Clinical Hypnosis in the Treatment of Postmenopausal Hot Flashes: A Randomized Controlled Trial

Gary R. Elkins, PhD; William I. Fisher, MA; Aimee K. Johnson, MA; Janet S. Carpenter, PhD, RN, FAAN; Timothy Z. Keith, PhD


Menopause. 2013;20(3):291-298. 

In This Article


Study Design

The study was conducted as a single-blind, randomized, clinical trial in central Texas evaluating the effectiveness of clinical hypnosis compared with active structured-attention control. The study was approved by the university institutional review board, and all participants provided a written informed consent form.

Participant Selection

Women were recruited from December 2008 to April 2012 via newspaper advertisements, professional referrals, and television and billboard advertisements. Participants were eligible if they were postmenopausal, were 18 years or older, and had no menstrual period in the past 12 months or had no menstrual period in the past 6 months and either (a) had a medically documented history of a follicle-stimulating hormone level greater than 40 mIU/ml or (b) had undergone bilateral oophorectomy. Furthermore, participants were required to have a self-reported history of seven hot flashes per day at minimum (or 50 hot flashes per week) at baseline.

Participants were required to have discontinued estrogen/progestin-containing products based on Food and Drug Administration draft recommendations for industry:[26] 1 week or longer for prior vaginal hormonal products (rings, creams,and gels), 4 weeks or longer for prior transdermal estrogen-alone or estrogen/progestin products, 8 weeks or longer for prior oral estrogen and/or progestin therapy, 8 weeks or longer for prior intrauterine progestin therapy, 3 months or longer for prior progestin implants and estrogen-alone injectable drug therapy, and 6 months or longer for prior estrogen pellet therapy or progestin injectable drug therapy.

Women were excluded if they were receiving any other treatment of hot flashes or using any complementary or alternative medical treatments of vasomotor symptoms (including soy, black cohosh, phytoestrogens, and any other mind-body techniques). Participants were excluded from the study if they had a history of psychosis, borderline personality disorder, or serious psychopathology because these diagnoses are considered to be contraindications for clinical hypnosis.

Data Collection

Potential participants were screened for eligibility by telephone. Women who met the eligibility criteria completed baseline measures, including the Hot Flash Symptoms Diary,[27,28] physiological monitoring of hot flashes, and other secondary outcome measures. Random assignment was made by the study biostatistician from a confidential computer generated list of permuted blocks of varying sizes. Participants were randomized by sealed envelope. The envelope was not opened by the research coordinator until the women had completed and provided their baseline data. Participants were scheduled for five weekly sessions of either clinical hypnosis or structured attention. Participants completed follow-up assessments on weeks 6 and 12. Participants were paid after three intervention or control sessions and at the final follow-up contact, for a possible total of US$300.


The clinical hypnosis intervention followed a treatment manual and was delivered by therapists who were specifically trained in clinical hypnosis according to established training standards.[29] The clinical hypnosis intervention consisted of hypnotic inductions and instruction in the practice of self-hypnosis toward the therapeutic goals of reduction of hot flashes and improved sleep.[30,31,32,33,34,35,36,37] In each 45-minute session, participants were provided specific suggestions for mental imagery for coolness, safe place imagery, and relaxation (individualized based on the women’s preference). Participants were also provided an audio recording of a hypnotic induction and tasked with the daily practice of self-hypnosis at home. Treatment fidelity and compliance were evaluated at each session by means of a checklist.

The structured-attention control condition, designed to match the clinical intervention in therapist exposure, therapeutic environment, interpersonal exchange, and encouragement, was used based on the recommendations for minimal-effect interventions.[38] Structured attention matched the hypnotic intervention in that it consisted of five sessions, where discussion of symptoms, attentive listening, interpersonal exchange, avoidance of negative suggestions, monitoring, measurement, and encouragement were provided in a therapeutic environment by a clinician trained to deliver the control intervention, using a standard structured-attention manual. However, no hypnotic inductions or cooling suggestions were provided. Each structured-attention session lasted approximately 45 minutes, and fidelity was determined by a checklist. Participants were given a CD that provided information about hot flashes and tasked with daily listening.


