Linking psychosocial stressors and childhood obesity
C Gundersen, Department of Agricultural and Consumer Economics, University of Illinois,324 Mumford Hall, 1301 West Gregory Dr., Urbana, IL 61801-3605, USA. E-mail: firstname.lastname@example.org
Research has established a wide array of genetic and environmental factors that are associated with childhood obesity. The focus of this review is on recent work that has established the relationship between one set of environmental factors, stressors and childhood obesity. These stressors are particularly prevalent for low-income children, a demographic group that has high rates of obesity in the USA and other developed countries. In this review, we begin by summarizing the psychosocial stressors faced by children followed by health outcomes associated with exposure to these stressors documented in the literature. We then summarize 11 articles which examined the connection between psychosocial stressors in the household and obesity and eight articles which examined the connection between individual psychosocial stressors and obesity. Policy recommendations emerging from this research include recognizing reductions in childhood obesity as a potential added benefit of social safety net programmes that reduce financial stress among families. In addition, policies and programmes geared towards childhood obesity prevention should focus on helping children build resources and capacities to teach them how to cope effectively with stressor exposure. We conclude with suggestions for future research.
Healthy People 2010(1) and President Obama (2) have identified childhood obesity as a national health priority as it has immediate consequences for a child's mental and physical health (1,3–5), as well as implications for future health (1,4–6). Addressing childhood obesity has therefore become an even higher profile concern than in past decades (7–9). To effectively address this growing concern, it is crucial to understand the relevant genetic and environmental factors associated with childhood obesity. In this review, we focus on one factor that has received increasing recognition – the role of psychosocial stressors.
We adopt the definition of psychosocial stressors as external events or conditions that threaten an individual's well-being (10–14) with a focus on the well-being of children and adolescents. Psychosocial stressors can include both household and individual-level events and conditions. Household level events and conditions include parental divorce (15,16), poor parental marital quality (17,18), poor parental mental health (19,20), chronic physical health conditions of family members (21,22), domestic violence (18,23,24), child abuse (24–27) and general relationship strain amongst family members (28). Events and conditions at the level of the individual include engaging in risky behaviours (29,30) and having poor mental health (31,32). In both instances, psychosocial stressors represent a barrier to the mental and physical capacities of an individual (10,11). Indeed, in addition to its effect on childhood obesity as covered below, extant literature shows that psychosocial stressors are associated with a wide array of negative health outcomes among children and adolescents (33–35).
The potential strain on the body because of stressors is particularly important to consider for children, who are still developing biologically, cognitively and socially (23,36). Additionally, understanding the impact of stressors on children's development, particularly their weight status, may be especially important for children growing up in low-income households as psychosocial stressors are more common among those living in poverty than for those who are not (19,37–39). The effects of stressors on low-income children is especially relevant in light of the substantially higher rates of childhood obesity found among low-income children across the globe (40–47). The relationship between low socioeconomic status and childhood obesity is complex as illustrated in the reviews available in the epidemiological literature (48–52). Thus, the role of socioeconomic status must be distinguished from the events and conditions associated with that particular position (39).
We emphasize psychosocial stressors as a way to differentiate the fact that this review addresses issues of stress rather than disadvantage more generally. This review opens with a brief discussion of the health outcomes commonly ascribed to experiencing these stressors. We then consider the connections between psychosocial stressors and childhood obesity. We conclude with policy recommendations and remarks on future research directions.
A multi-database search was conducted for studies examining various stressor indicators and processes related to the mental and physical health of children (2 years of age and older) and adolescents (under the age of 18) and links to obesity specifically. Searches were limited to publications within the last 20 years from journals in the fields of behavioural medicine, developmental psychology, family studies, health sciences, public health and sociology using PsychINFO, PubMed and Web of Science. The following keywords guided the searches: psychosocial stressors and child/adolescent health, allostatic load and child/allostatic health, family stress and child/adolescent health, stress and child/adolescent health, stress and obesity. Studies were then selected based on whether (i) the sample included human subjects, specifically families, children or adolescents; (ii) analyses examined how psychosocial stressors were related to mental or physical health outcomes (developmental disabilities were omitted as an outcome); (iii) the measures for stressors cohered with the definition adopted in this review and (iv) articles were published in an English language journal.
