University of Southern California Dissertations and Theses

Developmental trajectories of sensory patterns in young children with and without autism spectrum disorder: a longitudinal population-based study from infancy to school age

(USC Thesis Other)

Developmental trajectories of sensory patterns in young children with and without autism spectrum disorder: a longitudinal population-based study from infancy to school age

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content i

DEVELOPMENTAL TRAJECTORIES OF SENSORY PATTERNS IN YOUNG CHILDREN
WITH AND WITHOUT AUTISM SPECTRUM DISORDER: A LONGITUDINAL
POPULATION-BASED STUDY FROM INFANCY TO SCHOOL AGE
by
Yun-Ju (Claire) Chen

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(OCCUPATIONAL SCIENCE)

August 2021

Copyright 2021 Yun-Ju Chen

ii

ACKNOWLEDGEMENTS
First and foremost, I would like to express the deepest gratitude to my committee and
mentor, Dr. Grace T. Baranek, for her expert guidance and warm support throughout my doctoral
journey. I have learned so much from her not only about how to design and conduct research but
also about being a leader and collaborative member of a research team. I believe that her
mentorship will continue to influence my scholarly work for years to come. I also would like to
extend my appreciation to the other members of my dissertation committee, Dr. Mary C. Lawlor,
Dr. John Sideris, and Dr. Linda R. Watson. They have all kindly provided invaluable advice in
improving various aspects of this dissertation research.
It also has been such a pleasure and privilege for me to work with many intelligent and
kind scholars across sites at the University of Southern California and the University of North
Carolina at Chapel Hill. I am honored to be a member of the PEARLS and INSPIRE Lab teams,
growing and learning together with the investigators and other team members in a very
supportive environment. I owe a special debt of thanks to the families who participated in the
NCCDS study for over six years for their faithful dedication. Also, my doctoral study and this
research would not be possible without the generous support from many sources (as listed in the
section of Funding Sources). Last but not least, I would like to thank my family and friends in
Taiwan, Los Angeles, North Carolina, and other parts of the world, for their support over these
years with relentless encouragement and constant prayers.
Acknowledgement of Funding Sources: The current studies were built upon the North
Carolina Child Development Survey Cohort 2 (NCCDS-2) project, which was supported by the
Autism Speaks Foundation (Grant #5946) and partial funding from The Ireland Family
Foundation. The school-age follow-up was supported by the Organization for Autism Research

iii

Graduate Research Grant (Grant #2019G3) awarded to Y.C., the Program for Early Autism
Research, Leadership, and Service (PEARLS) funds at the University of North Carolina at
Chapel Hill, as well as Dr. Grace Baranek’s research funds. All costs associated with this
doctoral degree, including tuitions, training, and conference travel, were generously supported by
the US-Taiwan Fulbright Graduate Study Grants (2016-2018) and the research assistantship
stipend from the USC Mrs. T.H. Chan Division of Occupational Science and Occupational
Therapy (2017-2021).

iv

TABLE OF CONTENTS
Acknowledgements………………………………………………………………………….….... ii
List of Tables …...................…………………………………………………………………..… vi
List of Figures ………………………………………………………………………………...... vii
Abbreviations………………………………………………....................................................... viii
Abstract……………………………………..............................……………………………….... ix
Chapter 1: General Introduction
Literature Review………………............………………............………………................ 1
Present Studies………………............………………............……….............................. 15
Chapter 2: Developmental Trajectories of Sensory Patterns from Infancy to School Age in ASD
and Non-ASD Populations (Study 1)
Abstract…………............………………............……………............………………..... 18
Introduction…………............………………............……………...............………….... 20
Method…………............………………............……………............…………............... 23
Results…………............………………............……………............…………............... 27
Discussion…………............………………............……………............………….......... 30
Conclusion…………............………………............……………............…………......... 34
References…………............………………............……………............………...…...... 36
Tables and Figures…………............………………............……………......................... 44
Chapter 3: Early Developmental Profiles of Sensory Patterns in a Community Sample:
Associations with School-age Outcomes (Study 2)
Abstract…………............………………............……………............……………...….. 50
Introduction…………............………………............……………............…………........ 52
Method…………............………………............……………............…………............... 56
Results…………............………………............……………............…………............... 62
Discussion…………............………………............……………............…………......... 65
Conclusion…………............………………............……………............…………......... 70

v

References…………............………………............……………............…………......... 71
Tables and Figures…………............………………............……………........................ 80
Chapter 4: Impacts of Early Sensory Trajectories on School-age Outcomes: A Longitudinal Birth
Cohort Study (Study 3)
Abstract…………............………………............……………............…………..….….. 86
Introduction…………............………………............……………............……….…....... 88
Method…………............………………............……………............…………............... 92
Results…………............………………............……………............…………................97
Discussion…………............………………............……………............…………....... 100
Conclusion…………............………………............……………............…………....... 107
References…………............………………............……………............……….….......109
Tables and Figures…………............………………............……………....................... 120

Chapter 5: General Discussion.......……..................……………….................………….….… 134
References……..................………………..............……..................………………................. 144
Appendices
Appendix 1……………..............…………………..............………………….............. 160
Appendix 2……………..............…………………..............………………….............. 162
Appendix 3……………..............…………………..............………………….............. 167

vi

LIST OF TABLES
Chapter 1 (General Introduction)
Table 1.1. Common terms in the literature used to describe sensory patterns……………….....…2
Chapter 2 (Study 1)
Table 2.1. Demographics and descriptive statistics of measures by clinical outcome group……45
Table 2.2. FYI/SEQ items used for deriving sensory construct scores……………….....……… 45
Table 2.3. Longitudinal invariance testing results……………………………………………….46
Table 2.4. Latent growth factor estimates by clinical outcome group………………………..….48
Chapter 3 (Study 2)
Table 3.1. Previous evidence on sensory-based subtypes………………………………….…….80
Table 3.2. Sample demographics…………………………………………………………...…… 83
Table 3.3. Fit statistics for multivariate LCGA……………………………………………….… 83
Table 3.4. Age-6 distal outcome by trajectory class (subset sample) …………..……….……... 85
Chapter 4 (Study 3)
Table 4.1. Summary of findings on the associations between sensory patterns and other domains
of behavior…………………………………………………………………………………...… 121
Table 4.2. Descriptive statistics of school-age outcome data for the subsample……………….126
Table 4.3. Intercorrelations of school-age outcome variables………………….…………....… 127
Table 4.4. Total effects of individual growth (intercepts and slopes) of sensory patterns on
school-age outcomes………………….………………….………………….…………….…... 129
Table 4.5. Significant indirect effects………………….…………….....….……………….…. 131

vii

LIST OF FIGURES
Chapter 2 (Study 1)
Figure 2.1. Flow chart of diagnostic classification based on parent reports………………..……44
Figure 2.2. Latent growth factor associations……………………………………………………46
Figure 2.3. Trajectories of sensory patterns by child’s sex, race, and parent education levels….47
Figure 2.4. Trajectories of sensory patterns by clinical outcome group…………………………48
Figure 2.5. Associations between individual slopes of HYPER/HYPO & SRS T-scores at T3... 49
Chapter 3 (Study 2)
Figure 3.1. Latent trajectory classes of sensory patterns (5 classes) …………………………… 84
Chapter 4 (Study 3)
Figure 4.1. Hypothesized model……………….…………………………….……………....... 128
Figure 4.2. Significant prediction effects of latent growth factors on school-age outcomes......132

viii

ABBREVIATIONS
The following abbreviations were used throughout the manuscript and are provided here for
reference.
ASD Autism spectrum disorder
DCQ Developmental Concerns Questionnaire
DD Developmental delay
DIF Differential item functioning
FYI First Years Inventory
HYPER Hyperresponsiveness
HYPO Hyporesponsiveness
IRT Item response theory
LCGA Latent class growth analysis
LGCM Latent growth curve modeling
ND No diagnosis or concerns
OD Other diagnosis or concerns
PEM-CY Participation and Environment Measure-Children and Youth
RRBs Restrictive and repetitive behaviors
SEQ Sensory Experiences Questionnaire
SIRS Sensory interests, repetitions, and seeking behaviors
SPD Sensory processing disorder
SRS Social Responsiveness Scale
TD Typical development
VABS Vineland Adaptive Behavior Scales

ix

ABSTRACT
Past research has demonstrated the ubiquitous presence and early emergence of sensory
patterns, including hyperresponsiveness (HYPER), hyporesponsiveness (HYPO), and sensory
interests, repetitions, and seeking behaviors (SIRS) in children with autism spectrum disorder
(ASD), as well as the potential role of sensory processing as a key building block for higher-
level social and cognitive functions. Whereas previous findings highlighted the cross-sectional
differences of sensory patterns across age and diagnostic groups, it remains unclear how the
developmental trajectories of sensory patterns in ASD differ from those with other diagnostic
outcomes. More evidence is needed to understand the developmental and heterogeneous nature
of sensory patterns in young children with ASD among a general population, as well as how they
are associated with broader developmental outcomes, such as adaptive/maladaptive behaviors
and participation, in order to evaluate the contribution of sensory patterns to the early
identification, diagnosis and prognosis of children with ASD and/or other developmental
challenges.
To address these empirical gaps, this dissertation aimed to 1) model developmental
trajectories of sensory patterns from infancy to school age in a community sample and explore
the demographic and clinical factors that may account for their variability; 2) identify
developmental trajectory subtypes of sensory patterns as associated with school-age outcomes,
and 3) understand the specific longitudinal impact of sensory patterns on school-age outcomes.
We followed up on a longitudinal cohort of children (N=1,517) from a large community sample
whose caregivers completed surveys regarding their child’s sensory patterns and other
developmental concerns at three time-points: infancy (Time 1: 6-19 months), pre-school years
(Time 2: 3-4 years), and school years (Time 3: 6-7 years). At Time 3, we collected additional

x

outcome data on autism symptoms, adaptive/maladaptive behaviors, and activity participation in
a subsample (N=389) that included families who had reported any diagnoses or concerns
(N=312) and those who had not reported any diagnoses or concerns at previous time-points
(N=77).
In Study 1, we conducted multivariate latent growth curve modeling (LGCM) with
demographic covariates to estimate sensory trajectories from Time 1 to Time 3 with the full
sample (N=1,517) and the results revealed highly variable longitudinal patterns across the three
sensory patterns. Such variability could be partially explained by demographic characteristics
(i.e., child’s sex, race, and parent education) and clinical outcome status (ASD and non-ASD
conditions). Particularly, the slopes of HYPER and HYPO were better able to differentiate ASD
from other conditions, including non-ASD children with sensory issues. Parent education
accounted for more of the variations in trajectories of children’s sensory responsiveness than
child’s sex and race. Furthermore, the latent growth factors of the three sensory patterns were
associated with each other, indicating their co-occurrence and co-development over time. These
findings from Study 1 support the potential utility of longitudinal sensory patterns from infancy
for early detection of ASD and the pivotal role of family-tailored approaches to address young
children’s sensory challenges.
To further address the variability that could not be explained by the known factors (i.e.,
demographics and clinical outcome status) in Study 1, we performed latent class growth analysis
(LCGA) to identify the sensory trajectory subtypes from infancy to school age among the full
sample in Study 2. We also examined how children classified to these trajectory subtypes differ
in their demographic characteristics, clinical outcome status (ASD and non-ASD conditions),
and school-age outcomes, including autism symptoms, adaptive/maladaptive behaviors, and

xi

activity participation. As a result, we identified five distinct subtypes that vary in latent growth
factors across sensory patterns. Particularly, there was a subtype (3% of the sample)
characterized by elevated and worsening sensory patterns and highly associated with ASD and
poor school-age outcomes. The other four subtypes were characterized by low to moderate levels
of sensory patterns and generally stable/improving trajectories and were associated with
strengths and weaknesses in school-age outcomes. These results indicated that profiling children
based on their early sensory trajectories may help to identify children who are more likely to
experience developmental challenges at school age and may thus introduce opportunities for
early intervention.
In Study 3, we examined the specific impacts of sensory trajectories on each school-age
outcome variable. Multivariate LGCM was performed with demographic covariates and latent
growth factor regressions, and school-age outcome variables were included in the model by
being regressed onto the latent growth factor of sensory patterns. Overall, the change rate of
HYPER was the most significant predictor of school-age outcomes. The initial levels of sensory
patterns had indirect effects on some distal outcomes via the change rates of HYPER and HYPO.
Differential impacts of HYPER and HYPO were observed on maladaptive behavior: HYPER
was more associated with internalizing behavior while HYPO was more with externalizing
behavior. Also, the variations in autism symptoms and maladaptive behavior were explained to
larger extents by sensory trajectories as compared to other behavioral domains. These results
indicate that early sensory challenges may have cascading effects on other domains of behavior.
Thus, sensory responsiveness during this period may be an important target for early intervention
towards more optimal outcomes.
Overall, the findings of these three studies enhance our understanding of the

xii

developmental and heterogeneous nature of sensory patterns from infancy to school age in a
large community sample with various clinical characteristics, including ASD and non-ASD
conditions. The observed associations between early sensory trajectories and later school-age
outcomes may indicate the critical role of personalized intervention or services for children with
sensory challenges and their families across early development towards optimal outcomes.

1

CHAPTER 1
General Introduction
Literature Review
Sensory Patterns in Young Children with ASD
Since Kanner’s (1943) first description of autism as “reacting to loud noises or moving
objects with horror or panic”, atypical sensory behaviors have been widely reported in children
with ASD. Recent estimates of the prevalence of sensory patterns in children with ASD ranged
from 66 to 94% depending upon the methods used (Baranek et al., 2006; Leekam et al., 2007;
Klintwall et al., 2011; Zachor & Ben-Itzchak, 2014). The clinical significance of sensory patterns
has been reflected in the recent revision of the Diagnostic and Statistical Manual of Mental
Disorders (DSM-5; American Psychiatric Association, 2013) by including “hyper- or hypo-
reactivity to sensory input or unusual interest in sensory aspects of the environment” as one of
the core elements under Criterion B: restricted, repetitive patterns of behavior, interests, or
activities. It has been suggested that sensory features are primary characteristics of the
neurobiological mechanism of ASD (Robertson & Simmons, 2013), meriting further
investigations to understand their impact on other domains of behavior that characterize ASD.
As far, most of the previous literature related to sensory features in ASD has used parent-
report measures to characterize observable reaction patterns that impact a child in daily life, such
as the Sensory Profile (SP; Dunn, 1999b), the Short Sensory Profile (SSP; Dunn, 1999a), the
Sensory Experiences Questionnaire (SEQ; Baranek, 1999a), and the Sensory Processing Measure
(SPM; Parham, Ecker, & Kuhaneck, 2010). These instruments vary in the conceptual models for
measure development, environments assessed, target age range, and the availability of ASD-
1

2

specific normative data. Given these variations, the terminology and characterization of sensory
patterns across measures and studies have been inconsistent (Schaaf & Lane, 2015; Cascio et al.,
2016). In the current study, the following terms will be used to describe the three major sensory
patterns (Baranek et al., 2006): (1) hyperresponsiveness (HYPER), an exaggerated or avoidant
response to sensory stimuli; (2) hyporesponsiveness (HYPO), a lack of or delayed response to
sensory stimuli; (3) sensory interests, repetitions, and seeking behaviors (SIRS), a fascination
with or craving of sensory stimulation that is intense and may be repetitive in nature. Table 1.1
presents similar terms commonly used in the literature to describe these sensory patterns.
Table 1.1. Common terms in the literature used to describe sensory patterns

Although these sensory patterns are not specific to ASD, as they have been observed by
several studies to be present in children with other developmental disabilities (Wiggins et al.,
2009; Watson et al., 2011; Lane, Reynolds, & Dumenci, 2012), parents of children with ASD
have reported differences more frequently in how their children respond to sensory stimuli
(Rogers & Ozonoff, 2005; Baranek et al., 2006, 2007). HYPO was found to be the most
prevalent and more unique sensory pattern in ASD (Ben-Sasson, 2009b; Baranek et al., 2013),
Sensory Patterns Synonyms
Hyperresponsiveness
(HYPER)
Hyper-reactivity (DSM-5), sensory over-sensitivity/responsivity (Miller
et al., 2007; Ben-Sasson et al., 2009a; Zachor & Ben-Itzchak, 2014),
sensory sensitivity (Dunn, 1997), sensory defensiveness (Pfeiffer &
Kinnealey, 2003)
Hyporesponsiveness
(HYPO)
Hypo-reactivity (DSM-5), sensory under-sensitivity/responsivity (Miller
et al., 2007; Ben-Sasson et al., 2008; Zachor & Ben-Itzchak, 2014), hypo-
sensitivity (Robertson & Simmons, 2013), low registration (Dunn, 1997)
Sensory interests,
repetitions, and seeking
behaviors (SIRS)
Unusual sensory interest (DSM-5), sensory seeking (Miller et al. 2007;
Boyd et al. 2010; Robertson & Simmons 2013)

3

and seemed to emerge early at infancy, and also as a precursor of SIRS (Freuler et al., 2012;
Baranek et al., 2013). Different sensory patterns, such as HYPER and HYPO, may co-occur in
the same individual, and coexisting patterns have been reported to be more prevalent in young
children with ASD than those with other diagnoses (Baranek et al., 2006; Ausderau et al.,
2014b). These sensory patterns have been observed in children with ASD across modalities, such
as visual (Dunn, 1999b; Simmons et al., 2009), auditory (e.g., Tomchek & Dunn, 2007;
Reynolds & Lane, 2008), tactile (Baranek, Foster, & Berkson, 1997; Foss-Feig, Heacock, &
Cascio, 2012), taste/smell (Ledford & Gast, 2009; Cermak, Curtin, & Bandini, 2010),
proprioceptive/ vestibular (Gepner & Mestre, 2002; Kern et al., 2007b) and multisensory stimuli
(Iarocci & McDonald, 2006). In addition to parent reports, atypical sensory processing patterns
were also found at the neurophysiological and psychophysical levels across modalities (Marco et
al., 2011; Baum et al., 2015). Moreover, each individual with ASD can show distinct sensory
patterns in response to different sensory modalities, under different contexts, and across different
stages of life (Ausderau et al., 2014a). Given such variable presentations of atypical sensory
patterns, it is important to understand the developmental and heterogeneous nature of sensory
patterns in young children with ASD, as well as how they are associated with other ASD
symptoms and domains of functioning, in order to evaluate the potential contribution of sensory
patterns to the identification, diagnosis and developmental course of ASD.
The Heterogeneous Nature of Sensory Patterns in ASD
While there has been increasing evidence on differences in sensory patterns between
children with ASD and other developmental conditions, great variability was also observed in
sensory patterns within diagnostic groups, particularly in ASD (Baranek et al. 2014; Uljarević et
al., 2017). Many of the participants in previous studies showed behaviors reflecting more than

4

one type of sensory pattern and often in more than one modality. Thus, researchers have
endeavored to parse out subtypes of children with ASD based on their sensory patterns using
several analytic approaches. Using model-based clustering methods, Lane et al. (2014) reported
four sensory subtypes in children with ASD aged 2 to10 years, characterized as sensory adaptive
(typical), taste/smell sensitive, postural inattentive, and generalized sensory difference (high
levels of features across all sensory domains). HYPO and SIRS were observed across children
with ASD who showed clinically significant sensory patterns. Ben-Sasson et al. (2008) used
hierarchical clustering analysis to identify three subtypes for infants aged 18 to 33 months
diagnosed with ASD: low frequency, high frequency of overall sensory features, and mixed
patterns (high frequency of HYPER and HYPO, but low frequency of sensory seeking). A large-
scale study using latent profile analysis identified four subtypes among children with ASD aged
2 to 12 years: low, sensitive-distressed (high levels of HYPER and enhanced perception, but low
levels of HYPO and SIRS), attenuated-preoccupied (high levels of HYPO and SIRS, but low
levels of HYPO and enhanced perception), and extreme-mixed features across all patterns
(Ausderau et al.; 2014a). Moreover, a previous study subtyped children with ASD on sensory
patterns in combination with other non-sensory behaviors, such as autism symptoms and
attention/memory, thus yielding four complex subtypes. For example, the most low-functioning
subtype they identified was characterized by elevated HYPO and SIRS, as well as social-
communication deficits (Liss et al., 2006). Another study subtyped children with ASD aged 3 to
6 years based on both sensory patterns and developmental performance, including
communication skills, motor performance, and adaptive behavior (Tomchek et al., 2018). They
identified four subtypes that varied by age as well as differed by degree and quality of sensory
patterns and developmental skills. For instance, the largest “sensorimotor” subgroup,

5

characterized by HYPER to taste/smell, sensory seeking, and HYPO, had the youngest average
age and the lowest developmental scores.
In general, the emerging evidence of sensory-based subtyping has revealed that the
presentation of sensory patterns in ASD can be better characterized as a set of distinct sensory or
multifactorial subtypes beyond group-level descriptions (i.e., ASD as a single category).
Furthermore, researchers have found that children classified into different subtypes may have
various behavioral manifestations in anxiety (Green et al., 2012; Uljarević et al., 2016), affective
symptoms (Ben-Sasson et al., 2008), symptom severity (Ausderau et al., 2014b), adaptive
behavior (Lane et al., 2010; Ausderau et al., 2016), and activity participation (Little et al., 2015).
These findings all reflect the fact that various patterns of sensory responsiveness are not mutually
exclusive, and individuals with ASD could exhibit behaviors indexed by one or more sensory
patterns that may vary across modalities, contexts, and developmental stages, while interacting
with other individual and environmental factors.
The Developmental Nature of Sensory Patterns in ASD
The three core sensory patterns (i.e., HYPER, HYPO, and SIRS) have been observed in
retrospective reports of children with ASD (Baranek, 1999a; Dawson et al., 2000; Freuler et al.,
2012), as well as in prospective studies with high-risk infant siblings as early as 2 to 6 months of
age, potentially preceding many of the social communication symptoms associated with ASD
(Zwaigenbaum et al., 2005; Bryson et al., 2007; Deconinck, Soncarrieu, & Dan, 2013). As
discussed above, sensory patterns have been considered as potential early markers of ASD given
the early emergence of sensory patterns as precursors to developmental milestones in social
cognition (Robertson & Baron-Cohen, 2017). However, the manifestation of sensory differences
in ASD seems to vary across development, particularly during early childhood. As concluded in

6

a meta-analysis of studies before 2008 (Ben-Sasson et al., 2009b), HYPER and SIRS increased
from infancy until reaching a peak at 6 to 9 years of age, and gradually decreased thereafter.
Despite its early emergence as a differentiating feature between ASD versus non-ASD, HYPO
seemed to decrease as a function of mental age in studies with preschool-age and school-age
children with ASD (Baranek, Foster, & Berkson, 1997; Baranek et al., 2006; Kern et al, 2007a,
2007c). In contrast, HYPER may become more prominent as children transition from preschool
to school stage (Talay-Ongan & Wood, 2000; Liss et al., 2006). It is possible that external
sensory features, such as HYPER, become apparent as the environmental demands from school
settings increase, and children are better able to verbalize their negative sensory experiences to
caregivers. In contrast, the amelioration of sensory patterns observed later on may be related to
children acquiring coping strategies or the maturation of orienting/executive attention networks
(Kern et al., 2007c; Baranek et al., 2014).
These developmentally sensitive patterns identified by previous cross-sectional and meta-
analytic studies are partially supported by a recent longitudinal study that followed up a large
community sample of children with ASD aged 2 to 12 years at two time-points separated by
around 3 years (Baranek et al., 2019). They found that the group mean scores of HYPO and
SIRS declined over time, while intra-individual differences remained stable overall. A recent
longitudinal study with high-risk infant siblings indicated that HYPO (differentially for the
affected high-risk sibs) and HYPER became more pronounced between 12 and 24 months of age
(Wolff et al., 2019), while other studies did not find significant changes in overall sensory scores
over time in children with ASD aged 2 to 8 years (McCormick et al., 2016; Perez-Repetto et al.,
2017).

7

Current evidence on developmental changes of sensory patterns (and other behavioral
domains) is mostly based on the investigations of group-based instead of individual-based
trajectories, as well as arbitrary group comparison (e.g., ASD vs. TD) of patterns, and thus might
fail to detect individual deviations from typical trajectories over a continuum of behavioral
manifestation across clinical conditions. For instance, if HYPER was found to increase during
early childhood in children with ASD, would a similar pattern also be found in those with other
developmental challenges, such as developmental delay and/or ADHD? That is, is this pattern
unique to ASD? Would a subset of children with ASD show a distinct trajectory that is
associated with better developmental outcomes despite the elevated autism severity? When are
the potential points of divergence of trajectories that lead to different outcomes? These questions
that have not been clarified in the previous research highlighted the need to move from
heterogeneity to “chronogeneity” for better understanding the complexity of ASD from a
developmental perspective (Georgiades, Bishop, & Frazier, 2017). Particularly, the
developmental changes of sensory patterns are understudied compared to other autism traits such
as social-communication deficits, and thus call for a more systematic probing of the topic.
Sensory Patterns as Predictors of ASD
Previous research has linked specific sensory patterns to other core features of ASD such
as social communication deficits (Lane et al., 2010; Hilton et al., 2010; Watson et al., 2011) and
restrictive and repetitive behaviors (RRBs) (Baker et al., 2008; Chen, Rodgers, & McConachie,
2009; Boyd et al., 2010; Wolff et al., 2019). It has been hypothesized that altered sensory
processing during infancy may directly contribute to the accumulation of later-developed
functioning such as social communication (Ronconi, Molteni, & Casartelli, 2016; Thye et al.,
2018). This perspective has been supported by evidence that sensory differences have

8

precedence for, and may produce cascading effects on higher-order functions such as social-
emotional behaviors (Ben-Sasson et al., 2009a; Green et al., 2012), communication skills
(Watson et al., 2011; Baranek et al., 2013), and attention (Sabatos-Devito et al., 2016),
particularly in the early stages of development. However, most of the existing evidence is
correlational rather than causal in nature, and thus requires further validation with longitudinal
studies. Recently, longitudinal studies demonstrated that SIRS by 24 months predicts later social
difficulties at 36 months of age (Baranek et al., 2018; Damiano-Goodwin et al., 2018). These
findings support the potential role of sensory processing as a key building block for higher-level
social and cognitive functions, which are critical domains for ASD diagnosis and prognosis.
It is not until recently that there have been a few prospective studies using sensory
measures to predict a later ASD diagnosis. Germani and colleagues (2014) investigated the
predictive associations between the Infant/Toddler Sensory Profile (ITSP; Dunn 2002) scores at
24 months and 3-year diagnostic outcomes in a sample of high-risk infant siblings. The analyses
showed that high-risk infants diagnosed with ASD had more difficulties in auditory processing,
and significantly low registration, which is related to HYPO. Similarly, another study also found
that high-risk infants later diagnosed with ASD showed significantly low registration and more
difficulties in the tactile and vestibular domains, as measured by the ITSP at an early age (Van
Etten et al., 2017). Using the Sensory Processing Assessment (SPA; Baranek, 1999b), Baranek
and colleagues (2018) examined how SIRS and social orienting in infants at-risk for ASD at 13-
15 months and 20-24 months predict later diagnostic outcomes at 3 to 5 years of age. As a result,
SIRS behaviors at 20-24 months, but not at 13-15 months, were able to predict later social
symptom severity, as mediated by social orienting. Another study with a similar design
examined how SIRS measured by the SPA, along with EEG frontal asymmetry in high-risk

9

infant siblings at 18 months predict a later ASD diagnosis. They reported that high-risk infants
with a later ASD diagnosis did show elevated SIRS at infancy (Damiano-Goodwin et al., 2018).
A recent study included the SSP as one of the diagnostic assessment tools, in addition to the
ADOS and Repetitive Behavior Scale-Revised (RBS-R), to examine the factors that contribute to
a diagnosis of ASD in children aged 2.5 to 10 years (Duvekot et al., 2017). They found that high
scores on the SSP significantly predicted an ASD diagnosis regardless of child’s sex.
Furthermore, a recent study reported that early HYPER and HYPO across 14 to 23 months
predicted later autism severity at 3 to 5 years among children flagged as high-risk for ASD at
infancy (Grzadzinski et al., 2020). These findings indicate that sensory measures might be useful
to predict a later diagnosis of ASD. Further research is needed to validate the utility of sensory
measures in the early identification of ASD. Particularly, there are currently no longitudinal
studies of community samples investigating patterns of change specifically in sensory patterns
from infancy to school age that could be useful predictors of later diagnostic or developmental
outcomes.
Sensory Patterns and Adaptive/Maladaptive Behaviors in ASD
In the field of developmental science, a child’s functional outcome often refers to his or
her acquired adaptive skills. According to the American Association on Intellectual and
Developmental Disabilities (AAIDD), adaptive behavior is defined as, “the collection of
conceptual, social, and practical skills that have been learned by people in order to function in
their everyday lives, including communication, community use, functional academics,
home/school living, health and safety, leisure, self-care, self-direction, social, work, motor skills”
(Luckasson et al., 2002, p. 73). Another broadly accepted definition is proposed by Sparrow,
Cicchetti, and Balla (2016) as “the performance of daily activities required for personal and

10

social self-sufficiency across a variety of life situations including self-care, community mobility,
home maintenance, establishing and maintaining relationships, and communicating needs and
feelings.” The evaluation of degrees by which adaptive behavior is affected in children with
ASD is not only essential to understand their strengths and weaknesses, but also to establish
appropriate intervention plans (Bal et al., 2015).
Previous studies have documented associations between sensory features and adaptive
behavior in older children with ASD. Sensory patterns were associated with both strengths and
challenges in communication skills (Liss et al., 2006; Lane et al., 2010; Watson et al., 2011;
Tomchek, Little, & Dunn, 2015). Social impairments were found to be associated with HYPO
(Baranek et al., 2013) and atypical sensory modulation (Kern et al., 2007c, Hilton, Graver, &
LaVesser, 2007). In a recent longitudinal study, Williams and colleagues (2018) reported that
higher HYPO and HYPER in early childhood predicted lower overall adaptive behavior and
daily living skills in both children with ASD and developmental delay (DD). In contrast, another
longitudinal study found that sensory features were not predictive of adaptive behavior either as
a main effect or across time for young children aged 2 to 8 years with ASD, DD, or typical
development (TD) after controlling for intellectual ability and other autism symptoms
(McCormick et al., 2016). The authors argued that the non-significant results might be due to the
small sample size that did not allow them to examine potential subtypes among their sample.
This reflects the necessity of examining the association by subtypes, which should capture a
fuller range of variability instead of by the arbitrary diagnostic groups. This point was
underscored by subtyping studies of children aged 2 to 12 years that found significant
associations with specific sensory subtypes and adaptive/maladaptive behaviors as well as
increased caregiver stress (Lane et al., 2010; Ausderau et al., 2016). Thus, different subtypes of

11

children with ASD may have strengths or weaknesses when adapting to different environmental
demands, and thus would be benefited from more targeted interventions.
Sensory patterns were also found to be associated with maladaptive behavior. Children
with ASD who were more sensitive to environmental stimuli tended to display higher levels of
stress (Bakker & Moulding, 2012), anxiety (Kinnealey, Koenig, & Smith, 2011; Green et al.,
2012; Mazurek et al., 2013), emotional dysregulation (Ben-Sasson et al, 2009a) and inattention
(Dellapiazza et al., 2021). SIRS and HYPO were found to be associated with hyperactivity and
poor adaptability to contextual demands (Baker et al., 2008; Chuang et al., 2012). It is to be
noted that atypical sensory patterns may be a precursor of maladaptive behavior such as anxiety
symptoms, as demonstrated by a previous longitudinal investigation (Green et al., 2012). Also,
the associations with sensory patterns were found to be limited to maladaptive behavior instead
of adaptive behavior in some studies (O’Donnell et al., 2012; Nieto, Lopez, & Gandia, 2017).
Overall, existing evidence has supported the co-occurring associations between sensory patterns
and adaptive/maladaptive behaviors in ASD. However, it remains unclear whether sensory
patterns early in life predict later adaptive/maladaptive outcomes given the cross-sectional
designs adopted by most of the previous studies. The mixed results also call for the need to take
individual variability into account when examining these associations.
Sensory Patterns and Participation in ASD – What Children Actually Do in Their Daily Lives
Matters
The field of autism research has had an exclusive focus on better adaptive skills,
particularly social communication skills, to define optimal outcomes and often overlooks an
individual’s real-life functioning in certain tasks or jobs with or without support (Georgiades &
Kasari, 2018). Although current definitions of adaptive behavior have addressed its

12

developmental and transactive nature reflected by its dynamic relationship with different life
stages, contexts, and their various levels of demands, the measures widely used in the field of
psychology and education seem to have a disproportionate emphasis on what children can do
instead of what children actually do in their daily lives. Adaptive behavior, in and of itself, does
not fully explain levels of participation, which might be mediated or moderated by other factors
(Kileen et al., 2018). Kramer and colleagues (2012) interviewed professionals and parents of a
child with ASD regarding factors that influence their decisions when rating the child’s adaptive
behavior using an assessment of a child’s ability to perform activities required for self-
sufficiency at home and community settings. The parents noted great variability in their child’s
performance of activities across contexts that cannot be fully captured by measures of adaptive
behavior. For instance, some children may have the ability to execute certain activities without
understanding the purpose underlying them or appropriate levels of motivation. Professionals
also recognized the importance of a measure that taps behaviors in multiple environments with
multiple people. The authors pointed out that the completion of a certain activity may not
guarantee the meaningfulness of that activity to the child, as well as his or her level of
engagement in that activity. Thus, in order to obtain a comprehensive picture of a child’s
developmental outcomes as a result of individual capacities transacting with complex physical
and social contexts, it is necessary to assess children’s meaningful outcomes, which include their
actual participation in activities across various settings.
To achieve this goal, occupational science provides a critical theoretical basis for
occupational therapy and other disciplines to comprehensively understand what people do with
meaningfulness and purposefulness (Nelson, 1996). Occupational scientists perceive
environments as embedding challenges, difficulties, and demands but also as providing

13

affordances and opportunities (Yerxa, 2000). Occupations serve as a platform for individuals to
encounter these challenges and opportunities, and to develop adaptive responses for a specific
context at a given point in time in the real world (King et al., 2018). Furthermore, an individual’s
occupational engagement reflects both the process and outcome of developmental changes
(Coster, 1998), and being occupational is beyond acquiring universal actions or survival skills
(Wilcock, 2006). Thus, it is not possible to understand an individual’s functional status without
understanding the occupations in which he or she is participating. Regarding children’s
functional outcomes at school age, which is a life stage when children are given a variety of
choices of occupations while experiencing challenges and opportunities under different contexts,
it is important to assess children’s functional outcomes that include both adaptive behaviors and
participation in activities across various contexts. Such a comprehensive evaluation of school-
age outcomes would be a step toward the goal of understanding each child as an occupational
being while connecting occupational science and other disciplines in terms of communicating
meaningful outcomes in children with ASD or other developmental disabilities.
Existing evidence has demonstrated the impact of sensory patterns on participation in
daily activities and family occupations in children with ASD. Hochhauser and Engel-Yeger
(2010) found that atypical sensory patterns in children with ASD were associated with lower
participation, specifically in social, physical, and informal activities. Those with higher levels of
SIRS and HYPER tended to participate in a limited range of activities mostly at home with less
intensity. Zobel-Lachiusa and colleagues (2015) examined the relationship between mealtime
participation and sensory differences. They reported that children with ASD showed more
challenges in mealtime activities, which were associated with more pronounced sensory patterns.
Another study also demonstrated that children with more HYPER and sensory avoiding behavior

