Abstract: This pilot study explored the effects of a series of six weekly Guided Imagery and Music [GIM] sessions on mood state and plasma á-endorphin in healthy subjects. Eight healthy adults (21-45years old) randomly assigned to control and experimental conditions completed the Profile of Mood States and donated a 10 ml blood sample before and after an 8 -- week GIM intervention period. There were no group differences in potential confounding variables such as age, physical activity levels, drug use, restful sleep, imaging ability, health locus of control subscale scores, or hassles intensity at baseline and the two groups showed no pre-treatment differences in anxiety, depression, or beta -- endorphin (all p's > .10). After the intervention we found that, controlling for individual differences in self-report bias (Marlowe Crowne Social Desirability Scale), experimental subjects were significantly less depressed (F( 1,5) = 6.34, p < .05) though not significantly different from controls on á-endorphin levels (F < 1.0). These preliminary findings suggest that a time-limited GIM intervention may affect depression levels in healthy individuals.


The Bonny Method of Guided Imagery and Music [GIM] was begun by Helen Bonny about 20 years ago [Bonny, 1980]. Today, fewer than 100 therapists have completed advanced training in GIM. While no experimental studies have examined treatment effects of GIM on mood, several case studies have reported such effects. Goldberg, Hoss, and Chesna [1988] described an adaptation of GIM in the treatment of emotional lability and depression in a woman with severe deficits of short-term memory resulting from brain damage. Nolan [1983] reported the case of a forensic psychiatric inmate who experienced release from repressed trauma and feelings of anger following GIM treatment. Hale [1992] described a young woman who confronted and moved through grief and fears in recovery from a mastectomy through GIM.

GIM has also been reported to affect physiological processes. McDonald [1986] described the successful discharging of parasites previously resistant to physical/chemical treatment by 16 patients who each underwent a series of GIM sessions. In one of two cases described, Pickett [1987] reported that a client experienced complete remission of uterine fibroid tumors and the return of normal menstruation following life changes made in response to GIM. Merritt [1993] reported that a client with ankylosing spondilitis experienced relief from long-standing pain following two series of GIM sessions. In an experimental study of 30 unmedicated adults with essential hypertension, McDonald [1990] found that GIM significantly reduced blood pressure. The potential mechanisms for these effects have not yet been investigated.

The endogenous opioids, a group of more than 15 neurohormones, have been found to affect pain perception, neuroendocrine and immune regulation, neurotransmission, and mood [Kalin & Loevinger, 1983]. á-endorphin, one of the endogenous opioids, has been negatively associated with mood disturbance, especially depression, in a number of studies. á-endorphin levels increase with strenuous exercise [see Sforzo, 1988 for a review], which has also been associated with enhanced mood.

Although the effect of individual GIM on á-endorphin has not been examined, effects of a group music imaging session on á-endorphin has been explored. McKinney, Tims, Kumar, and Kumar [1994] studied the effects of music imaging, silent imaging, and music listening on plasma á-endorphin in 83 undergraduates. Each of four groups of subjects participated in a different experimental condition. Those in the two imagery conditions participated in identical physical relaxation and imagery inductions. The music imaging condition then heard music to facilitate their imagery while the silent imaging group engaged in imagery without music. The music listening group listened to the same piece of music as those in the music imaging group without the induction. The fourth group was a measurement -- only control group. Two blood samples were drawn approximately 2 hours apart, one at the beginning and one at the end of each experimental session. Statistically controlling for individual differences in pretest á-endorphin, the authors found that á-endorphin levels were lowered significantly in the music imaging group and that posttest levels of á-endorphin were significantly lower in the music imaging group than in any other group.

Goldstein [1980] suggested that endogenous opioids may be mediators for the "thrills" experienced by some music listeners since such thrills were diminished by naloxone, an opiate receptor antagonist. Halpaap et al. [cited by Spintge & Droh, 1987] found that women who listened to music during labor experienced lower levels of both á-endorphin and adrenocorticotropic hormone (ACTH), a neurohormone released by the pituitary under stress. Satoh et al. [1983], however, found no significant decline in á-endorphin or other plasma stress hormones in surgical patients who listened to music.

Although all of the studies of music imaging and music listening have examined only acute effects of these experiences, together they indicate that an intervention that synergistically combines music and imagery may affect neurohormone levels, including á-endorphin. Given the evidence that GIM affects both mood and physiologic processes, and the previously established relationship between depression and á-endorphin, we explored the simultaneous effect of GIM on depression and plasma á-endorphin. To this end, we hypothesized that subjects in the GIM condition would demonstrate significant reductions in depressed mood and resting levels of plasma á-endorphin as compared to those in an assessment -- only control condition.



