Cognitive StatusCognitive Status
Assessing older adults from ethnic minorities for cognitive impairment, dementia, and delirium presents a number of challenges including: finding suitable interpreters when the elder’s command of English is poor; the variable beliefs related to cognitive loss with age in different cultures; how to approach the decision to institutionalize; and the ethical issues pertaining to medical decision-making (Yeo, 1996 ). Typically, a cutpoint score is employed as a way to standardize when cognitive impairment is determined to be significant.
Two commonly used instruments to assess cognitive status are discussed below in respect to ethnicity:
- The Mini-Mental State Examination (MMSE)
- The Short Portable Mental Status Questionnaire (SPMSQ)
The MMSE has been critizized for its lack of sensitivity to lower education level individuals and studies have shown that certain items in the instrument are difficult to administer (Rosselli, Tappen, Williams & Salvatierra, (2006). Hispanic Americans and persons with less than 8 years of formal education tend to be falsely identified as possibly cognitively impaired when using the MMSE (Mouton & Esparza, 2001; Mungas, Marshall, Weldon, Haan, & Reed, 1996; Crum, Anthony, Bassett, & Folstein, 1993). Clock drawing have been used more often because they have been shown to be more sensitive to sociocultural influences (Le Rue, Romero, Ortiz, et al, 1999). Similarly, clock-drawing also has shown sensitivity of the MMSE to mild cognitive impairment syndromes in non Alzheimer’s type dementias (De Jager, Hogervorst, Combrinck, et al., 2003).
Findings from a study of Mexican American elderly in the Southwest show that the degree f affiliation with the Mexican-American ethnicity correlates with poor performance of cognitive impairment (Heller, Briones, Schiffer, Guerrero, et al, 2006), even after controlling for education, age and gender. The authors were unable to make conclusions to their findings however suggest that there are some unknown determinants of late life cognitive impairement in these populations (Heller et al, 2006).
Among older African-Americans, Hispanic-Americans, and persons with educational attainment less than high school, a lower threshold score for determination of cognitive impairment has been recommended (less than 18 out of a possible 30 points) to improve sensitivity (82%) and specificity (99%) for the diagnosis of dementia dementia (Bohnstedt, Fox, & Kohatsu, 1994; Baker, 1996; Leveille et al., 1998).
A standard cutpoint of 23 or less to determine cognitive impairment tends to overestimate of the number of Hispanics with true impairment of cognitive function. The Hispanic EPESE indicated that when the standard MMSE threshold score of 23 was used, 22.3% of Mexican-American older adults were classified as cognitively impaired, but this high rate of cognitive impairment may reflect the lack of schooling (Majurin et al., 2000; Marshall, Mungas, Weldon, Reed, & Haan, 1997).
The SPMSQ has been specifically validated in older African-American and Hispanic-American samples with excellent sensitivity and specificity (Pfeiffer, 1975; Fillenbaum, Heyman, Williams, Prosnit, & Burchett, 1990; Welsh et al., 1995; Mouton & Esparza., 2001). Spanish language versions are available for both measures. See Mungas, Marshall, Weldon, et al., 1996 and Marshall, Mungas, D., Weldon, M., 1997 for information on the Spanish language version of the MMSE.
The Spanish and English Neuropsychological Assessement Scale (SENAS) was initially developed specifically to address bias in measurement of cognitive abilities for different ethnic groups. (Mungas, Reed, Crane, Haan, Gonzalez, 2004). For more detailed descripton of the SENAS instrument see Mungas, Reed, Marshall, & Gonzales, (2000). In addition to the cognitive screening measures, a Neuropsychological Screening Battery for Hispanics (NeSBHIS) has been developed, standardized, and normed by age for Hispanics. See Pontón, Satz, Herrera et al.(1996).
Using the H-EPESE data researchers examined the association with cognitive decline and Mexican Americans with diabetes (Rotkiewicz-Piorun, Snih, Raji, Kuo, & Markides, 2006). Findings revealed that male Hispanics with high depressive symptoms and diabetic complications were likely to have greater declines in their MMSE scores over time.
Back To Top |Depression
In most cases, there is little information on how standardized assessment instruments for the measure of depression perform for older adults from ethnic minority groups. Two widely used measures are (1) the Centers for Epidemiologic Studies Depression Scale (CES-D) and (2) the Geriatric Depression Scale (GDS).
The CES-D is a 20 item questionnaire designed to measure depressive symptoms in a community-based samples (Radloff, 1977; Golding & Aneshensel, 1989; Golding, Aneshensel, & Hough, 1991). Reliability estimates for the CES-D are high, ranging from 0.84 to 0.92. Samples of African-Americans, Hispanic Americans, and other diverse groups have shown that the CES-D can usefully measure depression (Mouton, Johnson, & Cole, 1995). See Golding & Aneshensel (1989), and Golding, Aneshenel, & Hough (1991) for information on the Spanish language CES-D.
The GDS has good sensitivity and specificity in most samples, although it appears to be have poorer performance among African-Americans and Hispanic Americanswhen compared to whites(Baker, Espino, Robinson, & et al., 1993). The GDS was also less sensitive to significant depression among Hispanic-Americans (Baker et al., 1993; Baker & Espino, 1997) See Baker and Espino (1997) for information on the Spanish language GDS.
Functional Assessment
In the domain of functional assessment, we may find that the willingness to report difficulty taking care of oneself may be powerfully related to fear of admitting one’s dependence on others by older persons from certain groups. Observed differences in functional status across ethnic groups may represent true differences but could represent measurement error from the instrument used in the assessment of physical function. Physical function assessment generally employ self-report instruments that rely on the subjective response of patients. Performance-based measures provide more objective measures of function, but are harder to carry out in the clinical setting and may not always relate directly to performance at home (Guralnik, Branch, Cummings, & Curb, 1989; Guralnik, Reuben, Buchner, & Ferrucci, 1995). The choice of method generally relates to the time constraints on the clinician, the training of the clinician (and staff), and need for the most reliable and valid information.
