Understanding the etiology of an eating disorder is perhaps the most complicated issue surrounding the disease, as teasing apart cause and consequence can be extremely difficult. This problem becomes immediately apparent when examining family factors associated with eating disorders. Research over the past decade has focused largely on identifying family factors that potentially contribute to the development of an eating disorder in an individual, and further refining these characteristics into prototypes for the “anorexic family” or the “bulimic family.” Identifying a pattern of specific family risk factors would be an extremely useful tool in recognizing those vulnerable for developing an eating disorder. While the research has been unable to paint an entirely complete picture of family characteristics, certain traits surface as typical to the eating disordered family. Unfortunately, much of the existing literature on family factors and eating disorders relies upon correlational data, as controlled studies are difficult to conduct within a family setting. Caution must therefore be applied to such findings, as one cannot assume causality; based on strictly correlational studies alone, it cannot be determined whether the family environment caused the eating disorder, or whether the eating disorder led to family dysfunction. Nevertheless, it remains useful to examine any significant factors that emerge from the literature in order to increase understanding about each potential factor influencing the development of eating disorders.
Although they both fall into the common continuum of eating disorders, anorexia nervosa (restricting subtype) and bulimia nervosa have symptoms that are quite distinct. Anorexia is characterized by symptoms that include a refusal to maintain a normal weight, an intense fear of gaining weight, and body image disturbance. Conversely, bulimia is generally marked by the maintenance of a normal weight, engaging in recurrent episodes of binge eating that are often followed by purging or some other form of inappropriate compensatory behavior. These compensatory measures can include self-induced vomiting, laxative or diuretic use, or excessive exercise. Bulimics also tend place extreme importance on body shape and weight in self-evaluations (Walsh & Garner, 1997). Such differences in symptomology certainly warrant the prediction that, if indeed family influences play a role in the development of eating disorders, there will be discrepancies in family patterns among anorexics and bulimics. The literature indicates that this is, in fact, the case on measures of family climate, interaction, and characteristics. While it may be premature to claim the existence of an “anorexic family” or a “bulimic family,” there are some stable traits that consistently emerge, indicating a possibility of eventually identifying some specific family risk factors for anorexia and bulimia.
An early study by Johnson and Flach (1985) attempted to describe general family characteristics of bulimic women as compared to age-matched women not suffering from any eating disorders. A group of 105 subjects meeting DSM-III criteria for bulimia completed several questionnaires, including the Eating Disorders Inventory (EDI), a measure of attitudes and behaviors common to anorexia and bulimia. The EDI is composed of eight subscales: drive for thinness, bulimia, body dissatisfaction, ineffectiveness, perfectionism, interpersonal distrust, introceptive awareness, and fear of maturity (Johnson & Flach, 1985). As a measure of family factors, subjects took the Family Environment Scale (FES), a questionnaire designed to quantify the subject’s perception of his or her family environment. The FES consists of ten individual subscales, designed to measure specific aspects of family functioning: cohesion, expressiveness, conflict, independence, achievement orientation, intellectual-cultural orientation, active-recreational orientation, moral-religious emphasis, organization, and control (Johnson & Flach, 1985). Scores of bulimic subjects differed significantly from those of control subjects on virtually every subscale of the FES. Importantly, bulimics perceived their families as low in cohesiveness and high in conflict, yet also very low in independence and highly achievement-oriented. Bulimics also rated their families as placing a low emphasis on cultural, recreational, or intellectual pursuits. According to the authors,
in which they are expected to achieve at a high level while being encouraged to remain enmeshed in a family that does not encourage self-sufficient or expressive behavior and does not support intellectual or social behavior that might facilitate successful achievement (Johnson & Flach, 1985).
Johnson and Flach (1985) identify some primary factors associated with the families of bulimic women that consistently appear in the literature. The findings of this study were replicated with a smaller sample size by Ordmann and Kirschenbaum (1986); bulimics scored higher than controls on the conflict subscale of the FES and significantly lower on cohesion, expressiveness, and active-recreational orientation. However, caution must be applied to the results of both studies; although the FES has been demonstrated as a valid and reliable measure of family functioning, it remains based on the perceptions of an individual, rather than the actual state of the family. Additionally, it is known that bulimia distorts an individual’s perceptions of themselves and their bodies; it might therefore be possible that the disorder is also capable of distorting an individual’s perception of their family functioning as well.
