Investigating sex-related differences in brain structure and function in bipolar I disorder using multimodal MRI

Abstract

Background

Past research has highlighted that bipolar I disorder is associated with significant changes in brain structure and function. Notably, the manifestation and progression of bipolar I disorder have been known to differ between males and females. However, the relationship between sex-related differences and bipolar I disorder diagnosis affecting these changes was not fully understood. This study aimed to investigate the sex-by-diagnosis interactions concerning the structural and functional features of the brain in individuals with bipolar I disorder.

Methods

Both structural and functional MRI data were obtained from 105 individuals with bipolar I disorder (36 males and 69 females) and 210 healthy controls (72 males and 138 females). Voxel-wise analyses of gray matter volume and functional connectivity were conducted using a general linear regression model. This model included age, sex, diagnosis, and a sex-by-diagnosis interaction as predictors to explore potential sex-related differences in the brain features of participants with bipolar I disorder.

Results

The gray matter volume analysis revealed significant sex-by-diagnosis interactions in six brain regions: the left caudate (p < 0.001), left thalamus (p < 0.001), right caudate (p = 0.003), right thalamus (p < 0.001), left anterior cingulate gyrus (p = 0.022), and left middle/posterior cingulate gyrus (p = 0.015). Using these regions as seeds, we detected a significant sex-by-diagnosis interaction in the functional connectivity alteration between the left thalamus and right angular gyrus (p = 0.019).

Conclusions

Our findings revealed a noteworthy sex-by-diagnosis interaction, with male individuals with bipolar I disorder displaying larger gray matter volume and altered functional connectivity in the limbic system compared to female individuals with bipolar I disorder and healthy participants. These results hint at potential sex-related differences in the pathophysiology of the limbic system in bipolar I disorder, which may have significant implications for understanding the underlying mechanisms in bipolar I disorder. Our findings could contribute to developing more personalized treatment approaches for individuals with bipolar I disorder.

Introduction

Bipolar disorder (BD) is characterized by episodes of depression and either hypomania or mania, with prolonged mood disturbances that affect cognition and other life functions. BD is categorized into bipolar I disorder (BD-I) and bipolar II disorder (BD-II) based on the severity of manic symptoms. BD-I is characterized by manic episodes, while BD-II is characterized by hypomanic episodes. The prevalence of BD exceeds 1%, with BD-I accounting for a prevalence of 0.4–0.6% [1, 2]. Relevant studies have revealed no sex differences in the prevalence, age of onset, or severity of BD symptoms [3, 4]. However, in clinical presentation, female individuals with BD exhibit more frequent emotional state transitions than male individuals do, and female individuals often experience depression [3, 5, 6]. Additionally, BD is often accompanied by sex-related comorbid symptoms, with anxiety disorders more frequent in female individuals [7] and conduct and substance use disorders more common in male individuals [8, 9]. Therefore, the sex differences identified in the clinical statistics of BD, as reflected in the relevant literature, may support the idea that the brain structure and function of individuals with BD are influenced by sex factors [10].

Structural brain magnetic resonance imaging (MRI) studies have generally found no sex differences in total gray matter (GM) volume or white matter (WM) volume in individuals with BD [4, 10–13]. Nonetheless, the specifics of brain-related sex differences in individuals with BD remain unclear, with relevant literature presenting inconsistent findings about such differences in specific brain regions [10]. Aylward et al. observed an enlarged caudate volume in male individuals with BD than that of male healthy controls (HCs) [14]. Furthermore, compared to male HCs, male individuals with BD exhibited low left anterior cingulate cortex volumes, but no evidence of sex differences was observed in other cingulate cortical subregions [15].

A review study showed that the activation state of the right ventrolateral prefrontal cortex was lower in female individuals with BD compared to female HCs during the fearful facial affect categorization task. In addition, the activation state of the left caudate was higher in female individuals with BD than in female HCs during the Iowa Gambling Task [10]. A few research studies have conducted resting-state fMRI in this context [16]. However, resting-state fMRI studies involving psychiatric participants and healthy controls (HCs) are widely available and have been scientifically validated as reliable indicators of neural activity [17, 18]. Therefore, resting-state fMRI may be performed to ascertain whether sex-based differences in neural activity are present in individuals with BD.

