Self-Affirmation Activates Brain Systems Associated with Self-Related Processing and Reward and is Reinforced by Future Orientation

Abstract

Self-affirmation theory posits that people are motivated to maintain a positive self-view and that threats to perceived self-competence
are met with resistance. When threatened, self-affirmations can restore self-competence by allowing individuals to reflect
on sources of self-worth, such as core values. Many questions exist, however, about the underlying mechanisms associated with
self-affirmation. We examined the neural mechanisms of self-affirmation with a task developed for use in a functional magnetic
resonance imaging (fMRI) environment. Results of a region of interest analysis demonstrated that participants who were affirmed
(compared to unaffirmed participants) showed increased activity in key regions of the brain’s self-processing (MPFC+PCC) and
valuation (VS+VMPFC) systems when reflecting on future oriented core values (compared to everyday activities). Furthermore,
this neural activity went on to predict changes in sedentary behavior consistent with successful affirmation in response to
a separate physical activity intervention. These results highlight neural processes associated with successful self-affirmation,
and further suggest that key pathways may be amplified in conjunction with prospection.

Full text

RUNNING HEAD: NEURAL CORRELATES OF SELF-AFFIRMATION
1
Self-affirmation activates brain systems associated with self-related processing and reward and is
reinforced by future orientation
Christopher N. Cascio
1
, Matthew Brook O’Donnell
1
, Francis J. Tinney, Jr.
2
, Matthew D.
Lieberman
3
, Shelley E. Taylor
3
, Victor J. Strecher
2
, & Emily B. Falk
1
University of Pennsylvania
1
, University of Michigan
2
, University of California, Los Angeles
3
Correspondence:
Christopher N. Cascio
Annenberg School for Communication
University of Pennsylvania
3620 Walnut St.
Philadelphia, PA 19104
ccascio@asc.upenn.edu
Emily B. Falk
Annenberg School for Communication
University of Pennsylvania
3620 Walnut St.
Philadelphia, PA 19104
falk@asc.upenn.edu
Total words in main text <5332>
Acknowledgements: We thank Holly Derry, Ian Moore, and Michele Demers for assistance in
developing intervention materials, and the staff of the University of Michigan fMRI Center for
support and assistance with fMRI data acquisition. We thank Angela Fagerlin, Thad Polk,
Lawrence An, Kennith Resnicow and the Michigan CECCR for support in realizing this project
and Sonya Dal Cin and Sara Konrath for helpful discussions. This research was supported by The
Michigan Center of Excellence in Cancer Communication Research/NIH-P50 CA101451 (PI
Strecher), a NIH New Innovator Award/1DP2DA03515601 (PI Falk) and NIH/NCI
1R01CA180015-01 (PI Falk). The authors thank Kristin Shumaker, Nicolette Gregor, Alison
Sagon for assistance with data collection, and Jonathan Mitchell for advice regarding processing of
the accelerometer data.
© The Author (2015). Published by Oxford University Press. For Permissions, please email:
journals.permissions@oup.com
Social Cognitive and Affective Neuroscience Advance Access published November 5, 2015
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

2
Abstract
Self-affirmation theory posits that people are motivated to maintain a positive self-view and that
threats to perceived self-competence are met with resistance. When threatened, self-affirmations
can restore self-competence by allowing individuals to reflect on sources of self-worth, such as
core values. Many questions exist, however, about the underlying mechanisms associated with
self-affirmation. We examined the neural mechanisms of self-affirmation with a task developed
for use in a functional magnetic resonance imaging (fMRI) environment. Results of a region of
interest analysis demonstrated that participants who were affirmed (compared to unaffirmed
participants) showed increased activity in key regions of the brain’s self-processing
(MPFC+PCC) and valuation (VS+VMPFC) systems when reflecting on future oriented core
values (compared to everyday activities). Furthermore, this neural activity went on to predict
changes in sedentary behavior consistent with successful affirmation in response to a separate
physical activity intervention. These results highlight neural processes associated with successful
self-affirmation, and further suggest that key pathways may be amplified in conjunction with
prospection.
Keywords: self-affirmation, fMRI, reward, positive valuation, emotion regulation
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

3
Introduction
It is well documented that people seek to maintain a positive self-view and that threats to
perceived self-competence across many domains are met with resistance (Sherman & Cohen,
2006). A large body of literature, however, demonstrates that a class of interventions called self-
affirmations have benefits across threatening situations; affirmations can decrease stress,
increase well being, improve academic performance and make people more open to behavior
change (for a review, see (Cohen & Sherman, 2014)). Self-affirmations are acts that affirm one’s
self-worth, often by having individuals reflect on core values, which may give individuals a
broader view of the self. This in turn can allow individuals to move beyond specific threats to
self-integrity or self-competence (Cohen & Sherman, 2014; Steele, 1988).
Effects associated with self-affirmation interventions often occur without explicit
awareness (Sherman et al., 2009). This lack of awareness makes it difficult for individuals to
introspect on their experience and makes it difficult for researchers to examine specific
underlying mechanisms that lead from the affirmation experience to behavioral change.
Neuroimaging methods offer one way to examine a set of processes underlying self-affirmation
interventions at the point of actual affirmation exposure, without the need for individuals to
reflect on their experience (Falk et al., 2015); however, the neural mechanisms that underpin acts
of self-affirmation have not been studied (Cohen & Sherman, 2014). Understanding the
underlying neural mechanisms associated with self-affirmation will help to further expand our
theoretical understanding of the processes at play during self-affirmation and may contribute to
the development of more effective interventions. Thus, our first research question centers on the
neurocognitive processes associated with the act of self-affirmation. Furthermore, the core brain
systems involved in self-related processing and reward, that we hypothesize to be involved in
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

