Alzheimer’s disease burden in China (ABC) study: protocol for a nationwide multicentre cross-sectional and prospective cohort study Authors: Yixuan Wang, Qi Qin, Kun Yang, Mingzhi Yu, Yao Yao, Chao Gong, Jing Guo, Li Yang, Yi Tang Published: 2026 ABSTRACT AbstractIntroduction Alzheimer’s disease (AD) is imposing an increasing public health and socioeconomic burden. In China, rapid population ageing is sharply increasing disease burden. Previous studies have shown that AD-related costs are mainly driven by long-term informal care. However, evidence in China remains limited by an incomplete cost framework and insufficient consideration of caregivers’ burden and indirect costs. Notably, the National Dementia Action Plan (2024–2030), issued by the Chinese government, marks a major shift to early detection and comprehensive care of AD, highlighting the urgent need for nationally representative economic evidence to support policy implementation. This study aims to evaluate the economic burden and quality of life of patients with AD and their caregivers in mainland China, and is the first nationwide study to include individuals with amnestic mild cognitive impairment, providing foundational data for future health technology assessment of early AD interventions.Methods and analysisBaseline characteristics will be presented and compared using t-tests or χ2 tests. Economic burden will be estimated by calculating the per capita cost and weighted national total based on provincial numbers of patients with AD. Indirect costs will be assessed using locally adapted replacement cost approach and forgone wages approach. The analysis will be stratified by disease severity and age. Future burden will be projected by linking data from China Statistical Yearbook 2025 and the UN World Population Prospects 2024. Unmet care needs, AD-related catastrophic health expenditure and AD dependency ratio (ADDR) will also be assessed.Ethics and disseminationEthics approval was obtained from the Ethics Committee of Xuanwu Hospital, Capital Medical University (No. 2023132). The study has been registered at ClinicalTrials.gov (NCT05995418) and the Chinese Clinical Trial Registry (ChiCTR2300074723). The results from this study will be actively disseminated through research articles and conference presentations.Trial registration numbersNCT05995418; ChiCTR2300074723.Keywords: HEALTH ECONOMICS, Dementia, Quality of Life, China STRENGTHS AND LIMITATIONS OF THIS STUDY This study employs a two-stage probability sampling design across 602 cognitive centres in 31 provinces, ensuring strong national representativeness of the Chinese population. This study employs a two-stage probability sampling design across 602 cognitive centres in 31 provinces, ensuring strong national representativeness of the Chinese population. It is the first national health economics study to prospectively include patients with amnestic mild cognitive impairment, addressing a critical evidence gap in the early stage of disease and providing data for future evaluation of the cost-effectiveness of early screening and interventions. It is the first national health economics study to prospectively include patients with amnestic mild cognitive impairment, addressing a critical evidence gap in the early stage of disease and providing data for future evaluation of the cost-effectiveness of early screening and interventions. The integration of clinical characteristics, economic burden and quality of life scales provides a multidimensional framework for future policy evaluations and health technology assessment. The integration of clinical characteristics, economic burden and quality of life scales provides a multidimensional framework for future policy evaluations and health technology assessment. The calculation of indirect costs relies heavily on caregivers’ self-reported data regarding care time and missed work, which may introduce recall bias. The calculation of indirect costs relies heavily on caregivers’ self-reported data regarding care time and missed work, which may introduce recall bias. The use of a nested subsample for the 12-month follow-up may introduce loss-to-follow-up bias, although appropriate statistical weighting techniques will be applied to mitigate this. The use of a nested subsample for the 12-month follow-up may introduce loss-to-follow-up bias, although appropriate statistical weighting techniques will be applied to mitigate this. Introduction Background Alzheimer’s disease (AD) is the most common neurodegenerative disease. It causes progressive decline in cognitive function, ultimately leading to significant disability or death.1 In China, the prevalence of AD and related dementias is increasing rapidly in the context of population ageing. As the disease progresses, patients require growing levels of care and assistance, resulting in escalating informal caregiving time and non-medical expenditures.2 3 Families often face difficult trade-offs between employment, caregiving responsibilities and financial sustainability.4 As a result, AD burden extends beyond clinical impairment to encompass substantial social and economic consequences, particularly for family caregivers.5 Several health economic studies on AD have been conducted in China, among which the