15 July 2011. Can common, minor ailments of aging increase the risk of Alzheimer’s disease and hasten cognitive decline? Though this idea might seem counterintuitive, three papers, from separate groups, in the July 19 Neurology suggest that various markers of general health and aging may indeed be linked to dementia. In one study, researchers led by Raj Shah at Rush University, Chicago, Illinois, found that both high and low hemoglobin levels in older people were tied to greater AD risk and faster rates of cognitive decline. Clive Holmes and colleagues at the University of Southampton, U.K., focused on the role of proinflammatory cytokines, which tend to increase with age. The researchers found that people with AD who had higher levels of these cytokines also displayed more neuropsychiatric symptoms. This complemented the group’s previous finding that high levels of the same cytokines predicted faster cognitive decline. Meanwhile, Kenneth Rockwood’s group at QEII Health Sciences Centre, Halifax, Canada, took the broadest approach, combining a slew of aging-related health factors not normally associated with AD into one variable that predicted AD risk better than traditional dementia risk factors did.
In an editorial accompanying the three papers, Jean Francois Dartigues and Catherine Féart at the University Victor Ségalen of Bordeaux, France, hail the focus on novel AD risk factors “that involve new pathways and suggest innovative treatment strategies.” Nonetheless, they caution that these were exploratory studies conducted in observational cohorts not originally designed to measure these factors, and note that the data need replication.
Scientists have traditionally drawn a distinction between what happens in the brain and what happens in the periphery, but Holmes told ARF that growing evidence suggests that “these two systems are intimately related and they communicate with one another…. We should not ignore [peripheral events].” Or, as Rockwood put it, “Anything that is bad for you in general will end up being bad for your brain. It is all hooked up.”
Shah and colleagues were interested in hemoglobin because previous studies had shown an association between anemia and low cognitive scores (see Denny et al., 2006; Chaves et al., 2006; and Shah et al., 2009). Data were inconclusive on whether anemia was a risk factor for AD, however, with one prospective study finding that it was (see Atti et al., 2006), while another found it was not (see Beard et al., 1997).
To clarify the issue, Shah and colleagues analyzed data from almost 900 elderly participants in the Rush Memory and Aging Study. In contrast to most previous studies, the researchers looked at the whole range of hemoglobin levels. They found that both low and high levels were associated with a higher risk of developing AD, as well as a faster rate of cognitive decline. Participants with anemia, which is common in the elderly, had about 60 percent increased risk of developing AD over roughly three years, and had a rate of cognitive decline comparable to being 12 years older than their actual age. The effects of high hemoglobin appeared to be more dramatic than low levels, giving a greater than threefold increased risk of AD and a cognitive decline equivalent to someone 18 years older than the participants’ actual age. However, the high hemoglobin results are preliminary, Shah told ARF, because comparatively few seniors have high levels—about 10 people in this study.
The next steps are to replicate these findings and search for the mechanisms behind the effect, Shah said. He is currently examining brains donated after death by participants in the Rush Study, looking for correlations between hemoglobin levels and pathological changes. More research will be needed to determine if correcting hemoglobin abnormalities lowers the risk of AD, Shah said. It may be that hemoglobin itself does not drive the development of dementia, Shah suggested, but is instead a marker for another disease process. For example, low hemoglobin is associated with vascular disease, kidney disease, and dysregulation of iron that can lead to oxidative stress, all of which could place stress on the brain. High hemoglobin can indicate lung problems or overproduction of red blood cells, both of which can involve hypoxia. Even if hemoglobin does not directly contribute to disease processes, it may be useful as an AD risk marker, Shah noted, because primary care doctors commonly test hemoglobin levels.
Systemic inflammation is another common consequence of aging. In previous work, Holmes and colleagues showed that high levels of the proinflammatory cytokines tumor necrosis factor-α (TNFα) and interleukin 6 (IL-6) were linked to a twofold increase in rates of cognitive decline over six months in a prospective cohort of about 300 people with AD (see Holmes et al., 2009). To follow up on this study, the authors examined the same cohort for the development of neuropsychiatric symptoms such as depression, apathy, and anxiety, which are common in AD. They found about a twofold increase in these symptoms in people with higher levels of circulating TNFα and IL-6. This increase occurred independently from the development of delirium.
