Early visual foraging in relationship to familial risk for autism and hyperactivity/inattention: A preliminary study

Abstract

Information foraging is atypical in both Autism Spectrum Disorders (ASD) and
attention deficit hyperactivity disorder (ADHD), however while ASD is associated
with restricted exploration and preference for sameness, ADHD is characterized by
hyperactivity and increased novelty seeking. Here we ask whether similar biases are
present in visual foraging in younger siblings of children with a diagnosis of ASD
with or without additional high levels of hyperactivity and inattention. Fifty-four lowrisk
controls (LR) and 50 high-risk siblings (HR) took part in an eye-tracking study at
8 and 14 months and at 3 years of age. At 8 months, siblings of children with ASD
and low levels of hyperactivity/inattention (HR/ASD-HI) were more likely to return
to previously visited areas in the visual scene than were LR and siblings of children
with ASD and high levels of hyperactivity/inattention (HR/ASD+HI). Thus, we reveal
a paradoxical effect, in which additional family risk for ADHD core-symptoms
mitigates the effect of ASD risk on visual information foraging.

Full text

Journal of Attention Disorders
1 –9
© The Author(s) 2015
Reprints and permissions:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/1087054715616490
jad.sagepub.com
Article
Many human activities rely on information foraging, that is,
the seeking out and processing of information. Theoretical
accounts of information foraging draw heavily on general
models for resource foraging, which describe a trade-off
between effort or energy expenditure and the amount of
resources the organism manages to secure. Optimal foraging
requires shifting between exploitation, that is, taking advan-
tage of the resource at hand and exploration of the environ-
ment to discover new resources (Cohen, McClure, & Yu,
2007). There is no absolute best combination of exploration
and exploitation, but rather organisms gain from flexibly
adapting foraging to environmental conditions (i.e., it is better
to exploit the resources at hand if the further environment is
uncertain, and better to explore when faced with competition;
Kacelnik, Houston, & Krebs, 1981). Any strong inherent bias,
either toward exploratory behavior or for resource exploita-
tion, might therefore interfere with optimal foraging. This
article deals with the development of such potential biases in
information foraging, in infants at familial risk for autism
spectrum disorders (ASDs) or for attentional problems.
Atypical Foraging in ASD and ADHD
Individual differences in foraging behavior have been related
to the function of particular neural systems. For example,
biases for exploratory behavior were associated with differ-
ences in dopaminergic and noradrenergic function (Frank,
Doll, Oas-Terpstra, & Moreno, 2009; Jepma & Nieuwenhuis,
2011). Stimulating dopaminergic function in monkeys leads
to an increase in preference for novel stimuli over old stimuli
with larger reward value (Costa, Tran, Turchi, & Averbeck,
2014). Explorative choices, that is, choosing a new over an
old item, are preceded by increases in pupil diameter, which
index locus coeruleus/noradrenergic function (Jepma &
Nieuwenhuis, 2011). Both ASD and ADHD have been associ-
ated with atypicality in dopamine (e.g., Kriete & Noelle,
2015; Solanto, 2002) and noradrenaline functions (Biederman
& Spencer, 1999; Blaser, Eglington, Carter, & Kaldy, 2014) as
well as, separately, with atypical information foraging. In
ASD, decreased exploratory behavior was documented in a
task in which children had to discover objects hidden in
616490JADXXX10.1177/1087054715616490Journal of Attention DisordersGliga et al.
research-article2015
1Centre for Brain and Cognitive Development, Birkbeck, University of
London, UK
2Institute of Psychiatry, Psychology & Neuroscience, King’s College
London, UK
Corresponding Author:
Teodora Gliga, Birkbeck, University of London, Malet Street, WC1E
7HX, London, UK.
Email: t.gliga@bbk.ac.uk
Early Visual Foraging in Relationship
to Familial Risk for Autism and
Hyperactivity/Inattention
Teodora Gliga1, Tim J. Smith1, Noreen Likely1, Tony Charman2,
and Mark H. Johnson1
Abstract
Objective. Information foraging is atypical in both autism spectrum disorders (ASDs) and ADHD; however, while ASD is
associated with restricted exploration and preference for sameness, ADHD is characterized by hyperactivity and increased
novelty seeking. Here, we ask whether similar biases are present in visual foraging in younger siblings of children with
a diagnosis of ASD with or without additional high levels of hyperactivity and inattention. Method. Fifty-four low-risk
controls (LR) and 50 high-risk siblings (HR) took part in an eye-tracking study at 8 and 14 months and at 3 years of age.
Results. At 8 months, siblings of children with ASD and low levels of hyperactivity/inattention (HR/ASD-HI) were more
likely to return to previously visited areas in the visual scene than were LR and siblings of children with ASD and high levels
of hyperactivity/inattention (HR/ASD+HI). Conclusion. We show that visual foraging is atypical in infants at-risk for ASD.
We also reveal a paradoxical effect, in that additional family risk for ADHD core symptoms mitigates the effect of ASD risk
on visual information foraging. (J. of Att. Dis. XXXX; XX(X) XX-XX)
Keywords
infants, ASD, visual attention, eye-tracking
by guest on December 5, 2015jad.sagepub.comDownloaded from

2 Journal of Attention Disorders
different containers. Children with ASD spent less time in
active exploration and opened fewer containers and these
behaviors related to anatomical differences in the cerebellum,
such as differences in the size of the vermal lobules (Pierce &
Courchesne, 2001). In another similar foraging study,
Pellicano et al. (2011) reported less systematic searches and a
higher proportion of revisitations in children with ASD.
Differences were also found in free-viewing of visual scenes
(Elison, Sasson, Turner-Brown, Dichter, & Bodfish, 2012).
