Epidemiology and risk factors of Alzheimer’s disease and related dementias in South and Southeast Asia: a systematic review and meta-analysis protocol

Abstract

Background

Alzheimer’s disease (AD) impacts over 55 million individuals worldwide and remains the leading cause of dementia (60–70% of cases). By 2050, South and Southeast Asia are projected to have an older adult population more than double, bearing a major share of Alzheimer’s disease burden. This will exert a heavy strain on healthcare systems, particularly in resource-limited countries where support and infrastructure are already stretched. Despite this, no review has yet explored the regional epidemiology and associated risk factors in this context. Thus, this study protocol outlines to synthesise prevailing evidence from these densely populated regions, particularly low- and middle-income nations within South and Southeast Asia.

Methods

This review will include studies that reported epidemiological characteristics including prevalence, age of onset, mortality, and risk factors of AD and related dementias comprising in South and Southeast Asian regions. Studies published in any language from inception to date will be extracted from PubMed, Scopus, CINAHL, EMBASE and APA PsycNet, following Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) and Meta-Analysis of Observational Studies in Epidemiology (MOOSE) guidelines. We will also search grey literature sources and screen the reference lists of the articles selected for full-text review to identify additional relevant studies. Observational studies including case–control, cohort, and cross-sectional designs reporting desired outcomes will be included and appraised for quality assessment with the modified Newcastle-Ottawa Scale (mNOS). The included articles will be appraised by two independent reviewers, with a third resolving any conflicts. Pooled estimates of prevalence, age of onset and mortality will be analysed using random effect meta-analysis (REML) model. Associated risk factors, including modifiable and non-modifiable will be narratively synthesised. Forest plots will be used to visualise the findings, and heterogeneity across the included studies will be assessed using the I² and Cochrane’s Q statistics. Potential publication bias will be assessed using a funnel plot along with the Begg’s and Egger’s tests. Sensitivity and subgroup analyses will also be conducted to assess the robustness of pooled estimates and to explore potential sources of heterogeneity. Statistical analysis will be conducted using Rstudio (v.4.3.2) and GraphPad Prism V.9.0.2.

Ethics and disseminations

The systematic review is focused on the analysis of secondary data from published literature; thus, no ethical approval will be needed. The protocol will follow international standard guidelines, findings will be reported in a reputed journal and disseminated through (inter)national conferences, webinars and key stakeholders to inform policy, research and AD management strategies.

PROSPERO registration number

CRD 420251047105.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• This study will explore the epidemiology of Alzheimer’s disease and related dementias (AD/ADRDs), it’s associated factors, and subgroup variations across South and Southeast Asian regions.

• The review will adhere to high-quality international guidelines, including Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P), Meta-Analysis of Observational Studies in Epidemiology (MOOSE) and modified Newcastle-Ottawa Scale (mNOS), with independent, blinded review and oversight.

• Inter-rater reliability among reviewers will be evaluated at each review stage, using Cohen’s kappa and percentage agreement, with a third reviewer ensuring impartiality.

• Excluding studies that do not specify clinically diagnosed AD/ADRDs cases may limit data from South and Southeast Asia, potentially affecting comprehensive regional prevalence estimation of all-cause dementia and it’s clinical subtypes.

• In addition, variability in the study design, diagnostic criteria, assessment tools, and individual quality of studies may impact comparability and generalisability, introducing methodological bias.

