Posttraumatic stress disorder (PTSD) is a debilitating psychological illness
characterized by significant emotion dysregulation (American Psychiatric Association,
2013). To further refine our clinical
interventions, we must develop a more precise understanding of the mechanisms
underpinning PTSD. One promising candidate mechanism underlying the relationship between
trauma exposure and PTSD is emotion regulation (ER) flexibility (Levy-Gigi et al.,
2016). ER flexibility refers to the
ability to select and switch between different ER strategies according to dynamic
contextual demands (Aldao et al., 2015). ER
flexibility is conceptually aligned with existing theoretical frameworks, including
psychological flexibility (Kashdan & Rottenberg, 2010) and coping flexibility (Cheng, 2001), which emphasise the adaptive benefit of varying one’s
use of regulatory strategies. However, ER flexibility is distinct in its explicit
emphasis on the centrality of aligning one’s ER behaviour according to the
specific contextual demands of each situation. In this way, ER flexibility has been
defined as a multi-componential process encompassing one’s sensitivity to
situational context (context sensitivity),
one’s repertoire of ER strategies (strategy
repertoire), and one’s ability to monitor strategy efficacy
(feedback responsiveness) (Chen & Bonanno,
2021). The synchrony of these three
components underpins ER flexibility and drives adaptive psychological outcomes (Chen
& Bonanno, 2021).
Given the complexity of ER flexibility as a multi-componential process,
measuring ER flexibility in a way that accounts for the dynamic interplay between
strategy use and situational context has been a significant challenge. While there is no
consensus on the best way to measure ER flexibility, popular methodological approaches
have included survey-based questionnaires [e.g., Context Sensitive Index (Chen &
Bonanno, 2021); Flexible Regulation of
Emotional Expression Scale (Burton & Bonanno, 2016)], ecological momentary assessment or experience sampling
methodology (e.g., Chen et al., 2024;
English et al., 2020), and experimental
paradigms (e.g., Sheppes et al., 2011;
2014). Of these, the current study will
focus on experimental methodology. This provides a first step to understanding causal
associations between mechanisms and outcomes. One of the dominant experimental paradigms
used to measure ER flexibility is the regulatory selection paradigm (Sheppes et al.,
2011). This experimental paradigm is
informed by Aldao and colleagues’ (2015) conceptualisation of ER flexibility, which focuses on the
aforementioned components of context sensitivity and
strategy repertoire. In this paradigm,
participants are instructed to minimise their distress elicited by changing contextual
demands - low and high-intensity negative stimuli (e.g., negative images, electric
shocks, or distressing phrases) - by selecting from an available strategy repertoire of
either reappraisal1 or distraction2. In
line with empirical evidence, ER flexibility is operationalised in this paradigm as the
preferential selection of reappraisal when confronted with stimuli low in emotional
intensity, and distraction for stimuli high in emotional intensity (Fine et al.,
2021; Levy-Gigi et al., 2016; Sheppes et al., 2011). This operationalisation rests on prior experimental research
consistently demonstrating that healthy individuals spontaneously switch their ER
strategy based on changes in the emotional intensity of stimuli. Specifically,
individuals overwhelmingly choose to distract in high emotional intensity contexts and
reappraise in low emotional intensity contexts (Fine et al., 2021; Levy-Gigi et al., 2016; Specker et al., 2023). This pattern of regulatory selection has been corroborated
theoretically and using neurophysiological data (Sheppes, 2020). Distraction is thought to be suitable for managing emotions in
high-intensity contexts as it allows individuals to redirect their attention away from
the triggering stimulus (Sheppes & Meiran, 2008). In doing so, distraction prevents premature emotional
processing and the subsequent development of disproportionately negative emotion
(Sheppes, 2020). Using neuroimaging data,
distraction is associated with earlier and stronger modulation of the Late Positive
Potential (LPP; an electrocortical indicator of effective processing of emotionally
arousing information). Strong, early modulation of LPP results in greater activation of
the neural network suppressing amygdala activity associated with fear, making
distraction more suited to high-intensity situations (Kanske et al., 2011; McRae et al., 2010; Sheppes et al., 2011). Conversely, individuals exhibit lower negative emotionality
and emotional reactivity in low-intensity situations (Sheppes, 2020). Therefore, reappraisal may be more suitable
for low-intensity situations where individuals can successfully engage and accurately
reinterpret the meaning of the emotional stimuli to downregulate their emotional
response (Sheppes, 2020). In EEG studies,
reappraisal has been shown to allow for deeper, more meaningful processing of emotional
stimuli. This is evidenced by late modulation of the LPP, and was associated with lower
activation of the neural network governing fear-based responses to emotional situations
via amygdala activity (Kanske et al., 2011;
McRae et al., 2010; Sheppes et al.,
2011). Thus, the tendency to reappraise
low intensity situations and distract from high intensity situations is considered the
most adaptive pattern of ER flexibility.