How might inflammation affect the brain? As amyloid plaques accumulate with age, microglia may react to these deposits and become primed, Holmes told ARF. This leaves the brain vulnerable to the effects of a secondary infection that then triggers the microglia to begin damaging the brain. “It is a two-hit hypothesis,” Holmes noted. This model has been proposed by others in the field as well (see, e.g., ARF related news story). Holmes said they are currently testing a TNFα-blocking drug, Etanercept, in a small pilot study to see if it can slow AD progression. Etanercept is typically used to treat rheumatoid arthritis, but has generated interest in the AD field before when a controversial case study found promising results (see ARF related news story; also earlier findings: see ARF news story). Holmes is also applying for funding to conduct a prospective study to see if late-life infections and trauma can trigger AD in pre-symptomatic people. Holmes pointed out that many known risk factors for AD, such as head injury, diabetes, obesity, and surgeries, all involve inflammation, suggesting it could be a common underlying mechanism behind cognitive decline.
In contrast, Rockwood and colleagues concentrated on factors with no known relationship to AD. “In our aging research program we have been very struck by how small risks accumulate to turn into big risks,” Rockwood told ARF. Using data from several thousand elderly participants in the Canadian Study of Health and Aging, first author Xiaowei Song combined 19 general health complaints, none of which is traditionally associated with AD, into a “frailty index.” The health issues included such things as poor eyesight, foot problems, nasal congestion, and wearing dentures. The authors found that a higher frailty index at baseline was linked to a greater risk of developing AD five and 10 years later, with AD risk increasing by about 3 percent for each deficit present in a participant. The association held even after adjusting for age, sex, education, and other traditional risk factors, and the authors note that it outperformed the predictive value of traditional risk factors such as diabetes and heart disease in this study.
In ongoing work, Rockwood said he is combining traditional and nontraditional risk factors into one variable to see how well that performs in predicting AD. One implication of the finding, Rockwood suggested, is that researchers may be able to gain statistical power in their studies by rolling all confounding factors and effect modifiers into one big variable, rather than having to adjust separately for each one. “That will be a much more tractable model to use,” Rockwood noted. He is also interested in looking at how these general health factors might affect brain function. His group has analyzed data from the Alzheimer’s Disease Neuroimaging Initiative to look for structural brain changes that correlate with changes in the frailty index, and they will present their findings at the 2011 International Conference on Alzheimer’s Disease, Rockwood said. Although it is not yet clear how poor general health might relate to AD, Rockwood favors the idea that these events cause harm by placing stress on brains with impaired plasticity or faulty repair mechanisms. “Where some people can withstand these low-grade insults, other people cannot, because they have a brain that, as it starts to repair itself, does more harm than good,” he speculated. Rockwood stressed that he is not suggesting that any trivial health problem can cause AD, but rather is promoting the idea that “the key problem in AD may be an aberrant [brain] repair mechanism.”
In their editorial, Dartigues and Féart noted that the health measures examined in these three papers might not have a direct effect on AD processes, but instead might be related to the disease through various confounding variables, such as physical activity, nutrition, or other health behaviors. More study will be needed to clarify this. Nonetheless, the commentators find it intriguing that many health issues of aging seem to increase AD risk. “The results reported in the present issue suggest a new vision of preventive or curative treatments which, instead of targeting specific etiologic mechanisms, would instead aim at improving general health,” they wrote.—Madolyn Bowman Rogers.
Shah RC, Buchman AS, Wilson RS, Leurgans SE, Bennett DA. Hemoglobin level in older persons and incident Alzheimer disease. Neurology. 2011 Jul 19;77(3):219-26. Abstract
Holmes C, Cunningham C, Zotova E, Culliford D, Perry VH. Proinflammatory cytokines, sickness behavior, and Alzheimer disease. Neurology. 2011 Jul 19;77(3):212-18. Abstract
Song X, Mitnitski A, Rockwood K. Nontraditional risk factors combine to predict Alzheimer disease and dementia. Neurology. 2011 Jul 19;77(3):227-34. Abstract
Dartigues JF, Féart C. Risk factors for Alzheimer disease: Aging beyond age? Neurology. 2011 Jul 19;77(3):206-7. Abstract