Abstract
The findings from this cross-sectional study have implications for how preschool children with ASD perform their self-care activities and suggest that more than half of preschool children with ASD have a need for occupational therapy interventions that target self-care skills.
Preschool age is a critical time for children to acquire self-care skills and to learn independence in self-care. Autism spectrum disorder (ASD) is a common lifelong developmental disability that can affect a person in many ways (American Psychiatric Association [APA], 2013). The current prevalence of ASD is approximately 1 in 54, but this rate is expected to increase (Maenner et al., 2020), with a concomitant impact on caregivers. Almost 50% of preschool children with ASD cannot perform self-care independently (Jasmin et al., 2009; World Health Organization, 2021), which may lead to difficulties participating in later school life and social activities, thereby substantially decreasing their well-being (Bal et al., 2015; Carothers & Taylor, 2004; Jasmin et al., 2009; Suprajitno & Arisky, 2019).
Moreover, caregivers of children with ASD have been found to have more pressure and burden than caregivers of typically developing (TD) children or children with other diagnoses (Dabrowska & Pisula, 2010; Lounds et al., 2007; Seltzer & Krauss, 1989; Suprajitno & Arisky, 2019). Lower levels of independence in self-care among children with ASD are associated with higher levels of stress and depression among their caregivers (Cavkaytar & Pollard, 2009; Dabrowska & Pisula, 2010). For example, lower levels of independence in self-care among adolescents with ASD were associated with higher levels of caregiver burden (Lin, 2011) and resulted in poor adult outcomes (Gray et al., 2014). Most people are first diagnosed with ASD at preschool age, so it has a lifelong impact. Bal et al. (2015) suggested that early intervention may be beneficial to development of self-care skills for young children with ASD. Thus, increasing the degree of independence in self-care among preschool children with ASD is important.
As the most basic concept of independence, self-care is fundamental to living in a social world (Shepherd, 2015). The acquisition of self-care skills usually begins at preschool age and matures through the development of performance factors, such as gross motor, fine motor, cognition, executive function, and social–emotional skills acquired during the early years of life (Dosman et al., 2012; Shepherd, 2015). In daily living, through continuous application, practice, and interaction with contexts, children can develop the ability to integrate these performance factors in a functional way to fulfill and care for themselves (Dosman et al., 2012). TD children at preschool age are expected to perform self-feeding, dressing, toileting, personal hygiene, and grooming independently with different proficiency and maturity levels as per their age groups (Shepherd, 2015). However, Bal et al. (2015) indicated that people with ASD showed slow attainment of daily living ability and that their self-care performance was often considerably below age-level expectations over time.
Self-care performance among preschool children with ASD is not fully understood and has been neglected in the literature for a long time. Previous studies have focused mainly on exploring the core symptoms and emotional behaviors that result in poor self-care performance among children with ASD, such as repetitive behaviors, restrictive interests, and preference for specific textures (Lockner et al., 2008; Nadon et al., 2011; Provost et al., 2010). These core symptoms could lead to difficulties in performing self-care activities (Lockner et al., 2008; Nadon et al., 2011; Provost et al., 2010). However, actual self-care performance and relevant skills among children with ASD remains unclear because an actual skills evaluation is lacking (Bal et al., 2015; Carothers & Taylor, 2004; Cavkaytar & Pollard, 2009).
Previous studies have also collected information about self-care performance among people with ASD mainly from the viewpoint of caregivers (Jasmin et al., 2009; Zobel-Lachiusa et al., 2015). The Vineland Adaptive Behavior Scales (VABS; Sparrow et al., 2005), a caregiver rating instrument, are commonly used to measure self-care skills of children with ASD. Previous studies that used the VABS concluded that children with ASD had relative strengths in their daily living skills compared with their social and communication skills (Bölte & Poustka, 2002; Carter et al., 1998). However, Jasmin et al. (2009) reported inconsistent findings in two self-care measures based on caregiver report. Almost 50% of preschool children with ASD showed a significant delay (<−2 SDs) of self-care functioning on the WeeFIM™, whereas the daily living skills standard score of the VABS was moderately low (between −1 and −2 SDs). Jasmin et al. suggested that observational tests of self-care skills should be added to validate the caregivers’ responses, indicating that caregiver rating scales might lead to biased results. Caregiver rating scales are unable to show the actual self-care performance of children (Mlinac & Feng, 2016). An observational measure is therefore needed alongside a caregiver rating scale to provide a full understanding of self-care performance among preschool children with ASD.
