Thermosensation and emotion: Thermosensory accuracy in a dynamic thermal matching task is linked to depression and anxiety symptomatology

Interoception is related to the generation of bodily feelings and the awareness of ourselves as ‘sentient beings ’ , informing the organism about its bodily needs to guarantee survival. Previous studies have reported links among interoception, emotion processing, and mental health. For example, the alignment of interoceptive dimensions (i. e., accuracy, sensibility, awareness) can predict emotional symptoms, such as anxiety. Here, we aimed to investigate the relationship between the perception of a certain type of skin-mediated interoceptive signal, i.e., thermosensation, and self-reported depression, anxiety, and stress. One hundred seventy participants completed the Depression Anxiety Stress Scale (DASS-21) and a dynamic thermal matching task, a static temperature detection task, and a heartbeat counting task. Our results revealed that self-reported anxiety and depression were related to the perception of temperature on hairy and non-hairy skin, respectively: higher anxiety was related to better performance on the thermal matching task on the forearm, while higher depression was related to poorer performance on dynamic and static temperature tasks on the palm. Discrepancies between thermosensory accuracy and sensibility measures (‘trait prediction error ’ ) were related to heightened anxiety, in line with previous studies. No significant correlations were found between DASS-21 scores and heartbeat counting accuracy. In conclusion, this study suggests that individual differences in thermosensory perception in different areas of the body are associated with self-reported anxiety and depression.


Introduction
Influential theories suggest that emotional feeling states arise from physiological changes within the body (e.g., [1,2]).These proposals are at least partially supported by empirical data concerning the contribution of interoceptive mechanisms in emotion processing (e.g., [2][3][4][5][6]).Interoception refers to 'the process by which the nervous system senses, interprets, and integrates signals originating from within the body, providing a moment-by-moment mapping of the body's internal landscape across conscious and unconscious levels' [7].As such, the impact of interoception is thought to extend beyond homeostatic regulation and survival and to relate to the awareness of ourselves as 'sentient selves' at any given time [8][9][10].
The relationship between interoception and emotion processing is bidirectional: on the one hand, it has been suggested that changes in ascending interoceptive inputs from the heart and stomach modulate the subjective experience of emotions (e.g., how intensely an emotion is experienced, [11]).On the other hand, changes in emotional experiences activate physiological reactions in the body that in turn can modulate afferent interoceptive signals (e.g., [12,13]).Furthermore, behavioural, physiological, and neuroimaging studies have shown an overlap in individuals' abilities to attend to interoceptive stimuli and the way they process emotions (e.g., [14,15]).These studies showed that such bodily signals are closely linked to higher cognitive functions such as memory (e.g., [16]) and decision-making (e.g., [17,18] for an overview) as well as the experience of specific types of emotions (e.g., fear, Garfinkel et al. [19], see also see [2,3,4,6,20,21]).
Recent developments in the field of skin-mediated interoception have proposed that thermosensation might provide a novel method to quantify interoception [54,65,66].The skin, given its nature, is a sensory organ exposed to signals from the inside of the body, and, as part of one's own body, it provides a rich source of internal signals regarding the state of the bodily self.However, skin signals also provide information about the external environment, and temperature has been traditionally considered an aspect of touch.Thus, both interoceptive and exteroceptive sensory information are provided by skin signals, making it difficult to disentangle the two (Perspect Psychol Sci [67]).This idea is supported by anatomical, physiological, and experimental arguments showing that noxious, thermal, and affective-touch information is signaled by special classes of receptors (C-fibres) in the skin, which reach the brain via different anatomical pathways through the spinal cord and thalamus than discriminative tactile information [8,68].These C-fiber-mediated somatosensory signals target a different cortical area, the posterior insular cortex, which is crucial for interoception and processes visceral information [51].Crucially, the body and brain are both involved in thermoregulation processes, such as sweating or shivering, and these bodily functions are also closely linked to the experience of emotions such as stress and fear [69].Although the relationship between changes in body temperature and the experience of emotions has received increasing attention [69,70], less is known about the relationship between individual differences in the perception of thermal changes on the body and affective symptoms.Accordingly, here, we focused on the relationship between self-reported depression, anxiety, and stress and thermosensation.We combined data from three different studies where we recruited a total of 172 participants and asked them to complete the Depression Anxiety Stress Scale (DASS-21, [71]).To target different aspects of thermosensation, we tested the perception of moving (dynamic) thermal stimuli on the skin by means of the newly developed thermal matching task [72] and the perception of static temperature by means of a classic temperature detection task [73].In keeping with previous studies suggesting a link between the experience of emotions and cardiac interoceptive accuracy (e.g., [31,74]), we also employed a heartbeat counting task [75] as a measure of cardiac interoception, with the aim of replicating the findings that people with better cardiac interoceptive accuracy show better processing of affective symptoms, such as anxiety [2].We hypothesised that we would observe a similar relationship between emotion processing and thermosensation, quantified by means of the dynamic thermal matching task and the static thermal detection task (see also [72]).In particular, we hypothesised that the newly developed thermal matching task would be especially sensitive to detecting correlations with affective symptoms because it taps into interoceptive thermal comfort processes and involves a slow, caress-like application of thermal stimuli to the skin (see Crucianelli & Ehrsson, [67] and [72] for a full account of the interoceptive aspects of thermosensation quantified by means of the thermal matching task).Furthermore, we expected that people with higher discrepancies between interoceptive dimensions across tasks would also show higher depression and anxiety symptoms, in keeping with previous studies [31].