Typically developing (TD) and ASD participants aged 2 to 18
years were presented with large arrays of images. The number
of images visited per unit of time increased with age in both
groups, but with a steeper slope in TD participants (Elison
et al., 2012). Thus, across measures, ASD seems to be associ-
ated with less exploratory behavior. When findings could not
be explained by specific preferences related to the disorder, as
is the case in both Pellicano et al. (2011) and Pierce and
Courchesne (2001), returning to already explored locations/
objects could be driven by incomplete information processing
or poor memory. However, IQ was not a predictor of perfor-
mance in either Elison et al. (2012) or in Pierce and Courchesne
(2001), and groups were matched in IQ in Pellicano et al.
(2011). Thus, these biases seem to manifest independently of
concurrent information processing difficulties.
In contrast to ASD, ADHD is associated with extreme
novelty seeking, especially in the case of the hyperactive
and combined (hyperactive and inattentive) subtypes
(Salgado et al., 2008). Intriguing evidence that genes asso-
ciated with ADHD (e.g., DRD4 gene variants) are more fre-
quent in populations that have a history of migration (Chen,
Burton, Greenberger, & Dmitrieva, 1999; Matthews &
Butler, 2011) led to the suggestion that hyperactivity might
be an adaptation to the food-scarce and volatile environ-
ment our ancestors lived in (Jensen et al., 1997). Some
rather indirect evidence for a bias toward exploratory
behavior in ADHD comes from a study in which the num-
ber of regions visited when free-viewing a visual scene cor-
related with differences in curiosity (Risko, Anderson,
Lanthier, & Kingstone, 2012), which some have associated
with ADHD (Williams & Taylor, 2006).
The existing literature suggests ASD and ADHD might
be associated with opposing biases in foraging. However, as
ASD and ADHD often co-occur (approximately 20% of
U.K. 7-year-olds with ASD meet criteria for ADHD, and
vice versa; Russell, Rodgers, Ukoumunne, & Ford, 2014),
this raises the question of how ASD and ADHD-specific
foraging biases interact during development. Additive phe-
notypic effects were previously described in children with
comorbid ASD and ADHD, for example, neural processing
of human gaze in these children was similar to both profiles
of children with ASD only or with ADHD only (Tye et al.,
2013). An increase in symptom severity, compared with the
single diagnosis cases, was also documented (e.g., Goldin,
Matson, Tureck, Cervantes, & Jang, 2013). These findings
raise the intriguing possibility that, when disease-specific
phenotypes are at opposite ends of a spectrum (e.g.,
increased exploration in ADHD and decreased exploration
in ASD), risk for one disorder may mitigate the effects of
risk for the other disorder.
We were therefore interested in investigating the impact
ASD and ADHD risk has on information foraging during
development. Both occulo-motor behaviors and object
manipulation have been used to measure information forag-
ing in infants (Bornstein, Hahn, & Suwalsky, 2013). Of the
two measures, visual scanning is less confounded by gen-
eral motor development.
Visual Foraging and Its Development
It has been suggested that the type of cognitive processes that
derived from spatial foraging for resources, such as food,
may also be associated with the control of visual attention
(Hills, Todd, & Goldstone, 2008). For example, when freely
exploring new visual scenes, participants make shorter fixa-
tions and longer saccades during the first few seconds, cover-
ing the whole visual scene. However, as participants continue
to look, their fixations gradually become longer and their sac-
cades shorter (Fischer, Graupner, Velichkovsky, & Pannasch,
2013; Pannasch, Helmert, Roth, Herbold, & Walter, 2008).
This suggests that the optimum visual information strategy
relies on an initial exploratory phase, where the new scene is
mapped to discover potential interesting sights, followed by
information exploitation, where chosen locations (sources of
information) are repeatedly investigated (Krebs, Kacelnik, &
Taylor, 1978). There is still limited understanding of devel-
opmental changes in visual foraging, but the limited existing
evidence suggests an increase in exploration with age. During
the first month after birth, infants restrict their scanning to a
small portion of an image, but by 3 months they produce lon-
ger saccades and gaze patterns that are more systematically
distributed over visual scenes (Bronson, 1991). However,
even beyond 6 months of age, scanning patterns remain
restricted to particular locations of the visual scene
(Schlesinger & Amso, 2013). In the current study we will ask
whether measures of visual scanning capture information
foraging atypicalities early in development.
The Current Study
The current work is a reanalysis of a previously published
data set (Elsabbagh et al., 2013). The “face-pop out” para-
digm was designed to measure face orienting as an early
marker for ASD. Infants freely explored displays contain-
ing a face and four other objects while their eye movements
were measured with an eye tracker (Figure 1). Because
infants had 15 s to explore each new display, by analyzing
the sequences of visits to faces and other objects, this para-
digm can inform about visual foraging strategies.
