Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Apr 1.
Published in final edited form as: Int J Eat Disord. 2022 Jan 27;55(4):505–517. doi: 10.1002/eat.23687

Comorbidity Between Eating Disorders and Psychiatric Disorders

Natalie C Momen 1, Oleguer Plana-Ripoll 1,2, Zeynep Yilmaz 1,3,4, Laura M Thornton 3, John J McGrath 1,5,6, Cynthia M Bulik 3,4,7, Liselotte Vogdrup Petersen 1,8
PMCID: PMC9763547  NIHMSID: NIHMS1854362  PMID: 35084057

Abstract

Objective:

Previous literature has established an increased risk of eating disorders among individuals with other psychiatric disorders and vice versa. However, often studies have focused on eating disorders as a single diagnostic entity and/or investigated selected psychiatric comorbidities. We conducted a comprehensive study, exploring bidirectional associations between different types of eating disorders and broad groups of all other psychiatric disorders, to identify patterns of comorbidity.

Method:

We included all people born in Denmark 1963-2010. We collected information on eating disorders and considered the risk of subsequent psychiatric disorders using Cox-proportional hazards regression. Absolute risks were calculated using competing risks survival analyses. We also considered prior psychiatric disorders and subsequent eating disorders.

Results:

An increased risk was seen for almost all disorder pairs of diagnoses evaluated. Following an anorexia nervosa (AN) diagnosis, the median hazard ratio for the different subsequent psychiatric disorders was 3.80 (range 2.48 to 6.15); following an other eating disorder (OED) diagnosis, it was 3.16 (range 2.05 to 5.14). After different psychiatric disorder diagnoses, the median hazard ratio was 2.66 for later AN (range 1.21 to 5.31) and 2.51 for later OED (range 1.25 to 4.10). Absolute risk of eating disorders were also higher among those with other psychiatric disorders than those without.

Discussion:

In this broad examination, we identified bidirectional increases in risk of comorbidity for those with both eating disorder diagnoses and psychiatric disorder diagnoses. Although our findings indicate different patterns of comorbidity between eating disorders, these variations were generally small.

Keywords: eating disorders, anorexia nervosa, bulimia nervosa, comorbidity, multimorbidity, epidemiology, registries

Introduction

Comorbidity between different types of mental disorders has been described as “pervasive” (Plana-Ripoll et al., 2019). Within mental disorders, the nature of this comorbidity can vary, related to the age of onset of different mental disorders, and the tendency for certain types of mental disorders to co-occur (e.g. ‘internalizing’ disorders such as neurotic and mood disorders(de Jonge et al., 2018; Kessler et al., 2011).) Individuals with eating disorders are at increased risk of other psychiatric disorders, and those with psychiatric disorders are at increased risk of eating disorders (McGrath et al., 2020; Plana-Ripoll et al., 2019). As eating disorders differ somewhat from other psychiatric disorders (for example in terms of the age of onset, gender distribution, and increased mortality), in this study, we investigate eating disorder comorbidity in greater detail.

A comprehensive study, carried out using the Danish national registers, revealed the bidirectional nature of eating disorder-psychiatric disorder comorbidity (Plana-Ripoll et al., 2019). However, this study used broadly defined diagnostic groups based on International Classification of Diseases (ICD) subchapters, combining all eating disorder diagnoses into a single diagnostic category. Another recent Danish study focused only on comorbidity in anorexia nervosa (AN) patients (Steinhausen et al., 2021). A 2006 US study highlighted that the patterns of mental disorder comorbidity varied across different types of eating disorders (Blinder, Cumella, & Sanathara, 2006). Although there were similar levels of occurrence for mood or neurotic disorders in people with different types of eating disorders, other differences emerged. Schizophrenia and other psychotic disorders were most common in patients with AN compared to those with other eating disorders; whereas substance use disorders were more common in patients with bulimia nervosa (BN) compared to patients with other eating disorders. Others have also found differences in comorbidity between patients with different types of eating disorder, for example increased risk of substance use disorders in AN (Gadalla & Piran, 2007) and higher prevalence of post-traumatic stress disorder (Kaye, Bulik, Thornton, Barbarich, & Masters, 2004) among those with eating disorders other than AN. These differences are not surprising, as eating disorders represent distinct but overlapping psychiatric conditions with unique precipitating factors and sequelae.

In order to explore these research questions, we adapted the methods developed in our previous broad study of psychiatric disorder comorbidity in the Danish population (Plana-Ripoll et al., 2019) to investigate eating disorder and psychiatric comorbidity in greater detail. We aimed to conduct a large, comprehensive study to describe the comorbid psychiatric diagnoses made after an eating disorder diagnoses, and the comorbid eating disorder diagnoses made after psychiatric disorder diagnoses. To do this, we aimed to calculate bidirectional relative and absolute risks of comorbidity between specific types of eating disorders (i.e. AN and bulimia nervosa [BN]) and other broadly defined psychiatric disorders using the Danish nationwide registers. Due to the diagnostic coding systems used in the Danish national registers, which did not include a specific diagnostic code for BN until 1995, we were able to provide bidirectional estimates (i.e. the risk of diagnosis with a psychiatric disorder after diagnosis with an eating disorder and vice versa) for AN and a broad category of “other eating disorders” (OED), which includes all eating disorders other than AN. It was possible for us to consider BN as a later diagnosis (i.e. prior diagnoses of psychiatric disorders and later diagnosis of BN) as BN was available as a diagnosis in ICD-10. We expected to find increased risk of subsequent psychiatric disorders among both those with prior AN or OED, as well as increased risk of both subsequent AN and OED among those with other prior psychiatric disorders. Although we expected absolute risks for all later disorders to be higher among those who had a prior disorder compared to those in the matched reference groups, we expected cumulative incidence to be higher in the direction of prior eating disorders-later psychiatric disorder than in the opposite direction due to the prevalence of eating disorders relative to that of other psychiatric disorders. Additionally, as seen in previous studies, we expected this risk to be highest temporally close to the diagnosis of the first disorder, after which the risk would fall slightly (although remaining higher than that in people with no eating disorder diagnoses) (Plana-Ripoll et al., 2019; Steinhausen et al., 2021). In addition to providing all results for both sexes combined, we produced sex-specific results.

