Intake of Dietary Magnesium and the ... - Wiley Online Library

Human Physiology

Intake of Dietary Magnesium and the Prevalence of the Metabolic Syndrome among U.S. Adults Earl S. Ford,* Chaoyang Li,* Lisa C. McGuire,* Ali H. Mokdad,* and Simin Liu†

Abstract FORD, EARL S., CHAOYANG LI, LISA C. MCGUIRE, ALI H. MOKDAD, AND SIMIN LIU. Intake of dietary magnesium and the prevalence of the metabolic syndrome among U.S. adults. Obesity. 2007;15:1139 –1146. Objective: Limited data suggest that people with the metabolic syndrome have lower intakes or circulating concentrations of magnesium than those who do not have the syndrome. The aim of this study was to examine the associations between dietary intake of magnesium and the prevalence of the metabolic syndrome in a nationally representative sample of U.S. adults. Research Methods and Procedures: We used data for 7669 participants ⱖ20 years of age of the Third National Health and Nutrition Examination Survey (1988 to 1994). The metabolic syndrome was defined using the criteria of the National Cholesterol Education Program. Magnesium intake was determined from a single dietary 24-hour recall. Results: The unadjusted prevalences of the metabolic syndrome were 29.0% (quintile of lowest magnesium intake), 27.5%, 25.8%, 23.9%, and 21.8% for increasing quintiles of magnesium intake (p for trend ⫽ 0.002). After multiple adjustment, the odds ratios for the second through the fifth quintiles (highest intake) of magnesium intake among all participants included in the analysis were 0.84 [95% confidence interval (CI): 0.58, 1.23], 0.76 (95% CI: 0.54, 1.07), 0.62 (95% CI: 0.40, 0.98), and 0.56 (95% CI: 0.34, 0.92), respectively (p for trend ⫽ 0.029). The associations were similar for men and women.

Received for review March 1, 2006. Accepted in final form November 15, 2006. The costs of publication of this article were defrayed, in part, by the payment of page charges. This article must, therefore, be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. *Centers for Disease Control and Prevention, Atlanta, Georgia; and †Department of Epidemiology, University of California, Los Angeles, California. Address correspondence to Earl Ford, Centers for Disease Control and Prevention, 4770 Buford Highway, MS K66, Atlanta, GA 30341. E-mail: [email protected] Copyright © 2007 NAASO

Discussion: Our results showing an inverse association between dietary magnesium intake and the prevalence of the metabolic syndrome add to the evidence that adequate magnesium intake or a diet rich in magnesium may be important for maintaining good cardiometabolic health. Key words: epidemiology, health surveys, magnesium, nutrition, risk factors

Introduction The metabolic syndrome is highly prevalent in populations around the world (1,2). Components often included in definitions of this syndrome are obesity (generally abdominal obesity), high blood pressure, dyslipidemia (hypertriglyceridemia or low high-density lipoprotein cholesterol), and hyperglycemia. Regardless of how it is defined, people with the syndrome are at increased risk for cardiovascular disease and diabetes (3). Major determinants of the metabolic syndrome are obesity and sedentary behavior (4,5); the role of dietary constituents is less understood. Because components of the diet are important in many chronic diseases, including cardiovascular disease and diabetes, it is conceivable that nutrients contribute to the development of this syndrome. Many people with the metabolic syndrome have insulin resistance (6 –9). Research supports the notion that low intakes of magnesium may contribute to insulin resistance (10). Thus, inadequate intake of magnesium might also play a role in the pathogenesis of the metabolic syndrome. In a cross-sectional analysis of data from the Women’s Health Study, dietary intake of magnesium was inversely related to the prevalence of the metabolic syndrome (11). Subsequently, results from the prospective Coronary Artery Risk Development in Young Adults study showed an inverse association between intake of magnesium and incident metabolic syndrome (12). In addition, an inverse association between circulating concentrations of magnesium and the metabolic syndrome was found in another study (13). Whether low intake of magnesium is indeed associated with OBESITY Vol. 15 No. 5 May 2007

1139

Magnesium and Metabolic Syndrome, Ford et al.

an increased risk for the metabolic syndrome is important to determine in light of the syndrome’s increasing prevalence in the U.S. population (14) and the suboptimal intake of magnesium (15,16). Therefore, the aim of our study was to examine the cross-sectional associations between dietary intake of magnesium and the prevalence of the metabolic syndrome using a nationally representative sample of U.S. men and women.

