Parent-Based Sleep Education Workshops in Autism

Abstract

To determine if parents can successfully teach their children with autism spectrum disorders to become better sleepers, we piloted small group parent education workshops focused on behavioral sleep strategies. Workshops consisted of three 2-hour sessions conducted over consecutive weeks by 2 physicians. Curricula included establishing effective daytime and nighttime habits, initiating a bedtime routine, and optimizing parental interactions at bedtime and during night wakings. Baseline and treatment questionnaires and actigraphy were analyzed in 20 children, ages 3 to 10 years. Improvements after treatment were seen in the total scale and several insomnia-related subscales of the Children’s Sleep Habits Questionnaire. Actigraphy documented reduced sleep latency in children presenting with sleep onset delay. Improvements were also noted in measures of sleep habits and daytime behavior. Brief parent-based behavioral sleep workshops in children with autism spectrum disorders appear effective in improving subjective and objective measures of sleep, sleep habits, and daytime behavior.

Children with autism spectrum disorders are more likely to experience problems with sleep than typically developing children, with parent reports documenting a prevalence of 53% to 78% (vs 26-32% for typically developing).^1,2 The most prominent sleep problem is insomnia, characterized by difficulty falling asleep, difficulty maintaining sleep, and early morning awakenings.^3-5 In our prior work, using parentally completed questionnaires and polysomnographic measures of sleep, we found that children with autism spectrum disorders characterized as poor sleepers by their parents had prolonged sleep latency (time to fall asleep) and reduced sleep efficiency (time asleep/time in bed), on an initial night of polysomnography, as compared to children described as good sleepers.⁶ In addition, the poor sleepers exhibited more impaired scores on the affective problems subscale of the Child Behavior Checklist^7,8 and on the social communication portion of the Autism Diagnostic Observation Schedule.⁹ Other investigators have related shortened sleep duration to overall autism symptomatology and social interactions,¹⁰ and have correlated child behavior problems to parental reports of past and current child sleep problems.¹¹ Sleep problems in autism spectrum disorders have also been related to repetitive behaviors, although this relationship may be moderated by cognitive level.¹²

The causes of poor sleep in children with autism spectrum disorders appear multifactorial and include aberrations in neurotransmitters that promote sleep, such as serotonin and melatonin, as well as coexisting psychiatric symptoms, such as anxiety.¹³ Poor sleep hygiene (sleep habits) is a less emphasized contributor to insomnia in children with autism spectrum disorders; however, the core behavioral deficits associated with autism spectrum disorders may impede the establishment of sound bedtime behaviors and routines. For example, children with autism spectrum disorders may have difficulty with emotional regulation (eg, ability to calm self) or transitioning from preferred or stimulating activities to sleep. Due to deficits in communication skills, children with autism spectrum disorders may not readily understand the expectations of parents related to going to bed and falling asleep. Adjusting the sleep environment to promote sleep and conveying sleep expectations in an effective way may be particularly challenging for parents of children with autism spectrum disorders dealing with multiple other priorities and stressors.

A literature review of behavioral treatments for sleep problems in children with special needs (encompassing those with physical illness, psychological problems, intellectual disabilities, and neurodevelopmental disorders including autism) concluded that these techniques appear promising in improving sleep as well as daytime functioning of the child and family functioning.¹⁴ This review emphasized, however, that only small studies or case reports have been performed, with inclusion of children having a variety of diagnoses (not limited to autism spectrum disorders) and without collection of objective sleep data. Only small case studies^15-17 have been performed in this area, without sufficient evidence to assess efficacy.

The primary objective of the pilot project was to study the impact of a parent behavioral sleep education workshop on insomnia in children with autism spectrum disorders. A secondary objective was to determine if treating insomnia would improve aspects of the children’s daytime behavior as well as the stress level of their parents.

Methods

Participants

The study was approved by the Vanderbilt University Institutional Review Board. We recruited families having children between the ages of 3 and 10 years with a clinical diagnosis of autism spectrum disorders (autism, pervasive developmental disorder, not otherwise specified, or Asperger’s disorder) who expressed that sleep was a concern. These families were recruited from Vanderbilt University Medical Center subspecialty clinics as well as from the community (eg, local autism society, public schools).

In each child, medical and psychological records were reviewed and a comprehensive sleep and medical history was performed by a neurology sleep specialist (B.A.M.) to exclude primary sleep disorders such as sleep apnea and narcolepsy as well as neurological and medical conditions that may contribute to disordered sleep (eg, epileptic seizures). To be able to generalize our results to the broader population of autism spectrum disorders, children taking medications were included as were those with intellectual disability. The Autism Diagnostic Observation Schedule, performed by a trained psychologist, confirmed the clinical diagnosis of autism spectrum disorders.⁹ Children also received the Peabody Picture Vocabulary Test-III as a measure of receptive language.¹⁸ To avoid selection bias and maximize recruitment and retention of families, on-site child care by licensed personnel was offered to all participating families.

