Massage and relaxation therapies' effects on depressed adolescent mothers.

Field, Tiffany ; Grizzle, Nancy ; Scafidi, Frank 等

Depression is one of the most prevalent medical disorders and has been recognized as a distinct pathologic entity from early Egyptian times. Anxiety is one of the primary features of depression in adolescents (Goldman, 1988), and relaxation therapy (RT) is usually noted to decrease anxiety (Richter, 1984). Using the State-Trait Anxiety Scale, for example, anxiety levels were found to be lower in psychiatric patients following nine sessions of relaxation therapy (Hosmand, Helmes, Kazarian, & Tekatch, 1985). Even following one brief RT session, mood was elevated on the Profile of Mood States Scale (Matthew & Gelder, 1969).

In a longer term outcome study, RT was as effective as psychotherapy and pharmacotherapy in reducing anxiety (McLean & Hakistian, 1979) and even more effective than cognitive behavior therapy (Reynolds & Coats, 1986). Similarly, in a study of depressed child and adolescent psychiatric patients, both groups benefitted from as little as one hour of relaxation therapy (Platania-Solazzo, Field, Blank, Seligman, Kuhn, Schanberg, & Saab, 1992). In that study, self-reported anxiety as well as anxious behavior and fidgeting decreased, and increases were noted in positive affect.

Massage therapy (MT), in contrast, was used in only one study in the literature. In that study child and adolescent psychiatric patients had lower anxiety levels following five massage therapy sessions as well as more optimal affect and sleep patterns and lower stress hormones including cortisol and norepinephrine (Field, Morrow, Valdeon, Larson, Kuhn, & Schanberg, 1991).

The purpose of the present study was to compare the effects of massage and relaxation therapies on anxiety and depression in a sample of depressed adolescent mothers. Maternal depression ranges from 25-30% during the first three months after delivery (O'Hara, Neunaber, & Zekoski, 1984). Even mild depression and anxiety may affect the new mother's relationship with her child. For example, in one study, postpartum depressed mothers demonstrated less rocking, gaze, and positive regard toward their infants than did nondepressed mothers (Livingood, Dean, & Smith, 1983). Others have reported less frequent positive and more frequent negative states among depressed mother-infant dyads (Cohn, Campbell, Matias, & Hopkins, 1990; Field, 1992; Field, Healy, Goldstein, & Guthertz, 1990). Massage and relaxation therapy were expected to decrease the mothers' depressive and anxiety symptoms.

METHOD

The sample was comprised of 32 depressed adolescent mothers who had recently given birth at a large inner-city hospital and were recruited from the hospital's maternity ward. The subjects were randomly assigned to the massage therapy or relaxation therapy group. The groups did not differ on age (M = 18.1), years of education (M = 10.4), ethnicity (71% black, 29% Hispanic), or SES (M = 4.7 on the Hollingshead Index). To qualify for the study the mothers needed to have an elevated Beck Depression Inventory (BDI) score and to be free of current medication or other treatment for depression or related disorders. The depression classification was based on a diagnosis of dysthymia on the Diagnostic Inventory Schedule (Costello, Edelbrock, & Costello, 1985) and a score greater than 16 on the Beck Depression Inventory (Beck, Ward, Mendelson, Mock, & Erbaugh, 1961).

Treatment Procedures

Massage therapy. The massage therapy subjects (N = 16) received a 30-minute massage per day on two consecutive days per week for five consecutive weeks (10 massages). For the first 15 minutes the subjects were in a supine position for massage in the following four regions: (1) head/neck - slow lateral stroking of forehead, followed by long and slow stroking from the neck across the shoulders and from the shoulders to the neck; (2) arms/hands - long and slow stroking from above the shoulder to beyond the hand, followed by individual stroking of each hand; (3) torso - placing the hand gently on the solar plexus (base of chest) and adding a gentle rocking motion; and (4) legs/feet - long, slow stroking from the hip to beyond the foot, followed by stroking each foot. For the second 15 minutes the subject was in a prone position. This segment consisted of an achilles tendon (ankle) stretch (bend the knee to stretch), knee and calf shake, long and slow strokes from over the buttocks to the toes, lateral lumbar stretch (10 times) including strokes parallel to the spine from the base of the spine to the shoulders and along the arms to past the end of the hands, trapezius (shoulder blade) squeeze, friction alongside the spine with the ulnar (outside) edge of the hand from superior to inferior, posterior neck squeeze and stretch, and long, slow strokes from the head down the entire posterior surface of the body to the toes (3 times). The massage therapy was administered by trained massage therapists and performed at the same time of day (mid-afternoon) over the 5-week period.

