Paper 1 PDF

Clinical Child and Family Psychology Review (2019) 22:290–315
https://doi.org/10.1007/s10567-019-00279-z
The Mediating Effect of Self-Regulation in the Association Between

Poverty and Child Weight: A Systematic Review
Katherine A. Hails1 · Yiyao Zhou1 · Daniel S. Shaw1
Published online: 6 February 2019

© Springer Science+Business Media, LLC, part of Springer Nature 2019
Abstract
Deficits in self-regulation (SR) have been proposed as a potential contributor to child overweight/obesity, a public health
concern that disproportionately affects children living in poverty. Although poverty is known to influence SR, SR has not
been considered as a potential mechanism in the association between poverty and child obesity. The aim of the current
paper was to systematically review the current literature to determine whether SR is a viable mechanism in the relationship
between child exposure to poverty and later risk of overweight/obesity. We systematically review and summarize literature
in three related areas with the aim of generating a developmentally informed model that accounts for the consistent associa-
tion between poverty and child weight, specifically how: (1) poverty relates to child weight, (2) poverty relates to child SR,
and (3) SR is associated with weight. To quantify the strength of associations for each pathway, effect sizes were collected
and aggregated. Findings from the studies included suggest small but potentially meaningful associations between poverty
and child SR and between SR and child weight. The conceptualization and measurement of SR, however, varied across
literature studies and made it difficult to determine whether SR can feasibly connect poverty to child obesity. Although SR
may be a promising potential target for obesity intervention for low-income children, additional research on how SR affects
risk of obesity is crucial, especially based on the lack of success of the limited number of SR-promoting interventions for
improving children’s weight outcomes.
Keywords Poverty · Obesity · BMI · Self-regulation · Food insecurity
Introduction and Pollack 2003). Unfortunately, research on child obesity

prevention and treatment has only recently started to include
Pediatric obesity is a public health concern that affects chil- children from low-income families (Kimbro et al. 2007;
dren’s quality of life and their current and future health out- Shrewsbury and Wardle 2008), who tend to be at greater
comes. Seventeen percent of U.S. children ages 2–19 are risk of weight problems. Obesity prevention and treatment
estimated to be obese, and an additional 15% overweight may be especially challenging for children living in poverty
(Ogden et al. 2012), with these numbers reflecting a sub- because of increased exposure to environmental stress and
stantial increase over the past several decades (Ogden et al. chaotic living conditions (Evans and English 2002).
2016). Although the medical problems associated with There are many adverse psychosocial outcomes associ-
overweight may not become apparent until adulthood (e.g., ated with exposure to poverty in childhood. Family income
cardiovascular disease, metabolic syndrome), the risk of and other markers of financial hardship have been found to
adult obesity is known to be greater for children who are have detrimental effects on children’s academic achieve-
overweight in childhood (Simmonds et al. 2016). Further- ment (Votruba-Drzal 2006), behavior (Eamon 2002), and
more, overweight and obese children report lower quality mental (Dearing et al. 2006) and physical health (Law-
of life (Friedlander et al. 2003) and social isolation (Strauss lor et al. 2006). Poverty has also been established as a
risk factor for child overweight and obesity (Alaimo et al.
* Daniel S. Shaw 2001; Cameron et al. 2015). This association at first glance
[email protected] may appear paradoxical, as historically, poverty has been
associated with inability to afford food, leading to mal-
1
University of Pittsburgh, 210 South Bouquet Street, nutrition and underweight. Several mechanisms for the
Pittsburgh, PA, USA
13
Vol:.(1234567890)
Clinical Child and Family Psychology Review (2019) 22:290–315 291
association between poverty and obesity have been pro- be particularly critical for children living in poverty because
posed, with most focused on the premise that children liv- of their greater exposure to contexts that promote obesity.
ing in poverty tend to eat more high calorie, low-nutrient A critical attribute for children’s social and academic
foods and have less access to opportunities for physical success that has been linked to poverty is self-regulation.
activity (Drewnowski and Specter 2004; Sallis and Glanz Self-regulation (SR) encompasses one’s ability to modulate
2006). behavior and emotions in response to internal and external
Although both family-based (Epstein et al. 1990) and demands and is often conceptualized using a dual process
community interventions (Bleich et al. 2013) have been model consisting of “hot,” reactive processes (e.g., impul-
developed to address and prevent child overweight, few have sivity) and a “cool,” reasoned, cognitive system (Isasi and
been found to have lasting effects on child weight (Ebbeling Wills 2011; Metcalfe and Mischel 1999). Children living
et al. 2002). Most intervention efforts fall broadly in the in poverty have been postulated and found to show lower
category of “lifestyle interventions,” with a relatively nar- levels of SR because of their greater exposure to witness-
row focus on decreasing calorie intake and increasing physi- ing and experiencing violence and trauma, more frequent
cal activity (Brown and Summerbell 2009; McCallum et al. changes in residence, and both exposure to and genetic risk
2007; Nemet et al. 2005); prevention efforts tend to focus on of mental health problems among family members (Evans
similar objectives (Economos et al. 2007; Hollar et al. 2010). and English 2002; Lengua et al. 2007; Raver 2004). As SR is
However, targeting these areas has proven surprisingly and vital for helping children manage their emotions and behav-
frustratingly resistant to long-term change. Theoretically, ior, it is not surprising that deficits in SR have been linked
these behaviors may be even more challenging to modify to a variety of adverse outcomes for children and adults,
in low-income children because of lack of access to nutri- including peer rejection in middle childhood (Trentacosta
tionally sound food and limited opportunities for physical and Shaw 2009), antisocial behavior in adolescence (Sitnick
activity. et al. 2017), and importantly, physical health in adulthood
Historically, child obesity treatment and prevention (Moffitt et al. 2011). As maintenance of healthy body weight
efforts have not taken into account differences in children’s involves appropriate SR of energy intake, it follows that SR
developmental status. “Sensitive periods” during which has been found to be associated with child body mass index
excessive child weight gain is particularly strongly linked to (BMI), overweight, and obesity in both cross-sectional stud-
later obesity have been identified (i.e., during middle child- ies (Miller et al. 2016) and longitudinal ones (Francis and
hood in a period known as “adiposity rebound,” and during Susman 2009; Graziano et al. 2010; Seeyave et al. 2009).
adolescence when they go through puberty (Dietz 1994)). As child weight and poverty have been associated with
However, knowledge about development has infrequently multiple components of SR, SR may be especially impor-
been incorporated into designing targeted treatment and pre- tant in explaining how poverty affects child weight. In addi-
vention efforts for children at specific development stages, tion, as SR has been found to be malleable using school-,
nor has it been systematically evaluated as a moderator of family-, and individual-based approaches during multiple
treatment efficacy in reviews or meta-analyses (Ho et al. developmental periods (Johnson 2000; Raver et al. 2011;
2012). In addition, few researchers have evaluated whether Riggs et al. 2006; Shelleby et al. 2012), it represents a viable
the effect of poverty on child weight varies depending on the target for prevention and intervention. However, almost no
timing and duration of poverty exposure at different devel- studies have investigated the potential mediating role of SR
opmental stages of childhood. in associations between poverty and child weight.
Because of the lack of success in reducing children’s The current review summarizes the literature in three
weight by modifying their energy balance, it could prove related areas with the aim of generating a developmentally
valuable to investigate potential alternative mechanisms informed model that accounts for the consistent association
leading to child obesity as targets of prevention and/or between poverty and child weight, specifically how: (1) pov-
intervention, targets that might be more malleable than diet erty relates to child weight, (2) poverty relates to child SR,
or physical activity. Although addressing energy balance in and (3) SR is associated with weight. Although there is a
children is the ultimate goal to promote and sustain healthy relatively rich literature documenting each of these direct
weight maintenance, it has become clear that addressing paths, extremely few empirical papers have attempted to
childhood obesity by focusing narrowly and primarily on integrate these studies across literatures to investigate the
children’s calorie intake and physical activity has shown lim- possibility that SR might mediate associations between pov-
ited success. Identifying broader behavioral styles that may erty and child weight. Further, a developmental approach
underlie children’s tendencies to overeat and/or be physically to this literature can determine whether there are sensitive
inactive and exploring how such behavioral styles may exac- developmental periods during which child weight and SR are
erbate continued weight gain could prove helpful in improv- particularly vulnerable to the effects of poverty. In addition
ing obesity prevention efforts in particular. Such issues may to summarizing and consolidating the evidence from studies
13

292 Clinical Child and Family Psychology Review (2019) 22:290–315
across two separate literatures (i.e., obesity literature and reduced food intake” (i.e., with or without hunger) (Cole-
SR literature), the goal of this paper was also to determine man-Jensen et al. 2014). FI is most commonly measured
whether SR could be a viable target for obesity prevention using an 18-item scale developed by the USDA (Hamilton
for children living in poverty. and Cook 1997).
This review includes only longitudinal studies, which, When income or an income-to-needs ratio was used
in the absence of an experimental design, are necessary for as the indicator of poverty, the word “income” is used to
establishing a case for the presence of causal mechanisms encompass both. The term “SES” is used when the study
among variables of interest. Without a longitudinal design in employed a measure that included not only income but also
which baseline levels of the variable of interest are included either parental education, occupational status, or both. The
in the model, the possibility of other variables driving the term “poverty” is used to refer broadly to all measures of
association, or that variables simply co-occur without a poverty or poverty-related stressors included in this review,
causal relationship, is difficult to rule out. including income, SES, cumulative risk, and FI.
Self‑Regulation
Methods
The definition of SR and its various components varies
Operational Definitions of Variables greatly and is frequently inconsistent across studies. Con-
ceptually, the construct of SR is quite broad and is thus fre-
Poverty/Financial Hardship quently deconstructed into distinct but often closely related
components. However, these components (e.g., effortful
Although poverty can be defined and operationalized in control, inhibitory control, emotion regulation) may be dif-
many ways, the current review focused specifically on stud- ficult to differentiate, in part because there are no tasks that
ies that define poverty based on income and/or financial “cleanly” measure one without involving the others. Fur-
strain. Studies often use measures of poverty that aggregate thermore, the language used to label and categorize SR and
multiple indicators of socioeconomic status (SES), includ- its components is quite variable, with different terms being
ing income but also variables such as parental education used to describe the same processes, and vice versa, and
and occupation. This review included studies that used such different literatures tending to adopt divergent language for
composites as long as income or another proxy for finan- describing SR.
cial strain was included. Although parental education and In spite of these issues, overarching themes to character-
occupation may play some role in influencing children’s ize SR have emerged across disparate literatures. For one,
weight (Lamerz et al. 2005), there is a stronger rationale for it is common for researchers to conceptualize SR processes
focusing on income specifically, as it is more malleable and using some kind of a dual process model. A particularly
its measurement more straightforward than other aspects of common and well-accepted model conceptualizes SR as
SES. consisting of two systems referred to as “hot” and “cool”
This review also includes studies in which income/finan- (Metcalfe and Mischel 1999). Hot SR is thought to be
cial strain was incorporated into a cumulative risk index that engaged in emotionally arousing situations, including tasks
included other risks and stressors. The use of cumulative risk with “appetitive demands,” whereas cool SR is deployed
indices has strong theoretical and empirical support, with primarily in emotionally neutral contexts and involves more
the rationale being that although the effect of any single risk conscious, cognitive attentional processes and planful reac-
may be small, multiple risks in combination are particularly tions (Eisenberg and Zhou 2016). Cool SR overlaps with the
detrimental and more damaging than their individual addi- construct of executive functioning as it is typically concep-
tive effects (Forehand et al. 1998). Children living in poverty tualized (Willoughby et al. 2011).
are much more likely to be exposed to multiple, interrelated In the current review, regardless of how authors referred
risks, such as exposure to violence, substandard housing, to constructs and measures in their papers, behavioral
marital conflict, and maternal psychopathology (Evans and measures were categorized as “hot” or “cool” based on the
Kim 2007; Rutter et al. 1997). author’s evaluation of the SR domain each task intended
Lastly, food insecurity (FI) was also considered a measure to measure, with measures that involved regulation or inhi-
of poverty/financial strain, as a family’s FI status indicates bition with an emotional or appetitive component catego-
the presence or absence of an important resource (food) rized as “hot,” and emotionally neutral tasks as “cool.” One
that is closely tied to financial hardship. FI has been defined commonly used measure of hot SR is delay of gratification
by the United States Department of Agriculture (USDA) (DoG). In a DoG task, children are presented with some
as “reduced quality, variety, or desirability of diet,” with desired object or food (for example, a cookie) and are told
or without the presence of “disrupted eating patterns and that they can either eat one cookie now, or have two cookies
13
later. During the waiting period, children are usually left been observed in children as young as 5 years old (Carper
alone with the desired object/food with no other opportuni- et al. 2000).
ties for entertainment or distraction and they can alert the The aspects of SR featured in this review had to be via-
experimenter at any time to indicate that they are unable to ble candidates for mediating the association between pov-
wait any longer. Refer to Table 1 for a summary of how addi- erty and weight. Although there are alternative models for
tional common observed measures of SR were categorized understanding SR, the hot/cool distinction was selected to
as “hot” or “cool.” categorize SR measures because the SR of healthy weight
A variety of methodologies for the measurement of child maintenance likely involves both emotional and cognitive
SR is included in the current review, such as observed meas- processes.
ures of child SR and parent or teacher reports on question-
naires. Nearly all questionnaire measures assessing child SR Child Weight
capture aspects of both hot and cool domains. For exam-
ple, items from the inhibitory control subscale of the Child Two ways of operationalizing weight were used in the stud-
Behavior Questionnaire (Rothbart et al. 2001) measure ies included in this review. The first is using BMI, a continu-
both “hot” emotional, impulsive behaviors such as “has an ous variable, as the outcome of interest, and the second is
easy time waiting to open a present,” and “cool,” measured using weight status, a categorical outcome variable such as
behaviors or processes, as in “likes to plan carefully before the presence or absence of overweight and/or obesity. The
doing something.” Therefore, in the current review, ques- use of BMI has the advantage of allowing for trajectories of
tionnaire measures are not categorized as “hot” or “cool.” weight gain over time to be analyzed with respect to income
In her conceptualization of SR, Calkins (2007) adds a or income changes, while weight status offers the ability to
physiological component in addition to the systems she differentiate clinically significant from normal child weight.
refers to as “emotional” and “attentional” (which map onto There is some evidence to suggest that weight status
hot and cool SR, respectively). She highlights the impor- measured after approximately age two is relatively stable
tance of heart rate variability (HRV) as a measure of para- throughout childhood. Findings from one study suggest that
sympathetic nervous system activation, which is crucial in children categorized as normal weight or obese at as early
modulating emotion regulation and reactivity (Calkins and as age 3 were likely to remain the same weight status at
Dedmon 2000; Porges 2007). Thus, measures of HRV such age 15 (Tran et al. 2016). However, some research indicates
as respiratory sinus arrhythmia (RSA) are also considered that early childhood BMI is a poor predictor of later child-
a marker of SR in the current review and categorized as hood weight when it is measured prior to adiposity rebound,
“physiological.” Physiological SR is subsumed under the when normative increases in BMI start to occur in all chil-
category of hot SR, as it is thought to be more relevant in sit- dren, usually by the age of 6 (Rolland-Cachera et al. 2006).
uations that require the regulation of emotional or behavioral Therefore, it is important to consider the age that weight was
arousal. measured before drawing conclusions about the effects (or
Studies that used both observed and questionnaire meth- lack thereof) of environmental (e.g., exposure to poverty) or
ods that specifically assessed eating or appetite components child characteristics (e.g., SR) on child weight.
of SR (e.g., emotional overeating) were also included and
considered as a category separate from hot and cool SR, Age Periods Covered
as SR of appetite likely contains aspects of both hot and
cool SR. For example, some children may overeat as a way This review focuses on longitudinal studies in which the
to cope with and/or regulate negative emotions (Goossens final measurement of the independent variable (poverty or
et al. 2009), which would capture aspects of hot SR. Alterna- SR) was prior to age 10, as exposure to poverty during early
tively, features of cool SR may be employed when children and middle childhood was of primary interest because of the
consciously restrain themselves from eating large amounts possible implications for obesity prevention. Additionally,
of palatable but unhealthy foods, an eating behavior that has the review excluded studies in which the last measurement
Table 1 Examples of observed self-regulation measures categorized as hot versus cool

