Does Interoceptive Sensibility Mediate the Relationship Between Emotional Distress and Pain Severity? A Cross-Sectional Study in Chronic Pain and Pain-Free Individuals.

Or Katrieli: conceptualization, methodology, investigation, data curation, formal analysis, writing – original draft, writing – review and editing. Jemma Silvert: conceptualization, methodology, writing – review and editing. Yaron Haviv: conceptualization, methodology, project administration, resources, writing – review and editing, funding acquisition. Gadi Gilam: conceptualization, methodology, formal analysis, validation, project administration, resources, writing – original draft, writing – review and editing, funding acquisition, supervision.

This work was supported by funding from the Cabakoff Research Fund (G.G. and Y.H.) and partially by the Israeli Science Foundation (Grant 2127/24; G.G.).

For participants recruited from the community, all procedures were approved by the Hebrew University of Jerusalem Institutional Review Board (protocol #12092023). For participants recruited from the Orofacial Pain Clinic at Hadassah Medical Center, all procedures were approved by the Hadassah Medical Organization Helsinki Committee (protocol #HMO‐0470‐23). All participants electronically signed an informed consent as part of an online survey administered through Qualtrics ( https://www.qualtrics.com/ ). A convenience sample was obtained by distributing the survey through social media outlets and word of mouth for the community sample (between November 2023 and February 2024) and to past and current patients of the orofacial pain clinic (between January 2024 and December 2024). Following informed consent, participants completed demographic, general medical and pain‐related background information, and all other measures were subsequently administered in a randomized order per each participant. A raffle was conducted at the end of data collection amongst all participants who completed the survey in its entirety, awarding prizes of 200 ILS (about US $60) to each of five participants from the community‐based sample, and five from the clinic sample. The research plan, including design, aims, hypotheses and planned statistical analyses, was preregistered on the Open Science Framework website on November 27th, 2023 ( https://osf.io/maqpy/ ). Deviations from preregistration are noted throughout the methods section. Patients and stakeholders were not involved in any aspects of this study beyond as participants. A preregistered power analysis was conducted to calculate the targeted sample size (Hulley et al.  2001 ) of a chronic pain group, informed by previous findings of the correlation between interoceptive sensibility and pain intensity (Colgan et al.  2022 ; Valenzuela‐Moguillansky et al.  2017 ). A Pearson's r  = 0.25 and an α (2‐sided) of 0.05 required a sample size of n  = 141 for power of 0.85. A sample size of n  = 150 was thus aimed for and extended to guide the target sample size for each of three groups. Based on the below detailed final sample, the COFP group was underpowered to detect this correlation, yet power analysis for detecting a medium effect size of 0.25 using ANOVA and ANCOVA for group comparisons (Faul et al.  2009 ) was 99.9% and 97.7%, respectively. We included participants 18 years of age or above, living in Israel with a minimum intermediate level of Hebrew proficiency. The CP and COFP groups further required a pain intensity composite score (see below) equal to or larger than three (Horn‐Hofmann et al.  2018 ; Kimura et al.  2015 ; Simanski et al.  2014 ; Zebis et al.  2011 ), pain as an ongoing problem for at least 3 months and pain experienced in at least half of the days in the past 6 months (Deyo et al.  2015 ). Unlike our preregistered plan, we separated between the CP and COFP groups because they were recruited from different sources, a broad community‐dwelling population and a clinical population of treatment seeking patients; and they had different timings of data collection, which is particularly relevant to the macro‐environmental context of the Hamas‐Israel conflict. Thus, to prevent potential sampling frame, timing and etiological biases between the CP and COFP groups, participants recruited from the community who reported that their primary CP condition was orofacial were excluded. The PFC group required a pain intensity composite score lower than three, not currently suffering from any pain condition, and no current diagnosis of other medical or mental‐health conditions, especially those with somatic presentations or strongly comorbid with pain (e.g., cardio‐vascular disease, anxiety disorders). To ensure participants attention and concentration throughout the survey, three item‐checks were included in three different questionnaires (e.g., ‘Control question for your attention! This item tests your concentration, please choose the response neither agree nor disagree ’). Participants were excluded if they failed two or more of these item‐checks (Mor‐Elzas et al.  2026 ) or if they did not complete the primary measures, as detailed below. A total of 1462 and 490 entries were registered in the community‐based