Thus far, research on the self-care performance of preschool children with ASD is limited in Taiwan. Existing studies rely heavily on information provided by caregivers. Considering the importance of independence in self-care among preschool children with ASD, a need exists to determine children’s actual self-care performance. In this study, we focused on detecting self-care performance from different viewpoints by using caregiver report and observational measures. This study addresses the following questions: Do preschool children with ASD have poor self-care performance compared with norms? Do the Assessment of Motor and Process Skills (AMPS; Fisher & Jones, 2012) scores correlate with the self-care normative standard scores on the Pediatric Evaluation of Disability Inventory (PEDI; Haley et al., 1992)? What is the agreement between the AMPS and the PEDI in determining poor self-care performance?
Method
Participants
We recruited a sample of 60 preschool children with ASD from a medical center in Tainan, Taiwan, using the following inclusion criteria: (1) diagnosed with ASD by registered pediatric psychiatrists according to the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; APA, 2013); (2) scores of ≥30 on the Standard Version of the Childhood Autism Rating Scale–Second Edition (CARS2–ST; Schopler et al., 2010); (3) ages 48 to 71 mo; and (4) no sensory impairments, such as visual or hearing impairments. The self-care data for these children with ASD have been reported previously (Chi & Lin, 2021). Table 1 presents the characteristics of the children with ASD. Most (83.3%) were boys with a mean age of 61.5 mo. Twenty children were ages 48 to 59 mo (33.3%), and 40 children were ages 60 to 71 mo (66.7%). The mean CARS2–ST score was 33.6, and the mean Full Scale IQ on the Wechsler Preschool and Primary Scale of Intelligence (Chen & Chen, 2013) was 84.5 (50.0% had an IQ between 86 and 127).
Sample Characteristics and Self-Care Scores of Children With ASD (N = 60)
Note. Age-matched norms are reported in the test manuals (Fisher & Jones, 2012, pp. 15–54; Tseng & Chen, 2012, p. 28). — = not applicable; ASD = autism spectrum disorder; AMPS = Assessment of Motor and Process Skills; CARS2–ST = Standard Version of the Childhood Autism Rating Scale–Second Edition; PEDI–C = Chinese version of the Pediatric Evaluation of Disability Inventory; WPPSI = Wechsler Preschool and Primary Scale of Intelligence.
Instruments
Standard Version of the Childhood Autism Rating Scale–Second Edition
The CARS2–ST is an observational rating scale that identifies children with ASD and categorizes the severity of autistic features (Schopler et al., 2010). The raw scores are summed from the 15 items for children with ASD ages ≥2 yr. Higher scores indicate more severe symptoms. Three descriptive levels include likely nonautistic (scores ranging from 15 to 29.5), mild to moderate autism (30–36.5), and severe autism (≥37; Schopler et al., 2010). The CARS2–ST has great internal consistency and good content and criterion validity (Schopler et al., 2010).
Assessment of Motor and Process Skills
The AMPS is a standardized observation-based measure that a trained and certified rater can use to objectively assess several activities of daily living (ADLs; Fisher & Jones, 2012). The AMPS, which is suitable for use with any child older than age 2 yr with any functional limitation, has normative data on various population groups and for different age groups (ranging from 3 to 103 yr); it can detect problems with the efficiency, safety, and quality of performance of ADLs. Each item is scored on a 4-point scale ranging from 1 (severely deficit performance) to 4 (competent performance). The scoring is divided into two parts: ADL motor skills and ADL process skills, which we calculated using the AMPS computer software. The AMPS logit scores can be compared with the normative range of ADL ability typically seen among healthy people of the same age (Fisher & Jones, 2012). Higher logits indicate better self-care performance. Merritt (2011) reported cutoff scores of 1.5 logits for motor skills and 1 logit for process skills for determining the need for assistance.
The AMPS showed good reliability on the Rasch equivalent of Cronbach’s α of motor skills (.92) and process skills (.91). The scale validity and person response validity demonstrated an acceptable goodness of fit to the many-faceted Rasch model (Fisher, 1993; Merritt, 2011). For preschool-age children, it encompasses some important occupational activities that involve taking care of one’s own body, such as self-feeding, bathing, dressing, personal hygiene, and grooming (Shepherd, 2015). Table 2 provides the selection tasks and levels of difficulty for AMPS motor and process skills.