Research Methods and Procedures The Third National Health and Nutrition Examination Survey was conducted from 1988 to 1994. The survey used a multistage, stratified sampling design to recruit men and women from the civilian, non-institutionalized population. After an interview in the home, participants were invited to attend the medical examination center for a series of examinations and tests, to complete additional questionnaires, and to provide blood and other biological specimens. Details of the survey and its methods have been published (17,18). Metabolic Syndrome In accordance with the National Cholesterol Education Program report, participants who had three or more of five criteria were considered to have the metabolic syndrome (19,20): abdominal obesity (waist circumference ⬎102 cm in men or ⬎88 cm in women); concentration of triglycerides ⱖ150 mg/dL (1.7 mM); concentration of high-density lipoprotein cholesterol ⬍40 mg/dL (1.03 mM) in men or ⬍50 mg/dL (1.29 mM) in women; a systolic blood pressure ⱖ130 mm Hg or a diastolic blood pressure ⱖ85 mm Hg); and fasting glucose ⱖ100 mg/dL (ⱖ5.6 mM). Participants who reported current use of anti-hypertensive or anti-diabetic medication (insulin or oral agents) were counted as having high blood pressure or diabetes, respectively. Waist circumferences to the nearest 0.1 cm were obtained at the level of the right iliac crest. Serum triglycerides were measured enzymatically after hydrolyzation to glycerol on a Hitachi 704 Analyzer (Boehringer Mannheim Diagnostics, Indianapolis, IN). High-density lipoprotein cholesterol was measured after the precipitation of other lipoproteins with a heparin-manganese chloride mixture on a Hitachi 704 Analyzer. Three blood pressure readings were obtained in the mobile examination center. The average of the second and third systolic blood pressure and diastolic blood pressure readings was used in the analyses. Serum glucose concentration was measured using an enzymatic reaction (Cobas Mira Chemistry System; Roche Diagnostic Systems, Mountclair, NJ). Dietary Assessment Trained interviewers administered a single 24-hour dietary recall to participants using a microcomputer-based 1140

OBESITY Vol. 15 No. 5 May 2007

interview system known as the NHANES III Dietary Data Collection that was developed by the University of Minnesota Nutrition Coordinating Center. Interviews were conducted during all days of the week. Participants reported all foods and beverages except drinking water that they had consumed during the previous 24 hours. Food and beverages were quantified using food-specific units, abstract food models, special charts, and other measurement aids. Interviewers probed for specific brand names. Intakes of nutrients were calculated using the United States Department of Agriculture’s Survey Nutrient Database, which was updated during the course of the study. Included in the analysis were the following dietary variables: magnesium, calcium, potassium, total energy intake, fat, carbohydrate, and fiber. Covariates Sociodemographic covariates included in the analyses were age, sex, race or ethnicity (white, African-American, Mexican American, other), and educational attainment (highest completed grade or year of education). BMI (kg/ m2) was calculated from measured height and weight during the anthropometric examination. Smoking status was categorized as never smoked, former smoker, and current smoker. Sedentary was defined as no leisure-time physical activity. Participants were asked how many times per month they drank beer or low-calorie beer, wine, wine coolers, sangria, champagne, or hard liquor. The responses to these questions were summed. Participants were also asked about their use of vitamins or minerals with the question, “Have you taken any vitamins or minerals during the past 24 hours?” In addition, we included an indicator variable (yes/ no) for any blood relatives who had had a heart attack before age 50. C-reactive protein was measured using latex-enhanced nephelometry on a Behring Nephelometer Analyzer System (Behring Diagnostics, Somerville, NJ) with NA Latex CRP Kit (Behring Diagnostics). The lower detection limit was 3.0 mg/L. Plasma insulin was measured by radioimmunoassay using the Pharmacia Insulin RIA Kit (Pharmacia Diagnostics, Uppsala, Sweden). Data Analysis We limited our analyses to participants ⱖ20 years who had fasted ⱖ8 hours. We excluded pregnant women. We tested for linear trends in the means or percentages of the covariates across sex-specific quintiles of intake of magnesium by using linear orthogonal polynomial contrasts. We used logistic regression analysis to examine whether dietary intake of magnesium was significantly associated with the metabolic syndrome after adjusting for age, sex, race or ethnicity, education, smoking status, concentration of C-reactive protein, use of alcohol, physical activity, family history of early coronary heart disease, use of vitamins or supplements, and, by sex-specific quintile, percent calories


as fat, percent calories as carbohydrate, fiber intake, and total energy intake. Additional adjustment for intakes of dietary calcium and potassium was done selectively. For the logistic regression models, we calculated p values for linear trend by using the median value for each quintile of magnesium intake. To account for the complex survey design, we used the statistical software SUDAAN (Research Triangle Institute, Research Triangle Park, NC) and medical examination clinic sampling weights to produce our weighted estimates and standard errors.

