Psychiatry Research 291 (2020) 113257

Contents lists available at ScienceDirect

Psychiatry Research
journal homepage: www.elsevier.com/locate/psychres

Age-related changes in the association of resting-state fMRI signal variability T

and global functional connectivity in non-demented healthy people
Wanqing Xiea,b, Chung-Kang Penga, Jihong Shenb, Ching-Po Linc, Shih-Jen Tsaic,d,
⁎
Shujuan Wangb, Qianqian Chub, Albert C. Yanga,c,e,
a
Division of Interdisciplinary Medicine and Biotechnology, Beth Israel Deaconess Medical Center / Harvard Medical School, Boston, MA, United States
b
College of Mathematical Sciences, Harbin Engineering University, Harbin, China
c
Institute of Brain Sciences, National Yang-Ming University, Taipei, Taiwan
d
Department of Psychiatry, Taipei Veterans General Hospital, Taipei, Taiwan
e
Digital Medicine Center, National Yang-Ming University, Taipei, Taiwan

A R T I C LE I N FO A B S T R A C T

Keywords: Research suggests that the aging relates to variability of resting-state fMRI (rs-fMRI) signal and the functional
Age-related brain activity change connectivity. However, the association between the spatial and temporal activity of resting-state fMRI signal was
Voxel-wise functional connectivity less documented. We recruited 477 healthy Han Chinese participants, who were separated into young, middle
Brain signal variability and old groups to investigate the relationship between the variability and global functional connectivity (gFC) in
Resting-state fmri signal
different age ranges using standard deviation (SD) of time series and gFC, respectively. Our analysis revealed the
changing patterns during healthy aging: 1) 17 brain regions(Olfactory_L, Orbital_L etc.) were identified to have
significant association of age with both SD and gFC respectively by linear regression analysis; 2) Two typical
associations could be observed between SD and gFC: positive and negative correlations; 3) The variation ratio of
SD to gFC was changing with age at the voxel level by using unsupervised clustering method. It is the first time to
combine voxel-wise variability and gFC together for the study of age-related changes with rs-fMRI signal. This
study may provide a new clue for understanding the synchronization of human brain based on SD and gFC due to
the effect of aging.

1. Introduction variability of signal fluctuation, which may indicate a higher level of

Aging is a continuous process associated with inevitable physical
and cognitive decline (Naik et al., 2017), and normal aging is typically
presented with decline in attention, memory and executive functioning
(Hedden and Gabrieli, 2004; Whalley et al., 2004). In recent years,
studies have utilized fMRI to study the link between heathy aging and
changes in functional brain activity and connectivity. The fMRI can
acquire a high resolution of tridimensional images together with one
dimensional time series which can capture hemodynamics responses to
the resting-state neuronal activities by measuring the contrast between
oxygenated and deoxygenated red blood cells, termed blood oxygena-
tion level-dependent (BOLD) signal (Ferreira and
Busatto, 2013).Temporal signal variability and spatial functional con-
nectivity of rs-fMRI signal are two commonly used and popularly dis-
cussed markers for the study of aging (Ishii et al., 2018).
With regard to rs-fMRI signal variability, standard deviation (SD) is
a basic measurement (Garrett et al., 2010). Higher SD implies a greater
activation in a certain area of the brain (Garrett et al., 2011). Compared
to younger adults, older adults exhibited less variability in default mode
networks and several other regions (Garrett et al., 2010), while in-
creased variability can also be revealed in the medial frontal gyri and
superior frontal gyri, as well as anterior cingulate, left fusiform and
cerebellum, bilaterally (Kielar et al., 2016). Thus, it is not clear that the
variability of different brain regions with aging are inconsistent.
As to global functional connectivity (gFC), with the rs-fMRI signal
obtained in each voxel, pairwise correlations can be estimated, which
can depict connectivity strength of a given brain region to rest of the
brain areas. Several evidences indicate that significant resting-state
functional connectivity (RSFC) changes occur during the normal aging
process (Griffanti et al., 2018). Other evidence also suggests age-related
RSFC decline in attention networks, motor networks and other systems,
which are thought to be critically implicated in attention, memory and
executive functions (Andrews‐Hanna et al., 2014; Grady et al., 2016).
