Rates of Undiagnosed Cognitive Impairment and Performance on the Montreal Cognitive Assessment Among Older Adults in Primary Care.

Abstract

The number of adults in the USA with cognitive impairment is increasing; however, few studies report prevalence rates of undiagnosed cognitive impairment among older adults in primary care. To determine the prevalence of undiagnosed cognitive impairment among adults ages 55years and older in primary care settings and provide normative data for the Montreal Cognitive Assessment in this context. Single interview, observational study. English-speaking adults ages 55years and older without diagnoses of cognitive impairment recruited from primary care practices in New York City, NY, and Chicago, IL (n = 872). Montreal Cognitive Assessment (MoCA). Undiagnosed cognitive impairment was defined by age and education adjusted z-scores more than 1.0 and 1.5 standard deviations below published norms, corresponding to mild or moderate to severe cognitive impairment, respectively. The mean age was 66.8 (8.0) years, 44.7% were male, 32.9% were Black or African-American, and 29.1% were Latinx. Undiagnosed cognitive impairment was identified in 20.8% of subjects (mild impairment, 10.5%; moderate-severe impairment, 10.3%). Impairment at any level of severity was associated in bivariate analyses with several patient characteristics, most notably for race and ethnicity (White, non-Latinx, 6.9% vs. Black, non-Latinx, 26.8%, Latinx, 28.2%, other race, 21.9%; p < 0.0001), place of birth (US 17.5% vs. non-US 30.7%, p < 0.0001), depression (33.1% vs. no depression, 18.1%; p < 0.0001), and impairment in activities of daily living (≥ 1 ADL impairment, 34.0% vs. no ADL impairment, 18.2%; p < 0.0001). Undiagnosed cognitive impairment is common among urban dwelling older adults attending primary care practices, and was associated with several patient characteristics, including non-White race and ethnicity and depression. Normative data for the MoCA from this study may serve as a useful resource for studies of similar patient populations.