Structural characteristics of amygdala subregions in type 2 diabetes mellitus

Type 2 diabetes mellitus (T2DM) patients often suffer from depressive symptoms, which seriously affect cooperation in treatment and nursing. The amygdala plays a significant role in depression. This study aims to explore the microstructural alterations of the amygdala in T2DM and to investigate the relationship between the alterations and depressive symptoms. Fifty T2DM and 50 healthy controls were included. Firstly, the volumes of subcortical regions and subregions of amygdala were calculated by FreeSurfer. Covariance analysis (ANCOVA) was conducted between the two groups with covariates of age, sex


Introduction
The prevalence of type 2 diabetes mellitus (T2DM) is steadily escalating globally, with projections indicating a substantial rise to encompass approximately 783.2 million by 2045 [1].Individuals diagnosed with T2DM exhibit a heightened susceptibility to depression in comparison to those without diabetes or impaired glucose metabolism [2].Furthermore, patients with T2DM are twice as likely to experience depression compared to the healthy population [3].However, the relationship between structural changes in the brain and depressive symptoms in T2DM patients remains unclear.
Subcortical structures play a pivotal role in human cognition, emotion, and learning [4].Previous researches have shown that the volumes of the caudate [5][6][7], hippocampus [6,8], thalamus [7,8], and amygdala [9,10] are reduced in individuals with T2DM.The atrophy of hippocampus and thalamus has been associated with cognitive impairments, such as memory loss [11] and decreased cognitive processing speed [7].Moreover, a functional magnetic resonance imaging (fMRI)-based study found that the right hippocampus and putamen exhibit a higher degree of centrality in patients with depressive T2DM compared to those with non-depressive T2DM [12].Some studies have also pointed out that reduced volume of amygdala in T2DM patients is associated with anxiety and depression [8].The amygdala, an essential subcortical region responsible for processing emotional stimuli according to animal studies [13], has been implicated in the onset of depression.However, the relationship between structure of subcortical nuclei and depressive symptoms in T2DM remains to be fully elucidated.
The amygdala is closely associated with depression.Some studies have suggested that the basolateral amygdala is engaged in the consolidation of episodic memories through interactions with the prefrontal cortex and the hippocampus, while the primary function of the central nucleus is to interact with the hypothalamus to regulate emotional and physiological responses under acute stress [14,15].Therefore, the structural and functional heterogeneity of amygdala subregions should be thoroughly considered.Some studies have shed light on the structural alterations in the basal nucleus, lateral nucleus, cortical nucleus, and cortico-amygdaloid transition area in depression.They also found that the volume of the bilateral cortico-amygdaloid transition area, left accessory basal nucleus, and left lateral nucleus showed negative correlation with depressive symptoms [16,17].Additionally, some resting-state functional magnetic resonance imaging (rs-fMRI) studies have found abnormal alterations in the amygdala of T2DM patients.Patients with T2DM combined with mild cognitive impairment showed changes in the amplitude of low-frequency fluctuation in the amygdala that were related to cognitive performance [18].Reduced nodal topological properties in the amygdala were found in an rs-fMRI-based study with graph-theoretic network analysis [19].Another study explored the relationship between neuronal activity and the corresponding cerebral blood perfusion in patients with T2DM and found disturbed neurovascular coupling in the amygdala [20].However, most previous studies on T2DM have ignored the rich morphological information in the subregions of amygdala and the relationship between subregions of amygdala and depressive symptoms.
In this study, we initially explored the microstructural characteristics of subcortical regions and subregions of the amygdala in individuals with T2DM.Furthermore, we conducted structural covariance analysis for amygdala subregions to reveal changes in coordination between subregions.Finally, correlation analysis was performed to explore the relationship between the amygdala subregions and depressive symptoms in individuals with T2DM.We try to reveal the structural alterations and morphological coordination in subcortical regions and the amygdala subregions in T2DM and to provide novel biomarkers for detecting brain impairment associated with depressive symptoms in individuals with T2DM.

