Two weeks to tune in: Evaluating the effects of a short‐term body scan on interoception

Abstract

Mindfulness practices, such as the body scan, could enhance interoception. While prior research shows promise for its effects on interoceptive sensibility (IS; subjective experience of bodily cues) and accuracy (IAcc; accurate detection of bodily signals), studies often use a limited set of interoceptive variables or apply small samples. In two pre‐registered randomized trials, we examined a 2‐week daily auditory body scan intervention vs. active control (guided imagery; Study 1) or passive control (Study 2). Study 1 included N = 85 participants (M _age = 22.26, 71% women), randomized to body scan or guided imagery. Study 2 included N = 90 participants (M _age = 23.05, 80% women), randomized to body scan or passive control. IAcc was measured pre‐ and post‐intervention using the heartbeat tracking task (HTT) and heartbeat discrimination task (HDT), while IS was assessed via confidence ratings and the multidimensional assessment of interoceptive awareness (MAIA‐2). Study 1 showed significant time effects for IAcc (HTT: p < .001; HDT: p = .012), confidence ratings (p's < .001), and MAIA‐2 (p < .001). Study 2 found improvements following the body scan for IAcc (HTT: p = .050), confidence ratings (HTT: p = .006; HDT: p = .0496), and MAIA‐2 (p = .003). Findings suggest that body scan enhances interoception within two weeks, though guided imagery and similar approaches may yield comparable effects.

INTRODUCTION

Mindfulness describes a mindset of being aware of one's internal state and the environment in a non‐judgmental, open‐minded way (Kabat‐Zinn, 2003). Mindfulness practices have been discussed to benefit mental health (Galante et al., 2023; Hanley et al., 2015; Johnson et al., 2023; Keng et al., 2011; Remskar et al., 2024) with four neurobiologically based key mechanisms being involved: Attention regulation, body awareness, emotion regulation, and change in the perspective on the self (Hölzel et al., 2011). Together, these mechanisms may facilitate self‐regulation, which is a core competence for health (Cameron & Leventhal, 2003; de Ridder & de Wit, 2006). Importantly, self‐regulation seems to be related to the perception of organismic signals, commonly labeled as interoception (Rominger & Schwerdtfeger, 2023; Schultchen et al., 2019).

Interoception covers different domains, including visceroception (i.e., the perception of visceral organs, like the heart, stomach, or bladder), proprioception (i.e., the perception of posture and movement transmitted via proprioceptors), and other motivationally relevant physiological sensations such as pain and thermal or cutaneous sensations (e.g., Craig, 2002; Garfinkel et al., 2015). Interoception has been considered particularly important for health and self‐regulation (Khalsa et al., 2018). It can be subdivided into different facets, among which interoceptive sensibility (IS) refers to the subjective perception/sensing of bodily signals, interoceptive accuracy (IAcc) is the accuracy with which bodily signals can be perceived, and interoceptive awareness is a meta‐cognitive component specifying the correspondence between IS (e.g., in terms of confidence ratings) and IAcc (Garfinkel et al., 2015). Since recent approaches call for a refined conceptualization of interoception (e.g., Desmedt et al., 2025; Murphy, 2024; Schoeller et al., 2025), both IS and IAcc could be subsumed under the term interoceptive sensing. While IS can be assessed easily via self‐reports and questionnaires, such as the Multidimensional Assessment of Interoceptive Awareness (MAIA; Mehling et al., 2018) or confidence ratings to organismic detection paradigms, IAcc requires the recording of bodily signals while asking the individual about their sensation. Because of the demanding assessment of various bodily signals, research has mainly concentrated on the perception of cardiac activity (cardioception), which can be assessed with the help of an electrocardiogram (ECG; e.g., Schandry, 1981; Whitehead et al., 1977) and has been suggested to be of particular importance for emotional processing (e.g., Kindermann & Werner, 2014). It should be noted, though, that recent research questioned the comparability of measures between and within the domains (for non‐coherent associations between interoceptive domains, see Crucianelli et al., 2022; Schoeller et al., 2025; for heterogeneity within the cardioception domain, see Hickman et al., 2020; Murphy, 2025; for heterogeneity of questionnaires assessing IS, see Desmedt et al., 2022; Ferentzi et al., 2018; Todd et al., 2022; Vig et al., 2022). Hence, although in the following, we refer to IAcc and IS as two major domains of interoception, it should be noted that both imply quite heterogeneous operationalizations. Therefore, we decided to specify the exact measure of IAcc and IS, respectively, when referring to studies below.

Importantly, being aware of raising thoughts and bodily signals without judging them (i.e., being mindful) necessitates interoceptive processes (Fissler et al., 2016; Gibson, 2019; Todd & Aspell, 2022). In particular, mindfulness‐based stress reduction (MBSR; Kabat‐Zinn, 2013) aims to facilitate non‐judgmental awareness of bodily signals. MBSR is an eight‐week meditation program that includes sitting and walking meditations, yoga, body scan, and breathing meditation as intensive practices. Among those practices, the body scan has gained particular interest in interoception research (Gibson, 2019). During the body scan, a person systematically directs their attention throughout the entire body, consciously perceiving all sensations without judging or evaluating them. The ultimate goal of the body scan is to connect with bodily sensations without judgment.

