Fan Et Al 2023 - Right Amygdala-Right Precuneus Connectivity Is Associated With Childhood Trauma in Major Depression Patiens and Healthy Control
Fan Et Al 2023 - Right Amygdala-Right Precuneus Connectivity Is Associated With Childhood Trauma in Major Depression Patiens and Healthy Control
Fan Et Al 2023 - Right Amygdala-Right Precuneus Connectivity Is Associated With Childhood Trauma in Major Depression Patiens and Healthy Control
Abstract
The present study investigated the effect of childhood trauma (CT) on amygdala and hippocampus functional connectivity (FC) and the
association with clinical presentations of major depressive disorder (MDD). Participants included 73 MDD patients (42 with moderate-
to-severe CT and 31 with no or low CT) and 64 healthy controls (HC; 30 with moderate-to-severe CT and 34 with no or low CT). Seed-based
whole-brain resting-state FC analyses were performed with seeds located in amygdala and hippocampus. Individuals with moderate-
to-severe CT, irrespective of MDD diagnosis, had decreased right amygdala–right precuneus connectivity compared to those with no
or low CT. Right amygdala–right precuneus connectivity was significantly correlated with physical and social trait anhedonia in MDD.
Mediation effects of this FC on relationship between CT (specifically neglect but not abuse) and trait anhedonia in MDD were significant.
MDD patients demonstrated increased right amygdala–left middle frontal gyrus FC, decreased right amygdala–right medial superior
frontal gyrus (mSFG) FC and decreased right hippocampus–bilateral mSFG FC relative to HC. Findings highlight the effect of CT on right
amygdala–right precuneus FC irrespective of MDD diagnosis. FC of right amygdala–right precuneus may be involved in the mechanism
linking CT and depression through its association with trait anhedonia.
Introduction both cortical and subcortical brain regions (Ochsner et al., 2012).
The amygdala connectivity has been proposed to be involved in
Major depressive disorder (MDD) is one of the most common
critical functions related to depression including modulation of
psychiatric conditions, of which the core symptoms were low
sensory information, generation of emotion experience and emo-
mood and anhedonia. Previous studies have well established that
childhood trauma (CT) is one of the strongest risk factors in tion regulation (Banks et al., 2007; Morawetz et al., 2020). It was not
the development of later-life MDD (Tunnard et al., 2014); how- only a central region in the limbic circuit responsible for negative
ever, the precise mechanism by which CT leads to MDD remains emotion processing (Morawetz et al., 2020) but also an impor-
unclear. CT putting individuals at risk for mental health prob- tant area in the reward system involved in the positive emotion
lems is probably due to its effect on the developing brain, as processing (Murray, 2007). Hippocampus plays an important role
during childhood, the brain is immature, and the brain struc- in the processing and regulation of threat and fear and repre-
ture and function might be modified by the traumatic events to sents a vital position in the stress-related pathology (Mahar et al.,
a relatively great extent (Glaser, 2014). Hence, the exploration 2014). Previous studies have revealed that CT might have effects
of the effect of CT on brain associated with depression might on structure and function of both amygdala and hippocampus
be a potential strategy to reveal the mechanisms linking CT (Dannlowski et al., 2012; Herringa et al., 2013; Hanson et al., 2015;
and MDD. Bounoua et al., 2020; Nogovitsyn et al., 2020). These findings indi-
Both amygdala and hippocampus are important areas that cate that the connectivity of amygdala and hippocampus might
have been centrally implicated in MDD (Cullen et al., 2014; Mahar play a quite important role in the neural mechanisms via which
et al., 2014). Amygdala has wide functional connectivity (FC) with CT leads to depression.
Resting-state functional magnetic resonance imaging (fMRI) with no or low CT group classification (see details in the ‘Clini-
is a popular and reliable tool for probing neural FC based on cal assessments’ section). All participants were adults (≥18 years
a certain brain region. Some studies have explored the effect old) and right-handed. A subset of this data set has been previ-
of CT on amygdala and hippocampus FC. For example, previ- ously published (Fan et al., 2020), in which we investigated the
ous studies have found that a high level of CT was associated effects of CT on anhedonia in depression via reward system. This
with a decreased FC between hippocampus and subgenual ante- data set mainly enlarged the sample size of MDD with no or low
rior cingulate cortex (sgACC) and between amygdala and ventral CT and HC with moderate-to-severe CT to increase the statistical
medial prefrontal cortex, sgACC, insula, and posterior cingulate power. The main results of the previous study were replicated with
and increased connectivity between amygdala and dorsal pre- this data set, which we have detailed in the Supplement (Analysis
frontal cortex, hippocampus, parahippocampal gyrus, inferior of striatum-based FC, Tables S1–S3, Figure S1). The study was
temporal gyrus, orbitofrontal cortex, cerebellum, and brainstem approved by the Ethics Committee of the Second Xiangya Hospi-
(Herringa et al., 2013; Birn et al., 2014; Dean et al., 2014; Cancel tal of Central South University, and all the participants provided
et al., 2017). These findings were revealed in samples with dif- written informed consent.
