540646

research-article2014
CPXXXX10.1177/2167702614540646Picci, ScherfTwo-Hit Model of Autism

Theoretical/Methodological/Review Article

Clinical Psychological Science

A Two-Hit Model of Autism: Adolescence 1­–23

© The Author(s) 2014
Reprints and permissions:
as the Second Hit sagepub.com/journalsPermissions.nav
DOI: 10.1177/2167702614540646
cpx.sagepub.com

Giorgia Picci and K. Suzanne Scherf

Department of Psychology, Pennsylvania State University

Abstract
Adolescence brings dramatic changes in behavior and neural organization. Unfortunately, for some 30% of individuals
with autism, there is marked decline in adaptive functioning during adolescence. We propose a two-hit model of
autism. First, early perturbations in neural development function as a “first hit” that sets up a neural system that is “built
to fail” in the face of a second hit. Second, the confluence of pubertal hormones, neural reorganization, and increasing
social demands during adolescence provides the “second hit” that interferes with the ability to transition into adult
social roles and levels of adaptive functioning. In support of this model, we review evidence about adolescent-specific
neural and behavioral development in autism. We conclude with predictions and recommendations for empirical
investigation about several domains in which developmental trajectories for individuals with autism may be uniquely
deterred in adolescence.

Keywords
adolescent peer relations, autism, brain, risk taking, social behavior

Received 4/14/14; Revision accepted 6/2/14

Adolescence is a time of dramatic physical, cognitive, problems, including depression, anxiety disorders, and
emotional, behavioral, and social changes, which may bipolar disorder, as well as a broad range of domains of
have unique consequences for individuals with autism. problem behaviors that include risk taking, alcohol and
This is the developmental period surrounding the transi- substance use, aggression, and violence (Gardner &
tion from late childhood to early adulthood, which Steinberg, 2005; see Steinberg, 2008). Adolescents with
includes the awkward period between sexual maturation autism spectrum disorders not only share these same risks
and the attainment of adult roles and responsibilities with their non–autism spectrum disorder peers but also
(Dahl & Spear, 2004). With adolescence comes pubertal may face additional vulnerabilities during this develop-
maturation as well as the challenge of new developmen- mental period. For example, although the transition from
tal tasks (e.g., formation of high-quality friendships, childhood through late adolescence is generally marked
acquiring autonomy from parents, forming romantic and by improvements in core symptoms (Seltzer, Shattuck,
sexual relationships). Among typically developing (TD) Abbeduto, & Greenburg, 2004) and some social and cog-
individuals, success in the ability to accomplish these nitive skills (Anderson et al., 2007; Anderson, Oti, Lord, &
adolescent-specific developmental tasks predicts levels of Welch, 2009), an estimated 30% of children with autism
adaptive functioning in adulthood (e.g., work compe- experience deterioration in functioning for several years
tence, romantic competence; Roisman, Masten, or more with the onset of puberty ( J. Brown, 1969;
Coatsworth, & Tellegen, 2004). The ability to accomplish Eisenberg, 1956; Gillberg & Schaumann, 1982; Rutter,
such tasks may require drastic reorganization in the func- 1970). This “pubertal deterioration” is associated with
tional organization of underlying neural networks to sup-
port the emergence of new behaviors (see Scherf & Scott,
Corresponding Author:
2012; Scherf, Thomas, Doyle, & Behrmann, 2013). K. Suzanne Scherf, Department of Psychology, Pennsylvania State
In TD individuals, adolescence is a period during University, 113 Moore Hall, University Park, PA 16802
which we see the emergence of many social-emotional E-mail: [email protected]

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016
2 Picci, Scherf

concomitant neurological complications (Gillberg & Developmentally Sensitive Two-Hit

Steffenburg, 1987; Kanne, Gerber, Quirmbach, Sparrow, Model of Autism
Cicchetti, & Saulnier, 2011), a substantial increase in social
withdrawal (Anderson, Mayes, & Lord, 2011), and feelings Alfred Knudson (1971) originally formulated the two-hit
of loneliness (Billstedt, Gillberg, & Gillberg, 2005). hypothesis in an effort to explain a mechanism by which
These aspects of adolescent deterioration in autism multiple genetic mutations can lead to cancerous cell
are particularly concerning in light of findings that ado- development. In Knudson’s seminal model, an initial
lescent success in accomplishing developmental tasks is genetic mutation (i.e., a first hit) is inherited, which
highly predictive of emerging adulthood outcomes causes a predisposition for a second, later occurring
among TD individuals (i.e., success in work, romantic mutation (i.e., a second hit) that greatly increases the
relationships, and friendships; Seiffge-Krenke, Luyckx, likelihood of developing cancerous tumors. It is impor-
& Salmela-Aro, 2014; Yu, Branje, Keijsers, & Meeus, tant to note that the second hit does not necessarily have
2014). Autism is fairly unique as a developmental disor- to be a genetic mutation. It can present in the form of
der in terms of the marked failure of affected individu- biological mechanisms, environmental insults, or some
als to transition into autonomous adult social roles and combination of harmful factors.
levels of adaptive functioning. In contrast, individuals The two-hit hypothesis more recently has begun to
with other kinds of debilitating developmental disorders garner attention in fields outside of genetics. For exam-
(i.e., Down’s syndrome, intellectual disabilities) do ple, it has been used as a model for understanding the
achieve fairly successful adult outcomes in that they get developmental course of schizophrenia. Maynard, Sikich,
married, have children, live autonomously, and maintain Lieberman, and LaMantia (2001) proposed that an early
employment (Esbensen, Bishop, Seltzer, Greenberg, & disruption in cell-to-cell signaling during neural develop-
Taylor, 2010; Matson, Dempsey, & Fodstad, 2009; ment (e.g., neurogenesis, cell differentiation, and migra-
Whitehouse, Watt, Line, & Bishop, 2009). In this article, tion) is a first hit that makes the brain vulnerable to a
we propose that adolescence may be an especially vul- second hit, which they argued occurs in adolescence and
nerable period of development in autism that makes the further compromises the already vulnerable brain.
transition into adult social roles and adult levels of Maynard et al. suggested that the second hit is usually an
adaptive functioning overwhelmingly difficult. We frame environmental factor that triggers the onset of symptoms
this vulnerable developmental period in the context of in the primed system, which is “built to fail” (see Fig. 1
a “two-hit” model (e.g., Knudson, 1971) that integrates for an adaptation of the Maynard et al. two-hit model of
genetic, environmental, and developmental factors to schizophrenia). This two-hit model has provided a valu-
explain the onset and developmental course of autism able framework for integrating genetic, environmental,
through early adulthood. and developmental factors to explain the onset and
The article is organized as follows. We begin by fram- developmental course of schizophrenia. Here, we have
ing the notion of a two-hit model as it was first concep- adapted this model to predict that individuals with autism
tualized and has been used to explain and make experience two developmental insults that uniquely
predictions about disorders, such as schizophrenia. Then influence their ability to take on adult social roles and
we briefly review existing evidence for early perturba- levels of adaptive functioning.
tions in neural development as a potential “first hit” for We propose that a two-hit model will provide a frame-
individuals with autism that predisposes the neural sys- work for integrating the seemingly disparate literature on
tem to a “second hit.” Next, we explain how the conflu- causal factors contributing to the developmental course
ence of pubertal hormones, neural reorganization, and of autism and will also direct research toward under-
the increasing demands of adolescence may fundamen- standing how adolescence may be an especially vulner-
tally interfere with the ability to transition into adult able period for the developmental trajectory and
social roles and levels of adaptive functioning. In so outcomes of autism. We argue that for the model to have
doing, we review the scant evidence on adolescent-spe- predictive power, it must be sensitive to findings that the
cific neural and behavioral development in autism and behavioral and neural manifestations of autism differ
use our novel model to make predictions about how throughout development. For example, although failing
developmental trajectories for individuals with autism to respond to one’s own name is a hallmark feature of
may be uniquely deterred in adolescence. We conclude autism in infancy, it is not a diagnostic criterion in later
with suggestions about several areas of future study we childhood. Similarly, with respect to brain development,
consider essential for understanding whether and how the intrinsic functional connectivity of neural networks
adolescence is a second hit in the developmental course appears to be enhanced in children with autism com-
of autism. pared with TD children but reduced in adolescents and

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016
Two-Hit Model of Autism 3

Neurodevelopmental Priming
First Hit Second Hit

Prenatal Genetic and Environmental Environmental Factors

Disruptions (i.e., maternal nutritional (i.e., stressors)
deficiencies, viral infections in utero, genetic
predispositions)

Schizophrenia
Developing Central Symptoms
Vulnerable Neuronal Circuits
Nervous System

Fig. 1. Schematic illustration of a two-hit model of schizophrenia. The model proposes that there is a first hit via prenatal genetic or envi-
ronmental disruptions in the developmental central nervous system. As a result, the system is primed with vulnerable neuronal circuitry that
is subsequently exacerbated by a second hit to the system via potential environmental factors that lead to the onset of schizophrenia systems,
specifically during adolescence. Adapted from Maynard, Sikich, Lieberman, and LaMantia (2001).

adults with autism compared with TD adults (Uddin, (a) the relative stability (or lack thereof) of neural atypi-
Supekar, & Menon, 2013). calities (e.g., increased grey matter volume), without
Our two-hit model of autism accommodates these much regard for how these atypicalities affect levels of
developmental considerations and emphasizes the con- adaptive functioning or even behavior across develop-
tributions of developmental influences that accompany ment; or (b) atypical profiles of neural activation (e.g.,
the transitions into and out of adolescence, which are atypical intrinsic functional connectivity) or structural
predictive of adult outcomes for TD individuals. Figure 2 organization (e.g., deficient macrostructural properties of
provides a schematic representation of our model. We particular fiber tracts) within a specific developmental
suggest that the first hit likely results from early disrup- period (e.g., childhood or adulthood) but not as a func-
tions to neural development (even as early as prenatal tion of development. Existing research also has yet to
development) that fundamentally compromise develop- consider the adolescent deterioration that is often
ing neural circuits. In turn, these compromised neural reported in autism (e.g., Gillberg & Steffenburg, 1987).
circuits contribute to poor behavioral outcomes and atyp- The goal of our model is to provide a conceptual link
ical neural organization in infancy and childhood (Fig. 2, between these early neural atypicalities, the reported
Prediction 1). This compromised neural circuitry lays the adolescent deterioration, and the failure of individuals
foundation for a subsequent hit, which we argue is the with autism to successfully transition into adult levels of
developmental period of adolescence (Fig. 2, Predictions adaptive functioning (e.g., low rates of independence
2a and 2b). More specifically, we contend that the con- from parents, high rates of unemployment, and social
comitant pressures of adolescent-specific developmental isolation; Henninger & Taylor, 2013; Howlin, Moss,
tasks and pubertal maturation combine to create a sec- Savage, & Rutter 2013). Whereas previous explanations
ondary hit, thereby resulting in a failure to develop new of functional deterioration in autism have focused on the
behaviors that are essential for acquiring adult levels of influence of increasing environmental demands that alter
adaptive functioning. In sum, this developmentally sensi- or magnify the presentation of symptoms (Levy & Perry,
tive two-hit model strives to explain the mechanistic fac- 2011), we highlight the unique biological (i.e., pubertal
tors that prevent individuals with autism from transitioning development and neural reorganization) and environ-
into adult social roles and levels of adaptive functioning. mental (i.e., developmental tasks, increasing social
It is important to note that we are not the first research- demands and expectations) influences that uniquely
ers to emphasize the role of early neural atypicalities in combine during adolescence to alter developmental tra-
the developmental course of autism (e.g., Courchesne, jectories in autism. Also, unlike models of “regressive
Campbell, & Solso, 2011; Courchesne et al., 2007). autism,” which attempt to explain the acquisition and
However, with few exceptions, the vast majority of the subsequent loss of skills or behaviors (Hansen et al.,
work in this area has contributed to our understanding of 2008), our model is designed to explain why individuals

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016
4 Picci, Scherf

Neurodevelopmental priming
1st Hit 2nd Hit

Genetic and neurodevelopment Adolescent-related developmental

disruptions (e.g., migratory and cell tasks & pubertal hormones 2b
differentiation defects, brain
overgrowth)

Developing CNS Vulnerable neuronal circuits Poor Adolescent Failure to transition

Outcomes into adult roles

1 2a

Poor Childhood Outcomes

Fig. 2. Schematic illustration of a developmentally sensitive two-hit model of autism. We propose that the first hit (i.e., genetic and neurode-
velopmental disruptions) calibrates the system to be fundamentally vulnerable, which leads to poor childhood outcomes. A subsequent second
hit (adolescent-specific developmental tasks and pubertal hormones) compounds previous neural vulnerabilities in addition to creating new and
importantly different vulnerabilities, which lead to impediments in attaining adult levels of adaptive functioning. Colored lines within the diagram
are numbered by area of research. The blue line predicts links between vulnerable neural circuits and poor childhood outcomes, which has been
minimally characterized in the literature. The green lines are symbolic of two predictions that have yet to be addressed: how poor childhood out-
comes cascade into poor adolescent outcomes and how poor adolescent outcomes (as a result of a second hit) lead to a failure to transition into
adult social roles and levels of adaptive functioning.

with autism overwhelmingly fail to acquire adult levels of Benayed et al., 2005; Coe, Girirajan, & Eichler, 2012;
adaptive functioning. Tabuchi et al., 2007). Abnormal expression of these genes
Our hope is that this model will provide researchers during prenatal development leads to the disorganization
with a conceptual framework to investigate potential of neurons across multiple cortical layers, particularly in
adolescent-specific vulnerabilities that may contribute to the prefrontal and temporal lobes, which is likely related
deleterious outcomes for individuals with autism as they to neuronal migratory defects and gene-sequencing dis-
attempt to transition into adult social roles. In so doing, ruptions (Stoner et al., 2014).
we build on the emerging findings of adolescent-specific Longitudinal brain imaging studies have revealed that
neural and behavioral development in TD individuals, early brain development follows an altered developmen-
which have clearly identified adolescence as a crucial tal course in autism, particularly in infancy and toddler-
period that is predictive of adult levels of adaptive hood. Specifically, there are converging reports of early
functioning. accelerated growth of both grey and white matter in the
first years of life that is followed by a marked deceleration
in overall brain growth in late childhood (8–9 years of
What could be the first hit in autism? age; Courchesne et al., 2007). There is also evidence of a
We suggest that early (even prenatal) structural brain plateau or decrease in neural growth rate specifically dur-
abnormalities, which occur before the onset of behav- ing the adolescent years, which is dramatically different
ioral symptoms in autism (Courchesne, Campbell, & from the typical trajectory of brain development
Solso, 2011; Stoner et al., 2014), are a likely first hit in the (Courchesne, Webb, & Schumann, 2011; Hardan, Libove,
developmental course of autism. There is mounting evi- Keshavan, Melhem, & Minshew, 2009). Critically, this
dence from genetic, prenatal, infant, and longitudinal decline in neural growth during adolescence is arguably
neuroimaging studies that structural and functional con- not an extension of, but instead categorically different
nections in the brain are compromised in autism. For from, the early phase of brain overgrowth (Courchesne,
example, the genes that have been consistently impli- Campbell, & Solso, 2011; Schumann & Amaral, 2006; van
cated in autism etiology are involved in multiple aspects Kooten et al., 2008; Vargas, Nascimbene, Krishnan,
of neural circuitry development (i.e., neuronal migration, Zimmerman, & Pardo, 2005). More specifically, the
cell differentiation, signaling pathways, axon and syn- arrested and even decelerated patterns of brain develop-
apse form/function, etc.; Abrahams & Geschwind, 2010; ment during adolescence may be a reactive consequence

