Interoception is usually defined as sensing and processing of afferent signals from within the body and its visceral organs to the brain (Cameron, 2001; Craig, 2002). Such afferent signals include changes in heart rate, respiratory, or muscle contractions, which are typically classified into proprioception and visceroception (Vaitl, 1996). Proprioception describes the sensation of bodily signals from the skin and the musculoskeletal apparatus, whereas visceroception comprises signals arising from within the body or the inner organs (e.g., cardiovascular, respiratory, and gastrointestinal signals). The perception of these signals is related to bodily states such as thirst, hunger, desire, pain as well as the level of energy and stress (Vaitl, 1996; Herbert and Pollatos, 2008). However, the increasing interest in interoception is due to its close link to emotions, feelings, and the perception of disease-specific symptoms (Schandry et al., 1993; Tsakiris et al., 2011). Individuals with a more precise detection of internal bodily signals experience emotions more intensively, are better able to process such signals, respectively, and to regulate them (e.g., Craig, 2003; Barrett et al., 2004; Critchley et al., 2004; Herbert et al., 2011). In order to get a better understanding of the concept of interoception and the consequences of interindividual differences, it will be defined in more detail below.

Garfinkel and Critchley (2013) and Garfinkel et al. (2015) proposed a clear demarcation between the different nomenclatures and meanings of the term interoception. Recently, they developed a three-dimensional model of interoception and provided empirical support for the necessity to assess different aspects of interoceptive processes (Garfinkel et al., 2015). The model differentiates between interoceptive accuracy (IAc), interoceptive sensibility (IS), and interoceptive awareness. IAc has been proposed as the ability to detect signals from within the body and to objectively quantify the individual performance via behavioral testing. Typically, IAc is investigated using procedures related to cardiovascular perception. Most prominently, IAc is tested by the heartbeat detecting task (Schandry, 1981) or the heartbeat discrimination task (Whitehead et al., 1977; Katkin et al., 1983; Brener and Kluvitse, 1988). Another established test to investigate IAc is the water load test based on the sensation of gastrointestinal signals (Vaitl, 1996; Herbert et al., 2012). Actual methods include procedures to measure the sensitivity for respiratory signals (Garfinkel et al., 2016). However, the heartbeat detection task is widely used in research on interoceptive processes and represents the standard procedure to measure IAc (e.g., Eshkevari et al., 2014; Kindermann and Werner, 2014; Koch and Pollatos, 2014; Pollatos and Georgiou, 2016; Pollatos et al., 2016). Furthermore, Garfinkel et al. (2015) defined IS as the second dimension of interoceptive processes by the subjective belief and confidence in the internal bodily state, mostly assessed by self-report questionnaires (e.g., Body Perception Questionnaire; Porges, 1993). Another possibility to assess IS is to indicate one’s own subjective confidence in the performance of a specific IAc task (e.g., Ehlers et al., 1995; Khalsa et al., 2008; Parkin et al., 2014). Whereas the self-report questionnaires regarding bodily signals in general, the confidence rating is more specific because of its request directly after the IAc task. Otherwise, the confidence rating includes only one item, which is repeated assessed after every interval. In general, IS depends on cognitive processes such as body awareness and associated evaluations, memories as well as attitudes (Mehling et al., 2009). Lastly, the metacognitive or interoceptive awareness dimension combines objective and subjective measures. The aim is to analyze the accordance between confidence and accuracy. For analyses, Garfinkel et al. (2015) suggest receiver operating characteristics (ROC) curves or the trial-by-trial confidence-accuracy correlation.

To emphasize the importance of interoception, there is evidence of a disturbed IAc and/or IS in clinical and subclinical populations. For example, in patients with anorexia nervosa as well as overweight and obese people a reduced IAc and IS are observed (Fassino et al., 2004; Pollatos et al., 2008; Herbert and Pollatos, 2014; Fischer et al., 2016). On the other hand, patients with bulimia nervosa and healthy controls did not differ in IAc (Pollatos and Georgiou, 2016), but bulimic patients showed a reduced IS (Fassino et al., 2004; Pollatos and Georgiou, 2016). In addition, Eshkevari et al. (2014) found no differences in IAc between different patients with eating disorders and healthy controls, whereas IS was higher in patients with eating disorders than in the control group. Otherwise, people suffering from somatization disorder have a reduced IAc compared to healthy controls (Weiss et al., 2014). Similar results were found for patients with higher scores of alexithymia (Herbert et al., 2011; Pollatos and Gramann, 2012) and healthy participants with depression and anxiety symptoms as well as patients diagnosed with depression (Pollatos et al., 2009; Dunn et al., 2010; Terhaar et al., 2012). Moreover, findings concerning interoceptive abilities of patients with panic disorders are quite mixed. Ehlers et al. (1995) and Ehlers and Breuer (1996) demonstrated that patients with panic disorders (as well as generalized anxiety disorder) showed an enhanced ability to perceive cardiovascular signals. In contrast, subsequent studies on this topic found no reliable differences in interoceptive abilities in people suffering from panic or generalized anxiety (Barsky et al., 1994; Van der Does et al., 2000). According to Ehlers and Breuer (1996), those inconsistencies are due to different instructions and sample characteristics used in subsequent research (e.g., including patients with medication that might has affected their cardiovascular systems). Previous research proved this only for the accuracy of cardiovascular signals, but they found no evidence for other interoceptive abilities (Limmer et al., 2015). To conclude, there is evidence that interoceptive processes are disturbed in people with different mental problems. Because interoception has shown to be of importance for processes of emotion perception and regulation (e.g., Herbert et al., 2007; Wölk et al., 2014; Kever et al., 2015; Pollatos et al., 2015), it seems necessary to develop and study methods that possibly lead to changes in interoceptive processes.

