Introduction

Eating disorders (EDs) are among the most common causes of ill health in young people, affecting 3–4% of women and 0.3% of men [1]. They are associated with elevated rates of morbidity and mortality. Individuals with eating disorders have an elevated risk of dying by suicide compared to age-matched population estimates [2].

Binge eating (BE) behavior is characterized by uncontrollable consumption of a huge quantity of food in a certain time. It is a main feature of anorexia nervosa binge/purge subtype, bulimia nervosa, and binge eating disorder (BED) [3]. Patients may manifest BE behavior without fulfilling diagnostic criteria of any eating disorder recognized by Diagnostic and Statistical Manual of Mental Disorders (DSM), but it is still considered a serious mental health issue with major detrimental effects [4].

The most frequent eating disorder exhibiting BE behavior is BED [5]. A 2007 study found that 3.5% of women and 2.0% of men had binge eating disorder during their life. This makes BED three times more prevalent than anorexia and bulimia combined [6]. Binge eating was only listed in appendix B in the DSM-IV and had to be diagnosed with the non-specific Eating Disorder Not Otherwise Specified until recognized as a distinct eating disorder in the Fifth Edition of the DSM-5, published in 2013 [7].

Borderline personality disorder (BPD) is characterized by instability in the core symptom domains of impulsivity and negative affect as well as associated dysfunctional interpersonal relationships [8]. According to DSM-5, BPD is diagnosed by the presences of at least five of the following; unstable interpersonal relationships, distorted self-image, emotional instability, persistent feeling of emptiness, fear of abandonment, and at least two self-destructive impulsive behaviors (e.g., irrational spending, sex, drug addiction, reckless driving, BE), difficulty controlling anger, recurrent suicidal or self-mutilating behaviors, and stress-related paranoid ideation or dissociation [8].

BPD shows high comorbidity with depression (80%), alcohol and other substance use disorder, posttraumatic stress disorder, as well as eating disorders (56%), panic disorder (48%), and bipolar disorder (20%) [9].

The comorbidity of eating disorder with BPD is common as they share many of the same risk factors, phenotypes, and behavioral traits including: attachment disturbances, adverse childhood experiences, non-suicidal self-injury, interpersonal difficulties, and comorbid maladaptive mood and impulse regulation behaviors [10]. The common diagnostic features in both syndromes include distorted self-image, negative affect, impulsivity, emotional dysregulation, and interpersonal instability [11]. The relationship between BPD and disordered eating was explored by Al-Salom and Boylan study [12], they mentioned that among the symptoms of BPD, fear of being abandoned or rejected by others relates to the social psychology concept of rejection sensitivity. It is possible that disordered eating may be used by adolescents with BPD in their study to regulate their emotions in situations where rejection sensitivity is exacerbated [12]. Therefore, patients with BPD may be more sensitive to rejection, and these fears of rejection may result in increased emotion dysregulation and subsequent dysregulated eating behaviors [13].

Another study acknowledged a bi-directional relationship between emotional regulation difficulties and identity integration, and that both may reinforce each other leading to eating disorders. They stated that difficulties with affect regulation may be secondary to unclear and unstable representations of self in individuals, and that individuals who have difficulty making sense of their internal world may resort to behaviors like binging and purging to decrease the intensity of negative emotional states [14].

Hence, we can consider emotion dysregulation as a final product and the progression of the various BPD symptoms.

Also, several studies have shown associations between disordered eating and impulsivity [15] with impulsivity being lately considered as a transdiagnostic characteristic of individuals with eating disorders [16]. Meanwhile BE is considered one of impulsive behaviors in BPD patients [17]. Therefore, it is important to investigate the role of emotion dysregulation and impulsivity traits in the development of disordered eating behavior.

The prevalence of BPD is high in bulimia nervosa (21%), moderate in BED (9%), and low in anorexia nervosa (3%) [18]. People with BPD and co-occurring eating disorders have higher rates of suicidal behaviors, self-harming behaviors, anxiety disorders, and mood disorders than patients diagnosed with BPD without co-occurring eating disorders [19]. Therefore, it is a sign of severity and a predictor of poor prognosis and should raise concern for the high risk it holds.