Procedures

In each workshop, 3 to 5 families participated, with a total of 5 workshops conducted in this study. Each workshop consisted of 3 sessions conducted over 3 consecutive weeks, with each session lasting 2 hours. Prior to the first session, baseline data were collected from each family, which included demographics, questionnaires on sleep, behavior, and parental stress completed by the parents, and a week of actigraphy accompanied by sleep diaries from each child.

Workshops were led by a neurology sleep specialist (B.A.M.) and a pediatrician with expertise in autism spectrum disorders (S.M.) with assistance from a nurse educator (K.A.) and an educational consultant (K.F.). The curriculum consisted of presentations, breakout groups and group discussions, and question-and-answer periods (Table 1). Between sessions, parents completed homework in the form of daily data collection sheets, recording their child’s bedtimes, sleep onset times, and wake times as well as daily habits such as napping, exercise, and caffeine consumption. Homework was discussed in a group format at the beginning of sessions 2 and 3, with parents highlighting strategies that were successful as well as continued areas of difficulty.

Table 1.

Curriculum for Workshop

Workshop Session 1: Establishing effective daytime and nighttime habits and a bedtime routine: Introductions. Parents state their major sleep challenge and what they hope to achieve from the workshop. PowerPoint Presentation – Causes of sleep problems in children with autism spectrum disorder. – Goals of educational workshop. – Components of successful sleep: daytime and evening habits (eg, limited caffeine and sufficient exercise), sleep environment (eg, level of light), and bedtime routine. – Sleep amount/timing/regularity—How much sleep does a child need? Discussion of why a bedtime that is too early may hinder sleep, and how regular sleep and wake times promote sleep. Breakout Session – Half of the participants meet with one physician to discuss sleep habits and environment while the others meet with second physician to discuss sleep amount/timing/regularity. – After 15 minutes, participants switch groups. – Physicians target specific areas based on participants’ responses to the Children’s Sleep Habits Questionnaire and Family Inventory of Sleep Habits and discuss strategies related to improving sleep habits, environment, and sleep amount/timing/regularity. PowerPoint Presentation on bedtime routine – What does the bedtime routine accomplish? – Examples of visual schedules. – Use of a timer during the bedtime routine. Video illustrating a successful bedtime routine PowerPoint Presentation on sleep resistance – Causes—“child is not sleepy” vs “child is scared.” – Strategies for the “not sleepy” child—later bedtime, avoidance of late-day naps. Strategies for the “scared” child—adding a dim night light or emphasizing calming activities in the bedtime routine. – Effective parent interactions to minimize sleep resistance (setting expectations while relieving anxiety) including graduated extinction methods (eg, sitting in chair next to child’s bed with back to the child, and gradually moving closer to the door each night) coupled with reward systems. Breakout session on bedtime routine/sleep resistance – Parents complete a worksheet listing activities that make up bedtime routine. Each activity is labeled as to whether it is stimulating or relaxing, and hard or easy for the child. Activities are then ranked in terms of child preference. – Using this worksheet as a guide, staff complete personalized visual schedules for each family, tailored to the preferences (eg, favorite soothing activities are included) and cognitive levels of the child (eg, pictures are used for nonverbal children). – Parents complete a worksheet followed by a group discussion on sleep resistance, including determining whether their child is “not sleepy” or “scared” and strategies that they plan to try in the upcoming week to minimize sleep resistance. – Participants share strategies related to minimizing sleep resistance. Summary session – Parents share the 1 or 2 pearls they have learned this evening that they plan to apply at home. – Delivery of visual support boards and education on how to use the boards. – Wrap-up with discussion of homework for next week. Homework consists of completion of a datasheet, which the parent completes nightly, that includes the elements discussed in this session (eg, bedtime, wake time, components of bedtime routine, strategies for sleep resistance).

Workshop Session 2: Minimizing Night Wakings and Early Morning Awakenings Review of Homework—presentation by each family of what strategies were effective and ineffective during the week with discussion from the group. PowerPoint Presentation on night wakings and early morning awakenings – Concepts that (a) children who can fall asleep on their own at bedtime are more likely to fall back to sleep on their own during a night waking, and that (b) what children are exposed to immediately before sleep (eg, parent present) is what they will likely require to fall back to sleep when spontaneously awakened. The importance of teaching children to fall asleep on their own, without a parent present immediately before sleep, was emphasized. – Strategies for minimizing night wakings, including the bedtime pass,21 which limits the number of parental requests while providing comfort. – The importance of reinforcers, such as morning rewards (eg, wrapped presents from a dollar store or a warm blanket from the dryer) were emphasized. – Early morning awakenings and how they differ from night wakings (eg, child may have obtained the full amount of sleep needed for the night and wakes up ready for the day) – Strategies for minimizing early morning awakenings (eg, later bedtime) or making them less disruptive to the family, such as teaching early-rising children to entertain themselves. Group Discussion on night wakings/early morning awakenings – Parents complete a worksheet followed by a group discussion on night wakings and early morning awakenings and strategies that they plan to try in the upcoming week to minimize sleep resistance. – Participants share strategies related to minimizing sleep resistance. Summary session – Parents share the 1 or 2 pearls they have learned this evening that they plan to apply at home. – Delivery of bedtime passes. – Distribution of homework for upcoming week.