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016
Two-Hit Model of Autism 5

to the initial early disruption (Courchesne, Webb, & focused on investigating age-related changes in these
Schuman, 2011). These findings are compatible with our abnormalities of structural connectivity.
two-hit model in that the early developmental disruption The existing evidence overwhelmingly suggests that
in brain growth and organization may lead to early vul- early structural brain abnormalities in individuals with
nerabilities (e.g., poor childhood outcomes), which may autism introduce vulnerabilities in brain organization that
become compounded by later occurring developmental have severe consequences for neural organization and
processes that lead to further deterioration, specifically are related to symptom severity. We argue that these early
during adolescence. neural aberrations represent the first hit to individuals
Similarly, studies of functional and structural neural with autism, thereby setting up a neural system that is
connectivity in children and adults with autism are consis- built to fail in the face of a second hit.
tent with the notion that brain organization is atypical
throughout the full developmental course of autism.
The second hit: Adolescence
Functional organization, that is, the coordinated neural
activity between regions, is reportedly atypical during In our model, we propose that the concomitant pressures
cognitive tasks (e.g., mental rotation, face processing, of adolescent-specific developmental tasks together with
Tower of London, visuomotor) as well as during rest in all the surge of pubertal hormones create a secondary hit for
ages tested (i.e., school-age children through adulthood; individuals with autism that their compromised neural
Just, Cherkassky, Keller, Kana, & Minshew, 2007; Mizuno, circuitry cannot accommodate. The result is a failure to
Villalobos, Davies, Dahl, & Müller, 2006; Villalobos, acquire the critical new behaviors that are essential for
Mizuno, Dahl, Kemmotsu, & Müller, 2005). In some cases, the transition to adult social roles and levels of adaptive
the extent of atypicality in these patterns of functional functioning. In what follows, we provide examples of
organization has been linked to symptom severity (Assaf critical domains in which TD adolescents and adoles-
et al., 2010; Monk et al., 2009; Supekar et al., 2013; Weng cents with autism fundamentally diverge in their devel-
et al., 2010). One significant limitation of these studies is opmental trajectories in ways that are specific to
that they were not specifically designed to study age- adolescence. We contend that these diverging trajectories
related changes in functional brain organization to address provide initial evidence for considering adolescence as a
how developmental trajectories in individuals with autism second hit to individuals with autism. These findings
may differ from those in TD individuals. Uddin and col- begin to reflect the extensive ways in which adolescence
leagues (2013) have suggested that the nature of the atyp- may uniquely and negatively affect individuals with
icality in functional organization of neural circuits may be autism, thereby setting them on an altered trajectory in
fundamentally different in children with autism compared their transition into adulthood.
with adolescents and adults with autism, which is alto- There are particularly vulnerable neural regions in
gether different from typical comparison groups. Similar which reduced volume and increased neuron loss emerge
to the findings of early gross structural brain develop- in adolescence and extend into adulthood among indi-
ment, this claim suggests that there may be a categorical viduals with autism (Courchesne, Campbell, & Solso,
shift in the functional organization of neural networks 2011; Schumann & Amaral, 2006; van Kooten et al., 2008).
particularly during adolescence in autism, which lends In particular, the amygdala exhibits early overgrowth in
support to our argument that adolescence may be a par- childhood followed by a distinct decline in volume dur-
ticularly vulnerable time (i.e., the second hit) in the devel- ing adolescence for individuals with autism (Mosconi
opmental course of autism. et al., 2009; Nacewicz et al., 2006; Sparks et al., 2002). In
Finally, findings from structural imaging methodolo- contrast, the amygdala continues to grow in volume
gies, such as diffusion tensor imaging, have suggested throughout adolescence in TD individuals (Hu, Pruessner,
that there is reduced integrity in white matter tracts in Coupé, & Collins, 2013; Uematsu et al., 2012). In addi-
both children and adults with autism, especially with tion, the caudate nucleus of the striatum evinces an atypi-
respect to the corpus callosum, the largest white matter cal growth pattern; it shows an increase in volume during
tract connecting the right and left hemispheres (Barnea- adolescent years for individuals with autism that is linked
Goraly et al., 2004; Just et al., 2007; Kana, Keller, with more severe symptomology (Hollander et al., 2005;
Cherkassky, Minshew, & Just, 2006; Lee et al., 2007; Rojas et al., 2006). Conversely, in TD individuals, the cau-
Schaer et al., 2013; Shukla, Keehn, & Müller, 2011). White date decreases in volume during adolescent development
matter tracts connecting many other brain regions are (Giedd, 2006; Hu et al., 2013; Uematsu et al., 2012). These
disrupted as well (e.g., Nair, Treiber, Shukla, Shih, & findings illustrate that growth patterns in some neural
Müller, 2013; Schaer et al., 2013). Unfortunately, as with regions categorically diverge in adolescence for individu-
the functional connectivity studies, few studies have als with autism.

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016
6 Picci, Scherf

In the domain of visuoperceptual processing, TD indi- anxiety during adolescence (Brereton, Tonge, & Einfeld,
viduals become increasingly attuned to the global prop- 2006; Kuusikko et al., 2008; Mayes, Calhoun, Murray, &
erties of visual stimuli, particularly during adolescence, Zahid, 2011; McPheeters, Davis, Navarre, & Scott, 2011).
whereas individuals with autism do not (Scherf, Luna, Depression and anxiety are the most common disorders
Kimchi, Minshew, & Behrmann, 2008). Specifically, early that co-occur with autism; as many as 65% of individu-
in development, both TD children and those with autism als with autism are affected with depression, anxiety, or
are biased to detect more local features of objects. By both (Ghaziuddin, Ghaziuddin, & Greden, 2002; Green
early adolescence, TD individuals begin to become more & Cox, 2000). In one large-scale study of 11- to 17-year-
biased for detecting the global properties of visual stimuli olds with autism, anxiety and depression symptoms
(Scherf, Behrmann, Kimchi, & Luna, 2009); however, this were highly correlated with one another and were both
emerging sensitivity is not present among adolescents linked with the severity of autism symptoms (Mayes,
with autism. This puts individuals with autism at a disad- Calhoun, Murray, & Zahid, 2011). In contrast, other
vantage in processing the more global properties of research has suggested that higher functioning children
objects, which are especially relevant for distinguishing and adolescents with autism evince higher prevalence
perceptually homogenous visual objects, such as indi- rates of anxiety (79%) and depression (54%) than do
vidual faces (Caron, Mottron, Berthiaume, & Dawson, lower functioning individuals with autism (67% and
2006). In addition, the ability to recognize individual 42%, respectively; Mayes, Calhoun, Murray, Ahuja, &
unfamiliar faces diverges specifically in adolescence Smith, 2011).
between TD individuals and individuals with autism These findings converge on the notion that across
(O’Hearn, Schroer, Minshew, & Luna, 2010). Although TD many domains, developmental trajectories for individuals
individuals continue to improve in these skills, those with with autism plateau or worsen during adolescence, which
autism plateau in adolescence and remain relatively means they fail to acquire skills or make the same transi-
impaired as adults. tions as do TD adolescents. This may in turn put them at
Executive functioning and working memory abilities great risk for developing psychopathology, worsening of
also diverge during adolescence in individuals with basic cognitive abilities, and failing to accomplish the
autism. Specifically, individuals with autism worsen in core developmental tasks of adolescence.
their metacognitive abilities (i.e., working memory, plan-
ning and organization) in real-world executive function-
ing tasks during adolescence, whereas TD individuals Critical Developmental Domains That
improve (Rosenthal et al., 2013). Critically, these trajecto- May Be Especially Vulnerable for
ries of metacognitive skills distinctly diverge during early
adolescent development and continue to decline later in
Adolescents With Autism
adolescence and into emerging adulthood (between 11– In what follows, we review findings that support our
13 and 14–18 years of age). Metacognitive abilities are notion that the increasing social demands and chal-
critical for adaptive functioning in the social world; there- lenges of adolescence cannot be accommodated by the
fore, worsening of these important skills specifically dur- atypical and vulnerable neural systems of individuals
ing adolescence is especially problematic for individuals with autism. The resulting consequence is a severe limi-
with autism. tation in the ability to transition into adult social roles
Finally, adolescents with autism and TD adolescents and levels of adaptive functioning. We focus on findings
begin to diverge even in more basic biological domains, from three primary domains: peer relationships, risk
such as sleep. For instance, there is evidence for worsen- taking and sensation seeking, and cognitive control.
ing sleep problems that emerge disproportionately in Behavior in these domains directly contributes to meet-
individuals with autism compared with control individu- ing the challenges of adolescent-specific developmental
als in the early stages of adolescent development tasks for TD adolescents and facilitates the transition to
(between ages 11 and 13; Sivertsen, Posserud, Gillberg, adulthood. For this reason, our model specifically pre-
Lundervold, & Hysing, 2012). Given that disturbed sleep dicts that the behavioral and neural foundation of these
is related to emerging psychopathology specifically dur- domains is likely to be particularly disrupted for indi-
ing adolescence for TD individuals (Dahl, 1996; viduals with autism. For each domain, we briefly discuss
Fredriksen, Rhodes, Reddy, & Way, 2004; Roberts & the findings from current research on typical adoles-
Duong, 2014), this finding might suggest that adolescents cents, the evidence (if any) about how these domains
with autism are at even greater risk for developing ado- develop in adolescents with autism, and hypotheses
lescent-onset psychopathology. about how the domains may present specific challenges
In fact, individuals with autism may be at a height- and increase vulnerabilities for individuals with autism
ened risk for developing comorbid depression or during adolescence.

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016
Two-Hit Model of Autism 7

Peer relationships Unfortunately, very little is known about whether ado-

lescents with autism experience this kind of adolescent-
During adolescence, peers begin to take on new salience specific peer reorientation or increasing neural salience
and relevance for TD individuals, which amounts to a of peer-related information. Adults with autism clearly
social reorientation from parents toward peers (B. struggle to manage the demands of adult relationships
Brown, 2004; Nelson, Leibenluft, McClure, & Pine, and social roles. Specifically, they struggle in forming
2005).1 With this social reorientation emerges an unprec- romantic partnerships (Billstedt et al., 2005) and high-
edented drive for acceptance from social peers and quality friendships (Orsmond, Krauss, & Seltzer, 2004), as
hypersensitivity to peer evaluation. Adolescent peer rela- well as in gaining autonomy from parents (Eaves & Ho,
tionships are more elaborate than friendships at any ear- 2008). Here, we consider the possibility that individuals
lier developmental period (for review, see B. Brown, with autism fail to accomplish the developmental task of
2004). Peers become a critical source of social support forming intimate peer relationships in part because of a
(B. Brown, Eicher, & Petrie, 1986) as well as the focus of failure to undergo peer reorientation during adolescence
new romantic and sexual interests (see Collins, Welsh, & that may be related to atypical neural processing of infor-
Furman, 2009). Adolescents’ emotional responses to mation about peers.
social stimuli are intensified and are modulated by social
contexts involving peers. For example, in the presence
of peers, TD adolescents are more prone to erratic and Peer relationships differ in quantity, quality, and
emotionally influenced behavior, even as they are nature in autism. There is good evidence that the
achieving adultlike competence in many cognitive abili- quantity, quality, and nature of friendships are different
ties (Dahl & Spear, 2004). All of these facets of social among adolescents with autism. One study reported that
reorientation that are newly emerging in adolescence are roughly 24% of adolescents with autism do not have
critical for social competence with peers, which predicts reciprocal peer relationships (Mazurek & Kanne, 2010).
work and romantic competence in young adulthood Adolescents with autism, compared with TD adolescents,
(Roisman et al., 2004). spend less time with peers and more quality time with
This reorientation toward peers is also reflected in parents and other adults (Kuo, Orsmond, Cohn, & Coster,
changes in the neural processing of peer-related infor- 2013; Solish, Perry, & Minnes, 2010). The very notion of
mation in TD adolescents. Specifically, a number of friendship appears to be qualitatively different in autism.
studies have yielded converging evidence that process- Adolescents with autism describe a “friend” as someone
ing of peer-related information (e.g., peer presence, with whom they share common interests (Carrington,
socially evaluative contexts, peer acceptance or rejec- Templeton, & Papinczak, 2003). This characterization is
tion) at the neural level is categorically different in ado- supported by reports that their friendships are less social
lescence compared with childhood or adulthood, in nature and more focused on shared circumscribed
especially within regions of the “social brain” (Chein, interests (e.g., playing video games; Bauminger &
Albert, O’Brien, Uckert, & Steinberg, 2011; Jankowski, ­Schulman, 2003; Kuo et al., 2013). Adolescents with autism
Moore, Merchant, Kahn, & Pfeifer, 2014; Jones et al., are less likely to rely on peers for social and romantic
2014; Pfeifer et al., 2013). Specifically, regions support- learning than are TD individuals (Stokes, Newton, & Kaur,
ing reward processing (e.g., ventral striatum) and social 2007). This potentially deprives them of important con-
cognition (e.g., amygdala, temporoparietal junction, texts for learning about social relationships. For example,
temporal pole) tend to show increased or differential among TD adolescents, peer friendships foster vital social
activation patterns in adolescents in tasks that elicit pro- skills (e.g., conflict resolution, interpersonal competence,
cessing of peer information ( Jankowski et al., 2014; sharing, reciprocal communication; Hartup, 1996) that are
Jones et al., 2014; Peake, Dishion, Stormshak, Moore, & transferable to other social relationships (e.g., romantic
Pfeifer, 2013; Rodrigo, Padrón, de Vega, & Ferstl, 2014). relationships; Bagwell, Newcomb, & Bukowski, 1998).
These findings suggest that peers, compared with adults In addition, the nature of how peer friendships are
and children, are inherently more rewarding for TD initiated and evolve is different for adolescents with
adolescents, which may motivate the social reorienta- autism. In contrast with TD individuals, most friendships
tion toward peers during adolescent development. for adolescents with autism are prearranged and scaf-
Critically, the same regions (caudate and amygdala) that folded by parents. One study has suggested that roughly
are increasingly sensitive to peer-related information in 50% of friendships involving an adolescent with autism
TD adolescents show early and ongoing disrupted emerge in the context of a prearranged setting, and a
growth patterns in individuals with autism, which may minor 8% of friendships involve reciprocal activities out-
severely limit alterations in the neural processing of side of an organized or prearranged setting (Orsmond
peer-related information during adolescence. et al., 2004).

