Behind Each Bite: A smartphone EMA study on the associations between affect, food cravings, and intake

Background: Today’s obesogenic environment challenges everyone’s dietary choices as ultra-processed foods are everywhere. This study investigates the interplay of three key variables in this context: affect, food’s craving and intake in daily life. We hypothesized that momentary negative affect would heighten food cravings and intake. Moreover, links between EMA variables and baseline dietary restraint and impulsivity will be computed. Method: Healthy-weight women (n = 106; 18–35 y.o.) completed a 14-day smartphone Ecological Momentary Assessment (EMA) study (8 prompts/day; M compliance = 90.68%). Positive affect (PA), negative affect (NA), craving degree, craving amount, and food intake were assessed. A multi-level vector autoregressive model (mlVAR) estimated temporal, contemporaneous, between-subject, and random-effects networks.Results: Temporal: PA (β = .19) and NA (β = .16) showed positive autocorrelations. NA predicted lower PA (β = −.07) and craving degree (β = −.04). Craving amount predicted increased intake (β = .16), while intake negatively predicted itself (β = −.12). Contemporaneous: PA and NA were strongly negatively linked (r = −.70). Craving degree and amount correlated robustly (r = .96); all other edges were weak (r ≤ .03). Between-subject: craving degree and amount co-varied (r = .77); craving amount related positively to intake (r = .29), while craving degree related negatively (r = −.19); affect showed no BS associations. Random-effects revealed heterogeneity, especially for craving-intake and affect-craving edges. Conclusions: Quantity—not intensity—of cravings best predicted subsequent eating, while affect showed small, inconsistent effects on both craving and intake. Substantial interpersonal variability highlights the need for individually tailored and longitudinal interventions targeting eating behaviors.