Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Sep 1.
Published in final edited form as: J Affect Disord. 2021 Jun 5;292:319–327. doi: 10.1016/j.jad.2021.05.086

Emotion-Related Network Reorganization Following Fish Oil Supplementation in Depressed Bipolar Offspring: An fMRI Graph-Based Connectome Analysis

Wenbin Li 1,2,#, Du Lei 1,#, Maxwell J Tallman 1, L Rodrigo Patino 1, Qiyong Gong 2, Jeffrey R Strawn 1, Melissa P DelBello 1, Robert K McNamara 1
PMCID: PMC8282765  NIHMSID: NIHMS1711779  PMID: 34139404

Abstract

Introduction:

Mood disorders are associated with fronto-limbic structural and functional abnormalities and deficits in omega-3 polyunsaturated fatty acids including eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA). Emerging evidence also suggests that n-3 PUFA, which are enriched in fish oil, promote cortical plasticity and connectivity. The present study performed a graph-based connectome analysis to investigate the role of n-3 PUFA in emotion-related network organization in medication-free depressed adolescent bipolar offspring.

Methods:

At baseline patients (n=53) were compared with healthy controls (n=53), and patients were then randomized to 12-week double-blind treatment with placebo or fish oil. At baseline and endpoint, erythrocyte EPA+DHA levels were measured and fMRI scans (4 Tesla) were obtained while performing a continuous performance task with emotional and neutral distractors (CPT-END). Graph-based analysis was used to characterize topological properties of large-scale brain network organization.

Results:

Compared with healthy controls, patients exhibited lower erythrocyte EPA+DHA levels (p=0.0001), lower network clustering coefficients (p=0.029), global efficiency (p=0.042), and lower node centrality and connectivity strengths in frontal-limbic regions (p<0.05). Compared with placebo, 12-week fish oil supplementation increased erythrocyte EPA+DHA levels (p<0.001), network clustering coefficient (p=0.005), global (p=0.047) and local (p=0.023) efficiency, and node centralities mainly in temporal regions (p<0.05).

Limitations:

The duration of fish oil supplementation was relatively short and the sample size was relatively small.

Conclusions:

These findings provide preliminary evidence that abnormalities in emotion-related network organization observed in depressed high-risk youth may be amenable to modification through fish oil supplementation.

Keywords: Brain networks, Major depressive disorder, Bipolar I disorder, Omega-3 polyunsaturated fatty acids, Emotional regulation

1. Introduction

The initial onset of mood disorders, including bipolar I disorder and major depressive disorder (MDD), frequently occurs during the late childhood and adolescent period (Kessler et al., 2005; Perlis et al., 2009). This developmental period is associated with progressive structural and functional changes in fronto-limbic connectivity and reductions in amygdala (AMY) reactivity to emotional stimuli (Gee et al., 2013; Hare et al., 2008; Gerber et al., 2009; Glantz et al., 2007; Mills et al., 2014; Swartz et al., 2014; Sowell et al., 1999; Wu et al., 2016). Imaging evidence suggests that adolescents and adults with mood disorders exhibit fronto-limbic structural and functional abnormalities and elevated AMY reactivity to emotional stimuli (Altshuler et al., 2005; Beesdo et al., 2009; Connolly et al., 2017; Cullen et al., 2014; Cheng et al., 2018; Kaiser et al., 2015; Liu et al., 2019; Olsaysky et al., 2012; Tang et al., 2018; Rich et al., 2006; Pavuluri et al., 2007; Strakowski et al., 2011; Wang et al., 2009). These and other findings suggest that a disruption of typical fronto-limbic network developmental maturation may contribute to emotional dysregulation in mood disorders (Paus et al., 2008). Therefore, identifying modifiable risk factors could inform early intervention strategies for youth at high-risk for mood dysregulation.

Converging evidence suggests that the pathophysiology and potentially etiology of mood disorders is associated with a deficiency in omega-3 polyunsaturated fatty acids (n-3 PUFA), including eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) (Grosso et al., 2016; Hibbeln, 1998; Noaghiul and Hibbeln, 2003; Grosso et al., 2014; Mocking et al., 2016; Sublette et al., 2011; Sarris et al., 2012). This is supported in part by evidence that patients with MDD or bipolar I disorder exhibit lower blood EPA and/or DHA levels compared with healthy subjects (McNamara & Welge, 2016; McNamara et al., 2014, 2015, 2016; Lin et al., 2010; Pomponi et al., 2013). In the mammalian brain, DHA levels increase rapidly in the frontal cortex during peri-adolescent development (Carver et al., 2001), and animal studies have demonstrated that developmental deficits in brain DHA accrual reduce synaptic density, connectivity, and plasticity (Cao et al., 2009; Carbone et al., 2020; de Velasco et al., 2012; Grayson et al., 2014; Moreira et al., 2010). Moreover, a recent study of healthy older adults found that blood EPA+DHA levels were associated with resting-state functional connectivity and network efficiency in key emotion-regulating regions including the prefrontal cortex and AMY (Talukdar et al., 2019). Together, these findings suggest that n-3 PUFA deficiency may represent a plausible and modifiable risk factor for emotion-related network dysregulation associated with mood disorders.

Graph theoretical analysis is increasingly being used to characterize topological properties of brain network organization and connectivity (Bullmore and Sporns, 2009; Rubinov and Sporns, 2010; Zalesky et al., 2010), and recent graph-based studies have identified abnormal network organization in different psychiatric disorders (Chen et al., 2019; Yu et al., 2020; Wang et al., 2017). The present study performed a graph-based analysis of emotion-related network organization in antidepressant-free adolescents with a depressive disorder and a biological parent with bipolar I disorder i.e., at high-risk for progressive mood dysregulation (Axelson et al., 2015; Ratheesh et al., 2017; Mortensen et al., 2003) and healthy controls. Patients were then randomized to 12-weeks of, double blind, placebo-controlled fish oil supplementation trial. fMRI scans were acquired during performance of a continuous performance task with emotional and neutral distracters (CPT-END) which engages prefrontal and limbic brain regions (Yamasaki et al., 2002). Erythrocyte n-3 PUFA levels were measured at baseline and endpoint. We hypothesized that emotion-related network organization would be abnormal in depressed high risk youth compared with healthy controls, and that fish oil supplementation would increase emotion-related network organization and efficiency.

2. Materials and Methods

2.1. Participants

This study was approved by the Institutional Review Boards of University of Cincinnati Medical Center and was registered at clinicaltrials.gov with identifier NCT00917501. Written informed consent was obtained from each parent and written assent was obtained from patients younger than 18 years of age. Healthy controls did not have a personal or family history of a DSM-IV-TR (APA, 2000) Axis I disorder, and patients had a current DSM-IV-TR (diagnosis of MDD or Depressive Disorder NOS (operationalized as 4 of 5 criteria for a major depressive episode or meeting all MDD criteria except duration), determined by the Washington University in St. Louis Kiddie Schedule for Affective Disorders and Schizophrenia, WASH-U-KSADS (Geller et al., 2001), a Childhood Depression Rating Scale-Revised Version (CDRS-R) (Poznanski et al., 1979, 1983) raw score of ≥40, and at least one biological parent with bipolar disorder, type I, as determined by the Structured Clinical Interview for DSM Disorders (SCID)(First et al., 1995). Diagnostic instruments were administered by trained clinicians with established diagnostic reliability (κ>0.9).

Full patient inclusion and exclusion criteria are described elsewhere (McNamara et al., 2020). All study participants were between the ages of 9–21 years and were excluded by a diagnosis of substance dependence within the previous 3 months, major medical or neurological illness (e.g., head trauma resulting in loss of consciousness), if female a positive urine pregnancy test, and an IQ<70 estimated by the Wechsler Abbreviated Scale of Intelligence (WASI) (Wechsler, 1997). Patients were excluded by use of antipsychotics, other mood stabilizers, stimulants, or atomoxetine within 72 hours (aripiprazole within two weeks was exclusionary because of its long half-life) or antidepressants within 5 days (fluoxetine within one month was exclusionary because of its long half-life); concomitant use of other psychotropic medications or medications with central nervous system (CNS) effects within 5 half-lives from baseline or prior treatment with a medication with CNS effects that requires more than 5 days of a screening period; left-hand dominance; or a contraindication to MRI scans (e.g., metal clips, braces or claustrophobia).

