Scope of the Problem

Dementia is a loss of cognitive abilities in multiple domains that results in impairment in normal activities of daily living and loss of independence. There are a number of diseases that cause dementia, but the most common is Alzheimer’s disease (AD), which is responsible for 60 to 80 percent of all dementia.¹ AD is a neurodegenerative disorder that begins in the mesial temporal lobe causing memory loss, but pathology soon spreads into other brain regions causing dementia. AD is defined pathologically by the presence of cerebral atrophy, extracellular amyloid plaques, and intraneuronal neurofibrillary tangles.

The major risk factor for AD is age, with the prevalence doubling every 5 years after the age of 65.² The prevalence of AD in 2002 was estimated to be 2.3 million individuals over age 70, based on a national population-based sample.³ Another group estimated the prevalence of AD in 2000 at 4.5 million individuals aged 65 and older, based on a U.S. regional sample.⁴ This latter figure was updated to an estimated 5.3 million individuals with AD in 2008.¹ This translates into about one in 8 to 10 persons over the age of 65 suffering from AD. The worldwide prevalence of dementia is estimated to be 35.6 million in 2010, with the number exceeding 65 million in 2030 and 115 million in 2050, making it a pressing global health concern.⁵

The diagnosis of AD is challenging both clinically and pathologically. There are multiple sets of well-established criteria for the clinical diagnosis of AD.^6–9 There is some variation among the sets of criteria, but each requires evidence of cognitive and functional decline that impacts the individual’s ability to carry out routine daily activities. Even though significant progress has been made in identifying imaging, cerebrospinal fluid, and blood markers of disease, the diagnosis of AD during life currently is based primarily on the phenotypic presentation of cognitive and functional decline. However, the diagnosis of “definite AD” requires neuropathologic confirmation, which is complicated by the fact that AD-specific pathology infrequently occurs in isolation. Typically other pathology such as various types of vascular lesions or Lewy bodies are present, and these may increase with advancing age.¹⁰ The presence of these multiple pathologies suggest that dementia is often due to multiple causes and not solely AD or any other single etiology. The correlation between AD pathology and cognitive symptoms is limited, providing further support for the idea that the other pathologies present also contribute to the cognitive presentation.

The term cognitive decline covers a continuum of cognitive changes, some of which are considered to be within the spectrum of normal aging and others that exceed expected decline for normal aging and are categorized as mild impairment. Typically performance in one or more cognitive domains such as memory, orientation, language, executive function, and praxis are assessed to determine decline. The diagnostic threshold between normal and pathological cognitive changes is imprecise. Pathological cognitive decline is often referred to as mild cognitive impairment (MCI) or cognitive impairment not demented (CIND); each term has multiple subtypes reflecting the construct of multiple etiologies. The diagnostic criteria for MCI and CIND are still evolving, but guidelines generally include greater than expected cognitive decline for the individual’s age and education level, and no more than mild functional impairment insufficient to meet the threshold for a diagnosis of dementia.

The number of individuals with MCI or CIND exceeds that of AD. The prevalence of CIND in 2002 was estimated to be 5.33 million individuals over age 70 (22 percent), based on a nationally representative sample.¹¹ These individuals are at increased risk for development of AD, which makes them a prime target for intervention protocols. Longitudinal followup has shown that these individuals progress to dementia at a rate of 12 to 15 percent per year compared to 1 to 2 percent among cognitively healthy older adults.^11,12

AD has a wide-ranging impact on individuals, families, and healthcare systems. It causes severe suffering for patients, including progressive functional impairment, loss of independence, emotional distress, and behavioral symptoms. Dementia is associated with a greater burden of co-existing medical illness,¹³ nursing home placement,¹⁴ and increased mortality.^15,16 Families and caregivers often experience emotional and financial stress. Up to 50 percent of caregivers suffer from significant psychological distress¹⁷ and incur > $18,000 annually in unreimbursed caregiving expenses (in 1998 dollars).¹⁸ In 2005, the direct cost to Medicare and Medicaid for care of people with AD and other dementias was $111 billion, and indirect costs to business for employees who were caregivers for individuals with AD and other dementia was estimated at $36.5 billion.¹ According to the World Alzheimer Report 2009,⁵ AD accounts for 4.1 percent of all disability-adjusted life years (DALY) among those over 60, and cognitive impairment is the health condition that is most strongly associated with institutionalization.

AD is progressive, and it is the sixth leading cause of death for people of all ages in the United States.¹⁹ Current therapies provide only modest symptomatic benefit, so methods to delay onset and/or modify progression are crucially needed.