by guest on December 5, 2015jad.sagepub.comDownloaded from

Gliga et al. 3
Study participants were infants with an older sibling with
ASD. Approximately 20% of children at-risk will receive a
diagnosis of ASD themselves (Ozonoff et al., 2011), and
another 30% will manifest elevated levels of ASD symptoms
(Messinger et al., 2013). Thus, by comparing visual foraging
in this population with a low-risk cohort, we aim to capture
differences due to genetic susceptibility for ASD. As men-
tioned earlier, ASD and ADHD are often co-occurring and,
like ASD, ADHD has moderate heritability (Larsson, Chang,
D’Onofrio, & Lichtenstein, 2014). To investigate the effect
ADHD risk has on visual foraging, we characterized the
older siblings’ hyperactivity and inattention profile using the
Strengths and Difficulties Questionnaire (SDQ). The SDQ
has been extensively used as a screener for ADHD and has
good sensitivity, especially for children with the combined
subtype (hyperactivity/inattention; Carballo, Rodríguez-
Blanco, García-Nieto, & Baca-García, 2014; Ullebø,
Posserud, Heiervang, Gillberg, & Obel, 2011). To reflect the
fact that the SDQ is a screening, not a diagnostic, instrument
we refer to the risk conferred by a high score on the SDQ as
hyperactivity/inattention risk (HI risk). We can thus compare
the effect of ASD risk, and that of additional HI risk, on
visual foraging during early development. Infants contrib-
uted data at three age points (8 months, 14 months, and 3
years). We analyzed the temporal dynamics of attention as
participants scanned the different areas of interest (AOIs),
focusing in particular on the likelihood of images being revis-
ited. We expected ASD risk to be associated with decreased
exploration (i.e., higher likelihood of revisiting AOI) and that
additional HI risk may moderate this effect.
Method
Participants
Participants were 54 infants at high familial risk (HR) and
50 infants at low familial risk (LR) for ASD. Infants
attended lab-based testing 3 times, first between 6 and 10
months, a second time around 14 months, and a third time
around 3 years of age. Only a subset of participants contrib-
uted data at the 3-year visit (Table 1). At the time of enroll-
ment, none of the infants had been diagnosed with any
medical or developmental condition. HR infants all had an
older sibling (proband) with a community clinical diagnosis
of ASD. Proband diagnosis was confirmed by two expert
clinicians (GP, TC) based on information from the
Development and Well-Being Assessment (DAWBA) and
the parent-report Social Communication Questionnaire
(SCQ, Rutter, Bailey, & Lord, 2003). Infants in the low-risk
group were recruited from a volunteer database at the xxx.
All low-risk infants had at least one older sibling with typi-
cal development and no first-degree relatives with ASD.
None of the older siblings scored above instrument cutoff
for ASD on the SCQ (≥15, 1 score missing).
Assessing Hyperactivity/Inattention-Risk
Proband ADHD risk was attributed based on scores on the
SDQ, filled in by the parent of HR participants. The SDQ
is composed of 25 items that ask about behavioral attri-
butes of the child and are combined to form five sub-
scales. The hyperactivity/inattention subscale covers
restlessness, fidgeting, concentration, distractibility, and
impulsivity. Each item can be answered with not true,
somewhat true, or certainly true and they are scored 0, 1,
or 2, respectively, giving a total score out of a possible 10
for each subscale. Screening studies suggest a cutoff of
8/10 as providing the best ADHD diagnostic accuracy
(Carballo et al., 2014). Using a cutoff of 8, we separated
the HR-ASD group into a HR/ASD-HI group (n = 26),
including children whose probands scored below 8 (aver-
age = 5.19, SD = 1.7) and a HR/ASD+HI group (n = 20),
when probands scored 8 or above (average = 9.25, SD =
0.78). Hyperactivity and Inattention subscale scores did
not relate to proband social communication abilities (cor-
relation with SCQ, r = .246, p > .1).
Visual Scanning Task and Procedure
The same stimuli and procedure as previously described in
Elsabbagh et al. (2013) was used. At 8 and 14 months,
infants saw 14 different slides (example in Figure 1), each
for 15 s. At 3 years, only 10 slides were shown. Each slide
contained 5 images, one from each of the following catego-
ries: faces, mobile phones, birds, cars, and scrambled faces.
Figure 1. Example stimulus, AOIs, and visit coding.
Note. Successive fixations within an AOI were coded as one visit (e.g.
fixations 1 and 2 and fixations 3 and 4, in this example). The third visit, in
green, is a revisit to the face. AOI = area of interest.
by guest on December 5, 2015jad.sagepub.comDownloaded from

4 Journal of Attention Disorders
A central attention getter was presented before each slide to
reorient infant’s attention. Infants were seated on their par-
ent’s lap in front of the eye-tracking monitor (at 8 and 14
months) or on their own in front of the eye-tracking monitor
(3 years of age). Parents were asked to refrain from pointing
to the screen or naming any of the images.
Revisitation Analysis
Gaze data were recorded with a Tobii 1,750 eye tracker at
50 Hz. Data were parsed into fixations defined as “gaze
remaining within a 30 pixel radius (~.80° of visual angle)
for a minimum of 60 ms.” A list of fixations, in the order in
which they occurred and their corresponding AOI, was
extracted for each participant and used to compute visit
durations and order during each trial. A visit was defined as
the sum of all consecutive fixations made within an AOI.
Each visit was coded as either a first visit (the first visit to
that AOI, coded as 0) or a revisit (the AOI had been visited
before within the trial, coded as 1). The probability of
choosing an old item was always zero at the first visit as
none of the other items had been fixated at this point within
the trial. Likewise, the probability of choosing an old item
at the second visit was always zero because all of the pos-
sible target AOIs were new items. We coded up to 10 visits
per trial. For each participant, we calculated the revisitation
likelihood for each visit, by averaging across data obtained
from all slides.
