Maternal ADHD: Parent-Child Interactions and Relations With Child Disruptive Behavior

Abstract

This study examined how ADHD symptoms in mothers of children with ADHD relate to their behavior during parent-child interactions and to their children's disruptive behavior. Findings indicated that mothers’ retrospective self-ratings of ADHD symptoms were related to their present negativity during parent-led play. Mothers’ self-ratings of current inattentive symptoms were related to their impatience during child-led play. Maternal ADHD symptoms were also related to their ratings of their children's ADHD and oppositional-defiant behaviors. Identifying relations between maternal psychopathology, such as ADHD symptomatology, and behaviors during parent-child interactions may yield clues to additional parent behavioral changes that would enhance treatment outcomes for young children with ADHD.

KEYWORDS:

• adult Attention-Deficit/Hyperactivity Disorder
• child
• disruptive behavior
• parent-child interactions

Acknowledgments

This study, the Master's thesis of Alison R. Zisser (supervised by Sheila M. Eyberg), was funded by the National Institute of Mental Health (R01-MH-072780). The authors thank the members of the UF Child Study Laboratory for their contributions to this study.

Alison R. Zisser is now at the Kennedy Krieger Institute.

Notes

Note. The subscript _p indicates a parent category; the subscript _c indicates a child category. Category definitions from the Dyadic Parent Child Interaction Coding System (Eyberg, Nelson, Ginn, Bhuiyan, & Boggs, 2012).

Note. n = 55 mother-child dyads, except for correlations involving the ASR where n = 54 due to missing data for one family. CAARS = Conners’ Adult ADHD Rating Scale; ASR Attention Problems = Adult Self-Report, Attention Problems Syndrome Scale; WURS = Wender Utah Rating Scale.

**p < .01.

Note. Due to missing data, n values varied based on the measures correlated; n = 53 mother-child dyads for correlations involving the CAARS and CBCL, n = 51 mother-child dyads for correlations involving the CAARS and SNAP-IV, n = 52 mother-child dyads for correlations involving the ASR and CBCL, n = 50 for correlations involving ASR and SNAP-IV. CAARS = Conners’ Adult ADHD Rating Scale; ASR Attention Problems = Adult Self-Report, Attention Problems Syndrome Scale; WURS = Wender Utah Rating Scale.

^† p < .10. *p < .05. **p < .01.

Note. n = 54 mother-child dyads, except correlations involving ASR where n = 53 due to missing data for one family. CAARS = Conners’ Adult ADHD Rating Scale; ASR Attention Problems = Adult Self-Report, Attention Problems Syndrome Scale; WURS = Wender Utah Rating Scale.

†p < .10. *p < .05. **p < .01.

Due to missing data, sample size was smaller for analyses involving the SNAP-IV (n = 52) and the CBCL (n = 53). Analyses involving both the SNAP-IV and CBCL included only 50 families.

It is important to note the particularly low Cronbach's alpha value for the CBCL Attention Problem Syndrome Scale for the CBCL version completed by mothers of 4- and 5-year-old participants (i.e., Cronbach's alpha of −.04). This finding was surprising given the relatively high reliability coefficient reported in the normative sample and the prevalent use of the CBCL as a measure of child inattention and hyperactivity in the child clinical literature (Achenbach & Rescorla, 2001). There are several factors we believe contributed to the low reliability of this scale in our sample. First, all of the children in our sample had significant hyperactive symptomatology as reported by both parent and teacher. Thus, there was limited variability in responses to the Attention Problems Syndrome Scale. Additionally, the CBCL scale is restricted to three responses—i.e., 0 = Not True (as far as you know); 1 = Somewhat or Sometimes True; 2 = Very True or Often True. Limiting the scale to three points further restricts the range of responses on this scale. Given that alpha values are based on correlations among scale items, limited variability will affect alpha. Finally, analysis of Cronbach's alpha if single scale items were deleted shows the only item that significantly changes alpha if deleted is the item “Poorly Coordinated or Clumsy.” Deletion of this item improves Cronbach's alpha from −.04 to .26. Although this alpha value is still quite low, the magnitude of change indicates that this item was particularly problematic for this sample.

It is also important to note that although stronger than the CBCL Attention Problems Syndrome Scale, the reliability for the Attention Deficit/Hyperactivity Scale was also quite low. We believe that restricted variability in the sample as well as the limited range of the response options affected the reliability of this scale as well. Given the low reliability of the CBCL ADHD-related scales, results of analyses using these scales must be interpreted with caution.