This study was a randomized, placebo controlled trial, whose primary aim was to evaluate the efficacy of flaxseed in reducing hot flash activity in women. Secondary aims included the evaluation of the side effect profile for flaxseed and to evaluate its effects on secondary endpoints, namely activity interference from hot flashes, mood and other menopausal symptoms. Women eligible for this trial included those who were > 18 years, postmenopausal, with or without a history of breast cancer (currently without malignant disease and having completed chemotherapy/radiation), who wished to avoid estrogen based treatment. The women had to have experienced hot flashes for > 1 month and hot flashes had to occur ≥ 28 times per week. Women with a history of allergies to flaxseed, current or planned use of other agents for treating hot flashes (i.e.: gabapentin, clonidine, antidepressants, hormonal treatments), were excluded. Women could not be receiving anticoagulants, anti-hypertensives or have co-morbid gastrointestinal disorders or uncontrolled diabetes. Tamoxifen, raloxifene, or aromatase inhibitors were allowed if the participant had been on a consistent dose for ≥4 weeks and were not planning to stop the medication during the study period.
Eligible participants were randomly assigned by computer using dynamic allocation to either flaxseed or placebo, stratified by age (18–49 vs. ≥50), current use of tamoxifen/selective estrogen receptor modulator (SERM)/aromatase inhibitor (AI), duration of hot flashes (≤9 months vs. > 9 months), and number of daily hot flashes (4 to 9 vs. ≥10).
The study agent was administered as a bar, containing 7.5 grams of flaxseed (5% lignans) providing 410 mg of lignans, 6 grams of protein and 20% fiber. It contained 190 calories. The matching placebo contained 2 grams of protein, 20% fiber and 200 calories and did not have any flaxseed, lignans or soy products. Both active and placebo bars were provided by Glanbia Nutritionals, who also arranged product quality testing and content verification by an independent company. Good manufacturing processes were used.
Participants did not eat any study bars during the first study week, but kept daily diaries and completed baseline questionnaires. This baseline week was followed by 6 weeks in which participants were instructed to eat one bar daily, in keeping with the design of the positive pilot trial. This could be eaten all at once or throughout the day. The primary outcome measure was a prospective self report daily diary, in which participants recorded both the frequency and severity of hot flashes during each 24 hour period over the 7 week study period. Secondary endpoint measures included the Hot Flash Related Daily Interference Scale (HFDIS)25
, the Profile of Mood States (POMS)26
, the Menopause Quality of Life Scale (MenQOL) 27,28
, the Global Impression of Benefit 29
, and a side effect questionnaire. Side effect and menopausal symptom information were collected weekly, while the HFRDIS and POMS were completed at the end of the study. Study personnel called women during weeks 2, 3, 5 and 7 to answer questions, encourage adherence and grade toxicities per the Common Terminology Criteria.
The intra-patient difference in hot flash activity between baseline (study week 1) and the last treatment week was the primary endpoint. Hot flash activity was measured by the weekly average hot flash score 24
, which is a composite entity of both frequency and severity of hot flashes. The daily hot flash score was computed by multiplying the mean severity grade by the frequency over a 24 hour period. Taking the average of daily hot flash scores over a week produces the weekly average hot flash score. The independent two-sample t
-test or Wilcoxon rank-sum test was used to examine the change of weekly average hot flash score from baseline to treatment termination between flaxseed and placebo arms. All endpoints for the secondary analysis were assessed for changes from baseline to the last treatment week, except for Global Impression of Benefit which was only evaluated at the end of the study, before the patient was unblinded. Participants who did not provide baseline or week 7 data were excluded from the analysis. Patterns of missing data were evaluated and were found to be similar between groups.
From previous hot flash trials, a mean reduction in hot flash score for placebo was estimated to be 3.5 units, with an additional 3.5 units between the placebo and flaxseed arm considered to be clinically important 24
. Therefore, using a two sided independent sample t-test, an accrual goal of 64 patients per arm to provide 80% power with a 5% type I error rate was desired. An additional 13 patients per arm were accrued to account for missing data.