Hemispheric asymmetries and network dysfunctions in adolescent depression: A neuroimaging study using resting-state functional magnetic resonance imaging

Abstract

BACKGROUND

Currently, adolescent depression is one of the most significant public health concerns, markedly influencing emotional, cognitive, and social maturation. Despite advancements in distinguish the neurobiological substrates underlying depression, the intricate patterns of disrupted brain network connectivity in adolescents warrant further exploration.

AIM

To elucidate the neural correlates of adolescent depression by examining brain network connectivity using resting-state functional magnetic resonance imaging (rs-fMRI).

METHODS

The study cohort comprised 74 depressed adolescents and 59 healthy controls aged 12 to 17 years. Participants underwent rs-fMRI to evaluate functional connectivity within and across critical brain networks, including the visual, default mode network (DMN), dorsal attention, salience, somatomotor, and frontoparietal control networks.

RESULTS

Analyses revealed pronounced functional disparities within key neural circuits among adolescents with depression. The results demonstrated existence of hemispheric asymmetries characterized by enhanced activity in the left visual network, which contrasted the diminished activity in the right hemisphere. The DMN facilitated increased activity within the left prefrontal cortex and reduced engagement in the right hemisphere, implicating disrupted self-referential and emotional processing mechanisms. Additionally, an overactive right dorsal attention network and a hypoactive salience network were identified, underscoring significant abnormalities in attentional and emotional regulation in adolescent depression.

CONCLUSION

The findings from this study underscore distinct neural connectivity disruptions in adolescent depression, underscoring the critical role of specific neurobiological markers for precise early diagnosis of adolescent depression. The observed functional asymmetries and network-specific deviations elucidate the complex neurobiological architecture of adolescent depression, supporting the development of targeted therapeutic strategies.

Key Words: Adolescent depression; Brain network connectivity; Neuroimaging biomarkers; Functional magnetic resonance imaging; Default mode network; Salience network; Hemispheric asymmetry

Core Tip: This study explores the neural underpinnings of adolescent depression by investigating disrupted brain network connectivity through resting-state functional magnetic resonance imaging. Key findings highlight significant hemispheric asymmetries, with enhanced activity in the left visual and default mode networks and reduced engagement in the right hemisphere. Additionally, the overactivity of the right dorsal attention network and hypoactivity of the salience network were observed, indicating impairments in emotional and attentional regulation. These distinct neurobiological markers emphasize the importance of early diagnosis and personalized therapeutic interventions for adolescent depression.

INTRODUCTION

Adolescent depression has emerged as a significant public health concern due to its profound impact on emotional, cognitive, and social development[1-3]. The critical developmental stage of adolescents is characterized by significant neural maturation and synaptic pruning[4], which renders the adolescent brain susceptible to environmental stressors and genetic predispositions, potentially leading to the onset of depressive disorders[5-7]. Recent advancements in neuroimaging and network neuroscience have provided novel insights into the underlying neurobiological mechanisms of depression. In particular, research findings have emphasized the role of brain network connectivity[8,9] in understanding the pathophysiology of depression.

Brain network connectivity refers to the intricate interaction between different brain regions to facilitate communication and information processing across the brain[10]. This connectivity is categorized into structural connectivity, which refers to the physical connections between different brain regions, and functional connectivity, which refers to the temporal correlations in neural activity across different brain regions[11]. Research findings have revealed disruptions in both structural and functional connectivity among adolescents with depression[12-14], indicating that abnormalities in brain network connectivity significantly contribute to the cognitive and emotional disturbances associated with depression[15,16].

Neuroimaging techniques, such as functional magnetic resonance imaging (fMRI), have enabled the identification of key brain networks implicated in adolescent depression. These include the default mode network (DMN), which is associated with rumination[17]; the salience network, which is involved in response to emotionally salient stimuli[18]; the executive control network, which plays a critical role in cognitive control and emotional regulation[19]. Disruptions in the connectivity within and between these networks have been associated with the characteristic symptoms of depression, including persistent negative affect, cognitive impairment, and difficulty in emotion regulation.