A paper by Stern et al (1989) attempts to combat this potential problem by administering the FES to a parent of the eating disordered subject, in addition to the subject and a control. Both anorexic and bulimic subjects were used in this study. Stern et al (1989) discovered that, regardless of the eating disorder, families in the experimental group tended to rate themselves as being mutually unsupportive, experienced higher incidence of conflict, and were less prone to open expression of emotion. The family evaluations given by parents and their children tended to be fairly consistent with one another. However, it is worth noting that parents generally rated their families in a more positive manner than did their children; this finding makes it difficult to interpret the results of the FES, as it is difficult to determine which party is being most accurate in their appraisal. Only one significant difference between perceived family functioning in anorexics and bulimics is reported in Stern et al (1989). Both bulimic subjects and their parents rated their families as having higher orientation-orientation than either anorexic or control subjects. Although tempting to construe the findings from such Stern et al (1989) as signifying a family etiology for eating disorders, the authors note that comparable FES scores have been collected from generally distressed families (defined as a family with at least one psychiatrically disturbed member). Abnormal scores on the FES may simply be a consequence of familial stress, rather than unique to eating disordered families (Stern et al, 1989). Causality remains an issue as well; the disordered family environment may be an outcome, and not a source, of the eating disorder.
A study by Kog and Vandereycken (1989) attempts to further tease apart differences in anorexic and bulimic family functioning. Taking the parameters set by Stern et al (1989) a step further, Kog and Vandereycken (1989) include the patient’s entire family in the analysis. Rather than examining the entire FES, this study focuses on those factors most associated with eating disordered families, based on the literature. These factors, cohesion, adaptability, and conflict, were measured via both behavioral and self-report tasks. Families participated in two semi-structured tasks, one collaborative decision-making task, and one conflict-resolution task. Following the completion of these activities, family members individually completed a questionnaire similar to the FES (Kog & Vandereycken, 1989). Eating disordered families scored high on scales of conflict avoidance, implying that although conflict did exist in these families, discussion of disagreement was absent. Furthermore, eating disordered families exhibited more stability than controls, suggesting a maladaptive rigidity of functioning, particularly when coping with discord (Kog & Vandereycken, 1989). Anorexics and their families evaluated themselves as a “tightly knit structure, with interpersonal boundary problems, stability, and conflict avoidance” (Kog & Vandereycken, 1989). Bulimic families followed the same overall trend, yet demonstrated even higher ratings than anorexics in these areas.
A more complete analysis of differences in anorexic and bulimic family functioning is found in Humphrey (1988). In this study, family triads consisting of a teenage girl (diagnosed with anorexia, bulimia, or no history of eating disorder) and her parents completed the Structural Analysis of Social Behavior (SASB) questionnaire, designed to measure both interpersonal and intrapsychic relationships. Three concentrations of attention are measured by the SASB: focus on other, focus on self, and intrapsychic focus (Humphrey, 1988). Cluster analysis performed on the data revealed patterns in family functioning. Consistent with other studies on this topic, a substantial amount of dissatisfaction and distress was a characteristic of family functioning for anorexic and bulimics. Furthermore, bulimics, including the binge-purge subtype of anorexia in this particular study, perceived less understanding, nurturing behavior, and support in their families, while experiencing greater instance of mutual blaming, neglect, and rejection (Humphrey, 1988). One new finding is the tendency of the mothers of bulimic teenagers to blame their daughters for family problems and to deny their own role in their family’s unhappiness. Conversely, anorexic families displayed a more positive attitude towards parent-child relationships, with especially high affection reported between fathers and daughters. Interestingly, a primary source of distress in Humphrey’s (1988) anorexic families was poor marital relationships. Wives reported their husbands as neglectful, withdrawn, unsupportive, and “sulky;” husbands in the study seemed unaware of any problems, reporting their marriages as generally satisfactory (Humphrey, 1988).
Scalf-McIver and Thompson (1989) also report specific issues of family dysfunction in eating disordered families. In addition to the familiar correlate of low familial cohesion, Scalf-McIver and Thompson (1989) also examined the consistency of parental affection via the Parental Inconsistency of Love scale. The severity of bulimic symptoms, as measured by the Bulimia Test (BULIT) and the Bulimia Cognitive Distortion Scale (BCDS), was correlated with a perceived inconsistency of maternal affection. This inconsistent maternal expression of love was the best predictor of severity of bulimic symptoms, while lack of cohesion proved a strong predictor of dissatisfaction with physical appearance (Scalf-McIver & Thompson, 1989).
