Identifying Gaps and Barriers in Alzheimer’s Disease and Related Dementia Research and Management in Low- and Middle-Income Countries: A Survey of Health Professionals and Researchers

Abstract

Background:

The significant increase in Alzheimer’s disease and related dementia prevalence is a global health crisis, acutely impacting low- and lower-middle and upper-middle-income countries (LLMICs/UMICs).

Objective:

The objective of this study is to identify key barriers and gaps in dementia care and research in LLMICs and UMICs.

Methods:

We conducted an international, cross-sectional survey among clinicians and healthcare professionals (n = 249 in 34 countries) across LLMICs and UMICs, exploring patient demographics, use of clinical diagnosis, dementia evaluation, screening/evaluation tools, and care and treatment.

Results:

Significant disparities were found in diagnostic practices, access to assessments, and access to care. On average, clinicians in LLMICs saw more patients, had less time for evaluations, lower use of formal screening and tools, and less access to biomarkers. They were also under-resourced compared to UMICs.

Conclusions:

The findings provide insights for policymakers, healthcare organizations, and researchers to address the complex challenges associated with dementia care in diverse settings. Addressing these challenges requires a multipronged approach involving local, national, and international stakeholders.

Keywords

Alzheimer’s disease dementia disparity low and middle-income countries resources underserved

INTRODUCTION

Alzheimer’s disease and related dementias (ADRD) will affect 152 million individuals by 2050, 1 primarily due to increasing life expectancy. Nearly two-thirds of ADRD cases will be in low- and middle-income countries (LMICs) in the next three decades. Most LMICs are largely unprepared for the projected dementia prevalence, where the multidimensional impact will increase disability and dependency for older adults, the financial and emotional burden for caregivers, and community tensions associated with ADRD stigma. 2 Socioeconomic vulnerabilities, including low educational attainment, limited healthcare infrastructure, and geopolitical factors, are several putative barriers that pose unique complexities in the assessment and treatment of dementia in an aging global population in LMICs. 3

Socioeconomic vulnerabilities are often rooted in historical and documented periods of colonialism, relationships with foreign institutions and industries, conflict, natural disasters, or structural adjustment programs.4,5, 4,5 LMICs with limited budgets and resources face a dilemma in prioritizing healthcare spending. With the growing demand for acute medical services and infrastructure development, chronic neurodegenerative conditions like dementia often receive lesser attention. Economic constraints affect not only the availability of resources for care but also the development of skilled healthcare professionals specialized in dementia. In a similar vein, the healthcare infrastructure in many LMICs will face larger challenges to be effectively prepared to handle the complexities of screening, assessment, treatment, and long-term care associated with ADRD than in high-income countries (HIC). 6 Chief among these challenges is a stark shortage of neurologists, neuropsychologists, geriatricians, gerontologists, and mental health professionals trained in diagnosing and managing dementia. 7 Conventional diagnostic tools such as neuroimaging and biomarkers (biofluids [cerebrospinal fluid, plasma] and neuroimaging) are scarce and expensive and, if available, are concentrated in large urban environments. 8 Additionally, geopolitical factors (e.g., conflict, climate change, social inequity, internal displacement) significantly influence healthcare priorities and funding in LMICs, where some countries have a web of international politics and economic dependencies that shape their healthcare agendas.5,9, 5,9 For instance, foreign aid and international health initiatives have primarily focused on communicable and infectious diseases like HIV/AIDS and tuberculosis, overshadowing non-communicable chronic disorders such as dementia.10 - 12 This results in a disproportionate allocation of resources, leaving dementia research, clinical trials, and care underfunded and under-researched.12 - 16

Given the heterogeneity, diversity, and intersectionality of the aforementioned factors, a multimodal approach is required for LMICs. 17 Models, assessments, and methodologies from HICs are not generalizable or easily adaptable for LMICs. 18 Diverse languages and dialects, distinct ethnic backgrounds, and varying literacy levels hinder efforts to raise awareness and educate communities about dementia. 19 Consequently, developing effective, culturally sensitive educational programs that resonate with diverse populations is a significant challenge. The cultural perception of dementia in LMICs adds another layer of complexity. In many of these societies, symptoms of dementia are often attributed to normal aging or even supernatural causes, leading to stigmatization and social isolation. 20 This perception impedes timely medical intervention and reduces the quality of life for patients and caregivers, often leading to underdiagnosis or misdiagnosis, delaying critical care and support.

LMICs face the daunting task of revamping their healthcare systems, addressing cultural stigmas, and reallocating resources to meet the growing needs for equitable dementia care. This requires not only local and national efforts but also international support and collaboration; however, the onus is often shouldered by providers. This study’s primary objective was to identify gaps and barriers in dementia care, support, research, and resources in LMICs to inform targeted strategies to address obstacles encountered across similar contexts. This study was led by the LMIC Workgroup, part of the Diversity and Disparities Professional Interest Area (PIA) of the Alzheimer’s Association International Society to Advance Alzheimer’s Research and Treatment (ISTAART).

