Persistent infection with neurotropic herpes viruses and cognitive impairment.

BACKGROUND: Herpes virus infections can cause cognitive impairment during and after acute encephalitis. Although chronic, latent/persistent infection is considered to be relatively benign, some studies have documented cognitive impairment in exposed persons that is untraceable to encephalitis. These studies were conducted among schizophrenia (SZ) patients or older community dwellers, among whom it is difficult to control for the effects of co-morbid illness and medications. To determine whether the associations can be generalized to other groups, we examined a large sample of younger control individuals, SZ patients and their non-psychotic relatives (n=1852). Method Using multivariate models, cognitive performance was evaluated in relation to exposures to herpes simplex virus type 1 (HSV-1), herpes simplex virus type 2 (HSV-2) and cytomegalovirus (CMV), controlling for familial and diagnostic status and sociodemographic variables, including occupation and educational status. Composite cognitive measures were derived from nine cognitive domains using principal components of heritability (PCH). Exposure was indexed by antibodies to viral antigens. RESULTS: PCH1, the most heritable component of cognitive performance, declines with exposure to CMV or HSV-1 regardless of case/relative/control group status (p = 1.09 × 10-5 and 0.01 respectively), with stronger association with exposure to multiple herpes viruses (ß = -0.25, p = 7.28 × 10-10). There were no significant interactions between exposure and group status. CONCLUSIONS: Latent/persistent herpes virus infections can be associated with cognitive impairments regardless of other health status.

Alpha-2 macroglobulin is genetically associated with Alzheimer disease.

Alpha-2-macroglobulin (alpha-2M; encoded by the gene A2M) is a serum pan-protease inhibitor that has been implicated in Alzheimer disease (AD) based on its ability to mediate the clearance and degradation of A beta, the major component of beta-amyloid deposits. Analysis of a deletion in the A2M gene at the 5' splice site of 'exon II' of the bait region (exon 18) revealed that inheritance of the deletion (A2M-2) confers increased risk for AD (Mantel-Haenzel odds ratio=3.56, P=0.001). The sibship disequilibrium test (SDT) also revealed a significant association between A2M and AD (P=0.00009). These values were comparable to those obtained for the APOE-epsilon4 allele in the same sample, but in contrast to APOE-epsilon4, A2M-2 did not affect age of onset. The observed association of A2M with AD did not appear to account for the previously published linkage of AD to chromosome 12, which we were unable to confirm in this sample. A2M, LRP1 (encoding the alpha-2M receptor) and the genes for two other LRP ligands, APOE and APP (encoding the amyloid beta-protein precursor), have now all been genetically linked to AD, suggesting that these proteins may participate in a common neuropathogenic pathway leading to AD.

Linkage analysis of schizophrenia in African-American families.

While many studies have sought a window into the genetics of schizophrenia, few have focused on African-American families. An exception is the Project among African-Americans to Explore Risks for Schizophrenia (PAARTNERS), which seeks to identify novel and known risk variation for schizophrenia by genetic analyses of African-American families. We report a linkage study of diagnostic status in 217 African-American families using the Illumina Linkage Panel. Due to assumed incomplete and time-dependent penetrance, we performed linkage analysis using two different treatments of diagnosis: (1) treating both affected and unaffected individuals as informative for linkage (using the program SIBPAL) and (2) treating only affected individuals as informative (using the program MERLIN). We also explore three definitions of affected status: narrowly defined schizophrenia; one broadened to include schizoaffective disorder; and another including all diagnoses indicating psychosis. Several regions show a decrease in the evidence for linkage as the definition broadens 8q22.1 (rs911, 99.26 cM; SIBPAL p-value [p] goes from 0.006 to 0.02), 16q24.3 (rs1006547, 130.48 cM; p from 0.00095 to 0.0085), and 20q13.2 (rs1022689, 81.73 cM; p from 0.00015 to 0.032). One region shows a substantial increase in evidence for linkage, 11p15.2 (rs722317, 24.27 cM; p from 0.0022 to 0.0000003); MERLIN results support the significance of the SIBPAL results (p=0.00001). Our linkage results overlap two broad, previously-reported linkage regions: 8p23.3-p12 found in studies sampling largely families of European ancestry; and 11p11.2-q22.3 reported by a study of African-American families. These results should prove quite useful for uncovering loci affecting risk for schizophrenia.

Allele increasing susceptibility to human breast cancer may be linked to the glutamate-pyruvate transaminase locus.

