Abstract
Aims
Numerous studies have shown inverse associations between serum magnesium (Mg) and risk of cardiovascular disease (CVD), but studies of dietary Mg have not been consistent. To examine the association of a Mg-rich diet score with risks of CVD, coronary heart disease (CHD), and ischaemic stroke in the Atherosclerosis Risk in Communities (ARIC) study.
Methods and results
There were 15 022 Black and White adults without prevalent CVD at baseline (1987–89) included in this analysis. Diet was assessed at two visits 6 years apart using an interviewer-administered 66-item food frequency questionnaire. A Mg-rich diet score was created that included servings of whole grain products, nuts, vegetables, fruit, legumes, coffee, and tea. Cox proportional hazard regression evaluated associations of incident CVD, CHD, and stroke across quintiles of Mg-rich diet score, adjusting for demographics, lifestyle factors, and clinical characteristics. Over >30 years of follow-up, there were 3531 incident CVD events (2562 CHD, 1332 ischaemic stroke). Participants who consumed more Mg-rich foods were older, female, White, had lower blood pressure, fewer were not current smokers, and more reported being physically active. A Mg-rich diet was inversely associated with incident CVD (HRQ5 vs. Q1 = 0.87, 95% CI: 0.77–0.98, Ptrend = 0.02) and CHD (HRQ5 vs. Q1 = 0.82, 95% CI: 0.71–0.95, Ptrend = 0.01); however, the diet-stroke association was null (HRQ5 vs. Q1 = 1.00, 95% CI: 0.82–1.22, Ptrend = 0.97).
Conclusion
Consuming a diet including Mg-rich foods, such as whole grains, nuts, vegetables, fruits, legumes, coffee, and tea, is associated with lower risk of CVD and CHD, but not ischaemic stroke.
Keywords: Magnesium, Dietary intake, Cardiovascular disease, Coronary heart disease, Cohort study
Graphical Abstract
Graphical Abstract.
Introduction
Numerous studies have shown that low serum magnesium (Mg) concentrations are associated with an elevated risk of cardiovascular disease (CVD) and adverse levels of CVD risk factors.1–3 In healthy individuals, serum Mg is regulated by the kidneys, intestines, and bone with normal serum Mg concentrations ranging between 0.7 and 1.1 mM.4 In addition, serum Mg may be influenced by aging, medication use, and dietary intake.5 Because serum Mg is regulated, ∼40–60% of dietary Mg is usually absorbed in the intestines; but when circulating Mg is low, as much as 80% can be absorbed.6
Although the Recommended Daily Allowance (RDA) for women is 310–320 mg and 400–420 mg for men,7 women and men enrolled in the 2016–17 National Health and Nutrition Examination Survey consumed an average dietary Mg intake of only 271 and 343 mg per day,8 respectively. As many as 75% of adults in the USA do not meet the RDA for dietary Mg intake.9 Past studies of dietary Mg, including a 2016 meta-analysis, have been inconsistent, showing some associations with cardiovascular risk factors, but not incident CVD or coronary heart disease (CHD).10,11
Excellent food sources of Mg include whole grains, nuts, vegetables, fruits, legumes, coffee, and tea.7 Dietary intake of whole grains, nuts, and legumes has generally shown beneficial associations with CVD risk, while associations have been inconsistent for vegetables, fruits, coffee, and tea. Specifically, studies have shown whole grain intake to be inversely associated with CVD, CHD, and type 2 diabetes.12 Intake of more nuts as well as nuts in the context of a healthy diet pattern, such as the Mediterranean diet, is associated with lower risk of CVD and adverse cardiovascular events.13 Legumes consistently show an inverse association with CVD and CHD, while associations with stroke are less clear.14 Some studies report a beneficial association of fruit and vegetable consumption with CVD and CVD risk factors,15 while others do not.8 Finally, inconsistent associations have been reported for the associations between coffee and tea relative to CVD, CHD, and stroke.16
Diet patterns, as a more holistic assessment of dietary intake, have often shown consistent inverse associations with chronic disease outcomes, possibly due to the synergic effect resulting from the combinations of individual foods and beverages.17 To date, no studies have examined the association of a diet pattern rich in Mg with risk of CVD. Therefore, using data from the Atherosclerosis Risk in Communities (ARIC) study, we tested the hypothesis that a novel Mg-rich diet score would be inversely associated with incident CVD, including CHD and ischaemic stroke.
Methods
All participants signed informed consents, and the ARIC protocol was approved by all participating Institutional Review Boards.
Study design and population
The ARIC study is a community-based prospective cohort designed to examine CVD aetiology and outcomes in middle-aged and older adults. Recruitment occurred from 1987–89 enrolling 15 792 adults aged 45–64 years living in Forsyth County, NC, Jackson, MS, selected suburbs of Minneapolis, MN, and Washington County, MD.18
For this analysis, we excluded participants with prevalent CVD at the time their food frequency questionnaire (FFQ) was administered (n = 238), self-identified as neither White nor Black (n = 48), and Black individuals from Maryland and Minnesota field centres (n = 55). Participants who had missing diet data at baseline (n = 364) were also excluded, as were participants with implausible energy intake, defined as <500 kcal or >3500 kcal for women and <700 kcal and >4500 kcal for men (n = 442). The final sample in this analysis included 15 022 participants.
Diet assessment
Dietary intake was assessed at visits 1 (baseline, 1987–89) and 3 (1993–95), via a 66-item FFQ, which was a modified version of the 61-item instrument validated by Willett et al.19 This FFQ was administered by trained interviewers according to a study protocol to obtain food and beverage intake and their frequency from nine categories ranging from less than once per month to greater than six times per day. Daily servings of major food groups were created for whole grain products, nuts, vegetables, fruits, legumes, coffee, tea, refined grain products, red and processed meat, poultry, fish, eggs, dairy, diet beverages, and sugar-sweetened beverages. In this analysis, there were 15 022 participants with baseline diet data; including 12 345 participants with both baseline and visit 3 diet data. The average of nutrient and food group data for participants with both baseline and visit 3 data was used in the analysis; for participants missing visit 3 data (n = 2677), only baseline diet data were used.
Two de novo food scores were created: (1) a Mg-rich diet score, and (2) a score representing the rest of dietary intake. The Mg-rich diet score was created including the following foods: whole grains, nuts, vegetables, fruits, legumes, coffee, and tea. An individual was assigned a sum of scores 0–4 that corresponded to the quintile of intake for each of these Mg-rich foods (quintile 1 = 0, quintile 2 = 1, quintile 3 = 2, quintile 4 = 3, and quintile 5 = 4). Using principal components analysis (PCA), another food score was derived that included the remaining foods and beverages (refined grain products, red and processed meat, poultry, fish, eggs, dairy, diet beverages, and sugar-sweetened beverages). In addition, dietary Mg (mg/day) was calculated by multiplying the frequency of consumption by the amount of Mg in each food, then summed for all items in the FFQ. Spearman correlation between dietary Mg and Mg-rich diet score was r = 0.65 (P < 0.001).
Ascertainment of cardiovascular events
Recurring phone calls (annual before 2012; twice-yearly thereafter), community-wide hospital surveillance, and links to local and national death-certificate registries were used to identify hospitalizations and CVD events.18 Events evaluated herein include incident CVD, CHD, and ischaemic stroke. Hospital records for potential events were obtained and abstracted, then were adjudicated by physicians. Incident CVD (based on composite of CHD or ischaemic stroke) was defined as development of first CVD diagnosis, occurring after ARIC visit 3 and through 2019.18 Incident CHD was defined consistently with previously published parameters as first definite or probable myocardial infarction, silent myocardial infarction, fatal CHD, or coronary revascularization.18 Incident ischaemic stroke was defined following previous literature as the first definite or probable cardioembolic or thrombotic brain infarction.18
Other measurements
At baseline (visit 1), demographic measurements and other characteristics (age, sex, race, field centre, education) were obtained from an administered questionnaire, as were lifestyle characteristics (current smoking status, current drinking alcohol status) and current health information (medication use, diabetes). Physical activity (sports index) was assessed by the Baecke Physical Activity questionnaire.20 Height and weight were measured to the nearest cm and pound, respectively. Body mass index (BMI) was calculated as weight in kg divided by height in metres squared. Technicians measured participants’ systolic and diastolic blood pressures on the right arm three times, with the last two measurements averaged and included in the analysis.18 Total and high-density lipoprotein (HDL) cholesterol levels were measured from fasting blood samples drawn at the baseline clinic examination. Low-density lipoprotein (LDL) cholesterol was calculated using the Friedewald formula. Diabetes was defined as self-reported diagnosis of diabetes by a physician, use of diabetes medication, a fasting glucose ≥ 126 mg/dL, or non-fasting glucose ≥ 200 mg/dL.21 High blood pressure was defined as the use of blood pressure-lowering medication, systolic blood pressure ≥ 140 mmHg, or diastolic blood pressure ≥ 90 mmHg.21
Statistical analysis
Data were analysed using SAS software, version 9.4 (SAS Institute; Cary, NC). Multivariable regression analysis was used to assess the associations of baseline demographic, behavioural, and physiologic characteristics stratified across quintiles of Mg-rich diet score and reported as unadjusted means (standard deviation) and means (standard error) adjusted for age, sex, race, field centre, education, and energy intake. In addition, associations of averaged (visits 1 and 3) nutrient and food group intakes were stratified across quintiles of Mg-rich diet score. The Ptrend values were calculated using quintile numbers.
Cox proportional hazard regression evaluated the associations of Mg-rich diet score with incident CVD, and, separately, for incident CHD and incident ischaemic stroke. Several models were developed: model 1 was adjusted for age, sex, race, field centre, education, and energy intake; model 2 was adjusted for factors in model 1 plus lifestyle factors (current smoking status, current drinking alcohol status, physical activity), and PCA derived diet score representing the rest of the diet; and model 3 was adjusted for model 2 plus medications and comorbidities, i.e. anti-hypertensive medications and lipid-lowering medications, diabetes, BMI, high blood pressure, and high cholesterol. The proportional hazards assumption was assessed using Kaplan–Meier curves and a Log-Rank test. It was found not violated for ischaemic stroke and CHD, whereas CVD started to have slight violations at the end of follow-up; however, we do not expect this to considerably affect the results. The primary outcomes were examined for interaction with time for the proportional hazards analyses as a precautionary step. Multiplicative interactions by sex and by race were evaluated by including cross-product terms in the models (e.g. Mg-rich diet score quintiles ∗ sex), which were not significant (P for interaction > 0.10).
Results
Unadjusted and adjusted baseline characteristics by quintiles of Mg-rich diet score are shown in Table 1. Participants who consumed more servings of Mg-rich foods (quintile 5; Q5) were older, female, White, and more likely to have graduated from high school or a higher level of education than those who consumed less (quintile 1; Q1) after adjustment for demographics and energy intake. Study participants who consumed more Mg-rich foods (Q5) engaged in more healthy lifestyle behaviours, compared to participants consuming fewer Mg-rich foods (Q1), including higher physical activity score and fewer reporting current smoking. Participants with higher Mg-rich diet scores had lower blood pressure and fewer were hypertensive, while more reported taking lipid-lowering medications. Body mass index, cholesterol levels, and hypertension medication use were relatively consistent across quintiles. Prevalent diabetes was higher among those consuming more Mg-rich foods.
Table 1.