Burgeoning experimental evidence demonstrates a strong association between
ER flexibility and PTSD using the regulatory selection paradigm (Fine et al.,
2021; Levy-Gigi et al., 2016). Among university students and child sexual
assault survivors, those with higher levels of PTSD exhibited lower levels of ER
flexibility compared to matched controls (Fine et al., 2021). Further, Levy-Gigi and colleagues (2016) found that ER flexibility moderated the
relationship between PTSD and trauma exposure whereby greater trauma exposure only
predicted PTSD if war-exposed firefighters exhibited lower levels of ER flexibility.
Importantly, this well-established dose-response relationship between trauma exposure
and PTSD was non-existent for firefighters with high regulatory flexibility (Levy-Gigi
et al., 2016). These studies present two
critical findings: first, that deficits in ER flexibility are associated with more
severe PTSD symptomatology, and second, that ER flexibility may be protective against
the development of PTSD following trauma. Although these findings advance our current
understanding of the mechanisms underpinning PTSD development, the conclusions drawn
from these paradigms remain correlational in nature. Instead, an experimental paradigm
that manipulates, rather than measures, levels of ER flexibility would allow us to test
the mechanistic influence of ER flexibility in PTSD. One such paradigm is the instructed
ER flexibility paradigm (Specker & Nickerson, 2023).
In line with the prevailing conceptualization of ER flexibility (Aldao et
al., 2015), Specker and Nickerson’s
instructed paradigm (2023) operationalises
ER flexibility as the possession of (1) a repertoire of ER strategies and, (2) context
sensitive strategy use. Here, the ER strategy repertoire comprises both reappraisal and
distraction, and context sensitivity is operationalised as the matching of ER strategies
in line with changing contextual demands (changes in image intensity). Unlike
Sheppes’ (2011) regulatory selection
paradigm which measures the natural regulatory tendencies of participants (spontaneous ER), the instructed ER flexibility paradigm
experimentally manipulates ER flexibility. To do
this, participants are instructed to adopt either a flexible or inflexible ER approach
when confronted with negative images. Emotional responses are elicited via negative
images equally distributed across low and high emotional intensity. This creates two
distinct emotional contexts; high emotional intensity stimuli and low emotional
intensity stimuli. The paradigm comprises four conditions. To manipulate ER flexibility,
participants in the ER Flexible condition are
instructed to distract during high-intensity emotional stimuli and reappraise
low-intensity emotional stimuli. The ER Flexible
condition represents a wide ER strategy repertoire and high context sensitivity, in line
with the prevailing conceptualization of ER flexibility (Bonanno & Burton,
2013). To create a fully crossed design,
three additional control conditions are included: Context
Insensitive, Inflexible
Distraction, Inflexible Reappraisal.
In the first control condition, Context Insensitive,
participants are instructed to adopt the inverse pattern of the ER Flexible condition. That is, participants reappraise high-intensity
emotional stimuli and distract during low-intensity emotional stimuli. This condition
functions as a control to test whether the benefit of a flexible ER approach on
emotional responding is driven by both a wide strategy repertoire and high context
sensitivity (ER Flexible) or simply just a wide
strategy repertoire without any context sensitivity (Context
Insensitive). Two additional inflexible control conditions are included.
In these conditions, participants are instructed to either exclusively reappraise
(Inflexible Reappraisal) or distract (Inflexible Distraction) regardless of changing contextual
demands. Here, participants are instructed to adopt a narrow ER strategy repertoire and
poor context sensitivity. Following each image, emotional responding is indexed via
self-reported negative affect. Previous research using this paradigm found that
individuals with high anxiety reported lower negative affect when instructed to adopt an
ER flexible approach (Specker & Nickerson, 2023).
There is mounting evidence to suggest that lower levels of ER flexibility
are implicated in PTSD (Fine et al., 2021;
Levy-Gigi et al., 2016). However, there is
an apparent deficit of evidence causally relating ER flexibility to emotional responding
in PTSD. Considering the remaining gaps in the PTSD literature, the present study used
the instructed ER flexibility paradigm to investigate the mechanistic influence of ER
flexibility on emotional responding among trauma-exposed individuals with and without
probable PTSD. In line with Specker and Nickerson’s (2023) findings demonstrating a temporal link between
ER flexibility and anxiety, we predicted that being guided in how to adopt an ER
flexible approach when confronted with trauma-salient stimuli would be most effective at
downregulating negative affect for those with probable PTSD. This is because PTSD, like
anxiety, has been associated with marked deficits in ER flexibility relative to cohorts
without PTSD (Fine et al., 2021).
Specifically, we hypothesised that participants with probable PTSD in the ER Flexible condition would exhibit significantly lower
negative affect in response to trauma-salient stimuli than those in the inflexible ER
control conditions. For those without probable PTSD, we did not expect to detect a
pronounced difference in emotional responding between the instructed conditions.
