Overweight or obesity associations with physical aggression in children and adolescents

Abstract

Being overweight or obese (overweight/obesity) or physically aggressive in childhood and adolescence can have lifelong consequences, hence are important public health problems. Identifying a relationship between these problems would assist in understanding their developmental origins. The present paper sought to review previous studies and use meta-analysis to evaluate whether there is evidence of a relationship between overweight/obesity and physical aggression in children and adolescents. A systematic search of studies that reported the effect of overweight/obesity (in the form of body mass index) on physical aggression was conducted. A total of 23 studies were identified, representing data from 255,377 participants. The results indicate that children and adolescents who are overweight or obese are more physically aggressive than their normal-weight or underweight peers. The average weighted standardized mean difference (the effect size) for aggression in overweight and obese children and adolescents compared to others was found to be 0.27 (95% confidence interval [CI₉₅]: .17–.37), and was significant (p<.001). Gender sub-analysis indicated that higher physical aggression amongst overweight or obese compared to normal-weight or underweight peers is a slightly larger effect for boys (standardized mean difference of .35, CI₉₅: .18–.52, p<.001) than girls (standardized mean difference of .24, CI₉₅: .07–.42, p<.01). High levels of heterogeneity (94.41%) were found between study-level effect sizes. The developmental processes that may explain the association between overweight/obesity and physical aggression in children and adolescents are discussed.

Keywords

Obesity overweight physical aggression bullying meta-analysis

Introduction

The incidence of overweight and obesity in childhood and adolescence has increased globally. In 2012, approximately 24% of boys and 27% of girls in Australia aged 5–17 years were classified as overweight or obese (Australian Bureau of Statistics, 2013; National Health and Medical Research Council, 2014). In the United States, childhood obesity has quadrupled in the last 30 years, with the percentage of children aged 6–11 years who were obese increasing from 7% in 1980 to nearly 18% in 2012, and the percentage of adolescents aged 12–19 years who were obese increasing from 5% to nearly 21% over the same period (Centers for Disease Control and Prevention, 2014; Loeber, 1990). Similar figures have been identified in other parts of the world, such as in parts of Europe, Scandinavia, Africa, and Asia (Child Obesity, 2015).

These high prevalence rates are concerning as overweight and obesity are associated with an increased risk of developing cardiovascular diseases (Centers for Disease Control and Prevention, 2014; Dietz, 2004), social difficulties (such as social isolation or teasing from peers) (Daniels et al., 2005), and psychological issues (such as depression) (Daniels et al., 2005). A considerable burden is also placed on the national health care system as a result. In Australia in 2008, the total annual amount of obesity-related costs (including costs associated with loss of productivity and carers’ fees) was estimated at AUD$58 billion (Access Economics, 2008), and in the United States in the same period was estimated to be US$254 billion, including US$208 billion in lost productivity, and US$46 billion in direct medical costs (American Heart Association, 2013).

It is also possible that obesity may be linked to broader child behavior problems, such as that of physical aggression. Physical aggression in children and adolescents can be a precursor to a variety of developmental problems, such as general physical health problems (Bor, McGee, Hayatbakhsh, Dean, & Najman, 2010), mental health problems such as mood, anxiety, and psychotic disorders (Fergusson & Woodward, 2000; Samuelson, Hodgins, Larsson, Larm, & Tengstrom, 2010), and participation in criminal activity (Fergusson & Woodward, 2000; Samuelson et al., 2010). Nationally representative large-scale studies in Australia and the United States have indicated that the proportion of children and adolescents (aged 12 to 18 years) who engage in physically aggressive behavior ranges from 18% to 33% (Bond, Thomas, Toumbourou, Patton, & Catalano, 2000; Eaton et al., 2012; Vassallo et al., 2002; Williams, Toumbourou, Williamson, Hemphill, & Patton, 2009).

Much of the research involving overweight and obese children and adolescents has been aimed at interventions targeting physical health. Although more limited, there have also been some investigations into associations between being obese or overweight (obesity/overweight) and having mental health and/or behavioral problems in childhood or adolescence. For example, a number of studies have explored associations between body mass index (BMI) and: psychological adjustment in children (Banis et al., 1988); interactions with peers (Baum & Forehand, 1984; Gallup & Wilson, 2009; Zeller, Reiter-Purtill, & Ramey, 2008); generalized behavior problems and bullying behavior (Datar & Sturm, 2004; Griffiths, Wolke, Page, & Horwood, 2006; Janssen, Craig, Boyce, & Picket, 2004; Griffiths et al., 2006); health-risk behaviors (Berg, Simonsson, & Ringqvist, 2005; Farhat, Iannotti, & Simons-Morton, 2010; Halfon, Larson, & Slusser, 2013; Pasch, Nelson, Lytle, Moe, & Perry, 2008; Berg, Simonsson, & Ringqvist, 2005); and internalizing and externalizing behaviors (Eschenbeck, Kohlmann, Dudey, & Schurholz, 2009; Hwang et al., 2006; Eschenbeck et al., 2009; Lumeng et al., 2003; Luukkonen, Räsänen, Hakko, Riala, & The STUDY-70 Workgroup, 2010; Mustillo et al., 2003; Piko, Keresztes, & Pluhar, 2006; Pitrou, Shojaei, Wazana, Gilbert, & Kovess-Masfety, 2010; Sawyer et al., 2006; Seyedamini, Malek, Ebrahimi-Mameghani, & Tajik, 2012).

There have only been a small number of studies that have reported and investigated the possibility of an effect of being overweight or obese on physical aggression (Gallup & Wilson, 2009; Ikäheimo et al., 2007; Janssen et al., 2004; Piko et al., 2005; Raine, Reynolds, Venables, Mednick, & Farrington, 1998; Tremblay et al., 1998). Based on significance testing, it appears that the results of many studies investigating this relationship have been mixed. Some research has identified a significant relationship between being overweight or obese and engaging in physically aggressive behaviors (Banis et al., 1988; Baum & Forehand, 1984; Berg et al., 2005; Datar & Sturm, 2004; Eschenbeck et al., 2009; Griffiths et al., 2006; Halfon et al., 2013; Hwang et al., 2006; Ikäheimo et al., 2007; Janssen et al., 2004; Lumeng et al., 2003; Piko et al., 2005; Raine et al., 1998; Sawyer et al., 2006; Seyedamini et al., 2012; Tremblay et al., 1998; Zeller et al., 2008). Other studies, however, have not found a significant relationship (Farhat et al., 2010; Gallup & Wilson, 2009; Luukkonen et al., 2010; Pasch et al., 2008; Pitrou et al., 2010). Analyses relating to any gender differences have also yielded inconsistent results. Some studies that collected data for both boys and girls have found a significant relationship between overweight/obesity and physical aggression in girls but not boys (Datar & Sturm, 2004; Janssen et al., 2004), whilst others have found a significant relationship between overweight/obesity and physical aggression in boys but not girls (Griffiths et al., 2006; Piko et al., 2005; Sawyer et al., 2006).

A number of theories have been proposed to explain why overweight/obesity may influence physical aggression in children and adolescents. These are outlined in the conceptual model in Figure 1 below.

Figure 1.

Theories of how overweight/obesity influences physical aggression in children and adolescents.

Social Dominance Model

Some authors have argued that this relationship exists in children and adolescents as a result of evolutionary utility. A social dominance perspective posits that, over time, increased height and weight resulted in reinforcement of the use of aggression as an effective strategy not only for winning social conflicts, but also the securing of material resources and mating opportunities (Gallup & Wilson, 2009; Piko et al., 2005; Raine et al., 1998; Volk et al., 2012). Gallup and Wilson (2009) further suggested that an association between overweight/obese and peer aggression may exist among females as a means to diminish the reputation and appeal of female peers, in an attempt to make other females less desirable to available males.

Socioeconomic Status (SES)

An alternative theory holds that childhood disadvantage may be a common cause of both physical health and behavioral problems. In particular, low SES (as measured by parental income) has been found to correlate with overweight and obesity in children and adolescents (Dinsa, Goryakin, Fumagalli, & Suhrcke, 2012; Eschenbeck et al., 2009; Farhat et al., 2010; Mustillo et al., 2003; Pitrou et al., 2010; Sawyer et al., 2006; Wang & Lim, 2012; Wu et al., 2015), but only in developed countries and not in developing countries. Dinsa et al. (2012) and Wu et al. (2015) suggest that individuals of low SES from developing countries face food shortages (which prevent them from becoming overweight/obese) while those in developed countries are particularly exposed to the marketing of cheap processed, energy-dense foods (Dinsa et al., 2012; Wu et al., 2015). In developed nations, nutrient-rich low calorie foods (such as wholegrain cereals, fruits, and vegetables) are less affordable for those of lower SES, which would result in the consumption of more energy-dense foods that are more affordable (such as fast foods) (Dinsa et al., 2012; Wu et al., 2015). In addition, higher degrees of urbanization and technological progress (including increases in information dissemination) in developed nations render occupations less laborious, resulting in less energy expenditure and the leading of a more sedentary lifestyle (Dinsa et al., 2012; Wang & Lim, 2012).