and orofacial clinic surveys, respectively. Of these, 670 and 290 were excluded from these two surveys, respectively, since they did not complete the primary measures; 24 and 70 were excluded since they failed attention checks; and 2 were excluded due to their age, Hebrew level or not living in Israel. The remaining 766 and 128 of the community‐based and orofacial clinic surveys, respectively, included 400 for the CP and 128 for COFP groups, from which 161 and 39 for each of these two groups were excluded due to not meeting the pain criteria. In addition, 36 were excluded from the CP group since their primary condition was orofacial and would potentially confound the differentiation between the two pain groups. For the remaining 366 from the community‐based survey, reflecting the PFC group, 148 were excluded for pain criteria, and 48 for having a medical or mental health condition. The final sample included n  = 203 for the CP group, n  = 89 for the COFP group and n  = 170 for the PFC group. Demographic information included age, sex, years of education, employment, marital status and place of residence during the Hamas‐Israel war outbreak, which was scaled to a number reflecting proximal (1) to distal (4) distance from the war zones (Mor‐Elzas et al.  2026 ). These numbers indicated a combination of the proximity to the war front, the time to get to shelter in case of missile attacks, and the actual amount of missile attacks since the beginning of the war, all in accordance with Israel's Home Front Command guidelines ( https://www.oref.org.il/ ; see Figure  S1 for illustration). Three additional questions inquired whether pain, mental health and physical health changed in comparison to before the war, with participants responding that it either worsened, did not change or improved, scored −1, 0 and +1, respectively. These three scores were averaged to a composite score indicating the self‐reported impact of war on health. Cronbach's alpha internal consistency for these three items was 0.72. Clinical background for the pain groups included pain duration, pain frequency in the past 6 months, pattern of pain attacks and the primary pain condition. Participants also reported the presence of non‐pain medical and mental‐health conditions, as well as any pain and/or non‐pain medications. Pain severity measures included a composite score of pain intensity , calculated by averaging three self‐reported measures: current pain, average pain and worst pain in the past 7 days (Chen et al.  2012 ). These three items were assessed using a gold standard 0 (no pain) to 10 (worst pain imaginable) numerical rating scale (Dworkin et al.  2005 ; Safikhani et al.  2018 ). Cronbach's alpha for these three items was 0.96. The National Institute of Health's Patient‐Reported Outcomes Measurement Information System (PROMIS; Broderick et al.  2013 ) measure for pain interference was a secondary measure of pain severity. This measure assesses the extent to which pain interferes with daily life activities (e.g., ‘How much did pain interfere with your household chores?’) in the past 7 days, using eight items scored on a scale from 1 (not at all) to 5 (very much). Ratings are summarized for a total score ranging between 8 and 40. Cronbach's alpha was 0.97. Emotional distress was assessed as a composite score of three negative‐affect related factors, namely depression , anxiety and anger , all assessed using PROMIS measures (Cook et al.  2016 ; Pilkonis et al.  2011 ). Each measure assesses the frequency of feelings and behaviours associated with the factor (e.g., ‘I felt sad’, ‘I felt worried’ and ‘I felt annoyed’) in the past 7 days, using eight items scored on a scale from 1 (never) to 5 (always). Ratings are summarized for a total score ranging between 8 and 40. Cronbach's alpha for this sample was 0.95, 0.93 and 0.94, for depression, anxiety and anger, respectively, and 0.86 for the composite score. Interoceptive sensibility measures included the Multidimensional Assessment of Interoceptive Awareness (MAIA), version 2 (Mehling et al.  2018 ), which comprises 37 items pertaining to eight subscales. Participants are asked to indicate the frequency by which each item generally applies to them on a scale of 0 (never) to 5 (always). Ratings are averaged for a total score ranging between 0 and 5, that had a Cronbach's alpha of 0.87, as well as for each subscale separately: Noticing refers to being aware of body sensations (four items, e.g., ‘I notice when I am uncomfortable in my body’) and had a Cronbach's alpha of 0.83; Not‐distracting refers to the tendency to ignore unpleasant sensations (six items, e.g., ‘I distract myself from sensations of discomfort’) and had a Cronbach's alpha of 0.79; Not‐worrying refers to the tendency to not become distressed by painful or uncomfortable sensations (five items, e.g., ‘When I feel physical pain, I become upset’) and had a Cronbach's alpha of 0.79; Attention regulation refers to the ability to sustain and control attention to body sensations (seven items, e.g., ‘I can return awareness to my body if I am distracted’) and had a Cronbach's alpha of 0.85; Emotional awareness refers to being aware of the connection between body sensations and emotional states (five items, e.g., ‘I notice how my body changes when I am angry’) and had a Cronbach's alpha of 0.82; Self‐regulation refers to the ability to regulate psychological distress by attending body sensations (four items, e.g., ‘I can use my breath to reduce tension’) and had a Cronbach's alpha of 0.82; Body listening refers to actively attending the body for insight (three items, e.g., ‘I listen to my body to inform me about what to do’) and had a Cronbach's alpha of 0.85; and Trust refers to experiencing one's body as safe (three items, e.g., ‘I am at home in my body’) and had a Cronbach's alpha of 0.84. The very short form of the Body Perception Questionnaire (BPQ; Cabrera et al.  2018 ; Porges  1993 ) was a secondary measure of interoceptive sensibility. Participants are asked to indicate how often they are generally aware of 12 bodily experiences (e.g., ‘My mouth being dry’) in day‐to‐day life, on a scale of 1 (never) to 5 (always). Ratings are summarized for a total score ranging between 12 and 60. Cronbach's alpha was 0.92. Data processing and analyses were conducted using IBM SPSS version 30. Descriptive information for all measures was summarized and presented for each of the three groups. MAIA‐2 data was calculated as a total score and separately for each sub‐scale. Comparisons between groups were tested using parametric ( t ‐test/ANOVA) and/or non‐parametric (Chi 2 ) statistical tests as appropriate to the measure. Since the study occurred about 5 weeks after the Hamas‐Israel war erupted on October 7th, 2023, the measurement of proximity to the war zone and the self‐reported health‐related war‐dependent changes were used to control for potential war‐related effects by adding them as covariates in an additional ANCOVA comparing all primary and secondary measures between groups. Demographic factors that were statistically different between groups were further added as covariates to this ANCOVA. Next, Pearson correlation coefficients were used to test the associations between measures. The coefficients were calculated and presented separately for each group, and where relevant, statistically compared using Fisher's Z test. Generally, all tests were two‐sided and a p  ≤ 0.05 was considered as an initial statistical threshold for potential significance, followed by Bonferroni correction for multiple comparisons. Finally, a bootstrapped‐based mediation analysis was conducted using PROCESS (model 4; Hayes  2012 ), with bias‐corrected 95% confidence intervals and 10,000 bootstrap samples, to test within each relevant group whether interoceptive sensibility mediated the relationship between pain severity and emotional distress. Specifically, the primary and preregistered aim was to test mediation based on the MAIA‐2 sub‐scales that would reveal significant correlations with both emotional distress and pain intensity. These same analytic steps were repeated to explore, as preregistered, mediation models with the secondary measures, again based on a priori significant correlations between the MAIA‐2 subscales and/or the BPQ with both emotional distress, pain intensity and pain interference. Due to the cross‐sectional nature of this study and inspired by the mutual influence of emotional distress and CP on each other, unlike the preregistration, mediation was tested across both directional specifications of the models such that pain severity measures and emotional distress were modelled as either the X or the Y variables of mediation (Figure  1A ). To note again, since we deviated from our preregistration by creating two separate pain groups for CP and COFP, all COFP‐related analyses should be considered as exploratory.

Table  1 presents the demographic characteristics across the three groups. Most of the sample were female (67.32%) with an average middle‐age (47.66 ± 14.94 years old; M ± SD), highly educated (15.33 ± 3.02 years of education), living with a partner (67.32%) and working (59.88%). About 10% were living in close proximity to the war front lines (zone 1), while about half (52%) were living in the most densely populated region of the country (zone 2), and as might be expected, the self‐reported impact of war on health was generally negative (−0.28 ± 0.40). Some group differences indicated a slightly higher female‐to‐male ratio in the COFP group, more years of education in the PFC group, more people from zone 4 in the COFP group, which can be expected since the clinic is physically located in this zone, and a larger negative impact of the war on health in the CP and COFP groups, which could be expected and serves as a validation of this assessment. Demographic characteristics according to the three groups. Note: CP indicates a significant group difference compared to the Chronic Pain group, and similarly for COFP and PFC for the Chronic Oro‐Facial Pain and Pain Free Controls, respectively. Abbreviations: COFP, Chronic Oro‐Facial Pain; CP, Chronic Pain; PFC, Pain Free Controls. Uncorrected. Bonferroni corrected. Table  2 presents the clinical pain characteristics in the two pain groups. Most participants reported