Younger Sibling Outcome Characterization
A standard measure of mental development level, the
Mullen Scales for Early Learning (MSEL; Mullen, 1995),
was collected. The MSEL is a standardized direct develop-
mental assessment that yields a standardized score (M =
100, SD = 15) of overall intellectual ability (Early Learning
Composite standard score). The HR group was also
assessed with the parent-report Autism Diagnostic
Interview–Revised (ADI-R; Lord et al., 1994) and the
Autism Diagnosis Observation Schedule (ADOS; Lord
et al., 2000). A “best estimate clinical consensus” approach
to diagnosis was taken following a review by experienced
clinical researchers (TC) taking account of all information
about the child (i.e., MSEL, informal observation), in addi-
tion to information from the ADI-R and ADOS-G. Children
were included in the ASD group if they met International
Classification of Diseases–Revision 10 (ICD-10; World
Health Organization, 1993) criteria for any pervasive
developmental disorder (PDD). Given the young age of the
children, and in line with the changes to Diagnostic and
Statistical Manual of Mental Disorders (5th ed.; DSM-5;
American Psychiatric Association, 2013), no attempt was
made to assign specific subcategories of PDD/childhood
autism diagnosis. Children from the HR group were con-
sidered TD (high-risk-typical) if they (a) did not meet ICD-
10 criteria for an ASD, (b) did not score above the ASD
cutoff on the ADOS or ADI, and (c) scored within 1.5 SD
Table 1. Participant Characteristics.
LR HR/ASD-HI HR/ASD+HI
F:M 29:21 16:10 10:10
Time 1 (8 months)
Age months (SD) 7.87 (1.1) 8.01 (1.1) 7.56 (1.4)
Trial no. (min-max) 13.83 (12-14) 13.85 (13-14) 13.17 (4-14)
N148 26 18
Mullen ELC2 (SD) 104.42 (11.31)a,b 91.30 (11.21)a95.21 (14.24)b
Time 2 (14 months)
Age months (SD) 13.91 (3.1) 14.48 (1.2) 13.80 (1.6)
Trial no. (min-max) 13.67 (4-14) 13.71 (11-14) 13.94 (13-14)
N145 24 18
Mullen ELC (SD) 106.10 (15.72) 99.07 (17.31) 98.20 (18.94)
Time 3 (3 years)
Age months (SD) 38.22 (3.05) 37.42 (1.4) 37.20 (1.7)
Trial no. (min-max) 9.60 (7-10) 9.95 (5-14) 10 (10-10)
N126 20 13
Mullen ELC2 (SD) 115.79 (16.27)a100.92 (22.61)a111.94 (22.61)
ADOS SC — 9.42 (5.2) 7.40 (5.3)
Note. Superscript letters indicate significant paired comparisons (Bonferroni correction, p < .05). LR = low familial risk; HR = high familial risk;
ASD = autism spectrum disorder; HI = hyperactivity/inattention; Mullen ELC = Early Learning Composite standard scores; ADOS SC = Autism Diag-
nostic Observation Scale, Social Interaction and Communication Composite.
1Number of participants contributing eye-tracking data at each age.
2Indicates a main effect of group.
by guest on December 5, 2015jad.sagepub.comDownloaded from

Gliga et al. 5
of the population mean on the Mullen Early Learning
Composite (ELC) score (>77.5) and Receptive Language
(RL) and Expressive Language (EL) subscale T-scores
(>35). Children from the HR group were considered to
have atypical development if they did not fall into any of
the above groups. That is, they either scored above the
ADOS or ADI cutoff for ASD or scored <1.5 SD on the
Mullen ELC or RL and EL, but did not meet ICD-10 crite-
ria for an ASD. From the 47 at-risk participants taking part
in this task, 17 met criteria for an ASD diagnosis, 18 were
at-risk typical, and 12 were in the HR-atypical group (9
scoring above the ADOS ASD cutoff, 1 scoring above the
ADOS ASD cutoff and <1.5 SD Mullen ELC cutoff, 1
scoring above the ADI ASD cutoff, and 1 scoring <1.5 SD
Mullen ELC cutoff).
Results
Preliminary Analyses
We computed the average number of AOI visits per trial for
each age group. On average, participants contributed 7 vis-
its per slide at 8 months (min average 3, max average 9.9),
8 visits at 14 months (min 2.42, max 10), and 9 visits at 3
years (min 6.8 and max 10). Thus, to avoid data loss, we
restricted the analysis to the first three visits, which meant
restricting the analysis of revisitation likelihood to Visit 3
(as the likelihood of revisitation is zero at the first and sec-
ond visits). Given that at the third visit, all previous visits
had been on new AOIs, the one other advantage of this mea-
sure is that it is not affected by visit history (and by poten-
tial group differences in visit history). Because at both 8
months and 3 years, the HR/ASD-HI group had the lowest
scores on the Mullen Scales of Early Learning of the three
groups, (Table 1), we investigated the relationships between
revisitation likelihood and concurrent Mullen scores. None
of these relationships reached significance (see Suplemental
Online Material).
Does ASD and HI Risk Influence Visual
Foraging?
We carried out a repeated measures ANOVA on likelihood
of revisitation at the third visit, with age (8, 14 months, and
3 years) and group (24 LR, 20 HR/ASD-HI, 12 HR/
ASD+HI). This yielded a main effect of age, F(2, 106) =
17.97, p < .001, as well as a significant effect of group, F(2,
53) = 3.81, p = .028. With age, all groups became more
exploratory, in the sense that they were increasingly more
likely to visit a new AOI rather than returning to a previ-
ously seen one. Bonferroni-corrected post hoc t tests
revealed that LR participants were less likely to revisit than
the HR/ASD-HI group (p = .051), while the HR/ASD+HI
group was not significantly different from either the HR/
ASD-HI participants (p = .091) or the LR participants. We
followed up on a significant Age × Group interaction, F(4,
106) = 3.40, p = .012, with three repeated measures
ANOVAs for each age (Figure 2). At 8 months, the effect of
group was significant, F(2, 95) = 7.39, p = .001, HR/
ASD-HI were significantly different from both LR (p =
.009) and HR/ASD+HI (p = .001), the last two being indis-
tinguishable statistically. Likelihood of revisitation was
above chance level (.25) in all groups, LR: t(47) = 7.07,
p < .001; HR/ASD-HI: t(25) = 10.54, p < .001; HR/
ASD+HI: t(17) = 3.17, p = .006. At 14 months, the effect of
group was again significant, F(2, 86) = 4.65, p = .012; only
LR and HR/ASD-HI were significantly different from each
other (p = .012). Likelihood of revisitation was above
chance level in all groups, LR: t(43) = 3.17, p = .003; HR/
ASD-HI: t(23) = 4.53, p < .001; HR/ASD+HI: t(17) = 5.41,
p < .001. Groups did not significantly differ at 3 years and
for all groups, the likelihood of revisitation was not
Figure 2. Likelihood of revisitation at all ages depending on family risk; bars represent 1 Standard Error of the Mean; dimensional
relationship between revisit likelihood at 8 months and proband hyperactivity/inattention score.
