See corresponding editorial on page 269.
Dietary magnesium intake and risk of stroke: a meta-analysis of
Susanna C Larsson, Nicola Orsini, and Alicja Wolk
Background: Prospective studies of dietary magnesium intake in
relation to risk of stroke have yielded inconsistent results.
Objective: We conducted a dose-response meta-analysis to summa-
rize the evidence regarding the association between magnesium in-
take and stroke risk.
Design: Relevant studies were identified by searching PubMed and
EMBASE from January 1966 through September 2011 and review-
ing reference lists of retrieved articles. We included prospective
studies that reported RRs with 95% CIs of stroke for ?3 categories
of magnesium intake. Results from individual studies were com-
bined by using a random-effects model.
Results: Seven prospective studies, with 6477 cases of stroke and
241,378 participants, were eligible for inclusion in the meta-analy-
sis. We observed a modest but statistically significant inverse asso-
ciation between magnesium intake and risk of stroke. An intake
increment of 100 mg Mg/d was associated with an 8% reduction
in risk of total stroke (combined RR: 0.92; 95% CI: 0.88, 0.97),
without heterogeneity among studies (P = 0.66, I2= 0%). Magne-
sium intake was inversely associated with risk of ischemic stroke
(RR: 0.91; 95% CI: 0.87, 0.96) but not intracerebral hemorrhage
(RR: 0.96; 95% CI: 0.84, 1.10) or subarachnoid hemorrhage (RR:
1.01; 95% CI: 0.90, 1.14).
Conclusion: Dietary magnesium intake is inversely associated with
risk of stroke, specifically ischemic stroke.
Am J Clin Nutr
Dietary magnesium intake has been inversely associated with
risk factors for stroke such as hypertension (1, 2), metabolic
syndrome (3), and type 2 diabetes (2, 4). Randomized clinical
trials have shown that magnesium supplementation modestly
reduces diastolic blood pressure (5) as well as fasting C-peptide
studies have shown that high-magnesium diets have favorable
effects on plasma glucose and blood lipid concentrations (7), and
magnesium deficiency increases the susceptibility of lipoproteins
to peroxidation (8). Hence, a high dietary magnesium intake may
potentially reduce risk of stroke.
To our knowledge, the epidemiologic evidence on the relation
between dietary magnesium intake and risk of stroke has not yet
been summarized. Therefore, we performed a systematic review
and dose-response meta-analysis to assess the association be-
tween magnesium intake and risk of total stroke and stroke
Literature search and selection
We followed standard criteria for conducting and reporting
of meta-analyses of observational studies (9). We conducted
a literature search of PubMed (http://www.ncbi.nlm.nih.gov/
pubmed) and EMBASE (http://www.embase.com) from January
1966 through September 2011 by using the key words “mag-
nesium intake”combined with“stroke.” Inaddition, we reviewed
reference lists of retrieved articles to identify additional relevant
studies. No language restrictions were imposed.
Studies were included in this meta-analysis if they fulfilled the
following criteria: 1) had a prospective design, 2) the exposure
of interest was magnesium intake, 3) the outcome was stroke,
and 4) the studies reported RRs with 95% CIs for ?3 quanti-
tative categories of magnesium intake. RRs had to at least be
adjusted for age and sex (if applicable).
The following data were extracted from each study: first
author’s last name, publication year, name of the cohort study,
study location, years of follow-up, sex, age, sample size (number
of cases and total number of participants), magnesium intake
categories, covariates adjusted for in the multivariable analysis,
and RRs with their 95% CIs for each category of magnesium
intake. We extracted RRs that reflected the greatest degree of
adjustment for potential confounders. Data extraction was con-
ducted independently by 2 investigators (SCL and NO) with
disagreements resolved by consensus.
1From the Division of Nutritional Epidemiology, National Institute of
Environmental Medicine, Karolinska Institutet, Stockholm, Sweden.
2Funding sources had no role in the design and conduct of the study;
collection, management, analysis, and interpretation of the data; or prepara-
tion, review, or approval of the manuscript.
3Supported by a research grant from the Swedish Council for Working
Life and Social Research and a Research Fellow grant from Karolinska
4Address correspondence to SC Larsson, Division of Nutritional Epide-
miology, National Institute of Environmental Medicine, Karolinska Institutet,
Box 210, SE-17177 Stockholm, Sweden. E-mail: email@example.com.
