Dementia is a major cause of disability among the elderly. Alzheimer's disease (AD) accounts for approximately one-half the cases, and cerebrovascular-associated dementia (stroke and multi-infarct) underlies an additional 15% of all cases of dementia. [1] Until recently, age and family history have been the strongest risk factors for dementia. Apolipoprotein E type 4 allele (apoE element 4) is associated with AD in the late-onset familial form and in sporadic AD cases. [2-4] However, studies of apoE element 4 and dementia have focused primarily on case identification through AD clinics. Uncertainty remains regarding the frequency with which elderly apoE element 4 carriers develop dementia, sampled from a large, randomly ascertained population of elderly persons. There are two population-based studies of apoE. [5,6] The study by van Duijn et al. [5] is a population-based case-control study of 175 early-onset (onset less than equals age 65) AD cases and 159 age-matched nondemented persons. The study by Kuusisto et al. [6] also has a limited age representation and sampled only persons within a 10-year age interval (between the ages of 69 and 78 years). Although both studies confirm an increased risk for AD associated with the apoE element 4 allele, neither samples persons above age 78 and most are much younger than 78 years. The past studies have not established the cumulative incidence of dementia associated with apoE element 4 in a randomly sampled cohort of elderly individuals.
The association of apoE element 4 with dementia may extend beyond AD, and there are reports of an association between apoE element 4 and an increased risk for multi-infarct dementia. [7,8] We examined the association of apoE element 4 with dementia in the Framingham Study, a population-based longitudinal study of chronic diseases.
Methods.
Subjects.
The Framingham Study is a longitudinal study of 5,209 participants (2,336 men and 2,873 women) examined every 2 years since 1948. [9-11] The study participants were aged 28 to 62 years at the initial evaluation, and consequently all survivors are over age 70 years. Information is known on cause of death for more than 98% of the deceased participants in the Framingham Study. Family history of dementia is not known. The present study consists of 1,030 individuals who survived to have DNA collected at exam 19 (1986-1987) and who were free of dementia at exam 14 (1976-1977).
Framingham Dementia Study.
The documentation of cognitive function in the Framingham Study cohort began at exam cycle 14 in 1976 (age range, 56 to 90), when participants underwent a neuropsychological test battery. [1,12] A total of 2,083 nondemented persons were identified at exam cycle 14 and an additional 529 nondemented persons at exam 17, giving a total Dementia Cohort of 2,612 individuals. Prevalent cases of dementia were excluded from consideration. Beginning at exam 17 (1982), members of the Dementia Cohort were routinely administered the Mini-Mental State Examination (MMSE). [13] Persons with a score below the standard cut-off of 24 were called back for an in-depth neurologic and neuropsychological examination. In order to identify demented persons with higher levels of education who may be missed with the 24-point cut-off, persons who experienced a decline of three or more points on the MMSE from the most recent previous examination, were also examined. In addition, possible dementia cases were referred for dementia assessment by Framingham Study physicians when cognitive impairment was noted during the biennial examination.
Visits are made by physicians and other Framingham staff for examination of persons known to be in nursing homes. MMSE and dementia examinations are performed, and blood is collected for these study participants.
A comprehensive study of stroke in this cohort has been under way since 1981 through the Framingham Precursors of Stroke Incidence and Prognosis Study. CTs of the head have been obtained for more than 85% of the patients with stroke. Since 1991, MRI studies have been performed when possible for all stroke patients, although this represents only a small number of recent cases.
For each case of possible dementia, a detailed case review was undertaken by a panel consisting of two neurologists and a neuropsychologist. Neurologic and neuropsychological examinations were performed, outside medical records and imaging studies (CTs and MRIs of the brain) were reviewed, and pertinent Framingham Study records were examined as part of the process of establishing a diagnosis of dementia. Only persons with moderate or severe cognitive impairment persisting for at least 1 year as well as demonstrating cognitive decline were considered incident cases. The review panel estimated the age at onset based on the initial decline on MMSE, Framingham examination data, outside physician records, and information provided by family members. [1] Patients for whom definite dementia could not be established were reassessed at 2-year intervals, and thus persons with possible or mild dementia may be confirmed as cases on subsequent examination.
