BACKGROUND AND OBJECTIVES:

Parents often do not accurately perceive overweight and/or obesity in their children. Changing this is widely considered an essential first step to reducing child overweight, but recent research suggests that, in fact, this could promote greater weight gain. We aimed to determine the directionality over time between higher child adiposity and parental perception of child overweight.

METHODS:

Participants were from 2 cohorts of the population-based Longitudinal Study of Australian Children followed biennially since 2004. Repeated measures of BMI z scores and parental perceptions of overweight were available for the kindergarten cohort at 6 waves (ages 4–5, 6–7, 8–9, 10–11, 12–13, and 14–15 years; n = 4632) and for the birth cohort at 4 waves (ages 2–3, 4–5, 8–9, and 10–11 years; n = 4445). Bidirectionality between overweight perception and BMI z score was examined by using cross-lagged regression models.

RESULTS:

In both cohorts, wave-on-wave lagged effects were strong (all: P < .001) but much larger from BMI z score to parent perception. For every unit increase in the BMI z score, the odds of a child being perceived as overweight in the next wave ranged from 2.9 (birth cohort: age 2–3 years) to 10.4 (kindergarten cohort: age 6–7 years). These effects were ∼3 to 12 times larger than the reverse, whereby the perception of overweight predicted 0.2 to 0.5 higher BMI z score in the next wave.

CONCLUSIONS:

Higher child BMI z scores strikingly predicted a subsequent parental perception of child overweight. Parent-perceived overweight preceded rising (not falling) BMI, but these effects were small. Clinician efforts to make parents aware of overweight may not be harmful but seem unlikely to improve children’s BMI status.

What’s Known on This Subject:

It is widely stated that parents must be aware of their child’s overweight status to effect change. However, recent research has suggested this could in fact be harmful; when parents perceive overweight, children may actually gain more weight throughout childhood.

What This Study Adds:

In cohorts of children age 2 to 15 years with a biennial BMI and parent-perceived weight status, higher BMI strikingly predicted parent-perceived overweight, which only weakly preceded rising (not falling) BMI. Perception is thus a response to rising BMI; increasing perception may be neither helpful nor harmful.

Many parents do not consider their children with overweight or obesity to be overweight or obese.1,2 The prevailing ethos of many antiobesity activities for children (from primary to secondary care and/or prevention to tertiary treatment) is that parents must be aware of and concerned about their child’s overweight status before "success" can ensue. Prochaska et al’s3 transtheoretical model has therefore provided a convenient starting point for universal screening and primary care interventions over the last 2 decades.4,7 The reasoning is that when parents recognize and become appropriately concerned about their children’s overweight, then they will instigate effective changes such that children’s subsequent gains (or relative losses) in BMI should be smaller. In other words, correct parental perception of a child as overweight could be beneficial and could even be a prerequisite to healthier BMI.

In their recent article in Pediatrics, Robinson and Sutin8 challenge this assumption. Across 5 waves of biennial data from the Longitudinal Study of Australian Children (LSAC) (spanning ages 4–13 years), parental identification of child overweight did not protect against future weight gain; instead, it was consistently associated with greater weight gain throughout childhood. These findings held regardless of whether the parents were accurate or inaccurate in their perception of overweight. Similarly, within a smaller Dutch cohort of children, Gerards et al9 demonstrated that children with overweight whose parents accurately perceived them to be overweight gained more weight in the future. Rather than beneficial, these scenarios reframed parental perception of child overweight as actively harmful to efforts to curb obesity.

