Depressive symptoms and parenting competence: An analysis of 13 regulatory processes

Abstract

Mechanisms that lead depressive symptoms to undermine parenting are poorly understood. This review examines cognitive, affective, and motivational processes thought to be responsible for the impact of depressive symptoms on parenting. We present a five-step, action-control model and review 152 studies relevant to 13 regulatory processes. Evidence suggests that depressive symptoms undermine parenting because they reduce child-oriented goals, undermine attention to child input, increase negative appraisals of children and parenting competence, activate low-positive and high-negative emotion, and increase positive evaluations of coercive parenting. Yet, this review reveals significant limitations in knowledge of these processes. Evidence that they mediate depression-parenting relations is scare; important processes remain unstudied; conceptions of regulation are undifferentiated; children’s contributions are largely unexamined; moderating variables are largely unexplored; and methods fail to capture the dynamics of processing input from children. Rigorous testing of such process models holds promise for clarifying the basis of depression-related parenting problems.

Introduction

Depressive symptoms in mothers are consistently associated with low parenting competence. As depressive symptoms increase, mothers tend to become less positive and responsive to children, more negative and disengaged, and more hostile, manipulative, and inconsistent during discipline (Cummings and Davies, 1994, Downey and Coyne, 1990, Goodman, 1992, Lovejoy et al., 2000). The children of depressed mothers, in turn, are vulnerable to physical, cognitive, social, and emotional impairments of diverse kinds (Burke, 2003, Downey and Coyne, 1990, Lyons-Ruth et al., 2000). Despite considerable evidence linking depressive symptoms to low parenting competence, the mechanisms responsible for this relation are poorly understood. As a result, we cannot answer questions as fundamental as why mothers with depressive symptoms sometimes appear competent as parents and at other times profoundly incompetent; why they sometimes are withdrawn and at other times are intrusive and overly controlling; whether it is their negative thinking, emotional distress, or other processes that are the heart of their parenting deficits and, therefore, should be the target of intervention efforts. In this review, we attempt to understand the parenting deficits of these mothers by examining processes thought to regulate their parenting. First, to enumerate these processes, we integrate ideas from social, developmental, and clinical research into a process-oriented, action-control model. Second, we determine whether evidence supports the potential mediating role of 13 processes specified in the model. Third, we discuss limitations in current theories and methods that restrict our ability to understand the mechanisms by which depressive symptoms undermine parenting.

Depressive symptoms are common, particularly in women and in mothers of young children. For women of childbearing age, depression is the leading cause of disability (Stewart, 2006). Although estimates of its prevalence vary, they are uniformly high. At any time 20–25% of the population has depressive symptoms severe enough to be considered clinically significant (Gotlib & Hammen, 1992); 9–26% of women will experience a major depression in their lifetime (Gotlib and Hammen, 1992, Schwartz and Schwartz, 1993). Rates of depressive symptoms are higher for women with infants or young children. Twenty-two percent of mothers of 17-month-olds, for example, have elevated depressive symptomatology; 36% continue to have significant symptoms a year later (McLennan, Kotelchuck, & Cho, 2001). Prevalence is even higher if mothers are single, from low-income families, or have children with medical problems (e.g., Horowitz and Kerker, 2001, O’Brien et al., 1999). It is clear that substantial numbers of children are reared by mothers with significant depressive symptoms.

That depressive symptoms compromise parenting is established. They are associated with a variety of parenting behaviors that have been linked to problems in child development (Cummings and Davies, 1994, Downey and Coyne, 1990, Goodman, 1992, Lovejoy et al., 2000). First, depressive symptoms predict maternal withdrawal, that is, low responsiveness and lack of ongoing involvement with children. Second, they predict maternal intrusiveness (Cohn et al., 1986, Field et al., 1990). Third, they predict flat and negative emotional expression to children (e.g., Cohn, Campbell, Matias, & Hopkins, 1990) and low positive expression (e.g., Feng, Shaw, Skuban, & Lane, 2007). Fourth, they predict ineffective discipline, specifically discipline that is harsh, inconsistent, manipulative, and indulgent (Leung and Slep, 2006, Lovejoy et al., 2000). Each of these predicts poor child outcomes (Dix, 1991, Holden, 1997, Maccoby, 1980, Maccoby and Martin, 1983). For the purposes of this review these characteristics constitute low parenting competence.

Conceptions of why depressive symptoms lead to these parenting deficits are not well developed. Major reviews stress that often researchers think such deficits reflect simply incompatibility between depressive symptoms and the requirements of effective parenting (Downey and Coyne, 1990, Lovejoy et al., 2000). Intuitively, negative mood, low energy, withdrawal, and other symptoms seem to preclude sustained, positive, and responsive behavior with children. A few researchers, however, have attempted to delineate specific regulatory mechanisms. Three have been stressed in major reviews (Cummings and Davies, 1994, Downey and Coyne, 1990, Goodman, 1992, Lovejoy et al., 2000).

