Content uploaded by Gabriela Jagielska
Author content
All content in this area was uploaded by Gabriela Jagielska on May 28, 2017
Content may be subject to copyright.
Psychiatr. Pol. 2017; 51(2): 205–218
PL ISSN 0033-2674 (PRINT), ISSN 2391-5854 (ONLINE)
www.psychiatriapolska.pl
DOI: https://doi.org/10.12740/PP/64580
Outcome, comorbidity and prognosis in anorexia nervosa
Gabriela Jagielska1, Iwona Kacperska2
1 Department of Child Psychiatry, Medical University of Warsaw
2 Child and Adolescent Psychiatry, Independent Public Children’s Clinical Hospital in Warsaw
Summary
Anorexia nervosa (AN) is a relatively common disorder, especially in adolescent and
young adult women. The lifetime prevalence of AN in females ranges from 1.2 to 2.2%.
The prevalence in males is 10-times lower. The condition is associated with a high risk of
chronic course and poor prognosis in terms of treatment and the risk of death. Longer follow-
up periods seemed to correspond with increased improvement rates and increased mortality.
Onset of the disorder during adolescence is associated with better prognosis. It is reported that
as much as 70% to over 80% of patients in this age group achieve remission. Worse outcomes
are observed in patients who required hospitalization and in adults. Recent studies indicate
improved prognosis for cure and lower mortality rates than previously reported. However,
the recovery can take several years and AN is associated with high risk of developing other
psychiatric disorders during the patients’ lifetime, even after recovery from AN (mainly:
aective disorders, anxiety disorders, obsessive-compulsive disorders, substance abuse dis-
orders). Studies indicate that bulimic symptoms often occur in the course of anorexia nervosa
(especially within 2–3 years from the onset of AN). The authors present a review of literature
on the course, comorbidity, mortality, and prognostic factors in AN. Better knowledge of the
course of anorexia can contribute to more realistic expectations of the pace of symptomatic
improvement, as well as to a creation of therapeutic programs which are better adapted to
the needs of the patients.
Key words: anorexia nervosa, treatment outcome, mortality
Introduction
According to large population-based studies, the lifetime prevalence of anorexia
nervosa (AN) in females ranges from 1.2 to 2.2% and that of sub-threshold/atypical AN
(aAN) from 2.4 to 4.3% [1–3]. The estimated prevalence in males is 10-times lower
[4], however, there are studies indicating the prevalence of AN in the male population
to be somewhat higher. Bulik at al. [1] reported the prevalence of AN (according to
DSM-IV criteria) in females to be 1.2%, and in males 0.29%, and according to DSM-
Gabriela Jagielska, Iwona Kacperska
206
5 criteria AN prevalence in the population of 14-year-olds was estimated at 3.2% in
females, and 1.6% in males [5].
The onset of AN most commonly occurs during adolescence, with 15–19-year-
olds constituting 40% of newly diagnosed cases. Peak incidence is between 14 and
18 years of age. Approximately 85% of cases begin before the age of 20, and nearly
all of them before the age of 25 [6]. AN is relatively rare in the population under 13
years of age [4]. Some studies indicate increasing incidence of AN [7–9], especially in
females aged 15–24 years [8]. A meta-analysis of studies on AN incidence in Northern
Europe indicates an increasing AN incidence until the 1970s, with a subsequent plateau
at approximately 5–5.4 cases per 100,000 [8].
AN is a psychiatric disorder with a substantial comorbidity, chronic course, and the
highest mortality among all psychiatric conditions [10]. The disorder present a chal-
lenge for doctors and therapists as AN suerers often deny being aected, hide their
symptoms and signs, or avoid treatment. Studies demonstrate that only approximately
50% of AN cases are diagnosed [2] and only 1 in 3 of those diagnosed receive special-
ist care [11]. Out of those who begin treatment, 20–51% of inpatients and 23–73% of
outpatients avoid or drop out from their treatment program [12].
Better knowledge of the course of anorexia can contribute to more realistic expecta-
tions of the pace of symptomatic improvement, as well as to a creation of therapeutic
programs which are better adapted to the needs of the patients.
The course of anorexia nervosa
A meta-analysis of 119 studies published in the second half of the 20th century [13]
demonstrated that 46.9±19.7% of patients (range: 0–92%) recovered, 33.5±17.8% of
patients (range: 0–75%) improved, and 20.8 ±12.8% of patients (range: 0–79%) suf-
fered from chronic AN. Such considerable dierences in the course of this condition
are mostly due to variations in terms of the adopted recovery criteria, follow-up dura-
tion, duration of symptomatic remission, study group clinical characteristics, and the
type of management. Longer follow-up periods seemed to correspond with increased
improvement rates (and increased mortality). Follow-up durations of <4 years, 4–10
years, and >10 years corresponded to the recovery rates of 32.6±24.3%, 47±15.7%,
and 73.2±16.2%, respectively.
Long-term treatment outcomes are better when considered with respect to the
particular AN symptom normalization. For example, normalization of weight, men-
ses, and eating habits were observed in 59.6±15.3%, 57.0±17.2%, and 46.8±19.6% of
cases, respectively [13]. A study in a group of adolescents receiving family therapy,
with a 2–6-year-long follow-up, demonstrated that the rates of somatic recovery (de-
ned as reaching >85% of ideal body weight – IBW) and resumption of menses were
signicantly higher than those of psychological recovery (94.4%, 90.5%, and 57.1%,
respectively). In this study, the time necessary to reach weight improvement was less
than 1 year and psychological remission was achieved after another year [14].