Participants
Fifty patients diagnosed with T2DM were recruited from the Department of Endocrinology of the First Affiliated Hospital of Guangzhou University of Traditional Chinese Medicine from October 2021 to December 2022.This study was approved by the Medical Research Ethics Committee of the First Affiliated Hospital of Guangzhou University of Traditional Chinese Medicine (No. k [2020] 115), and all participants provided written informed consent.All T2DM patients met the latest standards of the American Diabetes Association (fasting blood glucose (FBG) ≥ 7.0 mmol/L, OGTT 2-h glucose ≥ 11.1 mmol/L, HbA1c ≥ 6.5 %, or random blood glucose ≥ 11.1 mmol/L) [21].Among those with clinical symptoms of hyperglycemia, all T2DM patients had a history of using oral antidiabetic drugs, insulin, or lifestyle measures.Additionally, fifty healthy controls (HC) with gender, age, and education matching T2DM were included, having undergone physical examinations within the past six months and having no history of diabetes.Exclusion criteria for all participants included: 1) age < 30years or > 70years; 2) presence of coronary heart disease, nephritis, tumor, gastrointestinal disease, or thyroid disease; 3) mental illness, including schizophrenia, bipolar disorder, or post-traumatic stress disorder; 4) contraindication of MRI; 5) obvious brain parenchyma lesion, including intracerebral hemorrhage, cerebral ischemic stroke or brain tumor; 6) left-handedness.

Clinical data
We collected basic information from all subjects, including gender, age, years of education, blood pressure (systolic and diastolic), and body mass index (BMI).Clinical data for diabetic subjects were obtained from the inpatient medical record system, including fasting blood glucose (FBG), fasting insulin (FIN), HbA1c, C-peptide, cholesterol (Cho), triglyceride (TG), and lactate dehydrogenase (LDH).The Homeostatic Model Assessment of Insulin Resistance (HOMA-IR) was calculated as：HOMA − IR = (FBG × FIN)/22.5.Depression symptoms of all subjects were estimated by the self-rating depression scale (SDS).

Magnetic resonance image data acquisition
All MRI scans were conducted using the same equipment to minimize variability.A 3 T MRI scanner (MAGNETOM Prisma, Siemens, Germany) equipped with a 64-channel head coil was used for experiments.Medical imaging specialists ruled out intracranial lesions using T1weighted imaging, T2-weighted imaging, and T2-fluid-attenuated inversion recovery.Three-dimensional T1-weighted imaging (3D-T1WI) was used in this study.More details of MRI scanning parameters are as follows: repetition time (TR) = 2530 ms, echo time (TE) = 2.98 ms, field of view (FOV) = 256 × 256 mm 2 , voxel size = 1 × 1 × 1 mm 3 , and number of slices = 192.

MRI processing
Three-dimensional T1-weighted images were analyzed by the reconall Freesurfer version 7.2.0 pipeline (http://surfer.nmr.mgh.harvard.edu/), including motion correction, skull stripping, field inhomogeneities correction, and registration in Talairach coordinate space.After pre-processing procedures, subcortical structure segmentation was performed, including thalamus, caudate, putamen, pallidum, hippocampus, amygdala, nucleus accumbens and ventral diencephalon (VentralDC).Subsequently, the amygdala subfield segmentation pipeline was applied on T1-weighted images based on a probabilistic atlas built with ultra-high-resolution ex vivo MRI data [22].Yielding 9 amygdala subfields, which include: anterior amygdaloid, cortico-amygdaloid transition area; basal, lateral, accessory basal, central, cortical medial, and paralaminar nuclei.subcortical gray matter structure and amygdala subregions of each participant for quality control，4 subjects with abnormal segmentation were excluded.

Statistical analyses
We conducted statistical analysis using SPSS 26.For normally distributed continuous variables, t-tests were utilized.Skewed continuous variables were assessed using the Mann-Whitney U-test, while categorical variables were analyzed using the chi-square test.Statistical significance was defined as P < 0.05.
Group comparisons between the T2DM and HC groups were conducted using ANCOVA to assess the volumes of eight subcortical regions and subregions of the amygdala, with age, sex, years of education, and estimated total intracranial volume (eTIV) as covariates.The Benjamini and Hochberg false discovery rate (FDR) correction was applied, and statistical significance was defined as P FDR < 0.05.Effect size was assessed using partial eta 2 , where a larger value indicates a greater difference between the means of the two groups.Critical values to distinguish between small, medium, and large effect sizes are 0.01, 0.06, and 0.14, respectively.MATLAB R2012b was used for structural covariate analysis of amygdala subregions.Partial correlation coefficients were calculated using Pearson's partial correlation analysis for the volumes of each amygdala subregion and other subregions within the two groups separately, while controlling for covariates such as eTIV, age, sex, and education.Altered structural covariance connectivity was determined by comparing differences in correlation coefficients of corresponding subregions of amygdala [23][24][25].The partial correlation coefficients were transformed by Fisher z transformation to obtain Z values, and then Z-tests were performed.P values were corrected by FDR and P FDR < 0.05 was considered to be statistically significant.
The partial correlation analysis was performed with age, sex, year of education, and eTIV as covariates to explore the correlation between the significantly reduced volume of subcortical and amygdala subregions and depression symptoms in T2DM.Statistical significance was determined with P FDR < 0.05.

