Elsevier

Neuropsychologia

Volume 108, 8 January 2018, Pages 135-146
Neuropsychologia

Brain activity elicited by viewing pictures of the own virtually amputated body predicts xenomelia

https://doi.org/10.1016/j.neuropsychologia.2017.11.025Get rights and content

Highlights

  • Confronting xenomelia subjects with pictures showing their desired body state.

  • Pleasantness/attractiveness ratings are higher for amputated bodies in xenomelia.

  • Brain activity predicts xenomelia when viewing pictures of one’s own amputated body.

  • Pictures of one’s own amputated body with prosthesis classifies xenomelia subjects.

  • Activity in parietal and visual areas showed highest classification accuracy.

Abstract

Background

Xenomelia is a rare condition characterized by the persistent desire for the amputation of physically healthy limbs. Prior studies highlighted the importance of superior and inferior parietal lobuli (SPL/IPL) and other sensorimotor regions as key brain structures associated with xenomelia. We expected activity differences in these areas in response to pictures showing the desired body state, i.e. that of an amputee in xenomelia.

Methods

Functional magnetic resonance images were acquired in 12 xenomelia individuals and 11 controls while they viewed pictures of their own real and virtually amputated body. Pictures were rated on several dimensions. Multivariate statistics using machine learning was performed on imaging data.

Results

Brain activity when viewing pictures of one's own virtually amputated body predicted group membership accurately with a balanced accuracy of 82.58% (p = 0.002), sensitivity of 83.33% (p = 0.018), specificity of 81.82% (p = 0.015) and an area under the ROC curve of 0.77. Among the highest predictive brain regions were bilateral SPL, IPL, and caudate nucleus, other limb representing areas, but also occipital regions. Pleasantness and attractiveness ratings were higher for amputated bodies in xenomelia.

Conclusions

Findings show that neuronal processing in response to pictures of one's own desired body state is different in xenomelia compared with controls and might represent a neuronal substrate of the xenomelia complaints that become behaviourally relevant, at least when rating the pleasantness and attractiveness of one's own body. Our findings converge with structural peculiarities reported in xenomelia and partially overlap in task and results with that of anorexia and transgender research.

Introduction

Xenomelia (McGeoch et al., 2011), also known as body integrity identity disorder (First, 2005, First and Fisher, 2012) or apotemnophilia (Brang et al., 2008), is an unusual and rare condition in which sufferers express the persistent desire for the amputation of limb(s) that are physically fully functioning. Xenomelia individuals typically report that their unwanted limb(s) do not belong to themselves and that they would feel better if the limb(s) were to be removed (Blanke et al., 2009). Psychiatric and neurocognitive assessments of xenomelia individuals reveal no known disease that could account for the amputation desire, and especially the absence of psychotic disorders has been repeatedly reported (Blom et al., 2012, Brugger and Lenggenhager, 2014, First, 2005, Hilti et al., 2013, van Dijk et al., 2013).

Since it has been proposed that xenomelia might be a neurological disorder (Brang et al., 2008, McGeoch et al., 2011), particularly a new right parietal lobe syndrome (McGeoch et al., 2011), there is growing evidence that both brain structure (Blom et al., 2016, Hänggi et al., 2016, Hänggi et al., 2017, Hilti et al., 2013) and brain function (Hänggi et al., 2017, McGeoch et al., 2011, van Dijk et al., 2013) differ between xenomelia individuals and controls as reviewed elsewhere (Brugger et al., 2016). There is evidence that social components might play a role in the condition as well (Brugger et al., 2013).

Structural and functional neuroimaging studies in xenomelia revealed that the right superior (SPL) and inferior parietal lobule (IPL) (Hänggi et al., 2017, Hilti et al., 2013, McGeoch et al., 2011, van Dijk et al., 2013), right anterior insula (Brang et al., 2008, Hilti et al., 2013), and right primary and secondary somatosensory cortex (SI and SII, respectively) (Hilti et al., 2013, van Dijk et al., 2013), particularly its left leg representation (Hänggi et al., 2017, Hilti et al., 2013) are altered in the condition compared with controls. The paracentral lobule, supplementary motor area, basal ganglia, thalamus, and the cerebellum (Blom et al., 2016, Hänggi et al., 2016, Hänggi et al., 2017) are additionally altered in xenomelia.

Taken together, there is strong evidence that many areas of the sensorimotor system and the insula are altered in xenomelia individuals as revealed by anomalies in cortical thickness and surface area derived from surface-based morphometry (Hilti et al., 2013). Differences in the magnetoencephalographic signal (McGeoch et al., 2011) and neuronal activity measured with functional magnetic resonance imaging (fMRI) (van Dijk et al., 2013) in response to tactile stimulation have been reported as well. Anomalies have also be found in structural and functional connectivity derived from diffusion tensor imaging-based fibre tractography and seed-to-seed-based correlation analyses of resting state fMRI data, respectively (Hänggi et al., 2017), and in thalamic, putaminal, and pallidal shape (Hänggi et al., 2016).