Tasks and Levels of Difficulty for Motor and Process Skills
Note. Levels of difficulty are taken from the tables of the 110 standardized tasks from the Assessment of Motor and Process Skills manual (Fisher & Jones, 2012).
Insufficient data were available for stable calibration of logit scores for this task.
Pediatric Evaluation of Disability Inventory
The PEDI is a parent-reported inventory for evaluating self-care performance of children between ages 6 mo and 7.5 yr (Haley et al., 1992). The PEDI includes three areas: self-care, mobility, and social function. Each area includes three scales: the Functional Skills Scale, the Caregiver Assistance Scale, and the Environment Modified Scale. The Functional Skills Scale indicates the child’s capability to participate in daily activities. The Caregiver Assistance Scale measures the amount of assistance that children need to carry out functional activities. The normative standard scores have a mean of 50 (SD = 10). Higher scores on the Functional Skills Scale and Caregiver Assistance Scale indicate better performance and increased independence (Haley et al., 1992). In this study, only the self-care area was used to measure children’s capability and performance, using the normative standard scores of the Functional Skills Scale and the Caregiver Assistance Scale. The Chinese version of the PEDI (PEDI–C; Tseng & Chen, 2012) was used for this study. In terms of validity, the PEDI–C has good construct validity and content validity in assessing the self-care of children (Tseng & Chen, 2012).
Procedures
Recruitment information, including the introduction of this research and contact information, was delivered through flyers and posters posted on the bulletin boards of the medical center and early intervention centers. When participants and their main caregivers expressed their willingness to participate in this study, the primary investigator (Lin) contacted them via phone and email for further introduction and evaluation. The primary investigator gave a clear explanation of the study’s purpose and methodology to the participants and their main caregivers. All caregivers signed a consent form to show their agreement to enroll in this study. The primary investigator used the CARS2–ST to determine the level of autistic behaviors of the children with ASD and to confirm that participants met the inclusion criteria. A registered occupational therapist (I-Jou Chi) interviewed the caregivers using the PEDI–C. All participants were asked to select three activities from the provided six ADL tasks (see Table 2) and to bring their own daily tools, such as utensils and clothes, to demonstrate their self-care performance with the AMPS, which was assessed by an AMPS-certified occupational therapist (Ling-Yi Lin).
Data Analysis
In this study, we used IBM SPSS Statistics (Version 22) to analyze the descriptive statistics using correlation coefficients. The level of significance was set at p < .05. The logits of the AMPS and the normative standard scores of the PEDI–C were used. Pearson correlation coefficients were used to examine the relationships between the subscales of the two measures. Agreement between the two measures generated raw agreement percentages. A simple agreement of ≥80% was considered acceptable (Sattler, 2002).
Results
Table 1 presents the AMPS and PEDI–C data for the children with ASD according to age-matched normative data. Preschool children with ASD had a mean score of 1.3 logits (SD = 0.5) for the AMPS motor skills and 0.5 logits (SD = 0.7) for the AMPS process skills. Approximately 56.7% of the children with ASD scored below the cutoff score of 1.5 logits for motor skills, and 73.3% scored below the cutoff score of 1 logit for process skills. About 53.3% of children with ASD scored both below 1.5 logits for the AMPS motor skills and below 1 logit for the process skills, indicating difficulties performing ADL tasks. Fourteen children with ASD scored higher than average for age-matched normative data for both motor and process skills.
The mean PEDI–C Functional Skills Scale score for all children with ASD was 36.5 (SD = 20.3); 53.3% scored 1 SD below the mean. Only 15 children with ASD scored higher than average compared with the normative standard scores. The mean PEDI–C Caregiver Assistance Scale score for all children with ASD was 32.8 (SD = 17.1); 61.7% scored 1 SD below the mean. Only 8 children with ASD scored higher than average compared with the normative standard scores. The average PEDI–C self-care normative standard scores were moderately low (between −1 and −2 SDs), indicating poor self-care performance. However, a discrepancy between the children’s Functional Skills Scale and Caregiver Assistance Scale was observed. Ten children with ASD scored higher than average for the Functional Skills Scale normative standard scores but lower than average compared with the Caregiver Assistance Scale normative standard scores. Eight children with ASD who were considered to have poor performance on the Functional Skills Scale were considered to be in the normal range on the Caregiver Assistance Scale.
