Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jul 23.
Published in final edited form as: J Am Acad Child Adolesc Psychiatry. 2025 Jul 23;65(7):917–928. doi: 10.1016/j.jaac.2025.07.421

Different Changes in Brain Functional Networks Following 12-Week Psychostimulant Treatment in Attention-Deficit/Hyperactivity Disorder Youth With and Without Familial Risk for Bipolar I Disorder

Kun Qin 1,2,#, Nanfang Pan 2,3,4,#, Du Lei 2,5, Ziyu Zhu 2,3, Maxwell J Tallman 2, L Rodrigo Patino 2, Qiyong Gong 3,6,7, John A Sweeney 2,3, Melissa P DelBello 2, Robert K McNamara 2
PMCID: PMC12645708  NIHMSID: NIHMS2110111  PMID: 40712681

Abstract

Objective:

Although psychostimulants are the first-line treatment for ADHD and can normalize associated brain network abnormalities, their neural effects in ADHD youth with a family history of bipolar I disorder (BD) have not yet been systematically investigated.

Method:

This study investigated the effects of 12-week mixed amphetamine salts-extended release (MAS-XR) treatment on brain functional networks and symptom changes in ADHD youth with (‘high-risk’, HR) and without (‘low-risk’, LR) a first-degree relative with BD. LR received 12-week open-label MAS-XR and HR were randomized to MAS-XR or placebo (PLA). Healthy controls (HC) were included for comparative purposes. Resting-state fMRI scans were acquired at baseline and week 12. Global and nodal network topological metrics were evaluated based on functional connectivity matrices using graph theoretical analysis.

Results:

A total of N=135 youth were included (HC, n=45; LR-MAS, n=46; HR-MAS, n=28; HR-PLA, n=16). For comparisons between HR-MAS and LR-MAS groups, significant group-by-time interactions were identified in global efficiency, characteristic path length, left superior parietal gyrus degree, and right amygdala efficiency. Relative to HC, all baseline abnormalities were normalized at week-12 in the LR-MAS group, whereas a decrease in right amygdala efficiency uniquely emerged in the HR-MAS group. Significant correlations were identified between changes in network topology and ADHD symptoms. Connectome-based predictive modeling analysis revealed that baseline network dysconnectivity predicted ADHD treatment outcomes in HR-MAS only.

Conclusion:

These findings provide novel evidence that ADHD youth with and without BD familial risk exhibit different changes in brain network topology and associated symptoms following 12 weeks of psychostimulant treatment.

Clinical trial registration information:

Neuroimaging Study of Risk Factors for Adolescent Bipolar Disorder (NERF); https://clinicaltrials.gov/study/NCT02478788

Keywords: attention-deficit/hyperactivity disorder, connectome, magnetic resonance imaging, graph theory, stimulants

INTRODUCTION

Attention-deficit/hyperactivity disorder (ADHD) commonly precedes the initial onset of bipolar I disorder (BD) and is a significant risk factor for developing BD.1,2 Youth with ADHD have a 9-fold greater risk of developing BD compared to their peers without ADHD.3 BD is highly heritable and having a first-degree relative with BD also robustly increases the risk of BD.4 Moreover, youth with familial risk for BD are more likely to develop ADHD prior to the onset of BD,5,6 and prevalence rates of ADHD in youth with a BD family history are significantly higher than those without a BD family history.7,8 Importantly, previous studies have shown that ADHD youth with a first-degree relative with BD exhibit more severe symptoms and an elevated prevalence for multiple psychopathologies.9,10 While these findings suggest that ADHD in conjunction with BD family history is associated with greater risk and may represent a distinctive disorder than ADHD alone, the underlying neural mechanisms remain poorly understood.

Resting-state functional magnetic resonance imaging has been widely used to characterize brain functional abnormalities in ADHD.11 More recently, the application of graph theoretical analysis, which models networks as mathematical graphs comprising nodes and edges, has enabled the quantitative characterization of topological properties at the whole-brain network level.12 Over the past decade, an increasing number of graph theory studies have investigated brain network disruptions in ADHD youth, revealing topological abnormalities including lower global integration, weaker small-worldness, and disrupted regional centralities of prefrontal, temporal and parietal regions.13–15 Most relevant to the present study, our group recently reported that medications-free ADHD youth with and without BD familial risk exhibit overlapping as well as unique patterns of functional network topological disruption.16 While our preliminary finding suggest that ADHD in youth with a familial risk for BD are characterized by different network-level abnormalities compared with ADHD alone, brain network changes in response to ADHD medications are not yet investigated.

Psychostimulants are currently the first-line treatment for ADHD regardless of BD familial risk.17 Although controversial, some evidence suggests that psychostimulant exposure can precipitate mood symptoms and accelerate the onset of mania in a subset of vulnerable ADHD youth.18,19 Emerging evidence further suggests that improvements in ADHD symptoms following psychostimulant treatment are associated with neuroplastic changes in brain networks. Preclinical studies have shown that repeated exposure to psychostimulants produce enduring synaptic remodeling in prefrontal and limbic regions,20,21 and clinical evidence also suggests that psychostimulants can normalize aberrant functional network connectivity in ADHD youth.22,23 Regarding network-level topological organization, Kaiser et al. reported acute effects of a single-dose methylphenidate challenge on network regional centrality in the striatum and thalamus in ADHD children.24 A prospective resting-state fMRI study found that 12-weeks of methylphenidate treatment significantly reduced functional network segregation in medication-naïve youth with ADHD.25 However, the latter studies did not control for BD family history, and the effects of psychostimulant treatment on brain network organization and connectivity in ADHD youth with a BD family history have never been investigated. However, previous studies did not account for genetic risk of BD which are known to influence brain structure and function.26,27 The effects of psychostimulant treatment on brain network organization and connectivity in ADHD youth with a BD family history remain unclear and require further investigation.

In the context of a larger 12-week psychostimulant treatment trial, this exploratory analysis aimed to evaluate changes in functional network organization in ADHD youth with and without a first-degree relative with BD. A healthy control (HC) group was included for comparative purposes. Functional brain network and derived network topological metrics were estimated using graph theoretical analysis. Based on the evidence reviewed above, we hypothesized that changes in global and nodal network topology following psychostimulant treatment would differ between ADHD youth with and without a BD family history, and that these alterations would correlate with changes in ADHD symptom ratings. In addition, an exploratory connectome-based predictive modeling (CPM) analysis was conducted to elucidate relationships between baseline connectome profiles and response to psychostimulant treatment in ADHD youth with and without a BD family history.28 It was hypothesized that baseline functional network profiles would demonstrate differential predictive capacity for ADHD symptom improvement between ADHD youth with and without a BD family history.

