Elsevier

Neuropsychologia

Volume 48, Issue 1, January 2010, Pages 26-37
Neuropsychologia

Touch and feel? Using the rubber hand paradigm to investigate self-touch enhancement in right-hemisphere stroke patients

https://doi.org/10.1016/j.neuropsychologia.2009.08.006Get rights and content

Abstract

Following stroke, a patient may fail to report touch administered by another person but claim that she feels touch when it is self-administered. We investigated three explanations for self-touch enhancement: (1) proprioceptive information from the administering hand, (2) attentional modulation, and (3) temporal expectation. Tactile sensation was assessed with vision precluded, and with the affected hand positioned in the left and right hemispace. In four of six experiments, the somatic rubber hand paradigm was used: the Examiner administered stimulation to the patient's affected left hand while guiding the patient's right hand to administer synchronous stimulation to a prosthetic hand. Even though the patient's two hands were not in contact, patients detected the same number of stimulations as when they touched their own hand directly (self-administered condition). Moreover, there was no decline in rates of detection when potentially informative movements of the administering hand were restricted. This demonstrates that patients feel rather than infer stimulation under conditions of self-touch. When patients received stimulation to the affected hand in the opposite hemispace to the hand administering touch to the prosthetic hand, all but one showed self-touch enhancement. Thus, neither proprioceptive information nor attentional modulation at the spatial region of the administering hand provided a sufficient explanation for self-touch enhancement. A follow-up experiment indicated an important role for temporal expectation: a delay, between the patient's stimulation of the prosthetic hand and the Examiner's stimulation of the patient's affected hand, eliminated the self-touch enhancement effect.

Introduction

Weiskrantz and Zhang (1987) were the first to report self-touch enhancement in a right-hemisphere stroke patient. On conventional sensory testing, the patient was unable to detect force of less than 35 g (filament of 0.6 mm diameter) administered to her left hand: healthy individuals detect intensities less than .07 g. Despite clear sensory deficits, the patient “quietly and insistently maintained that she had feeling in her left hand when it was touched by her own right hand” (p. 632). Following a systematic assessment, Weiskrantz and Zhang confirmed that the patient enjoyed an “impressive degree of residual cutaneous sensitivity” (p. 634) when she was actively involved in the administration of touch.

Valentini, Kischka, and Halligan (2008) have recently assessed self-touch enhancement in 39 patients (22 with right-hemisphere damage and 17 with left-hemisphere damage), who were between 20 and 3308 days post-stroke. Seventeen right-hemisphere patients and five left-hemisphere patients demonstrated self-touch enhancement, operationally defined as (a) an improvement in stimulus detection, (b) an improvement in stimulus localisation, or (c) subjectively higher ratings of stimulus intensity under conditions of self-administered touch. The laterality bias (i.e., the higher rate of self-touch enhancement among patients with right-hemisphere damage) cannot be explained by concomitant visuospatial neglect. Self-touch enhancement was exhibited by patients with (n = 5) and without (n = 17) visuospatial neglect. Valentini et al.’s (2008) systematic research programme confirms that self-touch enhancement is a pervasive phenomenon, particularly among patients with right-hemisphere damage. However, the precise mechanisms underlying self-touch enhancement are yet to be established. Here we consider explanations in terms of (1) the patient using proprioceptive information from the administering hand to infer touch, and (2) the patient's administering hand directing attention to the affected body part.

Under conditions of self-touch, a patient may use proprioceptive information from the administering hand to infer stimulation. By judging the position of her administering hand relative to the affected body part, the patient may infer that she is being touched and she may report stimulation based on knowledge rather than feeling. The pioneering studies on self-touch enhancement took important steps to restrict the patient's use of proprioceptive information. To eliminate cues based on skin-to-skin contact, the patient used an instrument to administer stimulation (Valentini et al., 2008, Weiskrantz and Zhang, 1987); to make localisation more difficult, the Examiner moved the patient's administering hand between trials (Valentini et al., 2008). Sham trials provided an additional measure for ensuring the reliability of patient responses (Valentini et al., 2008, Weiskrantz and Zhang, 1987). On sham trials, the Examiner guided the patient to administer stimulation to one of the Examiner's fingers, which were inter-digited with the fingers of the patient's affected hand. If the patient reported touch only on valid trials (and not on sham trials) this provided an indication that the patient was feeling touch when she administered stimulation to her own hand. But, it is also possible that the patient may have used subtle differences in the location of her administering hand to distinguish valid from sham stimulations, given that she was guided to a slightly different location on these two trial types.