207
Outcome, comorbidity and prognosis in anorexia nervosa
The course of anorexia nervosa in population-based studies
Based on the suppositions that patients requiring inpatient treatment may have
poorer prognosis and that some patients would never begin therapy, some studies were
conducted to assess the course of AN in groups comprising individuals selected from
the general population. One of such studies compared adolescents (aged up to 19 years)
selected during a population-based study (20% of whom had never had any contact
with a psychiatrist and only 61% of whom had ever received therapy) and a control
group of healthy adolescents. After 18 years of follow-up, 54% of individuals from
the study group (vs. 88% of the control group) were free from disturbed behaviors
and attitudes with respect to food and shape, while 22% (vs. 2% of the control group)
avoided eating. According to the Morgan–Russell criteria, treatment outcomes were
good, intermediate, and poor in 84%, 10%, and 6% of patients, respectively. Eighteen
years after receiving AN diagnosis 12% of patients still met eating disorder criteria.
AN persisted continually in 2% of patients [15].
In another group of females with the onset of AN during adolescence (57% of
whom received specialist care at some point) weight normalization was observed in
nearly all patients after 7 years, but in 41% of patients symptoms of eating disorder
persisted (mainly bulimia nervosa – BN). After 10 years of follow-up, the study
group did not dier from the control group in terms of body weight or BMI (Body
Mass Index), while 27% of females still met eating disorder criteria. One third of
patients reported running commentary voices with regards to eating-associated
behaviors during the period of AN symptoms, and 11.8% of patients underwent
a psychotic episode during follow-up. After a 10-year follow-up, Morgan–Russell’s
good, intermediate, and poor outcome was exhibited by 49%, 41%, and 10% of
patients, respectively. No deaths were reported during the study. The authors sug-
gested that the outcome in the case of treated patients might be very similar to the
natural outcome of the illness [16].
A population-based study in adolescent females [2] showed a symptomatic im-
provement in 66.8% of patients with AN and 69.1% of patients with atypical AN
(aAN) within 5 years. The outcomes did not dier in the treated and untreated groups.
Most patients achieved clinical improvement within 5 years and, from that point on,
improvement usually progressed until full recovery. The endpoint of follow-up saw
recovery in 70.9% of females with AN and in 76.6% of females with aAN. Five years
after clinical improvement, most patients achieved complete, or nearly complete, psy-
chological recovery (in Eating Disorders Inventory subscales – EDI) and their body
weight approached that of their healthy peers.
Another, similar population-based study of AN outcome, that adopted a relatively
high BMI criterion for good treatment outcome (BMI ≥19 kg/m2), included adolescents
and adults newly diagnosed at the level of primary healthcare. After an average of
approximately 5 years, the study showed good treatment outcomes in 55% of patients
with restrictive type AN and in 57% of patients with bulimic-type AN, while inter-
mediate outcomes were observed in 21% and 17% of patients, respectively, and poor
outcomes in 21% and 32%, respectively [17].
Gabriela Jagielska, Iwona Kacperska
208
The course of anorexia nervosa depending on the age of patients
AN treatment outcomes observed in adult groups, jointly evaluated groups of
adults and adolescents are worse than in adolescents [13, 18–21]. According to meta-
analysis by Steinhausen [13], treatment outcomes in adolescents were better than those
observed in the adolescent and adult populations evaluated jointly (with good treatment
outcomes observed in 57.1±15% vs. 44.2±21.8% of those populations, respectively,
and a chronic course observed in 16.9±7.5% vs. 23.5±14.9% of those populations,
respectively). More recent studies in adolescents with AN indicate that approximately
70%, and even over 80% of adolescents achieve remission [13, 18–21].
A 10-year-long prospective study by Herpertz-Dahlmann et al. [19] in adolescent
females hospitalized due to AN showed a complete recovery rate of 69%. Importantly,
the patients with long time since their improvement did not dier from the control group
in terms of other psychiatric disorders or psychosocial functioning. Another long-term
observational study in women who had received (as adolescents) an average of 12
months of comprehensive therapy at a specialist facility, showed increased remission
rates from 67.6% after 8 years to 85.3% after 16 years, with the achieved remissions
seemingly permanent. Between year 8 and 16 of follow-up, the study also demonstrated
a considerable improvement in the following symptoms: desire to be thin, body dis-
satisfaction and bulimic symptoms, while diculties in recognizing emotions and in
identifying hunger and satiety persisted [21]. A similarly good outcomes (remissions
after 3.5–14.5 years in 82% of cases) were achieved in a group of adolescent females
receiving intensive treatment involving brief periods of hospitalization (when strictly
necessary); with almost all patients receiving further outpatient care [18].
In another long-term study (follow-up of 10–15 years after hospitalization) in
a group of adolescent patients receiving intensive therapeutic program (including
individual, group and family therapy) 75.8% of patients reached complete recovery,
with a partial recovery rate of 10.5%, and chronic AN in 13.7%. Transient symptoms
aggravation was observed in 30% of patients. Following recovery, AN recurrences
were relatively rare. Patients with chronic AN demonstrated considerably poorer work
adjustment, social relations, and life satisfaction after a period of 5 and 10 years in
comparison with healthy individuals and those with partial improvement [22].
