As expected, individuals with AN demonstrated higher levels of body image distortion as indicated by higher BSE errors on the Light Beam Apparatus, which confirms our hypothesis that AN individuals would overestimate their body size more than healthy controls. Additionally the AN group showed higher levels of body dissatisfaction than the HC group, through a number of self report measures, supporting previous findings .
Contrary to our hypothesis, previously reported haptic perception impairments in AN were not supported by our findings, with no significant difference between the AN and HC groups.
This difference may be accounted for by the increased time taken by the AN group to complete the task, which increased accuracy. This contrasts with Grunwald et al.  finding, where the groups took an equal amount of time. The AN group in this study were more inaccurate on the task than in Grunwald et al.  study, but the HC group were also more inaccurate, comparable with the AN group. This may point to a more impulsive and therefore inaccurate style within the HC group.
In accordance with our hypothesis the AN group reported significantly higher levels of perfectionism compared to controls.
In addition, associations between body image dissatisfaction, perfectionism, haptic perception and body image distortion were explored. There was no evidence to support a relationship between the haptic perception tasks and the body size estimation task, thus failing to support the hypothesis that fundamental haptic perception ability affects body distortion in AN. However, all the attitudinal, self-report measures were significantly correlated with body image distortion. Body dissatisfaction showed the strongest negative relationship – the smaller the desired silhouette than the actual, the greater the overestimation of body size, which supports Cash and Deagle’s  proposal that body overestimation may contribute to body dissatisfaction, though we can make no claims as to the direction of the relationship between the two. This links the concepts of dissatisfaction and distortion in individuals with AN, despite proposed separate neural correlates of each .
Overall, the positive relationships between body dissatisfaction and perfectionism with body size estimation in the AN group would suggest that attitudes (cognitions and affect) and behaviours are significantly related to body image distortion in AN.
The findings suggest that attitudinal factors and perfectionism are related to body distortion, but that there is no significant relationship between body distortion and haptic perceptual performance, at least as measured on a ‘neutral’ task involving haptic perception. This suggests that there are no fundamental haptic perceptual problems underlying body image disturbances in AN, but does not negate findings of sub-optimal visual perceptual performance in AN. Nor does it negate findings of parietal lobe dysfunction in AN (e.g. ), related to body image issues, which may involve difficulties in integrating sensory information from different modalities, rather than separate deficits in perception per se, as suggested by Case et al.  in their investigation of performance on the size-weight illusion task. In this context, it may be that haptic perception in individuals with AN is intact, at the fundamental level, but is overridden by an increased sensitivity to visual input, as seen in the rubber hand illusion , which then leads to a visual distortion of body image, and a lack of attention to proprioceptive or interoceptive information. Equally, it may be that body dissatisfaction impacts on the visual mental image of the body, which then also affects tactile perception when related to one’s own body, as suggested by Keizer et al. . It will be of interest to use such “body-related” measures in future research to determine if a more salient focus does have an effect.
The strengths of this study are that, compared to previous work in the field, the two components of body image were measured separately, rather than treating them as a unitary concept. Confounding body dissatisfaction and body distortion has been argued to be a reason for mixed findings in the literature . The haptic perception task provided a ‘neutral’ measure of fundamental somatosensory perception, divorced as far as possible from body attitudes. The BSE task, whilst not free from attitudinal biases, was intended to give as clear a picture of body image distortion as possible, including re-evaluation of the completed gestalt silhouette, and avoiding the use of distressing images of the individual. Coupled with self-report measures, this was intended to give as rounded a picture of body image disturbance as possible.
Some limitations must be considered. Firstly, the AN group spent significantly longer on the haptic perception task compared to the HC group, which may have allowed them to be more accurate than they would have been if there was a time limit imposed. Secondly, analyses of the angle paradigm task were found to be underpowered (post hoc analysis) which may explain non-significant findings.
Future research should replicate the Angle Paradigm Task performance in AN (and its subtypes) compared to HC with greater numbers as this study may have been underpowered in this domain. Given that the AN group took longer to complete this task, a time limit would be useful to determine any impact on accuracy. The use of both uni- and bi-manual tasks would assist in identifying performance related to sensory integration, and to haptic perception. A battery of haptic tasks with and without visual feedback would also clarify the picture of sensory processing in the disorder, as would comparison of ‘body-neutral’ tasks, with ‘own-body focused’ paradigms, likely to activate different bodily representations. Further exploration of the role of perfectionism in body image disturbance is warranted, and whether this relates to a subgroup of individuals with obsessive-compulsive traits. It will be useful to relate this to the specific perfectionism dimensions linked to eating disorder symptomatology, and then to body dissatisfaction and distortion in particular.
Clinically, this study highlights the role of cognitive behavioural interventions in modifying beliefs relating to body image and the use of exposure (behavioural tasks, mirror exposure) in addressing body size distortion (see [51–53]). Additionally, given the high levels of perfectionism and its association with body distortion, techniques which focus on acceptance (e.g. mindfulness; ) or on adapting CBT for clinical perfectionism  with specific reference to body image may be beneficial.