Being treated badly predicted rising drinking -- unless self-control was quick

Following 133 college students for three years, those reporting more social mistreatment showed steeper increases in alcohol problems. Faster response inhibition on a lab task blunted that link -- but the crucial three-way test was not significant (p = 0.10).

Being belittled, excluded or treated unfairly is a reliable predictor of heavy drinking, but not everyone responds that way, and researchers led by Chia-Chuan Yu asked whether the difference lies in self-regulation. They followed 133 college students for three years, measuring alcohol problems annually with the AUDIT and social mistreatment alongside it, and at baseline put everyone through a Stop-Signal Task in an MRI scanner. That task yields two measures: stop-signal reaction time, essentially how fast someone can cancel an action already underway, and post-error slowing, how much they pause after getting one wrong. Drinking problems rose across the three years, and the rise was steeper for students reporting more mistreatment, a significant interaction adding about 2.67% of explained variance. Inhibition mattered too: students with faster stop-signal times showed flatter drinking trajectories, and there was a significant interaction between inhibition and mistreatment, suggesting quicker self-control blunted the effect of being treated badly. The interpretation needs restraint on exactly the point that makes the finding interesting. The three-way test -- whether inhibition changes how mistreatment shapes drinking over time -- came out non-significant at p = 0.10. So the moderation is visible in the cross-section but not confirmed in the trajectory, and the authors describe the brain-behaviour analyses as exploratory throughout. They also note their p-threshold was an uncorrected 0.05 across a set of candidate moderators, which inflates the chance of a false positive. Other limits are worth carrying into any reading of it: these were relatively healthy college students, not people with diagnosed alcohol use disorder; the sample was 64.7% female and the imbalance widened at follow-up; and the mistreatment measure does not distinguish its source, so sexism, racism and ordinary interpersonal unpleasantness are pooled into one score despite plausibly acting differently.