ApoE and the Risk of
Alzheimer's Disease

Joan W. was 70 years old and in good health, but she was worried. In recent months, she had begun to notice that she was occasionally unable to recall an important detail of the movie she had seen the night before, or couldn't find the word she needed in the middle of a sentence. Her older brother had suffered from Alzheimer's disease (AD), and - mercifully - had died of a heart attack after several difficult years of slow decline. Ms. W. had read in the newspaper that several genes for AD had been discovered, and thought she recalled hearing about a new test for the disease. Maybe the test would help her know if these little lapses were the beginning of something more serious. She decided to ask her doctor.

Her doctor began by telling her that the kinds of symptoms she was describing were normal, and did not sound like the harbingers of AD. On the other hand, he said, having a brother with the disease increased her risk somewhat, so she was right to be concerned. The doctor reported that he, too, had read about genes for AD, and had received a glossy pamphlet or two advertising a new genetic test for the disease. He pledged to look beyond the pamphlets and get back to her.

Four known genes for AD
The genetics of AD is quite complex. To date, four genes have been associated with the disease. Three of these (on chromosomes 21, 14, and 1) lead to the development of early-onset AD (onset prior to age 60 or 65), which is much less common than late-onset AD (onset after age 60 to 65).

One gene, the Apolipoprotein E (ApoE) gene on chromosome 19 has been associated primarily with late-onset AD, although more recent work suggests that its peak effect occurs earlier than originally thought. It has aroused intense interest because the great majority of the four million Americans diagnosed with AD developed the disease after age 65.

The role of ApoE
The ApoE gene makes a protein involved in carrying cholesterol in the bloodstream. It occurs in three forms, or alleles, designated "2," "3," and "4." The 4 allele is associated with AD. While the three early-onset AD genes are deterministic - everyone who carries an abnormal form develops AD - ApoE-4 appears instead to be a risk factor for AD.

Several years ago, it was first noted that the E-4 allele was over-represented in late-onset AD patients, both those with and those without a family history of the disease. Subsequent studies documented that ApoE-4 is also over-represented in early-onset AD, again including subjects with and without a family history. Recent work by our own group at Massachusetts General Hospital in collaboration with other sites suggests that ApoE4 is most strongly associated with AD that begins in the 60's (thus straddling early and late-onset studies), when the disease is relatively rare, and not in the 70's and 80's, when it is more common.

Most evidence suggests that ApoE-4 does not directly "cause" AD, but instead leads to the occurrence of the disease at an earlier age (although with a disease that occurs so late in the life span, the two are not entirely separable). The magnitude of ApoE's effect on age-at-onset is substantial. Our group looked at families in which at least two siblings were diagnosed with AD. We found that the average age of onset of individuals with two copies of ApoE-4 was 67, while that for individuals with one or no copies was 74.

Age-at-onset tends to be similar within families, but the similarity often breaks down when family members differ in their ApoE status. Within our group of families, if one affected sibling has two copies of ApoE-4, and the other has none, onset occurred at a younger age in the sibling with two copies 93 percent of the time, with an average difference of more than five years. Moreover, out of some 250 sibling pairs with AD, 10 pairs had a difference in age-at-onset of at least 15 years. Seven of the 10 pairs consist of an earlier-onset individual who has two copies of ApoE-4 and a later-onset individual with one copy. The ApoE status of the two siblings in the other three pairs is identical.

While the effect on age-at-onset is strong, several studies have shown that individuals with two copies of ApoE-4 can live well into their 80's without developing AD. For instance, among the brothers and sisters of the sibling pairs with AD in our study, there are 15 unaffected siblings who have no evidence of memory problems, are older than their affected siblings, and carry two copies of ApoE-4. Because these individuals all have two siblings with AD, we would consider them at elevated risk for the disease, even without knowing their ApoE status. Knowing that they carry two copies of ApoE-4 makes them appear at still higher risk, yet they have already survived beyond - often well beyond - the age at which their two siblings developed the disease.

Thus, ApoE appears primarily to affect when rather than whether one develops AD. Of course, when carries considerable personal and public health importance. First, if age-at-onset can be delayed sufficiently, the disease may be pushed right out of the life span. Second, although the disease is devastating at any age, its emotional, social, and economic toll tends to rise as the age-at-onset falls.

ApoE testing and the risk for AD
Since ApoE-4 is clearly a risk factor for AD, an ApoE test might appear to give people like Mrs. W. useful information about their risk for AD. However, two expert panels representing several organizations involved in AD advocacy, treatment, and research recommend against ApoE testing because its predictive value is poor: Many individuals with no copies of ApoE-4 develop the disease, and many individuals with two copies never do. Although it is clear that those with two copies (and perhaps one) of ApoE-4 are at greater risk of AD than those with none, the actual magnitude of the risk is unknown, and depends critically on longevity. In addition, assuming that the disease does develop, those with two copies of ApoE-4 will have, on average, an earlier age-at-onset than those with none or one, but it is impossible to predict at what age the first symptoms will appear.

ApoE and cardiovascular risk
Independent of concerns about AD, many people have already been tested for ApoE as part of a cardiovascular risk profile. Presumably related to ApoE's role in the transport of cholesterol, the ApoE-4 allele also contributes to the risk of atherosclerosis, and, thus, heart attack and stroke. Here, too, ApoE-4 acts as a risk factor rather than leading invariably to atherosclerosis, but there is a critical difference: Some of the environmental risk factors for atherosclerosis are well-characterized and amenable to intervention. Patients can act to reduce their risk by making dietary changes, instituting an exercise program, and stopping smoking.

While individuals concerned about their risk for AD might reduce their risk of dementia due to stroke by instituting these same changes (and would thereby decrease their risk of heart attack as well), given current knowledge of the environmental influences on AD, there is little that can be done to alter the risk of AD itself.

Other risk and protective factors for AD
Of course, the biggest determinant of whether or not a person will develop AD is longevity. The incidence of the disease rises dramatically with age. There is some suspicion that women may be at increased risk beyond the effect of their longer lifespan, but this remains unclear. The only factor besides age and family history that appears solidly associated with the development of AD is head injury.

As for protective factors, probably the best established is education. The mechanism of this effect is unclear, but some investigators think that educational achievement identifies individuals who are likely to use their brains more actively (the "use it or lose it" hypothesis), while others think it may relate to an increase in the reserve capacity of the brain.

Another factor that appears protective in some studies is smoking, which may act by stimulating a type of brain receptor known to be deficient in AD. However, smoking primarily decreases the risk of AD by contributing to early death, and it also increases the risk of dementia due to stroke.

Other factors believed to be protective but not yet clearly established include non-steroidal anti-inflammatory agents like aspirin, ibuprofen, and naproxen; post-menopausal estrogen supplementation; and antioxidants such as vitamins C and E. Physicians frequently suggest treatment with these agents to their patients for other reasons (e.g., estrogen supplements to prevent osteoporosis, or aspirin to prevent a second heard attack), and their possible protective effect against AD, while insufficient on its own to justify their use, may properly be included when weighing the risks and benefits of these medications.

Future directions
Studies are under way to more fully characterize environmental risk and protective factors and their interaction with ApoE, but, for now, only family history of longevity and of AD carry appreciable predictive value. Other studies are attempting to better characterize the risk due to ApoE, and its impact at different ages. Still others are trying to identify additional genes involved in the development of AD, especially the more common late-onset form, and to understand how the various genes involved might lead singly or in concert to the development of the disease. It is hoped that such understanding will lead to better treatments and preventive interventions for the disease.

The pace of discoveries in AD genetics is sufficiently rapid that these benefits may appear within the foreseeable future. However, other more troubling consequences may emerge as well. The possibility of genetic testing is of particular concern. A test for ApoE-4, for example, is the subject of fairly aggressive marketing, despite the caution urged by the scientific community. Similar issues have arisen around other common diseases of late life, notably breast and colon cancer, in which genes appear to play a major, but not deterministic, role.

As we learn more about the role of genes and other factors in these and other illnesses, we must also try to learn more about the needs of people like Ms. W. Ms. W was grateful to learn so much about the genetics of AD and satisfied that an ApoE test would not be very helpful in understanding her risk. But what kind of information would be helpful in relieving her anxiety or at least in planning for her future? And what else would she like to see emerge from research into the genetics of AD? A broader understanding of the needs and hopes of people like Ms. W. will help to inform - and inspire - our efforts to understand, predict, treat, cure, and prevent AD and other genetically complex diseases in the coming decades. *

Dr. Blacker is Assistant Professor of Psychiatry at Harvard Medical School and Massachusetts General Hospital, and Assistant Professor of Epidemiology, Harvard School of Public Health.