To the Editor: Since the original observation of Peter Kramer (1) that fluoxetine may change personality, various observations and statements about the influence of serotonergic antidepressants on personality have been made. Two interesting reports recently appeared in the Journal. First, Yevgenia Gelfin, M.D., and colleagues (2) reported that they did not observe significant effects attributable to fluoxetine on any psychological variables in 15 healthy volunteers. Then, in a double-blind study of 46 healthy volunteers, Brian Knutson, Ph.D., and colleagues (3) found that the administration of paroxetine may modulate a dimension of normal personality characterized by lower negative affect experience and higher affiliative behavior in the absence of baseline clinical depression or other psychopathology.

Any effort to explain these seemingly discrepant findings only raises more questions. First, how is a "change of personality" measured? Could the discrepancy in findings between these two studies (2, 3) be explained just by the use of different measurement tools? Second, is everyone equally affected by antidepressant treatment? According to Dr. Knutson and colleagues (3), volunteers were not equally affected. Does this mean that some may be affected and some may not be affected at all? Could this help define the biological substrate of personality? Third, would a design with the use of an active placebo help resolve the discrepancy between the two studies? Fourth, could it be that serotonergic antidepressants just help "normal" and other people tolerate stress better, as suggested by Norden (4)? Fifth, are these personality changes permanent? Neither of the two studies addresses the last issue, even though it could be easily done by repeating personality measures 1, 3, and any number of months after discontinuing the antidepressant. Maybe the authors could provide us with these data.

The only definite conclusion we can make is that we are very far from elucidating the alteration of personality (if any) by serotonergic antidepressants.