Skip to main content

Table 1 Characteristics of the 11 prospective studies included in the meta-analysis

From: Circulating magnesium levels and incidence of coronary heart diseases, hypertension, and type 2 diabetes mellitus: a meta-analysis of prospective cohort studies

Source

Age at baseline (year)

Men (%)

Duration of follow-up (year)

No. of participants /events

Exposure categories

Exposure assessment

Outcome and its assessment

Adjusted variables

Main results

Study quality

Coronary heart diseases

 

Gartside et al. (1995) [23], The NHANES Study, USA

25–74

NA

10

8251/492

Baseline serum magnesium level (mEq/L): <1.62; 1.62- < 1.74; ≥1.74.

NA

CHD events were determined by hospital records and death certificates (ICD-9 codes 410–414 for CHD).

Age, gender, quetelet index, exercise, sedimentation rate, dietary iron, maximum weight, cigarette smoking, exercise, riboflavin, and alcohol consumption at baseline.

RR (95% CI):1.00 (Referent);0.96 (0.78, 1.19);0.68 (0.54, 0.87).

9

Liao et al. _female. (1998) [24], The ARIC Study, USA

45–64

0

5.2

7767/96

Baseline serum magnesium level (mEq/L): ≤1.5; 1.6; 1.7; 1.8.

Serum magnesium was measured using the metallo-chromic dye calmagite, based on the procedure of Gindler and Heth.

CHD incidence was ascertained by self-report, medical records, and death certificates. Out-of-hospital deaths were confirmed both by the death certificate and an interview with the next of kin and questionnaires completed by the patient’s physicians.

Age, race, field center, waist/hip ratio, smoking status, ethanol drinking, education, sports index, fibrinogen, total and HDL cholesterol, triglycerides, diuretic use, hormone replacement, SBP, and diabetes status.

RR (95% CI):1.00 (Referent);1.02 (0.61, 1.71);0.59 (0.30, 1.20);0.55 (0.27, 1.14).Ptrend = 0.047.

9

Liao et al. _male. (1998) [24], The ARIC Study, USA

45–64

100

5.2

6155/223

Baseline serum magnesium level (mEq/L): ≤1.5; 1.6; 1.7; 1.8.

Serum magnesium was measured using the metallo-chromic dye calmagite, based on the procedure of Gindler and Heth.

CHD incidence was ascertained by self-report, medical records, and death certificates. Out-of-hospital deaths were confirmed both by the death certificate and an interview with the next of kin and questionnaires completed by the patient’s physicians.

Age, race, field center, waist/hip ratio, smoking status, ethanol drinking, education, sports index, fibrinogen, total and HDL cholesterol, triglycerides, diuretic use, SBP, and diabetes status.

RR (95% CI):1.00 (Referent);1.48 (1.03, 2.13);1.08 (0.72, 1.61);0.84 (0.53, 1.31).Ptrend = 0.23.

9

Ford et al. (1999) [25], The NHANES Study, USA

25–74

40.1

15.5

12,340/2637

Baseline serum magnesium level (mmol/L): 0.41- < 0.80; 0.80- < 0.84; 0.84- < 0.89; 0.89–1.45.

Atomic absorption spectrophotometry.

IHD events were determined by hospital records (ICD-9-CM codes 410–414).-

Age, gender, race, education, smoking, cholesterol, SBP, antihypertensive medication, self-reported diabetes, BMI, leisure-time and non-leisure time physical activity, and alcohol consumption.

HR (95% CI):1.00 (Referent);0.95 (0.79, 1.14);0.87 (0.73, 1.04);0.92 (0.79, 1.07).

9

Khan et al. (2010) [26], The Framingham offspring Study, USA

44.3 ± 10.0

48.3

20

3531/554

Baseline serum magnesium level (mg/dL): 1.40–1.77; 1.77–1.88; 1.88–1.98; 1.98–2.50.

Serum magnesium was measured using standard colorimetric assay.

CVD event was defined as angina pectoris, coronary insufficiency (prolonged angina with documented ECG changes), MI, stroke or transient ischemic attack, heart failure, intermittent claudication, or death secondary to CVD.

Age, sex, BMI, diabetes, SBP, total/HDL cholesterol ratio, smoking, hemoglobin, albumin, and estimated GFR.

HR (95% CI):1.00 (Referent);1.09 (0.86, 1.37);0.88 (0.69, 1.13);0.91 (0.72, 1.17).Ptrend = 0.23.Continuous(↑ 0.15 mg/dL):0.83 (0.49, 1.40).

9

Joosten et al. (2013 a) [27], The PREVEND study, The Netherlands

28–75

49

8.1

7764/435

Baseline plasma magnesium level (mmol/L): <0.77; 0.77–0.79; 0.80–0.82; 0.83–0.85; >0.85.

Plasma magnesium was measured by a xylidyl blue method.