Hot Cool
Delay of gratification (food/non-food) (e.g., Seeyave et al., 2009) Stroop and stroop-like tasks (e.g., Lengua et al., 2007)
Gift delay (e.g., Thompson et al., 2013) Motor control tasks (e.g., Balance Beam task) (e.g., Bassett et al., 2012)
High chair frustration task (Graziano et al., 2010) Opposite of commands task (e.g., Head-to-Toes) (e.g., Wanless et al., 2011)
Continuous performance test (e.g., Bub et al., 2016)
13

of the dependent variable (weight or SR) was after age 18, duplicates were removed. Each article was reviewed to
as the scope of the review was limited to childhood and ado- determine its appropriateness for inclusion in the system-
lescence. Studies that measured weight prior to age 2 were atic review. Inclusion criteria were as follows: (a) pub-
also excluded, as some research suggests that this early in lished in a peer-reviewed journal, (b) published or availa-
childhood, child overweight/obesity is a poor predictor of ble in English, (c) normative sample (i.e., excluded studies
later excess weight (Whitaker et al. 1997). that tested associations with specific populations like very
To gain an understanding of which developmental stages low birth weight infants or children prenatally exposed
have been most studied with regard to the pathways of inter- to substances, (d) longitudinal, with at least 6 months
est and whether there were differences in the strength of between assessments (several studies used assessments
associations depending on stage, age periods were catego- at the beginning and end of school years). Studies were
rized as follows: Early Childhood (0–5), Middle Childhood excluded if they were part of an intervention study (e.g.,
(6–11), and Adolescence (12–18). an early childhood intervention that tested SR as an out-
come and included family income as a covariate), unless
Inclusion of Studies it was a weight loss intervention that specifically targeted
SR. There were several other exclusion criteria that were
Articles for the current review were identified through unique to one of the pathways; these are described in the
searches using the databases PsycINFO and PubMed. Sepa- section specific to that pathway.
rate searches were conducted for each of the three pathways. For the two pathways in which poverty was the inde-
(1) SES search terms (appearing in title and/or abstract) were pendent variable, in some cases the poverty variable
as follows: socioeconomic, poverty, income, cumulative was measured when mothers were pregnant with the tar-
risk, adversity, and food insecurity. (2) SR keywords, also get child—these studies were included in the review. In
in the title or abstract, consisted of the following: regulat* addition, it was frequently the case that poverty (usually
(the asterisk denotes that a word stem was searched), execu- income) was operationalized using an average of family
tive function, inhibitory control, executive control, effortful income across multiple years of assessment. For example,
control, delay of gratification, eating in the absence of hun- one study averaged income over 3 years of assessment
ger, overeat*, appetit*, impulsiv*, and attent*. (3) Weight (when children were 5, 7, and 9 years old), with the out-
search terms appearing in the title or abstract were: body come for BMI at age 9 (Tiberio et al. 2014). These studies
mass index, BMI, overweight, obes*, weight. For all, the were included as long as there were at least two other
term child*, boy, or girl had to appear in the abstract. assessment points included in the SES variable in addi-
As an example, for the poverty → weight pathway, Psy- tion to the concurrent assessment (i.e., the age 9 assess-
cINFO was searched first using the terms “socioeconomic” ment in this example). However, studies were excluded if
and “body mass index.” The search was repeated until every they tested trajectories of family income or chronicity of
combination of the “poverty” and “weight” search terms poverty over a period of more than 5 years, as such analy-
had been entered. The citation for every article that resulted ses are not directly comparable to those testing poverty
from each search was downloaded to a file for review. This exposure at a single assessment point. Similarly, studies
search was then repeated in an identical fashion using the that tested growth in child weight or children’s weight tra-
PubMed database. This procedure was then repeated for the jectories over time were also excluded.
poverty → SR (using the “poverty” and “SR” keywords) and The review was primarily concerned with the relation-
SR → weight (using the “SR” and “weight” search terms) ship between poverty and obesity in the USA, although
searches. this phenomenon has been well documented in many other
Next, some studies were identified using reference lists high-income countries. Therefore, studies with non-U.S.
of the initial articles, and also searching for papers that samples were included if the World Bank defined the coun-
cited them. Only longitudinal studies, where the depend- try as a “high-income economy” as the USA is catego-
ent variable was measured after the independent variable, rized. This classification is based on gross national income
were included in the formal review. Experimental studies per capita; those with a GNI of $12,056 per capita are
where poverty or SR was manipulated as part of an interven- considered high income. As the pattern of the relation-
tion were also included. Although many child weight loss ship between poverty and weight for individuals living in
interventions involve some aspect of SR (Braet et al. 2004; low-income countries may differ compared to high-income
Epstein et al. 1981), only studies that specifically measured countries, and the relationship for individuals in middle-
SR pre- and post-intervention were included in the review income countries appears to be inconsistent (Dinsa et al.
on the SR to weight pathway. 2012), the attention of the current review was focused on
Upon completion of the searches in PsycINFO and high-income countries.
PubMed, the results for each pathway were compiled and
13
Effect Size Aggregation using BMI at age 9 (middle childhood), and Assari and col-
leagues (2018) testing effects on BMI at age 15 (adoles-
Articles that met criteria for the review were searched to cence). Therefore, the effect sizes for all three papers were
locate the unadjusted relationship between the two variables retained for the analysis by developmental stage, but not
of interest (e.g., bivariate, zero-order correlation, unadjusted for a separate “total effect size” that was calculated without
odds ratio). In addition to effect size, the following informa- consideration of age of outcome.
tion was also extracted from each paper that met inclusion Effect sizes were transformed such that all were consist-
criteria: age that the independent variable was measured, ently representing the relationship between poverty and
age(s) that the dependent variable was measured, how pov- weight. For example, if the variable for “poverty” used in
erty was operationalized, type of SR (if applicable), and the study was a continuous measure of family income, opera-
information about the sample, including the country where tionalized as dollars per year, the sign ( ± ) was reversed so
the study took place and whether it was described in the that it became an index of poverty rather than income (i.e., a
paper as predominantly low income. positive r indicates that high poverty/low income is associ-
For the poverty → weight pathway, an overall total aggre- ated with higher BMI). Similarly, for the poverty → SR and
gated effect size was calculated for each pathway, along with SR → weight pathways, the way in which SR was opera-
three different effect sizes stratified by developmental stage tionalized was considered prior to aggregating effect sizes.
(i.e., early childhood, middle childhood, adolescence), as When the variable of interest was operationalized in a study
one goal of the study was to determine whether the rela- as a lack of SR (e.g., impulsivity, eating more in the absence
tionship between poverty and weight differed depending on of hunger), the sign was reversed to ensure consistency
when weight was measured. Similarly, for the poverty → SR across studies.
and SR → weight pathways, three different effect sizes were
calculated based on which type of SR was tested (hot, cool,
or questionnaire). For the SR → weight pathways, it would Review of Literature
have been ideal to assess age of outcome in addition to SR
type, but the cell sizes would have been too small to detect Longitudinal Research Linking Poverty and Food
meaningful differences in effect sizes. In addition, a “total Insecurity to Childhood Weight
effect size” was calculated without consideration of devel-
opmental stage or SR type. Included in the review were studies in which the primary
If an unadjusted measure of effect size was not included research aim was to assess the relationship between poverty
in the manuscript, the first and/or corresponding author of and weight, and also studies in which the primary independ-
the paper was contacted by email to request the necessary ent variable of interest was not poverty or SES, but some
information. Only studies with an effect size were retained other variable (e.g., breastfeeding, parental smoking), with
for inclusion in the systematic review. After any one sample SES included as a covariate in analyses.
was represented in the effect size list for a given pathway,
authors of other papers that used the same sample were Results of Aggregated Effect Size Analysis for Longitudinal
no longer contacted if the effect size information was not Associations Between Poverty Exposure and Weight
included in the manuscript. If effect size information from
the same sample was available from more than one paper, Use of the search methodology described above resulted in
effect sizes were averaged into a single “entry” for that the identification of 4156 articles. Of those, 130 were found
sample. to meet criteria for inclusion in the review. Eighteen studies
One exception to the above rule was that if the develop- included effect size information in the paper. For 49 stud-
mental stage during which BMI was measured (for the pov- ies, authors were emailed. Of those, 17 provided effect size
erty → weight pathway) or the type of SR measured (for the information. Twenty-five unique samples were represented
poverty → SR and SR → weight pathways) differed across in the analysis, with an additional five studies being from
multiple papers that used the same sample, multiple effect duplicate samples in which more than one paper included an
sizes were retained. For example, for the poverty → weight effect size (in these cases, effect sizes were averaged across
pathway, three studies utilized the Fragile Families dataset studies). Fifty-seven papers were from duplicate samples
(Assari et al. 2018; Burdette and Whitaker 2007; Starkey (e.g., ECLS-K, n = 14; NICHD SECCYD, n = 9; NLSY79,
and Revenson 2014). Although in all three papers research- n = 9) where the effect size was not available in the paper.
ers used the same measure of poverty (income at birth), As described above, authors were not emailed to obtain
each assessed their outcome of obesity or weight at a differ- effect size information for studies that used duplicate sam-
ent age, with Burdette and Whitaker (2007) using obesity ples (i.e., those samples for which an effect size had already
at age 3 (early childhood), Starkey and Revenson (2014) been obtained via the paper itself or an email to the author).
13

Effect sizes were converted to r from odds ratios, risk Strength of Associations by Features of the Sample
ratios, spearman’s rho, and Cohen’s d (which was usually
calculated by the first author using a table included in the Eight papers used predominantly low-income samples, with
paper) using an effect size calculation Web site (Lenhard three from the Fragile Families study (Assari et al. 2018;
and Lenhard 2016). Studies that had effect sizes of r ≥ .10 Burdette and Whitaker 2007; Starkey & Revenson, 2015)
were enumerated, as r = .10 is generally considered to be and five from other samples (Baptiste-Roberts et al. 2012;
the lower limit for what should be considered a “small” Coley and Lombardi 2012; Collings et al. 2017; Janjua et al.
effect size (Cohen 1992). Out of the 25 effect sizes from 2012; Lim et al. 2011). Of those, five studies found little to
the 25 unique samples, 14 exceeded that threshold with no association between poverty and weight (with r’s rang-
effect sizes of r ≥ .10 in the expected direction (i.e., greater ing from − 0.06 to 0.03). For the sixth study, which used a
poverty associated with higher weight) (see Tables 2, 3). low-income sample of 322 mother–child dyads from Bos-
Two effect sizes were in the opposite direction, with cor- ton, Chicago, and San Antonio, (Coley and Lombardi 2012),
relation coefficients for the association between poverty the effect size was in the small to moderate range (r = .19).
and weight in the − .10 to − .20 range (i.e., higher poverty However, the relationship was in the opposite direction of
associated with lower weight) (Coley and Lombardi 2012; what was expected, with higher income in early childhood
Tiberio et al. 2014). The total average effect size for the (age 0–2) associated with higher rates of child overweight/
longitudinal association between poverty and weight was obesity at age 7.
r = .09. Eleven of the 25 unique samples represented were com-
prised of non-U.S. participants (predominantly from the
Table 2 Studies with a positive Author Sample r (converted)

relationship and effect size ≥
0.10 for poverty-weight pathway Bhargava et al. (2008) ECLS-K, N = 7635 0.16
Li et al. (2007) and Strauss (1999) NLSY79; N = 1739; 2913 0.24
Williams et al. (2018) and Gibbs (2014) ECLS-B; N = 7022 0.12
Evans et al. (2012) Community sample (NY); N = 244 0.19
Tandon et al. (2015) Community sample (WA); N = 306 0.24
Baldwin et al. (2016) E-Risk twins; N = 2232 0.16
Holm-Denoma et al. (2014) Community sample (OR); N = 38 0.16
Lee et al. (2014) and Lane et al. (2013) NICHD SECCYD; N = 1150; 1238 0.11
Hancock et al. (2014) Longitudinal Study of Australian Children 0.18
(LSAC); N = 2596
Sen et al. (2017) Project VIVA; N = 992 0.12
Bush et al. (2017) Peers and Wellness Study; N = 338 0.17
Van Hulst et al. (2015) QUALITY Quebec; N = 512 0.13
Lim et al. (2011) Community sample (MI); N = 317 0.39
Strobino et al. (2016) Healthy Steps for Young Children 0.19
National Evaluation; N = 4745
Table 3 Studies with a negative relationship or effect size < 0.10 for poverty-weight pathway

Author Sample r (converted)
Mészáros et al. (2008) Hungarian sample, N = 495 − 0.04

Janjua et al. (2012) Sample recruited from University of Alabama prenatal clinic, N = 740 0.06
Slopen et al. (2012) Avon Longitudinal Study (UK), N = 3286 0.04
Baptiste-Roberts et al. (2012) The National Collaborative Perinatal Project, N = 28358 0.05
Carter et al. (2012) Longitudinal Study of Child Development in Quebec, N = 1566 − 0.07
Collings et al. (2017) Born in Bradford (UK); N = 1338 − 0.03
van Rossem et al. (2010) Generation R (Netherlands); N = 2954 − 0.05
Kwok et al. (2010) Children of 1997 (Hong Kong); N = 8237 0.04
Bergmeier et al. (2014) Australian Research Council Discovery Grant Study; N = 201 0.03
13
U.K., but also Canada, Hungary, Australia, Korea, and Timing of Exposure to Poverty and Weight Outcome
Hong Kong). Of the 11 studies that used non-U.S. sam- by Developmental Period
ples, four found evidence for a small to moderate associa-
tion between poverty and weight: one of the three studies In estimating effect sizes by age at which the weight out-
from the U.K. (Baldwin et al. 2016), one of the two from come was measured, poverty was most strongly related to
Australia (Hancock et al. 2014), and one of the two stud- weight when weight was measured in adolescence (r = .15,
ies from Canada (Van Hulst et al. 2015). As an example, n = 5), versus weight in early (r = .04, n = 7) and middle
Hancock and colleagues (2014) used a large longitudinal childhood (r = .07, n = 14). However, none of the differences
dataset from Australia (N = 2596) to address their primary between the magnitudes of the correlation coefficients were
research question of whether maternal protectiveness was statistically significant (z = 0.08, ns and z = 0.15, ns, for ado-
associated with greater risk of child overweight/obesity. lescence versus early childhood and adolescence versus mid-
The study, which started when children were 4–5 years dle childhood, respectively).
old, with follow-up assessments 2, 4, and 6 years later, Two samples out of the seven that tested the association
included income quartile as a covariate in analyses. At between poverty exposure and BMI measured in early child-
each follow-up visit, the group with the lowest income hood found an effect size where r ≥ 0.10 for the association
at baseline was consistently found to have greater risk of between the two variables. Two studies used the ECLS-
overweight/obesity compared to the highest income group, B, a large, representative U.S. dataset (N ≅ 7000–8000).
with odds ratios ranging from 1.7 (at the age 8–9 visit) to Researchers assessed risk of overweight/obesity at age 2
2.3 (at the age 10–11 assessment). Of the other six stud- (Gibbs and Forste 2014) and age 4–5 (Williams et al. 2018)
ies in which the association between poverty and weight by SES at birth, both finding effect sizes of small magni-
was minimal, samples were drawn from the Netherlands tude; averaged across the two studies, the effect size for this
(van Rossem et al. 2010), Hungary (Mészáros et al. 2008), sample was r = .12. Although Gibbs and Forste found that
Korea (Lee et al. 2016), and Hong Kong (Kwok et al. family SES at 9 months was associated with obesity at age
2010). When the 11 non-U.S. samples were removed from 2, infant feeding practices (i.e., being formula-fed and early
the aggregated effect size calculation, the new aggregated introduction to solid foods) were a powerful mediator in
effect size (n = 15) was r = .12. this relationship. Another study using a very large sample
of children across the U.S. found that lower-income tertile
measured when infants were 2–4 months was associated
Food Insecurity and Child Weight with higher risk of being in the 90th or 95th percentile of
weight-for-length (a measure similar to BMI that is appro-
Upon review of the literature, a separate aggregated effect priate for infants from birth to age 2) at age 2 (Strobino
size was calculated for food insecurity (FI), as the evidence et al. 2016).
of an association between FI and child weight outcomes is It should be noted that as the majority of studiest tested
generally weaker than for income and SES-based measures exposure to poverty in early childhood (n = 20), it was not
of poverty (Gundersen et al. 2008). Five effect sizes from possible to assess differences in the strength of associations
five different samples were included in the FI → weight depending on when exposure to poverty occurred.
pathway, all but one of which overlapped with the sam-
ples included in the poverty → weight pathway (i.e., five
studies tested the associations between both income and Summary of Literature Testing Associations Between
weight and FI and weight separately). Of the five studies Poverty and Child Weight
that tested relationships between exposure to FI and child
weight, the average effect size was 0.07 (with r’s ranging The overall average effect size for the pathway from poverty
from − 0.15 to − 0.16). One study (Coley and Lombardi to child weight was small (r = .09). However, for 14 studies
2012, described above) found that early childhood FI was using 14 independent samples, the effect size was greater
associated with lower rates of overweight and obesity in than r = .10, compared with nine studies where r < .10. None
middle childhood (r = − .15). Three other studies found of the studies using a predominantly low-income sample
that FI was associated with higher weight in early child- (n = 6) found evidence for an association between poverty
hood with an association of r ≥ 0.10. For the studies that and weight in the expected direction. Of the studies that used
also tested the association between poverty and weight a non-U.S. sample, a slight majority did not find evidence
(Bhargava et al. 2008; Coley and Lombardi 2012), the for an association between poverty and child weight. The
direction and strength of the association between FI and average effect size was the strongest (r = .15) for studies that
weight were quite similar as they were for the association tested the relationship between poverty and weight when
between poverty and weight. weight was measured in adolescence. There was no evidence
13

