Chonchol Michel, Dale Rita, Schrier Robert W, Estacio Raymond
Division of Renal Diseases and Hypertension, University of Colorado at Denver, Health Sciences Center, Denver, CO 80262, USA.
Am J Med. 2009 Apr;122(4):380-6. doi: 10.1016/j.amjmed.2008.09.039.
Although serum phosphorus, calcium, and calcium-phosphorus product levels have been associated with cardiovascular events and mortality in patients with normal kidney function, most studies have not examined the association of these minerals with outcomes when collected repeatedly over time.
We evaluated the association of serum phosphorus, calcium, and calcium-phosphorus product levels with cardiovascular events and mortality in 950 participants of the Appropriate Blood Pressure Control in Diabetes trial by both time-dependent Cox regression models using the cumulative average of minerals measured over time and fixed covariate Cox regression models with only baseline values of these minerals.
There were 42 deaths and 193 cardiovascular events among the participants, who were followed for an average of 4.8 years following randomization. A significant association was noted between baseline serum phosphorus >3.9 mg/dL and baseline calcium-phosphorus product >36.8 mg(2)/dL(2) compared with the lowest referent category with the adjusted risk of cardiovascular death (hazard ratio [HR] 5.00; 95% confidence interval [CI], 1.70-14.72) and (HR 10.01; 95% CI, 2.55-39.31), respectively. However, in time-dependent models using mineral values repeated during the course of the study, only the average of serum phosphorus remains significant (HR 4.25; 95% CI, 1.15 to 16.65).
In the Appropriate Blood Pressure Control in Diabetes cohort, serum phosphorus, but not serum calcium or calcium-phosphorus product, was associated with cardiovascular mortality in time-dependent Cox regression models. Thus, serum phosphorus levels may be more reliable in predicting cardiovascular mortality in patients with type 2 diabetes.
尽管血清磷、钙及钙磷乘积水平与肾功能正常患者的心血管事件及死亡率相关，但大多数研究并未考察这些矿物质随时间重复测量时与预后的关联。
我们在糖尿病患者血压适度控制试验的950名参与者中，通过使用随时间测量的矿物质累积平均值的时间依赖性Cox回归模型以及仅使用这些矿物质基线值的固定协变量Cox回归模型，评估血清磷、钙及钙磷乘积水平与心血管事件及死亡率的关联。
参与者中共有42例死亡和193例心血管事件，随机分组后平均随访4.8年。与最低参照类别相比，基线血清磷>3.9mg/dL和基线钙磷乘积>36.8mg²/dL²与心血管死亡的校正风险显著相关（风险比[HR]分别为5.00；95%置信区间[CI]，1.70 - 14.72）和（HR 10.01；95%CI，2.55 - 39.31）。然而，在使用研究过程中重复测量的矿物质值的时间依赖性模型中，只有血清磷的平均值仍具有显著性（HR 4.25；95%CI，1.15至16.65）。
在糖尿病患者血压适度控制队列中，在时间依赖性Cox回归模型中，血清磷而非血清钙或钙磷乘积与心血管死亡率相关。因此，血清磷水平在预测2型糖尿病患者心血管死亡率方面可能更可靠。
