Greater Dynamic and Lower Static Functional Brain Connectivity Prospectively Predict Placebo Response in Pediatric Generalized Anxiety Disorder

Abstract

Objectives:

Placebo response is one of the most significant barriers to detecting treatment effects in pediatric (and adult) clinical trials focusing on affective and anxiety disorders. We sought to identify neurofunctional predictors of placebo response in adolescents with generalized anxiety disorder (GAD) by examining dynamic and static functional brain connectivity.

Methods:

Before randomization to blinded placebo, adolescents, aged 12–17 years, with GAD (N = 25) underwent resting state functional magnetic resonance imaging. Whole brain voxelwise correlation analyses were used to determine the relationship between change in anxiety symptoms from baseline to week 8 and seed-based dynamic and static functional connectivity maps of regions in the salience and ventral attention networks (amygdala, dorsal anterior cingulate cortex [dACC], and ventrolateral prefrontal cortex [VLPFC]).

Results:

Greater dynamic functional connectivity variability in amygdala, dACC, VLPFC, and regions within salience, default mode, and frontoparietal networks was associated with greater placebo response. Lower static functional connectivity between amygdala and dorsolateral prefrontal cortex, amygdala and medial prefrontal cortex, dACC and posterior cingulate cortex and greater static functional connectivity between VLPFC and inferior parietal lobule were associated with greater placebo response.

Conclusion:

Placebo response is associated with a distinct dynamic and static connectivity fingerprint characterized by “variable” dynamic but “weak” static connectivity in the salience, default mode, frontoparietal, and ventral attention networks. These data provide granular evidence of how circuit-based biotypes mechanistically relate to placebo response. Finding biosignatures that predict placebo response is critically important in clinical psychopharmacology and to improve our ability to detect medication-placebo differences in clinical trials.

Introduction

Placebo response is one of the most significant barriers to detecting treatment effects in pediatric (and adult) clinical trials focusing on affective and anxiety disorders (Walkup 2017). Rates of placebo response are increasing and now reach 40%–50% in some clinical trials of children and adolescents with depressive and anxiety disorders (Bridge et al. 2009; Emslie et al. 2014). Clinical improvement while patients receive placebo during a clinical trial (i.e., placebo response) remains poorly understood despite several recent reports of clinical predictors of its emergence in children and adolescents with depressive and anxiety disorders (Cohen et al. 2010; Dobson and Strawn 2016; Strawn et al. 2017a, 2017b). The extant data suggest that the magnitude of placebo response relates to the disorder being studied (e.g., anxiety disorder vs. depressive disorder), the age of patients being studied, funding source, and trial design (e.g., number of sites, number of subjects, randomization, and duration) (Cohen et al. 2010; Dobson and Strawn 2016).

Two meta-analyses examined clinical and trial design predictors of placebo response in pediatric anxiety disorders. In the first, placebo response was associated with race (with Caucasian participants having lower placebo responses), gender (with males having lower placebo response), and the presence of a study washout period (with longer washout periods associated with a lower placebo response) (Cohen et al. 2010). In the second meta-analysis, which examined 14 trials (9 medications) in pediatric patients (N = 2230) with anxiety disorders (Dobson and Strawn 2016), higher placebo response rates were associated with a greater number of study sites and fewer patients per site. Lower placebo response rates were associated with federally funded studies, studies conducted in the United States, and studies of anxiety disorders (compared to depressive disorders). In addition, one prior study examined predictors of placebo response among individual patients (N = 76) in the Child/Adolescent Anxiety Multimodal Study, a multisite, randomized controlled trial that examined the efficacy of cognitive-behavioral therapy, sertraline, their combination, and placebo for the treatment of separation and generalized and social anxiety disorders (Walkup et al. 2008). In this sample, placebo response was associated with having a diagnosis of separation anxiety disorder and with treatment expectation but response was not associated with age, race, or other demographics. In addition, placebo response typically occurred early in the course of treatment and then plateaued (Strawn et al. 2017a, 2017b).

The neurobiology of placebo response has received little attention in adult and pediatric patients (Mayberg et al. 2002; Benedetti et al. 2005; Brown and Pecina 2019). In fact, to our knowledge there are no studies examining neurofunctional predictors of placebo response in pediatric anxiety or depressive disorders. In adults with major depressive disorder (MDD) and co-occurring pain symptoms, one functional magnetic resonance imaging (fMRI) study identified decreased connectivity density within the pain network in placebo treated patients that inversely correlated with placebo-related improvement in depressive symptoms (Wang et al. 2019). In adults with MDD (N = 34), Pecina et al. (2015) found pretreatment μ-opioid receptor binding in the nucleus accumbens correlated with improvement in the Quick Inventory of Depressive Symptoms over 10 weeks of treatment (r = 0.47, p = 0.01). In addition, in patients who received placebo, changes in μ-opioid receptor binding in the thalamus, subgenual anterior cingulate cortex, and amygdala were associated with subsequent improvement in depressive symptoms (Pecina et al. 2015).

The neurobiology of pediatric anxiety disorders involves structural and functional abnormalities within salience, attention, and default mode networks (Blackford and Pine 2012; Sylvester et al. 2012; Guyer et al. 2013; Strawn et al. 2014a, 2014b; Williams 2016). Within these networks, gray matter volumes (Mueller et al. 2013; Strawn et al. 2013, 2015; Gold et al. 2017), cortical thickness (Strawn et al. 2014a, 2014b), functional activity (Monk et al. 2006, 2008; Beesdo et al. 2009; Strawn et al. 2012), and functional connectivity (Strawn et al. 2012; Roy et al. 2013; Kujawa et al. 2016) differ in the ventrolateral prefrontal cortex (VLPFC), dorsal anterior cingulate cortex (dACC), and amygdala. Furthermore, in youth who are at increased risk for developing anxiety disorders, the functional activity and connectivity of these structures are already altered (Rogers et al. 2017; Sylvester et al. 2018). Taken together, these studies suggest that altered connectivity of these networks is a central feature of pediatric anxiety disorders. Most studies of pediatric anxiety (and other internalizing) disorders have focused on static connectivity of these networks, which use the correlation coefficient of two time series across the entire scan and assume that the degree of connectivity strength between regions is constant over time. Temporal variations in connectivity strength are not captured (Hutchison et al. 2013). Yet, particularly in anxiety disorders, alterations in attentional and emotion processing are dynamic and continually shift as patients acquire and process environmental data with regard to their internal world and schemas. Functional connectivity within these networks likely fluctuates over time (Kucyi et al. 2017). But static connectivity analysis cannot provide information about connectivity fluctuation that may be especially important in complex psychiatric disorders. Dynamic functional connectivity analysis assesses functional connectivity variability over time and complements data from static connectivity. As complementary approaches, the combination of dynamic and static functional connectivity allows the variance and mean connectivity strength to be examined. Moreover, these dynamic approaches have provided important insights into the circuitry of affective disorders in adults (Liao et al. 2018; Li et al. 2019a, 2019b). This approach has potential to provide novel information about functional connectivity alterations for pediatric anxiety disorders (Hutchison et al. 2013). However, there are no studies of dynamic functional connectivity in adolescents with anxiety disorders, and specifically, there are no static or dynamic functional connectivity studies with regard to treatment (or placebo) response in youth.

