INVESTIGACIÓN
Distinct patterns of thought mediate the
link between brain functional connectomes
and well-being
Deniz Vatansever
1,2, Theodoros Karapanagiotidis2, Daniel S.. Margulies3,
Elizabeth Jefferies2, and Jonathan Smallwood2
1Institute of Science and Technology for Brain-inspired Intelligence, Fudan University, Shanghai, Porcelana
2Department of Psychology, University of York, york, Reino Unido
3Brain and Spine Institute, French National Centre for Scientific Research, París, Francia
un acceso abierto
diario
Palabras clave: Conectividad funcional, Functional magnetic resonance imaging, Graph theory, Mental
salud, Resting state, Patterns of thought
ABSTRACTO
Ongoing thought patterns constitute important aspects of both healthy and abnormal human
cognition. Sin embargo, the neural mechanisms behind these daily experiences and their
contribution to well-being remain a matter of debate. Aquí, using resting-state fMRI and
retrospective thought sampling in a large neurotypical cohort (norte = 211), we identified two
distinct patterns of thought, broadly describing the participants’ current concerns and future
planes, that significantly explained variability in the individual functional connectomes.
Consistent with the view that ongoing thoughts are an emergent property of multiple neural
sistemas, network-based analysis highlighted the central importance of both unimodal and
transmodal cortices in the generation of these experiences. En tono rimbombante, while state-dependent
current concerns predicted better psychological health, mediating the effect of functional
connectomes, trait-level future plans were related to better social health, yet with no
mediatory influence. Colectivamente, we show that ongoing thoughts can influence the link
between brain physiology and well-being.
RESUMEN DEL AUTOR
Occupying a considerable portion of our waking lives, spontaneous thoughts constitute the
foundations of our rich inner mental experiences and well-being. Sin embargo, the neural
mechanisms behind this cognitive process and its relation to our mental health remain
unresolved. In a large cohort of participants, we show that distinct dimensions of ongoing
thoughts emerge from a broad set of whole-brain functional interactions, with certain
patterns significantly mediating the link between brain connectivity and psychosocial health.
En general, these results highlight the heterogeneous nature of self-generated thoughts, el
content and form of which have a significant influence on the association between our brain
physiology and mental well-being.
INTRODUCCIÓN
Recent advances in functional magnetic resonance imaging (resonancia magnética funcional) methods and data anal-
ysis techniques have facilitated a new era in the characterization of the neural representa-
tions that underlie human cognition and behavior (Respeto, 2013). Big data-driven approaches
are utilized to explore whole-brain functional interactions during unconstrained states of rest,
Citación: Vatansever, D., Karapanagiotidis,
T., Margulies, D. S., Jefferies, MI., &
Smallwood, j. (2020). Distinct patterns
of thought mediate the link between
brain functional connectomes and
well-being. Neurociencia en red,
4(3), 637–657. https://doi.org/10.1162/
netn_a_00137
DOI:
https://doi.org/10.1162/netn_a_00137
Supporting Information:
https://doi.org/10.1162/netn_a_00137
Recibió: 18 Octubre 2019
Aceptado: 4 Marzo 2020
Conflicto de intereses: Los autores tienen
declaró que no hay intereses en competencia
existir.
Autor correspondiente:
Deniz Vatansever
deniz@fudan.edu.cn
Editor de manejo:
Olaf Sporns
Derechos de autor: © 2020
Instituto de Tecnología de Massachusetts
Publicado bajo Creative Commons
Atribución 4.0 Internacional
(CC POR 4.0) licencia
La prensa del MIT
yo
D
oh
w
norte
oh
a
d
mi
d
F
r
oh
metro
h
t
t
pag
:
/
/
d
i
r
mi
C
t
.
metro
i
t
.
/
/
t
mi
d
tu
norte
mi
norte
a
r
t
i
C
mi
–
pag
d
yo
F
/
/
/
/
/
4
3
6
3
7
1
8
6
7
2
6
4
norte
mi
norte
_
a
_
0
0
1
3
7
pag
d
t
.
F
b
y
gramo
tu
mi
s
t
t
oh
norte
0
7
S
mi
pag
mi
metro
b
mi
r
2
0
2
3
Linking thoughts, cerebro, and well-being
explaining considerable levels of population-wise variation in complex traits, including intel-
ligence (verde, gao, Scheinost, & Constable, 2018), personality (Toschi, Riccelli, Indovina,
Terracciano, & Passamonti, 2018), daily habits (Cheng et al., 2019), and self-perceived quality
de la vida (Kraft et al., 2018). In addition to helping researchers derive novel theories on healthy
brain processing, one important goal of these functional connectomic mapping initiatives is
to devise neural, cognitivo, and behavioral links to better understand mental health disor-
ders, and to identify “at-risk” groups for future preventative measures (Castellanos, Di Martino,
Craddock, Mehta, & Milham, 2013; VanEssen & Respeto, 2015). Sin embargo, most studies often
neglect the fact that periods of unconstrained states of rest can be characterized by patterns of
spontaneous thought that are also associated with well-being (Andrews-Hanna et al., 2013),
which may have unique neurocognitive correlates.
Research from psychology suggests that the ability to self-generate patterns of cognition is
a core element of our mental lives, occupying a considerable portion of our daily mentation
(Antrobus, Cantante, & Greenberg, 1966; Klinger, 1971; Cantante & Antrobus, 1963). críticamente,
these thoughts are particularly prevalent in situations with low external demands, como
periods of wakeful rest (Smallwood, Nind, & O’Connor, 2009), eso es, the conditions when
resting-state functional data are most commonly recorded. En tono rimbombante, measures of such ex-
periences have a wide range of momentary correlates including indicators of stress (Engert,
Smallwood, & Cantante, 2014), ongoing physiology (Konishi, Marrón, Battaglini, & Smallwood,
2017), and task performance (Smallwood, Beach, Schooler, & Handy, 2008), and also have
documented links to both beneficial and deleterious aspects of psychological functioning. Para
ejemplo, patterns of ongoing thought have been previously linked to individuals’ ability to plan
for future goals (Baird, Smallwood, & Schooler, 2011), wait for long-term rewards (Smallwood,
Ruby, & Cantante, 2013), and devise creative solutions to both personal (Baird et al., 2012) y
social problems (Ruby, Smallwood, Sackur, et al., 2013). Other forms of self-generated men-
tation are associated with unhappiness (Killingsworth & Gilbert, 2010), as well as poor perfor-
mance in sustained attention tasks (Allan Cheyne, Solman, Carriere, & Smilek, 2009)
or measures of fluid intelligence (Mrazek et al., 2012).
De hecho, disruptive thought patterns
are reported to underlie the absentminded mistakes in our everyday functioning (Carriere,
Cheyne, & Smilek, 2008; McVay & kane, 2009), including traffic accidents (Galera et al.,
2012) or medical malpractice (Smallwood, Mrazek, & Schooler, 2011), and to form a poten-
tial basis for cognitive impairments reported in attention-deficit/hyperactivity disorder (Seli,
Smallwood, Cheyne, & Smilek, 2015; Vatansever, Bozhilova, Asherson, & Smallwood, 2018)
and clinical depression (Marchetti, Van de Putte, & Koster, 2014). Tomados juntos, this body of
evidence underlines the vital importance of both stable and transient thought patterns in our
daily mentation and their variable influence on our psychological and social well-being.
Además, there is now emerging evidence that links distinct profiles of neural activity
to trait variance in specific patterns of thoughts. These investigations highlight that patterns of
ongoing thought have complex and often heterogeneous neural correlates (Smallwood et al.,
2016; Stawarczyk, Majerus, Comandante, Van der Linden, & D’Argembeau, 2011), which depend on
the functional interaction of multiple neural and cognitive components (Gorgolewski et al.,
2014; Poerio et al., 2017; Wang, Poerio, et al., 2018). Por ejemplo, Wang and colleagues
identified distinct patterns of ongoing thoughts that were linked to reductions in within- y
between-network connectivity for neural systems important for external attention (ventral, dorsal,
and frontoparietal systems), which was in turn related to reduced performance in tasks of cog-
nitive aptitude (Wang, Bzdok, et al., 2018). Other studies have illustrated that different types
of spontaneous thought depend on differential patterns of functional connectivity between
regions important for memory (p.ej., temporal lobe) and those that form the core associative
638
Well-being:
The holistic experience of a state
characterized by health and
happiness.
Conectividad funcional:
Statistical dependencies among
time series obtained from
neurophysiological data.
Neurociencia en red
yo
D
oh
w
norte
oh
a
d
mi
d
F
r
oh
metro
h
t
t
pag
:
/
/
d
i
r
mi
C
t
.
metro
i
t
.
/
/
t
mi
d
tu
norte
mi
norte
a
r
t
i
C
mi
–
pag
d
yo
F
/
/
/
/
/
4
3
6
3
7
1
8
6
7
2
6
4
norte
mi
norte
_
a
_
0
0
1
3
7
pag
d
t
.
F
b
y
gramo
tu
mi
s
t
t
oh
norte
0
7
S
mi
pag
mi
metro
b
mi
r
2
0
2
3
Linking thoughts, cerebro, and well-being
cortices (Golchert et al., 2017; Karapanagiotidis, Bernhardt, Jefferies, & Smallwood, 2017;
Smallwood et al., 2016). Notablemente, tasks such as creative idea generation (Beaty, Benedek,
Kaufman, & Silvia, 2015) and future planning (Spreng, stevens, Chamberlain, Gilmore, &
Schacter, 2010), which are collectively associated with unconstrained thought (Baird et al.,
2012; Medea et al., 2016), depend on similar patterns of interaction between regions of the
default mode and frontoparietal networks.
Converging bodies of evidence from contemporary cognitive neuroscience, por lo tanto, alto-
luz (a) that patterns of neural activity at rest have associated patterns of thought, y (b) eso
both neural and self-report descriptions of patterns of unconstrained activity are predictive of
a wide range of psychological features of an individual, including mental well-being. Collec-
activamente, these observations highlight the need to understand the extent to which relationships
between neural activity at rest and well-being are dependent upon the nature of patterns of
ongoing experience that emerges while neural activity is recorded. Our current study reflects
an attempt to address this issue by quantifying the relationship between patterns of thoughts
during rest and the associated neural activity, and then exploring whether descriptions of neu-
ral organization gained in this fashion are associated with trait variance in a cross-culturally
validated questionnaire that assessed both psychological and social well-being.
yo
D
oh
w
norte
oh
a
d
mi
d
F
r
oh
metro
h
t
t
pag
:
/
/
d
i
r
mi
C
t
.
metro
i
t
.
/
t
/
mi
d
tu
norte
mi
norte
a
r
t
i
C
mi
–
pag
d
yo
F
/
/
/
/
/
4
3
6
3
7
1
8
6
7
2
6
4
norte
mi
norte
_
a
_
0
0
1
3
7
pag
d
.
t
F
b
y
gramo
tu
mi
s
t
t
oh
norte
0
7
S
mi
pag
mi
metro
b
mi
r
2
0
2
3
Mediation:
A statistical method to assess the
mediatory influence of an external
variable on the relationship between
independent and dependent
variables.
