According to epidemiological studies on depression, women tend to experience depression at twice the rate of their male counterparts (e.g., Eaton et al. 1997). Not surprisingly, researchers have traditionally focused on maternal depression as a risk factor for emotional and behavioral disturbances in the child. Indeed, the link between maternal depression and maladjustment in children is supported by a wealth of research. Maternal depressive symptoms have been linked with children’s externalizing and internalizing behavior problems (Cummings et al. 2005; Marchand and Hock 1998; Weinfield et al. 2009), insecure attachment (Campbell et al. 2004), lower self-esteem (Politano et al. 1992), and more academic and behavior problems at school (Sinclair and Murray 1998).

Considerably less is known about the impact of fathers’ depression on children’s adjustment, largely due to the under-representation of fathers in child development research (Lamb 1975; Phares et al. 2005). Still, some research has shown that fathers of childrearing age, younger than 45 years, are more likely to experience depression than those who are 45 years or older (Blazer et al. 1994), suggesting that they, too, are an important target of research investigation. Yet, only a relatively small number of studies have considered paternal depression as risk factor for child maladjustment (Kane and Garber 2004), and most of these studies considered both maternal and paternal depression (e.g., Marchand and Hock 1998). Only rarely have child development researchers considered depressed fathers as a distinct focus of research investigation (Phares et al. 2005; Schatch et al. 2009). In order to better understand the unique role of paternal depression in children’s social and emotional development, the present study examined the links between fathers’ depressive symptoms and parenting behaviors and their children’s behavior problems.

Parenting behaviors have been identified as an important mechanism by which parents’ depressive symptoms contribute to the development and maintenance of behavior problems in their children (Cummings et al. 2005). Interpersonal perspectives on depression argue that depressed individuals tend to use more maladaptive interpersonal behaviors which hold consequences not only for the depressed individual but also may affect the well-being or social adjustment of other family members, especially the children (Cummings and Davies 1999; Downey and Coyne 1990). Parenting presents many challenges to the depressed parent and is one context in which their impaired interpersonal behaviors are likely to interfere with their ability to effectively manage their children (Kane and Garber 2004).

Indeed, developmental perspectives on child maladjustment support a theoretical model wherein parenting behaviors mediate the link between parental depressive symptoms and child emotional and behavioral disturbance (Cicchetti and Toth 1997). Ample research on child maladjustment has focused on the parenting behaviors of depressed mothers as a source of risk. Studies have shown that depressed mothers are more critical (Goodman et al. 1994), use more hostile and controlling behavior with their children (Marchand et al. 2002), and engage in more conflictual parent–child exchanges (Harnish et al. 1995). Further, mothers’ hostile and controlling behaviors and over-reactive discipline have been associated with children’s externalizing and internalizing behaviors (Harnish et al. 1995; Marchand et al. 2002; O’Leary et al. 1999), and several studies have identified mothers’ dysfunctional parenting behaviors as a mediator in the link between mother’s depressive symptoms and externalizing and internalizing behavior problems in the child (Burt et al. 2005; Bifulco et al. 2002; Harnish et al. 1995 ; Johnson et al. 2001).

It is likely that fathers’ depressive symptoms and negative parenting behaviors operate in similar fashion as mothers (Kane and Garber 2004). Certainly, a number of studies have provided support for fathers’ depressive symptoms as a source of risk for child maladjustment. In particular, fathers’ depressive symptoms have been linked with children’s externalizing and internalizing behavior problems (Carro et al. 1993; Cummings et al. 2005; Dave et al. 2008; Compas et al. 1991; Elgar et al. 2007; Gross et al. 2008; Jacob and Johnson 2001; Kane and Garber 2009; Marchand and Hock 1998; Marchand et al. 2004; Mulvaney et al. 2007; Schatch et al. 2009). Much less well understood is the role of fathers’ parenting behaviors in children’s social and emotional adjustment, especially in the context of paternal depression. Still, at least a few studies have shown that fathers’ over-reactive discipline behaviors, rejection, and decreased nurturance, monitoring, and positive parenting have been linked with children’s externalizing and internalizing behaviors (Elgar et al. 2007; O’Leary and Vidair 2005; Schatch et al. 2009). Further, one recent study by Elgar et al. (2007) showed that fathers’ nurturance, rejection, and monitoring mediated the relation between fathers’ depressive symptoms and children’s behavior problems.