Until now, it is not clear how to improve interoceptive processes most effectively. Recent studies proposed interoception – especially IAc – to be a trait-like sensitivity toward visceral signals (Herbert and Pollatos, 2012; Herbert et al., 2012). Thus, there is good reason to assume that interoceptive processes are positively affected by trainings of heartbeat or breathing sensitivity. Several empirical studies support the idea that heartbeat perception can be trained, especially through manipulation procedures such as food deprivation, physical effort, and changes of the autonomic cardiovascular system (Herbert et al., 2007, 2012; Herbert and Pollatos, 2012). Another opportunity are self-focused procedures with the increase of attention through a photo of oneself, self-relevant words, or the own mirror image (Tsakiris et al., 2011; Ainley et al., 2012; Pollatos et al., 2016). Mostly, these outcomes are only short-term effects and found in participants with a low IAc prior to the training. Whereas different self-focus procedures showed first evidence for improved interoceptive processes, previous research investigating techniques stemming from a Buddhism mindfulness tradition are rare and the evidence is mixed. Because mindfulness interventions are concerned with the idea of being present in the moment, focusing on the body and related signals in a non-judgmental way, such methods are of high interest in research on interoception.

One approach to train interoceptive processes might be the so called body scan (BS) as recently suggested (e.g., Parkin et al., 2014). The BS is – in addition to classical forms of meditation and mindful yoga – a central element of the mindfulness-based stress reduction program (MBSR) developed by Kabat-Zinn in 1979 (Kabat-Zinn, 1982; Kabat-Zinn and Hanh, 2009; Sauer-Zavala et al., 2013). Common to all three techniques is a mindfulness part, where attention is paid to different parts of the body (e.g., toes, back or head) as well as sensations (such as pain or muscle tension) in the present moment without any judgment or criticism of the moment to moment experience (Kabat-Zinn, 2013). Additionally, during the BS participants were also constantly instructed to focus on their breathing and heartbeat.

Empirically, many studies showed a positive influence of an 8-week MBSR training on different health related outcomes (e.g., stress reduction, cancer, depression, and heart disease) in people with and without psychiatric or somatic disorders (e.g., Astin, 1997; Davidson et al., 2003; Ramel et al., 2004; Eberth and Sedlmeier, 2012; Katterman et al., 2014; Ussher et al., 2014; la Cour and Petersen, 2015; Yang et al., 2016). To date, there is only limited evidence showing positive effects of the MBSR-program as well as the BS implemented separately on parameters of the cardiovascular system like heart rate, cardiac pre-ejection period, and cardiac respiratory sinus arrhythmia (Ditto et al., 2006; Zeidan et al., 2010). However, there is a growing body of research showing positive effects of practicing the BS continuously on body perception, concentration, awareness of different disease-specific symptoms, and stress reduction (Ditto et al., 2006; Weiss et al., 2010; Dreeben et al., 2013). It is presented here that there are no studies available exploring the single effect of the BS over 8 weeks on interoceptive processes which is especially important for the research.

So far, previous studies considering the relationship between meditation, MBSR/Mindfulness-Based Cognitive Therapy (MBCT), or other mindfulness programs on interoceptive processes showed contradictory results. With the focus on meditation, one study examined the influence of a 10-day Vipassana meditation course and found an increase of IAc in meditation novices compared to an inactive-control group (Krygier et al., 2015). In addition, it has been demonstrated recently that a mindfulness training can improve IS [subscales attention regulation, self-regulation, body listening and trusting of the multidimensional assessment of interoceptive awareness (MAIA); Mehling et al., 2012] in depressed patients compared to both an active-control group and a healthy control group (Fissler et al., 2016). The results of a neurophysiological study support these findings (Farb et al., 2013). Farb et al. (2013) observed structural and functional changes in brain areas (anterior insula regions and dorsomedial prefrontal cortex) linked to interoception through meditation practice. Nevertheless, expertise in meditation did not seem to have an influence on IAc (Khalsa et al., 2008). Accordingly, heartbeat perception accuracy was unaffected by whether participants were experienced meditators (Buddhist or Kundalini; experience ranging from 12 till 29 years) or novices. Similar results were found in studies of Nielsen and Kaszniak (2006) as well as Melloni et al. (2013). In both studies, the level of experience in meditation was not associated with differences in IAc. However, meditation experts felt significantly more confident in their interoceptive performance and experienced the task as easier compared to non-experts. Lastly, Parkin et al. (2014) investigated the relationship between trait mindfulness and IAc as well as IS. In line with the aforementioned results of Khalsa et al. (2008), no relationship between mindfulness and heartbeat perception (IAc) was found, but participants with higher mindfulness scores were more confident regarding their heartbeat perception performance (IS; see also Kok and Singer, 2016). There is only one study investigating the influence of an 8-week MBSR- or MBCT on IAc and IS (Parkin et al., 2014). Results showed no improvement in heartbeat perception, but participants felt more confident in their performance. Similar effects were observed in patients suffering from chronic pain and depression after 8 weeks of MBCT on the subscales self-regulation and not distracting of the MAIA (de Jong et al., 2016).

