Volume 73, No.12: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
786
Wongsawaeng et al.
Doonyaporn Wongsawaeng, M.D.*, Orasa Chawalparit, M.D.*, Siriwan Piyapittayanan, M.D.*, Tanyaluck
ientunyakit, M.D.*, Weerasak Muangpaisan, M.D.**, Kitikan ana-udom, M.D. ***, Panida Charnchaowanish,
B.Sc*, Chanon Ngamsombat, M.D.*
*Department of Radiology, **Department of Preventive and Social Medicine, ***Department of Psychiatry, Faculty of Medicine Siriraj Hospital,
Mahidol University, Bangkok 10700, ailand.
Magnetic Resonance Hippocampal Subeld
Volumetric Analysis for Differentiating among
Healthy Older Adults and Older Adults with Mild
Cognitive Impairment or Major Depressive Disorder
ABSTRACT
Objective: Depression among older adults is frequently an early symptom of cognitive decline, and is believed to
be a risk factor for Alzheimer’s disease (AD). Hippocampal subeld volume loss is found in both mild cognitive
impairment (MCI) and major depressive disorder (MDD). We aimed to investigate the potential of MR hippocampal
subeld volumetry for discriminating among healthy older adults (HOA) and older adults with MCI or MDD.
Materials and Methods: Seventy age-matched subjects (29 non-depressed MCI, 12 MDD, and 29 HOA) underwent
3-Tesla MR imaging (MRI) with high-resolution 3D-T1W-TFE whole brain. Hippocampal subeld volumetric
measurements were performed using FreeSurfer soware to distinguish among MCI, MDD, and HOA. Subgroup
analysis with amyloid PET result was also performed.
Results: Signicantly smaller bilateral hippocampal tail volume was observed in MCI compared to HOA (p=0.004
and p=0.04 on the le and right side, respectively). e same comparative nding was observed at le HATA
(hippocampus-amygdala-transition-area) of MCI (p=0.046). Other regions showed non-signicantly smaller size
in MCI than in HOA [le molecular layer HP (p=0.06), le whole hippocampus (p=0.06), and le CA1 (p=0.07)].
ere was a non-signicant trend toward smaller size in almost all 13 subeld hippocampal regions of MCI compared
to MDD, even in subgroup analysis with amyloid PET result.
Conclusion: MR hippocampal subeld volumetry may have value in routine clinical practice for screening individuals
with MCI, and may be a valuable adjunct to amyloid PET study for very early-stage diagnosis of AD.
Keywords: Magnetic resonance hippocampal subeld volumetric analysis, mild cognitive impairment (MCI), major
depressive disorder (MDD), healthy older adults (HOA) (Siriraj Med J 2021; 73: 786-792)
Corresponding author: Chanon Ngamsombat
E-mail: ngamsombatc@gmail.com
Received 9 March 2021 Revised 29 June 2021 Accepted 1 July 2021
ORCID ID: https://orcid.org/0000-0001-5055-0711
http://dx.doi.org/10.33192/Smj.2021.102
INTRODUCTION
Mild cognitive impairment (MCI) is diagnosed
when people have measurable changes in thinking ability
noticed by the person affected, family members, or
friends even though the observed impairment does
not aect the individual’s activities of daily living.
1
e
2011 recommendations from the National Institute on
Aging-Alzheimer’s Association diagnostic guideline for
Volume 73, No.12: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
787
Original Article
SMJ
Alzheimer’s disease (AD) working groups suggest that
some MCI cases reect the early stage of AD.
2
Depression,
especially in older adults, frequently develops concomitantly
with cognitive impairment, and it may be a psychological
reaction or a risk factor for AD.
3
One of the most mentioned structures in limbic
system is hippocampus, which is known to involve in both
neurodegenerative disease, especially AD, and emotional
regulation.
4
Hippocampal atrophy is usually detected
in late stage of AD. Previous study found that subeld
hippocampal atrophy evidenced by magnetic resonance
imaging (MRI) might be helpful for early detection of
mild cognitive impairment who have converted to AD
(MCI-c).
5
Concerning mood regulation, a previous meta-
analysis found more hippocampal volume loss in MDD
than in the control; however, the impact of illness on
hippocampal volume is probably related to duration
and severity.