that the relationship between FI and weight differed from the that threshold with effect sizes of r ≥ 0.10 in the expected
relationship between poverty and weight. direction (i.e., greater poverty associated with lower SR)
(see Tables 4, 5).
Longitudinal Research Linking Poverty to Childhood
Self‑Regulation Association Between Exposure to Poverty on Hot, Cool,
and Questionnaire‑Based Measures of Child SR
SR as a Mediator in the Poverty to Obesity Pathway
Separate rs were then calculated to compare the effect sizes
Although other potential mechanisms are certainly involved for the associations between poverty and each type of SR
in the pathway from poverty to child weight, SR was selected (hot, cool, questionnaire-based). When the SR variable used
as the focus of this review because there are relatively rich in a study was a composite of both observed and question-
literatures on (1) poverty predicting SR and (2) SR predict- naire measures, it was classified as questionnaire-based. In
ing child weight. In addition, SR may be more malleable addition, if there was one sample that tested more than one
than other mediators that have proven resistant to change in type of SR in different studies or in the same study, effect
earlier trials (e.g., diet, physical activity) (Webster-Stratton, sizes for each were included in this analysis. As an example,
Jamila Reid, & Stoolmiller (2008). of the six studies that used the NICHD SECCYD sample and
Similar to the poverty to weight pathway reviewed above, had an effect size included in the original paper, Crook and
the review included studies in which authors primarily Evans (2014), Dilworth-Bart, Khurshid, and Vandell (2007),
sought to test relationships between poverty and child SR. and Blums, Belsky, Grimm, and Chen (2017) all tested asso-
Also included were studies where the research question was ciations between poverty and cool SR, as measured by the
not specifically about relations between poverty and SR, Tower of Hanoi task (Blums et al. 2017; Crook and Evans
but about some other independent variable (e.g., maternal 2014) and the Continuous Performance Test (Dilworth-Bart
depression, family structure) and SR, with SES or family et al. 2007). The aggregated effect size for cool SR was cal-
income included as a covariate in analyses. culated using an average of the rs from these three studies.
Two of the studies (Blums et al. 2017; Nievar et al. 2014)
Results of Aggregated Effect Size Analysis for Longitudinal tested associations with hot SR using a delay task, and the
Associations Between Poverty Exposure and SR hot SR aggregated effect size was calculated as the average
from these two papers.
Searching PubMed and PsycINFO with the terms listed In comparing the magnitude of the aggregated effect
above for the poverty → SR pathway yielded 842 articles. sizes for the three types of SR, all were comparable to one
Of those, 95 were found to meet criteria for inclusion in the another. For studies that tested the relationship between
review. Forty-one studies already reported effect size infor- poverty and cool SR (n = 21), r = − .14; for poverty and hot
mation in the paper. For 22 studies, authors were emailed to SR (n = 11), r = − .13; and for poverty and questionnaire-
obtain this information. Of those, 15 provided data on effect based SR (n = 19), r = − .17. A Fisher’s z test indicated that
sizes. Thirty-seven unique samples were represented in the none of the correlation coefficients were significantly differ-
effect size aggregation analysis; an additional 19 papers used ent from one another. Although more studies used cool SR
an already included sample. Thus, for these already included measures than hot, there was little evidence to indicate that
samples, effect sizes were averaged across these multiple the strength of the relationship between poverty and SR was
studies. Twenty-four papers were from duplicate samples weaker when hot SR measures were used.
where the effect size was not available in the manuscript
(e.g., the Family Life Project, n = 10; ECLS-K, n = 2, the Features of the Sample and Measurement of Poverty
Early Head Start Research and Evaluation Project, n = 2).
Effect sizes for all but two of the studies in the pov- Most studies used income-based measures of poverty. Of
erty → SR pathway were already in the form of r and did not the eight studies from six independent samples that used a
need to be converted. A total r for the association between cumulative risk index that included a measure of poverty
poverty and SR was calculated. When a study tested hot, (Bocknek et al. 2014; Brophy-Herb et al. 2013; Chang et al.
cool, and questionnaire-measured SR (or any combination 2012; Conradt et al. 2014; Doan et al. 2012; Lengua et al.
of two or three of these), the highest of the r’s for any one 2007; Mistry et al. 2010; Wade et al. 2018), only Wade and
measure was used to represent the total. The total average colleagues used an observed measure of child SR, with the
effect size for the longitudinal association between poverty other studies using questionnaires. Using a relatively large
and SR was r = − .16. As in the poverty → weight pathway, Canadian sample, Wade and colleagues (2018) found that a
studies that had effect sizes of r ≥ 0.10 were tabulated. Out of cumulative risk index (including variables such as mater-
the 37 effect sizes from the 37 unique samples, 28 exceeded nal depression and single parenthood along with low family
13
Table 4 Studies with a negative relationship and effect size ≥ 0.10 for poverty-self-regulation (SR) pathway
Author Sample SR classification r (total)
Wanless et al. (2011) Community sample, N = 95 Cool − 0.26

McClelland (2012) Community sample, N = 34 Cool − 0.29
Brown (2013) Sample from PA head start, N = 205 Cool − 0.11
Nesbitt et al. (2013) Durham child health & development study, N = 206 Cool − 0.23
Sturge-Apple et al. (2015) Community sample, N = 140 Hot − 0.24
Lengua et al. (2007) Community sample, N = 70 Questionnaire − 0.26
Lengua (2006) Community sample, N = 120 Questionnaire − 0.22
Thompson et al. (2013), Tandon et al. (2015), Len- Community sample, N = 306 Hot; Cool − 0.24
gua et al. (2013, 2014, 2015)
Doan et al. (2012) Community sample, N = 265 Cool; Questionnaire − 0.31
Sektnan et al. (2010), Crook (2014), Dilworth-Bart NICHD SECCYD, N’s range from 1009–1364 Hot; Cool; Questionnaire − 0.26
et al. (2007), Nievar et al. (2014), Blums et al.
(2017), Prattet al. (2016)
Conradt et al. (2014) Maternal lifestyle study, N = 1121 Physiological − 0.10
Moilanen (2010) Early steps study, N = 731 Questionnaire − 0.13
King et al. (2013) Community sample, N = 214 Questionnaire − 0.17
Bernier et al. (2010, 2015), Rochette (2014) Canadian community sample, N’s range from Hot; Cool; Questionnaire − 0.27
60–114
Suor et al. (2018) Community sample, N = 160 Hot; Cool; Questionnaire − 0.21
Berry et al. (2014), Vernon-Feagans et al. (2016), Family life project, N’s range from 1037–1235 Cool; Questionnaire − 0.26
Gueron-Sela et al. (2018), Sulik et al. (2015)
Bockneck et al. (2014), Mistry et al. (2010), Brophy- Early head start research & evaluation project, N’s Cool; Questionnaire − 0.16
Herb et al. (2013) range from 1258–1851
Peredo et al. (2015) Community sample, N = 100 Hot − 0.19
Kidwell (2007) Community sample, N = 56 Questionnaire − 0.27
Sabol et al. (2018) National center for research on early childhood Cool − 0.15
education study, N = 211
Wade et al. (2018) Healthy babies healthy children (Canada), N = 501 Cool − 0.13
Eisenberg et al. (2010) Community sample, N = 209 Hot; Questionnaire − 0.18
Gartstein et al. (2013) Community sample, N = 147 Cool − 0.13
Martoccio et al. (2014) Sample from MI head start, N = 127 Cool − 0.12
Hartanto et al. (2018) ECLS-K Cool; Questionnaire − 0.19
Bassett et al. (2012) Community sample, N = 261 Hot; Cool − 0.15
Evans et al. (2005) Community sample, N = 329 Questionnaire − 0.29
Abenavoli et al. (2015) Community sample N = 207 Cool; Questionnaire − 0.15
Table 5 Studies with a negative relationship or effect size < 0.10 for poverty-SR pathway

Brown et al. (2013) Sample from PA head start; N = 205 Questionnaire 0.05
Roy et al. (2014a), Roy and Raver Chicago school readiness project; N’s range from 391–602 Cool; Questionnaire − 0.07
(2014b), Raver et al. (2013)
Choi et al. (2016) Sample from OK early head Start/Head Start; N = 169 Cool − 0.03
Marcovitch et al. (2015) Community sample; N = 226 Cool − 0.08
Razza et al. (2010) Fragile families; N = 1046 Cool − 0.07
Degol (2015) Community sample; N = 216 Hot − 0.06
Pearce et al. (2016) UK Millenium cohort study; N = 5107 Questionnaire − 0.05
13

income) in early infancy was associated with lower chil- intervals. In all studies, researchers found significant asso-
dren’s cool SR (as measured by tasks that tested inhibitory ciations between income and both hot and cool SR, with hot
control—the Bear/Dragon and cognitive flexibility—the SR measured with a gift delay task and cool SR measured
Dimensional Change Card Sort) at age 3 (r = − .13). using a battery that included verbal inhibitory control and
Fifteen studies (not including duplicate samples) used cognitive flexibility tasks. Importantly, Lengua and col-
predominantly low-income samples. The Family Life Pro- leagues (2014) found that the effect of income on both types
ject, a longitudinal study of children born in low-income, of SR was primarily explained by parenting variables (e.g.,
rural counties in North Carolina and Pennsylvania, was one scaffolding, limit-setting).
such sample that was used in four different studies included
in the aggregated effect size analysis. Using this sample, Effect of Poverty on Physiological SR
researchers consistently found an association between early
childhood income (when children were 0–3 years) and SR, Very few studies examined direct effects of poverty or pov-
also measured in early childhood. The observed SR battery erty-related stressors on children’s parasympathetic nervous
included multiple tasks assessing several aspects of cool SR, system functioning, often assessed using respiratory sinus
including working memory, attention shifting, and inhibi- arrhythmia (RSA). Baseline RSA is thought to provide an
tory control (Berry et al. 2014; Gueron-Sela et al. 2018; indication of physiological SR while at rest, and higher RSA
Sulik et al. 2015; Vernon-Feagans et al. 2016); teachers also is usually associated with more optimal health and behavior
reported on children’s “behavioral regulation” in kindergar- outcomes (Propper 2012). Children’s RSA is known to be
ten (Vernon-Feagans et al. 2016). To determine whether the vulnerable to several types of environmental stressors, such
magnitude of effect sizes was comparable, an aggregated as adolescents’ exposure to early life adverse events (Daches
effect size was calculated for the association between pov- et al. 2017). Only one study was found that assessed the
erty and SR for low-income (n = 15) versus representative longitudinal relation between poverty and/or cumulative risk
samples (n = 22). The magnitude was stronger for represent- and RSA. Researchers found that a cumulative risk index
ative samples (r = .18 versus r = .13), but not significantly so (which included poverty) from birth to age 3 was associated
(z = 0.14, ns). with less RSA reactivity and less growth in reactivity from
ages 3 to 6 (Conradt et al. 2014). This finding is opposite
Child Age of what was expected, as greater RSA reactivity is typically
associated with negative outcomes. However, by age 6,
The majority of studies (n = 27) tested the outcome of SR researchers found that there was a significant, positive asso-
in early childhood, usually starting at preschool-age (i.e., ciation between early cumulative risk and RSA reactivity
age 3–4), with just three studies using SR measured in ado- in the expected direction, indicating that perhaps there is a
lescence as a final outcome (Doan et al. 2012; Evans et al. “latent” effect of early adversity on physiological regulation.
2005; King et al. 2013). The study that used the earliest
outcome measure for SR, at age 18 months, was the only Effect of Food Insecurity on SR
one of all 37 studies in the analysis for which the effect size
was in the opposite of the expected direction (i.e., higher Only one study was found that tested the association between
poverty associated with higher SR, r = .13). In this study, child FI and SR longitudinally. Using the ECLS-K dataset,
Gartstein, Bridgett, Young, Panksepp, and Power (2013) Grineski, Morales, Collins, and Rubio (2018) found evi-
found that income in early infancy was negatively associated dence for a small, negative association between FI in kinder-
with parent-reported SR at 18 months using the effortful garten and SR one year later, as measured by teacher report
control subscale from a measure of early childhood tempera- (r = − .07) and observed cool SR (using a working memory
ment; maternal parenting stress was positively associated test and the Dimensional Change Card Sort test, which tests
with effortful control. cognitive flexibility, r = − .10).
Many studies, however, found significant associations
for the association between poverty and child SR when Summary of Research Linking Poverty Exposure to Child SR
children were as young as 3 and 4 years old (e.g., Lengua
et al. 2007; Sektnan et al. 2010; Sturge-Apple et al. 2016). Overall, exposure to poverty and poverty-related risk was
As an example, Thompson and colleagues (2013), Tandon, associated with lower SR, with several studies indicating
Thompson, Moran, and Lengua (2015), and Lengua in three that the association between poverty and diminished SR may
different studies (Lengua et al. 2014, 2015, 2013) all used be detected at as early as age 3 or 4. The majority of the liter-
the same economically representative community sample ature focused on cool rather than hot SR as an outcome, but
to test associations between age 3 income and both hot and there was no evidence that the association with poverty was
cool measures of child SR assessed four times at nine-month stronger for cool SR. More research is needed to understand
13
more about the effect of poverty on physiological SR and the measure different aspects of child SR in relation to eating
effect of food insecurity on SR in general. behavior than traditional hot and cool SR tasks.
Longitudinal Research Linking Self‑Regulation Results of Aggregated Effect Size Analysis for Longitudinal
to Childhood Weight Associations Between SR and Weight
The final pathway in the proposed model focuses on SR as Searching PubMed and PsycINFO with the terms listed
a predictor of child weight. In this model, SR is thought to above for the poverty → weight pathway yielded 1131 arti-
mediate the effects of poverty on child weight, and thus in cles. Of those, 33 were found to meet criteria for inclusion in
the following section SR is discussed specifically as a vari- the review. Twelve studies already included effect size infor-
able that predicts child weight outcomes. mation in the paper. For 13 studies, authors were emailed
to obtain this information. Of those, six provided effect size
Longitudinal Research on SR to Child Weight information. Seventeen unique samples were represented in
the effect size aggregation analysis. Two additional papers
Measurement and Overview Measures of SR in relation to that reported a correlation coefficient used a duplicate sam-
weight tended to resemble those used in studies unrelated ple (in both cases, the sample was drawn from the NICHD
to eating or weight. However, there are some domain-spe- study). Five other papers were from duplicate samples where
cific measures used to measure SR specifically related to the effect size did not appear in the manuscript.
food intake. One example is the Eating in the Absence of The total average affect size was calculated in the same
Hunger (EAH) paradigm. In EAH tasks, children are usu- manner as for the poverty → SR pathway, with the highest of
ally first provided with a standardized meal. When children the rs for any one measure of SR (i.e., hot, cool, or question-
report that they are full, they begin a “free-access” session naire-based) used to represent the total. The total aggregated
in which they are allowed to eat as much as they would like affect size for the SR → weight association was r = − .13,
from an assortment of palatable sweet and savory snacks. with greater SR associated with lower weight. Thirteen out
The amount of snacks that they eat is then recorded, with of the 17 unique samples had an effect size of r ≥ 0.10 (see
children who consume fewer calories in the absence of Tables 6, 7).
hunger assumed to be better at self-regulating their caloric
intake. In addition to the EAH protocol, there are also some Type of Self‑Regulation and Longitudinal Associations
questionnaire measures that assess child SR regarding eat- with Weight
ing behaviors (e.g., Children’s Eating Behavior Question-
naire (CEBQ); Wardle, Guthrie, Sanderson & Rapoport Again, a correlation coefficient for the association between
2001). each type of SR (cool, hot, questionnaire) and later weight
The extent to which domain-specific eating SR tasks was calculated. Of the four studies that tested the relation-
are associated with more global measures of SR remains ship between cool SR and child weight, the average r = − .04,
unclear, with some studies finding associations between with just one study finding an association in which r ≥ 0.10.
EAH and traditional SR measures (Tan and Holub 2010) Of the studies that evaluated associations between hot SR
and others not finding evidence of this association (Leung and child weight, all seven r’s were ≥ 0.10 and the average
et al. 2014). Interestingly, Hughes and colleagues (2015) effect size was r = − .14. In one study that tested multiple
found a significant correlation between EAH and a food types of SR (hot, cool, and questionnaire-based) within the
DoG task in low-income preschoolers in an unexpected same large sample (NICHD), researchers used a single SR
direction, with greater SR in the food DoG task associated factor comprised of five indicators (hot, cool, and question-
with more calories consumed in the absence of hunger. They naire-based, all measured at age 4) and found that higher
also found that only eating-specific SR tasks or question- SR predicted lower BMI at three different ages (8, 11, and
naires (EAH, parent-reported satiety and food responsive- 15) (Bub et al. 2016). In addition, researchers found that SR
ness on the CEBQ) were concurrently associated with BMI, accounted for more of the variance in BMI when children
while domain-general hot (food DoG, gift delay) and cool were older. However, in considering the bivariate correla-
(tapping task, flexible item selection task) SR tasks having tions between each SR indicator and BMI averaged across
no significant relation. In addition, a cross-sectional study the three ages, the strongest relations were for hot SR (meas-
found that for preschoolers, eating more in the absence of ured with a food DoG task) and the questionnaire measure
hunger was associated with greater BMI, while waiting time (parent report of inhibitory control and attention focusing),
on a food DoG task was not (Power et al. 2016). Together r = − .10 and r = − .09, respectively, while the association
these studies suggest that domain-specific food SR tasks may between cool SR (measured with a Stroop and Continuous
13