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016
8 Picci, Scherf

Finally, there is strong evidence that adolescents with likely stems from not only their existing impaired social
autism often fail miserably in their attempts to form peer skills, which are apparent even in childhood, and might
relationships, which is reflected in the excessively high be traced back to early aberrant development of neural
rates of bullying (Little, 2002), an extreme form of peer regions in the social brain, but also a lack of increasing
rejection. Several studies have reported that adolescents salience and relevance of peers during adolescence.
with autism are often victims of bullying, with rates
approaching 50% (Cappadocia, Weiss, & Pepler, 2012; Neural processing of peer-related information is
Rowley et al., 2012; Sterzing, Shattuck, Narendorf, atypical in autism. Consistent with our notions that
Wagner, & Cooper, 2012; van Roekel, Scholte, & Didden, early atypical neural development lays the foundation for
2010). Like TD adolescents, adolescents with autism a subsequent second hit in the developmental course of
report feelings of distress related to the experience of autism, some evidence has suggested that neural pro-
being socially excluded, even in laboratory simulations cessing of peer-related information is atypical in adoles-
of peer rejection (Bolling et al., 2011; Masten et al., 2011). cents with autism, specifically within regions of the social
These findings reveal that adolescents with autism are brain that exhibit early aberrant growth patterns. Much of
cognizant of the fact that they are being rejected; how- the work evaluating the neural processing of peer-related
ever, it is unclear whether these feelings of distress are information in autism has employed experimental peer
particularly meaningful to or motivating for them. Finally, exclusion and reward-based learning paradigms. In two
given their difficulties forming and maintaining high- studies, researchers measured brain activation while par-
quality peer friendships, adolescents with autism are not ticipants played a game of Cyberball (Bolling et al., 2011;
likely to be able to draw on peer-based social support Masten et al., 2011). In this game, participants play catch
that is protective against the effects of peer rejection for with two other unfamiliar peer “players” who eventually
TD adolescents (Hodges & Perry, 1999). Together, these exclude the participant and pass the ball only between
findings suggest that individuals with autism are at a loss themselves. During peer exclusion, adolescents with
in terms of forming and maintaining intimate peer rela- autism (aged 7–18 years) report similar levels of distress
tionships in adolescence—a critical developmental task resulting from the exclusion as do TD adolescents; how-
that foreshadows a successful transition into adult social ever, they evince weaker activation throughout a whole
roles and levels of adaptive functioning. network of neural regions that have been previously
This difficulty initiating and maintaining peer relation- implicated in processing social exclusion in adults,
ships may contribute to the awkward attempts to culti- including right posterior insula, left ventral anterior cin-
vate romantic relationships seen among individuals with gulate cortex, left posterior cingulate cortex, hippocam-
autism. According to some reports, 44% of adults with pus, and precuneus (Bolling et al., 2011). Moreover,
autism never date (Farley et al., 2009), which reflects the individuals who exhibited more severe autism symptoms
enormous challenge that sexual and romantic relation- evinced the weakest activation in these regions. Similarly,
ships pose for them (Henault, 2006). The adults with adolescents with autism (mean age = 14 years) displayed
autism who do manage to engage in romantic relation- less neural activation in the ventral anterior cingulate cor-
ships are often lacking in sexual knowledge (Byers, tex, anterior insula, ventral striatum, and ventrolateral
Nichols, Voyer, & Reilly, 2013). Although there is a pau- prefrontal cortex, which are regions that have been previ-
city of research exploring the nature and frequency of ously linked with the distressing responses of peer exclu-
romantic relationships among adolescents with autism, sion and with the process of regulating such responses
the work that does exist supports the notion that these during a game of Cyberball (Masten et al., 2011). These
individuals have immense difficulty initiating and engag- findings revealed that adolescents with autism report dis-
ing in these relationships. For example, adolescents with tress when being socially excluded by unfamiliar peers
autism tend to focus their romantic attention and efforts but that simulated social exclusion by peers may not be
toward strangers and engage in inappropriate courting as stressful for them as it is for TD adolescents. On one
behaviors (e.g., stalking behaviors; Stokes et al., 2007). hand, this relative insensitivity might be protective against
They also do not appear to be able to learn about notions the distress of peer rejection. On the other hand, a weak-
such as personal privacy and appropriate dating behavior ened drive for acceptance by peers and hyposensitivity
even in the context of formal sexual educational training to peer evaluation might also limit processing of impor-
(Stokes & Kaur, 2005). tant feedback signals about the relative success with
In sum, adolescents with autism struggle tremendously which adolescents with autism form high-quality peer
in their attempts to develop intimate peer relationships friendships.
(friendship as well as romantic), which is a core develop- In addition to becoming increasingly sensitive to the
mental task of adolescence that predicts adult levels of evaluation and acceptance of peers, the reward value of
adaptive social functioning. We propose that this difficulty peer-related information increases dramatically in TD

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016
Two-Hit Model of Autism 9

adolescents (Blakemore, 2008; Galvan, 2010), which is areas supporting reward learning (i.e., anterior cingulate
reflected in enhanced neural activation within the ventral cortex, ventral prefrontal cortex, and ventral striatum)
striatum and orbitofrontal cortex (Chein et al., 2011). To during social-reward learning than did TD adolescents.
our knowledge, no studies have specifically measured There were no group differences during monetary-
reward-related neural responses to peers among adoles- reward learning.
cents with autism. However, there is a related line of
research that has focused on potential dysfunction of Hypotheses about peer relationships among adoles-
neural response to social versus monetary rewards in cents with autism. On the basis of this literature, we
individuals with autism. Although this work does not use hypothesize that there is a significantly weakened (and
peer-related information as the social reward, it may potentially nonexistent) social reorientation toward peers
reveal atypicalities in the basic neural processing of social among adolescents with autism. Using our model, we pre-
reward that may be implicated in the failure among ado- dict that this limited social reorientation originates from
lescents with autism to socially reorient toward peers. early atypical neural organization, particularly within the
This emerging literature converges on the notion that mesocorticolimbic reward circuitry, that disproportion-
the neural basis of reward processing is disrupted in indi- ately disrupts processing of social rewards. The burden of
viduals with autism (Delmonte et al., 2012; Kohls et al., this disruption becomes especially apparent during ado-
2011; Kohls, et al., 2013; Lin, Rangel, & Adolphs, 2012); in lescence, when adolescents with autism are not driven to
comparison to responses observed in TD adolescents pursue high-quality peer friendships and navigate the
and adults, individuals with autism experience a dispro- increasingly complex nature of adolescent friendships. In
portionate impairment in the processing of social rewards turn, this severely limits the quality and nature of their
compared with monetary rewards. For example, com- peer relationships in adolescence, which ultimately
pared with TD individuals, adolescents and adults (13–25 restricts their ability to accomplish other core adolescent-
years old) with autism evinced reduced activity in the specific developmental tasks, such as forming romantic
dorsal striatum during the receipt of social rewards (i.e., and sexual relationships and becoming more autonomous
a smiling unfamiliar female face; Delmonte et al., 2012). from parents. In the Future Directions for Studying Ado-
Importantly, there were no such differences in the receipt lescence as a Second Hit for Autism section, we discuss
of monetary rewards, thereby suggesting that these defi- potential experimental paradigms and sets of findings that
cits are specific to social rewards. Also, decreased activa- would help evaluate these hypotheses.
tion to the social rewards was related to increased
repetitive behaviors in the autism group, which suggests
that the core behavioral deficits of autism may funda-
Risk taking and sensation seeking
mentally interfere with the ability to evaluate the salience During adolescence, TD individuals evince a marked
and relevance of social stimuli. Recent findings of typical increase in risk-taking behaviors, particularly in the con-
or even enhanced reward-related activation in response text of peer interactions (Chein et al., 2011). Although the
to nonsocial cues (e.g., food; Cascio et al., 2012) and heightened risk-taking behavior that is common in typi-
personalized stimuli of interest (Cascio et al., 2014; cal adolescents can have severe consequences (i.e. delin-
Dichter et al., 2012) among adolescents with autism have quency, injury, death), these changes in behavior are
demonstrated that the processing of social rewards may normative. This interest in exploring new experiences
be disproportionately disrupted in autism. (i.e., sensation seeking) is a hallmark feature of adoles-
Similarly, social rewards appear to be less motivating cence and is likely functionally important for motivating
or useful in learning paradigms for individuals with them to meet core developmental tasks of adolescence,
autism. Specifically, in a reward-learning task with mon- including taking on more adult social roles and relation-
etary (winning or losing money) and social (positive and ships and obtaining autonomy from parents (Spear, 2000;
negative faces) rewards, young adults with autism had a Steinberg, 2008).
slower learning rate for the social-reward stimuli com- Several models of brain development argue that the
pared with the monetary-reward stimuli despite being adolescent-specific increase in risk-taking behaviors is
able to accurately identify the valence of the social stim- enabled by a temporal asynchrony in the structural and
uli (Lin et al., 2012). In contrast, the age-matched TD functional development of two primary systems, the
participants showed the opposite pattern of behavioral motivational limbic systems (especially the amygdala and
results with better performance for the social-reward the striatum) and prefrontal control regions (e.g., Casey
stimuli than for the monetary-reward stimuli. Similarly, in et al., 2010; Ernst, Pine, & Hardin, 2006; Steinberg, 2010).
an implicit learning task executed in a functional MRI For example, according to the triadic model (Ernst et al.,
scanner with both social and monetary rewards, adoles- 2006), during early adolescence, there are impressive
cents with autism exhibited significantly less activation in changes in the dopaminergic system that differentially

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016
10 Picci, Scherf

influence the functional and structural development of There is only one study, to our knowledge, in which
the reward system (e.g., ventral striatum) when the harm- researchers have explicitly investigated risk-taking behav-
avoidant (e.g., amygdala) and cognitive control (e.g., iors in adolescents with autism. South, Dana, White, and
medial/ventral prefrontal cortex) systems are still rela- Crowley (2011) explored the relationship between age,
tively underdeveloped and weak. This imbalance argu- anxiety symptoms, and risky decisions in typical children
ably motivates the pursuit of rewards without much and adolescents (aged 8–18 years) and adolescents with
restraint but also allows adolescents to explore new autism in the context of the balloon analogue risk task. In
social contexts, roles, and environments that may other- this task, participants acquire money each time they
wise be frightening to them. pump a balloon without having it pop. It is important to
Much of the work on risk-taking behavior is conducted note that the balloon can explode at any time, thereby
in the context of decision-making (i.e., gambling) tasks causing participants to lose all their money. South and
that use money as a reward. In the Iowa Gambling Task colleagues reported that the TD adolescents and those
(IGT; Bechara, Damasio, Damasio, & Anderson, 1994), with autism exhibited similar levels of risk taking during
participants have to maximize monetary gain by choos- adolescence. However, there were important moderators
ing cards from multiple decks that vary with respect to that differentially influenced their risk behavior. Age was
their relative risk (e.g., high win/high loss, low win/low a moderator of risky decisions but only for TD individu-
loss). A large literature has shown that reward-seeking als. Thus, although the mean levels of risk were similar
behavior in this task increases from early to midadoles- across the two groups, the rate of risk increased over
cence, and risk aversion does not become a clear strategy time for the TD individuals. This suggests that compared
until late adolescence and early adulthood (Cauffman with TD individuals, individuals with autism may exhibit
et al., 2010; Steinberg, 2010). Also, neural activation in higher levels of risk as children but potentially lower lev-
reward-related regions, including the ventral striatum and els of risk taking as adolescents. Finally, as predicted, TD
the ventral medial prefrontal cortex, follow the same age- adolescents who had higher symptoms of anxiety were
related increase in activation (i.e., peaking in adoles- risk averse (i.e., made fewer risky decisions). However,
cence) during high-risk gambles compared with low-risk adolescents with autism who had higher levels of anxiety
gambles (Van Leijenhorst et al., 2010). This work collec- made riskier choices. In contrast to the predictions, they
tively suggests that adolescents evince uniquely strong were not risk averse. This complicated set of results is
reward-seeking behaviors and patterns of neural activa- difficult to interpret but suggests that adolescents with
tion, particularly in high-risk situations. autism exhibit comparable, and potentially lower, levels
There is a less clear understanding, across all develop- of risk-taking behavior to TD adolescents. Given this lim-
mental stages, of the profile of risk-taking behavior in ited set of findings, it is critical to understand the status
individuals with autism. For example, one study of ado- of the cognitive control system, which is largely mediated
lescents and adults with autism (mean age = 19.7 years) by the prefrontal cortex, in autism. If it is particularly
showed no differences in IGT performance between the compromised, individuals with autism may be especially
adolescents and adults with autism and TD individuals challenged in adolescence when they have to manage
with respect to choosing from the advantageous decks even slight increases in the motivation to engage in risky
(South et al., 2008). Another study reported mild risk behavior.
aversion for adults with autism in a slightly different ver-
sion of a gambling task (De Martino, Harrison, Knafo, Compromised cognitive control systems
Bird, & Dolan, 2008). There are even fewer studies
focused on risk-taking behavior specifically among ado-
in autism
lescents with autism. In an IGT study by Johnson, Cognitive control is central to goal-directed decision
Yechiam, Murphy, Queller, and Stout (2006), adolescents making in that it keeps impulses (e.g., to engage in
and adults with autism, compared with TD individuals, risky behavior) in check long enough to deliberate
did not show consistently different risk-taking behavior. alternate choices (Chein et al., 2011). This aspect of
It is interesting that the TD individuals tended to follow a cognitive control is empirically evaluated by asking par-
win-stay, loose-shift strategy, whereas the individuals ticipants to inhibit a prepotent behavioral response,
with autism were more likely to consistently shift across such as in a go/no-go or antisaccade task. In TD indi-
decks regardless of the outcome. These findings may viduals, there are marked improvements in cognitive
reflect differential sensitivity to rewards and losses or dif- control through adolescence (e.g., Somerville & Casey,
ferent learning strategies among the individuals with 2010). A large literature has suggested that cognitive
autism. They do not suggest that adolescents and adults control is relatively weakened in children and adoles-
with autism are more or less risk averse than are TD cents with autism (Geurts, van den Bergh, & Ruzzano,
individuals. 2014; for review, see O’Hearn, Asato, Ordaz, & Luna,