ferent mental illness diagnosis, such as samples with posttrau-
regressing out of nuisance covariates (six head motion param- controlled as covariates when investigating the effect of diagno-
eters, white matter signal, cerebrospinal fluid signal and global sis. The recent stress level as indicated by PSS scores was addi-
signal) and temporal band-pass filtering (0.01–0.08 Hz). We also tionally controlled when investigating the effects involving CT.
calculated the frame-wise displacement (FD) and scrubbed any Significance threshold for imaging analysis was set at cluster-
frame with FD >0.5 mm (Power et al., 2012). level family-wise error (FWE)–corrected P < 0.05, starting from
an uncorrected voxel-level P-value of 0.001. After establishing
Seed definition and FC analysis CT’s effects on clinical variables and FC, correlation analyses and
mediation analysis were carried out to examine the relationships
After preprocessing, the seed-based whole-brain FC was calcu-
among CT as well as dimensions of neglect and abuse, abnormal
lated. Seeds were located in the bilateral amygdala and bilat-
FCs and depression, anhedonia and anxiety levels in MDD and HC
eral hippocampus. These seeds were defined as spheres with
separately.
a 4 mm radius centered on the following coordinates (Montreal
Neurological Institute): left amygdala (−23, −6, −20), right amyg-
dala (21, −6, −20), left hippocampus (−31, −25, −11) and right Results
Demographic and clinical variables
Table 1. Demographic and clinical characteristics of MDD with moderate-to-severe CT, MDD with no or low CT, HC with moderate-to-
severe CT and HC with no or low CT groups
Statistics
MDD with moderate- MDD with no or HC with moderate-to- HC with no or low
Characteristics to-severe CT (N = 42) low CT (N = 31) severe CT (N = 30) CT (N = 34) F/t/𝜒2 P
Age (years) 23.83 (5.91) 22.97 (5.85) 20.47 (2.71) 20.82 (2.73) 4.32 0.006
Gender (female, %) 23 (54.76) 23 (74.19) 16 (53.33) 17 (50.0) 4.66 0.198
Verbal intelligence 47.52 (7.48) 45.48 (8.22) 49.13 (7.55) 47.68 (7.55) 1.23 0.300
Duration (weeks) 23.31 (5.88) 22.13 (5.93) – – 0.83 0.407
Age onset 22.24 (5.83) 21.20 (5.92) – – 0.74 0.462
CTQ total 54.79 (11.08) 35.94 (4.57) 47.27 (10.15) 32.76 (4.42) 54.75 <0.001
Physical neglect 11.79 (3.05) 7.48 (1.46) 11.43 (3.28) 6.76 (1.40) 37.95 <0.001
Physical abuse 8.07 (3.22) 5.32 (0.83) 6.73 (2.84) 5.42 (0.97) 11.57 <0.001
Emotional neglect 18.12 (3.96) 9.94 (3.03) 14.00 (4.48) 8.91 (2.70) 50.08 <0.001
Emotional abuse 10.93 (4.48) 7.45 (2.28) 8.73 (3.20) 6.58 (1.58) 13.14 <0.001
Sexual abuse 5.88 (1.66) 5.74 (1.23) 6.37 (1.96) 5.18 (0.39) 3.63 0.015
Neglect 29.90 (5.42) 17.42 (3.58) 25.43 (5.89) 15.56 (3.59) 73.36 <0.001
Abuse 24.88 (7.72) 18.52 (2.66) 21.83 (6.37) 17.21 (1.98) 14.92 <0.001
BDI 30.79 (9.63) 25.26 (9.46) 7.70 (4.88) 5.29 (4.50) 97.02 <0.001
RSAS 22.43 (7.19) 18.35 (7.00) 13.33 (4. 17) 8.91 (4.42) 35.78 <0.001
PAS 31.62 (11.44) 26.13 (12.27) 18.30 (6.35) 13.91 (7.12) 24.02 <0.001
SAI 60.05 (12.59) 54.94 (9.92) 39.67 (8.13) 35.68 (8.75) 47.29 <0.001
TAI 63.26 (8.85) 58.68 (7.15) 43.87 (7.09) 39.53 (9.43) 67.80 <0.001
PSS 27.50 (6.52) 25.29 (6.63) 16.00 (5.36) 13.26 (4.94) 48.39 <0.001
FD 0.07 (0.03) 0.07 (0.04) 0.08 (0.02) 0.08 (0.03) 0.43 0.735
Frames censored (%) 0.04 (0.07) 0.01 (0.01) 0.02 (0.04) 0.02 (0.04) 2.68 0.050
CT. Two patient groups had similar age onset and illness duration. right hippocampus–generated FCs. MDD patients had increased
There were no significant differences in head motion parame- right amygdala–left middle frontal gyrus (MFG) FC, decreased
ters as indicated by mean FD and percentage of censored frames right amygdala–right medial superior frontal gyrus (mSFG) FC
among the four groups (Table 1). and decreased right hippocampus–bilateral mSFG FC than HC
(P < 0.05 FWEcorr-cluster) (Table 2, Figure 1B–D). No significant
Effects of CT on FC with seeds in amygdala and diagnosis × CT interactions for all four seeds were found.
