Database search and data extraction

Figure 1 presents the data extraction records for each stage of the search. Of the initial 781 papers identified after exclusion of duplicates, 26 met the criteria for inclusion. Most papers reported on a single outcome, but two reported on more than one outcome. An overview of the key details of the 26 papers arranged by outcome measure is presented in table 1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Systematic review search flow diagram based on Preferred Reporting Items for Systematic Reviews and Meta-Analyses.9

Designs and methodologies

There were no randomised control trials within the corpus of studies. There was a heterogeneous group of observational studies that included experimental (n=5), quasi-experimental (n=6), longitudinal (n=6) and cross-sectional designs (n=9). The description of each study and their findings are summarised in table 1 and summary of quality and GRADE score in table 2. It should be noted that the sample sizes are generally small and therefore present challenges for generalisability. In addition, the studies using correlational designs had varying levels of statistical control. Where such factors present a major concern for interpretation this is noted in table 2.

Outcomes

The consistency of benefit or cost to the child was examined with reference to study quality. Three studies examining the association of napping with changes in salivary cortisol were not included in this process because the meaning of the outcomes could not be interpreted in terms of benefit or cost to the child. A summary of the direction of findings for all other outcomes is presented in figure 2.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Number of articles reporting health and development benefits or costs of napping.

The most consistent finding was an association between napping and night sleep. Table 3 provides a summary of the magnitude of findings and documents a moderate association of napping with night sleep patterns including later sleep onset,15 ,18 ,19 reduced duration5 ,16 ,17 ,20 and quality indices.16 ,19 ,20 Three studies examine differences in total 24 h sleep duration and report no difference between those who do and do not nap, indicating that daytime napping changes the distribution, rather than the duration, of sleep in a 24 h period.5 ,16 ,20 However, Cairns and Harsh22 in a study of transition to school report reduction in total sleep duration which they attribute to loss of naps. Notably, eight of the studies of nighttime sleep were of children aged >2 years, a point at which there is a significant decline in habitual napping.6 Three papers provide evidence of direction of association. Komada et al18 presents evidence of the direction of the relationship through a detailed study of nine consecutive 24 h periods in a sample of 967 children aged 0–5 years. The study reports that there were no significant differences in night sleep associated with duration of napping in children <2 years of age. Children >2 years of age, however, decreased sleep duration and later bedtime occurred on nights after napping. Fukuda and Sakashita15 report that among children attending Kindergarten, where napping is optional, the onset of night sleep was later on evenings after a nap compared with onset on days in which the same children had not napped. Napping was not found to be a response to shorter sleep duration on the night preceding a nap. Iwata et al39 present data on children who are non-habitual nappers attending nursery during weekdays, where naptime is compulsory and compares their weekday and weekend sleep. During weekends, most children did not nap and had earlier sleep onset and shorter sleep latency. A subgroup of children, however, had late unplanned naps that increased the onset and decreased the quality of their nighttime sleep. A likely explanation for these findings is that napping acts to dissipate homeostatic sleep pressure,10 with this mechanism evidence by increased sleep latency and later sleep onset. Together the evidence suggests that, beyond the age of 2 years napping can influence night sleep patterning.15 ,18 ,39 The meaning of these affects, however, remains uncertain.

Behavioural outcomes were examined by five studies. Berger et al24 report an experimental study of sleep restriction among children, aged 30–36 months, trained to a nap schedule. Under these conditions a reduction in positive facial emotional responses and increase in negative facial emotional responses to visual stimuli and problem solving were found. Three studies are non-experimental. Spruyt et al,23 in a detailed 12-month longitudinal study, report that children with easier temperament napped more readily, but at 12 months, decreased daytime sleep was associated with better emotional regulation as measured by the Bayley Scales of Infant Development. Two studies report association with internalising behavioural problems using the parent-form of the Child Behaviour Check List.17 ,25 Yokomaku et al17 found longer nap duration among children aged 4–6 years was associated with greater anxiety/depression, withdrawal and thought problems, whereas Hall et al25 found that among children aged 1–3 years difficulty settling for naps at home was associated with anxiety, depression, withdrawal and internalising behaviour. In addition, Hall et al25 obtained teacher report using the Preschool Adjustment Questionnaire and Preschool Behaviour Questionnaire. There were inconsistent outcomes; higher daily frequency of napping was associated with less prosocial behaviour and reduced ability to deal with challenges, whereas difficulty settling for naps was associated with poorer behaviour and poorer adjustment at preschool. The difference between the experimental study and non-experimental studies are most likely explained by two processes. First, although the experimental study focuses on disruption of habitual napping,24 the non-experimental studies focus on normally occurring nap behaviours.17 ,23 ,25 Second, two of the non-experimental studies were of children who attended preschool settings.17 ,25 The preschool setting may be a factor explaining the association of problems settling for naps and behavioural difficulties because these contexts often mandate naptime and may not be environmentally conducive to napping.40 Both the stage in developmental sleep transition and the environmental context of napping are significant factors in the interpretation of results.

Studies of cognitive outcomes also present a mixed picture. Three studies report that language learning is improved after a nap,9 ,26 ,27 whereas another reports that being awake, rather than napping, improves language generalisation.30 The age of children and the task used to assess learning may be implicated in these findings but there is currently insufficient evidence to speculate on cause of this inconsistency. An important factor in interpretation of effect, however, is the habitual napping status of the children. Kurdziel et al9 report a positive effect of napping on learning and memory consolidation in preschool aged children, but with benefit only for habitual nappers. Two further studies suggest that sleep consolidation (ratio of night-to-day sleep) is a key marker of cognitive maturity.20 ,28 Dionne et al,28 using a longitudinal twin design, mapped the association of sleep consolidation from 6 to 30 months on language development at 60 months. They report that later sleep consolidation is a risk factor for language delay, whereas more rapid sleep consolidation positively predicts language learning.28 Similarly, Lam et al20 report that daytime napping was associated with poorer neurocognitive function in children aged 3–5 years and suggest that cessation of napping may be a marker of brain maturation. The corpus of cognition studies directs attention to both the benefits of napping for learning and memory, especially among younger children,26 ,27 and the changing value of napping across time.20 Sleep consolidation may be a global marker of neurological maturity.

Evidence on the implications of napping for physical health was poor. Two studies report that, among preschoolers, napping serves to protect against accidental injury.31 ,35 Both studies, however, have the serious limitation of dependence on retrospective recall at the time of attending accident and emergency departments.31 ,35 The only other published studies pertaining to napping and health were three reporting on the association with childhood obesity.32–34 All reported shorter sleep duration as a predictor of excessive weight or obesity, but no independent effect for napping. These findings suggest that the function of napping with regard to weight status is potentially different from that of night sleep. Considerably more research is required before there can be any certainty about the effects of napping on children's physical health.

In summary, there is currently a low volume of evidence regarding the impacts of napping on children's development and health. Extant literature covers a range of outcomes with few using standard, comparable measures. The quality of studies reflects the relatively new focus of research on the independent effects of napping. There is an imperative for increased understanding about individual differences in napping trajectories and, indeed, sleep needs within this age group. Future studies should address the dual complexities of sleep transition across early childhood and the impacts of environmental manipulations at home and in the non-parental care context and should consider multiple child outcomes.