A relationship between low SES and aggression in childhood and adolescence has also been identified in many studies (Datar & Sturm, 2004; Jansen et al., 2012; Najman et al., 2010; Romano, Tremblay, Boulerice, & Swisher, 2005; Santiago, Wadsworth, & Stump, 2011), though results have been mixed. Najman et al. (2010) and Santiago et al. (2011) hypothesized that young people who experience chronic family poverty would eventually resign themselves to being unable to join a more middle-class, mainstream sector of society, and as a form of rebellion elect to engage in delinquent, criminal, and aggressive behavior, often with other like-minded peers. Living a healthy, non-violent lifestyle diminishes in priority as their social exclusion becomes a more evident feature of their day-to-day lives (Najman et al., 2010).

Negative Peer Interactions: Peer Rejection or Peer Victimization

An additional theory may be that overweight and obese children and adolescents experience more negative peer interactions (in the form of peer rejection or victimization by peers), which may lead to subsequent displays of physical aggression (Berg et al., 2005). There is evidence demonstrating support for a relationship between overweight/obesity and peer rejection (Gunnarsdottir, Njardvik, Olafsdottir, Craighead, & Bjarnason, 2012) and victimization by peers (Puhl, Luedicke, & Heuer, 2011; van Geel, Vedder, & Tanilon, 2014), as well as associations between physical aggression and peer rejection (Dodge et al., 2003) and victimization by peers (Ostrov, 2010; Sullivan, Farrell, & Kliewer, 2006) in children and adolescents. Children are heavily reliant on physical cues in their social interactions and are likely to be influenced by stereotypes associated with those cues (such as, that overweight or obese individuals tease others, are teased, fight, or are selfish and mean) and subsequently behave in a manner towards other normal-weight individuals that is congruent with those beliefs and stereotypes (Janssen et al., 2004). Besides conforming with negative stereotypes about themselves, overweight youngsters may also avoid physical activities in an effort to prevent being subjected to teasing and social rejection by others (Gunnarsdottir et al., 2012), thus beginning a vicious cycle by reducing social interaction opportunities, resulting in lower social competence and lower social skills, which may further predispose them to teasing and social rejection (Gunnarsdottir et al., 2012). Children who are victimized may also learn from those interactions and may model aggressive behavior towards others in future hostile interactions with peers, particularly if those subsequent actions are positively reinforced by being effective at reducing peer victimization and facilitating goal attainment such as social status (Ostrov, 2010). As children enter adolescence, achieving a high social status amongst peers takes on paramount importance, and they become increasingly sensitive to stressors that are deemed humiliating or embarrassing (Sullivan et al., 2006). Simultaneously, their coping skills undergo dramatic shifts as their coping repertoire increases, whereby they acquire and use more emotion-focused coping strategies, some of which are less adaptive than others (Sullivan et al., 2006). Aggression and delinquent behaviors are sometimes used as maladaptive coping strategies in order to deal with the emotional arousal created by peer victimization experiences (Sullivan et al., 2006).

Mental Health Issues

Both social disadvantage (and associated child development problems) and peer conflict can contribute to mental health problems, providing a further theory as to why overweight/obesity and physical aggression may be associated. Common mental health problems (such as depression and anxiety) have been found to correlate with both overweight/obesity and physical aggression (Sanders, Han, Baker, & Cobley, 2015). Young people suffering from anxious and depressive symptomatology may engage in more instances of emotional comfort eating or stress-induced eating as a form of emotional defense or as a coping mechanism (Pervanidou & Chrousos, 2011; Wilson & Sato, 2013). Wilson and Sato (2013) suggested that elevations in cortisol levels due to stress may also contribute towards stress-eating, and a common response for an individual experiencing high levels of stress would be to seek out highly palatable (often energy-dense, high-fat, and sweet) foods to provide relief in order to diminish the distress. Chronic stress and depressive symptomatology have also been linked to sedentary behavior because of reduced motivation levels to engage in daily activities (Berg et al., 2005; Datar & Sturm, 2004; Gundersen, Mahatmya, Garasky, & Lohman, 2010; Halfon et al., 2013; Hoare, Skouteris, Fuller-Tyszkiewicz, Millar, & Allender, 2014; Korczak, Lipman, Morrison, & Szatmari, 2013; Mustillo et al., 2003; Pasch et al., 2008; Sanders et al., 2015; Seyedamini et al., 2012; Wilson & Sato, 2013), which may result in an imbalance between energy intake and energy expenditure, increasing the likelihood of these individuals becoming or remaining overweight/obese (Wilson & Sato, 2013). Physical activity is also associated with increased self-efficacy and social integration, which are skills that are protective against symptoms of depression and anxiety (Hoare et al., 2014). Some studies have also found a relationship between stress, anxiety, and depressive symptomatology and physically aggressive behavior (Leschied, Chiodo, Nowicki, & Rodger, 2008; Priddis, Landy, Moroney, & Kane, 2014; Seah & Ang, 2008), though in-depth explorations into this link appear to be lacking. Priddis, Landy, Moroney, and Kane (2014) allude towards difficulties with emotion regulation in youth exhibiting these symptoms and traits, whilst Seah and Ang (2008) point towards anxious individuals experiencing higher levels of hypervigilance with respect to perceived hostile threats in social contexts, which then results in pre-emptive aggressive behavior.

Self-regulation Problems: Impulsive/Risky Behavior (Including Substance Use)

Childhood stress associated with prolonged exposure to social disadvantage, conflict (including both within the home and amongst peers), and mental health problems has been argued to undermine healthy neurological development, resulting in long-term “toxic stress,” which can lead to maladaptive behavioral characteristics such as impulsivity (Hornor, 2015; National Scientific Council on the Developing Child, 2005/2014). Both overweight/obesity and physical aggression in children and adolescents are associated with problems in self-regulation indicated by impulsive and risky behaviors, such as engaging in substance use. Motivations behind engaging in impulsive and risky behaviors (including substance use) seem to center on the increased importance of social standing within peer groups during this crucial developmental and transitional period for young people (Farhat et al., 2010; Huang, Lanza, Wright-Volel, & Anglin, 2013; Lanza, Grella, & Chung, 2014). The critical importance of a sense of belonging, in combination with a desire for peer acceptance and undergoing the development of self-regulatory processes (which are especially important for rational decision-making abilities), may increase a young person’s vulnerability towards engaging in impulsive and risky behaviors (Huang et al., 2013; Lanza et al., 2014). Early substance use may also encourage further social connections with others who engage in similar risky and impulsive behaviors, which may further isolate and distance the individuals from engaging in healthy, normative influences (Pasch, Velazquez, Cance, Moe, & Lytle, 2012). These early maladaptive behaviors may serve as an early indicator of the later development of a wide array of health-risk behaviors, which include sedentary behaviors, the acquisition of poor eating habits, and other habits which may be detrimental towards health (Pasch et al., 2008). Those who have difficulty with self-control and exhibit poor emotion regulation abilities (including anger regulation) may also be more likely to over-consume energy-dense food, which would contribute to an increased overweight/obesity risk (Harrist, Hubbs-Tait, Topham, Shriver, & Page, 2013; Lanza et al., 2014). Reverse effects have also been found in some studies, as increased alcohol intake can have an effect on energy balance and could lead to weight gain over time, or dull incentives to be physically active (Pasch et al., 2012). Aside from over-consumption of energy-dense food, associations have also been found between impulsive and risky behaviors and engaging in physically aggressive behaviors (Mercado-Crespo & Mbah, 2013; Moore et al., 2014; Piko et al., 2005; Raine et al., 1998; Timmermans, van Lier, & Koot, 2008). Moore et al. (2014) hypothesize that physical aggression can serve as a respite and escape from stressful situations, and Raine et al. (1998) purport that it may be a way for young people to seek attention from their caregivers and peers, or to obtain rewards and social status.

