Fewer children with autism spectrum disorder (ASD) with motor challenges receive physical and recreational therapies compared to standard therapies: A SPARK dataset analysis

Abstract

~85% of children with ASD from the SPARK study receive standard interventions such as speech-language (SLT) and occupational (OT) therapies. In contrast, only 32 to 13% of children with ASD received movement therapies such as physical and recreational therapies (PT and RT), respectively. Little is known about how service receipt changes as a function of children’s motor difficulties. Parents of children with ASD completed online surveys to provide a report of their child’s motor difficulties using the Developmental Coordination Disorder-Questionnaire (DCD-Q) as well as the various therapies their child received (SLT, OT, PT, ABA, SSI, RT) by location (at school, privately/in community, or both). While movement therapies (OT, PT, RT) increased with increasing motor risk and severity, there continued to be large disparities in PT (37-55%) and RT (15-19%) compared to OT (85-92%) across various settings indicating clear unmet needs for specific motor services given that 88.3% of this sample of children with ASD is at risk for motor difficulties / DCD. DCD-Q fine and gross motor subscale scores were fairly comparably affected yet PT/RT were less received versus OT. These findings explain the paucity of PT/RTs and physical activity programs for individuals with ASD and the need for movement experts to receive appropriate access and training to work with individuals with ASD.

Lay Summary

There are clear disparities in motor services provided to children with ASD (physical and recreational therapies) compared to their standard therapies (speech language and occupational therapies). Children with ASD need greater access to and funding for motor services (physical and recreational therapies) and physical activity (PA) programs. Movement experts including PTs, adapted physical educators, and community exercise / sports coaches / professionals need basic, specialized, and continuing education training to meet the needs of children and adults with ASD not only for providing early developmental and school-based services but also for ensuring appropriate built environments and providing general PT/adapted physical education services as well as PA programs.

Introduction

Autism spectrum disorder (ASD) has historically been considered a social communication disorder; but in the recent years has also been recognized as a multisystem disorder (Srinivasan & Bhat, 2013; Elsabbagh & Johnson, 2016). More recently, motor difficulties of children ASD have been highlighted in a large national study (i.e., the SPARK study) conducted by the Simons Foundation (Bhat, 2020a, 2020b, 2022, 2023; Bhat et al., 2022; Ketcheson et al., 2021). 87-88% of school-age children with ASD in the SPARK study were at risk for motor impairment using the Developmental Coordination Disorder Questionnaire (DCD-Q, Bhat, 2020a, 2020b; Ketcheson et al., 2021). However, only a small proportion of children in this sample were receiving movement therapies - 32% received Physical Therapy (PT) and 13% received Recreational Therapies (RT, Bhat, 2020a). Given that DCD-Q has ~85-92% positive predictive value against standard motor measures such as the Movement Assessment Battery for Children (M-ABC), a valid and reliable measure to determine motor service needs (van Damme et al., 2021; Miller et al., 2021; Green et al., 2009); it is estimated that ~75-79% (sample size x 0.88 x (0.85-0.9) of the SPARK sample is most likely in need of PT/RT services. Furthermore, motor skill development is known to have cascading effects on the overall social communication, cognitive, behavioral, and functional skill development of children (Iverson et al., 2023). Taken together, for children with ASD a lack of motor services / physical activity or a prioritization of standard of care, sedentary therapies (speech, behavioral, and occupational) will not only negatively impact their overall development and functional skill acquisition but also promotes long-term physical inactivity, motor dislike, and poor physical health/mental well-being and perhaps shortened life expectancy (Dawalt et al., 2019; Bhat, 2021, 2022; Weir et al., 2021; Dahlgren et al., 2022). In the current analysis, the proportion of various therapeutic services (motor and non-motor) provided to children with ASD in the SPARK sample were compared as a function of risk of motor delay (MI risk, yes/no risk of DCD) and severity of motor impairment (MI severity, 5 levels from very low to extremely high MI). Children with ASD often receive polytherapies (more than one service) given the complex symptomatology of ASD including Speech Language Therapy (SLT) for their communication difficulties and Occupational Therapy (OT) for their fine motor and sensory processing challenges (Srinivasan, Ekbladh, Freedman, & Bhat, 2021; Ames et al., 2021; Monz, Houghton, Law & Loss, 2019). In fact, SLT and OT are the most frequent school-based interventions for children with ASD (Srinivasan et al., 2021; Yingling & Bell, 2020) They also receive Applied Behavioral Analysis (ABA) / behavioral therapies throughout early and middle childhood and Social Skill Interventions (SSI) in middle to late childhood (Srinivasan et al., 2021; Becerra et al., 2017). Few studies have compared the receipt of various motor services (OT, PT, and RT) provided to children with ASD (Bhat, 2020a; Monz et al., 2019; Srinivasan et al., 2021; Zablotsky et al., 2015). Monz et al. (2019) found that 80% of children with ASD received either SLT or OT services and 52% received both whereas only 14% of children with ASD received PT or RT services. Children with ASD receive more school-based services (OT=41.7% vs. PT=18.6%) than community-based services (OT=11.4% vs. PT=6.4%) but in both settings, OT services are offered to a larger proportion of children than PT services (Zablotsky et al., 2015). These findings have been replicated by Becerra et al. (2017) who found that 55% of children with ASD received OT services whereas only 25% received PT services. Moreover, only few studies have compared receipt of OT and PT services to school/community-based RT (Bhat, 2020a; Monz et al., 2019) In a nationwide sample of children with ASD throughout the US, parents said that 80% received OT services, 32% received PT services, and only 13% received recreational therapies/activities (Bhat, 2020a).