We addressed these knowledge gaps by examining both static and dynamic functional connectivity and subsequent placebo response in unmedicated, largely treatment naive adolescents with generalized anxiety disorder (GAD). Specifically, we examined the resting-state functional connectivity of three hubs from the ventral attention (VLPFC) and salience networks (dACC and amygdala), which have been previously linked to anxiety disorder (Sylvester et al. 2018). We determined the relationship between seed-based functional connectivity maps from these nodes and placebo-related change in anxiety symptoms in addition to examining associations of clinical and demographic factors to placebo response in this study sample. We hypothesized that, based on the currently understood neurobiology of pediatric anxiety disorders, altered static and dynamic functional connectivity from the dACC, VLPFC, and amygdala to regions that subserve error sensitivity and attention/emotion regulation within the salience, ventral attention, and default mode networks would predict greater placebo response.

Methods

Participants and procedures

The protocol was approved by the Institutional Review Board for the University of Cincinnati and conducted in accordance with Good Clinical Practice guidelines. The study was conducted at a single academic site in the United States from February 2015 (first patient visit) to November 2018 (last patient visit). All patients and their parents (or legal guardians) provided written informed assent and consent before study participation.

Outpatients aged 12–17 years who met Diagnostic and Statistical Manual of Mental Disorders, 4th ed., Text Revision (DSM-IV-TR; American Psychiatric Association 2000) criteria for GAD, assessed using the Anxiety Disorders Interview Schedule (Silverman and Albano 1996), had a Pediatric Anxiety Rating Scale (PARS) score ≥15 at screening and baseline visits (Riddle et al. 2002), and had a Clinical Global Impressions-Severity (CGI-Severity) score ≥4 were eligible (Guy 1976). Patients were required to be medically stable and provide a negative urine drug screen and a negative urine pregnancy test (for girls) at screening. Patients with secondary diagnoses of separation or social anxiety disorder or panic disorder and/or agoraphobia were enrolled, provided that GAD was the primary diagnosis; however, patients with current MDD or any history of bipolar disorder, psychotic disorder, obsessive compulsive disorder, or posttraumatic stress disorder were excluded. Patients were randomized to receive a selective serotonin reuptake inhibitor (SSRI) or placebo (1:1). The study incorporated a 1-week screening period and an 8-week double-blind placebo-controlled treatment period.

Assessment of anxiety symptoms and response

The PARS score was the primary dimensional outcome measure for anxiety symptom severity. For patients who dropped out before week 8 (end point), a last observation carried forward approach was utilized for missing PARS scores. Any patient who had a 50% worsening in PARS score at two consecutive visits was discontinued from the study. Finally, the change from baseline to end point (week 8/early termination) provided a continuous measure of placebo response.

Neuroimaging and acquisition

Images were acquired on a 3-T scanner (Achieva; Philips) with a 32-channel phased-array head coil. Scanner noise was attenuated with earplugs and headphones; head motion was restricted with foam padding around head. Functional images were obtained using a single shot, fast Fourier echo–echo planar imaging sequence with the following parameters: repetition time (TR) = 2000 ms; echo time (TE) = 30 ms; number of axial slices = 40; resolution = 2.8 mm × 2.8 mm; slice thickness = 3 mm; flip angle = 75°; matrix = 80 × 80; and field of view = 224 mm × 224 mm (Weaver et al. 2013; Fonseca Pachi et al. 2019). High-resolution anatomical images were obtained using a three dimensional T1-weighted Turbo field echo sequence with the following parameters: TR = 6.8 ms; TE = 2.9 ms; number of sagittal slices = 160; resolution = 1 mm × 1 mm; slice thickness = 1 mm; flip angle = 9°; matrix = 256 × 256; and field of view = 256 mm × 256 mm. All images were reviewed by a pediatric neuroradiologist to assess any structural abnormalities and then examined by a second radiologist (L.L.) to assess for the presence of any artifacts.

fMRI data preprocessing

Data were preprocessed with the SPM12 package and DPABI toolbox (Yan et al. 2016). For each patient, the first 10 images were discarded to ensure signal stabilization. The remaining images were corrected for head motion by regression of 24 head motion parameters (Friston et al. 1996), and mean framewise displacement was <0.2 mm. Functional images were spatially normalized to standard Montreal Neurological Institute space using unified segmentation on individual T1 images (Ashburner and Friston 2005). The normalized images were then smoothed with a 6-mm full width at half-maximum Gaussian kernel. Linear trends and nuisance signals (six motion parameters, white matter signal, and cerebrospinal fluid signal) were removed with linear regression, and a temporal band pass filter (0.01–0.08 Hz) was utilized to exclude high and low frequency signals.

Seed-based dynamic functional connectivity

Three regions of interest from the salience network (amygdala, dACC) and the ventral attention network (VLPFC) were defined using the Brainnetome atlas (Fan et al. 2016). To detect dynamic functional connectivity of these regions, a sliding window approach (Allen et al. 2014) in DynamicBC toolbox (Liao et al. 2014) was used. A sliding window length of 20 TR (40 seconds), which captures resting-state functional connectivity fluctuations (Preti and Van de Ville 2017), and a step of 1 TR (2 seconds) were used to produce 121 temporal windows. This window length was chosen given that cognitive states and brain networks were found to be stabilized at window lengths of about 30–60 seconds (Jones et al. 2012; Shirer et al. 2012). In each sliding window, temporal correlation coefficients between the truncated time course of the seeds and all the other gray matter voxels were calculated. Thus, 121 correlation maps were created for each participant. A Fisher's r-to-z transformation was then applied to all the correlation maps to improve the normality of the distribution of correlation values. The variance of these correlation coefficients was computed by calculating the standard deviation of z values at each voxel to assess functional connectivity variability (i.e., dynamic functional connectivity).