Rationale for this Quantitative Review

Given the negative implications for the long-term developmental trajectory of children and adolescents who are overweight/obese or who engage in physically aggressive behavior, further investigation of this relationship appears warranted. Whilst many studies reported data for the required variables (i.e. BMI and incidences of physical aggression), many examined other associations rather than a direct relationship between overweight/obesity and physical aggression. Since significance testing is highly dependent on sample size, it is important to identify effects across pooled samples in order to synthesize findings from prior studies. Neither a systematic review nor a meta-analysis of the existing literature relevant to the relationship between overweight/obesity and physical aggression in children and adolescents has been previously published. The present paper aimed to complete a systematic review and a meta-analysis of the association between overweight/obesity and physical aggression in children and adolescents.

Method

The literature review and meta-analysis were undertaken in accordance with the recommended PRISMA guidelines (Moher, Liberati, Tetzlaff, Altman, & The PRISMA Group, 2009).

Inclusion and Exclusion Criteria

The definition of physical aggression included actions deliberately targeted at physically injuring other individuals. In order to widen the search for relevant articles, search terms with similar meanings (such as aggression, bullying, externalizing behavior, physical assault, and criminal offending) or conditions that include physically aggressive behavior in their symptomatology (such as conduct disorder (CD)) were also included.

The independent predictor involved measurements of overweight/obesity only (indicated by measures of BMI and bulk index), and did not include variants of measurements of body size such as body shape or muscle mass.

Participants up to the age of 18 years were included in the analyses, and studies of a cross-sectional and longitudinal nature were included.

Literature Search

An initial search for articles relevant to the review was conducted utilizing the following electronic databases: PsycInfo; Medline Complete; Web of Knowledge; Academic Search Complete; and SCOPUS. A comprehensive list of the search terms are listed in Table 1.

Table 1.

Summary of databases, search terms, limiters, number of results from each database, and the dates between which the research spanned.

Database	Search terms	Years between which the research spanned	Limiters	Total number of results
PsycInfo	(“body size” OR “obes” OR “overweight” OR “BMI” OR “body mass index”) AND (“aggressi” OR “antisocial behavi” OR “conduct disorder” OR “oppositional defiant disorder” OR “bull” OR “externali?ing behavi” OR “physical assault”) AND (“child” OR “adolescen*”)	1936–2013	English; Age Groups: Childhood (birth–12 yrs), Adolescence (13–17 yrs); Population Group: Human; Methodology: CLINICAL CASE STUDY, EMPIRICAL STUDY, -Experimental Replication, -Followup Study, -Longitudinal Study, ---Prospective Study, ---Retrospective Study, FIELD STUDY, LITERATURE REVIEW, -Systematic Review, MATHEMATICAL MODEL, -Meta Analysis, -Nonclinical Case Study, -Qualitative Study, -Quantitative Study, TREATMENT OUTCOME/CLINICAL TRIAL, TWIN STUDY	180
Medline Complete	As above	1963–2013	Human; Age Related: All Child: 0–18 years; Language: English	499
Web of Knowledge	As above	1943–2013	Article, Other, Case Report, Clinical Trial, Reference Material. Language: English	795
Academic Search Complete	As above	1987–2013	Document Type: Article, Book Chapter, Case Study, Interview; Language: English	374
SCOPUS	As above	1954–2013	English, Articles, Article Conference Paper, Article in Press, Undefined short survey	589

PsycINFO was also searched for unpublished dissertations using the same search terms and limiters. The abstracts of all dissertations registered since 1995 (the point at which full abstracts are available) featuring the search terms and limiters were checked for relevance to the meta-analysis. The pre-specified search criteria were adapted as necessary for each database, and articles published at any time from the beginning of the database records until the search cut-off date (August 2013) were included. This search produced an initial list of 2437 articles. Once duplicates were removed, the number of articles for screening was 1522. The volume of articles suggested by the searches reflected the sensitivity of the search terms used. In deciding which key words and phrases to specify in the search, sensitivity took precedence over specificity; thus, a number of broader terms were used in order to reduce the risk of missing any relevant studies. As a result, a large number of the articles that initially appeared relevant for inclusion were excluded at the title stage. For example, studies concentrating solely on perpetrators of verbal aggression or relational bullying (as opposed to those concerning physical aggression) were excluded at this stage. Manual searches were also made from reference lists of all included articles for potentially eligible studies to include in this review, which yielded an additional 11 articles. Once records deemed to be irrelevant were excluded, 33 articles remained to be assessed for eligibility.

The abstracts of these remaining 33 articles were then read in full and 23 studies were included. A summary of the selection process used in this meta-analysis is outlined in the PRISMA flowchart (Figure 2).

Figure 2.

PRISMA flowchart showing selection of articles included in the meta-analysis.

Hard copies of 33 articles were retrieved, and seven of the retrieved articles were excluded after initial reading as they did not meet the inclusion criteria. Two studies with relevant data (Anderson, Cohen, Naumova, & Must, 2006; Pinhey, 2002) were excluded because of inadequate sample and statistical information, and attempts to contact these authors for clarification were unsuccessful. Another study with relevant data (BeLue, Francis, & Colaco, 2009) was excluded because the statistical data was unclear. The narrative synthesis and data extraction was undertaken by the lead author. This method produced a final list of 23 articles that fell within the scope of the review. Characteristics of the studies included are detailed in Table 2.

Table 2.

Details of studies included in the meta-analysis.