Ten adult volunteers between the ages of 21 and 45 were recruited from a large metropolitan community. In order to minimize sources of extraneous variance in psychological and physiological measures, potential subjects were excluded from participation in the study on the following grounds: (a) use of antibiotics or tranquilizers in the 2 weeks prior to the onset of the study or cortisone or other anabolic steroids within six months prior to the study, (b) use of recreational drugs within 2 months prior to the study, (c) cigarette smoking, (d) alcohol intake exceeding ten drinks per week, (e) pregnancy or lactation at the time of the study or childbirth within 6 weeks prior to the onset of the study, (f) psychiatric or other significant medical illness within 3 months prior to the study, (g) history of chronic mental or physical illness, (h) skin reactive disease such as eczema, (i) symptoms of acute infection (e.g., fever, sore throat, nasal congestion), and (g) previous experience of 2 or more GIM sessions.

Subjects were assigned randomly to either the experimental or control condition. Of the initial 10,1 subject (from the control group) withdrew from the study for health reasons. A second subject (from the experimental group) attended all sessions, but refused to engage in the experimental procedure. Posttest data for this subject were not included in the analysis. The following data are based on 4 subjects in each of the experimental and control groups. The experimental group included 3 males and 1 female.


The experiment consisted of the following three phases: (a) pretest, (b) intervention, and (c) posttest. The control group provided pre- and posttest measurements only.

Pretesting occurred at 7:30 a.m., because plasma levels of á-endorphin have been shown to have a circadian rhythm with peak levels in the early morning [Iranmesh et al.,1989]. Subjects were instructed not to eat or drink anything except water after the previous midnight. At the pretest session, all subjects completed the Profile of Mood States [McNair, Lorr, & Droppleman, 1981], the Multidimensional Health Locus of Control Scale [Wallston, Wallston, & DeVellis, 1978], the Marlowe -- Crowne Social Desirability Scale [Crowne & Marlowe, 1964], the Hassles Scale, [Kanner et al., 1981], and the Creative Imagination Scale [Wilson & Barber, 1978]. The last is a 10-task instrument which measures a subject's imaging ability through a variety of imaginary experiences. In addition, blood samples were collected by venipuncture for á-endorphin assay (described in a later section).

Following collection of all pretest materials, each subject was informed of their group assignment. Subjects in the control group were given an appointment date for posttesting (approximately 8 weeks following pretesting) and were informed that they could undergo the treatment following that date. Subjects in the experimental group were given a weekly appointment time during which to participate in the experimental intervention.

Intervention. At the initial experimental session, GIM was defined as a method of self-exploration which involves listening to specifically designed sequences of selected classical music to allow images to come to conscious awareness and sharing the experience with a guide. It was explained that during this and subsequent sessions, subjects would listen to music while in a focused state to allow images to emerge and to experience these images as fully as they were able. During this session, subjects were informed of the variety of experiences possible and a broad definition of images which included feelings, physical sensations, sensory -- kinesthetic experiences, and memories was offered. GIM was differentiated from relaxation sessions which might also involve music and imagery. This was done by informing participants that although sometimes GIM is relaxing, it includes a broad range of experiences, some of which may not be comfortable. They were told that it is the purpose of GIM to be with these images to the extent that one is willing to allow change to occur in them.

Subjects participated in a total of six individual GIM sessions, approximately one per week, each lasting 1 1/2 to 2 hours. During these sessions, the subject spoke aloud, describing the imagery experience as it occurred. The experimenter kept a duplicate transcript of each session and provided the subject with a copy at the end of the session. Mandala drawing was used to facilitate postsession processing and integration of the experience.

The posttest for both groups occurred approximately 8 weeks following the pretest and was the same as the pretest.

Physiological Measure

Control variables. In order to control for variance in á-endorphin due to extraneous sources, the following variables were measured: age, sex, amount of restful sleep during the 3 nights preceding the drawing of blood samples, medication use, and point in the menstrual cycle for female subjects.

Sample collection procedures. At both pretest and posttest, one blood sample was collected from each subject in an EDTA tube and placed on ice. These samples then were centrifuged at 4øC and the plasma collected. The plasma was transferred into eppendorph tubes which contained 10 microliters 1 aprotinin, a protease inhibitor, as a stabilizer, and then frozen at ó-70øC until examination.