METHODS

Study population and design

Participants were eligible to participate if they were clinicians, healthcare professionals, and/or clinical researchers working in LMICs who predominantly saw or provided treatment to adults with cognitive impairment or dementia, irrespective of any specific etiology. The LMIC Workgroup developed a survey that employed a cross-sectional design. To accommodate a culturally and linguistically diverse respondent pool, the original English questionnaire was forward-translated into French, Mandarin Chinese, Portuguese, and Spanish and then back-translated to ensure clarity and parallel interpretation. Data collection occurred between January to September 2023. The link to the survey was disseminated online to targeted countries through the LMIC Workgroup networks and the Alzheimer’s Association ISTAART PIAs. This study was approved by the Human Research Protection Office at Washington University (202204115); all participants consented to data use prior to participation.

Questionnaire

The survey consisted of 38 questions. The first set of 12 questions probed about participants’ demographic information, including age, nationality, location of work, type of institution, professional background, years of working activities, and training background. The remaining 24 questions focused on clinical practice, including the average number of patients seen every month, patients’ demographics, dementia etiologies encountered, utilization of clinical guidelines and tools (e.g., cognitive, functional, and behavioral screening tools), utilization of assessments and tests (e.g., neuropsychological, language and functional assessments, structural neuroimaging exams, imaging biomarker analysis, genetic testing for ADRD), pharmacological treatments, and available education or activities for caregivers. Finally, two open-ended questions inquired about the main challenges respondents face in clinical or research work, as well as identifying resources or support considered beneficial to clinical or research activities by the respondents. The link to the survey was disseminated online, via email and social media, to targeted countries through LMIC Workgroup networks and ISTAART. Data were collected using REDCap, a secure platform for creating and curating online databases and surveys.

Statistical analysis

Survey responder’s primary country of practice were categorized based on World Bank group classification. 21 There were 297 total respondents; 48 were excluded due to employment in HICs, resulting in the inclusion of 249 participants in the final dataset. We employ the classification group to derive three income categories: low-income countries (seven participants), lower-middle-income countries (96 participants), and upper-middle-income countries (146 participants). Initially, we compared the low-income countries group to the middle-income countries group. However, due to the small sample size in the low-income countries group, we elected to combine low-income with the low- and lower-middle-income countries group (LLMICs) to ensure a balanced sample size for comparing various factors. For questions that involved multiple-choice and checkbox responses, we combined all responses from different language versions with the English version as the reference based on the labels. Responses to questions with comments and the final two open-ended questions in non-English languages were translated into English. Based on the question about the use of dementia guidelines, we compiled a comprehensive list of guidelines mentioned by respondents. Then, we categorized them into four groups: established guidelines, laboratory or biomarker guidelines, screening, clinical evaluation, neuropsychological guidelines, and unspecified guidelines. Responses were initially coded based on keywords for the last two open-ended questions and grouped according to different thematic categories. Independent t-test was used to compare continuous variables and Chi-square test for categorical variables to identify statistically significant differences (p < 0.05) between the study groups. Statistical tests were two-sided, and data were analyzed using R version 4.2.3.

RESULTS

Participant characteristics

In the LLMICs group, there were 103 participants working in 14 countries, while the upper-middle-income countries (UMICs) included 146 participants working in 20 countries (Table 1; Supplementary Figure 1). Among respondents employed in LLMICs, the majority were male (64.1%), between 18 and 44 years of age (58.3%) and originated from Asia (49.5%). In contrast, among respondents who worked in UMICs, most were female (57.5%), between 18 and 44 years of age (50.7%) and originated from South America (68.5%). Notably, the two groups had similarities in professional backgrounds and working settings. In both LLMICs and UMICs, a higher proportion of participants worked in urban areas (81.6% and 93.2%, respectively, p = 0.007), were employed in public facilities (54.4% and 56.2%, respectively, p = 0.391), held professional positions in medicine (MD) (62.1% and 84.2%, respectively, p < 0.001), and conducted clinical and research work (63.1% and 76.0%, respectively, p < 0.001). Participants who worked in UMICs had more years of work experience and held PhD degrees compared to those in LLMICs (41.8% in UMICs compared to 13.6% in LLMICs). In terms of years of experience working, there were statistically significant differences noted between participants in 0–10 years (40.8% in LLMICs versus 29.5% in UMICs) and 21–40 years (19.4% in LLMICs versus 33.6% in UMICs) while there were no differences in 11–20 (35.0% versus 33.6%, respectively) or 40 + (1.0% versus 0.7%, respectively). In LLMICs, 13.6% of survey respondents held PhDs, compared to 41.8% in UMICs.