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016
Two-Hit Model of Autism 11

2008). For example, Minshew, Luna, and Sweeney begin to show some evidence of risk aversion in adoles-
(1999) found that children, adolescents, and adults with cence. It is important to note that even modest increases
autism made significantly more errors than did age- in risk-taking behavior in adolescence may be especially
matched TD control individuals in an antisaccade task, problematic for individuals with autism (i.e., function as
which requires participants to resist looking at a primed a second hit), given the consistent findings of compro-
visual target and, instead, to look in the opposite direc- mised cognitive control abilities. We propose that the
tion. Among individuals with autism, this deficit exists underlying neural and behavioral vulnerabilities in cogni-
in childhood and never improves as a function of age in tive control may lead to an even more robust adolescent-
adolescence and adulthood as it does among TD indi- specific imbalance between the approach/reward and
viduals. A recent meta-analysis of cognitive control in regulatory systems for individuals with autism.
children and adolescents with autism concluded that In addition, we predict that adolescents with autism
the ability to inhibit proponent responses is significantly may be less susceptible to the influence of peer contexts
impaired in autism but actually reduces across the life in the pursuit of risky behavior as a result of decreased
span (Geurts et al., 2014). sensitivity to social rewards that stems from early atypical-
Despite the large behavioral literature on cognitive ity in the development of mesocorticolimbic neural cir-
control deficits in children and adolescents with autism, cuitry. This potential insensitivity to the rewarding nature
there are only a handful of neuroimaging studies on of peers may be protective, thereby making adolescents
atypical neural activation during inhibition of prepotent with autism less susceptible to the kinds of risk taking that
responses in children and adolescents with autism. In lead to dire consequences for TD individuals (e.g., car
one study, participants had to inhibit a regularized map- accidents, drug use, binge drinking). Critically, future
ping between a visual stimulus and a button press (i.e., research is necessary to understand how findings about
same side of screen). The adolescents with autism (aged altered reward processing in autism relate to risk-taking
12–18 years) exhibited less activation within and func- behaviors in adolescents with autism. For example, if the
tional connectivity (i.e., temporal coordination) between peer context does not necessarily elicit stronger risk-­
a network of neural regions implicated in cognitive con- taking behavior in adolescents with autism, it is important
trol (Solomon et al., 2009). In a subsequent study using to understand whether other contexts do, such as those in
the same task, Solomon et al. (2013) specifically investi- which they can indulge restricted interests. For example,
gated age-related changes in the behavioral and neural video games are a common focus of restricted interests
basis of cognitive control from early (12–15 years) to later for adolescent boys with autism (Mazurek & Engelhardt,
(16–18 years) adolescence in individuals with autism and 2013). It could be that sensation- and reward-seeking
age-matched TD control individuals. Although neither behaviors in these boys during adolescence might be
group improved in accuracy with age, both groups excessively focused on video games at the expense of
improved in the efficiency of their performance (i.e., time spent pursuing peer social interactions.
faster reaction time to reach the same level of accuracy) Finally, we suggest that for adolescents with autism,
from early to later adolescence. However, even though any blunting of adolescent social risk taking likely
there were no behavioral differences between the two impedes the development of social-emotional competen-
groups, the adolescents with autism differentially relied cies that are necessary for pursuing autonomy and inde-
on a subset of regions in the cognitive control network pendence from parents, which compromises their ability
(i.e., anterior cingulate cortex and ventrolateral prefrontal to transition into adult levels of adaptive functioning.
cortex) compared with the TD adolescents (i.e., dorsolat-
eral prefrontal cortex and parietal cortex). Solomon and
colleagues suggested that these findings might reflect Summary of vulnerable domains for
strategic differences between the two groups with respect
to how they manage cognitive control, particularly dur-
adolescents with autism
ing cognitively demanding experiences. We have identified several domains of development (i.e.,
peer reorientation, risk taking/sensation seeking, cogni-
Hypotheses about risk taking and cognitive con- tive control) that are critical for accomplishing the devel-
trol in adolescents with autism. The existing find- opmental tasks of adolescence (e.g., forming intimate
ings provide no evidence that individuals with autism friendships and romantic relationships, gaining auton-
show consistently higher levels of risk-taking behavior omy from parents) and will ultimately facilitate the transi-
than do TD individuals. What remains to be seen is tion from adolescence to adult levels of adaptive
whether there are similar gradual increases in risk-taking functioning. In addition, we have critically reviewed the
behavior as a function of age that peak in adolescence existing (albeit scant) evidence that indicates that these
(as in TD individuals) or whether adolescents with autism domains of behavior are measurably disrupted as

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016
12 Picci, Scherf

individuals with autism transition through adolescence. includes an understanding of the contributing (or inter-
We summarize this evidence as follows: fering) role of the mesocorticolimbic reward circuitry to
this process. Second, we argue that the developmental
1. Adolescents with autism struggle in their attempts to trajectory of risk-taking behavior needs to be studied in
develop intimate peer relationships, which may be autism with an emphasis on investigating the focus and
an indication that adolescent-specific social reorien- contexts that drive app­ roach and reward motivations
tation is significantly weaker in this population. (e.g., objects of restricted interests, peers). This method
2. Peer rejection is disproportionately higher for indi- will need to be complemented by a parallel investigation
viduals with autism in adolescence. This may lead of whether there is a disproportionate imbalance
to risk for developmental deterioration and increas- between approach and regulatory neural systems in
ing depression and anxiety during adolescence. autism, particularly during adolescence. Finally, we sug-
3. Compared with TD individuals, adolescents with gest that the role of pubertal development must be eval-
autism appear to exhibit similar, or slightly reduced, uated in the context of these other areas of research
levels of risk-taking behavior. However, nothing is given the impressive pubertal deterioration that has been
known about differences in the context in which reported in the literature (Billstedt et al., 2005; Gillberg
risk-taking behaviors occur for adolescents with & Steffenburg, 1987) and the reports of the influence of
autism (i.e., social contexts vs. contexts associated pubertal maturation on many of these developing sys-
with repetitive and restricted interests). tems in TD adolescents (Forbes & Dahl, 2010; Moore
4. The ability to inhibit a prepotent response is par- et al., 2012; Pfeifer et al., 2013).
ticularly difficult for individuals with autism and
may be especially true in adolescence. There are
also weaknesses in the contribution of the pre- Investigating the role of peers in
frontal cortex to cognitive control, which may adolescent development
contribute to a robust adolescent-specific imbal- We propose that early neural atypicalities contribute to
ance between the approach/reward and regula- poor childhood social outcomes that cascade to interfere
tory systems for individuals with autism. with peer reorientation in adolescents with autism. We
also suggest that increasing levels of restricted interests
In sum, we argue that the limited evidence reviewed that intensify during adolescence among individuals with
herein is consistent with our prediction that adolescence autism may interfere with motivation to form peer relation-
functions as a second hit to the already vulnerable neural ships. The resulting consequence is that the ability to pur-
system of individuals with autism; however, we also rec- sue friendships as well as romantic and sexual relationships
ognize that future research must be conducted to deter- is significantly compromised, which, in turn, leads to poor
mine (a) the extent and nature of the difficulties social competence in adulthood. These hypotheses can be
individuals with autism exhibit in accomplishing the evaluated by testing several empirical questions.
developmental tasks of adolescence, (b) a potential link First, to what extent do adolescents with autism reori-
between these difficulties and early alterations in the ent toward peers? We suggest that future research
development of neural systems and their functional and addressing this question should focus on understanding
structural organization, and (c) the relation between whether there is increasing (a) complexity of peer rela-
behavior in adolescence and the failure to take on adult tionships, (b) reliance on peers for social support,
social roles and reach adaptive levels of functioning dur- (c) quality of peer interactions, (d) emotional responses
ing adulthood. to social contexts involving peers, (e) drive for accep-
tance from peers, (f) sensitivity to peer evaluation, (g)
Future Directions for Studying neural reward-related responses to peers, and (h) moti-
Adolescence as a Second Hit for vation to engage in romantic or sexual relationships with
peers. These are all hallmarks of peer reorientation
Autism among TD adolescents.
In this final section, we recommend three areas of future We also recommend that future studies on adolescent
research that must proceed in order to understand changes in the nature of peer interactions and complex-
whether and how adolescence functions as a second hit ity of friendships go beyond using parent and adolescent
(i.e., a period of important vulnerability) to the develop- survey measures. Adolescents with autism and their par-
mental course of autism. First, we emphasize the need to ents appear to have fairly different notions of friendship
investigate whether and to what extent there is peer (Orsmond et al., 2004), which may challenge the validity
reorientation among adolescents with autism. This of such a measurement for use in autism. We suggest that

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016
Two-Hit Model of Autism 13

researchers employ more ecologically valid observational With regard to research on adolescent-specific devel-
methods of social interactions (e.g., social problem solv- opmental changes in the drive for acceptance from peers
ing) between peer dyads (i.e., an adolescent with autism and sensitivity to peer evaluation, we suggest that inves-
and a friend identified by the adolescent). It would also tigators consider using the Chatroom task (Guyer, Choate,
be useful to compare the nature of social interactions in Pine, & Nelson, 2012), which was specifically designed to
peer dyads with those elicited in parent-adolescent dyads simulate ecologically valid adolescent social interactions
(i.e., an adolescent with autism and his or her parent). (Bolling et al., 2011). In this task, participants first rank
In terms of evaluating the potential reward value of photographs of peers for social interest (i.e., how inter-
peer contexts, we recommend that this work proceed ested they are in interacting with the person) and are
using very basic tasks. For example, one existing func- made to believe that they will also be ranked by each of
tional MRI study has reported that children and adoles- the peers. In the return visit, participants are scanned
cents with autism exhibited levels of neural activation while they view each photograph that they rated previ-
similar to TD individuals during passive viewing of famil- ously and are given “feedback” about whether the peer
iar adult and same-aged unfamiliar peer faces (Pierce & was mutually interested. Then participants have to judge
Redcay, 2008). This particular study was focused on eval- how they feel after receiving the feedback. Thus, the par-
uating the neural signal in posterior visuoperceptual ticipant experiences both acceptance and rejection by
regions, such as the fusiform gyrus. However, a similar the virtual peers throughout the course of the experi-
approach could be used to investigate the reward-related ment. Behavioral and neural responses collected during
response to familiar and unfamiliar peer and adult faces the Chatroom task could help determine the relative con-
in the reward circuitry. This would help evaluate whether cern (or lack thereof) about peer acceptance and rejec-
peer faces are more rewarding to adolescents with autism tion among adolescents with autism. Furthermore, an
than are adult faces or whether familiarity modulates assessment of individual differences in the response to
such an effect. If adolescents with autism consistently peer rejection in early adolescence could be used to pre-
show stronger reward-related activation to familiar faces, dict longitudinal changes in the magnitude of symptom
regardless of age, than to peer faces regardless of famil- severity and depression symptoms in later adolescence
iarity, the notion that peers may not take on the same and early adulthood. This approach would provide foun-
reward value for adolescents with autism as for TD ado- dational information about peer reorientation and sensi-
lescents would be supported. tivity to peer rejection in autism and may also provide a
In addition, studies on more complex social process- clearer link between a failure to accomplish developmen-
ing of peer and adult faces, such as judging trustworthi- tal tasks of adolescence and worsening outcomes for
ness or attractiveness traits, could help determine the young adults with autism.
extent to which adolescents with autism are dispropor- For TD individuals, the formation of romantic relation-
tionately socially motivated to process information about ships during adolescence is an important antecedent to
peers (e.g., higher ratings overall for trustworthiness and successful romantic relationships in adulthood (i.e., num-
attractiveness for peer as compared with adult faces). A ber of romantic partners, duration of romantic relation-
set of comparison studies that involve familiar and unfa- ships; Rauer, Pettit, Lansford, Bates, & Dodge, 2013;
miliar objects that are and are not the focus of each par- Simpson, Collins, & Salvatore, 2011). Although there is an
ticipant’s own restricted interest would help investigate extremely sparse literature regarding romantic and sexual
the relative calibration of reward systems among adoles- partnerships in adolescents with autism, findings in adults
cents with autism (e.g., restricted interest objects > com- with autism suggest that outcomes are rather bleak.
mon objects > familiar faces). It also is critical to Recall that approximately 44% of adults with autism have
investigate these same responses in children and adults never dated (Farley et al., 2009). This does not appear to
with autism to examine whether there are adolescent- be due to a lack of sexual well-being or interest; adults
specific changes in these responses. This kind of approach with autism do report sexual desire as well as arousabil-
would enable researchers to assess the extent to which ity (Byers et al., 2013), and many express an interest in
developmental trajectories diverge between TD individu- marriage and in having intimate and sexual relationships
als and those with autism in terms of social-information (Haracopos & Pedersen, 1992; Newport, Newport, &
processing about peers. Finally, it is important that future Bolick, 2002; Ousley & Mesibov, 1991). However, their
research assess whether individual differences in the knowledge of sexual behavior and romantic relationships
magnitude of these reward-related responses to peer is usually limited (Ousley & Mesibov, 1991). These find-
faces and objects of restricted interests are at all related ings suggest that individuals with autism appear to be
to the abilities to form peer friendships during adoles- interested in pursuing romantic and sexual relationships
cence and to longer-term adult outcomes of social but do not have the social skills to be able to do so. This
competence. interpretation is confirmed by findings that have shown

