Alterations in serotonin transporter and body image-related cognition in anorexia nervosa

Abstract

The serotonin system has been implicated in the pathophysiology of anorexia nervosa (AN). A recent report proposed that body image distortion (BID), a core symptom of AN, may relate to abnormalities of the serotonin system, especially the serotonin transporter (5HTT). Positron emission tomography (PET) studies of underweight patients with active AN reported alterations in serotonin receptors, but not 5HTT. Here, we aimed to disclose the clinicopathophysiology of AN by focusing on 5HTT and cognitive functions, including BID, in groups with active AN. Twenty-two underweight female patients with AN (12 restricting-type AN (ANR); 10 binge-eating/purging-type AN (ANBP)) and 20 age-matched healthy female subjects underwent PET with a 5HTT radioligand [¹¹C]DASB. The binding potential (BP_ND) of [¹¹C]DASB was estimated semiquantitatively, and clinical data from Raven's colored progressive matrices for general intelligence, the Stroop test for focused attention, the Iowa gambling task for decision making and a dot-probe task designed for BID were compared with the levels of BP_ND in different groups. [¹¹C]DASB BP_ND was significantly decreased in the medial parietal cortex in patients with AN and in the dorsal raphe in patients with ANR compared with healthy subjects (p < .05 corrected). Patients with ANR showed a significantly negative correlation between [¹¹C]DASB BP_ND in the dorsal raphe and performance on the dot-probe task (p < .05 corrected). While reduced 5HTT in the medial parietal cortex (the somatosensory association area) is pathophysiologically important in AN in general, additional 5HTT reduction in the dorsal raphe as seen in ANR is implicated for the clinicopathophysiological relevance.

Keywords: Anorexia nervosa; Body image distortion; Positron emission tomography; Serotonin transporter.

Fig. 1

Design of the dot-probe task. (A) One trial of the dot-probe task consists of three phases (i.e., the fixation, perception, and decision phases). (B) The stimulus frames consist of a pair of computer-generated distorted images of another person's body.

Fig. 2

Regions of significantly decreased [¹¹C]DASB binding potential (BP_ND) values in the patients with AN, ANR, and ANBP compared with healthy subjects. The images show regions of significantly decreased [¹¹C]DASB binding potential (BP_ND) values with a significance threshold set at p < .001 uncorrected at the voxel level and p < .05 corrected for multiple comparisons using familywise error at the cluster level.

Fig. 3

[¹¹C]DASB binding potential (BP_ND) in the dorsal raphe, thalamus, dorsal caudate, antero-ventral striatum, subgenual cingulate, medial parietal cortex and cerebellar cortex among the subtypes of anorexia nervosa and the healthy subjects. The open circles represent healthy subjects, the closed triangles represent patients with restricting-type anorexia nervosa (ANR), and the closed inverted triangles represent patients with binge-eating-/purging-type anorexia nervosa (ANBP). Error bars represent the standard errors. An asterisk (*) indicates significance at p < .05 using one-way ANOVA.

Fig. 4

Correlations of the patients with restricting-type anorexia nervosa between [¹¹C]DASB binding potential (BP_ND) values and the reaction times on the dot-probe task in the dorsal raphe, thalamus, dorsal caudate, antero-ventral striatum and subgenual cingulate. The closed triangles represent patients with restricting-type anorexia nervosa (ANR) in the scattergrams. The significance threshold was set at p < .05 with the Bonferroni correction for multiple comparisons.