Can Certain Factors Protect Against or Increase Risk of AD?

Can Certain Factors Protect Against or Increase Risk of AD?

We’ve known for some time that certain genetic and nongenetic factors can increase the risk of developing AD. Recent evidence has suggested that other factors may actually help to reduce AD risk (see p. 14 for more on one of these factors—cognitive reserve). The combined weight of these advances is making scientists eager to understand how these risk and protective factors balance each other over the course of a lifetime. They also want to know more about the types of interventions that may be useful in changing this balance in the direction of healthy aging and the times in the life cycle at which these interventions might be most effective.

Ideas about what these risk and protective factors might be are derived from a variety of different types of studies, including genetics studies, studies of individual lifestyles and behavioral patterns, and studies of large groups or populations. Findings from these studies are important because they point the way to potential therapeutic approaches that might be worth investigating in controlled clinical trials. If confirmed in trials, they may suggest ways that people can change their lifestyles or environments to reduce risk. The genetics studies, in particular, will help identify pathways that affect the development or progression of AD. They also will help researchers develop new animal models to understand early events in the disease, as well as identify potential targets for treatment and prevention.

New Discoveries About AD Genetics
Genetic studies of complex neurodegenerative diseases such as AD have focused on two key issues—whether a gene might influence a person’s overall risk of developing a disease and whether a gene might influence some particular aspect of a person’s risk, such as the age at which the disease begins (“age at onset”). To date, only four of the approximately 30,000 genes in the human genetic map (the “genome”) have been conclusively shown to affect AD development. Mutations in three genes—the APP gene found on chromosome 21, the PS1 gene on chromosome 14, or the PS2 gene on chromosome 1—are linked to the rare early-onset form of familial AD. The APP gene is responsible for making APP, the precursor to beta-amyloid. The presenilin genes code for proteins that are components of enzymes that play an important part in cleaving APP to form beta-amyloid. Presenilin gene mutations promote the breakdown of APP, leading to increased production of harmful beta-amyloid.

A fourth gene on chromosome 19 encodes a protein called apolipoprotein E (ApoE). ApoE carries lipids in the bloodstream and is important in clearing lipids from the blood. APOE, the gene that encodes ApoE, has three common forms, or alleles—e2,e3, ande4. Thee4 allele is a risk factor gene for the common late-onset AD. Thee2 allele may provide some protection against AD ande3 is thought to play a neutral role.

Scientists estimate that an additional four to seven risk factor genes exist for late-onset AD. A study from researchers at Massachusetts General Hospital, who are participating in NIMH’s Alzheimer Disease Genetics Initiative and who are also supported by NIA funding, is shedding light on these genes. The researchers screened the entire genomes of a large number of families with AD using 382 genetic markers and complex statistical analysis to identify regions of the genome that were associated with AD and might contain additional susceptibility genes (Blacker et al., 2003). The study identified 12 additional chromosomal regions that might be linked to AD. Some of these regions will probably be found not to contribute to AD. However, other regions may well harbor genuine AD susceptibility genes, particularly the regions that yielded strong statistical evidence during the genome analysis. Even though it will be difficult to identify and characterize the genes in these chromosomal regions, the results will greatly facilitate the development of strategies for AD treatment, early intervention, and prevention.

Another team of researchers examined genetic associations with AD in a different population. This Boston University School of Medicine team worked with a population in Wadi Ara, an Arab community in northern Israel that has an unusually high prevalence of AD (Farrer et al., 2003). A genomic scan conducted on people from this community with and without AD revealed markers with significant AD allelic association on chromosomes 2, 9, 10, and 12. The researchers then analyzed the distribution of allele frequencies to narrow the potential regions on these chromosomes where the genes might be found. The unique characteristics of the Wadi Ara populations and the fact that findings from this analysis replicate those from other genome scans may help scientists more rapidly identify AD risk factor genes on these chromosomes.

