.

An educational DVD intervention produced significant reductions in stereotypy severity and impairment.

Initial reductions in symptoms and impairment appeared to be maintained over time.

Abbreviations

CMS

Complex motor stereotypies

ITT

Intent-to-treat (group)

LTF

Lost-to-follow-up (group)

MASC-2

Multidimensional Anxiety Scale for Children-2

PGI-I

Patient Global Impression of Improvement

SLAS

Stereotypy Linear Analog Scale

SSS

Stereotypy Severity Scale