老年人糖尿病发病风险模型的建立

发布时间：2018-05-10 16:21

本文选题：老年人 + 糖尿病　；参考：《苏州大学》2015年硕士论文

【摘要】：随着老年人数量及其比重的不断增加,老年人群体的健康日益受到社会的关注。糖尿病是影响老年人健康和生活质量的重要因素。因此,我们研究影响老年人糖尿病发病的因素,在此基础上建立、验证老年人糖尿病发病风险模型。根据所建立的模型确定重点干预人群,进而采取有效的、针对性的干预措施来降低老年人糖尿病发病风险。这对于提高老年人的生活质量、促进社会的发展和进步具有重要意义。目的通过对苏州市相城区渭塘镇血糖正常,无糖尿病发病史及重大疾病的老年人进行为期两年以糖尿病为结局的队列研究,分析影响老年人糖尿病发病风险的临床指标、建立及验证老年人糖尿病发病风险模型。为更好地降低老年人糖尿病发病风险、提高老年人的身体健康和生活质量提供一定的科学依据。方法苏州市相城区第三人民医院开展了针对老年人的免费健康体检活动。我们从老年人健康体检资料中选择基线血糖正常老年人进行为期两年以糖尿病为结局的队列研究,应用相关统计学软件分析老年人糖尿病发病风险影响因素,建立并验证老年人糖尿病发病风险模型。采用的统计学方法主要有单因素logistic回归、t检验、多因素logistic回归、ROC曲线(用来评价模型预测结果)。结果2011年基线血糖正常老年人共有1610人,到2012年有1610人,到2013年有1588人。失访人数(人)为22,比率为1.37%。老年人糖尿病发病影响因素单因素分析显示影响老年人群两年后糖尿病发病风险的因素为甘油三酯偏高、高血压、静息心率过快、超重、肥胖、空腹血糖损害及糖尿病家族史。它们的OR及95%置信区间分别为2.080(1.335,3.241)、2.112(1.309,3.407)、2.098(1.378,3.196)、2.474(1.589,3.851)、2.928(1.445,5.933)、11.281(7.234,17.592)及2.425(1.542,3.811)。胆固醇水平,高密度脂蛋白水平,血红蛋白水平,尿酸及肌酐水平对老年人两年后糖尿病发病风险的影响均无统计学意义。老年人糖尿病发病风险影响因素多因素分析表明影响老年人群两年后糖尿病发病风险的因素有甘油三酯偏高、胆固醇偏高、肥胖、超重、空腹血糖损害和糖尿病家族史。它们的OR和95%置信区间分别为1.788(1.065,3.122)、0.457(0.231,0.903)、2.302(1.041,5.087)、2.164(1.331,3.518)、9.944(6.249,15.825)和1.926(1.179,3.149)。静息心率,低密度脂蛋白水平,高密度脂蛋白水平,血红蛋白水平,尿酸及肌酐水平对老年人两年后糖尿病发病风险的影响均无统计学意义。老年人糖尿病发病风险模型为：将1588名老年人影响因素基线值带入糖尿病发病风险模型求出该老年人群不司个体的糖尿病发生概率。经过ROC曲线分析,结果显示曲线下面积(AUC)为).828,95%置信区间为(0.785,0.872),预测效果较好。结论：1.老年人糖尿病发病风险的影响因素甘油三酯偏高,胆固醇偏高,空腹血糖损害,超重,肥胖,糖尿病家族史。2.所建立的老年人糖尿病发病风险模型合理,能较好地预测老年人糖尿病的发病情况。
[Abstract]:With the increasing of the number and proportion of the elderly, the health of the elderly has been paid more and more attention. Diabetes is an important factor affecting the health and quality of life of the elderly. Therefore, we study the factors that affect the incidence of diabetes in the elderly, and establish a model to verify the risk of diabetes in the elderly. According to the established model, the key intervention population was determined, and then effective and targeted intervention measures were taken to reduce the risk of diabetes in the elderly. This is of great significance for improving the quality of life of the elderly and promoting the development and progress of the society. Objective to analyze the clinical indexes of diabetes risk in elderly patients with normal blood sugar, no history of diabetes and major diseases in Xiangcheng District of Suzhou City, and to conduct a cohort study on the outcome of diabetes mellitus for a period of two years. To establish and verify the risk model of diabetes in the elderly. It provides a scientific basis for reducing the risk of diabetes and improving the health and quality of life of the elderly. Methods the third people's Hospital of Xiangcheng District of Suzhou carried out free health examination for the elderly. We selected baseline blood glucose from the elderly health examination data to conduct a two-year diabetes outcome cohort study, using the relevant statistical software to analyze the risk factors of diabetes in the elderly. To establish and verify the risk model of diabetes in the elderly. The main statistical methods used are single factor logistic regression t test, multivariate logistic regression and ROC curve (used to evaluate the prediction results of the model). Results there were 1610 elderly people with normal blood glucose in 2011, 1610 in 2012 and 1588 in 2013. The number of people who lost the interview was 22, the ratio was 1.37. Univariate analysis showed that the risk factors of diabetes in the elderly were high triglyceride, high blood pressure, fast resting heart rate, overweight, obesity, impaired fasting blood glucose and family history of diabetes. Their OR and 95% confidence intervals were 2.080 / 1.335/ 3.241U / 2.1121.309/ 3.407/ 2.098 / 1.378/ 3.196 / 2.474 / 1.589 / 3.851/ 2.92828 / 1.4455.933 / 11.2817.234/ 17.592and 2.42551.542/ 3.811. respectively. The levels of cholesterol, high density lipoprotein, hemoglobin, uric acid and creatinine had no significant effect on the risk of diabetes in the elderly two years later. The multivariate analysis of the risk factors of diabetes in the elderly showed that the factors affecting the risk of diabetes were high triglyceride, high cholesterol, obesity, overweight, impaired fasting blood glucose and family history of diabetes. Their OR and 95% confidence intervals were 1.788 / 1.0656 / 3.122 / 0.4570.231/ 0.903 / 2.302/ 1.041 / 5.087 / 3.51818 / 9.9444.249/ 15.825) and 1.9261.179/ 3.149/ 3.149respectively. Resting heart rate, low density lipoprotein level, high density lipoprotein level, hemoglobin level, uric acid and creatinine levels had no significant effect on the risk of diabetes in the elderly two years later. The risk model of diabetes in the elderly was as follows: the baseline value of the influencing factors of 1588 elderly people was brought into the diabetes risk model to find out the probability of diabetes in the elderly population. Through the analysis of ROC curve, the result shows that the area under the curve has a confidence interval of 0.7850.872g, and the prediction effect is better. Conclusion 1. The risk factors of diabetes in the elderly were high triglyceride, high cholesterol, impaired fasting blood glucose, overweight, obesity, family history of diabetes. The established risk model of diabetes in the elderly is reasonable and can predict the incidence of diabetes in the elderly.
【学位授予单位】：苏州大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：R587.1

【参考文献】