Skip to main content
. Author manuscript; available in PMC: 2018 Jun 1.
Published in final edited form as: Am J Psychiatry. 2017 Mar 21;174(6):566–575. doi: 10.1176/appi.ajp.2017.16080897

Table 3.

Beta weights for prediction of age 16 outcome

A

Predictors Mean Betas
Demographic information

Age 0.101
Sex (male) −0.079
Site 0.105
Handedness (right-handed) −0.182
Symptoms

Psychotic-like experiences at baseline 0.577
Brain ROIs

dlPFC during failed inhibition 0.021
Caudate during failed inhibition −0.088
Cerebellum during angry faces processing 0.276
Hippocampus/amygdala during neutral faces processing 0.253
Middle temporal during neutral faces processing 0.070
Cerebellum during neutral faces processing −0.234
Inferior frontal during neutral faces processing −0.126
Lingual gyrus during neutral faces processing −0.142
Fusiform gyrus during neutral faces processing −0.077
ACC/MCC during anticipation of reward 0.066
Fusiform gyrus during anticipation of reward −0.216
B

Predictors Mean Betas
Demographic and substance use information

Cannabis use in the previous year 0.137
Lifetime cigarette use 0.150
Symptoms

Internalizing behaviors 0.307
Externalizing behaviors 0.084
Brain ROIs

Cerebellum during angry faces processing 0.092
Hippocampus/amygdala during neutral faces processing 0.073
Cerebellum during neutral faces processing −0.090
C

Predictors Mean Betas
Demographic and substance use information

Age 0.027
Sex (male) 0.188
Site 0.023
Handedness (right-handed) −0.020
Puberty development score 0.077
Cannabis use in the previous year 0.250
Lifetime cigarette use 0.268
Alcohol use in the previous year −0.171
Symptoms

Internalizing behaviors 0.297
Externalizing behaviors 0.098
Brain ROIs

dlPFC during failed inhibition 0.100
Caudate during failed inhibition −0.142
Cerebellum during angry faces processing 0.317
Hippocampus/amygdala during neutral faces processing 0.192
Middle temporal during neutral faces processing 0.100
Cerebellum during neutral faces processing −0.066
Inferior frontal during neutral faces processing −0.101
Lingual gyrus during neutral faces processing −0.158
Fusiform gyrus during neutral faces processing −0.135
ACC/MCC during anticipation of reward −0.005
Fusiform gyrus during anticipation of reward −0.306
D

Predictors Mean Betas
Demographic information

Age −0.024
Sex (male) −0.053
Site −0.097
Handedness (right-handed) 0.041
Puberty development score 0.096
Cannabis use in the previous year 0.140
Lifetime cigarette use −0.057
Alcohol use in the previous year 0.058
Symptoms

Internalizing behaviors 0.154
Externalizing behaviors 0.100
Brain ROIs

Caudate during failed inhibition 0.062
Cerebellum during angry faces processing −0.083
Hippocampus/amygdala during neutral faces processing −0.029
Middle temporal during neutral faces processing −0.025
Cerebellum during neutral faces processing −0.132
Inferior frontal during neutral faces processing 0.036
Lingual gyrus during neutral faces processing 0.071
Fusiform gyrus during neutral faces processing 0.039
ACC/MCC during anticipation of reward 0.025
Fusiform gyrus during anticipation of reward 0.088

Abbreviations: dlPFC, dorsolateral prefrontal cortex; ACC/MCC, anterior/middle cingulate cortex.

A, Mean beta weights (averaged over 10 outer folds) of the features that were present in at least 9 folds (out of 10) of the final model for classification of the mood and psychotic symptoms group versus the no mood symptoms group in the London-Dublin subsample. B, Mean beta weights (averaged over 10 outer folds) of the features that were present in at least 9 folds (out of 10) of the final model for classification of the mood and psychotic symptoms group versus the no mood symptoms group in the full IMAGEN sample. C, Mean beta weights (averaged over 10 outer folds) of the features that were present in at least 9 folds (out of 10) of the final model for classification of the mood and psychotic symptoms group versus the mood symptoms only group in the full IMAGEN sample. D, Mean beta weights (averaged over 10 outer folds) of the features that were present in at least 9 folds (out of 10) of the final model for classification of the mood symptoms only group versus the no mood symptoms group in the full IMAGEN sample.