Overall, there appears to be a clear disparity in motor services (PT and RT) targeting gross motor skills and physical activity for children with ASD, compared to fine motor / OT services. In this analysis, various therapeutic services offered to school-age children with ASD between 5 and 15 years from the SPARK sample were examined as a function of risk for motor impairment (MI risk) and as a function of increasing severity of motor difficulties (MI severity) based on total DCD-Q scores. It is hypothesized that a smaller proportion of children with ASD will receive motor (PT / RT) services compared to standard ASD services (SLT, OT, ABA) and while motor services will increase as a function of risk or severity of motor impairment there will continue to be a significant disparity in motor services (PT and RT) offered compared to standard therapies (OT, SLT, and ABA) across multiple settings (school, private/ community, or both). These findings also indicate that a large number of children with ASD are probably undertreated and have unmet motor needs due to a lack of access to gross motor/physical activity (PA)-based services (PT and RTs).

Methods

SPARK Study Procedures and Data Access

Families throughout the US with one or more children with ASD were recruited in the SPARK study through over 21 clinical sites across the US using a multi-method recruitment strategy including social media and fliers to national autism centers (Feliciano et al., 2018). Families voluntarily signed up for this study by completing the online questionnaires (https://sparkforautism.org/registration/account_information/) on the SPARK website. They also received information on studies in their nearby community to volunteer for local research studies. This author signed up with the SPARK study to utilize their study recruitment resources (i.e., SPARK Participant Match Resource) for different ongoing research studies approved by the University of Delaware (UD)’s Human Subjects Review Board. UD also signed an authorization agreement with the Simons Foundation; after which the author received access to version 3 of the SPARK study database (release date: February 2019).

SPARK Forms and Measures

The SPARK database comprises of multiple parent questionnaires such as the basic medical screening form, individual data form, and background history form. The basic medical screening form includes demographic information, birth history, professional diagnosis of ASD and other disorders, as well as other general medical conditions. The individual data form provides details on when the ASD diagnosis was made, which professional provided the diagnoses, whether there is a presence of a cognitive impairment, whether there is an Individualized Education Plan (IEP) for the child, and whether the child receives ASD services. The background history form lists the various intervention services received by the child in school, in private/community or both, as well as information regarding cognitive, language, and functional age level of each participant (i.e., above, at, slightly below, or significantly below same-age peers). Apart from these participant details, the author also analyzed data from the Developmental Coordination Disorder Questionnaire (DCD-Q, Schoemaker et al., 2006) and the Social Communication Questionnaire – Lifetime (SCQ, Berument et al., 1999) measure used to screen for motor and social communication delay in school-age children.

Table 1 shows the filters applied to the SPARK study sample to ensure that the children were included if they were between 5 and 15 years (i.e., DCD-Q validated for this age range) and held a professional ASD diagnosis or an individualized education plan (IEP) for ASD. Children were excluded if there were other neurological conditions unrelated to ASD leading to motor difficulties. Specifically, children with medical conditions/birth injuries (i.e., brain and spinal cord malformations, prenatal alcohol/drug exposure, and history of brain bleed) were excluded. Additionally, participants who met the SCQ cut-off of score ≥ 12 were included as it is often used as a measure of social communication delays associated with ASD in large sample survey datasets (Lee et al., 2010; Daniels et al., 2012; Marvin et al., 2017). According to publisher guidelines, if certain DCD-Q items were not completed, the child did not receive a total DCD-Q score and their data were excluded. After applying these criteria, the final sample included is 13,528 children. Compared to the sample analyzed in Bhat, 2020a, the current study includes participants with cognitive impairment/intellectual disability as this analysis focuses on the motor and other services data in children with ASD.

Table 1:

Filters used to remove samples based on birth injury criteria and criteria associated with the social communication questionnaire (SCQ) and Developmental Coordination Disorder – Questionnaire (DCD-Q) data.

Criterion for Removing Samples	Number of Samples Removed	Number of Samples Retained
Original dataset		45,319
Age from background history form was blank	22,044	23,275
Age at evaluation was <24 months or >215 months	1,880	21,395
CNS malformations (brain, spinal cord)	110	21,285
Prenatal alcohol or drug exposure	181	21,104
Intraventricular hemorrhage	169	20,935
Delays or impairment due to brain injury, stroke, lead poisoning, FAS, HIV, radiation, hydrocephalus, brain tumor, drug effects, etc.	817	20,118
No professional diagnosis of ASD	0	20,118
Diagnosis age <18 months and no IEP	120	19,998
Assigned as invalid using SCQ validity column	8	19,990
SCQ final score was not assigned due to missing items	520	19,470
SCQ final score <12	2,096	18,022
DCD-Q data not available	4,491	13,531
DCD-Q age <60 months	3	13,528
Final dataset		13,528

Sample Demographics

The sample had approximately 80.4% males and 19.6% females. The majority of participants (78.3%) were White, 10.8% were multi-racial, and 4.9% were African American. Additionally, 17% of the sample identified as Hispanic. The sample was evenly distributed for annual household income. Many children held formal co-occurring diagnoses – specifically, 41.6% had ADHD, 17.5% had motor delay or DCD, 22.6% had learning disability, 15% had cognitive impairment, and 61.5% had language disorders. Similar to Bhat, 2020a, 88.3% of children with ASD in this sample were at risk for motor impairment based on their DCD-Q performance. As mentioned earlier, the DCD-Q has 85-92% positive predictive value against standard motor measures such as the M-ABC which supports a child’s referral for motor services (van Damme et al., 2021; Miller et al., 2021; Green et al., 2009). In the current SPARK sample, 88.3% children with ASD failed on the DCD-Q and if ~85-90% of those would have been correctly identified using a follow-up M-ABC assessment, it can be estimated that ~75-79% (Sample size x 0.88 x (0.85-0.9)) of the current SPARK sample is most likely in need of PT/RT services.

