Global, regional, and national burden of eating disorders in 204 countries and territories, 1990–2021: findings from the Global Burden of Disease Study 2021 Authors: Yi Shen, Bo Chen, Hanwen Zhang, Wanying Shen, Cui Wu, Liying Jiang Published: 2025 ABSTRACT ObjectiveThe study is to examine the global, regional, and national burden of anorexia nervosa and bulimia nervosa from 1990 to 2021, and also the profile of eating disorders burden separated by the Socio-demographic Index (SDI) measurement.MethodsWe extracted age-standardized rates (ASRs) and their 95% uncertainty intervals (UIs) for the prevalence and disability-adjusted life years (DALYs) of anorexia nervosa (AN) and bulimia nervosa (BN) from the Global Burden of Disease Study 2021 (GBD 2021) through the period 1990–2021. Estimated annual percentage changes (EAPCs) were calculated to quantify secular trends of the burden of AN and BN. Cross-country inequality analysis was performed to assess absolute and relative inequalities of the burden between countries and territories.ResultsGlobally, the burden of AN and BN shows distinct geographic variations. In 2021, the age-standardized prevalence and DALYs rates of the diseases were generally higher in high-income regions such as North America, Western Europe, and Australasia, whereas Asia, Africa, and most countries across Latin America recorded relatively low levels. Although the burden was higher among females, the increase in ASRs was greater among males. In 2021, the highest ASRs of AN and BN were observed in Australasia globally, with prevalence rates of 184.36 (95% UI: 124.96-257.23) and 811.90 (95% UI: 629.68-1041.59), and DALYs rates of 39.46 (95% UI: 23.94–62.53) and 170.34 (95% UI: 106.42-262.05) per 100,000 population, respectively. However, East Asia exhibited the most significant ASRs increases. The burden estimates exhibited a positive correlation with SDI levels across most geographical regions. Furthermore, we observed that higher prevalence and DALYs rates were disproportionately concentrated in those countries with higher SDI.ConclusionSubstantial disparities exist in the burden of eating disorders across different SDI levels, regions, and countries, with marked variations in the rates of increase. Future research and interventions aimed at mitigating the burden should prioritize the development of tailored prevention and treatment strategies accounting for distinct regional and cultural contexts.Plain English summaryEating disorders (EDs), such as anorexia nervosa (AN) and bulimia nervosa (BN), are serious health conditions that affect millions of people around the world. Understanding their impact on different populations is crucial for organizing effective healthcare and support.In this study, we scrutinized the global burden of these disorders—meaning how common and disabling they are—across different countries and regions from 1990 to 2021. We analyzed data from over 200 countries to see how patterns have changed over time and whether they are linked to a country’s level of social and economic development.What we found:Global Patterns: The burden of eating disorders is not evenly spread. In 2021, they were much more common in high-income regions like North America, Western Europe, and Australasia. In contrast, countries across Asia, Africa, and much of Latin America reported relatively lower levels. A Growing Concern for Men: While females are still more affected overall, the rate of increase in cases among males has been greater over the past three decades. Link to Development: There is a clear connection between a nation's development level and eating disorder rates. Countries with higher socioeconomic development tend to have significantly higher prevalence and health impacts from these disorders.Why this matters:Our findings show that the challenge of eating disorders varies greatly around the world. This means that a one-size-fits-all approach will not work. To reduce the global impact of these illnesses, prevention and treatment strategies must be carefully tailored to fit the specific social, economic, and cultural contexts of different regions and countries.Supplementary InformationThe online version contains supplementary material available at 10.1186/s40337-025-01492-8. Introduction Eating disorders (EDs) are characterized by an overvaluation of weight and shape, psychosocial impairment, and dysregulated weight control [1]. This ultimately predisposes individuals to numerous physical and psychiatric complications, contributing to degressive physical health and cognitive deficits [2, 3]. EDs represent a major public health concern, adversely impacting both physical and psychological well-being [4]. Anorexia nervosa (AN), bulimia nervosa (BN), and binge eating disorders (BEDs) constitute the primary clinically recognized EDs within current nosology [2]. EDs have complex and multifactorial proponents, including interactions between genetic susceptibility and environmental factors such as sociocultural influences [3, 5]. Anorexia nervosa (AN) and bulimia nervosa (BN) are those serious eating disorders (EDs), and psychological factors are the most significant upsurge roles on the onset and development of the disease [6]. Previous evidence suggests that various eating disorders and related behaviors share common psychological features, represented by core cognitive preoccupations with body shape and weight [7]. Psychiatric comorbidity, particularly affective and anxiety disorders, is highly prevalent among affected individuals [8]. Globally, approximately 7.8% individuals are diagnosed EDs, while a substantially larger proportion (≥ 20%) experience subclinical EDs presentations [1]. The weighted average lifetime prevalence of EDs globally was estimated to be 8.4% in females (3.3%–18.6%) and 2.2% in males (0.8%–6.5%)9. By disorder, the lifetime prevalence was 1.4% for AN and 1.9% for BN among females, compared with 0.2% for AN and 0.6% for BN among males [9]. The global burden of disease (GBD) study also reported that 13.6 million (95% UI: 10.2–17.5) people had AN or BN in 2019, equivalent to 176.2 (131.7-225.6) per 100,000 people [10]. Importantly, a reported rise in EDs incidence was also observed during the COVID-19 pandemic [11, 12]. Since AN is one of the most common subtype of EDs, more than 3 million people experience the disease [13]. AN is involved in an elevated incidence of respiratory and cervical cancers [14], also connected with deteriorated skeletal health, leading to impaired bone development, increased