An eating disorder is a mental illness in which the sufferer’s daily life is dominated by thoughts and behaviour associated with food, weight and the body (
1). The most common symptoms of an eating disorder are intake of too little food (restrictive food intake), self-induced vomiting, over-exercising and use of diuretic or weight-loss drugs or laxatives ( 2). This may have a serious effect on the body and can result in death ( 3).
The complexity inherent in the illness is illustrated by the view of the Outpatient Clinic for Eating Disorders of what an eating disorder is: «In an attempt to repair and cope with life, food and the body become a self-regulating factor, a field of control and language itself. The illness is characterised by feelings of self-hate, shame, ambivalence and a lack of identity» (
4). This definition reveals factors specific to the illness that may affect communication at the GP surgery.
The illness generally develops in adolescence and early adulthood (
1). The patient often has a combination of symptoms of an eating disorder, usually varying from day to day or from one period to another ( 1, 5).
Eating disorders are classified into eight different diagnoses in ICD-10. Weight, weight change and symptoms of an eating disorder are the main diagnostic criteria (
6). A BMI of less than 17.5 m/kg² is a decisive indication of anorexia. The variation in symptoms and change in body weight mean that the patient may satisfy criteria for different eating disorder diagnoses during the period of illness ( 5).
A common diagnostic code has been selected for ICPC-2: P86 anorexia nervosa/bulimia (
7). The explanatory text describes conscious weight-loss for anorexia and repeated episodes of binge eating, followed by vomiting or use of laxatives, for bulimia.
The following ailments are typical consequences of the symptoms of eating disorders: lethargy, sleeplessness, dizziness, fainting, difficulties in concentrating, headache, muscle pains, disruptions of the menstrual cycle, oedema, and abdominal and digestive problems (
2). Reduced growth in height, reduced bone density, injuries due to over-exercising, fractures due to osteoporosis, renal failure and dental erosion are examples of partially irreversible complications ( 2).
There is approximately twice as great a risk of mortality for bulimia and six times as great a risk of mortality for anorexia as in the peer age group who do not have the illness (
3). Out of five deaths from anorexia, four have somatic causes whereas one is a suicide ( 3). Cardiovascular mortality is most frequent ( 8). Risk factors for mortality are severe underweight, long-term illness, chronic hypokalaemia, changes in ECG and chronically low serum albumin.
Patients with eating disorders visit their general practitioner (GP) more frequently than their peers in the same age group, but often present with somatic and other mental conditions without the eating disorder necessarily being discovered (
9 – 11). In particular, patients with bulimia need doctors who actively ask questions to begin a conversation about the eating disorder ( 11). A US study showed that two out of three patients with bulimia were never asked about their eating pattern by healthcare workers ( 10). Half stated that they would not answer truthfully if they were asked.
The purpose of this study was to investigate whether patients with serious eating disorders and their GPs discuss the symptoms of the eating disorder and the ailments that the patients have. Furthermore, clinical examinations of the eating disorder by the GP were surveyed, as well as the extent to which the topic of serious complications is broached.
Material and method
The study was conducted at two adult departments: the Outpatient Clinic for Eating Disorders, Department for Personality Psychiatry, Oslo University Hospital, and the Regional Centre for Eating Disorders, Nordland Hospital, Bodø (24-hour emergency unit and outpatient clinic). Almost all the patients who are treated at these departments have one of the following diagnoses (ICD-10): F.50.0 anorexia nervosa, F.50.1 atypical anorexia nervosa, F.50.2 bulimia nervosa or F.50.3 atypical bulimia nervosa.
Patients over the age of 18 years at the start of their treatment were requested to respond anonymously to a questionnaire on their use and experience of their GP. The data collection period was from 1 August 2009 (1 January 2010 for Bodø) to 30 June 2011. Five patients who were referred by the authors were not included for reasons of habilitation.
We developed the questionnaire based on clinical experience from the GP surgery and the specialist unit. It was tested on 45 patients in a pilot study at the Outpatient Clinic for Eating Disorders in Oslo in 2008. The questions were edited in line with the experience gained from this.
