Obsessions are strongly related to eating disorder symptoms in anorexia nervosa and atypical anorexia nervosa

Highlights

• Anorexia Nervosa (AN) and Obsessive-Compulsive Disorder (OCD) are highly comorbid.

• We tested which aspects of OCD are most relevant in AN and Atypical AN.

• Obsessing was most related to AN symptoms.

• There were no significant differences between AN and Atypical AN in OCD symptoms.

• Obsessions may need particular attention in treatments of comorbid AN-OCD.

Abstract

Anorexia nervosa (AN) and obsessive-compulsive disorder (OCD) are highly comorbid. However, little research has examined which specific cognitive-behavioral aspects (e.g., checking, obsessing) of OCD are most relevant in those with AN. Furthermore, there is no research examining aspects of OCD in Atypical AN. The current two studies (N = 139 and N = 115 individuals diagnosed with AN/Atypical AN) examined a) which aspects of OCD were most related to AN symptomatology and b) if there were differences in OCD between individuals diagnosed with AN vs Atypical AN. We found that obsessing was most related to AN symptoms. We also found that there were no substantial significant differences between AN and Atypical AN. These findings add to the literature suggesting minimal differences between AN and Atypical AN, specifically regarding OCD symptomatology. These findings clarify that obsessions (rather than compulsions) may be the specific aspect of OCD most warranting treatment intervention in AN and Atypical AN.

Introduction

Anorexia nervosa (AN) is a serious mental illness, carrying a high risk of mortality and causing extreme suffering and impairment (Klump, Bulik, Kaye, Treasure, & Tyson, 2009). Part of this impairment is due to the high rates of comorbidity, with rates estimated up to 85% (Pallister & Waller, 2008). One frequently co-occurring disorder is Obsessive-Compulsive Disorder (OCD). Estimates suggest that between 35 and 44% of individuals with AN also meet criteria for OCD (Halmi et al., 2005; LaSalle et al., 2004; Pinto, Mancebo, Eisen, Pagano, & Rasmussen, 2006; Rubenstein et al., 1993; Swinbourne & Touyz, 2007).

In addition to the frequent co-occurrence, AN-OCD has a strong, positive genetic correlation, (~55%; Anttila et al., 2018; Cederlöf et al., 2015; Mas et al., 2013), suggesting a shared etiology between AN and OCD. Further, other work has shown that eating disorder (ED) cognitions and behaviors serve a function similar to obsessions and compulsions (e.g., as a temporary relief for anxiety, while predicting longer term increases in anxiety and ED symptoms or temporary safety behaviors) (Levinson et al., 2018). Overall, there is strong evidence for both genotypic and phenotypic similarity between OCD and EDs. However, there is a lack of research exploring the relationship between specific ED symptoms (e.g., drive for thinness, bulimia symptoms) and specific dimensions of OCD (e.g., obsessions, checking behaviors).

OCD is a heterogeneous disorder consisting of many different cognitive-behavioral aspects, including; washing, checking, ordering, hoarding, obsessing, and neutralizing (Foa et al., 2002; Foa, Kozak, Salkovskis, Coles, & Amir, 1998). Obsessing is a cognitive aspect, whereas, washing, checking, ordering, hoarding, and neutralizing focus primarily on OCD behaviors in response to obsessions. Washing is focused on fears of contamination (e.g., repeatedly washing one's hands). Checking focuses on compulsions to check certain objects or places in order to ascertain that they are a certain way, (e.g., checking to ensure one has not run over an animal). Ordering focuses on a preference for an individual to have or do things a certain way or in a specific order (e.g., arranging books in a certain manner). Obsessing focuses on the tendency to have intrusive thoughts or repeated thoughts about a specific subject (e.g., intrusive violent thoughts). Hoarding focuses on the tendency to collect certain things that do not have value (e.g., collecting old envelopes). Finally, Neutralizing focuses on compulsions to do certain behaviors in order to “cancel out” certain other behaviors conceptualized as negative or bad.

Each of these dimensions of OCD are uniquely correlated with different types of psychopathology (Campos, Yoshimi, Simão, Torresan, & Torres, 2015; Torres, Cruz, Vicentini, Lima, & Ramos-Cerqueira, 2016). For example, obsessing, hoarding, and washing are associated with suicidal behaviors (Campos et al., 2015); the severity of alcohol dependence has been significantly associated with neutralizing and ordering (Campos et al., 2015); and obsessing is associated with psychosis severity (Fernandez-Egea, Worbe, Bernardo, & Robbins, 2018). However, the way in which these specific dimensions of OCD relate to EDs is not well established.