Even within the literature examining the associations between stress, mental and physical health outcomes, and obesity, there are aspects that we do not cover in this review such as the association of the built environment and obesity (8,53–56), stress and eating patterns (57–64), and stress and activity involvement (65–68). These are all important research areas, but they are beyond the scope of this review that focuses specifically on psychosocial stressors and childhood obesity.
The body's response to stressor exposure
Mental health outcomes
The disruptions caused by psychosocial stressors, such as the household and individual events and conditions previously mentioned, often lead to mental health concerns. Indeed, a common consequence of stressor exposure is poor mental health, particularly depression (24,30,69) and anxiety (29,70,71). In addition, mental health outcomes vary between socioeconomic groups, with low-income children having a greater likelihood of exhibiting psychological and behavioural problems as a result of overexposure to a variety of psychosocial stressors (33–35).
Mental health problems pose unique negative effects on a child's well-being, and may set the stage for the development of other negative outcomes. For example, Conger and colleagues' (37) family stress model illustrates how parents' exposure to stressors shapes parenting behaviours, which in turn influence child outcomes such as mental health. The strained family relationships that often accompany mental health issues create a stressful atmosphere in the home, which may exacerbate and perhaps exploit the mental health issues. How children and adolescents adjust to stressors has implications towards their mental health as well. For example, whether the individual perceives a stressor as controllable and their strategy for coping with it can diminish or enhance potential the negative effect of the stressor on mental health (10,11,71,72). Li and Rukavina (73) review different ways that overweight or obese children may cope with weight stigma.
Physical health outcomes
Reporting exposure to greater levels of psychosocial stressors has also been associated with higher levels of self-reported illness (32,74,75). Poor physical health is illustrative of how elevated cortisol responses wear away at the biological systems in place to effectively cope with stress (76,77). Another physical health outcome from exposure to stressors is dysregulated cardiovascular reactivity (22,78). In addition, allostatic load research indicates that poor health is a product of damage because of early and chronic exposure to stressors such as those described in the previous sections (21,36,78,79). Specifically, stressors trigger a response from the hypothalamic-pituitary-adrenal (HPA) axis, which regulates the corticotrophin releasing factor and the feedback of cortisol, an important stress hormone (27,80). Over time, an individual over-exposed to stressors has a blunted response to stress resulting in decreased cortisol levels and other associated neurobiological dysfunction (36,77).
These physical health reactions to stressors have also been linked to weight gain in children via direct metabolic changes and maladaptive coping behaviours, such as lack of exercise and over-eating (81,82). Moreover, research has found that chronic arousal of the HPA axis disrupts metabolism. As cortisol levels increase energy intake and central adiposity also increase despite elevated leptin levels (82–84). Leptin is involved in regulating satiety suggesting that increased cortisol levels may disrupt the leptin system and thus increase the risk of obesity (83,84). The majority of the research addressing the links among exposure to psychosocial stressors and childhood obesity has failed to include these physiological pathways in their models. As a result, we review literature examining the basic connections between exposure to psychosocial stressors and childhood obesity.
Exposure to stressors and childhood obesity
Psychosocial stressors in the household and obesity
While parents' body mass index (BMI) or overweight status is often the strongest predictor of a child's BMI (7,85), some families may have characteristics that make them more vulnerable to psychosocial stressors and subsequent health problems such as obesity. Indeed, researchers have found that households characterized as having less family cohesion, more conflict and disruptive home environments increase the child's risk of being overweight/obese (86,87). The association between psychosocial stressors and obesity in the family domain gives support for the family stress model (37) that links the psychosocial environment to child outcomes. While past research has looked at the connection between family mental health and interactions to child internalizing and externalizing outcomes, the literature reviewed in this paper suggests that physical health outcomes such as overweight or obesity can also be influenced by psychosocial stressors that are present in the family environment.
Eleven empirical articles examining psychosocial stressors in the household and obesity were reviewed and are shown in Table 1. These eleven articles represent the total literature that has examined this relationship specifically and cohered with the definition of stressors adopted in this review. Of the 11, six explicitly defined measures as household or family stressors (45,88,94–97); four defined measures as environmental or social factors (89–92) and one defined measures as indicators of household or family dynamics (93).
Table 1. Obesity studies with household stress antecedents
|Crossman et al. (2006) (92)||National Longitudinal Study of Adolescent Health|
n = 6378 (7th–12th graders)
USA (nationally representative)
1995; obesity measured 6 years later
Ordinal regression models
|Family closeness negatively related to overweight in males, negatively related to overweight in females|
Families that promote good weight-related habits have positive/protective effect on future weight
Females with lower self-esteem have a higher risk for excessive weight in young adulthood regardless of current weight status and family environment
|Garaskyet al. (2009) (88)||Panel Study of Income Dynamics: Child Development Supplement|
n = 1136 (5–12 years)
n = 1001 (15–18 years)
USA (nationally representative)
Ordered probit models
|Family disruption and conflict|
Caregiver's mental and physical health problems
Health care struggles
|For younger children, lack of cognitive stimulation and emotional support positively linked to being overweight/obese|
For older children, family financial strain and the mental and physical health problems of the caregiver were positively associated with being overweight/obese
|Gibson et al. (2007) (93)||Childhood Growth and Development Study|
n = 329 children (6–13 years old)
|Maternal mental and physical health|
Negative life events
|Maternal BMI and family structure significantly associated with child BMI|
Treatment seeking families more socially disadvantaged and higher levels of maternal depression
|Gundersen et al. (2008) (94)||National Health and Nutrition Examination Survey|
n = 841 (3–17 years)
USA (nationally representative)
|Maternal mental and physical health|
|Food secure children whose mothers experienced stress were more likely to be overweight or obese|
|Koch et al. (2008) (45)||All Babies in Southeast Sweden Project|
n = 7443
1997–1999 for initial data collection; obesity measured 5 years later
|Serious life events|
Parenting stress and worries
|Children from households that reported stress in at least 2 domains were more likely to be obese|
|Lissau and Sorensen (1994) (90)||Pupils in Copenhagen Municipality|
n = 881 (9–10 years)
1974 for initial data collection; obesity measured 10 years later
Perceived support by parents
|Children receiving no support had a higher risk of becoming obese in young adulthood|
Children considered to be neglected had about a 10-fold increase in the likelihood of becoming obese in young adulthood
|Lohman et al. (2009) (95)||Welfare, Children and Families: a Three City Study|
n = 1011 (10–14 years)
Boston, MA; Chicago, IL; San Antonio, TX
Family functioning, disruption, conflict
Maternal physical and mental health
|Food insecure child's probability of being obese increase as maternal stressors increase in the household|
|Moens et al. (2009) (96)||Overweight children seeking treatment plus control group|
n = 197 (6–14 years)
Location not specified
Year not specified
|No effect of stress on child's BMI|
|Stenhammar et al. (2010) (97)||Children born during 5-month time period|
n = 873
Uppsala County, Sweden
|Swedish Parenthood Stress Questionnaire||Child's probability of being overweight increase due to mother's reports of stress|
|Strauss and Knight (1999) (89)||National Longitudinal Survey of Youth|
n = 2913 (0–8 years)
USA (nationally representative)
1979 for initial data collection; obesity measured ten years later
|Demographic and socioeconomic status variables: low income and blue-collar parental occupation predictors of childhood obesity|
Low HOME cognitive scores increases risk for obesity: rearing in an environment with low-average cognitive stimulation had increased risk of developing obesity
Single mother headed households increase risk- Low-education mother (no high school degree) increase risk
|Zeller et al. (2007) (91)||Clients of a paediatric weight clinic plus control group|
n = 149
Undisclosed location in U.S.
Year of data collection not disclosed
|Family environment scale|
About Your Child's Eating-Revised
Parent psychological distress
|Obese children had mothers who reported higher levels of psychological distress|
Maternal distress increased a child's odds of being obese by twofold and mediated the relationship between child obesity status and poor family interactions
Mothers of obese children reported more challenging mealtime interactions and higher levels of family conflict
The outcome measure in all 11 articles was the risk of overweight or obesity as measured by age–gender derived BMI percentiles for children or by BMI. BMI cut-off points from the Center for Disease Control were used in (88–92,94–96), BMI cut-off points from the International Obesity Task Force were used in (45,97) and cut-off points were not specified in (90). Two (92,97) of the studies used self-reported measures of height and weight, while the remaining studies used measures that were collected by trained personnel.
With the exception of Moens (96), at least one measure of stress was associated with obesity in every study. The measures of stress associated with obesity included low self-esteem (92), financial strain (88), maternal depression (93), indices of stress which incorporates multiple dimensions (45,88,95), neglect (90), maternal self-reports of stress (97) and maternal distress (91,94). All except for (90) used established scales, and all but (88,90,93–94) discussed the construct validity of the stressor measures.
In evaluating the finding that stressors are associated with childhood obesity, several issues may be relevant. First, all of the studies controlled for relevant demographic confounders. Of particular importance is that (45,88,89,91–97) controlled for maternal or parental BMI, a variable that has consistently been found to have a major association with childhood weight status. Second, the association of stress and obesity were measured contemporaneously in (88,91,93–97) and with a lag of at least a few years in (45,89–91). Third, the data used in the models were collected using two main methods. In (45,88–90,92–95,97), the data were a sample that was designed to be representative of a population. In contrast, in (91,96), the data were collected from children who were seeking treatment for problems associated with obesity and a control group was then taken from similar, non-overweight children. While the quality of the articles used in this review has been well established, there are, as always, remaining issues that the literature may wish to address. We turn to those in our concluding remarks.
Individual psychosocial stressors and obesity
At an individual-level, some children may face substantial amounts of stressors, which may make them vulnerable to being overweight or obese. To date, very few papers have examined the association between individual child stress and obesity specifically. Much of the literature has focused on children's behavioural correlates, such as physical activity e.g. (7,41,56,98,99) or television viewing (7,100), and their association with obesity. We reviewed eight empirical articles that look specifically at the relationship between psychosocial stressors and obesity (31,32,46,95,101–104); a summary of the studies is displayed in Table 2.
Table 2. Obesity studies with child stress antecedents
|Falkner et al. (2001) (101)||Middle and high school students|
n = 9943 (7th–11th graders)
|Compared with non-obese adolescents, obese adolescents reported spending less time with friends, more serious emotional problems, and were more likely to consider themselves poor students|
|Franko et al. (2005) (31)||National Heart, Lung, and Blood Institute Growth and Health Study|
n = 1554
Berkeley CA; Cincinnati, OH; Rockville, MD
1987 for initial data collection; obesity measured 5 years later
OLS, logistic regression
|Depressive symptoms||Depressive symptoms at age 16 significantly associated with BMI/obesity at age 21|
Depressive symptoms at age 18 somewhat associated with BMI/obesity at age 21
Maximal depressive symptoms across age 16-18 sig. associated with obesity at age 21
|Goodman and Whitaker (2002) (32)||National Longitudinal Study of Adolescent Health|
n = 9374 (7th–12th graders)
1995; obesity measured 1 year later
|Depression at baseline significantly linked to future obesity even after controlling for adolescent smoking, self-esteem, delinquent behaviour and physical activity|
|Lohman et al. (2009) (95)||Welfare, Children and Families: a Three City Study|
n = 1011 (10–14 years)
Boston, MA; Chicago, IL; San Antonio, TX
|Poor academic performance|
Lack of future orientation
|Increase in adolescent stress had a significant and direct effect on the probability of obesity|