14

showed lower levels of competence in social activities and school performance (Reynolds et al.,
2011). Also, academic underachievement in children with ASD was found to be associated with
HYPO and SIRS (Ashburner, Ziviani, & Rodger, 2008). These findings reveal the potential
causal relationship between sensory differences and participation in daily activities, as well as
indicate the possibility of using this knowledge to improve participation and quality of life in
children with ASD. The evaluation of meaningful outcomes will ultimately help practitioners to
consider more individually tailored approaches to intervention for subgroups of children whom
sensory patterns impact most severely, as well as to help alleviate any cascading effects of
altered sensory processing on children’s long-term outcomes (Schauder & Bennetto, 2016).
Sensory Patterns, Adaptive and Participation Outcomes in Non-ASD Populations
Thus far, we have discussed the current evidence regarding the heterogeneous and
developmental nature of sensory patterns, as well as their associations with adaptive/maladaptive
behaviors and participation outcomes in children with ASD. However, these characteristics and
associations are not limited to ASD. A recent subtyping study on a large population-based
sample demonstrated the presence of sensory subtypes across ASD, ADHD, learning disabilities,
and TD (Little et al., 2017). Another subtyping study on children with SPD found three subtypes
which varied by levels of sensory patterns (Miller et al., 2017). These findings suggested that
elevated scores on sensory patterns are not unique to ASD but rather are reflections of children's
responses to environmental demands, which occur on a continuum for those with and without
developmental conditions (Baranek et al., 2014).
Past research has reported that atypical sensory patterns are associated with impaired
adaptive behavior, maladaptive behavior, and participation in daily activities in children with
non-ASD conditions. For instance, school-aged children with sensory processing disorder (SPD)

15

had poorer activity performance, lower level of enjoyment of the activity, and less frequency of
participation (Bar‐Shalita, Vatine, & Parush, 2008). Cosbey and colleagues (2010) found that
while children with SPD showed similar activity preference and use of free time, differences
were found in community-based social participation related to intensity and enjoyment of
involvement. In a subtyping study on children with SPD, all the subtypes showed challenges in
adaptive skills (Miller et al., 2017). Specifically, children with high HYPO/SIRS had the lowest
social skills, and those with high HYPER/SIRS showed more maladaptive behaviors, including
both internalizing and externalizing behaviors. Watson and colleagues (2011) reported that while
more HYPO and SIRS were associated with poorer language skills in both ASD and DD, SIRS
was associated with social-communicative symptom severity only in ASD. Children with ADHD
with sensory differences showed significantly lower preference to participate in leisure activities
than those without (Engel-Yeger & Ziv-On, 2011). Also, higher levels of HYPER and SIRS
were associated with sleep difficulties (Vasak et al. 2015) in TD infants. All these findings
indicate the importance of investigating the associations across a larger community sample that
may include various diagnostic conditions and levels of sensory challenges. Overall, a better
understanding of sensory subtypes across a general population and how sensory patterns
differentially predict later functional outcomes in children with ASD vs. non-ASD conditions
will eventually allow us to appropriately evaluate the utility of sensory patterns as a
distinguishing feature of ASD.
Present Studies
Given the lack of evidence on longitudinal variability of sensory patterns beyond ASD
populations as discussed above, this dissertation aimed to address the empirical gaps by
examining the heterogeneous trajectories of sensory patterns from infancy to school age among a

16

large community sample. We also investigated their associations with demographic and clinical
characteristics as well as school-age outcomes to understand the clinical relevance of the
identified heterogeneous sensory trajectories. The specific aims for each study are as follows.
Study 1: To model developmental trajectories of sensory patterns from infancy to school age in a
community sample
1) To establish comparable scores from changing sensory measures across the three time-
points
2) To model developmental trajectories of the three sensory patterns (i.e., HYPER, HYPO,
and SIRS) using parallel-process latent growth modeling
3) To examine the associations between latent growth factors (i.e., intercepts and slopes) of
sensory patterns
4) To examine the factors that may account for the variability of sensory trajectories,
including clinical outcome group status and demographic predictors
Study 2: To identify trajectory subtypes of sensory features from infancy to school age and their
associations with school-age outcomes in a community sample
1) To identify more homogeneous sensory trajectory subtypes, given the large variability of
trajectories observed in Study 1
2) To examine the demographic characteristics of the identified sensory trajectory subtypes
3) To examine to what extent the children who are classified to these subtypes differ in their
clinical characteristics and school-age outcomes, including autism symptoms,
adaptive/maladaptive behaviors, and activity participation

17

Study 3: To investigate the longitudinal impacts of sensory patterns on specific school-age
outcomes
1) To examine the direct and indirect effects of sensory growth factors along with
demographic covariates on school-age outcomes
2) To identify the school-age outcome domains which are more impacted by sensory
trajectories

18

CHAPTER 2
Developmental Trajectories of Sensory Patterns from Infancy to School Age in ASD and
Non-ASD Populations (Study 1)
Abstract
Background: Past research has demonstrated the early emergence of atypical sensory patterns in
infants at-risk for autism spectrum disorder (ASD) and their potential utility in predicting ASD
outcomes. The current study was the first attempt using person-based analytic approaches to
probe the longitudinal heterogeneity of sensory patterns, including hyperresponsiveness
(HYPER), hyporesponsiveness (HYPO), and sensory interests, repetitions, and seeking
behaviors (SIRS), from age 1 to 6 in a community sample.
Method: Caregivers of a birth cohort of infants (N=1,517) completed surveys regarding their
child’s sensory patterns and other developmental concerns during infancy (6-19 months), pre-
school years (3-4 years), and school years (6-7 years). Latent growth trajectories of sensory
patterns were estimated with child’s sex, race, and parent education levels as covariates. Latent
growth factor estimates were compared across children with various clinical outcomes.
Results: Models indicated highly variable trajectories of sensory patterns, with significant
decreases in HYPER/SIRS and increase in HYPO as mean trends. Non-white race and lower
parent education levels were associated with higher SIRS at baseline, while boys and higher
parent education levels were associated with higher HYPO at baseline. Higher parent education
levels predicted fewer increases in HYPO and HYPER over time. The trajectories by clinical
outcome group revealed differential patterns across children with ASD and non-ASD conditions.
Particularly, the slopes of HYPER and HYPO were highly associated with autism severity at
school age (r=.54-.55) and seemed to best differentiate ASD from other clinical outcomes,

19

including non-ASD children with sensory challenges.
Conclusion: Heterogeneous trajectories of sensory patterns during early childhood were
observed among a community sample, and this variability could be partially explained by child’s
sex, race, parent’s education, and long-term clinical outcome status. The utility of sensory
trajectories for differentiating ASD from non-ASD groups may have important implications for
early detection and intervention.
Keywords: Sensory; developmental trajectories; heterogeneity; community sample

20

Introduction
During recent decades, the concept of heterogeneity has greatly impacted our
understanding of autism spectrum disorder (ASD) and other neurodevelopmental disorders.
Despite the accumulated evidence of the potential utility of biomarkers for diagnosis, behavioral
markers such as parental concerns are still informative for early identification (Zwaigenbaum &
Penner, 2018). The behavioral manifestations of ASD have been found to vary from one
individual to another, as affected by factors such as cognitive abilities, comorbidity,
developmental changes, and demographics (Kim & Lord, 2013). Among these contributing
factors to the heterogeneity of ASD behavioral markers, developmental changes or trajectories
are particularly important to consider for the very young population, which is more impacted by
rapid brain development associated with risk manifestations of neurodevelopmental disorders
(LeBlanc & Fagiolini, 2011). As one step further from the cross-sectional investigations,
longitudinal subtyping studies have identified various developmental patterns in autism
symptoms (e.g., Landa et al., 2013; Kim et al., 2018), indicating the necessity of considering
developmental trajectories as autism phenotypes.
However, most of the previous evidence on developmental trajectories of behavioral
markers has been limited to the “core” domains of ASD such as social communication and
restricted/repetitive behaviors, despite the growing attention to other non-social ASD markers
such as sensory and motor behaviors (Sacrey et al., 2015; Canu et al., 2020). In the current study,
we will focus on the trajectories of one of the potential ASD markers: sensory patterns, defined
as atypical behavioral responses to sensory stimuli in the environment including
hyperresponsiveness (HYPER), hyporesponsiveness (HYPO), and sensory interests, repetitions,
and seeking behaviors (SIRS). These three major sensory patterns have been observed in

21

retrospective reports of children with ASD (Dawson et al., 2000; Freuler et al., 2012), as well as
in prospective studies with high-risk infant siblings within the first years of life, potentially
preceding many of the social-communication symptoms associated with ASD (Zwaigenbaum et
al., 2005; Deconinck, Soncarrieu, & Dan, 2013). Previous meta-analytic findings showed that
HYPER and SIRS increased from infancy until 6 to 9 years of age, and gradually decreased
thereafter (Ben-Sasson et al., 2009). Despite the early emergence of HYPO as a differentiating
feature between ASD and typical development (TD), its severity seemed to decrease as a
function of mental age in children with ASD (Kern et al, 2007; Baranek et al., 2013a). This may
be related to the acquired coping strategies or executive network maturation (Kern et al., 2007;
Baranek et al., 2013a). In contrast, HYPER tended to become more prominent as children
transition from preschool to school stage (Talay-Ongan & Wood, 2000; Liss et al., 2006). It was
suggested that external sensory symptoms, such as HYPER, may become apparent since the
environmental demands from school settings increase, and children are better able to verbalize
their sensory experiences to caregivers. A recent study followed up a large sample of children
with ASD aged 2 to 12 years at two time-points separated by around three years (Baranek et al.,
2019), reporting that the group means of HYPO and SIRS declined over time, while intra-
individual differences remained stable. Another longitudinal study on high-risk infant siblings
indicated that HYPO and HYPER became more pronounced between 12 and 24 months of age
(Wolff et al., 2019), while other studies did not find significant changes in overall sensory
symptoms over time in children with ASD aged 2 to 8 years (McCormick et al., 2016; Perez-
Repetto et al., 2017).
These findings, however, were often confounded by mixed factors such as child’s mental
or chronological age, IQ, measures used (parent-report or observational), study designs (cross-

22

sectional or longitudinal), and analytic approaches (variable-based or person-based). Particularly,
the previous evidence was mostly based on the “snapshots” of different time-points on varying
samples of children and thus might fail to detect individual deviations from typical trajectories.
Many questions are remaining to be answered. For instance, if HYPER was found to increase
during early childhood in children with ASD, could a similar pattern be also observed in those
with other developmental challenges, such as sensory processing disorder, ADHD, and
developmental delay? That is, is such a pattern unique to ASD? When could we observe a
divergence of trajectories that lead to different outcomes? These questions thus highlight the
need to move from heterogeneity to “chronogeneity” (i.e., heterogeneity as defined by variable
longitudinal patterns) to better understand ASD as a neurodevelopmental disorder (Geogiades et
al., 2017). Such information would ultimately contribute to the identification of more fine-
grained behavioral phenotypes and the intervening factors that promote the resilience of children
and their families during early childhood (Geogiades & Kasari, 2018). While there have been a
growing number of studies addressing this important issue, as summarized in a recent systematic
review of autism traits (Pender et al., 2020), there is currently no evidence on the chronogeneity
of sensory patterns in children with ASD and other conditions. Therefore, this study aimed to
explore the developmental variability of sensory patterns in a large community sample we
prospectively followed up from infancy, and to identify the trajectories toward ASD and other
clinical outcomes at school age. We used latent growth curve modeling (LGCM), an analytic
method that allows for estimating between-person differences in within-person patterns of
change over time (Curran, Obeidat, & Losardo, 2010), to address the empirical gaps.
The specific research questions were: 1) What are the developmental trajectories (i.e.,
stability or linear/non-linear change) of sensory patterns, including HYER, HYPO, and SIRS,

23

from infancy to school age across a community sample? 2) Are the latent growth factors (i.e.,
intercepts and slopes) of the three sensory patterns intercorrelated when they are estimated
simultaneously in a multivariate LGCM? 3) Are there significant between-person differences
among these trajectories? If so, to what extent could such variability be explained by
demographics (child’s sex, race, and parent education)? 4) Are the latent growth factor estimates
differ by clinical outcome groups and associated with autism severity at school age?
Method
Participants and Procedure
This study expanded upon a large cohort of families as a subset of a longitudinal research
study, the North Carolina Child Development Survey (NCCDS) project (see Appendix 1 for the
IRB approval). Families were initially ascertained from birth registries, with recruitment at the
age of 6-19 months (T1) for completing the First Years Inventory (FYI) version 3.1, and then
were re-contacted by a follow-up survey at 3-4 years (T2) to assess their developmental and
ASD outcomes, including the Developmental Concerns Questionnaire (DCQ) version 1.5,
Sensory Experiences Questionnaire (SEQ) version 2.1, and Social Responsiveness Scale (SRS)
2
nd
Edition. In the current study, the 2,236 families who returned their responses at T2 were
followed up via emails to complete new questionnaires regarding their child’s current diagnostic
status and/or any parent-report concerns (DCQ), as well as sensory patterns (SEQ) at 6-7 years of
age (T3, Phase-1). At the second phase of T3, which was about 5 months after Phase-1 responses
were returned, a subset of families (N=465) who reported any diagnosis/concerns at previous
time-points was again contacted via email invitations to complete the SRS. For data collection at
T3-Phase 1, participants who completed the survey could choose to be entered into a raffle at a
1/50 chance of winning a 50-dollar gift card. At T3-Phase 2, each participant with complete

24

responses was eligible to receive a 15-dollar gift card and a report of their child's results (see
Appendix 2 for the report template). We ended up receiving 1,519 responses at Phase-1 and 389
responses at Phase-2 (response rate=68-73%). After removing cases with incomplete data, 1,517
sets of longitudinal responses were retained for further analysis. Based on all the available
parent-report data collected at T2 and T3, children were classified into either ASD, sensory
processing disorder or sensory-relates issues (SPD), other non-ASD/non-SPD diagnoses or
concerns (OD), and no diagnosis or concerns (ND) group (see Figure 2.1 for the classification
criteria and flow). For analysis purposes, we combined children of non-white races (Black,
Asian, American Indian, Hawaiian, and multi-racial) into one group given their small numbers.
Three categories of parent education were created based on the reported mother’s and father’s
education at T1. The demographics by clinical outcome groups were shown in Table 2.1.
Measures
The FYI version 3.1 (FYIv3.1; Baranek et al., 2013b)
The FYIv3.1 is a newly revised parent-report measure (69 items) designed to identify
infants aged 6 to 16 months at risk for a later diagnosis of ASD, measuring the frequency of
behaviors across social communication, sensory-regulatory functions, and motor development
with a 5-point Likert scale. Its previous version FYIv2.0 (Reznick et al., 2007) has been well
validated and used in several studies (e.g., Turner-Brown et al., 2013; Ben-Sasson & Carter,
2012). For this study, fourteen items related to sensory patterns were extracted and decomposed
to constructs of HYPER, HYPO, and SIRS (see Table 2.2 for a list of items). The FYIv3.1 data
were collected at T1. Each participant randomly received either an A or B form (each form with
48 FYI items, 27 items in common) for survey completion.

25

The SEQ version 2.1 (SEQv2.1; Baranek, 1999)
The SEQv2.1 is a parent questionnaire (42 items) designed to measure the frequency of
responses to daily sensory experiences for children ages 1 to 12 years with a 5-point Likert scale
(higher scores indicate endorsement of more sensory features). It has excellent internal
consistency (alpha=.80) and test-retest reliability (ICC=.92) (Little et al., 2011), along with good
discriminative validity (Baranek et al, 2006). Fourteen items common to those in the FYIv3.1
(varied in item contents but intended to measure the same behavior) were extracted to establish
sensory construct scores, and the total score of all items was used for clinical outcome
classification. The SEQ data were collected at T2 and T3.
The DCQ version 1.5 (DCQv1.5; Reznick et al., 2005)
The DCQv1.5 is a parent-report measure with open-ended questions about whether a
parent or professional has been concerned about the child’s development and whether the child
has received any clinical diagnoses. It was used as one of the outcome measures in the validation
study of FYIv2.0 (Turner-Brown et al., 2013). The DCQ data at T2 and T3 were used for
outcome classification. Responses were coded to determine whether the child has had a
diagnosed developmental disability, including ASD, and/or any developmental concerns (see
Appendix 3 for the REDCap coding form).
The SRS-2 (Constantino & Gruber, 2012)
SRS-2 is a well-validated parent-report scale that measures deficits in social behavior
associated with ASD to determine levels of autism symptoms. The SRS data were collected at T2
(preschool-age version) and T3 (school-age version), and a cutoff of T-score ≥60 at either time-
point was used to determine ASD group membership.

26

Data Analyses
Establish Longitudinally Comparable Scores of Sensory Patterns
As the sensory items were extracted from two measures administered at different time-
points, it is important to ensure the comparability of scores (i.e., to meet the assumption of
longitudinal measurement invariance) before conducting the LGCM. We first tested whether
longitudinal invariance held at the configural level over time, followed by metric and scalar
invariance tests for each of the three sensory constructs (Millsap, 2012). Full-information
maximum likelihood (ML) estimation was used in Mplus 8.4 (Muthén & Muthén, 2018) to
account for the split-form missingness at T1. We ensured that there were less than 50% of the
items missing for each construct at each time-point so that the item-response-theory (IRT) scores
would be constructed upon enough items. Differences in fit indices between models were
evaluated to determine whether invariance held at different levels. A decrease in comparative fit
index (CFI) or Tucker–Lewis Index (TLI) >.01, or an increase in root-mean-square error of
approximation (RMSEA) >.01 indicates measurement non-invariance (Cheung & Rensvold,
2002). The purpose of invariance testing in the current study was to ensure that at least
configural invariance was met before constructing IRT scores that adjust for differential item
functioning (DIF). Next, DIF was evaluated to determine which non-DIF items could be used as
anchor items for scale equating, using the method developed by Stocking and Lord (1983). A
relatively conservative criterion (McFadden’s pseudo-R
2
change ≥0.02 between nested logistic
regression DIF models) was used to detect meaningful DIF (Paz et al., 2013; Wong et al., 2015).
It has been recommended to have at least one anchor item for every four non-common items to
avoid construct drift (Kolen & Brennan, 2004). By recalibrating group-specific item parameter
estimates (i.e., estimates specific to each time-point) for the DIF items, IRT trait scores of

27

HYPER, HYPO, and SIRS that accounted for DIF across time-points were generated (Choi,
Gibbons, & Crane, 2011). The DIF detection and IRT trait score computation were implemented
with R package lordif (Choi, Gibbons, & Crane, 2016).
Latent Growth Curve Modeling
We performed univariate LGCMs with robust ML estimation on HYPER, HYPO, and
SIRS separately to determine their functional forms over the three time-points. A CFI/TLI ≥.95
and an RMSEA <.08 indicate a good fit (Hu & Bentler, 1999). Upon confirming that the forms
of the three trajectories were consistent with satisfactory model fits, a multivariate LGCM was
conducted by estimating them simultaneously. Latent growth factor covariances were specified
to examine the interrelationships among intercepts and slopes across constructs. Next,
demographic variables (child’s sex, race, and parent education levels) were included as
predictors/time-invariant covariates to examine their effects on the growth factors (i.e., slopes
and intercepts) in a conditional LGCM. Lastly, we compared the latent growth factor estimates
by clinical outcome groups (ASD, SPD, OD, and ND) using one-way ANOVA with Bonferroni
corrections. All the LGCM analyses were performed in Mplus 8.4. Additionally, correlations
between the individual estimates of growth factors and SRS T-scores at T3 were examined for
the subsample (N=389) with complete T3 outcome data.
Results
Measurement Invariance Testing and DIF Adjustment
Longitudinal invariance testing on each of the constructs demonstrated invariance at least
at the configural level (see Table 2.3 for the model fits), indicating that the constructs to be
measured by the selected items held constant across time-points. To resolve the scalar non-

28

invariance, DIF detection process was implemented to identify the non-DIF items for scale
equating. Four out of fourteen items (2 in HYPER, 1 in HYPO, and 1 in SIRS; see Table 2.2)
were identified as anchor items (McFadden’s pseudo-R
2
change=.003 to .013) used for
producing DIF-adjusted scores. The descriptive statistics of the trait scores for HYPER, HYPO,
and SIRS at each time-point by clinical outcome group were shown in Table 2.1. There were
significant group differences in the trait scores across all time-points (F=7.15 to 149.31, p all
<.001).
Demographics by Clinical Outcome Group
As shown in Table 2.1, the proportions of boys in the ASD and OD groups were
significantly higher than the ND group (odds ratio [OR]=3.45 & 1.81, p<.001). More children in
the ASD and SPD groups were non-white (OR=1.71 & 2.18, p<.05). More parents of children in
the ASD and SPD groups reported lower education levels at T1 (OR=4.11 & 2.99, p<.001).
HYPER, HYPO, and SIRS Trajectories
Univariate linear LGCMs for each of the three sensory patterns indicated excellent model
fit (χ
2
(1)=.4 to .7, all CFI/TLI=1.00, RMSEA<.001). We then brought the three trajectories
together in an unconditional multivariate LGCM, which demonstrated a good fit: χ
2
(15)=27.4,
CFI=.995, TLI=.988, RMSEA=.023. The unconditional model indicated the intercepts of
HYPER and SIRS significantly different from zero (M=.22 & .49, SE=.04, p<.001). The slope
estimates indicated a significant increase in HYPO (M=.15, SE=.04, p<.001) and significant
decreases in HYPER and SIRS (M=-.26 & -.32, SE=.04, p<.001).
Latent Growth Factor Associations

29

As shown in Figure 2.2, strong correlations were found between the intercepts of HYPER
and HYPO (r=.72, p<.001), and HYPR and SIRS (r=.54, p<.001). Medium correlations were
observed between the slopes of HYPER and HYPO (r=.50, p<.001), as well as HYPER and
SIRS (r=.42, p<.001), indicating that these trajectories traveled together in the same direction
over time. Most of the significant intercept-slope correlations were weak and negative, ranging
from -.17 to -.26. The only positive correlation was between the intercept of SIRS and the slope
of HYPO (r=.14, p<.01).
Impact of Demographic Variables on Sensory Trajectories
The conditional model demonstrated satisfactory model fit: χ
2
(24)=66.0, CFI=.984,
TLI=.958, RMSEA=.035. The three demographic covariates explained the variability of the
latent growth factors to various extents (see Figure 2.3). Child’s sex was a significant predictor
of the intercepts of HYPO (β=-.13, p<.01). That is, boys tended to have higher HYPO scores at
baseline. Child’s race only predicted the intercept of SIRS (β=.17, p<.001), indicating that
families of non-white children tended to report more frequent SIRS behaviors at baseline. Parent
education levels predicted the intercepts of HYPO (β=.12, p=.011) and SIRS (β=-.15, p<.001),
and the slopes of HYPER (β=-.10, p<.01) and HYPO (β=-.19, p<.001). These results indicated
that children of parents with higher education levels had lower SIRS but higher HYPO scores at
baseline. Also, higher parent education levels predicted smaller increases in HYPER and HYPO
over time.
Sensory Trajectories by Clinical Outcome Group
Figure 2.4 shows the individual sensory trajectories by clinical outcome group
membership. As shown in Table 2.4, group differences were significant across all latent growth

30

factors (F=49.1 to 113.4, p all <.001), particularly in the slopes of HYPER and HYPO as well as
the intercept of SIRS. Children classified to ASD or SPD groups on average scored significantly
higher than their OD and ND counterparts at baseline and more increases across sensory
constructs (p all <.001). The ASD group showed steeper increases in HYPER and HYPO than
the other groups (p all <.001) but did not differ from SPD in the slope of SIRS. The OD group
differed from the ND group only in the slope of HYPER and the intercept of HYPO (p both
<.05). Overall, all the latent growth factors were able to differentiate ASD/SPD from OD/ND,
while the slopes of HYPER and HYPO were able to further differentiate ASD from SPD.
Associations Between Individual Growth Factor Estimates and Autism Severity at School Age
The correlation coefficients between the intercepts/slopes of the sensory patterns and
SRS T-scores at T3 were .31/.54 (HYPER), .36/.55 (HYPO), and .39/.34 (SIRS). Figure 2.5
showed the correlations specific to each clinical outcome group between the slopes of
HYPER/HYPO and SRS T-scores at T3. The correlations for the slopes of HYPER and HYPO in
ASD were specifically higher than the other groups (r=.43 & .55), demonstrating moderating
effects of clinical outcome status.
Discussion
This study was the first to investigate the developmental trajectories of sensory patterns
in a community sample spanning early childhood. There were several important and novel
findings. First, we demonstrated “chronogeneity” by identifying highly variable individual
trajectories of sensory patterns among this sample of children with diverse clinical outcomes. It
should be noted that the overall decreases in HYPER/SIRS and increase in HYPO were averages
across diverse individual trajectories. Past evidence on the development of sensory patterns (as

31

well as other behavioral traits) was often limited to ASD or ASD-TD comparisons, and thus
might fail to represent the broader severity continuum. Our findings showed the heterogeneous
trajectories as indexed by a wide range of severity and change rates as associated with various
clinical outcomes (Figure 2.4). For those with a parent-report ASD diagnosis, significant
increases were observed over time across all the sensory patterns, in contrast to the overall stable
or decreasing patterns in their non-ASD counterparts. These results were generally consistent
with the previous evidence on high-risk siblings later diagnosed with ASD versus non-ASD
controls, where significant group differences across sensory patterns were observed at baseline,
followed by a larger increase in HYPO among those with ASD during the first two years of life
(Wolff et al., 2019). Additionally, the current finding that HYPO scores decreased as children
aged in this longitudinal study complements the cross-sectional finding of an earlier study
(Baranek et al., 2013a) that HYPO was negatively correlated with mental age.
Importantly, the current study included children who were reported to have sensory
challenges but did not have an ASD diagnosis to understand the specific utility of sensory
patterns to differentiate ASD from sensory disorders. The ASD and SPD groups both showed
elevated sensory scores across this period with similar levels of challenges in the beginning.
However, more obvious ASD-SPD distinctions were found in HYPER and HYPO as early as
before age three since children classified to the SPD group did not show significant increases in
these two constructs as their ASD counterparts did. This indicated that the change rate of sensory
reactivity patterns might be a useful predictor of a later ASD outcome with good differentiability
from non-ASD conditions with sensory issues. The better utility of HYPER in differentiating
ASD versus non-ASD demonstrated in our study was consistent with the recent meta-analytic
results that HYPER generally demonstrated larger effect sizes across studies when comparing

32

ASD with other groups (Ben-Sasson et al., 2019). Our findings also partially agreed with a
previous comparison of sensory symptoms in children aged 2-15 years with ASD versus SPD,
where they found a significant group difference in HYPO, but not in HYPER and SIRS
(Tavassoli et al., 2018). Adding to the previous evidence, our study showed that sensory
differences between ASD and SPD were not only in severity but also in how much they change
over time. It is noteworthy that variable trajectories were also present “within” each of the
clinical outcome groups. For instance, decreasing/stable patterns in some constructs could also
be observed in some children classified to the ASD group. This thus calls for further subtyping
studies to unravel this chronogeneity which cannot be fully explained by the known grouping
variables (e.g., demographics and clinical outcomes).
We also found that the developmental variability of sensory patterns could be partially
explained by child’s sex, race, and parent education levels. Our findings showed that boys
seemed to initially score higher HYPO than girls did. This sex-related finding was partially
consistent with previous evidence that boys tended to have more frequent sensory abnormalities
(Jussila et al., 2020), while the sex differences in HYPER might not be obvious (Ben-Sasson et
al., 2009). It also highlighted the importance of including a general-population sample to reduce
potential biases from external factors such as sex, as many findings in autism research were
overrepresented by male participants (Pender et al., 2020). Furthermore, we reported the first
evidence on race differences in sensory patterns – non-white children were reported to have more
SIRS, which might be associated with race discrepancies in proxy ratings of children’s problem
behaviors (Harvey et al., 2013) that merit future research. Parent education seemed to have larger
impacts on sensory trajectories relative to child’s sex and race. The finding that higher parent
education levels predicted fewer increases in HYPER and HYPO may indicate the critical roles

33

of caregivers in their children’s development of sensory behaviors, particularly the sensory
reactivity patterns (i.e., HYPER and HYPO), which have been reported to be associated with
parent responsiveness (Kinard et al., 2017; Jaegermann & Klein, 2010). Another interesting
finding was that children of parents whom both had college degrees (or beyond) seemed to have
higher HYPO scores at infancy, while showing lower SIRS scores through school age. It is
possible that parents with higher education levels are more sensitive to their child’s atypical
sensory behaviors and thus tend to report more HYPO, which may be less observable than
HYPER and SIRS at infancy when the child is still nonverbal. While discrepancies between
parent-report and clinician-observational sensory behaviors have been reported in previous
studies (Zwaigenbaum et al., 2005; Baranek et al., 2008), further research is needed to examine
potential informant biases related to parent education or other family factors across different
sensory constructs particularly at infancy.
Finally, the significant associations among the growth factors of sensory trajectories
supported the notion that these sensory patterns are not mutually exclusive, and every child could
exhibit behaviors indexed by one or more sensory features across contexts and developmental
stages (Uljarević et al., 2017). The strong correlations between HYPER and HYPO in both
intercepts and slopes may reveal a dynamic process of up- and down-regulation of sensory input
during early development (Baranek, Reinhartsen, & Wannamaker, 2001). The negative
correlations between intercepts and slopes might be related to ceiling effects (i.e., a limited range
of potential increases for those with high baseline scores). The only positive correlation observed
between SIRS at baseline and change in HYPO may provide further temporal evidence (i.e., the
precedence of SIRS) on previous findings of their co-occurrence (Ausderau et al., 2016).

34

Limitations
A limitation of this study is the use of two different measures of sensory patterns across
time-points with a portion of items extracted from full measures. Although endeavors were made
to construct longitudinally comparable scores, measurement biases could not be completely
avoided. It is also possible that these items might not be comprehensive enough to capture the
full picture of sensory patterns of interest. Arguably the representativeness of the IRT scores
constructed based on partial forms was supported by the fact that those who were classified to
the SPD group based on total scores did show elevated IRT scores over time, but it would be
optimal to replicate the current findings with full-length and exactly repeated measures. Another
limitation is that only parent-report measures were used for the clinical-outcome classification.
Past studies have demonstrated a high level of agreement (96-98%) between parent reports of
their child receiving a definite ASD diagnosis and clinical reports (Daniels et al., 2012; Warnell
et al., 2015). The use of parent reports in some population-based studies of ASD might also
support its utility (e.g., Kogan et al., 2018; Turner-Brown et al., 2013). However, it is still
possible that our classifications with the parent-report data might not be consistent with
diagnoses made based on comprehensive in-person evaluations. Also, we did not examine the
influence of other factors, such as the types and intensity of interventions children received
during this period, on the trajectories, thus requiring future investigations to improve our
understanding of the interaction between risk and resilience.
Conclusion
This study provided the first evidence on the early trajectories of sensory patterns among
a community sample using person-centered analytical approaches. The child’s sex, race, parent
education, and long-term clinical outcome status accounted for the large variability of trajectories

35

to various extents. Notably, the latent growth factors, such as the change rates of HYPER and
HYPO, were able to differentiate children with ASD from those with other non-ASD conditions,
supporting their potential utility for early detection. The three sensory constructs were also found
to co-exist and influence each other over time. The significant impact of parent education on the
sensory trajectories may indicate that intervention strategies specifically addressing a child’s
developmental changes in sensory behaviors, with tailored family approaches, should be
considered in future research.

36

References
Ausderau, K. K., Sideris, J., Little, L. M., Furlong, M., Bulluck, J. C., & Baranek, G. T. (2016).
Sensory subtypes and associated outcomes in children with autism spectrum disorders.
Autism Research, 9(12), 1316-1327.
Baranek, G. T. (1999). Sensory Experiences Questionnaire (SEQ). Unpublished manuscript,
University of North Carolina at Chapel Hill.
Baranek, G. T., Carlson, M., Sideris, J., Kirby, A. V., Watson, L. R., Williams, K. L., & Bulluck,
J. (2019). Longitudinal assessment of stability of sensory features in children with autism
spectrum disorder or other developmental disabilities. Autism Research, 12(1), 100-111.
Baranek, G. T., David, F. J., Poe, M. D., Stone, W. L., & Watson, L. R. (2006). Sensory
Experiences Questionnaire: discriminating sensory features in young children with autism,
developmental delays, and typical development. Journal of Child Psychology and
Psychiatry, 47(6), 591-601.
Baranek, G. T., Little, L. M., Diane Parham, L., Ausderau, K. K., & Sabatos‐DeVito, M. G.
(2014). Sensory features in autism spectrum disorders. Handbook of Autism and Pervasive
Developmental Disorders, Fourth Edition.
Baranek, G. T., Reinhartsen, D., & Wannamaker, S. (2001). Play: Engaging children with
autism. Sensorimotor Interventions in Autism, 311-351.
Baranek, G. T., Roberts, J. E., David, F. J., Sideris, J., Mirrett, P. L., Hatton, D. D., & Bailey, D.
B. (2008). Developmental trajectories and correlates of sensory processing in young boys
with fragile X syndrome. Physical and Occupational Therapy in Pediatrics, 28(1), 79-98.

37

Baranek, G. T., Watson, L. R., Boyd, B. A., Poe, M. D., David, F. J., & McGuire, L. (2013a).
Hyporesponsiveness to social and nonsocial sensory stimuli in children with autism,
children with developmental delays, and typically developing children. Development and
Psychopathology, 25(2), 307-320.
Baranek, G. T., Watson, L. R., Crais, E. R., Turner-Brown, L. & Reznick, J. S. (2013). First
Years Inventory (FYI) 3.1. Chapel Hill, NC: University of North Carolina at Chapel Hill.
Ben-Sasson, A., & Carter, A. S. (2012). The application of the first year inventory for ASD
screening in Israel. Journal of Autism and Developmental Disorders, 42(9), 1906-1916.
Ben-Sasson, A., Gal, E., Fluss, R., Katz-Zetler, N., & Cermak, S. A. (2019). Update of a meta-
analysis of sensory symptoms in ASD: A new decade of research. Journal of Autism and
Developmental Disorders, 49(12), 4974-4996.
Ben-Sasson, A., Hen, L., Fluss, R., Cermak, S. A., Engel-Yeger, B., & Gal, E. (2009). A meta-
analysis of sensory modulation symptoms in individuals with autism spectrum disorders.
Journal of Autism and Developmental Disorders, 39(1), 1-11.
Canu, D., Van der Paelt, S., Canal-Bedia, R., Posada, M., Vanvuchelen, M., & Roeyers, H.
(2020). Early non-social behavioural indicators of autism spectrum disorder (ASD) in
siblings at elevated likelihood for ASD: a systematic review. European Child &
Adolescent Psychiatry, 1-42.
Cheung, G. W., & Rensvold, R. B. (2002). Evaluating goodness-of-fit indexes for testing
measurement invariance. Structural Equation Modeling, 9(2), 233-255.

