Table 2 Observational studies examining the relationship between serum Mg levels and mortality
| Authors (year) | Subjects | Dialysis duration | Dialysate Mg (mmol/L) | Follow-up period | Adjusted HR | Adjustments |
|---|---|---|---|---|---|---|
| Pre-dialysis CKD | ||||||
| Van Laecke et al. (2013) [12] | 1650 | – | – | Median 5.1 years |
All-cause mortality 0.930 per 0.1-mg/dL increase in serum Mg levels All-cause mortality 1.613 in the low Mg group (< 1.8 mg/dL) vs. the high Mg group (> 2.2 mg/dL) | Age, sex, DM, hypertension, obesity, smoking, eGFR, diuretics, RAAS blockade, UA, Na, K, P, CRP |
| Dialysis | ||||||
| Sakaguchi et al. (2014) [69] | 142,555, HD | Median 7 years | 0.5 | 1 year | All-cause mortality 1.28, cardiovascular mortality 1.24, non-cardiovascular mortality 1.32 in the lowest Mg sextile (< 2.3 mg/dL) vs. the second highest sextile (≥ 2.8, < 3.1 mg/dL) | Age, sex, HD duration, weekly HD time, BMI, DM, CVD, parathyroidectomy, hip fracture, BUN, albumin, ALP, hemoglobin, Ca, P, CRP, iPTH, VDRAs, PBs, cinacalcet |
| Kurita et al. (2015) [70] | 2165, HD | Median 8.3 years | 0.5 | 3 years |
All-cause mortality 1.734 in the lowest Mg quintile (≤ 2.3 mg/dL) vs. the middle quintile (> 2.5, ≤ 2.7 mg/dL) All-cause mortality 1.649 in the second lowest quintile (> 2.3, ≤ 2.5 mg/dL) vs.the middle quintile | Age, sex, HD duration, Kt/V, primary renal disease, BMI, CVD, lung disease, liver disease, malignancy, parathyroidectomy, albumin, hemoglobin, K, Ca, P, CRP, iPTH, serum iron, ferritin |
| de Roji van Zuijdewijn et al. (2015) [71] | HD 184, HDF 181 | Median 1.8 years | 0.5 | Mean 3.1 years | All-cause mortality 0.88, cardiovascular mortality 0.73, sudden death 0.78 per 0.1-mmol/L (0.24 mg/dL) increase in serum Mg levels | Age, sex, HD duration, weekly HD time, dialysis modality (HD/HDF), residual renal function, BMI, BP, DM, CVD, albumin, Ca, P, iPTH |
| Lacson et al. (2015) [33] | 27,544, HD | Median 2.5 years | Various | 1 year | All-cause mortality 0.89 in the highest Mg group (≥ 1.25 mmol/L) vs. the reference group (≥ 0.80, < 0.95 mmol/L) (not significant) |
Case mix: age, sex, race/ethnicity, HD duration, vascular access type, BSA Laboratory: Kt/V, DM, albumin, hemoglobin, Ca, P, iPTH |
| Li et al. (2015) [34] | 9359, HD | Incident patients | Not mentioned | Mean 19 months | No relationship between time-varying serum Mg levels and all-cause mortality after all adjustments | Age, sex, race/ethnicity, Kt/V, BMI, DM, hypertension, dyslipidemia, CVD, lung disease, liver disease, cancer, BUN, albumin, ALP, hemoglobin, K, Ca, P, iPTH, ferritin, nPCR |
| Cai et al. (2016) [35] | 253, PD | Incident patients | 0.25 | Median 29 months |
All-cause mortality 0.075 in the normomagnesemia group (≥ 1.7 mg/dL) vs. the hypomagnesemia group (< 1.7 mg/dL) Cardiovascular mortality 0.003 in the normomagnesemia group vs. the hypomagnesemia group | Age, sex, BMI, DM, BP, urine output, net UF, weekly Ccr, residual renal function, albumin, total cholesterol, triglycerides, hemoglobin, Na, Ca, P, iPTH, calcium carbonate, VDRAs |
| Yang et al. (2016) [72] | 10,692, PD | Incident patients | Not mentioned | Median 13 months |
Hospitalization risk 1.09 in the lowest Mg quintile (< 1.8 mg/dL) vs. the middle quintile (≥ 2.0, < 2.2 mg/dL) All-cause mortality 0.97 in the lowest Mg quintile vs. the middle quintile (not significant) | Age, sex, race/ethnicity, primary insurance, primary renal disease, total weekly Kt/V, residual renal function, 4-h D/P Cr ratio, DM, hypertension, CVD, albumin, hemoglobin, K, Ca, P, bicarbonate, iPTH, ferritin, iron saturation |