Patients' Dreams May Reveal Psychotherapy Progress

Why do we dream? To learn, to process our memories, to regulate our moods, to solve our problems. So, at least, imply some scientific studies.

Learning, memory processing, mood regulation, and problem solving, of course, are critical to the psychodynamic process and to the outcome of psychoanalysis. Thus, might the content of dreams reflect progress or lack thereof in psychoanalysis and in psychodynamically oriented psychotherapy?

Some anecdotal evidence has suggested that this might be the case. And now three small pilot studies do as well.

They were conducted by Myron Glucksman, M.D., a clinical professor of psychiatry at New York Medical College in Valhalla, N.Y., and Milton Kramer, M.D., a clinical professor of psychiatry at New York University School of Medicine in New York City. Glucksman presented the results at the 47th annual meeting of the American Academy of Psychoanalysis and Dynamic Psychiatry, held in San Francisco in May.

In the first study, 12 subjects who had completed either psychoanalysis or psychodynamically oriented psychotherapy were rated by Glucksman for clinical progress-say, symptom reduction, conflict resolution, or improved relationships with other people-and then rank-ordered from the most improved to the least improved.

After that, Glucksman turned the content of the first and last dream that each subject had dreamt during therapy over to Kramer. Kramer did not know anything about the patients or how Glucksman had ranked them. Kramer then evaluated the content of the first dream and last dream of each subject according to either negative to positive, or positive to negative, changes in mood, behavior, self-image, and problem solving, and, on the basis of this evaluation, rank-ordered the subjects from the most improved clinically to the least improved clinically.

Glucksman and Kramer then compared their rankings to see whether there was any correlation and found that the correlation came close to reaching statistical significance (p = .07).

In their second study, Glucksman and Kramer once again used the same 12 subjects and Glucksman's previous rankordering of them from the most clinically improved to the least clinically improved. But this time Glucksman picked a dream from early therapy and from late therapy for each subject that portrayed a particular psychodynamic issue or conflict that the subject was experiencing and which Glucksman believed might reflect the status of the subject in analysis at these time points. These dream pairs were then randomly ordered and turned over to Kramer, who evaluated their content and used his evaluation to rank-order the subjects from most improved clinically to least improved clinically.

The researchers then looked to see whether there was any correlation between Glucksman's rank-ordering of the subjects' clinical progress via clinical assessment and Kramer's rank-ordering of the subjects' clinical progress via dream assessment. The result of p=0.09 was not statistically significant or quite as good as in the previous study, but encouraging.

Finally, in the third study, Glucksman used his previous assessment of the 12 subjects' clinical progress to divide them into two groups-the six most improved clinically and the six least improved clinically. Glucksman then selected, for each subject, a dream from early in analysis and from late in analysis that pertained to psychodynamic issues with which he or she was grappling and which he believed might reflect the status of the subject in analysis at these time points. Glucksman then turned the 12 dream pairs, in random order, over to Kramer. Kramer evaluated the dreams' content and decided on the basis of his evaluation which six of the subjects had improved most clinically and which six had improved least clinically.

The researchers then looked to see whether there was any correlation between Glucksman's grouping based on clinical as sessment and Kramer's grouping based on dream assessment. The result was highly significant statistically (p = .001). "We were really encouraged by this," Glucksman said at the AAPDP meeting.

Glucksman and Kramer will be conducting more research to determine whether dreams are a reliable index of clinical progress in analysis, Glucksman told Psychiatric News. And if they can come up with enough evidence to demonstrate it, "then successive dreams selected during therapy may be an extremely useful tool for the clinician to evaluate the effectiveness of therapy," he and Kramer foresee.

Such evidence would also add to the growing scientific testimony that analysis and psychodynamically oriented psychotherapy are clinically effective (Psychiatric News, February 2, 2001; July 6, 2001; December 20, 2002). ▪