Study at a glance
- Authors: Ribera d’Alcalà, Chiara, Webster, Duncan G., Esteves, Jorge E.
- Year: 2014
- Journal: International Journal of Osteopathic Medicine
- DOI: 10.1016/j.ijosm.2014.08.003
- N: 59 (22 chronic pain outpatients, 37 controls); heartbeat data for 58
- Design: Case-control, cross-sectional; heartbeat counting task with ECG, Body Awareness Questionnaire, and self-reported years of proprioceptive and mind-body activities. Mann-Whitney U, Spearman correlations, backward stepwise regression. Based on a master’s dissertation; read from the accepted manuscript.
Key findings
- No difference in heartbeat counting accuracy between chronic pain patients and controls (mean 0.65 vs 0.70; U = 370, p = .77).
- No difference in Body Awareness Questionnaire scores (4.34 vs 4.53; p = .93).
- Heartbeat counting accuracy was unrelated to BAQ body awareness, in the whole sample (r = -.04) and within each group.
- The only significant correlation was negative: more years of mind-body practice (yoga, tai chi, meditation, Feldenkrais, Alexander technique) went with lower heartbeat counting accuracy (r = -.23, one-tailed p = .04). After removing outliers the authors report a positive but non-significant trend.
- No variable (age, sex, BMI, pain, BAQ, activity years) significantly predicted heartbeat counting accuracy in a backward stepwise regression; the final model kept age and pain, with pain carrying the larger positive weight (beta = .36, model p = .13).
Limitations
- Groups differ badly on age (48 vs 26 years) and BMI (26.9 vs 23.9), both known to affect heartbeat counting; controls were osteopathy students and staff, patients were clinic outpatients.
- Small sample for a null result (22 patients), with no power analysis, and one-tailed tests throughout.
- The abstract reports a ‘positive trend between body awareness and mindfulness based activities’, but the significant result in the paper is a negative correlation with heartbeat counting, and the positive trend appears only after post hoc outlier removal. The headline and the data do not match.
- All patients were receiving osteopathic manipulative treatment, which the authors themselves flag as a possible confound for body awareness.
- Anxiety and depression were exclusion criteria, which removes the affective component most pain-interoception accounts treat as central, and pain type, duration and severity were not analysed.
- Non-standard counting intervals and instructions; ‘interoceptive awareness’ is used for heartbeat counting accuracy (pre-Garfinkel 2015 terminology).
Connections
- Methods: Heartbeat detection task
- Concepts: Chronic pain, Interoceptive taxonomy (Farb et al. Box 1), Interoceptive sensitivity, Age-related interoceptive decline, Insular cortex
- Debates: contemplative training increase objective interoceptive accuracy?, Is the heartbeat-counting task a valid measure of cardiac perception?
- Related studies: Murphy et al. (2020), Khalsa et al. (2009), Desmedt et al. (2022), Bonaz et al. (2021)
A small case-control study from the British School of Osteopathy, and the wiki’s first primary data on interoception in chronic pain. Until now Chronic pain rested on theory (Craig’s reclassification of pain as interoceptive), lesion work, and one sentence in a review (Bonaz et al. 2021, citing Di Lernia et al. 2016) that accuracy deficits predict symptom severity. This paper tests the simpler version of that idea directly, and finds nothing.
What it found
Patients and controls did not differ on heartbeat counting accuracy or on the Body Awareness Questionnaire, and the two measures did not correlate with each other. The authors present this as “the first study investigating the role of interoceptive awareness in individuals affected by chronic pain” and call for more research.
How much weight it bears
Very little, for three reasons that stack.
- The groups are not comparable. Patients were on average 22 years older and heavier. Age lowers heartbeat counting accuracy (Khalsa et al. 2009, see Age-related interoceptive decline) and BMI is one of the few reliable correlates of the task (Desmedt et al. 2022). Those confounds could mask a real difference as easily as create a false one.
- The instrument is the contested one. Everything on Is the heartbeat-counting task a valid measure of cardiac perception? applies, and the authors also changed the standard instructions.
- The sample excludes what matters. Anxiety and depression were exclusion criteria, and pain severity was never analysed, so the study cannot speak to the severity claim the wiki currently holds.
So this is not a contradiction of the review finding on Chronic pain: that finding is about accuracy predicting severity, and this study tested only a group difference, in a confounded sample. It is recorded there as a tension. Di Lernia et al.’s review may well include this study; the wiki does not hold the review first-hand, so that is unchecked.
Two smaller contributions
A null that the 2×2 model predicts. Heartbeat counting did not correlate with the BAQ (r = -.04). Murphy et al. (2020) describe the BAQ as a scale that bundles believed accuracy with attention to the body, and their model predicts that attention-flavoured self-report will not track objective accuracy. This is one more instance of the wiki’s recurring self-report/accuracy null, and it fits Murphy’s boundary on it rather than challenging it. See Interoceptive taxonomy (Farb et al. Box 1).
A weak data point on mindfulness and accuracy. Years of mind-body practice went with slightly worse heartbeat counting, not better. Cross-sectional, self-reported, one-tailed, and reversed by outlier removal, so it is noise-level. But its direction matches the Khalsa side of contemplative training increase objective interoceptive accuracy? (practitioners are not more accurate), and it is worth noting that the paper’s abstract describes the opposite.
Placement
Adds the first empirical study to Chronic pain, a small entry to contemplative training increase objective interoceptive accuracy?, and a key-study link on Heartbeat detection task. The BAQ (Shields et al. 1989) has no method page; it appears here and in Murphy et al. (2020) only.