To the Editor: I would like to provide some background from the literature for the report by David J. Hellerstein, M.D., et al. (1). Although the authors’ finding appears to be solid and valid, it is not new. There is literature going back decades regarding how treatment of an axis I disorder can reduce measured personality dysfunction in both anxiety and depression (2–4). There is also literature indicating that these personality changes have clinical significance (5). Finally, there is also a growing literature on the effects of pharmacology on personality; an example is a review by Soloff (6).

A review of the literature shows the discussion moving forward from replicated findings to the meanings of the findings themselves. There is the question of whether the findings are measurement artifacts, treatment responses secondary to the treatment of an axis I disorder, or responses secondary to direct treatment of personality symptoms themselves. These are the more interesting questions. There was an empirical report (7) containing the suggestion that there may be an entity called a "stress-induced personality disorder," which, under the stress of an axis I illness, appears similar to a personality disorder but remits with the treatment of the axis I disorder. I applaud Dr. Hellerstein et al. for moving this topic into a useful area of research. I hope they find the previous literature helpful in their future endeavors.