Welcome to Your Child's Brain: How the Mind Grows From Conception to College (46 page)

The relationship between SES and health may be attributable to the effects of stress, which can damage the brain and the rest of the body (see
chapter 26
). In many species, life at the bottom of the dominance hierarchy involves chronic stress and a poorly functioning biological stress response system. You could imagine that animals with poor stress responses are just more likely to become subordinate, but researchers found that social subordination occurs first and causes poor stress responsiveness, and not the other way around.

It can be most stressful to be a high-ranking animal in some species or under special circumstances, for instance, when dominance can be maintained only by fighting a lot. But in people, it’s usually the low-ranking members of society who experience the most stress. Social status is so important to people that reducing the power or status of middle-SES adults in an experimental situation decreases their ability to concentrate, ignore distractors, and inhibit inappropriate behavior. We speculate that chronically low social status may have a similar effect on low-SES children. In one study, by age ten, children in Montreal already showed a sharp relationship between SES and cortisol, with blood levels twice as high in the lowest-SES children as in the highest-SES children.

How we interpret the circumstances of our lives also has a strong effect on our stress responses (see
chapter 26
)—often stronger than the effects of our actual economic circumstances. Low-SES people not only experience more chronic stresses and negative life events, but also experience ambiguous events as being more stressful, compared to higher-SES people. When people are asked to give their own position in society on a drawing of a ten-rung ladder, their ranking is a stronger predictor of health than their actual SES. People who are satisfied with their standard of living and feel financially secure are healthier, regardless of their actual income, occupation, and education, than people who are unsatisfied and anxious about the future. Along the same lines, countries, states, or cities with greater income inequality have steeper gradients of SES versus health. This may
be because income inequality interferes with the feeling of community, which provides many types of social support to counteract stress. Increased crime also correlates with income inequality—again, better than with absolute poverty. So the existence of strong inequality in society may be a major driver of stress.

Which parts of children’s brains are damaged by deprivation? We know from animal studies that chronic stress can cause structural changes in the hippocampus and amygdala (see
chapter 26
). In people, low subjective SES and other sources of chronic stress are linked to reduced hippocampal volume. Long-term memory, which depends on hippocampal function, is impaired in low-SES populations. In experimental animals, chronic stress can cause neurons to die, prevent new neurons from being born or surviving, and cause dendrites to become less complex (a change that is reversible) in the hippocampus. Scores on a variety of language tests also vary strongly with SES, perhaps due to the less complex language environment provided by low-SES parents (see
chapter 6
).

In people, the perception of low SES is associated with stronger activity in the amygdala in response to threats. That’s understandable; if you believe that you are low on the totem pole, it’s natural to feel vulnerable and therefore respond strongly to danger. Indeed such increased vigilance may reflect a sensible reaction to real dangers in the environment. The amygdala is important for rapid processing of events that induce fear and other emotions (see
chapter 18
), and it is extensively interconnected with the stress response system.

Across the life span, from infants to adults, low SES predicts decreased executive function, perhaps because the environment offers fewer opportunities to strengthen these abilities through practice. The medial prefrontal cortex (including the anterior cingulate and orbitofrontal regions) is an important inhibitor of the stress system. In experimental animals and people, chronic stress reduces the size of the prefrontal cortex. This brain region is involved in working memory and planning and organizing behavior (aspects of executive function), and it is also necessary for learned suppression of fearful reactions to situations that are no longer dangerous. People who perceive themselves as having low SES have reduced volume in one part of the anterior cingulate cortex. One promising intervention for low-income preschool children, Tools of the Mind, focuses on promoting behaviors that depend on the prefrontal cortex (see
Practical tip: Imaginary friends, real skills
).

The causes and possible solutions to the SES-health gradient are hotly debated,
within the scientific community as well as in society. The key problem for research is that people aren’t randomly assigned to be poor, so we can’t draw conclusions about causality by comparing the characteristics of low-SES and high-SES people (see
Did you know? Epidemiology is hard to interpret
).

The existence of strong inequality in society may be a major driver of stress.

Do people develop problems because they’re disadvantaged? Or do they become (or remain) disadvantaged due to poor health or other problems? There is evidence in favor of both positions. The health of adopted children is best predicted by their adopted parents’ income, not their biological parents’ income, suggesting that family income can influence health independently of genetics. Along the same lines, childhood SES predicts adult health, as we discuss below. On the other hand, the adult income and (particularly) education of adopted children does depend partly on their biological parents’ characteristics.

It’s important to remember that these two classes of explanations aren’t mutually exclusive. Indeed, the most likely relationship between poverty and achievement is a vicious cycle, in which starting life with few resources leads children to develop a variety of problems, which then make their life situation worse, reducing their resources (and their children’s resources) still further.

Some of the relationship between SES and cognitive achievement may be attributable to exposure to environmental hazards, more common in poor neighborhoods, that can cause substantial, lasting impairment in brain function. Children exposed to lead before or during elementary school age have lower IQs and impulse control, as well as higher aggression and delinquency, compared with children of the same SES. All these problems persist through adulthood. Mercury exposure also reduces IQ, along with attention, memory, and language development.

