See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/257795398

Affective Reactivity Mediates an Inverse Relation Between Mindfulness and

Anxiety

Article in Mindfulness · October 2014

DOI: 10.1007/s12671-013-0206-x

CITATIONS READS

33 524

3 authors, including:

Brian Ostafin Jessica Brooks

University of Groningen Georgia Southern University
87 PUBLICATIONS 4,121 CITATIONS 8 PUBLICATIONS 132 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Brian Ostafin on 20 January 2014.

The user has requested enhancement of the downloaded file.

Mindfulness
DOI 10.1007/s12671-013-0206-x

ORIGINAL PAPER

Affective Reactivity Mediates an Inverse Relation

Between Mindfulness and Anxiety
Brian D. Ostafin & Jessica J. Brooks & Magali Laitem

# Springer Science+Business Media New York 2013

Abstract Little is known about the mechanisms through used both interventions and individual differences methods to
which mindfulness is related to psychological symptoms examine the relation between mindfulness and psychological
such as anxiety. One potential mechanism consists of indi- distress. Intervention research has found reductions in psycho-
vidual differences in emotion-responding variables such as logical distress after mindfulness training ranging from a single
reactivity to aversive stimuli. The current research was session (Arch & Craske 2006) to 10-day intensive retreats
designed to examine whether affective reactivity may act (Ostafin et al. 2006) to weekly sessions administered over the
as a mechanism of mindfulness. Across two studies, an course of 2 months (Carmody & Baer 2008).
inverse relation between trait mindfulness (specifically, the These results are supported by findings of an inverse rela-
Nonjudging of Inner Experience and Acting with Awareness tion between individual difference measures of mindfulness
factors of the Five-Facet Mindfulness Questionnaire) and and psychological distress (Brown & Ryan 2003). Baer et al.
chronic anxiety was partially mediated by affective reactiv- (2006) have operationalized mindfulness as consisting of
ity, assessed with direct (self-report in study 1) and indirect the following five factors (assessed with the Five-Facet
(lexical decision task in study 2) measures. These results Mindfulness Questionnaire; FFMQ): Observing—the aware-
contribute to the understanding of the psychological mech- ness of internal and external phenomena; Describing—the
anisms through which mindfulness works. tendency to apply words to experience; Acting with
Awareness—engaging awareness to ongoing activities, rather
Keywords Mindfulness . Meditation . Anxiety . Emotion than allowing them to occur mindlessly; Nonjudging of Inner
regulation . Self-regulation Experience—an accepting, nonevaluative attitude toward
thoughts and feelings involving a letting go of an agenda to
maintain or change one’s experience; and Nonreactivity to
Introduction Inner Feelings—the ability to let thoughts and feelings pass,
rather than getting caught up in them. The importance of
Mindfulness, a central component of Buddhist contemplative assessing trait differences in mindfulness is demonstrated by
practice (Thera 1965), can be described as involving both an evidence that mindfulness training leads to increases in self-
awareness and a receptive acceptance of internal experience report measures of trait mindfulness (Brown & Ryan 2003)
(Bishop et al. 2004; Brown & Ryan 2003; Kabat-Zinn 1990). and that an increase in trait mindfulness mediates the relation
There is increasing evidence that mindfulness is inversely relat- between mindfulness training and psychological distress
ed to indices of psychological distress, such as depression and (Carmody & Baer 2008; Shapiro et al. 2008).
anxiety (Baer 2003; Brown & Ryan 2003). Researchers have Why is mindfulness related to lower levels of psycholog-
ical distress? Although there are a number of potential
mechanisms through which mindfulness may work (see
B. D. Ostafin (*)
Department of Psychology, University of Groningen, Brown et al. 2007; Shapiro et al. 2006), relatively little
Grote Kruisstraat 2/1, 9712 Groningen, Netherlands research has examined this question. One potential mecha-
e-mail: [email protected] nism of the relation between mindfulness and psychological
J. J. Brooks : M. Laitem
distress consists of differences in emotion responding vari-
Department of Psychology, North Dakota State University, ables, such as affective reactivity (the peak amplitude of the
Dept. 2765, P.O. Box 6050, Fargo, ND 58108, USA affective response—e.g., whether a fear response is mild or
 Mindfulness