Each of the measures introduced by Valentini et al. (2008) and Weiskrantz and Zhang (1987) reduces proprioceptive information. To discount completely the explanation of proprioceptive information playing a role in self-touch enhancement, we use the somatic rubber hand paradigm (Ehrsson, Holmes, & Passingham, 2005). In the somatic rubber hand paradigm, the Examiner administers stimulation to the patient's affected hand while guiding the patient's unaffected hand to administer synchronous stimulation to a prosthetic hand. Using this paradigm, neurologically healthy individuals have been shown to experience the powerful illusion of self-touch, even though the two hands are separated by 15 cm (Ehrsson et al., 2005). This illusion is driven by multisensory correlations between proprioceptive and tactile inputs (Ehrsson et al., 2005)—the participant administers stimulation (to the prosthetic hand) and receives corresponding stimulation on her passive hand. We can use this paradigm to investigate possible explanations for self-touch enhancement, by holding constant the conditions of direct self-touch (i.e., the action of the patient's administering hand, the stimulation to the patient's affected hand, and the temporal correspondence between these two events) but removing the proprioceptive correspondence between the patient's administering and receiving hand. An explanation for self-touch enhancement based on the patient using proprioceptive knowledge to infer touch is eliminated. The patient can no longer infer stimulation based on the relative position of her two hands because she is now administering and receiving touch in two distinct spatial locations, and this is true regardless of whether the patient has the subjective impression that she is touching her own hand.1 Moreover, on sham trials, no stimulation is administered to the patient's affected hand (by the Examiner) but the patient is guided to administer stimulation to the prosthetic hand (as she is on valid stimulation trials). Because the location of the patient's administering hand is identical on valid and sham trials, she cannot use subtle differences in hand position to distinguish these two trial types. Consequently, our first prediction: if the explanation for self-touch enhancement is that the patient uses relative hand position to determine whether she has been touched, detection will decline when the patient is no longer able to use hand position to infer stimulation.

Even if the patient does not use relative hand position to decide whether she is being touched, she may use subtle movements of her administering hand to focus attention. For example, if the patient detects a stimulation to her middle finger, and the next movement of the administering hand is leftward, she may use this information to focus attention now on the ring and little finger. But these potentially informative movements of the administering hand can be held to a minimum by guiding the patient to stimulate the same location on all trials (the back of the prosthetic hand), while each of her five digits continues to receive stimulation from the Examiner. Our second prediction: if the patient uses subtle movements of the administering hand to focus attention, detection should decline when movements of the hand are restricted. Alternatively, if the patient uses subtle movements to localise felt sensations, localisation (but not detection) should decline when movements of the hand are restricted.

Valentini et al. (2008) have proposed that self-touch enhancement may be due to attentional modulation. Impetus for this proposal comes from research conducted by Coslett and Lie (2004) demonstrating that the unaffected hand serves as an “attentional wand” which directs processing resources to the affected side of the body. Coslett and Lie (2004) present two patients with sensory deficits following right-hemisphere stroke. The first patient exhibited sensory extinction, failing to report touch to the left hand under conditions of bilateral stimulation. The second patient exhibited severe left-side sensory deficits affecting unilateral touch, proprioception and temperature discrimination. In both patients, report of Examiner-administered stimuli on the left hand improved when the patients’ right and left hands were in contact during stimulation. This enhancement was proposed to result from attentional modulation, driven by the behavioural salience of the unaffected hand (Coslett & Lie, 2004).

Reed, Garza, and Roberts (2007) review the role of the body in spatial attention. The researchers observe an important role for the hands in the allocation of attention, noting that this role depends on whether the hand is static (e.g., the unaffected hand in Coslett & Lie, 2004) or in action (e.g., the unaffected hand in Valentini et al., 2008, Weiskrantz and Zhang, 1987). When the hand is static, the “region near the hand may be prioritized so that the potential relevance of cues and targets appearing in that space is increased” (p. 47). By contrast, when the hand is moving, attention shifts to the “functional spatial range of the action” (p. 51) rather than the hand itself. Given this distinction, it is possible that the sensory enhancement exhibited by Coslett and Lie's patients (2004) is driven by a different attentional process to that exhibited by the patients with self-touch enhancement in the studies of Weiskrantz and Zhang (1987) and Valentini et al. (2008).

