Metaparenting: associations with parenting stress ... - Springer Link

ADHD Atten Def Hyp Disord (2012) 4:1–10 DOI 10.1007/s12402-011-0068-9

ORIGINAL ARTICLE

Metaparenting: associations with parenting stress, child-rearing practices, and retention in parents of children at risk for ADHD Leanne Tamm • George W. Holden • Paul A. Nakonezny • Sarah Swart • Carroll W. Hughes

Received: 6 September 2011 / Accepted: 19 November 2011 / Published online: 3 December 2011 Ó Springer-Verlag 2011

Abstract The aim of the study is to investigate metaparenting (effortful, deliberate cognition about parenting) in parents of children at risk for ADHD including predictors, correlates, and intervention outcomes. Parents (n = 68) of children with significant ADHD symptoms (i.e., C6 inattentive or hyperactive/impulsive symptoms with impairment in C2 settings, mostly un-medicated) provided ratings of metaparenting, parenting stress and practices, and child ADHD symptoms before and after parent training. Parents were predominantly Caucasian, in their upper thirties, and most had schooling beyond high school. We investigated the relation between metaparenting and baseline predictors, and whether metaparenting predicted (1) parenting behaviors at baseline, (2) attrition, and (3) parenting stress and parent/child behaviors at outcome. More educated mothers, with fewer people living in the home, and higher levels of parenting stress, reported more metaparenting. Parents with lower problem-solving and assessing scores reported more inconsistent parenting, and those with lower problemsolving scores were more likely to drop out of parent training. Higher problem-solving and reflecting scores at

L. Tamm (&) Behavioral Medicine and Clinical Psychology, Cincinnati Children’s Hospital Medical Center, 3333 Burnet Ave. MLC 10006, Cincinnati, OH 45229-3039, USA e-mail: [email protected] G. W. Holden Southern Methodist University, Dallas, TX, USA P. A. Nakonezny C. W. Hughes University of Texas Southwestern Medical Center at Dallas, Dallas, TX, USA S. Swart June Shelton School and Evaluation Center, Dallas, TX, USA

baseline were associated with more parental stress. Higher reflecting at baseline predicted child hyperactivity/impulsivity at outcome. Our findings indicate metaparenting is associated with parenting behaviors and decisions to complete parent training. Furthermore, metaparenting appears to be a complex, finely nuanced construct with both positive and negative associations with reports of parenting practices and stress. Keywords Metaparenting ADHD Retention Parent training

Introduction Parental cognitions are increasingly being appreciated as key determinants of parenting and therefore prime targets for intervention. For example, parental attributions regarding the causes of their children’s behavior directly affects their parenting choices and responses to misbehavior (Miller 1995). Parental feelings of self-efficacy are also associated with parenting behavior, such as their responsiveness to their child and the decision to attend parent training programs (Hoza et al. 2000). Furthermore, parental cognitions are associated with enrollment and engagement in child treatment programs (Mah and Johnston 2008) and recognized to be a fruitful focus of parenting intervention programs (Maloney and Altmaier 2007). Children with attention-deficit/hyperactivity disorder (ADHD) are particularly challenging to parents (Pimentel et al. 2011) due to the nature of the disorder itself. These children need frequent reminders to do the most basic of tasks because they are forgetful, disorganized, impulsive, and often require close monitoring to complete daily tasks; homework, mealtime, and morning/evening routines can be

123

2

particularly challenging. Therefore, studying how parental cognitions are affected is important. The domain of parenting is particularly crucial when considering intervention with this vulnerable population due to questions of parenting child-rearing strategies, treatment adherence and response, as well as child treatment outcomes. There is little research on the role parental cognitions play for parents of children with ADHD. To date, parental expectations and beliefs have been found to be associated with enrollment and engagement in treatment (Corkum et al. 1999; Harrison and Sofronoff 2002; Johnston et al. 2005), parental distress (Johnston and Mash 2001), as well as child treatment outcomes (Hoza et al. 2000). Notably, parental expectations regarding treatment predict treatment attendance and dropout more than socioeconomic status, parental stress/psychopathology, and severity of child’s problem (Nock and Kazdin 2001). Research investigating determinants of negative parenting (e.g., lax or inconsistent parenting or punitive parenting), known to characterize the parenting practices of parents of children with ADHD, is critical because negative parenting has been linked prospectively with behavior problems (McKee et al. 2004). Recently, a new parenting construct has been articulated. Labeled ‘‘metaparenting,’’ it refers to the deliberate thoughts or effortful cognitions parents have regarding parenting itself (Hawk and Holden 2006; Holden and Hawk 2003). Given that much of the parental social cognition literature has focused on implicit processes (i.e., those occurring with limited parental awareness) or attributions and attitudes, this new construct centers on deliberate thoughts about parenting itself (Hawk and Holden 2006). Metaparenting is theorized to promote effective parenting, to reduce child-rearing stress through seeking of social support and knowledge about parenting, and to drive parental transformations (Holden and Hawk 2003). However, it is a complex construct that is multiply determined by such factors as demographics of the parent (culture, age, gender), child characteristics (age, gender, difficultness), and the context (e.g., stress levels, resources). In addition, metaparenting can both drive parenting behaviors and reflect problems in the parent–child relationship. For example, parents who spend time investigating and testing solutions to common problems may be effective parents. On the other hand, parents who are dealing with challenging, perplexing child behaviors may also devote a comparable amount of time to thinking about the problem but without arriving at a successful strategy. Thus, time devoted to metaparenting may reflect a parent effect, a child effect, or a relationship effect. The metaparenting profile questionnaire (MPPQ) (Holden and Hawk 2003) was developed to measure anticipating, assessing, problem solving, and reflecting. Anticipating