Introduction

Alzheimer’s disease (AD) is an expanding global epidemic and a major contributor to dementia accounting for 60–70%. It is marked by progressive neurodegeneration, cognitive decline and functional impairment.¹ There are notable geographical variations in the incidence of AD globally.² The highest rates are found in Europe (19.4 per 1000 people),³ followed by the USA (15 per 1000),⁴ Brazil (7.79 per 1000)⁵ and India (3.29 per 1000).⁶ AD ranks as the fifth leading contributor to mortality globally.⁷ Within the USA alone, individuals living with the disease are likely to range from 6.9 to 13.8 million by the year 2060.⁸ Moreover, the associated economic burden is also substantial, with annual per capita costs ranging from US $ 468 to US $ 1,71, 284 depending on regional, cultural and social care differences.⁹ AD not only impairs individuals’ independence, but also significantly impacts caregivers, healthcare systems, and society at large.⁶ Both modifiable factors (e.g., vascular risk factors, comorbidities, malnutrition, diet, inflammation, educational level, psychiatric illness)^{7 10} and non-modifiable factors (e.g, age, sex, familial history, Apolipoprotein E [ApoE], ɛ4 allele, altered gene regulation, genetic factors)^{11 12} contribute independently or collectively to AD pathogenesis.^{13 14} Studies have suggested that modifiable factors contribute to~35% of the global population attributable fraction for Alzheimer’s disease and related dementias (AD/ADRDs), and addressing the risk determinants could delay up to 40% of cases.¹⁵ Moreover, geographical, socio-economic, socio-cultural, environmental and healthcare access disparities further influence the overall epidemiology, associated risk factors and burden of AD in different regions.^{16 17}

While population ageing is a global phenomenon, with every country witnessing a peak both in the proportion and number of older individuals. AD poses a significant and escalating public health challenge.¹⁸ Between 2015 and 2050, the global share of people aged over 60 is expected to double from 12% to 22% in future.¹⁹ Currently, AD has lower incidence rates in low- and middle-income countries (LMICs) compared with the western world and often starts at an earlier stage of life.¹⁰ The burden of AD will be particularly concerning in LMICs, with ~80% of cases expected to occur there by 2050, exerting a significant disease burden on the entire healthcare system.^{6 19 20} South and Southeast Asia, comprising 25.3% and 8.5% of the global population, respectively, are projected to undergo the most significant rise in older adult population.²¹ This projected rise in AD cases is anticipated to place a substantial disease burden on respective healthcare infrastructure.²² By the next 25 years, geriatric people in these regions are forecasted to be doubled, with an increase of 312 million.^{23 24} Consequently, LMICs are expected to bear nearly 37% of the world disability burden from AD/ADRDs, with South and Southeast Asia accounting for 20% and 17%, respectively.¹² In spite of the growing burden, research on AD in LMICs remains sporadic and limited compared with high-income countries (HICs).²⁵ To date, there is a scarcity of systematic exploration and no meta-analysis has previously focused on the epidemiology and associated risk factors particularly in South and Southeast Asian regions. This gap in evidence synthesis impedes the development and design of targeted interventions. Moreover, current global estimates are largely drawn from HICs, while these regions suffer from underdiagnosis, limited surveillance system, and scarcity of epidemiological data. AD case rates across the countries are depicted in figure 1, highlighting the disparity and the urgency for region-specific evidence. Therefore, this study protocol outlines a comprehensive methodology for systematic review and meta-analysis of epidemiological characteristics of AD/ADRDs including prevalence, age of onset, mortality, as well as associated risk factors among South and Southeast Asian population.

Figure 1. Distribution of Alzheimer’s disease rates across South and Southeast Asian countries. The map showing the geographic variation of Alzheimer’s disease prevalence rates (cases per 10 000 population) from the World Population Review Databases, 2025, highlighting differences across South and Southeast Asian countries as well as high-, upper-middle- and low-middle-income countries in these regions.

Study objectives

The primary objective of this study is to review relevant literature systematically from inception to 1^st May 2025 and conduct a meta-analysis to estimate the epidemiological characteristics of clinically diagnosed AD/ADRDs cases including pooled prevalence, age of onset and mortality among the population in South Asia and Southeast Asia.²⁶ In addition, the study aims to investigate the geographical variations in the prevalence, age of onset, and mortality of AD/ADRD cases in these regions. It will also explore both modifiable and non-modifiable risk factors associated with this disease.

By estimating the pooled prevalence, age of onset and mortality rates, geographical differences, and associated risk factors across different economic settings,²⁶ the study will allow for regional comparison and contribute to broader insights into the epidemiological burden of the disease (figure 1). The findings will emphasise the urge for addressing the growing disease burden, addressing all key determinants to guide future prevention, early detection, and intervention strategies in these regions of the world.