The subtype of dementia was determined by the review panel with the NINCDS/ADRDA criteria. [1] Individuals with cognitive impairment attributable to metabolic, psychiatric, or congenital causes were excluded as dementia cases, as were cases due to focal nonprogressive lesions resulting in an identifiable neurobehavioral syndrome such as aphasia. In nearly all cases of dementia, a brain imaging study (usually a CT or MRI) had been performed in recent years. Overall, 56% of dementia cases underwent brain imaging studies as part of a clinical workup for dementia. Non-AD types of dementia included multi-infarct dementia, dementia complicated by stroke (together accounting for 41% of non-AD cases), nonprogressive dementia caused by brain injury, probable dementia of Parkinson's disease, dementia complicated by parkinsonism, progressive dementia caused by a combination of etiologies, progressive dementia caused by another known or unknown etiology, and dementia that does not fit any other category. Criteria for dementia subtypes are listed in the article by Bachman et al. [1] Confirmed cases of dementia through exam 22 (1992-1993) are included, thereby providing longitudinal follow-up from 1976 to 1992. The diagnoses of dementia have been made continuously since 1976 and have been made blind to the apoE genotype of the individual.
Apolipoprotein E genotyping.
ApoE genotyping was performed in the original Framingham cohort from DNA samples collected during the period 1986 to 1987 (exam 19). A total of 1,078 participants were typed, but 48 of these were either not assessed for dementia or could not be determined to be nondemented at exam fourteen. Thus, a total of 1,030 persons are included in these studies. As the apoE genotype does not change over time, typing was performed on the earliest available blood samples, representing the largest sample. Demented and nondemented persons who died prior to exam 19 are not included in the analysis. Although additional blood may have been collected at subsequent exams from study participants who did not attend exam 19, DNA has not yet been extracted from these specimens.
ApoE genotypes were determined after DNA amplification and restriction isotyping as previously described. [14] The reaction mixture contained the following reagents: 14 mu L distilled water, 5 mu L 10 times reaction buffer, 0.5 mu L Taq polymerase, 0.5 mu L dATP, 1 mu L dCTP, 1 mu L dGTP, 1 mu L dTTP, 0.5 mu L primer F4 (ACAGAATTCGCCCCGGCCTGGTACAC), 0.5 mu L primer F6 (TAAGCTTGGCACGGCTGTCCAAGGA), 0.5 mu L DMSO, and 100 ng DNA. Each reaction mixture was heated to 94 degrees C for 5 minutes for denaturation, and subjected to 30 cycles of amplification by primer annealing (60 degrees C for 1 min), extension (70 degrees C for 2 min), and denaturation (95 degrees C for 1 min). After amplification, five units of Hha I were added directly to the reaction mixture and incubated for 3 hours at 37 degrees C. Each reaction mixture was loaded onto an 8% polyacrylamide nondenaturing gel and electrophoresed for 3 hours under constant current (45 mA). After electrophoresis, the gel was stained with ethidium bromide for 10 minutes, and DNA fragments were visualized by ultraviolet light.
The apoE measurement technique allowed the classification of participants into one of six genotypes: apoE element 2/ element 2, element 2/ element 3, element 3/ element 3, element 3/ element 4, element 2/ element 4, and element 4/ element 4.
Statistical methods.
The cumulative incidence of dementia was assessed by Kaplan-Meier survival analysis, and the cumulative incidence of AD for those typed for apoE was compared with those not typed using the logrank chi-square. Survival ages for the 2,612 persons in the Dementia Cohort were determined as the age at onset for demented persons, the age at death for nondemented deceased persons who had been examined within the preceding 2 years, the age at last examination for nondemented deceased persons not examined for more than 2 years, and the age at last examination for nondemented living individuals. Risk for developing AD or other dementia during the interval of follow-up since exam 14 among different genotypes was assessed by risk ratio with 95% confidence interval computed by Cox proportional hazard. The proportional hazards model controlled for the age at exam 14 and included dichotomous variables for three genotypes: apoE2 ( element 2/ element 2, element 2/ element or element 2/ element 4), apoE34 ( element 3/ element 4 only), and apoE44 ( element 4/ element 4 only), thereby making the apoE element 3/ element 3 genotype the baseline comparison group.