Despite the reports discussed above, our understanding of sex differences in the brains of individuals with BD-I remains limited. Therefore, in this study, we performed systematic imaging analysis and integrated data from both MRI types. This study aimed to (1) identify the sex-by-diagnosis interactions in brain regions from structural images of BD-I individuals; and (2) extend the investigation into the functional connectivity of the brain in individuals with BD-I, with a focus on regions where the sex-by-diagnosis interactions were evident in structural images.

Materials and methods

Participants

All participants were selected from the Taiwan Aging and Mental Illness (TAMI) cohort. We included 105 individuals with BD-I (36 males, mean age = 52.81 ± 15 years; 69 females, mean age = 45.93 ± 12.49 years). In addition, a total of 210 HCs (72 males, mean age = 47.68 ± 12.94 years; 138 females, mean age = 48.54 ± 14.63 years) were selected in this study from the TAMI cohort in a 1:2 ratio. The following exclusion criteria were applied: (1) participants who were left-handed; (2) a Mini-Mental State Examination (MMSE) score of less than 26 [19]; and (3) participants with other psychiatric or neurological disorders. The present study was approved by the Institutional Review Board of Taipei Veterans General Hospital and National Yang Ming Chiao Tung University. All participants provided written informed consent.

MRI data acquisition

The MRI data of all participants were obtained from a 3.0-T Siemens MRI scanner with a 12-channel head coil at National Yang Ming Chiao Tung University. Whole-brain structural MRI data were collected using a three-dimensional magnetization-prepared rapid gradient echo sequence with the following imaging parameters: repetition time (TR) = 2530 ms; echo time (TE) = 3.5 ms; inversion time = 1100 ms; matrix size = 256 × 256; 192 slices; slice thickness = 1 mm; voxel size = 1.0 × 1.0 × 1.0 mm³; and flip angle = 7°. Whole-brain resting-state fMRI data were collected using a T2*-weighted gradient-echo-planar imaging sequence with these parameters: TR = 2500 ms; TE = 27 ms; matrix size = 64 × 64; voxel size = 3.4 × 3.4 × 3.4 mm³; 200 time points; field of view = 200 mm; and flip angle = 77°. The scanning protocols were consistent with those in our previous study [20].

MRI data preprocessing

Structural MRI and resting-state fMRI were processed using DPARSF_V4.3_170105 [21] and SPM12 (https://www.fil.ion.ucl.ac.uk/spm/software/spm12) in MATLAB R2019b (MathWorks, Natick, MA, USA). For the structural MRI data, preprocessing was conducted through the following steps: (1) The images were manually reoriented based on the anterior commissure–posterior commissure (AC–PC) line; (2) All images were subsequently normalized to the Montreal Neurological Institute (MNI) standard space and segmented into GM, WM, and cerebrospinal fluid (CSF) images. It is noteworthy that the GM images contain cortical, subcortical and cerebellar regions. We chose the GM image for subsequent analysis. The raw resting-state fMRI data were preprocessed through the following sequential steps: Initially, the first five data points were removed; subsequently, all the images were corrected by implementing slice-timing, realignment, and manual reorientation. In the third step, the reoriented images were coregistered with T1-weighted images, normalized to the MNI standard space, and resampled to a voxel size of 3 × 3 × 3 mm³. In the fourth stage, covariates were regressed out, encompassing those relevant to the time courses of six head motions, WM, and CSF. Ultimately, we applied temporal low-pass filtering within the range of 0.01 to 0.1 Hz. The preprocessing procedures were consistent with those in our previous study [20].

Statistical analysis

Demographic and clinical characteristics

Group differences in demographic and clinical characteristics, including age, MMSE score, Hamilton Depression Rating Scale (HAM-D) score, and Young Mania Rating Scale (YMRS) score, were evaluated through an independent t test. The chi-squared test was also employed to compare the sex distributions within groups. All statistical analyses were performed in IBM SPSS 24.0 software. The significance level was set at 0.05.

