Predictors of Participation Difficulties in Autistic Children

Home Life

As with all of the outcome variables in the current study, home life is not clearly defined in the data set we used from the 2011 Survey of Pathways to Diagnosis and Services that was conducted by the Centers for Disease Control and Prevention (hereinafter referred to as “Pathways”; Child and Adolescent Health Measurement Initiative [CAHMI], 2015). However, the literature generally defines the term to include interactions with family members, family mealtime, meal preparation, care of other family members, self-care, chores, homework, personal care, cleanup, and care of animals or pets (Anaby et al., 2013). Home life spans several occupations from the Occupational Therapy Practice Framework: Domain and Process (4th ed.; OTPF–4; AOTA, 2020), including activities of daily living (ADLs), instrumental activities of daily living, education, and social participation. Autistic children participate less in home life than children with no diagnosis, particularly in the areas of personal care management, household chores, and school preparation (McCollum et al., 2016), and difficulties with home life continue into adulthood (Smith et al., 2012). Studies have identified lack of skills in daily living, communication, and socialization as major barriers to home life participation (Liss et al., 2001; Perry et al., 2009).

Classroom Learning

Previously identified areas of concern regarding classroom learning include behavioral and emotional differences, school performance, social activities, communication, following rules and directions, personal care awareness, task completion, positive social interactions, and safety (Sparapani et al., 2016). Classroom learning is part of the occupation of education in the OTPF–4. Many areas of concern are related to executive function, a frequent area of difficulty for autistic individuals (Demetriou et al., 2018). Sparapani et al. (2016) found that autistic children spent less time emotionally regulated and actively engaged in the classroom than other children. Autistic children experience behavioral and emotional differences from their peers without a diagnosis that are potential predictors of classroom learning, including shorter attention span, anxiety, depression, and behaviors that can be disruptive in the classroom (Ashburner et al., 2010). Aspects of the classroom environment contribute to these behaviors and may be potential targets for intervention (Martin, 2016).

Friendships

The friendships of autistic children without ID differ from those of children without an ASD diagnosis in self-, peer-, and teacher-rated quality and reciprocity (Kasari et al., 2011). Friendships are part of the occupation of social participation in the OTPF–4. In a study examining the friendships of young adults with ASD only, significantly more were likely to never see friends, to never be called by friends, to never be invited to activities, and to be socially isolated compared with young adults with other diagnoses who received special education services (Orsmond et al., 2013).

Leisure Activities

Leisure is defined as intrinsically motivated nonobligatory activity that is engaged in during time not committed to obligatory occupations such as work, self-care, or sleep (AOTA, 2020; Parham & Fazio, 1997). This definition is consistent with that of the occupation of leisure in the OTPF–4. Recreation, a category of leisure, consists of an individual’s preferred pleasurable and enjoyable activities engaged in during leisure time (Veal, 1992) and includes both physically active and solitary sedentary activities (e.g., watching television and videos; playing video games; using computers, cellphones, and other electronic devices). Although many studies have found lower levels of physically active recreational activity in autistic children (Askari et al., 2015; Hilton et al., 2008; Potvin et al., 2013), Ratcliff et al. (2018) found no significant difference in recreation participation between children with and without ASD, but they did not differentiate between physically active and solitary sedentary leisure.

Participants

We examined data from the 2011 Survey of Pathways to Diagnosis and Services data set (CAHMI, 2015). This survey is part of a repeated cross-sectional survey of parents and caregivers of children with special health care needs who have ever had a diagnosis of ASD, ID, or developmental delay from all 50 U.S. states; data were weighted to represent the U.S. population. Survey respondents were 6,090 parents or caregivers of children ages 6 to 17 yr, and phone interviews were conducted with 4,032 respondents. We excluded 1,112 respondents who did not complete questions about home life, friendships, classroom learning, and leisure activities. We analyzed data from reports having no missing observations (N = 2,920) for children with ASD only (n = 834), ASD and ID (n = 227), ID only (n = 508), developmental disorder only (n = 995), and no current diagnosis (n = 356; see Appendix Figure A.1). Missing responses for 236 children were imputed in the data analyses. This study did not require approval or oversight from the institutional review board at the University of Texas Medical Branch (No. 18–044) because we used publicly available, deidentified data.