Incident IHD was defined as follows: acute myocardial infarction (ICD-10 code I21), hospitalization for other acute ischemic heart disease (ICD-10 code I24), coronary artery bypassing grafting, or percutaneous transluminal coronary angioplasty.

Age, sex, BMI, smoking, diabetes, ratio of total to HDL cholesterol, parenteral history of IHD, alcohol intake, plasma levels of sodium, potassium, and calcium.

HR (95% CI):1.06 (0.79, 1.43);0.90 (0.66, 1.22);1.00 (Referent);1.08 (0.80, 1.45);1.07 (0.80, 1.43).Ptrend = 0.59.

9

Hypertension

 

Peacock et al. _female. (1999) [28], The ARIC Study, USA

52.8 (45–64)

0

6

4190/822

Baseline serum magnesium levels in quartiles (mEq/L): 0.7–1.5; 1.6; 1.7; 1.8–2.3.

Serum magnesium was measured using the metallo-chromic dye calmagite, based on the procedure of Gindler and Heth.

Incident hypertension was defined as the new occurrence of SBP ≥ 140 mmHg or DBP ≥ 90 mmHg, or currently taking antihypertensive medication at the follow up visits.

Age, race, field center, BMI, waist/hip ratio, diabetes, education, family history of hypertension, leisure activity score, hormone replacement therapy, and baseline SBP.

OR (95% CI):1.00 (Referent);0.82 (0.64, 1.05);0.93 (0.73, 1.19);0.76 (0.58, 0.99).Ptrend = 0.11.

9

Peacock et al. _male. (1999) [28], The ARIC Study, USA

53.5 (45–64)

100

6

3541/755

Baseline serum magnesium levels in quartiles (mEq/L): 0.7–1.5; 1.6; 1.7; 1.8–2.3.

Serum magnesium was measured using the metallo-chromic dye calmagite, based on the procedure of Gindler and Heth.

Incident hypertension was defined as the new occurrence of SBP ≥ 140 mmHg or DBP ≥ 90 mmHg, or currently taking antihypertensive medication at the follow up visits.

Age, race, field center, BMI, waist/hip ratio, diabetes, education, family history of hypertension, leisure activity score, and baseline SBP.

OR (95% CI):1.00 (Referent);0.91 (0.70, 1.19);0.92 (0.71, 1.19);0.90 (0.68, 1.18).Ptrend = 0.50.

9

Khan et al. (2010) [26], The Framingham offspring Study, USA

42.5 ± 9.5

45.0

8

2520/551

Baseline serum magnesium level (mg/dL): 1.40–1.77; 1.77–1.88; 1.88–1.98; 1.98–2.50.

Serum magnesium was measured using standard colorimetric assay.

Incident hypertension was defined as the new occurrence of SBP ≥ 140 mmHg or DBP ≥ 90 mmHg or currently taking antihypertensive medication, including diuretics.

Age, sex, BMI, diabetes, systolic blood pressure, total/HDL cholesterol ratio, smoking, hemoglobin, albumin, and eGFR.

HR (95% CI):1.00 (Referent);0.97 (0.72, 1.31);0.96 (0.70, 1.32);1.03 (0.75, 1.41).Ptrend = 0.89.Continuous(↑ 0.15 mg/dL):1.03 (0.92, 1.15).

9

Joosten et al. (2013 b) [29], The PREVEND study, The Netherlands

28–75

45.3

7.6

4625/1021

Baseline plasma magnesium level (mmol/L): 0.55–0.77; 0.78–0.80; 0.81–0.84; 0.85–1.04.

Plasma magnesium was measured by a xylidyl blue method.

Incident hypertension was defined as the new occurrence of SBP ≥ 140 mmHg, DBP ≥ ≥ 90 mmHg, or initiation of antihypertensive medication.

Age, sex, BMI, smoking, diabetes, parenteral history of hypertension, alcohol intake, study design, and plasma levels of sodium, potassium, and calcium.

HR (95% CI):1.00 (Referent);1.00 (0.84, 1.20);0.92 (0.77, 1.09);0.94 (0.79, 1.12).Continuous(mmol/L):0.94 (0.83, 1.05).

9

Diabetes mellitus

 

Kao et al. _black. (1999) [11], The ARIC Study, USA

53.0

36.3

5.3

2622/367

Baseline serum magnesium level (mEq/L): 0.50–1.40; 1.50; 1.60; 1.70; 1.80; 1.90–2.60.

Serum magnesium was measured using the metallo-chromic dyecalmagite, based on the procedure of Gindler and Heth.

Diabetes mellitus was defined as the presence of any of the following: 1) fasting glucose levels ≥7.0 mmol/L; 2) non-fasting glucose levels ≥11.1 mmol/L; 3) current use of diabetic medication; 4) a positive response to the question “Has a doctor ever told you that you had diabetes?” Individuals without diabetes at baseline who met any of these conditions at follow-up were considered to have incident cases of diabetes.

