Gender differences in reward-related decision processing under stress

Abstract

Recent research indicates gender differences in the impact of stress on decision behavior, but little is known about the brain mechanisms involved in these gender-specific stress effects. The current study used functional magnetic resonance imaging (fMRI) to determine whether induced stress resulted in gender-specific patterns of brain activation during a decision task involving monetary reward. Specifically, we manipulated physiological stress levels using a cold pressor task, prior to a risky decision making task. Healthy men (n = 24, 12 stressed) and women (n = 23, 11 stressed) completed the decision task after either cold pressor stress or a control task during the period of cortisol response to the cold pressor. Gender differences in behavior were present in stressed participants but not controls, such that stress led to greater reward collection and faster decision speed in males but less reward collection and slower decision speed in females. A gender-by-stress interaction was observed for the dorsal striatum and anterior insula. With cold stress, activation in these regions was increased in males but decreased in females. The findings of this study indicate that the impact of stress on reward-related decision processing differs depending on gender.

Fig. 1

Two versions of the adapted BART included (A) an active task (PLAY GAME screen) in which participants decided how large to inflate a series of balloons in order to earn money which accumulated with each ‘pump’ while trying to avoid ‘explosions’ which resulted in a loss of earnings for the current balloon and a balloon popping sound effect. At any point, participants could decision to ‘cash out’ their earnings for the current balloon, which transferred current balloon earnings to the permanent bank and resulted in a cash register sound effect. In (B) the passive task (KEEP CLICKING screen) participants could not earn or lose money but were instructed to press the pump balloon button throughout the block. In the passive task, the ‘Cash out $$$’ button was disabled and balloons appeared at random sizes while cash out and explosion sounds occurred randomly. Fixation periods (10 s) appeared in between alternating active and passive task blocks (four blocks per condition).

Fig. 2

Cortisol levels increased significantly in the stress group (cold pressor) but not in the control group (warm water). Cortisol responses to stress were not dependent on gender. Poststress cortisol is the average of cortisol at 21 and 35 min after the start of the hand immersion task (s2, s3, respectively); immediately before and just after the decision task. Error bars represent s.e.m.

Fig. 3

Gender-by-stress effects on behavior and earnings in the BART. Group means for number of balloons completed per active block (A), decision speed (B) and total earnings in US dollars across all four active blocks (C). Error bars represent s.e.m.

Fig. 4

Whole-brain analysis revealed significant gender–stress interactions for the active vs passive contrast in the left putamen and left anterior insula (A) with participant as a random factor (Z > 2.3; FWE-corrected cluster significance threshold of P < 0.05) in the mixed-effects analysis. To ensure we extracted signal change values separately for the putamen and insula, we used structurally defined regions of the putamen and insula from FSL’s MNI structural atlas (B) to mask these significant clusters from the mixed-effects analysis (C). Mean percent signal change was greater in the stress condition for males and diminished in the stress condition in females in the left putamen (D) and left anterior insula (E).