Effectiveness of group and self-help cognitive behavior therapy in reducing problematic menopausal hot flushes and night sweats (MENOS 2)

ArticleinMenopause (New York, N.Y.) 19(7):749-59 · February 2012with56 Reads
Impact Factor: 3.36 · DOI: 10.1097/gme.0b013e31823fe835 · Source: PubMed
Abstract

The aim of this study was to examine the effectiveness of group cognitive behavioral therapy (CBT) and guided self-help CBT in reducing hot flush and night sweat (HF/NS) problem rating at 6 and 26 weeks after randomization. This was a randomized control trial of 140 women having 10 or more problematic HF/NS a week for at least a month. The primary outcome was HF/NS problem rating (1-10) at 6 weeks after randomization. Secondary outcomes were physiologically measured HF/NS at 6 weeks; HF/NS problem rating at 6 weeks; and frequency, mood (Women's Health Questionnaire), and health-related quality of life (General Health Survey Short Form-36) at 6 and 26 weeks. Intention-to-treat analysis was used, and between-group differences were estimated using linear mixed models. Baseline mean (SD) HF/NS weekly frequency was 63.15 (49.24), and problem rating was 5.87 (2.28). Group and self-help CBT both significantly reduced HF/NS problem rating at 6 weeks-group CBT versus no treatment control (NTC; adjusted mean difference, 2.12; 95% CI, 1.36-2.88; P < 0.001) and self-help CBT versus NTC (adjusted mean difference, 2.08; 95% CI, 1.29-2.86; P < 0.001)-and at 26 weeks-group CBT versus NTC (adjusted mean difference, 1.33; 95% CI, 0.54-2.13; P = 0.001) and self-help CBT versus NTC (adjusted mean difference, 1.19; 95% CI, 0.36-2.02; P = 0.005). Group and self-help CBT significantly reduced night sweat frequency at 6 and 26 weeks. There were improvements in mood and quality of life at 6 weeks and improved emotional and physical functioning for group CBT at 26 weeks. These results suggest that CBT delivered in group or self-help format is an effective treatment option for women during the menopause transition and postmenopause with problematic HF/NS.

Full-text

Available from: Myra S Hunter, Feb 19, 2014
Menopause: The Journal of The North American Menopause Society
Vol. 19, No. 7, pp. 749/759
DOI: 10.1097/gme.0b013e31823fe835
* 2012 by The North American Menopause Society
Effectiveness of group and self-help cognitive behavior therapy in
reducing problematic menopausal hot flushes and night sweats
(MENOS 2): a randomized controlled trial
Beverley Ayers, DHealthPsych,
1
Melanie Smith, DClinPsych,
1
Jennifer Hellier, MSc,
2
Eleanor Mann, DPhil,
1
and Myra S. Hunter, PhD
1
Abstract
Objective: The aim of this study was to examine the effectiveness of group cognitive behavioral therapy (CBT)
and guided self-help CBT in reducing hot flush and night sweat (HF/NS) problem rating at 6 and 26 weeks after
randomization.
Methods: This was a randomized control trial of 140 women having 10 or more problematic HF/NS a week for
at least a month. The primary outcome was HF/NS problem rating (1-10) at 6 weeks after randomization. Secondary
outcomes were physiologically measured HF/ NS at 6 weeks; HF/NS probl em rating at 6 weeks; and frequency,
mood (Women’s Health Questionnaire), and health-related quality of life (General Health Survey Short FormY36)
at 6 and 26 weeks. Intention-to-treat analysis was used, and between-group differences were estimated using linear
mixed models.
Results: Baseline mean (SD) HF/NS weekly frequency was 63.15 (49.24), and problem rating was 5.87 (2.28).
Group and self-help CBT both significantly reduced HF/NS problem rating at 6 weeksVgroup CBT versus no
treatment control (NTC; adjusted mean difference, 2.12; 95% CI, 1.36-2.88; P G 0.001) and self-help CBT versus
NTC (adjusted mean difference, 2.08; 95% CI, 1.29-2.86; P G 0.001)Vand at 26 week sVgroup CBT versus NTC
(adjusted mean difference, 1.33; 95% CI, 0.54-2.13; P = 0.001) and self-help CBT versus NTC (adjusted mean
difference, 1.19; 95% CI, 0.36-2.0 2; P = 0.005). Group and self-help CBT significantly reduced night sweat fre-
quency at 6 and 26 weeks. There were improveme nts in mood and quality of life at 6 weeks and improved emot ional
and physical functioning for group CBT at 26 weeks.
Conclusions: These results suggest that CBT delivered in group or self-help format is an effective treatment
option for women during the menopause transition and postmenopause with probl ematic HF/NS.
Key Words: Hot flushes Y Night sweats Y Menopause Y Cognitive behavior therapy Y Group Y Self-help Y
Vasomotor symptoms.
H
ot flushes and night sweats (HF/NS) are problematic
for 20% to 25% of women during the menopausal
transition because they cause discomfort, sleep dis-
ruption, and social embarrassment
1
and are the main reason
for medical help seeking. HF/NS are associated with changes
in estrogen levels and neurotransmitters, which have an im-
pact on the thermoregulatory system,
2<4
leading to a narrowed
thermoneutral zone.
5
They are potentiated by general levels
of stress
6
and can be triggered by small elevations in core
body temperature, hot drinks, alcohol, physical activity, or
emotional reactions.
7<9
Hormone therapy (HT) is an effective
treatment of HF/NS, but the results of prospective trials have
suggested increased risks of breast cancer and stroke.
10,11
As
a resul t, HT use has declined since 2003,
12,13
and those who
have discontinued treatment often report recurrence of symp-
toms.
14
Selective serotonin reuptake inhibitors (SSRIs) are
nonhormonal medical treatments, but evidence of efficacy has
been mixed, particularly in trials of healthy women, and their
longer term efficacy is uncertain.
15<18
There is a wide range of
other medical and complementary therapy options available.