Intriguingly, all these brain regions are associated with the construction and the maintenance of a coherent body image (Berlucchi and Aglioti, 2010, Gerardin et al., 2003, Giummarra et al., 2008, Moseley et al., 2012, Tsakiris et al., 2007, Wolbers et al., 2003). Investigations of healthy human subjects provide strong evidence that both SPL and IPL mediate body ownership (Kammers et al., 2009) and are key for the integrative mental imagery of the configurations of limbs (Wolbers et al., 2003). In particular the right-hemispheric SPL was involved in monitoring illusory limb displacements, independent of the limb's laterality (Naito et al., 2005). In addition, primary and secondary somatosensory hand representations are crucial in mediating the illusion of a rubber hand, where the visual observation of a rubber hand being touched simultaneously with the experience of touch to one's own hand evokes the sensation of an incorporation of the rubber hand (Tsakiris et al., 2007) and leads to a diminished animation of the real hand (Moseley et al., 2008). With respect to the homeostatic state of the body and interoceptive awareness, the insula predominately in the right hemisphere is key for establishing and maintaining the sense of body ownership (Craig, 2009, Craig, 2011, Critchley et al., 2004, Karnath and Baier, 2010, Moseley et al., 2012).

Here, we investigated the neuronal response to different types of body-related pictures showing either one's own desired body state or one's own undesired body state. Similar paradigms as the one used here have already been applied to anorexia nervosa patients and transgender individuals (Feusner et al., 2016, Fladung et al., 2010). For that purpose, to our knowledge for the first time, individual stimuli comprising photos of one's own body (and a control body as well) in different positions were constructed and edited according to the individual amputation desire of the xenomelia subjects. We compared 12 xenomelia individuals with 11 male controls with respect to whole-brain functional activity when viewing these body-related pictures. Activity was measured with fMRI in response to six different categories of body-related pictures: the real own body/a control body, the own body virtually amputated/a control body virtually amputated, and the own body virtually amputated with prosthesis/control body virtually amputated with prosthesis. Subsequently to the fMRI session, these pictures were also rated with respect to pleasantness, intensity, attractiveness and sexual arousal.

We hypothesized that viewing the desired state of one's own body, i.e. that of an amputee in xenomelia and that of a non-amputated body in controls, evokes differential brain activity compared with brain activity when viewing one's own undesired body state, i.e. a non-amputated body in xenomelia and an amputated body in controls. Differential brain activity is expected to be present mainly in brain regions associated with the construction and maintenance of a coherent body image such as the parietal cortices (SPL and IPL), but also in brain regions housing limb representations such as the primary and secondary somatosensory cortex, basal ganglia, thalamus, premotor cortex, supplementary motor area as well as the insular cortex that is key for the homeostatic state of the body and interoceptive awareness. It has already been shown that all these brain regions are structurally and/or functionally conspicuous in xenomelia subjects as reviewed elsewhere (Brugger et al., 2016).

In addition to the brain regions associated with body representations, we also expected differential brain activity in limbic regions associated with emotional processing due to the fact that viewing the desired or undesired body state also elicits differential affective states. With respect to the direction of effects and based on the small corpus of the functional neuroimaging literature in this condition, we expected increased as well as reduced activity in xenomelia compared with healthy male controls (Hänggi et al., 2017, McGeoch et al., 2011, van Dijk et al., 2013).

Section snippets

Subjects

Thirteen male individuals with xenomelia were recruited via Internet. One subject quit the fMRI session and withdrew from the study and was therefore excluded. The mean age of the remaining 12 individuals was 46.5 years (standard deviation (SD) = 11.6 years, range 32–68 years) and eight of them achieved a university degree. Nine individuals whished an amputation of the left leg (at the height of the thigh), one of the right leg, and the remaining two of both legs. Seven individuals were

Demographic, psychiatric and psychological measures

There were no significant differences between groups with respect to age (Mann-Whitney-U-test, z = −1.17, p = 0.26). The distributions of their educational levels (8 subjects with a university degree in each group) and handedness (all right handers) were comparable between groups. Psychiatric and psychological measures were assessed in xenomelia individuals only and compared with normative population values (Franke, 2000, Franke, 2002) (Table 1).

There were no significant deviations from

Discussion

The present study provides first evidence for a dissociation in functional brain activations between xenomelia subjects and controls in response to individual pictures representing their desired or undesired body states. In particular, the pictures of the desired amputated body of xenomelia individuals elicited differential neuronal activations capable of classifying these sufferers from healthy control subjects based solely on neuronal activations. This divergent pattern of brain activation

Acknowledgments

We thank all subjects for their participation, and are grateful to those suffering from the condition under investigation for their willingness to contribute to research. We also acknowledge Patrick Weigand for his support during experimentation.

Conflict of interest

All authors declare that there are no competing interests.

Funding

This study was partially supported by a grant (no grant number available) from Asklepios Proresearch (https://www.asklepios.com/proresearch/).

References (53)

  • S. Lemm et al.

    Introduction to machine learning for brain imaging

    Neuroimage

    (2011)
  • G. Macauda et al.