In recent follow-up analyses, Robinson and Sutin10 show that in the same cohort, the children whose parents perceive them to be overweight have a negative view of their body size and are attempting to lose weight. They make a number of suggestions as to why parental perception of overweight could be actively harmful and could indeed increase future weight gain. Their possibilities include the stigma of parental perception of overweight counteracting children’s self-regulation (with attempted weight loss eventually leading to weight gain), child demotivation, maladaptive responses (such as overeating) in response to feeling judged, alterations to parent-child interactions (eg, providing food for comfort) to create a self-fulfilling prophecy, and/or that parents are not able or willing to manage their child’s overweight.8,10 

Like Robinson and Sutin,8,10 Parkinson et al’s11 analyses within the Gateshead Millennium Study (a population-based cohort study) did not reveal evidence that parental classification of child overweight at age 7 years was beneficial for children’s BMI change by age 15 years. (See also ref 12.) However, the analyses did not reveal it to be harmful either; neither mothers’ awareness of nor mothers’ concerns about the children’s weight at age 7 years had any impact on the subsequent BMI outcome. This aligns with findings from our own13,15 and others’6,16 trials in which baseline parental concern did not contribute to intervention effectiveness.

Here, we extend Robinson and Sutin’s8,10 analyses to examine new hypotheses within the same LSAC data set and seek to replicate findings across 2 cohorts, thus spanning effectively all of childhood (age range from 2–3 to 14–15 years). We examined multiple biennial waves of repeated measurements of both children’s BMI and parental perceptions of children’s weight statuses, looking to draw out the true pattern of any emerging responsive evolution that might be present. We hypothesized that perception of a child’s overweight is an appropriate (if imperfect) parental response to a child’s high and/or rising BMI. Because trajectories of rising child BMI are typically sustained over time,17,18 with few children resolving to normal weight once already on an upward overweight trajectory,19 we further hypothesized that the perception of overweight after weight gain would in turn precede further weight gain but that this direction of “effect” would be weaker.

Data were drawn from LSAC, which has followed 2 nationally representative cohorts of children with biennial home visits since 2004.20 At wave 1, the birth cohort (n = 5107; 64% response rate) was age 0 to 1 year, and the kindergarten cohort (n = 4983; 59% response rate) was age 4 to 5 years (Fig 1). Although remote areas of Australia were excluded from recruitment, LSAC is considered to be broadly population representative.20 Participants were initially sampled from the Medicare enrolment database, and recruitment involved a postcode-based, 2-stage clustered design, which is described in detail elsewhere.20 Each data collection wave is approved by the Australian Institute of Family Studies Ethics Committee, and parents provided written informed consent at wave 1.

FIGURE 1

Participant retention and analytic sample. a Denotes baseline wave of data collection for the current study.

FIGURE 1

Participant retention and analytic sample. a Denotes baseline wave of data collection for the current study.

Children were included in the current study if they had BMI and parent-reported overweight perception data in at least 2 waves (Fig 1). In total, 4445 children of the birth cohort and 4632 children of the kindergarten cohort were included.

Trained interviewers conducted 90-minute home visits at each wave. Children’s weight was measured in light clothing to 50 g by using glass bathroom scales (code 79985; Salter Australia, Springvale, Victoria, Australia), and children’s height was measured to 0.1 cm by using a portable rigid stadiometer (code IPO955; Invicta, Leicester, United Kingdom).21 The BMI z score was calculated from the BMI (kilograms per meter squared) by using Centers for Disease Control and Prevention growth reference data from 2000.22 

The parent (usually the mother) was asked to select a response to the question, “Which of these best describes your child?” The 4 response categories were dichotomized into parents not perceiving (categories of underweight and normal weight) versus perceiving (categories of somewhat overweight and very overweight) their child as being overweight. Parental perception of overweight was collected at each wave except at age 6 to 7 years in the birth cohort. Children in the birth cohort also lacked BMI data at wave 1 when aged 0 to 1 year.

Covariates included child sex, indigenous background, and whether a main language other than English was spoken at home. A composite continuous indicator of family socioeconomic position (based on parental reports of annual family income, occupation, and education) was provided in the LSAC data set; this was categorized into quintiles. Having special health care needs (which indicated whether children had any chronic developmental, behavioral, emotional, or physical condition needing more care than other children of the same age and/or ongoing medication [yes or no]) was also included as a covariate.

