Parent Outcomes Following Participation in Cognitive Behavior Therapy for Autistic Children in a Community Setting: Parent Mental Health, Mindful Parenting, and Parenting Practices

ABSTRACT

Parents of autistic children are at a higher risk for mental health problems, including anxiety, depression, and stress. Cognitive behavior therapy (CBT) that targets children's emotion regulation may have an indirect influence on parent outcomes, especially if they play a supporting role in their child's intervention. However, most CBT interventions have been carried out in highly controlled research settings and there are a few studies that examined parental outcomes after participating in autistic child‐focused CBT within a community setting. The current study examined parent outcomes (i.e., mental health problems, mindful parenting, and parenting practices) following a community‐based CBT program with concurrent parent involvement for autistic children, as well as associations between changes in parent and child outcomes (i.e., autism symptoms and emotion dysregulation). Participants included 77 parent–child dyads across seven community organizations in Ontario, Canada. Parents reported improved mindful parenting and positive parenting practices post‐intervention, and no significant changes in their mental health. Multiple mediation analyses revealed that positive changes in parent outcomes (i.e., mindful parenting and parenting practices) were associated with positive changes in child emotion regulation. These positive changes in parenting practices mediated the relationship between mindful parenting and child emotion regulation. Results suggest that participating in community‐based CBT is mutually beneficial for autistic children and their parents, particularly in improving parenting behaviors.

Summary.

• Parents of autistic children are at a higher risk for mental health problems, including anxiety, depression, and stress.

• This study examined parent outcomes (i.e., mental health, mindful parenting, and parenting practices) following their involvement in cognitive behavior therapy intervention for their autistic children.

• Parents reported improvement in their mindful parenting and their parenting practices.

• Parent involvement in interventions for autistic children can be beneficial for both of them.

Caregivers of autistic ¹ children often report higher rates of mental health difficulties, such as anxiety, depression, and stress (Bitsika and Sharpley 2004; Hayes and Watson 2013; Lai et al. 2015) and challenges with their approach to parenting (O'Nions et al. 2018; Osborne et al. 2008). The transactional model of development (Sameroff 2009) emphasizes the bidirectional interactions between parent and child, which extends to the relations between child and parent mental health (Neece, Green, and Baker 2012). Behaviors that are challenging or disruptive, and extreme emotional dysregulation in autistic children, have been associated with parents' mental health challenges, especially increased parental stress (Allik, Larsson, and Smedje 2006; Lecavalier, Leone, and Wiltz 2006; Mills et al. 2022; Neece, Green, and Baker 2012). These parent and child factors can have fluctuating effects on each other over time. Piro‐Gambetti et al. (2022) followed 188 families of autistic children for 4 years and found that maternal and paternal depression symptoms at the first timepoint predicted child internalizing issues 12 months later. They also found that children's internalizing problems in the third year predicted maternal depression symptoms in the fourth year. They concluded that child mental health can contribute to parent mental health, and that parent mental health can play an important role in increasing child internalizing mental health issues.

Mindful parenting is a key contributor of parent mental health and parenting practices (Parent et al. 2010; Raulston, Kosty, and McIntyre 2021). Mindful parenting is defined as parents' ability to pay attention to their child and their parenting in a non‐judgmental manner, focusing on each moment, especially during parent–child interactions (Kabat‐Zinn and Kabat‐Zinn 2021). As Bögels and Restifo (2014, 104) indicated, it includes “(1) greater awareness of a child's unique nature, feelings, and needs; (2) a greater ability to be present and listen with full attention; (3) recognizing and accepting things as they are in each moment, whether pleasant or unpleasant; and (4) recognizing one's own reactive impulses and learning to respond more appropriately and imaginatively, with greater clarity and kindness.” Parenting practices are the behaviors or approaches that the parents use during their interaction with their child, which can shape how a child develops. In the literature, categorical approach (i.e., four parenting styles: authoritative, authoritarian, indulgent, and neglecting) and continuous measurement of parent practices (i.e., support, involvement, warmth, and harsh punishment) are both prevalent. Continuous measures of parenting practices are dichotomized into positive and negative parenting practices. Positive parenting practices consist of warmth, quality time, child monitoring, child‐centered behaviors, and positive reinforcements, while negative practices are displayed through negative affect, rejection, hostility, intrusiveness, and neglect/distancing (Amato and Flower 2002). Mindful parenting is theoretically different from parenting practices as it involves parents' cognition, internal states, and behavior. In families of autistic children, mindful parenting has been related to improved parent mental health issues (Beer, Ward, and Moar 2013; Cheung, Leung, and Mak 2019; Conner and White 2014; Jones et al. 2014; Oñate and Calvete 2019). Cachia, Anderson, and Moore (2016) reviewed the benefits of mindfulness interventions for parents of autistic children and reported associations between improved mindful parenting and reduced stress, depression, and a higher sense of parenting competence. Furthermore, mindfulness interventions have shown positive effects on parenting practices in parents of autistic children including a decrease in laxness (permissive discipline), over‐reactivity (authoritarian discipline), and verbosity (less yelling and threatening statements) (De Bruin et al. 2015; Ridderinkhof et al. 2018).

