The way we perceive bodily sensations depends on a number of environmental and contextual variables that can enhance or reduce perceptive acuity and alter somatosensation (Longo & Sadibolova, 2013
). One of the variables that has been found to influence somatosensory perception is the vision of the body. Some studies have shown that directing visual attention towards a body part enhances tactile perception at that stimulated body part, both when the visual information provides useful information about the tactile stimulation (Halligan, Hunt, Marshall & Wade, 1996
; Làdavas, Pellegrino, Farnè & Zeloni, 1998
, Làdavas, Farnè, Zeloni & di Pellegrino, 2000
) and when it does not (Serino, Farnè, Rinaldesi, Haggard & Làdavas, 2007
), that is, when the visual input is non-informative.
Indeed, non-informative vision of the body was found to enhance tactile perception in grating orientation tasks by decreasing discrimination thresholds and increasing discrimination accuracy (Taylor-Clarke, Kennett & Haggard, 2004
; Cardini, Longo & Haggard, 2011
). Alongside, non-informative vision of the body was also found to enhance tactile spatial acuity in terms of reduced two-point discrimination, enhanced amplitude discrimination of above-threshold stimuli and reduced tactile detection thresholds (Tipper et al., 1998
; Kennett, Taylor-Clarke & Haggard, 2001
; Serino, Padiglioni, Haggard & Làdavas, 2009
; Keizer, Smeets, Dijkerman, van Elburg & Postma, 2012
; Harris, Arabzadeh, Moore & Clifford, 2007
). In these cases, vision of the body has been thought to improve tactile perception by sharpening tactile receptive fields in the primary cortical somatosensory map (Haggard, Christakou, & Serino, 2007
Moreover, the vision of the body has been found to facilitate not only the perception of exteroceptive tactile stimuli but also the perception of internal body signals (interoception). Ainley, Tajadura‐Jiménez, Fotopoulou and Tsakiris (2012
) and Ainley and Tsakiris (2013
) demonstrated that non-informative vision of one’s own face can enhance the detection of heartbeat sensations (namely, interoceptive accuracy). Participants were asked to perform a heartbeat perception task (HPT; Schandry, 1981
) while watching a photograph of their face or their reflection in a mirror as compared to a blank screen. Results showed that participants were more accurate in perceiving their heartbeat whilst looking at their face, although the face was a body part unrelated to the perceptual task, and vision of the self was not providing informative data for the completion of the task per se. In contrast, Serino et al. (2007
) showed that the vision of a rubber foot does not enhance tactile acuity on the hand. However, it should be noted that interoception and exteroception are two distinct ways to experience the body that correspond to different processes and therefore they cannot always be compared.
Other research has shown, however, that vision of the body can either enhance or diminish perceptive acuity depending on the body part targeted (Tipper et al., 1998
: Serino et al., 2007
), the type of task used (Longo & Sadibolova, 2013
; Press, Taylor-Clarke, Kennett & Haggard, 2004
), and participants’ characteristics (Costantini, 2014
; Eshkevari et al., 2012
). In this respect, Harris et al. (2007
) proposed that non-informative vision of the body does not simply enhance somatosensory processing, but rather it induces adaptive changes in tactile sensitivity within a bimodal sensory system. According to this reasoning, after adaptation of tactile receptive fields subsequent to the vision of the body, detection of near-threshold stimuli is impaired, while the discrimination of stimuli is enhanced.
According to the previous paragraph, Tipper et al. (1998
), for example, suggested that familiarity of the stimulated body part can modulate the effects of vision on somatosensory processes. In their study, non-informative vision of a familiar body site (such as the face) was found to facilitate the detection of supra-threshold tactile pulses; in contrast, the vision of a less familiar body part (such as the back of the neck) was found to have little impact (neither facilitatory nor inhibitory) on supra-threshold tactile detection.
Moreover, it has been shown that the vision of the body can either enhance or diminish tactile acuity depending on the type of task used. Longo and Sadibolova (2013
), for example, found that vision can actively distort touch perception, rather than increase accuracy when participants are asked to estimate the size of a tactile stimulus. In their study, vision of the stimulated body part significantly reduced the perceived size of a tactile stimulus, as compared to vision of an object or of a non-stimulated body part.
Alongside, Press et al. (2004
) showed that the complexity of the task can play a role in determining the effects of vision on tactile perception. In their study, non-informative vision only enhanced tactile perception when the task was both difficult and involving a spatial component (making speeded responses in an at-threshold two-point discrimination task). Performance on an easier non-spatial discrimination task (detecting a brief gap in a 250 ms above-threshold vibration) was actually worse when participants viewed the targeted body part as compared to a neutral object.