An examination of the correlations between the AMPS and PEDI–C showed that AMPS motor skills were significantly correlated with PEDI–C Functional Skills Scale scores (r = .44, p < .001) and Caregiver Assistance Scale scores (r = .41, p = .001). AMPS process skills were also significantly correlated with PEDI–C Functional Skills Scale scores (r = .36, p = .004) and Caregiver Assistance Scale scores (r = .27, p = .038). However, the relationships between the two measures were low. Table 3 presents the correlations between the AMPS and the PEDI–C.
Correlations of Two Self-Care Measures
Note. AMPS = Assessment of Motor and Process Skills; PEDI–C = Chinese version of the Pediatric Evaluation of Disability Inventory.
p < .05.
p < .01.
p < .001.
Additionally, AMPS scores (with a cutoff score of 1.5 logits for motor skills and 1 logit for process skills) and PEDI–C Functional Skills Scale scores (−2 SDs below the mean) were used to define poor self-care performance. Overall, the results for the 36 children with ASD were consistent with AMPS and PEDI–C scores; however, those for 24 children (40.0%) were discrepant. Among these 24 children with ASD, 21 scored at the level indicating age-appropriate performance on the PEDI–C but had poor performance scores on the AMPS. These 21 children with ASD had a mean score of 0.25 logits (SD = 0.52; range = −1.2 to 0.9) for AMPS process skills, with all children’s scores falling below 1 logit. Their mean PEDI–C Functional Skills Scale score was 46.5 (SD = 11.4). However, their mean PEDI–C Caregiver Assistance Scale score was 34.3 (SD = 15.9). Five children with ASD scored extremely low (<10) on the Caregiver Assistance Scale, indicating the need for maximal assistance.
Discussion
This study expands the knowledge on the self-care performance of Taiwanese preschool children with ASD. Our findings indicate that more than half of preschool children with ASD had poor self-care performance and needed assistance. The correlations between the AMPS and PEDI–C were low. In addition, 40% of participants yielded discrepant results on the AMPS and PEDI–C.
Compared with norms, preschool children with ASD have poor self-care performance. These results were consistent with the findings of several previous studies, which reported self-care dependence among children with ASD (Carothers & Taylor, 2004; Cavkaytar & Pollard, 2009; Franchini et al., 2018; Jasmin et al., 2009; Nadon et al., 2011). For instance, Jasmin et al. (2009) reported that preschool children with ASD had low scores (<−2 SDs) for self-care on the WeeFIM and for daily living skills on the VABS (between −1 and −2 SDs). The majority of those children had cognitive delays. In this study, only 8 children (13.3%) had a significant cognitive delay. The results indicate that the self-care performance of preschool children with ASD was poor compared with norms, suggesting that preschool children with ASD were significantly less independent and needed more assistance than their TD peers when performing self-care activities. Notably, previous research indicated that young children with ASD have relatively strong self-care skills compared with other developmental skills (Bölte & Poustka, 2002; Carter et al., 1998); however, many children with ASD had difficulties performing self-care tasks. The results show that poor self-care performance among preschool children with ASD should not be overlooked even though these children had average intellectual ability.
Consistent with previous findings, children with ASD had poor motor skills performance, including postural stability, coordination, and gross and fine motor skills (Jasmin et al., 2009; Liu & Breslin, 2013; Travers et al., 2017). Motor skills were related to daily living skills among young children with ASD (MacDonald et al., 2013) and were predictive of daily living skills later in life (Travers et al., 2017). However, there remain very few interventions targeting motor skills or daily living skills as a priority for young children with ASD (MacDonald et al., 2013). Regarding process skills, the performance of children with ASD was mainly influenced by skills associated with executive function (especially planning), sensorimotor skills, and behavior adaptation, which were reported as problems among children with ASD (Gardiner et al., 2017; Jasmin et al., 2009). A previous study indicated that executive function was a key moderator in the development of self-care independence (Tarazi et al., 2007). Thus, motor and process skills might lead to problems with performing self-care among children with ASD. This result shows that ADL motor and process skills should be strengthened in early intervention for preschool children with ASD.
The correlations between the two self-care measures, AMPS and PEDI–C, were low in this study, indicating a difference between the nature of the activities assessed by the observational instrument and the parent-reported measure. A review of the two measures separately showed that the PEDI–C provided a comprehensive picture of a child’s functional profile from the caregiver’s perspective to understand the relation between acquisition of self-care skills and the need for assistance (Tseng & Chen, 2012). Researchers and clinicians may consider other factors that may affect performance of self-care tasks (Haley et al., 2010). However, the PEDI mainly uses semistructured interviews to gather information on children’s self-care performances from the recall of caregivers. The uncertainty caused by recall in interviews may affect the accuracy of the information obtained about the children’s self-care performances (Haley et al., 2010).