long‐standing pain, with more than half experiencing pain for over 5 years (55.48%) and about a third for more than a year (31.51%), and the majority reported daily or almost daily pain during the past 6 months (69.52%). In addition, over a third indicated additional medical conditions (35.62%) and a third reported mental‐health conditions (40.41%), with more than half taking pain medications (57.19%) and more than half taking non‐pain medications (55.82%). There were no group differences in these characteristics, though, as intended, the primary pain conditions were clearly different. Musculoskeletal pain was the most prevalent in the CP group, while temporomandibular disorders and general orofacial pain were the most common conditions in the COFP group. Clinical characteristics in the two chronic pain groups. Abbreviations: BMS, Burning Mouth Syndrome; COFP, Chronic Oro‐Facial Pain; CP, Chronic Pain; CRPS, Complex Regional Pain Syndrome; PFC, Pain Free Controls; TMD, Temporomandibular Disorders. Table  3 presents a summary of descriptive information pertaining to the primary and secondary measures, separately for each group and results of the group comparison tests. As can be expected based on the application of the inclusion/exclusion criteria, the CP and COFP groups reported significantly higher pain intensity levels compared to the PFC group. Accordingly, and in line with hypotheses about the CP group and the exploratory analyses with the COFP group, both pain groups reported significantly higher levels of pain interference and emotional distress compared to the PFC group. Across these measures, there were no differences between the CP and COFP groups. In fact, this was the case for all the primary and secondary measures. Descriptive information (M ± SD) for the primary and secondary measures according to the three groups. Note: p ‐controlled refers to the p ‐value of an ANCOVA in which covariance of sex, education, proximity to the war zone and the health‐related war‐dependent changes were controlled for; CP indicates a significant group difference compared to the Chronic Pain group, and similarly for COFP and PFC for the Chronic Oro‐Facial Pain and Pain Free Controls, respectively. Abbreviations: BPQ, Body Perception Questionnaire; COFP, Chronic Oro‐Facial Pain; CP, Chronic Pain; MAIA‐2, Multidimensional Assessment of Interoceptive Awareness, version 2; PFC, Pain Free Controls. Uncorrected. Bonferroni corrected. In line with hypotheses about the CP group and the exploratory analyses with the COFP group, a differential pattern of interoceptive sensibility manifested between the pain groups and the PFC group. The pain groups reported significantly higher levels of noticing their bodily sensations, and significantly lower levels of not‐worrying about and trusting these sensations compared to the PFC group. The CP group also reported significantly higher levels of body perception compared to the PFC group. Two effects that did not meet the Bonferroni correction threshold indicated lower levels of not‐distracting and higher levels of emotional awareness in the CP compared to the PFC group. All results remained significant when controlling for sex, education, proximity to the war zone and the health‐related war‐dependent changes, to the extent that the two effects that did not meet the Bonferroni correction threshold now did. In contrast, no group differences were found for attention regulation, self‐regulation and for body listening. Together these results generally indicate that compared to people living without pain in their lives, people with persistent pain reported being more aware of their bodily signals and of the connection between these signals and their emotional states, more worried about and less able to ignore painful or discomforting signals and less trusting of their body. Table  4 presents Pearson correlation coefficients between primary and secondary measures, separately for each group. All groups showed significant medium‐to‐large positive correlation coefficients ( r  = 0.41–0.54, p  < 0.001) between pain intensity and pain interference. In partial alignment with hypotheses, emotional distress positively correlated with pain intensity across all groups with a significant medium sized coefficient ( r  = 0.30, p  < 0.001) in the PFC group, and marginal positive coefficients for the CP and COFP groups ( r  = 0.20 and 0.27, respectively, p  < 0.01 uncorrected for multiple comparisons). In contrast to hypotheses about the CP group and the exploratory analyses with the COFP group, these two groups did not demonstrate stronger coefficients than the PFC group, though this was not the case for most significant findings described below. In addition, emotional distress significantly correlated with pain interference across all groups with medium‐to‐large sized coefficients ( r  = 0.33–0.46, p  < 0.001). These results confirmed that emotional distress is generally associated with pain severity, both the sensory as well as the functional dimensions of pain. Correlation matrix according to the three groups. Note: CP indicates a significant difference in the correlation coefficients compared