The prevalence of eating disorders among American women has increased dramatically in the past decade. In turn, the psychological community has expanded it's scope of research and study by focusing more attention on eating disorders and concentrating on other extended issues related to eating disorders. The desire to distinguish and understand a possible relationship between bulimia nervosa and depression has become a major focus within the field. The purpose of this paper is to introduce the reader to literature from two different journal articles on this relationship.
Presently, there are two predominant hypotheses on the relationship between bulimia nervosa and depression. The first hypothesis states that bulimia nervosa is an affective variant of depression. (Hinz and Williamson, 1987) This idea came about due to early reports of a high prevalence of clinical depression in bulimics and a high lifetime prevalence of depression in the families of these patients. Recent studies, however, provide evidence that this type of relationship between bulimia and depression is still unconfirmed. (Levy et al., 1989) The second hypothesis constitutes that bulimia is a "distinctive diagnostic entity with a psychopathological process different from that of other mental disorders". (Hinz, et al., 1987) The following articles review the findings of research which effect the level of support for these two hypotheses.
Bulimia is characterized by a number of symptoms. Many of these symptoms are also common among depression. Schlesier-Carter, et al., 1989 state that for bulimia,
"the distinctive behaviors appear clinically to occur in the context
of (a) unrealistic beliefs and evaluations about eating and body weight,
and (b) dysphoric affective states... Cognitions manifested by bulimics
include dysfunctional attributions regarding weight gain, guilt, and self-deprecating
attributions centered on feeling out of control."
Schlesier, et al. (1987) state that depression:
"...carries characteristic cognitive features of clinical import.
Maladaptive thinking patterns related to depressive symptomatology include
negative automatic thoughts, dysfunctional attitudes, and causal attributions."
Therefore, there are connections between these two disorders. There is a link in the relationship between bulimia and depression. Understanding this link is essential for treatment and therapy.
Lisa Hinz and Donald Williamson (1987) reflect on the evidence presented in past studies relating bulimia and depression and discriminate between the idea that bulimia is an affective variant of depression or a separate diagnostic entity often associated with depression. In their attempt to understand the relationship, the authors point out the complications that arise in an evaluation of studies that deal with depression and other illnesses due to the common associations made between the two. Hinz and Williamson (1987) state that "evaluation of the affective variant hypothesis is difficult because it is well established that depression accompanies many chronic psychiatric and physical illnesses". In turn, knowing this common occurrence, in order to disprove the affective disorder theory there are four propositions that are made in an attempt to evaluate whether the syndromes associated with bulimia are equivalent to the syndromes of affective disorders. The propositions, or assumptions, that the authors make are that "a) depression should accompany bulimia with a very high frequency, and both syndromes should share a similar phenomenology; b) diagnostic tests for depression should be positive for bulimia, with the same frequency found in a known depressive population; c) studies of etiology should show very similar etiological factors for bulimia and affective disorders; and d) studies of bulimia as a distinctive syndrome should consistently fail to show psychopathological processes that are distinctly different from other diagnostic entities". (Hinz, et al., 1987) In coming to their conclusion about bulimia as an affective variant to depression, Hinz and Williamson analyze the findings of the proponents of this hypothesis in descriptive studies, structured interviews studies, controlled investigations using standardized psychological tests, and treatments with antidepressant medication.
Supporters of the affective variant of depression hypothesis state that five different studies performed by Johnson et al. (1982), Fairburn and Cooper (1982), Abraham and Beaumont (1982), Herzog (1982), and Hatsukami et al. (1984) show that the phenomenology of bulimia suggests that bulimia is a symptom of depression. The collaboration of these studies point to commonalities and crossovers between the DSM-III diagnosis requirements for bulimia and depression, as well as a statistical estimate that 27% to 75% of bulimics may have experienced an affective disorder at some time in their lives. (Hinz, et al., 1987) Hinz and Williamson, however, claim that the findings of the studies do not prove that bulimia is an affective illness. Instead, the findings are considered to be weak due to methodological reasons. The studies failed to use structured interviews or questionnaires, to report the reliability and validity of the instruments, and to use control or comparison groups. Also, the studies do not show a high frequency of depression in bulimia at the phase of initial evaluation. Higher rates of depression are only found in bulimics when occurrence of depression across a lifetime is considered and most people with psychiatric issues experience depression at some point in their lives. (Hinz and Williamson, 1987)
Herzog (1984), Walsh et al. (1985), and Fairburn and Cooper (1986) each performed structured interview studies in attempt to link bulimia as a symptom of depression while establishing reliability and validity for their assessment. While the combination of these studies and the descriptive studies were able to show an association and linkage between depression and bulimia, the evidence found does not support the hypothesis that bulimia is an affective variant of depression. Furthermore, Hinz and Williamson (1987) acknowledge that the Fairburn and Cooper's study, which directly compared the phenomenology of depression and bulimia, "found distinct differences in symptom patterns between the two groups...in particular, this study found that bulimics were less depressed and more obsessive and anxious". While many studies and many people do acknowledge a linkage between bulimia and depression, the above studies show that bulimia nervosa and depression do not share similar phenomenology.