Unadjusted mean (SD) and adjusted mean (SE) baseline characteristics stratified across quintiles of magnesium-rich diet score among ARIC participants, n = 15 022
| Characteristics | Quintiles of Mg-rich diet score unadjusted, mean (SD) | Quintiles of Mg-rich diet score adjusteda, mean (SE) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Q1 (n = 2984) | Q2 (n = 2959) | Q3 (n = 3248) | Q4 (n = 2986) | Q5 (n = 2845) | P trend | Q1 (n = 2984) | Q2 (n = 2959) | Q3 (n = 3248) | Q4 (n = 2986) | Q5 (n = 2845) | P trend | |
| Mg-rich diet score | 6.94 (1.93) | 11.06 (0.81) | 14.01 (0.81) | 16.92 (0.81) | 21.03 (0.88) | 7.10 (0.03) | 11.13 (0.02) | 14.01 (0.02) | 16.86 (0.02) | 20.86 (0.03) | <0.001 | |
| Demographics | ||||||||||||
| Age (y) | 53.5 (5.83) | 54.02 (5.65) | 54.15 (5.74) | 54.37 (5.74) | 54.62 (5.73) | <0.001 | 53.12 (0.11) | 53.85 (0.10) | 54.14 (0.10) | 54.54 (0.10) | 55.05 (0.11) | <0.001 |
| Female (%) | 55.9 (0.5) | 56.5 (0.5) | 55.6 (0.5) | 55.8 (0.5) | 55.0 (0.5) | 0.38 | 45.5 (0.01) | 52.1 (0.01) | 55.6 (0.01) | 59.8 (0.01) | 66.5 (0.01) | <0.001 |
| Black (%) | 38.7 (0.003) | 29.0 (0.003) | 24.5 (0.003) | 22.3 (0.003) | 17.2 (0.003) | <0.001 | 28.9 (0.003) | 26.8 (0.003) | 26.4 (0.003) | 25.7 (0.003) | 24.1 (0.003) | <0.001 |
| Education, >HS (%) | 35.6 (0.8) | 40.8 (0.8) | 45.4 (0.8) | 48.1 (0.8) | 51.3 (0.8) | <0.001 | 67.2 (0.01) | 73.0 (0.01) | 77.3 (0.01) | 79.8 (0.01) | 84.7 (0.01) | <0.001 |
| Lifestyle factors | ||||||||||||
| Sports index > 2 (%) | 48.5 (0.7) | 55.4 (0.8) | 58.9 (0.8) | 62.6 (0.8) | 68.0 (0.8) | <0.001 | 62.7 (0.9) | 69.4 (0.7) | 72.9 (0.8) | 74.1 (0.9) | 79.3 (0.9) | <0.001 |
| Current smoker (%) | 31.5 (0.5) | 27.3 (0.4) | 25.1 (0.4) | 23.5 (0.4) | 21.7 (0.4) | <0.001 | 32.2 (0.8) | 28.1 (0.8) | 25.3 (0.8) | 23.1 (0.8) | 20.1 (0.9) | <0.001 |
| Never alcohol drinking (%) | 26.7 (0.5) | 25.8 (0.5) | 25.5 (0.5) | 24.5 (0.5) | 23.8 (0.5) | 0.006 | 23.7 (0.8) | 25.2 (0.7) | 26.2 (0.7) | 25.2 (0.8) | 24.5 (0.8) | <0.001 |
| Physiologic characteristics | ||||||||||||
| BMI (kg/m2) | 28.17 (5.82) | 27.86 (5.31) | 27.63 (5.31) | 27.40 (5.20) | 27.41 (5.15) | <0.001 | 27.7 (0.1) | 27.7 (0.1) | 27.7 (0.1) | 27.5 (0.1) | 27.7 (0.1) | 0.96 |
| Systolic BP (mmHg) | 123 (19.8) | 122 (18.6) | 121 (18.7) | 120 (18.4) | 119 (17.5) | <0.001 | 122 (0.3) | 121 (0.3) | 121 (0.3) | 121 (0.3) | 120 (0.4) | <0.001 |
| Diastolic BP (mmHg) | 75 (12.0) | 74 (11.1) | 74 (10.7) | 73 (10.8) | 72 (10.8) | <0.001 | 74 (0.2) | 74 (0.2) | 74 (0.2) | 74 (0.2) | 73 (0.2) | 0.01 |
| HDL cholesterol (mg/dL) | 53.57 (17.54) | 53.27 (18.00) | 53.32 (17.33) | 53.15 (17.88) | 53.19 (17.51) | 0.27 | 53.31 (0.32) | 53.16 (0.31) | 53.39 (0.29) | 53.26 (0.30) | 53.53 (0.33) | 0.62 |
| LDL cholesterol (mg/dL) | 136.06 (40.74) | 134.97 (41.60) | 134.50 (39.56) | 134.15 (39.50) | 134.45 (38.87) | 0.07 | 135.39 (0.78) | 134.57 (0.75) | 134.51 (0.71) | 134.48 (0.74) | 135.22 (0.80) | 0.87 |
| Triglycerides (mg/dL) | 128.19 (87.34) | 132.02 (93.60) | 130.04 (86.39) | 134.97 (90.68) | 132.91 (89.98) | 0.02 | 132.02 (1.75) | 133.06 (1.67) | 129.75 (1.57) | 133.53 (1.65) | 129.16 (1.78) | 0.38 |
| HBP medication (%) | 28.5 (0.5) | 25.3 (0.5) | 26.2 (0.5) | 24.3 (0.5) | 21.5 (0.4) | <0.001 | 30.1 (0.9) | 29.8 (0.8) | 31.6 (0.8) | 30.8 (0.8) | 28.6 (0.9) | 0.47 |
| Lipid-lowering medication (%) | 1.8 (0.1) | 2.7 (0.2) | 2.6 (0.2) | 3.3 (0.2) | 3.8 (0.2) | <0.001 | 1.3 (0.3) | 2.4 (0.3) | 2.5 (0.3) | 3.6 (0.3) | 4.6 (0.3) | <0.001 |
| Diabetes (%) | 10.1 (0.3) | 10.0 (0.3) | 9.2 (0.3) | 9.7 (0.3) | 9.2 (0.3) | 0.24 | 7.4 (0.6) | 9.1 (0.5) | 9.3 (0.5) | 10.6 (0.5) | 11.4 (0.6) | <0.001 |
| HBP (%) | 34.0 (0.5) | 30.2 (0.5) | 29.4 (0.5) | 27.7 (0.4) | 24.4 (0.4) | <0.001 | 29.9 (0.8) | 28.8 (0.8) | 29.4 (0.8) | 28.9 (0.8) | 27.3 (0.9) | 0.08 |
BMI, body mass index; BP, blood pressure; HDL, high-density lipoprotein; LDL, low-density lipoprotein; HBP, high blood pressure.
aAdjusted for age, sex, race, field centre, education, and energy intake.