Participants
A total of 172 healthy, naïve participants (87 females and 85 males, mean age 26.297 ± 5.05 years) were recruited across three separate studies using social media and advertisements on the Karolinksa Institutet campus.The inclusion criteria included being 18-40 years old and being right-handed.The exclusion criteria were having a history of any psychiatric or neurological condition, taking any medications, having sensory or health conditions that might result in a skin condition (e.g., psoriasis), and having any scars or tattoos on the left forearm or hand.All participants provided signed written consent, and they received a cinema ticket or 200 SEK per hour as compensation for their time.The study was conducted in accordance with the provisions of the Declaration of Helsinki, as revised in 2008.Part of the data was analysed using a different statistical method for different purposes and has been published as part of two other manuscripts (N = 64; [72]; N = 64; [76]).

Self-reported measures and interoceptive sensibility
Participants were asked to provide demographic information, such as age, weight, height (to calculate the body mass index, BMI), and handedness.Next, participants were asked to complete the following self-report questionnaires: the Body Awareness Questionnaire (BAQ), an 18-item questionnaire assessing body awareness [77] as a measure of interoceptive sensibility, that is, participants' reported awareness of their bodily sensations, and the Depression, Anxiety and Stress Scale -21 Item (DASS, [71,78]).

Heartbeat counting task
Heartbeat frequency was recorded by means of a Biopac MP150 Heart Rate oximeter (Goleta, CA, United States) attached to the participant's nondominant index finger and connected to a Windows laptop with AcqKnowledge software (version 5.0).Participants were asked to silently count their heartbeats between two verbal signals of 'go' and 'stop', without manually taking their pulse or feeling their chest [75].They were encouraged to only count the heartbeats they were sure about but were also instructed to take into account weak sensations rather than making their best guess (as in Ferentzi et al., [36,79]).Participants completed a 15-s practice trial before proceeding to the actual counting task, with 108 participants completing three experimental trials (interval lengths of 25 s, 45 s, and 65 s, as in [53,72]) and 64 participants completing six experimental trials (interval lengths of 25 s, 30 s, 35 s, s, 45 s, and 50 s, as in [76]), which were presented in a randomized order.Short 30-s breaks were taken between each trial.

