Dialectical Behavior Therapy for Binge Eating and Bulimia (4 page)

In an experimental study of women with BED and weight-matched controls,
negative emotional states were associated with both loss of control and an eating
episode labeled as a binge (Telch & Agras, 1996). And an experimental study that
manipulated caloric deprivation and the conditions for inducing a negative or a
neutral mood noted the self-defned binges of obese women with BED to be signifcantly associated with negative mood rather than with caloric deprivation (Agras
& Telch, 1998). Other experiments describe similar fndings regarding the role of
negative mood in triggering binge eating (Chua, Touyz, & Hill, 2004).

A study that utilized experience sampling from participants who tracked their
moods and eating behaviors with handheld computers over 6 days reported that
more aversive moods preceded binge eating in participants with BED compared
with weight-matched controls without BED (Greeno, Wing, & Shiffman, 2000).
Furthermore, at random prompts throughout the course of the study, participants
with BED indicated experiencing a signifcantly worse mood on average than participants without BED. Greater diurnal fuctuations in depression and anxiety
were recorded among binge eaters assessed during a 2-week period (Lingswiler,
Crowther, & Stephens, 1987). The binge eaters, whether of normal weight or overweight, also described negative moods as more frequent during eating and binge-eating episodes than did non-binge-eating participants (Lingswiler et al., 1987). In
addition to being more likely to eat in response to negative moods, there is evidence
that binge eaters evaluate situations as being more stressful than do non-binge
eaters (Hansel & Wittrock, 1997).

A potential reason that binge eaters eat in response to negative moods and
judge situations as more stressful may be that they have defcits in the ability to
regulate their emotions. Supporting evidence for the role of emotion dysregulation
in binge eaters was found by Whiteside and colleagues (2007), who reported that
diffculties in identifying and making sense of emotional states, along with limited
access to emotion regulation strategies, were associated with binge eating over
and above the effects of gender, food restriction, and overevaluation of weight and
shape.

BULIMIA NERVOSA

BN is marked by a preoccupation with thinness and episodes of binge eating followed by compensatory behaviors (e.g., vomiting, extreme dietary restriction, lax—
ative or diuretic abuse, overexercise). To establish the diagnosis, these episodes
must have occurred at least twice a week on average during the prior 3 months.
The disorder is not diagnosed within the context of anorexia nervosa, and most
clients are of normal weight.

DIALECTICAL BEHAVIOR THERAPY FOR BINGE EATING AND BULIMIA

The onset of BN typically occurs in adolescence or early adulthood and is
often precipitated by dieting (American Psychiatric Association, 2000). Females
make up the majority (90%) of those diagnosed (American Psychiatric Association,
2000), with approximately 1 in 100 women meeting criteria for BN (Hoek & van
Hoeken, 2003; Hsu, 1996). Rates of BN may be higher within certain subgroups.
For instance of college women in the United States, 2 to 4 in 100 may be affected
(Katzman & Wolchick, 1984; Healy, Conroy, & Walsh, 1985; Pyle, Halvorson, Neuman, & Mitchell, 1986; Drewnowski, Yee, & Krahn, 1988; Pyle, Neuman, Halvorson, & Mitchell, 1991).

BN has a chronic course that is likely to be unstable (Fairburn et al., 2000),
with waxing and waning periods of binge eating and purging over time (Milos,
Spindler, Schnyder, & Fairburn, 2005; Wilson, Grilo, & Vitousek, 2007). Rates of
remission range from an estimated 31% to 74% (Ben-Tovim et al., 2001; Grilo et al.,
2003; Milos et al., 2005; Wilson et al., 2007). When remission takes place, it tends
to be time-limited, so that relapse is common (Ben-Tovim et al., 2001; Herzog et al.,
1999; Wilson et al., 2007).

Like BED, BN is associated with impairments in psychiatric, social, and phys—
ical functioning. These are elaborated in the following sections.

