The Zurich 3-Step Concept for the Management of Behavioral Sleep Disorders in Children: A Before-and-After Study

Abstract

Objectives:

Several strategies have been found to be effective for the treatment of childhood behavioral sleep disorders. One which has yet to be evaluated is the Zurich 3-step concept, which combines basic notions of the two-process model of sleep regulation (introducing a regular rhythm and adjusting bedtime to sleep need) with behavioral strategies. This uncontrolled before-and-after study describes our concept and its step-wise approach, assesses changes in sleep-wake variables and behavior problems, and also examines associations between changes in sleep-wake variables and behavior problems.

Methods:

A total of 79 children with sleep problems (age range 6–47 months, 42% females) were included. Sleep problems were assessed by the Infant Sleep Questionnaire, sleep-wake variables by diary and actigraphy, and behavior problems of children ≥ 18 months by the Child Behavior Checklist.

Results:

A significant decrease in nocturnal wake duration (Cohen's d = −0.34) and a significant increase in the duration of the longest continuous nocturnal sleep period (Cohen's d = 0.19) were found from before to after intervention (on average 2.7 months, SD 1.5). The variability for sleep onset and end time decreased, and actigraphically measured circadian rest-activity cycle measures improved. Parent-reported internalizing and total behavior problems also decreased (Cohen's d = 0.66).

Conclusions:

The findings of both objective and subjective assessment techniques suggest that the Zurich 3-step concept is effective. Thus, the intervention concept may be useful in clinical practice with sleep-disordered children.

Citation:

Werner H, Hunkeler P, Benz C, Molinari L, Guyer C, Häfliger F, Huber R, Jenni OG. The Zurich 3-step concept for the management of behavioral sleep disorders in children: a before-and-after study. J Clin Sleep Med 2015;11(3):241–249.

Behavioral sleep disorders in children, such as bedtime resistance, difficulties falling asleep, and frequent or prolonged nocturnal awakenings, are among the most common parental complaints to health care professionals in the first few years of a child's life.¹ Although many of these problems are transient in nature, they may have an impact on children's health and development and on the family, affecting parent-child interactions and parental well-being.^2–5 Thus, attention should be paid to these problems in clinical practice to prevent negative interactions between the child and the parents and to avoid the development of chronic sleep problems.⁶

The origins of behavioral sleep disorders are broad, encompassing biological, neurodevelopmental, and constitutional factors that may be influenced by environmental and cultural aspects (for an overview of involved factors, see the transactional model⁷). It has been suggested that the individual's own characteristics need to be in accord with the environment and its expectations and demands for optimal development and adaptive functioning.^8,9 Thus, in the context of childhood behavioral sleep disorders, how the child is encouraged to sleep should be congruent with aspects of the child's biological sleep regulation.^10,11

Sleep and wakefulness are regulated by two biological processes: homeostatic and circadian.^12,13 The homeostatic process is responsible for the accumulation of sleep pressure during wakefulness (the longer one is awake, the greater one's sleep pressure) and its reduction during sleep. In contrast, the circadian process is independent of the duration of previous sleep; it is a clock-like process promoting arousal during the day and sleep during the night. These two processes oppose each other, so that the increase in circadian alertness across the day counteracts the increase in homeostatic sleep pressure during wakefulness; during the night, an increase in circadian sleep tendency counteracts the declining sleep pressure, thus ensuring sleep continuity.

BRIEF SUMMARY

Current Knowledge/Study Rationale: The Zurich 3-step concept combines basic notions of the two-process model of sleep regulation (introducing a regular rhythm and adjusting bedtime to sleep need) with behavioral strategies to resolve learned maladaptive sleep habits for the management of behavioral sleep disorders in children. In this study, significant effects from before to after the intervention were found for objectively und subjectively measured sleep-wake variables and for child behavioral problems.

Study Impact: Our study may enrich the literature on interventions in sleep-disordered children by describing how aspects of the two-process model of sleep regulation may be integrated into a clinical intervention concept.

