Table 1.
Functional neuroimaging studies of youth antisocial behavior
| Study | Sample (groups contrasted; recruitment; age; gender; comorbidity and exclusions) | Key Methods (fMRI task and stimuli; measures of AB) | Contrast | Key Findings; coordinates (x, y, z)1[LWH1] |
|---|---|---|---|---|
| Emotion focused designs | ||||
| White, Marsh, et al., 2013 | 15 DBD+PT vs. 17 controls; community; age 10 – 17; mostly males; excluded substance use, anxiety and mood disorders; 8 DBD youth also had ADHD when excluded findings weakened but were in same direction |
“Bars task” – fear and neutral faces flanked by bars. Participants rate if bars are parallel. Amount of deviation from parallel decreases attention load; K-SADS + PT assessed via total APSD scores. |
Low load > high load | Control > DBD+PT : L amygdala/lentiform nucleus (−20, −14, −3) (DBD group shows reverse: > response to fear under high versus low load) |
| Fear > neutral | Control > DBD+PT: L middle temporal gyrus (BA 21) (−53, −3, −13 & −58, −8, −11) CU traits correlated with amygdala response to fearful expressions under low attentional load Main effect of group (control > DBD+PT: R middle frontal gyrus, bilateral superior frontal gyrus, L inferior frontal cortex, R posterior cingulate cortex, L inferior temporal cortex, L declive, L lingual gyrus[LWH2] |
|||
| Viding et al., 2012 | 15 CP+CU+ vs. 15 CP+CU− vs. 16 controls; community; age 10 – 16; males; controlling for ADHD, anxiety, and depression did not alter findings | Backward masked pictures of facial affect. Calm or fear were presented preattentively followed by calm faces; CASI-4R + CU traits assessed via the ICU (median split). |
Masked fear > masked calm | CP+CU− > control > CP+CU+ : R amygdala (−20, −3, −18); CP+CU− > control, CP+CU+ significant post-hoc; control and CP+CU+ not significant post-hoc; CU correlated negatively with 1 voxel in R amygdala (−24, −23, −15). |
| White, Williams, et al., 2012 | 17 DBD+ PT vs. 19 controls; community; age 10 – 17 ; mostly males; no substance use, anxiety or mood disorders; 9 DBD youth also had ADHD; when excluded some findings remained significant |
Eye gaze task with probes towards or away from the eyes using neutral, angry, and fearful faces; K-SADS+PT assessed via total APSD scores. |
Fear > Neutral during incongruent trials > congruent | Controls > DBD+PT : bilateral superior parietal lobule (BA 7) (23, −58, 44; −20, −60, 44), bilateral inferior parietal lobule (BA 40) (29, 41, 44; −44, −35, 40), R posterior cingulate cortex (BA 31) (17, −32, 40), L cuneus (BA 18) (−5, −80, 21). |
| Neutral > anger | DBD+PT > Controls: L superior frontal cortex (BA 9) (−29, 50, 28); R middle frontal cortex (BA6) (32, −5, 52). | |||
| Incongruent > congruent trials | Controls > DBD+PT: R middle temporal cortex (38, −11, −5); R thalamus (14, −16, 4). | |||
| Kalnin et al., 2011 | 22 CD or ODD vs. 22 controls (50% media violence exposed); community and clinic fliers; age 13 – 17; males and females; comorbidity not reported | Emotional stroop task; K-SADS |
“emotional” (violent) words > non-emotional words | AB youth who were also high on media exposure to violence have less activation in the R rostral ACC (13, 37, −8), R amygdala (22, 0, −21), and R posterior superior frontal gyrus (18, −1, 48). |
| Sebastian et al., 2012 | 31 CP+ vs. 16 controls; community; age 10 – 16; males; controlling for ADHD, GAD, MDD, and alcohol use disorder symptoms weakened but did not change the direction of the findings; overlapping subjects with Viding et al., 2012 | Cartoons asking participants to use cognitive or affective ToM or physical causality; CASI-4R + CU traits assessed via the ICU. |