study by Jia et al first highlighted the economic burden of AD in China, estimating direct medical, direct non-medical and indirect costs.6 This study laid a crucial foundation for domestic health economics research on AD. However, China’s population ageing accelerated significantly over the past decade, economic development progressed steadily and healthcare systems and cognitive diagnosis capabilities improved continuously. The National Dementia Action Plan (2024–2030) issued by the Chinese government in 2024, marks a shift in China’s AD prevention and control policy. It has shifted from a single disease treatment to a full-course management including screening, diagnosis, treatment and follow-up.7 Against this background, the ‘Alzheimer’s disease Burden in China’ (ABC) Study is designed as a nationwide, multicentre research integrating Hospital Information System data, structured questionnaire and follow-up data. The ABC Study aims to cover multiple dimensions including clinical characteristics, economic burden, psychological scales and policy impacts. This protocol describes the design, sampling strategy, variable framework and planned analyses of the ABC Study, with the aim of providing a transparent methodological reference for AD health economics research in China. We aim to provide solid data support for future evaluation of intervention measures, design of long-term care insurance (LTCI) policies and cross-disciplinary research. Current evidence gaps Since the pivotal study by Professor Jia’s team in 2018, studies of economic burden of AD in China have entered a phase of rapid growth. More research has begun to explore the impact of disease staging on cost structures, regional resource allocation disparities and the cost-effectiveness of new diagnostic technologies such as biomarkers. However, existing research exhibits several limitations that hinder the translation of study evidence into policy. First, representative data across all provinces are still lacking, hindering equitable resource allocation and policy formulation between regions.8 Second, insufficient stage-specific cost analysis fails to accurately reflect changes in burden structure across different disease stages.9 Third, inadequate attention to early-onset patients and those with multiple comorbidities may lead to underestimation of indirect costs and family risks. More critically, existing studies often overemphasise economic expenditures while neglecting patients’ psychological well-being, quality-of-life changes and caregivers’ emotional distress and life imbalances.10 Without systematic assessment of these human-centred dimensions, estimated economic burdens fail to authentically reflect the disease’s comprehensive impact on families and society. Policy urgency and public health significance In China, the National Dementia Action Plan (2024–2030) has set the goal of establishing a comprehensive, continuous prevention and control system for dementia by 2030. Therefore, policymakers require reliable, comprehensive, and regionally representative AD burden data support: How do we evaluate the incremental costs of population screening? How should Long-Term Care Insurance payment scope be tailored for individuals with cognitive impairment? How will the inclusion of novel AD drugs in the national medical insurance catalogue impact total societal expenditure? In response to this pressing need, the ABC Study emerges as a nationwide, multicentre investigation. This research will collect structured cost assessments, caregiver modules and quality-of-life evaluations covering all disease stages. By combining cross-sectional estimates with follow-up data, the study aims to provide data support for the prevention and treatment objectives under the ‘Healthy China 2030’ initiative. Methods and analysis Study population The ABC Study aims to conduct a comprehensive multicentre investigation into the economic burden of patients with AD and their families across various provinces and cities in mainland China. The study population not only includes patients diagnosed with AD at the dementia stage but also prospectively incorporates individuals in the stage of amnestic mild cognitive impairment (aMCI). This group is particularly important for exploring the economic value of early intervention in the disease. The inclusion and exclusion criteria for patients are shown in table 1. Table 1Inclusion and exclusion criteria of patientsInclusion criteriaExclusion criteria Diagnosed according to the National Institute on Aging-Alzheimer’s Association 2011 criteria for aMCI and AD.Clinical Dementia Rating (CDR) score greater than or equal to 0.5.Have good visual, auditory and language functions, or can complete neuropsychological assessments after correction.Participants or their legal representatives sign informed consent. History of stroke with neurological focal signs and imaging findings consistent with cerebral small vessel disease (Modified Fazekas score ≥2).Presence of mental or neurological developmental delay.Presence of other known conditions that may cause cognitive impairment.Diagnosis of a disease that prevents completion of cognitive assessments.Refusal to sign informed consent at baseline. AD, Alzheimer’s disease; aMCI, amnestic mild