Brain structural correlates of reward sensitivity and impulsivity in adolescents with normal and excess weight

Abstract

Introduction: Neuroscience evidence suggests that adolescent obesity is linked to brain dysfunctions associated with enhanced reward and somatosensory processing and reduced impulse control during food processing. Comparatively less is known about the role of more stable brain structural measures and their link to personality traits and neuropsychological factors on the presentation of adolescent obesity. Here we aimed to investigate regional brain anatomy in adolescents with excess weight vs. lean controls. We also aimed to contrast the associations between brain structure and personality and cognitive measures in both groups.

Methods: Fifty-two adolescents (16 with normal weight and 36 with excess weight) were scanned using magnetic resonance imaging and completed the Sensitivity to Punishment and Sensitivity to Reward Questionnaire (SPSRQ), the UPPS-P scale, and the Stroop task. Voxel-based morphometry (VBM) was used to assess possible between-group differences in regional gray matter (GM) and to measure the putative differences in the way reward and punishment sensitivity, impulsivity and inhibitory control relate to regional GM volumes, which were analyzed using both region of interest (ROI) and whole brain analyses. The ROIs included areas involved in reward/somatosensory processing (striatum, somatosensory cortices) and motivation/impulse control (hippocampus, prefrontal cortex).

Results: Excess weight adolescents showed increased GM volume in the right hippocampus. Voxel-wise volumes of the second somatosensory cortex (SII) were correlated with reward sensitivity and positive urgency in lean controls, but this association was missed in excess weight adolescents. Moreover, Stroop performance correlated with dorsolateral prefrontal cortex volumes in controls but not in excess weight adolescents.

Conclusion: Adolescents with excess weight have structural abnormalities in brain regions associated with somatosensory processing and motivation.

Figure 1. Clusters of significant gray matter volume increase in excess weight compared with normal weight subjects.

Peak coordinates were located in the right hippocampus (x, y, z_ 38, −13, −18; t = 4.21; pFWE-SVC<0.05). Results are overlaid on coronal and sagittal sections of a normalized brain, and the numbers correspond to the ‘y’ and ‘x’ coordinates in MNI space. Color bar represents t value. For demonstration purposes the images are displayed at p<0.001 (uncorrected, k>50).

Figure 2. Between-group interaction between regional gray matter volume and reward sensitivity.

A. Voxel-wise correlations between regional gray matter volume and reward sensitivity score specifically observed in normal weight subjects. Peak coordinate was located in the left secondary somatosensory cortex (SII, Brodmann area 43) (x, y, z = −60, −7, 11; t = 4.51; pFWE-SVC<0.05). Results are overlaid on coronal (left) and axial (right) sections of a normalized brain, and the numbers correspond to the ‘y’ and ‘z’ coordinates in MNI space, respectively. Color bar represents t value. For demonstration purposes the images are displayed at p<0.001 (uncorrected, k>100). B. Plot of the correlation between gray matter volume at the peak coordinate and the reward sensitivity score. Normal weight group (filled circles, solid line) showed a significant correlation between these two measures (r = −0.750; p<0.005), while in the excess weight group the correlation was not significant (r = 0.284; p>0.05).

Figure 3. Between-group interaction between regional gray matter volume and positive urgency.

A. Voxel-wise correlations between regional gray matter volume and positive urgency (UPPS-P) score specifically observed in normal weight subjects. Peak coordinate was located in the left secondary somatosensory cortex (SII, Brodmann area 43) (x, y, z = −63, −7, 15; t = 4.89; pFWE-SVC<0.05). Results are overlaid on coronal (left) and axial (right) sections of a normalized brain, and the numbers correspond to the ‘y’ and ‘z’ coordinates in MNI space, respectively. Color bar represents t value. For demonstration purposes the images are displayed at p<0.001 (uncorrected, k>100). B. Plot of the correlation between gray matter volume at the peak coordinate and the positive urgency score. Normal weight group (filled circles, solid line) showed a significant correlation between these two measures (r = −0.856; p<0.0005), while in the excess weight group the correlation was not significant (r = 0.058; p>0.05).

Figure 4. Between-group interaction between regional gray matter volume and response inhibition.

A. Voxel-wise correlations between regional gray matter volume and the Stroop response inhibition score specifically observed in normal weight subjects. Peak coordinate was located in the left dorsolateral prefrontal cortex (Brodmann area 9) (x, y, z = −61, 6, 24; t = 5.01; pFWE-SVC<0.05). Results are overlaid on coronal (left) and axial (right) sections of a normalized brain, and the numbers correspond to the ‘y’ and ‘z’ coordinates in MNI space, respectively. Color bar represents t value. For demonstration purposes the images are displayed at p<0.001 (uncorrected, k>100). B. Plot of the correlation between gray matter volume at the peak coordinate and the Stroop response inhibition score. Normal weight group (filled circles, solid line) showed a significant correlation between these two measures (r = 0.769; p<0.005), while in the excess weight group the correlation was not significant (r = −0.327; p>0.05).