It may be possible to shed further light on the attentional modulation theory using the somatic rubber hand paradigm. If the prosthetic hand were positioned in the opposite hemispace to the patient's affected hand, the patient would administer touch (to the prosthetic hand) in one hemispace and receive touch (on the affected hand) in the opposite hemispace. In this condition, the patient's affected hand would not fall “within the functional spatial range” (Reed et al., 2007, p. 57) of the administering hand action. Our third prediction: if self-touch enhancement occurs because attention is enhanced in the region of space in which a motor programme unfolds (see Rizzolatti et al., 1987, Umilta, 2000), detection should decline when the patient administers and receives stimulation in opposite sides of space.

Experiments 1 and 2 provide baseline rates for Examiner- and self-administered stimulation. Following on from these experiments, the somatic rubber hand paradigm is used to investigate an explanation for self-touch enhancement based on proprioceptive information from the administering hand. In Experiment 3, the patient administers stimulation to the digit of the prosthetic hand corresponding to the digit of her own hand receiving stimulation from the Examiner. In this experiment, the patient cannot use the relative position of her two hands to infer stimulation. In Experiment 4, the patient administers stimulation to the same location on the back of the prosthetic hand on all trials. In this experiment, potentially informative movements of the patient's administering hand are held to a minimum. In Experiment 5, the somatic rubber hand paradigm is used to investigate an explanation for self-touch enhancement based on attentional modulation. The prosthetic hand and the patient's affected hand are positioned so that the patient administers and receives stimulation in distinct spatial regions, one hand in each hemispace. In this experiment, the administering hand directs the patient's attention away from the affected hand.

Section snippets

Patient recruitment

Five right-hemisphere stroke patients were examined as Valentini et al. (2008) have shown that self-touch enhancement occurs more frequently among patients with right-hemisphere damage. Three patients (Patients NG, SM, SK) were recruited from the inpatient rehabilitation ward at the Oxford Centre for Enablement, and two patients (Patients CJ, CA) from the outpatient services of the Oxford Centre for Enablement. Patients CJ and CA were chosen specifically because they were attending therapy

Experiment 1: Examiner administers stimulation to the patient's hand

Experiment 1 was used to establish each patient's baseline performance for stimulation administered by the Examiner.

Experiment 2: Patient administers stimulation to the patient's hand

Experiment 2 was designed to replicate the self-touch enhancement effect (Valentini et al., 2008, Weiskrantz and Zhang, 1987). Positioning of the affected hand was not described in the previous studies. We assess sensation with the affected hand positioned first in the left hemispace and then in the right hemispace. Studies investigating somatosensory extinction have shown that detection of stimuli administered to a patient's left hand may be enhanced when the left hand is positioned to the

Experiment 3: Examiner administers stimulation to the patient's hand while the patient administers stimulation to the prosthetic hand—Somatic rubber hand paradigm

In Experiment 3, the somatic rubber hand paradigm (Ehrsson et al., 2005) was used to investigate whether the self-touch enhancement effect is maintained when the patient is unable to use knowledge about the relative position of each of the hands to infer stimulation.

Experiment 4: Examiner administers stimulation to the patient's hand while the patient administers stimulation to the prosthetic hand—Somatic rubber hand paradigm with localisation manipulation

Experiment 4 was designed to control for the possibility that movements of the administering hand are used as an aid to stimulus detection and localisation under conditions of self-administered touch. In Experiment 3, the patient administered touch to the digit of the prosthetic hand which corresponded to the digit of his or her hand receiving stimulation from the Examiner. Even though the patient could not use the relative position of each hand to judge whether the affected hand was being

Experiment 5: Examiner administers stimulation to the patient's hand while the patient administers stimulation to the prosthetic hand—Somatic rubber hand paradigm with hemispace manipulation

In Experiment 5, the prosthetic hand and the patient's affected hand were positioned so that the patient administered and received stimulation in distinct spatial regions, one hand in each hemispace. It has been argued that under conditions of self-administered touch, the hand administering the stimulation (i.e., the patient's unaffected hand) may direct the patient's attention to the affected hand (Valentini et al., 2008). According to this proposal, enhanced sensory performance is a product