However, findings of increased functional connectivity related to aging

⁎
Corresponding author at: Insitute of Brain Science / Digital Medicine Center, National Yang-Ming University, No.155, Sec.2, Linong Street, Taipei, 112.
E-mail address: accyang@gmail.com (A.C. Yang).

https://doi.org/10.1016/j.psychres.2020.113257
Received 30 March 2020; Received in revised form 24 June 2020; Accepted 25 June 2020
Available online 27 June 2020
0165-1781/ © 2020 Published by Elsevier B.V.
W. Xie, et al.
have also been described (Biswal et al., 2010). For instance, increased
functional interaction within frontal and parietal networks, which are
both involved in attention, evidenced by a higher degree of intra-net-
work connectivity in elderly individuals compared to young controls
(Toussaint et al., 2011). Furthermore, aging-related increase in both
long- and short-range functional connectivity were detected in the so-
matosensory and motor cortices, cerebellum and brainstem
(Tomasi and Volkow, 2012). These findings of increases and decreases
in RSFC suggests that the interpretation of age-related RSFC changes is
inconsistent.
Considering that both rs-fMRI temporal activity and gFC are related
to aging (Britz et al., 2008; Hutchison and Morton, 2016), it is rea-
sonable to assume the possibility of associations between the variability
of rs-fMRI signal and gFC. The correlation coefficients between dy-
namic amplitude of low-frequency fluctuation and dynamic functional
connectivity in 48 brain regions revealed that the two items were
correlated in time and their correlations are altered in schizophrenia
(Fu et al., 2017). However, until recently, previous work only focused
on some certain brain regions. The voxel-wise relationship between gFC
and SD of rs-fMRI signal in each brain region and whether such patterns
in brain BOLD activity are altered with aging remains poorly docu-
mented.
In this study, we hypothesized that aging is not only associated with
changes in SD and gFC of rs-fMRI signal, but also the relationship be-
tween SD and gFC differentially associated with aging in different brain
regions. To test this hypothesis, we proposed a region of interest ap-
proach using Automated Anatomical Labeling (AAL) atlas (Tzourio-
Mazoyer et al., 2002). The overall aim of this study was to characterize
changes in the variability and functional connectivity of spontaneous
neuronal signals associated with aging. In addition, we expected to
reveal the detailed relationship of SD-gFC and aging in voxel-wised
brain regions. The combination of rs-fMRI signal variability and gFC
may be of help to understand how the functionality of brain activity
changes spatially and temporally during healthy aging.
2. Methods
2.1. Study participants and image acquisition
The study cohort consisted of 477 Han Chinese participants from
communities in Northern Taiwan. All of the participants were healthy
volunteers and were separated into three age groups: young (age
20–39), middle (age 40–59), and old group (age ≥ 60). The age groups
were used to investigate the relationship between SD and gFC in dif-
ferent age ranges. Table 1 shows the demographic characteristics of the
normal aging cohort. Each participant was evaluated by a trained re-
search assistant using a mini-international neuropsychiatric interview
to exclude those with Axis I psychiatric disorders (Sheehan et al., 1998).
The cognitive function of all participants was assessed using the Mini-
Mental State Examination (MMSE) (Folstein et al., al.,1975) and
Wechsler Digit Span Task (DST) (Wechsler, 1997). The old group were
further assessed using the CDR scale (Hughes et al., 1982) to exclude
those with dementia. The overall exclusion criteria for all participants
consisted of the following: (a) the presence of dementia; (b) the pre-
sence of Axis I psychiatric disorders, such as schizophrenia, bipolar
disorders, or unipolar depression; and (c) a history of neurological
Table 1
Normal aging cohort characteristics.
Age group (year) Number of participants Females (%)
20–39 152 79 (52.0)
40–59 135 80 (59.3)
≥60 190 66 (34.7)
Key: DST-F, digit span forward test; DST-B, digit span backward test; MMSE, mini-mental state examination.
2
Psychiatry Research 291 (2020) 113257
conditions, such as head injury, stroke, or Parkinson's disease.
The study was conducted in accordance with the Declaration of
Helsinki and was approved by the Institutional Review Board of Taipei
Veterans General Hospital.