Materials and Methods
Participants
Participants were recruited via Amazon’s Mechanical Turk
(MTurk) across June and July 2023. MTurk is the most widely used platform for
online studies as it is low-cost and yields high-quality data (Aguinis et al.,
2021). Additionally, MTurk
yields similar prevalence rates of trauma exposure and psychopathology as
lab-based samples (van Stolk-Cooke et al., 2018). Eligibility criteria required participants to have
experienced a Criterion A traumatic event according to the Fifth Edition of the
Diagnostic Statistical Manual for Mental Disorders (American Psychiatric
Association, 2013), be fluent in
English, be at least eighteen years old, reside in the United States, and have
completed at least 1000 Human Intelligence Tasks (HIT) with a HIT approval ratio
(HAR) of at least 99%. Several response validity indicators were
incorporated into a Part 1 screener in line with established recommendations to
identify and exclude suspicious responders (e.g., bots) and inattentive
responders (See Supplement A of the Supplemental Materials). The recruitment and
screening process for parts 1 and 2 are depicted in Fig. 1. An a priori power analysis conducted via
GPower 3.1.9.7. was conducted to determine our target sample size. This
indicated that a sample size of 126 participants was required to detect a small
to medium effect size (f = 0.03) with 80% power (1-b) and an error
rate (a) of 0.05. The final sample comprised 148 participants.
Fig. 1
Flowchart depicting participant exclusions during the
recruitment and screening phases of Part 1 and 2
×
![]()
Measures
Demographics
Participants provided their age, gender, ethnic background, and
level of educational attainment.
Trauma Exposure
The Life Events Checklist – Fifth Edition (LEC-5)
(Weathers et al., 2013) was
used to screen for lifetime exposure to potentially traumatic events. The
LEC-5 comprises 16 potentially traumatic events according to the DSM-5
Criterion A definition of a traumatic event. For each event, participants
select any of the following response options that apply: (1) Happened to me,
(2) Witnessed it, (3) Learned about it, (4) Part of my job, (5) Not sure, or
(6) Doesn’t apply (Weathers et al., 2013). An item was considered endorsed if either (1), (2)
or (4) were selected. A total score was then computed by summing endorsed
items. To reduce demand characteristics during the initial screening phase,
20 ‘filler’ items describing neutral or positive events were
added to the existing items (Jobson et al., 2022). The LEC-5 has good convergent validity with the
gold-standard Clinician-Administered PTSD Scale (Gray et al., 2004) and is routinely used in
trauma-exposed US samples (Gray et al., 2004).
PTSD Symptoms
The Posttraumatic Stress Disorder Checklist for DSM-5 (PCL-5)
(Weathers et al., 2013) indexed
PTSD symptom severity. The PCL-5 is a 20-item measure that assesses DSM-5
PTSD symptoms (Weathers et al., 2013). Each item is measured on a 5-point Likert scale
(0 = Not at all,
4 = Extremely). Items
are added to generate a total score ranging from 0 to 80 (van Stolk-Cooke et
al., 2018). The PCL-5 provides
a clinical cut-off score, with scores above 33 suggesting a probable PTSD
diagnosis (Verhey et al., 2018). The PCL-5 is a gold-standard tool for assessing PTSD
symptoms, with good construct and convergent validity with other measures of
PTSD (McDonald & Calhoun, 2010). The PCL-5 demonstrated excellent internal
consistency in the current sample (α = 0.964).
Emotional Stimuli
Images presented in this study were identical to those used in
Sheppes and colleagues’ (2011) regulatory selection paradigm. This set comprises
30 trauma-related International Affective Picture System (IAPS; Bradley
& Lang, 2017) images
depicting interpersonal violence, deceased people and animals, civil unrest,
mutilated bodies, and insects. There were 15 low emotional intensity images
(Mean ValenceLow = 3.41; Mean
ArousalLow = 5.01) and 15 high emotional
intensity images (Mean ValenceLow = 1.99; Mean
ArousalLow = 6.12), categorised based on their
normative ratings for arousal (Sheppes et al., 2011). To improve validity, image
content was approximately matched across low- and high-emotional intensity
categories (Levy-Gigi et al., 2016).
Nature Images
A total of nine images depicting a natural landscape were used
in the present study to improve executive attention by minimising
participant fatigue (Gamble et al., 2014). A nature image was presented after every set of
three negative images. Images depicted scenes of waterfalls, lakes, and
beaches. All images were approximately matched in content.
Negative Affect
Consistent with past ER experimental paradigms (Sheppes et al.,
2014; Specker &
Nickerson, 2023), negative
affect was measured using a single self-report item following the
presentation of each negative image. Participants were asked to rate how
negatively they felt after viewing the experimental image and using their
assigned ER strategy (“How negative did the
picture make you feel?”) on a 9-point Likert scale
(1 = Not negative at
all, 9 = Very
negative) (Sheppes et al., 2014). A mean score was calculated to index overall
negative affect.
