铁路职工高血压病患病风险预测模型的建立与评价

周忠云; 马雯璐; 杜晓林

doi:10.16766/j.cnki.issn.1674-4152.004063

铁路职工高血压病患病风险预测模型的建立与评价

doi: 10.16766/j.cnki.issn.1674-4152.004063

山东中医药大学第二附属医院心内科，山东济南 250001

基金项目:

山东省中医药科技发展计划项目 Z-2022051

详细信息

通讯作者:
杜晓林，E-mail: zdey2023@126.com

中图分类号: R544.1
计量
- 文章访问数: 12
- HTML全文浏览量: 6
- PDF下载量: 2
- 被引次数: 0
出版历程
- 收稿日期: 2024-06-08
- 网络出版日期: 2025-09-04

Establishment and evaluation of a model for predicting the risk of hypertension among railroad workers

Department of Cardiology, the Second Affiliated Hospital of Shandong University of TCM, Jinan, Shandong 250001, China

摘要

摘要: 目的分析铁路职工高血压患病的危险因素，建立适合铁路职工的高血压患病风险预测模型。方法选取2022年1—12月于山东中医药大学第二附属医院进行体检的铁路职工2 717人为研究对象，通过问卷调查、现场体格检查和实验室检测等方法进行资料收集，按7∶3分为建模集和验证集，经过单因素分析及LASSO回归分析筛选变量，然后应用多因素logistic回归分析确定最终纳入的预测变量，构建高血压患病风险列线图模型并通过受试者工作特征(ROC)曲线、校准曲线和决策曲线分析法(DCA)对预测模型进行评价。结果调查2 717人，951人患有高血压，高血压患病率为35.0%。经过单因素和多因素分析，年龄、缺乏运动、失眠、高盐饮食、高血压家族史、BMI、心率(HR)、甘油三酯(TG)均被纳入预测模型。通过验证，模型显示出了很好的预测性能，ROC曲线下面积AUC为0.844(95% CI: 0.825~0.860)，验证集的AUC为0.869(95% CI：0.845~0.896)，校准曲线表现出良好的一致性，DCA分析表明，建模集和验证集显示出很大的正向收益率。结论本研究建立了预测高血压患病的列线图模型，该模型具有中等水平的预测效能，可用于指导铁路职工高血压患病风险预测。
- 高血压 /
- 铁路职工 /
- 危险因素 /
- 列线图 /
- 预测模型 /
- 决策曲线分析
Abstract: Objective To analyze the risk factors of hypertension among railroad workers and to establish an individualized risk prediction model for hypertension. Methods A total of 2 717 railroad workers who underwent the physical examination in the Second Affiliated Hospital of Shandong University of TCM from January 2022 to December 2022 were enrolled in this study, Data collection was done through face-to-face questionnaires, on-site physical examinations, and laboratory tests. According to the ratio of 7∶3, they were randomly divided into a modeling group and a validation group. After preliminary screening by univariate analysis and LASSO regression, the predictive variables were determined by multivariate analysis, and a nomogram model of hypertension was constructed. The predictive models were evaluated using the ROC curve, calibration curve, and decision curve analysis (DCA). Results A total of 2 717 railroad workers were enrolled in this study, and 951 railroad workers had hypertension, the incidence of hypertension was 35.0%, After a univariate and multivariate analysis, age, lack of exercise, loss of sleep, high salt diet, family history of hypertension, BMI, heart rate (HR), triglyceride (TG)were included in prediction model. The model showed a better performance, the area under the ROC curve was 0.844 (95% CI: 0.825-0.860), and the AUC for the validation group was 0.869 (95% CI: 0.845-0.896). The calibration curve showed a good agreement. DCA analysis showed that the modeling and validation groups showed large positive yields. Conclusion We developed a nomogram for predicting the risk of hypertension among railroad workers, which can be used as a tool to guide future patients.
- Hypertension /
- Railroad workers /
- Risk of illness /
- Prediction model /
- Nomogram /
- Decision curve analysis