Structural MRI

A general linear regression model (GLM) was used to examine group differences. The GM images were the dependent variable, with sex, age, diagnosis, total intracranial volume, and sex-by-diagnosis interaction as the predictors. Statistical significance was considered based on the following criteria: (1) voxel cluster > 30 voxels, (2) voxel-level p = 0.001, and (3) after adjustment for family-wise error rate, brain regions with a cluster-level p < 0.05. The sex-by-diagnosis interaction statistics were visualized using the xjView toolbox (https://www.alivelearn.net/xjview). Next, we selected the brain regions that showed significant differences in the sex-by-diagnosis interaction as the regions of interest (ROIs) and created masks to perform the volume analysis and resting-state fMRI analysis.

Resting-state fMRI

We used the masks created from the structural images, which showed significant differences in the sex-by-diagnosis interaction between the two groups, to extract the resting-state fMRI cerebral blood flow signal data for subsequent analysis. The functional connectivity (FC) of the cerebral blood flow signals between the ROIs and the whole-brain GM voxels was calculated by Pearson’s correlation coefficient. Subsequently, Fisher’s z-transformation was employed to improve the normality of the data distribution. Additionally, we applied a smoothing process using a 6-mm full-width at half-maximum Gaussian kernel. The calculated FC was used as the dependent variable by GLM, and sex, age, diagnosis, and sex-by-diagnosis interaction were the predictors. Statistical significance parameters were set the same as those for structural MRI. The sex-by-diagnosis interaction statistics were visualized using xjView and BrainNet Viewer [22]. The flowchart of this study is presented in Fig. 1.

Fig. 1.

Research process flowchart. Abbreviations: HC, healthy controls; BD-I, bipolar I disorder; MRI, magnetic resonance imaging; GM, gray matter; ROIs, regions of interest; FC, functional connectivity

Results

Demographic and clinical characteristics

In the BD-I group, male individuals had a higher mean age than their female counterparts (p = 0.01). Female individuals, on the other hand, displayed greater severity of depression, as gauged by the HAM-D, compared to male participants (p = 0.03). Notably, there were no significant sex differences in the mean severity of mania, assessed by the YMRS (p = 0.48), nor in the proportion of individuals with psychotic features (p = 0.20). Among the HCs, male participants had a higher education level than female participants (p = 0.002), though there were no significant sex differences in mean age (p = 0.67). When comparing HCs with BD-I individuals, HCs demonstrated higher education levels and MMSE scores (p < 0.001 and 0.001, respectively). However, the mean age remained consistent across both groups (p = 0.98). Table 1 presents the statistical findings pertaining to the demographic and clinical characteristics of BD-I individuals and HCs.

Table 1.

Demographic and clinical characteristics of bipolar I disorder and healthy controls

Characteristics	BD-I					HC					Group comparison
	Male (n = 36)	Female (n = 69)	t (χ2)	P	Total (n = 105)	Male (n = 72)	Female (n = 138)	t	p	Total (n = 210)	t	p
Age, year (range)	52.81 ± 15.00 (20–83)	45.93 ± 12.49 (22–69)	2.50	0.01 *	48.29 ± 13.73 (20–83)	47.68 ± 12.94 (22–80)	48.54 ± 14.63 (21–85)	-0.43	0.67	48.24 ± 14.05 (21–85)	0.03	0.98
Education level, year (range)	12.53 ± 3.26 (2–18)	12.62 ± 3.57 (6–21)	-0.13	0.89	12.59 ± 3.45 (2–21)	16.03 ± 3.98 (3–25)	14.22 ± 4.07 (0–24)	3.08	0.002 *	14.84 ± 4.12 (0–25)	-4.81	< 0.001 *
MMSE (range)	27.06 ± 2.78 (18–30)	27.04 ± 2.53 (16–30)	0.02	0.98	27.05 ± 2.60 (16–30)	27.69 ± 5.01 (23–30)	28.48 ± 1.57 (22–30)	-1.29	0.20	28.21 ± 3.20 (22–30)	-3.23	0.001 *
Duration of illness, year (range)	17.41 ± 10.54 (2–45)	18.29 ± 12.89 (0–58)	-0.33	0.74	17.76 ± 12.21 (0–58)	-	-	-	-	-	-	-
YMRS (range)	2.19 ± 3.23 (0–8)	2.77 ± 4.04 (0–8)	-0.70	0.48	2.53 ± 3.89 (0–8)	-	-	-	-	-	-	-
HAM-D (range)	3.97 ± 3.32 (0–13)	5.81 ± 5.12 (0–17)	-2.21	0.03 *	5.04 ± 4.59 (0–17)	-	-	-	-	-	-	-
Psychotic feature, n (%)	12 (33.33)	32 (46.38)	1.65	0.20 a	44 (41.91)	-	-	-	-	-	-	-
Treatment status, n (%)
Lithium	10 (27.78)	21 (30.43)	-	-	31 (29.52)
AntiEpileptic	28 (77.78)	50 (72.46)			78 (74.29)
Gen1AntiPsych	10 (27.78)	13 (18.84)			23 (21.90)
Gen2AntiPsych	24 (66.67)	45 (65.22)			69 (65.71)
Antidepressant	6 (16.67)	15 (21.74)			21 (20.00)