intense), the latency until peak response is reached and the factors of mindfulness (Baer et al. 2006). Research with this
rate of affective recovery toward baseline (Davidson 1998; measure suggests that several of the factors may be more
Koole 2009). Although emotion responding is influenced by important than others in modulating psychological distress.
genetics, evidence for the contribution of the environment Baer et al. (2006) found that, when entered simultaneously in
indicates that temperament is not destiny (Soussignan et al. a regression analysis, the Acting with Awareness, Nonjudging,
2009) and suggests the need for determining the types of and Nonreactivity factors demonstrated an inverse relation with
learning that can reduce aversive reactions to acute stressors. psychological symptoms whereas the Describe factor did not
Mindfulness may represent one such practice that can influ- (the Observe factor was not included in the analysis because it
ence emotion responding. was positively related to psychological distress).
One specific hypothesis is that an inverse relation between In order to establish a variable as a mechanism of mind-
mindfulness and chronic psychological distress is partially fulness, it must: (1) be related to mindfulness, (2) be related
explained by mindful individuals demonstrating less affective to an outcome variable (e.g., psychological distress), and (3)
reactivity to aversive events. This idea is supported by research account for variance of the relation between mindfulness
showing that mindfulness is related to less reactivity to aver- and the outcome variable (Baron & Kenny 1986). Evidence
sive stimuli. In one study, affective reactivity was examined by for affective reactivity as a mechanism of mindfulness
assessing self-reported affect after viewing aversive pictures would thus include demonstrating a relation between it and
(Arch & Craske 2006). The results indicated that compared to both mindfulness and psychological distress in the same
participants who received an induction to elicit worry, those study. The current research was designed to extend previous
who received a brief mindfulness training experienced less work on affective reactivity as a potential mechanism of
(self-reported) negative affect after viewing unpleasant pic- mindfulness by examining whether it would be related to
tures. In a second study, affective reactivity to unpleasant both mindfulness and chronic anxiety, thus demonstrating
pictures was assessed with both direct (self-report ratings of the first two requirements of mediation. Further, we exam-
intensity) and indirect (i.e., not relying on introspective self- ined the third requirement of mediation by assessing wheth-
report; in this case, skin conductance response) measures er affective reactivity accounted for a statistically significant
(Ortner et al. 2007; study 2). The results showed that compared portion of the relation between mindfulness and anxiety.
to a wait-list control group, a mindfulness training group
demonstrated less reactivity measured by both self-report and
skin conductance response. Further, the mindfulness group Pilot Study
showed a smaller skin conductance response to the unpleasant
pictures than did a relaxation training group. A third study Studies 1 and 2 were designed to examine whether the
examined the effects of a mindfulness intervention on affective inverse relation between mindfulness and chronic anxiety
reactivity in a sample of individuals diagnosed with social is explained, in part, by less affective reactivity to aversive
anxiety (Goldin & Gross 2010). Both before and after an 8- events. A pilot study was conducted in order to establish the
week mindfulness intervention, participants read negative so- validity of an emotion induction procedure and the use of a
cial anxiety-related statements (e.g., “People always judge lexical decision task (LDT) as an indirect measure (i.e.,
me”) and rated their affect immediately afterward (and again using reaction time performance rather than introspective
after 12 s of focusing attention on their breath). Contrary to the self-report) of affective reactivity. Evidence for using a LDT
two previous studies, the findings indicated that mindfulness to assess affective reactivity includes findings that eliciting
training did not reduce affective reactivity (but that it did emotion increases the speed of perceiving verbal stimuli that
improve the influence of the breath focus on reducing negative have the same valence as the elicited emotion and that
affect). The discrepancy between the positive results of the first intensity of self-reported emotion is correlated with reaction
two studies and the negative results of latter study may be due time on the LDT (Niedenthal & Setterlund 1994; Olafson &
to the different emotion induction methods. Whereas the third Ferraro 2001). We examined whether an emotion induction
study used self-referential statements to induce emotion, the using aversive pictures would influence direct (self-report)
first two studies used pictures, which may be a more reliable and indirect (LDT) measures of emotion state.
induction method (Göritz 2007; study 1).
The findings from Arch and Craske (2006) and Ortner et al.
(2007) suggest that mindful individuals exhibit less reactivity to Method
negative stimuli. To the extent that reactivity to particular
events contributes to chronic psychological distress (Mennin Participants
et al. 2005), affective reactivity may mediate the inverse rela-
tionship between mindfulness and chronic distress such as Forty (25 males) who were fluent in English participated for
anxiety. As noted above, one recent measure has proposed five partial fulfillment of a psychology course requirement. Mean
Mindfulness

