Introduction
Autism Spectrum Disorder (ASD) is neurodevelopmental syndrome with diverse phenotypes. Its main features are high social communication deficit and repetitive and restrictive behaviors [1]. Studies have reported that the prevalence of this disorder is increasing [2]. It affects all areas, especially social participation, sleeping, daily living activities, education and playing [2]. Children with ASD can play, although the game of their choice or its quality and duration may seem unusual to healthy children. Play-based therapeutic intervention is recommended for these children [2, 3]. Play is the best way to help a child with ASD. Parents of these children can approach their children through play [4, 5]. Studies show that mothers of children with ASD experience more stress than mothers with normal children [6]. It should be noted that it is not only the upbringing of a ASD child that creates a psychological burden for the family, but also the inability to establish mutual interactions, the child’s understanding, and communication between family members, which makes cause the family to face problems [7]. 
Mother’s mood affects the behavioral problems of ASD children and even their fathers’ mental health, and this can facilitate or disrupt the recovery process of these children [8, 9]. Therefore, it seems necessary to find an effective way to reduce the stress of mothers of ASD children and thus increase the child’s performance. In Iran, methods such as positive parenting program, parenting skills program, parental thinking program, resilience training program, and response-based methods have been used to investigate the effect of stress in mothers of these children [10-14]. These methods only examined the mothers and did not involve the children in the intervention. The best intervention method that can involve both mother and child is play therapy [1, 15]. Filial therapy is a family-centered play therapy that uses play as a primary tool to help parents and children communicate, work on their problems, and build a stronger relationship [16]. It was introduced by Bernard and Louise Guerney [23]. 
The ten-session filial therapy model developed by Gary Landreth [17] is one of the most widely used treatment protocols in this approach which is called Child Parent Relationship Therapy (CPRT). In this study, we used this protocol. Although there is evidence of high levels of stress in mothers of children with ASD, studies that have conducted in the field of ASD have focused more on the problems of ASD children and few interventions have been proposed to reduce maternal stress [18]. The aim of this study was to investigate the effect of filial therapy on reducing stress of mothers of children with ASD.
Materials & Methods
This is an experimental study with Pre-test/Post-test design using control group. The study population consists of all mothers of children with ASD aged 4-12 years referred to the occupational therapy and autism centers in Isfahan, Iran. First, the research objectives and method were explained to all mothers. Then, the mothers who volunteered to participate in the intervention signed a consent form, and then completed a demographic form, Gilliam Autism Rating Scale-2 (GARS-2), and Abidin’s Parenting Stress Index-Short Form (PSI-SF). Then they were randomly divided into two groups of intervention and control. The sample size was calculated 40 according to the Formula 1. Since filial therapy groups consist of 8 people, the sample size was determined 32. There were 4 groups of 8 assigned into two groups of intervention and control.

The PSI-SF has 36 items developed by Abidin [19]. It has three subscales of Parental Distress (PD), Parent-Child Dysfunctional Interaction (PCDI), and Difficult Child (DC). This questionnaire has acceptable reliability and validity to measure maternal stress [20]. The GARS-2 is a checklist designed by Gilliam (1994) to diagnose autism in the age group of 3-22 years. It has acceptable reliability and validity to measure autism. We used this tool to match children in the intervention and control groups.

Mothers participated in filial therapy sessions in groups of 8, each for 1-2 hours, once a week. After each session, there was a 30-min play session with the children; during this time, they followed what they had learned in the therapy sessions while playing with their child. Then, at the beginning of the next session, the therapist and the mothers reviewed the previous topics, answered the questions, reviewed the parenting booklets that should be completed for each session, and watched the parent-child play videos. At the end of the 10-week intervention period, the PSI-SF was completed again by the mothers in both groups. The mean scores of PSI-SF and its subscales before and after the intervention in each group were compared by using paired t-test and independent t-test. ANCOVA and Eta squared coefficient were used to determine the effect of intervention on the PSI-SF score. Data analysis was performed in SPSS by considering a significance level of P<0.05.
Results
Participants included 9 girls (5 in the intervention group and 4 in the control group) and 23 boys (11 in the intervention group and 12 in the control group). The subjects in the two groups were not significantly different in terms of mother’s occupation, mother’s age, number of children, child’s educational level, child’s GARS score, child’s gender and child’s age (Table 1). Shapiro-Wilk test was used to evaluate the normality of overall parenting stress data distribution in the Pre-test and Post-test phases. As shown in Table 2, the results of paired t-test showed no statistically significant difference between the two groups at baseline (P= 0.679) but a significant difference in the Post-test phase (P=0.010). The results of independent t-test showed that the difference in overall PSI-SF scores between Pre-test and Post-test phases in the control group was not significant (P=0.268) but there was a significant difference in the intervention group (P=0.001). Findings indicate that filial therapy caused significant changes in the intervention group compared to the control group.