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016
14 Picci, Scherf

that adults and adolescents with autism engage in inap- On the basis of the existing literature, we suggest that
propriate courting behaviors (e.g., stalking behaviors; there are four empirical questions that need to be
Stokes et al., 2007) and sexual behaviors (e.g., touching addressed: (a) What is the developmental trajectory of
private body areas in public, removing clothing in public, risk-taking/sensation-seeking behavior in autism? (b) Is
masturbating in public, and touching members of the there emerging risk aversion? (c) If so, does this influence
opposite sex inappropriately; Ruble & Dalrymple, 1993). the ability to gain independence from parents and pursue
Given that adolescence is the time when sexual behav- more adult social roles? and (d) What is the relative influ-
iors are learned and practiced among TD individuals ence of peers and other nonpeer contexts (e.g., involving
(Feldmann & Middleman, 2002), and that parents of ado- restricted interests) on risk-taking behaviors?
lescents with autism report that their child’s sexuality is a To address the first two questions, researchers should
major concern, it is imperative that future work evaluate use cross-sectional and longitudinal designs to investi-
the emerging nature of romantic and sexual relationships gate the trajectory of risk-taking behaviors from child-
among adolescents with autism. hood through early adulthood using established parent
First, it will be important to understand whether ado- and adolescent questionnaires and interviews, such as
lescents with autism are similarly motivated to pursue the Barratt Impulsiveness Scale (Patton, Stanford, &
romantic and sexual relationships and whether they have Barratt, 1995), Benthin Risk Perception Measure (Benthin,
the social skills to pursue such relationships. Part of this Slovic, & Severson, 1993), Youth Decision-Making Ques­
work needs to assess the frequency with which adoles- tionnaire (Ford, Wentzel, Wood, Stevens, & Siesfeld, 1989),
cents with autism are engaging in (or attempting to and Zuckerman Sensation-Seeking Scale (Zuckerman,
engage in) romantic relationships. Such work likely needs Eysenck, & Eysenck, 1978). These existing questionnaires
to be done using interviewing/survey methods, such as may need to be modified to dissociate risk taking in and
the Sexual Behavior Scale (Stokes & Kaur, 2005) and the out of peer contexts for use with adolescents with autism.
Dating Questionnaire (Connolly, Craig, Goldberg, & In addition, it is essential to evaluate the existence of
Pepler, 2004), and may require parental input as well. It risk-taking and risk-aversion behaviors depending on
is also critical for future work to understand the develop- whether the context is social (i.e., with peers) or nonso-
mental stages in the formation of romantic relationships cial (i.e., neutral, monetary, or focused on restricted inter-
in adolescents with autism. That is, to the extent that ests). Future researchers should also aim to evaluate the
these relationships exist, do they follow a similar devel- potential link between alterations in reward-seeking
opmental progression as in TD adolescent romantic rela- behaviors among adolescents and subsequent autonomy
tionships (i.e., mixed-gender group affiliations, then in later adulthood. In other words, there may be indi-
group dates, and eventually formation of romantic part- vidual differences in the degree to which individuals with
nerships; Connolly, Nguyen, Pepler, Craig, & Jiang, 2013)? autism engage in sensation-seeking and risk-taking
This approach may reveal critical information about the behaviors that are systematically related to the likelihood
specific aspects of forming romantic and sexual partner- that they accomplish some autonomy from parents (e.g.,
ships in adolescence that are particularly difficult for ado- moving to a group home, going to college, obtaining a
lescents with autism. As a result, it could also lead to job).
mechanistic explanations about why adolescents with To address the willingness of adolescents with autism
autism overwhelmingly fail to meet this core develop- to take risks to gain rewards, researchers should use new
mental task that is so critical for later adult relationships. or modified experimental paradigms. For example, we
recommend that researchers use simple decision-making
Investigating risk taking among tasks (such as the IGT) that manipulate peer presence
(i.e., in the room or nearby when the adolescent with
adolescents with autism autism is participating in the task) or peer-related stimuli
We suggest that compared with TD adolescents, adoles- (i.e., pictures of peers or peer contexts) and stimuli of
cents with autism may have a disproportionate imbalance restricted individualized interests. These tasks (e.g., IGT)
between their cognitive control and reward-processing are traditionally nonsocial in nature and likely dispropor-
systems. If this is the case, risk taking is likely to be a tionately motivate individuals who value monetary
considerable vulnerability for individuals with autism, rewards. Thus, paradigms that target adolescents with
especially during adolescence. Establishing the develop- autism may need to be tailored to the restricted interests
mental trajectory and contextual focus of risk-taking of each participant (i.e., one individual may have cars as
behavior in autism is essential to understanding whether a restricted interest, whereas another individual may have
and to what degree there is a disproportionate imbalance video games as a restricted interest). Therefore, research-
between the reward and regulatory systems in autism ers may want to contact parents before administering the
during adolescence. task to determine the stimuli that are most rewarding for

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016
Two-Hit Model of Autism 15

the adolescents with autism. The decision-making task influences behavioral and brain development in autism.
itself can be simple (e.g., guessing whether the value of For example, given the reports of pubertal deterioration
a card is higher or lower than 5; Delgado, Nystrom, (Billstedt et al., 2005; Gillberg & Steffenburg, 1987),
Fissell, Noll, & Fiez, 2000) and depend on rewards that pubertal development may uniquely complicate develop-
are focused on the adolescents’ restricted interests. By ment in these important adolescent domains. Even minor
evaluating the willingness of adolescents with autism to alterations to the timing or tempo of pubertal develop-
take risks with rewards focused on peers versus on ment may have significant consequences for emerging
restricted interests, researchers can better understand the and fine-tuning all of the domains of adolescent develop-
risk-taking behavioral patterns and trajectories across ment explored in this article.
contexts. This approach can dissociate whether, in ado- We suggest that in future studies, researchers must
lescence, the reward-processing system in autism is cali- begin to systematically investigate the influence of puber-
brated differently but risk-taking behavior is comparable tal development on adolescent-specific deterioration in
or whether both reward processing and risk taking are autism, including how it does or does not influence
importantly altered. behavior in the domains we have outlined as well as the
With respect to evaluating the degree to which cogni- structural and functional organization in the developing
tive control is impaired using peer-based versus non- brain. Recent evidence regarding human adolescence
peer-based stimuli, a simple go/no-go task is an effective provides support for the mountain of evidence obtained
way to assess both contexts. For instance, one could test from animal models (see Sisk & Zehr, 2005), which indi-
same-age peer faces versus adult faces, familiar peer cates that pubertal hormones sculpt both functional and
faces versus unfamiliar peer faces, peer faces versus structural organization of neural circuitry (Barnea-Goraly
objects that are the focus of restricted interests, and neu- et al., 2004; Burnett & Blakemore, 2009; Gogtay et al.,
tral objects versus objects that are the focus of restricted 2004; Ostby et al., 2009; Schmithorst & Yuan, 2010;
interests. This approach can be used to target both the Tamnes et al., 2010). With these findings in mind, and
behavioral and the neural basis of the ability to inhibit a using our two-hit model, we predict that pubertal hor-
prepotent response despite differing contexts that may mones contribute in important ways to disruptions in
elicit more or less reward-related and risk-­ taking adolescent-specific neural organization and behavioral
responses in adolescents with autism. Such research that outcomes in autism. We suggest that the influx of puber-
directly focuses on potential differences in the reward- tal hormones is forced to shape a pervasively disrupted
seeking and risk-taking behaviors of TD adolescents and neural system in individuals with autism. As a result,
adolescents with autism could inform m ­ eaningful forms puberty-related influences on brain and behavioral devel-
of intervention (e.g., sexual-health education, other opment are expected to be markedly different in indi-
social adaptive functioning skills). viduals with autism compared with TD individuals.
Furthermore, these differences are likely to contribute to
Pubertal development in adolescents the failure to develop new behaviors that are essential for
acquiring adult levels of adaptive functioning (e.g., inde-
with autism pendence from parents, intimate friendships, romantic
An area of research that has yet to be explicitly explored and sexual partnerships). Future research on the relation-
in adolescents with autism is the role of pubertal devel- ship between pubertal development and pubertal hor-
opment. Pubertal development is known to influence mones, neural functioning and organization, and behavior
behavior (Feinberg & Campbell, 2010; Halpern, 2006; in adolescents with autism will provide critical evidence
Marceau, Dorn, & Susman, 2012; Martin et al., 2002; to evaluate this hypothesis.
Quevedo, Benning, Gunnar, & Dahl, 2009; Varlinskaya,
Vetter-O’Hagen, & Spear, 2013; Wallen, 2001; Waylen &
Conclusions
Wolke, 2004) and neural activation and organization
(Blakemore, Burnett, & Dahl, 2010; Bramen et al., 2011; We have proposed a novel theoretical framework for
Forbes & Dahl, 2010; Forbes, Phillips, Silk, Ryan, & Dahl, conceptualizing the full developmental trajectory of
2011; Goddings, Burnett Heyes, Bird, Viner, & Blakemore, autism and have attempted to connect the most cutting-
2012; Herting, Maxwell, Irvine, & Nagel, 2012; Holm edge findings about atypicalities in early brain develop-
et al., 2009; Lenroot & Giedd, 2010; Romeo, 2003; Schulz, ment, adolescent-specific deterioration in functioning,
Molenda-Figueira, & Sisk, 2009; Sisk & Zehr, 2005; Spear, and poor levels of adaptive functioning in adulthood. It
2011) and is hypothesized to motivate and enable new is important to note that our framework is particularly
behaviors in TD adolescents (Dahl & Spear 2004; Forbes sensitive to the nonlinear atypicalities in the develop-
& Dahl, 2010; Scherf et al., 2013; Scherf & Scott, 2012). mental course of autism. We focus on the role of adoles-
There are no data that address whether or how puberty cence as a developmental period of particular vulnerability

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016
16 Picci, Scherf

that builds on the early first hit for individuals with autism interests. This ultimately may be a tool for targeting such
and uniquely contributes to atypical trajectories toward individuals for interventions designed to enhance the
adult outcomes. It is clear that adolescence is exception- reward value of peers. The creation of interventions that
ally difficult for individuals with autism, especially in involve hobbies or relate to restricted interests with same-
comparison to individuals with other developmental dis- age peers could encourage peer interaction and possibly
orders who fair reasonably well in taking on adult social enhance reward derived from peer interactions. We
roles and levels of adaptive functioning. For instance, believe that this two-hit model of the developmental
compared with adults with autism, adults with Down’s course of autism provides a framework that will guide
syndrome achieve far more independence in their resi- both basic and translational research about the central
dential living status and social contact with friends factors that uniquely interfere with the ability to achieve
(Esbensen, et al., 2010). adult levels of adaptive functioning in autism.
We provide evidence that suggests that the core
developmental processes that support the acquisition of Author Contributions
adult levels of social and adaptive functioning putatively G. Picci and K. S. Scherf jointly conceived and drafted the man-
fail to develop in autism, specifically during adolescence. uscript. Both authors approved the final version of the manu-
We appreciate that there are important individual differ- script for submission.
ences among individuals with autism in terms of adult-
hood outcomes and suggest that this variation can be Acknowledgments
used to evaluate some of the hypothesized trajectories.
We would like to thank Elisabeth Whyte and Sara Barth for
For instance, individuals with autism who evince more helpful comments on earlier versions of this manuscript.
typical levels of social-reward processing may be capa-
ble of forming higher quality peer relationships in ado-
Declaration of Conflicting Interests
lescence that lead to subsequent higher quality adult
relationships. Also, we suggest that there may be indi- The authors declared that they had no conflicts of interest
with respect to their authorship or the publication of this
vidual differences with respect to how well this model
article.
characterizes the development of any one individual
with autism and that there may be important moderating
factors that mitigate or exacerbate the effects of the Funding
­second hit of adolescence. This work was supported by Pennsylvania Department of
Specifically, individual differences in the manifestation Health SAP Grant 4100047862 to K. S. Scherf (co-principal
of the first hit on atypical neural network organization investigator), the Social Science Research Institute, and the
Department of Psychology at Pennsylvania State University, as
will likely moderate the extent to which the second hit
well as a Graduate Research Fellowship from the National
overwhelms the developing system in the face of new Science Foundation to G. Picci (DGE1255832). Any opinions,
demands during adolescence. For example, individuals findings, and conclusions or recommendations expressed in
with autism who show disproportionately early (i.e., pre- this material are those of the authors and do not necessarily
natal and infant development) and severe disruptions to reflect the views of the National Science Foundation.
the behaviors or neural networks that ultimately support
cognitive control might exhibit more difficulty with indis- Note
criminate risk taking (and lack of cognitive control) in
1. This is not to say that peers become the sole support system
adolescence, which could make them especially vulner-
for adolescents; parents maintain a critical presence in adoles-
able to more harmful consequences of risk taking. cents’ lives (see Lam, McHale, & Crouter, 2012).
Uncovering these potential outcomes of the first hit could
reveal important targets for intervention that could mod-
References
erate (or even prevent) subsequent consequences as
individuals transition into adolescence. Abrahams, B. S., & Geschwind, D. H. (2010). Connecting
Our goal is to provide a model that generally charac- genes to brain in the autism spectrum disorders. Archives
terizes the full developmental trajectory of autism from of Neurology, 67, 395–399. doi:10.1001/archneurol.2010.47
Anderson, D., Lord, C., Risi, S., DiLavore, P., Shulman, C.,
early infancy to adulthood, that motivates basic research
Thurm, A., . . . Pickles, A. (2007). Patterns of growth in ver-
on this trajectory, and that will ultimately inform treat- bal abilities among children with autism spectrum disorder.
ment and intervention approaches. For example, the go/ Journal of Consulting and Clinical Psychology, 75, 594–604.
no-go paradigm that we outlined to evaluate cognitive Anderson, D., Mayes, M., & Lord, C. (2011). Changes in mal-
control and approach tendencies may help identify ado- adaptive behaviors from midchildhood to young adult-
lescents with autism who exhibit enhanced approach hood in autism spectrum disorder. American Journal on
behaviors for objects that are the focus of their restricted Intellectual and Developmental Disabilities, 116, 381–397.