bone resorption, and bone marrow suppression [15, 16]. It is noteworthy that among individuals with EDs, AN has the highest standardized mortality rate [17]. Physical abnormalities in patients with BN are typically mild. However, purging behaviors, such as self-induced vomiting, misuse of laxatives, and diuretics can lead to electrolyte imbalances [18]. These may be esophageal damage caused by repeated vomiting. Apart from the profound impact on physical and mental health, EDs have brought considerable challenges to public health undertakings, for which effective prevention and treatment countermeasures are urgently needed [19]. Previous literature on EDs has largely relied on data from 1990 to 2019 or earlier, which poses limitations due to restricted time spans or outdated datasets. To the best of knowledge, existing reports have primarily focused on EDs burden in specific populations such as reproductive-aged women or adolescents, lacking systematic characterization across other age groups. Here, by analyzing global data from 1990 to 2021 for AN and BN, we present the first quantitative assessment of disparities in AN and BN prevalence and disability-adjusted life years (DALYs) between the highest and lowest income countries. These findings provide the information for better health care decision making and policy implementation. Our study aims to: (1) analyze the global, regional, and national burden of AN and BN from 1990 to 2021; (2) and examine associations between AN and BN burden and development levels by the Socio-demographic Index (SDI) measurement. Method Research data The Global Burden of Diseases (GBD) 2021 incorporated findings from nationally representative surveys, censuses, and meta-analyses, employing the latest epidemiological data and advanced standardized methodologies [20], representing a comprehensive assessment of health loss across 204 countries and territories for 371 diseases and injuries, alongside 88 risk factors [21]. General GBD 2021 methods are detailed on the official website (https://www.healthdata.org/research-analysis/gbd) [22]. For each estimated metric, GBD 2021 reports a 95% UIs. These UIs account for multiple sources of uncertainty, including sampling error, non-sampling error, and model uncertainty, providing a more robust evaluation of estimation uncertainty [23]. The composite development level of a nation or territory is quantified via the socio-demographic index (SDI), deriving from three components: lag-distributed income per capita, average educational attainment among individuals aged 15 years and older, and the total fertility rate among those under age 25 [24]. Countries and territories were subsequently stratified into five SDI based regions: low, low-middle, middle, high-middle, and high. Using data retrieved from the Global Health Data Exchange (GHDx) query tool (https://vizhub.healthdata.org/gbd-results/) [25], age-standardized rates (ASRs) per 100,000 population for the prevalence and DALYs of EDs, AN and BN were obtained between 1990 and 2021. In the GBD 2021, EDs were categorized into two distinct entities: anorexia nervosa and bulimia nervosa. Case definitions adhered to diagnostic criteria specified in the Diagnostic and Statistical Manual of Mental Disorders (DSM; encompassing versions DSM-III, DSM-III-R, DSM-IV, DSM-IV-TR, and DSM-5) and the International Classification of Diseases (ICD; versions ICD-9 and ICD-10) [26]. GBD 2021 also assumes that there are no cases of EDs under the age of 5 or over the age of 50. These settings are consistent with the corresponding cause of death models for EDs. Our analysis was restricted to individuals aged 5–49 years, being stratified into 5-year age bands, to evaluate sex- and age-specific burdens of eating disorders. Data analysis Trends in EDs prevalence and DALYs were quantified using ASRs and their corresponding estimated annual percentage changes (EAPCs). The ASR (per 100,000 population) was computed as the sum of the products of the age-specific ratios (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{\upalpha\:}}_{\text{I}}$$\end{document}, where i denotes the i-th age group) and the population size (or weights;\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{\upomega\:}}_{\text{i}}$$\end{document}) of age group i in the selected reference standard population and then dividing by the sum of the standard population weights. A represents the total number of age strata [27]. Age standardization employed the GBD world standard population.\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\text{A}\text{S}\text{R}=\frac{{\sum\:}_{\text{i}=1}^{\text{A}}{{\upalpha\:}}_{\text{i}}{{\upomega\:}}_{\text{i}}}{{\sum\:}_{\text{i}=1}^{\text{A}}{{\upalpha\:}}_{\text{i}}}\times\:100\:000$$\end{document} EAPC is often used to reflect the trend of ASR within a certain range. The regression line and the natural logarithm of the rates are fitted as follows:1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\text{ln}\left(y\right)=\alpha\:+\beta\:x+\epsilon\:$$\end{document}2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\text{E}\text{A}\text{P}\text{C}=100\times\:\left({\text{e}}^{{\upbeta\:}}-1\right)\:\:\:\:\:\:\:\:\:\:\:\:\:$$\end{document} Where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:y$$\end{document} denotes the rate, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:x$$\end{document} represents the calendar year, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:$$\end{document} indicates the slope of the trend segment, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\alpha\:$$\end{document} is the intercept, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\epsilon\:$$\end{document} signifies the error term. The estimated EAPC was calculated as \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:100\times\:\left({\text{e}}^{{\upbeta\:}}-1\right)$$\end{document}, with its corresponding 95% confidence interval (CI) derived from the linear regression model. An upward trend was defined when both the EAPC and the lower boundary of its 95% CI > 0. Conversely, a downward trend was identified when both the EAPC and the upper boundary of its 95% CI < 0. All other scenarios were classified as stable trends [28]. Cross-country inequality analysis In terms of cross-country inequality analysis, the slope index of inequality (SII) and concentration index represent standardized metrics, quantifying absolute and relative health gradients, respectively [29]. The SII is determined by regressing national