The patterns of the occurrence of breast cancer in 11 high-risk families were evaluated by segregation and linkage analysis. These patterns were consistent with the hypothesis that increased susceptibility to breast cancer was inherited as an autosomal dominant allele with high penetrance in women. The postulated susceptibility allele in these families may be chromosomally linked to the glutamate-pyruvate transaminase (E.C. 2.6.1.2, alanine aminotransferase) locus. Confirmation of this linkage in other families would establish the existence of a gene increasing susceptibility to breast cancer. Since there is no association in the general population between a woman's glutamate-pyruvate transaminase genotype and her cancer risk, the glutamate-pyruvate transaminase linkage cannot be used as a screening test for breast cancer.

Evidence for genetic linkage of Alzheimer's disease to chromosome 10q.

Recent studies suggest that insulin-degrading enzyme (IDE) in neurons and microglia degrades Abeta, the principal component of beta-amyloid and one of the neuropathological hallmarks of Alzheimer's disease (AD). We performed parametric and nonparametric linkage analyses of seven genetic markers on chromosome 10q, six of which map near the IDE gene, in 435 multiplex AD families. These analyses revealed significant evidence of linkage for adjacent markers (D10S1671, D10S583, D10S1710, and D10S566), which was most pronounced in late-onset families. Furthermore, we found evidence for allele-specific association between the putative disease locus and marker D10S583, which has recently been located within 195 kilobases of the IDE gene.

A polymorphism in SOD2 is associated with development of Alzheimer's disease.

Genes involved in cellular mechanisms to repair oxidative damage are strong candidates as etiologic factors for Alzheimer's disease (AD). One important enzyme involved in this mechanism is superoxide dismutase 2 (SOD2). The gene for this enzyme lies within a single haplotype block at 6q25.3, a region showing evidence for linkage to AD in a genome scan. We genotyped four single nucleotide polymorphisms (SNPs) in SOD2 in families of the National Institute of Mental Health-AD Genetics Initiative (ADGI): rs2758346 in the 5' untranslated region (UTR), rs4880 in exon 2, rs2855116 in intron 3 and rs5746136 in the 3'UTR. Under a dominant model, family-based association tests showed significant evidence for association of AD with the first three loci in a candidate gene set of families with individuals having age of onset of at least 50 years and two affected and one unaffected sibling, and in a late-onset subset of families (families with all affected individuals having age of onset of at least 65 years) from the full ADGI sample. The alleles transmitted more frequently to cases than expected under the null hypothesis were T, C, G, and G. Global tests of the transmission of haplotypes indicate that the first two loci have the most consistent association with risk of AD. Because of the high linkage disequilibrium in this small (14 kb) gene, and the presence of 100 SNPs in this gene, 26 of which may have functional significance, additional genotyping and sequencing are needed to identify the functionally relevant SNP. We discuss the importance of our findings and the relevance of SOD2 to AD risk.

Rapid determination of clonality by detection of two closely-linked X chromosome exonic polymorphisms using allele-specific PCR.

We reported two specific, reproducible, and quantitative clonality assays based on detection of exonic polymorphisms of the X chromosome genes p55 and G6PD using rtPCR-LDR. These assays are inconvenient for screening purposes. This study sought to develop a simple, reproducible assay, practical for screening genomic DNA samples for p55/G6PD genotypes, rapid clonality determination, and to determine the linkage relationship between these closely related loci. The salient feature of ASPCR is the performance of two PCR rounds. The first generates template; the second, using one aliquot of first-round products in two reaction tubes, each containing one allele-specific primer, detects each allele. ASPCR and rtPCR-LDR produced identical p55/G6PD results in 91 normal female genomic DNAs, and in 12 clonal hematopoietic disorder cDNAs, confirming assay validity. 209 female and 207 male genomic DNA samples were analyzed for p55/G6PD genotype by ASPCR; 60% of females were heterozygous at one or both loci. G6PD and p55 allelic frequencies were significantly different among African-American men and women, but were not significantly different among Caucasian men and women. These loci were in linkage equilibrium among African Americans, but not among Caucasians. ASPCR is a rapid, sensitive, and specific method for screening large numbers of genomic DNAs, and for rapid clonality determination.

Knowledge of and attitudes about Alzheimer disease genetics: report of a pilot survey and two focus groups.