Manipulation Check
Consistent with past ER experimental paradigms, participants
were asked to, “Briefly describe what you
thought about while viewing the image you just saw” to
verify the correct implementation of their assigned ER strategy (Sheppes et
al., 2011; Specker &
Nickerson, 2023). A total of 16
open-ended validity checks were included in the study. Validity checks were
presented after each of the eight training and practice items and validity
checks were also presented an additional eight times randomly throughout the
experimental phase. Consistent with previous studies (e.g., Fine et al.,
2021; Levy-Gigi et al.,
2016; Specker &
Nickerson, 2023), responses
were inspected by a single rater blinded to the participants’
instructed ER condition allocation. Responses were flagged if they
contradicted the experimental instructions, for example, if a participant
distracted despite being told to reappraise, or demonstrated a gross
misapplication of the assigned strategy, for example, not distracting or
reappraising. Participants who failed four or more checks were excluded from
the analysis (see Fig. 1).
Consistent with similar previous studies, 22% of participants were
excluded for failing this manipulation check (Specker & Nickerson,
2023).
Procedure
The current study was conducted online using Qualtrics. Data
collection occurred over two sessions – Part 1 and 2 – conducted
approximately seven days apart. Following informed consent, participants
completed the English proficiency test, demographic questionnaire and the
attention and validity indicator items (object check, conscientious responder
check, and image check; see Supplement A of the Supplemental Materials).
Participants then completed a modified version of the LEC-5 containing
additional neutral and positive events to assess levels of trauma exposure while
reducing demand characteristics. After Part 1, participants were debriefed and
received USD$1.50 in compensation.
Eligible participants were sent an advertisement and survey link
for Part 2 via MTurk. After consenting to participate in Part 2, trauma exposure
was re-assessed, this time using only the original LEC-5 (i.e., with no filler
items), and PTSD symptom severity was measured using the PCL-5.
Next, participants proceeded to the instructed ER flexibility
paradigm. The current study replicated the experimental procedure for the
instructed ER flexibility paradigm outlined by Specker and Nickerson
(2023), with the addition of
nature images between experimental blocks. Nature images were added to minimise
fatigue by improving executive attention (Gamble et al., 2014). The instructed ER flexibility
paradigm comprised three phases: teaching, practice, and experimental.
In the teaching phase, participants were briefly introduced to the
experimental task and oriented to the negative affect scale. Participants then
received instructions on distraction and reappraisal via written text and
instructional videos (links to the videos and instructions can be found in
Supplement B of the Supplemental Materials). The duration of each instructional
video was 2 min and 30 s. The videos provided a brief explanation
of the strategy and a demonstration of how to use the strategy whilst viewing a
negative image like those shown in the experimental phase. After viewing each
video, participants practised implementing the target strategy while viewing one
low-intensity image and one high-intensity image. The order in which
participants learnt the two ER strategies was counterbalanced.
In the practice phase, participants were provided further details
on the experimental procedure and reminded to only use the assigned strategy.
Then, participants completed two distraction trials and two reappraisal trials,
each comprising a low and high-intensity image. In each practice trial,
participants first previewed the image for one second. Then, participants were
informed in bold written text which strategy to employ when the image returned.
When the participant had finished reading the instruction, the same image
reappeared for six seconds whilst the participant attended to the image and
implemented their instructed strategy. Each participant was asked to complete an
open response item asking them to describe how they distracted or reappraised
the image. These responses were blindly coded after completion. Participants
then rated their negative affect.
In the experimental phase, participants were randomised to one of
four conditions: ER Flexible, Context Insensitive,
Inflexible Distraction, and Inflexible
Reappraisal. Participants assigned to the ER Flexible condition were instructed to use distraction during
high-intensity images and reappraisal during the presentation of low-intensity
images. Participants assigned to the Context
Insensitive condition were instructed to use reappraisal during
high-intensity images and distraction during low-intensity images (i.e., the
inverse pattern to the ER Flexible
condition). In Inflexible Distraction and
Inflexible Reappraisal conditions,
participants were instructed to employ a single ER strategy (either reappraisal
or distraction) regardless of changes in contextual demands (image intensity).
Participants were not informed that the experimental images varied in negative
emotional intensity. Images were arranged into ten “blocks”. Each
“block” contained three negative images of the same emotional
intensity, so participants used the same ER strategy when attending to images in
the same “block”. Images were arranged this way to minimise the
potential added cognitive load posed by being involved in the experimental
conditions that required participants to switch between the two strategies
(ER Flexible and Context Insensitive) (Specker & Nickerson, 2023). The order of the manipulation checks,
images, and blocks was randomised and then fixed across conditions, in line with
Specker and Nickerson’s (2023) experimental procedure. The structure and timing of the
experimental trials are shown in Fig. 2. Part 2 took approximately 40 min to complete. After
the experiment, participants were debriefed and compensated with USD$4.50. The
[De-Identified] Human Research Ethics Committee provided ethics approval
([De-identified]).
Fig. 2
Illustration of the instructed ER flexibility
experimental trial structure.
Ms = milliseconds
×
![]()
Data Analysis
Statistical analyses were conducted using IBM SPSS version 29.
