>> The analyzes are based on the anonymized research database with around 1.75 million insured people from AOK Nordost. The results presented capture the development between the years 2010 and 2016. All AOK Nordost insured persons between the ages of 6 and 54 who were fully insured within the year under consideration and lived in the three federal states of Berlin, Brandenburg and Mecklenburg-Western Pomerania were included included in the analysis. In 2010, the study population comprised around 722,000 insured people. By 2016, it had increased to around 746,000 insured people (see Table 1). All persons are considered to have an eating disorder if at least one confirmed outpatient or one inpatient main or secondary diagnosis was reported with an ICD code of type F50.x (F50.- Eating Disorders) in the year of analysis.
For the present study, subcategories such as anorexia and bulimia were not analyzed separately, but rather the ICD-10 group of eating disorders was considered as one category. In 2016, anorexia and bulimia (including the atypical forms) only accounted for a small proportion of eating disorder diagnoses among those insured by AOK Nordost in 2016: 15% among minors and 28% among adults. On the other hand, over 60% (adults) and over 80% (children and adolescents) of the diagnoses fell into the unspecific subcategories “unspecified eating disorders” and “other eating disorders”. This is a typical picture for clinical practice (Fairburn/Bohn 2005). The large number of non-specific diagnoses is due, on the one hand, to the fact that the first point of contact for those affected is usually family doctors or pediatricians who do not carry out more specific diagnostics. On the other hand, it is not uncommon in everyday life that not all diagnostic criteria are met (Fairburn/Bohn 2005). Nevertheless, these affected people also have a clinically relevant disorder of eating behavior, which can be associated with similarly serious physical, mental and social impairments as the “classic” diagnoses (Herpertz-Dahlmann et al. 2008; Swanson et al. 2011). If they were left out of the prevalence analysis, the epidemiological significance of eating disorders would be significantly underestimated.
Eating disorders – status quo and
Development trends
In 2010, 3,498 (0.48%) of 6- to 54-year-old insured persons in the Northeast region were diagnosed with a psychogenic eating disorder. Just 6 years later, in 2016, there were already 6,133 (0.82%) people affected (see Fig. 1). This corresponds to an increase of 71%. However, there are sometimes clear differences between the age groups and genders.

Prevalence by age and gender
As Figure 2 shows, in 2016, women who were insured by AOK Nordost were significantly more likely to be diagnosed with an eating disorder. Overall, the prevalence among them was 1.32%, compared to 0.37% among men - meaning that a clinically significant eating problem was documented by a doctor more than three times as often among female insured persons as among male insured persons. In Figure 2, gender-specific changes across age groups are also visible. From the age of 13 onwards, the prevalence among women initially increased significantly, peaked between the ages of 18 and 34 and then fell again. Among men, however, the proportion of eating disorder diagnoses decreases between the ages of 13 and 34 and then increases again. However, the proportion of diagnosed eating disorders in all age groups - with the exception of 6 to 12 year olds - is significantly higher among women than among men.
In contrast to the peak illness reported in the literature between around 15 and 25 years (Holtkamp/Herpertz-Dahlmann 2005; de Zwaan 2002), this is around ten years higher among women insured under AOK Nordost. It is possible that in the 30- to 34-year-old age group, symptoms that are still pronounced are associated with a higher level of insight into the illness and a greater willingness to seek medical help. It is also conceivable that the long-term physical consequences of eating disorders become more noticeable from this age onwards, so that those affected ultimately have to see a doctor.
What is also striking in both sexes is the still substantial prevalence in middle adulthood (see Figure 2). These age groups also show the greatest increases in prevalence when looking at the changes in the last 6 years (see Figure 3): Among 35 to 54 year olds, the proportion of diagnosed eating disorders increased by 0.37% to 0.50% between 2010 and 2016 . This corresponds to an increase by a factor of two to three. However, among the classic risk age groups of 13 to 17 year olds and 18 to 24 year olds, the increase was only 0.31% and 0.35% respectively.
Eating disorders are no longer just an issue for young people among those insured by AOK Nordost. Some of these could be chronic cases that are now also being diagnosed by doctors in older age groups due to increased sensitivity and decreased stigma regarding eating disorders in recent decades. This would clearly show how important early detection and intervention are.