DCD-Q

The DCD-Q is a 15-item parent questionnaire used to assess a child’s gross- and fine-motor coordination during everyday functional/play skills within their natural environment (Schoemaker et al., 2006). Each motor skills is rated on a scale of 1 to 5 indicating performance “not at all” similar to the stated skill, 2 indicating performance “a bit like” the stated skill, 3 indicating performance “moderately similar” to the stated skill, 4 indicating performance “quite similar” to the stated skill, and 5 indicating performance “extremely similar” to the stated skill. Note that clinicians would consider an item score of 3 or below an important indicator of motor concerns. The questionnaire focuses on various motor skills such as ball skills (e.g., hitting or catching a ball), complex body coordination skills (e.g., jumping, running, etc.), fine motor skills (e.g., writing, cutting, etc.), and general motor control abilities (e.g., quickness, clumsiness, fatigability, etc.). These skills are categorized into three subscales: control during movement, fine motor coordination, and general coordination. The total final score is calculated as a sum of the individual subscale scores with higher scores indicating better motor performance. Definite motor impairment (MI) or suspect DCD (<10^th percentile) is determined based on the final score cutoffs which differ for different age groups. For example, these cutoffs include a score < 47 for children between 5 years to < 8 years, a score below 56 for children between 8 years to < 10 years, and a score < 58 for children between 10-15 years.

Motor Impairment (MI) Risk:

Based on these criteria, an assignment of risk for DCD (1 = Yes, 0 = No) is provided for each participant. A DCD diagnosis is typically confirmed with a follow-up, standardized motor assessment, and clinical judgment of a trained movement clinician. The risk for DCD / motor impairment was 88.3% for this sample; however, this paper focuses on how motor and other ASD services change as a function of MI risk and MI severity based on DCD-Q scores.

MI Severity:

The full sample was categorized based on level of motor impairment (MI) defined by using the total DCD-Q score mean (μ) and standard deviation (δ) as shown below.

• Very low MI = DCD-Q total score > μ+2δ

OR 60.8 ≥ DCD-Q score ≤ 75, N = 528, 3.9% of sample.

• Low MI = DCD-Q total score between μ+δ to μ+2δ

OR 48.7 ≥ DCD-Q score < 60.8, N = 1,838, 13.6% of sample.

• High MI = DCD-Q total score between μ to μ+δ

OR 36.5 ≥ DCD-Q score < 48.7, N = 3,791, 28.0% of sample.

• Very high MI = DCD-Q total score between μ-δ to μ

OR 24.4 ≥ DCD-Q score < 36.5, N = 5,026, 37.2% of sample

• Extremely high MI = DCD-Q total score < μ-δ

OR 15 ≥ DCD-Q score < 24.4, N = 2,345, 17.3% of sample

It should be noted that 82.5% of the entire sample had high to extremely high risk for MI and these children would most likely be the subgroup needing PT/RT services. Note that 82.5% is close to the previously estimated proportion of 75-79% children who may become eligible for motor services upon receiving a follow-up, standardized motor assessment.

Gross, Fine, or General Motor Risk using Normalized DCD-Q subscale scores:

To compare gross motor (CDM or gross motor), fine motor (FM or fine motor), and general coordination (GC or general motor) risk of delay across the 5 aforementioned subgroups, each of the 3 subscale DCD-Q scores were divided by the number of items per subscale to obtain a normalized score that ranged from 1-5 for each DCD-Q subscale (Bhat, 2022) with 1 indicating low motor performance, 3 indicating moderately low performance, and 5 indicating typical/skilled motor performance. A score of 3 or below was considered a substantial domain-specific motor risk (gross, fine, or general motor risk).

Percent Services Received:

The number and percent of children receiving a particular service (motor: occupational therapy (OT), PT, RT and non-motor: Speech, Behavioral, and SST) was calculated as a function of MI risk and MI severity. Additionally, parents were asked where their child received the service (at school, privately/in community, or both); however, RT services were only tracked in the community and not at school.

Statistical Analyses

The primary outcome of this analysis was percent of services received for each service type (motor: OT, PT, RT and non-motor: Speech, Behavioral, and SST). Percent services received were examined as a function of MI risk based on DCD-Q score (Yes or No DCD risk), and MI severity subgroups (“very low” to “low” to “high” to “very high” to “extremely high” MI). To study associations between motor risk / motor severity and services received, Spearman rank correlations were conducted due to the categorical nature of services data. Chi-square tests and related effect size measures using phi coefficients were used to compare service proportions across categories of MI risk (yes vs. no) and MI severity (5 subgroups). Odds ratios were the ratio of percent services received by the subgroup of interest ((i) Yes DCD/MI risk or (ii) 4 of the 5 subgroups based on MI severity) divided by the percent services received by the comparison subgroup (No DCD/MI risk or the least affected subgroup of Very Low MI severity). Wilcoxon signed rank tests were used to compare normalized DCD-Q subscale scores in children with MI risk or with substantial MI severity. Statistical significance was set at p<0.05; however, the majority of the p values are <0.0001 and are stated throughout.

Results

Proportion of services: Overall services and motor / non-motor services as a function of MI risk

As shown in Figure 1 and Table 2‘s full sample data, amongst motor services, 83.2% of the entire SPARK sample received OT, 34.8% received PT, and 14.7% received RT. Among non-motor services, 85.8% of the children with ASD received SLT, 67.8% received behavioral services based on ABA or developmental approaches, and 47.7% received social skill interventions or SSI. There was a general trend for greater proportion of children receiving services as a function of increasing MI risk and MI severity based on the small positive correlations between services and motor risk/impairment except for SSI and RT (Correlations in Table 2).