123

L. Tamm et al.

refers to thinking about aspects of childrearing before they take place (e.g., researching schools to find a suitable school district before the child starts kindergarten); assessing is the appraisal/evaluation of the parent, child, and environment (e.g., parents monitoring child’s development of social/peer interactions); problem solving concerns the multi-step process combining the ability to recognize a problem, strategize a solution, execute the solution, and assess the outcome; and finally reflecting describes consideration of past experience—a ‘‘re-assessment’’ of parent–child behaviors and interactions (e.g., parents thinking about a previously imposed punishment and determining whether the desired change in behavior was achieved). In an initial investigation of the construct, using a sample of 116 mothers of young children (ages 4–6 years), metaparenting was found to be associated with several variables (Hawk and Holden 2006). Results revealed mothers reported engaging in frequent metaparenting and that it could be measured reliably. However, contrary to some of the hypotheses, mothers’ age and education were not significant predictors, although number of children, child age, and child gender were. Mothers with fewer children reported more problem solving, mothers of boys reported more anticipating, and mothers with younger children reported higher total metaparenting. Parental stress did not predict metaparenting, but mothers with more life stressors had higher levels of reflecting. Finally, childrearing laxness and over-reactivity were negatively associated with problem solving. Thus, the results showed that the MPPQ has utility in measuring metaparenting and that the construct of metaparenting was related to reported parenting behaviors. Given the argument that metaparenting may drive parental transformation and the initial findings showing relations between metaparenting and child-rearing behavior, we investigated the construct in parents of young children at risk for ADHD who participated in a behavioral parent training program. One goal of this study was to determine whether the findings of the previous study of metaparenting in parents of typically developing children (Hawk and Holden 2006) would be replicated in parents of young children at risk for ADHD. A second goal was to determine whether metaparenting at baseline predicted parenting stress, reported parenting practices, and child outcomes after participation in the parent training program. Finally, a third goal was to examine whether metaparenting predicted retention in the parent training program. Given that there have been no previous studies of metaparenting in parents of children with ADHD receiving parent training, this is viewed as an exploratory study, and no specific hypotheses were made. If initial levels of metaparenting are related to such issues as attrition or treatment success, then the construct could prove to be an important new

Metaparenting: associations with parenting stress

assessment variable to utilize with parents heading into interventions.

Methods Participants Participants consisted of 68 parents of children at risk for ADHD (mean age 5.18 years ±1.33; 69% male) who were enrolled in behavioral parent training programs conducted as part of a research study at the University of Texas Southwestern Medical Center. Participants were recruited through referrals, advertisements, and mailings. Data were collected from 63 mothers and 5 fathers (note: we did not collect information regarding marital status of parents). The mean age for the parents was 37.82 ± 5.78 years. Additional descriptive characteristics are reported in Table 1. Procedures To participate in the study, parents (n = 75) completed a phone screen consisting of questions about the child’s developmental history, family history, ADHD symptoms and impairment (based on the ADHD 18 symptoms listed in the DSM-IV), and medication status. Inclusion criteria consisted of parent report of six or more inattentive and/or six or more hyperactive/impulsive symptoms of ADHD with an age of onset prior to age 7 years and reported impairment in two or more settings, although no formal diagnostic procedures were conducted. Parents were also asked about previous or current psychological diagnoses. Children with a reported history of pervasive Table 1 Demographic characteristics of sample (n = 68)

3

developmental disorder or current mood or anxiety disorder were excluded. Eligible families (n = 68) attended one of four behavioral parent training programs utilizing the community parent education program (COPE) (Cunningham et al. 1998) model between April 2008 and March 2009. COPE is a community-based manualized program that is offered to large groups ranging in size from 10 to 14 people including one or both parents and sometimes other primary caregivers. The topics covered each week were various parenting strategies, including the following: focusing attention on positive behavior, balancing attention, ignoring, developing transitional strategies, implementing reward systems, and strategies for dealing with defiant behaviors. Groups met once a week for 2 h for a 10-week period. Prior to participation, parents provided informed consent in accordance with procedures established by the University of Texas Southwestern Medical Center Institutional Review Board. During the first and last sessions of COPE, one parent (typically the mother) completed self-report forms on metaparenting, parenting stress, parenting practices, and ADHD symptoms in her child. Parents also completed a demographics form at baseline. Measures Metaparenting The MPPQ (Holden and Hawk 2003) was used to analyze parental cognitions. Parents responded to 16 questions about metaparenting using a Likert-type scale (i.e., 1 = never/rarely to 5 = constantly). The four Likert-type subscales—assessing (five items evaluating how the child