Other studies have shown that a region on chromosome 10 is likely to harbor at least one AD risk factor gene and several research teams have made advances in this area. For example, scientists from the Karolinska Institute in Stockholm, Sweden, examined a stretch of genetic material on chromosome 10 that contains the insulin degrading enzyme (IDE) gene (Prince et al., 2003). This enzyme is of interest because it also degrades beta-amyloid. Genetically engineered mice that do not have the IDE gene develop high insulin levels, glucose intolerance, and increased brain levels of beta-amyloid (see p. 52 for more on the links between AD, insulin, and diabetes). The scientists compared genetic material from people with and without AD to assess the IDE gene and two other close-by genes. They were interested in finding differences between individuals at a single point in the genetic code (these points are called SNPs) and in stretches of DNA that are inherited in common among groups of people (these stretches of DNA are called haplotypes). Results strongly indicated that this region contains alleles and haplotypes that confer AD risk. These findings provide substantial evidence that genetic variation within or extremely close to IDE affects both disease risk and traits related to the severity of the disease. The study also indicated that an analysis of this type can be an effective way to assess genetic variation in complex diseases like AD.

Investigators from the Veterans Affairs Puget Sound Health Care System, in Seattle, explored the connection between IDE and APOE status in a study with cognitivelyhealthy older people and those with AD (Craft et al., 2003). The scientists gave participants infusions of insulin at five different levels and then measured cognitive performance and APP levels in the blood 2 hours later. The results supported a role for insulin in normal memory function and on APP levels and suggested that the APOE-e4 allele interacts with insulin to affect APP levels. This team of investigators explored these relationships further in a second study (Cook et al., 2003). In light of the possible association of IDE with late-onset AD and the evidence that APOE status may affect insulin metabolism, they hypothesized that the expression of the IDE gene may be altered in people with AD. To test this hypothesis, they measured the expression of IDE in hippocampal tissue from the brains of individuals who were cognitively healthy and others who had AD. They found that the hippocampal IDE protein was reduced by about half in those with AD who carried the APOE-e4 allele compared to the participants with AD who did not carry the APOE-e4 allele or to cognitively healthy participants. These findings show that reduced IDE expression is associated with APOE-e4 and suggest that IDE may interact with APOE-e4 status to affect beta-amyloid metabolism. Consistent with these findings, scientists at the University of California at Los Angeles found that deficient insulin signaling correlated with reduced IDE in the brains of people with AD, as well as in the brains of a mouse model of AD. The authors hypothesize that enhancing the effects of insulin in the brain may reduce amyloid accumulation and is a promising strategy to pursue against AD (Zhao et al., 2004).

Studies of chromosome 10 also are allowing scientists to explore another intriguing aspect of AD—age of onset. A research team at Duke University Medical Center began their study by focusing on a region of chromosome 10 that earlier studies suggested contained an area that influenced age of onset for both AD and Parkinson’s disease (Li et al., 2003b). This chromosomal region contains a large number of genes, however, so the investigators established a novel approach called “genomic convergence,” which allowed them to reduce and prioritize the number of candidate genes in the chromosome 10 region. By combining three independent but complementary lines of evidence, the team found four genes that mapped to a region of chromosome 10 that appeared to be important to the age of onset. These genes were expressed differently in people with and without AD. Though these findings are potentially valuable, additional studies must be conducted to determine whether, in fact, these genes do affect age of onset.

Other scientists have focused on chromosomes 12 and 19. In one study, conducted collaboratively by scientists from NIA and Celera Diagnostics of Alameda, California, the researchers examined regions of chromosomes 12 and 19 from three separate sample sets that included people with and without AD (Li et al., 2004). The researchers found that SNPs in the glyceraldehyde 3 phosphate dehydrogenase (GAPD) gene family were significantly associated with AD risk in all three sample sets. They also found that some GAPD SNPs on chromosome 12 were associated with an age of onset of 75 years and older, whereas some GAPD SNPs on chromosome 19 were associated with an age of onset of less than 75 years. Individually, the GAPD SNPs made different contributions to AD risk in each of the sample sets, and the investigators speculate that variants in functionally similar genes may account for this heterogeneity of AD risk.

Finally, knowledge gained from studying Down syndrome is revealing much about chromosome 21 and its possible role in AD. Most people have two copies of chromosome 21, but people with DS have three. Therefore, they have an extra copy of the APP gene. Every individual with DS who survives into his or her third decade develops the signature brain pathology of AD, although the location and distribution of these features are much more variable than in traditional AD. Research conducted by National Cancer Institute (NCI)-funded scientists demonstrates that those who survive into adulthood with DS also have a substantially increased risk of death from a number of causes, including cancer (Hill et al., 2003).

