慢性疾病风险评分与肿瘤发生风险的队列研究

Association between Chronic Disease Risk Score and Cancer Risk:a Cohort Study

背景肿瘤与其他慢性疾病一样均呈多维状态,多种慢性疾病与肿瘤有着共同的危险因素。单个慢性疾病指标与肿瘤发生风险之间的关联较弱,因此构建由多个慢性疾病指标组成的风险评分并分析其与肿瘤发生风险之间的关联有较大价值。目的探讨慢性疾病风险评分与肿瘤发生风险的关联。方法以“天津市慢病风险与健康管理队列”中2015年1月至2019年12月在天津医科大学总医院健康管理中心进行体检且符合研究标准的18009名体检者为研究对象。所有体检者完成健康风险评估问卷、体格检查、实验室检查指标检测。随访结束于肿瘤事件发生日期或队列随访结束日期(2019-12-31),以最先发生的日期为准。以慢性疾病相关指标[包括体质指数(BMI)、腰围、血压、血糖、总胆固醇、三酰甘油、尿酸、总胆红素、心率及评估肾小球滤过率]为自变量,以肿瘤是否发病为因变量,采用多因素Cox回归模型评估各慢性疾病指标与肿瘤发生风险,并计算慢性疾病风险综合评分。根据慢性疾病风险评分三分位数将体检者分为慢性疾病风险评分低分组(<6分)、中分组(6~8分)和高分组(≥9分)。采用Cox回归分析,探究不同慢性疾病风险评分组体检者肿瘤发生风险。结果本研究共随访71835人年(中位随访4.00年),确诊肿瘤患者91例(男33例、女58例)。与慢性疾病风险评分低分组(<6分,n=6403)相比,慢性疾病风险评分中分组(6~8分,n=6459)和高分组(≥9分,n=5147)体检者肿瘤发生风险的HR(95%CI)值分别为2.16[95%CI(1.20,3.90),P=0.011]和3.08[95%CI(1.72,5.50),P<0.001](P趋势<0.001),满足比例风险假定检验(χ2=1.98,P=0.371);慢性疾病风险评分每增加1分,体检者肿瘤发生风险增加17%[95%CI(9%,25%),P<0.001],满足比例风险假定检验(χ2=0.31,P=0.579)。女性体检者中,与慢性疾病风险评分低分组(<6分)相比,慢性疾病风险评分高分组(≥9分)肿瘤发生风险的HR(95%CI)值为3.00[95%CI(1.32,6.82),P=0.009],满足比例风险假定检验(χ2=1.24,P=0.538);慢性疾病风险评分每增加1分,女性体检者肿瘤发生风险增加21%[95%CI(10%,33%),P<0.001]。结论慢性疾病风险评分越高的体检者肿瘤发生风险越高,且在女性中明显,男性中无此关联。

Background As with other chronic diseases,cancer presents a multidimensional state.And many chronic diseases share common risk factors with cancers.The association between a single chronic disease indicator and the risk of cancer is weak,so it is significant to construct a risk score composed of multiple chronic disease indicators and analyze its association with the risk of cancer.Objective To investigate the association between chronic disease risk score and cancer risk.Methods This study selected a total of 18009 eligible individuals who had participated in Tianjin Chronic Disease Risk and Health Management Cohort Study and undergone physical examination in Health Management Center,Tianjin Medical University General Hospital from January 2015 to December 2019.All subjects completed the health risk assessment questionnaire,physical examination,and laboratory examination.Follow-up ended on the day of cancer occurrence or the day of follow-up termination(December 31,2019).Multivariate Cox regression model was used to evaluate the relationship of the risk of cancer(used as the dependent variable)with each of the chronic disease related indicators[consisting of BMI,waist circumference,blood pressure,blood glucose,total cholesterol,triglyceride,uric acid,total bilirubin,heart rate,and estimated glomerular filtration rate(used as independent variables)].And then the comprehensive score of chronic disease risk was calculated,and the cancer risk in tertile groups of the score[low-score group(<6 points),medium-score group(6-8 points),and high-score group(≥9 points)]was assessed by Cox regression analysis.Results A total of 71835 person-years(median 4.00 years)were followed up,and 91 cases(33 males and 58 females)were diagnosed with cancer.Compared with lowscore group(n=6403),the HR value(95%CI)of cancer risk was 2.16[95%CI(1.20,3.90),P=0.011]in mediumscore group(n=6459),and 3.08[95%CI(1.72,5.50),P<0.001]in high-score group(n=5147)(Ptrend<0.001),which satisfied the proportional hazards hypothesis test(χ~2=1.98,P=0.371).For per point increase in the chronic disease risk score,the risk of cancer increased by 17%[95%CI(9%,25%),P<0.001],which satisfied the proportional hazards hypothesis test(χ~2=0.31,P=0.579).Compared with females in low-score group,the females in high-score group had an increased risk of cancer,with HR value(95%CI)of 3.00[95%CI(1.32,6.82),P=0.009],which satisfied the proportional hazards hypothesis test(χ~2=1.24,P=0.538).For per point increase in the chronic disease risk score,the risk of cancer in females increased by 21%[95%CI(10%,33%),P<0.001].Conclusion The higher the chronic disease risk score,the higher cancer risk,and the association was significant in females,but not in males.