participant age was 20.2 years (SD=2.7). Self-reported eth- seated subject). After providing informed consent and de-
nicity consisted of 34 Whites (non-Hispanic), four mographics, participants completed the baseline direct (self-
Asian/Pacific Islanders, one Black (non-Hispanic), and one report) and indirect (LDT) measures of emotion state.
Hispanic. Participants were then randomly assigned to view either
the control or the negative affect pictures. After this, they
Measures completed the post-induction direct and indirect measures of
emotion state. Participants were then debriefed and left the
Emotion Induction Task The emotion induction task lab.
consisted of a computer presentation of a series of 10 neutral
valenced slides (e.g., a ship) or 10 negative valenced slides
(e.g., a mangled and bloody hand) from the International Results
Affective Picture System (Lang et al. 2005; see Appendix A
for a list of the slides used in this task) which were presented Treatment of LDT Data Only correct trials (i.e., categorizing
for 10 s each. Participants were instructed to think about the negative affect and neutral words as “words” by pressing
how each picture made them feel. The emotion induction the correct response key) were included in the data analyses.
was presented on a personal computer with Inquisit software One participant demonstrated a high error rate (>20 %) and
(Draine 2004). was excluded from analyses (Teachman et al. 2001). The
mean error rate for the remaining subjects was 5.9 % (SD=
Direct Measure of Affective Reactivity The five negative 5.2) for the first LDT and 4.4 % (SD=3.5) for the second
affect items (afraid, ashamed, hostile, nervous, upset) from LDT. Trials with a reaction time of 3.0 SD above or below
the short-form Positive and Negative Affect Schedule the mean for each participant were classified as outliers and
(Thompson 2007) were used to assess state negative affect recoded to the sum of the mean ±3.0 SD. In order to reduce
at baseline and post-emotion induction. The scale consisted the positive skew found in most reaction time data, the
of a seven-point Likert scale ranging from 1 (not at all) to 7 response times were log transformed (Fazio 1990). The
(very much) and demonstrated adequate internal consistency LDT score was calculated as the mean reaction time for
both at baseline (coefficient α=0.71) and post-emotion in- the neutral targets minus the mean reaction time for the
duction (coefficient α=0.90). negative targets so that larger LDT scores indicate greater
accessibility of negative affect words.
Indirect Measure of Affective Reactivity Each participant
completed LDT to assess the accessibility of negative Influence of Emotion Induction on Direct and Indirect Mea-
affect-related words. The LDT was administered on a per- sures of Affective Reactivity The main hypothesis was that
sonal computer with Inquisit software (Draine 2004). The negative valenced pictures would increase negative affect as
LDT consisted of categorizing strings of letters as a “word” measured by self-report and the LDT. The influence of the
(e.g., scared) or “nonword” (e.g., triphy) as quickly and emotion induction was examined with between group (neu-
accurately as possible by pressing one of two response keys tral vs. negative slides) by within group (pre- and post-
with the left and right index fingers. A 500-ms fixation point emotion induction) repeated measures analyses. The results
preceded each letter string. The letter string stayed on the indicated a significant interaction for the self-report measure
screen until the correct response was made. Each response of negative affect (F(1, 37)=20.8, p=0.00005) and the LDT
was followed by a 1,000 ms intertrial interval. The stimuli (F(1, 37)=4.2, p=0.046). Repeated measures analyses dem-
consisted of 10 negative affect words taken from the Positive onstrated (a) that the negative slides increased self-reported
and Negative Affect Schedule (Watson et al. 1988), 10 neutral negative affect from baseline (M=1.9, SD=0.86) to post-
words, and 20 nonwords (see Appendix B for the stimuli). affect induction (M=3.2, SD=1.4), t(1, 18)=−4.1, p=0.001,
Participants completed a 10-trial practice block that used which represents a large effect size (Cohen’s d=1.1); (b) the
neutral and nonword stimuli before completing a 40-trial negative slides increased accessibility of negative affect
critical block in which each stimulus from all three categories words in the LDT, from baseline (M=−0.04, SD=0.05) to
was randomly presented once. The LDT was presented both post-affect induction (M=−0.01, SD=0.04), t(1, 19)=−3.1,
before and after the emotion induction task. The second LDT p=0.001, which represents a large effect size (Cohen’s d=
did not include a practice block. 0.8); (c) that the neutral slides did not change self-reported
negative affect from baseline (M=1.9, SD=0.8) to post-
Procedure affect induction (M=1.7, SD=0.8), t(1, 19)=1.8, p=0.09;
and (d) the neutral slides did not change the accessibility of
Participants were run in groups of up to five in semiprivate negative affect words in the LDT, from baseline (M=−0.02,
workstations (cubicles with a heavy drape door behind the SD=0.04) to post-affect induction (M=−0.02, SD=0.04),
 Mindfulness

t(1, 19)=−0.01, p=0.93. These results suggest that (a) the (Never or very rarely true) to 5 (Almost always or always
emotion induction task was able to elicit negative affect and true). Example items from the scales and their internal
that (b) the LDT can be used as an indirect measure of consistency (coefficient α) follow: Nonjudge (e.g., “I tell
affective reactivity. myself that I shouldn’t be feeling the way I’m feeling”)=0.
86, Act with Awareness (e.g., “I find it difficult to stay
focused on what’s happening in the present”) = 0.82,
Study 1 Describe (e.g., “I’m good at finding the words to describe
my feelings”)=0.85, Nonreact (e.g., “In difficult situations, I
Study 1 was designed to examine the role of affective can pause without immediately reacting”) = 0.75, and
reactivity measured by self-report as a potential mechanism Observe (e.g., “I pay attention to sensations, such as the
of mindfulness. We had the following three hypotheses: (1) wind in my hair or sun on my face”)=0.79.
an inverse relation between mindfulness (specifically, the
three subscales of the FFMQ that may be most relevant to Chronic Anxiety Anxiety over the past month was
emotion responding—Acting with Awareness, Nonjudging, assessed with the anxiety scale from the Brief Symptom
and Nonreactivity; Baer et al. 2006) and affective reactivity, Inventory (Derogatis & Melisaratos 1983). This scale
(2) a positive relation between affective reactivity and consists of six items (e.g., “feeling fearful”) designed to
chronic anxiety, and (3) an inverse relation between mind- assess the extent to which participants experienced anxi-
fulness and chronic anxiety would be partially mediated by ety over the previous month using a Likert scale ranging
affective reactivity. from 1 (Not at all) to 5 (Extremely). The scale demon-
strated adequate internal consistency (coefficient α=0.76)
in this sample.
Method
Procedure
Participants
The procedure was similar to the pilot study, with partici-
Eighty volunteers (42 males, 38 females) who were fluent in pants beginning by completing several baseline measures,
English participated for partial fulfillment of a psychology including measures of trait mindfulness and chronic anxiety.
course requirement. Mean participant age was 20.05 years Participants next rated their baseline negative emotion state.
(SD=2.43). Self-reported ethnicity consisted of 70 White After this, participants completed the negative emotion in-
(non-Hispanic), 3 Asian/Pacific Islander, 3 Black (non- duction, after which they completed the post-induction rat-
Hispanic), 2 American Indian/Alaskan Native, and 2 ing of negative emotion state.
reporting “Other”.

