Impulse Control Is Differentially Modulated By Social Versus Non-Social Stimuli

Background: Extensive research has shown that typically developing individuals show a bias in impulse control towards faces. In autism, prior work has also shown increased impulsivity towards happy faces. However, it is unknown whether other types of salient cues, such as hobbies or interests, elicit a similar bias.

Objectives: The aim of this study was to investigate impulse control and underlying neural mechanisms towards different types of salient stimuli in typical development to provide a foundation for future studies in autism. We expected participants would show a similar bias towards interests as they do towards social stimuli.

Methods: Forty typically developing adults (17 females) aged 23.4 (± 3.3) years performed a novel go-nogo task during fMRI. For the social conditions, participants were presented with happy and calm faces. For the interest conditions, participants chose their favorite (interest) and least favorite hobby/activity (non-interest) from 23 options. There were 5 runs of go-nogo pairs (blue vs. yellow, happy vs. calm, calm vs. happy, interests vs. non-interests, non-interests vs. interests). Participants were instructed to press to the target cue (go) when it appeared on screen and not press to the distractor cue (nogo). False-alarm rates were calculated as the number of erroneous button presses to the nogo cue in each of the five conditions.
D-prime was calculated as the normalized hit rate (go-accuracy) – normalized false alarm rate. We used ex-Gaussian distribution parameters to characterize fast (mu and sigma) and slow (tau) reaction times (RT’s).

Results: Go-accuracy did not differ between conditions, showing participants attended to the different conditions equally well. Linear mixed-effects model analysis showed main effects for false-alarm rate (p = .006), d-prime (p = .006), reaction times to go-stimuli (p < .001), and the ex-Gaussian parameters mu (p < .001), and sigma (p < .001). Participants showed highest d-prime, lowest false alarm rate, and fastest RT’s to colors. There were no differences between calm and happy faces on any of the measures however; d-prime was significantly higher for interests compared to non-interests (p = .042), suggesting increased control towards interests. To assess an overall bias for social stimuli, comparing the collapsed social stimuli (faces) against the collapsed objects (interests/non-interests), showed d-prime was lower (p = .024) for social stimuli. Ex-Gaussian parameter mu also showed slower reaction times (p< .001) towards social stimuli. As expected, dorsolateral prefrontal cortex (dlPFC) showed heightened activation during inhibition. We will further investigate how cognitive control networks are modulated by distinct salient stimuli.

Conclusions: As expected, typically developing adults were more impulsive towards social stimuli. Interestingly, the difference in d-prime between social and non-social stimuli was driven by higher d-prime towards interests, suggesting that, in contrast to our hypothesis, one’s interest facilitates inhibition. Together these findings show impulse control is differentially modulated by distinct salient stimuli, which may be important for individuals with autism, where stimuli associated with their interests may interfere during impulse control. Future work examining the neural underpinnings of impulsivity to salient cues will constrain interpretations of the behavioral data.