Figure 1:

Service provided to children with ASD. OT: Occupational Therapy, PT: Physical Therapy, Behav/Dev: Behavioral/Developmental, RT: Recreational Therapies.

Table 2:

Percent motor services received and ratio of PT/OT, RT/PT, PT/RT, OT=Occupational Therapy, PT=Physical Therapy, RT=Recreational therapies, SLT=Speech Language therapy. Behav./Dev.= Behavioral/Developmental, SSI=Social Skill Interventions, RT=Recreational Therapy.

	Motor services received [%]			Ratio of % services received			Non-motor services [%]
	OT	PT	RT	PT/OT	RT/OT	RT/PT	Speech	Behav./Dev.	SSI
Full Sample	83.2%	34.8%	14.7%	0.42	0.18	0.42	85.8%	67.8%	47.7%
Motor risk No	72.1%	17.9%	10%	0.25	0.14	0.56	84.4%	64.6%	46.8%
Motor Risk Yes	84.7%	37.1%	15.3%	0.44	0.18	0.41	86%	68.2%	47.8%
Very low	68%	17.2%	11.6%	0.25	0.17	0.67	80.3%	64.2%	57.6%
Low	73.2%	21%	10.3%	0.29	0.14	0.49	82.4%	65.3%	53.4%
High	80.8%	27.5%	13.4%	0.34	0.17	0.49	84.3%	66%	50.8%
Very High	85.9%	38.1%	15.5%	0.44	0.18	0.41	86.5%	68.2%	46.1%
Extremely High	92.4%	54.7%	19.3%	0.59	0.21	0.55	90.4%	72.6%	39.4%

Table 3 states the chi-square, p, and phi coefficient (effect size measure) values for calculations comparing services as a function of motor risk. As shown in Figure 2, motor services increased as a function of MI risk (chi-square test p-values ranged between <0.0001 and 0.0043, Table 3). Children with ASD with greater MI risk received more motor services including OT (DCD=84.7% vs. no DCD=72.1%, p<0.0001, odds ratio = 2.14, Table 4), PT (DCD=37.1% vs. no DCD=17.9%, p<0.0001, odds ratio = 2.7), and RT (DCD=15.3% vs. no DCD=10%, p<0.0001, odds ratio = 1.6 Table 4) services. To be clear, children with greater MI risk (Yes DCD) received OT 2.14 times more, PT 2.7 times more, and RT 1.6 times more than children less/no MI risk (i.e., No DCD). Note that the effect sizes for receiving OT and PT services as a function of motor risk was medium and all other services reported small effects based on phi coefficient values (Table 3). However, even then the proportion of children with ASD and MI risk receiving PT (37.1%) and RT (15.3%) services was much lower compared to SLT (85.9%) and OT (84.7%) services. In terms of ratio of services received, PT/OT was 0.44 (or 2 out of 5 children receiving OT also received PT) for the DCD-risk subgroup, RT/OT was 0.18 (1 out of 5 children receiving OT also received RT) (Table 2). For non-motor services, children with ASD with a risk of MI showed a trend for more behavioral services (p=0.05, odds ratio = 1.2, Table 4); however, the other non-motor services (SLT and SSI, chi-square test ps>0.05) did not differ significantly as a function of MI risk.

Table 3:

Spearman correlations, Chi-Square, and phi coefficient values for percent services received based on MI risk and MI levels (very low to extremely high MI). Statistical significance for correlations was set at p < 0.05, however, the maximum p-value is 0.013. MI levels are based on DCD-Q scores such that higher DCD-Q scores indicate lower MI; hence, MI level and DCD-Q scores correlate in opposite directions with percent services received. Statistical significance for Chi-Square analysis was set at p < 0.05, however maximum p-value is 0.0043. Effect sizes are categorized as small (S, phi ≤ 0.1), medium (M, 0.1 > phi > 0.5), or large (L, phi ≥ 0.5). NS: Not significant. Significant correlations are listed and correlations ≥ 0.1 are highlighted. OT=Occupational Therapy, PT=Physical Therapy, RT=Recreational therapies, SLT=Speech Language therapy. Behav./Dev.= Behavioral/Developmental, SSI=Social Skill Interventions.

Received Therapy (Yes/No)	MI risk (yes/no DCD)	MI level (5 levels)	DCD-Q Total Score	DCD-Q CDM (Gross motor) Score	DCD-Q FM (Fine motor) Score	DCD-Q GC (General Motor) Score
Spearman Correlations
OT	0.109	0.169	−0.174	−0.159	−0.167	−0.103
PT	0.130	0.231	−0.243	−0.255	−0.160	−0.171
SLT	NS	0.075	−0.078	−0.089	−0.096	NS
Behav./Dev.	0.025	0.051	−0.050	−0.039	−0.057	−0.030
SSI	NS	−0.096	0.099	0.076	0.151	0.022
RT	0.048	0.075	−0.079	−0.080	−0.063	−0.050
Chi-Square Tests
OT	141.3	402.8
PT	253.4	753.4
SLT	NS	81.7
Behav./Dev.	8.2	39.3
SSI	NS	129.9
RT	34.6	81.1
Phi values for chi-square effect sizes
OT	0.10 (M)	0.17 (M)
PT	0.14 (M)	0.24 (M)
SLT	0.01 (S)	0.08 (S)
Behav./Dev.	0.02 (S)	0.05 (S)
SSI	0.01 (S)	0.10 (S)
RT	0.05 (S)	0.08 (S)

Figure 2:

Services received as a function of risk of motor impairment (MI)

Table 4:

Odds ratios for receiving a therapy service based on motor impairment (MI) risk and MI severity. Statistical significance for chi-square analysis was set at p < 0.05. NS: Not significant. Compared to no risk, odds of receiving more services were obtained for children with MI risk. Compared to very low MI severity subgroup, odds of receiving more services were obtained for the remaining 4 MI severity subgroups.