Measures Results

Negative Emotion Induction The negative emotion induc- Negative Emotion Induction A repeated measures analysis
tion consisted of the same task (presenting pictures) used in was performed to examine the extent to which the emotion
the pilot study. All participants viewed the negative pictures induction increased negative emotion. The results indicate
in this study. that negative emotion increased from the baseline (M=1.70,
SE=0.08) to post-induction (M=4.25, SE=0.21), F(1, 79)=
Direct Measure of Affective Reactivity A self-report assess- 133.52, p=1×10−18. This increase in negative affect repre-
ment was administered as a direct measure of emotion state. sents a large effect size (Cohen’s d=1.83).
The assessment consisted of a single item using a Likert
scale ranging from 1 (I do NOT feel negative) to 7 (I feel Mindfulness and Affective Reactivity The first hypothesis
VERY negative). Single-item measures of valence have dem- was that mindfulness would be inversely related to affective
onstrated good validity with other self-report measures of reactivity. In order to examine affective reactivity, correla-
affect (Bradley & Lang 1994) and have been used in other tion analyses were conducted between the mindfulness
emotion induction research (Colibazzi et al. 2010; Goldin & scales and post-induction negative emotion state. The results
Gross 2010; Larsen et al. 2009). indicate that the Nonjudge, Act aware, and Nonreact factors
predicted the post-induction self-report negative emotion
Mindfulness The FFMQ (Baer et al. 2006) was used to state, Nonjudge r= −0.28 (p=0.01), Act aware r=−0.30
assess individual differences in mindfulness. The FFMQ (p=0.006), and Nonreact r=−0.34 (p=0.002; a correlation
consists of 39 items that use a Likert-scale ranging from 1 matrix is reported in Table 1).
Mindfulness

Table 1 Zero-order correlations between mindfulness factors, past- predictor). A regression of anxiety on the Nonjudge scale
month anxiety, and affective reactivity
and post-induction negative affect indicated that the post-
Study 1 (N=80) Study 2 (N=68) induction negative affect rating was positively related to
anxiety, t(2, 77)=3.14, p=0.002 (for Act with Awareness,
Measure Anxiety Self-report Anxiety Lexical decision task t(2, 77)=3.09, p=0.003). Fourth, the relation between the
Affective Affective
Reactivity Reactivity predictor and criterion variables must be significantly reduced.
Using the same regression as for step 3, the results indicated
Nonjudge −0.50** −0.28* −0.38** −0.26* that although the relation between Nonjudge and anxiety
Act aware −0.46** −0.30** −0.53** −0.34** remained significant, t(2, 77)=−4.24, p=0.00006 (for Act
Nonreact −0.19 −0.34** −0.06 0.04 with Awareness, t(2, 77)=−3.63, p=0.001), a Sobel test indi-
Describe −0.17 −0.15 −0.13 −0.13 cated that the magnitude of the relation was significantly
Observe 0.29* −0.02 0.26* 0.05 reduced, z = −1.99, p = 0.046 (for Act with Awareness,
Anxiety – 0.42** – 0.52** z=−2.22, p=0.03), indicating a partial mediation of the rela-
tion between Nonjudging and anxiety and Act with
Anxiety anxiety experienced over the past month, Self-report Affective Awareness and anxiety.
Reactivity post-induction state negative affect, Lexical decision task
Affective Reactivity LDT score (mean reaction time for neutral words
minus mean reaction time for negative affect words such that larger
scores represents greater accessibility of negative affect words) Study 2
*p<0.05, **p<0.01
Study 2 was designed to replicate the findings of study 1, but
Chronic Anxiety and Affective Reactivity The second hypoth- with an indirect measure of affective reactivity. Thus, study 2
esis was that greater affective reactivity would be related to examined the role of affective reactivity measured by a LDT
greater chronic anxiety. A correlation analysis between anxiety as a potential mechanism of an inverse relation between
and post-induction negative affect indicated a positive relation mindfulness and chronic anxiety. We had the following three
between affective reactivity and anxiety, r=0.42 (p=0.0001). hypotheses: (1) an inverse relation between mindfulness and
affective reactivity (i.e., faster categorization of negative affect
Mediation of Chronic Anxiety The third hypothesis was that words relative to neutral words after viewing aversive pic-
that affective reactivity in response to an acute stressor would tures), (2) a positive relation between affective reactivity and
partially mediate the relation between trait mindfulness and chronic anxiety, and (3) an inverse relation between mindful-
anxiety. Using procedures outlined by Baron and Kenny ness and chronic anxiety would be partially mediated by
(1986) and significance tests of the effects (Sobel 1982), me- affective reactivity.
diation was considered to have occurred if the relation between
the predictor and criterion variables was partially or totally
accounted for by the hypothesized mediator(s) and if indirect Method
effects were observed (MacKinnon et al. 2007). Mediation
analyses were conducted with the Nonjudge and Act aware Participants
scales, as they were inversely correlated with anxiety.
Baron and Kenny (1986) proposed four steps to establish Seventy-one volunteers (42 males, 29 females) who were
mediation (these analyses are summarized in Fig. 1a). First, a fluent in English participated for partial fulfillment of a
relation between the predictor variable/s (mindfulness factors) psychology course requirement. Mean participant age was
and the criterion variable (chronic anxiety) must be 19.68 years (SD=3.32). Self-reported ethnicity consisted of
established. A regression analysis of anxiety on the 69 White (non-Hispanic) and 2 American Indian/Alaskan
Nonjudge scale indicated a relation between Nonjudge and Native.
anxiety, t(1, 78 = −5.06, p = 0.000003) (for the Act with
Awareness factor, t(1, 78)=−4.54, p=0.00002). Second, a Measures
relation between the predictor and hypothesized mediator
must be established. This relation was examined with a re- Negative Emotion Induction The negative emotion induc-
gression analysis of post-induction negative affect on the tion consisted of the same task (presentation of pictures)
Nonjudge scale. The results indicated a relation between used in study 1. As in study 1, all participants viewed the
Nonjudge and post-induction negative affect, t(1, 78) = negative pictures.
−2.59, p=0.01 (for Act with Awareness, t(1, 78)=−2.81,
p=0.006). Third, a relation between the mediator and criterion Indirect Measure of Affective Reactivity The LDT was ad-
variable must be established (while controlling for the ministered as an indirect measure of emotion state (i.e.,
 Mindfulness