Ethnicity of parent (%)

78.0% European American 4.4% African American 13.2% Hispanic

Education of parent (%)

4.4% Other 4.4% High school graduate 25% Partial college/vocational training 44.1% College graduate 26.5% Graduate degree

Parent demographic data describe parents who completed the ratings ADHD attention/deficithyperactivity disorder, ODD oppositional defiant disorder

Number of people living in the home (mean ± SD)

4.0 ± 1.0

Number of parent training sessions attended out of 10 (mean ± SD)

7.7 ± 2.4

Percentage of parents completing parent training program

72%

Parenting stress (raw score; mean ± SD)

82.78 ± 24.17

Parent rating of child ADHD symptoms—combined (mean ± SD)

1.95 ± .65

Parent rating of child ODD symptoms (mean ± SD)

1.29 ± .84

Percentage of children receiving ADHD medication

13%

123

4

is developing), anticipating (three items regarding thinking in advance about childrearing situations), problem solving (five items regarding generating solutions to problems), and reflecting (three items related to having concerns about the child or childrearing)—were scored by summing the items loading on that factor. A total metaparenting score was also calculated by summing responses to the 16 questions. Higher scores typically reflect more frequent cognitions in each domain. Parenting stress The parenting stress index-short form (PSI-SF; Abidin 1995) was used to assess parent stress. The PSI-SF is a parent self-report measure containing 36 items covering three domains of parent stress: parental distress, parent– child dysfunctional interaction, and difficult child. Total parenting stress was also measured. Items were rated on a Likert-type scale from 1 (Strongly Disagree) to 5 (Strongly Agree). Parenting practices The Alabama parenting questionnaire-preschool version (APQ-P; Clerkin et al. 2007) was used to assess childrearing practices. The APQ-P is a 24-item self-report questionnaire with answer choices on a Likert-type scale from 1 (Never) to 5 (Always) and is used to assess three domains of parenting: positive parenting, inconsistent parenting, and punitive parenting. Disruptive behavior The Swanson, Nolan, and Pelham (SNAP) ADHD rating scale (http://www.adhd.net) was used to assess the child’s ADHD symptoms. Parents rated how much the symptoms describe their child on the nine inattentive symptoms and nine hyperactive-impulsive symptoms using a 0 (never/not at all) to 3 (often/very much) scale. The SNAP rating scale also includes eight Oppositional Defiant Disorder (ODD) symptoms. An average score of 1.67 on the parent-rated ADHD combined items and an average score of 1.88 on the parent-rated ODD items are considered to be in the clinically significant range according to the scoring instructions. Statistical analysis Demographic characteristics for the overall sample are described using the sample mean and standard deviation for continuous variables and the percentage for categorical variables.

123

L. Tamm et al.

Do baseline demographics predict metaparenting? Multiple linear robust regression with MM estimation was conducted on each of the four MPPQ subscales (baseline values). For each multiple linear robust regression, three maternal variables (age, education, number of people living in the home—the latter being our proxy for parity), two child variables (age, gender), and one contextual variable (total parenting stress from the PSI-SF) were included as predictors (covariates) in each model; these variables were selected because they most closely matched predictors used in the previous study of the MPPQ (Hawk and Holden 2006). We excluded fathers from these analyses due to the limited sample size (n = 5). Does baseline metaparenting predict parenting practices? Multiple linear robust regression with MM estimation was used to examine the relation between each baseline MPPQ subscale (predictor) and parenting practices (inconsistent parenting, positive parenting, and punitive parenting). For each multiple linear robust regression, the four baseline MPPQ subscales (assessing, anticipating, problem solving, and reflecting) and total metaparenting were included as predictors and each measure of parenting practices was a dependent variable in a separate model. Is there a relationship between ADHD severity and metaparenting? To investigate the relationship between ADHD severity and metaparenting, the Pearson product–moment correlation coefficient (r) was used to assess the linear relationship between each metaparenting subscale and ADHD severity at baseline (i.e., average of responses to all 18 ADHD items on the SNAP rating scale). We also ran regression analyses similar to that described above, which included ADHD baseline severity as a covariate, but these results were essentially the same as those obtained without the inclusion of the ADHD severity covariate; thus, we report the results without the inclusion of the ADHD severity covariate. Does baseline metaparenting predict parent and child response to parent training at outcome? Simple linear robust regression with MM estimation was used to examine the relationship between each baseline MPPQ subscale (predictor) and each dependent variable of parent stress (parental distress, parent–child dysfunctional interaction, difficult child, and total parenting stress) and parenting practices (inconsistent parenting, positive parenting, and punitive parenting) at outcome (i.e., after parent


