The influence of age on the rubber hand illusion

Highlights

• The proportion of subjects responding to the RHI is higher in younger individuals.

• RHI onset time is shorter in younger individuals.

• Subjective responses to a RHI questionnaire reflect objective results.

• Subjects’ age should be carefully taken into account in experiments on the RHI.

Abstract

The rubber hand illusion is a perceptual illusion whereby a model hand is embodied during tactile stimulation. The aim of the present study is to investigate the onset time of the illusion in relation to age. We used two sensors, made using Arduino NANO, in order to record the onset time in which the participants said to start perceiving the illusion. The subjects involved in the experiment were divided in 3 age groups: 16–20, 21–50, and 51–88 years. We observed a clear-cut stronger manifestation of the illusion in the younger group both in terms of proportion of responders vs. non-responders, which was higher in the first age group, and in terms of illusion onset time which was shorter in the same group. These results were substantiated by the responses to a questionnaire measuring subjective perception of the illusion.

Introduction

How do we manage to possess the sense of ownership of our limbs and of our whole body? First published in Nature by Matthew Botvinick and Jonathan Cohen (Botvinick & Cohen, 1998), the rubber hand illusion (RHI) represents a useful tool that allows to clarify which processes underlie the sense of body ownership. In fact, the article develops the idea that it is possible to transfer the sense of ownership of a limb, and the consequent tactile perception that can be referred to it, in a foreign body. This occurs as a consequence of a triadic interaction between vision, touch and proprioception (Tsakiris & Haggard, 2005). During the experience, in fact, we act in such a way as to initially put into conflict these different sensory modalities. However, precisely because normally there is a correlation between them, there will be a simulated reconciliation of visual and tactile inputs, in such a way that the sensation of touch will gradually be referred to the point where the rubber hand is located, giving rise to the illusion (Maravita, Spence, & Driver, 2003). The literature in this field has recently greatly expanded and includes experimental designs that, starting from the main paradigm of Botvinick and Cohen, are going to investigate more extensively the sense of ownership not only of a single foreign limb, as in the rubber hand, but to an entire artificial body, like a mannequin (Petkova and Ehrsson, 2008, Petkova et al., 2011) or a doll (Van Der Hoort et al., 2011, Van Der Hoort and Ehrsson, 2014) or even an entire real body (Guterstam et al., 2015, Petkova and Ehrsson, 2008, Petkova and Ehrsson, 2010, Preston and Ehrsson, 2016, Preston and Ehrsson, 2018) belonging to another person (body swap illusion). The use of virtual reality techniques has in fact greatly extended the field of application of this experimental paradigm, opening up new interesting possibilities and ensuring the expansion of knowledge of body awareness.

So, many variations of the original experiment have been made in order to carry out a controlled manipulation of the possession of one's own body. However, among the variables manipulated, it remains unclear how the age factor can affect the RHI and, more generally, the sense of body ownership. Although age-related changes seem to be scarcely investigated in the field of the RHI, they were often examined in several studies related to multisensory integration processes.

The findings of these studies showed that the multisensory processing is somehow altered in aging, more specifically, it seems that the multisensory integration is enhanced in older individuals, who would present a wider temporal binding window (TBW; Diederich et al., 2008, Laurienti et al., 2006, Peiffer et al., 2007).

Laurienti et al. (2006) have evaluated the speed of discrimination in response to visual, auditory and combined visual-auditory (multisensory) stimuli in young and elderly participants. It turned out that older individuals were faster and had shorter reaction times in the multisensory task compared to younger participants, although in general both groups had shorter reaction time in the multisensory task than in the unisensory one. Similarly, Peiffer et al. (2007) found improvement in performance and response speed in the multisensory tasks (also visual-auditory paradigm) in older compared to younger participants. Finally, Laurienti et al. (2006) results also showed a broader integration window in the elderly compared to young adults. The same results are also confirmed by more recent studies, which demonstrate that multisensory integration changes throughout lifespan (Basharat et al., 2018, Setti et al., 2011, Mahoney et al., 2011).

