Twenty-seven RCTs were included in the review, which collectively presented data on 2,724 women.
Forty-one controlled trials were identified from the searches. Of these 14 were excluded for one or more of the following reasons: not randomised (6); published only in conference or dissertation abstract form, and not derived from peer-reviewed publications (5); only one septum investigated and not the syndrome as a whole (4); assessed health volunteers (3); the diagnosis of interest was primary dysmenorrhoea (1). The review, therefore, consisted of 27 studies.
Herbal medicine (7 RCTs).
Two independent investigations of the chaste tree (Vitex agnes castus) failed to find strong evidence to support the herbal preparation. In one of these trials, only one of the 20 symptoms was significantly improved, and this was mirrored by a similar improvement in patients treated with the control intervention. The other trial reported that the chaste tree was at least as effective as vitamin B6, but the trial was also methodologically flawed, partly because of it being underpowered. An assessment of ginkgo (Ginkgo biloba) found that only one of a range of measures related to breast pain was improved. In 4 studies investigating 3 to 6 mg doses of evening primrose oil (Oenothera biennis), no significant association was seen in favour of the oil over control medicines. These studies were, however, of low power.
Homeopathy (1 RCT).
One study that investigated homeopathy had such strict inclusion criteria that only 10 of the 205 patients who were screened actually participated in the study.
Dietary supplements (13 RCTs). Calcium supplements were studied in 2 RCTs, both led by the same investigator. Despite methodological flaws, the studies both reported that patients offered supplements had improvements in a range of symptoms, compared with patients in the control arm. In investigations of magnesium supplements, 2 separate small studies found improvements with magnesium, compared with control therapy, but for different symptoms. The investigators of one of these studies repeated their study to find no advantage of magnesium alone, but an improvement for magnesium when taken in combination with vitamin B6.
Two small studies, each by the same investigators, assessed vitamin E. One, a dose response study, found 300 IU/day to be optimal when compared with 150 and 600 IU/day. The subsequent study, using a dose of 400 IU/day, failed to find any association at the 5% level of significance.
In investigations of a nutritional multi-supplement containing large amounts of magnesium, vitamin B6 and the essential micro-nutrients, researchers used doses of 4 to 12 tablets/day. One study found that 6 tablets/day was significantly better than placebo in terms of the overall response to the MSQ scale, anxiety levels and cravings. Another study that compared 8 tablets with 4 tablets found, on subjective ratings, that significantly more women reported improvements on the higher dose. An additional study investigated 6 and 12 tablets and found both to be superior to placebo, but the reviewers raised ethical concerns owing to the 12-tablet dose exceeding the safe daily intake for two of its components. A study of another supplement consisting of magnesium, vitamin B6 and yeast, found significant improvements in 5 out of 7 septum clusters at 6 months, measured in terms of the MDQ scale.
Positive effects on appetite, memory and mood were found in study of a drink containing simple and complex carbohydrates, compared with one of two control drinks containing a protein-carbohydrate mixture or simple carbohydrates alone. However, only 24 of the 99 participants in this study were analysed owing to protocol breeches. Other interventions.
A small trial of progressive muscle relaxation, compared with reading or charting symptoms, found significant improvements in the physical symptoms of PMS. In those women with more severe PMS, the overall score on the two rating scales and the emotional septum cluster were also improved. Another trial compared massage therapy with progressive relaxation. Although improvements over baseline were reported, no comparisons between the groups were reported.
Sham-controlled reflexology was found to be effective in terms of the overall somatic and psychological symptoms reported in diary format by the patient. In this study, therapist blinding was not possible.
The initial results of a sham-controlled chiropractice crossover trial appeared to show a benefit of chiropractice. However, reanalysis showed the first intervention led to significant improvements over baseline, irrespective of the order in which the sham therapy and chiropractic were administered, whereas there was little further improvement associated with the second intervention.
Two related studies of biofeedback were reported by the same investigators. A significant reduction was seen in the overall condition of the patients, including significant results in relation to physiological and affective symptoms.