Relaxation therapy. The relaxation therapy subjects (N = 16) spent the same amount of time in RT as the massage therapy subjects spent in MT. They attended 30-minute relaxation therapy sessions on two consecutive days a week for five consecutive weeks (10 sessions). The subjects used exercise mats, and the first 15 minutes of the session consisted of yoga exercises. The second 15-minute segment consisted of progressive muscle relaxation. For this purpose the subjects were instructed to breathe deeply for several minutes. The instructor then asked the subjects to relax and tense eight different muscle groups beginning with the feet and ending with the head. The same measures as those used in the massage/control conditions were applied in the relaxation therapy over the same time frame.

Assessment Procedures

The effects of the massage and relaxation therapies were assessed on the first and last days of the assessment period as follows: (a) Behavior Observation Scales (Platania-Solazzo, Field, Blank, Seligman, Kuhn, Schanberg, & Saab, 1991) including seven behavior ratings (state, affect, activity level, anxiety level, fidgeting/nervous behavior, vocalizations, and cooperation) by a psychology graduate student; (b) Profile of Mood States (POMS; McNair, Lorr, & Droppleman, 1971); (c) State Anxiety Inventory for Children (STAIC; Spielberger, Gorsuch, & Lushene, 1970); (d) pulse rate; (e) saliva samples of cortisol; and (f) urine sample for cortisol. On the first and last days of the treatment period the baseline, during session and post-session assessment measures were collected according to the following schedule: (a) 30 minutes prior to the treatment condition: STAIC and POMS; (b) immediately prior to the treatment condition: pulse rate, saliva samples, and behavior observation ratings based on the previous 30 minutes of the subjects' behavior; (c) immediately after the session: pulse rate, STAIC, POMS, behavior ratings based on the previous 30 minutes of behavior, saliva, and urine samples were taken; (d) thirty minutes after the session: because of the 20-minute lag time, saliva samples were taken for cortisol assays 30 minutes after the session to tap the cortisol levels at the end of the session. These samples were frozen and sent to Dr. Saul Schanberg's laboratory at Duke University for analysis.

Behavior observation scales and activity level. Behavior observation ratings were completed three times based on behavior observed: (a) during the thirty minutes prior to the sessions (pre); (b) during the session (during); and (c) during the thirty minutes after the sessions were completed (post). This seven-behavior rating scale was originally used to assess behavior following relaxation therapy classes (Platania-Solazzo et al., 1991). In that study the rating scale reliably discriminated the behaviors of depressed and adjustment disorder children/adolescents before and after relaxation therapy classes. To complete this measure, the behavior of each subject is rated by an observer on a 3-point continuum on seven scales including state, affect, activity, anxiety, fidgeting, vocalization, and cooperation. Observers are given descriptors and examples of each level of the seven scales. Interobserver agreement was assessed by two independent observers (psychology graduate students) recording simultaneously across the observation period for one-fourth of the subjects. Interobserver reliability was calculated by Kappa to correct for chance agreement.

The Profile of Mood States (POMS: McNair, Lorr & Droppleman, 1971). The POMS is a 5-point adjective rating scale asking the subject to describe how well an adjective describes his/her feelings for the past two weeks. The POMS consists of 65 items such as friendly or lonely. Only the 14 items that comprise the depression factor were used in this study. This questionnaire was individually administered prior to and immediately after the massage/relaxation therapy session and required approximately five minutes to complete.

The State Anxiety Inventory for Children. The STAIC is an adaptation of the State/Trait Anxiety Inventory specifically designed for the study of anxiety in school-age children and adolescents who are below average in reading level. The inventory consists of 20 items such as "I feel: very relaxed, relaxed or not relaxed." The test requires approximately five minutes to complete and was administered prior to and immediately after the massage/relaxation therapy sessions.

Physiological and biochemical measures. Pulse rate was measured by taking the subject's radial pulse for thirty seconds prior to, following, and thirty minutes after the end of each session. Cortisol samples were collected at the same times. Due to the twenty-minute lag in cortisol change, saliva samples reflected cortisol levels at twenty minutes prior to the session, ten minutes into the session, and ten minutes after the end of the session.

Saliva cortisol samples were obtained by having subjects place a cotton dental swab dipped in sugar-free lemonade crystals along their gumline for thirty seconds. The swab was drawn into a syringe and then the saliva was inserted into a microcentrifuge tube. The samples were frozen and assayed for cortisol levels at Duke University.

Urine samples were collected under the supervision of a psychology graduate student. The samples were frozen and also sent to Duke University for an assay of cortisol.