HTML全文

图 1 基于LASSO回归筛选纳入预测模型的危险因素

注：A为全部危险因素的LASSO系数分布；B为基于LASSO模型最优参数(λ)选择危险因素。

Figure 1. LASSO regression-based screening of risk factors for inclusion in predictive models

下载: 全尺寸图片幻灯片

图 2 高血压患病风险预测模型列线图

Figure 2. Hypertension prevalence risk prediction model line graph

下载: 全尺寸图片幻灯片

图 3 高血压患病风险预测模型的校准曲线

注：A为建模集的校准曲线；B为验证集的校准曲线。

Figure 3. Calibration curve of predictive models for hypertension prevalence risk

下载: 全尺寸图片幻灯片

图 4 高血压患病风险预测模型的ROC曲线

注：A为建模集的ROC曲线；B为验证集的ROC曲线。

Figure 4. ROC curve of prediction models for hypertension prevalence risk

下载: 全尺寸图片幻灯片

图 5 高血压患病风险预测模型的DCA曲线

注：A为建模集的DCA曲线；B为验证集的DCA曲线。

Figure 5. DCA curve of predictive models for hypertension prevalence risk

下载: 全尺寸图片幻灯片

表 1 建模集和验证集基线资料比较

Table 1. Comparison of baseline data between the modeling set and the validation set

项目	建模集(n=1 901)	验证集(n=816)	统计量	P值
高血压[例(%)]			0.740^a	0.390
是	678(35.67)	277(33.95)
否	1 223(64.33)	539(66.05)
性别[例(%)]			0.078^a	0.780
男性	1 891(99.47)	811(99.39)
女性	10(0.53)	5(0.61)
年龄[例(%)]			0.006^a	0.997
18~＜40岁	1 424(74.91)	611(74.88)
40~＜50岁	329(17.31)	142(17.40)
≥50岁	148(7.78)	63(7.72)
身高[M(P₂₅, P₇₅), cm]	177(173, 181)	177(173, 181)	1.829^b	0.067
体重[M(P₂₅, P₇₅), kg]	81(72, 90)	82(73, 91)	1.861^b	0.063
BMI[M(P₂₅, P₇₅)]	25.90(23.20, 28.40)	26.20(23.30, 28.70)	1.508^b	0.131
缺乏运动[例(%)]			1.600^a	0.206
是	729(38.35)	292(35.78)
否	1 172(61.65)	524(64.22)
失眠[例(%)]			2.643^a	0.104
是	788(41.45)	311(38.11)
否	1 113(58.55)	505(61.89)
吸烟[例(%)]			3.670^a	0.055
是	671(35.30)	257(31.50)
否	1 230(64.70)	559(68.50)
饮酒[例(%)]			1.748^a	0.186
是	584(30.72)	230(28.19)
否	1317(69.28)	586(71.81)
高盐饮食[例(%)]			0.277^a	0.599
是	684(35.98)	285(34.93)
否	1 217(64.02)	531(65.07)
高血压家族史[例(%)]			0.353^a	0.552
是	406(21.29)	166(20.34)
否	1 495(78.64)	650(79.66)
糖尿病史[例(%)]			0.016^a	0.898
是	91(4.78)	40(4.90)
否	1 810(95.22)	776(95.10)
冠心病史[例(%)]			0.838^a	0.360
是	277(14.57)	108(13.24)
否	1 624(85.43)	708(86.76)
收缩压[M(P₂₅, P₇₅), mmHg]	135(126, 143)	135(126, 143)	1.243^b	0.214
舒张压[M(P₂₅, P₇₅), mmHg]	82(75, 89)	82(74, 89)	0.841^b	0.400
HR[M(P₂₅, P₇₅), 次/min]	82(74, 89)	81(72, 89)	1.290^b	0.197
WBC[M(P₂₅, P₇₅), ×10⁹/L]	6.85(5.71, 7.79)	6.82(5.61, 7.79)	0.609^b	0.543
RBC(x±s，×10¹²/L)	5.24±0.01	5.20±0.01	2.786^c	0.676
HB[M(P₂₅, P₇₅), g/L]	161(155, 167)	160(153, 167)	1.892^b	0.058
PLT[M(P₂₅, P₇₅), ×10⁹/L]	248(210, 282)	249(212, 282)	0.251^b	0.802
ALT[M(P₂₅, P₇₅), U/L]	31(17, 37)	31(17, 38)	0.473^b	0.636
AST[M(P₂₅, P₇₅), U/L]	26(20, 28)	25(20, 28)	0.227^b	0.821
BUN[M(P₂₅, P₇₅), μmol/L]	4.85(4.08, 5.49)	4.88(4.13, 5.55)	0.672^b	0.501
Scr[M(P₂₅, P₇₅), μmol/L]	76.30(69.90, 81.80)	75.90(69.70, 82.30)	0.306^b	0.759
UA[M(P₂₅, P₇₅), mmol/L]	389(336, 438)	388(336, 436)	0.219^b	0.827
FPG(x±s，mmol/L)	5.22±0.03	5.47±0.05	4.793^c	0.798
TC[M(P₂₅, P₇₅), mmol/L]	4.86(4.19, 5.44)	4.92(4.26, 5.52)	1.792^b	0.073
TG(x±s，mmol/L)	1.55±0.04	1.67±0.07	1.633^c	0.230
LDL-C[M(P₂₅, P₇₅), mmol/L]	2.59(2.13, 3.02)	2.62(2.18, 3.05)	1.873^b	0.061
注：^a为χ²值, ^b为Z值, ^c为t值。1 mmHg=0.133 kPa。

