Behavior-Based Interventions to Enhance Cognitive Functioning and Independence in Older Adults

Abstract

THE CURRENT AND PROJECTED FINANCIAL AND EMOtional burden of cognitive impairment to individuals and their families, as well as the financial burden to society, are staggering. There are 24 million individuals with dementia in the world, and 4.6 million new cases are diagnosed annually. The combined worldwide prevalence of age-associated cognitive impairment and cognitive disorders is predicted to reach 84 million individuals affected by 2040. In 2000, more than 4 million US adults had Alzheimer disease; by 2050, this will increase to 13 million. Furthermore, declines in specific cognitive domains (eg, memory, executive functions) are predictive of deficits in the performance of instrumental activities of daily living (IADL) in older adults, an outcome that seriously threatens the ability of the aging population to live independently. A growing number of studies of various pharmacologically based interventions hold promise for the postponement or limitation of cognitive decline and its consequences, and new drugs are under development. However, only modest success has been achieved to date, with available drugs producing generally low to moderate improvements in cognition and IADL, effects that are further limited by drug adverse effects and nonadherence. Other biological treatments formerly believed to be promising, such as hormone (estrogen) therapy, vitamin E, and nonsteroidal anti-inflammatory drugs, have proven to be neutral or detrimental. Less common than the studies of biological treatments are investigations of nonpharmacological interventions, such as memory training, other cognitive training, or increased physical activity. Although such interventions are recommended by physicians for general cognitive protection, supporting evidence for their efficacy is limited. Welldesigned randomized clinical trials are needed to identify additional nonbiological mechanisms that may prevent, reduce, or delay cognitive decline in older adults. The Advanced Cognitive Training for Independent and Vital Elderly (ACTIVE) study, reported by Willis and colleagues in this issue of JAMA, represents an important step in this direction. The ACTIVE study is, to date, the only long-term randomized controlled trial in which the effects of cognitive enhancement interventions were assessed after 5 years for both cognitive and functional status. The design of the ACTIVE study and its results provide insights into the potential promise of behavior-based interventions to improve cognition, as well as their consequences for improving functional status. The ACTIVE study also underscores the complexities and challenges inherent in this type of research, the difficulty in interpreting the results, and the problems associated with applying these results to clinical care. ACTIVE is a well-designed study in which a large community-based cohort of 2802 participants was enrolled and followed up for 5 years. The retention rate (67% at 5 years) is comparable with other trials with similar populations of older adults. The ACTIVE investigators found that a relatively short (10-session) training intervention for specific cognitive functions (memory, reasoning, and speed of processing) produced immediate improvements in these domains compared with a nonintervention control group, replicating prior research findings. More important, these cognitive improvements were sustained after 5 years of follow-up, although none of these improvements had effects beyond the specific cognitive domains of the intervention, either immediately or after the 5-year follow-up period. Also, results addressing the investigators’ primary hypothesis— that cognitive training would delay declines in functional status measured by self-reported IADL scores and performance assessments—were less clear and less compelling. Only participants who underwent reasoning training (verbal episodic) self-reported significantly higher IADL functioning compared with the control group. The remaining 2 intervention groups had higher, but nonsignificant, selfreported IADL scores than the control group. None of the original groups (intent-to-treat cohort) demonstrated significant improvements in the performance-based measures.