University of Southern California Dissertations and Theses

Neural and behavioral correlates of fear processing in first-time fathers

(USC Thesis Other)

Neural and behavioral correlates of fear processing in first-time fathers

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

Neural and Behavioral Correlates of Fear Processing in First-Time Fathers
by
Sarah Ann Stoycos, M.A.

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

August 2021

Copyright 2021 Sarah Ann Stoycos
ii
ACKNOWLEDGEMENTS
Financial support for the studies detailed in this project was provided by NSF CAREER
Award 1552452 (PI: Saxbe).

First and foremost, I want to thank my advisor, Dr. Darby Saxbe, for your mentorship
and encouragement. I feel very fortunate to have been with you during the formation of the
NEST Lab neuroimaging substudy and immensely privileged that you were willing to add in
measures so that I could pursue my interests of studying emotion perception and empathy in
special populations. Thank you for being with and supporting me through this journey—the ups
and downs, to Europe and back, and all the missed jokes in between.

I would like to thank my committee members, Drs. Darby Saxbe, Gayla Margolin, John
Monterosso, Jonas Kaplan, and Lisa Aziz-Zadeh, for their kindness in offering their time and
effort to reviewing and providing feedback for this project. I would also like to thank the many
NeuroEndocrinology of Social Ties Laboratory research assistants, lab managers, and graduate
students who helped make this project possible with their countless hours spent running visits
and managing and collecting data. I specifically would like to thank Dr. Hannah Khoddam, Dr.
Diane Goldenberg, Geoffrey Corner, Mona Khaled, Narcis Marshall, Alyssa Morris, Sofia
Cárdenas, Katelyn Horton, Nia Barbee, Ann Bryna Tsai, and Pia Sellery. I would also like to
thank the generous couples and infants who were willing to welcome us researchers into an
intimate period of their life to benefit science.

Most importantly, I am grateful for the strong women in my life that have mentored,
supported, believed in, and held me up throughout my life and on this journey. To Margie,
Lindsey, AJ, Kathy, and Anna—your grounding presence, strength, and unwavering belief in me
helped me remain authentic and in touch with what matters most to me in this life. Thank you
does not even begin to cover it. I would like to thank Gretchen Abell and Kelley O’Connell, for
instilling in me from a young age a work ethic and drive to pursue excellence in all that I do. I’d
like to thank Meredith, Steph, Natasha, Kelly, Tamar, Caroline and Eloise, for your friendship
and ongoing support.

Finally, I am exceptionally grateful for my wonderful and supportive partner, Kurt
Klinke, for staying by my side throughout this journey and encouraging me to persist when I
needed it most.

iii
TABLE OF CONTENTS
Acknowledgements ......................................................................................................................... ii
List of Tables ................................................................................................................................. iv
List of Figures ..................................................................................................................................v
Abstract .......................................................................................................................................... vi
INTRODUCTION ...........................................................................................................................1
Alloparenting and fathers .....................................................................................................2
Paternal brain as overlapping, yet distinct from the maternal brain ....................................3
Current Studies .....................................................................................................................4

STUDY 1: EXAMINING APPROACH-AVOIDANCE RESPONSES TO FEAR AND INFANT
STIMULI AND THEIR RELATION TO SELF-REPORTED EMPATHY IN FIRST-TIME
FATHERS ........................................................................................................................................6
Method ...............................................................................................................................13
Results ................................................................................................................................18
Discussion ..........................................................................................................................22

STUDY 2: NEURAL AND BEHAVIORAL CORRELATES OF FEAR PROCESSING IN
MEN TRANSITIONING TO PARENTHOOD ............................................................................27
Method ...............................................................................................................................35
Results ................................................................................................................................43
Discussion ..........................................................................................................................53

GENERAL DISCUSSION ............................................................................................................60
Conclusion .........................................................................................................................63

REFERENCES ..............................................................................................................................65

iv
LIST OF TABLES

Table 1.1 Study 1 Participant Characteristics ................................................................................14
Table 1.2 Example of IAT Condition Presentation .......................................................................17
Table 1.3 Mean Response Times to Face Stimuli During Congruent and Incongruent Trials ......20
Table 2.1 Study 2 Participant Characteristics ................................................................................36
Table 2.2 Peak MRI Activation Coordinates .................................................................................49
Table 2.3 Results from Multiple Linear Regression Models Predicting Changes in Fear
Recognition ....................................................................................................................................51

v
LIST OF FIGURES

Figure 1.1 Example of Kindchenschema .........................................................................................9
Figure 1.2 IAT Effects ...................................................................................................................19
Figure 1.3 Approach-Avoidance Response Times by Condition and Congruency .......................21
Figure 1.4 Approach-Avoidance Responses and Self-Reported Empathy ....................................22
Figure 2.1 fMRI Emotion Perception Task ....................................................................................39
Figure 2.2 fMRI Implicit Faces Behavioral Data for Fear .............................................................45
Figure 2.3 Prenatal Whole Brain Results .......................................................................................46
Figure 2.4 Prenatal PAG Activation and Fear Recognition Accuracy ..........................................47
Figure 2.5 Postpartum Whole Brain Results ..................................................................................48
Figure 2.6 Longitudinal Whole Brain Results ...............................................................................49
Figure 2.7 Prenatal PAG % Signal Change Predicts Change in Fear Recognition ........................51

vi
NEURAL AND BEHAVIORAL CORRELATES OF FEAR PROCESSING IN FIRST-TIME
FATHERS
Sarah A. Stoycos, M.A.
Thesis Advisor: Darby Saxbe, Ph.D.
ABSTRACT

Alloparenting, or parenting by anyone other than the biological mother, is relatively
understudied in the transition to parenthood literature despite the prevalence of alloparenting in
cooperative societies by grandparents, fathers, and close others. Studying alloparenting and
fathering presents an opportunity to understand how investment in childrearing may impact
adults’ behavioral and neural processing of emotion and other processes pivotal for caregiving.
This dissertation examines the behavioral and neural correlates of face emotion perception in
males transitioning to fatherhood, with a specific focus on whether studying adult distress cues
(i.e., fear faces) may provide valuable insights and nuance to understanding alloparental care and
the transition to fatherhood. Both Paper 1 and Paper 2 present data from a longitudinal study of
first-time fathers. In Paper 1, expectant fathers completed a combined approach-avoidance
implicit association test (IAT) designed to examine whether the adult fear face is more closely
associated with the infant neutral face than with other adult faces and to measure approach-
withdrawal action tendencies in response to these faces. This paper also examined associations
with self-reported empathic concern and perspective-taking. Results indicated that expectant
fathers responded faster when adult fear faces were paired with infant faces versus adult faces.
Relatedly, we found that infant faces elicited behavioral approach on an arm flexion task,
whereas responses to adult fear faces were context-dependent, such that fear faces elicited faster
behavioral approach when paired with infant faces, but faster behavioral avoidance when paired
vii
with other adult faces. Additionally, perspective-taking was positively associated with behavioral
approach responses for fear stimuli, but not infant stimuli, and empathic concern was positively
associated with all responses to infant stimuli, but not fear stimuli. Paper 2 is a longitudinal
examination of the neural and behavioral correlates of fear processing in fathers before and after
the birth of their first child and aimed to understand whether emotion perception changes with
the transition to fatherhood. We found preliminary support that this transition is accompanied by
neural changes that may be associated with inhibition of fear learning pathways and excitation of
social reinforcement learning and felt reward in response to fear faces. To our knowledge, this is
the first longitudinal investigation of the neural and behavioral characteristics of men before and
after the birth of their first child and the first paternal brain neuroimaging study to use non-infant
cues. This dissertation is the first examination of the neural and behavioral correlates of face
emotion processing in males before and after the birth of their first child and implications of
these findings, strengths and limitations, and future directions are discussed.

INTRODUCTION

“A father’s brain provides a model for great neural plasticity driven by acts of committed
daily caregiving that occur without the hormonal changes associated with pregnancy and
childbirth and are associated with father–child reciprocity. These neuroscientific findings may
give rise to the hope that, in a society in which adults are committed to allomothering
[alloparenting], the young may thrive and produce a more empathic, accepting and equal
society.”-Feldman, Braun, & Champagne, 2019

Over the last several decades, an emerging research literature on the human parental brain
has been enhanced by increasingly sophisticated neuroimaging approaches. The parenting brain
research has predominantly focused on the maternal brain and associated mother-child outcomes
but only recently has incorporated research on fathers (for review see Rilling, 2013; Leuner &
Sabihi, 2016 and Feldman, Braun, & Champagne, 2019; Lambert, 2012; Swain, Dayton, Kim,
Tolman, & Volling, 2014; Feldman, 2015, 2016; Numan & Young, 2016). The transition to
motherhood is accompanied by major changes in neural, behavioral, and emotional processing
that are likely subserved via neuroendocrine changes (Feldman, 2016). However, research on the
transition to fatherhood is nascent.
Recent literature has criticized the state of human parenting research, noting that many of
the parenting theories emerged from WEIRD (Western, Educated, Industrialized, Rich, and
Democratic) research at a time when the norm for parenting in Westernized cultures was heavily
dependent on the nuclear family with mother as caretaker and father as provider (Sear, 2016).
This led to an outgrowth of parenting theories and research almost exclusively focused on
mother-infant attachment and the transition to motherhood, which can be seen by the plethora of
research examining motherhood. However, more recently a culturally redefined perspective on
parenting has stressed the importance of humans as cooperative species with altricial young who
most often are raised with cooperative parenting by alloparents (Hrdy, 2009).
2
Alloparenting and fathers
Alloparenting refers to parenting by anyone other than the mother (Batson & Shaw,
1991) and is universal to the human experience (Sear, 2016; Hrdy, 2009). Alloparenting is
relatively understudied in the transition to parenthood literature despite the prevalence of
alloparenting in cooperative societies by grandparents, fathers, and close others. Alloparenting
provides a unique context to understand individual differences and mechanisms that impact
investment in child-rearing, and fathers offer a particularly interesting example given that
paternal investment is variable in humans. On the opposite direction of causality, studying
alloparenting and fathering also presents an opportunity to understand how investment in
childrearing may impact adults’ behavioral and neural processing of emotion and other processes
pivotal for caregiving, such as empathy (Parsons et al., 2019). Humans, more so than any other
species, show paternal engagement in offspring care in many (but not all) cases (Sear, 2016).
Only about 10% of mammalian species have males that engage in prolonged parental care,
making paternal investment in rearing offspring a highly specialized trait that is poorly
understood and likely more recently evolved than maternal care (Stockley & Hobson, 2016;
Clutton-Brock, 1989). It is well understood that human paternal involvement is linked with
increased survival rates of offspring and better social, educational, behavioral, and biological
outcomes for the offspring (Cabrera, Tamis-Lemonda, Bradley, Hofferth, & Lamb, 2000;
Sarkadi, Kristiansson, Oberklaid, & Bremberg, 2008). However, few studies have examined the
transition to fatherhood by assessing men before and after the birth of their first child.
Understanding mechanisms that subserve the successful transition to fatherhood and support
paternal investment in childrearing may inform later interventions targeting fathers at risk.
Furthermore, studying fathers as alloparents can extend research on motivated helping and
3
caregiving towards others and may contribute to broader literature including altruism and
empathy.
Paternal brain as overlapping, yet distinct from the maternal brain
Current research on the paternal brain and hormones indicates some overlap with the
maternal neural and hormonal networks; however, there also seem to be distinct neural and
hormonal characteristics of paternal care (Swain et al., 2014; Feldman, Braun, & Champagne,
2019). For example, Atzil, Hendler, Zagoory-Sharon, Winetraub, & Feldman, (2012) found that
mothers showed a positive correlation between oxytocin levels and neural activation within the
limbic system when undergoing a functional neuroimaging task contrasting images of their own
child versus an unknown child. However, fathers showed a negative correlation between
oxytocin and modulatory cortical brain areas. Fathers also appear to exhibit experience-
dependent changes, instead of primarily hormone-driven changes as seen in mothers, in neural
responses to infant stimuli, with primary-caregiver fathers showing different patterns of
activation than secondary-caregiver fathers (Abraham et al., 2014). Furthermore, a study of gray
matter changes in the postpartum period showed that mothers had reductions in gray matter
during the first two years postpartum, but fathers did not (Hoekzema et al., 2017). Despite these
potential differences in the paternal versus maternal brain, prevailing methodologies in the
paternal brain literature are often drawn directly from the maternal brain literature. Most studies
show own- and other- infant stimuli to postpartum fathers during functional neuroimaging, often
times with positive stimuli and passive viewing rather than specific task instructions. However, it
is not clear what mechanism this type of study design is elucidating. Comparison of own-infant
to other-infant stimuli may target preferential responding but may not tap into meaningful
distinctions in emotion processing. It may be useful to examine negatively valanced stimuli and
4
general emotion processing stimuli, beyond just infant stimuli (Parsons et al., 2019; Marsh,
2016). For example, empathy, as defined as an individual’s propensity to engage in motivated
and costly helping in response to other’s distress, is essential to successful engagement in
childrearing (Preston, 2013; Marsh, 2016) but has not been thoroughly explored in paternal brain
research.
Much of the transition to parenthood literature has been based on animal models.
Animals, rodents in particular, go through drastic changes from the prenatal to postpartum period
characterized by reorganization of avoidance or aggressive behavior towards young to that of
approach and provisional care (see Feldman, 2015 and Preston, 2013 for review). However, the
human transition to parenthood may be less delineated. Regardless of parenting status, humans
tend to show positive responses to infants (Senese et al., 2013), and parents compared to
nonparents show similar performance on tasks with infant stimuli (Parsons et al., 2017; Irwin,
2003). Given that positive and affiliative responses to infant cues are not specific to human
parents, the transition to parenthood may be harder to study and more nuanced in humans than in
animal models. Therefore, using a broader range of stimuli, beyond infant stimuli, may provide
useful insight into how the transition to parenthood is associated with adults’ socioemotional
functioning (Parsons et al., 2019). Despite this, most of the parenting brain literature has almost
exclusively used infant cues (vocalizations and infant facial stimuli). Incorporating non-infant
stimuli may be fruitful for understanding the neural bases of alloparental care, such as individual
differences in motivated and costly helping in response to distress.
Current studies
This dissertation examines the behavioral and neural correlates of emotion perception in
males transitioning to fatherhood, with a specific focus on whether studying adult distress cues
5
(i.e. fear faces) may provide valuable insights and nuance to understanding alloparental care.
First, given that most parenting brain research has used infant cues, we examined the relevance
of fear cues to infant cues by examining how closely associated the adult fear face may be to the
infant face in first-time fathers. The theoretical foundation for Paper 1 is based on Lorenz’ 1943
notion of kindchenschema or “baby schema,” which puts forth that typical features of the infant
face evolved specifically to elicit motivated caregiving from adults and the hypothesis that the
adult fear face morphologically capitalizes on kindchenschema. Expectant fathers completed a
combined approach-avoidance implicit association test (IAT) designed to examine whether the
adult fear face is more closely associated with the infant neutral face than with other adult faces
and to measure approach-withdrawal action tendencies in response to these faces. This paper also
examined whether self-reported empathic concern and perspective-taking, traits important to
fathering, were associated with approach-withdrawal patterns in response to adult and infant
faces. Paper 2 is a longitudinal examination of the neural and behavioral correlates of fear
processing in fathers before and after the birth of their first child aimed at understanding whether
emotion perception changes with the transition to fatherhood. To our knowledge, this is the first
longitudinal investigation of the neural and behavioral characteristics of men before and after the
birth of their first child and the first paternal brain neuroimaging study to use non-infant cues.

