Share this article on:

Music and Preoperative Anxiety: A Randomized, Controlled Study

Wang, Shu-Ming MD*,; Kulkarni, Lina MD†,; Dolev, Jackqulin MD‡, and; Kain, Zeev N. MD§

doi: 10.1213/00000539-200206000-00021

Music may decrease the anxiety experienced by patients before surgery. Previous studies of this issue were hindered with multiple methodological problems. In this investigation, we examined this hypothesis while using a rigorous study design and objective outcome measures. Adult patients undergoing anesthesia and surgery were randomly assigned to two study groups. Subjects in Group 1 (n = 48) listened to a 30-min patient-selected music session, and subjects in Group 2 (n = 45) received no intervention. By using self-report validated behavioral (State-Trait Anxiety Inventory) and physiological measures of anxiety (heart rate, blood pressure, and electrodermal activity and serum cortisol, epinephrine, and norepinephrine), patients were evaluated before, during, and after administration of the intervention. We found that after intervention, subjects in the Music group reported significantly lower anxiety levels as compared with the Control group (F1,91 = 15.4, P = 0.001). That is, the postintervention anxiety level of subjects in the Music group decreased by 16% as compared with the preintervention level, whereas the anxiety level of the Control group did not change significantly. Two-way repeated-measures analysis of variance performed for the electrodermal activity, blood pressure, heart rate, cortisol, and catecholamine data demonstrated no group difference and no time × group interaction (P = not significant). In conclusion, under the conditions of this study, patients who listened to music before surgery reported lower levels of state anxiety. Physiological outcomes did not differ, however, between the two study groups.

Departments of *†§Anesthesiology, §Pediatrics, and §Child Psychiatry, Yale University School of Medicine, New Haven, Connecticut; and ‡Department of Internal Medicine, Stanford University School of Medicine, Palo Alto, California

Supported in part by the National Institutes of Health (NICHD, Grant R01HD37007-01) and the Patrick and Catherine Weldon Donaghue Medical Research Foundations.

January 23, 2002.

Address correspondence and reprint requests to Zeev N. Kain, MD, Department of Anesthesiology, Yale School of Medicine, 333 Cedar St., New Haven, CT 06510. Address e-mail to kain@biomed.

Each year, approximately 50 million surgical procedures are performed in the United States (1). Reports indicate that a significant number of adult patients undergoing surgical procedures experience high levels of anxiety before surgery (2). The phenomena of perioperative anxiety are directly related to fear surrounding an unfamiliar environment, loss of control, and fear of death and disfigurement. Currently, modalities such as preoperative premedication and psychological preparation programs are being used to treat preoperative anxiety in adult patients.

Music has well established psychological effects, including the induction and modification of moods and emotions (3). Thus, it is no surprise that music has become a modality grouped with many alternative medical therapies that have evolved into a cluster known as cognitive therapies (3). Currently, music is being promoted by the American Music Therapy Association as a treatment modality for conditions such as dementia, stroke, Parkinson’s disease, various affective disorders, and pain (4). A recent article published in JAMA has emphasized, however, the strong need for objective outcome data to establish the therapeutic efficacy of this modality (4).

The role of music as a therapeutic modality for the treatment of preoperative anxiety in adult patients has been previously evaluated (5–8), and an editorial in the British Journal of Surgery concluded that the anxiolytic effects of music before surgery are well documented (9). A close examination of previous studies reveals, however, that these investigations are hindered by multiple methodological problems, such as small sample size, lack of a control group, no randomization, selection bias, nonobjective outcome mea-sures, and nonstandardized protocols. Thus, using an appropriate sample size, a randomized control study design, and objective behavioral and physiological outcome measures, we decided to reexamine the effects of preoperative music on preoperative anxiety.

Back to Top | Article Outline


This randomized, controlled study was conducted with consecutive patients aged 18–65 yr, ASA physical status I–III, scheduled to undergo anesthesia and elective outpatient surgery. Sedative premedication was not offered to patients participating in this study. The primary outcome was preoperative anxiety, as examined by a self-reported psychological instrument and physiological measures. Subjects were randomly assigned to receive either music delivered via headphones (Intervention group) or no music (Control group). For each patient, the experimental session lasted 30 min. A repeated-measures design was used in which each subject’s behavior was evaluated before, during, and after the intervention. The IRB approved the study protocol, and informed consent was obtained from all subjects.

The State-Trait Anxiety Inventory (STAI) is a 40-item self-report measure that contains 20 items measuring state anxiety and 20 items measuring trait anxiety (10). Total scores for state and trait sections separately range from 20 to 80, with higher scores denoting higher levels of anxiety.