Workshop Session 3: Addressing individualized sleep concerns Review of Homework—presentation by each family of what strategies were effective and ineffective during the week with discussion from the group. Question and answer session to address individual sleep concerns. Discussion of how to get back on schedule when an illness or change in the family routine occurs, description of symptoms of sleep apnea and restless legs syndrome, use of supplemental melatonin, and the role of diet in sleep are examples of some of the most frequently discussed topics. Completion of anonymous program evaluation survey.

Topics in Session 1 included establishing effective daytime and nighttime habits and a bedtime routine. Specific areas related to improving sleep habits were targeted based on participants’ responses to the Children’s Sleep Habits Questionnaire¹⁹ and the Family Inventory of Sleep Habits,²⁰ two validated instruments for assessing sleep concerns. For example, parents were given sheets detailing the amount of caffeine in common foods and beverages so that they could quantify their children’s caffeine intake and instructed to limit caffeine in the afternoon and evening. Ample exercise during the day, with calming, rather than stimulating activities, in the evening was also encouraged. Parents were provided with education on the importance of a regular bedtime that was not too early, and assisted in selecting a bedtime that coincided with the child’s optimal time for sleep. In addition to formal presentations on these topics, participants were provided with the opportunity to discuss these elements within a group setting, sharing with the other participants their individualized experiences. Each parent was individually guided in selecting a bedtime routine which would promote sleep in their child. Visual supports to reinforce the routine were provided and parents were trained in the use of the visual supports by the instructors; a video on the bedtime routine was also presented to the group. The causes of sleep resistance, and strategies to minimize it (eg, graduated extinction or gradually eliminating the presence of a parent in the room at bedtime) were presented and discussed in a group format.

Session 2 focused on minimizing night wakings and early morning awakenings. Presenters discussed the concepts that (a) children who can fall asleep on their own at bedtime are more likely to fall back to sleep on their own during a night waking and that (b) what children are exposed to immediately before sleep (eg, parent present) is what they will likely require to fall back to sleep when spontaneously awakened. The importance of teaching children to fall asleep on their own, without a parent present immediately before sleep, was emphasized. Strategies to reduce the number and duration of night wakings, including use of a bedtime pass^21,22 and morning rewards, were also discussed. Parents were given bedtime passes during the session to use at home. The causes of early morning awakenings and strategies for minimizing them, or making them less disruptive to the family, were also discussed.

Session 3 addressed individualized sleep concerns, including how to get back on schedule when an illness or change in the family routine occurred. Medical causes of sleep disorders, including symptoms of sleep apnea and the use of medication to promote sleep, were also discussed. Completion of an anonymous program evaluation survey was done at the end of Session 3.

Detailed descriptions of the concepts presented in the curriculum are outlined in Table 1 and discussed in depth in a review article¹³ and book.²³

Approximately 1 month after the final session of the workshop, parents were asked to repeat the baseline questionnaires, and a second week of posttreatment actigraphy accompanied by sleep diaries was collected from each child. Parents were compensated with a gift card after returning each set of questionnaires, and children received a gift card for each week they wore the actigraph.

Measures

Children’s Sleep Habits Questionnaire¹⁹

The Children’s Sleep Habits Questionnaire is a validated parentally completed questionnaire that has been used to examine sleep behavior in toddlers, preschool and school-aged children with a variety of conditions, including autism spectrum disorders.^6,24 Subscales of the Children’s Sleep Habits Questionnaire measure insomnia-related dimensions such as bedtime resistance, sleep anxiety, sleep onset delay, sleep duration, and night wakings, as well as other dimensions such as daytime sleepiness, sleep-disordered breathing, and parasomnias. A total score is also calculated, and we previously reported the use of a modified total score incorporating the insomnia domains (total of items comprising the bedtime resistance, sleep anxiety, sleep onset delay, sleep duration, and night wakings scales).²⁰ The Children’s Sleep Habits Questionnaire was used to measure changes in subscales and total scores after our behavioral intervention. In combination with the Family Inventory of Sleep Habits and the sleep history, the Children’s Sleep Habits Questionnaire was also used to define specific sleep challenges present in each child, which were targeted during the workshop sessions.

