Twenty RCTs with a total of 1,220 participants were included in the review. Twelve of the included studies used a parallel design (981 participants) and eight used a crossover design (239 participants). Sixteen trials were double-blind, whilst four did not report details of blinding. The number of participants in the included studies ranged from 13 to 461.
The authors stated that their funnel plot was symmetrical, minimising the possibility that publication bias was present.
Magnesium supplementation resulted in a small overall reduction in systolic BP of 0.6 mmHg (95% CI: -2.2, 1.0, P=0.051) and a small overall reduction in diastolic BP of 0.8 mmHg (95% CI: -2.1, 0.5, P=0.142). There was significant heterogeneity between studies for both systolic and diastolic BP changes (P<0.001), indicating that other factors could have had an impact on the efficacy of magnesium in reducing BP.
When limiting the analysis to the 16 double-blinded studies, a 10 mmol/day increase in magnesium intake was associated with a decrease in systolic BP of 3.4 mmHg (95% CI: 0.6, 6.1, P=0.02) and a decrease in diastolic BP of 2.0 mmHg (95% CI: -0.9, 5.0, P=0.18).
When limiting the analysis to the 14 trials of patients with hypertension, a 10 mmol/day increase in magnesium intake was associated with a decrease in systolic BP of 3.3 mmHg (95% CI: -0.1, 6.8, P=0.06) and a decrease in diastolic BP of 2.3 mmHg (95% CI: -1.0, 5.6, P=0.17).
For each 10 mmol/day larger dose of magnesium, systolic BP decreased by 4.3 mmHg (95% CI: 2.2, 6.3, P<0.001) and diastolic BP decreased by 2.3 mmHg (95% CI: 0.0, 4.9, P=0.09). In addition, the effect of magnesium on systolic BP was stronger in trials with lower baseline systolic BP, although this relationship was not found for diastolic BP.
Age, gender, trial design and trial duration were not significantly associated with magnesium efficacy.