The current study examined whether biologically-influenced temperamental traits implicated in the pathogenesis of disordered eating behaviors contribute to their development over and above current negative affect and dietary restraint. Participants (N = 276) were undergraduate psychology women who completed the BULIT-R, BDI-II, STAIS, RS, and the MPQ. Temperamental characteristics, particularly increased Negative Emotionality and decreased Positive Emotionality, were significantly associated with increased levels of bulimic symptomatology. Moreover, these dimensions accounted for small, but statistically significant amounts of the variance of bulimic symptomatology over and above current negative affect and dietary restraint. Contrary to expectations, impulsivity did not predict bulimic symptoms. This study provides evidence that temperamental dimensions related to mood rather than impulsivity are associated with bulimic symptomatology and contribute to bulimic symptoms over and above state mood and dietary restraint. Future empirical studies are needed to clarify relationships between temperament and other risk factors for disordered eating.
Numerous investigations have focused on dietary restraint and negative affect in the pathogenesis of bulimia nervosa. The dietary restraint model posits that a pattern of disinhibited eating is triggered and maintained by habitual dietary restriction (Herman and Mack, 1975 and Polivy and Herman, 1985). The negative affect model proposes that individuals binge eat to distract themselves from aversive emotions such as depression, anxiety, anger, or stress (Heatherton and Baumeister, 1991 and Leon et al., 1995). Highlighting the interrelationship among these risk factors, the dual-pathway model posits that both dietary restraint and negative affect contribute to the development of bulimic behaviors, although the complex nature of these relationships requires elucidation (Stice, 1994).
Despite substantial evidence documenting the relationship between eating disorders, dietary restraint and negative affect, available research evidence suggests that the etiology of eating disorders is multifactorial, and a combination of risk factors likely determines whether an individual develops bulimic symptomatology and whether those symptoms persist (Jimerson, Wolfe, Brotman, Metzger, 1996). Recent work documenting biological risk factors and heritability of eating disorders has suggested that specific temperamental traits also may increase risk for bulimic eating behaviors. A number of conceptualizations of temperament exist, and there is consensus that temperament is multidimensional and biologically-influenced (Tellegen, 1985 and Zuckerman, 1991). However, the use of differing assessments of temperament reflecting varying theoretical conceptions has hindered a precise understanding of the relationship between temperament and eating disorders.
In general, cross-sectional and longitudinal studies of temperament in clinical samples of individuals with bulimia nervosa (BN) have suggested that affective dysregulation, stress reactivity, and neuroticism are associated with BN (Casper et al., 1992, Lilenfeld et al., 2000 and Pryor and Wiederman, 1996). However, conflicting findings have been reported with regard to BN and other dimensions of temperament. For example, some studies have reported higher levels of impulsive tendencies in individuals with BN than in non-symptomatic control individuals (Brewerton et al., 1993, Bulik et al., 1995 and Waller et al., 1993), but others have not replicated this finding (Casper et al., 1992, Lilenfeld et al., 2000 and Pryor and Wiederman, 1996). Thus, a complete understanding of the relationship between temperament and BN remains a challenge.
Additionally, most studies of temperament and eating disorders have focused on clinical samples, and less is known about temperament and eating disorder symptoms in nonclinical populations. Research is needed to identify those individuals who may be at increased risk for eating disorders (Smolak, Levine, & Striegel-Moore, 1996), and examining temperamental characteristics in nonclinical samples might contribute to our understanding of similarities and differences between clinical and nonclinical groups. Furthermore, no studies have examined temperament in relation to dietary restraint and negative affect in either a clinical or a nonclinical sample.
Thus, the aim of the current investigation was to examine whether negative affect, dietary restraint, and specific temperamental factors were independently related to bulimic symptoms in a cross-sectional sample of college women. Consistent with previous findings, we predicted that negative affect and dietary restraint would be positively associated with bulimic symptomatology. With regard to temperament, we hypothesized that negative emotionality and characteristics of behavioral disinhibition and impulsivity would be positively related to bulimic symptomatology. Next, we examined whether a model that included restraint, negative affect, and temperament enhanced prediction of bulimic symptoms. Specifically, we hypothesized that temperament would contribute to the prediction of bulimic symptoms over and above restraint and negative affect. Finally, in a series of exploratory analyses, we examined the interactions among negative affect, dietary restraint, and temperament in predicting bulimic symptoms.
