Content uploaded by Anna Price
Author content
All content in this area was uploaded by Anna Price on Feb 12, 2014
Content may be subject to copyright.
Children’s sleep patterns from 0 to 9 years:
Australian population longitudinal study
Anna M H Price,
1,2
Judith E Brown,
3
Michael Bittman,
3
Melissa Wake,
1,2,4
Jon Quach,
1,2
Harriet Hiscock
1,2,4
1
Murdoch Childrens Research
Institute, Parkville, Victoria,
Australia
2
Centre for Community Child
Health, The Royal Children’s
Hospital, Parkville, Victoria,
Australia
3
School of Behavioural,
Cognitive and Social Sciences,
University of New England,
Armidale, North South Wales,
Australia
4
Department of Paediatrics,
The University of Melbourne,
Parkville, Victoria, Australia
Correspondence to
Dr Anna Price, Centre for
Community Child Health,
The Royal Children’s Hospital,
Flemington Road, Parkville,
VIC 3052, Australia;
anna.price@mcri.edu.au
Received 27 March 2013
Revised 16 October 2013
Accepted 29 October 2013
▸http://dx.doi.org/10.1136/
archdischild-2013-304083
To cite: Price AMH,
Brown JE, Bittman M, et al.
Arch Dis Child Published
Online First: [please include
Day Month Year]
doi:10.1136/archdischild-
2013-304150
ABSTRACT
Objective To provide accurate population normative
data documenting cross-sectional, age-specific sleep
patterns in Australian children aged 0–9 years.
Design and setting The first three waves of the
nationally representative Longitudinal Study of Australian
Children, comprising two cohorts recruited in 2004 at
ages 0–1 years (n=5107) and 4–5 years (n=4983), and
assessed biennially.
Participants Children with analysable sleep data for at
least one wave.
Measures At every wave, parents prospectively
completed 24-h time-use diaries for a randomly selected
week or weekend day. ‘Sleeping, napping’was one of
the 26 precoded activities recorded in 15-min time
intervals.
Results From 0 to 9 years of age, 24-h sleep duration
fell from a mean peak of 14 (SD 2.2) h at 4–6 months
to 10 (SD 1.9) h at 9 years, mainly due to progressively
later mean sleep onset time from 20:00 (SD 75 min) to
21:00 (SD 60 min) and declining length of day sleep
from 3.0 (SD 1.7) h to 0.03 (SD 0.2) h. Number and
duration of night wakings also fell. By primary school,
wake and sleep onset times were markedly later on
weekend days. The most striking feature of the centile
charts is the huge variation at all ages in sleep duration,
sleep onset time and, especially, wake time in this
normal population.
Conclusions Parents and professionals can use these
new centile charts to judge normalcy of children’s sleep.
In future research, these population parameters will now
be used to empirically determine optimal child sleep
patterns for child and parent outcomes like mental and
physical health.
INTRODUCTION
Insufficient or poor-quality sleep in childhood is
associated with serious negative consequences
including poorer emotional, behavioural and cogni-
tive functioning, increased injury and obesity, and
poorer parental mental and general health.
1–4
The
cost of childhood sleep problems is considerable.
For Australian families, the average cost associated
with seeking professional healthcare to manage
infant sleep problems in the second 6 months of
life totals $A380 per family (adjusted for inflation
to 2012).
5
Unpublished population data indicate
that sleep problems in children aged 0–7 years
(estimated population 1.14 million) are associated
with a $A15.3 million cost to government in add-
itional health services every year.
6
Matricciani’s recent systematic review verified
the common perception that sleep duration in
childhood (5–18 years) is decreasing.
7
Data from
218 studies (n=690 747 from 20 countries)
showed that the median decrease in children’s sleep
duration was 0.75 min per year since 1905. This
could be contributing to the rise in morbidities
such as childhood obesity and attention deficit dis-
order recorded over recent decades.
8
It is equally possible that too much sleep is detri-
mental to health. In a recent critical review, some
adult studies suggested that short (<7 h) and long
(≥8 h) nightly sleep duration could be associated
with obesity (ie, a non-linear association).
9
Although comparable studies with children suggest
only a negative linear relationship, more evidence
could reveal complex, non-linear relationships.
9
Finally, and independently of duration, sleep timing
and fragmentation may be important to children’s
health. Olds et al
10
, studying time diary data in
2200 Australians aged 9–16 years, compared two
groups with the same total sleep duration. Those
What is already known
▸Research interest in infant and child sleep has
rapidly increased because of their relevance to
‘modern’problems such as obesity and
attention deficit disorder.