Due to the bidirectional nature of parent and child factors, mindful parenting is thought to be associated with children's mental health by virtue of its impact on parent mental health and parenting practices (Duncan, Coatsworth, and Greenberg 2009; Kil et al. 2021). Engaging in mindful parenting during dyadic interactions may foster a more positive parental attitude toward their child, ultimately facilitating co‐regulation and contributing to enhanced child mental health. Parent et al. (2016) examined the relationship between parent dispositional mindfulness, mindful parenting, parenting practices, and youth psychopathology across neurotypical youth at three developmental ages: young childhood, middle childhood, and adolescence. Higher levels of mindful parenting and lower levels of negative parenting practices were associated with lower levels of child externalizing and internalizing symptoms. In autistic children, mindful parenting has been associated with decreased child externalizing behaviors (Bögels et al. 2008) and conduct problems (Raulston, Kosty, and McIntyre 2021) and improved parent–child interactions in mental health and community settings (Singh et al. 2014, 2021). More generally, in neurotypical samples, positive parenting practices (i.e., warmth, supportiveness, positive reinforcement) is negatively associated with child emotion dysregulation and mental health challenges (Goagoses et al. 2023). Similarly, in autistic children, positive parenting interventions aiming to increase parent skills, confidence, and improving parenting style, have been found to improve child challenging behaviors and mental health problems (Brian et al. 2021). In contrast, negative parenting practices (i.e., coercive parenting, high levels of criticism, and a negative relationship) can have a negative impact on child functioning (Day et al. 2021; Osborne et al. 2008) and is associated with high levels of challenging behavior in autistic children (Bader, Barry, and Hann 2015).

Interventions that target children's emotional regulation via cognitive behavior therapy (CBT) may have a secondary benefit of supporting parent outcomes. Escovar et al. (2019) has referred to these benefits as vicarious parent outcomes, where parents experience changes in their symptoms despite the child‐focused nature of the intervention. Studies that have examined CBT interventions for autistic children with anxiety and emotional dysregulation have shown that parents of treatment‐responsive children report reductions in trait anxiety (i.e., a general inclination to feel anxious across situations) (Conner, Maddox, and White 2013; Reaven et al. 2015), parent stress (Weiss, Viecili, and Bohr 2015), and improvements in family functioning (Keeton et al. 2013)). Further, Maughan and Weiss (2017) reported that, in a lab‐based randomized waitlist‐controlled trial of parent‐involved CBT for children, parents of children in the treatment immediate group reported a significant decrease in symptoms of depression compared to their counterparts in the waitlist group. There was also a general tendency toward improved scores for parents' anxiety and stress.

Parent involvement in children's interventions has been associated with improvements in parenting practices and strategies. Within the context of treatment, family accommodations, which are behaviors completed by family members to prevent the child from experiencing distress, have been suggested to risk stalling the benefits of CBT by maintaining children's symptoms and family impairment and may be related to parents' own anxiety around parenting (Storch et al. 2015). Parents who were involved in their autistic children's CBT have reported feeling more competent in supporting their children (Sofronoff, Attwood, and Hinton 2005) and to have reduced their family accommodations (Storch et al. 2015). Maughan and Weiss (2017) examined changes in parenting approaches following a child‐focused randomized CBT program and reported an increase in parents' positive comments (e.g., statements of praise, approval, or appreciation), a trend toward a decrease in negative parenting practices (e.g., hostility, laxness, or over‐reactivity) and improvements in self‐reported mindful parenting.

It is important to note that these intervention studies were carried out in highly controlled research efficacy trials, and to the best of our knowledge, very few studies have examined the parent outcomes of child‐focused CBT within a naturalistic community setting for autistic children. It is widely recognized that evidence‐based interventions developed in research settings do not necessarily yield the same results when implemented in community settings due to barriers of intervention delivery at different levels including at the patient, the clinicians, the organization, and/or the policy level (Breitenstein et al. 2010; Damschroder et al. 2009). It is important to examine the effects of child‐focused, parent‐involved CBT on parent mental health, mindful parenting, and parenting practices in a community setting to better support these families' needs. This study fills a gap in the literature by examining the relationship between the changes of parent and child outcome changes following participation in community‐based CBT intervention. This advances our understanding of the bidirectional nature of parent and child mental health in families of autistic children as well as the underlying contributing factors to this relationship. Furthermore, understanding the variability in treatment outcomes for autistic children and their parents is critical to informing intervention choice, design, and adaptations within community settings.

1. Current Study

The aim of the current study was to examine changes in parents' outcomes (parent mental health, mindful parenting, and parenting practices) after participating in a group‐based virtually delivered CBT program for verbally able autistic children within a community service setting and how these changes might be related to child improvements. First, it was hypothesized that parents would show significant improvements in their mental health, mindful parenting, and parenting behaviors from pre‐ to post‐intervention. Second, improvements in parent factors would be associated with improvements in children's emotion regulation and social skills (the targets of the intervention). Third, associations between improvements in mindful parenting and children's emotion regulation and social functioning would be mediated by improvements in parent mental health and parenting practices.

