Professional Documents
Culture Documents
John Hedberg
Carl E. Thoresen
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Stanford University
The authors evaluated an 8-week, 2-hr per week training for physicians, nurses, chaplains, and other
health professionals using nonsectarian, spiritually based self-management tools based on passage
meditation (E. Easwaran, 1978/1991). Participants were randomized to intervention (n 27) or waiting
list (n 31). Pretest, posttest, and 8- and 19-week follow-up data were gathered on 8 measures, including
perceived stress, burnout, mental health, and psychological well-being. Aggregated across examinations,
beneficial treatment effects were observed on stress ( p .0013) and mental health ( p .03). Treatment
effects on stress were mediated by adherence to practices ( p .05). Stress reductions remained large at
19 weeks (84% of the pretest standard deviation, p .006). Evidence suggests this program reduces
stress and may enhance mental health.
Keywords: meditation, spirituality, social cognitive theory, stress, mental health
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
715
professional education credits were offered. After completing all assessments, participants received stipends of $100. This project was approved
by the Institutional Review Boards of the host hospital and of the Public
Health Institute of Oakland, California.
Prospective participants were instructed to attend one of three registration sessions in September 2002, in which they gave consent, completed a
pretest assessment (Exam 1), and were then informed of their randomized
group assignment. One or 2 weeks later, those in the treatment condition
began the 8-week training. Additional assessments occurred after the last
training session (Exam 2), 8 weeks later (Exam 3), and 19 weeks later
(Exam 4). Dropouts (n 3) were more likely to be other or no
denomination than Protestant or Roman Catholic ( p .05). Neither
treatment condition nor dropout was significantly associated with covariables or pretest values of any of the eight outcome variables ( p .05).
Participant flow is summarized in Figure 2. Selected characteristics of the
58 final participants included in the intent-to-treat analyses are displayed in
Table 1. Further details on theoretical and empirical background, participant characteristics, recruitment, and randomization are available on the
Web at http://dx.doi.org/10.1037/0022-006X.74.4.714.supp.
Eight major stress and well-being outcome variables were measured at
each examination: Perceived stress was measured with a well-known
14-item measure from S. Cohen and Williamson (1988); mental health
(five items) and vitality (four items) with subscales from the Medical
Outcomes Study (Ware, Kosinski, & Gandek, 2002); three dimensions of
burnout by psychometrically sound subscales of the Maslach Burnout
Inventory (Maslach, Jackson, & Leiter, 1996); and life satisfaction with a
commonly used five-item scale from Diener, Emmons, Larsen, and Griffin
(1985). Job satisfaction was measured with responses to a single item:
Considering all aspects of my job, I would say that I am very satisfied
with my job. Responses were coded on a 7-point scale from 1 (strongly
disagree) to 7 (strongly agree).
At Exam 1, we assessed covariables presented in Table 1, as well as total
years of patient contact and tendencies toward socially desirable responding, using a 13-item version of the MarloweCrowne scale (Reynolds,
Method
The intervention was hosted at a large hospital in Colorado. Enrollment
was open to health professionals with current patient contact. Participants
(N 61) were recruited through in-service talks, flyers, and word of mouth
at two local hospitals. Promotional talks and materials emphasized potential gains in stress management and professional effectiveness. Continuing
716
Table 1
Selected Participant Characteristics by Treatment Status
No. (%)a
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Characteristic
Gender
Women
Men
Age, years
2639
4049
5059
6070
Occupation
Nurse
Physician
Other
Education, years
Less than 16
1618
19 or more
Marital status
Married now
Separated/divorced
Widowed/never married
Previous meditation
Yes
No
No. (%)a
All
Tx
Cx
50 (86)
8 (14)
23 (87)
4 (13)
27 (85)
4 (15)
7 (12)
20 (34)
22 (38)
9 (16)
2 (7)
10 (37)
10 (37)
5 (19)
5 (16)
10 (32)
12 (39)
4 (13)
37 (64)
7 (12)
14 (24)
20 (74)
2 (7)
5 (19)
17 (55)
5 (16)
9 (29)
13 (22)
32 (55)
13 (22)
8 (30)
13 (48)
6 (22)
5 (16)
19 (61)
7 (23)
36 (62)
14 (24)
8 (14)
15 (56)
9 (33)
3 (11)
21 (68)
5 (16)
5 (16)
24 (41)
34 (59)
10 (37)
17 (63)
14 (45)
17 (55)
Characteristic
Spiritual identity
Spiritual and religious
Spiritual, not religious
Not spiritual
Extent spiritual
Very
Moderate
Slightly/not at all
Extent religious
Very/moderate
Slightly
Not at all
Religious attendance
Weekly or more
Less than weekly
Never
Religious denomination
Protestant
Roman Catholic
Other
None
All (Total)
All
Tx
Cx
26 (45)
29 (50)
3 (5)
11 (41)
15 (56)
1 (4)
15 (48)
14 (45)
2 (6)
14 (24)
31 (53)
13 (22)
5 (19)
16 (59)
6 (22)
9 (29)
15 (48)
7 (23)
25 (43)
25 (43)
8 (14)
10 (37)
13 (48)
4 (15)
15 (48)
12 (39)
4 (13)
19 (33)
37 (64)
2 (3)
8 (30)
19 (70)
0 (0)
11 (35)
18 (58)
2 (6)
22 (38)
17 (29)
7 (12)
12 (21)
9 (33)
9 (33)
4 (15)
5 (19)
13 (42)
8 (26)
3 (10)
7 (23)
58
27
31
Results
No adverse effects from the EPP training were observed. Table
2 presents estimates and confidence intervals for changes on
outcome measures since pretest. Perceived stress showed strong
and significant reductions relative to pretest at all subsequent
examinations. At Exam 2 (posttest), stress in the treatment group
versus the control group was already reduced by more than 60% of
a pretest standard deviation ( p .05, d 0.63; J. Cohen, 1988).
