Is talk cheap? Correspondence between self‐attributions about changes in drinking and longitudinal changes in drinking during the 2019 coronavirus pandemic

BACKGROUND: There are concerns that the coronavirus disease 2019 (COVID-19) pandemic may increase drinking, but most accounts to date are cross-sectional studies of self-attributions about alcohol-related impacts and the accuracy of those perceptions has not been investigated. The current study examined the correspondence between self-attributions of pandemic-related changes in drinking and longitudinally-measured changes in drinking and alcohol-related consequences in a sample of emerging adults. METHODS: In an existing ongoing longitudinal study on alcohol misuse (≥1 heavy episodic drinking day/month) in emerging adults, 473 individuals (M_age  = 23.8; 41.7% male) received a supplemental assessment from June 17th to July 1st, 2020, during public health restrictions in Ontario, Canada. These intrapandemic data were matched to the most recent assessment prior to the pandemic (~8 months earlier). Self-attributions about changes in drinking were assessed globally (i.e., increases/decreases/no change) and with higher resolution questions clarifying the magnitude of changes. RESULTS: Global self-attributions about changes in drinking substantively paralleled longitudinal changes in weekly drinking days (DD). In the longitudinal data, individuals' who self-reported increases in drinking exhibited significant increases; individuals' who self-reported decreases exhibited significant decreases; and individuals who self-reported no change exhibited nonsignificant changes. Higher resolution items likewise revealed longitudinal patterns of weekly drinking that were substantively consistent with self-attributions. Heavy DD and alcohol-related consequences exhibited similar patterns, but only individuals who self-reported large increases in drinking exhibited increases on these outcomes. Individuals who reported large increases in drinking also exhibited significant increases in depression and posttraumatic stress disorder symptoms. CONCLUSIONS: Self-attributions about drinking closely corresponded to longitudinal changes in drinking, supporting the validity of self-attributions in population-level surveys, particularly in young adults. Notably, a subgroup was identified that exhibited pronounced increases for all alcohol-related outcomes and concurrent increases in internalizing psychopathology.