In recent years, DS has become a “natural model” for studying how AD develops and progresses. For example, several teams of scientists funded by the NICHD have attempted to find the genes on chromosome 21 that contribute to the development of dementia in individuals with DS and to understand the implications of these changes for the level of gene expression on normal brain development and maturation. At least 10 genes on chromosome 21 are involved in brain structure or function, but investigators have implicated only three of these genes to date in DS. Researchers at the University of Arkansas for Medical Sciences, Little Rock; the University of Connecticut, Farmington; and the Instituto de Médica in Córdoba, Argentina, are studying various aspects of these alterations in gene expression and the pathways these genes influence. Findings from these studies will help them understand why dendrites develop abnormally in infants with DS and will help them learn more about amyloid plaque formation and neuronal decline and death in middle-aged people with DS.

The University of Arkansas researchers discovered that in adults with DS, neuronal loss is dramatic and the plaques that develop are consistent with those found in AD (Mrak and Griffin, 2004). The researchers were able to demonstrate that two chromosome 21-based gene products—APP and S100B—are involved in these neuronal changes. S100B, which is over-expressed throughout the lifetime of people with DS, has been implicated in beta-amyloid plaque formation and correlates with the AD pathology that people with DS begin to experience as young adults. They also found that a protein called IL-1 (an inflammatory protein), which is not found on chromosome 21, is also over-expressed throughout life in people with DS. IL-1 increases expression of APP and S100B and drives numerous processes that contribute to the development of late-onset and DS-related AD.

Another genetics area of great interest to investigators is APOE-e4, the risk factor gene found on chromosome 19. A number of recent studies have shed additional light on the role of this APOE allele in AD. For example, studies using PET scans have found that people with AD have abnor-mally low rates of glucose metabolism in certain areas of the cerebral cortex (the outer layer of neurons in the brain that controls conscious thought, mental activity, and voluntary movement, and that processes sensory information from the outside world). Building on earlier PET studies that showed that cognitively healthy older carriers of the APOE-e4 allele had abnormally low rates of glucose metabolism in those same brain regions, investigators at the Banner Good Samaritan Medical Center in Phoenix, Arizona, examined whether this was true for relatively young adults as well (Reiman et al., 2004). The investigators performed PET and MRI scans and conducted neuropsychological tests on 24 healthy participants, 12 of whom were APOE-e4 carriers and 12 were not. The two groups did not differ significantly in gender, age, educational level, neuropsychological test scores, or other characteristics. Results of this study were consistent with previous studies in that the researchers found that APOE-e4 carriers had abnormally low rates of glucose metabolism in the selected brain regions. This study is important because it documents the earliest brain changes yet seen in living persons at risk of AD. These results also provide evidence that AD-like changes in the brains of APOE-e4 carriers can occur in cognitively healthy young adults. Tracking brain and cognitive changes over time will be necessary to determine how this pattern of AD-like brain changes relates to the likelihood that APOE-e4 carriers will develop AD at a later age. If these functional brain changes are eventually validated as an early predictor of AD, very early intervention and treatment will become a possibility.

Another study from the same group compared memory decline and new learning in APOE-e4 carriers and noncarriers aged 48 to 77 over a 2-year period. These investigators and colleagues from several other research sites in Arizona found that APOE-e4 carriers aged 60 and older showed a decline in new learning as compared with noncarriers (Baxter et al., 2003). No difference in cognitive performance was seen in those younger than 60 years old. These findings suggest that repeat testing of new learning over time may be a sensitive measure for detecting early cognitive changes in older people who are at increased risk of AD.

In a follow-up study with a larger group of participants and a longer study period, the Arizona researchers examined whether memory loss could be detected in carriers of APOE-e4 before the onset of amnestic MCI (Caselli et al., 2004). The cognitive abilities of a group of 212 cognitively healthy middle-aged to older adults were evaluated over approximately 33 months. In this study, the investigators did find that APOE-e4 carriers older than age 50 showed a modest decline in a number of memory skills compared to noncarriers of the same age, but they detected no differences in language, spatial skills, or executive functioning. As with the other studies, the authors emphasize that more work is needed before the relationship, if any, between memory decline and MCI or AD can be determined.