Study reference (including country, sample size, and sample characteristics)	Controlled variables	Outcome variables	Relevant findings
Banis et al. (1988) (USA, n=30, 7–12 years)*	Gender, age, Body Mass Index (BMI), ethnicity, socioeconomic status (SES)	Aggressive behavior (assessed by structured diagnostic interview and the Diagnostic Interview Schedule for Children (DISC)	Obese girls were significantly different in levels of aggression from normal-weight girls that were clinic-referred (p<.001) or non-clinic-referred (p<.001). Obese girls were significantly different in levels of cruelty to others from normal-weight girls that were clinic-referred (p<.001) or non-clinic-referred (p<.05). Obese boys were significantly different in levels of aggression from normal-weight boys that were clinic-referred (p<.001) or non-clinic-referred (p<.001)
Baum and Forehand (1984) (USA, n=48, grades 6–7)*	Gender, age, ethnicity, body size (thin/underweight, average, chubby/slightly overweight, fat/moderately to extremely overweight)	Behavioral observations of dispensing negative behaviors (including physical attacks and threats on others) by undergraduate psychology students	Both the chubby build and fat build groups received and dispensed a significantly greater percentage of negative interactions than the average weight group (p<.05)
Berg et al. (2005) (Sweden, n=959 boys, 15 years)*	BMI, social factors, eating habits and physical activity, body image and relations, school situation, use of alcohol/drugs/tobacco, somatic/psychological symptoms	Physical aggression (kicking or hitting)	Overweight boys were significantly more likely to have ever kicked or hit another student (p<.05). Obese boys were significantly more likely to have ever bullied another student (p<.05)
Datar and Sturm (2004) (USA, n=9949, grades 1–2)/*	Gender, BMI, ethnicity, mother’s education level, annual family income, family structure, parent–child interaction, birth weight, and maternal depressive symptoms	Externalizing behavior (including fighting and disturbing ongoing activities)	Significant association between overweight girls and teacher-reported externalizing behavior problems in kindergarten
Eschenback et al. (2009) (Germany, n=156,948, 6–14 years)*	Gender, age	Externalizing disorder (including conduct disorders)	Incidences of conduct disorder were significantly higher in obese children (p<.001)
Farhat et al. (2010) (USA, n=7825, 11–17 years)*	Age, BMI, ethnicity, family composition, family affluence	Physical fighting	No significant relationships between BMI and physical fighting in boys or girls
Gallup and Wilson (2009) (USA, n=155, 13–16 years)*	Gender, BMI, ethnicity	Physical aggression (hitting, kicking, physical intimidation)	No significant relationship between BMI and self-reported physical aggression
Griffiths et al. (2006) (United Kingdom, n=8210, 7.5–8.5 years)**	Age, gender, BMI, parental social class (manual vs non-manual occupations)	Overt bullying (including threatening/blackmailing, hitting, or beating up)	Weight category (underweight, average weight, overweight) at age 7.5 was found to predict overt bullying status at age 8.5 for boys (p<.05) but not for girls
Halfon et al. (2013) (USA, n=43,296, 10–17 years)*	Age, BMI, ethnicity, household income, family structure, parental education level	Externalizing problems (bullying or being cruel, assessed by the Behaviour Problems Index (BPI))	Significant relationships between both overweight (p<.05) and obese (p<.05) children and adolescents and the presence of externalizing problems and conduct disorders
Hwang et al. (2006) (South Korea, n=453, grades 4–6)	Gender, BMI	Aggressive behavior (assessed by the Aggressive Behaviour Subscale of the Child Behavior Checklist (CBCL))	Overweight children obtained significantly higher scores on the subscale measuring aggressive behavior compared to children who were “at risk of overweight” (p=.025)
Ikäheimo et al. (2007) (Finland, n=5589 boys, measured at birth, 12 months, 14 years, up to 32 years)**	Age, BMI, maternal age, SES at birth, mother’s parity, marital status, smoking status, wantedness of the pregnancy, gestational age of the child	Violent offending (homicide, attempted homicide, assault, robbery, arson, sexual crimes, violation of domestic peace)	Infants with high BMI at 12 months of age were significantly more likely to engage in violent offending behavior in adulthood (p=.001)
Janssen et al. (2004) (Canada, n=5479, 11–16 years)*	Age, gender, BMI	Physical bullying (hitting, shoving, kicking, pushing)	Physical bullying increased with increasing BMI category in girls (p<.05)
Lumeng et al. (2003) (USA, n=755, 8–11 years)*	Gender, ethnicity, mother’s marital status, mother’s education level, family poverty status, mother’s obesity, mother’s depressive symptoms, mother’s smoking status, use of behavior-modifying medication, hours of television per day, and history of academic grade retention	Aggression scores (assessed by the BPI)	Significant relationship between being overweight and clinically meaningful behavior (scores above the 90th percentile on the BPI) (p=.01)
Luukkonen et al. (2010) (Finland, n=508, 12–17 years)*	Gender, family type, repeated grade at school	Bullying (assessed by scores on the semi-structured Schedule for Affective Disorder and Schizophrenia for School-Age Children, Present and Lifetime)	No significant relationship between overweight status and being a bully or bully-victim
Mustillo et al. (2003) (USA, n=991, 9–13 years then annual follow-up until 16)**	Age, gender, BMI, family income, parental education, family structure, parental mental illness history, drug abuse or criminal convictions, harsh/overprotective parenting style, lax supervision, traumatic life events	Oppositional defiant disorder or conduct disorder diagnoses (assessed by the Child and Adolescent Psychiatric Assessment)	Diagnosis of oppositional defiant disorder was significantly more common in chronically obese boys and girls (p<.01)
Pasch et al. (2008) (USA, n=3010, grades 7–8)/*	Age, BMI, SES, family structure, ethnicity	Physical fighting behavior	No significant relationship between BMI in grade 7 and fighting behavior in grade 7 or grade 8
Piko, Keresztes, and Pluhar (2006) (Hungary, n=550, 10–15 years)*	Age⁺, BMI, gender, school achievement (grades from school)	Physical aggression (measured by the Aggression Questionnaire)	Significant relationship between BMI and physical aggression in boys (p<.05)
Pitrou et al. (2010) (France, n=2341, 6–11 years)*	Age⁺, gender, BMI, lifestyle behaviors (physical activity outside school, watching TV, playing video games, participation in organized sports), sociodemographic characteristics (parental education, parents’ employment status, household income, family structure), parental psychological distress and alcohol misuse, traumatic life events, parenting style (caring/punitive behaviors, autonomy-promoting, overprotective attitudes), academic achievement	Possible ODD and conduct disorder (CD) diagnoses (measured by the Dominic Interactive) and parent-reported conduct disorders (measured by the Strengths and Difficulties Questionnaire (SDQ))	No significant relationship between BMI and possible ODD or CD diagnoses or parent-reported conduct disorders
Raine et al. (1998) (Mauritius, n=1795, 3 and 11 years)/*	Gender, ethnicity, bulk index (measured using height and weight), SES	Physical aggression (assessed by the Aggression Scale from the CBCL)	Aggressive children at age 3 were significantly taller (p<.002), heavier (p<.02), and bulkier (p<.004) relative to non-aggressive children. Body size at age 3 years was related to aggression at age 11 (height p<.007, bulk p<.04, weight p<.09)
Sawyer et al. (2006) (Australia, n=4983, 4–5 years)**	Gender, BMI, ethnicity, indigenous status, maternal sociodemographics (highest education level, psychological distress), family sociodemographics (occupation, income), neighborhood sociodemographics (neighborhood disadvantage)	Self-, parent-, and teacher-reported conduct problems (measured by the SDQ)	Significant relationship between teacher-reported conduct problems and BMI in boys (p<.05)
Seyedamini et al. (2012) (Iran, n=300 girls, 7–12 years)*	BMI, sociodemographics (maternal education level, birth order, TV watching duration), crisis experience during last six months, previous disease history, birth weight	Aggressive behavior (assessed by the CBCL)	Significant relationship between aggression in overweight (p=.0001) and obese (p=.03) girls
Tremblay et al. (1998) (Canada, n=1037 boys, 6–13 years)**	Age, parents’ age (at child’s birth), parents’ educational/occupational status, family composition (at age 6), BMI, testosterone levels	Physical aggression (assessed using the Social Behaviour Questionnaire for teacher-reports, the Pupil Evaluation Inventory for peer-ratings, and a 27-item self-report delinquency questionnaire)	BMI significantly correlated with self-reported physical aggression (p<.001). Self-reported physical aggression at age 13 was correlated with BMI at age 12 and 13 (p<.001)
Zeller et al. (2008) (USA, n=166, 8–16 years)*	Age⁺, BMI, ethnicity, SES, family composition, family size, parental age/education	Aggressive-disruptive behavior (including engaging in physical fighting) rated by peers, teachers, and self	BMI significantly correlated with peer-reported aggressive-disruptive behavior (p=.03)

ODD: Oppositional Defiant Disorder.

* Cross-sectional design.

** Longitudinal design.

⁺ Separate sub-analyses for age differences was not reported.

Nineteen of the articles identified for inclusion in this meta-analysis measured some form of physically aggressive behavior. The types of behaviors included the following:

aggressive and cruel behaviors (as measured by the Aggression subscale of the Child Behavior Checklist) (Halfon et al., 2013; Hwang et al., 2006; Ikäheimo et al., 2007; Lumeng et al., 2010; Raine et al., 1998; Seyedamini et al., 2012);

physical attacks and fighting (Datar & Sturm, 2004; Farhat et al., 2010; Lumeng et al., 2010; Pasch et al., 2008; Raine et al., 1998; Tremblay et al., 1998);

threats (including threats to physically assault another) (Lumeng et al., 2010);

kicking, pushing, or hitting (Janssen et al., 2004);

bullying (including overt bullying) (Farhat et al., 2010; Griffiths et al., 2006; Halfon et al., 2013; Janssen et al., 2004; Lumeng et al., 2003);

conduct problems and disorders (Eschenbeck et al., 2009; Mustillo et al., 2003; Pitrou et al., 2010; Sawyer et al., 2006);

violent offending behavior (Ikäheimo et al., 2007).

Diverse data-recording methodologies were utilized, including:

psychometric testing, utilizing self-, parent-, and teacher-report measures such as the Child Behavior Checklist (Hwang et al., 2006; Ikäheimo et al., 2007; Raine et al., 1998; Seyedamini et al., 2012), the Bullying and Friendship Schedule (Griffiths et al., 2006), the Behavior Problem Index (Halfon et al., 2013; Lumeng et al., 2003), the Aggression Questionnaire (Piko et al., 2005), the Schedule for Affective Disorder and Schizophrenia for School-Age Children, Present and Lifetime (Luukkonen et al., 2010), and the Strengths and Difficulties Questionnaire (Pitrou et al., 2010; Sawyer et al., 2006);

behavioral observations (Lumeng et al., 2010);

national surveys (Griffiths et al., 2006; Halfon et al., 2013; Ikäheimo et al., 2007; Janssen et al., 2004; Lumeng et al., 2003; Mustillo et al., 2003; Pitrou et al., 2010; Sawyer et al., 2006);

health insurance surveys (Eschenbeck et al., 2009);

self-, parent-, and teacher-reported questionnaires (Pasch et al., 2008; Piko et al., 2005; Tremblay et al., 1998; Zeller et al., 2008);

criminal records (Ikäheimo et al., 2007).

