Sudha Seshadri, MD, Associate Professor of Neurology at the Boston University School of Medicine, discusses her recent paper, "Association of parental dementia with cognitive and brain MRI measures in middle-aged adults," which was recently published in Neurology® (2009;73:2071-2078). She spoke with Richard D. King, MD, PhD, member of the editorial board of the Science section of AAN.com.
AAN.com: Can you briefly summarize the methodology and major findings of the study?
Seshadri: In this paper, we present data on 717 middle-aged participants from the Framingham Offspring study. We compared performance on cognitive tests and volumetric brain MRI measures in two groups: participants with and without parental dementia. The diagnosis of dementia or normal cognitive status in the parents had been verified prospectively as part of a continuing follow-up of the Framingham original cohort. We observed that carriers of the ApoE ε4-allele who had a history of parental dementia and Alzheimer's disease (AD) had worse verbal and visuospatial memory and a higher rate of global brain atrophy, and that these features are detectable two decades before the average age at onset of dementia in the participants' parents. Poorer performance in memory tests is equivalent to the effect of approximately 15 years of brain aging.
To our knowledge, this is the first time that such cognitive and structural changes have been demonstrated in middle-aged offspring of parents with sporadic, late-onset (as opposed to autosomal-dominant, early-onset) dementia and AD. We believe that the public health implications of these observations are substantial. These findings reinforce prior data suggesting a prolonged subclinical phase before the onset of dementia, and support ongoing efforts to identify genetic and midlife environmental risk correlates of this subclinical phase and its progression to clinical dementia.
AAN.com: Given the large influence ApoE ε4 status had, do you think that routine screening should be done in offspring of people with dementia? Should the general population be screened?
Seshadri: At this time, I believe the risks of screening outweigh the benefits, since we are not aware of any effective, targeted preventive interventions. Measures such as physical activity, social engagement, maintaining a healthy weight, and controlling vascular risk factors should be encouraged in everyone—particularly in persons who have a parental history of AD and are APOE ε4 positive. The risks would be the anticipation of poorer memory performance that could lead to anxiety, depression, and stress. Although a recent paper in the New England Journal of Medicine by Robert C. Green, MD, MPH, and colleagues from the REVEAL study described no additional psychological distress from learning one's APOE status, this evidence appeared in the context of a study situation with good counseling and related only to the risk of future AD. Being aware of more proximate additional risks (such as poorer memory in one's sixties) could increase the potential for distress.
AAN.com: In sections of your paper, you distinguished "dementia" from "Alzheimer's." Given that everyone with AD by definition has "dementia," why did you decide to make this distinction? Was there insufficient evidence to make a more specific diagnosis, or did this pertain to age of onset?
Seshadri: The group of persons with AD comprises a subset of all persons with dementia. In our sample, AD was the most common type of dementia. Our general statistical approach is to first consider all types of dementia as one outcome. However, we recognize that analyses focused on "all dementia" are likely to be driven largely by effects associated with AD, which we verify with a separate analysis of AD only. If there were a dissociation of results between the "all dementias" versus "AD only" analyses, this would suggest that significant associations might be linked to other types of dementia, the most probable of which is vascular dementia, the second most commonly diagnosed dementia.
AAN.com: How has the use of neuroimaging changed since the study was started, and what effect has this had on the way you approach neurodegenerative disease?
Seshadri: The Framingham Heart Study began in 1948, the second generation was enrolled in 1971, and we have screened for incident dementia since 1975 (Professor Philip Wolf is the principal investigator of this study). Imaging came after the study began: CT scanning was introduced in the 1970s, and MRI in the late 1980s and early 1990s. More recently, we have been able to measure regional volumes and perform diffusion weighted imaging and diffusion tensor imaging (DTI) on brain MRI. We are now able to study subclinical markers (or endophenotypes) of Alzheimer-type and vascular pathology on volumetric brain MRI. Such imaging markers are being refined every day and we try to keep up: we have two sets of MRIs on all eligible and consenting second-generation participants and can apply new techniques to measuring and analyzing the MRI data we have stored.
Criteria for the diagnosis of Alzheimer's disease and vascular cognitive impairment dementia have also evolved over time. As accepted consensus criteria for these diseases emerged, we have adjusted our review process to apply these standard criteria to our subjects and to reassess previously reviewed cases based on current criteria.
AAN.com: Which neuropsychological tests do you think are most useful (sensitive and/or specific) to practicing neurologists in making a diagnosis for memory loss?
Seshadri: As a practicing neurologist, I find the portion of the Clinical Dementia Rating (CDR) in which both patient and informant are asked to recall specific details of an event that happened approximately a week earlier to be very helpful. I do not assign a score based on that test alone (although this information contributes to the CDR memory score), but it gives me a good sense of whether or not there is a problem. I also use a version of the logical memory paragraph and a word list, and sometimes a Rey-Osterrieth diagram for visual memory. I find these adequate in many patients, and of course I seek more comprehensive testing from a neuropsychologist colleague in others. To screen for overall cognitive status I find the clock drawing test and simultaneous letter/categorical fluency are also helpful. In general, people who can do serial subtraction of seven from 100 down to 72 tend to be the "worried well" rather than those with early dementia. And no test beats a good history!
AAN.com: What are the implications of your findings for clinical neurologists?
Seshadri: If our findings are replicated, the data suggests that persons who have a parent with AD and are ApoE ε4 positive need to be followed carefully, looking for evidence of cognitive decline. Studies of preventive interventions should target this population, and genetic studies should look for an interaction with APOE. People in this group should be encouraged to be physically active, perform challenging cognitive activities, stay socially engaged, maintain a healthy weight, and control vascular risk factors.
Within the past five years, Dr. King has received research grant support from the National Institute of Aging, the Alzheimer's Association, the UNCF*Merck foundation and the Robert Wood Johnson Foundation.