下载: 导出CSV

表 2 铁路职工高血压病患病风险的单因素分析

Table 2. Univariate analysis of the risk of hypertension among railway workers

项目	高血压组(n=678)	正常组(n=1 223)	统计量	P值
性别[例(%)]				0.510^c
男性	676(99.71)	1 215(99.35)
女性	2(0.29)	8(0.65)
年龄[例(%)]			93.832^a	＜0.001
18~＜40岁	426(62.83)	998(81.60)
40~＜50岁	157(23.16)	172(14.06)
≥50岁	95(14.01)	53(4.33)
身高[M(P₂₅, P₇₅)，cm]	176(172, 180)	177(173, 181)	4.612^b	0.032
体重[M(P₂₅, P₇₅)，kg]	85(75, 105)	79(70, 87)	86.682^b	＜0.001
体重指数[M(P₂₅, P₇₅)]	27.02(24.39，29.93)	24.97(22.72, 27.46)	117.394^b	＜0.001
缺乏运动[例(%)]			221.097^a	＜0.001
否	267(39.38)	905(74.00)
是	411(60.62)	318(26.00)
失眠[例(%)]			221.006^a	＜0.001
否	244(35.99)	869(71.05)
是	434(64.01)	354(28.95)
吸烟[例(%)]			2.165^a	0.141
是	254(37.46)	417(34.10)
否	424(62.54)	806(65.90)
饮酒[例(%)]			18.744^a	＜0.001
是	250(36.87)	334(27.31)
否	428(63.13)	889(72.69)
高盐饮食[例(%)]			236.188^a	＜0.001
是	398(58.70)	286(23.39)
否	280(41.30)	937(76.61)
高血压家族史[例(%)]			43.109^a	＜0.001
是	201(29.65)	205(16.76)
否	477(70.35)	1 018(83.24)
糖尿病史[例(%)]			23.348^a	＜0.001
是	54(7.96)	37(3.03)
否	624(92.04)	1186(96.97)
冠心病史[例(%)]			31.272^a	＜0.001
是	140(20.65)	137(11.20)
否	538(79.35)	1 086(88.80)
收缩压[M(P₂₅, P₇₅)，mmHg]	150(142, 155)	127(123, 135)	1 195.682^b	＜0.001
舒张压[M(P₂₅, P₇₅)，mmHg]	92(85, 98)	77(72, 82)	827.091^b	＜0.001
HR[M(P₂₅, P₇₅)，次/min]	84(76, 92)	80(73, 86)	54.030^b	＜0.001
WBC[M(P₂₅, P₇₅)，×10⁹/L]	7.11(5.89, 8.07)	6.71(5.62, 7.61)	24.626^b	＜0.001
RBC[M(P₂₅, P₇₅)，×10¹²/L]	5.27(5.00, 5.51)	5.22(4.90, 5.47)	5.840^b	0.016
HB[M(P₂₅, P₇₅)，g/L]	162(156, 168)	160(154, 166)	15.027^b	＜0.001
PLT[M(P₂₅, P₇₅)，×10⁹/L]	251(210, 285)	247(210, 281)	0.734b	0.392
ALT[M(P₂₅, P₇₅)，U/L])	35(19, 40)	30(17, 35)	28.311^b	＜0.001
AST[M(P₂₅, P₇₅)，U/L]	27(21, 30)	25(20, 27)	26.785^b	＜0.001
BUN[M(P₂₅, P₇₅)，μmol/L]	4.93(4.10, 5.60)	4.81(4.06, 5.43)	3.023^b	0.082
Scr[M(P₂₅, P₇₅)，μmol/L]	75.85(69.40, 81.97)	75.90(70.15, 81.60)	0.301^b	0.583
UA[M(P₂₅, P₇₅)，mmol/L]	401(343, 453)	383(332, 430)	19.213^b	＜0.001
FPG[M(P₂₅, P₇₅)，mmol/L]	5.42(4.78, 5.59)	5.11(4.66, 5.29)	-4.984^b	＜0.001
TC[M(P₂₅, P₇₅)，mmol/L]	5.07(4.32, 5.64)	4.74(4.13, 5.30)	45.411^b	＜0.001
TG[M(P₂₅, P₇₅)，mmol/L]	1.90(0.90, 2.09)	1.35(0.72, 1.52)	76.976^b	＜0.001
LDL-C[M(P₂₅, P₇₅)，mmol/L]	2.71(2.19, 3.14)	2.52(2.09, 2.91)	37.125^b	＜0.001
注：^a为χ²值, ^b为Z值, ^c为采用Fisher精确检验。