to the Chronic Pain group, and similarly for COFP and PFC for the Chronic Oro‐Facial Pain and Pain Free Controls, respectively. Abbreviations: COFP, Chronic Oro‐Facial Pain; CP, Chronic Pain; PFC, Pain Free Controls. p  < 0.01, uncorrected. p  < 0.001, Bonferroni corrected. Unlike hypotheses about the CP group and the exploratory analyses with the COFP group, and although in both CP and PFC groups the sample sizes were well powered to detect even the expected low effect sizes, there were no significant correlation coefficients between interoceptive sensibility measures, both primary and secondary and pain intensity. Several of the interoceptive sensibility metrics did significantly correlate with pain interference: Noticing had a significant medium‐to‐large positive coefficient ( r  = 0.39, p  < 0.001) in the COFP group, significantly larger than the non‐significant positive coefficients of this sub‐scale in the CP and PFC group ( r  = 0.14 and 0.06, respectively); Not‐worrying had significant medium‐to‐large negative coefficients in the CP and COFP groups ( r  = −0.26 and −0.41, respectively, p  < 0.001), with the PFC group having a non‐significant negative coefficient ( r  = −0.16) significantly smaller than the COFP's group; and trusting had a significant medium negative coefficient ( r  = −0.30, p  < 0.001) in the CP group, with the COFP and PFC groups having a non‐significant negative coefficient ( r  = −0.20 and −0.14, respectively). These results generally indicated that as people in the pain groups reported being more aware of their bodily signals, more distressed from painful or discomforting sensations, and that they did not experience their body as safe, they reported more pain interference. In line with hypotheses about the CP group and the exploratory analyses with the COFP group, there were several significant correlation coefficients between interoceptive sensibility measures and emotional distress: Noticing had a significant medium‐to‐large positive coefficient ( r  = 0.43, p  < 0.001) in the COFP group, significantly larger than the non‐significant coefficient in the PFC group ( r  = 0.02), not different than the marginally significant positive coefficient ( r  = 0.22, p  < 0.01 uncorrected for multiple comparisons) in the CP group; Not‐worrying had significant medium‐to‐large negative coefficients in all groups ( r  = −0.36 to −0.64, p  < 0.001), with the COFP group having significantly stronger coefficient than the CP and PFC groups; Self‐regulation had significant medium negative coefficients in the CP and PFC groups ( r  = −0.35 and −0.32, respectively, p  < 0.001) and a non‐significant negative coefficient ( r  = −0.26) in the COFP group; trusting had significant small‐to‐medium negative coefficients in all groups ( r  = −0.28 to −0.42, p  < 0.001); and the BPQ had significant small‐to‐medium positive coefficients in the CP and COFP groups ( r  = 0.27 and 0.34, respectively, p  < 0.001). These results generally indicated that as people in the pain groups reported being more aware and perceptive of their bodily signals, more distressed from painful or discomforting sensations and less able to regulate their distress by attending these sensations and that they did not experience their body as safe, they reported more emotional distress. The correlation coefficients between the two interoceptive sensibility measures are reported in Table  S1 . There were some divergences between the CP and COFP groups in the association between these two measures, specifically between the BPQ and the MAIA‐2 total score ( r  = 0.24 and −0.06, respectively, p  < 0.001) and the attention regulation subscale ( r  = 0.24 and −0.19, respectively), which might reflect features specific to a treatment‐seeking orofacial pain population or to the smaller COFP sample size. Considering the above correlation coefficients, the hypothesis related to the primary objective of this study was refuted. Interoceptive sensibility measured either with MAIA‐2 or BPQ did not mediate the relationship between pain intensity and emotional distress. However, in line with preregistration, several models were explored, based on MAIA‐2 and BPQ dimensions that correlated significantly with both pain interference and emotional distress. Namely, BPQ and not‐worrying was tested within both pain groups, while trusting was tested for the CP group and noticing was tested for the COFP group, all as potential mediators of the relationship between pain interference and emotional distress. Not‐worrying about painful or discomforting sensations significantly and partially mediated the relationship between pain interference and emotional distress in both CP and COFP groups (Figure  1B,C , respectively). BPQ only partially mediated the relationship between pain interference and emotional distress in the CP group (indirect effect a*b = 0.04, 95% CI = 0.01–0.08). Trusting in one's body as reported in the CP group (Figure  1D ) and noticing one's bodily sensations as reported in the COFP group (Figure  1E ), significantly and partially mediated the relationship between pain interference and emotional distress in both statistical directions.