2. Methods

2.1. Participants

This study was approved by the York University Research Ethics Board. Seven urban community provider agencies participated in the current study. These community organizations are publicly funded and had a mandate to deliver programming to autistic children, adapting to online delivery during the COVID‐19 pandemic. Parent–child dyads were recruited by each agency, resulting in a total of 87 dyads, but only 77 dyads completed the program. Reasons for attrition included scheduling issues with the group sessions (n = 2), the program required too much time commitment (n = 2), lack of interest in the intervention theme (n = 1), challenges with virtual format for their child (n = 3), and urgent family obligations (n = 2). Independent sample t‐tests did not show significant differences in child demographics between those who completed the program and those who did not (i.e., age and gender) or pre‐intervention child outcome measures (i.e., social functioning and emotion dysregulation), all ps > 0.05. Similarly, there were no significant differences in parent demographics (i.e., age, gender, household income, education, employment status, marital status) and pre‐intervention parent outcome measures (i.e., parental mental health issues, mindful parenting, parenting behaviors), all ps > 0.05 between those who completed the program and those who did not.

Agencies determined their own recruitment strategies; some recruited from their waitlists, others posted information about the program on their social media account, and some agencies recruited participants through therapist referrals. Most parents were female (95%; 5% male) and between the ages of 27 and 58 years (M = 42.47 years, SD = 5.71). Children were 8–13 years of age (M = 9.90 years, SD = 1.28), and 79% were male. Demographic characteristics of parent–child dyads were collected pre‐intervention through an online survey and are further described in Table 1. Inclusion criteria were (a) children between the ages of 8 and 13 years, who had a diagnosis of autism from regulated healthcare professionals as determined by the autism service provider agency, (b) parent‐reported goals of working on child emotion regulation and social functioning, and (c) having a parent who was able to participate in the parent sessions. Exclusion criteria were (a) the presence of child's intellectual disability via parent report or by screening, (b) child's diagnosis of acute psychosis or conduct disorder, or (c) any behaviors that would create a safety concern for child's group participation, especially when participating remotely from home.

TABLE 1.

Parent and child demographics (n = 77).

	M (SD) or %
Parent age (years)	42.47 (5.71)
Parent age range (years)	27–58 (fairly even/normal distribution)
Household income (CAD) a
< $49,999	13
$50,000–$99,999	20.8
$100,000–$149,999	20.8
$150,000–$200,000 >	22.1
Prefer not to disclose	15.6
Parent education b
No high school diploma	1.3
High school graduate or equivalent	22.3
Associate degree/diploma	21.1
Undergraduate degree or higher	55.2
Employment status
Full time	49.4
Part time	10.4
Unemployed	7.8
Other	32.5
Marital status c
Married/Common Law	77.6
Separated/Divorced	14.5
Single	7.9
Visible minority g (% Yes) d	14.8
Child age (years)	9.90 (1.28)
Child age range (years)	8 to 13 (normal distribution)
Child gender e
Female	20.8
Male	79.2
Child ethnicity f
Asian/South Asian	11.5
White	74.4
Other	14.1

^a 1 missing.

^b 6 missing.

^c 1 missing.

^d 1 missing.

^e 16 missing.

^f Additional categories of ethnicity were listed and were not endorsed.

^g Visible minority refers to “persons, other than Aboriginal peoples, who are non‐Caucasian in race or non‐white in color,” as per the Canadian. Employment Equity Act (SC 1995, c.44).

2.2. Measures

2.2.1. Parent Outcomes

2.2.1.1. Depression Anxiety Stress Scale (DASS‐21)

Parent mental health challenges were measured using the Depression Anxiety Stress Scale (DASS‐21; Lovibond 1995). The DASS‐21 is a self‐report measure composed of 21 items that are divided into three subscales: depression, anxiety, and stress. Parents are asked to rate how much each statement applies to them over the past week using a 4‐point Likert scale from 0 = Never—did not apply to me at all to 3 = Almost always—applied to me very much, or most of the time. Greater mental health problem severity was indicated by higher scores, with total scores ranging from 0 to 63. This measure is often used to assess parent psychopathology in families of autistic children (Lai et al. 2015; Maughan and Weiss 2017; Tajik‐Parvinchi et al. 2020) and has demonstrated excellent internal consistency with this population (α = 0.91 for total score; Maughan and Weiss 2017). For the current sample, internal consistency was excellent for the total score at both pre‐ and post‐interventions (α = 0.90), good for the depression and stress subscales at both timepoints, respectively (α = 0.84, α = 0.87; α = 0.82, α = 0.83), and acceptable for the anxiety subscale at both timepoints (α = 0.75, α = 0.71). Overall, DASS‐21 total scores at pre‐intervention ranged from 1 to 33 (M = 10.96, SD = 7.59), indicating that the majority of the parents in the current sample did not endorse clinical levels of psychopathology. Subscale scores can be categorized as falling within normal, mild, moderate, severe, and extremely severe ranges. For depression, scores in the 10–13 range are deemed mild symptoms (10.3% of current sample at pre‐intervention), scores in the 14–20 range indicate moderate symptoms (11.5%), and scores higher than 21 are considered severe or extremely severe (1.1%). For anxiety, scores in the 8–9 range are considered mild symptoms (14.1%), scores from 10 to 14 are moderate (7.1%), and scores higher than 15 are considered severe to extremely severe (7.1%). For stress, scores in the 15–18 range are considered mild (17.4%), scores in the 19–25 are moderate (12.8%), and scores 26 or above are severe to extremely severe (2.3%).