prevalence rates and DALY rates for all age groups on a relative position scale related to socio-demographic development. The concentration index is calculated by numerically integrating the area under the Lorenz curve, which is generated based on the cumulative population proportion classified by the SDI [30]. All statistical analysis were performed using R (version 4.4.2) and corresponding statistical software packages. P < 0.05 was considered statistically significant. Results The global burden of anorexia nervosa and bulimia nervosa The global burden and trends of AN and BN are presented in Tables 1 and 2, respectively. The global age-standardized prevalence rate of AN increased from 45.32 (95% UI: 31.56–63.43) per 100,000 population in 1990 to 47.75 (95% UI: 33.23–66.33) per 100,000 in 2021, with the EAPC of 0.24 (95% CI: 0.21–0.27; Fig. 1A). Correspondingly, the global age-standardized DALY rate for AN rose from 9.80 (95%UI: 6.11–15.53) per 100,000 in 1990 to 10.31 (95%UI: 6.42–16.25) per 100,000 in 2021, with the EAPC of 0.23 (95% CI: 0.21–0.26; Fig. 1C). For BN, the global age-standardized prevalence rate increased from 130.66 (95%UI: 86.28-184.95) per 100,000 in 1990 to 156.85 (95% UI: 103.39-221.39) per 100,000 in 2021, representing an annual increase of 0.66 (95% CI: 0.61–0.72; Supplementary Fig. 1A). The global age-standardized DALY rate for BN increased from 27.53 (95% UI: 15.53–45.89) per 100,000 in 1990 to 33.05 (95% UI: 18.64–55.29) per 100,000 in 2021, being equivalent to an annual increase of 0.67 (95% CI: 0.62–0.72; Supplementary Fig. 1C). Table 1Age-standardized rates of prevalence and disability-adjusted life years of anorexia nervosa in 2021, alongside their Temporal trend from 1990 to 2021, presented at global and regional levelsLocationPrevalence (95% UI)DALYs (95% UI)ASR in 1990(per 100,000 population)ASR in 2021(per 100,000 population)1990–2021EAPC (95% CI)ASR in 1990(per 100,000 population)ASR in 2021(per 100,000 population)1990–2021EAPC (95% CI)Global45.32 (31.56,63.43)47.75 (33.23,66.33)0.24 (0.21,0.27)9.80 (6.11,15.53)10.31 (6.42,16.25)0.23 (0.21,0.26)Gender Male23.71 (16.49,33.42)25.89 (17.97,36.29)0.34 (0.31,0.37)5.13 (3.13,8.20)5.60 (3.36,9.02)0.34 (0.31,0.36) Female67.46 (46.57,94.59)70.38 (48.70,97.10)0.21 (0.18,0.24)14.58 (9.12,23.09)15.18 (9.53,23.90)0.20 (0.18,0.23)Socio-demographic index High-middle SDI45.02 (31.44,62.83)52.91 (36.47,73.88)0.62 (0.56,0.68)9.78 (6.09,15.49)11.49 (7.03,18.14)0.63 (0.56,0.69) High SDI113.53 (78.40,155.75)125.74 (85.97,171.27)0.32 (0.29,0.34)24.81 (15.47,38.60)27.45 (17.27,42.79)0.31 (0.27,0.34) Low-middle SDI28.27 (19.50,39.66)34.18 (23.58,48.52)0.67 (0.64,0.71)6.00 (3.64,9.56)7.28 (4.46,11.54)0.69 (0.65,0.72) Low SDI25.72 (17.94,36.26)27.38 (19.11,38.61)0.29 (0.21,0.37)5.43 (3.28,8.70)5.82 (3.52,9.38)0.32 (0.24,0.40) Middle SDI30.11 (20.71,42.70)39.32 (27.20,55.43)0.95 (0.93,0.97)6.46 (3.91,10.40)8.45 (5.17,13.51)0.96 (0.94,0.98)Region East Asia23.26 (16.07,32.94)33.69 (23.32,47.70)1.39 (1.34,1.44)5.05 (3.03,8.05)7.41 (4.46,11.78)1.45 (1.39,1.50) Southeast Asia26.26 (18.39,37.02)32.87 (22.65,45.90)0.68 (0.65,0.72)5.65 (3.41,9.05)7.09 (4.35,11.31)0.70 (0.66,0.73) Oceania25.15 (17.52,35.50)25.60 (17.78,36.17)-0.05 (-0.11,0.00)5.39 (3.14,8.57)5.47 (3.22,8.84)-0.05 (-0.11,0.01) Central Asia39.50 (27.34,56.06)43.95 (30.63,62.26)0.63 (0.40,0.85)8.55 (5.22,13.65)9.48 (5.75,15.02)0.61 (0.39,0.84) Central Europe42.20 (29.31,59.67)57.23 (39.86,80.34)1.24 (1.15,1.33)9.03 (5.39,14.58)12.73 (7.92,19.98)1.41 (1.31,1.51) Eastern Europe53.63 (37.06,76.03)57.07 (39.50,80.05)0.46 (0.25,0.67)11.57 (7.01,18.34)12.43 (7.69,19.44)0.50 (0.29,0.72) High-income Asia Pacific117.32 (81.11,163.10)150.22 (103.32,204.40)0.75 (0.66,0.85)26.00 (15.87,40.81)33.62 (21.38,51.95)0.74 (0.62,0.86) Australasia159.72 (108.98,231.06)184.36 (124.96,257.23)1.03 (0.77,1.29)34.35 (20.54,53.54)39.46 (23.94,62.53)1.01 (0.75,1.26) Western Europe127.27 (90.18,173.16)143.10 (98.64,195.28)0.41 (0.36,0.45)28.15 (18.08,43.05)31.33 (19.83,48.46)0.37 (0.32,0.41) Southern Latin America74.35 (50.53,106.86)82.99 (57.07,117.07)0.33 (0.29,0.37)16.09 (9.61,25.88)17.73 (10.59,28.60)0.27 (0.21,0.32) High-income North America134.41 (91.19,187.71)137.69 (94.01,189.86)0.11 (0.08,0.13)28.85 (17.24,45.31)29.55 (18.03,46.11)0.11 (0.09,0.14) Caribbean40.50 (27.84,57.56)41.09 (28.42,57.79)0.14 (0.10,0.18)8.62 (5.24,13.80)8.72 (5.22,13.85)0.14 (0.10,0.18) Andean Latin America38.21 (26.27,53.39)44.29 (30.86,61.93)0.64 (0.57,0.71)8.19 (4.97,13.05)9.49 (5.67,15.29)0.65 (0.58,0.73) Central Latin America43.89 (30.12,61.78)46.48 (31.98,65.73)0.22 (0.20,0.24)9.38 (5.69,14.94)9.97 (6.11,16.08)0.22 (0.21,0.24) Tropical Latin America55.78 (38.51,78.51)62.39 (43.14,88.48)0.46 (0.43,0.49)11.89 (7.38,19.40)13.30 (8.26,21.29)0.47 (0.43,0.51) North Africa and Middle East38.13 (26.32,54.21)41.77 (28.72,59.80)0.41 (0.37,0.46)8.12 (4.92,13.12)8.89 (5.41,14.07)0.42 (0.38,0.47) South Asia26.11 (17.93,36.65)33.77 (23.15,47.63)0.88 (0.84,0.92)5.53 (3.34,8.86)7.19 (4.39,11.47)0.90 (0.86,0.93) Central Sub-Saharan Africa28.71 (19.96,41.02)26.84 (18.83,37.61)-0.04 (-0.21,0.12)6.01 (3.50,9.60)5.69 (3.49,9.55)-0.00 (-0.16,0.16) Eastern Sub-Saharan Africa24.99 (17.36,35.34)26.92 (18.65,37.62)0.33 (0.25,0.40)5.29 (3.21,8.35)5.73 (3.49,9.14)0.36 (0.29,0.44) Southern Sub-Saharan Africa39.23 (27.22,55.07)39.42 (27.30,55.90)0.06 (0.05,0.08)8.31 (4.98,13.16)8.33 (5.00,13.26)0.06 (0.04,0.07) Western Sub-Saharan Africa29.97 (20.60,42.31)31.69 (22.01,44.80)0.33 (0.24,0.41)6.34 (3.87,10.22)6.72 (4.10,10.72)0.35 (0.26,0.44)ASR age-standardized rate, DALYs disability-adjusted life-years, EAPC estimated annual percentage change, UI uncertainty interval, CI confidence interval, SDI sociodemographic index Table 1Age-standardized rates of prevalence and disability-adjusted life years of anorexia nervosa in 2021, alongside their Temporal trend from 1990 to 2021, presented at global and regional levelsLocationPrevalence (95% UI)DALYs (95% UI)ASR in 1990(per 100,000 population)ASR in 2021(per 100,000 population)1990–2021EAPC (95% CI)ASR in 1990(per 100,000 population)ASR in 2021(per 100,000 population)1990–2021EAPC (95% CI)Global45.32 (31.56,63.43)47.75 (33.23,66.33)0.24 (0.21,0.27)9.80 (6.11,15.53)10.31 (6.42,16.25)0.23 (0.21,0.26)Gender Male23.71 (16.49,33.42)25.89 (17.97,36.29)0.34 (0.31,0.37)5.13 (3.13,8.20)5.60 (3.36,9.02)0.34 (0.31,0.36) Female67.46 (46.57,94.59)70.38 (48.70,97.10)0.21 (0.18,0.24)14.58 (9.12,23.09)15.18 (9.53,23.90)0.20 (0.18,0.23)Socio-demographic index High-middle