In the majority of studies, overweight/obesity indicators were obtained by calculating the body mass index, a standard formula involving height and weight (Datar & Sturm, 2004; Farhat et al., 2010; Griffiths et al., 2006; Halfon et al., 2013; Hwang et al., 2006; Ikäheimo et al., 2007; Janssen et al., 2004; Lumeng et al., 2003; Mustillo et al., 2003; Pasch et al., 2008; Pitrou et al., 2010; Raine et al., 1998; Sawyer et al., 2006; Seyedamini et al., 2012; Tremblay et al., 1998). Current research suggests that BMI is an accurate and widely used measure of overweight/obesity (Aeberli, Gut-Knabenhans, Kusche-Ammann, Molinari, & Zimmermann, 2013; Laurson, Eisenmann, & Welk, 2011; Ralston, Walker, & Truby, 2012). In the remaining studies, observer ratings (of thin/underweight, average, chubby/slightly overweight, and fat/moderately to extremely overweight) (Lumeng et al., 2010), physician-diagnosed obesity (Eschenbeck et al., 2009), and a bulk index (calculated as the z-transformation of the product of height and weight, with a constant of 1 added to scores) (Raine et al., 1998) were used.

Of the 23 studies included in this meta-analysis, the results from three studies utilized unadjusted correlations and reported only the ages, genders, and BMIs of the participants (Eschenbeck et al., 2009; Hwang et al., 2006; Janssen et al., 2004); the remaining 20 controlled for factors such as socioeconomic status (including neighborhood sociodemographics (Banis et al., 1988; Datar & Sturm, 2004, Farhat et al., 2010; Griffiths et al., 2006; Halfon et al., 2013; Ikäheimo et al., 2007; Lumeng et al., 2003; Mustillo et al., 2003; Pasch et al., 2008; Pitrou et al., 2010; Raine et al., 1998; Sawyer et al., 2006; Zeller et al., 2008), mental illness (Berg et al., 2005; Mustillo et al., 2003; Seyedamini et al., 2012), and problematic alcohol or drug use (Berg et al., 2005)).

Effect Size Calculations and Data Extraction Procedures

From each study, data was extracted to allow an effect size to be computed (along with its variance) that quantified the effect of overweight/obesity on physical aggression, using a standardized mean difference (SMD) as the effect size measure. Positive values indicated that children and adolescents who were overweight or obese evidenced higher incidences of engaging in physical aggression compared to those who were underweight or of normal weight.

The Comprehensive Meta-Analysis Software Package (Borenstein et al., 2011) was used to convert the extracted data to a common effect size and variance.

Meta-Analytic Procedures

Effect sizes from each study were pooled and a weighted average effect size was computed using a random effects model as the differences between study level effect sizes were assumed to be the sum of sampling error (within-study variance) and “true” effect size differences (between-study variance), and additionally, the sample characteristics of each study and the outcome measures used were not presumed to be equivalent. Significance tests for the weighted average effect size were computed using an alpha level of .05.

The total amount of heterogeneity between studies was measured using the Q and I ² statistics, which described the presence and amount of heterogeneity (respectively) in effect sizes (as a percentage or proportion) that was attributable to between-study effects rather than within-study effects (e.g. participant or methodological characteristics) (Higgins & Thompson, 2002). As a guideline, Higgins, Thompson, Deeks, and Altman (2003) suggest that values of 25%, 50%, and 75% correspond to low, moderate, and high levels of heterogeneity respectively.

Publication Bias Analyses

Publication bias can occur due to problems such as non-publication of non-significant findings. To investigate this issue, funnel plots and Egger’s test of asymmetry were conducted to measure whether the average effect sizes were more often identified in the larger (higher precision) studies and whether the distribution was symmetric (Egger, Smith, Schneider, & Minder, 1997). Rosenthal’s Fail-Safe N was computed to assess whether possible non-significant studies were inadvertently omitted from the analysis, which would reduce the observed collective effect size.

Results

Overweight/Obesity and Physical Aggression: Adjusted Effect Sizes

A total of 23 studies were included. The effect sizes computed for each study and the weighted average effect size are presented as a forest plot in Table 3. Positive SMD values indicated that overweight or obese children or adolescents were more likely to be physically aggressive than their average-weight or underweight peers. The weighted average effect size computed was .27 (95% confidence interval [CI₉₅]: .17–.37), and highly significant (p<.001). This indicated that on average, children and adolescents who were overweight or obese were more likely to engage in physical aggression than their peers who were not overweight or obese. According to Cohen’s (1988) taxonomy, this corresponds to a small to medium effect size.

Table 3.

Summary of adjusted effect sizes of overweight/obesity on physical aggression across all studies.

CD: conduct disorder

Table 3 demonstrates that there was substantial variability in study level effect sizes. For example, the largest effect size observed was 1.22 and the smallest –0.30. The effects across studies were highly heterogeneous (Q(46)=823.15, p<.001). Calculation of the I ² statistic confirmed this, and provided further indication that 94.41% of variability between effect sizes represented true heterogeneity, a strong suggestion of the existence of genuine differences underlying the results of the studies, rather than the result of chance. The high levels of heterogeneity between the studies included in the analysis suggest that both non-systematic study influences and true effects accounted for the differences in study results, thus supporting the decision to utilize a random effects model.

Two sub-analyses controlling for gender were conducted in order to ascertain whether the overall significant effect between overweight/obesity and physical aggression was maintained. Of the 23 studies included in this meta-analysis, 13 studies were included in a sub-analysis investigating the relationship between overweight/obesity and physical aggression in boys, and 10 studies in a sub-analysis investigating the relationship between overweight/obesity and physical aggression in girls. The results demonstrated that the overweight/obesity statuses of both genders were significantly associated with physically aggressive behavior. Compared to their normal-weight peers, overweight or obese boys were more likely to be physically aggressive (SMD=.35, CI₉₅: .18–.52, p<.001) with this effect also significant but somewhat smaller for overweight or obese girls (SMD=.24, CI₉₅: .07–.42, p<.01). The results of these analyses are presented as forest plots in Tables 4 and 5.

Table 4.

Summary of adjusted effect sizes of overweight/obesity on physical aggression for boys.

Table 5.

Summary of adjusted effect sizes of overweight/obesity on physical aggression for girls.

Sub-analyses for age groups were unable to be conducted as the age group clusters were varied, and when all the available data was collated many of the age groups overlapped without a way of discerning the distribution of participants at each age.

Publication Bias Analyses

Evidence of publication bias was not found. The funnel plot appeared visually symmetrical, and this was supported by Egger’s test of asymmetry, which was not significant (intercept=0.53, t(45)=0.50, p=.31 (one-tailed)). Additionally, Rosenthal’s Fail-Safe N yielded a value of 5888, suggesting that an additional 5888 studies not yielding significant correlations between overweight/obesity and physical aggression in children and adolescents would be needed to increase the p-value for the meta-analysis to above 0.05 and nullify the current significant results.

Discussion

The aim of this paper was to systematically review prior studies in order to complete a meta-analysis to ascertain the strength of the relationship between overweight/obesity and physical aggression in children and adolescents. Pooling the data from 23 international studies, it was identified that physical aggression was significantly associated with overweight and obesity. Overweight and obese children and adolescents were more likely to engage in physically aggressive behaviors, compared to those who were not obese or overweight. These effects were found to apply across gender, although effects were slightly larger for boys compared to girls. Tests of heterogeneity demonstrated that the collection of studies compiled and included in this analysis were highly dissimilar (94% heterogeneity), suggesting that there may be underlying differences in study characteristics that contribute to the observed association between child and adolescent weight status and aggression.

Despite the wide variations between (and within) the studies included in this analysis (such as differences in participant populations, outcome measures, and study designs), the meta-analysis supported the view that there is an underlying relationship between overweight/obesity and physical aggression in children and adolescents. The current analysis included studies spanning a wide age group, varying forms of physical aggression, utilization of different study designs and outcome measures, as well as diverse data collection and reporting methodologies. Many of the included studies were controlled, with well-defined variables, which may have also contributed towards high levels of overall heterogeneity. The number and range of research presented in this review utilized a large number of participants (total n=255,377), covered a wide range of ages (from four to 17 years of age), geographical locations (including participants from Caucasian, Asian, Middle Eastern, European, Pacific Islander, and Scandinavian descent), and measures (such as psychometric tests, behavioral observations, surveys and questionnaires, criminal history records, and self-reports).

It was also unlikely that bias impacted the study selection process used in this analysis. As demonstrated in Table 3, although the width of the confidence intervals from the studies varied greatly, the combined effect size yielded from the analysis remained highly significant, strengthening the proposition that overweight/obesity could be a risk factor that uniquely contributes to physical aggression in children and adolescents.

Implications for Theory

The findings of this analysis reveal that a small to moderate association exists between overweight/obesity and physical aggression in children and adolescents. These findings raise the importance of investigating the theoretical mechanisms that may explain these associations.