下载: 导出CSV

表 3 铁路职工高血压患病风险多因素分析

Table 3. Multivariate analysis of the risk of hypertension among railway workers

变量	B	SE	Waldχ²	P值	OR值	95% CI
18岁≤年龄＜40岁	1.315	0.217	36.681	＜0.001	3.726	2.435~5.704
40岁≤年龄＜50岁	0.490	0.246	3.989	0.046	1.633	1.009~2.643
缺乏运动	-1.122	0.120	86.868	＜0.001	0.326	0.257~0.412
失眠	-1.152	0.119	93.286	＜0.001	0.316	0.250~0.399
高盐饮食	1.455	0.121	145.479	＜0.001	4.286	3.383~5.429
高血压家族史	0.640	0.137	21.724	＜0.001	1.897	1.449~2.483
BMI＜18	1.427	0.539	7.010	0.008	4.168	1.449~11.988
18≤BMI＜24	0.733	0.160	20.833	＜0.001	2.080	1.519~2.849
24≤BMI＜28	0.622	0.143	18.954	＜0.001	1.862	1.408~2.464
HR≥70次/min	0.430	0.191	5.058	0.025	1.537	1.057~2.235
TC＞6.5 mmol/L	0.343	0.279	1.510	0.219	1.410	0.815~2.437
TG＞1.8 mmol/L	0.463	0.148	9.827	0.002	1.589	1.190~2.123
LDL-C＞3.1 mmol/L	0.250	0.156	2.581	0.108	1.284	0.946~1.743

下载: 导出CSV

参考文献(24)