Descriptive statistics were calculated by using Stata 14.0 software (Stata Corp, College Station, TX). Next, we applied a cross-lagged modeling approach in Mplus version 7.2 to examine bidirectional associations in the kindergarten cohort across all 6 2-year age bands from age 4 to 5 to age 14 to 15 years. The model in the kindergarten cohort simultaneously included the following: (1) lagged linear regressions of parents’ perception of child overweight to child BMI z score in the following wave, (2) lagged logistic regressions of child BMI z score to parents’ perception in the following wave, and (3) regressions to account for continuity between BMI z score assessments (linear) and between perception assessments (logistic) across waves. With the inclusion of both directions of the BMI-perception associations, lagged paths within the same age period were accounted for each other and could be directly compared in strength. Wald tests were used to examine whether these opposite directions of effects from 1 wave to another differed significantly from each other. As a confirmatory analysis, a similar model was applied across 4 waves of the parallel birth cohort, followed from age 2 to 3 to age 10 to 11 years.

Models were estimated by using maximum likelihood estimation with robust SEs to account for any potential nonnormality of the data. Survey weights were applied to account for differential nonresponse at wave 1. Missing values in BMI, parental perception, and covariates were accounted for by full information maximum likelihood procedures available in Mplus. This method is used to estimate model parameters and SEs by using the available data and is equivalent to multiple imputations when the number of imputations performed approaches infinity.23 The uncertainty associated with missing data is reflected in wider confidence intervals. Of the 4632 children in the kindergarten cohort, 2772 children (59.8%) had full data on BMI and perception, 1377 children (29.7%) had 1 to 4 missing data points, and 483 children (10.4%) had 5 to 8 missing data points in BMI and/or perception data. Of the 4445 children in the birth cohort, 3146 children (70.8%) had full data on BMI and perception, 275 children (6.2%) had 1 missing data point, 563 children (12.7%) had 2 missing data points, 59 children (1.3%) had 3 missing data points, and 402 children (9.0%) had 4 missing data points in BMI and/or perception data.

The participant flow through the study is shown in Fig 1, and the baseline sample characteristics are shown in Table 1. Both cohorts contained equal proportions of boys and girls, and the average BMI z score was elevated among both cohorts. A small minority of participants were Indigenous Australians. Between 11% and 17% of the sample came from non-English–speaking backgrounds, were in the lowest socioeconomic position quintile, and reported having special health care needs. Table 2 shows the numbers of children with each of the key variables at each wave in each cohort. Altogether, 4632 and 4445 children were included in the cross-lagged models in the kindergarten and birth cohorts, respectively.

TABLE 1

Baseline Characteristics of Analyzed Sample (Children With At Least 2 BMI and 2 Parental Perception of Overweight Measures)

Characteristics (Wave 1)Kindergarten Cohort, n = 4632Birth Cohort, n = 4445
Age, y, mean (SD) 4.7 (0.2) 2.8 (0.2) 
Male sex, % 51.1 51.3 
BMI z score, mean (SD) 0.55 (1.0) 0.52 (1.1) 
Socioeconomic position, %a   
 Most disadvantaged 14.4 13.8 
 Most advantaged 25.5 24.8 
Indigenous status (yes), % 3.4 3.5 
English speaking background (no), % 16.7 14.4 
Special health care needs (yes), % 13.1 11.4 
Characteristics (Wave 1)Kindergarten Cohort, n = 4632Birth Cohort, n = 4445
Age, y, mean (SD) 4.7 (0.2) 2.8 (0.2) 
Male sex, % 51.1 51.3 
BMI z score, mean (SD) 0.55 (1.0) 0.52 (1.1) 
Socioeconomic position, %a   
 Most disadvantaged 14.4 13.8 
 Most advantaged 25.5 24.8 
Indigenous status (yes), % 3.4 3.5 
English speaking background (no), % 16.7 14.4 
Special health care needs (yes), % 13.1 11.4 
a

Middle quintiles not presented for space limitations.