To date, there have been two studies that examined these specific dimensions of OCD in ED samples (Davies, Liao, Campbell, & Tchanturia, 2009; Naylor, Mountford, & Brown, 2011). Naylor et al. (2011) found that women with AN and bulimia nervosa (BN) report significantly higher scores on each of the six dimensions of OCD than do healthy controls. They also found that overall OCD dimensions were associated with exercise and overall ED pathology, but did not examine the associations between each dimension of OCD separately (rather they used the total OCD score). A second study, Davies et al. (2009), also found that scores on each dimension of OCD, with the exception of hoarding, were significantly elevated in individuals with a diagnosis of AN and BN versus healthy controls. They also found a significant correlation between overall ED pathology and three dimensions of OCD: neutralizing, obsessing, and ordering.

Though these studies represent an important first step in understanding the relationship between dimensions of OCD and ED pathology, more research is needed to test the unique relationships between dimensions of OCD and specific ED symptoms, to understand the most important components of OCD cognitions and behaviors in the treatment of EDs. Further, there has been no research testing OCD symptoms in a sample diagnosed with Atypical AN. Atypical AN is a subtype of Other Specified Feeding and Eating Disorders (OSFED), in which all symptoms for AN are met with the exception that body mass index (BMI) is over 18.5 (in AN, an individual must have a BMI under 18.5) (American Psychiatric Association, 2013; Forney, Brown, Holland-Carter, Kennedy, & Keel, 2017). Atypical AN is highly prevalent (2.8% lifetime prevalence compared to 0.8% AN; Stice, Marti, & Rohde, 2013; Whitelaw, Gilbertson, Lee, & Sawyer, 2014), and research reports that Atypical AN is associated with equal or higher impairment and mortality rates than other EDs (Sawyer, Whitelaw, Le Grange, Yeo, & Hughes, 2016). However, little research shows other clear distinctions between AN and Atypical AN, including in terms of psychiatry comorbidity (Moskowitz & Weiselberg, 2017). The literature that does exist suggests that AN and Atypical AN share more similarities than differences, including medical complications, rates of psychiatric comorbidity, psychological symptoms, and restriction behaviors (Coniglio et al., 2017; Moskowitz & Weiselberg, 2017; Sawyer et al., 2016; Whitelaw et al., 2014). Indeed, it has been suggested that the only significant differentiating point may be the initial weight (e.g., normal versus overweight) when weight loss began that may have resulted in an AN (underweight) versus Atypical AN (other weight) diagnosis (Forney et al., 2017; Whitelaw et al., 2014). Despite its deadliness and high prevalence, no research has characterized OCD symptomatology in Atypical AN, despite the fact that such comorbidity might lead to increased severity within Atypical AN. However, given the high rates of OCD present in AN, we also expected that OCD dimensions would be highly relevant in Atypical AN. Thus, given the current state of the literature, we would hypothesize no differences between AN and atypical AN in terms of OCD symptom dimensions, given there is no literature to support that such a difference would exist.

In the current study, we surveyed a sample of individuals diagnosed with AN or Atypical AN (Study 1: N = 139; 39 AN/100 Atypical AN and Study 2: N = 115; 36 AN/79 Atypical AN), to assess both ED and OCD symptoms. We had two primary goals. First, to test if cognitive-behavioral dimensions of OCD differed between AN and Atypical AN. Second, to detect which unique dimensions of OCD (e.g., obsessing, checking) were related to specific ED symptoms (e.g., drive for thinness, bulimia symptoms, body dissatisfaction, and overall eating pathology). We also tested the relationship between OCD and ED behaviors across time in prospective data (Study 1). We hypothesized that there would be no significant differences in OCD symptoms between AN and Atypical AN. Second, given the strong research showing cognitive distress in EDs, as well as prior research showing a correlation between obsessing and ED symptoms, we hypothesized that obsessing would be the cognitive dimension of OCD most related to ED symptoms. We also hypothesized that obsessions would predict ED behaviors (binge eating, purging, fasting) across time, given high reports of disturbing thoughts in the ED population and their specific documented relationship with ED behaviors.