Examining subtypes, part II 1

Running Head: SUBTYPES OF BEHAVIOR PROBLEMS, PART II

Examining subtypes of behavior problems among 3-year-old children, Part II: Investigating differences in parent psychopathology, couple conflict, and other family stressors

Lauren H. Goldstein2, Elizabeth A. Harvey[1][2], Julie L. Friedman-Weieneth2, Courtney Pierce2, Alexis Tellert2, & Jenna C. Sippel2

This manuscript is in press at Journal of Abnormal Child Psychology


Abstract

This study examined family stressors among 3-year-old children who were classified as hyperactive (HYP), hyperactive and oppositional defiant (HYP/OD), and non-problem based on mothers’ reports of behavior. Children with HYP/OD were found to experience higher levels of family stressors than non-problem children on almost every family stressor variable. Compared to children with HYP, families of children with HYP/OD also tended to report more Axis II maternal psychopathology, Axis I paternal psychopathology, and high intensity couple conflict tactics. However, the HYP and HYP/OD group did not significantly differ on maternal Axis I psychopathology, paternal Axis II psychopathology, parental marital status, negative life events, frequency of couple conflict, or use of lower intensity couple conflict tactics. Parents of children with HYP and HYP/OD reported more negative life events, more maternal adult ADHD symptoms, and more maternal avoidance and verbal aggression during marital conflict than parents of non-problem children. Implications for treatment and etiology are discussed.

Key words: hyperactivity, oppositional defiance, preschool-aged children, parent psychopathology, couple conflict


Examining subtypes of behavior problems among 3-year-old children, Part II: Investigating differences in parent psychopathology, couple conflict, and other family stressors

Examining subtypes, part II 1

The present study is the second in a three-part series, which examines how biological factors, family stressors, and parenting vary across preschool-aged children with hyperactivity, comorbid hyperactivity/oppositional defiance, and no behavior problems.[3] In Part I (Harvey, Friedman-Weieneth, Goldstein, & Sherman, 2006), we presented a model in which family stressors were proposed to play a central role in the development of comorbid ADHD/ODD, but not in the development of ADHD. In particular, family stressors are thought to interact with early child characteristics and lead to the development and maintenance of behavior problems (Barkley, 1990). Child impulsivity and a family history of ADHD may also result in elevated levels of family stressors among children with ADHD who do not have ODD; however, theory and some research suggest that stressors should be significantly higher among families of children with ADHD/ODD and ODD than among families of children with pure ADHD (Counts, Nigg, Stawicki, Rappley, & Von Eye, 2005).

Much of the research linking family stressors with ADHD and ODD has been conducted with older children (Biederman, Milberger, & Faraone, 1995; Counts et al., 2005). A growing body of research supporting the validity of ADHD and ODD diagnoses among preschool-aged children (e.g., Keenan & Wakschlag, 2004; Lahey et al., 1998) highlights the need to examine the role of family stressors among young children. Moreover, research (Waschbusch, 2002) points to the importance of taking into account the considerable overlap between ADHD and ODD among both school-aged children (Biederman, Newcorn, & Sprich, 1991) and preschool-aged children (Keenan & Wakschlag, 2000; Wilens et al., 2002). The present study examines family stressors among children with comorbid hyperactivity/oppositional-defiance, hyperactivity alone, and no problems, to determine whether differences that are consistent with existing theory can be observed in children as young as 3 years old.

Parent Psychopathology

The link between parent psychopathology and child functioning has been well-documented (e.g., Lahey et al., 1988). However, few studies have examined children’s behavior problems multidimensionally. Results have been mixed for parental depression and anxiety with some studies finding associations with ADHD regardless of co-occurring ODD or CD (Chronis, Lahey, Pelham, Kipp, Baumann, & Lee, 2003; Cunningham & Boyle, 2002; Jensen et al., 2001; Johnston, 1996; Nigg and Hinshaw, 1998), while others have not found a clear link between parental depression or anxiety and pure ADHD (Barkley, Fischer, Edelbrock, & Smallish, 1991; Lahey et al., 1988). Research suggest that parental substance abuse and ASPD are associated with ADHD in children only when comorbid ODD or CD are present (e.g., Chronis et al., 2003; Faroane, Biederman, Keenan, & Tsuang, 1991; Frick, Lahey, Christ, Loeber, & Green, 1991). Parental history of childhood ADHD (Chronis et al., 2003; Faraone et al., 1991; Frick et al., 1991; Nigg & Hinshaw, 1998) and paternal adult ADHD symptoms have been associated with children’s ADHD regardless of comorbid conduct problems (Nigg Hinshaw, 1998).

