This paper describes the design and cohort profile of the Lebanon Study on Aging and HeAlth (LSAHA). LSAHA is the first large-scale, population-level study of Alzheimer’s Disease and Related Dementias (ADRD) and other aging-related health challenges in the Middle East and North Africa (MENA) region. The social and political circumstances in Lebanon over the last few decades render a systematic investigation of ADRD risk and other aging-related health conditions particularly relevant. Adults who are currently aging in Lebanon have experienced prolonged exposure to intense social and political upheaval throughout much of their life course—a civil war from 1975 to 1990, internal displacement, foreign occupation, recurring episodes of political instability, and chronic government mismanagement [1, 2]. These conditions have taxed the social and economic stability of the country, which was further aggravated by the severe economic crisis of 2019 that saw a significant portion of the population descend into poverty [3], the port blast in August 2020, and renewed hostilities in southern Lebanon and Beirut that started in October, 2023. Examining the effects of such destabilizing, stressful experiences throughout the life course on ADRD risk and other late-life health outcomes is a primary objective of the study.

There is a dearth of systematic population-level data on the magnitude of ADRD risk in Lebanon and Arab populations more generally. According to data from the Global Burden of Disease Study, the burden due to ADRD in the MENA region is among the highest globally, and especially Lebanon is projected to experience among the largest increases in ADRD in the next 25 years [4]. These projections reflect the growing burden due to ADRD that will disproportionally affect low- and middle-income countries (LMIC) in future decades [5, 6]. Insights from LSAHA on ADRD and its primary risk factors will therefore likely have relevance for ADRD risk in other LMIC, especially those that have gone through similar periods of social, political, and economic instability.

An important reason for the absence of population-level data on ADRD in Lebanon is the lack of well-validated, comprehensive measures of cognitive functioning and dementia outcomes that can be administered in large samples of older adults [7]. The few previous epidemiologic studies of dementia in the Arab region have relied primarily on short screening tests to identify likely ADRD cases in the general population [8–10]. An exception is a study of 502 older adults in Lebanon that relied on the 10/66 cognitive battery [11]. In LSAHA, we used the approach of the Harmonized Cognitive Assessment Protocol (HCAP) as the basis of our cognitive battery, which is being implemented in about a dozen studies globally [12, 13]. Using the HCAP model provides a comprehensive cognitive assessment among older adults in Lebanon and enables international comparisons regarding the prevalence and incidence of ADRD and its most important risk factors across different populations in the future.

The main objectives of LSAHA are to investigate contextually informed risk factors for ADRD in the Lebanese population, and to provide critical new data on aging-related health and care needs of older adults living with ADRD and their families. The design of LSAHA was aimed at establishing a cohort of adults aged 60 years and older that would be generally representative of the older adult population in Lebanon and of sufficient size to produce statistically valid and robust findings. It follows the general methodology of the Health and Retirement Study (HRS) in the U.S. [14], which has now been adopted in about 50 large-scale population studies of older adult populations in both high-income countries (HIC) and LMIC around the globe. These studies are connected with one another in the HRS Around the World Network [15], and many of these studies also include an HCAP for the assessment of cognitive functioning and ADRD.