A number of strategies have been developed for the treatment of behavioral sleep disorders in children. In a review of 52 studies utilizing Sackett's criteria for evidence-based treatment, empirical data from controlled group studies provide strong support for unmodified extinction (i.e., ignoring the child's crying and tantrums, also called the “cry it out” approach) and parental education (providing advice about positive bedtime routines, consistent sleep schedules, and parental responses to nocturnal awakenings).¹⁴ However, no study has yet been published assessing the effects of an intervention that firstly focuses on the basic concepts of the two-process model of sleep regulation to support the adjustment of parental expectations to the homeostatic and circa-dian processes, and secondly uses behavioral strategies to handle elements of maladaptive learned sleep behavior. Such a step-wise approach is used in the Zurich 3-step intervention concept (see Table 1), which has been used for more than 30 years in our outpatient clinic and was first described by Largo and Hunziker.¹⁵

Table 1.

Overview of sleep-related information provided to parents of children with behavioral sleep disorders.

The goal of the current study is to describe the three steps of our intervention and to assess changes in sleep-wake variables before and after the intervention using subjective and objective measures. For children older than 18 months, our study also assesses changes with respect to behavior problems and compares their level with healthy references before the intervention. Based on our clinical experience, we hypothesized that the duration of nocturnal awakenings would significantly decrease and that the longest continuous nocturnal sleep duration would increase during the course of our intervention. Furthermore, we expected more regular sleep onset and sleep end times and improved circadian rest-activity cycle measures after the intervention. Finally, we hypothesized that sleep-disordered children would generally show more behavior problems than healthy children, and that fewer behavior problems would be found after the intervention.

METHODS

Participants

Parents of children between 6 months and 4 years of age with behavioral sleep problems who attended outpatient sleep consultations at the Child Development Center of the University Children's Hospital Zurich, Switzerland, were consecutively recruited over a period of 2 years. Parents were referred to our center by health care professionals (e.g., pediatricians, psychologists). Additional inclusion criteria were Swiss residence and having at least one German-speaking parent. Children were excluded if they had other medical or developmental issues (physician ratings). Participant flow is illustrated in Figure 1.

Figure 1. Participant flowchart.

Of the 171 children treated during the study period, 38 children were excluded (on the grounds given in Figure 1). Comparisons between eligible and non-eligible children displayed no significant differences in terms of child's gender (χ² = 0.01, p = 1.00) or age (z = −1.26, p = 0.21). Of the 133 eligible chil -dren, 51 parents declined participation in the study. Participants did not differ from non-participants in terms of child's gender (χ² = 0.17, p = 0.72) or age (z = −0.21, p = 0.83). At baseline (T0), 3 children were dropped from the study due to missing or incomplete diary data. Thus, the final sample comprised 79 children, of whom 26 children were older than 18 months. At follow-up (T1), the sample consisted of 69 children. A total of 10 children were dropped from the study due to missing or incomplete diary data. Actigraphy data were collected from a subsample of 42 children with complete diary data. These 42 children did not differ from the 37 children excluded from the subsample (mainly due to missing or incomplete actigraphy data) in terms of gender (χ² = 1.88, p = 0.25), age (z = −0.32, p = 0.75), socioeconomic status (z = −0.58, p = 0.56), or parent-reported nocturnal awakenings (χ² = 1.03, p = 0.41).

Procedure

The study was approved by the Ethical Review Board of the Canton Zurich, Switzerland, and was performed in full accordance with the Declaration of Helsinki. Immediately after registration, parents were approached by a developmental pediatrician. Following oral consent, sleep problems were assessed by a standardized questionnaire in a phone interview with the parents. Afterwards, informed consent, a diary, an actigraph and standardized questionnaires were mailed to participating parents. They were requested to complete the diary for each day before the first consultation and to attach the actigraph to their child's non-dominant wrist. The device was to be worn all the time and removed only for activities in which it might become wet (e.g., showering). In addition, parents were asked to complete the questionnaires at home and to bring them together with the diary, actigraph and written informed consent to the hospital. After the first outpatient consultation, parents were asked to continue completing the diary and to keep the actigraph on their child until the second consultation. Sleep assessments at T0 were carried out immediately prior to the first consultation; at T1, on average 5 days after the first consultation. Child behavior problems were followed up after the end of all consultations, which was on average 2.7 months after the first consultation. Questionnaires for the follow-up assessment were mailed to participating parents.