All participants underwent the same fMRI scanning protocol per-
formed at the National Yang-Ming University using 3.0T Siemens MRI
scanner (Siemens, Erlangen, Germany) equipped with a 12-channel
head coil (Yang et al., 2014). The rs-fMRI images included 200 EPI
volume images, which were acquired along the AC-PC plane at the rate
of 2.5 s. The duration of the rs-fMRI scanning procedure, including
acquisition of structural MRI image, was approximately 15 mins for
each participant.
The rs-fMRI images were preprocessed through the Data Processing
Assistant for Resting-State fMRI toolbox (Chao-Gan and Yu-
Feng, 2010), and the preprocessing procedures were consistent with our
previous report (Yang et al., 2015). Briefly, we preprocess the image as
follows: removed first 5 time points, slice-time corrected, realigned
(head motion control), normalized into the standard stereotaxic space
and resampled to a 3 mm cubic voxel. Temporal low-pass filtering
(0.01–0.08 Hz) was performed to reduce the influence of high-fre-
quency noise from physiologic confounders. Covariates of the fMRI
time series were regressed out, including the time courses of six head
motion, white matter, and cerebrospinal fluid. To avoid introducing
distortions in the time series data, no global signal regression was
performed (Murphy et al., 2009).
2.2. Analysis of temporal variability and spatial connectivity of rs-fMRI
signal
Fig. 1 presents the data analysis flow of this study in Saitama
Prostate Symptom Score (SPSS 22.0). First, we estimated the temporal
variability of brain activity by calculating the SD of rs-fMRI signal in
each gray matter voxel. Second, we estimated the spatial connectivity
using gFC (Van Den Heuvel and Pol, 2010). Briefly, the pairwise de-
pendency between rs-fMRI signal in each gray matter voxel was cal-
culated using Pearson's correlation and the correlation coefficient was
further normalized by Fisher transformation (Anders et al., 2004). The
gFC value for a gray matter voxel i was then calculated by:
∑ Fz in − 1
gFCi =
n−1
Where n is the total voxel number in gray matter, Fz represents the
Fisher-transformed z-value in every other gray matter voxel.
Franzmeier et al. (2017) suggested that inclusion of both positive and
negative correlations may underestimate functional connectivity.
However, our study is to investigate the relationship between gFC and
local brain signal variability, we still include all pairwise correlations to
reduce potential bias due to arbitrary exclusion of negative functional
connectivity. Third, we derived the mean SD and gFC in each 90 brain
regions according to AAL.
2.3. Statistical analysis
We performed a linear regression analysis to identify the brain re-
gions that both their SD and gFC were significantly correlated to the
age, with corrections for multiple comparisons using the Bonferroni
EDU_YEAR DST-F DST-B MMSE
17.6 ± 2.4 15.3 ± 1.43 10.7 ± 2.9 29.3 ± 1.1
14.7 ± 3.4 13.6 ± 2.2 8.3 ± 3.0 28.5 ± 1.3
9.5 ± 6.3 12.6 ± 2.7 5.5 ± 3.4 27.0 ± 2.8
W. Xie, et al.
Fig. 1. The flowchart for the exploration of brain regions both have significant correlation between variability of rs-fMRI signal / gFC and age.
method and a P value less than 0.00056(0.05/90) was considered sig-
nificant. Next, the relationship between SD and gFC in each voxel was
visualized at the three age groups level for each identified aging-related
brain region, and visualized the change of brain voxel clusters using
BrainNet (Xia et al., 2013). We then used an unsupervised classification
method, k-means clustering with angle between vectors as the distance
measured to identify the potential clustering of the brain voxels in a
given brain region based on the relationship between SD and gFC.
Furthermore, we calculated the slope of linear regression as well as the
number of points (number of brain voxels) within each cluster of brain
voxel clusters and to evaluate if these brain voxel cluster properties
(i.e., slope and the number of voxels) would be significantly different
among three age groups using one-way analysis of variance (ANOVA).
Post-hoc test (Bonferroni) was then used to assess the between-group
difference and the P-value less than 0.00056 was considered significant.
In order to disregard the effects of the AAL template, we also scatter
plotted the SD-gFC for the whole brain based on the mean SD and gFC
of each age group and calculated the linear regression slope.