Accordingly, Mirams, Poliakoff, Brown and Lloyd (2010
) found that non-informative vision of the hand increased errors on a non-spatial touch detection task—the Somatic Signal Detection Task (SSDT, Lloyd, Mason, Brown & Poliakoff, 2008
). The SSDT involves detecting near-threshold vibrations delivered on the fingertip, where on 50% of trials there is a simultaneous LED flashing next to the targeted finger. During this task, the presence of the light increases incorrect detection of a vibration when it did not actually occur (Lloyd et al., 2008
). Mirams et al. (2010
) used the SSDT to assess the effects of a second visual variable, that was a vision and no vision of the stimulated hand, on touch perception. Specifically, the study analyzed whether non-informative vision of the hand compared to no vision of the hand would reduce or increase incorrect reports of feeling touch during the task. According to previous research (Press et al., 2004
; Harris et al., 2007
), vision of the hand would induce a reduced tactile detection, with less reports of touch (i.e., reduce ‘hits’). However, the study showed that during the vision condition, participants were more inclined to make false reports of feeling the touch, and especially on trials when the light flash occurred. The authors suggested that vision of the hand may have raised the focus on interoceptive information to a detrimental degree that led participants to misinterpret internal signals as external touch, which resulted in more errors during the SSDT. This led to the conclusion that non-informative vision of the body may lead to higher somatic interference and ultimately to a less accurate discrimination of touch during the SSDT.
Another variable that may influence the way the vision of the body impacts on perception refers to individual differences and personality characteristics. A fundamental assumption of cognitive approaches to personality and psychopathology is that individuals differ in their response to similar situations because of differences in the way they process incoming stimuli, in terms of both lower-level and higher-level information processing (Öhman, Lundqvist, & Esteves, 2001
; Mineka, Rafaeli, & Yovel, 2003
; Yovel, Revelle & Mineka, 2005
). Accordingly, psychiatric symptoms (i.e., schizophrenic, eating disorder and somatoform symptoms; Ferri et al., 2014
; Peled, Ritsner, Hirschmann, Geva & Modai, 2000
; Eshkevari et al., 2012
) and personality traits (i.e., emphatic abilities and the ability to describe personal experiences; Asai, Mao, Sugimori & Tanno, 2011
; Haans, Kaiser, Bouwhuis & IJsselsteijn, 2012
) have been related to individual differences in the way participants encode bodily related visual information (Costantini, 2014
Research in this context has placed a great focus specifically on Eating Disorders (EDs). EDs are a family of psychopathologies characterized by aberrant eating habits and rituals, fear of gaining weight, disturbances in body weight or shape perception, including unawareness of such perceptual disturbances (APA, 2013
Body perception in EDs has been linked to a greater focus on visual aspects of the body at the expense of other incoming information (Mehling et al., 2009
). On a phenomenological level, this heightened focus on visual bodily information manifests itself with excessive concerns and rumination about one’s own physical appearance and body image (Arciero & Guidano, 2000
). However, recent evidence suggests this shift of focus to be present also in the context of lower-level sensory processing, with a general over-investment on exteroceptive information coupled with a blunted perception of bodily information coming from within the body (interoceptive deficits). Simply put, ED patients have been deemed to have a preferential reliance on sensory data deriving from the outer world (exteroception) over interoceptive information (Mehling et al., 2009
; Arciero & Guidano, 2000
). In this regard, informative data come from studies analysing the integration of conflicting visual and internal somatic information about the body, for example using the Rubber Hand Illusion paradigm (RHI; Botvinick & Cohen, 1998
). During the RHI, an experimenter strokes a rubber hand placed in front of participants synchronously with their own hand, which is hidden from sight. The synchronous visuo-tactile stimulation induces participants to mislocate the position of their own hand as closer to the rubber hand, and especially ED patients who were found to be more inclined to perceive this illusion (Mussap & Salton, 2006
; Eshkevari et al., 2012
; Caglar-Nazali et al., 2014
). Therefore, these results provide experimental support to the hypothesis that people with EDs may have an increased sensitivity to the visual aspects of body perception (Eshkevari et al., 2012
According to these data, it could be argued that non-informative vision of the body may be encoded differently in participants presenting with high versus low ED symptoms and therefore have different effects on touch perception. To test this hypothesis, the current study investigated the effects of non–informative vision of the body on tactile perception in participants presenting with low and high ED symptoms. Tactile perception and visuo-tactile integration were analysed using the SSDT paradigm (Lloyd et al., 2008
). Replicating the design employed by Mirams et al. (2010
), participants underwent the SSDT in two experimental conditions: non-informative vision of the hand and no vision of the hand.