Risk Factors for Alzheimer’s Disease and Cognitive Decline

Findings from numerous epidemiological and clinical studies suggest that multiple biological, behavioral, social, and environmental factors may contribute to the delay or prevention of cognitive decline and AD. Studies of selected risk or protective factors for cognitive decline and AD have been published, but few systematic reviews have examined the quality of these studies and the conclusions that can be drawn from the data. It is not clear whether the results of these previous studies are of sufficient strength to warrant specific recommendations for behavioral, lifestyle, or pharmaceutical interventions/modifications targeted to these endpoints.

Researchers face several challenges in accurately identifying factors that alter the risk of cognitive decline and AD. The pathological changes for AD begin years prior to the overt clinical symptoms currently recognized as the hallmark of the disease,^20,21 suggesting that the relevant exposures may cover much of the lifespan. Some exposures may alter risk of later cognitive disease only at certain time points in life. For these factors, there may be a limited window of time when interventions will be effective. The quality of the measurement of exposures varies considerably across studies, and the effect of this is magnified when one considers the potential inter-relationship between many of the factors. The criterion standard procedures and criteria for the cognitive outcomes differ across studies. In addition, the domains assessed and the specific cognitive measures used differ across studies, making it difficult to synthesize findings from multiple sources. There is evidence that the effect of some putative risk factors (e.g., diabetes and psychological stress) may vary by subgroups such as race and/or sex,^22,23 necessitating careful consideration of sample characteristics in the search for interventions. Finally, not all exposures under consideration are readily modified. Identifying the association between cognitive outcomes and all of the factors investigated in this report is the goal, but an added benefit would be realized by identifying those factors that can realistically be modified.

Purpose of this Evidence Report

The National Institutes of Health (NIH) Office of Medical Applications of Research (OMAR) reviews and evaluates clinically relevant NIH research program information and promotes the effective transfer of this information to the health care community. OMAR accomplishes this objective through its Consensus Development Program. This includes major Consensus Development Conferences and State-of-the-Science Conferences when less definitive evidence is available.

As background for a State-of-the-Science Conference scheduled for April 2010, OMAR and the National Institute on Aging commissioned this evidence report on “Preventing Alzheimer’s Disease and Cognitive Decline” through the Agency for Healthcare Research and Quality (AHRQ). The key research questions were developed by a planning committee. The aim of the report is to summarize the available literature, frame the discussions regarding potential risk factors, highlight the limitations of the evidence base, and identify areas for future research. Through this report, OMAR seeks to increase the scientific rigor of its State-of-the-Science Conference. The focus of the this conference will be on evaluating existing data to determine whether there is sufficient quality of evidence to warrant any specific recommendations at this time for behavioral, lifestyle, or pharmaceutical interventions/modifications targeted at AD and cognitive decline in later life, and to identify needs for additional research.

The findings of our review clarify what is known about factors that modify the risk of AD or cognitive decline as a means of providing authoritative background information for participants at the State-of-the-Science Conference. More broadly, we expect that our findings will be useful to major stakeholders in this arena, including policymakers, advocacy groups, community organizations, health care providers, and mid- to late-life adults. We also identify future research priorities, which may be useful to government agencies and private sector funding organizations.

Role of the Technical Expert Panel

We identified experts in the field of Alzheimer’s disease and cognitive decline to serve as members of the project’s Technical Expert Panel (TEP). The TEP contributes to AHRQ’s broader goals of (1) creating and maintaining science partnerships and public-private partnerships; and (2) meeting the needs of an array of potential customers and users of this product. To ensure accountability and scientifically relevant work, we asked the TEP for advice at key stages of the project. More specifically, TEP members participated in conference calls and email exchanges to refine the analytic framework and key questions at the beginning of the project, refine the scope of the project, discuss inclusion and exclusion criteria, and provide expert advice on methodology.

Members of the TEP were:

• Jesse A. Berlin, Sc.D. – Research & Development at Johnson & Johnson Pharmaceuticals
• Ornit Chiba-Falek, Ph.D. – Duke University
• John Ioannidis, M.D. – University of Ioannina, Greece
• Dan Kaufer, M.D. – University of North Carolina, Chapel Hill
• Michael Marsiske, Ph.D. – University of Florida

These individuals represent a broad range of specialties relevant to our topic (including neurology, psychology, statistics, and genetics). Because of their extensive knowledge of the literature on these topics, TEP members were also invited to participate in the peer review of this draft report.

Organization of this Report

Chapter 2 describes the methods used to produce this report, including the key questions addressed, the analytic framework, our search strategies, and inclusion/exclusion criteria. In Chapter 3, we report on the numbers of publications reviewed and present the results of our literature search and synthesis on the six key questions that OMAR posed for this review. Chapter 4 discusses these findings further, highlights methodological shortcomings of the extant research, and offers recommendations for future research. Chapter 5 summarizes the major conclusions.