Seed-based static functional connectivity

Seed-based resting-state functional connectivity analyses were conducted for the same seeds (above) using the Resting-State fMRI Data Analysis Tool Kit. Specifically, the time series within each region of interest (averaging across all voxels) was extracted. Then, voxelwise correlation analyses were performed between each region of interest and the rest of the brain to obtain functional connectivity maps. Finally, the correlation coefficients were transformed to z-value images using the Fisher r-to-z transformation to improve the normality of the functional connectivity maps.

Post hoc comparisons of functional connectivity variability and strength in regions associated with placebo response in the entire sample were examined in patients who met response criteria for placebo response compared to those without “response,” based on >35% reduction in PARS score from baseline to end point (see below) (Caporino et al. 2013).

Statistical analyses

Demographic and clinical variables (e.g., age, sex, race) were incorporated into a multiple regression model to examine the relationship between these variables and change in PARS score from baseline to end point/early termination. Nonimaging statistical analyses were performed using R (version 3.1.2), and p-values <0.05 were considered statistically significant.

To detect the brain connectivity associated with placebo response, a whole-brain correlation analysis was conducted examining relationships between baseline functional connectivity measures and change in anxiety symptoms (i.e., change in PARS score from week 8 to baseline). An AlphaSim approach was applied to correct for multiple comparisons in these analyses, with a threshold of p < 0.005 at the voxel level and p < 0.05 at the cluster level (dynamic functional connectivity minimum cluster size: 78 voxels; static functional connectivity minimum cluster size: 41 voxels). In the post hoc analysis, we extracted the functional connectivity variability and strength from the clusters that were significantly associated with placebo response in the whole-brain correlation analysis and compared it between the placebo responders and nonresponders using independent t-tests, with p < 0.05 considered statistically significant. Finally, the sensitivity of the dynamic functional connectivity results was examined at a longer window length (i.e., TR of 25 [50 seconds]).

Results

Characteristics of placebo-treated patients

Twenty-five patients were randomized to placebo, and resting state functional connectivity data were available for 22 of them. Demographic and clinical characteristics of the imaging cohort are shown in Table 1. At the time of baseline scans, all patients had moderate-to-severe anxiety (mean baseline PARS score = 17 ± 2, mean CGI-Severity score = 4). Multiple linear regression analysis found no significant relationship between change in PARS score over the course of 8 weeks of receiving placebo and age, sex, baseline PARS, treatment expectation, secondary diagnosis, and race (Table 2). Furthermore, age, sex, race, full-scale intelligence quotient, baseline PARS, secondary diagnosis, and prior SSRI/serotonin-norepinephrine reuptake inhibitor treatment did not differ between patients with and without a 35% reduction in PARS score (Supplementary Table S1).

Table 1.

Demographic and Clinical Characteristics of Imaging Study Patients

Characteristics	Placebo, n = 22
Age, years	14.9 ± 1.6
Girls, n (%)	16 (73)
Race
Asian	2(9)
Black and African American	1 (5)
Caucasian	18 (82)
Other	1 (5)
Hispanic or Latino	0 (0)
Full scale IQ	104 ± 11
PARS score, baseline	17 ± 2
PARS score, week 8/ET	13 ± 5
CGI-S score, median	4
Diagnoses
Separation anxiety disorder	4 (18)
Panic disorder	14 (64)
Agoraphobia	7 (32)
ADHD	4 (18)
Specific phobia	2 (9)
Prior SSRI/SNRI treatment, n (%)	6 (27)

CGI-S, Clinical Global Impressions-Severity; PARS, Pediatric Anxiety Rating Scale; ADHD, attention-deficit/hyperactivity disorder; IQ, intelligence quotient; SSRI, selective serotonin reuptake inhibitor; SNRI, serotonin-norepinephrine reuptake inhibitor; ET, early termination.

Table 2.

Regression Model of Change in Pediatric Anxiety Rating Scale Score Over Eight Weeks in Adolescents Receiving Placebo (n = 25)

	Estimate	SE	z Value	p-Value
Age	1.030	0.807	1.277	0.234
Sex	−0.896	2.820	−0.317	0.758
Baseline PARS	−0.234	0.867	−0.270	0.793
Treatment expectations^a	−0.654	0.680	−0.961	0.361
Separation anxiety disorder	−5.229	5.008	−1.044	0.324
Social anxiety disorder	4.734	3.187	1.485	0.172
Panic disorder	2.164	3.034	0.713	0.494
ADHD	0.237	3.029	0.078	0.939
Race (nonwhite)	6.901	3.522	1.960	0.081

This scale was not collected for three of the patients.

PARS, Pediatric Anxiety Rating Scale; ADHD, attention-deficit/hyperactivity disorder; SE, standard error.

Placebo response and baseline dynamic functional connectivity

Improvement in anxiety symptoms was associated with more variable dynamic functional connectivity between: (1) left amygdala and left inferior parietal lobule (IPL), (2) left amygdala and contralateral anterior insula, (3) left amygdala and right angular gyrus, and (4) right amygdala and left fusiform gyrus at baseline, before placebo initiation (Fig. 1A, B; Table 3). More variable dynamic connectivity between left dACC and left angular gyrus and between right dACC and contralateral medial prefrontal cortex (mPFC)/dorsolateral prefrontal cortex, as well as between left VLPFC and ipsilateral IPL, correlated with greater subsequent placebo response (Fig. 1C–E; Table 3).

FIG. 1.

Increased dynamic FC associated with placebo response in adolescents with generalized anxiety disorder. For each seed (left column, red) significant whole-brain dynamic FC associations are shown (middle column). Placebo response (right column) was associated with decreased dynamic FC between the left amygdala and left IPL, between the left amygdala and right insula, and between the left amygdala and right angular (A). For the right amygdala, placebo response was associated with decreased dynamic FC to the left fusiform (B). Placebo response was also associated with decreased dynamic FC between left dACC and left angular gyrus (C), between the right dACC and left mPFC (D), and between the left VLPFC and IPL (E). dACC, dorsal anterior cingulate cortex; FC, functional connectivity; IPL, inferior parietal lobule; mPFC, medial prefrontal cortex; VLPFC, ventrolateral prefrontal cortex. Color images are available online.

