Abstract
Background: Health anxiety may exist with or without prominent somatic symptoms, but the impact of somatic symptoms on treatment response is unclear. The study objective was to examine this question further as symptom burden may impact choice of type of treatment. Methods: This exploratory study used a unique database from a prior trial of 193 individuals with DSM-IV hypochondriasis who had been randomly assigned to either cognitive behavioral therapy, fluoxetine, combined therapy, or placebo. Two subgroups were newly defined-no/low somatic burden (n = 42) and prominent somatic burden (n = 151). Response was defined by ≥30% improvement in hypochondriasis. Results: Among high somatic hypochondriacal participants, compared to placebo, the odds of being a responder were significantly greater among those who received fluoxetine, either alone (OR = 4.46; 95% CI: 1.38, 14.41) or with cognitive behavioral therapy (OR = 3.56; 95% CI: 1.19, 10.68); the estimated odds were not significantly different for those receiving cognitive behavioral therapy alone (OR = 1.81; 95% CI: 0.59, 5.54). In contrast, among low somatic hypochondriacal participants, compared to placebo, the observed odds of being a responder were similar in magnitude and direction for those who received cognitive behavioral therapy, either alone (OR = 3.00; 95% CI: 0.38, 23.68) or in combination with fluoxetine (OR = 3.60; 95% CI: 0.62, 21.03), compared to the odds for those receiving fluoxetine alone (OR = 0.90; 95% CI: 0.14, 5.65). High somatic hypochondriacal individuals assigned to any fluoxetine group had significantly greater odds of being a responder than those who had not received fluoxetine (OR = 2.70; 95% CI: 1.33, 5.48). Low somatic hypochondriacal individuals assigned to any cognitive behavioral therapy group had significantly greater odds of being a responder than those who had not received cognitive behavioral therapy (OR = 8.03; 95% CI: 1.41, 45.67). Conclusion: These findings indicate that somatic symptom burden may be important in guiding treatment selection among individuals with marked health anxiety, as hypochondriacal individuals with high somatic burden responded more often to fluoxetine while those with low somatic burden responded more often to cognitive behavioral therapy. Systematic replication with larger studies is needed.