[1]	刘明波, 何新叶, 杨晓红, 等. 《中国心血管健康与疾病报告2023》要点解读[J]. 中国心血管杂志, 2024, 29(4): 305-324. LIU M B, HE X Y, YANG X H, et al. Interpretation of report on cardiovascular health and diseases in china 2023[J]. Chinese Journal of Cardiovascular Medicine, 2024, 29(4): 305-324.
[2]	马超, 刘莉, 邓光璞, 等. 基于CHAID模型的广州市某街道高血压患者生活方式影响因素分析[J]. 中华全科医学, 2023, 21(5): 788-791. doi: 10.16766/j.cnki.issn.1674-4152.002984 MA C, LIU L, DENG G P, et al. CHAID model-based analysis of influencing factors of hypertensive patients' lifestyle on a street in Guangzhou[J]. Chinese Journal of General Practice, 2023, 21(5): 788-791. doi: 10.16766/j.cnki.issn.1674-4152.002984
[3]	中国高血压防治指南修订委员会, 高血压联盟(中国), 中华医学会心血管病学分会, 等. 中国高血压防治指南(2018年修订版)[J]. 中国心血管杂志, 2019, 24(1): 24-56. Writing Group of 2018 Chinese Guidelines for the Management of Hypertension, Chinese Hypertension League, Chinese Society of Cardiology, et al. Guidelines for prevention and treatment of hypertension in China (2018 revision)[J]. Chinese Journal of Cardiovascular Medicine, 2019, 24(1): 24-56.
[4]	中华医学会糖尿病学分会. 中国2型糖尿病防治指南(2020年版)[J]. 中华糖尿病杂志, 2021, 10(1): 52-67. Chinese Diabetes Society. Chinese guidelines for the prevention and treatment of type 2 diabetes (2020 edition)[J]. Chinese Journal of Diabetes, 2021, 10(1): 52-67.
[5]	孔博. 北京铁路辖区职工高血压检出率及其危险因素分析[J]. 中国社区医师, 2019, 35(6): 29-32. KONG B. Analysis of hypertension detection rate and risk factors among railway employees in Beijing area[J]. Chinese Community Doctors, 2019, 35(6): 29-32.
[6]	郑立强, 孙肇庆, 张欣刚, 等. Framingham高血压风险预测模型在中国农村人口中的预测价值[J]. 中华高血压杂志, 2014, 22(5): 498. ZHENG L Q, SUN Z Q, ZHANG X G, et al. Predictive value for the rural Chinese population of the Framingham hypertension risk model[J]. Chinese Journal of Hypertension, 2014, 22(5): 498.
[7]	CAREY R M, MORAN A E, WHELTON P K. Treatment of Hypertension[J]. JAMA, 2022, 328(18): 1849-1861.
[8]	UNGER T, BORGHI C, CHARCHAR F, et al. 2020 International society of hypertension global hypertension practice guidelines[J]. Hypertension, 2020, 38(6): 982-1004.
[9]	VANDERBURGH J A, REINHART-KING C A. The role of age-related intimal remodeling and stiffening in atherosclerosis[J]. Adv Pharmacol, 2018, 81: 365-391.
[10]	段晓阳, 李敏, 祁志刚, 等. 高血压家族史和行为危险因素聚集与高血压患病关系[J]. 中国公共卫生, 2022, 38(2): 172-176. DUAN X Y, LI M, QI Z G, et al. Associations of family hypertension history and clustered behavioral risk factors with hypertension among community adult residents in Nantong city[J]. Chinese Journal of Public Health, 2022, 38(2): 172-176.
[11]	LI A L, PENG Q, SHAO Y Q, et al. The interaction on hypertension between family history and diabetes and other risk factors[J]. Sci Rep, 2021, 11(1): 4716. DOI: 10.1038/s41598-021-83589-z.
[12]	ZHANG L, LI J L, ZHANG L L, et al. Association and interaction analysis of body mass index and triglycerides level with blood pressure in elderly individuals in China[J]. Biomed Res Int, 2018, 2018: 8934534. DOI: 10.1155/2018/8934534.
[13]	PEDRALLI M L, MARSCHNER R A, KOLLET D P, et al. Publisher correction: different exercise training modalities produce similar endothelial function improvements in individuals with prehypertension or hypertension: a randomized clinical trial[J]. Sci Rep, 2020, 10(1): 10564. DOI: 10.1038/s41598-020-67586-2.
