Cognitive differences in dementia patients with autopsy-verified AD, Lewy body pathology, or both.

OBJECTIVE: To examine the neuropsychological profile of dementia patients from a community-based autopsy sample of dementia, comparing Alzheimer disease (AD), Lewy body pathology (LBP) alone, and LBP with coexistent AD (AD/LBP). METHODS: The authors reviewed 135 subjects from a community-based study of dementia for whom autopsy and brain tissue was available. Diagnostic groups were determined according to standard neuropathologic methods and criteria, and the presence of LBs was determined using alpha-synuclein immunostaining. Neuropathologically defined diagnostic groups of AD, AD/LBP, and LBP were examined for differences on neuropsychological test performance at the time of initial study enrollment. RESULTS: There were 48 patients with AD alone, 65 with LB and AD pathology (AD/LBP), and 22 with LBP alone (LBP alone). There were no significant differences between groups demographically or on performance of enrollment Mini-Mental State Examination (MMSE) or Dementia Rating Scale (DRS). AD patients performed worse than the LBP patients on memory measures (Fuld Object Memory Evaluation Delayed Recall, Wechsler Memory Scale Logical Memory Immediate and Delayed Recall; p < 0.05) and a naming task (Consortium to Establish a Registry for Alzheimer's Disease Naming; p < 0.05). LBP patients were more impaired than AD patients on executive function (Trail Making Test Part B; p < 0.05) and attention tasks (Wechsler Adult Intelligence Scale-Revised Digit Span; p < 0.05). Decline in MMSE and DRS scores over time were greatest in the patients with AD/LBP. CONCLUSIONS: In a community-based sample of older, medically complicated patients with dementia, there are neuropsychological differences between dementia subtypes at the time of diagnosis. In particular, patients with Alzheimer disease (AD) alone and AD/Lewy body pathology (LBP) had more severe memory impairment than patients with LBP. LBP alone was associated with more severe executive dysfunction. Patients with AD/LBP had the most rapid rate of cognitive decline.

Head circumference and incident Alzheimer's disease: modification by apolipoprotein E.

BACKGROUND: The clinical expression of AD likely occurs when the accumulation of degeneration in specific brain regions leads to the descent below a critical threshold of "brain reserve" beyond which normal cognitive function cannot be maintained. The association between head circumference (HC), a measure of brain reserve, and the incidence of probable AD was examined in a large nondemented cohort that has been followed since 1992 and its modification by APOE epsilon 4 genotype. METHODS: Fifty-nine incident cases of probable AD were identified from 1,869 initially nondemented individuals seen at the baseline examination (1992 to 1994) and followed for a mean of 3.8 years. Variables measured at baseline included age, education, gender, HC, height, weight, and score on the National Adult Reading Test-Revised. APOE was genotyped at the time of the first biennial examination (1994 to 1996) and was available for 1,111 individuals in the cohort. Cox proportional hazard regression was performed to estimate hazard ratios (HR) for probable AD for HC and other covariates. RESULTS: Incident cases were significantly older, less educated, shorter, and lighter, had lower estimated verbal IQ scores, and were more likely to have at least one APOE epsilon 4 allele than unaffected individuals. The HR associated with the lowest tertile of HC (<21.4 inches) adjusted for education, gender, and APOE epsilon 4 was 2.3 (95% CI 0.7 to 6.9, p = 0.16). The HR for one or two APOE epsilon 4 alleles was significant (HR = 4.8, 95% CI 1.8 to 12.9, p = 0.002). The combination of low HC and APOE epsilon 4 strongly predicted earlier onset of AD with HR = 14.1 (95% CI 3.0 to 65, p = 0.0007). CONCLUSIONS: Smaller HC, in the presence of the APOE epsilon 4 allele, hastens the age at onset of AD. These results support the brain reserve hypothesis and its importance in precipitating the clinical expression of AD among genetically predisposed individuals.

Healthcare utilization and costs in managed care patients with Alzheimer's disease during the last few years of life.