6
To date, no study has compared subeld hippocampal
volume between MCI and MDD in older adults. Accordingly,
the aim of this study was to investigate the potential of
MR hippocampal subeld volumetry for discriminating
among older adults with non-depressed MCI, older
adults with treatment-naïve MDD, and healthy older
adults (HOA).
MATERIALS AND METHODS
Study population
is retrospective study reviewed the MRI DICOM
les, clinical information, and neuropsychological test
results of 72 subjects (30 MCI, 12 MDD, and 30 HOA)
who were recruited at a single national tertiary referral
center in ailand during January 2016 to September
2020. e protocol for this study was approved by the
Siriraj Institutional Review Board (SIRB) of the Faculty of
Medicine Siriraj Hospital, Mahidol University, Bangkok,
ailand (Si 1037/2020).
e 30 MCI and 30 HOA subjects, recruited from
neurology and geriatric clinics at our center, were part
from the SIRB-approved study (Si 137/2015). Clinical
evaluation of MCI and HOA subjects was performed by
a senior geriatric neurologist (WM) who specializes in
dementia.
e 12 MDD subjects, rst-diagnosed treatment-
naïve patients, recruited from the psychiatric clinic at
our center, were part from a dierent SIRB-approved
study (Si 239/2016). Diagnosis and severity of depression
were determined by a board-certied psychiatrist.
Two out of 72 subjects (1 MCI and 1 HOA) were
excluded due to aws in their MRI DICOM les. e
remaining 70 subjects (29 MCI, 12 MDD, and 29 HOA)
were included and analyzed. e amyloid PET result for all
of the 29 MCI patients were recorded and subcategorized
as PET positive MCI (PET+ve MCI; n=12) or PET negative
MCI (PET-ve MCI; n=17) patients. Age, gender, education
level, ai Mental State Examination (TMSE)
7
, Clinical
Dementia Rating Scale (CDR), and Hamilton Rating
Scale for Depression (HAM-D)
8
were also collected and
recorded. Two years of clinical follow-up among the
29 MCI subjects was achieved by the end of September
2020.
Operational denitions
1. Criteria for mild cognitive impairment (MCI)
1) Age equal to or greater than 60 years
2) Subjective memory complaint by the patient,
family member, or clinician with preserved activities of
daily living (ADL)
3) CDR score of 0.5
4) Absence of dementia by National Institute of
Neurological and Communicative Disorders and Stroke
and the Alzheimer’s Disease and Related Disorders
Association (NINCDS-ADRDA) criteria
5) TMSE score from 24 to 30
6) No history of depressive symptom
2. Criteria for major depressive disorder (MDD)
1) Age equal to or greater than 60 years
2) First diagnosed approaching fulllment of the
Diagnostic and Statistical Manual of Mental Disorders,
5
th
Edition (DSM-5) criteria for MDD
9
3) Depression severity was measured by HAM-D
4) TMSE score from 24 and 30
5) No other psychiatric disorders, antidepressant
drug use, currently unstable medical or neurological
condition
3. Criteria for healthy older adults (HOA)
1) Age equal to or greater than 60 years
2) TMSE score from 24 to 30
3) CDR score of 0
4) No neurological or psychiatric illness, non-
demented, and normal ADL
Magnetic resonance imaging (MRI) acquisition
All 70 enrolled subjects underwent 3T MRI scans
(Ingenia, Philips Medical System, Best, the Netherlands)
with a 32-channel head coil. e MRI protocol included
a 3D high-resolution T1W-TFE sequence covering whole
brain (eld-of-view (FOV) 230×230×172 mm
3
, matrix
size 352x352, voxel size 0.72×0.72×0.65 mm
3
, echo time
Volume 73, No.12: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
788
Wongsawaeng et al.
(TE)/repetition time (TR) 4.8/9.8 ms, ip angle 8°, scan
time 6 min). All MRI DICOM les were transferred to
hippocampal subeld segmentation process.
Hippocampal subeld segmentation
e FreeSurfer image analysis pipeline (version
6.0)
10
was used for automated hippocampal subeld
segmentation. e validated ultra-high resolution 13
subeld hippocampal regions (Fig 1) were, as follows:
presubiculum, subiculum, parasubiculum, cornu ammonis
(CA)1, CA2/3, CA4, molecular layer hippocampus (HP),
GC-ML-DG (granule cell layer and molecular layer
of dentate gyrus), HATA (hippocampus-amygdala-
transition-area), hippocampal tail, mbria, hippocampal
ssure, and the whole hippocampus, bilaterally.
e raw volume data each of subeld was displayed
and then normalized according to each subject’s intracranial
volume (ICV) derived from FreeSurfer soware via this
following formula: volume normalized = volume raw
data x 1,000/ICV in cm
3
.
