Testosterone administration in females modulates moral judgment and patterns of brain activation and functional connectivity

Abstract

Morality is defined as prescriptive norms regarding how people should treat one another, and includes concepts of fairness, justice, and rights. One recent study with moral dilemmas suggested that testosterone administration increases utilitarian judgments, which depends on second-to-fourth (2D: 4D) digit ratio, as a proxy of prenatal priming. However, the neural mechanism by which acute testosterone modulates moral reasoning remains to be determined. Using a placebo-controlled within-subject design, the current study examined the neuromodulatory effect of testosterone in young females by combining moral dilemmas, 2D: 4D, functional magnetic resonance imaging (fMRI), and subjective ratings of morally laden scenarios. Results showed that testosterone administration elicited more utilitarian responses to evitable dilemmas. The high 2D: 4D group scored more punishments for moral evaluation, whereas the low 2D: 4D group did the opposite. The activity in the amygdala, anterior insular cortex, and dorsolateral prefrontal cortex (dlPFC) was increased when participants evaluated morally unorthodox actions (intentional harm). The activity in the posterior superior temporal sulcus/temporoparietal junction (pSTS/TPJ) to accidental harm was decreased, specific to the high 2D: 4D group. The functional connectivity between the amygdala and dlPFC was reduced. The activity in the pSTS/TPJ to perceived agency predicted utilitarian responses to evitable dilemmas. The findings demonstrate the acute effect of testosterone on neural responses associated with moral judgment, and provide evidence to support that prenatal sex-hormones priming could be important for early neurodevelopment, which plays a crucial role in the neural and behavioral manifestations of testosterone on adult moral reasoning. Hum Brain Mapp 37:3417-3430, 2016. © 2016 Wiley Periodicals, Inc.

Keywords: 2D: 4D; fMRI; moral reasoning; testosterone; utilitarian judgments.

Figure 1

Experimental procedures. The design is placebo‐controlled, randomly assigned, double‐masked, and crossover. We sublingually administered either testosterone or placebo to participants on two separate days. Participant filled out the PANAS. Saliva was collected before and after administration. Hand scanning was employed to measure the 2D: 4D digit ratio. Participants read and responded to moral dilemmas at their own pace. During fMRI scanning, participants were presented with the stimuli depicting moral transgressions [agency (intentional vs. unintentional) and target (people vs. objects). After fMRI scanning, participants were requested to respond to questions probing moral reasoning.

Figure 2

Utilitarian judgments on moral dilemmas after testosterone administration. The y‐axis indicates proportions of 'yes' judgments. Error bar represents the 95% confidence intervals. There were four dilemma types: non‐moral, impersonal, personal‐inevitable, and personal‐evitable moral dilemmas. On personal‐evitable moral dilemmas, testosterone relative to placebo administration significantly increased the frequency of endorsing 'yes' responses (P = 0.002).

Figure 3

Response to perceived agency and target after testosterone administration. (a) To perceived agency (intentional vs. unintentional harm), the testosterone administration increased the activity in the amygdala (AMYG), periacqueductal gray (PAG), ventromedial prefrontal cortex (vmPFC), superior temporal gyrus (STG), caudate nucleus (CN), anterior cingulate cortex (ACC), and posterior cingulate cortex (PCC). Instead, the placebo was associated with the activity in the vmPFC only (thresholded at P < 0.01, uncorrected, k = 25 for visual purposes). (b) To perceived target (people vs. objects), the testosterone increased the activity in the anterior insular cortex (AIC) and posterior superior temporal sulcus/temporoparietal junction (pSTS/TPJ), whereas the placebo was associated with the activity in the amygdala (AMYG), ventromedial (vmPFC), dorsolateral (dlPFC), and ventrolateral prefrontal cortex (vlPFC) along with temporal pole (TP) and pSTS/TPJ (thresholded at P < 0.01, uncorrected, k = 25 for visual purposes).

Figure 4

Testosterone modulation on moral evaluations in the right pSTS/TPJ. (a) Neuro‐hemodynamic response to unintentional (accidental) harm to objects was reduced by testosterone administration, depending on the 2D: 4D. Testosterone relative to placebo administration decreased activity in the pSTS/TPJ in the high 2D: 4D group (testosterone vs. placebo: 3.1 ± 0.74 vs. 4.56 ± 0.8), but not in the low 2D: 4D group (4.61 ± 0.81 vs. 3.39 ± 0.55). Asteroid indicated P < 0.05, one‐tailed. (b) As a result of testosterone administration, the activity in the pSTS/TPJ in response to moral evaluations predicted moral utilitarian decisions. The endorsement of evitable dilemmas was positively correlated with the activity in the pSTS/TPJ to perceived agency. Abbreviations: TI, testosterone/intentional harm; TU, testosterone/unintentional harm; PI, placebo/intentional harm; PU, placebo/unintentional harm.

Figure 5

Functional connectivity after testosterone administration. Testosterone relative to placebo administration significantly decreased the functional connectivity of the amygdala with the rostral dorsolateral (dlPFC) and dorsomedial prefrontal cortex (dmPFC).