5

Table 2 Multiple linear robust regression models for metaparenting predictors (measured at baseline) Predictor

Total metaparenting v2(1) b (SE) R2 = .19

Assessing v2(1) b (SE) R2 = .04

Anticipating v2(1) b (SE) R2 = .09

Reflecting v2(1) b (SE) R2 = .32

Problem solving v2(1) b (SE) R2 = .13

Maternal age

v2 = .14

v2 = .01

v2 = .04

v2 = .03

v2 = .57

Maternal education People living in home Child age Child gender Total parenting stress

.07 (.20)

-.01 (.08)

.01 (.04)

.01 (.05)

.05 (.07)

v2 = 8.18**a

v2 = 1.10

v2 = 2.56

v2 = 4.78*

v2 = 5.80*

3.96 (1.38)

.56 (.54)

.43 (.27)

.69 (.32)

1.17 (.48)

v2 = 6.69**a

v2 = 1.00

v2 = 3.24

v2 = .41

v2 = 7.40**a

-2.88 (1.11)

-.45 (.45)

-.39 (.22)

-.16 (.25)

-1.07 (.39)

2

v = 1.41

2

v = .82

2

2

v = .03

v = 4.04*

v2 = .51

1.11 (.94)

.32 (.36)

.03 (.18)

.43 (.22)

-.23 (.32)

v2 = .23

v2 = .03

v2 = 1.33

v2 = .26

v2 = .10

1.24 (2.57)

-.17 (1.00)

.58 (.50)

.30 (.59)

-.28 (.90)

v2 = 6.78**a

v2 = .29

v2 = .21

v2 = 37.42**a

v2 = .03

.13 (.05)

.01 (.02)

-.004 (.01)

.07 (.01)

.003 (.02)

MPPQ metaparenting profile questionnaire, b parameter estimate, SE standard error * Uncorrected P \ .05; ** uncorrected P \ .01 a

Remained significant even after correcting for multiple testing via Bonferroni method

training), as well as ratings of children’s ADHD symptoms at outcome in separate models.

testing via the Bonferroni method (calculated within each analytic question), and we provide the R2 from each linear regression model as an effect size estimator.

Does baseline metaparenting predict attrition? Simple logistic regression was used to estimate the odds of attrition from each MPPQ subscale in a separate model. The odds ratio can be interpreted as the effect size estimator. The 95% Wald confidence intervals were calculated for each odds ratio, and the Wald Chi-square statistic was used to test for a significant association between each MPPQ subscale and attrition (binary). What changes were observed after parent training? To investigate whether the COPE intervention was associated with changes in parenting stress, child-rearing practices, and ADHD ratings of children, paired sample t tests (one-tailed) were performed comparing baseline and outcome ratings in these three domains. Because we expected directional mean differences here, the paired sample t tests were conducted as one-tailed hypothesis tests. We performed statistical analyses using SAS software, version 9.2 (SAS Institute, Inc., Cary, NC, USA) and SPSS, version 19. The level of significance was set at a = .05 (two-tailed for all tests except for the paired sample t tests). Because of the exploratory nature of the study and to assist with future hypothesis testing, we present the results (and P values) without correction for multiple testing; however, in Tables 2, 3, and 4 we also demarcate which results remained significant after correcting P values for multiple

Results Results are organized around four questions: (1) Are parent or child characteristics associated with differing levels of metaparenting at baseline?; (2) Was metaparenting related to parenting or child variables when training was completed?; (3) Was metaparenting associated with attrition?; and (4) What changes were observed in parenting? We first evaluated whether variations in metaparenting levels assessed at baseline were associated with five mother and child variables. Multiple linear robust regression revealed that metaparenting at baseline was significantly predicted by two maternal variables: maternal education (total metaparenting, problem solving, and reflecting) and number of family members living in the home (total metaparenting, problem solving), and one child variable— child age (reflecting; Table 2). Maternal education was related to more metaparenting though number of family members was associated with less metaparenting. Parents of older children engaged in more metaparenting. Neither maternal age nor child gender was found to be associated with the construct. Next, we examined whether metaparenting at baseline was related to child-rearing practices at baseline. Inconsistent parenting at baseline was significantly predicted by assessing and problem solving (Table 3). We observed significant moderate correlations

123

6 Table 3 Multiple linear robust regression models for parenting predictors at baseline

L. Tamm et al.

MPPQ predictor

Inconsistent parenting v2(1) b (SE) R2 = .27

Positive parenting v2(1) b (SE) R2 = .07

Punitive parenting v2(1) b (SE) R2 = .13

Total metaparenting

v2 = 31.19**a

v2 = 1.05

v2 = .77

.44 (.20)

-.31 (.31)