Methods

To provide estimates of the bidirectional associations between eating disorders and other psychiatric disorders, we examined: i) the risk of diagnosis of a psychiatric disorder after being diagnosed with a specific eating disorder (i.e., prior eating disorder-later psychiatric disorder) and ii) the risk of diagnosis with a specific eating disorder after being diagnosed with a psychiatric disorder (i.e., prior psychiatric disorder-later eating disorder). The methods are described below; additional details are provided in the Supplementary Material.

The Danish Data Protection Agency and the Danish Health Data Authority approved this study.

Study Population and Ascertainment of Disorders of Interest

This population-based cohort study included all individuals born in Denmark between 1963-2010 (N=3,186,333), identified in the Danish Civil Registration System (Pedersen, 2011). Additionally, individuals had to be residing in the country at the start of follow-up. When investigating relative risks, follow-up started on 01/01/1969, or their sixth birthday, whichever occurred later (Supplementary Figure S1). For calculation of absolute risks, follow-up started on the latter of 01/01/2000 or their sixth birthday (after outpatient and emergency appointments became available in the Danish National Patient Register (Lynge, Sandegaard, & Rebolj, 2011; Schmidt et al., 2015), to ensure we were capturing all types of admission).

Eating Disorders

We identified individuals with eating disorders by combining data from the Danish Psychiatric Central Research Register (Mors, Perto, & Mortensen, 2011) and the Danish National Patient Register (Lynge et al., 2011; Schmidt et al., 2015). For each person, the date of diagnosis for each eating disorder was ascertained from the admission date as an inpatient, outpatient appointment, or emergency visit during which the diagnosis was made (after 6 years of age).

We were able to identify individuals with AN throughout the entire study period (prior to 1994 ICD-8 (World Health Organization, 1967) code 306.50; 1994 onwards ICD-10 (World Health Organization, 1993) F50.0 and F50.1; see Figure S1), and could therefore consider AN as both a prior and a later disorder of interest. In order be able to consider other types of eating disorders throughout the entire study period, we created a category called other eating disorders (OED), which comprised all eating disorders other than AN (ICD-8 306.58 and 306.59; ICD-10 F50.2, F50.3, F50.8 and F50.9). This categorization was necessary given the diagnostic codes available in the Danish registers at the time that did not directly correspond with the DSM categories.

Moreover, individuals with BN could not be identified in ICD-8 as a specific diagnostic code for BN did not exist; however, it could be identified from 1995 onwards, when ICD-10 was introduced into the Danish registers (ICD-10 F50.2 and F50.3). Therefore, we were able to consider both AN and BN as a later disorder of interest. Additionally, we categorised all eating disorders other than AN or BN as eating disorders not otherwise specified (EDNOS; ICD-8 306.58 and 306.59; ICD-10 F50.8 and F50.9).

In summary, we were able to consider AN and OED (all eating disorders other than AN) as both prior and later disorders of interest; additionally, we could consider BN and EDNOS (all disorders other than AN or BN) as later disorders of interest. This additional breakdown for eating disorders as the later disorder was included to make full use of the data available.

Psychiatric Disorders

The Danish Psychiatric Central Research Register provided information on other psychiatric disorders. As has been done in previous publications based on Danish registers (Momen et al., 2020; Pedersen et al., 2014; Plana-Ripoll et al., 2019), we used the nine Mental, Behavioral and Neurodevelopmental subchapter categories (other than eating disorders) described in ICD-10 (e.g. F00-F09, F10-19, F20-29, etc.) and the corresponding diagnoses in ICD-8 (Table 1). For each individual in the study, the date of onset for each psychiatric disorder was defined as the date of first contact (inpatient, outpatient, or emergency visit).

Table 1.

Psychiatric disorders and eating disorders: International Classification of Disease codes used for definition and the frequency of cases among those born in Denmark between 1963-2010 (n=3,186,133)

Disorder Definition Cases (n)
ICD-10 ICD-8 equivalents
Psychiatric disorders
Organic F00-F09 290.09, 290.10, 290.11, 290.18, 290.19, 292.x9, 293.x9, 294.x9, 309.x9 2,892
Substance Use F10-F19 291.x9, 294.39, 303.x9, 303.20, 303.28, 303.90, 304.x9 56,168
Schizophrenia F20-F29 295.x9, 296.89, 297.x9, 298.29-298.99, 299.04, 299.05, 299.09, 301.83 40,302
Mood F30-F39 296.x9 (excluding 296.89), 298.09, 298.19, 300.49, 301.19 105,733
Neurotic F40-F48 300.x9 (excluding 300.49), 305.x9, 305.68, 307.99 183,557
Personality F60 301.x9 (excluding 301.19), 301.80, 301.81, 301.82, 301.84 63,526
Intellectual F70-F79 311.xx, 312.xx, 313.xx, 314.xx, 315.xx 13,589
Developmental F84 299.00, 299.01, 299.02, 299.03 22,449
Behavioral F90-F98 306·x9, 308·0x 62,368
Eating disorders
Anorexia nervosa F50.0, F50.1 306.50 11,811
OED F50.2, F50.3, F50.8, F50.9 306.58, 306.59 16,688
Bulimia nervosa F50.2, F50.3 - 7,947
EDNOS F50.8, F50.9 306.58, 306.59 10,737
Open in a new tab

Abbreviations: OED Other Eating Disorders, EDNOS Eating Disorders Not Otherwise Specified

Statistical Analysis

Our study examined the bidirectional associations between eating disorders (AN and OED) and nine other psychiatric disorders. We were also able to examine associations between prior psychiatric disorders and later BN and EDNOS. Depending on whether individuals had a diagnosis of the prior disorder of interest during follow-up, we estimated the relative and absolute risks of receiving a subsequent diagnosis of each later disorder during follow-up. All analyses were run in Stata 16.