6

STUDY 1: EXAMINING APPROACH-AVOIDANCE RESPONSES TO FEAR AND
INFANT STIMULI AND THEIR RELATION TO SELF-REPORTED EMPATHY IN
FIRST-TIME FATHERS

Facial expressions are powerful means of communication that shape social behavior
(Willis, Windsor, Lawson, & Ridley, 2015; Marsh & Ambady, 2007). Infants cannot
communicate verbally, so the ability to perceive and respond to emotional facial expressions may
underlie effective parenting (Rimé, 2009; Rilling & Young, 2014; Rutherford, Wallace, Laurent,
& Mayes, 2015). Individual differences in caregiver responses to distress cues may therefore act
as a mechanism underlying caregiver-child bonding (Dykas & Cassidy, 2011; Feldman, 2012;
Bowlby, 1969). The ability to perceive and respond to distress cues may also be associated with
higher order sociocognitive processes like social cognition and empathy. The current study
examined how expectant fathers respond to neutral infant and adult threat-related facial
expressions in conjunction with their self-reported empathy.
The human parental brain literature has primarily used infant faces as stimuli during
functional neuroimaging tasks (Rilling, 2013). However, this literature has generally not
differentiated between distress and non-distress facial cues, a distinction framed as essential to
motivated caregiving by Preston (2013) and Marsh (2016). Most commonly, parents passively
view neutral or happy infant stimuli during functional neuroimaging. When viewing such images
of their own child versus an unfamiliar child, mothers exhibited activation in the
thalamocingulate pathway, insula, and amygdala (Leibenluft, Gobbini, Harrison, & Haxby, 2004;
Bartels & Zeki, 2004), which may suggest that mothers found their own child more salient than
an unfamiliar child (Rilling, 2013). Studies of both mothers and fathers viewing neutral infant
cues have also reported greater activation to own-infant than other-infant in the medial prefrontal
cortex, inferior frontal gyrus, inferior parietal lobule, insula, and anterior cingulate cortex
7
(Abraham et al., 2014) and also in reward and approach-related regions (i.e. ventral tegmental
area and orbitofrontal cortex; Atzil, Hendler, Zagoory-Sharon, Winetraub, & Feldman, 2012).
Therefore, there is support for increased responding to one’s own infant as opposed to another
infant when they are exhibiting neutral or positive cues, but these results may tell us more about
the saliency and liking of a particular infant rather than motivated approach behavior to care for
offspring.
Infant stimulus perception and action tendencies have also been studied in non-parents
and provided valuable insights into caregiving motivations (Caria et al., 2012; Aradhye, Vonk, &
Arisa, 2015; Glocker et al., 2009; Bornstein et al., 2008). Infants most commonly use gestures,
body movements, or intense vocal distress sounds to indicate threat responses, rather than facial
expressions of fear (Sullivan, 2014). Aradhye, Vonk, & Arisa (2015) examined non-parent
adults’ responses to infants smiling, crying, or exhibiting neutral cues and linked it to the adults’
nurturing motivation (i.e. likelihood to adopt) and the adults’ own emotional response (i.e.
experienced distress). They found that videos of infants smiling or exhibiting neutral cues were
rated as being cuter and more likely to be adopted, unlike videos of crying children, which
elicited more felt distress in the adult and lower likelihood of adoption (Aradhye, Vonk, & Arisa,
2015). Furthermore, there were no differences in empathic concern or perspective taking in
adult’s likelihood to adopt (Aradhye, Vonk, & Arisa, 2015). In nonparent adults, neutral infant
faces were linked with greater caregiving motivation (Glocker et al., 2009; Jia, Ding, & Cheng,
2021).
The communicative meaning of adult emotional facial expressions has also been studied
extensively (Seidel, Habel, Kirschner, Gur, & Derntl, 2010; Vrijsen, van Oostrom, Speckens,
Becker, & Rinck, 2013). For example, happiness is understood to elicit appetitive responses and
8
anger to elicit aversive responses. Both fear and anger are evolutionarily salient emotional
expressions because they may communicate potential threat, which would require immediate
action from the perceiver. For example, an angry facial expression indicates to the perceiver that
the person displaying emotion is a potential threat and aversive responding would be indicated.
Historically, fear has been associated with anger as indicative of threat and therefore, aversive
responding (Morris et al., 1996). However, in recent decades, evidence supports that fear’s
communicative meaning may be more nuanced (Taylor & Whalen, 2014). Unlike anger where
the source of the potential threat is clear, someone displaying fear is indicating there is a threat
somewhere, but the source and nature of the threat remains ambiguous. This then allows for
more flexible responding to fear versus anger, where it may be valuable to consider the larger
context to help determine the best course of action and whether the threat is a shared threat in the
environment, versus an internal threat in the displayer.
Along these lines, Hammer and Marsh (2014) posit that fear may actually serve to
mitigate threat or to elicit helping behavior from the perceiver because it is morphologically
similar to kindchenschema or “baby schema.” Kindchenschema was a term first coined by
Lorenz in 1943 to reflect infant features that are particularly salient and powerful elicitors of
caregiving motives in perceivers to thereby ensure the survival of altricial young (Lorenz, 1943,
1971; Marsh & Ambady, 2007; Jia, Ding, & Chen, 2021). Mammalian infants tend to have large,
round heads and eyes with big sclera, small nose, big cheeks, and a slightly opened mouth
(Lorenz, 1943, 1971). Comparatively, the adult facial expression of fear is associated with raised
and drawn together eyebrows, a slightly opened mouth, large and round eyes showing a high
proportion of sclera (white part of eye) to cornea, and lower eyelids tensed (Matsumoto &
9
Ekman, 2008; Ekman & Friesen, 1976; see Figure 1.1), showing clear overlap with
kindchenschema.

Figure 1.1. Images of the adult fear face and neutral infant face exhibiting morphological
features consistent with kindchenschema, or “baby schema” such as larger sclera, open mouth,
and raised and drawn together eyebrows. Kindchenschema are associated with caregiving action
tendencies in perceivers, thought to have evolved as a mechanism to elicit caregiving for altricial
young.

The adult fear face’s morphological similarity to infant faces may coopt the evolutionarily salient
cues that effectively elicit sensitive caregiving and costly helping behavior from perceivers.
Indeed, hierarchical interactions across mammalian species are marked by appeasement displays
(Hammer & Marsh, 2015). For example, when a dog is scolded by its owner, it will often display
more puppy-like features to the upset owner, such as opening their eyes wider with raised and
drawn together eyebrows. This often serves to appease the owner and elicit caring motivations,
instead of eliciting increased anger from the owner. In certain contexts, the adult fear face may
function similarly and serve as an appeasement cue that capitalizes on infant facial features to
effectively elicit caregiving motives and disarm aggression in perceivers. Indeed, in a study that
investigated the approachability of emotions (happiness, sadness, anger, fear, disgust, neutral,
surprise) across contexts, they found that participants rated adult fearful faces as highly
approachable in the context of helping the person displaying fear, but not when the participants
needed help from someone displaying a fearful face or in the absence of any context (Willis,
10
Windsor, Lawson, & Ridley, 2015). Furthermore, fearful faces may be perceived with differing
levels of threat depending on the vulnerability of the perceiver (Hammer & Marsh, 2015).
In support of this, studies of non-parents have found that fearful facial expressions elicit
approach behavior (Marsh, Ambady, & Kleck, 2005; Hammer & Marsh, 2015). Hammer and
Marsh (2015) used an implicit association lever task in which participants responded via arm
flexion or extension (theorized to reflect approach or avoid behavior) to fearful and angry facial
expressions as well as to neutral infant and adult faces. Participants responded in a pattern
consistent with an implicit association between adult fearful faces and neutral infant faces and
both expressions elicited approach behavior rather than avoidance (Hammer & Marsh, 2015).
Similarly, Marsh, Ambady, & Kleck (2005) used a lever task to examine appetitive versus
aversive responses to anger and fearful faces in perceivers and found that anger facilitated
avoidance behavior and fear facilitated approach behavior.
To summarize, adult fear facial expressions elicit context-dependent responses that may
be associated with motivated approach, perhaps due to the adult fear face’s morphological
similarity to kindchenschema. Existing evidence on the neurobiology of fear processing supports
the contention that the fear facial expression can lead to flexible responses in perceivers. There
are discernable neural and behavioral differences in response to fear versus anger. In a meta-
analysis of 105 fMRI studies using emotional faces paradigms, Fusar-Poli et al., (2009) found
that viewing fear faces differentially recruits the amygdala (along with sad and happy faces)
while viewing anger faces recruits the insula (along with disgust faces). The amygdala, a limbic
region, is implicated in perceiving the salience and valence of perceived stimuli (LeDoux, 2007;
Adolphs, 2002). The amygdala has downstream projections to a small part of the brain stem, the
periaqueductal gray (PAG), which is linked with avoidance and retreat behaviors (Linnman et
11
al., 2012; Namburi et al., 2015). However, the amygdala also projects to frontal regulatory
regions, with the strongest connections to the orbitofrontal cortex (OFC). These upstream
projections allow for flexible responding, and when warranted, intentional overriding of
avoidance-pathways to allow for approach-motivated responding.
Self-reported empathy and face emotion processing
Given that the communicative meaning of fear may be context-dependent, yet fearful
faces are salient enough to elicit motivated responding, responses to fear processing provide a
unique context for studying individual differences in motivated helping. Fear processing and
associated action tendencies rely on complex social processes such as empathy and perspective-
taking in order to decide whether to respond with approach or withdrawal behavior. Empathy is a
multidimensional construct that, at its most rudimentary level, is split into cognitive empathy and
affective or emotional empathy (Blair, 2008; Marsh, 2016; Marsh, 2013; Decety, Lewis, &
Cowell, 2015; Zaki & Oschner, 2012). Cognitive empathy, while stemming from affective
empathy (Preston & de Waal, 2002) refers to the ability to mentally understand other peoples’
points of view, motives, and goals and is commonly referred to as perspective-taking or theory of
mind (Brune and Brune-Cohrs, 2006). Affective empathy refers to being able to understand
others’ emotions and emotional states (Blair, 2008), and several studies demonstrate individual
differences in affective empathy for different emotional states, such as empathy for fear, disgust,
or anger (Marsh, 2016). Differences in fear processing have been found to differentiate groups
known to be particularly high or low on motivated caregiving. For example, individuals who
exhibit highly altruistic behavior (compared to healthy controls) appear to show heightened
sensitivity to fearful faces, whereas deficits in fear recognition are also linked with deficits in
helping behavior and higher harming behavior (Marsh, et al., 2014; Lozier, Cardinale, VanMeter,
12
& Marsh, 2014). Examining associations between self-reported empathy and approach-
avoidance tendencies in response to fear and adult distress-related stimuli may also provide
valuable insights into individual differences in expectant fathers that may reflect preparation for
parenthood.
Embodiment as mechanism
To measure approach-avoidance responses, the implicit association task presented in the
current study utilizes a joystick lever, which capitalizes on the mechanism of embodiment.
Embodiment refers to how bodily states or movements are likely coupled with neurological
activation of the somatosensory and perceptual/visual cortex. Applying theories of embodiment
to emotion processing, perceiving others’ emotions may implicate actually feeling or moving as
if one feels the emotion themselves (Niedenthal, 2007; Chartrand & Bargh, 1999; Hatfield,
Cacciopo, & Rapson, 1994; Neuman & Strack, 2000). Niedenthal (2007) also postulates that
people make emotional judgments by embodying their automatic emotional response to stimuli
via subliminal facial activation. In the current study, we are using an implicit association test
(IAT) that uses a joystick lever to capitalize on embodied simulation in response to emotional
facial expressions. Furthermore, there is a plethora of data supporting the notion that movement
influences our evaluation of stimuli (see Schwarz, 2017 for summary), but only when the valence
of the movement and stimuli match (Förster, 1998). Therefore, we would expect that congruent
trials on this IAT should show faster lever movement (response times) than incongruent trials.
The Current Study
The current study used a combined approach-avoidance IAT (Hammer & Marsh, 2015)
with stimulus pairings across emotions (i.e. angry, fearful) and maturity (i.e. adult, infant). Adult
fear faces and neutral infant faces are hypothesized to be the congruous condition, along with
13
adult anger faces and neutral adult faces; the incongruous conditions are adult fear faces and
neutral adult faces, and adult anger faces and neutral infant faces. Participants are asked to sort
the pairs of faces using a joystick lever, with behavioral approach operationalized as arm flexion
and behavioral avoidance operationalized as arm extension. Participants also completed a self-
report measure of trait empathy in order to examine associations between task performance and
empathy. Our study aims and hypotheses are as follows:
Aim 1: Do expectant fathers exhibit an implicit association for fearful and infant faces?
Hypothesis 1: Consistent with other studies, we expect that expectant fathers will show
faster reaction times when fearful faces are paired with infant faces.
Aim 2: Do fear and infant faces elicit approach-based responses among expectant fathers?
Hypothesis 2: Both fearful faces and infant faces will elicit faster behavioral approach
(pulling the lever), rather than avoidance (pushing the lever).
Aim 3: Is self-reported empathy associated with approach-avoidance responses for infant and
fear faces among expectant fathers?
Hypothesis 3: We expect that self-reported empathic concern and perspective taking will
be inversely associated with approach-responses (pull) to fear and infant faces.
Methods
Participants
Data for the current study were drawn from an ongoing longitudinal study that follows
cohabiting, heterosexual couples in Los Angeles over the transition to parenthood. All study
procedures received university Institutional Review Board (IRB) approval. The IAT was only
administered to the MRI substudy sample of the larger, longitudinal HATCH study. Interested
expectant fathers were first phone-screened for potential MRI eligibility while they were also
14
undergoing their screening for the larger longitudinal study. All interested males underwent
informed consent and, within two weeks of the in-lab visit where they completed the IRI,
completed the IAT after undergoing neuroimaging at a separate scan visit. Participants were paid
for their time. The current study included data from 42 fathers who completed the IAT and IRI.
Eight fathers’ data were dropped from IAT analyses due to lack of completion of the task (N=4)
and an excessive error rate (>= 2 SD above or below the mean; N = 4). This criteria for dropping
subjects is consistent with Hammer & Marsh (2015). Therefore, the final N used in data analyses
was 34. Males (Mage = 30.59, SD = 3.71, range = 23-37 years) participated when their partners
were between 20.38-38.52 weeks pregnant (M = 27.78, SD = 5.12). Complete demographic data
are presented in Table 1.1.
Table 1.1.
Participant Characteristics

Participant Characteristics
Age 30.6 (3.7)
Race
Asian / Pacific Islander 7 (20.6%)
Black / African American 3 (8.8%)
Hispanic / Latinx 12 (35.3%)
White / Caucasian 10 (29.4%)
Other / Decline to State 2 (5.9%)
Educational Attainment
High School / GED 1 (2.9%)
Some College 7 (20.6%)
College Degree 14 (41.1%)
Graduate Degree 12 (35.3%)
Couple Characteristics
Married 29 (85.3%)
Weeks Pregnant (PN) 27.78 (5.12)
Self-Reported Empathy
Empathic Concern 2.81 (.57)
Perspective-Taking 2.86 (.53)

15
Combined Approach-Avoidance Implicit Association Test
Materials
The IAT was administered using an Inspiron 17, 5000 series computer running Microsoft
7 Home Premium with a 17.3-inch screen, using Eprime 2.0 Professional Software. Participants
responded using a Microsoft Sidewinder joystick lever that was placed directly in front of the
participant. This task was an exact replication of the Hammer and Marsh (2015) task, and code
and stimuli for the task was kindly shared by Dr. Marsh’s Laboratory on Social and Affective
Neuroscience, with only minor adaptations made for our computer’s specific processing system.
Angry and fearful stimuli were drawn from the Picture of Facial Affect series (Ekman & Friesen,
1976) and a previously validated set of neutral infant and adult faces were pulled from publicly
available datasets (Marsh, Yu, Pine, Gorodetsky, Goldman, & Blair 2012). All images were
presented in grayscale on a black background in the center of the screen and appeared on the
screen for 2,000 ms.
Procedure
The task used in the current study (Combined Approach-Avoidance Implicit Association
Test (IAT)) was conducted after the prenatal MRI scan. During administration of the IAT, the
lead research assistant read instructions aloud to participants while they were simultaneously
presented on-screen. Participants were instructed that they would be sorting faces into categories
by either pushing or pulling the lever in front of them. Participants were asked to stabilize the
joystick with their left hand and to respond with their right hand. Participants were instructed to
hold the lever during the entirety of the task, were given time to practice moving the lever and
familiarize themselves with the lever, and were notified that, at various points throughout the
task, more instructions would appear on the screen and to please read them carefully prior to
16
proceeding. Participants were instructed to respond as quickly as possible without making
mistakes.
Measures
Combined Approach-Avoidance Implicit Association Test (IAT). The IAT was
adapted by Hammer & Marsh from the IAT originally developed by Greenwald & Banaji (2003),
which has been widely used in a variety of implicit association studies. Two congruent and two
incongruent blocks were presented, with two single-category practice blocks preceding them
resulting in eight blocks total (see Table 1.2). During congruent blocks, fearful and infant faces
were paired, and angry and adult faces were paired, such that participants responded to each pair
with the same response (either push or pull). For example, during one congruent block,
participants were instructed to push the lever in response to fearful and infant faces and to pull
the lever in response to angry and adult faces, then the instructions were reversed for the second
congruent block (see Table 1.2). During incongruent blocks, anger and infant faces and fearful
and adult faces were paired such that participants responded to each pair with the same response.
For example, during one incongruent block, participants were instructed to pull the lever in
response to angry and infant faces and to push the lever in response to fearful and adult faces,
then the instructions were reversed for the second incongruent block. Each test block had 40
trials where stimuli were presented for 2,000 ms and reminder labels (i.e. category names) were
positioned to the left and right of stimuli throughout the entire block. The order of block
administration was randomly generated by the task. Error messages were not displayed during
the entirety of the task.
Consistent with classic IAT format, single-category practice blocks were presented prior
to test blocks. Each practice block was 20 trials long and stimuli were presented for 2,000 ms
17
with category reminders on-screen throughout the block. For example, during one practice block,
participants were instructed to pull in response to adult faces and push in response to infant faces
and in the next block, participants were asked to pull in response to angry faces and push in
response to fearful faces. Practice blocks were not used in final scores.
Table 1.2.
Example of IAT Condition Presentation
Block No.
of
Trials
Function Stimuli assigned to
Push response
Stimuli assigned to
Pull response
1 20 Practice Fear Anger
2 20 Practice Infant Adult
3 40 Test-congruent Infant + Fear Adult + Anger
4 40 Test-incongruent Infant + Anger Adult + Fear
5 20 Practice Anger Fear
6 20 Practice Adults Infant
7 40 Test-congruent Adult + Anger Infant + Fear
8 40 Test-incongruent Adult + Fear Infant + Anger
Note. Blocks highlighted in gray indicate the test conditions used in analyses.