First, researchers propose that depressive symptoms create negative biases in mothers’ appraisals both of children and of mothers’ own competence as parents. As depressive symptoms increase, mothers may infer that children posses negative intentions and dispositions and that mothers themselves are unable to parent effectively (Alloy, 1988, Beck, 1967, Abramson et al., 1978, Friedlander et al., 1986, Gotlib and Hammen, 1992, Richters, 1992, Teti and Gelfand, 1991). Negative appraisals, in turn, may activate negative emotion and lead to behavior that is harsh, unresponsive, or insensitive. Second, researchers propose that depressive symptoms reduce mothers’ tolerance for aversive stimulation (Forehand et al., 1987, Lahey et al., 1984, Bauer and Twentyman, 1985). As depressive symptoms increase, children’s demands and difficult behaviors may increasingly arouse distress in mothers and lead them either to avoid children or react forcefully to reduce aversive behaviors. Third, researchers propose that depressive symptoms induce motivation to minimize effort, that is, to select activities and responses that require low effort (Cummings and Davies, 1994, Downey and Coyne, 1990, Kochanska et al., 1987). This is thought to be responsible for tendencies of mothers with depressive symptoms to suppress forcefully, or simply avoid, difficult child behaviors (Kochanska et al., 1987).

Although important, these proposals are relatively undifferentiated. They comprise neither a comprehensive nor integrative analysis of how depressive symptoms affect the processes responsible for parenting behavior. One approach that may provide a more complete account is what we refer to as action-control theory (Carver and Scheier, 1990, Dix and Branca, 2003, Kuhl and Beckmann, 1985, Crick and Dodge, 1994). This perspective stresses the cognitive, affective, and motivational processes that control goal-directed action. Numerous theories propose that human action is the result of an internal guidance system that functions to promote individuals’ moment-to-moment goals. This emphasis is shared by perspectives as diverse as cognitive social learning theory (Bandura, 1997), information-processing theory (Crick & Dodge, 1994), basic emotion theory (Frijda, 1986, Lazarus, 1991), and a variety of goal-oriented theories in social psychology (Carver and Scheier, 1990, Gollwitzer and Bargh, 1996, Kuhl and Beckmann, 1985, Miller et al., 1960, Powers, 1973, Schank and Abelson, 1977). Each of these approaches depicts individuals as active agents involved in, “a continual process of moving toward various kinds of mental goal representations” (Carver & Scheier, 1990, p. 3). Typically, the heart of these theories is a feedback control system (Powers, 1973, Carver and Scheier, 1990, Miller et al., 1960, Schank and Abelson, 1977). Individuals achieve desired outcomes by using feedback from events to correct problems and adapt to changing circumstances. Thus, to promote their concerns, individuals instantiate immediate goals, attend to input, and use their understanding of input as feedback to adjust behavior so that their concerns are promoted. When goals are not achieved, they alter either the goals themselves or the action plans selected to promote them.

Fig. 1 depicts an action-control framework for conceptualizing how depressive symptoms influence parenting competence. The framework characterizes parenting in terms of five steps related to the regulation of goal-directed action: (a) activating concerns or goals, (b) encoding goal-relevant information, (c) appraising the significance of that information for achieving goals, (d) activating emotions to direct and motivate action, and (e) evaluating and selecting an immediate response. As her child builds with blocks, for example, a mother might have as a goal that the child succeed without frustration (Step 1). She might attend to signs of the child’s success and frustration (Step 2), appraise why the child is being successful or is failing (Step 3), activate worry when the child’s frustration mounts (Step 4), and evaluate options for how to reduce this frustration (Step 5). Depressive symptoms may disrupt parenting by undermining mothers’ ability to execute each of these processes.

With its emphasis on immediate cognitive, motivational, and affective processes, this approach is consistent with theories of depression, particularly cognitive and cognitive-behavioral approaches (Abramson et al., 1978, Alloy, 1988, Beck, 1976, Clark and Beck, 1999). Although immediate interaction goals are not central to these approaches, theories of depression universally acknowledge disturbances in motivation that alter what people want during interactions with others (Gotlib and Hammen, 1992, Hoberman and Lewinsohn, 1985), Step 1 of our model. Cognitive approaches in particular stress that depressive symptoms bias attention and encoding, induce negative appraisals of self and others (Beck, 1967, Beck, 1976, Bieling and Segal, 2004), and promote negative attributions for the causes of events (Beck, 1967, Beck, 1976, Bieling and Segal, 2004). These are the processes that define Steps 2 and 3. All major theories of depression stress the emotional processes central to Step 4, that is, the regulatory importance of low positive and high negative emotion. Interpersonal and behavioral theories also stress the tendency for depressive symptoms to undermine social problem solving and increase the difficulty of generating actions that achieve positive outcomes (Gotlib and Hammen, 1992, Lewinsohn, 1974, Nezu, 1987, Nezu et al., 1989). These skills are aligned with Step 5.

