Suppressor effects in the relations of psychological variables to resting blood pressure.

The role of suppressor effects in obscuring the relation of psychological variables and blood pressure was studied. Forty-five nonmedicated patients in a family practice rated themselves relative to their peers on a series of characteristics. Two of these, trapped and lonely, exhibited marked suppressor effects in their relationship with each other and resting blood pressure. Self-ratings of lonely exhibited a moderately strong positive association with ratings of trapped (r = .63) while only ratings of lonely showed a significant correlation with resting systolic blood pressure (r = .31). Yet when the suppressor effects of the ratings on feeling lonely were removed in the regression analyses, the ratings on feeling trapped showed a significant positive association with resting systolic pressure (r = .42). Similarly, while neither ratings on feeling lonely or on feeling trapped showed separately a significant association with resting diastolic pressure, when suppressor effects were removed in the regression analyses, the ratings of trapped were significantly associated with diastolic pressure in a positive direction (r = .34) and ratings of lonely were significantly associated in a negative direction (r = .33). The significance of this specific suppressor effect and the implications of suppression for psychosomatic research are discussed.