Family Inventory of Sleep Habits

The Family Inventory of Sleep Habits (Table 2) is a parent-report questionnaire which assesses sleep hygiene in their children (habits which promote sleep). This includes attention to developing a structured and consistent bedtime routine, sleep environment (eg, level of light and background noises), daytime habits (eg, caffeine use and exercise), and parental interactions at bedtime and upon night wakings. Parents are asked to rate the frequency of sleep habits over the last month on a 5-point scale. We previously validated a 12-item research version of the Family Inventory of Sleep Habits.²⁰ In our parent education classes, we used the total 22-item Family Inventory of Sleep Habits (Appendix[Pls Provide the appendix.]), which provides a comprehensive overview of sleep habits. Along with the Children’s Sleep Habits Questionnaire, we used the Family Inventory of Sleep Habits within the workshop setting to counsel parents on specific strategies to improve sleep hygiene.

Table 2.

Family Inventory of Sleep Habits: Full Version

Directions: For each item below, please indicate how often it was true within the last month: (1) Never (2) Occasionally, (3) Sometimes, (4) Usually, (5) Always. My child gets exercise during the day. My child naps more than 1 hour during the day. My child’s bedroom is used as a “time-out” area for discipline during the day. My child’s bedroom is used as a play area during the day. My child wakes up at about the same time each morning. In the hour before bedtime, my child engages in relaxing activities. My child has drinks or foods containing caffeine after 5 PM (examples: chocolate, Coca-Cola). In the hour before bedtime, my child engages in exciting or stimulating activities (examples: rough play, video games, sports). My child sleeps better when wearing pajamas/sleepwear made from certain fabrics. My child sleeps better with certain sheets or blankets on his/her bed. My child sleeps better when his/her room is a certain temperature at night (either warm or cool). My child’s room is dark or dimly lit at bedtime. My child’s room is quiet at bedtime. My child goes to bed at the same time each night. My child follows a regular bedtime routine that lasts between 15 and 30 minutes. My child has a favorite comfort object that he or she sleeps with. I stay in my child’s room until he or she falls asleep. After my child is tucked in, I check on him/her before he or she falls asleep. My child watches TV, videos, or DVDs to help him/her fall asleep. My child listens to music to help him/her fall asleep. If my child wakes up during the night, I keep our interactions brief. If your child does not wake up during the night, circle here: N/A If my child gets out of bed during the night, I return him/her to his/her own bed. If your child does not get out of bed during the night, circle here: N/A

Parental Concerns Questionnaire²⁵

The Parental Concerns Questionnaire is a validated parentally completed questionnaire that has been used to assess the presence and severity of 13 developmental and behavioral concerns expressed by parents of children with autism spectrum disorder. Domains on the Parental Concerns Questionnaire include those related to core symptoms of autism spectrum disorder (eg, language delay and social interactions) as well as related symptoms (eg, hyperactivity and compulsive behavior).

Repetitive Behavior Scale–Revised²⁶

Repetitive Behavior Scales–Revised is an observer-completed questionnaire validated in adults with autism spectrum disorders. A total score as well as subscale scores for stereotyped, self-injurious, compulsive, ritualistic, sameness, and restricted behaviors are determined.

Parenting Stress Index–Short Form²⁷

The Parenting Stress Index–Short Form is a 36-item abbreviated version of the Parenting Stress Index which provides a total stress score as well as the subdomain scores of parental distress, difficult child, and parent-child interactions. The Parenting Stress Index–Short Form has been used to measure parental stress in autism spectrum disorder.²⁸

Actigraphy

Actigraphy is a flexible, cost-effective, and minimally intrusive method for estimating sleep-wake patterns based on physical activity. This method is ideal for children with autism spectrum disorders, who are more prone to sensory sensitivities. Actiwatch-64 actigraphs (Mini-Mitter, Respironics, Bend, Oregon) were used. All movements greater than 0.5 g were sampled at the high sensitivity level within 1-minute epochs. Parents were instructed to place the watch on their child’s dominant wrist. For younger or lower-functioning children, however, we found that it was easier for the child to wear the watch on his or her dominant ankle, a method validated by prior investigators.²⁹

At the conclusion of each baseline and posttreatment week, the raw data were downloaded from the actigraph and scored with Actiware Software Version 5.0 (Mini-Mitter, Respironics). The following variables were averaged for each evaluation period: sleep latency, total sleep time, sleep efficiency, and wake time-after-sleep-onset. Sleep diaries and event markers were used to set lights out (time child first attempted sleep) and lights on (when child awakened for the day).