In one study AN outcomes and mortality were compared in three age groups
(<18 years old, aged 18–39, and ≥40 years old). The study showed the poorest treat-
ment outcome, social and interpersonal functioning, and quality of life in patients who
were in the 18–39 age group at the beginning of the follow-up. This age group also
showed the highest mortality [23].
Outcome of anorexia nervosa in hospitalized patients
Worse outcomes are observed in patients who required hospitalization.
A review of the literature from the period 2004–2009 showed that treatment out-
comes ranged from optimistic for patients of community health clinic (2% of them
continued to be diagnosed with AN after a 9-year follow-up), to much worse for
209
Outcome, comorbidity and prognosis in anorexia nervosa
inpatients (18% of inpatients met AN criteria after 12 years of follow-up). Authors
summarized that a majority of outpatients with AN are in remission ve or more
years following presentation and conversely, minority of patients with AN treated in
inpatient settings achieve remission irrespective of whether follow-up occurs 2 or 12
years following hospitalization [24].
Sullivan et al. [25] reported that 10% of patients continued to meet criteria of AN
12 years after their rst hospitalization, while 15.7% had aAN. The authors empha-
sized the fact that even females who did not meet the criteria of AN had a relatively
low body weight, cognitive functioning characteristic of AN, and signicantly lower
desired BMI in comparison to those parameters in females from the control group.
It should be emphasized that in the past the outcomes for hospitalized patients were
worse than in the present.
Despite its average follow-up of 20 years, one study that evaluated patients (ado-
lescents and adults with a relatively long duration of the illness and low body weight
before hospitalization) hospitalized in the sixties showed generally poorer treatment
outcomes with good treatment outcomes reported only in 30% of patients, intermediate
outcomes in 32.5% of patients, and poor treatment outcomes in 20% of patients (17.5%
of patients died) [26]. A more recent example of poor outcome of AN was a 21-year-long
prospective study in adolescent and adult females following their rst hospitalization
that was published in 2000. This study showed complete recovery in 50.6% of patients,
improvement in 20.8% of patients, and persistent fully symptomatic AN (according to
DSM-IV criteria) in 10.4% of patients. The mortality rate was 16.7% [27].
After hospital discharge health deterioration is a common phenomenon in AN
patients.
An observational study of patients treated in a hospital setting (undergoing in-
tensive treatment) showed moderate or slight deterioration in their condition within
the rst two years after discharge with respect to the improvement achieved during
hospitalization. However, progressive improvement was observed after 3–12 years
of follow-up [28].
Another study demonstrated good outcomes (according to Morgan–Russell criteria)
after an average of approximately 8 years follow-up in 75% of patients, intermediate
outcomes in 11% of patients, and poor outcomes in 14% of patients treated during
adolescence, despite frequent need for re-hospitalization (35% of the study group).
However, 38% of patients were still concerned about their weight and appearance
(including 62% who continued to limit their food intake) [29].
Chronic course of anorexia nervosa
A number of studies indicate that AN is a potentially chronic disorder both in the
treated and the general populations. According to Strober et al. [22], patients with
adolescent-onset AN who underwent intensive treatment required from 4.8 to 6.6
years to recover. The rates of partial and complete improvement 2 years after hospi-
talization were 10% and 0%, respectively, and 3 years after hospitalization – 21% and
1%, respectively. Improvement rate increased rapidly 4 years after hospital treatment.
Gabriela Jagielska, Iwona Kacperska
210
Partial improvement or complete recovery after 5 years were observed in 54.8% and
17.9% of patients, respectively, and increased in the subsequent years of follow-up (to
73.7% and 58.9%, respectively, after 7 years, and to 84.3% and 72.6%, respectively,
after 10 years). There were also sporadic cases of complete recovery after more than
10 years of follow-up.
The average duration of AN in patients identied in a population-based study of
adolescents was 3.4±2.4 years (range 0.9–14.7 years) [15]. Another, similar study in
the general population showed that the median time to recovery for the restrictive type
of AN was 3.3 years, and binge-eating/purging type of AN – 4.4 years [17]. Likewise,
a study by Råstam et al. [16] showed the mean duration of symptoms to be 3.3 years
(2.7–3.8 years).
Bulimic symptoms in the course of anorexia nervosa
Studies indicate that bulimic symptoms often occur in the course of anorexia
nervosa. Strober et al. [22] demonstrated that approximately 30% of patients with
restrictive type of AN exhibit binge-eating episodes (with an onset after an aver-
age of 24 months of follow-up). This symptom was more common in patients with
a hostile attitude toward their families and in the case of absent parental empathy and
tenderness toward the patient. Eckert et al. [30], who included only female patients
hospitalized for AN, demonstrated that a period of bulimic symptoms occurred in up
to 64% of cases. The high prevalence of bulimic symptoms in AN was also conrmed
in a population-based study with an 18-year follow-up of adolescents (some of them
treatment-naïve), which showed that only 24% of patients had never exhibited bulimic
symptoms [15]. Another study showed 15% of patients exhibiting symptoms of BN
after 20 years of follow-up [26].
Diagnostic crossover to BN among hospitalized adolescent and adult patients with
AN (binge-eating/purging type in 71% of patients) occurred both in the restrictive type
and binge-eating/purging type; however, in the longer perspective, this occurred in
patients initially diagnosed with binge-eating/purging type. After 12 years of follow-
up, crossover from AN to BN persisted only in 9.5% of patients [28].