Review question

The research question for this study is structured using the PECO framework,²⁷ which provides a systematic approach outlining the population (P), exposure (E), comparison (C) and outcome (O). The PECO elements are defined as follows:

• Population (P): individuals residing in South Asia and Southeast Asia.

• Exposure (E): being clinically diagnosed as AD/ADRDs.

• Comparison (C): there will be no comparison group in this study.

• Outcome (O): epidemiology of AD/ADRDs including pooled prevalence, age of onset, mortality along with associated risk factors (both modifiable and non-modifiable) across these regions.

Guided by the PECO framework, this review will answer the following research question: ‘What is the epidemiology of AD and it’s associated risk factors among South and Southeast Asian population?’.

Methods

Study design

This review will adhere to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Protocols (PRISMA-P 2015)²⁸ and the Meta-Analysis of Observational Studies in Epidemiology (MOOSE)²⁹ guidelines. It will strictly follow their respective checklists to ensure methodological rigour, transparency, and consistency throughout the review process including study selection, inclusion, data extraction, quality assessment, and data analysis. The review will include empirical longitudinal studies including case-control, cohort, and cross-sectional studies that will meet pre-defined eligibility criteria published from inception to date. Eligible studies must assess patients with AD/ADRDs using standardised, validated assessment tools, and report the epidemiology including prevalence of AD/ADRDs, age of onset of the disease, mortality and/or explore associated risk factors. Included articles quality will be assessed using modified Newcastle-Ottawa Scale (mNOS).³⁰ The protocol for this study has been registered with PROSPERO³¹ under the registration number CRD 420251047105.

Eligibility criteria

The eligibility (both inclusion and exclusion) criteria for selecting relevant articles for the review are outlined as follows in table 1.

Table 1. Eligibility criteria for the systematic review.

Criteria type	Details
Inclusion criteria
Study design	Empirical population-based observational studies with quantitative estimates (case-control, cross-sectional, longitudinal including cohort studies).
Study area	Studies conducted in South and/or Southeast Asian countries, or any other global studies that specifically report data for population from any of the South and Southeast Asian countries.
Outcome	Studies reporting prevalence of AD/ADRDs, onset of the disease, risk factors and subtype of dementia including AD.
Assessment tools	Use of any standardised and validated tools/criteria to assess AD/ADRDs (eg, MMSE, CDR, NIA-AA, NINCDS-ADRDA, NINDS-AIREN, DSM IV, HDS, MoCA).
Language and availability	Studies published in any language with free full-text availability, either through open-access platforms or institutional repositories.
Gender-specific studies	Gender-specific studies (including only men or women) reporting relevant outcomes on AD/ADRDs.
Exclusion criteria (studies)
Publication type	Review articles, media reports, interventional studies, case reports, case series, study protocols, commentaries, book chapters, preprints, conference abstracts, letters.
Data type	Studies lacking any quantifiable prevalence data on AD/ADRDs or related expected outcomes or associated risk factors.
Access	Unpublished studies or studies with inaccessible full texts.
Outcome reporting	Studies without using any comprehensive assessment tool or not using internationally accepted definitions/criteria to clinically diagnose AD/ADRDs.
Geographical specificity	Articles presenting data from multiple countries without a South and Southeast Asia-specific sub-group analysis.
Duplicate reporting	Multiple articles published from the same project using the same dataset (Only the most comprehensive, recent and methodologically robust article reporting AD/ADRDs cases will be considered).
Study relevance	Animal studies.
Specific subgroup focus	Studies with an improper use of multi-design (e.g., insufficient sample sizes or individual cases or narrowly defined subgroups).
Exclusion criteria (population)
Non-target population	Studies on patients with clinically diagnosed dementia that did not specify the exact number of clinical AD cases or provide any subgroup analyses (dementia or other clinical subtypes).

AD, Alzheimer’s disease; ADRDs, Alzheimer’s disease and related dementias ; CDR, Clinical Dementia Rating; HDS, Hasegawa Dementia Scale; DSM IV, Diagnostic and Statistical Manual of Mental Disorders, fourth edition; MMSE, Mini-Mental State Examination; NIA-AA, National Institute on Aging and Alzheimer’s Association; NINCDS-ADRDA, National Institute of Neurological and Communicative Disorders and Stroke-Alzheimer’s Disease and Related Disorders Association; NINDS-AIREN, National Institute of Neurological Disorders and Stroke-Association Internationale pour la Recherche et l’Enseignement en Neurosciences.

