Abstract
Background:
Alzheimer’s disease (AD) exerts a heavy burden on China. Substantial factors are found associated with high expenditure of AD in high-income countries. To date, few studies have been conducted in China.
Objective:
This study aimed to analyze the associated factors of the total annual costs of AD in China.
Methods:
Data were drawn from a multi-center, cross-sectional, socioeconomic study on the costs of AD conducted in China from October 2015 to March 2016. Generalized linear model (GLM) using gamma distribution with a log-link function was employed to examine the associated factors of the total cost.
Results:
Univariate analysis showed that the demographic and clinical characteristics of AD patients and their caregivers had a substantial impact on the total cost. In GLM analysis, age, monthly household income, AD severity, number of comorbidities, and treatment with memantine were associated with higher expenditure, while the use of a nursing home/care facility was associated with lower expenditure. The mean annual costs for patients with severe dementia were almost twice as high as those for patients with mild dementia (US$ 25,601 versus US$ 13,387, p < 0.001). The mean total cost of AD patients with at least five comorbidities (US$ 38,348) was almost three times than those with no comorbidities (US$ 13,744).
Conclusion:
In China, AD severity and comorbidities were the most critical factors impacting the total cost. Optimizing care patterns, delaying disease progression, and managing comorbidities comprehensively could decrease the heavy burden of AD.
INTRODUCTION
Alzheimer’s disease (AD) is the most common type of dementia. It was reported that 50 million people lived with dementia in 2018 worldwide, and the total cost of dementia was US$ 1 trillion [1]. In China, there were 8.75 million AD patients in 2015 [2], and with the rapid growth of an aging population, it is expected to exceed 20 million by 2050 [3]. According to our recent study, the annual socioeconomic cost of per AD patient in China was US$ 19,144, and the total cost was US$ 167.74 billion in 2015 [4]. Therefore, AD exerts a dramatically heavy economic burden and societal impact on China [5].
AD tends to have a progressive and long-term course, and substantial factors such as AD severity and comorbidities are associated with high expenditure. AD patients are prone to have a greater number of comorbidities. It has been reported that more than 70% of AD patients suffer from at least one comorbidity [6,7, 6,7]. Due to the aggregation and complexity of the comorbidities, they are inclined to decrease health-related quality of life, aggravate care burden, and increase socioeconomic costs. Most studies on the associated factors of AD costs have been conducted in high-income countries, and the results are inconsistent and conflicting [8], probably due to the differences in the health-care system and cultural characteristics. Thus, it was necessary to analyze the country-specific impact factors of the costs of AD. In 2015, we launched a nation-wide, multi-center, cross-sectional study to explore the socioeconomic burden of AD in China, but we did not analyze the associated factors. To our knowledge, few such studies have been conducted in China to date.
To address this issue, we explored the associated factors of total cost of AD on the basis of our previous study. The results will provide an insight into the possible impact factors that induce high costs, and provide substantial evidence for formulating a health policy to reduce the enormous expenditure of AD.
METHODS
Study design and population
This study was conducted in 30 provincial, municipal, and autonomous regions in mainland China from October 2015 to March 2016 [9]. All participants were aged ≥60 years, and the inclusion criteria were as follows: 1) a primary diagnosis of AD according to the National Institute of Neurologic and Communicative Disorders and Stroke and the Alzheimer’s Disease and Related Disorders Association criteria (NINCDS-ADRDA) [10]; 2) complete information about the costs of AD and comorbidities, obtained from electronic medical records systems and face-to-face interviews. The data source covered 81 research centers, including five different types of health facilities (Tier 3 hospital, psychiatric hospital/mental health center, gerontology hospital, nursing home/care facility, and urban/rural communities). We included eastern/central/western geographic areas to represent different economic levels of China. Details of the study design have been provided in our previous published papers [4, 9].
Characteristics of AD patients and caregivers
The demographics of AD patients including gender, age, marital status, number of children, educational level, occupation, living location (urban/rural), geographic area, and monthly household income were recorded. The demographics of the caregivers included gender, age, relationship of the caregiver to the AD patient, and number of caregivers. For caregivers, self-reported diagnosis of insomnia, depression, and anxiety were recorded as outcomes of caregiving or as conditions exacerbated by caregiving.