Moreover, emerging evidence indicate that the developmental trajectory of brain network connectivity during adolescence may be altered in individuals with depression[20]. Typically, adolescence is a period of increasing network integration, however, this process may be disrupted in depressed adolescents, leading to inefficient and sparse network connectivity[20,21]. These alterations serve as potential biomarkers for early diagnosis and treatment response prediction, thereby offering new insights for developing personalized interventions.

Given the growing recognition of the importance of brain network connectivity in adolescent depression[22-24], this study aims to investigate and identify the specific patterns of connectivity that characterize adolescent depression. We aim to reveal the neural correlates of depression in adolescents by using neuroimaging tools and analytical methods. This research will contribute to the broader field of adolescent mental health by providing deeper insights into the neural mechanisms underlying depression, thereby informing the development of effective therapeutic strategies.

MATERIALS AND METHODS

Ethics declaration

Written informed consents were obtained from the guardians of participants. The research was approved by the Medical Ethics Committee of the First Affiliated Hospital of Chongqing Medical University (No. 2017-157) and the Medical Ethics Committee of Chongqing People’s Hospital (No. KY S2024-007-01). The study was conducted in accordance with the ethical guidelines in the Declaration of Helsinki.

Participants

This prospective study recruited adolescents aged 12 to 17 years diagnosed with depression, along with healthy controls (HCs), between August 2020 and July 2022. The adolescents with depression were recruited from the inpatient clinics of the Department of Psychiatry at the First Affiliated Hospital of Chongqing Medical University, whereas the HCs were recruited through community awareness programs.

The diagnosis was based on the diagnostic and statistical manual of mental disorders, 5^th edition criteria, and participants were evaluated using the MINI International Neuropsychiatric Interview for Children and Adolescents, conducted by two experienced psychiatrists. Severity of depression was assessed using the 17-item Hamilton depression rating scale (HAMD-17), ensuring inclusion of only participants experiencing a clinically significant depressive episode (HAMD-17 score > 17). The inclusion criteria were as follows: Participants experiencing their first depressive episode, with no history of manic or hypomanic episodes, and no antidepressant treatment received. Han nationality, right-handed, and aged 12 to 17 years. Additionally, participants must not have taken psychotropic, anesthetics, sleeping, sedative, or analgesic drugs in the past month. The exclusion criteria were as follows: Participants diagnosed with neurological diseases, including epilepsy and multiple sclerosis, or other major physical health conditions, including heart disease and cancer. Presence of other mental and borderline personality disorders or intracranial mass. Participants with a family history of mental illness or self-harm behavior, a history of severe craniocerebral trauma causing loss of consciousness, and a history of drug or alcohol abuse or dependence. Additionally, participants who exhibited contraindications to MRI scanning, presence of foreign metal bodies affecting image quality (for instance pacemakers, certain metal implants, and dental braces), or inability to undergo complete MRI scanning were excluded.

For the HCs group, healthy volunteers were recruited from the community. The inclusion criteria were as follows: Participant with a HAMD-17 score below 7, of Han nationality, right-handed, aged 12 to 17 years, and with no other physical or serious mental diseases. The exclusion criteria for the control group were the same as those for the participants in the depression group.

Collecting resting-state-fMRI data

Resting-state fMRI (rs-fMRI) data were collected using a 3T GE Signa HDxt scanner (General Electric Healthcare, Chicago, IL, United States) equipped with an 8-channel head coil. Participants were instructed to relax with their eyes closed and remain awake, while minimizing their cognitive activity. None of the participants reported falling asleep during the scan. Foam pads and earplugs were used to reduce head motions and machine noise.

The echo-planar imaging pulse sequence parameters were as follows: Repetition time (TR) = 2000 ms, echo time (TE) = 40 ms, field of view (FOV) = 240 mm × 240 mm, matrix size = 64 × 64, flip angle = 90, slice count = 33, slice thickness/gap = 4.0 mm/0 mm, scan duration = 8 minutes, and a total of 240 volumes. Three-dimensional T1-weighted MRIs, used for co-registration of rs-fMRI data, exhibited the following parameters: TR = 24 ms, TE = 9 ms, FOV = 240 mm × 240 mm, matrix size = 240 × 240, flip angle = 90, and slice thickness/gap = 1.0 mm/0 mm.

