Eating Behaviours of Women with an Eating Disorder and Autism Spectrum Disorder

Rationale: A large body of literature has focused on the role and clinical implications of autism spectrum traits that have been repeatedly found in eating disorder samples. On the one hand, elevated ASD traits have been associated with a more severe presentation of the ED and subsequently with poorer treatment outcome. On the other hand, recent results indicate that women with diagnosed ASD experience significantly more eating problems than controls, such as having eating rituals and sensory sensitivities, while also reporting more disordered eating behaviours. Research into the eating behaviours of women with both diagnosed ASD and an ED however is scarce. This study aims to gain insight into the eating behaviours of women with an ED and comorbid ASD by comparing their behaviours to women with an ED but without ASD and to women with ASD but without an ED). In addition to that, we are investigating whether their attitudes and behaviours towards food are associated with other (co-morbid) problems.
Objective: To gain insight into the eating behaviours of women with an eating disorder (ED) and comorbid autism spectrum disorder (ASS) as well as to investigate whether these behaviours are different compared to those of women with an ED but without ASD and whether these behaviours are comparable to those of women with ASD but without an ED. Lastly, we want to investigate whether these eating behaviours are related to other psychiatric complaints and a lower quality of life.
Study design: Cross-sectional design with at least 80 participants: at least 20 (aiming at 30) with ED and ASD, 30 with ED and 30 with ASD. Administering questionnaires about picky eating behaviour, ARFID, eating disorder related behaviours, eating disturbances related to ASD, other psychiatric complaints and quality of life.
Study population: at least 20 adult female patients [18 years of age or older] with an ED and ASD, 30 adult female patients with an ED but without ASD and 30 adult female patients with ASD but without an ED.
Main study parameters/endpoints: Primary outcome: Eating behaviour including eating disorder symptoms, picky eating behaviour, eating disturbances related to ASD and how these relate to psychiatric complaints and lower quality of life.

Primary Objective: To gain insight into the eating behaviours of adult women with an ED and comorbid ASD and how they compare to the eating behaviours of adult women with an ED but without ASD and to those of adult women with an ASD but without an ED.

Secondary Objective(s): To investigate whether these eating behaviours relate to other psychiatric complaints (BSI – Brief Symptom Inventory), neurocognitive functioning (DFlex – Detail and Flexibility Questionnaire) and lower quality of life (MHQol – Mental Health Quality of Life).

Cross-sectional investigation with one time measurement (around the time of intake) of disturbed eating behaviours such as disordered eating (EDE-Q – Eating Disorder Examination Questionnaire), picky eating behaviour (APEQ – Adult Picky Eating Questionnaire), eating disturbances found in avoidant restrictive food intake disorder (ARFID) (NIAS – Nine Item ARFID Screen) and ASD specific eating disturbances (SWEAA – Swedish Eating Assessment for Autism Spectrum Disorders).

Primary outcomes will be investigated by means of these questionnaires which will be administered by means of a link:
Description of questionnaires:

 EDE-Q (Fairburn & Beglin, 1994) to measure eating disorder symptomatology. This measure consists of 28 items with different answer possibilities, e.g. Likert scales or open ended questions asking about the frequency of a certain behaviour. The questions concern the frequency in which the patient engages in behaviours indicative of an eating disorder over a 28-day period. There is a global score as well as four subscales: Restraint, Eating Concern, Shape Concern and Weight Concern.

 SWEAA (Karlsson et al., 2013) to measure eating disturbances specific to ASD. This measure consists of 60 items comprising eight subscales, two single items, five ASD-specific items as well as demographic and medical background variables by means of a Likert scale. The measure contains questions about traditional ED behaviours such as dieting, bingeing and purging behaviour as well as about medicines and somatic and psychiatric diagnoses that could affect eating habits, weight and appetite. Other questions ask about social situations during mealtimes, perception and behaviour as well as routines regarding food and mealtimes. Scores are added to subscores within their respective subscale: Perception, Motor Control, Purchase of Food, Eating Behaviour, Mealtime Surroundings, Social Situation at Mealtime, Other Behaviour associated with Disturbed Eating, Hunger/Satiety.

 APEQ (Ellis et al., 2017) to measure picky eating behaviour. This multidimensional measure of adult picky eating attitudes and behaviours consists of 16 items and can be answered by means of a Likert scale. There is a global score and four subscales: Meal Presentation, Food Variety, Meal Disengagement, and Taste Aversion. This measure contains questions about behaviours related to picky eating behaviours such as rigid food preparation and presentation preferences, limited food variety and avoidance of novel foods, the avoidance of mealtimes, and the rejection of bitter and sour tasting food.

