Poor-quality studies suggest that vitamin B₆ use is beneficial in premenstrual syndrome

QUESTION

Is the use of vitamin B₆ (pyridoxine hydrochloride) effective for relieving overall premenstrual symptoms and depressive premenstrual syndrome?

DATA SOURCES

Studies were identified by searching MEDLINE (1966-1998), EMBASE (1988-1996), PsychLIT (1974-1997), CINAHL (Cumulative Index to Nursing and Allied Health) (1982-1997), and the Cochrane Controlled Trials Register with the terms premenstrual syndrome, pyridoxine, premenstrual tension, late luteal phase dysphoric disorder, and premenstrual dysphoric disorder. Bibliographies of relevant articles were scanned, and pharmaceutical companies were contacted.

STUDY SELECTION

Published and unpublished studies in any language were selected if they were randomized, double-blind, placebo-controlled studies for which data could be acquired. Studies involving multivitamin supplements were included if the supplement contained ≥50 mg of vitamin B₆. Authors were contacted if additional information was required.

DATA EXTRACTION

Data were extracted on participants, dose and preparation of vitamin B₆, outcome measures, results, withdrawals, side effects, and quality scores. Main outcome was subjective lessening of overall premenstrual symptoms, and a secondary outcome was a decrease in depressive premenstrual symptoms. High-quality studies had scores of 3 on the Jadad scale (maximum: 5 points) and of 6 on the authors' quality rating scale (maximum: 8 points).

MAIN RESULTS AND CONCLUSIONS

Ten studies with 12 comparisons met the inclusion criteria. Sample sizes ranged from 31 to 434 patients (mean: 98 patients). Doses ranged from 50 mg per day (3 comparisons) to 600 mg per day (1 comparison). Quality scores on the Jadad scale were 1 (3 studies), 2 (1 study), and 3 (3 studies) points. On the authors' quality rating scale, scores were 3 (2 studies), 4 (5 studies), 5 (2 studies), and 6 (1 study) points. Quality scores were not used to exclude trials from the analysis because few high-quality studies were found.

Pooled results showed that vitamin B₆ relieved premenstrual symptoms (combined odds ratio: 2.12; 95% confidence interval: 1.80-2.48). One study led to heterogeneity (P<0.001), and the removal of this study results in a combined odds ratio of 1.69 (95% confidence interval: 1.39-2.06). The three high-quality studies according to the Jadad scale showed a benefit for the use of vitamin B₆, whereas the one high-quality study according to the authors' quality rating showed no effect; this study had low power for detecting any differences. No dose response was found in the nine trials. A meta-analysis of randomized, double-blind, placebo-controlled trials shows that the use of vitamin B₆ relieves overall premenstrual and depressive premenstrual symptoms. Most studies were of poor quality.