Several studies aimed to increase interoception via mindfulness practices and the body scan, in particular, with quite heterogeneous findings (Bornemann et al., 2014; Bornemann & Singer, 2017; Fischer et al., 2017; Lima‐Araujo et al., 2022; Parkin et al., 2014; Schroter et al., 2023; for a recent meta‐analysis revealing an overall small and quite fragile effect, see Schwartz et al., 2025). While some of the studies examined long‐term effects across 3 to 9 months (Bornemann et al., 2014; Bornemann & Singer, 2017) and found beneficial effects on self‐reported interoception (IS via the MAIA) and IAcc (using a cardioception paradigm), the latter indicated by a small‐to‐medium‐sized effect after 6 and 9 months), other research applied shorter time frames from a single 20‐minute session (Schroter et al., 2023) up to 8 weeks (Fischer et al., 2017; Parkin et al., 2014). Notably, while short‐term effects appeared more fragile (for effects on IS [MAIA] after 3 days of training, see Lima‐Araujo et al., 2022; no effects for both IS [confidence ratings] and IAcc [cardioception] after 2 weeks: Parkin et al., 2014), studies across 8 weeks found some effects on both IS (confidence ratings; Parkin et al., 2014) and IAcc (cardioception; Fischer et al., 2017), when compared to passive control groups.

It is important to note that most of the above‐mentioned research on short‐term effects suffered from comparably moderate sample sizes (usually N < 30) precluding detection of small to medium effect sizes. Another important factor to consider is the heterogeneity of the body scan intervention per se. For example, Bornemann et al. (2014) and Bornemann and Singer (2017) strived for quite broad mindfulness interventions and applied the body scan together with breath meditation, and added walking meditation and sound meditation on a weekly basis. Lima‐Araujo et al. (2022) used a 30‐minute body scan accompanied by breath awareness exercises. Conversely, Fischer et al. (2017), Parkin et al. (2014), and Schroter et al. (2023) focused exclusively on the body scan, which was presented via an audiotape for 15 to 20 minutes. Furthermore, the general pattern of findings seems to suggest that when compared to active control conditions (like attending to sounds; e.g. Parkin et al., 2014) or audiobook conditions (Fischer et al., 2017; Lima‐Araujo et al., 2022; Schroter et al., 2023), body scan practice might not prove superior for increasing interoception.

Even more, the assessment of interoception appears limited in previous research. Specifically, while IS was either assessed via confidence ratings to cardiac interoception tasks (Fischer et al., 2017; Parkin et al., 2014; Schroter et al., 2023) or the MAIA (Bornemann et al., 2014; Lima‐Araujo et al., 2022; Schroter et al., 2023), IAcc was solely assessed using the heartbeat tracking task (HTT; Schandry, 1981), which asks participants to count their heartbeats during different time intervals without taking their pulse or manipulating breath, etc. Of note, the HTT has been criticized due to questionable validity (e.g., Brener & Ring, 2016; Corneille et al., 2020; Jones, 1994; Ring & Brener, 2018; Ring et al., 2015; Zamariola et al., 2018; but see Ainley et al., 2020; Zimprich et al., 2020). The main points of criticism pertain to guessing strategies independent of interoception (time estimation or knowledge about the own heart rate; e.g., Desmedt et al., 2020), and individual differences in decision threshold (Zamariola et al., 2018). As an alternative, the heartbeat discrimination task (HDT) has been proposed (Whitehead et al., 1977). This task requires participants to judge whether an acoustic signal is synchronous or asynchronous to individual heartbeats. Because it is more difficult, complex (i.e., shifting attention between internal and external cues) and methodologically more demanding, the HDT has been applied less frequently in interoception research.

Hence, the aim of the present research was to evaluate whether a 2‐week body scan is sufficient to facilitate different indicators of interoception (IAcc and IS) using two paradigms (HTT, HDT) with sufficiently powered sample sizes. We applied a two‐study approach with Study 1 comparing the effects of a body scan to an active control intervention (guided imagery) and Study 2 comparing the effects to a passive control group. Both studies were pre‐registered, thus enhancing confidence in the findings. It was expected that a 2‐week body scan leads to improved IAcc and IS when compared to an active control (Study 1) and a passive control condition (Study 2).

STUDY 1

Methods

Participants

We conducted an a priori power analysis for a repeated measures ANOVA with a within‐between interaction in G*Power version 3.1.9.7 (Faul et al., 2007). Our goal was to obtain .80 power to detect an effect size of η ²  = .027 (interaction of time and group for IAcc in Study 1; Fischer et al., 2017) at the standard .05 alpha error probability, resulting in a required sample size of 74 individuals. Moreover, assuming a medium‐sized effect, a power analysis for a t‐test for paired samples at the standard .05 alpha error probability and a statistical power of .80 resulted in a sample of N = 34, suggesting a total sample size of at least N = 68.