Table 6 Studies with a negative relationship and effect size ≥ 0.10 for self-regulation (SR)-weight pathway
Graziano et al. (2010) Community sample; N = 57 Cool; Hot − 0.14

Anzman (2009) Community sample; N = 197 Questionnaire − 0.30
Evans et al. (2012) Community sample; N = 244 Hot (food) − 0.21
Mallan et al. (2014) Community sample; N = 37 Questionnaire (appetite-specific) − 0.38
Graziano et al. (2013) Community sample; N = 195 Hot − 0.19
Francis (2009), Seeyave et al. NICHD; N’s range from 805–1061 Hot (food); Cool; Questionnaire − 0.14
(2009), Bub et al. (2016)
Duckworth et al. (2010) Community sample; N = 105 Questionnaire/observed composite − 0.22
Tandon et al. (2015) Community sample; N = 306 Hot; Cool − 0.10
Steinsbekk (2015) Sample from Norway; N = 687 Questionnaire (appetite-specific) − 0.21
Asta et al. (2016) Sample from MI head start; N = 209 Eating in the absence of hunger task − 0.11
Parkinson et al. (2010) Gateshead millenium study (UK); N = 583 Questionnaire (appetite-specific) − 0.15
Carter et al. (2012) Longitudinal study of child development in Que- Questionnaire (appetite-specific) − 0.41
bec, N = 1566
Table 7 Studies with a negative relationship and effect size < 0.10 for self-regulation (SR)-weight pathway
Anderson et al. (2017) Millenium cohort study; UK, N = 10,955 Questionnaire 0.05
Piche et al. (2012) Longitudinal study of child development in Quebec; N = 966 Questionnaire 0.04
Stautz et al. (2016) Avon longitudinal study (UK); N = 6069 Cool; Questionnaire − 0.02
Fogel et al. (2018) Growing up in Singapore towards healthy outcomes; N = 158 Eating in the absence of − 0.05
hunger task
Bergmeier et al. (2014) Australian research council discovery grant study; N = 201 Questionnaire 0.05
Datar et al. (2018) ECLS-K; N = 7070 Questionnaire − 0.07
Performance Test) and average BMI was relatively lower and EAH at 27 months was much stronger than for EAH
(r = − .03). at 21 months. Asta and colleagues note that in their study
A total of 11 studies analyzed questionnaire-based meas- the number of calories consumed in the absence of hunger
ures of SR in relation to child weight, with an aggregated increased with age, particularly for sweet foods, support-
effect size of r = − .14. Of those, eight used questionnaires ing some prior research indicating that children’s eating SR
that reported on general aspects of child SR (e.g., Chil- declines with age (Johnson and Taylor-Holloway 2006).
dren’s Behavior Questionnaire; Rothbart et al. 2001), and However, Fogel and colleagues’ (2018) study focused on a
three used eating-specific questionnaires. The average effect group of children in Singapore who were slightly older, with
size for questionnaires that assessed eating-specific behav- an EAH task at age 4 used to predict age 6 BMI, and did not
iors was r = − .25, whereas for questionnaires that assessed find an association between EAH and BMI.
SR broadly, r = − .07. All three of the studies that used
eating-specific questionnaires used the CEBQ: some, but Developmental Timing
not all of the eight scales that comprise the CEBQ assess
children’s SR of eating (e.g., emotional over-eating, satiety In estimating effect sizes by age at which the weight outcome
responsiveness). was measured, overall SR (i.e., not separated by SR type)
Of the two studies that used an EAH task to predict was most strongly related to weight when measured in ado-
later BMI, one study found some evidence for an asso- lescence (r = − .18, n = 5), versus measuring weight in early
ciation (Asta et al. 2016), while the other did not (Fogel (r = − .10, n = 6) and middle childhood (r = − .11, n = 6).
et al. 2018). Asta and colleagues (2016), using a predomi- None of the differences between the magnitudes of the cor-
nantly low-income sample, tested EAH twice in very early relation coefficients were statistically significant (z = 0.09, ns
childhood (21 and 27 months) and BMI when children and z = 0.08, ns, for adolescence versus early childhood and
were 33 months. The association between 33 month BMI adolescence versus middle childhood, respectively).
13
Most studies (n = 13) measured SR beginning in early
Tapping task) associated with EF-training

lower % of overweight children compared
No effect of intervention on obesity at the
childhood. Six studies measured weight through adolescence
At 3-year follow-up, ECI associated with
for the control group while EF-training

working memory and meta-cognition
Improvements in cool SR (Corsi Block
subscales of teacher-reported BRIEF

group (r = .20). Increased deficits in
and found significant effects for the relation between SR
measured in early/middle childhood and adolescent weight.
group remained stable (r = .27)

end of academic year (r = .01)
The association between SR measured in early childhood
Intervention effects on weight

on adolescent weight was established using both observed
(Francis and Susman 2009) and questionnaire measures of
SR (Anzman and Birch 2009).
to ST (r = .30)
Effects of SR‑Promoting Interventions on Child Weight
Three studies were reviewed that used a weight loss inter-

vention that included an SR component and measured SR
pre- and post-treatment (see Table 8). One evaluated the
teacher-reported SR (average r = − .19)

pared to both other groups (r = .21 for
HS + POPS + IYS versus HS + POPS)
EF-training group 8 weeks post-inter-

but higher child-reported SR (r = .16)
12-week f/u (r = .05 for 12-week f/u)

HS + POPS + IYS group had improve-
vention, but not maintained through

ECI associated with lower parent and
effect of a general SR-building intervention as part of an
ments in teacher-reported SR com-
Better weight loss maintenance in

obesity prevention program for low-income preschoolers
(Lumeng et al. 2017) and two tested a weight loss interven-
tion for obese children ages 8–13 (Israel et al. 1994) and
Intervention effects on SR
8–14 (Verbeken et al. 2013).
An effect size (d) for the effect of each intervention on
SR and weight was calculated and then converted to r to be
consistent with the effect size measures used for the other
associations. The total aggregated effect size for the effect
of the interventions (N = 3) on SR was r = .13 in favor of
the SR intervention. For two of the three intervention stud-
ies (Lumeng et al. 2017; Verbeken et al. 2013), there was
Standard treatment (ST) vs. Enhanced

series (IYS) vs. head start (HS) only
(POPS) vs. POPS + Incredible years
Child Involvement (ECI), including

some evidence that the intervention promoted children’s
Executive Functioning (EF) training

training in self-management skills
Preschool obesity prevention series
SR, as measured by questionnaires. Only Verbeken and col-

leagues (2013) used an observed SR measure to investigate
changes in SR during the intervention period, finding that
the intervention was associated with increases in cool SR
versus care as usual

(r = .20). Israel and colleagues (1994), using parent, teacher,
and child report of SR, did not find consistent associations
Intervention
between the intervention, which included training in child

self-management and SR, and improvements in child SR;
in fact, averaging across parent, teacher, and child reports,
greater SR improvements were demonstrated in the compari-
son group (average r = − .10, ns).
Verbeken et al. (2013) N = 44 obese Dutch children age
N = 697 Head start preschoolers
8–14 in an inpatient treatment

N = 34 obese children age 9–13
For calculating the aggregated effect size for the effect of

Table 8 SR-Promoting Weight Intervention Studies
the interventions on child weight, if there were multiple fol-

low-up assessments of weight, the final outcome for weight
was used. The average r was 0.12 for the effect of the inter-
ventions on child weight compared to the control group, with
one study finding an effect on weight and the other two stud-
program
ies having no or minimal effect on weight change. Israel and

Sample
colleagues (1994) found that their “enhanced child involve-

ment” intervention was associated with a lower proportion
of children who were overweight at 3 years post-treatment
Lumeng et al. (2017)
Israel et al. (1994)
compared to the standard treatment group (r = .30). Interest-

ingly, as noted above, Israel and colleagues (1994) did not
find any effect of the intervention on SR, suggesting that the
Author
effects of the intervention on child weight at follow-up were

operating through a mechanism other than improvements to
13

children’s SR. For the other two intervention studies, there Although the aggregated effect size for the pathway from
was no evidence to suggest a meaningful effect of SR train- poverty to weight was particularly small, it is comparable to
ing on weight loss or weight loss maintenance. the effect sizes that previous systematic reviews and/or meta-
analyses have established for other risk factors for child obe-
Summary sity. Examples include the following: maternal smoking dur-
ing pregnancy and child overweight: OR = 1.5 (equivalent to
The aggregated average effect size from SR to child weight r = .11) (Oken et al. 2008); breast-feeding and child obesity:
was in the “small” range (r = − .13) and comparable to the OR = .78 equivalent to r = .07 (Arenz et al. 2004); and TV
magnitude of the correlation coefficients for the other two viewing and child body fat: r = .07 (Marshall et al. 2004).
pathways. There was some evidence that hot SR is a stronger For the pathway from poverty to SR, the effect size found
predictor of later weight than cool SR, with hot SR more in the current review (r = − .16) was identical to one found in
commonly tested. Of the different measures of SR, there a recent meta-analysis on the association between SES and
was also some evidence that eating-specific questionnaires behavioral executive functioning tasks (i.e., observed cool
were more predictive of later weight than domain-specific SR) (Lawson et al. 2018). There was minimal overlap of the
SR questionnaires, but only three studies used this measure. studies include in both analyses—of the 25 unique samples
There was little evidence that SR-focused obesity prevention in their meta-analysis and the 37 in the effect size analysis in
or intervention programs help reduce child weight, but only the current review, only four samples overlapped. Lawson,
three studies were found that tested a child weight interven- Hook, and Farah’s meta-analysis also included education
tion and also measured SR. and occupation-based measures of SES and was not limited
to longitudinal studies.
Although to our knowledge no other systematic reviews
Discussion or meta-analyses have been conducted on the relationship
between SR and weight in children, a recent meta-analysis
Synthesis of Literature and Presentation of Updated of the association between behavioral task and questionnaire
Model measures of impulsivity and adult BMI also found evidence
for a small but significant effect size (r = .07) (Emery and
At the outset of this review, it was proposed that child SR Levine 2017).
might be one explanatory mechanism by which poverty Based on the literature reviewed, Fig. 1 represents a theo-
influences children’s weight outcomes. In each of the path- retical model depicting relationships among the variables of
ways of interest, effect sizes were in expected direction, interest. The model includes both food insecurity and cumu-
albeit in the “small” range, with r = .09 (poverty → weight), lative risk as variables related to poverty that may influ-
r = − .16 (poverty → SR), and r = − .13 (SR → weight). ence child weight. The dotted lines from FI to SR and child
Fig. 1 Theoretical model of
associations between poverty,
SR, and child weight
13
weight and from cumulative risk to child weight reflect that SR to Weight
there is not yet enough evidence to suggest an association
for these pathways and the need for more research examin- Surprisingly little research has tested the extent to which
ing these paths. eating-specific SR measures, particularly EAH, predict later
child weight or weight gain over time, with only one obser-
vational EAH study evaluating this association in early to
Limitations of the Current Literature
middle childhood and three studies using parent report. It
would be helpful to learn more about relationships between
Poverty to Weight
different types of SR, including EAH and hot/cool observed
measures, and child weight longitudinally across develop-
The majority of studies reviewed in this pathway measured
mental stages.
exposure to poverty during early childhood. Although there
are compelling reasons to suspect that exposure to poverty
Self‑Regulation Across Literatures
in early childhood may be particularly harmful, such as
early adversity impacting reactivity to stress (Shonkoff et al.
A challenge of the proposed model that future research
2012), it remains necessary to specifically test this assump-
should address is the extent to which each type of SR (hot,
tion by comparing children’s exposure to poverty at various
cool, appetite-specific) is a viable candidate to mediate the
points in development in relation to later weight. It would be
association between poverty and weight. The current review
important to isolate the effects of poverty exposure during
revealed that although there is some overlap in how SR is
different developmental stages on child weight to determine
conceptualized in relation to the poverty to SR and the SR to
the extent to which early poverty has unique effects on child
weight pathways, there are also important differences. In the
BMI. Alternatively, more prospectively ascertained data are
poverty to SR pathway, more of the literature has focused on
also needed to examine the cumulative effects of experienc-
cool SR. The relatively greater emphasis on cool SR in most
ing poverty across multiple developmental periods (e.g.,
studies may be because of a greater emphasis in the litera-
early and middle childhood versus early childhood through
ture of identifying possible mechanisms for the relationship
adolescence).
between poverty exposure and academic achievement. As an
example, some research suggests that cool SR, rather than
Poverty to SR hot, is more strongly associated with academic performance
(Willoughby et al. 2011). In line with this explanation, for
A major limitation of the poverty to SR literature was that many of the studies that used exclusively cool SR measures,
only three studies tested SR in adolescence. Moreover, few authors emphasized the importance of SR in the context of
studies spanned multiple developmental periods. There- academic achievement (McClelland and Wanless 2012; Nes-
fore, it remains unclear whether early exposure to poverty bitt et al. 2013; Thompson et al. 2013; Wanless et al. 2011).
continues to predict SR later in childhood and adolescence. Conversely, researchers investigating the pathway from
Prior research suggests that early exposure to poverty may SR to weight have primarily used hot SR measures, with
be particularly detrimental for outcomes much later in devel- relatively little evidence for an association between cool SR
opment such as academic achievement and employment and child weight. It could be that hot and cool SR differ-
(Duncan and Magnuson 2013), but there is not yet enough entially affect various aspects of weight regulation or that
evidence to indicate that this is also the case for SR. the relative influence of each changes throughout develop-
Only one study evaluated the longitudinal relationship ment. For example, hot SR might be a better predictor of
between FI and SR, using both teacher report and cool weight during early childhood, and cool SR might become
SR measures (Grineski et al. 2018). For developing a bet- more important later in childhood (Metcalfe and Mischel
ter understanding of the relationship between poverty and 1999) when children become more responsible for their own
weight outcomes, it would be particularly interesting and nutritional choices. More studies are needed that evaluate
important to test the relationship between FI and SR as the effect of poverty and/or cumulative risk on hot SR, and
measured by DoG tasks. When children are food insecure, especially studies that compare how poverty affects cool
they may not know exactly when their next meal might be, versus hot SR. Few researchers across both literatures used
which could affect whether they are willing to wait for a big- the same questionnaire measures of SR, which could sim-
ger reward. Additionally, no studies were found that tested ply reflect the multitude of SR-related questionnaires that
the longitudinal relationship between poverty or FI and eat- are available rather than differences in their usage across
ing-specific child SR, although a cross-sectional study found literature studies. Neither poverty nor obesity researchers
that children living in homeless shelters reported overeating have sufficiently tested physiological SR (e.g., RSA) as an
as a coping strategy for FI (Smith and Richards 2008). outcome or predictor, and poverty researchers also have not
13