Data Analysis

Data were double-entered and verified in REDCap (Research Electronic Data Capture), a secure online database administered by Vanderbilt University Medical Center, and were analyzed in SPSS Version 15. Paired baseline and after-treatment data were compared using the Wilcoxon signed-rank test; this nonparametric test was used because the outcomes measured (ordinal variables) do not necessarily follow the normal distribution. Exact P-values (2-sided) less than .05 were considered statistically significant; nonsignificant trends less than .10 were also reported given the pilot nature of this study. Spearman correlations were used to determine the relation of Peabody Picture Vocabulary Test-III scores to Children’s Sleep Habits Questionnaire scores. We did not adjust for multiple comparisons, given the small sample size and the pilot nature of the study. However, we chose to interpret significant results and nonsignificant trends conservatively within the scientific context.

Results

Participants

In all, 25 families enrolled in the study, and 22 families completed the parent education training. Of the 3 noncompleters, 2 withdrew after enrollment due to other family priorities and 1 child did not meet Autism Diagnostic Observation Schedule criteria for autism spectrum disorder. Of the 22 families, 2 who completed the training were unable to be contacted for the return of the posttreatment data despite multiple attempts by phone and letter, leaving 20 families for whom baseline and posttreatment questionnaire data were available. The analysis of these 20 families is presented below.

Mean age of participants was 5.8 ± 2.7 years, with 15 children having Autism Diagnostic Observation Schedule scores in the autism range and 5 children in the autism spectrum range. The mean Peabody Picture Vocabulary Test standard score was 74.6 ± 25. Our participant population included 4 girls and 5 ethnic minorities (ie, 4 African-American and 1 Asian). Seven of our participating children came from single-parent homes (one of whom had a sibling), and 13 families came from dual-parent homes (10 of whom had at least 1 sibling). Six children were taking psychotropic medications, including risperidone, amphetamine/dextroamphetamine, dexmethylphenidate, lisdexamfetamine, oxycarbazepine, and clonidine.

Sleep Concerns

Difficulty falling asleep was the most common concern in our cohort, noted in 75% of children, followed by night wakings in 60%, and early morning awakenings in 15% of children. Cosleeping with parents occurred in 35% of children, most with night wakings, with their parents stating that cosleeping was not their preference but served as a solution to minimize night wakings. Parents described their children with difficulty falling asleep as having trouble “settling down” or “winding down,” with some children crying and screaming but others engaging in specific behaviors (eg, playing with mother’s hair or rubbing tags on clothing). Children who were verbal tended to speak incessantly about a favorite interest (eg, cartoon scripts), sing to themselves, or make repeated requests of their parents (eg, drink of water). Parents reported that sleep problems were present since birth in 50% of the children, with the other 50% occurring around the time of the autism diagnosis.

Intervention Results

Children’s Sleep Habits Questionnaire

In a paired group analysis comparing baseline measures with treatment measures, several of the insomnia subscales improved significantly on the Children’s Sleep Habits Questionnaire (Table 3), including sleep onset delay (children took less time to fall asleep), sleep duration (children slept longer), bedtime resistance, and sleep anxiety. Although parents reported that night wakings were shorter and their children were able to fall back to sleep more easily, night wakings still persisted in the majority of children, with the night wakings subscale not improving significantly with treatment. Both the modified and total scores on the Children’s Sleep Habits Questionnaire also showed improvement, while noninsomnia-related domains, including parasomnias, sleep-disordered breathing, and daytime sleepiness did not change. The Peabody Picture Vocabulary Test-III did not correlate significantly with improvements in the modified (r = −0.11; P = .66) and total scores (r = .06; P = .80).

Table 3.

Children’s Sleep Habits Questionnaire Subscales^a

	Baseline		Treatment			Normative Sample
	Mean	SD	Mean	SD	P-Valueb	Mean	SD
Bedtime resistance	10.50	3.65	8.45	2.84	.001	7.06	1.89
Sleep onset delay	2.30	0.80	1.50	0.61	.004	1.25	.53
Sleep duration	6.53	2.22	4.79	2.07	.003	3.41	.93
Sleep anxiety	7.00	2.27	5.85	1.73	.022	4.89	1.45
Night wakings	5.55	1.90	4.85	1.79	.508	3.51	.89
Parasomnias	10.50	2.24	10.05	2.37	.607	8.11	1.25
Sleep-disordered breathing	3.42	0.84	3.32	0.75	.625	3.24	.63
Daytime sleepiness	13.95	4.32	13.25	3.43	.096	9.64	2.80
Modified Children’s Sleep Habits Questionnaire total	28.63	6.15	22.84	5.43	.000
Total Children’s Sleep Habits Questionnaire score Normative data19	56.63	9.21	49.74	9.24	.004

^aThe Children’s Sleep Habits Questionnaire is a validated parentally reported measure of a child’s sleep, encompassing a variety of domains listed above. A higher score on a subscale means more difficulty (eg, a higher score on sleep duration means shorter sleep).

^bWilcoxon signed-rank tests with exact (2-sided) P-values were used.

