Skip to main content
. Author manuscript; available in PMC: 2021 Sep 1.
Published in final edited form as: J Electrocardiol. 2020 Jul 22;62:20–25. doi: 10.1016/j.jelectrocard.2020.07.008

Table 2.

Associations of serum magnesium with atrial arrhythmias: the ARIC study, 2016-2017

Serum magnesium, mmol/L
<0.75 0.75-<0.85 0.85->0.95 ≥0.95 Per 0.1 mmol/L decrement

AF Burden OR (95% CI)
N 478 986 1004 45
Continuous AF vs. no AF
Continuous AF, N 33 46 43 1
Model 1 1.90 (1.17-3.06) 1.12 (0.73-1.73) 1 (Ref) - 1.27 (1.02-1.57)
Model 2 1.70 (1.01-2.86) 0.96 (0.60-1.54) 1 (Ref) - 1.18 (0.93-1.50)
Model 3 1.48 (0.81-2.68) 0.90 (0.55-1.49) 1 (Ref) - 1.11 (0.84-1.46)
Model 4 1.51 (0.83-2.75) 0.91 (0.55-1.51) 1 (Ref) - 1.12 (0.85-1.47)
Intermittent AF vs. no AF
Intermittent AF, N 17 32 32 0
Model 1 1.30 (0.71-2.38) 1.07 (0.65-1.77) 1 (Ref) - 1.09 (0.83-1.43)
Model 2 1.11 (0.59-2.12) 0.96 (0.57-1.62) 1 (Ref) - 1.03 (0.77-1.37)
Model 3 1.24 (0.63-2.47) 0.94 (0.56-1.59) 1 (Ref) - 1.07 (0.78-1.47)
Model 4 1.29 (0.64-2.58) 0.98 (0.57-1.66) 1 (Ref) - 1.10 (0.80-1.51)

Ratio of Geometric Means (95% CI),§

Nc 400 851 870 42
Isolated PAC burden
Geometric Mean 226 234 192 159
Model 1 1.20 (0.98-1.48) 1.21 (1.03-1.43) 1 (Ref) 0.82 (0.48-1.40) 1.09 (0.99-1.19)
Model 2 1.16 (0.94-1.45) 1.24 (1.05-1.47) 1 (Ref) 0.81 (0.46-1.42) 1.07 (0.97-1.18)
Model 3 1.18 (0.93-1.49) 1.26 (1.06-1.49) 1 (Ref) 0.85 (0.48-1.51) 1.08 (0.97-1.19)
Model 4 1.16 (0.91-1.48) 1.26 (1.06-1.50) 1 (Ref) 0.86 (0.48-1.53) 1.07 (0.96-1.18)
SVT burden
Geometric Mean 1.9 2.2 2.2 1.9
Model 1 0.90 (0.82-0.99) 1.00 (0.92-1.07) 1 (Ref) 0.85 (0.66-1.08) 0.96 (0.92-1.00)
Model 2 0.91 (0.82-1.00) 1.01 (0.94-1.09) 1 (Ref) 0.89 (0.69-1.16) 0.97 (0.93-1.01)
Model 3 0.95 (0.85-1.06) 1.02 (0.95-1.11) 1 (Ref) 0.92 (0.71-1.20) 0.99 (0.94-1.04)
Model 4 0.95 (0.85-1.06) 1.02 (0.95-1.11) 1 (Ref) 0.92 (0.71-1.20) 0.99 (0.94-1.04)

Abbreviations: Atherosclerosis Risk in Communities, ARIC; standard deviation, SD; confidence interval, CI; atrial fibrillation, AF; premature atrial contraction, PAC; supraventricular tachycardia, SVT.

Model 1=age, sex, race-center

Model 2=Model 1+educational attaimnent, smoking status, drinking status, physical activity, magnesium or calcium supplement use, body mass index

Model 3=Model 2+diabetes, systolic and diastolic blood pressure, antihypertensive medication use

Model 4=Model 3+serum potassium

PAC and SVT burden were log-transformed due to right-skew distribution. The ratio of geometric mean refers to the exponentiated beta coefficient from linear regression, e.g., exp(β<0.75 vs 0.85-0.95 mmol/L)· An example interpretation, in Model 1, the geometric mean PAC burden among individuals with serum Mg <0.75 mmol/L was 1.20 fold the geometric mean PAC burden among those with serum Mg 0.85-0.95 mmol/L.

§

Excluding those with AF detected on the ECG patch monitor or AF identified by prior ARIC ascertainment