evaluated eating-specific SR as a potential outcome of low in children’s weight was driven by improvements in child
SES or poverty-related stress. SR or some other variable (e.g., perhaps the intervention
influenced parental food purchases).
Global Versus Domain‑Specific SR Unfortunately, the few studies that have tested SR-specific
interventions and monitored SR have not found particularly
Eating-specific SR has been limited to use as a predictor compelling evidence to suggest that they help treat or pre-
of child weight. No studies tested the influence of poverty vent obesity. Researchers in one recent study included in the
on SR specific to eating or appetite (e.g., observed EAH, present analysis (Lumeng et al. 2017) tested an SR interven-
parent-reported overeating or eating too quickly). How- tion in a manner closely aligned with the model proposed
ever, there are some theoretical reasons why an association in this review, adding a general SR component (the Incred-
between poverty and SR of eating might be expected. Prior ible Years Series) to an obesity prevention program for low-
research suggests that general exposure to stress (not spe- income preschoolers in Head Start. There are a variety of
cific to poverty-related stressors) may affect children’s eating possible explanations for why Lumeng and colleagues did
behaviors (Michels et al. 2012), and that cortisol output in not find evidence for an effect of the SR intervention on child
response to a laboratory-based stressor is associated with weight. One possibility is that although interventions that
eating more in the absence of hunger (Francis et al. 2013). target children’s global SR may be beneficial for improv-
This may be an interesting and important area for future ing hot/cool SR, a component specific to eating SR may be
investigation. necessary for obesity prevention. Another possibility could
Prior research demonstrates that correlations between be that children in the study were too young to benefit from
EAH and more global SR tasks, even DoG with food, are an SR intervention for obesity prevention, as some literature
not as strong as might be expected (Hughes et al. 2015). It suggests that younger children are better at self-regulating
may be that EAH primarily measures reward responsive- their energy intake than older children (Cecil et al. 2005).
ness, and a child could be responsive to reward (i.e., enjoy A third explanation is that effects of the SR intervention
the pleasurable taste of food) while still being able to suc- might have latent effects on BMI that may not be evident
cessfully delay with the promise of a better reward. Another until later in development during “sensitive” periods, such
possibility is that SR of eating in the context of an EAH task as adiposity rebound (~ age 7) and puberty. Lastly, it could
is more of an involuntary process and is tapping a different be that SR-promoting interventions are most effective for
aspect of SR than those that require volitional inhibition those children with the lowest levels of SR at baseline, and
(e.g., DoG tasks). It would be important for longitudinal that SR interventions may fail to demonstrate main effects
studies to further explore how different aspects and measures on weight for the whole sample while still having a moder-
of SR are affected by poverty and affect weight, particularly ate effect on specific subsamples (e.g., children with low
because eating-specific SR tasks have infrequently been used SR). Interestingly, both Brotman and colleagues’ (2012) and
as a predictor of later weight in methodologically rigorous, Smith and colleagues’ (2015) studies focused on youth with
longitudinal studies. high levels of conduct problems, and conduct problems are
associated with diminished SR. Testing eating-specific ver-
Feasibility of Model with Respect to SR sus general SR interventions for obesity treatment and pre-
Interventions vention for children at a variety of stages of development, as
well as potential moderators of treatment efficacy are areas
One goal of the current review was to evaluate the existing of research that future studies should systematically address.
evidence for an explanatory mechanism in the pathway There is an ongoing treatment study testing the effects
from poverty exposure to child risk of obesity with the of an SR-promoting intervention for obese 8–16-year-
intention of identifying a potential target of intervention. old children (Warschburger 2015), but findings are not
Findings from a recent meta-analysis indicate that not only yet available. The intervention, which will be compared
do SR-promoting interventions have beneficial effects on to a no-treatment control condition, involves a computer
child SR, but also on more distal outcomes, including program where children are taught to “push” high-calo-
academic achievement and behavior (Pandey et al. 2018). rie snacks away while learning to “approach” nutritious
With respect to weight specifically as an outcome, prior foods (e.g., vegetables), with the Stroop test being used to
research indicates that early childhood interventions that measure changes in SR throughout the intervention period.
implicitly or explicitly promote SR skills may be associ- This trial could add to the existing literature by testing the
ated with healthier BMIs later in childhood (Brotman et al. efficacy of appetite-specific SR training on child weight
2012; Smith et al. 2015). However, these studies did not reduction, whereas the three other treatment studies tested
directly measure children’s SR, so it is unclear whether the more general SR interventions.
association between the intervention and improvements
13
Research suggests that SR specific to energy intake is Another possibility that few studies have accounted for in
modifiable (Johnson 2000). In a study that was not a con- the relationship between SR and weight is a genetic pathway.
trolled trial and that did not test weight as an outcome, Perhaps there may a relationship between low SR and risk
Johnson found that after six weeks of an SR-promoting of overweight, but it is not a causal one, and both traits are
intervention, normal weight preschool children were bet- inherited. For example, when Seeyave and colleagues (2009)
ter able to regulate their eating in an EAH task, indicating included a proxy for maternal weight in their model, the lon-
that eating-specific SR behaviors may be malleable starting gitudinal association between SR and BMI was significantly
from an early age. Using doll play and skits, the intervention attenuated, which they attribute to “inherited disinhibition.”
modeled appropriate SR of eating and focused on identifying Future studies should investigate the heritability of SR of
physiological hunger and satiety cues. However, no studies eating behaviors.
were found that specifically incorporated eating-specific SR Inherent in the proposed model and the studies reviewed
training into an intervention and measured effects on both in the SR to child weight path is that SR promotes healthy
SR and weight. One study addressing this issue is currently weight because of SR in relation to eating and appetite.
ongoing, testing an intervention that focuses on promoting However, eating is one of two major components of child
parents’ knowledge of eating SR for their children (Eneli weight maintenance, with the other being physical activity.
et al. 2015). Although the relationship between SR and physical activ-
ity has been previously evaluated, usually the focus is on
Developmental Considerations the SR-promoting effects of physical activity rather than SR
being a necessary component for maintaining regular physi-
One of the aims of the current review was to determine cal activity.
whether effects of poverty and SR on weight are visible in
early childhood, prior to adiposity rebound. There was some Limitations of the Current Study
evidence that perhaps the relationship between poverty and
weight is weaker prior to adiposity rebound (~ age 5–7), with A limitation of the current review is its focus on financial
only two studies out of seven finding an effect size of greater aspects of poverty (e.g., income-based measures) rather than
than r = .10 when weight was measured in early childhood. social class (e.g., parental education, occupation). Studies
Additionally, some prior research suggests that younger were excluded from the review if the measure of “poverty”
children are better at self-regulating their energy intake did not include some aspect of income or deprivation at the
than older children (Johnson and Taylor-Holloway 2006), family or neighborhood level. Future researchers should pri-
with the beginning of middle childhood sometimes cited oritize education or social class-based measures of SES in
as a time in which children become less adept at modulat- the relationship to child SR and weight outcomes, as this was
ing their energy intake (Tsukayama et al. 2010). Therefore, outside of the scope of this review.
timing of measurement of SR of eating/appetite may be par- Another major limitation of the current systematic review
ticularly important, as EAH tasks administered prior to the and effect size analysis is that many studies were not able to
age of 7 may be less effective in predicting later adiposity. be included because no effect size was available. Also, par-
However, none of the studies in the current review tested ticularly for the poverty to weight and poverty to SR path-
EAH in middle childhood or adolescence. Randomized or ways, there was a great deal of variability in the definition
quasi-experimental studies where exogenous income is pro- and measurement of poverty, which resulted in the exclusion
vided to low-income families at various child ages could of some studies that may indeed have important implica-
prove especially useful for better understanding the effect of tions for understanding the relationship between poverty and
developmental timing of poverty on child weight outcomes. later outcomes. For example, as described in the Methods
section, studies were excluded if they tested trajectories of
Feasibility and Limitations of Proposed Model family income or chronicity of poverty because it would be
inappropriate to combine these effect sizes with those that
One important limitation is that the proposed model does not tested poverty exposure at a single assessment point. How-
consider the effect of poverty or poverty-related stressors on ever, studies that identify trajectories of family income over
specific child biological factors that may affect weight. For time (e.g., by categorizing families as “upwardly mobile” or
example, there is some evidence to suggest that children’s “downwardly mobile”) could be very helpful for identifying
exposure to chronic stress has consequences for metabolism whether there are sensitive developmental periods during
(Pervanidou and Chrousos 2011). It could be that poverty which child weight and SR are particularly vulnerable or
has both metabolic consequences and affects child SR, but malleable.
that it is the adverse effects on metabolism that drives the Another limitation is that, although we report the results
association between poverty and weight. from Fisher’s z tests assessing differences in correlation
13

coefficients across different aspects of the studies (e.g., hot are driving effects on child weight. Future studies should
versus cool SR in relation to child weight, differences in evaluate the hypothesis that the effect of SNAP on child
strength of association between poverty and weight in early weight operates through reducing FI, which could be inform-
childhood versus adolescence), there was little power to ative for further exploring a causal relationship between FI
detect these differences. For example, for the z test to be and weight.
significant with a difference between correlation coefficients
of 0.30 (an effect size of medium magnitude), there would
Parenting
need to be 60 studies in each group.
A promising mediator to explore in addition to or alongside
Future Directions
child SR in relation to weight may be parenting, as parenting
behaviors are known to be affected by poverty (Kiernan and
Future research must first determine whether there is suf-
Mensah 2011; McLeod and Shanahan 1993) explain some
ficient evidence across multiple studies with diverse sam-
of the relationship between poverty and SR (Lengua et al.
ples that SR mediates the relationship between poverty and
2014), and are predictive of child weight (Kakinami et al.
weight. If the model does indeed seem viable, a next step
2015). Few studies have assessed the association between
would be to explore the possible implications for interven-
poverty and parents’ feeding practices, although one cross-
tion. That is, by improving child SR, can the impact of early
sectional study found an association between FI and restric-
exposure to poverty on overweight/obesity be attenuated?
tive/pressuring feeding style (Gross et al. 2012). Similar to
SR, parenting may be a promising mediator to explore, as
Poverty and Food Insecurity
it is known to be malleable to intervention (Dishion et al.
2008). No longitudinal studies have previously tested
Future research should take advantage of quasi-experimental
whether SR mediates the relationship between parenting and
study designs where income was increased for some fami-
child weight, which would be an important area for future
lies but not others. For example, the effect of the Earned
research.
Income Tax Credit (EITC) on BMI has been tested in adults
(Schmeiser 2009), but has not yet been evaluated in children.
A quasi-experimental study where researchers evaluated the Gender
effect of a conditional cash transfer program on Mexican
children’s BMI found evidence for a beneficial effect of Although child gender was not tested as a moderator in
income increases in reducing the prevalence of child over- the relationship between poverty and weight in the current
weight (Fernald et al. 2008), although this study was not analysis, several studies that were included in the review
longitudinal. found some evidence that the association between pov-
More research on the impact of FI on child weight is erty and overweight/obesity may be more pronounced for
necessary, with quasi-experimental studies having the female children. More studies should test the moderating
potential to be particularly informative. Programs exist that effect of gender, as it could be that the stronger relationship
could theoretically reduce children’s exposure to FI by pro- between poverty and weight observed for girls is driving
viding additional nutrition. In fact, the USDA states that what appears to be a general effect of poverty on obesity for
the Supplemental Nutrition Assistance Program (SNAP) is all children. Child gender would be particularly important to
“designed to reduce food insecurity,” and there is indeed consider as a moderator when the outcome for weight is in
some evidence that it does (Bhattacharya and Currie 2001). adolescence, as puberty impacts boys and girls dissimilarly
However, there is mixed evidence for the effect of SNAP on at different ages.
child obesity, with some research suggesting that it may be
associated with an increase in obesity among certain popu-
Specificity of Outcome of Child Weight in Relation
lations, such as girls in middle childhood (Gibson 2004).
to Poverty/SR
Research by Schmeiser (2012), however, suggests that a fam-
ily’s participation in SNAP is associated with a reduction in
As SR is known to be associated with a variety of child out-
BMI and overweight/obesity for boys 5–18 and girls 5–11,
comes, it is unclear the extent to which the proposed model
but not girls from 12 to 18. Yet another study found that
would be specific to child weight as an outcome, or whether
Head Start children whose families participated in SNAP
one of several other outcomes (e.g., behavior problems, aca-
were no more or less likely to show increases in BMI over
demic achievement, delinquency) could replace it. Future
one year (Simmons et al. 2012). However, none of these
research should test this possibility.
studies directly measured FI, so there is no way to know
whether reductions in FI attributed to SNAP participation
13
Obesity is known to be comorbid with attention defi- Compliance with Ethical Standards
cit hyperactivity disorder (ADHD) (Cortese et al. 2008;
Cortese and Castellanos 2014), perhaps because of simi- Ethical Approval This article does not contain any studies with human
lar underlying mechanisms of difficulties with SR. One participants performed by any of the authors.
interesting question for future research might be to assess
whether ADHD treatment, which should improve atten-
tion and impulsivity, also has a beneficial effect on child
weight. Some research suggests that children medicated References
for ADHD have reduced risk of obesity compared with
children with ADHD who are not taking medication Abenavoli, R. M., Greenberg, M. T., & Bierman, K. L. (2015). Parent
support for learning at school entry: benefits for aggressive chil-
(Kim et al. 2011; Waring and Lapane 2008). However, it
dren in high-risk urban contexts. Early childhood research quar-
is important to note that another study failed to replicate terly, 31, 9–18. https://doi.org/10.1016/j.ecresq.2014.12.003.
this effect (Byrd et al. 2013), and in none of these studies Alaimo, K., Olson, C. M., & Frongillo, E. A. (2001). Low family
were children randomly assigned to ADHD medication income and food insufficiency in relation to overweight in US
children: is there a paradox? Archives of pediatrics & adolescent
treatment. Another potential confound is that simulant
medicine, 155(10), 1161–1167.
medication is known to suppress appetite, so this could Anderson, S., Sacker, A., Whitaker, R., & Kelly, Y. (2017). Self-
be driving the protective effects on weight rather than regulation and household routines at age three and obesity at
improvements to SR. The effect of psychosocial, non- age eleven: longitudinal analysis of the UK Millennium Cohort
Study. International Journal of Obesity, 41(10), 1459.
stimulant ADHD treatment on child weight, potentially
Anzman, S. L., & Birch, L. L. (2009). Low inhibitory control and
mediated by improvements to SR, would be an interesting restrictive feeding practices predict weight outcomes. The Jour-
area for further research. nal of pediatrics, 155(5), 651–656.
Arenz, S., Rückerl, R., Koletzko, B., & von Kries, R. (2004). Breast-
feeding and childhood obesity—a systematic review. Interna-
tional Journal of Obesity, 28(10), 1247.
Assari, S., Thomas, A., Caldwell, C. H., & Mincy, R. B. (2018).
Conclusions Blacks’ diminished health return of family structure and socio-
economic status; 15 years of follow-up of a national urban sam-
ple of youth. Journal of Urban Health, 95(1), 21–35.
Findings from the studies included in the current review
Asta, K., Miller, A. L., Retzloff, L., Rosenblum, K., Kaciroti, N. A., &
suggest meaningful associations between poverty and Lumeng, J. C. (2016). Eating in the absence of hunger and weight
child SR and between SR and child weight. However, gain in low-income toddlers. Pediatrics, 137(5), e20153786.
more research is warranted to better understand specific https://doi.org/10.1542/peds.2015-3786.
Baldwin, J. R., Arseneault, L., Odgers, C., Belsky, D. W., Matthews, T.,
factors that might underlie the association between pov-
Ambler, A.,.. . Danese, A. (2016). Childhood bullying victimi-
erty and child weight, especially based on the potential zation and overweight in young adulthood: a cohort study. Psy-
implications for prevention and treatment. Although SR chosomatic medicine, 78(9), 1094–1103. https: //doi.org/10.1097/
may be a promising potential target for obesity interven- PSY.0000000000000388.
Baptiste-Roberts, K., Nicholson, W. K., Wang, N.-Y., & Brancati,
tion for low-income children, additional research on how
F. L. (2012). Gestational diabetes and subsequent growth pat-
SR affects risk of obesity is crucial, especially based on terns of offspring: The national collaborative perinatal project.
the lack of success of the limited number of SR-promoting Maternal and Child Health Journal, 16(1), 125–132. https://doi.
interventions for improving children’s weight outcomes. In org/10.1007/s10995-011-0756-2.
Bassett, H. H., Denham, S., Wyatt, T. M., & Warren-Khot, H. K.
particular, aspects of SR that have traditionally been lim-
(2012). Refining the preschool self-regulation assessment for use
ited to the obesity literature (i.e., eating-specific SR) need in preschool classrooms. Infant and Child Development, 21(6),
to be tested in the context of poverty. Different aspects of 596–616. https://doi.org/10.1002/icd.1763.
SR (i.e., hot, cool, physiological, eating-specific) should Bergmeier, H., Skouteris, H., Horwood, S., Hooley, M., & Richardson,
B. (2014). Associations between child temperament, maternal
be tested in additional studies as mediators in the relation-
feeding practices and child body mass index during the preschool
ship between poverty and child weight, which will allow years: a systematic review of the literature. Obesity Reviews,
the weaknesses and limitations of the current model to 15(1), 9–18.
be tested systematically before being incorporated into Bernier, A., Beauchamp, M. H., Carlson, S. M., & Lalonde, G. (2015).
A secure base from which to regulate: attachment security in
interventions. In addition, a focus on developmental status
toddlerhood as a predictor of executive functioning at school
should be incorporated into future research on associations entry. Developmental Psychology, 51(9), 1177–1189. https: //doi.
between poverty, SR, and child weight. org/10.1037/dev0000032.
Bernier, A., Carlson, S. M., Bordeleau, S., & Carrier, J. (2010). Rela-
tions between physiological and cognitive regulatory systems:
infant sleep regulation and subsequent executive functioning.
Child Development, 81(6), 1739–1752. https://doi.org/10.111
1/j.1467-8624.2010.01507.x.
13