Other investigators have looked at the occurrence of APOE-e4 across a number of neurodegenerative diseases. A research team from the Mayo Clinic in Rochester, Minnesota, and Jacksonville, Florida, and the London School of Hygiene, assessed whether the presence of the APOE-e4 allele influenced the frequency of AD-like pathology in dementia with Lewy bodies, frontotemporal degeneration, progressive supranuclear palsy, corticobasal degeneration, and multiple system atrophy (Josephs et al., 2004). They found that this allele strongly influenced the occurrence of AD-like damage in these diseases, particularly in dementia with Lewy bodies and multiple system atrophy. Further study is needed to find out why APOE-e4 appears to have a larger effect in these diseases than in others.

A group of investigators from the Cache County Study has reported on recent genetic findings about whether AD is an inevitable consequence of aging (Khachaturian et al., 2004). The Cache County Study, a long-term study of 5,000 people in Cache County, Utah, provides a unique perspective on AD because the participating population is one of the longest lived in the United States. It includes about 700 individuals older than 85 and almost 250 who are 90 years old and older. This aspect of the study allowed the investigators to explore one AD puzzle: We know that the risk of AD increases with age, but we don’t know much about the proportion of people who would be likely to get AD over an extended lifetime, say 100 years, or about the effect of APOE-e4 over that time period. Using several types of analytic models, these investigators set out to estimate the risk of developing AD as a function of age and number of APOE-e4 alleles. The models estimated that 28 percent of individuals would not develop AD over any reasonable life expectancy. They confirmed that AD onset is accelerated in individuals with one, and especially two, APOE-e4 alleles, but did not see any meaningful difference in lifetime risk of developing AD related to number of APOE-e4 alleles. The authors concluded that this population contains individuals who are not susceptible to developing AD at an advanced age, regardless of which APOE allele they may carry. Discovering the genetic or environmental factors that are responsible for this resilience is clearly a high priority for future research.

Finally, scientists have used a genetics approach to examine late-life depression, a condition that is often associated with cognitive impairment and is a risk factor for the development of dementia and AD. To determine whether the presence of APOE-e4 could explain the linkage of this form of depression with dementia, University of Pittsburgh investigators funded by NIMH and NIA analyzed how frequently the different forms of APOE occurred in patients with depression compared to healthy older adults and people with AD (Butters et al., 2003). As expected, APOE-e4 occurred more frequently in individuals with AD. However, the frequency of APOE-e4 was not higher in those with depression compared to the healthy study participants. APOE-e4 frequency also did not differ among depressed participants with and without accompanying cognitive impairment. Unexpectedly, the onset of depression occurred at a younger age for APOE-e4 carriers compared to noncarriers. These results indicate that APOE-e4 is associated with the age of onset of late-life depression, but not cognitive functioning in that condition. Further studies will be required to elucidate the avenues through which late-life depression may contribute to the onset of AD.

Lifestyle and Dietary Patterns
The possible impact of environmental and lifestyle factors, such as intellectually stimulating activities, physical activity, and diet, on AD risk is becoming an increased focus of research. A number of studies over the past few years have provided intriguing hints that these factors may be linked to a reduced risk of AD, and they are consistent with what we know about other benefits associated with health-promoting behaviors throughout life. It is important to note, however, that these factors have been identified in observational and animal studies, and at present, they are only associated with changes in AD risk. Only further research, including clinical trials, will reveal whether, in fact, these factors can help prevent AD.

Recent studies have looked at these environmental and lifestyle issues from various aspects. For example, researchers from the Rush ADC found reduced AD risk among participants in a Chicago Health and Aging Project study who consumed fish frequently and whose diet was high in unsaturated, unhydrogenated fats (Morris et al., 2003a; Morris et al., 2003b). In this CHAP study, complete dietary intake and disease diagnosis data were available on 815 people, and the length of follow-up between dietary assessment and clinical evaluation was about 2.3 years. CHAP investigators found that participants who consumed one or more fish meals per week had a 60 percent reduced risk of AD compared to participants who seldom or never ate fish. People whose diets were higher in polyunsaturated and vegetable fats also had a reduced AD risk compared to those whose fat intake was predominantly saturated fats. Though these studies support other research showing health benefits related to fish and unsaturated fat consumption and provide some intriguing hints about AD, the authors caution that further studies are needed before dietary recommendations can be made based on a relationship to AD risk.