Five of seven children reported to cosleep with parents at baseline showed improvement, with fewer nights of cosleeping reported. Only 1 of 3 children with early morning wakings was reported to improve in this area.

Family Inventory of Sleep Habits

Several items showed significant improvement on the Family Inventory of Sleep Habits with treatment (Table 4), including those related to the child engaging in relaxing, rather than stimulating activities before bedtime. Having a regular bedtime routine and not using electronic sleep aids (TV and videos) also showed improvement after treatment. After treatment, the parents reported staying in their child’s room less often until he or she fell asleep, and returning their child to the child’s bedroom more often upon awakening.

Table 4.

Family Inventory of Sleep Habits Scores^a

	Baseline		Treatment
	Mean	SD	Mean	SD	P-Valueb
1. Exercise during the day	4.25	0.85	4.50	0.69	.063
2. Napping during the day	1.95	1.00	1.85	0.99	.687
3. Bedroom used for “time-out” during day	1.15	0.37	1.15	0.37	1.000
4. Bedroom used as play area during day	3.05	1.23	2.60	1.19	.070
5. Wakes up about same time each day	3.70	0.86	4.05	0.22	.289
6. Relaxing activities before bed	2.95	1.19	4.05	0.60	.001
7. Caffeine after 5 PM	1.70	0.86	1.50	0.51	1.000
8. Stimulating activities before bed	2.60	1.05	1.80	0.95	.001
9. Sleeps better with certain sleepwear	2.22	1.44	2.11	1.23	1.000
10. Sleeps better with certain sheets	2.37	1.38	2.47	1.47	1.000
11. Sleeps better when room certain temperature	2.44	1.15	2.50	1.34	.727
12. Room is dark or dimly lit at bedtime	4.61	1.04	4.78	0.43	1.000
13. Room is quiet at bedtime	4.17	1.15	4.56	0.51	.453
14. Goes to bed at same time each night	3.94	1.16	4.06	0.87	1.000
15. Follows regular bedtime routine (15-30 minutes)	3.39	1.24	4.17	0.79	.022
16. Has a favorite comfort object for sleep	2.94	1.86	2.83	1.62	.754
17. Parent stays in child’s room until sleep	2.56	1.65	1.94	1.30	.039
18. Parent checks on child before he or she falls asleep	3.11	1.60	2.72	1.32	.065
19. Child watches TV, videos, or DVDs to fall asleep	2.28	1.36	1.50	1.04	.004
20. Child listens to music to fall sleep	2.61	1.46	2.17	1.34	.344
21. Parent keeps interactions brief if child awakens	4.00	1.29	4.46	1.13	.289
22. Parents returns child to bed if child awakens	2.82	1.33	4.36	0.67	.008

^aThe Family Inventory of Sleep Habits is a validated parentally reported measure of a child’s sleep habits.

^bWilcoxon signed-rank tests with exact (2-sided) P-values were used.

Actigraphy

Baseline and treatment actigraphy data were available in 12 of these 20 families who completed the parent education training. The predominant cause of missing actigraphy data was the actigraph not worn due to poor tolerance; less commonly, watch malfunction occurred or there was insufficient diary information to interpret the actigraph. In 11 of 12 children with baseline and treatment actigraphy data, the actigraph was worn on the wrist; in only 1 child, the actigraph was worn on the ankle due to intolerance at the wrist site.

Of the 12 children for whom actigraphy data were available, 9 had problems with sleep initiation, taking at least 20 minutes to fall asleep on baseline actigraphy recordings. In this group of 9 children, the mean sleep latency (±standard deviation) decreased significantly from 62.2 ± 33.3 minutes at baseline to 45.6 ± 27.6 minutes with treatment (P = .039). The other 3 children fell asleep in less than 20 minutes on baseline actigraphy recordings, and were not reported by their parents to have difficulty falling asleep. Only 5 of the 12 children for whom actigraphy data were available had problems with night wakings; 2 of these 5 children improved by both parent report and actigraphy, 2 were not improved (with neither parent report nor actigraphy showing improvement) and 1 was improved by parent report but not by actigraphy. In this group of 5 children, the wake time after sleep onset was 24.5 ± 9.8 minutes at baseline and 32.2 ± 24.7 minutes with treatment (P = 1.0).

When the complete group of 12 children for whom actigraphy data were available were analyzed together, time in bed significantly improved (decreased) from 575.21 ± 53.5 to 541.91 ± 36.2 minutes (P = .039) without significant changes noted in sleep latency, sleep efficiency, sleep duration, or wake time after sleep onset (P > .1).

Behavior Questionnaires

On the Parental Concerns Questionnaire (Table 5), significant improvements were seen in hyperactivity and self-stimulatory behavior, in addition to sleep disturbance, with a compulsive behavior showing a nonsignificant trend toward improvement. On the Repetitive Behavior Scale, the restricted behavior scale total significantly improved with treatment (5.11 ± 3.6, baseline; 3.89 ± 3.1, treatment; P = .007), without improvement seen in the other scales.