A total of 127 individuals fully completed the online survey, which assessed exclusion criteria and collected sociodemographic data. Of these, 45 did not meet the participation requirements (i.e., diagnosed psychiatric or cardiovascular condition or psychoactive or cardiovascular medication, practicing more than 60 min/week body‐focused relaxation techniques) or did not respond to the invitation for the first laboratory session. Two participants did not attend the second laboratory session. The final sample consisted of 80 participants (57 female, 21 male, 2 non‐binary) aged between 18 and 47 years (M = 22.26, SD = 3.74). All participants were students. A total of 26 participants engaged in a body‐focused activity, most commonly yoga (n = 19). Nine participants practiced meditation, one regularly performed mindfulness exercises, and another practiced breathing meditation. One participant reported regularly engaging in another unspecified activity. Participants completed the intervention 12 times on average (M = 12.18, SD = 4.09) during the two weeks study duration. The study was approved by the local ethics committee (GZ. 39/65/63 ex 2021/22), and written informed consent was provided. The study was pre‐registered at OSF: https://osf.io/xk3av). A CONSORT flow‐chart is depicted in Figure 1A in the Appendix).

FIGURE 1.

Study 1. Top line: Trajectories for interoceptive accuracy (IAcc) for the heartbeat tracking task (HTT; A) and the heartbeat discrimination task (HDT; B). Confidence ratings are depicted at the bottom line (HTT: C; HDT; D). Solid lines depict the body scan group and dashed lines the control group (guided imagery). Whiskers indicate ± 1 SE.

Interventions

Following pre‐assessment at the laboratory, participants were randomized to either the body scan or the guided imagery intervention via a random number generator.

Body Scan

For the guided body scan, an audio file was provided. Additionally, heartbeat perception was incorporated, meaning that participants were instructed to perceive their own heartbeat without evaluating it (see Supplementary Material). The audio file was made available via a QR code link, through which participants had to enter their personal participant code using the LimeSurvey tool. They were then redirected to the audio file, which was hosted on a server of the university. The guided body scan audio file lasted approximately 18 minutes. For the audio file, a non‐commercially available calm background music track was used.

Guided imagery

Guided imagery can be defined as a technique for easily visualizing imaginary places, objects, or events to influence physiological and psychological states (Kohls et al., 2019; Utay & Miller, 2006). Research has shown that guided imagery can enhance both physiological and psychological relaxation in comparison to an inactive control group, with effects being similar to those of progressive muscle relaxation or breathing meditation (Toussaint et al., 2021). To the best of our knowledge, no study to date has directly compared guided imagery with a body scan intervention in terms of interoception.

For the guided imagery exercise, an adapted script by Kohls et al. (2019) was used (see Supplementary Material). During the guided imagery exercise, participants were instructed to take a walk on a fictional island in the sea. The script was supplemented with sensory stimuli (touch, sight, smell, sound). As with the guided body scan, the audio file was made available via a QR code link, and the same calm background music track as for the Body Scan was used. The audio file lasted approximately 16 minutes.

Variables and Instruments

For the assessment of IAcc, an ECG measurement was conducted using the BioPac MP150 amplifier system with a 3‐lead setup. The R‐waves of the ECG were identified using the Accusync172 ECG Trigger Monitor, which sent triggers to the computer corresponding to the R‐waves. Auditory stimuli were presented to the study participants via a loudspeaker located approximately two meters away. Study participants were tested in a separate, light‐protected, and quiet room while seated comfortably in a chair about one meter away from the screen on which the questionnaires and interoception tasks were displayed.

Heartbeat Tracking Task (HTT)

At the beginning of the HTT, participants were instructed to silently count each heartbeat they could perceive (Schandry, 1981). Across seven trials with randomized time intervals (20, 25, 35, 45, 55, 65, and 75 seconds), participants' heartbeats were recorded using ECG measurement. The start and end of each time interval were marked by an acoustic signal. After each time interval, participants were asked to report the number of perceived heartbeats and rate their confidence in their response on a scale from 0 (completely guessed, no heartbeat perception) to 100 (completely certain, full heartbeat perception). We applied the adapted instructions suggested by Desmedt et al. (2018) to count only those heartbeats that were actually perceived. During the task, participants were not allowed to manually check their pulse or hold their breath. Compliance with the instructions was monitored via two cameras.

Heartbeat Discrimination Task (HDT)

A total of 40 trials were presented, each consisting of 10 tones presented at 440 Hz for 50 ms, either synchronous or asynchronous with the participants' heartbeats (Whitehead et al., 1977). A randomized set of 20 asynchronous and 20 synchronous trials was used. The synchronous tones were presented with either a minimal delay of 230 ms, and the asynchronous tones with a longer delay of 540 ms (see Rominger et al., 2021). After each trial (i.e., 10 heartbeats), participants were asked whether they perceived their heartbeats as synchronous or asynchronous with the tones they heard. As in the HTT (Schandry, 1981), they also should rate their confidence on a scale from 0 (completely guessed, no heartbeat perception) to 100 (completely certain, full heartbeat perception). Again, participants were not allowed to manually check their pulse or hold their breath during the task.

MAIA‐2

The German version (Bornemann et al., 2014) of the Multidimensional Assessment of Interoceptive Awareness, Version 2 (MAIA‐2; Mehling et al., 2018) was used. The questionnaire is a widely used self‐report instrument for measuring subjectively perceived interoceptive ability. It consists of eight scales: Noticing, Not‐Distracting, Not‐Worrying, Attention Regulation, Emotional Awareness, Self‐Regulation, Body Listening, and Trust. Using 37 items, participants rate how often these experiences generally apply to them in daily life on a 6‐point Likert scale ranging from 0 (never) to 5 (always). In the present sample, internal consistencies for the total scale were McDonald's ω = .89 at both pre‐ and post‐assessment. For the subscales, reliabilities varied at pretest from ω = .62 (Noticing) to ω = .87 (Trust), and at posttest from ω = .70 (Noticing) to ω = .87 (Trust).