Plasma beta-endorphin assay . ( 1) Plasma fractions from subjects' blood samples were assessed by immunoradiometric assay using kits from Nichols Institute according to instructions provided by the suppliers. Undiluted samples (200 microliters 1 each) were incubated in duplicate at room temperature for 20 hours simultaneously with (125)I-labeled antibody and with antibody immobilized on a plastic bead. Beta-endorphin present in the sample binds between the two antibodies, forming a solid phase labeled antibody/á-endorphin/antibody complex. Beads were then washed thrice with a wash solution reconstituted from the wash reagent concentrate provided and 9 parts distilled water. After the third complete aspiration of all liquid from each tube, radioactivity bound to the bead was measured with an LKB 1272 CliniGamma counter. The average count per minute (CPM) of duplicate sets was recorded. A standard curve was generated by computer using the 6 pairs of standard tubes provided which ranged from 0 pg/ml to 5000 pg/ml. The average CPM of each standard level was plotted on the ordinate against the standard concentration on the abcissa of a semi-logarithmic graph. Results were calculated from the curve by an in-house computer and were expressed in picograms per milliliter. The sensitivity of this assay is 14 pg/ml. The coefficient of variation for the intraassay variance is 4.1-4.4% and for the interassay variance is 7.7-9.0%.


This study utilized a completely randomized one-way design with two levels to explore the effect of the Bonny Method of guided imagery and music on depressed mood and plasma á-endorphin. Dependent variables were scores on the Depression/DeJection scale of the Profile of Mood States [POMS] and plasma á-endorphin levels.

Control Variables

In order to rule out possible effects of confounding variables on á-endorphin, we computed a Pearson correlation coefficient between The á-endorphin and each of the following: sex, level of physical activity, frequency of daily stressors, intensity of daily stressors, and amount of restful sleep. No subject reported any current medication use. Correlations for these variables were all r < .30. However, the correlation between age and á-endorphin at pretest was r = .35 (N = 8) and between point in the menstrual cycle for female subjects and á-endorphin at the time of pretest was r = -.49 (N = 4).

An analysis of variance (ANOVA) was computed for each control variable to assure that there were no significant differences between groups prior to treatment. As shown in Table 1, the groups were statistically equivalent on age, intensity of daily hassles, imaging ability, health locus of control subscales and pretest level of anxiety (all p's >. 10). The control group reported a significantly higher frequency of hassles at the time of pretest, although this difference was not seen at posttest or in the pre- to posttest change scores.

Since there is empirical and theoretical justification for controlling the influence of environmental stressors and self-report response style on self-report measures of mood, both Hassles Frequency scores and Marlowe-Crowne Social Desirability Scale scores were used as covariates in the analyses of between group differences in depression. Homogeneity of regression slopes was confirmed for each covariate with pretest Depression/Dejection subscale scores (F's( 1,4) < 1.00, p's > .10). The groups showed no significant difference at pretest on either depression (F( 1,4) = .02, n.s.) or á-endorphin (F( 1,6) = .10, n.s.).

Means and standard deviations for dependent variables are presented in Table 2. ANOVA or ANCOVA as appropriate was computed to determine the effect, if any, of GIM on each dependent variable. In the following presentation of results, each hypothesis will be considered separately.

Hypothesis 1. As noted above, Hassles Frequency and Marlowe-Crowne Social Desirability Scale scores were used as covariates with Depression/Dejection subscale scores. The assumption of homogeneity of regression slopes was met for both covariates with posttest Depression/Dejection scores (F's( 1,4) < 1.00, p > .10). The ANCOVA for posttreatment group differences in depression and an examination of adjusted group means indicate that the experimental group demonstrated significantly less depression at posttest than the control group (F( 1,4) = 12.06, p = .03). Calculation of omega squared showed that the experimental condition accounted for 43% of the variance in postest level of depression.

Hypothesis 2. There was no significant difference between groups at posttest in levels of plasma á-endorphin (F( 1,6) = .75, n.s.). 2 A Pearson correlation coefficient computed between depression and á-endorphin revealed that these two measures correlated significantly at pretest (r = .77, p < .05). Although this relationship was not found at posttest (r = .19, n.s.), an examination of correlations for each sex revealed that the posttest correlation between á-endorphin and depression for males was significant (r = .98, p < .05, n = 4) while that for females was not (r = .23, n.s., n = 4).