Taken together, the aforementioned research suggests that the consideration of fathers’ depressive symptoms and parenting behaviors in children’s adjustment is an important avenue for further research investigation. The present study will extend previous research by focusing exclusively on the role of fathers’ depressive symptoms and parenting behaviors in children’s externalizing and internalizing behavior problems. With regard to parenting behaviors, fathers’ lax and over-reactive discipline behaviors were considered due to their theoretical relevance to children’s externalizing and internalizing behaviors. Symptoms of depression may include withdrawal and feelings of helplessness as well as irritability and negativity. Parents’ withdrawn or helpless behaviors may translate into parents’ use of more lax behaviors in the discipline context; whereas, parents’ irritability and negativity may translate into greater use of over-reactive behaviors—both of which may have an important bearing on children’s social and emotional adjustment.

To elaborate, by using a lax discipline approach a parent may fail to provide sufficient and clear guidelines for socially acceptable behavior which may lead to externalizing behavior problems in the child. Further, a child who has not been provided with clear guidelines for socially acceptable behavior may be left feeling socially incompetent and exhibit more anxious or withdrawn behaviors that are characteristic of internalizing behavior problems. As well, by using over-reactive parenting behaviors, a parent may inadvertently serve as a model for aggressive behaviors in their children, which may then lead to externalizing symptoms. Additionally, a parent’s over-reactive behaviors could cause children to withdraw or feel anxious and lead to internalizing symptoms.

With regard to children’s externalizing and internalizing behaviors, the present study considered both father- and mother-reported behavior problems. A commonly noted concern in the research literature on parents’ mental health and children’s social and emotional adjustment centers on the issue of parental negative bias. That is, the significant association between parents’ depressive symptoms and children’s behavior problems may in part be due to depressed parents providing more distorted or negatively biased ratings of their children’s behaviors (Richters 1992). Indeed, some research has shown that depressed parents tend to rate their children more negatively on behavior rating scales relative to other raters (Conrad and Hammen 1989; Mulvaney et al. 2007). However, it is also possible that the significant link between parents’ depressive symptoms and children’s behavior problems results from the sole reliance on the parent as an informant of both their own and their child’s attributes (Mulvaney et al. 2007). The reliance on both fathers’ and mothers’ reports of children’s behaviors, as in the present study, provides a more powerful test of the associations between fathers’ depressive symptoms and behavior problems in the child.

In the present study, it was expected that fathers’ depressive symptoms would be significantly related to children’s externalizing and internalizing behavior problems, with more depressive symptoms being associated with more father- and mother-reported externalizing and internalizing behaviors. It was also expected that fathers’ depressive symptoms would be significantly related to fathers’ lax and over-reactive discipline, with more depressive symptoms being associated with more use of lax and over-reactive discipline. Finally, fathers’ lax and over-reactive discipline would be significantly related to children’s externalizing and internalizing behaviors, with more use of lax and over-reactive discipline being related to more father- and mother-reported externalizing and internalizing behaviors in the child.

Method

Sample

Data were collected from a convenience sample of predominantly Caucasian families (95%) residing in several communities in South Central Pennsylvania. Families were recruited through advertisements in local community newspapers. Criteria for participation in the study were: (a) being the parent of a 4–8-year-old child who did not have a chronic illness, physical or mental disability, or behavioral or mental disturbance for which the child was receiving medication, (b) being married to one’s current spouse for 2 or more years, and (c) residing in the same household with one’s spouse and the focal child at the time of the study.