SDI45.02 (31.44,62.83)52.91 (36.47,73.88)0.62 (0.56,0.68)9.78 (6.09,15.49)11.49 (7.03,18.14)0.63 (0.56,0.69) High SDI113.53 (78.40,155.75)125.74 (85.97,171.27)0.32 (0.29,0.34)24.81 (15.47,38.60)27.45 (17.27,42.79)0.31 (0.27,0.34) Low-middle SDI28.27 (19.50,39.66)34.18 (23.58,48.52)0.67 (0.64,0.71)6.00 (3.64,9.56)7.28 (4.46,11.54)0.69 (0.65,0.72) Low SDI25.72 (17.94,36.26)27.38 (19.11,38.61)0.29 (0.21,0.37)5.43 (3.28,8.70)5.82 (3.52,9.38)0.32 (0.24,0.40) Middle SDI30.11 (20.71,42.70)39.32 (27.20,55.43)0.95 (0.93,0.97)6.46 (3.91,10.40)8.45 (5.17,13.51)0.96 (0.94,0.98)Region East Asia23.26 (16.07,32.94)33.69 (23.32,47.70)1.39 (1.34,1.44)5.05 (3.03,8.05)7.41 (4.46,11.78)1.45 (1.39,1.50) Southeast Asia26.26 (18.39,37.02)32.87 (22.65,45.90)0.68 (0.65,0.72)5.65 (3.41,9.05)7.09 (4.35,11.31)0.70 (0.66,0.73) Oceania25.15 (17.52,35.50)25.60 (17.78,36.17)-0.05 (-0.11,0.00)5.39 (3.14,8.57)5.47 (3.22,8.84)-0.05 (-0.11,0.01) Central Asia39.50 (27.34,56.06)43.95 (30.63,62.26)0.63 (0.40,0.85)8.55 (5.22,13.65)9.48 (5.75,15.02)0.61 (0.39,0.84) Central Europe42.20 (29.31,59.67)57.23 (39.86,80.34)1.24 (1.15,1.33)9.03 (5.39,14.58)12.73 (7.92,19.98)1.41 (1.31,1.51) Eastern Europe53.63 (37.06,76.03)57.07 (39.50,80.05)0.46 (0.25,0.67)11.57 (7.01,18.34)12.43 (7.69,19.44)0.50 (0.29,0.72) High-income Asia Pacific117.32 (81.11,163.10)150.22 (103.32,204.40)0.75 (0.66,0.85)26.00 (15.87,40.81)33.62 (21.38,51.95)0.74 (0.62,0.86) Australasia159.72 (108.98,231.06)184.36 (124.96,257.23)1.03 (0.77,1.29)34.35 (20.54,53.54)39.46 (23.94,62.53)1.01 (0.75,1.26) Western Europe127.27 (90.18,173.16)143.10 (98.64,195.28)0.41 (0.36,0.45)28.15 (18.08,43.05)31.33 (19.83,48.46)0.37 (0.32,0.41) Southern Latin America74.35 (50.53,106.86)82.99 (57.07,117.07)0.33 (0.29,0.37)16.09 (9.61,25.88)17.73 (10.59,28.60)0.27 (0.21,0.32) High-income North America134.41 (91.19,187.71)137.69 (94.01,189.86)0.11 (0.08,0.13)28.85 (17.24,45.31)29.55 (18.03,46.11)0.11 (0.09,0.14) Caribbean40.50 (27.84,57.56)41.09 (28.42,57.79)0.14 (0.10,0.18)8.62 (5.24,13.80)8.72 (5.22,13.85)0.14 (0.10,0.18) Andean Latin America38.21 (26.27,53.39)44.29 (30.86,61.93)0.64 (0.57,0.71)8.19 (4.97,13.05)9.49 (5.67,15.29)0.65 (0.58,0.73) Central Latin America43.89 (30.12,61.78)46.48 (31.98,65.73)0.22 (0.20,0.24)9.38 (5.69,14.94)9.97 (6.11,16.08)0.22 (0.21,0.24) Tropical Latin America55.78 (38.51,78.51)62.39 (43.14,88.48)0.46 (0.43,0.49)11.89 (7.38,19.40)13.30 (8.26,21.29)0.47 (0.43,0.51) North Africa and Middle East38.13 (26.32,54.21)41.77 (28.72,59.80)0.41 (0.37,0.46)8.12 (4.92,13.12)8.89 (5.41,14.07)0.42 (0.38,0.47) South Asia26.11 (17.93,36.65)33.77 (23.15,47.63)0.88 (0.84,0.92)5.53 (3.34,8.86)7.19 (4.39,11.47)0.90 (0.86,0.93) Central Sub-Saharan Africa28.71 (19.96,41.02)26.84 (18.83,37.61)-0.04 (-0.21,0.12)6.01 (3.50,9.60)5.69 (3.49,9.55)-0.00 (-0.16,0.16) Eastern Sub-Saharan Africa24.99 (17.36,35.34)26.92 (18.65,37.62)0.33 (0.25,0.40)5.29 (3.21,8.35)5.73 (3.49,9.14)0.36 (0.29,0.44) Southern Sub-Saharan Africa39.23 (27.22,55.07)39.42 (27.30,55.90)0.06 (0.05,0.08)8.31 (4.98,13.16)8.33 (5.00,13.26)0.06 (0.04,0.07) Western Sub-Saharan Africa29.97 (20.60,42.31)31.69 (22.01,44.80)0.33 (0.24,0.41)6.34 (3.87,10.22)6.72 (4.10,10.72)0.35 (0.26,0.44)ASR age-standardized rate, DALYs disability-adjusted life-years, EAPC estimated annual percentage change, UI uncertainty interval, CI confidence interval, SDI sociodemographic index Age-standardized rates of prevalence and disability-adjusted life years of anorexia nervosa in 2021, alongside their Temporal trend from 1990 to 2021, presented at global and regional levels ASR age-standardized rate, DALYs disability-adjusted life-years, EAPC estimated annual percentage change, UI uncertainty interval, CI confidence interval, SDI sociodemographic index Table 2Age-standardized rates of prevalence and disability-adjusted life years of bulimia nervosa in 2021, alongside their Temporal trend from 1990 to 2021, presented at global and regional levelsLocationPrevalence (95% UI)DALYs (95% UI)ASR in 1990(per 100,000 population)ASR in 2021(per 100,000 population)1990–2021EAPC (95% CI)ASR in 1990(per 100,000 population)ASR in 2021(per 100,000 population)1990–2021EAPC (95% CI)Global130.66 (86.28,184.95)156.85 (103.39,221.39)0.66 (0.61,0.72)27.53 (15.53,45.89)33.05 (18.64,55.29)0.67 (0.62,0.72)Gender Male98.03 (65.00,142.88)121.98 (81.38,177.10)0.80 (0.74,0.87)20.87 (11.72,34.84)26.01 (14.68,43.26)0.81 (0.75,0.88) Female164.09 (108.15,230.02)192.73 (125.99,269.49)0.58 (0.53,0.62)34.35 (19.24,56.42)40.30 (22.68,66.23)0.58 (0.54,0.63)Socio-demographic index High-middle SDI129.57 (87.51,181.74)173.19 (114.76,245.25)1.11 (1.01,1.22)27.41 (15.72,45.14)36.70 (20.43,60.71)1.13 (1.02,1.23) High SDI273.11 (182.59,381.40)311.26 (211.22,435.75)0.33 (0.28,0.38)57.53 (32.24,93.91)65.38 (37.29,106.61)0.32 (0.27,0.37) Low-middle SDI93.51 (60.84,134.77)127.72 (83.03,183.55)1.12 (1.06,1.19)19.62 (11.14,32.63)26.86 (14.97,45.07)1.14 (1.08,1.21) Low SDI83.52 (54.24,121.96)96.69 (62.85,140.31)0.62 (0.51,0.74)17.44 (9.79,29.13)20.31 (11.42,33.98)0.65 (0.54,0.77) Middle SDI96.20 (62.58,139.63)144.24 (94.08,205.37)1.42 (1.37,1.47)20.34 (11.34,33.97)30.48 (17.07,51.27)1.42 (1.37,1.47)Region East Asia58.73 (37.95,86.68)103.0 2 (66.60,153.25)1.91 (1.79,2.02)12.57 (6.90,21.14)22.12 (12.26,36.86)1.92 (1.81,2.04) Southeast Asia71.11 (46.28,103.62)101.72 (65.65,147.79)1.14 (1.07,1.20)15.11 (8.38,25.01)21.65 (12.01,36.82)1.15 (1.09,1.22) Oceania73.88 (48.18,108.01)78.38 (52.05,115.83)0.07 (0.01,0.14)15.59 (8.48,26.06)16.61 (9.11,27.48)0.09 (0.02,0.15) Central Asia104.02 (68.49,149.91)118.68 (77.28,170.62)0.81 (0.51,1.11)22.05 (12.50,36.28)25.14 (13.98,41.74)0.82 (0.52,1.13) Central Europe115.89 (77.34,164.23)150.60 (99.31,214.58)1.06 (0.97,1.14)24.56 (13.58,40.77)31.96 (17.85,52.57)1.07 (0.99,1.15) Eastern Europe133.31 (86.48,192.79)143.62 (93.38,205.80)0.59 (0.34,0.85)28.15 (15.78,47.33)30.33 (17.02,50.64)0.60 (0.35,0.86) High-income Asia Pacific241.96 (160.25,341.38)297.08 (197.53,419.61)0.61 (0.57,0.66)51.46 (28.78,84.81)63.25 (35.40,103.92)0.62 (0.57,0.66) Australasia582.50 (380.91,825.28)811.90 (629.68,1041.59)1.14 (1.01,1.27)121.78 (67.28,202.10)170.34 (106.42,262.05)1.14 (1.01,1.27) Western Europe310.63 (214.97,425.03)361.01 (246.60,499.69)0.53 (0.48,0.58)65.27 (37.93,105.01)75.78 (44.42,121.85)0.53 (0.48,0.59) Southern Latin America241.43 (161.52,336.70)291.93 (198.73,408.38)0.65 (0.59,0.70)50.87 (28.39,83.00)61.34 (34.84,99.61)0.65 (0.59,0.70) High-income North America339.36 (225.18,480.64)331.57 (218.72,477.48)-0.26 (-0.40,-0.13)71.28 (39.37,116.81)69.0 7 (38.67,114.63)-0.28 (-0.41,-0.14) Caribbean177.82 (117.24,254.02)187.13 (125.42,263.03)0.32 (0.27,0.36)37.40 (21.14,62.64)39.38 (22.18,65.44)0.32 (0.27,0.37) Andean Latin America215.61 (140.78,308.45)268.81 (177.88,379.71)0.78 (0.74,0.82)45.52 (25.64,76.29)56.65 (31.07,95.60)0.79 (0.74,0.83) Central Latin America197.93 (130.92,281.33)211.24 (140.01,301.41)0.26 (0.23,0.29)41.76 (23.06,69.04)44.53 (25.30,73.65)0.26 (0.23,0.29) Tropical Latin America169.20 (110.13,244.69)197.21 (128.54,283.47)0.62 (0.56,0.67)35.43 (19.55,59.26)41.23 (23.13,68.36)0.63 (0.57,0.68) North Africa and Middle East164.47 (110.65,230.78)196.47 (131.44,274.81)0.83 (0.75,0.90)34.55 (19.60,55.81)41.24 (23.27,67.46)0.83 (0.76,0.91) South Asia85.35 (55.34,124.57)129.10 (83.70,186.00)1.44 (1.37,1.51)17.89 (10.17,29.71)27.14 (15.24,45.72)1.46 (1.39,1.52) Central Sub-Saharan Africa94.10 (61.30,136.63)90.55 (58.59,130.23)0.10 (-0.11,0.32)19.50 (10.74,32.33)18.93 (10.75,31.62)0.15 (-0.07,0.36) Eastern Sub-Saharan Africa78.94 (51.21,115.37)92.43 (60.01,135.13)0.63 (0.53,0.73)16.51 (9.19,27.77)19.44 (10.90,32.67)0.67 (0.56,0.77) Southern Sub-Saharan Africa140.54 (90.76,202.89)149.61 (96.71,213.61)0.33 (0.29,0.36)29.48 (16.63,49.58)31.17 (17.42,52.07)0.31 (0.27,0.35) Western Sub-Saharan Africa100.27 (64.69,146.13)113.39 (73.40,164.14)0.65 (0.52,0.77)20.96 (11.78,35.43)23.87 (13.40,40.06)0.68 (0.55,0.80)ASR age-standardized rate, DALYs disability-adjusted life-years, EAPC estimated annual percentage change, UI uncertainty interval, CI confidence interval, SDI sociodemographic index Table 2Age-standardized rates of prevalence and disability-adjusted life years of bulimia nervosa in 2021, alongside their Temporal trend from 1990 to 2021, presented at global and regional levelsLocationPrevalence (95% UI)DALYs (95% UI)ASR in 1990(per 100,000 population)ASR in 2021(per 100,000 population)1990–2021EAPC (95% CI)ASR in 1990(per 100,000 population)ASR in 2021(per 100,000 population)1990–2021EAPC (95% CI)Global130.66 (86.28,184.95)156.85 (103.39,221.39)0.66 (0.61,0.72)27.53 (15.53,45.89)33.05 (18.64,55.29)0.67 (0.62,0.72)Gender Male98.03 (65.00,142.88)121.98 (81.38,177.10)0.80 (0.74,0.87)20.87 (11.72,34.84)26.01 (14.68,43.26)0.81 (0.75,0.88) Female164.09 (108.15,230.02)192.73 (125.99,269.49)0.58 (0.53,0.62)34.35 (19.24,56.42)40.30 (22.68,66.23)0.58 (0.54,0.63)Socio-demographic index High-middle SDI129.57 (87.51,181.74)173.19 (114.76,245.25)1.11 (1.01,1.22)27.41 (15.72,45.14)36.70 (20.43,60.71)1.13 (1.02,1.23) High SDI273.11 (182.59,381.40)311.26 (211.22,435.75)0.33 (0.28,0.38)57.53 (32.24,93.91)65.38 (37.29,106.61)0.32 (0.27,0.37) Low-middle SDI93.51 (60.84,134.77)127.72 (83.03,183.55)1.12 (1.06,1.19)19.62 (11.14,32.63)26.86 (14.97,45.07)1.14 (1.08,1.21) Low SDI83.52 (54.24,121.96)96.69 (62.85,140.31)0.62 (0.51,0.74)17.44 (9.79,29.13)20.31 (11.42,33.98)0.65 (0.54,0.77) Middle SDI96.20 (62.58,139.63)144.24 (94.08,205.37)1.42 (1.37,1.47)20.34 (11.34,33.97)30.48 (17.07,51.27)1.42 (1.37,1.47)Region East Asia58.73 (37.95,86.68)103.0 2 (66.60,153.25)1.91 (1.79,2.02)12.57 (6.90,21.14)22.12 (12.26,36.86)1.92 (1.81,2.04) Southeast Asia71.11 (46.28,103.62)101.72 (65.65,147.79)1.14 (1.07,1.20)15.11 (8.38,25.01)21.65 (12.01,36.82)1.15 (1.09,1.22) Oceania73.88 (48.18,108.01)78.38 (52.05,115.83)0.07 (0.01,0.14)15.59 (8.48,26.06)16.61 (9.11,27.48)0.09 (0.02,0.15) Central Asia104.02 (68.49,149.91)118.68 (77.28,170.62)0.81 (0.51,1.11)22.05 (12.50,36.28)25.14 (13.98,41.74)0.82 (0.52,1.13) Central Europe115.89 (77.34,164.23)150.60 (99.31,214.58)1.06 (0.97,1.14)24.56 (13.58,40.77)31.96 (17.85,52.57)1.07 (0.99,1.15) Eastern Europe133.31 (86.48,192.79)143.62 (93.38,205.80)0.59 (0.34,0.85)28.15 (15.78,47.33)30.33 (17.02,50.64)0.60 (0.35,0.86) High-income Asia Pacific241.96 (160.25,341.38)297.08 (197.53,419.61)0.61 (0.57,0.66)51.46 (28.78,84.81)63.25 (35.40,103.92)0.62 (0.57,0.66) Australasia582.50 (380.91,825.28)811.90 (629.68,1041.59)1.14 (1.01,1.27)121.78 (67.28,202.10)170.34 (106.42,262.05)1.14 (1.01,1.27) Western Europe310.63 (214.97,425.03)361.01 (246.60,499.69)0.53 (0.48,0.58)65.27 (37.93,105.01)75.78 (44.42,121.85)0.53 (0.48,0.59) Southern Latin America241.43 (161.52,336.70)291.93 (198.73,408.38)0.65 (0.59,0.70)50.87 (28.39,83.00)61.34 (34.84,99.61)0.65 (0.59,0.70) High-income North America339.36 (225.18,480.64)331.57 (218.72,477.48)-0.26 (-0.40,-0.13)71.28 (39.37,116.81)69.0 7 (38.67,114.63)-0.28 (-0.41,-0.14) Caribbean177.82 (117.24,254.02)187.13 (125.42,263.03)0.32 (0.27,0.36)37.40 (21.14,62.64)39.38 (22.18,65.44)0.32 (0.27,0.37) Andean Latin America215.61 (140.78,308.45)268.81 (177.88,379.71)0.78 (0.74,0.82)45.52 (25.64,76.29)56.65 (31.07,95.60)0.79 (0.74,0.83) Central Latin America197.93 (130.92,281.33)211.24 (140.01,301.41)0.26 (0.23,0.29)41.76 (23.06,69.04)44.53 (25.30,73.65)0.26 (0.23,0.29) Tropical Latin America169.20 (110.13,244.69)197.21 (128.54,283.47)0.62 (0.56,0.67)35.43 (19.55,59.26)41.23 (23.13,68.36)0.63 (0.57,0.68) North Africa and Middle East164.47 (110.65,230.78)196.47 (131.44,274.81)0.83 (0.75,0.90)34.55 (19.60,55.81)41.24 (23.27,67.46)0.83 (0.76,0.91) South Asia85.35 (55.34,124.57)129.10 (83.70,186.00)1.44 (1.37,1.51)17.89 (10.17,29.71)27.14 (15.24,45.72)1.46 (1.39,1.52) Central Sub-Saharan Africa94.10 (61.30,136.63)90.55 (58.59,130.23)0.10 (-0.11,0.32)19.50 (10.74,32.33)18.93 (10.75,31.62)0.15 (-0.07,0.36) Eastern Sub-Saharan Africa78.94 (51.21,115.37)92.43 (60.01,135.13)0.63 (0.53,0.73)16.51 (9.19,27.77)19.44 (10.90,32.67)0.67 (0.56,0.77) Southern Sub-Saharan Africa140.54 (90.76,202.89)149.61 (96.71,213.61)0.33 (0.29,0.36)29.48 (16.63,49.58)31.17 (17.42,52.07)0.31 (0.27,0.35) Western Sub-Saharan Africa100.27 (64.69,146.13)113.39 (73.40,164.14)0.65 (0.52,0.77)20.96 (11.78,35.43)23.87 (13.40,40.06)0.68 (0.55,0.80)ASR age-standardized rate, DALYs disability-adjusted life-years, EAPC estimated annual percentage change, UI uncertainty interval, CI confidence interval, SDI sociodemographic index Age-standardized rates of prevalence and disability-adjusted life years of bulimia nervosa in 2021, alongside their Temporal trend from 1990 to 2021, presented at global and regional levels ASR age-standardized rate, DALYs disability-adjusted life-years, EAPC estimated annual percentage change, UI uncertainty interval, CI confidence interval, SDI sociodemographic index Fig. 1The EAPC of prevalence (A, B) and DALYs (C, D) for anorexia nervosa at global and regional levels. EAPC estimated annual percentage change, DALYs disability-adjusted life years The EAPC of prevalence (A, B) and DALYs (C, D) for anorexia nervosa at global and regional levels. EAPC estimated annual percentage change, DALYs disability-adjusted life years During 1990–2021, females exhibited consistently higher age-standardized prevalence and DALY rates than males for AN and BN (Tables 1 and 2). However, males demonstrated significantly higher EAPC for AN and BN, suggesting a greater acceleration in disease burden among males. In 2021, high-SDI regions carried the highest age-standardized prevalence and DALY rates for AN and BN. Throughout 1990–2021, all SDI quintiles observed consistent increases in age-standardized prevalence and DALY rates. The most substantial increases occurred in middle-SDI regions for AN (Fig. 1) and BN (Supplementary Fig. 1). We further analyzed sex-specific disease burden for AN (Fig. 2) and BN (Supplementary Fig. 2) across age groups. For AN and BN, females exhibited higher age-standardized prevalence and DALY rates than males in each age subgroup. The peak pattern for BN occurred in the 20–29 years age group, whereas AN peaked in the 15–24 years cohort. Fig. 2The prevalence rates (A) and DALY rates (B) of anorexia nervosa in different age groups globally in 2021. DALYs disability-adjusted life year The prevalence rates (A) and DALY rates (B) of anorexia nervosa in different age groups globally in 2021. DALYs disability-adjusted life year The regional and National burden of anorexia nervosa and bulimia nervosa Regional level burden and trends for AN and BN are presented in Tables 1 and 2, respectively. In 2021, Australasia demonstrated the highest age-standardized prevalence and DALY rates for both disorders. During the observation period, AN showed increased ASRs across all regions apart from Oceania and Central Sub-Saharan Africa (Fig. 1B and D). BN exhibited climbed ASRs across all regions except high-income North America (Supplementary Figs. 1B, D). For both AN and BN, East Asia demonstrated the most pronounced ASRs increases in prevalence and DALY. The nation level burden and trends for AN and BN are presented in Supplementary Tables 1 and 2, respectively. In 2021, Monaco exhibited the highest age-standardized DALY and prevalence rates for AN (Figs. 3A and 4A), while Australia showed the highest corresponding rates for BN (Supplementary Figs. 3A, C). During 1990–2021, increasing trends in age-standardized DALY and prevalence rates were observed in 166 and 165 distinct countries/territories for AN, respectively. Similarly, 182 nations/territories demonstrated rising rates for BN. Equatorial Guinea recorded the most rapid EAPC acceleration in both age-standardized DALY and prevalence rates for AN (Figs. 3B and 4B) and BN (Supplementary Figs. 3B, D). Fig. 3Age-standardized DALY rates in 2021 (A), and the estimated annual percentage change of age-standardized DALY rates from 1990 to 2021 (B), for anorexia nervosa in 204 countries or territories. DALY disability-adjusted life year, EAPC estimated annual percentage change Age-standardized DALY rates in 2021 (A), and the estimated annual percentage change of age-standardized DALY rates from 1990 to 2021 (B), for anorexia nervosa in 204 countries or territories. DALY disability-adjusted life year, EAPC estimated annual percentage change Fig. 4Age-standardized prevalence rates in 2021 (A), and the estimated annual percentage change of age-standardized prevalence rates from 1990 to 2021 (B), for anorexia nervosa in 204 countries or territories. EAPC estimated annual percentage change Age-standardized prevalence rates in 2021 (A), and the estimated annual percentage change of age-standardized prevalence rates from 1990 to 2021 (B), for anorexia nervosa in 204 countries or territories. EAPC estimated annual percentage change Association between the burden of anorexia nervosa and bulimia nervosa and the SDI Associations between AN burden estimates and SDI levels across geographical regions were investigated from 1990 to 2021 (Fig. 5). Throughout the 32-year observation period, all 21 regions exhibited increasing SDI values. Most geographical regions demonstrated positive correlations between burden estimates and SDI levels. Below the SDI threshold of 0.7, burden estimates plateaued with incremental SDI growth; however, beyond this threshold, these estimates accelerated substantially with rising SDI. Fig. 5Age-standardized rates of prevalence (A) and DALYs (B) of anorexia nervosa were classified according to SDI in 21 global disease burden regions, 1990-2021. The blue line represents the average expected relationship between SDI and the prevalence (A) and DALYs (B) of anorexia nervosa, and for each region, the points on the left depict estimates for each year from 1990 to 2021. DALYs disability-adjusted life years, SDI socio-demographic index Age-standardized rates of prevalence (A) and DALYs (B) of anorexia nervosa were classified according to SDI in 21 global disease burden regions, 1990-2021. The blue line represents the average expected relationship between SDI and the prevalence (A) and DALYs (B) of anorexia nervosa, and for each region, the points on the left depict estimates for each year from 1990 to 2021. DALYs disability-adjusted life years, SDI socio-demographic index Analyzed the relationship between the prevalence and DALY rates of BN and SDI in 21 regions worldwide from 1990 to 2021 (Supplementary Fig. 4). Similar to the trend of AN, there is a significant positive correlation between the estimated burden of BN and SDI. The influential factors of EAPC The burden of AN is increasing in most countries (EAPC > 0). In total, the EAPC shows a mild positive correlation with age-standardized prevalence rates and DALY rates (Fig. 6). In 2021, the SDI demonstrated a positive correlation with the EAPC for both AN prevalence (r = 0.31, P < 0.05; Fig. 6B) and DALY rates (r = 0.29, P < 0.05; Fig. 6D). Fig. 6The correlation between EAPC in age-standardized prevalence rates and the age-standardized prevalence rates in 2021 (A); the correlation between EAPC in age-standardized prevalence rates and SDI in 2021 (B); the correlation between EAPC in age-standardized DALY rates and the age-standardized DALY rates in 2021 (C); the correlation between EAPC in age-standardized DALY rates and SDI in 2021 (D). The circle represents the country, and the size of the circle represents the age standardized prevalence rate or DALYs for that country or region in 2021. The R index and p-value in the figure were derived from Pearson correlation analysis. EAPC estimated annual percentage