Interim discussion

The results from Experiments 1–5 confirm that patients with impaired sensation are better at detecting sensory stimulation when they are involved in administration. We assessed four patients with impaired sensation and found that, in these patients, enhanced sensation under conditions of self-touch was not due to patients using proprioceptive information from the administering hand to infer touch (Experiment 3). Furthermore, enhanced detection of stimulation did not depend on patients using

Procedure

In Experiment 6, the Examiner administered stimulation to Patient NG's affected left hand while her right hand was guided by the Examiner to administer stimulation to the prosthetic hand. A 1-s delay was introduced: the Examiner guided the patient to administer stimulation to the prosthetic hand before the patient's affected hand was stimulated by the Examiner. The patient's affected hand and the prosthetic hand were positioned in the same side of space (as in Experiments 3 and 4). Sensation

General discussion

A stroke patient with impaired sensation may fail to detect stimulation when it is administered by another person but may detect identical stimulation when it is self-administered. We have used a novel method – the somatic rubber hand paradigm (Ehrsson et al., 2005) – to investigate self-touch enhancement. The Examiner administers stimulation to the patient's affected hand, while guiding the patient's unaffected hand to administer stimulation to a prosthetic hand. Neurologically healthy

Conclusion

The results of the current experimental series confirm that patients with impaired sensation may experience enhanced sensory perception under conditions of self-administered touch. By using the somatic rubber hand paradigm, we were able to demonstrate that the patient feels the stimulation under conditions of self-administered touch. The patient does not use the relative position of her two hands to infer stimulation, nor does she use subtle movements of the administering hand to focus

Acknowledgments

The authors would like to thank the patients who took part in this study and the staff at the Oxford Centre for Enablement for supporting this research. Gratitude is extended to Professor Derick Wade for invaluable discussions, and to two anonymous reviewers for helpful comments on the manuscript.

References (34)

  • S. Aglioti et al.

    Frames of reference for mapping tactile stimuli in brain-damaged patients

    Journal of Cognitive Neuroscience

    (1999)
  • C. Collin et al.

    Assessing motor impairment after stroke: A pilot reliability study

    Journal of Neurology, Neurosurgery, and Psychiatry

    (1990)
  • J.T. Coull et al.

    Where and when to pay attention: The neural systems for directing attention to spatial locations and to time intervals as revealed by both PET and fMRI

    Journal of Neuroscience

    (1998)
  • G. Demeurisse et al.

    Motor evaluation in vascular hemiplegia

    European Neurology

    (1980)
  • H.H. Ehrsson et al.

    Touching a rubber hand: Feeling of body ownership is associated with activity in multisensory brain areas

    Journal of Neuroscience

    (2005)
  • F. Goodenough

    A new approach to the measurement of intelligence of young children

    Journal of Genetic Psychology

    (1926)
  • S.J. Goodrich et al.

    Anti-extinction following unilateral parietal damage

    Cognitive Neuropsychology

    (1997)
  • Cited by (12)

    • EEG changes caused by spontaneous facial self-touch may represent emotion regulating processes and working memory maintenance

      2014, Brain Research
      Citation Excerpt :

      The term “spontaneous” means these kinds of movements, particularly the triggering of the movements, is not based on deliberate decision-making processes of the respecting person. Contrary to other self-touching actions in experimental settings (Ackerley et al., 2012; Guest et al., 2009; Schutz-Bosbach et al., 2009; van Stralen et al., 2011; White et al., 2010) spontaneous STG are performed without a request by any other person and without any prompt by an external stimulus. These spontaneous movements are fairly common in everyday life and may be observed as a ubiquitous behaviour in people from all social classes.

    • Cooling the thermal grill illusion through self-touch

      2010, Current Biology
      Citation Excerpt :

      Pressing the two hands against each other not only provides additional proprioceptive, thermal, and tactile input but also provides strongly correlated sensory input across different body parts. These correlations enhance sensory processing [16, 17] and strengthen the association between the touching body parts, thus creating a coherent body representation [4]. Therefore, we next combined the afferent inputs of the TGI with the distinctive experience of self-touch, to increase multisensory integration and coherence of body representation (Figure 1).

    • The touched self: Affective touch and body awareness in health and disease

      2016, Affective Touch and the Neurophysiology of CT Afferents
    View all citing articles on Scopus
    View full text