According to a number of authors, symptoms of bulimia develop within 2–3
years from the onset of AN [16, 22, 25, 30, 31]. Sullivan et al. [25] reported diag-
nostic crossover from AN to BN mainly within the rst 2 years of follow-up, while
development of BN after 5 years from AN onset was considered unlikely. This study
with an average of 12-year follow-up after the rst hospitalization, showed that 54%
of females met criteria for BN at some point. After 12 years, BN criteria were met
by 11.4% of patients.
Contrary to the ndings in the above-mentioned studies, a follow-up of patients
undergoing comprehensive treatment, mainly in outpatient setting, demonstrated
symptoms of bulimia in only 5% after 8 years and in 1.5% of patients after 16 years
[21]. One population-based study showed no conversion to symptoms of bulimia in
patients with a restrictive type of AN (diagnosed in primary care setting) [17].
211
Outcome, comorbidity and prognosis in anorexia nervosa
Mortality in anorexia nervosa
Mortality in AN depends on study group characteristics (general vs. hospitalized
population, disorder severity, diagnostic criteria), sample size, follow-up duration,
separate analysis of AN and other eating disorders, and the type of therapy.
Reported crude death rates (the proportion of deaths in a given population) range
from 0 to 22.5%, while the standardized mortality ratio (SMR), which quanties the
increase in mortality in a study population of the same age, ranges from 0 to 17.8.
The SMR for suicidal death in AN ranges from 23.1 to 56.9 (as cited in: [32]). Ac-
cording to Klump et al. [10] the SMR in AN is the highest of that in all psychiatric
disorders.
Patients who receive treatment during adolescence are at a lower risk of death.
Some studies even suggest no mortality after 10 years of follow-up in patients with
adolescent-onset AN [18, 19, 22, 33]. Moreover, after 10–15 years of follow-up, no
deaths were observed in patients receiving intensive therapeutic support and nutritional
treatment in a hospital setting [22] as well as during an 18-year-long prospective ob-
servational study in patients with AN identied in screening tests of adolescents aged
up to 19 years (despite the fact that most of the patients did not receive therapy and
20% of patients did not receive specialist care) [15].
The crude death rate in a prospective long-term population-based study in Finnish
teenagers, including also untreated cases, was 0.3% per decade [2]. Another popula-
tion-based study in cases of AN newly diagnosed in primary care setting (an average
follow-up of 5 years) reported a 0.7% mortality [17].
An analysis of national Swedish registry including female patients undergoing
treatment during adolescence in the period between 1987 and 1993 demonstrated total
mortality of 1.2%, and AN-related mortality of 0.8%. The SMR was 3 [34]. A similar
low mortality (1%) was observed in a group of patients treated during adolescence
(receiving intensive comprehensive therapy) [21].
Worse outcomes with higher mortality were observed in groups of adults, jointly
analyzed adults and adolescents, and patients requiring treatment in a hospital setting.
For example, after an average of 8-year follow-up, in the group of outpatient females
who were adolescents or adults at the time of qualication for treatment, the mortality
rate was 4.27%, and after excluding non-AN-related deaths SMR was 9.7. A meta-
analysis (including 2,240 patients) conducted by the same authors showed that the
mortality rate in AN was 5.25% and the SMR was 9.7 [35].
A prospective study in adolescents and adults requiring hospitalization (binge-
eating/purging type of AN constituted 71%) showed crude death rate of 4.9% after 2
years, 5.8% after 6 years, and 6.8% after 12 years. The SMR after 12 years was 8.85.
The patients who died had exhibited higher rates of depression, lower body weight,
higher rates of laxative abuse, hypersensitivity in social relations, feeling of inadequacy,
higher rates of introversion, lower self-esteem, and poorer social functioning at the
time of hospital admission [28].
Long-term treatment outcomes of other studies on the results of treatment in severe
cases from tertiary centers or those hospitalized showed high crude death rates of up
Gabriela Jagielska, Iwona Kacperska
212
to 16–18% [15, 26, 27, 36]. For example, one prospective study (average follow-up of
21years after the rst hospitalization) reported an AN-related mortality of 15.6% [27].
The SMR in a large cohort study in patients admitted to tertiary hospitals (treated in the
80s and in the 90s) was as high as 10.5 (with 41% of deaths being due to suicides) [37].
A meta-analysis of 42 studies (average follow-up of 7.6 years – from 1.7 to 33
years, 3,006 subjects) demonstrated crude death rate in AN of 5.9% (0.56% per year
and approximately 5.6% per decade). The SMR in females with AN aged 15–24 was
12 and was two times higher than that in the group of females hospitalized for other
psychiatric reasons. The causes of death were: AN complications, suicides, and un-
known in 54%, 27%, and 19% of cases, respectively. The suicide rates were 200-times
greater than those in the general population [38].
A more recent meta-analysis of studies published in the second half of the 20th
century (5,590 patients) yielded a crude death rate of 5.9±5.7% (range 0–22%).
The mortality rate was 0.9%±2% in the follow-up period of <4 years, 4.9%±4.3% in
the follow-up period of 4–10 years, and 9.4%±8.3% in the follow-up period of >10
years. The SMR ranged from 1.36 to 17.8. The highest SMR value (30) was observed
within the rst year after presentation. The mortality rate in patients treated during
adolescence were lower than those in studies where adolescents and adults were evalu-
ated jointly (1.8%±2.5% and 5.9%±5.7%, respectively) [13].