TABLE 2

BMI, BMI z Score, and Parental Perception of Child Overweight/Obesity in the Birth and Kindergarten Cohorts at Each Wave

CharacteristicsAge, y
2–34–56–78–910–1112–1314–15
Kindergarten cohort (n = 4632)        
 With BMI data, n — 4602 4418 4289 4023 3802 3275 
 BMI, mean (SD) — 16.3 (1.7) 16.5 (2.1) 17.6 (2.8) 19.0 (3.9) 20.5 (3.9) 21.9 (4.1) 
 BMI z score, mean (SD) — 0.55 (1.0) 0.38 (1.0) 0.38 (1.0) 0.36 (1.0) 0.35 (1.1) 0.36 (1.2) 
 Perceived as overweight/obese, n — 4621 4446 4323 4113 3849 3365 
 Child perceived as overweight/obese, % — 4.2 5.3 9.7 17.3 15.1 13.3 
Birth cohort (n = 4445)        
 With BMI data, n 4261 4282 — 3978 3563 — — 
 BMI, mean (SD) 16.8 (1.6) 16.3 (1.7) — 17.6 (2.9) 18.8 (3.6) — — 
 BMI z score, mean (SD) 0.52 (1.1) 0.53 (1.1) — 0.35 (1.1) 0.27 (1.2) — — 
 Perceived as overweight/obese, n 4323 4324 — 3990 3653 — — 
 Child perceived as overweight/obese, % 2.6 2.8 — 11.6 15.6 — — 
CharacteristicsAge, y
2–34–56–78–910–1112–1314–15
Kindergarten cohort (n = 4632)        
 With BMI data, n — 4602 4418 4289 4023 3802 3275 
 BMI, mean (SD) — 16.3 (1.7) 16.5 (2.1) 17.6 (2.8) 19.0 (3.9) 20.5 (3.9) 21.9 (4.1) 
 BMI z score, mean (SD) — 0.55 (1.0) 0.38 (1.0) 0.38 (1.0) 0.36 (1.0) 0.35 (1.1) 0.36 (1.2) 
 Perceived as overweight/obese, n — 4621 4446 4323 4113 3849 3365 
 Child perceived as overweight/obese, % — 4.2 5.3 9.7 17.3 15.1 13.3 
Birth cohort (n = 4445)        
 With BMI data, n 4261 4282 — 3978 3563 — — 
 BMI, mean (SD) 16.8 (1.6) 16.3 (1.7) — 17.6 (2.9) 18.8 (3.6) — — 
 BMI z score, mean (SD) 0.52 (1.1) 0.53 (1.1) — 0.35 (1.1) 0.27 (1.2) — — 
 Perceived as overweight/obese, n 4323 4324 — 3990 3653 — — 
 Child perceived as overweight/obese, % 2.6 2.8 — 11.6 15.6 — — 

—, not applicable.

Figure 2 shows the bidirectional wave-on-wave associations between the BMI z score and parental perception of child overweight in the kindergarten and birth cohorts separately. We present unadjusted model results because adjustments for sex, indigenous background, English language spoken at home, socioeconomic position, and having a special health care need made virtually no change to the effect estimates. Wave-on-wave direct predictions were strong and consistent between waves and between cohorts. As expected, the BMI z score at each wave was a powerful predictor of the BMI z score at subsequent waves; for example, in the kindergarten cohort, a 1 unit increase in the BMI z score at age 4 to 5 years predicted a 0.72 higher BMI z score in the next wave at age 6 to 7 years. Parental perception of overweight similarly predicted perception at subsequent waves; for example, for the kindergarten cohort, parental perception of a child as overweight at age 4 to 5 years was associated with 10.1 times higher odds of perceiving the child as overweight in the next wave at age 6 to 7 years.

FIGURE 2

Bidirectional wave-on-wave associations from BMI z score to parental perception of child overweight (odds ratio, OR) and from parental perception to BMI z score (beta, B) in the kindergarten and birth cohorts.

FIGURE 2

Bidirectional wave-on-wave associations from BMI z score to parental perception of child overweight (odds ratio, OR) and from parental perception to BMI z score (beta, B) in the kindergarten and birth cohorts.

