Loneliness is related to a person’s social bonds or the quality of an individual’s social network (Hawkley et al. 2008). Human social relationships, particularly the closest most intimate ones, have a very tactile side to them (Cascio et al. 2018; Dunbar 2010). However, the link between social touch and loneliness has seldom been addressed. In scientific research, proxemics studies (Hall 1969; Sorokowska et al. 2017; Sussman and Rosenfeld 1982) measured physical contact in different cultures rather than the social significance and the emotional implications of tactile behaviour, which inform and shape the actual practice. Suvilehto et al. (2015, 2019) reported that, despite broad overall similarities, there are significant differences in the patterns of permissible bodily touching across cultures. Recently, neuro-psychological approaches to social touch have made significant progress in our understanding of the underlying mechanisms and social implications of human touch behaviour (Cascio et al. 2018; Su and Su 2018). Analysing the effects of social touch on loneliness is relevant to elucidate the evolutionary, cultural, and psychological processes that come into play in this relationship.
There are at least two reasons why analysing the effects of social touch on loneliness is important. Firstly, social touch is a powerful bonding mechanism that humans share with other primates that could constitute one environmental trigger for loneliness under an evolutionary perspective and one mechanism we use to reconnect with our social network. Secondly, it is important to place the analysis of the evolutionary mechanism of loneliness in real-life cultural contexts, given that tactile behaviour —and social interaction in general— is always culturally shaped.
The evolutionary theory of loneliness (Cacioppo et al. 2006, 2014) posits that feeling lonely is an alarm against survival risks associated with social isolation. The rationale is that loneliness arises with relationship quality decay and motivates us to invest time and effort to restore social bonds. Although relationship quality has been identified as the signal that triggers the loneliness alarm (Hawkley et al. 2008; Hawkley and Cacioppo 2010), it is not yet clear exactly what features of social interaction humans monitor to establish the quality of their relationships or what are the underlying mechanisms that activate and deactivate the alarm. Evolutionary features respond to specific processes but the quality of a relationship —or the perception of the quality— is a general abstraction of a collection of events and perceptions rather than a straight-forward unitary trait. In other words, to provide adaptive fitness, loneliness needs to be connected in some way to the (social) environment and not just to a subjective evaluation (satisfaction) or perception of the social network and/or social interaction. Thus, more detailed analysis is needed to understand how the internal mechanism of feeling lonely is connected to the external environment, and previous research suggests that physical contact might be involved in that connection.
Bonding in primates relies on tactile social activities (such as halo-grooming) and humans evolved to incorporate additional more time-efficient mechanisms such as singing, dancing and storytelling (Dunbar 2012). But modern humans still use and respond to physical bonding mechanisms (Dunbar 2010; Nummenmaa et al. 2016); affectionate touch has been found to have positive effects on relational well-being and it is associated with emotional depth in the relationship (Jakubiak and Feeney 2017; Suvilehto et al. 2015, 2019). Therefore, social touch might be one specific signal of the quality or depth of a relationship that could be involved in triggering loneliness as a warning against social disconnection.
As mentioned before, tactile behaviour varies widely among cultural groups and differences in the amount of touch, the forms of touch and its meanings or acceptability are not universal (Burleson et al. 2019). However, if touch is involved in an evolutionary mechanism, some of its effects must remain even in the face of contradicting cultural practices, given that evolutionary features are not culture-specific. Therefore, it is important to discriminate cultural practices from evolutionary processes —which is not easy. However, one strategy to address this difficulty is to analyse the effects of touch on loneliness in a cultural context that does not promote physical contact to distinguish cultural expectations as potential confounders.
The aim of the research reported in this paper is to examine the relationship between touch and loneliness. This topic is relevant because if loneliness is taken as an evolutionary feature, it needs to be connected to the environment, respond to it, and provide adaptive fit. But the specific environmental cues that trigger loneliness are not clearly identified yet in the evolutionary approach. Thus, we experimentally tested whether touch —as an evolutionarily ancient mechanism for social bonding— still affects loneliness despite contradicting social practices and values. We firstly address the theoretical argument that touch (or lack thereof) could act as an environmental cue for social isolation by discussing existing literature on the connection between physical contact and loneliness. Secondly, we present and discuss the evidence from an experimental study, which showed reduced levels of loneliness (neglect item) among people who received physical contact, compared to those who did not.
This research adds to the current evolutionary theory on loneliness by suggesting a novel research line on the specific mechanisms that trigger or deactivate the loneliness alarm (i.e., physical contact) with an approach that considers the specificities of the cultural context in which individuals actually live and interact (e.g., cultural expectations of independency and self-reliance).
