Introduction: Constitutional thinness (CT) is an underweight state characterized by normal menstruations and no change in feeding behaviour. Thinness is the only resemblance between Anorexia Nervosa (AN) and CT. Removal of amenorrhea in the new DSM 5 definition of AN might lead to misdiagnosis between these two populations. The objective of this study was to compare CT, AN and Control subjects in terms of biological, anthropometric, and psychological markers in order to better distinguish AN from CT subjects.
Patients and method: This retrospective study conducted from 2000 to 2015 included three groups of young women: fifty-six CT, forty restrictive-type AN and fifty-four control subjects were included in the study. Basal evaluation was made in all groups with any intervention. Body composition, nutritional markers, pituitary hormones, bone markers and psychological scores were evaluated in three groups. For every markers, a receiver Operator Characteristics (ROC) curve was calculated to evaluate the accuracy of differentiation between AN and CT groups.
Results: For the majority of the studied parameters, CT subjects were similar to controls but dramatically different from AN subjects. According to ROC data, while psychological scores were unsuccessful to differentiate AN from CT, except for DEBQ Restrained Eating item, free-T3 and Leptin were strong tools for AN and CT distinction as they displayed high sensitivities and specificities with low P values (<0.0001).
Conclusions: Taking AN and CT distinction as a major objective, the exclusive use of psychological criteria is not sufficient and should be completed by at least a free T3 determination, which is a cheap and an accessible laboratory testing for general practitioners. The final goal is to avoid social stigmatization and excessive useless therapies for CT women.
20 - 23 May 2017
European Society of Endocrinology