老年人认知损害与抑郁症共病自然史多状态Markov模型及疾病负担研究

Cognitive impairment and late-onset depression are two major public health issues. They not only endanger health of the elderly, but put burdens heavily on patients, caregivers, societies and health care systems as well. Usually, these two common disease conditions occur frequently and simultaneously among the elderly. The relationship of mild cognitive impairment (MCI), Alzheimer's disease (AD) and late-onset depression is still complicated and not well understood until now. In addition, there are no quantitative research findings on the disease burden imposed by cognitive impairment and depression co-existing. To explore the relationship between cognitive impairment and depression and disease burden will contribute to clinical diagnosis and early intervention, social service systems burden decrease, and quality of life improvement among the elderly. Multi-state Markov model, which has great advantages in dynamically evaluating disease progression of chronic diseases with multi-states and multi-stages, can be used to imitate the natural history of disease. We have successfully applied multi-state time-homogeneous Markov model to MCI progression and have done some researches about disease burden imposed by AD. Considering multi-states, multi-stages of cognitive impairment and depression co-existing, we come to a conclusion that multi-state Markov model is proved to be particularly useful and is the best way to analyze the complex relationship between those states of disease process. Based on four years cohort follow-up among the community-dwelling elderly people, multi-state homogeneous Markov model and disease burden of AD, the purposes of this study are: (1) to fit piecewise homogeneous model and time-transformation model if transition intensities are time-dependent; (2) to apply semi-competing risk model considering that dementia and death are two competing risk events in which the death times are known for demented subjects but clearly for subjects who died before dementia no dementia times are known; and (3) to clarify quantitatively group, economic, caregivers and family burden imposed by cognitive impairment and depression co-occurrence. This study will provide scientific basis for strategies and health policies of prevention and early intervention for cognitive impairment and depression co-existing. Accordingly, systematic research methods about natural history and burden for multi-stage diseases could be developed.

痴呆与抑郁症是严重危害老年人生命健康并造成极大负担的重大公共卫生问题，二者通常合并发生。轻度认知损害（MCI）、阿尔茨海默病（AD）和抑郁症三者之间关系尚无定论；由认知损害与抑郁症共病造成的疾病负担定量研究未见报道。探讨认知损害与抑郁症关系及其疾病负担有助于临床诊断和及时干预，减轻社会服务系统负担，提高老年人口生命质量。在前期社区老年人MCI转归多状态齐次Markov模型和AD疾病负担研究基础上，本项目对研究队列进行4年随访观察，研究内容包括：拟合针对转移强度与时间有关的分段齐次Markov模型和时间转换Markov模型；针对死亡和痴呆构成竞争风险，研究半竞争风险模型在认知损害自然史中的应用；定量评估认知损害与抑郁症共病造成的群体、经济、照料者和家庭负担。研究结果可为认知损害和抑郁症共病预防和早期干预策略及其卫生政策制定提供科学依据，并藉此形成系统的多状态进程疾病自然史和疾病负担研究方法。

痴呆症与抑郁症是严重危害老年人生命健康并造成极大负担的重大公共卫生问题，二者通常合并发生。探讨认知损害与抑郁症关系及其疾病负担有助于临床诊断和及时干预，减轻社会服务系统负担，提高老年人口生命质量。在前期社区老年人轻度认知损害（mild cognitive impairment，MCI）转归多状态齐次Markov模型和阿尔茨海默病(Alzheimer’s disease, AD)疾病负担研究基础上，本项目对研究队列进行4年随访观察，研究内容包括：.（1）多状态模型在AD进程自然史研究中的应用：基于太原市600名社区老年人的队列随访资料，分别将非齐次Markov模型和时间转换Markov模型（适用于数据满足Markov性，不满足时齐性的情况）、竞争风险模型（以MCI作为暂态，AD与发生AD前死亡分别作为两个吸收态，其中发生AD前死亡为AD的竞争风险事件）和特定转移Cox模型应用于MCI向AD转归过程，结果表明AD进程不同疾病状态间的转移具有不同的影响因素，应据此开展疾病分状态的重点防治工作。.（2）认知损害与抑郁症关系探讨：多水平模型、贝叶斯半参数多状态模型和贝叶斯网络模型结果表明，老年抑郁可能会加剧老年人认知功能下降，有直接因果关系，老年认知损害与老年抑郁的发生发展有关。.（3）认知损害合并抑郁的疾病负担评价：将疾病病程与其费用相结合，构建认知损害合并抑郁疾病进展的Markov模型，结果表明认知损害合并抑郁症组的疾病发展结局及远期经济负担均重于仅认知损害组。抑郁可能加重了认知损害的疾病经济负担，加快了疾病进展速度。.（4）AD患者及其照料者负担评价：探讨了阿尔茨海默病患者因素与照料者因素可能对照料者负担产生的直接或间接影响，揭示了社会支持的调节效应和照料者积极感受的中介效应。.研究结果可为认知损害和抑郁症共病预防和早期干预策略及其卫生政策制定提供科学依据，并藉此形成系统的多状态进程疾病自然史和疾病负担研究方法。