Chi-square tests, a one-way Analysis of Variance (ANOVA), and an independent
samples t-test were conducted to identify between-group differences (i.e.,
between the probable PTSD groups and between the instructed ER conditions) in
demographic characteristics. A paired samples t-test was used to verify that the
experimental manipulation of image intensity to generate distinct
‘low’ and ‘high’ negative emotional intensity
contexts was successful. For the main analysis, a 2 (probable PTSD) x 4
(instructed ER conditions) Analysis of Covariance (ANCOVA) with planned
comparisons was used to examine whether instructing participants in ER
flexibility (ER Flexible condition) would
result in lower negative affect compared to the three control conditions
(Context Insensitive, Inflexible Distraction,
Inflexible Reappraisal), particularly among those with probable
PTSD. Instructed ER conditions (ER
Flexible, Context
Insensitive, Inflexible
Distraction, Inflexible
Reappraisal) and probable PTSD groups (probable PTSD, no PTSD) were
entered as between-subject factors. Since age differed between the instructed ER
conditions, it was entered into the model as a covariate. The Benjamini-Hochberg
false discovery rate procedure was used to statistically control for the use of
multiple comparisons (Benjamini & Hochberg, 1995). This procedure is less stringent than the Bonferroni
correction, and considered a valid alternative that balances both the objectives
of reducing the family wise error rate while maximising statistical power to
detect a true effect (Ioannidis, 2005; Perneger, 1998).
Results
Sample Characteristics and Probable PTSD
Sociodemographic characteristics are provided in
Table 1. Participants were
predominantly male (54.6%) and predominantly white (78.4%). 24
(16.2%) participants met the criteria for a probable PTSD diagnosis on
the PCL-5, while 124 (83.8%) participants did not met criteria for
probable PTSD (Ashbaugh et al., 2016). Frequencies of participant exposure to potentially
traumatic events can be found in Supplement C of the Supplemental Materials. The
most frequently endorsed traumatic events were transportation accidents
(77%), natural disasters (56.8%), physical assault (48.6%),
and a life-threatening illness or injury (43.2%).
Table 1
Demographic characteristics for the overall
sample
Demographic Variables | M (SD) or n
(%) |
|---|---|
Age | 42.2 (11.4) |
Gender | 81 (54.6%) male, 64 (43.2%)
female |
Other | 1 (0.7%) |
Prefer not to answer | 2 (1.4%) |
Cultural background | |
White or Caucasian | 116 (78.4%) |
Black or African American | 12 (8.1%) |
Latino or Hispanic | 5 (3.4%) |
American Indian or Alaska
native | 1 (0.7%) |
Asian or South Asian | 6 (4.1%) |
Native Hawaiian or Pacific
Islander | 1 (0.7%) |
Other | 4 (2.7%) |
Prefer not to answer | 3 (2.0%) |
Employment status | |
Employed full-time | 110 (74.3%) |
Employed part-time | 13 (8.8%) |
Employed casually | 4 (2.7%) |
Unemployed | 4 (2.7%) |
Homemaker | 3 (2.0%) |
Retired | 11 (7.4%) |
Highest educational attainment | |
Junior High | 1 (0.7%) |
Senior High | 16 (10.8%) |
Some college credit | 31 (20.9%) |
Undergraduate Degree | 75 (50.7%) |
Vocational Qualification | 4 (2.7%) |
Masters Degree | 17 (11.5%) |
Doctoral Degree | 2 (1.4%) |
Prefer not to answer | 2 (1.4%) |
Preliminary Analyses
To verify random allocation and test whether there were any
systematic differences between instructed ER conditions or probable PTSD groups,
the conditions and groups were compared across demographic variables.
Instructed ER Conditions
Chi-squared tests indicated no significant differences between
instructed ER conditions in gender, χ2 (3, N = 145) = 7.47, p = .058, ϕ = 0.04, ethnicity, χ2 (3, N = 148) = 0.473, p = .925, ϕ = 0.05, employment, χ2 (3, N = 148) = 0.62, p = .891, ϕ = 0.04, or education, χ2 (3, N = 146) = 3.06, p = .383, ϕ = 0.07. However, a one-way ANOVA
showed significant differences between the instructed ER conditions with
respect to age, F(3,
144) = 2.70, p = .048,
ηp2 = 0.05.
Thus, age was controlled for in the main analyses. Overall, these findings
show that random allocation was largely successful.
Probable PTSD Groups
Chi-square tests indicated that probable and non-probable PTSD
groups did not significantly differ in ethnicity, χ2 (31, N = 148) = 1.22, p = .270, ϕ = 0.09, employment, χ2 (1, N = 148) = 2.36, p = .124, ϕ = 0.13, education, χ2 (1, N = 146) = 0.81 p = .369, ϕ = 0.07, or gender, χ2 (3, N = 148) = 6.61, p = .085, ϕ = 0.21. An independent-samples t-test
showed that age did not significantly differ among those with and without
probable PTSD, t(148) = 0.35, p = .073, d = 0.08. Further, the distribution of
participants in the probable PTSD vs. non-probable PTSD groups did not
significantly differ between the instructed ER conditions, χ2 (3, N = 148) = 4.90, p = .179, ϕ = 0.12. These findings demonstrate
that there were no systematic differences between the probable and
non-probable PTSD groups.