The cardiovascular system is highly responsive to a variety of psychological and behavioral states (11). Both heart rate (HR) and blood pressure (BP) have been widely used as dependent variables in behavioral studies designed to alter levels of anxiety and are frequently cited as physiological indices of stress in psychology, aviation medicine, and anesthesia (11–13).

Electrodermal activity (EDA) was recorded as a measure of autonomic arousal in response to stress. EDA is a measure of change in skin conductance resulting from eccrine sweat gland activity, which is modulated by states of emotional stress (14). EDA was recorded continuously by using a Model 3992/2 Biolog® (UFI, Morro Bay, CA) ambulatory recording system. The Biolog has the capability to continuously measure five different physiological variables; for the purpose of this study, we continuously monitored EDA and HR. The EDA was measured with a constant-voltage (0.5 V) excitation skin conductance level signal conditioner and voltage as recommended by Lykken and Venables (15). EDA was sampled at 10 Hz, and skin conductance levels of 0–40.95 mmho were recorded. Recording was performed with two Ag-AgCl electrodes filled with BioGel electropotential medium (BioGel UFI 1090) and connected to the volar surface of the second and third fingers of the nondominant hand.

Blood samples were transferred immediately into a tube containing heparin. After mixing, the blood was centrifuged at 4000 rpm for 10 min, and the plasma was stored at −70°C. Plasma cortisol was determined in duplicate by use of radioimmunoassay kits from Diagnostic System Laboratories (Webster, TX). Samples were analyzed in a single large batch, duplicates agreed within 15%, and quality assessment samples were well within the manufacturer’s defined range. The time of day at which the blood sample for cortisol analysis was obtained was recorded as well.

Blood samples were transferred immediately into a tube containing antioxidant. After mixing, the blood was centrifuged at 4000 rpm for 10 min, and the plasma was stored at −70°C. Plasma catecholamines were determined by using high-performance liquid chromatography and an electrochemical detector (ESA Inc., Chelmsford, MA). Samples were analyzed in a single large batch, and quality assessment samples were well within the manufacturer’s defined range.

Subjects were contacted the night before surgery, and informed consent was obtained via the telephone. At this time, subjects were told that they might or might not have an opportunity to actually listen to music before surgery. All participants were asked to bring their favorite compact disk to the hospital the morning of surgery in case they were assigned to the Music group.

On the morning of surgery, all subjects arrived 1 h 15 min before the scheduled surgery and underwent a routine hospital admission process. After completion of the admission process by the nurses, baseline demographic data, anxiety score (STAI), and BP/HR were obtained by an investigator. An IV cannula was inserted, and a blood sample was obtained. The Biolog monitoring system was attached to the patient, and HR and EDA were monitored continuously throughout the experimental protocol. Next, patients were randomly assigned to either an Intervention group or Control group and were escorted to an isolated room. Throughout the duration of the experimental session, no hospital personnel were allowed to enter the room or communicate with the patient. All individuals accompanying the patient were allowed to stay with the patient throughout the duration of the experimental intervention, and patients were allowed to read or converse regardless of group assignment. An investigator was present outside the isolated room and ensured adherence to this protocol. Music was administered for 30 min to the Intervention group via headphones by a compact disk player (Discman ESP; Sony Corp., Tokyo, Japan). The type of music administered was selected by the patient. The Control group wore headphones with no music or white noise administered. Once the experimental session was completed, anxiety and BP were assessed again, and a second blood sample was obtained. At this point, the experimental protocol was completed, and the subjects underwent the standard-of-care anesthetic and surgical management.

On the basis of previous investigations (5–7), a sample size of 34 subjects in each group was able to detect a decrease of 20% in anxiety as assessed by changes in the STAI. This was based on a power of 90% and an α of 0.05. Because we appreciated a priori that previous investigations have overestimated the effects of music, we recruited approximately 45 patients to each experimental group. Normally distributed data are presented as mean ± sd; skewed data are presented as median and interquartile range (25%–75%). Baseline characteristics of the two groups were analyzed with Student’s t-test for continuous data and χ2 tests for categorical analysis. Two-way repeated-measures analysis of variance (ANOVA) with trait anxiety as a covariable was used to analyze the changes in the behavioral (STAI) and physiological (cortisol, EDA, HR, and systolic and diastolic BP) anxiety levels of patients along the various time points. All outcome data were normalized and expressed as a percentage from the baseline that was measured before the intervention. Because we appreciated that cortisol levels change on the basis of the time of day, we documented the time of day that the experiment was performed (Table 1). Comparisons were considered significant if P < 0.05.

Table 1

Table 1

Back to Top | Article Outline


Ninety-three patients were enrolled in this study. There were no significant differences between the two study groups regarding baseline demographics, types of surgical procedures, anesthetic techniques, or time of day the surgery was performed (Table 1). There were also no differences in the musical backgrounds or preferences of patients participating in the study (Table 2). Finally, no preintervention behavioral or physiological value differed significantly between the two study groups (Table 3).