2.2.1.2. Bangor Mindful Parenting Scale (BMPS)

The Bangor Mindful Parenting Scale (BMPS; Griffith and Hastings 2022; Jones et al. 2014) was used to assess mindful parenting. Parents rated 15 statements on a 4‐point Likert scale from 0 = Never true to 3 = Always true. The BMPS has five mindfulness domains: observing, describing, acting with awareness, non‐reactivity, and non‐judgment. A total score is calculated by summing all scales, with higher scores indicating greater mindful parenting (a score of 45 indicates the highest mindful parenting score). This measure was created for parents of children with intellectual disabilities and/or autism and has demonstrated good internal reliability across multiple studies (α = 0.79; Jones et al. 2014; α = 0.80; Raulston, Kosty, and McIntyre 2021). For the current sample, internal reliability estimate was good, with α = 0.80 and α = 0.82, for the total score at pre‐ and post‐intervention, respectively.

2.2.1.3. Parenting and Family Adjustment Scales (PAFAS)

Parenting approaches and parent and family adjustment were measured using the Parenting and Family Adjustment Scales (PAFAS; Sanders et al. 2014). Parents rated 30 statements on a 4‐point scale from 0 = Not true of me at all to 3 = True of me very much. There are two domains: the parenting domain (18 items) consisting of parental consistency, coercive parenting, positive encouragement, and parent–child relationship, and the family adjustment domain (12 items) incorporating parental emotional adjustment, family relationships, and parental teamwork. Items are summed to indicate levels of dysfunction related to parenting and family adjustment. Scores of 54 and 36 are the highest levels of dysfunction on the parenting domain and the family adjustment domain, respectively. This measure has been shown to be reliable and valid among families of children with autism and other developmental disabilities, including in terms of internal consistency (Mazzucchelli et al. 2018). For the current sample, internal consistency estimates were very good in the parenting domain, with α = 0.76 for both pre‐ and post‐interventions, and the family adjustment domain, α = 0.89 and α = 0.86 at pre‐ and post‐intervention, respectively.

2.2.2. Child Outcomes

2.2.2.1. Social Responsiveness Scale 2^nd Edition (SRS‐2)

To capture children's social functioning and assess autism specific social impairments, the Social Responsiveness Scale (SRS‐2; Constantino and Gruber 2012) was used. The SRS‐2 is a parent‐report measure comprising 65 items that are summed and converted to T‐scores to generate a total score of symptom severity, as well as five subscale scores: social awareness, social cognition, social communication, social motivation, and restricted interests and repetitive behavior. The Social Communication and Interaction T‐score was used in the current analysis by summing the four social subscales. This measure is one of the most widely used measures of children's actual social performance and can be expected to show moderate to large changes within the context of a successful clinical intervention (Wolstencroft et al. 2018). Pre‐intervention, the social communication interaction T‐scores ranged from 54 to 90 (M = 70.23, SD = 8.91).

2.2.2.2. Emotion Dysregulation Inventory (EDI)

Children's emotional regulation difficulties were measured using the Emotion Dysregulation Inventory (EDI; Mazefsky, Day, et al. 2018). The EDI is a brief parent‐report measure that was created specifically to capture emotion regulation impairments in autistic children and adolescents. It consists of two scales: the emotional reactivity scale (seven items) that reflects the inability to regulate intense and rapidly rising negative emotions and the dysphoria scale (six items) that captures the lack of positive affect and motivation and the presence of sadness and anxiousness. For each domain, the items are summed and converted to T‐scores. Pre‐intervention, emotional reactivity scores ranged from 30.10 to 66.70 (M = 50.18, SD = 7.25), and dysphoria scores ranged from 36.40 to 70.30 (M = 48.55, SD = 9.18). This measure has shown excellent validity and reliability (α = 0.94–0.99; Mills et al. 2022) in autistic samples and is sensitive to change (Beck et al. 2020; Mazefsky, Yu, et al. 2018). For the current sample, internal consistency estimates were excellent in the reactivity domain, with α = 0.92 for both pre‐ and post‐intervention, and the dysphoria domain, α = 0.87 and α = 0.92 at pre‐ and post‐intervention, respectively.

2.3. Procedure

2.3.1. Intervention

The Secret Agent Society: Small Group program (SAS:SG) intervention is a manualized, child‐targeted, parent‐involved adapted CBT program that aims at improving autistic children's emotion regulation and social skills (Beaumont 2013). SAS:SG is a multi‐component program consisting of (a) child group sessions, (b) between session skills practice “missions” and activities, (c) concurrent parent group sessions, and (d) school support. The current study was implemented during the COVID‐19 pandemic from September 2020 to December 2021, and the intervention was provided virtually by all agencies. Child sessions were either delivered in a 9‐session (90 min per session) or an 18‐session (45 min per session) format. The sessions were led by either one or two SAS facilitators with a group of 3–4 or 4–6 children. Sessions consisted of a combination of methods such as didactic teaching, in vivo practice, and positive reinforcement. Various topics were covered in sessions such as emotion detection, regulation, problem solving, conversation skills, coping with mistakes and losing, and understanding and managing bullies. Between meetings, children were assigned weekly tasks through the SAS digital headquarters, such as online gamified learning, skills practice, and journal reflections.

SAS parent sessions were delivered in three different group formats, and agencies had the autonomy to choose the schedule that worked the best for them. Possible formats included (1) 9‐week sessions of 45 min per week, (2) 18‐week sessions of 30 min per week, or (3) 2‐h sessions every 3 weeks over 9 weeks. All agencies offered a 2‐h parent information session prior to commencing the program. Parent sessions were delivered concurrently with the child sessions, and their purpose was to provide information to parents on the content of child sessions and to equip parents with the skills to support their children in applying their SAS skills in other settings. It is important to highlight that there was no content that focused on parent mental health, such as improving parent depression, anxiety or stress, or teaching techniques for mindfulness parenting or positive parenting behaviors.

