Introduction

Binge eating disorder (BED) is a newly recognized eating disorder characterized by recurrent, distressing episodes of binge eating without the inappropriate compensatory behaviors of bulimia nervosa (BN) (e.g., self-induced vomiting or laxative abuse) [1]. BED is more common than other eating disorders, with an estimated worldwide lifetime prevalence of 1.9 % [2, 3]. It is associated with psychiatric comorbidity, psychological distress, reduced quality of life, and functional impairment.

In addition to substantial psychiatric and psychological morbidity, growing research indicates that BED, as well as binge eating behavior in general, is also associated with obesity [2, 3]. Moreover, a growing number of prospective studies show that BED or binge eating behavior is associated with the development of obesity [4–6]. Finally, BED appears to be associated with greater health dissatisfaction [7], increased health service utilization and costs [8–11], and enhanced mortality [12].

Less is known, however, about the specific medical conditions associated with BED beyond obesity, and the relationship of these medical conditions to co-occurring obesity and psychiatric disorders. Moreover, while there are a number of authoritative reviews on the medical conditions associated with BN and AN [13–15], there have only been a few reviews of the medical comorbidity of BED [16–18].

In light of the increasing awareness of BED, we reviewed studies of the medical comorbidity of BED, focusing on medical conditions beyond obesity. We also attempted to assess the contribution of obesity and psychiatric comorbidity to the general medical morbidity of BED.

Methods

We searched PubMed and the Cochrane Library for all English-, Spanish-, and French-language articles published from database inception through March 2016 using the key terms binge eating and medical comorbidity, which revealed 216 articles. We evaluated only studies of BED or potential variants of BED, including non-purging BN and/or clinically significant binge eating behavior. We did not include studies of anorexia nervosa (AN), BN, or EDNOS unless rates of BED or clinically significant binge eating behavior were specified. We also evaluated only studies that assessed general medical conditions beyond obesity. From the 216 articles, 17 were appropriate for review. Based on these articles, we then conducted searches of the term binge eating disorder with the specific medical terms diabetes, cardiovascular, heart disease, metabolic syndrome, gastrointestinal, arthritis, asthma, and narcolepsy, which yielded 54 more suitable reports. The search was supplemented with a manual review of the reference lists of relevant articles.

Results

We found three broad categories of studies: cross-sectional studies of medical conditions in BED; prospective studies of medical conditions in BED; and studies of BED in specific medical conditions. These studies are summarized in Tables 1, 2, 3 and described below.

Table 1 Cross-sectional studies of medical conditions in individuals with BED
Full size table
Table 2 Prospective studies of medical conditions in BED
Full size table
Table 3 Studies of BED in specific medical conditions
Full size table

Cross-sectional studies of medical conditions in BED

In epidemiological samples, BED was associated with diabetes, hypertension, arthritis, chronic back/neck pain, other pain disorders (including chronic headaches and fibromyalgia), gastrointestinal conditions (gastrointestinal symptoms, ulcer, and irritable bowel syndrome), sleep problems, and asthma [3, 19–25]. Among women, BED was also associated with early menarche, menstrual dysfunction, delivery of higher birth weight babies, and long duration of the first and second stages of labor [23, 26–28]. Associations of BED with fibromyalgia, pain, irritable bowel syndrome, gastrointestinal symptoms, sleep problems, and early menarche persisted after controlling for obesity or body mass index (BMI) [20, 22, 25, 28]. Associations of BED with diabetes, hypertension, chronic back/neck pain, chronic headaches, other chronic pain, and asthma, but not with arthritis or ulcer persisted after controlling for psychiatric comorbidity [3, 19, 21, 24].

Several studies explored the temporal relationship between onset of BED and onset of co-occurring medical conditions. For example, in the World Health Organization (WHO), World Mental Health Surveys, over 24,000 respondents from 14 countries were evaluated with the WHO Composite International Diagnostic Interview and a checklist assessing 15 lifetime chronic physical conditions [3]. Survival analysis found temporally primary BED predicted subsequent arthritis, chronic back/neck pain, chronic headaches, other chronic pain, diabetes, hypertension, and ulcer. In addition, reciprocal survival models found generally insignificant associations between temporally primary physical conditions and subsequent onset of BED. In subsequent analyses of data from the WHO World Mental health Surveys, temporally primary BED was associated with the development of diabetes [21], hypertension [24], and adult-onset asthma [19]. These findings remained significant when analyses were adjusted for lifetime mental disorder comorbidity.

Two epidemiology studies found BED was not associated with subsequent heart attack and stroke [3] or with heart disease [23]. In another epidemiologic sample, binge eating behavior was not associated with heart disease [7]. In addition, BED was not associated with cancer or renal disease [23].

At least four studies evaluated diabetes in groups of patients with BED [23, 29–31] with rates ranging from 2 % [29] to 26 % [31]. Studies with non-BED comparison groups found higher [23, 30] or comparable [31] rates of diabetes in the BED group.

