In this issue of the Journal, Madden and colleagues report their prospective investigation of eating disorders in children across Australia (Madden et al).1 This study is an important “first” and investigates the putative increasing problem of early-onset eating disorders (EOEDs) in children aged 5–13 years. Over 3 years, detailed data were collected by the Australian Paediatric Surveillance Unit for 101 children who were managed either as outpatients or in hospital for EOEDs — mainly from paediatricians, but also from child psychiatrists. Most children were hospitalised for treatment. The study raises interesting issues and unanswered questions about eating disorders.

Although there are no earlier data for comparison (ie, conclusions cannot be drawn about whether or not the incidence of eating disorders in younger children is increasing), the annual incidence rate for EOEDs of 1.4 per 100 000 children aged 5–13 years accords with international figures. This is especially true of the even higher incidence rate in New South Wales, where there may have been more comprehensive reporting. Of particular concern were the high rates of severe, life-threatening medical complications (hypothermia, hypotension and bradycardia) in inpatients, suggesting under-referral or under-recognition of the problem, and thus delays in active specialist care. Further, most of the children received nasogastric feeding, and a third received psychotropic medication — treatments that may not have been required with earlier, more active intervention.

Turning to specific issues, there was a relatively high proportion of boys in this study — a quarter of the total. In contrast, men account for about one in 10 adult cases of anorexia nervosa and bulimia nervosa. When broader diagnostic groups, such as binge eating disorder, are considered, rates are higher, particularly in community samples,2 albeit men account for a minority (around 30%) of patients.3 However, the types of eating disorder reported in higher numbers in men (such as binge eating disorder) differ from the EOEDs reported in Madden and colleagues’ study. EOED was characterised by “determined food avoidance plus weight loss or a failure to gain weight during a period of growth, in the absence of any identifiable organic cause”1 — a variant of anorexia nervosa, if not full-spectrum anorexia nervosa.

Nevertheless, the finding that one in four EOED cases affected boys is consistent with results of other studies. For example, a Danish study found males to be younger than females at first presentation and more likely to re-present with psychotic disorder.4 In addition, a large recent study of United States high school students found that binge-eating symptoms were reported by 11.0% of girls and 3.3% of boys, and that recurrent serious purging behaviour (eg, vomiting, laxative use or excessive exercise) was reported by 9.4% of girls and 13.5% of boys.5

How EOEDs in children relate to the eating disorders that emerge later, in adolescence and adult years, is unknown. A 10-year follow-up of a community cohort of 1943 Australian 14–15-year-old adolescents found that partial anorexia nervosa and bulimia nervosa occurred in nearly one in 10 girls aged 15–17 years, and that these girls appeared to be psychologically vulnerable, with poorer functional outcomes and psychiatric morbidity. However, there was little evidence of progression to full anorexia nervosa or bulimia nervosa.6 In contrast, prepubertal children with eating disorders are thought to have a particularly poor prognosis, with high levels of physical and psychiatric morbidity.7 Madden and colleagues’ data support a hypothesis that EOEDs may differ in important ways — including sex distribution and course — from eating disorders with onset in adolescence and adulthood. Whether they have a differing outcome is unknown, and follow-up is imperative.

It is also important that an evidence base for treatments for prepubertal children is developed, despite the well known challenges of conducting controlled trials in an uncommon disorder in children. Notwithstanding this, the high (71%) rate of response to treatment reported by Madden and colleagues accords with other research that has demonstrated more positive treatment outcomes in older children and adolescents when compared with adults, particularly for anorexia nervosa.7

Madden and colleagues report that comorbidities, particularly anxiety disorders and depression, were common in the patients with EOEDs; they also found that around one in five patients were prescribed antidepressants, and one in 10 were prescribed antipsychotics. Given the concerns about the effects of these medications on the developing brain and the risks of antidepressant prescribing in youth, this level of use seems high. It is likely to reflect the severity of illness (especially anorexia nervosa), and the small evidence base supporting use of second-generation antipsychotics, such as olanzapine, in adults with anorexia nervosa.8

Lastly, these results highlight the dilemma of how to address concerns about the epidemic of obesity, while also avoiding contributing to the problems of the much smaller number of children with EOEDs that involve severe dietary restriction and weight loss. Although there is no similar “epidemic of eating disorders”,9 a recent South Australian study indicated that disordered eating in adults is increasing, mostly in the overweight population.10 This supports closer integration of prevention strategies and treatments for disordered eating and obesity, such as the promotion of healthy eating patterns and foods, rather than severe dietary restriction. Extreme weight control behaviour and weight disorders are both important health problems in young people, and Hippocrates’ aphorism that “a diet brought to the extreme point of attenuation is dangerous; and repletion, when in the extreme, is also dangerous”11 remains relevant today.

In conclusion, Madden and colleagues address an important and potentially increasing problem in prepubertal children. It is imperative that research attention is now directed towards understanding why such young children are developing severe eating disorders and how effective identification and treatment can be targeted earlier.