Study Procedure and Laboratory Testing

Participants first received information about the study and its procedure via the LimeSurvey tool, along with the informed consent form. After providing their consent, they generated their personal participant code. They then completed a survey, which collected sociodemographic data and screened for exclusion criteria. If all participation requirements were met, they received an email with a link to schedule their appointments for the two main assessments. Two laboratory assessments took place, one before the 2‐week intervention phase and one thereafter.

Procedure of the Laboratory Assessments (Pre‐ and Post‐Intervention)

At the beginning of the first laboratory session, all participants completed a COVID‐19 questionnaire to screen for symptoms or possible contact with COVID‐19‐positive individuals and signed the informed consent form. Then, body weight and height were taken to calculate the body mass index (BMI).

Afterward, the electrodes and sensors were attached, and participants completed questionnaires, including the MAIA‐2. A 3‐minute baseline measurement followed, during which participants viewed nature images on a screen. Thereafter, the HTT was administered, followed by the HDT. The order of the tasks was fixed to prevent transfer effects from the acoustic signals of the HDT to the HTT. At the end of the first laboratory session, participants were randomly assigned (using the random number generation function of MS Excel) to one of the two intervention groups. They received the QR code for their assigned intervention and were introduced to the procedure. During the second laboratory session after the intervention phase, the COVID‐19 questionnaire was completed again at the beginning of the session to check for symptoms. The procedure was identical to the first session. At the end of the second laboratory visit, participants were debriefed about the purpose of the study. Psychology students received course credit upon request. Additionally, all participants had the opportunity to enter a raffle for a €10 voucher. Each laboratory session lasted approximately 60 to 70 minutes.

Adherence to the protocol

In order to assess whether participants adhered to the study protocol, they were asked to report the number of exercises conducted during the two weeks. On average, exercises were conducted 12 times with no significant differences between groups (body scan: M = 12.03, SD = 0.70; guided imagery: M = 12.33, SD = 0.60; t[78] = 0.33, p = .745).

Data parametrization and analysis

To assess IAcc, performance measures from the HTT (Schandry, 1981) and the HDT (Whitehead et al., 1977) were used. For the IAcc score of the HTT, the sum of actually measured heartbeats (HB_actual) was subtracted from the sum of perceived heartbeats (HB_perceived). This difference was then divided by the sum of actual heartbeats, and the resulting value was subtracted from 1. Thus, the IAcc score for the HTT was calculated using the following formula (Michal et al., 2014):

$$\text{IAcc}\left(\mathrm{HTT}\right)=1-\mid {\mathrm{\Sigma HB}}_{\text{actual}}-{\mathrm{\Sigma HB}}_{\text{perceived}}\mid /{\mathrm{\Sigma HB}}_{\text{actual}}$$

For the IAcc score of the HDT, the d‐prime (d’) parameter from signal detection theory was calculated (Michal et al., 2014). Correctly identified synchronous trials were defined as hits, while asynchronous trials that were incorrectly identified as synchronous were defined as false alarms. IAcc was calculated using the following formula (Michal et al., 2014), where the z‐values refer to the inverse standard normal distribution:

$$\text{IAcc}\left(\mathrm{HDT}\right)={\mathrm{d}}^{’}={\mathrm{z}}_{\text{hits}}-{\mathrm{z}}_{\text{false alarm}}$$

For assessing IS, confidence ratings to both the HTT and HDT were used as well as the sum scores of the MAIA‐2 at pre‐ and post‐assessment, respectively. Mixed ANOVAs were calculated with group (body scan vs. guided imagery) as a between‐subjects factor and time (pre vs. post) as within‐subject factor. Partial Eta‐squared (η _p ² ) and Cohen's d were used to quantify effect sizes. Because previous research found limited evidence for strong relationships between cardiac interoception tasks (Hickman et al., 2020), we decided for separate analyses for both the HTT and HDT. The level of significance was fixed at p < .05 (two‐tailed).

Results

To determine significant group differences at pre‐test, independent samples t‐tests and Chi ² ‐tests were conducted (see Table 1 for an overview). There were no significant differences with respect to age (p = .790), gender (p = .294), and BMI (p = .532). Regarding the confidence ratings of the HDT, participants in the body scan group showed higher values (t[78] = 2.04, p = .045, Cohen's d = 0.46). No significant group differences were found for the other interoception‐related variables (all p's > .23; see Table 1). Notably, the number of minutes per week spent with body‐focused practice (e.g., yoga, Tai Chi, Qi Gong, breath meditation, etc.) at the time of study entry did not differ significantly between groups (t[70.10] = 1.10, p = .275, d = .25).

TABLE 1.

Test for differences between groups at pre‐assessment (Study 1).