Children who live in noisy environments, such as near airports or highways, are delayed in learning to read compared with other children of the same SES. Chronic noise exposure also causes deficits in attention and long-term memory, perhaps because it is known to increase stress hormone levels. Crowded or chaotic environments (at home or at school) impair cognitive development and academic performance and increase psychological distress in both parents and children, again independent of SES. These environmental conditions are all common in the lives of low-SES children and often occur together.

DID YOU KNOW? EPIDEMIOLOGY IS HARD TO INTERPRET

The tools of epidemiology, appropriately, are best suited to the study of epidemics, which are caused by a single factor (a germ). The same tools are increasingly used to study conditions like heart disease, which have far more complex causes. Epidemiological studies of this kind are far more difficult to interpret and should be approached with a skeptical eye.

In a typical epidemiology study, scientists collect data on a large group of people for years and then attempt to correlate risk factors, such as excessive drinking, with health outcomes, such as deaths due to injury. Studies of this sort have serious limitations, which are rarely taken into account in your local newspaper or when health agencies make lifestyle recommendations based on their findings.

It is almost impossible to draw reliable conclusions about cause and effect from correlation data. One pitfall is reverse correlation. For instance, obesity is correlated with poverty. Does poverty lead to poor diet and lack of exercise, which then cause obesity, as is commonly assumed? Or might obesity cause poverty due to wage discrimination against fat people? Another pitfall is that an additional (unstudied) factor might cause both parts of the correlation. Harsh parenting is correlated with later antisocial behavior. Does that mean harsh parenting causes antisocial behavior? Or could it be that some parents pass along a genetic tendency to antisocial behavior to their children, who then are likely to misbehave, evoking harsh parenting, even from adoptive parents? We did not invent these two examples. In both cases, there is good evidence for the second interpretation, at least as a partial explanation of the observed correlations (see
p. 151
for more information on the effects of harsh parenting).

Making interpretation even more difficult, risk factors tend to travel in packs. Postmenopausal women taking hormone replacement therapy have fewer heart attacks than other women, but they are also less likely to die from homicide or accidents—effects that are unlikely to be caused by hormones. The explanation is that women who take hormone replacement therapy typically have a variety of healthy characteristics: compared to other women, they pay more attention to their health, exercise more, and are richer, more educated, and thinner. When the risk factors are correlated with each other, it becomes very difficult to sort out causes from accidental “bystander qualities,” even if the observed correlations are strong.

Epidemiology can be very useful. The link between cigarette smoking and lung cancer was established through this technique because the correlation is large (heavy smokers have twenty or thirty times more risk than non-smokers) and the rate of lung cancer in nonsmokers is low. Many side effects of approved drugs have also been identified by epidemiology. But most lifestyle effects are small to moderate, and most of the common diseases in developed countries are influenced by multiple factors. Under those conditions, epidemiology can only generate hypotheses that must be tested by other means. Such studies should be interpreted with care and caution.

Growing up in a low-SES family predicts poor health even for children whose SES improves in adulthood. For example, in a group of nuns who had been living together since early adulthood, disease risk and longevity still varied depending on their education (whether or not they had gone to college). For more than fifty years, the nuns had shared meals, housing conditions, and a very similar lifestyle, but the traces of their early experiences were still substantial, with educated sisters living an average of 3.28 years longer than less educated sisters. In general, people whose SES improves later in life gain less advantage from the change than people whose SES improves in childhood.

Children whose families move out of poverty improve in some areas but not others. One study followed 1,420 poor children in North Carolina from 1993 (at ages nine to thirteen) through 2000. American Indian families were more than twice as likely as non-Indian families to be below the poverty line when the study began. In 1996, a casino opened and began to distribute some of its profits to every person on the reservation. Children whose families moved above the poverty line showed a 40 percent decrease in antisocial behaviors during the study, while children whose families remained poor showed no change in antisocial behaviors.
In contrast, moving out of poverty had no effect on symptoms of depression and anxiety, though children who had never been poor had fewer symptoms than always-poor or ex-poor children.

If indeed poverty leads to a vicious cycle like the one we’ve described, it should be easiest to break that cycle in young children, before they fall too far behind their peers. Intensive preschool enrichment programs can have positive effects that last into adulthood, substantially increasing the odds of a poor child graduating from high school, finishing college, getting a skilled job, and owning a home. These programs can also reduce the likelihood that a child will need special education or repeat a year of school.

Mostly these effects do not depend on increasing children’s IQs. Instead the positive outcomes seem to stem from improvements in social competence, including perseverance and motivation (see
chapter 13
) and emotional well-being. The programs that produce these results tend to be extensive, long-lasting interventions, which require a considerable commitment from both families and funding agencies. These programs are often still cost-effective for society in the long run if they reduce the likelihood that children will need special education or repeat a year of school or that they will receive welfare payments as adults.