Received October 5, 2011; revised and accepted November 1, 2011.
From the
1
Department of Psychology (at Guy’s), Institute of Psychiatry,
King’s College London, London, UK; and
2
Mental Health and Neuro-
science Clinical Trials Unit, Institute of Psychiatry, King’s College
London, London, UK.
Funding/support: This work was funded by the National Institute for
Health Research Biomedical Research Centre for Mental Health,
South London, and Maudsley National Health Service Foundation
Trust and the Institute of Psychiatry, King’s C ollege London.
Financial disclosure/conflicts of interest: None reported.
Authors’ contributions: B.A. was the trial coordinator on the study and
generated the first draft of this manuscript based on the study protocol.
M.S. was the clinical psychologist who conducted the group CBT and
guided self-help. J.H. was the trial statistician. E.M. analyzed the phys-
iological data, and M.S.H. was the principal investigator of this study and
wrote the original protocol and supervised the intervention. All authors
contributed to the final version of the manuscript.
Address correspondence to: Myra S. Hunter, PhD, Department of Psy-
chology, Institute of Psychiatry, King’s College London, 5th Floor Ber-
mondsey Wing, Guy’s Campus, London Bridge, London SE1 9RT,
England. E-mail: myra.hunter@kcl.ac.uk
Menopause, Vol. 19, No. 7, 2012 749
Copyright © 2012 The North American Menopause Society. Unauthorized reproduction of this article is prohibited.
Page 1
However, many lack adequate research evidence to support
their efficacy or are associated with significant adverse ef-
fects,
19<25
and reviews recommend further controlled evalua-
tion, particularly for complementary therapies and herbal
remedies.
19,22,26,27
There is a need for safe, acceptable, and
effective nonhormonal treatments for women during the meno-
pause transition and postmenopause with problematic symp-
toms, many of whom prefer nonmedical options.
28,29
Recent research suggests that psychological factors, includ-
ing emotional reactions such as anxiety or embarrassment, and
negative beliefs and catastrophic thoughts might moderate the
experience of HF/NS
29<35
and that relaxation and/or paced
breathing might reduce HF/NS.
36<38
Hunter and Liao
29
devel-
oped a cognitive behavioral treatment, including psychoedu-
cation, paced breathing/relaxation, and cognitive behavioral
therapy (CBT) to help women manage HF/NS, which was ac-
ceptable to women and showed promise in exploratory trials
of individual and group CBT, reducing HF/NS by 40% to
50%.
29,30,39
However, more evidence is needed to evaluate
CBT interventions using randomized controlled trials (RCTs).
31
Group and self-help methods of delivering CBT can be cost-
effective alternatives to one-to-o ne treatments.
39,40
For example,
in a recent systematic review of CBT interventions for depres-
sion and anxiety, guided self-help interventions (eg, using
telephone contacts) were found to be effective and compa-
rable with face-to-face treatments.
41
HF/NS can be measured by both subjective and physio-
logical measures.
33
Self-report measures include diary re-
cords of frequency and severity of HF/NS and measures of
the impact of HF/NS on daily life (problem-rating or bother-
someness).
1
Sternal skin conductance (SSC) monitoring is
considered the most valid physiological measure of HF/NS,
42
although there is discordance between subjective and phys-
iologically measured frequency, and optimal methods of mea-
surement are currently under investigation.
43
In general, HF/NS
frequency does not correlate highly with severity or bother
caused by HF/NS.
44
Consequently, HF/NS problem rating or
bother is now considered to be the most appropriate clini-
cal outcome measure because it is associated with help seek-
ing and reduced quality of life.
33,44
In the current study, HF/
NS problem rating is the primary outcome measure, and we
include a reliable subjective measure of HF/NS frequency
(Hot Flush Rating Scale
1
[HFRS]) and a physiological mea-
sure (SSC monitoring) to evaluate treatment effect s.
The aim of the current study was to evaluate the effective-
ness of group CBT and guided self-help CBT in alleviating
HF/NS in women wi th problematic menopausal symptoms in
an RCT. Group and self-help CBT are compared with a no
treatment control (NTC) group at 6 and 26 weeks after random-
ization. It is the first RCT to evaluate CBT for menopause
symptoms using both physiological (SSC) and self-report
measures of HF/NS.
METHODS
The trial is reported in accordance with the CONSORT
guidelines for randomized control trials
45
and is described in
detail in a trial protocol article
46
(trial registration: current con-
trolled trials: ISRCTN57302613). Ethics approval was obtained
from King’s College London Research Ethics Committee (Psy-
chiatry, Nursing, and Midwifery Research Ethics Subcommittee;
reference: PNM/08/09-42). Data were collected at King’s Col-
lege London, and all participants gave written informed consent
before taking part.
Study population and eligibility criteria
Women with variable mens trual cycles (menopausal tran-
sition) or those who were more than 1 year from their last
menstrual period (postmenopause)
47
with problematic HF/NS
were recruited from South London general practices, breast
screening clinics, a menopause Web site (Menopause Mat-
ters), and local newspaper advertisements between March
2009 and May 2010. Broad inclusion criteria were applied
because we were interested in the impact of self -help CBT
upon less problematic HF/NS. Inclusion criteria were English
speaking; 18 years or older; having problematic HF/NS (score
above 2 on the HFRS
1
) for at least 1 month (minimum weekly
frequency of 10); living within traveling dista nce of London;
and willingness to maintain, or report any change in, any
menopause treatments, remedies, or supplements at a similar
level during the trial. Exclusion criteria were nonYEnglish
speaking, history of breast cancer, and having medical or psy-
chiatric conditions that would affect the ability to participate.
