Köseoglu · Magnesium research 2008 · randomized controlled trial · n=40

The effects of magnesium prophylaxis in migraine without aura.

Cited 94 times in the scientific literature.

Level 2 - randomized trial

Double-blind randomized placebo-controlled trial

PubMed 18705538 · record verified 2026-08-31

What was done

A double-blind, randomized, placebo-controlled study evaluated the prophylactic effects of oral magnesium citrate (600 mg/day for 3 months) in 40 patients aged 20–55 years who experienced 2–5 migraine attacks per month without aura. Patients were assigned to either magnesium treatment (n = 30) or placebo (n = 10). Outcomes assessed before and after the 3-month period included clinical migraine attack frequency and severity, visual evoked potential (VEP) P1 amplitude, and cortical blood flow measured by statistical parametric mapping of brain single-photon emission computed tomography (SPECT).

What was found

Compared with pre-treatment values, patients receiving magnesium had significant decreases in attack frequency, attack severity, and VEP P1 amplitude (all p < 0.001). Comparing post/pre-treatment ratios between groups, the magnesium group showed significantly greater reductions than the placebo group in attack frequency (p = 0.005), attack severity (p < 0.001), and P1 amplitude (p < 0.05). Cortical blood flow increased significantly after magnesium treatment in inferolateral frontal (p < 0.001), inferolateral temporal (p = 0.001), and insular regions (p < 0.01), whereas no significant cortical blood flow changes were observed with placebo. Absolute baseline and post-treatment values (means/SDs) were not reported in the abstract.

Why it matters

This study provides randomized evidence that daily magnesium citrate supplementation reduces the frequency and severity of migraine attacks without aura, suggesting concurrent neurogenic and vascular mechanisms of action.

Limits

The overall sample size was small (n = 40) with an unequal 3:1 allocation ratio (30 magnesium vs. 10 placebo), limiting statistical power in the control arm. The abstract reports p-values and ratios but omits absolute numerical values, baseline statistics, confidence intervals, and adverse event data.

Cited by