All lagged associations were also highly significant (P < .001). Thus, for every unit increase in the BMI z score, the odds of parents perceiving children as overweight in the next wave ranged from 2.85 (birth cohort: age 2–3 years) to 10.4 (kindergarten cohort: age 6–7 years), whereas the perception of child overweight predicted a BMI z score that was ∼0.2 to 0.5 higher in the next wave. As we hypothesized, the obvious bidirectional lagged effects between 2 subsequent waves were much stronger from higher BMI z score to increasing parental perception of overweight than from parental perception of overweight to rising BMI. These findings were confirmed by Wald tests (all P values <.001).

To enhance the effect size comparison with the lagged direction for perception to BMI z score, Fig 3 shows the same findings as Fig 2 but with equivalent estimated β values replacing the odds ratios from the lagged direction for BMI z score to perception. Using these comparable metrics, the lagged effect sizes from BMI z score to parental perception of overweight ranged from ∼3 times larger (eg, birth cohort: age 2–3 years; β = 1.05 vs .39, respectively) to, at most, 12 times larger (kindergarten cohort: age 8–9 years, β = 2.16 vs .17, respectively) than the effect sizes from perception of overweight to BMI z score. When stratified and examined separately, results were similar for children who were normal weight at baseline and those who were overweight at baseline (data not shown).

FIGURE 3

Alternative to Fig 2 showing all lagged associations as B values. OR, odds ratio.

FIGURE 3

Alternative to Fig 2 showing all lagged associations as B values. OR, odds ratio.

There are strikingly large predictive associations from child BMI z score to parental perception of child overweight 2 years later. These far outweigh the consistent but small predictions from parental perception of overweight to higher BMI. Perception therefore seemed mainly to reflect an awareness of an already rising trajectory of BMI. At no point in either cohort did parental perception precede falling BMI. These findings held true for every 2-year age band from age 2 to 3 years to age 14 to 15 years and were highly replicable across 2 distinct cohorts of slightly different ages harmonized on all measures and covariates. Thus, although clinician efforts to make parents more aware of overweight may not be harmful, it also seems unlikely that it will help to improve children’s BMI status.

Within 2 population-representative cohorts of children, we obtained significant, although modest, relationships from greater parental perception of overweight to higher adolescent BMI. Not surprisingly (because we draw on 1 of the same cohorts), our results from greater parental perception to higher later BMI z score are consistent in direction and size with Robinson and Sutin’s8 work in the older of the 2 LSAC cohorts. They also resemble Gerards et al’s9 findings that accurate weight status perception at age 5 was a predictor of higher future BMI in 102 children with overweight followed to age 9. Parkinson et al11 and Jeffery et al12 found no clear relationship of BMI at age 15 to 16 years with parental perception of overweight at age 6 to 7 years. Small effects (commensurate with ours) could have been obscured by their smaller samples (n = 228 and n = 237, respectively) and the long gap (likely spanning much of puberty for most children) between their measurements. Importantly, neither we nor the authors of any of these 4 studies found evidence that greater parental perception of overweight predicted better (lower) BMI at any subsequent time point.

We go further than these existing studies in also examining the reverse association (high BMI to parental perception of overweight), which dominated our cross-lagged models in the strength and size of effects across multiple 2-year time windows. Authors of recent studies (including one in which this same data set was used but without the benefit of a bidirectional analysis) have concluded that parental perception of overweight could be harmful in that it precedes BMI gain.8,9 From our more nuanced findings and those of Parkinson et al11 and Jeffery et al,12 we think it more likely that parental perception of overweight occurs mostly in recognition that a child’s weight trajectory is already rising.