Information sources

Electronic search

The information for this study will be sourced from a comprehensive range of reputable databases, including PubMed (via NLM), Scopus (via Elsevier), CINAHL (via EBSCOhost), EMBASE (via Elsevier) and APA PsycNet (via APA). Additionally, to ensure the quality of the review, the research team will comprehensively search reference lists of included studies (citation chaining) as well as grey literature, to identify potential eligible studies reporting the desired outcomes.

Search strategy

To ensure a comprehensive and systematic search, a detailed search strategy has been formulated consulting with a librarian and expert in systematic reviews for selected bibliographic databases in combination with Medical Subject Headings, truncated and phrase-searched key-terms, appropriate filters for specified databases. The search strategy has integrated keywords focused on (“Alzheimer’s disease” OR “Alzheimer’s dementia”) (“Epidemiology” OR “Prevalence”) (“Risk factors” OR “Associated factors”) (“South Asia” OR “Southeast Asia”) (“Observational studies” OR “Cross-sectional studies” OR “Case-control studies”). A preliminary key search string for the predefined databases is outlined in table 2 and strings for each mentioned database are included as a supplementary file (online supplemental file 1).

Table 2. Key terms for article screening during developing search strategy.

Concept	Search terms (Combined with OR/AND)
Population group	(Adults) OR (Male) OR (Female) OR (Older adults) OR (Older population) OR (Older population) OR (Senior citizen) OR (Geriatric group) OR (Ageing population)
Geographical region (South Asia)	(South Asia*) OR (South Asian Region) OR (South Asian Country*) OR (Southern Asia) OR (Afghanistan*) OR (Bangladesh*) OR (Bhutan*) OR (India*) OR (Sri Lanka*) OR (Maldives*) OR (Nepal*) OR (Pakistan*) OR (Lanka) OR (Ceylon)
Geographical region (Southeast Asia)	(Southeast Asia*) OR (South East Asian Region) OR (Southeast Asian country*) OR (Brunei*) OR (Cambodia*) OR (Indonesia*) OR (Laos*) OR (Malaysia*) OR (Myanmar*) OR (Philippines*) OR (Singapore*) OR (Thailand*) OR (Timor-Leste*) OR (Vietnam*)
Disease terms	(Alzheimer’s disease) OR (Alzheimer disease) OR (Alzheimer’s dementia) OR (Dementia) OR (Alzheimer’s disease related dementia) OR (Cognitive decline) OR (Mild cognitive impairment) OR (Neurodegenerative disorder) OR (AD*) OR (ADRD)
Epidemiological concepts	(Epidemiology*) OR (Prevalence) OR (Incidence) OR (Trends) OR (Risk factors) OR (Associated factors) OR (Triggering factors) OR (Determinants) OR (Modifiable factors) OR (Non-modifiable factors) OR (Protective factors) OR (Contributing factors)

Boolean operators (AND, OR, NOT), truncations and database-specific filters will be applied to refine and enhance the search results. There will be no language restriction during the search of relevant literature to reduce the language bias and increase comprehensiveness of the included studies.

Condition/domain being studied

This review will aim to assess the epidemiology of AD/ADRDs, specifically focusing on prevalence, age of onset, mortality and associated risk factors, across South and Southeast Asian countries.

Population/participants

The review will include individuals from the general population of South and Southeast Asia, regardless of age, ethnicity or gender diagnosed for clinical AD/ADRDs, based on the clinical diagnosis of the expert neurologists and/or validated assessment tools.

Exposure

We will include studies that have investigated both modifiable and non-modifiable risk factors associated with AD/ADRDs as exposures.

Comparator(s)/control

In the designed study, we will have no comparator/control group.

Context

Understanding the epidemiology (prevalence, age of onset, mortality) and risk factors of AD/ADRDs among South and Southeast Asian population.