A meta-analysis of studies published between the year 1996 and 2010 (12,189
patients) indicated that the SMR for AN patients was 5.85. The higher mortality
rates were associated with age, low BMI, comorbidities (aective disorders, history
of suicidal and self-harm behaviors, alcohol abuse) and hospitalization due to other
psychiatric disorders. Few deaths were reported in the younger age group, while many
deaths were reported in those treated at the age of 20 and over. The SMR values in the
youngest patients (age at hospitalization 15–19 years), in those hospitalized at 20–29,
and in those treated at age 30 and over were 3, 10, and 6, respectively [39].
Analysis of 6,009 females hospitalized due to AN at least once in relation to the
cause of death registry in Sweden revealed high SMRs (the SMR for all causes 6.20,
with the SMRs for natural and unnatural causes 4.9 and 8.9, respectively). The highest
SMR (of 650.0) was for AN-related deaths, followed by the SMRs for substance abuse
(18.9) and suicide (13.6). The patients who were older at the time of hospitalization
were at a higher risk of early death (twofold higher in those aged 20–29 years and
fourfold higher in those aged 30–39 years in comparison to that for the age 10–19) and
underwent repeated hospitalizations including psychiatric hospitalizations for other
disorders. Authors emphasized that recent years have seen a decrease in the reported
mortality rates, which may be associated with improved care and therapy, treatment
of psychiatric comorbidities, better realimentation strategies, and a better control of
somatic disorders [32].
The recent ndings of lower mortality rates were conrmed by an analysis of
studies published between 2004 and 2009. In this period, the crude death rate ranged
from 0 to 8%, with the cumulative death rate of 2.8% [24].
213
Outcome, comorbidity and prognosis in anorexia nervosa
Comorbidities in anorexia nervosa
AN is associated with a high comorbidity. Micali et al. [5] indicated that co-
morbidity is a common problem during the acute phase of the illness. A 2-year
follow-up of adolescents aged 14 and 16 meeting the DSM-5 criteria for AN showed
an increased risk of depressive and anxiety disorders. Many studies show a high
risk of developing other psychiatric disorders during the patients’ lifetime, even
after recovery from AN. Additional psychiatric disorders are more likely to occur in
those still aected. Two thirds of patients with a persistent eating disorder and one
third of those who stopped meeting eating disorder criteria had at least one other
psychiatric disorder [18].
A meta-analysis of studies published in the second half of the 20th century [13]
indicated that aective disorders occurred in 24.1%±16.3% (2–67%), neurotic or anxi-
ety disorders in 25.5%±14.9% (4–61%), obsessive-compulsive disorders in 12%±6.4%
(0–23%), substance abuse disorder in 14.6%±10.4% (2–38%), and schizophrenia in
4.6%±5.7% (1–28%) of AN patients. A follow-up in patients with AN showed con-
comitant partially or fully symptomatic eating disorders or another psychiatric disorder
without eating disorders in over 50% of patients. A total of 31%±25.1% (0–76%) of
patients had an obsessive-compulsive personality disorder, 16.6%±19.9% (0–53%) of
patients had a histrionic personality disorder, and 17.4%±16.8% (0–69%) of patients
had other personality disorders including borderline states.
A 12-year follow-up of AN patients by Sullivan et al. [25] showed higher rates of
various psychiatric disorders during their lifetime versus the control group (respec-
tively: depression disorders in 51.4% and 35.7%; alcohol dependence in 27.1% and
10.2%; other dependencies in 30% and 12.2%; anxiety disorders in 60% and 32.7%).
The following disorders were signicantly more common in the aected group than
in the control group: generalized anxiety disorder, separation anxiety disorder, obses-
sive-compulsive disorder, and panic disorder. An 18-year follow-up of patients with
adolescent-onset AN versus the control group also showed the proportion of patients
with at least one additional psychiatric disorder to be over two times higher than in
the control group (39% vs. 16%, respectively) [18].
Sixteen years after the beginning of treatment (when 85.3% of patients no longer
had AN symptoms and 68% of patients — no eating disorders) 37% of patients con-
tinued to have symptoms of various psychiatric disorders (which did not disturb the
patients’ functioning in most cases), including most frequently symptoms of depres-
sion (56%), anxiety disorders (28%), and obsessive-compulsive disorders (12%) [21].
Another study with good treatment results (82% of patients in remission) after
a 3.5–14.5-year follow-up showed at least one axis I psychiatric disorder (according
to DSM-IV) in 41% of patients (most commonly a depressive or anxiety disorder),
with 10% of patients exhibiting symptoms of post-traumatic stress disorder (PTSD).
At the end of the follow-up period, 55% of patients exhibited eating disorders or other
psychiatric disorders [18].