It is possible that having an overweight or large body size could lead to future displays of physical aggression, or that a young person who engages in physically aggressive behavior may subsequently become overweight or obese. Longitudinal observations can identify temporal order. In the current analysis, data from six longitudinal studies were included and revealed overall support for theories (Figure 1) that suggest that overweight or large body size leads to future displays of physical aggression. Five studies demonstrated a significant longitudinal relationship showing overweight/obesity preceded physical aggression (Griffiths et al., 2006 (boys); Ikäheimo et al., 2007; Pasch et al., 2008; Raine et al., 1998; Tremblay et al., 1998), while one showed effects in a similar direction that were not significant (Mustillo et al., 2003). Griffiths et al. (2006) found no effect for girls. In two studies, predictive effects were found to apply from pre-school. Ikäheimo et al. (2007) found that infants with a high BMI at 12 months of age were significantly more likely to engage in violent offending behavior in adulthood. Raine et al. (1998) found that bulk index (measured using height and weight) at age three was related to physical aggression at age 11. Although Tremblay et al. (1998) found evidence that BMI preceded aggression, there was also evidence of reciprocal associations, with physical aggression at ages 10 and 11 a significant predictor of BMI at age 12.

Evidence of reciprocal relationships between both overweight/obesity and physical aggression suggests these variables could have common underlying determinants. As discussed in the Introduction, the research literature suggests that several theoretical processes (Figure 1) may contribute in some way to this relationship (such as SES, mental health issues (such as anxiety and depression), and poor self-regulation indicated by engaging in impulsive and risky behaviors (including substance use)). The nature of these influences and their potential role as mediator or moderator variables remains unclear.

Based on the present findings it is unlikely that SES fully explains the common association between overweight/obesity and physical aggression. When SES was adjusted for in eight of the included studies, the relationship between overweight/obesity and physical aggression remained significant (Banis et al., 1988; Datar & Sturm, 2004 (for girls only); Griffiths et al., 2006; Halfon et al., 2013; Ikäheimo et al., 2007; Lumeng et al., 2010; Mustillo et al., 2003; Raine et al., 1998). In four studies, effects were non-significant and remained so after adjustment for SES (Datar & Sturm, 2004 (for boys only); Farhat et al., 2010; Pasch et al., 2012; Pitrou et al., 2010). Only one study found a significant association was no longer significant after adjustment for SES (Sawyer et al., 2006). The results of the present analysis, together with the review of the literature in the Introduction, have provided strong indicators that a relationship between overweight/obesity and physical aggression exists amongst children and adolescents.

Directions for Future Research

Since overweight/obesity and physically aggressive behavior can all have dire consequences with respect to mental health outcomes and developmental trajectories, further research investigating this relationship in longitudinal and controlled overweight/obesity prevention trials should take place in order to create and implement early intervention strategies when early indicators are present. The role of common risk factors for physically aggressive behavior in children and adolescents (including socioeconomic status, health-risk behaviors, poor coping skills, mental health symptoms (such as anxiety and depression), impulsive and risky behaviors) should be controlled in future studies, while also examining the effects of sedentary behaviors and poor eating habits (Pasch et al., 2008).

Future research should be conducted utilizing large, representative samples with measures and methods standardized to be comparable to those most commonly identified in the current review for more consistent results.

Research into the underlying developmental processes that may explain a potential effect of overweight/obesity on physical aggression (see Figure 1) should be completed, together with an examination into any mediators or moderators that may influence the relationship. Future studies should seek to understand the reasons for heterogeneity and factors that may modify the association. Advancing theoretical and empirical understanding would enable the development of effective and targeted intervention strategies to promote overall healthy lifestyle characteristics during this critical age of development.

Limitations

Although results from this meta-analysis implied higher levels of physical aggression among overweight or obese children or adolescents, it cannot address the issue of causality. For example, there were no studies that used an intervention design to reduce levels of overweight or obesity and then demonstrate whether this subsequently reduced levels of physical aggression. The majority of data utilized from the included studies used cross-sectional, correlational research designs. However, the six longitudinal studies generally supported the view that overweight/obese children and adolescents subsequently became physically aggressive. To better establish causation, future longitudinal and intervention studies could examine the reciprocal effects of modifying either overweight/obesity or physical aggression or their common determinants.

More detailed sub-analyses (such as for age groups) or meta-regressions were unable to be conducted due to lack of reported data from the studies. Therefore, the present analysis was unable to include information relating to the contributions of covariates; given the variety of study designs included in this analysis, measures and results did not utilize a consistent, standardized format. As such, the accuracy of the results may have been affected.

The general and widely used measure of overweight/obesity in the developmental literature is BMI (Aeberli et al., 2013; Mercado-Crespo & Mbah, 2013). A limitation of using BMI as the sole indicator of overweight/obesity, however, is the fact that only measures of overall height and weight of the individual are taken. Factors that may affect an individual’s weight, yet are not an indicator of adiposity (such as muscle mass), are not considered. Indeed, some studies have demonstrated that BMI may be positively correlated with increased muscle strength (and therefore mass) (Ervin, Fryar, Wang, Miller, & Ogden, 2014; Gallup, White, & Gallup, 2007; Kader, Hasan, Kamal, & Hussein, 2016). As such, measures of waist circumference (Aeberli et al., 2013; Daniels, 2009; Freedman & Sherry, 2009) and skinfold thickness tests (Freedman & Sherry, 2009) conducted in addition to a calculation of BMI would increase the sensitivity and accuracy of measurements of overweight/obesity in future research.

A further limitation of this analysis was the variation in outcome measures. For example, the definition of “bullying” in some of the included studies consisted of both relational and physical aggression (Griffiths et al., 2006; Halfon et al., 2013; Luukkonen et al., 2010; Pitrou et al., 2010), and were not separated in the reported results, making it difficult to define the proportions of participants that engaged in relational and physical aggression. This variation in measures may have contributed to the large heterogeneity in the results.

Finally, it remains possible that the current findings are limited by publication bias. Given the high value of Rosenthal’s Fail-Safe N, non-reporting of non-significant associations is considered unlikely to have undermined the overall finding of a significant association.

Conclusion

Notwithstanding the limitations identified above, this meta-analysis represents an accurate picture of current research and concludes that there is a small significant relationship between overweight/obesity and physical aggression in children and adolescents. Given the many adverse physical, social, and psychological effects upon children and adolescents who are larger or physically aggressive, and the financial burden placed on society, it would be pertinent for further longitudinal and intervention research to be conducted to specify the potential causal models defined in Figure 1. This review has highlighted significant between-study heterogeneity, suggesting that future research should seek to better understand contextual variation and standardize study methods and measures in order to more accurately specify the causal effect of overweight/obesity on physical aggression in children and adolescents.

Footnotes

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Access Economics. (2008). The growing cost of obesity in 2008: Three years on. Retrieved from www.diabetesaustralia.com.au/PageFiles/7830/FULLREPORTGrowingCostOfObesity2008.pdf.

Aeberli

Gut-Knabenhans

Kusche-Ammann

R. S.

Molinari

Zimmermann

M. B.

(2013). A composite score combining waist circumference and body mass index more accurately predicts body fat percentage in 6- to 13-year-old children. European Journal of Nutrition, 52, 247–253.

American Heart Association. (2013). Statistical fact sheet 2013 update: Overweight and obesity. Retrieved from www.heart.org/idc/groups/heart-public/@wcm/@sop/@smd/documents/downloadable/ucm_319588.pdf.

Anderson

S. E.

Cohen

Naumova

E. N.

Must

(2006). Relationship of childhood behaviour disorders to weight gain from childhood to adulthood. Ambulatory Pediatrics, 6, 297–301.

Australian Bureau of Statistics. (2013). Children who are overweight or obese. Retrieved from www.abs.gov.au/AUSSTATS/abs@.nsf/Lookup/1301.0Chapter11062009%E2%80%9310#.

Banis

H. T.

Varni

J. W.

Wallander

J. L.

Korsch

Jay

S. M.

Adler

… Negrete

(1988). Psychological and social adjustment of obese children and their families. Child: Care, Health and Development, 14, 157–173.

Baum

C. G.

Forehand

(1984). Social factors associated with adolescent obesity. Journal of Pediatric Psychology, 9, 293–302.

BeLue

Francis

L. A.

Colaco

(2009). Mental health problems and overweight in a nationally representative sample of adolescents: Effects of race and ethnicity. Pediatrics, 123, 697–703.