    When less is more – implicit preference for incomplete bodies in xenomelia

    J. Psychiatr. Res.

    (2017)
  • Y. Miyake et al.

    Brain activation during the perception of distorted body images in eating disorders

    Psychiatry Res.

    (2010)
  • G.L. Moseley et al.

    Bodily illusions in health and disease: physiological and clinical perspectives and the concept of a cortical 'body matrix'

    Neurosci. Biobehav. Rev.

    (2012)
  • F. Pereira et al.

    Machine learning classifiers and fMRI: a tutorial overview

    Neuroimage

    (2009)
  • P. Sachdev et al.

    Brains of anorexia nervosa patients process self-images differently from non-self-images: an fMRI study

    Neuropsychologia

    (2008)
  • N. Tzourio-Mazoyer et al.

    Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain

    Neuroimage

    (2002)
  • R. Uher et al.

    Functional neuroanatomy of body shape perception in healthy and eating-disordered women

    Biol. Psychiatry

    (2005)
  • T.F. Beek et al.

    Gender incongruence/gender dysphoria and its classification history

    Int Rev. Psychiatry

    (2016)
  • G. Berlucchi et al.

    The body in the brain revisited

    Exp. Brain Res

    (2010)
  • A. Berti

    This limb is mine but I do not want it: from anatomy to body ownership

    Brain

    (2013)
  • O. Blanke et al.

    Preliminary evidence for a fronto-parietal dysfunction in able-bodied participants with a desire for limb amputation

    J. Neuropsychol.

    (2009)
  • R.M. Blom et al.

    Body integrity identity disorder

    PLoS One

    (2012)
  • R.M. Blom et al.

    The desire for amputation or paralyzation: evidence for structural brain anomalies in body integrity identity disorder (BIID)

    PLoS One

    (2016)
  • G. Bottini et al.

    Is the desire for amputation related to disturbed emotion processing? A multiple case study analysis in BIID

    Neurocase

    (2015)
  • D. Brang et al.

    Apotemnophilia: a neurological disorder

    Neuroreport

    (2008)
  • Cited by (17)

    • Is this my foot? Experimentally induced disownership in individuals with body integrity dysphoria

      2022, Consciousness and Cognition
      Citation Excerpt :

      Structural and functional alterations were found in a distributed cortical body representation network spanning the right superior parietal lobule, primary and secondary somatosensory cortices, supplementary motor area, premotor cortex, insula, and various subcortical areas (Blom et al., 2016; Hänggi et al., 2017; Saetta et al., 2020). Furthermore, differential neural activity the in above-mentioned network was found both in response to touch to their desired as compared to their undesired body parts (van Dijk et al., 2013; McGeoch et al., 2011), as well as in response to seeing their own body as amputated versus complete (Oddo-Sommerfeld et al., 2018). Along these lines, several behavioural investigations showed differential effects between the affected and unaffected limb in body-related unisensory and/or multisensory processing by showing, for example, altered temporo-spatial processing of tactile stimuli (Aoyama et al., 2012), or altered reactions to pain, pain anticipation (Brang et al., 2008; Romano et al., 2015), and top-down modulation of pain.

    • Not my body, not my pain? Pain perception and placebo analgesia in individuals with body integrity dysphoria

      2022, Cortex
      Citation Excerpt :

      Participants with BID were invited to a full day of testing, which included several unrelated tasks both prior and after the pain perception and placebo analgesia task. Although the sample size of 18 participants with BID could be considered small, and was not based on a formal power calculation, it is nevertheless a comparatively large sample in comparison to other studies (e.g., n = 5 (Aoyama et al., 2012), n = 7 (Bottini et al., 2015), n = 10 (Romano et al., 2015), n = 9 (Lenggenhager et al., 2015), n = 13 (Hänggi et al., 2016; Hänggi et al., 2016; Hänggi et al., 2017; Oddo-Sommerfeld et al., 2018), n = 15 (Oddo et al., 2014), n = 16 (Saetta et al., 2020)) when considering the rarity and secrecy of the disorder. The sample consisted of a homogeneous group of patients with a uniform unilateral amputation desire of either the left or right leg.

    • White matter abnormalities in the amputation variant of body integrity dysphoria

      2022, Cortex
      Citation Excerpt :

      Moreover, the left orbito-frontal cortex (lOrbFG) as a key hub involved in both limbic and reward systems showed reduced intrinsic functional connectivity (Saetta et al., 2020). Other brain regions functionally altered in BID are the left superior and middle temporal gyri (lTG, Oddo-Sommerfeld et al., 2018; Saetta et al., 2020). These areas have often been linked to the mirror network.

    • Neural Correlates of Body Integrity Dysphoria

      2020, Current Biology
      Citation Excerpt :

      Pretending behavior might thus correspond to acting out a possibly innate, individually shaped body image. Oddo et al. [24] recently presented BID individuals and controls with morphs depicting either the observer’s actual or amputated body. In that study, multivariate statistics using machine learning were able to discriminate between a BID group and a control group based on the BOLD signal in the rSPL.

    View all citing articles on Scopus
    1

    Shared first authorship.

    View full text