Toggle Main Menu Toggle Search

Open Access padlockePrints

Inclusion of crossover trials in meta-analysis of nutritional supplements for raised blood pressure.

Lookup NU author(s): Dr Heather Dickinson, Donald Nicolson, Dr Fiona CampbellORCiD, Professor James Mason


Full text for this publication is not currently held within this repository. Alternative links are provided below where available.


Objective: To inform a national guideline development group about the effectiveness of nutritional supplements to reduce hypertension. Methods: RCTs of parallel and crossover design were included, comparing potassium, calcium or magnesium supplements with placebo or no treatment. Participants had blood pressure ≥140/85 mm Hg and ≥8 weeks follow-up. Relevant studies were identified from existing systematic reviews and by searching electronic databases from 1998 onwards. Meta-analyses estimated weighted mean differences using a random effects model. Results: We found 23 studies (12 parallel and 11 crossover) of 1,126 patients (769 and 357 respectively). Crossover trials rarely reported the variance of the crossover treatment effect. We investigated various solutions: stochastic and deterministic imputation of this variance, use only of first periods of crossover trials, or exclusion of such trials. Conclusions: Omission of crossover trials from meta-analyses is a poor use of available evidence. However, choice of method of inclusion did not materially affect the conclusions. There was no evidence that these nutritional supplements reduced blood pressure.

Publication metadata

Author(s): Dickinson H, Nicolson DJ, Campbell F, Mason J

Publication type: Conference Proceedings (inc. Abstract)

Publication status: Published

Conference Name: XI Cochrane Colloquium: Evidence, Health Care and Culture

Year of Conference: 2003