Table 2

Table 2

Table 3

Table 3

For behavioral outcome, a two-way repeated-measures ANOVA performed for state anxiety data demonstrated a significant group × time interaction (F1,91 = 15.4, P = 0.001) (Table 4). That is, there was a significant change over time in the anxiety level of the patients on the basis of the group assignment. Post hoc analysis demonstrated that patients in the Music group were significantly less anxious after intervention as compared with patients in the Control group (P = 0.001). At the conclusion of the intervention, the anxiety level of subjects in the Music group decreased by 16% as compared with before intervention, whereas the anxiety level of the Control group did not change significantly (84% ± 15% vs 99% ± 20%) (Table 4).

Table 4

Table 4

For physiological outcomes, a two-way repeated-measures ANOVA performed for EDA data demonstrated no group difference (F1,82 = 1.7, P = 0.19) and no time × group interaction (F1,82 =1.4, P = 0.23) (Fig. 1). Similarly, no group differences (F1,82 = 0.45, P = 0.5) and no time × group interaction were found for HR data (F1,82 = 0.6, P = 0.44) (Fig. 1). Also, as can be seen in Table 4, there were no significant differences in systolic or diastolic BP after the intervention. Finally, cortisol, epinephrine, and norepinephrine levels did not change significantly as a result of the intervention (Table 4).

Figure 1

Figure 1

Back to Top | Article Outline


Under the conditions of this study, we found that patients who listened to music reported being less anxious after intervention. In contrast, there were no differences between the two study groups with regard to physiological signs of anxiety (such as BP, HR, and EDA) or neuroendocrine variables (such as cortisol, epinephrine, and norepinephrine).

The dichotomy found in this study between psychological and physiological measures in anxiety is not surprising considering existing medical-music literature. Although earlier studies reported significant changes in hemodynamic variables and EDA in response to relaxing music, later studies could not replicate this phenomenon (16). Also, similar to our findings, several investigators have reported that relaxing music is associated with lower state anxiety scores without any changes in the physiological variables assessed (17,18). This lack of correlation between psychological and physiological indicators of anxiety may be explained by trait-related individual responsiveness of the sympathetic system (16). It is interesting to note that our laboratory has observed a similar dichotomy between psychological and physiological signs of anxiety in a previous investigation that assessed the effectiveness of auricular acupuncture for the treatment of anxiety (19).

Previously, our study group reported that patients undergoing surgery under regional anesthesia required significantly less propofol for sedation if they were exposed to music during the operation (20). In this investigation, we have extended our earlier work to the preoperative period as well. Several other investigators have examined the effects of music on the anxiety of patients undergoing surgery (5–8). Unfortunately, studies conducted in this area were hindered by major methodological issues that make the conclusions of those articles invalid. The number of previous studies that have examined the effects of preoperative music on preoperative physiological outcomes is very limited. In two related studies, Miluk-Kolasa et al. (21,22) measured cortisol levels, BP, HR, skin temperature, and glucose levels in patients in conjunction with informing them that they would have to undergo surgery the following day. One group received music immediately after receipt of this information, whereas the other group received no music. These investigators found that the information about the impending surgery triggered a smaller increase of variables such as BP, HR, skin temperature, and cortisol in patients who listened to music (21,22). Although interesting, the results of these two studies may not be entirely relevant to the concept of preoperative anxiety, because patients who wait for surgery typically know about the surgery for days and weeks before surgery, and thus the psychobiology of patients on the morning of surgery is very different from that of patients who were just notified about the need to undergo surgery the next day.

The effects of music on patients undergoing procedures such as bronchoscopy and sigmoidoscopy are contradictory as well. Although some studies report that music is effective in reducing anxiety levels (23,24), others indicate that no change in the anxiety level was seen as a result of music (25,26). When one closely examines the studies that addressed the sigmoidoscopy and endoscopy population, one finds many of the methodological problems that exist in the preoperative anxiety literature. Thus, further valid research is needed in this area as well.

Several methodological issues related to this study have to be addressed. First, the study described in this article is a randomized trial that enrolled an appropriate number of subjects and used valid and objective behavioral and physiological outcome instruments. This is in contrast to most of the studies that examined this issue. Second, we have to note that, because of the type of intervention studied, subjects could not be blinded to group assignment. Third, we rejected a priori the idea of the Control group’s being exposed to white noise. Instead, we opted for subjects in both groups to be able to be involved in other activities, such as talking to their companions or reading. Both groups, however, were isolated, and no medical personnel were allowed to interact with the subjects during the study period. Also, blinding the observer was not a consideration with this study because we used physiological and self-report outcome measures. Thus, the observer could not influence the outcome of this study.