Table 1

Participant characteristics

	Respondents in low- and lower-middle income countries (n = 103)	Respondents in upper-middle income countries (n = 146)	p
Continental regions
Africa	49 (47.6%)	6 (4.1%)	< 0.001
Asia	51 (49.5%)	14 (9.6%)
Central America Region	2 (1.9%)	23 (15.8%)
Europe	0 (0%)	1 (0.7%)
North America	1 (1.0%)	1 (0.7%)
South America	0 (0%)	100 (68.5%)
Sex (Female), no. (%)	33 (32.0%)	84 (57.5%)	< 0.001
Age group (45–64), no. (%)	34 (33.0%)	59 (40.4%)	0.431
Location of work (Urban), no. (%)	84 (81.6%)	136 (93.2%)	0.007
Institution (Public facility), no. (%)	56 (54.4%)	82 (56.2%)	0.391
Profession (Neurology), no. (%)	45 (43.7%)	58 (39.7%)	< 0.001
Professional MD (MD), no. (%)	64 (62.1%)	123 (84.2%)	< 0.001
Capacity (Both clinical and research), no. (%)	65 (63.1%)	111 (76.0%)	< 0.001
Years of work (11-20), no. (%)	36 (35.0%)	49 (33.6%)	0.081
Training (MD), no. (%)	70 (68.0%)	37 (25.3%)	< 0.001
Country received training (Trained and working in the same country), no. (%)	88 (85.4%)	112 (76.7%)	< 0.001

Patient characteristics

The mean number of patients seen by a respondent monthly in LLMICs was 240 compared to 105 in UMICs. In LLMICs, most of the patients were from suburban or peri-urban areas (48.5%), while in UMICs, the majority were from urban areas (69.9%) (Table 2).

Table 2

Patients’ characteristics

	Low and lower-middle income country (N = 103)	Upper-middle income country (N = 146)	p
Number of patients, mean (SD)	240 (312)	105 (122)	< 0.001
Patients’ origin
Rural areas	12 (11.7%)	3 (2.1%)	< 0.001
Suburban/Peri-urban areas	50 (48.5%)	37 (25.3%)
Urban areas	40 (38.8%)	102 (69.9%)
Patients’ location
Private hospital or clinic	40 (38.8%)	52 (35.6%)	0.137
Public hospital or clinic	55 (53.4%)	67 (45.9%)
Residential care facility	3 (2.9%)	4 (2.7%)
Other	5 (4.9%)	20 (13.7%)

Clinical diagnoses

The average monthly number of cases varied across subjective cognitive decline (SCD [14.3; 13.1]), mild cognitive impairment (MCI [14.3, 21.4]), general dementia (12.8, 25.7), dementia of the Alzheimer’s type (DAT [1.67; 1.77]) and unspecified dementia (2.53; 2.78) in LLMICs and UMICs, respectively. Notably, there were statistically significant differences in the number of dementia and MCI cases per month between the two groups (p < 0.001 and p = 0.02, respectively). However, no statistically significant difference was observed in the number of monthly SCD cases, DAT cases, and general dementia cases (p = 0.66, p = 0.72, p = 0.83, respectively). Respondents from both LLMICs and UMICs indicated that they had seen cases of SCD, MCI, dementia, and DAT within one month, with fewer reporting cases of unspecified dementia. Participants from both LLMICs (52.4%) and UMICs (82.2%) endorsed using a specific guideline or criteria for diagnosing dementia. However, in LLMICs, approximately 45% indicated using no particular guidance for ascertaining a diagnosis, which was significantly higher (p < 0.001) compared to 15% UMICs respondents. A larger number of respondents in UMICs (58.9%) reported use of established guidelines compared to those in LLMICs (42.7%) (p < 0.001) (Fig. 1). The top two guidelines used in LLMICs were DSM and ICD-10, while the DSM, NIA-AA, and unspecified tools or guidelines were used in UMICs. Significant differences were observed in the use of ICD-10, NIA-AA, and unspecified tools or guidelines (p < 0.001, p = 0.02, p = 0.03,respectively).

Fig. 1

The number of reported uses of specified guidelines.

Dementia evaluation

Most participants in both LLMICs and UMICs affirmed using cognitive screening tools. Significant differences were observed in evaluation time and the use of cognitive screening tools between groups. The mean evaluation time in the LLMICs was 36.9 min, compared to 53.5 min in the UMICs group (p = 0.005). Compared to LLMICs, UMICs respondents endorsed greater use of screening tools such as MMSE or MoCA (79.6% versus 94.5%; p < 0.001). The MMSE, MoCA, and ACE-R or ACE-III were the most frequently used tools in both groups (Fig. 2), while tools such as MIS and GPCog were less commonly used. CERAD, RUDAS, and Brief Cognitive Screening Battery were frequently used in UMICs but were seldom used in LLMICs. Significant differences existed in the use of MMSE (p = 0.04), MoCA (p = 0.01), CERAD (p < .001), RUDAS (p = 0.04), and Brief Cognitive Screening Battery (p < 0.001) between the two groups.

Fig. 2

The number of reported uses of screening tools.