In this review our focus is not on the causes of depressive symptoms – whether they are due to genetic, biochemical, or social factors. Instead, we stress how immediate regulatory processes associated with depressive symptoms affect the quality of parents’ reactions to children. We assume that processes discussed in the literature that are less specific and immediate than those stressed here influence parents’ behavior by affecting in some way processes occurring at the moment a chosen behavior occurs. These processes are one link in the chain that connects global, non-specific variables such as social isolation or chemical imbalances with the immediate reactions that ultimately determine whether effective behavior with children occurs. Thus, the model has the potential to explain the mechanisms responsible for the impact on parenting of a wide variety of more distal depression-related variables and to characterize components of behavioral regulation that might be targeted by intervention efforts.

Fig. 1 also displays 13 processes and associated hypotheses related to why depressive symptoms undermine parenting. Each hypothesis concerns a process occurring within its respective step. Some of these processes are widely studied; others have been proposed, but are unstudied. In evaluating each, we examine evidence related to three questions, depicted as paths in Fig. 2. (a) Are depressive symptoms associated with the process proposed in that hypothesis (Path 1)? (b) Regardless of depressive symptoms, does that process predict parenting reactions associated with low competence? Does it predict, for example, unresponsiveness, negative emotional expression, or preferences for harsh discipline (Path 2)? (c) Is there evidence that the process specified in that hypothesis mediates the relation of depressive symptoms to reactions related to low parenting competence (Path 3)?

To evaluate the model and its principal proposals – the depression-related paths of each hypothesis (Paths 1 and 3) – we attempted to locate all studies that related cognitive, affective, or motivational processes to depressive symptoms in mothers. Using EBSCO Research Databases, we searched for relevant articles using dozens of search terms related to mothers (e.g., mother, parent, parental, maternal), depression (e.g., depressive, dyphoric, depressed), motivation (e.g., motives, goals, expectations), cognition (e.g., appraisal, attention, attitude, attribution, belief, competence, confidence, control, efficacy, encoding, evaluation), and emotion (e.g., affect, anger, anxiety, emotion, guilt, sad, worry). These searches yielded several thousand entries. We adopted three selection criteria. First, to be included, articles had to be published in peer-reviewed journals. Second, they had to address either depressive symptoms in non-clinical populations or diagnosed unipolar depression. We excluded studies of mothers with bipolar disorders. Third, articles had to report the relation of depressive symptoms or of having a depressive disorder to processes in mothers thought to regulate their reactions to children. Any process that could be activated during mother–child interaction – for example, a motive, expectation, emotion, appraisal – was included. These criteria led to a final set of 152 depression-related studies. Note that Path 2 of each hypothesis does not involve depressive symptoms directly, but instead simply links the components of processing specified in the model to parenting competence (see Fig. 1). To locate studies of these relations, we conducted searches with the same search terms but without depression-related terms. Because few studies have examined fathers and because their processing may differ from that of mothers (e.g., Slep & O’Leary, 2007), we included only research on mothers.

The mothers who participated in these studies were from diverse backgrounds. Their children ranged in age from newborns to adolescents. Most studies included infants (n = 43), preschoolers (n = 60), or elementary schoolers (n = 55), although 26 studies (17%) included preadolescents or adolescents. In some studies depressive symptoms were measured with depression inventories (e.g., Center for Epidemiological Study Depression Inventory; Beck Depression Inventory); in others, mothers were diagnosed using criteria set forth in the Diagnostic and Statistical Manual of Mental Disorders. Thirty-five studies (23%) examined mothers with a diagnosed depressive disorder; 117 studies (77%), mothers with no diagnosed disorder (see Table 1). Of the studies of undiagnosed mothers, 62 (53%) involved mothers from populations in which depressive symptoms are commonplace (e.g., single mothers, mothers from low SES backgrounds, mothers of children with behavioral problems). Lyons-Ruth et al. (2000) noted that the child outcomes associated with mothers’ depressive symptoms in clinical studies do not differ substantially from those observed in nonclincal studies. This is the case, as well, for research examined here on parenting processes, rather than child outcomes. Thus, we discuss the findings simply in terms of depressive symptoms generally. Table 1 summarizes the distribution of studies across the five processing steps, 13 specific processes, and diagnosed vs. undiagnosed samples.