Statistical analysis

The structural scans were processed using FreeSurfer 4.5.0[25], while the structure-function registration was conducted using FsFast[26]. Functional preprocessing steps included slice time correction, motion correction, low-pass temporal filtering retaining frequencies below 0.08 Hz, regression of motion parameters, and signals from ventricular, white matter, and whole-brain regions[27], as well as the removal of linear trends. The final data were projected onto the fsaverage6 surface space (with a vertex spacing of approximately 2 mm) and subsequently smoothened using a 6 mm full-width half-maximum kernel.

We categorized the brain into 200 regions[28] and for every region, the mean value was calculated across all voxels within that region, producing a single time-series per region. We estimated the temporal similarity for each region by computing Pearson’s correlation coefficients and subsequently computed the network connectivity values.

The unidirectional graph matrices were then vectorized into 19900 edges per subject[29]. We performed a paired t-test with false discovery rate correction (P < 0.05) to compare the functional connectivity differences between the patient group and the HCs. Subsequently, we calculated effect sizes (Cohen’s d) and 95% confidence intervals (CI) to quantify the magnitude of group differences.

To analyze the demographic data, we conducted a two-sample t-test for continuous data and a χ² test for categorical data using international business machines corporation statistical product and service solutions 22.0 software.

RESULTS

Study participants

A total of 77 adolescents with depression were initially recruited, with three being excluded due to incomplete MRI examinations. In contrast, a total of 61 HCs were initially recruited, with one being excluded due to the presence of a lesion at the pineal gland space and another for undergoing dental orthodontic treatment. Subsequently, 74 participants with depression (age: 14.99 ± 1.49 years; 50 females, 24 males) and 59 HCs (age: 15.05 ± 2.00 years; 19 females, 40 males) (mean ± SD) were included in this study (Figure 1).

Figure 1 A flowchart illustrating the selection process for adolescents with depression and healthy controls. HCs: Healthy controls; MINI-kid: MINI International Neuropsychiatric Interview for Children and Adolescents; HAMD-17: 17-item Hamilton depression rating scale; DSM-V: Diagnostic and statistical manual of mental disorders, 5^th edition criteria.

Table 1 summarizes the demographic characteristics of the adolescents with depression and HCs. There were no significant differences between the adolescent depression and HCs groups in terms of education level, gender, and age (P > 0.05).

Table 1 Demographics and clinical profiles of adolescent depression and healthy controls, mean ± SD.

Item	Depression group (n = 74)	HCs group (n = 59)	t/χ2 value	P value
Gender (F/M)	50/24	40/19	0a	1.00
Age (years)	14.99 ± 1.49	15.02 ± 2.00	-0.10b	0.92
Education (years)	9.09 ± 1.54	8.92 ± 2.09	0.57b	0.57
HAMD	26.16 ± 5.77	0.90 ± 1.67	32.58b	< 0.001

^aχ² test.

^bTwo sample t-test.

HCs: Healthy controls; F: Female; M: Male; HAMD: Hamilton depression rating scale.

Functional discrepancies across key brain networks

Significant differences in functional brain network activity were observed between the adolescent depression and HCs groups across several key networks (Figure 2 and Table 2). Notably, the left hemisphere visual network, exhibited significantly increased activity in the adolescent depression group compared to the HCs (t = 4.389, P < 0.001, Cohen’s d = 1.20, 95%CI: 0.85-1.55). Conversely, the right hemisphere visual network exhibited a significantly decreased activity in depression group compared to the HCs (t = -3.106, P = 0.002, Cohen’s d = 0.92, 95%CI: 0.47-1.37).

Figure 2  Functional connectivity differences across key brain networks in adolescents with depression group and healthy controls group.

