In the controlled trial with 100 antidepressant-refractory OCD patients, 80 percent of those who received CBT responded, compared to 23 percent of those who received the antipsychotic risperidone, and 15 percent of those who received placebo pills. Forty-three percent experienced symptoms reduced to a minimal level following CBT treatment, compared to 13 percent for risperidone and 5 percent for placebo.

The study, published September 11, 2013 in JAMA Psychiatry, was led by Helen Blair Simpson, M.D., of Columbia University, in New York City; and Edna Foa, Ph.D., of the University of Pennsylvania, Philadelphia.

In an accompanying editorial, grantees Kerry Ressler, M.D., and Barbara Rothbaum, Ph.D., of Emory University, Atlanta, note that antidepressants are effective in treating only a subset of OCD patients. They add that the targeted form of CBT works via different mechanisms -- such as retraining the brain's habit-forming circuitry to unlearn compulsive rituals.

Matthew Rudorfer, M.D., chief of the NIMH Somatic Treatments Program, which funded the study, said that in demonstrating how different patients respond best to different approaches, it helps to move the field toward the goal of more personalized treatment.