2.4. Data Analysis

Data were analyzed using the Statistical Package for the Social Sciences (SPSS) version 28. Statistical significance was evaluated at the α = 0.05 level. To test the hypothesis that parents would show significant improvements in their mental health, mindful parenting, and parenting behaviors from pre‐ to post‐intervention, paired samples t‐tests were calculated to compare scores on the different measures at pre‐ and post‐intervention, with Cohen's d as a measure of effect size. To test the hypothesis that improvements in parent factors would be associated with improvements in children's emotion regulation and social skills (the targets of the intervention), bivariate analyses using Pearson product moment correlations and 95% bias‐corrected bootstrapped (N = 1000) confidence intervals (CIs) examined the associations among parent and child change scores. Finally, to test the hypothesis that associations between improvements in mindful parenting and children's emotion regulation and social functioning would be mediated by improvements in parent mental health and parenting practices, the possibility of multiple mediators was tested using Hayes' PROCESS macro, embedded in SPSS (Hayes 2012). Pearson product moment correlations and Preacher and Hayes's (2008) bootstrapping procedure (5000 samples) were used to analyze the data (Farmer 2012; Preacher and Hayes 2008). Mediations were significant if the intervals between the lower and upper limits of a 95% CI did not contain zero (Preacher and Hayes, 2008). Two multiple mediation analyses were calculated, examining the relationship between changes in mindful parenting and changes in child emotional dysregulation; the first model focused on child reactivity, while the second model focused on child dysphoria. R ² was calculated as a measure of effect size for the regression models in the multiple mediation analyses.

3. Results

3.1. Parent Outcomes: Parent Mental Health, Mindful Parenting, and Parenting Behaviors

All paired sample t‐test comparisons are reported in Table 2. Paired sample t‐tests indicated no significant change from pre‐ to post‐intervention for parent mental health using the DASS total score, t(74) = 0.73, p = 0.47. Paired sample t‐tests indicated higher mindful parenting post‐intervention in comparison to pre‐intervention on the BMPS total score, t(74) = −3.46, p < 0.001, d = −0.40. The BMPS non‐reactivity and non‐judgment scales were significantly higher from pre‐ to post‐intervention t(75) = −3.50, p < 0.001, d = −0.40 and t(75) = −3.29, p = 0.002, d = 0.38, respectively. However, the BMPS observing, describing, and acting with awareness subscales did not show any significant change from pre‐ to post‐intervention.

TABLE 2.

Parent outcomes paired T‐tests from pre‐ to post‐intervention (n = 77).

Variable	Pre‐intervention M(SD)	Post‐intervention M(SD)	t(df)	Effect size
DASS total score	10.96 (7.59)	10.43 (7.71)	0.73 (74)	0.09
DASS stress	5.84 (3.32)	5.41 (3.28)	1.17 (75)	0.13
DASS anxiety	2.36 (2.71)	2.36 (2.76)	0.00 (74)	0.00
DASS depression	2.70 (2.76)	2.65 (2.93)	0.19 (76)	0.02
BMPS total score	32.39 (5.57)	33.89 (5.41)	−3.46*** (74)	−0.40
BMPS observing	6.46 (1.54)	6.71 (1.63)	−1.51 (75)	−0.17
BMPS describing	7.22 (1.65)	7.32 (1.66)	−0.59 (75)	−0.07
BMPS acting with awareness	7.34 (1.14)	7.49 (1.18)	−1.49 (76)	−0.17
BMPS non‐reactivity	5.09 (1.56)	5.66 (1.53)	−3.50*** (75)	−0.40
BMPS non‐judgment	6.26 (1.76)	6.75 (1.57)	−3.29** (75)	−0.38
PAFAS parenting	11.66 (5.19)	10.67 (4.86)	2.56* (76)	0.29
Coercive parenting	3.62 (2.15)	3.23 (1.82)	2.31* (76)	0.26
Lack of positive encouragement	2.29 (1.49)	1.87 (1.41)	2.34* (76)	0.27
Parental inconsistency	4.32 (2.38)	4.34 (2.32)	−0.07 (76)	−0.01
Poor parent–child relationship	1.43 (2.04)	1.23 (1.90)	1.32 (76)	0.15
PAFAS family adjustment	10.73 (6.52)	10.32 (6.26)	1.04 (59)	0.13
Lack of parental emotional adjustment	5.25 (2.76)	4.74 (2.57)	2.40* (76)	0.27
Poor family relationships	2.88 (2.24)	2.75 (2.04)	0.81 (76)	0.09
Poor parental teamwork	2.65 (2.64)	2.90 (2.84)	−1.41 (59)	−0.18

^* p < 0.05.

^** p < 0.01.

^*** p < 0.001.

In terms of parenting practices, paired sample t‐tests indicated improved scores in the PAFAS parenting domain, t(76) = 2.56, p = 0.013, d = 0.29, but no significant change in the PAFAS family adjustment domain. Within the parenting domain, the coercive parenting and the positive encouragement subscales were significantly improved at post‐intervention, t(76) = 2.31 p = 0.024, d = 0.26 and t(76) = 2.34, p = 0.02, d = 0.27. In terms of the family adjustment domain scales, only parental emotional adjustment significantly improved, t(76) = 2.40, p = 0.02, d = 0.27.