The majority of families were comprised of both biological parents and their children. Only 5% of families had one or more children from a previous marriage. The majority of families (55%) had one child residing in the home at the time of the study; only two families reported having more than two children. Family income ranged from US$15,000 to 160,000 (M = US$74,230), with 11% of families earning between US$15,000 and 39,000, 13% earning between US$40,000 and 59,000, 37% earning between US$60,000 and 79,000, 26% earning between US$80,000 and 100,000, and 13% earning more than US$100,000. On average, fathers were 37 years old (range: 21–59 years) and had 15 years of schooling (range: 12–18 years); mothers’ average as was 34 years old (range: 21–59 years) and their average years of schooling was 16 years (range: 12–20 years). The majority of fathers (69%) and mothers (88%) had some college education. Children’s average age was 77 months. Census data collected in 2000 indicated that the sample used in the present study was representative of the larger population in terms of racial diversity; however, in terms of socioeconomic status, the study sample had proportionately more middle-income families.

Procedure

In response to the newspaper advertisements, 59 families contacted the primary investigator by phone or postcard and provided their home address. Fathers and mothers were then mailed separate packets that included an informed consent form, self-administered questionnaires, explicit instructions for completing the questionnaires, and a self-addressed, stamped envelope for returning the study materials to the primary investigator. In addition, families were told that they would receive US$20 for returning completed questionnaires. The questionnaires were used to obtain data on parents’ socio-demographic characteristics and fathers’ depressive symptoms and child discipline behaviors. Fathers and mothers also completed a child behavior checklist; they were instructed to complete the child behavior checklist with reference to a child in their family who was between 4 and 8 years of age.

Forty-five families returned questionnaires. The present study is based on 36 families for whom complete data were available.

Measures

Depressive Symptoms

The Center for Epidemiological Studies Depression Scale (CES-D; Radloff 1977) was used to assess symptoms of depression in fathers. The CES-D Scale is a self-report measure that was designed to assess level of depressive symptomatology, with an emphasis on depression. The 20 items on this scale can be expected to be experienced by a healthy individual to a certain degree; however, more symptoms will be experienced by a seriously depressed individual. A score of 16 or above has been established as the cutoff that distinguishes individuals with a possible diagnosable depressive disorder from those with depressive symptoms below the clinical range (Meyers and Weissman 1980).

For completion of this scale, fathers were asked to indicate, for each item, the frequency with which they had experienced the symptom mentioned in the item during the past week. The rating scale options were: 0 = “rarely or none of the time (less than 1 day),” 1 = “some or a little of the time (1–2 days),” 2 = “occasionally of a moderate amount (3–4 days),” and 3 = “most or all of the time (5–7 days).” Weights for depressive symptomatology range from 0 to 3, yielding a total possible score ranging from 0 to 60, with lower scores indicating less depressive symptomatology and higher scores indicating greater depressive symptomatology. The internal consistency for this measure has been shown to be about .85 in the general population (Radloff 1977). The alpha coefficient in the present study was .89.

Child Behavior Problems

The Child Behavior Checklist (CBCL; Achenbach and Rescorla 2000, 2001) was used to assess father- and mother-reported child behaviors. The CBCL yields three scores, externalizing, internalizing, and total problem behaviors. For the purpose of this study, only the externalizing and internalizing behavior scores were utilized. Example items that demonstrate externalizing behaviors include “temper tantrums or hot temper,” “threatens people,” and “impulsive or acts without thinking.” While internalizing behaviors include “too fearful or anxious,” “shy or timid,” and “refuses to talk.” To complete the checklist, fathers and mothers responded to 113 items by rating the child’s behavior during the past 2 months. Each item is rated on a three-point Likert-type scale ranging from 0 = “not true at all” to 2 = “very true or often.” Scoring of the checklists was in accordance with the procedures advanced by Achenbach and Rescorla (2000, 2001). The parent-report form of the Child Behavior Checklist has demonstrated good reliability and validity (Achenbach and Rescorla 2000, 2001). In the present study, alpha coefficients for father- and mother-reported externalizing and internalizing behaviors ranged from .81 to .86.

Parent Discipline

Fathers completed the Parenting scale in order to assess their lax and over-reactive discipline (Arnold et al. 1993). The Parent scale is a self-report measure that assesses parents’ responses to child misbehavior; high scores indicate very dysfunctional parental discipline. Each question consists of two anchors that are placed on either end of Likert scales ranging from 1 to 7. The 11-item Laxness subscale and 10-item Over-reactivity subscale were used in the present study. Examples that represent items from the Laxness subscale are “I am the kind of parent that sets limits on what my child is allowed to do” versus “I am the kind of parent that lets my child do whatever he/she wants.” Examples that represent items from the Over-reactivity subscale are “When my child misbehaves I raise my voice or yell” versus “When my child misbehaves I speak calmly to my child.” Alpha coefficients have been previously reported as .83 and .82 for the Laxness and Over-reactivity subscales, respectively (Arnold et al. 1993). In the present study, alpha coefficients based on fathers’ reports were .84 for the Laxness subscale and .68 for the Over-reactivity subscale.