Study Variables

The Pathways data include 25 social–behavioral items categorized into five social–behavioral areas: emotional regulation problems, conduct problems, hyperactivity/inattention, social problems, and prosocial behavior (see Appendix Table A.1). Higher scores indicate greater social behavioral difficulties. We also examined four sensory domains—sensory seeking, sensory avoidant, low registration/disorganized/overreactive, and repetitive behaviors—using the 15 Pathways sensory items (Lee et al., 2019). Lee and colleagues (2019) calibrated the sensory items with item response theory models and found good psychometric properties, including structural and known group validity. Higher scores represent greater sensory differences.

ADL, skill-related, and behavioral variables from the Pathways data set included ADLs (eating, dressing, bathing), fine motor difficulty (using hands), learning (understanding, paying attention), anxiety or depression, and behavior problems (acting out, fighting, bullying, arguing). These items were rated using a 3-point scale (1 = no difficulty, 2 = a little difficulty, 3 = a lot of difficulty). Higher scores represent more difficulties. Contextual variables consisted of parent education level, race or ethnicity, household income, and health insurance coverage and were included as the study covariates.

Outcome Variables

The outcome variables were home life (Item C27CA), friendships (Item C27CB), classroom learning (Item C27CC), and leisure activities (Item C27CD). For those four items, parents or caregivers were asked, “Do the difficulties interfere with your child’s everyday life?” Respondents used a 4-point rating scale (1 = not at all, 2 = only a little, 3 = quite a lot, 4 = a great deal), which we operationally dichotomized as 0 = not at all/only a little, 1 = quite a lot/a great deal to estimate odds ratios.

Data Analysis

We conducted population-weighted descriptive data analyses for the demographic characteristics using t tests for the continuous variable (age) and chi-square tests for the categorical variables (sex, race or ethnicity, parent education level, household income, health insurance coverage). Next, we performed four individual multivariate logistic regression models for the outcome variables and identified significant predictor variables. For the multivariate logistic regression analyses, we examined whether the study variables violated the multicollinearity assumption, and we checked the model fit. To control for the multilevel sampling structures of the complex survey database, we conducted subpopulation approaches using the entire sample with the sampling weight, strata, and cluster information; however, the primary analyses were produced using data solely for the autistic children with and without ID. For the multivariate logistic regression models, we used the SAS MI and MIANALYZE procedure modules to conduct the fully conditional specification as multiple imputations for the 236 missing observations in the study variables (Allison, 2002). In the individual multivariate regression models, we calculated the point estimations as adjusted odds ratios and 95% confidence intervals. We used SAS Version 9.4 for data management and all study analyses.

Sensory Processing, Emotional Regulation, and Behavioral Variables

We hypothesized that sensory processing, emotional regulation, and behavioral variables would predict participation difficulties in home life, classroom learning, friendships, and leisure activities for autistic children with and without ID. This hypothesis was supported by some of the variables in these categories.

Sensory processing, emotional regulation, and behavioral variables are interconnected and consistent with theoretical understandings of sensory processing; sensory information registered from the environment is processed in the brain, which directs emotional and behavioral responses (Blair & Diamond, 2008; Smith Roley et al., 2007). Management of emotional and behavioral responses is governed by executive function systems that work to control attentional processes, cognitive flexibility, goal setting, and information processing and enable individuals to control themselves physically, cognitively, and emotionally (Monteiro, 2021). These processes support a person’s ability to participate in home life, classroom learning, friendships, and leisure activities and may play a role as predictors of participation in these activities.