In conclusion, patients scheduled to undergo elective outpatient surgery and who listened to music reported less anxiety as compared with patients who did not listen to music. Physiological indicators, however, did not differ significantly between the two study groups. We suggest that patient-selected music can be used to reduce the anxiety of patients who are about to undergo surgery.

The authors would like to thank Paul G. Barash, MD, for his critical review of this manuscript.

Back to Top | Article Outline


1. Sobel D, Ornstein R. The healthy mind, healthy body handbook. Los Altos, CA: DRX, 1996.
2. Badner N, Nielson W, Munk S, et al. Preoperative anxiety: detection and contributing factors. Can J Anaesth 1990; 37: 444–7.
3. Standley JM. Music research in medical/dental treatment: meta-analysis and clinical applications. J Music Ther 1986; 23: 56–122.
4. Marwick C. Music therapists chime in with data on medical results. JAMA 2000; 283: 731–3.
5. Cunningham MF, Monson B, Bookbinder M. Introducing a music program in the perioperative area. AORN J 1997; 66: 674–82.
6. Gaberson KB. The effect of humorous and musical distraction on preoperative anxiety. AORN J 1995; 62: 784–8.
7. Updike PA, Charles DM. Music Rx: physiological and emotional responses to taped music programs of preoperative patients awaiting plastic surgery. Ann Plast Surg 1987; 19: 29–33.
8. Kaempf G, Amodei ME. The effect of music on anxiety: a research study. AORN J 1989; 50: 112–8.
9. Thompson JF, Kam PC. Music in the operating theatre. Br J Surg 1995; 82: 1586–7.
10. Spielberger C, Gorsuch R, Lushene R. State-Trait Anxiety Inventory manual. Palo Alto, CA: Consulting Psychologists Press, 1970.
11. Krantz D, Falconer J. Measurement of cardiovascular responses. In: Cohen S, Kessler R, Gordon Underwood L, eds. Measuring stress. New York: Oxford University Press, 1995: 193–212.
12. Roscoe AH. Stress and workload in pilots. Aviat Space Environ Med 1978; 49: 630–6.
13. Scheinin M, Scheinin H, Ekblad U, Kanto J. Biological correlates of mental stress related to anticipated caesarean section. Acta Anaesthesiol Scand 1990; 34: 640–4.
14. Lim AT, Edis G, Kranz H, et al. Postoperative pain control: contribution of psychological factors and transcutaneous electrical stimulation. Pain 1983; 17: 179–88.
15. Lykken DT, Venables PH. Direct measurement of skin conductance: a proposal for standardization. Psychophysiology 1971; 8: 656–72.
16. Hanser S. Music therapy and stress reduction research. J Music Ther 1985; 22: 193–206.
17. Davis W, Thaut M. The influence of preferred relaxing music on measures of state anxiety, relaxation, and physiological responses. J Music Ther 1989; 26: 168–87.
18. Strauser J. The effects of music versus silence of measures of state anxiety, perceived relaxation, and physiological responses on patients receiving chiropractic interventions. J Music Ther 1997; 34: 88–105.
19. Wang S, Kain Z. Auricular acupuncture: a potential treatment for anxiety. Anesth Analg 2001; 92: 548–53.
20. Koch M, Kain Z, Ayoub C, Rosenbaum S. The sedative and analgesic sparing effect of music. Anesthesiology 1998; 89: 300–6.
21. Miluk-Kolasa B, Obminski Z, Stupnicki R, Golec L. Effects of music treatment on salivary cortisol in patients exposed to pre-surgical stress. Exp Clin Endocrinol 1994; 102: 118–20.
22. Miluk-Kolasa BM, Matejek M, Stupnicki R. The effects of music listening on changes in selected physiological parameters in adult pre-surgical patients. J Music Ther 1996; 33: 208–18.
23. Palakanis KC, DeNobile JW, Sweeney WB, Blankenship CL. Effect of music therapy on state anxiety in patients undergoing flexible sigmoidoscopy. Dis Colon Rectum 1994; 37: 478–81.
24. Bampton P, Draper B. Effect of relaxation music on patient tolerance of gastrointestinal endoscopic procedures. J Clin Gastroenterol 1997; 25: 343–5.
25. Colt HG, Powers A, Shanks TG. Effect of music on state anxiety scores in patients undergoing fiberoptic bronchoscopy. Chest 1999; 116: 819–24.
26. Stermer E, Levy N, Beny A, et al. Ambience in the endoscopy room has little effect on patients. J Clin Gastroenterol 1998; 26: 256–8.
© 2002 International Anesthesia Research Society