Results

Means and standard deviations for the study variables are presented in Table 1. Consistent with previous research (Achenbach 1991), children’s externalizing behavior problems were positively correlated with their internalizing behavior problems, as reported by fathers (r = .82, p < .01) and mothers (r = .55 = , p < .01). Further, fathers’ and mothers’ reports were positively correlated with one another for both externalizing (r = .82, p < .01) and internalizing (r = .64, p < .01) behavior problems, respectively. Although gender differences in the development of externalizing behaviors are commonly noted, with many studies showing a tendency toward higher scores in males (e.g., McMahon and Forehand 1988), our findings were consistent with research that has shown no significant differences in the externalizing behavior scores of males and females (e.g., Achenbach et al. 1987). In the present study, mean scores for boys’ (N = 17) and girls’ (N = 19) father-reported externalizing behaviors were 8.24 and 9.05, respectively. Mean scores for boys’ and girls’ father-reported internalizing behaviors were 7.53 and 7.26, respectively. With regard to mother-reported behaviors, mean scores for boys’ and girls’ externalizing behaviors were 9.13 and 8.33, respectively, and mean scores for boys’ and girls’ internalizing behaviors were 7.19 and 6.39, respectively. Thus, data for boys and girls were pooled.

Table 1 Correlations between the study variables (N = 36)
Full size table

The relations among the primary study variables were first examined using Pearson correlations. Results provide considerable support for the expected links between the study variables. Table 1 shows that fathers’ depressive symptoms were positively related to children’s father-reported externalizing and internalizing behaviors and mother-reported externalizing behaviors. Additionally, fathers’ depressive symptoms were positively related to fathers’ lax and over-reactive discipline. Lastly, fathers’ lax discipline was positively related to children’s father- and mother-reported externalizing and internalizing behaviors, and fathers’ over-reactive discipline was positively related to children’s father-reported externalizing and internalizing behaviors and mother-reported externalizing behaviors.

Mediational Analyses

In order to determine whether fathers’ lax and over-reactive discipline mediated the associations between fathers’ depressive symptoms and children’s externalizing and internalizing behaviors, a series of regression analyses recommended by Baron and Kenny (1986) were used. According to Baron and Kenny (1986), four conditions must be met in order to establish mediation: (1) the independent variable must be significantly related to the mediator; (2) the independent variable must be significantly related to the dependent variable; (3) the mediator must be significantly related to the dependent variable when the dependent variable is regressed on the independent variable and the mediator. Moreover, (4) the association between the independent variable and the dependent variable should no longer be significant when the mediator is added to the regression equation. Partial mediation occurs when the association between the independent variable and the dependent variable continues to be significant when the independent variable and the mediator are regressed on the dependent variable; however, the strength of the association is reduced.

Using the aforementioned regression series, several regression analyses were conducted. The models tested were based on results from the correlational analyses. Results are presented in Tables 2, 3 and 4 and provide some support for mediation. As shown in Table 3, the independent variable, fathers’ depressive symptoms, was significantly associated with the mediator, fathers’ lax discipline, supporting condition 1 of the Baron and Kenny test. Additionally, condition 2 was met; the independent variable, fathers’ depressive symptoms, was significantly associated with the dependent variable, children’s father-reported internalizing behaviors. Further, fathers’ lax discipline, the mediator, was significantly associated with the dependent variable, children’s father-reported internalizing behaviors, supporting condition 3. Finally, condition 4 was met; the association between the independent variable, fathers’ depressive symptoms, and the dependent variable, children’s father-reported internalizing behaviors, was no longer significant when the dependent variable, children’s father-reported internalizing behaviors, was regressed on the independent variable, fathers’ depressive symptoms, and the mediator, fathers’ lax discipline. Thus, mediation was supported.