METHOD

Participants

Three groups of psychostimulant-free youth (10–18 years) were enrolled: 1) youth with ADHD and at least one biological parent or sibling with BD (‘high-risk’, HR), 2) youth with ADHD and no first- or second-degree relative with a mood or psychotic disorder (‘low-risk’, LR), and 3) typically developing HC with no personal or family history of a DSM-5 Axis I psychiatric disorder. The Structured Clinical Interview for DSM-5 (SCID-5-CV) was administered to confirm a parental or sibling (≥18 years) diagnosis of BD, and the Kiddie Schedule for Affective Disorders and Schizophrenia - Present and Lifetime Version (KSADS-PL) was administered to confirm a sibling (< 18 years) diagnosis of BD. The Family Interview for Genetics Studies (FIGS) was used to confirm DSM-5 BD diagnoses in other first- or second-degree relatives. Healthy controls had no lifetime DSM-5 Axis I disorder as determined by the KSADS-PL, and had no first- or second-degree relatives with a mood or psychotic disorder as determined by the FIGS.

All participants had no contraindication to an MRI scan, had an IQ ≥ 80 as determined by the Wechsler Abbreviated Scale of Intelligence (WASI), had no major medical or neurological illness that could influence MRI results or any significant episode (> 10 minutes) of loss of consciousness, and had no lifetime DSM-5 substance use disorder. All ADHD youth had no current DSM-5 mood, conduct, eating or psychotic disorders, Tourette’s disorder, chronic tic disorder, or autism spectrum disorder, had no exposure to psychostimulants (prescription or recreational) or other medications used for the treatment of ADHD (e.g. atomoxetine, alpha-2 agonists) for at least 3 months prior to screening, had no lifetime exposure to mood-stabilizer or antipsychotic medications, and had no clinically significant ECG or blood pressure abnormalities. Written informed assent and consent were obtained from youth and their legal guardians, respectively. This study was approved by the University of Cincinnati Institutional Review Board and was registered at clinicaltrials.gov with the identifier NCT02478788.

Treatments

ADHD youth who completed initial screening and the baseline fMRI scan began their 12-week treatment with mixed amphetamine salts-extended release (Adderall XR, MAS-XR). LR ADHD (LR-MAS) received open-label MAS-XR, and high-risk youth were randomized to double-blind treatment with MAS-XR (HR-MAS) or placebo (HR-PLA) (Figure S1, available online). Additionally, a subgroup of HR youth (n=10) who received open-label MAS-XR treatment were included for the current analysis. Their symptom responses did not differ significantly from HR youth who received double-blind MAS-XR treatment. Furthermore, after excluding these n=10 subjects, our primary and secondary imaging findings remained statistically significant. In accordance with current MAS-XR dosage and administration guidelines, the starting dose was 10 mg/day and was increased based on effectiveness and tolerability to within a range of 10–40 mg/day. The MAS dose was increased within a range of 10–30 mg/day in subjects that were 10–12 years old. Medication adherence was monitored by subject interview, pill counts, legal guardian interviews, and a medication adherence diary. LR-MAS and HR-MAS groups did not differ in MAS-XR endpoint dose (LR-MAS: 19.7±6.7 vs. HR-MAS: 19.5±6.5, p = .93), median dose (LR-MAS: 17.7±5.2 vs. HR-MAS: 15.8±5.7, p = .12) or mean maximum dose (LR-MAS: 20.3±6.6 vs. HR-MAS: 20.9±5.7, p = .65), and the group-by-time interaction effect was not significant (p = .74). All ADHD youth were instructed to not take their daily MAS dose on the day of the endpoint MRI scan (i.e., 24h washout period) to obviate acute effects.

Symptom ratings

ADHD symptom ratings were obtained using the clinician-administered Attention-Deficit Hyperactivity Disorder Rating Scale (ADHD-RS), and inattention and hyperactivity/impulsivity subscale scores were analyzed separately. Remission was defined as an endpoint ADHD-RS total score ≤ 18. Clinician ratings were administered by a child and adolescent psychiatrist with established inter-rater reliabilities (kappa > 0.9).

MRI data acquisition and preprocessing

All participants received MRI scan at baseline and week 12. Both resting-state fMRI and high-resolution T1-weighted MRI data were collected using a Philips Ingenia 3T MRI scanner with a 32-channel head coil. The entire resting-state fMRI session lasted for 10 minutes, and detailed scanning parameters are available in Supplement 1, available online. During the resting-state fMRI session, subjects were presented with a white fixation cross on a black background and asked to stay still and awake and to watch the crosshair. Raw images underwent a series of preprocessing procedures that included slice timing correction, head motion correction, space normalization, smooth, removal of nuisance variables and scrubbing. Subjects showing excessive head motion (movement > 3mm or rotation > 3 degree or mean framewise displacement (FD) > 0.5) were excluded from our analysis. In addition, we conducted a head motion analysis to assess potential differences in mean FD between groups and sessions, as well as its influence on our findings. No significant group, time or group-by-time interaction effects were observed for individual mean FD (Table S1, available online). Furthermore, including mean FD as a covariate in our statistical models did not change our results (Table S2, available online). Details regarding image preprocessing and head motion analysis are presented in Supplement 1, available online.

Functional network construction and topological metrics estimation

Individual functional networks were constructed based on preprocessed resting-state fMRI scans. We used the automatic anatomical labeling (AAL) atlas to parcellate the whole brain into 90 regions (Table S3, available online). The functional connectivity between each pair of regions was calculated as Pearson correlation coefficients of regional blood oxygen level dependent (BOLD) signals. Fisher’s z-transformation was applied to each functional connectivity value, generating a 90 × 90 connectivity matrix that represents the individual functional network. To obtain functional network topological metrics, graph theoretical analysis was performed using the GRETNA toolbox.29 Network topological metrics were assessed at global and nodal levels. Global topological metrics included global efficiency (Eglob), local efficiency (Eloc), clustering coefficient (Cp) and characteristic path length (Lp). Nodal topological metrics included nodal degree, efficiency and betweenness of 90 regions. More details regarding the definition and calculation of above topological metrics are shown in Supplement 1, available online.