10,11
[18F] Florbetapir PET/CT to detect cerebral amyloid
deposition
All 29 MCI patients also underwent amyloid positron
emission tomography (PET) study with administration
of our proprietary [18F] orbetapir biomarker
12
shortly
before or aer MRI scan. Our specic PET/CT scan
(Discovery STE; GE Healthcare, Chicago, IL, USA)
acquisition and image protocols are described in ADNI
GO
13
and ADNI 2.
14
In the present study, two nuclear
medicine physicians who were blinded to patient clinical
information reached a consensus decision regarding who
was amyloid positive and who was amyloid negative
according to the published criteria.
15
(Fig 2)
Fig 2. e transaxial images of [F-18] orbetapir PET brain study in two dierent patients with mild cognitive impairment (MCI) showing
positive brain amyloid deposition due to mildly increased radiotracer uptake at bilateral temporal cerebral cortices (A), and negative amyloid
deposition due to clear gray-white matter discrimination without abnormal cortical uptake (B).
Fig 1. Coronal view MRI bilateral hippocampi of a 72-year-old male with mild cognitive impairment (amyloid PET positive) (A), and a
68-year-old male with rst diagnosis treatment-naïve MDD (B) shown in T1-weighted image (le), and T1-weighted image with subeld
hippocampal segmentations (right).
Volume 73, No.12: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
789
Original Article
SMJ
Statistical analysis
All statistical analyses were performed using SPSS
Statistics version 18.0 (SPSS, Inc., Chicago, IL, USA).
Continuous variables were analyzed by analysis of variance
(ANOVA) with Bonferroni post hoc comparisons, and
the categorical variables were analyzed by chi-square
test. A p-value <0.05 was considered to be statistically
signicant.
RESULTS
1. Demographics, clinical and neuropsychological data
Seventy age-matched subjects were included in
this study (MCI=29, MDD=12, HOA=29). e mean
± SD age of these 3 groups was 68.1±4.3, 70.8±6.0, and
68.7±4.8 years, respectively. As expected, there were no
statistically signicant dierences in TMSE score among
the 3 study groups (Table 1). Six of the 29 MCI patients
had clinically proven AD-converted MCI by the end of
the 2-year follow-up, and all 6 of those patients had an
initial amyloid PET result that was positive.
2. Hippocampal subelds
2.1 Comparison between MCI and HOA (Table 2)
e bilateral hippocampal tails showed signicantly
smaller volume in the MCI group compared to the HOA
group (p=0.004 and p=0.04 on the le and right side,
respectively), as well as at the le HATA (hippocampus-
amygdala-transition-area) (p=0.046). We also observed
a trend towards signicantly smaller size in the MCI
group compared to the HOA group for le molecular
layer HP (p=0.06), le whole hippocampus (p=0.06),
and le CA1 (p=0.07).
2.2 Comparison between MCI and MDD (Tables
2, 3)
ere was a non-signicant trend toward smaller
size in almost all of the 13 subeld hippocampal regions
when compared between MCI and MDD subjects – even
in subgroup analysis (MCI PET+ve and MCI PET-ve).
2.3 Comparison between MDD and HOA (Table 2)
ere was no signicant dierence between the
MDD and HOA groups for any subeld hippocampal
regions.
2.4 Comparison between MCI PET+ve and HOA
(Table 3)
In subgroup analysis combined with amyloid
PET result, we found that the bilateral hippocampal
tails showed a signicantly smaller volume in the MCI
PET+ve group than in the HOA group (p=0.002 and
p=0.02 on the le and right side, respectively). e le
whole hippocampus (p=0.05), le molecular layer HP
(p=0.07), and le subiculum (p=0.07) all demonstrated
smaller volume among MCI PET+ve subjects compared
to HOA subjects.
2.5 Comparison between MCI PET-ve and HOA
(Table 3)
No statistically signicant dierence in hippocampal
subeld volumes was observed between these two groups.