change, SDI socio-demographic index The correlation between EAPC in age-standardized prevalence rates and the age-standardized prevalence rates in 2021 (A); the correlation between EAPC in age-standardized prevalence rates and SDI in 2021 (B); the correlation between EAPC in age-standardized DALY rates and the age-standardized DALY rates in 2021 (C); the correlation between EAPC in age-standardized DALY rates and SDI in 2021 (D). The circle represents the country, and the size of the circle represents the age standardized prevalence rate or DALYs for that country or region in 2021. The R index and p-value in the figure were derived from Pearson correlation analysis. EAPC estimated annual percentage change, SDI socio-demographic index Significantly positive correlations were observed between 2021 SDI and EAPC for both age-standardized BN prevalence (r = 0.22, P < 0.05; Supplementary Fig. 5B) and DALY rates (r = 0.21, P < 0.05; Supplementary Fig. 5D). However, 2021 ASRs appeared to be unrelated with EAPC for prevalence (r = 0.1, P = 0.14; Supplementary Fig. 5A) or DALY rates (r = 0.093, P = 0.19; Supplementary Fig. 5C). Cross-country inequality analysis This study reveals absolutely relative disparities in AN and BN burden associated with SDI. From 1990 to 2021, the slope indices of inequality for AN prevalence (Fig. 7A) and DALY rates (Fig. 7C) between the highest and the lowest SDI countries increased from 38.98 to 47.26 per 100,000 and 8.36 to 10.29 per 100,000, respectively, representing disproportionate concentration of higher burden in high-SDI nations. Correspondingly, concentration indices (CI) reflecting relative gradient inequality also rose through this period (Fig. 7B and D). Fig. 7Health inequality regression curves (A) and concentration curves (B) for the prevalence and DALYs (Disability − Adjusted Life Years) of anorexia nervosa from 1990 to 2021 across the world Health inequality regression curves (A) and concentration curves (B) for the prevalence and DALYs (Disability − Adjusted Life Years) of anorexia nervosa from 1990 to 2021 across the world The slope indices of inequality for BN prevalence rose from 155.54 per 100,000 in 1990 to 204.73 per 100,000 in 2021. The concentration indices for prevalence increased from 0.239 in 1990 to 0.25 in 2021. Between 1990 and 2021, the slope indices of inequality for BN DALY rate rose from 33.06 to 43.21, with an absolute difference expansion of approximately 10.15. The concentration indices for DALY rate increased from 0.24 to 0.25, remaining relatively stable (Supplementary Fig. 6). Discussion Our study scrutinized the burden of eating disorders (EDs) caused by anorexia nervosa (AN) and bulimia nervosa (BN) globally, regionally, and nationally from 1990 to 2021. To our knowledge, this is the first study to use 2021 GBD data to examine the disease burden of eating disorders in individuals aged 5–49 years and the first to conduct a cross-country inequality analysis. We found that in most geographical regions, the burden of anorexia nervosa and bulimia nervosa rose with the increase in SDI, with the most significant increase seen in East Asia. Notably, the higher burden disproportionately concentrated in countries with higher SDI. Over the past 32 years, the global burden of AN and BN has been on the rise, reflecting the complex interplay of sociocultural, medical-diagnostic, and public health factors. Although the prevalence of body image and eating issues was once thought to be linked to Western culture, their increasing rate has become a global problem. A recent review showed that the proportion of body image and eating issues worldwide rose from 3.5% to 7.8% [31]. The “thin ideal” image of women in White- centered mainstream media is pervasive and widely promoted through mass and social media, contributing greatly to body image problems and eating pathology [32, 33]. On the social perspective, body shape and appearance are highly valued in certain cultures, where a slim body equals power, success, and happiness [32]. Enhanced diagnostic capacity is also the key factor as for the increased disease burden. Specifically, DSM-5 lowered the threshold for AN, such as removing the amenorrhea criterion and relaxing the weight cutoff, and for BN, such as reducing the required frequency of binge eating/purging from twice to once a week, leading to higher diagnostic sensitivity [34]. Furthermore, Studies from various countries have shown a rise in EDs referrals and hospital admissions shortly after the onset of the COVID-19 pandemic, particularly among younger populations [35, 36]. And, hospitalizations remained high even one year into the pandemic [37]. A rising trend in DALYs from AN and BN has been recorded globally, paralleling the changes in prevalence. The parallel rise indicates that the growing number of cases is accompanied by a significant escalation in total health loss, encompassing both years of life lost (YLLs) due to premature mortality and years lived with disability (YLDs). These findings further underscore that EDs are evolving into a pressing public health challenge, necessitating healthcare systems equipped to manage both acute manifestations and long-term disabling consequences of these conditions. The onset of EDs typically occurs in adolescence and early adulthood [38], also emerging in middle-aged and elderly individuals [39]. This is consistent with our findings. Our study revealed that the burden of AN and BN peaks reached in the age group of 20–24. This is a critical transitional period, such as receiving higher education, entering the workforce, and living independently, which may cultivate appropriate environment conducive to EDs behaviors [40]. In our study, the burden of AN and BN is higher in females than in males. Compared to males, females are involved in greater risk of body dissatisfaction and eating disorders [41, 42]. Moreover, females consistently pose a strong desire for being thinner. Social pressure to be thin, and efforts to achieve the ideal of thinness are associated with EDs behaviors and predict the occurrence of EDs [5, 43]. Although females have the preference to be thinner and are more likely to be affected by EDs, there’s a growing body of the rising incidence in males [44]. Previous studies have demonstrated that the prevalence of EDs for males has risen sharply over the past two decades, while males account for 10% of clinical eating disorders cases [45]. However, recent data suggests that this prevalence could reach as high as 25% [46, 47]. Other