In the analysis of a large neurotypical sample (norte = 211), we found differential profiles
of functional interactions among multiple neural systems that explained individual variation
among two distinct patterns of thought, broadly corresponding to the participants’ current con-
cerns and future plans. Using a test-retest sample (norte = 40), we found reasonable concordance
between the tendency to generate future plans at rest and their brain basis, Indicando que
these measures are likely to reflect neurocognitive traits. Por otro lado, current concerns
were not stable across individuals, but showed evidence of common changes in terms of the
pattern of thought and pattern of functional connectivity, suggesting that these neurocognitive
links reflect a more transient state. Finalmente, we found that while current concerns indirectly
mediated the effect of brain functional interactions on psychological well-being, future plans
showed no such mediation effect.
RESULTADOS
Dimensions of Variation in Ongoing Thought Patterns
With the aim of investigating the differential influence of distinct thought patterns on the link
between brain functional network topology and mental well-being, we collected 9 min of fMRI
data from a large cohort of neurotypical participants (norte = 211) during a period of wakeful rest.
This neural measure was complemented with a session of retrospective thought sampling, anuncio-
ministered immediately after the resting-state scanning, as well as self-assessed ratings of psy-
chological and social well-being on a cross-culturally validated questionnaire from the World
Health Organization Quality of Life (WHOQOL) grupo (WHOQOL Group, 1998), collected
outside the scanner in a separate behavioral session. The workflow for the data collection and
analysis techniques utilized in this study is presented in Figure 1, with further details provided
in the Methods section and the Supporting Information. Parts of this dataset have been pre-
viously employed in our prior investigations (Vatansever, Bozhilova, et al., 2018; Vatansever,
Bzdok, et al., 2017; Wang, Bzdok, et al., 2018; Wang, Poerio, et al., 2018).
Our first objective was to establish the dimensions of variation within the participants’ on-
going thought patterns. The employed retrospective thought sampling method required par-
ticipants to subjectively characterize their thoughts using 4-choice (escala Likert) ratings on a
Neurociencia en red
639
Linking thoughts, cerebro, and well-being
yo
D
oh
w
norte
oh
a
d
mi
d
F
r
oh
metro
h
t
t
pag
:
/
/
d
i
r
mi
C
t
.
metro
i
t
.
/
t
/
mi
d
tu
norte
mi
norte
a
r
t
i
C
mi
–
pag
d
yo
F
/
/
/
/
/
4
3
6
3
7
1
8
6
7
2
6
4
norte
mi
norte
_
a
_
0
0
1
3
7
pag
d
.
t
F
b
y
gramo
tu
mi
s
t
t
oh
norte
0
7
S
mi
pag
mi
metro
b
mi
r
2
0
2
3
Cifra 1. Experimental data collection and analysis pipeline. A large cohort of participants (norte =
211) were fMRI scanned during a state of wakeful rest for a total of 9 mín.. A comprehensive pipeline
of fMRI preprocessing and data denoising procedures were followed in order to ensure maximal
removal of nuisance variables. The thought sampling scores collected at the end of the resting-state
scanning session were first hierarchically clustered into two major groups, each of which was then
reduced to three patterns of thought using principal component analysis. Fully connected, weighted
correlation matrices based on the Power et al. (2011) parcellation (264 regiones de interés) scheme
and the individual component scores on the identified patterns of thought were then carried forward
on to network-based statistics (NBS) with the aim of identifying components of brain connections
that related to the participants’ thought patterns. For all six patterns of thought, t tests were carried
out with an initial T score of 3.2 and a significance level of p < 0.05 over 5,000 permutations
(mean connectivity, age, gender, and percentage of invalid scans based on the composite motion
score from the scrubbing procedure were entered as group-level nuisance regressors). The identified
brain components that significantly related to the participants’ thoughts were first characterized
at the group level and then used as mask graphs to create thresholded connectivity matrices for
each participant. Network metrics of positive, negative, total, and fractional strength (i.e., the ratio
of positive to negative strength) as well as betweenness centrality were measured on individual
thresholded matrices and were further used to characterize the identified neurocognitive profiles.
Linear regressions and mediation analyses were then employed to investigate the mediatory effect
of thought patterns on the link between brain connectivity and psychological and social well-being
as measured using a cross-culturally validated World Health Organization Quality of Life group
(WHOQOL-BREF) questionnaire.
set of questions derived from our previous investigations (Gorgolewski et al., 2014; Medea
et al., 2016; Ruby, Smallwood, Engen, & Singer, 2013; Ruby, Smallwood, Sackur, et al., 2013;
Smallwood et al., 2016; Supplementary Table S1). With the aim of reducing dimensionality
and improving interpretability, these ratings were first hierarchically clustered and then decom-
posed into distinct dimensions using principal component analysis (PCA), which established
the main patterns of thought reported in our sample (Figure 2). The total number of thought pat-
terns was selected using scree plots, based on the eigenvalue (>1) and the explanatory power
gained by each additional decomposition (Supplementary Figures S1 and S2). The robustness
assessments and typical ratings from participants who scored highest on a given thought pattern
are provided in Supplementary Figures S3–S5.
Network-based statistic:
An open-source toolbox to
statistically assess hypotheses related
to the human connectome.
Principal component analysis:
A method to reduce dimensionality
of the variable space into a smaller
number of orthogonal variables that
maximize explained variance.
Neurociencia en red
640
Linking thoughts, cerebro, and well-being
Hierarchical clustering:
A clustering algorithm aiming to form
a topological hierarchy among
clusters of observations.
Cifra 2. Decomposing distinct patterns of thought. Following the initial hierarchical clustering of
the participants’ ratings on the experience sampling questionnaire, a total of six patterns of thought
were identified using PCA. Three principal components in each of the two clusters explained 51%
y 35% of the variability in the data, respectivamente. The Varimax rotated component loadings are
visualized using word clouds. While the size of the text refers to the relative strength of the com-
ponent loadings, positive and negative loadings are indicated via red and blue fonts, respectivamente.
The components highlighted (A) important/specific thoughts, (B) perceptually decoupled/hard-to-
stop thoughts, (C) positive/spontaneous thoughts, (D) insightful/image-based thoughts, (mi) deliberate/
verbal thoughts, y (F) negative/past-related thoughts. The individual variation on these patterns
of thought were carried forward on to the NBS analysis as between-subject explanatory variables.
Functional connectome:
Wiring diagram of a comprehensive
set of functional connections in the
cerebro.
Principal component analysis, applied separately to each of the two initial hierarchical
clusters on the self-reported ratings, revealed three patterns of thought each (a total of six),
explicando 51% y 35% of the variance, respectivamente. The identified patterns highlighted
aspects of ongoing thoughts that are commonly characterized in the existing literature
(Smallwood & Schooler, 2015). These encompassed thoughts that were important and specific
(Figura 2A); perceptually decoupled and hard-to-stop thoughts about the self (Figura 2B); posición-
itive, spontaneous thoughts about others (Figura 2C); insightful and image-based thoughts
(Figura 2D); deliberate, verbal thoughts about the future (Figura 2E); and negative, past-related
thoughts (Figura 2F).
Distinct Patterns of Thought Link to Differential Functional Connectomic Profiles
Próximo, we determined the extent to which the human functional connectomes were associ-
ated with individual variation on the identified patterns of thought. We used a whole-brain
parcellation scheme (Power et al., 2011). Defining each of the 264 brain regions as nodes,
and Pearson correlation coefficients among them as edges, we constructed fully connected,
weighted functional connectomes for each participant. Utilizing network-based statistics (NBS;
Brilla, Proporcionó, & bullmore, 2010), we entered the individual component scores for all six
patterns of thought as variables of interest in a regression model, while removing the effects
of nuisance variables including mean connectivity, edad, género, and the composite motion
puntaje (es decir., percentage of invalid volumes identified in the motion artifact detection procedure;
Supplementary Figures S6–S7). This revealed two distinct patterns of thought that were signifi-
cantly related to differential profiles of brain functional connectomic organization (Tthreshold =
3.2, 5,000 permutations, un = 0.05). Although we report on the initial Tthreshold = 3.2 límite,
comparable results for Tthreshold = 3.1 and Tthreshold = 3.3 are presented in the Supporting
Information section (Supplementary Figure S9).
Neurociencia en red
641
yo
D
oh
w
norte
oh
a
d
mi
d
F
r
oh
metro
h
t
t
pag
:
/
/
d
i
r
mi
C
t
.
metro
i
t
.
t
/
/
mi
d
tu
norte
mi
norte
a
r
t
i
C
mi
–
pag
d
yo
F
/
/
/
/
/
4
3
6
3
7
1
8
6
7
2
6
4
norte
mi
norte
_
a
_
0
0
1
3
7
pag
d
t
.
F
b
y
gramo
tu
mi
s
t
t
oh
norte
0
7
S
mi
pag
mi
metro
b
mi
r
2
0
2
3
Linking thoughts, cerebro, and well-being
The first thought pattern that was significantly associated with neural patterns reflected high
loadings on “habitual,” “realistic,” “specific,” and “important” elements of ongoing cognition,
broadly corresponding to the construct of current concerns that has been argued to play an
important motivational role in the content and form of ongoing thoughts (Klinger, 2013). Popu-
lation variation in this experience was linked to a combination of positive and negative connec-
tions from a wide range of both unimodal and transmodal brain regions (Figura 3A) (Tthreshold =
3.2, 5,000 permutations, pag = 0.015). These included areas associated with visual, auditory,
somato-motor, dorsal/ventral attention, cingulo-opercular, prominencia, frontoparietal, default mode,
yo
D
oh
w
norte
oh
a
d
mi
d
F
r
oh
metro
h
t
t
pag
:
/
/
d
i
r
mi
C
t
.
metro
i
t
.
/
t
/
mi
d
tu
norte
mi
norte
a
r
t
i
C
mi
–
pag
d
yo
F
/
/
/
/
/
4
3
6
3
7
1
8
6
7
2
6
4
norte
mi
norte
_
a
_
0
0
1
3
7
pag
d
.
t
F
b
y
gramo
tu
mi
s
t
t
oh
norte
0
7
S
mi
pag
mi
metro
b
mi
r
2
0
2
3
Cifra 3. Distinct patterns of thought are linked to differential profiles of functional connectomic
organización. NBS identified two connected components of brain functional interactions at rest that
significantly related to the participants’ scores on distinct patterns of thought, which highlighted (A)
important, específico, realistic, and habitual thoughts about the self and others that were not spon-
taneous; así como (B) deliberate, verbal, thematic, abstract, problem-based, and future-oriented
thoughts in words that were not about the past or in images. The average connectivity patterns of
these brain graphs are visualized on an MNI152 smoothed glass brain, with the nodes color-coded
according to the original Power et al. (2011) parcellation scheme, and the positive/negative connec-
tions colored in red and blue, respectivamente, in which the size of edges represents average strength.
The average connectivity patterns of these brain graphs and the average graph theory metrics cal-
culated across participants are visualized in circular representations. The Automated Anatomical
Labeling (AAL) nomenclature, original network parcellation, average positive (rojo: [0–1] Pearson r
escala), negative (azul: [0–1] Pearson r scale), total (purple: [0–2] Pearson r scale), and fractional
fortalezas (naranja: [0–10] arbitrary scale) as well as betweenness centrality (verde: [0–6] arbitrary
escala) of each region in the identified brain components are listed around the rings. el promedio
positive and negative connections between these regions are represented by links color-coded red
and blue, respectivamente ([0–1] Pearson r scale).
Neurociencia en red
642
Linking thoughts, cerebro, and well-being
Graph theory:
A field of mathematics quantifying
the topological properties of graphs
formed by objects (nodos) and their
pairwise relationships (bordes).
subcortical networks, and a set of regions with memory-related functions, with the strongest
link observed between the right middle frontal and middle cingulate gyri (salience network).