	Body scan	Guided imagery	t(78)	p	d
	M (SD	M (SD)
Age	22.15 (4.42)	22.38 (2.96)	0.27	.790	0.06
BMI (kg/m2)	22.75 (4.48)	22.20 (3.35)	0.63	.532	0.14
IAcc (HTT)	0.46 (0.26)	0.45 (0.17)	0.20	.845	0.04
IAcc (HDT)	0.31 (0.92)	0.18 (0.54)	0.77	.445	0.17
IS (HTT)	43.66 (20.76)	38.31 (19.38)	1.19	.237	0.27
IS (HDT)	48.77 (13.99)	40.7 (20.67)	2.04	.045	0.46
MAIA‐2	2.70 (0.60)	2.56 (0.55)	1.02	.310	0.23
Contemplative practice (min./week)	8.91 (15.63)	13.63 (22.16)	1.10	.275	0.25

IAcc

For the HTT, the repeated‐measures ANOVA showed a significant main effect of time (F[1,78] = 31.30, p < .001, Wilk's λ = .714, η _p ²  = .29) qualified by a large‐sized effect, suggesting increasing scores from pre to post assessment. However, there was no significant effect for group (F[1,78] = 0.01, p = .923, η _p ²  = .00) and no interaction of time and group (F[1,78] = 0.08, p = .776, Wilk's λ = .999, η _p ²  = .00).

Likewise, for the HDT there was a significant main effect for time (F[1,78] = 6.70, p = .012, Wilk's λ = = .921, η _p ²  = .08) of medium effect size, but no significant effects for group (F[1,78] = 2.18, p = .143, η _p ²  = .03) and time by group interaction (F[1,78] = 1.31, p = .256, Wilk's λ = .983, η _p ²  = .02). Taken together, results indicate that participants improved in IAcc from the first to the second assessment, irrespective of group assignment. Figure 1 depicts the group trajectories for IAcc (A: HTT, B: HDT).

Confidence ratings

The repeated‐measures ANOVA for the HTT showed a significant main effect of time (F[1,78] = 16.47, p < .001, Wilk's λ = .826, η _p ²  = .17), indicating increasing scores with a large‐sized effect. The main effect of group (F[1,78] = 1.98, p = .635, η _p ²  = .03) and the interaction of time and group (F[1,78] = 0.01, p = .943, Wilk's λ = 1.000, η _p ²  = .00) were not significant. Overall, participants showed a significant improvement in confidence ratings from pre to post assessment, regardless of group assignment (Figure 1 C).

In a similar vein, analysis of the HDT showed a significant main effect of time (F[1,78] = 18.25, p < .001, Wilk's λ = .810, η _p ²  = .19), indicating a large‐sized effect. While the main effect of group was significant (F[1,78] = 5.80, p = .018, η _p ²  = .07), the time by group interaction proved not reliable (F[1,78] = 0.29, p = .595, Wilk's λ = .996, η _p ²  = .00). Like for the HTT, participants evidenced a significant increase in confidence ratings across the two weeks irrespective of group assignment, and the body scan group showed overall higher ratings (Figure 1 D).

MAIA‐2

The 2 × 2 mixed‐design ANOVA revealed a significant main effect of time (F[1,78] = 25.53, p < .001, Wilk's λ = .753, η _p ²  = .25) indicating a large‐sized effect and a significant main effect of group (F[1,78] = 4.38, p = .040, η _p ²  = .05), qualified by a small to medium effect size. The interaction of time and group showed a trend toward statistical significance, with a small to medium effect size (F[1,78] = 3.75, p = .056, Wilk's λ = .954, η _p ²  = .05). Both intervention groups exhibited increasing values in the MAIA‐2 from pre to post assessment (body scan: t(39) = 4.64, p < .001, d = 0.73; guided imagery: t(39) = 2.37, p = .023, d = 0.38). Importantly, while groups did not differ significantly from each other at pre‐test (t(78) = 1.02, p = .310), the body scan resulted in significantly higher scores at post‐test (t(78) = 2.91, p = .005) indicated by a medium effect size (d = 0.65; Figure 2).

FIGURE 2.

Group trajectories for the MAIA‐2 total score across time. Solid lines depict the body scan group and dashed lines the control group. Whiskers indicate ± 1 SE.

Discussion of Study 1

The aim of Study 1 was to evaluate the effectiveness of a 2‐week body scan intervention relative to an active control intervention (guided imagery) on different facets of interoception. Findings were generally not consistent with a beneficial effect for the body scan intervention relative to an active control group, but instead suggested significant main effects of time across all variables. Hence, practicing either body scan or guided imagery on a daily basis across two weeks was along with an increase in both IAcc and IS. The only (fragile) evidence for the superiority of the body scan was found for the MAIA‐2 with significantly higher values at post‐assessment in individuals of the body scan relative to the guided imagery group and a trend toward an interaction of group and time.

Taken together, the findings of Study 1 did not support the hypothesis that a 2‐week body scan intervention could increase interoceptive abilities beyond guided imagery. It should be noted in this respect that this pattern of findings aligns with research showing that body scan interventions of various dosages (from a single session to 8 weeks) were not superior relative to active control conditions (Fischer et al., 2017; Lima‐Araujo et al., 2022; Parkin et al., 2014; Schroter et al., 2023). Nonetheless, the trend toward increases in the MAIA‐2 is corroborated by previous research showing that a 3‐day body scan resulted in elevated MAIA scores (Lima‐Araujo et al., 2022). Taken together, the findings cumulate to suggest that specific interventions targeting at directing attention to internal cues, like the body scan, do not reliably increase interoception over and above other psychological activities such as guided imagery.