Unfortunately, comorbid eating disorders are often undiagnosed and unrecognized by primary, secondary, or specialist care despite those patients having high rates of attendance at health care facilities [11]. Therefore, it is highly recommended to detect and diagnose comorbid BPD and disordered eating behaviors early, for saving their lives, improve the quality of life and the prognosis of those patients [11].

BPD rates are found to be higher in bulimia nervosa and the anorexia binge‐purge subtype than patients with anorexia nervosa‐restrictive, suggesting impulsivity and emotion dysregulation as common characteristics between BPD and bulimic symptoms [20].

In addition, there has been a recent rise in the prevalence of EDs documented in Arab and Eastern countries despite that EDs are considered more related to the Western cultural [21]. But unfortunately, there are a limited number of studies addressing the association between BPD and BE in Arab and Middle Eastern countries.

Accordingly, we find it necessary to conduct this study to describe the frequency of BE behavior in a population of Egyptian BPD patients, and to investigate the relationship between impulsivity and emotional regulation with BE behavior as they are core problems in both disorders. We aim to highlight the comorbidity of BPD and eating disorders, and so help in their screening and enhancing treatment protocols provided to BPD.

Method

Participants

We collected a convenient sample of 70 participants calculated by the Power Analysis and Sample Size (PASS) program by setting alpha error at 5% and confidence interval width at 0.1. [22]. Patients were recruited from the outpatient clinic of Okasha Institute of psychiatry, Ain Shams University. Cairo, Egypt. The institute is located in western Cairo, and serves both urban and rural areas, including Greater Cairo and other governorates as well.

Participants enrolled met the following eligibility criteria: (a) fulfilling the diagnosis of BPD as outlined in the DSM-IV criteria using the Structured Clinical Interview for DSM-IV Axis II disorders (SCID-II), (b) 18–45 years, and (c) able to give informed consent for their participation in the study. Both genders were included. Participants were excluded if they had any comorbid diseases that affect eating behavior, diet or body mass index, e.g., diabetes mellitus (DM), thyroid disease, other endocrinal disorders, or other psychiatric disorder (schizophrenia, bipolar, or major depression).

Procedure

The cross-sectional observational study was conducted over a period of 6 months from November 2019 till April 2020. During this period, participants visiting the outpatient clinics fulfilling the study inclusion criteria were enrolled until total number of participants reached 70. Each participant was interviewed and assessed by the researchers, ensuring that all questionnaires were completed with no missing data.

Participants were interviewed using the Structured Clinical Interview for DSM-IV Personality Disorders (SCID-II) to diagnose BPD and exclude other personality disorders.

They were then assessed for impulsivity, and emotional regulation using the Barratt Impulsiveness Scale—version 11 (BIS) and the Trait Meta Mood Scale (TMMS), respectively. They were screened for BE behaviors using the Binge Eating Scale (BES) and the diagnosis of BED was further confirmed by DSM-5 criteria. We focused on BED as it is the most frequent eating disorder exhibiting BE behavior. In addition, the association between anorexia nervosa as well as bulimia nervosa with Borderline Personality Disorder is well studied, but fewer studies were done to explore the association between Binge Eating Disorder and Borderline Personality Disorder.

Measures

SCID-II was used as a diagnostic tool for personality disorders according to the DSM-IV criteria. The SCID-II is a 113-item structured clinical interview for Axis II disorders. The items are organized on a disorder by disorder basis. Individual items or criteria are scored as: 0, absent; 1, subthreshold; or 2, threshold [23]. The Arabic version was used [24]. SCID-II was used, since an Arabic validated version is available. The criteria of BPD were similar to that of DSM 5.

Diagnostic and Statistical Manual of Mental Disorders-Fifth Edition (DSM-5) criteria were used for diagnosis of binge eating disorder as it was only introduced by DSM 5 [8].

Binge Eating Scale (BES) assesses the presence or absence of recurrent BE behavior. The total score ranges from 0 to 46 and is positively correlated with the severity of binging behavior. A significant level of BE behavior is defined by a BES score ≥ 17 where none-binging (< 17), moderate (18–26), and severe (> 27) [25]. The Arabic version was used [26].

