Fig 6. Relationship between age and ventral striatum activation during total win > neutral trials in the MTwiNS sample.
k = number of voxels within the cluster. Significant clusters were identified in SPM12 using a mask of the ventral striatum [9], grey matter segmented using AFNI [36]. False positive rate is controlled across the ventral striatum using 3dClustSim for cluster-level correction (punc < .001, alpha < .05, k > 3). (a) Quadratic association between age and ventral striatum activation during total win > neutral trials (n = 464). Right: k = 7, T = 3.29, MNI 16, 8, -12. (b) Plot of the quadratic age effect with extracted ventral striatum activation (n = 464). VS = ventral striatum. To interpret the quadratic age effect, right VS activation was extracted from a mask of the significant cluster during total win > neutral trials and plotted by age in months with a quadratic curve. The association between age and VS activation followed a “U” shaped curve, with activation decreasing from early- to mid-adolescence and increasing from mid- to late adolescence. (c) Linear association between age and ventral striatum activation during total win > neutral trials in a subsample of MTwiNS youth ages 10 to 18 years old (n = 454). MTwiNS youth were excluded from analyses if they were younger than 10 (n = 9) or older than 18 years old (n = 1) to confirm that our results were not due to ages at the extreme ends of the distribution. Left: k = 6, T = 3.70, MNI -22, 12, -10.