Table 4. Cox proportional hazard analysis of associations of the Mg level and BMD with the risk of total incident fractures after transforming the Mg levels and BMD into binary categories using the median value or official cut-off value.
| Variables | Unadjusted | Adjusted | ||
|---|---|---|---|---|
| HR (95% CI) | P-value | HR (95% CI) | P-value | |
| Lower Mg (≤ 2.5 mg/dL) | 3.91 (1.86–9.22) | 0.0002 | 2.31 (1.03–5.17) | 0.030 |
| Total hip BMD T-score < -1.0 | 3.14 (1.12–13.10) | 0.026 | 1.49 (0.38–5.87) | 0.55 |
| L2-L4 BMD T-score < -1.0 | 3.13 (1.62–6.12) | 0.0007 | 2.59 (1.09–6.16) | 0.027 |
| 1/3 Radius BMD T-score < -1.0 | 2.14 (1.04–4.82) | 0.036 | 0.57 (0.21–1.52) | 0.27 |
Adjusted for age, BMI, HD duration, low serum Mg level (≤ 2.5 mg/dL), past incident fracture, use of phosphate binders, total hip BMD T-score < -1.0, L2-L4 BMD T-score < -1.0 and distal 1/3 radius BMD T-score < -1.0, which were the covariates with P<0.05 in the univariate analysis, as shown in Table 3.