A higher percentage of respondents in UMICs used informant-based or functional screening tools, compared to LLMICs used such tools (80.8% versus 40.8%; p < 0.001). Although CDR was the most frequently used functional screening tool in LLMICs, only 25.2% reported using it, compared to 58.9% in UMICs. Additionally, a higher percentage of participants in UMICs used the AD8, DAD, IQCODE, and Amsterdam IADL Questionnaire compared to the LLMICs group (Fig. 3). Differences were observed in the use of the CDR and IQCODE (p < 0.001, p = 0.01). Similar patterns were observed in the use of behavioral assessment tools. Most respondents in UMICs endorsed using at least one type of behavioral assessment tool, compared to those in LLMICs (71.2% versus 37.9%; p < 0.001). Behavioral assessment tools like the Geriatric Depression Scale (29.1% versus 56.2% [p < 0.001]) and Neuropsychiatric Inventory Questionnaire (28.2% versus 47.9% [p = 0.001]) were commonly used but were underutilized in LLMICs compared to UMICs. Additional tools like the Cornell Depression Scale were used but more infrequently across respondents in UMICs (15.1%) and LLMICs (6.8%) (Fig. 4).

Fig. 3

The number of reported uses of informant-based or functional screening tools.

Fig. 4

The number of reported uses of behavioral functional tools.

Reasons for not using screening tools

In cases where respondents did not use screening tools, specified reasons were investigated. The four main reasons selected by 20% to 30% of participants from LLMICs included no instrument available in the local language/dialect, lack of local norms or validation, lack of training, and lack of time (Supplementary Figure 2). In contrast, less than 10% of participants from UMICs agreed with the first three reasons, and only 11.6% of participants from UMICs selected a lack of time, compared to 29.1% of respondents from LLMICs. Statistically significant differences were observed across all chosen reasons between the two groups (p < 0.001).

Dementia assessment access and utilization

Participants in LLMICs and UMICs reported their access to various assessments including neuropsychological assessment (61.2% versus 78.8%), specialized language assessment (37.9% versus 50.7%), specialized functional assessment (31.1% versus 40.4%), structural neuroimaging exams (75.7% versus 87.0%), imaging biomarker analysis (17.5% versus 21.9%), cerebrospinal fluid analysis (25.2% versus 36.3%), and genetic testing (25.2% versus 32.9%) for ADRD (Supplementary Figure 3). The UMICs group reported higher access to all assessments, and there were statistically significant differences in the access to neuropsychological assessment (p = 0.001) and structural neuroimaging exams (p = 0.04). There was no statistically significant difference in the proportion of computed tomography (CT) or magnetic resonance imaging (MRI) scans used. However, a significant difference was observed in the average time to receive results between the two groups (p < 0.001), with 52.4% of respondents in LLMICs endorsed receiving results within four weeks, compared to 20.5% in UMICs. In imaging biomarker analysis, positron emission tomography with fluorodeoxyglucose (FDG-PET) was more accessible to participants in LLMICs (13.6%) compared to single-photon computed emission tomography (SPECT) (7.8%) and amyloid-PET (2.9%). SPECT (13.0%, FDG-PET (20.5%), and amyloid-PET (13.7%) were also more accessible in UMICs, while tau-PET was not accessible to either group. Although LLMICs had less access to imaging biomarker analysis, obtaining the results took a shorter time, with 13.6% endorsing having results within one month, compared to 2.1% in the UMICs group. Significant differences were observed in the use of amyloid-PET (p = 0.01) and the average time to receive results (p < 0.001) between groups.

For cerebrospinal fluid analysis, both groups had access to cell count only or general biochemistry and AD biomarkers. However, LLMICs had better access to cell count only or general biochemistry, while UMICs had better access to AD biomarkers. Approximately 17.5% of respondents in LLMICs endorsed getting results within a month, compared to 3.4% in the UMICs group. Significant differences were observed in the use of AD biomarkers and the average time to receive results between the two groups (p < 0.001).

Pharmacological/nonpharmacology treatments and caregiving

Respondents from both groups had access to pharmacological treatments, including cholinesterase inhibitors, memantine, antidepressants, antipsychotics, mood stabilizers, anti-seizure medications, and psychostimulants. The only significant difference was the availability of memantine, which was more accessible in UMICs compared to LLMICs (78.8% versus 58.3%; p < 0.001). Cognitive or neuropsychological rehabilitation, cognitive stimulation or training, occupational therapy, speech and language therapy, and physical therapy were available to patients in both groups. However, cognitive stimulation or training was less accessible in LLMICs compared to UMICs (33.0% versus 54.8%; p = 0.001). In both LLMICs and UMICs, more than half of the respondents indicated that centers provided training or activities for caregivers. General information, group training, and psychological support or psychotherapy were offered to caregivers without any statistically significant difference in access to caregivers between the two groups. However, there is a statistically significant difference in the availability of caregiver trials between the two groups (p < 0.001). Observational or interventional research trials were more available in UMICs (52.1%) than in LLMICs (29.1%).

Main challenges and helpful support

Respondents described daily challenges centered around a population-specific cultural perspective and from an institutional perspective. At the population-specific and cultural perspective level, macro-factors, such as a lack of awareness of dementia, dementia-related stigma, and inadequate health-related policies, as well as micro-factors, such as lack of money to cover medical expenses and caregiver burden, were common. At the institutional level, challenges were classified into business, clinical, or research. From the business platform, a lack of resources for the employee’s well-being, inadequate finances and decision-making for allocation, and a lack of support networks were prominent. The clinical level included inaccessibility and high cost of resources, such as tools and medications, and the lack of high-quality care. The research level included a lack of research opportunities, a lack of interest in research, a lack of research training, and difficulties in recruiting participants.