studies have reported that over the past 30 years, EDs prevalence has increased by 22% in males [48]. The accelerating growth of male DALYs is equally concerning. Between 1990 and 2021, the global age-standardized DALY rates among males increased from 5.13 to 5.60 per 100,000 for AN (EAPC = 0.34), and from 20.87 to 26.01 per 100,000 for BN (EAPC = 0.81). The data demonstrated the disabling impact of EDs on men is gradually worsening. These results concurrently indicate that the burden of male EDs has been seriously underestimated in the past. Globally, male EDs are gradually considered as a significant public-health problem, especially in Australia [49]. The underestimation of the male burden of EDs calls for a multi-dimensional explanation. Even though EDs are increasingly common in young males, the widespread cultural notion of EDs as a “female-only” condition means that males may fail to recognize EDs symptoms on themselves [50], leading to male patients being overlooked or even misdiagnosed. Besides, males tend to experience more stigma around EDs than females. This internalized stigma is likely one of the most prominent reasons for their reluctance to seek outpatient treatment [51]. Then, before 2013, DSM-IV excluded men from the diagnostic criteria for AN, leading to an underestimation of the burden of AN in men [52]. With the medical community’s growing recognition of male EDs and male patients’ heightened awareness of mental health conditions, diagnostic criteria and tools have gradually improved, resulting in more accurate diagnoses [53]. This has to some extent contributed to the increasing trend on the male EDs burden. Globally, the burden is greater in high-SDI regions, which is consistent with previous studies [26]. This could be linked to abundant medical resources, advanced diagnostic technology, and advanced health awareness. Also, the burden exhibits significant regional differences. The highest burden in Australasia exists, possibly due to the unique socio-cultural environment in this region. For instance, in New Zealand, the strive for body shape and appearance may increase the risk of eating disorders. One study on New Zealand women aged 18–24 reported that 65% of participants had significant body image distress, with 45% having uncontrolled diet, and with 36.5% presenting emotional eating tendencies, and these closely linked to body image dissatisfaction [54]. Remarkably, AN and BN have increased most significantly in East Asia. This could arise from rapid economic growth, industrialization, urbanization, and intensified globalization, as well as the penetration of dietary culture and aesthetic concepts [55]. At the national level, The DALYs burden of AN is highest in Monaco, while BN is highest in Australia. These national differences could be due to distinct genetic backgrounds and dietary behaviors, which may be influenced by environmental factors, including diet and gut microbiota [56, 57], and also relate to the allocation of medical resources and diagnostic levels. Moreover, varying cultural factors such as internet addiction, media influence, and lifestyle changes (stress, increased weight loss diets) increasingly impact our lives and affect EDs [9]. Our findings hold significant public health implications. The global burden of AN and BN continues to rise, necessitating enhanced capacity for screening and diagnosis within healthcare systems, particularly in primary care. The notably accelerated increase in the disease burden among males underscores the potential need for gender-sensitive assessment tools. At present, the treatment models for AN and BN in clinical practice are mainly developed based on female populations [58]. Previous study suggest that it is not imperative for men to prescribe fundamentally different treatment methods for eating disorders, but focus on making the treatment environment more inclusive, considering the unique needs of male patients, improving care, and ensuring positive long-term prognosis [44]. From an epidemiological perspective, the rapid growth in disease burden in middle-SDI regions, especially East Asia, calls for forward-looking policy making to mitigate the impact of “Westernized” aesthetic ideals and associated risk factors. Therefore, concerted efforts should be directed toward reducing the incidence of the disease with the most efficacious preventative measure. Future mechanistic studies are required to elucidate the targeted trigger behind the rising burden in male populations and middle-SDI regions. The study has some limitations that need to be considered. Firstly, data sources are lacking for some countries, and the GBD global modelling approach may not align with regional epidemiological patterns, which may underestimate disease burdens in low-income nations. Secondly, GBD 2021 failed to include binge eating disorder (BED) and other specified feeding or eating disorder (OSFED), thereby leading to the underestimation of the overall EDs burden. More importantly, evidence on women individuals aged 40–60 with perimenopause and menopause may be at high-risk period for EDs [59]. Also, this work highlights the need for future epidemiological research to investigate the prevalence of EDs aged 50 and above in large-scale nationally representative surveys, and to provide a more precise perspective on the disease burden. Conclusion This study used GBD 2021 data to provide a comprehensive description of the global, regional and national burden of eating disorders, revealing complex geographical and temporal trends in their prevalence worldwide. There are significant differences in disease burden between different SDI levels, regions, and countries. Public health education for improving healthy dietary habits and positive body image should optimally be integrated into primary prevention of the diseases in health care to mitigate the disease burden. New policies, financing, and implementation can put established best practices and numerous research findings into action, and concerted efforts could be developed from the regional perspective. Supplementary Information Supplementary material 1. Supplementary material 1. Supplementary material 2. Supplementary material 2. Supplementary material 3. Supplementary material 3. Supplementary material 4. Supplementary material 4. Supplementary material 5. Supplementary material 5. Supplementary material 6. Supplementary material 6. Supplementary material 7. Supplementary material 7. Supplementary material 8. Supplementary material 8.