Abstract
Currently, 5%–16.5% of school-age children are estimated to have sensory processing difficulties (SPD), which adversely affect a child’s ability to participate in daily activities (Ahn et al., 2004; Ben-Sasson et al., 2009; Miller et al., 2009). Many people with SPD may exhibit motor planning difficulties (Miller et al., 2007), problems with play skills (Koenig & Rudney, 2010), and impairments in activities of daily living (White et al., 2007). They may also experience difficulty with social participation (Cohn et al., 2000) and low academic achievement, all of which negatively affect emotional well-being (Koenig & Rudney, 2010). Young children with SPD have social–emotional problems, which increase over time and can affect social competence at school age (Ben-Sasson et al., 2009).
Early identification of SPD may enable early intervention, which has shown promising results in toddlers with developmental delays and SPD (Blanche et al., 2016). If services can be provided when symptoms first appear, deficits will potentially be minimized before characteristics of this condition are fully manifested (Blauw-Hospers et al., 2007). Many risk factors are associated with SPD, including prematurity and early environmental experiences in the neonatal intensive care unit environment (Crozier et al., 2016). Evidence also suggests a genetic predisposition to SPD (Keuler et al., 2011). In the current study, we were interested in this heritable link to SPD (Keuler et al., 2011), prospectively examining infants who are siblings of children with SPD. This research is likely to contribute to a greater understanding of the early signs of SPD.
Successfully implemented prospective studies on infants at high genetic risk for autism spectrum disorder (ASD) provide a model for our study of infants at high genetic risk for SPD. Longitudinal studies in ASD have been successfully conducted, finding early behaviors associated with eventual diagnosis. For example, studies of infants who have older siblings with ASD found that early behavioral features—such as increased perceptual sensitivity (Clifford et al., 2013), delayed gross motor and visual reception skills (Estes et al., 2015), and head lag at 6 mo (Flanagan et al., 2012)—may be early indicators of developmental disruption in at-risk infants later diagnosed with ASD. Decreasing gaze to another’s eyes (Jones & Klin, 2013) and abnormal visual orienting (Elison et al., 2013) may also be early behaviors associated with ASD.
To the best of our knowledge, researchers have not conducted between-groups comparisons of infants at risk for SPD versus those not at risk to characterize the early clinical symptoms of SPD. We hypothesized that signs and symptoms of SPD manifest early in a child’s life. We aimed to compare infants at low and high risk for SPD. Our goal was to characterize SPD risk factors in young children to assist health care professionals in identifying problems early and to determine which developmental assessments are sensitive to detecting problems early in the child’s development.
We compared behaviors in a group of high- and low-risk infants as a precursor to a larger study to characterize early signs of SPD. The aims were to answer the following research questions:
Can babies at high risk for SPD, ages 4–24 mo, be differentiated from those at low risk?
Which outcome measures are associated with risk status?
Method
A case–control observational pilot study was conducted to evaluate differences in developmental skills between infants at low risk and high risk for SPD. Study data were collected either in the clinic or during a home visit by the primary researcher (first author), a pediatric occupational therapy practitioner with 20 yr of experience administering developmental assessments. This researcher was blinded to the child’s risk status. Data collection sessions were videotaped. Informed parental consent was received before infants were admitted to the study. The study was approved by the Rocky Mountain University of Health Professions institutional review board.
Participants
Two groups of infants ages 4–24 mo with older siblings were recruited for this study. The high-risk (HR) group was composed of infants who had an older sibling with SPD. The older sibling’s SPD was based on confirmation of impairment in SPD defined by (1) the global clinical impression of an occupational therapy practitioner after standardized testing using the Miller Function and Participation Scales (Miller, 2006) or the Bruininks–Oseretsky Test of Motor Proficiency–2 (Bruininks & Bruininks, 2005) and atypical performance on the Sensory Processing Scale Assessment (Schoen et al., 2014), (2) abnormal scores on the Short Sensory Profile (McIntosh et al., 1999; <−2.5 standard deviations) and the Sensory Processing 3 Dimensions (SP3D) Inventory (Schoen et al., 2017), and (3) atypical performance during structured and unstructured observation in the clinic. The low-risk (LR) group consisted of typically developing infant siblings who had no older sibling or parental history of sensory problems or diagnoses in the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; American Psychiatric Association, 2013).