Relative risks

We calculated hazard ratios (HRs) and 95% confidence intervals (CIs), using Cox proportional hazards regression with age as the underlying time scale. For each pair, we compared the rate of diagnosis with each later disorder between those exposed (i.e., diagnosed with the prior disorder of interest) and unexposed (i.e., not diagnosed with the prior disorder of interest) to each prior disorder. The bidirectional analyses were run separately. Each association was calculated using two models: i) Model A examined the association between a prior disorder and a later disorder, adjusted for sex and birthdate, which, in combination with underlying age in the models, also adjusted for calendar time; ii) Model B also adjusted for additional preceding psychiatric comorbidity: models included psychiatric disorder comorbidity with onset before the prior-disorder, but not with onset after the prior-disorder.

Follow-up ended at the first of the following events: diagnosis with the later disorder of interest, death, emigration, or end of follow-up (31/12/2016). A hierarchy was applied to eating disorders, so when considering OED as prior disorders, AN also acted as a censoring event, like death or emigration. For EDNOS, both AN and BN acted as censoring events. This ensured that an observed association was ascribed to a specific eating disorder rather than to several (Larsen et al., 2017). The longest period an individual was followed-up for to ascertain whether a later diagnosis of a psychiatric disorder or eating disorder was made was 47 years (1969-2016); therefore, the interval between diagnosis of a prior disorder and later disorder of interest could reach a maximum of 47 years.

The prior disorders were treated as time-varying exposures. Individuals were “unexposed” to the prior disorder, moving to the ‘exposed’ group on the day they were diagnosed for the first time with this disorder.

When the rates of a later disorder are not proportional over time among the exposed and unexposed groups, the estimates can be interpreted as an average hazard ratio over the entire follow-up period (Xu & O’Quigley, 2000).

Absolute risks

We also estimated the absolute risks: competing risks survival analyses were used to calculate cumulative incidence of diagnosis with a later disorder in the 15 years following diagnosis of a prior disorder, accounting for individuals’ simultaneous risk of diagnosis with the later disorder, dying, or emigrating.

For each prior disorder, a matched cohort was generated, which enabled comparison of absolute risks among people with a diagnosis of the prior disorder versus people without the diagnosis. Using the example AN as the prior disorder: we identified everyone with a diagnosis of AN and randomly matched each of these people with up to three age- and sex-matched individuals who had not received an AN diagnosis at or before the age the index person was diagnosed. Cumulative incidence of diagnoses of each later disorder were calculated for those with AN and for the matched “reference” cohort. More detailed information can be found in the Supplementary Appendix.

Results

In total, 3,186,133 individuals were born in Denmark between 1963 and 2010. The study population used to provide estimates of relative risk consisted of 3,094,715 Danish residents (1,507,760 females and 1,586,955 males; see Supplementary Figure S2), each followed for up to 47 years, resulting in a total follow-up time of 74.99 million person years. During this time, 42,410 individuals died and 86,292 emigrated. The number of cases with each disorder in the population are presented in Table 1 and sex and year of birth proportions are presented in Table 2. The study population used to calculate absolute risks included 3,048,258 individuals (1,485,601 females and 1,562,657 males), each followed for up to 16 years; total follow-up was 42.10 million years.

Table 2.

Baseline characteristics of the population born in Denmark between 1963-2010 (n=3,186,133)

Baseline characteristics of the study population N %
Sex Male 1,635,716 51.3
Female 1,550,617 48.7
Birth year 1963-1970 627,036 19.7
1971-1980 672,913 21.1
1981-1990 558,990 17.5
1991-2000 677,175 21.3
2001-2010 650,219 20.4
Open in a new tab

General findings for AN and OED are described below. Bidirectional pairwise associations and absolute risks for all eating disorder-psychiatric disorder pairs are shown in Figures 1–4 and Supplementary Figures S3–S10. Results for both Model A (adjusted for age and sex) and Model B (adjusted for age, sex and prior psychiatric disorders) are shown in the supplement; however, here only results for Model B are discussed because these are also adjusted for prior comorbidity.

Figure 1. The bidirectional associations between anorexia nervosa and psychiatric disorders.

Figure 1.

Open in a new tab

The panels show the hazard ratios (HRs) and 95% confidence intervals (CIs) of the associations between anorexia nervosa and psychiatric disorders for all persons (men and women). Estimates were obtained via Cox proportional hazards models with age as the underlying time scale, adjusting for sex, calendar time and other psychiatric disorders with onset before the prior disorder under study. In the left panel, the HRs show the risk of being diagnosed with a psychiatric disorder following a diagnosis of anorexia nervosa (i.e., anorexia nervosa is the prior disorder), compared to people without a diagnosis of anorexia nervosa. The right panel displays the HRs for the risk of being diagnosed with anorexia nervosa following a psychiatric condition diagnosis (i.e., the psychiatric disorder of interest is the prior disorder), compared with people without the psychiatric disorder.

Footnote: The arrow shown for organic disorders when AN was a later disorder indicates that the 95% CI goes beyond the scale of the graph, i.e., is above 10.

Figure 4A and 4B. Absolute risks for psychiatric disorder and other eating disorders.

Figure 4A and 4B.

Figure 4A and 4B.