38

Choi, S. W., Gibbons, L. E., & Crane, P. K. (2011). Lordif: An R package for detecting
differential item functioning using iterative hybrid ordinal logistic regression/item
response theory and Monte Carlo simulations. Journal of Statistical Software, 39(8), 1.
Choi, S. W., Gibbons, L. E., & Crane, P. K. (2016). Lordif: Logistic ordinal regression
differential item functioning using IRT. https://CRAN.R-project.org/package=lordif, R
package version 0.3-3.
Constantino J. N. & Gruber C. P. (2012). Social Responsiveness Scale–Second Edition (SRS-2).
Torrance, CA: Western Psychological Services.
Curran, P. J., Obeidat, K., & Losardo, D. (2010). Twelve frequently asked questions about
growth curve modeling. Journal of Cognition and Development, 11(2), 121-136.
Daniels, A. M., Rosenberg, R. E., Anderson, C., Law, J. K., Marvin, A. R., & Law, P. A. (2012).
Verification of parent-report of child autism spectrum disorder diagnosis to a web-based
autism registry. Journal of Autism and Developmental Disorders, 42(2), 257-265.
Dawson, G., Osterling, J., Meltzoff, A. N., & Kuhl, P. (2000). Case study of the development of
an infant with autism from birth to two years of age. Journal of Applied Developmental
Psychology, 21(3), 299-313.
Deconinck, N., Soncarrieu, M., & Dan, B. (2013). Toward better recognition of early predictors
for autism spectrum disorders. Pediatric Neurology, 49(4), 225-231.
Freuler, A., Baranek, G. T., Watson, L. R., Boyd, B. A., & Bulluck, J. C. (2012). Precursors and
trajectories of sensory features: qualitative analysis of infant home videos. American
Journal of Occupational Therapy, 66(5), e81-e84.

39

Georgiades, S., Bishop, S. L., & Frazier, T. (2017). Editorial Perspective: Longitudinal research
in autism–introducing the concept of ‘chronogeneity’. Journal of Child Psychology and
Psychiatry, 58(5), 634-636.
Georgiades, S., & Kasari, C. (2018). Reframing optimal outcomes in autism. JAMA Pediatrics,
172(8), 716-717.
Harvey, E. A., Fischer, C., Weieneth, J. L., Hurwitz, S. D., & Sayer, A. G. (2013). Predictors of
discrepancies between informants’ ratings of preschool-aged children's behavior: An
examination of ethnicity, child characteristics, and family functioning. Early Childhood
Research Quarterly, 28(4), 668-682.
Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis:
Conventional criteria versus new alternatives. Structural Equation Modeling: a
Multidisciplinary Journal, 6(1), 1-55.
Jaegermann, N., & Klein, P. S. (2010). Enhancing mothers' interactions with toddlers who have
sensory‐processing disorders. Infant Mental Health Journal: Official Publication of The
World Association for Infant Mental Health, 31(3), 291-311.
Jussila, K., Junttila, M., Kielinen, M., Ebeling, H., Joskitt, L., Moilanen, I., & Mattila, M. L.
(2020). Sensory abnormality and quantitative autism traits in children with and without
autism spectrum disorder in an epidemiological population. Journal of Autism and
Developmental Disorders, 50(1), 180-188.
Kern, J. K., Trivedi, M. H., Grannemann, B. D., Garver, C. R., Johnson, D. G., Andrews, A.
A., ... & Schroeder, J. L. (2007). Sensory correlations in autism. Autism, 11(2), 123-134.

40

Kim, S. H., Bal, V. H., Benrey, N., Choi, Y. B., Guthrie, W., Colombi, C., & Lord, C. (2018).
Variability in autism symptom trajectories using repeated observations from 14 to 36
months of age. Journal of the American Academy of Child & Adolescent Psychiatry,
57(11), 837-848.
Kim, S. H., & Lord, C. (2013). The behavioral manifestations of autism spectrum disorders. The
Neuroscience of Autism Spectrum Disorders, 25-37.
Kinard, J. L., Sideris, J., Watson, L. R., Baranek, G. T., Crais, E. R., Wakeford, L., & Turner-
Brown, L. M. (2017). Predictors of parent responsiveness to 1-year-olds at-risk for autism
spectrum disorder. Journal of Autism and Developmental Disorders, 47(1), 172-186.
Kogan, M. D., Vladutiu, C. J., Schieve, L. A., Ghandour, R. M., Blumberg, S. J., Zablotsky,
B., ... & Lu, M. C. (2018). The prevalence of parent-reported autism spectrum disorder
among US children. Pediatrics, 142(6).
Kolen, M. J., & Brennan, R. L. (2014). Test Equating, Scaling, and Linking: Methods and
Practices. Springer Science & Business Media.
Landa, R. J., Gross, A. L., Stuart, E. A., & Faherty, A. (2013). Developmental trajectories in
children with and without autism spectrum disorders: the first 3 years. Child Development,
84(2), 429-442.
LeBlanc, J. J., & Fagiolini, M. (2011). Autism: a “critical period” disorder? Neural
Plasticity, 2011.
Liss, M., Saulnier, C., Fein, D., & Kinsbourne, M. (2006). Sensory and attention abnormalities in
autistic spectrum disorders. Autism, 10, 155–172.

41

Little, L. M., Freuler, A. C., Houser, M. B., Guckian, L., Carbine, K., David, F. J., & Baranek, G.
T. (2011). Psychometric validation of the sensory experiences questionnaire. American
Journal of Occupational Therapy, 65(2), 207-210.
McCormick, C., Hepburn, S., Young, G. S., & Rogers, S. J. (2016). Sensory symptoms in
children with autism spectrum disorder, other developmental disorders and typical
development: a longitudinal study. Autism, 20(5), 572-579.
Millsap, R. E. (2012). Statistical approaches to measurement invariance. Routledge.
Muthén, L. K., & Muthén, B. (2018). Mplus. The Comprehensive Modelling Program for
Applied Researchers: User’s Guide, 5.
Paz, S. H., Spritzer, K. L., Morales, L. S., & Hays, R. D. (2013). Evaluation of the patient-
reported outcomes information system (PROMIS) Spanish-language physical functioning
items. Quality of Life Research, 22(7), 1819–1830.
Pender, R., Fearon, P., Heron, J., & Mandy, W. (2020). The longitudinal heterogeneity of autistic
traits: A systematic review. Research in Autism Spectrum Disorders, 79, 101671.
Perez-Repetto, L., Jasmin, E., Fombonne, E., Gisel, E., & Couture, M. (2017). Longitudinal
study of sensory features in children with autism spectrum disorder. Autism Research and
Treatment, 2017.
Reznick, J. S., Baranek, G. T., Reavis, S., Watson, L. R., & Crais, E. R. (2007). A parent-report
instrument for identifying one-year-olds at risk for an eventual diagnosis of autism: the
first year inventory. Journal of Autism and Developmental Disorders, 37(9), 1691-1710.

42

Reznick, J. S., Baranek, G. T., Watson, L. R., & Crais, E. R. (2005). Developmental Concerns
Questionnaire. Chapel Hill NC: University of North Carolina.
Sacrey, L. A. R., Zwaigenbaum, L., Bryson, S., Brian, J., Smith, I. M., Roberts, W., ... &
Vaillancourt, T. (2015). Can parents’ concerns predict autism spectrum disorder? A
prospective study of high-risk siblings from 6 to 36 months of age. Journal of the
American Academy of Child & Adolescent Psychiatry, 54(6), 470-478.
Stocking, M. L., & Lord, F. M. (1983). Developing a common metric in item response theory.
Applied Psychological Measurement, 7(2), 201-210.
Talay-Ongan, A., & Wood, K. (2000). Unusual sensory sensitivities in autism: A possible
crossroads. International Journal of Disability, Development and Education, 47(2), 201-
212.
Tavassoli, T., Miller, L. J., Schoen, S. A., Brout, J. J., Sullivan, J., & Baron-Cohen, S. (2018).
Sensory reactivity, empathizing and systemizing in autism spectrum conditions and
sensory processing disorder. Developmental Cognitive Neuroscience, 29, 72-77.
Turner-Brown, L. M., Baranek, G. T., Reznick, J. S., Watson, L. R., & Crais, E. R. (2013). The
First Year Inventory: a longitudinal follow-up of 12-month-old to 3-year-old children.
Autism, 17(5), 527-540.
Uljarević, M., Baranek, G., Vivanti, G., Hedley, D., Hudry, K., & Lane, A. (2017).
Heterogeneity of sensory features in autism spectrum disorder: Challenges and
perspectives for future research. Autism Research, 10(5), 703-710.

43

Warnell, F., George, B., McConachie, H., Johnson, M., Hardy, R., & Parr, J. R. (2015).
Designing and recruiting to UK autism spectrum disorder research databases: do they
include representative children with valid ASD diagnoses? BMJ Open, 5(9), e008625.
Wolff, J. J., Dimian, A. F., Botteron, K. N., Dager, S. R., Elison, J. T., Estes, A. M., ... & IBIS
Network. (2019). A longitudinal study of parent‐reported sensory responsiveness in
toddlers at‐risk for autism. Journal of Child Psychology and Psychiatry.
Wong, A. W., Lai, J. S., Correia, H., & Cella, D. (2015). Evaluating psychometric properties of
the Spanish-version of the pediatric functional assessment of chronic illness therapy-
perceived cognitive function (pedsFACIT-PCF). Quality of Life Research, 24(9), 2289-
2295.
Zwaigenbaum, L., Bryson, S., Rogers, T., Roberts, W., Brian, J., & Szatmari, P. (2005).
Behavioral manifestations of autism in the first year of life. International Journal of
Developmental Neuroscience, 23(2-3), 143-152.
Zwaigenbaum, L., & Penner, M. (2018). Autism spectrum disorder: advances in diagnosis and
evaluation. BMJ, 361.

44

Tables and Figures
Figure 2.1. Flow chart of diagnostic classification based on parent reports (N=1,517)

Any diagnosis of
ASD was reported at
T2 or T3?
SRS total T-score ≥
60 at T2 or T3?

ASD/AT group
(N=101)
Yes
()
No
Yes
SEQ total score >
1 SD above the mean
at T2 or T3?
SPD/SF group
(N=136)
Any sensory-related
diagnosis/concern
was reported at T2 or
T3?
Yes
Yes
No
No
OD Group
(N=222)
Any non-ASD/
sensory-related
diagnosis/concern
was reported at T2 or
T3?
Yes
ND Group
(N=1,058)
No
n=31
()
n=70
()
n=124
()
n=12
()
No
n=1,486
()
n=1,416
()
n=1,292
()
n=1,280
()
n=222
()
n=1,058
()

45

Table 2.1. Demographics and descriptive statistics of measures by clinical outcome group
(N=1,517)

Table 2.2. FYI/SEQ items used for deriving sensory construct scores
Hyper Hypo

SIRS
▪ Refuse certain food textures ▪ Ignore when name called ▪ Look at objects in unusual ways
*

▪ Avoid looking at face ▪ Tune-out sounds/noises
*
▪ Fascinated with textures
▪ Distressed when touched ▪ Slow to notice new objects ▪ Fascinated with lights
▪ Distressed at loud sounds
*
▪ Slow to react to pain ▪ Flap arms/hands
▪ Distressed during grooming
*
▪ Mouth non-food objects

*Anchor items

ASD
(N=101)
SPD
(N=136)
OD
(N=222)
ND
(N=1,058)
Child’s Sex (male) 74 (73%) 69 (51%) 131 (59%) 468 (44%)
Child’s Race
White 82 (81%) 105 (77%) 196 (89%) 932 (88%)
Black 7 (7%) 11 (8%) 9 (4%) 38 (4%)
Other 12 (12%) 20 (15%) 17 (7%) 88 (8%)
Parent Education Levels at T1 (5% missing)
Both parents had a college
degree (or beyond)
35 (35%) 66 (49%) 131 (59%) 664 (63%)
One of the parents had a
college degree (or beyond)
26 (26%) 28 (21%) 50 (23%) 224 (21%)
None of the parents had a
college degree (or beyond)
33 (33%) 37 (27%) 22 (10%) 117 (11%)
IRT Trait (Sensory Patterns) Scores [Mean (SD)]
HYPER
T1 .26 (.69) .38 (.69) .04 (.64) .06 (.65)
T2 .93 (.98) .56 (.76) .01 (.66) -.18 (.63)
T3 1.12 (.96) .47 (.86) -.12 (.70) -.28 (.66)
HYPO
T1 .11 (.62) .06 (.65) .07 (.62) -.07 (.56)
T2 .69 (.78) .39 (.75) -.01 (.67) -.07 (.64)
T3 .93 (.82) .42 (.70) .10 (.69) -.07 (.63)
SIRS
T1 .45 (.81) .63 (.79) .02 (.81) .03 (.77)
T2 .66 (.99) .96 (.88) -.26 (.75) -.17 (.77)
T3 .80 (.93) .75 (.85) -.25 (.75) -.29 (.73)

46

Table 2.3. Longitudinal invariance testing results
HYPER HYPO SIRS
CFI TLI RMSEA CFI TLI RMSEA CFI TLI RMSEA
Configural .955 .934 .039 1.00 1.00 <.001 .982 .973 .021
Metric .951 .935 .038 .981 .971 .016 .978 .971 .022
Scalar .719 .662 .087 .571 .420 .072 .767 .722 .067

Figure 2.2. Latent growth factor associations

*p< .05; **p< .01; ***p< .001 (two-tailed)

47

Figure 2.3. Trajectories of sensory patterns by child’s sex, race, and parent education levels
[estimated means with 95% confidence intervals]

HYPER
HYPO SIRS
Child’s Sex
Child’s Race
Parent Education
+

+
Parent Education: PE1=None of the parents had college degrees (or beyond); PE2=One parent had college degrees (or
beyond); PE3=Both parents had college degrees (or beyond)
*p< .05; **p< .01; ***p< .001.
Intercept***
Intercept**
Intercept*** Intercept*, Slope*** Slope**

48

Figure 2.4. Trajectories of sensory patterns by clinical outcome group (bolded lines: mean
trajectories of each group)

Table 2.4. Latent growth factor estimates by clinical outcome group [mean (SD)]
ASD
(N=101)
SPD
(N=136)
OD
(N=222)
ND
(N=1,058)
F Test
(3, 1513)
Post-hoc Group
Comparisons
HYPER INT .33 (.35) .33 (.35) .09 (.33) .05 (.31) 49.1
***
ASD>OD
***
, ASD>ND
***
,
SPD>OD
***
, SPD>ND
***
(ASD=SPD, OD=ND)
SLP .35 (.42) .06 (.38) -.10 (.29) -.16 (.26) 111.6
***
ASD>SPD
***
, ASD>OD
***
,
ASD>ND
***
, SPD>OD
***
,
SPD>ND
***
, OD>ND
*

HYPO INT .17 (.24) .10 (.26) -.00 (.22) -.06 (.21) 51.3
***
ASD>OD
***
, ASD>ND
***
,
SPD>OD
***
, SPD>ND
***
,
OD>ND
**
(ASD=SPD)
SLP .33 (.26) .16 (.22) .04 (.21) .00 (.20) 98.0
***
ASD>SPD
***
, ASD>OD
***
,
ASD>ND
***
, SPD>OD
***
,
SPD>ND
***
(OD=ND)
SIRS INT .45 (.39) .54 (.40) .02 (.36) .03 (.36) 113.4
***
ASD>OD
***
, ASD>ND
***
,
SPD>OD
***
, SPD>ND
***
(ASD=SPD, OD=ND)
SLP .14 (.32) .07 (.29) -.13 (.25) -.15 (.23) 67.4
***
ASD>OD
***
, ASD>ND
***
,
SPD>OD
***
, SPD>ND
***
(ASD=SPD, OD=ND)
*p< .05; **p< .01; ***p< .001 (two-tailed). INT=intercept; SLP=slope.

49

Figure 2.5. Associations between individual slopes of HYPER/HYPO and SRS T-scores at T3
(N=389)

50

CHAPTER 3
Early Developmental Profiles of Sensory Patterns in a Community Sample: Associations
with School-age Outcomes (Study 2)
Abstract
Background: Previous sensory-based subtyping studies have demonstrated the heterogeneous
manifestation of sensory patterns in children with ASD. However, there is a lack of evidence on
sensory subtypes as defined by longitudinal variability beyond ASD populations. This study
aimed to identify subtypes of sensory pattern trajectories from age 1 to 6 years among a large
community sample of young children, and to characterize these subtypes in terms of their
demographics and associated school-age outcomes, including autism symptoms,
adaptive/maladaptive behaviors, and activity participation.
Method: Sensory patterns of 1,517 children were assessed at three time-points (infancy,
preschool-age, and school-age) with parent-report questionnaires. Latent class growth analysis
was used to identify distinct subtypes based on trajectories across multiple domains of sensory
patterns covarying child’s sex, race, and parent education. Associations between class
membership and school-age outcomes were examined on a subset sample with outcome data
(N=389).
Results: Five distinct trajectory classes were identified: an Adaptive-All Improving class with
very low sensory features and better school-age outcomes (35%); an Elevated-All Worsening
class (3%) associated with high autism severity and poor outcomes across almost all domains;
and three other classes (62% in total) generally characterized by moderate levels of sensory
features and stable or improving patterns, as well as better outcomes than the Elevated-All
Worsening class. Children in the Elevated-All Worsening class were more likely to be boys and

51

have parents with less education.
Conclusion: Significant longitudinal variability of sensory patterns was observed in a
community sample of children and could be parsed out into meaningful subtypes. These subtypes
varied not only in the intensity and change rates of sensory patterns, but also in their
demographics and school-age outcomes. These findings indicate the significant contribution of
sensory patterns in defining differential pathways towards various long-term outcomes, and thus
have important implications for early identification and intervention.
Keywords: Sensory patterns; latent-class trajectory; school-age outcomes; autism spectrum
disorder; community sample

52

Introduction
Atypical sensory behaviors are frequently observed in children with autism spectrum
disorder (ASD) with prevalence ranging from 66 to 94% (Baranek et al., 2006; Zachor & Ben-
Itzchak, 2014). Sensory patterns are defined as behavioral responses to sensory stimuli in the
environment, including hyperresponsiveness (HYPER), hyporesponsiveness (HYPO), and
sensory interests, repetitions, and seeking behaviors (SIRS). Previous evidence has revealed the
early emergence of atypical sensory behaviors within the first year in infants who later developed
ASD (Van Etten et al., 2017; Wolff et al., 2019), indicating their potential utility as early
behavioral markers of ASD. Like other autism traits, the manifestation of sensory patterns is
heterogeneous and may change over time (Uljarević et al., 2017; Ben-Sasson et al., 2019). Thus,
our better understanding of such variability would contribute to better use of sensory patterns for
early detection and intervention planning.
Sensory Subtypes
The past decades of research on young children with ASD has revealed individual
differences in observed sensory patterns to various sensory modalities under social or non-social
contexts, which may be suggestive of different clinically relevant behavioral phenotypes
(Uljarević et al., 2017). It has been demonstrated that the presence of sensory subtypes varied
across severity and combinations of sensory patterns and/or modalities, particularly in children
with ASD (see Table 3.1 for a summary of studies on sensory-based subtyping for children aged
≤12 years). While the defining characteristics of these subtypes varied across studies, most of the
studies were able to identify a subgroup of children with ASD who did not show clinically
significant sensory features (ranging from 24 to 77% of the samples), indicating a broad range of
behavioral manifestation during childhood among ASD populations. Furthermore, some studies

53

found demographic differences across the identified subtypes, such as IQ and chronological age
(Liss et al., 2006; Ausderau et al., 2014; Lane, Molloy, & Bishop, 2014; Tomchek et al., 2018).
For instance, some of the relatively severe subtypes in these studies seemed to be associated with
younger ages. It remains unclear, however, whether such age differences were due to
developmental changes, because the subtyping was done on samples within a wide age range
(i.e., age-heterogeneous samples). Overall, the inconsistent findings due to different measures
used for subtyping, study designs, subtyping methods, and participant characteristics across
studies make it challenging to draw generalizable conclusions.
Importantly, while sensory patterns are prevalent in ASD, they are also commonly
observed in children with other neurodevelopmental conditions, and even in typically developing
children (Lane, Reynolds, & Dumenci, 2012; Watson et al., 2011). Population-based studies
showed that 5-8% of the preschool-aged and school-aged children were reported to have sensory
challenges (Jussila et al., 2020; Ahn et al., 2004). There has been a lack of evidence on sensory
subtypes in non-ASD populations although such studies are important for understanding how and
to what extent atypical sensory behaviors are unique to ASD. Previous findings on children with
sensory processing disorder (SPD) and a population-based sample that included children with
ASD and other diagnoses (Table 3.1b) demonstrated that variability in sensory patterns was also
present in non-ASD populations. Thus, elevated sensory patterns might be better considered as
reflections of children's innate characteristics transacting with their ongoing responses to
environmental demands, which may occur on a continuum for those with and without
developmental conditions (Baranek et al., 2014).
Associations between Sensory Patterns and Other Domains of Behavior

54

As shown in Table 3.1, many studies reported differences across the identified sensory
subtypes in other domains of behaviors, such as autism symptoms (Liss et al., 2006; Ausderau et
al., 2016), affective symptoms (Ben-Sasson et al., 2008), and adaptive behavior (Lane et al.,
2010; Ausderau et al., 2016; Tomchek et al., 2018), as external validations of the proposed
classifications. Non-subtyping studies also have shown that sensory patterns were associated
with adaptive behavior (Williams et al., 2018), social communication (Watson et al., 2011;
Tomchek, Little, & Dunn, 2015), motor skills (Surgent et al., 2020), as well as maladaptive
behavior (Green et al., 2012; O’Donnell et al., 2012) in children with ASD. Moreover, atypical
sensory patterns were found to impact autistic children’s daily activity participation, such as
mealtime (Zobel-Lachiusa et al., 2015) and school activities (Reynolds et al., 2011) in terms of
frequency and range of activities they participated in. Such associations with other domains of
behavior were also observed in non-ASD children. For instance, elevated sensory features were
associated with lower social communication and adaptive skills not only in children with ASD
but also in those with other developmental disabilities (Watson et al., 2011; Williams et al.,
2018). Subtypes of children with SPD all showed challenges in different domains of adaptive
behavior (Miller et al., 2017). These findings supported the clinical relevance of identified
subtypes and revealed that sensory challenges could impact other areas of behavior in young
children regardless of an ASD diagnosis. However, there is currently a lack of evidence on the
utility of sensory subtypes in predicting long-term outcomes. Such evidence is important for
clinicians to apply the classification findings to the early detection of children who might be
vulnerable to poor developmental outcomes. It also has critical implications for early
intervention by supporting the cascading effects of atypical sensory processing on subsequent
social and adaptive development (Thye et al., 2018).

55

Longitudinal Variability of Sensory Patterns
Another major empirical gap in sensory-related research is the longitudinal variability of
sensory patterns and their utility in predicting later outcomes. The manifestation of sensory
patterns may vary across development (Ben-Sasson et al., 2009; Baranek et al., 2019) and the
differences in their manifestation between ASD and other conditions may also shift over time
(Wolff et al., 2019; Ben-Sasson et al., 2019). Past research on the stability/changes of autism
symptoms during early childhood has shed light on the heterogeneity as defined by
developmental patterns. Some children showed elevated symptoms early in life that attenuated
over time while others followed a regressive pattern (Ozonoff et al., 2018), and the latter was
often associated with poorer long-term outcomes (Kalb et al., 2010). However, it remains unclear
whether such varying developmental patterns from very early in life are also present in the
sensory domain. As shown in Table 3.1c, some evidence on the longitudinal variability of
sensory patterns has revealed that at least 90% of the children with ASD or other conditions aged
2-12 years showed stable sensory patterns across two time-points (Ausderau et al., 2014; Dwyer
et al., 2020), generally consistent with findings from another non-subtyping longitudinal study,
where they found stable inter-personal standing across sensory patterns (Baranek et al., 2019). It
is noteworthy that the latent transition analysis used in Ausderau et al.’s study (2014) provided
information on the stability of subgroup membership over time and thus the subtyping process
was not defined by longitudinal changes. On the other hand, Dwyer and colleagues (2020)
reported the first evidence on sensory subtypes based on their continuous changes across two
time-points during childhood. However, only two time-points of data are not optimal for
investigating individual heterogeneity in growth as the error variance could not be well
accounted for in growth mixture models (Hesser, 2015). Additionally, the longitudinal data in

56

both studies were collected from age-heterogeneous samples, and thus the results might be
confounded by age. Given these methodological concerns, there is still a need for expanding the
existing evidence on longitudinal variability of sensory patterns with age-homogeneous samples,
particularly starting very early in life.
The current study aimed to fill the empirical gaps mentioned above by identifying
subgroups based on trajectories of sensory patterns (i.e., HYPER, HYPO, and SIRS) from
infancy to school age in a community sample. We also assessed to what extent the broader
developmental outcomes at school age differed across these identified subtypes to understand
their clinical relevance. The specific research questions include: 1) How many longitudinal
sensory subtypes can be identified using latent class growth analysis (LCGA) among a
community sample (N=1,517) and how are these subtypes defined by latent growth factors (i.e.,
intercept and slope)? 2) Do these subtypes differ in their demographic characteristics, including
child’s sex, race, and parents’ education levels? 3) Do children in these subtypes differ across
their clinical characteristics (e.g., ASD or other non-ASD conditions) and distal outcomes at age
6, including autism symptoms, adaptive/maladaptive behaviors, and activity participation, in a
subset sample with outcome data (N=389)?
Method
Participants and Procedures
This study was an extension of a longitudinal research study, the North Carolina Child
Development Survey (NCCDS), on a large cohort of families. Families were initially ascertained
from birth registries, with recruitment at the age of 6-19 months (T1) for completing the First
Years Inventory (FYI) version 3.1, and then were re-contacted by a follow-up survey at 3-4 years

57

(T2) to assess their developmental and ASD outcomes, including the Developmental Concerns
Questionnaire (DCQ) version 1.5, Sensory Experiences Questionnaire (SEQ) version 2.1, and
Social Responsiveness Scale (SRS) 2
nd
Edition. In the current study, the 2,236 families who
returned their responses at T2 were followed up via emails to complete new questionnaires
regarding their child’s current diagnostic status and/or any parent-report concerns (DCQ), as well
as sensory patterns (SEQ) at 6-7 years of age (T3, Phase-1). At the second phase of T3, which
was about five months after Phase-1 responses were returned, a subset of families (N=465) was
again contacted via email invitations to complete the SRS, Vineland Adaptive Behavior Scales
(VABS) 3
rd
Edition, and Participation and Environment Measure-Children and Youth (PEM-
CY). This subset included families who reported any diagnosis/concerns at previous time-points
(N=359) and a random sample of families whose responses did not indicate concerns at any of
the previous time-points (N=106). For data collection at T3-Phase 1, participants who completed
the survey could choose to be entered into a raffle at a 1/50 chance of winning a 50-dollar gift
card. At T3-Phase 2, each participant with complete responses was eligible to receive a 15-dollar
gift card and a report of their child's results (see Appendix 2 for the report template). We ended
up receiving 1,519 responses at Phase-1 and 389 responses at Phase-2 (response rates=68-73%).
Out of the 389 responses at Phase-2, 312 (80%) had reported developmental diagnosis or
concerns at previous time-points. After removing cases with incomplete data, 1,517 sets of
longitudinal responses were retained for further analysis. The demographics were shown in
Table 3.2 by full sample (with complete sensory scores measured by FYI and SEQ over three
time-points) and subset sample (with additional T3 distal outcomes).
Measures
Sensory Patterns

58

Due to the longitudinal nature of the study, two parent-report measures appropriate to the
age of the children were used to operationalize sensory patterns at each age.
The FYI version 3.1 (FYIv3.1; Baranek et al., 2013b)
The FYIv3.1 is a parent-report measure revised from a previous version (FYIv2.0) that
has been validated in a large community sample (Reznick et al., 2007; Turner-Brown et al.,
2013). It was designed to identify infants aged 6-16 months at risk for a later diagnosis of ASD,
measuring the frequency of behaviors across social communication, sensory-regulatory
functions, and motor development with a 5-point Likert scale. The FYIv3.1 data were collected
at T1. For this study, fourteen items related to sensory patterns were extracted for establishing
sensory construct scores.
The SEQ version 2.1 (SEQv2.1; Baranek, 1999)
The SEQv2.1 is a 33-item parent questionnaire designed to measure the frequency of
responses to daily sensory experiences for children ages 1-12 years with a 5-point Likert scale
(higher scores indicate endorsement of more sensory features). It has good internal consistency
(Cronbach's α=.80) and test-retest reliability (r=.92) (Little et al., 2011), along with good
discriminative validity (Baranek et al, 2006). Fourteen items common to those in the FYIv3.1
(varied in item wordings but intended to measure the same behavior) were extracted to establish
sensory construct scores, while the total score was used for evaluating the presence of sensory
challenges. The SEQv2.1 data were collected at T2 and T3.
Clinical and School-age Outcomes

59

Three parent-report measures were administered via follow-up surveys to capture a
child’s clinical outcomes across development and functional outcomes at school age, including
autism symptoms, adaptive/maladaptive behaviors, and activity participation.
The DCQ version 1.5 (DCQv1.5; Reznick et al., 2005)
The DCQv1.5 is a parent-report measure with open-ended questions about whether a
parent or professional has been concerned about the child’s development and whether the child
has received any clinical diagnoses. It was used as one of the outcome measures in the validation
study of FYIv2.0 (Turner-Brown et al., 2013). The DCQ data at T2 and T3 were used for
outcome classification. Responses were coded to determine whether the child has had a
diagnosed developmental disability, including ASD, and/or any developmental concerns (see
Appendix 3 for the REDCap coding form).
The SRS-2 (Constantino & Gruber, 2012)
SRS-2 is a well-validated parent-report scale that measures deficits in social behavior
associated with ASD to determine levels of autism symptoms. It has excellent internal
consistency (Cronbach's α=.95) and test-retest reliability (r=.88-.95). The SRS data were
collected at T2 (preschool-age version) and T3 (school-age version). T-scores at T3 were used as
one of the distal outcome variables. A T-score ≥60 suggests clinically significant social
impairment, and therefore this cutoff was used to indicate an ASD diagnosis in the current study.
The VABS-3 (Sparrow, Cicchetti, & Saulnier, 2016)
The domain-level parent/caregiver form was used to assess children's adaptive behaviors
in three domains (communication, social, and daily living skills), as well as motor skills and
maladaptive behaviors (internalizing and externalizing behaviors). This form has an excellent

60

internal consistency coefficient (Cronbach's α=.97) and test-retest reliability (r=.87). The VABS
data were collected at T3 and the standardized scores in these six subdomains were used for
analysis.
The PEM-CY (Coster, Law, & Bedell, 2010)
The PEM-CY is a parent-report questionnaire that measures participation in 25 types of
daily activities under three contexts (home, school, and community). At T3, the caregiver was
asked to report on the child’s participation frequency (never=0 to daily=7) and their extent of
involvement (minimally involved=1 to very involved=5) for each activity. Average scores in the
variety of activities a child participated in, frequency, and involvement of participation for each
context were rescaled to have the same range of scores (0 to 10). The test-retest reliability was
moderate to good (r=.58 to .84), and large to significant differences were found between groups
with and without disabilities on all subscales (Coster et al., 2011).
Data Analyses
Trait scores of sensory patterns based on item response theory (IRT) were constructed
upon the items extracted from the FYI and SEQ, following a series of longitudinal invariance
testing and adjustment for differential item functioning (for details, see Chapter 2). We first
performed univariate latent growth curve modeling (LGCM) on HYPER, HYPO, and SIRS
separately to determine their growth form (i.e., linear or non-linear). After confirming that the
growth form was consistent across the three sensory constructs and their model fits were
satisfactory, the three trajectories were estimated simultaneously in a multivariate (or parallel-
process) latent growth model, controlling for child’s sex, race, and parents’ education levels.

61

Model fit indices of CFI/TLI ≥.95 and RMSEA <.08 are considered a good model fit (Hu &
Bentler, 1999).
Next, LCGA, a special case of growth mixture modeling, in which the within-class
variances and covariances of latent growth factors were constrained to zero to meet the
assumption of homogeneity within class, was used to identify clusters based on the trajectories of
HYPER, HYPO, and SIRS. Multivariate LCGA is considered useful to examine the
developmental process of two or more constructs simultaneously when they are related to each
other (Zhou et al., 2020). Models with different numbers of classes were estimated to determine
the optimal class solution based on several fit indices: Akaike information criterion (AIC),
Bayesian information criterion (BIC), sample size-adjusted Bayesian information criterion
(SABIC), entropy, adjusted Lo-Mendell-Rubin likelihood ratio test (LMR-LRT) and bootstrap
likelihood ratio test (BLRT) statistics. Practically, lower AIC/BIC values indicate better fit and a
higher entropy value reflects a higher degree of class distinction. The LMR-LRT and BLRT
compared fit statistics between neighboring models, with p-values <.05 indicating significant
improvement in the model fit by adding each additional class (Tein, Coxe, & Cham, 2013). Upon
determining the most interpretable and parsimonious solution with optimal fit statistics, each
child was assigned to the class with the highest posterior probabilities. The demographic
variables were included as the predictors of trajectory class membership using Vermunt’s three-
step approach (Vermunt, 2010) to evaluate whether children with certain demographic
characteristics were more likely to be in certain classes.
Finally, Bolck-Croon-Hagenaars (BCH) three-step approach, a robust and flexible method
for adjusting classification errors (Bakk & Vermunt, 2016), was applied to test differences
between classes on each distal outcome variable (i.e., SRS, VABS, and PEM-CY subscale

62

scores). Those without age-6 outcome data were treated as missing in the model. All the latent
growth analyses were conducted with robust maximum-likelihood estimation in Mplus 8.4
(Muthén & Muthén, 2018). Additionally, we examined the proportion of children who met the
following conditions in each identified class with the evaluation of odds ratios (OR):
1) ASD: Reported by parents to have an ASD diagnosis via DCQ and/or met the SRS cutoff
of elevated risk for ASD at T2 or T3.
2) Non-ASD with sensory issues: Reported by parents to have sensory issues via DCQ
and/or SEQ total score >1 standard deviation (SD) above the mean at T2 or T3, without
co-occurrence of ASD as defined above.
Results
Longitudinal Trends of Sensory Patterns
Univariate linear LGCMs showed strong evidence of linear growth across the three
sensory constructs (χ
2
(1)=.4 to .7, all CFI/TLI=1.00, RMSEA<.001). The multivariate model
with simultaneous estimation of the three trajectories also demonstrated a good fit (χ
2
(15)=27.4,
CFI=.995, TLI=.988, RMSEA=.023). Notably, significant variances were found for the
intercepts and slopes across trajectories (all p<.001), suggesting the presence of significant
individual differences in the initial levels and change rates of all the sensory patterns. Thus, it
was of great interest to further parse out more homogeneous patterns among the highly variable
trajectories using LCGA.
Trajectory Classes of Sensory Patterns
LCGA models were then fitted with two to six classes to determine the optimal number of
trajectory classes (see Table 3.3 for fit statistics). Based on the fit indices and clinical

63

interpretability of these class solutions, a five-class model was selected as the final model. While
LMR-LRT showed non-significant improvement from the four-class to five-class model, the
entropy did slightly improve for the five-class model in addition to the decreases in AIC, BIC,
and SABIC. Although the proportion of the smallest class was only around 3%, it might
represent the estimated 5-9% of the general population with the most severe sensory issues (Ahn
et al., 2004; Jusilla et al., 2020) and thus we considered this class as clinically relevant. The
average posterior probabilities ranged from 76 to 87%, indicating a satisfactory degree of
precision in classifying children into these trajectory classes. The five trajectory classes were
depicted as below and visualized in Figure 3.1:
1) Class 1 (Adaptive-All Improving, n=537, 35%) accounted for the largest portion and
showed the lowest levels of scores across constructs over time. The intercepts of the three
constructs were significantly lower than zero (M=-.32 to -.22, SE=.04 to .05, p all <.001).
The mean slopes of HYPER and SIRS indicated significant decreases (M=-.21 & -.17,
SE=.02, p<.001), and a slight decrease was observed in HYPO (slope M=-.05, SE=.02,
p=.012).
2) Class 2 (Moderate-HYPO Worsening, n=171, 11%) was characterized by elevated scores
across constructs (intercepts M=.32 to .70, SE=.10 to .13, p all <.01). The mean slope of
HYPO indicated a significant increase (M=.25, SE=.05, p<.001), while HYPER and SIRS
remained stable over time.
3) Class 3 (Moderate SIRS-HYPER Improving, n=316, 21%) specifically scored high in SIRS
over time (intercept M=.64, SE=.07, p<.001), while low in the other two constructs.
HYPO and SIRS remained stable while HYPER significantly decreased (slope M=-.22,
SE=.05, p<.001).