Reference management and articles screening platform

To manage the records obtained through the comprehensive literature search from predefined electronic databases, EndNote v.21.0³² will be applied to organise and compile all references. Duplicate entries will be systematically identified and removed using EndNote reference management software. Any additional studies identified through manual searches from grey literature will also be added to the EndNote library as necessary. After de-duplication, the existing records will be exported into Rayyan QCRI,³³ an available web-based platform designed to support collaborative article screening. Independent reviewers will conduct the initial screening phase, and full-text articles will be uploaded for subsequent assessment.

Study selection process

To ensure a comprehensive and unbiased inclusion of studies on epidemiology and risk factors on AD/ADRDs, the study team will go through a multistage screening process. Study selection will be mainly conducted in two stages: (1) titles and abstracts screening using Rayyan web-based application, and (2) full text of articles screening based on the eligibility. In each of the stages, conflicts will be discussed and mitigated by a third reviewer. A record of reasons for excluding individual studies will be documented. Studies screened during the first stage will only undergo the full-text screening phase. The entire article selection process will be documented in the PRISMA flow diagram²⁸ (figure 2). To ensure that only studies of sufficient quality, and direct relevance to the research question will be included for the final synthesis. The method will help streamlining the selection process, and improve co-ordination among the reviewers. A visual overview of the entire article screening and study selection process is illustrated in figure 3.

Figure 2. PRISMA flow diagram for overall study procedure. Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram outlining the selection process of studies included in the systematic review searched from six (PubMed, Scopus, CINAHL, EMBASE and APA PsycNet) electronic databases from inception to 1^st May 2025.

Figure 3. Systematic process of articles screening for final inclusion. The down-drop flowchart illustrating stepwise depiction of the article screening procedure leading to final study inclusion.

Data extraction and management

After identifying eligible studies, reviewers will extract desired data using a structured, predesigned data extraction form. This form will capture key study information, including author details, publication year, study methodology, sample characteristics, geographical aspects, study settings and participant demographics such as age, gender and diagnostic criteria for AD/ADRDs. The extraction will focus on the data related to prevalence, age of disease onset, mortality and associated risk factors of AD/ADRDs, encompassing modifiable and non-modifiable risk factors. For the main outcome of interest effect measure will be extracted as prevalence proportion (% and n/N; where, n = number with the outcome, and N = total sample size) along with 95% confidence intervals (CIs). For the associated risk and protective factors, odds ratios (ORs) or risk ratios (RRs) with 95% CIs (adjusted or unadjusted), including corresponding p-value, will be extracted. Relevant details regarding diagnostic criteria and assessment tools will also be extracted and categorised according to severity levels (e.g., mild/moderate/severe) as extracted. Where applicable, additional information on disease severity, age of onset, mortality, comorbidity and functional status will also be extracted. Two reviewers will extract data using standardised Microsoft Excel spreadsheet, with a third reviewer verifying accuracy and completeness. The final dataset will be prepared, cross-checked and analysed using R statistical software (v.4.3.2) to facilitate a thorough and transparent synthesis of results.

Patients and public involvement

This study protocol does not directly involve individual patients, caregivers, members or public organisations. The study relies exclusively on data from previously published studies.

Statistical analysis

Data analysis and statistical analysis

Finally, included studies will be meta-analysed to report the pooled prevalence, age of onset and mortality. Descriptive analysis will be presented to demonstrate the included study characteristics and regional variations. Given the expected variability in methodologies, and sample characteristics across studies, a random-effects meta-analysis (REML) model will be employed. The pooled prevalence estimates will then be computed using an REML to assess for heterogeneity across the studies.³⁴ The package entitled ‘meta’ will be used for meta-analysis.³⁵ To quantify the extent of heterogenicity among studies, we will calculate both the index of heterogeneity I² statistic, with values exceeding 50% considering substantial heterogeneity and Cochrane’s Q test (with a p<0.10 indicating significant heterogeneity).³⁶ Risk factors associated with AD/ADRDs will be summarised in a descriptive table, if multiple studies report on the same risk factors, the results will be pooled using meta-analysis. If quantitative synthesis is not feasible for meta-analysis, narrative synthesis will be conducted. To enhance clarity and interpretability, a forest plot will be used to visually represent effect sizes and their CIs (figure 4A), highlighting variations across the studies.³⁷

Figure 4. Forest and funnel plot for meta-analysis. Forest and funnel plots demonstrating the meta-analysis of selected outcomes across a sample of 10 included studies. (A) Forest plot showing Odds Ratios (ORs) comparison. (B) Funnel plot showing publication bias detection.