	Speech	OT	PT	Behav. / Dev.	SSI	RT
DCD / Motor risk
Yes vs. No	1.13	*2.14	*2.72	*1.17	1.04	*1.63
MI level
Low      vs. Very Low	1.15	*1.29	1.27	1.05	0.84	0.88
High      vs. Very Low	*1.31	*1.98	*1.82	1.08	*0.76	1.18
Very High    vs. Very Low	*1.58	*2.87	*2.95	1.20	*0.63	*1.41
Extremely High vs. Very Low	*2.30	*5.73	*5.80	*1.48	*0.48	*1.83

Proportion of services: Motor / non-motor services as a function of MI severity

Table 3 states the chi-square, p, and phi coefficient (for effect size) values for calculations comparing services as a function of MI severity. As shown in Figure 3, motor services increased as a function of MI severity. Children with ASD with extremely high MI (EHMI) clearly received more services compared to those who had very low MI (VLMI) (OT: EHMI=92.4% vs. VLMI=68%, p<0.0001, odds ratio (OR) =5.7), (PT: EHMI=54.7% vs. VLMI=17.2%, p<0.0001, odds ratio =5.8), and (RT: EHMI=19.3% vs. VLMI=11.6%, p<0.0001, odds ratio =1.83, Table 4). To be clear, compared to children with very low MI severity, children with EHMI received OT 5.7 times more, PT 5.8 times more, and RT 1.8 times more. Note that the effect sizes for receiving OT and PT services as a function of MI severity was medium and all other services reported small effects based on phi coefficient values (Table 3). Once again, the PT and RT services, for the EHMI group (OT=92.4%, PT= 54.7%, RT=19.3%) or the VHMI group (OT=85.9%, PT= 38.1%, RT=15.5%) were surprisingly lower compared to their OT services. In terms of ratio of services received, PT/OT ratio for the HMI to EHMI was 0.3-0.6, RT/OT for the HMI to EHMI was 0.17-0.2 (Table 2). In terms of non-motor services, children with EHMI severity received more SLT and behavioral services compared to those with VLMI severity (SLT: EHMI=90.4% vs. VLMI=80.3%, p<0.0001, odds ratio = 2.3), (Behavioral: EHMI=72.6% vs. VLMI=64.2%, p<0.0001, odds ratio = 1.48, Table 4). Note that SSI will reduce with increasing MI as it is often offered to children needing less support.

Figure 3:

Serviced received as a function of MI severity

Comparing gross, fine, and general motor performance as a function of MI risk / MI severity

Here two questions were asked: a) whether the normalized DCD-Q subscale scores differed between the FM, CDM, and GC subscales for the Yes motor risk/DCD subgroup (based on MI risk) or the more affected HMI, VHMI, and EHMI subgroups (based on MI severity) and b) whether the subscale differences FM vs. GM vs. GC parallel the service discrepancies seen between OT, PT and RT. Overall, Figure 4A & 4B as well as Table 5’s % sample column reveal that 82.5% of the sample is having high to extremely high MI. Based on Wilcoxon signed rank tests, the three subscale scores differ slightly between the FM, CDM, and GC subscales for Yes motor risk subgroup with highest performance on the CDM (Mean score =2.51±0.87) followed by FM (Mean score = 2.17±0.99) and lastly GC (Mean score = 2.00±0.69) (ps<0.0001 for all comparisons). Similar statistically significant differences were seen for the HMI, VHMI, and EHMI subgroups (Figure 4A, Table 5). Here too, notice that all three DCD-Q subdomains are affected in children with ASD with greater MI risk / severity with the normalized mean subscale scores being below 3 out of 5. Given the moderate or low fine motor and gross motor DCD-Q subscale scores one would expect similar levels of fine motor (OT) and gross motor (PT/RT) service receipt. Surprisingly, Figure 4B reveals stark differences between OT and PT/RT service receipt with OT services being disproportionately higher compared to PT/RT services.

Figure 4:

(A) Normalized DCD-Q subscale scores (FM, CDM, GC) for the very low, low, high, very high, and extremely high subgroups of children with ASD and MI. (B) Percent services received – OT, PT, RTs by the very low, low, high, very high, and extremely high subgroups of children with ASD and MI.

Table 5:

Mean and Standard deviation (SD) of normalized DCD-Q total and subdomain (CDM, FM, or GC) scores ranging from 1-5 with a mean of ~3 or below indicating a clear gross, fine or general motor coordination difficulty (note the means highlighted). 82.5% of the sample has clear gross, fine, and general motor difficulties. Fine motor service needs are met as 80-92% of the sample receive OT. However, gross motor service needs are not met with only 27-55% receiving PT, and only 13-19% receiving RT services. Note that CDM, FM, and GC subscale scores differ in the Yes MI risk and High, Very High, Extremely High subgroups for all three subscales: CDM score > FM score > GC score, ps<0.0001 based on Wilcoxon Signed Rank Tests.