LSAHA employed a probability-based multi-stage sampling design in the two pre-selected areas (strata) with a target sample size of 3,000, equally allocated by stratum, with the goal of recruiting 1,500 participants in each. We relied on gridded-population datasets created by WorldPop to provide estimates of population size within a uniform grid size [16]. Both strata were comprised of administrative areas, including areas where lack of local cooperation or security concerns would prohibit any field operations, which were, therefore, removed from the sampling frame. Beirut includes 13 administrative areas of which two areas were excluded for lack of residential population. Next, we created square-shape grids of 100 x 100 meters, or groups of such grids across the entire city, which were then proportionately allocated across the 11 remaining administrative areas. We then selected a simple random sample of grids from each administrative area to serve as primary sampling unit (PSU). In the Zahle stratum, we stratified the 58 administrative areas implicitly according to religion and socio-economic status and then selected the 25 administrative areas as PSUs using probability proportional to estimated size (PPeS), in which the estimated size was obtained by combining estimates provided by mayors of municipalities and WorldPop population estimates. We proportionately allocated grids across the 25 selected PSUs and selected a random sample of grids or groups of grids in each PSU as the secondary sampling unit (SSU). We created maps of all selected grids and sequentially numbered all buildings within each grid. Enumerators made up to three attempts to contact all residential units in each building, except those deemed to be unoccupied or abandoned. After contact with a unit was made, a household roster form was completed with an adult member of the household. We selected up to two age-eligible adults from each household; in households with more than two age-eligible adults, two adults were randomly selected with equal probability using a computer algorithm. Other than age, the inclusion criteria were ability to speak Arabic, having resided in the household for at least three months during the past year, and having no plans to leave Lebanon during the next six months.

We computed sampling weights based on the inverse of the probability of selection at all stages of the sampling design. These weights were subsequently adjusted to account for differential non-response at the SSU and housing unit levels. Finally, the weights were calibrated by age and sex according to population estimates provided by the Central Administration for Statistics – Lebanon [17].

Data collection included the following elements: a household interview with a member of the selected household as the participant (this could be the participant him/herself), an interview with the participant including anthropometric assessments (individual interview), an interview with a key informant of the participant (informant interview), and a blood draw. Recruitment and data collection for the participant interview started on October 9, 2023 and was completed on September 10, 2024. Data collection from the key informants was completed on September 30, 2024, and collection of blood samples was completed on April 10, 2025.

The individual interview involved the administration of a two-hour computer-assisted interview, that included a broad range of standardized tests and questions covering various domains such as demographic characteristics, social and economic conditions including retirement and pensions, physical and mental health, and war exposures (described below in further detail). If a participant elected to complete the interview via a proxy respondent, that proxy was asked to complete a shortened version of the interview without any of the cognitive or physical tests.

The key informant interview, completed via phone by someone designated by the participant, included standardized measures of the participant’s daily functioning and memory status. The household interview covered aspects of the physical conditions of the home environment, the financial assets and expenditures of the household, as well as questions about food and water security.

All participants were invited to donate a blood sample which was usually collected several weeks after the individual interview. Blood samples were immediately analyzed for several clinically relevant biomarkers (e.g., cholesterol, HbA1c) that were shared with the participant, while the remainder of the samples were aliquoted and stored in a biobank for future analysis.

The study questionnaire was developed to comprehensively capture health, social, economic, and lifestyle indicators relevant to aging and to incorporate contextual factors that are particularly important in the Lebanese setting, including economic hardships, exposure to war and conflict, and care needs arrangements. The selection of individual measures, scales and tests was informed by the availability of measures and tests previously used and validated in Lebanon and those used in HRS and its network of studies around the world. Cognitive tests were selected to align with HCAP assessments used in other studies. The final selection of questionnaire measures was based on generating inventories of instruments from HRS studies (with an emphasis on those in LMICs and comparable settings to Lebanon) and prior studies in Lebanon. Measurement alignment with the HRS and HCAP networks is intended to facilitate data harmonization and international comparative research in the future. Measures obtained across all the study components are summarized in Table 1.

This interview was composed of several instruments to evaluate the cognitive and functional status of the respondent as perceived by a key informant, including the Informant Questionnaire on Cognitive Decline among the Elderly (IQCODE) [33], the Blessed Dementia Scale [34], and the Informant Community Screening Interview for Dementia (IF-CSI’D) [20].

This interview included standardized questions on the household infrastructure (e.g., electricity, generators), income and assets, expenditures, food and water insecurity, and the impact of the economic crisis on the household.