3
Psychiatry Research 291 (2020) 113257
3. Results
3.1. Identify the association of SD and gFC with aging
We identified 17 brain regions that their mean SD and gFC were
both significantly correlated to age (list in Table 2). In general, age was
correlated positively with SD and negatively with gFC. The highest
correlation between SD and age was observed in right insula, followed
by left and right olfactory lobe, and left amygdala. On the other hand,
the highest correlation between gFC and age was observed in right and
left heschl gyrus, followed by left superior temporal pole and left
amygdala.
3.2. The relationship between SD and gFC and their association with aging
Based on the SD-gFC scatter plots of the three age groups in Fig. 2,
we found that most of the brain regions showed distinct relationships
and the correlation between SD and gFC changes with age: Besides the
positive correlation part which can be observed in all the scatter plots,
negative correlation can also be found in some parts of certain brain
regions; The slope of the part with a positive correlation between SD
and gFC increased in the old-age group, while the young group is bigger
W. Xie, et al. Psychiatry Research 291 (2020) 113257

Table 2 age-related brain regions. Fig. 2 shows the results of four representative
The P value of all the brain regions was corrected for multiple comparison with brain regions and the corresponding brain mapping, all identified re-
Bonferroni method at the significance level of P < 0.00056 (0.05/90). gions can be found in Figure 1 in supplementary materials of in-
Brain Area Correlation Coefficient Correlation Coefficient dividuals in each age group. The identified brain regions can be divided
Between SD and Age Between g FC and Age in two typical groups: Only positive correlation, such as left medial
prefrontal orbital; both positive and negative correlations, such as in-
Olfactory_L 0.455 −0.202
sular and right superior temporal, temporal pole. Because the voxel
Olfactory_R 0.455 −0.183
Medial Prefrontal, 0.247 −0.186 numbers in amygdala are not big enough, the association between SD
Orbital_L and gFC is not so obvious as others in this region. Moreover, the cor-
Rectus_L 0.304 −0.230 responding brain mapping shows that the anatomical structure of the
Rectus_R 0.340 −0.189
brain region can be separated into two organized parts, and changes of
Insula_R 0.458 −0.209
Anterior Cingulate_L 0.204 −0.217
voxel numbers and locations in the cluster with positive correlation can
Amygdala_L 0.443 −0.233 also be reflected in the anatomical brain mapping.
Amygdala_R 0.373 −0.193
Postcentral_L 0.062 −0.190
Postcentral_R 0.046 −0.202 3.3. Aging-related change in the relationship between temporal variability
Superior Parietal_R 0.028 −0.178 and spatial functional connectivity of rs-fMRI signal
Heschl_L 0.391 −0.258
Heschl_R 0.360 −0.283
Fig. 3 depicts between-the-group comparisons of the slopes and
Superior Temporal_L 0.230 −0.184
Superior Temporal, 0.371 −0.240 volume ratio in two clusters identified by the k-mean clustering of SD-
Temporal Pole_L gFC relationship in each brain region across the three age groups. Ex-
Superior Temporal, 0.324 −0.233 ample region left rectus is shown in Fig. 3(a), the clusters colored in
Temporal Pole_R
blue and red called the left and right clusters for convenience sepa-
L: left; R: right.
rately.
In general, most of the slopes were significantly increased in the old-
than the middle-age group. According to the consideration of Silhouette age group for left and right clusters. Specifically, for the left cluster, the
score, 2-mean cluster was selected. Table 2 in supplementary materials slopes were increased in the old-age group in olfactory, left medial
shows the Silhouette scores of k-means clustering (k = 2, 3, 4) for all 17 prefrontal, orbital, left and right postcentral brain regions, compared to
the young-or middle-age group. For the right cluster, the slopes were

Fig. 2. (a), (b), (c) and (d) are the results of Insula, Amygdala, left Medial prefrontal orbital, and right superior temporal, temporal pole respectively. Red and blue
represent two kinds of voxels after 2-mean cluster in SD-gFC scatter plots. The cluster colored red in FC-SD corresponds to the red area in the 3D brain.

4
W. Xie, et al.
Fig. 3. (a) is an example from the brain region left rectus which shows clearly the left cluster in blue and the right cluster in red. Fig. 3(b) shows the ANOVA statistical
results after post hoc for the slope of left and right cluster, and the volume ratio of left and right cluster for different age groups.
increased in the old-age group in all the 17 brain regions except the left
heschl. Of note, we observed that the slopes of the left cluster were
decreased in the right insula, amygdala, left heschl, left superior tem-
poral and left superior temporal, temporal pole brain regions, and none
of the slopes of the right cluster of SD-gFC relationship were increased
in any brain regions.