OBJECTIVES: In preparation for the development of an educational intervention on Alzheimer disease (AD) genetics, we undertook a pilot survey of knowledge in this area and attitudes toward genetic testing for AD among individuals with a family history of AD. METHODS: For the pilot study, we administered a 30-min questionnaire to 57 unaffected individuals from a genetic linkage study. For the focus groups, we interviewed two groups of subjects, ages 44-70 years, with a family history of AD, one of 10 Caucasians and the other of 6 African-Americans. RESULTS: The pilot study showed that there was limited knowledge of genetics overall and AD genetics in particular, considerable concern about personal risk, and little knowledge of or interest in genetic testing for the disease. The focus groups reinforced and fleshed out these impressions and highlighted the importance of caregiving experience in the attitudes toward personal risk for AD. CONCLUSIONS: These results underscore the value of genetics education for this and other complex diseases and suggest specific foci for educational interventions.

Genetic epidemiology of breast cancer and associated cancers in high-risk families. I. Segregation analysis.

Genetic and environmental hypotheses that might explain the patterns of occurrence of breast cancer and associated cancers in 18 large families at high risk of the disease were tested with the use of segregation analysis. For 16 pedigrees, results were consistent with the hypothesis that breast cancer has a genetic etiology. In 2 other families, breast cancer appeared more likely to have an environmental origin. Breast cancer susceptibility is best explained by hypotheses that postulate autosomal dominant susceptibility alleles in 10 families with primarily premenopausal breast cancer and ovarian cancer, in 4 families with primarily postmenopausal breast cancer, and in 2 families with breast cancer, brain tumor, sarcoma, leukemia, and adrenocortical carcinoma in children and young adults. In an accompanying paper, genetic susceptibility in the first 2 groups of families is further explored with the use of linkage analysis.

Genetic epidemiology of breast cancer and associated cancers in high-risk families. II. Linkage analysis.

Chromosomal locations of hypothetical alleles which increase susceptibility to human breast cancer in some families were investigated by genetic linkage analysis. In 7 families with primarily premenopausal breast cancer and (in 5 families) ovarian cancer, a dominant susceptibility allele may be linked to the genetic marker glutamicpyruvic transaminase or alanine aminotransferase (GPT; lod score 1.95 at zero recombination). The most positive lod score for linkage to a recessive susceptibility allele was for acid phosphatase (ACP; lod score 0.78 at 40% recombination), but ACP was informative in ony 1 family. In 3 families with primarily postmenopausal breast cancer, none of 21 genetic markers provided any evidence for linkage to either dominant or recessive susceptibility alleles. In the families with the possible GPT linkage, women who carry the hypothetical susceptibility allele would be at high risk of breast cancer, whereas their relatives who do not carry that allele would have no increased risk. GPT genotype is not associated with breast cancer risk in the general population, so GPT linkage cannot be used as a screening test for breast cancer.

Evidence for genetic admixture as a determinant in the occurrence of insulin-dependent diabetes mellitus in U.S. blacks.

In recent years, it has been proposed that genetic admixture may have played a role in the increased frequency of insulin-dependent diabetes mellitus (IDDM) in young U.S. blacks relative to African blacks. In support of this proposal, the similar associations of specific markers of the major histocompatibility complex (MHC) with IDDM in U.S. blacks with respect to U.S. whites have been cited. To determine whether racial admixture was a factor in the increased prevalence, we did three analyses of admixture. In the first we used nine genetic markers (ABO, Rh, Fy, Hp, Gc, Pl, OR, Tfr, and Gm) and determined that there was significantly greater than zero genetic contribution from whites in our sample of U.S. black IDDM patients (9.6 +/- 2.3%, P less than 0.01) when a sample of U.S. blacks without IDDM was used as one "parental" population. In the next two analyses, we estimated the amounts of genetic contribution from whites in the U.S. blacks with and without IDDM using reported gene frequencies for West African blacks for four genetic markers (ABO, Rh, Fy, and Hp). The estimate of admixture (21.4 +/- 2.8%) for the black IDDM sample was greater than that for the U.S. black controls (17.9 +/- 2.3%), although the difference was not significant. Our estimate of genetic contribution from whites, 21.4% for black IDDM patients, supports the assumptions of 20% admixture which MacDonald and Rotter and Hodge used to test their respective models for the inheritance of IDDM. These results support the hypothesis that admixture with the white population is, in part, responsible for the increase in prevalence of IDDM seen in U.S. blacks.

HLA-A, -B, and -DR associations in type I diabetes mellitus with onset after age forty.