Interpersonal Reactivity Index (IRI). The Interpersonal Reactivity Index is a 28-item
self-report, multidimensional measure of empathy designed to assess individual differences in
empathic ability (Davis, 1983). It has four, seven-item subscales measuring different aspects of
empathy. The perspective-taking scale measures an individual’s ability to cognitively understand
another person’s point-of-view or experience (“When I'm upset at someone, I usually try to "put
myself in his shoes" for a while”). The empathic concern scale measures an individual’s
tendency to respond to other’s distress with warmth, compassion, and concern (“When I see
someone being taken advantage of, I feel kind of protective toward them”). Participants rate
themselves using a five-point Likert scale ranging from 0 (does not describe me well) to 4
(describes me very well). Item responses are summed (reverse-worded items scored
appropriately) to create subscale and total scores. The IRI has been shown to have good internal
18
validity and external validity (Davis, 1983). This measure was collected at the prenatal in-lab
visit preceding the IAT session. Means and standard deviations for the perspective-taking and
empathic concern subscales of the IRI were used for this study.
Results
Data from 34 fathers are presented. Rates of nonresponses, errors, and accuracy are
comparable or slightly better than Hammer & Marsh (2015). Nonresponses constituted 15.64%
of all trials, 13.20% of congruent trials, 18.10% of incongruent trials, 17.72% of emotion trials
(i.e. fear, anger) and 8.22% of maturity trials (i.e. infant, adult). Nonresponses were omitted from
all analyses in line with Greenwald and colleagues (2003). Participants responded with 78.11%
accuracy overall, 81.51% accuracy for congruent trials, 74.71% for incongruent trials, 73.41%
accuracy for emotion trials, and 89.69% accuracy for maturity trials. The total error rate was
6.25%, with a congruent trial error rate of 5.29%, incongruent error rate of 7.21%, emotion error
rate of 8.87%, and maturity anger rate of 2.10%. Consistent with Greenwald and colleagues
(2003), error trials were included in analyses by adding 600ms to the average reaction time for
the trial.
Aim 1: IAT Effects
First, to examine overall task performance, we investigated the implicit association of
fear and infant faces and angry and adult faces. To measure the IAT effect, Greenwald et al.’s
(2013) D was calculated by subtracting all combined congruent trials from the combined
incongruent trials and dividing the difference by the pooled standard deviation. D-scores can
range from − 2 to + 2 and the higher the D-score, the more positive an implicit attitude. This
analysis yielded a medium effect for the task, Mincongruent = 1,331 ms, Mcongruent = 1,275 ms, SD =
157 ms, D = .35. These results indicate that, consistent with Hypothesis 1, participants were
19
faster to respond when fearful faces were paired with infant faces and when angry faces were
paired with adult faces, t(33) = 3.47, p = .001 (see Figure 1.2).
IAT effect sizes were also calculated separately for emotional faces and for maturity
(infant vs adult faces). We observed a medium effect size for fearful and angry faces (Mincongruent
= 1,469 ms, Mcongruent = 1,404 ms, SD = 156 ms, D = .42) such that participants responded
quicker to blocks when fearful faces were paired with infant faces than with adult faces and
when angry faces were paired with adult faces than infant faces, t(33) = 2.29, p = .029. We
observed a medium effect for faces varying in maturity (Mincongruent = 1,222 ms, Mcongruent = 1,164
ms, SD = 184 ms, D = .31) indicating participants responded faster when infant faces were paired
with fearful faces than with angry faces and when adult faces were paired with angry faces over
fearful faces, t(33) = 2.71, p = .01; see Figure 1.2).

Figure 1.2. IAT Effects.
Overall IAT effects (left panel), participants were faster to respond when fearful faces were
paired with infant faces and when angry faces were paired with adult faces. IAT Emotion effects
(middle panel), participants responded quicker to blocks when fearful faces were paired with
infant faces than with adult faces. IAT Maturity effects (right panel), participants responded
faster when infant faces were paired with fearful faces than with angry faces.

20
Aim 2: Approach-Avoidance Responses
To examine our second aim of whether fear and infant faces elicit approach-based
responses, we collapsed trials across congruency and calculated two, 2 (lever direction) x 2(face
type) RM ANOVAS for emotion (angry, fearful) and maturity (infant, adult). We hypothesized
that fearful faces would elicit approach-motivated responses. There was a main effect of
emotional face, F(1, 33) = 8.23, p = .007. Participants responded more quickly to fearful (M =
1404.46 ms, SEM = 28.64 ms) than angry faces (M = 1465.00, SEM = 24.62 ms). There was no
main effect of lever direction, F(1, 33) = 1.71, p = .20 and no interaction of expression*lever
direction, F(1,33) = .067, p = .797 and therefore, our hypothesis that fear elicits approach
responses over avoidance responses was not supported.
We also hypothesized that infant faces would elicit approach-based responses. There was
a main effect of face maturity, F(1, 33) = 110.69, p = .000 such that participants responded more
quickly to infant (M = 1109.06 ms, SEM = 29.49 ms) than adult faces (M = 1276.10, SEM =
31.55 ms). There was no main effect of lever direction, F(1, 33) = 1.98, p = .168. In support of
our hypothesis, there was an interaction of face*lever direction, F(1,33) = 10.14, p = .003.
Participants pulled the lever faster than they pushed it in response to infant faces, M difference =
103.21 ms, t(33) = 3.22, p = .003. There was no significant difference in reaction time to adult
pulling versus pushing, but participants did respond faster to pushing adult faces than pulling
them (see Figure 1.3).
Table 1.3.
Mean Response Times to Face Stimuli During Congruent and Incongruent Trials

Congruent Incongruent

M (SD) M (SD) M Difference
Expression

Angry

Pull 1,329 (138) 1,555 (278) -249
21
Push 1,452 (178) 1,525 (256) -74
Fear

Pull 1,387 (291) 1,422 (223) -11
Push 1,447 (224) 1,396 (221) 95
Age

Adult

Pull 1,261 (232) 1,315 (207) -56
Push 1,232 (247) 1,296 (257) -8
Infant

Pull 1,018 (161) 1,097 (224) -83
Push 1,144 (239) 1,177 (257) 11
Note. All values are in milliseconds. Negative mean differences indicate response times were
greater in the incongruent condition

Figure 1.3. Approach-Avoidance Response Times by Condition and Congruency.
Response times are presented in milliseconds.

Aim 3: Response times and Self-Reported Empathy
Participants reported an average perspective-taking score of 2.86 (SD = .53; Range: 2.00
– 3.86) and empathic concern score of 2.81 (SD = .57; Range: 1.43 – 3.71). There was no
association between the subscales (r(34) = .21, p = .24) or between subscales and participant age
(p’s > .652) or days pregnant (p’s > .512).
22
We hypothesized that self-reported empathic concern and perspective taking would be
inversely associated with approach reaction times to fear and infant faces. We found partial
support for our hypothesis in that empathic concern was associated with faster reaction times for
approach (r(34) = -.525, p = .001) and avoidance (r(34) = -.431, p = .011) responses to infant
faces but was not to fear faces. Perspective-taking was associated with faster reaction times for
approach responses to fear faces, r(34) = -.355, p = .04, but was not associated with avoidance
responses to fear or any other faces. See Figure 1.4.

Figure 1.4. Approach-Avoidance Responses and Self-Reported Empathy.
Left panel-Empathic concern was associated with faster reaction times for approach (r(34) = -
.525, p = .001) and avoidance (r(34) = -.431, p = .011) responses to infant faces. Right panel-
Perspective-taking was associated with faster reaction times for approach responses to fear faces,
r(34) = -.355, p = .04.

Discussion
The current study aimed to replicate Hammer and Marsh’s (2015) study of approach-
avoidance mechanisms in face processing, testing the theory that fearful faces and infant faces
would elicit similar responses, and to extend this work within a sample of expectant fathers.
Consistent with Hammer and Marsh’s (2015) findings, participants showed faster implicit
23
associations when fearful faces were paired with infant faces and when angry faces were paired
with adult faces. We also replicated the finding that infant faces elicited behavioral approach
over avoidance responses but failed to replicate the finding that fear faces elicited behavioral
approach over avoidance responses. Lastly, we found partial support for our hypothesis that self-
reported empathic concern and perspective-taking would predict faster approach reaction times
to infant and fear faces. Fathers who reported higher self-reported empathic concern showed
faster approach and avoidance responses to infant faces but not fear faces, whereas fathers who
reported greater perspective-taking showed faster approach responses to fear faces but not infant
faces.
These data support our contention that the fearful face more closely approximates the
infant neutral face than adult neutral face. One explanation for this is that the fear face has
similar morphological features to “baby schema,” which are features that are strongly associated
with eliciting caregiving motives to ensure the survival of the individual displaying the face (Jia,
Ding, & Cheng, 2021). Although fear perception has not been studied in the transition to
fatherhood, studies examining fear in populations exhibiting motivated caregiving and altruism
have studied fear perception and empathy for fear and often hypothesize that adults help
unrelated others exhibiting distress because the adult fear face coopts “baby schema.”
However, while our data support potential morphological similarity between the neutral
infant and adult fear face, our data did not support our expectation that fear would elicit approach
behavior over withdrawal behavior. Expectant fathers did show a preference for approach
responses to infant faces. However, expectant father’s responses to fear were more nuanced.
When fearful faces were paired with infant faces, expectant fathers showed a preference for
approach responses to fear. In contrast, when fearful faces were paired with adult neutral faces,
24
expectant fathers exhibited faster withdrawal behavior to fearful faces. While different than the
Marsh & Hammer (2015) results, these data may be because the adult neutral face is often
interpreted negatively by viewers and may support fear being a context-dependent, ambiguous
cue requiring other surrounding evidence in order to interpret whether the fearful face is
displaying threat versus a need for help.
Given that fear may approximate the infant face, which we know is a salient cue eliciting
caregiving and appeasement, but that action tendencies linked with adult faces may be more
nuanced and context dependent, fearful adult face stimuli represent a useful approach for
elucidating individual differences in processes related to helping others in need. Studying fear
perception may also be informative of individual difference constructs that can affect caregiving,
such as sociocognitive processes like empathy. Indeed, when looking at self-reported empathy in
our expectant fathers who completed this IAT, perspective taking was linked with faster
behavioral approach to fearful faces, and empathic concern was related to approach and
avoidance responses to infant cues. Perspective-taking is often considered a top-down cognitive
process involving intentional effort, which would be required for deciphering the meaning of
adult fear expressions given the evidence that they are context-dependent.
In the current study, approach was operationalized as arm flexion and avoidance as arm
extension. Participants were not given explicit instructions related to approach-avoidance
responses, but instead were asked to categorize the faces by either “pulling” or “pushing.” There
has been some debate about whether approach-avoidance tendencies elicit arm flexion and
extension (Chen & Bargh, 1999; Rotteveel & Phaf, 2004; Wilkowski & Meier, 2010), but the
current paradigm was originally created by Hammer and Marsh (2015) with this in mind. While
we cannot be certain of whether flexion was appropriately categorized as approach and
25
extension categorized as avoidance, the current study paired stimuli that have more ambiguous
responding (adult fear and neutral faces) with stimuli that have relatively robust associations
with approach and avoidance responses, namely infant and adult anger faces respectively.
Therefore, while we cannot be certain of the responses, it is highly likely that approach and
avoidance tendencies were properly categorized by arm flexion and extension given that a
preference for arm flexion was shown for infant faces and arm extension for angry faces.
Moreover, the approach-avoidance nature of the task was implicit, thereby avoiding activating
pre-existing approach-avoidance construals (Seibt, Neumann, Nussinson, & Strack, 2008).
Manufactured distinction between automatic and intentional responding. Newer models
of cognitive processing, simulation, and embodiment have questioned the distinction between
automatic responding prior to conscious awareness versus intentional, conscious responding
(Barsalou, 2016). The current study targets implicit, automatic responses and is based on the
IAT framework of these responses being automatic (Greenwald, Nosek, & Banaji, 2003).
However, we interpret the data with caution, taking into account Barsalou’s (2016) situated
conceptualization theory in that the automatic reactions on the IAT measured via embodiment of
motoric movement in response to emotional faces may not be all that distinct from cognitive
control.
Conclusion and future directions
The current study provides further support for the ideas that infant and adult fearful facial
cues are associated together, that approach-avoidance tendencies in response to fearful cues may
be context-dependent, and that individual differences in self-reported empathy are associated
with faster categorization and approach behavior to both infant and fearful faces. Almost all of
the literature on the parental brain has focused on adults’ behavioral and neural responses to
26
infant cues, including infant faces and vocalizations. Expanding this literature to include broader
assessment of overall changes in emotion processing, including responses to adult emotional
faces, may be of value. Specifically, given that humans overwhelmingly positively respond to
infant faces and that capturing facial expressions of infant threat is challenging, it may be
valuable to examine perception of adult threat-related facial cues to provide a more nuanced
understanding of responses to distress-specific stimuli in expectant fathers, in addition to
understanding the broader impacts of the transition to fatherhood on general emotion processing.