hippocampus
Significant main effects of CT were detected on right amygdala– Correlation and mediation analyses
generated FCs. Participants with moderate-to-severe CT showed Results showed that right amygdala–right precuneus FC was
decreased right amygdala–right precuneus FC than those with no significantly correlated with trait social anhedonia (r = −0.42,
or low CT (P < 0.05 FWEcorr-cluster) (Table 2, Figure 1A). Significant P < 0.001) and trait physical anhedonia (r = −0.36, P = 0.002)
main effects of diagnosis were revealed on right amygdala– and (Figure 1E), while it was not correlated with depression (r = −0.10,
Table 2. Diagnosis and CT effects on functional connectivity with seed located in the amygdala and hippocampus
BA, Broadmann area; x, y, z, coordinates of peak locations in the Montreal Neurological Institute space (MNI). P < 0.05, cluster-level FWE corrected with voxel-level
starting from P < 0.001 uncorrected.
Fig. 1. Effect of CT and diagnosis on amygdala and hippocampus FC and the association with trait anhedonia. Image analysis revealed that
participants with moderate-to-severe CT, irrespective of MDD diagnosis, had decreased FC between right amygdala seed and right precuneus
compared to those with no or low CT (A). MDD patients generally had increased FC between right amygdala seed and left MFG (B) and decreased FC
between right amygdala seed and right mSFG (C) and between right hippocampus and bilateral mSFG relative to HC (D). The right amygdala–right
precuneus connectivity was significantly correlated with physical and social trait anhedonia in MDD (E). Significance threshold for image analyses was
set at P < 0.05, FWE cluster-level corrected, starting from voxel-level P < 0.001 uncorrected.
J. Fan et al. 5
P = 0.399) and anxiety (SAI: r = −0.22, P = 0.060; TAI: r = −0.13, levels in MDD. The main findings were as follows: (i) individu-
P = 0.268) in MDD patients (Table S4). The mediation analy- als with moderate-to-severe CT, irrespective of MDD diagnosis,
sis revealed that the mediation effect of right amygdala–right had decreased right amygdala–right precuneus FC than those
precuneus FC on the relationship between CT and physical with no or low level of CT; (ii) decreased right amygdala–right
trait anhedonia [𝛽 = 0.08, bootstrapped 95% confidence inter- precuneus FC was significantly correlated with social and phys-
val (CI) = 0.0132–0.1780, P = 0.022] and between CT and social ical trait anhedonia in MDD, and the mediation effects of this
trait anhedonia (𝛽 = 0.09, bootstrapped 95% CI = 0.0110–0.1120, FC on relationship between CT, especially the neglect dimension
P = 0.017) in MDD was both significant. After controlling for while not the abuse, and trait anhedonia were significant; and (iii)
FC mediation effects, the direct effect of CT on physical trait MDD patients generally had increased right amygdala–left MFG
anhedonia (𝛽 = 0.03, bootstrapped 95% CI = −0.1827 to 0.2430, FC, decreased right amygdala–right mSFG FC and decreased right
P = 0.825) and on social trait anhedonia (𝛽 = 0.04, bootstrapped hippocampus–bilateral mSFG FC compared to HC. These findings
95% CI = −0.0436 to 0.2065, P = 0.225) was not significant (Table 3). highlight the effect of CT on right amygdala–right precuneus FC,
Correlation and mediation analyses further conducted to distin- suggesting that this FC and the associated trait anhedonia might
guish the different effects of neglect and abuse dimensions of CT be manifested as important mechanisms involved in depression
Table 3. Mediation effects of right amygdala–right precuneus FC on the relationship between CT as well as its dimensions (neglect vs.
abuse) and anhedonia in MDD patients
95% CI 95% CI
Effect 𝛽 Boot LLCI Boot ULCI P Effect 𝛽 Boot LLCI Boot ULCI P
Significant indirect effects are highlighted in bold. confidence interval (CI); lower limit of confidence interval (LLCI); upper limit of confidence interval (ULCI).