Berry, D., Blair, C., Ursache, A., Willoughby, M. T., & Granger, D. polymorphism. JAMA Pediatrics, 171(1), 61–67. https://doi.
A. (2014). & Family Life Project Key. Early childcare, execu- org/10.1001/jamapediatrics.2016.2332.
tive functioning, and the moderating role of early stress physiol- Byrd, H. M., Curtin, C., & Anderson, S. E. (2013). Attention-deficit/
ogy. Developmental Psychology, 50(4), 1250–1261. https://doi. hyperactivity disorder and obesity in US males and females, age
org/10.1037/a0034700. I. 8–15 years: National Health and Nutrition Examination Survey
Bhargava, A., Jolliffe, D., & Howard, L. L. (2008). Socio-economic, 2001–2004. Pediatric obesity, 8(6), 445–453.
behavioural and environmental factors predicted body weights Calkins, S. D. (2007). The emergence of self-regulation: Biological and
and household food insecurity scores in the early childhood behavioral control mechanisms supporting toddler competencies.
longitudinal study-Kindergarten. British Journal of Nutrition, Socioemotional development in the toddler years: Transitions
100(2), 438–444. https://doi.org/10.1017/S0007114508894366. and transformations, 261–284.
Bhattacharya, J., & Currie, J. (2001). Youths at nutrition risk: mal- Calkins, S. D., & Dedmon, S. E. (2000). Physiological and behavioral
nourished or misnourished? In Risky behavior among youths: regulation in two-year-old children with aggressive/destructive
an economic analysis (pp. 483–522): Chicago: University of behavior problems. Journal of abnormal child psychology, 28(2),
Chicago Press. 103–118.
Bleich, S. N., Segal, J., Wu, Y., Wilson, R., & Wang, Y. (2013). Sys- Cameron, A. J., Spence, A. C., Laws, R., Hesketh, K. D., Lioret, S.,
tematic review of community-based childhood obesity prevention & Campbell, K. J. (2015). A review of the relationship between
studies. Pediatrics, 132(1), e201–e210. socioeconomic position and the early-life predictors of obesity.
Blums, A., Belsky, J., Grimm, K., & Chen, Z. (2017). Building links Current obesity reports, 4(3), 350–362.
between early socioeconomic status, cognitive ability, and math Carper, J., Fisher, J. O., & Birch, L. L. (2000). Young girls’ emerging
and science achievement. Journal of Cognition and Development, dietary restraint and disinhibition are related to parental control
18(1), 16–40. https://doi.org/10.1080/15248372.2016.1228652. in child feeding. Appetite, 35(2), 121–129.
Bocknek, E. L., Brophy-Herb, H. E., Fitzgerald, H. E., Schiffman, R. Carter, M., Dubois, L., Tremblay, M., Taljaard, M., & Jones, B. (2012).
F., & Vogel, C. (2014). Stability of biological father presence Trajectories of childhood weight gain: the relative importance of
as a proxy for family stability: cross-racial associations with local environment versus individual social and early life factors.
the longitudinal development of emotion regulation in toddler- PloS ONE. 7(10). https: //doi.org/10.1371/journa l.pone.004706 5.
hood. Infant Mental Health Journal, 35(4), 309–321. https://doi. Cecil, J. E., Palmer, C. N., Wrieden, W., Murrie, I., Bolton-Smith, C.,
org/10.1002/imhj.21454. Watt, P.,.. . Hetherington, M. M. (2005). Energy intakes of chil-
Braet, C., Tanghe, A., Decaluwé, V., Moens, E., & Rosseel, Y. (2004). dren after preloads: adjustment, not compensation. The American
Inpatient treatment for children with obesity: weight loss, psy- journal of clinical nutrition, 82(2), 302–308.
chological well-being, and eating behavior. Journal of pediatric Chang, H., Shelleby, E. C., Cheong, J., & Shaw, D. S. (2012). Cumula-
psychology, 29(7), 519–529. tive risk, negative emotionality, and emotion regulation as pre-
Brophy-Herb, H. E., Zajicek-Farber, M. L., Bocknek, E. L., McKel- dictors of social competence in transition to school: a mediated
vey, L. M., & Stansbury, K. (2013). Longitudinal connections moderation model. Social Development, 21(4), 780–800. https
of maternal supportiveness and early emotion regulation to chil- ://doi.org/10.1111/j.1467-9507.2011.00648.x.
dren’s school readiness in low-income families. Journal of the Choi, J. Y., Castle, S., Williamson, A. C., Young, E., Worley, L., Long,
Society for Social Work and Research, 4(1), 2–19. https://doi. M., & Horm, D. M. (2016). Teacher-child interactions and the
org/10.5243/jsswr.2013.1. development of executive function in preschool-age children
Brotman, L. M., Dawson-McClure, S., Huang, K.-Y., Theise, R., Kam- attending Head Start. Early Education and Development, 27(6),
boukos, D., Wang, J., & Ogedegbe, G. (2012). Early childhood 751–769. https://doi.org/10.1080/10409289.2016.1129864.
family intervention and long-term obesity prevention among Cohen, J. (1992). A power primer. Psychological bulletin, 112(1), 155.
high-risk minority youth. Pediatrics, 129(3), e621–e628. Coleman-Jensen, A., Gregory, C., & Singh, A. (2014). Household food
Brown, E. D., Ackerman, B. P., & Moore, C. A. (2013). Family adver- security in the United States in 2013. Washington, D.C.: Eco-
sity and inhibitory control for economically disadvantaged chil- nomic Research Service.
dren: Preschool relations and associations with school readiness. Coley, R. L., & Lombardi, C. M. (2012). Early maternal employment
Journal of Family Psychology, 27(3), 443. and childhood obesity among economically disadvantaged fami-
Brown, E. D., & Sax, K. L. (2013). Arts enrichment and preschool lies in the USA. Early Child Development and Care, 182(8),
emotions for low-income children at risk. Early childhood 983–998. https://doi.org/10.1080/03004430.2012.678594.
research quarterly, 28(2), 337–346. https://doi.org/10.1016/j. Collings, P. J., Ball, H. L., Santorelli, G., West, J., Barber, S. E.,
ecresq.2012.08.002. McEachan, R. R., & Wright, J. (2017). Sleep duration and adi-
Brown, T., & Summerbell, C. (2009). Systematic review of school- posity in early childhood: Evidence for bidirectional associa-
based interventions that focus on changing dietary intake and tions from the Born in Bradford Study. Sleep. 40(2). https://doi.
physical activity levels to prevent childhood obesity: an update org/10.1093/sleep/zsw054.
to the obesity guidance produced by the National Institute Conradt, E., Abar, B., Sheinkopf, S., Lester, B., Lagasse, L., Seifer, R.,
for Health and Clinical Excellence. Obesity Reviews, 10(1), & Higgins, R. (2014). The role of prenatal substance exposure
110–141. and early adversity on parasympathetic functioning from 3 to
Bub, K. L., Robinson, L. E., & Curtis, D. S. (2016). Longitudinal asso- 6 years of age. Developmental Psychobiology, 56(4), 821–835.
ciations between self-regulation and health across childhood and https://doi.org/10.1002/dev.21155.
adolescence. Health Psychology, 35(11), 1235–1245. https: //doi. Cortese, S., Angriman, M., Maffeis, C., Isnard, P., Konofal, E., Lecen-
org/10.1037/hea0000401. dreux, M., & Mouren, M.-C. (2008). Attention-deficit/hyperac-
Burdette, H. L., & Whitaker, R. C. (2007). Differences by race and tivity disorder (ADHD) and obesity: a systematic review of the
ethnicity in the relationship between breastfeeding and obesity in literature. Critical reviews in food science and nutrition, 48(6),
preschool children. Ethnicity and Disease, 17(3), 467. 524–537.
Bush, N. R., Allison, A. L., Miller, A. L., Deardorff, J., Adler, N. Cortese, S., & Castellanos, F. X. (2014). The relationship between
E., & Boyce, W. T. (2017). Socioeconomic disparities in ADHD and obesity: implications for therapy. Expert review of
childhood obesity risk: association with an oxytocin receptor neurotherapeutics, 14(5), 473–479.
13
Crook, S. R., & Evans, G. W. (2014). The role of planning skills in the Eisenberg, N., & Zhou, Q. (2016). Conceptions of executive function
income–achievement gap. Child Development, 85(2), 405–411. and regulation: when and to what degree do they overlap? In J.
https://doi.org/10.1111/cdev.12129. A. Griffin, P. McCardle & L. S. Freund (Eds.), Executive function
Daches, S., Kovacs, M., George, C. J., Yaroslavsky, I., Kiss, E., Vetró, in preschool-age children: integrating measurement, neurode-
Á, & Halas, K. (2017). Childhood adversity predicts reduced velopment, and translational research (pp. 115–136). Washing-
physiological flexibility during the processing of negative affect ton, DC: American Psychological Association. Chapter xi, 362
among adolescents with major depression histories. International Pages).
Journal of Psychophysiology, 121, 22–28. Emery, R. L., & Levine, M. D. (2017). Questionnaire and behavioral
Datar, A., & Chung, P. J. (2018). Childhood self-control and adolescent task measures of impulsivity are differentially associated with
obesity: evidence from longitudinal data on a national cohort. body mass index: a comprehensive meta-analysis. Psychological
Childhood Obesity, 14(4), 238–247. https://doi.org/10.1089/ bulletin, 143(8), 868.
chi.2017.0217. Eneli, I. U., Tylka, T. L., Hummel, J., Watowicz, R. P., Perez, S. A.,
Dearing, E., McCartney, K., & Taylor, B. A. (2006). Within-child asso- Kaciroti, N., & Lumeng, J. C. (2015). Rationale and design of
ciations between family income and externalizing and internal- the Feeding Dynamic Intervention (FDI) study for self-regulation
izing problems. Developmental Psychology, 42(2), 237. of energy intake in preschoolers. Contemporary clinical trials,
Degol, J. L., & Bachman, H. J. (2015). Preschool teachers’ classroom 41, 325–334.
behavioral socialization practices and low-income children’s self- Epstein, L. H., Valoski, A., Wing, R. R., & McCurley, J. (1990). Ten-
regulation skills. Early childhood research quarterly, 31, 89–100. year follow-up of behavioral, family-based treatment for obese
https://doi.org/10.1016/j.ecresq.2015.01.002. children. JAMA, 264(19), 2519–2523.
Dietz, W. H. (1994). Critical periods in childhood for the development Epstein, L. H., Wing, R. R., Koeske, R., Andrasik, F., & Ossip, D. J.
of obesity. The American journal of clinical nutrition, 59(5), (1981). Child and parent weight loss in family-based behavior
955–959. modification programs. Journal of Consulting and Clinical Psy-
Dilworth-Bart, J. E., Khurshid, A., & Vandell, D. L. (2007). Do mater- chology, 49(5), 674.
nal stress and home environment mediate the relation between Evans, G. W., & English, K. (2002). The environment of poverty:
early income-to-need and 54-months attentional abilities? Infant Multiple stressor exposure, psychophysiological stress, and soci-
and Child Development, 16(5), 525–552. https: //doi.org/10.1002/ oemotional adjustment. Child development, 73(4), 1238–1248.
icd.528. Evans, G. W., Fuller-Rowell, T. E., & Doan, S. N. (2012). Childhood
Dinsa, G. D., Goryakin, Y., Fumagalli, E., & Suhrcke, M. (2012). Obe- cumulative risk and obesity: the mediating role of self-regulatory
sity and socioeconomic status in developing countries: a system- ability. Pediatrics, 129(1), e68–e73.
atic review. Obesity Reviews, 13(11), 1067–1079. Evans, G. W., Gonnella, C., Marcynyszyn, L. A., Gentile, L., & Sal-
Dishion, T., Shaw, D. S., Connell, A. M., Gardner, F., Weaver, C., & pekar, N. (2005). The role of chaos in poverty and children’s
Wilson, M. (2008). The family check-up with high-risk indigent socioemotional adjustment. Psychological Science, 16(7), 560–
families: preventing problem behavior by increasing parents’ 565. https://doi.org/10.1111/j.0956-7976.2005.01575.x.
positive behavior support in early childhood. Child development, Evans, G. W., & Kim, P. (2007). Childhood poverty and health cumula-
79, 1395–1414. https: //doi.org/10.1111/j.1467-8624.2008.01195 tive risk exposure and stress dysregulation. Psychological Sci-
.x. ence, 18(11), 953–957.
Doan, S. N., Fuller-Rowell, T. E., & Evans, G. W. (2012). Cumulative Fernald, L. C., Gertler, P. J., & Neufeld, L. M. (2008). Role of cash in
risk and adolescent’s internalizing and externalizing problems: conditional cash transfer programmes for child health, growth,
the mediating roles of maternal responsiveness and self-regula- and development: an analysis of Mexico’s Oportunidades. The
tion. Developmental Psychology, 48(6), 1529–1539. https://doi. lancet, 371(9615), 828–837.
org/10.1037/a0027815. Fogel, A., McCrickerd, K., Fries, L. R., Goh, A. T., Quah, P. L., Chan,
Drewnowski, A., & Specter, S. (2004). Poverty and obesity: the role M. J., & Forde, C. G. (2018). Eating in the absence of hunger:
of energy density and energy costs. The American journal of Stability over time and associations with eating behaviours and
clinical nutrition, 79(1), 6–16. body composition in children. Physiology & Behavior, 192,
Duckworth, A. L., Tsukayama, E., & Geier, A. B. (2010). Self-con- 82–89. https://doi.org/10.1016/j.physbeh.2018.03.033.
trolled children stay leaner in the transition to adolescence. Appe- Forehand, R., Biggar, H., & Kotchick, B. A. (1998). Cumulative risk
tite, 54(2), 304–308. across family stressors: short-and long-term effects for adoles-
Duncan, G. J., & Magnuson, K. (2013). The long reach of early child- cents. Journal of abnormal child psychology, 26(2), 119–128.
hood poverty. In Economic stress, human capital, and families Francis, L. A., Granger, D., & Susman, E. J. (2013). Adrenocortical
in Asia (pp. 57–70): New York: Springer. regulation, eating in the absence of hunger and BMI in young
Eamon, M. K. (2002). Poverty, parenting, peer, and neighborhood children. Appetite, 64, 32–38.
influences on young adolescent antisocial behavior. Journal of Francis, L. A., & Susman, E. J. (2009). Self-regulation and rapid
Social Service Research, 28(1), 1–23. weight gain in children from age 3 to 12 years. Archives of pedi-
Ebbeling, C. B., Pawlak, D. B., & Ludwig, D. S. (2002). Childhood atrics & adolescent medicine, 163(4), 297–302.
obesity: public-health crisis, common sense cure. The lancet, Friedlander, S. L., Larkin, E. K., Rosen, C. L., Palermo, T. M., &
360(9331), 473–482. Redline, S. (2003). Decreased quality of life associated with obe-
Economos, C. D., Hyatt, R. R., Goldberg, J. P., Must, A., Naumova, E. sity in school-aged children. Archives of pediatrics & adolescent
N., Collins, J. J., & Nelson, M. E. (2007). A community interven- medicine, 157(12), 1206–1211.
tion reduces BMI z-score in children: shape up somerville first Gartstein, M. A., Bridgett, D. J., Young, B. N., Panksepp, J., & Power,
year results. Obesity, 15(5), 1325–1336. T. (2013). Origins of effortful control: Infant and parent con-
Eisenberg, N., Vidmar, M., Spinrad, T. L., Eggum, N. D., Edwards, A., tributions. Infancy, 18(2), 149–183. https: //doi.org/10.111
Gaertner, B., & Kupfer, A. (2010). Mothers’ teaching strategies 1/j.1532-7078.2012.00119.x.
and children’s effortful control: a longitudinal study. Develop- Gibbs, B. G., & Forste, R. (2014). Socioeconomic status, infant feeding
mental Psychology, 46(5), 1294–1308. https://doi.org/10.1037/ practices and early childhood obesity. Pediatric obesity, 9(2),
a0020236. 135–146.
13