Intakes of nutrients, Mg-rich foods, and Mg-poor foods by quintile of Mg-rich diet score are shown in Table 2. Compared to participants with the lowest Mg-rich diet score, participants with higher Mg-rich diet score consumed more calories, fibre, and Mg, and less saturated and total fat. As expected, participants with higher Mg-rich diet scores consumed more servings per day of whole grain products, nuts, dark leafy greens, other vegetables, fruits, legumes, coffee, and tea, dairy, and poultry. Intakes of diet beverages were similar across quintiles. Participants with a lower Mg-rich diet score consumed more refined grains, red and processed meat, eggs, and sugar-sweetened beverages.
Table 2.
Averaged mean (SE) daily dietary intake stratified across quintiles of magnesium-rich diet score among ARIC participants, n = 15 022
| Characteristics | Quintiles of Mg-rich diet score | |||||
|---|---|---|---|---|---|---|
| Q1 (n = 2984) | Q2 (n = 2959) | Q3 (n = 3248) | Q4 (n = 2986) | Q5 (n = 2845) | P trend | |
| Mg-rich diet score, mean (SE) | 7.10 (0.03) | 11.13 (0.02) | 14.01 (0.02) | 16.86 (0.02) | 20.86 (0.03) | <0.001 |
| Nutrient intake, daily | ||||||
| Energy, kcal | 1278 (9.84) | 1470 (9.73) | 1612 (9.28) | 1765 (9.70) | 2031 (10.02) | <0.001 |
| Fibre, g | 11.59 (0.09) | 14.41 (0.09) | 17.04 (0.08) | 19.81 (0.09) | 23.98 (0.10) | <0.001 |
| Total fat, g | 57.55 (0.27) | 58.73 (0.26) | 58.99 (0.25) | 59.22 (0.26) | 59.53 (0.28) | <0.001 |
| Saturated fat, g | 21.54 (0.11) | 21.63 (0.11) | 21.46 (0.10) | 21.20 (0.11) | 20.78 (0.12) | <0.001 |
| Magnesium, mg | 183.77 (1.53) | 220.81 (1.45) | 249.87 (1.35) | 280.35 (1.42) | 329.01 (1.53) | <0.001 |
| Food group intake, sv/d | ||||||
| Mg-rich foods and beverages | ||||||
| Whole grains | 0.63 (0.02) | 0.98 (0.02) | 1.24 (0.02) | 1.51 (0.02) | 2.02 (0.02) | <0.001 |
| Nuts | 0.19 (0.01) | 0.28 (0.01) | 0.34 (0.01) | 0.41 (0.01) | 0.56 (0.01) | <0.001 |
| Dark leafy greens | 0.08 (0.004) | 0.11 (0.004) | 0.15 (0.004) | 0.20 (0.004) | 0.25 (0.004) | <0.001 |
| Other vegetables | 1.30 (0.02) | 1.71 (0.02) | 2.15 (0.02) | 2.57 (0.02) | 3.26 (0.02) | <0.001 |
| Fruits | 1.24 (0.02) | 1.73 (0.02) | 2.12 (0.02) | 2.51 (0.02) | 3.06 (0.03) | <0.001 |
| Legumes | 0.16 (0.005) | 0.22 (0.005) | 0.29 (0.004) | 0.36 (0.005) | 0.52 (0.005) | <0.001 |
| Coffee | 1.15 (0.03) | 1.62 (0.03) | 1.75 (0.03) | 1.90 (0.03) | 2.26 (0.03) | <0.001 |
| Tea | 0.25 (0.02) | 0.40 (0.02) | 0.53 (0.02) | 0.69 (0.02) | 0.91 (0.02) | <0.001 |
| Other foods and beverages | ||||||
| Refined grains | 2.64 (0.02) | 2.50 (0.02) | 2.39 (0.02) | 2.33 (0.02) | 2.15 (0.03) | <0.001 |
| Red/processed meat | 1.03 (0.01) | 1.02 (0.01) | 0.97 (0.01) | 0.93 (0.01) | 0.89 (0.01) | <0.001 |
| Poultry | 0.31 (0.006) | 0.34 (0.005) | 0.38 (0.005) | 0.42 (0.005) | 0.47 (0.006) | <0.001 |
| Fish | 0.19 (0.005) | 0.23 (0.005) | 0.27 (0.005) | 0.31 (0.005) | 0.39 (0.005) | <0.001 |
| Eggs | 0.30 (0.007) | 0.30 (0.006) | 0.29 (0.006) | 0.28 (0.006) | 0.27 (0.007) | <0.001 |
| Dairy | 1.50 (0.02) | 1.54 (0.02) | 1.62 (0.02) | 1.68 (0.02) | 1.76 (0.02) | <0.001 |
| Diet beverages | 0.53 (0.02) | 0.51 (0.02) | 0.52 (0.02) | 0.53 (0.02) | 0.54 (0.02) | 0.38 |
| Sugar-sweetened beverages | 0.64 (0.01) | 0.47 (0.01) | 0.37 (0.01) | 0.29 (0.01) | 0.15 (0.01) | <0.001 |
Adjusted for age, sex, race, field centre, education, and energy intake.