|Mellbin and Vuille (1989) (46)||Children in Sweden with records on birth weight and weight at age seven|
n = 971 (7 years)
1963; obesity measured 10, 13 and 15 years later
|Social, physical and psychological problems|
|Individuals with higher psychosocial stress levels were found to increase their weight by age 15|
|Pine et al. (2001) (102)||Children diagnosed with major depression plus control group|
n = 268 (6–17 years)
New York, NY
Year of data collection not disclosed; obesity measured 10–15 years later
|Depression||Individuals who were depressed in childhood have a larger BMI than individuals who were psychiatrically healthy in childhood|
Childhood depression predicted a two fold increase in risk for overweight status in adulthood
|Sweeting et al. (2005) (103)||West of Scotland 11 to 16 Study|
n = 2127 (11–15 years)
Year of data collection not disclosed; obesity measured 4 years later
|Stably obese had lower self-image , mood, and behavioural disorders, but higher anxiety|
Becoming obese was predicted by lower self-image and higher victimization
|Van Jaarsveld et al. (2009) (104)||Health and Behaviour in Teenagers Study|
n = 4065 (11–16 years)
1999–2003; obesity measured over the next 7 years
|Perceived stress scale|
|Experiencing persistent high or moderate stress over a 5-year period was significantly associated with higher waist and BMI standard deviation scores|
Although the high-stress group had the highest adiposity, findings did not lend support that higher perceived stress resulted in faster weight gain
The outcome measure in all eight articles was the risk of overweight or obesity as measured by age–gender derived BMI percentiles for children or by BMI. BMI cut-off points from the Center for Disease Control were used in (31,32,95,101,102), BMI cut-off points from the International Obesity Task Force were used in (103,104) and cut-off points were not specified in (46). All but one (101) of the studies used measures that were collected by trained personnel rather than self-reports.
Research investigating the influence of children's psychosocial stressors on obesity has often examined just the association between children's mental health and obesity (See Liem et al. (105) for a more in-depth review). Half of the papers included in this review follow this line of research and primarily investigate the influence of depression on obesity (31,32,102,103). In all four papers, childhood depression was positively linked to these children becoming overweight or obese as adults. Depression was measured using established scales, and the reliability of the measure was discussed in all but (31).
The remaining four papers (46,95,101,104) look beyond depression and more broadly examine a child's level of psychosocial stressors and its influence on obesity. In addition to measures of mental health, the indicators of psychosocial stressors in the papers included measures of academic performance (95,101), future life goals (95,101), substance use (95), physical health (46), and overall perceptions of stress (104). All four papers found that experiencing higher levels of psychosocial stressors was associated with an increased likelihood of being overweight or obese. Studies (95,104) discussed the reliability of the stressor measures, which were derived from established scales. In (46,101), the stressor measures were study-specific.
As with the articles examining household level stressors and obesity, several issues were considered when evaluating the finding that stressors are associated with childhood obesity. First, all the studies controlled for relevant demographic confounders such as age, race, socioeconomic status with (95,102,104) also controlling for health behaviours and parents' obesity status (32,95). Second, with the exception of (95,102), the data used in the studies were collected using school-samples designed to be representative of a population. All but (95,101) examined the longitudinal associations between stressors and obesity, which bears important implications in developmental research in identifying potential critical periods for stressor exposure and the development of childhood obesity
As noted above, in reviewing the literature on the connection between stressors and childhood obesity, our choice of studies is based on their use of measures of stress. In each of the studies we reviewed, there was at least some evidence of the relationship between stressors and childhood obesity. We do recognize, however, that there is often a publication bias with respect to studies finding statistically significant relationships. In light of this, there may be other studies that do not show an influence between stressors and childhood obesity that have not been published.