Statistical analyses

Our primary analysis investigated changes in topological metrics between HR-MAS and LR-MAS groups by fitting a linear mixed effects model. We used the “lmer” function from the “lme4” package in R (v4.4.1) to fit our linear mixed effects model. The fixed effects included the intercept, the main effects of time (i.e., baseline and week-12), group (i.e., LR-MAS and HR-MAS) and their interaction, and the covariates of age and sex. The random effect was specified as a random intercept for each subject, and an unstructured covariance structure was used. The false discovery rate (FDR) correction was applied to control for type I error of multiple comparisons (FDR-adjusted P < 0.05), and P values (unadjusted) that survived FDR correction are reported. The effect size for each significant finding was estimated using Cohen’s d. For regions showing significant interactions, post-hoc analyses were performed to examine within-group changes over time and between-group differences at baseline and week 12. In addition, we conducted exploratory partial correlation analyses between changes in ADHD symptom ratings and significant topological metrics, using age and sex as covariates. Correlations were investigated separately within the LR-MAS and HR-MAS groups, and the group-by-score interaction effects were examined to evaluate whether the correlations significantly differed between the two groups. Furthermore, correlation analyses were performed on the combined dataset from both groups to examine potential associations across the entire ADHD youth.

Our secondary analyses investigated the following group-by-time interactions: (1) (HR-MAS vs. HC) × (baseline vs. week 12); (2) (LR-MAS vs. HC) × (baseline vs. week 12); (3) (HR-MAS vs. HR-PLA) × (baseline vs. week 12). The first two analyses compared ADHD groups with HC to determine whether observed treatment-induced changes in topological metrics were in a normalizing direction. The third analysis aimed to exclude any treatment-related changes that might be attributed to placebo effects in HR ADHD youth. Linear mixed effects models controlling for age and sex were used, and the FDR correction was applied to control for multiple comparisons (FDR-corrected P < 0.05).

To evaluate the test-retest reliability of network topological metrics, we conducted intra-class correlation coefficient (ICC) analyses within the HC group (using two-way mixed single measures testing for consistency) for all the metrics showing significant interaction effects. The negative ICC values, which indicate negative reliability and are theoretically difficult to interpret, were set to zero to represent complete unreliability.

Connectome-based prediction of treatment response

We used connectome-based predictive modeling (CPM) to predict 12-week treatment response. Briefly, CPM is a framework that takes network connectivity matrices and behavioral measures as predictor and response variables, respectively, to generate a linear predictive model (Figure S2, available online).28 Edges are first selected from the entire connectivity matrix and then categorized into positive and negative prediction networks, making predictions independently. Positive prediction networks contain a set of significant edges showing positive correlations which indicate higher baseline connectivity predicts greater ADHD symptom improvement. Negative prediction networks contain edges with significant negative correlations which indicate lower baseline connectivity predicts greater ADHD symptom improvement. We built separate CPM models for LR-MAS, HR-MAS, and both groups combined (i.e., LR-MAS and HR-MAS). The baseline network connectivity matrix was the predictor variable, and baseline-endpoint percentage change in ADHD-RS total score was the response variable. Leave-one-out cross-validation was performed to split data into training and test sets. Model performance was evaluated in terms of coefficient of determination (R2 = 1 - (sum of squared residuals) / (total sum of squares)) and Pearson correlation between predicted and actual ADHD-RS percentage changes (r[actual, predicted]). Statistical significance of the predictive model was examined using a nonparametric permutation test. For significant CPM models, the mean squared error (MSE) was calculated, and the most contributive edges and nodes were further investigated. In addition, the prediction framework was repeated by directly setting endpoint ADHD-RS scores as response variables to examine the stability of our prediction model. The CPM was implemented using previously validated custom Python scripts (Python v3.8.0: https://github.com/esfinn/cpm_tutorial/blob/master/cpm_tutorial.ipynb). Additional details about CPM are provided in Supplement 1, available online.

RESULTS

Demographic and clinical characteristics

A total of 135 youth (HC: n = 45; LR-MAS: n = 46; HR-MAS: n = 28; HR-PLA: n = 16) with usable baseline MRI data were included in this study. After week 12, 118 MRI scans (HC: n = 38; LR-MAS: n = 44; HR-MAS: n = 23; HR-PLA: n = 13) were available due to subject attrition and MRI data quality control (Supplement 1, available online). No significant baseline differences were observed for age, sex, hand dominance, race, BMI, or prior stimulant exposure among the 4 groups. The LR-MAS group had higher rates of ADHD-inattention subtype, whereas HR groups had higher rates of ADHD-combined subtype (p = .011). There were no significant ADHD group differences in baseline ADHD-RS total (p = .495), inattention subscale (p = .572), or hyperactivity/impulsivity subscale (P =0.084) scores (p = .226) (Table 1).

Table 1.

Baseline demographic and clinical characteristics

Variables HR-MAS (n=28) HR-PLA (n=16) LR-MAS (n=46) HC (n=45) p value
Age, years 14.12±2.31 13.14±2.75 14.26±2.49 14.76±2.41 .156
Sex, female n (%) 12 (42.9%) 4 (25.0%) 16 (34.8%) 17 (37.8%) .684
Handedness, right-handed, n (%) 26 (92.9%) 12 (75.0%) 39 (84.8%) 43 (95.6%) .146
Race, White n (%) 18 (64.3%) 8 (50.0%) 30 (65.2%) 29 (64.4%) .407
BMI 24.28±7.63 23.17±4.45 24.17±6.91 22.33±4.37 .441
ADHD subtype
 ADHD-I, n (%) 9 (32.1%) 4 (25.0%) 28 (60.9%) - .011
 ADHD-C, n (%) 19 (67.9%) 12 (75.0%) 18 (39.1%) -
Prior psychostimulant use, n (%) 12 (42.9%) 7 (43.8%) 16 (34.8%) - .715
ADHD-RS total score 35.18±9.57 35.00±12.4 32.52±10.0 - .495
 Hyperactivity/Impulsivity 15.43±7.29 15.94±7.36 11.93±8.15 - .084
 Inattention 19.75±5.07 19.06±6.61 20.59±4.89 - .572
Open in a new tab

Note: Values are mean ± standard deviation or number of subjects (n) and percentage (%). ADHD = attention-deficit/hyperactivity disorder; ADHD-C = attention-deficit/hyperactivity disorder combined type; ADHD-I = attention-deficit/hyperactivity disorder inattentive type; ADHD-RS = attention-deficit/hyperactivity disorder rating scale; BMI = body mass index; HC = healthy control; HR-MAS = high-risk youth treated with mixed amphetamine salts; HR-PLA = high-risk youth treated with placebo; LR-MAS = low-risk youth treated with mixed amphetamine salts.