TABLE 1. Demographic, clinical and neuropsychological data of MCI, MDD, and HOA subjects.
Subject data
MCI MDD HOA
(n=29) (n=12) (n=29)
p
Gender (male/female), n 15/14 5/7 10/19 0.41
Age (years), (mean±SD) 68.1±4.3 70.8±6.0 68.7±4.8 0.26
Education, n (%) <0.0001
- High school or lower 2 (6.9%) 8 (66.7%) 16 (55.2%)
- Higher than high school 27 (93.1%) 4 (33.3%) 13 (44.8%)
TMSE (mean±SD) 27.3±1.6 26.8±2.0 27.9±1.9 0.17
HAM-D (mean±SD) NA 24.5±4.3 NA NA
A p-value<0.05 indicates statistical signicance
Abbreviations: MCI, mild cognitive impairment; MDD, major depressive disorder; HOA, healthy older adults; SD, standard deviation;
TMSE, ai Mental State Examination, HAM-D; Hamilton Rating Scale for Depression; NA, not applicable
Volume 73, No.12: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
790
Wongsawaeng et al.
TABLE 2. Normalized hippocampal subeld volume compared among MCI, MDD, and HOA subjects.
TABLE 3. Hippocampal subeld volume compared among MCI PET +ve, MCI PET -ve, MDD, and HOA subjects.
Parameters MCI (n=29) MDD (n=12) HOA (n=29) P (MCI vs HOA)
Left hippocampal tail 323.4±67.2 342.4±70.2 383.0±67.5 0.004
a
Right hippocampal tail 345.3±78.2 368.0±54.0 392.4±70.1 0.04
a
Left HATA 37.4±7.5 43.7±13.9 42.3±7.4 0.046
a
Left molecular layer HP 352.1±67.7 385.6±88.0 394.1±58.2 0.06
Left whole hippocampus 2,151.5±376.2 2,338.6±521.6 2,396.6±337.1 0.06
Left CA1 394.4±79.9 435.0±88.3 440.8±68.5 0.07
a
Statistically signicant dierence (p<0.05) between the MCI and HOA groups
Abbreviations: MCI, mild cognitive impairment; MDD, major depressive disorder; HOA, healthy older adults; HATA, hippocampus-
amygdala-transition-area; HP, hippocampus; CA1, cornu ammonis 1
MCI PET+ve MCI PET-ve MDD HOA
P (MCI
Parameters
(n=12) (n=17) (n=12) (n=29)
PET+ve
vs HOA)
Left hippocampal tail 297.5±79.1 341.7±52.4 342.4±70.2 383.0±67.5 0.002
a
Right hippocampal tail 318.2±87.5 364.4±67.0 368.0±54.0 392.4±70.1 0.02
a
Left whole hippocampus 2,035.4±380.4 2,233.5±361.8 2,338.6±521.6 2,396.6±337.1 0.05
Left molecular layer HP 334.3±66.8 364.6±67.5 385.6±88.0 394.1±58.2 0.07
Left subiculum 261.9±56.5 292.6±50.6 310.1±78.2 311.6±48.4 0.07
a
Statistically signicant dierence (p<0.05) between the MCI PET+ve and HOA groups
Abbreviations: MCI, mild cognitive impairment; MDD, major depressive disorder; HOA, healthy older adults; HP, hippocampus
DISCUSSION
Interestingly, the signicantly smaller volume of
the bilateral hippocampal tails in the MCI compared
to the HOA group, as well as in subgroup analysis, was
observed in the MCI PET+ve, but not in the MCI PET-
ve group. Previous study
16,17
reported some dierences
in functionality between the ventral (anterior) and the
dorsal (posterior) hippocampus in which the posterior
part primarily performs cognitive functions, such as
learning and memory, whereas the anterior part is more
related to stress and emotion. Our 2-year clinical follow-
up data showed that about 20% of our MCI patients (6/29
subjects) converted to clinically diagnosed Alzheimer’s
disease (AD). More importantly, all 6 of those AD-
converted MCI patients (MCI-c) also had an initial
amyloid PET result that was positive. We propose that
the structural change of the hippocampus demonstrated
by MRI volumetric analysis, especially the small size of
the hippocampal tail, might be a predictor of conversion
to AD among MCI patients.
e relatively smaller volume of the le molecular layer
HP, le CA1, le subiculum, and le whole hippocampus
in the MCI group (especially MCI PET+ve) compared
to HOA subjects suggests that other hippocampal
Volume 73, No.12: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
791
Original Article
SMJ
subeld regions might also be aected in the early stage
of neurodegenerative disease. Scharfman, et al. reported
that neurons in the entorhinal cortex, especially the
supercial layer, were believed to be particularly vulnerable
to adverse eect in the early stage of Alzheimer’s disease
(AD)
18
and have been found interconnecting to axons
within the hippocampal formation.