In a recent study of Bornemann et al. (2015), IS could be shown to improve after a 3 months BS intervention and breathing meditation compared to a retest control group, which was recruited to investigate effects of repeated testing. Findings of another recent study failed to provide support for a positive effect of a BS training on IAc and IS (Parkin et al., 2014). In detail, the researchers compared a group participating in 15-min BS over 1 week with a control group who listened to a relaxing beach sound for the same amount of time, and a waiting-list control group.

To conclude, evidence is mixed concerning the idea that mindfulness practice strengthens the ability to perceive bodily signals. Results show no difference between intensive programs (e.g., meditation, MBSR or MBCT) and the isolated performance of the BS. Nevertheless, it has been argued in all described studies that meditation or other specific mindfulness tasks should improve IAc and IS as well (e.g., Khalsa et al., 2008; Parkin et al., 2014). In most studies, an improvement of IS through meditation was found, MBSR/MBCT and BS measured by questionnaire or estimation of confidence. One reason for the BS not being able to change IAc and IS in previous studies might be the training interval. In most cases, the BS was conducted only for 1 week or even shorter (e.g., Parkin et al., 2014), and information about standardization is often missing. Furthermore, Parkin et al. (2014) discussed the differences between IAc and IS. They attributed, that IS will strengthen cognitive processes for the body perception due to the training, especially evaluations, memories as well as attitudes, which is independent from the real perception. This assumption is in line with the differentiation between IAc and IS by Garfinkel et al. (2015) and should be considered.

In an attempt to clarify mixed effects of BS practice on interoceptive processes, we conducted two studies where participants trained the BS over a period of 8 weeks with daily practice sessions. Healthy students were trained in the standard BS suggested by Kabat-Zinn (2013) and longitudinal effects on interoceptive processes were assessed subsequently. This is the first study using three time points to measure outcomes of a BS training (T1: at the beginning; T2: after 4 or 5 weeks and T3: after 8 or 9 weeks). We expected IAc and IS to improve from a BS training compared to participants listening to an audio book (control group in study 1) or an inactive (control group in study 2).

The purpose of the study was to examine the effects of a 20-min daily practiced BS on interoceptive processes. Congruent with other studies, showing mixed results regarding different interoceptive processes, we expected a higher improvement of these processes after an 8-week BS mindfulness training. Because of the mixed results in previous studies (e.g., Parkin et al., 2014; Bornemann et al., 2015; de Jong et al., 2016) and the use of only short trainings, we hypothesized that an 8-week practiced BS will show effects. This change of interoceptive processes is important because of the disturbed interoceptive processes in clinical and subclinical populations. Therefore, we conducted two studies in order to investigate effects of an insulated BS meditation on the different interoceptive processes – IAc and IS. The effects of a Body-Scan training were tested against two different control groups. In study 1, we used an active control group (audio book). Due to the null training results in study 1, we speculated that the audio book group, especially because of the similar instruction, has a mindfulness effect, too. To prevent this, we started study 2 with an inactive control group. We examined all groups three times over a period of 8 weeks. To our knowledge, this is the first study investigating effects of an 8-week BS intervention based solely on the BS on interoceptive processes.

Our data provide consistent evidence for the influence of a BS intervention over 8 weeks on interoceptive processes. There are heterogeneous findings and statements about interoception as a trait or state variable (e.g., Schandry and Weitkunat, 1990; Ainley et al., 2012, 2013; Herbert et al., 2012). However there is first evidence, that especially participants with a low interoception can benefit from such a self-focus training (Ainley et al., 2012). Patients with different mental disorders showed low interoception (e.g., Pollatos et al., 2008, 2009; Weiss et al., 2014). Consequently, more research in clinical settings is needed to precisely clarify interoceptive processes in such samples and to find interventions for the improvement of these processes.

Both studies were carried out in accordance with the recommendation of the Guideline by the Institutional Review Board of the Ulm University (Germany) with written informed consent from all subjects. All subjects gave written informed consent in accordance with the Declaration of Helsinki. The protocol was approved by the Institutional Review Board of the Ulm University (Germany).

DF, MM, and OP, designed the study, analyzed the data and interpreted them. DF collected data and wrote the first draft of the manuscript. DF, MM, and OP, have approved the final manuscript and critically discussed them.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.