ADHD Symptom response

Compared with HR-MAS, LR-MAS exhibited significantly greater baseline-endpoint reductions in ADHD-RS inattention subscale scores (group-by-time interaction: p = .001) but not hyperactivity/impulsivity subscale scores (p = .639), and there was a trend for ADHD-RS total score (p = .087) (Figure 1; Table S4, available online). Remission rates (endpoint ADHD-RS total score ≤18) for LR-MAS and HR-MAS groups were 93.2% and 73.9%, respectively (p = .028). Compared with HR-PLA, HR-MAS exhibited significantly greater baseline-endpoint reductions in ADHD-RS inattention subscale scores (p = .030), but not ADHD-RS total scores (p = .106) or ADHD-RS hyperactivity/impulsivity subscale scores (p = .513) (Table S5, available online).

Figure 1. Changes in attention-deficit/hyperactivity disorder symptom ratings following 12-week psychostimulant treatment.

Figure 1.

Open in a new tab

Note: Symptom changes are separately shown for (A) ADHD-RS total score, (B) ADHD-RS hyperactivity/impulsivity subscale score and (C) ADHD-RS inattention subscale score. Values are group mean ± SEM. P-values for the group × time interaction are presented, and asterisks indicate significant within-group baseline-endpoint changes. ADHD-RS = attention-deficit/hyperactivity disorder rating scale; HR-MAS = high-risk youth treated with mixed amphetamine salts; LR-MAS = low-risk youth treated with mixed amphetamine salts. SEM = standard error of the mean.

*p < .05, **p < 0.01, ***p < .001.

Differential network topological changes between LR-MAS and HR-MAS groups

Graphical comparison of changes in functional global and nodal network metrics in all groups is presented in Figure S3, available online. For the primary analysis (i.e., LR-MAS vs. HR-MAS), significant group-by-time interactions were observed for Eglob (p = .013), Lp (p = .005), degree of the left superior parietal gyrus (SPG; p < .001), and efficiency of the right amygdala (p < .001). After controlling for ADHD subtype, primary findings, including Eglob (p = .014), Lp (p = .006), degree of the left SPG (p < .001), efficiency of the right amygdala (p < .001), remained statistically significant. Post-hoc analyses of within-group changes revealed that LR-MAS group exhibited significant baseline-endpoint increases in Eglob (p = .012), decreases in Lp (p = .002), and increases in degree of the left SPG (p < .001), whereas HR-MAS group exhibited significantly decreased efficiency of the right amygdala (p < .001) (Figure 2; Table S6, available online). For between-group differences, at baseline LR-MAS group exhibited significantly lower degree of the left SPG compared with HR-MAS (p < .001), and at week-12 HR-MAS group showed significantly lower efficiency of the right amygdala compared with LR-MAS (p = .038). All above significant topological metrics exhibited relatively fair to moderate reliability with test-retest ICCs ranging from 0.302 to 0.544.

Figure 2. Changes in functional network topological metrics following 12-week psychostimulant treatment.

Figure 2.

Open in a new tab

Note: Network topological changes are separately shown for (A) global efficiency, (B) characteristic path length, (C) left SPG nodal degree, and (D) right amygdala nodal efficiency. Left SPG and right amygdala localization in axial, sagittal and coronal views are illustrated. Values are group mean ± SEM. P-values for the group × time interaction are presented, and asterisks indicate significant within-group baseline-endpoint changes. HR-MAS = high-risk youth treated with mixed amphetamine salts; LR-MAS = low-risk youth treated with mixed amphetamine salts. SEM = standard error of the mean. SPG = superior parietal gyrus.

*p < .05, **p < 0.01, ***p < .001.

Differential network topological changes between HR-MAS/LR-MAS and HC groups

Within the HC group, there were no significant changes in any global or nodal topological metrics over 12 weeks (all p > .05), supporting the temporal stability of graph theoretical metrics. Significant group-by-time interactions between LR-MAS and HC groups were found for Eglob (p = .005), Lp (p = .001), and degree of the left SPG (p < .001) (Figure S4 and Table S7, available online). Compared with HC, LR-MAS exhibited significantly lower Eglob (p = .004), higher Lp (p = .002) and lower degree of the left SPG (p < .001) at baseline, but no significant group differences were observed at week 12.

For comparisons between HR-MAS and HC groups, significant group-by-time interactions were found for efficiency of the right amygdala (p < .001) and betweenness of the right inferior frontal gyrus (IFG) (p < .001) (Figure S5 and Table S8, available online). There were no significant between-group differences at baseline, but the HR-MAS group exhibited significantly lower efficiency of the right amygdala (p < .001), and greater betweenness of the right IFG (p < .001), at week 12 compared with HC. However, the ICC of right IFG nodal betweenness was 0 indicating no test-retest reliability.

Differential network topological changes between HR-MAS and HR-PLA

Within the HR-PLA group, no significant changes were identified for global topological metrics, and we observed a significant increase in betweenness of the right IFG over 12 weeks (p < .001). No significant group-by-time interactions were observed between HR-MAS and HR-PLA groups for either global or nodal metrics.

Associations between changes in network topological metrics and ADHD-RS ratings

Within the LR-MAS group, baseline-week 12 changes in ADHD-RS hyperactivity/impulsivity subscale scores were inversely correlated with changes in Eglob (r = −0.42, p = .008) and positively correlated with changes in Lp (r = 0.43, p = .007). No significant correlations were found within the HR-MAS group (Figure 3). Correlations between changes in nodal degree of the left SPG and ADHD-RS inattention subscale scores differed significantly between LR-MAS and HR-MAS groups (interaction effects: p = .0006). Specifically, the HR-MAS group exhibited a positive correlation (r = 0.57, p = .017) whereas the LR-MAS group exhibited a negative correlation (r = −0.33, p = .024). No significant correlations were observed when LR-MAS and HR-MAS groups were combined.

Figure 3. Correlations between changes in functional network topological metrics and attention-deficit/hyperactivity disorder symptom ratings.

Figure 3.

Open in a new tab

Note: Associations between baseline-endpoint changes in (A) global efficiency and (B) characteristic path length and ADHD-RS hyperactivity/impulsivity subscale scores, and between (C) left SPG nodal degree and ADHD-RS inattention subscale scores. Within group correlation coefficients and group interaction terms are presented. ADHD-RS = attention-deficit/hyperactivity disorder rating scale; HR-MAS = high-risk youth treated with mixed amphetamine salts; LR-MAS = low-risk youth treated with mixed amphetamine salts. SPG = superior parietal gyrus.