Many researchers have used controlled investigations using standardized psychological tests in their attempt to establish bulimia nervosa as an affective variant of depression while preventing bias on the part of the interviewer. Two of these studies were performed by Weiss and Ebert (1983), Katzman and Wolchik (1984), and Williamson et al. (1985). Weiss and Ebert (1983) found that bulimics scored "significantly higher rates of substance abuse, stealing, depression, suicide attempts, anxiety, somatic complaints, irregular eating habits, lowered self-esteem, and external locus of control in the bulimic group, in comparison with the control group" after comparing their subjects' responses on seven different psychological tests. (Hinz and Williamson, 1987) Hinz and Williamson (1987) also cite evidence found that bulimic women score significantly higher than other groups on levels of dietary restraint, binge eating, demand for approval, depression, body-image dysphoria and lower self-esteem. Furthermore, Hinz and Williamson (1985) "concluded that bulimic women were more neurotic, anxious, depressed, obsessive, impulsive, and manipulative that obese and normal weight women. Hinz and Williamson (1987) interpret the findings in these studies to show that "depression is a significant problem that is associated with bulimia but is not the only form of psychopathology observed in eating disorders". More specifically, these studies show that people with bulimia nervosa tend to be more neurotic, obsessive, and anxious. There is also evidence that bulimics have higher levels of body-image distortion. Furthermore, there are findings that current levels of depression are moderate and that only a minority of bulimics are clinically depressed at the time of evaluation. The prevalence of depression increases when bulimics are questioned about depression across their lifespan instead of level of depression directly prior to or during a bulimic period. (Hinz and Williamson, 1987)
Several studies have been done comparing the use of antidepressants for bulimia nervosa. Supporters of the affective variant hypothesis have used the evidence from these studies as a major source of supporting data. Hinz and Williamson (1987) state that while "some antidepressants -phenelzine, desiprimine, and imiprimine- clearly are effective for reducing binge eating, two antidepressants -mianserin and amitriptyline- have been found to be ineffective. Therefore, the strong conclusions reached by advocates of the affective variant hypothesis appear to be premature". Furthermore, it is illogical to state that bulimia is an affective variant of depression simply because the medication used to relieve depression also affects symptoms found in bulimia. In doing so, researchers are "equat(ing) treatment and etiological factors without empirical demonstration of their association". (Hinz and Williamson, 1987)
Hinz and Williamson (1987) conclude that while depression is common among bulimics, stating that bulimia is an affective variant of depression is premature. The authors found that the many of the studies used by supporters of the affective variant theory have flaws in their research and findings. Furthermore the four assumptions made by Hinz and Williamson (1987) that are expected assuming that the affective hypothesis was correct were not reflected in the studies.
Along with Hinz and Williamson (1987), Levy et al. (1989) call into question the hypothesis that bulimia is an affective variant. In their article, Levy et al. (1989) analyze the relationship between bulimia and depression by examining clinical evidence, family studies, polysomnography, and neuroendocrine findings.