Verification of Image Intensity
A paired samples t-test indicated that there was a significant
difference between mean negative affect for images of low (M = 2.56, SD = 1.25) and high (M = 5.30, SD = 1.83) negative emotional intensity,
t(145) = 24.92,
p < .001. This
shows that the low- and high-intensity images elicited two distinct levels
of negative emotional responses in line with their pre-determined intensity,
thus indicating that the experimental manipulation of emotional context was
valid.
ANCOVA Analyses
A 2 (probable PTSD Group) x 4 (Instructed ER Condition)
between-subjects ANCOVA with planned comparisons was performed to test the
hypotheses of the current study. For completeness, the results of the overall
ANCOVA model are presented in Table 2 and all pairwise comparisons are presented in Supplement D
of the Supplemental Materials. As expected, there was a significant main effect
for probable PTSD group, F(1,
138) = 10.80, p = .001,
ηp2 = 0.05,
such that participants with probable PTSD (M = 4.93, SD = 0.32) reported significantly greater negative
affect than those without probable PTSD (M = 3.80, SD = 0.12). The main effect of instructed ER condition
on negative affect was not significant, F(3,
138) = 2.23, p = .087,
ηp2 = 0.05.
The interaction effect between instructed ER condition and probable PTSD groups
was significant, F(3,
138) = 6.46, p = .016,
ηp2 = 0.05.
This indicated that the impact of the instructed ER conditions on participant
negative affect differed according to the participant’s probable PTSD
status. Planned comparisons exploring the impact of the instructed ER conditions
on negative affect among the probable and non-probable PTSD groups were
separately conducted to disaggregate this effect.
Table 2
ANCOVA of the effect of instructed ER and PTSD on
negative affect
Variables | F |
p
| ηp2 |
|---|---|---|---|
Covariate | |||
Age | F(1,138) = 0.58 | 0.447 | 0.073 |
Main Effect | |||
Probable PTSD group | F(1,138) = 10.80 | 0.001 | 0.073 |
Instructed ER condition | F(3,138) = 2.23 | 0.087 | 0.046 |
Interaction | |||
PTSD x Condition | F(3,138) = 3.58 | 0.016 | 0.072 |
Planned Comparisons
Testing whether instruction in ER flexibility (ER Flexible
condition) results in lower negative affect for those with probable PTSD. For
participants with probable PTSD, mean negative affect significantly differed
between the instructed ER conditions, F(3,
138) = 3.50, p = .017,
ηp2 = 0.07.
Specifically, planned comparisons indicated that, among participants with
probable PTSD, those in the ER Flexible
condition (M = 3.53, SE = 0.45) exhibited significantly
lower mean negative affect when compared to those in the Inflexible Distraction (M = 5.46, SE = 0.51), p = .005, 95% CI [-3.28, -0.58], Context
Insensitive (M = 5.10, SE = 0.55), p = .028, 95% CI [-2.97, -0.18], and Inflexible
Reappraisal (M = 5.63, SE = 0.95) conditions, p = .047, 95% CI [-4.18, -0.03]. These comparisons remained significant after
evaluating them against corrected alpha significance levels, calculated using
the Benjamini-Hochberg procedure (adjusted α values were:
α = 0.0167 [comparison 1], α = 0.0333
[comparison 2], α = 0.05 [comparison 3]). Mean negative
affect did not significantly differ among the instructed ER conditions for
participants without probable PTSD, F(3,
138) = 1.69, p = .171,
ηp2 = 0.04.
The mean negative affect for those with and without probable PTSD in each
instructed ER condition is depicted in Fig. 3.
Fig. 3
Mean negative affect among probable PTSD groups across
instructed ER conditions. Note. Error bars
represent ± 1 standard error. Significance
brackets show differences between conditions among participants
with probable PTSD. The significance brackets show statistically
significant differences between the ER
Flexible condition and the Context Insensitive (p = .028), Inflexible Distraction (p = .005), and
Inflexible Reappraisal
(p = .047)
conditions
×
![]()
Discussion
The current study used an instructed ER flexibility paradigm to assess
whether ER flexibility functioned as a mechanism influencing emotional responding in
those with probable PTSD. Our results revealed that training trauma-exposed
participants with probable PTSD in ER flexibility was the most effective way to
manage negative affect relative to all other regulatory approaches tested. This
supported our hypothesis, which posited that individuals with probable PTSD would
derive the greatest benefit from being instructed in ER flexibility. It is also
noteworthy that the same pattern of findings was not found for those without
probable PTSD. Although the present study was preliminary, our findings may indicate
the potential protective effects of ER flexibility for those with PTSD. The present
findings thus build upon the existing literature. To date, the relationship between
PTSD and ER flexibility has only been demonstrated in a correlational manner (Fine
et al., 2021; Levy-Gigi et al.,
2016). Studies have shown
associations between greater ER flexibility deficits and greater severity of PTSD
symptomatology (O’Brien et al., 2023). Such ER flexibility deficits include limited access to ER
strategies, poor emotional awareness, and an inability to select appropriate
strategies to manage emotions (O’Brien et al., 2023; Spikol et al., 2024). However, an integral requirement for classifying ER
flexibility as a psychological mechanism underpinning PTSD is to demonstrate this
relationship causally (Ehring et al., 2022). Specifically, high levels of ER flexibility must be shown
to induce improvements in subsequent emotional responding. To our knowledge, this is
the first study to experimentally manipulate ER flexibility and quantify its
real-time impact on emotional responding in an analogue probable PTSD sample.