Affective ToM > Cognitive ToM | Controls > CP: R amygdala (driven by greater response to affective – cognitive ToM in control group; no differences in CP group) (24, −12, −8) & R anterior insula (32, 16, 8) Supressor effects in R amygdala showing divergent relations with CU traits (− partial correlation) and CPs (+ partial correlation) |
| Passamonti et al., 2010 | 27 EO CD vs. 25 AO CD vs. 23 healthy controls; schools, clinics, youth offending services vs. schools and colleges; age 16 – 21; males; ADHD and MDD past and present in some but not all participants; controlling for ADHD did not alter findings; Both CD groups higher on CU traits than controls |
Gender categorization using pictures of facial affect with additional fixation cross trials a baseline measure; K-SADS + CU traits assessed via the YPI; No correlations found between CU traits and any fMRI activations. |
Anger > neutral | Both CD groups < control: L & R amygdala (−24, −4, −18; 26, −4, −22), vmPFC (10, 30, −17), L insula (−26, 19, −6), L & R OFC (46, 42, −9; 42, 44, −12). |
| Sad > neutral | Both CD groups < control: L & R amygdala (−20, −6, −13; 22, −6, −11), vmPFC (0, 26, −20). EO-CD < AO-CD: L & R amygdala (−20, −7, −16; 18, −6, −11), R anterior superior temporal sulcus (53, −11, −16). | |||
| Anger > fixation | Results in L & R amygdala & L insula driven by increased response to neutral faces in CD group, other areas driven by differences in neutral and anger faces | |||
| Sad > fixation | Results driven by differences in both neutral and anger faces | |||
| Marsh et al., 2008 | 12 CU + CD or ODD vs. 12 ADHD vs.12 control; community; age 10 – 17; males and females; excluded mood or anxiety disorders. | Gender categorization using pictures of facial affect; K-SADS + CU traits via > 20 score on the APSD and PCL− YV. |
Fearful > neutral | C/U < ADHD, C: R Amygdala (20, −7, −26); Less functional connectivity in CU group between amygdala and vmPFC (12, 35, −25). |
| Angry > neutral | Only the ADHD group differed from controls and the CU group | |||
| Jones et al., 2008 | 17 CU+ CP vs. 13 age and IQ matched; part or larger community study of twins; age 10 – 12; males; findings consistent when controlling for hyperactivity symptoms. | Gender categorization using pictures of facial affect; Combination of SDQ and APSD scores. |
Fearful > neutral | Control > CU+/CP+: R amygdala (30, 2, −25) |
| Herpertz et al., 2008 | 22 early onset CD vs. controls (with follow up comparison 13 “pure” ADHD to controls); clinic vs. community; age 12 – 17; males; all of CD group was diagnosed with ADHD. | Passive viewing IAPS pictures (negative, positive, & neutral); K-SADS. |
Negative > neutral | CD > control - L Amygdala (−30, −7, −25); CD group found IAPS pictures less arousing Present when controlling for IQ, anxiety/depression. Measures of internalizing symptoms correlate with amygdala reactivity in the CD group. |
|
Sterzer et al., 2005 Also see: Stadler et al., 2007 |
13 CD vs. 14 age matched controls; inpatients admitted for abnormal aggressive behavior vs. community; age 9 – 15; males; 62% comorbid for ADHD and high levels of anxiety/depression noted; some CD group on mediction. | Passive viewing IAPS pictures (negative & neutral); DSPDCA. |
Negative > neutral | Control > CD – R Dorsal ACC (9, 36, 33); ACC activation correlated with aggressive behavior on CBCL; CD group found IAPS pictures less arousing; Control > CD when controlling for anxiety/depression symptoms – L Amygdala (−21, −12, −18); |
| Studies focused on attention and inhibition | ||||
| Rubia, Halari, et al., 2009 | 13 pure early onset CD (+ODD) vs. 20 pure ADHD vs. 20 control (all medication naïve); clinic, advertisements; parent support groups; age 9 – 16; males; all comorbid for ODD; Same sample as Rubia et al., 2008. | Simon task (CD group made most overall errors, ADHD group was most variable); Maudsley diagnostic interview + SDQ. |