We further quantified the percentage of sub-brain regions that were
occupied by the left and right clusters of voxels in each brain region. In
general, the sub-brain regions formed by the left cluster of voxels of SD-
gFC relationship were increased in the old-age group, compared to the
young-and middle-age group and vice versa for the right cluster.
According to the results in Fig. 4, positive correlation and the faint
trend of negative correlation can be observed in each age group on the
whole brain. The slope between SD and gFC in the old-age group is
larger than other groups, and the positive slope of young group is larger
than the middle-age group.
5
Psychiatry Research 291 (2020) 113257
4. Discussion
4.1. Aging related brain regions with the consideration of SD and gFC
We identified significant age-related changes both in SD and gFC in
17 brain regions, which include medial prefrontal, orbital, anterior
cingulate and superior temporal in the left hemisphere, insular and
superior parietal in the right hemisphere, and both side of olfactory,
rectus, amygdala, post central, heschl, and superior temporal, and
temporal pole. All the 17 brain regions have significant positive cor-
relation between SD and age, while have significant negative correla-
tion between gFC and age. The study found that the negative correla-
tions between the age and RSFC were identified either when the overall
DMN was inspected, or when the analysis was restricted to the anterior
cingulate and left medial prefrontal cortices (Esposito et al., 2008).
Significant negative correlations between age and RSFC are docu-
mented in the posterior cingulate cortex, as well as in the left medial
W. Xie, et al.
Fig. 4. (a), (b) and (c) are the voxel-wise scatter plots of mean SD and FC based on the young group, middle-age group and the old group, respectively.
prefrontal cortex (Bluhm et al., 2008). Recently, the study revealed that
decreased fMRI signal variability for older participants can be found in
posterior cingulate gyrus, precuneus, cuneus, lateral parietal regions,
and also in the left inferior frontal gyrus, left middle frontal gyrus, left
middle temporal, middle occipital gyrus, bilaterally. The study also
revealed areas of increased variability for the older adults in the left and
right medial frontal gyri and superior frontal gyri, as well as anterior
cingulate gyrus, left fusiform gyrus (Kielar et al., 2016).
The increase of variability in left anterior cingulate and decrease of
RSFC in left medial prefrontal are consistent with the previous studies,
while the decrease of variability and increase of REFC cannot be de-
tected in our study. Though different changes appear in different brain
areas, there are 17 brain regions that all have significant correlation
between SD/RSFC and age according to the results of our study. Thus, it
is reasonable to assume the possibility of associations between the
variability of rs-fMRI signal and gFC.
4.2. Association between SD and gFC affected by aging
Rs-fMRI can characterize spontaneous brain activity and identify
brain networks with co-varied patterns, which make it a powerful
technique for examine brain abnormalities with the affection of aging
without considering the difference in task performance between dif-
ferent age groups. In this paper, the relationship between SD and gFC at
a voxel wise level was studied in each identified brain area. According
to the scattered figure of gFC and SD, positive correlation surely exists
in some brain areas, such as the left medial prefrontal orbital. While in
other brain areas, such as insula, amygdala, and the right superior
temporal, and temporal pole, the relationship between gFC and SD is
not consistent within the brain region that part of voxels within a given
brain region showed a positive correlation between SD and gFC, while
the remaining voxels showed a negative correlation between SD and
gFC.
Moreover, the slope (gFC-SD) has significant differences: compare
to young and middle age groups, the slope of old group significantly
increases in most of the positive correlated areas. However, compared
to the young- and middle-age group, only both sides of postcentral,
right superior parietal, and left superior temporal have increased slope,
while the middle-age group has a significantly smaller slope than the
young group. This may conform the ‘Last-in first-out’ theory which
suggests that networks involving regions of higher-order cognitive tasks
mature later during the development, but start declining at a relatively
younger age (Douaud et al., 2014; Salat et al., 2004).