Fig. 1 Model of relation (a) (b)

among mindfulness, anxiety,
and affective reactivity for Self-report Lexical decision
study 1 (a) and study 2 measure of task measure of
(b). Values presented are affective reactivity affective reactivity
standardized regression -0.28* 0.30** -0.26* 0.46**
coefficients. The value in
parentheses represents the (-0.50**) -0.41** (-0.38**) -0.27*
Nonjudge Anxiety Nonjudge Anxiety
coefficient for the direct
(unmediated) path. *p<0.05,
**p<0.01

Self-report Lexical decision

measure of task measure of
affective reactivity affective reactivity

-0.30** 0.31** -0.34** 0.39**

(-0.46**) -0.36** (-0.53**) -0.40**

Act aware Anxiety Act aware Anxiety

measuring accessibility of negative affect-related words participant were classified as outliers and recoded to the sum
compared to neutral words). It consisted of the same task of the mean ±3.0 SD. The reaction times were log
used in the pilot study with the change that each stimulus transformed to reduce the positive skew (Fazio 1990). The
was presented two times, increasing the number of critical LDT score was calculated as the mean reaction time for the
trials from 40 to 80. neutral affect targets minus the mean reaction time for the
negative targets so that larger LDT scores indicate greater
Mindfulness The FFMQ (Baer et al. 2006) was used to assess accessibility of negative affect words (i.e., greater affective
individual differences in mindfulness. Each factor demonstrat- reactivity).
ed good internal consistency (coefficient α) except for the
Nonreact factor: Nonjudge=0.86, Act with Awareness= Mindfulness and Affective Reactivity The first hypothesis
0.83, Describe=0.83, Nonreact=0.62 and Observe=0.82. was that mindfulness would be inversely related to affective
reactivity. In order to examine this, zero-order correlation
Chronic Anxiety Anxiety over the past month was assessed analyses were conducted between the mindfulness scales
with the anxiety scale from the Brief Symptom Inventory and the LDT score. The results indicate that the Nonjudging
(Derogatis & Melisaratos 1983). The scale demonstrated ade- and Act with Awareness scales predicted accessibility of neg-
quate internal consistency (coefficient α=0.74) in this sample. ative affect words, Nonjudging r=−0.26 (p=0.03) and Act
aware r=−0.34 (p=0.005; a correlation matrix is reported in
Procedure Table 1).

The procedure was similar to the pilot study and the first Chronic Anxiety and Affective Reactivity The second hy-
study, with participants beginning by completing several pothesis was that affective reactivity would be positively
baseline measures, including measures of trait mindfulness related to chronic anxiety. A correlation between anxiety
and chronic anxiety. Participants next completed the nega- and the LDT score indicated a relation between affective
tive emotion induction and shortly after completed the LDT. reactivity and anxiety, r=0.52 (p=0.000004).