Measures

Sleep-Wake Variables

The diary is a widely used tool to assess sleep and wake within 24-h periods and has been shown to be a valid source of information for sleep patterns in healthy and sleep-disordered children.^16,17 Periods of daytime and nighttime sleep were recorded in 15-min intervals, indicated by a continuous line, sleep onset time by the beginning and sleep end time by the ending of the continuous line (see example¹⁶). Parents were also asked to indicate periods when the child was sick, when the actigraph was taken off, or when unusual external motion occurred that could mask sleep (e.g., sleeping in the buggy).

Actigraphy is a nonintrusive method for objectively estimating individual's sleep patterns based on physical activity recordings under non-laboratory conditions. In this study, the Actiwatch Plus (AW4, Cambridge Neurotechnology, Cambridge, UK) was used, and intensity and duration of physical activity were recorded in 1-min epochs. The Actiwatch Activity and Sleep Analysis 7 software (version 7.31) was applied to score each epoch of activity as either sleep or wake using a medium sensitivity threshold.¹⁸ As used in the standardized algorithm of this software package, a particular epoch was scored as wake or sleep by comparing activity counts for the epoch in question and those epochs surrounding it. If the number of counts exceeded the threshold sensitivity, the epoch was scored as wake; otherwise the epoch was scored as sleep. In our study, we corrected for 2 measurement errors. First, missing data points of 10 min up to 3 h during daytime were replaced by averaged activity value for the day in question correcting false daytime sleep.¹⁹ If the data loss was > 3 h, then the particular day was excluded from the analysis; gaps < 10 min were not adjusted. Second, when diary data indicated daytime sleep but unusual external motions were recorded by actigraphy, the data were replaced with 0 activity counts to correct a false record of daytime wakefulness. A scoring interval of 30 min before the diary-reported sleep onset time to 30 min after the diary-reported sleep end time was used to define sleep-wake variables. Sleep onset time was defined as the first minute of ≥ 3 consecutive minutes of scored sleep within the scoring interval, and sleep end time as the last minute of ≥ 5 consecutive minutes of scored sleep immediately prior to the end of the scoring interval.²⁰

Table 2 gives an overview of the definition of sleep-wake variables assessed by diary and actigraphy.

Table 2.

Definitions of sleep-wake variables assessed by diary and actigraphy.

Circadian Rest-Activity Cycle Measures

Nonparametric circadian rhythm analysis (NPCRA) was carried out for 7 consecutive days with complete and edited actigraphy data.²¹ This analysis yields circadian characteristics of the rest-activity cycle, such as inter-daily stability (IS), intra-daily variability (IV), and relative amplitude (RA). The IS index quantifies the stability of the rest-activity cycle between days; higher values indicate a more stable rhythm. The IV index gives an indication of the fragmentation of the rhythm, that is, the frequency and extent of transitions between rest and activity; higher values indicate a more fragmented rhythm. The RA is calculated from the ratio of the most active 10-h period to the least active 5-h period across the averaged 24-h profile. Higher values indicate a stronger rhythm with a better consolidation of daytime activity and nighttime sleep.