The stress reduction was even larger at Exam 3 (d 1.00, p
.001) and at Exam 4 (d 0.84, p .01). After a highly
conservative Bonferroni adjustment for the presence of 24
outcome-focused hypothesis tests in Table 2, Exam 3 stress reduction remained statistically significant ( p .05), clearly rejecting
the global null hypothesis of no real treatment effects on any
outcome.
Other than stress, all outcomes except depersonalization and life
satisfaction demonstrated one or more changes in the hypothesized
direction ( p .20). These changes were most common at Exam 3
(8-week follow-up), in which all outcomes except mental health
were more favorable than at Exam 2.
Despite consistently stronger effects at Exam 3, tests for heterogeneity of treatment effect between Exams 2, 3, and 4 failed for
all outcomes to reject the null hypothesis of a time constant
717
Table 2
Examination 1 Pretest Values and Observed Treatment Effects at Examinations 2, 3, and 4 (n 58)
Treatment effects
Exam 1
pretest values
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Tx change Cx change
SD
Exam
95% CI
p (2-tailed)
Stress
26.14
6.57
Mental health
64.76
15.72
Vitality
54.74
20.66
Emotional exhaustion
21.40
10.76
6.88
5.34
Personal accomplishment
37.62
6.34
Life satisfaction
24.29
5.99
Job satisfaction
5.28
1.59
Adherence
1.00
0.00
2
3
4
24
2
3
4
24
2
3
4
24
2
3
4
24
2
3
4
24
2
3
4
24
2
3
4
24
2
3
4
24
2
3
4
24
4.14
6.57
5.55
5.23
5.31
5.10
4.06
4.82
0.15
5.42
1.39
2.23
0.45
3.51
0.10
1.00
0.67
0.44
0.52
0.60
1.52
1.90
1.19
1.51
0.58
1.68
0.80
0.93
0.03
0.39
0.06
0.08
2.39
2.15
2.07
2.20
7.82, 0.45
10.49, 2.65
9.51, 1.59
8.39, 2.08
2.54, 13.14
2.77, 12.96
3.80, 11.93
1.56, 11.20
9.57, 9.27
4.02, 14.86
8.05, 10.83
5.46, 9.91
4.16, 3.26
7.69, 0.66
4.43, 4.23
4.33, 2.33
1.06, 2.40
1.60, 2.49
1.67, 2.71
0.99, 1.20
0.62, 3.67
0.48, 4.28
1.25, 3.63
0.38, 3.40
1.88, 3.05
0.97, 4.32
1.87, 3.48
1.18, 3.05
0.61, 0.55
0.26, 1.05
0.62, 0.74
0.44, 0.60
2.19, 2.59
1.91, 2.40
1.81, 2.34
2.01, 2.40
.03a
.0012a
.006a
.0013a
.11b
.16b
.31b
.03b
.49b
.18b
.57b
.25b
.81a
.099a
.96a
.55a
.67b
.99b
.90b
.97b
.30b
.19b
.16b
.18b
.64a
.21a
.56a
.38a
.92b
.199b
.49b
.50b
.0001a
.0001a
.0001a
.0001a
Depersonalization
Note. Exam 2 4 refers to combined analyses of Exams 2, 3, and 4. Tx treatment group; Cx control group; CI confidence interval; Exam 1
pretest; Exam 2 posttest; Exam 3 8-week follow-up; Exam 4 19-week follow-up.
a
t tests for group differences in mean change from hierarchical linear models. b Wilcoxon nonparametric tests of change scores for variables that were
nonnormally distributed ( p .10) on ShapiroWilk tests of Exam 1 distributions.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
718
Discussion
This randomized, controlled study of a comprehensive
meditation-based spiritual intervention found statistically significant effects on perceived stress, suggesting this programs possible
value for a variety of stress management and other health applications. Combined analyses of posttest and two follow-up examinations demonstrated statistically significant beneficial treatment
effects for perceived stress and mental health. These effects could
not be attributed to socially desirable responding. The large perceived stress reductions are clinically promising, as this measure
has been correlated with a wide range of health variables including
depression, health services utilization, and self-reported health and
symptoms (S. Cohen & Williamson, 1988).