Results Of the 16,573 participants who attended the medical examination center, 10,116 remained after excluding participants who had fasted ⬍8 hours and pregnant women. After eliminating participants with incomplete data for all of the study variables, 7669 were included in the analyses. The unadjusted prevalence of the metabolic syndrome in the analytic sample was 25.6%. The mean and median intakes of magnesium were 310 and 276 mg/d, respectively. The dietary intake of magnesium was significantly and positively associated with the percentage of white participants, educational attainment, alcohol use, use of vitamin or mineral supplements, total energy intake, and fiber intake, and significantly and inversely associated with concentrations of insulin, elevated concentrations of C-reactive protein, and percentage of calories from carbohydrates (Table 1). In addition, participants with low intakes of magnesium had a higher BMI and were more sedentary than those with higher intakes. The unadjusted prevalence of the metabolic syndrome decreased as the intake of dietary magnesium increased (Table 2). For all five components of the metabolic syndrome as defined by National Cholesterol Education Program, inverse trends were seen with greater dietary intake of magnesium, but only the trends for abdominal obesity, low high-density lipoprotein cholesterol, and hyperglycemia had p ⬍ 0.05. In the logistic regression analyses, the odds ratios (ORs)1 tended to decrease progressively as the intake of magnesium increased. After adjustment for multiple factors, the OR for participants with the highest intakes of magnesium compared with those who consumed the least magnesium was 0.56 [95% confidence interval (CI) ⫽ 0.34, 0.92; Table 3]. After additional adjustment for intakes of dietary calcium and potassium, the ORs were 0.77 (95% CI: 0.51, 1.16) for quartile 2, 0.66 (95% CI: 0.45, 0.97) for quartile 3, 0.53 (95% CI: 0.31, 0.89) for quartile 4, and 0.48 (95% CI: 0.27, 0.86) for quartile 5. The multiple-adjusted ORs for quintiles 2 through 5 of magnesium intake were 0.92 (95% CI: 0.52, 1.64), 0.69 (95% CI: 0.40, 1.22), 0.59 (95% CI: 0.29, 1.19), and 0.50 (95% CI: 0.25, 0.99; p for trend ⫽ 1

Nonstandard abbreviations: OR, odds ratio; CI, confidence interval.

0.029) in men and 0.71 (95% CI: 0.43, 1.16), 0.76 (95% CI: 0.43, 1.35), 0.58 (95% CI: 0.33, 1.05), and 0.55 (95% CI: 0.29, 1.03; p for trend ⫽ 0.079) in women (p interaction for sex and magnesium intake ⫽ 0.621). Among participants who did not use vitamin or mineral supplements, the multiple-adjusted ORs were 0.84 (95% CI: 0.58, 1.22), 0.75 (95% CI: 0.53, 1.06), 0.58 (95% CI: 0.35, 0.98), and 0.47 (95% CI: 0.28, 0.80; p for trend ⫽ 0.008). Magnesium intake was not significantly associated with any of the components of the metabolic syndrome after adjustment for all of the covariates.

Discussion In this nationally representative sample of U.S. adults, dietary intake of magnesium was inversely associated with the prevalence of the metabolic syndrome. Our results are consistent with findings from the Women’s Health Study and the Coronary Artery Risk Development in Young Adults study (11,12). Because the intake of magnesium through dietary sources in the U.S. population is suboptimal, the evidence from the studies to date suggests a potential avenue for primary prevention that deserves more careful scrutiny through the conduct of more definitive studies. In the Women’s Health Study, the OR for participants in the highest quintile of magnesium intake was 0.73 (95% CI: 0.60, 0.88) (11). In comparison, our estimate among women was 0.55 (95% CI: 0.29, 1.03). When we restricted our analysis to women 45⫹ years of age, the OR was 0.33 (95% CI: 0.13, 0.83). In addition, our analysis suggested that the association between dietary intake of magnesium and the prevalence of the metabolic syndrome was similar among men and women. More recently, the results from an analysis of prospective data from the Coronary Artery Risk Development in Young Adults study were very consistent with the findings from the cross-sectional studies (12). During 15 years of follow-up, 608 of 4637 participants 18 to 30 years of age developed the metabolic syndrome. Compared with those whose intake of magnesium placed them in the lowest quartile, the adjusted hazard ratios were 0.91 (95% CI: 0.73, 1.13) for the second quartile, 0.67 (95% CI: 0.52, 0.85) for the third quartile, and 0.63 (95% CI: 0.47, 0.84) for the fourth quartile. Of the components of the metabolic syndrome, a large body of research offers the strongest support for a favorable effect of magnesium intake on blood pressure levels. Two meta-analyses, one of observational studies and one of clinical trials, have suggested that magnesium is inversely associated with blood pressure levels (21,22). A few reports found that obese study participants had lower intakes or circulating concentrations of magnesium than leaner participants (23–25). Other reports have suggested that magnesium intake might favorably affect lipid concentrations (26 – 30). Several supplementation studies also reported that OBESITY Vol. 15 No. 5 May 2007