Table 2 Regression analyses testing the meditational effects of fathers’ over-reactive discipline in the relation between fathers’ depressive symptoms and children’s father-reported externalizing and internalizing behaviors
Full size table
Table 3 Regression analyses testing the meditational effects of fathers’ lax discipline in the relation between fathers’ depressive symptoms and children’s father-reported externalizing and internalizing behaviors
Full size table
Table 4 Regression analyses testing the meditational effects of fathers’ over-reactive and lax discipline in the relation between fathers’ depressive symptoms and children’s mother-reported externalizing behaviors
Full size table

A different pattern emerged for children’s father- and mother-reported externalizing behaviors. As shown in Table 3, the independent variable, fathers’ depressive symptoms, was significantly associated with the mediator, fathers’ lax discipline, supporting condition 1 of the Baron and Kenny test. Additionally, condition 2 was met; the independent variable, fathers’ depressive symptoms, was significantly associated with the dependent variable, children’s father-reported externalizing behaviors. Condition 3 was also met; fathers’ lax discipline, the mediator, was significantly related to the dependent variable, children’s father-reported externalizing behaviors. However, condition 4 was not met; the strength of the association between the independent variable, fathers’ depressive symptoms, and the dependent variable, children’s father-reported externalizing behaviors, was not significantly reduced with the entry of fathers’ lax discipline into the regression equation. Thus, neither mediation nor partial mediation was supported. Instead, fathers’ depressive symptoms and lax discipline were independent predictors, accounting for unique proportions of the variance in children’s father-reported externalizing behaviors.

As shown in Table 4, the same pattern of results emerged when children’s mother-reported externalizing behaviors were considered.

Discussion

The present study examined the role of fathers’ depressive symptoms and lax and over-reactive discipline in children’s externalizing and internalizing behavior problems in a nonclinical, community sample of fathers and their children. In the present study, fathers’ depressive symptoms were positively related to children’s externalizing behavior problems, as reported by both fathers and mothers. A commonly noted concern in the research literature on parents’ mental health and children’s social and emotional adjustment centers on the issue of negative bias in depressed parents’ reports of their children’s behaviors, which may at least partially account for significant associations between parents’ depressive symptoms and children’s behavior problems (Richters 1992). By considering both father- and mother-reported behavior problems, the present study provided a more powerful test of the link between fathers’ depressive symptoms and children’s behavior problems, at least with regard to their externalizing symptoms.

Interestingly, in the present study, fathers’ depressive symptoms were significantly related to children’s father-reported, but not mother-reported, internalizing behaviors. Findings may reflect actual differences in the level of children’s internalizing symptoms that result from interactional processes that are present in the father–child, but not mother–child, relationship. However, a negative bias in fathers’ reports of children’s internalizing behaviors cannot be ruled out. Still, other factors may also partially account for the findings. For example, it is possible that findings may have been influenced by differences in the accuracy with which fathers and mothers reported children’s internalizing behaviors. As the primary caregivers of children, mothers have greater opportunities to observe their children’s behaviors, perhaps making them more accurate reporters. This may be especially true for children’s internalizing behaviors that some researchers have argued are more difficult to accurately report than externalizing behaviors due to the considerable ambiguity in their behavioral characteristics (De Los Reyes and Kazdin 2005; Mulvaney et al. 2007). Unlike externalizing symptoms which involve more acting out or disruptive behaviors that are easily recognized by observers, internalizing symptoms reflect internal distress within the child which is not necessarily readily observable or easy to identify. Consistent with other studies (Briggs-Gowan et al. 1996; Mulvaney et al. 2007), findings from the present study indicated greater agreement between fathers and mothers on their reports of children’s externalizing behaviors compared to their internalizing behaviors.