Baseline functional connectome predicted psychostimulant treatment response

Since no significant positive correlations were identified between baseline connectivity and changes in ADHD-RS scores, CPM models were trained based only on negative prediction network (i.e., lower connectivity predicts greater reduction in the ADHD-RS total score). When combining LR-MAS and HR-MAS groups, baseline negative prediction network did not significantly predict 12-week changes in ADHD-RS total scores (r[actual, predicted] = −0.126, R2 = −0.271, p = .986). When examining LR-MAS and HR-MAS groups separately, baseline negative prediction network significantly predicted treatment response in the HR-MAS group (r[actual, predicted] = 0.421, R2 = 0.232, p = .018, MSE = 0.053) but not in the LR-MAS group (r[actual, predicted] = 0.067, R2 = −0.093, p = .192) (Figure 4A). For the significant prediction model for HR-MAS group, we found that connections among the prefrontal, occipital and anterior cingulate cortices (ACC) contributed most to the prediction (Figure 4B and C). The top 10 regions with most connections within the negative prediction network were mainly distributed in the medial orbitofrontal cortex (mOFC), medial prefrontal cortex (mPFC), ACC, and regions related to visual, auditory and sensory processing (Figure 4D; Table S9, available online).

Figure 4. Prediction of treatment responses based on baseline functional connectome.

Figure 4.

Open in a new tab

Note: (A) Prediction performance in LR-MAS and HR-MAS groups and both groups combined presented as correlations between actual response and predicted response. (B) The most contributive edges consistently selected by the prediction model over all cross-validation folds. The size of the nodes indicates the number of connections with the region. (C) Top 10 regions contributing most to the prediction. ACC = anterior cingulate cortex; CUN = cuneus; HES = Heschl gyrus; HR-MAS = high-risk youth treated with mixed amphetamine salts; LR-MAS = low-risk youth treated with mixed amphetamine salts; mSFG = medial superior frontal gyrus; PoCG = postcentral gyrus; smOFC = superior medial orbitofrontal cortex; SOC = superior occipital cortex.

When using endpoint ADHD-RS score as a response variable, similar prediction patterns were observed. Briefly, decreased baseline connectivity could significantly predict lower endpoint ADHD-RS score in HR-MAS (r[actual, predicted] = 0.416, R2 = 0.226, p = .024) but not in LR-MAS (r[actual, predicted] = −0.203, R2 = −0.193, p = .846) or combined ADHD groups (r[actual, predicted] = 0.062, R2 = −0.077, p = .277) (Figure S6, available online).

DISCUSSION

To our knowledge, this is the first study to directly compare the effects of psychostimulant monotherapy on resting-state functional network global and nodal topological organization in ADHD youth with and without a BD family history. Consistent with our primary hypothesis, significant group-by-time interactions were found for changes in global and nodal metrics between HR-MAS and LR-MAS groups. Relative to HC, these changes were normalizing in LR-MAS group, but not in HR-MAS group, and a decrease in right amygdala nodal efficiency uniquely emerged in the HR-MAS group. Significant correlations were identified between changes in network metrics and ADHD symptoms ratings. Moreover, we found that baseline network dysconnectivity patterns predicted ADHD treatment outcomes in HR-MAS but not LR-MAS group. Together, these findings demonstrate that ADHD youth with and without BD familial risk exhibit differential patterns of network changes in response to 12-week psychostimulant monotherapy.

The human brain connectome conforms to the small-world architecture which enables both global integration measured by Eglob and Lp, and localized segregation measured by Eloc and Cp, to provide efficient information processing.30 At baseline, LR ADHD youth exhibited lower Eglob and higher Lp compared with HC, indicating weaker global integration and a shift from small-world network towards regularized networks.31 This finding is consistent with previous studies which also identified impaired global integrity and abnormal small-world architecture in youth with ADHD.32,33 Additionally, 12-week MAS treatment normalized Eglob and Lp in LR ADHD youth relative to healthy youth, and increases in the global efficiency correlated with reductions in ADHD-RS hyperactivity/impulsivity scores within the LR ADHD group. These findings are generally consistent with prior evidence that stimulant treatment has normalizing effects on aberrant network functional connectivity in ADHD youth which are associated with improvements in ADHD symptoms.22,23,34 In contrast, we found that HR-MAS group did not exhibit significant global topological alterations compared with HC at either baseline or week 12, and global metrics did not correlate with changes in ADHD symptoms. This finding suggests that decreased global network integration is relevant to the pathophysiology and treatment of ADHD in LR youth but not in ADHD youth with a BD family history.

We also observed abnormal regional network topology including nodal degree, efficiency and betweenness. At baseline, LR ADHD youth exhibited lower left SPG nodal degree compared with both HC and HR-MAS, and this was normalized following 12-week MAS treatment. Moreover, within the LR-MAS group greater increases in SPG nodal degree were associated with greater reductions in ADHD inattention subscale scores. These findings are congruent with prior evidence that improvements in ADHD symptoms following repeated psychostimulant treatment are associated with changes in SPG connectivity in ADHD youth.22,25 Specifically, 8-week methylphenidate treatment elicited local connectivity alterations in the SPG which correlated with improvements in ADHD symptoms22, and a resting-state study found that improvements in ADHD symptoms following 12-week methylphenidate treatment were associated with increased amplitude of low frequency fluctuation in the bilateral SPG.25 In contrast, the HR-MAS group did not exhibit significant SPG topological abnormalities at baseline, and 12-week MAS treatment did not produce a significant change in left SPG degree, with heterogeneous individual responses showing both increases and decreases. However, we still found that increases in left SPG degree among HR-MAS subjects were associated with smaller inattention symptom improvements. This finding suggested a more complex neurobiological-clinical relationship in HR-MAS subjects, potentially influenced by additional factors such as genetic vulnerability, baseline network configuration or treatment responsivity. These distinct SPG topological profiles observed in the HR-MAS group may partially contribute to the observed group differences in treatment-related changes in inattention symptoms.

Another notable finding is that 12-week MAS treatment decreased right amygdala nodal efficiency in HR-MAS but not LR-MAS group. Compared with HC, no baseline differences in the right amygdala were observed for both LR-MAS and HR-MAS groups and lower right amygdala efficiency was found at week 12 in the HR-MAS group only. Prior studies have demonstrated that youth with familial risk for developing BD and varying clinical features (i.e., current mood disorder and/or ADHD) exhibit aberrant right amygdala resting state connectivity,35,36 and hyperactivation in response to emotional stimuli.37–39 Prior graph analyses have also found that right amygdala efficiency or clustering correlated positively with manic symptom severity in medication-free patients with BD or MDD.40,41 Moreover, treatment with the mood-stabilizer lithium decreased right amygdala clustering in BD patients who were manic/hypomanic at baseline.42 In the present study, HR-MAS group exhibited baseline-endpoint reductions in manic symptom severity suggesting that the decrease in right amygdala efficiency may be a protective response. Nevertheless, these findings provide evidence for unique right amygdala efficiency changes in ADHD youth with a BD family history in response to 12-week MAS treatment.