This study examined whether interoceptive sensibility mediated the relationship between pain severity and emotional distress in people with CP. In line with preregistered hypotheses, findings demonstrated that people with CP reported higher levels of pain intensity, pain interference and emotional distress, as well as a differential pattern of interoceptive sensibility, compared to PFC. Specifically, people with CP reported noticing and perceiving their bodily sensations more than PFC and at the same time being more worried and having less trust in these sensations. This was so for people with mixed‐aetiology CP recruited from the community and for treatment‐seeking patients with COFP recruited from a clinic. Although these interoceptive sensibility dimensions confirmed the expected associations with emotional distress, unlike our preregistered hypothesis, no significant associations were found between them and pain intensity, although sample size for the CP and PFCs groups were well powered to detect even a small correlation. Therefore, statistical mediation of interoceptive sensibility in the link between emotional distress and pain intensity was refuted. However, exploratory associations were confirmed with pain interference, and specifically for the level of trust in and perceiving bodily sensations in the CP group, for the level of noticing these sensations in the COFP group, and for the level of worrying about distressing and painful sensations, that were found to statistically mediate the relationship between pain interference and emotional distress. Caution should be taken in regards to these findings about the treatment‐seeking COFP group since it was underpowered and deviated from our original preregistered plan to pool this group together with the mixed‐aetiology community‐dwelling CP group. Nevertheless, these findings align with previous work and extend them to indicate a potential role for interoception in mediating the relationship between pain and emotions. Moreover, it may offer novel clinical implications into potential pathways of CP‐related functional disability and emotional distress, through one's beliefs about and reported experience of their bodily sensations. These findings should be considered within the unique context of this study being conducted in the macro‐environmental context of an ongoing conflict. Previous meta‐analytic evidence indicated that CP of various etiologies was not reliably associated with differences in overall interoceptive sensibility compared to PFC (Horsburgh et al.  2024 ), especially as measured with the MAIA questionnaire (Mehling et al.  2012 , 2018 ). Rather, CP was associated with a nuanced and differential pattern of interoceptive sensibility across different dimensions, largely matching the current study's findings. In addition, results here did show heightened perception of bodily signals in people with CP as measured with the BPQ (Porges  1993 ), unlike meta‐analytic findings pointing at non‐significant effects in the same direction (Horsburgh et al.  2024 ). Together, meta‐analytic and current findings suggest that potentially dysfunctional interoceptive processing is not necessarily associated with diminished or heightened dispositional tendencies to perceive physiological and bodily signals. Rather, it is about an altered way of experiencing and interpreting these signals. For people with CP, bodily sensations appear to be more salient, more emotionally loaded and experienced as less safe. Although these findings reflect interoceptive sensibility during a sustained period of stress, requiring cautious interpretation, they align with views on dysfunctional interoception in physical and mental‐health disorders (Bonaz et al.  2021 ; Khalsa et al.  2018 ). Accordingly, CP may be characterized by a state of interoceptive saturation, in which the system is constantly exposed to heightened bodily signals, potentially biased towards distressful and painful interpretations. In such a state, increased monitoring of internal sensations does not lead to better understanding or regulation of these sensations. Instead, they may become even more salient and emotionally loaded, while their interpretation becomes less adaptive, contributing to the overall distress and functional impairment. This suggests that while interoceptive sensibility, especially as measured with MAIA, has been considered beneficial (Eggart and Valdés‐Stauber  2021 ; Ferentzi et al.  2021 ; Mehling et al.  2012 , 2013 ), i.e., the more the better, there may be an optimal range, such that both excessive or insufficient tendencies to monitor internal signals can be maladaptive. In the present study, and while considering the macro‐environmental context of the national conflict, none of the various dimensions of interoceptive sensibility correlated with pain intensity in any of the groups. A relatively low effect size was expected based on previous studies informing the power analysis (Colgan et al.  2022 ; Valenzuela‐Moguillansky et al.  2017 ). However, updated meta‐analytic findings indicate no consistent associations between any domains of interoception and pain intensity in people with CP (Horsburgh et al.  2024 ). In PFC, there are mixed findings regarding the association between interoceptive processing and administered pain sensitivity (Pollatos et al.  2012 ; Werner et al.  2009 ). In contrast, clearer and more consistent associations were found in the current study, specifically between the dimensions of interoceptive sensibility that differed between people with CP and PFC and pain interference. This suggests that interoceptive sensibility may be more relevant for understanding