In another study, NIA-funded researchers used data from the Honolulu-Asia Aging Study to find out whether certain dietary antioxidant intakes could reduce oxidative stress and thereby reduce dementia risk (Laurin et al., 2004). This study of approximately 3,700 Japanese-Americans focused on possible relationships between vascular factors—such as blood pressure, blood cholesterol, and inflammation — and the later development of dementias such as AD. The team found no association between risk of dementia and mid-life intakes of antioxidants such as vitamins E and C and beta-carotene.

A considerable amount of research also has been devoted to examining the role of educational attainment on AD risk. One recent study, conducted by investigators at Harvard Medical School, used data from the long-term Nurses’ Health Study of 16,596 older female registered nurses to assess the relationship of educational attainment, husband’s education, household income, and childhood socioeconomic status to cognitive function and decline (Lee et al., 2003). The study participants had an initial cognitive assessment in the late 1990s and a second assessment about 2 years later. The investigators found that women with a graduate degree had substantially decreased odds of a low initial cognitive score and of cognitive decline over the 2 years compared to women with less education. The other measures of socioeconomic status considered in this study had little, if any, relation to cognitive function or decline in later life.

In the past few years, several research groups have attempted to link participation in leisure activities with a lower risk of dementia. However, the exact relationship remains unclear. Scientists do not know whether increased participation in leisure activities lowers the risk of dementia or whether people who later develop dementia participate less in leisure activities because they are already in the early phase of dementia, before symptoms are evident. Investigators from the Albert Einstein College of Medicine in New York City explored this issue in a group of 469 cognitively healthy people who were older than age 75 and still living in the community (Verghese et al., 2003). Over a period of about 5 years, 124 people developed dementia. Based on a statistical analysis, the investigators concluded that participation in leisure activities was associated with a reduced risk of dementia, even after adjusting for participants’ initial cognitive status and after excluding participants with possible preclinical dementia. Controlled clinical trials are needed to explore this issue further and assess the protective effect of leisure activities on the risk of dementia.

In addition to these studies, which suggest an association between particular lifestyle factors and actual AD risk reduction, other research provides indirect indications that lifestyle factors may be related in some way to AD risk. For example, a recent collaborative study by investigators at the University of Toronto and the University of California at Irvine assessed whether long-term treatment with a combination of “behavioral enrichment” (extra attention and lots of training and stimulation) and a diet rich in antioxidants, including vitamins E and C, and fruit and vegetable extracts could reduce age-related cognitive decline in dogs (Milgram et al., 2004). Dogs are a good model for studying AD because they can perform sophisticated and complex cognitive behaviors, their brains accumulate beta-amyloid plaques with age, and the degree of beta-amyloid deposition is related to the severity of cognitive decline. This study included both old and young dogs. Some received the fortified food and the enriched environment, some received one or the other enrichment, and some received neither. At the end of a year, the researchers tested the dogs on two learning tasks. Not surprisingly, the researchers found that the old dogs performed less well than theyoung dogs. However, the performance of the old dogs was improved by the fortified food and behavioral enrichment. The effects of the treatments were most evident in thedogs who received both interventions.

Accumulating evidence also suggests that being physically active may benefit more than just our hearts and waistlines. Research in animals has shown that aspects of both brain function and cognitive function improve with physical exercise. Several studies in aging adults have shown similar results. One study, conducted by researchers at the University of Illinois at Urbana-Champaign, used a form of MRI to measure changes in brain activity in healthy adults aged 58 to 78 before and after a 6-month program of brisk walking (Colcombe et al., 2004). The researchers found that the function of neurons in key parts of the brain increased along with improvements in the participants’ cardiovascular fitness. Compared to a physically inactive group, the walkers were able to pay attention better and focus more clearly on goals while disregarding unimportant information. Scientists working in this area speculate that physical activity may be beneficial because it may improve blood flow to the brain. Another possibility under investigation is that physical activity triggers cellular mechanisms that protect the brain from damage and promote its repair.