A smaller body of research has linked parental anxiety, depression, substance use, and antisocial behavior to externalizing problems in preschoolers (Puttler, Zucker, Fitzgerald, & Bingham, 1998; Shaw, Winslow, Owens, & Hood, 1998; Spieker, Larson, Lewis, Keller, & Gilchrist, 1999; West & Newman, 2003). However, only Cunningham and Boyle (2002) have examined subtypes of behavior problems in exclusively preschool-aged children.[4] They found that mothers of 4-year-old children with pure hyperactivity and with comorbid hyperactivity/ oppositional-defiance reported more depression than did mothers of non-problem children.

Historically, research on the role of parent psychopathology in the development of children’s behavior problems has focused on mothers (Phares & Compas, 1992), and this trend has continued over the past decade (Phares, Fields, Kamboukos, & Lopez, 2005). Nonetheless, research suggests that externalizing problems in children are equally related to maternal and paternal psychopathology (Connell & Goodman, 2002), though different patterns have emerged for mothers and fathers. Maternal depression (e.g., Befera & Barkley, 1985), but not paternal depression (Cunningham, Benness, & Siegel, 1988; Lahey et al., 1988; Nigg & Hinshaw, 1998; Stewart, DeBlois, & Cummings, 1980), has been linked with ADHD in children. On the other hand, substance use (e.g., Frick, Lahey, and Loeber, 1992; Lahey et al., 1988; Loukas, Zucker, Fitzgerald, & Krull, 2003), ASPD (Frick, 1994), and adult ADHD (Nigg & Hinshaw, 1998) have been linked with child behavior more strongly for fathers than for mothers.

Couple Conflict

The link between interparental conflict and children’s behavior problems has also been well-established (e.g., Calzada, Eyberg, Rich, & Querido, 2004; Frosch & Mangelsdorf, 2001). Marital conflict may affect children’s behavior problems through its effect on children’s emotional arousal (Davies & Cummings, 1994) or through the disruption in the parent-child relationship (Grych & Fincham, 1990). Identifying which specific characteristics of couple conflict are most harmful for children is critical to fully understanding this process. For example, conflict that is high in frequency and intensity (e.g., hostile, aggressive conflict), involves child-related content, and remains unresolved, may place children at risk for behavior problems (Cummings, Vogel, & Cummings, 1989; Grych & Fincham, 1990). A pattern of demand-withdrawal (one person approaches a partner on an issue and the partner avoids discussion) during marital conflict has been identified as a common cycle in dissatisfied relationships (Caughlin & Huston, 2002); however this pattern has been linked only with child internalizing problems (Katz & Gottman, 1993), and not with behavior problems in older children (Lindahl, 1998). Few studies have examined whether marital functioning is associated with hyperactivity, oppositional-defiance, or a combination of the two, with only one focusing on preschool-aged children. Most of these have found elevated levels of marital conflict among parents of children with ADHD/ODD, but not among parents of children with pure ADHD (e.g., Barkley et al., 1991; Lindahl, 1998; Stormont-Spurgin & Zentall, 1995), with some evidence that differences may vary across specific marital dimensions (Lindahl, 1998; Stormont-Spurgin Zentall, 1995).

Contextual Stressors: Negative Life Events, Marital Status, and Low SES

Family stressors may also take the form of structural/contextual stressors, such as negative life events, single parenthood, and low socioeconomic status. These factors are thought to influence children primarily through their effects on parents’ well-being and parenting (Crnic, Gaze, & Hoffman, 2005; Mistry, Vandewater, & Huston, 2002). Families of children with ADHD tend to be of lower socioeconomic status and are more likely to be headed by single parents than are families of children without ADHD; however, this appears to be true only for children with ADHD who have comorbid disorders (Counts et al., 2005; Hinshaw, 1987; Johnston, 1996). Although theory suggests that negative life events should be linked with comorbid ADHD/ODD rather than with pure ADHD, the few studies of older children that have addressed this question have found similar elevations among children with pure ADHD and children with ADHD/ODD (Barkley et al., 1991; Johnston, 1996).

The Present Study

Theory suggests that family stressors place children with ADHD at risk for developing ODD, and research on older children provides some, though mixed, support for this. The present study examined whether differences across 3-year-old children with comorbid hyperactivity/ oppositional-defiance (HYP/OD), hyperactivity alone (HYP), and non-problem children are consistent with theory regarding the role of family stressors in the development of ADHD/ODD. Examining such differences provides a step toward understanding the degree to which these behavior subtypes differentiate from each other at an early age and represent clinically significant disturbances that share characteristics with ADHD and ADHD/ODD. In particular, the present study addressed the following questions:

