Despite a wide range of proposed risk factors and theoretical models, prediction of eating disorder (ED) onset remains poor. This study undertook the first comparison of two machine learning (ML) approaches [penalised logistic regression (LASSO), and prediction rule ensembles (PREs)] to conventional logistic regression (LR) models to enhance prediction of ED onset and differential ED diagnoses from a range of putative risk factors.

Data were part of a European Project and comprised 1402 participants, 642 ED patients [52% with anorexia nervosa (AN) and 40% with bulimia nervosa (BN)] and 760 controls. The Cross-Cultural Risk Factor Questionnaire, which assesses retrospectively a range of sociocultural and psychological ED risk factors occurring before the age of 12 years (46 predictors in total), was used.

All three statistical approaches had satisfactory model accuracy, with an average area under the curve (AUC) of 86% for predicting ED onset and 70% for predicting AN v. BN. Predictive performance was greatest for the two regression methods (LR and LASSO), although the PRE technique relied on fewer predictors with comparable accuracy. The individual risk factors differed depending on the outcome classification (EDs v. non-EDs and AN v. BN).

Even though the conventional LR performed comparably to the ML approaches in terms of predictive accuracy, the ML methods produced more parsimonious predictive models. ML approaches offer a viable way to modify screening practices for ED risk that balance accuracy against participant burden.

To assess the differences in comorbid lifetime substance use (tobacco, alcohol and drug use) between eating disorder (ED) patients and healthy controls.

Participants were a consecutive series of 779 ED cases, who had been referred to specialised ED units in five European countries. The ED cases were compared to a balanced control group of 785 healthy individuals. Assessment: Participants completed the Substance Use Subscale of the Cross Cultural Questionnaire (CCQ), a measure of lifetime tobacco, alcohol and drug use. In the control group, also the GHQ-28, the SCID-I interview and the EAT-26 were used.

ED patients had higher lifetime consumption of tobacco and drugs (p <0.01). The only insignificant result was obtained for alcohol (OR= 1.29; δ =0.157; N.S.) and cannabis use (OR= 1.21; δ = 0.037, N.S.). Significant differences across ED sub diagnoses also emerged for all of the assessed variables (p<0.01), with the BN and AN-BP patients generally presenting the highest prevalence rates. The only exception was detected for alcohol consumption where EDNOS patients demonstrated the highest values (p=0.008). Only a few cultural differences between countries emerged (p<0.05).

Lifetime tobacco and drug use but not alcohol consumption are more prevalent in ED patients than healthy controls. While alcohol appears to be more common in EDNOS, smoking and drug use are more frequent in patients with bulimic symptomatology. The differential risk observed in patients with bulimic features might be related to differences in temperament or might be the result of increased sensitivity to reward.

To examine whether there is an association between individual and family eating patterns during childhood and early adolescence and the likelihood of developing an eating disorder (ED) later in life.

Participants were a consecutive series of 879 ED cases from five different European countries. The ED cases were compared to a control group of 785 healthy individuals. Assessment: Participants completed the Early Eating Environmental Subscale of the Cross-Cultural (Environmental) Questionnaire (CCQ), a retrospective measure, which has been developed to detect dimensions associated with EDs in different countries. In the control group, also the GHQ-28, the SCID-I interview and the EAT-26 were used.

Five individual CatPCA procedures revealed five predetermined dimensions which were labeled: 1.) food as individualization; 2.) control and rules about food; 3.) food as social glue; 4.) healthy eating and 5.) food neglect. Logistic regression analyses indicated that the domains with the strongest effects were: food used as individualization (p=0.001; OR=1.76) and control and rules about food (p=0.001; OR=1.76). Conversely, healthy eating was negatively related to a later ED (p=0.001; OR=0.629). The pattern of associated ED factors was found to very between countries. There was very little difference in early eating behavior on the subtypes of the ED.

The fragmentation of meals within the family and control and rules about food appears to be linked to the development of a subsequent ED. On the other hand mantaining a structured and balanced diet during infancy seems to protect from a later ED.

Aetiological studies of eating disorders would benefit from a solution to the problem of instability of eating disorder symptoms. We present an approach to defining an eating disorders phenotype based on the retrospective assessment of lifetime eating disorders symptoms to define a lifetime pattern of illness. We further validate this approach by testing the most common lifetime categories for differences in the prevalence of specific childhood personality traits.

Ninety-seven females participated in this study, 35 with a current diagnosis of restricting anorexia nervosa, 32 with binge/purging subtype of anorexia nervosa and 30 with bulimia nervosa. Subjects were interviewed by a newly developed EATATE Lifetime Diagnostic Interview for a retrospective assessment of the lifetime course of eating disorders symptoms and childhood traits reflecting obsessive–compulsive personality.

The data illustrate the extensive instability of the eating disorders diagnosis. Four most common lifetime diagnostic categories were identified that significantly differ in the prevalence of childhood traits. Perfectionism and rigidity were more common in groups with a longer duration of underweight status, longer episodes of severe food restriction, excessive exercising, and shorter duration of binge eating.

The assessment of lifetime symptoms may produce a more accurate definition of the eating disorders phenotype. Obsessive–compulsive traits in childhood may moderate the course producing longer periods of underweight status. These findings may have important implications for nosology, treatment and future aetiological studies of eating disorders.