Multidimensional scaling of pediatric sleep breathing problems and bio-behavioral correlates

Complex relationships exist between pediatric sleep disorders and daytime behavior. Using a multidimensional scaling model, we investigated these relationships in 126 children with sleep breathing disorders (SBD).

Validated questionnaires on nighttime behavior, daytime behavior, and respiratory health were administered to a large number of school children in Belgium. Children who met the criterion of having at least one sleep-related breathing problem (three or more times per week during the past six months) were selected for further analyses. A total of 26 indicators were defined and modeled, including sleep problems, sleep efficiency, sleep environment, sleep enuresis, internalised and externalised behavioral problems, respiratory health of the child and relatives, smoking exposure, and caffeine consumption.

From 3,045 questionnaire responses 4.1% of the children were reported to have a SBD symptom. SBD children differed on sleep and health domains from non-SBD children. Furthermore, through scaling of the (dis)similarities among the 26 indicators the SBD child was able to be modeled. By way of an internal analysis of the data-matrix the following indicators were eliminated: sleep correlates, health of the family, and behavior rated by teachers, followed by caffeine intake, drugs, and behavior rated by the parents. This revealed a two-dimensional model, consisting of primary SBD and secondary SBD.

Children with SBD differ on many domains from children without such disorders and an underlying two-fold SBD concept was found. Firstly, the SBD-indicator positioned in between investigated correlates with disorders of initiating and maintaining sleep and sleep hyperhydrosis on one hand and with respiratory-related illnesses on the other; this was labeled primary SBD. Secondly, the SDB-indicator not closely associated with any of the investigated correlates can be interpreted as secondary SBD.