Measurements in the survey
The clinical characteristics of AD included disease course (time since AD diagnosis), Mini-Mental State Examination (MMSE), symptomatic treatment (donepezil and memantine), and the behavioral and psychological symptoms of dementia (BPSD). The severity of AD was assessed according to the MMSE (for literate dementia: mild: 21–24 points; moderate: 11–20; severe: ≤10; for illiterate dementia: mild: 16–19 points; moderate: 8–15; severe: ≤7) [11]. BPSD included delusions, hallucinations, depression, anxiety, irritability, agitation/aggression, disinhibition, euphoria, aberrant motor behavior, apathy, and abnormalities in sleep and appetite.
Comorbidities
Comorbidities were recorded from the electronic medical records systems and residence health systems at each site, and face-to-face interviews were performed when necessary. In total, nine chronic comorbidities were recorded. Ischemia stroke and lacunar stroke were included as “stroke”. Type 1 and type 2 diabetes mellitus were included as “diabetes”. Coronary heart disease, rhythm disturbances, and other chronic heart conditions were included as “heart disease”. Hypertension with all stages was included as “hypertension”. High levels of cholesterol and/or low-density lipoprotein cholesterol (LDL-C) and/or triglyceride were included as “hyperlipidemia”. Chronic obstructive pulmonary disease, asthma, and other chronic lung diseases were included as “lung disease”. All types of cancers were included as “cancer”. Osteoarthritis and other types of arthritis were included as “arthritis”. Parkinson’s disease, epilepsy, and so on were included as “other neurological disorders”.
Cost estimation
The total cost per patient per year was determined by subgroup costs, including the annual direct medical costs, direct non-medical costs, and indirect costs for each patient [4, 9]. Direct medical costs consisted of outpatient costs, hospitalization costs, and out-of-pocket expenses for health care and medications. Direct non-medical costs included the cost of transportation, accommodation, and meals when visiting a physician; the cost of nourishment and health-care equipment in the patient’s daily life; and formal care fees at the nursing home, care facility, or at home. Indirect costs included monetary losses caused by the patient’s inability to work, reduction in the informal caregivers’ income, and intangible costs, which included treatment of mental suffering of caregivers and unexpected injuries in patients with AD or their caregivers. (Exchange rate: US$ 1≈6.4 RMB; Bank of China, December 2015 [12, 13]).
Statistical analysis
Descriptive statistics were used to summarize patient characteristics including variables of gender, age, marital status, number of children, education level, occupation, monthly household income, living location (urban/rural), geographic area, health facility, AD course, number of comorbidities, MMSE, intake of memantine or donepezil, and BPSD. We also examined variables related to caregivers including age, gender, number of caregivers, caregiver’s relationship to the patient, and caregiver diagnosis of insomnia, and psychiatric problems. Total cost and cost components were presented as mean (standard deviations, SD) for each variable.
First, we conducted a univariate analysis to examine the association with each variable and total annual cost. Nonparametric statistical methods were used considering the characteristics of cost data including 1) non-negative costs, 2) excess zeroes from individuals not using healthcare services during a particular period, 3) heavy right tails, and 4) substantial right-skewness. Statistical significance for univariate analysis was tested by the Mann-Whitney U test or Kruskal-Wallis test.
Second, for the multivariate analysis, generalized linear model (GLM) were developed using gamma distribution with a log-link function. Models were established with total cost as the dependent variable and independent variables selected, with p value less than 0.1 in univariate analysis, with patient age and gender forced entry method.
Finally, we confirmed the significant factors related to the total cost, with p < 0.05. The exponential of coefficients [Exp (B)] with 95% confidence interval (CI) were reported for each variable. Besides, we also explored the interaction effects of comorbidities and AD severity. All analyses were performed using the SPSS 22 software (IBM Corp. Released 2013. IBM SPSS Statistics for Windows, Version 22.0. Armonk, NY, USA).
RESULTS
Patient and caregiver characteristics
Of the 3,098 enrolled participants with AD, 3,046 patients had provided complete information with regard to the demographic characteristics described in Table 1. The average cost per patient per year was US$ 19,144 (RMB 122,523). The mean age of the AD patients was 75.27±9.40 years, and 54.20% of the patients were women. The average score of MMSE was 13.76±9.14. There were 2.62 caregivers per AD patient, and the mean age of the caregivers was 63.26±13.02 years.
Demographic Characteristics and Mean Annually Total Costs (US$)
Univariate analysis demonstrated that health facility, gender, educational level, job nature, household income, living region, geographic area, AD severity, number of comorbidities, intake of memantine or donepezil, BPSD, number of caregivers and caregiver with insomnia and/or psychiatry problem were associated with the total cost (p < 0.05); marital status and number of children, caregiver’s relationship to AD patient and caregiver’s gender were not associated with the total cost (p > 0.05). AD, Alzheimer’s disease; BPSD, behavioral and psychological symptoms of dementia.