This study provides at least five novel contributions to the
literature. First, we provide the first randomized evidence that the
EPP, a meditation-based intervention, was successfully used with
health professionals. Second, we are the first to study effects of
EPP training sustained over two follow-ups spanning almost 6
months. Third, compared with the only other randomized study of
meditation for health professionals (Shapiro, Astin, Bishop, &
Cordova, 2005), we attained a much higher treatment group retention rate (90% vs. 56%). Fourth, this is the first time that the EPP
has been examined with the well-established measures used here.
Finally, these findings provide an important foundation for further
research on how the EPP or similar nonsectarian spiritual practices
might foster health by drawing on spiritual and religious resources.
Caution, however, is recommended in interpreting these findings. The current study lacked an active comparison group that
could more firmly rule out competing explanations for reported
effects. We do not know if similar effects would have been found
if we had used an active control group that provided opportunities
to participate in other types of weekly activities on the topic of
stress management. Such a finding seems highly unlikely given
extensive experience in teaching the EPP, but it remains a possibility to control in future studies.
Evidence that EPP practice mediated treatment effects is consistent with findings from qualitative interviews and earlier quantitative studies (Oman, Hedberg, Downs, & Parsons, 2003; Richards, Oman, Hedberg, Thoresen, & Bowden, 2006). Such
evidence, however, fails to clearly demonstrate causality (reverse
causality is not ruled out) but remains suggestive at best.
Findings that stress and mental health benefits were larger for
participants who were only moderately or less spiritual, compared
with highly spiritual, are consistent with findings of moderation by
spirituality in a previous EPP study (Oman, Hedberg, Downs, &
Parsons, 2003). One explanation may be that participants with
lower initial spirituality have more to learn because they possess
less initial familiarity with spiritual self-management tools.
Generalizability
Results appear most likely to generalize to other populations
that are similarly self-selected, such as health professionals who
enroll in continuing professional education courses that use similar
promotional materials. Further research would be required to assess generalizability to non-self-selected groups, such as an entire
staff of a unit or a hospital. No evidence was found for moderation
by demographics, but smaller numbers of non-nurses and male,
non-middle-aged, and non-White participants suggest considerable
caution in generalizing to these groups.
Other Limitations
Another limitation was the studys relatively small sample size,
which provided statistical power above 90% only for testing larger
changes, that is, changes that approach or exceed 80% of the
pretest standard deviation (d 0.80; S. Cohen & Williamson,
1988). As a result, favorable changes ( p .20, 2-tailed) in several
outcomes are at best suggestive of possible benefits and may merit
further study. Clearly there is no basis for any firm conclusions.
Most importantly, the absence of a competing comparison group
treatment to control for generalized benefits of participating in a
group leaves open what specifically accounted for the observed
changes. The competency of the instructors was not formally
assessed. Possible prior participant adherence to EPP practices was
not assessed, nor whether waiting list group participants engaged
concurrently in other meditation or stress management programs
or increased their spiritual activities. We lacked adequate sample
size and detailed measures of prior experience with meditation to
more fully explore how they might moderate treatment effects. We
did not evaluate changes in spiritual strivings, religious coping, or
various other potential mediators of longer term effects. Finally,
we relied entirely on self-report measures that were more vulnerable to confounding with related constructs.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
References
Bandura, A. (2003). On the psychosocial impact and mechanisms of
spiritual modeling. The International Journal for the Psychology of
Religion, 13, 167174.
Baron, R. M., & Kenny, D. A. (1986). The moderatormediator variable
distinction in social psychological research: Conceptual, strategic, and
statistical considerations. Journal of Personality and Social Psychology,
51, 11731182.
Bishop, S. R. (2002). What do we really know about mindfulness-based
stress reduction? Psychosomatic Medicine, 64, 71 83.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences
(2nd ed.). Hillsdale, NJ: Erlbaum.
Cohen, S., & Williamson, G. M. (1988). Perceived stress in a probability
sample of the United States. In S. Spacapan, & S. Oskamp (Eds.), The
social psychology of health (pp. 31 67). Newbury Park, CA: Sage.
Curlin, F. A., Lantos, J. D., Roach, C. J., Sellergren, S. A., & Chin, M. H.
(2005). Religious characteristics of U.S. physicians: A national survey.
Journal of General Internal Medicine, 20, 629 634.
Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The
Satisfaction with Life Scale. Journal of Personality Assessment, 49,
7175.
Easwaran, E. (1991). Meditation: A simple eight-point program for translating spiritual ideals into daily life (2nd ed.). Tomales, CA: Nilgiri
Press. (Original work published 1978)
Goleman, D. (1988). The meditative mind: The varieties of meditative
experience. New York: Tarcher.
Kabat-Zinn, J. (2003). Mindfulness-based interventions in context: Past,
present, and future. Clinical Psychology: Science & Practice, 10, 144
156.
Maslach, C., Jackson, S. E., & Leiter, M. P. (1996). Maslach Burnout
Inventory manual (3rd ed.). Palo Alto, CA: Consulting Psychologists
Press.
McEwen, B. S. (1998). Protective and damaging effects of stress mediators. New England Journal of Medicine, 338, 171179.
Miller, W. R., & Thoresen, C. E. (2003). Spirituality, religion, and health:
An emerging research field. American Psychologist, 58, 24 35.
719