OBJECTIVES: To learn whether managed care patients with Alzheimer's disease (AD) are more or less costly to care for than patients with other forms of dementia or patients without dementia during the last few years of life. DESIGN: Case control study. SETTING: A health maintenance organization base population. PARTICIPANTS: Three groups of subjects (mean age 85) who were deceased members of a dementia registry obtained from a health maintenance organization base population: 263 subjects with clinically diagnosed probable AD, 133 subjects with other forms of dementia, and 100 cognitively intact controls. MEASUREMENTS: Utilization records were examined for the 3 years preceding death. RESULTS: In all subcategories and in aggregate, utilization and costs of care were either similar or lower for patients with AD than for the other groups, even after controlling for age, gender, and comorbidity. CONCLUSIONS: Persons with AD do not incur higher costs than persons with other types of dementia or age-matched persons without dementia in a mature health maintenance organization during the last few years of life, when utilization is likely to be highest.

Using census data and birth certificates to reconstruct the early-life socioeconomic environment and the relation to the development of Alzheimer's disease.

The early-life environment and its effect on growth and maturation of children and adolescents are associated with several adult chronic diseases, including Alzheimer's disease. Because it is not feasible to collect information prospectively over the average life span, methods to reconstruct the early-life environment of the aged are necessary to evaluate these associations. In a community-based case-control study conducted in the United States, we collected U.S. census records and birth certificates to reconstruct the early-life socioeconomic environment of each elderly subject. Information was found on 82% of the available Alzheimer's disease cases (239 of 292) and 87% of the available controls (245 of 282). We investigated risk of Alzheimer's disease associated with father's occupation, parental age, household size, sibship size, and birth order. Subjects whose fathers were unskilled manual workers or laborers were at higher risk for Alzheimer's disease (odds ratio = 1.80, 95% confidence interval = 1.19--2.73). The risk of Alzheimer's disease was increased with increasing number of people in the household. We also evaluated whether subjects with the apolipoprotein epsilon 4 allele (APOE epsilon 4), a strong genetic risk factor that is not a necessary cause or a sufficient cause by itself for the development of Alzheimer's disease, were at higher risk than subjects who did not carry this allele. Among subjects with the APOE epsilon 4 allele whose fathers held lower-socioeconomic level occupations, the odds of developing Alzheimer's disease were higher (odds ratio = 2.35, 95% confidence interval = 1.07--5.16) compared with subjects without the allele (odds ratio = 1.40, 95% confidence interval = 0.78--2.52). Subjects carrying the APOE epsilon 4 allele alone have a threefold increased risk of Alzheimer's disease (odds ratio = 3.17, 95% confidence interval = 1.99--5.04). Compared with subjects with neither risk factor, subjects with both the genetic and the environmental risk factors (household size of seven or more and father's occupation being manual) had a relatively high risk of Alzheimer's disease (odds ratio = 14.8, 95% confidence interval = 4.9--46). The data suggest that APOE epsilon 4 may modify the associations between father's occupation, other early-life environmental factors, and development of Alzheimer's disease in late life.

Ethnicity and cognitive performance among older African Americans, Japanese Americans, and Caucasians: the role of education.

This cross-sectional analysis evaluated the association between ethnicity and cognitive performance and determined whether education modifies this association for nondemented older people (103 African Americans, 1,388 Japanese Americans, 2,306 Caucasians) in a study of dementia incidence. African Americans scored lower (median 89 out of 100) than Japanese Americans (93) and Caucasians (94) on the Cognitive Abilities Screening Instrument (CASI). Education affected CA

Neuropsychological test performance in a cognitively intact sample of older Japanese American adults.

The purpose of this paper was to present population-based data showing the effects of age on cognitive test performance in a sample of older Japanese American adults. In addition, the relative effects of education, gender, and primary spoken language were compared to effects that have been reported in the literature for majority culture older adults. Subjects included 201 non-demented Japanese American adults age 70 and older currently enrolled in the Kame Project, a prospective study of aging and dementia in King County, WA. Cognitive tests included the Consortium to Establish a Registry for Alzheimer's Disease (CERAD) neuropsychological assessment battery, WAIS-R Digit Span and Digit Symbol subtests, Trail Making Test, Purdue Pegboard, and Finger Tapping. Older age was associated with significantly (p<0.05) lower scores on all tests; less than high school education was associated with lower scores on all tests except Digit Span, Finger Tapping, and the Purdue Pegboard. Women and English-speaking participants scored higher than men and Japanese speakers on various tests of memory, attention, and visuomotor ability. These data reinforce the importance of using appropriately corrected norms when interpreting results of cognitive screening tests with minority culture older adults.

Postmenopausal estrogen and estrogen-progestin use and 2-year rate of cognitive change in a cohort of older Japanese American women: The Kame Project.

