基于广义估计方程和潜变量增长曲线模型的阿尔茨海默病健康相关生命质量动态变化研究

发布时间：2018-08-04 17:00

【摘要】：目的将健康相关生命质量（health related quality of life，HRQOL，又称生命质量、生活质量）作为主要结局指标，在前期研究基础上，继续追踪观察，并借助恰当的纵向数据统计分析模型，对阿尔茨海默病进程进行较全面的定量研究。将阿尔茨海默病(Alzheimer’sdisease，AD)进程中的疾病状态和生命质量相结合，研究阿尔茨海默病对生命质量的影响、生命质量的主要影响因素、生命质量随状态转移的动态变化规律等。为社区老年人AD预防、早期干预，制定不同时期AD防治对策，从而提高生命质量提供理论依据。同时探讨纵向数据统计分析模型——广义估计方程和潜变量增长模型在纵向数据分析中的优势。方法抽取太原市三个社区的517名60岁及以上老年人作为研究对象，对其进行随访调查，调查间隔期为6个月，共获得3次随访的资料。针对3次随访的纵向资料，引入广义估计方程和潜变量增长模型，以认知正常老化为状态1、轻度认知障碍为状态2、中重度认知障碍为状态3、阿尔茨海默病为状态4，重点研究AD进程中健康相关生命质量随状态转移的动态变化规律及其影响因素。结果广义估计方程分析结果：对老年人健康相关生命质量有统计学意义的影响因素是夫妻关系、与子女关系、听力、做家务情况、参加体育锻炼、喝茶、饮酒、一级亲属痴呆、高血脂、MOCA、ADL、GDS。从各种认知转归结局生命质量的影响因素比较分析结果来看：对1组（状态1→状态2）健康相关生命质量有统计学意义的影响因素是教育水平、离休前职业、业余爱好、喝茶、吸烟、铝制炊具使用、控制食量、脑部疾患、ADL、GDS。对2组（状态2→状态3）有统计学意义的影响因素是婚姻、夫妻关系、出生胎次、参加体育锻炼、一级亲属痴呆、ADL、GDS。对3组（状态3→状态4）有统计学意义的影响因素是教育水平、退休后第二职业、做家务情况、参加体育锻炼、喝茶、一级亲属痴呆、高血脂、GDS。对4组（状态3→状态2）有统计学意义的影响因素是离休前职业、婚姻、参加娱乐公益活动、吸烟、GDS。潜变量增长模型分析结果：总的来说，初始状态平均生命质量为29.865，处于一般水平，并且存在显著的个体差异，在随访的这一段时间，生命质量有先升高后下降的趋势，但变化速度却不存在显著的个体间差异，而且初始时生命质量的高低与后来的变化速度之间相关不显著。各转移组初始状态生命质量存在着显著的差异，，但2组（状态2→状态3）和3组（状态3→状态4）这两个转移组老年人生命质量有明显的先升高后下降趋势，而且2组升高要较3组明显；生命质量的发展速度存在显著的个体之间的差异。而1组（状态1→状态2）和4组（状态3→状态2）这两个转移组均未达到显著水平。结论AD进程中健康相关生命质量随状态转移的动态变化规律各有特点并且与多种因素相关，为更好地预防AD的发生，应该根据不同认知转归结局各自的发展变化及其影响因素，开展疾病分阶段重点防治。广义估计方程和潜变量增长模型在纵向数据的分析中可以分别考虑到数据在个体内的相关性和个体间差异，在应用中显示出其特有的优势。
[Abstract]:Objective to keep track of the health related quality of life (health related quality of life, HRQOL, quality of life and quality of life) on the basis of previous studies, and to carry out a comprehensive quantitative study on the Alzheimer's disease process with the help of appropriate longitudinal data statistical analysis model. Blzheimer The disease (Alzheimer 'sdisease, AD) process is combined with the quality of life to study the effect of Alzheimer's disease on the quality of life, the main influencing factors of the quality of life, the dynamic change of the quality of life with the state, and so on. For the elderly people in the community, the prevention of AD, the early intervention, and the formulation of the Countermeasures for the prevention and control of the AD in different periods, thus improve the prevention and control of AD. A theoretical basis for the quality of life is provided. At the same time, the advantages of the longitudinal data statistical analysis model, the generalized estimation equation and the latent variable growth model, are discussed in the longitudinal data analysis.
Methods a total of 517 people aged 60 and above in three communities in Taiyuan were selected as the research subjects. They were followed up with a follow-up survey of 6 months. A total of 3 follow-up data were obtained. According to the longitudinal data of the 3 follow-up, the generalized estimation equation and the latent variable growth model were introduced in order to recognize the normal aging condition 1 and the mild cognitive impairment. State 2, moderate to severe cognitive impairment was 3, Alzheimer's disease was 4. The dynamic changes of health related quality of life with state transfer in the AD process and its influencing factors were studied.
Results the results of generalized estimation equation: the influence factors on the health related quality of life of the elderly were marital relationship, relationship with children, hearing, doing housework, drinking tea, drinking, first-degree kinship dementia, hyperlipidemia, MOCA, ADL, GDS. from the comparison of the factors of life quality of various cognitive outcomes. The results showed that the 1 groups (state 1 to state 2) had statistical significance in health related life quality, such as educational level, pre occupation, hobby, tea drinking, smoking, aluminum cooking utensils, control of food, brain disease, ADL, and GDS., which had statistically significant factors in 2 groups (state 2 to 3) were marital, marital relations, Birth birth, participation in physical exercise, first-degree kinship dementia, ADL, and GDS. to 3 groups (state 3 to 4) were statistically significant factors affecting the level of education, the second career after retirement, doing housework, taking part in physical exercise, drinking tea, first-degree relatives dementia, hyperlipidemia, and GDS. on 4 groups (status 3 to state 2) with statistical significance Former occupation, marriage, entertainment and public welfare activities, smoking, GDS.
The analysis results of the latent variable growth model: in general, the average quality of life in the initial state is 29.865, at the general level, and there is a significant individual difference. In this period of follow-up, the quality of life rises first and then decreases, but the rate of change does not exist between the individuals and the high quality of life at the beginning. There were significant differences in the quality of life in the initial state of the transfer groups, but the 2 groups (state 2, state 3) and 3 groups (state 3 > state 4) had a significant increase in the quality of life in the two groups, and the 2 groups were significantly higher than the 3 groups; the speed of the development of the quality of life was faster. There was a significant difference in degree between individuals, but neither group 1 (state 1 state 2) nor group 4 (state 3 state 2) reached a significant level.
Conclusion the dynamic changes of the health related quality of life with state transfer in the AD process have their own characteristics and are related to various factors. In order to prevent the occurrence of AD better, it should be based on the development and changes of different cognitive outcomes and its influencing factors, and carry out the prevention and control of the disease in stages. The generalized estimation equation and the latent variable growth model should be carried out. In the analysis of longitudinal data, we can consider the relativity of data within individuals and the difference between individuals, and show its unique advantages in application.
【学位授予单位】：山西医科大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：R749.16

【参考文献】