the functional impact and disabling aspects of CP, and the subsequent affective distress, rather than the sensory components of pain. Indeed, noticing, trusting and not‐worrying about bodily sensations were also most consistently associated with emotional distress. Pain interference is of particular clinical interest as it is central to the definition of high‐impact CP, a subgroup of CP characterized by the most severe and persistent functional limitations in work‐related, personal and social activities, subsequently leading to elevated healthcare utilization and costs (Falasinnu et al.  2023 , 2025 ; Lucas and Sohi  2024 ; Pitcher et al.  2019 ). It is therefore crucial not only to identify people with high impact CP or potential risk for it, but also to identify modifiable factors that may be targeted to ease their burden. Demonstrating that several interoceptive sensibility dimensions mediated the relationship between pain interference and emotional distress suggests that by targeting these dimensions through tailored interventions one might reduce the functional and emotional impact of CP. Meta‐analytic evidence suggests that various mind–body interventions, including movement‐based and mindfulness‐based treatments, alter interoceptive processing and show positive effects on pain outcomes (Gnall et al.  2024 ), with specific increases in not‐worrying and reductions in pain interference. More psychologically and behavioural oriented interventions, including versions of cognitive‐behavioural therapy, showed similar positive indications though with less meta‐analytic conclusiveness (Heim et al.  2023 ). Additional numerous technologies are being developed and tested for targeted interoceptive interventions (Schoeller et al.  2024 ). For example, stimulation of C‐tactile small unmyelinated fibres projects sensory information to the insula, at the so‐called interoceptive cortex, and led to significant reductions in pain intensity in a sample of mixed‐aetiology CP (Di Lernia et al.  2020 ). Taken Together with present findings, these offer a promising platform for future interoception‐informed intervention for CP, especially for high‐impact CP. The place and context in which this study was conducted should be noted as an important limitation on generalizability, since data collection began in mid‐November of 2023, 5 weeks after the outbreak of the Hamas‐Israel war that erupted on October 7th of that year and lasted till December 2024, during which the conflict continued. A recent study indicated that during the onset of the Russia‐Ukraine war, stress‐induced analgesia lasted for about 30 days, with limited impact on other pain, physical and affective symptoms (Kazlou et al.  2022 ). Findings also indicated a stronger effect in countries in close proximity to Ukraine. It was thus assumed that initiating data collection 5 weeks after the onset of the war would minimize its impact on the study's measurements. Findings remained robust after controlling for geographic proximity to and self‐reported health impact of war, two pre‐registered factors assessing the potential effects of war. Thus said, data collection spanned more than a year within the macro‐environmental context of a national conflict and the negative impact of war on chronic pain (e.g., Dmytriiev and Garland  2026 ; Aloush et al.  2025 ), especially long‐term outcomes, might still have compounded our assessments, reflecting effects of war that were not controlled for. An additional relevant limitation relates to the sample size. Although large in size, with a total of 462 participants, the COFP group was smaller compared to the targeted group size (89 instead of 141) and thus might reflect underpowered findings. This, in addition to the COFP group being sampled from one specific pain clinic, also limits generalizability. Future studies in non‐conflict settings are necessary to confirm and generalize the results. Another important limitation is that this study was cross‐sectional and correlational and thus warrants caution in determining causal relationships. Rather, mediation may serve as a basis for hypothesis‐based experimental and/or longitudinal studies that would corroborate and extend current results. Such efforts would also benefit by extending measurement from purely self‐reported metrics used here to neurophysiological and behavioural measures, referred to as interoceptive attention and accuracy, respectively (Suksasilp and Garfinkel  2022 ). Notably, while the two most common measures for self‐reported interoception were used, concern about the validity of these measures have been raised (Desmedt et al.  2022 ), and future studies could also benefit from the usage of other self‐reported instruments. As such, ample opportunities emerge for uncovering the processes by which interoceptive processing informs the development and maintenance of chronic pain.

Considering the interactions between biological, psychological and socio‐cultural factors in understanding individualized pathways of pain chronification and maintenance is crucial (Cohen et al.  2021 ). This study emphasized one's body—its internal sensory signalling and their meta‐cognitive interpretations—in trying to understand the relationship between pain severity and emotional distress. Findings offer a new trajectory for the relationship between the functional impact of pain and its emotional sequelae via specific dimensions of interoceptive sensibility. A focused approach towards interoceptive‐based interventions is thus suggested, targeting specific modifiable dimensions that hopefully support and improve the lives of people suffering from persistent debilitating pain.