27
STUDY 2: NEURAL AND BEHAVIORAL CORRELATES OF FEAR PROCESSING IN
MEN TRANSITIONING TO PARENTHOOD

Study 1 found support for the expectation that adult fear faces are more closely associated
with the infant face than with other adult faces, and that individuals may respond in context-
dependent ways with approach or avoidance (e.g., approach when adult fear faces and infant
faces were paired, avoidance when adult fear faces with paired with neutral adult faces).
Approach-avoidance responses to the adult fear face were also associated with perspective-
taking, a sociocognitive process that is historically associated with frontal-subcortical processing.
Early neuroimaging research on fathers suggests that the transition to fatherhood may be
facilitated by experience-dependent, context-specific change and linked to frontal-subcortical
processing that then engages a reinforcement learning loop (Abraham et al., 2014; Feldman,
Braun, & Champagne, 2019). Studying fear processing in males transitioning to parenthood may
provide valuable insight into neuroplasticity in response to emotion perception of threat-related
cues during major life transitions. The current study is the first study to examine males before
and after the birth of their first child using neuroimaging and behavioral paradigms assessing fear
emotion perception in the paternal brain.
Emotion perception
Emotion perception is the ability to identify emotional states or to discriminate among
emotional states from a variety of stimuli (Salzman & Fusi, 2010). Emotion perception is one
part of a more complex process that gives rise to our emotional experiences and associated
behaviors (Adolphs, 2002). The process of studying and operationalizing emotion perception is
complex and has been a topic of debate for the past century. There are various theories about
what occurs when the brain is confronted with a stimulus that evokes emotion. Broadly speaking,
most theories agree that when the brain is confronted with a stimulus that evokes emotion,
28
several component processes take place including neural, cognitive, somatic, affective, and
behavioral responses; however, they debate the order that these responses occur. For example,
the James-Lange theory hypothesizes when an emotional stimulus is perceived, it first leads to
visceral and behavioral responses and then a conscious experience of emotion (James, 1884,
1894; Lange, 1922). Another perspective posits that the conscious experience of emotion and
visceral and behavioral responses occur simultaneously (Cannon, 1927; Bard, 1928). A recent
theoretical framework stemming from neuroimaging research highlights that cognitive, visceral,
and emotional responses may not be separable processes, but instead occur as a complex network
of responses facilitated by the network connectivity between the prefrontal cortex and amygdala
(Salzman & Fusi, 2010).
Emotion perception and communication are pivotal to human relationships and essential
for successful parenting (Somerville, Fani, MaClure-Tone, 2011; Parsons et al., 2019). For
example, observing and accurately perceiving a person’s emotional responses tells us about that
person’s needs and their environment, and may guide our subsequent behavior towards that
person (Somerville et al., 2011; Marsh, 2016). Additionally, the ability to quickly and accurately
decode facial affect has long been linked with childrearing and caregiving (Babchuk, Hames, &
Thompson, 1985) and attachment processes (Ainsworth, 1979). Some data support that females
are faster than males at recognizing negatively-valanced emotional faces (Hampson, van Anders,
& Mullin, 2006) and that these sex differences exist because of the requirement for rapid
decoding of nonverbal cues for successful childrearing and survival of altricial young (this is
referred to as the “fitness threat hypothesis”, see Babchuk, Hames, & Thompson, 1985). If rapid
decoding of negatively-valanced emotional facial expressions is advantageous to successful
29
raising of offspring, then might alloparental involvement in childrearing lead to changes in
emotion perception for nonmaternal figures?
Neuroplasticity of face emotion processing
The face is often central to understanding others, so one common way to study emotion
perception is through emotional facial expressions. Face emotion processing abilities develop
and shift over the lifespan. Starting in infancy and continuing through adolescence to adulthood,
a “gradual refinement” of the ability to identify and distinguish between different emotional
facial expressions occurs with some variability in maturity trajectories between emotions
(Somerville et al., 2011). Typically, emotion perception for happiness develops earliest, with fear
and other negatively valanced expressions taking longer. Additionally, older adulthood has been
associated with decreases in accurate emotion perception of negatively-valanced faces including
fearful, sad, and angry expressions (Ruffman, Henry, Livingstone, & Phillips, 2008).
Therefore, there is evidence that face emotion perception shifts over time during the
course of the lifespan and the transition to parenthood may be another developmental milestone
in adulthood linked with plasticity in the neural and behavioral correlates of face emotion
processing. Changes in face emotion perception across the lifespan are likely due to several
different mechanisms, including neurodevelopment in infancy through early adulthood,
neurodegeneration in older adulthood, and neuroendocrine and social reinforcement learning
throughout adulthood (Somerville et al., 2011). One particular period of pronounced
neuroendocrine and social behavior change in adulthood is the transition to parenthood (Swain et
al., 2014; Feldman, Braun, & Champagne, 2019). Most studies of parenthood have focused on
the neural and behavioral correlates of infant face emotion processing (Kim, Strathearn, &
Swain, 2016), but there is reason to believe that the transition to parenthood and alloparental care
30
is linked with broader impacts to emotion processing beyond infant stimuli. Interestingly,
changes in fear processing have been linked with menstruation- and pregnancy-related
hormones. Higher progesterone levels during menstruation have been associated with accuracy
to identify fearful faces (Pearson & Lewis, 2005). Behavioral responses to fear faces over the
prenatal and postpartum periods tend to fluctuate with changes in reproduction-related hormones
(Pearson, Lightman, & Evans, 2009; Pearson & Lewis, 2005) and during the transition to
motherhood, women go through behavioral changes in sensitivity to infant cues, distress, and
environmental threat (Kim, 2016). Consistent with this, parents, compared to non-parents, are
better at discerning changes in degree of distress level on faces (Proverbio, Brignone, Matarazzo,
Del Zotto, & Zani, 2006).
Empathy for fear
Emotion perception specifically linked to fear, or the ability to perceive distress that is
exhibited as a fear facial expression, is sometimes referred to as empathy for fear and it is
considered a vital building block towards more complex social processes aimed at alleviating
others distress, such as motivated caregiving (Marsh, 2016; Preston, 2013). Empathy for fear is
specifically associated with amygdala activation, more so than empathy for any other emotion
(Fusar-Puli et al., 2009; Marsh, 2016). The amygdala is essential for the detection of distress
cues and salient features of the environment (Blair, 2008; Adolphs, Tranel, Damasio, &
Damasio, 1995). Adolphs, Tranel, Damasio, and Damasio (1995) hypothesized that the amygdala
is responsible for the perceptual detection of fear (e.g. fearful facial cues such as widened eyes)
and internal representations of fear (e.g. felt fear in the perceiver). Therefore, the ability to
perceive fear may be linked with individuals’ own ability to feel fear themselves (LeDoux, 2007)
which can be important for generating motivated approach behavior (Marsh, 2016). Empathic
31
fear allows the perceiver to understand the other’s distress which is essential for motivated
caregiving (Marsh, 2016; Preston, 2013; De Waal, 2009; Marsh, 2013; Decety, Lewis, & Cowell,
2015; Eisenberg & Miller, 1987).
The amygdala and the mammalian caregiving network
The amygdala is one of the most interconnected regions of the brain (LeDoux, 2007) and
highly implicated in alloparental and maternal care as part of a larger network of regions
coordinating approach-avoidance responses to distress stimuli and social reward processing
(Feldman, Braun, & Champagne, 2019; Kim, 2016). The amygdala is vital to avoidance-driving
mechanisms like fear conditioning (Duvarci and Pare, 2014), but is also responsive to reward
cues (Murray, 2007) and is linked with increased vigilance when outcome uncertainty is present
in an ambiguous context (Taylor and Whalen, 2014). The “caregiving network” that supports
parenting has been theorized to include the amygdala, medial preoptic area of the hypothalamus
(MPOA), ventral tegmental area (VTA), nucleus accumbens (NAcc), and the ventral pallidum
(Preston, 2013; Swain et al., 2012; Feldman, Braun, & Champagne, 2019). The amygdala has
downstream projections to the periaqueductal gray (PAG) and this pathway is robustly linked
with fear conditioning, aversive stimuli, and withdrawal behavior (LeDoux, 2007; Marsh, 2016;
Namburi et al., 2015). Amygdala-PAG coordination is needed for the onset of parenting behavior
because the PAG is involved in the coordination of avoidance-motivated and freezing behavior
and is essential for fear conditioning, making inhibition of this pathway essential to learning to
respond to distress and fear cues with motivated caregiving, instead of self-preservative
avoidance (Preston, 2013; Marsh, 2016; Salzman & Fusi, 2010; Brethel-Haurwitz et al., 2017).
However, inhibition of withdrawal behaviors in response to others’ distress is insufficient for
successful parenting. It must also be paired with the onset and maintenance of approach-
32
motivated responses to distress. The amygdala has projections to the MPOA, a region rich with
oxytocin and vasopressin receptors, which has been studied extensively in maternal pup retrieval
care in rats (Numan & Insel, 2003). The MPOA is thought to be activated via pregnancy
hormones, helping rats’ transition from avoidant responding to distress cues, to approach
responding and felt reward in response to distress cues (Numan & Insel, 2003; Preston, 2013).
This felt reward and subsequent reward expectancy to approaching distress cues is thought to be
subserved via the MPOA’s projections to the VTA, NAcc, and ventral pallidum (Preston, 2013).
However, research on these topics is still limited in humans, who have less drastic differences in
response to infant and distress cues than rodents given that humans tend to respond positively to
infant cues regardless of parental status.
Even though men also experience hormonal shifts in the postpartum period that coincide
with involvement in childrearing (Abraham et al., 2014), some studies support greater cortical,
vs. subcortical, involvement in the onset of paternal behaviors (Feldman, Braun, & Champagne,
2019). This may be due in part to cortical modulatory inputs to the amygdala of the orbitofrontal
cortex (OFC), dorsolateral prefrontal cortex (DLPFC), and anterior cingulate cortex (ACC)
which allow for intentional overriding of the amygdala-PAG pathway as a response to distress,
and activation of the approach pathways as a response to distress (Preston, 2013; Zaki, 2014;
FeldmanHall et al., 2015; Feldman, Braun, & Champagne, 2019). Approach behavior in response
to distress may be associated with reward, for example in retrieving a distressed infant and they
calm down or in helping a scared person realize they are safe. This felt sense of reward is likely
facilitated by the MPOA’s projections to reward-saliency and reinforcing regions like the NAcc
and ventral pallidum (Kim et al., 2014; Feldman, Braun, & Champagne, 2019; Preston, 2013).
Therefore, the transition to fatherhood may be subserved via an experience-dependent
33
reinforcing feedback loop between amygdala reactivity to detection of fear or distress and
cortical input (DLPFC, OFC) in evaluating the meaning of the distress cues (ACC), followed by
approach behavior (VTA) to alleviate the distress cues which then triggers felt reward (NAcc,
ventral pallidum).
The Current Study
The current paper investigates potential changes in the neural and behavioral correlates of
fear emotion processing that may occur across the transition to fatherhood based on experience-
dependent, or social reinforcement learning. To our knowledge, this is the first study to
longitudinally examine functional neural changes in men before and after the birth of their first
child and is the first study in the parenting brain literature to examine whether the transition to
parenthood is associated with changes in emotion processing of adult fear facial cues, which may
have broader implications for how involvement in alloparental care may impact processes like
empathy and motivated caregiving towards non-kin others. In both the prenatal and postpartum
periods, fathers underwent functional neuroimaging (fMRI) where they viewed static images of
emotional faces and were asked to identify the sex of the face, thereby making the emotion of the
face irrelevant to the fMRI task. Fathers were than administered a forced-choice emotion
recognition paradigm to assess accuracy in identifying emotional expressions. The current study
aims to characterize neural and behavioral responses to fear faces among expectant fathers
followed from the prenatal period into the postpartum period. We seek to understand whether
neural activation towards distressed others (adults exhibiting fearful faces) is dynamic across the
transition to parenthood. We planned to test prenatal, postpartum, and longitudinal hypotheses,
as presented below.
Prenatal Hypotheses:
34
1. We expect that men will exhibit activation in the amygdala, periaqueductal
gray, fusiform gyrus, inferior frontal gyrus, and medial orbitofrontal cortex
while viewing fearful faces (Fusar-Poli et al., 2009).
2. Prenatal fear recognition accuracy on the forced-choice emotion recognition
behavioral task paradigm will be positively associated with both amygdala
and periaqueductal gray percent signal change from the fear versus baseline
contrast.
Postpartum Hypotheses:
1. Following the birth of their infant, men will exhibit activation to fear faces in
the same regions found in the prenatal scan, with additional activation in areas
specifically associated with the mammalian caregiving network: the VTA,
NAcc, and ventral pallidum (Preston, 2013).
2. Postpartum fear recognition accuracy on the forced-choice emotion
recognition behavioral task paradigm will be positively associated with
percent signal change during fear versus baseline contrast in the amygdala,
and inversely associated with periaqueductal gray percent signal change
during the fear versus baseline contrast, consistent with a dampening of
avoidance pathways in favor of approach-motivated pathways.
Longitudinal Hypotheses:
1. Men will exhibit increased activation in social reinforcement learning and
reward processing regions (VTA, NAcc, and ventral pallidum) to fear faces in
the postpartum period, compared to the prenatal period, and will exhibit
35
decreased responding in the PAG to fear faces from the prenatal to postpartum
period.
2. There will be a significant increase in fear recognition accuracy from the
prenatal to postpartum period.
Method
Participants
Recruitment. Participants were drawn from a larger, ongoing, longitudinal study of 100
couples expecting their first child who were recruited via flyers (posted in obstetrician’s offices
and neighborhood destinations) and online (e.g., posting on neighborhood websites and
Facebook parenting groups). This study follows first-time, heterosexual couples from pregnancy,
through the birth, and into the postpartum period. The goal of the study is to investigate the
hormonal, neurological, behavioral, and social changes that males and females experience during
their transition from “dyad to triad” and how these changes relate to parenting.
Inclusion and exclusion criteria. Inclusion criteria included that the parents be
heterosexual first-time parents of a singleton fetus, who are cohabitating and planning to cohabit
after the birth. Participants also needed to be able to complete the visits in English, free of
medications (such as steroids) known to interfere with the hormones under study (cortisol,
testosterone, prolactin, or oxytocin), and free of conditions affecting these hormones (e.g.
Cushing’s syndrome, adrenal insufficiency). Exclusion criteria for the larger study included if a
partner was on medications or had any conditions interfering with hormones, history of
psychiatric illness requiring medication, or use of illegal drugs. The MRI substudy had additional
exclusion criteria specific to neuroimaging. Participants were excluded from the MRI if they ever
had a loss of consciousness for greater than 10 seconds, history of neurological illness, pervasive
36
developmental disorder, autism spectrum disorder, any movement disorder (e.g. tic disorder),
claustrophobia, metal implants, left handedness, or current use of psychotropic medication or any
medication affecting the nervous system (e.g. beta blockers). If participants were using a
psychotropic medication that could be safely withheld for 24 hours prior to scanning, then they
were still eligible for inclusion.
Participant Characteristics. Forty-three males were recruited and eligible from the
larger longitudinal study to participate in the MRI substudy. Of those 43 males, one could not
complete neuroimaging due to unexpected claustrophobia and was subsequently dropped from
the MRI substudy. Four men were dropped from neuroimaging analyses due to extensive artifact
(N = 1), excessive motion (mean framewise displacement >.3; N = 1) and poor scan quality (N =
2), resulting in a final N of 38 fathers used in prenatal analyses. Thirty-five of the 42 prenatal
fathers were scanned again postpartum (6 scanned after COVID-19 pandemic began). Of these
35 postpartum fathers, four were dropped from neuroimaging analyses due to excessive motion
(mean framewise displacement >.3), resulting in 31 fathers used in postpartum analyses. See
Table 2.1 for participant characteristics.
Table 2.1.
Participant Characteristics

Participant Characteristics
Prenatal (N=38) Postpartum (N=31)
Longitudinal
(N=28)
Age 31.9 (3.9) 32.32 (3.7) 32.96 (3.7)
Race
Asian / Pacific Islander 7 (20.6%) 9 (29.0%) 9 (32.1%)
Black / African American 2 (5.3%) 3 (9.7%) 2 (7.1%)
Hispanic / Latinx 11 (28.9%) 9 (29.0%) 7 (25.0%)
White / Caucasian 10 (26.3%) 8 (25.8%) 8 (28.6%)
Other / Decline to State 2 (5.3%) 2 (6.5%) 2 (7.1%)
Educational Attainment
37
High School / GED 1 (2.6%) - -
Some College 6 (15.8%) 4 (12.9%) 3 (10.7%)
College Degree 13 (34.2%) 11 (35.5%) 10 (35.7%)
Graduate Degree 18 (47.4%) 16 (51.6%) 15 (53.6%)
Couple Characteristics
Married 33 (86.8%) 27 (86.8%) 24 (92.3%)
Weeks Pregnant (PN) 27.98 (5.09) - 29.00 (5.15)
Infant Age- Months (PP) 8.05 (2.78) 8.12 (2.93)
Time Between MR-Scans in
months
9.32 (2.78)
Right Handedness 38 (100%) 31 (100%) 28 (100%)
Note. PN = prenatal, PP = postpartum
Procedures
All study procedures were approved by the University Institutional Review Board.
Interested males were first phone screened for potential MRI eligibility while they were also
undergoing their screening for the larger longitudinal study. If deemed eligible during the phone
screen, males were asked if they would be interested in an additional substudy involving
neuroimaging. If interested, they were given the MRI consent form, safety screening form, and
incidental findings form at the prenatal in-lab visit of the larger study and later brought in for the
MRI study and completed the IRB-approved informed consent form again, the safety screening
form, and incidental findings form. Then the lead graduate research assistant described the
scanning protocol to the father and ran representative practice runs of functional tasks. The
prenatal and postpartum MRI visits each involved informed consent, safety screening, an hour
and a half scanning protocol, and subsequent behavioral testing. Participants were paid for their
time and their radiology scan from the prenatal visit was sent out for anonymous review by a
radiologist in case of incidental findings. In total, the prenatal and postpartum visit each took two
to three hours. The tasks used in the current study were collected at both prenatal and postpartum
scans.
Measures
38
Demographics. As part of the larger longitudinal study, participants complete a two-hour
battery of questionnaires at both the prenatal and postpartum in-lab visits. This battery includes
questions about participant demographic information, including age, ethnicity, education, and
relationship status.
Forced-Choice Behavioral Face Emotion Recognition Task. Participants were shown
fear, anger, disgust, surprise, happy, and sad facial expressions (Ekman & Friesen, 1976) and
asked to categorize them by the emotion being displayed. This task has been used to assess
emotion recognition ability in adults undergoing similar neuroimaging paradigms as the fMRI
task administered in this study (Marsh, et al., 2014). The task features individual presentations of
static faces, from the Pictures of Facial Affect set (Ekman & Friesen, 1976), expressing “basic”
emotions (anger, disgust, fear, happiness, sadness, and surprise) displayed by 10 different
Caucasian adults. Participants therefore viewed 60 expressions total, presented in randomized
order. Each expression appeared for 2,000 ms and was followed by a screen that instructs
participants to make a forced choice among responses corresponding to the six possible
emotions. Responses were self-paced. Both response selections and latencies were recorded.
Given that the focus of the current paper is on fear-processing and we did not have hypotheses
about the other emotions, we only report data from the fear condition in this manuscript.
fMRI Face Emotion Processing Task. While in the MRI scanner, participants were
presented with pictures of adult males and females displaying angry, fearful, and neutral facial
expressions and pictures of infants displaying neutral facial expressions and were asked to
indicate the sex of the face via button-box. The emotion of the face was irrelevant to the
instructed task. Similar fMRI tasks have been used in other non-parenting-focused studies
examining the neural correlates of face emotion recognition and its relation to emotion
39
perception and empathy and helping or harming behavior in special populations (Marsh, Stoycos,
et al., 2014; Lozier et al., 2014; Fusar-Poli et al., 2009).

Figure 2.1. fMRI Emotion Perception Task.
Example TR of the event-related series fMRI Face Emotion Processing Task administered at the
prenatal and postpartum visit. Fathers were asked to indicate the sex of the face via button press,
thereby making the emotion of the face irrelevant to the instructed task.