Respondents also mentioned some solutions to address critical areas in response to these challenges. Regarding population-specific cultural challenges, they noted that support, including financial support for families, meetings/conferences in the LMIC community to raise awareness of dementia, and training and relevant materials for caregivers in local language versions, would be helpful to solve those issues. Solutions to institutional challenges included support via long-term mentorship programs, international collaborative networks, training of clinicians and researchers, and the development and availability of language-appropriate, validated, standardized, and locally adapted tools and treatments that would be very helpful to address core issues (Table 3).

Table 3

Key challenges that participants face in their clinical or research work and support they would like to receive from ISTAART

Challenges					Action steps
Population specific culture^*	Macro-factors	Dementia related stigma			1. Expand service coverage in health care policy;
		Lack of awareness of dementia			2. Provide financial support to patients and their families;
		Lack of transportation			3. Local transportation construction;
		Inadequate health related policy			4. Provide training to educate the public, caregivers, and patients;
	Micro-factors	Money (expenses for medication)			5. Provide information/education material/family guidance in local languages to caregivers, patients, and their families;
		Low educational attainment			6. Set regular meetings on topics of interest to the LMIC community to raise awareness in the public;
		Fear of participating in clinical research			7. Create a school or develop caregiver support programs for caregivers;
		Caregiver burden (behavioral disorders)			8. Develop a self-care protocol for caregivers.
					9. Obtain funding to employ local rural-area residents as navigators to educate residents regarding ways to promote brain health, while also conducting cognitive screenings
Institution^*	Business operations^#	Support	Poor employee well-being	Low salary	1. Develop long-term mentorship program in ISTAART;
				Heavy workload	2. Enable collaboration with international network;
		Capacity	Finances and decision making for allocation	Institution’s bureaucracy	3. Conduct multilingual seminars, webinars, workshops, training, and courses (online available);
				Priority differences between NGO and service providers	4. Develop an online platform for sharing the data and information (free journal articles and accessible statistics);
			Lack of support network	No dementia working group (or not well-organized)	5. Teleconsultation with multi-professional teams;
				Lack of integrated network of clinical trials for interested patients	6. Provide travel grants to attend ISTAART conferences;
					7. Increase employee salary.
	Clinical operations^#	Support	Lack access/high cost of screening/evaluation/diagnosis/rehabilitation tools		1. Develop and give access to linguistically appropriate, brief, standardized, validated, locally appropriate tools and treatments;
			Lack access to treatment and medication		2. Establish specialized memory clinics;
			Lack of cultural and linguistically appropriate/locally validated/globally accepted and standardized tools		3. Provide internship opportunities;
			Lack of clinical dementia care center		4. Provide training for primary health care providers to reduce burden on specialists;
			Lack of caregiver/clinician training and related materials		5. Provide training for professionals (Capacity building);
					6. Establishment of day care and long-term care facilities.
		Capacity	Lack of awareness for early referral and dementia screening, assessment, and treatment		1. Ensure enough time for assessment;
			Quality of care	Low quality of primary care and lack of specialty (e.g., neurology)	2. Timely follow up with participants;
				Lack of rehabilitation service (program)/nursing care	3. Provide evaluation for patients with major neurocognitive disorder in advanced stages of the disease.
				Cumbersome procedures for making appointment and transfer patients
				Lack of structured post-diagnosis management (Lack of follow up for patients (treatment, rehabilitation))
				Lack of multidisciplinary team for clinical care
	Research operations^#	Lack of research opportunities, and specific ADRD research			1. Availability of research results;
		Lack of researcher training (general training)			2. Conduct training courses involving molecular/biomarker analysis;
		Researchers lack interests in doing research			3. Conduct more research in LMICs;
		Difficulties in participant recruitment (hard to reach out participants from rural areas)			4. Provide funding for research projects;
					5. Support epidemiological research.

^*“Training”, “Awareness”, “Funding” are three themes that mentioned in both “Population specific culture” and “Institution” domains. ^#“Lack of financial resources/workforce (human power)”, “Lack of equipment/facilities/space”, “Lack of support network” are three themes that mentioned in all domains under the “Institution” domain.

DISCUSSION

The increasing prevalence of ADRD, particularly in LMICs, poses a significant threat to global health and demands a comprehensive understanding of how risk factors vary across heterogeneous populations and how to improve access to earlier screening, diagnosis, and effective treatments. The need for more research has been explicitly centralized by the World Health Organization’s Blueprint for dementia research,22,23, 22,23 the Lancet Commission on dementia prevention, intervention, and care,12,24, 12,24 and annual reports 25 from Alzheimer’s Disease International. The results of this study add to the extant literature and provide valuable insights into the challenges and disparities faced by clinicians, healthcare professionals, and clinical researchers working in LMICs who care for patients with ADRD and cognitive impairment.

