impulse control—seen in the peer-reared monkeys (Barr et al. 2004). Such studies may elucidate the neural circuits governing the stress response and how they interact with the environment and on some of the variation in response to stress.
In the 1950s, Harlow found that lack of maternal care and handling is a severe form of early-life stress. Rhesus monkeys that had been “raised” by artificial mothers (made of bare wires) were stricken with terror when they were placed in novel situations (Harlow 1974), and those raised by their peers showed more anxiety and hyperarousal than monkeys raised by their biologic mothers (Mineka and Suomi 1978).
Early-life stress has been associated with increases in cortisol and other markers of increased HPA-axis activity (Levine 1962). In rats, normal maternal handling during infancy leads to a life-long increase in the number of glucocorticoid receptors in the hippocampus, whereas a lack of maternal handling decreases the number of glucocorticoid receptors in the stressed animals, and the decreases persist into adulthood and old age (Meaney et al. 1985, 1988). A greater density of glucocorticoid receptors after normal handling would be expected to increase negative feedback between the hippocampus and the HPA (see Figure 4-1) and result in greater inhibition of the HPA axis after a stressful event, which in turn would lead to a less reactive HPA axis, lower cortisol concentrations, and more rapid initiation of the stress response. But the lower density of receptors in the heavily stressed offspring would be expected to reduce negative feedback and lead to greater reactivity of the HPA axis, higher cortisol concentrations, and more prolonged stress response. Higher cortisol concentrations, which persisted from youth through old age, were indeed found in nonhandled (stressed) animals. At greater ages, the excess secretion of cortisol was associated with structural changes in the hippocampus and with deficits in spatial memory (Meaney et al. 1988). Normal maternal care led to lower concentrations of corticotropin and cortisol, indications of a less reactive HPA axis (Liu et al. 1997; Meaney et al. 1985; Sapolsky et al. 1986).
Evidence that early-life stress results in an overreactive HPA axis has come also from studies of CRH. As described earlier, CRH is released by the hypothalamus under the regulatory influence of the hippocampus, prefrontal cortex, and amygdala, and it signals the pituitary gland to release corticotropin, which leads to a release of cortisol by the adrenal gland (see Figure 4-1). Primates reared under conditions of early-life stress (unpredictable conditions for the mother to find food) displayed persistent increases of CRH throughout adulthood (Coplan et al. 1996). The timing of exposure to the stressor was important: CRH decreased when the stressful condition occurred later in infancy (Mathew et al. 2002). Those studies have helped to establish that early-life stress has permanent effects on the regulation of the HPA axis.
Human studies have demonstrated associations that appear to be consistent with the findings on early-life stress in animal studies. In a study of the effects of childhood abuse in New Zealanders followed from birth until their 30s, an association was found between childhood abuse and chronic inflammation in adulthood. The inflammatory marker C-reactive protein—a better marker for myocardial infarction than cholesterol concentrations—was found to be above normal in many of the young adults known to have been abused. A dose-response relationship was seen between the level of abuse and the concentration of C-reactive protein (Danese et al. 2007).