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016
Two-Hit Model of Autism 17

Anderson, D., Oti, R., Lord, C., & Welch, K. (2009). Patterns Brown, B. (2004). Adolescents’ relationships with peers. In R.
of growth in adaptive social abilities among children with Lerner & L. Steinberg (Eds.), Handbook of adolescent psy-
autism spectrum disorders. Journal of Abnormal Child chology (pp. 363–394). Hoboken, NJ: John Wiley & Sons.
Psychology, 37, 1019–1034. Brown, B., Eicher, S., & Petrie, S. (1986). The importance of
Assaf, M., Jagannathan, K., Calhoun, V. D., Miller, L., Stevens, peer group (“crowd”) affiliation in adolescence. Journal of
M. C., Sahl, R., . . . Pearlson, G. D. (2010). Abnormal func- Adolescence, 86, 73–96.
tional connectivity of default mode sub-networks in autism Brown, J. (1969). Adolescent development of children with
spectrum disorder patients. NeuroImage, 53, 247–256. infantile psychosis. Seminars in Psychiatry, 1, 79–89.
doi:10.1016/j.neuroimage.2010.05.067 Burnett, S., & Blakemore, S. J. (2009). Functional connectiv-
Bagwell, C., Newcomb, A., & Bukowski, W. (1998). ity during a social emotion task in adolescents and in
Preadolescent friendship and peer rejection as predictors adults. European Journal of Neuroscience, 29, 1294–1301.
of adult adjustment. Child Development, 69, 140–153. doi:10.1111/j.1460-9568.2009.06674.x
Barnea-Goraly, N., Kwon, H., Menon, V., Eliez, S., Lotspeich, Byers, E. S., Nichols, S., Voyer, S. D., & Reilly, G. (2013). Sexual
L., & Reiss, A. L. (2004). White matter structure in autism: well-being of a community sample of high-functioning
Preliminary evidence from diffusion tensor imaging. adults on the autism spectrum who have been in a romantic
Biological Psychiatry, 55, 323–326. doi:10.1016/j.bio- relationship. Autism: The International Journal of Research
psych.2003.10.022 and Practice, 17, 418–433. doi:10.1177/1362361311431950
Bauminger, N., & Shulman, C. (2003). The development and Cappadocia, M. C., Weiss, J. A., & Pepler, D. (2012). Bullying
maintenance of friendship in high-functioning children experiences among children and youth with autism spec-
with autism. Autism: The International Journal of Research trum disorders. Journal of Autism and Developmental
and Practice, 7, 81–97. Disorders, 42, 266–277. doi:10.1007/s10803-011-1241-x
Bechara, A., Damasio, A., Damasio, H., & Anderson, S. (1994). Caron, M. J., Mottron, L., Berthiaume, C., & Dawson, M. (2006).
Insensitivity to future consequences following damage to Cognitive mechanisms, specificity and neural underpin-
human prefrontal cortex. Cognition, 50, 7–15. nings of visuospatial peaks in autism. Brain, 129, 1789–
Benayed, R., Gharani, N., Rossman, I., Mancuso, V., Lazar, 1802. doi:10.1093/brain/awl072
G., Kamdar, S., . . . Zimmerman, R. A. (2005). Support Carrington, S., Templeton, E., & Papinczak, T. (2003).
for the homeobox transcription factor gene ENGRAILED Adolescents with Asperger syndrome and perceptions of
2 as an autism spectrum disorder susceptibility locus. friendship. Focus on Autism and Other Developmental
American Journal of Human Genetics, 77, 851–868. Disabilities, 18, 211–218. doi:10.1177/10883576030180040
doi:10.1086/497705 201
Benthin, A., Slovic, P., & Severson, H. (1993). A psychometric Cascio, C. J., Foss-Feig, J. H., Heacock, J. L., Newsom, C. R.,
study of adolescent risk perception. Journal of Adolescence, Cowan, R. L., Benningfield, M. M., . . . Cao, A. (2012).
16, 153–168. Response of neural reward regions to food cues in autism
Billstedt, E., Gillberg, C., & Gillberg, C. (2005). Autism after spectrum disorders. Journal of Neurodevelopmental
adolescence: Population-based 13- to 22-year follow-up Disorders, 4, 9–21. doi:10.1186/1866-1955-4-9
study of 120 individuals with autism diagnosed in child- Cascio, C. J., Foss-Feig, J. H., Heacock, J., Schauder, K. B.,
hood. Journal of Autism and Developmental Disorders, 35, Loring, W. A., Rogers, B. P., . . . Bolton, S. (2014). Affective
351–360. doi:10.1007/s10803-005-3302-5 neural response to restricted interests in autism spectrum
Blakemore, S. J. (2008). The social brain in adolescence. Nature disorders. Journal of Child Psychology and Psychiatry, 55,
Reviews Neuroscience, 9, 267–277. doi:10.1038/nrn2353 162–171. doi:10.1111/jcpp.12147
Blakemore, S., Burnett, S., & Dahl, R. E. (2010). The role of Casey, B. J., Jones, R. M., Levita, L., Libby, V., Pattwell, S. S.,
puberty in the developing adolescent brain. Human Brain Ruberry, E. J., . . . Somerville, L. H. (2010). The storm and
Mapping, 31, 926–933. doi:10.1002/hbm.21052 stress of adolescence: Insights from human imaging and
Bolling, D. Z., Pitskel, N. B., Deen, B., Crowley, M. J., McPartland, mouse genetics. Developmental Psychobiology, 52, 225–
J. C., Kaiser, M. D., . . . Pelphrey, K. A. (2011). Enhanced 235. doi:10.1002/dev.20447
neural responses to rule violation in children with autism: Cauffman, E., Shulman, E. P., Steinberg, L., Claus, E., Banich,
A comparison to social exclusion. Developmental Cognitive M. T., Graham, S., & Woolard, J. (2010). Age differences in
Neuroscience, 1, 280–294. doi:10.1016/j.dcn.2011.02.002 affective decision making as indexed by performance on
Bramen, J., Hranilovich, J., Dahl, R., Forbes, E., Chen, J., Toga, the Iowa Gambling Task. Developmental Psychology, 46,
A., . . . Sowell, E. (2011). Puberty influences medial tem- 193–207. doi:10.1037/a0016128
poral lobe and cortical gray matter maturation differently Chein, J., Albert, D., O’Brien, L., Uckert, K., & Steinberg, L.
in boys than girls matched for sexual maturity. Cerebral (2011). Peers increase adolescent risk taking by enhanc-
Cortex, 21, 636–646. doi:10.1093/cercor/bhq137 ing activity in the brain’s reward circuitry. Developmental
Brereton, A. V., Tonge, B. J., & Einfeld, S. L. (2006). Science, 14, F1–F10. doi:10.1111/j.1467-7687.2010.01035.x
Psychopathology in children and adolescents with autism Coe, B. P., Girirajan, S., & Eichler, E. E. (2012). The genetic vari-
compared to young people with intellectual disability. ability and commonality of neurodevelopmental disease.
Journal of Autism and Developmental Disorders, 36, 863– American Journal of Medical Genetics Part C: Seminars in
870. doi:10.1007/s10803-006-0125-y Medical Genetics, 160C, 118–129. doi:10.1002/ajmg.c.31327

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016
18 Picci, Scherf

Collins, W. A., Welsh, D. P., & Furman, W. (2009). Adolescent Down ­ syndrome in adulthood. American Journal on
romantic relationships. Annual Review of Psychology, 60, Intellectual and Developmental Disabilities, 115, 277–290.
631–652. doi:10.1146/annurev.psych.60.110707.163459 doi:10.1352/1944-7558-115.4.277
Connolly, J., Craig, W., Goldberg, A., & Pepler, D. (2004). Farley, M. A., McMahon, W. M., Fombonne, E., Jenson, W. R.,
Mixed-gender groups, dating, and romantic relationships Miller, J., Gardner, M., . . . Coon, H. (2009). Twenty-year
in early adolescence. Journal of Research on Adolescence, outcome for individuals with autism and average or near-
14, 185–207. average cognitive abilities. Autism Research, 2, 109–118.
Connolly, J., Nguyen, H. N., Pepler, D., Craig, W., & Jiang, D. doi:10.1002/aur.69
(2013). Developmental trajectories of romantic stages and Feinberg, I., & Campbell, I. G. (2010). Sleep EEG changes dur-
associations with problem behaviours during adolescence. ing adolescence: An index of a fundamental brain reor-
Journal of Adolescence, 36, 1013–1024. doi:10.1016/j ganization. Brain and Cognition, 72, 56–65. doi:10.1016/j
.adolescence.2013.08.006 .bandc.2009.09.008
Courchesne, E., Campbell, K., & Solso, S. (2011). Brain growth Feldmann, J., & Middleman, A. B. (2002). Adolescent sexual-
across the life span in autism: Age-specific changes in ana- ity and sexual behavior. Current Opinion in Obstetrics and
tomical pathology. Brain Research, 1380, 138–145. Gynecology, 14, 489–493.
Courchesne, E., Pierce, K., Schumann, C., Redcay, E., Forbes, E. E., & Dahl, R. E. (2010). Pubertal development and
Buckwalter, J., Kennedy, D., & Morgan, J. (2007). Mapping behavior: Hormonal activation of social and motivational
early brain development in autism. Neuron, 56, 399–413. tendencies. Brain and Cognition, 72, 66–72. doi:10.1016/j
Courchesne, E., Webb, S. J., & Schumann, C. M. (2011). From .bandc.2009.10.007
toddlers to adults: The changing landscape of the brain in Forbes, E. E., Phillips, M. L., Silk, J. S., Ryan, N. D., & Dahl, R. E.
autism. In D. G. Amaral, G. Dawson, & D. H. Geschwind (2011). Neural systems of threat processing in adolescents:
(Eds.), Autism spectrum disorders (pp. 611–631). New Role of pubertal maturation and relation to measures of
York, NY: Oxford University Press. negative affect. Developmental Neuropsychology, 36, 429–
Dahl, R. (1996). The impact of inadequate sleep on chil- 452. doi:10.1080/87565641.2010.550178
dren’s daytime cognitive function. Seminars in Pediatric Ford, M. E., Wentzel, K. R., Wood, D., Stevens, E., & Siesfeld, G.
Neurology, 3, 44–50. A. (1989). Processes associated with integrative social com-
Dahl, R. E., & Spear, L. P. (2004). Adolescent brain develop- petence: Emotional and contextual influences on adoles-
ment: Vulnerabilities and opportunities. Annals of the New cent social responsibility. Journal of Adolescent Research,
York Academy of Sciences, 1021, 1–22. 4, 405–425.
Delgado, M. R., Nystrom, L. E., Fissell, C., Noll, D. C., & Fiez, J. Fredriksen, K., Rhodes, J., Reddy, R., & Way, N. (2004). Sleepless
A. (2000). Tracking the hemodynamic responses to reward in Chicago: Tracking the effects of adolescent sleep loss
and punishment in the striatum. Journal of Neurophysiology, during the middle school years. Child Development, 75,
84, 3072–3077. 84–95.
Delmonte, S., Balsters, J. H., McGrath, J., Fitzgerald, J., Brennan, Galvan, A. (2010). Adolescent development of the reward sys-
S., Fagan, A. J., & Gallagher, L. (2012). Social and monetary tem. Frontiers in Human Neuroscience, 4, 6. Retrieved
reward processing in autism spectrum disorders. Molecular from http://journal.frontiersin.org/Journal/10.3389/neuro
Autism, 3, 7. Retrieved from http://www.molecularautism .09.006.2010/full
.com/content/3/1/7 Gardner, M., & Steinberg, L. (2005). Peer influence on risk taking,
De Martino, B., Harrison, N. A., Knafo, S., Bird, G., & Dolan, risk preference, and risky decision making in adolescence
R. J. (2008). Explaining enhanced logical consistency dur- and adulthood: An experimental study. Developmental
ing decision making in autism. Journal of Neuroscience, 28, Psychology, 41, 625–635. doi:10.1037/0012-1649.41.4.625
10746–10750. doi:10.1523/JNEUROSCI.2895-08.2008 Geurts, H. M., van den Bergh, S. F., & Ruzzano, L. (2014).
Dichter, G. S., Felder, J. N., Green, S. R., Rittenberg, A. M., Prepotent response inhibition and interference control in
Sasson, N. J., & Bodfish, J. W. (2012). Reward circuitry func- autism spectrum disorders: Two meta-analyses. Autism
tion in autism spectrum disorders. Social Cognitive and Research. Advance online publication. doi:10.1002/aur.1369
Affective Neuroscience, 7, 160–172. Ghaziuddin, M., Ghaziuddin, N., & Greden, J. (2002). Depression
Eaves, L. C., & Ho, H. H. (2008). Young adult outcome of autism in persons with autism: Implication for research and clinical
spectrum disorders. Journal of Autism and Developmental care. Journal of Autism and Developmental Disorders, 32,
Disorders, 38, 739–747. doi:10.1007/s10803-007-0441-x 299–306.
Eisenberg, L. (1956). The autistic child in adolescence. American Giedd, J., Clasen, L., Lenroot, R., Greenstein, D., Wallace, G.,
Journal of Psychiatry, 112, 607–612. Ordaz, S., . . . Stayer, C. (2006). Puberty-related influ-
Ernst, M., Pine, D. S., & Hardin, M. (2006). Triadic model ences on brain development. Molecular and Cellular
of the neurobiology of motivated behavior in adoles- Endocrinology, 254, 154–162.
cence. Psychological Medicine, 36, 299–312. doi:10.1017/ Gillberg, C., & Schaumann, H. (1982). Infantile autism and
S0033291705005891 puberty. Journal of Autism and Developmental Disorders,
Esbensen, A. J., Bishop, S., Seltzer, M. M., Greenberg, J. S., 11, 365–371.
& Taylor, J. L. (2010). Comparisons between ­individuals Gillberg, C., & Steffenburg, S. (1987). Outcome and prog-
with autism spectrum disorders and individuals with nostic factors in infantile autism and similar conditions:

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016
Two-Hit Model of Autism 19