The effects of stressor exposure on a wide variety of health outcomes for children and adolescents have been well established in the literature. In addition, alongside the numerous other genetic and environmental factors that have consistently been shown to be central to whether a child is obese (7,8), this review covers an additional important factor that matters for childhood obesity, stress. While efforts to reduce the numbers and types of stressors that children are exposed to can be justified on many grounds, the results reviewed here demonstrate that one additional benefit to reducing stressor exposure is a reduction in childhood obesity.
This should be taken into consideration when judging the costs and benefits associated with different social policies and programmes. In particular, insofar as social safety net programmes help reduce financial stress, reductions in childhood obesity should be included as an added potential benefit of these programmes. Policies and programmes geared towards childhood obesity prevention should focus on helping children build resources and capacities to not only decrease the amount of stressors they are exposed to in childhood, but also teach them how to cope effectively with this stressor exposure. Finally, broad health education should be provided including information about the powerful impact exposure to stressors can have on a child's mental and physical health, including their weight. For instance, comprehensive wellness programmes in schools or community centres might also include health and nutrition education and services, healthier food options and activities that teach children adaptive coping strategies. As part of these wellness programmes, the critical importance of exercise should be emphasized. This emphasis is warranted in-and-of-itself (i.e. exercise improves well-being) but also helps to reduce the stress faced by children. Overall, the results of this research have implications for policy in terms of providing more incentive to reduce the stressors to which children are exposed on a daily basis and identifying another avenue for addressing childhood obesity.
The above suggestions deal with interventions at the child and household level. Additionally, as noted most recently by Wilkinson and Pickett (106), societies with greater inequality also have higher incidences of negative health outcomes. Consistent with this, the increasing economic inequality experienced in the USA over the last few decades, and the accompanying increases in stress for some families, may be one further reason for the increase in obesity among US children. As such, alongside the other potential benefits to macroeconomic efforts to reducing inequality, one further benefit may be the reduction in childhood obesity.
These policy conclusions are based on the research summarized here. Nevertheless, stress–obesity research is still in its infancy and more concrete policy conclusions require additional research. To this end, we offer five suggestions for additional research. First, research in this area has not sorted out the possible associations of stressors with unobserved factors across children. As a consequence, what is ascribed to stressor exposure may actually be attributable to unobserved factors such as unique traits among children that allow them to cope with stressors differently. The use of instrumental variable analyses and fixed-effects models can address this problem to some extent. Second, research in this area has implicitly assumed that stressors influence obesity, but the converse does not hold. Given the stigma that is often associated with obesity, one may imagine that having an overweight child can lead to more stressors in a household. This increase in stress may also occur if the level of overweight is such that additional medical expenses are incurred, leading to financial strain in the household. Simultaneous equation modelling may help untangle the joint relationship between stressor exposure and obesity. (Research in this area has been done for adults (107).) Third, research in this area has examined the static relationship between stressors and obesity. Future research must examine the potential trajectories of stressor exposure and obesity. When considering the consequences associated with stress, these trajectories may be especially important as children become young adults. Fourth, as many families currently face economic stress because of the current economic downturn, more needs to be learned about the magnitude of the relationships between stressors and childhood obesity. This is especially true for children experiencing extreme forms of economic stress including not eating for extended periods and homelessness. Fifth, as noted above, households do not all experience stress in the same way. The work reviewed here implicitly assumes, however, that the experience of stress may differ for households. As an example, households with better financial management skills may be better able to address financial stress than less-skilled households. Future research may wish to consider how the experience of stress differs by household. Finally, we urge future researchers to combine survey work with data collection efforts that include biomarkers of the physiological responses of a child to stressor exposure, so that the pathways among stressor exposure, metabolic manifestations and weight change in children and can be better understood.
Conflict of Interest Statement
No conflict of interest was declared.
This research is supported by the USDA, Cooperative State Research, Education, and Extension Service grant number 2007-35215-17871.