MI Risk / MI level	N	% of sample	Normalized Total		Normalized CDM		Normalized FM		Normalized GC
			Mean	SD	Mean	SD	Mean	SD	Mean	SD
No MI Risk	1589	18.45	3.82	0.47	4.09	0.61	3.83	0.89	3.49	0.72
Yes MI Risk	11939	81.55	2.25	0.65	2.51	0.87	2.17	0.99	2.0	0.69
Very Low	528	3.9	4.35	0.25	4.55	0.39	4.39	0.60	4.07	0.53
Low	1838	13.6	3.56	0.22	3.91	0.54	3.56	0.85	3.14	0.62
High	3791	28	2.81	0.23	3.13	0.59	2.79	0.90	2.42	0.57
Very High	5026	37.2	2.03	0.23	2.30	0.51	1.90	0.70	1.83	0.48
Extremely High	2,345	17.3	1.38	0.17	1.48	0.34	1.26	0.35	1.34	0.34
Full sample	13528	100	2.43	0.81	2.70	0.98	2.36	1.12	2.18	0.84

Differences in service receipt by location in children with ASD with MI risk or with high to extremely high MI severity, subgroups needing motor services

Supplementary Tables S1–S3 show the breakdown of percent services provided in school only, sought in private/community setting only, or received in both settings for the full sample (Table S1) or children with MI risk (Yes DCD, Table S2) or the children with substantial MI severity (sample includes children in high, very high, and extremely high MI subgroups, Table S3). Visual inspection of data presented in Tables S1–S3 indicate that for services that are highly accessible to children with ASD (e.g. OT, Speech, Behavioral, and SSI), most children received these services in both settings (school and private), followed by school only, and lastly, they were sought in private/community settings only. Similar trends were also seen for PT services. Note that the SPARK study team assumed recreational therapies to be community-based and did not track recreational therapies in school settings (e.g., Adapted Physical Education); hence, those data are unavailable for the current sample. Nevertheless, once again this analysis confirmed disproportionately lower access to PT/RT services (PT=34.8-38%, RT=14.7-15.6%) compared to OT services (83.2-85.5%) for all subgroups examined.

Discussion

This study examined the proportion of motor and non-motor therapeutic services offered to children with ASD as a function of motor impairment (MI) risk and MI severity. Amongst motor services, 83.2% of children with ASD received OT services whereas only 34.8% received PT and 14.7% received RT. Amongst non-motor services, 85.8% of the SPARK sample received SLT, 67.8% received services based on ABA or developmental approaches, and 47.7% received social skill interventions. Motor services improved as a function of motor impairment risk. Children with ASD with a greater risk of motor impairment (DCD/motor risk = Yes) received more OT, PT, RT, and behavioral services compared to children with ASD without any motor difficulties (DCD/motor risk = No). However, the disparity between OT (84.7%) and PT (37.1%) / RT (15.3%) services continued to be high even for children with ASD with DCD/MI risk. Children with ASD with increasing severity of MI received more motor services (i.e., children with ASD with extremely high MI vs. those with very low MI, OT: 92.4% vs. 68%, PT: 54.7% vs. 17.2%, RT: 19.3% vs. 11.6%). However, children with ASD within each subgroup of varying levels of MI severity including those with severe motor difficulties continued to show disparity in service proportions between OT (81-92.4%), PT (27-54.7%), and RT (11.5-19.3%) services. While DCD-Q gross motor scores were slightly better than DCD-Q fine motor scores which were in turn better than DCD-Q general motor scores for all subgroups including children with ASD in the high to extremely high MI subgroups; all 3 normalized subscales scores were consistently below 3 on average indicating moderately low to low motor performance for high to extremely high MI subgroups. Therefore, in spite of substantial delays in all three motor subdomains, services addressing motor difficulties continue to be highly disparate in the ASD population indicating a significant gap in access to certain motor services such as PT and RT compared to OT. Last but not the least, service location differences revealed that majority of services are generally received in both schools and private/community settings, followed by school only, and some seek them in private/community settings only. However, the trend for disproportionately lower access to PT/RT services compared to OT services was observed across all locations/settings.

Fewer children with ASD receive PT and RT compared to SLT, OT and behavioral services

Children with ASD primarily receive SLT (86%), OT (85%), and behavioral (68%) services followed by social skill interventions (48%). In contrast, only 37% children with ASD received PT and 15% received RT services. While PT and RT service proportions improve with greater risk or severity of motor difficulties (i.e., at-risk for DCD based on DCD-Q) in children with ASD, the service proportions are still disparate, only 37-55% receive PT and only 15-19% received RT compared to the 85-92% receiving SLT and OT services. Given that 88.3% of children with ASD in the SPARK sample are at risk for motor impairment, they are currently undertreated for their motor difficulties and most likely have unmet motor needs for PT and RT services. Another large sample study of children with ASD found that 71% received SLT, 60% received OT, and 56% received behavioral services, whereas only 14% received PT and RT services (Monz et al., 2019). Moreover, the proportion of children with ASD receiving SLT, OT, PT, and behavioral services reduced significantly with age indicating that older children with ASD are not receiving as many motor services compared to the younger children (Srinivasan et al., 2021; Becerra et al., 2017). At the same time, motor difficulties in school-age children with ASD remain stable into adolescence and physical inactivity seems to further increase with age (Bhat, 2022; Dahlgren et al., 2021). The lack of motor services for children with ASD with persistent motor risk and severity as well as physical inactivity is counterintuitive and points to clear deficiencies in our current health service systems.

Disproportionate motor services seen across multiple settings in spite of moderately low to low fine and gross motor impairments

While motor skills (fine, gross, and general motor) were substantially and comparably affected in children with ASD with a motor/DCD risk and with greater motor impairment severity; the gap between OT, PT, and RTs was highly disproportionate and biased to OT (85%) compared to PT (37%) or RTs (15%). In addition, this disproportionate access to OT over PT/RT services was seen across school and private/community settings. Note that additional physical activity may be accessed in schools through Adapted Physical Education; however, this was not tracked in the SPARK sample. Monz et al (2019) has reported similar service discrepancies in the past literature with 60% receiving OT services but only 14% receiving PT/RT services in the community. Together, these findings provide proof of service gaps for PT and RT services for children with ASD. We should find ways to address motor gaps in children with ASD because motor skills are a foundation for other forms of development such as social communication, cognitive, and behavioral skill development. Through movement skill acquisition, children create opportunities for exploration and interactions with the world and their caregivers who in turn facilitate social communication and cognitive skill development (Iverson et al., 2023). Poor motor skills and a lack of access to motor services directly affects a child’s functional independence and over time leads to motor dislike and physical inactivity as well as lost opportunities for peer social interactions. In the long-term, high levels of physical inactivity in individuals with ASD will contribute to their poor physical health outcomes such as greater risk for cardiovascular disease and diabetes, as well as social isolation/depression, an overall poor quality of life, and perhaps lower life expectancy (Srinivasan et al., 2014; Weir et al., 2022; Dawalt et al., 2019). These findings call for caregivers and clinicians to advocate for greater motor / physical activity opportunities for individuals with ASD throughout their lifespan.

Barriers to accessing PT services and what can be gleaned from PT models in other countries?

Only one study has explored the barriers to accessing PT services in young and older children or adolescents with ASD (Campos et al., 2019). Campos et al. (2019) conducted semi-structured interviews of 10 hospital-based, Canadian physical therapists (PTs) who were asked about their past and current experiences working with children with ASD (including knowledge of ASD strategies, resources, and education) as well as their thoughts on the role of PTs in treating children with ASD. Note that the participating clinicians / PTs worked at a rehabilitation hospital and not in early intervention or school-based settings, which is where majority of the motor services are offered to children with ASD in US and Canada. The clinicians felt that PTs should play a consultative role of supporting parents, teachers, and other clinicians / exercise professionals on developing movement therapy plans and recreational programs for daily PA. They were also open to connecting families to community-based resources that provided regular PA programs but did not want to be the primary provider of movement/physical activity. Campos et al. (2019) also found that while the participating PTs were supportive of exploring new roles for PT practice, they expressed the lack of training, expertise, and confidence in managing behavioral/sensory issues and in implementing ASD treatment strategies as the main reason for not working with children with ASD. Lastly, the PTs also expressed that they had received little to no training on how to work with children with ASD within their formal PT education.

In terms of systemic barriers, the clinicians felt that because ASD is often considered a psychiatric and not a physical diagnosis, PT is generally not recommended to individuals with ASD, institutional staffing/resources are diverted to other neurological, medical, or surgical diagnoses (pediatric/adult), and systematically service delivery models do not include PTs as part of the management team for children with ASD (Campos et al., 2019). Some PTs felt that addressing gross motor issues of children with ASD did not fit within the traditional role of pediatric PTs. In addition, families may prioritize communication, behavioral, and fine motor therapies over gross motor therapies / physical activity making it more difficult for children to work with PTs for an extended period. Some PTs also expressed concerns about adding to caregiver burden by adding another service for the child. Overall, many individual and systemic barriers as well as traditional views of ASD and PT services seemed to prevent access to PT / motor services for children with ASD. US and Canadian service delivery systems seem to be fairly similar as this analysis of the SPARK sample shows that few children with ASD receive PT services in the US.

PTs in US can learn from service delivery models in other countries such as Israel and Belgium, where PTs play a prominent role in the care of children with ASD from early intervention to school-based settings to community-based, private practice settings (Atun-Einy et al., 2013; Van Damme & Bhat, 2023). The Israeli PT service delivery model proposed by Atun-Einy et al. (2013) focuses on a dynamical systems approach addressing individual, environment, and task-related needs of the child with the goal of achieving functional independence and promoting child’s strengths and interests. PTs are responsible for administering motor assessments, providing movement / PA interventions that are functional in nature within individual and small group-based sessions, as well as providing parent / teacher guidance and supervision. Both, entry-level and advanced, specialized PT education as well as research opportunities exist for PT clinicians working with individuals with ASD in Israel.

VanDamme & Bhat (2023) reported a similar model in Belgium wherein PTs play a prominent role as motor service providers to conduct full motor assessments and further motor services, following physician referrals of young and older children as well as adults with ASD who failed motor screening. PTs in rehabilitation hospitals, school settings, or community-based, private practice can offer evaluations and treatments for children with ASD. PT clinicians having an entry-level Master’s degree can choose to gain further specialization in mental health or pediatrics to acquire advanced ASD expertise. US and Canada health service systems could emulate the Belgian or Israeli models when rethinking the role of PTs for individuals with ASD in hopes of providing better access to motor services for individuals with ASD and this has been suggested in contemporary literature (Campos et al., 2019; Mieres, Kirby, Armstrong, Murphy, & Grossman, 2012; Bhat, Iverson, Bishop, Thurm, & Licari, 2023; Miller et al., 2023).

If PTs agree that they are movement experts responsible for promoting movement and fitness by designing physical activity plans specific to the needs, challenges, and goals of individuals of all ages and abilities (see choosept.com) then their clinical practices and healthcare systems should not preclude access to PT for individuals with ASD. PTs should be able to provide screening, evaluation, and treatment for motor and PA services as well as assessments for reasonable accommodations for individuals with ASD through direct care (in early intervention, school, hospital, or private care settings) or in collaboration with other professionals. PTs need to advocate for motor services for individuals with ASD by asking to be part of interdisciplinary, management teams. Given the recognition for lack of ASD expertise among PT clinicians, there is a need to provide PTs and other movement experts better training and education in the area of understanding challenges associated with ASD as well as their evaluation and treatment. We also need to build greater awareness among ASD stakeholders (parents, clinicians, including diagnosticians and interventionists, and funding agencies) about the value of motor services and PA in promoting physical and mental well-being of children and adults with ASD.