Received June 23, 2011. Accepted for publication October 24, 2011.
First published online December 28, 2011; doi: 10.3945/ajcn.111.022376.
Am J Clin Nutr 2012;95:362–6. Printed in USA. ? 2012 American Society for Nutrition
by guest on October 26, 2015
Because the range of magnesium intake and the cutoffs for
the categories differed between studies, we estimated an RR
with 95% CI for an increase in intake of 100 mg Mg/d for
each study. We used the method proposed by Greenland and
Longnecker (10) and Orsini et al (11) to compute the trend from
the correlated log RR estimates across categories of magnesium
intake. For each study, the median or mean magnesium intake
for each category was assigned to each corresponding RR.
When the median or mean intake per category was not provided,
we assigned the midpoint of upper and lower boundaries in each
category as the average intake. If the lower or upper boundary
for the lowest and highest category, respectively, was not
reported, we assumed that the boundary had the same amplitude
as the closest category.
To examine a potential nonlinear association between
magnesium intake and stroke risk, we performed a 2-stage,
random-effects, dose-response meta-analysis. This was done
by modeling magnesium intake by using restricted cubic
splines with 3 knots at fixed percentiles 10%, 50%, and 90% of
the distribution (12). First, a restricted cubic-spline model was
estimated by using generalized least-square regression and
taking into account the correlation within each set of published
RRs as described by Orsini et al (11). Second, we combined the
study-specific estimates by using the restricted maximum
likelihood method in a multivariate random-effects meta-
analysis (13). A P value for nonlinearity was calculated by
testing the null hypothesis that the coefficient of the second
spline was equal to zero. We showed no significant depar-
ture from a linear-response model (P-spline transformation =
I2statistics (14). Three cutoffs of these I2values were consid-
ered as follows: ,30% (no or marginal between-study hetero-
geneity), 30–75% (mild heterogeneity), and .75% (notable
heterogeneity). We performed analyses stratified by study lo-
cation and stroke subtypes. Publication bias was assessed with
Egger’s regression test (15). All statistical analyses were con-
ducted with Stata software (version 10.1; StataCorp). P , 0.05
was considered statistically significant.
Our literature search identified 7 independent prospective
3, 16–20). The studies were published between 1998 and 2011
and included a total of 6477 stroke cases and 241,378 partic-
ipants (Table 1). Four studies were conducted in the United
States, 2 studies were conducted in Europe, and one study was
conducted in Taiwan. The median magnesium intake in the
study populations ranged from 242 mg/d (in US men and
women) to 471 mg/d (in Finnish men). All studies provided risk
estimates that were adjusted for age, smoking, and BMI. Most
studies also controlled for physical activity (6 studies), diabetes
(6 studies), history of hypertension or measured blood pressure
(6 studies), and alcohol consumption (6 studies); fewer studies
adjusted for other nutrients (3 studies).
Magnesium intake and stroke
Estimated RRs of total stroke for an increase in intake of 100
mg Mg/d for each of the included studies and the combined
estimate are shown in Figure 2. The combined RR of total stroke
was 0.92 (95% CI: 0.88, 0.97) without heterogeneity among
studies (P = 0.66, I2= 0%). In a sensitivity analysis in which one
study at the time was omitted and the rest of studies were an-
alyzed, combined RRs ranged from 0.91 (95% CI: 0.85, 0.97) to
0.93 (95% CI: 0.88, 0.99). Hypertension and diabetes were
potential confounders of the association between magnesium
intake and stroke risk. The results persisted when we excluded
the study (17) that did not adjust for both hypertension and di-
abetes (RR: 0.92; 95% CI: 0.87, 0.98). We showed no evidence
of publication bias (Egger’s test: P = 0.54).
Combined RRs of stroke for an increment in intake of 100 mg
Mg/d were 0.91 (95% CI: 0.85, 0.96) in studies conducted in the
United States and 0.95 (95% CI: 0.88, 1.02) in the European
studies. Four studies provided results for stroke subtypes (3, 17,
18, 20) and 2 studies reported results for ischemic stroke only (2,
19). For these 6 studies, combined RRs for an increase in intake
of 100 mg Mg/d were 0.91 (95% CI: 0.87, 0.96) for ischemic
stroke, 0.96 (95% CI: 0.84, 1.10) for intracerebral hemorrhage,
and 1.01 (95% CI: 0.90, 1.14) for subarachnoid hemorrhage.