Exclusion criteria for both groups included significant birth trauma or birth risk factors (e.g., birth weight < 2,500 g, prematurity < 37-wk gestation, interventricular hemorrhage at birth); head injury; prenatal drug or excessive alcohol exposure; genetic, orthopedic, or neurological conditions besides SPD in the HR group; and English as a second language for the family.
The HR group was recruited from families of children with SPD attending therapy at the Sensory Therapies and Research (STAR) Institute for Sensory Processing Disorder in Greenwood Village, Colorado. The LR group was recruited through the community (i.e., local physicians’ offices, caregiver–child playgroups, day care settings, places of worship) and through friends and family of the STAR Institute’s staff.
An initial phone screening was conducted by research assistants to verify that the infants met criteria for either the LR or the HR group. If they met inclusion criteria, an appointment was scheduled. Parents or guardians of children in the HR group signed a consent to release confidential information that confirmed impairment in sensory processing in an older sibling. Initially, 30 families enrolled in the study. One infant in the HR group was teething and irritable and so performance-based assessments could not be completed; in addition, 3 children had unreturned parent-report measures (2 LR; 1 HR). One infant in the LR group had received physical therapy secondary to motor concerns, which was revealed during the evaluation. This child was removed from the study secondary to exclusion criteria. Thus, the final sample consisted of 29 participants.
Procedures
Recruitment, screening, scheduling, and permission forms were gathered by research assistants to keep the primary researcher blind to risk status. Parents were asked by the research assistants not to discuss risk status during the evaluations.
Instrumentation
A demographic questionnaire and parent-report measures were given to the primary caregiver at the beginning of the evaluation. Infants were tested with the caregiver present. Assessment ranged from one to three sessions depending on infant tolerance. Infants were seated in an infant seat, toddler chair, or parent’s lap (depending on the child’s age) for fine motor and social measures. Infants’ gross motor tasks were observed on a gym mat covered with a blanket. Performance-based assessments were scored in real time and were verified by the primary researcher for accuracy through videotaped recordings. The Toddler and Infant Motor Evaluation (T.I.M.E.; Miller & Roid, 1994) Mobility Subtest was only scored from videotapes. Parent-report measures were scored by the research assistants. Data were analyzed by the primary researcher.
Performance-Based Measures.
The Bayley Scales of Infant and Toddler Development–Third Edition (Bayley–III; Bayley, 2006) is an assessment of developmental functioning in infants and children ages 0–42 mo. It has three scales measuring cognition, language, and motor skills. Scaled composite scores were found to have high internal consistency as demonstrated by Cronbach’s αs ranging from .91 to .93 and subtests ranging from .86 to .91 (Albers & Grieve, 2007). Strong validity is reported between the Bayley–III Cognitive and Language Composites and the Wechsler Preschool and Primary Scale of Intelligence–Third Edition (Wechsler, 2002) verbal performance and full-scale scores (.71–.83). The Bayley–III Language Composite was moderately correlated with the Preschool Language Scale—Fourth Edition (Zimmerman et al., 2002) and the Auditory Comprehension and Expressive Communication subscales (.51–.71; Albers & Grieve, 2007).
The T.I.M.E. (Miller & Roid, 1994) consists of five subtests assessing quality of movement in young children ages 4–42 mo. The following subtests were used in this study:
The Motor Organization Subtest assesses praxis and sequencing (i.e., the ability to perform motor skills requiring visual and spatial skills and balance). This subtest has high interrater reliability ranging from .90 to .99 and the ability to discriminate between children with and without motor delays.
The Mobility Subtest assesses quality of movement in developmental positions and transitions. High reliability and accuracy for correctly identifying children with and without motor delays have been reported.