Two event-related fMRI runs of the task were programmed and presented via Eprime 2.0
Professional. Each trial of the task was three seconds long (2000 ms face presentation, 1000 ms
fixation) and each TR consists of fearful, angry, neutral, baby, and a fixation-rest trial, totaling
15-seconds per repetition (See Figure 2.1). A total of 20 repetitions were presented per run,
making each run of the task 5-minutes long. Angry and fearful facial expressions were taken
from the Pictures of Facial Affect Series dataset (Ekman & Friesen, 1976). Neutral images were
taken from the Karolinska Directed Emotional Faces dataset (KDEF; Lundqvist Flykt, & Öhman,
1998). All stimuli were presented against a black background with the neck and hair masked.
The order of stimulus presentation was determined using the genetic algorithm program in
Matlab (Wager & Nichols, 2003), which is designed to optimize the statistical power and
psychological validity within the fMRI design parameters. The current task was optimized for
the fear condition. Therefore, the two runs have a fixed-order presentation and run order was
counterbalanced across all participants. Only data from the fear versus baseline (rest modeled
40
implicitly) contrast are presented in the current paper, given our lack of hypotheses about the
other emotions.
MRI data acquisition. Whole-brain images were acquired using a 3T Siemens
MAGNETOM Prisma scanner with a 20-channel phased-array head coil. High-resolution, T1-
weighted anatomical images were acquired (3D Magnetization Prepared Rapid Acquisition
Gradient Echo; repetition time, 2530 ms; echo time, 3.13 ms; flip angle, 10°), with an isotropic
voxel resolution of 1mm
3
. Functional data were collected using a T2*-weighted echo-planar
imaging (EPI) with an interleaved sequence (40 2.5 mm transversal slices; repetition time, 2000
ms; echo time, 25 ms; field of view, 192 mm
2
; 3.0 x 3.0 x 2.5 mm voxels, flip angle, 90 degrees).
Data Analysis
Forced-Choice Behavioral Face Emotion Recognition Task Analysis. Emotion
recognition accuracy in this task was measured using an unbiased hit rate (Hu; Wagner, 1993)
and raw accuracy (H) analysis calculated for fear. Unbiased hit rate (Hu) determines accuracy by
assessing both raw accuracy and differential accuracy allowing us to account for both stimulus
accuracy and response bias in the participant. Raw accuracy and differential accuracy are
multiplied together to create the unbiased hit rate (Wagner, 1993). Raw accuracy is calculated as
the number of hits (correct categorization) divided by the number of stimuli presented of that
emotion category. Differential accuracy is calculated using the number of hits divided by the
total number of participant uses of that type of response. The proportion of hits expected by
chance (p(chance)) was calculated by taking the total stimuli presented for each emotion divided
by all stimuli presented, then multiplied by the total chosen for a given emotion divided by all
stimuli chosen. Data for H, Hu, and p(chance) were arcsine transformed prior to inferential
statistical testing (Wagner, 1993).
41
To test whether participant responses for fear recognition exceeded those expected by
chance, paired t-tests for each emotion comparing means of H and p(chance) and Hu and
p(chance) were computed. Once determined to be above chance, arcsine transformed H and Hu
were used in paired t-tests to examine whether there were mean differences in H and Hu from
prenatal to postpartum.
Additionally, arcsine transformed unbiased hit rate data were used during fMRI ROI
analyses so that individual differences in fear recognition during implicit fear processing could
be assessed. Bivariate correlations were examined between bilateral amygdala and PAG percent
signal change and arcsine transformed fear recognition accuracy at both prenatal and postpartum.
In order to examine whether PAG activation was associated with change in fear recognition from
prenatal to postpartum, we computed change scores for explicit fear recognition and right PAG
activity by subtracting prenatal from postpartum scores. We used these change scores to examine
whether prenatal right PAG percent signal change and change in right PAG activation predicted
change in explicit fear recognition using multiple linear regression when controlling for prenatal
fear recognition. We ran two models, one with arcsine transformed data, and one with raw values
and mean-centered predictors to assist with interpreting the models.
fMRI Face Emotion Processing Task Analysis. Response accuracy (male versus
female) and response times were collected during the fMRI task as a proxy for assessing whether
the participant was paying attention during the task. Means and standard deviations are reported
for prenatal and postpartum data. Paired t-tests were run to compare whether accuracy and speed
of responding in response to fearful faces differed from prenatal to postpartum.
Neuroimaging Analyses. Functional data were preprocessed and analyzed according to
the general linear model in FSL (FMRIB, Oxford, UK). FMRI data processing was carried out
42
using FEAT (FMRI Expert Analysis Tool) Version 6.00, part of FSL (FMRIB's Software
Library, www.fmrib.ox.ac.uk/fsl). We tested both nonlinear and linear registration of high
resolution structural and standard space images and had better registration with the linear
approach, therefore FLIRT was used for registration (Jenkinson 2001, 2002). Pre-statistics
processing included motion correction using MCFLIRT (Jenkinson 2002), slice-timing
correction using Fourier-space time-series phase-shifting, non-brain removal using BET (Smith
2002), spatial smoothing using a Gaussian kernel of FWHM 6.0mm, grand-mean intensity
normalization of the entire 4D dataset by a single multiplicative factor, and high-pass temporal
filtering. The general linear model had four explanatory variables (fear, anger, neutral, infant)
with rest modeled implicitly, to account for each condition while minimizing multicollinearity.
Infant, neutral, and angry faces were modeled so that they are not defaulted into the baseline
condition, however, they were not examined as variables of interest in this study. Contrasts of
interest were created as linear combinations of the explanatory variables to look at overall task
effects (e.g., fear > baseline). Time-series statistical analysis was carried out using FILM with
local autocorrelation correction (Woolrich 2001) and Z (Gaussianised T/F) statistic images were
thresholded non-parametrically using clusters determined by Z > 3.1 and a corrected cluster
significance threshold of P = 0.05 (Worsley 2001). Higher-level and group analyses were carried
out using FLAME (FMRIB's Local Analysis of Mixed Effects; Beckmann 2003, Woolrich 2004,
Woolrich 2008). All coordinates reported are from the Montreal Neurological Institute (MNI)
standard brains in radiological view.
Whole brain analyses and a priori hypothesized region of interest analyses were
conducted to examine overall effect of the task and overall response to distress cues (fearful
faces). Region of interest analyses on the bilateral amygdala were anatomically defined from the
43
Harvard-Oxford cortical atlas. Region of interest analyses on the left and right periaqueductal
gray were defined by creating a 5mm spherical ROI centered at MNI coordinates of x = -4 (left)
or 4 (right), y = -29, z = -12, based off of extant literature (Linnman, Moulton, Barmettler,
Becerra, & Borsook, 2012). To compute longitudinal analyses, FSL repeated measures
experimental design for a single-group paired difference was used to examine group differences
in prenatal and postpartum data. Input files were from the contrast of fear > baseline. Parameter
estimates from activated regions were converted into percent signal change and extracted using
Featquery and then used in SPSS to compute bivariate correlations, paired t-tests, and multiple
linear regressions.
Results
Forced-Choice Behavioral Face Emotion Recognition Task
Prenatal (N = 38). Accuracy rates above chance level of responding were found across
all emotion categories for both H (t(37)’s > 16.88, all p’s < .001) and Hu (t(37)’s > 14.97, all p’s
< .001). Father’s prenatal fear recognition raw accuracy (H) was 75% (SD = 18%) and unbiased
hit rate (Hu) was 60% (SD = 21%).
Postpartum (N = 31). Accuracy rates above chance level of responding were found
across all emotion categories for both H (t(30)’s > 14.83, all p’s < .001) and Hu (t(30)’s > 10.37,
all p’s < .001). Father’s postpartum fear recognition raw accuracy (H) was 76% (SD = 18%) and
unbiased hit rate (Hu) was 63% (SD = 20%).
Longitudinal (N = 28). There was no mean difference between prenatal (M = 72%, SD =
18%) and postpartum (M = 76%, SD = 17%) raw accuracy (t(27) = -1.243, p = .225). There was
no mean difference between prenatal (M = 57%, SD = 21%) and postpartum (M = 64%, SD =
19%) unbiased hit rate accuracy (t(27) = -1.56, p = .130).
44
fMRI Face Emotion Processing Task
Prenatal. Participants (N = 38) responded to the task with an overall average of 95%
accuracy for the gender of the face (SD = 11%) and with an overall reaction time of 921.01 ms
(SD = 126.55 ms) indicating they paid attention during the display of emotional faces during the
fMRI task. In response to fear faces, participants responded with 96% accuracy (SD = 10%) and
an average reaction time of 874.12 ms (SD = 145.09 ms). There were no associations found
between days pregnant, participant age, and task performance (all p’s > .10).
Postpartum. Participants (N = 31) responded to the task with an overall average of 96%
accuracy for the gender of the face (SD = 7%) and with an overall reaction time of 874.09 ms
(SD = 115.56 ms) indicating they paid attention during the display of emotional faces during the
fMRI task. In response to fear faces, participants responded with 98% accuracy (SD = 6%) and
an average reaction time of 874.09 ms (SD = 157.33 ms). There were no associations found
between infant age and task performance (all p’s > .20).
Longitudinal. In order to rule out differences in attention to the fMRI task as an
explanation for neural activation in response to fear faces differing between timepoints, we
compared fMRI behavioral task performance from prenatal to postpartum. There were no
differences in task performance accuracy (t(27) = -.79, p = .437) or reaction time (t(27) = -.20, p
= .845; (see Figure 2.2).
45

Figure 2.2. fMRI Implicit Faces Behavioral Data for Fear.
Participants indicated the sex of the face during the fMRI task to ensure attention to face emotion
stimuli. Response times and accuracy were calculated. There were no differences in task
performance accuracy or reaction time from prenatal to postpartum.

Neuroimaging Results
Peak MNI coordinates and Z-statistics for activation clusters are in parentheses and a
summary of neuroimaging results are presented in Table 2.2.
Prenatal. Consistent with our hypotheses, when viewing fearful faces compared to
baseline, expectant fathers showed activation in the right occipital lobe (44, -78, -14, Z = 8.53;
36, -58, 46, Z = 4.86), right inferior frontal gyrus pars opercularis (44, 8, 22, Z = 5.67) and pars
triangularis (42, 28, 20, Z = 5.30), and the left post (-38, -26, 48, Z = 5.18) and precentral gyri (-
44, 4, 28, Z = 5.33). Fathers also showed activation in subcortical emotion processing regions
including the bilateral amygdala (-18, -6, -18, Z = 5.57; 20, -4, -18, Z = 5.16), bilateral
paracingulate/ACC (-4, 8, 48, Z = 5.07) as well as memory processing regions (right
hippocampus, 24, -30, -6, Z = 5.99), the insula (30, 18, 0, Z = 4.31) and lastly, the periaqueductal
gray of the brain stem (6, -28, -10, Z = 4.25). Fathers showed deactivation in the left lateral
occipital cortex, superior division extending to the right postcentral gyrus, superior parietal
46
lobule, and precuneus (-46, -76, 30, Z = 7.39), the left subcallosal cortex extending to the frontal
pole and anterior cingulate cortex (0, 24, -14, Z = 5.93), the left superior frontal gyrus and
middle frontal gyrus (-20, 16, 48, Z = 5.21), and the right middle and superior frontal gyrus and
frontal pole (26, 28, 46, Z = 5.6). See Figure 2.3.

Figure 2.3. Prenatal Whole Brain Results.
Activation is pictured in orange and deactivation in blue.

Parameter estimates were extracted from a priori ROI’s of the left and right PAG and
amygdala. Bivariate correlations were run with ROI’s and explicit fear recognition. Contrary to
our hypotheses, left (r(38) = .264, p = .11) and right (r(38) = -.080, p = .63) amygdala percent
signal change were not associated with explicit fear recognition. Consistent with our hypotheses,
47
both left (r(38) = .429, p = .007) and right (r(38) = .440, p = .006) PAG activation during fear >
baseline contrast were positively associated with explicit fear recognition. See Figure 2.4

Figure 2.4. Prenatal PAG Activation and Fear Recognition Accuracy.
Left panel- Bilateral periaqueductal gray (PAG) activation (6, -28, -10, Z = 4.25) during prenatal
fear > baseline contrast. Right panel- Plotting the relationship between fear recognition accuracy
and bilateral ROI periaqueductal gray percent signal change from the fear > baseline contrast.

Postpartum. When viewing fearful faces compared to baseline, postpartum fathers
showed increased activation in a cluster with peak activation in the bilateral occipital pole (22, -
100, 6, Z = 5.61) reaching to the bilateral temporal occipital fusiform, in the left putamen
extending to the amygdala and hippocampus (-24, 3, -10, Z = 5.36), the right putamen and
pallidum (18, 8, -2, Z = 4.68), right amygdala and hippocampus (22, -30, -12, Z = 5.38), and the
left occipital fusiform (-34, -60, -18, Z = 5.12). Fathers showed deactivation in the bilateral
precuneus and posterior cingulate (8, -48, 44, Z = 5.44), the left lateral occipital cortex, superior
and inferior division as well as the parietal operculum cortex (-40, -86, 26, Z = 5.13), the left
frontal pole and paracingulate gyrus (-6, 68, 6, Z = 4.34), right parietal operculum (48, -32, 24, Z
= 4.09), left superior frontal gyrus extending to the middle frontal gyrus (-20, 34, 38, Z = 4.64),
the right frontal pole (26, 36, 30, Z = 4.50), and the right middle frontal gyrus (18, -2, 46, Z =
4.16). See Figure 2.5. Two additional separate whole-brain analyses were run with infant age and
time between scans as a covariate to examine whether these variables should be included in
48
longitudinal analyses. Neither infant age nor time between scans was associated with brain
activity during the fear versus baseline contrast.

Figure 2.5. Postpartum Whole Brain Results.
Activation is pictured in orange and deactivation in blue
Parameter estimates were extracted from a priori ROIs of the bilateral PAG and
amygdala. Bivariate correlations were run with ROI’s and postpartum explicit fear recognition.
Contrary to our hypotheses, left (r(31) = -.092, p = .62) and right (r(31) = .048, p = .797)
amygdala, as well as left PAG (r(31) = -.306, p = .095) percent signal change were not
associated with explicit fear recognition. Partially consistent with our hypotheses, right PAG
percent signal change was marginally significantly, negatively associated with postpartum
explicit fear recognition (r(31) = -.328, p = .072).
49

Longitudinal. A paired t-test comparing activation during fear versus baseline across
prenatal and postpartum visits revealed that fathers had more activation in the right (30, -92, 20,
Z = 4.35) and left (-8, -96, 0, Z = 4.20) occipital poles during the prenatal scan than the
postpartum scan. Fathers showed increased activation in the left parietooccipital junction (-44, -
78, 34, Z = 4.02) in the postpartum scan, compared to the prenatal scan. See Figure 2.6.

Figure 2.6. Longitudinal Whole-Brain Results.
Fathers showed greater activation in the bilateral occipital poles during the prenatal, compared to
postpartum period (purple clusters). Fathers showed greater activation in the parieto-occipital
junction in the postpartum, compared to the prenatal period (red cluster).

Table 2.2.

Peak MRI Activation Coordinates
Region Peak x Peak y Peak z Z
Prenatal
Fear > Baseline
right occipital 44 -78 -14 8.53
right IFG pars o 44 8 22 5.67
left postcentral gyrus -38 -26 48 5.18
left precentral gyrus/MFG -44 4 28 5.33
bilateral paracingulate / ACC -4 8 48 5.07
left amygdala -18 -6 -18 5.57
right lateral occipital 36 -58 46 4.86
right amygdala 20 -4 -18 5.16
right hippocampus 24 -30 -6 5.99
right insula 30 18 0 4.31
periaqueductal gray 6 -28 -10 4.25
50
Baseline > Fear
parieto-occipital junction -46 -76 30 7.39
left frontal pole 0 24 -14 5.93
left superior frontal gyrus -20 16 48 5.21
right MFG 26 28 46 5.6
Postpartum
Fear > Baseline
bilateral occipital pole and temporal
occipital fusiform 22 -100 6 5.61
left putamen, amygdala, and
hippocampus -24 4 -10 5.36
right putamen and pallidum 18 8 -2 4.68
left occipital fusiform and temporal
fusiform -34 -60 -18 5.12
right hippocampus and amygdala 22 -30 -12 5.38
Baseline > Fear
bilateral precuneus 8 -48 44 5.44
parieto-occipital junction -40 -86 26 5.13
left frontal pole -6 68 6 4.34
right parietal operculum 48 -32 24 4.09
left superior frontal and MFG -20 34 38 4.64
right frontal pole 26 36 30 4.5
right MFG 18 -2 46 4.16
Longitudinal
Prenatal > Postpartum, Fear > Baseline
right occipital pole 30 -92 20 4.35
left occipital pole -8 -96 0 4.2
Postpartum > Prenatal, Fear > Baseline
parieto-occipital junction -44 -78 34 4.02

We ran a series of multiple linear regression models examining right prenatal PAG
activation and changes in PAG activation across the transition to parenthood as predictors of
changes in fear recognition. One model used arcsine transformed values of fear recognition, and
the other used raw scores. Both models controlled for prenatal fear recognition. All predictors
were mean-centered for interpretability of the intercept coefficient. Results were similar in
magnitude, directionality, and level of significance between the arcsine transformed model and
51
the raw fear recognition model; therefore, we report results from the model using raw fear
recognition for ease of interpretation. Sensitivity analyses were conducted examining the effects
of interest in separate models and with and without prenatal fear recognition as a covariate, and
results were similar in direction, magnitude, and level of significance.
The model was significant, F(24,3) = 10.02, p < .001, and prenatal right PAG activation
predicted change in fear recognition accuracy (see Table 2.3). A scatterplot of this association is
depicted in Figure 2.7. Change in right PAG activation did not predict change in fear recognition.
Table 2.3.