Procedure
Potential participants opted into the study in response to
posters and leaflets and contacted the trial coordinator by tele-
phone or e-mail. They were screened by telephone and, if eli-
gible, sent trial information. Those who were interested in
participating attended an assessment interview covering medi-
cal history, menop ausal symptoms, and treatment use. They
also completed a consent form and sociodemographic and
baseline questionnaires and were fitted with an SSC hot flush
monitor (Bahr Monitor; Simplex Scientific LLC), which they
wore for 24 hours and returned the next day. Participants who
consented were randomly allocated to treatment groups (group
CBT, self-help CBT, and NTC) through an Internet random-
ization site (www.random.org) in a 1:1:1 ratio with block sizes
of 18 to 30 participants.
Posttreatment assessments occurred at 6 weeks after ran-
domization. Participants in all three arms of the study com-
pleted questionnaires and 24-hour SSC hot flus h monitoring.
Follow-up assessments occurred at 26 weeks after random-
ization; the participants completed questionnaires and an
interview about their experience of the trial and the interven-
tions (to be reported elsewher e).
Intervention
Group CBT included 2-hour sessions, once a week, for
4 weeks (8 h in total). The sessions focused on psychoeduca-
tion, stress management, paced breathing, and CBT, including
individual goal setting and homework.
29,39,46,48
The CBT for
HF/NS is based on a theoretical model of HF/NS, which out-
lines specific components and hypothesizes how these might
750 Menopause, Vol. 19, No. 7, 2012 * 2012 The North American Menopause Society
AYERS ET AL
Copyright © 2012 The North American Menopause Society. Unauthorized reproduction of this article is prohibited.
Page 2
modify women’s experience of HF/NS.
32
The treatment is
manualized using PowerPoint presentations, a relaxation/paced
breathing CD, and handouts. Sessions were audio recorded,
and 10% were rated by a clinical psychologist for adherence
to the manual.
Self-help CBT includes a self-help book completed during
a 4-week period and two contacts with a clinical psychologist
(one introductory session and a Bguiding[ telephone call 2 wk
into treatment). The content of self-help CBT was identical
to group CBT, with participants receiving the CD for daily
practice and weekly homework tasks (see Ayers et al
46
for
more information).
Session 1 introduced participants to the CBT model of HF/
NS, including the physiological, cognitive, behavioral,
and emotional components of HF/NS. Strategies to reduce
stress were discussed, and relaxation/paced diaphragmatic
breathing were demonstrated.
Session 2 introduced cognitive and behavioral strategies for
dealing with HF; participants completed exercises to de-
velop helpful cognitive and behavioral strategies during
HF and were taught to use paced breathing and asked to
practice regularly, in potentially stressful situations as well
as to use at the onset of a hot flush or night sweat.
Session 3 focused on manag ing NS and sleep. Mechanisms
of sleep and cognitive and behavioral strategies for deal-
ing with NS and nighttime waking were discussed.
Session 4 reviewed the overall CBT model, and participants
set individual goals and developed a maintenance (relapse
prevention) plan.
Participants in the NTC condition did not receive CBT
during the treatment phase but were able to access their general
practitioner and other healthcare options. They were offered a
form of CBT off-trial at the end of the study. Any treatments
and services accessed during the treatment phase were logged
at 6 and 26 weeks after randomization.
Measures
Demographic information (age, height, weight, ethnicity,
education, marital and employment status), smoking, alcohol
intake and exercise behaviors, menopause status (as defined
by the Stages of Reproductive Aging Workshop
47
), and treat-
ment use were recorded at baseline. Data collection was con-
ducted by the trial coordinator; data analysis was conducted
by the trial statistician who was blind to treatment condition.
Primary outcome
The primary outcome was hot flush problem rating at
6 weeks after randomization, as measured by a subscale of the
HFRS, a scale derived using factor analysis with good inter-
nal consistency (Cronbach > = 0.9) and test-retest reliabilit y
(r = 0.8).
1
Problem rating is determined by the mean of three
items: BTo what extent do you regard your flushes/sweats as
a problem?[BHow distressed do you feel about you r hot
flushes?[BHow much do your hot flushes interfere with your
daily routine?[ Each item is measured on a 10-point scale,
from low to high, with high scores indicating more problem-
atic hot flushes; a 2-point change on this scale is considered
clinically significant.
Secondary outcomes
Secondary outcomes were hot flush problem rating (at
26 weeks after randomization), HF/NS frequency measured
subjectively and physiologically (SSC), mood, and health-
related quality of life (HrQOL).
HF/NS frequency was measured using the HFRS Frequency
subscale, which records the number of HF/NS experi-
enced in the previous week and has good test-retest reli-
ability (r = 0.8) and validity, with significant correlations
with daily diary recordings for HF (r = 0.97, P G 0.001)
and NS (r = 0.94, P G 0.001).
1
Ambulatory SSC mon-
itoring was the 24-hour physiological measure of fre-
quency using the Bahr monitor (Simplex Scientific LLC);
HF/NS were defined by an incre ase in conductance of
2 mmhos or more in 30 seconds, occurring no less than
15 minutes since the last SSC- defined HF/NSVa method
of analysis that has been used with various ambulatory
and laboratory SSC equipment.
49
Participants also re-
corded whenever they had HF/NS by passing a magnet
(event marker) over the monitor.
Mood and HrQOL: the Women’s Health Questionnaire
(WHQ)
50
is widely used to measure perceptions of phys-
ical and emotional symptoms in the evaluation of in-
terventions for women during the menopause transition
and postmenopause. The following subscales were used in
this study: depressed mood, anxiety/fears, somatic symp-
toms, sleep problems, and memory/concentration. The
range for WHQ subscales is 0.0 to 1.0, and a clinically
meaningful change on the WHQ subscales would typi-
cally be 0.10 to 0.20.
51
The WHQ has been standardized
in mid-aged women and has high internal reliability
(Cronbach > = 0.7 to 0.84) and test-retest reliability (0.78
to 0.96). The General Health Survey Short FormY36 (SF-
36)
52
was used to measure dimensions of HrQOL (phys-
ical functioning, physical role limitation, pain, general
health, vitality, social functioning, emotional role limi-
tation, and mental health). The range for SF-36 subscales
is 0 to 100.