This meta-analysis of 7 prospective studies showed a signifi-
cant inverse association between dietary magnesium intake and
risk of stroke. An increase in intake of 100 mg Mg/d was as-
sociated with an 8% and 9% decreased risk of total stroke and
ischemic stroke, respectively. Magnesium intake was not asso-
ciated with hemorrhagic strokes. However, the number of cases
was lower for hemorrhagic stroke than for ischemic and total
strokes, which led to a lower statistical power to estimate the
association between magnesium intake and risk of hemorrhagic
Rich food sources of magnesium are green leafy vegetables,
nuts, beans, and whole grains. The US Recommended Dietary
Allowance of magnesium for women and men 31–70 y of age is
FIGURE 1. Flowchart of study selection by using PubMed (http://www.
ncbi.nlm.nih.gov/pubmed) and EMBASE (http://www.embase.com).
MAGNESIUM INTAKE AND STROKE
by guest on October 26, 2015
Prospective studies of magnesium intake and risk of stroke1
Study, cohort study
No. of cases
[RR (95% CI) for highest
compared with lowest category
of intake] Adjustments
Ascherio et al, 1998 (16),
Follow-up Study (United
328 total strokes
8 M, 40–75 y Median: 452 compared
with 243 mg/d
[0.92 (0.58, 1.46)]
Age, smoking, profession,
histories of hypertension and
history of MI, BMI, physical
activity, and intakes of alcohol,
dietary fiber, potassium, and
Age, smoking, menopausal status,
postmenopausal hormone use,
BMI, exercise, histories of
diabetes and high cholesterol,
aspirin use, multivitamin use,
vitamin E use, and intakes of
alcohol, omega-3 fatty acids,
Age, randomized treatment
assignment, smoking, BMI,
exercise, postmenopausal hormone
use, multivitamin use, histories
of diabetes, hypertension, and
history of MI, intake of alcohol,
and total energy
Age, supplementation group,
cigarettes smoked daily, BMI,
physical activity, systolic and
diastolic BP, serum total and
HDL cholesterol, histories of
diabetes and ischemic heart
disease, intake of alcohol, and
Age, sex, age · sex, smoking,
sex · smoking, area, central
obesity, BMI, diabetes, physical
activity, hypertension, use of
antihypertensive drugs, self-
reported heart disease,
apolipoprotein B, plasminogen,
and alcohol intake
Age, sex, center, education,
smoking, BMI, diabetes, systolic
blood pressure, use of
LDL and HDL cholesterol,
fibrinogen, von Willebrand factor,
and total energy intake
Age, education, smoking, BMI,
physical activity, history of
diabetes, history of hypertension,
aspirin use, family history
of myocardial infarction, and
intakes of alcohol, protein,
cholesterol, total fiber,
folate, and total energy
Iso et al, 1999 (17),
Nurses’ Health Study
690 total strokes, 386 ISs,
74 ICHs, and 129 SHs
14F, 34–59 y Median: 381 compared
with 211 mg/d
[0.80 (0.63, 1.01)]
Song et al, 2005 (3),
Women’s Health Study
368 total strokes (35,601)10 F, 39–89 y Median: 433 compared
with 255 mg/d
[0.90 (0.65, 1.26)]
Larsson et al, 2008 (18),
Carotene Cancer Prevention
2702 ISs, 383 ICHs,
and 196 SHs (26,556)
13.6 M, 50–69 y Median: 575 compared
with 382 mg/d
[IS: 0.85 (0.76, 0.97)]
Weng et al, 2008 (19),
risk FACtor Two-township
132 ISs (1772)10.6 M and F,
with ,242.6 mg/d
[0.68 ( 0.45, 1.04)]
Ohira et al, 2009 (2),
Atherosclerosis Risk in
577 ISs (13,277)15M and F,
with ,186 mg/d
[0.80 (0.75, 1.13)]
Larsson et al, 2011 (20),
1680 total strokes, 1310 ISs,
154 ICHs, and 79 SHs
10.4F, 49–83 yMedian: 373 compared
with 267 mg/d
[1.02 (0.82, 1.27)]
1BP, blood pressure; ICH, intracerebral hemorrhage; IS, ischemic stroke; MI, myocardial infarction; SH, subarachnoid hemorrhage.