Vascular Disease and AD
In recent years, a number of studies have suggested that vascular diseases—heart disease and stroke—may contribute to the development of AD, the severity of AD dementia, or the development of multi-infarct dementia (a type of dementia that results from multiple strokes). If this is true, it may have great scientific importance for two reasons. First, it may help investigators understand the origins of AD by encouraging them to focus attention on particular aspects of damage to the brain, such as microscopic strokes. Second, new avenues for preventive or delaying strategies may be possible if it becomes clear that modifiable risk factors for vascular disease (such as high cholesterol and high blood pressure) also are risk factors of AD or of degree of dementia.

Recently, five research teams working in separate large-scale epidemiologic studies explored whether vascular disease affects the likelihood of developing AD:

• Using data from the Washington Heights-Inwood Columbia Aging Project, a study of aging and dementia in New York City that includes 1,766 Medicare recipients, investigators found that the risk of AD was 60 percent higher in individuals with a history of stroke compared with individuals without such a history (Honig et al., 2003). The risk of AD was particularly great if hypertension, type II diabetes, or heart disease also was present.
• Boston University School of Medicine researchers working on the long-standing Framingham Study studied a group that had not had a stroke and did not have dementia (Ivan et al., 2004). The researchers found that those participants who went on to have a nonfatal stroke were twice as likely to develop dementia as those who did not experience a stroke. This finding underscores the importance of stroke prevention in reducing the possible risk of dementia.
• A group of scientists analyzed data from the University of Washington/Group Health Cooperative Alzheimer’s Disease Patient Registry and found that people with the same level of cognitive decline in AD who also had vascular damage had less severe AD pathology in brain tissue than those with AD alone (Riekse et al., 2004). These findings suggest that cerebrovascular disease may act to increase the severity of the cognitive impairment in people with AD.
• An analysis of data from the Honolulu-Asia Aging Study showed that high blood pressure during mid-life increased the risk of dementia later in life. Study participants who had high diastolic blood pressure, but who were never treated for hypertension, had an increased risk of atrophy in the hippocampal region of the brain (Korf et al., 2004). The authors suggest that early treatment of hypertension may reduce this risk.
• Finally, researchers at the Rush ADC and Rush Institute for Healthy Aging in Chicago, conducted a long-term clinical and pathology study as part of the ongoing Religious Orders Study (Schneider et al., 2004). These scientists found that brain infarction (a stroke occurring in blood vessels in the brain) and AD pathology each contributed to an increased likelihood of dementia. In particular, the odds of developing dementia increased nearly three-fold if evidence of one or more brain infarctions was present. Brain infarctions and AD pathology contributed jointly to a greater risk of developing dementia than would be indicated by either condition alone.

Since 1988, the National Heart, Lung, and Blood Institute has funded the Cardiovascular Health Study (CHS), a long-term study of risk factors for the development and progression of coronary heart disease (CHD) and stroke in elderly adults. This study has provided valuable data about the relationships between cardiovascular risk factors and AD, and it includes cognitive decline and dementia related to vascular disease as a key element of its design. One of the distinguishing features of the study has been add-on components funded by other NIH Institutes, including NIA. These add-ons have provided a cost-effective opportunity for investigators to use existing study populations to explore issues that were not part of the original study design. In the past 2 years, CHS study investigators from the University of Washington and the University of Pittsburgh have made several advances important to AD research. These studies have laid the groundwork for future examinations of the interactions between cerebrovascular risk factors, AD, MCI, and dementia:

• The researchers have developed a methodology for conducting long-term dementia evaluations (Lopez et al., 2003a). In 1991, they conducted MRIs on approximately 3,600 CHS participants. This large and diverse study population is a particular strength of the CHS. This same group of participants received cognitive function evaluations every year thereafter, and, from 1998 to 2000, they also were evaluated for dementia through detailed neurological and neuropsychological exams. Not only was this the largest-ever study of dementia to include cognitive testing, MRI, and examination of genetic markers, but it provides a methodology that other researchers can use to study dementia in large groups.
• The investigators assessed the incidence of dementia, vascular dementia, and AD across the CHS population who had received the MRIs (Fitzpatrick et al., 2004). There was no indication of sex or racial differences, but strong associations with education and APOE-e4 status were apparent. Compared with results from other studies, the incidence of vascular dementia was higher than expected.
• Investigators found that particular measures of cognitive function, APOE-e4 status, and brain tissue characteristics as seen on MRI were strong predictors of both dementia and AD in the CHS population (Kuller et al., 2003).
• The research team found that 22 per-cent of the participants aged 75 or older had amnestic MCI. Most of these people had other health problems as well, which affected their cognitive function (Lopez et al., 2003b). The research team also discovered that the people who developed this form of MCI were more likely to suffer from depression, be African-American, have a relatively low educational level, carry the APOE-e4 allele, or have cerebrovascular disease (Lopez et al., 2003c). Presence of the APOE-e4 allele was linked only to amnestic MCI. Though these findings are intriguing, additional research is necessary to determine exactly how these factors are associated with MCI risk.