Table 3.

Brain Regions Significantly Related with Placebo Response in Pediatric Anxiety Disorder (n = 22)

ROIs	Brain regions	Cluster size (voxels)	Peak t value	Peak MNI coordinate
ROIs	Brain regions	Cluster size (voxels)	Peak t value	X	Y	Z
Dynamic FC
Left amygdala	Left IPL	71	4.86	−50	−34	45
	Right insula	59	4.54	36	14	−12
	Right angular	41	5.76	34	−73	33
Right amygdala	Left fusiform	58	6.06	−36	−62	−6
Left dACC	Left angular	56	4.83	−39	−53	33
Right dACC	Left mPFC	49	4.72	−20	14	54
Left VLPFC	Left IPL	51	5.50	−56	−45	33
Static FC
Right amygdala	Left DLPFC	123	5.39	−22	6	51
Right amygdala	Right mPFC	78	3.62	8	11	48
Left dACC	PCC	78	3.74	6	−56	36
Left VLPFC	Right IPL	129	3.85	36	−42	33

AlphaSim corrected, p < 0.05.

ROIs, region of interests; MNI, Montreal Neurological Institute; FC, functional connectivity; DLPFC, dorsolateral prefrontal cortex; mPFC, medial prefrontal cortex; dACC, dorsal anterior cingulate cortex; PCC, posterior cingulate cortex; VLPFC, ventrolateral prefrontal cortex; IPL inferior parietal lobule.

In the post hoc examination of categorical placebo response, compared with nonresponders, placebo responders had, at baseline, significantly increased dynamic functional connectivity between each of the seeds and targets described above (Supplementary Fig. S1). Regarding sensitivity analyses, the findings did not change significantly across 40–50 second window sizes (Supplementary Table S2).

Placebo response and baseline static functional connectivity

Placebo-related improvement in anxiety symptoms was associated with decreased static functional connectivity between right amygdala and left DLPFC, between right amygdala and right mPFC, and between dACC and bilateral posterior cingulate cortex (PCC) (Fig. 2A, B; Table 3). Placebo-related improvement in anxiety symptoms was also associated with greater static functional connectivity between left VLPFC and right IPL (Fig. 2C).

FIG. 2.

Static FC signatures associated with placebo response in adolescents with generalized anxiety disorder. For each seed (left column, red) significant, whole-brain static FC associations are shown (right column). (A). For the dACC (B), placebo response was associated with increased static FC to the PCC and for the VLPFC (C), placebo response was associated with decreased static FC to the IPL. dACC, dorsal anterior cingulate cortex; FC, functional connectivity; IPL, inferior parietal lobule; PCC, posterior cingulate cortex; VLPFC, ventrolateral prefrontal cortex. Color images are available online.

In the post hoc examination of categorical placebo response, placebo responders significantly differed in terms of baseline, static functional connectivity compared to nonresponders between each of the seeds and targets described above (Supplementary Fig. S1).

Discussion

To our knowledge, this is the first study to evaluate neurofunctional predictors of placebo response in pediatric patients with anxiety disorders and only the second to examine clinical or demographic predictors of placebo response in youth with anxiety (Strawn et al. 2017a, 2017b). Our findings demonstrate that more placebo-related improvement in anxiety symptoms is predicted by more variable dynamic functional connectivity and less static connectivity within and between components of salience, frontoparietal, ventral attention, and default mode networks. In other words, anxious adolescents who are more likely to respond to placebo have “weaker” but “more flexible” connectivity among these networks. These findings raise the possibility that functional connectivity patterns could serve as a biomarker for placebo response in pediatric anxiety disorders. Predicting which patients are more likely to have placebo response could allow clinical trials to be better designed to more efficiently and reliably evaluate novel treatments.

In our sample, more flexibility/variability in dynamic functional connectivity and lower strength of static connectivity predicted greater placebo response. Specifically, increased variability but decreased strength of functional connectivity between salience, frontoparietal, ventral attention, and default mode networks was associated with greater placebo response in adolescents with GAD. Importantly, these networks are pathophysiologically implicated in pediatric anxiety disorders (Sylvester et al. 2012) and subserve stimulus-driven attention, cognitive control, and emotion regulation processes that are impaired in anxiety disorders (Basten et al. 2011; Sylvester et al. 2012; Williams 2016). Weaker but more flexible connectivity across these networks may facilitate recovery from illness through more ready cognitive reappraisal and benefits from social supports. In addition, flexible connectivity could facilitate a better appraisal of risks in the environment and cognitive control over anxiety responses. Consistent with the potential for biomarkers reflecting neural circuit function to guide clinical interventions, these data raise the possibility that circuit-based biotypes could mechanistically relate to treatment response and specifically placebo response—a critically important process in psychopharmacology and clinical psychiatry.

Functional connectivity (1) within salience (amygdala-insula) network, (2) between the salience and frontoparietal (amygdala-DLPFC, amygdala-IPL) networks, and (3) between the salience and default mode (amygdala-mPFC, dACC–PCC, amygdala-angular, dACC-angular, dACC-mPFC) networks is associated with placebo response. The salience network detects salient stimuli and recruits relevant networks to respond to them in an appropriate context-relevant manner. Decreased static functional connectivity within and between salience and frontoparietal networks has been found in patients with GAD and individuals with high trait anxiety (Etkin et al. 2009; Basten et al. 2011). Our results suggest that more flexible connectivity within the salience network and from the salience network to the frontoparietal network is associated with increased placebo response.

Regarding attentional networks, we observed more variable dynamic functional connectivity between frontoparietal, ventral attention, and salience networks in patients with a more robust placebo response. Defined by prefrontal (e.g., mPFC, anterior insula), as well as parietal (e.g., IPL and precuneus) nodes, these networks subserve attentional processing that is closely linked with the neurobiology of anxiety disorders. Increased functional activity of these networks in response to shifts in attention and threat-oriented attention has been observed in task-based fMRI studies (Britton et al. 2013; Hardee et al. 2013; Clauss et al. 2016). Networks locked in to a state of increased activity and limited flexibility of activity in the network may have a reduced ability to respond during placebo treatment, while those with this neural network pattern may more often require pharmacological intervention to achieve clinical recovery.