Thermal matching task
Thermal stimuli were delivered using a 25 × 50 mm thermode attached to a thermal stimulator (Somedic MSA, SenseLab, Sweden).We followed the procedure fully described in Crucianelli, Enmalm & Ehrsson [72].The participants' left forearms and palms were stroked at reference temperatures of 30 • C, 32 • C or 34 • C, and they were instructed to focus on this reference temperature because the task consisted of verbally indicating when they felt the same reference temperature again.That is, participants were asked to tell the experimenter which temperature felt the same as the reference temperature among a range of warmer or cooler stimuli.The experimenter touched the participant's skin with the thermode set at different temperatures ranging from ± 8 • C (which is 25 % of the neutral temperature of 32 • C; whether the starting temperature was + 8 • C or -8 • C from the reference temperature was counterbalanced across participants) of the reference temperature (ranging from 22 to 38 • C for the reference temperature of 30 • C; 24-40 • C for the reference temperature of 32 • C; and 26-42 • C for the reference temperature of 34 • C).The task followed a staircase procedure, that is, the temperature was either increased (i.e., from cool to warm) or decreased (i.e., from warm to cool) towards the reference temperature in discrete steps of 2 • C. The temperature was increased or decreased until the participants verbally indicated they felt the reference temperature or until the maximum or minimum temperature was reached (± 8 • C from the reference temperature, opposing the starting temperature) for a total of 9 potential strokes per trial, with a break of 3 s between trials, during which the participants were given the opportunity to provide their answer (i.e., affirm whether the perceived temperature corresponded to the reference temperature).The correct answer was always the reference temperature, and the order in which the reference temperatures were presented as well as the order of increasing and decreasing trials varied across trials to avoid anchor effects of the initial values (e.g., if one participant started with increasing trials based on one reference temperature, then they started with decreasing trials for the following reference temperature).Two trials per reference temperature were repeated, with one increasing and one decreasing trial, for a total of 6 trials presented in randomized order.The duration of each stroke was kept constant at 3 s; the velocity of tactile stimulation was optimal to target the C-Tactile afferent system (3 cm/s, [80]), and the direction of movement was always proximal to distal with respect to the participant.No additional pressure was applied aside from the weight of the thermode.The same procedure was repeated on the outer forearm (hairy skin) and the palm (non-hairy skin) in areas of 9 × 4 cm.

Static temperature detection task
As in the dynamic thermal matching task, tactile stimuli were delivered using the Somedic MSA Thermal Stimulator.We followed the procedure of the Marstock methods of the limits [73,81] fully described in [72].The experimenter held the thermode on the area of interest (left forearm or palm) without applying any additional pressure.Participants were asked to hold a response button using their right hand and to press it as soon as they perceived a change in temperature of any kind (i.e., warmer or colder than the previous perceived temperature, [73]).The starting temperature was always neutral (32 • C); the maximum probe temperature was set to 50 • C, and the minimum temperature was set to 10 • C for safety reasons.As soon as the button was pressed, the temperature automatically changed in the opposite direction and returned to the baseline temperature of 32 • C; the temperature stayed at 32 • C for 5 s before the next trial was started.The temperature changed at a rate of 1 • C/s and returned to the baseline temperature at a speed of 4 • C/s.This method included a total of five warm and five cold trials, presented in two blocks (warm and cold blocks).The procedure was repeated twice in a randomized order: once on the left forearm and once on the left palm.The static temperature detection task has been generally shown to have high reliability (e.g., [73,[82][83][84][85][86]) and convergent validity (Heldestad [87]).

Experimental procedure
Participants were welcomed into the experimental room and asked to sit at a table opposite the experimenter.Upon arrival, they were asked to sign a consent form and to complete the demographic questionnaire, BAQ (128 people only) and DASS-21 presented in an online format.The questionnaires were always presented at the beginning of the experimental procedure to ensure that participants were given some time at rest before completing the heartbeat counting task; this was the first interoceptive task that all participants completed to avoid any influence by other activities (e.g., [47,88]) (for an overview of the procedures and tasks, see Fig. 1).The soft oximeter was carefully and firmly placed around the finger without being too tight to reduce the possibility that participants could perceive their pulse in their finger [43,53].Participants were instructed to place both hands on the table, to breathe normally and to not cross their legs.To ensure accuracy, the participants were given the choice to either keep their eyes closed or open, whichever helped them feel more comfortable.The aforementioned experimental procedure took approximately 30 min, giving participants the opportunity to acclimatize themselves before proceeding with the dynamic thermal matching task.Participants were asked to wear a disposable blindfold and to place their left arm on the table to complete the dynamic thermal matching task, following the method fully described in the Methods section above.Upon completion, participants removed the blindfold, and they were given a short break before starting the static temperature detection task.No break was taken between the cold and warm blocks, and participants were only allowed to remove the blindfold at the very end of the task.All the experimental tasks were conducted on the left, nondominant hand or forearm.The starting location for each task was alternated between the forearm and the palm (e.g., participants who started one task on the palm started the next task on the forearm; those who started one task on the forearm completed the following task on the palm).The order of the tasks was kept constant (with internal randomization) (Fig. 1).The entire experimental procedure lasted approximately 45 mins, and the participants were provided with a full debriefing at the end of the session.Testing took place in a testing room with a constant temperature and humidity level.

