Commentary on: Incidence of eating disorders in Danish psychiatric secondary healthcare 1970–2008

In this issue of the Journal, Gammelmark et al. (2015) report findings from a case-register study of eating disorders (EDs) incidence in Denmark, expanding on earlier work using the same method. Participants were Danish citizens with a first time, primary ED diagnosis receiving hospital-based psychiatric treatment between 1 January 1970 and 31 December 2008, as identified in the Danish Central Psychiatric Research Register. The research is notable for at least two reasons. First, data are captured for an entire population over some four decades, up to and including the present. Second, for the period from 1995 to 2008, data for both inpatient and outpatient treatment are available.

Several findings of interest emerged. First, the (age-standardised) incidence of ED receiving treatment increased significantly during the four decades covered by the study. Second, the steepest increase occurred in females aged 15–19 years, this also being the subgroup for which incidence was highest, during the mid–late 1990s. Third, for the duration of the period of observation, females accounted for the vast majority of individuals with ED receiving treatment. Fourth, outpatient treatment accounted for the majority of ED cases throughout the study period with this proportion increasing somewhat in recent years. Finally, cases of anorexia nervosa (AN) constituted approximately three-quarters (77%) of ED cases receiving inpatient treatment throughout the study period and approximately half (45%) of all cases from 1995 to 2008.

Sources of undetected cases will have included individuals receiving treatment in primary care or private practice only, individuals receiving treatment for a comorbid mental health problem but not an ED and individuals with ED treated in non-psychiatric specialist medical settings (Mond et al., 2007). Hence, even as estimates of treated incidence, the figures reported are underestimates of the true figures. Aside from any other considerations, however, a population-level increase in the occurrence of AN is unlikely simply because the proportion of the population that is underweight – a sine qua non for the diagnosis of AN – has decreased substantially in recent decades (Darby et al., 2009). The particularly steep increase in incidence observed during the mid–late 1990s may have been due, in part, to the change from International Classification of Diseases–Eighth Revision (ICD-8) to ICD-10 (ICD9 was never implemented in Denmark) in 1994 (bulimia nervosa – BN – did not have a separate code in the ICD system prior to ICD-10) and/or the introduction of mandatory registration of outpatient visits for public hospital psychiatry departments in 1995. The fact that males constituted only a tiny proportion of identified cases likely reflects not only lower base rates but also lower uptake of mental healthcare and classification schemes that remain female-centric (Mond et al., 2014).

The authors’ decision to report data for inpatients and for inpatients and outpatients combined, but not for outpatient data alone, and for all ED and AN including atypical AN, but not for BN alone, seems regrettable. The breakdown by diagnosis and sex of the 55% of individuals receiving outpatient treatment who did not receive a diagnosis of AN – presumably cases of BN and variants of BN – would have been of interest, as would information concerning whether and how this breakdown changed over time for both inpatient and outpatient treatment. As the authors note, heterogeneity in the use of diagnostic criteria over time complicates the interpretation of findings relating to diagnoses other than AN. So does the fact that ICD criteria for ED are less specific than those of the Diagnostic and Statistical Manual of Mental Disorders (DSM) (Uher and Rutter, 2012). Still, it would have been better to present these findings. Presumably, there were no cases of binge eating disorder (BED), despite this being the most common ED, as specialist treatment for an ED is uncommon among individuals with this disorder (Mond et al., 2007).

The availability of comparison data from the same registry for individuals with other primary diagnoses, on the other hand, is a plus. These data show that similar trends occurred among females with other primary diagnoses (e.g. borderline personality disorder). Data bearing on the treated (inpatient or outpatient) incidence of first contacts for all diagnoses, showing a marked increase for both males and females between 1995 and 2011, are also consistent with a general increase in the use of psychiatric services. However, rates of all first contacts receiving only inpatient psychiatric treatment remained stable for both males and females during this period. Hence, the observed increase in inpatient ED treatment does not appear to be solely due to a general increase in the use of inpatient services.

As the authors note, the most plausible explanation for the observed increases in treated incidence is improved public awareness and understanding of the nature and adverse consequences of ED and, in turn, of the need for specialist treatment, particularly among individuals with AN. In this respect at least, the findings are encouraging.

See Research by Gammelmark et al., 49(8): 724–730.