Open in a new tab

Figure A shows estimates of absolute risks for a later diagnosis of a psychiatric disorder, following a diagnosis of other eating disorders, for all persons. Figure B shows absolute risks for a later diagnosis of other eating disorders, following a diagnosis of a psychiatric disorder, for all persons. The cumulative incidence per 100 persons (solid lines for those with the prior disorder of interest, dashed lines for the matched reference groups) of receiving a diagnosis of each later disorder of interest, after a diagnosis of the prior disorder of interest. Shaded grey areas around the lines for those with the prior disorder of interest represent 95% CIs (in some panels obscured by the estimates line). The horizontal axes show the time since first diagnosis of the prior disorder. The vertical axes show the cumulative incidence per 100 persons (and 95% CI).

NB. Numbers for prior organic disorders-later other eating disorders were too small to meet the regulations for reporting Danish register data

Relative Risks

Model B HRs for all persons (both males and females) are shown for each direction for AN and OED in Figures 1 and 2. All model B estimates were attenuated compared with their corresponding model A estimates (Supplementary Tables S1 and S2). Sex-specific HRs can be seen in Supplementary Figures S3 and S4.

Figure 2. The bidirectional associations between other eating disorders and mental disorders.

Figure 2.

Open in a new tab

The panels show the HRs and 95% CIs of the associations between other eating disorders and mental disorder for all persons (men and women). Estimates were obtained via Cox proportional hazards models with age as the underlying time scale, adjusting for sex, calendar time and other psychiatric disorders with onset before the prior disorder under study. In the left panel, the HRs show the risk of being diagnosed with a psychiatric disorder following a diagnosis of other eating disorders (i.e., other eating disorders is the prior disorder), compared with people without a diagnosis of other eating disorders. The right panel displays the HRs for the risk of being diagnosed with other eating disorders following a psychiatric disorder diagnosis (i.e., the psychiatric disorder of interest is the prior disorder), compared with people without the psychiatric disorder.

Prior Eating Disorders-Later Psychiatric Disorders

Among persons who had a prior diagnosis of either AN or OED, risks of subsequently receiving a diagnosis of each psychiatric disorder of interest were elevated; all point estimates were above 1 and none of the 95% CIs included 1. The rate of diagnosis of any later psychiatric disorder was almost four times higher in those with an AN diagnosis compared to those with no AN diagnosis (HR 3.93; 95% CI 3.78, 4.09). For those with an OED diagnosis, the rate was over five times higher than for those with no OED diagnosis (HR 5.15; 95% CI 4.94, 5.37). For the different ICD-10 subchapter categories of psychiatric disorders, the median HR was 3.38 following an AN diagnosis (range 2.48 to 6.15) and 3.07 following an OED diagnosis (range 2.05 to 5.14).

Prior Psychiatric Disorders-Later Eating Disorders

For all persons, elevated risks for AN and OED were observed after diagnosis of psychiatric disorders of interest; however, risks did not differ significantly for later intellectual disabilities for individuals with either prior AN or prior OED, compared to people without these prior disorders. Following a diagnosis of any ICD-10 subchapter category of psychiatric disorder, the risk of an AN diagnosis was over four times greater (HR 4.31; 95% CI 4.10, 4.53) and the risk of an OED diagnosis was over five times greater (HR: 5.41; 95% CI 5.20, 5.64) than the risk in people with no psychiatric disorder diagnoses. For the different ICD-10 subchapter categories of psychiatric disorders, the median HR was 2.53 for later AN (range 1.21 to 5.31) and 2.21 for later OED (range 1.25 to 4.10).

The HRs for BN and EDNOS can be seen in Supplementary Figures S5 and S6. In summary, following a diagnosis of any ICD-10 subchapter category of psychiatric disorder, the risk of an BN diagnosis was over four times greater (HR 4.27; 95% CI 4.03, 4.53). The median HR was 2.98 for later AN (range 0.53 to 3.75).

Absolute Risks

Figure 3a shows the sex-specific absolute risk of receiving each psychiatric diagnosis after a previous diagnosis of AN. Conversely, Figure 3b shows the risk of receiving an AN diagnosis after a prior diagnosis of each psychiatric disorder. The cumulative incidences when OED were the prior and later disorder of interest are shown in Figures 4a and 4b. Sex-specific absolute risks, as well as graphs for BN and EDNOS as the later disorders of interest, are in the Supplementary Material.

Figure 3A and 3B. Absolute risks for psychiatric disorders and anorexia nervosa.

Figure 3A and 3B.

Figure 3A and 3B.

Open in a new tab

Figure A shows estimates of absolute risks for a later diagnosis of a mental disorder, following a diagnosis of anorexia nervosa, for all persons. Figure B shows absolute risks for a later diagnosis of anorexia nervosa, following a diagnosis of a psychiatric disorder, for all persons. The cumulative incidence per 100 persons (solid lines for those with the prior disorder of interest, dashed lines for the matched reference groups) of receiving a diagnosis of each later disorder of interest, after a diagnosis of the prior disorder of interest. Shaded grey areas around the lines for those with the prior disorder of interest represent 95% CIs (in some panels obscured by the estimates line). The horizontal axes show the time since first diagnosis of the prior disorder. The vertical axes show the cumulative incidence per 100 persons (and 95% CI).

NB. Numbers for prior organic disorders-later anorexia nervosa were too small to meet the regulations for reporting Danish register data

Prior Eating Disorders-Later Psychiatric Disorders

For all pairs of prior eating disorders and later psychiatric disorders, cumulative incidence for all psychiatric disorders was higher among those with prior eating disorders than those without. The highest cumulative incidence proportion for all persons was seen for subsequent neurotic disorders in those with prior AN: within 1 year of AN diagnosis, 6.7% had a diagnosis of a neurotic disorder (95% CI 6.2-7.4%), rising to 28.8% by 15 years (95% CI 27.6-30.1%). Cumulative incidence in the reference group, was 0.7% 1 year into follow-up (95% CI 0.7-0.8%); at 15 years, it was at 9.0% (95% CI 8.7-9.4%).