Sample size and power
Group and self-help CBT were compared with the NTC
group based on identical assumptions. A total sample size of
120 (40 in each group) has 90% power to detect a clinically
relevant difference in mean HF/NS problem rating of 2 points,
allowing for baseline value (estimated to have a mean of 5 and
SD of 2.4) and a correlation between baseline and posttreat-
ment values of 0.4 (giving a correction factor of 0.84) with
two-sided significance level of 0.05.
Statistical analysis
Analyses were pragmatic, based on an inte ntion-to-treat
sample. Significance was considered at the 5% level for all
Menopause, Vol. 19, No. 7, 2012 751
RCT OF CBT FOR MENOPAUSAL SYMPTOMS
Copyright © 2012 The North American Menopause Society. Unauthorized reproduction of this article is prohibited.
Page 3
analyses. For the primary analyses (HF/NS problem rating),
between-group differences (group CBT vs NTC and self-help
CBT vs NTC) were estimated using mixed linear regression
models. Covariates were treatment group, baseline HF/NS
problem rating score, and time. A time-by-treatment inter-
action term was included to allow estimates at the individ-
ual time points to be summarized. The model for the HF/NS
problem rating incorporated two levels of random effectsV
participant and cohort group. Cohort group refers to the ran-
domization group (ie, to the participants who were randomized
at the same time into the three arms of the trial). Model as-
sumptions were checked using diagnostic plots.
Missing data were assumed to be missing at random,
conditional in the variables included in the substantive models.
Secondary outcomes were analyzed using mixed linear re-
gression models, with random participant and cohort group
intercepts and a time-by-treatment interaction terms; covariates
in the model were treatment group, baseline value of outcome,
and time. Effect sizes were calculated based on Cohen d,where
d is the difference between two means divided by the pooled
SD. The results from all analyses were summarized at 6 and
26 weeks with two-sided 95% CIs. All analyses were com-
pleted in Stata 11.0.
RESULTS
A total of 295 women expressed an interest in the tri al: 73
declined, 43 did not meet inclusion criteria, and 39 could not
participate for practical reasons (eg, travel difficulties), leav-
ing 140 women who underwent baseline assessment and ran-
domization. Data for the primary outcome were collected for
140 participants (100%). Key secondary outcomes were avail-
able for 129 participants (92%) at 6 weeks and 111 participants
FIG. 1. CONSORT diagram showing participant flow through the trial. HF/NS, hot flushes and night sweats; CBT, cognitive behavior therapy; NTC, no
treatment control; HFPR, hot flush problem rating.
752 Menopause, Vol. 19, No. 7, 2012 * 2012 The North American Menopause Society
AYERS ET AL
Copyright © 2012 The North American Menopause Society. Unauthorized reproduction of this article is prohibited.
Page 4
(79%) at 26 weeks. Participant flow through the trial is shown
in Figure 1.
Table 1 presents demographic and clinical information at
baseline for participants allocated to each arm of the trial. De-
mographic variables were balanced over the treatment arms.
Participants were women with an average age of 53 years,
mostly married or cohabiting and of white ethnicity; 67% were
educated beyond 16 years of age, 74% were employed, and
they were either in the late menopausal transition or were post-
menopausal. They reported more than 60 HF/NS per week
and had had HF/NS for an average of 3.94 years (SD, 4.27 y);
HF/NS lasted approximately 7.82 min utes (SD, 17.61 min),
ranging from 1 to 60 minutes. Twenty-eight percent had used
HT in the past, and 3% were current users but were still
symptomatic and continued using HT during the trial.
Before randomization, the participants were asked about
their self-efficacy in relation to carrying out the treatments and
their treatment preferences. There were no differences between
the three groups for self-efficacy, treatment preference, or be-
liefs that the treatment would help menopausal symptoms.
A treatment manual was developed before data collection
and used to train the clinical psychologist who delivered all
sessions. Traini ng and clinical supervision was provided by a
consultant clinical psychologist throughout the study period.
Audio tapes of 10% of group CBT sessions were rated for ad-
herence to protocol by an independent psychologist. Compliance
with protocol was very good and ranged from 95% to 100%.
Outcomes
Primary outcome
There were clinically significant group differences for HF/
NS problem rating at 6 weeks (mean difference of 2 points
or greater on the 10-point scale). Group CBT (adjusted mean
difference group CBT vs NTC, 2.12; 95% CI, 1.36-2.88; P G
0.001) and self-help CBT participants (adjusted mean differ-
ence self-help CBT vs NTC, 2.08; 95% CI, 1.29-2.86; P G
0.001) had significantly less problematic HF/NS than did NTC
participants. There were no significant differences between
group and self-help CBT at 6 weeks after treatment (Table 2,
Fig. 2). At 6 weeks, the unadjusted proportion of participants
reporting a clinically significant improvement greater than a 2-
point reduction from baseline on the HF/NS problem rating
scale was 65% (95% CI, 51-77), 73% (95% CI, 57-84) , and
21% (95% CI, 11-35) for group and self-help CBT and NTC,
respectively (Fig. 3). Unadjusted effect sizes (Cohen d) were
