中国组织工程研究

中国组织工程研究 ›› 2023, Vol. 27 ›› Issue (31): 5046-5051.doi: 10.12307/2023.569

• 骨科植入物相关临床实践 Clinical practice of orthopedic implant • 上一篇    下一篇

中老年人群血小板计数与腰椎骨密度的相关性

张金鹏1，陈  畅2，潘秋予3，麦陈耀3，李印龙3，郝玉玺1，胡  军4   

  1. 1潍坊医学院，山东省潍坊市  261053；2山东大学齐鲁医院，山东省济南市  250012；3西藏大学，西藏自治区拉萨市  850000；4山东中医药大学管理学院，山东省济南市  250355
  • 收稿日期:2022-08-16 接受日期:2022-09-24 出版日期:2023-11-08 发布日期:2023-01-31
  • 通讯作者: 胡军，博士生导师，山东中医药大学管理学院，山东省济南市 250355
  • 作者简介:张金鹏，男，1996年生，山东省潍坊市人，汉族，硕士，主要从事中老年人群骨质疏松症的预测与防治研究。

Correlation between platelet count and lumbar bone mineral density in middle-aged and elderly people

Zhang Jinpeng1, Chen Chang2, Pan Qiuyu3, Mai Chenyao3, Li Yinlong3, Hao Yuxi1, Hu Jun4   

  1. 1Weifang Medical University, Weifang 261053, Shandong Province, China; 2Qilu Hospital of Shandong University, Jinan 250012, Shandong Province, China; 3Tibet University, Lasa 850000, Tibet Autonomous Region, China; 4School of Management, Shandong University of Traditional Chinese Medicine, Jinan 250355, Shandong Province, China
  • Received:2022-08-16 Accepted:2022-09-24 Online:2023-11-08 Published:2023-01-31
  • Contact: Hu Jun, Doctoral supervisor, School of Management, Shandong University of Traditional Chinese Medicine, Jinan 250355, Shandong Province, China
  • About author:Zhang Jinpeng, Master, Weifang Medical University, Weifang 261053, Shandong Province, China

PDF

31

摘要：


文题释义：

多元线性回归模型：指用2个或2个以上的影响因素作为自变量来解释因变量的变化，多个自变量与因变量之间是线性关系。此次研究以血小板计数为自变量，年龄、性别、种族、受教育程度、婚姻史、体质量指数、腰围、钙、磷、高密度脂蛋白胆固醇、三酰甘油、低密度脂蛋白胆固醇、空腹血糖、尿白蛋白和尿肌酐为协变量，腰椎骨密度为因变量，用3个模型探讨中老年人群血小板计数与腰椎骨密度之间的关系，结果显示中老年人群腰椎骨密度与血小板计数呈负相关。
平滑曲线拟合：使用广义相加模型检验结果变量和危险因素(暴露)之间的非线性关系，有助于发现非线性的关系，并判断有无阙值效应，确定用一般的线性回归是否适当。此次研究用平滑曲线拟合解决血小板计数与腰椎骨密度的非线性关系，按性别分层后发现男性血小板计数在正常范围内存在2个拐点，用二分段线性回归模型找出这2个拐点，结果显示2个拐点分别是141×109 L-1和246×109 L-1。

背景：近年研究发现，血小板计数与骨密度之间可能具有内在联系，但对于二者内在联系的结论不一。
目的：研究中老年人群腰椎骨密度与血小板计数的相关性。
方法：研究对象来自美国NHANES数据库，共纳入2 571名男性和2 929名女性，对样本加权后采用加权多元线性回归模型和平滑曲线拟合，评估腰椎骨密度与血小板计数之间的关联。

结果与结论：①加权多元线性回归模型结果显示，总人群腰椎骨密度与血小板计数呈负相关；将腰椎骨密度4分组后与血小板计数仍呈负相关；按性别分层后在男性中仍呈现负相关关系，男性β=-0.000 2，95%CI(-0.000 3，<-0.000 1)，按种族分层后只有其他种族美国人的3个模型呈现正相关关系[模型1：β=0.000 2，95%CI(< 0.000 1，0.000 4)；模型2：β=0.000 1，95%CI(-0.000 1，0.000 3)；模型3：β=0.000 1，95%CI(-0.000 1，0.000 3)]，而西班牙裔、非西班牙裔白人和非西班牙裔黑人3个模型呈现负相关关系；②采用平滑曲线拟合解决血小板计数与腰椎骨密度之间的非线性关系，发现男性正常范围内的血小板计数与腰椎骨密度存在2个拐点，使用二分段线性回归模型找出这2个拐点，结果显示2个拐点分别是141×109 L-1和246×109 L-1；③结果显示，中老年人群男性腰椎骨密度与血小板计数呈负相关，且正常范围内血小板计数< 141×109 L-1时腰椎骨密度与血小板计数正相关，血小板计数在(141-246)×109 L-1时腰椎骨密度与血小板计数负相关，血小板计数> 246×109 L-1时腰椎骨密度与血小板计数正相关；女性腰椎骨密度与血小板计数可能呈正相关。

https://orcid.org/0000-0003-2215-3000(张金鹏)

中国组织工程研究杂志出版内容重点：人工关节；骨植入物；脊柱；骨折；内固定；数字化骨科；组织工程

关键词: 腰椎骨密度, 血小板计数, 相关性分析, 加权多元线性回归模型, 平滑曲线拟合, 阈值效应分析

Abstract: BACKGROUND: Recent studies have found that there may be an intrinsic association between platelet count and bone mineral density. However, previous studies have reached different conclusions on the intrinsic association between the two. 
OBJECTIVE: To study the correlation between lumbar spine bone density and platelet count in middle-aged and elderly people. 
METHODS: A total of 2 571 men and 2 929 women were recruited from the NHANES database. Weighted multiple linear regression models and smooth curve fitting were used to evaluate the association between lumbar bone mineral density and platelet count. 
RESULTS AND CONCLUSION: (1) The weighted multiple linear regression model results showed that lumbar bone mineral density was negatively correlated with platelet count in the total population. There was still a negative correlation between lumbar bone density and platelet count after grouping into four groups. When stratified by sex, there was still a negative correlation in men, β=-0.000 2, 95% confidence interval (CI): (-0.000 3, < -0.000 1). When stratified by race, only the three models for Americans of other races showed a positive correlation [model 1: β=0.000 2, 95% CI: (< 0.000 1, 0.000 4); Model 2: β=0.000 1, 
95% CI: (-0.000 1, 0.000 3); Model 3: β=0.000 1, 95% CI: (-0.000 1, 0.000 3)], while the three models of Hispanic, non-Hispanic white and non-Hispanic black showed a negative correlation. (2) The smooth curve fitting was used to solve the nonlinear relationship between platelet count and lumbar bone mineral density. It was found that there were two inflection points between platelet count and lumbar bone mineral density within the normal range in men. The piecewise linear regression model was used to identify the two inflection points, which were 141×109/L and 246×109/L. (3) To conclude, there is a negative correlation between lumbar bone mineral density and platelet count in middle-aged and elderly men. There is a positive correlation between lumbar bone mineral density and platelet count in men when platelet count is < 141×109/L in the normal range, a negative correlation when platelet count is (141-246)×109/L, and a positive correlation when platelet count is > 246×109/L. There may be a positive correlation between lumbar bone density and platelet count in women.

Key words: lumbar vertebra bone density, platelet count, correlation analysis, weighted multiple linear regression model, smooth curve fitting, threshold effect analysis

中图分类号: