Why Socioeconomic Status Affects the Health of Children A Psychosocial Perspective

This article provides an overview of research on socioeconomic status (SES) and physical health in childhood. SES has a gradient relationship with children’s health, such that for each incremental increase in SES, there is a comparable benefit in children’s health. In this article, I discuss psychosocial mechanisms underlying this association and argue that it is important to utilize knowledge about how the relationship between SES and health changes with age to inform a developmentally plausible search for mediators of this relationship. Furthermore, SES at different points in a child’s lifetime may have different effects on health. I advocate an interdisciplinary approach to searching for mediators that would allow researchers to understand how characteristics of society, the neighborhood, the family, and the individual child are involved in the processes linking SES and children’s health. KEYWORDS—socioeconomic status; children’s health; psychosocial One of the most striking and profound findings in epidemiology is that individuals lower in socioeconomic status (SES) have poorer health than individuals higher in SES. This relationship holds true whether health is measured as the prevalence rate of illness, the severity of illness, or the likelihood of mortality, and it is true for most types of diseases, as well as for many risk factors for diseases. This finding has been reported for many countries, including those with and those without universal health care. And it has been demonstrated across the life span, from childhood to older adulthood (Adler et al., 1994; Anderson & Armstead, 1995; Chen, Matthews, & Boyce, 2002). One of the most intriguing aspects of the relationship between SES and health is that it exists as a gradient. That is, it is not just that poor people have poorer health than rich people. Rather, each step increase in SES is accompanied by incremental benefits in health. This gradient makes the search for underlying mechanisms a challenge for researchers. Obvious mechanisms, such as inadequate nutrition, housing, or health insurance, cannot explain why upper-middle-class individuals have slightly poorer health than upper-class individuals. In this article, I discuss psychosocial explanations for the SES-health relationship, with an emphasis on children’s health. I focus here on physical health; however, other researchers have explored these issues for children’s mental health and well-being (see Leventhal & BrooksGunn and McLoyd under Recommended Reading). POSSIBLE PSYCHOSOCIAL PATHWAYS Researchers have suggested many explanations for the effect of SES on health. For example, the effect may be due to genetic influences, environmental exposures to toxins, quality of medical care, and psychological-behavioral factors, just to name a few possibilities (Anderson & Armstead, 1995). Here I provide a brief overview of some of the primary psychological-behavioral factors. Research in this area has focused on individual characteristics that fall into four main categories: stress, psychological distress, personality factors, and health behaviors (Adler et al., 1994; Anderson & Armstead, 1995). With respect to stress, lower-SES children and adults experience more negative life events (stressors) than higher-SES individuals; in addition, they perceive greater negative impact from any given event (stress appraisal). In turn, a large body of literature has linked stress to a wide variety of negative biological and health outcomes in both children and adults. Evidence has documented that stress is one plausible mediator linking SES to health (Cohen, Kaplan, & Salonen, 1999). Thus, one theory is that as one moves down in SES, the amount of stress one experiences increases, which in turn takes a physiological toll on the body, putting one at greater risk for a variety of diseases. A second possibility is that psychological distress plays a role. Because of the social environments in which they grow up, lower-SES individuals may be more prone to experiencing negative emotional states than higher-SES individuals are, and if the experience of negative emotions has biological consequences, this could also lead to poorer health. Previous research has found support for the notion that lower-SES individuals are more likely to experience negative emotions such as depression and anxiety, and that these negative emotions are linked to illnesses, such as cardiovascular disease, as well as to mortality rates (Gallo & Matthews, 2003). Address correspondence to Edith Chen, University of British Columbia, Department of Psychology, 2136 West Mall, Vancouver, B.C. V6T 1Z4, Canada. CURRENT DIRECTIONS IN PSYCHOLOGICAL SCIENCE 112 Volume 13—Number 3 Copyright r 2004 American Psychological Society A third hypothesis is that lower-SES individuals are likely to possess personality traits that are detrimental to health. That is, lowerSES individuals may be more likely than higher-SES individuals to possess certain dispositional traits that are adaptive in the social environments in which they live, but have negative health consequences. For example, living in a dangerous neighborhood may make lower-SES individuals likely to mistrust others and to hold cynical attitudes toward others. Thus, one might expect lower-SES individuals to be more hostile and less optimistic about their future than higher-SES individuals are. In turn, such personality traits have been found to place individuals at increased risk for illnesses (Adler et al., 1994). Finally, compared with individuals of higher SES, those of lower SES may be less likely to engage in healthy behaviors, such as exercising, eating a healthy diet, and not smoking. In part, this may be because of available resources. For example, the availability of healthy products in grocery stores varies by the SES of neighborhoods (Williams & Collins, 2001); people with reduced access to healthy products in their neighborhood grocery stores will have increased difficulty maintaining a healthy diet. Lower-SES neighborhoods also are more dangerous than higher-SES neighborhoods, and less likely to have public parks and venues for exercise (Williams & Collins, 2001); thus, decreases in SES increase the barriers to engaging in regular exercise. These factors are promising possibilities for clarifying the psychosocial reasons why decreases in SES are associated with decreases in health. However, most of these factors focus on the individual. In trying to understand the health of children, it is particularly important to consider the role of factors in the family and the larger environment. In addition, given the vast social, cognitive, emotional, and biological differences between young children and older adolescents, it is important to consider whether the relevance of the various factors depends on the individual’s age. DEVELOPMENTAL TRAJECTORIES Exploring the strength of the SES-health relationship during different periods of childhood may provide insight into pathways linking SES with health. My colleagues and I have argued that the relationship between SES and health may be stronger in certain periods of childhood than others. In trying to understand why this is so, one should consider developmental factors that are important during each period of childhood. Previously, we proposed three models of how the relationship between SES and health may change across childhood (Chen et al., 2002). The childhood-limited model states that relationships between SES and health are strongest in early childhood, and weaken with age. This suggests that factors that are particularly important during early childhood may play a role in explaining health outcomes. For example, the quality of child care, attachment to parents, and housing conditions may be important factors during this period. Research has shown, for example, that injuries are strongly correlated with SES early in childhood, but not during adolescence (West, 1997). It may be the case that unsafe housing conditions are most relevant to young children, who do not have the ability to recognize and avoid danger in their homes, but that as children age and improve in cognitive abilities, they more easily recognize and avoid dangers at home, so the strength of the relationship between SES and injury decreases. The adolescent-emergent model states that relationships between SES and health are weak early in life, but strengthen with age. According to this model, factors that become important during adolescence, such as peer influence or certain personality characteristics, may play a role in the SES-health relationship. For example, physical activity is more strongly correlated with SES during adolescence than earlier in childhood (Chen et al., 2002). One explanation may be that earlier in life, health behaviors are shaped strongly by parents as role models, but as a child ages, peers begin to exert influence on his or her health behaviors. The combination of parent plus peer influence may lead to stronger relationships between SES and health behaviors during adolescence than earlier in childhood. Finally, the persistence model states that relationships between SES and health are similar throughout childhood and adolescence. In such cases, factors that would not be expected to change with children’s age may be important. For example, the correlation between severity of asthma and SES is similar across childhood and adolescence (Chen et al., 2002). One possible explanation for this correlation is that asthma severity is in part determined by a family’s trust in their health care provider. Compared with higher-SES families, lowerSES families may have greater mistrust of the medical community, which in turn may lead to poorer adherence to instructions and advice regarding medications and behaviors for managing asthma. If this psychosocial factor does not change significantly as a child ages, then one would expect to see the relationship between SES and asthma severity follow a persistence model. LONGITUDINAL RELATIONSHIPS In addition to considering the relationship between SES and health at different points during childhood, it is important to understand how SES may change over children’s lives, and what impact these changes have on children’s health. Family SES can fluctuate dramatically from year to year, and a child’s history of SES may affect health differently than current SES does. For example, current SES may affect the quality of health care a family has access to, as well as how they are treated in medical settings. In contrast, history of SES may play a role in the development of health problems. For example, SES effects may accumulate over time. Previous research has shown that amount of time spent in low SES is an important predictor of adult mortality rates (McDonough, Duncan, Williams, & House, 1997), young adults’ self-reported health (Power, Manor, & Matthews, 1999), and cognitive development and behavioral problems in children (Duncan, Brooks-Gunn, & Klebanov, 1994). These findings suggest that it takes time for SES to have effects on health. Some researchers have suggested that there may be critical periods in childhood when SES has its biggest effect. For example, early childhood experiences may program a pattern of biological and behavioral responses that has prolonged effects across the life span. Research has demonstrated that SES early in life is a predictor of adult health behaviors (Lynch, Kaplan, & Salonen, 1997), and that early childhood environments predict adult cardiovascular disease (Barker, 1992). In addition, these relationships persist even after accounting for the effect of adult SES. These findings suggest that it may be important to understand the characteristics of a child’s environment during critical windows in order to understand health consequences later in life. Volume 13—Number 3 113 Edith Chen