Toward an understanding of risk factors for bulimia.

The purpose of this article is to integrate diverse research efforts in an attempt to move toward an understanding of risk factors for bulimia. For this task, three questions in particular require attention. Because 90% of bulimics are women, a first question to address is, Why women? Second, despite the high prevalence of dieting and weight concerns among women in general, it is still a minority who evidence the clinical syndrome of bulimia, leading to the question, Which women in particular? These questions are considered from a range of perspectivesusociocultural, developmental, psychological, and biological Third, the rapidly increasing prevalence of bulimia in recent years raises yet another question, Why now? Our analysis points to research questions that must be examined before we can expand our understanding of the etiology of bulimia. In its end-of-the-year review, Newsweek referred to 1981 as "the year of the binge purge syndrome" (Adler, 1982, p. 29). This designation reflected the public's growing awareness of a significant sociocultural phenomenon, namely, the seemingly sudden and dramatic rise of bulimia. One year earlier, bulimia had become recognized as a psychiatric disorder in its own right in the Diagnostic and Statistical Manual of Mental Disorders (DSM-nI; American Psychiatric Association, 1980); this development has facilitated standardized assessment. In the last few years there has been a proliferation of literature on bulimia, as researchers and clinicians have attempted to describe the clinical picture of the disorder, to outline treatment approaches, and to identify factors associated with it. Even though the investigative forays into bulimia have really just begun, it now seems both possible and useful to draw together the current and sometimes disparate existing pieces of knowledge about the disorder and to propose working hypotheses about its etiology. A few efforts have already been made in this direction (Garner, Rockert, Olmsted, Johnson, & Coscina, 1985; Hawkins & Clement, 1984; Johnson, Lewis, & Hagman, 1984; Russell, 1979; Slade, 1982). Our own conceptualization of this disorder both permits better understanding of the risk factors already proposed and implicates additional variables in the etiology of bulimia. We hope that as we delineate possible risk factors of bulimia, it will become clearer where our current knowledge is most lacking and therefore where research is needed. An understanding of etiology will, we hope, also facilitate the clinical treatment of bulimia. As we think about bulimia and its recent rise, three questions in particular demand attention. First, bulimia is primarily a woman's problem, with research consistently indicating that approximately 90% of bulimic individuals are female ~ (Halmi, Falk, & Schwartz, 1981; Katzman, Wolchik, & Braver, 1984; Leon, Carroll, Chernyk, & Finn, 1985; Pope, Hudson, Yurgelun-Todd, & Hudson, 1984; Pyle et al., 1983; Wilson, 1984). Hence, a key factor that places someone at risk for developing bulimia is being a woman. One major question that demands an answer then is, simply, Why women? Second, it appears that weight concerns and dieting are so pervasive among females today that they have become normative (Rodin, Silberstein, & Striegel-Moore, 1985). An overwhelming number of women currently feel too fat (regardless of their actual weight) and engage in repeated dieting efforts (Drewnowski, Riskey, & Desor, 1982; Garner, Olmsted, & Polivy, 1983; Herman & Polivy, 1975; Huon & Brown, 1984; Mann et al., 1983; Moss, Jennings, McFarland, & Carter, 1984; Nielsen, 1979; Nylander, 1971; Polivy & Herman, 1985; Pyle et al., 1983; Wooley & Wooley, 1984). Despite the prevalence of dieting and weight concerns among women in general, it is still a minority who develop the clinical syndrome of bulimia, thus prompting another essential question: Which women in particular? 2 In our discussion, we will be conceptualizing a continuum ranging from unconcern with weight and normal eating, to "normative discontent" with weight and moderately disregulated/restrained eating, to bulimia (Rodin, Silberstein, & Striegel-Moore, 1985). The question of "which women in particular" can be seen, therefore, as a question of which women will move along this continuum from normative concerns to bulimia. Third, it is not women in all times and places but Correspondence concerning this article should be addressed to Judith Rodin, Department of Psychology, Yale University, P.O. Box 11A Yale Station, New Haven, CT 06520. i At present, there are insufficient data to discuss the etiology of the disorder in the 10% of bulimics who are men. Some of the risk factors specified for women may relate to men as well. Some speculation on what groups of men are most vulnerable will be considered briefly in the last section of the article. z Many investigators suggest that eating disorders should be conceptualized as a spectrum spanning anorexia