Over >30 years of follow-up, there were 3531 incident CVD events (2562 CHD, 1332 ischaemic stroke). As shown in Table 3, participants who consumed more Mg-rich foods had a lower hazard ratio (HR) for incident CVD compared to those with a lower Mg-rich diet score (model 2; HRQ5 vs. Q1 = 0.87, Ptrend = 0.02). After adjusting for factors in the causal pathway, the association did not change (model 3; HRQ5 vs. Q1 = 0.88, Ptrend = 0.03). For risk of incident CHD, participants with a higher intake of Mg-rich foods had a lower risk of incident CHD (model 2; HRQ5 vs. Q1 = 0.82, Ptrend = 0.01). This finding was consistent and statistically significant for all three models. There was no significant association observed between Mg-rich diet score and incident ischaemic stroke. The same analysis was performed examining dietary Mg associated with CVD, CHD, and stroke in place of the Mg-rich diet score for comparison. No significant results were observed for quintiles of dietary Mg with CVD (Ptrend = 0.24), CHD (Ptrend = 0.73), and ischaemic stroke (Ptrend = 0.22) (Table 4).
Table 3.
Adjusted hazard ratios (95% CIs) of incident cardiovascular disease, coronary heart disease, and ischaemic stroke stratified across quintiles of magnesium-rich diet score, the ARIC study (n = 15 022)
| Quintiles of Mg-rich diet score, HR (95% CI) | ||||||
|---|---|---|---|---|---|---|
| Q1 (n = 2984) | Q2 (n = 2959) | Q3 (n = 3248) | Q4 (n = 2986) | Q5 (n = 2845) | P trend | |
| Mg-rich diet score, mean (SE) | 7.10 (0.03) | 11.13 (0.02) | 14.01 (0.02) | 16.86 (0.02) | 20.86 (0.03) | <0.001 |
| Incident CVD | ||||||
| Cases | 716 | 692 | 773 | 689 | 661 | |
| Model 1 | 1 (ref) | 0.93 (0.84–1.04) | 0.92 (0.83–1.02) | 0.88 (0.78–0.98) | 0.88 (0.78–0.99) | 0.01 |
| Model 2 | 1 (ref) | 0.92 (0.83–1.02) | 0.91 (0.81–1.01) | 0.87 (0.78–0.97) | 0.87 (0.77–0.98) | 0.02 |
| Model 3 | 1 (ref) | 0.91 (0.81–1.01) | 0.90 (0.81–1.00) | 0.86 (0.76–0.96) | 0.88 (0.78–0.99) | 0.03 |
| Incident CHD | ||||||
| Cases | 523 | 501 | 560 | 508 | 470 | |
| Model 1 | 1 (ref) | 0.92 (0.81–1.04) | 0.90 (0.80–1.02) | 0.88 (0.77–1.00) | 0.84 (0.73–0.97) | 0.01 |
| Model 2 | 1 (ref) | 0.90 (0.80–1.02) | 0.88 (0.78–0.99) | 0.86 (0.76–0.98) | 0.82 (0.71–0.95) | 0.01 |
| Model 3 | 1 (ref) | 0.88 (0.78–1.00) | 0.88 (0.77–0.99) | 0.84 (0.74–0.96) | 0.83 (0.72–0.96) | 0.01 |
| Incident ischaemic stroke | ||||||
| Cases | 266 | 256 | 293 | 255 | 262 | |
| Model 1 | 1 (ref) | 0.93 (0.78–1.11) | 0.97 (0.82–1.15) | 0.89 (0.75–1.07) | 0.98 (0.81–1.19) | 0.76 |
| Model 2 | 1 (ref) | 0.94 (0.79–1.12) | 0.97 (0.81–1.15) | 0.91 (0.75–1.09) | 1.00 (0.82–1.22) | 0.97 |
| Model 3 | 1 (ref) | 0.93 (0.78–1.11) | 0.96 (0.80–1.14) | 0.91 (0.75–1.09) | 1.02 (0.84–1.25) | 0.86 |
Model 1 adjusted for age, sex, race, field centre, education, and energy intake. Model 2 adjusted for model 1 + lifestyle factors (smoking, alcohol, physical activity) and PCA food score representing the remaining dietary intake. Model 3 adjusted for model 2 + medications (anti-hypertensives, lipid-lowering medications), diabetes, BMI, hypertension, and high cholesterol.
Mg-rich, magnesium-rich; ARIC, Atherosclerosis Risk in Communities study; CVD, cardiovascular disease; CHD, coronary heart disease.
Table 4.
Hazard ratios (95% CIs) of incident cardiovascular disease, coronary heart disease, and ischaemic stroke stratified across quintiles of dietary magnesium, the ARIC study (n = 13 689)
| Quintiles of dietary magnesium, HR (95% CI) | ||||||
|---|---|---|---|---|---|---|
| Q1 (n = 2737) | Q2 (n = 2738) | Q3 (n = 2738) | Q4 (n = 2738) | Q5 (n = 2738) | P trend | |
| Dietary Mg, mean (SD) | 139.42 (24.70) | 196.21 (12.81) | 239.55 (12.60) | 290.71 (17.18) | 400.08 (78.73) | |
| Incident CVD | ||||||
| Cases | 579 | 514 | 559 | 589 | 584 | |
| 1 (ref) | 0.86 (0.76–0.97) | 0.95 (0.84–1.07) | 0.99 (0.88–1.12) | 1.01 (0.89–1.15) | 0.24 | |
| Incident CHD | ||||||
| Cases | 404 | 373 | 393 | 424 | 401 | |
| 1 (ref) | 0.88 (0.76–1.02) | 0.95 (0.82–1.01) | 1.00 (0.86–1.15) | 0.97 (0.83–1.13) | 0.73 | |
| Incident ischaemic stroke | ||||||
| Cases | 219 | 196 | 222 | 226 | 228 | |
| 1 (ref) | 0.90 (0.74–1.09) | 1.01 (0.84–1.23) | 1.04 (0.85–1.26) | 1.08 (0.88–1.33) | 0.22 | |
Adjusted for age, sex, race, field centre, education, energy intake, smoking, alcohol, and physical activity.
Mg-rich, magnesium-rich; ARIC, Atherosclerosis Risk in Communities study; CVD, cardiovascular disease; CHD, coronary heart disease.