Univariate analysis of factors associated with costs
Univariate analysis showed that the demographic and clinical characteristics of AD patients and their caregivers had a substantial impact on the total cost. AD patients who were male, with high educational levels, with high household income, working mentally, and living in urban regions and in eastern China were significantly associated with a high total cost (p < 0.001). Regarding medication, the total cost for a donepezil (a cholinesterase inhibitor) user was lower than that for a non-user by US$ 928. The total cost for a donepezil user (US$ 17,477) was lower than that for a memantine user (US$ 26,941) (p < 0.001). BPSD was recorded in 12.57% of AD patients, with significantly higher expenditure (US$ 24,884) than without (US$ 18,319). The total cost for AD patients with one caregiver (US$ 19,167) was higher than that for those without a caregiver (US$ 14,829). As 10.24% of caregivers had insomnia and psychiatric problems, it resulted in a more substantial economic burden of US$ 3,478 per year, unlike with healthy caregivers (Table 1).
Multivariate analysis of factors associated with costs
When GLM was used and other covariates were controlled, age, health facility, geographic area, monthly household income, AD severity, number of comorbidities, and treatment with memantine continued to affect the expenditure (Table 2). Patients from a nursing home/care facility (Exp (B): 95 CI 0.665 : 0.511–0.84) and from central China (Exp (B): 95% CI 0.796 : 0.684–0.925) had to bear lower total costs; while patients with a monthly household income of >10,000 RMB (US$ 1,563) (Exp (B): 95% CI 1.717 : 1.224–2.409) had to bear a higher total cost. AD patients recruited from the nursing homes/care facilities bore the lowest costs (US$ 12,383), which was almost one-third of the total cost for those recruited from gerontology hospitals (US$ 35,157). The subgroup costs were significantly different in patients recruited from health facilities (Table 3). Memantine (an N-methyl-D-aspartate receptor agonist) was used by 30.50% of AD patients, which resulted in significantly higher costs of US$ 8,028 per year.
Multivariate analysis of main factors associated with total costs of AD patient
Multivariate analysis demonstrated that health facility, age, household income, living region, geographic area, AD severity, number of comorbidities and intake of memantine were associated with the total cost (p < 0.05); while gender, educational level, job nature, intake of donepezil, BPSD, numbers of caregivers, caregiver’s age and caregiver with insomnia and/or psychiatry problems were not associated with the total cost (p > 0.05). AD, Alzheimer’s disease; BPSD, behavioral and psychological symptoms of dementia; Exp(B), Exponential of coefficients; CI, confidence intervals; SE, standard error.
Mean subgroup costs per year by health facility (US$)
Subgroup costs included direct medical cost, direct non-medical cost and total indirect cost. The direct medical cost, direct non-medical cost and total indirect cost were significantly different among five different health facilities respectively (p = 0.000). SD, standard deviations.
The association of AD severity and comorbidities with total cost
AD severity and the number of comorbidities were the important factors associated with higher expenditure across all levels independently. There were 1.62±1.57 comorbidities per person, and 70.29% of AD patients had at least one comorbidity. The Exp (B) of the number of comorbidities ranged from 1.203 (95% CI: 1.031–1.405) for one comorbidity to 1.954 (95% CI: 1.463–2.612) for at least five comorbidities (Table 2). Estimated coefficients were much more significant for more comorbidities, suggesting a substantial increase in costs depending on the number of comorbidities. The mean total cost of AD in patients with no comorbidities was US$ 13,744, and it was US$ 38,348 for those with at least five comorbidities, which was almost three times the cost of the former case. Direct medical costs contributed to the increasing costs due to comorbidities (Table 4). For each additional comorbidity, the mean total cost increased by 18%, and it mainly comprised of direct medical costs.
The mean subgroup costs per year by number of comorbidities (US$)
Subgroup costs included direct medical cost, direct non-medical cost and total indirect cost. Direct medical cost and direct non-medical cost were significantly increased with the number of comorbidity (p = 0.000). The indirect medical cost was non-significant different among different number of comorbidity (p = 0.340). AD, Alzheimer’s disease; SD, standard deviation.