Abbreviations: BD-I, bipolar I disorder, HC healthy control, SD standard deviation, MMSE Mini-Mental State Examination, YMRS Young Mania Rating Scale, HAM-D Hamilton Rating Scale for Depression, AntiEpileptic Anti-Epileptic Agents, Gen1AntiPsych First Generation Antipsychotics, Gen2AntiPsych Second Generation Antipsychotics

Data are mean value ± SD or n (%) unless specified otherwise

^aChi-squared test, significant level (α) = 0.05

*p < 0.05

Structural MRI

A GLM was used to determine the influence of the sex-by-diagnosis interaction on the brain structure of individuals with BD-I. Our findings revealed statistically significant differences in the interaction across two groups, specifically within the following brain regions: the left caudate nucleus (p < 0.001), left thalamus (p < 0.001), left anterior cingulate gyrus (p = 0.022), left middle/posterior cingulate gyrus (p = 0.015), right caudate (p = 0.003), and right thalamus (p < 0.001). In addition, the volumes of the caudate, thalamus, left anterior cingulate gyrus, and left middle/posterior cingulate gyrus were larger in male individuals with BD-I than in female individuals and HCs (see Table 2 and Fig. 2 for more details).

Table 2.

Sex-by-diagnosis interaction in voxel-wise statistics of brain volume among all participants

Brain regions	AAL	MNI coordinates of maximal voxel			t	Cluster-level
		x	y	z		k	pFWE−corr
L caudate	71	-8	14	-8	4.82	117	< 0.001
L thalamus	77	-6	-10	8	4.95	130	< 0.001
L anterior cingulate gyrus	31	-4	32	12	4.79	31	0.022
L middle/posterior cingulate gyrus	-	0	-30	28	4.41	34	0.015
R caudate	72	8	8	-12	4.14	50	0.003
R thalamus	78	8	-12	4	6.31	139	< 0.001

Abbreviations: AAL automated anatomical labelling, k voxel numbers, p^FWE−corr, p value corrected for multiple comparison using family-wise error at cluster level, L left, R right

Only clusters with k ≥ 30 and p^FWE−corr < 0.05 at cluster level were reported

Fig. 2.

Comparison of brain volume differences between the two groups by sex.The histogram illustrates brain regions that exhibited significant sex-by-diagnosis interactions among all participants. Abbreviations: HC, healthy controls; BD-I, bipolar I disorder

Resting-state fMRI

For the results of the FC analysis, we found that the FC between the left thalamus and the right angular gyrus exhibited a sex-by-diagnosis interaction that was higher in male individuals with BD-I than in their female counterparts and HCs (p = 0.019). According to the GLM statistics, no significant difference was observed in the FC between the other ROIs and the whole-brain voxels (see Table 3; Fig. 3 for more details).

Table 3.

Sex-by-diagnosis interaction in voxel-wise statistics of functional connectivity using the left thalamus as the seed region among all participants

Brain regions	AAL	MNI coordinates of maximal voxel			t	Cluster-level
		x	y	z		k	pFWE−corr
R angular gyrus	66	36	-60	36	3.55	32	0.019

Abbreviations: AAL automated anatomical labelling, k voxel numbers, p_FWE−corr, p value corrected for multiple comparison using family-wise error at cluster level, R right

Only clusters with k ≥ 30 and p^FWE−corr < 0.05 at cluster level were reported

Fig. 3.