LARSSON ET AL
by guest on October 26, 2015
320 and 420 mg/d, respectively (21). The approximate magne-
mg in 1 oz roasted cashew nuts, 134 mg in 1 cup canned white
brown rice (22).
There are several potential mechanisms by which magnesium
may reduce risk of ischemic stroke. The protective effect of
magnesium against risk of stroke may in part be related to its
blood pressure–lowering effect. Findings from a meta-analysis of
12 randomized clinical trials (median follow-up: 11 wk) showed
that magnesium supplementation (mean dose: ;400 mg/d) may
slightly decrease diastolic blood pressure by 2.2 mm Hg (5).
Most studies included in this meta-analysis controlled for his-
tory of hypertension or blood pressure levels. There may have
been overadjustment if the association between magnesium in-
take and stroke was mediated via a reduction in blood pressure.
Magnesium intake has also been inversely associated with risk
of type 2 diabetes (2, 4), which is a risk factor for ischemic
stroke (23, 24). Therefore, it is possible that type 2 diabetes is
a mediator of the association between magnesium intake and
ischemic stroke. Moreover, a high-magnesium diet has been
shown to have favorable effects on plasma glucose and blood
lipid concentrations in rats with chronic diabetes (7). In addition,
there are reports that showed that magnesium deficiency in-
creased the susceptibility of lipoproteins to peroxidation in an-
assessment was based on data from prospective studies, which
eliminated recall and selection biases. Furthermore, by com-
bining results from several studies, we had high statistical power
to estimate the relation between magnesium intake and risk of
stroke. Also, all studies were very consistent with no heteroge-
neity among study results. This meta-analysis also had several
potential limitations. First, as a meta-analysis of observational
studies, the possibility of confounding as a potential explanation
for the observed inverse association between magnesium intake
and stroke risk could not be excluded. Although the majority of
studies adjusted for known risk factors for stroke, such as age,
cigarette smoking, BMI, physical activity, histories of diabetes
and hypertension, and alcohol consumption, the possibility for
residual confounding remained. Moreover, it could not be ex-
cluded that other nutrients or dietary components that are cor-
related with dietary magnesium may have been responsible,
partly or in whole, for the observed association. A second lim-
itation was the misclassification of magnesium intake, which
would most likely have led to an underestimation of the true
analysis of published studies, publication bias may be a problem.
In conclusion, results from this meta-analysis indicate that
dietary magnesium intake is inversely associated with risk of
stroke. Although it is premature to recommend magnesium
supplementation to reduce risk of stroke, increased consumption
of magnesium-rich foods such as green leafy vegetables, beans,
nuts, and whole-grain cereals appears to be prudent.
The authors’ responsibilities were as follows—SCL: provided study over-
sight and wrote and took primary responsibility for the final content of the
manuscript; SCL and NO: undertook data collection and performed statistical
analyses; and all authors: designed the research, assisted in the interpretation
of analyses and revision of the manuscript, and read and approved the final
manuscript. None of the authors had a conflict of interest.
1. Ma J, Folsom AR, Melnick SL, Eckfeldt JH, Sharrett AR, Nabulsi AA,
Hutchinson RG, Metcalf PA. Associations of serum and dietary mag-
nesium with cardiovascular disease, hypertension, diabetes, insulin,
and carotid arterial wall thickness: the ARIC study. Atherosclerosis
Risk in Communities Study. J Clin Epidemiol 1995;48:927–40.
2. Ohira T, Peacock JM, Iso H, Chambless LE, Rosamond WD, Folsom
AR. Serum and dietary magnesium and risk of ischemic stroke: the
Atherosclerosis Risk in Communities Study. Am J Epidemiol 2009;
3. Song Y, Manson JE, Cook NR, Albert CM, Buring JE, Liu S. Dietary
magnesium intake and risk of cardiovascular disease among women.
Am J Cardiol 2005;96:1135–41.
4. Larsson SC, Wolk A. Magnesium intake and risk of type 2 diabetes:
a meta-analysis. J Intern Med 2007;262:208–14.