AD severity was found to have a significant association with total cost (p < 0.001), which was U$ 13,597 for mild, US$ 16,789 for moderate, and US$ 26,001 for severe AD. Within the different AD severity groups, expenditure was significantly higher for those with more comorbidities (p = 0.000). In mild AD, costs ranged from US$ 10,947 in those with no comorbidities to US$ 25,094 in those with at least five comorbidities. In moderate AD, costs ranged from US$ 11,554 in those with no comorbidities to US$ 28,848 in those with at least five comorbidities. In severe AD, costs ranged from US$ 20,319 in those with no comorbidities to US$ 48,241 in those with at least five comorbidities (Table 5). The interaction terms between the number of comorbidities and AD severity were statistically non-significant.
Mean total costs per year by AD severity and the number of comorbidities (US$)
Within each AD severity group, total costs per year were significantly higher for those with more comorbidities (p = 0.000). AD, Alzheimer’s disease; SD, standard deviation.
As for the different comorbidities, hypertension (40.71%) was the most frequent comorbidity, followed by stroke (23.01%), heart disease (20.29%), diabetes mellitus (18.91%), and hyperlipidemia (16.09%). All costs were significantly associated with different comorbidities (p < 0.01), except cancer, arthritis, and other neurological disorders. AD patients with lung disease had the highest total cost per year (US$ 31,382) with Exp(B) 1.744 (95% CI: 1.347–2.259) (Table 6).
Prevalence and mean total cost per year by comorbidity (US$)
All cost differences are significantly associated with each comorbidity (p < 0.05), except for cancer, arthritis and other neurological disorders. Exp (B): Exponential of coefficients; CI: confidence intervals; SD, standard deviation; Neurological disorders refer to other neurological disorders.
DISCUSSION
The main strength of the study lies in the data source we used. It was the first multi-center, nation-wide study to analyze the associated factors of the total cost of AD in China. Our study enrolled three geographic areas to represent economic disparities and five types of health facilities to represent various disease stages and healthcare practices. Therefore, it was significantly advantageous to comprehensively evaluate the cost associated factors in China. AD severity and comorbidity were strongly associated with the total cost. Optimizing care patterns, delaying disease progression, and controlling comorbidities properly could contribute to decreasing the heavy burden of AD in China.
In our study, the data source from different health facilities had a substantial impact on the costs. Patients from a nursing home or care facility had the lowest expenditure, mainly because of the lowest indirect costs. Many studies worldwide demonstrated that indirect cost was the primary driver of the AD burden. The professional care from a nursing home/care facility could help to decrease the informal care cost, which was an optimal substitute for home care by family members. In China, care for the elderly was traditionally considered the responsibility of the family. However, for AD patients, living in a nursing home or care facility might be cost-saving. Nowadays, China has a shortage of professional care facilities, and the establishment of care facilities/nursing homes and formulation of an appropriate policy to encourage AD patients to stay in these could reduce the care burden.
We found that AD severity was significantly correlated to the total cost, which was similar to the results of other studies. The total cost ranged from US$ 16,236–39,000 annually for mild AD to US$ 32,388–82,000 for severe AD, depending on patient selection and different calculation methods [14 –17]. In most cases, costs for severe dementia were more than double that those for mild dementia[18]. Every 3-point decrease in MMSE was associated with a 12.5% increase in the total cost [19], and the increasing indirect costs could be the main reason. With the progression of AD, cognitive impairment worsened, and activity of daily function decreased, inducing more intensive and longer care time, which was a heavy burden on caregivers. Thus, maintaining AD patients at milder stages as far as possible and reducing informal care time may decrease the total cost.
AD patients tend to have multiple comorbidities which exert a substantial impact on total cost. A recent study in China found that the comorbidity burden was high with AD and other types of dementia, and were associated with more frequent and longer hospitalization, and high daily expenses [20]. A similar study in the USA indicated that expenditure was ten times higher for those with at least three comorbidities than for those without comorbidities [21]. In our study, the total cost for AD patients with at least five comorbidities was almost three times that for those without comorbidities. With each additional comorbidity, the mean total cost increased by 18% annually. In case of patients with at least three comorbidities, the total cost will be increased over 35% of our estimate of US $167.74 billion annually [4]. We found that direct cost, other than indirect costs, contributed to the increasing economic burden due to a greater number of comorbidities. AD patients are liable to have communication difficulties with doctors and poor compliance with medical therapies, which may exaggerate the complexity of treatment, resulting in the need to adopt conservative therapy leading to poor prognosis. Maintaining a healthy lifestyle, routine screening, and comprehensive treatment of comorbidities may facilitate the reduction of the burden on AD patients.
