PROJECT SUMMARY/ABSTRACT
The national prevalence of diagnosed and undiagnosed Alzheimer’s disease and related dementias (ADRD)
among Arab Americans is unknown. Estimates are unknown because Arab Americans are considered non-
Hispanic (NH) whites in national health surveys. By using innovative methods to link National Health Interview
Survey (NHIS) and Medical Expenditure Panel Survey (MEPS) data, we can disaggregate Arab Americans
from the NH white classification to uncover ADRD estimates. Specific aims are to: 1) estimate the age- and
sex-adjusted prevalence of diagnosed ADRD among foreign-born Arab Americans compared to US-born NH
whites and foreign-born NH whites, NH blacks, NH Asians and Hispanics; 2) examine associations between
race, ethnicity and nativity status and the prevalence of diagnosed ADRD before and after controlling for
covariates; 3) estimate the age- and sex-adjusted prevalence of undiagnosed ADRD among Arab Americans
compared to the aforementioned groups; and 4) examine associations between race, ethnicity and nativity
status and the prevalence of undiagnosed ADRD before and after controlling for covariates. Secondary,
cross-sectional data from 2000-2016 NHIS and 2001-2017 MEPS data sources will be combined and
analyzed. Responses will be matched, and a linked dataset will be created by the Agency for Healthcare
Research and Quality. The sample will include adults >45 years. The independent variable will be created
by combining questions assessing race, ethnicity and nativity status from NHIS. Foreign-born Arab
Americans will be compared to US-born NH whites and foreign-born NH whites, NH blacks, NH Asians and
Hispanics. For Aims 1 and 2, the dependent variable is diagnosed ADRD, measured using data from
MEPS medical condition files. Adults with ICD-9 codes 290, 294, 331, 797 (2001-2015) or ICD-10 codes F03
or G30 (2016-2017) will be classified as has having diagnosed ADRD. For Aims 3 and 4, the dependent
variable is undiagnosed ADRD, measured using MEPS medical condition and household component
files. The MEPS asks three questions to estimate cognitive limitations, an indicator of ADRD. Adults will be
categorized as having undiagnosed ADRD if: 1) they responded “yes” to one or more questions on cognitive
limitations and 2) do not have diagnosed ADRD. Age and sex adjusted prevalence of diagnosed ADRD (Aim 1)
and undiagnosed ADRD (Aim 3) will be calculated for Arab Americans compared to US-born NH whites and
foreign-born NH whites, blacks, Hispanics and Asians. Logistic regression will be used to determine
associations between the combined race, ethnicity and nativity status variable and diagnosed ADRD (Aim 2),
and undiagnosed ADRD (Aim 4) before and after controlling for covariates. The innovative method used in this
study will contribute the first estimates of diagnosed and undiagnosed ADRD among Arab Americans to the
medical literature. Results will help to facilitate policy decisions and advocate for resources to be allocated for
interventions to improve the health of this underrepresented but increasingly visible vulnerable population.