Sleep Problems

Behavioral sleep problems were assessed by parent reports, using the Infant Sleep Questionnaire (ISQ).²² The ISQ asks about the frequency and/or duration of settling and night waking problems as well as co-sleeping in the last months. A severe settling problem was identified when issues were present ≥ 4 nights per week and when settling took > 20 min. The settling problem was classed as moderate when issues were present ≥ 4 nights per week but settling took < 20 min, or < 4 nights per week but with settling taking > 20 min. Settling problems for < 4 nights per week and taking < 20 min were defined as no/mild problems. A severe nocturnal awakening problem was identified when issues were present ≥ 4 nights per week, waking ≥ 3 times per night, and waking for > 10 min during the night; a night waking problem was classed as moderate when, the issues were present for ≥ 4 nights per week, waking only once or twice per night and taking ≥ 10 min to go back to sleep, or for < 4 nights per week, waking ≥ 3 times per night but taking < 10 min to go back to sleep. Waking up once or twice per night and taking < 10 min to go back to sleep each time were defined as no/mild problems.

Sociodemographic Information

Children's birth order and Swiss nationality (yes/no) and parents' sociodemographic information (e.g., occupation) were assessed using a short parent-reported questionnaire. Socioeconomic status (SES) was estimated by means of a 6-point scale of both current paternal occupation and maternal education. The scores ranged from 2 (lowest SES) to 12 (highest SES). Three social classes were defined: SES scores 2 to 5, lower class; SES scores 6 to 9, middle class; SES scores 10 to 12, upper class. This measure has been used in previous studies and has been shown to be a reliable and valid indicator of SES in our community.²³

Behavior Problems

The Child Behavior Checklist (CBCL) is a widely used, well-validated, standardized measure assessing parental reports of a child's behavior problems.^24,25 In this study, the authorized German version of the CBCL/1.5-5 was used for children with age ≥ 18 months.²⁶ The instrument yields scores for 2 broadband scales (internalizing and externalizing behavior problems) and an overall total behavior problem score. Higher scores indicate more problems. Reference T- scored values are provided by a community sample of 700 healthy U.S. children aged 18–71 months.²⁵ To date, no Swiss/German norms are available. Consequently, the U.S. sample was used as an external control. In the current study, the internal consistency of the internalizing (α = 0.84/0.67), externalizing (α = 0.94/0.90), and the total behavior problems (α = 0.92/0.86) was acceptable and excellent for before and after intervention, respectively.

Intervention

In 2–3 outpatient consultations, parents were guided through 3 steps: (1) introducing a regular sleep-wake rhythm; (2) adjusting the child's bedtimes to individual sleep need; and (3) assisting the child to fall asleep on his/her own. At each consultation, the third step of the 3-step concept was only addressed after achieving the first and second steps. The intervention was delivered on an individual rather than group basis, thereby allowing for more flexibility in addressing personal concerns and providing individualized sleep-related information. An overview of the sleep-related information provided is listed in Table 1.

The first protocol step focused on providing information about the circadian process of sleep regulation, identifying sleep-phase preferences within a 24-h period (e.g., times when the child can easily fall asleep or easily wakes up on his/her own), and introducing a regular sleep-wake rhythm as an important zeitgeber of the circadian process. Learning more about these issues was intended to enhance parents' understanding of child's preferred sleep times. The second protocol step focused on giving information on the homeostatic process of sleep regulation, the calculation of a child's sleep need over 24 h based upon diary reports, the evaluation of the child's sleep need as a percentile relative to current reference curves,²⁷ and the discussion of realistic estimates for time in bed for both nighttime and daytime sleep. Learning more about these issues was intended to provide parents with a better appreciation of appropriate times in bed for their children and possible adjustment of bedtimes to actual sleep need. It was also intended to help parents acquire a feeling of control over their situation and thus reduce their own feelings of stress and anxiety. The third protocol step focused upon establishing a relaxing and calming bedtime routine that would enable the child to fall asleep on his/her own, also return to sleep after nocturnal awakenings. Parents were asked to describe a normal bedtime routine, focusing on their own behavior around the time of sleep initiation (e.g., rocking). If the child needed a parent's physical closeness to help him/her to fall asleep, parents were asked to develop other strategies that were then gradually introduced. Thus, it was intended that parents would gain a better understanding of how their behavior at bedtime relates to their child's requirements after nocturnal awakenings. It was also intended to enhance their understanding of how they could train their child to self-soothe after nocturnal awakenings.