4
affirmation, overlap with past studies of temporal orientation (i.e., considering events in the past
and future; (D’Argembeau et al., 2010; D’Argembeau, Xue, Lu, Van der Linden, & Bechara,
2008)). Thus, our second research question focuses on whether the neural pathways to self-
affirmation might be amplified in conjunction with specific temporal orientations.
Potential pathways to self-affirmation
One account of why self-affirmations are successful is attributed to their ability to
broaden a person’s overall perspective and reduce the effect of negative emotions (Sherman,
2013; Cohen & Sherman, 2014). For example, researchers have suggested that self-affirmations
remind individuals of psychosocial resources that extend beyond a specific threat, which allows
them to focus on sources of positive self-worth that transcends the threat. This in turn is thought
to reduce reactivity to the threat and protect overall psychological wellbeing (Cohen, Garcia,
Purdie-Vaughns, Apfel, & Brzustoski, 2009; Cook, Purdie-Vaughns, Garcia, & Cohen, 2012;
Koole, Smeets, Van Knippenberg, & Dijksterhuis, 1999; Sherman et al., 2013).
Such effects might arise through several different pathways. First, affirmations may
increase focus on sources of positive value to individuals. Self-affirmation interventions often
rely on having participants reflect on personal core values and rewarding experiences. This
pathway would engage neural mechanisms associated with reward and positive valuation. A
recent meta-analysis demonstrates that brain regions most prominently involved in reward and
positive valuation include the ventral striatum (VS) and ventral medial prefrontal cortex
(VMPFC; (Bartra, McGuire, & Kable, 2013)).
Related to the pathway described above, affirmations could also work by focusing people
on sources of positive self-worth, such as personal successes. This may also involve specific
reflection on personal attributes outside of the threat (Sherman & Hartson, 2011; Sherman,
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

5
2013). Meta-analyses across a variety of tasks find that self-related processing is most often
associated with increased activity in the medial prefrontal cortex (MPFC) and posterior cingulate
cortex (PCC; (Denny, Kober, Wager, & Ochsner, 2012; Northoff et al., 2006)). Thus, if self-
affirmations succeed due to a boost in self-related processing prior to threat exposure, activity in
the MPFC and PCC should increase during affirmation.
Furthermore, self-affirmations may allow for more efficient use of psychological
resources needed to deal with the incoming threat (Sherman, 2013). This has been demonstrated
in studies that examine the success of self-affirmation interventions in counteracting
manipulations that reduce available cognitive and psychological resources (e.g., cognitive load
and ego-depletion manipulations; (Schmeichel & Vohs, 2009; Vohs & Faber, 2007; Burson et
al., 2012; Logel & Cohen, 2012)). Although these studies find evidence that affirmation
interventions can reduce threat, it is unclear which psychological resources are actually involved
in this process. One possible source of regulatory resources include the ventrolateral prefrontal
cortex (VLPFC) and anterior cingulate cortex (ACC), which have been implicated in regulation
of emotion and facilitating difficult choices (Marsh, 2007; Oschner et al., 2004; Wager et al.,
2008). Self-affirmations may work by priming these regions to regulate emotions.
Affirmation and temporal orientation
Activity within several of our key self-related processing and reward regions of interest
(ROIs) changes with manipulations of temporal focus. Although self-affirmation interventions
have been successfully carried out using manipulations that focus both on past experiences as
well as future goals (for a review, see; (McQueen & Klein, 2006)), temporal orientation has not
been a core focus of affirmation research. Given the overlap between brain systems hypothesized
to support affirmation effects and to support temporal orientation effects, however, we examined
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

6
whether neural responses in brain systems associated with successful self-affirmation might
change with or be reinforced by temporal focus.
For example, it has been found that imagining future personally relevant, emotionally
positive, and rewarding events is associated with changes in VMPFC, striatum, MPFC and PCC
(Benoit, Gilbert, & Burgess, 2011; Benoit, Szpunar, & Schacter, 2014; D’Argembeau et al.,
2010, 2008). Increased activity in the MPFC has also been shown to positively correlate with
imagining positive (vs. negative) future episodes (D’Argembeau et al., 2008) and such activation
is further associated with projected reward value of the imagined future (Benoit et al., 2011,
2014). In addition, a recent meta-analysis found that increased activity in the MPFC and PCC,
among other regions, was associated with thinking about hypothetical (e.g., future) compared to
past episodes (Benoit & Schacter, 2015). Furthermore, another recent meta-analysis examining
neural correlates associated with personal goals, future thinking, and mind wandering found that
the MPFC is consistently activated in all three domains (Stawarczyk & D’Argembeau, 2015).
These studies support the idea that mentally simulating future events, especially those relevant to
personal goals, involves key regions hypothesized to be involved in self-affirmation
interventions, including the VMPFC, MPFC, and PCC. Thus, if both future oriented thought and
self-affirmation rely on similar neural mechanisms, they may mutually reinforce one another.
Importantly, these differences are not limited to neural activity. For example participants
have better memory recall when encoding new information coupled with imagined scenarios that
plan for the future, in comparison to remembering past events or events that are considered
without a time relationship (Klein, Robertson, & Delton, 2010). In addition, mental simulations
focusing on future events have been shown to benefit goal planning and one’s psychological
wellbeing (for a review, see; Schacter, 2012). Taken together, all of these studies reinforce the
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