The role of sensory processing as an activity participation predictor was identified by Loh et al. (2021), who found that sensory processing was related to participation in childhood occupations, and by Choi and Jung (2021), who found sensory processing to be a predictor of leisure participation in early adolescents with no diagnosis. Werkman et al. (2020) examined the relationships among sensory processing, emotional and behavioral problems, and social participation and found that autistic individuals with higher cognitive abilities and sensory processing difficulties demonstrated more emotional and behavioral problems and that social participation was restricted by sensory processing difficulties regardless of cognitive abilities.

Other studies have linked emotional and behavioral responses to participation. In a study examining home life participation, Reynolds et al. (2011) found that fewer autistic children completed chores compared with unaffected peers and that sensory processing contributed to participation in both play and leisure activities and job or chore activities. Another study found that the social and cognitive demands of activities limited participation in home life (Egilson et al., 2018).

Sparapani et al.’s (2016) study supports our findings of the importance of emotional regulation and behavioral variables, finding less time spent being emotionally regulated and actively engaged in the classroom among the autistic children in comparison with the other children. Freeman et al. (2017) found that the autistic children who were rated by teachers as having poor initiation, working memory, and planning and organization skills spent more time engaging in solitary play when on the playground. Several studies found that lower emotional regulation and behavioral skills (executive function) contributed the most to lower participation in classroom learning (Kheirollahzadeh et al., 2021; Zingerevich & LaVesser, 2009). Sensory processing (Zingerevich & LaVesser, 2009) and motor proficiency (Kheirollahzadeh et al., 2021) were also identified as contributors to reduced participation in classroom learning.

The ability to manage emotional and behavioral responses has been shown to be important for friendship participation. Fong and Iarocci (2020) found the executive function skills of emotional regulation and behavioral regulation to be associated with social competence in a study of 77 autistic children. In another study, increased participation in organized activity and better friendship quality were demonstrated by youth with better emotional control (Bohnert et al., 2019).

Findings from the current study support those of previous research identifying sensory processing as a predictor of participation. Hochhauser and Engel-Yeger (2010) found lower leisure participation to be correlated with atypical sensory processing in autistic children, and Choi and Jung (2021) found sensory processing abilities to be predictive of participation in leisure activities for autistic adolescents. In a scoping review, Askari et al. (2015) identified a pattern of sensory, cognitive, and behavioral factors as a barrier to participation for autistic children.

Social Variables

We hypothesized that social problems and lower prosocial behaviors would predict participation difficulties in home life, classroom learning, friendships, and leisure activities for both groups. This hypothesis was supported for all outcomes except classroom learning.

Lower prosocial behaviors was a significant predictor of participation problems in home life and leisure activities for both groups and of friendships for the ASD-only group. Our finding regarding friendships supports the results of previous studies showing that successful peer relationships require social skills such as person perception, empathy, and prosocial behavior (Cillessen & Bellmore, 2011; Vetter et al., 2013). In other studies, social competence was shown to be important for obtaining and maintaining friendships (Flannery & Smith, 2017), and social impairment was found to be a mediator in the relationship between executive function and friendship quality (Lieb & Bohnert, 2017). Social problems, such as preferring to play alone, not being liked by other children, and being picked on or bullied, predicted friendship difficulties for both groups in our study. Salters et al. (2022) found that instructors reported social challenges as the largest barrier to participation, although they identified cognitive and fine motor skills as well. It is possible that parents in the Pathways study were not aware of the impact of social skills on difficulties in classroom learning or interpreted that outcome narrowly to focus only on school attendance and not on interactions with others.

Contextual Variables

We hypothesized that parent education level, household income, and health insurance coverage would predict participation difficulties in classroom learning, friendships, and leisure activities for autistic children with and without ID, but we did not expect race or ethnicity to be a predictor. This hypothesis was generally not supported.