Brain regions with significant functional connectivity in sex-by-diagnosis interactions between the two groups.This figure illustrates the functional connectivity between the left thalamus and the right angular gyrus, highlighting regions with notable sex-by-diagnosis interactions

Discussion

The present study explored potential sex-based brain structural and functional differences in individuals with BD-I. The main findings can be summarized as follows. First, the volumes of the caudate, thalamus, left anterior cingulate gyrus, and left middle/posterior cingulate gyrus were observed to be larger in males with BD-I than in females with BD-I and HCs. Second, a noticeable sex-by-diagnosis interaction was detected in the FC between the left thalamus and the right angular gyrus.

Differences in the clinical characteristics of individuals with BD-I

According to the statistical results, female individuals with BD-I exhibited higher severity of depression than their male counterparts. However, a relevant study has suggested that no sex differences exist in depression severity [3]. In this study, we found a higher proportion of female participants than male participants. This result is similar to a previous study, which analyzed recently published large sample studies in patients with BD. The results showed a consistent predominance of female gender in samples of patients with BD, including BD-I and BD-II. The authors suggested possible reasons for this phenomenon, including sex differences in help-seeking behavior, sampling bias, comorbidities, symptom presentation, and geographic and cultural differences [23]. These factors may also explain why our study has more female participants than male participants. We performed the HAM-D to evaluate the depression severity in individuals with BD-I. The HAM-D classifies depression severity with scores: no depression (0–7), mild depression (8–16), moderate depression (17–23), and severe depression (≥ 24) [24]. In the present study, most individuals with BD-I did not show symptoms of depression. However, we found some female individuals with BD-I were in the depressed phase. Approximately 20% of female individuals (i.e., 15 female individuals) were found to have mild or moderate depression. This finding is similar to previous findings. Previous studies have shown that female patients are more likely to experience rapid cycling, depressive symptoms, and mixed states [25, 26]. This may reflect the finding that some of the female individuals with BD-I in our study were in the depressed phase. Additionally, no significant sex differences were observed in either mania severity or the percentage of individuals presenting with psychiatric symptoms. These findings are consistent with relevant studies [3, 9].

Implication of sex differences in the brain structure of individuals with BD-I

The statistical results of the GLM indicated that the caudate, thalamus, left anterior cingulate gyrus, and left middle/posterior cingulate gyrus exhibited a sex-by-diagnosis interaction. The caudate, a part of the basal ganglia, is associated with cognitive and emotional processing [27]. Additionally, the thalamus and cingulate gyrus, which are both located in the limbic system, are also associated with emotional performance. The basal ganglia and limbic system are often used as key brain regions for BD research, and the brain ROIs discovered in the present study may also be key regions related to BD in terms of sex among individuals with BD-I. Our findings might suggest that the observed interactions could have implications for understanding the differential impact of sex on emotional functions in individuals with BD-I.

Our findings indicated that the caudate exhibited a sex-by-diagnosis interaction, which was consistent with a previous study [14]. Furthermore, the results revealed that the caudate volume in male individuals with BD-I was larger than that in female individuals and HCs. Kozicky et al. found that increased right caudate volume in individuals with BD-I was associated with poorer cognitive performance. The authors speculated that the increase in the right caudate volume might be a compensatory mechanism to alleviate cognitive impairment [28]. In this study, we only utilized the MMSE as an overall cognitive function test, so we could not comprehensively assess the cognitive function status of the participants. It is challenging to infer whether the increase in the caudate volume in male individuals with BD-I was reflected in cognitive function. However, some studies have demonstrated that the caudate volume was smaller in individuals with BD-I than in HCs [29–31]. Despite variations in findings, they all suggest a structural abnormality in the basal ganglia of individuals with BD.

We found that the left thalamus volume was larger in male individuals with BD-I than in female individuals and HCs, which was consistent with a prior study [32]. Adler and colleagues suggested that structural abnormalities of the left thalamus might occur early in individuals with BD, and this could result from abnormal neuronal proliferation or pruning [32]. In male individuals with BD-I, the vesicular monoamine transporter (VMAT2) binding in the thalamus was higher than that in female individuals and HCs [33]. These findings indicated that an elevation in VMAT2 expression, leading to an increased concentration of monoaminergic synaptic terminals, may represent a disease-related abnormality in individuals with BD-I. Interestingly, this abnormality exhibited variation between male and female individuals [33]. Based on these findings, we speculated that alterations in VMAT2 concentration might correlate with the increased thalamus volume in male individuals with BD-I. However, additional research is essential to corroborate this assumption.

The present study identified a sex-by-diagnosis interaction in the left anterior cingulate gyrus and left middle/posterior cingulate gyrus. Previous studies have shown inconsistent results regarding volume-based or thickness-based morphometry in the anterior cingulate gyrus of male individuals with BD-I [15, 34]. In addition, lithium remains the most effective treatment for BD, demonstrating pronounced effects on both acute mood episodes and long-term mood stabilization. Previous studies have shown that treatment with lithium might increase anterior cingulate gyrus volumes in individuals with BD [35–37]. One voxelwise meta-analysis study revealed that the anterior cingulate gyrus volume in individuals with BD was initially smaller than that in HCs. However, with the progression of the disease and under lithium treatment, there was a notable increase in anterior cingulate gyrus volume [34]. Lithium has the potential to manifest both neurotrophic and neuroprotective effects, contributing to the limitation of pathological atrophy of GM in critical brain regions in individuals with BD [36]. Both Baldassano et al. and Karanti et al. have noted that male individuals with BD were more likely to be prescribed lithium than were female individuals [5, 38]. This may be the reason for the larger cingulate gyrus observed in male individuals with BD in this study.

Implications of sex differences in the FC of individuals with BD-I

In the present study, we explored sex differences in BD-I using voxel-based analysis of resting-state fMRI. We identified a sex-by-diagnosis interaction in the FC between the left thalamus and the right angular gyrus. Previous research has indicated significant differences in the FC between the left thalamus and the right angular gyrus, which might be associated with the sensorimotor networks and could be reflected in the severity of manic episodes [39]. While our results did not indicate any sex-based disparities in the clinical severity of mania, we cannot rule out the possibility that changes in FC among individuals with BD might serve as compensation for the sensorimotor networks [40]. Ding et al. found that individuals with schizophrenia exhibited increased FC in the thalamocortical network (encompassing the thalamus and angular gyrus) [41]. The thalamocortical network is primarily associated with rhythmic activities, such as sleep [42]. Salvatore et al. demonstrated that individuals with BD-I exhibited abnormal circadian rhythms compared to HCs [43]. In addition, individuals with BD-I experienced sleep problems, which, in some studies, have shown a correlation with manic episodes, particularly in female individuals [44]. These findings suggested that the thalamocortical network might influence the potential sex differences in sleep patterns among individuals with BD-I. However, future studies are essential to validate our findings. Notably, previous studies have explored the relationship between brain structure and function in BD, with a particular focus on the left thalamus. Chen et al. reported a larger volume of the left thalamus in males with BD than that of HCs [45]. In addition, Deicken et al. further supported this, showing increased N-acetylaspartate and creatine in both the right and left thalamus of males with BD-I than those of HCs. Elevated levels of N-acetylaspartate in the bilateral thalamus may indicate neuronal hypertrophy or hyperplasia, decreased density of glial cells, or abnormal pruning. In addition, elevated levels of creatine in the bilateral thalamus may reflect changes in cellular energy metabolism [46]. These findings suggest a possible cause for the structural and functional abnormalities in the left thalamus in males with BD-I, but inferring a causal relationship is challenging.