Furthermore, hypoconnectivity within the attentional network in anxiety disorders (Qiu et al. 2011; Williams 2016) and trait anxiety (Tian et al. 2016) relates to threat bias, as well as behavioral inhibition (a risk factor for developing anxiety in youth) (Fu et al. 2017). That increased activity in attentional networks, including threat bias, is associated with placebo response is not surprising. In this regard, greater plasticity of attentional networks may subtend the relationship between threat bias and placebo response in that patients who attribute improvement to internal factors (e.g., change in cognition) as opposed to those who attribute improvement to external factors differ with regard to placebo response (Powers et al. 2008). Thus increased flexibility, as reflected by increased dynamic functional connectivity in attentional networks, may relate to a greater processing of and attention to external cues such as social supports and indications that an environment is safe.

In this study, increased dynamic connectivity between salience and default mode network (e.g., mPFC, angular gyrus) was associated with greater placebo response. Long implicated in the pathophysiology of pediatric anxiety disorders (Strawn et al. 2012; Roy et al. 2013), the default mode network processes self-referential cognition (Harrison et al. 2011) and generates mental representations of one's self and future actions, thoughts, and feelings in addition to assigning significance to thoughts about oneself (Andrews-Hanna et al. 2014; Li et al. 2014). In adults, with affective disorders, patients who fail to deactivate the default mode network during emotion processing tasks (compared to those who successfully deactivate this network) have less treatment-related improvement (Spies et al. 2017). This pattern may represent a functional competition between the two networks, as dynamic shifting from dominance of default mode to task-based networks is crucial for adaptive behavior (Royall et al. 2013; Backes et al. 2014).

Historically, clinical and demographic predictors of placebo response in anxious youth have identified a small proportion of patients who improve while receiving placebo (Walkup et al. 2008) and found that some features predict placebo response such as presence of separation anxiety disorder and treatment expectation (Strawn et al. 2017a, 2017b). Our findings are consistent with the earlier observation that placebo response is not associated with race, age, or other demographic features or baseline anxiety. In our sample, baseline functional connectivity profiles were strongly associated with subsequent placebo response, while clinical features that we examined all failed to identify placebo responders or predict the magnitude of placebo response. That youth with imaging biomarkers—not particular symptom or demographic features—might benefit more from selective interventions represents an important advance. If confirmed, the identification of a target functional connectivity profile of placebo response could inform ongoing Fast-Fail Trial initiatives (Grabb et al. 2016) which seek to (1) address the low success rate and high cost of novel psychopharmacologic intervention trials and (2) determine whether these interventions engage targets and how target engagement relates to clinical improvement (Javitt et al. 2018; Krystal et al. 2019). Reducing the rate of placebo response by excluding patients with specific functional connectivity alterations could significantly improve the success of new drug trials.

While this is the first exploration of the neurophysiology of placebo response in pediatric anxiety disorders, there are several limitations. First, the sample is small (N = 25) with three patients having incomplete functional connectivity data. However, we note that the sample size in this pediatric study is similar to all but one examination of the neurophysiology of placebo response in adults with anxiety or depressive disorders. Second, we are limited in our ability to resolve multicollinearity among variables as has been the case with other studies of placebo response (Strawn et al. 2017a, 2017b). As such, the complex relationships among variables related to placebo response are difficult to disentangle with our study sample. Third, the sample was relatively homogeneous (i.e., >80% Caucasian), which may limit the generalizability to other ethnicities. Fourth, characteristics other than those evaluated in the present study may influence response to pill placebo, including a number of psychological factors, therapeutic alliance, as well as genetic factors and biological features (e.g., sympathetic tone) and other features not evaluated in this sample. Fifth, we examined gray matter regions that are associated with change in anxiety symptoms; however, it may be possible that some white matter regions have functional roles in generating, predicting, or facilitating psychological processes (Li et al. 2019a, 2019b), including placebo response. Finally, we examined the change in symptoms in patients who received placebo, as is common in studies of placebo response; however, causal attribution would be enhanced if the study included a no-intervention comparison group.

Conclusions

Our study findings indicate that baseline neurofunctional (i.e., dynamic and static functional connectivity) signatures that can prospectively predict placebo response in anxious adolescents have potential implications for both clinical trials and clinical practice. Moreover, the findings that clinical and demographic factors are poor predictors of patient-specific placebo response highlight the potential value of functional connectivity predictors of placebo response. This study also provides a scaffold on which the neurobiology of placebo response can be better understood. Recently, there has been increased enthusiasm for using circuit function to guide clinical intervention (Williams 2016). These data provide evidence of how circuit-based biotypes might mechanistically relate to treatment response and specifically placebo response—a critically important process in psychopharmacology and specifically in child and adolescent psychiatry where background clinical history and utility of self-assessments can be limited. Harnessing this knowledge in the clinic could help to identify patients who are likely or unlikely to have a robust placebo response and those who might require additional interventions to maximize their improvement.

Clinical Significance

Additional pharmacologic interventions are urgently needed (Strawn et al. 2020), but the development of new treatments is impeded by high placebo response rates. Consequently, understanding the neurobiology of placebo response, particularly in pediatric populations, is critical to achieving this goal. Our study reveals that increased placebo response is predicted by increased dynamic and decreased static functional connectivity within and between salience, frontoparietal, ventral attention, and default mode networks. These findings suggest that increased neurofunctional flexibility in placebo responders could relate to more responsiveness to moment-to-moment shifts between external and internal stimuli.

Footnotes

Acknowledgments

The authors thank the patients and their families for participating in this study and the Data Safety Monitoring Board for their oversight of the study. In addition, we thank the MR technologists from the Imaging Research Center at Cincinnati Children's Hospital Medical Center and Blaise V. Jones, MD, Chief of Neuroradiology, Cincinnati Children's Hospital Medical Center for his over-read of the patients' neuroimaging.

Disclaimer

The views expressed within this article represent those of the authors and are not intended to represent the position of NIMH, NICHD, the National Institutes of Health (NIH), or the Department of Health and Human Services.

Disclosures

Dr. Strawn has received research support from Allergan, Neuronetics, Lundbeck, Otsuka and the National Institutes of Health. He receives royalties from Springer Publishing for two texts and has received material support from Myriad. He has consulted to the Food and Drug Administration and Myriad Genetics. Dr. Mills has received research support from the Yung Family Foundation and Dr. Cecil receives support from National Institute of Environmental Health Sciences (R01 ES027224, K.M.C.). Dr. DelBello 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. The remaining authors have nothing to disclose. The views expressed within this article represent those of the authors and are not intended to represent the position of NIMH, NICHD, the National Institutes of Health (NIH), or the Department of Health and Human Services.

Supplementary Material

Supplementary Figure S1

Supplementary Table S1

Supplementary Table S2

References

Allen

, Damaraju

, Plis

, Erhardt

, Eichele

, Calhoun

: Tracking whole-brain connectivity dynamics in the resting state. Cereb Cortex, 24:663–676, 2014.

American Psychiatric Association: Diagnostic and Statistical Manual of Mental

Disorders

, 4th ed., Text

Revision

. Washington, DC: American Psychiatric Association; 2000.

Andrews-Hanna

, Smallwood

, Spreng

: The default network and self-generated thought: Component processes, dynamic control, and clinical relevance. Ann N Y Acad Sci, 1316:29–52, 2014.

Ashburner

, Friston

: Unified segmentation. Neuroimage, 26:839–851, 2005.

Backes

, Dietsche

, Nagels

, Stratmann

, Konrad

, Kircher

, Krug

: Increased neural activity during overt and continuous semantic verbal fluency in major depression: Mainly a failure to deactivate. Eur Arch Psychiatry Clin Neurosci, 264:631–645, 2014.

Basten

, Stelzel

, Fiebach

: Trait anxiety modulates the neural efficiency of inhibitory control. J Cogn Neurosci, 23:3132–3145, 2011.

Beesdo

, Lau

JYF

, Guyer

, McClure-Tone

, Monk

, Nelson

, Fromm

, Goldwin

, Wittchen

, Leibenluft

, Erns

, Pine

: Common and distinct amygdala-function perturbations in depressed vs anxious adolescents. Arch Gen Psychiatry, 66:275–285, 2009.

Benedetti

, Mayberg

, Wager

, Stohler

, Zubieta

: Neurobiological mechanisms of the placebo effect. J Neurosci, 25:10390–10402, 2005.

Blackford

, Pine

: Neural substrates of childhood anxiety disorders: A review of neuroimaging findings. J Child Adolesc Psychiatr Clin N Am, 21:501–525, 2012.

10.

Bridge

, Birmaher

, Iyengar

, Barbe

, Brent

: Placebo response in randomized controlled trials of antidepressants for pediatric major depressive disorder. Am J Psychiatry, 166:42–49, 2009.

11.

Britton

, Grillon

, Lissek

, Norcross

, Szuhany

, Chen

, Ernst

, Nelson

, Leibenluft

, Shechner

, Pine

: Response to learned threat: An fMRI study in adolescent and adult anxiety. Am J Psychiatry, 170:1195–1204, 2013.

12.

Brown

, Pecina

: Neuroimaging studies of antidepressant placebo effects: Challenges and opportunities. Front Psychiatry, 10:669, 2019.

13.

Caporino

, Brodman

, Kendall

, Albano

, Sherrill

, Piacentini

, Sakolsky

, Birmaher

, Compton

, Ginsburg

, Rynn

, McCracken

, Gosch

, Keeton

, March

, Walkup

: Defining treatment response and remission in child anxiety: Signal detection analysis using the pediatric anxiety rating scale. J Am Acad Child Adolesc Psychiatry, 52:57–67, 2013.

14.

Clauss

, Benningfield

, Rao

, Blackford

: Altered prefrontal cortex function marks heightened anxiety risk in children. J Am Acad Child Adolesc Psychiatry, 55:809–816, 2016.

15.

Cohen

, Consoli

, Bodeau

, Purper-Ouakil

, Deniau

, Guile

, Donnelly

: Predictors of placebo response in randomized controlled trials of psychotropic drugs for children and adolescents with internalizing disorders. J Child Adolesc Psychopharmacol, 20:39–47, 2010.

16.

Dobson

, Strawn

: Placebo response in pediatric anxiety disorders: Implications for clinical trial design and interpretation. J Child Adolesc Psychopharmacol, 26:686–693, 2016.

17.

Emslie

, Prakash

, Zhang

, Pangallo

, Bangs

, March

: A double-blind efficacy and safety study of duloxetine fixed doses in children and adolescents with major depressive disorder. J Child Adolesc Psychopharmacol, 24:170–179, 2014.

18.

Etkin

, Prater

, Schatzberg

, Menon

, Greicius

: Disrupted amygdalar subregion functional connectivity and evidence of a compensatory network in generalized anxiety disorder. Arch Gen Psychiatry, 66:1361–1372, 2009.

19.

Fan

, Li

, Zhuo

, Zhang

, Wang

, Chen

, Yang

, Chu

, Xie

, Laird

, Fox

, Eickhoff

, Yu

, Jiang

: The human brainnetome atlas: A new brain atlas based on connectional architecture. Cereb Cortex, 26:3508–3526, 2016.

20.

Fonseca Pachi

, Torralbo

, Yamamoto

: Alzheimer's Disease and the Volume of Hippocampal Subfields: Cluster Analysis Using Neuroimaging Data. Cham, Switzerland: Springer International Publishing, 2019, pp. 296–306.

21.

Friston

, Williams

, Howard

, Frackowiak

RSJ

, Turner

: Movement-related effects in fMRI time-series. Magn Reson Med, 35:346–355, 1996.

22.

, Taber-Thomas

, Perez-Edgar

: Frontolimbic functioning during threat-related attention: Relations to early behavioral inhibition and anxiety in children. Biol Psychol, 122:98–109, 2017.

23.

Gold

, Steuber

, White

, Pacheco

, Sachs

, Pagliaccio

, Berman

, Leibenluft

, Pine

: Cortical thickness and subcortical gray matter volume in pediatric anxiety disorders. Neuropsychopharmacology, 42:2423–2433, 2017.

24.

Grabb

, Cross

, Potter

, McCracken

: Derisking psychiatric drug development: The NIMH's fast fail program, a novel precompetitive model. J Clin Psychopharmacol, 36:419–421, 2016.

25.

Guy

: The clinical global impression scale. In: ECDEU Assessment Manual for Psychopharmacology, Revised. Rockville, MD: U.S. Department of Health, Education & Welfare, 1976.

26.

