Health Variables Are Informative in Screening for Mild Cognitive Impairment Among Elderly African Americans

Abstract

To aid primary care providers in identifying people at increased risk for cognitive decline, we explored the relative importance of health and demographic variables in detecting potential cognitive impairment using the Mini-Mental State Examination (MMSE). Participants were 94 older African Americans coming to see their primary care physicians for reasons other than cognitive complaints. Education was strongly associated with cognitive functioning. Among those with at least 9 years of education, patients with more vascular risk factors were at greater risk for mild cognitive impairment. For patients with fewer than 9 years of education, those with fewer prescribed medications were at increased risk for dementia. These results suggest that in addition to the MMSE, primary care physicians can make use of patients' health information to improve identification of patients at increased risk for cognitive impairment. With improved identification, physicians can implement strategies to mitigate the progression and impact of cognitive difficulties.

Keywords: MMSE; cognitive impairment; mild cognitive decline.

Figure 1

Classification tree categorizing normal, MCI, and dementia cognitive functioning using demographic and characteristics (age, gender, education, number of medications, and number of vascular risk factors). Note. The numbers below each classification outcome indicate the number of observations in the normal, MCI, and dementia class, respectively. For example, 47/37/10 in the first node indicates that, among the 94 patients (47 + 37 + 10) in the study, 47 patients were classified as having normal cognitive functioning, 37 as having MCI, and 10 as having dementia. Normal: Normal cognitive functioning (MMSE > 26); MCI: concern for mild cognitive impairment (20 < MMSE ≤ 26); Dementia: Concern for dementia (20 ≤ MMSE). MMSE = Mini-Mental State Examination.

Figure 2

Classification tree categorizing normal and nonnormal cognitive functioning using demographic and health characteristics (age, gender, education, number of medications, and number of vascular risk factors). Note. The numbers below each classification outcome indicate the number of observations in the normal and nonnormal class, respectively. For example, 28/11 in the left-mode node indicates that, among the 39 patients (28 + 11) who have at least 12 years of education, 28 were classified as having normal cognitive functioning, whereas 11 were classified as having nonnormal cognitive functioning. Normal: Normal cognitive functioning (MMSE > 26); Nonnormal: Concern for mild cognitive impairment or dementia (MMSE ≤ 26). MMSE = Mini-Mental State Examination.

Figure 3

Variable importance estimated for each classification tree model. Model 1: Normal/MCI/dementia classification. Model 2: Normal/nonnormal classification. Note. Meds = number of medication; Edu = education level; Vrisks = number of vascular risk factors.