Demographic and clinical characteristics of participants
Compared with HC, the T2DM group had higher SDS scores (P = 0.002).There was no significant statistical difference between age, sex, years of education, BMI, total intracranial volume, systolic, and diastolic blood pressure (P > 0.05).The biochemical data of fasting glucose, insulin, and cholesterol in the T2DM group and more details are shown in Table 1.

Subcortical regions and subregions of amygdala
We found that the volume reduced in the left thalamus (P FDR = 0.032) and left hippocampus (P FDR = 0.048) in the T2DM group (Table 2).Furthermore, we explored the volume of subregions in amygdala.The volume significantly decreased in bilateral corticoamygdaloid transition area (left: P FDR = 0.018; right: P FDR = 0.022), left basal nucleus (P FDR = 0.042), bilateral accessory basal nucleus (left: P FDR = 0.022; right: P FDR = 0.022), and left anterior amygdaloid area (P FDR = 0.042) in the T2DM, as detailed in Table 3.Compared with HC, the structural correlation between left paralaminar nucleus and right central nucleus was reduced in patients with T2DM (Z = − 3.91, P FDR = 0.013) (Fig. 2).

Correlation analysis
Partial correlation analysis was conducted, with covariates of age, sex, years of education, and total intracranial volume (eTIV) (Fig. 3).We found that the SDS scores were negatively correlated with the volume of bilateral cortico-amygdaloid transition area (left: r = − 0.42, P FDR = 0.024; right: r = − 0.36, P FDR = 0.045) in T2DM patients.

Discussion
This study unprecedentedly comprehensively captures the volume of subcortical regions and subregions of amygdala in T2DM and explores the correlation between amygdala and depressive symptoms.We found that the volume of left thalamus and left hippocampus was reduced.Furthermore, the volume of subregions of amygdala was reduced, especially the bilateral cortico-amygdaloid transition area, bilateral accessory basal nucleus, left basal nucleus, and left anterior amygdaloid area.We also explored the structural covariate connectivity between subregions of the amygdala and found diminished structural covariance connectivity between the left paralaminar nucleus and the right central nucleus.Particularly, the altered volume of the bilateral corticoamygdaloid transition area was associated with depressed symptoms in T2DM, which suggested that the reduced volume of corticoamygdaloid transition area may be the potential biomarker of depressed mood in patients with T2DM.Our findings provide insight into the alterations of subcortical regions and subfields of the amygdala, and the relationship between amygdala structural impairment and depressive mood in patients with T2DM.
There was a reduction of volume in the left thalamus and left hippocampus in T2DM.In previous studies, reduced mean kurtosis was found in the bilateral thalamus in a brain microstructure study using diffusion kurtosis imaging [26].These findings suggest that there are dysfunction of neurons and fiber tracts in the thalamus of T2DM patients, potentially serving as underlying microstructural mechanisms contributing to the macroscopic atrophy and structural abnormalities observed in the thalamus of T2DM patients.According to previous research, the thalamus is commonly associated with cognitive functions [27].Previous studies have further confirmed that abnormalities in hippocampal volume, blood perfusion, and neural activity, and that reduced volume is associated with cognitive decline [28].However, our study did not establish a significant correlation between thalamic and hippocampal volume and depressive symptoms in individuals with T2DM.Future investigations may delve into the potential link between thalamic and hippocampal microstructural alterations and cognitive impairment in T2DM.
The amygdala is a core brain area involved in regulating mood, anxiety, attention, and emotional memory in the central nervous system [29].Previous studies utilizing diffusion tensor imaging have noted that decreased fractional anisotropy in the uncinate fasciculus in individuals with depression may be indicative of diminished structural connectivity between the amygdala and the prefrontal cortex [30], suggesting potential impairments in nerve fiber bundles associated with depression.Another study based on fMRI identified reduced functional connectivity   between the amygdala and the inferior frontal gyrus, cingulate cortex, precentral gyrus, and fusiform gyrus in T2DM patients with depression [31].These findings underscore the potential significance of the amygdala as a critical region for early detection of depressive symptoms in individuals with T2DM.
The amygdala has been categorized into several subregions based on cytoarchitectonics, histochemistry, and function.Each subnucleus exhibits structural independence while maintaining interconnectedness with others [32,33].We further explored the alterations of volume in amygdala subregions and the relationship with depression symptoms in T2DM.Our study showed that reduction of volume in the bilateral cortico-amygdaloid transition area was negatively correlated with SDS scores.Notably, these alterations were not observed in the entire amygdala.This implied that assessing subregional volumes may offer greater sensitivity in detecting depressive symptoms in T2DM.The cortico-amygdaloid transition area has been described as a significant region associated with social communication and avoidance behavior [34].Some studies suggest that the cortico-amygdaloid transition area plays a role in assessing negative emotions [35,36], and the volume of the bilateral cortico-amygdaloid transition area in individuals with major depression is negatively correlated with the severity of depressive symptoms [17].Furthermore, the cortico-amygdaloid transition area is implicated in olfactory, social, and reward-related processing [22,35].Highlighting the significance of superficial nuclei of amygdala, an fMRI investigation revealed that superficial nuclei elicit distinct responses to static and dynamic visual stimuli depicting facial emotional expressions [37].This emphasizes the crucial role of superficial nuclei in human perception of environmental information, emotional evaluation, and social interaction.It is worth noting that these superficial nuclei are also linked to olfactory processing, and olfactory impairment has been a potential marker of cognitive decline in T2DM and Alzheimer's disease [38,39].There was an extensive reduction of volume in the accessory basal nucleus, basal nucleus, lateral nucleus, and anterior amygdaloid area in patients with major depression [40].In our study, the volume of bilateral accessory basal nucleus and left basal nucleus was reduced in patients with T2DM.The basal nucleus and accessory basal nucleus belong to the basolateral nucleus and play an essential role in the acquisition and expression of fear-related behaviors [35].Sensory input received by the lateral nucleus then reaches the basolateral nucleus, which regulates the output of information from the central nucleus to evoke behaviors and autonomic responses during the fear conditioning [41].We speculate that the atrophy of the basal nucleus and accessory basal nucleus in T2DM patients may be a potential factor affecting emotional processing.
In our study, we observed reduced structural covariance connectivity between the left paralaminar nucleus and the right central nucleus in T2DM patients.The central nucleus is the main site of amygdala output information and is involved in regulating stress-related behavioral responses [42].Several animal studies have proposed that the paralaminar nucleus is involved in contextual fear learning with the hippocampus [43].The right amygdala participates in the faster processing of emotional stimuli, whereas the left amygdala processes cognitively relevant emotional stimuli, so that the bilateral amygdala work together in a coordinated way in the complex process of emotional processing [44].Our study found reduced connectivity between the left paralaminar nucleus and the right central nucleus, suggesting that the structural covariance connectivity between the left and right amygdala and the synergistic role of emotion regulation might be affected in T2DM.
There are several limitations in this study.Firstly, this is a crosssectional study with small samples.Furthermore, depressive symptoms in diabetic patients were not graded in this study, and the duration of depressive symptoms in patients was not recorded.Therefore, in the future, T2DM patients with depressive symptoms can be graded, and the relationship between the duration of depression and structural changes in the amygdala can also be explored.Our study used 3 T magnetic field imaging, which has a lower signal-to-noise ratio and may not be sensitive enough to distinguish each subregion with the same accuracy as ultra-high magnetic field imaging and future studies could use higher resolution imaging to improve the accuracy of segmentation.