1141

1142


1532 44.3 (0.8) 12.3 (0.1) 26.5 (0.2) 61.8 (2.0) 8.7 (0.6) 1886.8 (27.7) 34.3 (0.3) 48.9 (0.5) 12.4 (0.2) 50.9 (2.1) 77.4 (2.7) 30.2 (2.1) 14.5 (1.6) 24.0 (1.7)

17.9 (1.2) 18.7 (1.6)

7.0 (0.5) 1481.5 (15.7) 33.8 (0.6) 50.9 (0.6) 8.2 (0.1) 50.1 (2.0) 67.0 (2.4) 30.1 (1.7) 18.1 (1.7) 28.9 (1.9)

15.9 (1.3) 14.5 (1.8)

Quintile 2: men ⴝ 222 to 292 and women ⴝ 165 to 213

1835 43.6 (0.7) 11.6 (0.1) 26.8 (0.2) 65.7 (2.4)

Values represent means or percentages (standard error).

Number Age (yrs) Education (years) BMI (kg/m2) Insulin (pM) Alcohol intake (number of times per month) Total calories (kcal/d) Percent calories from fat (%) Percent calories from carbohydrates (%) Fiber (g/d) Male (%) White (%) Current smoker (%) No exercise (%) C-reactive protein ⬎3 mg/L (%) Use of vitamin or mineral supplements during past 24 hours (%) Family history of heart attack before age 50 years (%)

Quintile 1: men 337

0.138

⬍0.001

0.011 ⬍0.001 0.963 ⬍0.001 0.066 0.001 ⬍0.001

0.001 ⬍0.001 0.285

0.252 ⬍0.001 0.022 0.002

p for linear trend

Table 1. Unadjusted selected characteristics (means or percentages and standard error) by quintiles of dietary magnesium intake among 7669 U.S. adults ⱖ20 years of age, National Health and Nutrition Examination Survey III 1988 to 1994


40.6 (2.5) 33.8 (2.1) 28.6 (1.7)

1835 29.0 (1.8) 40.1 (1.6) 28.2 (1.9)

Values represent percentage (standard error).

Number Metabolic syndrome (%) Abdominal obesity (%) Hypertriglyceridemia (%) Low high-density lipoprotein cholesterol (%) High blood pressure (%) Hyperglycemia (%)

Quintile 1: men 337

⬍0.001 0.107 0.004

0.002 ⬍0.001 0.092

p for linear trend

Table 2. Unadjusted prevalence of the metabolic syndrome and its components by quintiles of dietary magnesium intake among 7669 U.S. adults ⱖ20 years of age, National Health and Nutrition Examination Survey III 1988 to 1994



1143

1144


0.89 (0.66, 1.21) 0.84 (0.58, 1.23)

0.86 (0.66, 1.13) 1.15 (0.87, 1.53) 0.96 (0.72, 1.29) 0.75 (0.54, 1.05) 1.14 (0.83, 1.56)

1.00 1.00

1.00 1.00 1.00 1.00 1.00


1.00 (0.79, 1.26) 0.64 (0.45, 0.92) 0.90 (0.71, 1.13)

0.81 (0.63, 1.05) 0.92 (0.71, 1.20)

0.83 (0.65, 1.05) 0.76 (0.54, 1.07)


0.86 (0.64, 1.15) 0.60 (0.39, 0.91) 1.02 (0.67, 1.55)

0.72 (0.50, 1.02) 0.85 (0.59, 1.24)

0.73 (0.55, 0.97) 0.62 (0.40, 0.98)


0.83 (0.54, 1.25) 0.61 (0.37, 1.03) 0.85 (0.57, 1.28)

0.84 (0.59, 1.22) 0.85 (0.54, 1.31)

0.70 (0.52, 0.93) 0.56 (0.34, 0.92)

Quintile 5: men >466 and women >337

0.257 0.169 0.371

0.877 0.296

0.074 0.029

p for linear trend

* Model 1: adjusted for age. † Model 2: adjusted for age, sex, race or ethnicity, education, smoking status, concentration of C-reactive protein, alcohol use, physical activity, family history of early coronary heart disease, use of vitamin or supplement, history of diabetes (except model for hyperglycemia), percent calories as fat (as quintiles), percent calories as carbohydrate (as quintiles), fiber intake (as quintiles), and total energy intake (as quintiles).

Metabolic syndrome Model 1* Model 2† Components of the metabolic syndrome Abdominal obesity† Hypertriglyceridemia† Low high-density lipoprotein cholesterol† High blood pressure† Hyperglycemia†

Quintile 1: men