▸Starting from infancy, there are steady
age-related declines in duration, number/length
of night wakes and length of daytime sleeps.
▸However, current reference values are largely
based on inaccurate parental summary or
‘stylised’recall of sleep parameters, rather than
accurately recorded population-level sleep data.
What this study adds
▸Time-use diaries provide population normative
centile curves for sleep duration and, for the
first time, sleep onset times and wake times
throughout infancy and childhood.
▸There is a striking range and steady decline in
sleep duration, number of sleep episodes,
number/length of night wakes and length of
day sleeps.
▸These population parameters can be used to
determine optimal sleep patterns for children’s
behavioural, emotional and cognitive
outcomes, and parent outcomes like mental
health.
Price AMH, et al.Arch Dis Child 2013;0:1–7. doi:10.1136/archdischild-2013-304150 1
Original article
ADC Online First, published on December 16, 2013 as 10.1136/archdischild-2013-304150
Copyright Article author (or their employer) 2013. Produced by BMJ Publishing Group Ltd (& RCPCH) under licence.
group.bmj.com on December 17, 2013 - Published by adc.bmj.comDownloaded from
who rose and went to bed early were more physically active,
while those who rose and went to bed late reported more screen
time and had higher body mass index z scores.
10
Studies are yet
to examine the effect of sleep fragmentation on children’s
health; as Jenni
11
notes, such measures are missing from existing
normative research.
Galland’s
12
meta-analysis of sleep patterns in children aged
0–12 years summarises what is known about child sleep globally.
However, much of the reviewed literature relies on parents’
summary recall
12 13
that are known to be inherently inaccur-
ate.
14
Williams et al
15 16
addressed this by developing US popu-
lation norms for 24-h and daytime sleep duration based on
time-use diaries in 0–18-year-olds. Such diaries approximate
detailed written descriptions of daily activities. They are known
to reliably and validly record daily activities
17
and produce
more precise and accurate recollection of time spent in particu-
lar activities than summary or ‘stylised’recall.
18–20
Despite the known importance of adequate sleep for health,
recommendations for optimal childhood sleep parameters are
traditionally based on opinion rather than empirical evidence.
8
It
is thus unknown whether these recommendations identify optimal
sleep patterns for child and parent outcomes. The first necessary
step is to extract an accurate, contemporary description of the
range of population-based sleep parameters from infancy through
childhood. This will allow future analyses to examine cross-
sectional associations between sleep and a range of morbidities
and generate evidence-based sleep recommendations. Such
population-level data will make it possible to delineate whether
there are longitudinal sleep trajectories that are associated with
good and poor outcomes. We draw parallels with the advances in
epidemiology that became possible when internationally agreed
paediatric body mass index cut-points were first developed a
decade ago.
21 22
The nationally representative Longitudinal Study of
Australian Children (LSAC)
23
allows us to extend Williams’
research and address this challenge for the first time. We there-
fore aimed to document the cross-sectional, age-specific sleep
patterns of Australian children aged 4 months to 9 years, to
produce population-based centile charts for clinical use.
METHODS
Design and setting
Data are from the first three waves of the nationally representa-
tive LSAC. A complete description of the design and sample is
published elsewhere.
24
In brief, a two-stage cluster sampling
design was used to create two cohorts, birth (B ‘Baby’cohort:
aged 0–1 years in 2004) and preschool (K ‘Kindergarten’
cohort: aged 4–5 years in 2004), who are assessed every 2 years.
Both cohorts were enrolled during the same period from the
same geographical postcodes, but were sampled independently.
In the first stage, postcodes (except the most remote) were
sampled after stratifying by the state of residence and urban
versus rural status to ensure proportional geographical and
socioeconomic representation. All children in the relevant age
ranges registered on the Australian Medicare Database (98% of
Australian children) were randomly selected within each post-
code to participate in the trial.
Of the 7980 families invited to join the B cohort and 8446
families invited to join the K cohort, 5107 (64%) infants and
4983 (59%) children, respectively, participated in wave 1. The
final LSAC sample was proportionally representative of
Australian children based on urban versus rural geographical
location and by state, except that mothers who had completed
high school were over-represented and families with low
incomes under-represented.
24
At wave 3 in 2008, 4386/5107
(86%) B cohort children (aged 4–5 years) and 4332/4983 (87%)
K cohort children (aged 8–9 years) were retained. Retention was
marginally lower for children with less highly educated parents
and from non-English-speaking background.