As the bodily self (that is the global, multimodal awareness of one’s own body; Blanke, 2012
) in EDs is mostly anchored to exteroceptive coordinates, we hypothesized that vision of the body, by increasing the focus on the bodily self, would lead high symptomatic participants to focus more on exteroceptive tactile information. Therefore, we expected high ED participants to be more sensitive to touch when performing the SSDT during the non-informative vision condition (i.e., make a higher number of ‘hits’ and a lower number of ‘false alarms’). Of relevance here is a previous study from our research group where we found that vision of one’s own face increased sensitivity to touch during the SSDT only in participants presenting with high ED but not low ED symptoms (Sacchetti, Mirams, McGlone & Cazzato, 2020
). Conversely, in low ED participants we expected to replicate the results found by Mirams et al. (2010
) with a vision of the body increasing false alarms of touch rather than increasing hits. Indeed, as suggested by the authors, in non-symptomatic participants (in the presence of a bodily self (Blanke, 2012
) anchored also to interoception), vision of the body is more likely to enhance the focus on interoceptive information, that are not relevant for the task, and are, therefore, erroneously misinterpreted as external touch.
Additionally, the study further investigated whether individual differences in self-reported interoceptive sensibility (EDI-3), body awareness (Body Perception Questionnaire, BPQ), dysmorphic concerns (Dysmorphic Concerns Questionnaire, DCQ) and body dissatisfaction (the body dissatisfaction subscale on the Eating Disorders Inventory-3) could explain participants responses (hits and false alarms) during the SSDT. Specifically, according to the somatic interference hypothesis (by Mirams et al., 2010
) that links false alarms to a misperception of interoceptive sensations, we expected participants self-reporting higher difficulties in recognizing inner bodily signals (as measured by the interoceptive sensibility-EDI-3 subscale and the BPQ) to report also higher false alarms during the SSDT. Moreover, we hypothesized that participants showing higher levels of concerns about their physical appearance (as measured by the body dissatisfaction-EDI-3 subscale and the DCQ) would be more strongly influenced by the vision of their body and therefore respond with higher hits in this condition.
The aim of this study was to investigate the effects of non-informative vision of the body on exteroceptive multisensory integration and touch perception in participants presenting with different levels of ED symptoms. Based on previous literature that linked EDs with a greater focus on exteroceptive bodily information (Mehling et al., 2009
; Arciero & Guidano, 2000
; Arciero et al., 2003
; Mazzola et al., 2014
), we expected high ED participants to be more sensitive to tactile stimuli during the SSDT while viewing their body (in the Vision condition). Conversely, in line with previous results found by Mirams et al. (2010
), we expected low ED participants to report a lower sensitivity to touch and higher false alarms in the Vision condition due to a higher level of somatic interference in this condition.
Supporting our expectations, high ED participants were better able to correctly detect the touch during the SSDT when their hand was visible as compared to when their hand was hidden from sight, having a significantly higher HR in the Vision compared to the No Vision condition. Moreover, in high ED participants there was an effect of the Light on FA only in the No vision condition but not in the Vision condition. Therefore, the presence of the light was found to induce false reports of touch only when their hand was hidden from sight. Conversely, when the hand was visible, high ED participants were found to be less affected by the influence of the light, ultimately leading to less FA and a more accurate perception of touch.
These results are in line with arguments that body perception in EDs is characterized by a differential processing of exteroceptive bodily information. Indeed, EDs have been described by the phenomenological psychology as having an “outward dispositional affective style”, which means that ED patients tend to anchor their bodily self to a greater extent to external bodily reference points in the service of visceral and internal somatic information (Arciero & Guidano, 2000
; Arciero et al., 2003
; Mazzola et al., 2014
). In other words, body perception in EDs has been described as showing an over-investment on sensory information deriving from the interactions with the outer world, such as exteroceptive visual and tactile information. Coherently, the vision of the body (as manipulated experimentally) is thought to exacerbate this dispositional perceptive style leading to a greater focus on information, such as touch in the context of the current study. Specifically, we suggest that in participants presenting with ED symptoms, the vision of their body increases attention only towards those dimensions of the bodily self that are already invested by a greater focus, that is exteroception. In turn, this shift of focus determines a greater accuracy in exteroception, and therefore also in detecting tactile stimuli. Interestingly, results of this study are in line with a previous study from our research group (Sacchetti et al., 2020
) in which vision of the face, instead of the hand, was found to increase correct detection of touch during the SSDT in high ED participants. Taken together the two studies suggest a consistent effect of vision of the body (across different body parts; i.e., the face and the hand) in enhancing touch detection in EDs.