Our ability to look within multiple short, sequential time windows at how each of the 2 constructs affects the other across 6 waves of data collection spanning 10 years in 2 cohorts (enabling conclusions to be drawn spanning from toddlerhood to adolescence) was this study’s main strength. Beyond this, we have confidence in our findings because they are large, robust, and consistent across waves and across 2 separate cohorts. Generalizability within Australia is supported by the nationally representative sampling frame and large sample size; generalizability elsewhere cannot be known but is likely high given that the epidemiology of childhood obesity and patterns of parental concern in Australia largely mirror those of other wealthy countries. Despite some underrepresentation of families from non-English–speaking backgrounds and greater losses to follow-up of more disadvantaged families, these groups were nonetheless present, and any possible imbalances are at least partly counteracted by the application of survey weights. Although we cannot predict how much the observed associations might differ in a disadvantaged sample, the strength, size, and consistency of the BMI-to-perception effect and its lack of attenuation on adjusting for socioeconomic covariates suggest that the broad conclusions would hold. Height and weight were objectively measured, and BMI z scores were calculated by using internationally accepted conventions. The parental perception of overweight measure was the same or similar to that used in numerous other reports1 and, in our experience, reflects questions that clinicians frequently ask of parents in usual practice.

As noted in Results, stratified analyses did not support a difference in conclusions for children who were initially normal weight versus those who were initially overweight or obese. To check the possibility that results would differ for the ∼5% of children who were initially obese versus not obese in each cohort, we undertook further exploratory stratified analyses. This again gave similar findings to the whole-group analyses, that is, the pattern was stronger from BMI to perception than the other way around (from perception to BMI). However, these are not firm conclusions because of low numbers in some cells.

Our interpretation is that parents (rightly) respond to high and rising BMI by increasingly perceiving their child to be overweight. The weaker predictive association from parental perception to higher BMI may simply reflect an intuitive recognition that the child is already on a rising weight trajectory. Humans are visually primed to detect change,24 and latent class modeling on this and other cohorts has confirmed that these high-rising trajectories do exist and last over time.17,18 

We doubt that parents’ perceptions, however aligned with actual BMI, are harmful at the population level. For example, the motivational interviewing and treatment trial,25 in which 4- to 8-year-old children screening positive for overweight and/or obesity were targeted, reported a modestly increased uptake from parents who believed their child to be overweight. However, the small resulting reduction in the BMI z score (on average 0.12 after 2 years) was not clearly related to baseline parental identification or concern.26 Similarly, in a pre-post study of 4- to 12-year-old American children enrolled in Medicaid, Perrin et al4 systematically increased parental recognition of the BMI status of their children with overweight and/or obesity from 57% to 74% over a 3-month period, but this did not translate to an improvement in BMI. Other trials in which parental recognition and education are incorporated at an intensity that can be widely implemented have had largely null impacts over time.6,16 

Collectively, this evidence does not suggest that clinicians or public health initiatives should be focused on making parents aware of their child’s overweight because this does not predict a reduction in their future overweight and obesity.9,16 This should perhaps come as a relief given that parental perception of the child being overweight has become less, not more, accurate over the last decade1 despite much greater parental awareness of childhood obesity as a major threat to children’s health in general.27 To end childhood obesity, the overarching goals are to prevent entry to an upward BMI trajectory and to effectively intervene and move those who are overweight back toward a healthier trajectory. Perhaps freeing researchers and clinicians from this construct could pave the way for faster discovery and more rapid advances. For example, “nudge” interventions do not require recognition of a problem to effect lasting change; their application both at the population and individual level for childhood obesity could be 1 avenue to explore with some urgency.

Strikingly large predictive associations from higher child BMI z scores to greater parental perceptions of child overweight far outweighed the reverse associations, from perception of overweight to higher BMI. Rather than driving BMI trajectories upward (or downward), the evolution of parents’ perceptions of overweight seemed responsive and appropriate to children’s wave-on-wave BMI changes. Taken in concert with other recent literature, clinician and public health efforts to make parents aware of the actual BMI status of their child seem unlikely to be harmful, but equally seem unlikely to reduce childhood obesity.