En general, this neural component was composed of more positive than negative connections
and involved a set of distributed brain regions that encompassed a diverse set of large-scale
redes cerebrales. Characterization of this brain component using graph theory measures re-
vealed that an area in the left middle frontal gyrus, belonging to the dorsal attention network,
showed the highest positive, negative, and total strength as well as betweenness centrality. Este
highlights the left middle frontal gyrus as a key node in the functional connectivity patterns
linked to ongoing thoughts about the individuals’ current concerns. Además, the highest
fractional strength (es decir., the ratio of positive to negative strength) was observed in a region
belonging to the salience network, namely the right middle cingulate gyrus.
The second pattern of thought that related to neural function reflected high loadings on
the “future” rather than the “past,” as well as “deliberate” and “verbal” focus on “problems,"
which is broadly consistent with studies of ongoing thought that emphasize a deliberate fo-
cus on future plans (Baird et al., 2011; Seli, Ralph, Konishi, Smilek, & Schacter, 2017). Este
thought pattern was related to the functional interaction of brain regions from a small set of
large-scale brain networks mainly belonging to the higher order transmodal cortices (Tthreshold =
3.2, 5,000 permutations, pag = 0.026; Figura 3B). Both positive and negative links between re-
gions belonging to the default mode, frontoparietal, prominencia, cingulo-opercular, and visual
regions correlated with higher scores reported on this pattern of thought, with the strongest
link observed between regions in the right inferior temporal and middle frontal gyri. Relativo
to the “current concerns” component, this brain component was composed of more nega-
tive connections and was drawn from a more localized set of brain regions from a smaller
number of large-scale brain networks. Network-level analysis indicated that a left superior
frontal gyrus region belonging to the default mode network showed the highest positive, neg-
ative, and total strength, while a left middle cingulate region (salience network) displayed the
greatest fractional strength. Betweenness centrality, sin embargo, was highest on a right inferior
temporal gyrus region.
Test-Retest Reliability of Thought Patterns and Functional Connectomic Organization
Próximo, we examined the stability of the two neurocognitive measures over time by exploring
their intraclass correlation across two sessions in 40 participants for whom a second assessment
was performed (es decir., resting-state scan and thought sampling). For “important and specific
thoughts about the self and others” (es decir., current concerns), individual variability on this thought
pattern displayed no significant intraclass correlation (ICC = 0.14, 95% BCI [−0.18, 0.43], pag =
0.20); sin embargo, the associated brain connectivity (natural log of the fractional strength) era
correlated between the first and second assessment sessions (ICC = 0.60, 95% BCI [0.36, 0.76],
pag < 0.0001). Importantly, there was a significant positive correlation between the change in
component scores on this thought pattern and the change in brain connectivity between the
two assessment sessions (Pearson r = 0.40, p = 0.011; Figure 4A). Together this pattern is
broadly consistent with a more transient state.
For “deliberate verbal thoughts about the future” (i.e., future plans) on the other hand, indi-
vidual variability on both the thought pattern (ICC = 0.48, 95% BCI [0.21, 0.69], p = 0.00065)
and the associated brain connectivity (ICC = 0.61, 95% BCI [0.38, 0.55], p < 0.0001) was
consistent across the two assessment sessions. However, there was no significant correlation
between the change in component scores on this thought pattern and the change in brain
connectivity (Pearson r = 0.24, p = 0.14; Figure 4B). Collectively, these analyses show that
Network Neuroscience
643
l
D
o
w
n
o
a
d
e
d
f
r
o
m
h
t
t
p
:
/
/
d
i
r
e
c
t
.
m
i
t
.
t
/
/
e
d
u
n
e
n
a
r
t
i
c
e
-
p
d
l
f
/
/
/
/
/
4
3
6
3
7
1
8
6
7
2
6
4
n
e
n
_
a
_
0
0
1
3
7
p
d
t
.
f
b
y
g
u
e
s
t
t
o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3
Linking thoughts, brain, and well-being
Figure 4. Test-retest reliability of the identified thought patterns and the associated functional con-
nectomic organization. A second session of resting-state fMRI scanning and thought sampling was
carried out for a total of 40 participants. (A) For important/specific thoughts, while individual com-
ponent scores on this thought pattern did not show a significant intraclass correlation (ICC = 0.14, CI
[−0.18, 0.43], p = 0.20), the associated brain connectivity pattern (natural log of fractional strength)
was largely consistent between the two repeated assessment sessions (ICC = 0.60, CI [0.36, 0.76],
p < 0.0001). Moreover, the change in thought pattern was positively related to the change in brain
connectivity (Pearson r = 0.40, p = 0.011). (B) For deliberate/verbal thoughts, both the component
scores on this thought pattern (ICC = 0.48, CI [0.21, 0.69], p = 0.00065) and the associated brain
connectivity (ICC = 0.61, CI [0.38, 0.55], p < 0.0001) showed significant intraclass correlations.
However, there was no significant link between the change in this thought pattern and the change
in brain connectivity between the two assessment sessions (Pearson r = 0.24, p = 0.14).
l
D
o
w
n
o
a
d
e
d
f
r
o
m
h
t
t
p
:
/
/
d
i
r
e
c
t
.
m
i
t
.
t
/
/
e
d
u
n
e
n
a
r
t
i
c
e
-
p
d
l
f
/
/
/
/
/
4
3
6
3
7
1
8
6
7
2
6
4
n
e
n
_
a
_
0
0
1
3
7
p
d
.
t
f
b
y
g
u
e
s
t
t
o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3
the patterns of thought captured by “deliberate and verbal thoughts about the future” and their
neural representations show greater trait-like stability over time than the participants’ state-like
“important specific thoughts about the self and others.”
Distinct Patterns of Thought Mediate the Effect of Brain Connectivity on Well-Being
Finally, having identified two patterns of ongoing thought, each with an associated profile of
complex brain functional interactions at rest, we tested whether these neurocognitive metrics
derived from our study had mediatory influences on measures of mental health and well-
being in daily life, as indicated by a cross-culturally validated World Health Organization
questionnaire (i.e., WHOQOL-BREF). Incorporating the relative importance of both positive
and negative connections (Fox, Zhang, Snyder, & Raichle, 2009; Keller et al., 2015), we first
assessed the predictive power of fractional strength on psychological and social well-being via
linear regressions, followed by mediation analyses to examine the indirect influence of brain
connectivity on well-being, mediated by the participants’ reported patterns of thought.
For “important and specific thoughts about the self and others” (i.e., current concerns),
the fractional strength (natural log) of the associated brain connectivity component signifi-
cantly predicted the participants’ psychological well-being score (β = 0.52, t(164,4) = 2.49, p =
0.014), while no significant link was observed with social well-being (β = 0.063, t(164,4) = 0.34,
p = 0.74). A mediation analysis indicated that there was a significant indirect effect of brain
Network Neuroscience
644
Linking thoughts, brain, and well-being
l
D
o
w
n
o
a
d
e
d
f
r
o
m
h
t
t
p
:
/
/
d
i
r
e
c
t
.
m
i
t
.
t
/
/
e
d
u
n
e
n
a
r
t
i
c
e
-
p
d
l
f
/
/
/
/
/
4
3
6
3
7
1
8
6
7
2
6
4
n
e
n
_
a
_
0
0
1
3
7
p
d
t
.
f
b
y
g
u
e
s
t
t
o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3
Figure 5. Model of the functional connectome as a predictor of psychological and social well-
being, mediated by patterns of thought. While brain connectivity refers to the natural log of the
fractional strength, pattern of thought represents the individual component scores on the identified
thought pattern, and the psychological/social well-being measures are self-reported ratings on the
WHOQOL-BREF questionnaire. Only a subset of the participant cohort who fully completed the
well-being questionnaire (n = 169) was utilized in this analysis. The calculation of confidence in-
tervals (CI) for the mediation effect was based on the percentile bootstrap estimation approach with
5,000 samples (corrected for age, gender, and percentage of motion-related invalid scans identified
by the scrubbing procedure). (A) A significant indirect effect of brain connectivity on psychological
well-being was observed, mediated through the participants’ important/specific thoughts, broadly
related to their current concerns. (B) There was no significant indirect effect of brain connectivity
on social well-being, mediated through the participants’ deliberate/verbal thoughts on their future
plans. Mediation results were comparable when using the total strength (sum of positive and nega-
tive strength) graph measure, albeit with a smaller effect size.
connectivity on psychological well-being, mediated by the individuals’ scores on the identi-
fied pattern of thought (β = 0.34, SE = 0.17, 95% BCI [0.010, 0.68], corrected for age, gender,
and in-scanner motion; Figure 5A). For “deliberate and verbal thoughts about the future” (i.e.,
Network Neuroscience
645
Linking thoughts, brain, and well-being
future plans) on the other hand, the fractional strength (natural log) of the identified compo-
nent of brain connectivity significantly predicted the participants’ social well-being score on
the WHOQOL-BREF (β = 0.38, t(164,4) = 2.34, p = 0.020), while no significant link was ob-
served with psychological well-being (β = −0.082, t(164,4) = −0.43, p = 0.67). Moreover,
a mediation analysis revealed no evidence for an indirect effect of brain connectivity on social
well-being through the individuals’ scores on the identified thought pattern (β = 0.085, SE =
0.096, 95% BCI [−0.099, 0.28], corrected for age, gender, and in-scanner motion; Figure 5B).
Utilizing a more conventional total strength (sum of positive and negative strength) graph
measure revealed comparable results, that is, β = 0.12 (SE = 0.045, 95% BCI [0.031, 0.22])
for current concerns, and β = −0.016 (SE = 0.019, 95% BCI [−0.062, 0.012]) for future plans,
albeit with a smaller effect size.
DISCUSSION
The aim of this study was to examine whether accounting for the patterns of ongoing thoughts
that individuals experience during periods of wakeful rest (e.g., resting-state fMRI scanning)
could allow for a more nuanced understanding of the relationships between neural organiza-
tion and mental well-being. To achieve this goal, we first identified patterns of neural con-
nectivity at rest that varied with aspects of self-reported experience during this period. One
dimension of variation was along patterns of thinking that were indicative of a focus on current
concerns. This variation was associated with functional connections from the middle frontal
gyrus, a region within the salience network. While the neural pattern was a stable feature
of the assessed individuals, the pattern of thinking did not depict such reliability. Neverthe-
less, both neural and self-report patterns changed concurrently across time, indicating that this
neurocognitive profile described a transient mapping between brain and experience. A sec-
ond pattern, associated with deliberate thoughts about the future, was dominated by functional
connections from the superior frontal gyrus, situated within the default mode network. Both
neural and experiential features of this mode were consistent across individuals and showed
little evidence of common changes over time, suggesting a neural pattern that was relatively a
stable trait. Importantly, these neural components had differential associations with measures
of mental health and well-being. The transient neurocognitive component, linked to the partic-
ipants’ focus on current concerns, was significantly associated with self-reported psychological
well-being. Mediation analysis indicated that this brain and well-being relationship was fully
mediated by the associated descriptions of ongoing experience. In contrast, while the dimen-
sion of brain connectivity variation was associated with social well-being, this relationship was
independent of the associated patterns of ongoing experience.