The questionnaire contained 99 questions in total. Altogether 54 were related to the presence of symptoms of an eating disorder and the ailments that could result from the eating disorder, the communication with the GP about this, as well as prevention of complications. Examples: «Have you significantly limited your food intake for the purpose of losing weight during your period of illness?» and «During your period of illness, have you had persistent problems with any of the following physical symptoms (abdominal and digestive problems, difficulties in concentrating, muscle pains, repeated and pronounced headache, problems with sleep, missing menstrual periods/changes in menstruation, lethargy and/or exhaustion, dizziness and/or fainting, or swelling of the body)?». There were only yes/no alternatives to these questions. The questionnaire has been published in full as an appendix (in Norwegian) to our second article in this issue of the Journal of the Norwegian Medical Association (
Since we wished to identify patients who had more than isolated consultations with different GPs, we introduced the definition of an own GP who is referred to as «your doctor» based on whether patients responded yes to the following question: «Do you, as of today, have a particular GP whom you have seem
more than three times and whom you generally use when you need medical services?»
The data were analysed using a chi-square test, Fisher’s exact test and a one-way ANOVA test with SPSS version 20.
Written consent was obtained from the participants, and the study has been approved by the Regional Ethics Committee.
Altogether 140 patients commenced treatment at the two specialist units in the study period (Fig. 1). Eight of these were not given the questionnaire due to an oversight or because they were too ill. One patient declined to participate and five were not included because the lead author had referred them. A total of 126 received the questionnaire and 114 (90 %) responded. Altogether 76 outpatients commenced outpatient treatment in Oslo (67 %) and 38 (33 %) in Bodø (of whom 20 had been admitted).
Figure 1 Overview of the 140 patients who commenced treatment in the period
Altogether 91 (80 %) of the respondents were patients who had their own GP according to the study’s definition and had spoken to him/her about their eating disorder in the past year. Three patients (3 %) were men. A total of 51 (56 %) were aged between 18 and 25 years.
Altogether 20 patients (22 %) had never previously undergone treatment for an eating disorder. Thirty-one (34 %) had only had outpatient treatment, whereas 39 (43 %) had had one or more admissions to a psychiatric ward, specialist unit for eating disorders or department of general medicine in addition to any outpatient treatment. A total of 24 (26 %) had been admitted to a medical ward. One patient declined to answer the question.
All the patients gave their height when asked about this on the form. Altogether 86 (91 %) gave their current weight, 87 (92 %) their lowest weight and 86 (91 %) their highest weight as an adult. Average BMI at start of treatment was 20.1 kg/m² (SD 3.6 kg/m², median 19.8 kg/m²). A total of 21 patients (24 %) were severely underweight (BMI < 17.5 kg/m²), and 57 (65 %) had been severely underweight at least once in adulthood. Average weight change in adulthood was 19.6 kg (SD 10.6 kg, median 16.0 kg).
Altogether 29 patients (32 %) had spoken to their GP about their illness 1 – 2 times in the past year, 37 (41 %) 3 – 6 times and 25 (27 %) more than six times.
Symptoms and communication about these
Table 1 shows that the majority of the patients (92 %) had restricted their food intake in the past year, had over-exercised (85 %) and/or had had episodes of vomiting (79 %). 41 % had taken laxatives, 20 % weight-loss drugs and 18 % diuretics. Five patients had had all seven of the symptoms of an eating disorder, and one had only one symptom. Two patients were not included in this analysis because they answered no to all the questions about symptoms of an eating disorder.
Table 1 Number of patients (per cent) who reported various symptoms of an eating disorder in the past year¹ and the extent to which they had spoken to their GP about these symptoms in the past year
With eating disorder symptom in the past year
Communicated about eating disorder symptom
Symptom of an eating disorder
Restrictive food intake
Two out of three patients with vomiting and restrictive food intake had spoken to their GP about the symptoms of their eating disorder. Every third patient had spoken about over-exercising and use of laxatives. None had spoken about their use of diuretics or weight-loss drugs.