2.2. Symptom ratings

Efficacy results of this trial are detailed elsewhere (McNamara et al., 2020), and only primary efficacy outcome measures were investigated in the current study. Depression symptom severity was determined using the CDRS-R (Poznanski et al., 1979, 1983) and global change in illness severity was assessed using the Clinical Global Impression-Severity Scale (CGI-S)(Guy, 1970). Scales were administered by a child and adolescent psychiatrist or psychologist who was blinded to treatment group and with established inter-rater reliabilities (kappa >0.9).

2.3. Intervention

Patients took three placebo or fish oil capsules daily which were identical in size, shape, and color to protect the blind. Each fish oil capsule contained 450 mg EPA, 40 mg DPA, and 260 mg DHA for a total daily dose of 2,130 mg EPA+DHA (1.7:1 EPA/DHA ratio) or 2,250 mg n-3 PUFAs (EPA+DPA+DHA). Each placebo capsule contained linoleic acid (92 mg), stearic acid (23 mg), palmitic acid (114 mg), and oleic acid (746 mg). At baseline and endpoint, erythrocyte (red blood cell) membrane fatty acid composition (mg fatty acid/100 mg fatty acids) was determined by gas chromatography (Shimadzu GC-2010, Shimadzu Scientific Instruments Inc., Columbia MD USA). Primary measures of interest were EPA+DHA (‘omega-3 index’), arachidonic acid (AA, 20:4n-6), and the AA/EPA+DHA ratio (McNamara et al., 2015).

2.4. fMRI

2.4.1. Image acquisition

fMRI scans were performed at the University of Cincinnati’s Center for Imaging Research using a 4.0 Tesla Varian Unity INOVA Whole Body MRI/MRS system (Varian Inc., Palo Alto, CA). Anatomical T1-weighted, 3-D brain scan was obtained using a modified driven equilibrium Fourier transform (MDEFT) sequence (TMD = 1.1 s, TR = 13 ms, TE = 6 ms, FOV = 256 × 256 × 192 mm, matrix 256 × 256 × 192 pixels, flip angle = 20 degrees). A midsagittal localizer scan was obtained to place 40 contiguous 4 mm axial slices that extend from the inferior cerebellum to encompass the entire brain. Subjects then completed a fMRI session while performing the (CPT-END)(Yamasaki et al., 2002) using a T2*-weighted gradient-echo echoplanar imaging (EPI) pulse sequence (TR/TE = 2000/30 ms, FOV = 256 × 256 mm, matrix 64 × 64 pixels, slice-thickness = 4 mm, flip angle = 75 degrees). The CPT-END is a visual oddball paradigm comprised of three task conditions and one control condition, and a fixation cross was presented for one second between cues. Visual stimuli were presented using high-resolution video goggles (Resonance Technologies, Inc., Northridge, CA). The emotional and neutral pictures were presented pseudorandomly. 70% of the visual cues are simple colored squares (controls), 10% are simple colored circles (targets), 10% are emotionally neutral pictures, and 10% are emotionally unpleasant pictures. The neutral and emotional pictures were taken from the International Affective Picture System (IAPS, University of Florida) based on the rating criteria (Yamasaki et al., 2002). Each visual cue required a response; the circles (targets) required the unique response (button 2), whereas the squares and pictures all required the same response (button 1). Each imaging session consisted of two runs of 158 visual cues per run presented at three-second intervals (stimulus-onset asynchronies) for two seconds each.

2.4.2. Data preprocessing

Preprocessing of the fMRI data was performed using Statistical Parametric Mapping (SPM12, Wellcome Center for Human Neuroimaging, London, UK) and SPM-based Conn Toolbox 2018b (McGovern Institute for Brain Research, MIT, Cambridge, MA) (Whitfield-Gabrieli and Nieto-Castanon, 2012), running in MATLAB (The Mathworks Inc.; MA, USA) platform. The steps included realignment, slice-timing correction, co-registration to structural T1 scan, spatial normalization to Montreal Neurological Institute coordinates (MNI) space, spatial smoothing (8-mm Gaussian kernel) and band-pass filtering (0.009 < f < 0.08 Hz). The T1 scans were segmented into gray matter (GM), white matter (WM), and cerebrospinal fluid (CSF) tissue classes. The component-based noise correction known as aCompCor (White matter and CSF ROIs, 5 components each) (Muschelli et al., 2014). Artifact Detection Tools (ART)-based scrubbing (as many regressors as identified invalid scans) and motion regression (12 regressors: 6 motion parameters plus 6 first-order temporal derivatives) were used for control of physiological and movement confounds.

Automated Anatomical Labeling (AAL) 90 template were used as the nodes to sample the whole brain. The regions of interests (ROIs) were defined as spheres, with a radius of 8 mm. We first performed voxel-wise general linear model (GLM) analysis to examine task activations and extract ROI time series. The task regressors were calculated by convolving a box-car function of the task design with the canonical hemodynamic response function (HRF) in SPM. After model estimation, a contrast was defined for CPT-END task to obtain task related activations for the emotional stimulus condition compared with the respective neutral stimulus condition. For each of the 90 ROIs, the first eigenvariate was extracted using SPM’s volume of interest function. We additionally regress out 12 head motion regressors to minimize the effects of head motion on subsequent analyses (Friston et al., 1996). The time series of a ROI was then deconvolved with the HRF, and point-by-point multiplied with the psychological design variable (Di and Biswal, 2017, 2019). The PPI (Psychophysiological Interactions) term of emotion – neutral stimulus condition was calculated. The neuronal level PPI terms were then convolved with the HRF to represent the BOLD level predictors of PPI effects. Similar GLM models were built to examine PPI effects, with 1 regressor of the time series of a ROI, 4 regressors of task conditions (circle, square, emotion and neutral), 1 regressors of PPI effects, 12 head motion regressors and 1 constant term. The contrasts of interest were the PPI effects. The PPI effects were calculated between each pair of the 90 ROIs, which resulted in a 90 × 90 matrix for each subject. The PPI effects matrix for each subject was symmetrized by averaging the upper and lower triangular matrix and diagonal elements of the matrix was set as 0. The PPI effects matrixes were used for next large-scale brain network analysis. The Graph Theoretical Network Analysis (GRETNA, http://www.nitrc.org/projects/gretna/) toolbox (Wang et al., 2015) and Network Based Statistic (NBS, https://sites.google.com/site/bctnet/comparison/nbs) toolbox (Zalesky et al., 2010) were used for the graph theoretical analysis.

2.4.3. Data analyses

The weighted connection matrix was used to calculate the graph metrics to characterize the topological architectures of the whole brain network. The weighted connection matrix was thresholded using a range of sparsity threshold (0.1 < S < 0.5) with a step of 0.05 which resulted 10 different network architectures for each subject. For the weighted network at each sparsity level, the global and nodal network metrics were calculated. The area under the curve (AUC) for the whole range of sparsity was calculated to represent the average value for 10 different network architectures. For the global metrics, the clustering coefficient (Cp), characteristic path length (Lp) and the efficiency including global (Eglobal) and local (Elocal) efficiency were calculated (Latora and Marchiori, 2001). For nodal metrics, the topological centrality including degree and betweenness centrality and the nodal efficiency were calculated. The global metrics were used to characterize the network topological properties and to measure how efficiently the information transformed in the network while the nodal metrics focused on the topological importance of each nodes in the network. A full description and definition of all graphic measures used in current study is provided as supplementary material. NBS was used to investigate edges representing connections among nodes in the network (Zalesky et al., 2010). A nonparametric permutation test (10,000 permutations, p<0.05 and t>3.1) was used to test the cross-sectional group-wise differences and longitudinal changes of the network connections.