Discussion
Consuming a Mg-rich diet was inversely associated with risk of CVD and CHD over 30 years of follow-up, after adjusting for confounding factors. However, there was no association observed across quintiles of Mg-rich diet score with stroke risk. Notably, dietary Mg intake, based on summing Mg in foods consumed, was not related to CVD, which is consistent with previous studies.10,11 Participants in the highest quintile of Mg-rich diet score consumed more fibre and calories, less total and saturated fat, and fewer refined grain products, red and processed meat, and sugar-sweetened beverages. These findings of the Mg-rich diet score and dietary Mg intake support the concept of food synergy in dietary patterns for promoting cardiovascular health.
Mg plays a variety of roles in the body that may be related to CVD risk and cardiovascular health, including enzyme activation, calcium regulation, and energy production.6 Traditional pathways may explain the association between Mg and CVD including high blood pressure, high blood sugar, inflammation, and/or reduced blood circulation.22,23 Mg homeostasis is complex, and dietary Mg has generally shown inconclusive relationships with CVD in observational studies, with poor correlations between dietary and serum Mg.10,24 Higher coronary artery disease (CAD) risk has been found to be associated with lower serum Mg.1 Supplementation with Mg increased serum Mg and decreased blood pressure in a meta-analysis of randomized control trials.25 Increased Mg intake through supplementation or diet could potentially be a tool for improving heart health and decreasing CVD risk.25,26
Potentially novel pathways between Mg and heart health include platelet formation, endothelial function, normal sinus rhythm maintenance, and reducing oxidative stress.4 Experimental studies in cardiac tissue have shown a relationship between hypomagnesaemia and oxidative DNA damage.27 Cardiac tissue retrieved during autopsies of males with fatal CAD events was found to have less Mg in the muscle compared to males with non-cardiac causes of death.28
Results from a 2016 meta-analysis showed inverse associations of dietary Mg intake with incident stroke and diabetes, but not with CHD or CVD risk.10 However, in the current study and previously published studies in ARIC, dietary Mg intake was not associated with incident CVD, CHD, or stroke.3,29,30 Interestingly, many studies, including the ARIC study, reported significant and inverse associations between dietary intake of Mg and CVD risk factors.11 Inconsistent results may be explained by differences in study design methodology, including instruments used to assess dietary intake, frequency of diet assessment, and study population (i.e. different age, race, and socioeconomic groups). Importantly, consuming foods rich in Mg may provide synergy among the nutrients and food compounds included in these foods such as synergy of Mg with fibre, antioxidants, and phytochemicals.17
Dietary patterns may provide greater benefit for prevention of CVD and its risk factors than individual nutrients, due to the synergistic effects of nutrients and/or foods within the diet pattern.17,31 Plant-based diet patterns, composed of vegetable fats, antioxidants, phytonutrients, and dietary fibre, include the current Mg-rich diet score, Healthy Eating Index (HEI), Dietary Approach to Stop Hypertension (DASH), Mediterranean, Nordic, Portfolio, and vegan/vegetarian diet patterns.32,33 These diet patterns typically alleviate inflammation and oxidative stress and lower CVD risk factors34,35 compared to more animal-based diets, such as the Western dietary pattern.31 A Western diet is associated with more proinflammatory biomarkers and an abnormal lipid panel, thus promoting greater risk for CVD, CHD, and stroke.31 Thus, eating foods and beverages rich in Mg contributes a myriad of benefits that are inversely associated with CVD, further supporting the idea of food synergy.17
Strengths and limitations
Our study had several strengths and limitations to report. One limitation is the use of self-reported dietary intake as it is susceptible to misclassification and recall bias. However trained and certified interviewers administered the validated 66-item FFQ,19 used food models to obtain portion size of food and beverage consumed and queried for frequency of intake. In addition, two measures of dietary intake, instead of one, were obtained that would improve the precision of dietary intake. Further, misclassification of dietary intake would likely bias our risk estimates towards the null and make it less likely that we would observe significant exposure–outcome associations. While ARIC assessed dietary intake in mid-life and not late-life, trends in dietary intake of older adults enrolled in national surveys showed little change over 30 years.36 Moreover, numerous studies have reported mid-life dietary intake an important predictor of health outcomes in later life.37,38 Although we adjusted our models for known and potential confounders, residual confounding is always a possibility. Participants who consumed a Mg-rich diet also had other healthy lifestyle habits, including greater physical activity and less current smoking, however we adjusted for these and other factors in our statistical models to address residual confounding to the best of our ability. Given that we were investigating Mg-rich foods rather than Mg alone, our analyses did not adjust for single nutrients or compounds, such as fibre. Nutrients and food compounds, such as fibre, phytochemicals, and antioxidants, within foods act synergistically providing more benefit than a single nutrient.17 However, evidence have shown benefits for each of fibre, phytochemicals, and antioxidant intake with CVD risk.39 One strength of this study is the large and diverse study population, including middle-aged to older adults in the USA who identified as either Black or White. Another strength is the large number of incident CVD, CHD, and ischaemic stroke events observed over 30 years of follow-up. Finally, the Mg-rich diet score is novel and reflects Mg intake in this population. This study also adds support for the idea that synergy between foods provides greater health benefits than individual foods or nutrients.17
Conclusion
We observed that a higher Mg-rich diet score, composed of whole grain products, nuts, vegetables, fruit, legumes, coffee, and tea, was associated with lower risk of incident CVD and CHD, but not incident ischaemic stroke. Mg-rich foods are also rich in other nutrients, including fibre, phytochemicals, and antioxidants, shown to provide CVD benefit. The beneficial associations of a Mg-rich diet observed in this study align with the 2020–25 Dietary Guidelines for Americans to ‘make every bite count’ by choosing nutrient-dense foods.40
Acknowledgements
The authors thank the staff and participants of the ARIC study for their important contributions.
Contributor Information
Katherine L Copp, University of Minnesota School of Public Health, Division of Epidemiology and Community Health, 1300 South Second St, Suite 300, Minneapolis, MN 55454, USA.
Lyn M Steffen, University of Minnesota School of Public Health, Division of Epidemiology and Community Health, 1300 South Second St, Suite 300, Minneapolis, MN 55454, USA.