At this regard, some RHI studies on children have found a greater proprioceptive drift towards the rubber hand in very young children (for example from 4 to 9 years), which would then join results similar to adults from 10 years onwards. This finding showed that the more the self-awareness increases, the more the proprioceptive drift decreases. The increase of proprioceptive drift is in fact the sign of a strong illusory perception (Cowie et al., 2013, Cowie et al., 2015). This would confirm that the perception of one's own body necessitates of a long period of development, such that certain multisensory connections and certain aspects of self-perception of the body may not be fully developed in childhood and have a broad and slow growth, which protracts in adulthood (Cowie et al., 2013).

As previously stated, it is known that RHI depends to a certain extent on multisensory integration (between vision, touch and proprioception), but the relationships between the temporal window of multisensory processes and the temporal window related to the RHI have not been thoroughly investigated yet. Only a recent study by Costantini et al. (2016) considered this issue and confirmed that the RHI time constraints reflect the individual temporal resolution in multisensory perception. To do this, the Authors measured the TBW of the participants in a simultaneity judgment task, where they used visual and tactile stimuli. Afterwards, Authors also measured the susceptibility to the RHI both in synchronous and asynchronous conditions, discovering that the wider the individual TBW is, the higher is the tolerated level of asynchrony. In fact, the individual TBW finding in the multisensory task was used to establish the asynchrony between the visual stimuli provided on the rubber hand and the tactile stimuli delivered on the real hand. This allowed to relate the individual temporal resolution in the multisensory perception with the temporal features related to the RHI.

However, recent studies taking in account age-related changes in the RHI found discordant results and only one case confirmed the results previously discussed in the literature on multisensory integration processes (Marotta, Zampini, Tinazzi, & Fiorio, 2018). In Kállai et al. (2017) for example, participants are recruited in two main age groups: from 20 to 41, identified as Junior Group, and from 56 to 72, identified as Senior Group. Authors found that, compared to younger subjects, older subjects have a less vivid experience of the illusion. However, this result seems to be in contrast with another recent study (Palomo et al., 2018) that on the contrary, dividing its participants into three age groups (20–35; 36–60; 61–80), found no significant differences between them. On the basis of the discussed literature on multisensory processes, one would thus expect from these studies that the RHI is more vivid and experienced earlier by older individuals, who have a larger TBW, however the cited discordant results do not always confirm this expectation. No one of these studies take into account the measurement of the temporal features of the RHI.

To measure the illusion, different methods have been used over time, including ad hoc questionnaires aimed at investigating different aspects of the administered test, and the proprioceptive drift (Botvinick & Cohen, 1998). A further method uses the galvanic response, which is however possibly influenced by some threat for the fake limb (Armel and Ramachandran, 2003, Tsuji et al., 2013). Few, however, are the examples in the literature that deal with the onset time of the RHI and the relative methodologies with which such information is obtained. In Ehrsson, Spence, and Passingham (2004), for example, subjects were instructed to press a key with their foot as soon as they had the impression of feeling the sense of ownership referred to the rubber hand. It turned out that, on average, the illusion begins to be felt around 11 s after the start of visuotactile stimulation. In Lloyd (2007) a mean onset time of 6.52 s was instead observed. A more recent study carried out by Kalckert and Ehrsson (2017) evaluated instead the onset time in the moving rubber hand illusion, distinguishing between active movement made by the subject and passive movement induced by the experimenter. They found that the reaction time was actually faster in the active condition (20.9 s compared to 23.7 s of the passive condition).

For the measurement of the onset time we introduce here a semi-automated set-up made using Arduino NANO, in addition to the subsequent use of the classic questionnaire. The principle of the registration method of the onset time is similar to Ehrsson (2004), but in our case the procedure is handled by the experimenter.

Due to the lack of studies focusing on the relation between age and temporal dimension (onset time) of the RHI, the purpose of the present research is to investigate if these parameters interact, evaluating possible effects across wide age ranges. In our hypothesis, if the temporal dimension related to the RHI reflects the temporal dimension of all multisensory processes, we should be able to expect a more rapid response to the illusion in older individuals.