Together these analyses indicate that important aspects of the commonly reported relation-
ships between brain and well-being can partly be understood through their associations with
patterns of ongoing thoughts that participants experience during fMRI scanning. First, our study
shows that neural profiles, identified through their association with self-reported experience,
have differential associations with well-being. Self-reports are often subject to factors that im-
pact upon their credibility, however, the neural patterns of activity identified in this fashion
have the advantage that they are embedded in a cognitive context without the need for reverse
inference (Poldrack, 2011). Thus, experience sampling provides a complementary method for
determining whether the source of observed neural patterns are cognitive in nature, or emerge
for other reasons, such as cardiovascular function or motion-related confounds (Shen et al.,
2017). Second, our study demonstrates that experience sampling is sensitive to patterns of
neural activity that are stable across time and others that are transient. Our approach, there-
fore, may have direct relevance to studies that aim to explore the long-term stability of patterns
Network Neuroscience
646
l
D
o
w
n
o
a
d
e
d
f
r
o
m
h
t
t
p
:
/
/
d
i
r
e
c
t
.
m
i
t
.
/
/
t
e
d
u
n
e
n
a
r
t
i
c
e
-
p
d
l
f
/
/
/
/
/
4
3
6
3
7
1
8
6
7
2
6
4
n
e
n
_
a
_
0
0
1
3
7
p
d
.
t
f
b
y
g
u
e
s
t
t
o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3
Linking thoughts, brain, and well-being
of functional organization, or those that explore long-term changes in neural function. Based
on the current data, for example, experience sampling may provide a reasonably direct way to
test hypotheses into why certain neural patterns vary in their consistency across time. Despite
the inherent weaknesses associated with retrospective experience sampling (Christoff, Irving,
Fox, Spreng, & Andrews-Hanna, 2016), our study shows that the information it provides can
address important shortcomings of the conceptual interpretations placed on functional con-
nectivity patterns derived from resting-state analysis. Given the negligible cost associated with
acquiring descriptions of experience after resting-state scans, and their prevalence as a tool
of cognitive neuroscience, we see no reason why this method should not be employed as a
community standard in similar studies moving forward.
As well as highlighting the value of experience sampling to studies investigating the rela-
tionship between functional organization and traits of well-being, our study provides valuable
information into the neural processes that contribute to different types of spontaneous thought.
Current concerns, in the form of either unfulfilled goals or personally relevant information, oc-
cupy a significant portion of the thoughts we experience in our daily lives, potentially consti-
tuting a determining factor in the functional outcomes of our ongoing cognition (Klinger, 2013;
Marchetti et al., 2014). In line with these results, a key dimension of thought pattern that was
reported by the participants in our cohort was related to “important and specific thoughts about
the self and others” or more generally their current concerns. Our results revealed that impor-
tant functional connections were observed in the salience network, commonly implicated in
the detection of behaviorally important internal or external stimuli for the coordination of neu-
ral resources (Uddin, 2015). This neural system has been shown to causally influence the
functional interaction between default mode and frontoparietal networks that are commonly
anticorrelated at rest (Menon & Uddin, 2010). We recently combined momentary experience
sampling with online neural activity and found that a prefrontal region of this network was as-
sociated with the ability to prioritize patterns of episodic social thoughts during periods when
external demands were reduced (Turnbull et al., 2018). Together with such evidence our study
suggests that the role of the salience network in patterns of ongoing thought emerges from its
general capacity for prioritizing patterns of sensory, memorial, and affective content (Christoff
et al., 2016) that is motivated by their contextual relevance (McMillan, Kaufman, & Singer,
2013; Smallwood, 2013). To this end, recent perspectives argue for the amplificatory influ-
ence of such transient thought patterns on functional outcomes under particular contexts (i.e.,
stress or anger; Marchetti, Koster, Klinger, & Alloy, 2016; Watkins, 2008). This may explain
the state-like characteristic of the brain and well-being link that was observed in this study,
and its significant mediation by the participants’ current concerns. Although our experiment
was not formally set up to test an intervention-based mediatory relationship and thus could
not assess the potential influence of other variables (Bullock, Green, & Ha, 2010), the findings
of our study underline the central importance of at least one thought pattern in explaining the
link between brain physiology and mental well-being.
A second component of “deliberate and verbal thoughts about the future” was linked to
the connectivity of a limited number of nodes in transmodal cortices including regions of the
default mode, cingulo-opercular, salience, and frontoparietal networks. This neural pattern
was dominated by connections from a region of the superior frontal gyrus within the default
mode network. A deliberate focus on future goals reflects our ability to simulate and envision
the consequences of our actions based on our prior experiences (Schacter, Addis, & Buckner,
2008). Importantly, similar interactions between the default mode network and systems linked
to executive control (e.g., frontoparietal and salience networks) are observed when participants
engage in tasks that mimic this type of thought, such as creative problem-solving (Beaty et al.,
Network Neuroscience
647
l
D
o
w
n
o
a
d
e
d
f
r
o
m
h
t
t
p
:
/
/
d
i
r
e
c
t
.
m
i
t
.
t
/
/
e
d
u
n
e
n
a
r
t
i
c
e
-
p
d
l
f
/
/
/
/
/
4
3
6
3
7
1
8
6
7
2
6
4
n
e
n
_
a
_
0
0
1
3
7
p
d
t
.
f
b
y
g
u
e
s
t
t
o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3
Linking thoughts, brain, and well-being
2015), visuospatial (Vatansever, Manktelow, Sahakian, Menon, & Stamatakis, 2018) as well
as autobiographical planning (Spreng et al., 2010), and imagining reward outcomes (Gerlach,
Spreng, Madore, & Schacter, 2014). This pattern of thought was predictive of better social well-
being, an observation consistent with previous studies linking patterns of ongoing thought to
social problem-solving (Ruby, Smallwood, Sackur, et al., 2013) and our ability to infer the
actions and mental states of others (Frith, 2007). Despite its significant link with the social
health measure, however, the lack of a mediation effect of this thought pattern indicates the
trait-like nature of future plans that the participants experienced inside the scanner. With recent
hypotheses postulating the vital role of “goal attainment” in spontaneous thoughts (Klinger,
2013) and emerging reports indicating their links to personality traits (e.g., negative affectivity
and neuroticism; Andrews-Hanna et al., 2013), we postulate that the observed brain and well-
being link may reflect stable neurocognitive traits, particularly within thoughts that concern
the participants’ social support network.
Furthermore, this component was also characterized by both positive as well as negative
brain connections. Negative or anticorrelations have been historically assumed to arise from
the analysis techniques employed, and head motion that is thought to lead to spurious con-
nectivity measures (Murphy, Birn, Handwerker, Jones, & Bandettini, 2009). However, recent
reports suggest a neurophysiological basis and a potential cognitive importance of such anticor-
relations in healthy brain processing (Fox et al., 2009; Keller et al., 2015; Vatansever, Menon,
& Stamatakis, 2017). Hence, our results raise the possibility that the tuning of interactions be-
tween neural systems may give rise to different types of ongoing experience—a hypothesis that
requires further investigation. In addition, recent perspectives on the generation and mainte-
nance of thought patterns highlight the vital importance of considering the dynamic nature of
ongoing cognition and the within-individual variation in this process (Christoff et al., 2016). Al-
though retrospective thought sampling methods provide the advantage of acquiring measures
related to an undisrupted period of unconstrained cognition, online thought sampling and the
assessment of the link between neural and experiential dynamics might constitute a fruitful
route in deciphering the within- and between-individual variability in ongoing cognition and
the underlying neural mechanisms (Kucyi, 2018). Taken together with the participants’ current
concerns, further research will be required to deduce the differential influence of such stable
and transitory thought patterns on functional outcomes and to assess their therapeutic potential
as modulatory targets for intervention (Marchetti et al., 2016).
In summary, we have shown that distinct patterns of thought are reflected in the underlying
brain functional connectomes at rest and that certain types of these experiences may mediate
the influence of intrinsic brain connectivity organization on well-being. Our results highlight
the importance of considering thought patterns when establishing predictive relationships be-
tween functional connectomes and complex traits, but also suggest that taking this aspect of
human cognition into consideration may lead to better characterization of neural fingerprints
of the connectome, with the potential for more useful clinical markers. In the future, it may
be possible to tailor self-reported questions based on their neural associations, allowing devel-
opment of measures targeting particular psychiatric populations with well-established neural
hypotheses, such as mood and neurodegenerative disorders (Takamura & Hanakawa, 2017).
Furthermore, the analysis method we employed could be useful in studies that test psycholog-
ical or pharmacological interventions designed to improve well-being (Khalili-Mahani et al.,
2017), allowing these investigations to disentangle whether their intervention targets the un-
derlying neural architecture, changes in patterns of thought, or a combination of both.
Network Neuroscience
648
l
D
o
w
n
o
a
d
e
d
f
r
o
m
h
t
t
p
:
/
/
d
i
r
e
c
t
.
m
i
t
.
/
/
t
e
d
u
n
e
n
a
r
t
i
c
e
-
p
d
l
f
/
/
/
/
/
4
3
6
3
7
1
8
6
7
2
6
4
n
e
n
_
a
_
0
0
1
3
7
p
d
t
.
f
b
y
g
u
e
s
t
t
o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3
Linking thoughts, brain, and well-being
METHODS
Participant Demographics
In accordance with the Declaration of Helsinki on the conduct of research involving human
participants, ethical approval was obtained for this study from the Department of Psychology
and York Neuroimaging Centre, University of York ethics committees. Following a standard
informed consent procedure, a total of 226 healthy, right-handed (one left-handed), native
English speaker undergraduate or postgraduate students with normal to corrected vision were
recruited from the University of York. All volunteers received monetary compensation or
course credit for their participation in line with the departmental policies. As per the ex-
clusion criteria, none of the participants had a history of psychiatric or neurological illness,
severe claustrophobia, anticipated pregnancy, or drug use that could alter cognitive function-
ing. Moreover, an extensive motion-correction procedure was followed (described in detail
below) that resulted in the exclusion of 12 participants because of excessive head motion in-
side the scanner, and three participants were removed because of the impartial completion of
the thought sampling method. In total, 211 participants’ imaging and thought sampling data
were used in this analysis. The average age for this group was 20.85 years (range = 18–31,
SD = 2.44) with a 129/82 female to male ratio.
l
D
o
w
n
o
a
d
e
d
f
r
o
m
h
t
t
p
:
/
/
d
i
r
e
c
t
.
m
i
t
.
/
t
/
e
d
u
n
e
n
a
r
t
i
c
e
-
p
d
l
f
/
/
/
/
/
4
3
6
3
7
1
8
6
7
2
6
4
n
e
n
_
a
_
0
0
1
3
7
p
d
t
.
f
b
y
g
u
e
s
t
t
o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3
Decomposition of Thought Patterns
To ascertain the principal dimensions of variation in the thought patterns of this participant
cohort, a retrospective thought sampling questionnaire was administered immediately after
In this session, the participants were asked to subjectively
the resting-state fMRI scanning.
rate their thoughts during the resting-state scan on a 4-choice Likert scale from “Not at all” to
“Completely” based on a randomly presented set of questions that probed the content and form
of thoughts. This set of questions and the accompanying analysis techniques have been exten-
sively utilized in various thought sampling reports previously published in the literature (Sup-
plementary Table S1; Gorgolewski et al., 2014; Medea et al., 2016; Ruby, Smallwood, Engen, &
Singer, 2013; Ruby, Smallwood, Sackur, et al., 2013; Smallwood et al., 2016). First, the ratings
from each participant were hierarchically clustered based on the similarity of responses using
the Ward linkage method (squared Euclidean distance; Supplementary Figure S1). This tech-
nique was utilized in order to partition the thought ratings into two distinct groups, thus reducing
the number of variables to be decomposed into interpretable patterns of thought (Andrews-
Hanna et al., 2013). Subsequently, both groups of ratings were reduced to three factors each
(six in total) using PCA in SPSS (Version 23; https://www.ibm.com/products/spss-statistics). The
number of components was chosen based on scree plots indicating the eigenvalue of each
subsequent decomposition, and its ability to explain variability in the data (Supplementary
Figure S2). The component loadings for the total number of six decompositions were then
rotated using the Varimax method and the resulting factors were visualized on word clouds
(Figure 2) and heat maps (Supplementary Figure S3). The component scores of each partici-
pant on these patterns of thought were then used as between-subject covariates of interest in
the subsequent analyses. Further results on the robustness of the employed hierarchical clus-
tering and PCA procedures are provided in the Supporting Information and Supplementary
Figures S1–S4.