6
Psychiatry Research 291 (2020) 113257
4.3. Age-related association between rs-fMRI signal variability and gFC
Age-related effects on brain activities are widely studied with the
consideration of fMRI signal variation or functional connectivity. Age is
associated with a marked reduction in episodic memory function
(Nyberg et al., 2012; Rönnlund et al., 2005). Encoding the activity of
amygdala has specifically been linked to memory durability when
emotional stimuli were considered (Liu et al., 2014; Uncapher and
Rugg, 2005), and anterior cingulate, frontal, and insular could be
considered in the process of memory-strength modifying re-encoding
(Liu and Reder, 2016; Wing et al., 2013). A potential clue from this is
that if finer sub-segmentation can be implemented to the very brain
area, more information could be revealed.
According to the results of our study, both positive and negative
correlations can be observed between the variability and gFC of rs-fMRI
signal, especially under the segmentation of AAL template. Moreover,
the regression slope between variability and gFC significantly changes
with different age groups. Considering both positive and negative cor-
relations, the variability in the old-age group has the least influence on
gFC, while the middle-age group has the strongest influence of all.
Furthermore, the results are not limited by the anatomical template. To
the best of our knowledge, this is the first study to investigate age-
related brain regions based on the voxel-wise relationship of variability
and gFC, hence limited literature can be found to endorse our results.
The findings in our study suggest a new way to comprehend the syn-
chronization of human brain, further studies are required to better
understand the finer associations between the variation and correlation
of the rs-fMRI signal affected by aging. Moreover, if other diseases or
age-related diseases could be studied with the method mentioned in our
study, more meaningful and reasonable results might be revealed.
Author statement
Thanks for taking your time to review this manuscript entitled “Age-
related changes in the association of resting-state fMRI signal variability and
global functional connectivity in non-demented healthy people” (ID:
PSY_2020_689). All authors took part in the learning and re-
searches about all the comments carefully, and we work hard for more
than one month to update our manuscript which we hope meet with
approval (resubmitted online). Thanks again for all the comments!
In addition, we will say again that our manuscript has not been
submitted elsewhere, in whole or in part, and will not be submitted
elsewhere while under consideration at Psychiatry Research.
W. Xie, et al.
All the authors have made substantial contributions to the re-
searches, including data analysis, methodological study and manuscript
preparation. All the authors listed here have approved the manuscript
that is enclosed. We confirmed that all authors have reviewed the
manuscript, approve it's the contents and validate the accuracy of the
data. And all authors reported no biomedical financial interests or po-
tential conflicts of interest.
Disclosure statement
The authors have no conflict of interest to disclose.
Declaration of competing interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Acknowledgments
The study is supported in part by grants from the Ministry of Science
and Technology (MOST) of Taiwan (MOST 103–2314-B-075–067-MY3;
MOST 104–2218-E-010–007-MY3; MOST 104–2633-B-400–001; MOST
104–2314-B-075 −078 -MY2); Taiwan Veterans General Hospital
(VGHUST102-G1–2–1, and VGHUST103-G1–4–1); Brain Research
Center, National Yang-Ming University, Taiwan; and the Ministry of
Education, Aim for the Top University Plan, Taiwan. Authors also ac-
knowledge the MRI support received from the MRI Core Laboratory of
National Yang-Ming University, Taiwan. Dr. Xie is supported by the
Oversea Academic Training Funds, University of Science and
Technology of China and the China Scholarship Council; National
Natural Science Foundation of China, and Grant Numbers: 61503356;
Anhui Provincial Natural Science Foundation, Grant Number:
1608085QF153.
Supplementary materials
Supplementary material associated with this article can be found, in
the online version, at doi:10.1016/j.psychres.2020.113257.
References
Anders, S., Lotze, M., Erb, M., Grodd, W., Birbaumer, N., 2004. Brain activity underlying
emotional valence and arousal: a response‐related fMRI study. Hum. Brain Mapp. 23
(4), 200–209.
Andrews-Hanna, J.R., Smallwood, J., Spreng, R.N., 2014. The default network and self-
generated thought: component processes, dynamic control, and clinical relevance.
Ann. N. Y. Acad. Sci. 1316 (1), 29–52.
Biswal, B.B., Mennes, M., Zuo, X.-.N., Gohel, S., Kelly, C., Smith, S.M., Beckmann, C.F.,
Adelstein, J.S., Buckner, R.L., Colcombe, S., 2010. Toward discovery science of
human brain function. Proc. Natl. Acad. Sci. 107 (10), 4734–4739.