2.5. Statistical analyses

Statistical analyses of tabulated results were performed using R software (Version 3.6.0, http://www.r-project.com). Differences in topological metrics between patients and healthy controls at baseline were analyzed using two sample permutation tests with DAAG package in R (10,000 permutations, p<0.05). Group differences in longitudinal changes in topological measurements were also tested using the same permutation algorithm. To control for multiple comparisons and enhance the robustness in nodal metrics results, we only report those nodal centrality abnormalities found in at least two centrality measurements and exhibit the same directionality. This correction method has been used in prior studies to balance risk of false positive and negative findings (Lei et al., 2020, 2021; Lei et al., 2021). Exploratory correlational analyses were performed to assess associations between baseline clinical measurements (symptom scores and fatty acid levels) and graph metrics. We also evaluated the relationship between changes in graph metrics and clinical measures in fish oil and placebo groups separately. Pearson’s correlation analyses were used for continuous variables (i.e., CDRS-R scores, fatty acid measures). For CGI-S, patients were stratified into two subgroups (moderately ill, CGI-S ≤4 vs. markedly or severely ill, CGI-S 5 or 6) according to their CGI-S scores, and two-sample permutation t-tests were used to evaluate relationships with graphic metrics. As a supplementary analysis, we performed whole brain voxel-wise two-sample t-tests between MDD patients and healthy controls at baseline and longitudinal activation changes between treatment groups using the emotion – neutral contrast. All statistical analyses were two-sided.

3. Results

3.1. Cross-sectional analysis

A total of 53 healthy controls and 53 patients were included in the baseline analysis, and demographic characteristics are presented in Table 1. Patients and controls were closely matched for age, gender, and race, and patients exhibited significantly lower erythrocyte EPA+DHA levels (−24%, p≤0.0001), and a significantly higher AA/EPA+DHA ratio (+18%, p=0.002), compared with controls. For CPT-END performance, there were no significant group differences for percent correct (Controls: 96±9.7% vs. Patients: 97±3.3%, p=0.52) or reaction time (Controls: 891±190 vs. Patients: 884±202 msec, p=0.87) across all cues. For global metrics, patients exhibited significantly lower Cp (p=0.029) and Eglobal (p=0.042) than controls and no significant differences were observed for Lp (p=0.055) or Elocal (p=0.081)(Fig. 1A). Patients exhibited significantly lower node topological centralities in bilateral medial superior frontal cortex, right amygdala, left medial orbital frontal cortex, bilateral lingual gyrus, left middle frontal cortex, left hippocampus, left inferior orbital frontal cortex, left fusiform, left calcarine, left putamen and left precentral cortex in at least two topological measurements, and patients did exhibit any node that had greater topological centralities compared with healthy controls (Table 2). For node connectivity, patients showed significant weaker connections in a subnetwork involving 57 nodes and 121 connections compared with healthy controls (p<0.05, NBS corrected), and there were no stronger node connectivity observed (Fig. 2). The majority (14/15) of nodes that exhibited lower topological centralities were localized to this subnetwork. Patients also exhibited lower brain activation in midline brain regions and greater activation mainly in motor and sensory-related regions during negative emotional stimulus compared with healthy controls (Fig. S1 and Table S1).

Table 1.

Case-Control Demographics

Variable1 HC (n=53) MDD (n=53) P-value2
Age, years 14.4±0.4 14.3±0.4 0.9
Gender, n (%) female 42 (79) 42 (79) 1.0
Race, n (%) White 35 (66) 35 (66) 1.0
BMI, kg/m2 22.5±5.0 25.1±7.8 0.1
WASI 105.7±14.0 101.3±11.9 0.1
Erythrocyte Fatty acids (wt % TTL)
 EPA+DHA 4.2±1.2 3.2±0.7 0.0001
 AA 18.2±1.3 17.6±1.5 0.14
 AA/EPA+DHA 4.6±1.3 5.6±1.0 0.002
Open in a new tab
1

Values are group mean ± S.D. or number of subjects (n) and percent (%).

2

t-tests or X2

Figure 1.

Figure 1.

Open in a new tab

Baseline global graphic metric differences between depressed patients (MDD) and healthy controls (HC)(A) in response to negative emotional stimuli, and longitudinal differences between patients receiving 12-week fish oil (FO) or placebo (PLC)(B). Abbreviations: Cp, clustering coefficients; Eglobal, global network efficiency; Elocal, local network efficiency.

Table 2.

Case-Control Differences in Nodal Centrality

Region Centrality type HC MDD P-Value*
R medial superior frontal cortex Node Degree 8.6 (7.1) 4.9 (3.7) 0.0006
R amygdala Node Degree 10.4 (7.6) 6.1 (5.9) 0.0014
L medial orbital frontal cortex Node Degree 9.3 (7.4) 5.9 (4.4) 0.0052
L lingual gyrus Node Degree 10.3 (8.5) 6.6 (4.7) 0.0056
L middle frontal cortex Node Degree 7.0 (4.5) 4.8 (4.3) 0.0124
L hippocampus Node Degree 8.3 (5.6) 5.7 (4.5) 0.0134
L inferior orbital frontal cortex Node Degree 8.1 (8.7) 5.2 (3.5) 0.0172
L fusiform gyrus Node Degree 7.4 (6.7) 5.1 (3.2) 0.0202
L pallidum Node Degree 10.7 (8.0) 7.3 (6.7) 0.0216
R postcentral gyrus Node Degree 7.5 (5.8) 5.4 (3.3) 0.0184
L calcarine gyrus Node Degree 8.9 (7.6) 6.3 (4.6) 0.0308
L putamen Node Degree 9.9 (10.2) 6.4 (6.2) 0.0318
L medial superior frontal cortex Node Degree 8.7 (6.9) 6.3 (5.0) 0.0344
R lingual gyrus Node Degree 9.0 (7.4) 6.5 (5.1) 0.045
L precentral cortex Node Degree 6.8 (5.3) 5.0 (3.7) 0.0476
L medial orbital frontal cortex Node Efficiency 0.24 (0.1) 0.19 (0.08) 0.0048
L lingual gyrus Node Efficiency 0.24 (0.11) 0.19 (0.08) 0.0066
R amygdala Node Efficiency 0.24 (0.1) 0.19 (0.09) 0.0074
R medial superior frontal cortex Node Efficiency 0.22 (0.1) 0.18 (0.08) 0.01
L calcarine gyrus Node Efficiency 0.23 (0.11) 0.19 (0.08) 0.011
L inferior orbital frontal cortex Node Efficiency 0.22 (0.11) 0.18 (0.07) 0.0124
L hippocampus Node Efficiency 0.23 (0.09) 0.18 (0.08) 0.0154
L pallidum Node Efficiency 0.25 (0.11) 0.2 (0.1) 0.0176
L putamen Node Efficiency 0.24 (0.12) 0.19 (0.1) 0.0182
R lingual gyrus Node Efficiency 0.24 (0.1) 0.19 (0.09) 0.019
L fusiform gyrus Node Efficiency 0.22 (0.1) 0.18 (0.07) 0.019
L middle frontal cortex Node Efficiency 0.2 (0.08) 0.17 (0.08) 0.0232
L precentral cortex Node Efficiency 0.21 (0.09) 0.17 (0.07) 0.0234
L postcentral cortex Node Efficiency 0.21 (0.09) 0.18 (0.08) 0.035
L medial superior frontal cortex Node Efficiency 0.23 (0.1) 0.2 (0.09) 0.0496
L medial superior frontal cortex Node Betweenness 28.9 (61.5) 10.1 (14) 0.0036
L putamen Node Betweenness 22.3 (43.5) 8.4 (14) 0.01
R amygdala Node Betweenness 52.5 (88.9) 20.9 (47) 0.0152
L medial superior frontal cortex Node Betweenness 28.3 (40.9) 14.2 (29) 0.0448
Open in a new tab
*

P values were calculated by two-sample permutation tests. Abbreviations: L left; R right.

Figure 2.

Figure 2.

