Page 112 - 南京医科大学自然版
P. 112
第44卷第12期
·1726 · 南 京 医 科 大 学 学 报 2024年12月
表2 按照基线糖尿病和血糖水平分组的结核发病密度
Table 2 Tuberculosis incidence in groups based on baseline diabetes and blood glucose levels
Number of morbidity Number of observation Incidence 95%CI
Group
(n) (person⁃year) (100 000 person⁃years) (100 000 person⁃years)
All subjects 79 308 850 25.6 20.4-31.7
Non⁃diabetes mellitus 62 279 939 22.1 17.1-28.2
Diabetes mellitus 17 028 911 58.8 35.4-92.2
FBG<7.0 mmol/L 04 013 506 29.6 21.4-33.1
FBG≥7.0 mmol/L 13 020 460 63.5 39.3-96.8
建列线图模型,预测结核病风险。所有变量单项总得 分增加18分(图2)。
分和所对应的概率为受试者不患结核病的概率,结果
3 讨 论
显示BMI预测能力最大,其次是年龄,FBG和性别预测
能力较小;FBG<7.0 mmol/L时评分增加33分;女性评 这项在中国东部地区开展的 60 283 例的队列
Viriable HR(95%CI) P
Sex
Fcmalc 1
Man 1.735(1.322-1.864) <0.001
Age 1.014(1.001-1.027) 0.041
BMI(Continious) 0.924(0.861-0.993) 0.030
BMI(kg/m )
2
<18.5 1
18.5-<24.0 2.371(0.975-5.948) 0.074
24.0-<28.0 5.731(1.805-7.655) 0.003
≥28.0 1.045(0.493-3.188) 0.794
Smoke
No 1
Yes 1.074(0.843-1.377) 0.830
Diabetes mellitus
No 1
Yes 1.422(1.220-1.642) 0.010
FBG
<7.0 mmol/L 1
≥7.0 mmol/L 3.092(1.836-5.914) 0.001
0 2 4 6 8 10
图1 结核病发病风险的单因素分析森林图
Figure 1 The forest map of tuberculosis risk factors by univariate analysis
表3 结核病发病风险的多因素Cox比例风险模型
Table 3 The multivariate Cox proportional risk model of tuberculosis risk factors
Model Ⅰ Model Ⅱ Model Ⅲ
Variable
HR(95%CI) P HR(95%CI) P HR(95%CI) P
Sex
Female 1 1 1
Male 1.511(1.321-1.814) <0.005 1.517(1.324-1.823) <0.005 1.506(1.318-1.807) <0.004
BMI(Continious) 0.887(0.823-0.956) <0.002 0.888(0.824-0.956) <0.002 0.900(0.836-0.969) <0.005
Diabetes mellitus
No 1
Yes 3.057(1.770-5.281) <0.001
FBG
<7.0 mmol/L 1
≥7.0 mmol/L 3.766(2.054-6.906) <0.001
Continuous variable of FBG 1.085(1.041-1.132) <0.001
Model Ⅰ:Diabetes was included in the model as a dichotomous variable(yes or no);Model Ⅱ:FBG was included in the model as a dichotomous
variable(FBG<7.0 mmol/L, FBG≥7.0 mmol/L);Model Ⅲ:FBG was included in the model as a continuity variable.