Incongruent vs. oddball trails (interference inhibition) | C > ADHD, CD – R superior/middle temporal lobe (BA 21, 42, 22) (54, −7, −13), R Precuneus (BA 7/19/31) (14, −66, 36). |
| Oddball vs. congruent trials (attention) | C, CD > ADHD – L inferior/middle PFC (BA 45/47/46/10) (−32, 52, −2); C > ADHD, CD – R DLPFC (BA 8/9/6) (28, 33, 48). | |||
| Rubia, Smith, et al., 2009 | 14 pure early onset CD (+ODD) vs. 18 pure ADHD vs. 16 control (all medication naïve); clinic, advertisements; parent support groups; age 9 – 16; males; all comorbid for ODD; Same sample as Rubia et al., 2008. | Rewarded continuous performance task; Maudsley diagnostic interview + SDQ. |
Non-rewarded target trials versus non-target trials (Sustained attention) | C, CD > ADHD – L & R vlPFC (BA47/45/44/10 & BA 47, 45, 11) (−47, 44, −13; 43, 41, −13); C, ADHD > CD - R insula (& premotor cortex; BA6) (36, −15, −13; 26, −15, −13), R hippocampus (32, 19, −7), L postgenual/dorsal anterior cingulate (BA[LWH3] 32) (−4, 19, 42). |
| Rewarded versus nonrewarded target trials (effect of reward) | C, ADHD> CD - R lateral and medial OFC (BA 47) (32, 37, −2); C, CD > ADHD - L precuneueus and posterior cingulate gyrus (BA 29) (−4, −52, 15). | |||
| Rubia, et al, 2008 | 13 pure early onset CD (+ODD) vs. 20 pure ADHD vs. 20 control (all medication naïve); clinic, advertisements; parent support groups; age 9 – 16; males; all comorbid for ODD; | Visual tracking stop task; Maudsley diagnostic interview + SDQ. |
Successful trial versus failed stop | C, CD > ADHD – L middle/inferior frontal gyrus (BA 46/10) (−25, 59, 15). |
| Failed trial versus go | C > CD, ADHD – R posterior cingulate/precuneus (BA 29/31/7) (11, −48, 26); C, ADHD > CD – L inferior parietal (BA 40) (−40, −33, 37), R posterior/superior temporal/inferior parietal lobe (BA 3/4/42/40) (58, −19, 20). | |||
| Studies focused on learning, reward and other | ||||
|
Sharp et al., 2011 See also White et al., 2013 |
10 externalizing vs. 10 controls; community (recruited from the boy scouts); age 11 – 16; males; no information on comorbidity. | “Trust game” involving interaction with a virtual peer in exchange of money; Combined parent, peer, and self-report of externalizing on the CBCL. |
Share decisions > keep decisions | Externalizing > non-externalizing in bilateral insula (44, −8, 4; −29, 25, 9 & −39, 17, 4). |
| Outcome phase > fixation x neutral > kind/mean peer | Non-externalizing > externalizing in bilateral caudate (7, 4, 12); Non-externalizing showed greater response to neutral (vs. kind/mean) peer; externalizing group shows no moderation in bilateral insula (54, 13, −12; −51, 19, −11). | |||
| Finger et al., 2011 | 15 DBD vs. 15 controls; community fliers & referrals; mean age 13–14; males and females; 67% of CU/CD group comorbid for ADHD. | Passive avoidance task; K-SADS + CU traits assessed via APSD & PCL-YV. |
Early > late trials | CU/CD group have less activation in “network of attention” including R OFC (23, 36, −13) and L caudate (−2, 16, 13) among others. |
| Rewarded correct hits > punished commission errors | CU/CD < controls: R OFC (5, 64, −8), L middle frontal gyrus (−44, 12, 49) and parahippocampal gyrus (35, −41, −10); Amygdala (20, −10, −26), insula (−41, 14, 0), fusiform gyrus (32, −81, −20) and caudate (−8, 7, 16)(& others) responsiveness lower throughout entire task for those in the CU/CD group. | |||
| Crowley et al., 2010 | 20 abstinent ASD vs. 20 controls; boys in current substance treatment vs. controls from community; age 14 – 18; males; All subjects diagnosed with CD and substance abuse or dependence. | Colorado Balloon Game (a decision and reward task); DISC, CBCL, CIDI-SAM. |