Open in a new tab

Baseline differences in network node connectivity in depressed patients compared with the healthy controls. Depressed patients exhibited a subnetwork involving 57 nodes (blue nodes) and 121 connections (green lines) that were significantly lower compared with healthy controls (p<0.05, NBS corrected). The majority (14/15) of the nodes that exhibited decreased topological centralities (red nodes) were in this subnetwork.

There were no significant correlations between graphic metrics and continuous measures of depressive symptoms (CDRS-R) or fatty acid levels. For CGI-S, compared with moderately ill (n=29) patients severally ill patients (n=24) exhibited decreased Eglob and Elocal; decreased node degree in left orbital inferior frontal cortex, left putamen and left precentral cortex, decreased node efficiency in left orbital medial frontal cortex, bilateral lingual gyrus, right amygdala, left calcarine cortex, left orbital inferior frontal cortex, left hippocampus, left pallidum, left putamen, right lingual gyrus and left precentral cortex (Fig. S2).

3.2. Fish oil supplementation

A total of 42 patients completed the 12-week trial (Placebo, n=21; Fish oil n=21), and a total of 39 patients had usable baseline and endpoint fMRI data (Placebo, n=18; Fish oil, n=21). At baseline there were no significant treatment group differences in demographic, clinical, and fatty acid variables (Table 3). For erythrocyte fatty acids, significant group by time (baseline, week 12) interactions were observed for EPA (p=0.001), DPA (p=0.001), DHA (p=0.001), EPA+DHA (omega-3 index)(p=0.001), arachidonic acid (AA) (p=0.009), and the AA/EPA+DHA ratio (p=0.0001)(Table 3). There were no significant treatment group by time interactions for CDRS-R total score (p=0.414)(baseline-endpoint change: Placebo: −41%, p≤0.0001; Fish oil: −45%, p≤0.0001), and a significant group by time interaction was observed for CGI-S scores (p=0.015) (Placebo: −30%, p≤0.0001; Fish oil: −44%, p≤0.0001)(Table 3). For CPT-END performance, across all cues there were no significant group differences in percent correct at baseline (Placebo: 96.3±4.2% vs. Fish oil: 97.3±2.3%, p = 0.34) or endpoint (Placebo: 95.2±4.8% vs. Fish oil: 96.5±3.4%, p = 0.34), or reaction time at baseline (Placebo: 888±185 vs. Fish oil: 852±222 msec, p=0.88) or endpoint (Placebo 828±175 vs. Fish oil: 819±194 msec, p=0.89).

Table 3.

Patient Demographic and Clinical Characteristics

Placebo
(n=18)		 Fish Oil
(n=21)
Variable1 Pre Post Pre Post P-value2 P-value3
Demographics
 Age, years 14.3 (3.2) 14.3 (3.2) 14.4 (2.8) 14.4 (2.8) 0.8 ―
 Gender, n (%) female 14 (78) 14 (78) 17 (80) 17 (80) 1.0 ―
 Race, n (%) White 11 (61) 11 (61) 14 (67) 14 (67) 0.6 ―
 WASI, IQ 101.1 (12.3) 101.1 (12.3) 101.5 (11.8) 101.5 (11.8) 0.9 ―
 Social Status 3 (1.2) 3 (1.2) 2.9 (1.4) 2.9 (1.4) 0.9 ―
 Diagnosis, MDD/NOS 13/5 13/5 16/5 16/5 1.0 ―
Symptom Ratings
 CDRS-R 47.0 (8.7) 28.95 (6.4) 46.0 (6.6) 25.3 (7.2) 0.7 0.81
 CGI-S 4.6 (0.6) 3.3 (0.8) 4.4 (0.6) 2.4 (1.0) 0.4 0.015
Erythrocyte Fatty Acids 4
 EPA 0.17 (0.2) 0.16 (0.67) 0.21 (0.2) 1.3 (0.7) 0.4 <0.001
 DPA 2.3 (0.6) 2.1 (0.5) 2.2 (0.4) 3.5 (1.3) 0.5 <0.001
 DHA 3.1 (0.7) 2.8 (0.4) 3.0 (0.6) 5.1 (1.2) 0.5 <0.001
 EPA+DHA 3.3 (0.7) 3.0 (0.4) 3.3 (0.6) 6.3 (1.7) 0.8 <0.001
 AA 17.9 (1.2) 18.2 (2.0) 17.3 (1.7) 16.2 (2.4) 0.2 0.009
 AA/EPA+DHA 5.5 (0.9) 6.1 (0.8) 5.4 (1.0) 2.8 (1.0) 0.7 <0.001
Open in a new tab
1

Values are group mean ± S.D. or number of subjects (n) and percent (%).

2

t-tests or X2, baseline Placebo vs. n-3 PUFA.

3

Group x Time Interaction

4

Expressed as composition (mg fatty acid/100 mg fatty acids)

For global network metrics, patients in the fish oil group exhibited significantly greater increases in Cp (p=0.005), Eglobal (p=0.047) and Elocal (p=0.023), but not Lp (p=0.34), compared with the placebo group (Fig. 1B). For nodal measurements, greater increases in nodal topological centrality were observed in the fish oil group in the left superior temporal pole, right middle temporal gyrus, right superior temporal gyrus, left precentral cortex, right rolandic operculum, left hippocampus, right putamen, left cuneus, left middle temporal pole, left inferior parietal gyrus, bilateral heschl gyrus and left precuneus compared with the placebo group (Table 4). No significant decreases of nodal centrality measures were observed in the fish oil group compared with the placebo group. For network connection measures, no significant differences were found in the fish oil group compared with the placebo group after NBS correction, and no significant correlations were observed between changes in graphic metrics and symptom or fatty acid measures in fish oil or placebo group. There were no differences in activation changes between fish oil and placebo group in response to negative emotional stimulus.

Table 4.

Longitudinal Changes in Nodal Centrality

Region Centrality type Placebo Fish oil P-Value*
L superior temporal pole Node Degree 1.47 (4.0) −4.39 (5.3) 0.0002
R superior temporal gyrus Node Degree 3.69 (7.4) −3.98 (8.1) 0.0032
R middle temporal gyrus Node Degree 1.87 (5.6) −4.82 (7.8) 0.0036
R putamen Node Degree 1.37 (6.6) −4.24 (6.1) 0.0076
L hippocampus Node Degree 1.02 (7.0) −6.81 (11.1) 0.0108
L inferior parietal gyrus Node Degree 1.08 (6.5) −3.66 (5.8) 0.0202
R rolandic operculum Node Degree 0.69 (6.5) −5.55 (9.5) 0.021
L cuneus Node Degree 0.44 (8.4) −7.94 (12.8) 0.022
L middle temporal pole Node Degree 1.69 (4.6) −2.92 (8.2) 0.0352
L precentral cortex Node Degree −1.23 (4.3) −5.54 (7.5) 0.0362
R heschl gyrus Node Degree −0.11 (8.4) −8.88 (16.8) 0.044
L heschl gyrus Node Degree −1.55 (5.6) −8.09 (13.1) 0.0482
L superior temporal pole Node Efficiency 0.018 (0.1) −0.098 (0.1) 0.0014
R middle temporal gyrus Node Efficiency 0.026 (0.1) −0.088 (0.1) 0.0018
R superior temporal gyrus Node Efficiency 0.036 (0.12) −0.09 (0.12) 0.0022
L precentral cortex Node Efficiency −0.016 (0.11) −0.11 (0.11) 0.009
R rolandic operculum Node Efficiency 0.011 (0.12) −0.10 (0.14) 0.009
L hippocampus Node Efficiency 0.0058 (0.13) −0.11 (0.13) 0.009
R putamen Node Efficiency 0.00018 (0.12) −0.098 (0.11) 0.0118
L cuneus Node Efficiency −0.010 (0.12) −0.12 (0.15) 0.0144
L middle temporal pole Node Efficiency 0.011 (0.091) −0.07 (0.11) 0.0154
L inferior parietal gyrus Node Efficiency 0.012 (0.11) −0.08 (0.13) 0.0184
L heschl gyrus Node Efficiency −0.03 (0.11) −0.12 (0.17) 0.0348
R heschl gyrus Node Efficiency −0.0049 (0.13) −0.12 (0.17) 0.0232
L precuneus Node Efficiency −0.022 (0.12) −0.11 (0.15) 0.0404
L hippocampus Node betweenness 43.56 (94.5) −58.66 (131) 0.0022
R putamen Node betweenness 34.36 (59.9) −15.64 (45) 0.0034
L precuneus Node betweenness 8.21 (19.9) −6.05 (15) 0.0038
Open in a new tab
*

P values were calculated by two-sample permutation tests. Abbreviations: L left; R right.