nervosa, bulimia, and compulsive overeating (Andersen, 1983; Szmukler, 1982; Yager, Landsverk, Lee-Benner, & Johnson, 1983). In this relatively early stage of conceptualization, it seems useful to limit our scope to bulimia. However, we will sometimes draw on the existing literature about other eating disorders when relevant. A task for the future is clearly to delineate more precisely the commonalities and differences among the eating disorders and to develop a conceptual framework that integrates them. 246 March 1986 9 American Psychologist Copyright 1986 by the American Psychological Association, Inc. 0003-066X/86/$00.75 Vol. 41, No. 3, 246-263 rather women of this era in Western society who are developing bulimia. Therefore, a third question is, Why now? This question has received very little empirical attention. However, the seemingly sudden and dramatic rise of bulimia over the past few years suggests that we need to consider the possible role of sociohistorical factors. One critical aspect to the challenge of developing an etiological model of bulimia is the heterogeneity of the women who develop the disorder. Bulimic women differ with regard to their eating behavior and body weight, with some women exhibiting anorexia nervosa as well as bulimia either in the past or at present, others maintaining weight within the normal range, and others currently or in the past being obese (Beumont, George, & Smart, 1976; Garfinkel & Garner, 1982; Garner, Garfinkel, & O'Shaughnessy, 1985; Gormally, 1984; Loro & Orleans, 1981). Bulimic women can be divided into those who purge (by means of vomiting or abuse of cathartics) and those who do not resort to purging as a way of controlling their weight (Casper, Eckert, Halmi, Goldberg, & Davis, 1980; Garfinkel, Moldofsky, & Garner, 1980; Grace, Jacobson, & Fullager, 1985; Halmi et al., 1981). Furthermore, bulimic women vary greatly regarding the nature and extent of associated psychopathology. Some bulimic women do not exhibit any other psychiatric symptoms aside from those subsumed under the diagnosis ofbulimia (Johnson, Stuckey, Lewis, & Schwartz, 1982), whereas others show multiple types of psychopathology (Garner & Garfinkel, 1985; Garner, Garfinkel, & O'Shaughnessy, 1985; Hudson, Laffer, & Pope, 1982; Hudson, Pope, & Jonas, 1984; Lacey, 1982; Wallach & Lowenkopf, 1984). The implications of this heterogeneity for identifying risk factors are crucial. A particular risk factor that may be central to the etiology of the disorder in some women may be minor or even irrelevant in the development of bulimia in other women. Furthermore, this heterogeneity argues against unidimensional models of bulimia. Any model ofbulimia (e.g., biochemical or addiction models) still must consider the three questions that we are now posing. The questions--Why women? Which women in particular? Why now?--compose the starting point for our discussion of factors placing individuals at risk for bulimia. These questions compel us to consider bulimia from a range of perSpectives--sociocultural, developmental, psychological, and biological. Examining each of these perspectives in turn, we will consider the first two questions in tandem. From each perspective, we must try first to identify factors that might place women at greater risk than men for bulimia and second to understand which women in particular might be at greatest risk. Subsequently, we will consider our third question, Why now? Sociocultural Variables Central to an etiological analysis are the sociocultural factors that place women at greater risk than men for bulimia. We and others have reviewed data suggesting that risk increases because our society values attractiveness and thinness in particular, therefore making obesity a highly stigmatized condition (Boskind-White & White, 1983; Garner, Rockert, Olmsted, Johnson, & Coscina, 1985; Hawkins & Clement, 1984; Johnson et al., 1984; Rodin, Silberstein, & Striegel-Moore, 1985; Russell, 1979). Numerous studies suggest that this attitude affects people of all ages and that these social norms are applied more strongly to women than to men (see Rodin, Silberstein, & Striegel-Moore, 1985, for review). We begin the present analysis by asking which women in particular are affected by these sociocultural attitudes regarding attractiveness and weight, and then we suggest other signifcant social norms, not previously discussed, that may enhance the risk for bulimia in women. Which Women in Particular? How might the high value placed on thinness and the stigmatization of obesity in women have a greater impact on some women than on others, thus placing them at greater risk for bulimia? At a basic level, women at greatest risk for bulimia should be those who have accepted and internalized most deeply the sociocultural mores about thinness and attractiveness. In other words, the more a woman believes that "what is fat is bad, what is thin is beautiful, and what is beautiful is good," the more she will work toward thinness an