25
Procedures
At all waves, trained researchers administered a face-to-face
caregiver interview and conducted direct child assessments in
the child’s home. Researchers then left two time-use diaries (see
Measures below) with primary caregivers to complete on one
randomly selected week and one weekend day, using a desig-
nated day approach where respondents were told which days to
do a diary. Parents returned the diaries by post.
Measures
For the first three waves, LSAC included a ‘light’time-use diary (see
figure 1 for a sample diary completed at 4–5 years).
26 27
Parents
(usually the mother) completed the 24-h diaries, beginning at 04:00
and ending at 04:00 the following day. Beginning at 04:00 is a con-
vention which, while arbitrary, is widely used because at that time
virtually the entire population is undertaking their overnight repose.
Unlike (say) midnight or 05:00, it thus gives a very clear separation
between1day’s activities and the next for all except night workers.
A diary could be completed prospectively through the day or all
at once at the end of the day. As any of the 26 precoded activities
occurred, mothers indicated its duration in 15-min time intervals.
Sleep was represented by the category ‘sleep, napping’.An
‘episode’was defined by a change in activity or context of any
amount of time. Light time-use diaries are derived from and are
equally valid to traditional full-length time-use diaries, which
require detailed written descriptions of daily activities.
17
For the purposes of this paper, we defined sleep parameters
as follows:
▸Sleep onset time: start of first sleep episode occurring after
19:00. If the child was asleep at 19:00, sleep onset time was
defined as the beginning of that sleep episode providing that
episode was at least 90 min in duration.
▸Wake time: distinguished from an interruption of sleep in the
morning period (from 04:00) if the child was awake for at
least 75 min before the next sleep episode.
▸Number of night wakes: sum of night wakes from 04:00
until wake time and from 19:00 to 04:00. The average dur-
ation of night wakes was calculated over the same periods.
▸Length of daytime sleep: any sleep occurring between wake
and sleep onset times.
▸24-h sleep duration: sum of morning sleep from 4:00 until
wake time, daytime sleep and night-time sleep from sleep
onset time until 4:00 the following morning.
Sleep parameters varied by age more in infancy than child-
hood, so we classified wave 1 of the B cohort in three monthly
age groups, that is, 4–6 months, 7–9 months, 10–12 months
and 13–15 months. At all other waves, we used six monthly age
groups. Given the age range of children at each wave, not all
the consecutive monthly age groups from 0 to 9 years are repre-
sented by these data (see table 2).
Statistical analyses
We applied conservative data cleaning strategies and removed
poor-quality diaries, defined as having more than 150 min of
missing data (excluding time at child care), <10 episodes and/or
>5 episodes where more than five activities took place simultan-
eously. Traditionally, with time-use data, a cut-off of 90 min is
used. However, this loses a large proportion of the sample. We
2 Price AMH, et al.Arch Dis Child 2013;0:1–7. doi:10.1136/archdischild-2013-304150
Original article
group.bmj.com on December 17, 2013 - Published by adc.bmj.comDownloaded from
chose 150 min as this is what our statistician (JEB) has used in a
previous analysis of time-use data. The analytic sample com-
prised families who returned at least one good-quality diary
across the three waves. We calculated means and SDs for the
sleep parameters, taking account of the complex survey design,
using Stata SE V.8.2 for Windows (Stata, College Station, Texas,
USA). We present the data without controlling for repeated mea-
sures because we want to present unadulterated data where pos-
sible, and colleagues’previous analysis of LSAC data suggests
little reason to expect that controlling for repeated measures
would substantially alter the patterns.
28
Existing papers reporting normative sleep data use a range of
techniques to analyse the data, including least mean square,
linear regression models and Kernel plots.
16 29
We chose to plot
Loess curves (smoothing factor 0.75) for sleep duration, and
sleep onset and wake times from Gaussian centiles calculated
for each age group. Centiles were calculated using the formula:
Centile ¼
xþzs
Where x is the sample mean, z is the standardised z score for
each centile (eg, the z score for the 2nd and 98th centile was
2.05) and s is the sample SD. We plotted mean sleep onset and
wake times by week and weekend days. A ‘weekend’was defined
using Friday and Saturday night sleep onset times and Saturday
and Sunday wake times. Loess curves and column graphs for
sleep onset and wake times were plotted using ‘R’(V.2.13). To
determine whether to present data by gender, we conducted a
simple sensitivity analysis to examine whether there were differ-
ences in sleep duration between males and females. As there
were few differences between genders across the age categories,
we report results for the full sample.