Sensitivity analysis

Sensitivity analyses will assess the robustness of pooled estimates for AD/ADRDs prevalence. Specifically, studies identified as low quality during the assessment process will be excluded to determine whether study quality significantly affects the overall findings. This approach will be applied to meta-analyses of the overall outcome by re-running the analyses without low-quality studies, to assess changes in the pooled estimates and CIs. REMLs³⁸ will be used to compare differences in pooled effect sizes. If substantial variations are observed, this may indicate that study quality has a significant influence on the overall results of the review.

Subgroup analysis

A subgroup analysis will also be conducted based on factors such as type of diagnostic criteria, age group, gender, study design, economic settings (HICs, upper income countries [UICs], LMICs), regions (South Asia vs. Southeast Asia), study quality assessed by mNOS score (low vs. high-quality) and risk factors (modifiable vs. non-modifiable). Moreover, secular trends of AD will be explored by clustering the studies according to the decade of data collection (e.g., before 2000, 2000–2010 and after 2010) depending on the data availability from the included studies. In addition, further subanalysis will be stratified by dementia type, including all-cause dementia, clinically diagnosed AD and other clinical subtypes (e.g., vascular dementia, frontotemporal dementia, auto-immune dementia and parkinsonism spectrum).

Risk of bias and quality assessment of individual studies

To ensure a rigour and quality in the assessment process, two reviewers will assess the risk of bias following pre-designed appropriate assessment tool. A mNOS tool,³⁰ with a total score of 10 allocated across relevant sections, will be incorporated to assess the study quality. According to the total score, included studies will be categorised as high (0 to ≤ 3), moderate (4 to 6) and low risk (≥ 7 to 10). However, studies rated either high, moderate or low risk will be considered for final inclusion.³⁰ Key domains of quality assessment include: (A) selection: sample representativeness, sample size, (B) comparability: study design or analysis and (C) outcome: assessment of outcome and statistical analysis. I² statistic, which is interpreted as low (≤ 25%), moderate (> 25% to < 75%) or high (≥ 75%) of heterogeneity. In cases of substantial heterogeneity, an REML will be used, and prediction intervals will be reported to account for between-study variability.³⁶ Furthermore, inter-rater reliability agreement between independent reviewers will be assessed at each stage of the review process (title/abstract screening, full-text review, extraction and risk of bias assessment) using both Cohen’s kappa statistic and percentage agreement. A kappa value ≥ 0.60 and percentage agreement ≥ 80% (above moderate agreement) will be considered acceptable, indicating substantial agreement among reviewers.³⁹ Any conflict at any stage will be resolved through discussion or adjudication by a third reviewer to ensure methodological rigour, comprehensiveness and overall consistency.

Dealing with missing data

For instances where relevant information will be missing, the study team will reach out to the corresponding author via email to obtain clarification of finally included studies. If efforts to retrieve the necessary data are unsuccessful, the incomplete data will be excluded from quantitative analysis, but available factors will be included for narrative synthesis. The lack of such data and its potential impacts will be addressed in the review’s discussion section.

Strategy for data synthesis

A comprehensive search strategy will be conducted systematically using relevant keywords and concepts. Boolean operators, truncations, co-occurrence terms, filters and explode functions will be used to optimise the efficiency and precision of the search (figure 5). A preliminary systematic search using the key search terms, conducted in PubMed on 1^st May 2025, retrieved 114 potentially relevant studies from this region. After applying eligibility criteria during the screening process, eligible studies will undergo systematic data extraction and organisation for quantitative synthesis. The summary findings will be demonstrated using a table consisting of all the extracted data. Where appropriate, a subgroup analysis will be performed in accordance with the outcome to address heterogeneity among the studies. In case, where meta-analysis will not be feasible, particularly regarding the risk factors, a narrative synthesis will be carried out to summarise and analyse the findings concerning targeted outcomes of AD/ADRDs across the regions.