Clearly, consistent factors emerge in the families of anorexic and bulimic individuals as compared to controls. Another approach, demonstrated in McNamara and Loveman (1990), is to identify those specifically at risk for an eating disorder and examine their family functioning as compared to individuals with full-blown cases of the disorder. Repeat dieters have been demonstrated at risk for developing bulimia, particularly those individuals engaging in frequent episodes of binge eating. After being assessed for eating habits and grouped into bulimic, repeat dieter, and non-dieter categories, subjects took the Family Assessment Device (FAD), a family-functioning questionnaire with subscales measuring family problem solving, communication, roles, affective responsiveness and involvement, and behavior control (McNamara & Loveman, 1990). Bulimics scored high on measures of affective involvement, and low on affective responsiveness and problem-solving scales. Repeat dieters and non-dieters were statistically indistinguishable from each other on all measures of the FAD, leading the authors to suggest family functioning as a possible mediating factor in developing full-blown bulimia; perhaps repeat dieting behavior never transformed into a binge-purge cycle due to constructive family functioning (McNamara & Loveman, 1990). Unfortunately, no follow-up study was conducted with these subjects to identify family traits of individuals in the repeat dieting group that may have eventually developed bulimia nervosa.
The literature concerning eating disorders and family functioning certainly indicates that families with difficulties in cohesion and high conflict are potentially at risk for eating disorders, especially bulimia. One cannot claim, however, that this specific combination of characteristics is a formula for the etiology of an eating disorder. It is more logical to assume that family dysfunction has a greater, and nonspecific, link to individual psychopathology, rather than to eating disorders alone. In fact, recent literature suggests that the disturbed family functioning may be more closely related to depression (i.e., Thienemann & Steiner, 1993). In a recent study, Laliberte et al (1999) point out that family distress creates an individual who is generally vulnerable to mental illness. The authors suggest that the specific content of what is “expressed, valued, and modeled” will be strongly associated with the specific symptoms expressed, and hypothesize a specific family climate containing certain content will yield a powerful predictor of disturbed eating behavior. Family functioning questionnaires, including the Family Achievement Emphasis Scale, the Family Social Appearance Orientation Scale, as well as body image and personality tests, were administered to a girl and her mother. Factor analysis revealed that families placing high emphasis on appearance, family reputation, family identity, and family achievement set the ideal climate for the development of eating disorders, as the appearance and achievement scales were excellent predictors of disturbed eating behaviors (Laliberte et al, 1999). The authors extended these findings into a psychiatric population, hoping that the content criteria would discriminate individuals with eating disorders from depressed and healthy control subjects. Specific family climate variables, including appearance (concern for weight and shape) and achievement concerns, distinguished the eating disorder group from both groups of controls. When using only family process factors (i.e., cohesion, conflict), the eating disorder group was discriminated from the healthy controls, but not from the depressed group (Laliberte et al, 1999). This very recent study suggests some important directions for future research with eating disorders and family factors. By comparing eating disordered families to a psychiatric population, it is possible to separate factors specific to eating disorder pathology from factors related to family distress alone.
One goal of researching family factors associated with eating disorders is to develop criteria from which those at risk for anorexia and bulimia can be identified. Early identification introduces the possibility of preventing eating disorders from even developing by combating specific environmental triggers, which might include family characteristics or functioning. Unfortunately, most of the existing literature is based upon correlational data, and does not yet allow this predictive power. While consistent factors do emerge as significant family characteristics associated with eating disorders, recent literature suggests that these factors, including low cohesion, lack of emotional expression, and high conflict, may simply be characteristics of a distressed family, rather than specifically characteristic of eating disordered families. Although difficult to draw definite conclusions from this occasionally vague data, all research that can further elucidate any aspect of an eating disorder is useful; each piece of information can contribute to a more complete understanding of these difficult disorders, and each potential factor influencing the etiology of eating disorders is therefore worth investigating.
Psychology DepartmentThe Health Psychology Home Page is produced and maintained by David Schlundt, PhD.
Vanderbilt Homepage | Introduction to Vanderbilt | Admissions | Colleges & Schools | Research Centers | News & Media Information | People at Vanderbilt | Libraries | Administrative Departments | Medical
|Return to the Health Psychology Home Page|
|Send E-mail comments or questions to Dr. Schlundt|