Our findings accord with prior research by Levy-Gigi and colleagues
(2016), which provided indicators
of the mechanistic influence of ER flexibility on the development of PTSD. Levy-Gigi
and colleagues found that the relationship between trauma exposure and subsequent
PTSD symptomatology among war-exposed individuals was moderated by ER flexibility.
Specifically, individuals with low ER flexibility exhibited a significant
correlation between trauma-exposure and PTSD symptoms, while those with high ER
flexibility did not exhibit a dose-response relationship between trauma exposure and
subsequent PTSD symptoms. Our findings extend this prior correlational research to
demonstrate that trauma-affected participants with probable PTSD who adopted an ER
flexible approach when confronted with trauma-salient stressors exhibited small yet
significant improvements in emotional responding. This highlights that
experimentally inducing higher ER flexibility in individuals with probable PTSD
conferred psychological benefits. Together, these findings may suggest that ER
flexibility may be a viable mechanism underpinning PTSD that warrants further
investigation.
Research investigating the role of ER flexibility in the development of
PTSD may be fruitful in elucidating components of emotion regulation that facilitate
recovery. Well-established cognitive and cognitive-behavioural models of PTSD
consistently identify deficits in emotion regulation as key maintaining factors
(Ehlers & Clark, 2000; Foa et
al., 1989). Although investigations of
ER flexibility in the context of PTSD are in their infancy, key characteristics of
PTSD typify ER flexibility deficits, including a limited ER repertoire (marked by an
over-reliance on avoidance-based emotion regulation strategies) and limited
contextual sensitivity (marked by overgeneralised threat perception and heightened
emotional reactivity) (McLean & Foa, 2017; Tull et al., 2020). It is logical then that the provision of explicit guidance
in how to employ a more varied ER repertoire that is sensitive to context may result
in improved emotional responding specifically among participants with such deficits.
This suggests that ER flexibility may be a useful clinical target in interventions
for PTSD. Replication of our study, particularly with larger samples of individuals
with clinically diagnosed PTSD, is thus an important next step. Although
preliminary, findings from this study suggest that ER flexibility training may be a
useful adjunct to PTSD treatments. Existing efficacious PTSD treatments,
particularly those with ER skills training components such as Skills Training in
Affect and Interpersonal Regulation (STAIR; Cloitre et al., 2002) and Dialectical Behaviour Therapy for PTSD
(DBT-PTSD; Bohus et al., 2019), may be
complemented by including explicit training on how and
when to use specific ER strategies. In particular, while traditional
ER skills training tends to predominantly focus on building a client’s
repertoire of effective ER strategies, ER flexibility may be fostered by
additionally building skills in perceiving contextual changes (i.e., building an
understanding of the different emotional stressors that the client routinely
encounters), and building skills in selecting and implementing an appropriate
strategy in response to specific stressors (i.e., developing an understanding of
which strategies from the client’s personal repertoire best assists with
managing symptoms relating to particular stressors). Although tentative, further
research may be warranted to explore the potential need for, and effectiveness of,
skills training in ER flexibility.
Several Limitations are Worth Noting
Firstly, it is noteworthy that several valid conceptualisations of
ER flexibility exist. The present study was informed by Aldao and
colleagues’ (2015)
conceptualisation, which predominantly focuses on one’s strategy
repertoire and context sensitivity. However, Bonanno and Burton (2013) have conceptualised ER flexibility as
the interplay between three key components, namely strategy repertoire, context
sensitivity and feedback responsiveness. Feedback responsiveness was not
investigated in the present study. Future expansions of this experimental
design, that measure a participant’s ability to monitor the efficacy of
the instructed ER strategy and adjust if necessary, would be a fruitful avenue
for future research. Secondly, given the experimental nature of the study, the
instructed ER flexibility paradigm (Specker & Nickerson, 2023) used in the present study employed a
narrow operationalisation of ER flexibility. The rationale for this
operationalisation was two-fold. First, operationalising ER flexibility using
distraction and reappraisal strategies is consistent with prior experimental ER
studies and relevant to the presentation of PTSD (Fine et al., 2021; Levy-Gigi et al., 2016; Sheppes, 2020). This is because distraction and
reappraisal serve as exemplars of disengagement and engagement strategies,
respectively. Thus, keeping this methodological aspect consistent with the
broader literature facilitates comparison with existing studies. Second, it was
essential to constrain the operationalisation of ER flexibility, both the ER
repertoire and the contextual demands of the stimuli, to ensure that we were
precisely measuring its impact on subsequent emotional responding. However, a
narrow operationalisation of ER limits the generalisability and ecological
validity of our findings. Future studies should employ different
operationalisations of ER flexibility, such as modifying the contextual demands,
or the ER strategies used, to improve the validity of the present findings.
