The best magazine
Methods
Participants
Postmenopausal women with hot flashes were invited to participate in this open randomized controlled study through advertisements in a local newspaper. Enrollment took place in one gynecological outpatient university-based clinic in Sweden between January 2007 and September 2010. The eligibility criteria were as follows: (a) postmenopausal women (at least 12 mo since the last menstrual bleeding occurred or, in previously hysterectomized women, serum follicle-stimulating hormone level categorized as postmenopausal according to references in a local laboratory); (b) seven or more moderate to severe hot flashes per 24 hours or 50 or more hot flashes per week according to a 2-week screening diary; (c) ability to understand and speak Swedish; and (d) freely given informed consent forms. The exclusion criteria were as follows: (a) unstable thyroid disease or other metabolic diseases; (b) treatment of menopause-related symptoms with HT or other complementary or alternative treatments; (c) treatment with psychopharmacological drugs and/or sedatives; (d) untreated psychiatric disease; (e) frequent exercise (≥2 h of high-intensity activities per week). All eligible women were offered medical examination and gynecological examination, including vaginal ultrasound, before inclusion in the study.
The trial was conducted in accordance with the Declaration of Helsinki. Approval of the Regional Ethics Review Board in Linköping was received before the start of the study. All women received oral and written information about the study. A written informed consent form was obtained from all study participants before inclusion.
Intervention
Eligible women were randomized to either AR, which required participation in 10 group sessions during a period of 12 weeks, or to participation in an untreated CG. The therapist (E.N.) met the women assigned to AR in a group consisting of six to eight women. The women were told to practice each component at home daily.
The weekly sessions based on a scheme from Ost, as we have previously described, lasted for 60 minutes each. During the first session, the women in the AR group were given a lecture about menopause and theories of AR mechanisms and how AR may be applied in vasomotor symptoms. The group was given a rationale for applying AR as a coping technique for handling sudden unanticipated symptoms by quickly calming down and thus gaining control of the situation. The women assigned to the CG were told that, apart from continuing the daily registration of hot flashes, they were to act as an untreated CG using none of the following: hormonal treatment, other alternative medications, and natural remedies for hot flashes (not even acupuncture, mind-body therapies, or intensive physical activity). After 6 months of participation in the CG, they were offered the AR program.
Outcomes
The primary endpoint was the average number of moderate and severe hot flashes per 24 hours recorded in self-registered diaries on the 12th week in the AR group compared with the CG. The secondary variable was the average number of moderate and severe hot flashes per 24 hours after the 3-month follow-up and in the subgroup of women in the AR group who were defined as responders. Responders were defined as women with a 50% or higher reduction in hot flashes from baseline to the 12th week and at 3 months after “end of treatment.” Moreover, the total scores in the Women’s Health Questionnaire (WHQ), as measured on the 12th week and after the 3-month follow-up, were recorded.
The third endpoint was to follow the secretion of cortisol in all women to determine if the daily pattern of excretion would change within the AR group compared with the CG.
Adverse events were documented by using open-ended questions at each contact.
Measurements of Study Variables
Hot Flashes
Paper-and-pencil hot flash diary was used to assess hot flash frequency during the 2-week screening period and during 12 weeks in the AR group or the CG. Both groups had a 3-month follow-up, with hot flash registration in the diary at 1 week per month. Severity was subjectively estimated on a scale consisting of the following: (1) mild hot flashes (hot flash without sweating, not bothersome); (2) moderate hot flashes (bothersome hot flash and sweating, but without interruption of daily/nightly activity); or (3) severe hot flashes (bothersome hot flash and sweating, with interruption of daily/nightly activity). The women were told to register twice daily (ie, in the morning, the women should register all hot flashes that occurred in the night; in the evening just before bedtime, they were told to register all hot flashes that occurred during daytime). The women were encouraged to register continuously during the day.
The baseline average number and severity of flashes per 24 hours were calculated from the 2-week screening period, and the 12th week figure was calculated as an average of the 7 days of the 12th week. Data from the diaries were missing in four women on the 12th week (three women in the AR group and one woman in the CG) and in three women at the 3-month follow-up (two women in the AR group and one woman in the CG). In addition, nine women (eight women in the AR group and one woman in the CG) did not complete every recording during the 12th week, and a carry-forward procedure was used (ie, registrations on the 11th week were carried forward to replace missing values on the 12th week). No data were missing at the 3-month follow-up.