TABLE 1. Baseline sociodemographics and health of participants in the three arms of the trial
Participant demographics Group CBT (n = 48) Self-help CBT (n = 47) No treatment control (n = 45) Total (N = 140)
Age, mean (SD), y 53.73 (5.9) 51.70 (4.4) 53.87 (5.7) 53.09 (5.4)
Menopause status
Menopausal transition 17 (35%) 24 (51%) 15 (33%) 56 (40%)
Postmenopausal 31 (65%) 23 (49%) 30 (67%) 84 (60%)
HT use
Using HT 2 (4%) 1 (2%) 1 (2%) 4 (3%)
Used HT in past 15 (31%) 10 (21%) 14 (31%) 39 (28%)
Surgical menopause (oophorectomy or hysterectomy) 7 (15%) 11 (23%) 10 (22%) 28 (20%)
Ethnicity
White 39 (82%) 41 (87%) 35 (78%) 115 (82%)
Asian 2 (4%) 1 (2%) 1 (2%) 4 (3%)
Black 5 (10%) 4 (9%) 6 (13%) 15 (11%)
Other 2 (4%) 1 (2%) 3 (7%) 6 (4%)
Marital status
Single 9 (18%) 13 (28%) 12 (27%) 34 (24%)
Married/partnered 39 (82%) 34 (72%) 33 (73%) 106 (76%)
Education
No qualifications 5 (10%) 1 (2%) 3 (7%) 9 (6%)
Up to 16 y 12 (25%) 12 (26%) 14 (31%) 38 (27%)
Up to 18 y 6 (12%) 8 (17%) 6 (13%) 20 (14%)
Degree 25 (53%) 26 (55%) 22 (49%) 73 (53%)
Employment
Employed 36 (76%) 38 (81%) 29 (65%) 103 (74%)
Not employed/retired/student 12 (24%) 9 (19%) 16 (35%) 37 (26%)
BMI (overweight/obese) 19 (43%) 22 (49%) 23 (57%) 64 (46%)
Smoking behavior
Nonsmoker 20 (42%) 26 (55%) 24 (53%) 70 (50%)
Current smoker 3 (6%) 2 (4%) 7 (16%) 12 (9%)
Past smoker 25 (52%) 19 (41%) 14 (31%) 58 (41%)
Alcohol intake
None 10 (21%) 10 (21%) 9 (20%) 29 (21%)
1-10 units 31 (65%) 26 (55%) 34 (76%) 91 (65%)
11+ units 7 (14%) 11 (24%) 2 (4%) 20 (14%)
Exercise
G2 times a week 21 (44%) 18 (38%) 25 (56%) 64 (46%)
Q2 times a week 27 (56%) 29 (62%) 20 (44%) 76 (54%)
HT, hormone therapy; CBT, cognitive behavior therapy; BMI, body mass index.
Menopause, Vol. 19, No. 7, 2012 753
RCT OF CBT FOR MENOPAUSAL SYMPTOMS
Copyright © 2012 The North American Menopause Society. Unauthorized reproduction of this article is prohibited.
Page 5
large for group CBT (1.18; 95% CI, 1.36-2.8 8) and for self-
help CBT (1.41; 95% CI, 1.29-2.86) compared with NTC.
Secondary outcomes: HF/NS problem rating (26 wk)
Group CBT (adjusted mean difference group CBT vs NTC,
1.33; 95% CI, 0.54-2.13; P = 0.001) and self-help CBT
(adjusted mean difference self-help CBT vs NTC, 1.19; 95%
CI, 0.36-2.02; P = 0.005) participants had significantly less
problematic HF/NS than did NTC participants at 26 weeks
after randomization. At 26 weeks, the unadjusted proportion
of participants reporting a clinically significant improvement
greater than a 2-point reduction from baseline on the HF/NS
problem rating scale was 63% (95% CI, 47-77), 72% (95% CI,
55-84), and 33% (95% CI, 20-48) for group and self-help CBT
and NTC, respectively (Fig. 3). There were no significant dif-
ferences between group and self-help CBT conditions (Table 2,
Fig. 2).
Secondary outcomes: HF/NS frequency (6 and 26 wk)
There was considerable variation in HF/NS frequency. There
were no significant differences between group CBT and NTC
or self-help CBT and NTC at either 6 or 26 weeks for total HF/
NS frequency or for HF frequency considered separately.
Total HF/NS frequency reduced by 40% for group CBT, 36%
for self-help CBT, and 23% for NTC at 26 weeks; for HF
separately, there were reductions of 34% for both group and
self-help CBT and 28% for NTC at 26 weeks (Fig. 4).
Group CBT participants reported less frequent night sweats
at 6 weeks (adjusted mean difference group CBT vs NTC,
5.13; 95% CI, 0.83-9.43; P = 0.019) and at 26 weeks (adjusted
mean difference, 6.69; 95% CI, 2.16-11.22, P = 0.004) com-
pared with NTC participants (Fig. 5). Self-help CBT partic-
ipants also had fewer night sweats at 26 weeks compared with
NTC participants (self-help CBT vs NTC, 6.01; 95% CI, 1.24-
10.78; P = 0.014), but the difference was not significant at
6 weeks. Percentage reductions from baseline in NS frequency
at 26 weeks were 53% for group CBT, 43% for self-help CBT,
and 12% for NTC.
Secondary outcomes: HF/NS physiological monitor and
event markers (6 wk)
The 24-hour SSC monitor was worn by participants at base-
line and 6 weeks. There were general reductions in freque ncy
(SSC and event marker) across all groups but no significant
group differences. However, the data were subject to large
variability (see CI in Table 3).
FIG. 2. Hot flash/night sweat problem rating scale. Unadjusted means and
95% CI at baseline and 6 and 26 weeks after randomization for group CBT,
self-help CBT, and no treatment control. CBT, cognitive behavior therapy.
FIG. 3. Percentages of participants with a clinically significant improve-
ment from baseline of 2 points on the HF/NS Problem Rating Scale at 6
and 26 weeks after randomization for group CBT, self-help CBT, and no
treatment control. HF/NS, hot flushes and night sweats; CBT, cognitive
behavior therapy.
FIG. 4. Hot flush frequency. Unadjusted means and 95% CI at baseline
and 6 and 26 weeks after randomization for group CBT, self-help CBT,
and no treatment control. CBT, cognitive behavior therapy.
TABLE 2. Primary outcomes: Hot Flush Problem Rating
Group CBT Self-help CBT
6wk
Mean difference from control j2.124 j2.078
95% CI j2.883 to j1.364 j2.864 to j1.292
P G0.001 G0.001
26 wk
Mean difference from control j1.331 j1.192
95% CI j2.127 to j0.536 j2.024 to j0.362
P 0.001 0.005
Comparisons are made of the differences at 6 and 26 wk with the control
group, with the results from the final adjusted linear mixed model allowing for
cohort as a random effect.