7
hypothesis that engagement of our key ROIs may differ by temporal focus, and that future
orientation may reinforce the effects of reflecting on personally relevant core values. Thus, we
examined temporal focus as a potential moderator of neural responses in our key ROIs during
affirmation.
The current study
In sum, the current study aims to elucidate the underlying mechanisms associated with
self-affirmation by examining participants’ neural activity during a self-affirmation task
specifically designed for functional magnetic resonance imaging (fMRI). We tested the extent to
which exposure to self-affirmation produced increases in brain systems associated with positive
valuation (VS+VMPFC), self-related processing (MPFC+ PCC), and emotion regulation
(rACC+rVLPFC). In addition, we examined whether the neural effects of affirmation are
moderated by temporal orientation (past vs. future). We validated our fMRI-compatible self-
affirmation intervention in relation to its ability to increase receptivity to a subsequent set of
health messages designed to reduce sedentary behavior in sedentary adults (Falk et al., 2015).
Methods
Participants
Participants (N=67; self-affirmed=33; unaffirmed=28) were adults between the ages of
18-64 (41 females; mean age=33.42 years, SD=13.04; 44 White, 12 Black, 3 Asian, 1 Hispanic,
7 Other), recruited as part of a study examining neural correlates of exposure to health messages
that encouraged physical activity behavior in sedentary adults. All participants were sedentary
(participants self-reported an estimated < 195 minutes of combined walking, moderate or
vigorous activity per week at the time of recruitment). This was defined by mean activity on
short-form International Physical Activity Questionnaire (IPAQ) at the time of recruitment. On
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

8
average, recruited participants reported 23.5 minutes/week of activity, SD=49.5; mean BMI =
27.99, SD=6.84, indicating that on average participants were in the overweight category.
Participants were right-handed, did not suffer from claustrophobia, were not currently taking any
psychoactive medications, had normal (or corrected to normal) vision, and did not have metal in
their body that was contraindicated for MRI (see supplemental materials for additional sample
details).
Study design
Participants completed a three-part study (see Figure 1). At baseline, participants ranked
a list of 8 personal values, completed self-report questionnaires and were fitted with an
accelerometer to measure physical activity behavior. One week later participants completed an
fMRI appointment in which they underwent the fMRI-compatible self-affirmation (or control)
intervention. All participants then saw potentially-threatening messages encouraging physical
activity and the success of the affirmation manipulation was validated based on objectively
measured physical activity/sedentary behavior change attributable to self-affirmation in the
subsequent month. Additional details on the sample and task session can be found in Falk et al.,
2015, however, the neural processes associated with the actual affirmation task have not been
previously examined.
Self-affirmation task
During the initial baseline appointment participants were asked to rank a list of 8 values
from least to most valued, “Please order the following values according to how important they
are to you”. The list of 8 values included, creativity, relations with family and friends, sense of
humor, independence, business or earning money, politics, religious values, and spontaneity or
living life in the moment. These values were then used in the MRI portion of the self-affirmation
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

9
task, such that participants in the affirmed condition reflected on their top ranked value and
participants in the control condition reflected on their lowest ranked value.
Although there are many approaches to self-affirmation, two of the most prominent
approaches ask individuals to write about a highly ranked personal value or to respond to
questionnaires containing questions relevant to a highly ranked personal value (Cohen &
Sherman, 2014; McQueen & Klein, 2006; Sherman, 2013). Typically self-affirmation writing
tasks instruct affirmed participants to write for a period of time on one of their core values;
control groups typically write on a topic that is not valued (McQueen & Klein, 2006; Napper,
Harris, & Epton, 2009). Similarly, value scales involve the completion of questionnaires that
allow participants to express their identification with the core value and why their core value is
important to them; control participants complete questionnaires about topics of lower personal
value and importance (Sherman, 2013). As in other widely used affirmation manipulations (see
Cohen & Sherman, 2015; McQueen & Klein, 2006 for reviews), there were some differences in
the values most consistently ranked as top and bottom values in the current study, however, there
is also substantial overlap in values used in the affirmation and control conditions (for
distribution, see; table 1).
To save time and standardize instructions, participants received task instructions for the
affirmation task during the structural scan, directly prior to the task. To start the task
preparation, participants were initially instructed during the structural scan to “Please think
about an experience you had involving [VALUE]”, where [VALUE] was replaced with their
assigned value. This was followed by instructions to “Try and visualize yourself in the
experience and remember as many specific details as possible”. Participants were then prompted
with phrases to help keep them prepare for the main affirmation task. Example statements
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

10
included “Think about when the experience occurred” and “Think about how you currently feel
about this experience”. Once participants had come up with scenarios relevant to their top (or
bottom) ranked value during the structural scan, they completed the main self-affirmation task
during functional scanning.
To test the interaction between affirmation and temporal orientation, the main fMRI
affirmation task instructed participants on different trials to think about a time when value-
relevant scenarios had occurred (past) and when parallel scenarios could occur (future).
Participants in both affirmation and control conditions were presented with prompts for scenarios
focused on value statements as well as everyday activities (as a within subjects control
condition). All participants were presented with the same control (everyday) activity scenarios.
Example statements relating to experience of a specific value in the past or future condition are
as follows [value = friends and family]: "Think about a time in the past when you had fun with
family and friends", "Think about a time in the future when you might be having fun with family
and friends". Example everyday statements included: "Think about a time in the past when you
charged your cell phone", "Think about a time in the future when you might charge your cell
phone". Importantly it should be noted that all statements (value and control) were focused on
oneself and not subject to factual knowledge. This was done in order to have the distinction
between high and low values pertain more to the importance placed on the topic rather than on
topic knowledge. For example, for those who were assigned to think about what we referred to as
“politics”, the statements were not about politicians but rather how political values might be
manifest in one’s life (e.g., Think about a time in the future when you might read about current
events; Think about a time in the future when you might be inspired by people taking political
action). The self-affirmation task used a 2X2 block design, (past vs. future)X(everyday vs.
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