Statistical Analyses

Data were analyzed using the SPSS statistical package, release 20.0 for Windows (SPSS Inc., Chicago, IL, USA). All statistical tests were 2-sided with a predefined significance level of p < 0.05. Descriptive results for nominal variables are presented as number of cases and percentages. Mean, standard deviation (SD), and range are given for continuous variables. For all sleep-wake variables, daily measurements were averaged over 7 to 10 successive days for each child. These individual “weekly” means were the units of analysis. Coefficient of variation (CV) was calculated for selected sleep-wake variables by the ratio of the SD to the mean, expressed as percentage. Chi-square tests and Mann-Whitney U-tests were used, where appropriate, to compare sociodemographic (e.g., gender, age, nationality) data between eligible and non-eligible participants and non-participants, and between included and excluded children from subsample analyses. Neither the sleep-wake variables nor the Child Behavior Checklist (CBCL) scales deviated significantly from a normal distribution according to the Kolmogorov-Smirnov test; thus we used parametric Student t-tests for paired samples for testing equality of means between T0 and T1. Comparison of CBCL T-scores with published reference data was determined using one-sample t-tests. Linear multiple regression analysis was used to predict diary-reported nocturnal wake duration. For all comparisons, effect sizes were computed by Cohen's d (0.20, small effect; 0.50, medium effect; > 0.80, large effect).²⁸

RESULTS

Sample Characteristics

Descriptive information about the study sample is presented in Table 3. Most children were first-born and belonged to middle or upper SES class. Sixty-two children (75.6%) had Swiss nationality. Of the total of 82 children, 51 (62.2%) children had a moderate or severe settling problem and 76 (92.7%) a moderate or severe night waking problem. The majority of parents and children were co-sleeping several times per week, and 15 of the 82 children (18%) were occasionally breastfed. All families received 2 outpatient consultations, with the exception of 7 families (9%) who also had a third consultation.

Table 3.

Demographic and diagnostic characteristics (n = 82).

A multiple regression analysis predicting the nocturnal wake duration at T0 indicated that the age of the child (B = −10.52, SE = 4.76, β = −0.28, p = 0.03) and whether the child was breastfed or not (B = 22.24, SE = 9.88, β = 0.26, p = 0.028) significantly predicted the child's nocturnal wake duration. Thus, younger children and children who were occasionally breast-fed had longer nocturnal wake duration. No effects were found for child's gender, the number of consultations, frequent co-sleeping (0–2 nights per week vs. 3–7 nights per week), or SES.

Changes in Sleep-Wake Variables

Means and SD of the sleep-wake variables assessed by diary and actigraphy at T0 and T1 and statistics for mean comparisons between T0 and T1 are provided in Table 4. Changes from T0 to T1 were verified with comparable effect sizes by actigraphy for all diary-reported sleep-wake variables. Therefore, only diary-reported results are described here. A significant change from T0 to T1 with a moderate effect size was found for children's sleep duration (nighttime, p ≤ 0.001; daytime, p ≤ 0.001) and for 24-h sleep duration (p ≤ 0.001). At T1, children had shorter sleep duration than at T0. A significant change with a moderate effect size from T0 to T1 was also observed for children's nocturnal wake duration (p ≤ 0.001). At T1, children had shorter nocturnal awakenings than at T0. These effects remain significant after Bonferroni correction. Furthermore, significant changes from T0 to T1 with small effect sizes were found for children's longest continuous nocturnal sleep duration (p = 0.019) and sleep onset time in the evening (p ≤ 0.001). At T1, children had a longer longest continuous nocturnal sleep duration and a later sleep onset time. Moreover, the variability of sleep onset time in the evening (p = 0.064) and sleep end time in the morning (p = 0.042) decreased with a moderate effect size from T0 to T1.