Barriers to accessing physical activity (PA) programs for children with ASD

Caregivers of autistic children and autistic youths across multiple participatory action research studies have expressed systemic, environmental, family-related, and individual barriers to PA participation (Jachyra et al., 2021; Gregor et al., 2018; Blagrave, Colombo-Dougovito, & Healy, 2021; Arnell, Jerlinder, Geidne, & Lundqvist, 2022). Individual challenges to PA participation include motor, sensory, cognitive/executive functioning, social, and behavioral difficulties of children with ASD. As found in the SPARK study, a large group of individuals with ASD report motor challenges with planning, whole-body coordination, and balance which affects their ability to engage in community-based PA and may lead to a lack of motivation or long-term dislike for PA due to its substantial demands (Srinivasan, Pescatello, Bhat; 2014; Bhat, 2020a, 2020b, 2022; Kaur et al., 2018; Ketcheson et al., 2021; Wang et al., 2022).

Additionally, if the built environment of the exercise / PT facility providing PA does not accommodate autistic children’s sensory or executive functioning needs that could result in sensory overload from loud noises and bright lights, or planning / organizational difficulties (Blagrave et al., 2021). Social challenges of fitting in with peers due to non-acceptance of their challenging behaviors / restricted interests or systematic bullying by peers due to them not meeting group expectations can also become major barriers to accessing PA opportunities (Gregor et al., 2018; Jachyra et al., 2021; Arnell et al., 2022).

Family-related barriers include lack of family identity to be physically active and prioritization of sedentary services (Speech, OT, and ABA) over PA (Gregor et al., 2018; Jachyra et al., 2021). Lastly, systemic barriers include poor access and availability of PA programs including lack of supportive / enjoyable activities, lack of community support and empathy by caregivers, peers and their families (i.e., poor understanding of children’s needs), competitive / high performance nature of sports / PA programs, the lack of funding to support the programs or the caregivers (i.e., lack of transportation and participation/modeling assistance) as well as poorly trained professionals who do not understand individual needs (e.g., autocratic styles of coaches to promote rigorous routines) and lastly, excessive caregiver burden to balance therapies and PA opportunities (Gregor et al., 2018; Jachyra et al., 2021; Arnell et al., 2022) Overall, access to PA services is a multidimensional problem in children and youth with ASD and will require a more collaborative approach to enhance PA through help of PTs, adapted physical educators, community coaches / professionals, paraprofessionals/support staff, and parents.

Limitations

Current analyses were limited to studying motor / non-motor service receipt as a function of risk and severity of motor impairment; however, future publications will focus on various demographic and other co-occurring factors such as age, mother’s education, etc. and their influence on service receipt in children with ASD. A clear limitation of this work is its reliance on parent reporting of motor difficulties and services for children with ASD. However, using surveys to collect large sample data is a common practice in healthcare. Future studies assessing clinical assessment as well as record-based, services data should validate SPARK study findings in smaller but significant samples of children with ASD. Access to Adapted Physical Education in school settings was not tracked in the current SPARK sample and should be included in the future.

Clinical Implications and Conclusions

While ~88% children with ASD in the SPARK study were at risk for motor impairment; only ~35% were receiving PT services and only ~15% were receiving RT services. While motor services improved with greater risk and severity of motor impairment, there were significant disparities between physically active (PT: 37-55% and RT: 15-19%) and sedentary (Speech/OT: 85-92%) therapies across school, private, and community settings. It was also confirmed that discrepancies in PT/RT vs. OT service proportions did not align with motor needs of children with ASD. Both, gross and fine motor performance was moderately low and substantially affected in children with ASD. Together, these findings indicate that there are children with ASD who are undertreated for their motor problems and have clear unmet motor needs due to lack of access to PT and RT grounded in functional, enjoyable, gross motor, whole-body movement interventions and due to over-prioritization of sedentary therapies. Movement clinicians (physical and occupational therapists, adapted physical educators, play/creative movement therapists) need to advocate for motor services for individuals with ASD to meet children’s unmet motor/PA needs just as they would improve motor abilities and PA levels for individuals with other neurological diagnoses, Stroke, Parkinson’s disease, Cerebral palsy, Down syndrome, etc. Additionally, general PT practice or PA programs should be sensitive to the sensory/cognitive needs of individuals with ASD and clinicians should know how to adapt their clinical built environments and regular movement training practices to meet the needs of individuals with ASD within general practice settings. PT, adapted physical educators, and exercise/coaching professionals should make greater efforts to train future clinicians in caring for the multisystem needs of individuals with ASD including knowledge of their behavioral, sensory, motor/functional, cognitive, and social communication needs. Clinicians should seek training to gain a fundamental understanding of various ASD treatment strategies grounded in ABA, Picture Exchange Communication Systems (PECS), Treatment and Education of Autistic and Related Communication Handicapped Children (TEACCH), Sensory Integration, Floortime, and Motor Learning Theory during their entry-level education, through advanced pediatric specialization, through continuing education training programs, or via educational lectures at conferences. Such ASD-specific training should include focused modules on ASD diagnostic criteria and co-occurring conditions, ASD-specific treatment strategies, and clear clinical routes for accessing motor screening, assessment, and intervention to meet the specific needs of individuals with ASD. Clinical pathways for accessing motor services including PT should be developed by adapting clinical frameworks to emulate PT service delivery models implemented in other countries like Israel and Belgium. Taken together, several school-age children with ASD are undertreated and have unmet motor needs despite of having motor difficulties. Movement experts (PTs, OTs and adapted physical educators) should advocate for appropriate motor services and whole-body, movement interventions to meet family and child goals as well as provide enjoyable, individual and/or group-based PA opportunities to improve physical and mental well-being of all individuals with ASD.