Severe forms of Anorexia Nervosa (AN) are characterized by medical complications, psychiatric comorbidity, and high mortality. This study investigated potential associations between clinical/biological factors and the severity of AN, measured by the Body Mass Index (BMI). Red and white blood cells, hemoglobin, platelets, iron, vitamins D and B12, folate, and total cholesterol were measured in a mixed sample of 78 inpatients and outpatients. Linear regressions and one-way analyses of variance (ANOVAs) were carried out to evaluate the relationship between BMI and clinical/biochemical variables. BMI was significantly lower in hospitalized patients (F = 4.662; p = 0.034) and in those under pharmacological treatment (F = 5.733; p = 0.019) or poly-therapy (F = 5.635; p = 0.021). Higher vitamin B12 (β = −0.556, p < 0.001), total cholesterol (β = −0.320, p = 0.027), and later age at onset (with a trend towards significance) (β = −0.376, p = 0.058) were associated with a lower BMI. Increased total cholesterol and vitamin B12, later age at onset, current pharmacological treatment, and poly-therapy might be distinctive in patients with a lower BMI. In clinical practice, these findings may contribute to the early identification of AN patients at higher risk of developing complicated or chronic forms of the disorder. Further studies on larger samples are needed to identify potential predictive factors of AN severity in the framework of precision medicine.

The aim of the present study was to investigate the potential associations between clinical/socio-demographic variables and the presence of purging/binge-eating episodes in eating disorders (EDs). Clinical/socio-demographic variables and psychometric scores were collected. Groups of patients were identified according to the presence or absence of purging or objective binge-eating episodes (OBEs) and compared through t-test and chi-square tests. Binary logistic regression analyses were run. A sample of 51 ED outpatients was recruited. Patients with purging behaviors had a longer duration of untreated illness (DUI) (t = 1.672; p = 0.019) and smoked a higher number of cigarettes/day (t = 1.061; p = 0.030) compared to their counterparts. A lower BMI was associated with purging (OR = 0.881; p = 0.035), and an older age at onset showed a trend towards statistical significance (OR = 1.153; p = 0.061). Patients with OBEs, compared to their counterparts, were older (t = 0.095; p < 0.001), more frequently presented a diagnosis of bulimia or binge-eating disorder (χ² = 26.693; p < 0.001), a longer duration of illness (t = 2.162; p = 0.019), a higher number of hospitalizations (t = 1.301; p = 0.012), and more often received a prescription for pharmacological treatment (χ² = 7.864; OR = 6.000; p = 0.005). A longer duration of the last pharmacological treatment was associated with OBE (OR = 1.569; p = 0.046). In contrast to purging, OBE was associated with a more complicated and severe presentation of ED. A lower BMI and a later age at onset, as well as long-lasting previous pharmacological treatments, may predict the presence of purging/binging. Further research is needed to thoroughly characterize ED features and corroborate our preliminary findings.