     
  • LSAC

    Longitudinal Study of Australian Children

Dr Wake led the conceptual planning of the study, drafted the initial manuscript, reviewed and revised the final manuscript, and critically reviewed the manuscript for interpretation and intellectual content; Dr Kerr critically reviewed and revised the initial and final manuscripts and critically reviewed the manuscript for interpretation and intellectual content; Dr Jansen completed all data analyses, revised the final manuscript, and critically reviewed the manuscript for interpretation and intellectual content; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

FUNDING: In this article, unit record data from Growing Up in Australia, the Longitudinal Study of Australian Children, are used. The study is conducted in partnership with the Australian Department of Social Services, the Australian Institute of Family Studies, and the Australian Bureau of Statistics. The findings and views reported are those of the authors and should not be attributed to the Australian Department of Social Services, the Australian Institute of Family Studies, or the Australian Bureau of Statistics. Dr Wake was supported by Australian National Health and Medical Research Council Senior Research Fellowship 1046518 and by Cure Kids New Zealand. Dr Jansen was supported by a grant from the Dutch Diabetes Foundation (grant 2013.81.1664). Research at the Murdoch Children’s Research Institute is supported by the Victorian Government’s Operational Infrastructure Support Program. The researchers were independent of the funders.

1
Tompkins
CL
,
Seablom
M
,
Brock
DW
.
Parental perception of child’s body weight: a systematic review.
J Child Fam Stud
.
2015
;
24
(
5
):
1384
1391
2
Lundahl
A
,
Kidwell
KM
,
Nelson
TD
.
Parental underestimates of child weight: a meta-analysis.
Pediatrics
.
2014
;
133
(
3
). Available at: www.pediatrics.org/cgi/content/full/133/3/e689
[PubMed]
3
Prochaska
JO
,
DiClemente
CC
,
Norcross
JC
.
In search of how people change. Applications to addictive behaviors.
Am Psychol
.
1992
;
47
(
9
):
1102
1114
[PubMed]
4
Perrin
EM
,
Jacobson Vann
JC
,
Benjamin
JT
,
Skinner
AC
,
Wegner
S
,
Ammerman
AS
.
Use of a pediatrician toolkit to address parental perception of children’s weight status, nutrition, and activity behaviors.
Acad Pediatr
.
2010
;
10
(
4
):
274
281
[PubMed]
5
Rhee
KE
,
De Lago
CW
,
Arscott-Mills
T
,
Mehta
SD
,
Davis
RK
.
Factors associated with parental readiness to make changes for overweight children.
Pediatrics
.
2005
;
116
(
1
). Available at: www.pediatrics.org/cgi/content/full/116/1/e94
[PubMed]
6
Sim
LA
,
Lebow
J
,
Wang
Z
,
Koball
A
,
Murad
MH
.
Brief primary care obesity interventions: a meta-analysis.
Pediatrics
.
2016
;
138
(
4
):
e20160149
[PubMed]
7
Chomitz
VR
,
Collins
J
,
Kim
J
,
Kramer
E
,
McGowan
R
.
Promoting healthy weight among elementary school children via a health report card approach.
Arch Pediatr Adolesc Med
.
2003
;
157
(
8
):
765
772
[PubMed]
8
Robinson
E
,
Sutin
AR
.
Parental perception of weight status and weight gain across childhood.
Pediatrics
.
2016
;
137
(
5
):
e20153957
[PubMed]
9
Gerards
SM
,
Gubbels
JS
,
Dagnelie
PC
, et al
.
Parental perception of child’s weight status and subsequent BMIz change: the KOALA birth cohort study.
BMC Public Health
.
2014
;
14
:
291
[PubMed]
10
Robinson
E
,
Sutin
AR
.
Parents’ perceptions of their children as overweight and children’s weight concerns and weight gain.
Psychol Sci
.
2017
;
28
(
3
):
320
329
[PubMed]
11
Parkinson
KN
,
Reilly
JJ
,
Basterfield
L
, et al
.