Results from Multiple Linear Regression Models Predicting Changes in Fear Recognition

Model with raw Hu fear recognition
Variable β b 95% C. I.
Prenatal PAG % Signal Change
-
.430* -0.644 -1.151 to -.137
Change in PAG % Signal Change -.235 -0.239 -0.574 to 0.096
Prenatal Hu Fear Recognition
-
.545* -0.612 -0.951 to -0.272
Intercept

0.070 0.004 to 0.136
Note. *p < .05. Predictors in the second model with raw Hu fear recognition are grand mean-
centered. The model accounted for 56% of the variance in change in fear recognition, F(24, 3) =
10.02, p < .001, R
2
= .56.

-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
-0.2 -0.15 -0.1 -0.05 0 0.05 0.1 0.15 0.2 0.25 0.3
Change in Unbiased Hit Rate Fear
Recognition
Unstandardized Residuals of Prenatal PAG % Signal
Change
52
Figure 2.7. Prenatal PAG % Signal Change Predicts Change in Fear Recognition.
Prenatal PAG percent signal change is adjusted for prenatal fear recognition and change in PAG
percent signal change.

Discussion
The current study aimed to characterize males’ neural and behavioral responses to fear
faces from the prenatal to postpartum period in order to examine whether neural activation
towards distressed others (specifically, adults displaying fear) is dynamic across the transition to
parenthood.
Prenatal fear processing
Consistent with our prenatal hypotheses, expectant fathers exhibited BOLD activation in
fear processing, emotional saliency, and aversive responding regions (bilateral amygdala and
PAG), face-processing regions (fusiform gyrus), visual processing (right occipital cortex),
somatosensory regions (pre and post central gyri), insula, and frontal regulatory regions (IFG,
ACC). The pattern of functional activation seen in the prenatal period in response to fear faces is
consistent with the extant literature examining fear processing in healthy populations during a
task that does not require emotion labeling (Dricu & Frühholz, 2016; Fusar-Poli et al., 2009).
Also consistent with our prenatal hypotheses, fear recognition accuracy was positively associated
with PAG activation, but contrary to our hypotheses it was not associated with amygdala
activation. In further exploration of this, there is some evidence that labeling an emotional facial
expression results in amygdala deactivation, not activation (Hariri, Bookheimer, Mazziotta,
2000). This may be due in part to that amygdala activation is linked with the implicit (passive
viewing without prompted attention to the emotion of the face) viewing of fear processing that
pulls for saliency/arousal detection, but the labeling of the emotion as fear requires lexical
knowledge that may be more associated with frontal and temporal regions (Dal Monte et al.,
53
2013) which can actually then dampen amygdala response (Stoycos et al., 2017; Adolphs, 2002).
The positive association found between PAG activation and fear recognition accuracy is likely
subserved by the amygdala-PAG pathway, which is robustly linked with fear conditioning,
aversive stimuli, and withdrawal behavior (Namburi et al., 2015). Overall, these data are
consistent with expectant fathers showing prenatal fear perception patterns that are consistent
with threat cue processing and fear conditioning which often lead to withdrawal behaviors.
Postpartum fear processing
In the postpartum period, in response to fear versus baseline, fathers exhibited neural
activation in regions similar to that of the prenatal period, but with the exclusion of the PAG and
addition of the putamen, pallidum, and caudate. This was consistent with our hypotheses that
fathers would show BOLD activation in response to fear faces in reward saliency and social
reinforcement learning regions such as the pallidum and putamen. The putamen and pallidum
have been found as essential to paternal pup retrieval behavior in male mice and implicated in
fatherhood (Bakermans-Kranenburg, Lotz, Alyousefi-van Dijk, & Ijzendoorn, 2019; Feldman,
Braun, & Champagne, 2019). These regions are not typically associated with fear perception,
fear learning, or threat perception in the extant literature, and instead have been associated with
“learned safety” paradigms (Kong, Monje, Hirsch, & Pollack, 2013) and the mammalian
caregiving network and neural responses to infant faces (Feldman, Braun, & Champagne, 2019;
Preston, 2013; Marsh, 2016; Swain et al., 2014; Hoekzema et al., 2020).
The “learned safety” paradigm presents a potentially useful framework for understanding
possible mechanisms. Learned safety and learned fear both follow associative learning
paradigms where a conditioned stimulus (CS) and unconditioned stimulus (US) are paired. In
safety learning, there is an inverse association between an aversive US and neutral CS whereby
54
the neutral CS can prevent fear learning to the CS when paired with a safe environment (Kong,
Monje, Hirsch, & Pollack, 2013). This process is referred to as “conditioned inhibition of fear
learning” (Pollack et al., 2008). This model fits directly with the notion that fear processing and
its related action tendencies may be context-dependent, as discussed in Paper 1. By contrast, fear
learning is when there is a positive association between an aversive US and neutral CS. The
underlying neurobiology of the amygdala supports both that of safety learning and fear learning.
A complex coordination can occur within nuclei of the amygdala and amongst the amygdala’s
many interconnections to other regions in order to coordinate the inhibition of fear conditioning
and avoidance pathways and excitation of associative learning processes reinforced via felt
reward (Kong, Monje, Hirsch, & Pollak, 2013; Adolphs, 2002; LeDeux, 2007; Feldman, Braun,
& Champagne, 2019; Preston, 2013; Marsh, 2016). This associative learning mechanism that
allows for flexible responding to context-dependent distressing stimuli provides a plausible
mechanism for the onset of fathering and alloparenting being precipitated and maintained via
experience-dependent change.
Longitudinal fear emotion processing
Whole-brain repeated measures comparisons of prenatal to postpartum indicated greater
activation in the bilateral occipital poles during the prenatal versus postpartum scan when
viewing fear faces versus baseline; and greater activation in the left parieto-occipital junction at
the postpartum scan compared to the prenatal scan when viewing fear faces versus baseline.
Contrary to our change hypotheses, there were no significant mean differences in neural
activation in the VTA, NAcc, pallidum, or PAG from the prenatal to postpartum period. The
prenatal versus postpartum activations are consistent with the literature on attentional visual
processing of threat-related stimuli that indicate threat cues recruit enhanced visual attention
55
(Pourtois, Schwartz, Seghier, Lazeyras, & Vuilleumier, 2006). The postpartum versus prenatal
findings had peak voxel activation at MNI coordinates of -44, -78, 34. These coordinates are
heterogeneously identified in the extant literature as the lateral occipital cortex, superior division
(Marshall et al., under review), the temporoparietal junction (TPJ; FeldmanHall, Mobbs, &
Dalgleish, 2014), and the parieto-occipital junction (POJ; Malaia, Talavage, & Wilbur, 2014).
The peak activation of our cluster is superior to the occipital sulcus, posterior to the
temporoparietal junction and inferior to the superior temporal gyrus indicating that POJ may be
the most appropriate anatomical distinction for this cluster. The heterogeneity of classification of
the area makes it somewhat challenging to interpret; however, these MNI coordinates are broadly
associated with mentalizing (Marshall et al., under review; FeldmanHall, Mobbs, & Dalgleish,
2014) and rapid visual object discrimination at the global level of face versus nonface (Nagy,
Greenlee, & Kovács, 2012) which may indicate enhanced visual attentional processing of fear
stimuli given its saliency. Notably, visual processing of emotional cues is a complex process that
involves rapid gestalt detection of the object-as-face, then distinction of static features on the
face, then more complex deciphering of what expression the static features are displaying. This
process does not happen simultaneously, but over the course of ~170ms (Sugase, Yamane, Ueno,
& Kawano, 1999) and is likely impacted at various stages of processing by modulation from
interconnected structures like the OFC and amygdala (Adolphs, 2002). Therefore, the
distinctions between prenatal and postpartum visual cortex activation may indicate changes in
the speed of processing of the stimuli, the familiarity or saliency of the stimuli, or the meaning of
the stimuli to the perceiver.
Further exploration of the relationship between PAG activation and fear recognition
accuracy from the prenatal to postpartum period indicated that prenatal PAG activation, but not
56
change in PAG activation, predicted change in fear recognition accuracy from prenatal to
postpartum. Prenatal PAG % signal change predicted change in unbiased hit rate fear
recognition, such that the average person is increasing in fear recognition from prenatal to
postpartum by 7%. Further examination of this model indicates that higher levels of prenatal
PAG activation are associated with less of an increase in fear recognition and at high enough
levels, a predicted decrease in fear recognition. These results control for how PAG activation
changes from prenatal to postpartum and where people start in terms of fear recognition;
therefore, it is unlikely that these results are simply a result of regression to the mean. Overall,
prenatal PAG activation is advantageous with regard to fear recognition prenatally, but higher
prenatal PAG activation is linked with less of an advantage, or even a disadvantage, when
becoming a parent. Replication of these results is important for understanding their implications;
however, these findings support that PAG activation in response to fearful facial expressions is
likely associated with dynamic changes that accompany the transition to parenthood (Brethel-
Haurwitz et al., 2017). Further research is needed with larger sample sizes and with
neuroimaging data linked to actual parenting behavior to further understand this association.
Limitations and Future Directions
It should be noted that the existing MRI analysis programs are limited in their ability to
compute complex higher-level statistics such as structural equation modeling or path analysis
(Madhyastha et al., 2017; King et al., 2017), both of which are ideal for taking a whole-brain
network-based approach to longitudinal MRI analyses. Most analysis packages compute voxel-
wise or cluster-wise statistics within the general linear model framework. The NIH’s Analysis of
Functional Neuroimaging software and FSL are capable of a simplified general mixed linear
model, which can allow for a repeated measures analysis of variance with time modeled as the
57
fixed effects within-subjects variable. However, this methodology is further limited by the
current study’s two-timepoint collection of data. Within the linear model and with the constraint
of two timepoints, the relationship will be constrained to linearity. Looking at within subject
change (random effects of time) would result in overestimation of the model (Madhyastha et al.,
2017; King et al., 2017). Therefore, we cannot compute within-subject individual differences of
change, but only whether between-subjects change (or stability) occurred over time. This limits
the information that can be gleaned from our data. Future research would benefit from
monitoring change over time with shorter timepoints between scans and multiple scans so as to
look at nonlinear trajectories of change within and across alloparents.
Additionally, the current study investigated change over a time period of about nine
months. When investigating change over time it is important to understand how developmental
processes may influence that change, what the change is that is expected over time, and when in
development this change might occur (King et al., 2017). No emotion perception studies to date
have focused on the male transition to first-time parenthood in order to study functional
neuroplasticity in adulthood, despite the overwhelming evidence that significant biological
changes occur (Feldman, Braun, & Champagne, 2019; Kim et al., 2016). This is likely due to the
fact that the parenting literature has most commonly focused on the mother, and volitional
neuroimaging during pregnancy is contraindicated. Therefore, most existing longitudinal studies
of the paternal brain have not spanned longer than a four-month period and have all been
conducted with postpartum samples only. As a result, it is unknown whether the current study’s
nine-month lag time between the two visits is optimized for emotion perception processes.
Lastly, given that the proposed study involves change over time, the following
methodological constraints were put in place in an effort to reduce measurement error
58
(Vijayakumar, Mills, Alexander-Bloch, Tamnes, & Whittle, 2017): the same MRI equipment,
20-channel head coil, button boxes, behavioral task presentation and software, and MRI
hardware, were used at both the prenatal and postpartum visits. Additionally, the neuroimaging
task chosen used stimuli that are stable over time. The stimuli lack ecological validity; however,
there is a plethora of data supporting the robustness and accuracy of their ability to serve as
emotional facial cues for the stated emotion. Choosing a stable task with robust findings was
optimized over ecological validity given the longitudinal design of this project (Herting, Gautam,
Chen, Mezher, Vetter, 2018). All data processing steps and computer processing operating
systems were kept the same (e.g. no software updates) until analyses for both prenatal and
postpartum data were completed. Even with these measures in place, the current study took place
over four years and had an interval of about nine months between visits where participants went
through an immense transition in life. Future studies may want to explore potential moderators
and mediators, such as sleep, history of psychopathology, and relational status, that may impact
the father’s engagement in alloparental care and the impact of that alloparental care on neural
plasticity and change.
Conclusions
The current study highlights the value in studying a variety of emotional stimuli, such as
emotional adult face perception, in order to understand mechanisms that may facilitate the onset
and maintenance of alloparental care. Importantly, future studies would benefit from including
measures of alloparental engagement in offspring rearing and relationship functioning with the
mother to understand how these neural and behavioral correlates of fear processing map on to
relational patterns with offspring and the coparent. Additionally, while further research is needed
to corroborate these data, these data support that the paternal brain may undergo context-specific,
59
experience-dependent neuroplasticity that may serve to support the onset and maintenance of
alloparental caregiving behaviors.

60
General Discussion
This dissertation is the first prospective, longitudinal neuroimaging study of expectant
fathers followed into the postpartum period. It is also the first study to specifically examine fear
perception in a sample of individuals transitioning into parenthood. In Paper 1 we found that
expectant fathers responded faster when adult fear faces were paired with infant faces than with
other face emotion pairings, supporting the theory that fear expressions may operate as
kindchenschema that resemble human infant faces. Relatedly, we found that infant faces elicited
behavioral approach on an arm flexion task, whereas responses to adult fear faces were context-
dependent, such that fear faces elicited faster behavioral approach when paired with infant faces,
but faster behavioral avoidance when paired with other adult faces. Additionally, this paper
identified individual differences in empathy and their impact on behavioral approach-avoidance
action tendencies in response to fear and infant stimuli. Expectant fathers who self-reported
higher perspective-taking had faster behavioral approach responses for fear stimuli. Expectant
fathers who self-reported higher empathic concern had faster overall responses to infant stimuli.
In Paper 2, we examined the neural and behavioral correlates of fear emotion processing
across the transition to fatherhood in a sample of first-time expectant fathers followed from the
prenatal to postpartum periods. Prenatally, expectant fathers exhibited neural activation to fear vs
baseline in regions typically associated with fear and threat-related processing in healthy
populations. A region-of-interest analysis of the PAG and explicit fear recognition accuracy
indicated a positive association, such that higher PAG activation was associated with higher fear
recognition accuracy. Postpartum whole-brain analyses indicated similar activation patterns to
the prenatal period, with the exclusion of the PAG and addition of the putamen, caudate, and
pallidum, regions associated with associative learning and reward. This pattern of response is
61
also similar to that seen in response to infant faces. A region-of-interest analysis of the PAG and
explicit fear recognition accuracy in the postpartum period revealed a marginally significant,
inverse association such that higher PAG activation was associated with lower fear recognition
accuracy. Given that the postpartum fathers included six fathers scanned post-COVID with
slightly longer between scan times, we ran two whole brain analyses examining time lag between
scans and infant age as covariates. No associations were found between time lag between scans,
infant age, and postpartum brain activation during fear vs baseline contrasts. Longitudinal
statistical comparisons indicated greater activation in the occipital poles during the prenatal
period compared to postpartum, and greater activation in the parieto-occipital junction during the
postpartum period compared to prenatal. To our knowledge, this is the first longitudinal
investigation of the neural and behavioral characteristics of men before and after the birth of
their first child and the first paternal brain neuroimaging study to use non-infant cues in an effort
to assess how men perceive emotional stimuli that may be relevant to alloparenting. We found
preliminary support that this transition is accompanied by neural changes that may be associated
with inhibition of fear learning pathways and excitation of social reinforcement learning and felt
reward in response to fear faces.
Contributions to the literature
In 1994, Cosmides and Tooby called for a more rigorous cognitive neuroscience, urging
the field to become a theoretically based discipline rooted in evolutionary theory. They
advocated that if cognitive neuroscience was the study of the mind and its biological function,
then understanding the evolutionary origins of specific aspects of the human mind selected for
over time was imperative for understanding the underlying structural foundations of neural
networks and behaviors (Cosmides & Tooby, 1994). More recently, Scott-Philips, Dickens, and
62
West (2011) argued that identifying why a behavior was naturally selected for (i.e. ultimate
mechanisms) and how it continues to persist (i.e. proximate mechanisms) can help to elucidate
human behavioral traits. Indeed, for over a century, psychologists, anthropologists,
sociobiologists, and ethologists have been puzzled by the question of why an individual might
engage in sustained, costly allocation of resources to others aside from the self (Darwin, 1871;
Clutton-Brock, 1989). One prevailing theory first put forth by Darwin and Lorenz is that the
parent-infant relationship is the primary driver of biological and behavioral evolution in humans,
which then serves as the foundation for the human ability to engage in sociocognitive processes
required for societal cooperation, such as empathy and compassion. The ability to empathize
with another person’s distress and to respond with a desire to alleviate their distress is essential
for effective caregiving. In understanding alloparental care, we may also learn more about
empathy, costly helping, and how to build more inclusive, cooperative societies (Marsh, 2016;
Preston, 2013; Feldman, Braun, & Champagne, 2019).
This dissertation contends that engaging in alloparental care may shape emotion
processing for a broader array of stimuli than just infant faces. Expanding our investigation of
fathers to include fear processing and emotion processing more broadly may provide unique
insight into the biological mechanisms subserving social cognition in alloparental species. These
data support a theoretically-driven call to use emotional stimuli that are well-validated in
evoking individual differences, such as fear faces. Incorporating such stimuli into prospective
parenting brain research may allow for greater variability in participant responses and better
understanding of the range of alloparental behavioral variation and biobehavioral plasticity.
Strengths, limitations, and future directions
Given that fearful facial expressions are highly salient, but often ambiguous as to the
63
source of threat, the context of the perceiver and displayer of emotions likely influences how the
perceiver interprets the expression. Therefore, these data represent fear responding in a sample of
males who are cohabiting with their partners through pregnancy and into the postpartum period
once the infant is born. Fathers’ exposure to a lactating mother and time with the infant has been
associated with the magnitude of hormonal change fathers will go through in the postpartum
period (Feldman, Braun, & Champagne, 2019) and has also been associated with differential
neural activity between primary caregiving and secondary caregiving fathers who presumably
send differing amounts of time with the infant (Abrahem et al., 2014). Primary fathers’ brain
activity more closely resembled primary caregiving mothers than other secondary caregiving
fathers (Abrahem et al., 2014). Therefore, findings from both Study 1 and Study 2 should be
interpreted with reference to the context of our participants: our sample was ethnically diverse,
well-educated, mostly married, all cohabiting in a major urban city in the U.S., and generally
happily together.
Strengths of this study include the longitudinal, prospective design, use of standardized
stimuli, a specific focus on fathers as alloparents, and the novelty of expanding the parenting
brain literature to include emotion processing beyond infant faces. This was the first parenting
brain study to investigate fear perception. Limitations include our sample size, delay of six
participants due to COVID-19 pandemic, and only having two timepoints thereby linearly
constraining the findings to group differences. Additionally, the stimuli used were chosen based
off of their robust investigation in the larger literature thereby ensuring specificity and stability
over time in displaying the intended emotions. However, it is strongly recommended that future
studies investigate emotion perception in alloparents using more racially and ethnically diverse
emotional facial stimuli. Lastly, the current study analyzed neuroimaging data within the general
64
linear model using linearly combined contrasts. Future studies may want to consider network
modeling and multivariate approaches, as well as diffusion tensor imaging and structural MRI in
order to capture the dynamic processing of emotion perception amongst brain regions and to
understand how connectivity and volume are related to changes in emotion perception across the
transition to parenthood.
Conclusion
Overall, this dissertation provides a novel contribution to the parenting brain research,
emphasizing the importance of fathers as alloparents. This is the first study, to our knowledge, to
present longitudinal data from men before and after the birth of their first child. Our data support
the transition to fatherhood as a developmental period in adulthood characterized by potentially
meaningful shifts in face emotion processing, a process pivotal for higher level functions like
social cognition and cooperative parenting. These findings can set the stage for investigating the
impacts of alloparental care on adults’ emotion perception more broadly (beyond fear faces) with
the hope that greater understanding of the impacts of alloparenting-on the alloparent and child-
may have direct benefits for not only increasing parental involvement and positive outcomes for
children, but also for increased social cooperation characterized by caring for others in distress
beyond the immediate family (Preston, 2013; Marsh, 2016; Feldman, Braun, & Champagne,
2019).