BACKGROUND: The relation between estrogen and cognition among postmenopausal women remains controversial. Also uncertain is whether the proposed association varies between women taking unopposed estrogen and those taking estrogen combined with progestin. OBJECTIVE: To determine whether unopposed estrogen and combined estrogen-progestin use were associated with the rate of cognitive change in a cohort of older, Japanese American, postmenopausal women. METHODS: A prospective observational study in a population-based cohort of older Japanese Americans (aged > or =65 years) living in King County, Washington. Cognitive performance was measured in 837 women at baseline (1992-1994) and 2-year follow-up (1994-1997) examinations using the 100-point Cognitive Abilities Screening Instrument (CASI). Least squares means general linear models were used to estimate the 2-year rate of cognitive change according to categories of postmenopausal estrogen use. RESULTS: Approximately half of this cohort (n=455) had never used estrogen at any time since menopause, 186 were past users, 132 were current unopposed estrogen users, and 64 were current estrogen-progestin users. The majority of current estrogen users were taking conjugated estrogens, and all women receiving combined therapy were taking medroxyprogesterone acetate. After adjusting for age, education, language spoken at the interview, surgical menopause, and baseline CASI score, women who had never used postmenopausal estrogen improved slightly on the CASI scale (mean adjusted change, 0.79; SEM, 0.19). This change was significantly greater for current unopposed estrogen users (mean adjusted change, 1.68; SEM, 0.36; P=.04) and significantly worse for current estrogen-progestin users (mean adjusted change, -0.41; SEM, 0.50; P =.02) compared with never users. The improvement observed in past users (mean adjusted change, 1.12; SEM, 0.29) was intermediate between the changes for never users and current unopposed estrogen users and not significantly greater than that for never users (P=.35). CONCLUSIONS: Our findings support a modest beneficial association between current unopposed estrogen use and the rate of cognitive change. We also observed a modest detrimental association between current estrogen-progestin use and the rate of cognitive change. The clinical significance of these modest differences, however, is uncertain. Data from large, long-term randomized trials are required before applying this information to the clinical setting.

Advance directive preferences among subpopulations of Asian nursing home residents in the Pacific Northwest.

OBJECTIVE: To study advance directives (code status) among subgroups of Asian nursing home residents. DESIGN: Cross-sectional design. PARTICIPANTS AND SETTING: A total of 423 residents of Asian descent (aged >55) from two ethnic nursing homes in Seattle, Washington. METHODS: Chart review was conducted on 423 residents (199 discharged between 1995 and 1998 and 244 current residents) to ascertain code status, age, gender, ethnicity, comorbidity (using the Charlson Index), and religion. RESULTS: Seventy percent of the residents were women, median age was 83 +/- 9, 43% were Chinese, 40% Japanese, and 17% other Asian (Korean, Filipino, Southeast Asian). The majority of the patients in any subgroup (72% overall) were 'no code'. In bivariate analysis, ethnicity, increased age, and comorbidity were correlated with no code status. In multivariable logistic regression, Japanese residents were more likely to be no code (OR 4.1 (95% CI, 3.1- 5.4)) controlling for age, comorbidity, gender, and religion. Chinese were more likely to be full code (OR 3.3 (95% CI, 2.6-4.2)). CONCLUSIONS: Code status differs significantly among Asian subgroups in these ethnic nursing homes. Whereas the majority of residents are no code, Japanese residents are more likely than Chinese or others to be no code. Higher age and comorbidity are also correlated with no code status.

The cognitive abilities screening instrument (CASI): data from a cohort of 2524 cognitively intact elderly.

OBJECTIVES: To describe the effects of age and education for the Cognitive Abilities Screening Instrument (CASI), a 25-item test of cognitive function. DESIGN: Cross-sectional descriptive study of the initial enrollment in a community-based prospective cohort study. PARTICIPANTS: A total of 2524 cognitively intact older adults over age 65 who were members of a major health maintenance organization, and who consented to participate in a longitudinal study. MEASUREMENTS: Summary scores for the CASI are given in the form of mean, median and percentile distributions specific for age and educational level. RESULTS: Based upon maximum likelihood analyses, age and education were significant (p<0.0001) predictors of total CASI score. Increased age and lower education were associated with a lower CASI score, as well as an increased spread in score distribution. Gender was also significantly related (p<0.01) to total CASI, with women having a slightly higher distribution of scores. Mean total scores ranged from CASI=82.2 (SD=9.0) in subjects aged 90-95 who had less than a high school degree to CASI=94.8 (SD=3. 8) in subjects aged 65-69 with at least a high school education. CONCLUSIONS: Like most cognitive screening instruments, performance on the CASI in non-demented persons is influenced by age and education. The reference values for 5-year age categories described in this article should be useful for clinicians and research investigators when using the CASI as a measure of cognitive function.