It remains to be convincingly demonstrated whether insulin-requiring, ketosis-prone, lean-at-onset, type I diabetics who develop their disease after age 40 have the same disease as the children with similar characteristics. To address this question, we examined the population HLA genetic associations of this group. One hundred forty white, insulin-using diabetics with onset of disease past age 40 yr and 268 normal white controls have thus far been analyzed for HLA type. In the group of patients who were lean-at-onset and/or ketosis-prone (N = 54), there was a significantly increased frequency of DR4 (RR = 4.63; P less than 0.01) and significantly decreased frequency of DR2 (RR = 0.18; P less than 0.05) after correction. DR4 was also significantly increased after correction (RR = 5.72; P less than 0.25) in the subgroup who were both lean and ketosis-prone (N = 23). No significant differences in HLA-DR frequencies were found between the obese and not-ketosis-prone group (N = 69) and controls. No significant associations of HLA-A or-B antigens with either group were observed after correction for the number of antigens tested. To our knowledge, this is the first such study in the United States, and the first demonstrating that late onset diabetics who are lean-at-onset and/or ketosis-prone exhibit HLA-DR antigen associations which are similar to early onset cases.

Prevalence and risk factors of microalbuminuria in a cohort of African-American women with gestational diabetes.

OBJECTIVE: To study the prevalence of microalbuminuria (MA) in African-American women with a history of gestational diabetes (GDM) who are at high risk for insulin resistance and renal dysfunction and to study MA's relation to insulin resistance, type 2 diabetes, and hypertension. RESEARCH DESIGN AND METHODS: MA was assessed using 24-h, timed, and/or random urine samples in a cross-sectional sample (n = 289) from a cohort of African-American women with a history of GDM and followed for a median of 11 years (range 3.0-18.4) since their diabetic pregnancy. Subjects with a urine albumin excretion rate of 30-300 g/24 h or 30-300 microg/mg creatinine in a random sample were classified as having MA if two of three samples over a 3- to 6-month period were positive. These women were evaluated for family history of diabetes, smoking and alcohol use, BMI, diabetes, hypertension, and lipid abnormalities. Insulin sensitivity was determined using the homeostasis model assessment (HOMA) estimates, which used fasting insulin and glucose measurements obtained at the same time as the MA urine sample. RESULTS: At MA assessment, the women ranged in age from 22 to 57 years, with a median of 39 years. The overall prevalence of MA was 20%; 36% in those with diabetes. Those women with MA had higher rates of diabetes (63.8 vs. 28.6%, odds ratio [OR] = 4.4, P < 0.05), hypertension (82.8 vs. 42.9%, OR = 6.4, P < 0.05), and family history of diabetes (85.7 vs. 61.7%, OR = 3.7, P < 0.05). The proportion of subjects with MA with a family history of hypertension was nonsignificantly increased (92.9 vs. 82.4%). Subjects with MA were more obese (BMI 37.2 +/- 8.9 vs. 34.4 +/- 8.6 kg/m(2)) and had higher levels of HbA(1c) (8.8 +/- 3.3 vs. 6.6 +/- 1.8%, P < 0.001) and systolic (144.3 +/- 25.9 vs. 122.8 +/- 17.2 mmHg, P < 0.0001) and diastolic (95.1 +/- 15.4 vs. 82.5 +/- 11.9 mmHg, P < 0.0001) blood pressures. Lipid fractions were similar in those with and without MA. Although fasting glucose was much higher in subjects with MA (10.3 +/- 5.8 vs. 7.1 +/- 4.2 mmol/l, P = 0.0002), insulin levels were not significantly higher in subjects with MA (17.4 +/- 21.2 vs. 15.2 +/- 12.4 pmol/l). Insulin sensitivity, as measured using log HOMA, was similar (1.5 +/- 0.6 vs. 1.6 +/- 0.6) in women with and without MA, respectively. Multivariable logistic regression analyses indicated that HbA(1c), OR = 1.16 (1.07, 1.27), and systolic blood pressure, OR = 1.27 (1.14, 1.41), were independent risk factors for MA. In those with diabetes, the subjects with MA had higher rates of hypertension-83.8 vs. 56.1%, OR = 4.1 (1.5, 11.10)-which was reflected by their higher systolic and diastolic blood pressures, 146.1 mmHg (P = 0.001) and 94.8 mmHg (P = 0.002), respectively, and lower levels of VLDL (0.45 +/- 0.22 vs. 0.61 +/- 0.33 mmol/l, P = 0.021). In the multivariable analyses of those with diabetes, the two independent risk factors for MA were similar: HbA(1c), OR = 1.13 (1.01, 1.28), and systolic blood pressure, OR = 1.21 (1.04, 1.41). CONCLUSIONS: African-American women with a history of GDM have one of the highest rates for MA. Presence of MA was not associated with insulin resistance but was significantly independently associated with HbA(1c) levels and hypertension. These results, taken in context of the literature, suggest that hypertension and glucose intolerance, in part, influence MA through different mechanisms. Because of the high prevalence of MA in this population and MA's relation to all-cause and cardiovascular mortality, screening for MA should be considered.