65
References
Adolphs, R. (2002). Neural systems for recognizing emotion. Current Opinion in Neurobiology,
12(2), 169–177. https://doi.org/10.1016/S0959-4388(02)00301-X
Adolphs, R. (2002). Recognizing emotion from facial expressions: Psychological and
neurological mechanisms. Behavioral and Cognitive Neuroscience Reviews, 1(1), 21–61.
Ainsworth, M. (1979). Infant-mother attachment. American Psychologist, 34, 932-937.
Aradhye, C., Vonk, J., & Arida, D. (2015). Adults’ responsiveness to children’s facial
expressions. Journal of Experimental Child Psychology, 135, 56–71.
https://doi.org/10.1016/j.jecp.2015.02.006
Atzil, S., Hendler, T., Zagoory-Sharon, O., Winetraub, Y., & Feldman, R. (2012). Synchrony and
specificity in the maternal and the paternal brain: Relations to oxytocin and vasopressin.
Journal of the American Academy of Child and Adolescent Psychiatry, 51(8), 798–811.
https://doi.org/10.1016/j.jaac.2012.06.008
Babchuk, W. A., Hames, R. B., & Thompson, R. A. (1985). Sex differences in the recognition of
infant facial expressions of emotion: The primary caretaker hypothesis. Ethology and
Sociobiology, 6(2), 89–101. https://doi.org/10.1016/0162-3095(85)90002-0
Bakermans-Kranenburg, M. J., Lotz, A., Alyousefi-van Dijk, K., & van IJzendoorn, M. (2019).
Birth of a Father: Fathering in the First 1,000 Days. Child Development Perspectives,
13(4), 247–253. https://doi.org/10.1111/cdep.12347
Bard P. 1928. A diencephalic mechanism for the expression of rage with special reference to the
sympathetic nervous system. American Journal of Physiology. 84, 490–515.
66
Barsalou, L. W. (2016). Situated conceptualization: Theory and applications. In Y. Coello & M.
H. Fischer (Eds.), Foundations of embodied cognition: Perceptual and emotional
embodiment (pp. 11–37). East Sussex: Psychology Press.
Bartels, A., & Zeki, S. (2004). The neural correlates of maternal and romantic love. NeuroImage,
21(3), 1155–1166. https://doi.org/10.1016/j.neuroimage.2003.11.003
Batson, C. D., & Shaw, L. L. (1991). Evidence for altruism: Toward a pluralism of prosocial
motives. Psychological Inquiry, 2(2), 107–122.
https://doi.org/10.1207/s15327965pli0202
Beckmann, C., Jenkinson, M., and Smith, S. (2003). General multi-level linear modelling for
group analysis in FMRI. NeuroImage, 20, 1052-1063.
Blair, R. J. (2008). Fine cuts of empathy and the amygdala: dissociable deficits in psychopathy
and autism. Quarterly Journal of Experimental Psychology, 61, 157–170.
Bornstein, M. H., Tamis-LeMonda, C. S., Hahn, C. S., & Haynes, O. M. (2008). Maternal
Responsiveness to Young Children at Three Ages: Longitudinal Analysis of a
Multidimensional, Modular, and Specific Parenting Construct. Developmental
Psychology, 44(3), 867–874. https://doi.org/10.1037/0012-1649.44.3.867
Bowlby, J., 1969. Attachment and Loss, vol. 1. Basic Books, New York
Brethel-Haurwitz, K. M., O’Connell, K., Cardinale, E. M., Stoianova, M., Stoycos, S. A., Lozier,
L. M., … Marsh, A. A. (2017). Amygdala–midbrain connectivity indicates a role for the
mammalian parental care system in human altruism. Proceedings of the Royal Society B:
Biological Sciences, 284(1865), 20171731. https://doi.org/10.1098/rspb.2017.1731
67
Brüne, M., & Brüne-Cohrs, U. (2006). Theory of mind-evolution, ontogeny, brain mechanisms
and psychopathology. Neuroscience and Biobehavioral Reviews, 30(4), 437–455.
https://doi.org/10.1016/j.neubiorev.2005.08.001
Cabrera, N. J., Tamis-LeMonda, C. S., Bradley, R. H., Hofferth, S., Lamb, M. E. (2000)
Fatherhood in the twenty-first century. Child Development, 71(1), 127-136.
Cannon W. 1927. The James-Lange theory of emotions: a critical examination and an alternative
theory. American Journal of Psychology, 39, 106–24.
Caria, A., de Falco, S., Venuti, P., Lee, S., Esposito, G., Rigo, P., … Bornstein, M. H. (2012).
Species-specific response to human infant faces in the premotor cortex. NeuroImage,
60(2), 884–893. https://doi.org/10.1016/j.neuroimage.2011.12.068
Chartrand, T. L., & Bargh, J. A. (1999). The chameleon effect: The perception-behavior link and
social interaction. Journal of Personality and Social Psychology, 76, 893–910.
doi:10.1037/0022-3514.76.6.893
Chen, M. & Bargh, J. (1999). Consequences of automatic evaluation: Immediate behavioral
predispositions to approach or avoid the stimulus. Personality and Social Psychology
Bulletin. 25(2):215– 224.
Clutton-Brock, T. H. (1989). Mammalian mating systems. Proceedings of the Royal Society of
London B, 236, 339-372.
Cosmides, L., & Tooby, J. (1994). Beyond intuition and instinct blindness: toward an
evolutionarily rigorous cognitive science. Cognition, 50(1–3), 41–77.
https://doi.org/10.1016/0010-0277(94)90020-5
Darwin C. (1871). The descent of man and selection in relation to sex. London, UK: John
Murray.
68
Davis, M. H. (1983). Measuring individual differences in empathy: Evidence for a
multidimensional approach. Journal of Personality and Social Psychology, 44(1), 113–
126. https://doi.org/10.1037/0022-3514.44.1.113
Decety, J., Lewis, K. L., & Cowell, J. M. (2015) Specific electrophysiological components
disentangle affective sharing and empathic concern in psychopathy. Journal of
Neurophysiology, 114, 493-504.
Dricu, M., & Frühholz, S. (2016). Perceiving emotional expressions in others: Activation
likelihood estimation meta-analyses of explicit evaluation, passive perception and
incidental perception of emotions. Neuroscience and Biobehavioral Reviews, 71, 810–
828. https://doi.org/10.1016/j.neubiorev.2016.10.020
Duvarci S, & Pare D., (2014). Amygdala microcircuits controlling learned fear. Neuron, 82,
966–980.
Dykas, M. J. & Cassidy, J. (2011). Attachment and the processing of social information across
the life span: theory and evidence. Psychological Bulletin, 137(1), 19-46.
Eisenberg, N., & Miller, P. A. (1987). The relation of empathy to prosocial and related
behaviors. Psychological Bulletin, 101(1), 91–119. https://doi.org/10.1037/0033-
2909.101.1.91
Ekman, P., & Friesen, W. (1976) Pictures of Facial Affect (Consulting Psychologists, Palo Alto,
CA).
Feldman, R. (2012). Oxytocin and social affiliation in humans. Hormones and Behavior, 61(3),
380–391. https://doi.org/http://dx.doi.org/10.1016/j.yhbeh.2012.01.008
69
Feldman, R. (2015). The adaptive human parental brain: implications for children’s social
development. Trends in Neurosciences, 38(6), 387–399.
https://doi.org/10.1016/j.tins.2015.04.004
Feldman, R. (2016). The neurobiology of mammalian parenting and the biosocial context of
human caregiving. Hormones and Behavior, 77, 3–17.
https://doi.org/10.1016/j.yhbeh.2015.10.001
Feldman, R., Braun, K., & Champagne, F. A. (2019). The neural mechanisms and consequences
of paternal caregiving. Nature Reviews Neuroscience, 20(4), 205–224.
https://doi.org/10.1038/s41583-019-0124-6
Feldmanhall, O., Dalgleish, T., Evans, D., & Mobbs, D. (2015). NeuroImage Empathic concern
drives costly altruism. NeuroImage, 105, 347–356.
https://doi.org/10.1016/j.neuroimage.2014.10.043
FeldmanHall, O., Mobbs, D., & Dalgleish, T. (2014). Deconstructing the brain’s moral network:
Dissociable functionality between the temporoparietal junction and ventro-medial
prefrontal cortex. Social Cognitive and Affective Neuroscience, 9(3), 297–306.
https://doi.org/10.1093/scan/nss139
Förster J (1998). Effect of motor perceptions on affective judgment of attractive and unattractive
portraits. Zeitschrift für Experimentelle Psychologie, 45, 167-182.
Fusar-Poli, P., Placentino, A., Carletti, F., Landi, P., Allen, P., Surguladze, S., … Politi, P.
(2009). Functional atlas of emotional faces processing : a voxel-based meta-analysis of
105 functional magnetic resonance imaging studies. Journal of Psychiatry &
Neuroscience, 34(6), 418–432.
70
Glocker, M. L., Langleben, D. D., Ruparel, K., Loughead, J. W., Valdez, J. N., Griffin, M. D., …
Gur, R. C. (2009). Baby schema modulates the brain reward system in nulliparous
women. Proceedings of the National Academy of Sciences, 106(22), 9115–9119.
https://doi.org/10.1073/pnas.0811620106
Glocker, M. L., Langleben, D. D., Ruparel, K., Loughead, J. W., Gur, R. C., & Sachser, N.
(2009). Baby Schema in Infant Faces Induces Cuteness Perception and Motivation for
Caretaking in Adults. Ethology, 115(3), 257–263.
Greenwald, A. G., Nosek, B. A., & Banaji, M. R. (2003). Understanding and using the implicit
association test: I. An improved scoring algorithm. Journal of Personality and Social
Psychology, 85(2), 197–216. https://doi.org/10.1037/0022-3514.85.2.197
Hammer, J. L., & Marsh, A. A. (2015). Why do fearful facial expressions elicit behavioral
approach? Evidence from a combined approach-avoidance implicit association test.
Emotion, 15(2), 223–231.
Hampson, E., van Anders, S. M., & Mullin, L. I. (2006). A female advantage in the recognition
of emotional facial expressions: test of an evolutionary hypothesis. Evolution and Human
Behavior, 27(6), 401–416. https://doi.org/10.1016/j.evolhumbehav.2006.05.002
Hariri, A. R., Bookheimer, S. Y., & Mazziotta, J. C. (2000). Modulating emotional responses:
Effects of a neocortical network on the limbic system. NeuroReport, 11(1), 43–48.
https://doi.org/10.1097/00001756-200001170-00009
Hatfield, E., Cacioppo, J.T., & Rapson, R.L. (1994). Emotional contagion. Cambridge, England:
Cambridge University Press.
71
Herting, M. M., Gautam, P., Chen, Z., Mezher, A., & Vetter, N. C. (2018). Test-retest reliability
of longitudinal task-based fMRI: Implications for developmental studies. Developmental
Cognitive Neuroscience, 33, 17–26. https://doi.org/10.1016/j.dcn.2017.07.001
Hoekzema, E., Barba-Muller, E., Pozzobon, C., Picado, M., Lucco, F., Garcia-Garcia, D., …
Vilarroya, O. (2017). Pregnancy leads to long-lasting changes in human brain structure.
Nature Neuroscience, 20(2), 287–296. https://doi.org/10.1038/nn.4458
Hoekzema, E., Tamnes, C. K., Berns, P., Barba-Müller, E., Pozzobon, C., Picado, M., …
Carmona, S. (2020). Becoming a mother entails anatomical changes in the ventral
striatum of the human brain that facilitate its responsiveness to offspring cues.
Psychoneuroendocrinology, 112(October 2019), 104507.
https://doi.org/10.1016/j.psyneuen.2019.104507
Hrdy, S. B. (2009). Mothers and Others: The Evolutionary Origins of Mutual Understanding.
Belknap Press: Cambridge, MA.
James W. (1884). What is an emotion? Mind, 9, 188–205.
James W. (1894). The physical basis of emotion. Psychology Reviews, 1, 516–29.
Jamil, Z. (2014). Empathy: A motivated account. Psychological Bulletin, 140(6), 1608–1647.
https://doi.org/10.1037/a0037679
Jenkinson, M. & Smith, S. (2001). A global optimisation method for robust affine registration of
brain images. Medical Image Analysis, 5(2), 143-156.
Jenkinson, M., Bannister, P., Brady, M., and Smith, S. (2002). Improved optimisation for the
robust and accurate linear registration and motion correction of brain images.
NeuroImage, 17(2), 825-84.
72
Jia, Y. C., Ding, F. Y., & Cheng, G. (2021). Adults ’ responses to infant faces : Neutral infant
facial expressions elicit the strongest baby schema effect. Quarterly Journal of
Experimental Psychology, 0(00), 1–19. https://doi.org/10.1177/1747021820981862
Kim, P. (2016). Human Maternal Brain Plasticity: Adaptation to Parenting. New Directions for
Child and Adolescent Development, 2016(153), 47–58.
https://doi.org/10.1002/cad.20168
Kim, P., Rigo, P., Mayes, L. C., Feldman, R., Leckman, J. F., & Swain, J. E. (2014). Neural
Plasticity in Fathers of Human Infants. Social Neuroscience, 9(5), 522–535.
https://doi.org/10.3174/ajnr.A1256.Functional
Kim, P., Strathearn, L., & Swain, J. E. (2016). The maternal brain and its plasticity. Hormones
and Behavior, 77, 113–123. https://doi.org/10.1016/j.yhbeh.2015.08.001
King, K. M., Littlefield, A. K., McCabe, C. J., Mills, K. L., Flournoy, J., & Chassin, L. (2018).
Longitudinal modeling in developmental neuroimaging research: Common challenges,
and solutions from developmental psychology. Developmental Cognitive Neuroscience,
(November).
Kong, E., Monje, F. J., Hirsch, J., & Pollak, D. D. (2014). Learning not to fear: Neural correlates
of learned safety. Neuropsychopharmacology, 39(3), 515–527.
https://doi.org/10.1038/npp.2013.191
Lambert, K. G. (2012). The parental brain: Transformations and adaptations. Physiology and
Behavior, 107(5), 792–800. https://doi.org/10.1016/j.physbeh.2012.03.018
Lange C. (1922). The Emotions. Baltimore, MD: Williams & Wilkins
LeDoux, J. (2007). The amygdala. Current Biology, 17(20), 440–452.
https://doi.org/10.1093/ojls/12.3.440
73
Leibenluft, E., Gobbini, M. I., Harrison, T., & Haxby, J. V. (2004). Mothers’ neural activation in
response to pictures of their children and other children. Biological Psychiatry, 56(4),
225–232. https://doi.org/10.1016/j.biopsych.2004.05.017
Leuner, B., & Sabihi, S. (2016). The birth of new neurons in the maternal brain: hormonal
regulation and functional implications. Frontiers in Neuroendocrinology, 41, 99–113.
https://doi.org/10.1126/science.1249098.Sleep
Linnman, C., Moulton, E. A., Barmettler, G., Becerra, L., & Borsook, D. (2012). Neuroimaging
of the periaqueductal gray: State of the field. Neuroimage, 60(1), 505–522.
https://doi.org/10.1016/j.neuroimage.2011.11.095.Neuroimaging
Lorenz, K. (1943). Die angeborenen Formenmöglicher Erfahrung. (Innate form of potential
experience). Z. Tierpsychol, 5, 235–309.
Lorenz, K. (1971) Studies in Animal and Human Behavior, vol. II. London: Methuen.
Lozier, L. M., Cardinale, E. M., VanMeter, J.W., Marsh, A. A. (2014). Mediation of the
relationship between callous–unemotional traits and proactive aggression by amygdala
response to fear among children with conduct problems. JAMA Psychiatry, 71, 627–636.
Lundqvist D, Flykt A, Öhman A. The Karolinska Directed Emotional Faces—KDEF, CDROM
from Department of Clinical Neuroscience, Psychology section, Karolinska Institutet,
ISBN 91-630-7164-9. 1998.
Madhyastha, T., Peverill, M., Koh, N., McCabe, C., Flournoy, J., Mills, K., … McLaughlin, K.
A. (2017). Current methods and limitations for longitudinal fMRI analysis across
development. Developmental Cognitive Neuroscience, (10), 0–1.
https://doi.org/10.1016/j.dcn.2017.11.006
74
Malaia, E., Talavage, T. M., & Wilbur, R. B. (2014). Functional connectivity in task-negative
network of the Deaf: Effects of sign language experience. PeerJ, 2, 1–12.
https://doi.org/10.7717/peerj.446
Marsh, A. A. (2012). Empathy and compassion: A cognitive neuroscience perspective. Empathy:
From Bench to Bedside, 191–206.
Marsh, A. A. (2013). What can we learn about emotion by studying psychopathy? Frontiers in
Human Neuroscience, 7, 181.
Marsh, A. A. (2016). Understanding amygdala responsiveness to fearful expressions through the
lens of psychopathy and altruism. Journal of Neuroscience Research, 94(6), 513–525.
https://doi.org/10.1002/jnr.23668
Marsh, A. A., & Ambady, N. (2007). The influence of the fear facial expression on prosocial
responding. Cognition and Emotion, 21(2), 225–247.
https://doi.org/10.1080/02699930600652234
Marsh, A. A., Ambady, N., & Kleck, R. E. (2005). The effects of fear and anger facial
expressions on approach- and avoidance-related behaviors. Emotion, 5(1), 119–124.
https://doi.org/10.1037/1528-3542.5.1.119
Marsh, A. A., Stoycos, S. A., Brethel-Haurwitz, K. M., Robinson, P., VanMeter, J. W.,
Cardinale, E. M. (2014). Neural and cognitive characteristics of extraordinary altruists.
Proceedings of the National Academy of Science, 111, 15036– 15041.
Marsh, A. A., Yu, H. H., Pine, D. S., Gorodetsky, E. K., Goldman, D., & Blair, R. J. R. (2012).
The influence of oxytocin administration on responses to infant faces and potential
moderation by OXTR genotype. Psychopharmacology, 224(4), 469–476.
https://doi.org/10.1007/s00213-012-2775-0
75
Matsumoto, D. & Ekman, P. (2008). Facial expression analysis. Scholarpedia, 3, 4237.
Monte, O. D., Krueger, F., Solomon, J. M., Schintu, S., Knutson, K. M., Strenziok, M., …
Grafman, J. (2013). A voxel-based lesion study on facial emotion recognition after
penetrating brain injury. Social Cognitive and Affective Neuroscience, 8(6), 632–639.
https://doi.org/10.1093/scan/nss041
Morris JS, Frith CD, Perrett DI, Rowland D, Young AW, Calder AJ, Dolan RJ. 1996. A
differential neural response in the human amygdala to fearful and happy facial
expressions. Nature, 383, 812–815.
Murray EA. 2007. The amygdala, reward, and emotion. Trends in Cognitive Science, 11, 489–
497.
Nagy, K., Greenlee, M. W., & Kovács, G. (2012). The lateral occipital cortex in the face
perception network: An effective connectivity study. Frontiers in Psychology, 3, 1–12.
https://doi.org/10.3389/fpsyg.2012.00141
Namburi, P., Beyeler, A., Yorozu, S., Calhoon, G. G., Halbert, S. A., Wichmann, R., … Tye, K.
M. (2015). A circuit mechanism for differentiating positive and negative associations.
Nature, 520(7549), 675–678. https://doi.org/10.1038/nature14366
Neumann, R., & Strack, F. (2000). ‘‘Mood contagion’’: The automatic transfer of mood between
persons. Journal of Personality and Social Psychology, 79, 211–223.
Niedenthal, P. M. (2007). Embodying Emotion. Science, 316, 1002–1006.
Numan, M., & Insel, T. R. (2003). The neurobiology of parental behavior. New York, NY:
Springer-Verlag.
76
Numan, M., & Young, L. J. (2016). Neural mechanisms of mother-infant bonding and pair
bonding: Similarities, differences, and broader implications. Hormones and Behavior, 77,
98–112. https://doi.org/10.1016/j.yhbeh.2015.05.015
Parsons, C. E., Nummenmaa, L., Sinerva, E., Korja, R., Kajanoja, J., Young, K. S., … Karlsson,
L. (2019). Investigating the Effects of Perinatal Status and Gender on Adults’ Responses
to Infant and Adult Facial Emotion. Emotion, 21(2), 337–349.
https://doi.org/10.1037/emo0000698
Pearson, R. M., Lightman, S. L., & Evans, J. (2009). Emotional sensitivity for motherhood: Late
pregnancy is associated with enhanced accuracy to encode emotional faces. Hormones
and Behavior, 56(5), 557–563. https://doi.org/10.1016/j.yhbeh.2009.09.013
Pearson, R., Lewis, M.B., 2005. Fear recognition across the menstrual cycle. Hormones and
Behavior, 47, 267–271.
Pollak, D.D., Monje, F.J., Zuckerman, L., Denny, C.A., Drew, M.R., Kandel, E.R. (2008). An
animal model of a behavioral intervention for depression. Neuron, 60, 149–161.
Pourtois, G., Schwartz, S., Seghier, M. L., Lazeyras, F., & Vuilleumier, P. (2006). Neural
systems for orienting attention to the location of threat signals: An event-related fMRI
study. NeuroImage, 31(2), 920–933. https://doi.org/10.1016/j.neuroimage.2005.12.034
Preston, S. D. (2013). The origins of altruism in offspring care. Psychological Bulletin, 139(6),
1305–1341. https://doi.org/10.1037/a0031755
Preston, S. D., & De Waal, F. (2002). Empathy: Its ultimate and proximate bases. Behavioral
and Brain Sciences, 25, 1–20.
Proverbio, A. M., Brignone, V., Matarazzo, S., Del Zotto, M., & Zani, A. (2006). Gender and
parental status affect the visual cortical response to infant facial expression.
77
Neuropsychologia, 44(14), 2987–2999.
https://doi.org/10.1016/j.neuropsychologia.2006.06.015
Rilling, J. K. (2013). The neural and hormonal bases of human parental care. Neuropsychologia,
51(4), 731–747. https://doi.org/10.1016/j.neuropsychologia.2012.12.017
Rilling, J. K., & Young, L. J. (2014). The biology of mammalian parenting and its effect on
offspring social development. Science, 345(6198), 771–776.
https://doi.org/10.1126/science.1252723
Rimé, B. (2009). Emotion elicits the social sharing of emotion: Theory and empirical review.
Emotion Review, 1(1), 60–85. https://doi.org/10.1177/1754073908097189
Rotteveel, M. & Phaf, R.H. (2004). Automatic affective evaluation does not automatically
predispose for arm flexion and extension. Emotion, 4(2), 156–172.
Ruffman,T., Henry, J.D., Livingstone, V., Phillips, L.H. (2008). A meta-analysis of emotion
recognition and aging: Implications for neuropsychological models of aging.
Neuroscience and Biobehavioral Reviews, 34(4), 863–881.
Rutherford, H. J. V, Wallace, N. S., Laurent, H. K., & Mayes, L. C. (2015). Emotion regulation
in parenthood. Developmental Review, 36, 1–14.
https://doi.org/10.1016/j.dr.2014.12.008.Emotion
Salzman, C. D., & Fusi, S. (2010). Emotion, cognition, and mental state representation in
amygdala and prefrontal cortex. Annual Review of Neuroscience, 33, 173–202.
https://doi.org/10.1146/annurev.neuro.051508.135256
Sarkadi, A., Kristiansson, R., Oberklaid, F., Bremberg, S. (2008). Fathers’ involvement and
children’s developmental outcomes: a systematic review of longitudinal studies. Acta
Paediatr, 97(2), 153–158
78
Schwarz, N. (2017). Embodied Cognition and the Construction of Attitudes. Handbook of
Attitudes, 1–57.
Seibt B, Neumann R, Nussinson R, Strack F. (2008). Movement direction or change in distance?
Self- and object-related approach–avoidance motions. Journal of Experimental Social
Psychology, 44(3), 713–720
Seidel, E.M., Habel, U., Kirschner, M., Gur, R.C., Derntl, B. (2010). The impact of facial
emotional expressions on behavioral tendencies in women and men. Journal of
Experimental Psychology: Human Perception and Performance, 36(2), 500–507.
Senese, V. P., De Falco, S., Bornstein, M. H., Caria, A., Buffolino, S., & Venuti, P. (2013).
Human infant faces provoke implicit positive affective responses in parents and non-
parents alike. PLoS ONE, 8(11). https://doi.org/10.1371/journal.pone.0080379
Smith, S. (2002). Fast robust automated brain extraction. Human Brain Mapping, 17(3), 143-
155.
Somerville, L. H., Fani, N., & MaClure-Tone, E. B. (2011). Behavioral and neural representation
of emotional facial expressions across the lifespan. Developmental Neuropsychology,
36(4), 408–428. https://doi.org/10.1080/87565641.2010.549865.Behavioral
Stockley, P., & Hobson, L. (2016). Paternal care and litter size coevolution in mammals.
Proceedings of the Royal Society B: Biological Sciences, 283(1829), 20160140.
https://doi.org/10.1098/rspb.2016.0140
Stoycos, S. A., Piero, L. Del, Margolin, G., Kaplan, J. T., & Saxbe, D. E. (2017). Neural
correlates of inhibitory spillover in adolescence: Associations with internalizing
symptoms. Social Cognitive and Affective Neuroscience, 12(10), 1637–1646.
https://doi.org/10.1093/scan/nsx098
79
Sugase, Y., Yamane, S., Ueno, S., Kawano, K. (1999). Global and fine information coded by
single neurons in the temporal visual cortex. Nature, 400, 869-872.
Sullivan, M. W. (2014). Infant expressions in an approach/withdrawal framework. The Journal
of Genetic Psychology, 175(6), 472–493. https://doi.org/10.1080/00221325.2014.964169
Swain, J. E., Dayton, C. J., Kim, P., Tolman, R. M., & Volling, B. L. (2014). Progress on the
paternal brain: Theory, animal models, human brain research, and mental health
implications. Infant Mental Health Journal, 35(5), 394–408.
https://doi.org/http://dx.doi.org/10.1002/imhj.21471
Swain, J. E., Konrath, S., Brown, S. L., Finegood, E. D., Akce, L. B., Dayton, C. J., & Ho, S. S.
(2012). Parenting and beyond: Common neurocircuits underlying parental and
altruistic caregiving. Parenting, 12(2–3), 115–123.
https://doi.org/10.1080/15295192.2012.680409
Taylor, J.M. & Whalen, P.J. (2014). Fearful, but not angry, expressions diffuse attention to
peripheral targets in an attentional blink paradigm. Emotion, 14, 462–468.
Vijayakumar, N., Mills, K. L., Alexander-Bloch, A., Tamnes, C. K., & Whittle, S. (2017).
Structural brain development: a review of methodological approaches and best practices.
Developmental Cognitive Neuroscience, 45, 1.
Vrijsen, J.N., van Oostrom, I., Speckens, A., Becker, E.S., Rinck, M. (2013). Approach and
avoidance of emotional faces in happy and sad mood. Cognitive Therapy and Research,
37(1), 1–6.
Wager, T. D., & Nichols, T. E. (2003). Optimization of experimental design in fMRI: A general
framework using a genetic algorithm. NeuroImage, 18(2), 293–309.
https://doi.org/10.1016/S1053-8119(02)00046-0
80
Wagner, H. L. (1993). On measuring performance in category judgment studies of nonverbal
behavior. Journal of Nonverbal Behavior, 17, 3-28. https://doi.org/10.1007/BF00987006
Wilkowski, B.M. & Meier, B.P. (2010). Bring it on: angry facial expressions potentiate
approach-motivated motor behavior. Journal of Personality & Social Psychology, 98(2),
201–210.
Willis, M. L., Windsor, N. A., Lawson, D. L., & Ridley, N. J. (2015). Situational context and
perceived threat modulate approachability judgements to emotional faces. PloS One,
10(6), e0131472. https://doi.org/10.1371/journal.pone.0131472
Woolrich, M. (2008). Robust group analysis using outlier inference. NeuroImage, 41(2), 286-
301.
Woolrich, M., Behrens, T., Beckmann, C., Jenkinson, M., and Smith, S. (2004). Multilevel linear
modelling for FMRI group analysis using Bayesian inference. NeuroImage, 21(4), 1732-
1747.
Woolrich, M., Ripley, B., Brady, J., and Smith, S. (2001). Temporal autocorrelation in univariate
linear modelling of FMRI data. NeuroImage, 14(6), 1370-1386.
Worsley, K. (2001). Chapter 14, Statistical analysis of activation images. In Functional MRI: An
Introduction to Methods, eds. P. Jezzard, P.M. Matthews and S.M. Smith. Oxford
University Press.
Zaki, J., & Ochsner, K. N. (2012). The neuroscience of empathy: progress, pitfalls and promise.
Nature Neuroscience, 15(5), 675-680. https://doi.org/10.1038/nn.3085

