Even though there were no differences in our predefined ROIS of left and right DLPFC when computing the contrast UG-DG, other regions of bilateral DLPFC were still preferentially engaged (Table S2), thus replicating previous findings, at least in the adult sample (Spitzer et al., 2007). In addition, we analyzed cortical thickness as a measure of brain structure in each individual (see Experimental Procedures for details). Performing AZD5363 a whole-brain assessment of cortical thickness in children, we observed

widespread thinning with increased age in bilateral prefrontal, cingulate, supramarginal, paracentral, and medial occipital regions (family-wise error [FWE] < 0.05, Figure S3). BMS387032 Although there was a small negative relationship between age and cortical thickness in our ROIs, effects failed to reach significance (p > 0.3 in both lDLPFC and rDLPFC; Figures 3A and 3D). Given that studies on structural brain development typically include samples of a greater age range (Gogtay et al., 2004 and Sowell et al., 2003), we also looked at age-related cortical thinning over the

entire range of children and adults in our two ROIs. Indeed, this revealed significant thinning in both lDLPFC (r = −0.385, p = 0.014; ρ = −0.412, p = 0.008;) and rDLPFC (r = −0.428, p = 0.006; ρ = −0.322, p = 0.043; Figure S4), confirming previous results (Gogtay et al., 2004, Sowell et al., 2003 and Sowell et al., 2004). We also assessed whether cortical thickness predicts individual differences in strategic behavior and impulse control, irrespective of any age-related cortical thinning. After statistically controlling for age effects, we found that thickness in lDLPFC correlated positively with both strategic behavior (r = 0.528, p = 0.007; Figure 3B) and negatively with SSRT scores (r = −0.630, p = 0.001; Figure 3C). Considering

age-corrected cortical thickness Sodium butyrate of rDLPFC, on the other hand, we neither observed correlations with strategic behavior (r = 0.347, p = 0.089; Figure 3D) nor with SSRT scores (r = −0.049, p = 0.816; Figure 3E). This latter finding suggests that greater thickness of lDLPFC is related to both increased strategic behavior and impulse control, irrespective of age. In the sample of adults, analysis of the cortical thickness revealed no correlation with age in either lDLPFC or rDLPFC (p > 0.3). Interestingly, like in the sample of children, analysis of an age-corrected relationship between cortical thickness and individual differences in strategic behavior in the sample of adults revealed a significant positive correlation in lDLPFC (r = 0.663, p = 0.014; Figures 4B) but not in rDLPFC (r = 0.159; p = 0.587; Figure 4D). These data provide a striking convergence with the age-corrected cortical thickness in the children, showing that greater thickness in lDLPFC is linked to increased strategic behavior.