Table 5.

Parental Concerns Questionnaire Subscales^a

	Baseline		Treatment
Subscale	Mean	SD	Mean	SD	P-Valueb
Language use/understanding	3.00	1.12	2.85	1.04	.375
Compulsive behavior	2.20	0.89	1.80	0.83	.065
Anxiety	2.30	0.98	2.05	0.94	.180
Sensory issues	2.85	0.99	2.70	0.73	.581
Sleep disturbance	3.40	0.94	2.20	1.01	.000
Aggression	2.15	1.04	1.95	0.94	.453
Hyperactivity	3.15	1.04	2.60	0.99	.006
Attention span	2.90	0.79	2.95	0.76	1.000
Mood swings	2.10	1.02	1.85	0.93	.388
Eating habits	3.00	1.12	2.70	1.26	.289
Social interactions	2.80	0.95	2.55	0.89	.388
Self-stimulatory behavior	2.47	1.17	2.05	0.91	.021
Self-injurious behavior	1.60	0.88	1.45	0.69	.375

^aThe Parental Concerns Questionnaire is a 13-item, parent-completed scale of specific problems that a child experiences, rated on a scale including none, mild, moderate, or severe. A higher value on a subscale means more difficulty in that problem area.

^bWilcoxon signed-rank tests with exact (2-sided) P-values were used.

Parenting Stress Index–Short Form

The total Parenting Stress Index–Short Form score and the subscales of parental distress and parent-child interaction did not change significantly with treatment, with a nonsignificant trend toward improvement noted on the difficult child subscale (Table 6). One item on the scale did demonstrate significant improvement with treatment—this item was related to ease in establishing a sleeping or eating schedule (Table 6).

Table 6.

Parenting Stress Index–Short Form Scores^a

	Baseline		Treatment
	Mean	SD	Mean	SD	P-Valueb
Total scale	96.10	23.40	94.00	23.00	.263
Parental distress	32.05	11.08	31.10	10.58	.503
Parent-child interaction	25.90	6.70	26.50	8.58	.648
Difficult child	38.15	8.80	36.40	7.79	.064
Ease in establishing sleeping/eating schedule	4.05	1.03	3.32	1.33	.021

^aThe Parenting Stress Index–Short Form is a 36-item, parent-completed scale of specific questions related to parenting stress. A higher value on a subscale means greater stress.

^bWilcoxon signed-rank tests with exact (2-sided) P-values were used.

End-of-Workshop Parent Survey

Of the 20 families participating in the workshop, 18 completed anonymous evaluations. Results from the evaluations revealed that all participating parents found the presenters in the workshop to be competent and knowledgeable and would recommend the workshop to other parents. Seventy-seven percent of parents believed that their children’s sleep habits improved. Ninety-four percent of parents felt that the conveying of information (orally and with the use of handouts and examples) was relevant and useful. Fiftyeight percent of the parents felt that the total duration of the workshop was sufficient.

Discussion

Previous reports using behavioral methods to improve sleep in autism spectrum disorders have been limited to case reports and small case series,^15,16 and relied on parental reports of improvement. Our findings are unique in showing that sleep education workshops for parents of children with autism spectrum disorders improved measures of sleep in their children, across a range of receptive language levels; children having low receptive language levels also responded to the intervention. Those taking psychotropic medications also responded to the intervention. This finding suggests that sleep education may minimize the potential adverse effects of medications commonly used in this population, broadening this program’s use. An objective sleep parameter, actigraphy, also showed improvement in a subset of children with sleep onset insomnia. Time in bed decreased in all children in whom actigraphy data were available, suggesting that parents were successfully implementing the recommendation to delay bedtime to promote sleep. In addition, parents reported improvement in their children’s sleep habits after treatment, as well as improvement in hyperactivity, self-stimulatory behavior, and restricted behaviors. Parental satisfaction with the workshops was high, with parents acknowledging more skill in establishing a regular sleeping and eating schedule.

Our work contributes to a growing body of literature supporting behavioral methods to improve sleep in neuro-developmental disorders. Behavioral treatments for sleep problems in typically developing children are effective, as summarized in an American Academy of Sleep Medicine review³⁰ and guideline.³¹ A task force of the American Academy of Sleep Medicine also advocated implementing medications in conjunction with empirically based behavioral treatment strategies and adequate sleep hygiene.³² In children with neurodevelopmental disorders, this approach is especially prudent given that these children may be more sensitive to adverse effects of medications and have difficulty communicating these adverse effects to caregivers.