Touch and Loneliness in Individualistic Low-Contact Societies
Touch and physical displays of warmth or caring are important for human social bonding and psychological wellbeing (Field 2010; Gallace and Spence 2010; Jakubiak and Feeney 2017). There is evidence suggesting that touch, emotional depth or relationship quality, and loneliness are interconnected. Loneliness is related to poor relationship quality, particularly within the core of the social network (Heatley Tejada et al. 2017; Jackson et al. 2000), and those relationships, deep emotional ties, seem to rely heavily on physical contact and tactile ways of displaying affection (Dunbar 2010; Suvilehto et al. 2019). In addition, psychological research has shown that activities involving sensory perception —touch in particular— have therapeutic value for children who feel lonely (Bullock 1993; De Chomaso 1971). Although the pathways seem straightforward, the connection between these variables has seldom been directly addressed.
Although their role in human life is crucial, relationships and touch are both affected by cultural differences. Cultural norms regulate touch behaviour (Suvilehto et al. 2015, 2019) and by doing so they could have an effect over bonding, the sense of connectedness, and loneliness feelings. Some “cold” or low-contact cultures associate independence and self-reliance with not needing physical contact, or rather they associate high contact needs with dependency or weakness. This is the case in Anglo-Saxon societies, which are considered low-contact and individualistic (Beaulieu 2004; Hofstede 2001; Sorokowska et al. 2017; Triandis 2001). In those individualistic settings, the ideal of independence and self-reliance might result in less social touch and, potentially, more loneliness. In fact, previous studies have found that individual level collectivism is associated with less loneliness (Heu et al. 2019).
Western lifestyle and contemporary urban life are associated with growing levels of loneliness and extended periods of people living alone, which increase vulnerability to loneliness (Twenge et al. 2019). According to The Mental Health Foundation’s report, most of the risks of experiencing loneliness associated to contemporary Western lifestyle relate to the impossibility of —or limitations to— obtaining close and intimate physical contact and bonding experiences (Griffin 2010). Loneliness is an increasing concern in individualistic and low-contact societies (Cacioppo and Cacioppo 2018; Griffin 2010; Murthy 2017)Footnote 1 even though people are used to reduced physical contact and most of them strive to become independent and self-reliant.
It thus seems that cultural training in low contact, self-reliance and independence does not “immunize” against feeling lonely. What needs to be further analysed is whether observable effects of touch over loneliness remain in the presence of cultural training in low-contact societies. To put the latter to test, we designed an experimental study where people from a low-contact individualistic society (UK) were exposed to physical contact with another person to examine if they reported lower loneliness scores than a control group. This would allow us to observe a) if touch has an observable effect over loneliness scores, and b) whether there is evidence of an underlying (potentially evolutionary) mechanism that persists despite (i.e. is not annulled by) a cultural context that discourages touch in usual social interaction.
To analyse the effects of physical contact on loneliness in a low-contact culture we designed an experimental study that aimed to assess potential differences in loneliness scores after a brief exposure to physical contact (light rub on the back of the participants’ hands).
42 volunteers were recruited in the UK and randomly assigned to two experimental groups (physical contact and control). Two participants had to be removed from the final sample due to inconsistent heart rate readings or not following given instructions. Final sample was formed by 40 participants (27 females, 13 males), aged 27 on average (SD = 13.61), of which 22 (55%) reported being in a romantic relationship. Relationship status and gender distribution by experimental groups is shown in Table 1.
Measures and Procedure
Upon arrival, participants were informed of the activities involved in the experiment, requested to provide informed consentFootnote 2 and wear a heart rate monitor throughout the length of the experiment. They were randomly assigned to the experimental condition (having a few drops of oil rubbed to the back of their hands by the researcher) or to a control group (rubbing the oil on themselves). After that, all participants completed a questionnaire containing social network profiling, attachment style and loneliness measures, and an emotion recognition task.
Heart rate (HR) was recorded with a Polar H7 monitor and results were processed using Kubios software. The aim of having HR recordings was to detect potential differences in unconscious physiological responses to the experimental condition. This was considered relevant, given that loneliness has been found to have significant effects over HR.Footnote 3
Loneliness was measured with the Multi-faceted Loneliness Inventory (IMSOL) (Montero y López Lena 1999). This instrument was previously tested in a sample of 217 British participants. Reliability coefficient was α = .830 and the correlation with the short version of the UCLA loneliness scale was r = .8168, p < .001 in that sample. Similar to the UCLA scale, the IMSOL measure consists of three questions on subjective feelings of loneliness (i.e., “How often do you feel neglected?”, “How often do you feel lonely?”, “How often do you feel emptiness in your life?”) on a 1 (never) to 5 (always) frequency scale. A simple 1 (very little) to 10 (extremely) intensity scale was additionally included (i.e., “When you feel lonely, how intense would you say your loneliness feelings can be?”) to explore potential differences in loneliness feelings’ strength as opposed to frequency.