11
value). Each block consisted of exposure to the scenario prompt for 12 seconds in which
participants thought about the given statement and responded by pressing a button with their
index figure each time they thought of a personally relevant example associated with the given
statement. Participants were instructed to think about as many examples as they could for each
scenario. Ten different scenarios were given for each condition (past value, past everyday, future
value, and future everyday) for a total of 40 blocks in the task. Value-specific scenarios were
created based on reflections one may have when engaging in a self-affirmation writing exercise,
whereas everyday scenarios were created to represent common events that occur on a daily basis.
Participants saw a fixation cross for 2 and 12 (every fifth trial) seconds between each block.
Validation of the fMRI self-affirmation intervention
Following their randomly assigned affirmation or control intervention, all participants were
exposed to the same health messages encouraging increased physical activity and decreased
sedentary behavior. The success of the affirmation intervention was validated using behavior
change effects attributable to the experimental manipulation of self-affirmation. More specifically,
aggregate measures of sedentary behavior were created measuring pre and post intervention
activity captured for one week prior and one month following the intervention using triaxial
accelerometers, and compared by condition. For further details on the health messaging task and
accelerometer data collection and analysis, please see (Falk et al., 2015).
fMRI data acquisition and data analysis
Imaging data were acquired using a 3 Tesla GE Signa MRI scanner. One functional run
was acquired for each participant (323 volumes total
1
). Functional images were recorded using a
1
Note: For the first six participants (1 control, 5 affirmed due to the randomizer), a slightly longer (2 run) version of
the task was used, in which the blocks were 16 seconds long instead of 12, and the affirmation task was split into
two runs of 209 volumes each. These participants were the first to do the study and we initially had longer scan time
that included an extra 4 seconds for each block.
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

12
reverse spiral sequence (TR=2000ms, TE=30ms, flip angle=90°, 43 axial slices, FOV=220mm,
slice thickness=3mm; voxel size=3.44x3.44x3.0mm). We also acquired in-plane T1-weighted
images (43 slices; slice thickness=3mm; voxel size=.86x.86x3.0mm) and high-resolution T1-
weighted images (SPGR; 124 slices; slice thickness=1.02x1.02x1.2mm) for use in coregistration
and normalization.
Functional data were pre-processed and analyzed using Statistical Parametric Mapping
(SPM8, Wellcome Department of Cognitive Neurology, Institute of Neurology, London, UK;
please see supplemental materials for details of preprocessing stream). Data were modeled using
the general linear model as implemented in SPM8. Four trial types were modeled: past value
scenarios, future value scenarios, past everyday scenarios, future everyday scenarios; fixation
trials were not modeled and constituted an implicit baseline. The six rigid-body translation and
rotation parameters derived from spatial realignment were also included as nuisance regressors.
Data were high-pass filtered with a cutoff of 128s.
Region of interest analysis
To test the balance of activity within brain networks involved in positive valuation and
reward (VS+VMPFC), self-related processing (MPFC+PCC) and regulating emotions
(rACC+rVLPFC), we first conducted a priori defined ROI analyses on each network of interest
independently. Percent signal change scores were extracted from each combined network ROI
contrasting the value > everyday scenarios; past value > past everyday scenarios; and future
value > future everyday scenarios for each participant (see supplementary materials for ROI
definitions and analysis details).
To investigate neural processes associated with self-affirmation that extended beyond our
main ROI analyses, we subsequently conducted whole brain analyses examining differences
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

13
between the affirmed and control participants for each of our main target contrasts: value >
everyday scenarios; past value > past everyday scenarios; and future value > future everyday
scenarios. All analyses are reported with a threshold of p=.005, K=35, corrected for multiple
comparisons based on a Monte Carlo simulation using AlphaSim (Ward, 2000). Furthermore,
based on a priori hypotheses linking valuation activity (VS+VMPFC) to self-affirmation
processes, the relatively small size of VS, and positive results from a priori planned ROI
analyses, additional analyses were run using a threshold of (p=.005, K=19), corrected for
multiple comparisons based on a Monte Carlo simulation for the VS+VMPFC mask (949 total
voxels) in order to maintain an appropriate balance of type I and type II error risk, given the
exploratory nature of the whole-brain analysis (Lieberman & Cunningham, 2009).
Results
Effects of affirmation: Region of interest analysis
Main effects of affirmation. First, we examined whether activity in our a priori
hypothesized ROIs associated with valuation (VS+VMPFC), self-related processing
(MPFC+PCC), and emotion regulation (rACC+rVLPFC) were differentially activated for those
in the affirmed versus control group as they reflected on value > everyday scenarios. Overall,
affirmed participants displayed significantly greater activity in the valuation/reward network
(M=.102) versus control participants (M=.012) when exposed to value versus everyday scenarios
(t(57)=2.43, p=.018). Activity in the self-processing network while viewing value versus
everyday scenarios was not significantly different for those in the affirmed versus control group,
when averaging across temporal orientations (t(57)=.88, p=.382). Activity in the emotion
regulation network also did not differ between affirmed and control groups when averaging
across temporal orientations (t(57)=.62, p=.540). All ROI results are summarized in table 2.
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