Our findings highlight the impact of economic vulnerabilities on healthcare systems where economic constraints not only limit the allocation of resources for dementia care but also impede the development of a skilled workforce specialized in managing ADRD. The shortage of healthcare professionals trained in dementia management poses a significant obstacle to providing timely screening, assessment, diagnosis, and care of ADRD patients. Moreover, the concentration of diagnostic tools in urban environments exacerbates disparities, leaving rural areas underserved. The stark contrast in the average number of patients seen in LLMICs compared to UMICs highlights the significant burden of dementia in resource-constrained settings. The geographical distribution of patients, with a higher percentage from suburban/peri-urban areas in LLMICs, reflects the challenges of delivering healthcare in diverse settings. The observed differences in the monthly patients diagnosed with MCI underscore structural disparities in early detection and intervention.

Many respondents in LLMICs reported not using specific guidelines, which raises concerns about standardized diagnostic practices and a lack of uniform assessment practices. The preference for established guidelines in HICs suggests a foundation for uniform diagnostic criteria in both LLMICs and UMICs. However, variations in the choice of guidelines, with DSM, and ICD-10, prominent in LLMICs, DSM, NIA-AA, and unspecified guidelines in UMICs, may indicate the need for consensus-building on diagnostic protocols. These differences could stem from cultural perceptions, resource availability, or familiarity with specific guidelines. The longer evaluation time in UMICs may reflect a more comprehensive evaluation process, but it raises concerns about the feasibility of their implementation in LLMICs. The variations in the utilization of different screening tools highlight the need for standardized training and guidelines to ensure consistent and accurate assessments across diverse settings. The identified reasons for not using screening tools, such as the absence of instruments in local languages, lack of local norms or validation, insufficient training, and limited time, underscore critical barriers in LLMICs. These challenges may contribute to delays in dementia diagnosis and hinder the implementation of early interventions. Tools like the RUDAS or Number Symbol Coding Task are brief and may serve to rapidly screen patients for impairments and further evaluations.26,27, 26,27 Discrepancies in access to various assessments, including neuropsychological evaluations and structural neuroimaging exams, highlight infrastructure challenges in LLMICs. While the availability of specific assessments may be lower, the shorter time to receive results in LLMICs suggests potential efficiency gains or regional differences in healthcare systems. However, the limited access to imaging biomarker analysis in LLMICs emphasizes the need for increased resource allocation and infrastructure development to ensure equitable diagnostic capabilities across different income settings.

The NIA-AA’s 2011 clinical guidelines 28 and 2018 research framework 29 will be accompanied by an upcoming Revised Criteria for Diagnosis and Staging of Alzheimer’s Disease: Alzheimer’s Association Workgroup. The 2024 revision will 1) add criteria for diagnosis and staging to inform research and clinical care since several targeted therapies have received regulatory approval, 2) incorporate plasma biomarkers for categorization, disease diagnosis, and staging, and 3) address nonequivalence between fluid and imaging biomarkers within a category. These essential changes reflect the rapid progress of therapeutics and biological staging, most of which occur in HICs. However, adopting these diagnoses and staging guidelines poses additional challenges for adoption and implementation in LMICs where a lack of screening and underdiagnosis are persistent gaps. Additionally, access to biomarkers is mostly limited to larger cities with resourced institutions in urban centers. Globally, ADRD research will need to measure the influence of structural and social determinants of health 30 on regulating the risk of dementia across different settings. Emerging research in LMICs is examining the direct effects of multidimensional poverty on dementia in fragile and conflict situations.31 - 33

Disparities in healthcare access and utilization were evident based on the consensus from the responses. LLMICs face challenges in obtaining timely results from diagnostic assessments. While some imaging biomarker analyses were more accessible, the overall availability of assessments, including neuropsychological evaluations, was lower compared to UMICs. These findings underscore the urgent need for infrastructure development and resource allocation to bridge these gaps in access. There were variations in the availability of pharmacological treatments and cognitive interventions between LLMICs and UMICs. Notably, the accessibility of memantine was significantly higher in UMICs. Additionally, caregiver support and trials were more prevalent in UMICs. These findings highlight the need for equitable treatment options and support services distribution, ensuring that caregivers in LMICs receive the necessary resources.

The identified challenges at both the population-specific cultural and institutional levels provide a comprehensive overview of the obstacles faced by clinicians and researchers in LMICs. The recognition of factors such as a lack of awareness, stigma, inadequate health policies, financial constraints, and caregiver burden highlights the multi-faceted and complex nature of ADRD across LMICs. This study’s findings have significant implications for developing targeted interventions and policies to improve dementia care in LMICs. Efforts to prioritize dementia assessment and care must consider the historical context, geopolitical factors, and economic infrastructure before shaping healthcare agendas in LMICs. The disparities identified in diagnostic practices, access to assessments, treatment options, and caregiver support call for a comprehensive and context-specific approach. Efforts should be directed towards establishing standardized diagnostic guidelines that are culturally relevant and ensuring the availability of assessment tools in local languages. Capacity-building initiatives, including free training programs for healthcare professionals, can address the identified barriers and enhance the quality of dementia care. Collaborative efforts between local, national, and international stakeholders are crucial for addressing institutional challenges and promoting resource allocation for dementia care.