Additional psychiatric disorders are more common in patients with binge-eating/
purging subtype of AN. Fichter et al. [28] conducted a 12-year follow-up of adoles-
Gabriela Jagielska, Iwona Kacperska
214
cents and adults with a history of hospital treatment, including a high rate of patients
with binge-eating/purging type of AN. The study showed the presence of psychiatric
disorders in as many as 76.6% of patients. The respective lifetime prevalence and
end-of-follow-up rates for mood disorders were 63.8% and 20.8%, major depres-
sion – 54.4% and 16.9%, dysthymia – 9.1% and 3.9%, anxiety disorders – 46.3% and
28.6%, panic attacks – 20.8% and 11.7%, obsessive-compulsive disorder – 18.2% and
13%, substance use disorders – 29.9% and 9.1%, drug dependence (mostly involv-
ing prescription drugs) – 16.9% and 2.6%, and psychotic disorders – 1.3% and 0%.
Borderline personality was observed in 9.1% of patients (the article does not state at
which point of observation).
A 9–14-year follow-up conducted in females born between 1968 and 1977 (in
Sweden) and hospitalized during adolescence due to AN showed that 8.7% of them had
persistent chronic psychiatric problems requiring inpatient treatment, while 21.4% of
them received social welfare. These patients had a 5.8-fold higher risk of poor health
(requirement for psychiatric hospitalization and/or death), and a 2.6-fold higher risk
of welfare dependency than that in the general population [34].
Råstam et al. [16] emphasized the fact that approximately 50% of patients suered
from premorbid symptoms of obsessive-compulsive disorder (OCD), particularly,
obsessive-compulsive personality disorder (OCPD) and that depressive symptoms
usually develop during the course of AN. Over one third of patients were diagnosed
with OCD within their lifetime. After 7 years of follow-up in females with adolescent-
onset AN (after almost all patients achieved weight normalization) aective disorders,
OCD, and avoidant personality disorder continued to be overrepresented, and one in
ve patients exhibited an autism spectrum disorder (ASD). After 10 years, the rates
of ASD or OCPD were over two times higher than in the control group, with aective
disorders (particularly depression and dysthymia) and anxiety disorders (particularly
OCD) more common than in the control group. Some authors go as far as to believe
that AN is a particular type of OCD, however, obsessions and compulsions are most
commonly associated with eating, appearance, and compulsive physical exercises,
although the content and severity of symptoms allow to establish an additional diag-
nosis only in some patients (as cited in: [40]).
Polish study in a large group of females treated in a hospital setting conrmed high
rates of OCD symptoms (in 25% of patients), with none of the patients meeting the
diagnostic criteria of OCD. Most of the patients reported the eect of OCD symptoms
on their functioning as low [41].
Prognostic factors
A meta-analysis of studies published in the second half of the 20th century [13]
showed that self-induced vomiting, purgative abuse, bulimic type of AN, chronic
character of symptoms, and obsessive-compulsive personality were associated with
poorer prognosis, whereas histrionic personality was associated with better outcomes.
The analysis was inconclusive as to the eect of the duration of inpatients treatment or
the degree of weight loss. Hyperactivity was found to have no prognostic signicance.
215
Outcome, comorbidity and prognosis in anorexia nervosa
A review of literature from the period 2004–2009 suggested that symptom sever-
ity with the necessity for inpatient treatment and illness duration before treatment
initiation can help predict poorer treatment outcomes [24]. A number of studies indi-
cated that the outcomes in terms of mortality, psychosocial functioning, psychiatric
disorders, and eating-related problem behaviors were better in adolescent-onset AN
than in adult-onset AN [20, 22, 42, 43]. Poor prognosis was reported in patients with
a very early illness onset (age 8–14) [44]. A more favorable prognosis in patients with
adolescent-onset AN and a poor prognosis in patients with childhood-onset AN were
conrmed in a meta-analysis by Steinhausen [13]. The meta-analysis also indicated that
a good parent-child relationship may protect against an unfavorable outcome. A large
population-based study (including some treatment-naïve patients) showed a higher
recovery rate in patients diagnosed before age 19 than in those ≥20 years old [17].
Favorable predictors in AN include early treatment initiation and good social
adjustment in the period prior to AN onset (as cited in: [45]). According to a study by
Wentz et al. [15] obsessive-compulsive personality traits preceding the onset of AN, age
of onset, and premorbid autistic traits are unfavorable prognostic factors in the group
of patients with adolescent-onset AN. Råstam, Gilberg and Wentz [16] reported that
concomitant OCD, OCPD, and ASD have a negative eect on AN outcome. Błachno
et al. [41] demonstrated that the presence of OCD symptoms was associated with the
risk of more rapid weight loss and a longer period necessary for weight restoration.
Strober et al. [22] indicated a greater likelihood of a chronic course of the illness
in patients with exacerbated compulsive exercising toward the end of hospitalization
and poor social relations prior to illness onset. A longer time to recovery was observed
in patients with a hostile attitude toward their families and very severe compulsive
behaviors in everyday activities. Early illness onset (before the age of 12) in this study
was not found to be associated with prediction of complete or partial recovery. Older
age at the beginning of follow-up, alcohol or drug abuse, endocrine disorders, nega-
tive family history for eating disorders were shown to be predictors of poor outcomes
or death [23].
Tozzi et al. [46] presented an interesting approach to factors associated with re-
covery and persistent AN symptoms. After an average of 15 years of AN onset (during
adolescence and adulthood), the most common patient-reported factors contributing
to recovery were supportive nonfamilial relationship or having a partner, maturation
or growing out of the illness, therapy or counseling, while the factors contributing to
persistent illness were the family’s focus on their weight and eating as well as their
own low self-esteem.