Different comorbidities significantly contribute to the total cost of AD patients. We found that lung disease, stroke, and heart disease were the first three high-cost comorbidities. Lung disease had the first noticeable impact, which was beyond our expectations. We found that 9.2% of lung diseases involved acute exacerbation, which was caused by the failure to manage chronic conditions appropriately (Supplementary Table 1). Such aggravation of chronic comorbidities compromised the quality of life of both the patients and the caregivers, and resulted in more emergency visits to the hospital, more frequent and extended hospitalizations, and more professional and longer nursing care, which had a serious influence on the health condition and a significant impact on expenditure. Managing comorbidities properly and minimizing acute exacerbations as far as possible may reduce the total cost of AD patients.
In our previous study, pharmacological treatment took up 32.51% of total cost and was associated with AD burden [4]. Treatment with donepezil and memantine was reported to be cost-effective in longitudinal pharmacologic economic studies by significantly decreasing the cost of caregiving [18 , 22–26]. On the contrary, we found that memantine was associated with high costs for AD patients, which was due to the higher proportion of patients with severe AD rather than direct medical costs. This is a cross-sectional study, and it could not capture the longitudinal effect of the medicine on total cost. Thus, the association of medicines with AD costs needs further longitudinal investigation.
BPSD is a common symptom of AD, which could result in a poor quality of life, and be a considerable physical and psychiatric burden on caregivers [27 –30]. BPSD did not have a significant association with the total cost in our study, which was controversial to the results of other studies [16 , 31]. The simple recordings of BSPD may explain this. A further in-depth study is needed to evaluate the influence of BPSD on indirect costs.
In our previous report, informal care costs composed a substantial part of the total cost, and AD caregivers bore the social and economic strain for a long term. However, we did not find an association between caregivers’ demographic characteristics and total cost. Thus, we should pay more attention to the caregivers’ burden and health condition in the future studies.
There are some limitations to our study. We only assessed the number of comorbidities and nine common chronic comorbidities in this study. We did not record the severity of the comorbidity and other chronic comorbidities such as chronic kidney disease, etc. The quantified method should be used to evaluate comorbidities entirely and systematically. Additionally, to overcome the common limitation of all cross-sectional studies, we need to conduct a longitudinal study to further evaluate the associated factors.
CONCLUSIONS
In this multi-center, cross-sectional study conducted in China, we found that AD severity and comorbidity conditions had a significant impact on the total cost of AD. Reduction of the heavy burden of AD care may be brought about by optimizing care patterns, delaying disease progression, and managing comorbidities properly.
Footnotes
ACKNOWLEDGMENTS
We thank all investigators of the 81 research centers: Daojun Xie, The First Affiliated Hospital of Anhui Medical University; Jihui Lv, Beijing Geriatric Hospital; Xiaomei Meng, Beijing Haidian Hospital (Haidian Hospital of No. 3 Hospital of Peking University); Jing Gao, Peking Union Medical College Hospital; Boyan Fang, Aerospace Center Hospital; Yue Wang, Beijing Chaoyang Hospital, Capital Medical University; Yanjun Guo, Beijing Friendship Hospital, Capital Medical University; Fang Li, Fuxing Hospital, Capital Medical University; Jianping Jia, Xuan Wu Hospital, Capital Medical University; Dantao Peng, China–Japan Friendship Hospital; Yongan Sun, Peking University First Hospital; Ning Wang, The First Affiliated Hospital of Fujian Medical University; Xiaodong Pan, Fujian Medical University Union Hospital; Renjing Zhu, Zhongshan Hospital Xiamen University; Yi Zhang, Gansu Provincial Hospital; Haiqun Xie, Foshan Hospital Affiliated to Sun Yat-sen University; Shuwen Xu, Guangdong General Hospital; Yingjun