Gibson, D. (2004). Long-term food stamp program participation is Israel, A. C., Guile, C. A., Baker, J. E., & Silverman, W. K. (1994).
differentially related to overweight in young girls and boys. The An evaluation of enhanced self-regulation training in the treat-
Journal of Nutrition, 134(2), 372–379. ment of childhood obesity. Journal of pediatric psychology,
Goossens, L., Braet, C., Van Vlierberghe, L., & Mels, S. (2009). Loss 19(6), 737–749.
of control over eating in overweight youngsters: the role of anxi- Janjua, N. Z., Mahmood, B., Islam, M. A., & Goldenberg, R. L.
ety, depression and emotional eating. European Eating Disorders (2012). Maternal and early childhood risk factors for over-
Review, 17(1), 68–78. weight and obesity among low-income predominantly
Graziano, P. A., Calkins, S. D., & Keane, S. P. (2010). Toddler self- black children at age five years: A prospective cohort study.
regulation skills predict risk for pediatric obesity. International Journal of Obesity, 2012, 457173–457179. https : //doi.
Journal of Obesity, 34(4), 633–641. org/10.1155/2012/457173.
Graziano, P. A., Kelleher, R., Calkins, S. D., Keane, S. P., & Brien, Johnson, S. L. (2000). Improving preschoolers’ self-regulation of
M. O. (2013). Predicting weight outcomes in preadolescence: energy intake. Pediatrics, 106(6), 1429–1435.
the role of toddlers’ self-regulation skills and the temperament Johnson, S. L., & Taylor-Holloway, L. A. (2006). Non-Hispanic white
dimension of pleasure. International Journal of Obesity, 37(7), and Hispanic elementary school children’s self-regulation of
937–942. https://doi.org/10.1038/ijo.2012.165. energy intake. The American journal of clinical nutrition, 83(6),
Grineski, S. E., Morales, D. X., Collins, T. W., & Rubio, R. (2018). 1276–1282.
Transitional dynamics of household food insecurity impact chil- Kakinami, L., Barnett, T. A., Séguin, L., & Paradis, G. (2015). Parent-
dren’s developmental outcomes. Journal of Developmental and ing style and obesity risk in children. Preventive medicine, 75,
Behavioral Pediatrics, 39(9), 715–725. https://doi.org/10.1097/ 18–22.
DBP.0000000000000598. Kidwell, S. L., & Barnett, D. (2007). Adaptive emotion regula-
Gross, R. S., Mendelsohn, A. L., Fierman, A. H., Racine, A. D., & tion among low-income African American children. Merrill-
Messito, M. J. (2012). Food insecurity and obesogenic maternal Palmer Quarterly, 53(2), 155–183. https://doi.org/10.1353/
infant feeding styles and practices in low-income families. Pedi- mpq.2007.0011.
atrics, 130(2), 254–261. Kiernan, K. E., & Mensah, F. K. (2011). Poverty, family resources and
Gueron-Sela, N., Camerota, M., Willoughby, M. T., Vernon-Feagans, children’s early educational attainment: the mediating role of par-
L., & Cox, M. J., & Family Life Project Key (2018). Maternal enting. British Educational Research Journal, 37(2), 317–336.
depressive symptoms, mother-child interactions, and children’s Kim, J., Mutyala, B., Agiovlasitis, S., & Fernhall, B. (2011). Health
executive function. Developmental Psychology, 54(1), 71–82. behaviors and obesity among US children with attention deficit
https://doi.org/10.1037/dev0000389. I. hyperactivity disorder by gender and medication use. Preventive
Gundersen, C., Lohman, B. J., Eisenmann, J. C., Garasky, S., & Stew- medicine, 52(3), 218–222.
art, S. D. (2008). Child-specific food insecurity and overweight Kimbro, R. T., Brooks-Gunn, J., & McLanahan, S. (2007). Racial and
are not associated in a sample of 10–15 year-old low-income ethnic differentials in overweight and obesity among 3-year-old
youth. The Journal of Nutrition, 138(2), 371–378. children. American Journal of Public Health, 97(2), 298–305.
Hamilton, W. L., & Cook, J. T. (1997). Household food security in King, K. M., Lengua, L. J., & Monahan, K. C. (2013). Individual dif-
the United States in 1995: technical report of the food security ferences in the development of self-regulation during pre-adoles-
measurement project. cence: Connections to context and adjustment. Journal of Abnor-
Hancock, K. J., Lawrence, D., & Zubrick, S. R. (2014). Higher mater- mal Child Psychology, 41(1), 57–69. https://doi.org/10.1007/
nal protectiveness is associated with higher odds of child over- s10802-012-9665-0.
weight and obesity: a longitudinal Australian study. PLoS ONE, Kwok, M. K., Schooling, C. M., Lam, T. H., & Leung, G. M. (2010).
9(6), e100686. https://doi.org/10.1371/journal.pone.0100686. Paternal smoking and childhood overweight: evidence from the
Hartanto, A., Toh, W. X., & Yang, H. (2018). Bilingualism narrows Hong Kong “Children of 1997”. Pediatrics, 2009
socioeconomic disparities in executive functions and self-reg- Lamerz, A., Kuepper-Nybelen, J., Wehle, C., Bruning, N., Trost-
ulatory behaviors during early childhood: Evidence from the Brinkhues, G., Brenner, H., & Herpertz-Dahlmann, B. (2005).
Early Childhood Longitudinal Study. Child Development, 54(1), Social class, parental education, and obesity prevalence in a
71–82. https://doi.org/10.1111/cdev.13032. study of six-year-old children in Germany. International Journal
Ho, M., Garnett, S. P., Baur, L., Burrows, T., Stewart, L., Neve, M., of Obesity, 29(4), 373.
& Collins, C. (2012). Effectiveness of lifestyle interventions in Lane, S. P., Bluestone, C., & Burke, C. T. (2013). Trajectories of BMI
child obesity: systematic review with meta-analysis. Pediatrics, from early childhood through early adolescence: SES and psy-
130(6), e1647–e1671. chosocial predictors. British journal of health psychology, 18(1),
Hollar, D., Messiah, S. E., Lopez-Mitnik, G., Hollar, T. L., Almon, M., 66–82.
& Agatston, A. S. (2010). Effect of a two-year obesity prevention Lawlor, D. A., Ronalds, G., Macintyre, S., Clark, H., & Leon, D. A.
intervention on percentile changes in body mass index and aca- (2006). Family socioeconomic position at birth and future car-
demic performance in low-income elementary school children. diovascular disease risk: findings from the Aberdeen Children
American Journal of Public Health, 100(4), 646–653. of the 1950s cohort study. American Journal of Public Health,
Holm-Denoma, J. M., Smith, A., Lewinsohn, P. M., & Pettit, J. W. 96(7), 1271–1277.
(2014). Psychosocial predictors of body mass index at late child- Lawson, G. M., Hook, C. J., & Farah, M. J. (2018). A meta-analysis
hood: A longitudinal investigation. Journal of Health Psychol- of the relationship between socioeconomic status and executive
ogy, 19(6), 754–764. https://doi.org/10.1177/135910531347962 function performance among children. Developmental Science,
6. 21(2), e12529.
Hughes, S. O., Power, T. G., O’Connor, T. M., & Fisher, J. O. (2015). Lee, H., Andrew, M., Gebremariam, A., Lumeng, J. C., & Lee, J. M.
Executive functioning, emotion regulation, eating self-regulation, (2014). Longitudinal associations between poverty and obesity
and weight status in low-income preschool children: How do they from birth through adolescence. American Journal of Public
relate? Appetite, 89, 1–9. Health, 104(5), e70–e76.
Isasi, C. R., & Wills, T. A. (2011). Behavioral self-regulation and Lee, I., Bang, K.-S., Moon, H., & Kim, J. (2016). Comparison of
weight-related behaviors in inner-city adolescents: a model of obesity rates in early childhood (4 to 80 months) by paren-
direct and indirect effects. Childhood Obesity, 7(4), 306–315. tal socioeconomic status using national cohort dataset in
13
Korea. Asian Nursing Research, 10(4), 305–311. https://doi. achievement. Research in Human Development, 9(4), 278–297.
org/10.1016/j.anr.2016.10.005. https://doi.org/10.1080/15427609.2012.729907.
Lengua, L. J., Honorado, E., & Bush, N. R. (2007). Contextual risk McLeod, J. D., & Shanahan, M. J. (1993). Poverty, parenting, and chil-
and parenting as predictors of effortful control and social com- dren’s mental health. American Sociological Review, 351–366.
petence in preschool children. Journal of Applied Developmen- Mészáros, Z., Mészáros, J., Szmodis, M. B., Pampakas, P., Osváth, P.,
tal Psychology, 28(1), 40–55. https://doi.org/10.1016/j.appde & Völgyi, E. (2008). Primary school child development—Issues
v.2006.10.001. of socioeconomic status. Kinesiology, 40(2), 154–162. https://
Lengua, L. J., Kiff, C., Moran, L., Zalewski, M., Thompson, S., Cor- hrcak.srce.hr/30815.
tes, R., & Ruberry, E. (2014). Parenting mediates the effects Metcalfe, J., & Mischel, W. (1999). A hot/cool-system analysis of delay
of income and cumulative risk on the development of effort- of gratification: dynamics of willpower. Psychological Review,
ful control. Social Development, 23(3), 631–649. https://doi. 106(1), 3.
org/10.1111/sode.12071. Michels, N., Sioen, I., Braet, C., Eiben, G., Hebestreit, A., Huybrechts,
Lengua, L. J., Moran, L., Zalewski, M., Ruberry, E., Kiff, C., & I., & De Henauw, S. (2012). Stress, emotional eating behaviour
Thompson, S. (2015). Relations of growth in effortful con- and dietary patterns in children. Appetite, 59(3), 762–769.
trol to family income, cumulative risk, and adjustment in pre- Miller, A. L., Rosenblum, K. L., Retzloff, L. B., & Lumeng, J. C.
school-age children. Journal of Abnormal Child Psychology, (2016). Observed self-regulation is associated with weight in
43(4), 705–720. https://doi.org/10.1007/s10802-014-9941-2. low-income toddlers. Appetite, 105, 705–712.
Lengua, L. J., Zalewski, M., Fisher, P., & Moran, L. (2013). Does Mistry, R. S., Benner, A. D., Biesanz, J. C., Clark, S. L., & Howes, C.
HPA-Axis dysregulation account for the effects of income (2010). Family and social risk, and parental investments during
on effortful control and adjustment in preschool children?: the early childhood years as predictors of low-income children’s
Income, HPA-Axis dysregulation and effortful control. school readiness outcomes. Early childhood research quarterly,
Infant and Child Development, 22(5), 439–458. https: //doi. 25(4), 432–449. https://doi.org/10.1016/j.ecresq.2010.01.002.
org/10.1002/icd.1805. Moffitt, T. E., Arseneault, L., Belsky, D., Dickson, N., Hancox, R.
Lenhard, W., & Lenhard, A. (2016). Calculation of effect sizes. Biber- J., Harrington, H., & Ross, S. (2011). A gradient of childhood
gau: Psychometrica. https://www.psychometrica.de/effect_size. self-control predicts health, wealth, and public safety. Proceed-
html. ings of the National Academy of Sciences of the United States of
Leung, C. Y., Lumeng, J. C., Kaciroti, N. A., Chen, Y. P., Rosenblum, America, 108(7), 2693–2698.
K., & Miller, A. L. (2014). Surgency and negative affectivity, Nemet, D., Barkan, S., Epstein, Y., Friedland, O., Kowen, G., & Eli-
but not effortful control, are uniquely associated with obesogenic akim, A. (2005). Short-and long-term beneficial effects of a com-
eating behaviors among low-income preschoolers. Appetite, 78, bined dietary–behavioral–physical activity intervention for the
139–146. treatment of childhood obesity. Pediatrics, 115(4), e443–e449.
Li, C., Goran, M. I., Kaur, H., Nollen, N., & Ahluwalia, J. S. (2007). Nesbitt, K. T., Baker-Ward, L., & Willoughby, M. T. (2013). Execu-
Developmental trajectories of overweight during childhood: tive function mediates socio-economic and racial differences in
Role of early life factors. Obesity, 15(3), 760–771. https://doi. early academic achievement. Early childhood research quarterly,
org/10.1038/oby.2007.585. 28(4), 774–783. https://doi.org/10.1016/j.ecresq.2013.07.005.
Lim, S., Zoellner, J. M., Ajrouch, K. J., & Ismail, A. I. (2011). Over- Nievar, M. A., Moske, A. K., Johnson, D. J., & Chen, Q. (2014). Par-
weight in childhood: the role of resilient parenting in African- enting practices in preschool leading to later cognitive compe-
American households. American Journal of Preventive Medicine, tence: A family stress model. Early Education and Development,
40(3), 329–333. https://doi.org/10.1016/j.amepre.2010.11.006. 25(3), 318–337. https: //doi.org/10.1080/104092 89.2013.788426 .
Lumeng, J. C., Miller, A. L., Horodynski, M. A., Brophy-Herb, H. Ogden, C. L., Carroll, M. D., Kit, B. K., & Flegal, K. M. (2012).
E., Contreras, D., Lee, H., & Peterson, K. E. (2017). Improving Prevalence of obesity and trends in body mass index among US
self-regulation for obesity prevention in head start: A randomized children and adolescents, 1999–2010. JAMA, 307(5), 483–490.
controlled trial. Pediatrics, 139(5), e20162047. Ogden, C. L., Carroll, M. D., Lawman, H. G., Fryar, C. D., Kruszon-
Mallan, K. M., Nambiar, S., Magarey, A. M., & Daniels, L. A. (2014). Moran, D., Kit, B. K., & Flegal, K. M. (2016). Trends in obesity
Satiety responsiveness in toddlerhood predicts energy intake and prevalence among children and adolescents in the United States,
weight status at four years of age. Appetite, 74, 79–85. 1988–1994 through 2013–2014. JAMA, 315(21), 2292–2299.
Marcovitch, S., O’Brien, M., Calkins, S. D., Leerkes, E. M., Weaver, Oken, E., Levitan, E., & Gillman, M. (2008). Maternal smoking during
J. M., & Levine, D. W. (2015). A longitudinal assessment of the pregnancy and child overweight: systematic review and meta-
relation between executive function and theory of mind at 3, analysis. International Journal of Obesity, 32(2), 201.
4, and 5 years. Cognitive Development, 33, 40–55. https://doi. Pandey, A., Hale, D., Das, S., Goddings, A.-L., Blakemore, S.-J., &
org/10.1016/j.cogdev.2014.07.001. Viner, R. M. (2018). Effectiveness of universal self-regulation–
Marshall, S. J., Biddle, S. J., Gorely, T., Cameron, N., & Murdey, I. based interventions in children and adolescents: a systematic
(2004). Relationships between media use, body fatness and phys- review and meta-analysis. JAMA pediatrics, 172(6), 566–575.
ical activity in children and youth: a meta-analysis. International Parkinson, K. N., Drewett, R. F., Le Couteur, A. S., & Adamson, A.
Journal of Obesity, 28(10), 1238. J., & Gateshead Milennium Study Core, T. (2010). Do maternal
Martoccio, T. L., Brophy-Herb, H. E., & Onaga, E. E. (2014). Road to ratings of appetite in infants predict later Child Eating Behav-
readiness: Pathways from low-income childrenʼs early interac- iour Questionnaire scores and body mass index? Appetite. 54(1),
tions to school readiness skills. Infants & Young Children, 27(3), 186–190. https://doi.org/10.1016/j.appet.2009.10.007.
193–206. https://doi.org/10.1097/IYC.0000000000000014. Pearce, A., Sawyer, A. C. P., Chittleborough, C. R., Mittinty, M. N.,
McCallum, Z., Wake, M., Gerner, B., Baur, L., Gibbons, K., Gold, L., Law, C., & Lynch, J. W. (2016). Do early life cognitive ability
& Riess, C. (2007). Outcome data from the LEAP (Live, Eat and self-regulation skills explain socio-economic inequalities
and Play) trial: a randomized controlled trial of a primary care in academic achievement? An effect decomposition analysis in
intervention for childhood overweight/mild obesity. International UK and Australian cohorts. Social Science & Medicine, 165,
Journal of Obesity, 31(4), 630–636. 108–118. https://doi.org/10.1016/j.socscimed.2016.07.016.
McClelland, M. M., & Wanless, S. B. (2012). Growing up with assets Peredo, T. N., Owen, M. T., Rojas, R., & Caughy, M. O. B. (2015).
and risks: The importance of self-regulation for academic Child vocabulary, maternal behavior, and inhibitory control
13