In view of observed differential changes in ADHD symptoms in LR-MAS and HR-MAS groups, we additionally evaluated whether ADHD treatment outcome could be predicted by baseline brain functional connectome properties. We found that baseline functional network dysconnectivity within the negative prediction network was a more robust predictor of higher treatment response in HR-MAS group. Primary regions that contributed to the prediction included mOFC, mPFC, ACC, visual, auditory and sensory processing areas, all of which are key components of the reward circuit.43,44 Previous studies have shown that increased functional connectivity in the reward circuit is closely related to the severity of ADHD symptoms45, and reward processing is known to involve dopamine and the activation of dopaminergic neurons.46 Given that primary pharmacological action of psychostimulants is to elevate synaptic dopamine levels, reduced connectivity within the reward circuit may indicate heightened sensitivity to dopamine-related modulatory pathways, which in turn is associated with improved therapeutic outcomes of psychostimulant treatment.

This study has several notable limitations. First, the MAS treatment duration was relatively short (i.e., 12 weeks) and more robust or different changes in connectome metrics may emerge following longer MAS exposure. Second, due to the challenges of movement control in ADHD youth, our study employed a relatively lenient head motion threshold which should be considered when interpreting the findings. Third, given the variability of test-retest reliability of fMRI-based network measures47,48 and small sample sizes, results should be interpreted with caution pending replication on larger datasets. Fourth, we constructed brain functional connectome using the AAL atlas, and the current findings should be replicated with multiple parcellation templates. Fifth, analyses in addition to resting-state fMRI, including task-based functional connectivity, are warranted to extend these findings. Sixth, correlational analyses were not corrected for multiple comparisons and should therefore be viewed as exploratory and hypothesis-generating. Seventh, the generalizability and stability of our treatment response prediction model is limited by the small sample size, and further validation using large-scale and independent datasets are warranted. Strengths of this study include a well-characterized cohort of psychostimulant-free ADHD youth with and without BD family history with similar group demographics, inclusion of a healthy comparison group, prospective study design, and stringent correction for multiple comparisons.

In summary, these findings provide novel evidence for differential changes in functional network organization in response to 12-week stimulant monotherapy in ADHD youth with and without a first-degree relative with BD. Specifically, stimulant-induced changes in global and nodal network topology observed in LR ADHD youth were normalizing compared with healthy youth, whereas these global and nodal network changes were not observed in HR ADHD youth. Moreover, only HR ADHD youth exhibited stimulant-induced decreases in right amygdala efficiency relative to both LR ADHD youth and healthy youth. We also found that stimulant-induced changes in global and nodal network topology metrics correlated with reductions in ADHD symptoms in LR ADHD youth only, and that baseline network dysconnectivity differentially predicted changes in ADHD symptoms in LR and HR ADHD groups. Collectively, these prospective findings add to a growing body of evidence that ADHD in conjunction with BD family history may represent a distinct neurobiology compared with ADHD alone.

Supplementary Material

1

Disclosure:

Kun Qin has received research funding from Hubei Provincial Natural Science Foundation of China (grant No. 2024AFB366). Nanfang Pan has received research funding from Natural Science Foundation of China. Du Lei has received research funding from Natural Science Foundation of Chongqing. L. Rodrigo Patino has received research funding from NIH, PCORI, Abbvie, Allergan, Janssen, Johnson and Johnson, Lundbeck, Lilly, Otsuka, Pfizer, and Sunovion. Qiyong Gong has received research funding from Natural Science Foundation of China. Melissa P. DelBello has received research support from NIH, PCORI, Acadia, Alkermes, Janssen, Johnson and Johnson, Lundbeck, Otsuka, Pfizer, Sage, Sunovion, and Vanda. She is also a consultant or on the advisory board for Alkermes, Allergan, Janssen, Johnson and Johnson, Lundbeck, Merck, Myriad, and Sage. Robert K. McNamara has received research support from Martek Biosciences Inc, Royal DSM Nutritional Products, LLC, Inflammation Research Foundation, Ortho-McNeil Janssen, AstraZeneca, Eli Lilly, NARSAD, and NIH, and previously served on the scientific advisory board of the Inflammation Research Foundation.

This study was supported in part by R01 NIMH grant 097818 to R.K.M and M.P.D (Co-PIs); NIH had no further role in study design, in the collection, analysis and interpretation of data, in the writing of the report, or in the decision to submit the paper for publication.

Dr. John Sweeney passed away in 2023, prior to the final version of the manuscript being accepted. The authors are grateful for his contributions to the conceptualization and critical revision of the work

The authors thank Thomas Blom, MS, of the University of Cincinnati for assistance with the statistical analyses.

Footnotes

Consent has been provided for descriptions of specific patient information.

Thomas Blom, MS, served as the statistical expert for this research.

Ziyu Zhu and John A. Sweeney have reported no biomedical financial interests or potential conflicts of interest.

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.

Data Sharing:

Researchers interested in accessing the data may contact the corresponding author to explore the possibility of sharing data under appropriate agreements that ensure compliance with ethical and legal guidelines.