Conclusions
This study comprehensively assesses structural alterations in subcortical regions and amygdala subregions in T2DM patients, revealing reductions in the volume of the left thalamus, left hippocampus, left anterior amygdaloid area, bilateral cortico-amygdaloid transition area, bilateral accessory basal nucleus, and left basal nucleus in T2DM.The observed negative correlation between the bilateral corticoamygdaloid transition area of the amygdala and depressive symptoms highlights that the cortico-amygdaloid transition area may be a neuroimaging marker for identifying depressive symptoms in T2DM.

Fig. 2 .
Fig. 2. The alterations of structural covariance within the amygdala subregions in T2DM, compared with HC.Compared with HC, the structural correlation between left paralaminar nucleus and right central nucleus was reduced in patients with T2DM (z = − 3.91, P FDR = 0.013).Color bar represents the Z values; Red color indicates decreased structural association; Yellow color indicates increased structural association.La, lateral nucleus; Ba, basal nucleus; AB, accessory basal nucleus; AAA, anterior amygdaloid area; Ce, central nucleus; Me, medial nucleus; Co, cortical nucleus; CAT, cortico-amygdaloid transition area; PL, paralaminar nucleus.

Fig. 3 .
Fig. 3.The correlations between the significant amygdala subregions in patients with T2DM and depression symptoms.a.In T2DM group, SDS score was negatively correlated with the volume of left cortico-amygdaloid transition area (r = − 0.421, P FDR = 0.024).b.SDS score was negatively correlated with the volume of right cortico-amygdaloid transition area (r = − 0.358, P FDR = 0.045).SDS, self-rating depression scale.

Table 1
Demographic data, clinical biochemical indicators and neuropsychological results of all subjects.

Table 2
Group analysis of subcortical structural volumes.

Table 3
Group analysis of the volumes of amygdala subregions.