With regard to fathers’ discipline, correlational analyses revealed that fathers who reported more depressive symptoms reported using more lax and over-reactive discipline. Additionally, fathers who reported using more lax discipline had children who were rated by fathers and mothers as having more externalizing and internalizing behavior problems, and fathers who reported using more over-reactive discipline had children who were rated by fathers and mothers as having more reported externalizing behavior problems and by fathers as having more internalizing behavior problems. These findings are consistent with a recent study by Elgar et al. (2007) that showed significant associations between fathers’ depressive symptoms and negative parenting behaviors and children’s behavior problems (Elgar et al. 2007) and with another study by O’Leary and Vidair (2005) that identified fathers’ over-reactive discipline, in particular, as a significant correlate of children’s father-reported externalizing and internalizing behaviors.

When a series of regression equations were applied to the data in order to more fully explore the links between the father and child attributes, results indicated that fathers’ lax discipline mediated the association between fathers’ depressive symptoms and children’s father-reported internalizing behaviors. In contrast, fathers’ depressive symptoms and lax discipline were independent predictors of children’s father- and mother-reported externalizing behaviors. Findings suggest that additional processes need to be considered in order to better understand the link between fathers’ depressive symptoms and children’s externalizing behaviors. Other parenting behaviors not examined in the present study may also be important. For example, in one study Elgar et al. (2007) identified low levels of father nurturance and high levels of father rejection as mediators in the association between fathers’ depressive symptoms and children’s externalizing behavior problems. Additionally, aspects of the father–child relationship, in particular, father–child conflict, have been identified as a mediator in the link between fathers’ depressive symptoms and children’s externalizing behaviors (Kane and Garber 2009).

Some methodological limitations of the present study must be acknowledged. To begin, findings from the present study are based on data collected from a small, homogeneous sample of volunteers. Studies based on more diverse samples may result in a pattern of findings that differs from the one found in the present study. For example, in the present study mean scores for fathers’ and children’s symptoms were within the nonclinical range. It is uncertain whether the same pattern of findings would be found in families wherein fathers’ and children’s symptoms are more severe. Future studies based on larger samples of fathers and children that exhibit symptoms in the clinical range can further our understanding of the links between fathers’ depressive symptoms and discipline behaviors and children’s behavior problems.

An additional limitation of the present study was that fathers were the sole informants on their own attributes, opening up the possibility that some of the study’s findings may have been influenced by shared method variance. That is, significant associations between the study variables may be in part due to the idiosyncratic way in which an individual responds to the measures (Sabatelli and Bartle 1995). Although, self-reports may be the most appropriate method for assessing indices of psychological well-being, such as depressive symptoms, other father attributes, such as their discipline behaviors, may have been better assessed through behavior observations.

Lastly, the model tested in the present study assumes that fathers’ attributes are important determinants of children’s behavior problems. However, the cross-sectional nature of the data does not permit causal connections to be tested. It is likely that the relationships between the fathers’ and children’s attributes are reciprocal in nature, with children’s behavior problems also influencing fathers’ depressive symptoms and maladaptive discipline. At least one study based on a sample of mothers and children has provided support for a reciprocal relation between parent and children attributes. Marchand et al. (2002) found that mothers’ earlier hostile-controlling behaviors explained a significant portion of the variance in children’s later externalizing behaviors; reciprocally, children’s earlier externalizing behaviors explained a significant portion of the variance in mothers’ later hostile-controlling behavior. Future research using longitudinal data can more fully address the nature of the relations between father and child attributes.

The present study does contribute to the literature in important ways. By focusing exclusively on fathers’ attributes, findings advance our understanding of the unique role of fathers’ depressive symptoms in children’s behavior problems. Additionally, findings identify fathers’ maladaptive discipline as an important correlate of children’s behavior problems. The most consistent patterns of findings were the associations between fathers’ depressive symptoms and lax and over-reactive discipline and children’s externalizing behaviors and between fathers’ lax discipline and children’s internalizing behaviors, as these associations were significant for both fathers’ and mothers’ reports of children’s behaviors. Also noteworthy were the findings that emerged from the regression analyses. Findings indicated that fathers’ lax discipline mediated the link between fathers’ depressive symptoms and children’s father-reported internalizing behaviors. In contrast, fathers’ depressive symptoms and lax discipline were independent predictors, explaining unique variance in children’s father- and mother-reported externalizing behaviors. Findings suggest that fathers’ depressive symptoms and lax discipline may operate differently in children’s externalizing and internalizing behaviors.