RESULTS
Respondent characteristics
The B cohort provided 6976 useable diaries (n=3837) for wave
1, 5924 diaries (n=3309) for wave 2 and 5139 diaries
(n=2856) for wave 3. The K cohort provided 6207 diaries
(n=3563) for wave 1, 5719 diaries (n=3247) for wave 2 and
4976 diaries (n=2802) for wave 3. Table 1 shows that, com-
pared with children without analysable diary data, children with
these data were similar in gender and age, but more socio-
economically advantaged.
Figure 1 Example of Longitudinal Study of Australian Children (LSAC) weekday light time-use diary from wave 3 B cohort.
Price AMH, et al.Arch Dis Child 2013;0:1–7. doi:10.1136/archdischild-2013-304150 3
Original article
group.bmj.com on December 17, 2013 - Published by adc.bmj.comDownloaded from
Sleep patterns of Australian children
Table 2 and figures 2–4 show a striking range in sleep patterns.
The number and length of night wakes, number of sleep epi-
sodes and length of daytime and total sleep duration were great-
est at age 4–6 months, showing a steep decline in the first
3 years before a flatter, continuing decline to 9 years. The vari-
ation in sleep parameters followed a similar pattern, except for
an increase in total sleep duration from 6–9 years and a brief
increase in length of night wakes at 6.5 years.
The range in sleep duration (figure 2) was large throughout,
though the width and timing of the span differed somewhat by age.
At the age of 4–6 months, more than 8 h of sleep per day separated
the 2nd and 98th centiles (spanning 10–18 h), falling to a difference
of just over 5 h at age 5 (spanning 9–14 h) before rising again to a
difference of around 8 h at age 9 (spanning 6–14 h).
This reduction in total sleep duration was driven mainly by
two factors: progressively later sleep onset times (figure 3),
coupled with a reduction then cessation of daytime sleep. In
contrast, mean wake time (figure 4) stayed relatively stable over
time, although its variability increased markedly in the older
children, to a range of around 9 h (3:45–15:45) at 9 years.
Some of this widening in centiles represents the differences in
wake times between week and weekend days, both of which are
included in these centile charts. Figure 5 shows that from com-
mencement of school, children progressively woke up and had a
later sleep onset on weekends. For these older children, the
weekend–weekday difference was greatest for wake rather than
sleep onset times, ranging from less than 15 min at age 3, climb-
ing to 50 min at age 7 and >60 min at age 9 years. Similarly,
the difference in mean sleep onset times between week and
weekend days was less than 15 min at age 3, increasing to 20,
30 and 35 min for ages 5, 7 and 9 years, respectively.
Sleep patterns were similar for the overlapping ages in B and
K cohorts, apart from night wakes and daytime sleep, which
Table 1 Comparing baseline demographic characteristics of families with versus without analysable sleep diary data in the two cohorts
Characteristics
B Cohort K Cohort
Whole cohort
n=4831–5107
Analysable sleep diary data
Whole cohort
n=4659–4983
Analysable sleep diary data
Yes
n=2507–2625
No