4. Discussion

This graph-based connectome analysis investigated whole brain network organization in antidepressant-free depressed adolescent bipolar offspring prior to and following 12-week double-blind treatment with fish oil or placebo. At baseline, patients exhibited lower erythrocyte n-3 PUFA levels, clustering coefficients, global network efficiency, and nodal centrality in fronto-limbic regions compared with healthy controls, as well as a significant subnetwork node disconnection pattern. Patients with greater illness severity, as assessed with the CGI-S, exhibited lower global and local efficiency and node centrality in fronto-limbic regions compared with moderately ill patients. Compared with placebo, fish oil supplementation significantly increased erythrocyte n-3 PUFAs and increased network clustering coefficients, global and local network efficiency and increased nodal centralities primarily in temporal regions. Together, these results suggest that depressed bipolar offspring exhibit low n-3 PUFA biostatus and emotion-related network abnormalities that differ from typically developing adolescents, and that increasing in n-3 PUFA biostatus through fish oil supplementation alter patient network organization by increasing network efficiency, function segregation, and nodal centralities primarily in temporal regions.

Consistent with prior cross-sectional studies (McNamara et al., 2014, 2015, 2016), youth with depression exhibited lower erythrocyte EPA+DHA levels, and a higher AA/EPA+DHA ratio, compared with healthy adolescents. In support of our first hypothesis, patients exhibited lower network clustering coefficient, global efficiency and node centrality in fronto-limbic regions compared with healthy controls. These findings are congruent with previous studies that observed abnormalities in networks related to emotion regulation in MDD and bipolar patients (Yu et al., 2020; Wang et al., 2017; Kaiser et al., 2015; Li et al., 2018). For example, Wang et al. (2017) found that both MDD and BD patients exhibited lower global efficiency compared with controls. Lower global network efficiency suggests disrupted information processing integration and less efficient communication between nodes in the network (Rubinov and Sporns, 2010). Consistent with abnormal emotion network organization, patients exhibited lower node centrality in fronto-limbic regions and a node disconnection pattern within a subnetwork that included lower node topological centralities. Lastly, patients with greater illness severity, as assessed with the CGI-S, exhibited lower global and local efficiency and node centrality in fronto-limbic regions compared with moderately ill patients. Together, these findings suggest that depressed bipolar offspring exhibit a pattern of emotion-related network abnormalities that is consistent with fronto-limbic hypoconnectivity.

In support of our second hypothesis, patients who received fish oil supplementation exhibited greater increases in network functional segregation and local and global efficiency compared with placebo. This result is generally congruent with prior evidence that resting-state network efficiency is associated with blood EPA+DHA levels in healthy older adults (Talukdar et al., 2019). While the observed pattern of changes in these measures were in the direction of healthy controls, it notable that most changes in nodal centrality were localized to temporal cortex regions thought to play a role in the integration, processing, and storage of emotional representations (Dzafic et al., 2019 Fusar-Poli et al., 2009; Olson et al., 2013). Although translational evidence suggests that higher n-3 PUFA levels are associated with greater synaptic structural and functional connectivity in temporal lobe regions including the hippocampus (Cao et al., 2009; Grayson et al., 2014; He et al., 2009; Talukdar et al., 2019), we did not observe differences between fish oil and placebo groups in node connectivity strengths. The latter suggests that network functional segregation and efficiency are more sensitive to the effects of increasing n-3 PUFA levels, and may precede longer-term adaptive changes in node connectivity strength. It is also notable that we did not observe significant linear relationships between graphic metrics and measures of depression symptom severity, and this may be due in part to the fact that both treatment groups exhibited similar and significant reductions in depression severity over the trial (McNamara et al., 2020). While additional research is needed to decipher the central mechanisms underlying the beneficial effects of n-3 PUFA on depressive symptoms, the present results are generally consistent with findings that other therapeutic interventions for depression, including antidepressant medications and family focused therapy (Abdallah et al., 2017; Brakowski et al., 2017; Li et al., 2018; Singh et al., 2020) are also associated with network reorganization.

The present study has several limitations. First, the sample size was relatively small, particularly for the correlation analyses, and larger studies are warranted to replicate and extend the current findings. Second, the functional brain networks constructed from the functional MR imaging data were largely restricted to anatomic pathways, and combining neuroimaging modalities could enhance our understanding of the structure-function relationships (Damoiseaux 2009; Honey et al., 2009). Third, the duration of fish oil supplementation was relatively short (12 weeks), and more robust or differential changes in network organization might emerge following longer treatment. Fourth, the impact of factors associated with adverse clinical features and illness course, including childhood maltreatment (Agnew-Blais et al., 2016; Serafini et al., 2017), were not evaluated. Study strengths include a well-characterized cohort of antidepressant-free depressed high-risk youth, demographically similar comparator groups, and the randomized double-blind placebo-controlled study design.

We present novel evidence that depressed bipolar offspring exhibit abnormal emotion-related network organization and that increasing n-3 PUFA biostatus increases emotion-related network transferring efficiency and functional segregation. To our knowledge, this is the first prospective controlled trial to investigate the effects of fish oil supplementation on whole brain network organization in depressed youth. Together these findings add to a growing body of evidence implicating n-3 PUFA in cortical circuit plasticity and preliminary support for the broader view that n-3 PUFA insufficiency may be relevant to the etiology of abnormal emotional network organization in depressed patients. These findings encourage additional research to determine whether fish oil supplementation prior to the onset of mood symptoms can mitigate abnormalities in emotion-related network organization and progressive mood dysregulation in high-risk youth.

Supplementary Material

1

Highlights.

  • Mood disorders are associated with structural and functional connectivity abnormalities.

  • Omega-3 polyunsaturated fatty acids (n-3 PUFA) promote cortical plasticity and connectivity.

  • A graph-based connectome analysis revealed that emotion-related network organization is altered in depressed bipolar offspring.

  • Supplementation with n-3 PUFA changes emotion-related network organization and connectivity in depressed bipolar offspring.

Acknowledgments

Role of the funding source

Funding agencies had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, writing, review, or approval of the manuscript; and decision to submit the manuscript for publication.

This trial was supported in part by R34 NIH/NIMH grant MH083924 to R.K.M and M.P.D (Co-PIs); the National Natural Science Foundation (Grants 81621003-81820108018); Program for Changjiang Scholars and Innovative Research Team in University (PCSIRT, Grant No. IRT16R52) of China, and the Functional and Molecular Imaging Key Laboratory of Sichuan Province (FMIKLSP, Grant: 2019JDS0044). Funding agencies had no further role in study design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the paper for publication. The authors thank the Inflammation Research Foundation, Marblehead, MA USA for providing the FO and placebo capsules.

R.K.M. has received research support from Martek Biosciences Inc, Royal DSM Nutritional Products, LLC, Inflammation Research Foundation, Ortho-McNeil Janssen, AstraZeneca, Eli Lilly, NARSAD, and national institutes of health (NIH), and previously served on the scientific advisory board of the Inflammation Research Foundation. J.R.S. has received research support from Edgemont, Shire, Neuronetics, Otsuka, Allergan and NIH and received material support from and served as a consultant to Assurex/Genesight. He receives royalties from Springer Publishing and UpToDate and has received honoraria from CMEology and Current Psychiatry. M.P.D. receives research support from NIH, PCORI, Acadia, Allergan, Janssen, Johnson and Johnson, Lundbeck, Otsuka, Pfizer, and Sunovion. She is also a consultant, on the advisory board, or has received honoraria for speaking for Alkermes, Allergan, Assurex, CMEology, Janssen, Johnson and Johnson, Lundbeck, Myriad, Neuronetics, Otsuka, Pfizer, Sunovion, and Supernus.