Mediation of Chronic Anxiety The third hypothesis was that

Results that affective reactivity after an acute stressor would partial-
ly mediate the relation between trait mindfulness and anxi-
Treatment of LDT Data Only correct trials were included in ety (these analyses are summarized in Fig. 1b). Using
the data analyses. Three participants demonstrated a high regression analyses to demonstrate the steps from Baron
error rate (>20 %) and were excluded from analyses and Kenny (1986) that were listed in study 1, the first step
(Teachman et al. 2001). The mean error rate for the indicated a relation between the Nonjudge scale and anxiety,
remaining subjects was 6.11 % (SD=5.24). Trials with a t(1, 66)=−3.38, p=0.001 (for the Act with Awareness factor,
reaction time of 3.0 SD above or below the mean for each t(1, 66)=−5.08, p=0.000003). The second step indicated a
Mindfulness

relation between the Nonjudge scale and the LDT, t(1, 66)= the Baer et al. (2006) mindfulness scale. The results support
−2.17, p=0.03 (for Act with Awareness, t(1, 66)=−2.94, a definition of mindfulness as consisting of both an aware-
p=0.005). The third step indicated a relation between the ness factor and a nonjudgmental acceptance factor, as has
LDT and anxiety, controlling for the Nonjudge scale, t(2, been proposed by researchers (Bishop et al. 2004) and by
65)=4.37, p=0.00005 (for Act with Awareness, t(2, 65)=3. traditional practitioners (Gunaratana 2002). Contrary to our
86, p=0.0003). The fourth step indicated although the rela- hypotheses, the Nonreactivity to Inner Feelings factor was
tion between the Nonjudge scale and anxiety remained not inversely correlated with anxiety. One potential reason
significant when controlling for the LDT, t(2, 65)=−2.56, for this is that this scale demonstrated lower internal consis-
p = 0.01 (for Act with Awareness, t(2, 65) = −3.95, tency (particularly in study 2) than the Nonjudging and Act
p=0.0002), a Sobel test indicated that the magnitude of the with Awareness scales.
relation was significantly reduced, z=−1.92, p=0.05 (for The findings make a contribution by providing evidence
Act with Awareness, z=−2.22, p=0.03), indicating a partial that less affective reactivity partly explains previous find-
mediation of the relation between Nonjudging and anxiety ings of an inverse relation between psychological symptoms
and Act with Awareness and anxiety. and the Acting with Awareness and Nonjudging factors of
mindfulness (Baer et al. 2006). Future research would ben-
efit by examining how these mindfulness factors influence
Discussion affective reactivity. There are a number of intriguing possi-
bilities. For example, the Nonjudging factor involves
Although there is evidence for an inverse relation between allowing mental content (thoughts, impulses, and feelings)
mindfulness and psychological distress, little is known to manifest in the mind without evaluating them as good or
about the mechanisms through which this relation occurs. bad and without trying to change them. Previous research
Mindfulness teachers have suggested that one potential has demonstrated that negative (i.e., fear) appraisals of anx-
mechanism may be changes in affective responding to acute iety symptoms are involved in the development of anxiety
stressors (Kabat-Zinn 1990). This idea is supported by: (a) disorders (Taylor 1999). Thus, individuals high in
theoretical (Rottenberg & Gross 2003) and empirical Nonjudging may demonstrate less affective reactivity be-
(Mennin et al. 2005) perspectives that affective reactivity cause they do not appraise negative experience as being
contributes to psychological distress and (b) evidence that dangerous.
mindfulness is related to reduced affective reactivity (Arch Although the current findings suggest that affective reac-
& Craske 2006; Ortner et al. 2007). The findings from the tivity represents a mechanism through which mindfulness
current research indicate that mindfulness is related to less leads to less chronic anxiety, the affective reactivity variable
affective reactivity and that this variable partially mediated did not fully mediate the relation between mindfulness and
an inverse relation between mindfulness and anxiety expe- anxiety. It may be that other mechanisms are involved. For
rienced in daily life. example, several recent reviews have suggested that the
The current findings support and extend earlier work on beneficial effects of mindfulness may be partly explained
affective reactivity as a potential mechanism of mindfulness. through exposure (Brown et al. 2007; Shapiro et al. 2006).
Both studies 1 and 2 support previous findings that mind- Exposure interventions are enhanced when individuals at-
fulness is related to less affective reactivity toward negative tend to rather than distract from the feared stimulus
stimuli (Arch & Craske 2006; Ortner et al. 2007). The (Grayson et al. 1982). Because mindfulness involves contact
current work extends this research by demonstrating that with the present moment and allowing it to “just be”, no
less affective reactivity partially explains an inverse relation matter how unpleasant it is, fear responses may be more
between mindfulness and chronic anxiety. An additional likely to extinguish in exposure, leading to reduced anxiety.
contribution of the current research is the use of an indirect The way we assessed our mediators represents another
measure of affective state. Indirect (implicit) measures such possible reason we did not find full mediation between
as the LDT can help to overcome potential problems with mindfulness and anxiety. Specifically, we assessed core
direct measures, such as experimenter demand or subjects affect (valence) through self-report and the LDT rather than
being unwilling or unable to report their affect state (Isen et assessing indices of anxiety (with self-report, physiological
al. 2004). The use of indirect measures of affective measures, etc.). Although core affect is central to anxiety
responding such as the LDT in this research and skin con- and other discrete emotions (Russell & Barrett 1999), it may
ductance in Ortner et al. (2007) strengthens the ability to be that using anxiety-related mediators will provide a clearer
infer an inverse relation between mindfulness and affective picture of how mindfulness leads to less chronic anxiety.
reactivity. Assessing mindfulness with individual differences mea-
The results of both studies centered on the Acting with sures represents another study limitation. Several authors
Awareness and Nonjudging of Inner Experience factors of have suggested that because Westerners are relatively
 Mindfulness