RESULTS

T-tests were first conducted to ensure that the groups were comparable on demographic variables including maternal age, SES, ethnicity, and initial BDI scores. It was found that the groups did not differ on any of these variables. Group by repeated measures MANOVAs were conducted to compare the immediate and longer term effects of the massage and relaxation therapies. The two repeated measures were the therapy periods (pre and post) and the first-last day assessments. Post hoc comparisons were made by Bonferroni t tests.

A significant MANOVA followed by several significant group by repeated measures interaction effects on the ANOVAs suggested the following: (1) state anxiety decreased for the massage therapy group following the first and last day therapy sessions and only on the first day for the relaxation group; (2) lower POMS depression scores following the massage on the first and last days; (3) group by repeated measures interaction effects were significant for all of the Behavior Observation Scale ratings except activity level. These favored the massage therapy group including: (a) a higher state rating after therapy on day 10; (b) higher vocalization ratings after therapy on days 1 and 10; (c) lower anxiety ratings after the massage therapy sessions on days 1 and 10; (d) higher cooperation ratings for the MT group after therapy on days 1 and 10; (e) fidgetiness decreased for the MT group on both the first and last days; (4) decrease in pulse after massage therapy on days 1 and 10; (5) lower salivary cortisol levels after massage therapy; and (6) lower urine cortisol levels on the last day versus the first day of massage therapy. Table 1 Means for Massage Therapy Group Day 1 Day 10 Pre Post Pre Post State Anxiety Scale 35.67(4) 28.67 33.89(3) 28.28 POMS Depression 19.44(1) 11.43 20.39(3) 9.06 Behavior Observation Scale -State 2.22 2.29 2.09 2.11 -Affect 1.74(4) 2.67 1.67(4) 2.83 -Activity 1.61 1.67 1.69 1.50 -Vocalization 1.44(1) 1.83 1.50(2) 1.94 -Anxiety 1.83(4) 1.06 1.94(4) 1.06 -Cooperation 2.50(2) 2.89 2.17(4) 2.89 -Fidgetiness 1.61(2) 1.11 1.61(3) 1.11 Pulse 85.06(4) 77.76 87.88(4) 77.65 Saliva Cortisol 1.21(2) .84 1.55(1) 1.16 Urine Cortisol 166.95(1) 120.08 Superscript indicates significant differences between adjacent means [at] 1 = .05, 2 = .01, 3 = .005, 4 = .001

DISCUSSION

The relaxation group reported lower anxiety levels on the first day but only the massage therapy group reported less anxiety and showed less anxious behaviour and lower stress hormone (cortisol) levels after their sessions. In addition, only the massage therapy group experienced a reduction in depression and in stress (as manifested by their lower urinary cortisol levels) across the course of the study. These results are consistent with several studies in the literature using self-reports to tap anxiety levels following relaxation therapy (McLean & Makistian, 1979; Platania-Solazzo et al., 1990; Reynolds & Coats, 1986). The massage therapy data are consistent with the only study published on massage therapy showing not only a reduction in anxiety and depression by self-report but by behavior observation and cortisol levels (Field et al., 1991). Possible explanations for the differential effects are that a reduction in self-reported anxiety may be a simple placebo effects or the subject may have felt less anxious after a therapy designed to make them relax but did not show the same effects in their behavior or stress hormones because they had to work during the relaxation therapy. Anecdotal reports from the adolescent mothers were that they did not enjoy the relaxation therapy because they had to "work too hard." This may explain why they had lower state and cooperation ratings. Table 2 Means for Relaxation Therapy Group Day 1 Day 10 Pre Post Pre Post State Anxiety Scale 34.73(1) 31.55 33.09 31.73 POMS Depression 18.37 19.40 18.56 17.12 Behavior Observation Scale -State 2.20 2.10 2.45 2.00 -Affect 2.55 2.45 2.64 2.55 -Activity 1.82 1.82 1.91 1.73 -Vocalization 1.82 1.64 2.09 1.86 -Anxiety 1.55 1.55 1.18 1.09 -Cooperation 2.82 2.55 2.55 2.64 -Fidgetiness 1.36 1.00 1.09 1.00 Pulse 84.00 85.56 73.78 75.78 Saliva Cortisol 1.55 1.65 1.87 1.30 Urine Cortisol 165.50 168.39 Superscript indicates significant differences between adjacent means [at] 1 = .05

The long-term effect of massage therapy on urinary cortisol but not on self-report or observed behavior might relate to urinary cortisol reflecting a cumulative effect of the massage therapy that cannot be controlled by the subject. In contrast, the behavior and self-report could be controlled by the subject and may have reflected their current state at the time of the assessment. Many of the subjects expressed disappointment when the massage therapy ended. This might explain their behavior and why their self-report did not show the cumulative benefit of the urinary cortisol levels.