Limitations

In this study, several limitations need to be addressed. First, while we controlled for age, sex, and total intracranial volume in our data on individuals with BD-I and HCs, we did not account for the effects of psychotropic medication. To the best of our knowledge, there is no gold standard for converting equivalent medication use in individuals with BD, making it challenging to address. Future studies should include more information on drug dosages and blood tests. This will help us try to convert equivalent drug use (e.g., chlorpromazine or haloperidol equivalent doses for the antipsychotics; mean serum levels for the mood stabilizers). Meanwhile, a more granular approach to group divisions might offer more precise insights. We believe that if the groups can be divided according to the disease duration, treatment duration, type of mood stabilizer, and dose, the purpose of precision medicine may be better achieved in interpreting the results. Second, the significance and mechanisms behind the increased volume observed in male individuals with BD-I remain unknown. The reasons for the variations in brain region volumes by sex and the implications of such increased volume—whether beneficial or detrimental to individuals with BD—warrant further investigation. Finally, as suggested by previous research, the brain regions exhibiting a sex-by-diagnosis interaction in our findings play a role in cognitive function [28, 33, 47]. Kozicky et al. found a correlation between decreased caudate volume and poor cognitive performance in individuals with BD [28]. Zubieta et al. revealed poor verbal learning abilities in male individuals with BD due to abnormalities in the thalamus [33]. Zimmerman et al. reported that individuals with BD who performed poorly on the Wisconsin Card Sorting Test and the Trail Making Test had related abnormalities in the anterior cingulate gyrus [47]. However, our reliance on the MMSE as the sole measure of cognitive function may not provide a comprehensive assessment of cognitive status in individuals with BD-I.

Conclusions

The present study established a systematic process for analyzing brain images to determine sex differences in brain structure and FC in individuals with BD-I. The caudate, thalamus, left anterior cingulate gyrus, and left middle/posterior cingulate gyrus volumes were significantly higher in male individuals with BD-I than in female individuals. Moreover, the FC between the left thalamus and the right angular gyrus exhibited a sex-by-diagnosis interaction, suggesting that alterations in brain function may be associated with sex differences in individuals with BD. Future research should incorporate multimodal brain imaging and execute various signal analysis techniques. Such an approach will better support and elucidate the hypothesis of sex differences in the brain structure and function of individuals with BD.

Abbreviations

AC–PC

anterior commissure–posterior commissure

BD

bipolar disorder

BD-I

bipolar I disorder

BD-II

bipolar II disorder

CSF

cerebrospinal fluid

FC

functional connectivity

fMRI

functional MRI

GLM

general linear regression model

GM

gray matter

HAM-D

Hamilton Depression Rating Scale

HC(s)

healthy control(s)

MRI

magnetic resonance imaging

MMSE

Mini-Mental State Examination

MNI

Montreal Neurological Institute

ROI(s)

region(s) of interest

TR

repetition time

TE

echo time

TAMI

Taiwan Aging and Mental Illness

VMAT2

vesicular monoamine transporter

WM

white matter

Authors' contributions

A.C.Y. devised the concept and supervised the study. A.C.Y. and S.J.T contributed to data collection. M.Y.L., J.D.Z., and H.J.T. contributed to data analysis and interpretation of data. M.Y.L. and J.D.Z. wrote the first draft of the manuscript. A.C.Y. revised and improved the manuscript. All authors have approved the final manuscript. M.Y.L. and J.D.Z. contributed equally to this work.

Funding

This work was supported by the National Science and Technology Council (NSTC), Taiwan (Grant No. 113-2321-B-A49-020-, 111-2634-F-A49-014-, 112-2321-B-A49-013, and 112-2823-8-A49-001), Taiwan Centers for Disease Control, Ministry of Health and Welfare, Taiwan (Grant No. MOHW112-CDC-C-114-000111), Taipei Veterans General Hospital, Taipei, Taiwan (Grant No. V112C-054). Dr. A.C.Y. was also supported by the Mt. Jade Young Scholarship Award from the Ministry of Education, Taiwan. This work was financially supported by the Brain Research Center, National Yang Ming Chiao Tung University from The Featured Areas Research Center Program within the framework of the Higher Education Sprout Project by the Ministry of Education (MOE) in Taiwan (Gran No. 112W32101).

Data availability

The datasets used and analyzed during the present study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the Declaration of Helsinki and was approved by the Institutional Review Board of Taipei Veterans General Hospital and National Yang Ming Chiao Tung University. All participants provided written informed consent.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.