The presence of depressive symptoms in bulimics have long been recognized. Early reports and clinical evidence document that anorexics with bulimic behavior "are more frequently given diagnosis of depression and are more suicidal than nonbulimic patients with anorexia nervosa. Normal weight patients with bulimia also appear to have a high frequency of suicidal attempts and depression." (Levy et al., 1989) Many studies have been performed which note a high prevalence of concurrent or lifetime depression in bulimics. Studies have found that of their subjects, 59% of normal-weight bulimics and 80% of anorexic bulimics experienced major depression at some point in their lives. Fieghner found that 79% of the bulimic subjects experienced concurrent depression. Herzog reported 24% of sample size of 55 bulimic nervosa patients met the criteria for concurrent depression and Walsh et al. reported that 30% of the 50 subjects experienced concurrent depression. (Levy, et al., 1989) While these studies prove to be beneficial and do suggest a link between depression and bulimia, it is important to recognize that none of these studies used a control group which was studied simultaneously. Therefore, there is possible bias in the findings. Furthermore, many researchers interpret their studies differently. For instance, different researchers consider the same phenomena differently. Reported findings often depend upon what the researcher considers primary. Levy et al. (1989) explains that "in some instances, depression may be secondary to the abnormal eating behavior and the ensuring psychosocial consequences. Indeed, an apparent high frequency of depressive symptoms in bulimia might to a degree reflect the fact that depressed mood and self-deprecating thoughts were required for the DSM-III diagnosis of bulimia. At other times, depressive symptoms may reflect a condition of malnutrition." It is also possible to assume that many of the bulimic subjects involved in studies have come to receive treatment. These subjects may be more depressed than other bulimics who have not seeked treatment. Therefore, findings may not be a fair generalization of the entire bulimic population. (Levy et al., 1989)
Family studies have also been used to determine the relationship between bulimia and depression. Many of the early studies suggest that bulimia is an affective variable of depression due to the prevalence of familial depression or other affective disorders in bulimic patients. However, most of these early studies failed to personally interview or test the bulimics' family members for depression. "All positive family history studies have used either the probands' self-report or, less often, nonblind interviews, the prevalence of familial depression is difficult to interpret because of possible bias." (Levy et al., 1989) Furthermore, many studies did not work with control groups which leads to a possible bias. Levy et al. (1989) recommend that more familial studies using structured interviews and blind family interviews are needed for a more precise understanding to the relationship.
Sleep abnormalities have also been investigated in the search for understanding the relationship between depression and bulimia. Abnormal REM latencies and densities have been reported to be common in both bulimia and depression. More recent studies, however, find that there are differences between the two. Originally, findings suggested like depressed patients, bulimics had shorter REM latencies, as well as, REM densities. This finding has been refuted by later research which suggests that REM latencies and densities of bulimic patients are more similar to those of the control groups.
Several studies using neuroendocrine tests have been performed in attempt to further analyze this relationship. After investigating hypothalamic-pituitary-adrenal axis, thyrotropin growth hormone, and prolactin levels in bulimic and depressed subjects, it has been concluded that while the neuroendocrine system is imbalanced in bulimics there are differences than that experienced by depressed subjects. (Levy, et al., 1989) Evidence was found that show bulimics have "elevated B-hydroxybutric acid, enlarged ventricles on CT scans, and failure of an elevated growth hormone level to result in elevated somatomedin C." (Levy, et al., 1989) These findings suggest that the type of nutrition experienced by bulimics affect the abnormalities of the neuroendocrine system. However, current evidence from the Levy, et al. (1989) study shows that "neuroendocrinopathis in bulimic women have not been found to correlate with depression".
Presently there is inadequate evidence and support for the affective disorder hypothesis. In turn, there is stronger evidence to support the second hypothesis which constitutes bulimia as a "distinctive diagnostic entity with a psychopathological process different from that of other mental disorders". (Hinz, et al., 1987) Further research and study on the relationship between bulimia and depression is highly encouraged. With the high percentage of depression found in bulimic patients, understanding the link between these disorders is essential for treatment development.
Hinz, L. and Williamson, D. (1987). "Bulimia and Depression: A
Review of the Affective Variant Hypothesis". Psychological Bulletin,
Levy, A., Dixon, K., and Stern, S. (1989). "How are Depression
and Bulimia Related". The American Journal of Psychiatry, 146(2),
Schlesier-Carter, B., Hamilton, S., O'Neil, P., Lydiard, R., and Malcolm,
R. "Depression and Bulimia: The Link Between Depression and Bulimic
Cognitions". Journal of Abnormal Psychology, 98(3), 322-325.
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|