Prior Psychiatric Disorders-Later Eating Disorders

For all pairs, cumulative incidence for eating disorders was higher among those with prior psychiatric disorders than without. The highest cumulative incidence for personality disorders-OED: at 0.6% (95% CI 0.6-0.7%) 1 year after diagnosis of the personality disorder, and at 2.49% (95% CI 2.33-2.65%) after 15 years. For the matched reference group, the cumulative incidence reached 0.05% (95% CI 0.04-0.6%) and 0.5% (95% CI 0.4-0.5%) at the respective time points.

Discussion

This population-based study, comprising over 3 million individuals, shows bidirectional comorbidity between specific types of eating disorders and a broad range of other psychiatric disorders. We found that relative risks of receiving an AN or an OED diagnosis are elevated among those with psychiatric disorders, and vice versa; the magnitude of the risks differed between the two types of eating disorders slightly, but not substantially, for most eating disorder-psychiatric disorder pairs. Absolute risks for each of the later disorders were also raised among those with each prior disorder of interest, for example personality, mood and neurotic disorders were all diagnosed within 15 years for more than 15% of those with an AN or OED diagnosis.

In line with Plana-Ripoll et al. (Plana-Ripoll et al., 2019), our findings demonstrate that eating disorders were associated with both prior and later psychiatric disorders. This is also supported by other studies: although not all have looked at ordering of diagnoses, several studies have reported that a range of psychiatric disorders are overrepresented in individuals with AN and BN, for example neurotic/anxiety disorders (Bulik, Sullivan, Fear, & Joyce, 1997; Deep, Nagy, Weltzin, Rao, & Kaye, 1995; Silberg & Bulik, 2005), mood disorders (Deep et al., 1995; Piran, Kennedy, Garfinkel, & Owens, 1985), and personality disorders (Nilsson, Gillberg, Gillberg, & Rastam, 1999). Additionally, we observed that the pattern seen for absolute risks, i.e., that the rise in cumulative incidence of psychiatric disorders occurs relatively quickly in those with an eating disorder diagnosis in the period following diagnosis and then slows, is supported by the pattern seen in other Danish studies considering eating disorder comorbidity (Plana-Ripoll et al., 2019; Steinhausen et al., 2021).

Our findings highlight differing patterns of comorbidity between the types of eating disorders. Although direct comparisons were not made between the different eating disorders, compared to the study population without the respective prior disorders, the HRs suggested some small differences in bidirectional associations with other psychiatric disorders between the types of eating disorders. For instance, we found risk of subsequent neurotic disorders to be similar in those with AN and those with OED (HRs of 3.04 and 3.20 respectively); however, for those with prior neurotic disorder diagnoses, subsequent risk of EDNOS (HR 4.60) was higher than risk of AN, OED, or BN (HRs 3.23, 3.64 and 2.79 respectively). This difference in risk associated with order of diagnosis supports the findings of a US national survey of > 10,000 adolescents, which reported lower comorbidity of psychiatric disorders among individuals with AN, suggesting that the comorbidity of psychiatric disorders may be subsequent to a primary AN diagnosis in younger individuals (Swanson, Crow, Le Grange, Swendsen, & Merikangas, 2011). However, in contrast to our findings, some previous studies have found associations to be strongest between BN and comorbid psychiatric conditions, particularly disorders included in our category of “neurotic disorders”. For example, lifetime neurotic disorders and post-traumatic stress disorder have been observed to be less prevalent in those with AN than in those with BN (Hudson, Hiripi, Pope, & Kessler, 2007; Kaye et al., 2004). Additionally, a previous study found risk of subsequent BN to be higher than risk of AN (HR 15.2 vs 8.9, although not reaching statistical significance) in those with specific phobia (McGrath et al., 2020). Furthermore, binge eating and/or purging behaviours, which are criteria for a BN diagnosis, have been associated with increased risk for other psychiatric disorders including neurotic disorders (Hudson et al., 2007). A 2007 systematic review concluded that there was increased prevalence of alcohol use disorders in those with BN (although individual study results were divergent), but not those with AN (Gadalla & Piran, 2007). Another systematic review found that 34% of BN patients had a substance use disorder in their lifetime; whereas among AN patients, the proportion was 13% (Bahji et al., 2019). A 2021 study observed that immediate absolute risks of subsequent psychiatric disorders were more than double among people with an AN diagnosis, compared to people without (Steinhausen et al., 2021); a similar pattern was seen in both directions in our study, with the gap existing between the cumulative incidence of later disorders among those with and without the prior disorder of interest straight after initial diagnosis, and generally converging in later years of follow-up.

Observed differences could potentially reflect underlying differences in temperament associated with different eating disorders. Prior research (Atiye, Miettunen, & Raevuori-Helkamaa, 2015) has shown that individuals with AN are more likely to be compulsive, over-controlled, socially avoidant, conflict-averse, and risk avoidant. In contrast, individuals with BN, binge-eating disorder, and EDNOS who present with binge eating and/or purging tend to be more dysregulated and novelty seeking, and they are more likely to engage in impulsive behaviours such as self-harm, substance abuse, and other addictive behaviours. We found risks for subsequent substance use disorders to be increased for both those with both AN and OED, but the HR was slightly higher for OED (which included BN). Absolute risks observed in our study were lower for both those with AN and OED, however this may be due to limited follow-up time. Further, those with substance use disorders were at a higher risk of all eating disorder types. In terms of personality disorders, our study found an increase in personality disorders for both AN (relative risk 4.49) and OED (5.14), similar to a meta-analysis that found the rates of different personality disorders to be comparable among AN and BN cases, with the exception of obsessive-compulsive personality disorder, which was more prevalent in AN than BN (Martinussen et al., 2017). Although our findings do not always match those of other studies with regard to risks for each type of eating disorder, they provide additional evidence of the common co-occurrence of eating disorders and other psychiatric comorbidity, such as mood disorders and personality disorders (Godart et al., 2015; Sansone, Levitt, & Sansone, 2005). However, it is important to note that these patterns of elevated lifetime comorbidity observed in other studies may not apply to the specific examination of temporal order, as per the main objective of our study.