In our study, children with sleep onset insomnia improved in both subjective (Children’s Sleep Habits Questionnaire scores) and objective measures (actigraphy) after treatment. We demonstrated ability to teach sleep expectations even to children with limited receptive language. In contrast, children with night wakings did not improve on Children’s Sleep Habits Questionnaire sub-scales or by actigraphy. Lack of improvement in this group on the Children’s Sleep Habits Questionnaire subscale or by actigraphy may be due to several factors including (a) the workshop may have been more effective for treating sleep onset delay, (b) the small sample size of children with night wakings may have precluded us from detecting an improvement, or (c) the instruments used may have been less sensitive for detecting night wakings than sleep latency. A recent paper found actigraphy to miss night wakings detected by video.²⁹ Larger studies containing more participants and actigraphic data will be important to determine if improvements in night wakings are detected as reliably as improvements in sleep latency.

Family stress has also been related to sleep problems in children,³³ with parents of children with autism spectrum disorders who have sleep problems experiencing higher levels of parenting stress than those whose children did not have sleep problems.²⁸ The impact of behavioral sleep interventions in children with autism spectrum disorders on family stress has not previously been assessed in a systematic fashion. Although we did not find statistically significant improvements on the Parenting Stress Index–Short Form total or subscales in our sample, it is possible that our sample size was too small to detect an improvement. Alternative explanations for the lack of significant improvement on the Parenting Stress Index–Short Form in our study include the possibilities that family stress contributes to the sleep problem (and therefore may not be significantly improved when the sleep problem is ameliorated) or that parenting stress in autism may be due to other factors besides sleep. Future interventional studies involving larger numbers of participants will be helpful in better defining the relation of sleep to family stress in autism spectrum disorders.

Apart from the small sample size, our study has several other limitations. First, it lacked a control group of children not receiving the intervention. As the literature on using behavioral techniques to treat sleep problems in children with neurodevelopmental disorders has been limited to small case series, we chose to perform this noncontrolled pilot study, in which all children were treated, as a starting point to provide support for future controlled studies. Second, we relied on parentally completed questionnaires to measure improvement in behavior and parental stress. Although this reliance limits our ability to draw conclusions on daytime behavior and family stress given a tendency of parents receiving any intervention to report improvements, the lack of significant improvement on Children’s Sleep Habits Questionnaire subscales related to sleep-disordered breathing and parasomnias indicates that parents were not noting global improvements in all subscales, but ones specific to their child’s insomnia. In future controlled studies, it will be important to obtain independent assessments of child behavior (eg, teacher reports). Third, to measure many domains of behavior, we used the Parental Concerns Questionnaire, an autism-specific instrument which samples different aspects of a child’s behavior broadly. This choice was made because we did not want to burden parents with multiple measures and instead, in this pilot study, wished to establish which domains (eg, hyperactivity) were most sensitive to change. In future studies, we will include measures designed to assess change in the identified domains over the period of the study such as targeted scales to measure improvements in symptoms of attention deficit-hyperactivity disorder. Fourth, apart from the small sample size, actigraphy data were not recovered in all participants. Future studies should take measures to ensure recovery of a higher proportion of actigraphy data. In our ongoing studies of sleep and autism, we are maximizing actigraphy data collection by obtaining 2 to 3 weeks of data and reviewing the data several days into collection to ensure that the device is functioning properly and the family is complying with the recommended procedures. These procedures include keeping the watch on the child most of the day or, if necessary to remove for the child’s tolerance, placing it back on the child at least 30 minutes prior to bedtime to detect sleep onset. We are also having children with tactile sensitivities who cannot tolerate wearing the watch on their wrist and wear it within a shirt pocket, a technique which in our pilot data and that of others produces reliable results when compared with wrist actigraphy (Dr Margaret Souders, personal communication, July, 2008). Finally, this study did not obtain polysomnography to evaluate for primary sleep disorders contributing to insomnia, although a sleep specialist took a comprehensive sleep history in each child. Furthermore, in prior work using polysomnography, we found that coexisting sleep disorders contributing to insomnia are uncommon in this population.⁶

In addition to performing larger controlled studies, future directions for this work include (a) expansion of the population to adolescents with autism spectrum disorders, with involvement of the adolescents in the program as well as their parents, (b) establishment of the parent education program in community settings, such as pediatrician offices, with a teaching role for nonphysician educators, including nurses and educational consultants, and (c) incorporation of children with other neurodevelopmental disorders that have prominent sleep components. In future studies, we would also like to determine the effect of improving sleep patterns on daytime cognitive and behavioral functioning and to document the length of time that nighttime and daytime effects are maintained after training is completed.

In summary, this pilot parent-based behavioral sleep intervention appears feasible and effective in improving sleep and daytime behavior in children with autism spectrum disorders. Larger, more controlled investigations of parent-based behavioral sleep interventions for children with autism spectrum disorders appear warranted.