Risk factors for gestational diabetes in black population.

In a long-term longitudinal study of gestational diabetes mellitus in Black women, risk factors that were identified were age, obesity, a family history of diabetes, and the presence of hypertension. Poor predictors were a history of a previous large-for-date infant, parity, and age at first pregnancy. The prevalence of smooth muscle and nuclear autoantibodies was higher in gestational diabetic subjects. Gestational diabetic subjects who required insulin for glycemic control were more obese, had a lower frequency of the Bf-F phenotype and a higher frequency of the Bf-F1 phenotype, and had a lower frequency of the type 2 allele at the polymorphic locus adjacent to the insulin gene. Restriction-fragment-length polymorphisms flanking the insulin and apolipoprotein A-I and C-III genes, although not associated with gestational diabetes mellitus, may be associated with hyperlipidemia and subsequent atherosclerosis.

Genes within the major histocompatibility complex predict NIDDM in African-American women in Alabama.

OBJECTIVE: To test the hypothesis that genes within the major histocompatibility complex (MHC) are associated with gestational diabetes mellitus (GDM) and, subsequently, non-insulin-dependent diabetes mellitus (NIDDM) in African-American women. RESEARCH DESIGN AND METHODS: African-American women who presented with GDM were compared with pregnant African-American control subjects. Following pregnancy, GDM patients were assessed at various intervals of time (median = 6 years) to determine whether they had developed diabetes. RESULTS: GDM patients who required insulin during pregnancy possessed a significantly higher frequency of A33, DR2, DR9, and BF-S phenotypes than control subjects. GDM patients who subsequently developed NIDDM had a significantly higher frequency of B41, DR2, and BF-S and a lower frequency of DR1 and DR6 phenotypes than control subjects. Even after controlling for age and body mass index, B41 and DR2 were independent predictors of developing insulin-requiring GDM and NIDDM in GDM subjects. CONCLUSIONS: These results suggest that either one or more genes within the MHC are involved in the etiology of NIDDM.

The role of TNF and its receptors in Alzheimer's disease.

Tumor necrosis factor (TNF) is an important proinflammatory cytokine that is upregulated in Alzheimer disease (AD) patients and involved with AD genes. Several TNF promoter polymorphisms that increase expression are associated with inflammatory and infectious diseases. We previously reported results that detected a AD associated region near the TNF gene. Using family-based association tests we also reported an association between AD and a TNF haplotype in sibling-pair families, and a significant increase in the mean age of onset for a group of African-American AD patients carrying this same haplotype. Previous reports have shown that that the chromosome 1p and chromosome 12p regions are linked to late-onset AD. These two regions harbor TNF receptors (TNFR) 2 and 1, respectively, and binding to them mediates biological effects of TNF. We found a significant asssociation of a TNFR2 exon 6 polymorphism with late-onset AD in families with no individuals possessing the APOE E4E4 genotype under a dominant model. We found no significant association of three polymorphisms in the TNFR1 gene to AD. These results provide further evidence for the involvement of TNF in the pathogenesis of AD.

A positive association between maternal serum zinc concentration and birth weight.

A study was conducted on a cohort of 476 women (364 black, 112 white), who attended the Jefferson County Health Department clinic for their prenatal care, to ascertain the relationship between maternal serum zinc concentration measured early in pregnancy and birth weight. For all subjects maternal serum zinc was significantly related to birth weight after various independent determinants of birth weight were controlled for. The data in this study indicate a threshold for maternal serum zinc concentration below which the prevalence of low birth weight increases significantly. Pregnant women who had serum zinc concentrations in the lowest quartile had significantly higher prevalence of low birth weight than did those mothers who had serum zinc concentrations in the upper three quartiles during pregnancy. These findings suggest that maternal serum zinc concentration measured early in pregnancy could be used to identify those women at higher risk of giving birth to a low-birth-weight infant.

Plasma homocyst(e)ine, folate, and vitamin B-12 concentrations and risk for early-onset coronary artery disease.