REFERENCES

  • 1.Meier SM, Pavlova B, Dalsgaard S, et al. Attention-deficit hyperactivity disorder and anxiety disorders as precursors of bipolar disorder onset in adulthood. Br J Psychiatry. 2018;213(3):555–560. doi: 10.1192/bjp.2018.111 [DOI] [PubMed] [Google Scholar]
  • 2.Singh MK, DelBello MP, Kowatch RA, Strakowski SM. Co-occurrence of bipolar and attention-deficit hyperactivity disorders in children. Bipolar Disord. 2006;8(6):710–720. doi: 10.1111/j.1399-5618.2006.00391.x [DOI] [PubMed] [Google Scholar]
  • 3.Brancati GE, Perugi G, Milone A, Masi G, Sesso G. Development of bipolar disorder in patients with attention-deficit/hyperactivity disorder: A systematic review and meta-analysis of prospective studies. J Affect Disord. 2021;293:186–196. doi: 10.1016/j.jad.2021.06.033 [DOI] [PubMed] [Google Scholar]
  • 4.Gordovez FJA, McMahon FJ. The genetics of bipolar disorder. Mol Psychiatry. 2020;25(3):544–559. doi: 10.1038/s41380-019-0634-7 [DOI] [PubMed] [Google Scholar]
  • 5.De la Serna E, Ilzarbe D, Sugranyes G, et al. Lifetime psychopathology in child and adolescent offspring of parents diagnosed with schizophrenia or bipolar disorder: a 2-year follow-up study. Eur Child Adolesc Psychiatry. 2021;30(1):117–129. doi: 10.1007/s00787-020-01500-z [DOI] [PubMed] [Google Scholar]
  • 6.Lau P, Hawes DJ, Hunt C, Frankland A, Roberts G, Mitchell PB. Prevalence of psychopathology in bipolar high-risk offspring and siblings: a meta-analysis. Eur Child Adolesc Psychiatry. 2018;27(7):823–837. doi: 10.1007/s00787-017-1050-7 [DOI] [PubMed] [Google Scholar]
  • 7.Birmaher B, Merranko J, Hafeman D, et al. A Longitudinal Study of Psychiatric Disorders in Offspring of Parents With Bipolar Disorder From Preschool to Adolescence. J Am Acad Child Adolesc Psychiatry. 2021;60(11):1419–1429. doi: 10.1016/j.jaac.2021.02.023 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Khoury E, Acquaviva E, Purper-Ouakil D, Delorme R, Ellul P. Meta-analysis of personal and familial co-occurrence of Attention Deficit/Hyperactivity Disorder and Bipolar Disorder. Neurosci Biobehav Rev. 2023;146:105050. doi: 10.1016/j.neubiorev.2023.105050 [DOI] [PubMed] [Google Scholar]
  • 9.Kim JW, Yu H, Ryan ND, et al. Longitudinal trajectories of ADHD symptomatology in offspring of parents with bipolar disorder and community controls. J Clin Psychiatry. 2015;76(5):599–606. doi: 10.4088/JCP.14m09095 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Chen C, Tallman MJ, Cecil KM, et al. Symptom Profiles, But Not Prefrontal Neurochemistry, Differentiate ADHD Youth With and Without a Family History of Bipolar I Disorder. J Atten Disord. 2022;26(13):1762–1773. doi: 10.1177/10870547221101645 [DOI] [PubMed] [Google Scholar]
  • 11.Liu N, Liu Q, Yang Z, et al. Different functional alteration in attention-deficit/hyperactivity disorder across developmental age groups: A meta-analysis and an independent validation of resting-state functional connectivity studies. CNS Neurosci Ther. 2023;29(1):60–69. doi: 10.1111/cns.14032 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Bullmore E, Sporns O. Complex brain networks: graph theoretical analysis of structural and functional systems. Nat Rev Neurosci. 2009;10(3):186–198. doi: 10.1038/nrn2575 [DOI] [PubMed] [Google Scholar]
  • 13.Wang Y, Zuo C, Xu Q, Liao S, Kanji M, Wang D. Altered resting functional network topology assessed using graph theory in youth with attention-deficit/hyperactivity disorder. Prog Neuropsychopharmacol Biol Psychiatry. 2020;98:109796. doi: 10.1016/j.pnpbp.2019.109796 [DOI] [PubMed] [Google Scholar]
  • 14.Wang L, Zhu C, He Y, et al. Altered small-world brain functional networks in children with attention-deficit/hyperactivity disorder. Hum Brain Mapp. 2009;30(2):638–649. doi: 10.1002/hbm.20530 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Chen Y, Huang X, Wu M, et al. Disrupted brain functional networks in drug-naïve children with attention deficit hyperactivity disorder assessed using graph theory analysis. Hum Brain Mapp. 2019;40(17):4877–4887. doi: 10.1002/hbm.24743 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Qin K, Lei D, Zhu Z, et al. Different brain functional network abnormalities between attention-deficit/hyperactivity disorder youth with and without familial risk for bipolar disorder. Eur Child Adolesc Psychiatry. 2024;33(5):1395–1405. doi: 10.1007/s00787-023-02245-1 [DOI] [PubMed] [Google Scholar]
  • 17.Greenhill LL, Pliszka S, Dulcan MK, et al. Practice parameter for the use of stimulant medications in the treatment of children, adolescents, and adults. J Am Acad Child Adolesc Psychiatry. 2002;41(2 Suppl):26S–49S. doi: 10.1097/00004583-200202001-00003 [DOI] [PubMed] [Google Scholar]
  • 18.Perugi G, Vannucchi G, Bedani F, Favaretto E. Use of Stimulants in Bipolar Disorder. Curr Psychiatry Rep. 2017;19(1):7. doi: 10.1007/s11920-017-0758-x [DOI] [PubMed] [Google Scholar]
  • 19.Moran LV, Skinner JP, Shinn AK, et al. Risk of Incident Psychosis and Mania With Prescription Amphetamines. Am J Psychiatry. 2024;181(10):901–909. doi: 10.1176/appi.ajp.20230329 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Zimmer L Contribution of Clinical Neuroimaging to the Understanding of the Pharmacology of Methylphenidate. Trends Pharmacol Sci. 2017;38(7):608–620. doi: 10.1016/j.tips.2017.04.001 [DOI] [PubMed] [Google Scholar]
  • 21.Parlatini V, Bellato A, Murphy D, Cortese S. From neurons to brain networks, pharmacodynamics of stimulant medication for ADHD. Neurosci Biobehav Rev. 2024;164:105841. doi: 10.1016/j.neubiorev.2024.105841 [DOI] [PubMed] [Google Scholar]
  • 22.An L, Cao XH, Cao QJ, et al. Methylphenidate normalizes resting-state brain dysfunction in boys with attention deficit hyperactivity disorder. Neuropsychopharmacology. 2013;38(7):1287–1295. doi: 10.1038/npp.2013.27 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Rubia K, Halari R, Cubillo A, Mohammad AM, Brammer M, Taylor E. Methylphenidate normalises activation and functional connectivity deficits in attention and motivation networks in medication-naïve children with ADHD during a rewarded continuous performance task. Neuropharmacology. 2009;57(7–8):640–652. doi: 10.1016/j.neuropharm.2009.08.013 [DOI] [PubMed] [Google Scholar]
  • 24.Kaiser A, Broeder C, Cohen JR, Douw L, Reneman L, Schrantee A. Effects of a single-dose methylphenidate challenge on resting-state functional connectivity in stimulant-treatment naive children and adults with ADHD. Hum Brain Mapp. 2022;43(15):4664–4675. doi: 10.1002/hbm.25981 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Yoo JH, Kim D, Choi J, Jeong B. Treatment effect of methylphenidate on intrinsic functional brain network in medication-naïve ADHD children: A multivariate analysis. Brain Imaging Behav. 2018;12(2):518–531. doi: 10.1007/s11682-017-9713-z [DOI] [PubMed] [Google Scholar]