Design and plan of analysis
All statistical analyses were performed in JASP (JASPTeam, 2019).First, we analysed each task separately, namely, the thermal matching task, the static temperature detection task, and the heartbeat counting task.We focused on the analysis of the outcome measures of each task (i.e., the participant's performance) and the interoceptive trait prediction error (i.e., ITPE; see below) for the interoceptive tasks.These data were analysed by means of repeated-measures ANOVAs.Bonferroni-corrected post hoc comparisons were used to follow up significant effects and interactions.All p values were 2-tailed unless otherwise specified.Then, we ran correlational analyses to investigate the relationships between task performance and the questionnaire scores (DASS and BAQ).Given the ordinal nature of the questionnaire data, these data were analysed using non-parametric statistics (i.e., Spearman correlations).

Task performance and task accuracy
Heartbeat data were analysed using the 'count peak' function of AcqKnowledge software (version 5.0), which enabled extraction of the actual number of heartbeats within the specified time windows.We calculated heartbeat counting task accuracy (HCT accuracy, i.e., cardiac interoceptive accuracy) by means of the following formula, which allowed us to compare the counted and recorded heartbeats [75]: For the other tasks, the focus was 1) to explore whether there was a significant effect of touch location (hairy vs. non-hairy skin) and 2) to obtain an accuracy value that resembled, and therefore could be compared to, the interoceptive accuracy measured by means of the heartbeat counting task.This was done to ensure that levels of accuracy were equated across the modalities.
For the dynamic thermal matching task (TMT), we used the following formula to obtain a score that corresponded to TMT accuracy, which thus reflects "thermal interoceptive accuracy" similar to how HCT accuracy captured cardiac interoception: where 12 represents the total number of options presented to the participants (regardless of directionoverestimation or underestimation of temperature) across the three trials.Both formulas normally provide a value between 0 and 1, with 0 suggesting poor performance and 1 indicating optimal performance on the task.We kept the order of the increasing and decreasing stimuli separate given the different mechanisms and skin responses known to be involved when perceiving cooling and warming stimuli (for a review, see [89]).Thus, for each subject, we obtained one increasing and one decreasing accuracy value for the forearm and the palm.
Next, for the static temperature detection task, in keeping with previous studies, we were interested in both the sensitivity (TDT sensitivity, i.e., the smallest change in temperature a person could detect) and the consistency or precision (TDT consistency, i.e., the variability in the individual responses across the different trials, quantified as standard deviations) of detection across trials.

Interoceptive trait prediction error (ITPE)
As in Garfinkel et al. [31], the interoceptive trait prediction error (ITPE) was operationalized as the difference between the objective interoceptive accuracy and subjective interoceptive sensibility.For each task's accuracy (HCT accuracy and TMT accuracy) and sensibility variable (BAQ score), scores were converted to standardized Z values.

Z = (x − mean )/SD
On a within-participant basis, ITPE values were calculated as the difference between interoceptive sensibility and interoceptive accuracy.ITPEs for cardiac interoception and thermal interoception were calculated separately using the accuracy scores from each task and the sensibility score determined by the BAQ.Positive ITPE values indicated a propensity for individuals to overestimate their interoceptive ability, while negative scores reflected a propensity for individuals to underestimate their interoceptive ability.

