Personality Traits and Symptom Reduction in a Group Treatment for Women With Histories of Childhood Sexual Abuse Nancy L. Talbot, Paul R. Duberstein, Jessica S. Butzel, Christopher Cox, and Donna E. Giles The influence of personality on symptom reduction has not been examined in research on treatments for women with childhood sexual abuse histories, al- though personality has demonstrated predictive value in other treatment contexts. This study exam- ined personality variables associated with symptom reduction in group therapy for hospitalized women with histories of sexual abuse. Personality was mea- sured with the NEO-Five-Factor Inventory (NEO-FFI), which yields scores on neuroticism, extraversion, openness to experience, agreeableness, and conscien- tiousness. Among 86 women who participated in ei- ther the Women’s Safety in Recovery (WSIR) group therapy or treatment as usual, 43 completed assess- ments of symptom reduction at discharge and 6-month follow-up. We hypothesized that extraver- sion, agreeableness, and openness to experience would be associated with treatment outcome. Our results showed that agreeableness and extraversion moderated the effect of treatment on symptom reduc- tion. WSIR participants who were less agreeable im- proved more at discharge and 6-month follow-up than more agreeable WSIR participants. Moreover, women in the WSIR group who were more introverted showed greater symptom improvement at discharge than more extraverted women. Our findings suggest that more introverted, less agreeable patients with sexual abuse histories may indeed benefit from struc- tured group treatments. © 2003 Elsevier Inc. All rights reserved. P SYCHOTHERAPY treatment research is in- creasingly concerned with the question of “what works for whom.” 1 Even when therapies are found to be effective for particular patient groups, variability in outcomes among participants highlights the significance of factors other than treatment type. Previously we reported on the effectiveness of a group treatment for women with childhood sexual abuse his- tories who were hospitalized on a psychiatric unit. 2 The group therapy, Women’s Safety in Recovery (WSIR), was more effective in reducing psychological symptoms than a treatment-as-usual (TAU) compari- son. The influence of personality on treatment outcome has not been examined among women with childhood sexual abuse histories, although personality has dem- onstrated predictive value in other treatment con- texts. 3-9 Personality variables associated with treatment outcome in the WSIR and comparison groups are the focus of this report. Personality was assessed with the NEO Five-Factor Inventory (NEO-FFI), 10 a comprehensive, empirically derived assessment of five personality traits: neuroti- cism, extraversion, openness to experience, agreeable- ness, and conscientiousness. Theoretical consider- ations and empirical research suggest that people who are less extraverted, agreeable, and open to experience may be better candidates for structured, goal-directed therapies; in contrast, exploratory or interpersonal ther- apies may be better suited to people who are interper- sonally oriented, agreeable, and open. 10-14 Two reports on highly structured treatments for insomnia and ago- raphobia suggest that patients who are low in extraver- sion, openness to experience, and agreeableness can respond to structured behavioral treatments. In a be- havioral treatment for insomnia, patients with lower scores on openness to experience and extraversion made greater gains in total sleep time than those who obtained higher scores. 12 In the second study, patients with lower scores on agreeableness had fewer anxious behaviors and less agoraphobic thinking than higher scorers following a computer-based exposure treat- ment for agoraphobia. 13 The WSIR group is highly structured, psycho- educational and skills-focused rather than open- ended and exploratory. 15 Given that low scorers on extraversion, agreeableness, and openness to expe- rience do better than high scorers in structured, goal-directed therapies, 10-14 we could predict that From the Departments of Psychiatry, Oncology, Biostatistics, and Neurology, University of Rochester School of Medicine and Dentistry, Rochester, NY. Supported in part by a Leonard F. Salzman ResearchAward from the Department of Psychiatry, University of Rochester, and US Public Health Service Grants No. R01-MH39531, MH60350, and MH51201. Address reprint requests to Nancy L. Talbot, Ph.D., Associate Professor of Psychiatry, University of Rochester School of Med- icine and Dentistry, 300 Crittenden Blvd, Rochester, NY 14642-8409. © 2003 Elsevier Inc. All rights reserved. 0010-440X/03/4406-0001$30.00/0 doi:10.1016/S0010-440X(03)00142-1 448 Comprehensive Psychiatry, Vol. 44, No. 6 (November/December), 2003: pp 448-453