  • 26.Qin K, Sweeney JA, DelBello MP. The inferior frontal gyrus and familial risk for bipolar disorder. Psychoradiology. 2022;2(4):171–179. doi: 10.1093/psyrad/kkac022 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Pan N, Qin K, Patino LR, et al. Aberrant brain network topology in youth with a familial risk for bipolar disorder: a task-based fMRI connectome study. J Child Psychol Psychiatry. 2024;65(8):1072–1086. doi: 10.1111/jcpp.13946 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Shen X, Finn ES, Scheinost D, et al. Using connectome-based predictive modeling to predict individual behavior from brain connectivity. Nat Protoc. 2017;12(3):506–518. doi: 10.1038/nprot.2016.178 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Wang J, Wang X, Xia M, Liao X, Evans A, He Y. GRETNA: a graph theoretical network analysis toolbox for imaging connectomics. Front Hum Neurosci. 2015;9:386. doi: 10.3389/fnhum.2015.00386 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Rubinov M, Sporns O. Complex network measures of brain connectivity: uses and interpretations. Neuroimage. 2010;52(3):1059–1069. doi: 10.1016/j.neuroimage.2009.10.003 [DOI] [PubMed] [Google Scholar]
  • 31.Suo X, Lei D, Li L, et al. Psychoradiological patterns of small-world properties and a systematic review of connectome studies of patients with 6 major psychiatric disorders. J Psychiatry Neurosci. 2018;43(6):427. doi: 10.1503/jpn.170214 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Wang L, Zhu C, He Y, et al. Altered small-world brain functional networks in children with attention-deficit/hyperactivity disorder. Hum Brain Mapp. 2009;30(2):638–649. doi: 10.1002/hbm.20530 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Chen Y, Huang X, Wu M, et al. Disrupted brain functional networks in drug-naïve children with attention deficit hyperactivity disorder assessed using graph theory analysis. Hum Brain Mapp. 2019;40(17):4877–4887. doi: 10.1002/hbm.24743 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Silk TJ, Malpas C, Vance A, Bellgrove MA. The effect of single-dose methylphenidate on resting-state network functional connectivity in ADHD. Brain Imaging Behav. 2017;11(5):1422–1431. doi: 10.1007/s11682-016-9620-8 [DOI] [PubMed] [Google Scholar]
  • 35.Singh MK, Chang KD, Kelley RG, Saggar M, Reiss AL, Gotlib IH. Early signs of anomalous neural functional connectivity in healthy offspring of parents with bipolar disorder. Bipolar Disord. 2014;16(7):678–689. doi: 10.1111/bdi.12221 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Lin K, Shao R, Lu R, et al. Resting-state fMRI signals in offspring of parents with bipolar disorder at the high-risk and ultra-high-risk stages and their relations with cognitive function. J Psychiatr Res. 2018;98:99–106. doi: 10.1016/j.jpsychires.2018.01.001 [DOI] [PubMed] [Google Scholar]
  • 37.Ladouceur CD, Diwadkar VA, White R, et al. Fronto-limbic function in unaffected offspring at familial risk for bipolar disorder during an emotional working memory paradigm. Dev Cogn Neurosci. 2013;5:185–196. doi: 10.1016/j.dcn.2013.03.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Chang K, Garrett A, Kelley R, et al. Anomalous prefrontal-limbic activation and connectivity in youth at high-risk for bipolar disorder. J Affect Disord. 2017;222:7–13. doi: 10.1016/j.jad.2017.05.051 [DOI] [PubMed] [Google Scholar]
  • 39.Manelis A, Ladouceur CD, Graur S, et al. Altered amygdala-prefrontal response to facial emotion in offspring of parents with bipolar disorder. Brain. 2015;138(Pt 9):2777–2790. doi: 10.1093/brain/awv176 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Cha J, Spielberg JM, Hu B, Altinay M, Anand A. Resting-state functional connectivity graph-properties correlate with bipolar disorder-risk in young medication-free depressed subjects: Bipolar-risk Resting State Functional Connectivity in Major Depression. J Affect Disord. 2022;301:52–59. doi: 10.1016/j.jad.2022.01.033 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Spielberg JM, Beall EB, Hulvershorn LA, Altinay M, Karne H, Anand A. Resting State Brain Network Disturbances Related to Hypomania and Depression in Medication-Free Bipolar Disorder. Neuropsychopharmacology. 2016;41(13):3016–3024. doi: 10.1038/npp.2016.112 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Spielberg JM, Matyi MA, Karne H, Anand A. Lithium monotherapy associated longitudinal effects on resting state brain networks in clinical treatment of bipolar disorder. Bipolar Disord. 2019;21(4):361–371. doi: 10.1111/bdi.12718 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Hiser J, Koenigs M. The Multifaceted Role of the Ventromedial Prefrontal Cortex in Emotion, Decision Making, Social Cognition, and Psychopathology. Biol Psychiatry. 2018;83(8):638–647. doi: 10.1016/j.biopsych.2017.10.030 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Rolls ET, Cheng W, Feng J. The orbitofrontal cortex: reward, emotion and depression. Brain Commun. 2020;2(2):fcaa196. doi: 10.1093/braincomms/fcaa196 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Tomasi D, Volkow ND. Abnormal functional connectivity in children with attention-deficit/hyperactivity disorder. Biol Psychiatry. 2012;71(5):443–450. doi: 10.1016/j.biopsych.2011.11.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Speranza L, di Porzio U, Viggiano D, de Donato A, Volpicelli F. Dopamine: The Neuromodulator of Long-Term Synaptic Plasticity, Reward and Movement Control. Cells. 2021;10(4):735. doi: 10.3390/cells10040735 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Braun U, Plichta MM, Esslinger C, et al. Test-retest reliability of resting-state connectivity network characteristics using fMRI and graph theoretical measures. Neuroimage. 2012;59(2):1404–1412. doi: 10.1016/j.neuroimage.2011.08.044 [DOI] [PubMed] [Google Scholar]
  • 48.Cao H, Plichta MM, Schäfer A, et al. Test-retest reliability of fMRI-based graph theoretical properties during working memory, emotion processing, and resting state. Neuroimage. 2014;84:888–900. doi: 10.1016/j.neuroimage.2013.09.013 [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
NIHMS2110111-supplement-1.docx (6MB, docx)

Data Availability Statement

Researchers interested in accessing the data may contact the corresponding author to explore the possibility of sharing data under appropriate agreements that ensure compliance with ethical and legal guidelines.

RESOURCES