capacities are developing, are particularly vulnerable to the negative effects of maternal depression (Dawson and Ashman, in press; Dawson et al., 1994; Goodman and Gotlib, 1999; Murray and Cooper, 1997; NICHD Early Child Care Research Network, 1998a; Weinberg and Tronick, 1998; Zeanah et al., 1997). While the severity and chronicity of maternal depression are clearly predictive of disturbances in child development (Campbell et al., 1995; Cummings and Davies, 1999; Frankel and Harmon, 1996; NICHD Early Child Care Research Network, 1998a), there is no definitive evidence regarding how the timing of maternal depression in the course of a child's life affects the child (Cummings and Davies, 1994b; Gelfand and Teti, 1990; Seifer et al., 1996). The neurobiological data suggest, however, that this is an area in which careful study of the timing question is warranted.

Efforts to understand the processes that underlie the developmental consequences of maternal depression have, not surprisingly, focused on mother-child interactions. Many depressed mothers show disrupted patterns of interaction with their infants. They also express self-doubts about their ability to parent well and are more likely than nondepressed mothers to perceive their children as being difficult (Teti et al., 1996b). Maternal depression affects both the emotional availability of the mother and the emotional tenor of her interactions with her child. Specifically, depressed mothers are more likely either to withdraw from their children and respond with little emotion or energy, or to become intrusive and hostile toward them (Frankel and Harmon, 1996; Tronick and Weinberg, 1997; Zeanah et al., 1997). These behaviors undoubtedly contribute to the higher rates of insecure attachment, as well as the withdrawal, reduced activity, and dysphoria that are observed in infants of depressed mothers (Cummings and Davies, 1994b, 1999; Dawson et al., 1992; Frankel and Harmon, 1996; Murray and Cooper, 1997; Seifer et al., 1996; van IJzendoorn et al., 1992).

These disrupted interactions and, more generally, the adverse effects of maternal depression are not seen uniformly. Many depressed women are very good mothers who raise children who are securely attached, do well in school, and do not misbehave (Cummings and Davies, 1994b, 1999; Frankel and Harmon, 1996). This raises important questions about the conditions that either prevent the damaging processes from occurring or protect children from their effects. Parenting by depressed mothers tends to be disrupted primarily when it occurs in conjunction with other sources of stress or adversity. Accordingly, a child of a depressed mother who also experiences poverty, marital discord, or maltreatment, or whose mother is also abusing substances or is an adolescent, is much more likely to exhibit some form of compromised development than is the child of a mother whose depression occurs in the context of an otherwise supportive environment (Cummings and Davies, 1994b; Seifer et al., 1996; Zeanah et al., 1997).