Abstract
This study adds to a body of literature suggesting that Ayres Sensory Integration® is effective in increasing functional goals in children with sensory processing and integration challenges.
A systematic review by Schaaf et al. (2018) found that as many as 95% of pediatric occupational therapists, across different settings, use sensory integration interventions when working with children who have difficulties processing sensory information. Despite its frequency of use, there continues to be a dearth of rigorous research to support the effectiveness of sensory integration treatment across diagnostic conditions. The purpose of this study was to contribute to this body of literature by testing the effectiveness of occupational therapy using an Ayres Sensory Integration® (ASI) approach for children with sensory-based motor disorders (SBMDs).
Background
A. Jean Ayres developed the theory of sensory integration to explain the necessity of adequate processing and integration of sensory information for adaptive behavior and functional skills (Ayres, 1972; Schaaf & Miller, 2005). She defined sensory integration as the neurological process of organizing sensory information for use; this process serves as a foundation for participation across all domains of daily life (Ayres, 1972). Ayres postulated that contributions of the vestibular, proprioceptive, and tactile systems, in particular, formed the foundation for postural–ocular control, body awareness, and motor planning through mechanisms of feed forward and feedback processing. End products of adequate sensory integration, therefore, include successful engagement in motor activities; adequate fine motor, gross motor, and visual–motor coordination; and appropriate levels of self-esteem and self-efficacy (Ayres, 1979 ; Bundy & Lane, 2020). According to Ayres’ theory, children with inadequate sensory processing and integration in bodily senses (i.e., proprioceptive, vestibular, somatosensory, and interoceptive) are at risk for deficits in coordination and motor planning.
Many terms have been used to describe children with deficits in motor coordination and motor planning that are not the result of identifiable brain damage (e.g., cerebral palsy, traumatic brain injury) or genetic conditions (e.g., muscular dystrophy); diagnostic terms include developmental coordination disorder (DCD); dyspraxia or developmental dyspraxia; and deficits in attention, motor control, and perception (DAMP). To help identify the specific subset of children whose motor deficits stemmed from difficulties with sensory processing and integration, Miller, Anzalone, et al. (2007) introduced the term sensory-based motor disorders (SBMDs) in their published classification system of sensory processing disorders.
Children with SBMDs are thought to have poor postural control or volitional movement as a result of deficits in sensory integration, and they may present with a postural disorder or dyspraxia as a result. Although a full review of diagnostic terms is beyond the scope of this article, some consensus has been reached that DCD, dyspraxia, DAMP, and SBMD overlap to some degree; that all are present at birth or early in development; and that clumsiness or incoordination across all conditions can affect a child’s ability to master developmentally appropriate functional skills (Bhoyroo et al., 2018; Farmer et al., 2017 ; Gibbs et al., 2007).
Several intervention approaches have been suggested for children with deficits in motor coordination, including sport and fitness programs (e.g., Fong et al., 2012; Menz et al., 2013), neuromotor training (e.g., Cheng et al., 2019; Fong et al., 2016), and cognitive task-oriented approaches (e.g., Sangster et al., 2005; Scammell et al., 2016; Zwicker et al., 2015). For children with foundational deficits in sensory processing and integration, Ayres developed an intervention approach that served to remedy those underlying sensory–motor deficits using active engagement and challenges within a sensory-enriched environment; this intervention approach has been trademarked as ASI to distinguish it from other sensory-based approaches.
The goal of ASI is to facilitate improved ability of the brain to process and integrate sensory information as a foundation for optimal functioning and participation in everyday occupations (Ayres, 1972). Rather than teaching specific skills in a repetitive manner or using cognitive strategies, the ASI approach (Schaaf & Mailloux, 2015) is focused on providing opportunities for active and child-directed engagement in sensory–motor activities that challenge a child’s sensory systems, with a particular focus on the tactile, vestibular, and proprioceptive systems (Schaaf & Miller, 2005). The outcome of this intervention is hypothesized to be a greater foundation from which all new motor skills can develop and from which the child can respond more appropriately to dynamic changes in their environment (Bundy & Lane, 2020).
Over the past decade, research into the effectiveness of ASI has primarily focused on children diagnosed with autism spectrum disorder (ASD)—most likely because of the prioritization of research funding in this area. Between 2007 and 2015, five research studies measured the effectiveness of ASI (also referred to as “occupational therapy using a sensory integration approach”) in children with ASD (Dunbar et al., 2012; Iwanaga et al., 2014; Pfeiffer et al., 2011; Preis & McKenna, 2014; Schaaf et al., 2014). These published studies provided evidence to support the use of ASI for improvement of functional skills and participation among children with ASD.
However, despite children with ASD being a population with known motor coordination deficits, studies to date have failed to sufficiently measure the impact of ASI intervention on motor outcomes. In addition to a lack of research on motor outcomes, only one study, published more than a decade ago (Miller, Schoen, et al., 2007), has tested the ASI approach among children without ASD; yet again, motor outcomes were not a focus of that study. Therefore, the purpose of our study was to fill existing gaps in the literature by testing the effects of ASI on motor outcomes in a population of children with idiopathic sensory processing and integration challenges, specifically SBMDs. We hypothesized that significant and meaningful improvements would be observed in children’s motor skills and functional goals after a 30-session dose of ASI intervention.
Method
Study Design
In this study, we used a quasi-experimental, nonconcurrent multiple-baseline, repeated-measures design with an A–B configuration. This design allows for enrollment flexibility while controlling for possible threats to external validity by staggering the introduction of the independent variable (Kennedy, 2005). The baseline (A) phase was a nonintervention condition in which stability of motor skills was established for each participant through repeated measurement of probes. On the basis of methods outlined by Kazdin (1982) and Kennedy (2005), stability had to be obtained before initiating intervention. Stability was defined in this study as at least three consecutive data points with less than 20% variance.
During the intervention phase (B), participants took part in ∼60-min ASI therapy sessions with a licensed occupational therapist. As in the baseline phase, probes were measured weekly during the intervention phase and were used as a formative mechanism for gauging participant improvement. Pre- and posttest measures of motor performance and goal attainment were also completed as part of this study. Study procedures were approved by the Virginia Commonwealth University institutional review board; the study was registered with ClinicalTrials.gov (NCT03805334).
Participants
Children were recruited between February and December 2018 from a wait list at a university-affiliated children’s hospital in an urban setting. Participants were included in the study if they met the following criteria: (1) between ages 4 yr, 11 mo and 8 yr, 11 mo; (2) diagnosed with “lack of coordination” (R27.8 medical billing code) by a physician; and (3) able to communicate verbally per parent report. Children were also screened to ensure they had a sensory processing impairment that affected their motor performance; the Sensory Processing Measure–Home form (SPM–H; Parham, Ecker, et al., 2007) was used to assess this facet of eligibility. To qualify, participants needed to score in the “definite dysfunction” range in at least one of three SPM–H subcategories of body awareness, balance and motion, and planning and ideas.
Children were excluded from the study if they had a diagnosis of ASD or any considerable physical disability (e.g., cerebral palsy, muscular dystrophy) or psychological impairment (e.g., bipolar disorder, schizophrenia). Children were not excluded for having attention deficit hyperactivity disorder or a diagnosis of a learning disorder; any diagnoses were recorded as part of the screening process. To qualify for the study, participants could not be receiving occupational therapy services in another capacity during the study.
Five children were screened for eligibility and met all inclusion and exclusion criteria. Two families withdrew from the study before the baseline phase because of scheduling conflicts. Three children (M age = 6.4 yr) were enrolled in and completed all aspects of the study. Participant demographics are presented in Table 1.
Participant Demographics
Note. ADHD = attention deficit hyperactivity disorder; SPM–H = Sensory Processing Measure–Home form.
Intervention
The intervention sessions were scheduled for 3 times per week for 10 wk (30 sessions total), with each session lasting ∼60 min. The intervention followed the principles of ASI as outlined by Ayres (1972, 1979) and described in detail by Schaaf et al. (2012, 2014) and Parham and colleagues (Parham, Cohn, et al., 2007 ; Parham et al., 2011). A data-driven decision-making process (Schaaf & Blanche, 2012; Schaaf & Mailloux, 2015) was used to develop hypotheses about sensory processing and integration factors affecting the child’s motor performance, and individually tailored sensory–motor activities were developed to address these factors. Intervention activities were performed within the context of play, and the treating therapist encouraged active involvement of the child in both the planning and the “doing” of the activity. During sessions, the treating therapist intentionally encouraged active engagement, facilitated adaptive responses, and presented the just-right challenge during motor tasks.
Intervention took place in one of two large treatment rooms that were housed within an urban children’s hospital. Each room was equipped with mats, a variety of suspended equipment (e.g., swings, hammocks, ladders), large balls, climbing structures, and other materials for sensory–motor play. A long hallway area was also used for scooter board activities and riding toys (e.g., bike, roller racer).
Interventionist
All interventions were conducted 1:1 with the same occupational therapist (Lauren Andelin), who has advanced training and mentorship in sensory integration intervention, including certification from the Sensory Therapies and Research Center’s (now called the STAR Institute) intensive mentorship program. All sessions were video recorded by a rehab technician and were used as part of our fidelity assessment (described in the next section). Parents were encouraged, but not required, to be present for each session; weekly debriefs were conducted with all families.
Fidelity
The Ayres Sensory Integration Fidelity Measure (ASIFM) was used to ensure consistency of the intervention and adherence to the key structural and process elements of the ASI intervention (Parham, Cohn, et al., 2007; Parham et al., 2011). The tool looks at aspects of the environment, therapist credentials, and evidence of parent–therapist collaboration in goal setting. The tool also measures 10 therapeutic strategies used in ASI. A score of 80 of 100 is considered acceptable fidelity for ASI.
For this study, 15 min of each intervention session was assessed for fidelity. The 15-min segment observed was determined by entering the start and stop times for each video into a random number generator (https://www.random.org/). For example, if the program generated the number 17, then Minutes 17–32 of the video were assessed for fidelity. All videos were assessed for fidelity by the principal investigator (Stacey Reynolds), who is trained in use of the ASIFM tool and certified in administration of the intervention.
At the completion of the study, 96.6% (86 of 89) of the intervention sessions were recorded and reviewed for fidelity; missing videos occurred because of technician unavailability. Of the sessions reviewed, 98.8% (85 of 86) met fidelity standards. The mean fidelity score across all intervention sessions was 89 (SD = 6.78, range = 56–100).
Measures
Probes
Three motor tasks were used as probes throughout the study: (1) finger-to-nose touches with eyes open, (2) jumping jacks, and (3) balancing on one foot with eyes open. These probes are common test items on motor assessments such as the Bruininks–Oseretsky Test of Motor Proficiency–Second Edition (BOT–2) Complete Form (BOT–2 CF; Bruininks & Bruininks, 2005) and the Quick Neurological Screening Test (Mutti et al., 2017); in addition, they are often incorporated into clinical observations related to sensory and motor processing (Blanche, 2010). Each probe was administered 3 times per week before initiation of intervention until stability of baseline was achieved. Each probe was also administered 1 time per week during the course of intervention. All probes were administered after the treatment session and were recorded by the treating therapist (Lauren Andelin).
For the finger-to-nose probe, the participant was directed to stand on two feet with eyes open and arms abducted bilaterally with palms facing forward. Participants were instructed to touch index finger to nose one arm at a time for 10 s without stopping the movement and only flexing one arm at a time. Participants were allowed one trial touch before starting. Proper touches were considered tip of middle phalanx of index finger touching tip of nose at center, and arms staying in proper position throughout the movement.
For the jumping jack probe, the participants were directed to complete jumping jacks for 10 s. Participants were allowed one trial jump before starting. Proper jumping jacks were considered jumping with the legs spread wide and hands touching overhead before returning to the position with feet together and arms at sides. The jumps were to be in smooth repetition (i.e., continuous movements without extended pauses) without any stopping and starting between jumps. The number of proper jumps were recorded over 10 s.
For the one-foot balance probe, the participants were directed to stand on the dominant foot with hands on hips and eyes open for as long as they could without touching the nondominant foot to the ground. Participants were allowed one practice trial before starting. Balance was recorded in seconds until the participant touched the nondominant foot to the ground or hopped on the dominant foot.
Goal Attainment Scaling
Goal attainment scaling (GAS) was used to develop individual goals for each client that could be measured over the course of intervention with a standardized scale (Kiresuk et al., 2014). Goals are rated on a 5-point scale, and criteria for each level are established before starting treatment. Although each client has individual goals that are meaningful to them and their family, the outcome measurement scale is standardized so that it can be compared with achievement of other client goals. The expected level of outcome is established at initial goal setting, and the number 0 is used to rate an outcome in which a client achieves the expected level.
If a client achieves a better than expected outcome, the score can be +1 (somewhat better) or +2 (much better). If the client achieves a worse outcome than expected, the score can be −1 (somewhat worse) or −2 (much worse). GAS goals are identified through a semistructured interview with the client or caregivers, and three to five goals are set and weighted by importance (McDougall & King, 2007). GAS has been used previously in sensory integration outcomes research (Miller, Schoen, et al., 2007; Pfeiffer et al., 2011; Schaaf et al., 2014) and has been shown to be a reliable method for measurement of progress among children with neurodevelopmental disabilities (Ruble et al., 2012).
In our study, an independent evaluator (a licensed occupational therapist) trained in GAS administration completed a standardized GAS interview with families before the start of intervention and set four to five goals with defined outcome levels for each participant. The evaluator rated the baseline levels, which are scored as −2, for each goal before the start of the study. After the intervention period, the same evaluator met with the families and asked them to rate their child’s goal achievement. Goals for each participant are presented in Table 2.
Goal Attainment Scaling
Note. Goal attainment scaling scoring criteria: −2 = baseline level of performance; −1 = less than expected level of attainment; 0 = expected level of attainment; +1 = better than expected level of attainment; +2 = much better than expected level of attainment. N/A = not applicable.
Bruininks–Oseretsky Test of Motor Proficiency–Second Edition Short Form
We tested participants in the study pre- and postintervention using the BOT–2 Short Form (BOT–2 SF; Bruininks & Bruininks, 2005). The BOT–2 is a standardized assessment that uses goal-directed motor activities to measure a wide array of gross and fine motor skills in people ages 4 to 21 yr. The assessment can be administered as a whole (BOT–2 CF), select composites (BOT–2 SF), or select subtests. For the purposes of this study, the BOT–2 SF was used as a pre- and posttest measure of a range of motor skills. The BOT–2 SF consists of 14 items and takes approximately 25 min to administer. It includes eight subtests with two items each, including fine motor precision, fine motor integration, manual dexterity, bilateral coordination, balance, running speed and agility, upper limb coordination, and strength. A high correlation has been found between the BOT–2 SF and the BOT–2 CF, indicating strong content validity (Rigby et al., 2020). Test–retest reliability is strong (r ≥ .80) when using the BOT–2 SF with knee pushups (vs. full pushups; Deitz et al., 2007), as was used in this study. For this study, the BOT–2 SF was administered to each child by the same independent evaluator who completed the GAS interviews.
Data Analysis
We first analyzed probe data using visual inspection. For each participant, the three motor probes were graphed to compare the baseline and intervention phases (Figures 1–3). Values on the y axis represent duration for the leg balance probe, number correct for the finger-to-nose probe, and number completed for the jumping jacks probe. We conducted visual analyses for each phase using change in mean level and rate of change (i.e., trend in the data). Nugent’s (2010) method was used to reflect the trend in the data by drawing a line between the first and last data points (Engel & Schutt, 2013). We conducted statistical analyses using the 2 SD band analysis to determine the percentage of data points during intervention that fell outside the expected baseline range. A rule of thumb offered by Gottman and Leiblum (1974) is that if at least two consecutive intervention phase data points fall outside the band, a meaningful change was observed. The logic for this rule of thumb is that the probability of this occurrence happening by chance is less than the criterion of p < .05 (Nourbakhsh & Ottenbacher, 1994).
Motor probes for Participant 1 from baseline to intervention.