Note. HR = high familial risk; ASD = autism spectrum disorder; HI = hyperactivity/inattention.
by guest on December 5, 2015jad.sagepub.comDownloaded from

6 Journal of Attention Disorders
different from chance, LR: t(25) < 1; HR/ASD-HI: t(19) <
1; HR/ASD+HI: t(13) < 1.
Given that a dimensional measure of HI risk was avail-
able (the SDQ scores vary between 0 and 10), and to
strengthen our findings, we investigated the quantitative
relationship between revisitation likelihood at 8 and 14
months and this measure. This analysis yielded a significant
negative correlation between revisitation likelihood at 8
months and SDQ scores (r = −.430, p = .003; Figure 2) but
no significant relationship with visual foraging at 14
months.
Were Group Differences Driven by Returns to
the Face?
As many revisits at the third visit are returns to the face
(about 50% of first looks are to the face, Elsabbagh et al.,
2013), it is important to know whether the group differences
described mainly reflect differences in returns to this stimu-
lus. To ask this question, we ran a repeated measures
ANOVA with type (Face, non-Face) and group (LR, HR/
ASD-HI, HR/ASD+HI), at each age point. At both 8 and 14
months, this analysis yielded significant main effects of
type, 8 months: F(1, 88) = 42.35, p < .001; 14 months: F(1,
83) = 23.14, p < .001, and group, 8 months: F(2, 88) = 3.64,
p = .030; 14 months: F(2, 83) = 3.47, p = .036, but a nonsig-
nificant interaction of Type × Group, 8 months: F(2, 88) =
2.06, p > .1; 14 months: F(2, 83) < 1, which suggests that
group differences in revisitation likelihood concern both
types of images. At 3 years, the effects of type and group
were not significant, type: F(1, 55) = 2.84, p > .05; group:
F(2, 55) < 1, but there was a marginally significant interac-
tion between type and group, F(2, 55) = 2.99, p = .058.
However, follow-up paired t tests yielded no significant
effects of type.
Is Foraging Related to the Child’s ASD
Diagnosis?
As ASD diagnosis was available for the younger sibling, we
also ran a repeated-measures ANOVA on the revisitation
likelihood, with age as a within participant variable and out-
come (LR, HR-TD, HR-Atypical, and HR-ASD) as between-
participants factor. This yielded a significant interaction
between age and outcome, F(6, 108) = 2.43, p = .030, and a
nonsignificant effect of outcome, F(3, 54) < 1. We followed
up this analysis with three univariate ANOVAs for each age
point. Although HR-ASD did have the highest likelihood of
revisitation at 8 months, outcome was not a significant pre-
dictor, F(3, 102) = 2.23, p = .089 (see Supplemental Figure
S1). Follow-up paired t tests against the LR group (Dunnett,
1955) yielded marginally significant effects (LR vs.
HR-ASD, p = .084). At 14 months, there was a significant
effect of outcome, F(3, 92) = 2.78, p = .045, but none of the
follow-up paired t tests were significant. The effect of out-
come was not significant at 3 years.
To explore whether proband HI affected all outcome
groups, at 8 months a second repeated measures ANOVA
was run on the high-risk group only, this time with both
outcome (HR-TD, HR-atypical and HR-ASD) and HI level
(low, high) as between-participants variables. This yielded
a significant effect of HI level, F(1, 43) = 10.51, p = .002,
which did not interact with outcome, F(2, 43) < 1 (see
Supplemental Figure S2). Thus, high proband HI was asso-
ciated with stronger exploratory biases in all high-risk chil-
dren, independent of their ASD diagnosis.
Discussion
Borrowing the foraging framework developed for appeti-
tive decision-making in animals (Cohen et al., 2007;
Kacelnik et al., 1981), we investigated the development of
exploratory biases in visual information foraging. We
observed age-related changes. Six-month-olds showed a
strong tendency to revisit previously seen AOIs, with 40%
of their third visits being revisits. In contrast, by 3 years of
age, when faced with a new visual scene, children randomly
choose whether to explore a new AOI or return to a previ-
ously seen one. Thus, although we observe a decrease with
age in exploitative choices (i.e., preferring to return to old
AOIs), not even at 3 years of age does foraging become
driven by exploration (i.e., consistently preferring to sam-
ple new AOIs). The main question that motivated this
research was whether background family risk for ASD and
hyperactivity and inattention (core symptoms of ADHD)
impact on visual foraging early in life. In terms of the likeli-
hoods of revisitation, the HR/ASD-HI group was more
likely to revisit old locations than were LR participants, at
both 8 and 14 months. When the sibling’s own ASD diagno-
sis was taken into account, a trend was found for higher
revisitation likelihood in 8-month-old HR-ASDs, compared
with LR controls. This is reminiscent of previous findings
in children with a diagnosis of ASD, who also show high
return rates when exploring visual scenes or real environ-
ments (Elison et al., 2012; Pellicano et al., 2011; Pierce &
Courchesne, 2001).
