A cross-sectional observational analytical study was performed. The participants were women who consecutively presented for treatment at the Eating Disorders Outpatient Unit of General University Hospital of Ciudad Real during a 6-month period. Inclusion criteria were (1) diagnosis of anorexia nervosa (AN), bulimia nervosa (BN), binge eating disorder (BED) or eating disorder not otherwise specified (EDNOS) according to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5); (2) female gender; (3) age between 18 and 50 years; (4) a current clinical state that permitted psychometric evaluation. Comorbid symptoms of depression and anxiety were permitted. Patients were excluded if they 1) spoke a native language other than Spanish; (2) did not give written informed consent; (3) were male; (4) had comorbid bipolar disorder, psychotic disorder, major depressive disorder, organic mental disorder, attention-deficit hyperactivity disorder, autism spectrum disorder, Tourette syndrome or chronic fatigue syndrome.
Participants
At the beginning of the study, 74 patients were included. The mean age was 28.97 years (standard deviation, SD, 9.30). The participants were informed about the procedure, and those who agreed to participate signed the informed consent. Clinical assessment sand neuropsychological tests were administered, and the participants were given psychopathological self-report questionnaires to complete. Twenty-three participants (31.08%) did not return the questionnaires.
Although the sample was finally reduced to 51 subjects for most of the psychopathological variables, neuropsychological and clinical features could be evaluated in the 74 patients who were initially included. When the patients who did not return the questionnaires were compared to the ones who did, no differences were found regarding age, body mass index (BMI), age at onset, course of the disorder, education level, use of drugs, diagnoses, compensatory symptoms or neuropsychological scores.
Procedure
All patients who participated in the study were offered one year of treatment. This included twelve personal sessions of motivational enhancement therapy, six psycho-educational group sessions and twenty cognitive-behavioural therapy sessions. Treatment was based in Fairburn’s program (43) with 20 50-minute sessions based on enhanced cognitive behaviour therapy.Pharmacological treatment was provided when necessary.
Each day, selected patients were informed about the study and its procedure, and informed consent was obtained. After this, the self-report questionnaires were provided. Maximum confidentiality was guaranteed by removing any identifying information from the patients’ clinical data and storing data in a password-protected Excel base. The local Clinical Research Ethics Committee approved the study.
Instruments
All psychopathological tests were given to the patients to complete at the hospital while they waited for their appointment. The evaluation started with the Letter Number Sequencing test, followed by the Stroop test, the Symbol Digit Modalities Test and the copying portion of the Rey-Osterrieth Complex Figure Test (ROCFT). As it was necessary to wait five minutes before performing the drawing from memory task, the clinical questionnaire was given to the patients after the figure copying task; when the questionnaire was completed, the patients were asked draw the complex figure from memory.
Clinical test:
Clinical interview
A clinical interview was completed to collect information about symptoms, body parameters, exclusion and inclusion criteria and diagnoses and to obtain data regarding age at onset and education level (necessary for neuropsychological evaluation).
Status
The patient’s clinical status was evaluated by the Psychiatric Status Rating Scale. This scale is based on Herzog’s work (30) and was updated with the DSM-5 criteria for the different disorders. Clinical status was divided into three groups, each of which included two different categories: scores of 5-6 indicated that the patient met the criteria for the disorder (“Definite criteria, severe” and “Definite criteria”); scores of 3-4 indicated partial recovery (“Marked” and “Partial remission”), and scores of 1-2 indicated total recovery (“Residual” and “Usual self”). Afterwards, to improve statistical power, these categories were combined to create a dichotomous variable: “No evidence of illness” (Usual self, Residual, Partial remission) and “Evidence of illness” (all other categories). The original scale is included in additional file 5.
Psychopathological tests:
Body Shape Questionnaire (BSQ)
The BSQ was validated for the Spanish population by means of a 34-item self-report questionnaire with a six-point Likert-style rating scale (1= never; 6= always) (44). It measures concerns about body weight and shape in ED. Its internal consistency using Cronbach’s index is good (0.97), as is its concurrent validity. It differentiates clinical from nonclinical subjects and people with more or less concern about their weight. A score of 105 is the cut-off point for the Spanish population (44).
Eating Disorders Inventory (2nd edition) (EDI-2)
The EDI-2 is a self-report questionnaire with 91 items (11 scales: “Drive for thinness”, “Bulimia”, “Body dissatisfaction”, “Ineffectiveness”, “Perfectionism”, “Interpersonal distrust”, “Interoceptive awareness”, “Maturity fear”, “Ascetism”, “Impulsiveness” and “Social insecurity”). The internal consistency (Cronbach’s alpha) is higher than 0.80 in ED samples. Reliability coefficients (alpha) were from 0.83 to 0.93 in samples of patients from the original studies. The EDI-2 is validated for the Spanish population (45,46).
Beck Depression Inventory (BDI)
The BDI is a self-reported questionnaire with 21 items for evaluating the existence and severity of depressive symptoms. Its internal consistency is higher than 0.85 (47,48).
State-Trait Anxiety Inventory (STAI)
The STAI is divided in two independent scales that measure state and trait anxiety. Cronbach’s alpha is 0.90 for trait anxiety and 0.94 for state anxiety (49–51).