In collaboration with NIA researchers, a group of scientists at the University of California at San Francisco took a somewhat different look at vascular risk factors and the risk of cognitive decline (Yaffe et al., 2004a). These researchers wanted to determine whether an earlier stage of the cardiovascular disease process—metabolic syndrome—increased the risk of cognitive decline. Metabolic syndrome, a constellation of factors that increases heart disease risk, includes abdominal obesity, high triglyceride levels, low HDL (“good cholesterol”) levels, high blood pressure, and insulin resistance (an impaired ability to use insulin). The researchers also wanted to see whether vascular inflammation—another heart disease risk factor—modified possible associations between metabolic syndrome and cognitive decline. The investigators worked with participants of the Health, Aging, and Body Composition study, a 1997 to 2002 study of 3,075 older adults living in Memphis and Pittsburgh. The researchers found that those participants with metabolic syndrome had an increased risk of cognitive impairment and decline, even adjusting for other demographic and lifestyle factors. The increased risk of cognitive impairment was seen primarily in participants who had high levels of inflammation markers in their blood. This suggests that at least some of the increased risk associated with metabolic syndrome is caused by inflammation. Though these results are consistent with the findings of other studies, further research is clearly needed to determine whether reducing risk factors for cardiovascular disease or lowering inflammation can reduce the risk of cognitive impairment in older adults.

Another aspect of heart disease risk that has been of continuing interest to both heart disease and AD researchers is homocysteine. Previous epidemiologic studies have shown that elevated levels of this amino acid, which is a risk factor for heart disease, are associated with an increased risk of developing AD. Basic science studies have shown that high homocysteine levels make some neurons vulnerable to dysfunction and death and that an enzyme called betaine-homocysteine methyltransferase (BHMT) may play a major role in regulating homocysteine levels in the blood. However, little is known about the properties of BHMT and the extent to which genetic variations in the enzyme might contribute to abnormally high levels of homocysteine. University of Illinois scientists funded by NIDDK have been trying to fill these knowledge gaps and recently found that BHMT functions only when it is in a certain structural form (Szegedi and Garrow, 2004). Additional studies will help to clarify the relationship between BHMT structure and function.

Back to Top

Diabetes and AD
The possible association of diabetes, insulin, and AD also is garnering increasing attention. Type II diabetes mellitus is a serious public health problem in the U.S. The condition affects about one in five people over age 65 and has been associated with a variety of adverse health effects. Evidence from a number of epidemiologic studies suggests a possible link between diabetes and cognitive impairment, and this has spurred researchers to examine this link on many levels, from test tube investigations to population studies. The idea behind this area of research is to determine whether or not diabetes is a risk factor for cognitive decline, and if so, whether therapies for diabetes may help lower risk of cognitive decline or AD.

One way that researchers are looking at the association between diabetes, cognitive problems, and AD is through epidemiologic studies. Recent long-term studies have linked diabetes to cognitive function in four distinct populations: Catholic nuns, priests, and brothers older than 55; female registered nurses aged 70 to 81; postmenopausal women with osteoporosis; and whites aged 42 to 89 years old and living in the community. In the first of these studies, conducted by investigators at the Rush ADC, participants of the Religious Orders Study underwent annual assessments of five cognitive “systems” that are commonly affected by aging, dementia, and AD (Arvanitakis et al., 2004). Over the 5.5-year study period, 151 people were diagnosed with AD, including 31 who had diabetes. An analysis of the study data revealed that the risk of developing AD was 65 percent higher in people with diabetes than in those who did not have diabetes. The study also showed that diabetes appears to affect different cognitive systems differently, with the area of perceptual speed (the speed with which simple comparisons can be made) being most affected.