We also predicted that proband hyperactivity/inattention
would be associated with a bias toward exploration. Indeed,
at 8 months, likelihood to revisit AOIs inversely relates to
proband SDQ hyperactivity and inattention scores.
However, when groups are compared, HR/ASD+HI partici-
pant performance was similar to that of the control group at
8 months and was not different from either the HR/ASD-HI
or LR at 14 months. Our findings are compatible with an
additive effect, where ASD and ADHD risk contribute
opposing biases. Additive effects of dual ASD and ADHD
diagnosis have been previously documented, but more often
then not, the literature reports increased symptom severity
by guest on December 5, 2015jad.sagepub.comDownloaded from

Gliga et al. 7
in children with a dual diagnosis of ASD and ADHD (e.g.,
Goldin et al., 2013; Yerys et al., 2009). This is the first evi-
dence for a potential moderating effect of one risk type on
another. However, because we are assessing the effects of
co-occurring ASD and HI risk and not those of co-occur-
rence of these symptoms in a particular individual, we are
unable to say at this point whether the “typical” perfor-
mance of the HR/ASD+HI group reflects an interaction
between the two risk backgrounds or whether it results from
some children having inherited the ASD risk factors and
other HI risk factors. Some indication that the former
hypothesis might be true comes from the fact that HR/
ASD+HI performance variance is comparable to those of
the other two groups, Levene’s F(2, 89) = 1.17, p > .1
(Supplemental Figure S1). It is important to highlight that
whether co-occurring conditions result in additive, multipli-
cative, or a completely new phenotype may vary from one
phenotypic trait to another and therefore the current find-
ings should not be generalized to imply that co-occurring
risk for both ASD and ADHD might be overall beneficial.
Why are the effects of proband hyperactivity/inattention
more prominent earlier in development? Transitory signa-
tures of risk have been reported previously. For example,
9-month-old performance in a visual search task is a predic-
tor of later ASD symptoms, but the performance of
15-month-olds does not relate to later outcome (Gliga,
Bedford, Charman, Johnson, & the BASIS Team, 2015).
Similarly, the amount of looking toward faces is a predictor
of later face recognition when measured at 8 months, but
not at 14 months (deKlerk, Gliga, Charman, & Johnson,
2014). It is possible that these developmental changes
reflect adaptive mechanisms following initial perturbations
in brain functioning, and that eventually mask the perturba-
tions, but not before development has been set on an atypi-
cal pathway (Johnson, Jones, & Gliga, 2015).
Our findings are also relevant for the debate surrounding
the origin of ASD and ADHD. Some scholars have sug-
gested a common developmental origin, which would also
explain the high rates of comorbidity (Rommelse, Geurts,
Franke, Buitelaar, & Hartman, 2011), but developmental
studies have revealed limited overlap in terms of early
markers for these disorders (Johnson, Gliga, Jones, &
Charman, 2015). The current findings may add to this list
and suggest an alternative view, according to which comor-
bidity may result from the expression of a few general vul-
nerability factors, such as poor executive functions
(Johnson, 2012). However, it remains possible that our find-
ings do not reflect heritable genetic background but envi-
ronmental effects. Growing up alongside a sibling with a
developmental disorder may in itself affect development.
Proband hyperactivity symptoms may be more perturbing
when the sibling is younger, thus explaining the stronger
relationship found at 8 months of age. Despite the uncer-
tainty regarding the mechanisms behind the effects we
document here, these findings remain novel and important
as they add to our understanding of both early learning and
psychopathology. How infants explore information has
long-term consequences on their learning abilities
(Bornstein et al., 2013, and see SOM). We also reveal para-
doxical mechanisms for resilience, where high levels of
hyperactivity/inattention in older siblings, through shared
genetic or environmental mechanisms, can have paradoxi-
cal effects, being associated with optimal information for-
aging in infants at-risk for ASD.
Author Contributions
T.G. designed the eye-tracking study, T.G., T.S., and N.G. ana-
lyzed the data; T.G. and T.S. wrote the article with contribution
from M.H.J. and T.C. M.H.J. and T.C. led the BASIS program.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect
to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for
the research, authorship, and/or publication of this article: This work
was supported by Medical Research Council (MRC) Program Grant
G0701484 and the BASIS funding consortium led by Autistica
(www.basisnetwork.org).
References
American Psychiatric Association. (2013). Diagnostic and sta-
tistical manual of mental disorders (5th ed.). Arlington, VA:
American Psychiatric Publishing.
Biederman, J., & Spencer, T. (1999). Attention-deficit/hyperactiv-
ity disorder (ADHD) as a noradrenergic disorder. Biological
Psychiatry, 46, 1234-1242.
Blaser, E., Eglington, L., Carter, A. S., & Kaldy, Z. (2014).
Pupillometry reveals a mechanism for the autism spectrum
disorder (ASD) advantage in visual tasks. Scientific Reports,
4, Article 4301.
Bornstein, M. H., Hahn, C. S., & Suwalsky, J. T. (2013). Physically
developed and exploratory young infants contribute to their
own long-term academic achievement. Psychological sci-
ence, 24(10), 1906.
Bronson, G. (1991). Infant differences in the rate of visual encod-
ing. Child Development, 62, 44-54.
Carballo, J. J., Rodríguez-Blanco, L., García-Nieto, R., &
Baca-García, E. (2014). Screening for the ADHD pheno-
type using the Strengths and Difficulties Questionnaire in a
clinical sample of newly referred children and adolescents.