Figure 5. PICO (P: Population, I: Intervention, C: Condition, and O: Outcome) framework and co-occurrence network of key search terms. Cluster illustration of PICO framework highlighting the key components of the research question, alongside a co-occurrence network of key search terms used during the systematic search process.

Publication bias assessment

Publication bias will be evaluated with a combination of statistical test and visual techniques. Specifically, rank correlation test (Begg’s test),⁴⁰ Egger’s linear regression test⁴¹ and visual inspection of funnel plots to detect potential asymmetry and overall distribution of effect sizes (figure 4B). Funnel plots will be generated when at least 10 studies are available for the given primary outcome. To further evaluate the impact of publication bias, offering insights into potential small-study effects or biases that may influence the overall interpretation, the ‘trim and fill’ method will be applied. A significant result from Egger’s test indicating funnel plot asymmetry, suggesting potential publication bias. Conversely, p-values (p> 0.05) from Egger’s tests will be interpreted as indicative of a low risk of publication bias.³⁷ All analysis related to bias assessment will be performed and visualised using the ‘meta’ package in R.³⁵

The status and timeline of the study

The systematic review is currently at it’s initial phase, which involves refining the research question, formulating the search strategy and identifying appropriate databases and searching potential grey literature. The authors are in the process of finalising the searches and titles, abstracts screening phases. However, the entire review process, including searching, data synthesis, data extraction, quality assessment, data analysis and manuscript preparation is expected to be completed within the next 10-12 months. This timeline has been designed to ensure a comprehensive and methodologically sound exploration of the available evidence. A detailed timeline outlining the key milestones of the systematic review project is presented in the accompanying table 3.

Table 3. Timeline of the systematic review (with phases).

Phases of review	Key milestones	Estimated timeline*
Planning phase	1. Formulate research question. 2. Define primary and secondary objectives. 3. Register protocol in PROSPERO.	Y1Q1-Y1Q2 (Completed)
Search strategy development	1. Identify key search terms. 2. Select databases (Pre-defined). 3. Develop search syntax.	Y1Q2 (Completed)
Literature search	1. Conduct electronic database searches. 2. Import references to citation manager and organising. (EndNote)	Y1Q2 (Completed)
Primary screening	1. Screening for titles and abstracts. 2. Full-text screening as per eligibility. (Rayyan QCRI)	Y1Q2-Y1Q3 (Not completed)
Data extraction	1. Design data extraction form (desired variables). 2. Extract key findings from included studies (Microsoft Excel).	Y1Q3 Not completed)
Quality assessment	Scoring of included studies according to mNOS for quality assessment.	Y1Q3 (Not completed)
Data extraction and synthesis	Conduct meta-analysis and/or narrative synthesis (R Statistical Software and GraphPad Prism v.9).	Y1Q4 (Not completed)
Analysis and manuscript writing	1. Draft primary manuscript. 2. Review and revision based on coauthors/ reviewers’ feedback. 3. Finalise for submission.	Y1Q4-Y2Q1 (Not completed)
Submission and revisions	Submit to a reputed international peer-reviewed journal. Address as peer-review feedback. Disseminate findings.	Y2Q2 (Not completed)

^*Y₁Q₁=Year (Y)₁, Quarter (Q)₁ (January to March, 2025) of the reporting period 2025–2026.

mNOS, Modified Newcastle-Ottawa Scale ; Q, Quarter; Y, Year.