It is important to note that the presence of a time-ordered association may not reflect a causal association between prior and later disorder. There may be several reasons for the observed increases in risk. Some disorder pairs may share risk factors or mechanisms, for example genetic factors (Trace et al., 2013; Wade, Bulik, Neale, & Kendler, 2000), or infections (which have been linked to eating disorders (Breithaupt et al., 2019) and other psychiatric disorders (Benros et al., 2011; Benros et al., 2013)). Additionally, once a patient is referred for one psychiatric disorder, the likelihood of identifying additional diagnoses increases (Berkson, 1946), which may provide an explanation for the larger gap between absolute risks of later disorders in those with and without the prior disorder of interest; or, as is good clinical practice, changes may be made to diagnoses assigned as more symptoms develop. This is especially important in the interpretation of our results pertaining to intellectual disability and pervasive developmental disorders, both of which onset in early childhood and therefore in reality most likely to precede AN onset. Similarly, although personality disorders are pervasive in nature and often present from childhood, they are much less likely to be diagnosed until after the individual is assessed for another psychiatric disorder (e.g., depression, neurotic disorders, eating disorders). Therefore, it is difficult to capture the true temporality when it comes to earlier onset pervasive disorders, and our results should be interpreted with caution.

Our register-based study has several strengths. It uses the nationwide registers of Denmark which provide data on the whole population, giving a large sample size and minimising selection bias. Danish citizens have free and equal access to health care, thus any effect related to ability to afford private insurance/access to health care is reduced. It is mandatory in Denmark for all hospitals to report discharge diagnoses to central registries. The study is not susceptible to recall bias and self-reporting bias (limitations of cross-sectional surveys which rely on retrospective recall in those individuals alive at the time of the survey). We advance the science by considering specific types of eating disorders and providing estimates of relative and absolute risks.

Limitations should also be considered. First, misclassification could occur regarding ascertainment of disorders. Diagnoses for psychiatric disorders could have been made by a hospital doctor of any discipline and validity may also vary across diagnoses. Additionally, we do not have data on those who do not seek treatment for their disorders or those treated entirely in primary care settings. Although it is difficult to speculate regarding the effect on HRs, underestimation of absolute risks is likely. Misclassification can also occur for onset: the date of administrative onset, as recorded in the registers, is likely to be later than the actual onset of eating disorders and many psychiatric disorders. This may lead to incorrect temporal ordering, which could be the case for personality and pervasive developmental disorders, as mentioned above. However, date of diagnosis is consistently used in register-based studies as a proxy for date of onset. We note that it should be remembered that by considering “prior” and “later” diagnoses of disorders, we do not imply causality. We are describing the patterns of comorbid diagnoses in people who have an eating disorder diagnosis or a psychiatric disorder diagnosis, compared to people without such a diagnosis. Third, the length of follow-up is limited by the period of time for which the relevant registers have been available. Studies on mental disorders have suggested that risk of comorbid diagnoses is greatest in the years succeeding the index diagnosis (supported by the steeper rises in cumulative incidence in the first years after prior diagnosis (Momen et al., 2020; Plana-Ripoll et al., 2019)), but longer follow-up, which can be made as more register data become available, will be informative. Additionally, as a BN diagnosis was not possible to differentiate in the registers prior to the use of ICD-10 (in 1994), some BN may have been diagnosed prior to the first time it could be ascertained in the registers. Fourth, we were unable to control for several factors associated with both eating disorders and psychiatric disorders such as socioeconomic status, genetic factors, or childhood adversity. Fifth, although this is a comprehensive study and we separate types of eating disorders, the other types of psychiatric disorders were considered in broad groups. There may be additional variation in comorbidity between specific types of eating disorders and specific types of psychiatric disorders that our categories were unable to dissect. Sixth, symptom-level data are not available in Danish registers; as a result, we were unable to better differentiate the characteristics of eating disorders groups and the potential association of certain key behaviours (for instance, the presence of binge eating and/or purging) with the temporality of psychiatric comorbidity. Finally, the generalizability of our findings outside of Denmark may be limited. Patterns of comorbidity may vary in other countries, especially those with different health care structures.

CONCLUSIONS

This study provides a broad examination of the comorbidity between specific eating disorders and psychiatric disorders, presenting relative risks and absolute risks of the associations in both directions. It is a comprehensive examination of comorbidity between eating disorders and other psychiatric disorders. In keeping with other studies, we found that those with eating disorders have increased risks all psychiatric disorders (with around 40% of those with an AN or OED diagnosis receiving any subsequent psychiatric disorder diagnosis), and those with almost all types of psychiatric disorders have increased risks of eating disorders. Our findings also highlighted some differences in risks between types of eating disorders, however, these were generally small.

Supplementary Material

Supplementary Appendix Text Tables and Figures

Funding

ZY acknowledges grant funding from the NIMH (K01MH109782) and Brain and Behavior Research Foundation NARSAD Young Investigator Award (grant no. 28799). CMB is supported by NIMH (R01MH120170; R01MH124871; R01MH119084; R01MH118278); Brain and Behavior Research Foundation Distinguished Investigator Grant; Swedish Research Council (Vetenskapsrådet, award: 538-2013-8864); Lundbeck Foundation (Grant no. R276-2018-4581). JM is supported by the Danish National Research Foundation (Niels Bohr Professorship). JM is employed by The Queensland Centre for Mental Health Research which receives core funding from the Department of Health, Queensland Government. OPR is supported by a Lundbeck Foundation Fellowship (R345-2020-1588) and has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No 837180. LVP received funding from Lundbeck Foundation (Grant no. R276-2018-4581) and The Novo Nordisk Foundation (Grant no. NNF20OC0064993). The investigators conducted the research independently. The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or the decision to submit the manuscript for publication.