Dissociative Experiences Scale(DES)
The DES is a self-assessment questionnaire based on a visual analogue scale composed of 28 dissociative experiences; it asks how often the subject experiences these dissociative events (0-100%). An average score of 30 or more indicates that dissociative disorder is expected (52–55). The DES comprises 3 scales: “Derealization and depersonalization”, “Absorption” and “Amnesia” (56). Its internal reliability (Cronbach’s alpha) is 0.93, and its retest-reliability is 0.87 (56).
Attitudes Towards Change in Eating Disorders Scale(ACTA)
The ACTA, a Spanish validated self-report questionnaire with 59 items, was used to evaluate the attitudes towards change in cognitive, affective, behavioural, and relational features in ED patients (46). It is based on Prochaska and DiClemente’s theoretical model concerning the phases of change (57). The ACTA comprises 6 scales: “Precontemplation”, “Contemplation”, “Decision”, “Action”, “Maintenance” and “Relapse”. The “Precontemplation” subscale refers to the refusal to consider the presence of a disorder. The “Contemplation” subscale reflects a state in which the subject recognizes his/her eating problem, although he/she may under-evaluate its importance and shows no motivation to change his/her behaviours. The "Decision" subscale reflects that the patient has decided on a date at which he/she will begin to make changes. The "Action" subscale shows evidence of change in different areas: cognitive (positive thoughts about the resolution of the disorder), behavioural (changing habits for other, healthier ones), and affective (the patient perceives the disorder as an unpleasant problem that must be modified). The "Maintenance" subscale evaluates the stability of the achievements obtained in the action phase. Finally, the "Relapse" subscale is a subjective assessment of any worsening that the subject may experience. The reliability of each scale is from 0.90-0.74, and their retest-reliability ranges from 0.86-0.64 (46). It is designed for ED patients. An additional file describes the scale in greater detail [see Additional file 1] (in this study, the Spanish version was used).
Neuropsychological variables:
Working memory
This variable was measured using the Letter Number Sequencing test (LNS III) included in Wechsler Adult Intelligence Scale (WAIS-III) (58). For this test, participants listen to verbal sequences of letters and numbers that gradually increase in length and must be memorized, ordered and repeated back to the tester. The reliability and validity of this test are high, but the available data are related to intelligence evaluation. Two indexes are obtained: Total score and SpanLN.
Inhibitory attention
“Stroop’s effect” (59) is defined as the difference in the processes of naming colours and reading words. It is examined using the Stroop test, which measures inhibitory attention. Test-retest reliability is high in all studies (0.69-0.89) (60). The test comprises three pages: Two of them represent the “congruous condition”, in which participants are required to read the names of colours (henceforth referred to as colour words) printed in black ink (W) and to name different colour patches (C). On the third page, which represents the colour word (CW) condition, colour words are printed in an inconsistent colour of ink (for instance, the word “red” is printed in green ink). Thus, in this incongruent condition, participants are required to name the colour of the ink instead of reading the word. In other words, the participants are required to perform a less automated task (i.e., naming ink colour) while inhibiting the interference arising from a more automated task (i.e., reading the word). An additional file describes this test in more detail [see Additional file 2] (in this study, a Spanish version was used). The indexes obtained were W (the number of words read on the first page), C (the number of colours named on the second page), CW (the number of colours named on the third page) and interference. All these indexes are corrected by age.
Sustained attention
The Symbol Digit Modalities Test (SDMT) (61,62) mainly evaluates attention (included sustained attention) (63), optical tracking, speed of mental processing and visuo-motor speed. The test is based on the pairing of meaningless geometrical images with numbers from 1 to 9 according to a previous model. The indexes obtained are the total score and the scored SDMT (corrected by age and education level). [Additional file 3]
Executive function
The ROCFT (64,65) consists of copying a complicated line drawing and then drawing it from memory. Several indexes are obtained: quantitative and qualitative scores of copied and memorized reproduction accuracy (scored using Osterrieth’s method, as explained in Additional file 4), style, order and central coherence (explained in Additional file 4) (66–69). A global approach, for example, is shown by a tendency to draw the main structural elements first and place local elements in relation to this framework. On the other hand, individuals may copy the figure by first drawing local details and failing to maintain their overall spatial organisation. An extreme piecemeal approach is often seen in very young children and individuals with right hemisphere damage, although healthy adults differ in the degree to which they use a local or global strategy. Scoring and the style, order and coherence indexes are explained in Additional file 4. The indexes obtained were copy and memory accuracy, copy and memory time, copy and memory type, order index, style index and central coherence index.
Data analysis
SPSS 19.0 was used to analyse the data obtained.
A descriptive analysis was performed, followed by an inferential analysis. The statistical level accepted as significant was 5% (p<0.05). Qualitative variables were statistically analysed by means of the chi-square test for independence. Quantitative variables were examined for normal distribution using the Kolmogorov-Smirnov test if there were more than 50 data and the Shapiro-Wilk if there were fewer than 50 data points). Student’s t-test and the Mann-Whitney U test were used for dichotomous independent variables, and ANOVA and the Kruskal-Wallis test were used for independent variables with more than two categories. Pearson’s and Spearman’s correlations were used for quantitative independent variables.
Multivariate analysis was performed using simple linear regression to evaluate models of quantitative dependent variables. The variables included in the models were those that were statistically significant in the bivariant analysis of our sample and those that were statistically significant in previous studies or were considered of interest for the current study due to their transdiagnostic meaning (e.g., BMI, dichotomous status, age at onset of ED, antidepressant use, benzodiazepine use, total DES score, anxiety, depression, Perfectionism score). Later, the models were simplified to improve their alignment.