High plasma homocyst(e)ine (Hcy) concentrations may be a determinant of coronary artery disease (CAD). Folate and vitamin B-12 are required for the primary metabolic pathway to reduce Hcy concentrations. The interrelationships of Hcy and these two vitamin cofactors were investigated in a case-control study of 101 white males aged 30-50 y with angiographically demonstrated CAD, and 108 white male, similarly aged, control subjects living in the same community as the patients. The odds ratio (OR) of CAD per quartile increase of plasma Hcy concentration based on control values was 1.6 (95% CI: 1.3, 2.1). After age, HDL and LDL cholesterol, body mass index, smoking, hypertension, and diabetes were controlled for, Hcy remained an independent risk factor (OR: 1.4; 95% CI: 1.0, 2.0). The OR change per quartile increase of folate concentration was 0.8 (95% CI: 0.6, 1.0). This difference was reduced (OR: 0.9; 95% CI: 0.7, 1.2) after Hcy adjustment. No difference in the geometric mean of vitamin B-12 concentration was found between patients and control subjects, both 5.8 nmol/L. However, after Hcy and the other CAD risk factors were controlled for, the OR per quartile increase in vitamin B-12 concentration was 1.5 (95% CI: 1.0, 1.8). Reduction in plasma Hcy by interventions to increase plasma folate concentration may decrease CAD risk.

Reliability and validity of NINCDS-ADRDA criteria for Alzheimer's disease. The National Institute of Mental Health Genetics Initiative.

OBJECTIVE: To assess interrater reliability and validity of NINCDS-ADRDA (National Institute of Neurological and Communicative Diseases and Stroke/Alzheimer's Disease and Related Disorders Association) criteria for Alzheimer's disease (AD). DESIGN: A multisite reliability and validity study in which clinicians from each site diagnosed 60 case summaries yielding a preconsensus estimate of reliability and validity. A consensus conference was conducted for each disagreement, leading to a postconsensus estimate of validity. The criterion standard was a diagnosis of AD by autopsy. SETTING: Three academic medical centers. SUBJECTS: A convenience sample of 60 detailed case summaries, 40 with AD and 20 with other dementing disorders. MAIN OUTCOME MEASURES: The kappa coefficient, sensitivity, and specificity. RESULTS: The kappa coefficient for preconsensus agreement on a diagnosis of probable or possible AD vs non-AD was 0.51; the sensitivity of a diagnosis of probable or possible AD for a pathological diagnosis of AD was 0.81, and the specificity was 0.73. The postconsensus sensitivity was 0.83, and the specificity was 0.84. CONCLUSIONS: The results support the reliability and validity of NINCDS-ADRDA criteria and show that the consensus process may improve diagnostic accuracy. The cases are reviewed with a focus on the sources of diagnostic disagreements and errors and possible changes that might improve the accuracy of the criteria.

ApoE-4 and age at onset of Alzheimer's disease: the NIMH genetics initiative.

OBJECTIVE: To explore the impact of apoE-4 on Alzheimer's disease (AD) and its age at onset. DESIGN: A genetic linkage study using affected relative pairs, predominantly siblings. SETTING: Three academic medical centers ascertained subjects from memory disorder clinics, nursing homes, and the local community. SUBJECTS: 310 families including 679 subjects with AD by NINCDS/ADRDA and/or Khachaturian criteria and 231 unaffected subjects. OUTCOME MEASURE: ApoE genotype. ANALYTIC METHODS: Association, affected pedigree member, sibling pair, and lod score analyses. RESULTS: ApoE-4 was strongly associated with AD in this sample (allele frequency = 0.46 vs. 0.14 in controls, p < 0.000001). Results of lod score, affected pedigree member analysis, and sib-pair analysis also supported apoE-4 as a risk factor for AD. When the sample was stratified on family mean age at onset, the risk conferred by apoE-4 was most marked in the 61 to 65 age group. Individuals with two copies of apoE-4 had a significantly lower age at onset than those with one or no copies (66.4 vs. 72.0, p < 0.001), but individuals with one copy did not differ from those with none. Within families, the individual with the earliest age at onset had, on average, significantly more apoE-4 alleles (p < 0.0001) than the individual with the latest onset. DISCUSSION: This work supports previous reports of an association between apoE-4 and the development of AD and demonstrates that apoE-4 exerts its maximal effect before age 70. These findings have important implications for the potential use of apoE genotyping for diagnosis and prediction of disease. They also underscore the need to identify additional genetic factors involved in AD with onset beyond age 70 years.