中性粒细胞与淋巴细胞比值与心血管疾病:基于倾向性评分匹配的队列研究

Neutrophil to lymphocyte ratio and cardiovascular disease:a cohort study with propensity score matching

目的研究中性粒细胞与淋巴细胞比值(NLR)与心血管疾病(CVD)发病风险之间的关系,为CVD早期干预和机制探索提供数据支持。方法于2017年4—5月,采用多阶段整群抽样方法选取天津市两所社区卫生服务中心进行健康体检的4514名无既往CVD病史的40岁及以上居民为研究对象,回顾性收集2010—2017年首次体检时的NLR水平作为基线数据,并前瞻性随访至2019年12月,观察新发CVD情况。基线及随访数据的收集方式包括问卷调查、体格检查及实验室检测。按照基线NLR的第75百分位数将研究人群分为高NLR组(NLR≥2.62)和低NLR组(NLR<2.62)。采用SPSS 26.0进行倾向性评分匹配,采用Cox比例风险回归模型分析NLR与CVD发病风险之间的关系,计算风险比(HR)及其95%可信区间(CI)。采用SAS 9.4进行交互作用分析。结果剔除129名失访者后,共4385名研究对象纳入分析。随访期间(平均随访期为3.1年),新发CVD事件150例,累积发病率为3.42%,发病密度为10.96/1000人年。倾向性评分匹配后Cox比例风险回归模型显示,随着NLR水平升高,CVD发病风险增加(HR=1.13,95%CI:1.05~1.21)。高NLR组发生CVD的风险是低NLR组的2.05倍(HR=2.05,95%CI:1.29~3.23)。交互作用分析发现高NLR与高血压(HR=2.32,95%CI:1.39~3.87)、糖尿病(HR=2.23,95%CI:1.09~4.54)对CVD发病存在相乘交互作用。结论高NLR会增加CVD的风险,且与高血压、糖尿病存在相乘交互作用。

Objective To explore the relationship between Neutrophil to lymphocyte ratio and the risk of cardiovascular disease(CVD),and provide the data support for early intervention and mechanism exploration of CVD.Methods From April to May2017,the multistage cluster sampling method was used to select 4514 residents(≥40 years old)without CVD who underwent physical examination in two community health service centers of Tianjin as the subjects.NLR levels of the first physical examination from 2010 to 2017 were retrospectively collected served as the baseline data,which were prospectively followed up to December 2019 to observe the situation of new CVD.The methods of collecting the baseline data and follow-up data included the questionnaire,physical examination and laboratory test.The subjects were divided into high NLR group(NLR≥2.62)and low NLR group(NLR<2.62)according to the 75 th percentile(P75)of baseline NLR.SPSS 26.0 software was used for propensity score matching.Cox proportional risk regression model was used to analyze the relationship between NLR and CVD risk,and to calculate the risk ratio(HR)and its 95%confidence interval(CI).SAS 9.4 software was used for interaction analysis.Results After excluding 129 missing subjects,a total of 4385 subjects were included in the analysis.During follow-up(average 3.1 years),150 new CVD cases occurred,the cumulative incidence was 3.42%,and the incidence density was 10.96 per 1000 person-years.Cox proportional risk regression model after propensity score matching showed that CVD risk increased with NLR level(HR=1.13,95%CI:1.05~1.21).The risk of CVD in the high NLR group was 2.05 times higher than that in the low NLR group(HR=2.05,95%CI:1.29~3.23).The interaction analysis showed that high NLR and hypertension(HR=2.32,95%CI:1.39~3.87)and diabetes(HR=2.23,95%CI:1.09~4.54)had multiplicative interaction effects on CVD.Conclusion High NLR could increase the risk of CVD,NLR with hypertension or diabetes had a multiplicative interaction.