Therefore, our estimate is based predominately on research carried out in other countries (primarily other western European countries, Northern European countries, the USA and Australia). Of course, this is not ideal, given the different social and cultural contexts and access to treatment between countries.
We will revise this estimate as new data becomes available. We are in the process of attempting to obtain unpublished data that would reduce the uncertainty in this estimate.
To calculate the most accurate estimate with the data available to us, we looked at a series of studies that investigated prevalence of eating disorders among different genders, age ranges, and diagnoses, and sought to combine these into one total. We have broken these categories down below.
To determine the percentage of sufferers among girls aged 5 – 10, we took the prevalence rate for eating disorders estimated in the last national study of the prevalence of mental health conditions in children and adolescents conducted in the UK [1].
For those aged 11 – 34, we used the average prevalence rates for Anorexia and Bulimia reported in the last major academic review [2]. The paper cited reports these estimates as relating to “young women”. We applied them to under-35s based on the ages of the participants who took part in the studies they reviewed.
For those aged 11 – 24 with Eating Disorders Not Otherwise Specified (EDNOS), we used the prevalence rate from a large study of high school students in Portugal [3].
For those aged 35 and over with Anorexia, we applied an estimate derived from a review of studies on the outcome of Anorexia Nervosa [4], which suggested that 20% of patients develop a chronic condition. Therefore, we applied this ratio to the estimate of 0.3% for females aged 11 – 34:
For those aged 35 and over with Bulimia, we applied estimates derived by comparing the difference in prevalence of Bulimia as age increases, which we calculated from some major US studies [5].
For those over 25 with EDNOS, we applied estimates derived by comparing the difference in prevalence of Binge Eating Disorder (BED) as age increases, which we calculated from some major US studies [5].
To determine the percentage of sufferers among boys aged 5 – 10, we took the prevalence rate for eating disorders estimated by the last national study of the prevalence of mental health conditions in children and adolescents conducted in the UK [1].
For those aged 11 – 34 with Anorexia, we assumed that males account for 25% of cases [6]. We applied this ratio to the estimate for females with Anorexia aged 11 – 34, which is 0.3%.
For those aged 11 – 34 with Bulimia, we assumed that males account for 25% of cases [6]. We applied this ratio to the estimate for females with Bulimia aged 11 – 34, which is 1%:
For those aged 11 – 24 with Eating Disorders Not Otherwise Specified (EDNOS), we assumed that males account for 25% of cases [6]. We applied this ratio to the estimate for females with EDNOS aged 11 – 24, which is 2.37%:
For 35-year-olds and over with Anorexia, we assumed that males account for 25% of cases [6]. We applied this ratio to the estimate for females aged 35+ with Anorexia of 0.075%.
For 35-year-olds and over with Bulimia, we applied estimates derived by comparing the difference in prevalence of Bulimia as age increases which we calculated from some major US studies [5].
For 25-year-olds and over with EDNOS, we applied estimates derived by comparing the difference in prevalence of Binge Eating Disorder (BED) as age increases which we calculated from some major US studies [5].
When applied to recent data about the UK population [7], this leads to a total estimate of 1,248,663 ~ 1.25 million.
The studies reviewed to create this estimate calculated prevalence using the diagnostic definitions of DSM-IV. There is still relatively little evidence about the prevalence of eating disorders in the community based on the new DSM-5 diagnostic categories, which no longer include the umbrella of EDNOS, but instead include Binge Eating Disorder as a separate diagnosis, and OSFED (Other Specified Feeding or Eating Disorder) where a patient’s symptoms do not exactly match the expected criteria for Anorexia, Bulimia, or Binge Eating Disorder.
An alternative calculation based on the DSM-5 criteria using estimates derived from Machado et. al. (2013), Micali et. al. (2017) and Conceicao et. al. (2017) resulted in a similar total of 1.37 million. However, these three studies do not, when taken together, cover the entire age range, and so assumptions had to be made that their estimates were valid beyond the ages of the samples they recruited.
References
[1] Unfortunately, this study only assessed for the prevalence of anorexia or bulimia, and did not report separate prevalence rates by diagnosis. Source: Green et. al., 2005[2] Smink, van Hoeken and Hoek, 2012[3] Machado et. al., 2007[4] Steinhaussen (2002)[5] Swanson et. al. (2011); Hudson et. al. (2007) [6] Based on a review of community-based prevalence studies by Sweeting et. al. (2015). [7] ONS (2017)