The “Reading the Mind in the Eyes” test (RMET) (Baron-Cohen et al. 2001) was also applied as part of the broader research project. This test measures emotion recognition abilities and consists of a series of 36 images showing the eyes area of the face only and participants are asked to choose the emotion that each image evokes from a list of four available options.
After obtaining informed consent, participants were asked to wear the heart rate monitor. Firstly, a 2.5 min test was taken as test reading in a sitting position with two aims: verifying that the HR monitor was working properly, and making sure that all participants’ readings started from a resting position, allowing the effect of any previous physical activity on their heart rate (climbing the stairs to the experimental room, for instance) to fade out. For these reasons, this initial reading was not part of the analysis of HR results. Participants were able to see their on-line readings and the autonomic balance report as given by Elite HRV app (“morning readiness” test) at the end of this initial reading but they were not able to see their bpm or any other result during the rest of the experiment.
After this, participants received physical contact in the form of a light rub to the back of their hand. A female researcher (the same person in all cases) held the participant’s hand and applied the oil with a light massage, lasting approximately one minute, until it was evenly distributed and absorbed. They were told that the oil was applied as part of the experiment to test if it enhanced autonomic balance during the task they were about to solve. The oil was simply baby oil and participants were debriefed on this at the end of the session. The researcher rubbed the oil to the hand of the participants randomly assigned to the experimental condition. Participants assigned to the control group were asked to rub the oil themselves.
Next, the participant was left alone to answer the questionnaire and task. The HR was recorded until they finished with all the questions. When they were finished, the HR recording was stopped, and participants were asked to remove the monitor. The session ended with debriefing and thanking them for taking part in the experiment.
Touch for apparently pragmatic reasons was selected for this study to avoid, firstly, making participants feel uncomfortable when experiencing physical contact from a complete stranger. And secondly, to avoid disclosing up front the purpose of the experimental condition, which would very likely bias participant’s reaction to the stimulus and their responses to the questionnaire. The limitations of this strategy will be addressed in the “Discussion” section.
Analysis of Variance (ANOVA) was performed to examine the effects of the two conditions on loneliness scores and on heart rate. A strong artefact correction algorithm was appliedFootnote 4 to clean HR data before the analysis. The HR recording lasted for as long as it took the participant to answer all questions and tasks, and thus recording length varied form one participant to another. Therefore, to analyse comparable recordings, we only considered the first 20 min for the analysis, which was the time it took most participants to finish. The first half was divided into 2.5 min long periods to analyse in more detail the immediate reaction to the physical contact, given that the experimental condition occurred at the beginning of the session and we expected changes (if any) to occur during or immediately after the hand rub. Then, the overall average of the full 20 min of recording was calculated separately (see Table 4).
Descriptive statistics and loneliness, attachment and HR results can be found in Table 2.
We found significant differences between experimental and control groups (condition) and by relationship status in one of the loneliness scale items that reflects perceptions of neglect. Although there were no significant differences in the loneliness global average score, ANOVA showed that mean differences in the neglect item were significant between conditions F(1, 36) = 5.034 p = .031 and relationship status F(1, 36) = 9.274 p = .004. Post-hoc t-tests revealed that single participants that were exposed to physical contact reported significantly lower scores than single participants who did not receive physical contact (p < .05) (see Fig. 1 below). Participants in a relationship had lower scores than single participants (p < .05) and participants in the experimental condition had lower scores than the control group (p < .1). Table 3 shows mean and standard deviation by experimental group and by relationship status.
Regarding heart rate analysis, participants in the physical contact condition showed a significant difference (F(1, 38) = 3.759 p = .017) in heart rate mean in the overall average versus the first 2.5 min period. That is, pairwise comparisons showed that the average HR of people who received physical contact was significantly lower compared to the first reading of the experiment (p < .05). There were no significant differences observed in the control group in any of the periods or the overall average. Table 4 gives mean beats per minute by experimental condition and by period, followed by the corresponding graph in Fig. 2.
The observed differences show that the physical contact condition is associated with a faster and more persistent reduction of HR. Decreasing beats per minute was expected given that the experiment was conducted in a sitting position and did not involve physical activity. However, participants who received physical contact had a steeper downward slope than the control group, which resulted in the average HR being significantly different (lower) than the first analysed reading in the experimental group.