cognitive impairment. Inclusion and exclusion criteria of patients Diagnosed according to the National Institute on Aging-Alzheimer’s Association 2011 criteria for aMCI and AD. Diagnosed according to the National Institute on Aging-Alzheimer’s Association 2011 criteria for aMCI and AD. Clinical Dementia Rating (CDR) score greater than or equal to 0.5. Clinical Dementia Rating (CDR) score greater than or equal to 0.5. Have good visual, auditory and language functions, or can complete neuropsychological assessments after correction. Have good visual, auditory and language functions, or can complete neuropsychological assessments after correction. Participants or their legal representatives sign informed consent. Participants or their legal representatives sign informed consent. History of stroke with neurological focal signs and imaging findings consistent with cerebral small vessel disease (Modified Fazekas score ≥2). History of stroke with neurological focal signs and imaging findings consistent with cerebral small vessel disease (Modified Fazekas score ≥2). Presence of mental or neurological developmental delay. Presence of mental or neurological developmental delay. Presence of other known conditions that may cause cognitive impairment. Presence of other known conditions that may cause cognitive impairment. Diagnosis of a disease that prevents completion of cognitive assessments. Diagnosis of a disease that prevents completion of cognitive assessments. Refusal to sign informed consent at baseline. Refusal to sign informed consent at baseline. AD, Alzheimer’s disease; aMCI, amnestic mild cognitive impairment. Sampling strategy Participant recruitment and baseline cross-sectional data collection were conducted from July 2023 to August 2025 using a two-stage probability cluster sampling design. The first stage used medical institutions as sampling units. Considering the professionalism of AD diagnosis and treatment, the study relied on 602 officially certified cognitive centres distributed across 31 provinces, autonomous regions and municipalities as the sampling frame.8 This centre-based sampling not only ensured the reliability of diagnosis but also covered the main distribution areas of China’s AD medical resources. Since Guangxi, Qinghai and Ningxia each have only one cognitive centre qualified for AD diagnosis, these hospitals are directly included in the study through certainty selection. Xizang currently lacks hospitals with standardised diagnostic capabilities for AD. We recruited and trained locally qualified neurologists to conduct patient enrolment and data collection locally. This subset of samples will be separately labelled and subjected to sensitivity analysis. In the sample size calculation, the study used the average annual medical expenses of patients with AD as the core outcome indicator, employing a mean estimation method with relative precision calculation. To offset the loss of independence among samples caused by cluster sampling, the Design Effect (DEFF) is adjusted using the intracluster correlation coefficient (r). The study preset rho to be 0.005, which is a relatively robust parameter selection aimed at ensuring that the sample remains sufficiently representative even when there is a certain degree of homogeneity among hospitals. Under these assumptions, the DEFF is approximately 1.50. Detailed parameters are presented in table 2. Table 2Parameters definition and selectionParametersValueConfidence level (1−α)95%According to the standard normal distribution quantile.Relative error (δ)3%Lower than 3%.Population mean (μ) of medical costs20 893.39According to the relevant research on average economic burden of patients with AD in China.Population standard deviation (σ) of medical costs25 463.5Reflecting the degree of dispersion of medical costs among patients with AD, derived from the same referenced literature.8Intraclass correlation coefficient (r)0.005Measures the homogeneity of medical costs among patients with AD within the same hospital (cluster). As ICC in socioeconomic research usually applies as 0.001–0.005, we took a higher value to make the population more representative.Average cluster size (m)100We applied a small-scale preinvestigation among hospitals to arrive at this average cluster size. Cluster size in each hospital is allowed to fluctuate between 40 and 120. The estimated average number of patients enrolled per centre during the survey period is approximately 100.AD, Alzheimer’s disease; ICC, intracluster correlation coefficient. Parameters definition and selection AD, Alzheimer’s disease; ICC, intracluster correlation coefficient. Based on the China Alzheimer’s Disease Report 2024 and the China Statistical Yearbook 2025,11 we estimate the number of patients with AD in each administrative region, and calculate the regional distribution proportion accordingly. The number of involving hospitals in each region is allocated proportionally according to the patient scale. The largest remainder method is used to balance the deviation caused by the integer rounding to calculate the regional hospital numbers. The final numbers of hospitals selected in each region are shown in table 3. Table 3Numbers of hospitals across regionsRegionProportion of AD cases (%)Hospital