Journal of Attention Disorders. Advance online publication.
doi:10.1177/1087054714561858
Chen, C., Burton, M., Greenberger, E., & Dmitrieva, J. (1999).
Population migration and the variation of dopamine D4 recep-
tor (DRD4) allele frequencies around the globe. Evolution &
Human Behavior, 20, 309-324.
Cohen, J. D., McClure, S. M., & Yu, A. J. (2007). Should I
stay or should I go? How the human brain manages the
by guest on December 5, 2015jad.sagepub.comDownloaded from

8 Journal of Attention Disorders
trade-off between exploitation and exploration. Philosophical
Transactions of the Royal Society B: Biological Sciences,
362, 933-942.
Costa, V. D., Tran, V. L., Turchi, J., & Averbeck, B. B. (2014).
Dopamine modulates novelty seeking behavior during deci-
sion making. Behavioral Neuroscience, 128, 556-566.
deKlerk, C. C., Gliga, T., Charman, T., & Johnson, M. H. (2014).
Face engagement during infancy predicts later face recog-
nition ability in younger siblings of children with autism.
Developmental Science, 17, 596-611.
Dunnett, C. W. (1955). A multiple comparison procedure for
comparing several treatments with a control. Journal of the
American Statistical Association, 50, 1096-1121.
Elison, J. T., Sasson, N. J., Turner-Brown, L. M., Dichter, G. S.,
& Bodfish, J. W. (2012). Age trends in visual exploration
of social and nonsocial information in children with autism.
Research in Autism Spectrum Disorders, 6, 842-851.
Elsabbagh, M., Gliga, T., Pickles, A., Hudry, K., Charman, T.,
& Johnson, M. H., & the BASIS Team. (2013). The devel-
opment of face orienting mechanisms in infants at-risk for
autism. Behavioural Brain Research, 251, 147-154.
Fischer, T., Graupner, S. T., Velichkovsky, B. M., & Pannasch,
S. (2013). Attentional dynamics during free picture viewing:
Evidence from oculomotor behavior and electrocortical activ-
ity. Frontiers in Systems Neuroscience, 7, Article 17.
Frank, M. J., Doll, B. B., Oas-Terpstra, J., & Moreno, F. (2009).
Prefrontal and striatal dopaminergic genes predict indi-
vidual differences in exploration and exploitation. Nature
Neuroscience, 12, 1062-1068.
Gliga, T., Bedford, R., Charman, T., & Johnson, M. H., & the BASIS
Team. (2015). Enhanced visual search in infancy predicts
emerging autism symptoms. Current Biology, 25, 1727-1730.
Goldin, R. L., Matson, J. L., Tureck, K., Cervantes, P. E., &
Jang, J. (2013). A comparison of tantrum behavior profiles in
children with ASD, ADHD and comorbid ASD and ADHD.
Research in Developmental Disabilities, 34, 2669-2675.
Hills, T. T., Todd, P. M., & Goldstone, R. L. (2008). Search in
external and internal spaces evidence for generalized cogni-
tive search processes. Psychological Science, 19, 802-808.
Jensen, P. S., Mrazek, D., Knapp, P. K., Steinberg, L., Pfeffer, C.,
Schowalter, J., & Shapiro, T. (1997). Evolution and revolu-
tion in child psychiatry: ADHD as a disorder of adaptation.
Journal of the American Academy of Child & Adolescent
Psychiatry, 36, 1672-1681.
Jepma, M., & Nieuwenhuis, S. (2011). Pupil diameter pre-
dicts changes in the exploration–exploitation trade-off:
Evidence for the adaptive gain theory. Journal of Cognitive
Neuroscience, 23, 1587-1596.
Johnson, M. H. (2012). Executive function and developmental dis-
orders: The flip side of the coin. Trends in Cognitive Sciences,
16, 454-457.
Johnson, M. H., Gliga, T., Jones, E., & Charman, T. (2015).
Annual research review: Infant development, autism, and
ADHD—Early pathways to emerging disorders. Journal of
Child Psychology and Psychiatry, 56, 228-247.
Johnson, M. H., Jones, E., & Gliga, T. (2015). Brain adaptation
and alternative developmental trajectories. Development and
Psychopathology, 27, 425-442.
Kacelnik, A., Houston, A. I., & Krebs, J. R. (1981). Optimal for-
aging and territorial defense in the great tit (Parus major).
Behavioral Ecology and Sociobiology, 8, 35-40.
Krebs, J. R., Kacelnik, A., & Taylor, P. (1978). Test of optimal
sampling by foraging great tits. Nature, 275, 27-31.
Kriete, T., & Noelle, D. C. (2015). Dopamine and the develop-
ment of executive dysfunction in autism spectrum disorders.
PLoS ONE, 10(3), e0121605.
Larsson, H., Chang, Z., D’Onofrio, B. M., & Lichtenstein, P.
(2014). The heritability of clinically diagnosed attention defi-
cit hyperactivity disorder across the lifespan. Psychological
Medicine, 44, 2223-2229.
Lord, C., Rutter, M., & Le Couteur, A. (1994). Autism Diagnostic
Interview-Revised: a revised version of a diagnostic interview
for caregivers of individuals with possible pervasive develop-
mental disorders. Journal of autism and developmental disor-
ders, 24(5), 659-685.
Lord, C., Risi, S., Lambrecht, L., Cook, E. H., Jr., Leventhal,
B. L., DiLavore, P. C., & Rutter, M. (2000). The Autism
Diagnostic Observation Schedule—Generic: A standard mea-
sure of social and communication deficits associated with the
spectrum of autism. Journal of Autism and Developmental
Disorders, 30, 205-223.
