A systematic review of the health-related quality of life and economic burdens of anorexia nervosa, bulimia nervosa, and binge eating disorder

Anorexia nervosa (AN), bulimia nervosa (BN), and binge eating disorder (BED) are prevalent psychiatric disorders that are characterized by different symptoms. Based on mixed gender surveys, the lifetime prevalence of AN is estimated to be approximately 0.5–0.6 %, that of BN 0.5–1 %, and that of BED 1.1–2.3 % [1, 2, 3]. The onset of the majority of eating disorders (EDs) occurs between the ages of 10–20 years; however, in contrast to AN and BN, BED often occurs in older cohorts [3]. AN, BN, and BED are associated with numerous physical (e.g., diabetes, hypertension, ulcers) and mental health conditions (e.g., anxiety disorder, depression) [2, 4, 5, 6, 7]. Furthermore, patients with EDs have significantly elevated mortality rates compared with the standard population norms as well as higher suicide rates [8, 9].

Despite the major public health burden of AN, BN, and BED, there is lack of evidence on the differences in the health-related quality of life (HRQoL) and economic impact of different EDs [10, 11]. The objective of this study is to perform a systematic review of the published literature on the HRQoL and economic burdens of AN, BN, and BED. To our knowledge, no systematic review has been published on this topic to date.

This systematic review was based on an extensive literature search on the epidemiology, and HRQoL and economic burdens of EDs (i.e., AN, BN, BED, and eating disorders not otherwise specified [EDNOS]). The literature search was conducted in 2013 using Medline, Embase (via Scopus), PsycINFO, PsycARTICLES, Academic Search Complete, CINAHL Plus, Business Source Premier (via Ebsco Host), and Cochrane Library. The search terms are detailed in Online Resource 1. The search results were processed in two steps: first, the titles and abstracts of all the articles were screened; next, all the potentially relevant articles were analyzed in full text. The references of the included articles were screened for additional eligible studies. The literature screening was conducted by two independent reviewers and disagreements were resolved by the principal researcher.

The reasons for exclusion were as follows for the articles: (1) was not written in English; (2) was not published in a peer-reviewed journal; (3) was an editorial, letter, case report, or review; (4) was not specific to AN, BN, or BED; and (5) had an objective other than studying the HRQoL and/or economic burden (i.e., healthcare resource utilization, healthcare costs, and societal costs) of AN, BN, or BED. This review includes only publications in which AN, BN, and BED samples were clearly separated.

The following information was extracted from each included study: (1) the first author and year of publication; (2) the country; (3) the study design; (4) the study year; (5) the included ED(s); (6) the diagnostic criteria for EDs; (7) the characteristics of the study sample [sample size, % of female, age, and body mass index (BMI)]; (8) data on the HRQoL burden (HRQoL instrument, main findings); (9) data on healthcare utilization (i.e., resource use data on primary care, outpatient services, and inpatient care including emergency services); and (10) data on healthcare costs (i.e., the costs generated by health service utilization and medication use) and/or societal costs (i.e., costs related to the loss of productivity and symptom-related food costs) (perspective of the analysis, year of pricing, resource use and/or cost categories, and resource use and/or cost data).

To compare costs across studies, the costs were extrapolated to annual costs per case and inflated to year 2014 values, using country-specific GDP (gross domestic product) deflators. The costs were converted into United States (US) dollars where necessary, using purchasing power parities and then exchanged to euro (€) using the average annual US dollar/euro exchange rate for 2014 (0.7536). If the year of pricing was not referenced, the midpoint in the observation period was assumed to be the base year; if no observation period was reported, the year of publication was adopted as the base year.

A comprehensive review on the differences in the HRQoL and economic impact of AN, BN, and BED was missing from the literature. We analyzed 69 studies: data on HRQoL were reported in 41 studies (18 for AN, 17 for BN, and 18 for BED), on healthcare utilization in 20 studies (14 for AN, 12 for BN, and 8 for BED), and on healthcare costs in 17 studies (9 for AN, 11 for BN, and 2 for BED).