numbersPatient numbersNorth China11.681111110North East China8.0728768East China31.801303024Central South China27.250262592South West China15.207151446North West China5.9896570Total100.00096 (rounded) 9510 The numbers of hospitals in each region may have slight deviations due to rounding of integers. During study implementation, adjustments may be permitted without changing the regional stratification structure to ensure the overall sample size meets statistical requirements.AD, Alzheimer’s disease. Numbers of hospitals across regions The numbers of hospitals in each region may have slight deviations due to rounding of integers. During study implementation, adjustments may be permitted without changing the regional stratification structure to ensure the overall sample size meets statistical requirements. AD, Alzheimer’s disease. Nested phase 2 and prospective follow-up During the baseline study period, a nested 12-month phase 2 was conducted. As caregiving time and productivity loss are cumulative and cannot be accurately captured through a single cross-sectional assessment, a nested phase 2 was designed. The nested phase 2 was designed to include approximately 20% of participant–caregiver dyad from the planned baseline sample (planned N=9510). A stratified random sampling strategy was applied within clusters to maintain representativeness across regions and institutions. Data on caregiving hours, labour participation, and productivity loss were collected to estimate the indirect economic burden of AD. During the analysis phase, potential bias arising from selective participation in or attrition from the nested phase 2 will be addressed using weighting methods or sensitivity analyses. Following completion of baseline data collection, the ABC study is planned to continue as a prospective cohort. The first follow-up wave is planned to commence approximately 3 years after completion of the baseline (around 2028), with subsequent follow-up waves planned at approximately 3-year intervals. As continued follow-up of the ABC cohort is planned, no definitive final study end date has currently been decided. Study variables The ABC Study includes a multidimensional variable framework covering six domains, as shown in table 4. The detailed Baseline Case Report Form and Core Questionnaire and the Second-Phase Follow-up Questionnaire are provided in online supplemental appendix 1. Table 4Variable framework and analytical structureDomainSubdomainVariablesMeasurementAnalytical purposeSociodemographic characteristicsPatient-levelAge, sex, education, marital status, ethnicity, occupation, income sourceQuestionnaireCovariates, subgroup analysisHousehold-levelNumber of children, family income, residence (urban/rural)QuestionnaireEquity and financial risk analysisInsuranceType of medical insurance, LTC insurance, commercial insuranceQuestionnairePolicy impact analysisClinical characteristicsCognitive functionMMSE scoreStandardised scaleDisease severity classificationFunctional statusADL scoreStandardised scaleDependency levelDementia stagingCDR, CDR-SOBStandardised scaleAnalysis between stagesPresent illnessClinical diagnosis detailsCRFDisease profilingComorbidities and medicationChronic diseasesHypertension, diabetes, stroke, CAD, COPD, CKD, etcHospital systemMultimorbidity burden analysisMedication useAntidementia drugs, psychotropics, other chronic medicationsHospital systemCost driversEconomic burden(including costs for PET and blood biomarkers)Direct medical costsOutpatient, hospitalisation, medication costHospital systemCost estimationDirect non-medical costsCaregiving salary, transport, home modificationQuestionnaireHousehold expenditureIndirect costsProductivity loss, informal caregiving hoursQuestionnaire and follow-upReplacement and forgone wages approachQuality of life and psychological statusQoLQoL-AD (patients only), EQ-5D-5L (both patients and caregivers)Standardised scaleUtility estimation, HRQoLPatient and caregiver mental healthPHQ-9, CBF-PI-15Standardised scaleDepression burdenCaregiver burdenZarit Burden InterviewStandardised scaleCaregiver strainCaregiver-specific informationEconomic impactIncome loss, employment changeFollow-upDynamic indirect costEmotional statusPsychological distressFollow-upHuman-centred burdenCare contextWillingness to outsource careFollow-upPolicy simulationAD, Alzheimer’s disease; ADL, Activities of daily living; CAD, Coronary artery disease; CBF-PI-15, Chinese Big Five Personality Inventory-15; CDR, Clinical Dementia Rating; CKD, Chronic kidney disease; COPD, Chronic obstructive pulmonary disease; CRF, case report form; EQ-5D-5L, EuroQol 5-Dimension 5-Level questionnaire; HRQoL, Health-related quality of life; LTC, long-term care; MMSE, Mini-Mental State Examination; PET, Positron emission tomography; PHQ-9, Patient Health Questionnaire-9; QoL, quality of life. Variable framework and analytical structure AD, Alzheimer’s disease; ADL, Activities of daily living; CAD, Coronary artery disease; CBF-PI-15, Chinese Big Five Personality Inventory-15; CDR, Clinical Dementia Rating; CKD, Chronic kidney disease; COPD, Chronic obstructive pulmonary disease; CRF, case report form; EQ-5D-5L, EuroQol 5-Dimension 5-Level questionnaire; HRQoL, Health-related quality of life; LTC, long-term