Well-Being Assessment
For the assessment of the participants’ self-perceived well-being, we employed the brief version
of a health questionnaire previously established by the World Health Organization Quality of
Life (WHOQOL) group (WHOQOL Group, 1998). Termed WHOQOL-BREF, this quality of life
Network Neuroscience
649
Linking thoughts, brain, and well-being
assessment has been developed to provide a more comprehensive index of overall health of
nations that extends beyond measures of mortality and morbidity; hence, it was designed to be
readily administered across cultures and countries with different economic status. Extensively
validated in a large number of centers, WHOQOL-BREF consists of 26 questions that broadly
group into four domains of physical, psychological, social, and environmental health. All par-
ticipants were asked to complete this questionnaire at a separate session administered outside
the scanner. Out of 211 participants, 169 fully completed the questionnaire. The standardized
responses for the psychological and social health domains were then used as covariates of
interest in subsequent linear regression and mediation analyses.
MRI Data Acquisition
All MRI data acquisition was carried out at the York Neuroimaging Centre, York, with a 3T GE
HDx Excite magnetic resonance imaging (MRI) scanner using an eight-channel phased array
head coil. A single run of 9-min resting-state fMRI scan was carried out using single-shot 2D
gradient-echo-planar imaging. The parameters for this sequence was as follows: TR = 3 s,
TE = minimum full, flip angle = 90◦
, matrix size = 64 × 64, 60 slices, voxel size = 3 × 3 ×
3 mm3, 180 volumes. During resting-state scanning, the participants were asked to focus on
a fixation cross in the middle of the screen. Subsequently, a T1-weighted structural scan with
three-dimensional fast spoiled gradient echo was acquired (TR = 7.8 s, TE = minimum full, flip
angle = 20◦
, matrix size = 256 × 256, 176 slices, voxel size = 1.13 × 1.13 × 1 mm3).
MRI Data Preprocessing
All preprocessing and denoising steps for the MRI data were carried out using the SPM
software package (Version 12.0; http://www.fil.ion.ucl.ac.uk/spm/) and CONN functional con-
nectivity toolbox (Version 17.f; https://www.nitrc.org/projects/conn; Whitfield-Gabrieli &
Nieto-Castanon, 2012), based on the MATLAB platform (Version 16.a; https://uk.mathworks.
com/products/matlab.html). The first three functional volumes were removed in order to achieve
steady-state magnetization. The remaining data were first corrected for motion using six de-
grees of freedom (x, y, z translations and rotations), and adjusted for differences in slice time.
Subsequently, the high-resolution structural images were coregistered to the mean functional
image via rigid-body transformation, segmented into gray/white matter and cerebrospinal fluid
probability maps, and spatially normalized to Montreal Neurological Institute (MNI) space
alongside all functional volumes using the segmented images and a priori templates. This
indirect procedure utilizes the unified segmentation–normalization framework, which com-
bines tissue segmentation, bias correction, and spatial normalization in a single unified model
(Ashburner & Friston, 2005). No smoothing was employed, complying with recent reports on
the negative influence of this procedure on the construction of functional connectomes and
graph theoretic analyses (Alakorkko, Saarimaki, Glerean, Saramaki, & Korhonen, 2017).
Furthermore, a growing body of literature indicates the potential impact of volunteer head
motion inside the scanner on the subsequent estimates of functional connectivity and graph
theory metrics (Power, Barnes, Snyder, Schlaggar, & Petersen, 2012; (Van Dijk, Sabuncu, &
Buckner, 2012); Yan, Cheung, et al., 2013; Yan, Craddock, He, & Milham, 2013). In order to
ensure that motion and other artifacts did not confound our data, we have employed an ex-
tensive motion-correction procedure and denoising steps, comparable to those reported in the
literature (Ciric et al., 2017; Power et al., 2014). In addition to the removal of six realignment
parameters and their second-order derivatives using the general linear model (GLM; Friston,
Williams, Howard, Frackowiak, & Turner, 1996), a linear detrending term was applied as well
as the CompCor method that removed five principal components of the signal from white
Network Neuroscience
650
l
D
o
w
n
o
a
d
e
d
f
r
o
m
h
t
t
p
:
/
/
d
i
r
e
c
t
.
m
i
t
.
/
/
t
e
d
u
n
e
n
a
r
t
i
c
e
-
p
d
l
f
/
/
/
/
/
4
3
6
3
7
1
8
6
7
2
6
4
n
e
n
_
a
_
0
0
1
3
7
p
d
.
t
f
b
y
g
u
e
s
t
t
o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3
Linking thoughts, brain, and well-being
matter (WM) and cerebrospinal fluid (CSF; Behzadi, Restom, Liau, & Liu, 2007). Moreover,
the volumes affected by motion were identified and scrubbed based on the conservative set-
tings of motion greater than 0.5 mm and global signal changes larger than z = 3. A total of 12
participants, who had more than 15% of their data affected by motion, were excluded from this
study (Power et al., 2014). Though recent reports suggest the ability of global signal regression
to account for head motion, it is also known to introduce spurious anticorrelations, and was
thus not utilized in our analysis (Chai, Castanon, Ongur, & Whitfield-Gabrieli, 2012; Murphy
et al., 2009; Saad et al., 2012). Nevertheless, the composite motion score (i.e., percentage of
invalid scans) for each participant was also added as a covariate in group-level analyses to fur-
ther account for the potential influence of head motion on functional connectome estimations.
Finally, a band-pass filter between 0.009 Hz and 0.08 Hz was employed in order to focus on
low-frequency fluctuations (Biswal, Yetkin, Haughton, & Hyde, 1995; Fox et al., 2005). The
maximum, and mean motion parameters and global signal change, the percentage invalid
volumes that were scrubbed, and the distribution of correlation coefficients before and after
denoising steps are provided in Supplementary Figure S6. In addition, with the aim of ensur-
ing that our results were not confounded by motion artifacts, we calculated mean framewise
displacement using the Jenkinson formulation (Jenkinson, Bannister, Brady, & Smith, 2002),
which showed no significant associations with the main variables of interest employed in this
study (Supplementary Figures S7–S8).
Functional Connectome Analysis
Brain parcellation. We adopted a set of 264 regions based on the Power et al. (2011) brain
parcellation scheme that has been previously shown to produce reliable network topolo-
gies at rest and task conditions (Cole et al., 2013; Power et al., 2011; Vatansever, Menon,
Manktelow, Sahakian, & Stamatakis, 2015). The network partitions outlined by Cole et al.
(2013) were utilized to preassign each one of the 264 ROIs to one of the 13 large-scale networks
documented in the original publication (Power et al., 2011). Namely, 10 well-established net-
works covering dorsal (DAN) and ventral attention (VAN), salience (SAN), cingulo-opercular
(CON), frontoparietal control (FPN), default mode (DMN), visual (VN), auditory (AN), somato-
motor (hand and mouth) (SMN), and subcortical networks (SCN), as well as 3 networks that
fall into memory retrieval, cerebellum, and a network of uncertain function were used as the
13 network partitions (Power et al., 2011).
Fully connected, undirected, and weighted matrices (264 ROI ×
Connectome construction.
264 ROI) of bivariate correlation coefficients (Pearson r) were constructed for each participant
using the average BOLD signal time series obtained from the 6-mm (radius) spheres placed on
the MNI coordinates of all the 264 ROIs described above. The matrices reflected both positive
and negative weighted correlations. The arbitrary thresholding and binarization processes in
graph theoretic analysis often lead to loss of information, especially in the case of negative
correlations (Rubinov & Sporns, 2011). Given recent reports suggesting a neurophysiological
basis and potential cognitive importance of such anticorrelations in healthy brain function (Fox
et al., 2009; Keller et al., 2015; Spreng, Stevens, Viviano, & Schacter, 2016), we focused on
fully connected, weighted connectomes.
Network-based statistic. Next, we aimed to ascertain components of individual functional
connectomes that significantly predicted the participants’ between-subject variation on the
identified patterns of thought. For that purpose, we employed the network-based statistic (NBS)
toolbox (Version 1.2; https://www.nitrc.org/projects/nbs/; Zalesky et al., 2010), which provides
enhanced power to identify connected brain components formed by suprathreshold edge links
that are associated with a covariate of interest, while controlling for family-wise error (FWE)
Network Neuroscience
651
l
D
o
w
n
o
a
d
e
d
f
r
o
m
h
t
t
p
:
/
/
d
i
r
e
c
t
.
m
i
t
.
t
/
/
e
d
u
n
e
n
a
r
t
i
c
e
-
p
d
l
f
/
/
/
/
/
4
3
6
3
7
1
8
6
7
2
6
4
n
e
n
_
a
_
0
0
1
3
7
p
d
t
.
f
b
y
g
u
e
s
t
t
o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3
Linking thoughts, brain, and well-being
at the component level. Utilizing this method, we entered the individual component scores
for all six patterns of thought as variables of interest, while accounting for the effects of mean
connectivity, age, gender, and percentage of motion-related invalid scans identified by the
scrubbing procedure. Using these regressors, t tests were first carried out on fully connected
whole-brain network edges for each pattern of thought to assess the relationship between the
strength of an edge link and component scores on patterns of thought, storing the size of
the connected components that survived the chosen T threshold. Next, over a total of 5,000
permutations in which the outcome measures were randomized, random null distribu-
tions of maximal component size above the chosen threshold were generated. The number of
permutations in which the maximal component size was greater than the empirical compo-
nent size, normalized by the total number of permutations, was used to estimate p values (0.05
level of significance). While the initial Tthreshold = 3.2 was used for the main analysis, compa-
rable results for Tthreshold = 3.1 and Tthreshold = 3.3 are reported in the Supporting Information
section (Supplementary Figure S9). The resulting connected brain components, the links of
which showed a significant relationship with individual variability in thought patterns, were
then defined as mask graphs to threshold individual functional connectomes, which were then
carried forward onto graph theoretic analyses (Xia, Wang, & He, 2013).