Berg

Simonsson

Ringqvist

(2005). Social background, aspects of lifestyle, body image, relations, school situation, and somatic and psychological symptoms in obese and overweight 15-year-old boys in a county in Sweden. Scandinavian Journal of Primary Health Care, 23, 95–101.

10.

Bond

Thomas

Toumbourou

Patton

Catalano

(2000). Improving the lives of young Victorians in our community: A survey of risk and protective factors. Parkville, Australia: Centre for Adolescent Health.

11.

Bor

McGee

T. R.

Hayatbakhsh

Dean

Najman

J. M.

(2010). Do antisocial females exhibit poor outcomes in adulthood? An Australian cohort study. Australian and New Zealand Journal of Psychiatry, 44, 648–657.

12.

Borenstein

Hedges

Higgins

Rothstein

(2011). Comprehensive meta-analysis version 2.0 [Computer Software]. Englewood, NJ: Biostat.

13.

Centers for Disease Control and Prevention. (2014). Childhood obesity facts. Retrieved from www.cdc.gov/healthyyouth/obesity/facts.htm.

14.

Child Obesity . (2015). Retrieved from www.hsph.harvard.edu/obesity-prevention-source/obesity-trends/global-obesity-trends-in-children/.

15.

Cohen

(1988). Statistical power analysis for the behavioural sciences. Hillsdale, NJ: Lawrence Erlbaum Associates.

16.

Daniels

S. R.

(2009). The use of BMI in the clinical setting. Pediatrics, 124, s35–s41.

17.

Daniels

S. R.

Arnett

D. K.

Eckel

R. H.

Gidding

S. S.

Hayman

L. L.

Kumanyika

… Williams

C. L.

(2005). Overweight in children and adolescents: Pathophysiology, consequences, prevention, and treatment. Circulation, 111, 1999–2002.

18.

Datar

Sturm

(2004). Childhood overweight and parent- and teacher-reported behaviour problems. Archives of Pediatrics and Adolescent Medicine, 158, 804–810.

19.

Dietz

W. H.

(2004). Overweight in childhood and adolescence. New England Journal of Medicine, 350, 855–857.

20.

Dinsa

G. D.

Goryakin

Fumagalli

Suhrcke

(2012). Obesity and socioeconomic status in developing countries: A systematic review. Obesity Reviews, 13, 1067–1079.

21.

Dodge

K. A.

Lansford

J. E.

Burke

V. S.

Bates

J. E.

Pettit

G. S.

Fontaine

Price

J. M.

(2003). Peer rejection and social information-processing factors in the development of aggressive behaviour problems in children. Child Development, 74, 374–393.

22.

Eaton

D. K.

Kann

Kinchen

Shanklin

Flint

K. H.

Hawkins

… Wechsler

(2012). Youth risk behavior surveillance – United States, 2011. Morbidity and Mortality Weekly Report, 61, 1–59.

23.

Egger

Smith

G. D.

Schneider

Minder

(1997). Bias in meta-analysis detected by a simple, graphical test. British Medical Journal, 315, 629–634.

24.

Ervin

R. B.

Fryar

C. D.

Wang

Miller

I. M.

Ogden

C. L.

(2014). Strength and body weight in US children and adolescents. Pediatrics, 134, e782–e789.

25.

Eschenbeck

Kohlmann

Dudey

Schurholz

(2009). Physician-diagnosed obesity in German 6-14 year olds. Obesity Facts, 2, 67–73.

26.

Farhat

Iannotti

R. J.

Simons-Morton

B. G.

(2010). Overweight, obesity, youth, and health-risk behaviours. American Journal of Preventive Medicine, 38, 258–267.

27.

Fergusson

D. M.

Woodward

L. J.

(2000). Educational, psychosocial, and sexual outcomes of girls with conduct problems in early adolescence. Journal of Child Psychiatry, 41, 779–792.

28.

Freedman

D. S.

Sherry

(2009). The validity of BMI as an indicator of body fatness and risk among children. Pediatrics, 124, s23–s34.

29.

Gallup

A. C.

White

D. D.

Gallup

G. G.

(2007). Handgrip strength predicts sexual behavior, body morphology, and aggression in male college students. Evolution and Human Behavior, 28, 423–429.

30.

Gallup

A. C.

Wilson

D. S.

(2009). Body mass index (BMI) and peer aggression in adolescent females: An evolutionary perspective. Journal of Social, Evolutionary, and Cultural Psychology, 3, 356–371.

31.

Griffiths

L. J.

Wolke

Page

A. S.

Horwood

J. P.

(2006). Obesity and bullying: Different effects for boys and girls. Archives of Disease in Childhood, 91, 121–125.

32.

Gundersen

Mahatmya

Garasky

Lohman

(2010). Linking psychosocial stressors and childhood obesity. Obesity Reviews, 12, e54–e63.

33.

Gunnarsdottir

Njardvik

Olafsdottir

A. S.

Craighead

L. W.

Bjarnason

(2012). Teasing and social rejection among obese children enrolling in family-based behavioural treatment: Effects on psychological adjustment and academic competencies. International Journal of Obesity, 36, 35–44.

34.

Halfon

Larson

Slusser

(2013). Associations between obesity and comorbid mental health, developmental, and physical health conditions in a nationally representative sample of US children aged 10 to 17. Academic Pediatrics, 13, 6–13.

35.

Harrist

A. W.

Hubbs-Tait

Topham

G. L.

Shriver

L. H.

Page

M. C.

(2013). Emotion regulation is related to children’s emotional and external eating. Journal of Developmental and Behavioral Pediatrics, 34, 557–565.

36.

Higgins

Thompson

S. G.

(2002). Quantifying heterogeneity in a meta-analysis. Statistics in Medicine, 21, 1539–1558.

37.

Higgins

Thompson

S. G.

Deeks

J. J.

Altman

D. G.

(2003). Measuring inconsistency in meta-analyses. British Medical Journal, 327, 557–560.

38.

Hoare

Skouteris

Fuller-Tyszkiewicz

Millar

Allender

(2014). Associations between obesogenic risk factors and depression among adolescents: A systematic review. Obesity Reviews, 15, 40–51.

39.

Hornor

(2015). Childhood trauma exposure and toxic stress: What the PNP needs to know. Journal of Pediatric Health Care, 29, 191–198.

40.

Huang

D. Y. C.

Lanza

H. I.

Wright-Volel

Anglin

M. D.

(2013). Developmental trajectories of childhood obesity and risk behaviours in adolescents. Journal of Adolescence, 36, 139–148.

41.

Hwang

J. W.

Lyoo

I. K.

Nyun

Shin

M. S.

Kim

S. J.

Cho

S. C.

(2006). The relationship between temperament and character and psychopathology in community children with overweight. Developmental and Behavioral Pediatrics, 27, 18–24.

42.

Ikäheimo

Räsänen

Hakko

Hartikainen

Laitinen

Hodgins

Tiihonen

(2007). Body size and violent offending among males in the Northern Finland 1966 birth cohort. Social Psychiatry and Psychiatric Epidemiology, 42, 845–850.

43.

Jansen

P. W.

Verlinden

Berkel

A. D.

Mieloo

van der End

Veenstra

… Tiemeier

(2012). Prevalence of bullying and victimization among children in early elementary school: Do family and school neighbourhood socioeconomic status matter? BioMed Central Public Health, 12, 1–10.

44.

Janssen

Craig

W. M.

Boyce

W. F.

Picket

(2004). Associations between overweight and obesity with bullying behaviours in school-aged children. Pediatrics, 113, 1187–1194.

45.

Kader

N. A.

Hasan

A. K.

Kamal

H. M.

Hussein

Z. A.

(2016). Relation between body mass index percentile and muscle strength and endurance. The Egyptian Journal of Medical Human Genetics, 17, 367–372.

46.

Korczak

D. J.

Lipman

Morrison

Szatmari

(2013). Are children and adolescents with psychiatric illness at risk for increased future body weight? A systematic review. Developmental Medicine and Child Neurology, 55, 980–988.

47.

Lanza

H. I.

Grella

C. E.

Chung

P. J.

(2014). Does adolescent weight status predict problematic substance use patterns? American Journal of Health Behavior, 38, 708–716.

48.

Laurson

K. R.

Eisenmann

J. C.

Welk

G. J.

(2011). Body mass index standards based on agreement with health-related body fat. American Journal of Preventive Medicine, 41, s100–s105.

49.

Leschied

Chiodo

Nowicki

Rodger

(2008). Childhood predictors of adult criminality: A meta-analysis drawn from the prospective longitudinal literature. Canadian Journal of Criminology and Criminal Justice, 50, 435–467.

50.