A restrictive food intake was equally common in patients who were not severely underweight as in those who were. The severely underweight had spoken to their GP about their restrictive food intake to a greater extent (90 % versus 57 %, p = 0.01).
Altogether 85 % of the patients who were not severely underweight had had episodes of vomiting in the past year, whereas 57 % of the severely underweight had vomited (p = 0.007). Every third patient had not spoken to their GP about this (with no difference between the underweight and the others).
The extent to which the patient had previously received treatment yielded no difference in the prevalence of eating disorder symptoms or in communication about these at the GP surgery.
Altogether 84 patients (92 %) reported that a type of conversation in which the GP asked specifically about the various symptoms of their eating disorder would function best in eliciting the extent of these symptoms. A total of 17 of these patients stated that they would not have managed to add anything beyond what the GP asked about specifically. Six of the patients (7 %) would have supplied complete information without the GP asking questions.
Ailments and communication about ailments
The majority of the patients, irrespective of their weight, had persistent ailments that could be associated with their eating disorder (Table 2). At least half of the patients had spoken to their GP about a possible association between the eating disorder and these ailments.
Table 2 Number of patients (per cent) with long-term ailments that could be linked to the eating disorder, and whether these patients had spoken to their GP about a possible association between the ailment and the eating disorder
Patients who have the ailment
Communicated about a possible association between the ailment and the eating disorder
Difficulties with concentration
Injuries from exercising
Half of the patients were weighed at the GP surgery, and four out of five had blood tests taken (Table 3). Altogether 26 patients (29 %) were weighed several times. The severely underweight were weighed more often (one or more times) than those who were not severely underweight (81 % versus 45 %, p = 0.004). Those who felt that they were supported by their GP were weighed to a greater extent than those who did not feel that they received support (69 % versus 42 %, p = 0.01). Those who were satisfied with how they were received by their GP were weighed to a greater extent (63 % versus 35 %, p = 0.02). Whether the patient had previously entered treatment had no influence on the frequency of testing.
Table 3 Clinical examinations undertaken in the past year and whether possible serious complications and prevention of these have ever been spoken about, given as number of patients (per cent), n = 91
Clinical examinations conducted at the GP surgery in the past year
Blood pressure and pulse¹
Conversation about serious complications
Conversation about preventive measures
Signs that may indicate somatic exacerbation¹
Vitamin and mineral supplements¹
Conversation about serious complications
A maximum of every third patient had spoken to their GP about complications or measures to prevent these (Table 3). Those who had spoken to their GP about vomiting had spoken about electrolyte imbalance more often than those who had not spoken to their GP about vomiting (49 % versus 14 %, p = 0.005). The severely underweight had not spoken about osteoporosis to any greater extent than others.
The majority of the patients had had a restrictive food intake and episodes of vomiting in the past year – irrespective of body weight. This finding is recognised by the clinic, where it is observed that the patients of normal weight also permit themselves too little food, except for in the periods in which they eat and vomit. The energy in the food that is absorbed during the episodes of vomiting may be so substantial that it can compensate for the otherwise restricted food intake of the patient (
13). Consequently their weight can be at a normal level. The food cravings and hunger that occur due to the meagre food intake create a trigger for episodes of (binge) eating followed by vomiting. An important change would be to increase food intake in order to then reduce or avoid episodes of (binge) eating and vomiting. This does not thereby need to result in the patient gaining weight. Precisely this correlation is experienced as alien and creates insecurity.
It is worrying that two out of five patients who are not severely underweight have refrained from speaking to their GP about their low food intake. We interpret this as a need for the GP to ask more specifically about what the patient actually eats. This should then be compared to what the body uses, viewed in light of the level of activity. If the patient needs to gain weight, it must be clarified how much additional food is actually needed for weight to increase. Based on experience, weighing should be carried out in the most secure surroundings possible, particularly in the period in which the patient dares to change his/her intake. It is our experience that weighing can be helpful in reducing the illusion or the perception of body weight changing more than it actually does.
Fewer than half of the patients had spoken with their GP about over-exercising and binge eating. Every third patient tells about use of laxatives, none about diuretics or weight-loss drugs. At the clinic we observe that the above-mentioned drugs are often used in much higher doses than the recommended daily dose for those who are in need of them because of various illnesses.