Future research might assess relaxation and massage therapy devotees to avoid the bias against "relaxation therapy as work." A more passive form of relaxation therapy could also be used as a comparison. In addition, measures could be taken at the penultimate session to avoid the disappointment bias of the last session. The use of convergent self-report, behavioral, physiological, and biochemical measures led to different conclusions than the literature would suggest. But additional research would be needed before concluding that massage therapy has more positive and more long-lasting effects than relaxation therapy for reducing depression and anxiety in depressed adolescent mothers.

The authors thank the mothers who participated in this study and the research assistants who helped collect the data. This research was supported by an NIMH Research Scientist Award (#MH00331), an NIMH Research Grant (#MH46586) to Tiffany Field and by funds from Johnson and Johnson.

REFERENCES

Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory for measuring depression. Archives of General Psychiatry, 4, 561-571.

Cohn, J. F., Campbell, S. B., Matias, R., & Hopkins, J. (1990). Face-to-face interactions of postpartum depressed and nondepressed mother-infant pairs at two months. Developmental Psychology, 26, 185-193.

Costello, E. J., Edelbrock, C. S., & Costello, A. J. (1985). Validity of the NIMH Diagnostic Interview Schedule for children: A comparison between psychiatric and pediatric referrals. Journal of Abnormal Child Psychology, 13, 579-595.

Field, T. (1992). Infants of depressed mothers. Development and Psychopathology, 4, 49-66.

Field, T., Healy, B., Goldstein, S., & Guthertz, M. (1990). Behavior state matching in mother-infant interactions of nondepressed vs. depressed mother-infant dyads. Developmental Psychology, 26, 7-14.

Field, T., Morrow, C., Valdeon, C., Larson, S., Kuhn, C., & Schanberg, s. (1991). Massage reduces anxiety in child and adolescent psychiatric patients. Journal of the American Academy of Child and Adolescent Psychiatry, 31, 125-131.

Goldman, H. H. (1988). General psychiatry (p. 340). Norwalk, CT: Appleton & Lange.

Hosmand, L., Helmes, E., Kazarian, S., & Tekatch, G. (1985). Evaluation of a relaxation training program under medical and nonmedical conditions. Journal of Clinical Psychology, 41, 23-29.

Livingood, A. B., Dean, P., & Smith, B. D. (1983). The depressed mother as a source of stimulation for her infant. Journal of Clinical Psychology, 39, 369-375.

Matthew, A. M., & Gelder, M. G. (1969). Psychophysiological reactions to stress. In N. Schneiderman, & J. Tapp (Eds.), Behavioral medicine: The biopsychosocial approach (pp. 220-238). Hillsdale, NJ: Erlbaum.

McLean, P. D., & Hakistian, R. A. (1979). Clinical depression: Comparative efficacy of outpatient treatments. Journal of Consulting and Clinical Psychology, 47, 818-836.

McNair, D., Lorr, M., & Droppleman, L. (1971). Profile of Mood States Manual. San Diego, CA: Educational & Industrial Testing Services.

O'Hara, M. W., Neunaber, D. J., & Zekoski, E. M. (1984). Prospective study of postpartum depression: Prevalence, course, and predictive factors. Journal of Abnormal Psychology, 93, 158-171.

Platania-Solazzo, A., Field, T., Blank, J., Seligman, F., Kuhn, C., Schanberg, S., & Saab, P. (1992). Relaxation therapy reduces anxiety in child/adolescent psychiatry patients. Acta Paedopsychiatrica, 55, 115-120.

Reynolds, W., & Coats, K. (1986). A comparison of cognitive-behavioral therapy and relaxation training for treatment of depression in adolescents. Journal of Consulting and Clinical Psychology, 54, 653-660.

Richter, N. C. (1984). The efficacy of relaxation therapy with children. Journal of Abnormal Child Psychology, 12, 319-344.

Spielberger, C. D., Gorsuch, R. L., & Lushene, R. E. (1970). The State-Trait Anxiety Inventory. Palo Alto, CA: Consulting Psychologists Press.

Nancy Grizzle, B.A., Research Associate, Touch Research Institute.

Frank Scafidi, Ph.D., Assistant Research Professor, Touch Research Institute. University of Miami School of Medicine.

Saul Schanberg, Ph.D., M.D., Professor of Pharmacology/Biological Psychiatry, Duke University Medical Center.