development among Spanish-speaking children. Early Edu- developmental psychopathology. Development and psychopa-
cation and Development, 26(5–6), 749–769. https : //doi. thology, 9(02), 335–364.
org/10.1080/10409289.2015.1009319. Sabol, T. J., Bohlmann, N. L., & Downer, J. T. (2018). Low-income
Pervanidou, P., & Chrousos, G. P. (2011). Stress and obesity/metabolic ethnically diverse children’s engagement as a predictor of school
syndrome in childhood and adolescence. Pediatric obesity, 6(S1), readiness above preschool classroom quality. Child Development,
21–28. 89(2), 556–576. https://doi.org/10.1111/cdev.12832.
Piché, G., Fitzpatrick, C., & Pagani, L. S. (2012). Kindergarten self- Sallis, J. F., & Glanz, K. (2006). The role of built environments in
regulation as a predictor of body mass index and sports participa- physical activity, eating, and obesity in childhood. The Future
tion in fourth grade students. Mind, Brain, and Education, 6(1), of Children, 89–108.
19–26. Schmeiser, M. D. (2009). Expanding wallets and waistlines: the impact
Porges, S. W. (2007). The polyvagal perspective. Biological Psychol- of family income on the BMI of women and men eligible for
ogy, 74(2), 116–143. the Earned Income Tax Credit. Health economics, 18(11),
Power, T. G., Olivera, Y. A., Hill, R. A., Beck, A. D., Hopwood, V., 1277–1294.
Garcia, K. S., & Hughes, S. O. (2016). Emotion regulation strate- Schmeiser, M. D. (2012). The impact of long-term participation in
gies and childhood obesity in high risk preschoolers. Appetite, the supplemental nutrition assistance program on child obesity.
107, 623–627. Health economics, 21(4), 386–404.
Pratt, M. E., McClelland, M. M., Swanson, J., & Lipscomb, S. T. Seeyave, D. M., Coleman, S., Appugliese, D., Corwyn, R. F., Bradley,
(2016). Family risk profiles and school readiness: A person-cen- R. H., Davidson, N. S., & Lumeng, J. C. (2009). Ability to delay
tered approach. Early childhood research quarterly, 36, 462–474. gratification at age 4 years and risk of overweight at age 11 years.
https://doi.org/10.1016/j.ecresq.2016.01.017. Archives of pediatrics & adolescent medicine, 163(4), 303–308.
Propper, C. (2012). The early development of vagal tone: Effects of Sektnan, M., McClelland, M. M., Acock, A., & Morrison, F. J. (2010).
poverty and elevated contextual risk. New York: Oxford Uni- Relations between early family risk, children’s behavioral regu-
versity Press. lation, and academic achievement. Early Childhood Research
Raver, C. C. (2004). Placing emotional self-regulation in sociocultural Quarterly, 25(4), 464–479.
and socioeconomic contexts. Child development, 75(2), 346–353. Sen, S., Rifas-Shiman, S. L., Shivappa, N., Wirth, M. D., Hebert, J.
Raver, C. C., Jones, S. M., Li-Grining, C., Zhai, F., Bub, K., & Pressler, R., Gold, D. R., & Oken, E. (2017). Associations of prenatal and
E. (2011). CSRP’s impact on low-income preschoolers’ preaca- early life dietary inflammatory potential with childhood adipos-
demic skills: self-regulation as a mediating mechanism. Child ity and cardiometabolic risk in Project Viva. Pediatric Obesity,
development, 82(1), 362–378. 13(5), 292–300. https://doi.org/10.1111/ijpo.12221.
Raver, C. C., McCoy, D. C., Lowenstein, A. E., & Pess, R. (2013). Pre- Shelleby, E. C., Shaw, D. S., Cheong, J., Chang, H., Gardner, F., Dish-
dicting individual differences in low-income children’s executive ion, T. J., & Wilson, M. N. (2012). Behavioral control in at-
control from early to middle childhood. Developmental Science, risk toddlers: The influence of the Family Check-Up. Journal of
16(3), 394–408. Clinical Child & Adolescent Psychology, 41(3), 288–301.
Razza, R. A., Martin, A., & Brooks-Gunn, J. (2010). Associations Shonkoff, J. P., Garner, A. S., Siegel, B. S., Dobbins, M. I., Earls,
among family environment, sustained attention, and school readi- M. F., McGuinn, L., & Care, D. (2012). The lifelong effects of
ness for low-income children. Developmental Psychology, 46(6), early childhood adversity and toxic stress. Pediatrics, 129(1),
1528–1542. https://doi.org/10.1037/a0020389. e232–e246.
Riggs, N. R., Greenberg, M. T., Kusché, C. A., & Pentz, M. A. (2006). Shrewsbury, V., & Wardle, J. (2008). Socioeconomic status and adipos-
The mediational role of neurocognition in the behavioral out- ity in childhood: a systematic review of cross-sectional studies
comes of a social-emotional prevention program in elementary 1990–2005. Obesity, 16(2), 275–284.
school students: Effects of the PATHS curriculum. Prevention Simmonds, M., Llewellyn, A., Owen, C., & Woolacott, N. (2016). Pre-
Science, 7(1), 91–102. dicting adult obesity from childhood obesity: a systematic review
Rochette, É, & Bernier, A. (2014). Parenting, family socioeco- and meta-analysis. Obesity Reviews, 17(2), 95–107.
nomic status, and child executive functioning: A longitudinal Simmons, S., Alexander, J. L., Ewing, H., & Whetzel, S. (2012).
study. Merrill-Palmer Quarterly, 60(4), 431–460. https://doi. SNAP Participation in Preschool-Aged Children and Prevalence
org/10.13110/merrpalmquar1982.60.4.0431. of Overweight and Obesity. Journal of School Health, 82(12),
Rolland-Cachera, M., Deheeger, M., Maillot, M., & Bellisle, F. (2006). 548–552.
Early adiposity rebound: causes and consequences for obesity Sitnick, S. L., Shaw, D. S., Weaver, C. M., Shelleby, E. C., Choe, D. E.,
in children and adults. International Journal of Obesity, 30, Reuben, J. D., & Taraban, L. (2017). Early Childhood Predictors
S11–S17. of Severe Youth Violence in Low-Income Male Adolescents.
Rothbart, M. K., Ahadi, S. A., Hershey, K. L., & Fisher, P. (2001). Child development, 88(1), 27–40.
Investigations of temperament at three to seven years: The Slopen, N., Kubzansky, L. D., McLaughlin, K. A., & Koenen, K. C.
Children’s Behavior Questionnaire. Child development, 72(5), (2012). Childhood adversity and inflammatory processes in
1394–1408. youth: A prospective study. Psychoneuroendocrinology, 38(2),
Roy, A. L., McCoy, D. C., & Raver, C. C. (2014). Instability ver- 188–200. https://doi.org/10.1016/j.psyneuen.2012.05.013.
sus quality: residential mobility, neighborhood poverty, and Smith, C., & Richards, R. (2008). Dietary intake, overweight status,
children’s self-regulation. Developmental Psychology. 50(7), and perceptions of food insecurity among homeless Minnesotan
1891–1896. youth. American Journal of Human Biology, 20(5), 550–563.
Roy, A. L., & Raver, C. C. (2014). Are all risks equal? Early experi- Smith, J. D., Montaño, Z., Dishion, T. J., Shaw, D. S., & Wilson, M.
ences of poverty-related risk and children’s functioning. Journal N. (2015). Preventing weight gain and obesity: Indirect effects
of Family Psychology, 28(3), 391–400. https://doi.org/10.1037/ of the family check-up in early childhood. Prevention Science,
a0036683. 16(3), 408–419.
Rutter, M., Dunn, J., Plomin, R., Simonoff, E., Pickles, A., Maughan, Starkey, L., & Revenson, T. A. (2014). Early Changes in Socioeco-
B., & Eaves, L. (1997). Integrating nature and nurture: Impli- nomic Status Do Not Predict Changes in Body Mass in the First
cations of person–environment correlations and interactions for Decade of Life. Annals of Behavioral Medicine, 49(2), 212–220.
13
Stautz, K., Pechey, R., Couturier, D.-L., Deary, I. J., & Marteau, T. M. childhood obesity using recursive partitioning based on individ-
(2016). Do executive function and impulsivity predict adolescent ual, familial, and neighborhood environment factors. The Inter-
health behaviour after accounting for intelligence? Findings from national Journal of Behavioral Nutrition and Physical Activity,
the ALSPAC Cohort. PLoS ONE, 11(8), e0160512. https://doi. 12(1), 17–17. https://doi.org/10.1186/s12966-015-0175-7.
org/10.1371/journal.pone.0160512. van Rossem, L., Silva, L. M., Hokken-Koelega, A., Arends, L. R.,
Steinsbekk, S., & Wichstrøm, L. (2015). Predictors of change in BMI Moll, H. A., Jaddoe, V. W., & Raat, H. (2010). Socioeconomic
from the age of 4 to 8. Journal of pediatric psychology, 40(10), status is not inversely associated with overweight in preschool
1056–1064. children. The Journal of Pediatrics, 157(6), 929–935.e921. https
Strauss, R. S., & Knight, J. (1999). Influence of the home environment ://doi.org/10.1016/j.jpeds.2010.06.008.
on the development of obesity in children. Pediatrics, 103(6), Verbeken, S., Braet, C., Goossens, L., & Van der Oord, S. (2013).
e85–e85. Executive function training with game elements for obese chil-
Strauss, R. S., & Pollack, H. A. (2003). Social marginalization of over- dren: A novel treatment to enhance self-regulatory abilities for
weight children. Archives of pediatrics & adolescent medicine, weight-control. Behaviour research and therapy, 51(6), 290–299.
157(8), 746–752. Vernon-Feagans, L., Willoughby, M., & Garrett-Peters, P. (2016).
Strobino, D. M., Ahmed, S., Mistry, K., Wasserman, E., Hossain, M. Predictors of behavioral regulation in kindergarten: Household
B., & Minkovitz, C. (2016). Maternal depressive symptoms and chaos, parenting, and early executive functions. Developmen-
attained size among children in the first 2 years of life. Academic tal Psychology, 52(3), 430–441. https://doi.org/10.1037/dev00
pediatrics, 16(1), 75–81. 00087.
Sturge-Apple, M. L., Suor, J. H., Davies, P. T., Cicchetti, D., Skibo, Votruba-Drzal, E. (2006). Economic disparities in middle childhood
M. A., & Rogosch, F. A. (2016). Vagal tone and children’s development: Does income matter? Developmental Psychology,
delay of gratification: Differential sensitivity in resource-poor 42(6), 1154.
and resource-rich environments. Psychological Science, 27(6), Wade, M., Madigan, S., Plamondon, A., Rodrigues, M., Browne, D.,
885–893. https://doi.org/10.1177/0956797616640269. & Jenkins, J. M. (2018). Cumulative psychosocial risk, parental
Sulik, M. J., Blair, C., Mills-Koonce, R., Berry, D., Greenberg, M., socialization, and child cognitive functioning: A longitudinal
& Vernon-Feagans, L.. The Family Life Project, I. (2015). cascade model. Developmental Psychology, 54(6), 1038–1050.
Early parenting and the development of externalizing behavior https://doi.org/10.1037/dev0000493.
problems: Longitudinal mediation through children’s executive Wanless, S. B., McClelland, M. M., Tominey, S. L., & Acock, A. C.
function. Child Development. 86(5), 1588–1603. https://doi. (2011). The influence of demographic risk factors on children’s
org/10.1111/cdev.12386. behavioral regulation in prekindergarten and kindergarten.
Suor, J. H., Sturge-Apple, M. L., & Jones-Gordils, H. R. (2018). Pars- Early Education & Development, 22(3), 461–488. https://doi.
ing profiles of temperamental reactivity and differential routes org/10.1080/10409289.2011.536132.
to delay of gratification: A person-based approach. Development Wardle, J., Guthrie, C. A., Sanderson, S., & Rapoport, L. (2001).
and Psychopathology, 1–20. https: //doi.org/10.1017/S09545 7941 Development of the children’s eating behaviour questionnaire.
7001894. The Journal of Child Psychology and Psychiatry and Allied Dis-
Tan, C. C., & Holub, S. C. (2010). Children’s self-regulation in eating: ciplines, 42(7), 963–970.
Associations with inhibitory control and parents’ feeding behav- Waring, M. E., & Lapane, K. L. (2008). Overweight in children and
ior. Journal of pediatric psychology, 36(3), 340–345. adolescents in relation to attention-deficit/hyperactivity disorder:
Tandon, P., Thompson, S., Moran, L., & Lengua, L. J. (2015). Body Results from a national sample. Pediatrics, 122(1), e1–e6.
mass index mediates the effects of low income on preschool Warschburger, P. (2015). SRT-Joy–computer-assisted self-regulation
children’s executive control, with implications for behavior and training for obese children and adolescents: Study protocol for a
academics. Childhood Obesity, 11(5), 569–576. https://doi. randomized controlled trial. Trials, 16(1), 566.
org/10.1089/chi.2014.0071. Webster-Stratton, C., Reid, J., M., & Stoolmiller, M. (2008). Preventing
Thompson, S. F., Lengua, L. J., Zalewski, M., & Moran, L. (2013). conduct problems and improving school readiness: Evaluation
Income and the development of effortful control as predictors of the incredible years teacher and child training programs in
of teacher reports of preschool adjustment. Early childhood high-risk schools. Journal of Child Psychology and Psychiatry,
research quarterly, 28(4), 784–793. https://doi.org/10.1016/j. 49(5), 471–488.
ecresq.2013.07.006. Whitaker, R. C., Wright, J. A., Pepe, M. S., Seidel, K. D., & Dietz, W.
Tiberio, S. S., Kerr, D. C. R., Capaldi, D. M., Pears, K. C., Kim, H. H. (1997). Predicting obesity in young adulthood from child-
K., & Nowicka, P. (2014). Parental monitoring of children’s hood and parental obesity. New England Journal of Medicine,
media consumption: The long-term influences on body mass 337(13), 869–873.
index in children. JAMA Pediatrics, 168(5), 414–421. https:// Williams, A. S., Ge, B., Petroski, G., Kruse, R. L., McElroy, J. A., &
doi.org/10.1001/jamapediatrics.2013.5483. Koopman, R. J. (2018). Socioeconomic status and other factors
Tran, M. K., Krueger, P. M., McCormick, E., Davidson, A., & Main, associated with childhood obesity. The Journal of the American
D. S. (2016). Body mass transitions through childhood and early Board of Family Medicine, 31(4), 514–521.
adolescence: a multistate life table approach. American journal Willoughby, M., Kupersmidt, J., Voegler-Lee, M., & Bryant, D. (2011).
of epidemiology, 183(7), 643–649. Contributions of hot and cool self-regulation to preschool disrup-
Trentacosta, C. J., & Shaw, D. S. (2009). Emotional self-regulation, tive behavior and academic achievement. Developmental neu-
peer rejection, and antisocial behavior: Developmental asso- ropsychology, 36(2), 162–180.
ciations from early childhood to early adolescence. Journal of
Applied Developmental Psychology, 30(3), 356–365. Publisher’s Note Springer Nature remains neutral with regard to
Tsukayama, E., Toomey, S. L., Faith, M. S., & Duckworth, A. L. jurisdictional claims in published maps and institutional affiliations.
(2010). Self-control as a protective factor against overweight
status in the transition from childhood to adolescence. Archives
of pediatrics & adolescent medicine, 164(7), 631–635.
Van Hulst, A., Roy-Gagnon, M.-H., Gauvin, L., Kestens, Y., Hender-
son, M., & Barnett, T. A. (2015). Identifying risk profiles for
13
Clinical Child & Family Psychology Review is a copyright of Springer, 2019. All Rights
Reserved.

Paper 1 PDF

Uploaded by

Copyright:

Available Formats

Paper 1 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Paper 1 PDF

Uploaded by

Copyright:

Available Formats

Clinical Child and Family Psychology Review (2019) 22:290–315

The Mediating Effect of Self-Regulation in the Association Between

Published online: 6 February 2019

Keywords Poverty · Obesity · BMI · Self-regulation · Food insecurity

Introduction and Pollack 2003). Unfortunately, research on child obesity

Table 1 Examples of observed self-regulation measures categorized as hot versus cool

Table 2 Studies with a positive Author Sample r (converted)

Table 3 Studies with a negative relationship or effect size < 0.10 for poverty-weight pathway

Mészáros et al. (2008) Hungarian sample, N = 495 − 0.04

Wanless et al. (2011) Community sample, N = 95 Cool − 0.26

Table 5 Studies with a negative relationship or effect size < 0.10 for poverty-SR pathway

Graziano et al. (2010) Community sample; N = 57 Cool; Hot − 0.14

Most studies (n = 13) measured SR beginning in early

Tapping task) associated with EF-training

At 3-year follow-up, ECI associated with

for the control group while EF-training

subscales of teacher-reported BRIEF

group remained stable (r = .27)

Intervention effects on weight

Three studies were reviewed that used a weight loss inter-

teacher-reported SR (average r = − .19)

EF-training group 8 weeks post-inter-

12-week f/u (r = .05 for 12-week f/u)

vention, but not maintained through

ments in teacher-reported SR com-

Better weight loss maintenance in

Standard treatment (ST) vs. Enhanced

Child Involvement (ECI), including

Executive Functioning (EF) training

SR, as measured by questionnaires. Only Verbeken and col-

versus care as usual

between the intervention, which included training in child

8–14 in an inpatient treatment

For calculating the aggregated effect size for the effect of

the interventions on child weight, if there were multiple fol-

ies having no or minimal effect on weight change. Israel and

colleagues (1994) found that their “enhanced child involve-

Israel et al. (1994)

compared to the standard treatment group (r = .30). Interest-

effects of the intervention on child weight at follow-up were

You might also like

Table 1 Examples of observed self-regulation measures categorized as hot versus cool

Table 2 Studies with a positive Author Sample r (converted)

Table 3 Studies with a negative relationship or effect size < 0.10 for poverty-weight pathway

Table 5 Studies with a negative relationship or effect size < 0.10 for poverty-SR pathway