Health-related Quality of Life
HRQoL was assessed using the WHQ at baseline; on the 12th week in the AR group and the CG, respectively; and at the 3-month follow-up. The questionnaire was used as a self-administered instrument validated and used to assess middle-aged women’s physiological and emotional health. It has been widely used in multinational clinical trials and has high internal reliability (Cronbach’s α = 0.70-0.84) and test-retest reliability (Cronbach’s α = 0.78-0.96).
The WHQ has been validated for Swedish conditions and contains 36 items assessing nine subscales (depressed mood, anxiety/fear, vasomotor symptoms, sleep problems, somatic symptoms, menstrual symptoms, sexual behavior, memory/concentration, and attractiveness). Each item is rated on a four-point scale to reflect frequency (yes, definitely; yes, sometimes; no, not much; no, not at all) and reduced to a binary option ranging between 0 and 1. The subscales are summarized and divided by the number of items for each subscale. The value 0 is the most positive option, and the value 1 is the most negative option. The WHQ is sensitive to change, and a meaningful clinically significant change on the subscales 34 is suggested as a difference of 0.10 to 0.20.
In total, 4 (2%) of 180 questionnaires were missing.
Salivary Cortisol
Before collecting their saliva, the women were instructed not to eat, drink, or smoke later than 60 minutes before sampling. For women working nighttime, the instruction was to collect the sample on a “free day.” Brushing of teeth was not allowed for 60 minutes before sample collection to minimize the risk of blood contamination. The women were not allowed to perform physical exercise within 30 minutes before sample collection and were instructed to rest (preferably lying down) for at least 30 minutes.
Saliva was collected with Salimetrics Oral Swab (http://www.salimetrics.com) three times during a 24-hour period: at awakening, 30 minutes after awakening, and at bedtime. The procedure was repeated by all women three times during the study: at baseline, on the 12th week, and after the 3-month follow-up (end of the study). The analyses were performed at the laboratory of the Linköping University Hospital according to the recommendations of the manufacturer. All samples were analyzed in one batch after the end of the study. A commercial enzyme immunoassay kit designed for the analysis of salivary cortisol was used (http://www.salimetrics.com). The measurable range was 0.3 to 82 nmol/L. Calibrations and quality control in two different levels were performed at the beginning and at the end of each assay. The coefficient of variation was below 6%.
Adverse Events
At every follow-up contact, all women were given open-ended questions on whether any adverse events had occurred since the last visit.
Randomization
An independent statistician prepared a computer-generated randomization list. The randomization process was unknown to the study team. In total, at least 60 women were planned for randomization to either the AR group or the CG.
If the inclusion and exclusion criteria were met, a sealed opaque envelope was opened in consecutive order by the investigator together with the women.
Statistical Methods
Sample size assumption was based on the assumption that using AR for vasomotor symptoms would reduce both the severity and the frequency of hot flashes by at least 50% from baseline. Standard deviation was estimated to be one hot flash per 24 hours. To obtain a power of 80%, we would need 20 women per group to detect a significant difference between groups (P < 0.05, two-sided test).
The study was also designed to analyze variation in 24-hour salivary cortisol levels and a possible relation to hot flashes. Because there have been no previous studies in this area and to compensate for dropouts, we estimated that we would need approximately 50% more women for this analysis and thus decided to include about 30 women in each group.
Data were analyzed according to “intention to treat” (ie, including all women who fulfilled the inclusion criteria, had no exclusion criteria, initiated their treatment or control arm according to randomization, and had at least one hot flash measurement registered).
Baseline and demographic data were described as mean (SD). To compare differences between the two groups, we used Student’s t test for hot flashes and Mann-Whitney U test for analysis of differences in the change of scores in the WHQ. Repeated-measures analysis of variance was used to compare hot flashes per 24 hours between the two groups from baseline to the 3-month follow-up.
Descriptive statistics were used to analyze salivary cortisol in each studied group, and Mann-Whitney U test was used for measurements between groups.
P < 0.05 (two-sided) was considered significant. Data analyses were performed by Statistica from Stat Soft, standard version 9. All statistical analyses were performed with assistance from an independent statistician.
Source: ...