Table 4.

Sleep-wake variables measured by diary and actigraphy of sleep-disordered children at baseline (T0), and follow-up (T1).

Changes in Circadian Rest-Activity Cycle Measures

From T0 to T1, significant changes with a moderate effect size were found for 2 circadian rest-activity cycle measures. At T1, the rest-activity cycle was more stable between the days than at T0 (IS index at T0: mean = 0.70, SD = 0.093; IS index at T1: mean = 0.73, SD = 0.081; p = 0.013, d = 0.41) and daytime activity and nighttime sleep were both more consolidated (RA index at T0: mean = 0.89, SD = 0.05; RA index at T1: mean = 0.90, SD = 0.05; p = 0.015, d = 0.23). No significant change was found for the IV index (p = 0.46, d = 0.07).

Changes in Behavior Problems

Means and SD of the 2 broadband scales and the overall total behavior problem scale at T0 and T1 and statistics for mean comparisons between T0 and T1 are provided in Table 5. At T0, sleep-disordered children had more behavior problems than healthy controls (internalizing: p = 0.040, d = 0.42; externalizing: p = 0.14, d = 0.33; total problems: p = 0.003, d = 0.66). A significant change from T0 to T1 was found for both the internalizing (p = 0.031) and the total behavior problem scales (p = 0.018).

Table 5.

Behavior problems (CBCL T-scores) of sleep-disordered children with age ≥ 18 months at baseline (T0) and follow-up (T1).

Associations between Changes in Sleep-Wake Variables and Behavior Problems

We also hypothesized that the reduction in 24-h sleep duration from T0 to T1 would be associated with changes in behavior problems. For diary-reported 24-h sleep duration, although not reaching significance, a larger decrease of the 24-h sleep duration was associated with a greater reduction in behavior problems from T0 to T1 (internalizing: r = −0.37, p = 0.11; total problems: r = −0.34, p = 0.15). For actigraphically measured 24-h sleep duration, associations reached significance with respect to internalizing behavior problems (r = −0.63, p = 0.02) and marginal significance with respect to total problems (r = −0.53, p = 0.06).

DISCUSSION

This study describes the Zurich 3-step concept for the management of behavioral sleep problems in children and evaluates changes from before to after the intervention using subjective and objective measures. In line with our expectations, children's nocturnal wake duration significantly decreased and the duration of children's longest continuous nocturnal sleep increased during the course of the intervention. These effects may result from a decrease in sleep duration which led to more consolidated sleep (findings not presented). Theoretical considerations and the results of previous studies^12,29 lead to the expectation that sleep pressure increases with increasing wakefulness; this is then reflected in an increase in sleep depth with less frequent awakenings during subsequent sleep (sleep consolidation). Indeed, our results indicated a significant reduction for nighttime (about 14 min) and daytime sleep duration (about 10 min) from T0 to T1. For nighttime, the reduced sleep duration results from a delay of sleep onset time in the evening while keeping sleep end time in the morning constant. In other words, parents have adjusted the child's evening bedtimes to his/her sleep need. Because behavior problems also improved from before to after intervention, there is no reason to believe that this approach leads to any detrimental effect on children's behavior.

Significant changes from T0 to T1 were also found for children's sleep-wake rhythm. First, we examined the variability of sleep onset time in the evening, sleep end time in the morning and first daytime sleep onset time, measured by the coefficient of variation. Results indicated a reduction with a moderate effect size from T0 to T1 for the variability of sleep onset times in the evening and sleep end times in the morning. Second, we examined changes in circadian characteristics of the rest-activity cycles measured by actigraphy. In a free-running condition, an individual's circadian clock has a rhythm with a period slightly shorter or longer than 24 h and a phase relative to the time of day.³⁰ Thus, the internal circadian clock must be synchronized with night and day by external timing cues (e.g., by regular bedtimes). Our results indicated that children had a more stable (higher IS index) and a stronger rhythm (higher RA) at T1 than at T0. Thus, the coupling of the internal circadian rhythm to stable environmental zeitgebers was better, and children were more active during daytime and more inactive during nighttime. These findings are based on the supposition that parents had truly introduced a more regular sleep-wake rhythm and not only reported more regular sleep times in the diary.

