In Brief

Psychological research published in the October Archives of General Psychiatry (Vol. 61, No. 10) is challenging many behavioral scientists' long-held belief that personality disorders, or PDs, are fixed deficits that persist throughout people's lives.

The study--based on four years of data from the Longitudinal Study of Personality Disorders and funded by the National Institute of Mental Health--used a different statistical method than previous research to find that PD indicators decreased over time.

Such changes counter the views of some that such disorders as antisocial PD and narcissistic PD are inflexible and therefore largely untreatable, says the study's lead author, psychologist Mark F. Lenzenweger, PhD, a psychology professor at the State University of New York at Binghamton.

In the study, Lenzenweger, fellow faculty member Matthew D. Johnson, PhD, and Harvard education professor John B. Willet, PhD, used individual growth curve analysis--in which investigators study each participant's level and rate of change instead of averaging all participants together--to track 250 study participants who were college freshmen when they entered the longitudinal study in 1990. About 25 percent of the participants sought mental health treatment on their own before or during the study.

The researchers evaluated each person three times over four years for the PD features listed in the then-current edition of the DSM--the Diagnostic and Statistical Manual of Mental Disorders (Third Edition, revised). To avoid evaluator bias, participants saw a different evaluator each time and the evaluator was blind to previous assessments.

After the fourth year, the researchers plotted the number of participants' DSM personality disorder features at the three time points. They found that:

  • At the outset, about 50 percent of the sample was identified as at-risk for a PD; about 10 percent received a diagnosis during the study--a number similar to other prevalence estimates.

  • Some participants' PD features increased over time, while others decreased. But on average, the total number of PD features decreased at a rate of 1.4 features per year.

  • Participants identified as at-risk for a PD were more likely to have a change in their PD features, and, on average, those features declined.

"For something that's not supposed to be changing at all according to the DSM system, we saw a substantial amount of change," says Lenzenweger. Moreover, he notes, additional analyses ruled out possible explanations for what might be causing the change: In fact, participants' biological sex, a history of another mental health problem such as depression or schizophrenia, or whether a participant sought mental health treatment did not affect the rate of change.

However, Lenzenweger notes, it's unclear whether the treatments participants sought were effective for PD problems, since the study didn't distinguish between emerging therapies designed specifically for PDs--such as transference-based therapy or dialectical behavior therapy--and less-effective forms. Moreover, since the age and socioeconomic status of the study participants isn't reflective of the U.S. population, the study's findings need to be replicated, he says.

Caveats considered, Lenzenweger concludes the study's findings are still a "bright light" for those who treat and research PDs: "The flexibility has been documented, and now we have to develop treatments that can tap into that."