Network neuroscience analysis. Graph theoretic metrics in this study were calculated using
MATLAB functions obtained from the publicly available Brain Connectivity Toolbox (https://sites.
google.com/site/bctnet/). Commonly used in the identification of hub regions that greatly influ-
ence the efficiency of a network in distributing information (Rubinov & Sporns, 2010), network
strength denotes one of the most fundamental measures in weighted functional connectomes
and thus formed the basis of our network neuroscience analysis approach. Calculated as the
sum of all neighboring link weights (Rubinov & Sporns, 2010), we measured the strength of
connected components across all participants that were previously identified as illustrating
significant relations to individual variability in thought patterns. Based on recent reports sug-
gesting the importance of anticorrelations, we calculated positive, negative, as well as total
strength for each individual. Furthermore, given recent evidence suggesting a contribution of
the balance between both positive and negative correlations to healthy brain processing (Fox
et al., 2009; Keller et al., 2015; Spreng et al., 2016), we aimed to utilize a metric that incorpo-
rated the importance of the interplay between these links when assessing its predictive power
for explaining individual variability in self-reported mental well-being. Thus, we defined frac-
tional strength as the ratio of the sum of positive to negative links, which was later used as
the graph metric of interest in subsequent linear regression and mediation analyses. Finally, to
identify central nodes in the functionally connected components that significantly related to
the thought structures, betweenness centrality—the fraction of all shortest paths in the network
that pass through a given node—was calculated. The average positive and negative links and
the employed graph theoretic metrics were visualized on circular plots using Circos (Irimia,
Chambers, Torgerson, & Van Horn, 2012).
Test-retest reliability analysis. Our next aim in this study was to determine the reliability of the
identified patterns of thought and brain connectivity measures. This would not only establish
the generalizability of our results but would also allude to the potential differences in the state
versus trait-level variability of our neurocognitive measures. For that purpose, a second session
of resting-state scan and experience sampling was carried out for 44 participants using the same
parameters outlined above. The thought ratings and brain imaging data for this second session
were preprocessed using the same procedures, resulting in the exclusion of four participants
because of excessive motion. For the thought decomposition scores, the hierarchical clustering
and PCA decompositions obtained from the initial cohort were imposed on the ratings obtained
Network Neuroscience
652
l
D
o
w
n
o
a
d
e
d
f
r
o
m
h
t
t
p
:
/
/
d
i
r
e
c
t
.
m
i
t
.
/
t
/
e
d
u
n
e
n
a
r
t
i
c
e
-
p
d
l
f
/
/
/
/
/
4
3
6
3
7
1
8
6
7
2
6
4
n
e
n
_
a
_
0
0
1
3
7
p
d
.
t
f
b
y
g
u
e
s
t
t
o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3
Linking thoughts, brain, and well-being
from the second session (Wrigley, Albert, Deluzio, & Stevenson, 2006). Intraclass correlation
coefficients (ICC) were employed to assess the reliability of thought component scores and
the associated brain connectivity (as denoted by fractional strength) from the first and second
sessions. Furthermore, we assessed a potential link between the change in thought patterns
between the two sessions and the change in the associated brain functional connections using
Pearson correlations.
Finally, we used linear regressions to identify the relationship between
Mediation analysis.
component scores on the identified patterns of thought, the fractional strength (natural log)
of connected components, and the participants’ self-reported scores on the psychological and
social domains of the WHOQOL-BREF questionnaire. The relationships with health were Bon-
ferroni corrected for multiple comparisons across the two health domains. After establishing
linear relationships between all three measures, subsequent mediation analyses were carried
out with the aim of determining the indirect effect of brain connectivity on psychological and
social well-being through participants’ thought patterns. As it is suggested for small to medium
sample sizes (Shrout & Bolger, 2002), the percentile bootstrap estimation approach with 5,000
samples was used to ascertain the presence of a mediation effect.
l
D
o
w
n
o
a
d
e
d
f
r
o
m
h
t
t
p
:
/
/
d
i
r
e
c
t
.
m
i
t
.
/
/
t
e
d
u
n
e
n
a
r
t
i
c
e
-
p
d
l
f
/
/
/
/
/
4
3
6
3
7
1
8
6
7
2
6
4
n
e
n
_
a
_
0
0
1
3
7
p
d
.
t
f
b
y
g
u
e
s
t
t
o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3
ACKNOWLEDGMENTS
The authors extend their gratitude to Mladen Sormaz, Charlotte Murphy, Hao-Ting Wang, and
Giulia Poerio for their invaluable contribution to the scanning of participants. In addition, the
authors thank Andre Gouws, Ross Devlin, Jane Hazell, and the rest of the York Neuroimaging
Centre staff for their support in setting up the imaging protocol and scanning. Finally, we
thank all the participants for their time and effort in taking part in this study. The funders had
no role in study design, data collection and analysis, decision to publish, or preparation of the
manuscript.
SUPPORTING INFORMATION
Supporting information for this article is available at https://doi.org/10.1162/netn_a_00137.
AUTHOR CONTRIBUTIONS
Deniz Vatansever: Conceptualization; Data curation; Formal analysis; Investigation; Method-
ology; Project administration; Validation; Visualization; Writing - Original Draft; Writing -
Review & Editing. Theodoros Karapanagiotidis: Data curation; Methodology; Validation;
Visualization. Daniel S Margulies: Conceptualization. Elizabeth Jefferies: Conceptualiza-
tion; Investigation; Project administration; Resources; Supervision; Writing - Review & Editing.
Jonathan Smallwood: Conceptualization; Data curation; Funding acquisition; Investigation;
Methodology; Project administration; Resources; Supervision; Visualization; Writing - Review
& Editing.
FUNDING INFORMATION
Jonathan Smallwood, European Research Council (http://dx.doi.org/10.13039/501100000781),
Jonathan Smallwood, John Templeton Foundation (http://dx.doi.org/10.
Award ID: 646927.
13039/100000925). Deniz Vatansever, Science and Technology Commission of Shanghai Mu-
nicipality and ZJLab (http://dx.doi.org/10.13039/501100003399), Award ID: 2018SHZDZX01.
Network Neuroscience
653
Linking thoughts, brain, and well-being
REFERENCES
Alakorkko, T., Saarimaki, H., Glerean, E., Saramaki,
J., &
(2017). Effects of spatial smoothing on func-
Korhonen, O.
tional brain networks. European Journal of Neuroscience, 46(9),
2471–2480. https://doi.org/10.1111/ejn.13717
Allan Cheyne, J., Solman, G. J., Carriere, J. S., & Smilek, D. (2009).
Anatomy of an error: A bidirectional state model of task engage-
ment/disengagement and attention-related errors. Cognition, 111(1),
98–113. https://doi.org/10.1016/j.cognition.2008.12.009
Andrews-Hanna, J. R., Kaiser, R. H., Turner, A. E., Reineberg, A. E.,
Godinez, D., Dimidjian, S., & Banich, M. T.
(2013). A penny
for your thoughts: Dimensions of self-generated thought content
and relationships with individual differences in emotional well-
being. Frontiers in Psychology, 4, 900. https://doi.org/10.3389/
fpsyg.2013.00900
Antrobus, J. S., Singer, J. L., & Greenberg, S. (1966). Studies in the
stream of consciousness: Experimental enhancement and sup-
pression of spontaneous cognitive processes. Perceptual and
Motor Skills, 23, 399–417.
Ashburner, J., & Friston, K. J. (2005). Unified segmentation. Neu-
roImage, 26(3), 839–851. https://doi.org/10.1016/j.neuroimage.
2005.02.018
Baird, B., Smallwood, J., Mrazek, M. D., Kam, J. W., Franklin, M. S.,
& Schooler, J. W. (2012). Inspired by distraction: Mind wander-
ing facilitates creative incubation. Psychological Science, 23(10),
1117–1122. https://doi.org/10.1177/0956797612446024
Baird, B., Smallwood, J., & Schooler, J. W. (2011). Back to the fu-
ture: Autobiographical planning and the functionality of mind-
wandering. Consciousness and Cognition, 20(4), 1604–1611.
https://doi.org/10.1016/j.concog.2011.08.007
Barch, D. M. (2013). Brain network interactions in health and dis-
ease. Trends in Cognitive Sciences, 17(12), 603–605. https://doi.
org/10.1016/j.tics.2013.09.004
Beaty, R. E., Benedek, M., Kaufman, S. B., & Silvia, P. J. (2015). De-
fault and executive network coupling supports creative idea pro-
duction. Scientific Reports, 5, 10964. https://doi.org/10.1038/
srep10964
Behzadi, Y., Restom, K., Liau, J., & Liu, T. T. (2007). A component
based noise correction method (CompCor) for BOLD and perfu-
sion based fMRI. NeuroImage, 37(1), 90–101. https://doi.org/
10.1016/j.neuroimage.2007.04.042
Biswal, B., Yetkin, F. Z., Haughton, V. M., & Hyde, J. S.
(1995).
Functional connectivity in the motor cortex of resting human
brain using echo-planar MRI. Magnetic Resonance in Medicine,
34(4), 537–541.
Bullock, J. G., Green, D. P., & Ha, S. E. (2010). Yes, but what’s the
Journal of Person-
mechanism? (Don’t expect an easy answer).
ality and Social Psychology, 98(4), 550–558. https://doi.org/10.
1037/a0018933
Carriere, J. S., Cheyne, J. A., & Smilek, D.
(2008). Everyday at-
tention lapses and memory failures: The affective consequences
of mindlessness. Consciousness and Cognition, 17(3), 835–847.
https://doi.org/10.1016/j.concog.2007.04.008
Castellanos, F. X., Di Martino, A., Craddock, R. C., Mehta, A. D., &
Milham, M. P. (2013). Clinical applications of the functional con-
nectome. NeuroImage, 80, 527–540. https://doi.org/10.1016/
j.neuroimage.2013.04.083
Chai, X. J., Castanon, A. N., Ongur, D., & Whitfield-Gabrieli, S.
(2012). Anticorrelations in resting state networks without global
signal regression. NeuroImage, 59(2), 1420–1428. https://doi.
org/10.1016/j.neuroimage.2011.08.048
Cheng, W., Rolls, E. T., Robbins, T. W., Gong, W., Liu, Z., Lv, W., . . .
Feng, J.
(2019). Decreased brain connectivity in smoking con-
trasts with increased connectivity in drinking. eLife, 8. https://
doi.org/10.7554/eLife.40765
Christoff, K., Irving, Z. C., Fox, K. C., Spreng, R. N., & Andrews-
Hanna, J. R. (2016). Mind-wandering as spontaneous thought:
A dynamic framework. Nature Reviews: Neuroscience, 17(11),
718–731. https://doi.org/10.1038/nrn.2016.113
Ciric, R., Wolf, D. H., Power, J. D., Roalf, D. R., Baum, G. L.,
Ruparel, K., . . . Satterthwaite, T. D.
(2017). Benchmarking
of participant-level confound regression strategies for the control
of motion artifact in studies of functional connectivity. Neuro
Image, 154, 174–187.
https://doi.org/10.1016/j.neuroimage.
2017.03.020
Cole, M. W., Reynolds, J. R., Power, J. D., Repovs, G., Anticevic,
A., & Braver, T. S. (2013). Multi-task connectivity reveals flexible
hubs for adaptive task control. Nature Neuroscience, 16(9),
1348–1355. https://doi.org/10.1038/nn.3470
Engert, V., Smallwood, J., & Singer, T. (2014). Mind your thoughts:
Associations between self-generated thoughts and stress-induced
and baseline levels of cortisol and alpha-amylase. Biological
Psychology, 103, 283–291. https://doi.org/10.1016/j.biopsycho.
2014.10.004
Fox, M. D., Snyder, A. Z., Vincent, J. L., Corbetta, M., Van Essen,
D. C., & Raichle, M. E. (2005). The human brain is intrinsically
organized into dynamic, anticorrelated functional networks.