Impaired olfaction as a marker for cognitive decline: interaction with apolipoprotein E epsilon4 status.

OBJECTIVE: To determine whether olfactory status predicts cognitive decline (CD) over a 2-year follow-up period. METHODS: The authors enrolled individuals in a community-based longitudinal study of memory and aging in the Japanese-American community in King County, WA, between 1992 and 1994. At baseline they screened 1,985 persons using the Cognitive Abilities Screening Instrument (CASI) and the 12-item Cross-Cultural Smell Identification Test (CC-SIT). Of these 1,985 people, 1,836 were found not to be demented. Two years later the authors rescreened 1,604 participants with the CASI. They defined CD as a 2-year loss of > or =5.15 points/100 on the CASI. They genotyped 69% of the 1,604 people completing both examinations for apolipoprotein E (apoE). RESULTS: After adjusting for age, CASI score at baseline, education, smoking, sex, and follow-up time, the authors determined an odds ratio (OR) for CD of 0.90 (95% CI, 0.84 to 0.97) for an increase in each correct point on the CC-SIT (range, 0 to 12). Compared with normosmics, the OR for persons with impaired olfaction (microsmics) was 1.25 (95% CI, 0.83 to 1.89) and for anosmics the OR was 1.92 (95% CI, 1.06 to 3.47). Persons who were anosmic at baseline and who had at least one APOE-epsilon4 allele had 4.9 times the risk of CD (95% CI, 1.6 to 14.9) compared with normosmics without the epsilon4 allele. The estimated relative risk among women was 9.7 (95% CI, 1.3 to 70.4), and for men the risk was 3.2 (95% CI, 0.8 to 12.6). Receiver operating characteristic (ROC) curves showed that although the area under the curve (AUC) for baseline CASI was only 0.51, the AUC for CC-SIT alone was 0.62. Adding CC-SIT to the ROC model with CASI improved the AUC curve from 0.51 to 0.62. CONCLUSIONS: Unexplained olfactory dysfunction in the presence of one or more APOE-epsilon4 alleles is associated with a high risk of cognitive decline. Cross-Cultural Smell Identification Test classifies people with cognitive decline correctly to a greater degree than a global cognitive test.

Characteristics of sleep disturbance in community-dwelling Alzheimer's disease patients.

This study examined the frequency, predictors, and impact of sleep problems in a population-based sample of 205 Alzheimer's disease (AD) patients. Sleeping more than usual and early morning awakenings were the most common sleep problems reported but were the least disturbing behaviors for caregivers. Night-time awakenings were less common but were most disturbing to caregivers. Using logistic regression analyses, the factors most strongly associated with night awakenings among patients were male gender, greater memory problems, and decreased functional status. Patient depression increased the risk for caregivers to rate patient sleep problems as more disturbing overall. Cluster analyses revealed three characteristic groups of patients who awakened caregivers: one group was inactive during the day but had few other behavior problems; one group had increased levels of fearfulness, fidgeting, and occasional sadness; and the third group had multiple behavior problems, including frequent episodes of sadness, fearfulness, inactivity, fidgeting, and hallucinations. These findings indicate that the nature of sleep problems in AD is multifaceted; future research on the occurrence and treatment of sleep disturbance in dementia patients should consider the patterns of individual differences that may influence its development.

Anxiety of Alzheimer's disease: prevalence, and comorbidity.