n=2324–2482 p Value
Yes
n=3386–3563
No
n=1273–1420 p Value
Child
Male, % 51.2 50.1 51.9 0.4 50.9 51.5 49.5 0.2
Age in months, mean (SD) 9.2 (2.6) 9.1 (2.6) 9.4 (2.5) <0.0001 57.4 (2.6) 57.3 (2.5) 57.5 (2.8) <0.0001
Born in Australia/New Zealand, % 81.3 82.9 79.7 0.004 77.7 86.1 71.8 <0.0001
Primary caregiver
Age in years, mean (SD) 31.0 (5.5) 31.2 (5.3) 30.8 (5.7) <0.0001 34.7 (5.5) 35.0 (5.2) 34.0 (6.1) <0.0001
Born in Australia/New Zealand, % 81.4 82.9 79.9 0.004 77.7 80.1 71.8 <0.0001
English mainly spoken at home, % 85.6 87.7 83.4 <0.0001 84.4 87.9 75.6 <0.0001
Education status, % <0.0001 <0.0001
Did not complete high school 31.7 28.5 35.2 39.6 35.2 50.8
Completed high school 35.4 36.5 34.3 32.2 33.2 30.0
Completed university degree 32.8 35.1 30.5 28.1 31.7 19.2
Equivalised yearly household
income ($A), mean (SD)
31 745 (17 053) 32 558 (16 523) 30 868 (17 569) <0.0001 31 857 (16 612) 33 478 (16 346) 27 543 (16 550) <0.0001
Married/de facto, % 90.5 92.5 88.5 <0.0001 85.9 88.7 78.8 <0.0001
Table 2 Child sleep patterns (mean (SD)) by age groupings, in the two cohorts
Wave
Age
years (months) N Sleep (h)
No of sleep
episodes
No of
night wakes
Night
wakes (min)
Day
sleep (h)
Day
sleep* (h)
Wake
time†(am)
Sleep onset
time†(pm)
1 0.5 (4–6) 554 14.0 (2.2) 6.1 (1.9) 1.1 (1.2) 26.9 (33.8) 3.0 (1.7) 3.0 (1.7) 7:30 (1:30) 8:00 (1:15)
1 0.75 (7–9) 1573 13.6 (2.1) 5.4 (1.7) 1.0 (1.2) 20.3 (30.5) 2.7 (1.5) 2.8 (1.4) 7:15 (1:30) 8:00 (1:15)
1 1.0 (10–12) 1306 13.4 (2.0) 4.7 (1.5) 0.7 (1.0) 14.3 (22.5) 2.5 (1.4) 2.6 (1.3) 7:00 (1:15) 8:00 (1:15)
1 1.25 (13–15) 388 13.4 (1.9) 4.2 (1.4) 0.5 (0.8) 11.7 (21.7) 2.4 (1.3) 2.5 (1.3) 7:00 (1:15) 8:00 (1:15)
2 2.5 (28–33) 1275 11.9 (1.6) 2.8 (0.6) 0.2 (0.6) 4.4 (14.1) 1.0 (1.1) 1.2 (1.1) 7:15 (1:15) 8:15 (1:00)
2 3.0 (34–39) 1929 11.7 (1.6) 2.6 (0.7) 0.2 (0.5) 3.8 (16.1) 0.8 (1.0) 1.0 (1.1) 7:15 (1:15) 8:15 (1:00)
3 (B) 4.5 (52–57) 1251 11.1 (1.4) 2.2 (0.5) 0.1 (0.3) 1.8 (12.8) 0.2 (0.5) 0.2 (0.6) 7:15 (1:00) 8:15 (1:00)
1 (K) 4.5 (52–57) 1905 11.2 (1.5) 2.4 (0.7) 0.1 (0.4) 2.8 (12.9) 0.3 (0.7) 0.4 (0.8) 7:15 (1:30) 8:30 (1:00)
3 (B) 5.0 (58–63) 1549 11.1 (1.3) 2.2 (0.5) 0.1 (0.3) 1.4 (10.1) 0.1 (0.5) 0.2 (0.6) 7:15 (1:45) 8:15 (1:00)
1 (K) 5.0 (58–63) 1635 11.0 (1.3) 2.3 (0.7) 0.1 (0.3) 2.1 (13.1) 0.2 (0.6) 0.3 (0.7) 7:15 (1:30) 8:30 (1:00)
2 6.5 (76–81) 1292 10.5 (1.8) 2.0 (0.4) 0.1 (0.3) 2.3 (21.2) 0.04 (0.3) 0.1 (0.4) 7:30 (2:45) 8:45 (1:00)
2 7.0 (82–87) 1864 10.4 (1.8) 2.0 (0.4) 0.1 (0.3) 2.1 (16.3) 0.04 (0.3) 0.1 (0.4) 7:30 (2:45) 8:45 (0:45)
3 8.5 (100–105) 1278 10.3 (2.0) 2.0 (0.4) 0.04 (0.2) 1.4 (12.4) 0.03 (0.3) 0.04 (0.3) 7:45 (2:45) 9:00 (1:00)
3 9.0 (106–111) 1467 10.0 (1.9) 2.0 (0.4) 0.04 (0.2) 1.7 (15.5) 0.03 (0.2) 0.04 (0.3) 7:45 (3:00) 9:00 (1:00)
*Sleep time for those not attending day care on the diary day between wake and sleep onset times.
†SD for wake and sleep onset times is hours:minutes.
4 Price AMH, et al.Arch Dis Child 2013;0:1–7. doi:10.1136/archdischild-2013-304150
Original article
group.bmj.com on December 17, 2013 - Published by adc.bmj.comDownloaded from
were slightly longer for the K cohort, and sleep onset time,
which was 15 min later.