LSAHA identified a total of 3,546 potentially eligible participants, of whom 318 refused to participate, 5 we were unable to schedule a revisit, and 159 did not pass the decision competency test. Of the remaining 3,064 participants, 16 elected to have a proxy interview and 21 stopped the interview before reaching the section with cognitive tests. This left a final sample size of 3,027 participants who completed the cognitive test battery section (see Fig. 1 – sampling diagram). The total response rate was 69 percent, using standard definitions from the American Association for Public Opinion Research [35], after accounting for refusals and potentially eligible participants in households we were unable to screen. The response rate was 65 percent in Beirut and 78 percent in Zahle. We collected the Informant Questionnaire for 2373 participants (78%), and the Household Questionnaire for 1997 (66%) participants. A total of 2392 (79%) of the participants donated blood samples, of which 2264 (75%) were stored in a biobank.

The basic demographic and socio-economic variables are presented in Table 2. The average age of the sample was 71.7 (SD 8.1), with almost half the sample being between 60 and 69 years old while 20% was 80 or older. Although almost two-thirds of the participants were female (n = 1904), they accounted for 55.3 percent after applying the sampling weights. About two-thirds (65.1%) of the participants were married, and 21.6% were widowed. The large majority was of Lebanese origin, with 5.3 percent being of Syrian origin. Over 40% (43.1%) of the sample had attained primary education or less, and one in five (19.9%) had received at least some university training. The average number of assets in a household, out of 10 assets assessed, was 6.7 (SD 2.1). Close to 1 out of 10 (8.8%) of the participants reported having 3 assets or less, while more than half (59.7%) reported having 7 or more of these assets. Slightly more than 10 percent (12.3%) were still working, about a quarter were retired (26.7%), while the remainder were either not currently working or had never worked.

The sample from Beirut was slightly older, more likely to be female, while the Zahle sample was more likely of Syrian origin. The Beirut sample also had higher levels of education, more assets, and was more likely to be retired, while the Zahle sample was more likely to not be working or have never worked (66.7%).

** Sum of the following 10 assets: functional water heater, washing machine, satellite dish or cable, microwave, dishwasher, air conditioner, oven, refrigerator, car or truck, and access to internet.

A summary of health- and cognition-related characteristics is presented in Table 3. Almost half the sample (48.7%) were current smokers, and three-quarters were either overweight (34.6%) or obese (37.7%). Hypertension (57.2%) and high cholesterol (49.0%) were the most prevalent chronic health conditions, followed by diabetes (32.7%) and heart disease (32.4%). About half (51.4%) and a third (34.1%) reported moderate to severe symptoms of depression and anxiety, respectively. Almost one in ten (9.1%) of the participants reported difficulties in at least one ADL, and one in seven (13.9%) reported difficulties in at least one IADL. More than two-thirds of the sample reported to be in fair (48.1%) or poor (18.7%) self-rated health.

The health profiles were fairly similar, although the Beirut sample reported a somewhat higher prevalence of high cholesterol, heart disease, diabetes, cancer and psychiatric conditions, possibly reflecting more access to the health care system. In contrast, the Zahle sample reported a considerably higher prevalence of moderate to severe symptoms of depression and anxiety, ADL or IADL disability, and poor subjective health.

[Insert Table 3]

In terms of markers of cognitive functioning, over half the sample (52.1%) reported having fair or poor memory while only about 1 in 6 (16.0%) reported having excellent or very good memory. About half (49.8%) also reported having worse or much worse memory compared to 2 years ago while 45.3 percent described their memory to be about the same as two years ago. Data from informants indicated a decline in mental health functioning in 7.6 percent and difficulties in remembering things in 30.4 percent of the participants. Self-rated and informant-rated evaluations of memory and mental health functioning were considerably lower in the Zahle sample than the Beirut sample.

