A s we well know, depressiveillness representsa major healthproblem intheUnited States.Gold et al.’estimatethat13% to 20% of the populationhave depressivesymptoms atany one time,and that2% to3% ofthepopulationarehospitalizedor seriously impaired due todepression.Of thosehospitalized, 40% to60% are diagnosed ashaving melancholicdepressions.The best clinical estimateoflifetimeriskforthedevelopment of unipolardepressionis3% to4% inmen and 5% to9.3% in women. Bipolardepressionoccurs at significantly lowerratesinboth men and women (0.65%to0.88%). In both psychiatricand generalmedical settings, the initial diagnosisof depressioncan be inaccurate.A review of the literature by Feinberg and Goodman2 provided convincing evidence that 2.6% to 3% of patients given an initial diagnosis of major depression were given a diagnosis of dementia at follow-up. Moreover, a clinically significant percentage (not specified in their review) of thoseinitially diagnosed as having depressionwith secondarydementiawere later found tohaveareversible dementia syndrome of depression. The diagnoses of 5% to15% ofthosepatientswho had been diagnosedinitially as having dementia were laterchanged todepression. Last,only 11% to50% of thosepatientsinitially diagnosed as having dementia with secondary depression