Footnotes

Disclosures

The remaining authors do not have disclosures.

Declaration of Competing Interest

All authors disclosed no specific conflicts of interests.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

REFERENCES

  1. Abdallah CG, Averill LA, Collins KA, et al. Ketamine treatment and global brain connectivity in major depression. Neuropsychopharmacology. 2017;42:1210–1219. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Agnew-Blais J, Danese A. Childhood maltreatment and unfavourable clinical outcomes in bipolar disorder: a systematic review and meta-analysis. Lancet Psychiatry. 2016;3:342–349 [DOI] [PubMed] [Google Scholar]
  3. Altshuler L, Bookheimer S, Proenza MA, Townsend J, Sabb F, Firestine A, Bartzokis G, Mintz J, Mazziotta J, Cohen MS. Increased amygdala activation during mania: a functional magnetic resonance imaging study. Am J Psychiatry. 2005;162:1211–1213. [DOI] [PubMed] [Google Scholar]
  4. American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders. (DSM-IV-TR) (Text Revision), 4th Edition 2000; Washington, DC: American Psychiatric Press. [Google Scholar]
  5. Axelson D, Goldstein B, Goldstein T, Monk K, Yu H, Hickey MB, Sakolsky D, Diler R, Hafeman D, Merranko J, Iyengar S, Brent D, Kupfer D, Birmaher B. Diagnostic precursors to bipolar disorder in offspring of parents with bipolar disorder: A longitudinal study. Am J Psychiatry. 2015;172:638–646. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Beesdo K, Lau JY, Guyer AE, McClure-Tone EB, Monk CS, Nelson EE, Fromm SJ, Goldwin MA, Wittchen HU, Leibenluft E, Ernst M, Pine DS. Common and distinct amygdala-function perturbations in depressed vs anxious adolescents. Arch Gen Psychiatry. 2009;66:275–285. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Brakowski J, Spinelli S, Dörig N, Bosch OG, Manoliu A, Holtforth MG, Seifritz E. Resting state brain network function in major depression - Depression symptomatology, antidepressant treatment effects, future research. J Psychiatr Res. 2017;92:147–159. [DOI] [PubMed] [Google Scholar]
  8. Bullmore E, Sporns O. Complex brain networks: graph theoretical analysis of structural and functional systems. Nat Rev Neurosci. 2009;10:186–198. [DOI] [PubMed] [Google Scholar]
  9. Cao D, Kevala K, Kim J, Moon HS, Jun SB, Lovinger D, Kim HY. Docosahexaenoic acid promotes hippocampal neuronal development and synaptic function. J Neurochem. 2009;111:510–521. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Carbone BE, Abouleish M, Watters KE, Vogel S, Ribic A, Schroeder OH, Bader BM, Biederer T. Synaptic connectivity and cortical maturation are promoted by the ω−3 fatty acid docosahexaenoic acid. Cereb Cortex. 2020;30:226–240. [DOI] [PubMed] [Google Scholar]
  11. Carver JD, Benford VJ, Han B, Cantor AB. The relationship between age and the fatty acid composition of cerebral cortex and erythrocytes in human subjects. Brain Res Bull. 2001;56:79–85. [DOI] [PubMed] [Google Scholar]
  12. Chen Y, Huang X, Wu M, Li K, Hu X, Jiang P, Chen L, He N, Dai J, Wang S, He M, Guo L, Sweeney JA, Gong Q. Disrupted brain functional networks in drug-naïve children with attention deficit hyperactivity disorder assessed using graph theory analysis. Hum Brain Mapp. 2019;40:4877–4887. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Cheng W, Rolls ET, Qiu J, Xie X, Lyu W, Li Y, Huang CC, Yang AC, Tsai SJ, Lyu F, Zhuang K, Lin CP, Xie P, Feng J. Functional connectivity of the human amygdala in health and in depression. Soc Cogn Affect Neurosci. 2018;13:557–568. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Connolly CG, Ho TC, Blom EH, LeWinn KZ, Sacchet MD, Tymofiyeva O, Simmons AN, Yang TT. Resting-state functional connectivity of the amygdala and longitudinal changes in depression severity in adolescent depression. J Affect Disord. 20171;207:86–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Cullen KR, Westlund MK, Klimes-Dougan B, Mueller BA, Houri A, Eberly LE, Lim KO. Abnormal amygdala resting-state functional connectivity in adolescent depression. JAMA Psychiatry. 2014;71:1138–1147. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Damoiseaux JS, Greicius MD. Greater than the sum of its parts: a review of studies combining structural connectivity and resting-state functional connectivity. Brain Struct Funct. 2009;213:525–533. [DOI] [PubMed] [Google Scholar]
  17. de Velasco PC, Mendonça HR, Borba JM, Andrade da Costa BL, Guedes RC, Navarro DM, Santos GK, Faria-Melibeu Ada C, Campello Costa P, Serfaty CA. Nutritional restriction of omega-3 fatty acids alters topographical fine tuning and leads to a delay in the critical period in the rodent visual system. Exp Neurol. 2012;234:220–229. [DOI] [PubMed] [Google Scholar]
  18. Di X, Biswal BB. Psychophysiological interactions in a visual checkerboard task: Reproducibility, reliability, and the effects of deconvolution. Front Neurosci. 2017;11:573. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Di X, Biswal BB. Toward task connectomics: Examining whole-brain task modulated connectivity in different task domains. Cereb Cortex. 2019;29:1572–1583. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Dzafic I, Oestreich L, Martin AK, Mowry B, Burianová H. Stria terminalis, amygdala, and temporoparietal junction networks facilitate efficient emotion processing under expectations. Hum Brain Mapp. 2019;40:5382–5396. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. First MB, Spitzer RL, Gibbon M, Williams JB: Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-IV). New York State, 1996. [Google Scholar]
  22. Friston KJ, Williams S, Howard R, Frackowiak RS, Turner R. Movement-related effects in fMRI time-series. Magn Reson Med. 1996;35:346–355. [DOI] [PubMed] [Google Scholar]
  23. Fusar-Poli P, Placentino A, Carletti F, Landi P, Allen P, Surguladze S, Benedetti F, Abbamonte M, Gasparotti R, Barale F, Perez J, McGuire P, Politi P. Functional atlas of emotional faces processing: a voxel-based meta-analysis of 105 functional magnetic resonance imaging studies. J Psychiatry Neurosci. 2009;34:418–432. [PMC free article] [PubMed] [Google Scholar]
  24. Gee DG, Humphreys KL, Flannery J, Goff B, Telzer EH, Shapiro M, Hare TA, Bookheimer SY, Tottenham N. A developmental shift from positive to negative connectivity in human amygdala-prefrontal circuitry. J Neurosci. 2013;33:4584–4593. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Geller B, Zimerman B, Williams M, Bolhofner K, Craney JL, DelBello MP, Soutullo C: Reliability of the Washington University in St. Louis Kiddie Schedule for Affective Disorders and Schizophrenia (WASH-U-KSADS) mania and rapid cycling sections. J Am Acad Child Adolesc Psychiatry 40:450–455, 2001. [DOI] [PubMed] [Google Scholar]
  26. Gerber AJ, Peterson BS, Giedd JN, Lalonde FM, Celano MJ, White SL, Wallace GL, Lee NR, Lenroot RK. Anatomical brain magnetic resonance imaging of typically developing children and adolescents. J Am Acad Child Adolesc Psychiatry. 2009;48:465–470. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Glantz LA, Gilmore JH, Hamer RM, Lieberman JA, Jarskog LF. Synaptophysin and postsynaptic density protein 95 in the human prefrontal cortex from mid-gestation into early adulthood. Neuroscience. 2007;149:582–591. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Grayson DS, Kroenke CD, Neuringer M, Fair DA. Dietary omega-3 fatty acids modulate large-scale systems organization in the rhesus macaque brain. J Neurosci. 2014;34:2065–2074. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Grosso G, Micek A, Marventano S, Castellano S, Mistretta A, Pajak A, Galvano F. Dietary n-3 PUFA, fish consumption and depression: A systematic review and meta-analysis of observational studies. J Affect Disord. 2016;205:269–281. [DOI] [PubMed] [Google Scholar]