Generally, AN, BN, or BED was associated with significantly impaired HRQoL compared with the healthy population [18, 34, 37, 39, 40, 43]. Evidence also suggests that patients with EDs have impaired HRQoL compared to those with other psychiatric conditions [80]. Studies showed that obesity was associated with further deterioration of HRQoL of patients with BED [36, 37, 40, 43, 46]. In addition, other psychopathology factors (e.g., depression and anxiety) were associated with lower HRQoL of patients with AN, BN, and BED [19, 21, 24, 39, 43]. AN, BN, and BED may have differential effects on HRQoL compared with each other, depending mainly, but not exclusively, on the study design and the HRQoL domains investigated [15, 17, 22, 27]. The reviewed studies provided a range of results, at least partially due to the heterogeneity in the patient population and the research design. Consequently, the study results and conclusions seemed to be quite heterogeneous. It is difficult to draw a consistent conclusion on the relative effects of the assessed EDs on HRQoL, which may be because AN, BN, and BED share certain psychopathology symptoms (e.g., loss of control of eating appears both in BN and BED). In a recently published meta-analysis, Winkler et al. [81] could also not identify any specific differences between the HRQoL effects of different ED diagnostic groups. The HRQoL of the siblings and caregivers of AN, BN, and BED patients was also found to be impaired [49, 50, 51, 52, 53].

The economic impacts of EDs has been receiving increasing research attention in recent years [82, 83, 84]; however, a comparative review on the economic burden of AN, BN, and BED was lacking in the literature. Based on the reviewed studies, AN, BN, and BED are associated with increased healthcare utilization and higher healthcare costs. The evaluated EDs are associated with a high rate of hospitalization, outpatient care, and emergency department visits [3, 54, 58, 59, 60, 61, 62, 64, 65, 66, 69]. However, inpatient care seems to be more common in AN [54, 58, 60, 61, 64], which could primarily be due the severity of the condition. It was also notable that patients rarely receive specific treatment for their ED and mostly received treatment for general mental health, comorbid psychiatric symptoms, and/or weight loss [2, 3, 59, 66]. Improvement in diagnosis, reporting/coding and management of AN, BN, and BED may improve treatment for these patients.

The cost estimates varied considerably for AN, BN, and BED, as well as within particular disorders. Several differences in the design and analysis of the cost of illness in the studies reduced the comparability of the cost data. The annual direct healthcare costs of AN and BN appeared to be similar to those of other major psychiatric conditions [61]. Compared with BN, AN was associated with higher healthcare costs in most assessed studies [58, 60, 72]. Healthcare cost data showed variation across countries; however, an investigation of the factors behind the cross-country differences was beyond the scope of this review (differences in treatment patterns and evaluated cost factors as well as national characteristics of the healthcare systems are likely to be key determinants).

Compared to BN, obesity appears to be more common in BED [2]. Due to the limited number of studies, it is not possible to definitively separate burden specific to BED from those associated with obesity. Namely, a higher level of BMI decreases the physical HRQoL of individuals (e.g., worst physical functioning) [85]. Nevertheless, BED is distinct from obesity; there is a unique psychopathology and medical risk profile associated with BED [86]. Patients with BED have greater functional impairment, lower HRQoL, and more subjective distress compared to weight-matched obese individuals without BED [36, 37, 40, 44, 46]. This suggests that at least part of the burden of BED is unique to the disorder and provides a rationale for not only treating weight issues, but the disorder itself to fully address the burden of BED. However, obesity may have severe health consequences over time (e.g., diabetes and hypertension), which may have an impact on the long-term burden of BED. Although BED is the most prevalent ED, there is scarcity of information on the economic burden of BED; those studies that do report the economic burden of BED, most reported it in combination with BN or overall EDs (only 2 studies reported healthcare cost data for BED [39, 68]). Further research is needed on this topic, because data on the long-term health and economic impact of BED are essential to better establish the burden of illness in this patient population.

Our findings should be considered in light of the following limitations. The current systematic literature review involved searching only studies that were published in English and indexed in the selected databases. Despite the broad database selection, the inclusion of unpublished material and studies from other resources was precluded. As a further effort, we screened the references of the included articles for additional eligible studies. No formal quality assessment of the identified publications was performed, because of the large heterogeneity in the study types of the included studies.

In summary, AN, BN, and BED markedly impair HRQoL and are associated with increased healthcare use and costs compared with healthy individuals. The HRQoL impairment of AN, BN, and BED patients was found to be similar; however, healthcare utilization (especially, inpatient care) and costs were reported to be higher for AN patients than for BN and BED patients. Compared to BN, obesity appears to be more common in patients with BED, which may contribute to significant impairment of their HRQoL and higher healthcare use, especially in the long term. Thus, the burden of BED should be examined separately from that of BN. The limited evidence indicates that further research is warranted to better understand the differences between the HRQoL and economic burdens of BED in contrast to AN and BN in the long term.