Reflective functioning in anorexia nervosa: Does it differ from healthy controls and how is its relation to psychopathology?

Background: Previous studies found an impairment in the capacity to mentalize (operationalized as " reflective functioning ", RF) in patients with anorexia nervosa (AN), but only few studies used a validated interview procedure. The aim of this study was an assessment of RF in patients with anorexia nervosa in comparison to healthy subjects, using both an expert-rated measure as well as a self-report measure. Further, the study aimed to explore the relationship between RF and various aspects of psychopathology. Methods: 30 patients with AN and 30 matched healthy control subjects were assessed and compared regarding their level of RF using the Brief Reflective Functioning Interview (BRFI) and the Mentalization Questionnaire (MZQ). The correlations between RF values and eating disorder psychopathology (BMI, EDI, EDE-Q), body experience (DKB-35), general psychopathology (PHQ-9, PHQ-15, GAD-7) and impairment in personality functioning (OPD-SQS) were explored using simple linear regression analyses. Results: Regarding mentalizing on the RF-Scale (RF-BRFI), RF was M = 3.63 (SD = .67) in the patient group and M = 4.13 (SD = .94) in the healthy controls (HC). It showed to be significantly lower for patients with AN than for HCs (t(52.5) = 2.38; p = .011). Mentalizing in the MZQ was M = 3.24 (SD = .71), demonstrating significantly worse self-reported mentalizing for the patients (t(49.1) = − 9.38; p < .0001) compared to HCs (M = 1.73; SD = .45). Higher expert-rated RF was associated with better self-reported mentalizing (r = − .38; p = .006), but no significant correlation for the individual groups of patients ond controls could be found. RF as measured with the BRFI was only correlated to the EDI-subscale “weight concern”, while RF values of the MZQ were correlated with various aspects of psychopathology except depressive symptoms and BMI. Conclusion: The study could replicate the finding that patients with AN show impaired mentalizing. However, it seems likely, that observer-rated and self-report measures for mentalizing cover different aspects of the construct of mentalizing and should be regarded as complementary rather than interchangable. This may also explain different associations with features of psychopathology. Trial registration DRKS00031108