care; MMSE, Mini-Mental State Examination; PET, Positron emission tomography; PHQ-9, Patient Health Questionnaire-9; QoL, quality of life. This variable framework is intended to support both primary analyses of the current study and future secondary analyses, including health utility estimation, caregiver burden assessment and policy evaluation related to LTCI and dementia care financing. Data analysis plan The baseline characteristics of the study population will be described as mean±SD or median (IQR) for continuous variables and number (percentage) for categorical variables. T-test or non-parametric test will be used, and the χ2 test will be performed for comparing. Continuous cost variables will be analysed using generalised linear models to account for potential skewness. Regression models will incorporate sampling weights, stratification and clustering within primary sampling units (hospitals) using appropriate survey modules (eg, generalised estimating equations). These baseline characteristics will include age, sex, education, marital status, geographic region, place of residence (urban/rural), number of children, time of AD diagnosis, Mini-Mental State Examination (MMSE) scores, Activities of daily living (ADL) scores, sleep status and the number of comorbidities. Additional descriptive analyses will also be conducted for caregiver-level data collected in phase 2. Missing data patterns will be examined. For variables with more than 5% missingness, multiple imputation by chained equations will be considered. Otherwise, complete-case analysis will be performed. In the second phase, indirect cost data are collected from a subset of participants who completed follow-up interviews. To account for potential selection bias due to non-random attrition, inverse probability weighting will be applied.12 A logistic regression model will be used to estimate the probability of completing the second-phase survey conditional on baseline characteristics. Covariates included region, urban/rural residence, age, sex, education, marital status, household income, disease severity (Clinical Dementia Rating stage), functional status (ADL score), comorbidity burden, caregiver availability, unmet care needs, baseline direct medical and non-medical costs, and hospitalisation status. Stabilised inverse probability weights will be calculated as the marginal probability of follow-up divided by the predicted probability of follow-up. These weights will be applied to the second-phase sample to obtain estimates representative of the baseline population. Extreme weights will be examined and truncated in sensitivity analyses to assess robustness. Economic burden estimation The per capita cost will be calculated by summing up the costs across the mentioned categories. Nationwide costs will be calculated using provincial numbers of patients with AD as weights. Subgroup analysis will be used, stratified by severity (aMCI, mild, moderate, severe), age (40–59, 60–69, 70–79, ≥80). Caregiving time (hours) that absent from work will be collected by caregiving items. When estimating the economic impact on informal caregivers, two mostly used approaches (replacement approach and forgone wages approach) will be used.13 14 The replacement approach model is as follows: Unlike in other countries, formal care professionals in China are typically hired on a full-time, monthly salary basis. Therefore, the conventional ‘time×wage’ approach does not fully capture the true nature of replacement costs. Based on this, we adjust our model as described above. Annual wage of market caregiver is based on the 2024 average wage for the ‘Resident Services, Repair, and Other Services’ sector in each province.11 This standard is used to adjudicate nursing costs.13 15 The forgone wages approach model is as follows: Occupational average wage will be investigated by this study, and labour force participation rate (LFPR) will be estimated using data from the National Statistic Bureau.16 Nationwide grand total costs will be calculated using national costs per capita plus the number of patients with AD, and projection will be made based on the data of population structure using the World Population Prospects 2024 by the UN, presuming an unchanged prevalence according to other research, or using an estimated future prevalence.14 As the study population was recruited from clinics and hospitals with diagnostic capacity, the observed sample mainly represents patients who have already accessed medical care. However, a substantial proportion of individuals with AD in China remain undiagnosed or untreated in the community. Direct extrapolation of observed per-capita costs to the national AD population may therefore overestimate direct medical costs. To improve the national representativeness of cost estimates, a post-stratification cost adjustment model will be applied to account for undiagnosed and untreated patients.17 A conventional form of this adjustment model is expressed as: Because nationally representative cost data for undiagnosed patients with AD are not currently available in China, this study will use the ratio of the cost of undiagnosed patients to that of diagnosed patients derived from a study in Hong Kong, which may better reflect the real-world situation in