The Atypical Position Subtest assesses atypical movement patterns, including poor weight shifting, atypical extension and flexion, hypertonicity or hypotonicity, and presence of primitive reflexes. Discriminant validity was found between children with and without motor delays. Internal consistency reliability was .88 for infants ages 0–6 mo and .97 for infants ages 7–12 mo. Data indicated strong interrater reliability (r = .99).
Parent-Report Measures.
The Bayley Social–Emotional and Adaptive Behavior Questionnaire–Parent Caregiver Rating Form (Bayley, 2006) is used to measure adaptive behavior and social–emotional development. For the Adaptive Behavior Scale, reliability coefficients ranged from .79 to .98. Test–retest stability coefficients were .80 or higher for domain scores and the General Adaptive Composite scores, but they were slightly lower for specific skill areas. Interrater reliability ranged from .73 to .82. The Social–Emotional Scale has strong internal consistency, with coefficients ranging from .83 to .94.
The Infant/Toddler Sensory Profile–2 (birth–6 mo; Dunn, 2014) and the Toddler Sensory Profile–2 (7–35 mo; Dunn, 2014) are caregiver questionnaires designed to measure sensory processing in children. Studies have supported discriminant validity of the scale between children with and without conditions such as developmental delays, prematurity, low birth weight, reflux, and otitis media. Test–retest reliability has ranged from .83 to .97, and interrater reliability has ranged from .49 to .89.
The Communication and Symbolic Behavior Scales Developmental Profile™ (CSBS DPTM) Infant–Toddler Checklist (Wetherby & Prizant, 2002) is a parent-report, norm-referenced measure of communication skills, yielding three composites (Social, Speech, and Symbolic) designed for young children ages 6–24 mo. It has a high positive predictive value, estimated to be .75, for detecting ASD or developmental delay (Pierce et al., 2011).
The SP3D Inventory for Infants is an unpublished parent questionnaire of sensory modulation in infants. This inventory is an extension of the SP3D Inventory (Schoen et al., 2017). It has three subscales: Sensory Overresponsivity, Sensory Underresponsivity, and Sensory Craving. The sum of each of the three subscales was used in this study.
Data Analysis
IBM SPSS Statistics (Version 23; IBM Corp., Armonk, NY) was used for data analysis. Statistical significance was set at p < .05 a priori. When analyzing the demographic data, we made comparisons using t tests for age and a χ2 or Fisher’s exact probability test for gender, ethnicity, and socioeconomic status.
When analyzing the standardized measures, the scores were normally distributed as assessed by the Shapiro–Wilk’s test (p > .05). We used t tests to evaluate the difference between risk status (LR vs. HR) groups on outcome measures.
For the categorical outcome variables, a χ2 analysis or Fisher’s exact test was used to evaluate the difference between risk status (LR vs. HR) groups and atypical positions (dichotomized as none or presence of atypical positions) and scores’ classification columns for the Toddler Sensory Profile–2 quadrants (i.e., Much Less Than Others, Less Than Others, Just Like the Majority of Others, More Than Others). Descriptive statistics were also used to highlight important comparisons in outcome measures between the two groups, including means and standard deviations. Because this was a pilot study, we did not adjust for multiple comparisons. To determine which outcome measures were associated with risk status, we used a point biserial correlation.
Results
Participants
Thirteen HR infants (6 boys, 7 girls) ranging in age from 4 to 24 mo (11.96 ± 6.771) and 16 LR infants (6 boys, 10 girls) ranging in age from 4 to 24 mo (13.59 ± 6.37) participated in this study. No group differences were found in demographic data, including age, gender, race, and socioeconomic status (parents’ level of education attained; Table 1).
Demographics
Note. Education refers to the frequency (%) of parents with a college or graduate school education.
Not all participants in this group had data available; comparisons were made with χ2 or Fisher’s exact test.
Outcome Measures
Performance-Based Measures.