Guyer

, Masten

, Pine

: Pediatric anxiety disorders: A clinical guide. In: Pediatric Anxiety Disorders: A Clinical Guide. Edited by Vasa

, Roy AK. New York: Springer Science + Business Media, 2013, pp. 23–46.

27.

Hardee

, Benson

, Bar-Haim

, Mogg

, Bradley

, Chen

, Britton

, Ernst

, Fox

, Pine

, Perez-Edgar

: Patterns of neural connectivity during an attention bias task moderate associations between early childhood temperament and internalizing symptoms in young adulthood. Biol Psychiatry, 74:273–279, 2013.

28.

Harrison

, Pujol

, Contreras-Rodriguez

, Soriano-Mas

, Lopez-Sola

, Deus

, Ortiz

, Blanco-Hinojo

, Alonso

, Hernandez-Ribas

, Cardoner

, Menchon

: Task-induced deactivation from rest extends beyond the default mode brain network. PLoS One, 6:e22964, 2011.

29.

Hutchison

, Womelsdorf

, Allen

, Bandettini

, Calhoun

, Corbetta

, Penna

, Duyn

, Glover

, Gonzalez-Castillo

, Handwerker

, Keilholz

, Kiviniemi

, Leopold

, Pasquale

, Sporns

, Walter

, Chang

: Dynamic functional connectivity: Promise, issues, and interpretations. Neuroimage, 80:360–378, 2013.

30.

Javitt

, Carter

, Krystal

, Kantrowitz

, Girgis

, Kegeles

, Ragland

, Maddock

, Lesh

, Tanase

, Corlett

, Rothman

, Mason

, Qiu

, Robinson

, Potter

, Carlson

, Wall

, Choo

, Grinband

, Lieberman

: Utility of imaging-based biomarkers for glutamate-targeted drug development in psychotic disorders: A randomized clinical trial. JAMA Psychiatry, 75:11–19, 2018.

31.

Jones

, Vemuri

, Murphy

, Gunter

, Senjem

, Machulda

, Przybelski

, Gregg

, Kantarci

, Knopman

, Boeve

, Petersen

, Jack

: Non-stationarity in the “resting brain's”. modular architecture. PLoS One, 7:e39731, 2012.

32.

Krystal

, Pizzagalli

, Mathew

, Sanacora

, Keefe

, Song

, Calabrese

, Goddard

, Goodman

, Lisanby

, Smoski

, Weiner

, Iosifescu

, Nurnberger

, Szabo

, Murrough

, Shekhar

, Potter

: The first implementation of the NIMH FAST-FAIL approach to psychiatric drug development. Nat Rev Drug Discov, 18:82–84, 2019.

33.

Kucyi

, Hove

, Esterman

, Hutchison

, Valera

: Dynamic brain network correlates of spontaneous fluctuations in attention. Cereb Cortex, 27:1831–1840, 2017.

34.

Kujawa

, Swain

, Hanna

, Koschmann

, Simpson

, Connolly

, Fitzgerald

, Monk

, Phan

: Prefrontal reactivity to social signals of threat as a predictor of treatment response in anxious youth. Neuropsychopharmacology, 41:1983–1990, 2016.

35.

, Biswal

, Wang

, Duan

, Cui

, Chen

, Liao

: Exploring the functional connectome in white matter. Hum Brain Mapp, 40:4331–4344, 2019a.

36.

, Duan

, Cui

, Chen

, Liao

: More than just statics: Temporal dynamics of intrinsic brain activity predicts the suicidal ideation in depressed patients. Psychol Med, 49:852–860, 2019b.

37.

, Mai

, Liu

: The default mode network and social understanding of others: What do brain connectivity studies tell us. Front Hum Neurosci, 8:74, 2014.

38.

Liao

, Li

, Duan

, Cui

, Chen

: Static and dynamic connectomics differentiate between depressed patients with and without suicidal ideation. Hum Brain Mapp, 39:4105–4118, 2018.

39.

Liao

, Wu

, Xu

, Ji

, Zhang

, Zang

, Lu

: DynamicBC: A MATLAB toolbox for dynamic brain connectome analysis. Brain Connect, 4:780–790, 2014.

40.

Mayberg

, Silva

, Brannan

, Tekell

, Mahurin

, McGinnis

, Jerabek

: The functional neuroanatomy of the placebo effect. Am J Psychiatry, 159:728–737, 2002.

41.

Monk

, Nelson

, McClure

, Mogg

, Bradley

, Leibenluft

, Blair

RJR

, Chen

, Charney

, Ernst

, Pine

: Ventrolateral prefrontal cortex activation and attentional bias in response to angry faces in adolescents with generalized anxiety disorder. Am J Psychiatry, 163:1091–1097, 2006.

42.

Monk

, Telzer

, Mogg

, Bradley

, Mai

, Louro

HMC

, Chen

, McClure-Tone

, Ernst

, Pine

: Amygdala and ventrolateral prefrontal cortex activation to masked angry faces in children and adolescents with generalized anxiety disorder. Arch Gen Psychiatry, 65:568–576, 2008.

43.

Mueller

, Aouidad

, Gorodetsky

, Goldman

, Pine

, Ernst

: Gray matter volume in adolescent anxiety: an impact of the brain-derived neurotrophic factor Val(66)Met polymorphism?. J Am Acad Child Adolesc Psychiatry, 52:184–195, 2013.

44.

Pecina

, Bohnert

ASB

, Sikora

, Avery

, Langenecker

, Mickey

, Zubieta

: Association between placebo-activated neural systems and antidepressant responses neurochemistry of placebo effects in major depression. JAMA Psychiatry, 72:1087–1094, 2015.

45.

Powers

, Whitley

, Telch

, Smits

JAJ

, Bystritsky

: The effect of attributional processes concerning medication taking on return of fear. J Consult Clin Psychol, 76:478–490, 2008.

46.

Preti

, Van de Ville

: Dynamics of functional connectivity at high spatial resolution reveal long-range interactions and fine-scale organization. Sci Rep, 7:12773, 2017.

47.

Qiu

, Liao

, Ding

, Feng

, Zhu

, Nie

, Zhang

, Chen

, Gong

: Regional homogeneity changes in social anxiety disorder: A resting-state fMRI study. Psychiatry Res, 194:47–53, 2011.

48.