A number of studies assessed metabolic syndrome (or metabolic syndrome components) in groups of patients with BED, with rates of metabolic syndrome ranging from 24 % [32] to 60 % [33]. Some rates may be falsely low, as several studies excluded patients with medical conditions, including hypertension or diabetes [32, 34–36]. Three studies found metabolic syndrome to be more common in men with BED than women with BED [33, 37, 38], but this was not found in two other studies [29, 36]. Succurro et al. [39] compared obese patients with and without BED and found that BED patients had higher hemoglobin A1C and fasting insulin levels, lower HDL levels, and greater inflammatory markers. However, BED patients had a higher BMI than non-BED patients. In a study of 2225 bariatric surgery candidates, after adjusting for age, sex, education, and BMI, BED was associated with impaired glucose and triglyceride levels [40]. Similarly, Rosenbaum et al. [41] found women veteran primary care patients with BED had higher rates of lipid disorders and hypertension than those without BED. In contrast, in a study of 128 children seeking obesity treatment, those with clinically significant binge eating had similar glucose and lipid levels and systolic and diastolic blood pressures to those without binge eating [42]. In addition, Barber et al. [32] found comparable rates of MetS among overweight or obese patients with and without DSM-5 BED. Of note, the factor structure of metabolic syndrome in obese individuals with BED was similar to those in normative population studies [37].

Two studies assessed gastrointestinal disorders in patients with BED. In a comparison of 100 bariatric surgery candidates with 100 healthy controls evaluated for functional gastrointestinal disorders (FGIDs) according to Rome III criteria and for binge eating behavior, the prevalence of FGIDs was the same in the two groups [43]. However, patients who also had binge eating behavior had elevated rates of postprandial distress syndrome. In a study of liver function tests in a group of patients with eating disorders, levels of aspartate aminotransferase and alanine aminotransferase were directly associated with BMI in those with BED [44]. Of note, gastric dilation with or without rupture has been described in patients with BN and those with AN [45, 46]. However, we were unable to find a description of gastric dilation in an individual with clearly documented BED.

Prospective studies of medical conditions of BED

We found three prospective studies of medical conditions beyond obesity in individuals with BED (summarized in Table 2). In one study, overweight or obese adults with BED were more likely than those without BED to receive new diagnoses of dyslipidemia, hypertension, or diabetes over a 5-year period [47]. Individuals with BED were statistically significantly more likely to receive a new diagnosis of dyslipidemia, at least one of these diagnoses, as well as two or more of these diagnoses. As analyses were adjusted for baseline BMI and interval BMI change, the authors concluded that BED may confer a risk of the development of components of the metabolic syndrome over and above the risk attributable to obesity alone.

In a study of 180 children aged 5–12 years at risk for adult obesity conducted from 1996 to 2010, the presence of self-reported binge eating was associated with a 5.33 greater odds of metabolic syndrome at follow up [48]. In a recent Finnish study of 2342 eating disorder patients followed for 16 years and 9368 matched community controls [49], a substantially increased risk of type 2 diabetes was reported among individuals with BED at treatment onset. By the end of the study, every third BED patient had type 2 diabetes, and the risk was even higher in males. In contrast, 4.4 % of those with BN had type 2 diabetes.

Studies of BED in medical samples

We found 28 studies evaluating BED or binge eating behavior in groups of patients with a particular medical disorder (see Table 3). We review these studies below according to medical disorder.

Diabetes

We found 12 studies that evaluated BED or binge eating behavior in patients with type 2 diabetes [50–61], with rates of BED by DSM-IV criteria ranging from 1.4 % [50] to 26 % [52]. In the two studies that employed healthy control groups, there were no differences in BED prevalence rates between diabetes patients and controls [56, 58], but one study found a higher rate of binge eating behavior among diabetic patients than controls [56]. Five studies evaluated BED or binge eating behavior in patients with type 1 diabetes [62–66] with rates of binge eating behavior ranging from 3 % [62] to 17 % [63]. In the only study that had a control group, 12.5 % of females with type 1 diabetes had DSM-IV BED compared with 5.3 % of control females [66]. In a study that compared type 1 diabetic patients with and without BED, BED patients had a higher BMI and greater HgA1C levels [67].

Sleep problems/disorders

Questionnaire data from 72,435 women found that BED symptoms were associated with sleep problems during the first 18 weeks of pregnancy and greater dissatisfaction with sleep 18 months after childbirth [68].

High rates of binge eating have also been found in individuals with narcolepsy [57, 69]. One study found higher rates of binge eating in people with narcolepsy and cataplexy than narcolepsy alone [70]. However, in a study of 116 patients with narcolepsy and 80 controls, eating disorders, including BED, were not more common in patients with narcolepsy [71].

Gastrointestinal disorders

One study assessed eating psychopathology in 100 adults with newly diagnosed celiac disease and 100 controls and found the two groups had similar rates of clinically significant binge eating behavior (defined as a Binge Eating Scale score ≥17) [72]. However, these results are limited by the fact that people with eating disorder diagnoses were specifically excluded from participation.