Trait Meta Mood Scale (TMMS) was designed to assess how people reflect upon their moods, and determine the extent to which people attend to and value their feelings (Attention), feel clear rather than confused about their feelings (Clarity), and use positive thinking to repair negative moods (Repair). It consists of 30 items on a 5-point Likert-type scale with choices ranging from 1 = strongly disagree to 5 = strongly agree. The TMMS can be subdivided into three subscales which are: attention to feeling [range from 12 (low) to 60 (high)], clarity of feelings (range from 12 for low to 60 for high), and repair mood [range from 6 (low) to 40 (high)] and the total score ranges from 30 (low) to 150 (high), where high scores indicate better ability to deal with emotions [27]. The Arabic version was used as well [28].

Barratt Impulsiveness Scale—version 11(BIS) measures facets of impulsivity. The scale consists of three separate subscales: non-planning (BISnp), motor impulsivity (BISm), and attentional impulsivity (BISa). The scale contains 30 items which are rated on a 4-point Likert-type scale with choices ranging from (rarely/never) to (almost always) mild (< 70), moderate (70–80), and severe (> 80) [29]. The Arabic version was used [30].

Statistical analyses

The collected data were revised, coded, tabulated, and entered into data sheets using Statistical package for Social Science (SPSS 20). Data were presented and the appropriate analysis was done according to the type of data obtained for each parameter. Epidemiological data were represented as mean and standard deviation (± SD) for numerical data, while frequency and percentage were used to represent categorical data. Student T test was used to assess the statistical significance of the difference between the binge eating behavior group and the non-binging group means (classified based on their binge eating scale cut-off score of > 17). Chi-square test was used to examine the relationship between SCID-II criteria of BDP and binging. Fisher’s exact test was used to examine the relationship between SCID-II criteria of BDP and binging when the expected count is less than 5 in more than 20% of cells. Correlation analysis (using Pearson’s method) was used to assess the strength of association between TMMS, BIS, and BES.

Results

Participants’ socio-demographic and clinical characteristics

Table 1 showed that the mean age of the participants’ was (25.81 ± 6.34), the majority (68.6%) were women, with (68.6%) attained higher education, and 37.1% were working.

Table 1 Sociodemographic data of the participants
Full size table

According to SCID-II, most of the cases had impulsivity and affective instability (98.6%), followed by fear of abandonment, unstable interpersonal relationships and identity disturbance (92.9%), then chronic feelings of emptiness (84.3%) and difficulty controlling anger (80%). Recurrent suicide and paranoid ideations were slightly less frequent (77.1%).

BIS showed that all participants had impulsivity with various degree, with a 60% of cases classified as moderate impulsivity. For emotional regulation, their means on TMMS classified them as moderate degree of emotional dysregulation in all aspects, as shown in Table 2.

Table 2 Barratt impulsiveness scale and Trait Meta Mood Scale among participants
Full size table

Finally, Table 3 summarizes the frequency of BE behavior among our participants using BES; 37 (52.9%) cases suffered from this problem as scored by a threshold > 17.

Table 3 Binge Eating Scale (BES)
Full size table

Confirmation of BED diagnosis by DSM-5 criteria

DSM-5 criteria for BED, used to confirm diagnosis in patients with BE behavior, revealed that 26 (70%) patients out of 37 patients with BE behavior, and 37.14% out of the total 70 participants fulfilled DSM-5 criteria of BED.

Correlation between TMMS, BIS, and BES

Table 4 illustrates that binge eating was inversely correlated with age, clarity of feeling, and total score of TMMS. Thus, there was an inverse relationship between emotional regulation and binging. On the other hand, no significant correlation between binging and impulsivity.

Table 4 Correlation between TMMS, BIS & BES
Full size table

Comparison between the binging and no binging groups

Participants were classified based on their BES score into two groups, a group with binging behavior and a group without binging behavior. We then compared between both groups as regard socio-demographic data, SCID-II criteria of BPD, BIS, and TMMS. Each variable was treated as an independent variable, and that is why, we did not consider changing the cut-off point with p value < 0.05.

