Short summary

Anorexia nervosa is a syndrome, i.e. collections of symptoms, that it is not defined by etiology. The severity varies from mild subclinical cases to chronic enduring and fatal cases. The disease profoundly distresses the whole family which may request hospitalization of the patient. Prolonged hospitalization (in some cases, several years) has negative consequences itself. No overall knowledge of trends of treatment activities of anorexia nervosa in Denmark is available. This study is a multidimensional study of the treatment career across long-term course related to age of onset, psychiatric and somatic comorbidity, socioeconomic and demographic characteristics of a complete Danish cohort of patients with AN based on national registry data from 1995 to 2015.

Rationale

Background: Anorexia nervosa (AN) is a syndrome, i.e. collections of symptoms, that it is not defined by etiology. It is characterized by disturbed body image, ego-syntonic neglect, ambivalence, self-starvation, loss of body weight, obsessive thoughts of food, ritualistic patterns of food intake, elevated physical activity, depression, anxiety and emotional rigidity. The severity varies from mild subclinical cases to chronic enduring and fatal cases. The diagnostic criteria were revised 2013 for the Fifth Edition of the Diagnostic and Statistical Manual leading to markedly increased diagnostic heterogeneity and 30-50% increase in lifetime prevalence which is estimated to 1-4%. The severe cases are intractable, and there is no evidence that the prognosis has improved throughout the 20th century. The mortality rate remains high, and less than half of patients recover from the disease. Prolonged hospitalization (in some cases, several years) has negative consequences. In addition to the resource usage, hospitalizations involve profound implications of the patient's daily life with loss of social networks and competencies. Therefore, there is a consensus that AN should, as far as possible, be treated in an outpatient setting. However, AN profoundly distresses the family and relatives, which in many cases strongly request hospitalization of the patient.

The incidence of AN are changing over time along with changing diagnostic criteria, attention and request of treatment. As in other mental disorders, comorbidity plays a role in the manifestation of the disease during various stages of its course. After hospitalization for AN the occurrence of lifetime anxious or depressive comorbidities are very high. However, no overall knowledge of trends of treatment activities of AN in Denmark is available.

Purpose: To provide nationwide overview and transparency of time trends in treatment activities, age of onset and comorbidity in AN.

Methods: This study is a multidimensional study of the long-term course of a complete Danish cohort of patients with AN based on national registry data. The incidences and treatment activities in terms of hospitalization, outpatient contacts, medication and social health services for AN from the year 1995 to 2015 are related to psychiatric and somatic comorbidity, socioeconomic and demographic characteristics, and outcome. In the entire set of analyses, a specific focus is directed on the effect of age at manifestation of AN by considering three separate groups, namely subjects with pre-adolescent/childhood, adolescent, and adult age at the time of first diagnosis of the disorder. Furthermore, there is another major focus on the duration of follow-up by considering various subgroups with different follow-up periods. Data is extracted from the National Registers. The study is a matched case-control follow-up study. The statistical analyzes will be performed with robust cluster estimation methods and generalized linear mixed models that allow analysis of non-independent data. Multiple imputation will be used for missing or inconsistent information about run times.

Description of the cohort

A national cohort of all patients diagnosed for the first time with an eating disorder (ED) according to the International Statistical Classification of Diseases and Related Health Problems, 10th edition (ICD-10), code F50 in the Danish Patient Register (DPR) covering the period 1995 -2015. Using a matched case-control design at a 1:5 ratio, controls were identified in the Danish Civil Registry (DCR). The controls had to be alive at follow-up and were born at the same time as the cases, had the same sex, were living in the same community at the time of the first F50 diagnosis of the cases, and did not have any ED diagnosis themselves at any time of the observation period. The age range at first time diagnosis was set at a minimum of eight years and the empirical maximum which was 32 years. The definite sample of patients with AN (F 50.0) amounted to N=6077 and there were N=30 038 controls. According to the age at first diagnosis of AN, the patient sample was divided into three subgroups, namely, 8-13, 14-17, and 18-32 years at diagnosis. 

Data and biological material

Danish nationwide register data.

Collaborating researchers and departments

Center for Eating Disorders, Odense University Hospital & Psychiatric Services in the Region of Southern Denmark. Elite Research Centre for medical endocrinology, Institute of Clinical Research, University of Southern Denmark

• Professor René Klinkby Støving, PhD

Department of Child and Adolescent Psychiatry, Psychiatric University Clinic Zurich, Zürich, Switzerland

• Professor Hans-Christoph Steinhausen, PhD, DMSc

Child and Adolescent Psychiatric Research Unit, Psychiatry in the Region of Southern Denmark

• Professor Niels Bilenberg, PhD

Center for Eating Disorders, Odense University Hospital & Psychiatric Services in the Region of Southern Denmark, 

• Senior researcher Laura Al-Dakhiel Winkler, PhD

Steno Diabetes Center Odense, Odense University Hospital

• Michael Ejnar Røder, DMSc

Child and adolescent psychiatric research unit, Central Denmark Region

• Psychologist, senior researcher Loa Clausen, PhD

Child and Adolescent Psychiatric Research Unit, Psychiatry in the Region of Southern Denmark

• Kirsten Hørder

Institute for Epidemiology, Biostatistics & Biodemography, University of Southern Denmark

• Associate professor Pia Veldt Larsen