.12 (.14)

Assessing

v2 = 4.04* -.52 (.26)

v2 = .82 .34 (.38)

v2 = 1.37 -.21 (.18)

MPPQ metaparenting profile questionnaire, b parameter estimate, SE standard error

Anticipating

v2 = .30

v2 = .03

v2 = .24

-.16 (.29)

-.07 (.43)

.10 (.20)

* Uncorrected P \ .05; ** uncorrected P \ .01

Reflecting

v2 = .06

v2 = .38

v2 = .00

.39 (.63)

.001 (.29)

a


-.11 (.42) Problem solving

2

v = 13.08** -1.03 (.28)

between ADHD severity at baseline and assessing (r = .26, P \ .05), anticipating (r = .38, P \ .05), reflecting (r = .43, P \ .05), and total metaparenting (r = .39, P \ .05); problem solving was not significantly correlated with ADHD severity (r = .21, P = .08). We then examined whether metaparenting at baseline was related to parenting stress or child-rearing practices at outcome. Simple linear robust regression investigating the relation between each baseline metaparenting variable and parenting stress and parenting practices variables at outcome revealed significant relations for two metaparenting subscales (Table 4). Problem solving at baseline predicted stress of a difficult child at outcome and reflecting at baseline predicted all four PSI-SF scales (parental distress, parent–child dysfunctional interactions, difficult child and total parenting stress). Simple linear robust regressions investigating the relation between each baseline metaparenting variable and ratings of children’s ADHD at outcome showed that reflecting at baseline predicted ratings of child hyperactivity/impulsivity (Table 4). Results were not significant for the assessing and anticipating subscales. In terms of predicting attrition, simple logistic regression estimating the odds of dropping out from the intervention showed problem solving significantly predicted retention. Specifically, we found that a 1-unit increase in the problem-solving subscale was associated with 0.85 times (or 15% decrease in) the predicted odds of not completing parent training [v2(1) = 4.47, P \ .04]. This finding indicates that parents with higher problem-solving scores at baseline had lower odds of dropping out of the program. Logistic regression results for the other MPPQ subscales predicting attrition were not significant [assessing: v2(1) = .03, P = .86; anticipating: v2(1) = .02, P = .89; reflecting: v2(1) = .34, P = .56]. The final question we examined was whether there were significant changes in child rearing and reported child

123

a

2

v = 3.37

v2 = 3.13

.79 (.43)

-.34 (.19)

behavior from baseline to completion of training. Results from the paired sample t tests revealed that there was a significant reduction in reports of inconsistent parenting [t(49) = 2.52, P \ .05, d = .36], parental distress [t(48) = 1.69, P = .05, d = .25], child inattention [t(49) = 2.39, P \ .05, d = .38] and hyperactivity/impulsivity [t(49) = 2.33, P \ .05, d = .32]. Significant changes were not found in the APQ subscales positive parenting or punitive parenting or the PSI-SF subscales parent–child dysfunctional interactions, difficult child, or total stress.

Discussion The purpose of this study was to examine the role that parental thinking about their own child-rearing played in families with children at risk for ADHD. Metaparenting is theorized to be multiply determined by background variables, parent and child characteristics, contextual variables, as well as child behavior variables (Holden and Hawk 2003), and this study adds supporting evidence to that claim. In this sample of treatment-seeking parents, we found differing levels of baseline metaparenting were associated with two parent variables and one child variable. First, more educated parents as well as parents with fewer family members living at home reported engaging in more problem solving and reflecting. These relations were investigated in a previous study (Hawk and Holden 2006) but not observed, perhaps due to the lack of variability in that sample. Years of formal education has only rarely been linked to parental cognitions (Bornstein 2002) including promoting more deliberate thought that is a goal of parent training (Fine 1989). When fewer people are in the home, parents may engage in more reflecting likely due to fewer time constraints (or that it is easier to focus on one child vs. multiple children). In addition, child age was positively

Metaparenting: associations with parenting stress Table 4 Simple linear robust regression models for MPPQ predicting parenting practices/ stress at outcome

7

MPPQ predictor (preparent training)

Outcome variable (postparent training)

Assessing

Positive parenting

0.11 (0.21)

-0.30–0.53

-0.12 (0.14)

-0.39–0.15

Punitive parenting

-0.11 (0.08)

-0.28–0.06

Parental distress

0.45 (.33)

-0.20–1.10

P–C Dys. interaction

0.27 (0.28)

-0.28–0.81

Difficult child

0.64 (0.49)

-0.32–1.59

Total parenting stress

1.20 (1.95)

-0.66–3.07

Child inattention

0.01 (0.03)

-0.04–0.07

-0.002 (0.04)

-0.08–0.07

Positive parenting Inconsistent parenting Punitive parenting Parental distress P–C Dys. interaction

Problem solving

0.02 (0.36)

-0.70–0.74

-0.45 (0.24) -0.08 (0.14)