Conversely, as the bodily self in EDs is not anchored to interoceptive information, the vision of the body may not enhance interoceptive sensibility. Supporting this theory, the high ED group self-reported more confusion and difficulties in recognizing and responding to internal bodily states, as indicated by higher scores in the Interoceptive Deficit scale of the EDI-3.
Furthermore, it should be noted that results in the overall sample showed a positive correlation between HR in Light trials of the Vision condition and the Body Dissatisfaction subscale of the EDI-3. This indicates that participants presenting with a greater dissatisfaction towards their body were also more accurate in perceiving touch when multiple visual information accompanied the stimulation (the presence of the light and the non-informative vision of the hand). These results may suggest that Body Dissatisfaction specifically, among the different subscales of the EDI-3 accounts for the fact that vision of the hand increased HR only in the high ED but not in the low ED group.
For the low ED group, non-informative vision of the body was found not to impact participants’ ability to correctly detect touch. Replicating previous results by Mirams et al. (2010
), low ED participants were found to report comparable HR in the Vision and in the No Vision conditions. Moreover, results showed that for the low ED group the presence of the light was more likely to induce false reports of touch only when they performed the task while their hand was visible. Therefore, conversely to the high ED group, for the low ED group, non-informative vision of the hand was found to reduce tactile accuracy by increasing FA. However, conversely to Mirams et al. (2010
), and against our expectations, there was no overall difference in FA between the Vision and the No Vision condition. Nevertheless, our results partially support the somatic interference hypothesis, according to which non-informative vision of the body may increase somatic interference arising from internal bodily signals that are mistaken for the external touch.
With this regard, we found a positive correlation between scores on the interoceptive deficit subscale of the EDI-3 and FAs, suggesting that difficulties in recognizing interoceptive information are associated with an increased tendency to erroneously report touch. This is also in line with previous findings that FA during the SSDT are associated with lower levels of interoceptive accuracy as assessed using a Heartbeat Perception Task (HPT; Durlik et al., 2014
). This explanation is also consistent with Lloyd et al. (2008
) attentional account of touch misperception during the SSDT, according to which attention to the body can increase somatic disturbances by raising awareness of subtle bodily sensations that are confused with external tactile stimuli. A similar process has been used to explain somatoform symptoms, that is physical symptoms experienced in the absence of any apparent physical abnormality (APA, 2013
). Different lines of research have linked somatoform symptoms to a heightened and maladaptive awareness of the body that causes an increased salience of benign bodily sensations that are then mistaken for evidence of serious illness (Mehling et al., 2009
; Brown et al., 2012
). Coherently, previous research on this topic has shown that participants experiencing higher levels of somatoform symptoms are also more inclined to report false sensation of touch during the SSDT (Brown et al., 2010
), possibly due to a hypervigilance towards inner body signals.
Overall, in Mirams et al. (2010
), and in the low ED group of the current study, it might be possible that vision of the body amplified and distorted the focus on interoceptive sensations, which were therefore mistakenly confused for exteroceptive signals. This in turn would lead to greater false reports of touch. Interestingly, this effect was specific for the low ED group, and it was not found for the high ED group, where the perception of interoceptive information has been characterized as blunted in EDs (Mehling et al., 2009
In conclusion, the results of this study indicate that non-informative vision of the body can have different effects on touch perception depending on participants’ level of ED symptoms. Previous research has shown that non-informative vision of the body can either enhance or be detrimental to touch perception depending on the type of task that participants are required to performed, and the familiarity of the body part stimulated (Longo & Sadibolova, 2013
; Tipper et al., 1998
). However, our results indicate that top-down mechanisms involving participants’ relationship with their body (and specifically body dissatisfaction) can also play a role in determining the effects of non-informative vision of the body on touch perception. Specifically, whereas vision of the body was found to increase the correct detection of touch in participants presenting with high ED symptoms; in participants presenting with low ED symptoms, vision of the body was found to diminish tactile accuracy, by increasing the effect of the Light on false alarms, possibly due to a higher somatic interference of internal body sensations.
Results of this study not only inform current phenomenological models on body perception, but also suggest indications for the clinical field. Indeed, body misperception is still scarcely addressed by different national and international guidelines for the treatment of EDs (Cuzzolaro & Fassino, 2018
). It is possible that recovery from EDs could benefit from a partial shift of focus from external to internal bodily information, therefore leading to a rebalance between exteroception and interoception.