Demographics and self-report measures
The mean scores and standard deviations for age, BMI, interoceptive sensibility (as measured by the BAQ), and the DASS score are reported in Table 1.No effect of sex on any of these measures was found.

Relationship between dynamic thermal matching task accuracy and self-report measures
As mentioned in the Methods section, we obtained one increasing (staircase) temperature accuracy value for the forearm and one for the palm and one decreasing (staircase) value for the forearm and one for the palm.The results of the 2 (skin site: palm vs. forearm) x 2 (staircase: increasing vs. decreasing) repeated-measures ANOVA revealed a significant main effect of location (F (1, 171) = 28.15,p < 0.001, η p 2 = 0.057, Fig. 2), with participants being more accurate in the detection of dynamic temperature on the forearm (M = 0.829; SD = 0.134) than on the palm (M = 0.777; SD = 0.156).No main effect of staircase (F (1, 171) = 0.368; p = 0.545, η p 2 = 0.0059) or significant interaction (F (1, 171) = 0.01; p = 0.922, η p 2 = 0.00018) was found.Thus, we replicated previous findings in the thermal matching task, suggesting better  performance on hairy skin than on non-hairy skin [72,76].
The results of the correlational analyses investigating the relationship between TMT accuracy and depression, anxiety, stress and BAQ scores are reported in Table 2 and Fig. 3 below.Notably, we found a positive significant correlation between TMT accuracy on the forearm with increasing temperature and anxiety scores and a negative significant correlation between TMT accuracy on the palm with increasing temperature and depression score.
In terms of ITPE, the results of the correlational analyses investigating the relationship between ITPE based on TMT accuracy and depression, anxiety, and stress scores are reported in Table 3 and Fig. 4. Notably, we found negative significant correlations between the anxiety score and ITPE for the TMT performed on the forearm with decreasing temperature and on the palm with both increasing and decreasing temperature.

Table 2
Spearman's rho and p values for the correlations between questionnaire scores and thermal matching task (TMT) accuracy.
The results of the correlational analyses investigating the relationship between TDT sensitivity and consistency scores on the temperature detection task and depression, anxiety, stress and BAQ scores are reported in Table 4 and Table 5, respectively.See also Figs. 10, 11, 12, and 13.Notably, we found a negative significant correlation between sensitivity to the detection of warm temperatures on the forearm and the BAQ score.In terms of consistency in the temperature detection task, we found a significant correlation between consistency in the detection of cold temperatures on the palm and depression and BAQ scores.