-0.91–0.02 -0.36–0.20

0.49 (0.58)

-0.65–1.64

-0.08 (0.49)

-1.03–0.87

Difficult child

1.08 (0.81)

-0.51–2.67


1.22 (1.63)

-1.97–4.42

Child inattention

0.05 (0.04)

-0.04–0.14

Child hyperactivity/impulsivity

0.03 (0.07)

-0.10–0.16

0.08 (0.25)

-0.40–0.57

Inconsistent parenting

Positive parenting

-0.08 (0.15)

-0.37–0.21

Punitive parenting

-0.08 (0.10)

-0.27–0.12

Parental distress

0.44 (0.38)

-0.31–1.18


0.12 (0.32)

-0.49–0.74

Difficult child

Reflecting

95% confidence interval


Child hyperactivity/impulsivity Anticipating

b (SE)

1.23* (0.50)a

0.25–2.21

R2

.12


1.77 (1.04)

-0.26–3.80

Child inattention Child hyperactivity/impulsivity

0.03 (0.03) 0.06 (0.04)

-0.02–0.09 -0.02–0.14

-0.17 (0.30)

-0.76–0.42

0.03 (0.20)

-0.35–0.42

Punitive parenting

-0.08 (0.12)

-0.31–0.15

Parental distress

1.13* (0.46)a

0.24–2.02


0.90* (0.39)

0.16–1.64

.09

Positive parenting

MPPQ metaparenting profile questionnaire, P–C Dys. parent– child dysfunctional interaction, b parameter estimate, SE standard error


* Uncorrected P \ .05; ** uncorrected P \ .01

Difficult child

2.72** (0.52)a

1.71–3.74

.33


4.63** (1.19)a

2.28–6.97

.20

a

Child inattention


Child hyperactivity/impulsivity

associated with reflecting. It is not surprising that parents of older children, who had longer histories with their children, were more prone to think back about their child-rearing and their children. We did not replicate previous findings showing child age and gender predicted total metaparenting or that parity (i.e., number of people living in the home) predicted problem solving (Hawk and Holden 2006) despite the fact that the samples were quite similar demographically. It may be partly due to the fact that we did not have measures of life stress, maternal need for cognition, or social desirability to include in the model and thus could not fully

0.04 (0.04)

-0.04–0.11

0.12* (0.05)

0.02–0.22

.09

.10

replicate the previous study. However, a more likely explanation is that metaparenting is quite different in parents of children with significant ADHD symptoms who are seeking treatment, as demonstrated by the strong positive relation we observed between parenting stress and metaparenting (total metaparenting and problem solving), and the positive relation between metaparenting and ADHD severity at baseline. Metaparenting is more likely to occur when child-rearing problems are encountered (Holden and Hawk 2003). This is also supported by the observation that the average total parenting stress and metaparenting subscale scores in our sample are significantly higher than in

123

8

the Hawk and Holden sample (e.g., mean reflecting score in our sample at baseline was 10.7 vs. 2.8 in the original sample). We did, however, replicate findings that the problemsolving subscale is negatively related to inconsistent parenting (our proxy for laxness). Engaging in more problem solving likely reflects better ability to self-regulate and deliver appropriate disciplinary responses (Hawk and Holden 2006; Holden 2007). Parents with more difficult to manage children (i.e., those with ADHD symptoms) are more likely to engage in inconsistent discipline. It is particularly difficult to be consistent and to problem solve about how to respond to misbehavior if the child is frequently misbehaving. We also showed a negative relation between assessing and inconsistent parenting, suggesting that parents who engage in less appraisal or evaluation of themselves, the child, and environment, are more likely to engage in erratic parenting. The finding that higher problem solving at baseline was positively related to retention in the parent training program suggests that metaparenting has a direct relationship with treatment adherence and adds to the literature showing a relation between parental cognitions and enrollment and engagement in treatment (Johnston et al. 2005; Corkum et al. 1999; Harrison and Sofronoff 2002; Nock and Kazdin 2001). Lower problem-solving scores at baseline might indicate that parents are less committed and engaged in trying to figure out a solution to their children’s behavior problems. Consequently, they may be more prone to drop out of the program. Future investigations need to replicate and examine in greater depth the relation of problem solving and attrition from intervention programs. Metaparenting at baseline was found to be associated with several outcome variables. Reports of problem solving appeared to be a particularly revealing aspect of metaparenting. Parents who reported engaging in more problem solving were more likely to remain in the program, less likely to report inconsistent parenting, and more likely to experience stress at outcome. Thus, higher problem solving appears to be a promising indicator for adherence to parent training, but that the parent might be in particular need of assistance due to the stressful nature of the parent–child relationship. The other metaparenting subscale that revealed several significant associations at outcome was reflecting. Parents who engaged in more baseline reflecting had higher ratings of parental distress, parent–child dysfunctional interactions, difficult child, and total stress as well as child hyperactivity/impulsivity at outcome. One might have anticipated that higher metaparenting scores at baseline would be associated with better outcomes (i.e., lower rates of parenting stress) if higher levels of metaparenting are considered better than lower levels of metaparenting. However, too much metaparenting may be detrimental to