14
Affirmation and temporal orientation. Second, we tested whether affirmation effects
differed by temporal orientation within our key ROIs (see Table 2). On average, affirmed
participants displayed significantly greater activity in the valuation network (M=.133) than
control participants (M=-.029) when viewing future oriented value scenarios versus future
oriented everyday scenarios (t(57)=3.26, p=.002); the difference between responses to future and
past oriented value scenarios was also significantly different between affirmed and control
participants (t(57)=3.09, p=.003). Additionally, affirmed participants displayed significantly
greater activity in the self-processing network (M=.100) than control participants (M=.032) when
viewing future oriented value scenarios versus future oriented everyday scenarios (t(57)=2.50,
p=.015); the difference between responses to future and past oriented value scenarios was also
significantly different between affirmed and control participants (t(57)=3.48, p=.001).
Participants in the affirmation and control conditions did not differ in their activity in the
emotion regulation network when reflecting on future oriented value and everyday scenarios
(t(57)=1.30, p=.200), however the difference between responses to future and past oriented value
scenarios was significantly different between affirmed and control participants (t(57)=2.39,
p=.02).
Next, we examined whether affirmation effects differed by past orientation within our
key ROIs. No significant differences were observed between those in the affirmation versus
control condition for activity in regions associated with valuation (t(57)=.34, p=.738), self-
related processing (t(57)=-.97, p=.337) or emotion regulation (t(57)=1.77, p=.540) when
reflecting on past oriented value versus everyday scenarios.
Finally, within the affirmation group paired samples t-tests were run to examine whether
neural activity within our hypothesized ROIs were differently activated depending on temporal
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

15
orientation (past versus future). Neural activity within the valuation network (VS+VMPFC) was
significantly greater when viewing future oriented value scenarios (M=.108) compared to
viewing past oriented value scenarios (M=.003), t(29)=3.83, p<.001. Similarly, neural activity
with our self-processing network (MPFC+PCC) was also significantly greater when viewing
future oriented value scenarios (M=.114) compared to viewing past oriented value scenarios
(M=.044), t(29)=2.79, p=.009. Finally, neural activity within our emotion regulation network
(rACC+rVLPFC) was not significantly different when viewing future oriented value scenarios
(M=.056) compared to viewing past oriented value scenarios (M=.025), t(29)=1.65, p=.111.
Whole brain analysis
Following our hypothesis-driven ROI analyses, we ran a series of exploratory whole
brain analyses that examined differences in neural activity between the affirmed and control
groups for key contrasts of interest to explore regions outside of those covered by our ROI
analyses. Results of the whole brain contrast of value > everyday scenarios did not yield
significant results; future value > future everyday scenarios are reported in table 3, figure 2; past
value > past everyday scenarios did not yield significant results; and future > past value
scenarios are reported in table 4. Significant results from the whole brain analysis reinforce
effects observed in the ROI analyses. We observed increased activity within VMPFC and VS
when affirmed (relative to control) participants reflected on future-oriented (but not past-
oriented) value scenarios highlighting the role of activity within the valuation system,
particularly during prospection.
Neural activity during affirmation and subsequent behavior change
Validating the downstream effect of our affirmation manipulation on behavior change
following exposure to health messages, participants in the self-affirmation condition showed
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

16
steeper declines in their levels of sedentary behavior over time compared to control participants,
p=.008 (Falk et al., 2015). Given that the effects of affirmation within the brain during the
affirmation task were strongest in our hypothesized valuation and self-processing regions, we
next examined whether this activity was related to target behavior change. Increased activity in
the VS+VMPFC (figure 3) and MPFC+PCC ROIs during value > everyday scenarios were
associated with decreased average post intervention sedentary behavior, controlling for age,
gender, education, body mass index (BMI), and pre intervention sedentary behavior (β=-.26,
t(33)=-2.27, p=.030; β=-.27, t(33)=-2.15, p=.039; respectively). Next, a follow up analysis was
run examining temporal orientation differences. Results indicate that increased neural activity in
the valuation network during future value versus future everyday scenarios was marginally
associated (β=-.22, t(33)=-1.97, p=.057) and the self-processing network was significantly
associated (β=-.25, t(33)=-2.39, p=.023) with decreased sedentary behavior following the
affirmation intervention, controlling for age, gender, BMI, and pre intervention sedentary
behavior. No significant results were found for past oriented scenarios, p>.05.
Finally, we tested the indirect relationship between group assignment (affirmation versus
control) and changes in one’s sedentary behavior (post – pre intervention) through neural activity
in valuation and self-processing systems. Significant indirect effects were found for both the
valuation and self-processing ROIs (average causal mediation effect (ACME); B=-.07, CI=[-.13,
-.02], p=.01; B=-.03, CI=[-.07, -.00], p=.04; respectively), such that those in the affirmed
condition displayed greater activity in valuation and self-processing networks relative to those in
the control condition; in turn participants who displayed greater activation in the valuation and
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

17
self-processing networks also displayed significantly greater decreases in sedentary behavior
following the affirmation intervention, controlling for age, gender, years educated, and BMI.
2
Discussion
Results from the current study provide initial evidence of neural processes associated
with the act of self-affirmation. First, our hypotheses regarding the relationship between
affirmation and neural reward pathways were supported. ROI analyses revealed that affirmed
relative to control participants showed significantly greater activity in the hypothesized positive
valuation regions (VS+VMPFC), and that this effect was driven by affirmations focusing on
future rather than past experiences. In addition, increased activity in reward/valuation regions
during self-affirmation was associated with decreases in sedentary behavior following the
affirmation intervention. Furthermore, we observed a significant indirect effect, such that those
in the affirmed condition displayed greater activity in the valuation network, which was
associated with greater change in sedentary behavior following the affirmation intervention.
Thus, our results are consistent with the hypothesis that systems associated with positive
valuation play an important role in successful affirmation and are consistent with the broadened
value account of why self-affirmation interventions succeed (Cohen et al., 2009; Cook et al.,
2012; Koole et al., 1999; Sherman et al., 2013). The VS and VMPFC are brain regions that are
most commonly associated with the expectation and receipt of positively valued or rewarding
outcomes (Bartra et al., 2013). Importantly, this system encodes not only primary rewards (such
as food) but also more abstract rewards (Bartra et al., 2013), of the type that are called to mind
by personally meaningful values in self-affirmation.
2
Note: In addition to examining changes in sedentary behavior using difference scores, a test of indirect effects was
also run that examined post intervention sedentary behavior controlling for pre intervention sedentary behavior
(ACME; B=-.08, CI=[-.15, -.03], p<.01). Results were consistent for future oriented statements (ACME; B=-.04,
CI=[-.09, -.01], p=.01); past oriented statements were not significant (ACME; B=-.02, CI=[-.06, .01], p=.23).
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