Bluhm, R.L., Osuch, E.A., Lanius, R.A., Boksman, K., Neufeld, R.W., Théberge, J.,
Williamson, P., 2008. Default mode network connectivity: effects of age, sex, and
analytic approach. Neuroreport 19 (8), 887–891.
Britz, J., Landis, T., Michel, C.M., 2008. Right parietal brain activity precedes perceptual
alternation of bistable stimuli. Cereb. Cortex 19 (1), 55–65.
Chao-Gan, Y., Yu-Feng, Z., 2010. DPARSF: a MATLAB toolbox for “pipeline” data analysis
of resting-state fMRI. Frontiers in systems neuroscience 4.
Douaud, G., Groves, A.R., Tamnes, C.K., Westlye, L.T., Duff, E.P., Engvig, A., Walhovd,
K.B., James, A., Gass, A., Monsch, A.U., 2014. A common brain network links de-
velopment, aging, and vulnerability to disease. Proc. Natl. Acad. Sci. 111 (49),
17648–17653.
Esposito, F., Aragri, A., Pesaresi, I., Cirillo, S., Tedeschi, G., Marciano, E., Goebel, R., Di
Salle, F., 2008. Independent component model of the default-mode brain function:
combining individual-level and population-level analyses in resting-state fMRI.
Magn. Reson. Imaging 26 (7), 905–913.
Ferreira, L.K., Busatto, G.F., 2013. Resting-state functional connectivity in normal brain
aging. Neuroscience & Biobehavioral Reviews 37 (3), 384–400.
Psychiatry Research 291 (2020) 113257
Folstein, M.F., Folstein, S.E., McHugh, P.R., 1975. “Mini-mental state”. A practical
method for grading the cognitive state of patients for the clinician. J. Psychiatr. Res.
12, 189e198.
Franzmeier, N., Caballero, M.Á.Araque, Taylor, A.N.W., Simon-Vermot, L., Buerger, K.,
Ertl-Wagner, B., Mueller, C., Catak, C., Janowitz, D., Baykara, E., Gesierich, B.,
Duering, M., Ewers, M., 2017. Resting-state global functional connectivity as a bio-
marker of cognitive reserve in mild cognitive impairment. Brain imaging and beha-
vior 11 (2), 368–382.
Fu, Z., Tu, Y., Di, X., Du, Y., Pearlson, G., Turner, J., Biswal, B.B., Zhang, Z., Calhoun, V.,
2017. Characterizing dynamic amplitude of low-frequency fluctuation and its re-
lationship with dynamic functional connectivity: an application to schizophrenia.
Neuroimage.
Garrett, D.D., Kovacevic, N., McIntosh, A.R., Grady, C.L., 2010. Blood oxygen level-de-
pendent signal variability is more than just noise. J. Neurosci. 30 (14), 4914–4921.
Garrett, D.D., Kovacevic, N., McIntosh, A.R., Grady, C.L., 2011. The importance of being
variable. J. Neurosci. 31 (12), 4496–4503.
Grady, C., Sarraf, S., Saverino, C., Campbell, K., 2016. Age differences in the functional
interactions among the default, frontoparietal control, and dorsal attention networks.
Neurobiol. Aging 41, 159–172.
Griffanti, L., Stratmann, P., Rolinski, M., Filippini, N., Zsoldos, E., Mahmood, A.,
Zamboni, G., Douaud, G., Klein, J.C., Kivimäki, M., 2018. Exploring variability in
basal ganglia connectivity with functional MRI in healthy aging. Brain imaging and
behavior 1–6.
Hutchison, R.M., Morton, J.B., 2016. It's a matter of time: reframing the development of
cognitive control as a modification of the brain's temporal dynamics. Developmental
cognitive neuroscience 18, 70–77.
Ishii, R., Canuet, L., Aoki, Y., Hata, M., Iwase, M., Ikeda, S., Nishida, K., Ikeda, M., 2018.
Healthy and Pathological Brain Aging: from the Perspective of Oscillations,
Functional Connectivity, and Signal Complexity. Neuropsychobiology.
Kielar, A., Deschamps, T., Chu, R.K., Jokel, R., Khatamian, Y.B., Chen, J.J., Meltzer, J.A.,
2016. Identifying dysfunctional cortex: dissociable effects of stroke and aging on
resting state dynamics in MEG and fMRI. Frontiers in aging neuroscience 8, 40.