123

L. Tamm et al.

parents. For example, parents might become locked into an unproductive, overly ruminative approach regarding their child’s behavior resulting in inaction or perhaps become over-focused on their own needs as opposed to those of their child. Alternatively, the focus on parenting management strategies during the parent training course may have resulted in more awareness of their child’s problems and the difficulties parenting children with ADHD, thereby resulting in increased stress—particularly since the class was ending and parents were going to be on their own. The association between reflecting and child hyperactivityimpulsivity suggests that these children, who are much more difficult to manage, may result in perplexed parents. Unsure how to handle their children, these parents may devote considerable time reflecting on the misbehavior (e.g., what caused it, when did it begin, etc.). Thus, parents who have more challenging children, partly evidenced by higher ADHD ratings at outcome, are likely showing a child effect on metaparenting. It should be noted that although baseline metaparenting was positively correlated with ADHD severity, our results did not change substantially after controlling for ADHD severity. It is interesting that the effect sizes were small for the changes we observed in parenting practices, parenting distress and child ADHD ratings after participation in COPE. A relatively recent meta-analysis reported moderate (parenting) to large (ADHD symptoms) effect sizes in pre–postdesign studies investigating behavioral treatments for ADHD including parent management training (Fabiano et al. 2009), and large effect sizes were reported for parent training in general on reducing problem behavior and more moderate effects on parenting stress (Chronis et al. 2006). However, other studies utilizing the COPE program have also reported modest effects on ADHD symptoms (Thorell 2009; Cunningham et al. 1995). The program was developed for use in at-risk community populations and may not have as strong of an effect with clinical populations that may have more severe presenting problems. It should also be noted that the ratings were obtained on the last session of the parent training intervention which involved an active lesson. It would have been ideal to collect ratings at a later point to investigate whether there was delayed evidence of improvement after parents had time to practice and implement all the skills taught. Regardless, it is clear from the ratings that this group was demonstrating clinically significant ADHD symptomatology and parenting stress, and thus the MPPQ ratings shed light on metaparenting in parents of children at risk for ADHD. There are some limitations to this exploratory study. First, the fact that it is the first effort at studying metaparenting in parents with children with significant ADHD symptomatology meant that we did not have specific hypotheses in the absence of any prior research on the topic.


Further, the MPPQ, although shown to be reliable, still requires validation studies, and as yet, there are no normative data on this instrument. These studies are particularly necessary since metaparenting is a complex construct that can both drive and reflect parent–child issues. Another limitation is that we did not have a full diagnostic interview to confirm diagnosis and assess for other comorbidities that might contribute to the findings. It should be noted that less than 30% of the sample were categorized as at risk for ODD based on the SNAP-IV (i.e., score [1.87 on the ODD items). The sample is also relatively limited in size. However, we did utilize robust regression methodology that provides resistant (stable) estimates and results (in the presence of outliers and departure from normality). Also, we did not measure actual parental behavior or include measures of parent psychopathology. Given that parents of children with ADHD often have ADHD themselves, are at risk for higher rates of anxiety and depression (Evans et al. 2008), and parental ADHD symptoms affect child treatment outcomes (Sonuga-Barke et al. 2002), future research should include those variables. In conclusion, the findings contribute to our understanding of the role metaparenting plays in parents of children at risk for ADHD with parenting behaviors, parenting stress, and retention in behavioral parent training. The results reveal that metaparenting is indeed associated with parenting behaviors and whether parent training was completed. Further, metaparenting appears to be a complex, finely nuanced construct that relates both positively and negatively to demographic characteristics, parenting practices, and stress. Additional research will be needed to replicate these findings and to further examine the bidirectional effects of parental cognitions on other dimensions of parenting. Although this study should be regarded as preliminary, it does indicate that metaparenting may be a promising construct for clinicians to use as we found that it was related to attrition and treatment outcome. Acknowledgments We gratefully acknowledge the funding for the Center for Advanced ADHD Research, Treatment, and Education from the Sparrow Foundation. We thank Chad Barnes for data management, and Amanda Gray, Tabatha Melton, and Aleksandra Foxwell for contributions to earlier versions of this manuscript. We are grateful to the parents who participated in our parenting program and provided these ratings. Conflict of interest of interest.