3.2. Child Outcomes: Social Functioning and Emotion Regulation

Paired sample t‐test indicated significant change from pre‐ to post‐ intervention for children's social functioning as measured by the SRS‐2 SCI score (t(76) = 4.61, p < 0.001, d = 0.53). In terms of emotion regulation, paired sample t‐tests indicated improved scores in the child reactivity and child dysphoria (t(75) = 3.40, p = 0.001, d = 0.40; t(75) = 2.33, p = 0.02, d = 0.27), respectively. For further details on child outcomes, please refer to our previous study (Lee et al. 2024).

3.3. Bivariate Correlations Among Parent and Child Outcomes

3.3.1. Parent Outcomes and Child Social Functioning

Changes in parent mental health, mindful parenting, and parenting practices were not associated with changes in children's social functioning as measured by the SRS‐2 SCI score (all p's > 0.05). Even at the subscale level, only parental consistency (one of the PAFAS subscales) was associated with changes in children's SCI scores, r(75) = 0.27, p = 0.02, 95% CI [0.05, 0.46].

3.3.2. Parent Outcomes and Child Emotion Dysregulation

As shown in Table 3, Pearson product moment correlations revealed that improvements in parent mental health symptoms measured using DASS change scores were associated with reductions in child emotional dysregulation measured using EDI change scores in child reactivity (95% CI [0.02, 0.45]) and child dysphoria (95% CI [0.09, 0.50]). Greater improvements in overall mindful parenting measured using BMPS change scores were associated with reductions in child emotional dysregulation measured using EDI change scores in child reactivity (95% CI [−0.49, −0.07]) and child dysphoria (95% CI [−0.52, −0.11]). At the subscale level, improvements in caregivers' acting with awareness and non‐reactivity scores were associated with reductions in child reactivity (95% CI [−0.53, −0.12], 95% CI [−0.44, −0.01]). Moreover, improvements in caregivers' non‐reactivity, observing, and describing child emotional state scores were associated with reductions in child dysphoria (95% CI [−0.44, −0.01]), (95% CI [−0.47, −0.04]), and (95% CI [−0.44, −0.01]), respectively.

TABLE 3.

Bivariate correlations among child emotion dysregulation change scores and parent outcomes change scores.

	Child emotion dysregulation
	Child reactivity	Child dysphoria
Parent mental health	0.25*	0.31**
Parent stress	0.23	0.36**
Parent anxiety	0.21	0.15
Parent depression	0.15	0.19
Mindful parenting	−0.30*	−0.33**
Observing	−0.04	−0.27*
Describing	−0.09	−0.24*
Acting with awareness	−0.34**	−0.18
Non‐reactivity a	−0.24*	−0.23*
Non‐judgment	−0.21	0.05
Parenting	0.37**	0.24*
Coercive parenting	0.21	0.11
Lack of positive encouragement	0.13	0.12
Parental inconsistency	0.40***	0.18
Poor parent–child relationship	0.10	0.16
Family adjustment	0.09	0.18
Lack of parental emotional adjustment	0.11	0.26*
Poor family relationships	0.07	0.15
Poor parental teamwork	0.05	−0.04

Note: Child emotion dysregulation, parent mental health, parenting, and family adjustment variables are negatively valenced, and mindful parenting is positively valenced.

^* p < 0.05.

^** p < 0.01.

^*** p < 0.001.

^a Using regression analysis that controlled for baseline level, all significant correlational results remained significant except for the parenting non‐reactivity subdomain, which was not significantly related to changes in child reactivity and dysphoria (ps < 0.10).

Similarly, greater improvements in parenting behaviors measured using the PAFAS change scores were associated with reductions in child reactivity (95% CI [0.16, 0.55]) and child dysphoria (95% CI [0.02, 0.44]). More specifically, improvements in parental consistency were associated with reductions in child reactivity (r(74) = 0.40, p < 0.001, 95% CI [0.19, 0.57]). No other parenting subdomains were associated with changes in child dysregulation. In addition, changes in the family adjustment domain were not associated with changes in child dysregulation. However, improvements in parental emotional adjustment were associated with reductions in child dysphoria (95% CI [0.04, 0.46]).

3.3.3. Mediators of the Relationship Between Changes in Mindful Parenting and Child Dysregulation

As shown in Table 4, correlations among predictor (i.e., mindful parenting change scores), mediators (i.e., parenting behaviors and parent mental health change scores), and outcome variables (i.e., child reactivity and dysphoria change scores) were small to moderate in size. Changes in parenting behaviors were not associated with changes in parent mental health.

TABLE 4.

Bivariate correlations among predictor, mediators, and outcome variables.

	Mindful parenting	Parental mental health	Parenting behaviors	Child reactivity
Mindful parenting	—	—	—	—
Parental mental health	−0.39**	—	—	—
Parenting behaviors	−0.39**	0.19	—	—
Child reactivity	−0.30*	0.25*	0.37**	—
Child dysphoria	−0.33**	0.31**	0.24*	0.39**

^* p < 0.05.

^** p < 0.01.

