Changes in Dream Content After Drug Treatment

Nili T. Kirschner

University of California, Santa Cruz

NOTE: If you use this paper in research, please use the following citation, as this on-line version is simply a reprint of the original article:
Kirschner, N. (1999). Changes in dream content after drug treatment. Dreaming, 9, 195-200.


This case study examines the effects of sertraline (Zoloft) on the dream content of a young woman with generalized anxiety disorder and panic attacks. The study uses the major categories of Hall and Van de Castle's (1966) system of content analysis to compare dream reports before and after drug treatment. Prior to diagnosis and treatment, the dreamer had high levels of aggression and low levels of friendliness in her dreams. The post-medication dreams more closely approximate the female norms. This pilot study suggests a new direction for research on the effects of medication on dream content.


The purpose of this study is to present a case study of 33 pre-medication and 40 post-medication dream reports from a young woman who entered outpatient psychological treatment at age 18 and at age 20 was placed on sertraline, an anti-depressant selective serotonin reuptake inhibitor. Since the study involves only one subject and somewhat less than the 75-100 dream reports for each condition that would be ideal in terms of sample size (Domhoff, 1996), it is best viewed as a pilot study. Nonetheless, it is the first study on the influence of medication on dream content indicators to detect any changes that may have occurred. The study is also worthy of note because several of the changes in the dreams after medication are in the medium to large range in terms of the magnitude of effect sizes (Domhoff, 1996, 1999).

Perhaps due to the great difficulties of carrying out systematic studies on how drugs might effect dreams, the literature on this topic is not large. Roth, Kramer, and Salis (1979) provide an excellent review and critique of the nine studies that had been published up to the time of their overview. They conclude that "some sedative-hypnotic and antidepressant agents may effect the quality of dreams, but the precise nature of the effect is yet to be determined" (Roth et al., p. 220). They further note that "Few if any of the existing studies were more than pilot studies," and that "there has been a failure to examine even one drug in depth..." (Roth et al., p. 221). Moreover, it is their conclusion that "the lack of standardization of methods of assessing quality of dream content has resulted in isolated bits of information that do not yet form a coherent picture" (Roth et al., p. 221).

There have been only two new studies since the critique of the literature cited in the previous paragraph. The first does not deal directly with dream content, but with self-ratings of the pleasantness of dream reports by depressed outpatients at various stages in their treatment with trimipramine; the patients reported after four weeks that their dreams had fewer negative emotions (Riemann, Low, Schredl, Weigand, Dippel, & Berger, 1990).

The second study compares 27 pre-medication dream reports from a group of patients suffering major, nonpsychotic depression with 32 dreams from the same group after they began taking antidepressant medication, either fluoxetine (Prozac) or nefazndone (Serzone). The study is described as an "add-on to studies of the effects of antidepressants on sleep physiology" (Armitage, Rochlen, Fitch, Trivedi, & Rush, 1995, p. 191), so the dreams were obtained after a set time of awakening each morning in the sleep laboratory. The major finding is the low level of dream recall both before and during medication, with only 21 of 89 patients reporting at least one dream from at least one of the two conditions.

The two small sets of dreams are rated on 0-5 scales for 19 dimensions of dreams, such as vividness, active-passive participation, and presence of others. The scales for emotionality, sexuality, violence, and success/failure were adapted from Hall/Van de Castle categories (Armitage et al., 1995, p. 192). Both sets of dream reports were "short, relatively bland with little emotion" (Armitage et al., p. 193). There were only two differences: "On treatment, when depression symptoms were improved, dreams were less vivid with fewer scenes" (Armitage et al., p.196).

Given the tentative nature of most findings in this literature and the sometimes inconsistent results (Armitage et al., 1995, p.196; Roth et al., 1979, p. 219), perhaps this brief case study can suggest a new direction for future studies.

Participant and Method

The dreamer is a female college student. She recorded the first series of 33 dreams (pre-medication condition) as part of a psychotherapeutic process when she was 18 and in her first year of college. At that time she was suffering from anxiety and mild depression. After one year of psychotherapy through her college health services, the patient terminated her therapy because she felt that it was not helping her. One year after ending psychotherapy, at age 20, the patient consulted a doctor about heart palpitations and light-headedness and was referred to a psychiatrist who diagnosed her with generalized anxiety disorder and panic attacks. The psychiatrist prescribed sertraline (Zoloft), an antidepressant selective serotonin reuptake inhibitor that is sometimes prescribed to ease panic attacks and anxiety. The patient began taking 25 milligrams of sertraline nightly, and gradually worked up to 100 milligrams before the panic attacks stopped. One year later, at the age of 21 and still taking medication, she recorded 40 more dreams (post-medication condition) for possible use in a research project in conjunction with a college course in psychology, but without any knowledge of the Hall/Van de Castle system.