In the second study, investigators from the Harvard School of Public Health, working with participants in the Nurses’ Health Study, found that women with diabetes performed worse on a range of cognitive tests than did women without diabetes (Logroscino et al., 2004). The odds of cognitive difficulties were particularly high for those women who had had diabetes for a long time. Interestingly, women who were taking oral diabetes medications performed as well on the cognitive performance tests as did women without diabetes.

Researchers from the University of California at San Francisco investigated whether diabetes and impaired fasting glucose (IFG, an indicator of increased diabetes risk) were linked to cognitive function (Yaffe et al., 2004b). The data for the study came from the Multiple Outcomes of Raloxifene Evaluation (MORE), a long-term osteoporosis study of more than 7,000 women between the ages of 31 and 80 (see p. 61 for information on MORE). The investigators found that the cognitive abilities of women with IFG were lower than those of women with normal blood glucose levels but higher than those with diabetes. They also found a higher risk of developing cognitive impairment among the women with IFG or diabetes than among the women with normal blood glucose levels.

Another University of California at San Francisco team of researchers examined the change in cognitive performance over a 4-year period among older adults participating in a study called the Rancho Bernardo Study (Kanaya et al., 2004). Participants, who were divided into three groups—normal glucose tolerance, impaired glucose tolerance, and diabetic—were given three different cognitive tests at the beginning of the study and again 4 years later. The researchers found that scores on all three cognitive function tests did not differ across the groups at the beginning of the study, but that the women with diabetes had a more rapid decline in performance on the verbal fluency test over the study period compared with women in either of the other two groups.

These findings are complemented by additional epidemiologic research conducted by researchers from the Mt. Sinai School of Medicine and Israel’s Tel Aviv University and Sheba Medical Center (Schnaider Beeri et al., 2004). These investigators, working with more than 10,000 male participants in the Israeli Ischemic Heart Disease study, also found evidence of diabetes as a risk factor for dementia. In contrast to other studies, however, these researchers linked diabetes in midlife to dementia that emerged more than 35 years later in very old members of the study group.

An analysis of data from the Honolulu-Asia Aging Study provides intriguing evidence about the possible associations between insulin levels and dementia risk (Peila et al., 2004). These data showed that increasing levels of insulin in the blood were associated with an increased risk of dementia. Too-high insulin levels can be controlled through dietary changes and increased physical activity, so this may be a modifiable risk factor if future research confirms these findings.

Other researchers, such as one team from New York University School of Medicine, are studying diabetes and cognition through studies with animals and small numbers of individuals (Convit et al., 2003). We know that age is a risk factor for both impaired cognition and impaired glucose metabolism, and a relationship between these deficits is suspected. It is not clear, however, how poor glucose metabolism may exert its effect on memory function. One possibility is through the hippocampus, the brain structure that is important for learning and memory and is one of the earliest regions damaged by AD. Rodent studies have shown that during performance of a memory task, the hippocampus is activated, and hippocampal glucose levels drop in specific locations depending on the difficulty of the task. In older rodents, the drop in hippocampal glucose levels is more profound and lasts longer, perhaps forming a basis for impairments in memory performance seen in some older animals. Because of this vulnerability, subtle metabolic insults may lead to damage and volume loss. Using these findings as a starting point, the New York University investigators conducted memory tests with 30 nondiabetic middle-aged and older individuals who did not have dementia. Simultaneously, they administered a glucose tolerance test to measure the participants’ ability to regulate glucose. The size of the hippocampus and other brain regions was measured with MRI. A decreased ability to regulate glucose was associated with decreased general cognitive performance, memory impairments, and atrophy of the hippocampus. This association was not related to the participants’ ages. The investigators also found no association between the volume of other brain regions and glucose regulation. They concluded that a decreased ability to regulate glucose is associated with modest impairments of memory and that the impairments may be the result of a direct impact of poor glucose metabolism on brain structures important for memory. Based on these findings, the researchers suggest that memory deficits among older people with poorer glucose tolerance may be caused by an inability to compensate for the drops in hippocampal glucose levels that normally occur with activation of brain circuits during performance of a memory task. Perhaps better lifetime management of blood sugar may help maintain memory function in old age and perhaps even reduce the risk of hippocampal damage.

Back to Top

<< Back | Next >>