Footnotes

AVAILABILITY OF DATA, MATERIALS AND CODE

Access to individual-level Denmark data is governed by Danish authorities. These include the Danish Data Protection Agency, the Danish Health Data Authority, the Ethical Committee, and Statistics Denmark. Each scientific project must be approved before initiation, and approval is granted to a specific Danish research institution. Researchers at Danish research institutions may obtain the relevant approval and data. International researchers may gain data access if governed by a Danish research institution having needed approval and data access.

The code used in Stata can be provided on request to the corresponding author.

Conflicts of interest

CMB reports: Shire (grant recipient, Scientific Advisory Board member); Idorsia (consultant); Lundbeckfonden (grant recipient); Pearson (author, royalty recipient); Equip Health, Inc. (clinical advisory board).

REFERENCES

  1. Atiye M, Miettunen J, & Raevuori-Helkamaa A (2015). A meta-analysis of temperament in eating disorders. Eur Eat Disord Rev, 23(2), 89–99. doi: 10.1002/erv.2342 [DOI] [PubMed] [Google Scholar]
  2. Bahji A, Mazhar MN, Hudson CC, Nadkarni P, MacNeil BA, & Hawken E (2019). Prevalence of substance use disorder comorbidity among individuals with eating disorders: A systematic review and meta-analysis. Psychiatry Res, 273, 58–66. doi: 10.1016/j.psychres.2019.01.007 [DOI] [PubMed] [Google Scholar]
  3. Benros ME, Nielsen PR, Nordentoft M, Eaton WW, Dalton SO, & Mortensen PB (2011). Autoimmune diseases and severe infections as risk factors for schizophrenia: a 30-year population-based register study. Am J Psychiatry, 168(12), 1303–1310. doi: 10.1176/appi.ajp.2011.11030516 [DOI] [PubMed] [Google Scholar]
  4. Benros ME, Waltoft BL, Nordentoft M, Ostergaard SD, Eaton WW, Krogh J, & Mortensen PB (2013). Autoimmune diseases and severe infections as risk factors for mood disorders: a nationwide study. JAMA Psychiatry, 70(8), 812–820. doi: 10.1001/jamapsychiatry.2013.1111 [DOI] [PubMed] [Google Scholar]
  5. Berkson J (1946). Limitations of the application of fourfold table analysis to hospital data. Biometrics, 2(3), 47–53. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/21001024 [PubMed] [Google Scholar]
  6. Blinder BJ, Cumella EJ, & Sanathara VA (2006). Psychiatric comorbidities of female inpatients with eating disorders. Psychosom Med, 68(3), 454–462. doi: 10.1097/01.psy.0000221254.77675.f5 [DOI] [PubMed] [Google Scholar]
  7. Breithaupt L, Kohler-Forsberg O, Larsen JT, Benros ME, Thornton LM, Bulik CM, & Petersen L (2019). Association of Exposure to Infections in Childhood With Risk of Eating Disorders in Adolescent Girls. JAMA Psychiatry. doi: 10.1001/jamapsychiatry.2019.0297 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Bulik CM, Sullivan PF, Fear JL, & Joyce PR (1997). Eating disorders and antecedent anxiety disorders: a controlled study. Acta Psychiatr Scand, 96(2), 101–107. doi: 10.1111/j.1600-0447.1997.tb09913.x [DOI] [PubMed] [Google Scholar]
  9. de Jonge P, Wardenaar KJ, Lim CCW, Aguilar-Gaxiola S, Alonso J, Andrade LH, … Scott K (2018). The cross-national structure of mental disorders: results from the World Mental Health Surveys. Psychol Med, 48(12), 2073–2084. doi: 10.1017/S0033291717003610 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Deep AL, Nagy LM, Weltzin TE, Rao R, & Kaye WH (1995). Premorbid onset of psychopathology in long-term recovered anorexia nervosa. Int J Eat Disord, 17(3), 291–297. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/7773266 [PubMed] [Google Scholar]
  11. Gadalla T, & Piran N (2007). Co-occurrence of eating disorders and alcohol use disorders in women: a meta analysis. Arch Womens Ment Health, 10(4), 133–140. doi: 10.1007/s00737-007-0184-x [DOI] [PubMed] [Google Scholar]
  12. Godart N, Radon L, Curt F, Duclos J, Perdereau F, Lang F, … Flament MF (2015). Mood disorders in eating disorder patients: Prevalence and chronology of ONSET. J Affect Disord, 185, 115–122. doi: 10.1016/j.jad.2015.06.039 [DOI] [PubMed] [Google Scholar]
  13. Hudson JI, Hiripi E, Pope HG Jr., & Kessler RC (2007). The prevalence and correlates of eating disorders in the National Comorbidity Survey Replication. Biol Psychiatry, 61(3), 348–358. doi: 10.1016/j.biopsych.2006.03.040 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Kaye WH, Bulik CM, Thornton L, Barbarich N, & Masters K (2004). Comorbidity of anxiety disorders with anorexia and bulimia nervosa. Am J Psychiatry, 161(12), 2215–2221. doi: 10.1176/appi.ajp.161.12.2215 [DOI] [PubMed] [Google Scholar]
  15. Kessler RC, Ormel J, Petukhova M, McLaughlin KA, Green JG, Russo LJ, … Ustun TB (2011). Development of lifetime comorbidity in the World Health Organization world mental health surveys. Arch Gen Psychiatry, 68(1), 90–100. doi: 10.1001/archgenpsychiatry.2010.180 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Larsen JT, Munk-Olsen T, Bulik CM, Thornton LM, Koch SV, Mortensen PB, & Petersen L (2017). Early childhood adversities and risk of eating disorders in women: A Danish register-based cohort study. Int J Eat Disord, 50(12), 1404–1412. doi: 10.1002/eat.22798 [DOI] [PubMed] [Google Scholar]
