The neural correlates of emotional conflict monitoring as an early manifestation of affective and cognitive declines in persons with Type 2 diabetes

Figures

• 

  Figure 1. Behavioural performance on the emotional Stroop task between the diabetes and control groups. (A) A three-way ANOVA (group, congruency and emotion type) for accuracy yielded significant main effects of the factor group (F1,68 = 4.69, P = 0.034, pη2 = 0.065), congruency (F1,68 = 10.19, P = 0.002, pη2 = 0.13) and emotion type (F1,68 = 22.90, P < 0.001, pη2 = 0.252). While the accuracy was lower for fearful faces (fearful versus happy: 95.3% ± 0.8% versus 97.5% ± 0.6%) and for incongruent conditions (incongruent versus congruent: 94.5% ± 1.2% versus 98.3% ± 0.3%), persons with Type 2 diabetes were more error-prone than were the controls (diabetes versus controls: 95% ± 0.9% versus 97.8% ± 1%). (B) In terms of emotional interference (Δcon) and differences in the relative reaction times between incongruent trials and congruent trials, patients experienced significantly stronger interference effects (114 ± 14 versus 63 ± 11 ms, t = 2.781, P = 0.007).

• 

  Figure 2. Associations between emotional interference (Δcon) with the dorsal anterior cingulate cortex (dACC) and anterior insula (AI) activity. (A) The dACC activity was positively correlated with Δcon in the controls (r = 0.443, P = 0.014) but none in the diabetes (r = 0.266, P = 0.096). (B) The AI activity was positively correlated with Δcon in the controls (r = 0.364, P = 0.048) but none in the diabetes (r = 0.210, P = 0.193).

• 

  Figure 3. The functional connectivity between the dorsal anterior cingulate cortex (dACC) and amygdala (dACC–AMY). The psychophysiological interaction analysis showed a main effect of factor congruency (F1,68 = 5.91, P = 0.018, pη2 = 0.08) as participants exhibited negative functional coupling between the dACC and amygdala (x 26, y 4, z 16, k = 18; dACC–AMY) during incongruent trials [M(INCON) = −0.186 (0.059), M(CON) = 0.042(0.06), respectively, Bonferroni-corrected P < 0.05]. Pre hoc planned group-wise comparisons indicated that the dACC–AMY connectivity differed by congruency in diabetes but not in controls, with diabetes showing more-negative dACC–AMY connectivity in incongruent trials than in congruent trials [diabetes: M(INCON) = −0.241(0.069), M(CON) = 0.02(0.079), t = −2.061, P = 0.046; controls: M(INCON) = −0.13(0.099), M(CON) = 0.064(0.091), t = −1.439, P = 0.161].

• 

  Figure 4. The functional connectivity mediates glucose levels, behavioural interference, cognitive and affective declines. (A) The dorsal anterior cingulate cortex (dACC)–dorsolateral prefrontal cortex (DLPFC) functional connectivity and dACC–ventrolateral PFC (VLPFC) functional connectivity mediate the effects of behavioural interference (Δcon) on the prediction of Montreal Cognitive Assessment (MoCA) scores in individuals with diabetes. (B) The dACC–amygdala functional connectivity mediates the association between glycated haemoglobin (HbA1c) and the prediction of Beck Anxiety scores in individuals with diabetes. Functional connectivity was visualized by the lateral, top and back views of glass brain visualization.