Age, gender, education, BMI, family history of diabetes, waist/hip ratio, sports index, alcohol consumption, diuretic use, serum calcium and potassium levels, and fasting insulin and glucose levels at baseline.

OR (95% CI):0.90 (0.48, 1.68);0.82 (0.45, 1.51);1.01 (0.56, 1.83);0.94 (0.51, 1.73);0.94 (0.48, 1.81);1.00 (Referent).Ptrend = 0.71.

9

Kao et al. _white. (1999) [11], The ARIC Study, USA

54.2

46.0

5.6

9506/739

Baseline serum magnesium level (mEq/L): 0.50–1.40; 1.50; 1.60; 1.70; 1.80; 1.90–2.60.

Serum magnesium was measured using the metallo-chromic dye calmagite, based on the procedure of Gindler and Heth.

Diabetes mellitus was defined as the presence of any of the following: 1) fasting glucose levels ≥ 7.0 mmol/L; 2) non-fasting glucose levels ≥11.1 mmol/L; 3) current use of diabetic medication; 4) a positive response to the question “Has a doctor ever told you that you had diabetes?” Individuals without diabetes at baseline who met any of these conditions at follow-up were considered to have incident cases of diabetes.

Age, gender, education, BMI, family history of diabetes, waist/hip ratio, sports index, alcohol consumption, diuretic use, serum calcium and potassium levels, and fasting insulin and glucose levels at baseline.

OR (95% CI):1.55 (1.01, 2.37);1.11 (0.76, 1.63);1.05 (0.73, 1.50);1.10 (0.78, 1.57);1.07 (0.74, 1.57);1.00 (Referent).Ptrend = 0.10.

9

Everett et al. (2006) [30], NHANES I, USA

25–74

NA

15.0

9784/690

Baseline serum magnesium level in quartiles (mEq/L): ≤1.60; 1.60–1.68; 1.69–1.77; ≥1.78.

Atomic absorption spectrophotometry

Incident diabetic cases were identified from nursing home and hospital records and death certificates. ICD-9 (250.0–250.9) representing diabetes was used to identify individuals who were admitted to a health care facility.

Age, gender, race, education, exercise, BMI, hypertension, and total cholesterol.

HR (95% CI):1.51 (1.12, 2.03);1.20 (0.88, 1.62);0.99 (0.73, 1.34);1.00 (Referent).

9

Guerrero-Romero et al. (2008) [31], The Mexican Diabetes Prevention study, Mexico

56.7 ± 11.9 (20–65)

NA

6.6

817/78

Baseline serum magnesium level in two groups (mmol/L): <0.74; ≥0.74.

Serum magnesium was measured by colorimetric method with the Data Pro Plus Random Access Clinical Analyzer (Arlington, TX, USA).

Diabetes mellitus was defined as the presence of any of the following: 1) 2-h post load serum glucose levels ≥11.1 mmol/L; 2) current use of diabetic medication(hypoglycemic drugs or insulin).

Age, gender, family history of diabetes, waist circumference, and HOMA-IR index.

RR (95% CI):2.54 (1.1, 4.1);1.00 (Referent).

7

Kieboom et al. (2017) [32], The Rotterdam Study, the Netherlands

64.7 ± 9.7

42.2

6.7

8555/806

Baseline serum magnesium level in two groups (mmol/L): ≤0.72; >0.72.

Serum magnesium was measured using a colorimetric endpoint method with the Roche/Hitachi Cobas c501 Analyzer (Roche Diagnostics, Indianapolis, IN, USA)

Incident diabetic cases were ascertained from general practioners’ records, hospital discharge letters and glucose measurements. Diabetes was defined as the presence of any of the following: 1) fasting glucose levels ≥ 7.0 mmol/L; 2) non-fasting glucose levels ≥11.1 mmol/L; 3) current use of blood glucose lowering medication.

Age, age2, gender, BMI, smoking status, alcohol use, total cholesterol: HDL-cholesterol ratio, history of hypertension, stroke and CHD, eGFR, serum levels of calcium and potassium, as well as use of diuretics

HR (95% CI):1.79 (1.16, 2.77);1.00 (Referent).Continuous(↓0.1 mmol/L):1.18 (1.04, 1.33).

9

  1. Abbreviations: ARIC Atherosclerosis Risk in Communities, BMI body mass index, CHD coronary heart disease, CI confidence interval, CVD cardiovascular diseases, DBP diastolic blood pressure, eGFR estimated glomerular filtration rate, HDL high-density lipoprotein, HOMA-IR homeostasis model assessment for insulin resistance, HR hazards ratio, IHD ischemic heart disease, MI myocardial infarction, NA not available, NHANES National Health and Nutrition Examination Survey, OR odds ratio, PREVEND THE Prevention of Renal and Vascular End-Stage Disease Study, RR relative risk, SBP systolic blood pressure, USA the United States of America
  2. *It was evaluated with the Newcastle-Ottawa scale