Pupil Dilation Indexes But Does Not Causally Influence Conscious Error Detection: A Double-Blind Placebo-Controlled...

Conscious error detection is accompanied by error-related changes in phasic autonomic activity. This autonomic response is diminished in older age, accompanied by impairments in the conscious detection of action errors, i.e., increased error blindness. The degree to which the autonomic response to errors declines across the lifespan is correlated with the increase in error blindness. However, the direction of causality-whether changes in autonomic reactivity are a consequence or cause of increased error blindness-is still being debated. In the current study, we experimentally modulated the phasic autonomic response to action errors in healthy older adults while measuring their conscious error detection.

Across 2 sessions, 30 healthy older adults (ages 60-80 years) were given atomoxetine or placebo in a double-blind fashion. In each session, they performed an antisaccade task, which is commonly used to test conscious error detection. The autonomic response to errors was measured via changes in pupil dilation. A novelty-oddball task was also used as a manipulation check.

Atomoxetine reduced phasic pupil dilation to both novel stimuli in the novelty-oddball task and to action errors in the antisaccade task. However, despite this apparent blunting of the phasic autonomic response to errors, there were no significant differences in conscious error awareness between atomoxetine and placebo. Primary task performance was also unaffected.

Despite its apparent effects on phasic autonomic activity after errors, atomoxetine had no effect on conscious error detection in healthy older adults. This suggests that phasic autonomic activity is a consequence of, rather than a contributing factor to, conscious error awareness. It also suggests that changes to phasic autonomic activity are unlikely to explain increased error blindness in older age.