[14]	HALL M E, COHEN J B, ARD J D, et al. Weight-loss strategies for prevention and treatment of hypertension: a scientific statement from the American heart association[J]. Hypertension, 2021, 78(5): e38-e50.
[15]	HE D, FAN F, JIA J, et al. Lipid profiles and the risk of new-onset hypertension in a Chinese community-based cohort[J]. Nutr Metab Cardiovasc Dis, 2021, 31(3): 911-920.
[16]	赵连成, 武阳丰, 周北凡, 等. 不同体重指数和腰围人群的血压均值及高血压患病率调查[J]. 中华流行病学杂志, 2003, 24(6): 471-475. ZHAO L C, WU Y F, ZHOU B F, et al. Mean level of blood pressure and rate of hypertension among people with different levels of body mass index and waist circumference[J]. Chinese Journal of Epidemiology, 2003, 24(6): 471-475.
[17]	YANG Q, VERNOOIJ R W M, ZHU H, et al. Impact of sodium intake on blood pressure, mortality and major cardiovascular events: an umbrella review of systematic reviews and meta-analyses[J]. Crit Rev Food Sci Nutr, 2024, 12(3): 1-11.
[18]	CHEN C M, ZHAO W, YANG Z, et al. The role of dietary factors in chronic disease control in China[J]. Obes Rev, 2008, 9(7): 100-103.
[19]	张瑜, 王苗云, 徐淑雯. 中等强度有氧联合低强度抗阻运动对老年冠心病合并高血压患者心肺功能的影响[J]. 心电与循环, 2024, 43(5): 508-512. ZHANG Y, WANG M Y, XU S W. Effects of moderate-intensity aerobic exercise combined with low-intensity resistance exercise on lung and heart functions of elderly patients with coronary heart disease and hypertension[J]. Journal of Electrocardiology and Circulation, 2024, 43(5): 508-512.
[20]	郭香. 有氧运动联合渐进性放松训练对老年原发性高血压患者血压控制及生活质量的影响[J]. 中国实用医药, 2024, 19(17): 159-162. GUO X. The effect of aerobic exercise combined with progressive relaxation training on blood pressure control and quality of life in elderly patients with primary hypertension[J]. China Practical Medicine, 2024, 19(17): 159-162.
[21]	邸军军, 沈亮. 嘉兴市某三甲医院体检中老年人群高血压患病的驱动因素模型构建[J]. 中华全科医学, 2023, 21(11): 1880-1883. doi: 10.16766/j.cnki.issn.1674-4152.003248 DI J J, SHEN L. Construction of a driving factor model of hypertension among physical examinatop three hospital in Jiaxing City[J]. Chinese Journal of General Practice, 2023, 21(11): 1880-1883. doi: 10.16766/j.cnki.issn.1674-4152.003248
[22]	HEIN M, LANQUART J P, LOAS G, et al. Risk of high blood pressure associated with objective insomnia and self-reported insomnia complaints in major depression: a study on 703 individuals[J]. Clin Exp Hypertens, 2019, 41(6): 538-547.
[23]	郭一霆, 陈利平, 田秋慧, 等. 高血压患者夜间时段睡眠质量与血压变异性的相关性[J]. 中日友好医院学报, 2021, 35(5): 294-296. GUO Y T, CHEN L P, TIAN Q H, et al. The correlation between sleep quality during nocturnal periods and blood pressure variability in patients with hypertension[J]. Journal of China-Japan Friendship Hospital, 2021, 35(5): 294-296.
[24]	吴月惟, 綦苗苗, 孔月琼, 等. 海南省居民高血压患病风险预测模型建立与验证[J]. 中国预防医学杂志, 2023, 24(2): 93-101. WU Y W, QI M M, KONG Y Q, et al. Construction and validation of a risk prediction model for hypertension among residents in Hainan Province[J]. Chinese Preventive Medicine, 2023, 24(2): 93-101.