Background: Smith-Magenis syndrome (SMS) is a rare neurodevelopmental disorder characterized by severe sleep disturbances, circadian rhythm dysregulation, and prominent behavioral difficulties. Objective: To examine associations between subjective and objective sleep and circadian characteristics and behavioral difficulties in children with SMS in real life. Methods: Children aged 5-13 years with genetically confirmed SMS were included in a prospective single-center cohort study. Sleep and circadian rhythms were assessed using parent-reported questionnaires (Sleep Disturbance Scale for Children (SDSC), Horne and Östberg Morningness-Eveningness Questionnaire (H&O)) and 15-day home-based wrist actigraphy. Behavioral difficulties were evaluated using the Aberrant Behavior Checklist-Community (ABC-C). Associations between subjective and objective sleep and circadian measures with behavioral difficulties were studied using Spearman correlations for each ABC-C subscale. Results: Among the 20 children included (55% female, mean age 10 ± 2.5 years), 75% were carriers of a 17p11.2 microdeletion and 25% had an RAI1 mutation. Overall, 95% of children were treated with melatonin, 55% with psychostimulants and 20% with beta-blockers. According to the SDSC results, pathological sleep disturbances were present in 14/20 (70%) children and concerned mainly complaints of insomnia (10/20, 50%) and excessive daytime sleepiness (8/20, 40%). The H&O chronotypes confirmed the morning type in 70% of patients. Sleep and circadian measures obtained by actigraphy showed reduced total sleep time in 71% of the children, reduced sleep efficiency in 88%, long wake after sleep onset (WASO > 60 min) in 59%, as well as an early L5 onset (5 h period with the least movements < 23:38) and M10 onset (10 h period with the highest activity levels < 09:12) in 92% and 62% of children, respectively. The behavioral domains of the ABC-C with the highest scores were hyperactivity (mean 52.1, range 2.1-97.9) and irritability, agitation, and crying (mean 46.9, range 17.8-82.2). In the analysis of the relationship between behavior and sleep, a positive correlation was found between sleep disorders and stereotyped behaviors (r = 0.472, p = 0.048), insomnia and irritability, social withdrawal and stereotyped behaviors (r = 0.472, p = 0.048; r = 0.701, p = 0.001; and r = 0.648, p = 0.004, respectively), and non-restorative sleep and inappropriate speech (r = 0.476, p = 0.046). No significant association was found between objective sleep measures and behaviors. While subjective assessments suggested that a stronger morning chronotype was associated with lower irritability (r = -0.565, p = 0.015), actigraphy-derived circadian rhythm analyses revealed that an earlier L5 onset was associated with greater social withdrawal and hyperactivity (r = -0.556, p = 0.048 and r = -0.560, p = 0.049, respectively). Conclusions: The present cohort study of children with SMS studied in real-life conditions, using both objective and subjective measures, shows that patients continued to experience sleep and behavioral disturbances despite treatment. Exploratory analyses identified distinct associations between subjective sleep disturbances, circadian rhythm characteristics, and specific behavioral domains. While subjective sleep disturbances were associated with irritability, stereotypic behavior, and inappropriate speech, circadian rhythm parameters are specifically associated with hyperactivity. Both were also associated with social withdrawal. As these findings arise from an exploratory observational secondary analysis, they should be considered hypothesis-generating and require confirmation in larger prospective studies. If confirmed, they may help identify sleep and circadian rhythm characteristics as potential targets for future interventions in children with SMS.