Regarding the AMPS, only certified therapists can directly observe children’s performance, and the results can only be generated with a specific scoring software package (Fisher & Jones, 2012). The information collection process of the AMPS is rigorous to ensure that it reflects actual self-care performance. Moreover, the AMPS provides an objective viewpoint on children’s performance and excludes some factors (e.g., unnecessary assistance, unfamiliar tools) that may interfere with the evaluation. However, the skill profiles obtained with the AMPS mainly focus on performance skills that are small units of engagement in daily life occupations (American Occupational Therapy Association, 2020) rather than a broader functional profile. In addition, the AMPS is not suitable for use with people who are less motivated to participate in daily life tasks. To sum up, clinicians should not use PEDI scores to represent ADL performance; the PEDI cannot perfectly reflect ADL performance in real life and vice versa. Although the AMPS outlines children’s strengths and weaknesses in the context of real-life tasks, caregivers’ concerns about their children’s ADL performance cannot be captured solely with the AMPS.
Low agreement between the AMPS and PEDI–C in determining poor self-care performance was also indicated. In this study, 40.0% of children had discrepant results. All had a low score on the AMPS process skills. Merritt (2011) indicated that ADL process skills are the most accurate predictor of the need for assistance. If a child had a score below 1 logit, they would need assistance to carry out functional activities. From the caregivers’ perspective, these children had age-appropriate performance on the Functional Skills Scale but required certain levels of assistance. Their mean PEDI–C Caregiver Assistance Scale scores were moderately low (between −1 and −2 SDs). This result suggests that these children needed assistance to carry out functional activities.
Notably, a discrepancy between scores on the PEDI–C Functional Skills Scale and Caregiver Assistance Scale was observed. Haley et al. (2010) indicated that different parent expectations and different parenting experiences in valued and important functional activities may present challenges when applying the PEDI. Fournier et al. (2010) also reported that the child’s age has a direct effect on caregivers’ perceptions of the need to provide care. Caregivers may underestimate their child’s ability and may not know how to respond helpfully or promptly (Couchenour & Chrisman, 2016). The other possible reason might relate to accommodating the rigid behaviors and emotional problems of children with ASD (Schaaf et al., 2011). Many caregivers reported that excessive planning and structured routines were effective strategies to meet the needs of the child with ASD, especially with the time constraints they were under (Schaaf et al., 2011). Caregivers may have assumed that their child had the ability to perform self-care tasks; however, they provided much assistance to prevent any meltdowns from getting out of hand. Thus, a parenting education program addressing children’s self-care development could be offered to caregivers of preschool children with ASD.
Limitations
This research has some limitations. First, the participants enrolled in this study were mainly recruited from one medical center and early intervention centers in Taiwan. Additionally, the CARS2–ST category of preschool children with ASD used in this study was only mild to moderate and did not include the whole spectrum. The standardized procedure of the AMPS was modified by selecting six ADL tasks. These activities may not have provided the child with a sufficient level of challenge. Caution should thus be taken when generalizing the findings to preschool children with ASD. Another limitation is that the results of this study only represent the children’s current situation. Preschool age is a phase in which both development and performance fluctuate. The findings of this study provide little evidence to support the self-care developmental trajectory. Despite these limitations, this study established the self-care performance of preschool children with ASD. These results will help researchers understand more about preschool children with ASD.
Implications for Occupational Therapy Practice
The findings of this study have the following implications for pediatric occupational therapy practice: Preschool children with ASD had poor self-care performance compared with norms; thus, early intervention should target strengthening the motor and process skills related to self-care activities for preschool children with ASD. Both the AMPS and PEDI–C provided valuable information from different perspectives when assessing the self-care skills of preschool children with ASD.
Conclusion
On average, preschool children with ASD had difficulties performing self-care tasks. The poor self-care performance was related to their motor and process skills. Occupational therapy practitioners could focus on strengthening the motor and process skills of preschool children with ASD to aid in their performance of self-care tasks. Moreover, we have demonstrated that in addition to using self-report measures, an observational measure may provide valuable information on children’s strengths and weaknesses in performing real-life, self-care tasks.
Footnotes
Acknowledgments
We thank the children and parents who participated in this study.