Hence, the evidence that body scan relative to active control interventions results in improved interoception is fragile to date, thus questioning the specificity of the effects. The general improvement across time in Study 1 might either suggest that active control interventions might be equally useful to facilitate interoception or that mere order (e.g., training) effects were evident. The latter interpretation would suggest that individuals might train their skills in interoception with repeated assessment, irrespective of any treatment applied in between. In order to verify this reasoning, a direct comparison with a passive control group seems mandatory. Precisely, if interoception improves only in the body scan group, but not in a group of individuals without any treatment (passive control), the effect could be ascribed to the intervention. Finding a main effect of time again, without a significant interaction of group and time would favor the training effect hypothesis. Such a result would strongly argue against the positive effects of body scan interventions for interoception.

Therefore, Study 2 aimed to examine if the very same 2‐week body scan intervention as applied in Study 1 is beneficial for interoception when compared to a passive control condition. In this pre‐registered trial, we hypothesized that the body scan results in higher IAcc and IS (i.e., confidence ratings and MAIA‐2) than a passive control condition.

STUDY 2

Methods

Participants

Based on Study 1, we opted to retain a small to medium‐sized interaction effect (f ~ .15) with a power of .80 at p < .05 (two‐tailed), resulting in a recommended sample size of N = 90 according to G*Power (Faul et al., 2007). The online pre‐screening was fully completed by a total of 122 individuals, of whom 92 participated in the laboratory assessments. Exclusion criteria were diagnosis of mental or cardiovascular disorders, drugs affecting the cardiovascular system, more than 60 min/week of body‐focused relaxation techniques, pregnancy, or lactation. One participant had to be excluded due to a cardiac arrhythmia, potentially biasing IAcc. Thus, a final sample of 91 participants completed the entire study, including the online pre‐screening and both laboratory sessions. The study included 73 women and 18 men. The intervention group (n = 45) consisted of 37 women and eight men, while the control group (n = 46) included 36 women and ten men. Participants ranged in age from 18 to 43 years (M = 23.05, SD = 3.84), and n = 29 (31.9%) reported having a university degree, 59 (64.8%) had completed high school, and three (3.3%) had a secondary school diploma as their highest level of education. Seventy‐five participants indicated that they were primarily students. A CONSORT flow chart is depicted in Figure 2A in the Supplementary Materials.

Intervention

Similar to Study 1, participants were randomly assigned to either the body scan or the control group. The study was pre‐registered at OSF (https://doi.org/10.17605/OSF.IO/D2HCK) and approved by the local ethics committee (GZ. 39/81/63 ex 2023/24). Participants gave written informed consent prior to study enrollment. The same body scan intervention and protocol were used as in Study 1. The control group did not receive any treatment, but the same measures as those for the body scan group were applied at pre‐ and post‐test.

Variables and Instruments

The same measures as in Study 1 were used. Reliabilities for the MAIA‐2 were comparable to Study 1. Specifically, for the total scale McDonald's ω was .88 at pre‐assessment and .89 at post‐assessment, indicating reliable assessment. Reliabilities for the subscales ranged between .47 (noticing at pre assessment) and .90 (trust at post assessment).

Adherence to the protocol

On average, participants of the intervention group practiced the body scan 11.53 (SD = 2.23) times, which was comparable to Study 1; however, with a larger SD.

Results

There were no group differences in age (p = .851), BMI (p = .708), and gender (p = .635). Furthermore, groups did not differ with respect to IAcc (HTT: p = .349; HDT: p = .444), confidence ratings (HTT: p = .082; HDT: p = .318), and the MAIA‐2 (p = .511) at pre‐test. Groups did not differ in the duration of contemplative practices per week (Table 2).

TABLE 2.

Test for differences between groups at pre‐assessment (Study 2).

	Body scan	Control	t(89)	p	d
	M (SD	M (SD)
Age	22.98 (4.51)	23.13 (3.10)	0.19	.851	0.04
BMI (kg/m2)	22.46 (4.02)	22.19 (2.70)	0.38	.708	0.08
IAcc (HTT)	0.46 (0.25)	0.51 (0.24)	0.94	.349	0.20
IAcc (HDT)	0.30 (0.60)	0.41 (0.68)	0.77	.444	0.16
IS (HTT)	40.91 (21.44)	48.26 (18.33)	1.76	.082	0.37
IS (HDT)	44.50 (15.86)	47.98 (17.11)	1.00	.318	0.21
MAIA‐2	2.81 (0.55)	2.89 (0.56)	0.66	.511	0.14
Contemplative practice (min./week)	1.78 (5.24)	2.07 (9.81)	0.17	.862	0.04

IAcc

The ANOVA for IAcc revealed a marginally significant and small to medium‐sized time by group interaction (F[1, 89] = 3.93, p = .050, Wilk's λ = .958, η _p ²  = .04) for the HTT (Figure 3 A) with no significant main effects for time (F[1, 89] = 2.41, p = .124, Wilk's λ = .974, η _p ²  = .03) and group (F[1, 89] = 0.09, p = .771, η _p ²  = .00), respectively. Post hoc t‐test indicated that while the control group did not differ from pre to post assessment (t[45] = −0.29, p = .771), the body scan group showed a significant increase with a small to medium effect size (t[44] = 2.62, p = .012, d = 0.39). Regarding the HDT (Figure 3 B), there were no significant main effects for time (F[1, 89] = 1.66, p = .200, Wilk's λ = .982; η _p ²  = .02) and group (F[1, 89] = 0.75, p = .390; η _p ²  = .01), respectively, and no significant interaction between time and group (F[1, 89] = 0.00, p = .991, Wilk's λ = 1.000; η _p ²  = .00).