Liu, Q., Dong, Q., Chen, C., Xue, G., 2014. Neural processes during encoding support
durable memory. Neuroimage 88, 1–9.
Liu, X.L., Reder, L.M., 2016. fMRI exploration of pedagogical benefits of repeated testing:
when more is not always better. Brain and behavior 6 (7).
Murphy, Kevin, Birn Rasmus, M., Handwerker Daniel, A., Jones Tyler, B., Bandettini
Peter, A., 2009. The impact of global signal regression on resting state correlations:
are anti-correlated networks introduced. Neuroimage 44 (3), 893–905.
Naik, S., Banerjee, A., Bapi, R.S., Deco, G., Roy, D., 2017. Metastability in senescence.
Trends Cogn. Sci. 21 (7), 509–521.
Nyberg, L., Lövdén, M., Riklund, K., Lindenberger, U., Bäckman, L., 2012. Memory aging
and brain maintenance. Trends Cogn. Sci. 16 (5), 292–305.
Rönnlund, M., Nyberg, L., Bäckman, L., Nilsson, L.-.G., 2005. Stability, growth, and de-
cline in adult life span development of declarative memory: cross-sectional and
longitudinal data from a population-based study. Psychol. Aging 20 (1), 3.
Salat, D.H., Buckner, R.L., Snyder, A.Z., Greve, D.N., Desikan, R.S., Busa, E., Morris, J.C.,
Dale, A.M., Fischl, B., 2004. Thinning of the cerebral cortex in aging. Cereb. Cortex
14 (7), 721–730.
Tomasi, D., Volkow, N.D., 2012. Aging and functional brain networks. Mol. Psychiatry 17
(5), 549.
Toussaint, P.-.J., Maiz, S., Coynel, D., Messé, A., Perlbarg, V., Habert, M.O., Benali, H.,
2011. Characterization of the default mode functional connectivity in normal aging
and Alzheimer’s disease: an approach combining entropy-based and graph theoretical
measurements, Biomedical Imaging: from Nano to Macro. In: 2011 IEEE International
Symposium on. IEEE, pp. 853–856.
Tzourio-Mazoyer, N., Landeau, B., Papathanassiou, D., Crivello, F., Etard, O., Delcroix, N.,
Mazoyer, B., Joliot, M., 2002. Automated anatomical labeling of activations in SPM
using a macroscopic anatomical parcellation of the MNI MRI single-subject brain.
Neuroimage 15 (1), 273–289.
Uncapher, M.R., Rugg, M.D., 2005. Encoding and the durability of episodic memory: a
functional magnetic resonance imaging study. J. Neurosci. 25 (31), 7260–7267.
Van Den Heuvel, M.P., Pol, H.E.H., 2010. Exploring the brain network: a review on
resting-state fMRI functional connectivity. Eur. Neuropsychopharmacol. 20 (8),
519–534.
Wechsler, D., 1997. Wechsler Adult Intelligence Scale, third ed. The Psychological
Corporation, San Antonio, TX.
Wing, E.A., Marsh, E.J., Cabeza, R., 2013. Neural correlates of retrieval-based memory
enhancement: an fMRI study of the testing effect. Neuropsychologia 51 (12),
2360–2370.
Xia, M., Wang, J., He, Y., 2013. BrainNet Viewer: a network visualization tool for human
brain connectomics. PLoS One 8 (7), e68910.
Yang, A.C., Hong, C.J., Liou, Y.J., Huang, K.L., Huang, C.C., Liu, M.E., Lo, M.T., Huang,
N.E., Peng, C.K., Lin, C.P., Tsai, S.J., 2015. Decreased resting-state brain activity
complexity in schizophrenia characterized by both increased regularity and ran-
domness. Hum Brain Mapp 36 (6), 2174–2186.
Yang, A.C., Huang, C.C., Liu, M.E., Liou, Y.J., Hong, C.J., Lo, M.T., Huang, N.E., Peng,
C.K., Lin, C.P., Tsai, S.J., 2014. The APOE ɛ4 allele affects complexity and functional
connectivity of resting brain activity in healthy adults. Hum. Brain Mapp. 35 (7),
3238–3248.