 NIAS (Ellis et al., 2018) to measure ARFID-associated eating behaviours.
This brief multidimensional construct consists of 9 items and can be answered by means of a Likert scale. There is again a global score and scores can be added to subscores within their respective subscale: Fear, Appetite and Picky Eating. These scales measure behaviour related to the avoidance of food or eating due to a lack of appetite/interest and fear of choking/vomiting, or GI distress or the avoidance of food because of its sensory characteristics.

Secondary outcomes will be investigated by means of questionnaires which will be administered by means of a link:
Description of questionnaires:

 BSI (Derogatis, 1993) to measure general psychiatric complaints and distress. This measure contains 53 items that are answered by means of a Likert scale. The scores are calculated according to their respective subscales: Somatization, Obsession-Compulsion, Interpersonal Sensitivity, Depression, Anxiety, Hostility, Phobic Anxiety, Paranoid Ideation, Psychoticism. As an addition to the subscale scores, scores are also added up to derive a Global Severity Index, indicating the respondents distress level and is a combination of information about the number of symptoms and the intensity of distress. A Positive Symptom Total, indicating how many symptoms the respondent is experiencing, and a Positive Symptom Distress Index, providing information about the average level of distress the respondent experiences.

 MHQoL (van Krugten et al., 2019) to measure quality of life. This measure is a standardized measure of quality of life for use in people with mental health problems. It contains two parts, a descriptive system and a visual-analogue scale. The descriptive system comprises seven questions that cover seven dimensions: Self Image, Independence, Mood, Relationships, Daily Activities, Physical Health and Future, each with four response levels (ranging from very satisfied to very dissatisfied). The descriptive system generates an MHQoL index score. The MHQoL Visual-analogue scale records the self-esteemed general psychological well-being of the respondent on a horizontal scale ranging from zero (“worst imaginable psychological well-being”) to ten (“best imaginable psychological well-being”).

 DFlex (Roberts et al., 2011) to examine participants’ inflexible daily behaviour. The DFlex entails 24 items that rated on a six-point Likert scale (with anchors strongly agree and strongly disagree) with statements comprising the management of unexpected challenges, changes of daily routines and adapting their plans to accommodate others. The DFlex also comprises a cognitive rigidity subscale and attention to detail subscale. Higher sum scores on this scale imply higher cognitive rigidity and more attention to detail (i.e. lower flexibility).

Other diagnostic or screening instruments:
See below for a description of the diagnostic or screening instruments that are used in the process to determine whether a patient can be included to the study.

 AQ-50 (Baron-Cohen et al., 2001) as an additional screen for the presence of autistic traits. The AQ-50 is a self-reported questionnaire that consists of 50 statements rated on a 4-point Likert scale from 1 = “Definitely agree” to 4 = “Definitely Agree”. While the original version of the questionnaire (Baron-Cohen et al., 2001a) has been divided into two “agree” and “disagree” scales, the Dutch version of the AQ-50 still uses the 4-point Likert scale (Hoekstra et al., 2008). The 50 statements can be separated into five subscales: Attention to detail, Social skills, Attention Switching, Communication, and Imagination. The total score on all 50 items can be calculated, resulting in a minimum score of 50 and a maximum score of 200.
The AQ-50 is not used as a diagnostic instrument in this study, but merely functions as an extra screening instrument to make sure that the ED group consists of participants that do not exhibit ASD traits. That is why we use the suggested cut-off score of 112 on the AQ-50 according to Wouters & Spek (2011) in addition to the questions about the socio-communicative development of the patient and the behavioural observation with regard to reciprocity and cognitive and behavioural flexibility.

 SCID-5-CV (First, Williams, Karg, & Spitzer, 2016) to determine an ED or ARFID diagnosis for participants at the Autism Expert Centrum. The SCID-5-CV is a semi structured interview guide to determine DSM-5 diagnoses. It is administered by a clinician or trained mental health professional who is familiar with the DSM-5 classification and diagnostic criteria. For the purposes of this study, the ED chapter of the SCID-5-CV will be used, as well as the additional ARFID module.

 DSM-5 interview ASD (Spek, 2015) to assess the presence of an ASD in particpants at Rintveld. This semi-structured interview is based on the criteria of the DSM-5 and is administered by a clinician or trained mental health professional who is familiar with the DSM-5 classification and diagnostic criteria.

 SCOFF Questionnaire (Morgan et al, 1999) to screen for core features of eating disorders in participants at the Autism Expert Centrum. It contains five questions (see under Procedure at the Autism Expert Centrum) to detect eating disordered symptomatology. Each ”yes” equals 1 point, whereas a score of 2 points indicates a likely diagnosis of an eating disorder, warranting further assessment.

 ARFID screen to screen for core features of ARFID in participants at the Autism Expert Centrum. It contains three questions to detect ARFID related symptomatology. Each ”yes” indicates a likely diagnosis of ARFID, warranting further assessment.