So-Yun Yi, University of Minnesota School of Public Health, Division of Epidemiology and Community Health, 1300 South Second St, Suite 300, Minneapolis, MN 55454, USA.
Pamela L Lutsey, University of Minnesota School of Public Health, Division of Epidemiology and Community Health, 1300 South Second St, Suite 300, Minneapolis, MN 55454, USA.
Casey M Rebholz, Johns Hopkins University Bloomberg School of Public Health, Department of Epidemiology, Baltimore, MD 21287, USA.
Mary R Rooney, Johns Hopkins University Bloomberg School of Public Health, Department of Epidemiology, Baltimore, MD 21287, USA.
Author contribution
K.L.C. and L.M.S. conceived and designed the research and drafted the manuscript. K.L.C., S.-Y.Y., and L.M.S. contributed to the acquisition, analysis, or interpretation of the data. All authors critically revised the manuscript and gave approval of the final manuscript.
Funding
Supported by National Heart, Lung, and Blood Institute (NHLBI) of the National Institutes of Health (NIH) award T32HL007779 (to K.L.C.). The Atherosclerosis Risk in Communities study has been funded in part from the National Heart, Lung, and Blood Institute, NIH, Department of Health and Human Services, under contract nos. (75N92022D00001, 75N92022D00002, 75N92022D00003, 75N92022D00004, 75N92022D00005). C.M.R. was supported by grants from the National Institute of Diabetes and Digestive and Kidney Disease (NIDDK, R03 DK128386) and the National Heart, Lung, and Blood Institute (NHLBI, R01 HL153178). P.L.L. was supported by National Heart, Lung, and Blood Institute grant K24 HL159246.
Data availability
Data sharing can be conducted through the ARIC Coordinating Center: https://sites.cscc.unc.edu/aric/distribution-agreements.
References
- 1. Rooney MR, Alonso A, Folsom AR, Michos ED, Rebholz CM, Misialek JR, et al. Serum magnesium and the incidence of coronary artery disease over a median 27 years of follow-up in the Atherosclerosis Risk in Communities (ARIC) study and a meta-analysis. Am J Clin Nutr 2020;111:52–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Liao F, Folsom AR, Brancati FL. Is low magnesium concentration a risk factor for coronary heart disease? The atherosclerosis risk in communities (ARIC) study. Am Heart J 1998;136:480–490. [DOI] [PubMed] [Google Scholar]
- 3. 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;169:1437–1444. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. de Baaij JHF, Hoenderop JGJ, Bindels RJM. Magnesium in man: implications for health and disease. Physiol Rev 2015;95:1–46. [DOI] [PubMed] [Google Scholar]
- 5. Zimmerman M, Snow B. An introduction to nutrition: Independent; 2012. https://open.umn.edu/opentextbooks/textbooks/711. [Google Scholar]
- 6. Byrd-Bredbenner C, Berning JR, Kelley DS, Abbot J, Moe G, Beshgetoor D. Wardlaw's perspectives in nutrition. New York: McGraw-Hill Education; 2019. [Google Scholar]
- 7. Institute of Medicine (US) Standing Committee on the Scientific Evaluation of Dietary Reference Intakes . Dietary reference intakes for calcium, phosphorus, magnesium, vitamin D, and fluoride. Washington (DC): National Academies Press (US); 1997. [PubMed] [Google Scholar]
- 8. U.S. Department of Agriculture, Agricultural Research Service . 2020. Nutrient Intakes from Food and Beverages: Mean Amounts Consumed per Individual, by Gender and Age, What We Eat in America, NHANES 2017–2018.
- 9. Lutsey PL, Alonso A, Michos ED, Loehr LR, Astor BC, Coresh J, et al. Serum magnesium, phosphorus, and calcium are associated with risk of incident heart failure: the Atherosclerosis Risk in Communities (ARIC) study. Am J Clin Nutr 2014;100:756–764. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Fang X, Wang K, Han D, He X, Wei J, Zhao L, et al. Dietary magnesium intake and the risk of cardiovascular disease, type 2 diabetes, and all-cause mortality: a dose–response meta-analysis of prospective cohort studies. BMC Med 2016;14:210. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Rosique-Esteban N, Guasch-Ferré M, Hernández-Alonso P, Salas-Salvadó J. Dietary magnesium and cardiovascular disease: a review with emphasis in epidemiological studies. Nutrients 2018;10:168. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Steffen LM, Jacobs DR, Stevens J, Shahar E, Carithers T, Folsom AR. Associations of whole-grain, refined-grain, and fruit and vegetable consumption with risks of all-cause mortality and incident coronary artery disease and ischemic stroke: the Atherosclerosis Risk in Communities (ARIC) study. Am J Clin Nutr 2003;78:383–390. [DOI] [PubMed] [Google Scholar]
- 13. Liu X, Guasch-Ferré M, Drouin-Chartier JP, Tobias DK, Bhupathiraju SN, Rexrode KM, et al. Changes in nut consumption and subsequent cardiovascular disease risk among us men and women: 3 large prospective cohort studies. J Am Heart Assoc 2020;9:e013877. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Afshin A, Micha R, Khatibzadeh S, Mozaffarian D. Consumption of nuts and legumes and risk of incident ischemic heart disease, stroke, and diabetes: a systematic review and meta-analysis. Am J Clin Nutr 2014;100:278–288. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. He K, Hu FB, Colditz GA, Manson JE, Willett WC, Liu S. Changes in intake of fruits and vegetables in relation to risk of obesity and weight gain among middle-aged women. Int J Obes 2004;28:1569–1574. [DOI] [PubMed] [Google Scholar]
- 16. Ding M, Bhupathiraju SN, Satija A, Van Dam RM, Hu FB. Long-term coffee consumption and risk of cardiovascular disease: a systematic review and a dose–response meta-analysis of prospective cohort studies. Circulation 2014;129:643–659. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Jacobs DR, Steffen LM. Nutrients, foods, and dietary patterns as exposures in research: a framework for food synergy. Am J Clin Nutr 2003;78:508–513. [DOI] [PubMed] [Google Scholar]
