Risk factors
Univariable risk factors for patient- and parent reported sleep outcomes are presented in Supplementary Table 1. Higher sleep disturbances were reported by children more shortly after diagnosis (B=-.12, 95%CI -.20;-.04, P <.01), and by those with cerebral or posterior fossa tumors, compared to children with supratentorial midline tumors (B=-4.56, 95%CI 10.01;-.88, P= .10). Parents reported higher child sleep disturbances more shortly after diagnosis (B=-.10, 95%CI -.18;-.01, P= .03) and before radiotherapy (B=-6.41, 95%CI -12.46;-.35, P= .04). However, in the multivariable models (Table 4) shorter time after diagnosis (B=-.11, 95%CI -.19;-.03,P =<.01) remained the only independent significant determinant for self-reported sleep disturbance. For sleep related impairment, no significant risk factors were identified on both the self- and proxy-reports.
For actigraphic outcomes, univariable analyses (Supplementary Table 2) showed females had higher sleep efficiency (B=3.60, 95%CI -.60;-7.78,P =.09). The history of an obstructive hydrocephalus was associated with shorter sleep onset latency (B=-8.92, 95%CI -16.83;-1.01, P= .03) and longer total sleeping time (B=32.86, 95CI 1.10;64.62, P =.04). More nighttime awakenings were related to neurosurgery (B=5.37, 95% CI -.54;11.29, P= .07) and radiotherapy (proton or photon, B=4.27, 95%CI -.74;9.28, P= .09). Younger children had higher sleep onset latency (B=-1.96, 95%CI -3.36;-.56, P <.01), more total sleep time (B=-7.35, 95%CI -12.98;-1.73, P= .01) and spent more time in bed (B=-10.59, 95%CI 15.33;-5.85, P <.01). Multivariable analyses (Table 5) showed the history of an obstructive hydrocephalus independently predicted longer sleeping times (B=41.04, 95%CI 11.41;70.68, P< .01). Finally, younger age remained associated with longer sleep onset latency (B=-1.73, 95%CI -3.12;-.35,P= .02) and more total sleeping time (B=-8.63, 95%CI -14.00;-3.26, P <.01).
Body mass index, parental education level, start of chemotherapy, hormone deficiency and epilepsy were not significantly associated with any of the sleep outcomes.