Hot flash frequency and hot flash score were obtained using the Hot Flash Symptoms Diary.[27,28] Participants recorded their hot flashes for 7 days by daily frequency and severity (mild, moderate, severe, and very severe), and the diary was administered at baseline, on weeks 2 to 6, and at 12-week follow-up. This instrument provides a measure of hot flash frequency and hot flash score (product of frequency x severity).

A sternal skin conductance monitoring system was used to objectively measure hot flash frequency. The Biolog ambulatory recorder[39] recorded skin conductance levels using Biopac EL-507 silver/silver chloride electrodes for electrodermal activity and a 0.5 constant voltage circuit.[40] Electrodes are 1.0 cm in diameter and filled with 0.5% chloride solid gel.[41] Electrodes are attached 1.5 in. below the collar bones and 2 in. on either side of the sternal midline. The Biolog is programmed to sample 12-bit skin conductance data at 1 Hz (once per second). Customized software (FlashTrax, version 1.2; UFI, Morro Bay, CA) was used to evaluate hot flashes. Based on published norms,[10,11,12,13] hot flashes were flagged if there was an increase in sternal skin conductance of at least 2 Kmho within a 30-second period, with a 15-minute postevent lockout. Each skin conductance track was evaluated by a trained expert, verifying that each skin conductance event matched published norms for vasomotor events. Physiological monitoring of hot flashes was recorded for 24 hours at baseline, on week 6, and on week 12.

Secondary outcomes included measures of hot flash daily interference and sleep quality. To investigate the impacts of hot flashes on women’s overall quality of life, we used the Hot Flash Related Daily Interference Scale in this study.[42] This measure has been shown to be internally consistent and valid.[42] Data from this instrument were collected at baseline, on week 6, and at 12-week follow-up. Participants were given the Pittsburgh Sleep Quality Index (PSQI)[43] at baseline, at 6-week follow-up, and at 12-week follow-up to investigate quality of sleep. Treatment satisfaction was assessed at 12-week follow-up using a 0 to 10 rating (0 = "dissatisfied" to 10 = "completely satisfied"). Adverse events were assessed at each contact by women’s self-reports, and events were logged and reported to the institutional review board.

Statistical Analyses

Sample size was based on data from a pilot study that compared the hypnosis intervention to no-treatment control.[21] We estimated power calculations cognizant of the placebo effect in hot flash treatment literature, which reported effects of as large as a 30% reduction in hot flash frequencies and hot flash severity scores.[27,28] Calculations were conducted using G*Power, which takes into account expected effect size, desired power, correlation between pretest and posttest main effects, and sphericity.[44] Given the effect size (d = 0.488) in pilot data,[21] an α of 0.05, and a power of 0.90, a total sample of 146 (73 in each arm) was determined.

The modified intention-to-treat analysis included all randomized participants who provided diary data, which were analyzed regardless of adherence to protocol. Where appropriate, data were assessed for normality and homogeneity. Missing data were accounted for using maximum likelihood imputation.[45,46] To test primary outcomes, we performed two analyses of covariance (ANCOVA): one between-subjects factor (experimental condition) and one covariate (pretest). Perceived frequency and physiologically recorded frequency were analyzed separately. Analyses using a repeated measure (posttest and follow-up) were also conducted with the 12-week follow-up data to examine if the effects of the hypnosis intervention were maintained. Four separate analyses were conducted. The hypotheses associated with secondary outcomes were tested using a series of ANCOVA with pretest scores as covariate; follow-up data were analyzed with repeated-measures ANCOVA conducted using the SPSS statistical software package, version 19 (SPSS Inc., Chicago, IL).