  30. Grosso G, Pajak A, Marventano S, Castellano S, Galvano F, Bucolo C, Drago F, Caraci F. Role of omega-3 fatty acids in the treatment of depressive disorders: a comprehensive meta-analysis of randomized clinical trials. PLoS One. 2014;9:e96905. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Guy W: National Institute of Mental Health. In: ECDEU Assessment Manual for Psychopharmacology. Revised 12–1-12–6, 1970. [Google Scholar]
  32. Hare TA, Tottenham N, Galvan A, Voss HU, Glover GH, Casey BJ. Biological substrates of emotional reactivity and regulation in adolescence during an emotional go-nogo task. Biol Psychiatry. 2008;63:927–934. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. He C, Qu X, Cui L, Wang J, Kang JX. Improved spatial learning performance of fat-1 mice is associated with enhanced neurogenesis and neuritogenesis by docosahexaenoic acid. Proc Natl Acad Sci U S A. 2009;106:11370–11375. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Hibbeln JR. Fish consumption and major depression. Lancet. 1998;351(9110):1213. [DOI] [PubMed] [Google Scholar]
  35. Honey CJ, Sporns O, Cammoun L, Gigandet X, Thiran JP, Meuli R, Hagmann P. Predicting human resting-state functional connectivity from structural connectivity. Proc Natl Acad Sci U S A. 2009;106:2035–2040. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Kaiser RH, Andrews-Hanna JR, Wager TD, Pizzagalli DA. Large-scale network dysfunction in major depressive disorder: A meta-analysis of resting-state functional connectivity. JAMA Psychiatry. 2015;72:603–611. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, Walters EE. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry. 2005;62:593–602. [DOI] [PubMed] [Google Scholar]
  38. Latora V, Marchiori M. Efficient behavior of small-world networks. Phys Rev Lett. 2001;87:198701. [DOI] [PubMed] [Google Scholar]
  39. Lei D, Pinaya WHL, van Amelsvoort T, Marcelis M, Donohoe G, Mothersill DO, Corvin A, Gill M, Vieira S, Huang X, Lui S, Scarpazza C, Young J, Arango C, Bullmore E, Qiyong G, McGuire P, Mechelli A. Detecting schizophrenia at the level of the individual: relative diagnostic value of whole-brain images, connectome-wide functional connectivity and graph-based metrics. Psychol Med. 2020;50:1852–1861. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Lei D, Li W, Tallman MJ, Patino LR, McNamara RK, Strawn JR, Klein CC, Nery FG, Fleck DE, Qin K, Ai Y, Yang J, Zhang W, Lui S, Gong Q, Adler CM, Sweeney JA, DelBello MP. Changes in the brain structural connectome after a prospective randomized clinical trial of lithium and quetiapine treatment in youth with bipolar disorder. Neuropsychopharmacology. 2021. [Epub ahead of print]. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Li BJ, Friston K, Mody M, Wang HN, Lu HB, Hu DW. A brain network model for depression: From symptom understanding to disease intervention. CNS Neurosci Ther. 2018;24:1004–1019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Lin PY, Huang SY, Su KP. A meta-analytic review of polyunsaturated fatty acid compositions in patients with depression. Biol Psychiatry. 2010;68:140–147. [DOI] [PubMed] [Google Scholar]
  43. Liu C, Pu W, Wu G, Zhao J, Xue Z. Abnormal resting-state cerebral-limbic functional connectivity in bipolar depression and unipolar depression. BMC Neurosci. 2019;20:30. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. McNamara RK, Jandacek R, Tso P, Blom TJ, Welge JA, Strawn JR, Adler CM, Strakowski SM, DelBello MP. Adolescents with or at ultra-high risk for bipolar disorder exhibit erythrocyte docosahexaenoic acid and eicosapentaenoic acid deficits: a candidate prodromal risk biomarker. Early Interv Psychiatry. 2016;10:203–211. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. McNamara RK, Jandacek RJ, Tso P, Blom TJ, Welge JA, Strawn JR, Adler CM, DelBello MP, Strakowski SM. First-episode bipolar disorder is associated with erythrocyte membrane docosahexaenoic acid deficits: Dissociation from clinical response to lithium or quetiapine. Psychiatry Res. 2015;230:447–453. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. McNamara RK, Strawn JR, Tallman MJ, Welge JA, Patino LR, Blom T, DelBello MP. Effects of fish oil monotherapy on depression and prefrontal neurochemistry in adolescents at high risk for bipolar I disorder: A 12-week placebo-controlled 1H MRS trial. J Child Adolesc Psychopharmacol. 2020;30:293–305. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. McNamara RK, Strimpfel J, Jandacek RJ, Rider T, Tso P, Welge J, Strawn JR, DelBello MP. Detection and treatment of long-chain omega-3 fatty acid deficiency in adolescents with SSRI-resistant major depressive disorder. Pharma Nutrition. 2014;2:38–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. McNamara RK, Welge JA. Meta-analysis of erythrocyte polyunsaturated fatty acid biostatus in bipolar disorder. Bipolar Disord. 2016;18:300–306. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Mills KL, Goddings AL, Clasen LS, Giedd JN, Blakemore SJ. The developmental mismatch in structural brain maturation during adolescence. Dev Neurosci. 2014;36:147–160. [DOI] [PubMed] [Google Scholar]
  50. Mocking RJ, Harmsen I, Assies J, Koeter MW, Ruhé HG, Schene AH. Meta-analysis and meta-regression of omega-3 polyunsaturated fatty acid supplementation for major depressive disorder. Transl Psychiatry. 2016;6:e756. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Moreira JD, Knorr L, Ganzella M, Thomazi AP, de Souza CG, de Souza DG, Pitta CF, Mello e Souza T, Wofchuk S, Elisabetsky E, Vinadé L, Perry ML, Souza DO. Omega-3 fatty acids deprivation affects ontogeny of glutamatergic synapses in rats: relevance for behavior alterations. Neurochem Int. 2010;56:753–759. [DOI] [PubMed] [Google Scholar]
  52. Mortensen PB, Pedersen CB, Melbye M, Mors O, Ewald H: Individual and familial risk factors for bipolar affective disorders in Denmark. Arch Gen Psychiatry. 2003;60:1209–1215. [DOI] [PubMed] [Google Scholar]
  53. Muschelli J, Nebel MB, Caffo BS, Barber AD, Pekar JJ, Mostofsky SH. Reduction of motion-related artifacts in resting state fMRI using aCompCor. NeuroImage. 2014;96:22–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Noaghiul S, Hibbeln JR. Cross-national comparisons of seafood consumption and rates of bipolar disorders. Am J Psychiatry. 2003;160:2222–2227. [DOI] [PubMed] [Google Scholar]