Discussion

This study aims to synthesise existing evidence from longitudinal and observational studies, to estimate the epidemiology and pooled prevalence of AD. It will also investigate and analyse demographics, clinical and contextual risk factors associated with AD/ADRDs. This area has remained under-researched in this region compared with HICs.⁴² Studies involving patients across different age groups, assessment tools and settings both urban and rural from all countries within South and Southeast Asia will be included, regardless of healthcare infrastructure or institutional affiliation. While global prevalence estimates of AD continue to rise (13.8 million by 2060),⁸ particularly in LMICs, region-specific evidence remains limited, especially under-represented and under-invested in Asian countries.^{25 43} Geographical disparities contribute to~1.05% of the variation in mortality related to AD/ADRDs, along with other environmental exposures.⁴⁴ However, both non-modifiable factors,⁴⁵ and modifiable ones, including pre-existing conditions, unhealthy lifestyle choices, environmental exposures⁷ and limited access to AD care are expected to play significantly influencing the disease progression in these complex and multifactorial disorder.⁶ Gender disparities have also been observed globally, with women often showing higher prevalence rates of cognitive decline.⁴⁶ By comparing regional data to the global trends and including the secular trend, this review seeks to determine whether the epidemiological patterns align with or diverge from broader global findings. Thus, the findings can provide actionable insights for comparing AD/ADRDs burdens and prioritising future research tailored to the unique demographic and socio-cultural context.

Strengths and limitations of the review

This study will comprehensively evaluate and synthesise evidences from longitudinal and observational studies, while analysing subgroup variations within populations across the regions. This study will adhere to internationally recognised methodological standards, to ensure a rigorous approach to study selection, extraction of outcome variables, reporting and assessment for quality. Two reviewers will conduct articles screening and data extraction in a blinded manner, with a third reviewer providing oversight to maintain the objectivity and resolve discrepancies.

Several potential limitations may also warrant consideration. First, there might be substantial heterogeneity in study designs, diagnostic criteria and methodological quality across included studies. This may pose challenges to statistical pooling and compromise the uniformity of findings. To address this, we will conduct subgroup analysis. Second, we may face variations in the sample sizes, assessment tools and associated factors which may influence reporting pooled prevalence rates and identify risk factors, potentially limiting the generalisability of results across diverse cultural and healthcare settings, which will be addressed using sensitivity analysis by excluding studies with high risk of bias. Thus, sensitivity analyses and subgroup comparisons will be conducted to explore the sources of heterogenicity, robustness of the findings and address studies’ limitations. Third, since the review tends to exclude studies reporting all causes of dementia without specifying the number of clinically diagnosed AD/ADRDs cases, there may exist a paucity of eligible studies regarding epidemiology and risk factors of AD from desired regions or countries. This could be another potential limitation in comprehensively ascertaining the overall region-specific prevalence rate. Additionally, the reliance on self-reported or caregiver-reported data in many studies may introduce recall or reporting biases. To enhance the robustness and credibility of the findings, studies with missing or incomplete data will be excluded.

Research ethics, dissemination of the research findings and publication policy

As the study will not involve human subjects directly and is based on secondary data, no institutional approval will be required. The study findings will be compiled into a manuscript for submission to an international peer-reviewed journal. Beyond this, key findings will be shared through national and international scientific conferences (e.g., Alzheimer’s Association International Conference),⁴⁷ webinars, symposia and other relevant platforms. To enhance the overall impact of this review, results will also be shared with governments, policymakers, clinicians, neurologists, mental health professionals, academic institutions and other key stakeholders to inform health policy, explore future research projects, improve clinical practices and support regionally tailored strategies for the prevention and early detection of AD/ADRDs.

Potential impact and future direction

Summary findings of the study will have potential impact for future research and clinical practices. This study will provide region-specific evidences to better understand and quantify the disease burden on epidemiology and key risk factors of AD/ADRDs. By synthesising data on key outcomes, this study may exert valuable remarks into the growing burden of AD/ADRDs, where~66% of patients live in LMICs, yet only around 10% of studies have adopted a population-based design to date.²⁵ The findings of the study may attract global funders like Global Brain Health Institute and Alzheimer’s Society (UK) to conduct more explorative research in LMICs including South and Southeast Asian regions to compare with the global diversity. Moreover, the findings will highlight the critical gaps in AD research within LMICs and emphasize the need for increased funding and focused attention. It will also encourage the global dementia research community to prioritise these under-represented regions, reinforcing their importance to the broader research agenda for the associations. Ultimately, this study supports regional efforts to address AD/ADRDs by addressing key risk factors and improving health-related quality of life for both affected individuals and their respective family members and caregivers, in line with the United Nations Sustainable Development Goals (SDG 3), on health and well-being for all.