Conclusions
Despite the fact that recent studies report improved outcome for patients with
anorexia nervosa, it remains a serious disorder associated with high risk of chronic
course and death. More favorable outcome is observed in adolescent population.
The need for longstanding follow up and frequent psychiatric comorbidity should
be considered in treatment of anorexia nervosa.
Gabriela Jagielska, Iwona Kacperska
216
Future studies should focus on evaluation of dierent therapeutic programs, which
may further improve the eectiveness of treatment of AN.
References
1. Bulik CM, Sullivan PF, Tozzi F, Furberg H, Lichtenstein P, Pedersen NL. Prevalence, heritabililty,
and prospective risk factors for anorexia nervosa. Arch. Gen. Psychiat. 2006; 63(3): 305–312.
2. Keski-Rakhonen A, Hoek HW, Susser ES, Linna MS, Sihvola E, Raevuori A et al. Epidemiol-
ogy and course of anorexia nervosa in the community. Am. J. Psychiat. 2007; 164: 1259–1265.
3. Wade TD, Bergin JL, Tiggemann M, Bulik CM, Fairburn CG. Prevalence and long-term course
of lifetime eating disorders in an adult Australian twin cohort. Aust. NZ J. Psychiat. 2006;
40(2): 121–128.
4. Smink FRE, Hoeken D van, Hoek RW. Epidemiology of eating disorders: Incidence, prevalence
and mortality rates. Curr. Psychiat. Rep. 2012; 14: 406–414.
5. Micali N, Solmi F, Horton NJ, Crossby RD, Edyy KT, Calza JP et al. Adolescent eating disorders
predict psychiatric, high-risk behaviours and weight outcome in young adulthood. J. Am. Acad.
Child Psych. 2015; 54(8): 652–659.
6. Herpertz-Dahlmann B. Adolescent eating disorders: denitions, symptomatology, epidemiology
and comorbidity. Child Adol. Psych. Cl. 2008; 18: 31–47.
7. Lukas AR, Beard CM, O’Fallon WM, Kurland LT. 50-year trends in the incidence of anorexia
nervosa in Rochester, Minn.: a population-based study. Am. J. Psychiat. 1991; 148: 917–922.
8. Hoek HW. Incidence, prevalence and mortality of anorexia nervosa and other eating disorders.
Curr. Opin. Psychiatr. 2006; 19(4): 389–394.
9. Son GE van, Hoeken D van, Bertelds A, Furth EF van, Hoek HW. Time trends in the incidence of
eating disorders: a primary care study in Netherlands. Int. J. Eat. Disorder 2006; 39: 565–569.
10. Klump KL, Bulik CM, Kaye WH, Treasure J, Tyson E. Academy for eating disorders position
paper: eating disorders are serious mental illnesses. Int. J. Eat. Disorder 2009; 42(2): 97–103.
11. Hoek HW, Hoeken D van. Review of the prevalence and incidence of eating disorders. Int. J.
Eat. Disorder. 2003; 34: 383–396.
12. Abbate-Daga G, Amianto F, Delsedime N, De-Bacco C, Fassino S. Resistance to treatment and
change in anorexia nervosa: a clinical overview. BMC Psychiatry 2013; 13: 294–312.
13. Steinhausen HCh. The outcome of anorexia nervosa in the 20th century. Am. J. Psychiat. 2002;
159: 1284–1293.
14. Couturier J, Lock J. What is recovery in adolescent anorexia nervosa. Int. J. Eat. Disorder
2006; 39(7): 550–555.
15. Wentz E, Gillberg IC, Anckarsäter H, Gillberg CH, Råstam M. Adolescent-onset anorexia
nervosa: 18-year outcome. Brit. J. Psychiat. 2009; 194: 168–174.
16. Råstam M, Gillberg C, Wentz E. Outcome of teenage-onset anorexia nervosa in a Swedish
community-based sample. Eur. Child Adoles. Psych. 2003; 12(Suppl. 1): 178–190.
17. Son GE van, Hoeken D van, Furth EF van, Donker GA, Hoek HW. Course and outcome of eating
disorders in a primary care-based cohort. Int. J. Eat. Disorder 2010; 43: 130–138.
18. Halvorsen I, Andersen A, Heyerdahl S. Good outcome of adolescent onset anorexia nervosa
after systematic treatment. Eur. Child Adoles. Psych. 2004; 13: 295–306.
217
Outcome, comorbidity and prognosis in anorexia nervosa
19. Herpertz-Dahlmann B, Müller B, Herpertz S, Heussen N, Hebebrand J, Remschmidt H. Prospec-
tive 10-year follow-up in adolescent anorexia nervosa-course, outcome, psychiatric comorbidity
and psychosocial adaptation. J. Child Psychol. Psychiatry 2001 Jul;42(5): 603-612.
20. Steinhausen HC, Boyadjieva S, Grigorieu-Serbanescu M, Neumärker KJ. The outcome of ado-
lescent eating disorders, nding from an international collaborative study. Eur. Child Adoles.
Psych. 2003; 12: 91–98.
21. Nilsson K, Hägglöf B. Long-term follow-up of adolescent onset anorexia nervosa in Northern
Sweden. Eur. Eat. Disord. Rev. 2005; 13: 89–100.
22. Strober M, Freeman R, Morrell W. The long-term course of severe anorexia nervosa in ado-
lescents; Survival analysis of recovery, relapse, and outcome predictors over 10–15 years in
a prospective study. Int. J. Eat. Disorder 1997; 22(4): 339–360.