Mothers’ perceptions of child weight status and the subsequent weight gain of their children: a population-based longitudinal study.
Int J Obes (Lond)
.
2017
;
41
(
5
):
801
806
[PubMed]
12
Jeffery
AN
,
Metcalf
BS
,
Hosking
J
,
Mostazir
MB
,
Voss
LD
,
Wilkin
TJ
.
Awareness of body weight by mothers and their children: repeated measures in a single cohort (EarlyBird 64).
Child Care Health Dev
.
2015
;
41
(
3
):
434
442
[PubMed]
13
Wake
M
,
Baur
LA
,
Gerner
B
, et al
.
Outcomes and costs of primary care surveillance and intervention for overweight or obese children: the LEAP 2 randomised controlled trial.
BMJ
.
2009
;
339
:
b3308
[PubMed]
14
Wake
M
,
Lycett
K
,
Clifford
SA
, et al
.
Shared care obesity management in 3-10 year old children: 12 month outcomes of HopSCOTCH randomised trial.
BMJ
.
2013
;
346
:
f3092
[PubMed]
15
McCallum
Z
,
Wake
M
,
Gerner
B
, et al
.
Outcome data from the LEAP (Live, Eat and Play) trial: a randomized controlled trial of a primary care intervention for childhood overweight/mild obesity.
Int J Obes (Lond)
.
2007
;
31
(
4
):
630
636
[PubMed]
16
Neumark-Sztainer
D
,
Wall
M
,
Story
M
,
van den Berg
P
.
Accurate parental classification of overweight adolescents’ weight status: does it matter?
Pediatrics
.
2008
;
121
(
6
). Available at: www.pediatrics.org/cgi/content/full/121/6/e1495
[PubMed]
17
Jansen
PW
,
Mensah
FK
,
Nicholson
JM
,
Wake
M
.
Family and neighbourhood socioeconomic inequalities in childhood trajectories of BMI and overweight: longitudinal study of Australian children.
PLoS One
.
2013
;
8
(
7
):
e69676
[PubMed]
18
Pryor
LE
,
Tremblay
RE
,
Boivin
M
, et al
.
Developmental trajectories of body mass index in early childhood and their risk factors: an 8-year longitudinal study.
Arch Pediatr Adolesc Med
.
2011
;
165
(
10
):
906
912
[PubMed]
19
Hesketh
K
,
Wake
M
,
Waters
E
,
Carlin
J
,
Crawford
D
.
Stability of body mass index in Australian children: a prospective cohort study across the middle childhood years.
Public Health Nutr
.
2004
;
7
(
2
):
303
309
[PubMed]
20
Soloff
C
,
Lawrence
D
,
Johnstone
R
. Sample design. 2005. Available at: www.growingupinaustralia.gov.au/pubs/technical/tp1.pdf. Accessed March 1, 2017
21
Australian Institute of Family Studies
. Data user guide: Growing up in Australia - The Longitudinal Study of Australian Children. Available at: https://growingupinaustralia.gov.au/sites/default/files/userguide.pdf. Accessed March 1, 2017
22
Kuczmarski
RJ
,
Ogden
CL
,
Guo
SS
, et al
.
2000 CDC growth charts for the United States: methods and development.
Vital Health Stat 11
.
2002
;(
246
):
1
190
[PubMed]
23
Graham
JW
,
Olchowski
AE
,
Gilreath
TD
.
How many imputations are really needed? Some practical clarifications of multiple imputation theory.
Prev Sci
.
2007
;
8
(
3
):
206
213
[PubMed]
24
Öhman
A
.
Has evolution primed humans to “beware the beast”?
Proc Natl Acad Sci U S A
.
2007
;
104
(
42
):
16396
16397
[PubMed]
25
Taylor
RW
,
Williams
SM
,
Dawson
AM
,
Taylor
BJ
,
Meredith-Jones
K
,
Brown
D
.
What factors influence uptake into family-based obesity treatment after weight screening?
J Pediatr
.
2013
;
163
(
6
):
1657
1662.e1
[PubMed]
26
Taylor
RW
,
Cox
A
,
Knight
L
, et al
.
A tailored family-based obesity intervention: a randomized trial.
Pediatrics
.
2015
;
136
(
2
):
281
289
[PubMed]
27
The Royal Children’s Hospital Melbourne
. Top ten child health problems: what the public thinks. 2015. Available at: https://www.childhealthpoll.org.au/polls/top-10-child-health-problems:/. Accessed May 15, 2017

Competing Interests

POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.

FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.