  55. Olsavsky AK, Brotman MA, Rutenberg JG, Muhrer EJ, Deveney CM, Fromm SJ, Towbin K, Pine DS, Leibenluft E. Amygdala hyperactivation during face emotion processing in unaffected youth at risk for bipolar disorder. J Am Acad Child Adolesc Psychiatry. 2012;51:294–303. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Olson IR, McCoy D, Klobusicky E, Ross LA. Social cognition and the anterior temporal lobes: a review and theoretical framework. Soc Cogn Affect Neurosci. 2013;8:123–133. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Paus T, Keshavan M, Giedd JN. Why do many psychiatric disorders emerge during adolescence? Nat Rev Neurosci. 2008;9:947–957. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Pavuluri MN, O’Connor MM, Harral E, Sweeney JA. Affective neural circuitry during facial emotion processing in pediatric bipolar disorder. Biol Psychiatry. 2007;62:158–167. [DOI] [PubMed] [Google Scholar]
  59. Perlis RH, Dennehy EB, Miklowitz DJ, Delbello MP, Ostacher M, Calabrese JR, Ametrano RM, Wisniewski SR, Bowden CL, Thase ME, Nierenberg AA, Sachs G. Retrospective age at onset of bipolar disorder and outcome during two-year follow-up: results from the STEP-BD study. Bipolar Disord. 2009;11:391–400. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Pomponi M, Janiri L, La Torre G, Di Stasio E, Di Nicola M, Mazza M, Martinotti G, Bria P, Lippa S, Natili R, Pomponi MF. Plasma levels of n-3 fatty acids in bipolar patients: deficit restricted to DHA. J Psychiatr Res. 2013;47:337–342. [DOI] [PubMed] [Google Scholar]
  61. Poznanski EO, Cook SC, Carroll BJ, Corzo H: Use of the Children’s Depression Rating Scale in an inpatient psychiatric population. J Clin Psychiatry. 1983;44:200–203. [PubMed] [Google Scholar]
  62. Poznanski EO, Cook SC, Carroll BJ: A depression rating scale for children. Pediatrics. 1979;64:442–450. [PubMed] [Google Scholar]
  63. Ratheesh A, Davey C, Hetrick S, Alvarez-Jimenez M, Voutier C, Bechdolf A, McGorry PD, Scott J, Berk M, Cotton SM: A systematic review and meta-analysis of prospective transition from major depression to bipolar disorder. Acta Psychiatr Scand. 2017;135:273–284. [DOI] [PubMed] [Google Scholar]
  64. Rich BA, Vinton DT, Roberson-Nay R, Hommer RE, Berghorst LH, McClure EB, Fromm SJ, Pine DS, Leibenluft E. Limbic hyperactivation during processing of neutral facial expressions in children with bipolar disorder. Proc Natl Acad Sci U S A. 2006;103:8900–8905. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Rubinov M, Sporns O. Complex network measures of brain connectivity: uses and interpretations. NeuroImage. 2010;52:1059–1069. [DOI] [PubMed] [Google Scholar]
  66. Sarris J, Mischoulon D, Schweitzer I: Omega-3 for bipolar disorder: meta-analyses of use in mania and bipolar depression. J Clin Psychiatry 73:81–86, 2012. [DOI] [PubMed] [Google Scholar]
  67. Serafini G, Canepa G, Adavastro G, Nebbia J, Belvederi Murri M, Erbuto D, Pocai B, Fiorillo A, Pompili M, Flouri E, Amore M. The relationship between childhood maltreatment and non-suicidal self-injury: A systematic review. Front Psychiatry. 2017;8:149. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Singh MK, Nimarko AF, Garrett AS, Gorelik AJ, Roybal DJ, Walshaw PD, Chang KD, Miklowitz DJ. Changes in intrinsic brain connectivity in family-focused therapy versus standard psychoeducation among youth at high risk for bipolar disorder. J Am Acad Child Adolesc Psychiatry. 2020. [Epub ahead of print]. [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Sowell ER, Thompson PM, Holmes CJ, Batth R, Jernigan TL, Toga AW. Localizing age-related changes in brain structure between childhood and adolescence using statistical parametric mapping. Neuroimage. 1999;9:587–597. [DOI] [PubMed] [Google Scholar]
  70. Strakowski SM, Eliassen JC, Lamy M, Cerullo MA, Allendorfer JB, Madore M, Lee JH, Welge JA, DelBello MP, Fleck DE, Adler CM. Functional magnetic resonance imaging brain activation in bipolar mania: evidence for disruption of the ventrolateral prefrontal-amygdala emotional pathway. Biol Psychiatry. 2011;69:381–388. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Sublette ME, Ellis SP, Geant AL, Mann JJ. Meta-analysis of the effects of eicosapentaenoic acid (EPA) in clinical trials in depression. J Clin Psychiatry. 2011;72:1577–1584. [DOI] [PMC free article] [PubMed] [Google Scholar]
  72. Swartz JR, Carrasco M, Wiggins JL, Thomason ME, Monk CS. Age-related changes in the structure and function of prefrontal cortex-amygdala circuitry in children and adolescents: a multi-modal imaging approach. Neuroimage. 2014;86:212–220. [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Talukdar T, Zamroziewicz MK, Zwilling CE, Barbey AK. Nutrient biomarkers shape individual differences in functional brain connectivity: Evidence from omega-3 PUFAs. Hum Brain Mapp. 2019;40:1887–1897. [DOI] [PMC free article] [PubMed] [Google Scholar]
  74. Tang S, Lu L, Zhang L, Hu X, Bu X, Li H, Hu X, Gao Y, Zeng Z, Gong Q, Huang X. Abnormal amygdala resting-state functional connectivity in adults and adolescents with major depressive disorder: A comparative meta-analysis. EBioMedicine. 2018;36:436–445. [DOI] [PMC free article] [PubMed] [Google Scholar]
  75. Wang F, Kalmar JH, He Y, Jackowski M, Chepenik LG, Edmiston EE, Tie K, Gong G, Shah MP, Jones M, Uderman J, Constable RT, Blumberg HP. Functional and structural connectivity between the perigenual anterior cingulate and amygdala in bipolar disorder. Biol Psychiatry. 2009;66:516–521. [DOI] [PMC free article] [PubMed] [Google Scholar]
  76. Wang J, Wang X, Xia M, Liao X, Evans A, He Y. GRETNA: a graph theoretical network analysis toolbox for imaging connectomics. Front Human Neurosci. 2015;9:386. [DOI] [PMC free article] [PubMed] [Google Scholar]
  77. Wang Y, Wang J, Jia Y, Zhong S, Zhong M, Sun Y, Niu M, Zhao L, Zhao L, Pan J, Huang L, Huang R. Topologically convergent and divergent functional connectivity patterns in unmedicated unipolar depression and bipolar disorder. Transl Psychiatry. 2017;7:e1165–e1165. [DOI] [PMC free article] [PubMed] [Google Scholar]
  78. Wechsler D: Wechsler Adult Intelligence Scale-third edition: administration and scoring manual. San Antonio, TX: Psychological Corportation. 1997. [Google Scholar]
  79. Whitfield-Gabrieli S, Nieto-Castanon A. Conn: a functional connectivity toolbox for correlated and anticorrelated brain networks. Brain connectivity. 2012;2:125–141. [DOI] [PubMed] [Google Scholar]
  80. Wu M, Kujawa A, Lu LH, Fitzgerald DA, Klumpp H, Fitzgerald KD, Monk CS, Phan KL. Age-related changes in amygdala-frontal connectivity during emotional face processing from childhood into young adulthood. Hum Brain Mapp. 2016;37:1684–1695. [DOI] [PMC free article] [PubMed] [Google Scholar]
  81. Yamasaki H, LaBar KS, McCarthy G. Dissociable prefrontal brain systems for attention and emotion. Proc Natl Acad Sci U S A. 2002;99:11447–11451. [DOI] [PMC free article] [PubMed] [Google Scholar]
  82. Young RC, Biggs JT, Ziegler VE, Meyer DA: A rating scale for mania: reliability, validity and sensitivity. Br J Psychiatry. 1978;133:429–435. [DOI] [PubMed] [Google Scholar]
  83. Yu Z, Qin J, Xiong X, Xu F, Wang J, Hou F, Yang A. Abnormal topology of brain functional networks in unipolar depression and bipolar disorder using optimal graph thresholding. Prog Neuropsychopharmacol Biol Psychiatry. 2020;96:109758–109758. [DOI] [PubMed] [Google Scholar]
  84. Zalesky A, Fornito A, Bullmore ET. Network-based statistic: identifying differences in brain networks. NeuroImage. 2010;53:1197–1207. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1711779-supplement-1.docx (749.7KB, docx)

RESOURCES