Preserving Cognition in Older African Americans with Mild Cognitive Impairment

Abstract

To the Editor: Twenty-five percent of older adults in the United States have type 2 diabetes mellitus, which increases their risk for cognitive decline.1, 2 This risk disproportionately affects African Americans because they have higher rates of diabetes mellitus than whites and worse glycemic control, which magnifies their risk.3 These disparities contribute to why African Americans have twice the rate of Alzheimer's disease as whites.4 There are now 1 million older African Americans with diabetes mellitus in the United States, and their number will double by 2030.5 This will multiply the burden of cognitive impairment in older African Americans and necessitates preventive interventions. A randomized controlled trial is being conducted to test the efficacy of behavioral activation to reduce cognitive decline in African Americans with mild cognitive impairment. Behavioral activation is a standardized, evidence-based treatment often used to treat depression.6 This approach has been applied to increase participation in physical, cognitive, and social activities to preserve cognition. Controls received supportive therapy, which facilitates personal expression but contains no other active elements. Race-concordant community health workers deliver six in-home treatment sessions of both interventions. Because the relationship between diabetes mellitus and cognition is of interest, treatment effects on a subgroup of participants with diabetes mellitus who have had 12-month assessments were examined (n = 35). The Hopkins Verbal Learning Test—Revised, Delayed Recall (HVLT) and the Uniform Data Set (UDS) Neuropsychological Battery, which includes tests assessing verbal memory, attention, semantic memory, language, processing speed, visuospatial ability, and executive function, were administered (masked to treatment assignment) at baseline and 12 months. Standardized z scores from the component UDS tests yield the UDS composite score (UDS-CS), which reflects global cognition. The pilot data indicate that a decrement of 0.45 UDS-CS points translates clinically to inability to balance a checkbook, do laundry, shop, or prepare meals. The average age of participants was 75.1 ± 6.8; 28 (80%) were female. Table 1 shows that the scores of participants who received behavioral activation improved over 12 months in two cognitive domains: episodic memory (HVLT score) and executive function and processing speed (Digit Symbol Substitution Test). Supportive therapy controls had no comparable improvements. A greater proportion of participants who received behavioral activation than of controls improved by 0.45 points or more on the UDS-CS (40.0% vs 21.1%; odds ratio (OR) = 2.50, 955 CI = 0.55–11.33), and fewer declined by 0.45 points or more (6.7% vs 21.1%; OR = 0.27, 95% CI = 0.03—2.70). There were no significant changes in other neuropsychological test scores. The treatment trends observed for three cognitive measures in this sample of older African Americans with diabetes mellitus and mild cognitive impairment suggest that behavioral activation may preserve cognition in this high-risk group, although the sample was small, and the analysis was underpowered to detect treatment group differences. Nevertheless, these preliminary findings concur with the results of previous studies indicating that physical, cognitive, and social activity may prevent cognitive decline and extend these findings to older African Americans with diabetes mellitus.7 The latter finding is important because the number and longevity of this population, and their risk of progressive cognitive decline and dementia, are increasing. The behavioral activation treatment approach targets two entrenched health problems in older African Americans simultaneously (diabetes and impaired cognition) and may prevent cognitive decline by increasing cognitive (e.g., problem-solving) and physical (e.g., exercise) activities and perhaps by improving glycemic control. These two mechanisms are plausible because physical and cognitive inactivity, as well as poor glycemic control, increase the risk of cognitive decline.8, 9 Because there is no disease-modifying treatment for Alzheimer's disease, if an early nonpharmacological treatment such as behavioral activation delays cognitive decline even by 1 year, it will greatly reduce the personal and societal costs of Alzheimer's disease in older African Americans.10 This research is supported by Grant 1R01AG035025 from the National Institute on Aging. ClinicalTrials.gov Identifier: NCT01299766 Conflict of Interest: The authors declare no competing interests. Author Contributions: Rovner, Casten: study concept and design, data analysis and interpretation, manuscript preparation. Sponsor's Role: None.