There was no significant relation between the two groups regarding the socio-demographic data of the patients (p value > 0.05).

We found that only fear of abandonment among other criteria of SCID-II had a significant relation with binging (p value 0.02) (Table 5).

Table 5 Relation between borderline personality disorder criteria in SCID-II & binging
Full size table

According to data shown in Table 6, participants in the binging group had lower total score on TMMS and on the clarity of feeling subscale with significant statistical difference p value of 0.021 and 0.004, respectively. Finally, there were no statistically significant differences between both groups regarding impulsivity (p value 0.143) either on the level of the degree or on the subscales of BIS.

Table 6 Relation between binging, components of BIS subscale and TMMS
Full size table

Discussion

As the occurrence of BE behavior in patients with BPD is not uncommon, and research is lacking regarding the comorbidity of BE behavior with BPD especially in Arab & Middle East countries, we examined the frequency of BE behavior in a sample of Egyptian patients with BPD, to determine the relation between BE behavior, impulsivity, and emotional regulation in those patients.

Seventy participants diagnosed with BPD according to DSM-IV criteria were recruited, and their mean age was 25.81 ± 6.34. All BPD features reported by SCID-II were fulfilled by at least 77.1% of our participants where the vast majority of cases had impulsivity and affective instability (98.6%), which are the predominant underlying criteria driving other BPD symptoms explained by the results of [31]. The majority participants were women (68.6%). This finding appears to be in line with previous studies which stated that women with BPD are more likely to seek treatment than men [32].

According to BIS assessment for impulsivity, participants showed highest scores on the motor scale (mean 29.47 ± 5.05) compared to other scales, which means that most of them tend to act without thinking, this is similar to the findings of [33]. On the other side, various studies [34,35,36] showed opposite findings. This discrepancy necessitates further studies.

Scores of TMMS showed a moderate degree of impairment in emotional regulation among the participants with decrease in their abilities to pay attention to their inner emotional states, to understand and discriminate between them, and to repair negative mood emotional experiences to positive state. This was in line with the following studies [37,38,39].

BES was used to determine the frequency of BE behavior in the studied participants. 37 (52.9%) participants were found to suffer from BE behavior. Similar percentage was found by a study done in a Palestinian group reporting 50% of the sample had BE behavior [40]. However, lower percentage revealed in United Arab Emirates by [41] reporting 24% to 36% and in Jordan, results showed 16.9% [42].

Higher percentage was stated in a study that took place in Saudi Arabia by Rabie et al. [43] which found that the prevalence of BE behavior was 68.8%. These variations could be explained by the use of different self-reported screening tools among different group of subjects which is subjective and liable for bias.

The high frequency of BE in our study and in other Arab countries might be influenced by the transition of cultural background with the help of the digital media making popular culture and the arts more widely accessible. This might have led a wide spread adoption of Western styles and standards of beauty among the residents of Arab countries [21].

The DSM-5 criteria were used for diagnosis of BED after using the BES, and we found that 26 (70%) participants out of the 37 with BE behavior fulfilled DSM-5 criteria of BED which represented (37.14%) of the whole sample. This was consistent with the study done by Marino and Zanarini [44] who reported that 37% of BPD had BED. However, lower percentage reported in a group of studies, as that of Chen et al. [19] who reported that only 5.2% of cases had BED. Shenoy & Praharaj [45] reported 22.4% and another study found 12.5% had BED [46]. The differences could be explained by the APA, declaring that binge eating is a prevalent disorder in Middle East countries [8]. And this can be attributed to the impact of the current social and economic evolution on the culture of Arab and Eastern countries [47]. The inflation of the Egyptian pound led to increased economic hardship placed on the majority of Egypt’s population and driving up the cost of living together with the uncertainty about future goals’ attainment. These factors caused frustration and negative emotions to most of the youth population who are exposed in the same time to challenging high standard demands, increasing the likelihood of suffering from a wide range of psychological problems including eating disorders.

