Abstract Introduction Effects of child temperament on biopsychosocial outcomes, including sleep, are known to differ based on parent-child relationship quality (PCRQ). However, previous studies have primarily focused solely on early childhood, relied on subjective reports of PCRQ and sleep, and have neglected to examine variability in sleep, which is an important dimension of sleep health. We examined whether there were interactions between temperament and objectively measured PCRQ during early childhood in association with actigraphy measured sleep (average and variability) in middle childhood. Methods Participants included 90 preschoolers (mean age=4.24 years, SD=0.81) who were 47% female; 37% White, 17% Black, 47% other racial groups; and 46% Hispanic. Thirty-four percent experienced moderate to severe maltreatment. PCRQ was coded from a video-recorded free-play task between each child and their primary caregiver. Child temperament dimensions including Activity Level, Anger, Social Fear, Inhibitory Control, and Appropriate Attentional Allocation were measured with the parent-report Toddler Behavior Assessment Questionnaire. Child Mean and SDs of total sleep time (TST) and sleep efficiency (SE) were generated from actigraphy data collected for two weeks during a follow-up study when participants were ages 8-11. Results There were no direct effects of any of the temperament variables or PCRQ on sleep. In models adjusted for demographic characteristics and number of adversities experienced, greater attentional allocation was associated with greater average TST only among those children high in PCRQ (estimate = 20.59, 95% CI = 5.19, 35.99). For every one unit increase in attentional allocation score (possible range 1:7), there is a more than 20 min increase in mean TST. In contrast, greater activity levels were associated with better average sleep efficiency, only among those low in PCRQ (estimate = 2.35, 95% CI = 0.46, 4.24). For each unit increase in activity level score, there was a 2% increase in sleep efficiency. Conclusion Findings extend cross-sectional research in early childhood and provide support for models showing that children have differing susceptibility to the effects of temperament on sleep, depending on PCRQ. Support (if any) NIMH R01 MH083704, NICHD R01 HD086487, NICHD, R01 HD095837, SRCD Small Grant for Early Career Scholars, T32HD101392, NIGMS P20 GM139767.
Read full abstract