Table 2 Functional connectivity differences across key brain networks in adolescents with depression and healthy controls groups.

Item	X	Y	Z	t value	P value (FDR)
LH_Visual Network_14	120	19	135	4.389	< 0.001
RH_Visual Network_11	124	18	140	-3.106	0.002
LH_Default Mode Network_PFC_9	205	63	82	3.370	< 0.001
RH_Default Mode Network_Par_3	209	62	80	-5.397	< 0.001
RH_Default Mode Network_PFCdPFCm_5	208	62	81	-5.339	< 0.001
RH_Dorsal Attention Network_Post_9	4	119	21	3.119	0.002
RH_Salience Network_Med_1	200	57	250	-3.496	< 0.001
LH_Somatomotor Network_8	70	130	182	-3.149	0.002
LH_Somatomotor Network_11	70	130	185	-3.325	0.001

LH: Left hemisphere; RH: Right hemisphere; FDR: False discovery rate.

Particularly, the DMN demonstrated significant hemispheric asymmetries in functional activity. Within the left hemisphere, the prefrontal cortex (PFC) region of the DMN exhibited a significant upregulation in activity in the adolescent depression group compared to the HCs (t = 3.370, P = 0.001, Cohen’s d = 1.05, 95%CI: 0.68-1.42). Contrastingly, in the right hemisphere, the same region exhibited a significant downregulation in activity in the depression group compared to the HCs (t = -5.339, P < 0.001, Cohen’s d = 1.47, 95%CI: 1.08-1.86).

Further differences in activity were noted in the attention-related networks, with the right hemisphere dorsal attention network exhibiting a significantly increased activity in the adolescent depression group compared to HCs (t = 3.119, P = 0.002, Cohen’s d = 0.95, 95%CI: 0.51-1.39). In contrast, the right hemisphere salience network showed decreased activity in the adolescent depression group compared to the HCs (t = -3.496, P < 0.001, Cohen’s d = 0.99, 95%CI: 0.55-1.43). Additionally, both somatomotor networks in the left hemisphere were significantly less active in the adolescent depression group compared to HCs (t = -3.149, P = 0.002, Cohen’s d = 0.88, 95%CI: -1.40 to -0.36), indicating extensive functional disruptions across several critical brain networks in adolescents with depression.

DISCUSSION

The findings from this study indicate existence of significant functional differences across several key brain networks in adolescents with depression, thereby providing deeper insights into the neurobiological mechanisms underlying this disorder. The observed lateralized alterations in network activity, particularly within the visual network and the DMN, underscore the complex interaction of neural circuits that are disrupted in adolescent depression.

The increased activity in the left hemisphere visual network, compared to the decreased activity in the right hemisphere, suggests a potential compensatory mechanism or an imbalance in hemispheric dominance that underlies the perceptual and cognitive distortions characteristic of depression. This hemispheric asymmetry reflect a differential engagement of visual processing pathways, potentially associated with altered attentional biases toward negative stimuli-a phenomenon commonly exhibited in depressive disorders[30-33]. This differential activation affect basic visual perception, as well as contributing to broader cognitive deficits observed in the adolescent depression population[34], including impairments in attention, memory, and emotional regulation.

The DMN, which is vital in self-referential thought and the regulation of emotional processes, exhibited significant hemispheric asymmetries[35-37]. The increased activity in the left hemisphere PFC region of the DMN, alongside decreased activity in the right hemisphere, suggests that adolescents with depression experience heightened internal focus and rumination-a key cognitive hallmark of depression[38,39]. The reduced activity in the right hemisphere indicate a diminished capacity for adaptive disengagement from self-focused thought, thereby exacerbating the maladaptive cognitive patterns observed in depression. These findings are consistent with results from previous research[40] that associate DMN dysfunction with mood disorders, thereby underscoring the need for targeted interventions that aim to modulate DMN activity.

