Within-Day Variability in Negative Affect Moderates Cue Responsiveness in High-Calorie Snacking.

Background:Many discretionary foods ("snacks") contribute both to individual health risks and to global issues, in particular through high carbon footprints and water scarcity. Snacking is influenced by the presence of snacking cues such as food availability, observing others eating, and negative affect. However, less is known about the mechanisms underlying the effects of negative affect. This study examines whether the individual odds of consuming high-calorie snacks as a consequence to being exposed to known snacking cues were moderated by experiencing (i) higher or lower total negative affect per day or (ii) higher or lower negative affect variability per day. Methods: Secondary analysis of an ecological momentary assessment study of 60 participants over 14 days with food logs and randomly timed assessments of known snacking cues. High total daily negative affect levels (daily within-participant means) and negative affect variability (daily within-participant SDs) were examined as moderators to predict high-calorie snacking in three-level hierarchical random effects logistic regressions. Results: Consistent with previous studies, the odds of snacking increased when food was available (OR = 5.05, 95% CI 3.32, 7.66), when others were eating (OR = 5.11, 95% CI = 3.73, 6.99), and when participants experienced more negative affect (OR = 1.02, 95% CI = 1.01, 1.03). Associations for food availability (OR = 0.92, 95% CI 0.86, 0.99) and others eating (OR = 0.95, 95% CI 0.91, 0.99) were significantly moderated by negative affect variability such that associations between cues and high-calorie snacking were weaker on days with higher negative affect variability, but not negative affect levels. Conclusion: The relationship between cues to high-calorie snacking and snacking behavior varies with variability in negative affect, suggesting a complex relationship between affect and high-calorie snacking. Clearer conceptualizations on the relation between affect and eating are needed.