DISCUSSION
Principal findings
This is the largest and most detailed prospective population-
based study to document child sleep using accurate time-diary
data, and this is the first to present sleep onset and wake time
centiles based on this method. Duration of most sleep para-
meters decreased with age, although the variation in sleep dur-
ation and number of night wakes increased around school age.
Weekend and weekday sleep onset and wake times became less
synchronous in school-aged children.
Study strengths
These time-use data provide finer-grained, more accurate mea-
sures of child sleep than parent estimates/summaries,
10
on
which much of the existing literature is based. The prospective
nature of the sleep measure limits recall bias while capturing
fluctuations in sleep patterns. The population-based sampling
could allow these findings to generalise to families in many
advantaged countries with school schedules and cultural prac-
tices comparable to Australia.
Study limitations
We collected subjective parent report rather than an objective
sleep measure like actigraphy. As such, our findings may under-
estimate the number of sleep episodes and night wakes and the
length of night wakes in children aged 0–9 years.
14
Although
not perfect, sleep diaries approximate actigraphy better than
summary data for parent-reported child sleep and, while actigra-
phy is considered more objective, it is not a ‘gold standard’and
is difficult to collect for large-scale population-based research.
We are not sure of the reason for the differences between
cohorts but, as they are small, we do not expect them to be
meaningful. A further limitation is that our findings may not
fully generalise to the most vulnerable families for whom, given
their greater exposure to chaotic living circumstances, have
more sleep problems.
30
Finally, undoubtedly individual children
vary from day to day, and some children vary more than others.
However, because most children contributed only one with a
maximum of two diaries, we did not explore this further. With
our large sample sizes, we do not expect that this would alter
our cross-sectional population norms.
Interpretation in the light of other studies
Sleep duration for the current sample matched the largest exist-
ing Australian and New Zealand survey of sleep patterns in
0–3-year-olds
12 31
and was similar to the largest English cohort
studied from 6 months to 11 years.
29
However, it was consist-
ently (approximately 1 h) longer than Williams’normative US
data and Galland’s meta-analysis.
The number of night wakes for the current sample remained
low from infancy to 9 years. This suggests that it is the length of
night wakes (and accompanying disruption), rather than the
number that contributes to the high proportion of sleep
Figure 3 Centiles for sleep onset
times (pm) in the two cohorts.
Figure 2 Centiles for total sleep
duration per 24 h by age in the two
cohorts.
Price AMH, et al.Arch Dis Child 2013;0:1–7. doi:10.1136/archdischild-2013-304150 5
Original article
group.bmj.com on December 17, 2013 - Published by adc.bmj.comDownloaded from
problems reported by parents in the first year of life.
232
Most
children learn to self-settle by 12 months,
33
which may explain
why the length of night wakes dropped considerably after
infancy.
Wake times for Australian children are similar to English and
Swiss children from 0 to 5 years, with a greater pattern of differ-
ences emerging from 5 years of age.
13 29
These differences may
be related to difference in school start hours. Interestingly, week
and weekend day differences for school-aged children were
greater for wake rather than sleep onset times. Reasons for this
are unclear, but it could be that parents are more aware of when
their children wake up than when they go to sleep or when chil-
dren catch up on sleep during weekends.
Unanswered questions and future research
The ‘problematisation’of sleep, according to Matricciani et al
8
,
is the tendency for children’s sleep to be considered inadequate,
despite a lack of evidence. Our data provide a useful empirical
starting point from which our future research will determine
which sleep patterns impact most on child and parent outcomes,
whether such effects are linear or non-linear and whether clear
thresholds emerge for sleep duration, sleep onset time and/or
wake time beyond which certain outcomes are less optimal.
Ideally, when collecting population-level sleep data, an objective
measure like actigraphy would be collected from a
representative subsample and extrapolated to the full sample to
estimate normative sleep patterns as accurately as possible.
Implications
There is a wide range in ‘normal’child sleep from 0 to 9 years.
Practitioners can use these centile charts to better counsel fam-
ilies about the normalcy or otherwise of their child’s sleep. We
hope these data will lead directly to research identifying adap-
tive child sleep patterns, so practitioners could accurately target
sleep interventions to families most at risk of the adverse effects
of non-optimal child sleep.
Acknowledgements This paper uses confidentialised unit record files from the
Longitudinal Study of Australian Children (LSAC) survey. The LSAC project was
initiated and is funded by the Commonwealth Department of Families, Housing,
Community Services and Indigenous Affairs (FaHCSIA) and is managed by the
Australian Institute of Family Studies. The findings and views reported in this paper,
however, are those of the authors and should not be attributed to either FaHCSIA or
the Australian Institute of Family Studies.
