We sought to determine the association of limited English proficiency (LEP) with suspected undiagnosed Alzheimer’s disease and related dementias (ADRD).
Individuals with LEP are less likely to have usual sources of medical care, referral to specialty services like neurology, and to receive appropriate diagnostic testing compared to the English-proficient. ADRD underdiagnosis is well-established among minoritized populations and women, but the association among patients with LEP remains unexplored.
We used 2003-2021 Medical Expenditure Panel Survey (MEPS) data restricted to individuals older than 65 years without ICD-9 and ICD-10 diagnoses of ADRD. Suspected undiagnosed ADRD was determined if a respondent reported a cognitive limitation impacting daily functioning without an ADRD diagnosis. LEP status was determined if the clinical visit occurred in a language other than English. Weighted multivariate logistic regression models were performed between LEP and suspected undiagnosed ADRD. We additionally performed stratified analyses by race, ethnicity, insurance status, and healthcare setting type (i.e., ambulatory visits only).
In descriptive analysis, there was a greater prevalence of undiagnosed ADRD in those with LEP status versus the English-proficient (24.6% vs 12.7%; p<0.01). In multivariate logistic regression, LEP was significantly associated with 27% greater odds of suspected undiagnosed ADRD (OR: 1.27, 95%= 1.20, 1.33, p<0.001). In analyses stratified by race, LEP was linked to the highest odds of suspected undiagnosed ADRD among Black individuals (OR: 3.21, 95%= 2.75, 3.76, p<0.001). LEP status remained associated with an increased likelihood of undiagnosed ADRD in the publicly insured (OR: 1.23, 95%=1.16, 1.30, p<0.001) but not privately insured (OR: 0.95, 95% 0.84, 1.08, p>0.05).