Discussion
The present study investigated the relationship between thermosensation and cardiac interoception and state self-reported depression, anxiety, and stress.Our results regarding performance on the thermal matching task [72] revealed that anxiety and depression were related to the perception of dynamic temperature on hairy and non-hairy skin, respectively, but only when the temperature was increasing.That is, higher anxiety was related to better performance on the thermal matching task performed on the forearm, while higher depression was related to worse performance on the task performed on the palm.We also found a significant positive correlation between depression only and consistency in the temperature necessary to perceive a static cold temperature on the palm.These findings might suggest that dynamic thermosensation (i.e., an object moving on the skin at a pleasant speed), especially on hairy skin, might play a more salient role in emotion regulation than static thermosensation [72].We did not replicate previous findings suggesting a link between cardiac interoception and ITPE and depression, anxiety, and stress [31]; therefore, our negative findings are consistent with those of Desmedt et al. [32]; nevertheless, this inconsistency in the literature calls for further investigations on this topic.In particular, the use of novel tasks to quantify cardiac   interoception will likely be valuable because the reliability of the heartbeat counting task has been discussed and questioned in recent years (e.g., [32,[34][35][36]38,[40][41][42][43][44][45][46][47][48][49]88,93], but see [33,39]).Nevertheless, in the current study, we followed the best methodological practice to employ the heartbeat counting task (as suggested by [40]).Our choice was motivated by the aim of replicating previous studies that used this task to explore relationships between individual variability in heartbeat counting accuracy and anxiety and depression (see the Introduction).
Our results are the first to show a link between the ability to promptly perceive moving thermal changes on the body surface (such as those typically occurring in social exchanges with conspecifics, which are very close to thermoneutrality, as in the case of the thermal matching task; see Crucianelli, Enmalm & Ehrsson [72] for more details) and self-reported symptoms of depression and anxiety.Interestingly, we found a dissociation between skin sites and depression and anxiety, which might relate to the specific function of touch in social encounters as well as the more passive and active roles that these body parts (forearm vs. palm) play in social touch.Specifically, we found a relationship between performance on the thermal matching task performed on the forearm and anxiety symptoms.We know that specific types of touch on the forearm can communicate arousal and warning [94]; our findings suggest that people with higher anxiety might be particularly reactive and tuned into this type of social signal in the context of thermal comfort.In contrast, higher depression scores were related to worse performance on both the dynamic thermal matching task and the static detection task only when performed on the palm.We know that people with higher depressive symptoms are also more likely to experience social isolation [95].Thus, a lower reaction to perceived (i.e., performance on the thermal matching task) and detected (i.e., performance on the temperature detection task) thermosensory stimuli on the palm could be linked to reduced engagement in social tactile interactions since the palm is usually the body part used to actively touch other people.Importantly, we recently showed a significant link between self-reported depressive symptoms and the perception of affective touch [63].Taken together, this evidence points towards a generalizable link between affective and thermosensory interoceptive signals and depressive symptoms, whereby the social dimension conveyed by dynamic, affective, and thermal comforting tactile stimulation seems to play an important role.
We also found a relationship between discrepancies in performance and confidence in the thermal matching task and anxiety, in line with previous studies conducted in the context of cardiac interoception [31].This finding confirms the idea that the lack of alignment of interoception dimensions might be related to psychopathological symptoms by providing novel evidence regarding thermal interoception.This finding combined with the lack of a significant relationship between thermal and cardiac interoceptive accuracy scores and the interoceptive sensibility score (i.e., BAQ score) are in line with the idea that interoception can be conceptualized as a multidimensional construct, showing high independency between the dimensions [30,31,72,96].Nevertheless, in the present study, we used the BAQ to calculate the interoceptive prediction error (ITPE) and as a measure of interoceptive sensibility.However, the BAQ assesses a broader range of interoceptive dimensions and body domains than the experimental task used in this study.Furthermore, recent evidence suggests that BAQ scores do not converge with other measures of interoceptive sensibility to a satisfactory extent [97], thus resulting in low construct validity for our measure of ITPE.Thus, our results are in line with recent evidence showing issues in the interpretation and replicability of findings in self-reported interoception research and the need for better constructs and measures [97].
It is important to point out that we measured self-reported depression, anxiety, and stress in healthy participants, and their scores were within the healthy range.Therefore, our findings should be interpreted  in this context and be generalized with caution.Future studies should investigate their clinical translational potential and whether skinmediated interoception can be considered a biomarker to identify people at risk of developing mood symptomatology disorders.Furthermore, in this study, we used depression, anxiety, and stress as a proxy for emotion processing, and we did not quantify emotion processing by any other means.However, future studies could investigate the relevance of thermosensation to emotions more specifically, for example, in the case of thermal sensations commonly observed during the experience of anger or fear (e.g., [98]).In this context, it would also be interesting to investigate the relationship between internal body temperature (i.e., core temperature) and skin temperature (i.e., peripheral temperature) in response or in relation to different emotions.Similarly, it would be very interesting to explore spontaneous fluctuations in temperature in the body occurring in response to different emotional experiences.This approach would be highly ecologically valid and provide further information about the best temperatures to use in thermal tasks, such as in the thermal matching task.
It should be noted that in the present study, we pooled data across three different experiments as a post hoc explorative investigation.Therefore, the present study and its plan of analysis were not preregistered, and we did not conduct an a priori power analysis; as such, these are two methodological limitations.However, the methods and order of tasks were the same across the three experiments, which provides support for our approach.Furthermore, the final sample size was appropriate to reach conclusions regarding the relationship between thermosensation and self-reported measures of interoceptive sensibility (i.e., BAQ scores, compared to [72]).Nevertheless, our main findings of the correlations between the thermal matching task and anxiety and depression are novel but should be tested for replicability in future  studies.In addition, it could be argued that the participants' performance on the thermal matching task may have been influenced by practice or learning effects; nevertheless, we are confident that this was not the case because (1) we observed a main effect of location on task performance and this would not be the case if the participants had learned how to successfully complete the task across trials; (2) we did not observe a learning effect over time (difference in performance between the first and second block, regardless of location and temperature, p = 0.299); and (3) unpublished data from our laboratory showed a high level of test-retest reliability of the task across two independent testing sessions, thus minimizing the risk that performance on the thermal matching task could be influenced by a learning effect.To provide further support for the reliability of our task, we conducted additional control correlational analyses that show that the static and dynamic thermal tasks measure two distinct constructs (all p s between 0.061 and 0.848), as also suggested by differences in performance for different body locations (forearm vs. palm) and experimental conditions (cooling vs. warming) across tasks.Thus, future studies investigating thermosensation should take this evidence into consideration, avoid interpreting total scores that would aggregate different conditions and have different hypotheses based on the conditions.
Taken together, the present work supports a link between skin-based interoception, specifically thermosensation, and emotions and mental health [72].As such, the current findings can pave the way for further investigations on the use of thermal imaging as a potential measure of emotional arousal and anxiety [99,100].For example, future studies could investigate physiological responses in terms of thermosensation to emotion processing by using thermal cameras, as well as emotional responses to manipulations of body temperature, whereby thermal stimuli can be varied in the range of thermoneutrality just as in the context of heartbeats (see [101] for similar approaches to manipulate the heart rate).Such insight can shed light on conditions characterized by disorders in the experience of emotions such as alexithymia, which is also characterized by deficits in interoceptive processing [88] and thermosensation [102].