64

4) Class 4 (Mild-SIRS Improving, n=449, 30%) was characterized by elevated scores in
HYPER and HYPO (intercepts M=.30 & .16, SE=.06 & .04, p<.001) while SIRS did not
differ from zero at baseline and was followed by a significant decrease (slope M=-.19,
SE=.04, p<.001).
5) Class 5 (Elevated-All Worsening, n=44, 3%) was the most severe group, characterized by
elevated HYPER at baseline (intercept M=.58, SE=.18, p=.001), and large increases in
HYPER and HYPO (slopes M=.72 & .59, SE=.17 & .14, p<.001). The increase in SIRS
was smaller but still significant (slope M=.47, SE=.19, p=.015).
Demographic Differences across Trajectory Classes
A series of multinomial logistic regression models was conducted using Vermunt’s three-
step approach on the full sample to examine the impact of child’s sex, race, and parent education
on the trajectory class membership. The demographic characteristics by class were shown in
Figure 3.1. Compared to Class 1, children in Class 5 were almost five times more likely to be
boys (OR=4.9, 95% confidence interval/CI=[1.7, 14.3], p=.003). Children in Class 3 were about
three times more likely to be non-white (OR=2.8, 95% CI=[1.6, 4.6], p<.001). Additionally,
parents of children in Class 2 and 5 were more likely to report lower education levels (OR=1.5 &
3.7, 95% CI=[1.2, 2.0] & [2.3, 5.9], p<.01). Class 4 did not differ from Class 1 in any of the
demographic variables.
Differential Age-6 Outcomes across Trajectory Classes
Using BCH three-step approach, the equality tests of means across classes on distal
outcomes revealed that children in Class 5 had significantly worse age-6 outcomes in almost all
areas (see Table 3.4; class 1 was chosen as the reference group for comparisons given that it is

65

the most “adaptive” class). Their average SRS T-score was particularly higher than the other
groups (Cohen’s d=2.7 compared to Class 1). Children in Class 2 and 4 also showed elevated
SRS and challenges in some of the adaptive behaviors: Class 2 had lower social communication
and motor skills while Class 4 had lower motor and daily living skills. Both groups were
associated with more maladaptive behaviors but overall had better adaptive outcomes than Class
5. Class 1 and 3 overall had better outcomes than the other three groups, and no significant
difference was found between them across outcome variables except for SRS.
Regarding the participation outcomes, children in Class 5 participated in a smaller variety
of activities with less frequency and involvement across all contexts, except for the frequency of
participating in community activities, as compared to Class 1. Children in Class 2 had overall
lower home participation, less involvement in school activities, and participation in fewer types
of community activities. Class 4 showed less involvement in home and school activities.
Associations with Parent-report Clinical Outcomes
Children in Class 5 were 122.9 times more likely than those in Class 1 to have an ASD
condition based on parent-report measures (Cohen’s d=2.7). Class 2 and 4 were also associated
with elevated probabilities of an ASD condition (OR=15.2 & 5.2, p<.01, Cohen’s d=1.5 & .9).
Children in Class 2, 3, and 5 were respectively 25.5, 20.4, and 8.6 times more likely to have
sensory issues despite the absence of ASD compared to those in Class 1. All children in Class 5
were reported to have an ASD condition and/or sensory issues.
Discussion
The current study was the first to subtype a large community sample of children based on
the developmental trajectories of sensory patterns across three time-points beginning in the

66

infancy period. We adopted a person-based analytic approach that takes both inter-person and
intra-person differences into account, and performed the subtyping in a sample with diverse
clinical outcomes beyond ASD. Notably, we estimated parallel-process trajectories of HYPER,
HYPO, and SIRS, given the previous evidence of their co-occurrence (Ausderau et al., 2014;
Lane et al., 2010), and identified five distinct classes which not only varied in intensity and rates
of change over time across sensory patterns, but also differed in distal outcomes by the time
children reached 6 years of age. Past findings on the developmental course of sensory patterns
were mostly mean-level changes and were often limited to ASD populations despite the presence
of sensory patterns in individuals with other conditions. Among our population-based sample, we
identified a class characterized by highly elevated sensory scores and dramatic worsening
patterns over time as well as significant autism severity and impairments across developmental
domains at school age. Such divergence from typical trajectories seemed to become evident
starting in infancy, indicating that sensory patterns might be a useful early behavioral marker of
ASD and associated challenges.
We also identified another two milder-severity classes (Class 2 and 4) associated with
elevated autism severity despite the relatively better adaptive and participation outcomes.
Approximately 59% of the children with an ASD outcome were classified to these two classes in
addition to 32% in the most severe class (Class 5). Thus, at least 91% of our sample with ASD
showed moderate to elevated levels of sensory challenge, which is consistent with the previously
estimated prevalence of sensory symptoms in children with ASD (e.g., Baranek et al., 2006).
These findings highlighted the concept of “chronogeneity” (i.e., heterogeneity as defined by
longitudinal patterns) and may help us to better understand ASD as a neurodevelopmental
disorder (Georgiades, Bishop, & Frazier, 2017). Importantly, such longitudinal heterogeneity of

67

sensory patterns was also observed in children with other non-ASD conditions. Around 89% of
the non-ASD children with reported sensory issues were classified to mild-to-moderate classes,
and 8% were in Class 5. Other children without reported ASD or sensory issues could be
classified into any of the first four classes. All these findings indicated that the manifestation of
sensory patterns is a continuum across ASD and non-ASD conditions not only in severity, as
indicated by the previous subtyping study on a population-based sample (Little et al., 2017), but
also in how they change or develop over time. Recent research on autism traits has recognized
this continuum view, which may impact how we set appropriate clinical thresholds for diagnosis
and identify clinically relevant phenotypes (Constantino & Charman, 2016). Here our findings
further revealed the importance of examining such continuum beyond ASD as defined by
developmental changes to better identify deviations from typical pathways.
Regarding the demographic differences across trajectory classes, the only sex difference
was found in Class 5. Children in this class were almost five times more likely to be boys
compared to those in Class 1. This is not surprising as the majority (82%) comprising this class
had an ASD outcome and the prevalence of ASD is highly male-biased (Werling & Geschwind,
2013). Another demographic variable of interest was race. While the proportion of non-white
participants in our study was low (around 13%), we found that 22% of the children in Class 3
were of a non-white race. This class was characterized by more SIRS across the period.
Furthermore, we found that the two most severe classes (Class 2 and 5) were associated with
lower parent education levels. While there is a lack of evidence on differences in children’s
sensory behaviors as affected by child ethnicity and parent education levels, discrepancies in
parent ratings of children’s challenging behavior were found to be associated with family
socioeconomic status (SES) and race (Harvery et al., 2013; Lawson et al., 2017). Further

68

research is needed to clarify to what extent such differences can be explained by informant
differences in cultural interpretation and reporting of their children’s behaviors versus true
differences in children’s behavioral manifestation.
Another significant contribution of the current study lies in the examination of the
associations between the developmental course of sensory patterns and long-term outcomes,
which were defined not only by diagnostic categories and symptom severity, but also by
children’s adaptive functions and participation. The findings revealed significant group
differences across these outcomes. In general, the subtypes characterized by relatively stable or
improving sensory trajectory patterns tended to show better outcomes; in contrast, the subtype
with worsening patterns (i.e., Class 5) seemed to have worse school-age outcomes. An
interesting finding was that children in Class 3 did not differ from those in Class 1 across almost
all school-age outcomes despite their elevated SIRS. Previous evidence has shown that more
SIRS was associated with better adaptive behavior and more frequent home participation in
children with ASD and other developmental disabilities (Little et al., 2015; Williams et al.,
2018). These findings suggest that the impact of SIRS may be context-specific and may not
impair children’s adaptive and participation outcomes as much as atypical sensory reactivity
(i.e., hypo/hyperresponsiveness) does up to 6 years of age. Overall, larger group differences were
found in autism symptoms and maladaptive behavior, indicating that these domains might be
closely associated with and potentially impacted by sensory development during early childhood.
The significant group differences observed in autism-related social impairments and social-
communication outcomes may reflect the cascading effects of atypical responses to sensory input
on social development (Thye et al., 2018). Particularly, the increase in hypo-responsiveness
observed in Class 2 and 5 might be most relevant to later social deficits, as indicated in previous

69

research (Baranek et al., 2013a). On the other hand, group differences in participation outcomes
were less significant, indicating that there might be other factors such as family SES and
environmental barriers/support contributing to children’s participation in daily activities, which
merits further research. On the other hand, the associations found between rates of change and
later outcomes suggest the sensory patterns may be malleable, potentially through early
interventions, and that a reduction in their growth may translate into better long-term outcomes.
Also, the identification of fine-grained phenotypes would allow clinical practitioners to apply
tailored approaches that target the most impacted behavioral domains for children and families
with various needs.
Limitations
A limitation of this study was that we were only able to examine the associations between
class membership and distal outcomes with the subset whereas the subtyping analysis was
conducted on the full sample. The 82% of the full sample we did not reach at age 6 were more
likely children without significant issues, and we made efforts to randomly sample a portion of
them for a follow-up. We also used robust statistical methods to account for this missingness.
Another limitation is the sole use of parent-report measures for categorizing children into ASD
versus non-ASD outcomes. Although parent-report measures have been well applied in large-
scale research for outcome ascertainment (e.g., Turner-Brown et al, 2013; Kogan et al., 2018)
and the agreement between parental and clinical reports of an ASD diagnosis could be as high as
98% (Daniels et al, 2012), gold-standard observational measures would allow for more accurate
diagnostic classifications. Nevertheless, the comprehensive assessments of children’s
developmental outcomes in the current study, including adaptive/maladaptive behaviors and

70

activity participation, may inform us about the clinical relevance of the subtypes beyond
diagnostic classifications.
Conclusion
In summary, this study was able to parse out more homogeneous trajectory subtypes of
sensory patterns among a large community sample of young children. The identified subtypes
were clinically relevant given their differential associations with long-term outcomes, such as
autism symptoms, adaptive/maladaptive behaviors, and activity participation. This indicated that
profiling children based on their early sensory trajectories may help to identify those who are
more likely to experience developmental challenges at school age, and may thus introduce
opportunities for early intervention. The presence of a trajectory profile that was highly
associated with ASD outcomes indicated the potential utility of sensory trajectories for early
identification of ASD. More evidence is needed to understand the longitudinal variability of
sensory patterns over a longer lifespan within and beyond ASD, as well as to decipher the roles
of early interventions and other factors, such as environmental support and accesses to services,
in directing the trajectories toward improved long-term outcomes. The current study responded
to the call for more research that addresses multidimensional sensory features and examines the
potential value of early sensory variability as indicators of later outcomes under prospective
designs (Uljarević et al., 2017). We expect future research to further expand this roadmap for
better understanding individual differences in sensory patterns across development as a critical
step forward toward personalized intervention.

71

References
Ahn, R. R., Miller, L. J., Milberger, S., & McIntosh, D. N. (2004). Prevalence of parents’
perceptions of sensory processing disorders among kindergarten children. American
Journal of Occupational Therapy, 58(3), 287-293.
Ausderau, K. K., Furlong, M., Sideris, J., Bulluck, J., Little, L. M., Watson, L. R., ... & Baranek,
G. T. (2014). Sensory subtypes in children with autism spectrum disorder: latent profile
transition analysis using a national survey of sensory features. Journal of Child
Psychology and Psychiatry, 55(8), 935-944.
Ausderau, K. K., Sideris, J., Little, L. M., Furlong, M., Bulluck, J. C., & Baranek, G. T. (2016).
Sensory subtypes and associated outcomes in children with autism spectrum disorders.
Autism Research, 9(12), 1316-1327.
Bakk, Z., & Vermunt, J. K. (2016). Robustness of stepwise latent class modeling with
continuous distal outcomes. Structural Equation Modeling: A Multidisciplinary Journal,
23(1), 20-31.
Baranek, G. T. (1999). Sensory Experiences Questionnaire (SEQ). Unpublished manuscript,
University of North Carolina at Chapel Hill.
Baranek, G. T., Carlson, M., Sideris, J., Kirby, A. V., Watson, L. R., Williams, K. L., & Bulluck,
J. (2019). Longitudinal assessment of stability of sensory features in children with autism
spectrum disorder or other developmental disabilities. Autism Research, 12(1), 100-111.
Baranek, G. T., David, F. J., Poe, M. D., Stone, W. L., & Watson, L. R. (2006). Sensory
Experiences Questionnaire: discriminating sensory features in young children with autism,

72

developmental delays, and typical development. Journal of Child Psychology and
Psychiatry, 47(6), 591-601.
Baranek, G. T., Little, L. M., Diane Parham, L., Ausderau, K. K., & Sabatos‐DeVito, M. G.
(2014). Sensory features in autism spectrum disorders. Handbook of Autism and Pervasive
Developmental Disorders, Fourth Edition.
Baranek, G. T., Watson, L. R., Boyd, B. A., Poe, M. D., David, F. J., & McGuire, L. (2013a).
Hyporesponsiveness to social and nonsocial sensory stimuli in children with autism,
children with developmental delays, and typically developing children. Development and
Psychopathology, 25(2), 307.
Baranek, G. T., Watson, L. R., Crais, E. R., Turner-Brown, L. & Reznick, J. S. (2013). First
Years Inventory (FYI) 3.1. Chapel Hill, NC: University of North Carolina at Chapel Hill.
Ben‐Sasson, A., Cermak, S. A., Orsmond, G. I., Tager‐Flusberg, H., Kadlec, M. B., & Carter, A.
S. (2008). Sensory clusters of toddlers with autism spectrum disorders: Differences in
affective symptoms. Journal of Child Psychology and Psychiatry, 49(8), 817-825.
Ben-Sasson, A., Gal, E., Fluss, R., Katz-Zetler, N., & Cermak, S. A. (2019). Update of a meta-
analysis of sensory symptoms in ASD: A new decade of research. Journal of Autism and
Developmental Disorders, 49(12), 4974-4996.
Ben-Sasson, A., Hen, L., Fluss, R., Cermak, S. A., Engel-Yeger, B., & Gal, E. (2009). A meta-
analysis of sensory modulation symptoms in individuals with autism spectrum disorders.
Journal of Autism and Developmental Disorders, 39(1), 1-11.

73

Constantino, J. N., & Charman, T. (2016). Diagnosis of autism spectrum disorder: reconciling
the syndrome, its diverse origins, and variation in expression. The Lancet Neurology,
15(3), 279-291.
Constantino J. N. & Gruber C. P. (2012). Social Responsiveness Scale–Second Edition (SRS-2).
Torrance, CA: Western Psychological Services.
Coster, W., Bedell, G., Law, M., Khetani, M. A., Teplicky, R., Liljenquist, K., ... & Kao, Y. C.
(2011). Psychometric evaluation of the Participation and Environment Measure for
Children and Youth. Developmental Medicine & Child Neurology, 53(11), 1030-1037.
Coster W., Law M., & Bedell G. (2010). Participation and Environment Measure for Children
and Youth. Boston, MA: Boston University.
Daniels, A. M., Rosenberg, R. E., Anderson, C., Law, J. K., Marvin, A. R., & Law, P. A. (2012).
Verification of parent-report of child autism spectrum disorder diagnosis to a web-based
autism registry. Journal of Autism and Developmental Disorders, 42(2), 257-265.
Dwyer, P., Saron, C. D., & Rivera, S. M. (2020). Identification of Longitudinal Sensory
Subtypes in Typical Development and Autism Spectrum Development Using Growth
Mixture Modelling. Research in Autism Spectrum Disorders, 78, 101645.
Georgiades, S., Bishop, S. L., & Frazier, T. (2017). Editorial Perspective: Longitudinal research
in autism–introducing the concept of ‘chronogeneity’. Journal of Child Psychology and
Psychiatry, 58(5), 634-636.

74

Green, S. A., Ben-Sasson, A., Soto, T. W., & Carter, A. S. (2012). Anxiety and sensory over-
responsivity in toddlers with autism spectrum disorders: bidirectional effects across
time. Journal of Autism and Developmental Disorders, 42(6), 1112-1119.
Harvey, E. A., Fischer, C., Weieneth, J. L., Hurwitz, S. D., & Sayer, A. G. (2013). Predictors of
discrepancies between informants’ ratings of preschool-aged children's behavior: An
examination of ethnicity, child characteristics, and family functioning. Early Childhood
Research Quarterly, 28(4), 668-682.
Hesser, H. (2015). Modeling individual differences in randomized experiments using growth
models: Recommendations for design, statistical analysis and reporting of results of
internet interventions. Internet Interventions, 2(2), 110-120.
Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis:
Conventional criteria versus new alternatives. Structural Equation Modeling: A
Multidisciplinary Journal, 6(1), 1-55.
Jussila, K., Junttila, M., Kielinen, M., Ebeling, H., Joskitt, L., Moilanen, I., & Mattila, M. L.
(2019). Sensory Abnormality and Quantitative Autism Traits in Children With and
Without Autism Spectrum Disorder in an Epidemiological Population. Journal of Autism
and Developmental Disorders, 1-9.
Kalb, L. G., Law, J. K., Landa, R., & Law, P. A. (2010). Onset patterns prior to 36 months in
autism spectrum disorders. Journal of Autism and Developmental Disorders, 40(11),
1389-1402.

75

Kogan, M. D., Vladutiu, C. J., Schieve, L. A., Ghandour, R. M., Blumberg, S. J., Zablotsky,
B., ... & Lu, M.C. (2018). The prevalence of parent-reported autism spectrum disorder
among US children. Pediatrics, 142(6).
Lane, A. E., Molloy, C. A., & Bishop, S. L. (2014). Classification of children with autism
spectrum disorder by sensory subtype: a case for sensory‐based phenotypes. Autism
Research, 7(3), 322-333.
Lane, S. J., Reynolds, S., & Dumenci, L. (2012). Sensory overresponsivity and anxiety in
typically developing children and children with autism and attention deficit hyperactivity
disorder: cause or coexistence? American Journal of Occupational Therapy, 66(5), 595-
603.
Lane, A. E., Young, R. L., Baker, A. E., & Angley, M. T. (2010). Sensory processing subtypes in
autism: Association with adaptive behavior. Journal of Autism and Developmental
Disorders, 40(1), 112-122.
Lawson, G. M., Nissley-Tsiopinis, J., Nahmias, A., McConaughy, S. H., & Eiraldi, R. (2017).
Do parent and teacher report of ADHD symptoms in children differ by SES and racial
status? Journal of Psychopathology and Behavioral Assessment, 39(3), 426-440.
Liss, M., Saulnier, C., Fein, D., & Kinsbourne, M. (2006). Sensory and attention abnormalities in
autistic spectrum disorders. Autism, 10, 155–172.
Little, L. M., Ausderau, K., Sideris, J., & Baranek, G. T. (2015). Activity participation and
sensory features among children with autism spectrum disorders. Journal of Autism and
Developmental Disorders, 45(9), 2981-2990.

76

Little, L. M., Dean, E., Tomchek, S. D., & Dunn, W. (2017). Classifying sensory profiles of
children in the general population. Child: Care, Health and Development, 43(1), 81-88.
Little, L. M., Freuler, A. C., Houser, M. B., Guckian, L., Carbine, K., David, F. J., & Baranek, G.
T. (2011). Psychometric validation of the sensory experiences questionnaire. American
Journal of Occupational Therapy, 65(2), 207-210.
Miller, L. J., Schoen, S. A., Mulligan, S., & Sullivan, J. (2017). Identification of sensory
processing and integration symptom clusters: a preliminary study. Occupational Therapy
International, 2017.
Muthén, L. K., & Muthén, B. (2018). Mplus. The Comprehensive Modelling Program for
Applied Researchers: User’s Guide, 5.
O’Donnell, S., Deitz, J., Kartin, D., Nalty, T., & Dawson, G. (2012). Sensory processing,
problem behavior, adaptive behavior, and cognition in preschool children with autism
spectrum disorders. American Journal of Occupational Therapy, 66(5), 586-594.
Ozonoff, S., Gangi, D., Hanzel, E. P., Hill, A., Hill, M. M., Miller, M., ... & Iosif, A. M. (2018).
Onset patterns in autism: Variation across informants, methods, and timing. Autism
Research, 11(5), 788-797.
Reynolds, S., Bendixen, R. M., Lawrence, T., & Lane, S. J. (2011). A pilot study examining
activity participation, sensory responsiveness, and competence in children with high
functioning autism spectrum disorder. Journal of Autism and Developmental
Disorders, 41(11), 1496-1506.

77

Reznick, J. S., Baranek, G. T., Reavis, S., Watson, L. R., & Crais, E. R. (2007). A parent-report
instrument for identifying one-year-olds at risk for an eventual diagnosis of autism: the
first year inventory. Journal of Autism and Developmental Disorders, 37(9), 1691-1710.
Reznick, J. S., Baranek, G. T., Watson, L. R., & Crais, E.R. (2005). Developmental Concerns
Questionnaire. Chapel Hill NC: University of North Carolina.
Simpson, K., Adams, D., Alston-Knox, C., Heussler, H. S., & Keen, D. (2019). Exploring the
sensory profiles of children on the autism spectrum using the short sensory profile-2 (SSP-
2). Journal of Autism and Developmental Disorders, 49(5), 2069-2079.
Sparrow, S. S., Cicchetti, D. V., & Balla, D. A. (2016). Vineland-III: Vineland Adaptive
Behavior Scales (3rd ed.). San Antonio, TX: Pearson.
Surgent, O. J., Walczak, M., Zarzycki, O., Ausderau, K., & Travers, B. G. (2020). IQ and
sensory symptom severity best predict motor ability in children with and without autism
spectrum disorder. Journal of Autism and Developmental Disorders, 1-12.
Tein, J. Y., Coxe, S., & Cham, H. (2013). Statistical power to detect the correct number of
classes in latent profile analysis. Structural Equation Modeling: A Multidisciplinary
Journal, 20(4), 640-657.
Thye, M. D., Bednarz, H. M., Herringshaw, A. J., Sartin, E. B., & Kana, R. K. (2018). The
impact of atypical sensory processing on social impairments in autism spectrum disorder.
Developmental Cognitive Neuroscience, 29, 151-167.

78

Tomchek, S. D., Little, L. M., & Dunn, W. (2015). Sensory pattern contributions to
developmental performance in children with autism spectrum disorder. American Journal
of Occupational Therapy, 69(5), 6905185040p1-6905185040p10.
Tomchek, S. D., Little, L. M., Myers, J., & Dunn, W. (2018). Sensory subtypes in preschool
aged children with autism spectrum disorder. Journal of Autism and Developmental
Disorders, 48(6), 2139-2147.
Turner-Brown, L. M., Baranek, G. T., Reznick, J. S., Watson, L. R., & Crais, E. R. (2013). The
First Year Inventory: a longitudinal follow-up of 12-month-old to 3-year-old children.
Autism, 17(5), 527-540.
Uljarević, M., Baranek, G., Vivanti, G., Hedley, D., Hudry, K., & Lane, A. (2017).
Heterogeneity of sensory features in autism spectrum disorder: Challenges and
perspectives for future research. Autism Research, 10(5), 703-710.
Van Etten, H. M., Kaur, M., Srinivasan, S. M., Cohen, S. J., Bhat, A., & Dobkins, K. R. (2017).
Increased prevalence of unusual sensory behaviors in infants at risk for, and teens with,
autism spectrum disorder. Journal of Autism and Developmental Disorders, 47(11), 3431-
3445.
Vermunt, J. K. (2010). Latent class modeling with covariates: two improved three‐step
approaches. Political Analysis, 18, 450‐469.
Watson, L. R., Patten, E., Baranek, G. T., Poe, M., Boyd, B. A., Freuler, A., & Lorenzi, J.
(2011). Differential associations between sensory response patterns and language, social,
and communication measures in children with autism or other developmental
disabilities. Journal of Speech, Language, and Hearing Research, 54(6), 1562-1576.

79

Werling, D. M., & Geschwind, D. H. (2013). Sex differences in autism spectrum disorders.
Current Opinion in Neurology, 26(2), 146.
Williams, K. L., Kirby, A. V., Watson, L. R., Sideris, J., Bulluck, J., & Baranek, G. T. (2018).
Sensory features as predictors of adaptive behaviors: A comparative longitudinal study of
children with autism spectrum disorder and other developmental disabilities. Research in
Developmental Disabilities, 81, 103-112.
Wolff, J. J., Dimian, A. F., Botteron, K. N., Dager, S. R., Elison, J. T., Estes, A. M., ... & IBIS
Network. (2019). A longitudinal study of parent‐reported sensory responsiveness in
toddlers at‐risk for autism. Journal of Child Psychology and Psychiatry.
Zachor, D. A., & Ben-Itzchak, E. (2014). The relationship between clinical presentation and
unusual sensory interests in autism spectrum disorders: A preliminary investigation.
Journal of Autism and Developmental Disorders, 44(1), 229-235.
Zhou, Y., Zheng, H., Liang, Y., Wang, J., Han, R., & Liu, Z. (2020). Joint developmental
trajectories of bullying and victimization from childhood to adolescence: a parallel-
process latent class growth analysis. Journal of Interpersonal Violence,
0886260520933054.
Zobel-Lachiusa, J., Andrianopoulos, M. V., Mailloux, Z., & Cermak, S. A. (2015). Sensory
differences and mealtime behavior in children with autism. American Journal of
Occupational Therapy, 69(5), 6905185050p1-6905185050p8.

80

Tables and Figures
Table 3.1. Previous evidence on sensory-based subtypes
a. Cross-sectional findings on young children with ASD
Study Sample Age Measure Subtyping
Method
Identified Subtypes (%) Associations with Other
Variables
Liss et al.,
2006
144
children
with ASD
102.4
months
(mean)
Sensory Profile
(and other
behavioral
measures)
Hierarchical
agglomerative
cluster analysis
Sensory-related characterizations:
S1) High overreactivity and sensory
seeking (12%)
S2) Few sensory symptoms (25%)
S3) High underreactivity and sensory
seeking (30%)
S4) Moderate sensory overreactivity
(33%)
• S1 and S3 had higher autism
severity; S3 had the lowest
adaptive functioning and more
deficits in social communication
• S2 and S4 were less impaired in
adaptive behavior and had lower
autism severity; children in S2
were younger, while those in S4
were older.
Ben-
Sasson et
al., 2008
170
toddlers
with ASD
18-33
months
Infant Toddler
Sensory Profile
(construct
scores)
Ward’ s
minimum
variance
hierarchical
cluster analysis
S1) Low frequency of sensory symptoms
(26%)
S2) High frequency of sensory symptoms
(29%)
S3) Mixed: high under-and over-
responsivity + low seeking (45%)

S2 & S3 scored higher on negative
emotionality, depression, and
anxiety symptoms than S1.
Lane et al.,
2010
54 children
with ASD
33-115
months
Short Sensory
Profile
(construct
scores)
Model-based
cluster analysis
S1) Sensory-based inattentive seeking
(44%)
S2) Sensory modulation with movement
sensitivity (32%)
S3) Sensory modulation with taste/smell
sensitivity (24%)

• Sensory processing patterns were
associated with communication and
adaptive behavior.
• S3 had significantly greater
communication impairment than
S2.
Lane et al.,
2014
228
children
with ASD
2-10
years
Short Sensory
Profile
(construct
scores)
Model-based
cluster analysis
S1) Sensory adaptive (38%)
S2) Taste/Smell sensitive (40%)
S3) Postural inattentive (10%)
S4) Generalized sensory difference (12%)

• Children in S2 were younger and
had lower NVIQ.
• No group difference was found in
autism severity.
Tomchek
et al., 2018
400
children
with ASD
3-6
years
Short Sensory
Profile
(construct
scores)
Latent profile
analysis
S1) Sensorimotor: elevated scores in
taste/smell sensitivity, sensory seeking,
and hypo-responsivity (51%)
S2) Selective-complex: high levels of
• Subtypes differed in age (S1 &
S3 younger, S2 & S4 older
• S1 had the worst adaptive
functions while S4 had better
80

81

seeking with hypo-responsivity (15%)
S3) Perceptive-adaptable (24%)
S4) Vigilant-engaged: elevated sensitivity
and seeking behaviors (10%)

developmental performance than
the other subtypes.
Simpson et
al., 2019
271
children
with ASD
4-11
years
Short Sensory
Profile
(construct
scores)
Cluster analysis S1) Uniformly elevated (67%)
S2) Raised avoiding and sensitivity
(33%): elevated sensitivity/avoiding and
typical range on seeking and registration

No difference was found in age and
autism characteristics across
subtypes.

b. Cross-sectional findings on young children with other conditions
Study Sample Age Measure Subtyping
Method
Identified Subtypes (%) Associations with Other Variables
Little et al.,
2017
1,132 children with
typical
development, ASD
or other diagnoses
(population-based
sample)
3-14
years
Sensory Profile
(construct
scores)
Latent
profile
analysis
S1) Balanced (79%): low sensory features
S2) Interested (7%): elevated sensory
seeking
S3) Intense (5%): elevated sensory features
across all domains
S4) Mellow Until… (4%): elevated
registration and avoidance
S5) Vigilant (5%): elevated avoiding and
sensitivity
• Children with different
developmental conditions were
classified across subtypes. 22% of
ASD vs. 2% of TD children were
classified to S3.
• Children in S2 were significantly
younger than the other subtypes.
Miller et
al., 2017
252 children with
sensory processing
disorder
4-14
years
Sensory
Processing 3-
Dimension
Inventory
(construct
scores)
Ward’s
agglomerati
ve
hierarchical
cluster
analysis
S1) High sensory overresponsivity only
(46%)
S2) High sensory under + overresponsivity
(29%)
S3) High sensory craving + overresponsivity
(25%)
• There was no significant group
difference in adaptive behavior.
• S3 showed more challenging
behaviors than the other subtypes.

c. Longitudinal findings on young children with ASD or other conditions
Study Sample Age Measure Subtyping
Method
Identified Subtypes (%) Associations with Other Variables
Ausderau
et al.,
2014, 2016
Children with
ASD (n=1,294
2-12 years Sensory
Experiences
Questionnaire
Latent profile
transition
analysis
91% of the sample remained in the same
class across time:
S1) Mild (29%): low sensory features
• S3 was associated with younger age
and IQ on average and showed lowest
levels of adaptive behavior.
81

82

at T1, 884 at
T2)
(construct
scores)
S2) Sensitive-distressed (27%): low
HYPO/SIRS + high HYPER/enhanced
perception
S3) Attenuated-preoccupied (17%): high
HYPO/SIRS + low HYPER/enhanced
perception
S4) Extreme-mixed (17%): high across all
sensory features

• S4 was associated with higher autism
severity and lower SES, as well as
highest levels of maladaptive behavior
and parenting stress.
Dwyer et
al., 2020
Children with
ASD and TD
(n=245 at T1,
n=132 at T2)
2-5 years
at T1; 4-
10 years
at T2
Short Sensory
Profile
(total score)
Growth
mixture
modeling
S1) Stable Mild (55%)
S2) Stable Intense (43%)
S3) Increasingly Intense (2%)
• 64% of the ASD group were
classified to S2 and 3% to S3. 2% of
TD were in S2, none in S3.
• Children in S2 showed elevated
anxiety; S3 was associated with higher
cognitive functions.

82

83

Table 3.2. Sample demographics

Table 3.3. Fit statistics for multivariate LCGA
No. of
classes
AIC BIC SABIC Entropy LMR-LRT
(p)
BLRT
(p)
Latent class
proportion (%)
2 28883.1 28984.2 28923.9 .74 <.001 <.001 69/31
3 28530.8 28669.2 28586.6 .72 .015 <.001 44/48/8
4 28228.5 28404.2 28299.4 .69 <.001 <.001 39/23/28/10
5 28128.8 28341.8 28214.7 .70 .137 <.001 35/21/30/11/3
6 28047.1 28297.3 28148.0 .67 .244 <.001 34/18/18/16/11/3

Full Sample
(N=1,517)
Subset Sample
w/ Age-6
Outcomes
(N=389)
Sex (male) 742 (49%) 233 (60%)
Race
+

White 1,315 (87%) 341 (88%)
Black 65 (4%) 11 (3%)
Asian 16 (1%) 4 (1%)
American Indian/Hawaiian 11 (1%) 4 (1%)
Multi-racial/Other 110 (7%) 29 (7%)
Parent Education
++
(5% missing)
Both parents had a college
degree (or beyond)
896 (59%) 205 (53%)
One of the parents had a
college degree (or beyond)
328 (22%) 95 (24%)
None of the parents had a
college degree (or beyond)
209 (14%) 69 (18%)
+ For the analysis purposes, non-white races (Black and other races) were combined into
one group given the small sample sizes.
++ Categorization was based on the reported mother’s and father’s education at T1.