inexperienced with the concept and experience of mindful- Appendix B

ness, questionnaires may not be valid assessments of the
construct (Davidson 2010; Grossman 2008; Rosch 2007). In Stimuli used in the lexical decision task
defense of self-report mindfulness measures, a number of Negative words: distressed, scared, upset, guilty, hostile,
studies have found that these measures predict psychologi- irritable, ashamed, nervous, jittery, afraid
cal distress even after covarying other personality traits Neutral words: bowl, chair, contents, door, errand, gla-
related to affective reactivity (e.g., neuroticism; Brown & cier, seat, theory, time, window
Ryan 2003), that they correlate with amount of meditation Nonwords: conclide, bos, fethom, fragmunt, luquid,
practice (Baer et al. 2006), that they change after meditation meisure, parkying, tiste, wuekly, zune, tasp, botter, datest,
training (Brown & Ryan 2003), and that the influence of flibby, hestage, lohe, pilmow, sikkness, triphy, violeat
meditation training on psychological distress is mediated by
self-report mindfulness (Carmody & Baer 2008; Shapiro et
al. 2008). However, future research would benefit from
examining whether mindfulness training leads to similar References
results as found in the current research. Positive findings
would add confidence to the idea that affective reactivity is Arch, J. J., & Craske, M. G. (2006). Mechanisms of mindfulness:
an important mechanism of mindfulness and would have emotion regulation following a focused breathing induction. Be-
implications for clinical interventions. haviour Research and Therapy, 44, 1849–1858.
Baer, R. A. (2003). Mindfulness training as a clinical intervention: a
An analytical limitation of the study consists of the in- conceptual and empirical review. Clinical Psychology: Science
ferences that are allowed by mediation analyses using a and Practice, 10, 125–143.
cross-sectional design. The significant results from the anal- Baer, R. A., Smith, G. T., Hopkins, J., Krietemeyer, J., & Toney, L.
yses indicate that the affective reactivity variable is a medi- (2006). Using self-report assessment methods to explore facets of
mindfulness. Assessment, 13, 27–45.
ator in that it conceptually and statistically “accounts for the Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator vari-
relation between the predictor and the criterion” variables able distinction in social psychological research: conceptual, stra-
(Baron & Kenny 1986, p.1176). However, cross-sectional tegic, and statistical considerations. Journal of Personality and
mediation does not allow for causal inferences because the Social Psychology, 51, 1173–1182.
Bishop, S. R., Lau, M., Shapiro, S., Carlson, L., Anderson, N. D.,
direction of the relationships cannot be determined. From Carmody, J., et al. (2004). Mindfulness: a proposed operational
this perspective, the current results from the mediation anal- definition. Clinical Psychology: Science and Practice, 11, 230–
yses should be interpreted as preliminary and should be 241.
examined with experimental designs in future research. Bradley, M. M., & Lang, P. J. (1994). Measuring emotion: the self-
assessment manikin and the semantic differential. Journal of
In sum, the current research contributes to developing an Behavior Therapy and Experimental Psychiatry, 25, 49–59.
understanding of the mechanisms through which mindful- Brown, K. W., & Ryan, R. M. (2003). The benefits of being present:
ness exerts its beneficial effects. Previous work has found mindfulness and its role in psychological well-being. Journal of
that mindfulness is related to affective reactivity (Arch & Personality and Social Psychology, 84, 822–848.
Brown, K. W., Ryan, R. R., & Creswell, J. D. (2007). Mindfulness:
Craske 2006; Ortner et al. 2007), providing initial evidence theoretical foundations and evidence for its salutary effects. Psy-
that this variable may mediate the effects of mindfulness on chological Inquiry, 18, 211–237.
psychological distress. The current work extends this re- Carmody, J., & Baer, R. A. (2008). Relationships between mindfulness
search by providing evidence that the proposed mediator practice and levels of mindfulness, medical and psychological
symptoms and well-being in a mindfulness-based stress reduction
of affective reactivity is related to the criterion variable of program. Journal of Behavioral Medicine, 31, 23–33.
psychological distress, which is a condition in establishing Colibazzi, T., Posner, J., Wang, Z., Gorman, D., Gerber, A., Yu, S., et
mediation (Baron & Kenny 1986). The current research also al. (2010). Neural systems subserving valence and arousal during
extends previous work by demonstrating that the variance the experience of induced emotions. Emotion, 10, 377–389.
Davidson, R. J. (1998). Affective style and affective disorders: per-
shared by mindfulness and chronic anxiety is partially spectives from affective neuroscience. Cognition and Emotion,
accounted for by shared variance with affective reactivity. 12, 307–330.
Davidson, R. J. (2010). Empirical explorations of mindfulness: con-
ceptual and methodological conundrums. Emotion, 10, 8–11.
Derogatis, L. R., & Melisaratos, N. (1983). The brief symptom inven-
Appendix A tory: an introductory report. Psychological Medicine, 13, 595–
605.
IAPS slides used in affect induction task. Draine, S.C. (2004). Inquisit 2.0.50401 [Computer software]. Seattle,
Negative slides: 2800, 3150, 3170, 3230, 3350, 9040, WA: Millisecond Software
Fazio, R. H. (1990). A practical guide to the use of response latency in
9300, 9405, 9561, 9594 social psychological research. In C. Hendrick & M. S. Clark
Neutral slides: 2191, 2214, 2374, 5395, 7000, 7043, (Eds.), Review of personality and social psychology (Vol. 11,
7052, 7053, 7207, 7249 pp. 74–97). Newbury Park, CA: Sage.
Mindfulness