Psychiatric and Eating-Disorder-Specifc Symptomatology in BN

People with BN frequently suffer from co-occurring Axis I (e.g., mood, anxiety,
substance use) and Axis II (e.g., borderline personality, obsessive–compulsive) psychiatric disorders. Although estimates of prevalence vary across studies, reviews
of the literature have found higher rates of lifetime depression (e.g., 60–70%) in
patients with BN (Godart et al., 2007) than in the general community (16.2%;
Kessler et al., 2003), particularly when larger sample sizes are studied. Comor—
bid anxiety in individuals with BN is also more common than in controls, with
lifetime prevalence rates of least one anxiety disorder (e.g., obsessive–compulsive
disorder, social phobia, posttraumatic stress disorder) ranging from 25 to 75%
(Swinbourne & Touyz, 2007). Interestingly, the onset of the anxiety disorder in
most cases precedes that of the BN (Swinbourne & Touyz, 2007). Co-occurring
alcohol abuse or dependence is also very common, with a rate of 46% found by
Bulik and colleagues (2004). Rates of lifetime substance abuse also appear ele—
vated in women with BN, ranging from 18% (Herzog et al., 2006) to 28% (Lacey,
1993). Lifetime rates of alcohol abuse and substance use disorders in individuals
with BN are higher than in the general community (13.2% and 14.6%, respec—
tively; Kessler et al., 2005).

Rates of personality disorders, particularly Cluster B types, have also been
shown to be greater in people with BN. In a review of prevalence rates derived from
diagnostic interviews (a more stringent assessment method than self-report), borderline personality disorder rates were 6–37%, and avoidant personality disorder
rates were 2–36% (Cassin & von Ranson, 2005). In comparison, prevalence rates
in community samples have been reported at 0.5% for borderline personality disorder and 1.8% for avoidant personality disorder (Samuels et al., 2002).

As to eating-disorder-specifc psychopathology, people with BN, compared
with non-eating-disordered controls, report signifcantly higher levels of concerns

Why Dialectical Behavior Therapy?

regarding body shape, body weight, eating (e.g., preoccupation with food and calo—
ries), and dietary restriction (Cooper, Cooper, & Fairburn, 1989).

Social Impairment and BN

BN affects social adjustment. For instance, individuals with BN, compared with
non-eating-disordered controls, depict more overall social impairment, specifcally
in the areas of work, leisure, and family relationships (Herzog, Keller, Lavori, &
Ott, 1987). Women whose BN is active, compared with women whose BN is in
remission and with non-eating-disordered controls, report signifcantly less emotional support (Rorty, Yager, Buckwalter, & Rossotto, 1999). And both those with
active or remitted BN, compared with non-eating-disordered controls, expressed
signifcant dissatisfaction with the quality of emotional support provided by rela—
tives (Rorty et al., 1999). The social impairment experienced by patients with BN
appears to be enduring. Ten years after being diagnosed with BN, women contin—
ued to experience diffculties in their interpersonal relationships (Keel, Mitchell,
Miller, Davis, & Crow, 2000).

Studies of quality of life in BN show that individuals with a history of BN
versus those without such a history report more diffculties, particularly with
emotional functioning (Doll, Petersen, & Stewart-Brown, 2005). Compared with
a group of individuals with mood disorders, eating-disordered patients, including
patients with BN, reported having a worsened quality of life (de la Rie, Noorden—
bos, & van Furth, 2005). These differences in quality of life seem long-standing,
with former eating-disordered patients having a poorer quality of life than a control reference group (de la Rie et al., 2005).

Physiological Consequences of BN

BN is associated with serious physiological consequences, especially among those
who regularly vomit or engage in laxative abuse. Though mortality due to BN is
low, it is not insignifcant. Crude mortality rates due to all causes range from 0.3%
to 2% (Fichter, Quadfieg, & Hedlund, 2008; Keel & Mitchell, 1997). Potentially
life-threatening complications include low potassium (hypokalemia), esophageal
ruptures, cathartic colon, impaired kidney function, cardiac arrythmias, and cardiac arrest (Kaplan & Garfnkel, 1993; Sansone & Sansone, 1994).