84

Figure 3.1. Latent trajectory classes of sensory patterns (5 classes) [estimated means with 95% confidence
intervals]

Avg. Posterior Probabilities 86% 78% 76%
Boys 47% 54% 49%
Non-white races 11% 16% 22%
None of the parents had a
college degree (or beyond)

12%

21%

17%
Avg. Posterior Probabilities 78% 87%
Boys 46% 77%
Non-white races 9% 11%
None of the parents had a
college degree (or beyond)

11%

48%
Class 1 (n=537, 35%)
Adaptive - All Improving
Class 2 (n=171, 11%)
Moderate - HYPO Worsening
Class 3 (n=316, 21%)
Mod SIRS - HYPER Improving
Class 4 (n=449, 30%)
Mild - SIRS Improving
Class 5 (n=44, 3%)
Elevated - All Worsening
Sensory Patterns

85

Table 3.4. Age-6 distal outcome by trajectory class (subset sample N=389)

Mean (S.E.)
Class 1
(n=82)
Class 2
(n=90)
Class 3
(n=68)
Class 4
(n=115)
Class 5
(n=34)
2 vs. 1 3 vs. 1 4 vs. 1 5 vs. 1
χ
2
(p-value:
*
<.05,
**
<.01,
***
<.001)

SRS-2 T-score 44.0 (1.0) 57.7 (1.2) 47.6 (1.3) 50.7 (1.0) 74.1 (2.4) 83.9
***
4.5
*
18.1
***
133.3
***

VABS-3 Standardized Score
Communication Skills 104.6 (1.8) 97.9 (2.0) 106.6 (3.1) 101.2 (1.7) 77.7 (2.6) 6.7
*
.3 1.5 75.1
***

Social Skills 102.8 (1.6) 94.0 (1.6) 106.5 (2.8) 98.9 (1.4) 78.1 (2.0) 16.2
***
1.2 2.6 89.9
***

Daily Living Skills 101.0 (1.5) 97.3 (1.7) 101.9 (2.9) 97.3 (1.7) 80.5 (2.3) 2.9 .1 5.7
*
54.0
***

Motor Skills 106.0 (1.6) 99.7 (1.7) 104.5 (2.3) 98.2 (1.3) 85.2 (2.6) 7.7
**
.2 11.6
**
46.2
***

Internalizing Behavior 13.9 (.3) 17.5 (.3) 14.6 (.5) 16.4 (.4) 19.7 (.4) 58.5
***
1.0 19.3
***
102.9
***

Externalizing Behavior 13.9 (.4) 17.7 (.3) 14.4 (.6) 15.8 (.3) 19.0 (.6) 56.0
***
.3 9.5
**
55.1
***

PEM-CY Score
Home Variety 9.9 (.1) 9.7 (.1) 9.7 (.1) 9.6 (.1) 8.9 (.3) 4.4
*
3.0 2.6 13.3
***

Frequency 9.1 (.1) 8.7 (.1) 9.0 (.1) 8.9 (.1) 8.6 (.1) 10.2
**
1.1 2.4 13.0
***

Involvement 8.5 (.2) 7.8 (.1) 8.3 (.1) 7.8 (.1) 7.1 (.2) 8.2
**
1.0 9.0
**
27.1
***

School Variety 8.5 (.3) 8.0 (.2) 8.3 (.2) 8.2 (.2) 6.3 (.3) 1.9 .3 .5 24.4
***

Frequency 7.0 (.2) 6.8 (.1) 7.0 (.2) 6.7 (.2) 6.1 (.2) .8 .0 1.5 7.5
**

Involvement 9.3 (.2) 8.4 (.2) 8.8 (.2) 8.2 (.2) 6.7 (.4) 7.6
**
1.9 11.2
**
31.8
***

Community 7.3 (.3) 6.4 (.2) 7.6 (.2) 7.1 (.2) 5.6 (.3) 5.4
*
.4 .4 16.2
***

Frequency 6.0 (.1) 6.0 (.1) 6.2 (.1) 6.0 (.1) 5.9 (.3) .7 .0 .5 .9
Involvement 8.5 (.2) 8.0 (.2) 8.3 (.2) 7.9 (.2) 7.2 (.3) 2.5 .4 3.4 11.5
**

Parent-report Clinical Outcome
+
(%) Odds Ratio [95% CI]

ASD

15.2
***

[4.5, 52.2]

2.1
[.5, 9.1]

5.2
**

[1.5, 18.3]

122.9
***

[28.8, 524.6]

Non-ASD with Sensory Issues

25.5
***

[5.9, 110.2]

20.4
***

[4.6, 90.7]

3.4
[.7, 16.2]

8.6
*

[1.6, 45.0]

Variety
+ Based on DCQ (for reported ASD diagnosis or sensory-related concerns/diagnosis), SRS (T score ≥60 for ASD), and SEQ (total score >1SD above mean reflects
sensory issues/concerns)
85

86

CHAPTER 4
Impacts of Early Sensory Trajectories on School-age Outcomes: A Longitudinal Birth
Cohort Study (Study 3)
Abstract
Background: Sensory features are widely observed in but not exclusive to children with autism.
Past decades of research revealed associations between three specific sensory patterns (i.e.,
hyperresponsiveness [HYPER], hypo-responsiveness [HYPO], and sensory interests, repetitions,
and seeking behaviors [SIRS]) and other domains of behavior, such as adaptive/maladaptive
behaviors but the findings vary due to methodological differences. Moreover, there is a lack of
evidence on the longitudinal impact of sensory patterns on other behavioral domains. Thus, the
current study applied latent growth curve modeling, a person-centered approach, to
systematically examine the longitudinal effects of early sensory patterns on various school-age
outcomes in a birth cohort of young children.
Method: 1,517 caregivers completed surveys regarding their child’s sensory patterns across
three time-points from infancy to school age. A subsample (N=389) reported their child’s school-
age outcomes, including autism symptoms, adaptive/maladaptive behaviors, and activity
participation. Growth trajectories of sensory patterns were estimated with child’s sex, race, and
parent education levels as covariates. School-age outcomes were regressed onto latent growth
factors of sensory patterns to examine their direct, indirect, and total effects on distal outcomes.
Results: The change rate of HYPER was the most significant predictor of school-age outcomes,
particularly autism symptoms, adaptive/maladaptive outcomes, and children’s involvement
during participation. Differential impacts of HYPER and HYPO were observed on maladaptive

87

behavior: HYPER was more associated with internalizing behavior, while HYPO was more with
externalizing behavior. There were also indirect effects of sensory patterns at infancy via
developmental changes of HYPER and HYPO on several outcomes. Overall, the variations in
autism symptoms, maladaptive behavior, and socialization were explained to larger extents by
sensory trajectories as compared to other behavioral domains.
Conclusion: Children’s school-age outcomes were more impacted by the early trajectories of
sensory responsiveness, and their differential impacts highlighted the importance of early
detection and potential of tailored sensory intervention for improving specific long-term
outcomes. The significant impact of early sensory trajectories on later social-communication
outcomes may support the cascading effects theory, which has critical implications for early
intervention for children with autism.
Keywords: Developmental trajectories; sensory processing; school-age outcomes; birth cohort;
autism spectrum disorder

88

Introduction
Atypical sensory behaviors have been considered key features of autism spectrum
disorders (ASD) given their prevalence as high as 94% among young children with ASD
(Baranek et al., 2006; Ben-Sasson et al., 2009). However, like other autistic symptoms such as
social communication deficits, sensory features are also present among non-ASD populations,
and are continuously distributed across the general population (Constantino & Charman, 2016;
Jusilla et al., 2020). The estimated prevalence of elevated sensory features in population-based
samples of school-aged children was 5 to 8% (Ahn et al., 2004; Jusilla et al., 2020), which was
higher than the approximate 1% prevalence of social deficits among a school-aged general
population (Constantino & Todd, 2003). Moreover, the presence of sensory features, including
hyperresponsiveness (HYPER), hyporesponsiveness (HYPO) and sensory interests, repetitions,
and seeking behaviors (SIRS), has been empirically supported by psychometric validations in
populations of children with ASD and other developmental conditions, as well as typically
developing children (Little et al., 2011; Lee et al., under review). Given the ubiquitous presence
of sensory features in young children, especially those with various developmental disorders, it is
of great interest to understand their impact on other behavioral domains. In fact, the hypothesized
cascading effects of altered sensory processing on later social communication have been
empirically supported by behavioral evidence in children with ASD (Baranek et al., 2018;
Stevenson et al., 2018). Such findings may have important implications for providing more
effective tailored interventions or services that improve long-term outcomes for children with
sensory challenges and their families.
Over the past two decades, many studies have documented associations between sensory
features and other domains of behavior. In Table 4.1, we summarized the previous studies that

89

reported relevant correlation and/or regression findings by behavioral domains to get a general
picture of what we have known about the potential impact of sensory patterns among young
children. As shown in the tables, the existing evidence seems inconsistent. Take overall adaptive
behavior as an example, many studies reported negative associations with overall sensory
features (Rogers et al., 2003), as well as any of the three sensory patterns (Liss et al., 2006;
Ashburner, Ziviani, & Rodger, 2008; Williams et al., 2018) in children with ASD and other
conditions, although some studies showed null findings (Baker et al., 2008; Lane et al., 2010;
O'Donnell et al., 2012; McCormick et al., 2016). Mixed findings were also reported on other
domains of behavior, such as the subdomains of adaptive behavior, motor skills, maladaptive
behavior, autism symptoms, and activity participation. Given the methodological differences in
study designs, purposes, measures, and participant characteristics across studies, it is challenging
to draw conclusions from these findings. Notably, most of the findings were simple correlations
conducted using cross-sectional data, and thus they only provided evidence for the
interdependency of the two variables at one time-point, without considering the changes over
time, or potential effects of other variables such as demographics and other specific sensory
patterns. Previous evidence has indicated that the manifestation of sensory patterns could be
affected by children’s age (Ben-Sasson et al., 2009), sex (Jussila et al., 2020), race, and family
socioeconomic status (Gouze et al., 2009). Also, the three sensory patterns are often reported to
co-occur within children, resulting in various sensory subtypes (i.e., subgroups of children with
mixed patterns) (Ausderau et al., 2016; Uljarević et al., 2017). Therefore, it would be more
informative to include all sensory patterns as well as demographic covariates simultaneously in
predictive modeling (e.g., linear regressions and path analysis) for identifying the most
significant early sensory predictors of children’s later developmental outcomes.

90

Longitudinal evidence is critical for determining the impact of sensory patterns on later
developmental outcomes since the absence of temporal precedence of sensory patterns under
cross-sectional designs cannot establish causality. Among the few longitudinal studies, Williams
and colleagues (2018) found that increases in HYPO and HYPER in early development predicted
lower adaptive behaviors in both children with ASD and developmental delay (DD).
Interestingly, this same study found a reverse pattern for the impact of SIRS on adaptive
behavior, and only in children with DD. Another longitudinal study that included children with
ASD and other conditions (ages 2 to 8) found no significant association of sensory symptoms, as
measured by parent report questionnaires, to adaptive behavior after controlling for intellectual
ability and autism severity (McCormick et al., 2016). A recent study reported that early sensory
patterns and their changes across 14 to 23 months predicted later autism severity at 3 to 5 years
among children flagged as high-risk for ASD at infancy (Grzadzinski et al., 2020). Notably, they
found that autism severity was predicted by observed measures of hyporesponsiveness (along
with its change) and parent-reported measures of hyperresponsiveness, potentially indicating a
moderating effect of informant type. Despite these important findings, these studies adopted
variable-centered approaches to examine changes and their effects on later outcomes under the
assumption that the population is homogeneous for the main effects of interest, and thus, these
methodologies may not be able to fully address the variability in trajectories and their differential
associations with other variables (Laursen & Hoff, 2006).
To address the methodological issues discussed above, the current study aimed to
estimate individual trajectories of the three sensory patterns with latent growth curve modeling
(LGCM), a person-centered approach that accounts for both inter- and intra-person variability,
and to examine their effects on later school-age outcomes. Since we are interested in the parallel-

91

process trajectories of multiple sensory patterns (i.e., HYPER, HYPO, and SIRS) to reflect their
co-occurrence, LGCM serves as a more appropriate and flexible tool to examine such
simultaneous growth processes (Curran, Obeidat, & Losardo, 2010). It also allows us to examine
the predictive relationship among latent growth factors (e.g., slopes as predicted by intercepts) to
further investigate the indirect effects of intercepts and demographic predictors. Also,
corresponding to the concept that sensory patterns are continuously distributed across the general
population, the current study followed up on a birth cohort sample with various clinical
characteristics diagnosed at school-age, such as ASD or other developmental diagnoses, as well
as a typical-development comparison group. Since the manifestation of sensory patterns has been
shown to vary by chronological age, the age-homogeneous nature of our longitudinal sample,
studied at three specific time-points, would reduce the confounding effect of child’s age on the
result interpretations. We were specifically interested in the prediction effects of latent growth
factors of the three sensory patterns on school-age outcomes, including autism symptoms,
adaptive/maladaptive outcomes, and activity participation, while covarying demographic factors
(i.e., child’s sex, race, and parent education). In addition, the current study comprehensively
evaluated children’s school-age outcomes by considering children’s participation variety,
frequency, and involvement across home, school, and community contexts. The systematic
examination of sensory trajectories on these broader developmental outcomes may have clinical
implications for early interventions.
We explored the following questions: 1) Is there any predictive relationship among latent
growth factors (i.e., slopes as predicted by intercepts)? 2) What are the most significant sensory
predictors of children’s school-age outcomes, and how do their impacts vary across school-age
outcomes? Which aspects of school-age outcomes are most impacted? 3) Are there indirect

92

effects of the initial levels of sensory patterns and demographic variables on school-age
outcomes?
For the first question, results in Study 1 have revealed significant associations among
intercepts and slopes. Thus, we expected to further observe predictive relationship among the
intercept-slope pairs with higher associations (e.g., within-domain intercept-slope pairs). For the
remaining questions, we hypothesized that the growth factors of the three sensory patterns would
have differential impacts on school-age outcomes, according to the existing evidence reviewed in
Table 4.1. We also expected that sensory trajectories might account for more variability in
atypical behavior (e.g., autism symptoms and maladaptive behavior) than in functional abilities
outcomes (e.g., adaptive behavior and participation) given the shared nature of sensory patterns
and atypical behavior or symptoms. Furthermore, it was expected that parent education might
have indirect effects on distal outcomes via latent growth factors, given its significant impact on
sensory trajectories observed in Study 1.
Method
Participants and Procedure
The 1,517 participants included in the current study were initially ascertained and
recruited from birth registries in North Carolina. The caregivers were contacted at three time-
points (6-19 months T1; 3-4 years, T2; 6-7 years, T3) to report their child’s development on
various domains of behavior using paper forms and/or online questionnaires. Sensory patterns
were assessed with the First Years Inventory (FYI) version 3.1 at T1, and Sensory Experiences
Questionnaire (SEQ) version 2.1 at T2 and T3-Phase 1. At T2 and T3-Phase 1, parents also
completed the Developmental Concerns Questionnaire (DCQ) version 1.5 and Social

93

Responsiveness Scale (SRS) 2
nd
Edition to report their child’s developmental status including
ASD outcomes. At the second phase of T3, which was about 5 months after Phase-1 responses
were returned, a subset of families (N=465) was again contacted via email invitations to
complete the SRS, Vineland Adaptive Behavior Scales (VABS) 3
rd
Edition, and Participation
and Environment Measure-Children and Youth (PEM-CY). This subset included families who
reported any diagnosis/concerns at previous time-points (N=359) and a random sample of
families whose responses did not indicate concerns at any of the previous time-points (N=106).
We ended up receiving 389 responses (the current sample) across all the time-points. Out of the
389 responses at Phase-2, 312 (80%) had reported developmental diagnosis or concerns at
previous time-points. The demographic characteristics of the full sample and subsample can be
found in Table 3.2. Table 4.2 shows the descriptive statistics of the school-age outcome
measures for the subsample (N=389).
Measures
The FYI version 3.1 (FYIv3.1; Baranek et al., 2013)
The FYIv3.1 is a parent-report measure revised and expanded from a previous version
(FYIv2.0; Baranek et al., 2003) that was validated in a large community sample (Reznick et al.,
2007; Turner-Brown et al., 2013). The FYIv3.1 was designed to identify infants aged 6-16
months at risk for a later diagnosis of ASD, measuring the frequency of behaviors across three
broader domains, including social communication, sensory-regulatory functions, and motor
development. It uses a 5-point Likert scale. For this study, fourteen items (see Table 2.2 in
Chapter 2) related to sensory patterns were extracted for establishing sensory construct scores.
The SEQ version 2.1 (SEQv2.1; Baranek, 1999)

94

The SEQv2.1 is a 33-item parent questionnaire designed to measure the frequency of
behavioral responses to daily sensory experiences for children ages 1-12 years with a 5-point
Likert scale (higher scores indicate endorsement of more sensory features). It has good internal
consistency (Cronbach's α=.80) and test-retest reliability (r=.92) (Little et al., 2011), along with
good discriminative validity (Baranek et al, 2006). Fourteen items common to those in the
FYIv3.1 (varied in item contents but intended to measure the same behavior; see Table 2.2 in
Chapter 2) were extracted to establish sensory construct scores, while the total score was used for
evaluating the presence of sensory challenges.
School-age Outcomes
Three parent-report measures were administered via follow-up surveys to capture a
child’s clinical outcomes across development and functional outcomes at school age, including
autism symptoms, adaptive/maladaptive behaviors, and activity participation.
The DCQ version 1.5 (DCQv1.5; Reznick et al., 2005)
The DCQv1.5 is a parent-report measure with open-ended questions about whether a
parent or professional has been concerned about the child’s development and whether the child
has received any clinical diagnoses. It was used as one of the outcome measures in the validation
study of FYIv2.0 (Turner-Brown et al., 2013). The DCQv1.5 data at T2 and T3 were used for
outcome classification. Responses were coded to determine whether the child has had a
diagnosed developmental disability, including ASD, and/or any developmental concerns (see
Appendix 3 for the REDCap coding form).
The SRS-2 (Constantino & Gruber, 2012)

95

SRS-2 is a well-validated parent-report scale that measures deficits in social behavior
associated with ASD to determine levels of autism symptoms. It has excellent internal
consistency (Cronbach's α=.95) and test-retest reliability (r=.88-.95). The SRS data were
collected at T2 (preschool-age version) and T3 (school-age version). T-scores at T3 were used as
one of the distal outcome variables. A T-score ≥60 suggests clinically significant social
impairment, and therefore this cutoff was used to indicate an ASD diagnosis in the current study.
The VABS-3 (Sparrow, Cicchetti, & Saulnier, 2016)
The domain-level parent/caregiver form was used to assess children's adaptive behaviors
in three domains (communication, social, and daily living skills), as well as motor skills and
maladaptive behaviors (internalizing and externalizing behaviors). This form has an excellent
internal consistency coefficient (Cronbach's α=.97) and test-retest reliability (r=.87). The VABS
data were collected at T3 and the standardized scores in these six subdomains were used for
analysis.
The PEM-CY (Coster, Law, & Bedell, 2010)
The PEM-CY is a parent-report questionnaire that measures participation in 25 types of
daily activities under three contexts (home, school, and community). At T3, the caregiver was
asked to report on the child’s participation frequency (never=0 to daily=7) and their extent of
involvement (minimally involved=1 to very involved=5) for each activity. Average scores in the
variety of activities a child participated in, frequency, and involvement of participation for each
context were rescaled to have the same range of scores (0 to 10). The test-retest reliability was
moderate to good (r=.58-.84), and large to significant differences were found between groups
with and without disabilities on all subscales (Coster et al., 2011).

96

Data Analysis
First of all, trait scores of HYPER, HYPO, and SIRS based on item response theory
(IRT) were constructed upon the overlapping items extracted from the FYIv3.1 and SEQv2.1,
following a series of longitudinal invariance testing and adjustment for differential item
functioning (for details, see Chapter 2). Using the sensory trait scores as indicators, a series of
LGCMs was performed to ensure the appropriateness of the measurement (i.e., latent growth)
model: unconditional univariate models for each of the sensory construct followed by a
conditional multivariate model with demographic covariates and latent growth covariances (for
details, see Chapter 2). A CFI/TLI ≥.95 and an RMSEA <.06 were used to determine the
goodness of model fit (Hu & Bentler, 1999). Upon confirming that the measurement model fit
was satisfactory, school-age outcome variables were regressed on the latent growth factors of
sensory patterns as well as demographic covariates to examine their direct, indirect, and total
effects. The participants without school-age outcome data collected were treated as missing.
Since there was a total of 17 distal outcome variables, we estimated the LGCM model separately
for autism symptoms (SCI and RRB), adaptive behavior (COM, SOC, DLS, and MOT),
maladaptive behavior (INT and EXT), and participation variety, frequency, and involvement (in
home, school, and community contexts), while accounting for the effects of demographic
variables on these distal outcomes. The distal outcome variables that came from the same
measure, shared the same scale, and/or demonstrated high correlations (see Table 4.3) were
placed under the same model to account for their interdependence by specifying error
covariances among the endogenous variables.
For each model, we re-parameterized the covariances among latent growth factors so that
slope factors were regressed on intercept factors (Curran, Stice, & Chassin, 1997). This allowed

97

us to test the extent to which the starting point of each sensory construct in part predicts rates of
change as well as to examine the following indirect effects: 1) the effects of demographic
variables on distal outcomes via intercepts and slopes, and 2) the effects of intercepts on distal
outcomes via slopes. Finally, the pseudo-R
2
value for each distal outcome variable derived from
the models was used to evaluate to what extent its variation could be explained by the latent
growth factors and demographic covariates. The overall hypothesized model specification was
shown in Figure 4.1. All the analyses were conducted with robust maximum-likelihood
estimation in Mplus 8.4 (Muthén & Muthén, 2018).
Results
Latent Growth Trajectories of Sensory Patterns and Growth Factor Regressions
Univariate linear LGCMs showed strong evidence of linear growth across the three
sensory constructs (χ
2
(1)=.4 to .7, all CFI/TLI=1.00, RMSEA<.001) with large variability
(intercepts and slopes all <.001). The conditional multivariate model also demonstrated a good
fit (χ
2
(15)=27.4, CFI=.995, TLI=.988, RMSEA=.023). Weak to strong correlations (|r|=.17
to .72, all <.05) were observed among almost all the latent growth factors. When the growth
factor covariances were re-parametrized to be regressions (intercepts → slopes), two paths were
found significant: 1) the intercept of HYPER negatively predicted the slope of HYPER (β=-.45,
p=.001); 2) the intercept of SIRS positively predicted the slope of HYPO (β=.32, p=.002).
Another three paths were close to significance level: the intercepts of HYPER and HYPO
predicted the slope of SIRS in opposite directions (β=-.37 & .30, p<.07), and the intercept of
HYPO predicted the slope of HYPER (β=.21, p=.07).

98

Impacts of Sensory Trajectories on School-age Outcomes
The model fit statistics indicated good fits of the models that include school-age
outcomes (CFI=.97 to .99, TLI=.93 to .96, RMSEA=.03 to .04). The standardized beta
coefficients, standard errors, and approximate p values of the total effects of the predictors (i.e.,
latent growth factors and demographic variables) on school-age outcomes are reported in Table
4.4. The specific indirect effects are shown in Table 4.5. In addition, the significant effects of
latent growth factors by larger outcome domains (except for the variety of participated activities)
are visualized in Figure 4.2.
Overall, the growth factors of HYPER and HYPO posed larger impacts on school-age
outcomes. The slope of HYPER was a significant predictor of autism symptoms and all
adaptive/maladaptive outcomes (|β|=.16 to .44, p all <.05), as well as the frequency of home
participation and involvement in school/community activities (β=-.28 to -.22, p all <.01). The
slope of HYPO predicted autism symptoms, externalizing behavior, and frequency of school
participation (|β|=.35 to .61, p all <.05). The intercept of HYPO specifically predicted the social
communication domain of autism symptoms and the communication and socialization domains
of adaptive functioning (total effects |β|=.35 to .61, p all <.05). In addition, the intercept of
HYPO had indirect effects on socialization and internalizing behavior via the slope of HYPER
(indirect effects β=-.07 & .11, p<.05). Regarding the intercept of HYPER, its total effects on any
of the outcome variables were non-significant; however, its indirect effects via the slope of
HYPER were significant on autism symptoms, all adaptive outcomes, internalizing behavior,
frequency of home participation, and involvement in school/community activities (indirect
effects |β|=.10 to .21, p all <.05).

99

Compared to HYPER and HYPO, the growth factors of SIRS had relatively low impacts
on school-age outcomes. The only significant total effect was observed between the intercept of
SIRS and RRB (total effect β=.22, p <.05) but its direct effect was non-significant. Also, there
were indirect effects of the intercept of SIRS on the social-communication domain of autism
symptoms and externalizing behavior via the slope of HYPO (indirect effects β=.09 & .13,
p<.05).
Impacts of Demographic Variables on School-age Outcomes
Child’s sex significantly predicted autism symptoms, adaptive behavior, externalizing
behavior, as well as participation frequency and involvement across contexts (girls tended to
have better adaptive skills, more frequent participation with a higher level of involvement, as
well as fewer autism symptoms and externalizing behavior; total effects |β|=.11 to .23, p all
<.05). The only significant effect of child’s race was that non-white children tended to have more
social-communication impairments (total effect β=-.11, p<.05). Parent education was a
significant predictor of almost all outcome variables (total effects |β|=.11 to .25 for those p <.05)
with both direct and indirect effects observed. The major mediators of parent education on
school-age outcomes were the slopes of HYPER and HYPO (indirect effects |β|=.03 to .06, p all
<.05).
Total Variances of School-age Outcomes Explained
The pseudo-R
2
statistics revealed that the sensory growth factors along with demographic
covariates accounted for more variations in autism symptoms (SCI: 48.8%, RRB: 43.4%),
maladaptive behavior (INTB: 30.2%, EXTB: 35.7%), and socialization (35.3%) as a subdomain

100

of adaptive behavior. Less than 25% of the variances of the other adaptive and participation
outcomes were explained.
Discussion
The current study provided the first systematic investigation into the longitudinal impact
of early sensory patterns on school-age outcomes in a birth cohort sample of children with
various clinical conditions. LGCM revealed highly variable trajectories across sensory
constructs, supporting the notion that sensory patterns and their changes are continuously
distributed among children with diverse clinical characteristics. The multivariate LGCM allowed
us to examine the associations among latent growth factors. In the current study, we are
interested in the predictive relationship between intercepts and slopes and found that 1) higher
initial HYPER was followed by a smaller increase in HYPER, and 2) higher initial SIRS was
followed by more increase in HYPO. The first predictive relationship might reflect the overall
decreasing pattern of HYPER revealed by LGCM and might be related to ceiling effects (i.e., a
limited range of potential increases for those with high baseline scores). The second predictive
relationship is consistent with the significant association between the intercept of SIRS and the
slope of HYPO observed in Study 1. By constraining their relationship to be unidirectional, we
further demonstrated the potential precedence of SIRS. Such predictive relationships also
allowed us to examine how early sensory patterns impact later outcomes as mediated by their
change rates, which will be discussed below. In the following sections, we would unfold the
discussion of our findings on the impacts of sensory trajectories along with demographic
covariates by larger outcome domains.
Effects on Autism Symptoms

101

Previous studies using concurrent measures had reported that any of the sensory patterns
could be associated with autism symptoms, including social communication and RRBs (e.g., Liss
et al, 2006; Watson et al., 2011; Feldman et al., 2020). The current study provided longitudinal
evidence that sensory patterns beginning at infancy and their changes over time predicted autism
severity at school age. Our finding that more increases in HYPER and HYPO over time
predicted higher severity in both domains was consistent with the previous evidence that high-
risk siblings with a later diagnosis of ASD tended to show more increases in HYPER and HYPO,
but not SIRS, across the second year of life (Wolff et al., 2019). We also found that higher
HYPO and SIRS at infancy predicted more social-communication deficits, consistent with the
previous evidence that HYPO and SIRS at 20-24 months predicted later social deficits (Baranek
et al., 2018; Nowell et al., 2020); however, their impacts could be indirect (i.e., mediated by the
rates of change across early childhood). Specifically, the relationship between early SIRS and
later social-communication deficits was mediated by the change rate of HYPO, indicating the
potential use of SIRS as an early marker of later ASD-related challenges, which might be
mitigated by interventions targeting HYPO. Interestingly, we found that higher HYPER at
baseline predicted a smaller increase in HYPER, which further resulted in fewer autism
symptoms. This effect might be limited to children with non-ASD outcomes who tended to show
a decreasing pattern, as demonstrated in Study 1. Future research may further explore the
moderation effect of diagnostic groups. Across all school-age outcomes, autism symptoms
seemed to be best explained by sensory trajectories given the higher pseudo-R
2
values. These
findings together indicated that early sensory patterns and the change rates of sensory
responsiveness over time might be useful markers of ASD-related risk, adding to the previous

102

evidence of the precedence of altered sensory processing to later social-communication deficits
in children with ASD (Baranek et al., 2018; Thye et al., 2018).
Regarding the demographic covariates, boys tended to show more autism symptoms
(particularly RRBs given the significant direct effect of child’s sex) at school-age, generally
consistent with a previous population-based finding on school-aged children that boys showed
more RRBs but no sex difference was observed in social-communication deficits (Haraguchi et
al., 2019). The impacts of parents’ education levels on autism severity were both direct and
indirect: children of parents with lower education levels tended to show higher severity, which
was partially accounted for by the increase in atypical sensory responsiveness over time that was
predicted by lower parent education levels. Therefore, this might indicate the critical role of
family factors in shaping children’s developmental pathways toward ASD-related outcomes and
the necessity of family-centered approaches for addressing children’s sensory challenges early
on.
Effects on Maladaptive Behavior
The relationship between sensory hyperresponsiveness and internalizing behavior, such
as anxiety and depression, has been well supported by both behavioral and neurological evidence
from human and animal studies (Green & Ben-Sasson, 2010). Our finding also supported the
specific contribution of HYPER to internalizing behavior by showing that the increase of
HYPER over time predicted more internalizing behavior at school age. Also, the impact of the
initial level of HYPER and HYPO on internalizing behavior was mediated by the increase of
HYPER, highlighting HYPER as an important target of early intervention for preventing or
reducing internalizing behavior.

103

Interestingly, although the slope of HYPER also predicted externalizing behavior, the
slope of HYPO seemed a more significant predictor. The relationship between sensory patterns
and externalizing behavior has been relatively understudied as compared to internalizing
behavior. The connection between HYPO and externalizing behavior observed in the current
study might be related to the common nature between a lack of response to sensory stimuli,
inattention/failure to disengage and follow instructions. Previous physiological evidence also
indicated an association between under-arousal (low electrodermal responses to stimuli) and
externalizing behavior in children with ASD (Baker et al., 2018), which might be a potential
underlying mechanism for the observed behavioral association. Overall, our finding did support
the connection between atypical sensory responsiveness and overall maladaptive behavior as
supported by previous literature on children with problem behaviors (Ausderau et al., 2016; Ben-
Sasson et al., 2017; Lane & Reynolds, 2019). The differential associations (i.e., HYPER-
internalizing versus HYPO-externalizing) observed in the current study merit further
investigations of their underlying etiology, which will eventually contribute to more targeted
interventions.
Regarding the demographic predictors, it is not surprising that boys tended to show more
externalizing behavior, as demonstrated by previous research (Eme, 2016). The finding that
higher parent education directly predicted less child’s internalizing behavior was consistent with
the previous evidence that higher parent education, but not family income, was associated with
their child’s lower internalizing symptoms (Merz, Tottenham, & Noble, 2018). We also found
that the more increase of HYPER as predicted by lower parent education was associated with
more child’s internalizing behavior at school age. Such an indirect effect of parent education was
also observed in externalizing behavior, with the increase of HYPO as a mediator. As discussed

104

in the previous section, parent education seems to play a critical role in shaping children’s
trajectories of sensory responsiveness, which further impact their children’s maladaptive
outcomes. Thus, it is critical for early sensory intervention to take family factors into
considerations to mitigate children’s maladaptive outcomes.
Effects on Adaptive Behavior and Motor Skills
Overall, we found that larger increases of HYPER predicted lower adaptive and motor
skills at school age across the three subdomains. Communication and socialization were
additionally predicted by the initial level of HYPO, in line with our finding with the social-
communication domain of autism symptoms. Previous studies have documented the relationship
between HYPO and reduced social communication (Watson et al., 2011; Baranek et al., 2013),
and our current finding further supports that HYPO might be an important early predictor of later
social-adaptive functioning. Furthermore, the intensifying trajectory of HYPER (including both
direct and indirect effects) might interfere with children’s abilities to engage in activities and
acquire adaptive skills under sensory-demanding contexts. Thus, HYPER might be a critical
intervention target for ensuring children’s learning opportunities.
Our results showed that girls tended to have better adaptive and motor skills; such sex
difference was also demonstrated in previous studies on children with ASD and other conditions
at this age (Mandic-Maravic et al., 2015; Mahendiran et al., 2019). The finding that higher parent
education directly and indirectly predicted most of the adaptive skills of their child is consistent
with the evidence that family factors are important to children’s adaptive development (Hauser-
Cram et al., 1999; Hosokawa & Katsura, 2017), although we found such effects were specific to
the social-communication and motor domains.

105

Effects on Activity Participation
Compared to other outcome domains, sensory trajectories along with demographic
covariates accounted for less variations in the participation outcomes across contexts (smaller
than 25% of variations). This indicated that other explanatory factors not included in the current
study, such as family income as another critical indicator of family socioeconomic status (SES)
in addition to parent education, intervention or services received, and environmental
support/barrier, may be more important contributors (Law et al., 2006; Bedell et al., 2013) to be
examined in future research.
The variety of activities in which children participated was not significantly predicted by
any of the sensory growth factors and demographic variables, except that higher parent education
predicted more variety of community activities. In contrast, the frequency and involvement of
children’s participation were more impacted by the change rates of HYPER/HYPO, child’s sex,
and parent education. Particularly, more increase in HYPER over time predicted less frequent
home participation and less involvement in school/community activities. As discussed above, the
intensifying HYPER might prevent children from engaging in the participated activities and such
negative impact might be more evident in school/community contexts with higher environmental
demands. This finding was partially consistent with previous longitudinal evidence that the
increase of HYPER predicted reduced participation in home activities (routine errands) and
community activities among children with ASD (Little et al., 2015). However, our study further
demonstrated that the involvement, instead of frequency, might be more impacted by HYPER,
and thus it is critical to evaluate different aspects of children’s participation. Also, we found that
more increase in HYPO over time predicted less frequent school participation. A previous study
also found that HYPO, particularly auditory filtering difficulties, was associated with poor

106

school performance in children with ASD (Ashburner et al., 2008). The failure to follow
instructions at school, which is often a sensory-demanding environment, might account for
children’s less frequent participation. Thus, environmental adaptations or support at school may
help facilitate the participation of children with elevated HYPO. Furthermore, higher initial
HYPER predicted decreases in HYPER over time, and thus resulted in fewer challenges in
participation. Again, this indirect effect might be limited to children with non-ASD outcomes
that merit further investigations. Regardless, these findings together indicated that sensory
responsiveness during early childhood might be more critical targets of interventions for
improving children’s participation outcomes.
Regarding the demographic predictors, we observed that parent education was a
significant predictor of children’s participation outcomes by school-age. Previous cross-sectional
evidence has demonstrated connections between family SES and participation in physical
activities, suggesting the critical role of parental support in broadening children’s opportunities
for participation (Brockman et al., 2009; Mutz & Albrecht, 2017). The predictive relationship
between parent education and later participation outcomes might be related to parental
beliefs/expectations and associated parenting behaviors across children’s development that
contribute to children’s later overall performance (Davis-Kean et al., 2005). Finally, we found
that on average girls had more frequent participation and more involvement in activities across
contexts. Their better adaptive skills might contribute to higher levels of participation, meriting
further research examining the mediating effect of adaptive functions on such associations.
Limitations
A limitation of the current study is the use of only parent-report measures, although all
measures used in this large sample were well-validated. Regarding the associations between

107

sensory patterns and other domains of behavior, previous studies have demonstrated that the
results might depend upon whether the sensory features were assessed by parent-report or
clinician-observational measures (Williams et al., 2018; Grzadzinski et al., 2020). Thus, the
current findings might not be generalizable to sensory features measured via other types of
informants. Also, a larger sample of children with school-age outcome data is needed to further
understand the moderating effect of diagnosis. While we attempted to reflect the concept of
sensory patterns as continuously distributed within a population, associations moderated by
diagnostic grouping might be present and thus have implications for clinical practices that still
rely on diagnostic categories. Besides, the diagnosis could be further ascertained with gold-
standard diagnostic measures in future studies to derive more robust results. Another limitation is
not being able to include some important covariates into the analysis, such as children’s IQ,
family income, and factors related to environmental support and intervention, which might
explain more variations in higher-level outcomes such as activity participation. These factors
might also be important moderators or mediators of the association between sensory trajectories
and distal outcomes that could be explored in future research.
Conclusion
The current study provided the first systematic evidence on the effects of early
longitudinal sensory patterns on broader school-age outcomes in a community sample of children
with and without ASD and developmental disorders. Autism symptoms and maladaptive
behavior were observed to be more impacted by sensory trajectories as compared to other
behavioral domains. Also, the significant impacts of sensory patterns on social-communication
deficits and social-adaptive functioning might support the notion that sensory features produce
cascading effects on social development in children with ASD (Baranek et al., 2018). In general,

108

increases in atypical sensory responsiveness, especially HYPER, during early childhood
predicted poor school-age outcomes and thus might be targeted for early intervention.
Furthermore, we observed their differential impacts on challenging behaviors: HYPO tended to
be more associated with externalizing behavior while HYPER was more with internalizing
behavior. This thus highlights the potential of tailored interventions for addressing sensory
development as well as improving specific long-term outcomes. The indirect effects of early
sensory patterns on school-age outcomes via the developmental changes of HYPER and HYPO
might indicate the potential of early detection followed by intervention to “flatten” the
trajectories of atypical sensory responsiveness, further leading to more optimal school-age
outcomes. Further research is needed to clarify the differential roles of sensory patterns and their
underlying etiology for providing more effective interventions, while considering child and
family factors, across development among children with various clinical characteristics.