Chinese populations.18 In addition, a crude diagnostic ratio will be estimated using the 1-year consultation rate reported in the China Alzheimer Report 2025 and postconsultation diagnostic accuracy reported in previous studies.19 20 Based on published evidence, the 1-year consultation rate was 32.6%, and postconsultation diagnostic accuracy was assumed to range from 61% to 73%; the mean value of 67% will be used in the main analysis. Sensitivity analyses will be conducted under alternative assumptions regarding postconsultation diagnostic accuracy and the ratio of the cost of undiagnosed patients to that of diagnosed patients. The crude diagnostic ratio will be calculated as follows: In which D denotes the crude diagnostic ratio, C denotes the 1-year consultation rate, and A denotes postconsultation diagnostic accuracy. The adjusted total cost will be calculated as: In which N denotes the total number of patients with AD, D denotes the crude diagnostic ratio, C−obs denotes the mean per-capita cost among observed diagnosed or treated patients, and θ denotes the cost ratio of undiagnosed patients relative to diagnosed patients. Unmet value of market care Caregiving status (needing care and being cared for, needing care and not being cared for, not needing care) will be collected in patients. Regional caregiving gap (percentage of unmet caregiving) will be calculated based on investigation with provincial prevalence as weight. By multiplying caregiving gap with regional number of patients and regional indirect costs estimated by approaches in 2.5.1, regional and national unmet value of market care will be calculated. Projection will be made based on the data of population structure using the UN World Population Prospects 2024, presuming an unchanged prevalence according to other research or using an estimated future prevalence.5 21 AD dependency ratio Following the extension from the conventional dependency ratio (DR) and the old-age DR, the ABC Study explores the ADDR as a disease-specific indicator of potential care and economic pressure associated with AD.22 23 The ADDR will be calculated as follows: The numerator will be the estimated and projected number of individuals living with AD, as described in section Economic burden estimation, and the denominator will be the provincial or regional population aged 15–64 years. Provincial and regional population estimates and projections will be generated using a simplified cohort component technique based on the accessible data from the UN, National Bureau of Statistics and Baidu Migration Bigdata Platform.14 21 24 25 AD-related catastrophic health expenditure This study will also use costs and income data to calculate AD-related catastrophic health expenditure (CHE), which is defined as AD-related medical costs exceeding 40% of household income, as follows: Sensitivity analysis will be performed by adjusting the percentage of CHE and replacing the household income by an estimated household capacity to pay.26 Quality of life and psychological status QoL and psychological status of patients and caregivers will be analysed using validated instruments, including EuroQol 5-Dimension 5-Level questionnaire (EQ-5D-5L), QoL-AD, Patient Health Questionnaire-9 (PHQ-9) and the Zarit Burden Interview. Multivariate regression models will be used to identify factors related to quality of life and psychological outcomes. Changes in the economic burden of longitudinal follow-up of patients and caregivers Longitudinal follow-up is designed to assess the longitudinal changes in the economic burden of patients with AD and their caregivers over time. The analysis will focus on changes in the total economic burden during disease progression and explore the impact of disease staging and care burden on different types of cost. In addition, we will assess changes in the impact of care on families, including changes in caregiver productivity, income and care-related expenditure. Patient and public involvement Patients and the public are not involved in the design, conduct, reporting, or dissemination plans of this research. Discussion By constructing a comprehensive variable database covering 31 provinces, the ABC Study is designed to provide a structured evidence base for policy-making, resource allocation and clinical pathway optimisation in dementia care in China. Compared with the data estimation of China’s AD health economics published by Jia et al, this study improves the methodology and reflects the current situation of the health economics burden in China and the evolution of policy and social background.6 7 The ABC Study adopts a national multicentre cluster sampling design and plans to collect data covering all provinces in the country through standardised data collection to improve the representativeness and comparability of research samples. In addition, it incorporates a more comprehensive cost framework system, covering direct medical costs, non-medical costs and indirect costs, and includes caregiver-related outcome indicators and quality of life assessment indicators in detail. Unlike previous studies that focused mainly on the elderly population, this study is designed to include a group of young patients with AD aged 45–60, some of whom had not yet retired, thereby