Contributors MB, MW and HH conceived the original analyses. AMHP, JEB, MB,
MW, JQ and HH wrote the manuscript. JEB had full access to all the data in the
study and takes responsibility for the integrity of the data and the accuracy of the
data analysis.
Funding AMHP and JQ were supported by NHMRC Population Health Capacity
Building Grant #436914; MW by NHMRC Population Health Career Development
Award #546405; and HH by NHMRC Career Development Award 607351. The
Murdoch Childrens Research Institute (MCRI) administered the grants and provided
infrastructural support to its staff but played no role in the conduct or analysis of the
trial. MCRI research is supported by the Victorian Government’s Operational
Infrastructure Support Program.
Competing interests None.
Ethics approval The study was approved by the Australian Institute of Family
Studies Ethics Committee, and parents provided written informed consent.
Provenance and peer review Not commissioned; externally peer reviewed.
REFERENCES
1 Hiscock H, Canterford L, Ukoumunne OC, et al. Adverse associations of sleep
problems in Australian preschoolers: national population study. Pediatrics
2007;119:86–93.
2 Martin J, Hiscock H, Hardy P, et al. Adverse associations of infant and child sleep
problems and parent health: an Australian population study. Pediatrics
2007;119:947–55.
3 Quach J, Hiscock H, Canterford L, et al. Outcomes of child sleep problems over the
school-transition period: Australian population longitudinal study. Pediatrics
2009;123:1287–92.
4 Chen X, Beydoun MA, Wang Y. Is sleep duration associated with childhood
obesity? A systematic review and meta-analysis. Obesity (Silver Spring, Md
2008;16:265–74.
5 Hiscock H, Bayer J, Gold L, et al. Improving infant sleep and maternal mental
health: a cluster randomised trial. Arch Dis Child 2007;92:952–8.
6 Quach J, Gold L, Hiscock H, et al. Health care costs associated with sleep problems
up to age 7 years: Australian population-based study. (Unpublished).
7 Matricciani L, Olds T, Petkov J. In search of lost sleep: Secular trends in the sleep
time of school-aged children and adolescents. Sleep Med Rev 2012;16:203–11.
Figure 5 Sleep onset times and wake times by week/weekend day
and age in the two cohorts.
Figure 4 Centiles for wake times
(am) in the two cohorts.
6 Price AMH, et al.Arch Dis Child 2013;0:1–7. doi:10.1136/archdischild-2013-304150
Original article
group.bmj.com on December 17, 2013 - Published by adc.bmj.comDownloaded from
8 Matricciani LA, Olds TS, Blunden S, et al. Never enough sleep: a brief history of
sleep recommendations for children. Pediatrics 2012;129:548–56.
9 Marshall NS, Glozier N, Grunstein RR. Is sleep duration related to obesity? A critical
review of the epidemiological evidence. Sleep Med Rev 2008;12:289–98.
10 Olds TS, Maher CA, Matricciani L. Sleep duration or bedtime? Exploring the
relationship between sleep habits and weight status and activity patterns. Sleep
2011;34:1299–307.
11 Jenni OG. How much sleep is “normal”in children and adolescents? JAMA
Pediatrics 2012;167:91–2.
12 Galland BC, Taylor BJ, Elder DE, et al. Normal sleep patterns in infants and
children: a systematic review of observational studies. Sleep Med Rev
2012;16:213–22.
13 Iglowstein I, Jenni OG, Molinari L, et al. Sleep duration from infancy to
adolescence: reference values and generational trends. Pediatrics 2003;111:302–7.
14 Werner H, Molinari L, Guyer C, et al. Agreement rates between actigraphy, diary,
and questionnaire for children’s sleep patterns. Arch Pediatr Adolesc Med
2008;162:350–8.
15 Williams J, Zimmerman FJ, Bell JF. Norms and trends of sleep time among US
children and adolescents. JAMA Pediatrics 2012;167:55–60.
16 Williams JA, Zimmerman FJ, Bell JF. Norms and trends of sleep time among US
children and adolescents. JAMA Pediatrics 2013;167:55–60.
17 Gatenby R. Comparison of time use data from pre-coded diaries and fully written
diaries. The 25th IATUR Conference on Time Use Research;17–19 September
2003, Belgium: Free University of Brussels, 2003.
18 Niemi I. Systematic error in behavioural measurement: comparing results from
interview and time budget studies. Soc Indicators Res 1993;30:229–44.