While this study provides valuable insights, certain limitations should be acknowledged. The survey’s reliance on self-reported data introduces the potential for bias, cross-sectional design impacts change over time, and the sample size in low-income countries was small. Despite making the survey available via any web browser to ensure specific developer applications were not blocked, we were unable to recruit any respondents from China. Most patients are seen by family physicians and/or allied health professionals rather than neurologists. The preponderance of evidence was based on a large number of neurologists completing the questionaire. As a result, the results may reflect a specialist bias, which is more sensitive to dementia symptoms and approaches for assessment and management. Additionally, the study did not explore nuances of cultural beliefs around aging, stigma, and respect for older adults that could impact dementia care. The survey was distributed within the ISTAART network and by professionals in the LIMIC dementia working group, which may lead to a selection bias by including professionals with more awareness of existing resources and knowledge on dementia care. Despite these limitations, the findings contribute to the growing literature on dementia care in LMICs. We recruited diverse ages, sexes, professional backgrounds, and geographic locations to reflect the wide-ranging care providers in LMICs. Translation of the survey into multiple languages underscores the commitment to inclusivity, ensuring that language barriers do not hinder participation. The identified gaps and barriers emphasize the urgency of implementing targeted strategies, including international collaborations, policy reforms, and cultural and linguistic adaptation of screening, assessment, and diagnostic tools, to address the complex challenges associated with ADRD in LMICs.

This study sought to comprehensively examine the challenges and opportunities in dementia care in LMICs. The identified disparities in diagnostic practices, access to assessments, and treatment options underscore the urgency of addressing healthcare infrastructure and resource allocation. A concerted, multiprong approach is required to address economic challenges, ranging from policy reforms at individual institutions (e.g., clinics and hospitals) to international collaboration and support. The proposed solutions, informed by the experiences and perspectives of clinicians and researchers in LMICs, provide a roadmap for policymakers, healthcare organizations, and researchers to enhance dementia care in diverse and resource-constrained settings. The findings contribute to the global knowledge on dementia care and highlight the importance of context-specific strategies to bridge the gap in dementia care in LMICs.

AUTHOR CONTRIBUTIONS

Ganesh Babulal (Conceptualization; Data curation; Formal analysis; Funding acquisition; Investigation; Methodology; Project administration; Supervision; Writing – original draft; Writing – review & editing); Wenqing Zha (Methodology; Software; Visualization); Jean-Francois Trani (Methodology; Writing – review & editing); Jorge Llibre Guerra (Methodology; Writing – review & editing); Boon Lead Tee (Methodology; Writing – review & editing); Yiqi Zhu (Methodology; Writing – review & editing); Yaohua Chen (Methodology; Writing – review & editing); Ling Chen (Methodology; Writing – review & editing); Michael Bubu (Methodology; Writing – review & editing); Sylvia Josephy-Hernandez (Methodology; Writing – review & editing); Stephen Wandera (Methodology; Writing – review & editing); Wambūi Karanja (Conceptualization; Formal analysis; Investigation; Methodology; Supervision; Writing – review & editing); Ratnavalli Ellajosyula (Conceptualization; Formal analysis; Investigation; Methodology; Supervision; Writing – review & editing); Paulo Caramelli (Conceptualization; Formal analysis; Investigation; Methodology; Project administration; Supervision; Writing – original draft; Writing – review & editing).

Footnotes

ACKNOWLEDGMENTS

We would like to thank the respondents for lending their voices and lived experiences to this research and the staff and members of ISTAART who provided support in completing the data collection.

FUNDING

This work was partially supported by the National Institute of Health (NIH) and National Institute on Aging (NIH/NIA) grants R01AG068183 (GMB), R01AG074302 (GMB), and R01AG067428 (GMB).

CONFLICT OF INTEREST

Drs Ganesh Babulal and Paulo Caramelli are Editorial Board Members of this journal but were not involved in the peer-review process of this article nor had access to any information regarding its peer review.

DATA AVAILABILITY

The data supporting the findings of this study are available upon request from the contact author.

The supplementary material is available in the electronic version of this article: .

References

Nichols

, Steinmetz

, Vollset

, et al. Estimation of the global prevalence of dementia in 2019 and forecasted prevalence in 2050: an analysis for the Global Burden of Disease Study 2019. Lancet Public Health 2022; 7: e105–e125.

Wimo

, Seeher

, Cataldi

, et al. The worldwide costs of dementia in 2019. Alzheimers Dement 2023; 19: 2865–2873.

Mattap

, Devi

, McGrattan

, et al. The economic burden of dementia in low-and middle-income countries (LMICs): a systematic review. BMJ Glob 2022; 7: e007409.

Walsh

, Govia

, Peters

, et al.

What would a population-level approach to dementia risk reduction look like, and how would it work?

Alzheimers Dement 2023; 19: 3203–3209.