- 18. Wright JD, Folsom AR, Coresh J, Sharrett AR, Couper D, Wagenknecht LE, et al. The ARIC (Atherosclerosis Risk In Communities) study. J Am Coll Cardiol 2021;77:2939–2959. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Willett WC, Sampson L, Stampfer MJ, Rosner B, Bain C, Witschi J, et al. Reproducibility and validity of a semiquantitative food frequency questionnaire. Am J Epidemiol 1985;122:51–65. [DOI] [PubMed] [Google Scholar]
- 20. Baecke JA, Burema J, Frijters JE. A short questionnaire for the measurement of habitual physical activity in epidemiological studies. Am J Clin Nutr 1982;36:936–942. [DOI] [PubMed] [Google Scholar]
- 21. Saeed A, Nambi V, Sun W, Virani SS, Taffet GE, Deswal A, et al. Short-term global cardiovascular disease risk prediction in older adults. J Am Coll Cardiol 2018;71:2527–2536. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Costello RB, Elin RJ, Rosanoff A, Wallace TC, Guerrero-Romero F, Hruby A, et al. Perspective: the case for an evidence-based reference interval for serum magnesium: the time has come. Adv Nutr 2016;7:977–993. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Rude RK. Magnesium. In: Coates PM, Betz JM, Blackman MR, Cragg GM, Levine M, Moss J, White JD, eds. Encyclopedia of dietary supplements. 2nd ed. New York, NY: Informa Healthcare; 2010. p527–537. [Google Scholar]
- 24. Misialek JR, Lopez FL, Lutsey PL, Huxley RR, Peacock JM, Chen LY, et al. Serum and dietary magnesium and incidence of atrial fibrillation in whites and in African Americans—Atherosclerosis Risk in Communities (ARIC) study. Circ J 2013;77:323–329. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Zhang X, Li Y, Del Gobbo LC, Rosanoff A, Wang J, Zhang W, et al. Effects of magnesium supplementation on blood pressure: a meta-analysis of randomized double-blind placebo-controlled trials. Hypertension 2016;68:324–333. [DOI] [PubMed] [Google Scholar]
- 26. Veronese N, Watutantrige-Fernando S, Luchini C, Solmi M, Sartore G, Sergi G, et al. Effect of magnesium supplementation on glucose metabolism in people with or at risk of diabetes: a systematic review and meta-analysis of double-blind randomized controlled trials. Eur J Clin Nutr 2016;70:1354–1359. [DOI] [PubMed] [Google Scholar]
- 27. Shah NC, Shah GJ, Li Z, Jiang XC, Altura BT, Altura BM. Short-term magnesium deficiency downregulates telomerase, upregulates neutral sphingomyelinase and induces oxidative DNA damage in cardiovascular tissues: relevance to atherogenesis, cardiovascular diseases and aging. Int J Clin Exp Med 2014;7:497–514. [PMC free article] [PubMed] [Google Scholar]
- 28. Johnson CJ, Peterson DR, Smith EK. Myocardial tissue concentrations of magnesium and potassium in men dying suddenly from ischemic heart disease. Am J Clin Nutr 1979;32:967–970. [DOI] [PubMed] [Google Scholar]
- 29. Peacock JM, Folsom AR, Arnett DK, Eckfeldt JH, Szklo M. Relationship of serum and dietary magnesium to incident hypertension: the Atherosclerosis Risk in Communities (ARIC) study. Ann Epidemiol 1999;9:159–165. [DOI] [PubMed] [Google Scholar]
- 30. Kao WHL, Folsom AR, Nieto FJ, Mo J-P, Watson RL, Brancati FL. Serum and dietary magnesium and the risk for type 2 diabetes mellitus. Arch Intern Med 1999;159:2151. [DOI] [PubMed] [Google Scholar]
- 31. Trautwein EA, McKay S. The role of specific components of a plant-based diet in management of dyslipidemia and the impact on cardiovascular risk. Nutrients 2020;12:2671. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Casas R, Castro-Barquero S, Estruch R, Sacanella E. Nutrition and cardiovascular health. Int J Mol Sci 2018;19:3988. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. O’Connor LE, Hu EA, Steffen LM, Selvin E, Rebholz CM. Adherence to a Mediterranean-style eating pattern and risk of diabetes in a U.S. prospective cohort study. Nutr Diabetes 2020;10:8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Mayhew AJ, De Souza RJ, Meyre D, Anand SS, Mente A. A systematic review and meta-analysis of nut consumption and incident risk of CVD and all-cause mortality. Br J Nutr 2016;115:212–225. [DOI] [PubMed] [Google Scholar]
- 35. Marventano S, Izquierdo Pulido M, Sánchez-González C, Godos J, Speciani A, Galvano F, et al. Legume consumption and CVD risk: a systematic review and meta-analysis. Public Health Nutr 2017;20:245–254 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Johnston R, Poti JM, Popkin BM. Eating and aging: trends in dietary intake among older Americans from 1977–2010. J Nutr Health Aging 2014;18:234–242. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Talegawkar SA, Jin Y, Xue QL, Tanaka T, Simonsick EM, Tucker KL, et al. Dietary pattern trajectories in middle age and physical function in older age. J Gerontol A Biol Sci Med Sci 2021;76:513–519. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Samieri C, Sun Q, Townsend MK, Chiuve SE, Okereke OI, Willett WC, et al. The association between dietary patterns at midlife and health in aging: an observational study. Ann Intern Med 2013;159:584–591. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Tucker KL. Dietary intake and coronary heart disease: a variety of nutrients and phytochemicals are important. Curr Treat Options Cardio Med 2004;6:291–302. [DOI] [PubMed] [Google Scholar]
- 40. U.S. Department of Agriculture and U.S . Department of Health and Human Services. Dietary guidelines for Americans, 2020–2025. 9th Ed. Dietary Guidelines for Americans website; 2020. Available at DietaryGuidelines.gov. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Data sharing can be conducted through the ARIC Coordinating Center: https://sites.cscc.unc.edu/aric/distribution-agreements.