Loeber

(1990). Development and risk factors of juvenile antisocial behaviour and delinquency. Clinical Psychology Review, 10, 1–41.

51.

Lumeng

J. C.

Gannon

Cabral

H. J.

Frank

D. A.

Zuckerman

(2003). Association between clinically meaningful behaviour problems and overweight in children. Pediatrics, 112, 1138–1145.

52.

Lumeng

J. C.

Forrest

Appugliese

D. P.

Kaciroti

Corwyn

R. F.

Bradley

R. H.

(2010). Weight status as a predictor of being bullied in third through sixth grades. Pediatrics, 125, 1138–1145.

53.

Luukkonen

Räsänen

Hakko

Riala

The STUDY-70 Workgroup. (2010). Bullying behaviour in relation to psychiatric disorders and physical health among adolescents: A clinical cohort of 508 underage inpatient adolescents in Northern Finland. Psychiatry Research, 178, 166–170.

54.

Mercado-Crespo

M. C.

Mbah

A. K.

(2013). Race and ethnicity, substance use, and physical aggression among U.S. high school students. Journal of Interpersonal Violence, 28, 1367–1384.

55.

Moher

Liberati

Tetzlaff

Altman

D. G.

, & the PRISMA Group. (2009). Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. Annals of Internal Medicine, 151, 264–269.

56.

Moore

S. E.

Norman

R. E.

Sly

P. D.

Whitehouse

A. J. O.

Zubrick

S. R.

Scott

(2014). Adolescent peer aggression and its association with mental health and substance use in an Australian cohort. Journal of Adolescence, 37, 11–21.

57.

Mustillo

Worthman

Erkanli

Keeler

Angold

Costello

E. J.

(2003). Obesity and psychiatric disorder: Developmental trajectories. Pediatrics, 111, 851–860.

58.

Najman

J. M.

Clavarino

McGee

T. R.

Bor

Williams

G. M.

Hayatbakhsh

M. R.

(2010). Timing and chronicity of family poverty and development of unhealthy behaviours in children: A longitudinal study. Journal of Adolescent Health, 46, 538–544.

59.

National Health and Medical Research Council. (2014). Obesity and overweight. Retrieved from www.nhmrc.gov.au/your-health/obesity-and-overweight.

60.

National Scientific Council on the Developing Child. (2005/2014). Excessive stress disrupts the architecture of the developing brain. Working paper 3. Center on the Developing Child at Harvard University, 1–9.

61.

Ostrov

J. M.

(2010). Prospective associations between peer victimization and aggression. Child Development, 81, 1670–1677.

62.

Pasch

K. E.

Nelson

M. C.

Lytle

L. A.

Moe

S. G.

Perry

C. L.

(2008). Adoption of risk-related factors through early adolescence: Associations with weight status and implications for causal mechanisms. Journal of Adolescent Health, 43, 387–393.

63.

Pasch

K. E.

Velazquez

C. E.

Cance

J. D.

Moe

S. G.

Lytle

L. A.

(2012). Youth substance use and body composition: Does risk in one area predict risk in the other? Journal of Youth Adolescence, 41, 14–26.

64.

Pervanidou

Chrousos

G. P.

(2011). Stress and obesity/metabolic syndrome in childhood and adolescence. International Journal of Pediatric Obesity, 6, 21–28.

65.

Piko

B. F.

Keresztes

Pluhar

Z. F.

(2006). Aggressive behaviour and psychosocial health among children. Personality and Individual Differences, 40, 885–895.

66.

Pinhey

T. K.

(2002). A research note on body mass, physical aggression, and the competitiveness of Asian-Pacific Islander adolescents in Guam. Biodemography and Social Biology, 49, 90–98.

67.

Pitrou

Shojaei

Wazana

Gilbert

Kovess-Masfety

(2010). Child overweight, associated psychopathology, and social functioning: A French school-based survey in 6- to 11-year-old children. Obesity, 18, 809–817.

68.

Priddis

L. E.

Landy

Moroney

Kane

(2014). An exploratory study of aggression in school-age children: Underlying factors and implications for treatment. Australian Journal of Guidance and Counselling, 24, 18–35.

69.

Puhl

R. M.

Luedicke

Heuer

(2011). Weight-based victimization toward overweight adolescents: Observations and reactions of peers. Journal of School Health, 81, 696–703.

70.

Raine

Reynolds

Venables

P. G.

Mednick

Farrington

(1998). Fearlessness, stimulation-seeking, and large body size at age 3 years as early predispositions to childhood aggression at age 11 years. Archives of General Psychiatry, 55, 745–751.

71.

Ralston

R. A.

Walker

K. Z.

Truby

(2012). A review of the indices and references used to assess overweight and obesity in Australian children and adolescents. Nutrition & Dietetics, 69, 300–308.

72.

Romano

Tremblay

R. E.

Boulerice

Swisher

(2005). Multilevel correlates of childhood physical aggression and prosocial behaviour. Journal of Abnormal Child Psychology, 33, 565–578.

73.

Samuelson

Y. M.

Hodgins

Larsson

Larm

Tengstrom

(2010). Adolescent antisocial behaviour as predictor of adverse outcomes to age 50. A follow-up study of 1,947 individuals. Criminal Justice and Behaviour, 37, 158–174.

74.

Sanders

R. H.

Han

Baker

J. S.

Cobley

(2015). Childhood obesity and its physical and psychological co-morbidities: A systematic review of Australian children and adolescents. European Journal of Pediatrics, 174, 715–746.

75.

Santiago

C. D.

Wadsworth

M. E.

Stump

(2011). Socioeconomic status, neighbourhood disadvantage, and poverty-related stress: Prospective effects on psychological syndromes among diverse low-income families. Journal of Economic Psychology, 32, 218–230.

76.

Sawyer

M. G.

Miller-Lewis

Guy

Wake

Canterford

Carlin

J. B.

(2006). Is there a relationship between overweight and obesity and mental health problems in 4- to 5-year-old Australian children? Ambulatory Pediatrics, 6, 306–311.

77.

Seah

S. L.

Ang

R. P.

(2008). Differential correlates of reactive and proactive aggression in Asian adolescents: Relations to narcissism, anxiety, schizotypal traits, and peer relations. Aggressive Behavior, 34, 553–562.

78.

Seyedamini

Malek

Ebrahimi-Mameghani

Tajik

(2012). Correlation of obesity and overweight with emotional-behavioral problems in primary school age girls in Tabriz, Iran. Iran Journal of Pediatrics, 22, 15–22.

79.

Sullivan

T. N.

Farrell

A. D.

Kliewer

(2006). Peer victimization in early adolescence: Association between physical and relationship victimization and drug use, aggression, and delinquent behaviours among urban middle school students. Development and Psychopathology, 18, 119–137.

80.

Timmermans

van Lier

P. A. C.

Koot

H. M.

(2008). Which forms of child/adolescent externalizing behaviours account for late adolescent risky sexual behaviour and substance use? Journal of Child Psychology and Psychiatry, 49, 386–394.

81.

Tremblay

R. E.

Schaal

Boulerice

Arseneault

Soussignan

R. G.

Paquette

Laurent

(1998). Testosterone, physical aggression, dominance, and physical development in early adolescence. International Journal of Behavioral Development, 22, 753–777.

82.

van Geel

Vedder

Tanilon

(2014). Are overweight and obese youths more often bullied by their peers? A meta-analysis on the relation between weight status and bulling. International Journal of Obesity, 38, 1263–1267.

83.

Vassallo

Smart

Sanson

Dusseyer

McKendry

Toumbourou

J. W.

… Oberklaid

(2002). Patterns and precursors of adolescent antisocial behaviour (first report). Melbourne, Victoria: Crime Prevention Victoria.

84.

Wang

Lim

(2012). The global childhood obesity epidemic and the association between socio-economic status and childhood obesity. International Review of Psychiatry, 24, 176–188.

85.

Williams

Toumbourou

Williamson

Hemphill

Patton

(2009). Violent and antisocial behaviours among young adolescents in Australian Communities. West Perth, Australia: Australian Research Alliance for Children & Youth.

86.

Wilson

S. M.

Sato

A. F.

(2013). Stress and paediatric obesity: What we know and where to go. Stress Health, 30, 91–102.

87.

Ding

Hou

Mao

(2015). Socio-economic position as an intervention against overweight and obesity in children: A systematic review and meta-analysis. Scientific Reports, 5, 1–11.

88.

Zeller

M. H.

Reiter-Purtill

Ramey

(2008). Negative peer perceptions of obese children in the classroom environment. Obesity, 16, 755–762.