As shown in Figure 1 (path c), the total direct effect of changes in mindful parenting was a significant predictor of changes in child reactivity, prior to entering the mediator variables, t = −2.58, p = 0.01, CI = −0.71 to −0.09, accounting for 8% variance. With the inclusion of the potential mediators, the overall model accounted for 18% of the variance in child reactivity change, F(3, 68) = 5.11, p = 0.003. As illustrated in Figure 2 (path b), change in parenting behaviors (t = 2.59, p = 0.01, ηp2 = 0.27) was an independent predictor of child reactivity change, while differences in parent mental health (t = 1.08, p = 0.28) were not. The multiple mediation results indicated a significant total indirect effect (point estimate = −0.24 CI = −0.45 to −0.04), accounted for by the indirect effect of parenting behaviors (point estimate = −0.16, CI = −0.38 to −0.02) and not parent mental health (point estimate = −0.07, CI = −0.20 to 0.05). Increases in mindful parenting were associated with better parenting behaviors (path a), which in turn was associated with decreases in child reactivity (path b). The direct association of changes in mindful parenting and changes in child reactivity was no longer significant, suggesting a full mediation.

FIGURE 1.

Parallel mediation analysis of the effect of parenting behaviors and parent mental health on the mindful parenting and child reactivity relationship. **p < 0.05, **p < 0.01, ***p < 0.001.

FIGURE 2.

Parallel mediation analysis of the effect of parenting behaviors and parent mental health on the mindful parenting and child dysphoria relationship. *p < 0.05, **p < 0.01, ***p < 0.001.

As illustrated in Figure 2 (path c), the total direct effect of changes in mindful parenting was a predictor of changes in child dysphoria, before entering the mediator variables, t = −3.02, p = 0.004, CI = −0.84 to −0.17, accounting for 11% of the variance. With the inclusion of the potential mediators, the overall model accounted for 18% of the variance in changes in child dysphoria, F (3, 68) = 5.18, p = 0.003. As shown in path b, neither changes in parenting behaviors nor in parent mental health were significantly associated with changes in child dysphoria, though parenting behaviors were trending toward significance (t = 1.75, p = 0.08). The multiple mediation results indicated a significant total indirect effect (point estimate = −0.24, CI = −0.42 to −0.08), accounted for by the indirect effect of changes in parenting behaviors (point estimate = −0.12, CI = −0.32 to −0.01) but not changes in parent mental health (point estimate = −0.12, CI = −0.27 to 0.02).

4. Discussion

To our knowledge, this study is one of the few to examine parent outcomes following parent participation in a community‐based cognitive behavioral intervention for autistic children. Despite previous studies in controlled settings reporting improvements in parent stress (Weiss, Viecili, and Bohr 2015), depression (Maughan and Weiss 2017), and anxiety (Reaven et al. 2015) following child‐focused CBT programs, the current study, which was implemented within community settings, did not. It is important to highlight that most parents in the current study were not experiencing high levels of mental health challenges, with only 1%–12% of parents at pre‐intervention falling into the “moderate,” “severe,” and “extremely severe” ranges of depression, anxiety, or stress. It is possible that with less severe levels of mental health problems, there is less opportunity to experience change. In fact, at pre‐intervention, the mean levels of mental health symptoms were in the average range (not reaching the threshold for the mild level of symptom severity) across the three DASS subscales, which is in contrary to what has been reported previously (Bitsika and Sharpley 2004; Hayes and Watson 2013; Lai et al. 2015). Our sample may have been different from other samples due to the time commitment required by parents to participate in the intervention, which may have made it difficult for parents with higher levels of mental health problems to participate. Another reason may be that this virtual intervention attracted parents of autistic children who benefited from online programming during the COVID‐19 pandemic (Lee et al. 2021). The COVID‐19 pandemic negatively impacted the mental health of children with neurodevelopmental disabilities and their families (Nonweiler et al. 2020). According to Charalampopoulou et al. (2022), autistic youth and their parents who experienced mental health deterioration during the pandemic also experienced high levels of material deprivation (e.g., loss of medical and academic services). It is possible that having access to this program during that time acted as a protective factor. Parents in our current study maintained their current level of mental health, which is in itself a positive outcome considering the number of studies indicating that the parental mental health showed a steep decline during the pandemic (Lee et al. 2021; Racine et al. 2021). Regardless, given the lack of meaningful change in DASS scores, we suggest any interpretation of the correlation between the parent mental health symptom change and child emotion regulation improvements should be tempered.

As expected, parents became more mindful in their parenting from pre‐ to post‐intervention. These results are similar to Maughan and Weiss' (2017) findings, which suggested that alongside the cognitive behavioral strategies, mindfulness exercises (e.g., breathing exercises, body scans) that are completed by the child during the session and at home might contribute to parents' increased observation of their child and reflections of their own parenting. In the current program, parents reviewed children's learned strategies including mindfulness activities, with the aim to support their child's generalization of strategies at home. The BMPS allowed us to examine the five domains (i.e., observing, describing, acting with awareness, non‐reactivity, and non‐judgment) that underly mindful parenting. The specific improvements in parents' non‐judgment with medium‐sized effect further support Maughan and Weiss' (2017) suggestion. We may see changes in mindful parenting by the sheer nature of asking parents at multiple time points about their parenting experience, which can prompt them to engage in deeper reflections on their parenting practices. Additionally, it is important to note that observing, describing, and acting with awareness might be best addressed during parent‐targeted mindfulness interventions (Bazzano et al. 2015), and it would be of interest to examine whether larger effects could be achieved if parents were actively informed about mindful parenting.