A statistically significant difference was found between HR and LR groups on the Bayley–III Cognition, Receptive Communication, and Language Composite. Compared with the LR group, the HR group demonstrated significantly lower scores, indicating poorer performance in Cognition, t(26) = −2.92, p = .007; Receptive Communication, t(26) = −2.87, p = .008; and Language Composite, t(26) = −2.90, p = .007 (Table 2). Group differences in atypical positions (T.I.M.E.) were also found, with the HR group demonstrating more statistically significant atypical positions than the LR group (p = .016). Eight HR (62%) and 2 LR (13%) infants demonstrated atypical positions. No significant group differences were found in motor organization and motor stability (T.I.M.E.; see Table 2), although the HR group performed worse than the LR group on these subtests.
Means on the Performance-Based Measures and Effect Sizes
Note. Comparisons were made with t tests for all measures, except a Fisher’s exact test was used for atypical positions. Bayley–III = Bayley Scales of Infant and Toddler Development—Third Edition; M = mean; NA = not applicable; SD = standard deviation; T.I.M.E. = Toddler and Infant Motor Evaluation.
Parent-Report Measures.
No significant group differences were found in adaptive behavior and social–emotional development (Bayley Social–Emotional Questionnaire score; Bayley–III General Adaptive Composite, Conceptual, Social, and Practical scores), sensory modulation (the SP3D Inventory for Infants), and communication skills (the CSBS DP). A significant group difference in the sensory-seeking quadrant of the Toddler Sensory Profile–2 (Dunn, 2014) was found, with the HR group (n = 8) engaging less than the LR group (n = 10; p = .039) in sensory-seeking behaviors. Compared with 20% of the LR toddlers, 75% of the HR toddlers scored less than or much less than typical performance in sensory seeking. The majority of the LR toddlers (70%) scored in the typical range, compared with 25% of the HR toddlers. No group differences were found in the other sensory quadrants.
Association Between Outcome Measure and Risk Status.
A moderate correlation was found between risk status and Bayley–III Cognition (r pb = .496; n = 28; p = .007), Receptive Communication (r pb = −.491; n = 28; p = .008), and Language Composite (r pb = −.495; n = 28; p = .007). A fair correlation was observed between risk status and motor organization (T.I.M.E.: r pb = −.374; n = 28; p = .050). No other significant correlations were found.
Discussion
On the basis of the findings of published research, we conducted this pilot study to evaluate whether early identification of infants with sensory challenges was possible. In this preliminary study, we divided participants into two groups according to their risk status for SPD on the basis of findings of significant familial associations in previous research (Keuler et al., 2011). Our findings suggest group differences between infants at HR and LR for SPD in cognitive and language skills, atypical movement, and sensory processing and may provide valuable information for future studies of early indicators of SPD.
The HR group demonstrated more atypical positions than the LR group. Some of the children in this study demonstrated an asymmetrical tonic neck reflex (ATNR) posture beyond age 6 mo, the time by which it should have been integrated. The presence of ATNR may be predictive of neurodevelopmental problems and thus a possible indicator of concern (Hamer & Hadders-Algra, 2016). These preliminary findings regarding atypical movement are consistent with those of ASD infant studies, which found atypical movement patterns in infants later diagnosed with ASD (Phagava et al., 2008; Teitelbaum et al., 1998). The presence of atypical and restricted movement patterns in infants may be an early indicator of a developmental disruption and has been shown to predict not only later motor impairments but also later cognitive and behavioral dysfunction (Hadders-Algra, 2008; Hamer & Hadders-Algra, 2016).
Sensory processing differences were also found between the HR group and LR group on the Toddler Sensory Profile–2. The HR group in this study engaged in fewer sensory-seeking behaviors than the LR group. In a previous study, adolescent siblings of children with ASD scored significantly lower than controls on sensation seeking (De la Marche et al., 2012) but significantly higher than their siblings diagnosed with ASD, suggesting that atypical sensory processing may be an endophenotype or intermediate phenotype. Studies of infants at risk ASD found HR infants to be less sensory seeking than the LR group (Ben-Sasson et al., 2009; Mulligan & White, 2012); the authors of these studies have proposed that HR infants may have less capacity to explore their environment. Similarly, the HR group of the current study had significantly more atypical positions and, although not significant, had lower scores on the Motor Organization and the Mobility Subtests. We propose that infants in this study had subtle motor difficulties, adversely affecting their ability to interact with and explore the environment. An infant’s awareness of the environment and skills to actively explore it affect participation as well as social and cognitive development (de Groot, 2000; DeSantis et al., 2011).