23. Ackard DM, Richter S, Egan A, Cronemeyer C. Poor outcome and death among youth, young
adults, and midlife adults with eating disorders: an investigation of risk factors by age AT as-
sessment. Int. J. Eat. Disorder 2014; 47(7): 825–835.
24. Keel PK, Brown TA. Update on course and outcome in eating disorders. Int. J. Eat. Disorder
2010; 43: 195–204.
25. Sullivan PF, Bulik CM, Fear JL, Pickering A. Outcome of anorexia nervosa: a case-control
study. Am. J. Psychiat. 1998; 155(7): 939–946.
26. Ratnasuryia RH, Eisler I, Szmukler GI, Russell GF. Anorexia nervosa: outcome and prognostic
factors after 20 years. Brit. J. Psychiat. 1991; 158: 495–502.
27. Zipfel S, Löwe B, Reas DL, Deter H-CH, Herzog W. Long-term prognosis in anorexia nervosa:
lessons from 21-year follow-up study. Lancet 2000; 355: 721–722.
28. Fichter MM, Quadieg N, Hedlund S. Twelve-year course and outcome predictors of anorexia
nervosa. Int. J. Eat. Disorder 2006; 39: 87–100.
29. Komender J, Popielarska A, Tomaszewicz-Libudzic C, Jagielska G, Brzozowska A, Wolańczyk T.
Odległe wyniki leczenia dorastających chorych na jadłowstręt psychiczny. Psychiatr. Pol. 1998;
32(6): 759–769.
30. Eckert ED, Halmi KA, March P, Grove W, Crosby R. Ten-year follow-up of anorexia nervosa:
clinical course and outcome. Psychol. Med. 1995; 25(1): 143–156.
31. Bulik CM. Eating disorders in adolescents and young adults. Child Adol. Psych. Cl. 2002; 11:
201–218.
32. Papadopoulos FS, Ekbom, Brandt L, Ekselius L. Excess mortality, causes of death and prognostic
factors in anorexia nervosa. Brit. J. Psychiat. 2009; 194: 10–17.
33. Wentz E, Nilsson E, Gillberg C, Gillberg IC, Råstam M. Ten-year follow-up of adolescent-onset
anorexia nervosa: psychiatric disorders and overall functioning scales. J. Child Psychol. Psych.
2001; 42(5): 613–622.
34. Hjern A, Lindberg L, Lindblad F. Outcome and prognostic factors for adolescent female inpa-
tients with anorexia nervosa: 9 – to 14-year follow-up. Brit. J. Psychiat. 2006; 189: 428–432.
35. Signorini A, De Filippo E, Panico S, De Caprio C, Pasanisi F, Contaldo F. Long-term mortality
in anorexia nervosa: a report after 8-year follow-up and review of the most recent literature.
Eur. J. Clin. Nutr. 2007; 61: 119–122.
36. Theander S. Outcome and prognosis in anorexia nervosa and bulimia: some results of previous inves-
tigations, compared with those of a Swedish long-term study. J. Psychiatr. Res. 1985; 19: 493–508.
37. Birmingham CL, Su J, Hlynsky JA, Goldner EM, Gao M. The mortality rate from anorexia
nervosa. Int. J. Eat. Disorder 2005; 38: 143–146.
Gabriela Jagielska, Iwona Kacperska
218
38. Sullivan PF. Mortality in anorexia nervosa. Am. J. Psychiat. 1995; 152: 1073–1074.
39. Arcelus JA, Mitchell AJ, Wales J, Nielsen S. Mortality rates in patients with anorexia nervosa and
other eating disorders: a meta-analysis of 36 studies. Arch. Gen. Psychiat. 2011; 68(7): 724–731.
40. Błachno M, Bryńska A. Comorbidity and characteristic of obsessive-compulsive symptoms in
anorexia nervosa. Psychiatr. Pol. 2012; 46(6): 1019–1028.
41. Błachno M, Bryńska A, Tomaszewicz-Libudzic C, Jagielska G, Srebnicki T, Wolańczyk T.
The inuence of obsessive compulsive symptoms on the course of anorexia nervosa. Psychiatr.
Pol. 2014; 48(3): 429–439.
42. Hsu LKC. Outcome and early onset anorexia nervosa. What do we know? J. Youth Adolesc.
1996; 25: 563–568.
43. Theander S. Anorexia nervosa with early onset: Selection, gender, outcome, and results of
longterm follow-up study. J. Youth Adolescence 1996; 25: 419–430.
44. Lask B, Bryant-Waugh B. Early-onset anorexia nervosa and related eating disorders. J. Child
Psychol. Psych. 1992; 33: 281–300.
45. Cygankiewicz P, Solecka D, Pilecki MW, Józek B. Predyktory poprawy objawowej w zaburze-
niach odżywiania się. Analiza wstępna. Psychiatr. Pol. 2012; 46(2): 201–212.
46. Tozzi F, Sullivan PF, Fear JL, McKenzie J, Bulik CM. Causes and recovery in anorexia nervosa:
the patient’s perspective. Int. J. Eat. Disorder 2003; 33: 143–154.
Address: Gabriela Jagielska
Department of Child Psychiatry
Medical University of Warsaw
02-691 Warszawa, Żwirki i Wigury Street 63A