FIGURE 3.

Study 2. Top line: Trajectories for interoceptive accuracy (IAcc) for the heartbeat tracking task (HTT; A) and the heartbeat discrimination task (HDT; B). Confidence ratings are depicted at the bottom line (HTT: C; HDT; D). Solid lines depict the body scan group and dashed lines the control group (passive control). Whiskers indicate ± 1 SE.

Confidence ratings

Confidence ratings for the HTT showed a significant time by group interaction (F[1, 89] = 7.83, p = .006, Wilk's λ = .919, η _p ²  = .08; Figure 3 C) of medium effect size, while no main effects for time (F[1, 89] = 0.33, p = .566, Wilk's λ = .996; η _p ²  = .00) and group (F[1, 89] = 0.18, p = .669; η _p ²  = .00) emerged. Post hoc t‐tests revealed that while the control group showed a trend toward decreasing values (t[45] = −1.76, p = .085, d = −0.26), the body scan group evidenced a significant increase of small to medium effect size (t[44] = 2.17, p = .036, d = 0.32). In a similar vein, there was a significant time by group interaction for the HDT (F[1, 89] = 3.96, p = .0496, Wilk's λ = .957, η _p ²  = .04; Figure 3 D) as well as a significant main effect for time (F[1, 89] = 4.49, p = .037, Wilk's λ = .952, η _p ²  = .05) approaching a medium‐sized effect. Notably, the control group showed no significant change from pre to post assessment (t[45] = 0.11, p = .914), whereas the body scan group exhibited a significant increase indicated by a small to medium effect size (t[44] = 2.54, p = .015, d = 0.37).

MAIA‐2

Analysis of the MAIA‐2 revealed a significant main effect of time (F[1, 89] = 9.66, p = .003, Wilk's λ = .902; η _p ²  = .10), which was further qualified by a significant, medium‐sized time by group interaction (F[1, 89] = 9.15, p = .003, Wilk's λ = .907, η _p ²  = .09), depicted in Figure 4. Post hoc t‐tests revealed that while the control group showed no change from pre to post assessment (t[45] = 0.06, p = .951), the body scan group exhibited a significant increase (t[44] = 4.14, p < .001) of medium effect size (d = 0.62). The main effect for group was not significant (F[1, 89] = 0.01, p = .941).

FIGURE 4.

Group trajectories for the MAIA‐2 total score across time. Solid lines depict the body scan group and dashed lines the control group. Whiskers indicate ± 1 SE.

GENERAL DISCUSSION

The aim of Study 2 was to extend the findings of Study 1 by examining whether a 2‐week body scan intervention on a daily basis impacts facets of interoception when compared to a passive control group. In agreement with Study 1, which indicated general increases in interoception independent of the intervention applied, Study 2 found (marginally significant) improvements for IAcc (HTT), significant effects for confidence ratings for both cardiac interoception paradigms and the MAIA‐2. However, while Study 1 could not confirm unique effects of the body scan and thus raised concerns that mere training effects were responsible for the generally improved interoception performance at post assessment, Study 2 could rule out this explanation by showing the expected effects for the body scan group (with the exception of IAcc during the HDT) and no training effects within the passive control group. Hence, Study 2 confirmed most of the preregistered hypotheses that practicing the body scan daily across two weeks results in higher IAcc and IS (i.e., confidence ratings and MAIA‐2) than a passive control condition.

It should be emphasized that findings are generally in line with previous research that applied interventions across longer time periods (Bornemann et al., 2014; Bornemann & Singer, 2017; Fischer et al., 2017) comparing body scan or more comprehensive mindful interventions to passive control conditions. However, effect sizes appear small to medium in size, with larger effects found for subjective indicators of interoception (i.e., IS). The present research could substantiate and extend previous studies' findings by showing that effects were evident across a broad range of measures, with the only exception for IAcc in the HDT. Notably, we applied the HDT because recent critique questioned the validity of the HTT (e.g., Brener & Ring, 2016; Corneille et al., 2020; Jones, 1994; Ring et al., 2015; Ring & Brener, 2018; Zamariola et al., 2018). Direct comparisons between paradigms (Figure 1, Figure 3) suggest generally poorer performance on the HDT as compared to the HTT, which has also been reported by others (Eichler & Katkin, 1994; Khalsa et al., 2009; Knapp‐Kline & Kline, 2005; Parkin et al., 2014). Due to the more blurred assessment of interoception involving somatosensory integration and higher cognitive load (Knapp et al., 1997; A. Schulz et al., 2013; Suzuki et al., 2013), the HDT might not have proved sufficiently sensitive to track changes in IAcc in both studies.