CBT, cognitive behavior therapy.
754 Menopause, Vol. 19, No. 7, 2012 * 2012 The North American Menopause Society
AYERS ET AL
Copyright © 2012 The North American Menopause Society. Unauthorized reproduction of this article is prohibited.
Page 6
Secondary outcomes: WHQ (6 and 26 wk)
Group CBT participants reported less depressed mood (ad-
justed mean difference, 0.13; 95% CI, 0.06-0.20; P G 0.001)
and reduced anxiety/fears (adjusted mean difference, 0.17; 95%
CI, 0.07-0.28; P = 0.001) than NTC participants at 6 weeks
(Figs. 6 and 7). Self-help CBT participants reported reduc-
tions in depressed mood (adjusted mean difference self-help
CBT vs NTC, 0.08; 95% CI, 0.01-0.15; P = 0.045), memory/
concentration (0.15; 95% CI, 0.01-0.28; P = 0.036), and sleep
(0.17; 95% CI, 0.03-0.31; P = 0.018) compared with NTC
participants at 6 weeks (Table 4, Fig. 6).
Group CBT participants reported fewer memory/concen-
tration problems (0.15; 95% CI, 0.02-0.28; P = 0.020) and
fewer somatic symptoms (0.16; 95% CI, 0.08-0.25; P G 0.001)
than NTC participants at 26 weeks. Self-help participants re-
ported fewer memory/concentration problems (0.16; 95% CI,
0.03-0.30; P = 0.020) than NTC participants at 26 weeks.
There were no significant differences between group or self-
help CBT and NTC or between treatment groups for any other
subscales at 6 or 26 weeks.
Secondary outcomes: SF-36 (6 and 26 wk)
Group CBT participants reported better mental health
(adjusted mean difference group CBT vs NTC, 5.55; 95%
CI, 0.30-0.79; P = 0.038) and vitality levels (8.02; 95% CI,
0.93-15.11; P = 0.027) than did NTC participants at 6 weeks.
Self-help participants reported better general health (7.78;
95% CI, 1.71-13.84; P = 0.012) than did NTC participants
at 6 weeks. There were no significant differences for group
or self-help CBT for any other subscales at 6 weeks after
treatment.
Group CBT participants reported better physical function-
ing (adjusted mean difference, 6.56; 95% CI, 0.89-12.22; P =
0.023) and better emotional role functioning (14.48; 95% CI,
1.06-27.90; P = 0.035) than did NTC participants from base-
line at 26 weeks (Fig. 8).
TABLE 3. Physiological sternal skin conductance monitor and event marker data at baseline and 6 wk
Hot flush monitor Group CBT (n = 48) Self-help CBT (n = 47) NTC (n = 45)
Summary
Total HF/NS frequency
Baseline, mean (SD) 8.04 (6.51) 7.19 (6.78) 7.73 (7.64)
6 wk, mean (SD) 5.73 (7.28) 3.77 (6.27) 3.38 (4.40)
LMM Coefficient 2.266 0.524 -
Total HF/NS frequency SE 1.216 1.223
6wk
a
95% CI j0.12 to 4.96 j1.87 to 2.92
P 0.062 0.668
Event marker Group CBT (n = 48) Self-help CBT (n = 47) NTC (n = 45)
Summary
Total HF/NS frequency
Baseline, mean (SD) 7.42 (5.50) 8.26 (5.46) 7.73 (5.47)
6 wk, mean (SD) 3.90 (5.20) 4.15 (4.88) 4.20 (5.10)
LMM Coefficient j0.226 j0.207 -
Total HF/NS frequency SE 0.990 0.984
6wk
a
95% CI j2.166 to 1.714 j2.135 to 1.722
P 0.820 0.834
LMM, linear mixed model; HF/NS, hot flushes and night sweats; CBT, cognitive behavior therapy; NTC, no treatment control.
a
Cohort number (ie, randomization group) and participant identification as random effects compared with NTC.
FIG. 5. Night sweat frequency. Unadjusted means and 95% CI at base-
line and 6 and 26 weeks after randomization for group CBT, self-help
CBT, and no treatment control. CBT, cognitive behavior therapy.
FIG. 6. Women’s Health Questionnaire Depressed Mood. Unadjusted
means and 95% CI at baseline and 6 and 26 weeks after randomization
for group CBT, self-help CBT, and no treatment control. CBT, cognitive
behavior therapy.
Menopause, Vol. 19, No. 7, 2012 755
RCT OF CBT FOR MENOPAUSAL SYMPTOMS
Copyright © 2012 The North American Menopause Society. Unauthorized reproduction of this article is prohibited.
Page 7
There were no other significant differences for group or
self-help CBT for any other subscales at the 26-week follow up.
For all three groups, use of healthcare services and treat-
ments were logged at 6 weeks and at 26 weeks after random-
ization. 63% of group CBT, 49% of self-help CBT, and 60%
of control participants visited healthcare service providers
during the study, and for those who visited healthcare service
providers, the mean number of visits was 4.55, 3.50, and 4.77,
respectively. These visi ts were unrelated to menopause. There
were no significant differences between groups for health ser-
vices and treatments used. Most control participants took up
the offer of CBT off-trial, suggesting that they felt that they still
needed treatment.
DISCUSSION
This study demonstrates that when compared with an NTC
group, group and self-help CBT significantly reduced HF/NS
problem rating and NS frequency in symptomatic women dur-
ing the menopause transition and postmenopause. Approxi-
mately 65% and 73% of group and self-help CBT participants,
respectively, reported a clinically significant improvement of
2 points or more on the HF/NS problem rating scale at 6 weeks,
and changes were maintained at 26 weeks. Group CBT partic-
ipants benefited from a 53% reduction in NS frequency, as did
43% of self-help participants, compared with 12% of the NTC
group at 26 weeks. Hot flush frequency reduced by 36% to
40% in treatment arms, but this was not significantly different
from the 23% reduction in the NTC arm.