Surprisingly, higher household income was a strong predictor of greater participation difficulties in home life and classroom learning for autistic children with and without ID. Poverty has been shown to slow cognitive and language development and is associated with lower academic achievement (Hernandez, 2011), and household income has been associated with autistic children’s physical and play activities (Memari et al., 2015). It is possible that, in our analysis, parents and caregivers with higher household income were more able to focus on their children’s needs and to identify areas of participation difficulty than those with income at or below the poverty level, who likely experienced more money-related distractions.

Having no health insurance coverage predicted greater participation problems for children with ASD only in classroom learning and for autistic children with ID in friendships. It is possible that the families with no health insurance felt they had inadequate access to services for their children and therefore more frequently identified participation problems. Only 2% of families in this data set did not have health insurance, so the data may not adequately represent the opinions of that group.

Contrary to our hypothesis, identification with a race or ethnicity other than non-Hispanic White was a significant predictor for classroom learning participation problems for autistic children with and without ID but was a protective factor for leisure activities. It is possible that non-Hispanic White parents have higher expectations for their children in the classroom and consider a wider range of activities as leisure than other parents. A previous study found that some Black parents were concerned about institutional racism in schools (Fields-Smith & Williams, 2009), which may have increased the observations of problems with classroom learning in this portion of our sample.

Parent education level was not a significant predictor of any of the participation areas for either group. This was surprising because higher income and highly educated parents are more likely to be involved in their children’s education, which is a key factor in educational success (Cabrera et al., 2018). It is possible that the parent perspective on their child’s difficulty in the data we used is not the same as actual school performance measured in other studies. It is also possible that the high representation of educated and higher income parents in this sample meant that too few lower educated (less than high school) and poverty-level income parents were surveyed to adequately represent the differences in lower education and lower income conditions.

We also hypothesized that sensory processing and emotional regulation variables would be stronger predictors of participation difficulties than contextual factors. This hypothesis was supported; sensory processing and emotional regulation factors were strong predictors in all participation areas for autistic children both with and without ID. Previous studies have identified emotional regulation (Pugliese et al., 2016) and sensory responsiveness (Hilton et al., 2010) as potential contributors to participation. Memari et al. (2015) identified financial burden and lack of opportunities as stronger barriers than specific autistic characteristics, but they examined only physical activity participation.

Activities of Daily Living and Functional and Adaptive Skills

We hypothesized that the group of skills consisting of poorer ADLs and functional and adaptive skills would predict participation difficulties in home life and classroom learning. This hypothesis was supported, although only fine motor skills were a protective factor for leisure participation among autistic children with ID and learning difficulties were a predictor for classroom learning.

Difficulty learning, understanding, and paying attention predicted classroom learning in children with ASD only. Fine motor skills, including handwriting, manipulation of materials, and scissor skills, are required in the classroom environment, and difficulty with fine motor skills curtails a child’s ability to perform these classroom tasks and thus their classroom learning.

We expected the ability to complete ADLs such as dressing and toileting to be a predictor of home life and classroom learning participation. Holloway et al. (2021) found that better gross motor skills were associated with greater participation in self-care, leisure, and social interactions. In a study of adolescents, physical demands were among the factors that parents identified as barriers to participation in home and school life (Lamash et al., 2020). Surprisingly, we found that fine motor skills and ADLs were not predictive of participation in home life or classroom learning. Fine motor difficulties were actually protective for leisure activity difficulties for the autistic children with ID. It is possible that, because parents completed the survey and did not often observe their children in the classroom, they may not have been aware of the impact of these skills; in addition, they may have been accustomed to managing ADLs at home and did not think of fine motor problems as a limiter in home life. In addition, the question asked whether using their hands interfered with the child’s classroom participation. Parents may have interpreted the question as referring to school attendance rather than measures of success such as good writing skills.

Many leisure activities do not require fine motor skills; research findings indicate that many autistic children and adolescents play computer and video games and watch TV, videos, and DVDs (Lamash et al., 2020; Must et al., 2014; Stiller & Mößle, 2018). The fine motor skills required for watching TV and videos or playing computer and video games may be less demanding than the fine motor skills required for ADLs and other home life activities.