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Neurobiological correlates of fathers’ transition to parenthood

PDF

Inhibitory spillover in adolescence: a fMRI study

PDF

Predictors and outcomes across the transition to fatherhood

PDF

Theory of mind processing in expectant fathers: associations with prenatal oxytocin

PDF

Fathers’ social-cognitive network connectivity: associations with prenatal father-infant attachment and postpartum parenting behavior

PDF

Biological and behavioral correlates of emotional flexibility and associations with exposure to family aggression

PDF

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

PDF

Pregnancy in the time of COVID-19: effects on perinatal mental health, birth, and infant development

PDF

Prenatal predictors of parental sensitivity in first-time mothers and fathers

PDF

Inhibitory control in first-time fathers: Neural correlates and associations with paternal mental health

PDF

Spatial and temporal patterns of brain activity associated with emotions in music

PDF

The old ball and linkage: couples’ prenatal conflict behavior, cortisol linkage, and postpartum depression risk

PDF

Couple conflict during pregnancy: Do early family adversity and oxytocin play a role?

PDF

Relationships and meaning: examining the roles of personal meaning and meaning-making processes in couples dealing with important life events

PDF

You always say that: physical aggression perpetration, linguistic behavior, and conflict intensity in young adult dating couples

PDF

Decoding the neural basis of valence and arousal across affective states

PDF

Family aggression exposure and community violence exposure associated with brain volume in late adolescence: a comparison of automated versus manual segmentation

PDF

The emotional arousal effects on visual processing

PDF

Young adult dating couple interactions in daily life: links to family aggression and physiological processes

PDF

Core, social and moral disgust processing in youth with autism spectrum disorder (ASD)

Asset Metadata

Creator Stoycos, Sarah Ann (author)

Core Title Neural and behavioral correlates of fear processing in first-time fathers

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Psychology

Degree Conferral Date 2021-08

Publication Date 07/28/2021

Defense Date 05/13/2021

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

Tag fear processing,OAI-PMH Harvest,Parenting,social affective neuroscience

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Saxbe, Darby (committee chair), Aziz-Zadeh, Lisa (committee member), Kaplan, Jonas (committee member), Margolin, Gayla (committee member), Monterosso, John (committee member)

Creator Email stoycos.sarah@gmail.com,stoycos@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC15659362

Unique identifier UC15659362

Legacy Identifier etd-StoycosSar-9912

Document Type Dissertation

Format application/pdf (imt)

Rights Stoycos, Sarah Ann

Type texts

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

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

Repository Name University of Southern California Digital Library

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

Repository Email cisadmin@lib.usc.edu