The increased activity in the right hemisphere dorsal attention network is consistent with the heightened vigilance and difficulties in disengaging from negative stimuli observed in depressive patients[39]. This hyperactivity represents an overactive attentional bias toward external cues, potentially acting as a maladaptive response to perceived environmental stressors[41]. Conversely, the decreased activity in the right hemisphere salience network indicate a reduced ability to identify and prioritize emotionally relevant stimuli, which could impair adaptive emotional responses[42]. This overactive attentional engagement and diminished salience processing demonstrate the broader attentional and emotional dysregulation typical in adolescent depression.

The decreased activity observed in the left hemisphere somatomotor networks further implicates motor function alterations within the pathophysiology of depression. These disruptions may manifest as psychomotor slowing[43,44] or decreased physical coordination[45-47], both of which are commonly observed in depressive disorders. The lateralized nature of these findings indicates that the integration between sensory inputs and motor outputs is differentially affected in adolescents with depression, potentially influencing daily functioning and overall quality of life.

Adolescence is a period of profound brain maturation, characterized by processes such as synaptic pruning[48], and increased network integration[49]. These developmental changes likely interact with the connectivity disruptions observed in adolescent depression. The variability in developmental stages among participants may have contributed to the observed patterns of connectivity disruption. Younger adolescents, in earlier phases of neural maturation, may exhibit different connectivity profiles compared to older adolescents closer to adulthood. This variability underscores the importance of interpreting the findings within the context of developmental trajectories.

The observed hemispheric asymmetries and network-specific disruptions in adolescent depression reflect complex neurobiological mechanisms. The enhanced activity in the left visual network, coupled with diminished activity in the right hemisphere, suggests differential engagement of hemispheric visual processing pathways. This could stem from developmental factors such as synaptic pruning and the maturation of neural circuits, which are particularly pronounced during adolescence[50]. Similarly, the observed DMN dysregulation highlights an imbalance in self-referential processing, potentially driven by altered neurotransmitter dynamics and structural connectivity deficits.

The findings from this study hold significant promise for advancing the diagnosis and treatment of adolescent depression. The observed connectivity disruptions within key brain networks, such as the DMN and salience network, highlight potential neuroimaging biomarkers for early detection. For instance, hemispheric asymmetries in the DMN could be used to identify adolescents at high risk for depression, enabling early intervention. Machine learning algorithms integrating these biomarkers with clinical data could improve diagnostic accuracy and stratification of patients. In terms of treatment, the network-specific disruptions identified in this study provide a foundation for personalized therapeutic approaches. Furthermore, longitudinal monitoring of connectivity changes via rs-fMRI could serve as an objective measure of treatment efficacy, facilitating timely adjustments to therapeutic plans. Future research should prioritize integrating these biomarkers into machine learning frameworks and clinical trials to validate their predictive and therapeutic value.

This study, however, has several limitations that influence the validity and generalizability of the findings. These limitations include potential biases and imprecision, which could influence the interpretation and the significance of the results presented. For instance, the conclusions on hemispheric asymmetries and network dysfunction highlight distinct neural patterns associated with adolescent depression, however, the cross-sectional design of this study restricts causal inferences. This limitation requires longitudinal data to determine if these observed neural patterns are consequences or precursors of depressive symptoms. Consequently, future studies should incorporate more diverse participant samples and longitudinal study designs to deepen our understanding of the complex neural mechanisms involved in adolescent depression.

CONCLUSION

This study underscores the critical role of brain network connectivity in adolescent depression, with the results revealing significant functional discrepancies across key brain networks. The findings highlight the potential of using neuroimaging techniques to identify neural biomarkers that could facilitate early diagnosis and guide the development of personalized treatment approaches. The hemispheric asymmetries observed in the visual, DMN, attention, and somatomotor networks suggest that adolescent depression is characterized by a complex pattern of network- and hemisphere-specific functional changes. These findings enhance our understanding of the neural mechanisms underlying adolescent depression and underscore the importance of considering hemispheric specialization in the development of targeted therapeutic interventions.

ACKNOWLEDGEMENTS

The authors thanks all the individuals recruited as study participants, as well as anonymous commentators for their criticisms and proposals.