Fig. 2 .
Fig. 2. Main effect of skin site (forearm vs. palm) on performance on the thermal matching task.Single point data distribution, box plot with means and standard deviations, and group data density.

Fig. 3 .
Fig. 3. Density distribution of the TMT scores and their relationship with self-reported depression, anxiety, and stress scores.The red outline indicates a significant correlation.(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 4 .
Fig. 4. Density distribution of the ITPE scores on the TMT and their relationship with self-reported depression, anxiety, and stress scores.The red outline indicates a significant correlation.(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 5 .
Fig. 5. Density distribution of the heartbeat counting task (HCT) and its relationship with self-reported depression, anxiety, and stress scores.

Fig. 6 .
Fig. 6.Density distribution of the heartbeat counting task (HCT) and its relationship with self-reported depression, anxiety, and stress scores.

Fig. 7 .Fig. 8 .
Fig. 7. Main effect of temperature on sensitivity in the static temperature detection task.

Fig. 9 .
Fig. 9. Main effect of temperature on consistency in the static temperature detection task.

Fig. 10 .
Fig. 10.Correlations between depression, anxiety, and stress scores and sensitivity in the temperature detection task performed on the forearm (hairy skin) (warm on the top row and cold on the bottom row).

Fig. 11 .
Fig. 11.Correlations between depression, anxiety, and stress scores and sensitivity in the temperature detection task performed on the palm (non-hairy skin) (warm on the top row and cold on the bottom row).

Fig. 12 .
Fig. 12. Correlations between depression, anxiety, and stress scores and consistency in the temperature detection task performed on the forearm (hairy skin) (warm on the top row and cold on the bottom row).

Fig. 13 .
Fig. 13.The red outline indicates a significant correlation.(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Table 1
Means and standard deviations for the participants' age, body mass index (BMI), Body Awareness Questionnaire (BAQ) score, and scores on the depression, anxiety, and stress scales.Data are reported for 87 females and 63 males.For the BAQ, data are reported for 65 females and 63 males.

Table 3
Spearman's rho and p values for the correlations between questionnaire scores and interoception trait prediction error (ITPE) in the thermal matching task (TMT).

Table 4
Spearman's rho and (p values) for the correlations between questionnaire scores and sensitivity in the temperature detection task.

Table 5
Spearman's rho and p values for the correlations between questionnaire scores and consistency in the temperature detection task.