Our hypothesis that sleep-disordered children older than 18 months of age would have more behavior problems than healthy controls was also supported. The literature provides ample evidence that sleep problems in children and adolescents are linked to a host of emotional and behavioral difficulties.³¹ However, these problems are frequently bidirectional and complex, and thus causal inferences cannot be made. Our data show that the Zurich 3-step intervention concept may also positively affect children's behavior problems. Our results regarding associations between changes in sleep-wake variables and behavior problems indicate that a larger decrease of 24-h sleep duration from T0 to T1 was associated with a greater reduction in behavior problems.

The strength of the current study is the well-structured Zu-rich 3-step intervention concept. This is held in an individual format which allows personal concerns to be addressed and individualized sleep-related information to be provided. Each intervention was conducted by the same clinically experienced pediatrician (P.H.). The study occurred in appropriate statistical conditions, with no sociodemographic differences between study participants and non-participants nor between included and excluded children from subsample analyses and the use of actigraphy to confirm diary-reported sleep measures. Nevertheless, certain limitations have to be considered when interpreting the findings of our study. First, at study initiation, our 3-step intervention protocol was already well established in the clinical practice at our outpatient clinic, and we were not able to set up a randomized/non-randomized controlled trial for ethical reasons. We had to treat urgent cases, and no waiting list control group was possible. Moreover, introducing an additional and different intervention would not have been without substantial biases in clinical practice. Thus, we cannot rule out the possibility that improvements from T0 to T1 resulted from a time or age effect or another effect of the intervention (e.g., listening to parents' account of their struggles). However, we believe that the rather short intervals between the sleep assessment periods may have reduced the probability of a significant time or age effect. Moreover, the direction of the effects gives reason to believe that an intervention effect occurred. Thus, although our study may not be rated on a high level of evidence (I or II) as suggested by the AASM Classification of Evidence, we still reach a recommendation grade C, level IV (see Table 3 in Mindell,¹⁴ where 13 studies with a similar design were included). Second, our response rate was about 60% at T0 and 52% at T1. This rather low participation rate may have been caused by our study design. It was our intention to conduct out the study strictly within the limits of the clinical practice. Parents were recruited on the phone immediately after registration at our Center and had to complete a rather time-consuming 24-h sleep diary during a period when their psychological strain might have been highest. Thus, it is possible that parents did not want to have any additional burden, which reduced the participation rate. Third, the appropriateness of using U.S. norms for the CBCL/1,5-5 can be questioned. Slight differences in the age ranges of the normative samples and cultural differences may compromise comparability. However, cross-cultural bias seems unlikely, given previous findings supporting the use of U.S. norms for the CBCL in German samples.³²

Our findings suggest the Zurich 3-step intervention concept can effectively reduce behavioral sleep problems of children, firstly by focussing on basic concepts of the two-process model of sleep regulation (introducing a regular rhythm and adjusting bedtime to sleep need), and secondly by using behavioral strategies to handle elements of maladaptive learned sleep behavior. Thus, the Zurich 3-step intervention concept may be used in clinical practice with sleep-disordered children.

DISCLOSURE STATEMENT

This study was supported by a research grant from the Anna Muller Grocholski Foundation (Reachout II Project) and the Clinical Research Priority Program (CRPP) “Sleep and Health” of the University of Zurich. The authors have indicated no financial conflicts of interest.

ABBREVIATIONS

AASM

American Academy of Sleep Medicine

h

hours

ISQ

Infant Sleep Questionnaire

IS

interdaily stability

IV

intradaily variability

min

minutes

NPCRA

nonparametric circadian rhythm analysis

RA

relative amplitude

SES

socioeconomic status