Motor probes for Participant 2 from baseline to intervention.
Motor probes for Participant 3 from baseline to intervention.
GAS goal scores were compared pre- and postintervention using the Wilcoxon matched-pairs signed-rank test; in addition, a calculation of percentage of goals met was generated. Hedges’ g statistic was calculated to assess the magnitude of change between pre- and postintervention scores on the BOT–2. Hedges’ g is similar to Cohen’s d except when sample sizes are less than 20, when Hedges’ g outperforms Cohen’s d (Cohen, 1977). The statistic was manually calculated by dividing the mean difference between pre- and posttest scores by the weighted and pooled standard deviation for the two groups.
Results
Intervention Attendance and Fidelity
Participant 1 (P1) completed 29 intervention sessions over 9 consecutive wk. His last session was cancelled because of a family medical issue and was not rescheduled. Participant 2 (P2) and Participant 3 (P3) each completed 30 sessions over 10 (P2) or 11 (P3) consecutive wk. P3 had 3 wk with only two sessions because of participant illness or national holiday and made up the missed sessions during the 11th wk of intervention. Means and standard deviations from which the 2 SD analyses were conducted can be found in Table 3.
Baseline and Intervention Means and Standard Deviations
Repeated Measure: Motor Probes
The graphs for P1 appear in Figure 1 and show strong effects for the three motor probes (i.e., leg balance, jumping jacks, and finger to nose). The mean level from baseline to intervention increased, and the 2 SD analyses were strong for all three tasks. The 2 SD band showed little or no overlap for each variable (e.g., 67% for leg balance, 78% for jumping jacks, and 100% for finger to nose). The trend increased for finger to nose but was relatively unchanged for leg balance and jumping jacks.
The data for P2 appear in Figure 2 and show moderate to strong effects. The mean level for each variable showed increases, and the 2 SD analysis was moderately strong for leg balance, weak for jumping jacks, and strong for finger to nose. P2 was slower to change as evidenced by level changes not increasing until Session 9, after which the trend increased during the rest of the intervention.
Data for P3 appear in Figure 3 and show a positive effect for the finger-to-nose task. Leg balance and jumping jacks were relatively unchanged from baseline through intervention. Finger to nose showed change on all variables, including an increase in mean level, 100% nonoverlapping data as evidenced by the 2 SD analysis, and an increase in rate of change during intervention (i.e., trend). None of the variables for leg balance showed change. However, a slight increase occurred in the mean level for jumping jacks; this increase was not sustained beyond Session 9.
Goal Attainment Scaling
Five GAS goals were established for P1 and P2, and 4 goals were established for P3, resulting in 14 GAS goals. Of these 14 goals, 5 goals were related to feeding or oral–motor skills, 3 goals focused on postural control or core strength, 4 goals were related to dressing or toileting, and 2 goals were focused on attention or self-regulation. The outcomes of each goal are presented in Table 2. Progress was made toward 100% of the established goals, with 11 out of 14 (78.6%) goals achieving the expected level of attainment (0) or exceeding the expected level of attainment (+1 or +2). Results of the Wilcoxon matched-pairs signed-rank test revealed significant differences between pre- and posttest scores (p = .001).
Bruininks–Oseretsky Test of Motor Proficiency–Second Edition
BOT–2 data trends were examined visually, and scores are presented in Table 4. All 3 participants showed improvement in their BOT–2 total score and percentile rank, indicating improvement in motor performance over the intervention period. The Hedges’ g statistic was 1.05, indicating a large effect.
BOT–2 Outcomes
Note. BOT–2 = Bruininks–Oseretsky Test of Motor Proficiency–Second Edition.
Discussion
This study provides new evidence supporting the use of an ASI approach with children who have deficits in sensory processing and integration that affect their motor coordination. Similar to previous studies in the ASD population, significant and meaningful changes were observed across all GAS goals, suggesting high potential for ASI to lead to improvements in functional performance such as self-care and play participation. Although neurological changes were not measured directly in this study, researchers have hypothesized that these functional improvements are related to the impact of therapy on processing and integration of sensory input contributing to improvements in motor coordination. Because of our small sample size, significance testing of BOT–2 scores was not performed; however, the effect sizes for this measure were large, suggesting that the 30-session dose of ASI intervention was sufficient to elicit meaningful improvements in children’s developmental motor skills.