Heart/cardiovascular disease

Bankier et al. [73] found that 10 % of a sample of patients with stable coronary heart disease had current BED. Interestingly, two studies have found reductions in high-frequency heart rate variability in response to mental stress (a marker of reduced cardiac parasympathetic activity) in obese people with BED as compared to similarly obese people without BED [74, 75]. Another study found non-purging BN to be associated with longer QT intervals and increased QT dispersion [76]. Decreases in high-frequency heart rate variability are associated with an increased risk of coronary heart disease [77], while prolonged QT interval dispersion is predictors of cardiovascular mortality [78]. Finally, in an echocardiography study of 40 eating disorder patients, BED patients had lower left ventricular end-diastolic stroke volume (LvEDV) and cardiac output compared with normal controls [79]. After normalization for body surface area, BED patients had significantly lower LvEDV and left ventricular end-systolic volume than controls, AN patients, and BN patients.

Pain

In a study of 131 patients with fibromyalgia, obese patients had more binge days than non-obese patients [80].

Neurological disease

Raggi et al. [81] found that 15 % of 33 patients with idiopathic intracranial hypertension had BED and obesity [81]. In a large epidemiological sample, Kolstad et al. [82] found that pregnant women with epilepsy had a significantly higher rate of BED compared with pregnant women without epilepsy (6.5 vs. 4.7 %, respectively). In contrast, in a study of 400 patients with Parkinson’s disease, only 3 % had binge eating behavior [83].

Endocrine disease (other than diabetes)

Hollinrake et al. [84] evaluated 103 women with polycystic ovary syndrome, and found 13 % had BED compared with 2 % of controls.

Treatment of medical comorbidity in BED

We found no randomized-controlled treatment trials of patients with BED with medical comorbidity beyond obesity. Of note, in a double-blind, randomized trial in BED, chromium picolinate at two different doses (600 mcg and 900 mcg/day) significantly reduced fasting glucose without significantly reducing binge eating frequency or body weight [85].

Discussion

Growing research suggests BED is associated with a range of medical conditions in addition to obesity (see Table 4). Many of these conditions are related to obesity, including type 2 diabetes, hypertension, and dyslipidemia [86]. Several studies suggest that BED or BE behavior may increase the risk of these conditions, and of metabolic syndrome, above that of obesity or psychiatric comorbidity, but this needs further study. BED or binge eating behavior may also be associated with asthma, gastrointestinal symptoms and disorders (especially functional disorders), sleep disorders and problems, pain disorders, and among women, menstrual dysfunction, pregnancy complications, intracranial hypertension, and polycystic ovary syndrome.

Table 4 Main medical comorbidities of BED (beyond obesity)
Full size table

Though preliminary, these findings indicate patients with BED should receive comprehensive medical evaluations with particular focus on diabetes, hypertension, dyslipidemias, pain, sleep disorders, functional gastrointestinal disorders, and asthma. Women should additionally receive evaluation of reproductive function and for polycystic ovary syndrome. Conversely, patients with diabetes, hypertension, dyslipidemias, pain, and sleep disturbances should be evaluated for binge eating behavior despite their BMI, since BED could be an independent risk factor for some of these medical conditions.

The major limitation of this review is the paucity of epidemiologic and prospective studies of medical conditions in persons with BED. Also, the majority of studies of BED used DSM-IV criteria [87], and it is unknown if their findings would generalize to BED as defined by DSM-5 criteria. Another limitation is that most studies did not evaluate the relationship of medical comorbidity in BED with obesity, co-occurring psychiatric disorders, or degree of psychological distress. Thus, though BED is associated with substantial medical comorbidity, it remains unknown if this is due to BED itself, obesity, and/or the associated psychiatric comorbidity of BED. In addition, due to the lack of prospective studies, the chronological relationship of BED with its comorbid medical disorders is largely unknown. Yet, another limitation is the paucity of studies exploring the treatment of medical disorders in persons with BED. It is, therefore, unknown if diabetes, hypertension, dyslipidemias, pain, or sleep disturbances need to be differentially managed in individuals with BED. It is also unknown to what degree successful treatment of BED symptomatology is associated with resolution of medical comorbidity.

Further studies of medical comorbidity in BED are greatly needed. These include prospective studies of development of medical conditions in persons with BED, and the relationship of these conditions to eating disorder psychopathology, other forms of co-occurring psychopathology, and to baseline weight, weight change, and other measures of adiposity. A major priority for future research is to conduct longitudinal studies to determine if the development of BED predicts subsequent obesity-related conditions independent of the development of obesity. It is unknown whether BED affects treatment response of comorbid medical conditions, including diabetes, hypertension, or dyslipidemia. Conversely, the effects of BED in response to treatments for dyslipidemia, hypertension, diabetes, and pain are largely unknown. Prospective, randomized-controlled trials of treatment of medical comorbidities in persons with BED are, therefore, also needed.