People suffering from chronic pain (CP) experience considerable emotional distress (Cohen et al.  2021 ; Nicholas et al.  2019 ) and have high comorbidity rates with emotional psychopathologies, such as depression and anxiety (Goral et al.  2010 ; Hooten  2016 ). Emotions and emotion‐related brain systems have an important role in pain perception, modulation and chronification (Baliki and Apkarian  2015 ; Lumley et al.  2011 ; Malfliet et al.  2017 ). Although pain and emotions are inherently linked at the phenomenological, neurophysiological and clinical levels (Gilam et al.  2020 ), the mechanisms that mediate these relationships remain unclear. Interoception, the representation and regulation of the body's internal milieu, integrates information from all biological systems involved in bodily homeostasis and is considered important for generating and regulating various affective states, from thirst, hunger and pain to emotions such as sadness, fear and anger (Barrett  2017 ; Chen et al.  2021 ; Craig  2002 ; Pace‐Schott et al.  2019 ). Recent evidence supports the association between interoception and emotions (Parrinello et al.  2022 ) and indicates a potential deficit in interoceptive processing in various mental health conditions, including mood and anxiety disorders (Khalsa et al.  2018 ). A complex pattern of associations emerges also between interoception and pain, similarly indicating differential interoceptive processing in CP (Garfinkel and Eccleston  2025 ). For example, although in pain‐free controls (PFC), higher accuracy in monitoring interoceptive signals has been associated with higher pain sensitivity (Pollatos et al.  2012 ), people with CP generally demonstrate reduced accuracy in perceiving their interoceptive signals (Horsburgh et al.  2024 ). Interoception may thus be involved in mediating the relationship between pain and emotion (Gilam et al.  2020 ) and when altered, might lead to heightened pain severity and/or emotional distress. This study aimed to examine whether interoceptive sensibility—the self‐reported multi‐dimensional tendency to perceive, monitor and interpret bodily sensations—statistically mediated the relationship between pain severity and emotional distress (Figure  1A ). To this end, individuals with CP and PFC recruited from the community, and patients with chronic orofacial pain (COFP) recruited from an orofacial pain clinic, completed a survey measuring self‐reported pain intensity and pain interference as primary and secondary measures of pain severity, respectively, emotional distress and interoceptive sensibility. Preregistered hypotheses were that pain groups would have significantly higher levels of pain severity and emotional distress compared to the PFC group and that significant group differences would emerge across dimensions of interoceptive sensibility. Furthermore, pain severity, emotional distress and interoceptive sensibility were hypothesized to be significantly intercorrelated within the pain groups and more so compared to the PFC group. Finally, interoceptive sensibility was hypothesized to statistically mediate the relationship between pain severity and emotional distress within the pain groups. While preregistration originally combined the pain groups into one, they were ultimately separated (see Methods) and thus all analyses with the COFP group should be considered as exploratory. Since this study was conducted 5 weeks into the Hamas‐Israel war, two war‐related factors, namely proximity to the war zone and self‐reported impact of war on health, were preregistered as measures to control for their potential impact on findings. Interoceptive sensibility as a mediator between pain severity and emotional distress. The conceptual framework hypothesized in this study was that interoceptive sensibility will mediate the relationship between pain severity and emotional distress (A), and specifically with pain intensity. The hypotheses focusing on pain intensity were refuted, while exploratory analyses with pain interference provided some support for this hypothesis. Not‐worrying about painful or discomforting sensations partially mediated this relationship for both CP (B) and COFP (C) groups, and specifically in the directionality in which emotional distress is the outcome of pain. Trusting one's' bodily sensations in the CP group (D) and noticing bodily sensations in the COFP group (E) partially mediated this relationship in both directional specifications of the model. CP, Chronic Pain; COFP, Chronic Oro‐Facial Pain; → and ← indicate the direction of pain interference to emotional distress and the reverse, respectively; CI, confidence interval.

The authors declare no conflicts of interest.

Table S1: Correlation coefficients between interoceptive sensibility measures in each of the three groups. Figure S1: A map of Israel divided into four zones reflecting participants' place of living during the break‐out of the Hamas‐Israel war on October 7th, 2023 (adapted from https://www.oref.org.il ). Area code 1: Gaza Envelope (30 s to get to shelter before impact), western and central Negev (45 s and 1 min, respectively), Golan Heights (15 s) and upper Galilee (30 s). Area code 2: Lower Galilee (1 min), Haifa Bay Area (1 min), Sharon (1.5 min), Gush‐Dan (central Israel; 1.5 min), Shfela (1.5 min) and Shfelat Yehuda (1.5 min). Area code 3: Southern Negev (1.5 min), Menashe (1.5 min), the Valleys (1.5 min), Carmel (1 min) and Eilat (30 s). Area code 4: Jerusalem (1.5 min), Judea and Samaria (1.5 min), the Dead Sea (1.5 min) and Arava desert (1.5 min).