Abstracts of Published Research Papers from the CFM

Kabat-Zinn, J., Chapman, A. and Salmon, P. The relationship of cognitive and somatic components of anxiety to patient preference for alternative relaxation techniques. Mind/Body Medicine (1997) 2:101-109.


Background: The relationship between cognitive/somatic response pattern for anxiety and preference for different relaxation techniques was evaluated in an exploratory study of 135 medical patients referred for mindfulness-based stress reduction training, in which they practiced three major stress-reduction techniques.

Method: Following intervention, patients rated on visual analogue scales, how much they liked each of three techniques: sitting meditation, a body scan meditation, and Hatha Yoga, which differed in primary cognitive/somatic orientation but shared the unifying attentional stance characteristic of mindfulness meditation.

Results: Of the 74 patients who showed pre-treatment levels of anxiety above the mean for the entire group, 29 (39%) showed a pattern in which either the cognitive or the somatic component of anxiety predominated. The high cognitive/low somatic anxiety subgroup (n = 9) showed a significant preference for the most somatic technique (Hatha Yoga) and liked least the most cognitive technique (sitting meditation). The high somatic/low cognitive anxiety subgroup (n = 20) showed the inverse response. The body scan, with both cognitive and somatic qualities, was preferred to an intermediate degree by both groups. Irrespective of an individual’s mode of anxiety expression or technique preference, participation in the mindfulness-based stress reduction program appeared to be effective in reducing overall anxiety levels.

Conclusions: These findings differ from several previous studies of anxiety modality (cognitive or somatic) and relaxation technique preference that used nonclinical populations, and appear inconsistent with Davidson and Schwartz’ hypothesis that treatment of anxiety is best oriented toward the mode in which it is expressed.