The results from the 24-hour physiological HF/NS fre-
quency measure were somewhat inconclusive because the data
were subject to large variability. The 24-hour time frame may
not have been long enough, and the data analysis methods,
which follow rule-based methods, may need to be refined.
53,54
Pattern-based analyses
55
are being developed for the Bahr
monitor and will be applied to the data when available. How-
ever, the lack of significant change in SSC frequency parallels
the nonsignificant difference in subjective frequency measured
using the HFRS, which also had high variability. Overall, these
results support the hypothesis that the CBT intervention might
work by influencing symptom perception and appraisal and
possibly mood, rather than having an impact on physiological
mechanisms of HF/NS
32
; further analyses of moderators and
mediators of HF/NS are planned to clarify this issue, which will
include beliefs, behavior, and mood.
There were some additional benefits to mood and HrQOL
at posttreat ment and follow-up. Group CBT participants re-
ported improvements in depressed mood and anxiety/fears
(WHQ), mental health, and vitality (SF-36) at 6 weeks com-
pared with NTC, and at 26 weeks, they reported fewer memory/
concentration problems and somatic symptoms (WHQ) and
better physical and emotional role functioning (SF-36) com-
pared with NTC. Self-help CBT participants reported better
general health (SF-36), less depressed mood, and fewer memory/
concentration and sleep problems (WHQ) at 6 weeks and fewer
memory/concentration problems (WHQ) at 26 weeks com-
pared with NTC. Overall, group CBT had more impact on mood
and HrQOL than self-help CBT.
Our findings support previous exploratory work
29,39,47
and
are particularly important, given the frequent and problematic
nature of HF/NS in this sample, despite the fairly broad in-
clusion criteria used. Many women reported having chronic
symptoms and/or had taken HT in the past but still had sig-
nificant symptoms. We did not expect guided self-help CBT
to be as effective as group CBT, but the nature of the effects
on problem ratings and NS frequency, being very similar in
response to both interventions, suggests that the CBT may be
having a specific impact.
Comparison with other studies
Earlier CBT studies
29<31,39
suggested that CBT can reduce
the frequency of HF/NS and reduce HF/NS problem rating/
interference but were not sufficiently powered. We found that
group and self-help CBT can produce clinically significant
changes in HF/NS problem rating in 8 and 1.5 hours, respec-
tively (1-1.5 h per participant), compared with 12 hours
31
in a
previous group CBT study. In a recent study, Carmody et al
56
found that 20 hours of mindfulness training reduced HF/NS
bother by 15% at 9 weeks (posttreatment) and 22% at 20 weeks
(follow-up), with no effect on frequency; group and self-help
CBT resulted in an approximately 50% reduction in problem
rating at both 6 and 26 weeks. This level of improvement is
also greater than that found for HF/NS bother in a recent study
of escitalopram, a selective serotonin reuptake inhibitor.
18
Sys-
tematic reviews of acupuncture treatments have provided mixed
evidence for improvements in HF/NS.
57,58
Therefore, compared
with other inte rvention studies, group and self-help CBT are
brief and effective interventions in the treatmen t of HF/NS.
Previous studies
29,39
suggest that CBT might improve mood
and HrQOL. In the mindfulness study of Carmody et al
56
,
improvements were seen for HrQOL, which were maintained
at a 3-month follow-up. Mindfulness might be expected to
have a more general impact on mood because it is an inter-
vention focuse d on stress reduction rather than specifically on
focus in managing HF/NS. We found some improvements
FIG. 7. Women’s Health Questionnaire Anxiety/Fears. Unadjusted means
and 95% CIs at baseline and 6 and 26 weeks after randomization for group
CBT, self-help CBT, and no treatment control. CBT, cognitive behavior
therapy.
756 Menopause, Vol. 19, No. 7, 2012 * 2012 The North American Menopause Society
AYERS ET AL
Copyright © 2012 The North American Menopause Society. Unauthorized reproduction of this article is prohibited.
Page 8
in mood and HrQOL following group and self-help CBT. Im-
provements in reported concentration and memory are interest-
ing and warrant further study, given the high rates of memory
complaints during the menopausal transition. Memory prob-
lems are not necess arily associated with hormone levels or
stage of menopause but are more strongly associated with stress
and mood.
59<64
Arguably, CBT might affect memory and
concentration and other physical symptoms by reducing stress
and improving mood and teaching strategies to manage symp-
toms; however, having less problematic HF/NS and fewer night
sweats might also lead to improved concentration and memory.