Are there differences across subtypes on maternal and paternal psychopathology? Theoretical models suggest that while genetic/biological factors underlie both ADHD and ODD, family stressors also play a significant role in the etiology of ODD. Thus, parent psychopathology dimensions, other than adult ADHD, should be higher among children with HYP/OD than among children with HYP and children without problems. While establishing differences in parent psychopathology would not rule out other causal mechanisms including shared genetics, it would provide a first step in determining whether early preschool HYP and HYP/OD share etiological correlates with ADHD and ADHD/ODD. It is critical to examine dimensions separately given research that mothers may differ more on depression, whereas fathers may differ more on substance use, ASPD, and adult ADHD. As a result of shared genetics, adult ADHD should be higher among parents of children with HYP and HYP/OD than among non-problem children. Elevations in other types of psychopathology may be evident among parents of children with HYP due to child effects and potential third variables, but should be smaller than among parents of children with HYP/OD.

Are there differences across subtypes on marital conflict? Research and theory suggest that compared to parents of HYP and non-problem children, parents of children with HYP/OD should report more conflict that is high in frequency and intensity, involves child-related content, is avoidant, and remains unresolved. Although marital conflict is not thought to cause ADHD, parents of children with pure ADHD may show elevated levels of marital conflict as a result of child effects or parental ADHD symptomatology, though this comparison was exploratory due to limited research with conflicting results (Barkley et al., 1991; Johnston, 1996).

Are there differences across subtypes on negative life events, marital status, and socioeconomic status? It was predicted that parents of children with HYP/OD would more likely be single, of lower socioeconomic status, and report more negative life events than parents of HYP and non-problem children. Significant differences were not predicted between HYP and non-problem children. In Part I, we (Harvey et al., 2006) examined differences in socioeconomic status in order to provide context for interpreting results, and as predicted, children in the HYP/OD group had mothers with significantly lower education than mothers of children in the HYP and non-problem groups. Because these analyses were already completed in Part I, they will not be repeated, but will be discussed with other family stressors in this paper. Our prediction regarding negative life events was tentative because theory suggests that they should occur more frequently among children with HYP/OD than among children with HYP, but empirical studies have not supported this.

Do race/ethnicity and gender moderate differences across groups in parent psychopathology, marital conflict, and negative life events? There are thought to be cultural differences in psychopathology (Hall, Bansal, & Lopez, 1999), marital conflict (McLoyd, Cauce, Takeuchi, & Wilson, 2000), and stressful life events (Kilmer, Cowen, Wyman, Work, & Magnus, 1998). Few studies have examined whether the impact of family stressors on child psychopathology varies across ethnicity, but there is some evidence that it may (Costello, Keeler, & Angold, 2001). Because ethnic minority groups are underrepresented in research on behavior disorders in children (Gingerich, Turnock, Litfin, & Rosén, 1998), it is critical to examine whether relations that are predicted by existing models are supported across different ethnic groups. There also has been evidence that some family stressors may be more strongly related to behavior problems in boys than in girls (e.g., Davies & Lindsay, 2001), pointing to the need to examine whether gender moderates the relation between family stressors and child behavior.

Examining subtypes, part II 1

Method

Participants and Procedure

Participants were drawn from 258 children and their 258 mothers and 178 fathers who were participating in the first year of a longitudinal study of young children’s behavior problems. One hundred ninety-nine of these children had significant externalizing (hyperactivity and/or aggression) problems at the time of screening and 59 children did not have behavior problems. Children were all 3 years old at the time of initial screening and were 36 to 50 months at the time of the first home visit. In Part I of this series, children were classified into behavior subgroups using hyperactivity and oppositional-defiance indexes that were created by aggregating across several rating scales and an interview completed with the mother. Forty-one of these children were classified as HYP, 96 children as HYP/OD, and 59 children as non-problem. Details about participants and procedures can be found in the first paper of this series (Harvey et al., 2006).

Measures

Marital status. Families were classified as married/living together if the child lived with two parents (including stepparents) and as single if the child lived with just one parent.

Millon Clinical Multiaxial Inventory – III (MCMI-III). Parent psychopathology was measured using the MCMI–III (Millon, Davis, & Millon, 1997), a 175-item questionnaire that assesses symptoms of DSM-IV disorders. The following Axis I scales were used in the present study: anxiety, somatoform, bipolar: manic, dysthymia, major depression, posttraumatic stress (PTSD), and alcohol and drug abuse disorders. The following Axis II scales were included: schizoid, avoidant, depressive, dependent, histrionic, narcissistic, antisocial, compulsive, schizotypal, borderline, and paranoid personality disorders. These subscales have demonstrated good internal consistencies (average α = .82), test-retest reliability (average r = .91), and have been found to have generally good sensitivity in detecting clinician-based diagnoses (Millon & Davis, 1997). Base rate (BR) scores were used, which are standard scores tied to empirically derived population prevalence rates (Millon, Davis, & Millon, 1997).