BACKGROUND: Anxiety may be associated with psychiatric morbidity, disability, increased health care utilization, and mortality in Alzheimer's disease (AD) patients as it is in the general adult population. However, the phenomenology of anxiety symptoms in AD and its relationship to dementia progression, comorbid depression, and the presence of other problematic behaviors have not yet been examined. METHOD: Data on anxiety symptoms and their coexistence with other factors were obtained in 523 community-dwelling AD patients through interviews with their caregivers and direct physical examination. The prevalence of anxiety symptoms and their association to patient depression, other behavioral problems, gender, and age was investigated. RESULTS: Anxiety symptoms were common, occurring in 70% of subjects. Anxiety symptoms were significantly correlated with ADL impairment and other behavioral disturbances, including wandering, sexual misconduct, hallucinations, verbal threats, and physical abuse. Comorbidity of anxiety-depression was also prevalent: 54% of the sample had both anxiety and depression symptoms. ADL impairment and problem behaviors were significantly associated with comorbidity; however, the latter association was explained entirely by the presence of anxiety. CONCLUSION: Anxiety symptoms were common and significantly related to ADL and additional neuropsychiatric problems in this sample. These results indicate the need for additional research into the phenomenology of anxiety and comorbid anxiety-depression in AD and for the development and investigation of effective assessment and treatment of anxiety in AD clinical practice.

Estimation and sample design in prevalence surveys of dementia.

Population prevalence rates of dementia using stratified sampling have previously been estimated using two methods: standard weighted estimates and a logistic model-based approach. An earlier study described this application of the model-based approach and reported a small computer simulation comparing the performance of this estimator to the standard weighted estimator. In this article we use large-scale computer simulations based on data from the recently completed Kame survey of prevalent dementia in the Japanese-American residents of King County, Washington, to describe the performance of these estimators. We found that the standard weighted estimator was unbiased. This estimator performed well for a sample design with proportional allocation, but performed poorly for a sample design that included large strata that were lightly sampled. The logistic model-based estimator performed consistently well for all sample designs considered in terms of the extent of variability in estimation, although some modest bias was observed.

Cognitive decline and Japanese culture in a cohort of older Japanese Americans in King County, WA: the Kame Project.

OBJECTIVES: The prevalence of Alzheimer's disease in studies of Japanese show generally lower rates when compared with those of Caucasians. We hypothesized that among a cohort of Japanese Americans lifestyle differences would act to modify progression of the Alzheimer pathologic process over many years, resulting in a slower cognitive decline among persons whose lifestyle is more characteristically Japanese. METHODS: One thousand, eight hundred and thirty-six nondemented persons were screened with the Cognitive Abilities Screening Instrument (CASI) at baseline, and 1,604 were rescreened 2 years later. Baseline questions included migration status, exposure to Japanese culture in early life and maintenance of such culture in adulthood, and other risk factors. Cognitive decline was defined as a 2-year loss of > or = 5.15 points/100 on CASI. RESULTS: In multivariable logistic regression, variables relating to reading, writing, and speaking Japanese, being born or having lived in Japan in early life, and having friends who are only/mostly Japanese were inversely associated with cognitive decline (odds ratios ranged between 0.28 and 0.64, with p < .05). Two factors emerged in a factor analysis of these variables. The strongest explained 49% of the variance for acculturation and loaded heavily on knowledge of the Japanese language and having spent one's early years in Japan. When this factor was dichotomized into the top 20th percentile, it predicted cognitive decline with an odds ratio of 0.12 (95% CI 0.03-0.49). DISCUSSION: These results show that a Japanese lifestyle may decrease the risk of expressing cognitive decline over a 2-year follow-up period. Lower cardiovascular disease rates among Japanese may also predispose them to lower rates of cognitive decline. The greater social support characteristic of Japanese culture as well as the role that Japanese language and culture may play in neural connectivity during brain development and/or in mental stimulation in adult life may also explain our findings.

Alzheimer's disease symptom severity in blacks and whites.