Proceedings of the National Academy of Sciences, 102(27),
9673–9678. https://doi.org/10.1073/pnas.0504136102
Fox, M. D., Zhang, D., Snyder, A. Z., & Raichle, M. E. (2009). The
global signal and observed anticorrelated resting state brain net-
works. Journal of Neurophysiology, 101(6), 3270–3283. https://
doi.org/10.1152/jn.90777.2008
Friston, K. J., Williams, S., Howard, R., Frackowiak, R. S., &
Turner, R. (1996). Movement-related effects in fMRI time-series.
Magnetic Resonance in Medicine, 35(3), 346–355.
Frith, C. D.
(2007). The social brain? Philosophical Transactions
of the Royal Society of London. Series B: Biological Sciences,
362(1480), 671–678. https://doi.org/10.1098/rstb.2006.2003
Galera, C., Orriols, L., M’Bailara, K., Laborey, M., Contrand, B.,
(2012). Mind wandering
Ribereau-Gayon, R., . . . Lagarde, E.
and driving: Responsibility case-control study. BMJ, 345, e8105.
https://doi.org/10.1136/bmj.e8105
Gerlach, K. D., Spreng, R. N., Madore, K. P., & Schacter, D. L.
(2014). Future planning: Default network activity couples with
frontoparietal control network and reward-processing regions
during process and outcome simulations. Social Cognitive and
Affective Neuroscience, 9(12), 1942–1951. https://doi.org/10.
1093/scan/nsu001
Network Neuroscience
654
l
D
o
w
n
o
a
d
e
d
f
r
o
m
h
t
t
p
:
/
/
d
i
r
e
c
t
.
m
i
t
.
/
/
t
e
d
u
n
e
n
a
r
t
i
c
e
-
p
d
l
f
/
/
/
/
/
4
3
6
3
7
1
8
6
7
2
6
4
n
e
n
_
a
_
0
0
1
3
7
p
d
t
.
f
b
y
g
u
e
s
t
t
o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3
Linking thoughts, brain, and well-being
Golchert, J., Smallwood, J., Jefferies, E., Seli, P., Huntenburg, J. M.,
Liem, F., . . . Margulies, D. S.
Individual variation
in intentionality in the mind-wandering state is reflected in the
integration of the default-mode, fronto-parietal, and limbic net-
works. NeuroImage, 146, 226–235. https://doi.org/10.1016/
j.neuroimage.2016.11.025
(2017).
Gorgolewski, K. J., Lurie, D., Urchs, S., Kipping, J. A., Craddock,
R. C., Milham, M. P., . . . Smallwood, J. (2014). A correspondence
between individual differences in the brain’s intrinsic functional
architecture and the content and form of self-generated thoughts.
PLoS ONE, 9(5), e97176. https://doi.org/10.1371/journal.pone.
0097176
Greene, A. S., Gao, S., Scheinost, D., & Constable, R. T.
(2018).
Task-induced brain state manipulation improves prediction of in-
dividual traits. Nature Communications, 9(1), 2807. https://doi.
org/10.1038/s41467-018-04920-3
Irimia, A., Chambers, M. C., Torgerson, C. M., & Van Horn, J. D.
(2012). Circular representation of human cortical networks for
subject and population-level connectomic visualization. Neuro-
Image, 60(2), 1340–1351. https://doi.org/10.1016/j.neuroimage.
2012.01.107
Jenkinson, M., Bannister, P., Brady, M., & Smith, S.
Im-
proved optimization for the robust and accurate linear registra-
tion and motion correction of brain images. NeuroImage, 17(2),
825–841. https://doi.org/10.1016/s1053-8119(02)91132-8
(2002).
Karapanagiotidis, T., Bernhardt, B. C., Jefferies, E., & Smallwood, J.
(2017). Tracking thoughts: Exploring the neural architecture of
mental time travel during mind-wandering. NeuroImage, 147,
272–281. https://doi.org/10.1016/j.neuroimage.2016.12.031
Keller,
J. B., Hedden, T., Thompson, T. W., Anteraper, S. A.,
(2015). Resting-state
Gabrieli, J. D., & Whitfield-Gabrieli, S.
anticorrelations between medial and lateral prefrontal cortex:
Association with working memory, aging, and individual differ-
ences. Cortex, 64, 271–280. https://doi.org/10.1016/j.cortex.
2014.12.001
Khalili-Mahani, N., Rombouts, S. A., van Osch, M. J., Duff, E. P.,
Carbonell, F., Nickerson, L. D., . . . van Gerven, J. M.
(2017).
Biomarkers, designs, and interpretations of resting-state fMRI in
translational pharmacological research: A review of state-of-the-
art, challenges, and opportunities for studying brain chemistry.
Human Brain Mapping, 38(4), 2276–2325. https://doi.org/10.
1002/hbm.23516
Killingsworth, M. A., & Gilbert, D. T. (2010). A wandering mind is
an unhappy mind. Science, 330(6006), 932. https://doi.org/10.
1126/science.1192439
Klinger, E. (1971). Structure and functions of fantasy. New York, NY:
Wiley-Interscience.
Klinger, E. (2013). Goal commitments and the content of thoughts
and dreams: Basic principles. Frontiers in Psychology, 4, 415.
https://doi.org/10.3389/fpsyg.2013.00415
Konishi, M., Brown, K., Battaglini, L., & Smallwood, J.
(2017).
When attention wanders: Pupillometric signatures of fluctuations
in external attention. Cognition, 168, 16–26. https://doi.org/10.
1016/j.cognition.2017.06.006.
J. B., Sato,
J. R., Sommer,
I., Balardin,
Kraft,
J., Tobo, P.,
(2018). Quality of life is re-
Barrichello, C., . . . Kozasa, E. H.
Network Neuroscience
lated to the functional connectivity of the default mode network
at rest. Brain Imaging and Behavior. https://doi.org/10.1007/
s11682-018-9954-5
Kucyi, A.
(2018).
Just a thought: How mind-wandering is repre-
sented in dynamic brain connectivity. NeuroImage, 180(Pt. B),
505–514. https://doi.org/10.1016/j.neuroimage.2017.07.001
Marchetti, I., Koster, E. H. W., Klinger, E., & Alloy, L. B.
(2016).
Spontaneous thought and vulnerability to mood disorders: The
dark side of the wandering mind. Clinical Psychological Science,
4(5), 835–857. https://doi.org/10.1177/2167702615622383
Marchetti, I., Van de Putte, E., & Koster, E. H. (2014). Self-generated
thoughts and depression: From daydreaming to depressive symp-
toms. Frontiers in Human Neuroscience, 8, 131. https://doi.org/
10.3389/fnhum.2014.00131
McMillan, R. L., Kaufman, S. B., & Singer, J. L. (2013). Ode to pos-
itive constructive daydreaming. Frontiers in Psychology, 4, 626.
https://doi.org/10.3389/fpsyg.2013.00626
McVay, J. C., & Kane, M. J. (2009). Conducting the train of thought:
Working memory capacity, goal neglect, and mind wandering in
Journal of Experimental Psychology:
an executive-control task.
Learning, Memory, and Cognition, 35(1), 196–204. https://doi.
org/10.1037/a0014104
Medea, B., Karapanagiotidis, T., Konishi, M., Ottaviani, C.,
Margulies, D., Bernasconi, A., . . . Smallwood, J. (2016). How
do we decide what to do? Resting-state connectivity patterns
and components of self-generated thought linked to the devel-
opment of more concrete personal goals. Experimental Brain
Research, 236, 2469–2481.
https://doi.org/10.1007/s00221-
016-4729-y
Menon, V., & Uddin, L. Q.
(2010). Saliency, switching, attention
and control: A network model of insula function. Brain Struc-
ture and Function, 214(5–6), 655–667. https://doi.org/10.1007/
s00429-010-0262-0
Mrazek, M. D., Smallwood, J., Franklin, M. S., Chin, J. M., Baird,
B., & Schooler, J. W. (2012). The role of mind-wandering in mea-
surements of general aptitude. Journal of Experimental Psychology:
General, 141(4), 788–798. https://doi.org/10.1037/a0027968
Murphy, K., Birn, R. M., Handwerker, D. A.,
Jones, T. B., &
Bandettini, P. A. (2009). The impact of global signal regression
on resting state correlations: Are anti-correlated networks intro-
duced? NeuroImage, 44(3), 893–905. https://doi.org/10.1016/
j.neuroimage.2008.09.036
Poerio, G. L., Sormaz, M., Wang, H. T., Margulies, D., Jefferies,
E., & Smallwood, J.
(2017). The role of the default mode net-
work in component processes underlying the wandering mind.
Social Cognitive and Affective Neuroscience, 12(7), 1047–1062.
https://doi.org/10.1093/scan/nsx041
Poldrack, R. A. (2011). Inferring mental states from neuroimaging
data: From reverse inference to large-scale decoding. Neuron,
72(5), 692–697. https://doi.org/10.1016/j.neuron.2011.11.001
Power, J. D., Barnes, K. A., Snyder, A. Z., Schlaggar, B. L., &
Petersen, S. E.
(2012). Spurious but systematic correlations
in functional connectivity MRI networks arise from subject mo-
tion NeuroImage, 59(3), 2142–2154. https://doi.org/10.1016/
j.neuroimage.2011.10.018
655
l
D
o
w
n
o
a
d
e
d
f
r
o
m
h
t
t
p
:
/
/
d
i
r
e
c
t
.
m
i
t
.
/
/
t
e
d
u
n
e
n
a
r
t
i
c
e
-
p
d
l
f
/
/
/
/
/
4
3
6
3
7
1
8
6
7
2
6
4
n
e
n
_
a
_
0
0
1
3
7
p
d
.
t
f
b
y
g
u
e
s
t
t
o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3
Linking thoughts, brain, and well-being
Power, J. D., Cohen, A. L., Nelson, S. M., Wig, G. S., Barnes, K. A.,
Church, J. A., . . . Petersen, S. E. (2011). Functional network or-
ganization of the human brain. Neuron, 72(4), 665–678. https://
doi.org/10.1016/j.neuron.2011.09.006
Power,
J. D., Mitra, A., Laumann, T. O., Snyder, A. Z.,
Schlaggar, B. L., & Petersen, S. E. (2014). Methods to detect, char-
acterize, and remove motion artifact in resting state fMRI. Neu-
roImage, 84, 320–341. https://doi.org/10.1016/j.neuroimage.
2013.08.048
Rubinov, M., & Sporns, O. (2010). Complex network measures of
brain connectivity: Uses and interpretations. NeuroImage, 52(3),
1059–1069. https://doi.org/10.1016/j.neuroimage.2009.10.003
Rubinov, M., & Sporns, O. (2011). Weight-conserving characteriza-
tion of complex functional brain networks. NeuroImage, 56(4),
2068–2079. https://doi.org/10.1016/j.neuroimage.2011.03.069
Ruby, F. J., Smallwood, J., Engen, H., & Singer, T. (2013). How self-
generated thought shapes mood—The relation between mind-
wandering and mood depends on the socio-temporal content of
thoughts. PLoS ONE, 8(10), e77554. https://doi.org/10.1371/
journal.pone.0077554
Ruby, F. J., Smallwood, J., Sackur, J., & Singer, T.
Is self-
(2013).
Fron-
generated thought a means of social problem solving?
tiers in Psychology, 4, 962. https://doi.org/10.3389/fpsyg.2013.