19 Robinson JP. The Validity and Reliability of Diaries versus Alternative Time Use
Measures. In: Juster TF, Stafford FP, eds. Time, goods, and well-being. Ann Arbor,
MI: University of Michigan, 1985:33–62.
20 van der Ploeg HP, Merom D, Chau JY, et al. Advances in population surveillance for
physical activity and sedentary behavior: reliability and validity of time use surveys.
Am J Epidemiol 2010;172:1199–206.
21 Cole T, Bellizi M, Flegal K, et al. Establishing a standard definition for child
overweight and obesity worldwide: international survey. BMJ 2000;320:
1240–3.
22 Kuczmarski RJ, Ogden CL, Grummer-Strawn LM,, et al CDC growth charts: United
States. Adv Data 2000;(314):1–27.
23 Sanson A. Growing Up in Australia: The first 12 months of landmark study. Family
Matters. (64):40.
24 Soloff C, Lawrence D, Johnstone R. LSAC technical reference paper number 1:
sample design 2005 [cited 2013 November 19]. http://www.growingupinaustralia.
gov.au/pubs/technical/tp1.pdf
25 Sipthorp M, Misson S. Wave 3 weighting and non-response. LSAC technical paper
no.6 2009 [cited 2013 November 19]. http://www.growingupinaustralia.gov.au/
pubs/technical/tp6.pdf
26 Tey C, Wake M, Campbell M, et al. The Light Time-Use Diary and preschool activity
patterns: exploratory study. Int J Pediatr Obes 2007;2:167–73.
27 Gershuny J. Time-use surveys and the measurement of national well-being. Oxford,
UK: Centre for Time-use Research, Department of Sociology, University of Oxford,
2011; 12 September 2011.
28 Wake M, Clifford SA, Patton GC, et al. Morbidity patterns among the underweight,
overweight and obese between 2 and 18 years: population-based cross-sectional
analyses. Int J Obes (Lond) 2013;37:86–93.
29 Blair PS, Humphreys JS, Gringras P, et al. Childhood sleep duration and associated
demographic characteristics in an English cohort. Sleep 2012;35:353–60.
30 Arman AR, Ay P, Fis NP, et al. Association of sleep duration with socio-economic
status and behavioural problems among schoolchildren. Acta Paediatrica
2011;100:420–4.
31 Teng A, Bartle A, Sadeh A, et al. Infant and toddler sleep in Australia and New
Zealand. J Paediatr Child Health 2012;48:268–73.
32 Hiscock H, Wake M. Infant sleep problems and postnatal depression: a
community-based study. Pediatrics 2001;107:1317–22.
33 Henderson JMT, France KG, Owens JL, et al. Sleeping through the night: the consolidation
of self-regulated sleep across the first year of life. Pediatrics 2010;126:e1081–7.
Price AMH, et al.Arch Dis Child 2013;0:1–7. doi:10.1136/archdischild-2013-304150 7
Original article
group.bmj.com on December 17, 2013 - Published by adc.bmj.comDownloaded from
doi: 10.1136/archdischild-2013-304150
published online December 16, 2013Arch Dis Child
Anna M H Price, Judith E Brown, Michael Bittman, et al.
Australian population longitudinal study
years: Children's sleep patterns from 0 to 9
http://adc.bmj.com/content/early/2013/11/25/archdischild-2013-304150.full.html
Updated information and services can be found at:
These include:
References http://adc.bmj.com/content/early/2013/11/25/archdischild-2013-304150.full.html#ref-list-1
This article cites 25 articles, 10 of which can be accessed free at:
P<P Published online December 16, 2013 in advance of the print journal.
service
Email alerting the box at the top right corner of the online article.
Receive free email alerts when new articles cite this article. Sign up in
Collections
Topic
(1035 articles)Epidemiologic studies (38 articles)Press releases
Articles on similar topics can be found in the following collections
Notes
(DOIs) and date of initial publication.
publication. Citations to Advance online articles must include the digital object identifier
citable and establish publication priority; they are indexed by PubMed from initial
typeset, but have not not yet appeared in the paper journal. Advance online articles are
Advance online articles have been peer reviewed, accepted for publication, edited and
http://group.bmj.com/group/rights-licensing/permissions
To request permissions go to:
http://journals.bmj.com/cgi/reprintform
To order reprints go to:
http://group.bmj.com/subscribe/
To subscribe to BMJ go to:
group.bmj.com on December 17, 2013 - Published by adc.bmj.comDownloaded from