109

References
Ashburner, J., Ziviani, J., & Rodger, S. (2008). Sensory processing and classroom emotional,
behavioral, and educational outcomes in children with autism spectrum
disorder. American Journal of Occupational Therapy, 62(5), 564-573.
Ausderau, K. K., Furlong, M., Sideris, J., Bulluck, J., Little, L. M., Watson, L. R., ... & Baranek,
G. T. (2014). Sensory subtypes in children with autism spectrum disorder: latent profile
transition analysis using a national survey of sensory features. Journal of Child
Psychology and Psychiatry, 55(8), 935-944.
Ausderau, K. K., Sideris, J., Little, L. M., Furlong, M., Bulluck, J. C., & Baranek, G. T. (2016).
Sensory subtypes and associated outcomes in children with autism spectrum disorders.
Autism Research, 9(12), 1316-1327.
Baker, J. K., Fenning, R. M., Erath, S. A., Baucom, B. R., Moffitt, J., & Howland, M. A. (2018).
Sympathetic under-arousal and externalizing behavior problems in children with autism
spectrum disorder. Journal of Abnormal Child Psychology, 46(4), 895-906.
Baker, A. E., Lane, A., Angley, M. T., & Young, R. L. (2008). The relationship between sensory
processing patterns and behavioural responsiveness in autistic disorder: A pilot
study. Journal of Autism and Developmental Disorders, 38(5), 867-875.
Baranek, G. T. (1999). Sensory Experiences Questionnaire (SEQ). Unpublished manuscript,
University of North Carolina at Chapel Hill.
Baranek, G. T., David, F. J., Poe, M. D., Stone, W. L., & Watson, L. R. (2006). Sensory
Experiences Questionnaire: discriminating sensory features in young children with autism,

110

developmental delays, and typical development. Journal of Child Psychology and
Psychiatry, 47(6), 591-601.
Baranek, G. T., Watson, L. R., Crais, E. R., Turner-Brown, L. & Reznick, J. S. (2013). First
Years Inventory (FYI) 3.1. Chapel Hill, NC: University of North Carolina at Chapel Hill.
Baranek, G. T., Woynaroski, T. G., Nowell, S., Turner-Brown, L. M., DuBay, M., Crais, E. R.,
& Watson, L. R. (2018). Cascading effects of attention disengagement and sensory
seeking on social symptoms in a community sample of infants at-risk for a future
diagnosis of autism spectrum disorder. Developmental Cognitive Neuroscience, 29, 30-40.
Bar‐Shalita, T., Vatine, J. J., & Parush, S. (2008). Sensory modulation disorder: a risk factor for
participation in daily life activities. Developmental Medicine and Child
Neurology, 50(12), 932-937.
Bedell, G., Coster, W., Law, M., Liljenquist, K., Kao, Y. C., Teplicky, R., ... & Khetani, M. A.
(2013). Community participation, supports, and barriers of school-age children with and
without disabilities. Archives of Physical Medicine and Rehabilitation, 94(2), 315-323.
Ben-Sasson, A., Hen, L., Fluss, R., Cermak, S. A., Engel-Yeger, B., & Gal, E. (2009). A meta-
analysis of sensory modulation symptoms in individuals with autism spectrum disorders.
Journal of Autism and Developmental Disorders, 39(1), 1-11.
Ben-Sasson, A., Soto, T. W., Heberle, A. E., Carter, A. S., & Briggs-Gowan, M. J. (2017). Early
and concurrent features of ADHD and sensory over-responsivity symptom
clusters. Journal of Attention Disorders, 21(10), 835-845.

111

Boyd, B. A., Baranek, G. T., Sideris, J., Poe, M. D., Watson, L. R., Patten, E., & Miller, H.
(2010). Sensory features and repetitive behaviors in children with autism and
developmental delays. Autism Research, 3(2), 78-87.
Brockman, R., Jago, R., Fox, K. R., Thompson, J. L., Cartwright, K., & Page, A. S. (2009). "Get
off the sofa and go and play": Family and socioeconomic influences on the physical
activity of 10-11 year old children. BMC Public Health, 9(1), 1-7.
Carr, J. L., Agnihotri, S., & Keightley, M. (2010). Sensory processing and adaptive behavior
deficits of children across the fetal alcohol spectrum disorder continuum. Alcoholism:
Clinical and Experimental Research, 34(6), 1022-1032.
Chuang, I. C., Tseng, M. H., Lu, L., & Shieh, J. Y. (2012). Sensory correlates of difficult
temperament characteristics in preschool children with autism. Research in Autism
Spectrum Disorders, 6(3), 988-995.
Constantino, J. N., & Charman, T. (2016). Diagnosis of autism spectrum disorder: reconciling
the syndrome, its diverse origins, and variation in expression. The Lancet Neurology,
15(3), 279-291.
Constantino J. N. & Gruber C. P. (2012). Social Responsiveness Scale–Second Edition (SRS-2).
Torrance, CA: Western Psychological Services.
Constantino, J. N., & Todd, R. D. (2003). Autistic traits in the general population: a twin study.
Archives of General Psychiatry, 60(5), 524-530.

112

Coster, W., Bedell, G., Law, M., Khetani, M.A., Teplicky, R., Liljenquist, K., ... & Kao, Y.C.
(2011). Psychometric evaluation of the Participation and Environment Measure for
Children and Youth. Developmental Medicine & Child Neurology, 53(11), 1030-1037.
Coster W., Law M., & Bedell G. (2010). Participation and Environment Measure for Children
and Youth. Boston, MA: Boston University.
Curran, P. J., Obeidat, K., & Losardo, D. (2010). Twelve frequently asked questions about
growth curve modeling. Journal of Cognition and Development, 11(2), 121-136.
Curran, P. J., Stice, E., & Chassin, L. (1997). The relation between adolescent alcohol use and
peer alcohol use: a longitudinal random coefficients model. Journal of Consulting and
Clinical Psychology, 65(1), 130.
Davis-Kean, P. E. (2005). The influence of parent education and family income on child
achievement: the indirect role of parental expectations and the home environment. Journal
of family psychology, 19(2), 294.
Dellapiazza, F., Michelon, C., Oreve, M. J., Robel, L., Schoenberger, M., Chatel, C., ... &
Baghdadli, A. (2020). The impact of atypical sensory processing on adaptive functioning
and maladaptive behaviors in autism spectrum disorder during childhood: results from the
ELENA cohort. Journal of Autism and Developmental Disorders, 50(6), 2142-2152.
Eme, R. (2016). Sex differences in the prevalence and expression of externalizing behavior. In T.
P. Beauchaine & S. P. Hinshaw (Eds.), Oxford Library of Psychology. The Oxford
Handbook of Externalizing Spectrum Disorders (p. 239–263). Oxford University Press.

113

Engel-Yeger, B., & Ziv-On, D. (2011). The relationship between sensory processing difficulties
and leisure activity preference of children with different types of ADHD. Research in
Developmental Disabilities, 32(3), 1154-1162.
Feldman, J. I., Cassidy, M., Liu, Y., Kirby, A. V., Wallace, M. T., & Woynaroski, T. G. (2020).
Relations between Sensory Responsiveness and Features of Autism in Children. Brain
Sciences, 10(11), 775.
Goldsmith, H. H., Van Hulle, C. A., Arneson, C. L., Schreiber, J. E., & Gernsbacher, M. A.
(2006). A population-based twin study of parentally reported tactile and auditory
defensiveness in young children. Journal of Abnormal Child Psychology, 34(3), 378-392.
Green, S. A., & Ben-Sasson, A. (2010). Anxiety disorders and sensory over-responsivity in
children with autism spectrum disorders: is there a causal relationship? Journal of Autism
and Developmental Disorders, 40(12), 1495-1504.
Green, S. A., Ben-Sasson, A., Soto, T. W., & Carter, A. S. (2012). Anxiety and sensory over-
responsivity in toddlers with autism spectrum disorders: Bidirectional effects across
time. Journal of Autism and Developmental Disorders, 42(6), 1112-1119.
Gourley, L., Wind, C., Henninger, E. M., & Chinitz, S. (2013). Sensory processing difficulties,
behavioral problems, and parental stress in a clinical population of young
children. Journal of Child and Family Studies, 22(7), 912-921.
Grzadzinski, R., Donovan, K., Truong, K., Nowell, S., Lee, H., Sideris, J., ... & Watson, L. R.
(2020). Sensory reactivity at 1 and 2 years old is associated with ASD severity during the
preschool years. Journal of Autism and Developmental Disorders, 50(11), 3895-3904.

114

Gunn, T. E., Tavegia, B. D., Houskamp, B. M., McDonald, L. B., Bustrum, J. M., Welsh, R. K.,
& Mok, D. S. (2009). Relationship between sensory deficits and externalizing behaviors in
an urban, Latino preschool population. Journal of Child and Family Studies, 18(6), 653-
661.
Haraguchi, H., Stickley, A., Saito, A., Takahashi, H., & Kamio, Y. (2019). Stability of autistic
traits from 5 to 8 years of age among children in the general population. Journal of Autism
and Developmental Disorders, 49(1), 324-334.
Hauser‐Cram, P., Warfield, M. E., Shonkoff, J. P., Krauss, M. W., Upshur, C. C., & Sayer, A.
(1999). Family influences on adaptive development in young children with Down
syndrome. Child Development, 70(4), 979-989.
Hilton, C., Graver, K., & LaVesser, P. (2007). Relationship between social competence and
sensory processing in children with high functioning autism spectrum disorders. Research
in Autism Spectrum Disorders, 1(2), 164-173.
Hochhauser, M., & Engel-Yeger, B. (2010). Sensory processing abilities and their relation to
participation in leisure activities among children with high-functioning autism spectrum
disorder (HFASD). Research in Autism Spectrum Disorders, 4(4), 746-754.
Hosokawa, R., & Katsura, T. (2017). A longitudinal study of socioeconomic status, family
processes, and child adjustment from preschool until early elementary school: the role of
social competence. Child and Adolescent Psychiatry and Mental Health, 11(1), 1-28.
Jasmin, E., Couture, M., McKinley, P., Reid, G., Fombonne, E., & Gisel, E. (2009). Sensori-
motor and daily living skills of preschool children with autism spectrum
disorders. Journal of Autism and Developmental Disorders, 39(2), 231-241.

115

Jussila, K., Junttila, M., Kielinen, M., Ebeling, H., Joskitt, L., Moilanen, I., & Mattila, M.L.
(2019). Sensory Abnormality and Quantitative Autism Traits in Children With and
Without Autism Spectrum Disorder in an Epidemiological Population. Journal of Autism
and Developmental Disorders, 1-9.
Kojovic, N., Ben Hadid, L., Franchini, M., & Schaer, M. (2019). Sensory processing issues and
their association with social difficulties in children with autism spectrum
disorders. Journal of Clinical Medicine, 8(10), 1508.
Laursen, B., & Hoff, E. (2006). Person-centered and variable-centered approaches to
longitudinal data. Merrill-Palmer Quarterly (1982-), 377-389.
Lane, S. J., & Reynolds, S. (2019). Sensory over-responsivity as an added dimension in
ADHD. Frontiers in Integrative Neuroscience, 13, 40.
Lane, A. E., Young, R. L., Baker, A. E., & Angley, M. T. (2010). Sensory processing subtypes in
autism: Association with adaptive behavior. Journal of Autism and Developmental
Disorders, 40(1), 112-122.
Law, M., King, G., King, S., Kertoy, M., Hurley, P., Rosenbaum, P., ... & Hanna, S. (2006).
Patterns of participation in recreational and leisure activities among children with complex
physical disabilities. Developmental Medicine & Child Neurology, 48(5), 337-342.
Lee H., Chen Y., Sideris J., Watson L. R., Crais E. R., & Baranek G. T. (under review: minor
revision). Sensory Features of Young children from a Large Community Sample: Latent
Factor Structures of Sensory Experiences Questionnaire (v2.1). American Journal of
Occupational Therapy.

116

Little, L. M., Ausderau, K., Sideris, J., & Baranek, G. T. (2015). Activity participation and
sensory features among children with autism spectrum disorders. Journal of Autism and
Developmental Disorders, 45(9), 2981-2990.
Little, L. M., Freuler, A. C., Houser, M. B., Guckian, L., Carbine, K., David, F. J., & Baranek, G.
T. (2011). Psychometric validation of the sensory experiences questionnaire. American
Journal of Occupational Therapy, 65(2), 207-210.
Liss, M., Saulnier, C., Fein, D., & Kinsbourne, M. (2006). Sensory and attention abnormalities in
autistic spectrum disorders. Autism, 10(2), 155-172.
Mahendiran, T., Dupuis, A., Crosbie, J., Georgiades, S., Kelley, E., Liu, X., ... & Brian, J.
(2019). Sex differences in social adaptive function in autism spectrum disorder and
attention-deficit hyperactivity disorder. Frontiers in Psychiatry, 10, 607.
Mandic-Maravic, V., Pejovic-Milovancevic, M., Mitkovic-Voncina, M., Kostic, M., Aleksic-Hil,
O., Radosavljev-Kircanski, J., ... & Lecic-Tosevski, D. (2015). Sex differences in autism
spectrum disorders: does sex moderate the pathway from clinical symptoms to adaptive
behavior? Scientific Reports, 5(1), 1-8.
Mazurek, M. O., Vasa, R. A., Kalb, L. G., Kanne, S. M., Rosenberg, D., Keefer, A., ... &
Lowery, L. A. (2013). Anxiety, sensory over-responsivity, and gastrointestinal problems
in children with autism spectrum disorders. Journal of Abnormal Child Psychology, 41(1),
165-176.
McCormick, C., Hepburn, S., Young, G. S., & Rogers, S. J. (2016). Sensory symptoms in
children with autism spectrum disorder, other developmental disorders and typical
development: a longitudinal study. Autism, 20(5), 572-579.

117

Merz, E. C., Tottenham, N., & Noble, K. G. (2018). Socioeconomic status, amygdala volume,
and internalizing symptoms in children and adolescents. Journal of Clinical Child &
Adolescent Psychology, 47(2), 312-323.
Mikami, M., Hirota, T., Takahashi, M., Adachi, M., Saito, M., Koeda, S., ... & Yamada, J.
(2021). Atypical Sensory Processing Profiles and Their Associations With Motor
Problems In Preschoolers With Developmental Coordination Disorder. Child Psychiatry
and Human Development, 52(2), 311-320.
Muthén, L. K., & Muthén, B. (2018). Mplus. The Comprehensive Modelling Program for
Applied Researchers: user’s guide, 5.
Mutz, M., & Albrecht, P. (2017). Parents’ social status and children’s daily physical activity: The
role of familial socialization and support. Journal of Child and Family Studies, 26(11),
3026-3035.
Nowell, S. W., Watson, L. R., Crais, E. R., Baranek, G. T., Faldowski, R. A., & Turner-Brown,
L.M. (2020). Joint attention and sensory-regulatory features at 13 and 22 months as
predictors of preschool language and social-communication outcomes. Journal of Speech,
Language, and Hearing Research, 63(9), 3100-3116.
O’Donnell, S., Deitz, J., Kartin, D., Nalty, T., & Dawson, G. (2012). Sensory processing,
problem behavior, adaptive behavior, and cognition in preschool children with autism
spectrum disorders. American Journal of Occupational Therapy, 66(5), 586-594.
Pfeiffer, B., Kinnealey, M., Reed, C., & Herzberg, G. (2005). Sensory modulation and affective
disorders in children and adolescents with Asperger’s disorder. American Journal of
Occupational Therapy, 59(3), 335-345.

118

Reynolds, S., Bendixen, R. M., Lawrence, T., & Lane, S. J. (2011). A pilot study examining
activity participation, sensory responsiveness, and competence in children with high
functioning autism spectrum disorder. Journal of Autism and Developmental
Disorders, 41(11), 1496-1506.
Reznick, J. S., Baranek, G. T., Watson, L. R., & Crais, E. R. (2005). Developmental Concerns
Questionnaire. Chapel Hill NC: University of North Carolina.
Rogers, S. J., Hepburn, S., & Wehner, E. (2003). Parent reports of sensory symptoms in toddlers
with autism and those with other developmental disorders. Journal of Autism and
Developmental Disorders, 33(6), 631-642.
Schulz, S. E., & Stevenson, R. A. (2019). Sensory hypersensitivity predicts repetitive behaviours
in autistic and typically-developing children. Autism, 23(4), 1028-1041.
Sparrow, S. S., Cicchetti, D. V., & Balla, D. A. (2016). Vineland-III: Vineland Adaptive
Behavior Scales (3rd ed.). San Antonio, TX: Pearson.
Stevenson, R. A., Segers, M., Ncube, B. L., Black, K. R., Bebko, J. M., Ferber, S., & Barense,
M. D. (2018). The cascading influence of multisensory processing on speech perception in
autism. Autism, 22(5), 609-624.
Surgent, O. J., Walczak, M., Zarzycki, O., Ausderau, K., & Travers, B. G. (2020). IQ and
sensory symptom severity best predict motor ability in children with and without autism
spectrum disorder. Journal of Autism and Developmental Disorders, 1-12.

119

Thye, M. D., Bednarz, H. M., Herringshaw, A. J., Sartin, E. B., & Kana, R. K. (2018). The
impact of atypical sensory processing on social impairments in autism spectrum
disorder. Developmental Cognitive Neuroscience, 29, 151-167.
Tomchek, S. D., Little, L. M., & Dunn, W. (2015). Sensory pattern contributions to
developmental performance in children with autism spectrum disorder. American Journal
of Occupational Therapy, 69(5), 6905185040p1-6905185040p10.
Uljarević, M., Baranek, G., Vivanti, G., Hedley, D., Hudry, K., & Lane, A. (2017).
Heterogeneity of sensory features in autism spectrum disorder: Challenges and
perspectives for future research. Autism Research, 10(5), 703-710.
Watson, L. R., Patten, E., Baranek, G. T., Poe, M., Boyd, B. A., Freuler, A., & Lorenzi, J.
(2011). Differential associations between sensory response patterns and language, social,
and communication measures in children with autism or other developmental disabilities.
Journal of Speech, Language, and Hearing Research, 54(6): 1562–1576.
Wigham, S., Rodgers, J., South, M., McConachie, H., & Freeston, M. (2015). The interplay
between sensory processing abnormalities, intolerance of uncertainty, anxiety and
restricted and repetitive behaviours in autism spectrum disorder. Journal of Autism and
Developmental Disorders, 45(4), 943-952.
Williams, K. L., Kirby, A. V., Watson, L. R., Sideris, J., Bulluck, J., & Baranek, G. T. (2018).
Sensory features as predictors of adaptive behaviors: A comparative longitudinal study of
children with autism spectrum disorder and other developmental disabilities. Research in
Developmental Disabilities, 81, 103-112.

120

Wolff, J. J., Dimian, A. F., Botteron, K. N., Dager, S. R., Elison, J. T., Estes, A. M., ... & Gu, H.
(2019). A longitudinal study of parent‐reported sensory responsiveness in toddlers at‐risk
for autism. Journal of Child Psychology and Psychiatry, 60(3), 314-324.

121

Tables and Figures
Table 4.1. Summary of findings on the associations between sensory patterns and other domains of behavior (+/- indicate positive/
negative associations)
a. Adaptive Behavior/Skills
Studies Samples Age Analysis Adaptive
Behavior
(Overall)
Communication
Skills
Social
Skills
Daily Living
Skills
Motor
Skills
Rogers et al.,
2003
ASD, Fragile X
Syndrome, DD,
TD
12-50
months
Correlation Total score (-) -- -- -- --
Jasmin et al.,
2009
ASD 3-4 years Correlation -- -- -- HYPER (-) SIRS (-) with
gross motor,
HYPER (-) with
fine motor
O'Donnell et
al., 2012
ASD 3-4 years Correlation No
association
-- -- -- --
Kojovic et al.,
2019
ASD, TD 3-6 years Correlation Total score (-) No
association
Total score
(-)
Total score
(-)
No association
Tomchek et
al., 2015
ASD 50 months
(mean)
Regression -- HYPO/HYPER
(-), SIRS (-;
receptive)
SIRS (-) -- SIRS (-) with
gross/fine
motor
Watson et al.,
2011
ASD, DD 52 months
(mean)
Regression
(covarying sex,
diagnostic group &
nonverbal MA)
-- HYPO (-), SIRS
(-)
HYPO (-) -- --
McCormick et
al., 2016
ASD, DD, TD 2-8 years Regression
(covarying verbal MA
& diagnostic group)
No
association
-- -- -- --
Baker et al.,
2008
ASD 33-101
months
Correlation No
association
No
association
No
association
No
association
--
Lane et al.,
2010
ASD 33-115
months
Correlation No
association
Low
Energy/Weak (-)
No
association
No
association
No association
121

122

Williams et al.,
2018
ASD, DD 2-12 years
(2 time-
points; 3
years apart)
Regression
(covarying sex, CA,
IQ, family SES, and
service use)
HYPO (-; more
significant in DD);
HYPER (-); SIRS
(- ASD; + DD)
SIRS (- ASD; +
DD)
HYPO (-);
SIRS
(- ASD; +
DD)
HYPO (-),
HYPER (-);
SIRS (- ASD;
+ DD)
--
Dellapiazza et
al., 2020
ASD 3-11 years Regression
(covarying CA & IQ)

Null results of ANCOVA across ABC domains; group analysis revealed associations (-)
between SIRS and ABC domains
Carr et al.,
2010
Fetal Alcohol
Spectrum Disorder
3-14 years Correlation -- -- -- Total score
(-)
--
Mikami et al.,
2021
Developmental
coordination
disorder (DCD), TD
5 years Regression -- -- -- -- HYPER (-),
SIRS (+) in
DCD
Liss et al.,
2006
ASD 102 months
(mean)
Correlation HYPO (-), SIRS
(-)
SIRS (-) HYPER (-) HYPO (-),
SIRS (-)
--
Ashburner et
al., 2008
ASD 6-10 years Correlation HYPO (-) -- -- -- --
Surgent et al.,
2020
ASD, TD,
Intermediate
behavioral profile
6-10 years Correlation,
regression
-- -- -- -- Total score (-)
Feldman et al.,
2020
ASD, non-ASD 8-18 years Correlation,
regression
-- HYPO (-),
SIRS (-)
HYPO (-),
HYPER (-)
HYPO (-),
HYPER (-)
--

b. Autism symptoms and maladaptive behavior
Studies Samples Age Analysis Autism Symptoms Maladaptive
Behavior
(Overall)
Internalizing
Behavior
Externalizing
Behavior
Nowell et
al., 2020
High-risk infants 13-22
months
Regression (path
analysis)
HYPO (+), SIRS (+)
with social
communication
-- -- --
Grzadzinski
et al., 2020

Children at high
risk for ASD (44%
with ASD or other
diagnosis at 3-5 years)
14-23
months
Correlation,
regression
Observed HYPO (+),
parent-report HYPER
(+) with RRB
-- -- --
Goldsmith et
al., 2006
Population-based
sample
11-36
months
Correlation -- -- HYPER (+) HYPER (+)
122

123

Rogers et al.,
2003
ASD, Fragile X
Syndrome, DD, TD
12-50
months
Correlation Total score (+; with
RRB in ASD; with social
communication in FXS)
-- -- --
Green et al.,
2012
ASD 28 months
(mean)
Correlation,
regression (path
analysis)
HYPER (+) -- HYPER (+) --
O'Donnell et
al., 2012
ASD 3-4 years Correlation -- Total score (+) Total score (+) Total score (+)
Gourley et
al., 2013
Clinical sample 3-5 years Correlation -- -- Total score (+) Total score (+)
Kojovic et
al., 2019
ASD, TD 3-6 years Correlation Total (+), HYPO (+),
SIRS (+)
-- -- --
Gunn et al.,
2009
TD 4 years
(mean)
Correlation -- -- -- All constructs (+)
Boyd et al.,
2010
ASD, DD ≈ 50 months
(mean)
Regression
(covarying MA)
HYPER (+) with RRB -- -- --
Chuang et
al., 2012
ASD 48-84
months
Regression
(covarying
demographics)
-- HYPER (+),
HYPO (+),
SIRS (+)
-- --
Watson et
al., 2011
ASD, DD 52 months
(mean)
Regression
(covarying sex,
diagnostic group &
nonverbal MA)
HYPO (+), SIRS (+;
ASD)
-- -- --
Baker et al.,
2008
ASD 33-101
months
Correlation HYPO (+), SIRS (+) HYPER (+),
HYPO (+),
SIRS (+)
HYPER
[visual/auditory/mov
ement sensitivity] (+)
HYPER
[visual/auditory
sensitivity] (+)
Lane et al.,
2010
ASD 33-115
months
Correlation -- All sensory
constructs (+)
-- --
Liss et al.,
2006
ASD 102 months
(mean)
Correlation HYPER (+; with social
& RRBs, not with
communication), HYPO
(+), SIRS (+)
-- -- --
Dellapiazza
et al., 2020
ASD 3-11 years Regression
(covarying CA &
IQ)
HYPER (+), HYPO (+)
with stereotypy
HYPER (+) -- HYPER (+), SIRS
(+) with
hyperactivity
123

124

Pfeiffer et
al., 2005
ASD 6-10 years Correlation -- -- HYPER (+) --
Hilton, 2007 ASD 6-10 years Correlation HYPER (+), HYPO (+),
SIRS (+)
-- -- --
Ashburner et
al., 2008
ASD 6-10 years Correlation -- -- HYPO (+) SIRS (+), HYPO
(+), HYPER/
tactile (+) with
inattention
Schulz &
Stevenson,
2019
ASD, TD 6-20 years
(M=11.3
years)
Correlation,
regression
HYPER (+) with RRB -- -- --
Mazurek et
al., 2013
ASD 2-17 years Correlation -- -- HYPER (+) --
Wigham et
al., 2015
ASD 8-16 years Correlation,
regression (path
analysis)
HYPO/HYPER (+) with
RRB
-- HYPER (+), HYPO
(+)
--
Feldman et
al., 2020
ASD, non-ASD 8-18 years Correlation,
regression
HYPO/HYPER (+) with
social communication,
HYPO/HYPER/SIRS (+)
for RRB
-- HYPER (+) --

c. Activity participation
Studies Samples Age Analysis Variety Frequency Involvement
Little et al.,
2015
ASD 5-12 years Regression
(covarying autism
severity, CA & MA)
-- HYPER (-; community &
routines); HYPO (+;
community); SIRS (+;
parent-child activity)
--
Pfeiffer et al.,
2005
ASD 6-10 years Correlation HYPER (-), HYPO (-) for community use (performance)
Ashburner et
al., 2008
ASD 6-10 years Correlation HYPO (-), SIRS (-) for school/academic performance
Engel-Yeger &
Ziv-On, 2011
ADHD 6-10 years Correlation HYPO (-) for preference to participate in social/recreational activities,
SIRS (+) for physical activities
Bar‐Shalita et
al., 2008
SPD 6-11 years Correlation Total score (-) for level of performance Total score (-)
124

125

Hochhauser &
Engel-Yeger,
2010
ASD 6-11 years Correlation SIRS (+), HYPER/motor
sensitivity (+) for home
activities
HYPER (+/-; depending
on sensory modality &
types of activity)
HYPER (-)
Reynolds et al.,
2011
ASD, TD 6-12 years Regression
(covarying IQ & sex)
HYPER (-) for activity competence
125

126

Table 4.2. Descriptive statistics of school-age outcome data for the subsample

Subsample (N=389)
Sensory Patterns (Trait Scores) Mean (SD)
HYPER Time 1 .15 (.68)
Time 2 .29 (.88)
Time 3 .30 (.93)
HYPO Time 1 .07 (.63)
Time 2 .23 (.76)
Time 3 .34 (.77)
SIRS .21 (.83)
Time 2 .19 (.98)
Time 3 .24 (.96)
School-age Outcomes Mean (SD)
SRS T-score
Social Communication and
Interaction (SCI)
52.80 (11.14)
Restricted, Repetitive Behaviors
and Interests (RRB)
53.21 (11.46)
VABS Standardized Score
Communication (COM) 99.66 (15.85)
Socialization (SOC) 97.57 (13.59)
Daily Living Skill (DLS) 96.59 (13.43)
Motor Skill (MOT) 99.89 (12.80)
Internalizing Behavior (INT) 16.22 (2.99)
Externalizing Behavior (EXT) 16.02 (3.13)
PEM-CY Average Score
Variety (VAR) Home 9.62 (.79)
School 8.07 (1.66)
Community 6.90 (1.63)
Frequency (FREQ) Home 8.88 (.59)
School 6.78 (1.13)
Community 6.08 (1.07)
Involvement (INV) Home 7.94 (1.00)
School 8.41 (1.49)
Community 8.04 (1.32)
Time 1

127

Table 4.3. Intercorrelations of school-age outcome variables (p all <.01)

Variable 1 2 3 4 5 6 7 8 9 10 11
SRS
1. SCI --
2. RRB .85 --
VABS
3. COM -.58 -.47 --
4. SOC -.68 -.57 .70 --
5. DLS -.45 -.40 .70 .72 --
6. MOT -.47 -.42 .59 .56 .61 --
7. INT .60 .54 -.29 -.38 -.26 -.32 --
8. EXT .57 .48 -.34 -.53 -.29 -.27 .59 --
PEM-CY
+

9. VAR -.43 -.36 .52 .51 .48 .37 -.21 -.22 --
10. FREQ -.29 -.24 .21 .27 .22 .30 -.20 -.15 .16 --
11. INV -.55 -.43 .38 .44 .33 .35 -.38 -.37 .26 .38 --
+ Average across contexts (home, school, and community)

128

Figure 4.1. Hypothesized model

Intercepts

HYPER
HYPO
SIRS
Latent Growth Parameters
Demographic
Covariates
Child’s Sex
Child’s Race
Parent Education
Autism Severity
Adaptive Behavior
Maladaptive Behavior
Participation
School-age
Outcomes
Slopes
HYPER
HYPO
SIRS
Full Sample (N=1,517)
Subsample (N=389)

129

Table 4.4. Total effects of individual growth (intercepts and slopes) of sensory patterns on school-age outcomes
*Bolded β coefficients indicate significant direct effects.
a. Autism symptoms and adaptive/maladaptive behavior

Predicting
Variables
Autism Symptoms Adaptive Behavior Maladaptive Behavior
SCI RRB COM SOC DLS MOT INTB EXTB
Standardized Coefficients (Standard Errors)
HYPER

Intercept -.11
†
(.19) -.01
†
(.19) .21
†
(.24) .39
†
(.28) .12
†
(.23) -.08
†
(.21) .10
†
(.18) -.07 (.19)
Slope .34
***
(.06) .34
***
(.08) -.20
**
(.08) -.28
**
(.09) -.27
**
(.08) -.31
**
(.08) .44
***
(.07) .16
*
(.07)
HYPO

Intercept .35
*
(.16) .13 (.16) -.44
*
(.21) -.61
*†
(.25) -.32 (.20) -.09 (.17) .18
†
(.16) .20 (.17)
Slope .29
***
(.07) .25
**
(.08) -.11 (.11) -.09 (.12) -.03 (.11) -.05 (.10) .06 (.08) .44
***
(.09)
SIRS

Intercept .19
†
(.11) .22
*
(.11) .05 (.12) -.15 (.13) .06 (.12) .13 (.12) .06 (.11) .02
†
(.10)
Slope .03 (.08) .14 (.11) -.02 (.09) .02 (.09) -.01 (.09) -.06 (.08) -.03 (.07) .04 (.07)

Child’s Sex -.11
*
(.04) -.13
**
(.04) .15
**
(.05) .23
**
(.05) .16
**
(.05) .14
**
(.05) -.06 (.05) -.16
**
(.05)
Child’s Race .10
*
(.04) .02

(.05) -.04

(.06) -.06

(.05) -.00 (.05) .03 (.06) .06 (.05) -.00 (.05)
Parent Edu -.25
***†
(.04) -.20
***†
(.05) .19
***
(.05) .09
†
(.05) .04 (.05) .17
***†
(.05) -.15
**†
(.05) -.06
†
(.05)
Pseudo R
2
48.8% 43.4% 23.6% 35.3% 16.3% 19.0% 30.2% 35.7%
Model Fit Statistics
x
2
(df) 76.12 (30) 78.07 (36) 72.41 (30)
CFI .99 .99 .99
TLI .96 .96 .96
RMSEA .03 .03 .03
129

130

b. Activity participation

Predicting
Variables
Variety Frequency Involvement
HOM SCH COMM HOM SCH COMM HOM SCH COMM
Standardized β (SE)
HYPER
Intercept -.04 (.19) -.39 (.23) .01 (.20) .02
†
(.19) -.12 (.19)

-.11 (.20) .11 (.20) .05
†
(.21) .10
†
(.13)
Slope -.11 (.08) -.05 (.09) -.03 (.09) -.22
**
(.08) -.12 (.08) -.07

(.08) -.16 (.09) -.31
**
(.10) -.28
***
(.06)
HYPO
Intercept -.16 (.17) .26 (.20) -.12 (.16) -.17 (.18) -.07 (.17) -.08 (.17) -.25 (.18) -.26 (.18) -.28
*
(.12)
Slope -.03 (.09) -.14 (.12) .02 (.10) -.06 (.10) -.29
**
(.09) -.13 (.09) -.02 (.10) .14 (.11) -.09 (.06)
SIRS
Intercept .15 (.12) .10 (.13) .12 (.12) -.10 (.11) -.01 (.11) .05 (.12) -.01 (.11) .03 (.12) -.08 (.07)
Slope -.04 (.09) -.11 (.09) -.10 (.09) .05 (.08) .09 (.08) -.06 (.09) .07 (.09) .03 (.09) -.10
*
(.05)

Child’s Sex .05 (.05) -.04 (.05) .00 (.05) .19
***
(.05) .17
***
(.05) .15
**
(.05) .11
*
(.05) .18
***
(.05) .15
***
(.03)
Child’s Race .11 (.05) -.10 (.06) -.06 (.05) .02 (.05) -.10 (.06) -.07

(.06) -.07 (.06) -.04 (.05) -.05 (.03)
Parent Edu -.07 (.05) .11
*
(.05) .19
***
(.05) .12
*
(.05) .13
**†
(.05) .14
**
(.05) .19
***
(.05) .21
***†
(.05) .14
***†
(.03)
Pseudo R
2
7.5% 13.2% 7.2% 12.7% 20.1% 10.7% 11.0% 18.7% 23.4%
Model Fit Statistics
x
2
(df) 71.61 (33) 74.11 (33) 107.69 (33)
CFI .99 .99 .97
TLI .96 .96 .93
RMSEA .03 .03 .04
Note: Child's sex is coded 0=male, 1=female; Child's race is coded 0=white, 1=non-white; Parent education is coded 0=None of the parents had a college degree (or beyond),
1=One of the parents had a college degree (or beyond), 2=Both parents had a college degree (or beyond). HOM=home; SCH=school; COMM=community.
†
presence of significant indirect effect (paths shown in Table 4.5)
*p<.05. **p<.01. ***p<.001 (two-tailed)
130

131

Table 4.5. Significant Indirect Effects
Paths Standardized β (SE)
INT HYPER → SLP HYPER → SCI -.16
**
(.06)
INT SIRS → SLP HYPO → SCI .09
*
(.04)
EDU → SLP HYPO → SCI -.04
*
(.02)
EDU → SLP HYPER → SCI -.04
*
(.01)
INT HYPER → SLP HYPER → RRB -.17
**
(.06)
EDU → SLP HYPER → RRB -.04
*
(.02)
INT HYPER → SLP HYPER → COM .10
*
(.04)
INT HYPER → SLP HYPER → SOC .14
**
(.05)
INT HYPO → SLP HYPER → SOC -.07
*
(.04)
EDU → SLP HYPER → SOC .03
*
(.01)
INT HYPER → SLP HYPER → DLS .13
*
(.05)
INT HYPER → SLP HYPER → MOT .15
*
(.06)
EDU → SLP HYPER → MOT .03
*
(.01)
INT HYPER → SLP HYPER → INTB -.21
**
(.07)
INT HYPO → SLP HYPER → INTB .11
*
(.06)
EDU → SLP HYPER → INTB -.05
**
(.02)
INT SIRS → SLP HYPO → EXTB .13
*
(.06)
EDU → SLP HYPO → EXTB -.06
*
(.02)
INT HYPER → SLP HYPER → FREQ HOM .11
*
(.05)
EDU → SLP HYPO → FREQ SCH .04
*
(.02)
INT HYPER → SLP HYPER → INV SCH .14
*
(.06)
EDU → SLP HYPER → INV SCH .03
*
(.02)
INT HYPER → SLP HYPER → INV COMM .13
**
(.04)
EDU → SLP HYPER → INV COMM .03
*
(.01)

*p<.05. **p<.01. ***p<.001 (two-tailed)

132

Figure 4.2. Significant prediction effects of latent growth factors on school-age outcomes
a. Effects on autism symptoms

b. Effects on maladaptive behavior

c. Effects on adaptive behavior and motor skills

133

d. Effects on frequency of participation

e. Effects on involvement during participation

Note: Dashed lines indicate significant direct effects on the mediators (i.e., significant "a" paths) but non-significant
total effects (i.e., non-significant “c” paths) on the outcome variables of interest.
*p<.05. **p<.01. ***p<.001 (two-tailed)

134

CHAPTER 5
General Discussion
In a series of studies, we demonstrated that developmental trajectories of sensory patterns
from infancy to school age were highly variable among a large community sample by adopting
person-centered approaches (i.e., latent growth curve modeling and latent class growth analysis),
reflecting the concept of “chronogeneity” (Georgiades, Bishop, & Frazier, 2017). Also, these
early sensory trajectories were differentially associated with later clinical and broader
developmental outcomes. Our findings highlight the complexity of the early development of
sensory patterns as well as the importance of considering individual differences. Below we
unfold the discussion by the key findings across studies, followed by limitations and future
directions.
Sensory Patterns are Developmentally Variable Among a General Population
Past research has documented changes in the behavioral manifestation of sensory patterns
across early childhood; however, the mixed findings derived from previous studies with many
differences in methodological approaches and a lack of longitudinal evidence highlight the need
for more systematic investigations. To resolve such empirical inconsistencies, Study 1 served as
the very first attempt to model the multidimensional longitudinal course of sensory patterns in an
age-homogeneous community sample. While we found that HYPER and SIRS decreased and
HYPO increased over time as mean trends for children from infancy through six years of age,
individual trajectories were highly variable, partially depending on their demographics and
clinical outcome status.