In this study we explored the effect of GIM on depression and plasma á-endorphin in healthy adults. We found that a series of GIM sessions significantly mitigates depressed mood. There was no significant effect by GIM on á-endorphin as measured in this study.

The fact that depression was significantly changed while plasma levels of á-endorphin were not may be due to the sampling compartment for á-endorphin. Studies assaying á-endorphin from simultaneously collected samples of cerebrospinal fluid and plasma have failed to show any relationship between the central and peripheral levels [Kalin & Loevinger, 1983]. While peripheral levels have been reported to reflect acute effects of a variety of stressors including strenuous exercise, pregnancy, and surgical stress [Kalin & Loevinger, 1983], as well as of music imaging [McKinney et al., 1994] and music listening [Halpaap et al., cited by Spintge & Droh, 1987], the relationship between these influences and the central levels of á-endorphin is unknown. Naber and Pickar [1983] point out that studies of persons with clinical depression have failed to show a consistent relationship between depression and peripheral levels of á-endorphin. A relationship has been demonstrated between cerebrospinal fluid (CSF) levels of endorphins and abnormalities of the hypothalamic-pituitary-adrenocortical axis, abnormalities known to be associated with depressive illness. Thus, CSF may be the sampling compartment of choice for examining treatment effects of a psychotherapeutic intervention on á-endorphin and depression.

Two other variables which may have contributed to the absence of significant treatment effects for á-endorphin are the effect of the menstrual cycle on peripheral levels of á-endorphin and the unequal gender distribution between the experimental and control groups. While data were collected so that the variance caused by the menstrual cycle could be statistically controlled, the small number of women in the study prohibited meaningful use of these data.

This pilot study indicates it would be fruitful to pursue the effect of GIM on depression. The present preliminary findings for healthy adults warrant replication in a larger sample. In addition, the significant reduction in depressed mood for subjects in the present study points toward investigation of antidepressant effects of GIM for individuals with nonpsychotic depressive disorders.

Further work is needed to explore the effect of GIM on plasma á-endorphin. In order to take advantage of the most consistently documented relationship between á-endorphin and depression, á-endorphin assays ideally would be conducted from samples of CSF. Two directions might be taken to control for the variance in á-endorphin caused by the menstrual cycle of females. One approach would be to select only men. This possibility has some support given the change in depression effected by GIM and the high correlation between posttest depression and á-endorphin in the male subjects in this study. However, this approach would limit generalizability of any findings to men. An alternative approach would be to select a sufficiently large sample of subjects to allow for statistical control of the influence of the menstrual cycle on á-endorphin.

Additional studies examining other mood variables in both healthy and clinical populations are warranted. Well-controlled studies of effects of GIM on neuroendocrine measures in a larger sample may also prove fruitful. Given the well-documented relationship between clinical depression and cortisol [Gold, Goodwin, & Chrousos, 1988; Sachar, 1967] and the antidepressant effects found in the small sample of healthy individuals examined in this study, effects of GIM on cortisol also may be indicated.

Because of the small sample size, all of the findings of this study must be approached with caution. Nevertheless, these results suggest that GIM may improve depressed mood in healthy individuals. In addition, they offer both direction and encouragement for further exploration of treatment effects of GIM.

1 Since the antibodies used in this assay cross-react with á-lipotropin, the precursor protein of á-endorphin, levels of á-endorphin should be considered to be "á-endorphin-like immunoreactivity."

2 Although self-report data indicating day of menstrual cycle were collected from female participants, we were unable to analyze a statistical relationship between this variable and á-endorphin in any meaningful way due to the small number of women in


confounding variables - are potential extraneous contributors to the variance in a dependent variable

cortisol - is a hormone produced by the cortex of the adrenal gland at the direction of the hypothalamus (in the brain) via the pituitary gland. Cortisol is a product of the physiological stress response associated with stressors over which the individual has or perceives him/herself to have no active coping methods available.

endogenous opioids - are naturally occurring chemical substances that "act on the brain to decrease the sensation of pain" [Thomas, 1989, p. 1260].

neurohormone - a chemical substance produced by the body and transported in the blood stream which either is secreted as a result of nervous stimulation, or which affects nervous system function [Thomas, 1989].

opiate receptor antagonist - is a substance which blocks activity of both opium-related drugs and endogenous opioids at the receptor sites on cell surfaces.

protease inhibitor - is a substance which slows the action of protein-splitting enzymes.


