Background: Seventeen to 43% of patients with non-cardiac chest pain suffer from anxiety/panic disorders. Cognitive behavioural therapy (CBT) is effective in reducing non-cardiac chest pain. However, no data are available indicating that pain reduction following CBT may be cognitively mediated or whether success of CBT is dependent on the presence of panic. The aim of the study was threefold: (1) does CBT have a differential effect on cognitive measures; (2) does a relationship exist between improvement in non-cardiac chest pain and changes in cognitive measures, and (3) can panic be established as a moderator of the effect of treatment?
Methods: Sixty-five patients with non-cardiac chest pain completed a randomised trial comparing study CBT with 'care as usual'. Dependent measures were: frequency of chest pain, anxiety, the fear of bodily sensations, attributions and catastrophic cognitions.
Results: CBT had a differential effect on most of the cognitive measures. Pain reduction was associated with the development of more adequate cognitions with respect to chest pain, independent of anxiety reduction. Although panic patients reported higher baseline scores on the cognitive measures, no differences in treatment results were found between panic and no-panic patients.
Conclusions: Pain reduction following CBT may be cognitively mediated. The presence of panic did not affect the outcome of treatment, implying a broad applicability of the cognitive model for treatment of patients with non-cardiac chest pain.