18
In addition, our findings suggest that positive affirmations may have especially strong
effects within the reward system in conjunction with future orientation. This finding converges
with prior studies demonstrating that increased activity in the VMPFC is associated with
imagining positive rather than negative future events (D’Argembeau et al., 2008) and increases
when anticipating future rewards (Benoit et al., 2011, 2014). This account is also consistent with
a role of the reward system in guiding reinforcement learning and future behavioral decisions
through computation of the “incentive value of a contemplated behavioral act“ (Knutson &
Cooper, 2005; McClure, Daw, & Read Montague, 2003).
Furthermore, although not directly addressed by our data, past research suggests that self-
transcending values and goals may be particularly powerful. For example, affirmation of self-
transcending values is more powerful in reducing behaviors associated with ego depletion than
affirmation of self-enhancing values (Burson, Crocker, & Mischkowski, 2012). Our neural data
provide a possible link between such behavioral results and research examining neural reward
activity in response to prosocial (eudaimonic) versus selfish (hedonic) decisions, which finds that
VS activity differentially predicts later mental health outcomes. More specifically, increased
activity in the VS in response to potential prosocial rewards, relative to self-focused rewards is
associated with later positive outcomes (Telzer, Fuligni, Lieberman, & Galván, 2014). These
findings along with the findings from the current study support potential synergy between
prospection and value affirmation in eliciting the types of reward response that can prime
positive behavior.
Second, we found support for the hypothesis that future-oriented affirmations activated
brain regions implicated in self-related processing. In particular, the MPFC is often implicated in
reflecting on one’s own preferences, motivations and in the process of self-insight (for a review,
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

19
see; Lieberman, 2010). During future oriented affirmation, affirmed, relative to control,
participants displayed significantly greater activity in our MPFC and PCC ROIs. Importantly,
MPFC and PCC are consistently implicated in both self-related processing (Denny et al., 2012;
Northoff et al., 2006) and imagining personally relevant future events (D'Argembeau et al.,
2010), as well as remembering past events (for a review, see; Schacter, 2012). Furthermore,
increased activity in the MPFC is associated with imagining positive rather than negative future
events (D’Argembeau et al., 2008) and increases while anticipating future rewards (Benoit et al.,
2011, 2014). Thus, the current data and recent meta-analytic evidence suggests that in addition to
a strong role in self-related processing (Denny et al., 2012; Northoff et al., 2006), the MPFC is
more active when thinking about future compared to past episodes (Benoit & Schacter, 2015)
and when thinking about personal goals, future thinking, and mind wondering (Stawarczyk &
D’Argembeau, 2015). Successful self-affirmation interventions bring together several of these
components and our neural data suggest a new way in which these paths may mutually reinforce
one another. In other words, we find novel evidence that a future frame may act synergistically
with value-based self-affirmations to bolster a sense of self prior to threat exposure. This may
occur by calling to mind desired future states or motivations, also consistent with the broadened
value account of why self-affirmation interventions succeed.
Finally, the current study reports on the successful development of an fMRI-compatible
self-affirmation task, which can be used to examine neural mechanisms associated with self-
affirmation in other behavioral or theoretical contexts, and combined with other subsequent tasks
of interest to affirmation researchers. One strength of this task is that all aspects of the task
(including the within subjects control condition and instructions) are identical for affirmed and
control participants, differing only in the importance of the focal value to participants. This rules
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

20
out many confounds related to differing tasks. An additional strength is that our objective
behavioral results indicate that the self-affirmation manipulation was successful in decreasing
sedentary behavior in at-risk (sedentary) adults, which was mediated by activity in the valuation
network during affirmation and by activity in both valuation and self-related processing systems
during future oriented affirmations.
This adds to our understanding of affirmation from both basic science and applied
perspectives. The current results 1) highlight novel pathways to affirmation through neural
reward and self-processing pathways; and 2) suggest that these mechanisms may be reinforced or
augmented by prospection. It is possible that future oriented affirmations may be more
successful than past oriented affirmation, though between subjects follow up studies are needed
to test this hypothesis. Finally, the creation of a scanner compatible affirmation task opens
future research possibilities to explore the neural effects of affirmation in other contexts.
In addition to the primary strengths of the study addressed above, it should be noted that
each of our primary ROIs serves functions that go beyond those hypothesized in the current
investigation and thus should be taken as one of several possibilities (Poldrack, 2006). However,
in the current study the use of a priori hypothesized and theoretically driven ROIs helps reduce
problems with reverse inference. Furthermore, there are confines associated with the scanning
environment, such that we cannot know the specific scenarios envisioned by each participant in
response to our prompts at the time of affirmation exposure, we can only examine neural
processing that takes place during that time. Therefore, it is likely that variability in how
important the “lowest” ranked value was to participants existed, which may have allowed for
affirming benefit to some of those in the control condition, resulting in a conservative test of our
hypotheses. In addition, self-affirmation interventions are often confounded with value and
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