From neuroanatomy, the subiculum is the grey
structure that is located above the parahippocampal
gyrus, which is part of the entorhinal cortex, and it
superolaterally connects to the CA1 region. We postulate
that the change in the entorhinal cortex in early AD
might also propagate eect to the subiculum and CA1,
as well as to the molecular layer HP adhering to both
subiculum and CA1.
In older adults, depression oen develops concomitantly
with cognitive impairment. is is likely a psychological
reaction to cognitive decline, so it may manifest as an
early symptom in early-stage dementia patients. However,
recent data suggests that depression, particularly late-life
depression, can also be a risk factor for AD.
3
Two prior studies
19,20
reported signicant volume
change in some subeld hippocampal regions in MDD
patients with some specic conditions, such as recurrent
episode of depressive symptom (decreased volume as the
number of prior episodes increased)
19
, or continuous
remission of drug-naive disease (increased volume in
MDD patients who were in remission at least 6 months).
Concerning our result, there was no statistically signicant
dierence in volumetric analysis compared between
rst-diagnosed and untreated MDD and either MCI or
HOA subjects. is may suggest that the hippocampus
has some plasticity, especially relative to volumetric
change in depressive condition, but not in early or late-
stage neurodegenerative disease, which known to be
associated with progressive permanent neuronal loss.
Strengths and limitations
e strengths of our study were: 1) Clinical evaluation
of MCI and HOA subjects was performed by a senior
geriatric neurologist (WM) who specializes in dementia;
2) Amyloid-PET result was available for all MCI patients;
and, 3) All MDD patients had rst-diagnosed and untreated
status without any confounding factors, such as repeated
episode of depressive symptom or treatment-related
issues.
Limitations of the present study include 1) A lack
of data specic to depressive illness duration, which may
aect hippocampal subeld volume change as found
from prior study
21
; 2) e fact that our MDD patients
had only mild to moderate depressive severity, which
may not clearly demonstrate alteration of hippocampal
volume; 3) Our study’s single-center retrospective design;
and, 4) our overall small size and small group sample
sizes may have limited the statistical power of our study
to identify all signicant dierences between and among
groups.
CONCLUSION
MR hippocampal subeld volumetry may have
value in routine clinical practice for screening individuals
with MCI, and may be a valuable adjunct to amyloid
PET study for very early-stage diagnosis of AD. Future
study in subeld hippocampal volumetry compared
between MCI patients with and without codepressive
symptoms will further clarify the inuence of depression
on hippocampal atrophy, especially in some specic
subeld regions. is information will improve our
understanding of the underlying pathophysiology, and
will help us to better guide disease management in the
future.
ACKNOWLEDGEMENTS
e authors gratefully acknowledge Dr. Orawan
Supapueng for assistance with statistical analysis and
Mrs. Angkana Jongsawaddipatana for assistance with
data collection.
Conict of interest declaration: All authors declare
no personal or professional conicts of interest, and no
nancial support from the companies that produce and/
or distribute the drugs, devices, or materials described
in this report.
Funding disclosure: DW, OC, SP, TT, WM, KT, and
CN were each supported by a Chalermprakiat Grant
from the Faculty of Medicine Siriraj Hospital, Mahidol
University, Bangkok, ailand.
REFERENCES
1. Alzheimer’s Association. 2014 Alzheimer’s disease facts and
gures. Alzheimers Dement 2014;10(2):e47-92.
2. Albert MS, DeKosky ST, Dickson D, Dubois B, Feldman HH,
Fox NC, et al. e diagnosis of mild cognitive impairment due
to Alzheimer’s disease: Recommendations from the National
Institute on Aging-Alzheimer’s Association workgroups on
diagnostic guidelines for Alzheimer’s disease. Alzheimer Dement.