Sensorimotor dysfunction may be the earliest clinical indicator of abnormal connectivity and brain network aberrations. This dysfunction has been observed across many disorders before they have been fully manifested clinically, such as impairments in motor coordination and sensory integration and regulation (Levit-Binnun et al., 2013). Estes et al. (2015) suggested that sensorimotor differences at age 6 mo are associated with ASD at age 24 mo and precede the social communication, cognitive, and adaptive deficits that unfold later in infants diagnosed with ASD. In the current study, we found sensorimotor (atypical positions, atypical sensory processing, and motor difficulties) as well as cognitive and language differences between our HR and LR groups for SPD. Because sensory and motor processes share indirect neurological circuitry of later cognitive systems (Denckla, 2005; DeSantis et al., 2011), it is not surprising that our HR group also had lower scores in language and cognition than the LR group.
Limitations and Future Directions
The current study had several limitations. We used a small convenience sample, which was homogeneous and limited to a group of primarily well-educated White families. We also combined a wide range of ages (4–24 mo), and as such, we are not able to comment on different stages of development. Although parents in our LR group reported that their infants were typically developing, two parents expressed concerns about their children’s development during the evaluation. These children had lower scores in some developmental domains. It is not uncommon for minor delays to be overrepresented in infants who are enrolled in developmental studies such as this one (Hadders-Algra et al., 2010).
Future studies should include larger samples and longitudinal data to look at developmental trajectories of sensory processing and other domains of development explored in this pilot study. This information would be useful to determine the relation between early developmental skills and later diagnoses. Inclusion of a comparison group of children at risk for ASD would allow investigation of differences in the trajectories of early developmental skills among HR groups for other diagnoses. Direct observational assessments by trained researchers are recommended to assess quality of skills not routinely observed by parents (DeSantis et al., 2011), which may not be captured by parent-report measures. With the exception of the Toddler Sensory Profile–2, the parent-report measures used in this study tended not to show statistically significant group differences.
Early identification of SPD can lead to earlier intervention to minimize the manifestation of symptoms. In a study on early intervention using sensory-enriched environments, young children with developmental delays and SPD showed improvements on the Bayley–III in all domains except fine motor skills (Blanche et al., 2016). This finding suggests that future studies of at-risk infants are warranted because children with SPD can benefit from early intervention.
Implications for Occupational Therapy Practice
Results of the current study have important clinical implications for occupational therapy practitioners:
Early motor, sensory processing, language, and cognitive skills should be monitored in HR infants.
Early identification can facilitate early intervention to minimize disability and improve participation in early occupations for infants at high risk for SPD.
Measures of cognition, language, atypical positions, motor planning, and sensory processing show potential for use in future studies of LR and HR infants.
Conclusion
The results of this study provide preliminary evidence that babies ages 4–24 months at high risk for SPD can be differentiated from those at low risk. Potential risk factors were identified in infants who had an older sibling with SPD to assist health care professionals in identifying problems early. Motor, sensory processing, language, and cognitive abilities should be monitored in HR infants to facilitate early identification and intervention. These measures can also be included in future longitudinal studies to characterize the early signs of SPD. Earlier identification could lead to earlier intervention to optimize occupational outcomes and participation.
Footnotes
Acknowledgments
We thank the children and families who participated in this research; the research assistants, Andrea Valdez and Shannon Hampton, who helped to collect and manage the data; and our research students, Stacey Lau, Nishaka Shrestha, Katrina Begtson, Jess Wood, and Alex Lucas, for their help. We also thank Dr. Timothy A. Reistetter for editing and statistical advice. Finally, we thank the Wallace Research Foundation for their generous support of this study.