21
content making it difficult to distinguish which aspects of the intervention are driving results.
Future neuroimaging studies should attempt to untangle these differences in order to better
understand the underlying mechanisms associated with self-affirmation interventions.
Conclusion
The current results demonstrate that activity in hypothesized reward/valuation regions
(VS+VMPFC (Bartra et al., 2013) are primary pathways associated with self-affirmation.
Furthermore, regions associated with self-related processing (MPFC+PCC (Northoff et al., 2006;
Denny et al., 2012) and prospection (D’Argembeau et al., 2010, 2008) are associated with self-
affirmations that are future oriented. These neural correlates of self-affirmation were further
associated with objectively measured behavior change, suggesting the external validity of the
affirmation task. Taken together, our results highlight ways in which brain systems implicated in
positive valuation and self-related processing may be reinforced by prospection and suggest
novel insight into the balance of processes supporting affirmation. These results also introduce a
task for understanding the underlying mechanisms associated with self-affirmation and hence
provide a tool for future studies to examine effects of self-affirmation interventions across a wide
range of potential applications and outcomes.
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

22
References
Bartra, O., McGuire, J. T., & Kable, J. W. (2013). The valuation system: A coordinate-based
meta-analysis of {BOLD} fMRI experiments examining neural correlates of subjective
value. NeuroImage, 76(0), 412 – 427.
Benoit, R. G., Gilbert, S. J., & Burgess, P. W. (2011). A neural mechanism mediating the impact
of episodic prospection on farsighted decisions. The Journal of Neuroscience, 31(18),
6771–6779.
Benoit, R. G., & Schacter, D. L. (2015). Specifying the core network supporting episodic
simulation and episodic memory by activation likelihood estimation. Neuropsychologia,
75, 450–457.
Benoit, R. G., Szpunar, K. K., & Schacter, D. L. (2014). Ventromedial prefrontal cortex supports
affective future simulation by integrating distributed knowledge. Proceedings of the
National Academy of Sciences, 111(46), 16550–16555.
Burson, A., Crocker, J., & Mischkowski, D. (2012). Two types of value-affirmation implications
for self-control following social exclusion. Social Psychological and Personality Science,
3(4), 510–516.
Cohen, G. L., Garcia, J., Purdie-Vaughns, V., Apfel, N., & Brzustoski, P. (2009). Recursive
processes in self-affirmation: Intervening to close the minority achievement gap. Science,
324(5925), 400–403.
Cohen, G. L., & Sherman, D. K. (2014). The psychology of change: Self-affirmation and social
psychological intervention. Annual Review of Psychology, 65, 333–371.
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

23
Cook, J. E., Purdie-Vaughns, V., Garcia, J., & Cohen, G. L. (2012). Chronic threat and
contingent belonging: protective benefits of values affirmation on identity development.
Journal of Personality and Social Psychology, 102(3), 479.
D’Argembeau, A., Stawarczyk, D., Majerus, S., Collette, F., Van der Linden, M., Feyers, D., …
Salmon, E. (2010). The neural basis of personal goal processing when envisioning future
events. Journal of Cognitive Neuroscience, 22(8), 1701–1713.
D’Argembeau, A., Xue, G., Lu, Z.-L., Van der Linden, M., & Bechara, A. (2008). Neural
correlates of envisioning emotional events in the near and far future. NeuroImage, 40(1),
398–407.
Denny, B. T., Kober, H., Wager, T. D., & Ochsner, K. N. (2012). A meta-analysis of functional
neuroimaging studies of self-and other judgments reveals a spatial gradient for
mentalizing in medial prefrontal cortex. Journal of Cognitive Neuroscience, 24(8), 1742–
1752.
Falk, E. B., Cascio, C. N., O’Donnell, M. B., Shumaker, K., Tinney, F., Kang, Y., … Strecher,
V. (2015). Self affimration alters the brain’s response to health messages and subsequent
behavior change. Proceedings of the National Academy of Sciences.
Klein, S., Robertson, T., & Delton, A. (2010). Facing the future: Memory as an evolved system
for planning future acts. Memory & Cognition, 38(1), 13–22.
http://doi.org/10.3758/MC.38.1.13
Knutson, B., & Cooper, J. C. (2005). Functional magnetic resonance imaging of reward
prediction. Current Opinion in Neurology, 18(4), 411–417.
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

24
Koole, S. L., Smeets, K., Van Knippenberg, A., & Dijksterhuis, A. (1999). The cessation of
rumination through self-affirmation. Journal of Personality and Social Psychology,
77(1), 111.
Lieberman, M. D. (2010). Social Cognitive Neuroscience. Handbook of Social Psychology.
Lieberman, M. D., & Cunningham, W. A. (2009). Type I and Type II error concerns in fMRI
research: re-balancing the scale. Social Cognitive and Affective Neuroscience, 4(4), 423–
428.
McClure, S. M., Daw, N. D., & Read Montague, P. (2003). A computational substrate for
incentive salience. Trends in Neurosciences, 26(8), 423–428.
McQueen, A., & Klein, W. M. (2006). Experimental manipulations of self-affirmation: A
systematic review. Self and Identity, 5(4), 289–354.
Napper, L., Harris, P. R., & Epton, T. (2009). Developing and testing a self-affirmation
manipulation. Self and Identity, 8(1), 45–62.
Northoff, G., Heinzel, A., de Greck, M., Bermpohl, F., Dobrowolny, H., & Panksepp, J. (2006).
Self-referential processing in our brain—a meta-analysis of imaging studies on the self.
Neuroimage, 31(1), 440–457.
Poldrack, R. A. (2006). Can cognitive processes be inferred from neuroimaging data? Trends in
Cognitive Sciences, 10(2), 59.
Schacter, D. L. (2012). Adaptive constructive processes and the future of memory. American
Psychologist, 67(8), 603.
Sherman, D. K. (2013). SelfAffirmation: Understanding the Effects. Social and Personality
Psychology Compass, 7(11), 834–845.
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