135

Study 1 revealed that while child’s sex and race only affected the initial levels of certain
sensory patterns, parent education particularly impacted their child’s rate of changes in HYPER
and HYPO. In Study 2, we also found that the two most severe/worsening subtypes (Elevated -
All Worsening and Moderate - HYPO Worsening) were associated with lower parent education
levels. These may indicate the critical moderating role of parents in their child’s sensory
development, particularly in sensory responsiveness, which is often integral to parent-child
interactions (e.g., calling child’s name, eye/tactile contacts during daily routines). Thus, tailored
family approaches may be useful to consider as an important aspect of early interventions for
children with sensory challenges. Aside from demographic factors, children in different clinical
outcome groups also showed differences in their sensory trajectories. In general, children with
ASD and/or SPD were characterized by elevated and worsening patterns, while those in the OD
and ND groups tended to show low intensity and stable/improving patterns over time. However,
while most children with ASD showed elevated/worsening patterns, some of them were
characterized by trajectories more like those in the OD/ND groups. This within-group variability
was also observed in other clinical outcome groups.
To explain such variability that could not be fully explained by demographic and clinical
outcome status, Study 2 applied mixture modeling to identify the unknown homogeneous groups
among this community sample. As a result, we identified five fine-grained sensory trajectory
subtypes that differed by intensity and change rates across sensory patterns. These subtypes were
also characterized by various demographics and later school-age outcomes, thus supporting their
clinical relevance. Specifically, the two most severe/worsening subtypes were observed to have
more challenging behaviors and more difficulties in social communication. Therefore, profiling
children based on their early sensory trajectories may help to identify those who are more likely

136

to experience developmental challenges at school age and may further introduce opportunities
for early intervention.
Early Sensory Trajectories are Potential Markers of Later ASD
Expanding upon the previous finding that infant siblings with ASD showed worsening
patterns within the first two years of life (Wolff et al., 2019), Study 1 demonstrated that a
community sample of children with an ASD outcome generally had more elevated sensory
patterns from infancy, followed by larger increases in intensity (worsening patterns) over time.
Although the SPD group also showed elevated sensory patterns across this period as compared to
the other diagnosis/concerns and no diagnosis/concerns groups, their increases in HYPER and
HYPO were less dramatic. Thus, the slopes of sensory responsiveness might be useful ASD
markers with good specificity. Study 3 also revealed the specifically large impact of sensory
trajectories on autism symptoms at school age, including both social-communication deficits and
RRBs. These findings all support the strong relationship between sensory patterns (especially
atypical sensory responsiveness) and autism symptoms as well as indicate the importance of
continuous surveillance during early childhood for early identification of ASD.
As further support of the utility of sensory trajectories in early detection of ASD, the
largest differences across subtypes were found in autism severity at school age in Study 2. Three
percent of our sample was classified to the Elevated-All Worsening subtype, which was
associated with much higher autism severity and probability for an ASD outcome as compared to
other subtypes. The longitudinal sensory patterns of this subtype are generally consistent with
what we observed among children with ASD in Study 1 by showing elevated intensity and
significant increases over time across sensory patterns. This subtype further represented the cases
with the most elevated sensory challenges and poor school-age outcomes in the ASD and SPD

137

groups (32% of the ASD group and 8% of the SPD group were classified to this subtype). The
other subtypes, except for the Adaptive-All Improving subtype, are all characterized by medium
to elevated levels of intensity and varying rates of change in certain sensory patterns, as well as
were associated with weaknesses in certain school-age outcomes. Approximately 65% of the
ASD group and 89% of the SPD group were distributed across these three “medium-severity”
subtypes. These differences across subtypes associated with clinical outcome status indicate that
while sensory patterns are not unique to ASD, they are more prevalent and tend to be more
severe among ASD populations. Thus, more attention should be paid to the early identification
and interventions of sensory challenges for reducing their potential negative long-term impact in
children with ASD.
Sensory Patterns Co-occur and Co-develop From Infancy and Together Pose Differential
Downstream Effects on Later School-age Outcomes
In Study 1, the significant associations among latent growth factors across sensory
patterns not only indicate their co-occurrence, as demonstrated by previous subtyping studies
(e.g., Ausderau et al., 2014a), but also their co-development over time. Such associations were
specifically strong between HYPER and HYPO, which might be due to a dynamic process of up-
and down-regulation of sensory input during early development (Baranek, Reinhartsen, &
Wannamaker, 2001). We also reported the close relationship between early SIRS and later
changes in HYPO by demonstrating their significant intercept-slope association. This
relationship was further supported by the factor-regression finding in Study 3 that early SIRS
predicted later changes in HYPO. These results highlight the necessity of considering the
interactions of sensory domains across development and operationalizing sensory patterns as
multidimensional.

138

Thus, in Study 3 we further examined the effects of the parallel-process sensory
trajectories on school-age outcomes while accounting for their mutual influence. As a result, the
growth factors of HYPER tended to be the most robust predictors across school-age outcomes,
showing both direct and indirect effects. We also observed differential impacts of HYPER and
HYPO on maladaptive behavior, indicating that tailored sensory intervention is necessary for
improving children’s long-term outcomes. Across all school-age outcomes, more variations in
autism severity, maladaptive behavior, and social functioning were explained by sensory
trajectories (especially HYPER and HYPO). Existing longitudinal evidence has shown cascading
effects of early sensory patterns on later social communication (Baranek et al., 2018) and
internalizing behavior (Green et al., 2012). The current findings added to the existing evidence
that early sensory patterns and their changes over time predicted later broader functional
outcomes to various extents, as well as indicate that sensory responsiveness might be important
targets for early intervention towards more optimal long-term outcomes.
Although the impacts of sensory trajectories on children’s activity participation,
including variety, frequency, and involvement across home, school, and community contexts,
were found less significant as compared to adaptive/maladaptive behaviors and autism symptoms
in Study 2 and 3, we did observe that children’s frequency and involvement of participation were
more impacted by sensory trajectories than the variety of activities they participated in.
Children’s better engagement in activities is associated with more learning opportunities (Dunst
et al., 2002), and therefore helping children to cope with sensory challenges might facilitate their
participation in activities that introduces more learning opportunities. The finding that parent
education was a significant predictor of children’s overall participation also highlights the
contribution of family factors to what children actually do in their daily lives. Given that

139

participation is a more complex concept that involves transactions between person, environment,
and occupation (Law, 1996), other factors such as environmental support/barrier, access to
resources, cultural differences might account for more variations in such complex outcomes.
Nevertheless, the inclusion of specific dimensions of participation and the consideration of
contexts in the current research serves as a preliminary step toward a better understanding of
children’s comprehensive and meaningful outcomes at school age as associated with their early
sensory development.
Implications for Occupational Science and Occupational Therapy
Although child development has been a popular topic over decades across many
disciplines, the study conceptualization and methodology adopted, such as cross-sectional
design/analysis and diagnostic group comparisons without considering within-group differences
often fail to address the complex and dynamic qualities of the process. Our findings on the
developmental and heterogeneous nature of sensory patterns during early childhood revealed that
every child is a unique occupational being as characterized by diverse developmental pathways,
which are shaped by both child and family factors. Particularly, significant associations between
parent education and children’s development of sensory responsiveness were observed across the
current studies, indicating potential parent-child transaction that impacts children’s response to
and coping with the sensory environment (Dickie et al., 2009). Additionally, the observed
differential impacts of early sensory development on children’s later outcomes, including
participation in different contexts indicated a potential developmental cascade mechanism that
merits more future investigations. Through the comprehensive evaluation of children’s school-
age outcomes, we also demonstrated that what a child can do does not necessarily reflect what he
or she does in everyday life. The latter might be more impacted by environmental or socio-

140

contextual factors that are often addressed in the occupational science literature (Davis,
Polatajko, & Ruud, 2002; Humphry & Wakeford, 2006). This indicates that to better meet the
needs of the children and families in real lives, more emphasis should be placed on enhancing
children’s participation in meaningful activities through identifying the contributing factors and
barriers to be overcome via adaptations (Reynolds et al., 2017). Overall, the current studies
endeavored to understand children’s early sensory development from a perspective enriched by
the foundations and legacy of occupational science, as well as to provide interdisciplinary links
with other fields, such as psychology and developmental science.
The current studies also have important implications for occupational therapy. A better
understanding of the developmental and heterogeneous nature of children’s sensory behaviors
and their long-term impacts could help occupational therapy practitioners to effectively facilitate
behavior change and work with the family or stakeholders to implement adaptations based on
children’s specific needs. Our findings on the sensory development profiles and differential
impacts of sensory patterns on school-age outcomes highlighted the importance of identifying
children’s strengths and weaknesses in order to provide customized and targeted interventions
for achieving their optimal outcomes. It is also critical to consider child and family
characteristics given the demographic differences across sensory profiles. Particularly, the
observed associations between parent education and child’s sensory responsiveness may indicate
that parent-child transactions have impacts on children’s development over time. For instance, a
parent’s ability to respond to a child’s cues and create a sensory environment for optimal
engagement will help the child to cope with challenging sensory situations. Furthermore, the co-
occurrence of sensory patterns and potential interactions of one another, as observed across the
current studies, indicated the necessity of comprehensively assessing and treating atypical

141

sensory behaviors with an awareness of their mutual impact on children’s functional
performance. Ultimately, such fine-grained characterizations of sensory patterns across
development and populations of young children with and without developmental challenges may
help practitioners to serve children and families with a better grasp of “what works for whom,
and why.”
Limitations and Future Directions
The current research is not without limitations. Measurement biases due to the changing
measures of sensory patterns (the FYI at Time 1 and the SEQ at Time 2 and 3) might be
inevitable despite our efforts to construct comparable scores through invariance testing and scale
equating procedures. Exactly repeated measures with more representative items are always
encouraged in longitudinal studies. Also, three time-points of data only allow for modeling linear
changes despite the possibility of non-linear changes. Thus, we expect future studies to expand
the investigations of sensory trajectories by including more time-points, potentially further into
the school years. Furthermore, although the parent-report measures adopted in the current
research provided useful information regarding children’s developmental outcomes, we were not
able to ascertain ASD diagnosis with gold-standard measures given the large sample size in the
current studies. Future research might consider confirming the diagnosis at least for those who
meet the SRS cutoffs for more accurate diagnostic grouping among a population-based sample.
Aside from overcoming these limitations related to study design, we suggest future
studies to include intervention types or dosages, and more family/environment factors to
examine their potential moderating or mediating effects on children’s developmental trajectories
towards various long-term outcomes. As discussed above, these factors might account for more
variations in children’s participation in daily activities. Given the significant effects of parent

142

education revealed in the current research, parent-implemented or family-centered intervention
approaches might be of particular interest for the examination of their influence on children’s
sensory development. A further limitation in the current analysis is that parent education data
were collected only at Time 1, yet it is possible that education levels may have changed over a
five-year timespan for some parents. Thus, more details about parent characteristics such as
education levels or SES, might be necessary to measure at multiple timepoints for a more
accurate examination of the changing impacts of family factors on child outcomes. In addition, it
is possible that parents who enrolled in the study were motivated to participate because of
specific concerns about their children and thus, may have reported more sensory differences.
Therefore, it is important for future studies to collect information from other informants with
observational or more objective measures to avoid potential informant biases.
In addition, we modeled the parallel-process trajectories of specific sensory patterns
instead of the overall sensory symptoms, to understand the multidimensionality of sensory
development. Future studies might further examine the parallel-process trajectories of multiple
behavioral domains (e.g., social communication and sensory patterns) in order to examine their
emergence or timeline relative to each other and mutual impact across development (e.g.,
developmental cascades). It is also important to apply multimodal approaches (e.g., parent-
report, observational, and biobehavioral measures) to measure sensory patterns and/or other
behavioral domains of interest to account for potential informant biases and capture a complete
picture of early development.
Overall, the current studies responded to the call for future sensory-related research from
Uljarević and colleagues (2017) by 1) considering sensory patterns as dimensional constructs, 2)
applying analytic approaches that focus on individual differences under a longitudinal

143

prospective design, and 3) examining the potential value of early individual variability of sensory
patterns as indicators of later outcomes, including both diagnostic status and broader
developmental/functional outcomes. Our systematic investigations represent critical steps
towards a better understanding of the longitudinal variability of sensory patterns during early
childhood and its potential contribution to early detection of ASD and other developmental
challenges. We hope that such an improved understanding of heterogeneity across early
development would facilitate the delivery of interventions and services to children with needs
and their families in a timely and targeted manner.

144

REFERENCES
American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders
(DSM-5
®
). American Psychiatric Pub.
Ashburner, J., Ziviani, J., & Rodger, S. (2008). Sensory processing and classroom emotional,
behavioral, and educational outcomes in children with autism spectrum
disorder. American Journal of Occupational Therapy, 62(5), 564-573.
Ausderau, K., Furlong, M., Sideris, J., Bulluck, J., Little, L. M., Watson, L. R., ... & Baranek, G.
T. (2014a). Sensory subtypes in children with autism spectrum disorder: latent profile
transition analysis using a national survey of sensory features. Journal of Child
Psychology and Psychiatry, 55(8), 935-944.
Ausderau, K., Sideris, J., Furlong, M., Little, L. M., Bulluck, J., & Baranek, G. T. (2014b).
National survey of sensory features in children with ASD: Factor structure of the sensory
experience questionnaire (3.0). Journal of Autism and Developmental Disorders, 44(4),
915-925.
Ausderau, K. K., Sideris, J., Little, L. M., Furlong, M., Bulluck, J. C., & Baranek, G. T. (2016).
Sensory subtypes and associated outcomes in children with autism spectrum disorders.
Autism Research, 9(12), 1316-1327.
Baker, A. E., Lane, A., Angley, M. T., & Young, R. L. (2008). The relationship between sensory
processing patterns and behavioural responsiveness in autistic disorder: A pilot
study. Journal of Autism and Developmental Disorders, 38(5), 867-875.

145

Bakker, K., & Moulding, R. (2012). Sensory-processing sensitivity, dispositional mindfulness
and negative psychological symptoms. Personality and Individual Differences, 53(3), 341-
346.
Bal, V. H., Kim, S. H., Cheong, D., & Lord, C. (2015). Daily living skills in individuals with
autism spectrum disorder from 2 to 21 years of age. Autism, 19(7), 774-784.
Baranek, G. T. (1999a). Sensory Experiences Questionnaire (SEQ). Unpublished manuscript,
University of North Carolina at Chapel Hill.
Baranek, G. T. (1999b). Sensory Processing Assessment for Young Children (SPA). Unpublished
manuscript, University of North Carolina at Chapel Hill.
Baranek, G. T., Carlson, M., Sideris, J., Kirby, A. V., Watson, L. R., Williams, K. L., & Bulluck,
J. (2019). Longitudinal assessment of stability of sensory features in children with autism
spectrum disorder or other developmental disabilities. Autism Research, 12(1), 100-111.
Baranek, G. T., Boyd, B. A., Poe, M. D., David, F. J., & Watson, L. R. (2007). Hyperresponsive
sensory patterns in young children with autism, developmental delay, and typical
development. American Journal on Mental Retardation, 112(4), 233-245.
Baranek, G. T., David, F. J., Poe, M. D., Stone, W. L., & Watson, L. R. (2006). Sensory
Experiences Questionnaire: discriminating sensory features in young children with autism,
developmental delays, and typical development. Journal of Child Psychology and
Psychiatry, 47(6), 591-601.
Baranek, G. T., Foster, L. G., & Berkson, G. (1997). Tactile defensiveness and stereotyped
behaviors. The American Journal of Occupational Therapy, 51(2), 91-95.

146

Baranek, G. T., Little, L. M., Diane Parham, L., Ausderau, K. K., & Sabatos‐DeVito, M. G.
(2014). Sensory features in autism spectrum disorders. Handbook of Autism and Pervasive
Developmental Disorders, Fourth Edition.
Baranek, G. T., Reinhartsen, D., & Wannamaker, S. (2001). Play: Engaging children with
autism. Sensorimotor Interventions in Autism, 311-351.
Baranek, G. T., Watson, L. R., Boyd, B. A., Poe, M. D., David, F. J., & McGuire, L. (2013).
Hyporesponsiveness to social and nonsocial sensory stimuli in children with autism,
children with developmental delays, and typically developing children. Development and
Psychopathology, 25(2), 307-320.
Baranek, G. T., Woynaroski, T. G., Nowell, S., Turner-Brown, L., DuBay, M., Crais, E. R., &
Watson, L. R. (2018). Cascading effects of attention disengagement and sensory seeking
on social symptoms in a community sample of infants at-risk for a future diagnosis of
autism spectrum disorder. Developmental Cognitive Neuroscience, 29, 30-40.
Bar‐Shalita, T., Vatine, J. J., & Parush, S. (2008). Sensory modulation disorder: a risk factor for
participation in daily life activities. Developmental Medicine & Child Neurology, 50(12),
932-937.
Baum, S. H., Stevenson, R. A., & Wallace, M. T. (2015). Behavioral, perceptual, and neural
alterations in sensory and multisensory function in autism spectrum disorder. Progress in
Neurobiology, 134, 140-160.
Ben-Sasson, A., Carter, A. S., & Briggs-Gowan, M. J. (2009a). Sensory over-responsivity in
elementary school: prevalence and social-emotional correlates. Journal of Abnormal Child
Psychology, 37(5), 705-716.

147

Ben‐Sasson, A., Cermak, S. A., Orsmond, G. I., Tager‐Flusberg, H., Kadlec, M. B., & Carter, A.
S. (2008). Sensory clusters of toddlers with autism spectrum disorders: Differences in
affective symptoms. Journal of Child Psychology and Psychiatry, 49(8), 817-825.
Ben-Sasson, A., Hen, L., Fluss, R., Cermak, S. A., Engel-Yeger, B., & Gal, E. (2009b). A meta-
analysis of sensory modulation symptoms in individuals with autism spectrum disorders.
Journal of Autism and Developmental Disorders, 39(1), 1-11.
Boyd, B. A., Baranek, G. T., Sideris, J., Poe, M. D., Watson, L. R., Patten, E., & Miller, H.
(2010). Sensory features and repetitive behaviors in children with autism and
developmental delays. Autism Research, 3(2), 78-87.
Bryson, S. E., Zwaigenbaum, L., Brian, J., Roberts, W., Szatmari, P., Rombough, V., &
McDermott, C. (2007). A prospective case series of high-risk infants who developed
autism. Journal of Autism and Developmental Disorders, 37(1), 12-24.
Cascio, C. J., Woynaroski, T., Baranek, G. T., & Wallace, M. T. (2016). Toward an
interdisciplinary approach to understanding sensory function in autism spectrum
disorder. Autism Research, 9(9), 920-925.
Cermak, S. A., Curtin, C., & Bandini, L. G. (2010). Food selectivity and sensory sensitivity in
children with autism spectrum disorders. Journal of the American Dietetic
Association, 110(2), 238-246.
Chen, Y. H., Rodgers, J., & McConachie, H. (2009). Restricted and repetitive behaviours,
sensory processing and cognitive style in children with autism spectrum disorders. Journal
of Autism and Developmental Disorders, 39(4), 635-642.

148

Chuang, I. C., Tseng, M. H., Lu, L., & Shieh, J. Y. (2012). Sensory correlates of difficult
temperament characteristics in preschool children with autism. Research in Autism
Spectrum Disorders, 6(3), 988-995.
Cosbey, J., Johnston, S. S., & Dunn, M. L. (2010). Sensory processing disorders and social
participation. American Journal of Occupational Therapy, 64(3), 462-473.
Coster, W. (1998). Occupation-centered assessment of children. The American Journal of
Occupational Therapy, 52(5), 337-344.
Damiano-Goodwin, C. R., Woynaroski, T. G., Simon, D. M., Ibañez, L. V., Murias, M., Kirby,
A., ... & Cascio, C. J. (2018). Developmental sequelae and neurophysiologic substrates of
sensory seeking in infant siblings of children with autism spectrum
disorder. Developmental Cognitive Neuroscience, 29, 41-53.
Davis, J. A., Polatajko, H. J., & Ruud, C. A. (2002). Children's occupations in context: The
influence of history. Journal of Occupational Science, 9(2), 54-64.
Dawson, G., Osterling, J., Meltzoff, A. N., & Kuhl, P. (2000). Case study of the development of
an infant with autism from birth to two years of age. Journal of Applied Developmental
Psychology, 21(3), 299-313.
Deconinck, N., Soncarrieu, M., & Dan, B. (2013). Toward better recognition of early predictors
for autism spectrum disorders. Pediatric Neurology, 49(4), 225-231.
Dellapiazza, F., Michelon, C., Vernhet, C., Muratori, F., Blanc, N., Picot, M. C., & Baghdadli,
A. (2021). Sensory processing related to attention in children with ASD, ADHD, or
typical development: results from the ELENA cohort. European Child and Adolescent
Psychiatry, 30(2), 283-291.

149

Dickie, V. A., Baranek, G. T., Schultz, B., Watson, L. R., & McComish, C. S. (2009). Parent
reports of sensory experiences of preschool children with and without autism: A
qualitative study. American Journal of Occupational Therapy, 63(2), 172-181.
Dunn, W. (1997). The impact of sensory processing abilities on the daily lives of young children
and their families: A conceptual model. Infants and Young Children, 9, 23-35.
Dunn, W. (1999a). The Short Sensory Profile. San Antonio, TX: The Psychological Corporation.
Dunn, W. (1999b). The Sensory Profile: Examiner’s Manual. San Antonio, TX: The
Psychological Corporation.
Dunn, W. (2002). Infant/toddler Sensory Profile: User's Manual. Pearson.
Dunst, C. J., Hamby, D., Trivette, C. M., Raab, M., & Bruder, M. B. (2002). Young children's
participation in everyday family and community activity. Psychological Reports, 91(3),
875-897.
Engel-Yeger, B., & Ziv-On, D. (2011). The relationship between sensory processing difficulties
and leisure activity preference of children with different types of ADHD. Research in
Developmental Disabilities, 32(3), 1154-1162.
Foss-Feig, J. H., Heacock, J. L., & Cascio, C. J. (2012). Tactile responsiveness patterns and their
association with core features in autism spectrum disorders. Research in Autism Spectrum
Disorders, 6(1), 337-344.
Freuler, A., Baranek, G. T., Watson, L. R., Boyd, B. A., & Bulluck, J. C. (2012). Precursors and
trajectories of sensory features: qualitative analysis of infant home videos. American
Journal of Occupational Therapy, 66(5), e81-e84.

150

Georgiades, S., Bishop, S. L., & Frazier, T. (2017). Editorial Perspective: Longitudinal research
in autism–introducing the concept of ‘chronogeneity’. Journal of Child Psychology and
Psychiatry, 58(5), 634-636.
Georgiades, S., & Kasari, C. (2018). Reframing optimal outcomes in autism. JAMA
Pediatrics, 172(8), 716-717.
Gepner, B., & Mestre, D. R. (2002). Brief report: postural reactivity to fast visual motion
differentiates autistic from children with Asperger syndrome. Journal of Autism and
Developmental disorders, 32(3), 231-238.
Germani, T., Zwaigenbaum, L., Bryson, S., Brian, J., Smith, I., Roberts, W., ... & Vaillancourt,
T. (2014). Brief report: assessment of early sensory processing in infants at high-risk of
autism spectrum disorder. Journal of Autism and Developmental Disorders, 44(12), 3264-
3270.
Green, S. A., Ben-Sasson, A., Soto, T. W., & Carter, A. S. (2012). Anxiety and sensory over-
responsivity in toddlers with autism spectrum disorders: bidirectional effects across
time. Journal of Autism and Developmental Disorders, 42(6), 1112-1119.
Grzadzinski, R., Donovan, K., Truong, K., Nowell, S., Lee, H., Sideris, J., ... & Watson, L. R.
(2020). Sensory reactivity at 1 and 2 years old is associated with ASD severity during the
preschool years. Journal of Autism and Developmental Disorders, 50(11), 3895-3904.
Hilton, C., Graver, K., & LaVesser, P. (2007). Relationship between social competence and
sensory processing in children with high functioning autism spectrum disorders. Research
in Autism Spectrum Disorders, 1, 164–173.

151

Hilton, C. L., Harper, J. D., Kueker, R. H., Lang, A. R., Abbacchi, A. M., Todorov, A., &
LaVesser, P. D. (2010). Sensory responsiveness as a predictor of social severity in
children with high functioning autism spectrum disorders. Journal of Autism and
Developmental Disorders, 40(8), 937-945.
Hochhauser, M., & Engel-Yeger, B. (2010). Sensory processing abilities and their relation to
participation in leisure activities among children with high-functioning autism spectrum
disorder (HFASD). Research in Autism Spectrum Disorders, 4(4), 746-754.
Humphry, R., & Wakeford, L. (2006). An occupation-centered discussion of development and
implications for practice. American Journal of Occupational Therapy, 60(3), 258-267.
Iarocci, G., & McDonald, J. (2006). Sensory integration and the perceptual experience of persons
with autism. Journal of Autism and Developmental Disorders, 36(1), 77-90.
Kanner, L. (1943). Autistic disturbances of affective contact. Nervous Child, 2(3), 217-250.
Kern, J. K., Garver, C. R., Carmody, T., Andrews, A. A., Trivedi, M. H., & Mehta, J. A. (2007a).
Examining sensory quadrants in autism. Research in Autism Spectrum Disorders, 1(2),
185-193.
Kern, J. K., Garver, C. R., Grannemann, B. D., Trivedi, M. H., Carmody, T., Andrews, A. A., &
Mehta, J. A. (2007b). Response to vestibular sensory events in autism. Research in Autism
Spectrum Disorders, 1(1), 67-74.
Kern, J. K., Trivedi, M. H., Grannemann, B. D., Garver, C. R., Johnson, D. G., Andrews, A.
A., ... & Schroeder, J. L. (2007c). Sensory correlations in autism. Autism, 11(2), 123-134.

152

Killeen, H., Shiel, A., Law, M., O'Donovan, D. J., Segurado, R., & Anaby, D. (2018).
Relationships between adaptive behaviours, personal factors, and participation of young
children. Physical & Occupational Therapy in Pediatrics, 38(4), 343-354.
King, G., Imms, C., Stewart, D., Freeman, M., & Nguyen, T. (2018). A transactional framework
for pediatric rehabilitation: shifting the focus to situated contexts, transactional processes,
and adaptive developmental outcomes. Disability and Rehabilitation, 40(15), 1829-1841.
Kinnealey, M., Koenig, K. P., & Smith, S. (2011). Relationships between sensory modulation
and social supports and health-related quality of life. American Journal of Occupational
Therapy, 65(3), 320-327.
Klintwall, L., Holm, A., Eriksson, M., Carlsson, L. H., Olsson, M. B., Hedvall, Å., ... & Fernell,
E. (2011). Sensory abnormalities in autism: a brief report. Research in Developmental
Disabilities, 32(2), 795-800.
Kramer, J. M., Coster, W. J., Kao, Y. C., Snow, A., & Orsmond, G. I. (2012). A new approach to
the measurement of adaptive behavior: development of the PEDI-CAT for children and
youth with autism spectrum disorders. Physical & Occupational Therapy in Pediatrics,
32(1), 34-47.
Lane, A. E., Molloy, C. A., & Bishop, S. L. (2014). Classification of children with autism
spectrum disorder by sensory subtype: a case for sensory‐based phenotypes. Autism
Research, 7(3), 322-333.
Lane, S. J., Reynolds, S., & Dumenci, L. (2012). Sensory overresponsivity and anxiety in
typically developing children and children with autism and attention deficit hyperactivity

153

disorder: cause or coexistence? American Journal of Occupational Therapy, 66(5), 595-
603.
Lane, A. E., Young, R. L., Baker, A. E., & Angley, M. T. (2010). Sensory processing subtypes in
autism: Association with adaptive behavior. Journal of Autism and Developmental
Disorders, 40(1), 112-122.
Law, M., Cooper, B., Strong, S., Stewart, D., Rigby, P., & Letts, L. (1996). The person-
environment-occupation model: A transactive approach to occupational
performance. Canadian Journal of Occupational Therapy, 63(1), 9-23.
Ledford, J. R., & Gast, D. L. (2006). Feeding problems in children with autism spectrum
disorders: A review. Focus on Autism and Other Developmental Disabilities, 21(3), 153-
166.
Leekam, S. R., Nieto, C., Libby, S. J., Wing, L., & Gould, J. (2007). Describing the sensory
abnormalities of children and adults with autism. Journal of Autism and Developmental
Disorders, 37(5), 894-910.
Liss, M., Saulnier, C., Fein, D., & Kinsbourne, M. (2006). Sensory and attention abnormalities in
autistic spectrum disorders. Autism, 10, 155–172.
Little, L. M., Ausderau, K., Sideris, J., & Baranek, G. T. (2015). Activity participation and
sensory features among children with autism spectrum disorders. Journal of Autism and
Developmental Disorders, 45(9), 2981-2990.
Little, L. M., Dean, E., Tomchek, S. D., & Dunn, W. (2017). Classifying sensory profiles of
children in the general population. Child: Care, Health and Development, 43(1), 81-88.

154

Luckasson, R., Borthwick-Duffy, S., Buntinx, W. H., Coulter, D. L., Craig, E. M. P., Reeve,
A., ... & Tasse, M. J. (2002). Mental Retardation: Definition, Classification, and Systems
of Supports. American Association on Mental Retardation.
Marco, E. J., Hinkley, L. B., Hill, S. S., & Nagarajan, S. S. (2011). Sensory processing in autism:
a review of neurophysiologic findings. Pediatric Research, 69(5 Pt 2), 48R-54R.
Mazurek, M. O., Vasa, R. A., Kalb, L. G., Kanne, S. M., Rosenberg, D., Keefer, A., ... &
Lowery, L. A. (2013). Anxiety, sensory over-responsivity, and gastrointestinal problems
in children with autism spectrum disorders. Journal of Abnormal Child Psychology, 41(1),
165-176.
McCormick, C., Hepburn, S., Young, G. S., & Rogers, S. J. (2016). Sensory symptoms in
children with autism spectrum disorder, other developmental disorders and typical
development: a longitudinal study. Autism, 20(5), 572-579.
Miller, L. J., Anzalone, M. E., Lane, S. J., Cermak, S. A., & Osten, E. T. (2007). Concept
evolution in sensory integration: A proposed nosology for diagnosis. The American
Journal of Occupational Therapy, 61(2), 135.
Miller, L. J., Schoen, S. A., Mulligan, S., & Sullivan, J. (2017). Identification of sensory
processing and integration symptom clusters: a preliminary study. Occupational Therapy
International, 2017.
Nelson, D. L. (1996). Therapeutic occupation: A definition. American Journal of Occupational
Therapy, 50(10), 775-782.

155

Nieto, C., Lopez, B., & Gandia, H. (2017). Relationships between atypical sensory processing
patterns, maladaptive behaviour and maternal stress in Spanish children with autism
spectrum disorder. Journal of Intellectual Disability Research, 61(12), 1140-1150.
O’Donnell, S., Deitz, J., Kartin, D., Nalty, T., & Dawson, G. (2012). Sensory processing,
problem behavior, adaptive behavior, and cognition in preschool children with autism
spectrum disorders. American Journal of Occupational Therapy, 66(5), 586-594.
Parham, L. D., Ecker, C., & Miller Kuhaneck, H. (2010). SPM Sensory Processing Measure
Manual. Los Angeles, CA: Western Psychological Services.
Perez-Repetto, L., Jasmin, E., Fombonne, E., Gisel, E., & Couture, M. (2017). Longitudinal
study of sensory features in children with autism spectrum disorder. Autism Research and
Treatment, 2017.
Pfeiffer, B., & Kinnealey, M. (2003). Treatment of sensory defensiveness in
adults. Occupational Therapy International, 10(3), 175-184.
Reynolds, S., Bendixen, R. M., Lawrence, T., & Lane, S. J. (2011). A pilot study examining
activity participation, sensory responsiveness, and competence in children with high
functioning autism spectrum disorder. Journal of Autism and Developmental
Disorders, 41(11), 1496-1506.
Reynolds, S., Glennon, T. J., Ausderau, K., Bendixen, R. M., Kuhaneck, H. M., Pfeiffer, B., ... &
Bodison, S. C. (2017). Using a multifaceted approach to working with children who have
differences in sensory processing and integration. American Journal of Occupational
Therapy, 71(2), 7102360010p1-7102360010p10.