The dreams were coded separately by the author and Adam Schneider using Hall and Van de Castle's (1966) major content categories. The percentage of perfect agreement was above 90 percent for each category. When the codings differed, the codings by Schneider were used. The individual codings were entered into DreamSAT, which calculated a subset of the Hall/Van de Castle indicators along with significance levels and effect sizes (Domhoff, 1996, 1999; Schneider & Domhoff, 1999).


The post-medication dreams differed significantly from the pre-medication dreams in several of the content categories (see Table 1). The effect sizes were often above .40, which is considered large for dream content studies (Domhoff, 1996). The most dramatic effect is the post-medication decrease in the A/C Index, the ratio of aggressive acts to characters (h = -.82, p < .001). The aggression/friendliness percent also decreased (h = -.48, p < .003).

In addition to the decrease in aggression, there was a corresponding increase in friends and friendliness. The ratio of friendliness to characters (F/C Index) increased (h = +.41, p < .003), along with the number of characters who were friends of the dreamer (h = +.44, p < .003).

Another positive effect was the dreamer's increase in good fortune (h = +.58, p < .013). Finally, the number of elements from the dreamer's past decreased dramatically (h = -.71, p < .01), as did familiar settings (h = -.75, p < .002).

Table 1. Pre-Medication vs. Post-Medication Content Category Percentages
 Animal percent11%06%-.20.135
Male/Female percent54%60%+.10.545
Familiarity percent76%75%-.02 .913
Friends percent°42%64%+.44.003b
Group percent29%27%-.03 .820
Social interaction ratios
 A/C index°.78.39-.82.000b
F/C index°.18.35+.41.003b
Aggression/Friendliness percent°75%52%-.48.003b
Befriender percent80%63%-.39.197
Aggressor percent°53%41%-.23.330
Physical aggression percent°26%13%-.35.054
Other content categories
 Indoor setting percent53%59%+.13.544
Familiar setting percent 88%55%-.75.002b
Negative emotions percent°92%84%-.28.111
Dreamer-involved success percent°18%23%+.14.714
Bodily misfortunes percent°32%18%-.33.241
Percentage of dream reports with at least one:
Good fortune03%20%+.58.013a
Success09%08%-.05 .836
Elements from past 43%13%+.71.010b
NOTE: Categories followed by ° are possible indicators of psychopathology.
a Significant at the .05 level.
b Significant at the .01 level.

Most of the changes in the post-medication dream content brought the dreamer closer to the female norms established by Hall and Van de Castle (see Figure 1). In the cases where she moved further from the norms, such as friendliness and good fortune, it was on dream content that can be considered positive in nature.

Figure 1. h-profile of pre-medication and post-medication dreams compared to female norms.


The large effects in this study could be due to the passage of time, but research on age and dream content shows that there is very little change in dream content beyond young adulthood (Domhoff, 1996). The change also may be due in part to psychotherapy, but one unpublished study by Hall showed no effect of psychotherapy on dream content (Domhoff, 1993). It is therefore most likely that the changes in dream content are related to the medication or the reduction of anxiety that ensued.

One possible explanation for the disparity in findings between this study and previous studies is that the latter tended to focus on the dreams of depressed patients. Several researchers have found the dreams of depressives to be short, bland, and lacking in detail (Armitage et al., 1995; Barrett & Loeffler, 1992, Riemann et al., 1990). It may be harder to detect changes in content in less detailed dreams.

Domhoff (1996) has suggested that several of the Hall and Van de Castle content categories may be useful in predicting psychological disturbances. Before the dreamer was diagnosed and treated, she showed many of the potential indicators of psychopathology: high percentages of aggression, misfortune, and negative emotions, and low percentages of known characters, friendly interactions, and dreamer-involved success.

The dreamer's pre-medication scores are consistent with the previous findings that the dreams of anxious patients are low in friendly interactions (Gentil & Lader, 1978), and that patient populations tend to have more negative emotions in their dreams (Bollea, Carbonetti, Donini, Marrucci, & Vella, 1978). Her post-medication dreams more closely approximate the female norms in most categories, including those that might reflect psychological disturbance. In categories that are not indicators of psychopathology (such as the percentage of characters that are individuals or groups, the male/female percent, and the dream setting), the dreamer showed more consistency - even when she differed from the female norms - between the pre- and post-medication dreams. The changes in the pertinent categories suggest that the drug treatment did indeed affect dream content.

It is noteworthy that 43 percent of the pre-medication dreams contained elements from the dreamer's past, while only 13 percent of the post-medication dreams contained such references. This may be related to the large decrease in familiar settings in the post-medication dreams. Although Domhoff (1996) does not list a high percentage of elements from the past as an indicator of psychopathology, he does mention that people suffering post-traumatic stress disorder (PTSD), a type of anxiety disorder, tend to have dreams in which distressing events are relived again and again. It may be that other anxiety disorders invoke a similar response in which the dreamer has a tendency to dwell on past events. This possibility merits further research.

A final observation is that the results of this study provide support for Hartmann's (1984) biological model of the effects of drugs on dreams. An early study which focused mainly on long-term sleep patterns found little change in dream content associated with psychotropic drug administration (Hartmann & Cravens, 1974), but a later study conducted in Hartmann's laboratory indicated that increased levels of dopamine resulted in more vivid, nightmarish dreams (Hartmann, Russ, Oldfield, Falke, & Skoff, 1980). Based on his own research and the literature on drugs and nightmares, Hartmann (1984) proposed that drugs that increase the neurotransmitters dopamine or acetylcholine, or decrease norepinephrine or serotonin, produce nightmares and more vivid bizarre dreams. Drugs that have the opposite effects would decrease the incidence of disturbing dreams. The dreamer in this study was taking a serotonin reuptake inhibitor, which served to increase the effects of serotonin. According to the biological model, with the onset of medication the dreamer should have experienced a decrease in nightmares, or, in Hall and Van de Castle's terms, lower aggression, negative emotions, and other unpleasant factors. This was, in fact, the case.


This study would have benefited from several additions, most notably a larger sample of dreams from both conditions. It is possible that the dramatic effects in this study would have diminished or even disappeared in a larger sample of dreams. Although these data are from a small case study, they nonetheless point toward a new area of dream research. If dream reports could be obtained from a wide range of patients before and after undergoing drug treatment for psychological disturbances, it might be possible to assess the effects of medication on dream content in a systematic way.


  • Armitage, R., Rochlen, A., Fitch, T., Trivedi, M., & Rush, A. J. (1995). Dream recall and major depression: A preliminary report. Dreaming, 5 (3), 189-198.
  • Barrett, D., & Loeffler, M. (1992). Comparison of dream content of depressed vs nondepressed dreamers. Psychological Reports, 70, 403-406.
  • Bollea, E., Carbonetti, P., Donini, G., Marrucci, M., & Vella, G. (1978). Dream activities of depressives. Archivo de Psicologia, Neurologia e Psichiatria, 39, 473-501.
  • Domhoff, G. W. (1993). The repetition of dreams and dream elements: A possible clue to a function of dreams. In A. Moffitt, M. Kramer, & R. Hoffman (Eds.), The Functions of Dreams (pp. 293-320). Albany, NY: State University of New York Press.
  • Domhoff, G. W. (1996). Finding Meaning in Dreams: A Quantitative Approach. NY: Plenum Press.
  • Domhoff, G. W. (1999). New directions in the study of dream content using the Hall/Van de Castle coding system. Dreaming, 9, in press.
  • Gentil, M., & Lader, M. (1978). Dream content and daytime attitudes in anxious and calm women. Psychological Medicine, 8, 297-304.
  • Hall, C. S. & Van de Castle, R. (1966). The Content Analysis of Dreams. NY: Appleton-Century-Crofts.
  • Hartmann, E. (1984). The nightmare: The psychology and biology of terrifying dreams. NY: Basic Books, Inc.
  • Hartmann, E. & Cravens, J. (1974). Long-term psychotropic drug administration: Effects on dream content and on REM density. Sleep Research, 3, 119.
  • Hartmann, E., Russ, D., Oldfield, M., Falke, R., & Skoff, B. (1980). Dream content: Effects of L-Dopa. Sleep Research, 9, 153.
  • Riemann, D., Low, H., Schredl, M., Weigand, M., Dippel, B., & Berger, M. (1990). Investigations of morning and laboratory dream recall and content in depressive patients during baseline conditions and under antidepressive treatment with trimipramine. Psychiatric Journal of the University of Ottawa, 15, 93-99.
  • Roth, T., Kramer, M., & Salis, P. J. (1979). Drugs, REM sleep, and dreams. In B.B. Wolfman (Ed.), Handbook of Dreams (pp. 203-225). NY: Van Nostrand Reinhold Co.
  • Schneider, A., & Domhoff, G. W. (1999). The quantitative study of dream content. Available: http://www.dreamresearch.net

Go back to the Dream Library index.

dreamresearch.net home page dreamresearch.net contact info