Bonny, H. [1980]. GIM therapy: Past, present, and future implications. Salina, KS: The Bonny Foundation.

Crowne, D.P, & Marlowe, D. [1964]. The approval motive: Studies in evaluative dependence. New York: John Wiley.

Gold, P.W., Goodwin, E K., Chrousos, G. P. [1988]. Clinical and biochemical manifestations of depression: Relationship to the neurobiology of stress, part 2. New England Journal of Medicine, 319, 413-420.

Goldberg, E, Hoss, T. M., & Chesna, T. [1988]. Music and imagery as psychotherapy with a brain damaged patient: A case study. Music Therapy Perspectives, 5, 41-45.

Goldstein, A. [1980]. Thrills in response to music and other stimuli. Physiological Psychology, 8, 126-129.

Hale, S. E. [1992]. Wounded woman: The use of guided imagery and music in recovering from a mastectomy. Journal of the Association for Music and Imagery, 1, 99-106.

Horowitz, M., Wilner, N., & Alvarez, W. [1979]. Impact of Events scale: A measure of subjective stress. Psychosomatic Medicine, 41, 209-218.

Kalin, N.H., & Kalin, N.H. & Leovinger, B. L. [1983]. The central and peripheral opiod peptides: Their relationships and functions. Psychiatric Clinics of North America, 6, 415-428.

Kanner, A.D., Coyne, J.C., Schaefer, C., & Lazarus, R.S. [1981]. Comparison of two models of stress measurement: Daily hassles and uplifts versus major life events. Journal of Behavioral Medicine, 4, 1-39.

McDonald, R. G. [1986]. Healing parasitic infection through the partnership of GIM and applied kinesiology. Unpublished fellow's paper. Salina, KS: The Bonny Foundation.

McDonald, R. G. [1990]. The efficacy of guided imagery and music as a strategy of self-concept and. blood pressure change among adults with essential hypertension. Unpublished doctoral dissertation. Mpls, MN: Walden University.

McKinney, C. H., Tims, E C., Kumar, A., & Kumar, M. [1994]. Music, imagery. and plasma beta endorphin. Manuscript submitted for publication.

McNair, D., Lorr, M., Droppleman, L.E [1981]. EITS manual for the profile of mood states. San Diego: Educational and Industrial Testing Service.

Merritt, S. [1993]. The healing link: Guided imagery and music and the body/mind connection. Journal of the Association for Music and Imagery, 2, 11-28.

Nichols Institute Diagnostics. [1989]. Allegra beta-endorphin: Immunoassay for the quantitative determination of human beta endorphin levels inserum or plasma. San Juan Capistrano, CA: Author.

Nolan, P. [1983]. Insight therapy: Guided imagery and music in a forensic psychiatric setting. Music Therapy, 3, 43-51.

Pickett, E. [1987]. Fibroid tumors and response to guided imagery and music: Two case studies. Imagination. Cognition and Personality, 7, 165-176.

Sachar, E.J. [1967]. Corticosteroids in depressive illness. Archives of General Psychiatry, 17, 544-567.

Spingte, R. & Droh, R. [1987]. Effects of anxiolytic music on plasma levels of stress hormones in different medical specialties. In R. R. Pratt (Ed.), The fourth international symposium on music: Rehabilitation and human well-being, pp. 88-101. Lanham, MD: University Press of America.

Satoh, Y., Nagao, H., Ishihara, H., Oyama, T., & Spintge, R. [1983]. Effects of music on plasma stress hormones in surgical patients. Masui, 32, 1211.

Sforzo, G.A. [1988]. Opioids and exercise: An update. Sports Medicine, 7, 109-124.

Thomas, C. L. (Ed.). [1989]. Taber's Cyclopedic Medical Dictionary (16th ed.). Philadelphia: F.A. Davis.

Wallston, K.A., Wallston, B.S., & DeVellis, R. [1978]. Development of the multidimensional health locus of control (MHLC) scales. Health Education Monographs, 6, 160-170.

Wilson, S.C. & Barber, T.X. [1978]. The Creative Imagination Scale as measure of hypnotic responsiveness: Applications to experiemental and clinical huypnosis. The American Journal of Clinical Hypnosis, 20, 235-249.

AMI Publications.


By Cathy H. McKinney; Michael H. Antoni; Adarsh Kumar and Mahendra Kumar

Share this with your friends