enabling a more comprehensive assessment of productivity loss and early disease burden. A further strength of this study is that it is the first nationwide disease burden study in China to include individuals with aMCI; this may help fill an evidence gap in early-stage AD and provide data for future health technology assessment (HTA) of early screening and interventions. In addition, previous studies often estimated replacement costs using hourly wage-based assumptions, which may not be appropriate in the Chinese caregiving market, where caregivers are more commonly employed on a full-time or monthly salary basis rather than an hourly basis.6 27 This consideration has been incorporated into the questionnaire design, and the replacement cost approach and the forgone wages approach are designed to calculate the indirect cost. Building on previous research of ABC, the ABC Study may also provide a structured variable framework for future interdisciplinary studies, spanning health economics, caregiver research, dementia policy and long-term care financing. The potential value of the ABC Study data is mainly reflected in the following strategic research directions. Future research on caregiver burden and family resilience Future research could use the Zarit scale and family income structure data to construct a multiple regression model for caregiver burden.28 By analysing the nonlinear relationship between caregiver psychological stress and family financial risk across different disease stages, families at the highest risk of burden can be identified. Furthermore, based on 12-month follow-up data, it is possible to investigate whether caregivers’ psychological states may influence patients’ cognition through changes in care quality. HTA and medical insurance access analysis With the approval and launch of new biomarker-guided drugs such as Aβ monoclonal antibodies in China, healthcare insurance decision-makers are facing financial pressure and challenges in access decisions.29 The medical expenses at different stages provided by the ABC Study are essential parameters for conducting pharmacoeconomic models. Policy optimisation and expansion of LTCI Currently, China’s LTCI pilots mainly focus on physical disability and have insufficient coverage for disability caused by cognitive impairment.30 The ABC Study records the impairment of patients’ daily lives and the corresponding care time. Future studies can use this data to evaluate the differences in the effectiveness of different payment ratios and assessment tools in reducing family burdens by comparing the LTCI coverage models of different provinces.31 CHE and the risk of poverty due to illness As a long-term chronic disease, AD imposes a cumulative economic burden on families. Using the detailed household income and expenditure structure variables, the incidence of CHE among AD families in different provinces can be accurately calculated.32 ADDR as indicator of future AD-related care pressure ADDR may represent a useful extension of dependency-based indicators from general demographic ageing to disease-specific burden. The conventional DR was originally developed to reflect the pressure imposed by dependent populations on the working-age population and was later extended to old-age dependency measures in response to population ageing.22 23 Building on this, the ABC Study further focuses on AD-related burden through the ADDR. This may be particularly critical in China, where rapid population ageing and regional migration are reshaping the balance between care needs of patients with AD and the capacity of the working-age population to provide support. Ethics and dissemination The original protocol received approval from the Ethics Committee of Xuanwu Hospital, Capital Medical University, Beijing, China (No.2023132). The study protocol was registered at ClinicalTrials.gov (NCT05995418) and the Chinese Clinical Trial Registry (ChiCTR2300074723). All participants or their legal representatives provided written informed consent. For participants with cognitive impairment, decision-making capacity was assessed by trained physicians using the MMSE. Written informed consent was provided by legally authorised representatives for participants who lacked decision-making capacity. Supplementary material Acknowledgements We would like to express our sincere thanks to all medical workers, patients and caregivers participating in this study. This protocol manuscript was prepared with reference to the STROBE checklist and CHEERS. The reporting checklists are provided as online supplemental files. Footnotes Funding: This work was supported by the National Science and Technology Major Project for Non-communicable Chronic Diseases (Grant No. 2025ZD0546200) and the Beijing Outstanding Young Scientist Programme (Grant No. JWZQ20240101023). Prepublication history and additional supplemental material for this paper are available online. To view these files, please visit the journal online (https://doi.org/10.1136/bmjopen-2026-122035 ). Provenance and peer review: Not commissioned; externally peer reviewed. Patient consent for publication: Not applicable. Patient and public involvement: Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research. References Supplementary Materials