Stites

, Midgett

, Mechanic-Hamilton

, et al. Establishing a framework for gathering structural and social determinants of health in Alzheimer’s disease research centers. Gerontologist 2022; 62: 694–703.

Al-Rousan

, Tsoy

, Sd

, et al. Facilitators and barriers to dementia assessment and diagnosis: perspectives from dementia experts within a global health context. Front Neurol 2022; 13: 769360.

Maestre

. Assessing dementia in resource-poor regions. Curr Neurol Neurosci Rep 2012; 5: 511–519.

Parra

, Butler

, McGeown

, et al. Globalising strategies to meet global challenges: the case of ageing and dementia. J Glob Health 2019; 9: 020310.

Persaud

, Bhat

, Ventriglio

, et al. Geopolitical determinants of health. Ind Psychiatry J 2018; 27: 308–310.

10.

Oni

and Unwin

. Why the communicable/non-communicable disease dichotomy is problematic for public health control strategies: implications of multimorbidity for health systems in an era of health transition. Int Health 2015; 7: 390–399.

11.

Lopez

, Williams

, Levin

, et al. Remembering the forgotten non-communicable diseases. BMC Med 2014; 12: 200.

12.

Livingston

, Huntley

, Sommerlad

, et al. Dementia prevention, intervention, and care: report of the Lancet Commission. Lancet 2020; 396: 413–446.

13.

Llibre-Guerra

, Heavener

, Brucki

SMD

, et al. A call for clinical trial globalization in Alzheimer’s disease and related dementia. Alzheimers Dement 2023; 19: 3210–3221.

14.

Sexton

, Snyder

, Chandrasekaran

, et al. Expanding representation of low and middle income countries in global dementia research: commentary from the Alzheimer’s association. Front Neurol 2021; 12: 633777.

15.

Oliveira

, Deckers

, Zheng

, et al. The career development of early-and mid-career researchers in dementia should be a global priority: a call for action. Aging Ment Health 2022; 26: 439–441.

16.

James

, Mukadam

, Sommerlad

, et al. Culturally tailored therapeutic interventions for people affected by dementia: A systematic review and new conceptual model. Lancet Healthy Longevity 2021; 2: e171–e179.

17.

Alladi

and Hachinski

. World dementia: One approach does not fit all. Neurology 2018; 91: 264–270.

18.

Stephan

BCM

, Pakpahan

, Siervo

, et al. Prediction of dementia risk in low-income and middle-income countries (the 10/66 Study): an independent external validation of existing models. Lancet Global Health 2020; 8: e524–e535.

19.

Swaffer

. Dementia: Stigma, language, and dementia-friendly. London: Sage Publications Sage UK: 2014, p. 709–716.

20.

Johnston

, Preston

, Strivens

, et al. Understandings of dementia in low and middle income countries and amongst indigenous peoples: a systematic review and qualitative meta-synthesis. Aging Ment Health 2019; 24: 1183–1195.

21.

World Bank. World Bank Country and Lending Groups, https://datahelpdesk.worldbank.org/knowledgebase/articles/906519-world-bank-country-and-lending-groups (2024, accessed 1/28/24 2024).

22.

World Health Organization. A blueprint for dementia research. 2022.

23.

Cataldi

, Sachdev

, Chowdhary

, et al. A WHO blueprint for action to reshape dementia research. Nat Aging 2023; 3: 469–471.

24.

Livingston

, Sommerlad

, Orgeta

, et al. Dementia prevention, intervention, and care. Lancet 2017; 390: 2673–2734.

25.

Long

, Benoist

and Weidner

. World Alzheimer Report 2023. Reducing dementia risk: never too early, never too late. London: Alzheimer’s Disease International, 2023.

26.

Galvin

, Tolea

, Moore

, et al. The Number Symbol Coding Task: A brief measure of executive function to detect dementia and cognitive impairment. PLoS One 2020; 15: e0242233.

27.

Storey

, Rowland

JTJ

, Conforti

, et al. The Rowland universal dementia assessment scale (RUDAS): a multicultural cognitive assessment scale. Int Psychogeriatr 2004; 16: 13–31.

28.

Jack

Jr , Albert

, Knopman

, et al. Introduction to the recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 2011; 7: 257–262.

29.

Jack

Jr , Bennett

, Blennow

, et al. NIA-AA research framework: toward a biological definition of Alzheimer’s disease. Alzheimers Dement 2018; 14: 535–562.

30.

Adkins-Jackson

, George

, Besser

, et al. The structural and social determinants of Alzheimer’s disease related dementias. Alzheimers Dement 2023; 19: 3171–3185.

31.

Trani

, Moodley

, Maw

MTT

, et al. Association of multidimensional poverty with dementia in adults aged 50 years or older in South Africa. JAMA Netw Open 2022; 5: e224160.

32.

Trani

J-F

, Zhu

, Park

, et al. Multidimensional poverty is associated with dementia among adults in Afghanistan. EClinicalMedicine 2023; 58: 101906.

33.

Trani

J-F

, Zhu

, Park

, et al. Is multidimensional poverty associated to dementia risk? The case of older adults in Pakistan. Innov Aging 2024; 8: igae007.