Goldin, P. R., & Gross, J. J. (2010). Effects of mindfulness-based stress Ostafin, B. D., Chawla, N., Bowen, S., Dillworth, T. M., Witkiewitz,
reduction (MBSR) on emotion regulation in social anxiety disor- K., & Marlatt, G. A. (2006). Intensive mindfulness training and
der. Emotion, 10, 83–91. the reduction of psychological distress: a preliminary study. Cog-
Göritz, A. S. (2007). The induction of mood via the www. Motivation nitive and Behavioral Practice, 13, 191–197.
and Emotion, 31, 35–47. Rosch, E. (2007). More than mindfulness: when you have a tiger by the
Grayson, J. B., Foa, E. B., & Steketee, G. (1982). Habituation during tail, let it eat you. Psychological Inquiry, 18, 258–264.
exposure treatment: distraction versus attention-focusing. Behav- Rottenberg, J., & Gross, J. J. (2003). When emotion goes wrong:
iour Research and Therapy, 20, 323–328. realizing the promise of affective science. Clinical Psychology:
Grossman, P. (2008). On measuring mindfulness in psychosomatic and Science and Practice, 10, 227–232.
psychological research. Journal of Psychosomatic Research, 64, Russell, J. A., & Barrett, L. F. (1999). Core affect, prototypical emotional
405–408. episodes and other things called emotion: dissecting the ele-
Gunaratana, B. H. (2002). Mindfulness in plain English. Somerville, phant. Journal of Personality and Social Psychology, 76, 805–
MA: Wisdom. 819.
Isen, A. M., Labroo, A. A., & Durlach, P. (2004). An influence of Shapiro, S. L., Carlson, L. E., Astin, J. A., & Freedman, B. (2006).
product and brand name on positive affect: implicit and explicit Mechanisms of mindfulness. Journal of Clinical Psychology, 62,
measures. Motivation and Emotion, 28, 43–63. 373–386.
Kabat-Zinn, J. (1990). Full catastrophe living: using the wisdom of your Shapiro, S. L., Oman, D., Thoresen, C. E., Plante, T. G., & Flinders, T.
body and mind to face stress, pain, and illness. New York: Dell. (2008). Cultivating mindfulness: effects on well-being. Journal of
Koole, S. L. (2009). The psychology of emotion regulation: an inte- Clinical Psychology, 64, 840–862.
grative review. Cognition and Emotion, 23, 4–41. Sobel, M. E. (1982). Asymptotic confidence intervals for indirect effects
Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (2005). International in structural equation models. Sociological Methodology, 13, 290–
affective picture system (IAPS): Instruction manual and affective 312.
ratings. Gainesville, FL: Technical Report A-6. University of Soussignan, R., Boivin, M., Girard, A., Perusse, D., Liu, X., & Tremblay,
Florida. R. E. (2009). Genetic and environmental etiology of emotional and
Larsen, J. T., Norris, C. J., McGraw, A. P., Hawkley, L. C., & Cacioppo, J. social behaviors in 5-month-old infant twins: influence of the social
T. (2009). The evaluative space grid: a single-item measure of context. Infant Behavior & Development, 32, 1–9.
positivity and negativity. Cognition and Emotion, 23, 453–180. Taylor, S. (1999). Anxiety sensitivity: theory, research, and treatment of
MacKinnon, D. P., Fairchild, A. J., & Fritz, M. S. (2007). Mediation the fear of anxiety. Mahwah, NJ: Erlbaum.
analysis. Annual Review of Psychology, 58, 593–614. Teachman, B. A., Gregg, A. P., & Woody, S. R. (2001). Implicit associ-
Mennin, D. S., Heimberg, R. G., Turk, C. L., & Fresco, D. M. (2005). ations for fear-relevant stimuli among individuals with snake and
Preliminary evidence for an emotion dysregulation model of spider fears. Journal of Abnormal Psychology, 110, 226–235.
generalized anxiety disorder. Behaviour Research and Therapy, Thera, T. (1965). The heart of Buddhist meditation. York Beach, ME:
43, 1281–1310. Weiser.
Niedenthal, P. M., & Setterlund, M. B. (1994). Emotion congruence in Thompson, E. R. (2007). Development and validation of an interna-
perception. Personality and Social Psychology Bulletin, 20, 401– tionally reliable short-form of the Positive and Negative Affect
411. Schedule (PANAS). Journal of Cross-Cultural Psychology, 38,
Olafson, K. M., & Ferraro, F. R. (2001). Effects of emotional state on 227–241.
lexical decision performance. Brain and Cognition, 45, 15–20. Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and
Ortner, C. N. M., Kilner, S. J., & Zelazo, P. D. (2007). Mindfulness validation of brief measures of positive and negative affect: the
meditation and reduced emotional interference on a cognitive PANAS scales. Journal of Personality and Social Psychology, 54,
task. Motivation and Emotion, 31, 271–283. 1063–1070.

View publication stats