156

Reynolds, S., & Lane, S. J. (2008). Diagnostic validity of sensory over-responsivity: A review of
the literature and case reports. Journal of Autism and Developmental Disorders, 38(3),
516-529.
Robertson, C. E., & Baron-Cohen, S. (2017). Sensory perception in autism. Nature Reviews
Neuroscience, 18(11), 671.
Rogers, S. J., Hepburn, S., & Wehner, E. (2003). Parent reports of sensory symptoms in toddlers
with autism and those with other developmental disorders. Journal of Autism and
Developmental Disorders, 33(6), 631-642.
Ronconi, L., Molteni, M., & Casartelli, L. (2016). Building blocks of others' understanding: a
perspective shift in investigating social-communicative deficit in autism. Frontiers in
Human Neuroscience, 10, 144.
Sabatos-DeVito, M., Schipul, S. E., Bulluck, J. C., Belger, A., & Baranek, G. T. (2016). Eye
tracking reveals impaired attentional disengagement associated with sensory response
patterns in children with autism. Journal of Autism and Developmental Disorders, 46(4),
1319-1333.
Schaaf, R. C., & Lane, A. E. (2015). Toward a best-practice protocol for assessment of sensory
features in ASD. Journal of Autism and Developmental Disorders, 45(5), 1380-1395.
Schauder, K. B., & Bennetto, L. (2016). Toward an interdisciplinary understanding of sensory
dysfunction in autism spectrum disorder: an integration of the neural and symptom
literatures. Frontiers in Neuroscience, 10, 268.
Simmons, D. R., Robertson, A. E., McKay, L. S., Toal, E., McAleer, P., & Pollick, F. E. (2009).
Vision in autism spectrum disorders. Vision Research, 49(22), 2705-2739.

157

Sparrow, S. S., Cicchetti, D. V., & Balla, D. A. (2016). Vineland-III: Vineland Adaptive
Behavior Scales (3rd Ed.). San Antonio, TX: Pearson.
Talay-Ongan, A., & Wood, K. (2000). Unusual sensory sensitivities in autism: A possible
crossroads. International Journal of Disability, Development and Education, 47(2), 201-
212.
Thye, M. D., Bednarz, H. M., Herringshaw, A. J., Sartin, E. B., & Kana, R. K. (2018). The
impact of atypical sensory processing on social impairments in autism spectrum
disorder. Developmental Cognitive Neuroscience, 29, 151-167.
Tomchek, S. D., & Dunn, W. (2007). Sensory processing in children with and without autism: a
comparative study using the short sensory profile. The American Journal of Occupational
Therapy, 61(2), 190-200.
Tomchek, S. D., Little, L. M., & Dunn, W. (2015). Sensory pattern contributions to
developmental performance in children with autism spectrum disorder. American Journal
of Occupational Therapy, 69(5), 6905185040p1-6905185040p10.
Tomchek, S. D., Little, L. M., Myers, J., & Dunn, W. (2018). Sensory subtypes in preschool
aged children with autism spectrum disorder. Journal of Autism and Developmental
Disorders, 48(6), 2139-2147.
Uljarević, M., Baranek, G., Vivanti, G., Hedley, D., Hudry, K., & Lane, A. (2017).
Heterogeneity of sensory features in autism spectrum disorder: Challenges and
perspectives for future research. Autism Research, 10(5), 703-710.
Van Etten, H. M., Kaur, M., Srinivasan, S. M., Cohen, S. J., Bhat, A., & Dobkins, K. R. (2017).
Increased prevalence of unusual sensory behaviors in infants at risk for, and teens with,

158

autism spectrum disorder. Journal of Autism and Developmental Disorders, 47(11), 3431-
3445.
Vasak, M., Williamson, J., Garden, J., & Zwicker, J. G. (2015). Sensory processing and sleep in
typically developing infants and toddlers. American Journal of Occupational
Therapy, 69(4), 6904220040p1-6904220040p8.
Watson, L. R., Patten, E., Baranek, G. T., Poe, M., Boyd, B. A., Freuler, A., & Lorenzi, J.
(2011). Differential associations between sensory response patterns and language, social,
and communication measures in children with autism or other developmental
disabilities. Journal of Speech, Language, and Hearing Research, 54(6), 1562-1576.
Wiggins, L. D., Robins, D. L., Bakeman, R., & Adamson, L. B. (2009). Breif report: sensory
abnormalities as distinguishing symptoms of autism spectrum disorders in young children.
Journal of Autism and Developmental Disorders, 39(7), 1087-1091.
Wilcock, A. A. (2006). An Occupational Perspective of Health. Slack Incorporated.
Williams, K. L., Kirby, A. V., Watson, L. R., Sideris, J., Bulluck, J., & Baranek, G. T. (2018).
Sensory features as predictors of adaptive behaviors: A comparative longitudinal study of
children with autism spectrum disorder and other developmental disabilities. Research in
Developmental Disabilities, 81, 103-112.
Wolff, J. J., Dimian, A. F., Botteron, K. N., Dager, S. R., Elison, J. T., Estes, A. M., ... & IBIS
Network. (2019). A longitudinal study of parent‐reported sensory responsiveness in
toddlers at‐risk for autism. Journal of Child Psychology and Psychiatry.
Yerxa, E. J. (2000). Occupational science: A renaissance of service to humankind through
knowledge. Occupational Therapy International, 7(2), 87-98.

159

Zachor, D. A., & Ben-Itzchak, E. (2014). The relationship between clinical presentation and
unusual sensory interests in autism spectrum disorders: A preliminary
investigation. Journal of Autism and Developmental Disorders, 44(1), 229-235.
Zobel-Lachiusa, J., Andrianopoulos, M. V., Mailloux, Z., & Cermak, S. A. (2015). Sensory
differences and mealtime behavior in children with autism. American Journal of
Occupational Therapy, 69(5), 6905185050p1-6905185050p8.
Zwaigenbaum, L., Bryson, S., Rogers, T., Roberts, W., Brian, J., & Szatmari, P. (2005).
Behavioral manifestations of autism in the first year of life. International Journal of
Developmental Neuroscience, 23(2-3), 143-152.

160

APPENDICES
Appendix 1. Time 3 Follow-up IRB Approval

Proposal #HS-19-00651
University of Southern California Institutional Review Board
1640 Marengo Street, Suite 700
Los Angeles, California 90033-9269
Telephone: (323) 442-0114
Fax: (323) 224-8389
Email: irb@usc.edu
Date: Aug 27, 2019, 10:27am
To: Grace Baranek, PhD
Professor and Chair
OCCUPATIONAL SCIENCE AND OCCUPATIONAL THERAPY (DIVISION 7)
From: University of Southern California Institutional Review Board

TITLE OF PROPOSAL:
North Carolina Child Development Survey Cohort 2 Follow-up (NCCDS2 Follow-up)

Action Date: 8/27/2019 Action Taken: Acknowledged
Committee: Institutional Review Board
Note: The University of Southern California (USC) relies on the University of North Carolina at
Chapel Hill Institutional Review Board (FWA00004801) for IRB review of this study. IRB
review is ceded under the SMART IRB Master Reliance Authorization Agreement.

The USC IRB staff reviewed the following documents to ensure compliance with
institutional requirements:
1. iStar IRB Application HS-19-00651
__________________________________________________

FUTURE SUBMISSIONS TO THE USC IRB:

161

Amendments - Submit amendments to the USC IRB ONLY for the following local
changes:
• Addition or removal of investigators
• New conflict of interest for investigators
• Addition of special populations (adults who are not competent to consent or minors)
• Addition of LAC+USC Medical Center as a study location
• Changes in HIPAA authorization forms or waivers
• Changes in funding or addition of research procedures that require review by other
USC committees (Clinical Trials Office, Department of Contracts & Grants, and
Radiation Safety Committee)

If changes to the study do not involve these local context issues, do not submit an
amendment to the USC IRB. The Central IRB is responsible for reviewing and approving
all other changes.

Reportable Events - You must report adverse events, unanticipated problems, protocol
deviations, participant complaints, and other events to the Central IRB according to the
Standard Operating Procedure for Central Institutional Review Board Reporting. If any
unanticipated problems occur at your site, you must also report them to the USC IRB.
Report the event in iStar as soon as possible, but no later than 10 working days after
you become aware of the event.
__________________________________________________

NOTE TO PI - The SMART IRB LOA has been fully signed and is attached. Please provide
a copy to UNC Chapel Hill.
Attachments:
13-2648 University of Southern Cal relyon UNC SmartIRB LOA - USC signed 8-27-
19.pdf

162

Appendix 2. Age-6 Follow-up Report of Results to Parents (Template)

Dear Parent(s) of OOOOO,
We thank you for your continued participation in the North Carolina Child Development Survey
(NCCDS). This study has been following a cohort of children from infancy to the school years in
an effort to better understand individual differences in children’s development. As part of the
recent survey, you indicated that you wished to receive a report of the findings from the three
questionnaires you completed about your child. The three questionnaires were: the Participation
& Environment Measure – Children & Youth (PEM-CY), Social Responsiveness Scale (SRS),
and Vineland Adaptive Behavior Scales (VABS). Please note that these questionnaires were
conducted as part of a research study, and thus, are not intended as a clinical assessment. The
results are reported below for each of the measures. If you have any questions or concerns about
your child’s development after reviewing this report, or at any point in time, we encourage you to
contact your child’s pediatrician/primary health care provider. You may choose to share this
information with your provider. In addition, general information on some resources related to child
development can be found at the end of this report. If you have specific questions about the study,
please email us at yunjchen@med.unc.edu.

*Date of survey completion: March 30, 2020
Participation & Environment Measure – Children & Youth (PEM-CY)
The PEM-CY evaluates participation, which is about a child’s level of engagement in the
important activities of everyday life – at home, in school, and in the community. These results
are based on your answers to the questionnaire. The scores below indicate your child’s level of
participation in the three different settings, which is defined by the variety of activities he/she
participated in, frequency of his/her participation, and his/her involvement in these activities.
You can compare your child’s scores to the average scores for a normative sample of children
aged 5-11 years of age (maximum score=10). Generally speaking, a higher score indicates a
higher level of participation (i.e., he/she participates in a broader variety of activities, more
often participates in these activities, or is more involved in these activities) within a given
setting. The percentile rank indicates your child’s scores in relation to other children their age.
For example, a percentile rank of 85 means your child scored higher than 85% of the age-
matched normative sample.
(See next page for the scores)

163

++ Most children in the norm sample scored close to maximum (10) on the variety of participated home activities, so an
“average” level would be given even though your child scored maximum or close to maximum on this aspect of participation.

*Notes:
 PEM-CY Percentile ranks & Levels of participation:
PEM-CY
Percentile Rank

Level of Participation
Above 75 High
25 to 75 Average
Below 25 Low

 A list of activities in each setting:
• Home: Computer & video games; Indoor play & games; arts, crafts, music & hobbies; Watching TV,
videos & DVDs; Getting together with people; Socializing using technology; Household chores; Personal
care management; Homework; School preparation
• School: Classroom activities; Field trips & school events; School-sponsored teams, clubs & organizations;
Getting together with peers outside of class; Special roles at school
• Community: Neighborhood outings; Community events; Organized physical activities; Unstructured
physical activities; Classes and lessons; Organizations, groups, clubs & volunteer/leadership activities;
Religious/spiritual gatherings; Getting together with children; Working for pay; Overnight visits or trips

Settings
Aspects of
Participation

Home

School

Community
Variety OOOOO's Score 9.20 8.40 8.92
Percentile Rank
(Level)
60
(Average
++
)
70
(Average)
89
(High)
Frequency OOOOO's Score 8.10 8.70 7.15
Percentile Rank
(Level)
62
(Average)
91
(High)
82
(High)
Involvement OOOOO's Score 8.70 9.20 8.65
Percentile Rank
(Level)
84
(High)
78
(High)
71
(Average)

164

Social Responsiveness Scale, 2
nd
Edition (SRS-2)
The SRS is a parent-report questionnaire for children ages 4 to 18 that assesses challenges or
impairments in everyday social interactions. OOOOO's results can be found below as compared
to a normative sample of school-aged children across the United States. A total T-score is a
standardized total score that indicates your child’s overall level of social challenges as compared
to the normative sample. A total T-score greater than or equal to 60 may indicate social
challenges or impairments that warrant further consultation with your child’s
pediatrician/primary health care provider.

*Notes:
 SRS-2 Total T-score & Levels of social challenge/impairment:
SRS-2
Total T-score
Level of Social
Challenge/Impairment
76 to 90 Severe
66 to 75 Moderate
60 to 65 Mild
38 to 59 Typical/No impairment

OOOOO's
Total T-score
Level of Social
Challenge/Impairment

70

Moderate

165

Vineland Adaptive Behavior Scales, 3
rd
Edition (VABS-3)
The VABS is a parent-report instrument that measures adaptive behaviors, which are the things
that people need to do to function in their everyday lives. These important everyday behaviors
can be grouped into the broad areas of communication, practical daily living skills, and relating
to other people. The specific adaptive behaviors that are needed in life may change as a child
grows older and is less dependent upon the help of others. However, at every age, certain
behaviors and skills are important to daily life functioning in the home, school, and community.
OOOOO's results were compared to those of a normative sample, which is a representative group
of children of the same age as your child from across the United States. The labels below
describe OOOOO's standing in the broad areas described above, plus an overall adaptive
behavior summary score. A higher score indicates a greater level of adaptive skills. The
percentile rank indicates your child’s scores in relation to other children their age. For example, a
percentile rank of 85 means your child scored higher than 85% of the age-matched normative
sample.

Aspects of Adaptive Behavior
OOOOO's
Standard Score
Percentile Rank
(Level)
Communication Skills 120 95
(High)
Daily Living Skills 108 77
(High)
Social Skills and Relationships 125 99
(High)
Overall Adaptive Behavior 111 82
(High)
Motor Skills 120 97
(High)

*Notes:
 VABS-3 Standard scores & Levels of skills:
VABS-3
Standard Score

Level of Skills
130 to 140 High
115 to 129 Moderately High
86 to 114 Adequate
71 to 85 Moderately Low
20 to 70 Low

166

*General Information for Child Development Resources*
Family Support Network of NC (FSN-NC) 🌐 http://www.fsnnc.org
An agency serving parents, families and professionals statewide. Includes information about:
• Diseases, disabilities and chronic illnesses
• Community agencies
• Parent to parent support programs
• Locations of various support programs for parents throughout the state and nation
• Community contact persons for:
o Early Intervention services
o Exceptional Children's Programs through the Public Schools
o Local Interagency Coordinating Councils
Children with Special Health Care Needs Help Line 800-737-3028
The Help Line is an information and referral Help Line for those living with, caring for and concerned about a child
with special health care needs. Callers can learn about health care programs as well as funding resources available to
North Carolina residents.
Exceptional Children’s Division 🌐 https://ec.ncpublicschools.gov
Provides services under the Individual’s with Disabilities Education Act to North Carolina Children and youth ages
3-21 who have disabilities. The Division assures that students with disabilities, and those who are academically or
intellectually gifted, develop mentally, physically, emotionally and vocationally through the provision of an
appropriate individualized education in the least restrictive environment.
*Information provided by North Carolina Department of Health and Human Services

Again, we thank you for your dedicated participation since 2014, which has greatly contributed
to our understanding of child development! If you have any specific questions about the study,
please contact the NCCDS project coordinator, Claire Chen via email at yunjchen@med.unc.edu
or Dr. Grace Baranek at chair@chan.usc.edu. If you have any specific questions or concerns
about your child’s development, we encourage you to reach out directly to your primary health
care provider/pediatrician.

Sincerely,
The NCCDS Team, which includes:
Dr. Linda Watson (Professor, UNC Division of Speech and Hearing Sciences)
Dr. Elizabeth Crais (Professor, UNC Division of Speech and Hearing Sciences)
Dr. Grace Baranek (Professor, USC Chan Division of Occupational Science and Occupational Therapy; Adjunct
Professor, UNC Division of Occupational Science and Occupational Therapy)
Project Coordinator Claire Chen (PhD Candidate, USC Chan Division of Occupational Science and Occupational
Therapy)

167

Appendix 3. Time-2 and Time-3 DCQ Coding Form (REDCap)
#
Variable / Field
Name
Field Label
Field Note
Field Attributes (Field Type, Validation, Choices,
Calculations, etc.)
1 id Coding ID text
2 fyi_id FYI ID text, Required
3 coder Coder's Initials text, Required
4 fam Section Header: Section I. Family
History
Presence of any family history of
autism spectrum disorder (ASD)
or other developmental
disabilities? [fh]
Yes/No, Required
1 Yes
0 No

5 fam_1st
Show the field
ONLY if:
[fam] = '1'
Presence of family history for at
least one 1st-degree relative
(siblings/parents) to the child?
[fm1-6]
Yes/No
1 Yes
0 No

6 fam_1st_re
Show the field
ONLY if:
[fam_1st] = '1'
Please specify the relationship to
the child.
[fm1-6]
checkbox
1 fam_1st_re___1 Siblings
2 fam_1st_re___2 Parents
3 fam_1st_re___3 Other

7 fam_1st_dx
Show the field
ONLY if:
[fam_1st] = '1'
Please specify the type(s) of
diagnosis.
[fd1-6]
checkbox
1 fam_1st_dx___1 ASD
2 fam_1st_dx___2 ADHD/ADD
3 fam_1st_dx___3 Learning Disabilities (e.g., dyslexia)
4 fam_1st_dx___4 Genetic Disorders (e.g., Down syndrome,
Williams syndrome, Fragile X syndrome)
5 fam_1st_dx___5 Intellectual Disability / Developmental
Delay
6 fam_1st_dx___6 Speech/Language Delay/Disorder
7 fam_1st_dx___7 Motor Delay/Disorder
8 fam_1st_dx___8 Sensory Integration/Processing Disorder
9 fam_1st_dx___9 Tourette Syndrome / Tic Disorders / OCD
10 fam_1st_dx___10 Other Mental Disorders (e.g., Depression,
Bipolar disorder, Schizophrenia)
11 fam_1st_dx___11 Other

8 fam_2nd
Show the field
ONLY if:
[fam] = '1'
Presence of family history for at
least one 2nd-degree relative
(aunts/ uncles/ nieces/ nephews/
grandparents/ half-siblings) to the
child? [fm1-6]
Yes/No
1 Yes
0 No

9 fam_2nd_re
Show the field
ONLY if:
[fam_2nd] = '1'
Please specify the relationship to
the child.
[fm1-6]
checkbox
1 fam_2nd_re___1 Aunts/Uncles
2 fam_2nd_re___2 Nieces/Nephews
3 fam_2nd_re___3 Grandparents

168

4 fam_2nd_re___4 Half-siblings
5 fam_2nd_re___5 Other

10 fam_2nd_dx
Show the field
ONLY if:
[fam_2nd] = '1'
Please specify the type(s) of
diagnosis.
[fd1-6]
checkbox
1 fam_2nd_dx___1 ASD
2 fam_2nd_dx___2 ADHD/ADD
3 fam_2nd_dx___3 Learning Disabilities (e.g., dyslexia)
4 fam_2nd_dx___4 Genetic Disorders (e.g., Down syndrome,
Williams syndrome, Fragile X syndrome)
5 fam_2nd_dx___5 Intellectual Disability / Developmental
Delay
6 fam_2nd_dx___6 Speech/Language Delay/Disorder
7 fam_2nd_dx___7 Motor Delay/Disorder
8 fam_2nd_dx___8 Sensory Integration/Processing Disorder
9 fam_2nd_dx___9 Tourette Syndrome / Tic Disorders / OCD
10 fam_2nd_dx___10 Other Mental Disorders (e.g., Depression,
Bipolar disorder, Schizophrenia)
11 fam_2nd_dx___11 Other

11 fam_3rd
Show the field
ONLY if:
[fam] = '1'
Presence of family history for at
least one 3rd-degree relative (first
cousins/great-grandparents) to
the child? [fm1-6]
Yes/No
1 Yes
0 No

12 fam_3rd_re
Show the field
ONLY if:
[fam_3rd] = '1'
Please specify the relationship to
the child.
[fm1-6]
checkbox
1 fam_3rd_re___1 First cousins
2 fam_3rd_re___2 Great-grandparents
3 fam_3rd_re___3 Other

13 fam_3rd_dx
Show the field
ONLY if:
[fam_3rd] = '1'
Please specify the type(s) of
diagnosis.
[fd1-6]
checkbox
1 fam_3rd_dx___1 ASD
2 fam_3rd_dx___2 ADHD/ADD
3 fam_3rd_dx___3 Learning Disabilities (e.g., dyslexia)
4 fam_3rd_dx___4 Genetic Disorders (e.g., Down syndrome,
Williams syndrome, Fragile X syndrome)
5 fam_3rd_dx___5 Intellectual Disability / Developmental
Delay
6 fam_3rd_dx___6 Speech/Language Delay/Disorder
7 fam_3rd_dx___7 Motor Delay/Disorder
8 fam_3rd_dx___8 Sensory Integration/Processing Disorder
9 fam_3rd_dx___9 Tourette Syndrome / Tic Disorders / OCD
10 fam_3rd_dx___10 Other Mental Disorders (e.g., Depression,
Bipolar disorder, Schizophrenia)
11 fam_3rd_dx___11 Other

14 fam_oth
Show the field
ONLY if:
[fam] = '1'
Presence of family history for
other relatives to the child (which
cannot be classified to either of
the categories above)? [fm1-6]
Yes/No
1 Yes
0 No

169

15 fam_oth_dx
Show the field
ONLY if:
[fam_oth] = '1'
Please specify the type(s) of
diagnosis.
[fd1-6]
checkbox
1 fam_oth_dx___1 ASD
2 fam_oth_dx___2 ADHD/ADD
3 fam_oth_dx___3 Learning Disabilities (e.g., dyslexia)
4 fam_oth_dx___4 Genetic Disorders (e.g., Down syndrome,
Williams syndrome, Fragile X syndrome)
5 fam_oth_dx___5 Intellectual Disability / Developmental
Delay
6 fam_oth_dx___6 Speech/Language Delay/Disorder
7 fam_oth_dx___7 Motor Delay/Disorder
8 fam_oth_dx___8 Sensory Integration/Processing Disorder
9 fam_oth_dx___9 Tourette Syndrome / Tic Disorders / OCD
10 fam_oth_dx___10 Other Mental Disorders (e.g., Depression,
Bipolar disorder, Schizophrenia)
11 fam_oth_dx___11 Other

16 con Section Header: Section II.
Concerns
Has the parent or others had
concerns about the child's
development? [c1+c2]
Yes/No, Required
1 Yes
0 No

17 con_lan
Show the field
ONLY if:
[con] = '1'
Presence of concerns related to
language?
[c3+c4]
Yes/No
1 Yes
0 No

18 con_lan_spe
Show the field
ONLY if:
[con_lan] = '1'
Please choose the specific type(s)
of concern related to language.
[c3+c4]
checkbox
1 con_lan_spe___1 Language NOS (Not Otherwise Specified)
2 con_lan_spe___2 Articulation/pronunciation (lisp, stutter,
dropping letter sounds)
3 con_lan_spe___3 Language delay
4 con_lan_spe___4 Reading

19 con_soc
Show the field
ONLY if:
[con] = '1'
Presence of concerns related to
social functioning?
[c3+c4]
Yes/No
1 Yes
0 No

20 con_soc_spe
Show the field
ONLY if:
[con_soc] = '1'
Please choose the specific type(s)
of concerns related to social
functioning.
[c3+c4]
checkbox
1 con_soc_spe___1 Social NOS (Not Otherwise Specified)
2 con_soc_spe___2 Social (not interested in other kids or shy)
3 con_soc_spe___3 Anxiety/Fearfulness (other than social
shyness)
4 con_soc_spe___4 Unusual play with toys

21 con_beh
Show the field
ONLY if:
[con] = '1'
Presence of concerns related to
problematic behaviors?
[c3+c4]
Yes/No
1 Yes
0 No

170

22 con_beh_spe
Show the field
ONLY if:
[con_beh] = '1'
Please choose the specific type(s)
of concerns related to
problematic behaviors.
[c3+c4]
checkbox
1 con_beh_spe___1 Behavior NOS (Not Otherwise Specified)
2 con_beh_spe___2 Rigid behavior (trouble with transitions,
insistence on sameness)
3 con_beh_spe___3 Aggressive behavior/ tantrums / emotion
dysregulation
4 con_beh_spe___4 Overactivity/disruptive/compliance/attention
problem
5 con_beh_spe___5 Sensory behavior (staring at lights, sifting
items)

23 con_med
Show the field
ONLY if:
[con] = '1'
Presence of concerns related to
medical or motor problems?
[c3+c4]
Yes/No
1 Yes
0 No

24 con_med_spe
Show the field
ONLY if:
[con_med] = '1'
Please choose the specific type(s)
of concerns related to
medical/motor problems.
[c3+c4]
checkbox
1 con_med_spe___1 Medical/ Motor NOS (Not Otherwise
Specified)
2 con_med_spe___2 Motor problems (fine and gross motor)
3 con_med_spe___3 Medical problems (heart problem, etc.)
4 con_med_spe___4 Ear/hearing problems (tubes in ears)
5 con_med_spe___5 Other (potty training, etc.)

25 con_oth
Show the field
ONLY if:
[con] = '1'
Presence of other concerns (e.g.,
learning, academics, adaptive
skills)?
[c3+c4]
Yes/No
1 Yes
0 No

26 prof Section Header: Section III.
Diagnosis
Has the parent consulted with
doctors or other professionals?
[cp+dpro1-3]
Yes/No
1 Yes
0 No

27 prof_spe
Show the field
ONLY if:
[prof] = '1'
What type(s) of professional has
the parent consulted with?
[dpro1-3]
checkbox
1 prof_spe___1 Speech Language Pathologist
2 prof_spe___2 Psychologist/Psychiatrist
3 prof_spe___3 Primary Health Care Provider (Pediatrician/
Family Practioner)
4 prof_spe___4 School Teacher/Counselor
5 prof_spe___5 Early Intervention Program (CDSA/DEC)
6 prof_spe___6 Nurse
7 prof_spe___7 Social Worker
8 prof_spe___8 Occupational Therapist/Physical Therapist
(OT/PT)
9 prof_spe___9 Other

28 age_eva
Show the field
ONLY if:
[prof] = '1'
Age in months the child was
evaluated (*Note: Enter the
youngest age if the child was
evaluated at multiple times)
[dage1-3]
text

171

29 rec
Show the field
ONLY if:
[prof] = '1'
Recommended treatment or
intervention
[drt1-3]
checkbox
1 rec___1 None
2 rec___2 Monitor and re-evaluate if needed
3 rec___3 Evaluation in progress/recommended
4 rec___4 Recommend services
5 rec___5 Other

30 dx
Show the field
ONLY if:
[prof] = '1'
Diagnosis given by professional
[dx1-3]
checkbox
0 dx___0 No concern/diagnosis
1 dx___1 ASD
2 dx___2 ADHD/ADD
3 dx___3 Learning Disabilities (e.g., dyslexia)
4 dx___4 Genetic Disorders (e.g., Down syndrome, Williams
syndrome, Fragile X syndrome)
5 dx___5 Intellectual Disability / Developmental Delay
6 dx___6 Speech/Language Delay/Disorder
7 dx___7 Motor Delay/Disorder
8 dx___8 Sensory Integration Disorder
9 dx___9 Tourette Syndrome / Tic Disorders / OCD
10 dx___10 Other

31 dcq_sum Summary of Concern/Diagnosis
Status (based on information
from Section II & III)
checkbox
1 dcq_sum___1 No concern/diagnosis
2 dcq_sum___2 Presence of behavior NOT of specific interest
(not unambiguously autistic). For example,
articulation or behavioral problem
3 dcq_sum___3 Presence of behavior of specific interest
indicative of autism (one domain: social,
language or behaviors)
4 dcq_sum___4 Presence of behavior of specific interest strongly
indicative of autism (at least two domains:
social, language or behaviors).
5 dcq_sum___5 Diagnosis of motor delay or other global
developmental delay (other than autism or
language disorder)
6 dcq_sum___6 Diagnosis of autism
7 dcq_sum___7 Diagnosis of language or sensory integration
disorder
8 dcq_sum___8 Other diagnoses

32 notes Notes notes
33 complete Section Header: Form Status
Complete?
dropdown
0 Incomplete
1 Unverified
2 Complete

Abstract (if available)

Abstract Past research has demonstrated the ubiquitous presence and early emergence of sensory patterns, including hyperresponsiveness (HYPER), hyporesponsiveness (HYPO), and sensory interests, repetitions, and seeking behaviors (SIRS) in children with autism spectrum disorder (ASD), as well as the potential role of sensory processing as a key building block for higher-level social and cognitive functions. Whereas previous findings highlighted the cross-sectional differences of sensory patterns across age and diagnostic groups, it remains unclear how the developmental trajectories of sensory patterns in ASD differ from those with other diagnostic outcomes. More evidence is needed to understand the developmental and heterogeneous nature of sensory patterns in young children with ASD among a general population, as well as how they are associated with broader developmental outcomes, such as adaptive/maladaptive behaviors and participation, in order to evaluate the contribution of sensory patterns to the early identification, diagnosis and prognosis of children with ASD and/or other developmental challenges. ? To address these empirical gaps, this dissertation aimed to 1) model developmental trajectories of sensory patterns from infancy to school age in a community sample and explore the demographic and clinical factors that may account for their variability; 2) identify developmental trajectory subtypes of sensory patterns as associated with school-age outcomes, and 3) understand the specific longitudinal impact of sensory patterns on school-age outcomes. We followed up on a longitudinal cohort of children (N=1,517) from a large community sample whose caregivers completed surveys regarding their child’s sensory patterns and other developmental concerns at three time-points: infancy (Time 1: 6-19 months), pre-school years (Time 2: 3-4 years), and school years (Time 3: 6-7 years). At Time 3, we collected additional outcome data on autism symptoms, adaptive/maladaptive behaviors, and activity participation in a subsample (N=389) that included families who had reported any diagnoses or concerns (N=312) and those who had not reported any diagnoses or concerns at previous time-points (N=77). ? In Study 1, we conducted multivariate latent growth curve modeling (LGCM) with demographic covariates to estimate sensory trajectories from Time 1 to Time 3 with the full sample (N=1,517) and the results revealed highly variable longitudinal patterns across the three sensory patterns. Such variability could be partially explained by demographic characteristics (i.e., child’s sex, race, and parent education) and clinical outcome status (ASD and non-ASD conditions). Particularly, the slopes of HYPER and HYPO were better able to differentiate ASD from other conditions, including non-ASD children with sensory issues. Parent education accounted for more of the variations in trajectories of children’s sensory responsiveness than child’s sex and race. Furthermore, the latent growth factors of the three sensory patterns were associated with each other, indicating their co-occurrence and co-development over time. These findings from Study 1 support the potential utility of longitudinal sensory patterns from infancy for early detection of ASD and the pivotal role of family-tailored approaches to address young children’s sensory challenges. ? To further address the variability that could not be explained by the known factors (i.e., demographics and clinical outcome status) in Study 1, we performed latent class growth analysis (LCGA) to identify the sensory trajectory subtypes from infancy to school age among the full sample in Study 2. We also examined how children classified to these trajectory subtypes differ in their demographic characteristics, clinical outcome status (ASD and non-ASD conditions), and school-age outcomes, including autism symptoms, adaptive/maladaptive behaviors, and activity participation. As a result, we identified five distinct subtypes that vary in latent growth factors across sensory patterns. Particularly, there was a subtype (3% of the sample) characterized by elevated and worsening sensory patterns and highly associated with ASD and poor school-age outcomes. The other four subtypes were characterized by low to moderate levels of sensory patterns and generally stable/improving trajectories and were associated with strengths and weaknesses in school-age outcomes. These results indicated that profiling children based on their early sensory trajectories may help to identify children who are more likely to experience developmental challenges at school age and may thus introduce opportunities for early intervention. ? In Study 3, we examined the specific impacts of sensory trajectories on each school-age outcome variable. Multivariate LGCM was performed with demographic covariates and latent growth factor regressions, and school-age outcome variables were included in the model by being regressed onto the latent growth factor of sensory patterns. Overall, the change rate of HYPER was the most significant predictor of school-age outcomes. The initial levels of sensory patterns had indirect effects on some distal outcomes via the change rates of HYPER and HYPO. Differential impacts of HYPER and HYPO were observed on maladaptive behavior: HYPER was more associated with internalizing behavior while HYPO was more with externalizing behavior. Also, the variations in autism symptoms and maladaptive behavior were explained to larger extents by sensory trajectories as compared to other behavioral domains. These results indicate that early sensory challenges may have cascading effects on other domains of behavior. Thus, sensory responsiveness during this period may be an important target for early intervention towards more optimal outcomes. ? Overall, the findings of these three studies enhance our understanding of the developmental and heterogeneous nature of sensory patterns from infancy to school age in a large community sample with various clinical characteristics, including ASD and non-ASD conditions. The observed associations between early sensory trajectories and later school-age outcomes may indicate the critical role of personalized intervention or services for children with sensory challenges and their families across early development towards optimal outcomes.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Developmental trajectories, maternal-infant transactions, and child outcomes of social-emotional competence and sensory processing

PDF

Examining parent responsiveness to sensory reactivity and regulation cues of infants with developmental risk factors

PDF

Sleep problems, sensory difficulties, and social participation in children with autism spectrum disorders

PDF

Sociodemographic influences on the early caregiving environment and likelihood of autism

PDF

Behavioral and neural influences of interoception and alexithymia on emotional empathy in autism spectrum disorder

PDF

Autonomic and behavioral responses of children with autism to auditory stimulation

PDF

The impacts of the COVID-19 pandemic on therapy utilization among racially/ethnically and socio-economically diverse children with autism spectrum disorder

PDF

Sleep health and variability in youth: a real-time data capture study to examine influences on daily dietary intake patterns and longitudinal weight trajectories

Asset Metadata

Creator Chen, Yun-Ju (Claire) (author)

Core Title Developmental trajectories of sensory patterns in young children with and without autism spectrum disorder: a longitudinal population-based study from infancy to school age

School School of Dentistry

Degree Doctor of Philosophy

Degree Program Occupational Science

Degree Conferral Date 2021-08

Publication Date 07/28/2021

Defense Date 06/04/2021

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag autism,community sample,developmental trajectories,heterogeneity,OAI-PMH Harvest,school-age outcomes,sensory

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Baranek, Grace T. (committee chair), Lawlor, Mary C. (committee member), Sideris, John (committee member), Watson, Linda R. (committee member)

Creator Email chenyunj@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC15659360

Unique identifier UC15659360

Legacy Identifier etd-ChenYunJuC-9916

Document Type Dissertation

Format application/pdf (imt)

Rights Chen, Yun-Ju (Claire)

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu