Negative urgency and central adiposity in a community sample: Moderated mediation by depressive symptoms and eating behaviors

Abstract

Negative urgency - acting rashly in response to negative emotions - is a risk factor for central adiposity. We examine whether the relationship between negative urgency and waist-to-height ratio (WHtR) is mediated by eating behaviors (emotional eating, external eating, and cognitive restraint) and moderated by depressive symptom severity, factors that could be targeted to reduce risk associated with negative urgency. Using baseline data from the Nathan Kline Institute-Rockland Sample (N = 872; mean age = 42.4 years, SD = 15.3; 65% female; 27% non-White; mean body mass index = 27.9 kg/m², SD = 5.9), we conducted a series of mediation and moderated mediation analyses controlling for age, sex, race, and socioeconomic status. Overall, there was a positive association between negative urgency and WHtR. Emotional eating (ab = 0.02, SE = 0.003, 95% CI [0.02, 0.03]) and uncontrolled eating (ab = 0.008, SE = 0.002, 95% CI [0.004, 0.01]) were partial mediators of the relationship between negative urgency and WHtR, while cognitive restraint was not. In a parallel mediation model, emotional eating remained significant, while uncontrolled eating did not. Depressive symptom severity moderated the indirect effect of negative urgency on WHtR through emotional eating (b_int = -0.08, p < .001) but not the direct effect of negative urgency on WHtR. Our results indicate that emotional eating is a viable potential mechanism explaining the relationship between negative urgency and WHtR, and the indirect effect of negative urgency on WHtR through emotional eating becomes stronger as depressive symptom severity decreases.

Keywords: Depressive symptom severity; Emotional eating; Impulsivity; Negative urgency; Waist-to-height ratio.

Figure 1. Parallel mediation model of the relationship between negative urgency and WHtR by emotional eating and uncontrolled eating (n = 856)

Note. For each path, unstandardized regression coefficients are presented. WHtR = waist-to-height ratio. ps are as follows: a₁ < .001, a₂ < .001, b₁ < .001, b₂ = .162, c’ = .277 c = .004.*p < .05, **p < .001

Figure 2. Moderated mediation of the relationship between negative urgency and WHtR through emotional eating and as a function of BDI-II scores (n = 854)

Note. For each path, unstandardized regression coefficients are presented. Results for paths a, b, and c’ are reported for BDI-II = 0 (1 SD below the mean). The interaction of negative urgency and BDI-II has a significant effect on the relationship with emotional eating but not with WHtR. The relationship between negative urgency and emotional eating is significant at all levels of BDI-II, but the relationship between negative urgency and WHtR is not significant at any level of BDI-II. The moderated mediation is signficiant. However, the direct effect of negative urgency on WHtR is not significant. BDI-II = Beck Depression Inventory-II; WHtR = waist-to-height ratio. b_int = unstandardized coefficient of the interaction between negative urgency and BDI-II. ps are as follows: a < .001, b < .001, c’ = .953, c = .002, b_int1 < .001, b_int2 = .137.*p < .05, **p < .001

Figure 3. Visualization of the significant interaction observed in the moderated mediation analysis: BDI-II as a moderator of the relationship between negative urgency and emotional eating (n = 854).

Note. The relationship between negative urgency and emotional eating is signficiant at all levels of BDI-II. Low = 1 SD below the mean on negative urgency; High = 1 SD above the mean on negative urgency; BDI-II = Beck Depression Inventory-II. **all ps < .001