Evaluation of visual food stimuli paradigms on healthy adolescents for future use in fMRI studies in anorexia nervosa.

Background: Mostly, visual food stimuli paradigms for functional Magnetic Resonance Imaging are used in studies of eating disorders. However, the optimal contrasts and presentation modes are still under discussion. Therefore, we aimed to create and analyse a visual stimulation paradigm with defined contrast.

Methods: In this prospective study, a block-design fMRI paradigm with conditions of randomly altering blocks of high- and low-calorie food images and images of fixation cross was established. Food pictures were rated in advance by a group of patients diagnosed with anorexia nervosa to address the dedicated perception of patients with eating disorders. To optimize the scanning procedure and fMRI contrasts we have analysed neural activity differences between high-calorie stimuli versus baseline (H vs. X), low-calorie stimuli versus baseline (L vs. X) and high- versus low-calorie stimuli (H vs. L).

Results: By employing the developed paradigm, we were able to obtain results comparable to other studies and analysed them with different contrasts. Implementation of the contrast H versus X led to increased blood-oxygen-level-dependent signal (BOLD) mainly in unspecific areas, such as the visual cortex, the Broca´s area, bilaterally in the premotor cortex and the supplementary motor area, but also in thalami, insulae, the right dorsolateral prefrontal cortex, the left amygdala, the left putamen (p < .05). When applying the contrast L versus X, an enhancement of the BOLD signal was detected similarly within the visual area, the right temporal pole, the right precentral gyrus, Broca´s area, left insula, left hippocampus, the left parahippocampal gyrus, bilaterally premotor cortex and thalami (p < .05). Comparison of brain reactions regarding visual stimuli (high- versus low-calorie food), assumed to be more relevant in eating disorders, resulted in bilateral enhancement of the BOLD signal in primary, secondary and associative visual cortex (including fusiform gyri), as well as angular gyri (p < .05).

Conclusions: A carefully designed paradigm, based on the subject's characteristics, can increase the reliability of the fMRI study, and may reveal specific brain activations elicited by this custom-built stimuli. However, a putative disadvantage of implementing the contrast of high- versus low-calorie stimuli might be the omission of some interesting outcomes due to lower statistical power. Trial registration NCT02980120.