Prevalence by region and community type
Routine data makes it possible to evaluate prevalences broken down according to the regional origin of the insured. The regional differences determined in this way can then be used to specifically improve the supply situation in particularly hard-hit regions. In order to ensure comparability of the regions and years of analysis, the analysis groups were directly age and gender standardized to the standard population of Germany in 2011 between the ages of 6 and 54 (Destatis 2014).
As Figure 4 illustrates, Brandenburg and Mecklenburg-Western Pomerania were about the same in terms of diagnosed eating disorders in both 2010 and 2016. In Brandenburg these rose from 0.40% to 0.61% in the period under consideration, and in Mecklenburg-Western Pomerania from 0.43% to 0.59%. In 2010, the prevalence in Berlin (0.61%) was approximately 1.5 times higher than in the other two federal states. In 2016, the proportion in the federal capital rose by 0.5 percentage points to 1.10% - this means that the prevalence rate in Berlin is currently almost twice as high as in Brandenburg and Mecklenburg-Western Pomerania.
Berlin therefore occupies a special position in the northeast area. There may be a connection between the degree of urbanization and the number of diagnosed eating disorders, as Berlin is the only one of the three federal states considered that is completely urban. In order to examine this factor in more detail, the municipalities in the Northeast region were further divided into rural, semi-urban and urban municipalities based on the Eurostat municipality types and compared for the analysis years 2010 and 2016 (see Fig. 5).
In fact, especially for 2016, the prevalence tends to be lower the more rural the region under consideration (see Fig. 5). In semi-urban areas the probability of an eating disorder is increased by a factor of 1.3 compared to rural areas, and in urban areas even by a factor of 2.1. The higher risk in urban communities results not insignificantly from the high prevalence of eating disorder diagnoses in Berlin (see Fig. 4) and its high proportion of the urban population. However, if you only look at Brandenburg and Mecklenburg-Western Pomerania, the risk in urban areas is still 1.5 times higher than in rural areas.
An urban-rural divide is also evident at the level of Berlin's urban districts, independent cities and district regions (see Figure 6). In addition to the districts of Berlin, the cities of Schwerin, Cottbus, Rostock, Frankfurt (Oder) and Potsdam are among the areas with the highest prevalence rates (0.72% to 0.92%). The only non-purely urban region that was also in this area was the Oberhavel district (0.85%). This is followed by the districts of Barnim (0.70%), Dahme-Spreewald (0.70%) and Elbe-Elster (0.69%). Among the districts, they are among the regions with the highest prevalence, with three of these four districts bordering directly on Berlin and being spatially intertwined with the city in the area close to Berlin. Seen this way, it is not surprising that the districts with the highest prevalence are almost all close to Berlin.