OBJECTIVE: In order to determine whether there are racial differences in Alzheimer's Disease (AD) symptom severity and vascular comorbidities, we compared African-American (black) and Caucasian (white) patients with AD from similar socioeconomic backgrounds at the time the disease was first recognized. DESIGN: Cross-sectional observational study from a population-based dementia registry. PARTICIPANTS: Patients were enrolled from an HMO base population of 23,000 persons more than age 60 in Seattle, Washington. This study examines 453 subjects with probable AD (38 blacks (mean age 76.5, SD 6.4), and 415 whites (mean age 79.7, SD 6.7)). MEASUREMENTS: Measured were patient demographics, age at onset of AD, AD symptom duration, Mini-Mental State Exam (MMSE) score, Blessed Dementia Rating Scale, presence of psychiatric symptoms, and vascular comorbidities. RESULTS: Blacks had significantly lower mean cognitive scores (MMSE = 17.2, SD 5.6) compared with whites (MMSE = 20.2, SD 5.2, unpaired t test P < .01). The significant racial difference in MMSE scores persisted after controlling for education, duration of AD symptoms, age, and ADL impairment. Blacks and whites did not differ significantly regarding gender distribution, education level, income, or percent with early age of onset of AD. No statistically significant race-related differences were found in impairments in activities of daily living or symptoms of paranoia, hallucinations, or agitation. Blacks had significantly higher rates of hypertension (56%) compared with whites (34%) (Fisher's exact test, P = .013), but the rates of stroke and ischemic heart disease were similar. CONCLUSIONS: Despite uniform detection methods and controlling for reported duration of dementia symptoms, measured cognitive impairment is significantly more severe when AD is recognized in blacks compared with whites. The significantly higher prevalence of hypertension among black AD cases was not associated with excess cerebrovascular disease comorbidity. This study highlights a need for normative measurements of cognitive function in minority AD groups in order to distinguish differential cognitive symptom severity from possible measurement bias.

Coronary heart disease moderates the relationship of chronic stress with the metabolic syndrome.

Metabolic syndrome levels (MSLs) were compared in caregivers (CGs) of spouses with Alzheimer's disease who had diagnoses of coronary heart disease (CHD; n = 27) with non CGs with CHD diagnoses (n = 18), and CGs (n = 44) to non CGs (n = 52) free of CHD. MSLs were greater for CGs than non CGs, but only in persons with CHD (CHD, B for CG status = -.41; non CHD, B = .12; p < .05) at study entry (Time 1 = T1) and CHD, B = -.32; non CHD, B = .14; p < .05) 15-18 months later (Time 2 = T2). In the CHD group, MSLs were associated with poorer health habits at T1 (r = .39, p < .01), uplifts (r = -.37, p < .01) at T2, and CG status (p < .05) at T1 and T2. Relationships of CG status and MSLs declined in the presence of poor health habits at T1 and uplifts at T2. Poorer health habits and fewer uplifts may be associated with elevated MSLs in CGs with CHD.

Accelerated decline in apolipoprotein E-epsilon4 homozygotes with Alzheimer's disease.

BACKGROUND: The apolipoprotein E-epsilon4 (APOE-epsilon4) allele is a powerful genetic risk factor for the development of Alzheimer's disease (AD). AD patients who are APOE-epsilon4 homozygotes have an earlier age at onset, increased amyloid burden, and decreased acetylcholine levels--findings that suggest differences in disease severity or rate of progression. Studies of genotype differences in rate of decline, however, have produced negative results that may be due to methodologic biases. The current study examined rate of decline in the largest sample of APOE-genotyped AD patients for whom longitudinal cognitive data have been reported. METHODS: Newly diagnosed patients with probable AD (n = 201) comprised four genotype groups: epsilon2/3 (n = 14), epsilon3/3 (n = 75), epsilon3/4 (n = 82), and epsilon4/4 (n = 30). The Dementia Rating Scale (DRS) was administered at baseline and then annually for 1 to 6 years (mean, 2.5 years). For each subject, a DRS slope was calculated reflecting annual rate of decline. Rate of decline as measured by DRS slope differed according to genotype, with the effect modified by DRS score (p < 0.014). At the mean DRS score observed in our sample (DRS = 105), the epsilon4/4 group had an increased rate of decline (11.9 points per year) relative to the epsilon2/3 (5.8 points per year; p < 0.003), epsilon3/3 (9.3 points per year; p < 0.076), and epsilon3/4 (9.6 points per year; p < 0.055) groups. At a lower DRS score (DRS = 80), even larger differences were observed among genotypes; the epsilon4/4 group had a increased rate of decline (22.2 points per year) relative to the epsilon2/3 (9.7 points per year; p < 0.0006), epsilon3/4 (15.8 points per year; p < 0.020), and epsilon3/3 (18.2 points per year; p < 0.173) groups. The epsilon2/3 group had a significantly slower rate of decline than all other groups at DRS scores of 80 or 105. CONCLUSIONS: APOE-epsilon4 homozygosity is associated with a faster rate of cognitive decline, whereas the epsilon2 allele slows disease progression. These findings suggest that APOE plays a mechanistic role in the progression of AD, and is not simply related to disease onset.