5. Dickinson HO, Nicolson DJ, Campbell F, Cook JV, Beyer FR, Ford
GA, Mason J. Magnesium supplementation for the management of
essential hypertension in adults. Cochrane Database Syst Rev 2006;3:
6. Chacko SA, Sul J, Song Y, Li X, LeBlanc J, You Y, Butch A, Liu S.
Magnesium supplementation, metabolic and inflammatory markers,
and global genomic and proteomic profiling: a randomized, double-
blind, controlled, crossover trial in overweight individuals. Am J Clin
7. Soltani N, Keshavarz M, Dehpour AR. Effect of oral magnesium sul-
fate administration on blood pressure and lipid profile in streptozocin
diabetic rat. Eur J Pharmacol 2007;560:201–5.
8. Rayssiguier Y, Gueux E, Bussiere L, Durlach J, Mazur A. Dietary
magnesium affects susceptibility of lipoproteins and tissues to perox-
idation in rats. J Am Coll Nutr 1993;12:133–7.
9. Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D,
Moher D, Becker BJ, Sipe TA, Thacker SB. Meta-analysis of obser-
vational studies in epidemiology: a proposal for reporting. Meta-
analysis Of Observational Studies in Epidemiology (MOOSE) group.
10. Greenland S, Longnecker MP. Methods for trend estimation from
summarized dose-response data, with applications to meta-analysis.
Am J Epidemiol 1992;135:1301–9.
11. Orsini N, Bellocco R, Greenland S. Generalized least squares for trend
estimation of summarized dose-response data. Stata Journal 2006;6:
12. Harrell FE Jr, Lee KL, Pollock BG. Regression models in clinical
studies: determining relationships between predictors and response. J
Natl Cancer Inst 1988;80:1198–202.
FIGURE 2. RRs of total stroke for an increment in intake of 100 mg Mg/
d. Squares represent study-specific RR estimates [the size of the square
reflects the study-specific statistical weight (ie, the inverse of the
variance)], horizontal lines represent 95% CIs, and the diamond represents
the combined RR estimate with its 95% CI.
MAGNESIUM INTAKE AND STROKE
by guest on October 26, 2015
13. Jackson D, White IR, Thompson SG. Extending DerSimonian and
Laird’s methodology to perform multivariate random effects meta-
analyses. Stat Med 2010;29:1282–97.
14. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-
analysis. Stat Med 2002;21:1539–58.
15. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-
analysis detected by a simple, graphical test. BMJ 1997;315:629–34.
16. Ascherio A, Rimm EB, Hernan MA, Giovannucci EL, Kawachi I,
Stampfer MJ, Willett WC. Intake of potassium, magnesium, calcium,
and fiber and risk of stroke among US men. Circulation 1998;98:1198–
17. Iso H, Stampfer MJ, Manson JE, Rexrode K, Hennekens CH, Colditz
GA, Speizer FE, Willett WC. Prospective study of calcium, potassium,
and magnesium intake and risk of stroke in women. Stroke 1999;30:
18. Larsson SC, Virtanen MJ, Mars M, Ma ¨nnisto ¨ S, Pietinen P, Albanes D,
Virtamo J. Magnesium, calcium, potassium, and sodium intakes and
risk of stroke in male smokers. Arch Intern Med 2008;168:459–65.
19. Weng LC, Yeh WT, Bai CH, Chen HJ, Chuang SY, Chang HY, Lin BF,
Chen KJ, Pan WH. Is ischemic stroke risk related to folate status or
other nutrients correlated with folate intake? Stroke 2008;39:3152–8.
20. Larsson SC, Virtamo J, Wolk A. Potassium, calcium, and magnesium
intake and risk of stroke in women. Am J Epidemiol 2011;174:35–43.
21. Dietary Reference Intakes for calcium, phosphorus, magnesium, vita-
min D, and fluoride. Washington, DC: National Academy of Sciences.
Institute of Medicine. Food and Nutrition Board, 1997.
22. US Department of Agriculture, Agricultural Research Service. USDA
national nutrient database for standard reference, release 24, 2011.
Available from: http://www.ars.usda.gov/ba/bhnrc/ndl.
23. Andersen KK, Olsen TS, Dehlendorff C, Kammersgaard LP. Hemor-
rhagic and ischemic strokes compared: stroke severity, mortality, and
risk factors. Stroke 2009;40:2068–72.
24. Janghorbani M, Hu FB, Willett WC, Li TY, Manson JE, Logroscino G,
Rexrode KM. Prospective study of type 1 and type 2 diabetes and risk
of stroke subtypes: the Nurses’ Health Study. Diabetes Care 2007;30:
LARSSON ET AL
by guest on October 26, 2015