The authors declare that they have no conflict

References Abidin RR (1995) Parenting stress index—short form. Psychological Assessment Resources, Inc., Lutz Bornstein MH (ed) (2002) Handbook of parenting, 2nd edn. Lawrence Erlbaum Associates, Mahway

9 Chronis AM, Jones HA, Raggi VL (2006) Evidence-based psychosocial treatments for children and adolescents with attentiondeficit/hyperactivity disorder. Clin Psychol Rev 26(4):486–502 Clerkin SM, Marks DJ, Policaro KL, Halperin JM (2007) Psychometric properties of the Alabama parenting questionnairepreschool revision. J Clin Child Adolesc Psychol 36(1):19–28 Corkum P, Rimer P, Schachar R (1999) Parental knowledge of attention-deficit hyperactivity disorder and opinions of treatment options: impact on enrolment and adherence to a 12-month treatment trial. Can J Psychiatry 44:1043–1048 Cunningham CE, Bremner R, Boyle M (1995) Large group community-based parenting programs for families of preschoolers at risk for disruptive behavior disorders: utilization, cost effectiveness, and outcome. J Child Psychol Psychiatry 36(7):1141–1159 Cunningham CE, Bremner R, Secord M (1998) COPE the community parent education program: a school based family systems oriented workshop for parents of children with disruptive behavior disorders. COPE Works, Hamilton Evans SW, Schultz BK, Sadler JM (2008) Psychosocial interventions used to treat children with ADHD: safety and efficacy. J Psychosoc Nurs Ment Health Serv 46(8):49–57 Fabiano GA, Pelham WE Jr, Coles EK, Gnagy EM, Chronis-Tuscano A, O’Connor BC (2009) A meta-analysis of behavioral treatments for attention-deficit/hyperactivity disorder. Clin Psychol Rev 29(2):129–140. doi:S0272-7358(08)00156-6 Fine MJ (1989) The second handbook on parent education. Academic, New York Harrison C, Sofronoff K (2002) ADHD and parental psychological distress: role of demographics, child behavioral characteristics, and parental cognitions. J Am Acad Child Adolesc Psychiatry 41(6):703–711. doi:10.1097/00004583-200206000-00010 Hawk CK, Holden GW (2006) Metaparenting: an initial investigation into a new parental social cognition construct. Parent Sci Pract 6(4):321–342 Holden GW (2007) Thinking about child rearing: meta-parenting as a key component of effective parenting. In: Holden GW, Hawk CK (eds) Social cognition models of sensitive/effective parenting: competing or complementary? Symposium presented at the meeting of the Society for Research in Child Development, Boston, MA Holden GW, Hawk CK (2003) Metaparenting in the journey of child rearing. In: Kuczynski L (ed) Handbook of dynamics in parentchild relations. Sage, Thousand Oaks, pp 189–210 Hoza B, Owens JS, Pelham WE, Swanson JM, Conners CK, Hinshaw SP, Arnold LE, Kraemer HC (2000) Parent cognitions as predictors of child treatment response in attention-deficit/ hyperactivity disorder. J Abnorm Child Psychol 28(6):569–583 Johnston C, Mash EJ (2001) Families of children with attentiondeficit/hyperactivity disorder: review and recommendations for future research. Clin Child Fam Psychol Rev 4(3):183–207 Johnston C, Seipp C, Hommersen P, Hoza B, Fine S (2005) Treatment choices and experiences in attention deficit and hyperactivity disorder: relations to parents’ beliefs and attributions. Child Care Health Dev 31(6):669–677. doi:10.1111/j.1365-2214.2005.00555.x Mah JW, Johnston C (2008) Parental social cognitions: considerations in the acceptability of and engagement in behavioral parent training. Clin Child Fam Psychol Rev 11(4):218–236. doi: 10.1007/s10567-008-0038-8 Maloney R, Altmaier E (2007) An initial evaluation of a mindful parenting program. J Clin Psychol 63(12):1231–1238. doi: 10.1002/jclp.20395 McKee TE, Harvey E, Danforth JS, Ulaszek WR, Friedman JL (2004) The relation between parental coping styles and parent-child interactions before and after treatment for children with ADHD and oppositional behavior. J Clin Child Adolesc Psychol 33(1):158–168. doi:10.1207/S15374424JCCP3301_15

123

10 Miller SA (1995) Parents’ attributions for their children’s behavior. Child Dev 66(6):1557–1584 Nock MK, Kazdin AE (2001) Parent expectancies for child therapy: assessment and relation to participation in treatment. J Child Fam Stud 10:155–180 Pimentel MJ, Vieira-Santos S, Santos V, Vale MC (2011) Mothers of children with attention deficit/hyperactivity disorder: relationship among parenting stress, parental practices and child behaviour. Atten Defic Hyperact Disord 3(1):61–68. doi:10. 1007/s12402-011-0053-3

123

L. Tamm et al. Sonuga-Barke EJ, Daley D, Thompson M (2002) Does maternal ADHD reduce the effectiveness of parent training for preschool children’s ADHD? J Am Acad Child Adolesc Psychiatry 41(6): 696–702. doi:10.1097/00004583-200206000-00009 Thorell LB (2009) The community parent education program (COPE): treatment effects in a clinical and a community-based sample. Clin Child Psychol Psychiatry 14(3):373–387. doi:10.1177/ 1359104509104047