北京城乡老年人群健康生活方式评分与全因死亡的关联研究

Correlation between healthy lifestyle score and all-cause death in rural and urban elderly of Beijing

目的探讨北京市城乡老年人群健康生活方式评分与全因死亡的关联,为不同居住地老年人健康行为干预提供科学依据。方法数据来源于北京城乡老年人群健康综合研究(BECHCS)。于2009年和2014年采用多阶段整群随机抽样方法分别选取北京市海淀区万寿路街道和密云区巨各庄镇及冯家峪镇60岁及以上4499名老年居民作为研究对象,对其进行问卷调查、体格检查和实验室检测,每2~3年随访1次,随访截至2021年3月31日。根据吸烟情况、饮酒情况、锻炼情况和体型构建健康生活方式评分(0~4分),依据评分分为低风险健康生活方式(4分)、中风险健康生活方式(2~3分)和高风险健康生活方式(0~1分)3组。使用SPSS 26.0和R 4.2.1软件进行t检验、χ2检验及非参数检验,采用Kaplan-Meier法绘制不同健康生活方式评分下城乡老年人群的累积风险曲线,采用log-rank检验法比较不同健康生活方式评分下生存率及生存时间的差异。利用多因素Cox比例风险回归模型分析城乡老年人健康生活方式评分与全因死亡之间的关联,并对单一健康生活方式及交互作用进行效应分析。结果4499名老年居民中农村居民2397人(53.3%),城市居民2102人(46.7%),中位随访时间为6.76年(P25~P75:6.63~11.27年)。健康生活方式评分0~4分的老年居民分别为201人(4.5%)、823人(18.3%)、1442人(32.0%)、1468人(32.6%)和565人(12.6%)。随访期间共667人死亡,死亡密度为182.26/10000人年,其中农村死亡密度为191.06/10000人年,城市死亡密度为177.46/10000人年。多因素Cox回归分析结果显示,每增加1种健康生活方式,总人群全因死亡风险降低13.7%(HR=0.863,95%CI:0.747~0.997),城市老年人全因死亡风险降低24.1%(HR=0.759,95%CI:0.631~0.915),均有统计学意义(P<0.05,P<0.01)。单一健康生活方式Cox回归分析结果显示,总人群中从不吸烟导致全因死亡风险降低26.2%(HR=0.738,95%CI:0.610~0.893,P<0.01)。城市老年人从不吸烟与锻炼导致全因死亡风险分别降低31.0%(HR=0.690,95%CI:0.544~0.875)和24.7%(HR=0.753,95%CI:0.600~0.946),均有统计学意义(P<0.05,P<0.01)。城市老年人中存在吸烟状况与健康体型的相乘交互作用(HR=0.639,95%CI:0.415~0.986,P交互<0.05);农村老年人中存在吸烟状况与锻炼的相乘交互作用(HR=0.557,95%CI:0.343~0.904,P交互<0.05)。结论北京城市老年人群中,低风险健康生活方式评分与全因死亡降低呈正相关,但在农村人群中其关联尚不明确,需进一步进行探究。

Objective To explore the correlation between healthy lifestyle score and all-cause death in rural and urban elderly of Beijing,and provide the scientific basis for health behaviour intervention for elderly in different residence areas.Methods The data were from the Beijing Elderly Comprehensive Health Cohort Study(BECHCS).In 2009 and 2014,the multi-stage cluster random sampling method was used to selecct 4499 elderly residents(≥60 years old)from Wanshou road of Haidian district,Jugezhuang town and fengjiayu town of Miyun district in Beijing as the subjents.The investigation was performed with questionnaires,physical examinations and laboratory testing.Follow-up visits were conducted every 2 to 3 years until March 31,2021.According to smoking,drinking,exercise and physique,healthy lifestyle score(0-4 scores)was established.The healthy lifestyle score was divided into low risk healthy lifestyle group(4 scores),medium-risk healthy lifestyle group(2-3 scores)and high risk healthy lifestyle group(0-1 scores).The t test,χ2 test and non-parametric test were used to analyze the data,the used softwares were SPSS 26.0 and R 4.2.1.Cumulative risk curves of the urban and rural elderly population under different healthy lifestyle scores were drawn by the Kaplan-Meier method.The log-rank test was used to compare survival rates and survival time under different healthy lifestyle scores.The multivariable Cox proportional risk regression model was used to analyze the correlation between healthy lifestyle score and all-cause death of urban and rural elderly residents.Effect analysis was conducted on single healthy lifestyle and interaction.Results Among 4499 elderly residents,numbers of rural residents and urban residents were 2397 cases(53.3%)and 2102 cases(46.7%),respectively.Median follow-up time was 6.76 years(P25-P75:6.63-11.27 years).The numbers of elderly residents with healthy lifestyle score 0-4 scores were 201 cases(4.5%),823 cases(18.3%),1442 cases(32.0%),1468 cases(32.6%)and 565 cases(12.6%),respectively.A total of 667 cases died during follow-up.The death density was 182.26/104 person-years,the rural death density was 191.06/104 person-years,and urban death density was 177.46/104 person-years.Multivariate Cox regression analysis revealed that when one healthy lifestyle increased,the risk of all-cause death in the total elderly reduced by 13.7%(HR=0.863,95%CI:0.747-0.997,P<0.05);the risk of all-cause death in urban elderly decreased by 24.1%(HR=0.759,95%CI:0.631-0.915,P<0.01).The Cox regression analysis of a single healthy lifestyle factor showed that the risk of all-cause death in never smoking subjects decreased by 26.2%(HR=0.738,95%CI:0.610-0.893,P<0.01).The risk of all-cause death in urban never smoking and exercise subjects decreased by 31.0%(HR=0.690,95%CI:0.544-0.875)and 24.7%(HR=0.753,95%CI:0.600-0.946,P<0.05 or P<0.01).There were multiplicative interactions between smoking status and healthy body shape in urban elderly residents(HR=0.639,95%CI:0.415-0.986,P<0.05),and between smoking status and exercise in rural elderly residents status(HR=0.557,95%CI:0.343-0.904,P<0.05).Conclusion The low-risk healthy lifestyle score positively correlated to the reduced all-cause death in the urban elderly residents of Beijing.However,there is no significant corrlation between low-risk healthy lifestyle score and all-cause death.It should take the investigation in further.