Since 1991, there have been nine randomized controlled trials that have shown the benefits of DBT. In an early study, Linehan et al. (1991) found that the DBT treatment group, in comparison with the treatment as usual in the community (TAU) group, exhibited a reduction in the frequency and medical risk of parasuicidal behavior, an elevated retention rate in therapy, and a decreased number of days of inpatient psychiatric hospitalization. Although the two treatment conditions did not vary in efficacy in terms of improving the patients' depression, hopelessness, suicidal ideation, or reasons for living, Koons et al. (2001) did find such improvement with DBT when studying a cohort of women veterans with BPD against a TAU group. There have been claims that other variables, such as economic status and other confounding factors, lead to statistically significant differences in treatment group outcomes. Researchers, however, refute these claims, arguing that it is the DBT that is driving the results (Linehan and Heard 1993).