ated with two dimensions of child temperament: attention and emotion regulatory difficulties (West and Newman, 2003).

Some of the findings are informative of the developmental perspective, especially of the role of transactional processes. In particular, some researchers have tested the hypothesis that infants’ difficult temperament may increase the likelihood of postpartum depression. In one study, infants’ difficult temperament was related to maternal depression in mothers of 3-month-olds, and this association was mediated by perceived efficacy in the parenting role (Cutrona and Troutman, 1986). In another study, maternal depression and infant negative emotionality interacted to predict maternal reactivity/sensitivity (Pauli-Pott et al., 2000). Positive infant emotionality was not a predictor. Good marital support was directly associated with maternal reactivity/sensitivity, but not as a moderator.

In a direct test of the role of transactional processes, one group of researchers found that maternal depression and difficult infant temperament, but also fear/shyness, may interact to predict negative outcomes for children as adolescents, but the results vary by temperament factor, the child outcome, and gender. For example, in a longitudinal study that spanned ages 5 through 17, exposure to maternal depression in early childhood predicted increases in boys’ externalizing behavior problems over time only among boys whose temperament factor of impulsivity was low (Leve, Kim, and Pears, 2005). Others have found that children with more difficult temperament are more vulnerable to the effects of inadequate parenting, such as that found to be associated with depression in mothers (Goldsmith, Buss, and Lemery, 1997). Mothers of more difficult infants also perceive their parenting to be less efficacious, which in turn is linked to depression in mothers (Cutrona and Troutman, 1986; Porter and Hsu, 2003). Future research not only needs to continue to address the role of child temperament but also needs to include measures of temperament, parenting, and parents’ depressive symptoms that are not limited to maternal self-report for all of these variables.

A few researchers have addressed potential mediators of associations between depression in parents and children’s temperament, focusing on influences on fetal development. Antidepressant medication treatment during pregnancy, at least in one study, did not predict temperament (Nulman et al., 2002). However, other prenatal or fetal processes may matter. One study found that elevated maternal cortisol at 30–32 weeks of gestation, but not earlier in pregnancy, was significantly associated with greater maternal report of infant negative reactivity, with additive predictions from prenatal maternal anxiety and depression, even after controlling for postnatal maternal psychological state (Davis et al., 2007). In another study, at 2 and 6 months postpartum, mothers who had been depressed in pregnancy and/or postpartum, compared with nondepressed mothers, reported more difficult



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement