Using data from a longitudinal study of nearly 2000 Finnish adults, researchers have devised a simple scoring scheme to judge if the middle aged are at risk of developing dementia later in life. The study, from Miaa Kivipelto at the Karolinska Institute in Stockholm, Sweden, and colleagues there and at the University of Kuopio and University of Helsinki, both in Finland, shows that the presence of modifiable vascular risk factors in middle age, including high blood pressure, cholesterol and obesity, predict the occurrence of dementia (due to Alzheimer disease in most cases) 20 years later. A score based on these and other risk factors (age, education, physical activity, and ApoE status) was a good predictor of dementia by age 70, with higher scores indicating higher risk. With further improvement and validation, the novel dementia risk score, analogous to assessments now used for cardiovascular disease or diabetes, could help patients by allowing doctors to target AD education and prevention to the people most at risk. The work appeared online August 2 in Lancet Neurology.
In a slightly different vein, another longitudinal study, this one from Clifford Jack and colleagues at the Mayo Clinic in Rochester, Minnesota, shows that changes in metabolite levels in the brain, detected by proton magnetic resonance spectroscopy (MRS) correlate with cognitive decline. Their work, currently in press, but available since July 24 in Neurobiology of Aging online, suggests another potential tool to track the progress of AD and gauge the effectiveness of treatments.
Knowing as early as possible who has AD is the current focus of a tremendous research effort, much of it involving brain imaging. MRI-based volumetric measurements show changes in brain volume, for example, and FDG PET scans reveal decreases in glucose uptake that track with progression of clinical dementia. But to try to get a handle on AD before it starts, the Scandinavian researchers investigated mid-life risk factors among participants in the Cardiovascular Risk Factors, Aging and Dementia (CAIDE) study. Between 1982 and 1987, thousands of Finnish residents were interviewed and examined for a variety of cardiovascular risk factors as part of the study. Kivipelto and colleagues followed up on 1,409 participants (original mean age 50). When the same participants were re-examined 20 years later (mean age 71.3), 61 were clinically diagnosed as demented, with most of those (48) having AD (Kivipelto et al., 2005).
Comparisons of the demented and nondemented subjects showed that future dementia was predicted by middle age measures of education and cardiovascular risk factors including hypertension, hypercholesterolemia and obesity. As expected, ApoE status also correlated strongly with dementia, as did age. Regression analysis was used to assign a whole number to each of these risk factors based on its relative contribution to dementia risk. In addition, sex, physical activity level, and ApoE status were also scored. (Scores denoted the presence or absence of a risk factor, or in the cases of age and education, one of three possible levels). Adding the individual risk scores gave an overall dementia risk score: higher tallies were associated with higher risk, which varied from 1 percent or less at scores in the lowest quintile, to approximately 20 percent in the highest quintile.
Of course, the formula could be improved by adding in information on more risk factors which were not covered in the Finnish study, including family history of dementia, serum triglycerides, HDL and LDL concentrations, waist-hip ratio, presence of diabetes or glucose intolerance, and C-reactive protein, the authors note. In addition, the scheme will need to be validated in other populations, and at later follow-up ages, as most dementia develops after age 70.
While the individual predictive value of the dementia rating score was considered good (77 percent sensitivity and 63 percent specificity), its main use will be in targeting prevention and education efforts, the authors say. “The risk score provides a quantitative estimation of the probability of becoming demented, but it cannot definitely state whether a person will develop dementia. Therefore, the score should be mainly used to target the preventive measures to those most at risk, and it should not be used to label individuals as being demented or nondemented in the future,” they write.
The bright side is that many of the contributing risk factors are modifiable by medication and/or lifestyle changes. Indeed, the results highlight that many of the same primary preventions (reducing blood pressure and cholesterol levels and controlling body weight, for example) should reduce the risk of both cardiovascular disease and Alzheimer disease. The results of trials to assess whether the effects of mitigating cardiovascular risk factors like hypertension can reduce dementia in reality have shown some positive indications, but the results have not been conclusive.
In the imaging study, Mayo researcher and first author Kejal Kantarci measured changes in metabolite levels in brain over the course of several years in normal elderly or patients with mild cognitive impairment or AD. The ratio of N-acetylaspartate (NAA)/creatine, a marker for neuronal integrity, declined more in the posterior cingulate gyrus (a region affected early in AD) in people with MCI and AD, compared to controls. Importantly, the change over time correlated with measurements of dementia over the course of the study, and correlated with clinical progression similar to changes in ventricular volume.
The choline/creatine ratio has been reported to be elevated with aging and in AD, and the investigators found that changes in that marker predict changes in the Mini-mental State Exam scores in patients with AD and with MCI progressing to AD. Interestingly, in patients with stable MCI, the choline/creatine ratio declined, which the authors suggest could represent a compensatory mechanism in some people to upregulate acetyl choline synthesis and slow cognitive decline.
The results suggest that both the NAA/creatine and choline/creatine ratios could be useful as disease markers and surrogates for therapeutic effects of new treatments. Changes may indicate either enhanced neuronal survival (NAA and choline) or normalization of choline metabolism, measures of brain physiology that could be “equally useful” as volumetric measurements in monitoring drug effects in patients with AD, the authors conclude.—Pat McCaffrey