Importantly, the present research as well as previous studies also suggest that when compared to active control conditions, the body scan shows no superiority for increasing interoception (Fischer et al., 2017; Parkin et al., 2014). Although this finding is surprising at first sight, it suggests that improvements in interoception might not be exclusively tied to mindfulness‐based practices, as other contemplative activities also appear to facilitate attention to bodily cues equally well. Thus, any practice involving focused internal (such as guided imagery) or external attention (listening to auditory stimuli; e.g., Fischer et al., 2017; Lima‐Araujo et al., 2022; Parkin et al., 2014) may be similarly effective. This aligns with broader theories of neuroplasticity, where directing attention to bodily states—regardless of the specific method—may enhance interoceptive processing (Gibson, 2024). Specifically, regular contemplation may reinforce neural pathways related to interoception, leading to a more efficient and refined internal representation of bodily states (Farb et al., 2015; Gibson, 2024; Lazzarelli et al., 2024; M. Schulz, 2016). In this respect, meditation, body awareness training, and even imagery could shape interoceptive networks (e.g., anterior insula, ACC, somatosensory cortex; Lazzarelli et al., 2024; Menon & Uddin, 2010; Quadt et al., 2018; Silvanto & Nagai, 2025). Through experience‐dependent plasticity, focused attention may reorganize brain circuits, making interoceptive signals more salient and integrated into conscious experience (Farb et al., 2015). Importantly, predictive coding models of interoception have been introduced (Ainley et al., 2016; Petzschner et al., 2021) suggesting that contemplative practices may provide an interoceptive simulation map that facilitates perception of bodily cues (Farb et al., 2015). It should be noted in this respect that, besides the interventions applied in both studies, participants in Study 1 evidenced more expertise in contemplative practices than in Study 2 (i.e., more hours spent with other contemplative practices per week), thus possibly undermining the contrast between the body scan and guided imagery groups in Study 1.

Limitations and Outlook

Although applying comparably large sample sizes in two independent studies and using diverse methods to operationalize interoception, the present research has limitations that need to be acknowledged. First, we decided to conduct two separate studies instead of one large study with three separate conditions. We opted for the two study approach because of time economic reasons and the replication of the approach – we applied the body scan intervention in two independent samples. Nonetheless, future studies should apply more complex factorial intervention designs to verify – or falsify – our findings. Second, it should be noted that placebo‐like or general engagement effects might have played a role—meaning that just doing an activity that involves bodily processes or attention may be enough to boost interoception. Hence, future studies should aim to evaluate whether contemplative practices increase interoception beyond mere placebo effects. Therefore, various additional types of interventions such as slow‐paced breathing should be tested against the body scan (and guided imagery) to further specify the ingredients of the observed enhancing intervention effects. Third, although we consider it a strength of this research to have covered a wider range of interoceptive variables in both studies, thus striving for a rather comprehensive assessment of interoception, the applied measures are not without limitations. As outlined previously, interoception may widely differ across domains (e.g., Crucianelli et al., 2022; Murphy, 2025; Schoeller et al., 2025), and solely focusing on cardioception for quantifying IAcc falls short. In this respect, we also need to mention that we applied the Whitehead paradigm with only two fixed intervals, which has been criticized (for a review, see Körmendi & Ferentzi, 2024). Hence, future research is advised to strive for other physiological systems to assess interoceptive accuracy, like respiration (e.g., Nikolova et al., 2022), other measures to quantify IS (e.g., Desmedt et al., 2022), and adding novel paradigms to assess cardioception (e.g., Plans et al., 2021). Finally, we want to question whether focusing solely on IS and IAcc is an ample goal in interoception research. The valence of interoceptive processing (i.e., interoceptive interpretation as suggested by Desmedt et al., 2025) seems to be key in mindfulness practices, but has rarely been considered in previous research (see Feldman et al., 2024; MacVittie et al., 2024). Thus, not only strengthening the perception of organismic cues, but also modifying the valence of bodily cues (from negative to positive) should be of utmost interest in future intervention research.

Conclusions

Two independent studies on the effects of a 2‐week body scan intervention revealed that daily practice of the body scan across two weeks resulted in elevated interoception, both with respect to indicators of IS and cardiac IAcc. Notably, the beneficial effects were evident when comparing the intervention group to a passive control group, but did not reach effects beyond another active condition (guided imagery). Results suggest that practicing contemplative training daily for about 15 minutes may indeed facilitate the perception of bodily cues, which could ultimately benefit self‐regulation and health.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

ETHICS STATEMENT

Written permissions were obtained from the ethics committee of the University of Graz (GZ. 39/65/63 ex 2021/22 and GZ. 39/81/63 ex 2023/24). The ethical principles of the current Declaration of Helsinki were followed throughout the study.

Supporting information

Data S1. Body Scan Transcript

Data S2. Guided Imagery Transcript

Data S3. CONSORT 2010 Flow Diagram

Data S4. CONSORT 2010 Flow Diagram

Schwerdtfeger, A. R. , Weber, B. , & Rominger, C. (2025). Two weeks to tune in: Evaluating the effects of a short‐term body scan on interoception. Applied Psychology: Health and Well‐Being, 17(5), e70073. 10.1111/aphw.70073

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author, upon reasonable request.