LSAHA represents the first large-scale epidemiological study in the Arab region that includes a comprehensive, validated battery to characterize cognitive functioning and identify cases of ADRD in the general older adult population. LSAHA will address the paucity of robust epidemiologic data from well-designed population studies in the MENA region (El Metwally, 2019) which is projected to face some of the highest rates of increase in ADRD globally [4]. Most existing data on ADRD in the Arab region come from a few cross-sectional surveys, mostly in Egypt, relying primarily on short screening instruments to identify ADRD cases, or on hospital-based case series on nursing homes residents [7, 8]. Building on a previous pilot study of about 500 older adults in Lebanon [11, 36], LSAHA is designed as a longitudinal study to provide critically new information on the prevalence and incidence of ADRD, with a comprehensive and well-validated battery of cognitive tests and a large sample size to produce sufficient power to establish the role of the primary risk factors in this population. Our objective is not only to examine the role of established ADRD risk factors, such as age, educational attainment and chronic medical conditions including hypertension and diabetes, but also to investigate novel risk factors that are relevant in this setting, in particular exposure to political instability and violence and its economic ramifications during prolonged periods of the life course.

We achieved our recruitment goal of 3,000 participants aged ≥ 60 years, enrolling a total of 3,027 participants. We faced two important challenges in developing a sampling design that would yield a representative sample of the older adult population in Lebanon, mainly due to the lack of a population census and the security situation in certain areas of the country. Moreover, data collection in the field started on October 9, 2023, two days after the start of the conflict in Gaza. This required us to make further adjustments to the sampling strategy, as additional areas that originally were deemed safe were no longer available for the sampling plan. We also invested increased efforts to inform political and security stakeholders and conducted general media and local community outreach activities to announce the study in sampled study areas. To maximize diversity, our sampling plan focused on two purposefully selected study areas in Beirut and Zahle. Our final sample shows ample variability in the basic socio-demographic and health-related characteristics, providing confidence that we obtained a sample that is likely to be reasonably representative of the overall older adult population in Lebanon.

We found a high level of poor mental health in the sample, with a high prevalence of moderate to severe symptoms of depression and anxiety. These symptom levels fall in the upper range of what has been reported for other LMIC populations [37, 38], while being substantially lower in HIC populations [37, 39–41]. It seems likely that the renewed hostilities in Lebanon in the fall of 2023, coinciding with the start of our data collection, and the impact of the severe economic crisis of 2019 that left many older adults significantly impoverished, may have contributed to these findings. Both these factors are well-established determinants of poor mental health [42–46].

An important strength of LSAHA is that it has adopted the rigorous survey methods of the around 50 studies in the HRS Around The World Network [15]. LSAHA has also implemented the detailed assessment of multiple cognitive function domains as developed by HCAP in order to produce reliable ascertainment of ADRD cases in the general population [12, 13] and to provide a comprehensive assessment of baseline cognitive functioning for future studies of cognitive changes and decline. Through these international networks, we will be able to participate in comparative analyses regarding the epidemiology of ADRD across the globe. A particularly exciting opportunity in this regard relates to LSAHA’s unique focus on the role of prolonged exposure to political conflict and violence in dementia risk. This feature will enable us to identify contextually relevant risk factors and compare them with data on ADRD risk factors from other populations, particularly those with older generations that have experienced similar exposures in their life course, such as those in Northern Ireland [47] and Vietnam [48].

An important goal for LSAHA is to share these data with fellow scientists and stakeholders in Lebanon and around the world. A curated dataset will be made available on the Gateway to Global Aging platform (https://g2aging.org/home) in the near future. LSAHA data are expected to inform several immediate and long-term clinical care and policy goals in Lebanon and across the MENA region. Clinically, the validated cognitive test battery assessing multiple domains may provide the basis for new screening methods that can be integrated in health care settings to identify older adults at risk of cognitive decline and dementia. They will also contribute to a better understanding of the primary risk factors for dementia in this population, which may serve as a basis for screening and prevention strategies. From a policy perspective, LSAHA data will add to the evidence base for a national dementia strategy that will serve to guide workforce training, resource allocations and development of appropriate interventions for vulnerable and at-risk subgroups in the population. Finally, new information on ADRD risks associated with stressful exposures throughout the life course may be useful to the development of screening and prevention policies in other LMICs facing similar demographic, social, and political challenges.