In one study of 275 women with BN, the most common complaints were weak—
ness (84%), bloating (75%), cheek puffness (50%), dental symptoms (36%), and
fnger calluses (27%; Mitchell, Hatsukami, Eckert, & Pyle, 1985). Even when not
directly life threatening, bulimic behaviors profoundly affect the body in terms of
oral complications, gastrointestinal symptoms, renal and electrolyte abnormalities, cardiovascular symptoms, and negative consequences to the endocrine system
(see, e.g., reviews by Mehler, Crews, & Weiner, 2004; Mitchell & Crow, 2006). Ero—
sion of teeth enamel, for example, is usually seen within 6 months of self-induced
vomiting and is always evident in those suffering for 5 or more years (Althshuler,
Dechow, Waller, & Hardy, 1990). Parotid gland enlargement is estimated to affect
between 10 and 50% patients with BN (Mehler et al., 2004). Hypokalemia occurs
in approximately 14% of bulimic patients. This serious electrolyte disturbance

DIALECTICAL BEHAVIOR THERAPY FOR BINGE EATING AND BULIMIA

may potentially lead to other complications, such as cardiac arrythmias and the
degeneration of cardiac muscle (Casiero & Frishman, 2006). In terms of endocrine
system involvement, although evidence for a causal relationship between bulimia
and diabetes is mixed, it is clear that the presence of an eating disorder in addition
to diabetes is linked to a worsened diabetic course—including the development of
end-organ damage at a younger age (Rydall, Rodin, Olmsted, Devenyi, & Dane—
man, 1997).

Emotions, Affect Regulation, and BN

Aversive emotions may bring about, maintain, and be a consequence of BN behaviors. Bulimic behaviors are frequently related to negative affective states, including
anxiety, depression, and anger (e.g., Abraham & Beumont, 1982; Arnow, Kenardy,
& Agras, 1995; Stice, Killen, Hayward, & Taylor, 1998). In a study recording 2
weeks of data from participants with BN who were given handheld computers,
lower mood (less positive affect, more negative affect, higher anger/hostility, and
higher stress) was reported on days when binge eating and vomiting occurred
(Smyth et al., 2007). Within single days, the researchers noted a worsened mood
trajectory over the hours prior to a binge–purge episode and a sharply improved
mood trajectory following the event. These fndings help explain the persistence of
BN behaviors in the short run despite their not being an effective overall coping
strategy. In other words, the average “best” mood on a binge–purge day was still
more negative than the mood on days on which no binge–purge occurred. How—
ever, within a few hours after the event, binge and purge behaviors are strongly
negatively reinforced by allowing escape or avoidance of strongly negative affective
states (Smyth et al., 2007). These results are supported by other researchers (e.g.,
Lingswiler, Crowther, & Stephens, 1989; Lynch, Everingham, Dubitzky, Harman,
& Kassert, 2000; Powell & Thelen, 1996; Steiger et al., 2005). Likewise, purging
or laxative use in individuals with BN has been shown to reduce negative affect
prompted by binge eating (e.g., Powell & Thelen, 1996).

Self-report studies suggest that, in addition to increased negative mood on
binge–purge days, individuals with BN have higher levels of depression (Bulik,
Lawson, & Carter, 1996) and anger (Waller et al., 2003) and more fuctuating
moods (Johnson & Larson, 1982). Individuals with BN may also have defcits in
the processing of emotions. For instance, participants with BN, compared with
controls, show attentional defcits, including paying selective attention to emotionally laden words (e.g., body shape or weight and food-related words) when using the
Stroop paradigm (Dobson & Dozois, 2004), the visual probe paradigm (Rieger et
al., 1998), and the dichotic listening task paradigm (Schotte, McNally, & Turner,
1990). When presented food cues in experimental paradigms, participants with
BN, compared with those without BN, indicate greater anxiety (Bulik et al., 1996)
and have a potentiated startle refex, suggesting strong negative affect (Mauler,
Hamm, Weike, & Tuschen-Caffer, 2006). In addition, participants with BN compared with normal controls, endorse greater diffculties with self-awareness of
emotions (Legenbauer, Vocks, & Ruddel, 2008). Along with other eating-disordered
populations, they report more diffculties with distress tolerance than do women