Shandong Provincial Hospital, Shandong First Medical University & Shandong Academy of Medical Sciences, Jinan, 250021, China.
Beijing Institute of Brain Disorders, Laboratory of Brain Disorders, Ministry of Science and Technology, Collaborative Innovation Center for Brain Disorders, Capital Medical University, Beijing, 100069, China.
Clin Nutr. 2021 May;40(5):2726-2733. doi: 10.1016/j.clnu.2021.03.012. Epub 2021 Mar 25.
BACKGROUND & AIMS: Mendelian randomization (MR) studies have reported the causal association between serum calcium levels and bone mineral density (BMD). The results showed that genetically increased serum calcium levels in individuals with normal calcium levels did not increase BMD and could even reduce BMD. However, whether there are differences in the association between serum calcium and BMD in different age strata remains unclear.
We selected eight serum calcium genetic variants with genome-wide significance (P < 5.00E-08) as the potential instrumental variables. We conducted an MR analysis to evaluate the impact of serum calcium levels on total body BMD in five age strata, 0-15, 15-30, 30-45, 45-60, and ≥60 years, using large-scale serum calcium (61,079 individuals) and total body BMD genome-wide association study (66,628 individuals) datasets. For pleiotropy analysis, we used a manual method and four common statistical methods, namely the MR-Egger intercept, MR-PRESSO, heterogeneity, and Steiger filtering tests. For MR analysis, we selected four MR methods, namely inverse-variance weighted, weighted median, MR-Egger, and MR-PRESSO. In addition to the univariable MR analysis, we conducted a multivariate MR analysis taking into account the effect of serum parathyroid hormone levels.
Univariable MR analysis using the inverse-variance weighted method indicated that per 0.5-mg/dL increase (about 1 standard deviation) in serum calcium levels was statistically significantly associated with reduced total body BMD only in the ≥60 years stratum (effect estimate (beta) = -0.545, 95% confidence interval (CI): -0.892 to -0.198, P = 0.002). The weighted median regression (beta = -0.446, 95% CI: -0.821 to -0.094, P = 1.40E-02) and MR-PRESSO (beta = -0.545, 95% CI: -0.892 to -0.198, P = 0.022) MR methods further supported this suggestive association. The multivariable MR analysis also found a significant association between increased serum calcium levels and reduced total body BMD in the ≥60 years stratum (beta = -0.547, 95% CI: -0.934 to -0.16, P = 0.006).
Our results provide genetic evidence that increased serum calcium levels did not improve BMD in the general population and that the elevated serum calcium levels in generally healthy populations, especially in adults older than 60 years, may even reduce the BMD. Our results are comparable with those of recent MR findings.
孟德尔随机化（MR）研究报告了血清钙水平与骨密度（BMD）之间的因果关联。结果表明，在血钙水平正常的个体中，遗传上增加的血清钙水平并不能增加 BMD，甚至可能降低 BMD。然而，不同年龄组血清钙与 BMD 之间的关联是否存在差异尚不清楚。
我们选择了 8 个具有全基因组意义的血清钙遗传变异（P < 5.00E-08）作为潜在的工具变量。我们使用大型血清钙（61079 人）和全身 BMD 全基因组关联研究（66628 人）数据集，在五个年龄组（0-15、15-30、30-45、45-60 和≥60 岁）中进行了 MR 分析，以评估血清钙水平对全身 BMD 的影响。对于 pleiotropy 分析，我们使用了手动方法和四种常见的统计方法，即 MR-Egger 截距、MR-PRESSO、异质性和 Steiger 过滤检验。对于 MR 分析，我们选择了四种 MR 方法，即逆方差加权、加权中位数、MR-Egger 和 MR-PRESSO。除了单变量 MR 分析外，我们还进行了多变量 MR 分析，考虑了甲状旁腺激素水平的影响。
使用逆方差加权法的单变量 MR 分析表明，血清钙水平每增加 0.5mg/dL（约 1 个标准差），与全身 BMD 降低仅在≥60 岁组有统计学显著关联（效应估计（β）=-0.545，95%置信区间（CI）：-0.892 至-0.198，P=0.002）。加权中位数回归（β=-0.446，95%CI：-0.821 至-0.094，P=1.40E-02）和 MR-PRESSO（β=-0.545，95%CI：-0.892 至-0.198，P=0.022）MR 方法进一步支持了这一提示性关联。多变量 MR 分析也发现血清钙水平升高与全身 BMD 降低之间存在显著关联，尤其是在≥60 岁的年龄组（β=-0.547，95%CI：-0.934 至-0.16，P=0.006）。
我们的研究结果提供了遗传证据，表明血清钙水平升高并不能改善一般人群的 BMD，而在一般健康人群中升高的血清钙水平，尤其是 60 岁以上的成年人，甚至可能降低 BMD。我们的研究结果与最近的 MR 研究结果相似。
