Borderline personality disorder and the body: what 4 weeks of DBT changed

Patients with BPD paid markedly less attention to internal body signals than controls (MAIA-2 2.10 vs 2.87), and that gap tracked their interpersonal problems. After four weeks of residential DBT, only their self-reported attention improved -- the brain and heartbeat measures did not budge.

Interoception is the sense of your own body from the inside: a racing heart, a tight chest, the physical signature of an emotion. Because borderline personality disorder involves intense emotional swings and a disturbed sense of self, researchers have long suspected interoception is part of the picture, but "interoception" is really several different abilities. Researchers led by Jella Voelter separated two of them in BPD patients and healthy controls: accuracy, tested with a heartbeat discrimination task, and attention, measured by questionnaire and by fMRI while participants focused on their heartbeat. The two came apart cleanly. Accuracy was no different between groups overall, though it fell as BPD symptom severity rose. Attention was clearly reduced: BPD patients scored 2.10 on the MAIA-2 attention measure against 2.87 for controls, and the lower their score, the more interpersonal problems they reported. The brain data matched: in the insula and dorsal anterior cingulate, regions that integrate body signals, BPD patients showed less separation between attending inward to the heartbeat and attending outward, as though the switch between the two were blunted. Then 37 patients completed a four-week residential DBT programme and were retested. Self-reported interoceptive attention improved. Nothing else did -- not the brain response, not the heartbeat task. The authors read this as evidence that interoceptive disturbance in BPD is specific to attention, and that self-report and physiology respond to treatment on different timelines, noting that mindfulness effects on interoception usually take about eight weeks to appear while this programme ran four. The design limits what follows from it: there was no waiting-list patient group, so improvement over four weeks cannot be separated from time and residential care generally. The patients were a naturalistic sample with comorbidities and psychiatric medication, included as a statistical control rather than excluded, and the heartbeat task was difficult even for controls, which may hide real differences rather than show their absence.