Utilization of treatments
The willingness to seek therapy is a major hurdle for those affected by an eating disorder. For them, starting treatment means admitting their own problematic behavior, disclosing it to others and dealing with the underlying reasons. Studies show that younger patients have a higher chance of being cured or improving their symptoms, whereas a long-term illness leads to a worse prognosis (Steinhausen 2002). It is therefore important to treat those affected as early as possible. As the method of first choice, the S3 guidelines recommend treatment within the framework of the guideline psychotherapies (cognitive behavioral therapy, depth psychology-based psychotherapy and analytical psychotherapy) in both outpatient and inpatient settings and, in the case of children and adolescents, involving the family. Additional components should be offered in the inpatient area, such as nutritional, physical and occupational therapy.
In order to determine the willingness to treat a diagnosed eating disorder among AOK Nordost insured persons, incident confirmed outpatient and inpatient main or secondary diagnoses with an ICD code of type F50.x (F50.- Eating Disorders) from the years 2012 to 2014 were first identified. A diagnosis was classified as incident if the insured person had not received any other outpatient or inpatient main or secondary diagnosis in the area of eating disorders in a period of eight quarters prior to the quarter in which the diagnosis was made.
Then, starting from the quarter of the initial diagnosis, it was determined for a subsequent period of twelve quarters (including the quarter of the initial diagnosis) whether treatment was used. Treatments were operationalized in the outpatient area using the fee schedule items (GOPs) and in the inpatient area using the operation and procedure codes (OPS) and each had to be associated with an eating disorder diagnosis. In addition, the analysis only took into account insured persons who had been insured with AOK Nordost without interruption during the 20 quarters under consideration (eight quarters before and twelve quarters with or after the initial diagnosis).
The low treatment rate for eating disorder problems described in the literature (Hoek 2006) is also confirmed among those insured by AOK Nordost: only one in eleven of the initial diagnoses made between 2012 and 2014 was billed for treatment in a subsequent period of three years. The proportion of those who seek treatment decreases with the time that has passed since the initial diagnosis (see Figure 7). For the outpatient area, this analysis only takes into account the guideline psychotherapies that require application, in which a therapeutic unit (individual or group therapy) lasts 50 minutes. If psychotherapeutic treatment services that do not require an application are also included, one in six affected people can be treated. Given the high risk of chronicity and the long-term consequences associated with an eating disorder, the low treatment rate is worrying. It makes it clear how important it is to look for approaches to increase willingness to therapy - as early as possible - and how long waiting times for therapy places can be avoided.
Conclusions and measures
The prevalence of psychogenic eating disorders increased among AOK Nordost insured persons between the ages of 6 and 54 - between 2010 and 2016 alone it increased by 71%. At the same time, only a few patients seek treatment. Little is known about how high the spontaneous remission rate is for eating disorders. However, for those affected who start therapy, the disease usually lasts several years and is characterized by numerous setbacks (DGPM/DKPM 2010). Those affected without treatment may only be noticed years later in view of the long-term consequences of the eating disorder in the health system - this would be a conceivable explanation for the fact that in 2016 there were still substantial prevalence rates beyond the "classic" risk age groups. Eating disorders have significant negative consequences for both the individual and society. Given the upward trend in the number of cases, this burden is likely to increase in the future, so there is an urgent need for action.
In order to prevent initial symptoms from developing into a manifest eating disorder or even becoming chronic, people at risk must be identified early. On the other hand, rapid intervention is then necessary. The AOK Nordost has addressed both points and, together with experts such as the professional association of pediatricians and adolescent doctors, the Dick & Dünn advice center and the startup Jourvie (startup with an app of the same name to support the treatment of eating disorders), is the first statutory health insurance company to offer a care program for early detection and early treatment of psychogenic eating disorders in children and adolescents (7-17 years). This has been implemented in the Berlin region since January 2018. <<

• Underweight in AN, measured using the body mass index in adults or age-dependent percentile curves in children. BED can be associated with obesity.
• Thoughts and feelings influenced by weight and figure: How much and what can be eaten? These thoughts influence actions, resulting in binge eating or compensatory behavior (“purging”).
• Restrictions on calorie intake
• Binge eating in which sufferers lose control over what and how much they eat (BN and BED)
• Compensatory (purging) behavior includes measures that serve to get rid of ingested calories or fluids as quickly as possible. This includes self-induced vomiting as well as excessive exercise, the abuse of thyroid hormones to increase the basal metabolic rate, laxatives or diuretics. The so-called insulin purging, in which patients with type I diabetes mellitus deliberately do not inject themselves with insulin in order to prevent glucose absorption from food, is also considered a compensatory behavior.

ICD codes of type F50.x (F50.- Eating Disorders) include the following ICD-10 codes:
• F50.0 Anorexia nervosa
• F50.1 Atypical anorexia nervosa
• F50.2 Bulimia nervosa
• F50.3 Atypical bulimia nervosa
• F50.4 Binge eating in other psychological disorders
• F50.5 Vomiting in other mental disorders
• F50.8 Other eating disorders
• F50.9 Eating disorders, unspecified