00962
Saad, Z. S., Gotts, S. J., Murphy, K., Chen, G., Jo, H. J., Martin,
A., & Cox, R. W. (2012). Trouble at rest: How correlation pat-
terns and group differences become distorted after global signal
regression. Brain Connectivity, 2(1), 25–32. https://doi.org/10.
1089/brain.2012.0080
Schacter, D. L., Addis, D. R., & Buckner, R. L. (2008). Episodic simu-
lation of future events: Concepts, data, and applications. Annals
of the New York Academy of Sciences, 1124, 39–60. https://doi.
org/10.1196/annals.1440.001
Seli, P., Ralph, B. C. W., Konishi, M., Smilek, D., & Schacter, D. L.
(2017). What did you have in mind? Examining the content of
intentional and unintentional types of mind wandering. Con-
sciousness and Cognition, 51, 149–156. https://doi.org/10.1016/
j.concog.2017.03.007
Seli, P., Smallwood, J., Cheyne, J. A., & Smilek, D. (2015). On the
relation of mind wandering and ADHD symptomatology. Psy-
chonomic Bulletin and Review, 22(3), 629–636. https://doi.org/
10.3758/s13423-014-0793-0
Shen, X., Finn, E. S., Scheinost, D., Rosenberg, M. D., Chun, M. M.,
Papademetris, X., & Constable, R. T. (2017). Using connectome-
based predictive modeling to predict individual behavior from
brain connectivity. Nature Protocols, 12(3), 506–518. https://
doi.org/10.1038/nprot.2016.178
Shrout, P. E., & Bolger, N.
(2002). Mediation in experimental
and nonexperimental studies: New procedures and recommen-
dations. Psychological Methods, 7(4), 422–445.
Singer, J. L., & Antrobus, J. S.
(1963). A factor-analytic study of
daydreaming and conceptually related cognitive and personality
variables. Perceptual and Motor Skills, 17(1), 187–209. https://
doi.org/10.2466/pms.1963.17.1.187
Smallwood, J. (2013). Distinguishing how from why the mind wan-
ders: A process-occurrence framework for self-generated mental
activity. Psychological Bulletin, 139(3), 519–535. https://doi.org/
10.1037/a0030010
Smallwood, J., Beach, E., Schooler, J. W., & Handy, T. C.
(2008).
Going AWOL in the brain: Mind wandering reduces cortical
Journal of Cognitive Neuroscience,
analysis of external events.
20(3), 458–469. https://doi.org/10.1162/jocn.2008.20037
Smallwood, J., Karapanagiotidis, T., Ruby, F., Medea, B., de Caso, I.,
Konishi, M., . . . Jefferies, E. (2016). Representing representation:
Integration between the temporal lobe and the posterior cingu-
late influences the content and form of spontaneous thought.
PLoS ONE, 11(4), e0152272. https://doi.org/10.1371/journal.
pone.0152272
Smallwood, J., Mrazek, M. D., & Schooler, J. W. (2011). Medicine
for the wandering mind: Mind wandering in medical practice.
Medical Education, 45(11), 1072–1080. https://doi.org/10.1111/
j.1365-2923.2011.04074.x
Smallwood, J., Nind, L., & O’Connor, R. C. (2009). When is your
head at? An exploration of the factors associated with the tempo-
ral focus of the wandering mind. Consciousness and Cognition,
18(1), 118–125. https://doi.org/10.1016/j.concog.2008.11.004
(2013). Letting go of the
present: Mind-wandering is associated with reduced delay dis-
counting. Consciousness and Cognition, 22(1), 1–7. https://doi.
org/10.1016/j.concog.2012.10.007
Smallwood, J., Ruby, F. J., & Singer, T.
Smallwood, J., & Schooler, J. W.
(2015). The science of mind
wandering: Empirically navigating the stream of consciousness.
Annual Review of Psychology, 66, 487–518. https://doi.org/10.
1146/annurev-psych-010814-015331
Spreng, R. N., Stevens, W. D., Chamberlain,
J. P., Gilmore,
A. W., & Schacter, D. L. (2010). Default network activity, coupled
with the frontoparietal control network, supports goal-directed
cognition. NeuroImage, 53(1), 303–317. https://doi.org/10.1016/
j.neuroimage.2010.06.016
Spreng, R. N., Stevens, W. D., Viviano, J. D., & Schacter, D. L.
(2016). Attenuated anticorrelation between the default and dor-
sal attention networks with aging: Evidence from task and rest.
Neurobiology of Aging, 45, 149–160. https://doi.org/10.1016/j.
neurobiolaging.2016.05.020
Stawarczyk, D., Majerus, S., Maj, M., Van der Linden, M., &
D’Argembeau, A. (2011). Mind-wandering: Phenomenology and
function as assessed with a novel experience sampling method.
Acta Psychologica, 136(3), 370–381. https://doi.org/10.1016/j.
actpsy.2011.01.002
Takamura, T., & Hanakawa, T. (2017). Clinical utility of resting-state
functional connectivity magnetic resonance imaging for mood
and cognitive disorders. Journal of Neural Transmission (Vienna),
124(7), 821–839. https://doi.org/10.1007/s00702-017-1710-2
I., Terracciano, A., &
Indovina,
Passamonti, L. (2018). Functional connectome of the five-factor
model of personality. Personality Neuroscience, 1. https://doi.
org/10.1017/pen.2017.2
Toschi, N., Riccelli, R.,
Turnbull, A., Wang, H. T., Schooler, J. W., Jefferies, E., Margulies,
D. S., & Smallwood, J.
(2018). The ebb and flow of attention:
Between-subject variation in intrinsic connectivity and cogni-
tion associated with the dynamics of ongoing experience. Neuro-
Image. https://doi.org/10.1016/j.neuroimage.2018.09.069
Network Neuroscience
656
l
D
o
w
n
o
a
d
e
d
f
r
o
m
h
t
t
p
:
/
/
d
i
r
e
c
t
.
m
i
t
.
/
t
/
e
d
u
n
e
n
a
r
t
i
c
e
-
p
d
l
f
/
/
/
/
/
4
3
6
3
7
1
8
6
7
2
6
4
n
e
n
_
a
_
0
0
1
3
7
p
d
t
.
f
b
y
g
u
e
s
t
t
o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3
Linking thoughts, brain, and well-being
Uddin, L. Q. (2015). Salience processing and insular cortical func-
tion and dysfunction. Nature Reviews: Neuroscience, 16(1),
55–61. https://doi.org/10.1038/nrn3857
Van Dijk, K. R., Sabuncu, M. R., & Buckner, R. L.
(2012). The
influence of head motion on intrinsic functional connectivity
MRI. NeuroImage, 59(1), 431–438. https://doi.org/10.1016/j.
neuroimage.2011.07.044
Van Essen, D. C., & Barch, D. M. (2015). The human connectome in
health and psychopathology. World Psychiatry, 14(2), 154–157.
https://doi.org/10.1002/wps.20228
Vatansever, D., Bozhilova, N. S., Asherson, P., & Smallwood, J.
(2018). The devil is in the detail: Exploring the intrinsic neural
link attention-deficit/hyperactivity disorder
mechanisms that
symptomatology to ongoing cognition. Psychological Medicine,
1–10. https://doi.org/10.1017/S0033291718003598
Vatansever, D., Bzdok, D., Wang, H. T., Mollo, G., Sormaz, M.,
(2017). Varieties of semantic cog-
Murphy, C., . . . Jefferies, E.
nition revealed through simultaneous decomposition of intrin-
sic brain connectivity and behaviour. NeuroImage, 158, 1–11.
https://doi.org/10.1016/j.neuroimage.2017.06.067
Vatansever, D., Manktelow, A., Sahakian, B. J., Menon, D. K., &
(2018). Default mode network engagement
Stamatakis, E. A.
beyond self-referential internal mentation. Brain Connectivity,
8(4), 245–253. https://doi.org/10.1089/brain.2017.0489
Vatansever, D., Menon, D. K., Manktelow, A. E., Sahakian, B. J.,
(2015). Default mode dynamics for
& Stamatakis, E. A.
Journal of Neuroscience, 35(46),
global functional integration.
15254–15262. https://doi.org/10.1523/jneurosci.2135-15.2015
(2017). De-
fault mode contributions to automated information processing.
Proceedings of the National Academy of Sciences, 114(48),
12821–12826. https://doi.org/10.1073/pnas.1710521114
Vatansever, D., Menon, D. K., & Stamatakis, E. A.
Wang, H. T., Bzdok, D., Margulies, D., Craddock, C., Milham, M.,
Jefferies, E., . . . Smallwood, J. (2018). Patterns of thought: Popu-
lation variation in the associations between large-scale network
organisation and self-reported experiences at rest. NeuroImage,
176, 518–527. https://doi.org/10.1016/j.neuroimage.2018.04.
064
Wang, H. T., Poerio, G., Murphy, C., Bzdok, D., Jefferies, E., &
(2018). Dimensions of experience: Exploring
Smallwood, J.
the heterogeneity of the wandering mind. Psychological Science,
29(1), 56–71. https://doi.org/10.1177/0956797617728727
Watkins, E. R.
(2008). Constructive and unconstructive repetitive
thought. Psychological Bulletin, 134(2), 163–206. https://doi.
org/10.1037/0033-2909.134.2.163
Whitfield-Gabrieli, S., & Nieto-Castanon, A.
(2012). Conn: A
functional connectivity toolbox for correlated and anticorrelated
brain networks. Brain Connectivity, 2(3), 125–141. https://doi.
org/10.1089/brain.2012.0073
WHOQOL Group. (1998). Development of the World Health Or-
ganization WHOQOL-BREF quality of life assessment. Psycho-
logical Medicine, 28(3), 551–558.
Wrigley, A. T., Albert, W. J., Deluzio, K. J., & Stevenson, J. M.
(2006). Principal component analysis of lifting waveforms. Clin-
ical Biomechanics (Bristol, Avon), 21(6), 567–578. https://doi.
org/10.1016/j.clinbiomech.2006.01.004
Xia, M., Wang, J., & He, Y. (2013). BrainNet Viewer: A network vi-
sualization tool for human brain connectomics. PLoS ONE, 8(7),
e68910. https://doi.org/10.1371/journal.pone.0068910
Yan, C. G., Cheung, B., Kelly, C., Colcombe, S., Craddock, R. C.,
Di Martino, A., . . . Milham, M. P. (2013). A comprehensive as-
sessment of regional variation in the impact of head micromove-
ments on functional connectomics. NeuroImage, 76, 183–201.
https://doi.org/10.1016/j.neuroimage.2013.03.004
Yan, C. G., Craddock, R. C., He, Y., & Milham, M. P. (2013). Ad-
dressing head motion dependencies for small-world topologies
in functional connectomics. Frontiers in Human Neuroscience,
7, 910. https://doi.org/10.3389/fnhum.2013.00910
Zalesky, A., Fornito, A., & Bullmore, E. T. (2010). Network-based
statistic: Identifying differences in brain networks. NeuroImage,
53(4), 1197–1207. https://doi.org/10.1016/j.neuroimage.2010.
06.041
Network Neuroscience
657
l
D
o
w
n
o
a
d
e
d
f
r
o
m
h
t
t
p
:
/
/
d
i
r
e
c
t
.
m
i
t
.
/
/
t
e
d
u
n
e
n
a
r
t
i
c
e
-
p
d
l
f
/
/
/
/
/
4
3
6
3
7
1
8
6
7
2
6
4
n
e
n
_
a
_
0
0
1
3
7
p
d
.
t
f
b
y
g
u
e
s
t
t
o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3