The current study also found improvements in overall parenting practices, including increased positive encouragement and decreases in coercive parenting. In this child‐targeted intervention, it is encouraging to observe changes in parenting practices similar to those observed in parenting interventions, specifically when measured by the same scale (Sumargi, Sofronoff, and Morawska 2015). Changes in family adjustment, including in family relationships and parental teamwork, were not observed from this intervention, which is reasonable given that there was only the participation of one parent and no relevant content about family relationships or coparenting.

As hypothesized, improvements in mindful parenting and parenting practices were associated with reductions in both aspects of children's emotion dysregulation. Improvements in parents' acting with awareness and non‐reactivity scores, and parenting consistency, were associated with greater reductions in child reactivity. This may be attributable to parents utilizing non‐reactive responses to children's emotionality, which may serve to break the pattern of escalating negative emotionality during parent–child interactions (Coatsworth et al. 2018; Duncan, Coatsworth, and Greenberg 2009). Improvements in child dysphoria were also associated with improvements in parents' non‐reactivity, observing, and describing of child emotional states, which is consistent with the literature on expressed emotion that highlights the association among parental criticism, hostility and emotional over‐involvement, and child internalizing problems (Hastings and Lloyd 2007).

This study did not find significant associations between changes in parenting variables and changes in children's social functioning. Given that all measures were parent‐reported, the fact that there were significant associations between parent outcomes and child emotion regulation, and not social functioning, supports the possibility that the observed changes were not merely a function of shared method variance. These results are in line with previous research that suggests that child emotion regulation skills are more proximally related to parent mental health and parenting behavior than child social skills due to the reciprocal relationship between parent and child emotion regulation and the parental role in supporting the development of children's emotion regulation (Gulsrud, Jahromi, and Kasari 2010; Ting and Weiss 2017).

Many studies have examined associations between mindful parenting with child functioning and behavioral challenges (Aydin 2022; Beer, Ward, and Moar 2013; Weiss et al. 2012; Zhang, Wang, and Ying 2019). These studies suggest a negative association, in which an increase in mindful parenting was associated with a decrease in child challenging behaviors/emotion dysregulation and vice versa. The current study is one of the few that tries to explain the relationship between mindful parenting and child well‐being. Findings support the notion that changes in parenting practices could mediate the relationship between changes in mindful parenting and child emotion regulation. Ren et al. (2021) found comparable findings, indicating that in neurotypical children, heightened maternal mindfulness was associated with an increase in child emotion regulation, and that positive parenting practices mediated this relationship. Moreover, in another neurotypical sample, Zhang, Wang, and Ying (2019) found that an increase in general mindfulness (i.e., focusing awareness on the present with non‐judgmental and receptive attention) was associated with higher levels of child emotion regulation and that mindful parenting mediated this relationship. Taken together, an increase in general mindfulness outside of the parenting context is associated with an increase in mindful parenting, which, in turn, is associated with positive parenting behaviors that are associated with decreases in emotion dysregulation.

The current study has several limitations. First, the non‐experimental methods do not allow causal inferences. Despite the parent–child relationship being bidirectional and likely influencing each other, future studies will require more sophisticated methods to inform the direction and causality of this relationship. Future studies can examine the observed results by conducting an RCT within a community‐based setting. Second, using change scores can pose some challenges in interpretation, such as shared measurement error and regression to the mean. We addressed some of the challenges by running regression analyses controlling for baseline levels that showed the same significant results of our correlational analyses, with the exception of the parent non‐reactivity subdomain of the Bangor Mindful Parenting Scale. Third, despite the original efforts of this study to represent a community sample, due to the COVID‐19 pandemic and subsequent lockdowns, the current sample only reflects certain demographic categories (i.e., majority White families) and those who had the resources and capacity to engage in a child‐focused, parent‐involved CBT intervention during this time. Similarly, the sample parents in this study had low levels of mental health difficulties at pre‐intervention. Future studies may conduct targeted recruitment of parents with higher or more prominent mental health challenges to examine whether the same benefits for these parents would be replicated and whether parent mental health would play a role in mediating the relationship between mindful parenting and child emotion dysregulation in these families. Third, as all measures were based on a single informant (i.e., the parent, primarily the mother), and emotion regulation findings are known to vary based on the informant (Weiss, Thomson, and Chan 2014), a multi‐modal assessment approach could enhance the validity of the results (Beck et al. 2020). Moreover, future studies could include general mindfulness outside of the parenting context since it influences mindful parenting. Finally, we did not collect data regarding parent participation in their own therapy or their participation in mindful activities. It is possible that some parents have shown increased mindful parenting or changes in their behaviors due to other factors.

5. Conclusion

Our findings suggest that parents can experience positive changes in their mindful parenting and parenting practices following participation in child‐targeted intervention for autistic children implemented in the community. These positive parent changes were associated with child positive changes in terms of their emotion regulation, with changes in parenting practices mediating the relationship between mindful parenting and child emotion regulation. These results provide mutually beneficial outcomes for autistic children, parents, community organizations, and have implications for public policy as well. Offering a child‐targeted intervention that can help both children and parents might address some of the intervention implementation barriers (e.g., high caseload, cost) faced in the community. It can also lend support to advocacy efforts to include evidence‐based interventions in the routine implementation of public health services.

Ethics Statement

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Consent

Informed consent was obtained from all individual participants included in the study.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author.