Fig. 2. Mediation effects of right amygdala and right precuneus connectivity on relationship between neglect and physical and social trait
anhedonia. (A) Mediation effect of the right amygdala and right precuneus FC on the relationship between neglect and physical trait anhedonia was
significant (𝛽 = 0.09, bootstrapped 95% CI = 0.0273–0.3080). (B) Mediation effect of the right amygdala and right precuneus FC on the relationship
between neglect and social trait anhedonia was significant (𝛽 = 0.10, bootstrapped 95% CI = 0.0231–0.2024). After controlling for indirect effects, the
direct effect of CT on trait anhedonia was not significant. Mediation analyses were generated by using the bootstrap method from 5000 bootstrapped
samples. *P < 0.05; **P < 0.01; ***P < 0.001.
6 Social Cognitive and Affective Neuroscience, 2023, Vol. 00, No. 00
Precuneus was a heterogeneous brain region, of which the dif- previous studies (van Harmelen et al., 2010; Chaney et al., 2014;
ferent subregions may have different connectivity patterns and be Opel et al., 2014; Lu et al., 2019; Meinert et al., 2019), indicating
involved in distinct cognitive functions such as visuospatial pro- that the neural alterations following CT may be a vulnerabil-
cessing (Nicole et al., 2005), mental imagery (Cavanna and Trimble, ity to future psychiatric conditions, and some brain alterations
2006), retrieval of episodic information (Dorfel ̈ et al., 2009), reflec- previously attributed to MDD diagnosis might rather be character-
tive self-related processing (Lou et al., 2004) and awareness and ized as a function of childhood maltreatment. We failed to detect
conscious information processing (Vogt and Laureys, 2005). The the significant correlation between this FC and trait anhedonia
precuneus identified in the present study was located in the in HC. The result of HC generally having lower trait anhedonia
dorsal-posterior subregion. It was proposed that this subregion as compared with MDD might help explain this phenomenon.
was more likely to be involved in the visuospatial information Further longitudinal studies are needed to examine the specific
processing and motor imagery (Zhang and Li, 2012). FC between relationship between brain and behavioral alterations in HC with
amygdala and this precuneus subregion has been revealed in moderate-to-severe CT.
previous task fMRI studies and suggested to be involved in emo- Besides the effects of CT, our analysis also revealed that MDD
tion regulation through directing attention toward or away from patients generally had increased right amygdala–left MFG (dor-
Liu, W., Ge, T., Leng, Y., et al. (2017). The role of neural plastic- Nicole, W., Filiep, D., Stefan, S., Swinnen, S.P. (2005). The role of ante-
ity in depression: from hippocampus to prefrontal cortex. Neural rior cingulate cortex and precuneus in the coordination of motor
Plasticity, 2017, 6871089. behaviour. European Journal of Neuroscience, 22, 235–46.
Lou, H.C., Bruce, L., Michael, C., et al. (2004). Parietal cortex and rep- Nogovitsyn, N., Addington, J., Souza, R., et al. (2020). Childhood
resentation of the mental self. Proceedings of the National Academy trauma and amygdala nuclei volumes in youth at risk for mental
of Sciences U S A, 101, 6827–32. illness. Psychological Medicine, 1–8.
Lu, S., Xu, R., Cao, J., et al. (2019). The left dorsolateral prefrontal Ochsner, K.N., Silvers, J.A., Buhle, J.T. (2012). Functional imaging stud-
cortex volume is reduced in adults reporting childhood trauma ies of emotion regulation: a synthetic review and evolving model
independent of depression diagnosis. Journal of Psychiatric Research, of the cognitive control of emotion. Annals of the New York Academy
112, 12–7. of Sciences, 1251, E1–24.
Luking, K.R., Repovs, G., Belden, A.C., et al. (2011). Functional con- Opel, N., Redlich, R., Zwanzger, P., et al. (2014). Hippocampal atro-
nectivity of the amygdala in early-childhood-onset depression. phy in major depression: a function of childhood maltreatment
Journal of the American Academy of Child and Adolescent Psychiatry, rather than diagnosis? Neuropsychopharmacology, 39, 2723–31.
50, 1027–41.e3. Power, J.D., Barnes, K.A., Snyder, A.Z., Schlaggar, B.L., Petersen, S.E.