A p­ opulation-based study of 46 cases followed through Holm, S. M., Forbes, E. E., Ryan, N. D., Phillips, M. L., Tarr,
puberty. Journal of Autism and Developmental Disorders, J. A., & Dahl, R. E. (2009). Reward-related brain function
17, 273–287. and sleep in pre/early pubertal and mid/late pubertal ado-
Goddings, A.-L., Burnett Heyes, S., Bird, G., Viner, R. M., & lescents. Journal of Adolescent Health, 45, 326–334.
Blakemore, S.-J. (2012). The relationship between puberty Howlin, P., Moss, P., Savage, S., & Rutter, M. (2013). Social out-
and social emotion processing. Developmental Science, 15, comes in mid- to later adulthood among individuals diag-
108–111. doi:10.1111/j.1467-7687.2012.01174.x nosed with autism and average nonverbal IQ as children.
Gogtay, N., Giedd, J. N., Lusk, L., Hayashi, K. M., Greenstein, Journal of the American Academy of Child and Adolescent
D., Vaituzis, A. C., . . . Thompson, P. M. (2004). Dynamic Psychiatry, 52, 572-581.e1. doi:10.1016/j.jaac.2013.02.017
mapping of human cortical development during child- Hu, S., Pruessner, J. C., Coupé, P., & Collins, D. L. (2013).
hood through early adulthood. Proceedings of the National Volumetric analysis of medial temporal lobe struc-
Academy of Sciences, USA, 101, 8174–8179. tures in brain development from childhood to ado-
Green, J., & Cox, A. (2000). Social and psychiatric function lescence. NeuroImage, 74, 276–287. doi:10.1016/j
in adolescents with Asperger syndrome compared with .neuroimage.2013.02.032
conduct disorder. Journal of Autism and Developmental Jankowski, K. F., Moore, W. E., Merchant, J. S., Kahn, L. E.,
Disorders, 30, 279–293. & Pfeifer, J. H. (2014). But do you think I’m cool?
Guyer, A. E., Choate, V. R., Pine, D. S., & Nelson, E. E. (2012). Developmental differences in striatal recruitment during
Neural circuitry underlying affective response to peer direct and reflected social self-evaluations. Developmental
feedback in adolescence. Social Cognitive and Affective Cognitive Neuroscience, 8, 40–54. doi:10.1016/j.dcn.2014
Neuroscience, 7, 81–92. doi:10.1093/scan/nsr043 .01.003
Halpern, C. T. (2006). Integrating hormones and other biologi- Johnson, S. A., Yechiam, E., Murphy, R. R., Queller, S., &
cal factors into a developmental systems model of ado- Stout, J. C. (2006). Motivational processes and auto-
lescent female sexuality. New Directions for Child and nomic responsivity in Asperger’s disorder: Evidence from
Adolescent Development, 112, 9–22. doi:10.1002/cd.159 the Iowa Gambling Task. Journal of the International
Hansen, R. L., Ozonoff, S., Krakowiak, P., Angkustsiri, K., Jones, Neuropsychological Society, 12, 668–676. doi:10.1017/
C., Deprey, L. J., . . . Hertz-Picciotto, I. (2008). Regression in S1355617706060802
autism: Prevalence and associated factors in the CHARGE Jones, R. M., Somerville, L. H., Li, J., Ruberry, E. J., Powers,
Study. Ambulatory Pediatrics, 8, 25–31. doi:10.1016/j A., Mehta, N., . . . Casey, B. J. (2014). Adolescent-specific
.ambp.2007.08.006 patterns of behavior and neural activity during social
Haracopos, D., & Pedersen, L. (1992). Sexuality and autism: A reinforcement learning. Cognitive, Affective, & Behavioral
nationwide survey in Denmark. Unpublished manuscript. Neuroscience. Advance online publication. doi:10.3758/
Hardan, A. Y., Libove, R. A., Keshavan, M. S., Melhem, N. M., s13415-014-0257-z
& Minshew, N. J. (2009). A preliminary longitudinal mag- Just, M. A., Cherkassky, V. L., Keller, T. A., Kana, R. K., &
netic resonance imaging study of brain volume and corti- Minshew, N. J. (2007). Functional and anatomical corti-
cal thickness in autism. Biological Psychiatry, 66, 320–326. cal underconnectivity in autism: Evidence from an FMRI
doi:10.1016/j.biopsych.2009.04.02 study of an executive function task and corpus callosum
Hartup, W. (1996). The company they keep: Friendships and morphometry. Cerebral Cortex, 17, 951–961. doi:10.1093/
their developmental significance. Child Development, 67, cercor/bhl006
1–13. Kana, R. K., Keller, T. A., Cherkassky, V. L., Minshew, N. J.,
Henault, I. (2006). Asperger’s syndrome and sexuality from & Just, M. A. (2006). Sentence comprehension in autism:
adolescence through adulthood. London, England: Jessica Thinking in pictures with decreased functional connectiv-
Kingsley. ity. Brain, 129, 2484–2493. doi:10.1093/brain/awl164
Henninger, N. A., & Taylor, J. L. (2013). Outcomes in adults Kanne, S. M., Gerber, A. J., Quirmbach, L. M., Sparrow, S. S.,
with autism spectrum disorders: A historical perspective. Cicchetti, D. V., & Saulnier, C. A. (2011). The role of adap-
Autism: The International Journal of Research and Practice, tive behavior in autism spectrum disorders: Implications for
17, 103–116. doi:10.1177/1362361312441266 functional outcome. Journal of Autism and Developmental
Herting, M. M., Maxwell, E. C., Irvine, C., & Nagel, B. J. (2012). Disorders, 41, 1007–1018. doi:10.1007/s10803-010-1126-4
The impact of sex, puberty, and hormones on white matter Knudson, A. G. (1971). Mutation and cancer: Statistical study
microstructure in adolescents. Cerebral Cortex, 22, 1979– of retinoblastoma. Proceedings of the National Academy of
1992. doi:10.1093/cercor/bhr246 Sciences, USA, 68, 820–823.
Hodges, E. V., & Perry, D. G. (1999). Personal and interpersonal Kohls, G., Peltzer, J., Schulte-Rüther, M., Kamp-Becker, I.,
antecedents and consequences of victimization by peers. Remschmidt, H., Herpertz-Dahlmann, B., & Konrad, K.
Journal of Personality and Social Psychology, 76, 677–685. (2011). Atypical brain responses to reward cues in autism
Hollander, E., Anagnostou, E., Chaplin, W., Esposito, K., as revealed by event-related potentials. Journal of Autism
Haznedar, M. M., Licalzi, E., . . . Buchsbaum, M. (2005). and Developmental Disorders, 41, 1523–1533. doi:10.1007/
Striatal volume on magnetic resonance imaging and repeti- s10803-011-1177-1
tive behaviors in autism. Biological Psychiatry, 58, 226– Kohls, G., Schulte-Rüther, M., Nehrkorn, B., Müller, K., Fink,
232. doi:10.1016/j.biopsych.2005.03.040 G. R., Kamp-Becker, I., . . . Konrad, K. (2013). Reward

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016
20 Picci, Scherf

system dysfunction in autism spectrum disorders. Social Mayes, S. D., Calhoun, S. L., Murray, M. J., & Zahid, J. (2011).
Cognitive and Affective Neuroscience, 8, 565–572. Variables associated with anxiety and depression in chil-
doi:10.1093/scan/nss03 dren with autism. Journal of Developmental and Physical
Kuo, M. H., Orsmond, G. I., Cohn, E. S., & Coster, W. J. (2013). Disabilities, 23, 325–337.
Friendship characteristics and activity patterns of ado- Mayes, S. D., Calhoun, S. L., Murray, M. J., Ahuja, M., & Smith,
lescents with an autism spectrum disorder. Autism: The L. A. (2011). Anxiety, depression, and irritability in children
International Journal of Research and Practice, 17, 481– with autism relative to other neuropsychiatric disorders
500. doi:10.1177/1362361311416380 and typical development. Research in Autism Spectrum
Kuusikko, S., Pollock-Wurman, R., Jussila, K., Carter, A. S., Disorders, 5, 474–485.
Mattila, M. L., Ebeling, H., . . . Moilanen, I. (2008). Social Maynard, T. M., Sikich, L., Lieberman, J. A., & LaMantia, A.
anxiety in high-functioning children and adolescents with S. (2001). Neural development, cell-cell signaling, and
autism and Asperger syndrome. Journal of Autism and the “two-hit” hypothesis of schizophrenia. Schizophrenia
Developmental Disorders, 38, 1697–1709. doi:10.1007/ Bulletin, 27, 457–476.
s10803-008-0555-9 Mazurek, M. O., & Engelhardt, C. R. (2013). Video game use
Lam, C. B., McHale, S. M., & Crouter, A. C. (2012). Parent- in boys with autism spectrum disorder, ADHD, or typi-
child shared time from middle childhood to late adoles- cal development. Pediatrics, 132, 260–266. doi:10.1542/
cence: Developmental course and adjustment correlates. peds.2012-3956
Child Development, 83, 2089–2103. doi:10.1111/j.1467- Mazurek, M. O., & Kanne, S. M. (2010). Friendship and inter-
8624.2012.01826.x nalizing symptoms among children and adolescents with
Lee, J. E., Bigler, E. D., Alexander, A. L., Lazar, M., DuBray, ASD. Journal of Autism and Developmental Disorders, 40,
M. B., Chung, M. K., . . . Lainhart, J. E. (2007). Diffusion 1512–1520. doi:10.1007/s10803-010-1014-y
tensor imaging of white matter in the superior temporal McPheeters, M. L., Davis, A., Navarre, J. R., & Scott, T. A.
gyrus and temporal stem in autism. Neuroscience Letters, (2011). Family report of ASD concomitant with depres-
424, 127–132. doi:10.1016/j.neulet.2007.07.042 sion or anxiety among US children. Journal of Autism and
Lenroot, R. K., & Giedd, J. N. (2010). Sex differences in the Developmental Disorders, 41, 646–653. doi:10.1007/s10803-
adolescent brain. Brain and Cognition, 72, 46–55. 010-1085-9
doi:10.1016/j.bandc.2009.10.008 Minshew, N. J., Luna, B., & Sweeney, J. A. (1999). Occulomotor
Levy, A., & Perry, A. (2011). Outcomes in adolescents and evidence for neocortical systems but not cerebellar dys-
adults with autism: A review of the literature. Research function in autism. Neurology, 52, 917–921. doi:10.1212/
in Autism Spectrum Disorders, 5, 1271–1282. doi:10.1016/ WNL.52.5.917
j.rasd.2011.01.023 Mizuno, A., Villalobos, M. E., Davies, M. M., Dahl, B. C., &
Lin, A., Rangel, A., & Adolphs, R. (2012). Impaired learning of Müller, R. A. (2006). Partially enhanced thalamocortical
social compared to monetary rewards in autism. Frontiers functional connectivity in autism. Brain Research, 1104,
in Neuroscience, 6, 143. Retrieved from http://journal 160–174. doi:10.1016/j.brainres.2006.05.064
.frontiersin.org/Journal/10.3389/fnins.2012.00143/full Monk, C. S., Peltier, S. J., Wiggins, J. L., Weng, S. J., Carrasco, M.,
Little, L. (2002). Middle-class mothers’ perceptions of peer Risi, S., & Lord, C. (2009). Abnormalities of intrinsic functional
and sibling victimization among children with Asperger’s connectivity in autism spectrum disorders. NeuroImage, 47,
syndrome and nonverbal learning disorders. Issues in 764–772. doi:10.1016/j.neuroimage.2009.04.069
Comprehensive Pediatric Nursing, 25, 43–57. Moore, W. E., Pfeifer, J. H., Masten, C. L., Mazziotta, J. C.,
Marceau, K., Dorn, L. D., & Susman, E. J. (2012). Stress Iacoboni, M., & Dapretto, M. (2012). Facing puberty:
and puberty-related hormone reactivity, negative Associations between pubertal development and neural
emotionality, and parent–adolescent relationships. responses to affective facial displays. Social Cognitive and
Psychoneuroendocrinology, 37, 1286–1298. Affective Neuroscience, 7, 35–43. doi:10.1093/scan/nsr066
Martin, C. A., Kelly, T. H., Rayens, M. K., Brogli, B. R., Brenzel, Mosconi, M. W., Cody-Hazlett, H., Poe, M. D., Gerig, G.,
A., Smith, W. J., & Omar, H. A. (2002). Sensation seeking, Gimpel-Smith, R., & Piven, J. (2009). Longitudinal study of
puberty, and nicotine, alcohol, and marijuana use in ado- amygdala volume and joint attention in 2- to 4-year-old
lescence. Journal of the American Academy of Child and children with autism. Archives of General Psychiatry, 66,
Adolescent Psychiatry, 41, 1495–1502. 509–516.
Masten, C. L., Colich, N. L., Rudie, J. D., Bookheimer, S. Y., Nacewicz, B. M., Dalton, K. M., Johnstone, T., Long, M. T.,
Eisenberger, N. I., & Dapretto, M. (2011). An fMRI inves- McAuliff, E. M., Oakes, T. R., . . . Davidson, R. J. (2006).
tigation of responses to peer rejection in adolescents Amygdala volume and nonverbal social impairment in ado-
with autism spectrum disorders. Developmental Cognitive lescent and adult males with autism. Archives of General
Neuroscience, 1, 260–270. doi:10.1016/j.dcn.2011.01.004 Psychiatry, 63, 1417–1428.
Matson, J. L., Dempsey, T., & Fodstad, J. C. (2009). The effect of Nair, A., Treiber, J. M., Shukla, D. K., Shih, P., & Müller, R.
autism spectrum disorders on adaptive independent living A. (2013). Impaired thalamocortical connectivity in autism
skills in adults with severe intellectual disability. Research spectrum disorder: A study of functional and anatomical
in Developmental Disabilities, 30, 1203–1211. doi:10.1016/ connectivity. Brain, 136, 1942–1955. doi:10.1093/brain/
j.ridd.2009.04.001 awt079

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016
Two-Hit Model of Autism 21