We found BES to be a useful screening tool, especially in recognizing non-binging individuals [48], and many investigators approved the validity of BES in general and clinical settings [49]. Grupski et al. [50] found that BES showed acceptable specificity and high sensitivity in BED diagnosis in patients seeking bariatric surgery. Another study demonstrated that the BES had a Cronbach's alpha of 0.89 for the test re-test reliability, and when the cut-off point of 17 was compared with SCID-defined BED, the BES demonstrated a sensitivity of 0.98, and a specificity of 0.48 [51].

In the current study, we find that BES scale showed moderate sensitivity as it detected 70% of individuals having BED. Therefore, we recommend using it as a screening tool among mental health patients.

Binge eating behavior was inversely correlated with age, clarity of feeling, and total score of TMMS. Younger participants with greater inability to understand their emotional states tended to exhibit more binge eating behavior. This might be related to that younger age exhibit greater difficulties in emotion regulation and use binging as a way of coping as, the frontal lobe, responsible for judgment and inhibition is underdeveloped relative to the limbic system, which is responsible for emotional processing [52]. Therefore, with the more maturation and development of frontal lobe with age, the ability to control the limbic system and emotions increases.

When we compared binging group (52.9%) with non-binging group (47.1%) in relation to SCID-II criteria, we found a significant correlation between binging and fear of abandonment. This is in agreement with the previous study of Patton [53] who confirmed the using of BE to adjust or diminish fear of abandonment.

Regarding TMMS, the binging group had lower total and clarity of feeling scores—signifying their inability to regulate and understand their own emotion. This is corroborated with Khosravi’s [46] postulation that the progression and establishment of BE depends on poor emotional regulation.

This finding was in line with a previous study of Svaldi et al. [54], in which BED cases showed less emotional awareness and clarity, and that of [55] who suggested that half of binge eating episodes in patients with BE behavior are motivated by trying to cope with lack of emotion regulation. Meanwhile, their dysfunctional emotion regulation strategies directed them toward binging [56] to deal with the elevated negative emotion as it serves as a way of distracting attention away from the upsetting thoughts to focusing on the concrete physical stimuli associated with the binge eating [57].

Also, De Paoli et al.’s [58] study stated a transdiagnostic risk model for the development of ED and/or BPD symptoms, where the shared underlying factors of emotion dysregulation and abandonment may predispose individuals to develop symptoms for either disorder.

Accordingly, in the management of patients with BPD and eating disorders, good attention must be given to regulate their emotional dysfunctions as a core aspect in the management plan as it will help in the control of both disorders.

Although, Blasco-Fontecilla et al. [59] reported that BPD patients depend on impulsive behaviors such as BE to overcome feeling of emptiness, we did not find a significant association between binging and impulsivity. However, [60,61,62,63,64] found association between BE and different facets of impulsivity.

Therefore, our findings could be explained by Pearson et al.’s [65] hypothesis that over time, BE changes from an impulsive act (e.g., binging without knowing or thinking about its consequences) to a compulsive nature (e.g., binging regardless the consequences). Hence, the impulsivity is not necessarily the underlying cause for binging.

Strength and limitations

Research work focused on the comorbidity of binge eating with borderline disorder is scarce especially in Egypt. This study revealed that impulsivity is not the main association of binge eating in borderline patients. Multiple valid and reliable tools were used in the assessment.

However, using being a hospital-based sample rather than a community-based sample may affect the generalization of the results. The sample was recruited from treatment seeking patients. This might be a confounding factor which can affect the interpretation of results. The self-report BE is subjective and liable for bias. Also, excluding comorbid psychiatric and other personality disorders which are highly prevalent among BPD could affect the generalization of the study. Meanwhile, we only focused on diagnosing BED rather than other eating disorders showing binging behavior.

What is already known on this subject?

Given the extensive literature suggesting the relevance of binge eating to borderline personality psychopathology and emotional dysregulation, also proved by the present study, there are clear implications for screening and treating eating disorders owing to their potential for causing poor treatment outcome among BPD.

What this study adds?

The current study revealed that there are a substantial number of patients with BPD suffering from BED. We also found a significant association between binging and fear of abandonment.

Patients exhibiting binge eating are less capable of regulating and understanding their own emotions in comparison to the non-binging group of patients. On the other side, no significant correlation between binging and impulsivity was found.