Ouyang, Guangzhou First People’s Hospital; Muni Tang, Guangzhou Huiai Hospital; Cansheng Zhu, The Third Affiliated Hospital, Sun Yat-sen University; Jun Liu, Sun Yat-sen Memorial Hospital, Sun Yat-sen University; Shengliang Shi, The First Affiliated Hospital of Guangxi Medical University; Caiyou Hu, The Guangxi Zhuang Autonomous Region Jiangbin Hospital; Lan Chu, The Affiliated Hospital of Guizhou Medical University; Guilan Lu, Hainan Province Nongken General Hospital; Guoqiang Wen, Hainan General Hospital; Xifa Lan, First Hospital of Qinhuangdao; Peiyuan Lv, Hebei General Hospital; Huiying Zhao, No. 1 People’s Hospital in Shijiazhuang; Ping Gu, The First Hospital of Hebei Medical University; Ronghuan Yu, Kaifeng Central Hospital; Jiewen Zhang, Henan Provincial People’s Hospital; Youlong Zhou, The Fifth Affiliated Hospital of Zhengzhou University; Shanshan Yang, Daqing Oilfield General Hospital; Shurong Duan, The First Affiliated Hospital of Harbin Medical University; Junjian Zhang, Zhongnan Hospital of Wuhan University; Qiuyun Tu, The Third Xiangya Hospital of Central South University; Lu Shen, Xiangya Hospital Central South University; Li Sun, The First Bethune Hospital of Jilin University; Jun Xu, Northern Jiangsu People’s Hospital; Ying Huang, The First Affiliated Hospital of Gannan Medical University; Kunnan Zhang, Jiangxi Provincial People’s Hospital; Xiujie Han, Anshan Changda The Hospital; Qiang Ma, Affiliated Zhongshan Hospital of Dalian University; Cui Wang, Xiaohong Wang, Dalian Municipal Central Hospital Affiliated of Dalian Medical University; Yunpeng Cao, The First Hospital of China Medical University; Jianfei Xian, Shengjing Hospital of China Medical University; Furu Liang, Baotou City Central Hospital; Xuelian Ji, Inner Mongolia People’s Hospital; Li Mei, Xining No. 1 People’s Hospital; Yifeng Du, Shandong Provincial Hospital; Lan Tan, Qingdao Municipal Hospital; Jintao Zhang, Chinese PLA eighty-eighth hospital; Jinbiao Zhang, Weihai Municipal Hospital; Fengyun Hu, Shan Xi Provincial People’s Hospital; Yuling Tian, First Hospital of Shanxi Medical University; Zhirong Liu, Wei Ge, Xijing Hospital; Qiumin Qu, The First Affiliated Hospital of Xi’an Jiaotong University; Qihao Guo, Huashan Hospital, Fudan University; Huidong Tang, Ruijin Hospital, Shanghai Jiao- tong University School of Medicine; Guanjun Li, Shanghai Mental Health Center; Qin Chen, West China Hospital, Sichuan University; Ying Ma, Affiliated Hospital of North Sichuan Medical College; Yongjun Wang, Tianjin Anding Hospital; Yong Ji, Tianjin Huanhu Hospital; Jiong Shi, Tianjin Medical University General Hospital; Yongbin Song, Urumqi General Hospital of Lanzhou Military Area Command; Xiaoying Zhang, Xinjiang Bingtuan Hospital; Xinling Meng, Traditional Chinese Medicine Hospital of Xinjiang Autonomous Region; Xiufeng Xu, The First Affiliated Hospital of Kunming Medical University; Benyan Luo, The First Affiliated Hospital, Zhejiang University; Wei Chen, Sir Run Run Shaw Hospital, School of Medicine, Zhejiang University; Enyan Yu, Tongde Hospital of Zhejiang Province, Ying Zhang; Zhejiang Provincial People’s Hospital; Wenguang Liu, Kangning Hospital Affiliated to Wenzhou Medical University; Zhen Wang, The First Affiliated Hospital of Wenzhou Medical University; Tong Zhou, The Second Affiliated Hospital of Zhejiang University School of Medicine; Yanjiang Wang, Daping Hospital, Third Military Medical University; Xiaomei Wang, Southwest Hospital, Third Military Medical University; Yang Lv, The First Affiliated Hospital of Chongqing Medical University; Qing Zhang, General Hospital of Ningxia Medical University.
This study was supported by the Key Project of the National Natural Science Foundation of China (81530036); the National Key Scientific Instrument and Equipment Development Project (31627803); Mission Program of Beijing Municipal Administration of Hospitals (SML20150801); Beijing Scholars Program; Beijing Brain Initiative from Beijing Municipal Science & Technology Commission (Z161100000216137); Innovation Base Training and Development Special Program (Z171100002217007); CHINA-CANADA Joint Initiative on Alzheimer’s Disease and Related Disorders (81261120571), and Project for Outstanding Doctor with Combined Ability of Western and Chinese Medicine.