Instead of emotional intensity, contextual demands could include one’s
personal goals, the perceived controllability of the situation (Troy et al.,
2013), or the presence and absence of emotion-evoking cues (Spikol et al.,
2024). One potential way to
achieve this is to vary the images which trigger emotional responding based on
personal relevance to the participant. Furthermore, to improve the ecological
validity of the emotional triggers, richer more evocative emotional stimuli such
as a video or personally relevant stimuli could be used. Alternatively,
different ER strategies may be used instead of distraction and reappraisal, such
as suppression and acceptance (Aldao et al., 2010).
Thirdly, there may be limitations in the external validity of our
sample. The sample of participants recruited in the present study contained a
small percentage of those with probable PTSD. Additionally, our participants
were recruited via MTurk. While MTurk samples can recruit reliable and
comparable trauma-exposed samples (Engle et al., 2020), we note that our sample largely comprised
highly-educated white males. This differs from the demographic characteristics
typically seen in the general PTSD population in the US, which tends to comprise
a larger proportion of female, non-white, and lower-income individuals (Schein
et al., 2021). Accordingly, the
present sample may not be generalisable to the broader PTSD population in the
US. A potential reason for this discrepancy may be the stringent data validity
checks used in the present study, which required participants to have completed
a large number of prior MTurk studies (1000 Human Intelligence Tasks [HITs]) to
be eligible for the present study. This criterion was included to safeguard a
high level of data quality, but may have had the unintended consequence of
biasing the sample towards individuals that had more spare time to dedicate to
online research. Considering our findings, it may be worthwhile to replicate the
current study using purposive sampling to obtain a larger, and potentially more
representative, sample of individuals with PTSD. Additionally, while our sample
comprised participants with a diversity of traumatic experiences, the specific
influences of different types of traumatic events (e.g., transportation accident
vs. physical assault) or different levels of exposure (e.g., directly
experienced vs. witnessed) on the relationship between instructed ER
flexibility, probable PTSD and negative affect were not investigated. Owing to
the small sample size of the current study, it was not feasible to conduct such
fine-grain analysis. However, such investigations are important to advancing our
understanding of when, and for whom, instructed ER flexibility may be most
helpful, and thus constitute a worthwhile avenue for future research.
Finally, measurement of key constructs could also be improved. The
present study used the PTSD Checklist for the DSM-5 to screen participants with
a probable diagnosis of PTSD(PCL-5; American Psychiatric Association,
2013) as it is the gold-standard
measure of probable PTSD. However, the PCL-5 is a self-report questionnaire
which may limit its response validity. To improve this, the Clinician
Administered PTSD Scale for DSM-5 may be a more suitable measure (CAPS-5;
Weathers et al., 2018). The CAPS-5
is a widely used structured diagnostic interview that is implemented by trained
clinicians to diagnose posttraumatic stress disorder and measure symptom
severity (Weathers et al., 2018).
Additionally, a single-item measure of negative affect was used to measure the
impact of ER flexibility on emotional responding. Although this accords with
conventions in the field (Sheppes, 2020), the impact of ER flexibility on emotional responding
could be more comprehensively captured by multidimensional methods of
assessment. For instance, psychophysiological data such as skin conductance may
be collected as an objective, biological marker of emotion dysregulation via
autonomic nervous system activity (Fitzgerald et al., 2022). Alternatively, a more comprehensive
measure of negative emotion may be used, such as the negative affect scale of
the Positive and Negative Affect Schedule (PANAS; Watson et al., 1988). The negative items of the PANAS
capture emotions such as anger and guilt (Watson et al., 1988), which are both highly associated with
the persistent negative emotional states experienced by those with PTSD (Badour
et al., 2017).
Notwithstanding these limitations, the present findings offer
preliminary evidence that ER flexibility is a viable mechanism implicated in
PTSD. We sought to investigate whether experimentally manipulating ER
flexibility would influence emotional responding among trauma-affect
participants. Our results demonstrated that trauma-exposed individuals with
probable PTSD benefitted from being instructed in using an ER flexible approach,
characterised by a wider strategy repertoire and improved context sensitivity.
These findings implicate ER flexibility as a potential mechanism underpinning
PTSD and illuminate ER flexibility as potentially relevant clinical target in
PTSD treatment.
Acknowledgements
This work was supported by the UNSW Science Social Good Fund (2023) and P.
Specker was supported by an MQ: Transforming Mental Health Postdoctoral Scholarship
(MPSIP\15). Angela Nickerson was supported by an Australian National Health and
Medical Research Council Investigator Leadership Grant. We gratefully acknowledge
Joshua Wong and Vivian Mai (School of Psychology, University of New South Wales) for
their advice regarding study programming and participant recruitment. We wish to
thank all participants for their contributions to this study.
Declarations
Competing Interests
We have no conflicts of interest to declare with respect to our
authorship or the publication of the present article.
Open Access
This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.