Riddle

, Ginsburg

, Walkup

, Labelarte

, Pine

, Davies

, Greenhill

, Sweeney

, Klein

, Abikoff

, Hack

, Klee

, Bergman

, March

, Compton

, Robinson

, O'Hara

, Baker

, Ritz

, Roper

, Anxiety

: The Pediatric Anxiety Rating Scale (PARS): Development and psychometric properties. J Am Acad Child Adolesc Psychiatry, 41:1061–1069, 2002.

49.

Rogers

, Sylvester

, Mintz

, Kenley

, Shimony

, Barch

, Smyser

: Neonatal amygdala functional connectivity at rest in healthy and preterm infants and early internalizing symptoms. J Am Acad Child Adolesc Psychiatry, 56:157–166, 2017.

50.

Roy

, Fudge

, Kelly

, Perry

JSA

, Daniele

, Carlisi

, Benson

, Castellanos

, Milham

, Pine

, Ernst

: Intrinsic functional connectivity of amygdala-based networks in adolescent generalized anxiety disorder. J Am Acad Child Adolesc Psychiatry, 52:290–299, 2013.

51.

Royall

, Palmer

, Vidoni

, Honea

: The default mode network may be the key substrate of depressive symptom-related cognitive changes. J Alzheimers Dis, 34:547–560, 2013.

52.

Shirer

, Ryali

, Rykhlevskaia

, Menon

, Greicius

: Decoding subject-driven cognitive states with whole-brain connectivity patterns. Cereb Cortex, 22:158–165, 2012.

53.

Silverman

, Albano

: Anxiety Disorders Interview Schedule for DSM-IV-Child Version: Clinician Manual. San Antonio, TX: Psychological Corporation, 1996.

54.

Spies

, Kraus

, Geissberger

, Auer

, Klobl

, Tik

, Sturkat

, Hahn

, Woletz

, Pfabigan

, Kasper

, Lamm

, Windischberger

, Lanzenberger

: Default mode network deactivation during emotion processing predicts early antidepressant response. Transl Psychiatry, 7:e1008, 2017.

55.

Strawn

, Bitter

, Weber

, Chu

, Whitsel

, Adler

, Cerullo

, Eliassen

, Strakowski

, DelBello

: Neurocircuitry of generalized anxiety disorder in adolescents: A pilot functional neuroimaging and functional connectivity study. Depress Anxiety, 29:939–947, 2012.

56.

Strawn

, Compton

, Robertson

, Albano

, Hamdani

, Rynn

: Extended release guanfacine in pediatric anxiety disorders: A pilot, randomized, placebo-controlled trial. J Child Adolesc Psychopharmacol, 27:29–37, 2017a.

57.

Strawn

, Dobson

, Mills

, Cornwall

, Sakolsky

, Birmaher

, Compton

, Piacentini

, McCracken

, Ginsburg

, Kendall

, Walkup

, Albano

, Rynn

: Placebo response in pediatric anxiety disorders: Results from the child/adolescent anxiety multimodal study. J Child Adolesc Psychopharmacol, 27:501–508, 2017b.

58.

Strawn

, Dominick

, Patino

, Doyle

, Picard

, Phan

: Neurobiology of pediatric anxiety disorders. Curr Behav Neurosci Rep, 1:154–160, 2014a.

59.

Strawn

, Hamm

, Fitzgerald

, Monk

, Phan

: Neurostructural abnormalities in pediatric anxiety disorders. J Anxiety Disord, 32:81–88, 2015.

60.

Strawn

, Lu

, Peris

, Levine

, JTW: Pediatric anxiety disorders: What have we learnt in the last 10 years?. J Child Psychol Psychiatry [Epub ahead of print]; DOI: 10.1111/jcpp.13262.

61.

Strawn

, Wegman

, Dominick

, Swartz

, Wehry

, Patino

, Strakowski

, Adler

, Eliassen

, DelBello

: Cortical surface anatomy in pediatric patients with generalized anxiety disorder. J Anxiety Disord, 28:717–723, 2014b.

62.

Strawn

, Wehry

, Chu

, Adler

, Eliassen

, Cerullo

, Strakowski

, DelBello

: Neuroanatomic abnormalities in adolescents with generalized anxiety disorder: A voxel-based morphometry study. Depress Anxiety, 30:842–848, 2013.

63.

Sylvester

, Corbetta

, Raichle

, Rodebaugh

, Schlaggar

, Sheline

, Zorumski

, Lenz

: Functional network dysfunction in anxiety and anxiety disorders. Trends Neurosci, 35:527–535, 2012.

64.

Sylvester

, Smyser

, Kenley

, Ackerman

, Shimony

, Petersen

, Rogers

: Cortical functional connectivity evident after birth and behavioral inhibition at age 2. Am J Psychiatry, 175:180–187, 2018.

65.

Tian

, Wei

, Du

, Wang

, Yang

, Liu

, Meng

, Liu

, Qiu

: Assessment of trait anxiety and prediction of changes in state anxiety using functional brain imaging: A test–retest study. Neuroimage, 133:408–416, 2016.

66.

Walkup

: Antidepressant efficacy for depression in children and adolescents: Industry- and NIMH-funded studies. Am J Psychiatry, 174:430–437, 2017.

67.

Walkup

, Albano

, Piacentini

, Birmaher

, Compton

, Sherrill

, Ginsburg

, Rynn

, McCracken

, Waslick

, Iyengar

, March

, Kendall

: Cognitive behavioral therapy, sertraline, or a combination in childhood anxiety. N Engl J Med, 359:2753–2766, 2008.

68.

Wang

, Bernanke

, Peterson

, McGrath

, Stewart

, Chen

, Lee

, Wall

, Bastidas

, Hong

, Rutherford

, Hellerstein

, Posner

: The association between antidepressant treatment and brain connectivity in two double-blind, placebo-controlled clinical trials: A treatment mechanism study. Lancet Psychiatry, 6:667–674, 2019.

69.

Weaver

, Chaovalitwongse

, Novotny

, Poliakov

, Grabowski

, Ojemann

: Local functional connectivity as a pre-surgical tool for seizure focus identification in non-lesion, focal epilepsy. Front Neurol, 4:43, 2013.

70.

Williams

: Precision psychiatry: A neural circuit taxonomy for depression and anxiety. Lancet Psychiatry, 3:472–480, 2016.

71.

Yan

, Wang

, Zuo

, Zang

: DPABI: Data processing and analysis for (resting-state) brain imaging. Neuroinformatics, 14:339–351, 2016.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.12 MB

0.02 MB