25
Sherman, D. K., & Cohen, G. L. (2006). The psychology of self-defense: Self-affirmation theory.
Advances in Experimental Social Psychology, 38, 183.
Sherman, D. K., Cohen, G. L., Nelson, L. D., Nussbaum, A. D., Bunyan, D. P., & Garcia, J.
(2009). Affirmed yet unaware: Exploring the role of awareness in the process of self-
affirmation. Journal of Personality and Social Psychology, 97(5), 745.
Sherman, D. K., Hartson, K. A., Binning, K. R., Purdie-Vaughns, V., Garcia, J., Taborsky-Barba,
S., … Cohen, G. L. (2013). Deflecting the trajectory and changing the narrative: How
self-affirmation affects academic performance and motivation under identity threat.
Journal of Personality and Social Psychology, 104(4), 591.
Stawarczyk, D., & D’Argembeau, A. (2015). Neural correlates of personal goal processing
during episodic future thinking and mindwandering: An ALE metaanalysis. Human
Brain Mapping.
Steele, C. M. (1988). The psychology of self-affirmation: Sustaining the integrity of the self.
Advances in Experimental Social Psychology, 21(2), 261–302.
Telzer, E. H., Fuligni, A. J., Lieberman, M. D., & Galván, A. (2014). Neural sensitivity to
eudaimonic and hedonic rewards differentially predict adolescent depressive symptoms
over time. Proceedings of the National Academy of Sciences, 111(18), 6600–6605.
Ward, B. D. (2000). Simultaneous inference for fMRI data. AFNI 3dDeconvolve Documentation,
Medical College of Wisconsin.
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

26
Tables
Table 1. Value rankings.
Overall Frequency Group Frequency
Value Highest Value Lowest Value Affirm Control
Money 2 4 2 3
Creativity 3 4 1 3
Independence 8 1 3 0
Politics 1 34 1 16
Friends_Family
33 0 14 0
Religion 8 20 7 5
Humor 9 1 4 1
Spontaneity 3 3 1 0
Table 2. ROI analysis summary for the contrasts value > control, value future > control future,
and past value > past control.
ROI (value > control) Affirmed Mean Control Mean t(57) p
VS & VMPFC 0.102 0.012 2.43 0.018
MPFC & PCC 0.12 0.094 0.87 0.387
rACC & rVLPFC 0.035 0.018 0.62 0.54
ROI (future value > future control)
Affirmed Mean Control Mean t(57) p
VS & VMPFC 0.133 -0.029 3.26 0.002
MPFC & PCC 0.147 0.048 2.37 0.021
rACC & rVLPFC 0.04 -0.01 1.3 0.2
ROI (past value > past control) Affirmed Mean Control Mean t(57) p
VS & VMPFC 0.071 0.052 0.34 0.738
MPFC & PCC 0.092 0.14 -0.94 0.353
rACC & rVLPFC 0.03 0.046 -0.45 0.657
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

27
Table 3. Whole brain analysis comparing the contrast (future value > future control scenarios)
for the affirmed group subtracted from the control group (p=.005, K=35). *Results based on
cluster correction for multiple comparisons using the VS+VMPFC mask (p=.005, K=19).
Region x y z k t
VMPFC (bilateral) 8 56 -11 172 4.19
Posterior Cingulate (left) -9 -60 4 95 3.98
Thalamus (right) 15 -23 16 37 3.76
Supplimentary Motor Area (left) -30 8 52 144 3.5
Supplimentary Motor Area (right) 29 15 40 99 3.77
Calcarine (bilateral) 1 -102 -8 39 3.87
Brainstem (bilateral) -2 -33 -17 46 3.57
Cerebelum (right) 32 -47 -50 44 3.64
*VS (left) -13 22 1 20 3.52
Table 4. Whole brain analysis comparing the contrast (future value > past value scenarios) for
the affirmed > control group (p=.005, K=35).
Region x y z k t
VMPFC (bilateral) -2 43 -11 444 3.88
Precuneus (bilateral) -2 -60 67 46 3.38
Precunues/PCC (bilateral) 1 -54 4 1355
5.57
VS (bilateral) 10 0 12
DLPFC (right) 29 15 43 127 4.60
DLMPFC (left) -23 26 40 248 4.60
Occipital (right) 39 -81 34 148 4.87
Occipital (left) -37 -81 37 349 4.83
Cerebelum (right) 11 -50 -47 85 4.28
Cerebelum (right) 46 -71 -38 74 4.21
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

28
Figures
Figure 1. Study design
Figure 2. Whole brain analysis comparing the contrast (future value > future control scenarios)
for the affirmed group > control group.
Figure 3. Scatter plot of the residualized percent signal change activity in the valuation network
(VS+VMPFC) ROI from the contrast (value > control scenarios) predicting post intervention
sedentary behavior, controlling for age, gender, education, BMI, and pre intervention sedentary
behavior.
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

254x190mm (72 x 72 DPI)
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

254x190mm (72 x 72 DPI)
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from

254x190mm (72 x 72 DPI)
at University of Pennsylvania Library on November 10, 2015http://scan.oxfordjournals.org/Downloaded from