Nelson, E. E., Leibenluft, E., McClure, E., & Pine, D. S. (2005). Rodrigo, M. J., Padrón, I., de Vega, M., & Ferstl, E. C. (2014).
The social re-orientation of adolescence: A neuroscience Adolescents’ risky decision-making activates neural networks
perspective on the process and its relation to psychopa- related to social cognition and cognitive control processes.
thology. Psychological Medicine, 35, 163–174. doi:10.1017/ Frontiers in Human Neuroscience, 8, 60. Retrieved from
S0033291704003915 http://journal.frontiersin.org/Journal/10.3389/fnhum.2014
Newport, J., Newport, M., & Bolick, T. (2002). Autism-Asperger’s .00060/full
& sexuality: Puberty and beyond. Arlington, TX: Future Roisman, G., Masten, A., Coatsworth, J., & Tellegen, A. (2004).
Horizons. Salient and emerging developmental tasks in the transition
O’Hearn, K., Asato, M., Ordaz, S., & Luna, B. (2008). to adulthood. Child Development, 75, 123–133.
Neurodevelopment and executive function in autism. Rojas, D., Peterson, E., Winterrowd, E., Reite, M., Rogers, S.,
Development and Psychopathology, 20, 1103–1132. & Tregellas, J. (2006). Regional gray matter volumetric
doi:10.1017/S0954579408000527 changes in autism associated with social and repetitive
O’Hearn, K., Schroer, E., Minshew, N., & Luna, B. (2010). Lack behavior symptoms. BMC Psychiatry, 6, 56. Retrieved from
of developmental improvement on a face memory task http://www.biomedcentral.com/1471-244X/6/56
during adolescence in autism. Neuropsychologia, 48, 3955– Romeo, R. (2003). Puberty: A period of both organizational and
3960. doi:10.1016/j.neuropsychologia.2010.08.024 activational effects of steroid hormones on neurobehav-
Orsmond, G. I., Krauss, M. W., & Seltzer, M. M. (2004). Peer ioural development. Journal of Neuroendocrinology, 15,
relationships and social and recreational activities among 1185–1192.
adolescents and adults with autism. Journal of Autism and Rosenthal, M., Wallace, G. L., Lawson, R., Wills, M. C., Dixon, E.,
Developmental Disorders, 34, 245–256. Yerys, B. E., & Kenworthy, L. (2013). Impairments in real-
Ostby, Y., Tamnes, C. K., Fjell, A. M., Westlye, L. T., Due- world executive function increase from childhood to ado-
Tønnessen, P., & Walhovd, K. B. (2009). Heterogeneity in lescence in autism spectrum disorders. Neuropsychology,
subcortical brain development: A structural magnetic reso- 27, 13–18.
nance imaging study of brain maturation from 8 to 30 years. Rowley, E., Chandler, S., Baird, G., Simonoff, E., Pickles, A.,
Journal of Neuroscience, 29, 11772–11782. doi:10.1523/ Loucas, T., & Charman, T. (2012). The experience of
JNEUROSCI.1242-09.2009 friendship, victimization and bullying in children with
Ousley, O., & Mesibov, G. (1991). Sexual attitudes and knowl- an autism spectrum disorder: Associations with child
edge of high-functioning adolescents and adults with characteristics and school placement. Research in
autism. Journal of Autism and Developmental Disorders, Autism Spectrum Disorders, 6, 1126–1134. doi:10.1016/j
21, 471–481. .rasd.2012.03.004
Patton, J. H., Stanford, M. S., & Barratt, E. S. (1995). Factor struc- Ruble, L., & Dalrymple, N. (1993). Social/sexual awareness of
ture of the Barratt Impulsiveness Scale. Journal of Clinical persons with autism: A parental perspective. Archives of
Psychology, 51, 768–774. Sexual Behavior, 22, 229–240.
Peake, S. J., Dishion, T. J., Stormshak, E. A., Moore, W. E., & Rutter, M. (1970). Autistic children: Infancy to adulthood.
Pfeifer, J. H. (2013). Risk-taking and social exclusion in ado- Seminars in Psychiatry, 2, 435–450.
lescence: Neural mechanisms underlying peer influences Schaer, M., Ottet, M. C., Scariati, E., Dukes, D., Franchini, M.,
on decision-making. NeuroImage, 82, 23–34. doi:10.1016/j Eliez, S., & Glaser, B. (2013). Decreased frontal gyrifica-
.neuroimage.2013.05.061 tion correlates with altered connectivity in children with
Pfeifer, J. H., Kahn, L. E., Merchant, J. S., Peake, S. J., Veroude, autism. Frontiers in Human Neuroscience, 7, 750. Retrieved
K., Masten, C. L., . . . Dapretto, M. (2013). Longitudinal from http://journal.frontiersin.org/Journal/10.3389/fnhum
change in the neural bases of adolescent social self- .2013.00750/full
evaluations: Effects of age and pubertal development. Scherf, K., Behrmann, M., Kimchi, R., & Luna, B. (2009).
Journal of Neuroscience, 33, 7415–7419. doi:10.1523/ Emergence of global shape processing continues through
JNEUROSCI.4074-12.2013 adolescence. Child Development, 80, 162–177.
Pierce, K., & Redcay, E. (2008). Fusiform function in children with Scherf, K. S., Luna, B., Kimchi, R., Minshew, N., & Behrmann,
an autism spectrum disorder is a matter of “who.” Biological M. (2008). Missing the big picture: Impaired development
Psychiatry, 64, 552–560. doi:10.1016/j.biopsych.2008.05.013 of global shape processing in autism. Autism Research, 1,
Quevedo, K. M., Benning, S. D., Gunnar, M. R., & Dahl, R. 114–129. doi:10.1002/aur.17
E. (2009). The onset of puberty: Effects on the psycho- Scherf, K. S., & Scott, L. S. (2012). Connecting developmen-
physiology of defensive and appetitive motivation. tal trajectories: Biases in face processing from infancy to
Development and Psychopathology, 21, 27–45. doi:10.1017/ adulthood. Developmental Psychobiology, 54, 643–663.
S0954579409000030 doi:10.1002/dev.21013
Rauer, A. J., Pettit, G. S., Lansford, J. E., Bates, J. E., & Dodge, Scherf, K. S., Thomas, C., Doyle, J., & Behrmann, M. (2013).
K. A. (2013). Romantic relationship patterns in young adult- Emerging structure-function relations in the developing
hood and their developmental antecedents. Developmental face processing system. Cerebral Cortex. Advance online
Psychology, 49, 2159–2171. doi:10.1037/a0031845 publication. doi:10.1093/cercor/bht152
Roberts, R. E., & Duong, H. T. (2014). The prospective asso- Schmithorst, V., & Yuan, W. (2010). White matter development
ciation between sleep deprivation and depression among during adolescence as shown by diffusion MRI. Brain and
adolescents. Sleep, 37, 239–244. doi:10.5665/sleep.3388 Cognition, 72, 16–25. doi:10.1016/j.bandc.2009.06.005

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016
22 Picci, Scherf

Schulz, K. M., Molenda-Figueira, H. A., & Sisk, C. L. (2009). and Developmental Disorders, 41, 55–65. doi:10.1007/
Back to the future: The organizational–activational hypoth- s10803-010-1021-z
esis adapted to puberty and adolescence. Hormones and South, M., Ozonoff, S., Suchy, Y., Kesner, R. P., McMahon,
Behavior, 55, 597–604. doi:10.1016/j.yhbeh.2009.03.010 W. M., Lainhart, J. E., . . . Passingham, R. (2008). Intact
Schumann, C. M., & Amaral, D. G. (2006). Stereological emotion facilitation for nonsocial stimuli in autism: Is amyg-
analysis of amygdala neuron number in autism. dala impairment in autism specific for social information?
Journal of Neuroscience, 26, 7674–7679. doi:10.1523/ Journal of the International Neuropsychological Society, 14,
JNEUROSCI.1285-06.2006 42–54. doi:10.10170S1355617708080107
Seiffge-Krenke, I., Luyckx, K., & Salmela-Aro, K. (2014). Work Sparks, B. F., Friedman, S. D., Shaw, D. W., Aylward, E. H.,
and love during emerging adulthood: Introduction to the Echelard, D., Artru, A. A., . . . Dager, S. R. (2002). Brain
special issue. Emerging Adulthood, 2, 3–5. structural abnormalities in young children with autism
Seltzer, M. M., Shattuck, P., Abbeduto, L., & Greenberg, J. spectrum disorder. Neurology, 59, 184–192. doi:10.1212/
S. (2004). Trajectory of development in adolescents and WNL.59.2.184
adults with autism. Mental Retardation and Developmental Spear, L. P. (2000). The adolescent brain and age-related behav-
Disabilities Research Reviews, 10, 234–247. doi:10.1002/ ioral manifestations. Neuroscience & Biobehavioral Reviews,
mrdd.20038 24, 417–463.
Shukla, D. K., Keehn, B., & Müller, R. A. (2011). Tract-specific Spear, L. P. (2011). Rewards, aversions and affect in adoles-
analyses of diffusion tensor imaging show widespread cence: Emerging convergences across laboratory animal
white matter compromise in autism spectrum disorder. and human data. Developmental Cognitive Neuroscience, 1,
Journal of Child Psychology and Psychiatry, 52, 286–295. 390–403. doi:10.1016/j.dcn.2011.08.001
doi:10.1111/j.1469-7610.2010.02342.x Steinberg, L. (2008). A social neuroscience perspective on
Simpson, J. A., Collins, W. A., & Salvatore, J. E. (2011). The adolescent risk-taking. Developmental Review, 28, 78–106.
impact of early interpersonal experience on adult roman- doi:10.1016/j.dr.2007.08.002
tic relationship functioning: Recent findings from the Steinberg, L. (2010). A dual systems model of adolescent
Minnesota Longitudinal Study of Risk and Adaptation. risk-taking. Developmental Psychobiology, 52, 216–224.
Current Directions in Psychological Science, 20, 355–359. doi:10.1002/dev.20445
doi:10.1177/0963721411418468 Sterzing, P. R., Shattuck, P. T., Narendorf, S. C., Wagner, M.,
Sisk, C. L., & Zehr, J. L. (2005). Pubertal hormones orga- & Cooper, B. P. (2012). Bullying involvement and autism
nize the adolescent brain and behavior. Frontiers spectrum disorders: Prevalence and correlates of bullying
in Neuroendocrinology, 26, 163–174. doi:10.1016/j involvement among adolescents with an autism spectrum
.yfrne.2005.10.003 disorder. Archives of Pediatrics and Adolescent Medicine,
Sivertsen, B., Posserud, M. B., Gillberg, C., Lundervold, A. J., & 166, 1058–1064. doi:10.1001/archpediatrics.2012.790
Hysing, M. (2012). Sleep problems in children with autism Stokes, M. A., & Kaur, A. (2005). High-functioning autism
spectrum problems: A longitudinal population-based study. and sexuality: A parental perspective. Autism: The
Autism: The International Journal of Research and Practice, International Journal of Research and Practice, 9, 266–289.
16, 139–150. doi:10.1177/1362361311404255 doi:10.1177/1362361305053258
Solish, A., Perry, A., & Minnes, P. (2010). Participation of chil- Stokes, M., Newton, N., & Kaur, A. (2007). Stalking, and social
dren with and without disabilities in social, recreational and and romantic functioning among adolescents and adults
leisure activities. Journal of Applied Research in Intellectual with autism spectrum disorder. Journal of Autism and
Disabilities, 23, 226–236. Developmental Disorders, 37, 1969–1986. doi:10.1007/
Solomon, M., Ozonoff, S. J., Ursu, S., Ravizza, S., Cummings, s10803-006-0344-2
N., Ly, S., & Carter, C. S. (2009). The neural substrates Stoner, R., Chow, M. L., Boyle, M. P., Sunkin, S. M., Mouton,
of cognitive control deficits in autism spectrum disor- P. R., Roy, S., . . . Courchesne, E. (2014). Patches of disor-
ders. Neuropsychologia, 47, 2515–2526. doi:10.1016/j ganization in the neocortex of children with autism. New
.neuropsychologia.2009.04.019 England Journal of Medicine, 370, 1209–1219. doi:10.1056/
Solomon, M., Yoon, J. H., Ragland, J. D., Niendam, T. A., NEJMoa1307491
Lesh, T. A., Fairbrother, W., & Carter, C. S. (2013). The Supekar, K., Uddin, L. Q., Khouzam, A., Phillips, J., Gaillard, W.
development of the neural substrates of cognitive control D., Kenworthy, L. E., . . . Menon, V. (2013). Brain hypercon-
in adolescents with autism spectrum disorders. Biological nectivity in children with autism and its links to social defi-
Psychiatry. Advance online publication. doi:10.1016/j cits. Cell Reports, 5, 738–747. doi:10.1016/j.celrep.2013.10.001
.biopsych.2013.08.036 Tabuchi, K., Blundell, J., Etherton, M. R., Hammer, R. E., Liu,
Somerville, L. H., & Casey, B. J. (2010). Developmental neu- X., Powell, C. M., & Südhof, T. C. (2007). A neuroligin-3
robiology of cognitive control and motivational sys- mutation implicated in autism increases inhibitory synap-
tems. Current Opinion in Neurobiology, 20, 236–241. tic transmission in mice. Science, 318, 71–76. doi:10.1126/
doi:10.1016/j.conb.2010.01.006 science.1146221
South, M., Dana, J., White, S. E., & Crowley, M. J. (2011). Failure Tamnes, C. K., Ostby, Y., Fjell, A. M., Westlye, L. T., Due-
is not an option: Risk-taking is moderated by anxiety and Tønnessen, P., & Walhovd, K. B. (2010). Brain maturation
also by cognitive ability in children and adolescents diag- in adolescence and young adulthood: Regional age-related
nosed with an autism spectrum disorder. Journal of Autism changes in cortical thickness and white matter volume and

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016
Two-Hit Model of Autism 23

microstructure. Cerebral Cortex, 20, 534–548. doi:10.1093/ behavioral alterations. Hormones and Behavior, 64, 343–
cercor/bhp118 349. doi:10.1016/j.yhbeh.2012.11.012
Uddin, L. Q., Supekar, K., & Menon, V. (2013). Reconceptualizing Villalobos, M. E., Mizuno, A., Dahl, B. C., Kemmotsu, N., &
functional brain connectivity in autism from a develop- Müller, R. A. (2005). Reduced functional connectivity
mental perspective. Frontiers in Human Neuroscience, between V1 and inferior frontal cortex associated with
7, 458. Retrieved from http://journal.frontiersin.org/ visuomotor performance in autism. NeuroImage, 25, 916–
Journal/10.3389/fnhum.2013.00458/full 925. doi:10.1016/j.neuroimage.2004.12.022
Uematsu, A., Matsui, M., Tanaka, C., Takahashi, T., Noguchi, Wallen, K. (2001). Sex and context: Hormones and primate
K., Suzuki, M., & Nishijo, H. (2012). Developmental tra- sexual motivation. Hormones and Behavior, 40, 339–357.
jectories of amygdala and hippocampus from infancy to Waylen, A., & Wolke, D. (2004). Sex ’n’ drugs ’n’ rock ’n’ roll:
early adulthood in healthy individuals. PLoS ONE, 7(10), The meaning and social consequences of pubertal timing.
e46970. Retrieved from http://www.plosone.org/article/ European Journal of Endocrinology, 151, U151–U159.
info%3Adoi%2F10.1371%2Fjournal.pone.0046970 Weng, S. J., Carrasco, M., Swartz, J. R., Wiggins, J. L., Kurapati,
van Kooten, I., Palmen, S., Von Cappeln, P., Steinbusch, H., N., Liberzon, I., . . . Monk, C. S. (2011). Neural activation
Korr, H., Heinsen, H., . . . Schmitz, C. (2008). Neurons in to emotional faces in adolescents with autism spectrum
the fusiform gyrus are fewer and smaller in autism. Brain, disorders. Journal of Child Psychology and Psychiatry, 52,
131, 987–999. doi:10.1093/brain/awn033 296–305. doi:10.1111/j.1469-7610.2010.02317.x
Van Leijenhorst, L., Gunther Moor, B., Op de Macks, Z. A., Whitehouse, A. J., Watt, H. J., Line, E. A., & Bishop, D. V.
Rombouts, S. A., Westenberg, P. M., & Crone, E. A. (2010). (2009). Adult psychosocial outcomes of children
Adolescent risky decision-making: Neurocognitive devel- with specific language impairment, pragmatic lan-
opment of reward and control regions. NeuroImage, 51, guage impairment and autism. International Journal of
345–355. doi:10.1016/j.neuroimage.2010.02.038 Language & Communication Disorders, 44, 511–528.
van Roekel, E., Scholte, R. H., & Didden, R. (2010). Bullying among doi:10.1080/13682820802708098
adolescents with autism spectrum disorders: Prevalence and Yu, R., Branje, S., Keijsers, L., & Meeus, W. (2014). Brief report:
perception. Journal of Autism and Developmental Disorders, How adolescent personality moderates the effect of love
40, 63–73. doi:10.1007/s10803-009-0832-2 history on the young adulthood romantic relationship qual-
Vargas, D. L., Nascimbene, C., Krishnan, C., Zimmerman, A. W., ity. Journal of Adolescence. Advance online publication.
& Pardo, C. A. (2005). Neuroglial activation and neuroin- doi:10.1016/j.adolescence.2014.02.006
flammation in the brain of patients with autism. Annals of Zuckerman, M., Eysenck, S. B., & Eysenck, H. J. (1978).
Neurology, 57, 67–81. doi:10.1002/ana.20315 Sensation seeking in England and America: Cross-cultural,
Varlinskaya, E. I., Vetter-O’Hagen, C. S., & Spear, L. P. (2013). age, and sex comparisons. Journal of Consulting and
Puberty and gonadal hormones: Role in adolescent-typical Clinical Psychology, 46, 139–149.

Downloaded from cpx.sagepub.com at UNIV CALIFORNIA SAN DIEGO on February 11, 2016