Primary Objectives:
To investigate the prevalence and nature of sleep problems in 4- to 10-year-old children in both out-of-home care (long-term residential and foster care) and regular family settings.
Research questions:
- What is the prevalence and nature of sleep problems in Flemish children from the general population?
- What is the prevalence and nature of sleep problems in Flemish children placed in outof-home care?
Hypotheses:
- Based on findings from the original American Children Sleep Habits Questionnaire (CSHQ) validation study , and in the absence of Dutch percentile norms, the investigators hypothesize that approximately 23% of children in the general population will obtain a total CSHQ score of 41 or higher.
- The prevalence of sleep problems in Flemish children placed in out-of-home care is higher than the prevalence of sleep problems in the general population.
To investigate the differences in the prevalence and nature of sleep between groups of children with different life experiences and upbringing contexts.
Research question: Are there differences in the prevalence and nature of sleep problems between Flemish foster children, children placed in long-term residential care and children from the general population?
Hypotheses:
- Children placed in out-of-home care experience significantly more sleep-related problems compared to children from the general population.
- Children placed in residential out-of-home care experience significantly more sleep-related problems compared to children in foster care.
To evaluate the effectiveness of a brief trauma-based sleep intervention and thereby contribute to evidence-based practices that promote healthy sleep and psychological well-being in high-risk populations.
Research question: What is the effect of a brief trauma-sensitive sleep intervention in foster care? Hypothesis: Foster children who are randomized into the intervention group will experience a significant improvement in sleep health (e.g., increased total nighttime sleep and fewer sleep problems) based on caregiver reports, in comparison to foster children in the control group at both post-intervention and follow-up.
Secondary Objectives:
To identify associated individual and contextual factors, including factors associated with the placement such as placement duration, age at placement, contact with biological parents etc. linked to sleep in both children placed in out-of-home care and children from the general population.
Research questions:
- What characteristics are sleep problems of Flemish children in the general population associated with?
- What characteristics are sleep problems of children placed in out-of-home care associated with?
Hypotheses:
- Sleep-related problems in 4- to 10-year-old children (across all subgroups) are positively associated with adverse childhood experiences.
- Sleep-related problems in 4- to 10-year-old children (across all subgroups) are positively associated with behavioral problems.
- Sleep-related problems in out-of-home placed children are associated with placement characteristics including factors associated with the placement such as placement duration, age at placement, contact with biological parents, etc.
- Sleep-related problems in 4- to 10-year-old (foster) children are positively associated with levels of (foster) parenting stress.
To investigate the differences in associated factors of sleep between groups of children with different life experiences and upbringing contexts.
Research question: Are there differences in the associations between sleep and behavior, adverse childhood experiences and parenting stress in children placed in out-of-home care compared to children from the general population? Hypothesis: There are differences in the link between sleep and associated factors in children placed in out-of-home care compared to children from the general population.
To evaluate outcomes of a brief trauma-based sleep-intervention. Research question: What are outcomes of a trauma-focused sleep-intervention in foster children and their families? Hypothesis: A brief trauma-focused sleep intervention will result in a significant reduction of foster child behavior problems and parenting stress in the intervention group compared to the control group both at post-intervention and follow-up.
Data Collection:
The surveys of component 1 and 2 will be developed in Qualtrics and distributed to respondents via a web link and/or QR code. Data collection will be conducted anonymously. Participants may withdraw from the study at any time without consequences and without being required to provide a reason.
An a priori power analysis was conducted for all studies using G*Power 3.1 to determine the required minimum sample size. Based on previous literature, assuming a medium effect size, a target sample size of approximately 180 to 200 participants was set for each cross-sectional study (540 - 600 participants in total).
For the RCT study, assuming a medium intervention effect (Hedges' g = 0.50), α = .05, power = .80,three assessment points, and a within-subject correlation of .50, a minimum sample of approximately 120 families is required. To account for potential attrition, the investigators aim to recruit 130 families (65 per group).
Statistical Considerations and Data Analysis Cross-sectional survey studies Data will be analyzed using IBM SPSS Statistics (version 31). Prior to conducting the analyses, the dataset will be screened for missing values and outliers.Cases with substantial missing data or extreme outliers will be handled appropriately and, if necessary, excluded from further analyses. Assumptions for parametric tests (normality, homogeneity of variance, linearity) will be evaluated.