The result of ANCOVA presented in Table 3 showed that the Pre-test score of parenting stress had an effect on the overall parenting stress after intervention, such that it could explain 82% of the total parenting stress variation. After controlling the Pre-test score, the mean overall parenting stress score became significantly different between the two groups (P=0.001). According to Eta squared value, the group filial therapy can explain 54% of the changes in the total parenting stress of mothers. Overall, it can be said that group filial therapy can reduce the parenting stress of mothers.
Discussion
The purpose of this study was to investigate the effect of filial therapy on parenting stress in mothers of children with ASD. Statistical findings showed that the mean difference in parenting stress between the two groups after the intervention was significantly different, and indicated that filial therapy can be effective in reducing their parenting stress. Few studies have been conducted in Iran using filial therapy; however, the interest in research in this field is increasing. For example, Mirzaei [22] compared the effects of filial therapy and therapist-child play therapy on the parenting stress of the mothers of children with attention deficit hyperactivity disorder. Both methods reduced the stress of mothers, but filial therapy was more effective. Although the study population was different, his results are consistent with our findings. In another study, Nejati et al. [23] investigated the effectiveness of CPRT on improving maternal parenting styles in hyperactive children. Samples were 30 mothers divided randomly into two groups of intervention and control. Findings showed that the mothers in the intervention group had a significant increase in using authoritative parenting style compared to the control group. Their study revealed that filial therapy has a positive effect on parenting styles of mothers. 
Soltani [24] examined the effect of child-parent play therapy on 20 mothers of children with autism in Mashhad. Play therapy increased mothers’ self-efficacy and self-confidence. In his study, mothers’ training sessions were conducted individually with their children. Play therapy should be conducted in group; however, positive results were reported which is consistent with our findings. Moqavam [25] studied the effectiveness of CPRT on the stress, anxiety, and depression of 30 mothers with a mild mental disability child. The results showed that filial therapy reduced stress in mothers which is consistent with our findings. Becklof [26] evaluated the effect of filial therapy on children with pervasive development disorder. Twenty-two parents of children aged 3-10 years with this disorder were divided into control and intervention groups. The intervention group received 10 CPRT training sessions (one 2-hour session per week) and 7 play therapy sessions with their children (one 30-minute session per week). The results showed that the intervention l group was significantly improved in accepting their child’s need for autonomy and independence as well as their child’s overall acceptance. This is in agreement with the results of present study.
Conclusion
Autism is a disorder with a complex nature and various signs and symptoms that sometimes it even makes it difficult for psychiatrists and specialists to diagnose. Therefore, finding the best and most effective treatment for these children is difficult and its costs are high. On the other hand, due to the special characteristics of these children (stereotyped movements, not starting to communicate with others or continuing it, sensory problems, etc.), family members and especially mothers due to closer proximity to the child, face with many problems such as: stress, depression, increased costs, lack of social understanding, no participation in social activities, and spending a lot of time educating the child. Most mothers think that their children and their development are the first priority and sometimes they forget about themselves and their mental and even physical health. This has a negative effect on the growth and treatment of children with autism. With a focus on CPRT, filial therapy is a technique without limitations that in a time process can move both mother and child on the path to treatment. It can reduce the burden of stress and unhappiness of mothers by creating more acceptance of the child and positively affect the developmental and therapeutic process of children, directly and indirectly. It has no side effects and all people with different cultures can benefit from this treatment protocol.
The present study, like any other studies, had inevitable limitations. For example, since the questionnaires were completed by parents, they may have evaluated the symptoms better or worse than the actual situation. Moreover, due to the label of autism, many mothers were afraid to participate in the group therapy and introduce their child, and were reluctant to express their real opinions. Further studies are suggested on the mothers of other exceptional children or examining the effect of filial therapy on the stress of the fathers of ASD children. There was no follow-up assessment in the present study. More studies with a follow-up are recommended to determine the time effect on the results of the intervention. The present study was related to mothers in Isfahan; further studies can be conducted in other cities of Iran.
Ethical Considerations
Compliance with ethical guidelines
This study was approved by the Ethics Committee of University of Social Welfare and Rehabilitation Sciences (Code: IR.USWR.REC.1396.97).
Funding
The present paper was extracted from the MSc. thesis of the first author, Department of Occupational Therapy, School of Rehabilitation, University of Social Welfare and Rehabilitation Sciences. 
Authors' contributions
Ideation and conceptualization: Houshang Mirzaei and Zahra Kiani; Research analysis: All authors; Review of sources: Hossein Sortiji, Zahra Kiani, Elahe Ebrahimi; Drafting, editing and finalization: Houshang Mirzaei, Hossein Sortiji, Zahra Kiani; Supervision: Houshang Mirzaei, Hossein Sortiji, Samaneh Hosseinzadeh; Management: Houshang Mirzaei.
Conflicts of interest
The authors declared no conflict of interest.
Acknowledgements
The authors would like to thank the staffs of Mani and Borna occupational therapy clinics as well as the autism centers Yavaran Zainab and Ordibehesht.