Matthews, L. J., & Butler, P. M. (2011). Novelty-seeking DRD4
polymorphisms are associated with human migration distance
out-of-Africa after controlling for neutral population gene
structure. American Journal of Physical Anthropology, 145,
382-389.
Messinger, D., Young, G. S., Ozonoff, S., Dobkins, K., Carter, A.,
Zwaigenbaum, L., & Sigman, M. (2013). Beyond autism: A
Baby Siblings Research Consortium study of high-risk chil-
dren at three years of age. Journal of the American Academy
of Child & Adolescent Psychiatry, 52, 300-308.
Mullen, E. (1995). Mullen Scales of Early Learning (AGS ed.).
Bloomington, MN: Pearson Assessments.
Ozonoff, S., Young, G. S., Carter, A., Messinger, D., Yirmiya,
N., Zwaigenbaum, L., & Stone, W. L. (2011). Recurrence
risk for autism spectrum disorders: A Baby Siblings Research
Consortium study. Pediatrics, 128, e488-e495.
Pannasch, S., Helmert, J. R., Roth, K., Herbold, A. K., & Walter,
H. (2008). Visual fixation durations and saccade amplitudes:
Shifting relationship in a variety of conditions. Journal of Eye
Movement Research, 2(2), Article 4.
Pellicano, E., Smith, A. D., Cristino, F., Hood, B. M., Briscoe,
J., & Gilchrist, I. D. (2011). Children with autism are neither
systematic nor optimal foragers. Proceedings of the National
Academy of Sciences, 108, 421-426.
Pierce, K., & Courchesne, E. (2001). Evidence for a cerebellar role
in reduced exploration and stereotyped behavior in autism.
Biological Psychiatry, 49, 655-664.
Risko, E. F., Anderson, N. C., Lanthier, S., & Kingstone, A.
(2012). Curious eyes: Individual differences in personality
predict eye movement behavior in scene-viewing. Cognition,
122, 86-90.
Rommelse, N. N., Geurts, H. M., Franke, B., Buitelaar, J. K., &
Hartman, C. A. (2011). A review on cognitive and brain endo-
phenotypes that may be common in autism spectrum disorder
and attention-deficit/hyperactivity disorder and facilitate the
by guest on December 5, 2015jad.sagepub.comDownloaded from

Gliga et al. 9
search for pleiotropic genes. Neuroscience & Biobehavioral
Reviews, 35, 1363-1396.
Russell, G., Rodgers, L. R., Ukoumunne, O. C., & Ford, T. (2014).
Prevalence of parent-reported ASD and ADHD in the UK:
Findings from the Millennium Cohort Study. Journal of
Autism and Developmental Disorders, 44, 31-40.
Rutter, M., Bailey, A., & Lord, C. (2003). The Social
Communication Questionnaire (SCQ). Torrance, CA:
Western Psychological Services.
Salgado, C. A., Bau, C. H., Grevet, E. H., Fischer, A. G., Victor, M.
M., Kalil, K. L., & Belmonte-de-Abreu, P. (2008). Inattention
and hyperactivity dimensions of ADHD are associated with
different personality profiles. Psychopathology, 42, 108-112.
Schlesinger, M., & Amso, D. (2013). Image free-viewing as intrin-
sically-motivated exploration: estimating the learnability of
center-of-gaze image samples in infants and adults. Frontiers
in psychology. Epub ahead of print 31 October 2015. DOI:
10.3389/fpsyg.2013.00802.
Solanto, M. V. (2002). Dopamine dysfunction in AD/HD: Integrating
clinical and basic neuroscience research. Behavioural Brain
Research, 130, 65-71.
Tye, C., Mercure, E., Ashwood, K. L., Azadi, B., Asherson, P.,
Johnson, M. H., & McLoughlin, G. (2013). Neurophysiological
responses to faces and gaze direction differentiate chil-
dren with ASD, ADHD and ASD+ADHD. Developmental
Cognitive Neuroscience, 5, 71-85.
Ullebø, A. K., Posserud, M. B., Heiervang, E., Gillberg, C., & Obel,
C. (2011). Screening for the attention deficit hyperactivity disor-
der phenotype using the Strength and Difficulties Questionnaire.
European Child & Adolescent Psychiatry, 20, 451-458.
Williams, J., & Taylor, E. (2006). The evolution of hyperactiv-
ity, impulsivity and cognitive diversity. Journal of the Royal
Society Interface, 3, 399-413.
World Health Organization. (1993). The ICD-10 classification of
mental and behavioural disorders: Diagnostic criteria for
research. Geneva, Switzerland: Author.
Yerys, B. E., Wallace, G. L., Sokoloff, J. L., Shook, D. A., James,
J. D., & Kenworthy, L. (2009). Attention deficit/hyperactiv-
ity disorder symptoms moderate cognition and behavior in
children with autism spectrum disorders. Autism Research, 2,
322-333.
Author Biographies
Teodora Gliga is Programme Leader for the Infat Sibling Studies
at the Centre for Brain and Cognitive Development, Birkbeck,
University of London.
Tim J. Smith is a Senior Lecturer in Psychological Sciences at
Birkbeck, University of London.
Noreen Likely is a Trainee Clinical Psychologist at the University
of Limerick, Ireland. She contributed to the current work as part of
her Masters degree, at Birkbeck, University of London.
Tony Charman holds the Chair in Clinical Child Psychology in
Clinical Child Psychiatry at the Institute of Psychiatry, Psychology
& Neuroscience, King’s College London.
Mark H. Johnson is an MRC Programme Leader and Director of
the Centre for Brain and Cognitive Development, Birkbeck,
University of London.
by guest on December 5, 2015jad.sagepub.comDownloaded from