2011;7(3):270-9.
3. Korczyn AD, Halperin I. Depression and dementia. J Neurol
Sci. 2009;283(1-2):139-42.
4. Palazidou E. e neurobiology of depression. Vol. 101, British
Medical Bulletin. 2012. p.127-45.
5. Khan W, Westman E, Jones N, Wahlund LO, Mecocci P, Vellas B,
et al. Automated Hippocampal Subeld Measures as Predictors
Volume 73, No.12: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
792
of Conversion from Mild Cognitive Impairment to Alzheimer’s
Disease in Two Independent Cohorts. Brain Topogr. 2015;28(5):
746-59.
6. Mckinnon MC, Yucel K, Nazarov A, Macqueen GM, Mckinnon P.
Presented as a poster at the 58
th
Annual Conference of the
Canadian Psychiatric Assocation. J Psychiatry Neurosci. 2009;34:
7. Muangpaisan W, Md A, Sitthichai Bsc K, Richardson Phd K,
Brayne C. e Distribution of ai Mental State Examination
Scores among Non-Demented Elderly in Suburban Bangkok
Metropolitan and Associated Factors. J Med Assoc ai. 2015;
98(9):916-24.
8. Lotrakul M, Sukanich P, Sukying C. e Reliability and Validity
of ai version of Hamilton Rating Scale for Depression. J
Psychiat Assoc ai. 1996;41(4):235-46.
9. American Psychiatric Association. Diagnostic and Statistical
Manual of Mental Disorders, Fih Edition. American Psychiatric
Association; 2013.
10. Iglesias JE, Augustinack JC, Nguyen K, Player CM, Player A,
Wright M, et al. A computational atlas of the hippocampal
formation using ex vivo, ultra-high resolution MRI: Application
to adaptive segmentation of in vivo MRI. Neuroimage. 2015;
115:117-37.
11. Westman E, Aguilar C, Muehlboeck JS, Simmons A. Regional
magnetic resonance imaging measures for multivariate analysis
in Alzheimer’s disease and mild cognitive impairment. Brain
Topogr. 2013;26(1):9-23.
12. ientunyakit T, Sethanandha C, Muangpaisan W, Chawalparit
O, Arunrungvichian K, Siriprapa T, et al. Implementation of
[18F]-labeled amyloid brain PET imaging biomarker in the
diagnosis of Alzheimer’s disease. Nucl Med Commun. 2018;
39(2):186-92.
13. ADNI-GO PET Technical Procedures Manual for FDG and
AV-45 ADNI-GO PET Technical Procedures Manual, 2011.
14. ADNI 2 PET Technical Procedures Manual for FDG and AV-
45 ADNI 2 PET Technical Procedures Manual AV-45 (Florbetapir
F 18) & FDG. 2011.
15. Minoshima S, Drzezga AE, Barthel H, Bohnen N, Djekidel M,
Lewis DH, et al. SNMMI procedure standard/EANM practice
guideline for amyloid PET imaging of the brain 1.0. J Nucl
Med. 2016;57(8):1316–22.
16. Fanselow MS, Dong HW. Are the Dorsal and Ventral Hippocampus
Functionally Distinct Structures? Neuron. 2010;65(1):7-19.
17. Samuels BA, Hen R. Neurogenesis and aective disorders. Eur
J Neurosci. 2011;33(6):1152-9.
18. Scharfman HE, Chao MV. e entorhinal cortex and neurotrophin
signaling in Alzheimer’s disease and other disorders. Cogn
Neurosci. 2013;4(3-4):123-35.
19. Samuels BA, Leonardo ED, Hen R. Hippocampal subelds and
major depressive disorder. Vol. 77, Biological Psychiatry.
Elsevier USA; 2015. p. 210-1.
20. Katsuki A, Watanabe K, Nguyen L, Otsuka Y, Igata R, Ikenouchi
A, et al. Structural changes in hippocampal subelds in patients
with continuous remission of drug-naive major depressive
disorder. Int J Mol Sci. 2020;21(9):3032.
21. Han KM, Kim A, Kang W, Kang Y, Kang J, Won E, et al.
Hippocampal subeld volumes in major depressive disorder
and bipolar disorder. Eur Psychiatry. 2019;57:70-77.
Wongsawaeng et al.