The inclusion of the repeated measurement of movement probes to examine motor outcomes of ASI intervention was a novel aspect of this study. The motor probes of finger-to-nose touching, jumping jacks, and one-foot balance were chosen as proxies for motor coordination, with the intent that they could be easily measured and that changes would be observed shortly after initiating intervention. The probe that most clearly and consistently demonstrated change after the initiation of the intervention was the finger-to-nose probe, with strong positive effects seen across all 3 participants. Reflecting on these outcomes, it is possible that the other two probes (i.e., jumping jacks, one-foot balance) are more advanced skills, requiring greater levels of neurological integration, and are not appropriate to use as a measure of short-term change.
In addition to considering differences in our motor probes, results of this study must also be considered within the context of the individual participants. P1 had the highest scores on the BOT–2 at baseline, with total scores in the average normative range. The fact that this participant scored “much better than expected” on 3/5 GAS goals and had strong improvements across all three motor probes may be because of his high baseline level of function and the foundational skills needed to show improvement with a relatively short dose of intervention.
P2, however, was the youngest participant, with BOT–2 scores well below average at baseline. This participant showed much smaller changes, and changes on motor probes specifically appeared toward the end of the intervention. It is possible that children who have lower baseline motor scores will need a longer duration of intervention to see optimal impact.
Finally, P3 also had low baseline motor scores on the BOT–2 and showed little to no change on the motor probes of jumping jacks and one-foot balance. Examining his scores graphically shows a high rate of variability and inconsistency in his performance. One possible explanation is that this child had a known history of trauma and was currently in the process of integrating into his new adoptive family; variability in his motor performance may have been related to factors such as inattention, anxiety, or fluctuating arousal levels that were not measured but could have influenced his performance from week to week. These considerations are important for future trials aimed at better understanding dosage and factors mediating intervention effects.
Limitations
The limitations of this study are related primarily to measurement. Although an independent evaluator was used for the BOT–2 and GAS goals, that evaluator was not blinded to the study aims, potentially leading to examiner bias. In addition, motor probes were administered at the end of a therapy session each week, which may have slightly inflated the results. Last, the interventionist also performed the probe measurements for pragmatic reasons, further enhancing the risk of examiner bias in our study. In future studies, probes should be video recorded before intervention each week and then scored by a blind evaluator to increase internal validity; all other tests should be performed by an independent examiner blinded to study aims.
Implications for Occupational Therapy Practice
The results of this study have the following implications for occupational therapy practice: Clinicians should consider using an intensive dose of ASI intervention to improve foundational motor abilities necessary for the performance of functional skills among children with SBMDs. Children with higher baseline levels of motor skills may show faster gains than children with lower baseline scores. Children with lower levels of motor skills at baseline may need a longer duration of the intervention to see maximal benefits.
Conclusion
Findings from this study suggest improved motor skill and high goal achievement using a 3 times per week model of therapy with ASI. This study adds to a body of literature suggesting that ASI is effective in increasing functional goals among children with sensory processing and integration challenges. It also adds new information to the literature that suggests ASI can increase motor abilities among children with SBMDs. Future studies using multiple baseline designs are feasible in the clinic setting and can help build the evidence base for ASI.
Footnotes
Acknowledgments
Support for this project was provided by the C. Kenneth and Dianne Wright Center for Clinical and Translational Research (Grant UL1TR002649). This study was approved by the Virginia Commonwealth University (VCU) institutional review board (Approval HM20012008) and is registered with ClinicalTrials.gov (NCT03805334). The authors report no conflicts of interest. The authors acknowledge the Children’s Hospital of Richmond at VCU for their support with participant recruitment and facility use. The authors also acknowledge Allison Wolf, who conducted the Bruininks–Oseretsky Test of Motor Proficiency–Second Edition testing and goal attainment scaling goal setting during this study. The principal investigator, Stacey Reynolds, is Editor-in-Chief of the American Journal of Occupational Therapy.