The sensitivity (number of AD cases that are apoE element 3/ element 4 or element 4/ element 4 / total number of cases of AD), specificity (number of non-AD individuals that are not apoE element 3/ element 4 or element 4/ element 4 / total number of non-AD individuals), positive predictive value (number of AD cases that are apoE element 3/ element 4 or element 4/ element 4 / total number of persons that are apoE element 3/ element 4 or element 4/ element 4), and negative predicted value (number of non-AD individuals that are not apoE element 3/ element 4 or element 4/ element 4 / total number of individuals that are not apoE element 3/ element 4 or element 4/ element 4) were computed.
Results.
The Dementia Cohort contained 2,612 persons identified as nondemented. A total of 372 persons were not seen again for dementia assessment in 1982 or later; 221 of these died and 151 did not return for dementia follow-up. The survival ages by dementia type for the 2,612 Dementia Cohort members, 1,030 typed for apoE and 1,582 not typed are presented in the Table 1. Although the age at exam 14 is significantly younger for the 1,030 apoE typed (mean age equals 64.0 plus minus 5.7) than for those not typed (mean age equals 68.5 plus minus 7.7, p equals 0.0001), the mean survival age for typed individuals (mean equals 76.51, SD equals 5.65) is not different from that of those not typed (mean equals 76.85, SD equals 7.55; p equals 0.20). A Kaplan-Meier comparison showed no significant difference (log-rank X2 equals 2.77, p equals 0.1) in cumulative incidence of AD in the typed versus the nontyped groups, and age-specific cumulative incidences of AD were similar (age 80: apoE typed equals 0.046, nontyped equals 0.030; age 85: apoE typed equals 0.132, nontyped equals 0.087), with a slight increase in AD among the typed group. Thus, although the apoE typed individuals were younger at entry into the study, they did not survive to older ages than those not typed.
Table 1. Framingham Study dementia cohort: Exam 14 (1976, age 55-84) to Exam 22 (1992, age 71-100)
The frequency of the 4 allele in the sample of 1,030 is 11.84%, which is only modestly below the frequency for apoE element 4 among younger individuals. [15-18] Of the 16 apoE element 4 homozygotes, five were confirmed as having AD and two had a stroke. Two additional apoE element 4 homozygotes had an MMSE score below the cutoff of twenty-four. One of these has refused formal diagnostic evaluation, and the second is currently under evaluation.
The risk ratio for AD is 30.06 (95% CI equals 10.7 to 84.4) among the apoE element 4 homozygotes, 3.7 (95% CI equals 1.9 to 7.5) among the apoE element 3/ element 4 individuals, and 0.25 (95% CI equals 0.04 to 1.89) among those with apoE element 2 in comparison with the apoE element 3/ element 3 individuals.
Cumulative incidence of AD is presented in the Figure 1. Partly because of the small number of apoE element 4 homozygotes, these data can only be meaningfully interpreted up to age 80, with 55% of apoE element 4 homozygotes having AD at that age. Thus, apoE element 4 homozygosity alone does not appear to be sufficient to cause AD up to age 80 for some individuals. The apoE element 3/ element 4 heterozygotes were adequately represented to age 80 (n equals 65) and marginally represented to age 85 (n equals 22). The incidence of AD dementia was 8.1% at age 80 and was 27.3% by age 85 years. Of the 43 cases of AD, 21 were either heterozygous or homozygous for apoE element 4 Table 1.
Figure 1. The cumulative incidence of all Alzheimer's disease in 1,030 participants in the Framingham Study Dementia Cohort is significantly increased with the presence of either one or two apoE element 4 alleles. However, 45% of the apoE element 4 homozygotes have no evidence of dementia up to age 80 years. Sixteen apoE element 4 homozygotes, 194 apoE element 3/ element 4 heterozygotes, and 686 apoE element 3 homozygotes are presented. Only one of the 134 apoE element 2 carriers ( element 2/ element 2, element 2/ element 3, or element 2/ element 4) had AD, and this group is not presented. The apoE element 4 homozygotes are only represented to age 80 years. The apoE element 3/ element 4 heterozygotes are represented to age 85 (n equals 22), and the apoE element 3 homozygotes are represented to age 90 (n equals 27).
Only one of the 134 persons either heterozygous or homozygous for the apoE element 2 allele, including none of the 18 apoE element 2/ element 4 heterozygotes, was AD affected (table). The survival ages for the apoE element 2 carriers (table) are not younger than those of the other genotypes, and although the confidence interval encompasses 1, the apoE element 2 allele appears to be protective for AD. Among the apoE element 3 homozygotes, 3.1% developed AD by age 80 and 9.1% by age 85 years.
Although the specificity was 0.81, the sensitivity was only 0.49. Of concern is that only 21 of the 210 individuals who were either heterozygous or homozygous for apoE element 4 developed AD, and thus the positive predicted value is 0.10. The positive predicted value is 0.31 for apoE element 4 homozygosity. The negative predicted value is 0.97, indicating that AD is uncommon among those without an apoE element 4 allele.
Very few of the 1,030 had other dementias (N equals 25; 15 multi-infarct dementia or stroke, 5 with Parkinson's disease-related dementia, 2 other, and 3 combined dementia). However, 3.1% of the apoE element 3/ element 4 heterozygotes had other dementias compared with 2.1% of those without an apoE element 4 allele. A risk ratio of 2.3 (95% CI equals 0.9 to 6.1) for non-AD dementia was found for apoE element 3/ element 4 heterozygosity when compared with the apoE element 3/ element 3 genotype.
Discussion.
The finding of genetic linkage to chromosome 19 for a late-onset form of familial AD on chromosome 19q13 [19] led to studies of genes in the region. The apoE gene was a strong candidate since it was present in AD-related plaques and neurofibrillary tangles [20] and was localized to 19q13.2. [21] ApoE element 4 was reported to be associated with multi-infarct dementia [8] when an association of apoE element 4 to late-onset familial AD and to sporadic AD was reported. [2,3]
Since the initial reports, many other investigations have confirmed an association between apoE element 4 and late-onset familial AD, [4,22-24] and apoE element 4 appears to be the most significant known factor in the etiology of AD. However, most previous analyses of apoE element 4 and AD have been case-control studies. The Framingham Study permits an assessment of 1,030 elderly survivors from a randomly ascertained population. The frequency of the apoE element 4 allele among these survivors (0.1184) is only modestly different from that observed in the younger Framingham Offspring Study cohort (0.135). [25] In addition, the 16-year longitudinal evaluation of dementia, from 1976 to 1992, includes not only AD but other forms of dementia. The apoE element 4 allele is significantly associated with AD for both apoE element 4 homozygotes and element 3/ element 4 heterozygotes in these elderly survivors of the Framingham Study (risk ratio for apoE element 3/ element 4 equals 3.7 compared with apoE element 3/ element 3). Homozygosity dramatically increased risk for AD (risk ratio equals 30.1 compared with apoE element 3/ element 3). One-half of the apoE element 4 homozygotes developed AD by age 80, whereas 8.1% of apoE element 3/ element 4 heterozygotes and 3.1% of apoE element 3 homozygotes developed AD by age 80 years. By age 85, 27.3% of the apoE element 3/ element 4 heterozygotes and 9.1% of the apoE element 3 homozygotes developed AD. Conversely, the apoE element 2 allele was associated with a reduced risk for AD and for all dementia. These findings are consistent with a previously proposed protective effect for the apoE element 2 allele. [26] However, the protective effect for the apoE element 2 allele has recently been questioned. One study suggests that AD-affected persons with the apoE element 2 allele have a more rapid disease and are therefore underrepresented in samples of AD cases, [27] whereas a second study finds no association of apoE genotype with survival. [28]
Two of the 16 apoE element 4 homozygotes had strokes, suggesting that risk for stroke may be increased in this group. The apoE element 4 allele is a significant risk factor for coronary artery disease, [29] and thus this allele may increase risk for cerebrovascular disease as well. Of the apoE element 3/ element 4 heterozygotes, 3.09% (table) had other dementias, primarily stroke and multi-infarct dementia. However, this sample of 1,030 study participants contains only 25 cases of other dementia, and the ability to detect an association of apoE element 4 with other dementia may be reduced. In addition, some of these individuals may have unrecognized concurrent AD, which may account for apoE element 4 association.
The association of apoE element 4 with other forms of dementia is an area of continuing interest. The association of apoE element 4 with vascular dementia was not supported in a recent study. [30] ApoE element 4 may also be associated with Lewy body dementia. [31,32] Thus, the apoE element 4 allele may not be specific to AD.
We have previously reported the finding of increased risk for dementia other than AD among the parents of Framingham Study offspring with one or two 4 alleles. [25] Thus, the finding of two cases of stroke among 16 apoE element 4 homozygotes, and that 3.09% of the apoE element 3/ element 4 heterozygotes had other dementias compared with 2.1% of non-4 carriers, supports the view that the increased risk for dementia associated with the 4 allele is not specific to AD.
In the present study, we found an increased frequency of apoE element 4 in AD. Although the diagnoses of dementia are not autopsy confirmed, the AD criteria were essentially those of the NINCDS/ADRDA. [33] The present study emphasizes that most persons carrying one or two apoE element 4 alleles do not develop dementia. Although it has been argued that apoE element 4 has a primary involvement in the pathogenesis of AD, it remains to be determined whether apoE element 4 is a cause of AD [34] or a risk factor for the disease. [23] ApoE element 4 has reduced binding to tau, [35] a protein that promotes microtubule assembly, and to beta A4, [36] a peptide constituent of senile plaques. Thus, an apoE element 4-associated loss of function may increase risk for AD. [35,36] Additional studies of apoE element 4 carriers who do not develop dementia may reveal the role of this substrate in brain function as well as the relationship of apoE element 4 and other constituents of cholesterol [37] to the etiology of AD.
Footnotes
-
.AB.-Apolipoprotein E type 4 allele (apoE element 4) is associated with Alzheimer's disease (AD) in the late-onset familial form and in sporadic cases, but the age-associated risk in a randomly sampled elderly population is not established. We examined the association of apoE element 4 with AD and other dementias (mainly multi-infarct or dementia following stroke) in 1,030 persons aged 71 to 100 years in the population-based Framingham Study cohort. Kaplan-Meier survival analysis revealed that 55% of the apoE element 4/ element 4 homozygotes developed AD by age 80, whereas 27% of apoE element 3/ element 4 heterozygotes developed AD by age 85, and 9% of those without a 4 allele developed AD by age 85 years. In comparison with persons without a 4 allele, the risk ratio for AD was 3.7 (95% CI equals 1.9 to 7.5) for apoE element 3/ element 4 heterozygotes and 30.1 (95% CI equals 10.7 to 84.4) for apoE element 4 homozygotes. ApoE element 2 (2/2, 2/3, or 2/4 genotypes) was associated with an absence of AD. One-half (n equals 21) of the 43 AD patients were either homozygous or heterozygous for apoE element 4. We found evidence for an association of apoE element 4 with other dementia, primarily multi-infarct dementia and stroke. The risk ratio is 2.3 (95% CI equals 0.9 to 6.1) for non-AD dementias among persons with apoE element 3/ element 4. Although the apoE element 4 allele is a potent risk factor for AD and may be associated with other forms of dementia, most apoE element 4 carriers do not develop dementia, and about one-half of AD is not apoE element 4 associated. The low positive predictive value of this marker (0.10) suggests that use of apoE genotyping as a screening test for AD is not supported.| NEUROLOGY 1996;46:673-677
- Copyright 1996 by the Advanstar Communication Inc.
REFERENCES
Letters: Rapid online correspondence
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
You May Also be Interested in
Hemiplegic Migraine Associated With PRRT2 Variations A Clinical and Genetic Study
Dr. Robert Shapiro and Dr. Amynah Pradhan