Decision making > direction following | ASD < control for many areas including OFC, DLPFC, ACC, basal ganglia, insula, amygdala, & hippocampus (coordinates not provided as many regions were overlapping and part of same large cluster). |
| Wins > Loss | ASD < control in ACC, temporal regions, & cerebellum. | |||
| Loss > Win | ASD > control in OFC, DLPFC, brain stem & cerebellum. | |||
| Decety et al., 2009 | 8 childhood onset CD vs. 8 age, sex, & race matched control; part of larger community study; age 16 – 18; males and females; 88% of CD group comorbid for ADHD, | Animations of Painful situations: cause by accident, on purpose, help, non-painful; DISC + CADS. |
Pain by accident > non-painful Pain on purpose vs. accident |
CD group differential activation of “pain matrix” with greater activation in: L Amygdala (−18, −8, −8), Bilateral temporal pole (62, −2, −5; −50, −4, 0), ACC (2, 10, 23), R Caudate (12, 18, 3). CD > C- L anterior insula (−26, −23, −2), R SMA (12, −7, 56), L Precentral gyrus (−30, 18, 49); C > CD - L DLPFC (−42, 42, 18), R superior frontal gyrus (18, 64, 0). CD youth have less functional connectivity between amygdala and prefrontal areas while watching pain inflicted on others. Amygdala activity correlated with number of aggressive acts and ratings of daring. |
| Gatzke-Kopp et al., 2009 | 19 CD + ADHD boys vs. 11 controls; community fliers; age 12 – 16; males; almost all boys had ADHD and most had CD in “EXT” group. | Reward task (rewarded for reporting which side a light was on); DISC. |
Rewarded vs. non-rewarded blocks | CD/ADHD group continued to activate the caudate during non-reward blocks while controls shifted to bilateral dorsal ACC (14, 36, 24) activation; (CD/ADHD < Controls ACC during non-reward). |
| Finger et al., 2008 | 14 CU +CD ODD, 14 ADHD, 14 control; community fliers & referrals; age 10 – 17; males; excluded mood and anxiety disorders. | Probabilistic learning reversal task; K-SADS + CU traits assessed via the APSD. |
Punished reversal errors vs. all correct responses | C/U > C, ADHD – Bilateral medial frontal gyrus (BA 10) (−17, 47, 5; 23, 46, 9); C/U > C – caudate (20, −23, 26). |
1
All Coordinates are from the system of Talairach and Tournoux (Talairach & Tournoux, 1988). All coordinates that were reported in other coordinate systems (e.g., MNI) were converted to Talairach using the Wake Forest PickAtlas (http://fmri.wfubmc.edu/software/PickAtlas). Note that anatomical labels provided in this table are the ones used by the original authors regardless of the labels assigned by the Talairach daemon.
Abbreviations: ACC – Anterior Cingulate Cortex, AO-CD – Adolescent-Onset Conduct Disorder, APSD – Antisocial Process Screening Device, ASD – Antisocial Substance Disorder, BA – Brodmann’s Area; CASI-4R - Child and Adolescent Symptom Inventory – 4 R, CBCL – Child Behavior Checklist, CD – Conduct Disorder, CP – Conduct Problems, CIDI-SAM – Compsite International Diagnostic Interview – Substance Abuse Module, CU – Callous/Unemotional, DBD – Disruptive Behavior Disorder (CD and/pr ODD), DISC – Diagnostic Interview Schedule for Children, DPSPDCA – Diagnostic System for Psychiatric Disorders in Childhood and Adolescence, EO-CD – Early Onset Conduct Disorder, ICU – Inventory of Callous-Unemotional Traits; KSADS-PL – Schedule for Affective Disorders and Schizophrenia for School aged children; MNI – Coordinates based on the system of the Montreal Neurological Institute, PCL-YV – Psychopathy Checklist: Youth Version, PFC – Prefrontal Cortex, PT – Psychopathic Traits; SDQ – Strengths and Differences Questionnaire, SMA – Supplementary Motor Area; ToM – Theory of Mind; YPI – youth Psychopathic Traits Inventory.