The primary objective of the cross-sectional studies is to investigate the prevalence and nature of sleep problems in Flemish children aged 4-10 years from the general population and children placed in out-of-home care.
Descriptive statistics will be used to characterize the study sample and the distribution of the study variables. Sleep problems will primarily be operationalized as the total score on the Children's Sleep Habits Questionnaire (CSHQ), with higher scores indicating more sleep-related problems. The prevalence of clinically relevant sleep problems will additionally be estimated using the recommended cut-off score of 41 on the total CSHQ score. Prevalence estimates will be reported with 95% confidence intervals. The nature of sleep problems will be described using the total CSHQ score and the scores on the CSHQ subscales, allowing for the identification of the most prevalent types of sleep difficulties.
The secondary research questions aim to identify individual, family-related, and placement-related characteristics associated with sleep problems in children from the general population and children placed in out-of-home care. Univariable analyses will initially be conducted to describe the associations between individual predictors and sleep problems. Depending on the measurement level and distribution of the variables, Pearson or Spearman correlations, independent-samples t-tests, or analyses of variance (ANOVAs) will be used.
To examine the independent associations between the hypothesized individual, family-related, and placement-related factors and sleep problems, multivariable regression analyses will subsequently be conducted. The total CSHQ score will be used as the dependent variable in linear regression models.
Candidate variables for the multivariable models will be selected a priori based on theoretical considerations, previous empirical research, and the study hypotheses, rather than solely based on statistical significance in the univariable analyses. These prespecified candidate variables will be included in the initial multivariable models. Where appropriate, backward elimination will subsequently be used to obtain a more parsimonious final model. Prespecified variables considered theoretically important or relevant potential confounders will be retained irrespective of their statistical significance.
Separate multivariable models will be considered for children from the general population and children placed in out-of-home care, given that some predictors are specific to the out-of-home care context. For children placed in out-of-home care, placement-related characteristics such as placement duration, age at placement, and contact with biological parents will additionally be examined as potential explanatory variables. Across the groups, adverse childhood experiences, behavioral problems, and (foster) parenting stress will be examined as hypothesized correlates of sleep problems.
Regression coefficients, 95% confidence intervals, and p-values will be reported. Model assumptions will be assessed before interpreting the regression results, including linearity, homoscedasticity, normality of residuals, and multicollinearity.
Comparative study Descriptive statistics will be computed to compare the different samples. ANCOVA analyses will be conducted to investigate group differences in sleep and associated factors between different respondent groups.
RCT study Baseline characteristics of participants will be summarized descriptively for each study group.
Randomization To allocate foster parents to either the control or the intervention group the investigators will use minimization, controlling for important variables such as age, gender, sleep, ACEs, kinship/non-kinship and duration of placement.
Intervention effects A repeated-measures mixed-effects model (Linear Mixed Model) will be used to examine the effects of the intervention on sleep problems, behavioral problems, and parenting stress over time. Time, condition, and the time × condition interaction will be included as fixed effects. Participants will be included as a random intercept only, to account forwithin-subject dependency across repeated measurements. Significant time × condition interactions will be followed up using simple effects analyses and pairwise comparisons.
Drop-outs Between-group ANOVAs will be conducted to compare participants who completed the study with those who did not (retained vs. non-retained participants).
Analyses will be performed according to the intention-to-treat (ITT) principle.
Missing data The linear mixed-effects model will be used as the primary approach for handling missing outcome data. This approach allows participants with incomplete repeated measurements to contribute to the analyses using the data they provided, under the assumption that data are missing at random (MAR).
The extent and patterns of missing data will be examined. Where appropriate, multiple imputation (MI) will be considered as a sensitivity analysis to assess whether the findings are robust to alternative approaches for handling missing data.
Further design considerations The design and implementation of the RCT intervention study and comparative study will be further informed by the findings of the preceding studies.
All respondents of components 1 and 2 will remain anonymous while completing the questionnaire. The collected data will be used solely for research purposes and will be treated confidentially. The informed consent form states that participants may withdraw from the study at any time without consequences and without needing to provide a reason. All data will be securely stored on a VUB Pixiu platform that is accessible only to the researchers and will be retained for a maximum period of 10 years.
Given the inclusion of potentially sensitive questions (e.g., adverse childhood experiences), contact information for professional support services will be provided at the end of the questionnaires.