Strengths and limitations of study
We had high rates of follow-up, low refusal and dropout
rates, and no adverse event or deaths during the study period,
and the trial was adequately powered. However, we did not
TABLE 4. Mean (SD) and significant differences (group CBT vs NTC, self-help CBT vs NTC): Hot Flush Rating Scale
(frequency and problem rating), WHQ, and SF-36 at baseline and at 6 and 26 wk after randomization
Group, mean (SD) Self-help, mean (SD) Control, mean (SD)
HF problem rating (1-10) Baseline 6.00 (2.15) 5.84 (1.93) 5.79 (2.76)
6 wk 3.01 (2.11)
a
2.96 (1.76)
a
4.97 (2.44)
26 wk 2.86 (2.11)
a
3.07 (1.93)
b
4.18 (2.45)
HF frequency Baseline 43.75 (34.31) 53.34 (50.21) 38.80 (43.41)
6 wk 33.85 (36.39) 36.38 (30.21) 34.67 (41.23)
26 wk 29.18 (47.30) 35.00 (37.21) 28.30 (33.22)
NS frequency Baseline 18.08 (12.29) 17.34 (12.16) 17.89 (13.04)
6 wk 10.00 (9.62)
c
12.83 (11.85) 15.00 (12.85)
26 wk 8.59 (11.83)
b
9.94 (8.78)
c
15.75 (18.92)
Total HF/NS frequency Baseline 61.83 (38.17) 70.68 (57.49) 56.69 (50.43)
6 wk 43.85 (42.16) 49.20 (39.24) 49.67 (48.55)
26 wk 36.77 (50.71) 44.94 (42.70) 44.05 (45.18)
WHQ Depressed Mood Baseline 0.27 (0.22) 0.33 (0.23) 0.30 (0.28)
6 wk 0.16 (0.20)
a
0.21 (0.19)
c
0.28 (0.24)
26 wk 0.19 (0.20) 0.15 (0.18) 0.23 (0.20)
WHQ Somatic Symptoms Baseline 0.55 (0.23) 0.55 (0.21) 0.60 (0.26)
6 wk 0.45 (0.22) 0.41 (0.25) 0.51 (0.27)
26 wk 0.35 (0.23)
a
0.29 (0.25) 0.54 (0.28)
WHQ Anxiety/Fears Baseline 0.46 (0.31) 0.43 (0.28) 0.43 (0.31)
6 wk 0.23 (0.29)
b
0.29 (0.25) 0.36 (0.34)
26 wk 0.26 (0.29) 0.26 (0.29) 0.33 (0.33)
WHQ Sleep Problems Baseline 0.70 (0.30) 0.64 (0.31) 0.70 (0.31)
6 wk 0.49 (0.36) 0.36 (0.30)
c
0.57 (0.35)
26 wk 0.53 (0.32) 0.41 (0.31) 0.57 (0.36)
WHQ Memory/Concentration Baseline 0.73 (0.35) 0.79 (0.32) 0.67 (0.38)
6 wk 0.60 (0.37) 0.57 (0.41)
c
0.63 (0.39)
26 wk 0.57 (0.33)
c
0.60 (0.40)
c
0.67 (0.33)
SF-36 Physical Functioning Baseline 83.19 (18.28) 87.23 (13.51) 74.67 (27.97)
6 wk 81.43 (18.88) 90.47 (12.53) 80.38 (18.08)
26 wk 86.92 (13.55)
c
86.50 (20.56) 73.59 (28.68)
SF-36 Role-Physical Baseline 80.32 (36.09) 80.32 (29.46) 60.80 (42.55)
6 wk 82.14 (32.33) 83.59 (28.12) 68.59 (37.04)
26 wk 80.77 (34.63) 82.50 (32.92) 62.82 (45.11)
SF-36 Bodily Pain Baseline 65.53 (23.39) 65.74 (22.82) 55.78 (22.41)
6 wk 67.14 (20.52) 70.63 (20.94) 58.21 (26.44)
26 wk 68.21 (19.04) 66.33 (23.27) 55.64 (24.37)
SF-36 General Health Baseline 68.83 (20.28) 68.09 (17.59) 69.09 (20.01)
6 wk 69.76 (18.64) 74.84 (15.89)
c
67.95 (22.03)
26 wk 72.95 (20.28) 73.17 (15.28) 68.59 (19.87)
SF-36 Vitality Baseline 49.26 (21.72) 48.83 (17.76) 46.44 (20.02)
6 wk 58.21 (22.95)
c
55.00 (19.92) 51.03 (21.74)
26 wk 57.18 (24.78) 58.00 (19.01) 53.21 (19.31)
SF-36 Social Functioning Baseline 77.66 (25.17) 74.20 (23.37) 70.28 (28.49)
6 wk 84.53 (20.81) 85.16 (20.93) 80.13 (24.12)
26 wk 86.86 (22.39) 87.50 (19.14) 78.53 (28.53)
SF-36 Role-Emotional Baseline 67.38 (39.00) 70.92 (36.53) 70.46 (41.43)
6 wk 80.16 (31.29) 77.08 (34.33) 73.50 (38.37)
26 wk 82.05 (34.92)
c
86.67 (28.50) 68.23 (42.84)
SF-36 Mental Health Baseline 69.02 (19.64) 64.77 (15.37) 65.24 (21.57)
6 wk 76.48 (14.39)
c
72.25 (12.61) 69.95 (19.68)
26 wk 76.31 (19.88) 72.80 (14.80) 70.26 (16.64)
Results from adjusted linear mixed model.
HF, hot flush; NS, night sweat; WHQ, Women’s Health Questionnaire; SF-36, General Health Survey Short FormY36; CBT, cognitive behavior therapy; NTC,
no treatment control.
a
Significant at P G 0.001.
b
Significant at P G 0.01.
c
Significant at P G 0.05.
Menopause, Vol. 19, No. 7, 2012 757
RCT OF CBT FOR MENOPAUSAL SYMPTOMS
Copyright © 2012 The North American Menopause Society. Unauthorized reproduction of this article is prohibited.
Page 9
include an activ e control group and therefore cannot deter-
mine whether improvements occurred because of the addi-
tional attention CBT participants received through interaction
with others and with the psychologist during the intervent ion.
Placebo effects are reported in other HF/NS treat ment studies
and are higher in medication trials (up to 30%) than for other
interventions.
21
In the current study, 21% of the NTC partic-
ipants had a clinically significant reduction in HF/NS problem
rating at 26 weeks, which is similar to an expected placebo
effect and may be because of the added attention of being in a
trial or spontaneous remission. Moreover, the fact that guided
self-help CBT was as effective as group CBT sugges ts that the
additional attention time of approximately 6 hours that the
group CBT participants received did not influence HF/NS
outcomes. In other respects, the two treatment conditions of-
fered exactly the same active components and resulted in the
same specific outcomes (problem rating and NS frequency).
The sample included women from a range of ethnic and
educational backgrounds, and women were recruited from a
variety of accessible community and primary care settings. Our
analyses report fairly small CIs, indicating reliability within the
results. In addition, as a sensitivity analysis, the primary anal-
ysis was repeated for those participants who met protocol com-
pliance (ie, those who attended two or more therapy sessi ons).
The analyses reported the same results. For these reasons, we
can have confidence<