The fact that not all symptoms of an eating disorder are discussed can result in the GP being given an insufficient basis to assess the cause of the symptoms that the patients present with, which clinical tests should be taken, how frequently the clinical examinations should be repeated and the measures that should be implemented.
In this study half of the patients have spoken about the association between their eating disorder and the ailments they have. This is a surprisingly high proportion. Other studies show that patients go to their GP with corresponding ailments that are not interpreted in the context of the eating disorder (
9 – 11). This difference can presumably be attributed to the fact that all the patients described in this article have a GP who already knows about their eating disorder, in addition to the fact that the patient and the GP actually talk about the illness. Thus these patients and doctors have laid the foundations for being able to conduct a respectful conversation about the bodily aspects of the illness. The patients’ clear statements that the GP must ask directly about each individual symptom of an eating disorder in order for these to be revealed, is therefore a concrete piece of advice for the GPs. It is to be hoped that this advice can also help other patients and their GPs to come to an earlier realisation that the patient is suffering from an eating disorder.
Half of the patients are not weighed when visiting their GP, and we recognise at the clinic that patients may resist being weighed. However, the severely underweight were generally weighed. Severe underweight seems to be a sign that makes both the patient and the doctor convinced of the need for weighing.
There are two additional reasons why those who are not severely underweight should also be weighed – to detect weight loss early and to reassure the patient that weight seldom changes as radically as she/he imagines when the symptoms of an eating disorder are reduced. The patients are weighed more often if they experience support from or are satisfied with their GP. This indicates that weighing is easier to achieve if a relationship of trust is established between the patient and the GP.
The study reveals that little attention is paid to the presence and prevention of serious complications such as reduced bone density and reduced fertility, electrolyte imbalance and dental erosion. Thus GPs do not discuss the illness in such depth that they choose to address its complications with the patient. Many patients enter treatment with therapists who have no training in somatic medicine, and this finding is therefore considered serious. In this respect the GP has an important task to which greater attention should be paid.
Those patients who had never previously entered treatment for their eating disorder had not communicated differently or had more clinical examinations performed than those who had entered treatment. This finding indicates that some patients with no experience of treatment have possibly never undergone a somatic assessment, and that patients who had experienced speaking about their illness during previous treatment did not have more open communication at the GP surgery.
Strengths and weaknesses of the study
The objective of the study was to investigate the extent to which the patients had spoken with their GP about the symptoms of their eating disorder. In line with the authors’ clinical experience before the start of the study, previous studies (
9 – 11) have shown that the various symptoms and ailments of an eating disorder are not all necessarily a topic in GP consultations. The best solution was therefore to ask patients directly and anonymously about both the symptoms and ailments associated with the eating disorder and whether they had spoken to their GP about these.
As we could not find questionnaires for this type of research, we developed a questionnaire form ourselves which we tested in a pilot study. Because the symptoms vary considerably over the course of a year (
5), we did not ask in detail about the frequency of the different symptoms and ailments associated with an eating disorder. This choice means that the study cannot necessarily be compared with studies based on more detailed recording of symptoms and ailments.
The information retrieved in the study is not suitable for assessing patients’ somatic condition or dividing them into different diagnostic subgroups. Nevertheless, we do not believe that this diminishes our possibilities for elucidating patients’ communication with their GPs.
Since the patients in the study are exclusively recruited from specialist units, they may be more seriously ill and more motivated for treatment than other patients with eating disorders.
We interpret the results as indicating that the GP may have insufficient background information for assessing the patient’s somatic condition. This is a finding that we consider valuable for the assessment and planning of the treatment patients are offered as of today. The high response rate shows that patients are concerned about communication with their GPs.
The study does not reveal anything about GPs’ thoughts with regard to communication. Further studies of doctors’ experiences of their encounters with patients with eating disorders will be useful in order to discover further factors that may optimise communication. Studies in which both parties in the dialogue are described in parallel would presumably be particularly useful and informative.