  17. Lynge E, Sandegaard JL, & Rebolj M (2011). The Danish National Patient Register. Scandinavian Journal of Public Health, 39(7 Suppl), 30–33. doi: 10.1177/1403494811401482 [DOI] [PubMed] [Google Scholar]
  18. Martinussen M, Friborg O, Schmierer P, Kaiser S, Overgard KT, Neunhoeffer AL, … Rosenvinge JH (2017). The comorbidity of personality disorders in eating disorders: a meta-analysis. Eat Weight Disord, 22(2), 201–209. doi: 10.1007/s40519-016-0345-x [DOI] [PubMed] [Google Scholar]
  19. McGrath JJ, Lim CCW, Plana-Ripoll O, Holtz Y, Agerbo E, Momen NC, … de Jonge P. (2020). Comorbidity within mental disorders: a comprehensive analysis based on 145 990 survey respondents from 27 countries. Epidemiol Psychiatr Sci, 29, e153. doi: 10.1017/S2045796020000633 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Momen NC, Plana-Ripoll O, Agerbo E, Benros ME, Borglum AD, Christensen MK, … McGrath JJ (2020). Association between Mental Disorders and Subsequent Medical Conditions. N Engl J Med, 382(18), 1721–1731. doi: 10.1056/NEJMoa1915784 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Mors O, Perto GP, & Mortensen PB (2011). The Danish Psychiatric Central Research Register. Scandinavian Journal of Public Health, 39(7 Suppl), 54–57. doi: 10.1177/1403494810395825 [DOI] [PubMed] [Google Scholar]
  22. Nilsson EW, Gillberg C, Gillberg IC, & Rastam M (1999). Ten-year follow-up of adolescent-onset anorexia nervosa: personality disorders. J Am Acad Child Adolesc Psychiatry, 38(11), 1389–1395. doi: 10.1097/00004583-199911000-00013 [DOI] [PubMed] [Google Scholar]
  23. Pedersen CB (2011). The Danish Civil Registration System. Scand J Public Health, 39(7 Suppl), 22–25. doi: 10.1177/1403494810387965 [DOI] [PubMed] [Google Scholar]
  24. Pedersen CB, Mors O, Bertelsen A, Waltoft BL, Agerbo E, McGrath JJ, … Eaton WW (2014). A comprehensive nationwide study of the incidence rate and lifetime risk for treated mental disorders. JAMA Psychiatry, 71(5), 573–581. doi: 10.1001/jamapsychiatry.2014.16 [DOI] [PubMed] [Google Scholar]
  25. Piran N, Kennedy S, Garfinkel PE, & Owens M (1985). Affective disturbance in eating disorders. J Nerv Ment Dis, 173(7), 395–400. doi: 10.1097/00005053-198507000-00002 [DOI] [PubMed] [Google Scholar]
  26. Plana-Ripoll O, Pedersen CB, Holtz Y, Benros ME, Dalsgaard S, de Jonge P, … McGrath JJ (2019). Exploring Comorbidity Within Mental Disorders Among a Danish National Population. JAMA Psychiatry, 76, 259–270. doi: 10.1001/jamapsychiatry.2018.3658 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Sansone RA, Levitt JL, & Sansone LA (2005). The eating disorders. In Klykylo WM &Kay J(Eds.), Clinical Child Psychiatry (2nd ed., pp. 311–326). London: John Wiley. [Google Scholar]
  28. Schmidt M, Schmidt SA, Sandegaard JL, Ehrenstein V, Pedersen L, & Sorensen HT (2015). The Danish National Patient Registry: a review of content, data quality, and research potential. Clin Epidemiol, 7, 449–490. doi: 10.2147/CLEP.S91125 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Silberg JL, & Bulik CM (2005). The developmental association between eating disorders symptoms and symptoms of depression and anxiety in juvenile twin girls. J Child Psychol Psychiatry, 46(12), 1317–1326. doi: 10.1111/j.1469-7610.2005.01427.x [DOI] [PubMed] [Google Scholar]
  30. Steinhausen HC, Villumsen MD, Horder K, Winkler LA, Bilenberg N, & Stoving RK (2021). Comorbid mental disorders during long-term course in a nationwide cohort of patients with anorexia nervosa. Int J Eat Disord. doi: 10.1002/eat.23570 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Swanson SA, Crow SJ, Le Grange D, Swendsen J, & Merikangas KR (2011). Prevalence and correlates of eating disorders in adolescents. Results from the national comorbidity survey replication adolescent supplement. Arch Gen Psychiatry, 68(7), 714–723. doi: 10.1001/archgenpsychiatry.2011.22 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Trace SE, Thornton LM, Baker JH, Root TL, Janson LE, Lichtenstein P, … Bulik CM (2013). A behavioral-genetic investigation of bulimia nervosa and its relationship with alcohol use disorder. Psychiatry Res, 208(3), 232–237. doi: 10.1016/j.psychres.2013.04.030 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Wade TD, Bulik CM, Neale M, & Kendler KS (2000). Anorexia nervosa and major depression: shared genetic and environmental risk factors. Am J Psychiatry, 157(3), 469–471. doi: 10.1176/appi.ajp.157.3.469 [DOI] [PubMed] [Google Scholar]
  34. World Health Organization. (1967). International classification of diseases: Manual of the international statistical classification of diseases, injuries and causes of death (ICD-8). Retrieved from Geneva: [Google Scholar]
  35. World Health Organization. (1993). The ICD-10 Classification of Mental and Behavioural Disorders: Diagnostic Criteria for Research. Retrieved from Geneva: [Google Scholar]
  36. Xu R, & O’Quigley J (2000). Estimating average regression effect under non-proportional hazards. Biostatistics, 1(4), 423–439. doi: 10.1093/biostatistics/1.4.423 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Appendix Text Tables and Figures
NIHMS1854362-supplement-Supplementary_Appendix_Text_Tables_and_Figures.docx (1.9MB, docx)

RESOURCES